Blaustein's Pathology of the Female Genital Tract [7th ed.] 978-3-319-46333-9;978-3-319-46334-6

Citation preview

Robert J. Kurman Lora Hedrick Ellenson Brigitte M. Ronnett Editors

Blaustein’s Pathology of the Female Genital Tract

Seventh Edition

Blaustein’s Pathology of the Female Genital Tract

Robert J. Kurman • Lora Hedrick Ellenson Brigitte M. Ronnett Editors

Blaustein’s Pathology of the Female Genital Tract Seventh Edition

With 1479 Figures and 121 Tables

Editors Robert J. Kurman Department of Gynecology Obstetrics, Pathology and Oncology Division of Gynecologic Pathology Johns Hopkins University School of Medicine Baltimore, MD, USA

Lora Hedrick Ellenson Department of Pathology and Laboratory Medicine Division of Gynecologic Pathology Weill Cornell Medical College and New York Presbyterian Hospital New York, NY, USA

Brigitte M. Ronnett Department of Pathology Division of Gynecologic Pathology Johns Hopkins University School of Medicine Baltimore, MD, USA

ISBN 978-3-319-46333-9 ISBN 978-3-319-46334-6 (eBook) ISBN 978-3-319-46335-3 (print and electronic bundle) https://doi.org/10.1007/978-3-319-46334-6 Library of Congress Control Number: 2018963748 1st to 6th edition: © Springer Science+Business Media, LLC 1977, 1982, 1987, 1994, 2002, 2011 © Springer Nature Switzerland AG 2019 This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. The publisher, the authors, and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty, express or implied, with respect to the material contained herein or for any errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. This Springer imprint is published by the registered company Springer Nature Switzerland AG The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland

Dedication by Dr. Kurman: To Carole C. Kurman for her constant support and encouragement throughout my career. Dedication by Dr. Ellenson: With immeasurable gratitude to Richard, Thomas, and Taite for providing endless joy, support, and inspiration. Dedication by Dr. Ronnett: To Lisa, for her unwavering support and unconditional love, and in appreciation of the Johns Hopkins Gynecologic Pathology fellows—past, present, and future—for their invaluable role in our clinical service, which enables our academic pursuits. Dedication by the editorial team: We are indebted to our coauthors for their contributions to the current and prior editions, as well as to our fellows and residents, all of whom have made this project possible.

Preface

In science, technological innovation leads to progress. The application of the microscope, most notably the contributions of Virchow, introduced the theory of cellular pathology which led him to propose that cancer developed from otherwise normal cells. Similarly, the application of microscopic analysis to gynecologic tumors by Cullen revolutionized our understanding of gynecologic disease. Today the explosion in the fields of molecular biology and bioinformatics has provided novel information that once again is transforming the field of pathology. The arc of progress in gynecologic pathology can be appreciated by comparing the contents of the first edition of Blaustein’s Pathology of the Female Genital Tract which appeared in 1977 to this current edition. In the first edition, the etiology of cervical cancer was linked to herpes simplex virus-2 (HSV-2) based on seroepidemiologic studies and electron micrographs showing HSV-2 capsids in cervical carcinoma cells. The role of unopposed estrogenic stimulation in the development of endometrial carcinoma was described but still debated, and carcinoma of the endometrium was classified as endometrial adenocarcinoma, adenoacanthoma, adenosquamous carcinoma, and squamous carcinoma. In the fallopian tube chapter, one half a page was devoted to carcinoma in situ (intraepithelial carcinoma). Gestational trophoblastic disease was classified into three categories, hydatidiform mole, invasive mole, and choriocarcinoma. We now recognize that cervical cancer is due to human papillomavirus (HPV), and with universal adoption of vaccination for HPV, cervical cancer can conceivably be eradicated in the future. The histopathologic classification of endometrial cancer has been expanded, and the most recent next-generation sequencing studies suggest that it can be classified into four major molecular subtypes (Kandoth et al. 2013). Our entire concept of epithelial ovarian carcinoma has undergone a paradigm shift with attention now directed at the fallopian tube as the site of origin. A recent study that incorporates assays for mutations on liquid cervical samples and an assay for aneuploidy on circulating tumor DNA from plasma was shown to have high sensitivity and specificity for the detection of ovarian and endometrial cancer, demonstrating the potential for diagnosing ovarian and endometrial cancer based entirely on molecular genetic studies, at an early stage, when these carcinomas are more likely to be cured. Clearly, a lot has changed in the ensuing years. We think how simplistic and naive our understanding of gynecologic disease was 40 years ago, but future generations, 40 years from now, will look back at this edition and have the same reaction. Assessing where we are today and speculating on what the future may hold in store, we vii

viii

Preface

recognize that pathology stands at a crossroads as morphologic diagnosis gives way to molecular diagnosis. Given the pace of personalized medicine, it is possible that in the not too distant future classification systems may become obsolete as each individual’s tumor will be classified based on its unique molecular alterations. To guide this transition, it is critical that traditional morphologic analysis be carefully correlated with immunohistochemical and molecular biologic findings. This is one of the major contributions that the current edition of the Blaustein text aims to fulfill. In the preface of the first edition, Ancel Blaustein stated that “the text is written for obstetricians, gynecologists, pathologists and for residents training in these disciplines.” We recognize that this goal has been more than fulfilled as countless numbers of pathologists, gynecologists, and a host of physicians in other specialties have turned to this textbook to provide the most up-to-date information on the nature of gynecologic disease including discussion of diagnosis and treatment. This edition carries on this long tradition. The present edition has been rewritten to include advances in the field which, as noted earlier, are now largely based on the contributions of molecular biology. Since the last edition, the WHO Classification has been revised, and the classification systems used herein are those proposed by the WHO. Dr. Blaustein also had the prescience to realize that even in 1977 a comprehensive text necessitated multiple authors stating that “the expansion of information in the field of gynecologic pathology renders single authorship obsolete.” Accordingly, that first edition was the first textbook in pathology to have multiple contributors. With the passage of time, contributors have passed on or retired, and we have therefore recruited young, up-and-coming experts to replace previous contributors and to reinvigorate the textbook with new ideas and concepts. In that vein Dr. Kurman, having edited this textbook beginning with the third edition, will be stepping down and turning over the editorship to Drs. Ellenson and Ronnett who will succeed him and carry on the long tradition of this text as an invaluable educational resource in the field of obstetrics and gynecology. Robert J. Kurman, M.D. Lora Hedrick Ellenson, M.D. Brigitte M. Ronnett, M.D.

References Kandoth C, Schultz N et al (2013) Integrated genomic characterization of endometrial carcinoma. Nature 497(7447):67–73 Wang Y, Li L, Douville C et al (2018) Evaluation of liquid from the Papanicolaou test and other liquid biopsies for the detection of endometrial and ovarian cancers. Sci Transl Med 10:eaap8793, 1–9

Contents

1

Benign Diseases of the Vulva . . . . . . . . . . . . . . . . . . . . . . . . . . Demaretta S. Rush and Edward J. Wilkinson

1

2

Precursor Lesions and Malignant Tumors of the Vulva . . . . Edward J. Wilkinson and Demaretta S. Rush

65

3

Diseases of the Vagina . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131 Marisa R. Nucci, Richard J. Zaino, and Robert J. Kurman

4

Benign Diseases of the Cervix . . . . . . . . . . . . . . . . . . . . . . . . . 193 Thomas C. Wright and Brigitte M. Ronnett

5

Precancerous Lesions of the Cervix . . . . . . . . . . . . . . . . . . . . 239 Thomas C. Wright, Brigitte M. Ronnett, and Robert J. Kurman

6

Carcinoma and Other Tumors of the Cervix . . . . . . . . . . . . . 315 Edyta C. Pirog, Thomas C. Wright, Brigitte M. Ronnett, and Robert J. Kurman

7

Benign Diseases of the Endometrium . . . . . . . . . . . . . . . . . . . 375 Ricardo R. Lastra, W. Glenn McCluggage, and Lora Hedrick Ellenson

8

Precursors of Endometrial Carcinoma . . . . . . . . . . . . . . . . . 439 Lora Hedrick Ellenson, Brigitte M. Ronnett, and Robert J. Kurman

9

Endometrial Carcinoma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 473 Lora Hedrick Ellenson, Brigitte M. Ronnett, Robert A. Soslow, Ricardo R. Lastra, and Robert J. Kurman

10

Mesenchymal Tumors of the Uterus . . . . . . . . . . . . . . . . . . . . 535 Esther Oliva, Charles J. Zaloudek, and Robert A. Soslow

11

Diseases of the Fallopian Tube and Paratubal Region Russell Vang

12

Nonneoplastic Lesions of the Ovary . . . . . . . . . . . . . . . . . . . . 715 Julie A. Irving and Philip B. Clement

. . . . . 649

ix

x

Contents

13

Diseases of the Peritoneum . . . . . . . . . . . . . . . . . . . . . . . . . . . 771 Julie A. Irving and Philip B. Clement

14

Epithelial Tumors of the Ovary . . . . . . . . . . . . . . . . . . . . . . . 841 Jeffrey D. Seidman, Brigitte M. Ronnett, Ie-Ming Shih, Kathleen R. Cho, and Robert J. Kurman

15

Sex Cord-Stromal, Steroid Cell, and Other Ovarian Tumors with Endocrine, Paraendocrine, and Paraneoplastic Manifestations . . . . . . . . . . . . . . . . . . . . . . . . 967 Paul N. Staats and Robert H. Young

16

Germ Cell Tumors of the Ovary . . . . . . . . . . . . . . . . . . . . . . . 1047 Kruti P. Maniar and Russell Vang

17

Nonspecific Tumors of the Ovary, Including Mesenchymal Tumors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1125 Lauren E. Schwartz and Russell Vang

18

Metastatic Tumors of the Ovary . . . . . . . . . . . . . . . . . . . . . . . 1151 Melinda F. Lerwill and Robert H. Young

19

Diseases of the Placenta . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1223 Rebecca N. Baergen, Deborah J. Gersell, and Frederick T. Kraus

20

Gestational Trophoblastic Tumors and Related Tumorlike Lesions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1307 Ie-Ming Shih, Brigitte M. Ronnett, Michael Mazur, and Robert J. Kurman

21

Hematologic Neoplasms and Selected Tumorlike Lesions Involving the Female Reproductive Organs . . . . . . . 1377 Judith A. Ferry

22

Soft Tissue Lesions Involving Female Reproductive Organs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1405 John F. Fetsch and William B. Laskin

About the Editors

Dr. Robert J. Kurman, M.D., is the Emeritus Richard W. TeLinde Distinguished Professor of Gynecologic Pathology at the Johns Hopkins University School of Medicine and former Director of the Division of Gynecologic Pathology at the Johns Hopkins Hospital where his career was devoted to diagnosis, research, and teaching in the field of gynecologic pathology. His research activities began in the early 1970s with studies of germ cell tumors of the ovary and testis and gestational trophoblastic disease at which time he pioneered the application of immunohistochemistry (IHC) on formalin-fixed, paraffin-embedded tissue. These studies were among the first describing how IHC could be applied to surgical pathology. During this time, he also undertook studies on the relationship of endometrial hyperplasia to carcinoma which led to the development of a classification system of endometrial hyperplasia that was later adopted by the World Health Organization. In the late 1970s and 1980s, his work on establishing the link between HPV and cervical cancer played a role in the application of molecular testing for HPV as a screening tool. This also was instrumental in the development of “The Bethesda System (TBS) for Reporting Cervical/Vaginal Cytologic Diagnoses” that replaced the previous Papanicolaou classification system and is now the standard cytology xi

xii

classification system in the USA and abroad. He has also been a Consultant for Merck in their clinical HPV vaccine trials. In the last 15 years, he has focused on elucidating the pathogenesis of epithelial ovarian cancer. By collaborating not only with other pathologists but also with molecular biologists and epidemiologists, he has demonstrated the value of a multimodal approach to ovarian cancer research. His vision has led to the proposal of a new disease model, which synthesizes clinical observations and pathobiological mechanisms and validates conceptual hypotheses with molecular data, thereby bringing new insights to the field. For example, based on morphologic and molecular genetic studies, a dualistic model of ovarian carcinogenesis was developed, which has now become widely accepted in the field. In addition, the studies implicating a precursor lesion in the fallopian tube as the origin of many so-called ovarian carcinomas have dramatically changed our thinking on this subject, with important implications for ovarian cancer screening and prevention. His research has resulted in the publication of nearly 300 original peer-reviewed papers and over 150 review articles and book chapters. Dr. Kurman’s influence extends well beyond these research efforts. He has recruited and mentored pathologists and researchers who have become distinguished gynecologic pathologists. Pathologists know him as an author and editor through his significant educational publications, including Blaustein’s Pathology of the Female Genital Tract (third, fourth, fifth, and sixth editions), Diagnosis of Endometrial Biopsies and Curettings: A Practical Approach (two editions), the AFIP Fascicles on Tumors of the Cervix, Vagina, and Vulva (third and fourth series) and Tumors of the Uterine Corpus and Gestational Trophoblastic Disease (third series), and the 2014 World Health Organization Classification of Tumours of the Female Reproductive Organs. He is sought after as a lecturer worldwide and has contributed to the advancement of the field through his leadership in professional societies, including being President of the International Society of Gynecologic Pathologists, participation in international committees, and membership on editorial boards of numerous journals. In recognition of his scholarship and leadership activities, he was elected as an Honorary Fellow of the Royal College of Pathologists and the Austrian Society of Pathologists.

Lora Hedrick Ellenson, M.D., is Professor and Director of Gynecologic Pathology in the Department of Pathology and Laboratory Medicine at Weill Cornell Medical College in New York City. Her undergraduate work was done at the University of California, Berkeley, followed by medical school at Stanford University School of Medicine and residency training at the Johns

About the Editors

About the Editors

xiii

Hopkins Medical Institutions. Upon completing the Anatomic Pathology program, she joined the laboratory of Drs. Bert Vogelstein and Ken Kinzler as a postdoctoral fellow. She simultaneously trained as a Gynecologic Pathologist with Dr. Robert Kurman. Following her initial work studying the molecular biology of colon cancer, at the request of Dr. Kurman, she joined the Department of Pathology in the Division of Gynecologic Pathology where she established an independent research program to study the molecular genetics of endometrial carcinoma. Dr. Ellenson’s laboratory was one of the first to document the high frequency of TP53 mutations in uterine serous carcinoma and microsatellite instability and PTEN mutations in sporadic endometrioid carcinoma. In 1998, after becoming an Associate Professor at Johns Hopkins, she moved to Weill Cornell Medical College to oversee the Division of Gynecologic Pathology. Reflecting her expansive interests, Dr. Ellenson has, throughout her career, maintained an NIH-funded laboratory and simultaneously practiced diagnostic gynecological pathology. Since 2001, she has been an Associate Editor for The American Journal of Pathology. She has also been on the Editorial Committee for Annual Reviews of Pathology: Mechanisms of Disease for over 10 years. Dr. Ellenson has contributed chapters on the female genital tract for the last three editions of Robbins and Cotran Pathologic Basis of Disease and more recently for Basic Pathology. More recently, she became Editor in Chief of the International Journal of Gynecological Pathology and is on the editorial committee for the next edition of the WHO Classification of Tumours of Female Reproductive Organs.

Brigitte M. Ronnett, M.D., is Professor of Pathology and Gynecology and Obstetrics at the Johns Hopkins Medical Institutions. She graduated from Northwestern University and received her medical degree from the University of Chicago Pritzker School of Medicine. She completed residency training in pathology at the Johns Hopkins Hospital, a surgical pathology fellowship at Memorial Sloan-Kettering Cancer Center, and both surgical pathology and gynecologic pathology fellowships at the Johns Hopkins Hospital. She has been a member of the Division of Gynecologic Pathology at Johns Hopkins since 1995. Her clinical efforts are focused on a large gynecologic pathology consultation practice. Her research efforts have focused on ovarian mucinous tumors, uterine cervical and endometrial pathology, and hydatidiform moles.

Contributors

Rebecca N. Baergen Department of Pathology and Laboratory Medicine, Weill Cornell Medicine, New York-Presbyterian Hospital, New York, NY, USA Kathleen R. Cho Department of Pathology, University of Michigan Medical School, Ann Arbor, MI, USA Philip B. Clement Department of Pathology, Vancouver General Hospital, Vancouver, BC, Canada Judith A. Ferry James Homer Wright Pathology Laboratories of the Massachusetts General Hospital, Department of Pathology, Harvard Medical School, Boston, MA, USA John F. Fetsch Soft Tissue Pathology, The Joint Pathology Center, Silver Spring, MD, USA Deborah J. Gersell Department of Laboratory Medicine, St. John’s Mercy Medical Center, St. Louis, MO, USA Lora Hedrick Ellenson Department of Pathology and Laboratory Medicine, Division of Gynecologic Pathology, Weill Cornell Medical College and New York Presbyterian Hospital, New York, NY, USA Julie A. Irving Department of Laboratory Medicine, Pathology, and Medical Genetics, Royal Jubilee Hospital, Victoria, BC, Canada Frederick T. Kraus Perinatal Biology Laboratory, Department of Obstetrics and Gynecology, Washington University School of Medicine, St. Louis, MO, USA Robert J. Kurman Department of Gynecology, Obstetrics, Pathology and Oncology, Division of Gynecologic Pathology, Johns Hopkins University School of Medicine, Baltimore, MD, USA William B. Laskin Department of Pathology, Yale Surgical Pathology, Yale School of Medicine, New Haven, CT, USA xv

xvi

Ricardo R. Lastra Department of Pathology, University of Chicago, Chicago, IL, USA Melinda F. Lerwill Department of Pathology, Massachusetts General Hospital and Harvard Medical School, Boston, MA, USA Kruti P. Maniar Department of Pathology, Division of Surgical Pathology, Northwestern University Feinberg School of Medicine, Chicago, IL, USA Michael Mazur Department of Pathology, SUNY Upstate Medical University, Syracuse, NY, USA W. Glenn McCluggage Department of Pathology, Royal Group of Hospitals Trust, Belfast, UK Marisa R. Nucci Division of Women’s and Perinatal Pathology, Department of Pathology, Brigham Women’s Hospital, Boston, MA, USA Esther Oliva Massachusetts General Hospital, Boston, MA, USA Department of Pathology, Harvard Medical School, Boston, MA, USA Edyta C. Pirog Department of Pathology, Weill Cornell, Medical College and New York Presbyterian Hospital, New York, NY, USA Brigitte M. Ronnett Department of Pathology, Division of Gynecologic Pathology, Johns Hopkins University School of Medicine, Baltimore, MD, USA Demaretta S. Rush Department of Pathology, University of Arizona College of Medicine, Tucson, AZ, USA Lauren E. Schwartz Department of Pathology and Laboratory Medicine, Perelman School of Medicine at the University of Pennsylvania, Philadelphia, PA, USA Jeffrey D. Seidman Center for Devices and Radiological Health, Office of In Vitro Diagnostics and Radiological Health, Food and Drug Administration, Silver Spring, MD, USA Ie-Ming Shih Gynecologic Pathology Laboratory in the Department of Gynecology and Obstetrics, Johns Hopkins University School of Medicine, Baltimore, MD, USA Robert A. Soslow Department of Pathology, Memorial Sloan-Kettering Cancer Center, New York, NY, USA Department of Pathology and Laboratory Medicine, Weill Cornell Medical College and New York Presbyterian Hospital, New York, NY, USA Paul N. Staats Department of Pathology, University of Maryland School of Medicine, Baltimore, MD, USA

Edyta C. Pirog has retired.

Contributors

Contributors

xvii

Russell Vang Department of Pathology, Division of Gynecologic Pathology, The Johns Hopkins Medical Institutions, Baltimore, MD, USA Edward J. Wilkinson Department of Pathology, Immunology and Laboratory Medicine, University of Florida College of Medicine, Gainesville, FL, USA Thomas C. Wright Department of Pathology and Cell Biology, Columbia University, New York, NY, USA Robert H. Young Anatomic Pathology, James Homer Wright Pathology Laboratories, Massachusetts General Hospital, Harvard Medical School, Boston, MA, USA Richard J. Zaino Hershey, PA, USA Charles J. Zaloudek Department of Pathology, University of California, San Francisco, San Francisco, CA, USA

1

Benign Diseases of the Vulva Demaretta S. Rush and Edward J. Wilkinson

Contents Anatomy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

3

Developmental Abnormalities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Congenital . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Acquired . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

5 5 5

Infectious Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 Bacterial . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 Viral . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 Fungal . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18 Parasitic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19 Inflammatory Dermatoses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Spongiotic Pattern . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Allergic/Irritant Contact Dermatitis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Atopic Dermatitis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Acanthotic Pattern . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Psoriasis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Lichen Simplex Chronicus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Lichenoid Pattern . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Lichen Sclerosus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Lichen Planus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Fixed Drug Eruption . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Dermal Homogenization/Sclerosus Pattern . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Vesiculobullous Pattern . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Pemphigoid . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Pemphigoid Gestationis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Linear IgA Disease . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

19 20 20 21 22 22 23 25 25 27 28 29 30 30 32 32

D. S. Rush Department of Pathology, University of Arizona College of Medicine, Tucson, AZ, USA e-mail: [email protected] E. J. Wilkinson (*) Department of Pathology, Immunology and Laboratory Medicine, University of Florida College of Medicine, Gainesville, FL, USA e-mail: [email protected]fl.edu # Springer Nature Switzerland AG 2019 R. J. Kurman et al. (eds.), Blaustein’s Pathology of the Female Genital Tract, https://doi.org/10.1007/978-3-319-46334-6_1

1

2

D. S. Rush and E. J. Wilkinson Pemphigus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Bullous Systemic Lupus Erythematosus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Acantholytic Pattern . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Hailey–Hailey Disease . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Darier Disease . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Papular Acantholytic Dyskeratosis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Granulomatous Pattern . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Crohn Disease . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Melkersson–Rosenthal Syndrome and Granulomatous Vulvitis . . . . . . . . . . . . . . . . . . . . . . . . Sarcoidosis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Vasculopathic Pattern . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Vulvar Aphthosis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Behçet Disease . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Plasma Cell Vulvitis (Vulvitis of Zoon) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

32 34 35 35 36 37 37 37 38 38 39 39 40 40

Miscellaneous Dermatoses Lacking a Dominant Histologic Pattern . . . . . . . . . . . . . . . . Amicrobial Pustulosis of the Folds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Erythema Multiforme/Stevens–Johnson Syndrome . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Pyoderma Gangrenosum . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

41 41 42 42

Vulvodynia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43 Pigment Disorders and Benign Melanocytic Lesions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Postinflammatory Alterations in Pigmentation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Lentigo Simplex . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Vulvar Melanosis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Common Nevi . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Atypical Genital Nevi . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Dysplastic Nevi . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

44 44 44 45 45 46 47

Benign Squamous Proliferations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Fibroepithelial Polyp (Acrochordon) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Vestibular Papilloma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Seborrheic Keratosis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

47 47 47 48

Keratoacanthoma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49 Cysts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Epithelial Inclusion Cyst . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Bartholin Cyst and Abscess . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Mucous Cyst . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Ciliated Cysts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Gartner Duct Cyst . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Mammary-Like Cysts (Hidrocystoma) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Cysts of the Canal of Nuck . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Skene Gland Cyst . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

49 49 50 50 51 51 51 51 51

Benign Glandular Lesions of the Vulva . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Lesions of the Anogenital Mammary-Like Glands . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Lesions of Sweat Gland Origin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Solid Lesions of Bartholin Gland Origin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Prostate-Like Tissue and Solid Lesions of Skene’s Gland Origin . . . . . . . . . . . . . . . . . . . . . . . Adenoma of Minor Vestibular Glands . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Endometriosis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

51 52 53 54 54 54 55

Benign Lesions of Folliculo-Sebaceous Origin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55 Benign Lymphovascular Lesions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Hemangiomas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Angiokeratoma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Lymphangioma Circumscriptum . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

55 55 56 56

1

Benign Diseases of the Vulva

3

Miscellaneous Tumor-Like Lesions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Verruciform Xanthoma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Idiopathic Vulvar Calcinosis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Vulvar Amyloidosis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

57 57 57 58

Benign Lesions of the Urethra . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Urethral Prolapse . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Urethral Caruncles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Malakoplakia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Condyloma Acuminatum . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

58 58 58 58 58

References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59

Anatomy The external female genitalia include the mons pubis, labia majora and minora, clitoris, and vestibule. The mons is the portion of hair-bearing skin and associated subcutaneous tissue overlying the pubic symphysis. Inferior to the pubis, the hairbearing skin of the mons divides into two folds, the labia majora, which join posteriorly at the perineal body, just anterior to the anus. Medial to the labia majora are a second set of folds, the labia minora, each of which divides at the anterior end into two sets of smaller folds. The superior folds fuse in the midline anterior to the clitoris, forming the prepuce, or clitoral hood, and the inferior folds fuse in the midline posterior to the clitoris, forming the frenulum. Posteriorly, the labia minora fuse without further divisions, at the fourchette, posterior to the introitus and anterior to the perineal body. The clitoris consists of a bundle of erectile tissue situated in the midline, with just the tip, or glans, visible between the prepuce and frenulum. Beneath the surface is the body of the clitoris which branches at the base into two crurae running along the pubic rami in the deep soft tissue. The vestibule is a roughly diamond-shaped area bounded anteriorly by the frenulum of the clitoris, laterally by the medial edges of the labia minora, and posteriorly by the fourchette. Posterior to the frenulum on the vestibule is the urethral orifice, and posterior to that lies the vaginal opening, bounded by the hymen or remnants thereof. The external anatomy is illustrated in Fig. 1. Histologically, the hair-bearing skin of the mons is similar to the nongenital skin of the rest of the body, consisting of keratinizing squamous

epithelium with all of the usual adnexal skin structures, including hair follicles, sebaceous glands, eccrine sweat glands, and sensory receptors. The subcutaneous tissue of the mons is predominantly adipose tissue. The composition of the labia majora is nearly identical, but with an additional component of smooth muscle in the subcutaneous tissue and apocrine sweat glands deep to the epithelium. On the medial portion of the labia majora, the skin becomes hairless, and there is a corresponding absence of hair follicles, as well as an absence of sweat glands, although sebaceous glands remain and may be rather prominent in appearance, forming slightly elevated, pale areas known as Fordyce spots. The epithelium becomes thinner and keratinization decreases on the medial surface of the labia majora as well. The labia minora are composed of connective tissue rich in elastic fibers and blood vessels but without adipose tissue. The mucosa of the labia minora is similar to the medial portion of the labia majora, with which it is continuous, with sebaceous glands disappearing toward the medial side. The theoretical line of Hart, which runs along the medial edge of the labia minora, marks the junction between the keratinizing epithelium of the labia minora and the nonkeratinizing squamous mucosal lining of the vestibule. The mucosa of the vestibule is glycogenated in women of reproductive age, or under estrogen influence, and resembles vaginal mucosa. This epithelium merges with the transitional epithelium at the urethral meatus and with the duct openings of various submucosal glands.

4

D. S. Rush and E. J. Wilkinson

Interlabial sulcus

Prepuce Frenulum Labia minora

Labia majora Skene duct Hymeneal tagmentum Vestibule Bartholin duct Fourchette

Fig. 2 Skene gland. The gland is lined by a mixture of urothelium and mucin-secreting columnar epithelium

Perineum

Fig. 1 Anatomy of the vulva. The vestibule is the diamond-shaped area between the medial edges of the labia minora, extending from the frenulum to the fourchette, and containing the urethral and vaginal openings

The glandular structures of the vulva are not normally visible on external exam, or even on careful dissection, and many can only be seen microscopically unless enlarged by disease. Aside from the sweat glands in the hair-bearing portions of the vulva, which are no different from those elsewhere on the skin, there are a number of glandular structures specific to the vulvar area. Situated in the interlabial sulcus, surrounding the entire labia minora and clitoris, and on the perineum and surrounding the anus, are specialized apocrine glands, the anogenital mammary-like glands. These glands have a simple columnar epithelium with apical snouts and myoepithelium surrounding the glandular epithelium (van der Putte 1991). They may be located as far as 3.9 mm from the surface (Konstantinova et al. 2017a) and have long and wide coiled ducts that open to the surface. The paired Skene’s glands, homologues of the male prostate, are composed of pseudostratified mucus-secreting columnar epithelium and open to the external surface on both sides of the urethral meatus and along the posterior and lateral aspects of the urethra itself.

The ducts are lined by transitional epithelium (Fig. 2). The Bartholin glands are bilateral racemose, tubuloalveolar glands, with acini composed of simple, columnar, mucus-secreting epithelium (Fig. 3). The Bartholin duct, approximately 2.5 cm in length, has three types of epithelial linings depending on the location within the duct. It is lined proximally by mucussecreting epithelium, distally by transitional epithelium, and, at its exit, just external to the hymen ring of the vestibule posterolaterally, by squamous epithelium. The minor vestibular glands ring the vestibule and extend from the frenulum on both sides of the urethral meatus around the external base of the hymenal ring to the fourchette, lying within 1–2.5 mm of the superficial epithelium and communicating with the vestibular surface. They are composed of acini lined by simple columnar mucus-secreting epithelium. The major blood supply to the vulva is provided via the anterior and posterior labial branches of the superficial and deep external pudendal arteries, which branch from the femoral artery, and the internal pudendal arteries, which branch from the internal iliac arteries. The clitoris, including the crura and corpora cavernosa, is supplied separately by the deep arteries of the clitoris, whereas the anterior vaginal artery supplies blood flow to the vestibule and the Bartholin glands. The venous return parallels the arterial supply. The nerve supply to the vulva includes sensory nerves,

1

Benign Diseases of the Vulva

5

the glans clitoris joins the lymphatics of the urethra, traverses the urogenital diaphragm, and merges with the lymphatic plexus on the anterior surface of the bladder. From there, drainage is into the internal iliac, obturator, and external iliac nodes.

Developmental Abnormalities Congenital

Fig. 3 Bartholin gland and duct. The glandular acini are composed of simple mucin-secreting columnar epithelium which merges with the transitional epithelium lining the duct

special receptors, and autonomic nerves to the vessels and various glands. The major nerves of the vulva derive from the anterior (ilioinguinal) and posterior (pudendal) labial nerves. The clitoris is innervated by the dorsal nerve of the clitoris and the cavernous nerves of the clitoris, which also supply the vestibule. The entire vulva, with the exception of the clitoris, drains to the femoral and inguinal lymph nodes. Delicate intercommunicating lymphatic vessels extend to the labia minora, clitoral prepuce, and vestibule, bypassing the clitoris. The lymphatic bed of the labia majora drains in an anterosuperior direction toward the mons, joining the lymphatic vessels from the labia minora and prepuce, and then into the ipsilateral inguinal and femoral nodes. Some contralateral flow also may occur into the superior medial nodes of the femoral group. The superficial inguinal lymph nodes, consisting of 8–10 nodes on each side, divided into a superior oblique and an inferior ventral group, are the major nodes that drain the vulva and therefore are included in a radical vulvectomy. The superior oblique group is found about the Poupart ligament, and the inferior ventral group lies above the junction of the saphenous vein and fascia lata. Lymphatic drainage from the clitoris and midline perineum proceeds bilaterally in more than 67% of cases and may bypass the superficial nodes. A second minor lymphatic pathway from

Congenital anomalies of the vulva may include absence, hypoplasia, hyperplasia, or dupilication of various portions of the anatomy. Congenital absence of the clitoris and external genitalia has been described. In Müllerian agenesis, the external genitalia is largely intact, but the hymen and vagina are absent, usually represented only by a depression in the vestibular area. True hypoplasia of the labia minora occurs infrequently and may be a sign of defective steroidogenesis. Hypertrophy of the labia minora is more common, usually becoming more evident at puberty, and is defined as a measurement of more than 4 cm from the base to the outer edge (Margesson 2006). Clitoral enlargement may be seen in newborns with adrenogenital syndrome or who have been exposed in utero to exogenous maternal androgen therapy, as well as in hermaphroditism and, rarely, with lipodystrophy (Ridley and Neill 1999). Labial fusion may also may be present with intersex disorders, although slight fusion of the labia minora may be seen in infants without apparent cause and typically responds to topical estrogen cream. Urethral anomalies may result in aberrant locations of the urethral opening in the vagina or adjacent to the hymen rather than in the upper portion of the vestibule (Kaufman 1994). Duplication of the vulva is extremely rare and usually is associated with duplication of the internal Müllerian system and rectum as well.

Acquired Many vulvar conditions may result in altered anatomy later in life. In some cases, labial hypertrophy

6

may develop over time in association with chronic irritation, as from indwelling catheters. Labial fusion may also be an acquired abnormality, secondary to adhesions and scarring in the course of lichen sclerosus (LS), lichen planus (LP), or other inflammatory conditions, or female genital mutilation procedures. Female genital mutilation is practiced in parts of Africa, Asia, and the Middle East, and affected patients are increasingly encountered in Western medical practice, where they present with a variety of alterations to the vulvar anatomy which may be complex (AbdulCadir et al. 2016). These procedures involve removal of various portions of the vulva and may include reapproximations of the cut margins to partially or completely obscure the vaginal opening. In addition to destroying the normal anatomy, female genital mutilation may lead to complications which can further distort the vulva. Epidermal inclusion cysts of the vulva seem to be a particularly common complication and often present many years after the procedure as large, frequently pedunculated masses which may measure up to 7 cm and may be mistaken for clitoral enlargement due to hormonal factors or neoplasia (Riszk et al. 2007, Osarumwense 2010, Asante et al. 2010). A variety of tumors including granular cell tumors, hemangiomas, and vascular, neural, and smooth muscle tumors may also cause acquired clitoral enlargement.

Infectious Conditions In developed nations, the most prevalent infections of the vulva are sexually transmitted viral diseases due to human papillomavirus (HPV), herpes simplex virus (HSV), and molluscum contagiousum and syphilis, a bacterial disease. Many other pathogens, however, not all of them sexually transmitted, may involve the vulva. Most vulvar infections are readily diagnosed on clinical grounds or by ancillary testing. Biopsies of these conditions are generally performed only when the presentation is atypical or the diagnosis is otherwise in doubt. Although the histopathologic findings are frequently nonspecific, there may be features at least suggestive of a

D. S. Rush and E. J. Wilkinson

particular etiologic agent, which can guide the selection of additional testing to establish a definitive diagnosis. A summary of the principal vulvar infections presented in this chapter is provided in Table 1.

Bacterial Syphilis For many years, the rates of syphilis infection had been declining in the West, reaching a historic low in 2000, but began to climb again, particularly in the human immunodeficiency virus (HIV)-positive population (Cohen et al. 2013; Hope-Rapp et al. 2010) and in pregnant women, in subsequent years. The disease is caused by Treponema pallidum, a spirochete which does not stain with Gram stain and which cannot be cultured, making it particularly important to be aware of the characteristic clinical and microscopic features.

Clinical Features The disease manifests in phases. A week to 3 months after initial exposure, the primary lesion, or chancre, appears as a papule which develops into a painless, indurated, shallow, clean-based ulcer with raised edges. Chancres are usually single but may be multiple, especially in HIV-positive patients (Cohen et al. 2013). They may occur on inconspicuous surfaces, such as the cervix, anal mucosa, or oropharynx, and it is not uncommon for the primary lesion to go unnoticed and untreated. Typically, the chancre will heal on its own within 2–6 weeks. The timing of the development of the secondary stage of syphilis is variable but typically manifests 2–8 weeks after the chancre. At this point, the patient may present with a skin rash that often involves mucous membranes as well as the palms of the hands and soles of the feet, as well as with systemic symptoms of fever, headache, pharyngitis, and lymphadenopathy. Secondary syphilis may be accompanied by new painless cutaneous lesions known as condylomata lata and by mucous patches. Condylomata lata appear as elevated papules or plaques with a velvety papillary surface,

1

Benign Diseases of the Vulva

7

Table 1 Summary of significant vulvar infections, their clinical manifestations, and suggested ancillary tests to assist in their diagnosis Infectious conditions of the vulva Diagnosis Etiologic agent Bacterial Syphilis Treponema pallidum

Clinical manifestation(s)

Ancillary tests

Primary: Chancre (ulcer) Secondary: Condylomata lata (papule)

Dark field examination Immunofluorescence Immunohistochemistry Polymerase chain reaction (PCR) Serology Culture Warthin-Starry or gram stain (“Donovan bodies”) Serology Culture PCR Culture PCR Gram or Giemsa stain Culture Acid-fast stain

Granuloma inguinale

Klebsiella granulomatis

Ulcer

Lymphogranuloma venereum (LGV)

Chlamydia trachomatis, types L1, L2, L3

Ulcer, +/ inguinal lymphadenopathy

Chancroid

Haemophilus ducreyi

Ulcer

Tuberculosis

Mycobacterium tuberculosis

Ulcers, swellings, exophytic lesions

HPV, usually types 6 and 11 HSV, types 1 and 2

Condyloma acuminatum (genital warts) Ulcers, atypical lesions in immunosuppressed patients

Herpes zoster

Unilateral vesicles and ulcers

Viral Condyloma acuminatum Herpes

Varicella (vulvar shingles) Molluscum contagiosum Fungal Various

Parasitic Various

Molluscum contagiocsum

Not necessary Culture PCR Immunohistochemistry PCR Immunohistochemistry Not necessary

Various

Scaly plaques (most commonly)

Skin scrapings for KOH prep PAS or silver stains Culture

Various

Various

Usually not necessary

involving the vulva, perianal, and inguinal area and measuring up to 3 cm in diameter (Fig. 4), while mucous patches appear as gray-white erosions on the non-hair-bearing squamous mucosa of the inner labia. Both condylomata lata and mucous patches are heavily populated with spirochetes and are highly infectious. The tertiary phase of syphilis, which develops after many years of latency, is extremely uncommon today. Its most significant manifestations involve the cardiovascular and central nervous systems, although cutaneous or mucosal granulomatous lesions, or gummas, may also develop in this phase.

Microscopic Findings The primary chancre is characterized by ulceration of the epithelium with intense acute and chronic submucosal and perivascular inflammation, characterized by the presence of large numbers of plasma cells. Granulomatous inflammation may also be present. Histologic examination of condylomata lata reveals marked acanthosis and hyperkeratosis, patchy parakeratosis, and superficial intraepithelial microabscesses. Also present is a chronic inflammatory response within the dermis with a perivascular distribution, similar to that in the primary chancre but with a greater

8

predominance of plasma cells. The arteritis in the lesions of both primary and secondary syphilis may be sufficiently severe to result in obliteration of the smaller vessels. Other patterns of inflammation may also be seen in secondary syphilis, including a psoriasiform pattern, a lichenoid form, and a pustular form, in which the prominence of plasma cells in the inflammatory infiltrate and the endarteritis should raise the suspicion for syphilis (Fig. 5). Dieterle, Warthin–Starry, or Steiner stains for spirochetes may be of some use to identify the organisms, which are characteristically located at

Fig. 4 Condyloma lata. The lesions are exophytic, but with a relatively smooth, uniform surface

Fig. 5 (a) Secondary syphilis. This lesion shows prominent pseudoepitheliomatous hyperplasia overlying a dense dermal inflammatory infiltrate containing abundant plasma

D. S. Rush and E. J. Wilkinson

the dermal-epidermal junction and in a perivascular distribution in the superficial dermis, but these stains may be negative even with active infection. Immunohistochemical staining is now available and considerably easier to interpret (Fig. 6), and the organisms may also be identified on dark-field examination of fresh sera from an active lesion, by a fluorescent conjugated antibody technique, which employs a dried smear preparation, and by PCR. These tests are all much more sensitive and specific for detection of the organisms than silver staining techniques but are also more expensive and may not be available in all clinical settings. For confirmation of the diagnosis, serologic tests, which are widely available and may be used in place of or in addition to microscopic examination of clinical samples, are most commonly used. Serology is very sensitive in secondary syphilis but may be falsely negative in primary syphilis, in which case the test may need to be repeated at a later time. Clinical Course and Treatment Approximately 30% of patients with primary syphilis will undergo spontaneous remission of the disease. Those who are not treated or who do not achieve spontaneous remission may progress to tertiary syphilis, which, if it continues to go untreated, may prove to be fatal in 10% of those afflicted. Penicillin or another appropriate systemic antibiotic is the treatment of choice for all stages of the disease.

cells. (b) Deeper in the dermis, the inflammatory infiltrate shows a perivascular distribution

1

Benign Diseases of the Vulva

Fig. 6 Immunohistochemical stain for Treponema pallidum in the lesion illustrated in Fig. 5. The organisms are concentrated at the dermo-epidermal junction, and the characteristic spiral structure is evident

Granuloma Inguinale Granuloma inguinale, also known as donovanosis or granuloma venereum, is a sexually transmitted disease endemic to Papua New Guinea, South Africa, India, Brazil, and Australia. Occasional local outbreaks of the disease are seen in the West. The disease is caused by a gram-negative, heavily encapsulated rod formerly known as Calymmatobacterium granulomatis but recently reclassified as Klebsiella granulomatis (O'Farrell and Moi 2016). Clinical Features In women, the primary lesions occur on the vulva, vagina, or cervix. The lesions usually appear within 1 week to 1 month of exposure; anal coitus or fecal contamination of the vulva or vagina has been posited as the mode of transmission (Wilkinson and Stone 2008). Four types of lesions have been described: ulcerogranulomatous, hypertrophic, necrotic, and sclerotic/cicatricial, but the typical lesion begins as a papule which develops into an ulcer which then progressively increases in size. Despite its name, granuloma inguinale involves the inguinal region in only 10% of cases (O'Farrell and Moi 2016). Microscopic Findings Histologically, the main portion of the lesion consists of granulation tissue associated with an extensive chronic inflammatory cell infiltrate and endarteritis. The ulcer is usually covered with a

9

Fig. 7 Granuloma inguinale. Giemsa stain shows numerous intracytoplasmic “Donovan bodies” with a characteristic halo surrounding the organisms

fibrinous exudate, while the surface epithelium adjacent to the ulcer may show prominent pseudoepitheliomatous hyperplasia. Necrosis and microabscesses may be seen within the epidermis. The granulation tissue is accompanied by a dense mixed inflammatory cell infiltrate, consisting predominantly of plasma cells and mononuclear cells with few lymphocytes, which extends into the dermis. All of these findings, however, are nonspecific, and the diagnosis cannot be made with certainty without identification of the offending organisms. These show a characteristic bipolar staining pattern, likened to a “safety-pin,” on Warthin–Starry stain or Giemsa stain and are referred to as “Donovan bodies.” They are found within the cytoplasm of large vacuolated histiocytes in the lesion, as well as intracellularly (Fig. 7). Organisms may also be identified by preparing smears from the active lesion or a touch preparation of a biopsy from the edge of the ulcer, allowing the preparation to air dry, then fixing in methanol and staining with Giemsa stain, or by culture. Clinical Course and Treatment The lesions of granuloma inguinale grow more rapidly during pregnancy (O'Farrell and Moi 2016) and in patients with concurrent HIV (BastaJuzbasic and Ceovic 2014), in whom the lesions may persist longer and require a longer duration of treatment. In rare cases, the organisms may disseminate to involve other organs, most commonly liver and bone (O'Farrell and Moi 2016). Treatment with

10

D. S. Rush and E. J. Wilkinson

appropriate antibiotics, continued until the lesion is completely healed, is curative.

Lymphogranuloma Venereum (LGV) LGV is a sexually transmitted disease endemic to Africa, Asia, and Central and South America, caused by Chlamydia trachomatis types L1, L2, and L3. It is uncommon in the West, but sporadic outbreaks have been noted in recent years, principally involving HIV-positive homosexual male patients (French et al. 2005). Clinical Features The disease classically manifests in three phases, following an incubation period of 3–30 days. The initial lesion is a painless papule which may progress to an ulcer and heals within a week. Two to 6 weeks later, a second phase may develop, usually characterized by painful inflammation of inguinal lymph nodes which may progress to the development of draining sinus tracts. This classic presentation with inguinal involvement has become increasingly uncommon, however, especially in women (Basta-Juzbasic and Ceovic 2014), with proctitis, vaginitis, and cervicitis accompanied by constitutional symptoms now more frequently observed. A third phase of the disease may develop in patients who continue untreated, consisting of progressively worsening proctocolitis and abscess development with lymphatic obstruction and fibrosis and stricture of the vagina and rectum. Microscopic Findings The initial lesion of LGV heals rapidly and is very rarely biopsied. Even if it is, the histology is nonspecific, showing giant cells along with lymphocytes and plasma cells in the inflammatory infiltrate surrounding the ulcer (Fig. 8). Older lesions may exhibit extensive fibrosis of the dermis and sinus tracts. Although the organisms are extremely difficult to identify on light microscopy, evaluation with special stains or other techniques is important to rule out other infectious disease with a similar presentation. The diagnosis rests on the results of serologic testing as well as identification of the organism in culture or by PCR testing.

Fig. 8 LGV An intense superficial and deep chronic inflammatory infiltrate composed predominantly of lymphocytes and plasma cells is present

Clinical Behavior and Treatment Treatment is with antibiotics and may also include aspiration or incision and drainage of buboes to prevent progression to deep ulcers and fistulas and aid in healing. Timely treatment will prevent progression to the third, potentially disfiguring, phase of disease.

Chancroid Endemic in Africa, Asia, and the Caribbean, where it may be responsible for up to 56% of genital ulcer disease (Mohammed and Olumide 2008), chancroid, a sexually transmitted disease caused by Haemophilus ducreyi, is relatively rare in the West. Clinical Features The disease presents after a 3- to 7-day incubation period, initially as a small papule which progresses to a pustule and then to a soft painful ulcer. Lesions may be single or multiple and tend to be small, measuring approximately

1

Benign Diseases of the Vulva

1–2 mm in diameter, but multiple lesions may coalesce to form ulcers approaching 3 cm in diameter. In women, the ulcers may involve the fourchette, labia, vestibule, clitoris, and perianal area and are often subclinical. In 40–50% of patients, painful inguinal adenopathy develops a few days to 2 weeks following the ulcer (Mohammed and Olumide 2008; Basta-Juzbasic and Ceovic 2014). Microscopic Findings Histologic examination of the skin lesions shows a three-layered structure to the ulcer. Superficially, there is an ulcer bed containing abundant neutrophils, beneath which is a layer of granulation tissue. In the deepest part of the lesion is a chronic inflammatory infiltrate consisting primarily of lymphocytes and plasma cells. Gram or Giemsa stains may reveal the gram-negative organisms, which may be present in large numbers in pairs and in parallel chains in the more superficial portion of the lesion, but culture and PCR are more sensitive and specific methods of diagnosis. Clinical Course and Treatment Treatment with antibiotics is curative in immunocompetent individuals. The disease is currently much more frequently encountered in HIV-positive patients, however, in whom the lesions are more numerous, heal poorly, and may fail to respond adequately to treatment (Mohammed and Olumide 2008).

Tuberculosis Tuberculosis of the female genital tract is a common cause of pelvic inflammatory disease and infertility in some parts of the world but is very uncommon in most developed countries. It most commonly affects the fallopian tubes and endometrium (Manoj et al. 2008); vulvar involvement is exceedingly rare, present in less than 2% of tuberculosis cases with pelvic involvement (Shen et al. 2011; Manoj et al. 2008). It is usually the result of hematogenous spread from a primary pulmonary infection with Mycobacterium tuberculosis, which has often healed by the time the pelvic disease is detected, but autoinoculation is thought to be responsible in some cases.

11

Immunosuppression may play a role in susceptibility, as a case of vulvar tuberculosis has been described in a renal transplant patient (Wilkinson and Stone 2008), and hormones may influence development as well, as most cases present in the reproductive age range (Manoj et al. 2008) Clinical Features Vulvar tuberculosis may present as ulcerative lesions or swellings with multiple draining sinuses or as bulky exophytic lesions with associated lymphatic obstruction. Microscopic Findings Diagnosis usually can be made by biopsy of the involved tissues, which will reveal the characteristic caseating granulomas with Langhans giant cells. The mycobacteria can be identified on acid-fast stain, but this is far less reliable than isolation of the organism in culture. Clinical Course and Treatment Resection of lesions with a 6-month course of antitubercular drugs is curative (Manoj et al. 2008).

Miscellaneous Bacterial Infections Erythrasma is a superficial skin infection caused by Corynebacterium minutissimum, which presents as an asymptomatic macular pink-brown rash. Because it has a predilection for the skin folds, the vulvar area may be involved, but the diagnosis is generally made on clinical grounds and biopsy is rarely necessary. Erysipelas is a manifestation of infection of the skin with hemolytic streptocci or Staphylococcus aureus. It presents as a sharply marginated area of erythema, often associated with fever, malaise, chills, and nausea. The infection may progress to involve the subcutis, resulting in cellulitis, in which case the erythema is less well-defined and the involved area becomes edematous and painful. Skin infection with hemolytic streptococci or Staphylococcus aureus can also result in impetigo. On the vulva, impetigo is usually restricted to the hair-bearing skin, where it appears as small vesiculopustules which quickly rupture and develop a golden brown crust.

12

Rare cases of botryomycosis, an ulcerative infection of the skin caused by S. aureus, P. aeruginosa, E. coli, Streptococci, or Proteus species, have been reported to involve the vulva (Elas et al. 2014). This disease is best recognized by examination of the purulent drainage from the ulcer, in which the characteristic “granules” of bacteria may be identified. Necrotizing fasciitis may occur in the vulva and perineum in the skin damaged by recent surgical intervention or trauma. Most vulvar cases are caused by polymicrobial infection, including anaerobic species (Nakayama and Busse 2010). Predisposing factors include diabetes mellitus, immunosuppression, peripheral vascular disease, increased age, hypertension, obesity, and radiation exposure. Initially, necrotizing fasciitis may appear as mild cellulitis or edema with inflammation, frequently associated with severe pain out of proportion to the degree of apparent tissue damage. Fever may or may not be present. The disease typically progresses rapidly despite treatment with antibiotics and must be recognized quickly, as a delay in diagnosis without therapy carries a nearly 50% mortality rate (Stephenson et al. 1992). Prompt, aggressive surgical debridement radical excision of the infected tissue and broadspectrum systemic antibiotic therapy offers the only chance of cure, but even with appropriate treatment, mortality rates are reported to range from 20 to 40% (Sultan et al. 2012).

Viral Condyloma Acuminatum Condyloma acuminatum is an exophytic lesion of the skin, or, less commonly, the mucous membranes, caused by infection with low-risk subtypes of HPV, most commonly types 6 and 11. The frequency of vulvar condyloma acuminatum varies according to the population but is generally over 1%. Many risk factors for the development of vulvar condyloma have been identified. As a sexually transmitted disease, the risk of condyloma acuminatum is increased with increasing numbers of sexual partners. The risk is also increased in patients with HPV-related lesions elsewhere in the lower anogenital tract; up to 50% of women with

D. S. Rush and E. J. Wilkinson

vulvar condyloma acuminatum also have past, concurrent, or subsequent diagnoses of cervical or vaginal squamous intraepithelial lesions (Mittal et al. 2013). Other commonly associated conditions are vaginitis, pregnancy, diabetes, oral contraceptive use, and poor hygiene. Immunosuppression is an increasingly common predisposing condition, and women with HIV, organ transplants, or autoimmune diseases (Santana et al. 2011; Lyrio et al. 2013) often struggle with widespread lesions throughout the lower anogenital tract that can be very difficult to eradicate. When condyloma acuminatum is detected in a child, sexual abuse must be considered, but other modes of transmission have also been demonstrated. Vertical transmission of the virus, either in utero or intrapartum, is possible (Jayasinghe and Garland 2006), and when it occurs, the virus may remain dormant for years before visible lesion develops (Hornor 2004). Many pediatric cases have been shown to contain types 1 and 2, subtypes more common in common skin warts, and frequently the patient or caregiver is found to have common warts (Allen and Siegfried 1998; Stefanaki et al. 2012), suggesting transmission by autoinnoculation and or nonsexual contact. There is some evidence that the disease may also be transmissible by fomites (Jayasinghe and Garland 2006). Early data on the effects of HPV vaccination strongly suggests that condylomata acuminata may become increasingly less common. In Denmark and Australia, where robust vaccination programs have reached 70–85% of the target population, marked reductions in the diagnosis have already been reported (Ali et al. 2013; Baandrup et al. 2013; Read et al. 2011), and recent data suggests rates are dropping in the USA as well (Flagg et al. 2013). Clinical Features The lesions are usually asymptomatic and frequently multiple and multifocal. They often come to clinical attention because of considerable enlargement during pregnancy (Garland et al. 2009; Hoy et al. 2009). Any area of the vulva, as well as the vagina, cervix, urethra, and anal canal, may be affected. The lesions may be red, white, gray, or brown and vary in size from millimeters to several centimeters (Fig. 9). Application of a dilute

1

Benign Diseases of the Vulva

Fig. 9 Perianal condylomata. The exophytic lesions have a corrugated, irregular surface. (Photograph courtesy of Dr. Jaqueline Castagno, University of Florida)

Fig. 10 Condyloma acuminatum. This low-power magnification shows the complex papillary architecture, with central fibrovascular stalks lined by thickened epithelium

solution of acetic acid will impart a white appearance to the abnormal epithelium, which may aid in identification. On occasion, multiple lesions may coalesce into large plaques, an occurrence more common in diabetic or immunosuppressed patients. Microscopic Findings On histologic examination, the lesion consists of complex branching fibrovascular cores covered with acanthotic squamous epithelium, frequently with accompanying hyperkeratosis and parakeratosis (Fig. 10). The pathognomonic feature is koilocytic atypia, characterized by enlarged, hyperchromatic nuclei with irregular, wrinkled nuclear membranes accompanied by a region of perinuclear clearing or “halo,” usually located in

13

Fig. 11 Condyloma acuminatum. At higher power magnification, the characteristic koilocytic change, with perinuclear clearing surrounding enlarged, irregularly shaped hyperchromatic nuclei, is seen in the superficial layers of the epithelium

the upper third of the epithelium (Fig. 11). These cellular changes are the morphologic manifestation of HPV production in the infected cells but may be minimal to absent in some cases (Medeiros et al. 2005). Basal and parabasal hyperplasia, with increased mitotic figures confined to the lower third of the epithelium, and cytoplasmic maturation beginning to appear in the middle third and relatively normal maturation of the upper third are characteristic, and prominent intracellular bridges may be noted. Dyskeratotic cells may be seen, and a subepithelial chronic inflammatory infiltrate is commonly identified. Differential Diagnosis Other benign exophytic lesions of the vulva, such as fibroepithelial polyp, vestibular papilloma, and seborrheic keratosis, will lack the basal hyperplasia and koilocytic atypia typical of condyloma. In cases where the morphology is not sufficiently distinctive, immunohistochemical staining for Ki-67 may also be of further assistance in differentiating these lesions from condyloma. Because HPV infection activates the cell cycle in order to accomplish viral reproduction, a process which occurs in the maturing squamous cells, Ki-67 will be reactive in some cells in the upper levels of the epithelium in condylomata/low-grade squamous intraepithelial lesions (LSIL) while in normally epithelium and other benign squamous lesions, expression of Ki-67 is limited to the basal and parabasal cells.

14

D. S. Rush and E. J. Wilkinson

higher viral load (Koo et al. 2016) and lower rates in patients treated with surgical excision.

Although high-grade squamous intraepithelial lesions (HSIL) and squamous cell carcinomas with warty morphology may show marked koilocytic atypia in the superficial cells, they are distinguished from condylomata by greater immaturity of the epithelium, the presence of abundant, frequently atypical, mitotic figures, particularly if present in the upper layers of the epithelium, and, in the case of carcinoma, the presence of invasion into the underlying tissue. It should be remembered that, although uncommon, lesions with the typical morphologic features of condyloma can contain areas of associated HSIL. Such lesions are almost exclusively seen in immunosuppressed patients (Maniar et al. 2013), and for this reason, it is advised that even lesions which appear to be benign condylomata in such patients ought to be biopsied to ensure the absence of a high-grade component. Even when these lesions are biopsied, the extensive condylomatous component may be so distant from other abnormal areas that the high-grade component may be missed on sampling. The same may occur on a microscopic level when a small focus of adjacent high-grade lesion is missed when the slides are examined due to the relative abundance of condylomatous lesion. It is speculated that lesions like these may be responsible for previously reported cases of condyloma containing high-risk HPV.

Herpesvirus Genital infection with HSV was once almost exclusively caused by HSV type 2, and HSV type 1 was limited to oral lesions. Today, probably due to changing patterns of sexual behavior, genital infection with HSV type 1 is increasingly common, especially in younger cohorts, comprising approximately half of new cases in some developed countries (Gupta et al. 2007), although vulvar infection with HSV type 2 is still approximately six times more common than with HSV type 1. Primary genital infection with HSV type 1 is more frequent in women, and more often symptomatic (Fatahzadeh and Schwartz 2007), but also less likely to recur. Antibodies to one type of HSV provide some protection against the other, leading to a decreased frequency of infection with a second type, and may result in decreased severity and duration of the newly acquired infection (Fatahzadeh and Schwartz 2007). Although approximately 20% of the US population has been infected by HSV 2 by age 40, and up to 85% have been infected by HSV 1 by age 60, the frequency of vulvar involvement is unknown, and the majority of infections appear to be subclinical (Fatahzadeh and Schwartz 2007; Maccato and Kauffman 1992; Nettina 1998).

Clinical Course and Treatment: Condylomata acuminata may regress spontaneously but usually persist and may increase in size or number over time. They are not considered premalignant and do not progress to HSILs or carcinoma. Topical application of dilute podophyllin, imiquimod, concentrated halogenated acetic acid (trichloroacetic acid), or sinecatechins (Lacey et al. 2013) can be used for the treatment of small vulvar condylomata. Response to therapy may be decreased in immunosuppressed patients and patients with concurrent cervical HPV infection (Koo et al. 2016). Larger lesions and those refractory to topical treatments may be removed or eradicated by electrosurgery, cryosurgery, laser ablation, or surgical excision (Lacey et al. 2013). The overall recurrence rate is reported as 20–30% (Lacey et al. 2013), with higher rates reported in patients with a

Clinical Features The disease typically presents 4–7 days after exposure, with a prodrome of constitutional symptoms including fever, headache, muscle aches, and a painful, erythematous swelling of the vulva, followed by the eruption of clusters of papules and vesicles which evolve into exquisitely painful ulcers (Fig. 12). Lesions may involve the anus, urethra, bladder, cervix, and vagina, as well as the vulva, and may be accompanied by dysuria and vaginal discharge. Concurrent HIV infection may drastically alter the presentation of HSV-related disease. HIV-positive patients tend to have more severe and more frequent outbreaks, and to develop more extensive lesions which may take longer to resolve (Domfeh et al. 2012; Fatahzadeh and Schwartz 2007). They may also

1

Benign Diseases of the Vulva

Fig. 12 Herpes simplex infection of the vulva. Numerous small, shallow ulcers on an erythematous base are visible on the inferior labia majora, labia minora, and perineum

develop lesions with atypical symptomatology or morphology, including painless lesions, fissures, patchy erythema, furuncles (boils), wart-like lesions, or diffusely and deeply ulcerating lesions (Domfeh et al. 2012; Fatahzadeh and Schwartz 2007). Of most concern to the pathologist may be the hypertrophic, mass-forming lesions sometimes referred to as hypertrophic chronic vegetative lesion (HCVL). Clinically, these lesions present as exophytic masses, measuring several centimeters in greatest dimension (Fig. 13), which may mimic warts or squamous cell carcinoma clinically (Domfeh et al. 2012; Mosunjac et al. 2009; Gomes do Amaral et al. 2009; Strehl et al. 2012; Tangjitgamol et al. 2013). Microscopic Findings A sample from the margin of the ulcer is most likely to reveal the pathognomonic findings. Infected cells show homogenization of the nuclear chromatin resulting in a “ground glass” appearance, which then progresses to the more typical eosinophilic intranuclear inclusion body. Multinucleation is also a characteristic feature (Fig. 14). Cytologic evaluation of the scraping of the base and edges (Tzank preparation) of a fresh ulcer, or freshly opened vesicle, may also

15

Fig. 13 Hypertrophic chronic vegetative lesion (HCVL) and atypical manifestation of herpes infection. The lesion is present on the medial surface of the labium minus and shows minute surface ulcerations. (Reprinted with permission from Selim et al. 2015)

Fig. 14 The cellular changes of herpes simplex infection. Several multinucleated forms showing the characteristic homogenized glassy cytoplasm are present at the edge of this herpetic ulcer

show the characteristic cellular changes. Over time, the infected cells undergo karyorrhexis and lysis, and samples taken in the late ulcerative phase may not, therefore, always show the intranuclear inclusions.

16

D. S. Rush and E. J. Wilkinson

Fig. 15 Microscopic features of the hypertrophic chronic vegetative herpes lesion illustrated in Fig. 13. There is pseudoepitheliomatous hyperplasia (a) overlying a dense lymphoplasmacytic infiltrate (b). The cytopathologic

changes of HSV are seen focally (c), and immunohistochemical stain for HSV confirms the diagnosis (d). (Reprinted with permission from Selim et al. 2015)

The histologic features of the atypical lesions seen in HIV-positive patients are significantly different than those of the typical lesion. The large exophytic HCVL show a markedly thickened, hyperkeratotic epithelium, overlying a fibrotic,

thickened dermis, with a dense inflammatory infiltrate of plasma cells and lymphocytes. Only small areas of ulceration, in which the characteristic viral changes may be identified, are present in such lesions (Fig. 15).

1

Benign Diseases of the Vulva

17

Morphologic changes seen with HSV infection are not reliable in separating primary from secondary infection or in distinguishing HSV type 1 from type 2 infection, nor can they differentiate the lesions of herpes zoster, which may involve the vulva as well, though only rarely, but immunohistochemical stains can be used to distinguish the type of virus present in tissue, cytologic preparations, or cultures when necessary. Viral culture has a relatively low sensitivity, however, and only about 80% of primary infections and 25–50% of recurrent infections can be identified this way (Gupta et al. 2007). PCR is the preferred method to identify the virus, as it is more sensitive and faster than culture (Fatahzadeh and Schwartz 2007; Gupta et al. 2007; Hope-Rapp et al. 2010).

formations, made up of 50–100 individual clustered lesions, also have been described. In children, the lesions may develop anywhere in the skin, and the route of transmission is from close contact. Genital involvement is unusual in children (Zhuang et al. 2015). In adults, however, the genitals are usually the only site involved and the disease is almost exclusively transmitted by sexual contact (Bast-Juzbasic and Ceovic 2014). On the vulva, the keratinized surfaces of the labia majora, labia minora, and mons are most frequently affected. Lesions are usually asymptomatic, but may be pruritic, and excoriation from excessive scratching may facilitate secondary bacterial infection, which may mask the underlying condition and confound the diagnosis.

Clinical Course and Treatment Untreated, the ulcers of the initial episode heal in approximately 2–6 weeks, after which the virus lies dormant in regional sensory and autonomic ganglia. A 7–10-day course of systemic treatment with the antiviral agents acyclovir, valacyclovir, or famcyclovir can speed healing, decrease viral shedding, and decrease the incidence of new lesions, but these drugs do not prevent or eradicate latent infection, and they are not curative. Periodic reactivation of the virus is likely, leading to subsequent recurrences, the rate of which decreases with time since the primary infection (Gupta et al. 2007). Atypical lesions associated with concurrent HIV infection may not respond to conventional treatment; these patients may require higher doses of antiviral agents and longer durations of treatment, and some may require the use of alternative antiviral agents. Surgery may also be considered for large atypical lesions refractory to therapy.

Microscopic Findings Clinical diagnosis usually does not require biopsy, but when biopsy is performed, the histomorphology is distinctive (Fig. 16). The central dimple of the lesion can be seen histologically if the lesion is carefully bisected. Within the dermis, there often is a marked vascular response with endothelial proliferation and perivascular inflammation. In recent infections, the lesions demonstrate marked acanthosis and the characteristic eosinophilic intracytoplasmic viral inclusions (Henderson–Patterson bodies), which may also be identified in scrapings from the interior of the lesions.

Molluscum Contagiosum Clinical Features Molluscum contagiosum is a viral infection of the skin which manifests after an incubation period of 14–50 days as small, smooth papules (3–6 mm in diameter) with a central punctum or umbilication. They generally are multiple and separate, although they may be single. Rare plaque

Fig. 16 Molluscum contagiosum. The center of the lesion shows the characteristic eosinophilic viral inclusions (Henderson–Patterson bodies)

18

Clinical Course and Treatment The lesions are infectious as long as they are present, and although most lesions of molluscum contagiosum regress spontaneously within months to years, many patients are anxious to be rid of them sooner. Numerous treatment options are available to speed resolution, including curettage, cryosurgery, and topical agents. Responses are variable, and more than one treatment modality may be needed to successfully eradicate the disease.

Varicella (Herpes Zoster) Vulvar shingles, caused by the involvement of the vulva by varicella, the etiologic agent of chicken pox, is rare. The lesions represent reactivation of virus which has been dormant in the sacral ganglia. The prodromal vulvar pain, without apparent physical findings, may at first simulate vestibulitis, but the subsequent eruption of vesicles and ulcers soon distinguishes it. Patients are usually postmenopausal and/or immunosuppressed and the vesicles are characteristically unilateral. The histologic and cytologic findings are indistinguishable from HSV, but immunohistochemistry with virus-specific antibody or PCR can differentiate them when necessary. Cytomegalovirus Cytomegalovirus (CMV) is a rare cause of ulcerative cervicitis and vulvovaginitis which may mimic HSV infection clinically (Abou and Dallenbach 2013). The histopathologic findings are similar, except that the cytologic changes caused by CMV are both intranuclear and cytoplasmic, multinucleated forms do not occur, and the viral inclusions also may be seen involving vascular endothelial cells, as well as the epithelial cells. The diagnosis can be confirmed in tissue by immunohistochemical staining using specific antibodies to CMV, PCR of swabs collected from active lesions, or by isolation in culture. Epstein–Barr Virus Primary infection with Epstein–Barr virus (EBV) is occasionally the cause of ulcerations of the labia minora which may or may not be accompanied by

D. S. Rush and E. J. Wilkinson

the systemic symptoms characteristic of infectious mononucleosis. The ulcers are usually more than a centimeter in diameter (Halvorsen et al. 2006), necrotic, deep, and painful. The median age of affected patients is 14.5 years old, and in most patients there is no history of recent sexual activity (Halvorsen et al. 2006). Involvement of the vulva is presumed to occur via hematogenous spread, but sexual transmission cannot be completely ruled out in all cases. The ulcers appear very early in the disease, usually before serology can detect the infection, and heal spontaneously in 3–4 weeks (Halvorsen et al. 2006; Sand and Thomsen 2017; Taylor et al. 1998). Viral culture or PCR may identify the virus (Halvorsen et al. 2006; Sand and Thomsen 2017). Histological findings are nonspecific, and the lesion is extremely unlikely to be encountered as a biopsy specimen.

Fungal Fungal infection of the vulva may cause chronic inflammatory conditions of the vulvar and perianal skin. The most common agents are Candida species. Recurrent or chronic vulvar candidiasis may cause the involved mucosa to become atrophic and painful, with a glazed red appearance. Painful fissuring in the periclitoral area, interlabial sulcus, and around the introitus may develop and may be a clue to the diagnosis (Margesson, 2006). Dermatophytes are less common, and often of zoonotic origin, with Microsporum canis and Trichophyton mentagrophytes the most commonly encountered (Sand and Thomsen 2017). Dermatophytosis usually involves the hairbearing skin, producing follicular papulopustular lesions which may coalesce into elevated erythematous scaly plaques. Pityriasis versicolor (tinea versicolor), caused by Malassezia species, has also been reported on the vulva (Day and Scurry 2014). Fungal infections of the vulva rarely require biopsy, and accurate diagnosis generally can be accomplished by microscopic examination of skin scrapings placed in 10% potassium hydroxide or by appropriate culture methods. When biopsy is performed, the presence of neutrophils in the epidermis may suggest a fungal

1

Benign Diseases of the Vulva

19

infestation of the larval form of the muscoid fly and Sarcophaga, has also been reported and can be diagnosed by recognition of the larva extracted from the vulvar tissues (Cilla et al. 1992; Koranantakul et al. 1991).

Inflammatory Dermatoses

Fig. 17 Fungal infection of the vulvar skin. Numerous intraepithelial neutrophils are present, which can be a useful clue to a fungal etiology. The differential diagnosis includes psoriasis, which can look very similar (see Fig. 18a, b), and definitive diagnosis requires identification of the fungal organisms on special stains

etiology (Fig. 17), and performance of silver or periodic acid–Schiff (PAS) stain may reveal the organisms. Topical antifungal creams are the usual therapy.

Parasitic Pediculosis pubis, or pubic lice, caused by infestation of Pthirus pubis, is not uncommon, but rarely generates a specimen for histologic examination, as the nits, nymphs, and adult lice may be visualized with the naked eye or with the assistance of a magnifying glass. Other parasitic infections of the vulva are quite rare. Children infected with Enterobius vermicularis (pinworm) frequently experience severe vulvovaginal pruritus, which may awaken them at night, thought to be related to migrating worms. Examination of the vulvar vestibule and vagina in such cases reveals marked inflammation, but only rarely is the parasite identified in the vulvar tissue. A granuloma secondary to Enterobius eggs has been reported involving the vulva (Sun et al. 1991). Skin lesions on the vulva from penetration of the infective cercariae of schistosomiasis, usually Schistosoma mansoni, may be encountered in endemic areas, in which the parasite may be found within the epidermis on biopsied samples. Cutaneous myiasis of the vulva, secondary to

The inflammatory dermatoses are among the most common dermatologic disorders, and vulvar involvement is frequent. The recognition of these disorders, both clinically and histologically, is significantly more difficult on the vulvar skin than in extragenital sites, due in large part to the particular local conditions. The vulvar skin is confined in an occlusive environment, in which it is subject to high levels of moisture, friction, and other irritations, which frequently combine to induce an aberrant appearance of inflammatory dermatoses in this region. Features which are characteristic on nongenital skin may be obscured or even absent. At the same time, reactive changes and superinfection are much more likely to be superimposed on vulvar lesions, creating a complex tangle of symptoms and findings which can be difficult to tease apart. Biopsies of inflammatory dermatoses of the vulva, consequently, are frequently inconclusive, and arriving at a specific diagnosis often requires careful clinicopathologic correlation as well as observation over time. The job of the pathologist in these cases is not so much to establish an unequivocal diagnosis as to narrow the differential. Recognition of specific histologic patterns has long been used by dermatopathologists as the first step in classification of dermatologic conditions, leading the International Society for the Study of Vulvovaginal Diseases (ISSVD) to establish its first classification scheme based for vulvar dermatoses based on histologic pattern (Lynch et al. 2006). The original 2006 ISSVD classification of the inflammatory dermatoses is simplified to include only those disorders most likely to be encountered on vulvar biopsies, and is the basis for Table 2, which also includes some less common disorders worthy of consideration. Following this framework, the disorders are presented below in groups classified by the dominant histologic pattern.

20

D. S. Rush and E. J. Wilkinson

Table 2 A summary of vulvar dermatoses classified by histologic pattern. The conditions on the left are those included in the 2006 ISSVD classification. On the right are additional conditions presented in this chapter Less common entities (not included in the 2006 ISSVD classification)

2006 ISSVD classification of vulvar dermatosesa Spongiotic pattern Atopic dermatitis Allergic/irritant contact dermatitis Acanthotic pattern Psoriasis LSC (primary or secondary) Lichenoid pattern LS Fixed drug eruption LP Dermal homogenization/sclerosis pattern LS Morphea Vesiculobullous pattern Pemphigoid, cicatricial Bullous pemphigoid type (mucous membrane Pemphigoid gestationis pemphigoid) Pemphigus Linear IgA disease Bullous systemic lupus erythematosus (SLE) Acanthotic pattern Hailey–Hailey disease Darier’s disease Papular genitocrural acantholysis (papular acantholytic dyskeratosis) Granulomatous pattern Crohn’s disease Sarcoid Melkersson–Rosenthal disease (granulomatous vulvitis) Vasculopathic pattern Aphthous ulcers Behçet disease Plasma cell vulvitis a

Adapted from Lynch et al. 2006

Spongiotic Pattern Epithelial spongiosis is the result of intraepithelial edema. Clinically, spongiotic lesions present as eczematous dermatitides, with a wet, oozing surface. On the vulva, because it is usually closely covered, trapping of the moisture may result in maceration of the surface, which may disguise the underlying condition. Microscopically, spongiosis is evidenced by increased space between epithelial cells, representing the area of fluid accumulation.

Table 3 Common vulvar allergens and irritants. Many substances can act as either an allergen or an irritant, depending on the sensitivity of the patient Common vulvar allergens and irritants Fragrances Topical anesthetics Preservatives Topical antifungals and other antibiotics Emollients Metals (nickel, gold) Body fluids Soaps and detergents Lubricants Excessive heat

Allergic/Irritant Contact Dermatitis Clinical Features Contact dermatitis is the most commonly encountered spongiotic dermatitis of the vulva, affecting approximately 15–54% of women (Ball et al. 2015; Connor and Eppsteiner 2014). Vulvar contact dermatitis may be of an allergic type (allergic contact dermatitis, ACD) or an irritant type (irritant contact dermatitis, ICD), with the latter being the more common of the two. The allergic type is a cellmediated response to sensitizing agents, which can include a number of soaps, topical medications and remedies, or components thereof (Moyal-Barracco and Wendling 2014; O’Gorman and Torgerson 2013; Foote et al. 2013; Bauer et al. 2005). The irritant type is due not only to the presence of an irritant, but also to underlying skin damage and subsequent loss of barrier function, as may be seen in urinary incontinence, and occasionally in association with sanitary napkin use during menstruation (Wakashin 2007). A list of selected common potential allergens and irritants is presented in Table 3. The presentation of ICD and ACD is variable, depending on the severity and duration of the process. Acute ICD develops within minutes to hours of the exposure, while those of ACD take 24–48 h to develop. The lesions of ICD tend to be wellcircumscribed, confined to the area of contact, more likely to be painful, and less likely to develop vesicles and bullae, while those of ACD are more poorly demarcated, more likely to be pruritic, and more likely to develop vesicles and bullae. Superficial erosion or ulceration may be present in both.

1

Benign Diseases of the Vulva

Not infrequently, the clinical appearance is normal or only minimally altered (Ball et al. 2015), despite significant symptomatology.

Microscopic Findings The pathologic findings are also variable and depend on the age of the lesion. Spongiosis may be minimal at first, progressing to pronounced dermal edema with the formation of microvesicles, and regressing again with more time. In long-standing contact dermatitis, LSC (see section “Lichen Simplex Chronicus”) often supervenes, with prominent acanthosis and hyperkeratosis. Clinical Course and Treatment The symptoms of ICD and ACD will continue until contact with the offending agent can be eliminated, a process which may take some time to achieve. All possible exposures must be eliminated by the patient, which can necessitate quite extensive changes and can be quite disruptive, demanding the elimination of every soap, shampoo, laundry product, lotion, lubricant, topical ointments and creams, and even many items of clothing, among other things, until each can be exonerated by the result of patch testing or by slow reintroduction one at a time. A variety of therapies may be used for symptomatic control until the allergen or irritant can be identified. Vaseline or zinc oxide may be used to form a barrier to potential exposures, oral over-the-counter antihistamines may be used to control itching, and nonsteroidal anti-inflammatory agents may be used to control pain. Sitz baths and cold compresses may also provide relief. Unresponsive pain may be treated with tricyclic antidepressants and anticonvulsants. Finally, topical corticosteroids, or in severe cases localized injection or systemic steroid administration, are also usually part of the management, keeping in mind that in rare cases topical corticosteroids may themselves be the cause of the allergen or irritation.

Atopic Dermatitis Clinical Features With 85% of patients presenting before age 5, and the majority of cases remitting by adolescence, atopic dermatitis is predominantly a disease of

21

childhood. The disease first manifests in adulthood in only 2–8% of patients (Arkwright et al. 2013). The disorder may involve the vulvar skin, but very few cases have been reported in the literature to date, and the frequency of vulvar involvement is unknown. The physical findings may be limited to dryness and scaling, but thickening of the skin with localized excoriation may be evident if the vulva has been irritated by scratching.

Microscopic Findings Because the diagnosis is usually established by clinical findings and the appearance on the nongenital skin, vulvar biopsies are rarely performed on these patients. When they are, the pathologic findings are usually nonspecific. Spongiosis may be present. Within the dermis, lymphocytes and macrophages are present, and the density of the infiltrate tends to correlate with the severity and chronicity of the process. Eosinophils and mast cells also may be identified. Most often, superimposed lichen simplex chronicus (LSC) (see section “Lichen Simplex Chronicus”) has developed in response to the chronic itching and scratching, masking the initiating condition. Clinical Course and Treatment Although 70% of children with atopic dermatitis show spontaneous remission before they reach adulthood, a minority of patients suffer lifelong disease with periodic exacerbations. There is no cure for the disease, but most patients’ symptoms can be controlled using emollients and topical corticosteroids. In patients who do not show an adequate response, topical calcineurin inhibitors may be used in place of steroids. Rarely, severe, refractory disease requires systemic treatment with calcineurin inhibitors such as cyclosporine. Differential Diagnosis of Spongiotic Dermatitis of the Vulva Histology cannot always distinguish between ICD, ACD, and atopic dermatitis, but certain features favor a specific diagnosis. The presence of balloon cell change and dyskeratosis favors ICD, with the caveat that if such changes are severe fixed drug eruption (see section “Fixed Drug Eruption”) and erythema multiforme (see section “Erythema Multiforme/Stevens–Johnson Syndrome”) must

22

D. S. Rush and E. J. Wilkinson

Table 4 Select features of use in the differential diagnosis of common spongiotic dermatoses of the vulva Differential diagnosis of spongiotic dermatoses of the vulva Diagnosis Pain Pruritis Demarcation of lesions

a

ACD

Rare

Common

Poor

Presence of balloon cell change +/ dyskeratosisa No

ICD Atopy

Common Rare

Rare Invariably

Sharp Poor

Sometimes No

Formation of vesicles+/ bullaeb

Concurrent involvement of nongenital skin No

Sometimes (containing eosinophils and Langhans cells) No No No Yes

Rule out erythema multiforme Rule out vesiculobullous disease

b

be considered as well. Prominent vesicles containing eosinophils and aggregates of intraepidermal Langhan cells favor ACD (Moyal-Barracco and Wendling 2014; Hoang et al. 2014), although the former may also be seen early in the development of some vesiculobullous disorders, in which case immunofluorescent studies may be used to make the distinction. Spongiosis may be seen as a component of reactions to fungal infection, some drug reactions, arthropod bites, as well as in contact and atopic dermatitis. A careful search for fungal organisms, with the use of special stains such as PAS or silver stains if necessary, will easily establish the presence or absence of infection, and clinical history will help determine whether drug reactions or arthropod bites need to be considered. Spongiosis may also be seen in acantholytic disorders, but as acantholysis is not a feature of atopic or contact dermatitis, this finding will point to a different set of diseases altogether (see section “Acantholytic Pattern”). The differential diagnosis of spongiotic dermatoses is summarized in Table 4.

Acanthotic Pattern Acanthosis refers to thickening of the epithelium, which presents clinically as thick white plaques. Histologically, two patterns of acanthosis may be recognized. In “regular acanthosis,” sometimes

also called “psoriasiform hyperplasia,” the rete ridges are uniformly elongated, with all of them having the same length and width. In “irregular” acanthosis,” the rete ridges are variable in length and width from one to the next. Whether regular or irregular, acanthosis takes time to develop and is therefore an indication of a chronic, rather than an acute, condition.

Psoriasis Clinical Features Psoriasis is a chronic immune-mediated disease which affects approximately 3.2% of the population of the USA (Young et al. 2017) and involves the genitalia in 30–40% of patients (Andreassi and Bilenchi 2014). The median age of onset in women is 25 years (Young et al. 2017), and the severity of disease often fluctuates with hormone levels, with exacerbations developing in puberty, postpartum, and menopause. Vulvar psoriasis may present in the classic form, most commonly on the mons, with sharply demarcated erythematous papules and plaques covered with silvery scale, which show punctate points of bleeding when the superficial scale is removed (Auspitz sign). More often, however, on other vulvar sites, is the inverse form, in which the scale is absent, and the lesions appear as flat red patches which may also be eroded or ulcerated. Symptoms of vulvar psoriasis may include itching, burning, and pain.

1

Benign Diseases of the Vulva

23

Fig. 18 (a) Classic psoriasis. The epidermis shows regular acanthosis, thinning of the suprapapillary plates, loss of the granular layer, and confluent surface parakeratosis with superficial collections of neutrophils (Munro

microsabscesses). (b) Inverse psoriasis. This lesion shows the same features as classic psoriasis but with more even dispersion of neutrophils and a thicker parakeratotic surface

Microscopic Findings In the classic form, well-developed lesions show the typical diagnostic features of confluent superficial parakeratosis with intraepidermal collections of neutrophils (Munro microabscesses), epidermal hyperplasia, loss of the granular layer, spongiotic pustules, and thinning of the suprapapillary plates (Fig. 18a). In the inverse form (Fig. 18b), diagnostic features may be less prominent or even lacking on biopsy (Andreassi and Bilenchi 2014; Moyal-Barracco and Wendling 2014).

Lichen Simplex Chronicus

Clinical Course and Treatment A variety of treatments are available for psoriasis, but none is curative. Mild disease can be treated with a wide variety of topical agents, while more severe disease requires systemic immunosuppressive therapy with agents such as methotrexate or cyclosporine. Phototherapy is another effective adjunct or alternative to topical agents, but its use on the vulva, where the anatomy may make it difficult, is limited. Newer agents which target directly the cytokines responsible for the aberrant immune response are also available for use in moderate to severe disease, and have significantly improved outcomes.

Clinical Features LSC is a reactive pattern induced by persistent rubbing or scratching in response to local itching. It may occur without a diagnosed cause of the initial itch (primary LS) or as the end point of many chronic pruritic conditions such as atopic or contact dermatitis, chronic fungal infections, LS, or LP. It most commonly involves the hair-bearing skin of the labia majora, mons pubis, and perianal region and may be confined to a focal area, appearing gray-white or reddened on clinical exam (Fig. 19). The skin markings are often accentuated, a sign of intradermal edema and chronic rubbing; excoriation and fissures are common. Microscopic Findings In addition to prominent acanthosis of the irregular type, the histopathologic features may include hyperkeratosis and hypergranulosis. Fibrosis of the papillary dermis, characterized by vertical orientation of the collagen fibers, is characteristic, and the capillaries in the papillary dermis also show a vertical orientation (Fig. 20). This orientation of the collagen fibers and capillaries of the papillary dermis is distinctive and a useful clue to the diagnosis (Ball et al. 2015). A chronic inflammatory infiltrate in the superficial dermis is usually present, but the

24

Fig. 19 The gross appearance of LSC. The left labium majus shows a thickened, grayish surface. (Reprinted with permission from Ball et al. 2015)

presence of eosinophils and neutrophils is not to be expected. Parakeratosis may be present in LSC but is generally not a prominent feature. These changes are nonspecific, and the diagnosis is usually one of exclusion.

Clinical Course and Treatment Paradoxically, and much to the frustration of those afflicted, scratching the lesions of LSC only worsens the itch, setting up a vicious cycle which can be very difficult to control. Symptoms can be managed with topical steroids, calcineurin inhibitors, and, of course, avoidance of scratching, but recurrences are frequent. Differential Diagnosis of Acanthotic Dermatoses of the Vulva Clinical history will usually confirm a diagnosis of psoriasis, as most patients will have concurrent or previously diagnosed characteristic lesions elsewhere on the skin. In the absence of a clear history, or even when the patient has an

D. S. Rush and E. J. Wilkinson

Fig. 20 LSC. The epidermis is thickened and shows focal hypergranulosis. The vertical orientation of the collagen fibers and capillaries in the papillary dermis seen here is a characteristic finding. (Reprinted with permission from Ball et al. 2015)

established diagnosis of psoriasis, the prominent neutrophilic infiltrate seen in the disease is similar to that seen with fungal infection, and if other histologic features are at all equivocal, it is advisable to perform special stains to rule out fungal infection prior to making a diagnosis of vulvar psoriasis. The same should be done for cases of LSC showing significant inflammation, as chronic fungal infections are a common underlying cause of LSC. Although both psoriasis and LSC are characterized by pronounced acanthosis, they are usually easily distinguished from each other. Again, a clinical history of psoriasis can be extremely helpful, but histologic features are relatively reliable in this distinction as well. The pattern of acanthosis is regular in psoriasis but irregular in LSC, and the granular layer is prominent in LSC but lost in psoriasis. In addition, thinning of the suprapapillary dermis and neutrophilic microabscesses, prominent features in psoriasis, are not present in LSC.

1

Benign Diseases of the Vulva

25

Table 5 Selected features of use in differentiating common acanthotic dermatoses of the vulva Diagnosis

Feature Acanthosis pattern Granular layer Concurrent involvement of extragenital skin Spongiosis Eosinophils present Lichenoid infiltrate Basal vacuolar change Necrotic keratinocytes

Psoriasis Regular Lost Yes

LSC (primary) Irregular Prominent No

LSC secondary to contact dermatitis or atopy Irregular Prominent No

No No No No None

No No No No Rare

Yes Yes No No Rare

A more difficult problem is to determine whether LSC is of the primary or secondary type. Chronic contact dermatitis commonly evolves to LSC (see section “Allergic/Irritant Contact Dermatitis”), which may be suggested by the presence of associated spongiosis or an inflammatory infiltrate containing eosinophils. When LSC is superimposed on LS (see section “Lichen Sclerosus”), a lichenoid infiltrate or vacuolar changes at the dermo-epidermal junction may suggest the underlying diagnosis, as may the identification of hyalinization of the superficial dermis. The presence of necrotic keratinocytes, which may be numerous in LS but are scarce, if present at all, in LSC (Weyers 2015), may also be helpful in this differential. Acanthosis, hyperkeratosis, and hypergranulosis are common features in intraepithelial neoplasias of the vulva as well, but the presence of cytologic atypia, increased mitotic activity with atypical forms, and abnormal maturation patterns should distinguish them. The differential diagnosis of acanthotic dermatoses of the vulva is summarized in Table 5.

Lichenoid Pattern The lichenoid pattern of dermatitis is characterized by a band-like infiltrate confined to the papillary dermis and basal epidermis which obscures the dermo-epidermal junction and causes focal

LSC secondary to LS Irregular Prominent No No No Yes Yes Common

necrosis and vacuolization of the basal keratinocytes. The lesions in this category of vulvar dermatoses may be exceptionally difficult to differentiate from one another on histologic grounds alone.

Lichen Sclerosus LS is more common in the premenarchal and postmenopausal years and is diagnosed in an estimated 1–2% of patients in general gynecologic practice (Goldstein et al. 2005). It is currently considered to be an autoimmune condition, occurring in genetically predisposed patients (Fistarol and Itin 2013; Sherman et al. 2010). On the vulva, it involves the non-hair-bearing portions, predominantly affecting the keratinizing epithelium of the medial labia majora, interlabial sulci, labia minora, clitoris, perineum, and perianal area. The disease often begins around the clitoral hood (Fistarol and Itin 2013). Involvement may be limited to a small, single area, or may involve all of these regions in a figure-eightshaped distribution surrounding the introitus and anus. Early lesions may appear as white papules, which typically evolve into white plaques, or as sharply demarcated and slightly elevated erythematous patches with edema. Fissuring may be seen, especially between the clitoris and urethra, in the interlabial sulci and on the perineum

26

D. S. Rush and E. J. Wilkinson

Fig. 22 LS with minimal sclerosis. The epidermis shows pronounced hyperparakeratosis, and focal dyskeratotic cells can be seen. There is a dense lymphocytic infiltrate in the dermis, as well as collagen fibrosis, with linear formations of lymphocytes between collagen fibers Fig. 21 The gross appearance of LS. The skin is thin, shiny, pale, and wrinkled (parchment-like), with focal ecchymosis. There is loss of distinction between the labia majora and minora

posterior to the fourchette (Fistarol and Itin 2013) as well as erosions and ecchymoses. As the lesions age, they typically become dry, hypopigmented, and finely wrinkled, an appearance likened to parchment, cellophane, or tissue paper (Fig. 21). The disease does not proceed at the same rate in all areas, so lesions of many different stages may be present in the same patient at the same time. The most common symptom in patients with intact epithelium is pruritus. Patients with associated erosions or fissures may also experience pain, dysuria, and dyspareunia. Only 9% of patients are asymptomatic (Sherman et al. 2010).

Microscopic Findings LS was formerly known as LS et atrophicus, because in the classic, well-developed lesions, there is striking sclerosis in the papillary dermis and atrophy of the overlying epithelium. It has since become apparent that the microscopic findings of LS are extremely variable, and the classic atrophic, sclerotic appearance, while the easiest pattern to diagnose, is not always present. This accounts for the unique placement of LS into two categories the ISSVD classification of vulvar dermatoses (see Table 2).

Cases of LS with diminished or absent subepithelial sclerosis combined with epithelial hyperplasia rather than atrophy are not uncommon (Fig. 22), and they are significantly more difficult to distinguish from other conditions. It has long been maintained that such lesions, which show a prominent lichenoid infiltrate in the dermo-epidermal junction, with vacuolar changes and intraepidermal lymphocytes in the basal layer and little to no stromal homogenization (lichen sclerosus sine sclerosis), represented early lesions, which would progress in time to more recognizable sclerotic and atrophic lesions. Yet not all cases lacking stromal homogenization necessarily represent early cases, as recent studies have found these features in lesions of long standing (Weyers 2015). Features identified in these cases which may help point to the correct diagnosis include marked thickening of the papillary dermis due to thickening of collagen fibers, lymphocytes in linear formations between the thickened fibers, and the presence of tiny foci of homogenization in the dermal papillae (Weyers 2015). Extravasation of red blood cells in the dermis may be seen in LS regardless of the degree of sclerosis or atrophy, which accounts for the ecchymotic appearance sometimes seen. An absence of melanosomes in the keratinocytes and a disappearance of the melanocytes are also characteristics common to all LS lesions, and this lack of pigment, as well as associated edema, contributes to the white clinical appearance. In long-standing

1

Benign Diseases of the Vulva

disease, however, postinflammatory pigmentation or melanosis (see section “Postinflammatory Alterations in Pigmentation”) may occur, as well as superimposed LSC (see section “Lichen Simplex Chronicus”), features which can confound the diagnosis significantly.

Clinical Course and Treatment LS diagnosed in childhood may improve somewhat at puberty, but the majority of cases will persist into adulthood (Fistarol and Itin 2013). Untreated, adhesions and scarring from LS can lead to marked changes in the vulvar architecture, with obliteration and/or fusion of the labia minora, stenosis of the introitus, and obscuring of the clitoris. Timely and adequate therapy is imperative to prevent such changes. LS can be controlled with treatment, and symptoms may be relieved, but complete resolution is rare. The majority of patients experience repeated relapses and remissions. Therapy using highpotency topical corticosteroids produces symptomatic relief in a majority of patients and, in some cases, complete resolution of the disease (Fistarol and Itin 2013). Close clinical follow-up is always necessary, however, regardless of response to treatment, and any area of change developing within the LS should be promptly biopsied, as there is a small but significant risk of differentiated (simplex) type vulvar intraepithelial neoplasia (VIN) and subsequent squamous cell carcinoma in postmenopausal women.

Lichen Planus Clinical Features LP is currently understood as an autoimmune disease, in which T-cells react against basal keratinocytes (Goldstein and Metz 2005). Half of female patients with LP are reported to have vulvar involvement (Moyal-Barracco and Wendling 2014). Most patients are between 30 and 60 years old at onset, with a peak incidence in the 50s (Cooper and Wojnarowska 2006). Vulvar pain, pruritus, dyspareunia, and burning are common symptoms, although some patients may be asymptomatic.

27

Three categories of lesions are described, the classical papulosquamous, the erosive-atrophic, and the hypertrophic, and patients may have more than one type simultaneously. The erosive form is the most common on the vulva, usually involving the labia minora and introitus, and appears as welldelineated, red, eroded areas. The papulosquamous form is uncommon on the vulva, and when it occurs it is usually in the setting of generalized disease, while the hypertrophic form is only very rarely seen on the vulva. Both the papulosquamous and hypertrophic forms, when present on the vulva, are usually associated with the more common erosive lesions as well. Papulosquamous lesions present as single or multiple poorly demarcated pink, papules, rather than the well-delineated violaceous flat-topped papules typical of the lesions as described on the extragenital skin (Goldstein and Metz 2005). On the vulva, they usually involve the hair-bearing skin of the labia majora. Hypertrophic lesions present as single or multiple roughened plaques, usually in the perineal or perianal area. All types of lesion may be associated with a lacy, reticulated appearance of the surrounding epithelium, known to dermatologists as Wickham’s striae.

Microscopic Findings The histopathologic features of LP, not surprisingly, vary according to the category and location of the gross lesion. Classically, papulosquamous lesions show a sawtooth pattern of rete ridges with a wedge-shaped area of hypergranulosis, usually without hyperkeratosis. There is a dense band-like infiltrate, consisting predominantly of lymphocytes located in the upper dermis and obscuring the dermo-epidermal junction (Fig. 23). Liquefactive degeneration of the basal epithelial cells is present, and scattered necrotic keratinocytes which form eosinophilic colloid or civatte bodies, which can be seen in the basal epithelium and dropping down into the dermis. On the vulva, classic features may be less well-developed. Hypertrophic LP is similar to the papulosquamous form, but exaggerated acanthosis is present as well, which may cause significant confusion with LSC clinically and microscopically (Moyal-Barracco and Wendling 2014).

28

D. S. Rush and E. J. Wilkinson

striae, symptoms of pain and burning, involvement of other mucosal surfaces, and the presence of concurrent vaginal inflammation. Histologic features were the presence of a well-defined band of inflammation at the dermo-epidermal junction consisting predominantly of lymphocytes and signs of basal layer degeneration.

Fig. 23 LP, papulosquamous type. The lesion shows hyperkeratosis and hypergranulosis, with irregular acanthosis (“sawtooth” rete ridges) and a band-like inflammatory infiltrate of lymphocytes at the dermo-epidermal junction

Clinical Course and Treatment Like LS, untreated or unresponsive erosive LP can result in scarring and agglutination of the labia minora, severe introital and vaginal adhesions, and resultant stenosis or even obliteration of the vaginal canal (Lewis 1998). Topical corticosteroids are the first line of treatment for vulvar LP. Occasionally, other topical agents may be preferred. If topical treatments fail, systemic treatment with corticosteroids may be required (Goldstein and Metz 2005). Although treatment provides significant relief of symptoms in most patients, vulvar LP can be difficult to eradicate, particularly the erosive form. Only 9% of patients had complete resolution in one prospective study (Cooper and Wojnarowska 2006). As with LS, there is a risk of progression to differentiated VIN and squamous cell carcinoma, and close clinical follow-up is essential.

Fixed Drug Eruption Fig. 24 LP, erosive type. The epithelium is thin and predominantly eroded, and an obvious lichenoid infiltrate is lacking. (Reprinted with permission from Ball et al. 2015)

Erosive LP (Fig. 24) frequently lacks diagnostic features on biopsy, particularly if from the central portion of the lesion where the epithelium is entirely absent. If erosive LP is suspected, it is important that the biopsy be from the edge of the eroded area. An attempt to establish a set of diagnostic criteria reached no consensus on any particular indication as being essential to diagnosis (Simpson et al. 2013) but was able to identify several clinical and histologic features considered “supportive” of the diagnosis. The clinical features include location of the eroded areas at the vaginal introitus, the presence of surrounding Wickham’s

Clinical Features The vulva is a favored site for fixed drug eruption, a recurrent type IV hypersensitivity reaction (Andreassi and Bilenchi 2014). Classically associated with pyramidone, sulfonamides, antibiotics, and Nonsteroidal anti-inflammatory drugs (NSAIDs), an increasing number of antibacterial, antifungal, psychoactive, and analgesic drugs are now recognized as potential agents. On the keratinized vulvar skin, fixed drug eruptions are usually single, erythematous, round lesions involving a distinct, clearly delineated area, which may progress to erosion. On the nonkeratinized mucosa, the lesions often appear as erosions with irregular borders. The symptoms are generally mild, consisting of itching and burning.

1

Benign Diseases of the Vulva

29

Microscopic Features Histologically, the lesions demonstrate spongiotic epithelium overlying a dermis with a perivascular and interstitial infiltrate of lymphocytes admixed with eosinophils and neutrophils. The inflammation may extend upward to involve the basal layer of the epithelium, causing basal vacuolar change. In the nongenital skin, the healing phase is characterized by prominent post-inflammatory pigment incontinence, but this is much less common in the vulva. Clinical Course and Treatment The lesions typically appear at the same place on reexposure to the responsible agent, often within minutes to hours (Ball et al. 2015). Avoidance of the responsible agent eliminates the condition. Differential Diagnosis of Lichenoid Dermatoses of the Vulva It can be very difficult to distinguish early LP of any type from LS without significant atrophy or sclerosis, as both are characterized by a lichenoid inflammatory infiltrate and basal degeneration. This differential diagnosis is summarized in Table 6. Clinical features may be helpful, as LS is rarely painful, only rarely involves the extragenital skin, and never involves the vagina or other mucous membranes,

Table 6 Selected features of use to distinguish LS from LP of the vulva Vaginal involvement Parakeratosis

Extravasated red blood cells Stromal hyalinization Location of necrotic keratinocytes, if present

LS No

LP Yes

Common, may be present in vertical columns Common

Rare

Rare

Common

Rare

All layers of the epithelium, clustering may be present

Basal epidermis and upper dermis, no clusters

whereas LP is usually painful and frequently involves the extragenital skin and mucous membranes as well as the vagina. Useful histopathologic features in this differential diagnosis include the lack of melanophages in LS, which may be numerous in LP, the increased presence of lymphocytes in the epidermis in LS, and the difference in distribution of necrotic keratinocytes, which are confined to the basal layer or upper dermis in LP but extend from the basal layer up into all layers of the epidermis, sometimes forming clusters, in LS (Weyers 2015). The pattern of acanthosis, if present, may be helpful, as it is irregular in LP but usually regular in LS. Vertical columns of parakeratosis are recently described features of LS not seen in LP (Weyers 2013; Weyers 2015). Parakeratosis of any sort is uncommon in LP. With blistering, erosive LP may mimic mucous membrane pemphigoid (see section “Pemphigoid”). Mucous membrane pemphigoid can be distinguished by the identification of the characteristic subepithelial blisters and the presence of abundant eosinophils in the vesicles and in the dermis. Direct immunofluorescence studies will further confirm the diagnosis, showing linear IgG and C3 deposits, which is not a feature of LP. A lichenoid pattern may also be seen in fixed drug eruption, Stevens–Johnson syndrome, systemic lupus erythematosus, and graft versus host disease. The presence of eosinophils in the inflammatory infiltrate in fixed drug eruption helps to distinguish it from LS and LP, and the other entities are all extremely rare on the vulva and associated with other clinical findings which should establish the diagnosis. Plasma cell vulvitis may mimic a lichenoid dermatitis to a certain extent, but does not demonstrate damage to the basal epithelium, which will aid in this distinction.

Dermal Homogenization/Sclerosus Pattern In the dermal homogenization/sclerosus pattern, the papillary dermis is thickened by the deposition

30

D. S. Rush and E. J. Wilkinson

Differential Diagnosis of Sclerotic Dermatoses of the Vulva The differential diagnosis includes morphea, which is extremely rare in the vulva and in which the dermal fibrosis extends deeper into the reticular dermis, and radiation dermatitis, in which lymphocytes are rare. Occasional neoplasms of the skin may have associated dermal sclerosis, but unless the biopsy is extremely limited, the accompanying features of malignancy should be evident.

Fig. 25 LS. The well-developed lesion shows the characteristic features of thinning of the epithelium with loss of the rete ridges and a dense, homogenized layer beneath the epithelium. Deep to the homogenized tissue is the residual layer of chronic (lymphocytic) inflammation

of dense hyalinized material. LS is the only entity in this category in the ISSVD 2006 classification (see Table 2).

Clinical Features Clinical findings are as described for the nonsclerotic type of LS. Microscopic Findings When the characteristic features of dermal homogenization and sclerosis are well-developed, the diagnosis is relatively straightforward. The epithelium is typically atrophic, and the rete ridges are lost. Basal vacuolization is common, and there may be sparse lymphocytes in the basal layer. Overlying hyperkeratosis may be present. The dermal homogenization is seen as a subepidermal band of paucicellular, amorphous pink hyaline material, often with edema (Fig. 25), believed to be derived from deposits of protein derived from leaky microvessels with or without inadequate venous and lymphatic drainage (van der Avoort et al. 2010). Beneath the hyaline layer is a band-like infiltrate of lymphocytes which may become less prominent as the lesion ages. As in LS without sclerosis, there may be extravasated red blood cells in the dermis, and melanocytes and melanophages are absent.

Vesiculobullous Pattern The vesiculobullous dermatoses are characterized by fluid-filled spaces within the epidermis or between the epidermis and the dermis, in the absence of associated acantholysis. These disorders are frequently autoimmune in nature, and immunofluorescence is a necessary tool in their evaluation. On the extragenital skin, vesiculobullous lesions are often intact, tense fluid-filled blisters, but on the vulva local conditions are such that most lesions are rapidly ruptured and collapsed, and more likely to present as erosions, which may confound the clinical impression.

Pemphigoid Clinical Features Bullous pemphigoid, the most common of the autoimmune blistering diseases, involves the skin and, in a minority of patients, the mucous membranes. A significant number of patients have associated neurologic disease (Schiavo et al. 2013). Mucous membrane pemphigoid, formerly known as cicatricial pemphigoid, is a similar disease, but involves the mucous membranes exclusively. Both are caused by autoantibodies, sometimes developed in the context of a drug reaction, viral infection, or other inducing factor, to the components of hemidesmosomes and to type VII collagen, leading to loss of adhesion between the basal epithelium and the basement membrane. Although bullous pemphigoid is significantly more common than mucous membrane pemphigoid, vulvar involvement is more common

1

Benign Diseases of the Vulva

31

Fig. 26 Bullous pemphigoid. At low power, the separation of the epidermis from the basement membrane is evident

with the latter (Moyal-Barracco and Wendling 2014). Patients are usually older women, in the sixth to seventh decades of life, presenting with erosions, erythema, and small blisters of the labia minora, majora, and perianal mucosa which heal with scarring. A positive Nikolsky phenomenon (slippage and detachment of the superficial epidermis from the underlying dermis when the examining finger is slid over the skin surface) may be observed. Bullous pemphigoid presents with symptoms weeks to months before the eruption of blisters, which may consist of itching, erythema, or urticaria. When the bullae develop on the vulva, they rupture rapidly, leaving shallow erosions. The Nikolsky sign is absent in bullous pemphigoid.

Fig. 27 Mucous membrane pemphigoid. There is separation of the epithelium from the basement membrane, with inflammation in the dermis, but negligible inflammation within the blister itself. (Reprinted with permission from Hoang et al. 2015a)

Microscopic Findings In both types of pemphigoid, microscopic examination shows subepidermal blister formation with the blisters containing variable numbers of eosinophils and neutrophils (Figs. 26 and 27). Bullous pemphigoid shows a mixed inflammatory cell infiltrate within the dermis consisting predominantly of eosinophils with lesser components of lymphocytes and histiocytes. Direct immunofluorescence demonstrates linear IgG and complement C3 along the basement membrane in both types of pemphigoid (Fig. 28), and serologic studies will identify the circulating autoantibodies. Clinical Course and Treatment Bullous pemphigoid is usually a self-limited disease which regresses in months to years. Systemic corticosteroids are the principal treatment, and topical corticosteroids may be an alternative in

Fig. 28 Direct immunofluorescence of bullous pemphigoid shows linear deposition of C3 along the basement membrane. (Reprinted with permission from Hoang et al. 2015a)

32

limited disease (Ruocco et al. 2013b). Immunosuppressive drugs, such as azathioprine and cyclophosphamide, may be used in combination with or as alternatives to steroids, and dapsone and tetracycline-nicotinamide are also effective alternatives. Intravenous immunoglobulin therapy and plasmapheresis can also be effective but are reserved for severe, refractory cases. If the condition is drug related, the offending medication should be discontinued, which in some cases will allow for complete resolution. In mucous membrane pemphigoid, scarring and fibrosis can result in significant distortion of the vulvar anatomy, similar to advanced vulvar LS or LP, and more aggressive treatments may be required to minimize this risk.

Pemphigoid Gestationis Pemphigoid gestationis, formerly known as herpes gestationis, is another subepidermal bullous dermatosis, with morphology similar to other forms of pemphigoid, but immunofluorescence which shows predominantly C3, with associated IgG in only a minority of cases. It occurs in approximately one in 40–50 thousand pregnancies, presenting in the second or third trimester or immediately postpartum as an intensely pruritic eruption of small vesicles. It involves the mucosal membranes in 20% of cases, (Kneisel and Hertl 2011) but involves the vulva in only 10% (Hoang et al. 2015a). Treatment is geared toward the control of symptoms and usually consists of topical corticosteroids and oral antihistamines, with addition of systemic steroids for more severe cases (Kasperkiewicz et al. 2012). Most cases resolve shortly after delivery.

Linear IgA Disease Clinical Features Linear IgA disease is the most common autoimmune blistering disorder in children, but may also present in adulthood, usually in patients between 20 and 40 years old or over 60 years old. The disease commonly involves the lower abdominal, pelvic, inguinal, and genital areas, presenting as

D. S. Rush and E. J. Wilkinson

clusters of annular lesions that usually are pruritic and typically evolve over the course of 24 h into ulcerated, crusted lesions. Mucosal surfaces are involved in 50% of patients (Kneisel and Hertl 2011). In some cases, the eruption is preceded by a bacterial or viral infection (Egan and Zone 1999; Wojnarowska and Frith 1997) and other cases are drug induced, with vancomycin being one of the most commonly implicated agents (Klein and Callen 2000).

Microscopic Findings Biopsy of an early bullous lesion reveals subepithelial vesicles that contain predominantly neutrophils, sometimes admixed with eosinophils, within the vesicular fluid. Neutrophilic microabscesses may occur within the papillary dermis and epidermis, but more often, the neutrophils are evenly distributed along the basement membrane. The diagnostic finding is the identification of a linear deposition of IgA in the basement membrane on direct immunofluorescence studies. Clinical Course and Treatment First-line therapy is usually dapsone or sulfapyridine, sometimes with the addition of prednisone as well, with supplementary topical treatments to prevent superinfection and speed reepithelialization. Refractory cases may be managed with erythromycin, colchicine, flucloxacillin, or intravenous immunoglobulin therapy (Kasperkiewicz et al. 2012)

Pemphigus Pemphigus is an immunologically mediated blistering disease of the skin caused by autoantibodies directed against desmogleins (Hoang et al. 2015a). About 25% of cases are associated with other autoimmune diseases (Ruocco et al. 2013a) Many factors appear to trigger the development of the autoantibodies in individuals with a genetic predisposition, including drugs, viruses, and others (Ruocco et al. 2013a). Three forms of the disease are recognized, pemphigus vulgaris, pemphigus foliaceous, and pemphigus vegetans, all of which may involve the vulva.

1

Benign Diseases of the Vulva

Clinical Features The vesiculobullous lesions of pemphigus vulgaris may involve the skin, oral mucosa, and other mucosal surfaces. Studies have shown 22–44% of patients with pemphigus vulgaris have genital involvement (Kavala et al. 2015; Barbosa et al. 2012), and the vulva is the second most common mucosal site of involvement (Barbosa et al. 2012). In some cases, the genital area is the only site of disease (Barbosa et al. 2012). The labia majora and minora are the most common sites on the vulva (Kavala et al. 2015; Barbosa et al. 2012), and the lesion usually appears as an erosion rather than a blister, due to the location. Pemphigus foliaceous is similar to pemphigus vulgaris but involves only the skin, never the mucous membranes. Genital involvement is not uncommon and occurs exclusively on the keratinizing surfaces, involving either the labia minora or majora with equal frequency (Barbosa et al. 2012). Like pemphigus vulgaris, the lesions present predominantly as erosions on the vulva. A rare variant, pemphigus vegetans, shows a predilection for the flexural areas, and a small number of cases with involvment of the vulva and groin have been reported (Zaraa et al. 2010). The lesions may present as a localized, indurated, inflamed area with oozing vesicles on keratinized skin and as erosive plaques on mucosal surfaces. Most patients have concurrent oral involvement and multifocal disease, but rare cases may involve only one site (Ruocco et al. 2015; Zaraa et al. 2010). Microscopic Findings In pemphigus vulgaris, there is an intraepidermal separation in the suprabasilar layer, leaving the basal keratinocytes attached to the floor of the blister lending an appearance likened to a “row of tombstones,” (Fig. 29), and there may be prominent acantholysis in the follicular epithelium (Fig. 30). In pemphigus foliaceous, the separation occurs in the granular layer of the superficial epidermal layers, leading to the development of subcorneal blisters, without associated acantholysis of follicular epithelium. In both pemphigus vulgaris and pemphigus foliaceous, there

33

Fig. 29 Pemphigus vulgaris. The acantholysis begins in the suprabasilar epithelium, with basal keratinocytes aligned like “tombstones” along the base of the blister. (Reprinted with permission from Hoang et al. 2015a)

is a perivascular and interstitial infiltrate in the dermis consisting of lymphocytes, neutrophils, and, to a lesser degree, eosinophils. Pemphigus vegetans appears similar to pemphigus vulgaris, with the added findings of eosinophilic microabscesses and verrucous epidermal hyperplasia (Fig. 31), and with eosinophils forming a more prominent component of the dermal inflammatory infiltrate (Moyal-Barracco and Wendling 2014). In all three types of pemphigus, direct immunofluorescent studies show intercellular deposition of IgG and C3.

Clinical Course and Treatment Untreated, pemphigus may spread to involve increasing amounts of the body surface and may ultimately result in death from fluid and protein loss. With current therapy, the mortality rate is between 5 and 10% (Ruocco et al. 2013a) and represents a combined effect of the disease and complications of the high doses of systemic glucocorticoids and other drugs used to treat it. The

34

utility of the addition or substitution of additional drugs and treatments such as etretinate, azathioprine, cyclophosphamide, cyclosporine, and intravenous immunoglobulin to the traditional steroid

D. S. Rush and E. J. Wilkinson

regimen is still not entirely clear but appears to be useful in reducing the risk of relapse (Atzmony et al. 2015).

Bullous Systemic Lupus Erythematosus Bullous systemic lupus erythematosus is a very rare manifestation of lupus, seen in less than 5% of patients. Involvement of the vulva is even less common, but rare cases have been reported (Miziara et al. 2013). It develops when autoantibodies to the basement membrane and other antigens deposit in the skin, resulting in subepidermal blisters similar to those of bullous pemphigoid. The two can be distinguished by immunofluorescence; in lupus, there is a characteristic “full house” pattern on immunofluorescence, showing deposition of IgG, IgM, IgA, and C3 at the dermo-epidermal junction.

Fig. 30 Pemphigus vulgaris. The acantholysis in this disease involves the follicular epithelium as well, as shown here. (Reprinted with permission from Hoang et al. 2015a)

Differential Diagnosis of Vesiculobullous Dermatoses of the Vulva The erosions of the vesiculobullous dermatoses may appear similar to the ulcerations caused by many vulvar infections, but judicious use of special stains and cultures for suspected pathogens can easily distinguish them. When intact

Fig. 31 Pemphigus vegetans. In this variant of pemphigus, there is marked epidermal hyperplasia (a), with formation of eosinophilic microabscesses (b). (Reprinted with permission from Hoang et al. 2015a)

1

Benign Diseases of the Vulva

35

Table 7 Selected features of use to distinguish between vesiculobullous dermatoses of the vulva Diagnosis Bullous pemphigoid

Location of blister Subepidermal

Mucous membrane pemphigoid Pemphigoid gestationis Linear IgA disease

Subepidermal

Pemphigus vulgaris

Suprabasal (intraepidermal) Subcorneal (intraepidermal) Suprabasal (intraepidermal) Subepidermal

Pemphigus foliaceous Pemphigus vegetans Bullous SLE

Dermatitis herpetiformis

Subepidermal Subepidermal

Subepidermal

Typical findings on direct immunofluorescence Linear IgG and C3 along the basement membrane Linear IgG and C3 along the basement membrane Linear C3 along the basement membrane Linear IgA along the basement membrane Intercellular IgG and C3

Other useful clues Prominent eosinophils Significant scarring Concurrent or recent pregnancy Prominent neutrophilic infiltrate along the basement membrane Acantholysis in follicular epithelium

Intercellular IgG and C3 Intercellular IgG and C3 Linear IgG, IgM, IgA, and C3 along the basement membrane (full house) Granular IgA at tips of the papillary dermis

blisters are present and can be biopsied, establishing the level of separation in the epidermis is the first step in narrowing the differential diagnosis. In pemphigoid and bullous lupus erythematosus, the separation occurs at the basement membrane, while in pemphigus the separation is intraepidermal. Immunofluorescent studies will be of use even when the vesiculobullae are ruptured and will demonstrate distinctive patterns which are usually definitive. The characteristic association with concurrent pregnancy and resolution following delivery should distinguish pemphigoid gestationis from other entities. Dermatitis herpetiformis very rarely involves the vulva but may mimic other vesiculobullous dermatoses, particularly linear IgA disease, as it may also show microabscesses in the papillary dermis. The diagnosis of dermatitis herpetiformis may be suggested by the finding of fibrin deposits in the papillary dermis as well, but it is more definitively identified by its distinctive pattern of granular IgA at the dermal papillae on direct immunofluorescence (Table 7). The acantholytic disorders can likewise be distinguished from vesiculobullous diseases by the results of direct immunofluorescent studies,

Eosinophilic microabscesses Other manifestations of SLE Fibrin in dermal papillae

which are uniformly negative in acantholytic disorders (see section “Acantholytic Pattern”).

Acantholytic Pattern Acantholysis occurs when there is deficient cohesion between cells due to loss of desmosomes in the epithelium. Clinically, this dyscohesion manifests as small, flaccid blisters in the epithelium, and microscopically, the cells appear disorganized and separated into clusters or individual cells surrounded by empty spaces. In acantholytic disorders, the basal layers remain attached to the basement membrane, and the disruption is confined to the suprabasal layers of epithelium.

Hailey–Hailey Disease Clinical Features Hailey–Hailey disease is caused by a mutation in ATP2C1 and is inherited as an autosomal dominant trait. In approximately one-third of patients, however, the mutation is sporadic, and there is no family history of the disease. The disease

36

D. S. Rush and E. J. Wilkinson

respond completely. In severe cases, surgery, botulinum toxin, dermabrasion, or laser ablation may be used and can result in long-term remission (Farahnik et al. 2017).

Darier Disease

Fig. 32 Hailey–Hailey disease. The epithelium shows suprabasilar acantholysis, with clusters of keratinocytes maintaining cohesion within the acantholytic space

typically presents in the second or third decade of life. Intertriginous areas usually are involved, with symmetric distribution, but several cases in which the lesions are confined exclusively to the vulva have been reported (Wieselthier and Pincus 1993). The usual clinical presentation is recurrent clusters of vesicles that develop, rupture, and leave crusted, moist papules that later coalesce to form plaques.

Microscopic Findings The disease is characterized by intradermal acantholysis involving at least half of the epithelium above a suprabasilar cleft, with resultant suprabasilar lacunae. Acantholytic cells that maintain their nuclear details remain attached to each other in the acantholytic space, giving a dilapidated brick-wall appearance (Fig. 32). Mild dyskeratosis in the acantholytic cells is characteristic, and rarely, corps ronds, distinctive forms of dyskeratosis with pyknotic nuclei and a clear perinuclear halo, are seen in the granular layer. Chronic inflammation in the dermis is prominent, and direct immunofluorescent studies are negative. Clinical Course and Treatment The disease runs a chronic, relapsing course. Treatment modalities include oral and topical corticosteroids and antibiotics, cyclosporine, dapsone, and methotrexate for refractory disease (Farahnik et al. 2017), but most patients fail to

Clinical Features Darier disease is a genetic disease closely related to Hailey–Hailey and also inherited as an autosomal dominant trait, although nearly half of cases are sporadic (Takagi et al. 2016), caused by mutation in ATP2A2. The onset of disease occurs by 20 years of age, with a peak during adolescence. Factors which appear to exacerbate disease include high temperature, high humidity, excessive sweating, and mechanical irritation. It is not surprising, then, that vulvar or groin involvement is common (Moyal-Barracco and Wendling 2014). Other areas typically involved are the chest, neck, back, and ears. On clinical examination, the lesions are crusted, hyperkeratotic papules that often appear darker than the surrounding skin. Microscopic Findings The microscopic findings include acantholysis of the suprabasal epithelial cells in an acanthotic epidermis, resulting in clefts that extend from the basal layer through the granular layer. Columns of parakeratosis are characteristic, and dyskeratosis is prominent, with the formation of corps ronds and grains, another form of dyskeratotic cell distinguished by an elongated nucleus and scant cytoplasm without a perinuclear halo, throughout the granular layer. Dermal inflammation is usually minimal. Clinical Course and Treatment Like Hailey–Hailey, Darier disease is a chronic disease with frequent relapses. Treatment is aimed at controlling the symptoms and may include topical steroids or vitamin D3 ointment, oral retinoids or cyclosporine, topical antibiotics and antifungals to prevent superinfection, and avoidance of exacerbating factors. In severe cases, laser ablation may be used (Takagi et al. 2016).

1

Benign Diseases of the Vulva

37

Table 8 Summary of features useful to distinguish between acantholytic dermatoses of the vulva

Age of onset Areas involved Columns of parakeratosis Dyskeratosis Pattern of acantholysis Degree of acantholysis

Hailey–Hailey disease 2nd–3nd decade

Papular acantholytic dyskeratosis 2nd–5th decade

All intertriginous areas No

Darier disease Before age 20, with peak in adolescence Chest, back, neck, ears, groin Yes

Genital area folds No

Minimal Diffuse Severe

Prominent Diffuse Moderate

Prominent Localized Moderate

Papular Acantholytic Dyskeratosis Clinical Features Papular acantholytic dyskeratosis, or papular genitocrural acantholysis, is a rare chronic disorder which typically presents in the second to fifth decade of life. It presents as white to skin-colored smooth papules involving the perineum, labia majora, and inguinal folds. The lesions are usually asymptomatic but may be associated with pain or itching. Microscopic Findings Histologic examination reveals acanthosis, hyperkeratosis, hypergranulosis, and focal parakeratosis, with suprabasal acantholysis. Dyskeratotic cells are a prominent feature, forming corps ronds and grains throughout the thickness of the epidermis. Some cases may show a superficial dermal perivascular lymphocytic infiltrate as well. The histologic appearance is similar to Hailey–Hailey disease and Darier disease, and there is some evidence of a genetic relationship with Hailey–Hailey disease as well (Pernet et al. 2012; Yu et al. 2016). Clinical Course and Treatment Treatment is not required if asymptomatic but may consist of topical steroids or retinoids, tacrolimus, cryotherapy, or laser therapy. Response to treatment is variable. Complete resolution is not usually achieved, but symptoms are greatly reduced (Yu et al. 2016). Differential Diagnosis of Acantholytic Dermatoses of the Vulva Clinical features such as family history, age of onset, and whether there is involvement of cutaneous sites outside of the groin and vulva may be

useful in differentiation of the acantholytic dermatoses involving the vulva. Histologic features are also useful. The acantholysis is more marked in Hailey–Hailey disease than in Darier disease, and in papular acantholytic dyskeratosis, the acantholysis is localized rather than more diffuse as in Hailey–Hailey or Darier disease. In contrast to Darier disease and papular acantholytic dyskeratosis, there is minimal, if any, dyskeratosis and only very rare corps ronds or grains in Hailey–Hailey disease. Discrete columns of parakeratosis are seen only in Darier disease. Warty dyskeratoma is also characterized by prominent acantholysis, and may enter the differential, but unlike the acantholytic dermatoses, it is a solitary lesion. The differential diagnosis of acantholytic dermatoses of the vulva is summarized in Table 8.

Granulomatous Pattern Granulomatous inflammation is defined by the presence of clusters of epithelioid histiocytes and multinucleated giant cells with an associated inflammatory infiltrate consisting predominantly of lymphocytes. When the skin is involved, the inflammatory process involves the dermis and or subcutaneous tissue, while the epithelium is spared.

Crohn Disease Clinical Features Vulvar involvement by Crohn disease may occur as a direct extension of perianal disease continuous with gastrointestinal tract involvement, or, more commonly, as “metastatic” disease, distant

38

and separate from gastrointestinal tract involvement. In most cases, concomitant gastrointestinal involvement is present, but in approximately 25% of patients, particularly in children, vulvar lesions may precede the gastrointestinal diagnosis by many years (Duan et al. 2014; Moyal-Barracco and Wendling 2014). Four types of vulvar manifestations have been described. The most common is asymptomatic swelling of the labia minora and/or majora, reported in 67% of patients in one series (Barret et al. 2013). Deep, linear, “knife-like” ulcerations are also common. Hypertrophic lesions, presumed to be related to impaired lymphatic drainage, may also develop and may progress to form quite large acquired lymphangiomas (see section “Lymphangioma Circumscriptum”). The least common lesion is chronic suppuration or abscess. Despite the sometimes striking clinical findings, vulvar Crohn disease is typically asymptomatic. Only a minority of patients complain of symptoms, which may include pain, pruritis, discharge, dyspareunia, and dysuria (Barret et al. 2013).

Microscopic Findings Histologic findings understandably differ according to the type of lesion but typically include a subacute or chronic inflammatory infiltrate and epidermal ulceration in addition to caseating and noncaseating granulomata which may involve any region of the dermis. Small abscesses of neutrophils may also be present in the dermis. Hypertrophic lesions demonstrate dilation of lymphatic vessels and varying degrees of dermal fibrosis. Clinical Course and Treatment The clinical course is unpredictable; lesions may regress spontaneously or may stubbornly persist, requiring surgical resection to finally achieve resolution. The development of fistulas and draining sinus tracts is a common complication, particularly with involvement of the anus and rectum. In addition to the systemic therapy with steroids and immunomodulators used for the management of the intestinal disease, treatment of Crohn disease of the vulva may involve the addition of topical antibiotics and steroids to achieve a local response.

D. S. Rush and E. J. Wilkinson

Melkersson–Rosenthal Syndrome and Granulomatous Vulvitis Clinical Features Melkersson–Rosenthal syndrome is a poorly understood disease typically characterized by orofacial swelling, facial palsy, and abnormal furrows of the tongue. Some cases may be accompanied by granulomatous cheilitis, a swelling and inflammation of the lips, or with a similar process on the vulva known as granulomatous vulvitis; only very rarely do both cheilitis and vulvitis manifest in the same patient (Sbano et al. 2007). It is hypothesized that cases of granulomatous cheilitis or vulvitis occurring in the absence of other symptoms of Melkersson–Rosenthal syndrome represent formes frustes of the disease. Clinically, granulomatous vulvitis presents with painless erythema and swelling of the labia majora. Unlike Crohn disease, ulceration is very rare.

Microscopic Findings Histologically, the most striking finding is of non-necrotizing granulomata extending deep into the dermis. The findings may be indistinguishable from Crohn disease, and clinical correlation is imperative. Clinical Course and Treatment There is no definitive therapy for the disorder, which runs a chronic, relapsing course. Treatment is similar to that of Crohn disease; systemic corticosteroids and immunomodulating agents alone or in combination with topical or intralesional steroids and antibiotics have all been reported to be effective in treating symptoms (Ghosh et al. 2011). As in Crohn disease, surgery may be required as a last resort for refractory cases (Ghosh et al. 2011).

Sarcoidosis Clinical Features Genitourinary involvement with sarcoidosis is very uncommon, and only rare cases of vulvar involvement have been reported (Pereira and

1

Benign Diseases of the Vulva

Khan 2017). Patients complain of itching, and involved areas may have a papular, nodular, or plaque-like appearance. Less commonly, the involved areas are ulcerated.

Microscopic Findings Histologic examination shows florid granulomatous inflammation in the dermis beneath a mildly acanthotic and hyperkeratotic epithelium. There is minimal inflammation aside from the abundant non-necrotizing granulomata. Clinical Course and Treatment Treatment with topical steroids has been effective in other cutaneous sites, but therapy is not necessary for small asymptomatic skin lesions, and lesions may regress on their own in time (Pereira and Khan 2017). Differential Diagnosis of Granulomatous Dermatoses of the Vulva The clinical history may explain the findings, as the majority of patients will already have an established diagnosis of a granulomatous disease with systemic involvement. In the absence of such history, granulomatous response to foreign material or keratin from a ruptured epidermal inclusion cyst, a much more common finding on the vulva than granulomatous dermatoses, should be ruled out before these more unusual conditions are considered. Regardless of the clinical history, an investigation to rule out mycobacterial or fungal organisms is warranted to exclude infectious

39

causes. Granulomatous inflammation may also be seen in pyoderma gangrenosum (PG) (see section “Pyoderma Gangrenosum”), but the presence of an associated neutrophilic inflammation, not a feature of the granulomatous dermatoses, will readily distinguish this entity. The differential diagnosis of granulomatous inflammation of the vulva is summarized in Table 9.

Vasculopathic Pattern The disorders classified as vasculopathic in the ISSVD 2006 classification present as erosions or ulcerations which are attributed to local disruption of blood supply. Histologic examination in these conditions shows evidence of blood vessel damage in a background of widespread dermal inflammation.

Vulvar Aphthosis Vulvar aphthosis manifests as painful, shallow, well-demarcated ulcers typically less than 5 mm in diameter with a gray-white base. The most common site is the inner aspect of the labia minora. The onset is acute, and most ulcers heal within 7–10 days. As with oral aphthosis, the etiology is unclear. Risk factors for these ulcers include stress, infections, vitamin deficiency, and family history (Huppert et al. 2006). The lesions are rarely biopsied, and the histologic features are

Table 9 Selected features of use in the differential diagnosis of granulomatous inflammation of the vulva

Usual type of granuloma Ulceration Systemic disease Other useful features

Infection Necrotizing

Foreign body reaction Non-necrotizing

Crohn disease Necrotizing and non-necrotizing

Often Sometimes

No No

Organisms identified on special stains or cultures

Presence of foreign material (suture, keratin), multinucleated foreign body-type giant cells

Often Often (gastrointestinal) Edema, linear ulcerations

Melkersson–Rosenthal syndrome/ granulomatous vulvitis Non-necrotizing

Sarcoid Nonnecrotizing

No Yes (orofacial)/no

Occasional Yes

–

Pruritis

40

nonspecific, but in the initial stages, a dense neutrophilic infiltrate with a leukocytoclastic vasculitis involving the superficial capillaries may be seen.

D. S. Rush and E. J. Wilkinson

colchicine, azathioprine, cyclosporine, interferon, and tumor necrosis factor (TNF)-blocking agents (Yazici et al. 2010).

Plasma Cell Vulvitis (Vulvitis of Zoon) Behçet Disease Clinical Features Behçet disease is a multisystemic vasculitis of small and large vessels. The onset of disease is typically in the third decade of life, and it very rarely presents in childhood or in patients over 50 years old (Yazici et al. 2010). Clinical findings may include mucosal ulcers, a variety of cutaneous lesions, arthritis, uveitis, thrombophlebitis, and gastrointestinal and central nervous system lesions. Almost all patients have recurrent oral apthous ulcers, and 50–85% have genital ulcers as well (Mat et al. 2013). Genital ulcers are more common in female patients, and involvement of the vulva alone is thought to occur as a limited form of the disease (Moyal-Barracco and Wendling 2014). The most common location of the vulvar lesions is the labium majus, followed by the labium minus. The ulcers are painful and necrotic, with a sharp border. Microscopic Findings Histologic findings are nonspecific. Early lesions show a neutrophilic infiltration, which evolves into a lymphoplasmacytic one. In half of patients, a lymphocytic vasculitis may be seen, but a leukocytoclastic vasculitis is rare in Behçet disease (Mat et al. 2013). Clinical Course and Treatment The ulcers heal in 2–4 weeks, and unlike most infectious ulcers, lesions of Behçet disease over 1 cm are likely to leave a scar (Mat et al. 2013; Yazici et al. 2010). The diagnosis is made based on the constellation of clinical findings and tends to follow a course of repeated relapses and remissions, although in many patients, the disease resolves over time. Active disease is managed based on the systems involved. Genital ulcers are generally treated with topical steroids as needed to control symptoms. More severe disease may be treated with systemic therapies including

Clinical Features Plasma cell vulvitis is relatively rare and typically presents with a single shiny, red to orangebrown plaque, usually of the labia minora or introitus. Most lesions are associated with severe pruritus or burning pain (Virgili et al. 2015). The etiology is unknown, and trauma, viral infection, and autoimmune response have been proposed as possible triggers (Moyal-Barracco and Wendling 2014). Microscopic Findings Microscopic examination of early lesions may show slightly thickened epithelium with parakeratosis and a lichenoid lymphoplasmacytic infiltrate. In more advanced lesions, the epithelium is thinned, with flattening of rete ridges and lack of a granular layer or keratinized surface. Spongiosis is a common finding. There may be a neutrophilic infiltrate in the epithelium, with occasional superficial erosion, but ulceration is rare. The more prominent inflammation is in the dermis and consists predominantly, but not exclusively, of plasma cells (Fig. 33). This infiltrate is diagnostic when it consists of 50% plasma cells or more, and the diagnosis is precluded when the infiltrate is less than 25% plasma cells. When the infiltrate is more than 25% plasma cells but less than 50% plasma cells, additional features must be present which support the diagnosis. Prominent dilation of dermal blood vessels with associated intradermal red cell extravasation and hemosiderin-laden macrophages is a particularly useful supportive feature, and this finding is the rationale for its classification as a vasculopathic disorder in the ISSVD system (see Table 2). Clinical Course and Treatment The clinical course is one of recurrent relapses and remissions. Perineal hygiene, supportive care, and

1

Benign Diseases of the Vulva

41

topical and intralesional corticosteroids are the usual treatment (Yoganathan et al. 1994). Other options include etretinate (Robinson et al. 1998) and topical tacrolimus (Virgili et al. 2008). Some patients are fortunate enough to resolve spontaneously, while in others, the disease may be refractory to all treatments, and surgery may be the only option for resolution (Gurumurthy et al. 2009).

Differential Diagnosis of Vasculopathic Dermatoses of the Vulva The ulcers of vulvar aphthosis must be distinguished from infectious causes of ulceration, which can be achieved by the use of special stains and cultures when indicated. The perivascular plasma cell infiltrate in plasma cell vulvitis may also be suggestive of syphilis, which can be ruled out with serologic studies and by the failure to demonstrate spirochetes by the usual techniques.

Fig. 33 Plasma cell vulvitis. An inflammatory infiltrate composed predominantly of plasma cells is present beneath the epithelium

Clinical features are of the most use in distinguishing aphthosis from Behçet disease, as the histologic features are variable and nonspecific, but the finding of a leukocytoclastic vasculitis on biopsy favors aphthosis. A summary of the differential diagnosis of vasculopathic dermatoses of the vulva is provided in Table 10.

Miscellaneous Dermatoses Lacking a Dominant Histologic Pattern Amicrobial Pustulosis of the Folds Amicrobial pustulosis of the folds (APF) is a rare neutrophilic dermatosis affecting predominantly young female patients. Most cases develop in patients with a previous diagnosis of some type of autoimmune disease, but cases have been reported in which APF presented prior to the diagnosis of the autoimmune disease (Marzano et al. 2008), indicating the need for close clinical follow-up and testing of individuals without any previous diagnosis. The onset of APF is typically sudden, and it follows a course of chronic remissions and relapses. It presents as small pustules, some but not all of which are associated with hair follicles, which may coalesce into erosive, crusting plaques. The anogenital area is always involved, as well as the inguinal or other skin folds. Despite the name, however, involvement is often not limited to the skin folds alone but may involve multiple other areas of the skin as well (Schneider et al. 2016; Wang et al. 2017).

Table 10 Selected features of use to differentiate vasculopathic dermatoses of the vulva Ulcerateda Leukocytoclastic vasculitis Lymphocytic vasculitis Prominent plasma cellsb Mouth ulcers Intradermal red cell extravasation, hemosiderin deposition a

Aphthous ulcers Yes Yes, in early lesions Yes No Yes Rare

Rule out infectious causes, especially viral Rule out syphilis

b

Behcet disease Yes No Yes No Yes Rare

Plasma cell vulvitis No No Yes No Frequent

42

Microscopic examination shows small pustules within the epidermis, with an associated neutrophilic infiltrate in the epidermis, dermis, and perivascular space, often accompanied by pronounced spongiosis. Cultures of closed pustules are uniformly negative, but eroded areas may become superinfected. The differential diagnosis includes infection, pustular psoriasis, and other blistering diseases of the skin, and these disorders must be ruled out by appropriate studies and clinical correlation to establish a definitive diagnosis. Treatment options include topical and systemic steroids, cyclosporine, and dapsone. Recently, treatment with cimetidine in combination with ascorbic acid has been reported to be effective (Marzano et al. 2008).

Erythema Multiforme/ Stevens–Johnson Syndrome Erythema multiforme is an acute, self-limited hypersensitivity reaction. Vulvar involvement is not uncommon, but the disease is very rarely confined to the vulva. The instigating agent is often infection, particularly with HSV (Schofield et al. 1993) or mycoplasma (Saitoh et al. 1995). Other causes include drug reaction, pregnancy, malignancy, or radiotherapy. Clinically, erythema multiforme presents as painful red areas that quickly evolve into blisters and culminate in painful vulvar ulcers. The histopathologic features vary according to the age of the lesion. Early lesions show nonspecific changes of dermal edema and chronic inflammation in the dermis, erythrocyte extravasation, and focal interface dermatitis. The fully developed lesion shows marked vacuolar degeneration of the basal layer and necrosis of individual keratinocytes. In the late stages of the lesion, the necrotic epidermis detaches, forming a subepidermal blister which proceeds to ulceration and subsequent reepithelialization. Stevens–Johnson syndrome is similar to erythema multiforme but involves the skin more extensively, as well as the mucous membranes, particularly the oral mucosa and conjunctiva. It is

D. S. Rush and E. J. Wilkinson

also distinguished by involvement of other organ systems and an association with high fever and other systemic symptoms not seen in erythema multiforme. The most common cause of the Stevens–Johnson syndrome is drug reaction. Vulvar involvement is extremely rare with Stevens–Johnson syndrome, and the histologic findings are identical to those of erythema multiforme.

Pyoderma Gangrenosum PG is a progressive necrotic and ulcerative condition of the skin which appears to be autoimmune in nature (Ahronowitz et al. 2012). Most cases involve the lower extremities, and 50% of cases occur in association with systemic disease, most commonly inflammatory bowel disease or arthritis (Dabade and Davis 2011). In these cases, the associated disease is usually diagnosed prior to the skin disease, but in some cases, the PG precedes or is part of the initial presentation (Ahronowitz et al. 2012). In 25% of cases, the disease arises in an area of recent trauma, and when the vulva is involved, it may present at the site of previous surgery or healed obstetric lacerations (Reed et al. 2013). Occasional cases involving the vulva have been reported to have no other associated disease or previous trauma (Satoh and Yamamoto 2013). Clinically, PG presents as a painful, sharply demarcated and often deep ulceration with a characteristic raised, purple, and undermined margin to the ulceration. The microscopic findings are not specific, but biopsy is nonetheless advisable if only to rule out infectious causes of ulceration. In addition to the epithelial ulceration with a hyperkeratotic, hyperplastic margin, severe dermal inflammation consisting predominantly of neutrophils may be identified, but infectious organisms are not present in PG. Lymphocytic vasculitis may also be seen. Treatment of the associated autoimmune disease, if there is one, may lead to resolution of the PG as well, but the activity of the PG is frequently independent of the associated disease, necessitating specific intervention. Treatment usually consists of a combination of local wound care and

1

Benign Diseases of the Vulva

topical and/or systemic steroids or immunomodulating agents. The lesions are often refractory to treatment and require a trial-and-error approach to management until an effective regimen is identified. Surgical treatment may be considered and can be successful but should be used with caution in patients whose disease appears to have been induced by local trauma, as in these patients, surgery may only worsen the disease (Ahronowitz et al. 2012).

Vulvodynia Clinical Features The latest ISSVD consensus definition of vulvodynia is “vulvar pain of at least three months duration, without a clear identifiable cause, which may have potential associated factors” (Bornstein et al. 2016). The ISSVD classification of vulvodynia is based on the site of pain, whether it is localized or generalized, and whether it is provoked, spontaneous, or mixed. Vulvodynia may also be classified as primary, occurring with the first attempt at vaginal penetration, or secondary, developing after a period of normal functioning. The estimated prevalence in the USA is 3–7% (Xie et al. 2012), and the median age of onset is 28 years (Leclair et al. 2011). The etiology of vulvodynia is not well understood and is likely multifactorial. Patients frequently have a history of chronic or recurrent vulvovaginal infection, with persistent pain following resolution of the infectious process. In older terminology, the disease was commonly called vestibulitis, suggesting an inflammatory etiology, and this term may still be applied in certain cases. However, it has subsequently been recognized that many cases are not associated with significant inflammation, or not solely with inflammation, and are not confined to the vestibule, in which case the term “vestibulitis” is not appropriate. Whether or not to biopsy patients with significant pain but without visible lesions has been somewhat controversial, particularly as the histologic findings are variable and nonspecific.

43

Currently, it is recommended to biopsy patients unresponsive to empirical therapy and with no definite cause of pain, if only to rule out subclinical infection or inflammatory disease. Microscopic Features Recent studies suggest certain findings may be associated with vulvodynia in the absence of other histopathologic findings. The most significant findings include increased numbers of mast cells and neural hypertrophy and hyperplasia, which are suggested to represent evidence of abnormal immune response and neuromodulation leading to abnormal sensory perception in the area (Akopians and Rapkin 2015; Hoffstetter and Shah 2015; Leclair et al. 2011; Regauer et al. 2015). Other reported findings include nonspecific lymphocytic inflammation, increased progesterone receptor expression, and lymphoid tissue with germinal center formation (Tommola et al. 2015). The latter finding suggests an altered immune response may play a role in some cases, as such tissue was not found in controls. Changes to central neurologic processing of sensory stimuli in the region have also been hypothesized as an explanation for patients’ symptoms (Akopians and Rapkin 2015). Clinical Course and Treatment The combination of the lack of distinctive physical and histologic findings and poorly understood etiology make diagnosis and management of these patients extremely difficult. Treatments aim at symptom reduction and include sitz baths, topical emollients, topical anesthetics, topical and oral antidepressants, oral neuropathic pain medications such as gabapentin, the elimination of possible irritants and allergens, and adjunctive psychotherapy. Often a combination of these interventions is used. Spontaneous remission has been seen in some patients, but the majority does not achieve complete resolution despite multiple attempted therapies (Hofstetter and Shah 2015). Vestibulectomy is undertaken in some women with localized vestibulodynia, and short-term follow-up shows 60–85% of patients so treated experience significant relief (Hofstetter and Shah 2015).

44

D. S. Rush and E. J. Wilkinson

Pigment Disorders and Benign Melanocytic Lesions Physiologic hyperpigmentation of the vulvar skin, especially of the perineal body, perianal area, and the tips of the lateral labia minora and posterior vestibule, is usual in adult women. It is thought to occur as the result of hormonal influences and may be more prominent in women with naturally dark complexions. Darkening of the vulvar skin is not, therefore, in itself a cause of concern or an indication for biopsy. Hypopigmentation of the vulva is much less common but may occur in patients with vitiligo and albinism, or sometimes as postinflammatory change (see below), conditions which rarely necessitate biopsy. However, pigmented vulvar lesions, which will appear as well-demarcated, focal areas of discoloration, are estimated to occur in 10–12% of women (Murzaku et al. 2014)), and these often do require histologic examination for diagnosis. In fact, biopsy is more frequently necessary for the evaluation of localized pigmented lesions on the vulva than for those on the extragenital skin, as the gross appearance of vulvar pigmented lesions is not as reliable an indication of the diagnosis and expected behavior as in other skin sites.

Fig. 34 Lentigo simplex. Increased basilar pigmentation is evident, with abundant melanophages in the superficial dermis

such as LSC, LS, and LP. Histologic examination will show pigment-laden macrophages in the dermis or submucosa, as well as extracellularly (pigment incontinence), in addition to chronic changes related to the associated disorder.

Postinflammatory Alterations in Pigmentation Lentigo Simplex Both hypo- and hyperpigmentation may occur as postinflammatory changes in healing skin. Biopsy is rarely necessary, particularly in the typical clinical setting. In areas of previous ulceration, recently healed skin will temporarily lack a normal population of melanocytes, a condition referred to as postinflammatory depigmentation, or leukoderma. This disorder is common after herpes infection, syphilitic ulceration, burns, and deep laser or cryotherapy. Histologically, the skin appears thinned and lacks the usual amount of pigment, but on careful microscopic inspection, some melanin can usually be identified. Postinflammatory hyperpigmentation is a common finding, especially in patients with chronic inflammatory conditions

Lentigo simplex, a benign melanocytic proliferation, may occur on the mucous membranes as well as on the skin. It is relatively common in the vulva. The lesion is typically small, 4 mm or less in diameter, flat, and uniformly pigmented. Histologically, lentigo simplex is a localized circumscribed area of slightly hyperplastic epidermis that contains an increased number of normalappearing melanocytes along the side and tips of the elongated rete ridges associated with basilar hyperpigmentation (Fig. 34). Extreme degrees of epidermal pigmentation may be present, with numerous squamous cells exhibiting cytoplasmic melanin granules, usually in highest concentration

1

Benign Diseases of the Vulva

near the epithelial–stromal junction. There may be mild acanthosis and slight clubbing of the rete ridges, and heavily pigmented melanophages may be present in the upper dermis. At times, a minimal superficial dermal inflammatory cell infiltrate is noted, but this is by no means constant.

Vulvar Melanosis Vulvar melanosis, the most common pigmented lesion in women of reproductive age (Murzaku et al. 2014), is characterized by prominent brown to black pigmented macular areas with irregular borders that may be solitary or multiple, ranging from a few millimeters to several centimeters in size. The labia minora are the most common site of involvement (Murzaku et al. 2014), and the hair-bearing portions of the vulva are not affected. The cause is unknown; it is speculated that it may be a form of postinflammatory hyperpigmentation or a defect in melanin transport (Oliveira et al. 2011). Grossly, the lesions cannot be reliably distinguished from melanoma, and biopsy is imperative. Histologic examination shows a normal or slightly increased number of melanocytes but with increased dendritic processes and an increased amount of pigment within them, especially at the tips of the rete ridges. There is usually also elongation of the rete ridges and increased dermal melanophages. Unless the lesion is of significant size, a feature favoring melanosis over lentigo simplex, distinction between the two entities may be extremely difficult, both on clinical examination and histologically. The distinction lies in the fact that the melanocytes increase in number in lentigines but not in melanosis, a feature which is not always easily discernable and is somewhat subjective. Of greater importance is to distinguish the lesion from melanoma, which is, fortunately, usually considerably easier, as in melanoma, there is an even greater degree of melanocytic proliferation, in addition to cellular atypia, pagetoid upward spread of the pigmented cells, and an infiltrative growth pattern, none of which should be present in melanosis or lentigo.

45

Common Nevi Vulvar melanocytic nevi, like those elsewhere on the skin, may be junctional, compound, or intradermal. Clinically, nevi are usually well-defined, papular, uniformly pigmented, and typically less than 10 mm in diameter (Rock et al. 1990). They are most often found on the labia majora, labia minora, and the clitoral hood (Murzaku et al. 2014). These nevi are histologically identical to those found on the non-genital skin. Most nevi biopsied in the vulva are either compound or intradermal in type. In pure junctional nevi, which are identified relatively infrequently on the vulva, the nevus cells are located within the epidermis and at the dermal–epidermal junction. Histologic examination shows a proliferation of nevus cells, which are somewhat larger than melanocytes and have round or ovoid nuclei. The cytoplasm may contain melanin or appear clear without granules or fibrils. Dendrites are not present, and intercellular connections are not visible. The cells may lie singly within the dermis, but more commonly they tend to form nests. Individual cells, or cell nests, bulge downward from the tips of the rete ridges. In compound nevi, the nevus cells are located in both the epidermis and the dermis (Fig. 35). The basement membrane of the epidermis surrounding the nests disappears, and collagen and elastic fibers surround

Fig. 35 Compound nevus. Nests of nevus cells, many showing heavy pigmentation, are present at the dermoepidermal junction and within the dermis

46

D. S. Rush and E. J. Wilkinson

Atypical Genital Nevi

Fig. 36 Intradermal nevus. The nevus cell nests are completely enclosed within the dermis, with no activity at the dermo-epidermal junction

the nests, pushing the epidermis upward, accounting for the fact that the lesion is clinically elevated above the level of the surrounding skin. In intradermal nevi, the nevus cells and nests are entirely located within the connective tissue of the dermis (Fig. 36); no activity is seen at the dermo-epidermal junction. Common vulvar nevi do not demonstrate cellular atypia or architectural distortion and show a characteristic pattern of zonal maturation, with the cells becoming more spindled and neural-like with deeper penetration into the dermis. Melanocytic nevi in the setting of LS may display unusual and concerning histologic features worthy of special mention (Edwards 2010; Carlson et al. 2002). Grossly, such lesions are usually very dark and rather small. Histologically, the rete ridges tend to be elongated, with clusters of melanocytes present at the dermoepidermal junction. Melanocytes trapped in sclerosis may show mild cytologic atypia, and focal pagetoid upward spread of the nests and individual cells may be seen, mimicking melanoma. A lichenoid inflammatory infiltrate disrupting dermal nests and increased melanophages with pigment incontinence may also be seen, mimicking changes of regressing melanoma. Follow-up of these nevi has shown no evidence of malignant behavior, and recognition that these changes may occur in the setting of lichen sclerosus is important to prevent overcalling such lesions as melanoma.

Atypical genital nevi show a peculiar morphology with features in common with dysplastic nevi but without any associated risk for developing melanoma. As the name suggests, they are an unusual pigmented lesion occurring only on the genital skin, and they have distinctive histologic features not seen in nevi in other locations. About 5% of vulvar nevi fall into this category (Murzaku et al. 2014). Patients are usually young women, with the reported median age ranging from 21 to 26 years old (Gleason et al. 2008; Ribe 2008). They occur more often on the labia minora than do common nevi and, especially in children, are prone to develop in mucosal sites (Murzaku et al. 2014). The lesion may be concerning on clinical exam due to the large size, irregular borders, dark pigmentation, or any combination of these features. On microscopic examination, the lesions usually have a component of a typical common intradermal nevus, which is useful in the diagnosis, but numerous unusual features are also present. The nests of nevus cells are often much larger than those of common nevi (Fig. 37). There may be marked junctional activity, sometimes with a continuous band of nests between the epidermis and the dermis, which can obscure the dermo-epidermal junction (Gleason et al. 2008; Brenn 2011). Often there is cytologic atypia. Other worrisome features, including focal pagetoid spread of single nevus cells, adnexal involvement, nests along elongated rete ridges, and dense fibrosis of the papillary dermis, may be seen (Gleason et al. 2008; Ribe 2008; Murzaku et al. 2014; Brenn 2011). These lesions were once thought to be malignant and are still frequently misdiagnosed as such due to these features, but it has since become clear that these lesions are entirely benign and have virtually no risk of malignant transformation. Characteristics that help to distinguish atypical genital nevus from melanoma include the presence of dermal maturation, the sparsity of mitotic activity, and the absence of necrosis or ulceration (Murzaku et al. 2014).

1

Benign Diseases of the Vulva

47

and patients with multiple dysplastic nevi are at higher risk for development of melanoma generally, necessitating careful, lifelong dermatologic surveillance.

Benign Squamous Proliferations Fibroepithelial Polyp (Acrochordon)

Fig. 37 Atypical genital nevus. An unusually large nevus cell nest with cellular atypia is present at the dermoepidermal junction

Dysplastic Nevi Dysplastic melanocytic nevi are seen most often in young women of reproductive age and may show overlapping gross and microscopic features with atypical genital nevi but are fortunately very rare on the vulva. They present as pigmented, elevated lesions greater than 0.5 cm in diameter with irregular borders. Microscopic examination reveals elongation and bridging of the rete ridges, with nests of large epithelioid or spindle-shaped nevus cells with nuclear pleomorphism and prominent nucleoli. The nests extend into adnexal structures, including hair shafts and the ducts of sweat glands. The low-power impression is frequently that of a large junctional nevus, with a dermal component that has spindle- or epithelioid-type nevus cells in nests or isolated within the papillary and reticular dermis. Lamellar fibrosis of the upper dermis is a characteristic feature. Dysplastic nevi have a variably increased risk of transformation to malignant melanoma,

The fibroepithelial polyp, or “skin tag,” is a relatively uncommon benign polypoid tumor of the vulva. Fibroepithelial polyp occurs on the keratinized skin of the vulva and vary in appearance from small, flesh-colored, or hyperpigmented papillomatous growths resembling condylomata to large pedunculated tumors that often are hypopigmented. On cut section, fibroepithelial polyps are soft and fleshy. Small tumors may resemble intradermal nevi; large lesions may present cosmetic problems but generally are clinically insignificant. They usually arise in hair-bearing skin but may be found on the labia minora. Histologically, fibroepithelial polyps may be of two types, one that is predominantly epithelial and another that is primarily stromal. The epithelial surface varies from a thickened layer with papillomatosis and hyperkeratosis to an attenuated flattened layer exhibiting multiple primary folds. The connective tissue stalk is composed of loose bundles of collagen with a moderate number of blood vessels. The stroma may be edematous and hypocellular. The stromal cells usually have relatively uniform nuclei; however, marked atypia may be seen in some cases (Carter et al. 1992). They are distinguished from condyloma by the lack of cellular atypia (Fig. 38).

Vestibular Papilloma Vestibular papilloma presents as a small papillary lesion, usually less than 5 mm in length, with a diameter of 1–2 mm, which, unlike condyloma, does not appear white upon application of acetic acid. In contrast to a fibroepithelial polyp, vestibular

48

D. S. Rush and E. J. Wilkinson

Fig. 38 Fibroepithelial polyp. A fibrovascular stalk is covered by a lining of keratinized squamous epithelium showing no abnormalities of maturation or cellular atypia

Fig. 39 (a) Seborrheic keratosis. The thickened epithelium shows several horn cysts. Cytologic atypia is absent. (b) A squamous “eddy” is shown at the center of this rete ridge

papilloma typically occurs on the nonkeratinizing squamous mucosa of the vestibule and may be single or multiple. When multiple, the diagnosis is vestibular papillomatosis, a condition found in up to 33% of normal women of reproductive age (Hoang et al. 2015b). Single lesions may be asymptomatic, but vulvar papillomatosis is usually associated with pruritus, pain, and burning (Chan and Chiu 2008; Sangueza and Saenz 2007). The etiology of vestibular papillomatosis is unclear. On microscopic examination, squamous papilloma is composed of a delicate fibrovascular connective tissue core covered by nonkeratinized squamous epithelium, which is glycogen-rich in women of reproductive age. They may have a thin keratin layer. Inflammation is rarely present. Distinction from condyloma acuminatum is based on the lack of cellular atypia.

Seborrheic Keratosis Seborrheic keratosis is a benign exophytic lesion of the skin. Common on the sun-exposed skin of the elderly, they are less commonly seen on the vulva, where they may occur on the hair-bearing skin of the labia majora or mons. The lesions are frequently pigmented and well-demarcated from the surrounding skin, often described as appearing to be “stuck on” the surface, giving the impression they could be easily dislodged. On microscopic examination, there is prominent acanthosis and pronounced surface hyperkeratosis. Pigmentation, if present, is usually evident in the basal and parabasal cells. Follicular plugging with hyperkeratosis is a typical feature, as is the formation of “horn cysts,” accumulation of hyperkeratotic scale in intraepithelial spaces formed by

1

Benign Diseases of the Vulva

invaginations of the surface epithelium (Fig. 39a). “Squamous eddies,” clusters of squamous cells which give the impression of swirling currents, are another characteristic feature (Fig. 39b). The papillary architecture and hyperkeratosis of seborrheic keratosis can be suggestive of condyloma, particularly at low power. In fact, as seborrheic keratoses are frequently found to contain HPV DNA (Gushi et al. 2003; Bai et al. 2003), some authors have maintained that these lesions are variants of condyloma (Li and Ackerman 1994), although this remains controversial. The absence of koilocytic atypia in seborrheic keratosis and of keratin horn cysts in condylomata will usually resolve the diagnosis. The acanthosis and irregularity of the epithelium may also suggest a squamous intraepithelial lesion, but the lack of significant nuclear atypia or mitotic activity in seborrheic keratosis should distinguish it. The lesions may be treated with topical agents, destructive treatments, or excision.

Keratoacanthoma Keratoacanthoma presents as a rapidly growing, firm, dome-shaped lesion with central umbilication. Like seborrheic keratosis, it is common on the sun-exposed skin of the elderly but rarely reported on the vulva (Chen and Koenig 2004; Gilbey et al. 1997; Ozcan et al. 2006; Nascimento et al. 2005). On microscopic examination, the lesions show an endophytic growth pattern, with a broad, pushing border. A collarette of the epidermis surrounding a central keratin-filled crater corresponds to the central umbilication seen grossly. Toward the center of the crater, the squamous cells are larger, containing abundant amounts of eosinophilic glassy cytoplasm. During the early stage of rapid growth, there may be mild to moderate nuclear atypia, and mitotic figures may be numerous. As the lesion matures, however, these features regress, with only minimal atypia remaining in a well-developed lesion. A dense inflammatory infiltrate is usually present at the base of the lesion. Occasionally, small, irregular nests of cells may be present focally adjacent to the pushing border. Less commonly, perineural or intravascular invasion may be

49

apparent. These features have led many to believe that keratoacanthoma may actually be a low-grade variant of squamous cell carcinoma (Karraa and Khachemoune 2007; Schwartz 2004; Weedon et al. 2010), but in the absence of other features, particularly on the vulva, where malignant behavior has never been reported, a more conservative interpretation is advisable. Untreated, keratoacanthoma typically increases rapidly in size over a period of weeks to months and then may persist for months until finally undergoing spontaneous involution, usually within 6 months of eruption. Alternatively, excision may be performed and is usually curative, with rare recurrences reported.

Cysts Epithelial Inclusion Cyst Epithelial inclusion cysts frequently develop on the vulva. They may arise from occluded sebaceous glands that have undergone squamous metaplasia or as a complication from previous surgical intervention. They may develop anywhere on the vulva but are more common on the labia majora and clitoris. Epithelial inclusion cysts of the clitoris may be pedunculated and become quite large, with lesions measuring 8 cm or more (Al-Ojaimi and Abdulla 2012), causing them to be mistaken clinically for malignancy. Such lesions are more commonly seen as sequelae of female genital mutilation procedures, usually developing in adolescence (Asante et al. 2010; Riszk et al. 2007). It is believed that the hormonal milieu may play some role in the growth of the lesions at this time. The phenomenon is not limited to patients with a history of genital mutilation, however, as numerous similar cases have also been reported in women and girls without such a history, in patients ranging from very young children to the elderly (AndersonMueller et al. 2009; Cetinkursun et al. 2009; Paulus et al. 2010; Schober et al. 2014). On gross examination, excised lesions are found to contain thick white to yellow, grumous or cheesy material. On microscopic examination,

50

Fig. 40 Epidermal inclusion cyst. The lining of the cyst is thinned, and abundant keratotic debris fills the cystic cavity

the cyst lining appears as a relatively flattened, stratified squamous epithelium lacking adnexal structures (Fig. 40). Foreign body-type giant cells may be seen in the tissue adjacent to the cyst wall, a response to keratinous material leaking into the adjacent dermis. Epithelial inclusion cysts are benign lesions, although rare cases of squamous cell carcinoma may arise within them. Treatment of asymptomatic small cysts usually is not necessary; however, surgical excision may be necessary for diagnosis or if the cyst is enlarging, symptomatic, or secondarily infected.

Bartholin Cyst and Abscess The estimated lifetime risk for the development of a cyst or abscess of the Bartholin gland is 2% (Marzana and Haefner 2004), with abscess being approximately three times more common than cysts (Lee et al. 2014). The lesions develop as a result of obstruction of the vestibular orifice of the Bartholin ducts, leading to subsequent accumulation of secretions and associated cystic dilation of the duct. The normal Bartholin duct contains three types of epithelium: mucinous columnar epithelium distally, transitional epithelium centrally, and squamous epithelium toward the orifice. This accounts for the variable appearance of the epithelial lining of a Bartholin gland cyst, in which any of the three

D. S. Rush and E. J. Wilkinson

Fig. 41 Bartholin gland cyst. The lining of this cyst shows a mixture of mucinous, transitional, and squamous epithelium, mirroring the normal structure of the Bartholin gland duct

types may appear, often in combination (Fig. 41). A case of melanocytic colonization, in which melanocytes with cytoplasmic pigment were found in the cyst lining, has been reported (Nigam et al. 2017), but pigmentation is not a usual finding. As in other cystic lesions, in some cases the lining may be flattened and not classifiable. If the cyst is not infected, there is minimal, if any, inflammatory response within the adjacent tissue. Secondary infection of the cyst is common, however, in which case the lesion is classified as a Bartholin gland abscess, and associated inflammation may be a prominent feature. Marsupialization, drainage, and antibiotics, if infected, are the initial treatments of choice. Lesions which are not excised may recur, necessitating additional treatment and surgical excision may be performed for definitive treatment and to preclude further recurrence.

Mucous Cyst Mucous cysts of the vulva are most commonly seen in the vestibule or labia minora (Heller 2015). Their origin is unclear and may be different for different sites. It has been proposed that at least some lesions develop from occlusion of minor vestibular glands, but other proposed origins include Mullerian heterotopia or metaplasia,

1

Benign Diseases of the Vulva

Wolffian or urogenital sinus remnants, and the Bartholin gland (Heller 2015). The cysts are lined by a single layer of mucinsecreting cuboidal to columnar epithelium resembling the endocervix. Squamous metaplasia may be present. Special stains may distinguish these cysts from those of mesonephric origin, as the epithelium stains with both Alcian blue and Mayer mucicarmine, whereas the epithelial lining of cysts of mesonephric origin does not (Newland and Fusaro 1991), although this distinction is not critical, as neither lesion is of particular clinical significance. Excision is usually performed for diagnostic purposes or to relieve associated symptoms and is curative.

Ciliated Cysts Cysts lined with ciliated epithelium of tuboendometrioid type may also occur on the vulva. They are usually found in the vulvar vestibule and labia minora. Patients are typically between 25 and 35 years old and present with cysts measuring 1–3 cm (Kuniyuki et al. 2008). The proposed origin of the cysts is heterotopic Mullerian epithelium, a concept which is supported by the finding of estrogen and progesterone receptor expression in them (Kuniyuki et al. 2008). They are distinguished from endometriosis by the absence of associated endometrial stroma or hemosiderinladen macrophages. As with mucinous cysts, excision is not necessary, unless the lesion is symptomatic, but is usually performed to establish a diagnosis.

Gartner Duct Cyst Gartner duct cysts, presumably derived from remnants of the mesonephric (Wolffian) ducts, are encountered occasionally on the lateral aspects of the vulva and the vagina. They are thin walled, translucent, and contain clear fluid. The lining epithelium is cuboidal to columnar and is not ciliated, and they are filled with eosinophilic secretions.

51

Mammary-Like Cysts (Hidrocystoma) Hidrocystoma originates in the anogenital mammary-like glands and is found in the distribution of those glands, predominantly in the interlabial sulcus. It is characterized by a peripheral basal layer of myoepithelial cells and a luminal layer of cuboidal to columnar cells showing apical decapitation secretion, typical of apocrine glands. Anogenital mammary-like glands and dartos muscles may be identified in the surrounding tissue.

Cysts of the Canal of Nuck The canal of Nuck is formed by the peritoneum of the processus vaginalis which accompanies the round ligament through the inguinal canal to its insertion in the deep vulvar tissue. Cysts of the canal of Nuck are generally found in the superior aspect of the labia majora or inguinal canal and are believed to arise from inclusions of the peritoneum at the inferior insertion of the round ligament into the labia majora. As such, they are analogous to the hydrocele of the spermatic cord. These cysts can achieve substantial size and must be distinguished from an inguinal hernia, with which they are associated in approximately one-third of cases (Schneider et al. 1994). Microscopic examination reveals a cyst lining of flattened mesothelial cells.

Skene Gland Cyst Cysts derived from the paraurethral Skene’s glands are most common in women of reproductive age (Heller 2015). Though often asymptomatic, they may present with a sizeable mass or associated pain, discharge, dysuria, or urinary obstruction. Histologically, the cysts are lined by transitional or squamous epithelium.

Benign Glandular Lesions of the Vulva The various glands of the vulvar region may all develop hyperplastic or neoplastic mass lesions. Occasionally, also, glandular epithelium may be

52

found to replace the squamous epithelium, producing a sharply demarcated, red velvety area (Coghill et al. 1990; Horn et al. 2014). The phenomenon is not well-understood and can lead to significant diagnostic difficulty. Most patients are postmenopausal at the time of diagnosis (Horn et al. 2014). Vaginal adenosis with extension to the vestibule may explain some cases, but lesions confined to the vestibule or other vulvar sites at the time of presentation require a different explanation. The presence of mucinous epithelium on the vulva has been reported in association with re-epithelialization following Stevens–Johnson syndrome, topical medications like 5-fluorouracil, or laser treatment, and some cases of plasma cell vulvitis (Coghill et al. 1990; Heller 2015; Marquette et al. 1985), but other patients have no history of any sort of predisposing factor. Possible explanations which have been proposed include metaplastic changes or embryonic displacement of stem cells (Horn et al. 2014). Very rarely, glandular epithelium with an intestinal phenotype is identified on the vulva, possibly of cloacogenic origin (Horn et al. 2014), and this may be the origin of rare tubulovillous adenomas and gastro-intestinal-like carcinomas reported arising on the vulva (Vitrey et al. 2003; Fox et al. 1988).

Lesions of the Anogenital MammaryLike Glands The presence of mammary-like tissue in the vulva was once widely held to be ectopic in nature, representing caudal remnants of the embryonic “milk line” thought to extend from the axilla and pectoral region through the abdomen and pelvis, terminating in the vulvar region. Subsequent work has cast doubt on this theory, and it is now wellestablished that mammary-like tissue found in the vulva is derived from the anogenital mammarylike glands (El-Khoury et al. 2016; Kurashiga et al. 2014; Konstantinova et al. 2017a, b; van der Putte 1991; van der Putte 1994). These glands are morphologically and immunohistochemically remarkably similar to breast tissue except for the lack of a lobular architecture and a greater frequency of columnar cell change and columnar cell

D. S. Rush and E. J. Wilkinson

hyperplasia. Normally, the glands are small and not clinically visible, but they are the source of a number of proliferative lesions, most often benign, which produce palpable subcutaneous masses. By far the most common lesion of the anogenital mammary-like glands is hidradenoma papilliferum, which is also the most common benign glandular lesion of the vulva (Baker et al. 2013). Patients with hidradenoma papilliferum range in age from 25 to 90 years in age, with an average age of 49–52 years (Konstantinova et al. 2016; Scurry et al. 2009). The lesions are found in the distribution of the anogenital mammary-like glands, with the majority arising in the interlabial sulcus and involving the labia majora or minora. They may be cystic, solid, or pedunculated and are usually asymptomatic, although some patients may present with bleeding, pruritus, or complaints of an enlarging mass. Only rarely are the lesions painful. Grossly, the lesions present as an elevated area with a palpable subcutaneous mass and may appear red, blue, or skin-colored (Scurry et al. 2009). All but one of the reported cases have been solitary (Konstantinova et al. 2016) and usually small, ranging in size from 1 to 20 mm in greatest dimension with a mean size of 6–7 mm (Konstantinova et al. 2016; Scurry et al. 2009). Rare cases may exceed 2 cm in size (Konstantinova et al. 2016). Microscopically, there is considerable heterogeneity in the appearance of these tumors, though all invariably consist of a mixture of papillary and tubular structures (Fig. 42). The tubules and papillary structures are lined by predominantly cuboidal to columnar epithelial cells surrounded by myoepithelial cells, often reminiscent of intraductal papilloma of the breast (Fig. 43). The myoepithelial cells are most often flattened and inconspicuous and may require immunohistochemical stains such as smooth muscle actin (SMA), S100, calponin, or p63 to distinguish them (Fig. 44). Less commonly, the myoepithelial cells are rounded and demonstrate clear cytoplasm. Detection of the myoepithelial cell layer, with immunohistochemical confirmation, if necessary, is critical to the diagnosis and important to distinguish it from malignant glandular entities. The appearance of the epithelial cells can be quite variable,

1

Benign Diseases of the Vulva

Fig. 42 Hidradenoma papilliferum. At low magnification, the lesion is seen as a subcutaneous mass with a tubulopapillary architecture

Fig. 43 Hidradenoma papilliferum. A high-power magnification of the cells lining the tubules and papillae shows a distinct two layers: an outer myoepithelial layer and an inner layer of columnar cells with apical “snouts”

showing a spectrum of changes analogous to fibrocystic change of the breast, but typically there is at least some part of the lesion showing the typical apocrine features of columnar cells with eosinophilic cytoplasm and prominent apical snouts. Oxyphilic (apocrine) metaplasia is very common, and less commonly the epithelium may show clear or foamy cells. Squamous metaplasia and mucinous metaplasia have also been reported but are not common findings (Konstantinova et al. 2016; Scurry et al. 2009). Architectural features are similarly variable, with areas of solid or streaming growth of the epithelium, intraluminal tufting, arching, or micropapillary formations occurring admixed

53

Fig. 44 Hidradenoma papilliferum. Immunohistochemical staining for p63 highlights the myoepithelial cell layer

with the papillary fronds and tubules. Mitoses may be seen in both the epithelial and myoepithelial cell components and may be as high as 13 per 10 high-power fields, but even such high mitotic counts have no impact on clinical behavior (Sington et al. 2006), which is always benign, with no reported cases of recurrence or metastasis. Many other benign glandular lesions may arise in the anogenital mammary-like glands, most of which closely resemble lesions of the breast, and are termed accordingly. These include lactating adenoma, fibroadenoma, phyllodes tumor, pseudoangiomatous stromal hyperplasia, tubular adenoma, and erosive adenomatosis (Kazakov et al. 2011; Konstantinova et al. 2009; Scurry et al. 2009). Lesions diagnosed as syringocystadenoma papilliferum are presumably derived from the anogenital mammary-like glands as well (Steshenko et al. 2014).

Lesions of Sweat Gland Origin Occasional cases of nodular hidradenoma, eccrine hidrocystoma, poroma, spiradenoma, and syringoma have also been reported on the vulva (Baker et al. 2013; Dereli et al. 2007; Kazakov et al. 2010a; Ozcan 2009; Hoang and Kazakov 2015). They may arise from the anogenital mammary-like glands or other eccrine and apocrine sweat glands in the region,

54

Fig. 45 Syringoma. Comma-shaped glandular structures are seen in the fibrotic dermis

and the clinical and histologic features are identical to those of cases presenting on the nonvulvar skin. Syringoma, a benign tumor of eccrine origin, is the most common of these lesions and reportedly the second most common adnexal lesion of the vulva (Baker et al. 2013). Syringoma presents as multiple papules on the labia majora, often with associated pruritis. Histologic examination reveals a small, wellcircumscribed lesion in the upper dermis, consisting of small solid nests and ductal structures, often with a characteristic comma or tadpole shape, embedded in a collagenous stroma (Fig. 45).

Solid Lesions of Bartholin Gland Origin The most common benign solid lesion of the Bartholin gland is nodular hyperplasia (Kazakov et al. 2007). The lesions are usually asymptomatic, but may be associated with mild pain, and present as a firm, sometimes lobulated mass measuring from 2 to 4 cm, which appears gray on cut sections. Histologic examination demonstrates an irregular but not encapsulated proliferation of glandular acini composed of cuboidal or columnar cells with abundant mucin-filled cytoplasm, which maintain a normal acinar-duct relationship. Nuclei are basally located and without atypia. Associated dilated ducts filled with inspissated secretions form a lesser component of the lesion.

D. S. Rush and E. J. Wilkinson

Less common than nodular hyperplasia is the Bartholin gland adenoma. The distinction between nodular hyperplasia and adenoma of the Bartholin gland is not well-defined or uniformly accepted. In fact, the diagnosis of Bartholin gland adenoma itself is not universally accepted, and it remains a question whether such lesions are better considered as hamartomas than neoplasias (Heller and Bean 2014). Compounding the issue is the finding that one case of nodular hyperplasia of the Bartholin gland has been found to be monoclonal (Kazakov et al. 2007). The proposed distinction is that an adenoma is encapsulated, sharply circumscribed, and lacks the normal acinar-duct relationship. Rare cases of pleomorphic adenoma (mixed tumor of the vulva) have been reported arising in the Bartholin gland (Heller and Bean 2014), as have malignant tumors of salivary gland type suggesting that reports of the finding of salivary gland-like tissue in the vulva (Marwah and Bergman 1980) may also be derived from the Bartholin gland and are the source of such tumors.

Prostate-Like Tissue and Solid Lesions of Skene’s Gland Origin Prostatic-type tissue has been described in the vulva, typically as an incidental finding presenting as a lobular arrangement of glands and nests located in the superficial dermis (Kazakov et al. 2010b; Kelly et al. 2011). The tissue stains positively for markers of prostatic differentiation on immunohistochemistry. The origin of this tissue is not entirely clear, but recently “misplaced” Skene’s glands have been proposed as a possible explanation (Kelly et al. 2011). Eutopic Skene’s glands may also develop adenomatous hyperplasia, leading to nodular masses of prostatic-type tissue in the paraurethral area (Kazakov et al. 2010b).

Adenoma of Minor Vestibular Glands Adenoma of minor vestibular glands is a rare benign tumor. The lesions are small, ranging from 1–2 mm to 1 cm and composed of multiple

1

Benign Diseases of the Vulva

55

lobular clusters of small glands lined by mucinsecreting columnar epithelium. Most cases have been found incidentally in vestibulectomy specimens excised for vulvar vestibulitis (Prayson et al. 1995)

Endometriosis Endometriosis, in which endometrial glands and stroma are present in areas outside of the endometrial cavity, is uncommon on the vulva. Most reported cases have developed in the area of a previous episiotomy (Heller 2015; Li et al. 2015). In such cases, it is presumed that the endometrial tissue became implanted in the incision at or near the time of delivery. The disease has also been reported in the areas of vulvar incisions for other procedures (Buda et al. 2008), and rare cases of spontaneous perineal endometriosis with no previous surgical therapy have been reported (Nasu et al. 2013). In one unusual case, bilateral endometriomas developed in the Bartholin glands with no history of prior intervention (Aydin et al. 2011). Endometriosis involving the vulva, while unusual, looks no different histologically than endometriosis anywhere else, and the observation of benign-appearing endometrial glands and stroma is diagnostic. Associated fibrosis, hemosiderin-laden macrophages, and lymphocytic inflammation are commonly present. Excision with a wide margin is usually curative.

Benign Lesions of Folliculo-Sebaceous Origin Benign lesions of follicular origin are rare on the vulva and necessarily identified only on the hairbearing skin of the labia majora and mons. Trichoepithelioma, trichofolliculoma, pilomatricoma, cylindroma, and warty dyskeratoma have all been reported on the vulva (Baker et al. 2013; Heller et al. 2009; Lora et al. 2015). Their appearance and behavior do not differ from that of such lesions identified elsewhere on the body.

Fig. 46 Capillary hemangioma. The dermis is filled by lobules of capillaries, while the overlying epithelium is uninvolved

Benign sebaceous lesions are even more rare on the vulva. Sebaceous glands encountered in the vulvar tissue are most likely to represent Fordyce spots; sebaceous hyperplasia is reportedly extremely rare on the vulva (Baker et al. 2013), and sebaceoma is also only rarely reported (Baker et al. 2013).

Benign Lymphovascular Lesions Hemangiomas Cavernous hemangiomas are the most common type of hemangioma on the vulva (Hoang and Sangueza 2015). Small capillary hemangiomas are also reported to be common on the vulva, especially in elderly patients (Heller 2015). Arteriovenous hemangiomas occur less commonly. All of these lesions present as wellcircumscribed collections of blood-filled vessels involving the reticular dermis (Fig. 46), morphologically identical to those occurring in other body sites. Pyogenic granuloma, or lobular capillary hemangioma, although common in the vagina, is very rare on the vulva (Abreu-dos-Santos et al. 2016), and those that do arise are usually associated with

56

D. S. Rush and E. J. Wilkinson

Fig. 47 Pyogenic granuloma. This hemangioma variant is characterized by inflammation in the loose connective tissue between the proliferation of capillaries

Fig. 48 Angiokeratoma. Strands of squamous epithelium surround endothelial-lined vascular spaces containing red blood cells

pregnancy. Histologically, a thin ulcerated epidermis is noted covering a mass of granulation tissue. Capillaries are numerous, and secondary inflammatory changes frequently are found within the stroma (Fig. 47). Around the periphery of the lesion, there may be a downward growth of the epidermis producing a “collarette.” The overall architecture appears lobulated with a few thickened fibrous septae separating the vascular proliferation.

papillomatosis are present, along with a mild inflammatory reaction in the deep dermis.

Angiokeratoma Angiokeratoma is a variant of hemangioma quite common on the vulva. The usual site of involvement is the labia majora, where the lesion appears as a red, purple, or black, slightly elevated area. They may be solitary or multiple and unilateral or bilateral (Nomelini et al. 2010; Yigiter et al. 2008). They are rarely symptomatic but may result in bleeding, pain, or pruritus. Most patients are under 50 years of age. Their peculiar appearance often prompts excisional diagnostic biopsy, often for concern for melanoma. Histologic examination reveals endotheliallined and blood-filled dilated channels, separated by strands and cords of squamous epithelial cells representing downgrowth from the overlying epithelium, which is often hyperkeratotic (Fig. 48). Varying degrees of acanthosis and

Lymphangioma Circumscriptum Almost all lesions of the lymphatic derivation on the vulva are acquired, secondary to obstructed lymphatic flow which may result from chronic inflammatory conditions, surgery, radiation treatment, trauma, infection, or obesity (Chang et al. 2016; Heller 2015; Lawrance et al. 2015; Zhu et al. 2014). Unlike in other areas of the body, acquired lymphangioma circumscriptum of the vulva tends to be markedly exophytic and hyperplastic and may grossly suggest warts or even carcinoma (Lawrance et al. 2015). The surface of the lesion is usually covered in small oozing vesicles, and secondary infection may result in ulceration, pain, and cellulitis, which may confound the underlying diagnosis. Microscopic examination reveals dilated, tortuous thin-walled lymphatic vessels in the papillary and reticular dermis (Fig. 49). The overlying epithelium usually is unremarkable but may be eroded or hyperkeratotic. Treatment includes surgical excision, laser therapy, cryotherapy, electrocautery, or radiotherapy (Chang et al. 2016), but eradication may be difficult, especially with persistence of the underlying cause.

1

Benign Diseases of the Vulva

57

Fig. 50 Calcinosis. A well-circumscribed nodule of calcification is present in the superficial dermis

Fig. 49 Lymphangioma. Dilated lymphatic channels are prominent in the superficial dermis

Histologically, the lesion shows acanthosis and parakeratosis of the epithelial layer, with elongated rete ridges and the diagnostic finding of collections of foamy histiocytes (xanthoma cells) in the papillary dermis. There may be a varying degree of acute or chronic inflammation in the dermis as well. The lesion does not respond to topical steroids or topical wart treatments and may recur if incompletely removed.

Miscellaneous Tumor-Like Lesions Verruciform Xanthoma

Idiopathic Vulvar Calcinosis

Verruciform xanthoma is an uncommon benign lesion of mucosal surfaces occasionally reported on the vulva. Clinically, the lesion may be concerning for condyloma or squamous carcinoma. In reported cases, most patients have been postmenopausal, and most have been single lesions involving the labia minora or majora, clitoris, or fourchette (Fite et al. 2011; Frankel et al. 2011). Grossly the lesions present as asymptomatic, slow-growing, sharply demarcated plaques with a verrucous surface and a yellowish-orange hue, with reported dimensions ranging from 2 to 20 mm (Fite et al. 2011). Most patients have had associated disorders of the vulvar epithelium, usually LS, but also LP, Paget disease, and radiodermatitis. It is hypothesized that the lesions may be a reactive condition developing in response to damage at the dermo-epidermal junction (Fite et al. 2011).

Vulvar calcinosis is a rare benign condition which presents as firm, bilateral, subcutaneous nodules measuring 2–5 mm, involving the labia majora or fourchette, which may develop cystic change progressing to draining sinuses. Histologic examination demonstrates normalappearing overlying epithelium with basophilic acellular superficial subcutaneous nodules measuring from less than 0.1 mm to approximately 2 mm, sometimes associated with a chronic inflammatory infiltrate, mast cells, and foreign body giant cells (Fig. 50). The acellular material stains with Von Kossa stain and contains acid mucopolysaccharides (Balfour and Vincenti 1991). The etiology is unclear; the disease has been reported in patients with abnormal calcium metabolism and without (Biswas et al. 2007; Tomazzini et al. 2008), and never with evidence of calcium deposition in other sites.

58

Vulvar Amyloidosis Localized vulvar amyloidosis is extremely rare and almost always occurs in association with a HSIL. The amyloid deposits have been found to consist mainly of cytokeratins (Quddus et al. 2014).

Benign Lesions of the Urethra

D. S. Rush and E. J. Wilkinson

common finding and helps to distinguish a caruncle from urethral prolapse, which is otherwise very similar on gross and microscopic appearance. The stroma contains a variable amount of chronic inflammation and may be fibrotic or edematous, containing prominent blood vessels which are frequently dilated and congested. The etiology is unknown; like prolapse, estrogen deficiency is believed to play a role. Other possible factors include irritation, trauma, and local congestion (Conces et al. 2012).

Urethral Prolapse Prolapse of the urethral mucosa may occur at any age, but it is most common in premenarchal children and in postmenopausal women. Redundancy of the mucosa and laxity of the supporting periurethral fascia contribute to the formation of prolapse, which is aggravated by increased abdominal pressure; it may be related to relative lack of estrogen. The prolapsed urethra may present as a large red polypoid mass covered with urethral mucosa with edematous vascular submucosa, protruding from the urethra and mimicking a urethral neoplasm. Histologically, the urethral mucosa may exhibit ulceration, and the underlying connective tissue is generally filled with an inflammatory infiltrate. Vascular engorgement usually is present. Cryosurgery is an effective method of treatment (Kaufman 1994).

Urethral Caruncles Caruncles are fleshy, friable, sessile, or polypoid masses that arise in the posterior urethra near the meatus. Most patients are postmenopausal, with an average age of 68 years, but cases have been reported in adult women of all ages (Conces et al. 2012). Caruncles often are asymptomatic but may cause bleeding or dysuria. The lesions may be quite large, up to 3 cm in size, but the average lesion is less than 1 cm (Conces et al. 2012). Superficial ulceration is present in over one-third of lesions (Conces et al. 2012). On histologic examination, the epithelium is typically a mixture of urothelial and squamous. Invagination of the epithelium into the stroma, producing rounded nests or gland-like spaces, is a

Malakoplakia Malakoplakia of the urethra is a chronic granulomatous inflammatory process which presents as a polypoid mass at the urethral meatus. When the urethra is involved, it is usually as a result of spread from bladder involvement. On microscopic examination, the lesion contains foamy histiocytes, lymphocytes, granulocytes, and plasma cells. The diagnostic Michaelis–Gutmann bodies are seen within the cytoplasm of the histiocytes as inclusions having a blue-gray color. The inclusions may appear laminated or targetoid. With PAS stains, the Michaelis–Gutmann bodies usually stain pink to red. Many of the adjacent histiocytes also contain PAS-positive cytoplasmic material. Excision may be diagnostic and curative for small lesions within the urethra, although recurrences are not uncommon. Antibiotics also may be of value.

Condyloma Acuminatum Condyloma acuminatum may involve the urethra, and the presentation may mimic urethral prolapse, caruncle, or urethral carcinoma. Urethral condyloma, however, is usually seen in women of reproductive age and is rare in postmenopausal women. It is usually associated with condyloma acuminatum of other anogenital sites, particularly the vulvar vestibule. When condylomata acuminata are present in the mid or upper urethra, patients may have associated symptoms of urethritis, but otherwise the lesions are usually asymptomatic. The histologic appearance is the same as in other locations in the lower anogenital tract.

1

Benign Diseases of the Vulva

References Abdul-Cadir J, Catania L, Hindin MJ, Say L, Petignat P, Abdulcadir O (2016) Female genital mutilation: a visual reference and learning tool for health care professionals. Obstet Gynecol 128:958–963 Abou M, Dallenbach P (2013) Acute cervicitis and vulvovaginitis may be associated with cytomegalovirus. BMJ Case Rep. https://doi.org/10.1136/bcr-2013-008884 Abreu-dos-Santos F, Camara S, Reis F, Freitas T, Gaspar H, Cordeiro M (2016) Vulvar lobular capillary hemangioma: a rare location for a frequent entity. Case Rep Obstet Gynecol. https://doi.org/10.1155/2016/ 3435270 Ahronowitz I, Harp J, Shinkai K (2012) Etiology and management of pyoderma gangrenosum. Am J Clin Dermatol 13:191–211 Akopians AL, Rapkin AJ (2015) Vulvodynia: the role of inflammation in the etiology of localized provoked pain of the vulvar vestibule (vestibulodynia). Semin Reprod Med 33:239–245 Al-Ojaimi EH, Abdulla MM (2012) Giant epidermoid inclusion cyst of the clitoris mimicking clitoromegaly. J Lower Gen Tract Dis 17:58–60 Ali H, Guy RJ, Wand H, Read TH, Regan DG, Grulich AE, Fairley CK, Donovan B (2013) Decline in in-patient treatments of genital warts among young Australians following the national HPV vaccination program. BMC Infect Dis 13:140. http://www.biomedcentral. com/17471-2334/13/140 Allen AL, Siegfried EC (1998) The natural history of condyloma in children. J Am Acad Dermatol 39:951–955 Anderson-Mueller BE, Laudenschlager MD, Hansen KA (2009) Epidermoid cyst of the clitoris: an unusual cause of clitoromegaly in a patient without history of previous female circumcision. J Pediatr Adolesc Gynecol 22: e130–e132 Andreassi L, Bilenchi R (2014) Non-infectious inflammatory genital lesions. Clin Dermatol 32:307–314 Arkwright PD, Cassim M, Subramanian H, Spergel J, Schneider LC, Wollenberg A (2013) Management of difficult-to-treat atopic dermatitis. J Allergy Clin Immunol 1:142–151 Asante A, Omurtag K, Roberts C (2010) Epidermal inclusion cyst of the clitoris 30 years after female genital mutilation. Fertil Steril 94:e1–e3 Atzmony L, Hodak E, Leshem YA, Rosenbaum O, Gdalevich M, Anhalt GJ, Mimouni D (2015) The role of adjuvant therapy in pemphigus: a systemic review and meta-analysis. J Am Acad Dermatol 73:264–271 Aydin Y, Atis A, Polat N (2011) Bilateral endometrioma of Bartholin glands accompanying ovarian endometrioma. J Obstet Gynaecol 31:187–189 Baandrup L, Blomberg M, Dehlendorff C, Sand C, Andersen K, Kjaer S (2013) Significant decrease in the incidence of genital warts in young Danish women after implementation of a national human papillomavirus vaccination program. Sex Transm Dis 40:130–135

59 Bai H, Cviko A, Granter S, Yuan L, Betensly R, Crum C (2003) Immunophenotype and viral (human papillomavirus) correlates of vulvar seborrheic keratosis. Hum Path 34:559–564 Baker GM, Selim A, Hoang MP (2013) Vulvar adnexal lesions: a 32 year, single institution review from Massachusetts General Hospital. Arch Pathol Lab Med 137:1237–1246 Balfour FJT, Vincenti AC (1991) Idiopathic vulvar calcinosis. Histopathology 18:183–184 Ball RA, Edwards L, Reutter JC, West KL, Selim MA (2015) Inflammatory disorders affecting the epidermis of the vulva. In: Hoang, Selim (eds) Vulvar pathology. Springer, New York, pp 31–69 Barbosa NDF, de Aguiar LM, Maruta CW, Aoki V, Sotto MN, Labinas GHO, Perigo AM, Santi CG (2012) Vulvo-vaginal manifestations and evaluation of Papanicolou smears in pemphigus vulgaris and pemphigus foliaceous. J Am Acad Dermatol 67: 409–416 Barret M, de Parades V, Battistella M, Sokol H, Lemarchand N, Marteau P (2013) Crohn's disease of the vulva. J Crohn’s Colitis 8:563–570 Bast-Juzbasic A, Ceovic R (2014) Chancroid, lymphogranuloma venereum, granuloma inguinale, genital herpes simplex infection and molluscum contagiosum. Clin Dermatol 32:290–298 Bauer A, Rodiger C, Greif C, Kaatz M, Elsner P (2005) Vulvar dermatoses-irritant and allergic contact dermatitis of the vulva. Dermatology 210:143–149 Biswas A, Cooper J, Latifaj B (2007) Metastatic calcinosis cutis presenting as bilateral vulval cysts. Brit J Dermatol 157:622–624 Bornstein J, Goldstein AT, Stockdale CK, Bergeron S, Pukall C, Zolnoun D, Coady D (2016) 2015 ISSVD, ISSWSH and IPPS consensus terminology and classification of persistent vulvar pain and vulvodynia. Obstet Gynecol 127:745–751 Brenn T (2011) Atypical genital nevus. Arch Pathol Lab Med 135:317–320 Buda A, Ferrari L, Marra C, Passoni P, Perego P, Milani R (2008) Vulvar endometriosis in surgical scar after excision of the Bartholin gland: report of a case. Arch Gynecol Obstet 277:255–256 Carlson JA, Mu XC, Slominski A, Weismann K, Crowson AN, Malfetano J, Prieto V, Mihm M (2002) Melanocytic proliferations associated with lichen sclerosus. Arch Dermatol 138:77–87 Carter J, Elliott P, Russel P (1992) Bilateral fibroepithelial polyps of labium minus with atypical stromal cells. Pathology 24:37–39 Cetinkurson S, Narci A, Sahin O, Ozcara E (2009) Epidermoid cyst causing clitoromegaly in a child. Int J Gynaecol Obstet 105:64 Chan C, Chiu H (2008) Vestibular papillomatosis. NEJM 358:1485 Chang M, Newman C, Davis MDP, Lehman JS (2016) Acquired lymphangiectasia (lymphangioma circumscriptum) of the vulva: clinicopathologic study

60 of 11 patients from a single institution and 67 from the literature. Int J Dermatol 55:e482–e487 Chen W, Koenig C (2004) Vulvar keratoacanthoma: a report of two cases. Int J Gynecol Path 23:284–286 Cilla G, Pico F, Peris A, Idigoras P, Urbieta M, PerezTrallero E (1992) Human genital myiasis due to Sarcophaga. Rev Clín Esp 190:189–190 Coghill SB, Tyler X, Shaxted EJ (1990) Benign mucinous metaplasia of the vulva. Histopathol 17:373–375 Cohen SE, Klausner JD, Engelman J, Philip S (2013) Syphylis in the modern era. Infect Dis Clin N Am 27:705–722 Conces MR, Williamson SR, Montironi R, Lopez-Beltran A, Scarpelli M, Cheng L (2012) Urethral caruncle: clinicopathologic features of 41 cases. Hum Path 43:1400–1404 Connor CJ, Eppsteiner EE (2014) Vulvar contact dermatitis. Proc Obstet Gynecol 4: 1. Available from http://ir. uiowa.edu/pog/ Cooper SM, Wojnarowska F (2006) Influence of treatment on erosive lichen planus of the vulva and its prognosis. Arch Dermatol 142:289–294 Dabade TS, Davis MDP (2011) Diagnosis and tretment of the neutrophilic dermatoses (pyoderma gangrenosum, Sweet's syndrome). Dermatol Therapy 24:273–284 Day T, Scurry J (2014) Vulvar Pityriasis versicolor in an immunocompetent woman. J Lower Gen Tract Dis 18: e71–e73 Dereli T, Turk BG, Kazandi AC (2007) Syringomas of the vulva. Int J Gynaecol Obstet 99:65–66 Domfeh AB, Silasi DA, Lindo F, Parkash V (2012) Chronic hypertrophic vulvar herpes simulating neoplasia. Int J Gyneco Pathol 31:33–37 Duan D, Stevenson ML, Malter LB, Pomeranz MK (2014) Cutaneous Crohn’s disease of the vulva. BMJ Case Rep 2014. https://doi.org/10.1136/bcr-2014-204507 Edwards L (2010) Pigmented vulvar lesions. Dermatol Ther 23:449–457 Egan CA, Zone JJ (1999) Linear IgA bullous dermatosis. Int J Dermatol 38:818–827 Elas D, Swick B, Stone MS, Miller M, Stockdale C (2014) Botryomycosis of the vulva: a case report. J Lower Gen Tract Dis 18:e80–e83 El-Khoury J, Renald M, Plantier F, Avril M, MoyalBarracco M (2016) Vulvar hidradenoma papilliferum is located on the sites of mammary-like anogenital glands (MLAGs): analysis of the photographs of 52 tumors. J Am Acad Dermatol 75:380–384 Farahnik B, Blattner CM, Mortazie MB, Perry BM, Lear W, Elston DM (2017) Interentional treatments for HaileyHailey disease. J Am Acad Dermatol 76:551–558 Fatahzadeh M, Schwartz RA (2007) Human herpes simplex virus infections: epidemiology, pathogenesis, symptomatology, diagnosis, and management. J Am Acad Dermatol 57:737–763 Fistarol SK, Itin PH (2013) Diagnosis and treatment of lichen sclerosus. Am J Clin Dermatol 14:27–47 Fite C, Plantier F, Dupin N, Avril M, Moyal-Barracco M (2011) Vulvar verruciform xanthoma: ten cases associated with lichen sclerosus, lichen planus, or other conditions. Arch Dermatol 147:1087–1092

D. S. Rush and E. J. Wilkinson Flagg EW, Schwartz R, Weinstock H (2013) Prevalence of anogenital warts among participants in private health plans in the United States, 2003-2010: potential impact of human papillomavirus vaccination. Am J Pub Health 103:1428–1435 Foote CA, Brady SP, Brady KL, Clark NS, Mercurio MG (2013) Vulvar dermatitis from allergy to moist flushable wipes. J Lower Gen Tract Dis 18:E16–E18 Fox H, Wells M, Harris M, McWilliam LJ, Anderson GS (1988) Enteric tumours of the lower female genital tract: a report of three cases. Histopathol 12:167–176 Frankel MA, Rhodes HE, Euscher ED (2011) Verruciform xanthoma in an adolescent: a case report. J Lower Gen Tract Dis 16:70–74 French P, Ison CA, Macdonald N (2005) Lymphogranuloma venereum in the United Kingdom. Sex Transm Infect 81:97–98 Garland SM, Steben M, Sings HL, James M, Lu S, Railkar R, Barr E, Haupt RM, Joura EA (2009) Natural history of genital warts: analysis of the placebo arm of 2 randomized phase III trials of a quadrivalent human papillomavirus (types 6,11,16, and 18) vaccine. J Infect Dis 199:805–814 Ghosh D, Woodrow S, Mathew J, Lopes A, Das N (2011) Chronic granulomatous inflammation of the vulva: an unusual presentation with diagnostic and therapeutic difficulties. J Lower Gen Tract Dis 15:322–324 Gilbey S, Moore D, Look K, Sutton G (1997) Vulvar keratoacanthoma. Obstet Gynecol 89:848–850 Gleason BC, Hirsch MS, Nucci MR, Schmidt BA, Zembowicz A, Mihm MC, McKee PH, Brenn T (2008) Atypical genital nevi: a clinicopathologic analysis of 56 cases. Am J Surg Pathol 32:51–57 Goldstein AT, Marinoff SC, Chrostopher K, Srodon M (2005) Prevalence of vulvar lichen sclerosus in a general gynecology practice. J Reprod Med 50:477–480 Goldstein AT, Metz A (2005) Vulvar Lichen planus. Clin Obstet Gynecol 48:818–823 Gomes do Amaral RL, Giraldo PC, Cursino K, Goncalves AK, Eleuterio J, Giraldo H (2009) Nodular vulvar herpes in an HIV-positive woman. Int J Obstet Gynecol 107:255 Gupta R, Warren T, Wald A (2007) Genital herpes. Lancet 370:2127–2137 Gurumurthy M, Cairns M, Cruickshank M (2009) Case series of zoon vulvitis. J Lower Gen Tract Dis 14:56–58 Gushi A, Kanekura T, Kanzaki T, Eizuru Y (2003) Detection and sequences of human papillomavirus DNA in non-genital seborrheic keratosis of immunocompetent individuals. J Derm Sci 31:143–149 Halvorsen JA, Brevig T, Aas T, Skar AG, Slevolden EM, Moi H (2006) Genital ulcers as initial manifestation of Epstein-Barr virus infection: two new cases and a review of the literature. Acta Derm Venerol 86:439–442 Heller DS, Bean S (2014) Lesions of the Bartholin gland: a review. J Lower Gen Tract Dis 18:351–357 Heller DS (2015) Benign tumors and tumor-like lesions of the vulva. Clin Obstet Gynecol 58:526–535 Heller J, Roche N, Hameed M (2009) Trichoepithelioma of the vulva: report of a case and review of the literature. J Lower Gen Tract Dis 13:186–187

1

Benign Diseases of the Vulva

Hoang MP, Reuter J, Papalas JA, Edwards L, Selim MA (2014) Vulvar inflammatory dermatoses: an update and review. Am J Dermatopathol 36:689–704 Hoang MP, Sangueza OP (2015) Vascular lesions of the vulva. In: Huang MP, Selim MA (eds) Vulvar pathology. Springer, New York, pp 411–438 Hoang MP, Fernandez-Figueras MT, Mihm M (2015a) Blistering disorders and acantholytic processes affecting the epidermis of the vulva. In: Huang MP, Selim MA (eds) Vulvar pathology. Springer, New York, pp 71–94 Hoang MP, Kazakov DV (2015) Lesions of Anogenital mammary-like glands, adnexal neoplasms and metastases. In: Huang MP, Selim MA (eds) Vulvar pathology. Springer, New York, pp 327–354 Hoang MP, Kazakov DV, Selim MA (2015b) Cysts, glandular lesions and others. In: Huang MP, Selim MA (eds) Vulvar pathology. Springer, New York, pp 355–383 Hofstetter S, Shah M (2015) Vulvodynia. In: Clin Obstet Gynecol 58–536-545 Hope-Rapp E, Anyfantakis V, Fouere S, Bonhomme P, Louison JB, Tandeau de Marsac T, Chaine B, Vallee P, Casin I, Scieux C, Lassau F, Janier M (2010) Etiology of genital ulcer disease. A prospective study of 278 cases seen in an STD clinic in Paris. Sex Transm Dis 37:153–158 Horn L, Bauerfeind UK, Straub H, Karbe I (2014) Vulval intestinal enteric heteropia with a 10-year follow-up: a case report and review of the literature. Int J Gynecol Pathol 33:258–262 Hornor G (2004) Ano-genital warts in children: sexual abuse or not? J Pediatr Health Care 18:165–170 Hoy T, Singhal PK, Willey VJ, Insinga RP (2009) Assessing incidence and economic burden of genital warts with data from a US commercially insured population. Curr Med Res Opin 25:2343–2351 Huppert JS, Gerber MA, Dietch HR, Mortensen JE, Staat MA, Adams HPJ (2006) Vulvar ulcers in young females: a manifestation of apthosis. J Pediatr Adolesc Gynecol 19:195–204 Jayasinghe Y, Garland SM (2006) Genital warts in children: what do they mean? Arch Dis Child 91:696–700 Karaa A, Khachemoune A (2007) Keratoacanthoma: a tumor in search of a classification. Int J Derm 46:671–678 Kasperkiewicz M, Zillikens D, Schmidt E (2012) Pemphigoid diseases: pathogenesis, diagnosis and treatment. Autoimmunity 45:55–70 Kaufman RH (1994) Benign diseases of the vulva and vagina, 4th edn. Mosby, St. Louis Kavala M, Demir FT, Zindanci I, Can B, Turkoglu Z, Zemheri E, Cam OH, Teksen A (2015) Genital involvement in pemphigus vulgaris (PV): correlation with clinical and cervicovaginal pap smear findings. J Am Acad Dermatol 73:655–659 Kazakov DV, Curik R, Vanecek T, Mukensnabl P, Michal M (2007) Nodular hyperplasia of the Bartholin gland: a clinicopathological study of two cases, including detection of clonality by HUMARA. Am J Dermatopathol 29:385–387 Kazakov DV, Bouda J, Kacerovska D, Michal M (2010a) Vulvar syringomas with deep extension: a potential

61 histopathologic mimic of microcystic adnexal carcinoma. Int J Gynecol Pathol 30:92–94 Kazakov DV, Stewart CJR, Kacerovska D, leake R, Kreuzberg B, Chudacek Z, Hora M, Michal M (2010b) Prostatic-type tissue in the lower female genital tract: a morphologic spectrum including vaginal tubulosquamous polyp, adenomyomatous hyperplasia of paraurethral skene glands (female prostate) and ectopic lesion in the vulva. Am J Surg Pathol 34:950–955 Kazakov DV, Spagnolo DV, Kacerovska D, Michal M (2011) Lesions of anogenital mammary-like glands: an update. Adv Anat Pathol 18:1–28 Kelly P, McBride HA, Kennedy K, Connolly LE, McCluggage WG (2011) Misplaced Skene’s glands: glandular elements in the lower female genital tract that are variably immunoreactive with prostate markers and that encompass vaginal tubulosquamos polyp and cervical ectopic prostatic tissue. Int J Gynecol Pathol 30:605–612 Klein PA, Callen JP (2000) Drug-induced linear IgA bullous dermatosis after vancomycin discontinuance in a patient with renal insufficiency. J Am Acad Dermatol 42(Pt 2):316–323 Kneisel A, Hertl M (2011) Autoimmune bullous skin diseases. Part 1: clinical manifestations. JDDG 9: 844–857 Konstantinova AM, Michal M, Kacerovska D, Spagnolo D, Stewart C, Kutzner H, Zelger B, Plaza J, Denisjuk N, Hejda V, Shelekhova K, Bisceglia M, Danis D, Scurry J, van der Putte SC, Pyman J, Chetty N, Szabo R (2009) Mammary-like gland adenoma of the vulva: review of 46 cases. Pathology 41:372–378 Konstantinova AM, Michal M, Kacerovska D, Spagnolo DV, Stewart CJ, Kutzner H, Zelger B, Plaza JA, Denisjuk N, Hejda V, Shelekhova K, Bisceglia M, Danis D, Zamecnik M, Kerl K, Guenova E, Kazakov D (2016) Hidradenoma papilliferum: a clinicopathologic study of 264 tumors from 261 patients, with emphasis on mammary-type alterations. Am J Dermatopathol 38:374–383 Konstantinova AM, Kyrpychova L, Belousva IE, Spagnolo DV, Kacerovska D, Michal M, Kerl K, Kazakov DV (2017a) Anogenital mammary-like glands: a study of their normal histology with emphasis on glandular depth, presence of columnar epithelial cells, and distribution of elastic fibers. Am J Dermatopathol 39:663–67 Konstantinova AM, Stewart CJR, Kyrpychova L, Belousva IE, Michal M, Kazakov DV (2017b) An immunohistochemical study of anogenital mammarylike glands. Am J Dermatopathol 39:599–605 Koo YJ, Kim YS, Min KJ, Hong JH, Lee JK (2016) High -risk human papillomavirus infection and the prediction of poor treatment response and disease recurrence in patients with vulvovaginal condyloma. Int J Gynecol Obstet 132:20–24 Koranantakul O, Lekhakula A, Wansit R, Koranantakul Y (1991) Cutaneous myiasis of vulva caused by the muscoid fly (Chrysomyia genus). Southeast Asian J Trop Med Public Health 22:458–460

62 Kuniyuki S, Fukushima Y, Yoshida Y, Yamanaka K, Maekawa N, Inoue T (2008) Ciliated cyst of the vulva: oestrogen and progesterone receptors. Acta Derm Venereol 88:514–515 Kurashige Y, Kishida K, Kurashige K, Minemura T, Nagatani T (2014) Hidradenoma papilliferum of the vulva in association with an anogenital mammary-like gland. J Dermatol 41:411–413 Lacey CJN, Woodhall SC, Wikstrom A, Ross J (2013) 2012 European guideline for the management of anogenital warts. J Europ Acad Dermatol Vener 27: e263–e270 Lawrance NJ, Mowatt D, Yell JA (2015) Massive localized vulval lymphedema with lymphangiomas of the labia majora, misdiagnosed as viral warts. J Lower Gen Tract Dis 19:e64–e65 Leclair CM, Goetsch MF, Korcheva VB, Anderson R, Peters D, Morgan TK (2011) Differences in primary compared with secondary vestibulodynia by immunohistochemistry. Obstet Gynecol 117:1307–1313 Lee MY, Dalpiaz A, Schwamb R, Miao Y, Waltzer W, Khan A (2014) Clinical pathology of Bartholin’s glands: a review of the literature. Curr Urol 8:22–25 Lewis FM (1998) Vulval lichen planus. British J Dermatol 138:569–575 Li J, Ackerman AB (1994) Seborrheic keratoses that contain human papillomavirus are condyloma acuminata. Am J Dermatopathol 16:398–405 Li J, Shi Y, Zhou C, Lin J (2015) Diagnosis and treatment of perineal endometriosis: review of 17 cases. Arch Gynecol Obstet 292:1295–1299 Lora V, Scarabello A, Cota C (2015) Warty dyskeratoma as a cutaneous horn on the Mons pubis. Am J Dermatopathol 37:801–804 Lynch PJ, Moyal-Barracco M, Bogliatto F, Micheletti L, Scurry J (2006) ISSVD classification of vulvar dermatoses: pathologic subsets and their clinical correlates. J Reprod Med 2007:3–9 Lyrio LDC, Grassi MFR, Santana IU, Olavarria VG, Gomes A, Costa Pinto L, Oliveira RP, Aquino R, Santiago MB (2013) Prevalence of human papillomavirus infection in women with systemic lupus erythematosus. Rheumatol Int 33:335–340 Maniar KP, Ronnett BM, Vang R, Yemelyanova A (2013) Coexisting high-grade vulvar intraepithelial neoplasia (VIN) and condyloma acuminatum-independent lesions due to different HPV types occurring in immunocompromised patients. Am J Surg Pathol 37:53–60 Manoj K, Soma M, Ajay L, Ashish A, Rakesh S, Paliwal RV (2008) Tubercular sinus of labia majora: rare case report. Infect Dis Obstet Gynecol 2008:81753. https:// doi.org/10.1155/2008/87515 Margesson LJ (2006) Vulvar disease pearls. Dermatol Clin 24:145–155 Marquette GP, Su B, Woodruff JD (1985) Introital adenosis associated with Stevens-Johnson syndrome. Obstet Gynecol 66:143–145 Marwah S, Bergman ML (1980) Ectopic salivary gland in the vulva (choristoma): report of a case and review of the literature. Obstet Gynecol 56:389–391

D. S. Rush and E. J. Wilkinson Marzano DA, Haefner HK (2004) The Bartholin gland cyst: past, present and future. J Lower Gen Tract Dis 8:195–204 Marzano A, Ramoni S, Caputo R (2008) Amicrobial pustulosis of the folds: report of 6 cases and a literature review. Dermatology 216:305–311 Mat C, Yurdakul S, Sevim A, Ozyazgan Y, Tuzun Y (2013) Behcet’s syndrome: facts and controversies. Clin Dermatol 31:352–361 Medeiros F, Nascimento AF, Crum CP (2005) Early vulvar squamous neoplasia: advances in classification, diagnosis and differential diagnosis. Adv Anat Pathol 12:20–26 Mittal P, Prakash V, Gupta R, Dewan R, Singhal S, Suri J (2013) Giant condyloma acuminatum of vulva treated by surgical excision and reconstruction of defect. Arch Gynecol Obstet 287:1047–1048 Miziara ID, Mahmoud A, chagury AA, Alves RD (2013) Bullous systemic lupus erythematosus: case report. Int Arch Otorhinolayngol 17:344–346 Mohammed TT, Olumide YM (2008) Chancroid and human immunodeficiency virus infection-a review. Int J Dermatol 47:1–8 Mosunjac M, Park JP, Wang YF, Tadros T, Siddiqui M, Bagirov M, Little J (2009) Genital and perianal herpes simplex simulating neoplasia in patients with AIDS. AIDS Patient Care STDs 23. https://doi.org/10.1089/ apc.2008.0143 Moyal-Barracco M, Wendling J (2014) Vulvar dermatosis. Best Prac Res Clin Obstet Gynecol 28:946–958 Murzaku EK, Penn LA, Hale CS, Pomeranz MK, Polsky D (2014) Vulvar nevi, melanosis, and melanoma: an epidemiologic, clinical and histopathologic review. J Am Acad Dermatol 71:1241–1249 Nakayama J, Busse R (2010) An analysis of vulvar necrotizing fasciitis in the unique and ethnically diverse Hawaiian population. Hawaii Med J 69:13-16 Nascimento M, Cominos D, Davies N, Obermair A (2005) Vulval keratoacanthoma. Gyn Onc 97:674–676 Nasu K, Okamato M, Nishida M, Narahara H (2013) Endometriosis of the perineum. Obstet Gynaecol Res 39:1095–1097 Nettina SM (1998) Herpes genitalis. Lippincotts Prim Care Pract 2:303–306 Newland JR, Fusaro RM (1991) Mucinous cysts of the vulva. Nebr Med J 76:307–310 Nigam JS, Deshpande AH, Bharti JN, Nair V (2017) Melanocytic colonization of Bartholin cyst: a rarely observed phenomenon. J Lower Gen Tract Dis 21:1–2 Nomelini RS, Pansani PL, Guimaraes PDN, MartinsFilho A, Barcelos ACM, Murta EFC (2010) Vulvar angiokeratoma. J Obstet Gynaecol 30:418–419 O'Farrell N, Moi H (2016) 2016 European guideline on donovanosis. Int J STD AIDS 27:605–607 O'Gorman SM, Torgerson RR (2013) Allergic contact dermatitis of the vulva. Dermatitis 24:64–72 Oliveira A, Lobo I, Selores M (2011) Asymptomatic vulvar pigmentation. Clin Exp Dermatol 36:921–922 Osarumwense DO (2010) Post genital mutilation giant clitoral epidermoid inclusion cyst in Benin City. Nigeria J Pediatr Adolesc Gynecol 23:336–340

1

Benign Diseases of the Vulva

Ozcan F, Bilgic R, Cesur S (2006) Vulvar keratoacanthoma. APMIS 114:562–565 Ozcan Z (2009) Multiple eccrine hidrocystomas of the vulva. Int J Gynaecol Obstet 1:65 Paulus YM, Wong AE, Chen B, Jacobson MT (2010) Preputial epidermoid cyst: an atypical case of acquired pseudoclitoromegaly. J Lower Gen Tract Dis 14:382–386 Pereira IB, Khan A (2017) Sarcoidosis rare cutaneous manifestations: vulval and perianal involvement. J Obstet Gynaecol 37:539–540 Pernet C, Bessis D, Savagnac M, Tron E, Guillot B, Hovnanian A (2012) Genitoperineal papular acantholytic dyskeratosis is allelic to Hailey-Hailey disease. Brit J Dermatol 167:211–213 Prayson RA, Stoler MH, Hart WR (1995) Vulvar vestibulitis. A histopathologic study of 36 cases, including human papillomavirus in situ hybridization analysis. Am J Surg Path 19:154–160 Quddus MR, Sung CJ, Simon RA, Lawrence WD (2014) Localized amyloidosis of the vulva with and without vulvar intraepithelial neoplasia: report of a series. Hum Path 45:2037–2042 Read TR, Hocking JS, Chen MY (2011) The near disappearance of genital warts in young women 4 years after commencing a national human papillomavirus (HPV) vaccination program. Sex Transm Infect 87:544–547 Reed BG, Shippey S, Kremp A, Belin E (2013) Vulvar pyoderma gangrenosum originating from a healed obstetric laceration. Obstet Gynecol 122:452–454 Regauer S, Eberz B, Beham-Schmid C (2015) Mast cell infiltrates in vulvodynia represent secondary and idiopathic mast cell hyperplasia. APMIS 123:452–445 Ridley CM, Neill SM (1999) The vulva, 2nd edn. Blackwell, Oxford, pp 1–36 Ribe A (2008) Melanocytic lesions of the genital area with attention given to atypical genital nevi. J Cutan Pathol 35:24–27 Riszk DEE, Mohammed KH, Joshi SU, Al-Shabani AY, Bossmar TR (2007) A large clitoral epidermoid inclusion cyst Forst presenting in adulthood following childhood circumcision. J Obstetr Gynecol 27:446–448 Robinson JB, Im DD, Simmons-O’Brien E, Rosenshein NB (1998) Etretinate: therapy for plasma cell vulvitis. Obstet Gynecol 51:347–351 Rock B, Hood AF, Rock JA (1990) Prospective study of vulvar nevi. J Am Acad Dermatol 22:104–106 Ruocca V, Ruocco E, Schivao AL, Brunetti G, Guerrera LP, Wolf R (2013) Pemphigus: etiology, pathogenesis, and inducing or triggering factors: facts and controversies. Clin Dermatol 31(374):381 Ruocco E, Wolf R, Ruocco V, Brunetti G, Romano F, Schiavo AL (2013a) Pemphigus: associations and management guidelines: facts and controversies. Clin Dermatol 31:382–390 Ruocco E, Wolf R, Cacavale S, Brancaccio G, Ruocco V, Schiavo AL (2013b) Bullous pemphigoid: associations and management guidelines: facts and controversies. Clin Dermatol 31:400–412 Ruocco V, Ruocco E, Caccavale S, Gambardella A, Schivao AL (2015) Pemphigus vegetans of the folds (intertriginous areas). Clin Dermatol 33:471–476

63 Satoh M, Yamamoto T (2013) Genital pyoderma gangrenosusm: report of two cases and published work review of Japanese cases. J Dermatol 40:840–843 Saitoh A, Ohya T, Yoshida S, Hosoya R, Nishimura K (1995) A case report of Stevens-Johnson syndrome with mycoplasma infection. Acta Paediatr Jpn 37:113–115 Sand FL, Thomsen SF (2017) Skin diseases of the vulva: infectious diseases. J Obstet Gynaecol. https://doi.org/ 10.1080/01443615.2017.1306696 Sangueza JM, Saenz ML (2007) Challenge. Am J Dermatopathol 29:210 Santana IU, Gomes A, Lyrio LDC, Grassi MFR, Santiago MB (2011) Systemic lupus erythematosus, human papillomavirus infection, cervical pre-malignant and malignant lesions: a systematic review. Clin Rheumatol 30:665–672 Sbano P, Rubegni P, Risulo M, De Nisi MC, Fimiani M (2007) A case of idiopathic granulomatous cheilitis and vulvitis. Int J Dermatol 46:720–721 Schiavo AL, Ruocco E, Brancaccio G, Caccavale S, Ruocco V, Wolf R (2013) Bullous pemphigoid: etiology, pathogenesis and inducing factors: facts and controversies. Clinics in Dermatol 31:391–399 Schneider CA, Festa S, Spillert CR, Bruce CJ, Lazaro EJ (1994) Hydrocele of the canal of Nuck. N J Med 91:37–38 Schneider M, Cardones A, Selim MA (2016) Amicrobial pustulosis of the folds and autoimmunity. J Lower Gen Tract Dis 20:e57–e58 Schober MS, Hendrickson BW, Alpert SA (2014) Spontaneous clitoral hood epidermal inclusion cyst mimicking clitoromegaly in a pediatric patient. Urology 84:206–208 Schofield JK, Tatnall FM, Leigh IM (1993) Recurrent erythema multiforme: clinical features and treatment in a large series of patients. Brit J Dermatol 128: 542–545 Schwartz R (2004) Keratoacanthoma: a clinico-pathologic enigma. Dermatol Surg 30:326–333 Scurry J, van der Putte SCJ, Pyman J, Chetty N, Szabo R (2009) Mammary-like gland adenoma of the vulva: a review of 46 cases. Pathology 41:372–378 Selim MA et al (2015) Infectious diseases and infestations of the vulva. In: Hoang MP, Selim MA (eds) Vulvar pathology. Springer, New York, pp 139–193 Shen HP, Chang WC, Hsieh CH, Yang TC, Hung YC (2011) Vulvar tuberculosis. Taiwan J Obstet Gynecol 50:106–108 Sherman V, McPherson T, Baldo M, Salim A, Gao XH, Wojnarowska F (2010) The high rate of familial lichen sclerosus suggests a genetic contribution: an observational cohort study. J Eur Acad Dermatol Venereol 24:1031–1034 Simpson RC, Thomas KS, Leighton P, Murphy R (2013) Diagnostic criteria for erosive lichen planus affecting the vulva: an international electronic-Delphi consensus exercise. Brit J Dermatol 169:337–343 Sington J, Chandrapala R, Manek S, Hollowood K (2006) Mitotic count is not predictive of clinical behavior in hidradenoma papilliferum of the vulva: a clinicopathologic study of 19 cases. Am J Dermatopathol 28:322–326

64 Stefanaki C, Barkas G, Valari M, Bethimoutis G, Nicolaidou E, Vosynioti V, Kontochristopoulos G, Papadogeorgaki H, Verra P, Katsarou A (2012) Condyloma acuminata in children. Pediatr Infect Dis J 31:422–424 Stephenson H, Dotters DJ, Katz V, Droegemueller W (1992) Necrotizing fasciitis of the vulva. Am J Obstet Gynecol 166:1324–1327 Steshenko O, Chandrasekaran N, Lawton F (2014) Syringocystadenoma papilliferum of the vulva: a rarity in gynaecology. BMJ Case Reports. https://doi.org/ 10.1135/bcr-2014-203902 Strehl JD, Mehlhorn G, Koch MC, Harrer EG, Harrer T, Beckmann MW, Agaimy A (2012) Int J Gynecol Pathol 31:286–293 Sultan HY, Boyle AA, Sheppard N (2012) Necrotising fasciitis. BMJ 345:e4274. https://doi.org/10.1136/bmj.e4274 Sun T, Schwartz NS, Sewell C, Lieberman P, Gross S (1991) Enterobius egg granuloma of the vulva and peritoneum: review of the literature. Am J Trop Med Hyg 45:249–253 Tangjitgamol S, Loharamtaweethong K, Thawaramara T, Chanpanitkitchot S (2013) Vulvar pseudoepitheliomatous hyperplasia mimicking cancer in an immunocompromised patient. J Obstet Gynecol Res 40:255–258 Takagi A, Kamijo M, Ikeda S (2016) Darier Disease. J Dermatol 43:275–279 Taylor S, Drake SM, Dedicoat M, Wood MJ (1998) Genital ulcers associated with acute Epstein-Barr virus infection. Sex Transm Infect 74:296–297 Tomazzini E, Giraldo P, Amaral R, Eleuterio J, Cintra ML, Giraldo HPD (2008) Vulvar calcinosis in childhood. Int J Gynaecol Obstet 103:263–264 Tommola P, Butzow R, Unkila-Kallio L, Paavonen J, Meri S (2015) Activation of vestibule-associated lymphoid tissue in localized provoked vulvodynia. Am J Obstet Gynecol 212:476.e1–476.e8 Van der Avoort IAM, van der Laak JAWM, Otte-Holler I, van de Nieuwenhof HP, Massuger LFAG, de Hullu J, van Kempen LCLT (2010) The prognostic value of blood and lymph vessel parameters in lichen sclerosus for vulvar squamous cell carcinoma development: an immunohistochemical study. Am J Obstet Gynecol 203-167:e1–e8 Van der Putte SC (1991) Anogenital "sweat" glands. Histology and pathology of a gland the may mimic mammary glands. Am J Dermatopathol 13:557–567 Van der Putte SC (1994) Mammary-like glands of the vulva and their disorders. Int J Gynecol Pathol 13:150–160 Virgili A, Mantovani L, Lauriola MM, Marzola A, Corazza M (2008) Tacrolimus 0.1% ointment: is it really effective in plasma cell vulvitis? Report of four cases. Dermatology 216:243–246 Virgili A, Corazza M, Minghetti S, Borghi A (2015) Symptoms in plasma cell vulvitis: first observational cohort study on type, frequency and severity. Dermatology 230:113–118

D. S. Rush and E. J. Wilkinson Vitrey D, Frachon S, Balme B, Golfier F (2003) Tubulovillous adenoma of the vulva. Obstet Gynecol 102:1160–1163 Wang MZ, Camilleri MJ, Guo R, Wieland C (2017) Amicrobial pustulosis of the folds: report of 4 cases. J Cutan Pathol 44:367–372 Wakashin K (2007) Sanitary napkin contact dermatitis of the vulva: location dependent differences in skin surface conditions may play a role in negative patch test results. J Dermatol 34:834–837 Weedon D, Malo J, Brooks D, Williamson R (2010) Keratoacanthoma: is it really a variant of squamous carcinoma? ANZ J Surg 80:129–130 Wieselthier JS, Pincus SH (1993) Hailey-Hailey disease of the vulva. Arch Dermatol 129:1344–1345 Weyers W (2013) Hypertrophic lichen sclerosus with dyskeratosis and parakeratosis- a common presentation of vulvar lichen sclerosus not associated with a significant risk of malignancy. Am J Dermatopathol 35:713–721 Weyers W (2015) Hypertrophic lichen sclerosus sine sclerosis: clues to histopathologic diagnosis when presenting as psoriasiform lichenoid dermatitis. J Cutan Pathol 42:118–129 Wilkinson EJ, Stone IK (2008) Atlas of vulvar disease, 2nd edn. Lippincott Williams & Wilkins, Baltimore Wojnarowska F, Frith P (1997) Linear IgA disease. Dev Ophthalmol 28:64–72 Xie Y, Shi L, Xiong X, Wu E, Veasly C, Dade C (2012) Economic burden and quality of life of vulvodynia in the United States. Curr Med Res Opinion 28:601–608 Yazici Y, Yurdakul S, Yazici H (2010) Behcet’s Syndrome. Curr Rheumatol Rep 12:429–435 Yigiter M, Arda IS, Tosum E, Celik M, Hicsonmez A (2008) Angiokeratoma of clitoris: a rare lesion in an adolescent girl. Urology 71:604–606 Yoganathan S, Bohl TG, Mason G (1994) Plasma cell balanitis and vulvitis (of zoon). A study of 10 cases. J Reprod Med 39:939–944 Young M, Aldridge L, Parker P (2017) Psoriasis for the primary care practioner. J Amer Nurse Pract 29:157–178 Yu WY, Ng E, Hale C, Hu S, Pomeranz MK (2016) Papular acantholytic dyskeratosis of the vulva associated with familial Hailey-Hailey disease. Clin Exp Dermatol 41:628–631 Zaraa I, Sellami A, Bougerra C, Sellami MK, Chelly I, Zitouna M, Makni S, Hmida AB, Mokni M, Osman AB (2010) Pemphigus vegetans: a clinical, histological, immunopathological and prognostic study. J Euro Acad Dermato Venereol 25:1160–1167 Zhu J, Lu Z, Zheng M (2014) Acquired progressive lymphangioma in the inguinal area mimicking gioant condyloma acuminatum. Cutis 93:316–319 Zhuang K, Xu F, Ran Y, Lama J (2015) Atypical infantile genital Molluscum contagiosum. An Bras Dermatol 90:403–405

2

Precursor Lesions and Malignant Tumors of the Vulva Edward J. Wilkinson and Demaretta S. Rush

Contents Squamous Intraepithelial Lesions (SIL) (Vulvar Intraepithelial Neoplasia (VIN)) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . General Features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Clinical Features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Microscopic Findings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Microscopic Grading . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Histologic Subtypes of Vulvar SIL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Differential Diagnosis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Adjunctive Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Clinical Behavior and Treatment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

66 66 67 69 69 71 73 74 75

Differentiated Vulvar Intraepithelial Neoplasia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76 General Features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76 Clinical Features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76 Squamous Cell Carcinoma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Stage IA Invasive Squamous Cell Carcinoma (AJCC T1a, M0, N0; FIGO Stage IA) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Invasive Squamous Cell Carcinoma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Histologic Subtypes of Vulvar Squamous Cell Carcinoma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Basaloid Carcinoma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Warty Carcinoma (Condylomatous Carcinoma) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Verrucous Carcinoma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Giant Cell Squamous Carcinoma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Spindle Cell Squamous Cell Carcinoma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Lymphoepithelioma-Like Carcinoma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Plasmacytoid Squamous Carcinoma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

78 80 85 89 89 91 92 93 94 95 95

E. J. Wilkinson (*) Department of Pathology, Immunology and Laboratory Medicine, University of Florida College of Medicine, Gainesville, FL, USA e-mail: [email protected]fl.edu; [email protected] D. S. Rush Department of Pathology, University of Arizona College of Medicine, Tucson, AZ, USA e-mail: [email protected]; [email protected] # Springer Nature Switzerland AG 2019 R. J. Kurman et al. (eds.), Blaustein’s Pathology of the Female Genital Tract, https://doi.org/10.1007/978-3-319-46334-6_2

65

66

E. J. Wilkinson and D. S. Rush Skin Adnexal-type Carcinomas (Basal Cell, Adenoid Basal Cell, Basosquamous [Metatypical Basal Cell], Sebaceous Cell) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Glandular Tumors of the Vulva (Primary Adenocarcinomas of the Vulva) . . . . . . . Paget Disease . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Intestinal-Type Mucinous Adenocarcinoma (Villoglandular Mucinous Adenocarcinoma, Cloacogenic Carcinoma, Adenocarcinoma of Cloacogenic Origin) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Bartholin Gland Tumors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Skene Gland and Duct Adenocarcinomas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Mammary Gland-Like Adenocarcinomas, Including Phyllodes Tumor, Arising in Vulvar Specialized Anogenital Mammary-Like Glands . . . . . . . . . . . . . . . . . . . . Carcinomas of Sweat Gland Origin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

97 98 99

104 105 107 107 108

Malignant Melanoma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . General Features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Microscopic Findings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Differential Diagnosis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Staging, Clinical Behavior, and Treatment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

108 108 112 112 113

Other Malignant Tumors of the Vulva . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Carcinosarcoma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Malignant Blue Nevus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Yolk Sac Tumor (Endodermal Sinus Tumor) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Primary Malignant Lymphoma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . High-Grade Neuroendocrine Carcinoma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

114 114 114 115 115 116

Tumors of the Urethra . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118 Urethral Carcinoma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118 Other Malignant Tumors of the Urethra . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120 Gross Description, Processing, and Reporting of Vulvar Specimens . . . . . . . . . . . . . . 120 Vulvar Biopsies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120 Large Operative Specimens . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Wide Local Excision (Partial Deep Vulvectomy) and Superficial Vulvectomy . . . . . . . . Total Vulvectomy: Superficial or Deep . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Total Deep Vulvectomy with En Bloc Lymphadenectomy (Radical Vulvectomy, with Lymphadenectomy) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

121 121 121 122

References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122

Squamous Intraepithelial Lesions (SIL) College of American Pathologists (CAP) and a (Vulvar Intraepithelial Neoplasia (VIN)) number of leading medical organizations General Features Terminology for vulvar SILs has undergone considerable evolution as the biology of these lesions becomes better understood. The current terminology of “squamous intraepithelial lesions” replaces the previously used terms of dysplasia, carcinoma in situ, VIN, Bowen disease, Bowenoid dysplasia, and others. These changes were accepted by the World Health Organization (WHO) in concurrence with the

(American College of Obstetricians and Gynecologists (ACOG), American Society for Colposcopy and Cervical Pathology (ASCCP), International Society for the study of Vulvovaginal Disease (ISSVD), and others) and are summarized in Table 1 (Bornstein et al. 2016; Crum et al. 2014a; Darragh et al. 2013; Wilkinson et al. 2015). The WHO classification of vulvar SIL is subdivided into two grades: low-grade squamous intraepithelial lesion (LSIL) encompassing LSIL (VIN 1) and mild dysplasia and high-grade

2

Precursor Lesions and Malignant Tumors of the Vulva

Table 1 Classification of SIL of the vulva (VIN) Low-grade squamous intraepithelial lesion, LSIL (VIN 1): a SIL in which nuclear abnormalities are confined to the lowest one-third of the epithelium High-grade squamous intraepithelial lesion, HSIL (VIN 2-3/VIN 3): a squamous intraepithelial lesion in which nuclear abnormalities involve at least the lower two-thirds, to full thickness, of the epithelium Differentiated VIN (dVIN): is not included in the SIL category, because it is not a human papillomavirus (HPV)-related lesion (Crum et al. 2014a; Darragh et al. 2013)

squamous intraepithelial lesion (HSIL) encompassing high-grade VIN (VIN 2/VIN 3), moderate and severe dysplasia, and carcinoma in situ. Differentiated (simplex) VIN is termed “differentiated VIN” (dVIN) and is not included in the SIL terminology because it is not human papillomavirus (HPV) associated (Crum et al. 2014a; Darragh et al. 2013). This terminology is now widely accepted and is applicable provided one’s clinicians and those involved in patient care and funding for care understand its use. The incidence of vulvar SIL and vulvar squamous cell carcinoma in women younger than 60 years of age has significantly increased over the past 50 years, although this increase has not been observed in women over 60 years of age (Barlow et al. 2015; Meltzer-Gunnes et al. 2017). Women with HIV infection, and immunosuppressed women, have a significantly higher incidence of vulvar SIL. The true incidence of SIL/VIN probably is higher, because generally only a subset of HSIL cases are reported. Approximately 50% of these women have other HPV-related squamous lesions involving the genital tract, most often cervical SIL. Women with a prior history of cervical HSIL/CIN3 have an increased risk, for possibly 20 years or longer, of subsequent vulvar HSIL and vulvar carcinoma, as well as increased risk of HPV-related intraepithelial or invasive squamous neoplastic lesions of the vagina, anus, and oropharynx (Ebisch et al. 2017). Approximately one-half of these women have a history of a preexisting or concomitant sexually transmitted disease, of which condylomata acuminata is the most frequent. Nearly all

67

vulvar HSIL are oncogenic HPV positive, with HPV 16 being the most common and accounting for 80% or more of the cases. HPV 33 accounts for approximately 10% of cases, and types 59, 45, and 18, in decreasing frequency of detection, are also identified (Gargano et al. 2012; Saraiya et al. 2015). The estimated mean time from oncogenic HPV vulvar infection to HSIL is 18.5 months in one study (Garland et al. 2009). There is a recognized association between vulvar SIL and cigarette smoking, including recurrence and progression of the lesions (Khan et al. 2009).

Clinical Features Vulvar pruritus and irritation are the most common presenting symptoms. It has become increasingly common for patients to notice the lesions themselves and bring them to medical attention. Approximately one-third of the patients are asymptomatic. The lesions of HSIL typically have a raised surface and are macular or papular (Figs. 1, 2, 3, and 4). The most common sites of vulvar HSIL at presentation involve the labia minora and the perineum.

Fig. 1 Vulvar SIL, clinical appearance (HSIL/VIN 2-3). Multiple macular and plaquelike white areas are present on the labium majus

68

Fig. 2 Vulvar SIL, clinical appearance (HSIL/VIN 2-3). Multiple aceto-white macules are present on the labia majora in this young woman

Fig. 3 Vulvar SIL, clinical appearance (HSIL/VIN 2-3). Multiple warty aceto-white papular VIN lesions are present on the vulva and perianal area

Approximately two-thirds of HSIL lesions are multifocal (Wilkinson and Stone 2012). About one-third of cases of HSIL have perianal involvement, and in such cases the lesions may extend into the anus. When perianal HSIL are identified and there is concern that the HSIL

E. J. Wilkinson and D. S. Rush

Fig. 4 Vulvar SIL, clinical appearance (HSIL/VIN 2-3). Confluent distribution of pigmented, slightly raised, roughsurfaced areas involving the labia majora and minora

have extended higher in the anus, anal cytology can be performed (Darragh et al. 2013). If anal cytology is abnormal, or if otherwise clinically indicated, high-resolution anoscopy can be used to evaluate the extent of the lesions in the anal/ rectal area and perform directed biopsies (Hillman et al. 2016). Although the clinical appearance of vulvar HSIL can be highly variable at presentation, approximately one-half of the patients with HSIL have white lesions, or lesions that are distinctly aceto-white after the application of topical 3–5% acetic acid. Approximately one-quarter are pigmented (Figs. 3 and 4). HSIL is the second most common pigmented vulvar lesion. Pigmented vulvar HSIL usually occurs in keratinized vulvar skin. The remainder of HSIL may be pink, gray, or red. Lesions involving the nonkeratinized mucous membrane of the vestibule usually appear red. Such red lesions have been called erythroplasia of Queyrat, but like Bowenoid papulosis are not designated separately and are included within the SIL category. The lesions may be macular (Fig. 1) or papular (Fig. 2). In approximately three-quarters of cases, they are multiple, and in the majority of the remainder of cases, HSIL is a solitary lesion. Solitary lesions appear to be more common in older women and are more commonly associated with invasive squamous carcinoma.

2

Precursor Lesions and Malignant Tumors of the Vulva

Confluent growth of HSIL is relatively uncommon (Wilkinson and Stone 2012) (see Figs. 1 and 4). The anal skin and squamous mucosa of the anal canal are the most frequently involved secondary sites.

Microscopic Findings The epithelial cells of HSIL have a high nuclearcytoplasmic ratio and lack cytoplasmic maturation above the basal and parabasal layers. Mitotic activity is present above the basal layer, and the mitotic figures are often abnormal in appearance. Multinucleation and dyskeratosis, including formation of intraepithelial squamous pearls, may be seen (Figs. 5 and 6a, b). Nuclear pleomorphism and hyperchromasia are present; however, nucleoli are

69

uncommon. Radial dispersion of nuclear chromatin and coarse nuclear chromatin is seen in the epithelial cells of HSIL, corresponding to an increased number of interchromatinic and perichromatinic granules. The so-called individual cell keratinization seen within the epithelium is attributed to the presence of aggregated intracytoplasmic tonofilaments that may be produced in the process of abnormal cell division. Parakeratosis is present when keratinocytes fail to form granules of prekeratin and retain nuclear material at the epithelial surface. Both intracellular and extracellular pigment granules are distributed throughout the epidermis. In pigmented HSIL lesions, dermal melanophages often are prominent beneath the basal layer and within the dermal papillae, and both intracellular and extracellular pigment granules are distributed throughout the epidermis. The thickness of the epithelium involved by HSIL may range from 0.10 to 1.90 mm, with a mean of 0.52 mm (Benedet et al. 1991). HSIL involves the skin appendages in more than 50% of the cases studied. Skin appendage involvement by HSIL should be differentiated from early invasion (compare Figs. 6b, c, and 7).

Microscopic Grading

Fig. 5 Vulvar HSIL (VIN 2), warty type. In the lower two-thirds of the epithelium, the cells are crowded and lack maturation. Nuclei are hyperchromatic and pleomorphic. Within the upper one-third of the epithelium, there is cellular maturation with prominent koilocytosis

Grading of SIL is performed based on the findings in the most severe, highest grade, area. When the cellular squamous epithelial abnormalities are proliferation and lack of maturation is confined to the lower third of the epithelium, the lesion designated as LSIL/VIN 1 is reported (Fig. 8). LSIL of the vulva is an uncommon and controversial lesion, sometimes designated “atypical condyloma acuminatum” or “flat condyloma acuminatum,” although the use of such terms is not recommended. These lesions may have koilocytosis and minimal evidence of proliferation, but lack the exophytic growth pattern and more pronounced cellular atypia of condyloma acuminatum. Flat condyloma acuminatum of the vulva may be included in the LSIL category because the biologic difference between LSIL and flat condyloma is unknown and the morphologic distinction is unreliable. In addition,

70

Fig. 6 (a) Vulvar HSIL (VIN 3) with superficially invasive squamous cell carcinoma. Squamous differentiation within the HSIL is manifested by rounded foci of cells with eosinophilic cytoplasm near the basal layer. This is a useful feature in identifying early invasion. (b) HSIL with superficial invasion. Keratinization is present within the VIN.

E. J. Wilkinson and D. S. Rush

Small clusters of invasive squamous cell carcinoma are present in the superficial dermis. (c) HSIL with superficial invasion. A small focus of invasive squamous carcinoma is present at the base of the dermal papillae. Note the loss of palisading of the basal cells, as compared to those in the adjacent dermal papillae

2

Precursor Lesions and Malignant Tumors of the Vulva

Fig. 7 HSIL (VIN 3), involving a skin appendage. Cellular disarray with lack of maturation is present within the epithelium. Part of a sebaceous gland is present at the base of the lesion

71

typical condyloma acuminatum associated with LSIL or HSIL, a diagnosis of “LSIL with condyloma acuminatum” or “HSIL with condyloma acuminatum” is appropriate (Crum et al. 2014a). HSIL (encompassing VIN 2-3) is diagnosed when the cellular evidence of proliferation and lack of maturation extends through at least the basal third of the epithelium and may involve essentially the full thickness of the epithelium. Cellular changes may, or may not, include the surface layers above the granular zone (Figs. 5, 9, and 10). Abnormal mitoses are nearly always present in HSIL and may be seen in all but the most superficial layers of the epithelium. Lack of abnormal mitoses, or lack of mitosis above the basal layer, should raise the question as to whether a lesion belongs in the HSIL category. Vulvar SIL are predominately HSIL, whereas LSIL are relatively rare. With HSIL, Ki-67 (MIB-1) demonstrates nuclear reactivity throughout most of the epithelium (Fig. 9b). Immunohistochemical study with p16INK4a demonstrates “blocklike” reactivity in nearly the full thickness of the neoplastic epithelium in most cases (Darragh et al. 2013; Crum et al. 2014a).

Histologic Subtypes of Vulvar SIL

Fig. 8 LSIL (VIN 1). Crowding of the basal and parabasal cells with some cellular disarray and loss of maturation within the lower one-third of the epithelium. Koilocytotic cells are present on the surface

although LSIL are associated with high-risk HPV in approximately 40% of cases studied (Srodon et al. 2006), they appear to progress to high-grade lesions rarely, if ever. In the occasional case with

Vulvar SIL have several histologic types including warty (condylomatous), basaloid, and mixed warty and basaloid, classified on the basis of cellular features. Different morphologic types are occasionally found in one patient, and more than one type may be found in a single lesion. Mixtures of basaloid and warty patterns are particularly frequent. These “mixed” cases are classified according to the predominant component or simply as HSIL, basaloid/warty type. There is also a rare HSIL with pagetoid features (Raju et al. 2003). Some HSIL are not readily classifiable into one of these groups. Interobserver variability in diagnosing SIL is recognized (Preti et al. 2000). Reproducibility for subtyping SIL lesions is fair (kappa, 0.31–0.42) (Trimble et al. 1999). Warty (condylomatous) HSIL/VIN 2/3 lesions have an epithelial surface that is typically undulating or “spiked” and is often keratinized with

72

E. J. Wilkinson and D. S. Rush

Fig. 9 (a) HSIL (VIN 3), basaloid type. Prominent acanthosis and basaloid-type neoplastic keratinocytes involve nearly the full thickness of the epithelium. (b) HSIL (VIN 3), basaloid type. Complete replacement of

the epithelium with overlying parakeratosis similar to HSIL/CIN 3 of cervix. (Photograph originally published in Rush and Wilkinson (2016))

hypergranulosis and parakeratosis. There is usually prominent acanthosis with rete pegs that are relatively wide and extend deeply into the dermis. Dermal papillae are thin and can be close to the surface. The cells show evidence of maturation with prominent parabasal hyperplasia. The epithelial cells have well-defined cell membranes and prominent eosinophilic cytoplasm. Individual cell keratinization and small cells with eosinophilic cytoplasm, “corps ronds,” are often present. Multinucleated epithelial giant cells may also be present. The nuclei are enlarged and pleomorphic, with coarsely granular chromatin and increased nuclear-cytoplasmic ratio. Nucleoli are not prevalent. Koilocytes, characterized by cells with hyperchromatic, shrunken nuclei surrounded by a clear halo separating the nucleus from the cytoplasm, are characteristic (Figs. 5 and 10). The warty type

of HSIL has larger cells with greater nuclear pleomorphism than basaloid HSIL. Abnormal mitosis is usually identifiable (Fig. 10b). The basaloid type of HSIL has a thickened epithelium with a relatively smooth and flat surface without the undulated and spiked appearance of warty HSIL. Hyperkeratosis is often present, but less extensive than that seen in warty HSIL. The epithelium lacks cellular maturation and is often nearly entirely composed of atypical parabasal-type cells. These cells are relatively small, uniform cells with hyperchromatic and coarse nuclear chromatin. Nucleoli are rare, mitosis is common, and abnormal mitoses are usually present. Although there is little or no maturation of the keratinocytes, some keratinization, or parakeratosis, as well as koilocytosis may be present near the surface (see Fig. 9).

2

Precursor Lesions and Malignant Tumors of the Vulva

Fig. 10 (a) HSIL (VIN 3), warty (condylomatous) type. The surface of the lesion is spiked with marked hyperkeratosis. A prominent granular layer is evident. (Photograph originally published in Rush and Wilkinson (2016). (b) HSIL (VIN 3), warty (condylomatous) type. Marked cellular disarray with prominent nuclear pleomorphism. Several multinucleated keratinocytes are present

Both warty and basaloid HSIL may have intraepithelial growth that may extend into adjacent hair follicles and other skin adnexal structures. HSIL involves the skin appendages in more

73

than 50% of the cases studied. Skin appendages as deep as 2.7 mm in hair-bearing areas may be involved with HSIL, and care should be taken to differentiate this from early invasion (compare Figs. 6b, c and 7). In non-hair-bearing areas, the skin appendages and minor vestibular glands are more superficial. Basaloid and/or warty HSIL may be present adjacent to and/or associated with invasive squamous cell carcinoma (Kurman et al. 1993). The invasive carcinoma may be of the basaloid or warty types corresponding to the HSIL type, although the converse may occur with basaloid HSIL adjacent to warty carcinoma or warty HSIL adjacent to basaloid carcinoma. Basaloid HSIL and basaloid carcinoma are typically associated with oncogenic HPV, but can resembled dVIN. Nine (7%) of reported “differentiated” VIN cases in one study were initially interpreted as basaloid HSIL; however, they were negative for HPV and did not express p16INK4a, but did express p53 and were reinterpreted as variants of dVIN (Fuste et al. 2009). Lesions that resemble basaloid HSIL, but are HPV negative, are consistent with a variant of dVIN and are better classified as dVIN (Ordi et al. 2009). Pagetoid VIN is a very rare lesion with reports limited to a few cases (Raju et al. 2003). It has a pagetoid growth pattern of the neoplastic intraepithelial epithelial cells that are found in groups and nests within otherwise normal appearing epithelium. The cells of pagetoid VIN may resemble Paget disease of cutaneous type, or superficially spreading melanoma, having relatively pale cytoplasm and larger nuclei than the adjacent normal keratinocytes with relatively clear chromatin and evident nucleoli. Pagetoid VIN is distinguished from Paget disease or melanoma in situ by immunohistochemical study (see discussions below and Table 2). (Raju et al. 2003; Wilkinson and Brown 2002).

Differential Diagnosis The differential diagnosis of warty or basaloid SIL includes basal cell carcinoma, superficial

74

E. J. Wilkinson and D. S. Rush

spreading malignant melanoma, Paget disease, pagetoid urothelial intraepithelial neoplasia (PUIN), multinucleated atypia of the vulva, vulvar acanthosis with altered differentiation (VAAD), and podophyllin treatment effect (Wilkinson and Brown 2002; Nascimento et al. 2004; Al-Bannai et al. 2015). Basaloid HSIL may resemble dVIN; however, dVIN is HPV negative (Ordi et al. 2009) (Table 2). Immunohistochemical methods may be useful to assist in distinguishing melanoma and Paget disease from HSIL as listed in Table 2. Multinucleated atypia of the vulva has multinucleated keratinocytes, without significant nuclear atypia, within the lower to middle epithelial layers. VAAD has verruciform growth with wide and deep rete, with some maturation disorder of the keratinocytes with cytoplasmic pallor but without significant keratinocyte atypia. In a study of ten cases, seven had associated lichen simplex chronicus and one had lichen sclerosus. These authors propose that it may represent a percursor lesion to verrucous carcinoma of the vulva (Nascimento et al. 2004); however, in one case it was associated with progression to poorly differentiated vulvar carcinoma (Al-Bannai et al. 2015). The possibility that podophyllin effect on condylomata acuminata of the vulva could change these lesions and result in misinterpretation as HSIL is highly improbable because the changes from a single application of podophyllin regress within 1–2 weeks. Mitotic arrest with cells in metaphase is present after podophyllin and contrasts with the abnormal mitotic figures seen in HSIL. Nuclear karyorrhexis is rarely present in HSIL,

whereas it is present in most of the treated condyloma cases. In contrast to podophyllin effect, in HSIL nuclear size tends to be variable, and the nuclear chromatin usually is coarse with little cellular swelling (see ▶ Chap. 1, “Benign Diseases of the Vulva”). Despite the association with keratinizing squamous cell carcinoma, there are no longitudinal studies demonstrating that squamous cell hyperplasia alone is a precursor of squamous cell carcinoma (Kim et al. 1996). Women with vulvar lichen sclerosus are at risk for developing vulvar squamous cell carcinoma (Micheletti et al. 2016). In such cases it is not unusual to find prominent acanthosis with hyperkeratosis involving vulvar epithelium near the tumor site. dVIN is associated with lichen sclerosus and is a recognized precursor to squamous cell carcinoma in some cases (Micheletti et al. 2016; Rush and Wilkinson 2016; Yang and Hart 2000).

Adjunctive Studies Evidence for HPV, most commonly type 16, is present in most SIL warty/basaloid lesions, based on molecular biologic studies (Gargano et al. 2012; Samama et al. 2006; Saraiya et al. 2015). Alterations in the p16/pRb/cyclin D1 pathway occur in HSIL as demonstrated by immunohistochemical methods. Epigenetic silencing of p16INK4a occurs in approximately two-thirds of basaloid or warty HSIL (Santos et al. 2004). Immunohistochemical expression of p16INK4a was observed in 92% of HSIL in one study, and

Table 2 Immunohistochemical studies to differentiate vulvar Paget disease of skin or rectal origin, PUIN, and melanoma

As a primary skin neoplasm Related to anorectal carcinoma PUIN related to urothelial carcinoma Melanoma

CK7 + + + 0

CK20 0 + +(0) 0

GCDFP-15 + 0 0 0

CEA + + 0 0

S-100, HMB-45, and Melan-A 0 0 0 +

UPK II or III 0 0 + 0

CK7 cytokeratin 7, CK20 cytokeratin 20, GCDFP-15 gross cystic disease fluid protein-15, CEA carcinoembryonic antigen: UPK uroplakin II or III (Wilkinson and Brown 2002; Newsom et al. 2015)

2

Precursor Lesions and Malignant Tumors of the Vulva

these cases had nearly the full thickness expression of p16INK in the involved epithelium. In LSIL/VIN 1 lesions, two of ten were immunoreactive, and in these cases the reactivity was limited predominately within the lower half of the epithelium (Rufforny et al. 2005). Oncogenic HPV is typically not associated with dVIN and, if identified, suggests that the lesion is more probably a basaloid HSIL and not dVIN (Ordi et al. 2009). Approximately a third of cases demonstrated overexpression of cyclin D1 in a series of 13 cases of HSIL. None of these cases demonstrated lack of pRb protein (Lerma et al. 2002). Differentiated VIN lesions usually express p53 in the basal, and some parabasal cells and are not associated with oncogenic HPV (Yang and Hart 2000). Cellular proliferation analysis employing immunohistochemical studies for BCL-2 and Ki-67 (MIB-1) does not appear to be of much value in the diagnosis of LSIL and HSIL although Ki-67 may have some value in grading and identifying low-grade lesions (Logani et al. 2003). Proliferation studies using Phh3 in 18 cases of dVIN did not find such study to be of value in diagnosis (Loch et al. 2016). Most HSIL contain DNA aneuploid populations of cells (Wilkinson et al. 1981). DNA analysis by microspectrophotometry of multifocal lesions suggests that separate lesions arise from separate stem cells, forming distinguishable clones. Large confluent lesions may result from centrifugal growth from a single cell line or by confluence of separate and distinct clones. In single HSIL, approximately half have different stem cells by DNA microspectrophotometry, suggesting that such lesions may undergo clonal evolution (Wilkinson et al. 1981).

Clinical Behavior and Treatment The clinical course of HSIL following treatment is well studied; however, there are relatively few long-term studies of untreated HSIL. There is evidence that untreated HSIL will progress to invasive squamous cell carcinoma, and this occurs with invasion observed within 8 years of the diagnosis of HSIL (Jones et al. 2005).

75

HSIL associated with vulvar invasive squamous cell carcinoma diagnosed within excised HSIL occurs in 2–20% of lesions. These HSIL cases with associated invasion usually do not have a typical clinical appearance, even when vulvoscopy is employed prior to biopsy (Preti et al. 2017). Findings in the HSIL cases that were associated with invasive carcinoma included involvement of the clitoris, size of 20 mm or larger, a nodular lesion, and a lesion in older women (Preti et al. 2017). In evaluating epithelial changes adjacent to vulvar squamous cell carcinoma, 60–80% of superficially invasive squamous cell carcinomas and 25% of deeply invasive carcinomas have adjacent HSIL (Yoder et al. 2008). dVIN was reportedly associated with vulvar squamous cell carcinoma in 7 of 12 (58%) patients in one study (Yang and Hart 2000). In a study of 18 patients with dVIN, 14 (78%) had associated invasive squamous cell carcinoma (Loch et al. 2016). These and other studies suggest that there is a stronger association between dVIN and squamous cell carcinoma than there is between the usual recognized HSIL types and invasion (Bigby et al. 2016). There are no known differences in clinical behavior between warty and basaloid HSIL. Squamous cell carcinoma of the vulva associated with HSIL occurs most commonly in postmenopausal women, although it may occur in women of reproductive age and immunecompromised women. Vulvar HSIL is associated with vulvar condyloma acuminatum and HPV-related squamous cell carcinoma. In contrast, dVIN is associated with vulvar dermatoses including lichen sclerosus, and lichen planus, and is associated with carcinoma. In a study of 584 women with vulvar lichen planus, 1.6% had associated HSIL, and one of these patients subsequently had vulvar carcinoma. In contrast, among these women with lichen planus, 16 had dVIN, and of these 9 of the 16 had progression to invasive squamous cell carcinoma (Ragauer et al. 2016). Regression of HSIL appears to be most common in young women and those who are pregnant at the time that the HSIL are identified. In contrast women of advanced age, those who are severely immunosuppressed, and

76

women with Fanconi anemia are at a greater risk for invasion (Saraiya et al. 2008; Wilkinson et al. 1984). Conservative therapy for HSIL such as local superficial excision or laser ablation may be appropriate for selected patients, especially when the HSIL involves the non-hairy skin or mucous membrane areas (e.g., the vulvar vestibule and perianal areas). Topical imiquimod has gained some acceptance in the treatment of vulvar condyloma acuminatum and HSIL (off label) (ACOG Committee on Gynecologic Practice and ASCCP 2016; Ragauer et al. 2016; van Seters et al. 2008). Women who are cigarette smokers appear to have a higher frequency of recurrence of HSIL, and cessation of smoking is advised (Khan et al. 2009). Recurrence after treatment of vulvar HSIL occurs in approximately one-third of the patients, and approximately three-quarters of the recurrences occur within 3 years. A preliminary study suggested that those with HPV 16 antibodies had a lower rate of recurrence (Madeleine et al. 2016).

Differentiated Vulvar Intraepithelial Neoplasia

E. J. Wilkinson and D. S. Rush

degree of acanthosis, with thickened epithelium, prominent intercellular bridges, and elongation and anastomosis of rete pegs. The keratinocytes of dVIN are large and pleomorphic with a relatively large amounts of eosinophilic cytoplasm, as compared to other types of HSIL (Fig. 11), including in the basal and parabasal layers and sometimes limited to the base of rete ridges (Rush and Wilkinson 2016; Lerma et al. 2002; Yang and Hart 2000). Basal layer atypia was essential for the diagnosis in one survey of pathologists (Reutter et al. 2016). The nuclear chromatin is vesicular rather than coarse, and the nuclei have prominent nucleoli, usually most prominent in the basal and parabasal keratinocytes (Reutter et al. 2016; Mulvany and Allen 2008) (Fig. 11). Keratin pearl formation is common in dVIN but is not common in other VIN lesions.

Differential Diagnosis dVIN should be distinguished from associated lichen simplex chronicus or lichen sclerosus, repair or reaction related to erosion or superficial ulceration including pseudo-epitheliomatous (pseudo-carcinomatous) hyperplasia, vulvar

General Features Differentiated VIN (dVIN, VIN of simplex type) is significantly less common than HSIL. Patients with dVIN are usually postmenopausal and often suffer from associated lichen sclerosus or lichen simplex chronicus. It is often not recognized until invasive carcinoma has developed, when it is then identified adjacent to the carcinoma.

Clinical Features Microscopic Findings dVIN is difficult to identify, largely because, unlike in HSIL, the changes in dVIN are limited to the lower portion of the epithelium and the more superficial portions may appear relatively normal. It is usually associated with some

Fig. 11 dVIN with invasive squamous carcinoma. The epithelium has slight cellular disarray, and the keratinocytes contain abundant cytoplasm. Within the parabasal area of the epithelium, the keratinocytes have increased eosinophilic cytoplasm and are dyskeratotic, and the nuclei are enlarged and contain prominent nucleoli

2

Precursor Lesions and Malignant Tumors of the Vulva

acanthosis with altered differentiation (VAAD), trapped epithelium at a prior biopsy site, and granulomatous or decidual change in the immediately adjacent underlying epithelium (Reutter et al. 2016). A similar lesion to dVIN may occur overlying melanocytic nevi (Michalova et al. 2017). Another similar lesion has been designated by some as atypical squamous cell hyperplasia because this lesion has similar nuclear atypia but histologically resembles squamous cell hyperplasia more closely than it does dVIN (Kurman et al. 1993). Despite the association with keratinizing squamous cell carcinoma, there are no longitudinal studies demonstrating that squamous cell hyperplasia alone, as already defined here, is a precursor of squamous cell carcinoma.

Adjunctive Studies dVIN is typically not associated with HPV (see Fig. 6) (Crum et al. 2014a; Darragh et al. 2012) and does not stain for p16 on immunohistochemistry. dVIN lesions have been reported to express p53 in the basal and some parabasal cells in most cases (Yang and Hart 2000), but this staining pattern is relatively subtle and interpretation is subjective. The search for a better immunohistochemical marker is still ongoing. Proliferation studies using Phh3 in 18 cases of dVIN did not find such study to be of value in diagnosis (Loch et al. 2016). Immunohistochemical study for cytokeratin 17 (CK-17) in a study of 29 cases with dVIN demonstrated that the dVIN lesion was strongly and diffusely immunoreactive in the suprabasal layer or throughout the epithelium in 27 (93%) of the cases (Podoll et al. 2017). This finding was of value in distinguishing dVIN from lichen simplex chronicus, which expressed only superficial or focal immunoreactivity with CK-17. An immunohistochemical study of p53 in 14 dVIN lesions, ten of which were associated with vulvar invasive carcinoma, demonstrated loss of p53 expression in the invasive carcinomas in all cases. Corresponding loss of p53 was observed in the adjacent dVIN lesions in five cases. In eight cases p53 loss was observed in epithelium adjacent to the p53-positive dVIN or carcinoma, but did not fulfill the diagnosis of

77

dVIN, suggesting cellular changes not identifiable by histopathology. In one case the adjacent epithelium at the margin of excision had patchy p53 expression and on review was reinterpreted as not dVIN, and the margin reclassified as negative (Singh et al. 2015). These findings suggest that p53 and CK-17 immunohistochemical studies can be reasonable adjuncts to evaluate the intraepithelial changes suggesting dVIN and to evaluate margins where a dVIN lesion is in the differential diagnosis.

Clinical Behavior and Treatment dVIN is a neoplastic lesion with a high risk for developing into squamous cell carcinoma (Bigby et al. 2016). dVIN was reportedly associated with vulvar squamous cell carcinoma in 7 of 12 (58%) patients in one study (Yang and Hart 2000). In another study of 18 patients with dVIN, 14 (78%) had associated invasive squamous cell carcinoma (Loch et al. 2016). These and other studies suggest that there is a stronger association between dVIN and squamous cell carcinoma than there is between the usual recognized HSIL types and invasion (Bigby et al. 2016). Invasive squamous carcinomas found in association with dVIN are usually of keratinizing type. The association of lichen sclerosus and dVIN has yet to be completely understood. Women with vulvar lichen sclerosus are known to be at risk for developing vulvar squamous cell carcinoma (Micheletti et al. 2016). In such cases it is not unusual to find prominent acanthosis with hyperkeratosis involving vulvar epithelium near the tumor site, and in some cases lichen sclerosus is found in the background of dVIN adjacent to squamous cell carcinoma (Micheletti et al. 2016; Rush and Wilkinson 2016; Yang and Hart 2000). Recent studies suggest that lichen sclerosus may progress to carcinoma through a dVIN intermediary. In a retrospective study of 976 women with vulvar lichen sclerosus, 34 women developed vulvar neoplasia over time, with 10 having dVIN. Of these ten, four had lichen sclerosus with dVIN alone, five had associated superficially invasive vulvar carcinoma, and one had associated

78

keratinizing carcinoma. With the follow-up of these women, the probability of progression to neoplasia rose from 1.2% at 24 months to 36.8% at 300 months. The overall vulvar neoplasia incidence rate was estimated at 3.5% (Micheletti et al. 2016). Effective treatment and management may reduce the carcinoma risk associated with lichen sclerosus (Lee et al. 2015). Vulvar lichen planus may also be associated with dVIN and vulvar carcinoma risk. In a study of 584 women with vulvar lichen planus, 1.6% had associated HSIL, and one of these patients subsequently had vulvar carcinoma. In contrast, among these women with lichen planus, 16 had dVIN, and of these, 9 had progression to invasive squamous cell carcinoma (Ragauer et al. 2016). Treatment for dVIN is the same as for HSIL.

Squamous Cell Carcinoma Epidemiologic data on vulvar carcinomas is somewhat limited due to nonspecific coding and classification in some registries (Siegel et al. 2017). In a study of 39 population-based cancer center registries, squamous cell carcinoma accounted for 75% of invasive malignant tumors of the vulva and HSIL accounted for 77% of the intraepithelial neoplasms (Saraiya et al. 2008). For 2016, the American Cancer Society (ACS) estimates 5,950 new cases of vulvar squamous cell carcinoma annually in the United States and 1,110 deaths (Siegel et al. 2016), a significant increase over 2008, when the ACS had estimated 3,400 new cases (American Cancer Society 2017). Incidence rates for oncogenic HPV-related tumors are increasing in the oropharynx, anus, and vulva, and current low rates of HPV vaccination in the United States, of under one-third of the eligible population, do not favor that this incidence will decrease over time (Jemal et al. 2013). In a large retrospective study, 69% of the vulvar carcinoma cases were HPV positive and 48% were HPV 16 positive, with approximately 10% being HPV 33 positive and types 52, 18, and 31 being other types in decreasing order of frequency (Saraiya et al. 2015). In addition, other mucosal oncogenic HPV types have

E. J. Wilkinson and D. S. Rush

been reported including, but not limited to, HPV 26, 66, 67 68, 70, and 73 (Halec et al. 2017). For black or white women under 50 years of age, the age-specific rates of vulvar squamous cell carcinoma are approximately the same; however, beyond 50 years of age, the rate of vulvar squamous cell carcinoma increases, with the rate increase greater in white than in black women. Overall white women have the highest incidence of vulvar carcinoma; the incidence rates for black and Hispanic women are approximately one-third that of white and non-Hispanic women (Saraiya et al. 2008). Some European countries have also reported their data on vulvar squamous cell carcinoma and have observed an increasing incidence with age. Epidemiologic data on vulvar carcinomas is somewhat limited due to nonspecific coding and classification in some registries (Siegel et al. 2017). The mean age at presentation is 60–74 years (Kurman et al. 2010). Vulvar carcinoma is found to present at a younger age in black women than white women with a 10-year age difference between the two groups; however, black women had a better overall survival (Rauh-Hain et al. 2013). In addition, non-white women were found to be at lower risks for vulvar neoplasia than white women (Brinton et al. 2017). Vulvar squamous cell carcinoma has been described at 12 years of age (Rabah and Farmer 1999; Al-Ghamdi et al. 2002). Epidemiologic studies of women with vulvar carcinoma have identified an increased risk with prior or current vulvar HSIL, condyloma acuminatum, and cervical carcinoma, indicating that oncogenic HPV is usually a significant factor (Gargano et al. 2012). Vulvar lichen sclerosus is also associated with an increased lifetime risk of vulvar carcinoma and the risk increases with age (Micheletti et al. 2016). These lichen sclerosusassociated tumors do not appear directly related to oncogenic HPV, but may be influenced by autoimmune issues related to lichen sclerosus in older women. Lichen planus is also a recognized factor, and although not as common as lichen sclerosus, in a study of 584 women with vulvar lichen planus, 10 subsequently developed vulvar carcinoma, of which only 1 patient had HSIL. Of the 584 patients, 16 had associated dVIN, and 9 of the 10 developed

2

Precursor Lesions and Malignant Tumors of the Vulva

vulvar invasive squamous cell carcinoma (Ragauer et al. 2016). Additional related factors include other chronic inflammatory conditions and dermatoses such as lichen simplex chronicus, hidradenitis suppurativa, chronic genital granulomatous disease, and granuloma inguinale. Other factors include older age, number of lifetime sexual partners, cigarette smoking, immunodeficiency, diabetes mellitus, achlorhydria, poor perineal hygiene, and obesity (Brinton et al. 2017; Redman et al. 2005; Short et al. 2005). Vulvar carcinoma may occur during pregnancy, although parity does not appear to be a significant risk factor. Occupational risks are also recognized including, but not limited to, women cotton mill workers, women with industrial exposure to industrial oils, and those working in the dyeing and cutlery industries. Topical exposure to arsenicals also increases risk. A possible increased frequency has been suggested for women with blood group A, but findings remain inconclusive (Redman et al. 2005). Evidence supports the view that women at risk for vulvar carcinoma are in one of two main groups: first, those who have oncogenic HPV-associated squamous lesions, specifically HSIL, that usually express p16INK4a and, second, those with vulvar disease not associated with HPV and that usually do not express p16INK4a, including dVIN, those with a vulvar dermatoses including lichen sclerosus and lichen planus, and those with a chronic inflammatory vulvitis, such as chronic granulomatous disease (Micheletti et al. 2016: Ragauer et al. 2016). In a retrospective study using molecular HPV testing methods, of 88 women with vulvar HSIL/VIN 3, the median age was 57 years old, and of these 64 (94.1%) had HSIL that had oncogenic HPV. Of these, 55 (80.9%) were HPV 16 positive and 9 had other HPV types including 8.8% with HPV 33 and 2.9% with HPV 59 (Gargano et al. 2012). In a large European study, of 587 vulvar SIL/VIN cases, 86.7% had detectable HPV (de Sanjosé et al. 2013). The frequency of oncogenic HPV-associated vulvar squamous cell carcinoma is different between young and older women. HPV, usually type 16, is reported in less than one-fifth of vulvar carcinomas in older women (mean age 77 years),

79

whereas approximately four-fifths of the vulvar squamous carcinomas identified in younger women (mean age 50 years) are HPV associated. The histopathologic type of invasive carcinoma also differs. In the older women, vulvar squamous cell carcinomas are usually well differentiated and highly keratinized. In younger women the tumors are more commonly warty or basaloid carcinomas (Kurman et al. 1993). In a study of 177 women with invasive vulvar carcinoma, the mean age was 75 years of age, in contrast to the HSIL group with a mean age of 57. Of the 118 tumors studied from these women, 66.7% had oncogenic HPV detected; of these 86 (48.6%) had HPV 16 detected. Other types of oncogenic HPV identified in the additional cases included HPV 33 (10.2%) and HPV 52 (2.8%) (Gargano et al. 2012). In a similar study of 1709 invasive squamous carcinoma cases, 28.6% had associated HPV, and of these HPV 16 was detected in 72.5%, with HPV 33 in 6.6% and HPV 18 in 4.6% (de Sanjosé et al. 2013). In a study of 177 women with vulvar carcinoma, those with associated oncogenic HPV-related tumors were significantly younger (mean age 64 years) than women whose vulvar carcinomas were not HPV associated (mean age 81 years) (Gargano et al. 2012). In another study of 197 vulvar carcinoma patients, using p16 (CDKN2A) as a surrogate biomarker for oncogenic HPV, 79 had p16-positive tumors and were considered HPV associated (mean age 58.8 years); and 118 had p16-negative tumors (mean age 71.6 years). In addition, the women with p16-positive, oncogenic HPV-associated tumors had a significantly better overall survival, disease-specific survival, and progression-free survival as compared to those with p16-negative tumors (McAlpine et al. 2017). In a study of 97 women with vulvar squamous carcinoma, in addition to early clinical stage, p16 immunoreactivity and absent p53 reactivity were independent predictors of better overall survival (Dong et al. 2015). Older women with vulvar carcinoma (mean age, 77 years) typically do not have associated HSIL or a history of heavy cigarette smoking. Their tumors rarely contain HPV and usually are

80

well-differentiated keratinizing squamous cell carcinomas (Table 3) (Kruse et al. 2008; Kurman et al. 1993). Suppressed immune competence can be a risk factor in these women. The WHO classifies vulvar squamous carcinomas into five subtypes including keratinizing, nonkeratinizing, basaloid, warty, and verrucous (Crum et al. 2014a). In addition to these, other types have been identified and these are listed with the WHO types in Table 3. Women with non-HPV-related, vulvar squamous cell carcinoma often have associated vulvar dermatoses, especially lichen sclerosus (Carlson et al. 1998; Loch et al. 2016; Micheletti et al. 2016). Primary evidence based on p53 mutation analysis and clonal studies suggests that squamous cell hyperplasia of the vulva is probably not a precursor of non-HPV-related vulvar squamous cell carcinomas (Kim et al. 1996). Women with vulvar lichen sclerosus who develop squamous cell carcinoma tend to be older; their primary tumors more commonly involve the clitoris and typically are rarely associated with HSIL of the warty or basaloid types, but may have dVIN. It is recognized that there is an underreporting of vulvar skin diseases related to vulvar carcinoma. Squamous cell carcinomas associated with lichen sclerosus express tumor suppressor gene product p53 in nearly one-half of the cases

E. J. Wilkinson and D. S. Rush

and cytokine transforming growth factor-beta (TGF-b) in one-third of the cases as compared to 19% and 9%, respectively, for non-lichen sclerosusrelated tumors. Approximately one-half of these lichen sclerosus-associated tumors are associated with a prominent fibromyxoid stromal response, as compared to non-lichen sclerosus-related tumors (Carlson et al. 1998). Of 976 women with lichen sclerosus, 34 (3.5%) were subsequently found to have vulvar neoplasia. Ten of these women had dVIN and six of the ten had associated invasive carcinoma, of which five were superficially invasive. Progression to neoplasia rose from 1.2% at 24 months to 36.8% at 300 months (Micheletti et al. 2016).

Stage IA Invasive Squamous Cell Carcinoma (AJCC T1a, M0, N0; FIGO Stage IA)

General Features The majority of vulvar squamous cell carcinomas in the United States are stage I, being 2 cm or less in diameter, and clinically confined to the vulva without evidence of extension to other sites or lymph node metastasis (Tables 4, 5, and 6). Stage I vulvar carcinomas are divided into two subgroups based on depth of invasion. Stage 1A vulvar carcinoma is a superficially invasive squamous cell carcinoma Table 3 Histologic subtypes of squamous vulvar with a depth of invasion of 1 mm or less and a carcinoma diameter of 2 cm (20 mm) or less. The staging of Keratinizing carcinoma vulvar carcinoma as recommended by the AmeriNonkeratinizing carcinoma can Joint Committee on Cancer (AJCC) is summaBasaloid carcinoma rized in Tables 3, 4 and 5 (Gibb et al. 2016). In that Warty (condylomatous) carcinoma vulvar squamous cell carcinoma is relatively Verrucous carcinoma uncommon, there is evidence that pathology Giant cell carcinoma review can be of value to identify clinically releSpindle cell carcinoma vant pathologic findings. In one study, pathology Acantholytic squamous cell carcinoma (adenoid review of 316 cases where the original and a squamous carcinoma) review report were available and compared, Lymphoepithelioma-like carcinoma Plasmacytoid squamous cell carcinoma 55 (17%) of reports had at least 1 change related Basal cell carcinoma to reporting of the presence of invasion, depth of Metatypical basal cell carcinoma (basosquamous invasion, vascular space invasion, and/or margin carcinoma) status (Barbera et al. 2017). Adenoid basal cell carcinoma A stage Ia ((International Federation of GyneSebaceous cell carcinoma cology and Obstetrics FIGO) stage IA; AJCC T1a) NOS, not otherwise specified carcinoma of the vulva is defined as a single lesion

2

Precursor Lesions and Malignant Tumors of the Vulva

81

Table 4 AJCC 2017 staging of vulvar carcinoma When pathologic staging is designated “pT”; clinical staging is designated “T” Definition of primary tumor (T) T category FIGO stage T criteria TX: Primary tumor cannot be assessed T0: No evidence of primary tumor T1: I Tumor confined to the vulva, and/or perineum multifocal lesions should be designated as such. The largest lesion or the lesion with the greatest depth of invasion will be the target lesion identified to address the highest PT stage. Depth of invasion is defined in the measurement of the tumor from the epithelial-stromal junction of the adjacent-most superficial dermal papilla to the deepest point of invasion T1a IA Lesions 2 cm or less, confined to the vulva and/or perineum, and with stromal invasion of 1.0 mm or less T1b IB Lesions more than 2 cm or any size with stromal invasion of more than 1.0 mm confined to the vulva and/or perineum T2: II Tumor of any size with extension to adjacent perineal structures (lower/distal third of vagina, anal involvement. T3: IVA Tumor of any size with extension to any of the following: upper/proximal two-thirds of the urethra, upper proximal two-thirds of the vagina, bladder mucosa, or rectal mucosa or fixed to pelvic bone From AJCC Cancer Staging Manual, 8th ed. New York: Springer Science + Business; 2017 (Gibb et al. 2016, p. 636). Used with permission of the American Joint Committee on Cancer (AJCC), Chicago, Illinois. The original and primary source for this information is the AJCC Cancer Staging Manual, Eighth Edition (2017) published by Springer International Publishing

Table 5 Lymph node status and metastasis staging of vulvar tumors using the AJCC staging system Definition of regional lymph node (N) N category FIGO stage N criteria NX: Regional lymph nodes cannot be assessed N0: No regional lymph node metastasis NO(i+): Isolated tumor cells in regional lymph node(s) not greater than 0.2 mm N1: III Regional lymph node metastasis with one or two lymph node metastases each less than 5 mm or one lymph node metastasis = 5 mm N1aa IIIA One or two lymph node metastases each less than 5 mm N1b IIIA One lymph node metastasis = 5 mm N2: Regional lymph node metastasis with three or more lymph node metastases each less than 5 mm, or two or more lymph node metastases = 5 mm, or lymph node(s) with extranodal extension N2aa IIIB Three or more lymph node metastases each less than 5 mm N2b IIIB Two or more lymph node metastases = 5 mm N2c IIIC Lymph node(s) with extranodal extension N3 IVA Fixed or ulcerated regional lymph node metastasis Note: The site, size, and laterality of lymph node metastases should be recorded From AJCC Cancer Staging Manual, 8th ed. New York: Springer Science + Business; 2017 (Gibb et al. 2016, p. 637). Used with permission of the American Joint Committee on Cancer (AJCC), Chicago, Illinois. The original and primary source for this information is the AJCC Cancer Staging Manual, Eighth Edition (2017) published by Springer International Publishing a Includes micrometastasis, N1mi and N2mi

82

E. J. Wilkinson and D. S. Rush

Table 6 With the AJCC system, the presence or absence of metastasis: recorded as (M) Definition of distant metastasis (M) M category FIGO stage M criteria M0: No distant metastasis (no pathological M0; use clinical M to complete stage group)) M1: IVB Distant metastasis (including pelvic node metastasis) From AJCC Cancer Staging Manual, 8th ed. New York: Springer Science + Business; 2017 (Gibb et al. 2016, p. 637). Used with permission of the American Joint Committee on Cancer (AJCC), Chicago, Illinois. The original and primary source for this information is the AJCC Cancer Staging Manual, Eighth Edition (2017) published by Springer International Publishing

measuring 2 cm or less in diameter with a depth of invasion of 1 mm or less regardless of the presence of vascular invasion. Patients with tumors with more than one site of invasion, the tumor with the deepest invasion and highest stage is the stage assigned (Gibb et al. 2016). The term “microinvasive carcinoma” is not recommended by the WHO, the ISSVD, or the AJCC because there is no agreement on the definition of the term. In a multivariate retrospective analysis of 78 cases considered as “microinvasive” vulvar squamous carcinoma using a tumor depth of 5 mm or less, 28 (36%) were pathologic stage IA, 40 (51%) pathologic stage IB, 6 (8%) pathologic stage II, and 4 (5%) pathologic stage III (Yoder et al. 2008). The term “superficially invasive carcinoma,” adapted by the Lower Anogenital Squamous Terminology (LAST) committee of the CAP, is not specific and not appropriate for staging, but is applied to tumors with a depth of invasion of 1 mm or less but where the pathologic stage cannot be determined from the specimen submitted for various reasons. Problems such as the tumor extending to the margins of excision, the diameter of the tumor is unknown or cannot be measured, and other reasons (Darragh et al. 2013; Wilkinson 1991). Reporting the diameter, or horizontal growth, of the tumor, the depth of invasion, the thickness of the tumor, the presence or absence of vascular space involvement, and the status of surgical margins clearly defines the tumor and its extent. These findings will influence treatment options and are included in the CAP Vulvar Cancer protocol for synoptic reporting (Gibb et al. 2016).

Gross Findings Stage I invasive vulvar squamous cell carcinoma may present as an ulcer; a red, brown, or black

macule or papule; or a white hyperkeratotic plaque. The invasive carcinoma may be associated with an HSIL, or dVIN lesion, and clinically present as a HSIL, sometimes with associated lichen sclerosus, lichen planus, lichen simplex chronicus, condyloma acuminatum, or other vulvar conditions. The presence of invasion associated with HSIL, or dVIN, is suspected on clinical examination by the finding of an associated ulcer, irregularly contoured elevated mass, abnormal vascularity, or marked hyperkeratosis. No specific clinical findings definitively separate pure HSIL or dVIN from HSIL or dVIN with squamous carcinoma (Preti et al. 2017).

Microscopic Findings The identification of an invasive squamous cell carcinoma arising in a HSIL, or in dVIN, rests on the recognition of (1) isolated neoplastic squamous cells with increased eosinophilic cytoplasm and atypical nuclei with prominent nucleoli within the adjacent dermis, (2) loss of the orderly palisaded orientation of the basal keratinocytes of the dermal papillae, (3) irregular nests of neoplastic squamous cells with disorderly orientation within the dermis, (4) dyskeratosis and keratin pearl formation in the basal and parabasal layers, and (5) focal dermal reaction with fibrosis (dermal desmoplasia) or edema locally evident in the area of invasion (Kurman et al. 2010). Immunohistochemical study of laminin may also be of some value in that continuous reactive laminin is present about nests of HSIL, but reactivity is discontinuous about invasive squamous epithelium (Kurman et al. 2010; Rush et al. 2005). Proper measurement of the depth of invasion is extremely important to establish prognosis. In a study of 78 cases of vulvar squamous cell carcinoma with invasion of 5 mm or less, of 40 patients

2

Precursor Lesions and Malignant Tumors of the Vulva

83

with inguinal lymph node dissections, 5 had concurrent lymph node metastases. Concurrent lymph node metastases were compared to tumor depth of invasion, tumor thickness, tumor horizontal spread, estimated tumor volume, tumor histologic squamous tumor type, tumor grade, primary pattern of invasion, multifocality of tumor of the vulva, presence of perineural invasion, presence of angio-lymphatic invasion, histologic type of HSIL, as well as the presence of lichen sclerosus. In this detailed study, only tumor depth of invasion was statistically significant in correlating with the presence of lymph node metastases ( p = 0.027, n = 78, analysis of variance (ANOVA)). All additional correlations between these findings and lymph node metastasis were not statistically significant ( p > 0.05) (Yoder et al. 2008). In that the staging of a stage Ia invasive vulvar carcinoma is based upon the “depth of invasion of the tumor” (Fig. 12), this measurement requires a calibrated ocular or comparable measuring device, as does the measurement of the thickness of the tumor.

The measurement of the “depth of invasion” for staging stage I vulvar carcinoma is defined as the measurement from the epithelial dermal (stromal) junction of the most superficial adjacent dermal papillae to the deepest point of invasion (Fig. 12). Measurement from the basement membrane of the deepest adjacent tumor-free rete ridge to the deepest point of invasion has been studied in a series of 148 cases and, using this definition of depth of invasion, did result in 13 (9%) of the cases being restaged 1A group as compared to the conventional depth of invasion (van den Einden et al. 2015). This measurement, however, could underestimate the depth of tumor as it relates to the underlying valved lymphatics of the dermis or mucosa and access of the tumor to regional lymph nodes. This measurement is not the standard or recommended method of either measuring depth of invasion or staging. The measurement from the surface of the tumor, or from the base of the granular layer if a keratin layer is present to the deepest point of invasion, is defined as the “thickness of the tumor.” Both measurements are valuable, because

Fig. 12 (a) Squamous cell carcinoma; measurement for the depth of invasion. The depth of invasion is measured from the epithelial-dermal junction of the adjacent-most superficial dermal papillae to the deepest point of invasion. This measurement is applicable whether or not the surface epithelium is ulcerated or keratinized. This is the AJCC recommended method of measuring vulvar squamous cell carcinomas in determining if a tumor is stage T1a or T1b. (From AJCC Cancer Staging Manual, 8th ed. Springer; 2017, with permission. Figure # E.J. Wilkinson, 2007). Squamous cell carcinoma; measurement for the thickness of the tumor when the epithelial surface is intact. If the tumor is keratinized, the thickness of tumor is measured from the granular layer to the deepest point of invasion. For squamous cell carcinomas, the convention is to measure from the bottom of the granular layer. For melanoma the

convention is to measure from the top of the granular layer. If the epithelium is not keratinized, the thickness of the tumor is measured from the surface of epithelium overlying the tumor to the deepest point of invasion. (Figure # E.J. Wilkinson, 2007). (b) Measurements for tumor thickness when the tumor is ulcerated. For melanomas the tumor thickness is measured from the surface of the ulcerated tumor to the deepest point of invasion. For squamous cell carcinomas the depth of invasion is a more accurate measurement of the true depth of the tumor, as measured from the epithelial-dermal junction of the adjacent dermal papillae to the deepest point of invasion. (Figure # E.J. Wilkinson, 2007) (From Wilkinson EJ and Stone IK, Atlas of Vulvar Disease. Wolters Kluwer/. Lippincott-Williams Wilkins, 2012

84

in cases in which the tumor is ulcerated the thickness may be 1 mm or less and the depth of invasion may be well beyond 1 mm, and a thickness measurement alone would underestimate the depth of invasion (Fig. 12). Methods of measurement require a description along with the measurement within the pathology report (Darragh et al. 2013; Gibb et al. 2016). There can be some difficulties in measuring the depth of invasion in some cases, especially if there is tangential sectioning of the specimen or the surface epithelium is distorted, disrupted, or folded. When there is a question as to whether invasion is present and additional sectioning studies and review do not resolve the question, it is recommended that invasion not be diagnosed. In some cases the depth of invasion can be more readily calculated by measuring the thickness of the tumor and subtracting the epithelial measurement from the surface to the epithelial dermal (stromal) junction of the immediately adjacent dermal papillae. The measurement should be made from “. . .the epithelial-stromal junction of the adjacent most superficial dermal papilla. . .” and in some cases the immediately adjacent dermal papillae is not applicable for various reasons, including that it is not the most superficial dermal papillae, which may be the second papillae over on either side of the tumor. The issue is also not of such importance when the depth of invasion is obviously over 1 mm. In some cases levels of the block containing the section of tumor with the apparent deepest point of invasion may be worth obtaining to identify the best oriented section with the deepest point of invasion. In rare cases the involved epithelium may not have dermal papillae, and in such cases a measurement from the epithelial-dermal (stromal) junction of the adjacent epithelium uninvolved by tumor should be representative. In general, with a welloriented specimen with a tumor less than 3 mm in depth, and totally excised with intact noninvolved epithelium adjacent to the tumor, the measurements of depth of invasion and tumor thickness are made reliably (Figs. 12 and 13). In larger tumors, the diameter and dimensions of the tumor may be too great to include an adjacent dermal papilla; however, this often can be overcome by

E. J. Wilkinson and D. S. Rush

Fig. 13 Squamous cell carcinoma. The depth of invasion is 2.7 mm from the most superficial adjacent dermal papillae to the deepest point of invasion

appropriate sectioning. With a partially excised tumor, or when the tumor is superficially biopsied, these tumor measurements cannot be made reliably. When marked acanthosis is present, the thickness of the epithelium may give an overestimate of the depth of invasion. If the tumor is ulcerated, the thickness measurement may underestimate the true tumor size. The CAP has recommended that the information needed for pathologic staging be included in the surgical pathology report and the pathologic staging be reported when feasible. Staging is summarized in Tables 4, 5, and 6 (Gibb et al. 2016).

Clinical Behavior and Treatment Stage 1A carcinoma is usually curable by surgery alone, but patients who experience recurrence or develop a second primary may progress to more deeply invasive disease, with the associated risk of regional nodal metastases and death from disease. In a study of 28 women with stage 1A vulvar carcinoma, none had recurrence of squamous cell carcinoma in 240 months of follow-up (Yoder et al. 2008). In a series of 26 women with HSIL and associated superficially invasive squamous cell carcinoma, 10 (38%) of these patients experienced recurrence of HSIL or superficially invasive squamous cell carcinoma, all within 36 months of treatment (Herod et al. 1996). Three of these ten women subsequently presented with frankly invasive vulvar squamous cell carcinoma. Of the patients without recurrence of HSIL

Precursor Lesions and Malignant Tumors of the Vulva

or superficially invasive carcinoma, none had regional lymph node or distant metastases and none died of tumor. In another study of 40 patients with T1a (stage IA) vulvar carcinomas, none had regional lymph node metastasis; however, two patients had vulvar recurrence, and one of them had a groin node metastasis associated with the recurrence (Magrina et al. 2000). Surgical margin assessment correlates with recurrence in women treated with vulvar squamous cell carcinoma. The presence of HSIL at a surgical margin has borderline statistical significance related to recurrence of HSIL. Of 19 cases with HSIL at a surgical margin, 6 (58%) had recurrence of HSIL. In contrast only 4 of 59 (91%) women with margins free of HSIL or squamous cell carcinoma had recurrence of HSIL. In this study, two cases had invasive squamous cell carcinoma at a surgical margin, and both had recurrent, or persistent, squamous cell carcinoma (Yoder et al. 2008). This study identified a strong association between the type of surgery performed and surgical margin involvement by either HSIL or invasive squamous cell carcinoma. Of 19 cases with HSIL at the margin of excision, 15 (79%) were found in cases treated with wide local excision. Of the 78 cases in this study, 44 cases were treated by wide local excision, and of these, 15 had HSIL at a surgical margin, and 2 had invasive squamous cell carcinoma at the margin (Yoder et al. 2008). For now, it is safe to say there is a diminishing small risk of inguinal lymph node metastasis with tumors that are stage IA (T1a1) with a depth of invasion of 1 mm or less. For patients with stage IA carcinoma of the vulva, the recommended therapy is wide local excision without vulvectomy. Total excision of a lesion suspected of being a stage Ia invasive carcinoma is necessary to assure that an associated deeper squamous carcinoma is not immediately adjacent to the apparently superficially invasive focus. The surgical specimen typically encompasses the apparent HSIL and any associated hyperkeratotic or ulcerative lesions. The specimens are usually not more than 2–3 cm in greatest dimension, usually with 1 cm or less clinically negative margins. Sampling of the ipsilateral groin nodes, or bilateral groin nodes if the tumor is midline and/or

85

(15.2%)(769*)

15

Percent with inginual lymph node metastasis

2

14 13 (12.1%)(190*)

12 11 10 9 8 7 6 5 4 3 2 1 0

(0.0%)(69*) 0

1

2

3

4

5

Depth of invasion (mm)

Fig. 14 Percent of women who underwent lymphadenectomy with inguinal lymph node metastasis plotted against the depth of invasion of their tumor. The frequency of lymph node metastasis rises rapidly with depth of invasion beyond 1 mm. (Reprinted by permission of E.J. Wilkinson)

involves the clitoris with either sentinel lymph node sampling or lymph node biopsy, is an employed method of evaluation (Maroney et al. 2013). For stage IA (T1a) vulvar carcinomas, the probability of node metastasis is extremely small, and node sampling or resection is not contributory in most cases (Kurman et al. 2010; Yoder et al. 2008). Based on analysis of published series, the current treatment of these patients is partial deep vulvectomy (deep local excision) without lymphadenectomy (Magrina et al. 2000; Yoder et al. 2008) (Fig. 14).

Invasive Squamous Cell Carcinoma Clinical Features Women presenting with vulvar carcinoma may have a wide variety of presenting complaints relevant to the vulvar tumor, especially if the tumor is more advanced. Pruritus, burning, pain, bleeding, discharge, dyspareunia, dysuria, unpleasant odor, and palpation or observation of a mass are

86

reported. A past medical history of vulvar HSIL, dVIN, condyloma acuminatum, lichen sclerosus, or other chronic inflammatory diseases of the vulva may be present. Mental confusion and disorientation related to hypercalcemia have been reported associated with vulvar squamous cell carcinoma. After the ovary, the vulva is the second most common gynecologic tumor site associated with hypercalcemia. Vulvar squamous carcinomas with associated hypercalcemia usually are large, well differentiated, and without bony metastasis. Surgical excision of the tumor results in the serum calcium levels returning to normal and, if mental symptoms related to the hypercalcemia were present, also regresses. The hypercalcemia results from secretion by the tumor of parathyroid-related peptide (PTH-rP) or PTH-like substance. Invasive squamous carcinoma may present as a lesion associated with HSIL, as a focal ulcer or hyperkeratotic area in a field of vulvar lichen sclerosus, an exophytic papillomatous mass, or as an endophytic ulcer. The tumor usually is located on the labium minus or majus; however, the clitoris is primarily involved in less than one-fifth of cases. Typically the tumor is solitary (Figs. 15 and 16); less than 10% are multifocal.

Microscopic Findings Vulvar squamous cell carcinoma is typically contiguous with adjacent vulvar skin or mucosa, and that epithelium may be involved with HSIL, lichen sclerosus, chronic inflammation, or other changes. The invasive tumor typically lacks the usually orderly palisaded orientation of the basal epithelial cells with the underlying dermis. The adjacent dermis often has edema, desmoplastic change, and/or an associated inflammatory cell infiltrate that is typically not in direct contact with the invasive epithelial cells, but within the reactive dermis immediately below and adjacent to the tumor. Some prominent vascularity may be present within the dermis immediately below the tumor. The tumor may invade with a broad pushing pattern, or have variable degrees of “fingerlike” growth within the dermis, including single tumor cells within the dermis. The tumor growth pattern influences the risk of regional metastasis in

E. J. Wilkinson and D. S. Rush

Fig. 15 Squamous cell carcinoma. The tumor involves the medial aspect of the left anterior labium majus and clitoris

Fig. 16 Invasive squamous carcinoma presenting as a nodular mass in the right posterior labium majus

tumors exceeding 1 mm in depth of invasion (Yoder et al. 2008). A grading system for infiltrative patterns of vulvar carcinoma was proposed (Yoder et al. 2006). With this grading system, tumor features are classified in four histologic patterns of invasion. Pushing (P) lesions have broad smooth fronts of invasion along the tumor

2

Precursor Lesions and Malignant Tumors of the Vulva

dermal interface. Fingerlike (F) invasion has a pattern of invasion with bands of tumor attached to the tumor mass within the dermis. Small group (G) lesions are similar to fingerlike lesions, but with groups of neoplastic cells detached from the tumor surface and infiltrated into the dermis. Single cell (S) lesions have sharply angulated foci of invasion with single tumor cells along the tumor dermal interface. Lesions with a fingerlike pattern of invasion and multiple foci of invasion might correlate with disease recurrence. Capillary-like vascular space invasion by the tumor may be evident, and in such cases the tumor can usually be found attached to the lining of the capillary-like space. The tumor is composed of neoplastic squamous epithelial cells that may have variable eosinophilic cytoplasm with nuclei that exhibit variable degrees of anisonucleocytosis with nuclear chromatin abnormalities including clumped and radially dispersed chromatin. Nucleoli are usually present, and mitotic figures including abnormal mitotic figures may be evident, especially near the epithelial-dermal junction. A number of grading systems for vulvar squamous cell carcinomas have been proposed; however, in this work we will use the AJCC 2017 grading system. The AJCC recommends the following histopathologic grading for vulvar squamous carcinomas, designated “G” as follows: GX, grade cannot be assessed; G1, well differentiated; G2, moderately differentiated; G3, poorly differentiated; and G4 is no longer used (Gibb et al. 2016; American Joint Committee on Cancer 2017). Grade 1 tumors have no undifferentiated cells (Figs. 16, 17, and 18), grade 2 tumors contain less than 50% undifferentiated cells, and grade 3 tumors (Fig. 19) have greater than 50% to being entirely composed of undifferentiated cells. The risk of recurrence is reportedly higher with increasing grade (Homesley et al. 1991). In addition to tumor staging (as summarized in Tables 4 and 5), the AJCC staging system reports regional lymph node status (as N) and the presence or absence of metastases (as M) (Table 6). All lymph node tissue should be submitted for microscopic analysis. The microscopic evaluation of lymph nodes for detection of metastatic cell squamous carcinoma may be augmented by

87

Fig. 17 Invasive squamous cell carcinoma, well differentiated. Tongues of well-differentiated squamous epithelium with keratinization are evident

Fig. 18 Invasive keratinizing squamous cell carcinoma, well differentiated. The cells have abundant cytoplasm and large, round nuclei with prominent nucleoli

immunohistochemistry using a polyclonal cytokeratin antibody. Pathologic findings included in the CAP guidelines for evaluation of vulvar squamous cell carcinoma are the same for deep and superficially invasive squamous cell carcinomas (Greene et al. 2016). The CAP Surgical Pathology Cancer Case Summary/Checklist provides a synoptic approach to reporting vulvar cancer cases for either excisional biopsies or resections. In addition to tumor staging, the checklist includes macroscopic observations including specimen type, type of lymphadenectomy if performed, specific tumor site, and tumor size. Microscopic features to record include the histologic type of tumor,

88

E. J. Wilkinson and D. S. Rush

histologic grade, pathologic stage (pTMN/FIGO), tumor depth of invasion, tumor infiltrative pattern, surgical margin status, and the presence or absence of capillary-like vascular invasion by tumor. Additional pathologic findings can be added including pathologic findings in the adjacent vulvar skin or mucosa and any additional comments (Greene et al. 2016; Kurman et al. 2010).

lichen sclerosus, it has been observed that Ki-67 expression is increased in the squamous hyperplasia adjacent to the squamous cell carcinoma. These findings may imply premalignancy or a reactive process related to the carcinoma. In addition, p53 is expressed in the neoplastic component when dVIN is associated (Yang and Hart 2000). Nearly all vulvar HSIL and approximately one-third of vulvar squamous cell carcinomas are associated with oncogenic HPV, and the majority of these lesions will express p16INK4a (McAlpine et al. 2017; Riethdorf et al. 2004; Rufforny et al. 2005). In a meta-analysis of 17 studies including 2309 patients, overexpression of p16INK4a in vulvar squamous cell carcinomas correlated with lower tumor stage, negative node metastasis, younger patient age (under 55 years), and higher survival (Cao et al. 2016). Immunohistochemical expression of p16INK4a was identified in 80% of vulvar squamous cell carcinomas that were HPV positive by PCR analysis in a study of 57 patients (Lee et al. 2016). In this study, both better survival and lower infield relapse rates occurred in women with vulvar carcinomas that expressed p16 and were treated by radiotherapy as compared to the p16-negative cases.

Adjunctive Studies Cytogenetic studies on vulvar carcinomas have demonstrated that they are genetically complex and have karyotypic abnormalities (Crum et al. 2014a). A study of vulvar carcinomas employing flow cytometry and image analysis demonstrated a high frequency of aneuploidy with a predominance of tumors within the hypotetraploid range (Drew et al. 1996). Molecular genetic studies have demonstrated changes in the p16/pRB cyclin D1 pathway (Kurman et al. 2010). Both development and progression are apparent sequelae of altered gene expression. Immunohistochemical studies employing monoclonal antibodies to MIB-1 (Ki-67), a proliferation-associated marker, have demonstrated two distinct tumor labeling patterns, diffuse and localized, that appear to be associated with prognosis. The diffuse pattern is associated with poor prognosis (Hendricks et al. 1994). In vulvar squamous cell carcinoma associated with

Clinical Behavior and Treatment There are relatively limited data specifically evaluating tumors with a depth of invasion of 1.1–2 mm. In two relatively small studies specifically examining tumors with invasion of 1–2 mm, without vascular space involvement, none had node metastasis (Preti et al. 1993; Yoder et al. 2008). In a multivariate analysis performed by Yoder et al., 6 of 19 (31%) cases with 1.1 mm–2.0 mm depth of invasion had recurrence of vulvar squamous cell carcinoma (Yoder et al. 2008). In these cases, however, recurrence did not influence survival, and those women with tumors having a depth of invasion of 2.0 mm or less of invasion had 100% overall survival at 240 months (Yoder et al. 2008). Until more data are available on tumors that invade between 1 and 2 mm, the evidence is that women with such tumors have limited risk of regional lymph node metastasis and sentinel lymph node sampling rather than lymphadenectomy should be

Fig. 19 Invasive squamous cell carcinoma, moderately to poorly differentiated. Small nests and cords of invasive squamous cell carcinoma are present without keratinization

2

Precursor Lesions and Malignant Tumors of the Vulva

considered when node assessment is considered in such patients (Klapdor et al. 2017a; Klapdor et al. 2017b; Moore et al. 2003). Recurrence is a recognized risk and treatment with wide local excision/ partial vulvectomy in these patients, with at least 1 cm tumor-free margins is generally accepted (Maroney et al. 2013). Women with tumors having a depth of invasion of 2.1–3 mm or less have significant risk of regional lymph node metastasis, recurrence of tumor, and reduced survival. With this depth of invasion, there is inguinal lymph node metastasis in approximately 10% of cases. In one study, women with tumors having a depth of invasion between 2.1 and 3.0 mm, 5 of 16 (31%) cases had recurrent squamous cell carcinoma. These women had a 93% survival at 258 months, whereas women with tumors greater than 3.1 mm of invasion had 79% survival at 231 months (Yoder et al. 2008). Selected patients with tumors with a depth of invasion of 2.1–3 mm can be treated by wide local excision with ipsilateral regional node dissection. In such cases, and with more advanced tumors, gynecologic oncology consultation is usually advisable (Maroney et al. 2013). Vulvar squamous cell carcinomas with a depth of invasion exceeding 3.1 mm are associated with an increased risk of recurrence and death. In one study of tumors exceeding 3 mm invasion, 5 of 15 (33%) cases had recurrence of tumor with an overall survival of 79% at 231 months (Yoder et al. 2008). Women with vulvar squamous carcinomas with a 5 mm depth of invasion have been found to have inguinal lymph node metastasis in approximately 15% of cases (Fig. 14). In addition to the depth of invasion, microscopic findings that are associated with lymph node metastasis include a fibromyxoid stromal, or dermal, response to the infiltrative tumor and perineural invasion. Tumors with a fibromyxoid stromal response were found to be more deeply invasive and more likely to have lymph node metastasis, extranodal extension by tumor, and perineural invasion (Jeffus et al. 2015). Perineural invasion by tumor is identified as an independent predictor of recurrence (Holthoff et al. 2015). The type of surgery for vulvar squamous cell carcinomas with a depth of invasion less than

89

5 mm does not appear to influence recurrence of squamous cell carcinoma. In a study of 78 patients with tumors of 5 mm in depth of invasion or less, 44 (56%) underwent wide local excision (partial deep vulvectomy) of which 3 had adjuvant radiation, 7 (9%) had total (simple) vulvectomy, and 27 (35%) had total vulvectomy with inguinal femoral lymph node dissection (radical vulvectomy), 1 of whom had adjuvant radiation therapy. No association was identified between recurrent squamous cell carcinoma and the type of surgical procedure performed (Yoder et al. 2008).

Histologic Subtypes of Vulvar Squamous Cell Carcinoma Squamous cell carcinomas of the vulva are of several morphologically distinct subtypes (Table 3). Histologically, invasive squamous cell carcinomas that are not otherwise specified (NOS) usually are well-differentiated tumors, but moderately and poorly differentiated varieties are found in 5–10% of the cases (see Figs. 13, 16, 17, 18, and 19).

Basaloid Carcinoma An increased prevalence of HPV, mainly type 16, is associated with certain types of invasive squamous carcinomas of the vulva. Among these are basaloid carcinomas, which occur in younger women (mean age, 54 years), compared with typical keratinizing squamous cell carcinomas (mean age, 77 years) (Fig. 22b). HPV 16 is detected in approximately 70% of basaloid squamous carcinomas (Fig. 22b) (Kurman et al. 1993). HPV-associated vulvar squamous cell carcinomas are not reliably distinguished from nonHPV-related tumors by histopathology. Molecular methods are the most sensitive; however, detection of p16INK4a expression by immunohistochemical methods is more commonly clinically employed (McAlpine et al. 2017). Basaloid carcinomas are frequently associated with adjacent HSIL, usually of the basaloid type. In contrast to typical keratinizing squamous cell carcinomas, basaloid carcinomas are associated

90

Fig. 20 Basaloid desmoplastic stroma

E. J. Wilkinson and D. S. Rush

carcinoma

with

a

prominent

with synchronous or metachronous squamous neoplasms of the cervix and vagina (Kurman et al. 1993). On gross examination, basaloid carcinomas are similar to typical keratinizing squamous cell carcinomas. Microscopically, they are characterized by variable-sized nests of immature squamous cells with little, if any, squamous maturation. Some tumors are composed of small, irregularly shaped nests and cords of cells surrounded by a densely hyalinized stroma. The basal-type cells within the nests and cords resemble those in the classic type of carcinoma in situ of the cervix (Figs. 20, 21, and 22). Characteristically, the cells are ovoid and relatively uniform in size, with scant cytoplasm and a high nuclear cytoplasmic ratio, and therefore they appear undifferentiated. Nuclei contain evenly distributed coarsely granular chromatin, creating a stippled appearance. A moderate degree of mitotic activity usually is evident. Occasionally, the cells in the center of a nest show evidence of maturation and contain more abundant cytoplasm. Small foci of keratinization may be evident in the center of the nests, and keratin pearls occasionally are present. Desmosomes usually are not evident. The differential diagnosis of basaloid carcinoma includes metastatic small cell carcinoma, Merkel cell tumor, and basal cell carcinoma. These tumors have a more diffusely infiltrative pattern characterized by poorly defined nests, trabeculae, and individual cells invading the stroma rather than the broad anastomosing bands and well-defined nests typical of basaloid carcinoma. Small cell tumors

Fig. 21 Basaloid VIN with basaloid carcinoma. The basaloid VIN within the overlying epithelium has similar cellular features to the basaloid invasive tumor within the superficial and deep dermis. The tumor is composed of immature-appearing keratinocytes without significant maturation or keratinization

Fig. 22 Basaloid carcinoma. The tumor is composed of relatively small cells with hyperchromatic, slightly pleomorphic nuclei. There is cellular disarray throughout the neoplasm; however, some keratinization is evident

usually are immunoreactive for neuroendocrine markers; Merkel cell tumors have a characteristic perinuclear cytoplasmic “dot” demonstrated by immunohistochemical study with cytokeratins. Basaloid squamous carcinoma must be distinguished from basal cell carcinoma, but at times this may be difficult. In contrast to basaloid carcinoma, basal cell carcinomas tend to be more circumscribed and have a lobular appearance. The characteristic palisading of the outermost layer of cells in the nests of basal cell carcinoma is lacking in basaloid carcinoma.

2

Precursor Lesions and Malignant Tumors of the Vulva

91

The prognosis of basaloid carcinoma does not appear to be significantly different from squamous carcinoma of the usual type; no evidence of decreased survival was observed in relatively large study (Kurman et al. 1993).

Warty Carcinoma (Condylomatous Carcinoma) Warty carcinoma is found predominantly in younger women (mean age, 55 years) and presents clinically as a verrucoid or papillary tumor that may resemble condyloma acuminatum. Warty carcinoma, like basaloid carcinoma, is found predominantly in younger women (mean age, 55 years) and is associated with HPV. It may be associated with adjacent warty and/or basaloid HSIL and with other genital tract squamous neoplasms (Kurman et al. 1993). It presents clinically as a verrucoid or papillary tumor that may resemble condyloma acuminatum. Both warty squamous cell carcinoma and verrucous carcinoma may arise in association with vulvar condyloma acuminatum. On microscopic examination, warty carcinoma has multiple papillary projections with a keratinized epithelial surface and fibrovascular cores (Figs. 23 and 24). Cytologic atypia is present, especially within the basal and parabasal cells, where there is nuclear pleomorphism and nuclear hyperchromasia. Multinucleation may be present. Mitotic figures usually can be found and sometimes may be atypical. Cytoplasmic perinuclear clearing similar to koilocytosis in HSIL is present in a substantial number of cells; it is the most characteristic feature. At the junction between the exophytic portion of the tumor and the underlying stoma, irregularly shaped nests of epithelium are present that may be associated with keratin pearls and dyskeratotic cells. In this area the tumor resembles a keratinized squamous cell carcinoma. In some cases these areas are small and focal. Warty carcinoma is frequently associated with HPV type 16 (Kurman et al. 1993; Kurman et al. 2010). The clinical course of warty carcinoma appears generally good; however, lymph node metastasis may occur. The prognosis appears intermediate between that of verrucous carcinoma and

Fig. 23 (a) Warty (condylomatous) carcinoma. The tumor has well-differentiated neoplastic keratinocytes with keratinization. At the deep margin, the tumor is composed of irregularly shaped, varying-sized nests that infiltrate the stroma in a haphazard fashion. (b) Warty (condylomatous) carcinoma. The cords of neoplastic cells are supported by a fibrovascular stroma. Keratinization is present

squamous cell carcinoma of the usual type. Approximately 80% of warty and basaloid carcinomas have adjacent warty or basaloid HSIL. About one-quarter of the warty and basaloid carcinomas are associated with other genital tract squamous neoplasias (Kurman et al. 1993).

92

E. J. Wilkinson and D. S. Rush

Fig. 24 Warty (condylomatous) carcinoma. Cells with pleomorphic nuclei and vacuolated cytoplasm resembling koilocytes are present within the neoplastic epithelium. (Courtesy of R.J. Kurman, M.D., Baltimore, MD)

Verrucous Carcinoma Verrucous carcinoma is a highly differentiated squamous carcinoma that has an exophytic gross appearance and may be associated with pruritus and/or pain. The tumor invades with a pushing border in the form of bulbous pegs of neoplastic cells (Brisigotti et al. 1989; Liu et al. 2016). The term giant condyloma of Buschke-Lowenstein is considered to be a synonym for verrucous carcinoma, but it is confusing and therefore is not recommended (Kurman et al. 2010). Verrucous carcinoma is a papillary exophytic growth that may have the appearance of an exophytic, broadbased condyloma acuminatum and distort or completely obscure the vulva (Fig. 25). Secondary infection may be associated with a malodorous discharge. Regional lymph nodes usually are not enlarged. VAAD characterized as a noninvasive squamous lesion with variable verrucoid growth with acanthosis, parakeratosis, loss of the granular layer, and pale cytoplasm of the superficial keratinocytes is proposed as a probable precursor of vulvar verrucous carcinoma. These changes were reported adjacent to vulvar verrucous carcinoma in seven cases in a study of nine verrucous carcinomas from seven patients. Neither this lesion nor the verrucous carcinomas in this study were associated with HPV (Nascimento et al. 2004). Other studies have indicated that verrucous

Fig. 25 Verrucous carcinoma in cross section. This tumor is 5 cm in diameter and has a broad, well-defined margin of infiltration involving the underlying fibrofatty tissue

carcinoma may be associated with HPV, typically type 6 or variants of type 6. The microscopic features of verrucous carcinoma include prominent acanthosis with a pushing tumor-dermal interface and bland cytologic features (Fig. 26) (see ▶ Chap. 3, “Diseases of the Vagina”). There is minimal nuclear pleomorphism, with the greatest degree of nuclear atypia nearest the dermal interface. The nuclei may have coarse chromatin and variable-sized nucleoli, distinguishing them from normal adjacent keratinocytes. Mitotic figures are rare and when present are normal. The abundant cytoplasm of the tumor cells is eosinophilic, without dyskeratosis. Koilocytosis is not a feature of this tumor. Parakeratosis or hyperkeratosis usually is present and may be prominent. The tumor invades with a pushing border in the form of bulbous pegs of neoplastic cells (Brisigotti et al. 1989; Liu et al. 2016). There is an absence of fibrovascular cores separating the bulbous epithelial downgrowths. An

2

Precursor Lesions and Malignant Tumors of the Vulva

93

Fig. 27 Giant cell carcinoma of the vulva. Multinucleated tumor giant cells are evident. The cells contain nuclei with prominent nucleoli and abundant eosinophilic cytoplasm

Fig. 26 Verrucous carcinoma. There is a well-defined tumor-stromal interface with a pushing growth pattern. The keratinocytes are well differentiated, and some mitotic figures are present

inflammatory infiltrate within the dermis usually is present. These tumors typically are diploid. Invasive squamous carcinoma may be associated with verrucous carcinoma in 5–17% of cases reported (Liu et al. 2016). The differential diagnosis includes the typical variety of squamous cell carcinoma, warty carcinoma, and condyloma acuminatum. Squamous cell carcinomas at times may have some of the architectural features of verrucous carcinoma, but if they lack a high degree of differentiation, or have a nonpushing pattern of invasion, they should not be designated verrucous carcinoma. Squamous cell carcinoma of the usual type (keratinizing squamous carcinoma) has greater nuclear pleomorphism and a more irregular pattern of infiltration of the stroma compared with the bulbous nests of verrucous carcinoma. Warty carcinoma, despite its verruciform appearance, has fibrovascular cores within the papillary fronds, unlike verrucous carcinoma. In addition, these

tumors display greater nuclear atypia, “koilocytosis,” and, at their deep margin, invade like typical squamous cell carcinomas. Condyloma acuminatum is characterized by a complex branching papillary architecture with vascular papillae, lacks bulbous downgrowths, and typically shows koilocytosis, although in vulvar condylomas koilocytosis may be quite subtle. Verrucous carcinomas may recur locally after excision. Lymph node metastasis is extremely rare, and its presence should prompt reevaluation of the lesion for areas of the usual type of squamous cell carcinoma. Wide local excision and total vulvectomy without lymph node dissection are the most common methods of therapy. If the tumor is excised completely, the prognosis is excellent, although the recurrence is reported near 20% (Liu et al. 2016). The role of radiotherapy in vulvar verrucous carcinomas is not well studied, but it may be applicable in very advanced cases.

Giant Cell Squamous Carcinoma Squamous cell carcinoma with tumor giant cells is a variant of squamous cell carcinoma characterized by multinucleated tumor giant cells, large nuclei with prominent nucleoli, and prominent eosinophilic cytoplasm (Fig. 27). This tumor variant is relatively rare and is associated with a poor prognosis. The most important differential

94

diagnosis is malignant amelanotic melanoma, which commonly forms multinucleated tumor giant cells (Wilkinson et al. 1988). Melanomas typically have intranuclear inclusions and prominent nucleoli. Unlike giant cell carcinoma, melanomas are typically immunoreactive for S-100, melanoma antigen (HMB-45), and Melan-A and negative for cytokeratin.

Spindle Cell Squamous Cell Carcinoma Spindle cell carcinoma of the vulva/sarcomatoid squamous cell carcinoma is a rare variant of squamous cell carcinoma that may mimic a sarcoma or be associated with sarcoma-like stroma (Fig. 28). The tumors may be biphasic and may have giant cells (Bigby et al. 2014). It can be associated with vulvar squamous cell carcinoma and has been identified in one case as an invasive component involving vulvar anogenital mammary-like glands and associated with ductal carcinoma in situ in the involved gland (Tran et al. 2015). In a study of four cases, all four had associated lichen sclerosus, three of which had dVIN with the tumor and one had associated malignant heterologous elements in the tumor. None of these cases were associated with HPV (Bigby et al. 2014). These tumors express keratin, reflecting their epithelial origin, but may also express myogenin (myf4) by immunohistochemistry, and diagnosis cannot rest on

Fig. 28 Spindle cell (sarcomatoid) squamous cell carcinoma. The tumor has fascicles of spindle-shaped neoplastic squamous cells. There is a moderate nuclear pleomorphism and mitotic figures are evident

E. J. Wilkinson and D. S. Rush

this finding alone (McCluggage et al. 2013). Spindle cell carcinomas must be distinguished from mesenchymal spindle cell tumors, including leiomyosarcoma, malignant fibrous histiocytoma, fibrosarcoma, and myoepithelial carcinoma (Meenakshi and McCluggage 2009). Other tumors with spindle cells to consider include spindle cell malignant melanoma and transitional cell carcinoma with spindle cell features. The neoplastic cells of spindle cell carcinoma, unlike all the mesenchymal tumors and melanoma, are immunoreactive for keratin.

Acantholytic Squamous Cell Carcinoma (Adenoid Squamous Carcinoma; Pseudoangiomatous Carcinoma) An acantholytic squamous cell carcinoma forms rounded spaces, or pseudoacini, lined with a single layer of squamous cells. Dyskeratotic and acantholytic cells are sometimes present in the central lumen (Fig. 29). These changes are focal

Fig. 29 Acantholytic squamous carcinoma. Nests of poorly differentiated squamous cell carcimoma are arranged in a crude acinar manner. Some of the acinar like areas are vacuolated

2

Precursor Lesions and Malignant Tumors of the Vulva

in most cases and may occur within otherwise well-differentiated squamous tumors. Acantholytic squamous cell carcinomas have a prognosis similar to squamous carcinoma of the usual type; however, some reports describe a more aggressive behavior for this tumor type and specifically for those tumors that mimic angiosarcoma and are subclassified as a pseudoangiosarcomatous carcinoma variant of acantholytic squamous cell carcinoma (Horn et al. 2008). Adenosquamous carcinoma is the major entity in the differential diagnosis, but it is a distinct tumor from acantholytic squamous carcinoma and has both glandular and squamous components.

95

mixed with, and surrounded by, a dense lymphocytic infiltrate (Fig. 30) (see ▶ Chap. 6, “Carcinoma and Other Tumors of the Cervix”). The epithelial cells are immunoreactive for high molecular weight cytokeratins, which distinguish them from inflammatory processes and malignant large cell lymphomas. Lymphomas, including Ki-1 lymphomas, which are immunoreactive for lymphocytic markers, contain an immunophenotypic monoclonal population of neoplastic lymphocytic cells. The therapy is wide local excision with or without local radiation therapy (Carr et al. 1992).

Plasmacytoid Squamous Carcinoma Papillary Squamous Cell Carcinoma This tumor is very rare and has similar morphologic features to papillary squamous cell carcinoma of the cervix. The tumor is exophytic with an expansile and deep pushing infiltrative pattern within the dermis. Although there is very limited experience with this tumor, with negative lymph nodes it can be treated by deep wide excision (Lomo and Crum 2004).

Lymphoepithelioma-Like Carcinoma These tumors may occur rarely on the vulva in older individuals. They are composed of nests or syncytial groups of epithelioid-appearing cells

Fig. 30 Lymphoepithelioma-like carcinoma. The tumor within the dermis is composed of nests and syncytial groups of epithelioid neoplastic cells mixed with a dense lymphocytic infiltrate

Plasmacytoid squamous carcinoma is a rare tumor of the vulva. One case is reported in a 92-year-old woman who had been treated previously for multiple vulvar squamous and verrucous carcinomas. The patient’s prior tumors were all treated by local excision. The plasmacytoid tumor presented as a 3 cm polypoid vulvar ulcerated mass superior but adjacent to the urethra (Tran and Carlson 2008). Grossly the tumor was submucosal, somewhat yellow in color, and abutted the overlying epithelium. No evidence of HSIL or other related HPV changes were identified adjacent to the tumor. On microscopic examination approximately 60% of the tumor cells had plasmacytoid features with cells having epithelioid differentiation. These cells lacked the nuclear characteristic “clockface” appearance of plasma cells, but did have prominent amphophilic cytoplasm with an eccentrically placed nucleus. In addition, these cells were more than twice the size of normal plasma cells. The tumor cells were immunoreactive for cytokeratin 5, AE3, high molecular weight cytokeratin 903, p63, CD138, and VS38, but negative for kappa or lambda light chains. Some cells expressed AE1. Other immunohistochemical studies performed were negative. The tumor metastasized and was fatal within 1 year after diagnosis. The differential diagnosis of vulvar plasmacytoid squamous cell carcinoma includes metastatic or primary melanoma, myoepithelial tumors, neuroendocrine tumors, plasmacytoma, metastatic plasmacytoid urothelial carcinoma,

96

and lobular breast carcinoma. These tumors are distinguished by immunohistochemical studies, and clinical history is often contributory in metastatic cases.

Clinical Behavior and Treatment of Vulvar Squamous Cell Carcinoma Factors that may be of significance in prognosis and in the probability of lymph node metastasis include the diameter of the tumor, the presence of vascular space invasion, and tumor ulceration. Confluent growth, defined as anastomosing cords or tumor, or tumor in the dermis exceeding 1 mm3, does not correlate with the occurrence of node metastasis but is not found in tumors having 1 mm or less of invasion (Wilkinson 1991). When controlled for age, survival with vulvar carcinoma decreased with advancing age, higher stage and grade, increasing tumor thickness, prominent fibromyxoid dermal response, infiltrative pattern of growth, and basaloid tumor type (Pinto et al. 1999). Inguinofemoral lymph node status and the diameter of the tumor are independent prognostic factors. The relative 5-year relative survival for vulvar carcinoma, related to stage, as reported by the ACS is as follows: localized, including stage I and II, 86%; regional, including stage III and stage IVA, 54%; and distant, stage IVB, 16% (American Cancer Society 2017). Five-year survival for women with stage IA tumors treated with appropriate surgical excision approaches 100%; however, women with tumors with a depth of invasion greater than 1 mm require more extensive surgery, including groin node dissection (Maroney et al. 2013; Wilkinson 1991). Terminology for surgical procedures and characterization of depth of invasion of vulvar tumors is developed by the ISSVD and is shown in Table 7 (Iversen et al. 1990). Vulvar carcinoma typically spreads by direct extension and lymphatic metastasis and tends to recur locally, with distant metastasis being less common. Direct extension may include involvement of the bone, and metastasis may include bone metastasis. The reliability of clinical evaluation in the determination of whether tumor is present in inguinal nodes has a false-positive rate

E. J. Wilkinson and D. S. Rush Table 7 Surgical procedures and characterization of extent and depth of excision of vulvar tumors Vulvectomy: Partial vulvectomy: removal of a part of the vulvar/ perineal integument independent of depth Total vulvectomy: removal of the whole vulva and appropriate integument of the perineum independent of depth Depth of excision: Superficial excision: removal of the most superficial layer with a variable amount of dermis and subcutaneous tissue Deep excision: removal of the vulva to the superficial aponeurosis of the urogenital diaphragm and/or pubic periosteum ISSVD (Iversen et al. 1990)

less than 10% but a false-negative rate of approximately 20%. The pathologic evaluation of lymph lymph nodes for metastasis also may be falsely negative and is dependent on the lymph node sampling method and observations of the pathologist. In assessing groin nodes, fine needle aspiration may be the first step if clinically suspicious lymph nodes are present because the technique is rapid, safe, and cost-effective and will detect gross metastasis. Current therapy attempts to define high- and low-risk groups and requires individualization of therapy. Sentinel lymph node identification methods employing technetium-99mlabeled noncolloid are often employed to attempt to select and sample sentinel lymph nodes for excision and evaluation and possibly avoid inguinofemoral lymph node resection if the sentinel lymph node or nodes are free of metastatic tumor (Klapdor et al. 2017b; Moore et al. 2003). Tumor recurrence within the groin after negative sentinel lymph node evaluation was reported in 5 of 30 patients (16.7%), and in both cases the primary tumor was central and exceeded 2 cm in size (Klapdor et al. 2017a). In a study of 377 patients with stage T1 vulvar carcinoma, isolated groin recurrence occurred within a median follow-up of 105 months in 2.5% of the sentinel lymph node-negative cases and 8.0% of sentinel lymph node-positive cases (Te Grootenhuis et al. 2016). Radiation techniques are now available

2

Precursor Lesions and Malignant Tumors of the Vulva

97

that allow skin-sparing treatment of tumor involving groin nodes and has been used successfully in both primary and adjunctive treatment.

Skin Adnexal-type Carcinomas (Basal Cell, Adenoid Basal Cell, Basosquamous [Metatypical Basal Cell], Sebaceous Cell) Basal Cell Carcinoma Although basal cell carcinomas of the extragenital skin are extremely common, they are extremely uncommon on the vulva and account for less than 10% of vulvar carcinomas. They are found primarily in elderly white women (mean age, 70–76 years, usually presenting as an ulcer, an area of pigmentation or depigmentation, or as a mass). Pruritus is the most common presenting symptom (Elwood et al. 2014; Mulayim et al. 2002; Pleunis et al. 2016). Most of these tumors are confined to the labia majora, and approximately one-half are of infiltrative type. The histologic pattern resembles that of basal cell carcinomas occurring elsewhere on the skin (Fig. 31a). The tumor is composed of small elongated cells with deeply basophilic nuclei and may have a large variety of architectural patterns, ranging from slight palisading of the basal layer of the epidermis to the formation of large club-shaped masses of pleomorphic basal cells. The connective tissue adjacent to the tumor frequently contains a chronic inflammatory cell infiltrate and occasionally shows a mucoid or myxomatous change. Basal cell carcinomas are not associated with HPV. Tumors commonly have patchy p16 expression with about one-half of the cells being reactive, which can assist in distinguishing basal cell carcinoma from basaloid squamous cell carcinoma, which is HPV-associated. In addition, BerEP4 is typically expressed in basal cell carcinomas, but may be negative in those tumors with a prominent squamous component, and is negative in usual HPV-related squamous cell carcinomas (Elwood et al. 2014). Primary treatment is wide local excision (Pleunis et al. 2016). Local recurrence may

Fig. 31 (a) Basal cell carcinoma. The tumor cells are small and uniform, lack maturation, and show characteristic palisading at the periphery of the involved rete ridges. (b) Basosquamous carcinoma (metatypical basal cell carcinoma). The rete ridges are branched and extend in a pushing manner into the dermis. The cells show increased cytoplasm in the areas of squamous differentiation. (Courtesy of R.J. Kurman, M.D., Baltimore, MD)

occur in up to one-fifth of the cases. Metastasis to regional lymph nodes occurs but is rare. The overall prognosis of patients with these tumors is excellent, with only a rare metastasis or death reported (Benedet et al. 1997; Pleunis et al. 2016).

98

Adenoid Basal Cell Carcinoma Adenoid basal cell carcinoma is a variant of basal cell carcinoma in which tubular and gland-like differentiation is seen within a tumor that otherwise is a characteristic basal cell carcinoma. Basosquamous Carcinoma (Metatypical Basal Cell Carcinoma) Basosquamous carcinoma represents a mixture of both squamous and basal cell neoplastic elements (Fig. 31b). BerEP4 may be negative in these tumors, unlike typical basal cell carcinoma (Elwood et al. 2014). There is very limited information on the metastatic potential of basosquamous carcinoma. Experience is limited; however they are known to be locally aggressive, can recur locally, and may metastasize. They are managed like squamous cell carcinoma. Sebaceous Cell Carcinoma Sebaceous cell carcinoma is a rare tumor of the vulva that may be associated with vulvar squamous intraepithelial lesion (Escalonilla et al. 1999). The tumor has features of basosquamous cell carcinoma with sebaceous differentiation, a lobulated appearance, and cells with a cytoplasmic “bubbly” appearance. The cells express CAM 5.2 and some express nuclear p53; however, they are not HPV associated. Experience with these tumors is limited, but they may have aggressive behavior (Pusiol et al. 2011) (Fig. 32).

Fig. 32 Sebaceous carcinoma. The tumor is composed of cords and nests of basaloid-appearing cells. These cells are associated with sebaceous cells in pagetoid nests in the parabasal areas and in larger clusters near the epithelial surface. (Courtesy of R.J. Kurman, M.D., Baltimore, MD)

E. J. Wilkinson and D. S. Rush

Glandular Tumors of the Vulva (Primary Adenocarcinomas of the Vulva) Glandular malignancies of the vulva may originate from a variety of sources. Paget disease is an intraepithelial lesion with glandular differentiation, which may give rise to invasive adenocarcinoma. Adenocarcinoma in situ has been reported arising within vulvar papillary hidradenoma (Shah et al. 2008), but most adenocarcinomas of the vulva arise as primary malignant tumors of the Bartholin gland. Less commonly, such tumors may arise from sweat glands or from other skin appendages including those about the clitoris. Tumors including sclerosing ductal carcinoma with microcystic adnexal carcinoma-like features and pilomatrix carcinoma can occur (DuPont et al. 2009; Gazic et al., 2011). Other adenocarcinomas may arise from the urethra, the Skene gland, the paraurethral ducts or cysts, and from Paget disease (Heller and Bean 2014; Heller 2015).

Fig. 33 Paget disease, primary type. An eczematoid, slightly raised, and white area is present on the medial anterior surface of the right labium majus

2

Precursor Lesions and Malignant Tumors of the Vulva

99

Fig. 35 Paget disease, primary type. Large numbers of Paget cells within the basal and parabasal areas with isolated Paget cells extending into the upper epithelium

Table 8 Classification of primary and secondary vulvar Paget disease

Fig. 34 Paget disease, primary type, total deep vulvectomy. The white and eczematoid epithelial changes reflect the Paget disease that involves the perineal body and left labia majus. (Photo courtesy R. Foss, PA, Gainesville, Fl)

Paget Disease Clinical Features Clinically, vulvar Paget disease can present as either an erythematous lesion, often involving the vestibule and adjacent areas, or an eczematous lesion that appears as a red to pink area with white islands of hyperkeratosis and usually involving hair-bearing skin (Fig. 34). The extent of involvement can be very focal or extensive, extending to the anus, medial aspects of the upper thigh, or other contiguous sites. Because of its clinical resemblance to a dermatosis, these patients may be treated with various topical medications for some time before the diagnosis is made by biopsy (Fig. 35). Pruritus is present in more than half the patients and was present for a median duration of 2 years before the diagnosis in a study of 100 patients (Fanning et al. 1999). Almost all patients are postmenopausal Caucasian women, with a median age

Primary Paget disease (primary cutaneous Paget disease) Paget disease as an intraepithelial neoplasm/in situ Paget disease Paget disease as an intraepithelial neoplasm with invasion/invasive primary Paget disease Paget disease as a manifestation of an underlying cutaneous neoplasm Secondary Paget disease (Paget disease of non-cutaneous origin) Paget disease as a manifestation of anal-rectal adenocarcinoma Paget disease related to other adenocarcinomas Paget disease as a manifestation of urothelial in situ or invasive carcinoma PUIN Wilkinson and Brown (2002)

of 70 years (Fanning et al. 1999). Paget disease may occur in younger women of reproductive age. Vulvar pruritus or pain may bring the patient to the attention of the physician. Vulvar Paget disease has been subclassified by Wilkinson and Brown into two distinct types, specifically cutaneous Paget disease and non-cutaneous Paget disease based upon the origin of the neoplastic cells as summarized in Table 8 (Crum et al. 2014a; Wilkinson and Brown 2002).

Primary Cutaneous Paget Disease Primary vulvar cutaneous Paget disease is subdivided into three distinct groups (Table 8).

100

Primary cutaneous Paget disease may be a primary vulvar intraepithelial lesion originating in the vulvar epithelium, or it may represent spread from an underlying cutaneous neoplasm. As a primary intraepithelial proliferation of atypical glandular-type cells, histopathologic features are characterized by relatively large cells, with prominent cytoplasm, that are scattered within the squamous epithelium. These cells are generally larger than the adjacent keratinocytes and have large nuclei with prominent nucleoli. Their cytoplasm is finely granular and amphophilic to basophilic, and may be vacuolated, in contrast to the eosinophilic more homogeneous cytoplasm of keratinocytes. The neoplastic cells are typically clustered in groups or dispersed as single cells within the basal and parabasal areas and also in various layers within the epithelium. They may form small acinar groupings within the surface squamous epithelium. Intraepithelial involvement of underlying hair follicles, eccrine ducts, or other skin appendages extending as deep as 3.6 mm may occur (Konstantinova et al. 2016). Mitotic figures may be present but are not frequent (Fig. 36). Paget cells can be identified on cytologic examination of scrapings from salinemoistened involved areas. Proliferative epithelial lesions may be associated with Paget disease,

Fig. 36 Paget disease, primary type. Paget cells are present singly and in nests. Their pale cytoplasm differentiates them from surrounding keratinocytes. The nuclei of the Paget cells are larger, and their nuclear chromatin coarser, than in the adjacent keratinocytes

E. J. Wilkinson and D. S. Rush

including manifesting with a marked verrucoid papillomatous appearance of the epithelium that may resemble warty changes (Brainard and Hart 2000) (Table 8). Radial growth of intraepithelial Paget disease can occur where the lesion extends to involve significant portions of the labia and vestibule epithelium as well as the contiguous vaginal and perianal mucosa. In some cases invasion of the dermis or submucosa by the Paget cells may occur. Dermal, or submucosal, invasion from overlying intraepithelial Paget cells occurred in 12% in a series of 100 cases (Fanning et al. 1999). Vulvar cutaneous Paget disease that is intraepithelial is a form of intraepithelial neoplasia and can be considered an in situ adenocarcinoma that may become invasive. Primary cutaneous Paget disease can also present as a manifestation of a primary underlying adenocarcinoma of the vulva. Primary vulvar adenocarcinoma is present beneath Paget disease in approximately 4–20% of the cases reported. Its origin may be from primary invasive Paget disease or adenocarcinoma of the Bartholin gland, specialized anogenital glands, or other vulvar glandular structures. Early investigators noted underlying adnexal adenocarcinoma beneath the skin involved with Paget disease in many extramammary cases. This finding led some to conclude that Paget cells in the epidermis represented an intradermal migration of neoplastic cells from an underlying cutaneous tumor, as occurs in the breast. Eccrine carcinoma of the vulva with pagetoid spread of the tumor cells within the vulvar skin presented as Paget disease of the involved skin (Grin et al. 2008). This is a rare occurrence and is not the usual finding. That the origin of most underlying adenocarcinomas associated with Paget disease is not associated with sweat gland tumors was demonstrated in a study of 17 cases that showed that epidermal growth factor (EGF), which is typically present in approximately three-fourths of eccrine or apocrine sweat gland tumors, was absent in vulvar Paget disease (Al-Salameh et al. 2000). Overall mortality related to primary vulvar Paget disease is low, estimated as under 10%; however, when

2

Precursor Lesions and Malignant Tumors of the Vulva

there is associated invasion, or underlying adenocarcinoma, the prognosis will be influenced by the depth of invasion and stage of the tumor. Vulvar Paget disease may also be of non-cutaneous origin and there are two major groups of such cases. One group is Paget disease as a manifestation of an associated adjacent primary anal, rectal, or other non-cutaneous adenocarcinoma. The second group of non-cutaneous Paget disease represents those cases that are a manifestation of bladder or other urothelial carcinoma. Wilkinson and Brown have proposed that this Pagetlike neoplasm of the vulva be classified as PUIN to characterize this neoplasm as a manifestation of bladder (urothelial) neoplasia, rather than of glandular origin (Wilkinson and Brown 2002). Cases of Paget disease as a manifestation of an associated adjacent primary non-cutaneous adenocarcinoma include Paget disease associated with in situ or invasive rectal adenocarcinoma or colonic adenocarcinoma (Fig. 37) or intraepithelial extension from an adjacent cervical adenocarcinoma. Perianal involvement by Paget disease is associated with a high frequency of adenocarcinoma or squamous carcinoma of the rectum and may present with perianal pruritus, pain and burning, or bleeding after defecation (MacLean et al. 2004).

Fig. 37 Paget disease, secondary type, of anal-rectal adenocarcinoma origin. The intraepithelial Paget cells are related to the underlying adenocarcinoma and are present in the overlying epithelium. The Paget lesion is focally contiguous with the invasive anal adenocarcinoma that invades the dermis and focally involves the overlying epithelium

101

Primary perianal Paget disease may also occur and has an increased association with invasive adenocarcinoma as compared to the usual cutaneous vulvar Paget disease. Associated underlying anorectal carcinoma occurs in approximately one third of cases with primary perianal Paget disease. Primary perianal Paget disease may involve the vulva and should be distinguished from primary cutaneous vulvar Paget disease. Although cutaneous primary vulvar Paget disease involving the perianal area cannot be distinguished from Paget disease of anorectal origin by routine histologic examination when involving the vulva, the distinction can be made employing immunohistochemistry. Primary non-cutaneous perianal Paget disease is immunoreactive for CK 20 and negative for gross cystic disease fluid protein-15 (GCDFP-15) and is also associated with rectal Paget disease and, in most reported cases, with rectal adenocarcinoma (Table 7) (Nowak et al. 1998). Adenocarcinoma of the cervix manifesting as vulvar Paget disease is reported (Crawford et al. 1999). In such cases immunohistochemistry would reflect a cervical adenocarcinoma, depending on cellular type, and would have a high probability of containing HPV, especially types 16 or 18, unlike Paget cells of primary cutaneous origin that are typically HPV negative. Paget disease of urothelial origin, PUIN, is a cutaneous manifestation of bladder (urothelial) neoplasia: urinary tract malignancy has been reported with genital Paget disease, and the original report of extramammary Paget disease by Crocker in 1889 was of a man with penile and scrotal Paget disease associated with bladder carcinoma (Nowak et al. 1998). Paget-like vulvar mucosal and dermal changes associated with urothelial neoplasia can be recognized as a distinct entity (Wilkinson and Brown 2002). In these cases, there is involvement of the vagina, vulvar vestibule, including the periurethral area that is typically manifested as an erythematous lesion (Lu and Liang 2015). The cells of PUIN have the cytologic features of urothelial carcinoma in situ, with which PUIN is most commonly associated (Figs. 38 and 39) (Brown and Wilkinson 2002; Wilkinson and Brown 2002; Newsom et al. 2015). Of key

102

E. J. Wilkinson and D. S. Rush

importance regarding the recognition of PUIN is that this vulvar neoplasm is a manifestation of bladder urothelial neoplasia and is not associated with underlying adenocarcinoma.

Fig. 38 PUIN; Paget disease, secondary type, of urothelial origin. The epithelium has pagetoid urothelial cells throughout the full thickness, but most concentrated in the basal and parabasal areas. The “clefts” between the neoplastic urothelial cells and the adjacent epithelial keratinocytes is characteristic of PUIN

Fig. 39 PUIN; Paget disease, secondary type, of urothelial origin. Large neoplastic intraepithelial urothelial cells are present throughout the epithelium

Differential Diagnosis and Adjunctive Studies The PUIN cells do not contain mucin, PAS-positive material, CEA, or GCDFP-15 as do the Paget cells of primary cutaneous Paget disease, or non-cutaneous Paget disease related to adenocarcinoma anorectal area, or other nonvulvar primary sites (Wilkinson and Brown 2002). The cells of PUIN are immunoreactive for CK 7 and may be CK 20 positive and uroplakin II and uroplakin III positive, as are one-half of the primary urothelial carcinomas (Table 2) (Wilkinson and Brown 2002). Paget cells of primary cutaneous Paget disease, or non-cutaneous Paget disease related to adenocarcinoma anorectal area, or other nonvulvar primary sites are positive for PAS (diastase resistant), mucicarmine, aldehyde fuchsin, and Alcian Blue. The cells stain pink against a background of greenish-blue with Movat stain. In addition, Paget cells of cutaneous type, as well as the cells and secretions of normal eccrine and apocrine glands, are rich in CEA and GCDFP-15 and typically are negative for S-100 protein, HMB-45, and Melan-A (Newsom et al. 2015; Wilkinson and Brown 2002). Paget cells of cutaneous type are immunoreactive for CK 7 (Crawford et al. 1999; Wilkinson and Brown 2002). These reactions distinguish typical cutaneous Paget cells from PUIN, HSIL, superficial spreading malignant melanoma, and pagetoid reticulosis. They may also be useful in evaluation of margins in permanent sections. Melanoma is distinguished by being immunoreactive for S-100, HMB-45, and Melan-A, which are typically not identified in Paget cells (Hill et al. 2008; Piura et al. 1999). Paget cells, however, may contain granules of melanin, demonstrable with Fontana-Masson stain, probably produced by neighboring melanocytes and engulfed secondarily by the Paget cell.

2

Precursor Lesions and Malignant Tumors of the Vulva

Molecular subtyping and immunohistochemical studies of seven cases of invasive vulvar Paget disease demonstrated that, like breast adenocarcinomas, they can include all four intrinsic molecular subtypes: three were luminal B, two were luminal A, one was HER2 enriched, and one was basal like. Of these seven, three were classified as luminal B, HER2 amplified. Of the ten cases of noninvasive Paget cases none of these intraepithelial cases were HER2 enriched or basal-like (Tessier-Cloutier et al. 2017).

Clinical Behavior and Treatment The treatment and prognosis of Paget disease of the vulva depends on the type as described herein. For primary cutaneous Paget disease, prognosis depends on whether the lesion is intraepithelial only or if there is an associated invasive Paget disease. Primary intraepithelial Paget disease is usually a slowly progressive, indolent, superficial process. Accordingly, local excision of the visible lesion to the fascia, with 2 cm clinically visibly clear margins of excision, is sufficient. In cases associated with invasive Paget disease, or with underlying skin appendage or vulvar glandular adenocarcinoma, as determined by histopathologic evaluation, treatment includes ipsilateral inguinal-femoral lymphadenectomy. If the tumor extends to the margins of excision, or the excision is inadequate in the face of invasive disease, more extended partial or total vulvectomy may be needed (Baehrendtz et al. 1994; Crawford et al. 1999; Fanning et al. 1999). Recurrent vulvar Paget disease, occurring peripheral to an excised intraepithelial Paget lesion, does not appear to be associated with any significant risk of underlying adenocarcinoma and can be treated by a more conservative approach, such as superficial excision or topical therapy such as imiquimod. With topical imiquimod 5% cream, applied twice weekly for 3 months, complete remission was reported in 70% of 70 patients, with partial remission being reported in 16% patients (Dogan et al. 2017). Despite seemingly adequate excision, recurrence of the disease is frequent. This may be explained by the finding that in the patient with

103

vulvar primary cutaneous Paget disease, clinically normal appearing skin may contain Paget cells. A careful topographic study (Gunn and Gallager 1980) demonstrated that the outline of the histologically involved areas was highly irregular and of much greater extent than the visible lesion. In addition, multicentric foci of Paget cells were found, some occurring in grossly normal-appearing skin. This accounts for the frequent “recurrences” of disease despite seemingly adequate excision. These clinically normal-appearing areas of skin, however, are not associated with underlying skin appendage adenocarcinoma. In surgically treating vulvar Paget disease, excision to the fascia of the clinically involved area is limited to the visible area involved with Paget disease only and is sufficient to excise the Paget disease and a potential underlying adenocarcinoma, if present. The depth of invasion influences prognosis when invasion is present. When the depth of dermal invasion was 1 mm or less, no nodal metastasis or death from tumor was observed in seven such patients, whereas all three patients with tumor invasion beyond 1 mm had inguinofemoral node metastasis (Crawford et al. 1999). Frozen section evaluation of clinically normal-appearing skin margins adjacent to the primary cutaneous Paget disease has not been demonstrated either to improve survival or to reduce recurrence. In a study of 12 patients with involved margins following surgery, 7 (58%) had recurrence, whereas recurrence was seen in 1 of 4 patients (25%) with negative margins (Crawford et al. 1999). Recurrences of primary cutaneous intraepithelial Paget disease after excision of the primary lesion have not demonstrated a significant risk of associated underlying adenocarcinoma. Recurrences at the site of the original tumor or remote from it are treated conservatively by local superficial excision or topical imiquimod. Intraepithelial Paget disease has no significant risk of lymph node metastasis or death (Crawford et al. 1999; Fanning et al. 1999). The issues regarding margin assessment specifically relate to primary cutaneous Paget disease. Margin assessment for Paget

104

disease of non-cutaneous origin, including PUIN, needs to be addressed. In cases of perianal Paget disease, it is necessary to evaluate the rectum and anus to determine if the lesion is a manifestation of underlying rectal Paget disease or rectal or anal adenocarcinoma. In such cases, therapy is directed toward treatment of the rectal or anal carcinoma, and the vulvar Paget disease can be treated as an intraepithelial neoplasm, with local superficial excision, or more conservatively. In vulvar Paget disease associated with other adjacent adenocarcinomas, such as cervical adenocarcinoma, the treatment is focused on the primary adenocarcinoma, and the associated Paget disease is treated as an intraepithelial neoplastic process, with superficial excision. The prognosis in such cases is dependent on the stage and behavior of the associated adenocarcinoma. Therapy for vulvar PUIN is directed toward the bladder urothelial neoplasm, and the vulvar pagetoid intraepithelial urothelial neoplastic process is treated conservatively, as an intraepithelial neoplasm. Total vulvectomy and excision to the deep fascia is not indicated in this circumstance (Wilkinson and Brown 2002).

Intestinal-Type Mucinous Adenocarcinoma (Villoglandular Mucinous Adenocarcinoma, Cloacogenic Carcinoma, Adenocarcinoma of Cloacogenic Origin) These rare tumors of the vulva all have features of glandular tumors of the colon and are of uncertain origin. They are usually solitary, but may involve more than one site. The tumor has been reported as an in situ lesion on the mucosa of the hymen (Dubé et al. 2006). The tumor presents as a cutaneous mass and may be polypoid, excoriated, and/or inflamed (Willen et al. 1999). They appear to be primary mucinous tumors of the vulvar epithelium. Benignappearing mucinous glandular elements have been reported in the dermis deep to the tumor

E. J. Wilkinson and D. S. Rush

in one case (Zaidi and Conner 2001). It is also possible that they arise from some remnant of the early cloaca within the vulva, as the older term implies. Current opinion is that they arise probably from misplaced Skene periurethral glandular elements (McCluggage 2016). On microscopic examination the tumor may be in situ, entirely confined to the epithelium, with a villoglandular, papillary appearance with neoplastic colonic-type glandular epithelial cells on the surface of the tumor. The neoplastic mucinous epithelium is in continuity with the overlying epithelium and may resemble colonic villous adenoma when a predominant intraepithelial component is present. When invasion is present, it is in continuity with the overlying neoplastic epithelium. The tumor is not associated with an underlying dermal glandular neoplasm as may be seen with Paget disease. The invasive adenocarcinoma resembles mucinous colonic carcinoma, with neoplastic colonic-type epithelium with goblet cells and intracytoplasmic mucin. Apocrine differentiation is not observed (Willen et al. 1999; Zaidi and Conner 2001). In one study these tumors are reported as being reactive with mucicarmine and Alcian Blue pH 5 and immunoreactive for polyclonal CEA, CK- 17, CAM 5.2, and p53 antigen. In addition, S-100 and chromogranin are also reported as immunoreactive in some neoplastic epithelial cells, evidence of the presence of endocrine cells. The tumors are not reactive for monoclonal CEA and ER and PR. Electron microscopic studies have provided additional evidence of a colonic-type epithelium within these tumors (Dubé et al. 2006; Willen et al. 1999; Zaidi and Conner 2001). Although experience with these lesions is limited due to their rarity, the appropriate treatment of these tumors is currently felt to be the same as for similar lesions of colorectal mucosa, with more conservative treatment for those that are entirely intraepithelial. For deeper tumors, risk of lymph node metastasis exists, although regional lymph node involvement is apparently rare. Treatment with partial deep vulvectomy (deep local excision) has reportedly been effective (Willen et al. 1999).

2

Precursor Lesions and Malignant Tumors of the Vulva

Bartholin Gland Tumors General Features A wide variety of tumors may arise from the Bartholin gland. The criteria for the diagnosis of a tumor of Bartholin gland origin are that the neoplasm must (1) arise at the site of Bartholin gland, (2) be consistent histologically with a primary neoplasm of Bartholin gland, and (3) not be metastatic. Adenocarcinomas account for approximately 40% of Bartholin gland carcinomas, but others include squamous cell carcinoma (40%), adenoid cystic carcinoma (15%), transitional cell carcinoma (less than 5%), adenosquamous carcinoma (less than 5%), and poorly differentiated adenocarcinomas (Obermair et al. 2001; Ouldamer et al. 2013; Heller and Bean 2014). Clinical Features Carcinoma of the Bartholin gland usually presents as an enlargement in the gland area and may present as an apparent Bartholin cyst. The average age of women with this tumor is 50 years, with most between 40 and 70 years.

105

Other Primary Bartholin Gland Carcinomas Squamous cell carcinomas arising in the Bartholin gland have the same microscopic appearance as those arising elsewhere in the vulva. These tumors are typically immunoreactive for CEA. In some cases HPV 16 has been identified (Heller and Bean 2014). Adenoid cystic carcinomas arising in the Bartholin gland are similar to those occurring in the salivary glands, upper respiratory tract, and skin. There are two types of adenoid cystic carcinoma recognized in the lower genital tract (Xing et al. 2016). The pure adenoid cystic carcinoma is composed of uniform, small cells arranged in cords and nests with a cribriform pattern. Variable-sized cysts filled with amphophilic or eosinophilic acellular basement membrane-like material also may be encountered (Figs. 40 and 41). Keratin and S-100 antigen are detectable by immunohistochemical techniques. The S-100 reactivity may demonstrate a

Gross Findings Bartholin gland tumors are typically solid, deeply infiltrative and occupy the site of the gland. Occasionally a Bartholin cyst may obscure the associated deeper tumor. Tumors can range from 1 to 7 cm in diameter or larger. Microscopic Findings Adenocarcinomas of the Bartholin gland usually are nonspecific in type, but mucinous and papillary types are described. The tumors usually contain intracytoplasmic mucin and are immunoreactive for CEA. Fine needle aspiration cytology may be of value in diagnosis (Heller and Bean 2014). Differential Diagnosis The differential diagnosis of Bartholin gland adenocarcinoma includes adenocarcinoma of skin appendage origin and metastatic adenocarcinoma. Metastatic tumors typically do not involve the Bartholin gland, and the metastatic tumor type may not be consistent with a primary tumor of the Bartholin gland.

Fig. 40 Adenoid cystic carcinoma. Relatively small but somewhat pleomorphic cells surrounding sharply punched-out cystic spaces

106

Fig. 41 Adenoid cystic carcinoma. The tumor is composed of relatively small hyperchromatic cells arranged in well-circumscribed nests within the stroma. Well-defined cystic spaces are evident in this tumor. The surrounding stroma is desmoplastic. (Courtesy of R.J. Kurman, M.D., Baltimore, MD)

myoepithelial cell element. This pure type of tumor is not associated with HPV. The second type is associated with HPV and found in tumors of the cervix, rather than the vulva, and is a mixed type of tumor with a squamous cell carcinoma component. Gene rearrangement studies demonstrate that both types of genital-related adenoid cystic carcinomas can have chromosomal translocation of the genes encoding transcription factors MYB and NFIB functions. Six of the nine vulvar adenoid cystic carcinomas demonstrated NFIB rearrangement, and two of the nine had MYB rearrangement (Xing et al. 2016). The differential diagnosis of adenoid cystic carcinoma includes adenocarcinoma, basal cell carcinoma, metastatic atypical carcinoid, and

E. J. Wilkinson and D. S. Rush

small cell carcinoma. Metastatic small cell carcinoma from the vagina can also present as a Bartholin mass (Mirhashemi et al. 1998). Adenocarcinomas lack the uniform acinar arrangement and intraluminal basement membrane material of adenoid cystic carcinoma. Basal cell carcinomas are more solid and lack the cystic spaces and the intracystic basement membrane-like material. Metastatic carcinoids and small cell carcinomas are more solid, have fewer lumens, contain argyrophil cells, and often stain for neuroendocrine markers. Adenosquamous carcinoma of the Bartholin gland contains a mixture of squamous cells with intracellular bridges and glandular cells that typically contain mucin. Origin in a hidradenoma has been observed. Transitional cell carcinoma arising in the Bartholin gland comprises under 5% of cases. (Ouldamer et al. 2013). The tumor is composed of uniform polyhedral or rounded epithelial cells often lining broad papillary fronds. Rare areas of glandular or squamous differentiation may be present. The differential diagnosis of primary transitional cell carcinoma of the Bartholin gland includes poorly differentiated squamous cell carcinoma and adenocarcinoma. If more than rare foci contain glands or show keratinization, the tumor is of mixed cell type and should be so designated, listing the different tumor types. Rare examples of other primary, but rare, epithelial and neuroendocrine malignant tumors have been reported arising in the Bartholin gland including epithelioid-myoepithelial carcinoma, small cell neuroendocrine carcinoma (SCNEC), Merkel cell carcinoma, and lymphoepithelioma-like carcinoma (Heller and Bean 2014). Low-grade epithelial-myoepithelial carcinoma of the Bartholin gland has also been reported (McCluggage et al. 2009). Rare malignant soft tissue tumors including leiomyosarcoma and epithelioid sarcoma, as well as non-Hodgkin’s lymphoma, have also been reported arising in the Bartholin gland (Heller and Bean 2013).

2

Precursor Lesions and Malignant Tumors of the Vulva

Clinical Behavior and Treatment Approximately 20% of carcinomas of the Bartholin gland are associated with metastases to the inguinofemoral lymph nodes. The overall 5-year survival of patients with Bartholin gland carcinomas is approximately 50% when the groin nodes are free of tumor, but decreases to under 20% when two or more nodes are involved. If the groin nodes are involved, there is a 20% probability that pelvic lymph node metastasis also will be involved, but if the groin nodes are free of metastasis, there is essentially no risk of pelvic node metastasis. The primary treatment for Bartholin gland carcinomas is dependent on the stage of the tumor. Primary surgery may be wide local excision to the fascia or larger excision depending on clinical judgment. Ipsilateral or bilateral inguinofemoral lymph node dissection is often necessary, regardless of the type of primary excision (Ouldamer et al. 2013). Adjunctive radiation therapy to the vulva and regional lymph nodes may also be needed. Therapy for adenoid cystic carcinoma of the Bartholin gland is wide local excision with ipsilateral inguinofemoral lymphadenectomy (Copeland et al. 1986). Local recurrence is well documented, and, when the tumor involves the margin, adjuvant radiotherapy may be beneficial. Survival is better with adenoid cystic carcinoma than with other forms of carcinoma of the Bartholin gland. The treatment of vulvar adenosquamous carcinoma is similar to that of squamous cell carcinoma. Adenosquamous carcinomas have a poorer prognosis than squamous cell carcinomas, partly because of the higher frequency of lymph node metastasis. Merkel cell carcinoma is a rare tumor of the vulva and may occur in vulvar sites other than the Bartholin gland. Mean age at presentation was 59.6 years in a review of 17 cases, with the tumor presenting as a firm, mobile vulvar mass that was ulcerated in some cases. Reported therapy was predominately surgical excision with adjuvant radiation therapy. Following therapy, recurrence was observed in 70.6% of the cases, with a mortality of 47% at 7.8 months (Nguyen et al. 2017).

107

Skene Gland and Duct Adenocarcinomas Skene duct cysts, adenomatous hyperplasia, benign tumors including leiomyoma, adenofibroma (fibroadenoma), and other benign tumors may result in a mass in the Skene gland and/or duct. Primary adenocarcinomas arising in Skene glands and a periurethral cyst have also been reported (Murphy et al. 2004; Nagano et al. 2002). Skene gland adenocarcinomas are rare tumors, accounting for less than 0.003% of female genital tract malignant neoplasms (Heller 2015), and when identified often involve the urethral mucosa as well. Skene gland tumors are thought to arise from the luminal secretory cells of these glands, and most have been demonstrated to express prostate-specific antigen (PSA) and prostate acid phosphatase and may morphologically have the appearance of prostate adenocarcinoma as well, reflecting the homology between Skene periurethral glands and the prostate (Pongtippan et al. 2004). Such cases may also be associated with elevated PSA serum levels. Clear cell adenocarcinoma has also been reported, as has adenocarcinoma with neuroendocrine differentiation and adenoid cystic carcinoma (Heller 2015; Korytko et al. 2012). Treatment for benign Skene gland/duct tumors is local excision. The treatment of malignant Skene gland/duct tumors depends on the stage but is usually surgical excision. Radiation therapy may also be employed. Serum PSA levels can be followed to monitor therapy for those tumors that express PSA (Korytko et al. 2012).

Mammary Gland-Like Adenocarcinomas, Including Phyllodes Tumor, Arising in Vulvar Specialized Anogenital Mammary-Like Glands Specialized anogenital mammary-like glands, found typically in the interlabial sulcus, are thought to be the origin of vulvar mammary-like

108

(breast-like) adenocarcinomas (Van der Putte 1994). The origin of these tumors was once thought to be ectopic breast tissue; however, current opinion is that these glands are not ectopic tissue, but rather normal anatomic elements. The adenocarcinomas arising from these glands have occurred predominately in postmenopausal women. The histopathologic features are very similar to those of primary mammary adenocarcinomas, and a few have been associated with breast adenocarcinoma. Most have a pattern of growth of infiltrating, well-differentiated adenocarcinoma. In some cases an intraductal carcinoma component has been identified. Malignant phyllodes tumors have also been reported arising in these glands (Fu et al. 2011). These tumors are grossly and microscopically identical to those arising in the breast. Growth is in a pushing-like pattern against the adjacent normal tissues. As in the breast, these tumors may be of high or low grade based on the stromal cellularity and cellular atypia. The most prominent stromal cellularity, mitotic activity, and atypia, if present, are seen immediately adjacent to the epithelial elements. Complete surgical excision is the usual treatment. Local recurrences may occur; however, the prognosis is usually very good (Crum et al. 2014a; Kazakov et al. 2010). Molecular subtyping and immunohistochemical studies of seven mammary-like vulvar tumors from women with a median age of 72 years demonstrated that, like breast adenocarcinomas, they can include all four intrinsic molecular subtypes: three were luminal B, two were HER2 enriched, and one each of luminal A or basal-like (TessierCloutier et al. 2017). Metastases to inguinal lymph nodes from adenocarcinomas of vulvar mammary-like tumors may contain secretory material similar to that of breast adenocarcinomas, including alpha-lactalbumin and milk fat globulin protein, and estrogen and progesterone receptors may also be expressed. Metastatic adenocarcinoma to the vulva can be distinguished from primary adenocarcinomas if mammary-like glands or in situ adenocarcinoma is present. In the absence of mammary-like tissue, the distinction may not be possible. Treatment of primary adenocarcinoma arising within vulvar

E. J. Wilkinson and D. S. Rush

specialized anogenital glands is extended wide local excision and ipsilateral inguinofemoral lymphadenectomy. Treatment is similar to that for breast adenocarcinoma, with complete surgical excision, when possible, followed by chemotherapy and anti-estrogen therapy if estrogen receptors are present (Crum et al. 2014a; Kazakov et al. 2010).

Carcinomas of Sweat Gland Origin Carcinomas of vulvar sweat gland origin are rare, comprising less than 1% of all vulvar carcinomas. Patients typically present with a painless vulvar mass. In addition to undifferentiated sweat gland adenocarcinomas, ductal eccrine carcinoma, eccrine porocarcinoma, eccrine hidradenocarcinoma, clear cell hidradenocarcinoma, apocrine carcinomas, and spiradenocarcinoma are documented (Baker et al. 2013). Eccrine carcinoma of the vulva may have a “Pagetoid” appearance (Grin et al. 2008). Adenosquamous carcinoma can arise in a vulvar hidradenoma. Sebaceous carcinoma of the vulva can be associated with HSIL (Baker et al. 2013; Escalonilla et al. 1999). Mucinous adenocarcinoma with focal squamous and neuroendocrine differentiation is documented arising in the labium majus. This tumor expressed chromogranin A, protein gene product 9.5, serotonin, and vasoactive intestinal polypeptide (Graf et al. 1998).

Malignant Melanoma General Features Melanomas are the second most common vulvar malignancy, after squamous cell carcinoma, and account for approximately 3% of all melanomas in women and approximately 8–10% of all vulvar malignant neoplasms (De Simone et al. 2008; Ivan and Prieto 2015; Irvin et al. 2001) (Table 9). Melanomas are a complex group of tumors and are discussed in more detail in other references sites (Elder and Murphy 2010). Melanomas occur predominantly in white women. The mean age at diagnosis, in a large retrospective literature study,

2

Precursor Lesions and Malignant Tumors of the Vulva

109

Table 9 AJCC staging for melanomas is based on tumor thickness (T) Definition of primary tumor (T) T category Thickness Ulceration status TX primary tumor thickness cannot be assessed (e.g., diagnosis by curettage) T0 no evidence of primary tumor (e.g., unknown primary or completely regressed melanoma) Ts (melanoma in situ) Not applicable Not applicable T1 1.0–2.0 mm With ulceration T3 >2–4 mm Unknown or unspecified T3a >2–4 mm Without ulceration T3b >2–4 mm With ulceration T4 >4 mm Unknown or unspecified T4a >4 mm Without ulceration T4b >4 mm With ulceration From AJCC Cancer Staging Manual, 8th ed. New York: Springer Science + Business; 2017 (Gershenwald et al. 2017, p577). Used with permission of the American Joint Committee on Cancer (AJCC), Chicago, Illinois. The original and primary source for this information is the AJCC Cancer Staging Manual, Eighth Edition (2017) published by Springer International Publishing

was 62.2 years, with a range from 53 to 80 years (Heinzelmann-Schwarz et al. 2014). The age-specific incidence in a comprehensive Swedish study was 1.28 per 100,000 in women aged 75 or older. In women 60–74 years old, the incidence was 0.56 and in those 45–59, 0.19. In those 30–44 it was 0.08 and 0.02 or less in women under 29 years of age (Ragnarsson-Olding et al. 1999). Rare cases of vulvar melanoma have been reported in children (Egan et al. 1997). Cases reported in children have mostly been associated with lichen sclerosus, and such cases present the difficult problem of differentiating a genital nevus, or a compound melanocytic nevus overgrowing lichen sclerosus from melanoma, or from an atypical genital melanocytic nevus in the presence of lichen sclerosus (Mulcahy et al. 2013; Ivan and Prieto 2015). Vulvar bleeding was the most common presenting symptom of 198 women with vulvar melanoma, being recorded in 35% of the patients. A vulvar mass was observed in 28%, an ulcer in 5%, and a mole in 5% (Ragnarsson-Olding et al. 1999). Pruritus occurred in 15%, irritation and or burning in 14%, and discomfort with urination in 12%. In a

retrospective study of 33 cases, 72.2% of the women detected the lesion by self-examination, and the average time elapsed between detection of the melanoma and seeking medical help was 28.2 months. Of these patients, 53.3% had clinically ulcerated lesions, with a mean tumor size of 21.9 mm (5–50 mm) (Heinzelmann-Schwarz et al. 2014). An epidemiologic study of 762 women with vulvar/vaginal melanoma compared outcomes to a large cutaneous melanoma group and identified the mean age at diagnosis of 68 years for the vulva/ vaginal group and 52 years for the cutaneous group. Negative prognostic factors for survival included older age, race, advanced stage, involved lymph nodes, and history of radiation therapy (Mert et al. 2013). Vulvar melanomas may arise from a preexisting benign or atypical pigmented lesion. The melanoma presents as a mass in approximately 40% of cases, but change in a known pigmented lesion, or ulcer, may also be the initial finding. Pruritus, bleeding, or drainage from the lesion may be noted. Although generally pigmented, vulvar melanomas are amelanotic (nonpigmented) in approximately one-quarter of the cases (De Simone et al. 2008;

110

E. J. Wilkinson and D. S. Rush

Fig. 43 Superficial spreading malignant melanoma. Pagetoid spread of the melanoma cells into the upper third of the epidermis. Markedly atypical melanocytic cells are at the epithelial-dermal junction. No invasion is present

Fig. 42 Malignant melanoma. A superficial spreading malignant melanoma with vertical growth within the field of a superficially spreading malignant melanoma. (Courtesy of Linda S. Morgan, M.D., Gainesville, FL)

Moxley and Fader 2011; Ragnarsson-Olding et al. 1999). The majority of vulvar melanomas present as poorly circumscribed lesions with irregular contours on vulvar mucosa or mucosal-cutaneous sites, with less than one-quarter of the cases presenting on vulvar keratinized skin (Ivan and Prieto 2015). In a study of 33 cases, 31% of the melanomas presented on the labia minora (HeinzelmannSchwarz et al. 2014). A significant number of vulvar melanomas are multifocal and represented 26.3% of the cases in one large study (Heinzelmann-Schwarz et al. 2014). Melanomas of the vulva are classified into three distinct histopathologic types. These three types, in increasing order of frequency, are superficial spreading melanoma (Figs. 42, 43, and 44), nodular melanoma (Figs. 45, 46, and 47), and mucosal/acral lentiginous melanoma (Fig. 48). About one-quarter of the cases may be mixed or

Fig. 44 Superficial spreading melanoma with vertical, invasive growth

Fig. 45 Nodular melanoma, partial deep vulvectomy. The tumor is deeply pigmented and clearly demarcated from the adjacent vulvar epithelium. The tumor is deeply invasive. (Courtesy R. Foss, PA, Gainesville, Fl)

2

Precursor Lesions and Malignant Tumors of the Vulva

Fig. 46 Nodular melanoma. The tumor is deep within the dermis, with overlying epithelium that does not have an intraepithelial melanocytic lesion. The melanoma cells are large and polygonal and arranged in nests, sheets, and cords within the dermis

Fig. 47 Malignant melanoma, amelanotic, spindle cell type. The overlying epithelium has in situ melanoma with neoplastic cells within the basal and parabasal epithelium. The invasive melanoma consists of spindle-shaped neoplastic cells. Melanin pigment is not seen and the tumor is deeply invasive

otherwise unclassifiable (Ragnarsson-Olding et al. 1999). Thin melanomas also occur (de Giorgi et al. 2005). The relative frequency of these types differs in various reports (De Simone et al. 2008; Verschraegen et al. 2001; Wechter et al. 2004). Mucosal/acral lentiginous melanoma was the most common type identified in the Karolinska series, observed in 52% of the cases, whereas nodular melanoma accounted for 20%

111

Fig. 48 Mucosal/acral lentiginous melanoma. This variant of malignant melanoma has a radial component with a lentiginous pattern at the mucosal-stromal interface. In this case, spindle cells are present within the submucosa near the junctional zone and within the deeper dermis with an associated desmoplastic response. The cytologic uniformity is characteristic. No pagetoid spread is evident. This type of malignant melanoma occurs within the mucosa of the vulvar vestibule

and superficial spreading melanoma 4% of the 198 cases reported (Ragnarsson-Olding et al. 1999). This predominance of mucosal/acral lentiginous melanomas reflects the mucosal site of origin on the vulva. There is some variation in the frequency of melanoma type related to the vulvar anatomic site of the tumor. In one large series, mucosal/acral lentiginous melanomas occurred with similar frequency in glabrous, hairy-glabrous, and hairy skin, whereas nodular melanomas were seen primarily in glabrous and hairy-glabrous skin and less frequently in hairy skin. In contrast, superficial spreading melanomas occurred predominately in hairy skin (Ragnarsson-Olding et al. 1999). Some of the variation in the type of vulvar melanoma reported may relate to differences in the criteria used to distinguish superficial spreading melanoma from nodular melanoma. Superficial spreading melanoma can be differentiated from nodular melanoma by evaluating the adjacent epithelium. If the radial growth of a melanoma or atypical melanocytes involves three or more adjacent rete ridges, the tumor is classified as superficial spreading melanoma (see Fig. 44) (Smoller et al. 2016).

112

Microscopic Findings The histopathologic features of vulvar melanoma vary considerably, and certain features correlate with the subtype of the melanoma. Malignant melanomas may consist predominantly of epithelioid, dendritic (nevoid), or spindle cell types, either pure or mixed, within a given tumor. The cells may contain no melanin, or variable amounts, ranging from minimal to very large quantities. Mitotic figures are usually identifiable in both the intraepithelial and invasive melanocytes. When reporting mitotic count per square millimeter, only mitotic figures in the invasive component are counted. Capillary-like vascular and perineural invasion by tumor is a common finding. Mucosal/acral lentiginous melanomas occur within the mucosal areas, predominately the vestibule of the vulva. They have both vertical growth, as seen in nodular melanoma, and radial growth, consisting of extension of atypical melanocytes within the adjacent epithelium, as is seen adjacent to acral lentiginous and superficial spreading melanomas (Fig. 48). Within the mucosal/acral lentiginous melanoma, junctional melanocytes are numerous and often have a spindle to oval shape and are found predominately in the basal and parabasal epithelium and are found confluent with the invasive tumor and in the adjacent portion of the intraepithelial tumor reflecting the radial growth of the tumor. The cells may be in nests but pagetoid spread is minimal or absent. With invasion, the invasive tumor cells have features like the intraepithelial cells and lack maturation. A desmoplastic submucosal response may be present about the invasive tumor. Superficial spreading melanomas occur within the skin and hair-bearing areas. Pigmentation may be highly variable. The neoplastic melanocytic cells are in the junctional, basal, and parabasal areas and usually have a nested distribution. Pagetoid distribution within the epithelium is often present and may be prominent. The tumor cells are relatively large with nuclei that are fairly uniform shape with prominent nuclei. With invasion of the tumor, the neoplastic melanocytes in the dermis are similar in appearance to the

E. J. Wilkinson and D. S. Rush

neoplastic melanocytic cells within the epithelium. The invasive tumor cells may have variable cellular features but do not mature in the deeper portions of the tumor. Mitotic figures may be evident. Capillary-like vascular and perineural invasion by tumor is a common finding. In nodular melanomas, an intraepithelial prominent component may be present in addition to an invasive component, but without an adjacent radial growth phase, involving less than three rete ridges. The cells of nodular melanomas maybe polygonal (epithelioid) or spindle shaped. The polygonal cells contain abundant eosinophilic cytoplasm, large nuclei, and prominent nucleoli. The dendritic cells have tapering cytoplasmic extensions resembling nerve cells and show moderate nuclear pleomorphism. Spindle cells have smaller, oval nuclei and can be arranged in sheets or bundles.

Differential Diagnosis Superficial spreading malignant melanoma must be distinguished from a number of vulvar lesions including, but not limited to, Paget disease, squamous intraepithelial lesion (HSIL/VIN), dysplastic and atypical vulvar nevi, melanocytic nevus overgrowing lichen sclerosus, PUIN, and pagetoid/Bowenoid reticulosis (Ivan and Prieto 2015). The cells of Paget disease usually are larger than superficial spreading melanoma, have more cytoplasm, and are clustered with occasional gland formation. Squamous cell carcinomas with tumor giant cells, or those predominantly composed of spindle cells, may resemble malignant melanoma. Typical squamous cell carcinoma may be identifiable adjacent to the giant cell or spindle cell component, which may establish the diagnosis. Spindle cell tumors of soft tissue origin, large cell lymphomas, pagetoid reticulosis, Kaposi’s sarcoma, and metastatic tumors, including choriocarcinoma, may be included in the differential diagnosis. In these cases, review of the clinical history and physical and radiologic findings, as well as thorough sectioning of the submitted tissue and appropriate immunohistochemical studies, usually will contribute evidence to establish

2

Precursor Lesions and Malignant Tumors of the Vulva

the diagnosis. It is emphasized that when faced with a poorly differentiated vulvar tumor that defies classification on initial microscopic examination, melanoma should be placed first on the list of the differential diagnosis. Paget disease, HSIL, and PUIN can be distinguished from melanoma with histochemical and immunohistochemical studies for mucin, CEA, S-100 protein, HMB-45 (or Melan-A), and cytokeratins. Paget cells, HSIL, and PUIN, regardless of origin, are immunoreactive for cytokeratin markers such as CAM 5.2, CK7, and epithelial membrane antigen (EMA). Primary cutaneous Paget disease and Paget disease of anorectal origin also typically contain cytoplasmic mucin that stains with mucicarmine stain and are immunoreactive for CEA, whereas melanomas do not (see Table 2). Pagetoid reticulosis and Kaposi’s sarcoma will typically express CD45 and related lymphoproliferative markers. Melanomas usually are immunoreactive for S-100 protein, HMB-45, MART1, Melan-A, MITF, and Sox10, whereas the non-melanocytic tumors, including Paget, PUIN, HSIL, squamous cell carcinoma, pagetoid/ Bowenoid reticulosis, and Kaposi’s sarcoma, are negative (Ivan and Prieto 2015; Wilkinson and Brown 2002). Atypical genital nevi and a compound melanocytic nevus overgrowing lichen sclerosus are both usually well circumscribed and have minimal upward pagetoid migration of the melanocytes. Within the epidermis both have maturation of nevus cells in the deeper portions of the nevus, and mitotic figures are typically absent in these mature deeper melanocytes. Immunohistochemical studies for Ki-67 demonstrate low or no proliferation in the deeper melanocytes with HMB-45 expression in melanocytes in the epidermis and dermis (Ivan and Prieto 2015).

Staging, Clinical Behavior, and Treatment The AJCC staging for melanomas is based on tumor thickness (T) with or without ulceration, regional lymph node status (N), and presence of distant metastasis (M). Current staging for melanomas of the vulva uses the staging for melanoma

113

of the skin, and the staging criteria are the same regardless of growth pattern (Gershenwald et al. 2017; Gibb et al. 2017). Histologic grading of melanomas is not included in the AJCC staging system. For melanoma staging: “Tumor thickness is measured from the top of the granular layer of the epidermis to the deepest invasive cell across the broad base of the tumor” or if the tumor is ulcerated, from the “. . ..base of the ulcer to the deepest invasive cell across the broad base of the tumor” (Gershenwald et al. 2017). These currently defined AJCC measurements do not state how the measurement is made when a granular layer is not present and the lesion is not ulcerated, as is sometimes the case with mucosal melanomas. A common practice in that situation is to measure from the surface of the epithelium to the deepest invasive cell, as defined above (Table 9). Current AJCC lymph node (N) staging includes assessment of the presence or absence or tumor involving the lymph nodes, but also includes reporting the presence of in-transit, satellite, and/or microsatellite metastases. Clinically detected metastases by sentinel lymph node biopsy are included in those metastases classified as occult (Gershenwald et al. 2017). Distant metastasis (M) reporting includes reporting the presence or absence of evidence of distant metastasis to the skin, soft tissue, as well as muscle and non-regional lymph nodes, lung, central nervous system or other sites. Serum LDH levels are considered and included in staging and recorded as follows: not applicable, not recorded, unspecified, normal, or elevated (Gershenwald et al. 2017). The 2017 AJCC staging system does not use the Clark levels of invasion for staging because it applies only to keratinized skin and is not applicable to the mucosal sites of the vulva where the majority of melanomas are identified. Prior studies, however, have reported that both the level of invasion of a malignant melanoma (Clark level when applicable) and its thickness have prognostic significance (Raspagliesi et al. 2000). Factors that can adversely influence survival include a tumor thickness exceeding 1 mm, a mitotic count exceeding 10 per square mm, surface ulceration, and a minimal or absent

114

inflammatory reaction. When reporting mitotic count per square millimeter, only mitotic figures in the invasive component are counted. Vascular space invasion and tumor necrosis also are associated with a poorer prognosis and are seen more commonly with large melanomas. An excellent prognosis has been associated with melanomas with a thickness of 1.49 mm or less. A tumor volume of less than 100 mm3 also correlates with an excellent prognosis (Trimble 1996). The usual treatment for vulvar melanomas with a thickness of 0.75 mm or less is wide local excision with a 1 cm circumferential and 1–2 cm deep margins. This is also acceptable treatment for some melanomas with a thickness of 1 mm or less, although there is limited data available for vulvar melanomas with a thickness between 0.76 and 1.0 mm. Melanomas 1.1–4 mm thick require 2 cm surgical margins with deep margins of at least 1–2 cm (Trimble 1996). The usual initial treatment for thicker vulvar invasive melanomas is wide, deep excision of the lesion with an adequate tissue margin (partial deep vulvectomy). Unilateral inguinal-femoral lymphadenectomy is commonly performed on the same side of the vulva as the primary tumor (HeinzelmannSchwarz et al. 2014). In a multicenter study of 77 patients, 73% had stage I or II disease, and Breslow thickness was found associated with recurrence but not survival (Moxley and Fader 2011). In that study, radical surgery (total deep vulvectomy with bilateral inguinal lymph node dissection) did not improve survival, but did increase morbidity. Wide local excision was associated with a higher survival rate in one study (Tcheung et al. 2012). In addition, surgical tumor-free margins of 1 cm or greater were also reported to be a predictor of better prognosis and survival (Heinzelmann-Schwarz et al. 2014). Molecular studies are presently favored as testing to determine therapy, and the role of immunohistochemical studies in assessment for therapy is being studied (Ivan and Prieto 2015). Immunohistochemical study for c-KIT (antihuman CD117 antibody) has demonstrated some correlation between strong expression and disease-free and relapse-free survival (Heinzelmann-Schwarz et al. 2014). Mutation analysis of 33 vulvar melanomas

E. J. Wilkinson and D. S. Rush

demonstrated that C-KIT and NRAS mutations were more frequently identified than expected in melanomas of other sites, but BRAF mutations were less common (Rouzbahman et al. 2015). Molecular studies on vulvar melanomas to detect increased copy numbers and/or mutations of gene encoding the receptor tyrosine kinase KIT have identified a KIT mutation rate of 35–40% in vulvar mucosal melanomas. This gives promise that targeted therapy with KIT inhibitor-targeted therapies, such as imatinib mesylate, will significantly contribute to therapy as presently understood. Vulvar melanomas may recur locally or in the cervix, urethra, vagina, or rectum. In cases with recurrence, the median disease-free survival is highly variable but may be under 12 months and median survival under 3 years (Ferraioli et al. 2016). Distant metastasis may be the first sign of recurrence. Metastases to the lungs, brain, urinary bladder, bone marrow, and abdominal wall have all been observed. The prognosis after recurrence is guarded, with a 5-year survival of approximately 5%.

Other Malignant Tumors of the Vulva Carcinosarcoma Carcinosarcoma is a rare tumor of the vulva with only a few reported cases, and of these three have been associated with a primary vulvar squamous cell carcinoma (Lordello et al. 2017). Vulvar carcinosarcoma with spiradenocarcinoma associated with an eccrine spiradenoma has been documented (Chen et al. 2011). A recent case presented as a rapidly enlarging vulvar cyst composed of mucinous adenocarcinoma and anaplastic spindle cell carcinoma with chondrosarcomatous and osteosarcomatous components. This was associated with recurrence and death from tumor within months after surgery (Lordello et al. 2017).

Malignant Blue Nevus Malignant blue nevus is a rare tumor but has been reported as a primary vulvar tumor, arising in the labium majus of a 28-year-old woman. This

2

Precursor Lesions and Malignant Tumors of the Vulva

patient had an ovarian metastasis from this tumor 15 years after therapy. Malignant blue nevus typically has significant nuclear atypia, but a low mitotic rate index (Spatz et al. 1998).

Yolk Sac Tumor (Endodermal Sinus Tumor) Primary extragonadal yolk sac tumor is rare in extragonadal sites, but has been reported in a series of 15 cases. Of these, 1 was primary labial mass of the vulva, 11 were in the uterus, and 1 each in the vagina, bladder, and peritoneum (Ravishankar et al. 2017). In the vulva, vagina, and pelvis, the tumor is reported primarily in children and young women (Flanagan et al. 1997). The characteristic histopathologic features are quite variable, and although Schiller-Duval bodies and eosinophilic hyaline droplets are classic features, they are not always found. These features may not be evident in cases with other patterns including reticular, micropapillary, microcystic, and hepatoid. The eosinophilic droplets are PAS positive and diastase resistant and express alpha-fetoprotein on immunohistochemical study. A second germ cell component may also be identified. The histologic appearance does not appear to influence prognosis in tumors of the ovary, but there are insufficient cases within the vulva to evaluate this relationship. The variable patterns of growth may resemble adenocarcinoma, which is the primary differential diagnosis. Immunohistochemical studies of 14 genital cases demonstrated alpha-fetoprotein reactivity in one-half of the cases (Ravishankar et al. 2017). Additional studies reported as reactive include SALL 4, 12/12; CDX2, 10/12; glypican-3, 9/10; CK 20, 5/9. CK 7 and PAX8 were reactive is under one-half of the cases. If identified in the serum, elevated alpha-fetoprotein (AFP) can be used in monitoring follow-up. Therapy for vulvar yolk sac tumor is wide local excision of the tumor and chemotherapy in most cases (Ravishankar et al. 2017). Platinum-based chemotherapy has markedly improved survival in patients with this tumor, and better chemotherapy treatments will continue to improve survival.

115

Primary Malignant Lymphoma Although few cases are reported in the literature, the vulva is the second most common site, after the cervix, for malignant lymphoma involving the genital tract. These tumors occur predominately in women of reproductive age and postmenopausal women, but may occur in children. Rarely lymphoma may arise in a prior area of radiation therapy (Curtis et al. 2006). When identified the lymphoma may be primary, but may represent metastatic involvement. In such cases the lymphoma is often disseminated (Wang et al. 2017). The tumor may present as a Bartholin mass, clitoral enlargement, subcutaneous mass, and ulcerated or destructive neoplasm or mimic other vulvar tumors (Nucci et al. 2014; Vang et al. 2002). Diffuse large B-cell lymphoma is recognized to be the most common lymphoma involving the vulva (Nucci et al. 2014). There are exceptions as in a review of 29 cases of primary non-Hodgkin’s lymphoma involving the vulva, where only 8 of the 29 cases were of the diffuse large B cell type (Clemente et al. 2017). Other tumor types reported include, but are not limited to, diffuse mixed cell lymphoma, follicular large cell lymphoma and peripheral T cell lymphoma, Kappa-positive lymphoplasmacytic lymphoma and angiocentric small and large mixed cell lymphoma, plasmacytoma, and Burkitt lymphoma (Wang et al. 2017). The diagnosis of lymphoma is supported by immunohistochemical and/or flow cytometric studies including specific markers for lymphocytes, such as CD45, and T and B cellspecific markers, as well as specific molecular studies for gene rearrangements to identify the neoplastic cell population as lymphoma. The differential diagnosis includes inflammatory conditions and dermatoses as well as lymphoepithelioma-like carcinoma and other small blue cell tumors that can usually be distinguished by immunohistochemical studies (Wang et al. 2017). Dermatoses and benign inflammatory processes, unlike lymphomas, contain mixed populations of lymphocytes and other inflammatory cells. Lymphoepithelioma-like carcinomas contain epithelial cells that express high molecular weight cytokeratins and lack evidence of a

116

E. J. Wilkinson and D. S. Rush

monoclonal lymphocytic population (Carr et al. 1992). Large cell lymphomas, including Ki-1 lymphomas, may mimic poorly differentiated carcinoma. Immunohistochemical studies, including CD45 (leukocyte common antigen, LCA), and specific lymphocyte markers, as well as epithelial markers are of value to distinguish these tumors (Nucci et al. 2014). Appropriate aggressive chemotherapy is the treatment of choice for most lymphomas, although radiation therapy is employed in some cases (Wang et al. 2017). Prognosis is dependent on the lymphoma type, stage of the tumor, and response to therapy.

High-Grade Neuroendocrine Carcinoma In the vulva, the high-grade tumors that exhibit neuroendocrine differentiation include small cell neuroendocrine carcinoma, large cell neuroendocrine carcinoma, and Merkel cell tumors. Most vulvar SCNEC are Merkel cell tumors, with few exceptions (Crum et al. 2014b; Gardner et al. 2011).

Fig. 49 Merkel cell tumor. The tumor is diffusely and deeply infiltrated within the dermis and involves and abuts the overlying epithelium

Merkel Cell Tumor Merkel cell tumor of the vulva is a rare and aggressive tumor. In a review of 17 cases, the mean age was 59.6 years (Nguyen et al. 2017). These tumors typically present as an intradermal nodule or nodules and may be painful, with erythema of the overlying skin, and may be associated with HSIL or squamous cell carcinoma as well as both squamous and glandular differentiation (Gil-Moreno et al. 1997: Nguyen et al. 2017, Scurry et al. 1996). Three distinctive histopathologic types of Merkel cell tumor are recognized, namely, the trabecular or Fig. 50 Merkel cell tumor. The invasive tumor within the carcinoid-like type, the intermediate cell type, dermis includes ill-defined nests of tumor cells as well as and the small cell or oat cell-like type (Figs. 49, individual tumor cells 50 and 51). The distribution of these types among vulvar tumors has not been determined. small, uniform, hyperchromatic cells usually The histopathologic features are that of a poorly without prominent nucleoli. The tumor may differentiated neoplasm within the dermis and have pagetoid growth. Immunohistochemistry a distinctive perinuclear composed of a diffuse population of relatively demonstrates

2

Precursor Lesions and Malignant Tumors of the Vulva

117

mean of 7.8 months following surgery, a 47% mortality (Nguyen et al. 2017). For local disease, therapy usually includes wide local excision (partial deep vulvectomy) with a 2 cm margin of excision with sentinel lymph node biopsy and usually adjuvant radiotherapy (Nguyen et al. 2017). If the sentinel node demonstrates metastatic tumor, regional lymphadenectomy and postoperative radiation therapy to the primary and regional sites is usually indicated. With systemic disease chemotherapy is used when applicable. Fig. 51 Merkel cell tumor. The tumor cells are diffusely infiltrated within the dermis. The cells are relatively small with little cytoplasm and have nuclei with hyperchromatic chromatin

cytoplasmic dot with low molecular weight cytokeratin-specific antibodies such as AE1/AE3, Cam 5.2, and CK 20. The tumors are usually immunoreactive for NSE (Gil-Moreno et al. 1997). Chromogranin may be negative. These tumors usually contain membrane-bound neurosecretory granules by electron microscopy and may have adrenocorticotropic hormone (ACTH) secretion. The differential diagnosis includes other small cell tumors, of either primary or metastatic origin, including but not limited to neuroendocrine tumors, primitive neuroectodermal tumor (PNET), basal cell carcinoma, and basaloid squamous cell carcinoma (Wilkinson et al. 2014). When pagetoid spread is present, Paget disease, melanoma, and lymphoma are also considerations. Immunohistochemical findings will not distinguish Merkel cell tumor from a metastatic tumor of similar reactivity and morphology, but clinical history usually will. PNET expresses CD99 and lymphomas usually express CD45; neither of these are expressed in Merkel cell tumor. Merkel cell tumors are clinically very aggressive, with regional node metastasis and subsequent widespread metastasis often occurring within a year of diagnosis (Gil-Moreno et al. 1997; Scurry et al. 1996). In a review of 17 cases, 70.6% had recurrence with a mean of 6.3 months and within a

Metastatic Tumors Metastatic tumors comprise approximately 5–10% of all tumors of the vulva (Neto et al. 2003). The metastasis is usually a subcutaneous or dermal mass or may be ulcerated and is identified on average 3 years after treatment. Late metastasis can occur years post-therapy, as recognized with breast carcinoma where recurrence 20 years or more post-therapy can occur (Alligood-Percoco et al. 2015; Neto et al. 2003; Nucci et al. 2014). Tumors from other sites of the genital tract are the most common tumors that metastasize to the vulva, with squamous carcinoma of the cervix being the most frequent, followed by carcinomas of the endometrium and ovary. Other common primary tumor sites that may metastasize to the vulva include the bladder and urethra. Vulvar involvement can result from direct extension of tumors arising in the vagina, urethra, bladder, or rectum. Additional primary tumors that have metastasized to the vulva include, but are not limited to, malignancies of breast, colon, appendix, kidney, lung, stomach, gestational choriocarcinoma, melanoma, and neuroblastoma (Cheung and Cheung 2014; AlligoodPercoco et al. 2015; Neto et al. 2003; Rocconi et al. 2004; Ren et al. 2015). Low-grade endometrial stromal sarcoma can be metastatic to the clitoris (Androulaki et al. 2007). Malignant lymphomas, and rarely Hodgkin’s disease, also may metastasize to the vulva and Bartholin gland (Heller and Bean 2014; Wang et al. 2017). Acute myeloid leukemia can present as a mass of the

118

labium majus as metastatic myeloid sarcoma on evaluation (Erşahin et al. 2007). Metastatic tumors typically involve the dermis and overlying epithelium and consequently are often associated with ulceration. Patients with metastatic carcinoma to the vulva have a poor prognosis. Treatment is primarily palliative and radical surgery is not applicable.

Tumors of the Urethra Urethral Carcinoma Urethral carcinoma constitutes less than 1% of malignancies affecting the female genitalia and has an estimated annual incidence of 0.7 per million women. It occurs almost exclusively in elderly women, with a peak incidence of 80–84 years of age, but has been reported in women as young as 15 years old (Derksen et al. 2013). Urethral bleeding, urinary frequency, and dysuria are the most frequent presenting complaints. Tumors in the distal urethra, where most of these tumors arise, usually give rise to symptoms earlier in their course (Table 8). Most of these tumors arise in the distal urethra, and in a series of 91 women with urethral carcinoma, 45% were urothelial carcinomas and 29% were adenocarcinomas. Nineteen percent were squamous cell carcinomas and 6% were undifferentiated, or unknown, carcinoma types (Derksen et al. 2013). Squamous cell carcinomas and transitional cell carcinomas may be papillary in growth, forming papillomas or papillary carcinomas, or non-papillary, presenting as carcinoma in situ of urothelial or squamous type or as solid high-grade urothelial carcinomas or squamous cell carcinomas. Urothelial (transitional cell) carcinomas occur in the distal as well as proximal urethra and have been described arising within a urethral diverticulum (Murphy et al. 2004; Amin and Young 1997). The pagetoid variant of urothelial carcinoma in situ primarily involves the bladder but also can involve the urethra (Orozco et al. 1993). Radial growth of this intraepithelial neoplasm can involve the vulva and is referred to as PUIN. It can involve

E. J. Wilkinson and D. S. Rush

the urethral meatus, the vulvar vestibule, and contiguous vulvar mucosa and skin. PUIN clinically resembles vulvar cutaneous Paget disease (Wilkinson and Brown 2002). Primary urothelial carcinomas express CK 7 and usually CK 20. In addition, uroplakin III was identified in more than one-half of primary urinary tract urothelial carcinoma cases studied and in approximately two-thirds of the metastasis from urinary tract urothelial carcinomas (Brown and Wilkinson 2002). Uroplakin II is somewhat more sensitive than uroplakin III in the evaluation of PUIN lesions (Newsom et al. 2015). Adenocarcinomas of the urethra occur in the proximal urethra as well as within urethral diverticuli (Amin and Young 1997). Histopathologic types of adenocarcinoma include columnar/ mucinous, clear cell, and colloid (Murphy et al. 2004; Young and Scully 1985). Of these, clear cell adenocarcinoma of the urethra is of special interest in that this tumor occurs in adults with a wide age range, has distinct immunohistochemical and morphologic features suggesting Müllerian differentiation, and appears to have a generally better prognosis, even with more advanced stages (Drew et al. 1996; Oliva and Young 1996). These tumors, unlike Skene gland adenocarcinomas, do not express PSA or prostate acid phosphatase and have tubulocystic, papillary, and diffuse patterns of growth similar to clear cell carcinomas of the female genital tract (Drew et al. 1996). In one series, two-thirds of the cases arose in a urethral diverticulum (Oliva and Young 1996). The differential diagnosis of these clear cell tumors includes metastatic tumor, mesonephric carcinoma, and nephrogenic metaplasia (Murphy et al. 2004). Both urethral squamous cell carcinomas and adenocarcinomas are usually immunoreactive for CEA. The differential diagnosis of urethral clear cell adenocarcinoma includes metastatic clear cell carcinoma from the female genital tract, which is morphologically indistinguishable from urethral clear cell carcinoma. Endometriosis and nephrogenic adenoma can involve the urethra and may present as tumor, but lack the morphologic and nuclear features of clear cell carcinoma and are not associated with urethral clear cell carcinoma.

2

Precursor Lesions and Malignant Tumors of the Vulva

Primary mucinous adenocarcinoma of the urethra is rare. In a series of five cases, all presented as a urethral polypoid or papillary mass. The mean age of the women was 67 years old. Metastatic mucinous carcinoma was the main differential and excluded in all cases. Three of the five cases were stage pT4 at presentation. CDX2, CK20, and CK7 were positive or focally positive in four or the five cases (Harari et al. 2016). The prognosis related to urethral carcinoma is relatively poor. The 5-year survival is influenced by stage of the tumor and tumor type. The staging system for tumors of the urethra of the AJCC is summarized in Tables 10, 11, and 12 (American Joint Committee on Cancer 2017; McKenney et al. 2017). Of 91 women with clear cell adenocarcinoma, survival by stage

119

was 67% for stage 0–II, 53% for stage III, and 17% for stage IV. Overall 5-year survival was approximately 60% for the women with urethral urothelial or squamous cell carcinomas and 17% for the adenocarcinomas (Derksen et al. 2013). In a study of 13 women with clear cell adenocarcinoma of the urethra, 6 were alive and well 5 years or more after surgery, and 7 had recurrence of which 4 died of disease from 5 to 42 months following surgery. Three of the seven were alive with disease over 5 years after surgery (Oliva and Young 1996). For non-clear cell tumors, overall, survival is better with tumors in the distal urethra, and tumors in the proximal portions of the urethra have a poorer prognosis. Those patients with tumors involving the entire urethra have the poorest prognosis (Murphy et al.

Table 10 AJCC staging of urethral carcinoma primary tumor (T), urethra (2017) (Hansel et al. 2017) Definition of primary tumor (T) Male penile urethra and female urethra T category TX T0 Ta Tis T1 T2 T3 T4

T criteria Primary tumor cannot be assessed No evidence of primary tumor Noninvasive papillary carcinoma Carcinoma in situ Tumor invades subepithelial connective tissue Tumor invades any of the following: corpus spongiosum, periurethral muscle Tumor invades any of the following: corpus cavernosum, anterior vagina Tumor invades other adjacent organs (e.g., invasion of the bladder wall)

From AJCC Cancer Staging Manual, 8th ed. New York: Springer Science + Business (2017) (Hansel et al. 2017, pp. 771–772). Used with permission of the American Joint Committee on Cancer (AJCC), Chicago, Illinois (Hansel et al. 2017, pp. 771–772). The original and primary source for this information is the AJCC Cancer Staging Manual, Eighth Edition (2017) published by Springer International Publishing

Table 11 AJCC staging of urethral carcinoma: regional lymph nodes (N) (2017) (Hansel et al. 2017) Definition of regional lymph node (N) N category N criteria NX Regional lymph nodes cannot be assessed NO No regional lymph node metastasis N1 Single regional lymph node metastasis in the inguinal region or true pelvis [perivesical, obturator, internal (hypogastric) and external iliac], or presacral lymph node N2 Multiple regional lymph node metastasis in the inguinal region or true pelvis [perivesical, obturator, internal (hypogastric) and external iliac], or presacral lymph node From AJCC Cancer Staging Manual, 8th ed. New York: Springer Science + Business (2017) (Hansel et al. 2017, pp. 771–772). Used with permission of the American Joint Committee on Cancer (AJCC), Chicago, Illinois (Hansel et al. 2017, pp. 771–772). The original and primary source for this information is the AJCC Cancer Staging Manual, Eighth Edition (2017) published by Springer International Publishing

120 Table 12 AJCC staging of urethral carcinoma primary tumor (M), urethra (2017) (Hansel et al. 2017) Definition of distant metastasis (M) M category M criteria M0 No distant metastasis M1 Distant metastasis From AJCC Cancer Staging Manual, 8th ed. New York: Springer Science + Business (2017) (Hansel et al. 2017, pp. 771–772). Used with permission of the American Joint Committee on Cancer (AJCC), Chicago, Illinois (Hansel et al. 2017, pp. 771–772). The original and primary source for this information is the AJCC Cancer Staging Manual, Eighth Edition (2017) published by Springer International Publishing

E. J. Wilkinson and D. S. Rush

these cases direct mucosal-related metastasis and/or lymphatic metastasis may occur. Bladder neck involvement by the primary bladder carcinoma is the most significant risk factor for urethral involvement (Chen et al. 1997). Metastatic tumors from vulvar, vaginal, cervical, or anal carcinomas, as well as endometrial and, rarely ovarian carcinomas, may also occur and may initially mimic a urethral caruncle (Hammadeh et al. 1996).

Gross Description, Processing, and Reporting of Vulvar Specimens 2004). Survival in urethral carcinoma is influenced by the fact that up to one-half of the women with urethral carcinoma have metastasis to superficial or deep pelvic nodes when first seen. Improved early detection and individualized surgical and radiotherapy techniques promise to substantially increase survival.

Other Malignant Tumors of the Urethra Non-Hodgkin’s lymphoma (Atalay et al. 1998; Zahrani et al. 2012), carcinosarcoma (Konno et al. 1997), and sarcomas have all been reported arising within the urethra. Urethral caruncles with atypical stromal cells, and a florid proliferation of reactive lymphoid cells, are the primary differential diagnoses regarding lymphomas and sarcomas in this location. Immunohistochemical and/or molecular studies distinguish these benign processes from lymphoma. The atypical stromal cells in the urethral caruncles are immunoreactive for vimentin in approximately two-thirds of the cases and express alpha-smooth muscle actin in one-half of cases (Young et al. 1996). A number of metastatic tumors may involve the urethra, either by direct mucosal growth or by lymphatic or vascular metastasis (Murphy et al. 2004). Metastatic involvement of the urethra from bladder carcinoma occurs in 8% to 16% of cases (Chen et al. 1997; Maralani et al. 1997). In

Vulvar Biopsies Vulvar biopsies may be diagnostic, where only a sampling of the lesion in question is made, or complete excisional, where the clinician attempts to excise the entire lesion. Diagnostic biopsies may be punch biopsies, such as performed with a Keyes punch biopsy, or shave biopsies, or partial excisional biopsies, where a scalpel is used to excise a representative section of the lesion in question. Punch biopsies are usually preferred to evaluate inflammatory processes, such as lichen sclerosus, whereas excisional biopsies are preferred if the lesion is considered neoplastic and issues regarding possible invasive tumor are made. Shave biopsies are also useful for neoplastic lesions, provided issues such as depth of a tumor are not an issue. In general, vulvar biopsies should be handled like skin biopsies, and in all cases effort should be made to keep the specimen well orientated so that right-angle sections of the epithelial-dermal interface are made, to clearly identify the epithelial-dermal, or mucosal-stromal interface. Several methods are in common use to assess both deep and lateral resection margins. Orientation is aided if the surgeon places a suture on one edge of the epithelium for orientation or uses India ink, or tattoo dye, to facilitate the recognition of the surface of the lesion. For lateral and deep

2

Precursor Lesions and Malignant Tumors of the Vulva

margins, using one or more colored inks or dyes can be used to define the margins. Specimens should be sectioned into approximately 2 mm slices, with the plane of section at right angles to the surface of the specimen and the entire specimen submitted for evaluation, unless it is very large in which case it is handled as any other partial vulvectomy. In most cases several slices can be placed into a single cassette. Small punch biopsies are submitted totally and, when necessary, bisected perpendicular to the epithelial surface. It is useful to request at the time of submission that the paraffin block(s) be cut at multiple levels, for example, cutting three sections from each block. This will usually give an excellent view of the lesion in question. This practice also shortens the turnaround time, since without it small biopsies frequently are sectioned inadequately the first time they are cut, often with only one section made. Remounting the block and cutting again to make additional slides take more time than obtaining multiple levels at the onset and may result in loss of valuable tissue.

Large Operative Specimens Wide Local Excision (Partial Deep Vulvectomy) and Superficial Vulvectomy Most specimens are highly variable in their composition, since these procedures are tailored to the extent of the lesion. Recently, wide local excision has been performed for stage Ia (30 years, pure AIS (e.g., no coexistent HSIL), and larger lesions (>8 mm) (Munro et al. 2017; Costa et al. 2012). Based on these studies, conservative management by cone biopsy alone is now considered to be an option in women with AIS desirous of maintaining their fertility, if the cone biopsy margins are negative.

Cervical Cytology Strengths and Limitations of Cervical Cytology Although it was introduced over a half century ago, cervical cytologic screening continues to be one of the most effective cancer prevention test available. Over the half century since it was introduced, so much epidemiological and modeling data have accumulated demonstrating the effectiveness of cytology that it has become the index by which all other cancer screening tests are compared. Cytologic screening performed only twice in a woman’s lifetime can reduce her risk for invasive cervical cancer by up to 43% and yearly screening is estimated to reduce a woman’s risk by over 90% (Parkin 1991; Goldie et al. 2004). However, despite the effectiveness of cytologic screening, it is important to remember that no screening,

288

diagnostic, or therapeutic technique used in medicine is perfect, and cervical cytology is no exception. Some women will develop invasive cervical cancer, despite routine cytologic screening. Over the last decades numerous advances have been made in cervical cytology collection techniques, how cytological preparations are evaluated, and the classification systems used for reporting cytologic diagnosis. One of the most important advances has been the introduction of liquid-based cytology. With liquid-based cytology, the cells collected from the cervix are transferred directly to a liquid fixation solution that is shipped to the cytology laboratory where the slide is prepared. One of the primary advantages of liquid-based cytology is that molecular testing for sexually transmitted infections such as HPV DNA, chlamydia, N. gonorrhea can be performed directly from liquid-based specimens. HPV testing is particularly useful when a diagnosis of atypical squamous cells-undetermined significance (ASC-US) is made (i.e., “reflex” HPV DNA testing).

The Bethesda System (TBS) Terminology In 1988, TBS for reporting cervical/vaginal cytological diagnoses was developed to provide uniform guidelines for reviewing and reporting gynecological Papanicolaou tests (The Bethesda 1988). TBS classification was subsequently modified in 1991 and 2001 (Workshop 1991; Solomon et al. 2002). In 2014 in response to additional experience with liquid-based cytology, additional insights into the biology of HPV, the widespread introduction of HPV vaccination, and adoption of HPV testing either alone or in combination with cytology the 2001 Bethesda System was reviewed and updated (Nayar and Wilbur 2015). TBS is now the standard classification for cervical cytology used in the USA. There are multiple distinct parts to the report, but a statement of specimen adequacy and the Interpretation/Results are the key ones, Table 10. Other parts of the report can provide additional information. This is designed

T. C. Wright et al.

to assist clinicians by answering three basic questions: (1) was the sample adequate? (2) was the cytology test normal? and (3) if the test was not completely normal, what specifically was wrong?

Cytological Appearance of Cervical Cancer Precursors TBS classification categorizes precursors to cervical cancer as “epithelial cell abnormalities.” The category “epithelial cell abnormalities” is subdivided into abnormalities of squamous cells and abnormalities involving glandular cells, either endocervical or endometrial. Cytological changes previously classified as mild squamous cytological atypia and atypical endocervical cells are also included in this category. Squamous Cell Abnormalities Atypical Squamous Cells (ASC)

The ASC category is used to designate cytological changes suggestive of SIL that are quantitatively or qualitatively insufficient for a definitive diagnosis of SIL (Workshop 1991; Solomon et al. 2002; Nayar and Wilbur 2015). There are several points that need to be made with respect to ASC. First, a diagnosis of ASC is one of exclusion; the cells are abnormal, but they do not warrant a diagnosis of SIL. Second, a diagnosis of ASC should not be used when the underlying process is inflammatory or reactive, such slides should be carefully reviewed and classified as “negative for intraepithelial lesion or malignancy” whenever possible rather than ASC. Third, although the ASC category is sometimes disparagingly referred to as a “cytological wastebasket,” there are specific criteria that should be used for making this diagnosis. If these criteria are adhered to, the median rate of ASC in US laboratories is approximately 5% of all cytology specimens and the ASC rate should be no more than twice the SIL rate (Eversole et al. 2010). The 2014 Bethesda System subdivides the ASC category into two subdivisions: ASC-US refers to samples in which the

5

Precancerous Lesions of the Cervix

289

Table 10 TBS 2014 Classification (Nayar and Wilbur 2015) Specimen type: Indicate conventional smear (Pap smear) versus liquid-based preparation versus other Specimen adequacy: Satisfactory for evaluation Unsatisfactory for evaluation (specify reason) General categorization (optional) Negative for intraepithelial lesion or malignancy Other: See interpretation/result Epithelial cell abnormality: See interpretation/result (specify “squamous” or “glandular” as appropriate) Interpretation/result Negative for intraepithelial lesion or malignancy Nonneoplastic findings (optional to report) Nonneoplastic cellular variations Reactive cellular changes Glandular cells status posthysterectomy Organisms Other Endometrial cells (in a women > 45 yrs of age) Epithelial cell abnormalities Squamous cell ASC – Of undetermined significance (ASC-US) – Cannot exclude HSIL (ASC-H) LSIL (encompassing: HPV/milddysplasia/CIN1) HSIL (encompassing: moderate and severe dysplasia, CIS: CIN 2 and CIN 3) Squamous cell carcinoma Glandular cell Atypical – Endocervical cells (NOS or specify in comments) – Endometrial cells (NOS or specify in comments) – Glandular cells (NOS or specify in comments) Atypical – Endocervical cells, favor neoplastic – Glandular cells, favor neoplastic Endocervical AIS Adenocarcinoma – Endocervical – Endometrial – Extrauterine – Not otherwise specified (NOS) Other malignant neoplasms: (specify) Adjunctive testing: Provide a brief description of the test method(s) and report the result so that it is easily understood by the clinician Computer-assisted interpretation of cervical cytology If case examined by an automated device, specify device, and result Educational notes and comments appended to cytology reports (optional)

290

T. C. Wright et al.

Table 11 Criteria used to diagnosis ASC ASC-US Cells resemble superficial or intermediate squamous cells in size and configuration Nuclei are 2.5–3 times the size of a normal intermediate cell nuclei or 2 times the size of a metaplastic cell nucleus Nuclei are round to oval with minimal irregularities Nuclei are normochromatic to slightly hyperchromatic ASC-H Small cells with high N/C ratios Cells resemble parabasal or basal cells in size and configuration, but the nucleus is 1.5–2.5 times larger resulting in a high N/C ratio Cells occur singularly or in small groups Nuclei often have uneven chromatin and are hyperchromatic Nuclear contour is often irregular Crowded sheet pattern Crowded clusters of cells resembling parabasal or basal cells with atypical nuclear features including hyperchromasia and high N/C ratios Cells have a loss of polarity and can be difficult to visualize

cytological changes are suggestive of LSIL but lack sufficient cytological abnormalities to allow a definitive diagnosis, and Atypical Squamous Cells – Cannot Exclude an HSIL (ASC-H) refers to samples in which the cytological changes are suggestive of HSIL, but the cytological abnormalities are insufficient to allow a definitive interpretation (Sherman et al. 1999). The specific criteria used to diagnose ASC are given in Table 11. One of the major criteria used to distinguish ASC-US from benign cellular changes is nuclear size. In ASC-US, the nuclei are typically 2.5–3 times the size of a normal intermediate cell or twice the size of a squamous metaplastic cell nucleus, Fig. 69. In addition there can be a slightly increased N/C ratio. Other features that are found in ASC-US are minimal nuclear hyperchromasia and irregularity in chromatin distribution or nuclear shape (Nayar and Wilbur 2015). ASC-US can include “atypical parakeratosis” which has cells with nuclear atypia associated with dense orangeophilic cytoplasm. A diagnosis of ASC-US is sometimes made when there are cells with some, but not all of the criteria necessary for a diagnosis of LSIL, Fig. 70. This typically occurs when there are cytoplasmic changes that suggest HPV effect (nuclear halos), but the cells have minimal nuclear changes. It should be emphasized that the degree of nuclear changes considered sufficient to warrant a diagnosis of ASC-US is highly subjective and

Fig. 69 ASC-US. Intermediate squamous epithelial cells demonstrate nuclear enlargement and hyperchromasia. No organisms or inflammatory changes were identified on the smear

varies between cytologists. This introduces a degree of uncertainty with respect to a diagnosis of ASC-US, and studies have shown that a diagnosis of ASC-US is the least reproducible of all cytological diagnoses (Confortini et al. 2003, 2007; Gatscha et al. 2001). Approximately 3–5% of women with a diagnosis of ASC-US will have histologic HSIL (CIN 3) when colposcopy is performed (Stoler et al. 2013; Stoler et al. 2011; Tewari et al. 2017).

5

Precancerous Lesions of the Cervix

Fig. 70 ASC-US. These cells are suggestive but not diagnostic of a LSIL since they have a considerable degree of nuclear enlargement and perinuclear halos. However, the findings are not sufficient to allow a diagnosis LSIL and there only a limited number of such cells were present so a diagnosis of ASC-US was rendered

The second category of ASC is ASC-H in which the cells resemble parabasal or basal cells in size and configuration. ASC-H appears as two different types of patterns. One is small cells with high N/C ratios. These cells occur singly or in small groups. The size of the cells is similar to that of metaplastic cells, but the nuclei are 1.5–2.5 times larger than normal which results in a high N/C ratio. The cells frequently are hyperchromatic with irregular nuclear contours and uneven chromatin, Fig. 71. The differential diagnosis in such cases is between atypical immature squamous metaplasia and HSIL. The other pattern of ASC-H is the “crowded sheet pattern” (Nayar and Wilbur 2015). This pattern shows a crowded cluster of squamous cells with atypical nuclear features including hyperchromasia and high N/C ratios, Fig. 72. The cells have a loss of polarity and can be difficult to visualize. These crowded clusters of cells can represent HSIL that has grown into endocervical glands or crypts, reactive or neoplastic endocervical cells, or atrophy. ASC-H is an uncommon finding and typically accounts for less than 10% of all ASC diagnoses.

291

Fig. 71 ASC-H. A cluster of atypical immature metaplastic type cells is present. The cells have an increased N/C ratio, hyperchromasia, and slightly irregular nuclei. However, the cells have more cytoplasm than usually associated with HSIL and only a few clusters were present

The median reporting rate of ASC-H in the US laboratories in 2006 was 0.3% according to the CAP survey (Eversole et al. 2010). The majority of women with ASC-H are high-risk HPV DNA positive, and histologic HSIL is identified at the time of colposcopy in 12–68% of women with ASC-H (Sherman et al. 1999; Bandyopadhyay et al. 2008; Liman et al. 2005). Because of the high prevalence of histologic HSIL in women with ASC-H, it has been suggested that ASC-H would be more appropriately referred to as “equivocal HSIL” (Wright et al. 2007). LSIL

The LSIL category in TBS includes both HPV effects and mild dysplasia (CIN 1). The cells of LSIL are of the superficial or intermediate cell type and are found either as individual cells or as sheets of cells with well-defined cell borders. The cells are typically enlarged with abundant cytoplasm. Nuclei are usually enlarged to >3 times the size of a normal intermediate cell nucleus and in some instances can be quite large, Fig. 73. There are usually prominent perinuclear halos, and the nuclei are usually

292

T. C. Wright et al.

Fig. 73 LSIL. The cells are of the intermediate type with nuclear enlargement and prominent koilocytosis. One of the nuclei is more than 10 times the size of a normal intermediate cell

Fig. 72 ASC-H. This crowded sheet of cells is hyperchromatic and appears to have high N/C ratios. However, it is difficult to see the individual cells so one does not want to diagnose it as HSIL. Instead a diagnosis of ASC-H should be given

hyperchromatic and multinucleation is common, Fig. 74. The chromatin is finely granular and uniformly distributed. With cytology methods such as liquid-based cytology and computerized imaging systems, the rate of LSIL appears to be increasing in the US. In surveys taken in the 1990s the median reporting rate of LSIL in US laboratories was 1.6%, but by 2006 this has increased to 3% (Jones and Davey 2000; Eversole et al. 2010). A recent report of the impact of implementing a computerized cytology imaging system in a tertiary military center reported that after implementation the rate of LSIL increased from 2.6% to 3.9% (Duby and DiFurio 2009). Age is an important factor in determining the prevalence of LSIL. In a large US screening trial, the prevalence of LSIL decreased from 6.5% in women 21–24 years of age to 3.8% in women

Fig. 74 LSIL. Cells from this lesion demonstrate considerable koilocytosis with multinucleation and prominent halos

25–29 years and then to 1.4% in women 40–49 years (Wright et al. 2012). HSIL

Because TBS combines moderate and severe dysplasia together with carcinoma in situ in the HSIL category, there is a wide variation in the cytological appearance of HSIL. As the severity of the lesion increases, the degree of differentiation and the amount of cytoplasm

5

Precancerous Lesions of the Cervix

293

Table 12 Criteria used to diagnosed squamous epithelial cell abnormalities Bethesda System CIN terminology Older terminology Cell type Cell arrangement Number abnormal cells Koilocytosis Nuclear size Hyperchromasia N/C ratio

LSIL CIN 1 Mild dysplasia Superficial or intermediate Singly or sheets

HSIL CIN 2 Mod. dysplasia Parabasal

CIN 3 Severe dysplasia Basal

Singly or sheets

Singly or sheets

+

++

+++

CIS Basal, spindle, pleomorphic Singly or sheets or syncitia ++++

+++ +++ + +

+ ++ ++ ++

+/ + +++ +++

+/ + ++++ ++++

decreases, the N/C ratio increases, and the degree of nuclear atypia increases, Table 12. Although cellular size varies considerably, the cells of HSIL are typically smaller and have less cytoplasm than do the cells of LSIL. Some HSIL are quite small and are the size of basal cells, The degree of nuclear enlargement also varies considerably, Fig. 75. In some cases, the nucleus is as large as that of the typical LSIL, but since there is less cytoplasm the N/C ratio is higher than in LSIL. In cases in which the overall size of the HSIL cells is small, the nuclei may not much larger than that of a typical intermediate cell, but the N/C ratio is quite high, Fig. 76. The nucleus is typically quite hyperchromatic and the contour of the nucleus is usually irregular and indentations are often present. The chromatin can be fine or coarsely granular and is evenly distributed. Nucleoli are generally not present. The cytoplasm of HSIL also varies considerably. In some cases, it can be thin or lacey, in other cases it is densely metaplastic, and in others it can be densely keratinized, Fig. 77. The cells can occur singly, in sheets and clusters, or as syncytial aggregates. The number of abnormal cells that are present can vary dramatically case to case. When only a few small HSIL cells of the basal cell type are present, it can be quite challenging to correctly classify the case as a HSIL, and these cases account for a disproportionate percentage of false negative cervical cytology. According to the CAP 2006 survey, the median reporting rate

Fig. 75 HSIL. There is considerable variability in nuclear size. Many of the cells have a considerable amount of cytoplasm, but the N/C ratio is higher than usually seen in LSIL

of HSIL in US laboratories was 0.6% (Eversole et al. 2010). The rate of HSIL varies with age. In a large US screening trial, the prevalence of HSIL decreased from 0.7% in women 21–24 years of age to 0.4% in women 25–29 years and then to 0.2% in women 40–49 years (Wright et al. 2012). A diagnosis of HSIL connotes a high risk for significant cervical disease. Histologic HSIL is found in approximately 60% of women with HSIL (Massad et al. 2013b).

294

Fig. 76 HSIL. These cells are the size of parabasal cells, but the nucleus is the same size as an intermediate cell. Therefore, the N/C ratio is greatly increased

Fig. 77 HSIL. Many of the HSIL cells are keratinized and have quite hyperchromatic nuclei. It is often difficult to distinguish between HSIL of this type and keratinizing invasive squamous cell carcinoma

Invasive Squamous Cell Carcinoma

Squamous cell carcinomas of the cervix are subdivided into keratinizing and non-keratinizing types. Nonkeratinizing carcinomas typically

T. C. Wright et al.

Fig. 78 Invasive squamous cell carcinoma. The cells of this nonkeratinizing squamous cell carcinoma are polygonal and arranged in syncytial sheets. They are highly atypical, but smaller than the cells of many intraepithelial lesions

have large numbers of malignant cells that form loose cell sheets and syncytial arrangements. The cells are usually somewhat smaller than HSIL but have most of the features of HSIL. The nuclei have coarsely clumped chromatin and focal chromatin clearing and prominent macronucleoli may be present, Fig. 78. Squamous cell carcinomas often have a “dirty” background containing blood, cellular debris, fibrin, and necrotic material. This is often referred to as a tumor diathesis. This characteristic background is less prominent in liquid-based cytology specimens. However, in liquid-based cytology, there is often a distinctive necrotic background that is easy to recognize since it surrounds the cellular material in a “clumped” appearance and large, necrotic tissue fragments are sometimes present, Fig. 79. Cytology specimens from women with keratinizing carcinomas contain malignant cells demonstrating a variety of cell shapes and sizes, Fig. 80. Some of the cells are pleomorphic or tadpole-shaped. These cells have abundant orangophilic cytoplasm. There is frequently abundant hyperkeratosis and parakeratosis. The nuclei

5

Precancerous Lesions of the Cervix

295

endocervical and endometrial cells, are combined together in a single entity referred to as glandular cell abnormalities. Benign appearing endometrial cells occurring in postmenopausal women are classified as “other.” Glandular cell abnormalities are divided into three categories: atypical glandular cells, either unqualified or favor neoplastic; AIS; and invasive adenocarcinoma.

Atypical Glandular Cells (AGC)

Fig. 79 Tumor diathesis. This liquid-based cytology preparation from a woman with squamous cell carcinoma shows necrotic material surrounding malignant cells giving it a “clumped” appearance

Fig. 80 Invasive squamous cell carcinoma. The cells of this keratinizing squamous cell carcinoma are quite pleomorphic and include spindle-shaped, elongate, and caudate forms. The nuclei of some cells are extremely hyperchromatic

are irregular in shape and quite hyperchromatic. Sometimes the nuclei are degenerated, appearing as opaque masses or “ink blots.” Unlike nonkeratinizing squamous cell carcinoma, keratinizing squamous cell carcinomas often do not have “dirty” background or evidence of tumor diathesis. Glandular Cell Abnormalities In the 2014 Bethesda System, all types of glandular cell abnormalities, including both atypical

All AGC lacking the diagnostic features of adenocarcinoma, irrespective of whether they are of endometrial or endocervical origin, are classified by the 2014 Bethesda System as AGC with a specification as to whether they are endocervical, endometrial, or of uncertain origin. There are two categories of AGC. The first is AGC (either endocervical, endometrial, or unclassified) that are not qualified and the second is atypical glandular cells; favor neoplastic. Glandular cytological abnormalities are considerably less common than squamous abnormalities, and most cytologists tend to be less comfortable recognizing and diagnosing them. In addition, the criteria used to differentiate reactive endocervical changes, endocervical dysplasia, endocervical AIS and invasive endocervical adenocarcinoma are less well established than those used for squamous lesions. Cytologists even have difficulty in differentiating atypical endocervical cells from cases of HSIL that have extended into endocervical crypts. This accounts for the high prevalence of squamous abnormalities (approximately 30%) detected in women referred for AGC to colposcopy (Kim et al. 1999; Ronnett et al. 1999). The cytological features of atypical endocervical cells vary depending on the degree of the underlying histopathologic abnormality. Cases of the type designated by cytopathologists as atypical endocervical cells – not otherwise specified (NOS) – have variability in nuclear size and shape, Fig. 81. These cells occur in sheets or strips with some cell crowding and nuclear overlap. Nuclei are typically enlarged compared

296

Fig. 81 Atypical glandular cells – endocervical type (AGC-EC). These endocervical cells have enlarged nuclei and prominent nucleoli and vary somewhat in size and shape. The smear was obtained 6 weeks postpartum and follow-up examination was completely negative

to normal endocervical cells with up to 3–5 times greater nuclear area. There is mild hyperchromasia and mild chromatin irregularity. Mitoses are rare and there are only occasional nucleoli. Atypical endocervical cells favor neoplasia includes those cases where the cytological features are suggestive of AIS, but are insufficient to allow a definitive diagnosis. These cases typically have more nuclear hyperchromasia, variability in nuclear size, and granularity of the chromatin than is observed in cases of atypical endocervical cells, NOS, Fig. 82. When the cells occur in strips, they often are pseudostratified.

T. C. Wright et al.

Fig. 82 AGC – favor neoplasia. These endocervical cells are somewhat suggestive of AIS. The nuclei are hyperchromatic and the chromatin is coarsely clumped. There is variation in nuclear size and the cells form a three-dimensional aggregate

AIS

In cases of AIS, there are usually a larger number of AGC that form crowded cellular clusters. The sheets are usually three-dimensional and sometimes retain the architecture of the underlying glands, Fig. 83. The cells within these sheets occasionally form rosettes and have extensive feathering of the cells at the periphery. Individual endocervical cells are highly atypical with enlarged round, oval, or elongated nuclei that vary in size from cell to cell. In most cases, the chromatin is coarsely clumped and multiple mitoses are seen, Fig. 84. Sometimes it is difficult for the cytologist to determine whether the atypical

Fig. 83 AIS These endocervical cells form a tight threedimensional structure that is similar to the outline of an endocervical gland. A large number of these formations are often present in specimens from AIS

cells represent AGC or HSIL cells that have extended into an endocervical crypt or “gland.” In these cases, highly atypical nuclei are identified in the center of a cell aggregate and some of the cells at the periphery of the aggregate appear to be endocervical cells.

5

Precancerous Lesions of the Cervix

Fig. 84 AIS. The individual atypical endocervical cells are hyperchromatic with coarsely clumped chromatin. They show the characteristic feathering of the nuclei at the edge of the cluster which is typical of AIS

Adenocarcinoma

TBS subclassifies invasive adenocarcinomas into “adenocarcinoma-endocervical type,” “adenocarcinoma-endometrial type,” and “adenocarcinoma-not otherwise specified”. The cytological diagnosis of invasive adenocarcinoma is relatively straightforward. Adenocarcinoma cells from either an endocervical or an endometrial primary have enlarged nuclei, high N/C ratios, coarsely clumped chromatin, and prominent nucleoli, Fig. 85. They can occur singly or in clusters. Cytologists should try to distinguish between endometrial and endocervical primary adenocarcinomas whenever possible. Key features that allow discrimination between endometrial and endocervical origin in cytology include number of abnormal cells, size of the cells, retention of columnar configuration, appearance of cytoplasm, and nuclear structure (Ng 1993). Typically, adenocarcinoma of the cervix shows considerably larger numbers of cells than does endometrial adenocarcinoma. The cells of endometrial adenocarcinoma typically occur singly or as small clusters, whereas the cells of endocervical adenocarcinoma occur as larger

297

Fig. 85 Endocervical adenocarcinoma. These endocervical cells have the features of frank adenocarcinoma. The nuclei are quite enlarged, the chromatin is coarsely clumped and marginated, and there are prominent nucleoli. The background shows inflammation and necrosis indicating tumor diathesis is present

2 dimensional sheets, 3 dimensional clusters, or syncytial sheets. Cells derived from endocervical adenocarcinoma typically retain a columnar configuration which is lost in most endometrial carcinomas. The cytoplasm of cells exfoliated from endocervical adenocarcinoma is typically finely vacuolated, whereas the cytoplasm from cells of endometrial adenocarcinoma is typically scant and cyanophilic. The nuclei of cells of endometrial adenocarcinoma vary in size becoming larger with higher grade tumors. The chromatin is less granular and the nuclei are less hyperchromatic than the nuclei of cells of endocervical adenocarcinoma. They also less frequently have multiple nucleoli.

Management of Cytologic Abnormalities and Cervical Cancer Precursors In 2012 the ASCCP sponsored a consensus workshop to update prior Consensus Guidelines for the Management of Women with Cytological Abnormalities and Cervical Cancer Precursors (Wright et al. 2007; Massad et al. 2013a). These guidelines are widely used in the US and are evidence-

298

based with each recommendation accompanied by a grading of both the strength of the recommendation and the strength of the data supporting the recommendation. What follows is a brief synopsis of the guidelines. The complete recommendations and management algorithms are available at www.asccp.org.

ASC The prevalence of biopsy-confirmed HSIL among women undergoing colposcopy for an ASC cytology varies from 5% to 17% (Wright et al. 2007; Stoler et al. 2011, 2013). The prevalence of HSIL in women with ASC depends on a number of factors including the patient’s age, history, and the subclassification of the ASC result. Overall, it appears that approximately half of women with histologic HSIL have ASC as their initial abnormal cervical cytology result (Lonky et al. 1999; Kinney et al. 1998). However, it should be noted that the risk that a woman with ASC has invasive cervical cancer is quite low (about one per thousand).

T. C. Wright et al.

ASC-US In Kaiser Northern California, the 5 year cumulative risk for HSIL in women 30–64 years of age with ASC-US is 6.9% (Katki et al. 2013a). Two methods are considered acceptable for managing women in the general population with ASC-US: high-risk HPV DNA testing and repeating the cervical cytology at 1 year, Fig. 86 (Massad et al. 2013a). HPV DNA testing identifies more cases of HSIL than does a single repeat cervical cytology, but refers approximately equivalent numbers of women for colposcopy (Arbyn et al. 2006). Moreover, cost-effectiveness modeling has demonstrated that HPV DNA testing for women with ASC-US is highly attractive when the initial ASC-US cytology was obtained from a liquidbased sample (Kulasingam et al. 2006; Pedersen et al. 2016). Thus, high-risk HPV DNA testing is the preferred approach to managing women with ASC-US whenever liquid-based cytology is used for screening (Massad et al.

Fig. 86 ASCCP Consensus Conference algorithm for managing women with ASC-US

5

Precancerous Lesions of the Cervix

2013a). Women found to be high-risk HPV DNA positive should be referred to colposcopy, whereas HPV DNA negative women should undergo repeat cotesting in 3 years. Since the prevalence of HPV DNA positivity is much higher in young women with ASC-US than in older women, HPV DNA testing is not recommended for young women with ASC-US (Sherman et al. 2002; Boardman et al. 2005; Stoler et al. 2011, 2013). Instead, women 21–24 years of age with ASC-US are managed using annual repeat cytological examinations and only referred to colposcopy if the repeat Pap tests are diagnosed as ASC-H, HSIL, or AGC or are persistently abnormal for a period of 2 years, Fig. 87 (Massad et al. 2013a). Management options for pregnant patients with ASC-US are identical to those for nonpregnant patients with the exception that it is acceptable to defer the colposcopic examination until the patient is 6 weeks post-partum.

299

ASC-H ASC-H is a much more concerning cytology result than ASC-US since biopsy-confirmed HSIL is identified in 13% to 66% of women with ASC-H (Xu et al. 2016). Thus for the purposes of management, ASC-H should be considered to be an equivocal HSIL result and all women with ASC-H should be referred for a colposcopic evaluation (Massad et al. 2013a). If after colposcopy the patient has histologic LSIL or less, follow-up utilizing either repeat cotesting at 12 and 24 months, a LEEP, or review of the cytological, histological, and colposcopic findings is acceptable.

LSIL In Kaiser Northern California, the 5 year cumulative risk for histologic HSIL in women 30–64 years of age with LSIL is 16% (Katki et al. 2013b). The risk of HSIL varies

Fig. 87 ASCCP Consensus Conference algorithm for managing young women with either ASC-US or LSIL

300

considerably depending on the patient’s age and HPV status. The 5 year cumulative risk of HSIL is 19% for HPV-positive LSIL and only 5.1% for HPV-negative LSIL. In women 30–34 years with LSIL the risk of histologic HSIL is 17%, whereas it drops to 7.3% in women 60–64 years. Therefore, the management of women with a cytological result of LSIL varies depending on HPV status and age. Women with LSIL who either are of unknown HPV status or HPV positive should be referred to colposcopy, Fig. 88 (Massad et al. 2013a). The preferred approach to managing HPV negative women with LSIL is to do a repeat cotest in 12 months. Since invasive cervical cancer is very uncommon in young women and prospective studies have shown that over 90% of LSIL will spontaneously clear, young women with LSIL should not be referred to colposcopy, but should be followed using yearly cytology tests for a period of 2 years, Fig. 87 (Moscicki et al. 2001, 2006; Massad et al. 2013a). Another “special population” is postmenopausal women with LSIL.

T. C. Wright et al.

Both the prevalence of HPV DNA and the prevalence of histologic HSIL are lower in postmenopausal women with LSIL than in women in the general population with LSIL. Therefore, postmenopausal women with LSIL and no HPV test can be managed using either HPV testing, repeat cytologic testing at 6 and 12 months, and colposcopy.

HSIL Histologic HSIL is identified in 53–97% of women with a cytological result of HSIL and invasive cervical cancer is found in approximately 2% (Wright et al. 2007). In Kaiser, the 5 year cumulative detection rate of histologic HSIL in HPV negative women with cytologic HSIL is 49%. This increases to 71% in HPV-positive women with cytologic HSIL. Therefore, women with a cytological result of HSIL irrespective of HPV status should be referred for either a colposcopic evaluation or an immediate loop electrosurgical excisional procedure, Fig. 89 (Massad et al. 2013a). If after colposcopy the patient has

Fig. 88 ASCCP Consensus Conference algorithm for managing women with LSIL

5

Precancerous Lesions of the Cervix

301

Fig. 89 ASCCP Consensus Conference algorithm for managing women with HSIL

histologic LSIL or less, followed-up utilizing either repeat cotesting at 12 and 24 months, a LEEP, or review of the cytological, histological, and colposcopic findings is acceptable. For women 21–24 years with HSIL, initial colposcopy is recommended. If histologic HSIL is not identified, observation for up to 24 months using 6 monthly cytology and colposcopy is recommended. Nonpregnant women with HSIL who have an unsatisfactory colposcopic examination require a diagnostic excisional procedure.

Histologic LSIL LSIL is the histological manifestation of a HPV infection. The majority of histologic LSIL will spontaneously regress in the absence of therapy and few cases progress to histologic HSIL (Moscicki et al. 2004; Cox et al. 2003; Trimble et al. 2005). Therefore, it is recommended that histologic LSIL preceded by ASC-US, LSIL cytology, HPV 16 or 18 positivity, or HPV persistence undergo conservative follow-up consisting of cotesting at 1 year, Fig. 90 (Massad et al.

2013a). If histologic LSIL persists for at least 2 years, either continued follow-up or treatment is acceptable. Since the risk of an undetected histologic HSIL or glandular lesion is expected to be higher in women referred for the evaluation of an ASC-H, HSIL, or AGC on cytology, acceptable approaches to women with histologic LSIL preceded by HSIL or AGC cervical cytology includes follow-up utilizing either repeat cotesting at 12 and 24 months, LEEP, or review of the cytological, histological, and colposcopic findings provided the colposcopic examination is satisfactory and endocervical sampling is negative.

Histologic HSIL Women with histologic HSIL (CIN 3) are at significantly high risk of progressing to invasive cervical cancer and therefore treatment is recommended. Provided the colposcopic examination is satisfactory and there is no suggestion of invasive disease (e.g., by either colposcopy, cytology, or histology), both ablative or excisional treatment modalities are considered acceptable

302

T. C. Wright et al.

Fig. 90 ASCCP Consensus Conference algorithm for managing women with histologic LSIL

forms of treatment (Massad et al. 2013a). A diagnostic excisional procedure is recommended for all women with histological HSIL and unsatisfactory colposcopic examination or with recurrent disease. The management of women with histologic HSIL (CIN 2) is controversial and recommendations vary between countries. It is clear that regression rates and progression rates to invasive cervical cancer are lower in women with histologic HSIL (CIN 2) than in women with HSIL (CIN 3) (Tainio et al. 2018; Ostor 1993; Moscicki et al. 2010; Silver et al. 2018). In the US follow-up is recommended for women desirous of maintaining childbearing capacity with histologic HSIL or HSIL (CIN 2), but treatment is considered acceptable (Massad et al. 2013a). For all other women, treatment is recommended.

Endocervical Curettage ECC is performed to evaluate lesion distribution and morphology within the endocervical canal and

to exclude the presence of invasive carcinoma, and unsuspected cervical AIS and invasive adenocarcinoma. Over the last decade the utility of ECC has become the subject of considerable debate (Driggers and Zahn 2008). In the ASCUS/LSIL Triage Study (ALTS), the ECC provided only a minimal 2.2% increase in the detection of HSIL when performed in women under the age of 40 years, but provided a 13% increased detection of histologic HSIL when performed in women 40 years and older (Solomon et al. 2007). A recent NCI study found that ECC detected histologic HSIL in 14% of women undergoing colposcopy (Liu et al. 2017). It was more likely to find HSIL in women with a high-grade cytologic abnormality, those who were HPV 16 positive and those with a high-grade colposcopic impression. In women with ASC-US or LSIL cytology with an unsatisfactory colposcopy, the ECC detected histologic HSIL in 13% of women, whereas when the colposcopic examination was normal or the examination was

5

Precancerous Lesions of the Cervix

303

procedure. Frequently the second, carefully performed curettage yields no atypical epithelium and the patient may be managed on a conservative, outpatient basis. Conversely, the pathologist should be careful not to discount or overlook a few or even a single fragment of atypical epithelium in an ECC. In a review of 21 women who developed invasive cancer after cryotherapy, 7 out of 18 ECCs taken before cryotherapy were found on review to contain SIL that had been missed at the time of original diagnosis (Schmidt et al. 1992). Fig. 91 HSIL present in an endocervical curetting

References satisfactory it detected less than 5% HSIL. Many clinicians routinely perform an ECC during colposcopy. The ECC specimen consists of endocervical tissue fragments, blood, mucus, and, when positive, strips of atypical epithelium, Fig. 91. To avoid the loss of tiny tissue fragments during processing, the clinician should collect and concentrate the sample, including mucus and blood, on a small square of lens paper or using a cytobrush and immediately place it in the fixative (Hoffman et al. 1993). By this method, even the smallest tissue fragments can be recovered easily in the laboratory, embedded, and sectioned entirely. In most instances, when atypical epithelium is detected in the ECC, it lacks underlying stroma and orientation is not possible. As a result, the pathologist can neither rule out underlying invasion nor grade an intraepithelial lesion. In other cases, where the atypical epithelium is well oriented, the pathologist is able to grade the lesion and can, if desired. It is also helpful if the pathologist conveys an estimate of the amount of atypical epithelium that is present in the ECC. If only a few small fragments of atypical epithelium are present in the ECC, these may represent “pick-ups” from a lesion that is actually confined to the portio and does not extend into the endocervical canal. In such cases it may be preferable to reexamine the patient with the colposcope rather than proceeding directly with a diagnostic excisional

Abdul-Karim FW, Fu YS, Reagan JW, Wentz WB (1982) Morphometric study of intraepithelial neoplasia of the uterine cervix. Obstet Gynecol 60:210–214 ACOG (2013) Practice bulletin no. 140: management of abnormal cervical cancer screening test results and cervical cancer precursors. Obstet Gynecol 122(6): 1338–1367 Anciaux D, Lawrence WD, Gregoire L (1997) Glandular lesions of the uterine cervix: prognostic implications of human papillomavirus status. Int J Gynecol Pathol 16:103–110 Andersen ES, Arffmann E (1989) Adenocarcinoma in situ of the uterine cervix: a clinico-pathologic study of 36 cases. Gynecol Oncol 35:1–7 Arbyn M, Sasieni P, Meijer CJ, Clavel C, Koliopoulos G, Dillner J (2006) Chapter 9: Clinical applications of HPV testing: a summary of meta-analyses. Vaccine 24(Suppl 3):S78–S89 Ault KA, Joura EA, Kjaer SK, Iversen OE, Wheeler CM, Perez G et al (2011) Adenocarcinoma in situ and associated human papillomavirus type distribution observed in two clinical trials of a quadrivalent human papillomavirus vaccine. Int J Cancer 128(6):1344–1353. https://doi.org/10.1002/ijc.25723 Baker AC, Eltoum I, Curry RO, Stockard CR, Manne U, Grizzle WE et al (2006) Mucinous expression in benign and neoplastic glandular lesions of the uterine cervix. Arch Pathol Lab Med 130 (10):1510–1515 Baldur-Felskov B, Dehlendorff C, Junge J, Munk C, Kjaer SK (2014) Incidence of cervical lesions in Danish women before and after implementation of a national HPV vaccination program. Cancer Causes Control 25(7):915–922. https://doi.org/10.1007/s10552-0140392-4 Bandyopadhyay S, Austin RM, Dabbs D, Zhao C (2008) Adjunctive human papillomavirus DNA testing is a useful option in some clinical settings for disease risk

304 assessment and triage of females with ASC-H Papanicolaou test results. Arch Pathol Lab Med 132(12):1874–1881 Bergeron C, Ferenczy A, Shah K et al (1987) Multicentric human papillomavirus infections of the female genital tract. Correlation of viral types with abnormal mitotic figures, colposcopic presentation, and location. Obstet Gynecol 69:736–742 Bergeron C, Ordi J, Schmidt D, Trunk MJ, Keller T, Ridder R (2010) Conjunctive p16INK4a testing significantly increases accuracy in diagnosing high-grade cervical intraepithelial neoplasia. Am J Clin Pathol 133(3):395–406. Bernard H-U, Chan S-Y, Delius H (1994) Evolution of papillomaviruses. Curr Top Microbiol Immunol 186:33–54 Bertrand M, Lickrish GM, Colgan TJ (1987) The anatomic distribution of cervical adenocarcinoma in situ: implications for treatment. Am J Obstet Gynecol 157:21–25 Boardman LA, Stanko C, Weitzen S, Sung CJ (2005) Atypical squamous cells of undetermined significance: human papillomavirus testing in adolescents. Obstet Gynecol 105(4):741–746 Bocking A, Nguyen VQ (2004) Diagnostic and prognostic use of DNA image cytometry in cervical squamous intraepithelial lesions and invasive carcinoma. Cancer 102(1):41–54 Boddington MM, Spriggs AI, Cowdell RH (1976) Adenocarcinoma of the uterine cervix: cytological evidence of a long preclinical evolution. Br J Obstet Gynecol 83:900–903 Bollmann R, Mehes G, Torka R, Speich N, Schmitt C, Bollmann M (2003) Human papillomavirus typing and DNA ploidy determination of squamous intraepithelial lesions in liquid-based cytologic samples. Cancer 99(1):57–62 Bollmann R, Mehes G, Speich N, Schmitt C, Bollmann M (2005) Aberrant, highly hyperdiploid cells in human papillomavirus-positive, abnormal cytologic samples are associated with progressive lesions of the uterine cervix. Cancer 105(2):96–100 Boon ME, Baak JPA, Kurver PJH, Overdiep SH, Verdonk GW (1981) Adenocarcinoma in situ of the cervix: an underdiagnosed lesion. Cancer 48:768–773 Bosch FX, de Sanjose S (2007) The epidemiology of human papillomavirus infection and cervical cancer. Dis Markers 23(4):213–227 Bousfield L, Pacey F, Young Q, Krumins I, Osborn R (1980) Expanded cytologic criteria for the diagnosis of adenocarcinoma in situ of the cervix and related lesions. Acta Cytol 24(4):283–296 Boyle DP, McCluggage WG (2015) Stratified mucinproducing intraepithelial lesion (SMILE): report of a case series with associated pathological findings. Histopathology 66(5):658–663. https://doi.org/10.1111/ his.12498 Broders AC (1932) Carcinoma in situ contrasted with benign penetrating epithelium. J Am Med Assoc 99:1670–1674

T. C. Wright et al. Brotherton JM (2016) HPV vaccines: so much learnt, so many more lessons to come. Lancet Oncol 17(1):8–9. https://doi.org/10.1016/S1470-2045(15)00456-8 Brotherton JM, Gertig DM, May C, Chappell G, Saville M (2016) HPV vaccine impact in Australian women: ready for an HPV-based screening program. Med J Aust 204(5):184–1e1 Brown LJR, Wells M (1986) Cervical glandular atypia associated with squamous intraepithelial neoplasia: a premalignant lesion? J Clin Pathol 39:22–28 Brown DR, Shew ML, Qadadri B, Neptune N, Vargas M, Tu W et al (2005) A longitudinal study of genital human papillomavirus infection in a cohort of closely followed adolescent women. J Infect Dis 191(2): 182–192 Burchell AN, Winer RL, de Sanjose S, Franco EL (2006) Chapter 6: Epidemiology and transmission dynamics of genital HPV infection. Vaccine 24(Suppl 3):S52–S61 Burghardt E (1991) Colposcopy-cervical pathology, 2nd edn. Thieme Medical Publishers, New York Burk RD, Harari A, Chen Z (2013) Human papillomavirus genome variants. Virology 445(1–2):232–243. https:// doi.org/10.1016/j.virol.2013.07.018 Cameron RI, Maxwell P, Jenkins D, McCluggage WG (2002) Immunohistochemical staining with MIB1, bcl2 and p16 assists in the distinction of cervical glandular intraepithelial neoplasia from tubo-endometrial metaplasia, endometriosis and microglandular hyperplasia. Histopathology 41(4):313–321 Carreon JD, Sherman ME, Guillen D, Solomon D, Herrero R, Jeronimo J et al (2007) CIN2 is a much less reproducible and less valid diagnosis than CIN3: results from a histological review of population-based cervical samples. Int J Gynecol Pathol 26(4):441–446. https://doi.org/10.1097/pgp.0b013e31805152ab. 00004347-200710000-00013 [pii] Castle PE, Schiffman M, Gravitt PE, Kendall H, Fishman S, Dong H et al (2002) Comparisons of HPV DNA detection by MY09/11 PCR methods. J Med Virol 68(3):417–423 Castle PE, Stoler MH, Solomon D, Schiffman M (2007) The relationship of community biopsy-diagnosed cervical intraepithelial neoplasia grade 2 to the quality control pathology-reviewed diagnoses: an ALTS report. Am J Clin Pathol 127(5):805–815 Castle PE, Rodriguez AC, Burk RD, Herrero R, Wacholder S, Hildesheim A et al (2011) Long-term persistence of prevalently detected human papillomavirus infections in the absence of detectable cervical precancer and cancer. J Infect Dis 203(6):814–822. https://doi.org/10.1093/infdis/jiq116 Choschzick M, Woelber L, Gieseking F, Oosterwijk E, Tennstedt P (2014) Carbonic anhydrase IX is strongly overexpressed in adenocarcinoma in situ of the cervix uteri. Histopathology 64(4):600–602. https://doi.org/ 10.1111/his.12288 Clifford GM, Rana RK, Franceschi S, Smith JS, Gough G, Pimenta JM (2005) Human papillomavirus genotype distribution in low-grade cervical lesions: comparison

5

Precancerous Lesions of the Cervix

by geographic region and with cervical cancer. Cancer Epidemiol Biomarkers Prev 14(5):1157–1164 Clifford G, Franceschi S, Diaz M, Munoz N, Villa LL (2006) Chapter 3: HPV type-distribution in women with and without cervical neoplastic diseases. Vaccine 24(Suppl 3):S26–S34 Colgan TJ, Lickrish GM (1990) The topography and invasive potential of cervical adenocarcinoma in situ, with and without associated squamous dysplasia. Gynecol Oncol 36:246–249 Confortini M, Carozzi F, Dalla Palma P, Ghiringhello B, Parisio F, Prandi S et al (2003) Interlaboratory reproducibility of atypical squamous cells of undetermined significance report: a national survey. Cytopathology 14(5):263–268 Confortini M, Bondi A, Cariaggi MP, Carozzi F, Dalla Palma P, Ghiringhello B et al (2007) Interlaboratory reproducibility of liquid-based equivocal cervical cytology within a randomized controlled trial framework. Diagn Cytopathol 35(9):541–544 Costa S, Venturoli S, Negri G, Sideri M, Preti M, Pesaresi M et al (2012) Factors predicting the outcome of conservatively treated adenocarcinoma in situ of the uterine cervix: an analysis of 166 cases. Gynecol Oncol 124(3):490–495. https://doi.org/10.1016/j. ygyno.2011.11.039 Cox JT, Schiffman M, Solomon D (2003) Prospective follow-up suggests similar risk of subsequent cervical intraepithelial neoplasia grade 2 or 3 among women with cervical intraepithelial neoplasia grade 1 or negative colposcopy and directed biopsy. Am J Obstet Gynecol 188(6):1406–1412 Crocker J, Fox H, Langley FA (1968) Consistency in the histological diagnosis of epithelial abnormalities of the cervix uteri. J Clin Pathol 21:67–70 Crum CP, Egawa K, Fu YS et al (1983) Atypical immature metaplasia (AIM): a subset of human papillomavirus infection of the cervix. Cancer 51:2214–2219 Cullen TS (1900) Cancer of the uterus. Appleton and Company, New York Cullimore JE, Luesley DM, Rollason TP, Byrne P, Buckley CH, Anderson M et al (1992) A prospective study of conization of the cervix in the management of cervical intraepithelial glandular neoplasia (CIGN) – a preliminary report. Br J Obstet Gynecol 99:314–318 Darragh TM, Colgan TJ, Cox JT, Heller DS, Henry MR, Luff RD et al (2012) The lower Anogenital squamous terminology standardization project for HPV-associated lesions: background and consensus recommendations from the College of American Pathologists and the American Society for Colposcopy and Cervical Pathology. J Low Genit Tract Dis 16(3):205–242. https://doi.org/10.1097/ LGT.0b013e31825c31dd Datta SD, Koutsky LA, Ratelle S, Unger ER, Shlay J, McClain T et al (2008) Human papillomavirus infection and cervical cytology in women screened for cervical cancer in the United States, 2003–2005. Ann Intern Med 148(7):493–500

305 de Sanjose S, Diaz M, Castellsague X, Clifford G, Bruni L, Munoz N et al (2007) Worldwide prevalence and genotype distribution of cervical human papillomavirus DNA in women with normal cytology: a meta-analysis. Lancet Infect Dis 7(7):453–459 de Sanjose S, Quint WG, Alemany L, Geraets DT, Klaustermeier JE, Lloveras B et al (2010) Human papillomavirus genotype attribution in invasive cervical cancer: a retrospective cross-sectional worldwide study. Lancet Oncol 11(11):1048–1056. https://doi. org/10.1016/S1470-2045(10)70230-8 de Sanjose S, Brotons M, Pavon MA (2018) The natural history of human papillomavirus infection. Best Pract Res Clin Obstet Gynaecol 47:2–13. https://doi.org/ 10.1016/j.bpobgyn.2017.08.015 de Villiers EM, Fauquet C, Broker TR, Bernard HU, zur Hausen H (2004) Classification of papillomaviruses. Virology 324(1):17–27 Denehy TR, Gregori CA, Breen JL (1997) Endocervical curettage, cone margins, and residual adenocarcinoma in situ of the cervix. Obstet Gynecol 90(1):1–6 Denny LA, Franceschi S, de Sanjose S, Heard I, Moscicki AB, Palefsky J (2012) Human papillomavirus, human immunodeficiency virus and immunosuppression. Vaccine 30(Suppl 5):F168–F174. https://doi.org/10.1016/j. vaccine.2012.06.045 Denslow SA, Rositch AF, Firnhaber C, Ting J, Smith JS (2014) Incidence and progression of cervical lesions in women with HIV: a systematic global review. Int J STD AIDS 25(3):163–177. https://doi.org/10.1177/ 0956462413491735 Doorbar J (2007) Papillomavirus life cycle organization and biomarker selection. Dis Markers 23(4):297–313 Doorbar J (2018) Host control of human papillomavirus infection and disease. Best Pract Res Clin Obstet Gynaecol 47:27–41. https://doi.org/10.1016/j. bpobgyn.2017.08.001 Doorbar J, Quint W, Banks L, Bravo IG, Stoler M, Broker TR et al (2012) The biology and life-cycle of human papillomaviruses. Vaccine 30(Suppl 5):F55–F70. https://doi.org/10.1016/j.vaccine.2012.06.083 Driggers RW, Zahn CM (2008) To ECC or not to ECC: the question remains. Obstet Gynecol Clin North Am 35 (4):583–597; viii Duby JM, DiFurio MJ (2009) Implementation of the ThinPrep imaging system in a tertiary military medical center. Cancer Cytopathol 117(4):264–270 Duggan MA, Benoit JL, McGregor SE, Inoue M, Nation JG, Stuart GCE (1994) Adenocarcinoma in situ of the endocervix: human papillomavirus determination by dot blot hybridization and polymerase chain reaction amplification. Int J Gynecol Pathol 13(2):143–149 Duska LR (2009) Can we improve the detection of glandular cervical lesions: the role and limitations of the Pap smear diagnosis atypical glandular cells (AGC). Gynecol Oncol 114(3):381–382 Elfgren K, Elfstrom KM, Naucler P, Arnheim-Dahlstrom L, Dillner J (2017) Management of women with human papillomavirus persistence: long-term follow-up of a

306 randomized clinical trial. Am J Obstet Gynecol 216(3):264 e1–264 e7. https://doi.org/10.1016/j. ajog.2016.10.042 Eversole GM, Moriarty AT, Schwartz MR, Clayton AC, Souers R, Fatheree LA et al (2010) Practices of participants in the college of american pathologists interlaboratory comparison program in cervicovaginal cytology, 2006. Arch Pathol Lab Med 134(3):331–335. https://doi.org/10.1043/1543-2165-134.3.331 Franquemont DW, Ward BE, Anderson WA, Crum CP (1989) Prediction of “high-risk” cervical papillomavirus infection by biopsy morphology. Am J Clin Pathol 92(5):577–582 Friedell GH, McKay DG (1953) Adenocarcinoma in situ of endocervix. Cancer 6:887–897 Fu YS, Reagan JW (2002) Pathology of the uterine cervix, vagina, and vulva, 2nd edn. W.B. Saunders Company, Philadelphia Fu YS, Reagan JW, Richart RM (1983) Precursors of cervical cancer. Cancer Surv 2:359–382 Fu YS, Huang I, Beaudenon S, Ionesco M, Barrasso R, de Brux J et al (1988) Correlative study of human papillomavirus DNA, histopathology and morphometry in cervical condyloma and intraepithelial neoplasia. Int J Gynecol Pathol 7:297–307 Gage JC, Schiffman M, Hunt WC, Joste N, Ghosh A, Wentzensen N et al (2013) Cervical histopathology variability among laboratories: a population-based statewide investigation. Am J Clin Pathol 139(3):330–335. https://doi.org/10.1309/AJCPSD3ZXJXP7NNB Gatscha RM, Abadi M, Babore S, Chhieng D, Miller MJ, Saigo PE (2001) Smears diagnosed as ASCUS: interobserver variation and follow-up. Diagn Cytopathol 25(2):138–140 Genest DR, Stein L, Cibas E, Sheets E, Zitz JC, Crum CP (1993) A binary (Bethesda) system for classifying cervical cancer precursors: criteria, reproducibility, and viral correlates. Hum Pathol 24(7):730–736 Geng L, Connolly DC, Isacson C, Ronnett BM, Cho KR (1999) Atypical immature metaplasia (AIM) of the cervix: is it related to high-grade squamous intraepithelial lesion (HSIL)? Hum Pathol 30(3): 345–351 Gertig DM, Brotherton JM, Budd AC, Drennan K, Chappell G, Saville AM (2013) Impact of a population-based HPV vaccination program on cervical abnormalities: a data linkage study. BMC Med 11:227. https://doi.org/10.1186/1741-7015-11-227 Gloor E, Hurlimann J (1986a) Cervical intraepithelial glandular neoplasia (adenocarcinoma in situ and glandular dysplasia). A correlative study of 23 cases with histologic grading, histochemical analysis of mucins and immunohistochemical determination of the affinity for four lectins. Cancer 58:1272–1280 Gloor E, Hurlimann J (1986b) Cervical intraepithelial glandular neoplasia (adenocarcinoma in situ and glandular dysplasia). A correlative study of 23 cases with histologic grading, histochemical analysis of mucins,

T. C. Wright et al. and immunohistochemical determination of the affinity for four lectins. Cancer 58(6):1272–1280 Goldie SJ, Kim JJ, Wright TC (2004) Cost-effectiveness of human papillomavirus DNA testing for cervical cancer screening in women aged 30 years or more. Obstet Gynecol 103(4):619–631 Goldstein NS, Mani A (1998) The status and distance of cone biopsy margins as a predictor of excision adequacy for endocervical adenocarcinoma in situ. Am J Clin Pathol 109(6):727–732 Goldstein NS, Ahmad E, Hussain M, Hankin RC, PerezReyes N (1998) Endocervical glandular atypia: does a preneoplastic lesion of adenocarcinoma in situ exist? Am J Clin Pathol 110(2):200–209 Gravitt PE (2011) The known unknowns of HPV natural history. J Clin Invest 121(12):4593–4599. https://doi. org/10.1172/JCI57149 Guido RS, Jeronimo J, Schiffman M, Solomon D (2005) The distribution of neoplasia arising on the cervix: results from the ALTS trial. Am J Obstet Gynecol 193 (4):1331–1337 Hanselaar AG, Vooijs GP, Oud PS, Pahlplatz MM, Beck JLM (1988) DNA ploidy patterns in cervical intraepithelial neoplasia grade III, with and without synchronous invasive squamous cell carcinoma: measurements in nuclei isolated from paraffin-embedded tissue. Cancer 62:2537–2545 Hariri J, Oster A (2007) The negative predictive value of p16INK4a to assess the outcome of cervical intraepithelial neoplasia 1 in the uterine cervix. Int J Gynecol Pathol 26(3):223–228 Helper TK, Dockerty MB, Randall LM (1952) Primary adenocarcinoma of the cervix. Am J Obstet Gynecol 63:800–808 Herbst LH, Lenz J, Van Doorslaer K, Chen Z, Stacy BA, Wellehan JF Jr et al (2009) Genomic characterization of two novel reptilian papillomaviruses, Chelonia mydas papillomavirus 1 and Caretta caretta papillomavirus 1. Virology 383(1):131–135. https://doi.org/10.1016/j. virol.2008.09.022 Herfs M, Yamamoto Y, Laury A, Wang X, Nucci MR, McLaughlin-Drubin ME et al (2012) A discrete population of squamocolumnar junction cells implicated in the pathogenesis of cervical cancer. Proc Natl Acad Sci U S A 109(26):10516–10521. https://doi.org/10.1073/ pnas.1202684109 Higgins GD, Phillips GE, Smith LA, Uzelin DM, Burrell CJ (1992) High prevalence of human papillomavirus transcripts in all grades of cervical intraepithelial glandular neoplasia. Cancer 70:136–146 Ho GY, Studentsov YY, Bierman R, Burk RD (2004) Natural history of human papillomavirus type 16 virus-like particle antibodies in young women. Cancer Epidemiol Biomarkers Prev 13(1):110–116 Hoffman MS, Sterghos S Jr, Gordy LW, Gunasekaran S, Cavanagh D (1993) Evaluation of the cervical canal with the endocervical brush. Obstet Gynecol 82(4 Pt 1):573–577

5

Precancerous Lesions of the Cervix

Hoffmann R, Hirt B, Bechtold V, Beard P, Raj K (2006) Different modes of human papillomavirus DNA replication during maintenance. J Virol 80(9):4431–4439. https://doi.org/10.1128/JVI.80.9.4431-4439.2006 Holl K, Nowakowski AM, Powell N, McCluggage WG, Pirog EC, Collas De Souza S et al (2015) Human papillomavirus prevalence and type-distribution in cervical glandular neoplasias: results from a European multinational epidemiological study. Int J Cancer 137(12):2858–2868. https://doi.org/10.1002/ijc.29651 Holowaty P, Miller AB, Rohan T, To T (1999) Natural history of dysplasia of the uterine cervix. J Natl Cancer Inst 91(3):252–258 Hopkins MP, Roberts JA, Schmidt RW (1988) Cervical adenocarcinoma in situ. Obstet Gynecol 71:842–844 Human Papillomaviruses (2007) IARC monographs on the evaluation of carcinogenic risks to humans. IARC, Lyon Hurlimann J, Gloor E (1984) Adenocarcinoma in situ and invasive adenocarcinoma of the uterine cervix. An immunohistologic study with antibodies specific for several epithelial markers. Cancer 54(1):103–109 Ioffe OB, Sagae S, Moritani S, Dahmoush L, Chen TT, Silverberg SG (2003) Symposium part 3: should pathologists diagnose endocervical preneoplastic lesions “less than” adenocarcinoma in situ?: point. Int J Gynecol Pathol 22(1):18–21 Isaacson Wechsler E, Wang Q, Roberts I, Pagliarulo E, Jackson D, Untersperger C et al (2012) Reconstruction of human papillomavirus type 16-mediated early-stage neoplasia implicates E6/E7 deregulation and the loss of contact inhibition in neoplastic progression. J Virol 86(11):6358–6364. https://doi.org/10.1128/JVI.07069-11 Isacson C, Kessis TD, Hedrick L, Cho KR (1996) Both cell proliferation and apoptosis increase with lesion grade in cervical neoplasia but do not correlate with human papillomavirus type. Cancer Res 56(4):669–674 Ismail SM, Colelough AB, Dinnen JS, Eakins D, Evans DM, Gradwell E et al (1989) Observer variation in histopathological diagnosis and grading of cervical intraepithelial neoplasia. BMJ 298:707–710 Jakobsen A, Kristensen PB, Poulsen HK (1983) Flow cytometric classification of biopsy specimens from cervical intraepithelial neoplasia. Cytometry 4:166–169 Jaworski RC (1990) Endocervical glandular dysplasia, adenocarcinoma in situ, and early invasive (microinvasive) adenocarcinoma of the uterine cervix. Semin Diagn Pathol 7(3):190–204 Jaworski RC, Pacey NR, Greenberg ML, Osborn RA (1988) The histologic diagnosis of adenocarcinoma in situ and related lesions of the cervix uteri. Adenocarcinoma in situ. Cancer 61:1171–1181 Jenson AB, Rosenthal JD, Olson C, Pass F, Lancaster WD (1980) Immunologic relatedness of papillomavirus from different species. J Natl Cancer Inst 64:495–500 Jones BA, Davey DD (2000) Quality management in gynecologic cytology using interlaboratory comparison. Arch Pathol Lab Med 124(5):672–681

307 Kahn JA (2009) HPV vaccination for the prevention of cervical intraepithelial neoplasia. N Engl J Med 361:273 Kashimura M, Shinohara M, Oikawa K, Hamasaki K, Sato H (1990) An adenocarcinoma in situ of the uterine cervix that developed into invasive adenocarcinoma after 5 years. Gynecol Oncol 36:128–133 Katki HA, Schiffman M, Castle PE, Fetterman B, Poitras NE, Lorey T et al (2013a) Five-year risks of CIN 3+ and cervical cancer among women with HPV testing of ASC-US Pap results. J Low Genit Tract Dis 17 (5 Suppl 1):S36–S42. https://doi.org/10.1097/ LGT.0b013e3182854253 Katki HA, Schiffman M, Castle PE, Fetterman B, Poitras NE, Lorey T et al (2013b) Five-year risks of CIN 2+ and CIN 3+ among women with HPV-positive and HPV-negative LSIL Pap results. J Low Genit Tract Dis 17(5 Suppl 1):S43–S49. https://doi.org/10.1097/ LGT.0b013e3182854269 Keating JT, Cviko A, Riethdorf S, Riethdorf L, Quade BJ, Sun D et al (2001) Ki-67, cyclin E, and p16INK4 are complimentary surrogate biomarkers for human papilloma virus-related cervical neoplasia. Am J Surg Pathol 25(7):884–891 Kelly H, Weiss HA, Benavente Y, de Sanjose S, Mayaud P, ART and HPV Review Group et al (2018) Association of antiretroviral therapy with high-risk human papillomavirus, cervical intraepithelial neoplasia, and invasive cervical cancer in women living with HIV: a systematic review and meta-analysis. Lancet HIV 5(1):e45–e58. https://doi.org/10.1016/S2352-3018(17)30149-2 Kennedy AW, Biscotti CV (2002) Further study of the management of cervical adenocarcinoma in situ. Gynecol Oncol 86(3):361–364 Kim TJ, Kim HS, Park CT, Park IS, Hong SR, Park JS et al (1999) Clinical evaluation of follow-up methods and results of atypical glandular cells of undetermined significance (AGUS) detected on cervicovaginal Pap smears. Gynecol Oncol 73(2):292–298 Kinney WK, Manos MM, Hurley LB, Ransley JE (1998) Where’s the high-grade cervical neoplasia? The importance of minimally abnormal Papanicolaou diagnoses. Obstet Gynecol 91(6):973–976 Kirkland JA (1963) Atypical epithelial changes in the uterine cervix. J Clin Pathol 16:150–154 Kjaer SK, Frederiksen K, Munk C, Iftner T (2010) Longterm absolute risk of cervical intraepithelial neoplasia grade 3 or worse following human papillomavirus infection: role of persistence. J Natl Cancer Inst 102(19):1478–1488. https://doi.org/10.1093/jnci/ djq356 Klaes R, Friedrich T, Spitkovsky D, Ridder R, Rudy W, Petry U et al (2001) Overexpression of p16(INK4A) as a specific marker for dysplastic and neoplastic epithelial cells of the cervix uteri. Int J Cancer 92(2):276–284 Kolstad P, Klem V (1976) Long-term followup of 1121 cases of carcinoma in situ. Obstet Gynecol 48(2):125–129

308 Koss LG (1978) Dysplasia. A real concept or a misnomer? Obstet Gynecol 51:374 Koss L, Durfee GR (1956) Unusual patterns of squamous epithelium of uterine cervix: cytologic and pathologic study of koilocytotic atypia. Ann N Y Acad Sci 63:1245–1261 Koss LG, Stewart FW, Foote FW, Jordan MJ, Bader GM, Day E (1963) Some histological aspects of behavior of epidermoid carcinoma in situ and related lesions of the uterine cervix. Cancer 16(9):1160–1211 Kudo R, Sagai S, Hayakawa O, Ito E, Horimoto E, Hashimoto M (1991) Morphology of adenocarcinoma in situ and microinvasive adenocarcinoma of the uterine cervix. Acta Cytol 35:109–116 Kulasingam SL, Kim JJ, Lawrence WF, Mandelblatt JS, Myers ER, Schiffman M et al (2006) Cost-effectiveness analysis based on the atypical squamous cells of undetermined significance/low-grade squamous intraepithelial lesion Triage Study (ALTS). J Natl Cancer Inst 98(2):92–100 Kurman RJ, Ronnett BM, Sherman ME, Wilkinson EJ (2010) Tumors of the cervix, vagina and vulva. AFIP atlas of tumor pathology, vol fourth series. American Registry of Pathology in conjunction with Armed Forces Institute of Pathology, Washington, DC Kurman R, Carcangiu ML, Herrington CS, Young RH (2014) WHO classification of tumors of female reproductive organs. World Health Organization Classification of Tumors. International Agency for Research on Cancer, Lyon Leary J, Jaworski R, Houghton R (1991) In-situ hybridization using biotinylated DNA probes to human papillomavirus in adenocarcinoma in-situ and endocervical glandular dysplasia of the uterine cervix. Pathology 23:85–89 Lee KR (2003) Symposium part 4: should pathologists diagnose endocervical preneoplastic lesions “less than” adenocarcinoma in situ?: counterpoint. Int J Gynecol Pathol 22(1):22–24 Lee KR, Sun D, Crum CP (2000) Endocervical intraepithelial glandular atypia (dysplasia): a histopathologic, human papillomavirus, and MIB-1 analysis of 25 cases. Hum Pathol 31(6):656–664 Lee S, Rose MS, Sahasrabuddhe V, Zhao R, Duggan M, Afiero A (2016) Tissue-based immunohistochemical biomarker accuracy in the diagnosis of malignant glandular lesions of the uterine cervix: a systematic review of the literature and meta-analysis. Int J Gynecol Pathol 36:310–322 Liao SY, Brewer C, Zavada J, Pastorek J, Pastorekova S, Manetta A et al (1994) Identification of the MN antigen as a diagnostic biomarker of cervical intraepithelial squamous and glandular neoplasia and cervical carcinomas. Am J Pathol 145(3):598–609 Liao SY, Rodgers WH, Kauderer J, Darcy KM, Carter R, Susumu N et al (2013) Endocervical glandular neoplasia associated with lobular endocervical glandular hyperplasia is HPV-independent and correlates with carbonic anhydrase-IX expression: a Gynaecological

T. C. Wright et al. Oncology Group Study. Br J Cancer 108(3):613–620. https://doi.org/10.1038/bjc.2012.578 Liman AK, Giampoli EJ, Bonfiglio TA (2005) Should women with atypical squamous cells, cannot exclude high-grade squamous intraepithelial lesion, receive reflex human papillomavirus-DNA testing? Cancer 105(6):457–460 Liu AH, Walker J, Gage JC, Gold MA, Zuna R, Dunn ST et al (2017) Diagnosis of cervical Precancers by Endocervical curettage at colposcopy of women with abnormal cervical cytology. Obstet Gynecol 130(6):1218–1225. https://doi.org/10.1097/AOG.000 0000000002330 Lonky NM, Sadeghi M, Tsadik GW, Petitti D (1999) The clinical significance of the poor correlation of cervical dysplasia and cervical malignancy with referral cytologic results. Am J Obstet Gynecol 181(3):560–566 Loureiro J, Oliva E (2014) The spectrum of cervical glandular neoplasia and issues in differential diagnosis. Arch Pathol Lab Med 138(4):453–483. https://doi. org/10.5858/arpa.2012-0493-RA Luesley DM, Jordan JA, Woodman CBJ, Watson N, Williams DR, Waddell C (1987) A retrospective review of adenocarcinoma-in-situ and glandular atypia of the uterine cervix. Br J Obstet Gynaecol 94:699–703 Malpica A, Matisic JP, Niekirk DV, Crum CP, Staerkel GA, Yamal JM et al (2005) Kappa statistics to measure interrater and intrarater agreement for 1790 cervical biopsy specimens among twelve pathologists: qualitative histopathologic analysis and methodologic issues. Gynecol Oncol 99(3 Suppl 1):S38–S52. https://doi.org/ 10.1016/j.ygyno.2005.07.040 Marques T, Andrade LA, Vassallo J (1996) Endocervical tubal metaplasia and adenocarcinoma in situ: role of immunohistochemistry for carcinoembryonic antigen and vimentin in differential diagnosis. Histopathology 28(6):549–550 Massad LS (2006) More is more: improving the sensitivity of colposcopy. Obstet Gynecol 108(2):246–247 Massad LS, Einstein MH, Huh WK, Katki HA, Kinney WK, Schiffman M et al (2013a) 2012 updated consensus guidelines for the management of abnormal cervical cancer screening tests and cancer precursors. Obstet Gynecol 121(4):829–846. https://doi.org/10.1097/ AOG.0b013e3182883a34 Massad LS, Einstein MH, Huh WK, Katki HA, Kinney WK, Schiffman M et al (2013b) 2012 updated consensus guidelines for the management of abnormal cervical cancer screening tests and cancer precursors. J Low Genit Tract Dis 17(5 Suppl 1):S1–S27. https://doi.org/ 10.1097/LGT.0b013e318287d329 McCann MF, Irwin DE, Walton LA, Hulka BS, Morton JL, Axelrad CM (1992) Nicotine and cotinine in the cervical mucus of smokers, passive smokers, and nonsmokers. Cancer Epidemiol Biomarkers Prev 1:125–129 McCluggage WG (2007) Immunohistochemistry as a diagnostic aid in cervical pathology. Pathology 39(1):97–111

5

Precancerous Lesions of the Cervix

McCluggage WG, Maxwell P, McBride HA, Hamilton PW, Bharucha H (1995) Monoclonal antibodies Ki-67 and MIB1 in the distinction of tuboendometrial metaplasia from endocervical adenocarcinoma and adenocarcinoma in situ in formalin-fixed material. Int J Gynecol Pathol 14(3):209–216 McCredie MR, Sharples KJ, Paul C, Baranyai J, Medley G, Jones RW et al (2008) Natural history of cervical neoplasia and risk of invasive cancer in women with cervical intraepithelial neoplasia 3: a retrospective cohort study. Lancet Oncol 9(5):425–434 McDonald AC, Tergas AI, Kuhn L, Denny L, Wright TC Jr (2014) Distribution of human papillomavirus genotypes among HIV-positive and HIV-negative women in cape town. S Afr Front Oncol 4:48. https://doi.org/ 10.3389/fonc.2014.00048 Meisels A, Fortin R (1976) Condylomatous lesions of the cervix and vagina. I. Cytologic patterns. Acta Cytologica 20:505–509 Melnikow J, Nuovo J, Willan AR, Chan BK, Howell LP (1998) Natural history of cervical squamous intraepithelial lesions: a meta-analysis. Obstet Gynecol 92(4 Pt 2):727–735 Mikami Y, McCluggage WG (2013) Endocervical glandular lesions exhibiting gastric differentiation: an emerging spectrum of benign, premalignant, and malignant lesions. Adv Anat Pathol 20(4):227–237. https://doi. org/10.1097/PAP.0b013e31829c2d66 Mitchell MF, Tortolero-Luna G, Wright T, Sarkar A, Richards-Kortum R, Hong WK et al (1996) Cervical human papillomavirus infection and intraepithelial neoplasia: a review. J Natl Cancer Inst Monogr 21:17–25 Mittal KR, Chan W, Demopoulos RL (1990) Sensitivity and specificity of various morphological features of cervical condylomas. Arch Pathol Lab Med 114:1038–1041 Moscicki AB, Hills N, Shiboski S, Powell K, Jay N, Hanson E et al (2001) Risks for incident human papillomavirus infection and low-grade squamous intraepithelial lesion development in young females. J Am Med Assoc 285:2995–3002 Moscicki AB, Shiboski S, Hills NK, Powell KJ, Jay N, Hanson EN et al (2004) Regression of low-grade squamous intra-epithelial lesions in young women. Lancet 364(9446):1678–1683 Moscicki AB, Schiffman M, Kjaer S, Villa LL (2006) Chapter 5: Updating the natural history of HPV and anogenital cancer. Vaccine 24(Suppl 3):S42–S51 Moscicki AB, Ma Y, Wibbelsman C, Darragh TM, Powers A, Farhat S et al (2010) Rate of and risks for regression of cervical intraepithelial neoplasia 2 in adolescents and young women. Obstet Gynecol 116(6):1373–1380. https://doi.org/10.1097/AOG.0b013e3181fe777f. 00006250-201012000-00021 [pii] Moscicki AB, Schiffman M, Burchell A, Albero G, Giuliano AR, Goodman MT et al (2012) Updating the natural history of human papillomavirus and anogenital cancers. Vaccine 30(Suppl 5):F24–F33. https://doi.org/ 10.1016/j.vaccine.2012.05.089

309 Mourits MJE, Pieters WJ, Hollema H, Burger M (1992) Three-group metaphase as a morphologic criterion of progressive cervical intraepithelial neoplasia. Am J Obstet Gynecol 167(3):591–595 Munoz N, Castellsague X, De Gonzalez AB, Gissman L (2006) Chapter 1: HPV in the etiology of human cancer. Vaccine 24S3:S1–S10 Munro A, Codde J, Spilsbury K, Stewart CJ, Steel N, Leung Y et al (2017) Risk of persistent or recurrent neoplasia in conservatively treated women with cervical adenocarcinoma in situ with negative histological margins. Acta Obstet Gynecol Scand 96(4):432–437. https://doi.org/10.1111/aogs.13110 Muntz HG (1996) Can cervical adenocarcinoma in situ be safely managed by conization alone? Gynecol Oncol 61 (3):301–303. https://doi.org/10.1006/gyno.1996.0146 Murphy N, Ring M, Killalea AG, Uhlmann V, O’Donovan M, Mulcahy F et al (2003) p16INK4A as a marker for cervical dyskaryosis: CIN and cGIN in cervical biopsies and ThinPrep smears. J Clin Pathol 56(1):56–63 Murphy N, Heffron CC, King B, Ganuguapati UG, Ring M, McGuinness E et al (2004) p16INK4A positivity in benign, premalignant and malignant cervical glandular lesions: a potential diagnostic problem. Virchows Arch 445(6):610–615 Nasiell K, Nasiell M, Vaclavinkova V (1983) Behavior of moderate cervical dysplasia during long-term followup. Obstet Gynecol 61:609–614 Nayar R, Wilbur DC (2015) The Bethesda system for reporting cervical cytology. Definitions, criteria and explanatory notes, 3rd edn. Springer International Publishing, Heidelberg/New York/Dordrecht/London Negri G, Egarter-Vigl E, Kasal A, Romano F, Haitel A, Mian C (2003) p16INK4a is a useful marker for the diagnosis of adenocarcinoma of the cervix uteri and its precursors: an immunohistochemical study with immunocytochemical correlations. Am J Surg Pathol 27(2):187–193 Ng A (1993) Glandular diseases of the uterus. In: Keebler CM, Somrak TM (eds) The manual of Cytotechnology. American Society of Clinical Pathologists, Chicago Ostor AG (1993) Natural history of cervical intraepithelial neoplasia: a critical review. Int J Gynecol Pathol 12:186–192 Ostor AG, Pagano R, Davoren RAM, Fortune DW, Chanen W, Rome R (1984) Adenocarcinoma in situ of the cervix. Int J Gynecol Pathol 3:179–190 Ostor AG, Duncan A, Quinn M, Rome R (2000) Adenocarcinoma in situ of the uterine cervix: an experience with 100 cases. Gynecol Oncol 79(2):207–210 Papillomavirus Episteme [database on the Internet] (2018). Available from: http://pave.niaid.nih.gov/#home. Accessed Park JJ, Genest DR, Sun D, Crum CP (1999) Atypical immature metaplastic-like proliferations of the cervix: diagnostic reproducibility and viral (HPV) correlates. Hum Pathol 30(10):1161–1165

310 Park JJ, Sun D, Quade BJ, Flynn C, Sheets EE, Yang A et al (2000) Stratified mucin-producing intraepithelial lesions of the cervix: adenosquamous or columnar cell neoplasia? Am J Surg Pathol 24(10):1414–1419 Parkin DM (1991) Screening for cervix cancer in developing countries. In: Miller AB, Chamberlain J, Day NE, Hakama M, Prorok PC (eds) Cancer screening. Cambridge University Press, Cambridge, pp 184–198 Patten SF (1978) Diagnostic cytopathology of the uterine cervix, 2nd edn. S. Karger, Basal Pedersen K, Burger EA, Sy S, Kristiansen IS, Kim JJ (2016) Cost-effective management of women with minor cervical lesions: revisiting the application of HPV DNA testing. Gynecol Oncol 143(2):326–333. https://doi.org/10.1016/j.ygyno.2016.08.231 Pemberton FA, Smith GV (1929) The early diagnosis and prevention of carcinoma of the cervix: a clinical pathologic study of borderline cases treated at the free hospital for women. Am J Obstet Gynecol 17:165 Petry KU, Cox JT, Johnson K, Quint W, Ridder R, Sideri M et al (2016) Evaluating HPV-negative CIN2+ in the ATHENA trial. Int J Cancer 138(12):2932–2939. https://doi.org/10.1002/ijc.30032 Pirog EC, Isacson C, Szabolcs MJ, Kleter B, Quint W, Richart RM (2002) Proliferative activity of benign and neoplastic endocervical epithelium and correlation with HPV DNA detection. Int J Gynecol Pathol 21(1):22–26 Plaxe SC, Saltzstein SL (1999) Estimation of the duration of the preclinical phase of cervical adenocarcinoma suggests that there is ample opportunity for screening. Gynecol Oncol 75(1):55–61 Pretorius RG, Zhang X, Belinson JL, Zhang WH, Ren SD, Bao YP et al (2006) Distribution of cervical intraepithelial neoplasia 2, 3 and cancer on the uterine cervix. J Low Genit Tract Dis 10(1):45–50 Pretorius RG, Belinson JL, Burchette RJ, Hu S, Zhang X, Qiao YL (2011) Regardless of skill, performing more biopsies increases the sensitivity of colposcopy. J Low Genit Tract Dis 15(3):180–188. https://doi.org/ 10.1097/LGT.0b013e3181fb4547 Queiroz C, Silva TC, Alves VA, Villa LL, Costa MC, Travassos AG et al (2006a) P16(INK4a) expression as a potential prognostic marker in cervical pre-neoplastic and neoplastic lesions. Pathol Res Pract 202(2):77–83 Queiroz C, Silva TC, Alves VA, Villa LL, Costa MC, Travassos AG et al (2006b) Comparative study of the expression of cellular cycle proteins in cervical intraepithelial lesions. Pathol Res Pract 202(10): 731–737 Quint KD, de Koning MN, van Doorn LJ, Quint WG, Pirog EC (2010) HPV genotyping and HPV16 variant analysis in glandular and squamous neoplastic lesions of the uterine cervix. Gynecol Oncol 117(2):297–301. https:// doi.org/10.1016/j.ygyno.2010.02.003. S0090-8258 (10)00131-9 [pii] Ratnam S, Franco EL, Ferenczy A (2000) Human papillomavirus testing for primary screening of cervical cancer

T. C. Wright et al. precursors. Cancer Epidemiol Biomarkers Prev 9(9):945–951 Renshaw AA, Mody DR, Lozano RL, Volk EE, Walsh MK, Davey DD et al (2004) Detection of adenocarcinoma in situ of the cervix in Papanicolaou tests: comparison of diagnostic accuracy with other high-grade lesions. Arch Pathol Lab Med 128(2):153–157 Reuschenbach M, Wentzensen N, Dijkstra MG, von Knebel Doeberitz M, Arbyn M (2014) p16INK4a immunohistochemistry in cervical biopsy specimens: a systematic review and meta-analysis of the interobserver agreement. Am J Clin Pathol 142(6):767–772. https://doi. org/10.1309/AJCP3TPHV4TRIZEK Richart RM (1973) Cervical intraepithelial neoplasia: a review. In: Sommers SC (ed) Pathology annual. Appleton-Century-Crofts, East Norwalk, pp 301–328 Richart RM (1990) A modified terminology for cervical intraepithelial neoplasia. Obstet Gynecol 75(1):131–133 Riethdorf L, Riethdorf S, Lee KR, Cviko A, Loning T, Crum CP (2002) Human papillomaviruses, expression of p16, and early endocervical glandular neoplasia. Hum Pathol 33(9):899–904 Robertson AJ, Anderson JM, Beck JS, Burnett RA, Howatson SR, Lee FD et al (1989) Observer variability in histopathological reporting of cervical biopsy specimens. J Clin Pathol 42(3):231–238 Ronnett BM, Manos MM, Ransley JE, Fetterman BJ, Kinney WK, Hurley LB et al (1999) Atypical glandular cells of undetermined significance (AGUS): cytopathologic features, histopathologic results, and human papillomavirus DNA detection. Hum Pathol 30(7):816–825 Roura E, Travier N, Waterboer T, de Sanjose S, Bosch FX, Pawlita M et al (2016) The influence of hormonal factors on the risk of Developing Cervical Cancer and Pre-Cancer: results from the EPIC Cohort. PLoS One 11(1):e0147029. https://doi.org/10.1371/journal. pone.0147029 Ruba S, Schoolland M, Allpress S, Sterrett G (2004) Adenocarcinoma in situ of the uterine cervix: screening and diagnostic errors in Papanicolaou smears. Cancer 102(5):280–287. https://doi.org/10.1002/cncr.20600 Salani R, Puri I, Bristow RE (2009) Adenocarcinoma in situ of the uterine cervix: a metaanalysis of 1278 patients evaluating the predictive value of conization margin status. Am J Obstet Gynecol 200(2):182. e1–182.e5 Sanati S, Huettner P, Ylagan LR (2010) Role of ProExC: a novel immunoperoxidase marker in the evaluation of dysplastic squamous and glandular lesions in cervical specimens. Int J Gynecol Pathol 29(1):79–87. https:// doi.org/10.1097/PGP.0b013e3181ae81a0 Schiffman M, Castle PE, Jeronimo J, Rodriguez AC, Wacholder S (2007) Human papillomavirus and cervical cancer. Lancet 370(9590):890–907. https://doi.org/ 10.1016/S0140-6736(07)61416-0 Schiffman M, Wentzensen N, Wacholder S, Kinney W, Gage JC, Castle PE (2011) Human papillomavirus

5

Precancerous Lesions of the Cervix

testing in the prevention of cervical cancer. J Natl Cancer Inst 103(5):368–383. https://doi.org/10.1093/ jnci/djq562 Schmidt C, Pretorius RG, Bonin M, Hanson L, Semrad N, Watring W (1992) Invasive cervical cancer following cryotherapy for cervical intraepithelial neoplasia or human papillomavirus infection. Obstet Gynecol 80(5):797–800 SEER Program – National Cancer Institute, USA (2001). http://www-seer.ims.nci.nih.gov/ScientificSystems/. Accessed 20 Mar 2003, 2009 Sherman ME, Tabbara SO, Scott DR, Kurman RJ, Glass AG, Manos MM et al (1999) “ASCUS, rule out HSIL”: cytologic features, histologic correlates, and human papillomavirus detection. Mod Pathol 12(4):335–342 Sherman ME, Schiffman M, Cox JT, ASCUS LSIL Triage Study Group (2002) Effects of age and HPV load on colposcopic triage: data from the ASCUS LSIL Triage Study (ALTS). J Natl Cancer Inst 94:102–107 Shin CH, Schorge JO, Lee KR, Sheets EE (2000) Conservative management of adenocarcinoma in situ of the cervix. Gynecol Oncol 79(1):6–10 Sigurdsson K, Sigvaldason H (2007) Is it rational to start population-based cervical cancer screening at or soon after age 20? Analysis of time trends in preinvasive and invasive diseases. Eur J Cancer 43(4):769–774 Silver MI, Gage JC, Schiffman M, Fetterman B, Poitras NE, Lorey T et al (2018) Clinical outcomes after conservative Management of Cervical Intraepithelial Neoplasia Grade 2 (CIN2) in women ages 21–39 years. Cancer Prev Res (Phila) 11(3):165–170. https://doi. org/10.1158/1940-6207.CAPR-17-0293 Smith JS, Green J, Berrington de Gonzalez A, Appleby P, Peto J, Plummer M et al (2003) Cervical cancer and use of hormonal contraceptives: a systematic review. Lancet 361(9364):1159–1167 Smith JS, Lindsay L, Hoots B, Keys J, Franceschi S, Winer R et al (2007) Human papillomavirus type distribution in invasive cervical cancer and high-grade cervical lesions: a meta-analysis update. Int J Cancer 121(3):621–632 Snijders PJ, Steenbergen RD, Heideman DA, Meijer CJ (2006) HPV-mediated cervical carcinogenesis: concepts and clinical implications. J Pathol 208(2): 152–164 Solomon D, Davey D, Kurman R, Moriarty A, O’Connor D, Prey M et al (2002) The 2001 Bethesda system: terminology for reporting results of cervical cytology. JAMA 287(16):2114–2119 Solomon D, Stoler M, Jeronimo J, Khan M, Castle P, Schiffman M (2007) Diagnostic utility of endocervical curettage in women undergoing colposcopy for equivocal or low-grade cytologic abnormalities. Obstet Gynecol 110(2):288–295 Stanley MA (2012) Epithelial cell responses to infection with human papillomavirus. Clin Microbiol Rev 25(2):215–222. https://doi.org/10.1128/CMR.05028-11 Stoler MH, Schiffman M (2001) Interobserver reproducibility of cervical cytologic and histologic

311 interpretations: realistic estimates from the ASCUSLSIL Triage Study. JAMA 285(11):1500–1505 Stoler MH, Wright TC Jr, Sharma A, Apple R, Gutekunst K, Wright TL (2011) High-risk human papillomavirus testing in women with ASC-US cytology: results from the ATHENA HPV study. Am J Clin Pathol 135(3):468–475. https://doi.org/10.1309/ AJCPZ5JY6FCVNMOT.135/3/468 Stoler MH, Wright TC Jr, Cuzick J, Dockter J, Reid JL, Getman D et al (2013) APTIMA HPV assay performance in women with atypical squamous cells of undetermined significance cytology results. Am J Obstet Gynecol 208(2):144.e1–144.e8. https://doi.org/ 10.1016/j.ajog.2012.12.003 Stoler MH, Wright TC Jr, Ferenczy A, Ranger-Moore J, Fang Q, Kapadia M et al (2018) Routine use of adjunctive p16 immunohistochemistry improves diagnostic agreement of cervical biopsy interpretation: results from the CERTAIN Study. Am J Surg Pathol. https:// doi.org/10.1097/PAS.0000000000001072 Szarewski A, Cuzick J (1998) Smoking and cervical neoplasia; a review of the evidence. J Epidemiol Biostat 3:229 Szarewski A, Maddox P, Royston P, Jarvis M, Anderson M, Guillebaud J et al (2001) The effect of stopping smoking on cervical Langerhans’ cells and lymphocytes. BJOG 108(3):295–303 Tainio K, Athanasiou A, Tikkinen KAO, Aaltonen R, Cardenas J, Hernandes J et al (2018) Clinical course of untreated cervical intraepithelial neoplasia grade 2 under active surveillance: systematic review and meta-analysis. BMJ 360:k499. https://doi.org/ 10.1136/bmj.k499 Talia KL, McCluggage WG (2018) The developing spectrum of gastric-type cervical glandular lesions. Pathology 50(2):122–133. https://doi.org/10.1016/j.pathol. 2017.09.009 Talia KL, Stewart CJR, Howitt BE, Nucci MR, McCluggage WG (2017) HPV-negative gastric type adenocarcinoma in situ of the cervix: a Spectrum of rare lesions exhibiting gastric and intestinal differentiation. Am J Surg Pathol 41(8):1023–1033. https://doi. org/10.1097/PAS.0000000000000855 Tase T, Okagaki T, Clark BA, Twiggs LB, Ostrow RS, Faras AJ (1989a) Human papillomavirus DNA in glandular dysplasia and microglandular hyperplasia: presumed precursors of adenocarcinoma of the uterine cervix. Obstet Gynecol 73(6):1005–1008 Tase T, Okagaki T, Clark BA, Twiggs LB, Ostrow RS, Faras AJ (1989b) Human papillomavirus DNA in adenocarcinoma in situ, microinvasive adenocarcinoma of the uterine cervix and coexisting cervical squamous intraepithelial neoplasia. Int J Gynecol Pathol 8(1):8–17 Tavassoli FA, Devilee P (eds) (2003) Pathology and genetics of Tumours of the breast and female genital organs. World Health Organization classification of tumors. IARC Press, Lyon Tewari D, Novak-Weekley S, Hong C, Aslam S, Behrens CM (2017) Performance of the cobas HPV test for the

312 triage of atypical squamous cells of undetermined significance cytology in cervical specimens collected in SurePath. Am J Clin Pathol 148(5):450–457. https:// doi.org/10.1093/ajcp/aqx091 The (1988) Bethesda system for reporting cervical/vaginal cytologic diagnoses. Developed and approved at a National Cancer Institute Workshop, Bethesda, 12–13, 1988 Dec. 1988;11(5):291–297 Tjalma WA, Fiander A, Reich O, Powell N, Nowakowski AM, Kirschner B et al (2013) Differences in human papillomavirus type distribution in high-grade cervical intraepithelial neoplasia and invasive cervical cancer in Europe. Int J Cancer 132(4):854–867. https://doi.org/ 10.1002/ijc.27713 Trimble CL, Piantadosi S, Gravitt P, Ronnett B, Pizer E, Elko A et al (2005) Spontaneous regression of highgrade cervical dysplasia: effects of human papillomavirus type and HLA phenotype. Clin Cancer Res 11 (13):4717–4723. https://doi.org/10.1158/1078-0432. CCR-04-2599 Tringler B, Gup CJ, Singh M, Groshong S, Shroyer AL, Heinz DE et al (2004) Evaluation of p16INK4a and pRb expression in cervical squamous and glandular neoplasia. Hum Pathol 35(6):689–696 Trowell JE (1985) Intestinal metaplasia with argentaffin cells in the uterine cervix. Histopathology 9:561–569 Tsoumpou I, Arbyn M, Kyrgiou M, Wentzensen N, Koliopoulos G, Martin-Hirsch P et al (2009) p16 (INK4a) immunostaining in cytological and histological specimens from the uterine cervix: a systematic review and meta-analysis. Cancer Treat Rev 35(3):210–220 Umezawa T, Umemori M, Horiguchi A, Nomura K, Takahashi H, Yamada K et al (2015) Cytological variations and typical diagnostic features of endocervical adenocarcinoma in situ: a retrospective study of 74 cases. Cytojournal 12:8. https://doi.org/10.4103/ 1742-6413.156081 Ursin G, Pike MC, Preston-Martin S, d’Ablaing G 3rd, Peters RK (1996) Sexual, reproductive, and other risk factors for adenocarcinoma of the cervix: results from a population-based case-control study (California, United States) [see comments]. Cancer Causes Control 7(3):391–401 van der Horst J, Siebers AG, Bulten J, Massuger LF, de Kok IM (2017) Increasing incidence of invasive and in situ cervical adenocarcinoma in the Netherlands during 2004–2013. Cancer Med 6(2):416–423. https://doi.org/ 10.1002/cam4.971 Volgareva G, Zavalishina L, Andreeva Y, Frank G, Krutikova E, Golovina D et al (2004) Protein p16 as a marker of dysplastic and neoplastic alterations in cervical epithelial cells. BMC Cancer 4:58. https://doi.org/ 10.1186/1471-2407-4-58 Wang SS, Sherman ME, Hildesheim A, Lacey JV Jr, Devesa S (2004a) Cervical adenocarcinoma and squamous cell carcinoma incidence trends among white women and black women in the United States for 1976–2000. Cancer 100(5):1035–1044

T. C. Wright et al. Wang SS, Trunk M, Schiffman M, Herrero R, Sherman ME, Burk RD et al (2004b) Validation of p16INK4a as a marker of oncogenic human papillomavirus infection in cervical biopsies from a population-based cohort in Costa Rica. Cancer Epidemiol Biomarkers Prev 13(8):1355–1360 Ward BE, Burkett BA, Peterson C, Nichols M, Brennan C, Birch LM et al (1990) Cytological correlates of cervical papillomavirus infection. Int J Gynecol Pathol 9:297–305 Waxman AG, Conageski C, Silver MI, Tedeschi C, Stier EA, Apgar B et al (2017) ASCCP colposcopy standards: how do we perform colposcopy? Implications for establishing standards. J Low Genit Tract Dis 21(4):235–241. https://doi.org/10.1097/ LGT.0000000000000336 Williams J (1888) Cancer of the uterus: Harveian lectures for 1886. H.K. Lewis, London Winer RL, Lee SK, Hughes JP, Adam DE, Kiviat NB, Koutsky LA (2003) Genital human papillomavirus infection: incidence and risk factors in a cohort of female university students. Am J Epidemiol 157(3):218–226 Winer RL, Hughes JP, Feng Q, O’Reilly S, Kiviat NB, Holmes KK et al (2006) Condom use and the risk of genital human papillomavirus infection in young women. N Engl J Med 354(25):2645–2654 Witkiewicz A, Lee KR, Brodsky G, Cviko A, Brodsky J, Crum CP (2005) Superficial (early) endocervical adenocarcinoma in situ: a study of 12 cases and comparison to conventional AIS. Am J Surg Pathol 29(12):1609–1614 Wolf JK, Levenback C, Malpica A, Morris M, Burke T, Mitchell MF (1996) Adenocarcinoma in situ of the cervix: significance of cone biopsy margins. Obstet Gynecol 88(1):82–86. https://doi.org/10.1016/00297844(96)00083-X Woodman CB, Collins SI, Young LS (2007) The natural history of cervical HPV infection: unresolved issues. Nat Rev Cancer 7(1):11–22. https://doi.org/10.1038/ nrc2050 Workshop NCI (1991) The revised Bethesda System for reporting cervical/vaginal cytologic diagnoses. Report of the 1991 Bethesda workshop. J Am Med Assoc 267:1892 Wright TC Jr (2006) Chapter 3: Pathology of HPV infection at the cytologic and histologic levels: basis for a 2-tiered morphologic classification system. Int J Gynaecol Obstet 94(Suppl 1):S22–S31 Wright TC Jr, Massad LS, Dunton CJ, Spitzer M, Wilkinson EJ, Solomon D (2007) 2006 consensus guidelines for the management of women with abnormal cervical screening tests. J Low Genit Tract Dis 11(4):201–222 Wright TC Jr, Stoler MH, Behrens CM, Apple R, Derion T, Wright TL (2012) The ATHENA human papillomavirus study: design, methods, and baseline results. Am J Obstet Gynecol 206(1):46 e1–46e11. https://doi.org/ 10.1016/j.ajog.2011.07.024

5

Precancerous Lesions of the Cervix

Wright TC, Kuhn L (2006) Immunosuppression and the cervix; human immunodeficiency virus (HIV). In: Jordan JA, Singer A (eds) The cervix. Blackwell, Malden, pp 450–517 Wright TC, Kurman RJ (1994) A critical review of the morphologic classification systems of preinvasive lesions of the cervix: the scientific basis of the paradigm. Papillomavirus Rep 5:175–181 Wright TC, Schiffman M (2003) Adding a test for human papillomavirus DNA to cervical-cancer screening. N Engl J Med 348(6):489–490 Wright TC, Stoler MH, Behrens CM, Sharma A, Zhang G, Wright TL (2015) Primary cervical cancer screening with human papillomavirus: end of study results from the ATHENA study using HPV as the first-line screening test. Gynecol Oncol 136(2):189–197. https://doi. org/10.1016/j.ygyno.2014.11.076 Xu L, Verdoodt F, Wentzensen N, Bergeron C, Arbyn M (2016) Triage of ASC-H: a meta-analysis of the accuracy of high-risk HPV testing and other markers to

313 detect cervical precancer. Cancer Cytopathol 124(4):261–272. https://doi.org/10.1002/cncy.21661 Young RH, Scully RE (1989) Atypical forms of microglandular hyperplasia of the cervix simulating carcinoma. Am J Surg Pathol 13:50–56 Zhang Q, Kuhn L, Denny LA, De Souza M, Taylor S, Wright TC Jr (2007) Impact of utilizing p16INK4A immunohistochemistry on estimated performance of three cervical cancer screening tests. Int J Cancer 120(2):351–356 zur Hausen H (1977) Human papillomaviruses and their possible role in squamous cell carcinomas. Curr Top Microbiol Immunol 78:1–30 zur Hausen H (2002) Papillomaviruses and cancer: from basic studies to clinical application. Nat Rev Cancer 2(5):342–350 zur Hausen H (2009) Papillomaviruses in the causation of human cancers – a brief historical account. Virology 384 (2):260–265. https://doi.org/10.1016/j.virol.2008.11.046. S0042-6822(08)00772-1 [pii]

6

Carcinoma and Other Tumors of the Cervix Edyta C. Pirog, Thomas C. Wright, Brigitte M. Ronnett, and Robert J. Kurman

Contents Invasive Carcinoma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 316 Squamous Cell Carcinoma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Superficially Invasive Squamous Cell Carcinoma (SISCC) . . . . . . . . . . . . . . . . . . . . . . . . . . . . Invasive SCC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Keratinizing and Nonkeratinizing Carcinoma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Basaloid Carcinoma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Warty (Condylomatous) Carcinoma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Verrucous Carcinoma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Papillary SCC and Squamotransitional Carcinoma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Lymphoepithelioma-Like Carcinoma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

317 317 323 329 335 336 336 337 338

Adenocarcinoma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 338 Prevalence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 339 Pathogenesis and Epidemiologic Risk Factors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 339

Edyta C. Pirog has retired. E. C. Pirog Department of Pathology, Weill Cornell, Medical College and New York Presbyterian Hospital, New York, NY, USA e-mail: [email protected] T. C. Wright Department of Pathology and Cell Biology, Columbia University, New York, NY, USA e-mail: [email protected] B. M. Ronnett (*) Department of Pathology, Division of Gynecologic Pathology, Johns Hopkins University School of Medicine, Baltimore, MD, USA e-mail: [email protected] R. J. Kurman Department of Gynecology, Obstetrics, Pathology and Oncology, Division of Gynecologic Pathology, Johns Hopkins University School of Medicine, Baltimore, MD, USA e-mail: [email protected] # Springer Nature Switzerland AG 2019 R. J. Kurman et al. (eds.), Blaustein’s Pathology of the Female Genital Tract, https://doi.org/10.1007/978-3-319-46334-6_6

315

316

E. C. Pirog et al. Clinical Presentation and Gross Findings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Early Invasive Adenocarcinoma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Endocervical Adenocarcinoma, Usual Type . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Endometrioid Adenocarcinoma of the Cervix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Intestinal Adenocarcinoma of the Cervix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Villoglandular Adenocarcinoma of the Cervix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Gastric Adenocarcinoma of the Cervix, Including Minimal Deviation Adenocarcinoma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Signet Ring Cell-Type Adenocarcinoma of the Cervix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Clear Cell Adenocarcinoma of the Cervix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Serous Adenocarcinoma of the Cervix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Mesonephric Carcinoma of the Cervix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

340 340 342 347 349 349 351 356 356 357 358

Other Epithelial Tumors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Adenosquamous Carcinoma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Glassy Cell Carcinoma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Adenoid Basal Tumors, Including Epithelioma and Carcinoma . . . . . . . . . . . . . . . . . . . . . . . Adenoid Cystic Carcinoma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Mucoepidermoid Carcinoma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Neuroendocrine Tumors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Pathologic Findings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Immunohistochemical and Molecular Genetic Findings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Differential Diagnosis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Clinical Behavior and Treatment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

359 359 360 361 363 363 364 366 366 366 367

Mesenchymal and Mixed Epithelial-Mesenchymal Tumors . . . . . . . . . . . . . . . . . . . . . . . . 367 Miscellaneous Tumors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 368 Secondary Tumors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 369 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 369

Invasive Carcinoma The World Health Organization (WHO) 2014 classification recognizes three general categories of invasive carcinoma of the cervix: squamous cell carcinoma (SCC), adenocarcinoma, and “other epithelial tumors” (Table 1) (Kurman et al. 2014). The “other epithelial tumors” include adenosquamous carcinomas, adenoid basal and adenoid cystic carcinomas, undifferentiated carcinoma, as well as neuroendocrine tumors. The relative frequency of these different tumor types varies between countries; in general, SCC is the most common histologic subtype accounting for 76–89% of invasive carcinomas. Adenocarcinoma and adenosquamous carcinoma comprise 10–24% of cervical cancers, and all other categories are relatively rare, adding up to less than 5% (de Sanjose et al. 2010).

The most widely accepted staging system for tumors of the cervix is the four-stage system of the International Federation of Gynecology and Obstetrics (FIGO) (Table 2). Stage I includes tumors confined to the cervix and is divided into two subcategories: those that invade 5 mm or less into the stroma (and also are no larger than 7.0 mm in horizontal extent) and are macroscopically not visible (stage IA) and those that either invade more than 5 mm or are macroscopically visible (stage IB). Beyond stage I, staging of cervical cancer is based on clinical examination and imaging. Stage II tumors extend outside the cervix, but not to the pelvic sidewall, and do not invade the lower third of the vagina. Stage III tumors include those that extend to the pelvic sidewall, cause hydronephrosis, or invade the lower third of the vagina. Stage IV tumors extend beyond the true pelvis or clinically involve the mucosa of the bladder or rectum.

6

Carcinoma and Other Tumors of the Cervix

Table 1 Modified WHO histological classification of invasive carcinomas of the uterine cervix

317 Table 2 2009 modification of FIGO staging of carcinoma of the uterine cervix Stage I

Squamous cell carcinoma Keratinizing Nonkeratinizing Basaloid Warty Verrucous Papillary Squamotransitional Lymphoepithelioma-like carcinoma Adenocarcinoma Endocervical, usual-type adenocarcinoma Endometrioid adenocarcinoma Villoglandular adenocarcinoma Gastric adenocarcinoma Intestinal adenocarcinoma Signet ring cell adenocarcinoma Clear cell adenocarcinoma Serous adenocarcinoma Mesonephric adenocarcinoma Other epithelial tumors Adenosquamous carcinoma Glassy cell carcinoma Adenoid basal carcinoma Adenoid cystic carcinoma Undifferentiated carcinoma Neuroendocrine tumors Carcinoid Atypical carcinoid Small cell carcinoma Large cell neuroendocrine carcinoma

IA

IA1 IA2

IB IB1 IB2 II IIA IIA1 IIA2 IIB III

IIIA IIIB IV

Squamous Cell Carcinoma Superficially Invasive Squamous Cell Carcinoma (SISCC) The Lower Anogenital Squamous Terminology (LAST) Standardization Project for HPV-Associated Lesions introduced a new term – “superficially invasive squamous cell carcinoma” (SISCC) – for microscopic, clinically unapparent cervical SCC (Darragh et al. 2012). SISCC is defined as a lesion that is not grossly visible, has

IVA IVB a

Definition Cervical carcinoma confined to uterus (extension to the corpus should be disregarded) Invasive carcinoma diagnosed only by microscopy; all macroscopically visible lesions, even with superficial invasion, are stage IB Stromal invasion no greater than 3.0 mm in depth and 7.0 mm or less in horizontal spread Stromal invasion more than 3.0 mm and not more than 5.0 mm with a horizontal spread of 7.0 mm or lessa Clinically visible lesion confined to the cervix or microscopic lesion greater than IA2 Clinically visible lesion 4.0 cm or less in greatest dimension Clinically visible lesion more than 4.0 cm in greatest dimension Tumor invades beyond the uterus but not to pelvic wall or to lower third of the vagina Without parametrial invasion Clinically visible lesion 4.0 cm in greatest dimension Clinically visible lesion >4 cm in greatest dimension With parametrial invasion Tumor extends to the pelvic wall and/or involves lower third of vagina and/or causes hydronephrosis or nonfunctioning kidneyb Tumor involves lower third of vagina with no extension to pelvic wall Tumor extends to pelvic wall and/or causes hydronephrosis or nonfunctioning kidney The carcinoma has extended beyond the true pelvis or has involved (biopsy proven) the mucosa of the bladder or rectum. A bullous edema, as such, does not permit a case to be allotted to stage IV Spread of the growth to adjacent organs Spread to distant organs

The depth of invasion should not be more than 5 mm taken from the base of the epithelium, either surface or glandular, from which it originates. Vascular space involvement, venous or lymphatic, does not affect classification b On rectal examination, there is no cancer-free space between the tumor and the pelvic wall. All cases with hydronephrosis or nonfunctioning kidney are included, unless they are known to be due to another cause

318

an invasive depth of less than or equal to 3 mm from the basement membrane of the point of origin, and has a horizontal extent of less than or equal to 7 mm in maximum dimension, with margins negative for carcinoma. The presence of HSIL at the margin does not exclude a tumor from this category but should be noted. The presence of lymphatic invasion and whether there is multifocal invasion should be also included in the report (Darragh et al. 2012). SISCC represents the earliest phase of invasion from the background of highgrade squamous intraepithelial lesion (HSIL), and being an early lesion, it has a favorable prognosis. In the past, early invasive carcinoma was termed “microinvasive carcinoma” (MICA). The Society of Gynecologic Oncology (SGO) in the United States defined it as a microscopic tumor that invaded to a depth of no greater than 3 mm beyond the base of the epithelium of origin, either surface or glandular, with no criterion for lesion width. Lymphatic invasion was not allowed, and margins were required to be negative for carcinoma (a requirement for the margin status with regard to the presence of squamous intraepithelial lesion (SIL) was not specified). The FIGO had its own criteria and subdivided MICA into those that invaded no more than 3 mm in depth (stage IA1) and those that invaded more than 3 mm but not more than 5 mm in depth (stage IA2) and specified that horizontal extent could not exceed 7 mm (see Table 2). The presence of lymphvascular invasion did not exclude a tumor from FIGO stage IA (Berek and Hacker 2010). Due to the variability in definitions of MICA, this terminology is not recommended, and the term SISCC as proposed by the LAST Project is currently endorsed. Clinical evidence demonstrates that lesions categorized as SISCC and without lymph-vascular invasion have an extremely low rate of lymph node metastases, recurrences, or risk of death. Such patients may be offered conservative treatment, usually conization or trachelectomy, while those with tumors measuring >3 mm in depth or with lymphatic invasion should be considered for more radical therapy (see subsection “Treatment”) (Berek and Hacker 2010; Eskander et al. 2015). Single institution studies report a rate of SISCC as 3.6–4.7% in patients treated with cone excisions after biopsy diagnosis of HSIL (Killackey et al. 1986; Matseoane et al. 1992).

E. C. Pirog et al.

Clinical Features The average age of patients with SISCC is 39–42 years. Most patients with SISCC are asymptomatic with a grossly normal cervix or “friable cervix” with abnormal capillaries prone to bleeding. Neither cytologic results nor colposcopic exam can accurately predict the presence of early invasion. In cytology, abnormally prominent nucleoli may be seen in dysplastic cells, suggesting an invasive process. However, the positive predictive value of cytology for the presence of early invasion is reported at just 27.3% (Andersen et al. 1995). In colposcopic examination, the early invasion is typically seen within an acetowhite area, consistent with the background of HSIL. Within this area an abnormal vascular pattern is observed consisting of vessels with irregular and haphazard distribution, increased intercapillary distance, marked variations of caliber, and abrupt changes in direction forming acute angles. The sensitivity of colposcopic exam for identification of SISCC is reported at 30–50%. As expected, colposcopy predicts invasion more accurately with increasing depth of the invasive lesion (Berek and Hacker 2010). Pathologic Features The diagnosis of SISCC is based on identification of neoplastic squamous cells extending from HSIL into the underlying stroma. Superficial invasion may be a unifocal or multifocal process. The overlying squamous epithelium typically demonstrates extensive presence of HSIL, and in most instances, the underlying endocervical glands are also replaced by the intraepithelial lesion. Commonly, the foci of superficial invasion are seen as irregular tongues or buds that show a loss of the palisade-like cell arrangement typical of the epithelium at the epithelial-stromal junction. The invading cells are large and characterized by presence of abundant eosinophilic cytoplasm, low nuclear to cytoplasmic ratio, and nuclei with pale chromatin and prominent nucleoli (Figs. 1 and 2). This appearance is referred to as “paradoxical maturation” or “pseudomaturation” and contrasts with the background of HSIL that demonstrates immature, basaloid cell morphology, higher nuclear to cytoplasmic ratio, and nuclei with

6

Carcinoma and Other Tumors of the Cervix

Fig. 1 Superficially invasive focus showing neoplastic epithelium transversing the basement membrane of HSIL. The invasive cells show “pseudomaturation” with abundant eosinophilic cytoplasm and lower nuclear to cytoplasmic ratio

Fig. 2 SISCC. Invasive nests showing irregular, angulated contours. Central keratin pearls are present

coarse, granular chromatin. Occasionally, small areas of keratinization may be seen within the microinvasive foci. Because of focal disruption of the basement membrane, the margins of the invading nests are ragged, flanked by intact basement membrane on either side. The irregular contours of the tumor nests at the invasive tumor front are the most reliable criteria for the diagnosis of early invasion. Typically, there is also an inflammatory and stromal reaction in the area of invasion. The inflammatory response consists of a conspicuous lymphoplasmacytic infiltrate surrounding the tips of the invasive epithelial tongues. The stromal reaction presents as a

319

Fig. 3 SISCC. The tumor is composed of invasive nests with rounded contours, mimicking endocervical gland involvement by HSIL; however, the spacing is much denser than normal endocervical crypt spacing, and the nests are elongated and focally branching

desmoplastic response, in addition to stromal edema and, in some cases, capillary angiogenesis. Rarely, the invasive process may present as smooth and rounded contoured tumor nests infiltrating the stroma and mimicking gland involvement by HSIL. In such cases the invasive tongues and round to oval tumor nests are markedly crowded and tightly spaced as compared to the spacing of the adjacent benign endocervical glands (Fig. 3). The invasive nests may also extend beneath the level of normal endocervical crypts. Careful comparison with the architecture and layout of adjacent endocervical glands is paramount in diagnosing this deceptive pattern of invasion. The findings in HSIL associated with, or predictive of, superficial invasion include extensive involvement of surface epithelium by HSIL, extension of HSIL to deep endocervical crypts, lumenal necrosis, and intraepithelial squamous pseudomaturation. When such findings are present in a biopsy or cone specimen, extra attention should be paid to the possibility of invasion, and deeper sections may be obtained to thoroughly evaluate any suspicious areas. Ultrastructural electron microscopy study of early invasion in cervical SCC has demonstrated that there is a disappearance of basal lamina of the basement membrane and pseudopod-like cytoplasmic protrusions of the cancer cells are seen in a direct contact with the underlying stroma

320

(Kudo et al. 1990). The protrusions of the invading neoplastic cells contain abundant cytoplasmic vesicles, 70–90 nm in size, some of which are seen open directly into the extracellular matrix of the stroma. These vesicles are not observed in adjacent HSIL, which is supported by the intact basement membrane. These findings suggest that the substances contained in the vesicles may play a role in basement membrane destruction. In addition, the neoplastic cells traversing the basement membrane gaps show accumulation of actin filaments in the pseudopod protrusions. These local aggregates of cytoskeletal structures, not observed in adjacent HSIL, are thought to facilitate an ameba-like movement of the cancer cells. Finally, neoplastic cells in the areas of superficial invasion show decreased number of desmosomal junctions, enabling cellular discohesion and migration (Kudo et al. 1990). Subsequent immunohistochemical studies using antibodies directed against basement membrane constituents, such as laminin or type IV collagen, enhanced the recognition of early stromal invasion in squamous cervical lesions. These studies have demonstrated that normal squamous epithelium and SILs are supported by continuous, intact basement membrane with only occasional small basement membrane disruptions in areas with severe inflammatory reaction. Positive immunostaining for basement membrane is also observed in a proportion of invasive carcinomas, with more extensive staining seen in well-differentiated tumors. In addition, cases of metastatic SCC in the lymph nodes surrounded by basement membrane were reported (Antonelli et al. 1991). Since invasion and metastasis require breakage of the basement membrane, Liotta proposed a plausible explanation for these seemingly inconsistent findings (Liotta 1984). According to this hypothesis, which was subsequently validated by experimental studies (Antonelli et al. 1991), the cancer nests proceed through cycles of growth surge with basement membrane destruction and stromal invasion, followed by quiescence and basement membrane formation. During the quiescent phase, basement membrane may remain intact until a new surge of growth, during which it is focally dissolved allowing the tumor to bud out. To visualize this process better, a more complex approach was developed

E. C. Pirog et al.

Fig. 4 SISCC, double immunostaining for cytokeratin (red) and collagen IV (brown). Discrete basement membrane is seen underneath HSIL. The microinvasive tumor nests in the stroma lack the basement membrane cuff

with immunohistochemistry utilizing double immunostaining for cytokeratin and basement membrane components (Fig. 4) (Rush et al. 2005). With this method, the nests of cytokeratin-positive tumor cells are seen traversing the basement membrane at the invasive tumor fronts in invasive and superficially invasive cervical carcinomas.

Measurement and Significance of Depth of Invasion The measurement of the depth of stromal invasion may be difficult and the following guidelines are recommended. The depth of neoplastic projections should be measured from the initial site of invasion, either from the basal lamina of the surface epithelium or from endocervical glands replaced by the squamous intraepithelial lesion (Fig. 5). There are cases, however, in which direct continuity between invasive foci and SIL cannot be demonstrated, even in deeper levels of the paraffin block. In such instances, it is assumed that the invasion originated from basal cells of the overlying SIL. Therefore, the depth of invasion is arbitrarily measured from the basal lamina of the surface SIL. The depth of invasion also depends on the angle at which sections are prepared, and therefore efforts should be made to secure vertically sectioned tissue samples. Depth of stromal invasion is a major factor in determining the outcome of patients with SISCC with respect to risk of lymph node metastasis, risk

6

Carcinoma and Other Tumors of the Cervix

Table 4 Residual invasive tumor in postconization hysterectomy specimens according to lateral extent of carcinoma showing up to 5 mm depth of invasion

HSIL

a

b

c

Fig. 5 Methods of measuring depth of invasion in carcinoma of the cervix. The pattern of stromal invasion determines the stromal depth measurement that is most appropriate. (a) Origin of invasion from surface HSIL: depth of stromal invasion is measured from point of origin of invasion downward to the last cell of the invasive focus. (b) Origin of invasion from HSIL with gland involvement: depth of stromal invasion is measured from site of origin downward to the last cell of the invasive focus. (c) Origin of invasion not seen: depth of stromal invasion is measured from basal lamina of surface HSIL downward to the last cell of the invasive focus Table 3 Percentage of pelvic node metastases, recurrence, and death from disease in relation to depth of invasion in early invasive SCC. Number of reported patients shown in parenthesis

Lymph node metastasis Recurrence

Death from disease

321

Depth of cervical stromal invasion IB1, computed tomography (CT) or

E. C. Pirog et al.

magnetic resonance imaging (MRI) of the abdomen and pelvis is typically done to identify metastases, estimate tumor volume, and determine parametrial involvement. If MRI and CT are not available, cystoscopy, sigmoidoscopy, and IV urography, when clinically indicated, may be used for staging. Cervical cancer is the most common gynecologic cancer in pregnancy, but the incidence is low (1.5–12 per 100,000 pregnancies) (Jones et al. 1996). The mean age of pregnant women with invasive cervical cancer is 32 years, which is considerably lower than that of women in the general population with invasive cervical cancer. Women who are pregnant usually present with early-stage tumors; in one series 83% were stage I (Jones et al. 1996). Most women present with abnormal Pap findings as cervical cytology exam is a routine part of the initial prenatal evaluation. One-third of patients are diagnosed during first, second, and third trimester each, respectively (Jessup et al. 1996). The treatment of pregnant patients depends on the clinical stage and gestational age. Approximately 37% of patients carry the pregnancy to fetal maturity (Jessup et al. 1996). In general, prognosis is not altered by pregnancy.

Gross Findings The gross appearance of invasive SCC varies widely. Early lesions may be focally indurated, ulcerated, or present as a slightly elevated, granular area that bleeds readily. Colposcopic examination usually reveals atypical, tortuous vessels varying widely in size and configuration. Approximately 98% of early carcinomas are localized within the transformation zone, with variable degrees of encroachment onto the neighboring native portio. Advanced tumors are endophytic or exophytic. Endophytic carcinomas are ulcerated or nodular; they tend to develop within the endocervical canal and frequently invade deeply into the cervical stroma to produce an enlarged, hard, barrel-shaped cervix. In some patients with endophytic carcinomas, the cervix appears grossly normal. The exophytic varieties of cervical carcinoma have a polypoid or papillary, friable appearance.

6

Carcinoma and Other Tumors of the Cervix

Histologic Typing The current WHO classification divides invasive SCCs into two groups, keratinizing and nonkeratinizing (Kurman et al. 2014). In addition, it lists separately rare histologic variants of SCC, namely, basaloid, warty, verrucous, papillary, squamotransitional, and lymphoepithelioma-like carcinoma (LELC). The most common tumor types are keratinizing and nonkeratinizing, and clinical references pertaining to cervical carcinoma refer mainly to these two tumor types.

329

Microscopic Findings Microscopically, invasive SCC is characterized by anastomosing tongues or nests of neoplastic squamous epithelium infiltrating the stroma (Fig. 7). Characteristically, the contour of the infiltrating nests and clusters is irregular and ragged. In some cases, the tumor invades either as individual cells or almost completely replaces the stroma with large masses of neoplastic squamous cells. Cells in the center of the invading nests frequently become necrotic or undergo extensive keratinization. Individual cells are generally polygonal or oval with eosinophilic cytoplasm and prominent cellular membranes. Intracellular

bridges may or may not be visible. The nuclei may be relatively uniform or quite pleomorphic. In most cases, the chromatin is coarse and clumped, and mitotic figures, including abnormal forms, commonly are encountered. Keratinizing carcinomas are characterized by the presence of well-differentiated squamous cells that are arranged in nests that vary in size and configuration. The defining feature of keratinizing carcinomas is the presence of keratin pearls within the epithelium (Fig. 7). Keratin pearls are composed of clusters of squamous cells that have undergone keratinization and are arranged in a concentric nest. The neoplastic squamous cells not forming keratin pearls frequently have abundant eosinophilic cytoplasm and prominent intracellular bridges. The nuclei are often enlarged, but mitotic figures are not numerous. Nonkeratinizing SCC is characterized by nests of neoplastic squamous cells that frequently undergo individual cell keratinization but, by definition, do not form keratin. The cells have relatively indistinct cell borders. The nuclei tend to be round to oval with either prominent nucleoli or coarsely clumped chromatin. Mitotic figures are numerous. In poorly differentiated nonkeratinizing tumors, squamous differentiation may be difficult to ascertain (Fig. 8). The histologic subtype of

Fig. 7 Keratinizing SCC of the cervix, well differentiated, composed of islands and nests of neoplastic squamous epithelium with central keratin pearls

Fig. 8 Nonkeratinizing SCC of the cervix, poorly differentiated, composed of relatively small cells with high nuclear to cytoplasmic ratio, indistinct cytoplasm, and large, markedly atypical nuclei. Numerous mitoses are present

Keratinizing and Nonkeratinizing Carcinoma

330

Fig. 9 SCC with clear cell morphology. Sheets of carcinoma composed of polygonal cells with clear, glycogenated cytoplasm; this tumor variant is solid, without glandular architectural features of clear cell carcinoma

SCC has no prognostic significance in relation to predicting nodal spread or progression-free interval (Zaino et al. 1992). In some cases SCC is composed of solid sheets of cells with clear cytoplasm (Fig. 9). If the tumor is composed mostly of these highly glycogenated tumor cells, it may be diagnosed as SCC with the comment describing that the clear cell morphology of the cells is interpreted as a variant of SCC and not as a form of clear cell carcinoma. Rare nonkeratinizing SCCs assume a spindleshaped configuration resembling spindle cell squamous cancers of the larynx (Fig. 10). Immunohistochemical staining for epithelial membrane antigen and cytokeratins demonstrates the epithelial nature of the spindle cells in these cases. In addition to these variants, there are reports of rare cases of highly keratinized variant of squamous cell carcinomas of the cervix. These tumors are negative with HPV testing and resemble HPV-negative, keratinizing vulvar cancers. The sections show extensive keratin formation; an infiltrative, destructive pattern of growth; and only minimal cytologic atypia (Fig. 11). There is extensive hyperkeratosis and parakeratosis adjacent to the tumor but no evidence of squamous intraepithelial lesion (Morrison et al. 2001).

E. C. Pirog et al.

Fig. 10 SCC, spindle cell variant. Poorly differentiated carcinoma shows cells with spindled morphology

Fig. 11 Keratinizing SCC, well differentiated. Rare example of HPV-negative tumor with only mild cytologic atypia and infiltrative, destructive growth pattern

Tumor Grading The most commonly used grading system for SCC divides tumors into three groups, well differentiated (grade 1), moderately differentiated (grade 2), and poorly differentiated (grade 3). Most SCCs are moderately differentiated (grade 2), followed by poorly differentiated (grade 3) and well differentiated (grade 1). Tumors are graded as well differentiated (grade 1) when the cells appear mature, with abundant eosinophilic cytoplasm (Fig. 7). Typically, there

6

Carcinoma and Other Tumors of the Cervix

are keratin pearls in the center of neoplastic nests. Individual cell keratinization (dyskeratosis) characterized by intense cytoplasmic eosinophilia may also be present. The cells are tightly packed and have well-developed intercellular bridges. The nuclei are large, irregular, and hyperchromatic. Mitotic figures are present most notably at the periphery of the advancing epithelial nests. The stroma is often infiltrated by chronic inflammatory cells, and occasionally a foreign body giant cell reaction is observed. Tumors are graded as moderately differentiated (grade 2) SCCs when the neoplastic cells are more pleomorphic than in grade 1 tumors, have large irregular nuclei, and have less abundant cytoplasm. The cellular borders and intercellular bridge appear indistinct. Keratin pearl formation is rare, but individual cell keratinization is seen in the center of nests of tumor cells. Mitotic figures are more numerous than in grade 1 carcinomas. Poorly differentiated (grade 3) SCCs are diagnosed when the tumors are composed of cells showing little, if any, squamous maturation (Fig. 8). The nuclear to cytoplasmic ratio is high, the cytoplasm is scant, and indistinct and the cellular bridges are not seen. Tumor cells have hyperchromatic oval nuclei with coarse chromatin. Mitoses are abundant, and areas of central necrosis in tumor nests are present. Poorly differentiated tumors are occasionally composed of large, highly pleomorphic cells with giant, bizarre nuclei and abnormal mitotic figures. Although in some studies tumor grade was associated with patient’s survival, most studies have failed to confirm that histopathologic grade influences clinical outcome (Zaino et al. 1992). A study by the GOG evaluated a number of different tumor-grading systems including those proposed by Warren, Reagan, and Broder in surgically treated stage IB cervical cancers. Although there was good reproducibility between observers, none of the grading systems had prognostic significance. Nuclear grade, degree of keratinization, mitotic activity, pattern of infiltration, and degree of lymphoid response all lacked prognostic significance (Zaino et al. 1992).

331

Immunohistochemical Staining and HPV In Situ Hybridization SCCs of the cervix show immunohistochemical positivity for p16, manifested as diffuse/strong expression in essentially all tumor cells, in 90–100% of cases (Tringler et al. 2004; Nemejcova et al. 2015). P16 upregulation in cervical cancer is a reflection of cell infection with high-risk HPV and inactivation of pRB through binding to HPV oncoprotein E7 (Fig. 12). Cervical squamous cell tumors are positive for a range of cytokeratins, including CK7, as well as CK 4, 5, 6, 8, 13, 14, 16, 17, 18, and 19 (Smedts et al. 1992). Cervical SCCs are in general negative for PAX-8, in contrast to adenocarcinomas which are positive in 87% of cases (Tacha et al. 2011; Ozcan et al. 2011). Cervical squamous tumors are also positive for p63 and its isoform, p40, in close to 100% of cases (Nemejcova et al. 2015). In a study of 250 invasive carcinomas, a strong, diffuse p63 expression was present in 97% of SCCs, including 91% of small cell nonkeratinizing squamous cell carcinomas. In contrast, small cell neuroendocrine carcinomas either did not show p63 staining or only had focal expression ( PR (some ER+/PR-, some ER+/PR+) p53 wild type = nondiffuse expression (80% of tumor cells) or complete absence (“null”) d p16 varies from limited to extensive but usually is not diffuse (few exceptions) e Suboptimal sensitivity a

b

has diffuse p16 expression. Some endometrial endometrioid carcinomas with prominent mucinous/metaplastic-type differentiation can have more extensive to occasionally diffuse p16 expression, but the staining intensity is usually not as strong as is seen in high-risk HPV-related endocervical adenocarcinomas, and some negative patches are usually present if the sample is not too small (Ronnett, unpublished observations). Endometrioid carcinomas typically express hormone receptors (both ER and PR), but some tumors, particularly but not exclusively highgrade carcinomas, can lose expression. Therefore, analysis of p16 expression, alone or in combination with ER/PR, distinguishes high-risk HPV-related endocervical adenocarcinomas from low-grade endometrial endometrioid carcinomas in most cases (Yemelyanova et al. 2009). Some pathologists find CEA and vimentin to be of some value in the distinction of high-risk HPV-related (usual-type) endocervical adenocarcinomas from endometrial endometrioid carcinomas, but others do not. Most endometrial endometrioid carcinomas are vimentin-positive and CEA-negative, and most HPV-related usualtype endocervical adenocarcinomas are vimentinnegative and CEA-positive. However, in practice use of these markers is problematic for several reasons: (1) expression of these markers can be focal; (2) CEA staining can be difficult to interpret because squamous elements (commonly seen in endometrial endometrioid adenocarcinoma) can be positive, and in addition, there can be apical/ glycocalyceal staining in endometrioid

adenocarcinomas, while some usual-type endocervical adenocarcinomas are negative; (3) it may be difficult to ascertain whether vimentin expression is actually within glands versus in closely apposed stroma; and (4) tumors with mucinous differentiation usually express CEA, regardless of their origin, and may be vimentinnegative. Expression of p16 is also useful for identifying endocervical adenocarcinoma metastatic to the ovary. These ovarian metastases have a propensity to simulate primary ovarian endometrioid and mucinous tumors (atypical proliferative (borderline) tumors and carcinomas). Primary ovarian endometrioid and mucinous tumors, with few exceptions, are characterized by generally patchy p16 expression (or lack of expression), whereas metastatic high-risk HPV-related endocervical adenocarcinomas are diffusely/strongly positive. In the absence of a known primary endocervical adenocarcinoma, demonstrating high-risk HPV by in situ hybridization in metastatic lesions can be used for arriving at a definitive diagnosis (Fig. 24). Metastatic adenocarcinoma to the cervix usually occurs in the setting of a patient with a known, widely metastatic primary lesion and is histologically characterized by a lack of surface involvement and widespread lymph-vascular involvement. In assessing whether a carcinoma is of primary endocervical origin or is metastatic in the cervix, the pathologist should evaluate the following morphologic features: (1) neoplastic growth pattern, (2) coexistent in situ changes, (3) cell type, and (4) immunohistochemical characteristics. Transition

6

Carcinoma and Other Tumors of the Cervix

between in situ and invasive carcinoma provides evidence for a primary origin and is found in approximately 50% of primary cervical adenocarcinomas. Pax-8 positivity may be used to confirm primary müllerian origin of the tumor.

Prognostic Risk Factors A recent study (Roma et al. 2016) found a correlation between the histologic pattern of invasion in usual-type adenocarcinoma and risk of lymph node metastasis and recurrences. In that study the tumors were subclassified into three groups based on their pattern of growth (Silva classification). The tumors with well-demarcated glands and lacking destructive stromal invasion or lymphatic invasion were classified as pattern A. Tumors with diffuse destructive stromal infiltration or solid tumors were classified as pattern C. Tumors with early destructive invasion from well-demarcated glands were classified as pattern B. The study consisted of 352 cases, and all pattern A and B tumors as well as 83% of pattern C tumors were stage I. No lymph node metastases or recurrences were seen in any of pattern A cases regardless of their stage (IA1, IA2, or IB) or depth of invasion. Metastases and recurrences were recorded in 4.4% and 1.2% of tumors with pattern B and 23.8% and 22.1% of tumors with pattern C, respectively. Based on these results, the authors proposed that patients with pattern A growth might be spared lymph node dissection, patients with pattern B growth could be considered for sentinel node sampling, and patients with pattern C were the only group requiring lymph node dissection (Roma et al. 2016). Other reported prognostic indicators for cervical adenocarcinoma include tumor size, depth of invasion, involvement of lymph-vascular spaces, parametrial involvement, stage, age, and presence or absence of lymph node metastases (Eifel et al. 1990). Clinical Behavior and Treatment Adenocarcinoma of the cervix spreads in a fashion similar to SCC, and, in general, both squamous and adenocarcinomas are treated similarly. The most commonly used therapeutic modalities for stage I and II adenocarcinoma are radiation alone,

347

radiation with concurrent chemotherapy, radiation followed by simple hysterectomy, or radical surgery. Only a few studies have directly compared the therapeutic results achieved with invasive squamous cell and adenocarcinomas over the same time period and from the same institution; and these studies have produced conflicting data. Some studies have found that the overall 5-year survival rates are lower for adenocarcinoma (48–65%) than for SCC patients (68%) (Eifel et al. 1990). Other comparison studies, as well as several population-based studies, have failed to confirm that prognosis and survival are affected by histologic type (Anton-Culver et al. 1992). Therefore, the prognosis of cervical adenocarcinoma relative to squamous carcinoma remains a controversial issue.

Endometrioid Adenocarcinoma of the Cervix General Features It is not clear whether endocervical adenocarcinomas with endometrioid-type differentiation are truly a distinct entity. Tumors with this appearance in which high-risk HPV is detected are most consistent with mucin-depleted forms of usual-type endocervical adenocarcinoma and are thus not a distinct entity. According to the literature on tumors reported as this type, the average age of patients is 50 years (Pirog et al. 2000). In different series, endometrioid carcinomas accounted for 7–50% of endocervical adenocarcinomas (Young and Clement 2002). Studies have also reported that HPV DNA is identified in 78–100% of endometrioid adenocarcinomas that arise from the cervical squamocolumnar junction (Pirog et al. 2000; Jones et al. 2013). However, endometrioid adenocarcinomas arising from upper endocervix and lower uterine segment are typically HPV-negative (Holl et al. 2015). These observations indicate that there are problems in classification related to difficulty in the distinction of true primary endocervical adenocarcinomas of usual type with mucin depletion from true primary endometrial endometrioid carcinomas. Use of ancillary techniques (discussed above) can

348

E. C. Pirog et al.

resolve this problem. Most experts agree that true endometrioid adenocarcinomas of the cervix are rare, and only tumors with scant eosinophilic cytoplasm without apparent intracytoplasmic mucin should be classified as endometrioid (Young and Clement 2002).

Microscopic Findings The tumor cells are characterized by lack of mucin and scant, deeply eosinophilic cytoplasm, resembling endometrial-type epithelium (Fig. 29). Architecturally, these tumors typically grow as small, round, or tubular glands with either smooth luminal contours or intraluminal cribriform growth. Some tumors show papillary architecture with thick fibrovascular cores that are either exophytic or infiltrate the cervical wall. Cervical endometrioid adenocarcinoma of minimal deviation type has been described (Young and Scully 1993). Only few case reports are available. The reported cases occurred in women 32–64 years of age. The tumor is composed of bland, tubular, or cystically dilated endometrial-type glands infiltrating the cervical wall with no, or only minimal, stromal response (Fig. 30). The cytologic atypia is minimal and mitoses are rare.

Fig. 29 Endocervical adenocarcinoma of usual type with endometrioid features. Crowded glands have columnar cells with elongated nuclei and little apical cytoplasm, resembling a primary endometrial endometrioid carcinoma (detection of high-risk HPV established this as primary endocervical adenocarcinoma)

Immunohistochemical Staining The pattern of immunostaining of cervical endometrioid adenocarcinoma is similar to that of usual-type adenocarcinoma, with tumor cells demonstrating strong and diffuse p16 positivity and often negative staining for hormone receptors (Jones et al. 2013).

Fig. 30 Endometrioid adenocarcinoma of the cervix, minimal deviation type. The tumor is composed of tubular glands with minimal cytologic atypia

Differential Diagnosis Differentiating cervical from uterine endometrioid adenocarcinoma may require use of immunostains (ER-/PR-/diffuse p16+ result supports the diagnosis of cervical primary). Rare cases may show mixed/overlapping immunostaining patterns. Careful assessment of imaging studies may be of help. Primary uterine corpus tumors are usually bulky tumors that have invaded the myometrium by the time they extend to the cervix and therefore cause uterine enlargement. In contrast, primary cervical adenocarcinomas often cause cervical

enlargement in the absence of uterine enlargement. In addition, multiple sections may reveal atypical endometrial hyperplasia in primary endometrial tumors and either adenocarcinoma in situ or HSIL or foci with features of a typical endocervical carcinoma in primary endocervical tumors. In certain cases establishing the correct diagnosis may require knowing the HPV status of the tumor. If a tumor appears endometrioid and is truly HPV-negative and present in the lower uterine segment, then it is likely truly endometrial.

6

Carcinoma and Other Tumors of the Cervix

349

Intestinal Adenocarcinoma of the Cervix

lesions is stronger than the staining for enteric markers (Park et al. 2009).

General Features The tumor accounts for approximately 8% of cervical adenocarcinomas. The average age of patients is 47 years (range 26–69) (Pirog et al. 2000). HPV DNA is identified in 83% of tumors (Pirog et al. 2000). The precursor lesion is adenocarcinoma in situ, intestinal type. Recent reports have described cases of intestinal adenocarcinoma in situ which were HPV-negative and occurred in older patients (Talia et al. 2014).

Differential Diagnosis The differential diagnosis includes direct extension or metastasis from colonic adenocarcinoma. The presence of large, garland-shaped glands with incomplete lining and intraluminal necrosis is highly suggestive of a metastasis from the gastrointestinal tract. In equivocal cases immunohistochemistry should be used; negative immunostaining for PAX-8 and CK7 and positive enteric markers are indicative of spread from colonic adenocarcinoma.

Microscopic Findings The intestinal-type adenocarcinoma is composed of cells similar to those present in adenocarcinomas of the large intestine (Fig. 31). These tumors frequently contain goblet cells and more rarely argentaffin cells and Paneth cells. They can either form glands with papillae or infiltrate throughout the stroma in a pattern similar to that of colonic adenocarcinoma. Immunohistochemical Staining Intestinal adenocarcinoma displays immunohistochemical profile similar to adenocarcinoma of usual type (Saad et al. 2009). In addition, the malignancy may show an enteric immunophenotype with at least focal positivity for CDX2 and/or CK20 (Saad et al. 2009; Park et al. 2009). It has been observed, however, that CK7 positivity in these

Fig. 31 Intestinal adenocarcinoma of the cervix. The tumor is composed of glands with goblet cells

Villoglandular Adenocarcinoma of the Cervix General Features Villoglandular adenocarcinoma is a welldifferentiated variant of endocervical, endometrioid, or intestinal adenocarcinoma that occurs predominantly in young women and has an excellent prognosis (Jones et al. 1993b). The tumor accounts for approximately 3–6% of all adenocarcinomas. The entity was established as a separate diagnostic category based on a prognosis that is more favorable than that of usual-type endocervical adenocarcinoma. The average age of patients is reported at 33–41 years (range 21–61) with most patients younger than 40 years (Jones et al. 1993b). HPV DNA is identified in 100% of villoglandular adenocarcinomas (Jones et al. 2000). Microscopic Findings The characteristic features of this tumor are a surface component that is composed of papillae lined by epithelium that has only mild cytologic atypia (Fig. 32). The epithelial cells lining the papillae can display endocervical, endometrioid, or intestinal features (Figs. 33 and 34). Because of the large number of surface papillae, these tumors frequently form an exophytic, friable tumor mass. Most of the papillae have central cores containing spindle-shaped stromal cells resembling those of the normal cervical stroma and a variable number of inflammatory cells.

350

E. C. Pirog et al.

The papillae can be either long and thin or thick and short. Small papillary tufts composed entirely of epithelial cells of the type characteristic of serous carcinomas of the ovary are absent. Beneath the papillary surface, the infiltrating portion of the tumor is composed of irregular branching glands that are typically surrounded by only a minimal desmoplastic response. In the majority of cases, the tumor is superficially invasive, although deep invasion with extension into the uterine corpus may occur. The pattern of immunostaining is similar to that of usual-type endocervical adenocarcinoma (Jones et al. 2000). Fig. 32 Villoglandular adenocarcinoma. Low-power appearance of thin, elongated villous papillary structures

Fig. 33 Villoglandular adenocarcinoma, intestinal type. Epithelium lining papillae exhibit intestinal mucinous and goblet cell differentiation

Fig. 34 Villoglandular adenocarcinoma, endocervical type. Epithelium lining papillae exhibit typical cytologic features of usual-type endocervical adenocarcinoma

Differential Diagnosis The diagnosis of villoglandular adenocarcinoma may pose a challenge on biopsy. The differential diagnosis of this tumor includes papillary endocervicitis, papillary adenofibromas of the cervix, and müllerian papillomas. All three of these lesions lack the degree of cellular atypia and mitotic activity that is apparent in villoglandular adenocarcinomas. The müllerian papilloma is a lesion of children, whereas the average age of patients with villoglandular adenocarcinomas is in the fourth and fifth decade. In contrast to villoglandular adenocarcinoma, adenofibroma of the cervix and müllerian papilloma have a more prominent stromal component. Villoglandular adenocarcinoma must also be distinguished from usual adenocarcinoma with papillary architecture. Usual endocervical adenocarcinoma demonstrates irregular, thick fibrovascular cores and higher degree of cytologic atypia. Examination of multiple sections typically uncovers the conventional appearance with infiltrative glands or cribriform growth. While the rare serous carcinoma of the cervix is theoretically in the differential diagnosis, some doubt its existence and believe such tumors are more likely secondary to a serous carcinoma of endometrial or tubal/ovarian origin. Immunohistochemical analysis of p53 expression is useful for this rare situation, as villoglandular adenocarcinoma has non-aberrant (heterogeneous) p53 expression similar to usual-type endocervical adenocarcinoma, whereas serous carcinomas have aberrant p53 expression (either diffuse overexpression or complete absence of expression (“null”)).

6

Carcinoma and Other Tumors of the Cervix

Prognosis and Treatment In most of the cases published to date, the clinical outcome of patients with villoglandular adenocarcinomas has been excellent. In the two largest series, all patients, including those who were treated by cone biopsy, were alive and well with no evidence of recurrent disease after 7–77 months of follow-up (Jones et al. 2000). Conservative treatment should be considered only if the tumor is superficial and does not involve lymph-vascular spaces and there is no disease on the cone margins. In a series of cases from Japan, the presence of lymph-vascular space involvement was associated with lymph node metastases (Kaku et al. 1997). Rarely, villoglandular carcinomas may be mixed with other types of carcinomas. The authors are aware of two such cases in which lymph node metastasis has occurred (unpublished observations).

Gastric Adenocarcinoma of the Cervix, Including Minimal Deviation Adenocarcinoma General Features Gastric adenocarcinoma of the cervix is a recently described diagnostic entity (Kojima et al. 2007) that was included in the updated WHO 2014 tumor classification. The tumor spans a spectrum of morphologic appearances with the common feature of expression of gastric-type mucins. Minimal deviation adenocarcinoma with mucinous morphology was identified to be a part of that spectrum, and this tumor is now considered to represent a well-differentiated form of gastric adenocarcinoma (Kojima et al. 2007). The true incidence of gastric adenocarcinoma is not currently known as this lesion has been previously misclassified as usual, intestinal, or clear cell adenocarcinoma, but it is thought to be rare in Western countries, while in Japan it may account for over 20% of all cervical adenocarcinomas (Kojima et al. 2007). The pathogenesis of gastric adenocarcinoma is still being investigated. Notably, gastric adenocarcinoma, including minimal deviation type, is not related to HPV infection (Park et al. 2011). Some cases were reported to

351

be associated with Peutz-Jeghers syndrome, an autosomal dominant disorder caused by germline mutation of STK11, a serine threonine kinase gene (Gilks et al. 1989). In addition, somatic mutations of STK11 are reported in over a half of the sporadic cases (Kuragaki et al. 2003). Mutation of the p53 gene is suspected in a proportion of cases based on mutational-type immunostaining pattern of some tumors (Carleton et al. 2016). The putative precursor lesion is LEGH, a benign proliferation of endocervical glands with gastric phenotype. The average age of the patients is between 42 and 50 years (range 25–84) (Kojima et al. 2007; Karamurzin et al. 2015), which is similar to that of patients with usual adenocarcinoma. Many patients present with profuse watery vaginal discharge or enlarged, barrel-shaped cervix. Cytologic screening has low sensitivity for minimal deviation adenocarcinoma, and HPV testing has no role; therefore many patients present with high-stage tumors. The prognosis for patients with gastric adenocarcinoma, including that of minimal deviation adenocarcinoma, is significantly worse in comparison to usual type (Gilks et al. 1989). The reported overall 5-year disease-specific survival for gastric versus usual adenocarcinoma is 30–42% vs. 74–91%, respectively (Park et al. 2011; Karamurzin et al. 2015). The tumors have a predisposition for metastatic spread to the ovaries, abdominal cavity, and extraperitoneal sites (Karamurzin et al. 2015). In some cases the metastases may deceptively mimic primary ovarian mucinous tumors, including cystadenomas and atypical proliferative/borderline tumors, due to a maturation phenomenon in which the ovarian metastases can be even more differentiated than the primary cervical tumor. Sometimes the primary cervical neoplasm is identified only when staging surgery is performed.

Microscopic Findings Gastric adenocarcinoma is notable for diffuse cervical infiltration resulting in cervical enlargement (barrel cervix) without a distinct gross mass. The invasive glands show marked variation of sizes and shapes – from simple, tubular glands to cystically dilated glands (Fig. 35) or markedly

352

Fig. 35 Gastric adenocarcinoma, minimal deviation type. Low-power appearance of dilated, abnormally shaped, and crowded glands

Fig. 36 Gastric adenocarcinoma, minimal deviation type. Highly differentiated glands on left lack a stromal reaction, whereas deeper glands on right elicit a desmoplastic response at the invasive tumor front

complex and branched glands with intraluminal papillary infoldings. The tumor is usually deeply invasive, and a desmoplastic stromal response surrounding the invasive glands may be seen (Fig. 36). Cytologically, the glands are lined by mucinous epithelium with voluminous, clear, or pale eosinophilic cytoplasm showing distinct cellular borders (Fig. 37). The nuclei appear pale in comparison to the hyperchromatic nuclei of usual-type adenocarcinoma and are typically present basally in a single row. The nuclei are round to oval, in contrast to elongated and stratified nuclei of usual-

E. C. Pirog et al.

Fig. 37 Gastric adenocarcinoma. Tumor cells have voluminous pale eosinophilic cytoplasm, prominent cell borders, pale round vesicular nuclei, and prominent nucleoli. The lower right shows tumor with minimal deviation features; the upper left shows more atypical tumor with single cell infiltration of the stroma

Fig. 38 Gastric adenocarcinoma. High-power magnification demonstrates tumor cells with voluminous pale, eosinophilic cytoplasm, prominent cell borders, pale round vesicular nuclei, and prominent nucleoli

type adenocarcinoma, and have delicate, diffuse chromatin and distinct nucleoli, unlike usual-type adenocarcinoma that typically has coarse chromatin (Fig. 38). Mitotic figures are seen but are not abundant. Gastric adenocarcinoma may show a spectrum of differentiation from very well-differentiated areas lacking atypia and categorized as minimal deviation adenocarcinoma (Figs. 39 and 40) to areas with moderate to marked cytologic atypia

6

Carcinoma and Other Tumors of the Cervix

353

Fig. 39 Gastric adenocarcinoma, minimal deviation type. Very well-differentiated glands have minimal atypia, pale, round nuclei with conspicuous nucleoli

Fig. 41 Gastric adenocarcinoma. The tumor shows voluminous cytoplasm and conspicuous nucleoli and mild to moderate nuclear atypia

Fig. 40 Gastric adenocarcinoma, minimal deviation type. Well-differentiated glands with secretory granules

Fig. 42 Gastric adenocarcinoma. The tumor shows abundant eosinophilic cytoplasm, round nuclei, and marked nuclear atypia. Single cell infiltration, frequently seen in this tumor, is present in the center of the microphotograph

characterized by variation of cell size, loss of polarity, nuclear enlargement, variation of nuclear shapes, and the presence of macronucleoli (Figs. 41, 42, and 43). Other morphologic appearances of these tumors include papillary forms (Fig. 44) and tumors with glands with diminished cytoplasm (Fig. 45). Minimal deviation adenocarcinoma of gastric type is an extremely well-differentiated variant of gastric adenocarcinoma in which the neoplastic epithelium shows a high degree of maturation. These tumors were originally referred to as “adenoma malignum.” Because of their close cytologic resemblance to normal endocervical glands, in 1975 Silverberg and Hurt introduced the term

“minimal deviation adenocarcinoma” (MDA) for these lesions (Silverberg and Hurt 1975). Minimal deviation adenocarcinomas are uncommon tumors and account for only 1–3% of all cervical adenocarcinomas. The characteristic microscopic features of MDA include low-grade cytology but markedly atypical architecture (Figs. 39 and 40). The glands are lined by epithelium that usually has minimal, if any, nuclear atypia, but the nuclei are larger in comparison to adjacent benign glands. The nuclei are rounded, slightly vesicular with conspicuous nucleoli. Desmoplasia may be present surrounding the angular outpouchings or in the

354

E. C. Pirog et al.

diagnosis cannot be made on a superficial cervical biopsy, but instead requires either a cone biopsy or hysterectomy specimen.

Fig. 43 Gastric adenocarcinoma. The tumor shows voluminous pale cytoplasm, round nuclei, and marked nuclear atypia

Fig. 44 Gastric adenocarcinoma, minimal deviation type. Papillary tumor variant

deep portion of the tumor. However, in some cases, large areas of invasive tumor may be devoid of any stromal reaction. The presence of glands adjacent to thick-walled blood vessels is helpful in determining that stromal invasion is present. The most reliable criterion is the haphazard arrangement of glands that extend beyond the level of normal endocervical glands. Minimal deviation adenocarcinoma often involves more than two-thirds of the thickness of the cervical stroma, while the normal endocervical crypts and tunnels do not extend as deep. Because the depth of penetration of the glands is an essential histologic feature of minimal deviation adenocarcinoma, in most cases, the

Immunohistochemical Staining The cytoplasm of gastric adenocarcinoma characteristically shows immunopositivity for gastric mucin markers MUC-6 and HIK-1083 (Carleton et al. 2016; Mikami et al. 2004), however, in some cases such staining may be only focal. The staining has to be interpreted in the proper context because benign endocervical glands may be positive for HIK-1083 and MUC-6 in 2% and 8% of cases (Mikami et al. 2004). Similar to usual-type adenocarcinoma, gastric adenocarcinoma is negative for ER and PR (Carleton et al. 2016) and demonstrates cytoplasmic positivity for CEA (Carleton et al. 2016). The staining for p16 is typically negative (Fig. 45b), correlating with the negative HPV status; however, rare cases may show strong, blocklike p16 positivity (Carleton et al. 2016). Less than a half of cases show mutation-type p53 staining with either diffuse strong positivity (Fig. 45c), or a complete lack of p53 expression (null pattern) (Carleton et al. 2016). In addition, gastric adenocarcinomas are reported to express CK7, PAX-8, CA19.9, and hepatocyte nuclear factor-1 beta (Carleton et al. 2016). Differential Diagnosis Gastric minimal deviation adenocarcinoma has to be differentiated from benign conditions, of which the most important are LEGH, deep nabothian cysts, endocervical tunnel clusters, endocervicosis of the cervical wall, diffuse laminar endocervical glandular hyperplasia and adenomyomas of endocervical type. Lack of ER/PR staining distinguishes this tumor from all of the benign conditions except for LEGH (Carleton et al. 2016). Similar to gastric adenocarcinoma, LEGH shows immunopositivity for gastrin mucin markers and is negative for hormone receptors (Mikami et al. 2004). Differentiation of minimal deviation adenocarcinoma from non-invasive LEGH may be extremely difficult. A haphazard gland arrangement, extension deep into the cervical stroma and a desmoplastic response are all in favor of minimal deviation adenocarcinoma.

6

Carcinoma and Other Tumors of the Cervix

355

Fig. 45 (a) Gastric adenocarcinoma. Glands with diminished cytoplasm. Elsewhere in the sections the tumor showed typical voluminous cytoplasm. (b) Gastric adenocarcinoma same as (a), p16 immunostaining. Gastric

adenocarcinoma is typically negative with p16 immunostaining. (c) Gastric adenocarcinoma same as (a), p53 immunostaining. Tumor showing aberrant (diffuse) p53 expression

Desmoplasia can be better visualized with smooth muscle actin (SMA) immunostaining of the stromal cells. Positive SMA staining has been described in cervical stromal cells immediately adjacent to foci of tumor invasion, but not around LEGH (Mikami et al. 2005). In addition, minimal deviation adenocarcinoma was reported to be negative for PAX-2, with approximately 40% showing mutation type staining for p53, whereas LEGH exhibited positive PAX-2 staining and non-aberrant (“wild-type”) p53 expression (Carleton et al. 2016; Mikami et al. 2009; Rabban et al. 2010). Among the malignant lesions, gastric adenocarcinoma may be difficult to distinguish from usual-type and clear cell adenocarcinomas. Expression of MUC-6 (HIK-1083 is not routinely available outside of Japan) and a negative p16 stain favor gastric differentiation, whereas usual-type adenocarcinoma has diffuse p16

expression and lacks MUC-6 expression. In contrast to gastric adenocarcinoma, clear cell carcinomas typically show variable positivity for p16 and lack MUC-6 expression (Park et al. 2011). Interestingly, both clear cell carcinomas and gastric adenocarcinomas demonstrate cytoplasmic positivity for hepatocyte nuclear factor-1 beta (Carleton et al. 2016) and napsin A (unpublished observation). Gastric-type cervical adenocarcinomas show a predisposition for metastatic spread to the ovaries, abdominal cavity, and extraperitoneal sites, and the initial presentation may be with metastases which must be differentiated from metastatic gastrointestinal and pancreatobilliary tract adenocarcinomas. The most useful immunostain in such cases is PAX-8, as 68% of primary cervical gastric adenocarcinomas were shown to express this marker (Carleton et al. 2016).

356

Signet Ring Cell-Type Adenocarcinoma of the Cervix General Features Primary signet ring cell tumors of the cervix in the pure form are exceedingly rare and more often are admixed with intestinal, gastric, or endocervicaltype mucinous adenocarcinomas. A proportion of signet ring cell adenocarcinomas found in the cervix are metastatic lesions, most commonly from a primary gastric carcinoma (Imachi et al. 1993). Only few case reports of primary cervical tumors are available (Sal et al. 2016). Tumors arising in a background of usual-type adenocarcinoma were shown to be HPV-positive (Sal et al. 2016). Microscopic Findings The tumor shows characteristic infiltration by clusters or single cells distended by a mucin vacuole. Of the reported cases, the tumors were positive for p16, CK7, and CEA (Sal et al. 2016). A metastasis from primary gastric, breast, colonic, or appendiceal carcinoma has to be ruled out using immunohistochemical analysis in correlation with clinical findings and imaging studies.

Clear Cell Adenocarcinoma of the Cervix General Features Clear cell adenocarcinomas account for 2–7% of cervical adenocarcinomas and comprise a heterogeneous group of malignancies. Clear cell adenocarcinoma has a bimodal age distribution. The first peak occurs between 17 and 37 years of age (mean age 26 years); in addition, rare cases of clear cell adenocarcinoma in children have been reported (Hanselaar et al. 1997; Liebrich et al. 2009). The second peak occurs in women who are 44–88 years of age (mean age 71 years). The pathogenesis of clear cell adenocarcinomas of the cervix is not well understood. With rare exceptions, clear cell adenocarcinomas are negative for HPV DNA (Park et al. 2011; Holl et al. 2015; Ueno et al. 2013). In the past, cases occurring in younger patients were linked to diethylstilbestrol (DES) exposure in utero

E. C. Pirog et al.

(Robboy et al. 1984). In these patients the tumors were developing on the ectocervix, rather than in the endocervical canal. More recent data on patients born outside the DES exposure period still show a bimodal age distribution with some cases occurring in young, virginal women without exposure to either DES or HPV (Liebrich et al. 2009). It is thought that clear cell adenocarcinomas may develop from either adenosis of the ectocervix, cervical endometriosis, or cervical tuboendometrioid metaplasia (Hiromura et al. 2009). Immunostaining analysis of tumors from a group of older patients showed increased expression of EGFR (75% cases) and HER2 (25% cases) and loss of PTEN expression (50% cases). In addition, 58% of cases demonstrated expression of p-AKT, and 50% of cases had expression of p-mTOR, suggesting involvement of the PI3K-AKT pathway (Ueno et al. 2013). The prognosis of surgically treated patients with stage IB–IIB clear cell carcinomas without exposure to DES is similar to patients with non-clear cell adenocarcinomas (Thomas et al. 2008).

Microscopic Findings The histologic features of clear cell adenocarcinoma are similar to those of the more common clear cell adenocarcinomas of the endometrium or ovary. The appearances of clear cell carcinoma developing in women with or without DES exposure are the same. There are three basic microscopic patterns: solid, tubulocystic, and papillary. The cells comprising the tumor have abundant clear cytoplasm due to the accumulation of glycogen or granular eosinophilic cytoplasm, with prominent nuclei that can be quite hyperchromatic and pleomorphic and project into the lumen of the cysts and tubules to form “hobnail cells” (Fig. 46). The papillae often have hyalinized cores. Immunohistochemical Staining Cervical clear cell carcinomas are positive for hepatocyte nuclear factor-1 beta and napsin A similar to their ovarian and uterine counterparts (Park et al. 2011). The staining for both of these markers however is not entirely specific, as a proportion of gastric adenocarcinomas also shows positivity with both stains. Despite the

6

Carcinoma and Other Tumors of the Cervix

Fig. 46 Clear cell adenocarcinoma of the cervix. Glands are lined by markedly atypical cells with clear cytoplasm

negative HPV status, the tumors may show positivity for p16 with a spectrum from focal to diffuse. Clear cell carcinomas are negative for CEA, ER, and PR, and the majority show non-aberrant (“wild-type”) p53 expression (Park et al. 2011; Ueno et al. 2013).

Differential Diagnosis A metastasis from clear cell adenocarcinoma of the endometrium or ovary has to be excluded based on tumor location on imaging studies or histologic findings. In addition, clear cell adenocarcinoma has to be differentiated from the HPV-related clear cell variant of SCC and clear cell adenosquamous carcinoma of the cervix, as well as HPV-negative gastric and mesonephric adenocarcinomas. The clear cell variant of SCC shows a diffuse solid growth pattern, whereas clear cell adenocarcinoma displays a spectrum of architectural patterns within the same lesion. Clear cell adenosquamous carcinoma is composed of solid sheets of clear cells in addition to glandular spaces lined by columnar cells which are positive for mucicarmine (Fujiwara et al. 1995). Clear cell adenocarcinoma with tall, columnar cells may be difficult to distinguish from gastric adenocarcinoma on routine sections, and immunostains for p16, hepatocyte nuclear factor-1 beta, and napsin A may not be helpful due to some overlap in expression patterns; however, CEA was reported to be negative in clear cell carcinomas while showing variable expression in gastric tumors (Park et al. 2011). Clear cell adenocarcinoma with a

357

tubulocystic pattern and flattened epithelium must be differentiated from mesonephric adenocarcinomas, which typically express GATA-3 staining. In a small biopsy or curettage sample, the tumor may be difficult to distinguish from benign conditions such as florid or solid microglandular hyperplasia, florid Arias-Stella reaction, atypical oxyphilic metaplasia, and radiation-induced atypia. In all these conditions, lack of significant proliferative activity and expression of hormone receptors favors the benign processes. Clinical history of use of progestagenic hormones, current or recent pregnancy, or history of radiation can help to support the respective benign diagnoses in the differential diagnosis.

Serous Adenocarcinoma of the Cervix General Features Primary cervical serous adenocarcinoma is exceedingly rare. The majority of serous tumors involving the cervix are malignancies spreading directly from the endometrium or metastases from a primary ovarian, tubal, or peritoneal serous adenocarcinoma. Patients with primary cervical serous cancer range in age from 27 to 79 years (mean 52 years) (Zhou et al. 1998; Nofech-Mozes et al. 2006). Of the few cases reported in the literature, only rare tumors were positive for HPV DNA (Park et al. 2011). It is speculated that cervical serous adenocarcinoma may be caused by p53 gene mutations, similar to its endometrial counterpart (Nofech-Mozes et al. 2006). In the largest reported series, this histologic variant has been associated with a poor prognosis when diagnosed at an advanced stage, but the outcome for patients with stage I tumors is similar to that of patients with cervical adenocarcinomas of the usual type (Zhou et al. 1998).

Microscopic Findings Serous adenocarcinoma of the cervix is histologically identical to serous adenocarcinomas arising in the endometrium. The diagnosis of primary serous cervical adenocarcinoma should be made only if spread from other gynecologic

358

sites has been excluded. These tumors are composed of papillary tufts and complex papillae lined by cells with pleomorphic, hobnailed, high-grade nuclei. Alternatively, they infiltrate the stroma with glands with slit-like lumina and lined by markedly atypical cells, some with “smudgy” nuclei. High mitotic activity may be seen. In the reported series of serous adenocarcinomas, almost half of the cases were mixed with another histologic type, most commonly with low-grade villoglandular adenocarcinoma (Zhou et al. 1998). The tumors are positive for p16 while negative for ER and PR. Aberrant (“mutation-type”) p53 expression was seen in roughly half of the reported cases (NofechMozes et al. 2006).

Mesonephric Carcinoma of the Cervix General Features Mesonephric adenocarcinoma is a rare cervical tumor which develops from the mesonephric duct remnants located deep in the lateral cervical stroma. Mesonephric duct remnants are detected in up to 20% of cervices removed during routine hysterectomy, and adenocarcinomas can rarely develop in these remnants. Patients range in age from 34 to 72 years without apparent peak age, and most present with abnormal vaginal bleeding (Clement et al. 1995; Silver et al. 2001). In a recent study of targeted next-generation sequencing, 81% of mesonephric adenocarcinomas had either a KRAS (n = 12) or NRAS (n = 1) mutation. Mutations in chromatin remodeling genes (ARID1A, ARID1B, or SMARCA4) were present in 62% of cases. No mutations of PIK3CA or PTEN genes were identified. In addition, 1q gain was found in 75% of cases (Mirkovic et al. 2015). HPV DNA is not detected in this tumor type (Kenny et al. 2012). Stage I mesonephric carcinomas seem to have a more indolent behavior than other types of cervical adenocarcinoma (Clement et al. 1995). However, several high-stage tumors have had an aggressive course, and several tumors with a sarcomatoid component have metastasized (Silver et al. 2001).

E. C. Pirog et al.

Microscopic Findings Mesonephric carcinomas are very rare and in the past were confused with clear cell carcinomas of the cervix. In contrast to the superficial location of cervical clear cell adenocarcinomas, true mesonephric adenocarcinomas develop deep in the lateral wall of the cervix, in a site corresponding to the location of mesonephric duct remnants. Therefore, they often extend into the outer third of the cervical wall. Numerous architectural growth patterns can be seen, including tubular (Fig. 47), ductal, papillary, retiform, sex-cord-like, solid, and sarcomatoid, and these patterns may be seen in various proportions within the same tumor (Silver et al. 2001). The characteristic feature of tumors forming glandular spaces is the presence of periodic acid-Schiff (PAS)-positive, diastase-resistant, deeply eosinophilic intraluminal secretions similar to those present in benign mesonephric proliferations (Fig. 48). The ductal pattern consists of large tubular or dilated glandular spaces with occasional intraluminal infoldings or papillae lined by tall columnar cells with large hyperchromatic nuclei. This pattern may mimic endometrioid adenocarcinoma. In the tubular pattern the tumor grows as small, round to oval, tightly packed glands lined by low columnar, cuboidal, or flattened cells. The retiform pattern is characterized by branching, zigzag-shaped glandular spaces

Fig. 47 Mesonephric carcinoma. Tubular glands with eosinophilic secretions invade deep cervical stroma in a haphazard fashion

6

Carcinoma and Other Tumors of the Cervix

Fig. 48 Mesonephric carcinoma. Densely packed small tubular glands are lined by atypical cells and contain eosinophilic intraluminal secretions

resembling rete ovarii. The papillary pattern resembles the papillary growth of clear cell or serous adenocarcinoma; however, the nuclei are bland and uniform and lack atypia. In the sex-cord pattern, the tumor grows in long cords and trabeculae of cells with scant eosinophilic cytoplasm. Cytologically, the tumors are composed of relatively uniform columnar or cuboidal cells with scant to moderate amount of dark eosinophilic cytoplasm. The nuclei are oval, hyperchromatic with mostly stippled chromatin showing minimal to moderate atypia. Marked nuclear atypia is not seen. The mitotic index is highly variable and may range from 1 to 50 mitoses per 10 high-power field (HPF) (Silver et al. 2001). Mesonephric hyperplasia of the lobular or diffuse pattern is seen adjacent to the invasive lesions in the majority of cases.

Immunohistochemical Staining The immunohistochemical profile of mesonephric carcinomas is similar to that of normal mesonephric duct remnants. Epithelial markers including CK7 and EMA are positive. The tumors are characterized by expression of PAX-8 (diffuse/ strong), GATA-3 (variable extent and intensity), calretinin (may be focal), and CD10 (in the apical portion, luminal edge of the cells). Tumors typically have patchy p16 expression and are negative for ER, PR, and CK20 (Howitt et al. 2015; Silver et al. 2001; Kenny et al. 2012).

359

Differential Diagnosis Mesonephric adenocarcinoma has a wide differential diagnosis. On the benign spectrum, it has to be differentiated from hyperplasia of mesonephric remnants. The presence of cytologic atypia and variable architectural patterns favors adenocarcinoma. In addition, assessment of proliferative activity with a Ki-67 immunostain may be helpful, as hyperplasia was reported to show positivity in only 1–2% of cells, as compared to 5–36% of cells in adenocarcinoma (Silver et al. 2001). Mesonephric carcinomas with a prominent spindle cell component have to be distinguished from cervical carcinosarcoma. The carcinomatous component in the latter entity is often squamous or basaloid in contrast to tubules and glands present in mesonephric carcinomas with a prominent spindle cell component. The ductal variant of the tumor has to be differentiated from endometrioid adenocarcinoma. The presence of other distinct architectural patterns, intraluminal eosinophilic secretions, adjacent benign mesonephric remnants, and the pattern of ER-/PR-/calretinin+/ GATA-3+ staining confirm the diagnosis of mesonephric adenocarcinoma. The tubular pattern of mesonephric adenocarcinoma may simulate clear cell carcinoma, while the papillary pattern may mimic either clear cell or serous adenocarcinoma; however, in contrast to these two latter tumors, mesonephric adenocarcinoma typically shows lower nuclear grade, adjacent mesonephric remnants, and expression of GATA-3 (Howitt et al. 2015).

Other Epithelial Tumors Adenosquamous Carcinoma Adenosquamous carcinomas are defined as tumors that contain an admixture of histologically malignant squamous and glandular component. Just like SCC, adenosquamous tumors are linked to infection with high-risk HPVs and are most frequently positive for either HPV 18 or HPV 16 (de Sanjose et al. 2010; Holl et al. 2015). Adenosquamous carcinomas account for

360

approximately 2–3% of all cervical cancers (de Sanjose et al. 2010). The average age of patients is 50 years (Holl et al. 2015). The squamous component generally includes areas that are well differentiated and contain either keratin “pearls” or sheets of cells with individual cell keratinization. To make the diagnosis of adenosquamous carcinoma, the glands must be histologically recognizable (Fig. 49). The welldifferentiated tumors are usually easily identified; however, when the adenocarcinoma component is less well differentiated and is present only focally, it can easily be overlooked. Adenosquamous carcinomas are thought to arise from the pluripotential subcolumnar reserve cells of the endocervical mucous epithelium and represent biphasic differentiation. Squamous and glandular components were shown to be monoclonal in origin and had identical types of HPV supporting origin from a common precursor cell (Ueda et al. 2008). The prognosis in patients with adenosquamous carcinoma has been reported as worse than that in patients with SCC and adenocarcinoma, although not all studies have confirmed this finding (Lee et al. 2014). A rare tumor variant of adenosquamous carcinoma in which at least 70% of tumor cells have vacuolated, clear cytoplasm containing large amounts of glycogen has been referred to as clear cell adenosquamous carcinoma (Fujiwara et al. 1995). The cohesive sheets of tumor cells

Fig. 49 Adenosquamous cell carcinoma. Invasive nests of carcinoma display both squamous and glandular differentiation

E. C. Pirog et al.

are frequently subdivided by connective tissue septa, which can have a prominent lymphocytic infiltrate that produces a lobulated appearance. The tumors demonstrate focal gland formation and stain positively with a mucin stain such as mucicarmine. In some clear cell adenosquamous carcinomas, there are spindle-shaped cells suggesting squamous differentiation. Clear cell adenosquamous carcinomas need to be distinguished from clear cell carcinomas and glassy cell carcinoma of the cervix. Unlike clear cell carcinomas, clear cell adenosquamous carcinomas lack papillary or tubulocystic areas and hobnail cells. Clear cell adenosquamous carcinoma is associated with HPV 18 and has an aggressive clinical course. In one series, 7 of the 11 patients died of their disease, including 3 of 5 patients with stage IB disease (Fujiwara et al. 1995).

Glassy Cell Carcinoma Glassy cell carcinoma is a poorly differentiated adenosquamous carcinoma with distinct microscopic features. It comprises 10 mitotic figures/10 HPF). The tumor cells usually stain for

Fig. 56 Neuroendocrine carcinoma. Solid islands of carcinoma exhibit some cords and palisading typical of neuroendocrine differentiation; nuclei are atypical and numerous mitotic figures are evident

365

chromogranin A, synaptophysin, and CD56. The tumors show diffuse p16 positivity owing to association with high-risk HPV. Thyroid transcription factor-1 (TTF-1) positivity is seen in proportion of cases (McCluggage et al. 2010). Cytokeratin stains are often positive, and CEA is expressed in 70% of cases (Gilks et al. 1997). Large cell neuroendocrine carcinomas are often associated with glandular lesions. In one series, 66% of large cell neuroendocrine carcinomas had a coexisting adenocarcinoma in situ, and 25% had a coexisting adenocarcinoma (Gardner et al. 2011). Atypical carcinoids and large cell neuroendocrine carcinomas can be distinguished based on mitotic activity, nuclear atypia, and degree of necrosis (Table 8) (Albores-Saavedra et al. 1997). It is more difficult to differentiate between large cell neuroendocrine carcinoma and poorly differentiated cervical adenocarcinomas or squamous cell carcinomas. It is important that trabecular and insular growth patterns be looked for in poorly differentiated cervical tumors and that stains for neuroendocrine markers be used whenever there is an indication of neuroendocrine differentiation. It should be cautioned, however, that occasional typical cervical adenocarcinomas and adenosquamous carcinomas can stain focally with neuroendocrine markers or contain occasional argyrophilic cells. In contrast, large cell neuroendocrine carcinomas have evidence of neuroendocrine differentiation by routine light microscopy and show more diffuse expression of neuroendocrine markers. Finally, large cell neuroendocrine carcinoma has to be distinguished from malignant melanoma. The presence of melanin pigment and immunoreactivity for S100, HMB-45, and Sox10 facilitates the distinction between these tumors. Several studies have explored the role of HPV in large cell neuroendocrine carcinomas. In the largest series, HPV 16 was detected by in situ hybridization or polymerase chain reaction in 58% of cases, and HPV 18 was detected in 16% of cases (Grayson et al. 2002). Large cell neuroendocrine carcinomas are highly aggressive neoplasms. In the review of 31 published cases, 65% of patients died of disease within 3 years of diagnosis (Grayson et al. 2002).

366

Small Cell Carcinoma Small cell carcinomas of the cervix are histologically identical to their counterparts at other sites such as the lung. In most series these tumors account for 1–2% (range 0.5–5%) of all cervical tumors (Abeler et al. 1994). The age of the patients ranges from the second to ninth decade, with mean and median age in the fifth decade (Abeler et al. 1994). Most patients present with abnormal vaginal bleeding and have an obvious mass on pelvic examination. In rare cases, patients present with abdominal symptoms related to ovarian metastases. The number of patients presenting with abnormal cytologic examination is smaller than in patients with SCC. This results from lack of an in situ component and rapid growth of the tumor (Ambros et al. 1991).

Pathologic Findings Grossly, small cell carcinomas range in size from small, clinically unapparent lesions to large ulcerated tumors. Microscopically tumors are composed of sheets and cords of closely packed, small, scant cells with inconspicuous cytoplasm. The cells have hyperchromatic nuclei with finely stippled chromatin, inconspicuous nucleoli, and high nuclear to cytoplasmic ratio. The mitotic index is high, with three or more mitotic figures present in most HPF (Fig. 57). The nuclear shape varies from round to spindled, and nuclear molding is a characteristic feature. Smudging of the nucleus and extensive crush artifact frequently obscure nuclear detail and nucleoli. Small areas of either squamous or glandular differentiation can be present, but for tumor classification as small cell carcinoma, these elements should account for less than 5% of the tumor volume.

Immunohistochemical and Molecular Genetic Findings Neuroendocrine dense-core granules can be detected using Grimelius stains or electron microscopy in most cases. Although small cell carcinomas have been associated with ectopic

E. C. Pirog et al.

Fig. 57 Small cell carcinoma. Tumor is composed of cells with enlarged, atypical hyperchromatic nuclei, numerous mitotic figures, and scanty cytoplasm. Cellular molding also is present

adrenocorticotropic hormone (ACTH), insulin, and gastrin production, clinical symptoms related to ectopic hormone production are uncommon. By immunohistochemistry, neuroendocrine markers such as chromogranin or synaptophysin are present in many cases; however, staining might be very focal. Small cell carcinomas show variable expression of cytokeratins, epithelial membrane antigen, as well as a variety of hormones and polypeptides including ACTH, calcitonin, serotonin, gastrin, substance P, VIP, and somatostatin (Ueda et al. 1989). In addition, the tumors show diffuse p16 expression and TTF-1 positivity (McCluggage et al. 2010). Virtually all cervical small cell carcinomas have been associated with high-risk HPV types 18 and 16, with type 18 being the most prevalent, detected in 82–100% of the cases (Wang and Lu 2004).

Differential Diagnosis Differentiation between small cell carcinoma and nonkeratinizing squamous carcinoma with small cells can be difficult. The diagnosis of small cell carcinoma should be reserved for tumors composed of small cells in which squamous or glandular differentiation is absent or minor. Histologically, cells of nonkeratinizing squamous carcinoma with small cells resemble those of HSIL and lack the nuclear molding and

6

Carcinoma and Other Tumors of the Cervix

extensive crush artifact present in most small cell carcinomas. Small cell carcinomas invade the stroma diffusely in trabeculae and poorly defined nests. In contrast, nonkeratinizing squamous carcinomas with small cells invade the stroma in discrete nests. In an individual case, immunohistochemistry for neuroendocrine markers may not be helpful because 40% of nonkeratinizing squamous carcinomas with small cells are positive for neuroendocrine markers and 40% of small cell carcinomas are positive for cytokeratins (Ambros et al. 1991). Nuclear staining for p63 confirms squamous differentiation in nonkeratinizing small cell carcinomas, while neuroendocrine-type small cell carcinomas are negative for this marker (Wang et al. 2001). Due to the presence of high-risk HPV in both tumor types, p16 expression is similar. Expression of TTF-1 has been reported in cervical small cell carcinomas and can cause diagnostic problems in tumors that metastasize to the lungs. Immunohistochemical staining with antibodies against leukocyte common antigen and neuroendocrine markers can be useful for differentiating small cell carcinoma from lymphoproliferative disorders.

Clinical Behavior and Treatment Small cell carcinoma of the cervix is a highly aggressive tumor. Lymph-vascular invasion is present in 90% of cases and is often extensive (Abeler et al. 1994). The prognosis of small cell carcinoma of the cervix is worse than that of stage-comparable, poorly differentiated squamous carcinoma (Ambros et al. 1991; Zivanovic et al. 2009). Five-year survival rate is reported at 14% (Abeler et al. 1994). Combined modality treatments with chemotherapy and radiotherapy are currently used in management of small cell carcinoma. In a recent series, patients with earlystage disease treated with platinum-based chemotherapy in addition to radiation had significantly better overall and disease-free survival when compared to patients who did not receive chemotherapy as part of their initial treatment (Zivanovic et al. 2009).

367

Mesenchymal and Mixed EpithelialMesenchymal Tumors Malignant mesenchymal tumors that can arise in the cervix include leiomyosarcoma, endometrial stromal sarcoma, embryonal rhabdomyosarcoma (botryoid type) (Figs. 58 and 59), alveolar soft part sarcoma, malignant schwannomas, and osteosarcomas (see ▶ Chap. 10, “Mesenchymal Tumors of the Uterus”). Primary cervical sarcomas are rare, of which the most common is leiomyosarcoma.

Fig. 58 Embryonal rhabdomyosarcoma. Polypoid tumor has a cambium layer of hypercellular stroma immediately beneath the benign endocervical epithelium

Fig. 59 Embryonal rhabdomyosarcoma. Hypercellular foci within background edematous hypocellular stroma are composed of immature atypical cells with scant cytoplasm; mitotic figures and apoptotic bodies are usually evident

368

Primary cervical mixed epithelial and mesenchymal tumors include malignant müllerian mixed tumor (MMMT) and müllerian adenosarcoma. Cervical MMMT are less common than their much more common uterine counterparts, and in contrast to uterine MMMT, cervical MMMTs are related to infection with high-risk HPV (Grayson et al. 2001). Both cervical and uterine tumors usually occur in postmenopausal women, and both typically form polypoid or pedunculated masses. The mean age of patients was 65 years in the largest published series of cervical tumors (Clement et al. 1998). Histologically, cervical MMMTs differ in their appearance from MMMTs arising in the uterus. The carcinomatous component in cervical MMMT is often a basaloid tumor composed of anastomosing densely cellular trabeculae of small cells with scant cytoplasm and peripheral palisading. Other epithelial patterns include typical SCC and endometrioid adenocarcinoma. Adenoid basal and adenoid cystic components have also been reported in several cases (Mathoulin-Portier et al. 1998). The sarcomatous element is typically homologous and frequently has the appearance of a fibrosarcoma or endometrial stromal sarcoma. The sarcomatous element is frequently high-grade and may have myxoid change. Extension of uterine MMMT to the cervix is in the differential diagnosis of cervical MMMT. The correct diagnosis is based on the dominant location of the tumor, the appearance of the carcinomatous component, and the detection of high-risk HPV. In a study of eight patients with cervical MMMT, HPV DNA was detected by polymerase chain reaction in all cases. Using in situ hybridization, HPV 16 DNA was detected in both the epithelial and sarcomatous components in three cases (Grayson et al. 2001). Although the number of reported cases is small, cervical MMMTs may have a better prognosis than their uterine counterparts (Clement et al. 1998). Only a small number of müllerian adenosarcomas of the cervix have been reported (Jones and Lefkowitz 1995). Adenosarcomas occur in women between the ages of 14 and 67 years, with a mean age of 38 years (Jones and Lefkowitz 1995). Women typically present with vaginal bleeding or recurrent cervical polyps. Microscopically, tumors usually demonstrate

E. C. Pirog et al.

thick papillae covered with a typical endocervical-type epithelium. The appearance of the sarcomatous component can vary considerably. In some tumors, it consists of mitotically active, plump spindle cells that form periglandular cuffs and a cambium layer under the surface epithelium. At least two mitotic figures per 10 HPF are required to make a diagnosis of adenosarcoma, but in most cases the mitotic index exceeds 4 per 10 HPF. In other tumors, the stromal component contains foci that are more embryonic in appearance, with small, undifferentiated round cells that are mitotically active. Stains for desmin and myogenin as well as Ki-67 are useful to identify such foci as rhabdomyoblastic differentiation which can occur in adenosarcomas of both endometrial and cervical origin. Heterologous sarcomatous elements, including strap cells (skeletal muscle differentiation), lipoblasts, cartilage, and osteoid, can be present. The differential diagnosis includes adenofibroma, atypical endocervical polyp, adenomyoma of the cervix, and MMMT. Adenofibroma is also a biphasic lesion, but both the epithelium and stroma are benign. Atypical endocervical polyps show increased stromal cellularity and reactive nuclear atypia, but these changes are often focal and mitotic activity is absent. Adenomyomas can be distinguished from adenosarcoma by the presence of a well-defined bland myomatous component. The prognosis of cervical adenosarcoma is usually good after surgical therapy, although several patients have died of disease or developed recurrent tumor (Jones and Lefkowitz 1995).

Miscellaneous Tumors Primary malignant melanoma is among the least common of the malignant tumors that arise in the cervix (Clark et al. 1999). Common presenting signs include vaginal bleeding, frequently of short duration. In most instances, the lesion is pigmented and dark brown. The diagnosis of primary melanoma of the cervix is based on the histologic demonstration of junctional changes in the squamous epithelium and the absence of similar lesions elsewhere in the body.

6

Carcinoma and Other Tumors of the Cervix

Morphologically, it is identical to melanoma arising in the skin and extragenital mucous membranes; it frequently contains intracytoplasmic melanin pigment granules. Some tumors are amelanotic and have to be distinguished from undifferentiated carcinoma. Rarely cervical malignant melanoma can be composed of clear cells. Immunohistochemical staining for HMB 45, Melan-A (MART-1), S100 protein, and Sox10 and absence of epithelial markers are helpful to exclude clear cell carcinoma. Spindle cell malignant melanoma has to be distinguished from leiomyosarcoma or malignant peripheral nerve sheet tumor. In contrast to malignant melanoma, leiomyosarcoma expresses smooth muscle markers and is negative for melanocytic markers. The cell pigmentation, nesting, and the presence of an atypical epidermal or junctional component, together with diffuse, strong reactivity for S100 protein and positivity for other melanocytic markers, help to differentiate melanoma from malignant peripheral nerve sheet tumor. The prognosis of primary malignant melanoma is poor, with only 25% survival rate for patients with stage I disease (Clark et al. 1999). Primary choriocarcinoma and epithelioid trophoblastic tumors in the cervix are rare. The gross and microscopic appearance, as well as the clinical course, is identical with those found in the uterine corpus. Primary cervical germ cell tumors have been described: these include both the mature teratomas and yolk sac tumors. There are also case reports of primitive neuroectodermal tumors (PNET) of the cervix (Snijders-Keilholz et al. 2005). These tumors appear to be identical to PNETs occurring at other sites and in some cases have expressed the restricted surface antigen MIC-2, showed positive staining for CD99, and contained the EWS/FLI-1 chimeric mRNA transcript characteristic of PNET/Ewing sarcoma family (Masoura et al. 2012).

Secondary Tumors Direct extension from a pelvic tumor is the most common source of cervical involvement by secondary carcinoma, often originating in the endometrium, rectum, or bladder. Intrapelvic and

369

intragenital, lymphatic, or vascular metastases to the cervix occur less frequently. These lesions are usually associated with ovarian carcinoma and endometrial adenocarcinoma and less commonly with transitional cell carcinoma of the bladder. Another lesion that has a relatively high rate of cervical metastasis is choriocarcinoma. Sarcomas of the uterine corpus may also involve the cervix. Metastases to the cervix from distant primary sites are rare, the most common being the gastrointestinal tract (the colon and stomach), ovary, and breast (Perez-Montiel et al. 2012). Instances of metastatic carcinoma from the kidney, gallbladder, pancreas, lung, thyroid, and malignant melanoma have also been described. On occasion, metastases may occur primarily as cervical involvement and pose a differential diagnostic problem. Unusual gross appearance or histologic patterns, e.g., signet ring cell carcinoma or clear cell carcinoma, provide a clue to the possibility of origin from a distant primary site.

References Abeler VM et al (1994) Small cell carcinoma of the cervix. A clinicopathologic study of 26 patients. Cancer 73(3): 672–677 Albores-Saavedra J et al (1976) Carcinoid of the uterine cervix: additional observations on a new tumor entity. Cancer 38(6):2328–2342 Albores-Saavedra J et al (1992) The solid variant of adenoid cystic carcinoma of the cervix. Int J Gynecol Pathol 11(1):2–10 Albores-Saavedra J et al (1997) Terminology of endocrine tumors of the uterine cervix: results of a workshop sponsored by the College of American Pathologists and the National Cancer Institute. Arch Pathol Lab Med 121(1):34–39 Ambros RA et al (1991) Evaluation of histologic, morphometric, and immunohistochemical criteria in the differential diagnosis of small cell carcinomas of the cervix with particular reference to human papillomavirus types 16 and 18. Mod Pathol 4(5):586–593 Andersen ES, Nielsen K, Pedersen B (1995) The reliability of preconization diagnostic evaluation in patients with cervical intraepithelial neoplasia and microinvasive carcinoma. Gynecol Oncol 59(1):143–147 Anton-Culver H et al (1992) Comparison of adenocarcinoma and squamous cell carcinoma of the uterine cervix: a population-based epidemiologic study. Am J Obstet Gynecol 166(5):1507–1514

370 Antonelli AR et al (1991) Basement membrane components in normal, dysplastic, neoplastic laryngeal tissue and metastatic lymph nodes. Acta Otolaryngol 111(2): 437–443 Appleby P et al (2006) Carcinoma of the cervix and tobacco smoking: collaborative reanalysis of individual data on 13,541 women with carcinoma of the cervix and 23,017 women without carcinoma of the cervix from 23 epidemiological studies. Int J Cancer 118(6): 1481–1495 Appleby P et al (2007) Cervical cancer and hormonal contraceptives: collaborative reanalysis of individual data for 16,573 women with cervical cancer and 35,509 women without cervical cancer from 24 epidemiological studies. Lancet 370(9599):1609–1621 Averette HE et al (1976) Diagnosis and management of microinvasive (stage IA) carcinoma of the uterine cervix. Cancer 38(1 Suppl):414–425 Averette HE et al (1993) Radical hysterectomy for invasive cervical cancer. A 25-year prospective experience with the Miami technique. Cancer 71(4 Suppl):1422–1437 Baalbergen A, Smedts F, Helmerhorst TJ (2011) Conservative therapy in microinvasive adenocarcinoma of the uterine cervix is justified: an analysis of 59 cases and a review of the literature. Int J Gynecol Cancer 21(9): 1640–1645 Bais AG et al (2005) Lymphoepithelioma-like carcinoma of the uterine cervix: absence of Epstein-Barr virus, but presence of a multiple human papillomavirus infection. Gynecol Oncol 97(2):716–718 Benedetti-Panici P et al (2000) Early cervical carcinoma: the natural history of lymph node involvement redefined on the basis of thorough parametrectomy and giant section study. Cancer 88(10):2267–2274 Berek JS, Hacker NF (2010) Berek & Hacker’s gynecologic oncology. Wolters Kluwer/Lippincott Williams & Wilkins Health, Philadelphia, p 1 online resource (xv, 895 pages) Brainard JA, Hart WR (1998) Adenoid basal epitheliomas of the uterine cervix: a reevaluation of distinctive cervical basaloid lesions currently classified as adenoid basal carcinoma and adenoid basal hyperplasia. Am J Surg Pathol 22(8):965–975 Brinton LA et al (1993) Risk factors for cervical cancer by histology. Gynecol Oncol 51(3):301–306 Burghardt E, Holzer E (1977) Diagnosis and treatment of microinvasive carcinoma of the cervix uteri. Obstet Gynecol 49(6):641–653 Burghardt E et al (1991) Microinvasive carcinoma of the uterine cervix (International Federation of Gynecology and Obstetrics Stage IA). Cancer 67(4):1037–1045 Butler D et al (2002) Loss of Fhit expression as a potential marker of malignant progression in preinvasive squamous cervical cancer. Gynecol Oncol 86(2):144–149 Cancer Research Institute 2019, https://www.cancerresearch. org/immunotherapy/cancer-types/cervical-cancer Carleton C et al (2016) A detailed immunohistochemical analysis of a large series of cervical and vaginal gastrictype adenocarcinomas. Am J Surg Pathol 40(5):636–644

E. C. Pirog et al. Castellsague X et al (2006) Worldwide human papillomavirus etiology of cervical adenocarcinoma and its cofactors: implications for screening and prevention. J Natl Cancer Inst 98(5):303–315 Chen TD, Chuang HC, Lee LY (2012) Adenoid basal carcinoma of the uterine cervix: clinicopathologic features of 12 cases with reference to CD117 expression. Int J Gynecol Pathol 31(1):25–32 Cho NH et al (1998) Detection of human papillomavirus in warty carcinoma of the uterine cervix: comparison of immunohistochemistry, in situ hybridization and in situ polymerase chain reaction methods. Pathol Res Pract 194(10):713–720 Clark KC, Butz WR, Hapke MR (1999) Primary malignant melanoma of the uterine cervix: case report with world literature review. Int J Gynecol Pathol 18(3):265–273 Clement PB et al (1995) Malignant mesonephric neoplasms of the uterine cervix. A report of eight cases, including four with a malignant spindle cell component. Am J Surg Pathol 19(10):1158–1171 Clement PB et al (1998) Malignant mullerian mixed tumors of the uterine cervix: a report of nine cases of a neoplasm with morphology often different from its counterpart in the corpus. Int J Gynecol Pathol 17(3): 211–222 Copeland LJ et al (1992) Superficially invasive squamous cell carcinoma of the cervix. Gynecol Oncol 45(3): 307–312 Darragh TM et al (2012) The Lower Anogenital Squamous Terminology Standardization Project for HPV-Associated Lesions: background and consensus recommendations from the College of American Pathologists and the American Society for Colposcopy and Cervical Pathology. J Low Genit Tract Dis 16(3): 205–242 Day E et al (2016) Multifocal FIGO stage IA1 squamous carcinoma of the cervix: criteria for identification, staging, and its good clinical outcome. Int J Gynecol Pathol 35(5):467–474 de la Rochefordiere A et al (2015) PIK3CA pathway mutations predictive of poor response following standard radiochemotherapy +/ cetuximab in cervical cancer patients. Clin Cancer Res 21(11):2530–2537 de Sanjose S et al (2010) Human papillomavirus genotype attribution in invasive cervical cancer: a retrospective cross-sectional worldwide study. Lancet Oncol 11(11): 1048–1056 del Pino M et al (2012) Comprehensive analysis of human papillomavirus prevalence and the potential role of low-risk types in verrucous carcinoma. Mod Pathol 25(10):1354–1363 Diaz JP et al (2008) Oncologic outcome of fertility-sparing radical trachelectomy versus radical hysterectomy for stage IB1 cervical carcinoma. Gynecol Oncol 111(2): 255–260 Dugue PA et al (2013) Immunosuppression and risk of cervical cancer. Expert Rev Anticancer Ther 13(1): 29–42

6

Carcinoma and Other Tumors of the Cervix

Eifel PJ et al (1990) Adenocarcinoma of the uterine cervix. Prognosis and patterns of failure in 367 cases. Cancer 65(11):2507–2514 Eskander RN, Bristow RE, SpringerLink (Online Service) (2015) Gynecologic oncology a pocketbook. Springer, New York Fujiwara H et al (1995) Clear cell adenosquamous carcinoma of the cervix. An aggressive tumor associated with human papillomavirus-18. Cancer 76(9):1591–1600 Gaffney DK et al (2007) A phase II study of acute toxicity for Celebrex (celecoxib) and chemoradiation in patients with locally advanced cervical cancer: primary endpoint analysis of RTOG 0128. Int J Radiat Oncol Biol Phys 67(1):104–109 Gardner GJ, Reidy-Lagunes D, Gehrig PA (2011) Neuroendocrine tumors of the gynecologic tract: a Society of Gynecologic Oncology (SGO) clinical document. Gynecol Oncol 122(1):190–198 Gilks CB et al (1989) Adenoma malignum (minimal deviation adenocarcinoma) of the uterine cervix. A clinicopathological and immunohistochemical analysis of 26 cases. Am J Surg Pathol 13(9):717–729 Gilks CB et al (1997) Large cell neuroendocrine [corrected] carcinoma of the uterine cervix: a clinicopathologic study of 12 cases. Am J Surg Pathol 21(8):905–914 Globocan 2012 http://globocan.iarc.fr/old/FactSheets/can cers/cervix-new.asp Grayson W, Cooper K (2002) A reappraisal of “basaloid carcinoma” of the cervix, and the differential diagnosis of basaloid cervical neoplasms. Adv Anat Pathol 9(5): 290–300 Grayson W, Taylor LF, Cooper K (1999) Adenoid cystic and adenoid basal carcinoma of the uterine cervix: comparative morphologic, mucin, and immunohistochemical profile of two rare neoplasms of putative ‘reserve cell’ origin. Am J Surg Pathol 23(4):448–458 Grayson W, Taylor LF, Cooper K (2001) Carcinosarcoma of the uterine cervix: a report of eight cases with immunohistochemical analysis and evaluation of human papillomavirus status. Am J Surg Pathol 25(3):338–347 Grayson W et al (2002) Detection of human papillomavirus in large cell neuroendocrine carcinoma of the uterine cervix: a study of 12 cases. J Clin Pathol 55(2):108–114 Guitarte C et al (2014) Glassy cell carcinoma of the cervix: a systematic review and meta-analysis. Gynecol Oncol 133(2):186–191 Hanselaar A et al (1997) Clear cell adenocarcinoma of the vagina and cervix. An update of the Central Netherlands registry showing twin age incidence peaks. Cancer 79(11):2229–2236 Hiromura T et al (2009) Clear cell adenocarcinoma of the uterine cervix arising from a background of cervical endometriosis. Br J Radiol 82(973):e20–e22 Holl K et al (2015) Human papillomavirus prevalence and type-distribution in cervical glandular neoplasias: results from a European multinational epidemiological study. Int J Cancer 137(12):2858–2868

371 Hou MM et al (2014) Targeted PI3K/AKT/mTOR therapy for metastatic carcinomas of the cervix: a phase I clinical experience. Oncotarget 5(22):11168–11179 Howitt BE et al (2015) GATA3 is a sensitive and specific marker of benign and malignant mesonephric lesions in the lower female genital tract. Am J Surg Pathol 39(10): 1411–1419 Imachi M et al (1993) Metastatic adenocarcinoma to the uterine cervix from gastric cancer. A clinicopathologic analysis of 16 cases. Cancer 71(11):3472–3477 International Collaboration of Epidemiological Studies of Cervical Cancer (2009) Cervical carcinoma and sexual behavior: collaborative reanalysis of individual data on 15,461 women with cervical carcinoma and 29,164 women without cervical carcinoma from 21 epidemiological studies. Cancer Epidemiol Biomarkers Prev 18(4):1060–1069 Jessup JM et al (1996) Clinical highlights from the National Cancer Data Base: 1996. CA Cancer J Clin 46(3):185–192 Jones MW, Lefkowitz M (1995) Adenosarcoma of the uterine cervix: a clinicopathological study of 12 cases. Int J Gynecol Pathol 14(3):223–229 Jones WB et al (1993a) Early invasive carcinoma of the cervix. Gynecol Oncol 51(1):26–32 Jones MW, Silverberg SG, Kurman RJ (1993b) Welldifferentiated villoglandular adenocarcinoma of the uterine cervix: a clinicopathological study of 24 cases. Int J Gynecol Pathol 12(1):1–7 Jones WB et al (1996) Cervical carcinoma and pregnancy. A national patterns of care study of the American College of Surgeons. Cancer 77(8):1479–1488 Jones MW et al (2000) Well-differentiated villoglandular adenocarcinoma of the uterine cervix: oncogene/tumor suppressor gene alterations and human papillomavirus genotyping. Int J Gynecol Pathol 19(2):110–117 Jones MW et al (2013) Immunohistochemistry and HPV in situ hybridization in pathologic distinction between endocervical and endometrial adenocarcinoma: a comparative tissue microarray study of 76 tumors. Int J Gynecol Cancer 23(2):380–384 Kaku T et al (1997) Adenocarcinoma of the uterine cervix with predominantly villogladular papillary growth pattern. Gynecol Oncol 64(1):147–152 Kalhor N et al (2009) Immunohistochemical studies of trophoblastic tumors. Am J Surg Pathol 33(4):633–638 Karamurzin YS et al (2015) Gastric-type endocervical adenocarcinoma: an aggressive tumor with unusual metastatic patterns and poor prognosis. Am J Surg Pathol 39(11):1449–1457 Kenny SL et al (2012) Mesonephric adenocarcinomas of the uterine cervix and corpus: HPV-negative neoplasms that are commonly PAX8, CA125, and HMGA2 positive and that may be immunoreactive with TTF1 and hepatocyte nuclear factor 1-beta. Am J Surg Pathol 36(6): 799–807 Killackey MA, Jones WB, Lewis JL Jr (1986) Diagnostic conization of the cervix: review of 460 consecutive cases. Obstet Gynecol 67(6):766–770

372 Koenig C et al (1997) Papillary squamotransitional cell carcinoma of the cervix: a report of 32 cases. Am J Surg Pathol 21(8):915–921 Kojima A et al (2007) Gastric morphology and immunophenotype predict poor outcome in mucinous adenocarcinoma of the uterine cervix. Am J Surg Pathol 31(5):664–672 Kudo R, Sato T, Mizuuchi H (1990) Ultrastructural and immunohistochemical study of infiltration in microinvasive carcinoma of the uterine cervix. Gynecol Oncol 36(1):23–29 Kuragaki C et al (2003) Mutations in the STK11 gene characterize minimal deviation adenocarcinoma of the uterine cervix. Lab Investig 83(1):35–45 Kurman RJ, International Agency for Research on Cancer, World Health Organization (2014) WHO classification of tumours of female reproductive organs, 4th edn. World Health Organization classification of tumours. International Agency for Research on Cancer, Lyon, 307 p Lee KB et al (2006) Lymph node metastasis and lymph vascular space invasion in microinvasive squamous cell carcinoma of the uterine cervix. Int J Gynecol Cancer 16(3):1184–1187 Lee JY et al (2014) Prognosis of adenosquamous carcinoma compared with adenocarcinoma in uterine cervical cancer: a systematic review and meta-analysis of observational studies. Int J Gynecol Cancer 24(2):289–294 Lennerz JK et al (2009) Mucoepidermoid carcinoma of the cervix: another tumor with the t(11;19)-associated CRTC1-MAML2 gene fusion. Am J Surg Pathol 33 (6):835–843 Liang L et al (2016) Assessment of the utility of PAX8 immunohistochemical stain in diagnosing endocervical glandular lesions. Arch Pathol Lab Med 140(2): 148–152 Liebrich C et al (2009) Primary cervical cancer truly negative for high-risk human papillomavirus is a rare but distinct entity that can affect virgins and young adolescents. Eur J Gynaecol Oncol 30(1):45–48 Lininger RA et al (1998) Human papillomavirus type 16 is detected in transitional cell carcinomas and squamotransitional cell carcinomas of the cervix and endometrium. Cancer 83(3):521–527 Liotta LA (1984) Tumor invasion and metastases: role of the basement membrane. Warner-Lambert Parke-Davis Award lecture. Am J Pathol 117(3):339–348 Ma YY et al (2000) PIK3CA as an oncogene in cervical cancer. Oncogene 19(23):2739–2744 Mao TL et al (2008) HSD3B1 as a novel trophoblastassociated marker that assists in the differential diagnosis of trophoblastic tumors and tumorlike lesions. Am J Surg Pathol 32(2):236–242 Masoura S et al (2012) Primary primitive neuroectodermal tumor of the cervix confirmed with molecular analysis in a 23-year-old woman: a case report. Pathol Res Pract 208(4):245–249 Mathoulin-Portier MP et al (1998) Malignant mullerian mixed tumor of the uterine cervix with adenoid cystic component. Int J Gynecol Pathol 17(1):91–92 Matseoane S et al (1992) Diagnostic value of conization of the uterine cervix in the management of cervical

E. C. Pirog et al. neoplasia: a review of 756 consecutive patients. Gynecol Oncol 47(3):287–291 McCluggage WG, Kennedy K, Busam KJ (2010) An immunohistochemical study of cervical neuroendocrine carcinomas: neoplasms that are commonly TTF1 positive and which may express CK20 and P63. Am J Surg Pathol 34(4):525–532 McIlwaine P, Nagar H, McCluggage WG (2014) Multifocal FIGO stage 1A1 cervical squamous carcinomas have an extremely good prognosis equivalent to unifocal lesions. Int J Gynecol Pathol 33(3):213–217 McLaughlin-Drubin ME, Munger K (2009) Oncogenic activities of human papillomaviruses. Virus Res 143(2): 195–208 Melnikow J et al (1998) Natural history of cervical squamous intraepithelial lesions: a meta-analysis. Obstet Gynecol 92(4 Pt 2):727–735 Meyer JE et al (2013) HIV positivity but not HPV/p16 status is associated with higher recurrence rate in anal cancer. J Gastrointest Cancer 44(4):450–455 Mikami Y et al (2004) Gastrointestinal immunophenotype in adenocarcinomas of the uterine cervix and related glandular lesions: a possible link between lobular endocervical glandular hyperplasia/pyloric gland metaplasia and ‘adenoma malignum’. Mod Pathol 17(8):962–972 Mikami Y et al (2005) Immunophenotypic alteration of the stromal component in minimal deviation adenocarcinoma (‘adenoma malignum’) and endocervical glandular hyperplasia: a study using oestrogen receptor and alpha-smooth muscle actin double immunostaining. Histopathology 46(2):130–136 Mikami Y et al (2009) Ki67 labelling index and p53 status indicate neoplastic nature of atypical lobular endocervical glandular hyperplasia (ALEGH). Histopathology 55(3):362–364 Mikuta JJ, Celebre JA (1969) Adenocarcinoma of the cervix. Obstet Gynecol 33(6):753–756 Mills AM et al (2017) HR-HPV E6/E7 mRNA in situ hybridization: validation against PCR, DNA in situ hybridization, and p16 immunohistochemistry in 102 samples of cervical, vulvar, anal, and head and neck neoplasia. Am J Surg Pathol 41(5):607–615 Mirhashemi R et al (2003) Papillary squamous cell carcinoma of the uterine cervix: an immunophenotypic appraisal of 12 cases. Gynecol Oncol 90(3):657–661 Mirkovic J et al (2015) Targeted genomic profiling reveals recurrent KRAS mutations and gain of chromosome 1q in mesonephric carcinomas of the female genital tract. Mod Pathol 28(11):1504–1514 Monk BJ et al (2009) Phase II trial of bevacizumab in the treatment of persistent or recurrent squamous cell carcinoma of the cervix: a gynecologic oncology group study. J Clin Oncol 27(7):1069–1074 Morrison C et al (2001) Highly differentiated keratinizing squamous cell cancer of the cervix: a rare, locally aggressive tumor not associated with human papillomavirus or squamous intraepithelial lesions. Am J Surg Pathol 25(10):1310–1315 Munoz N et al (2003) Epidemiologic classification of human papillomavirus types associated with cervical cancer. N Engl J Med 348(6):518–527

6

Carcinoma and Other Tumors of the Cervix

NCI 2019, https://seer.cancer.gov/statfacts/html/cervix.html Nemejcova K, Cibula D, Dundr P (2015) Expression of HNF-1beta in cervical carcinomas: an immunohistochemical study of 155 cases. Diagn Pathol 10:8 Noel J et al (2001) Evidence of human papilloma virus infection but lack of Epstein-Barr virus in lymphoepithelioma-like carcinoma of uterine cervix: report of two cases and review of the literature. Hum Pathol 32(1):135–138 Nofech-Mozes S et al (2006) Immunohistochemical characterization of endocervical papillary serous carcinoma. Int J Gynecol Cancer 16(Suppl 1):286–292 Ojesina AI et al (2014) Landscape of genomic alterations in cervical carcinomas. Nature 506 (7488):371–375 Ostor AG (1995) Pandora’s box or Ariadne’s thread? Definition and prognostic significance of microinvasion in the uterine cervix. Squamous lesions. Pathol Annu 30 Pt 2:103–136 Ostor AG (2000) Early invasive adenocarcinoma of the uterine cervix. Int J Gynecol Pathol 19(1):29–38 Ozcan A et al (2011) PAX 8 expression in non-neoplastic tissues, primary tumors, and metastatic tumors: a comprehensive immunohistochemical study. Mod Pathol 24(6):751–764 Park KJ et al (2009) Immunoprofile of adenocarcinomas of the endometrium, endocervix, and ovary with mucinous differentiation. Appl Immunohistochem Mol Morphol 17(1):8–11 Park KJ et al (2011) Unusual endocervical adenocarcinomas: an immunohistochemical analysis with molecular detection of human papillomavirus. Am J Surg Pathol 35(5):633–646 Parwani AV et al (2005) Cervical adenoid basal tumors comprised of adenoid basal epithelioma associated with various types of invasive carcinoma: clinicopathologic features, human papillomavirus DNA detection, and P16 expression. Hum Pathol 36(1):82–90 Pereira TC et al (2011) Can we tell the site of origin of metastatic squamous cell carcinoma? An immunohistochemical tissue microarray study of 194 cases. Appl Immunohistochem Mol Morphol 19(1):10–14 Perez CA et al (1998) Tumor size, irradiation dose, and long-term outcome of carcinoma of uterine cervix. Int J Radiat Oncol Biol Phys 41(2):307–317 Perez-Montiel D et al (2012) Adenocarcinoma metastatic to the uterine cervix: a case series. J Obstet Gynaecol Res 38(3):541–549 Petry KU et al (1994) Cellular immunodeficiency enhances the progression of human papillomavirus-associated cervical lesions. Int J Cancer 57(6):836–840 Pirog EC et al (2000) Prevalence of human papillomavirus DNA in different histological subtypes of cervical adenocarcinoma. Am J Pathol 157(4):1055–1062 Poynor EA et al (2006) Clinicopathologic features of early adenocarcinoma of the cervix initially managed with cervical conization. Gynecol Oncol 103(3): 960–965 Prempree T, Villasanta U, Tang CK (1980) Management of adenoid cystic carcinoma of the uterine cervix (cylindroma): report of six cases and reappraisal of all

373 cases reported in the medical literature. Cancer 46(7): 1631–1635 Pretorius R et al (1991) Presentation of cervical cancer. Gynecol Oncol 42(1):48–53 Quint W et al (2012) One virus, one lesion–individual components of CIN lesions contain a specific HPV type. J Pathol 227(1):62–71 Rabban JT et al (2010) PAX2 distinguishes benign mesonephric and mullerian glandular lesions of the cervix from endocervical adenocarcinoma, including minimal deviation adenocarcinoma. Am J Surg Pathol 34(2): 137–146 Randall ME et al (1986) Papillary squamous cell carcinoma of the uterine cervix: a clinicopathologic study of nine cases. Int J Gynecol Pathol 5(1):1–10 Robboy SJ et al (1984) Atypical vaginal adenosis and cervical ectropion. Association with clear cell adenocarcinoma in diethylstilbestrol-exposed offspring. Cancer 54(5):869–875 Roma AA et al (2016) New pattern-based personalized risk stratification system for endocervical adenocarcinoma with important clinical implications and surgical outcome. Gynecol Oncol 141(1):36–42 Rush D et al (2005) Detection of microinvasion in vulvar and cervical intraepithelial neoplasia using double immunostaining for cytokeratin and basement membrane components. Arch Pathol Lab Med 129(6): 747–753 Russell MJ, Fadare O (2006) Adenoid basal lesions of the uterine cervix: evolving terminology and clinicopathological concepts. Diagn Pathol 1:18 Saad RS et al (2009) CDX-2 expression is a common event in primary intestinal-type endocervical adenocarcinoma. Am J Clin Pathol 132(4):531–538 Sal V et al (2016) Primary signet ring cell carcinoma of the cervix: a case report and review of the literature. Int J Surg Case Rep 21:1–5 Schwartz SM et al (2001) Human papillomavirus and prognosis of invasive cervical cancer: a populationbased study. J Clin Oncol 19(7):1906–1915 Sedlis A et al (1979) Microinvasive carcinoma of the uterine cervix: a clinical-pathologic study. Am J Obstet Gynecol 133(1):64–74 Sevin BU et al (1992) Microinvasive carcinoma of the cervix. Cancer 70(8):2121–2128 Silver SA et al (2001) Mesonephric adenocarcinomas of the uterine cervix: a study of 11 cases with immunohistochemical findings. Am J Surg Pathol 25(3): 379–387 Silverberg SG, Hurt WG (1975) Minimal deviation adenocarcinoma (“adenoma malignum”) of the cervix: a reappraisal. Am J Obstet Gynecol 121(7):971–975 Smedts F et al (1992) Keratin expression in cervical cancer. Am J Pathol 141(2):497–511 Smith HO et al (2000) The rising incidence of adenocarcinoma relative to squamous cell carcinoma of the uterine cervix in the United States–a 24-year population-based study. Gynecol Oncol 78(2):97–105 Smith HO et al (2002) Is there a difference in survival for IA1 and IA2 adenocarcinoma of the uterine cervix? Gynecol Oncol 85(2):229–241

374 Snijders-Keilholz A et al (2005) Primitive neuroectodermal tumor of the cervix uteri: a case report – changing concepts in therapy. Gynecol Oncol 98(3):516–519 Spoozak L et al (2012) Microinvasive adenocarcinoma of the cervix. Am J Obstet Gynecol 206(1):80.e1–80.e6 Sullivan LM et al (2008) Comprehensive evaluation of CDX2 in invasive cervical adenocarcinomas: immunopositivity in the absence of overt colorectal morphology. Am J Surg Pathol 32(11):1608–1612 Tacha D, Zhou D, Cheng L (2011) Expression of PAX8 in normal and neoplastic tissues: a comprehensive immunohistochemical study. Appl Immunohistochem Mol Morphol 19(4):293–299 Takeshima N et al (1999) Assessment of the revised International Federation of Gynecology and obstetrics staging for early invasive squamous cervical cancer. Gynecol Oncol 74(2):165–169 Talia KL, Cretney A, McCluggage WG (2014) A case of HPV-negative intestinal-type endocervical adenocarcinoma in situ with coexisting multifocal intestinal and gastric metaplasia. Am J Surg Pathol 38(2):289–291 Tambouret R, Clement PB, Young RH (2003) Endometrial endometrioid adenocarcinoma with a deceptive pattern of spread to the uterine cervix: a manifestation of stage IIb endometrial carcinoma liable to be misinterpreted as an independent carcinoma or a benign lesion. Am J Surg Pathol 27(8):1080–1088 Thomas MB et al (2008) Clear cell carcinoma of the cervix: a multi-institutional review in the post-DES era. Gynecol Oncol 109(3):335–339 Thomas LK et al (2014) Chromosomal gains and losses in human papillomavirus-associated neoplasia of the lower genital tract – a systematic review and metaanalysis. Eur J Cancer 50(1):85–98 Tjalma WA et al (2013) Differences in human papillomavirus type distribution in high-grade cervical intraepithelial neoplasia and invasive cervical cancer in Europe. Int J Cancer 132(4):854–867 Tringler B et al (2004) Evaluation of p16INK4a and pRb expression in cervical squamous and glandular neoplasia. Hum Pathol 35(6):689–696 Tseng CJ et al (1997) Lymphoepithelioma-like carcinoma of the uterine cervix: association with Epstein-Barr virus and human papillomavirus. Cancer 80(1):91–97 Ueda G et al (1989) An immunohistochemical study of small-cell and poorly differentiated carcinomas of the cervix using neuroendocrine markers. Gynecol Oncol 34(2):164–169 Ueda Y et al (2008) Clonality and HPV infection analysis of concurrent glandular and squamous lesions and adenosquamous carcinomas of the uterine cervix. Am J Clin Pathol 130(3):389–400 Ueno S et al (2013) Absence of human papillomavirus infection and activation of PI3K-AKT pathway in cervical clear cell carcinoma. Int J Gynecol Cancer 23(6):1084–1091 Vizcaino AP et al (1998) International trends in the incidence of cervical cancer: I. Adenocarcinoma and adenosquamous cell carcinomas. Int J Cancer 75(4):536–545

E. C. Pirog et al. Walboomers JM et al (1999) Human papillomavirus is a necessary cause of invasive cervical cancer worldwide. J Pathol 189(1):12–19 Wang HL, Lu DW (2004) Detection of human papillomavirus DNA and expression of p16, Rb, and p53 proteins in small cell carcinomas of the uterine cervix. Am J Surg Pathol 28(7):901–908 Wang TY et al (2001) Histologic and immunophenotypic classification of cervical carcinomas by expression of the p53 homologue p63: a study of 250 cases. Hum Pathol 32(5):479–486 Wang X et al (2014) The significant diagnostic value of human telomerase RNA component (hTERC) gene detection in high-grade cervical lesions and invasive cancer. Tumour Biol 35(7):6893–6900 Weichert W et al (2009) Molecular HPV typing as a diagnostic tool to discriminate primary from metastatic squamous cell carcinoma of the lung. Am J Surg Pathol 33(4):513–520 Wright AA et al (2013) Oncogenic mutations in cervical cancer: genomic differences between adenocarcinomas and squamous cell carcinomas of the cervix. Cancer 119(21):3776–3783 Xing D et al (2016) Lower female genital tract tumors with adenoid cystic differentiation: P16 expression and high-risk HPV detection. Am J Surg Pathol 40 (4):529–536 Yemelyanova A et al (2009) Endocervical adenocarcinomas with prominent endometrial or endomyometrial involvement simulating primary endometrial carcinomas: utility of HPV DNA detection and immunohistochemical expression of p16 and hormone receptors to confirm the cervical origin of the corpus tumor. Am J Surg Pathol 33(6):914–924 Yoneda JY et al (2015) Surgical treatment of microinvasive cervical cancer: analysis of pathologic features with implications on radicality. Int J Gynecol Cancer 25(4): 694–698 Yorganci A et al (2003) A case report of multicentric verrucous carcinoma of the female genital tract. Gynecol Oncol 90(2):478–481 Young RH, Clement PB (2002) Endocervical adenocarcinoma and its variants: their morphology and differential diagnosis. Histopathology 41(3):185–207 Young RH, Scully RE (1993) Minimal-deviation endometrioid adenocarcinoma of the uterine cervix. A report of five cases of a distinctive neoplasm that may be misinterpreted as benign. Am J Surg Pathol 17(7): 660–665 Zaino RJ et al (1992) Histopathologic predictors of the behavior of surgically treated stage IB squamous cell carcinoma of the cervix. A Gynecologic Oncology Group study. Cancer 69(7):1750–1758 Zhou C et al (1998) Papillary serous carcinoma of the uterine cervix: a clinicopathologic study of 17 cases. Am J Surg Pathol 22(1):113–120 Zivanovic O et al (2009) Small cell neuroendocrine carcinoma of the cervix: analysis of outcome, recurrence pattern and the impact of platinum-based combination chemotherapy. Gynecol Oncol 112(3):590–593

7

Benign Diseases of the Endometrium Ricardo R. Lastra, W. Glenn McCluggage, and Lora Hedrick Ellenson

Contents Embryology and Anatomy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 377 Vascular Anatomy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 378 Congenital Defects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 378 Fusion Defects of the Müllerian Ducts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 378 Atresia of the Müllerian Ducts and the Vagina . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 378 Normal Cyclical Endometrium . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Proliferative Phase . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Secretory Phase . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Menstrual Phase . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

379 380 380 383

Lower Uterine Segment Endometrium . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 384 Steroid Hormone, Steroid Hormone Receptor, and Immunopeptide Interactions in the Endometrial Cycle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 384 Immunohistochemistry of the Normal Endometrium . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 386 Hematopoietic Cells Within the Endometrium . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 386 Gestational Endometrium . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 387 Endometrium Associated with Ectopic Pregnancy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 390

R. R. Lastra (*) Department of Pathology, University of Chicago, Chicago, IL, USA e-mail: [email protected] W. G. McCluggage Department of Pathology, Royal Group of Hospitals Trust, Belfast, UK e-mail: [email protected] L. Hedrick Ellenson Department of Pathology and Laboratory Medicine, Division of Gynecologic Pathology, Weill Cornell Medical College and New York Presbyterian Hospital, New York, NY, USA e-mail: [email protected] # Springer Nature Switzerland AG 2019 R. J. Kurman et al. (eds.), Blaustein’s Pathology of the Female Genital Tract, https://doi.org/10.1007/978-3-319-46334-6_7

375

376

R. R. Lastra et al. Postmenopausal Endometrium . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 390 Endometrial Sampling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 391 Criteria for Adequacy of Endometrial Sample . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 392 Artifacts in Endometrial Biopsy Specimens . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 393 Contaminants and Other Elements in Endometrial Biopsies . . . . . . . . . . . . . . . . . . . . . . 394 Extrauterine Tissues in Endometrial Biopsy Specimens . . . . . . . . . . . . . . . . . . . . . . . . . . . . 395 Endometritis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 396 Specific Forms of Endometritis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Chlamydia Trachomatis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Cytomegalovirus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Herpes Simplex Virus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Mycoplasma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Actinomyces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Fungi and Parasites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Malakoplakia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Lymphoma-Like Lesion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Endometrial Granulomas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

398 398 398 399 399 399 400 400 401 401

Ligneous (Pseudomembranous) Endometritis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 402 Dysfunctional Uterine Bleeding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 402 Estrogen-Related DUB, Including Endometrium Associated with Anovulatory Cycles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 404 Progesterone-Related DUB: Luteal Phase Defects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 405 Effects of Exogenous Hormonal Agents and Drugs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Estrogen-Only HRT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Combined Estrogen and Progestin HRT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Progestin-Only Compounds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Progestin-Like Effect Without Exogenous Hormone Use . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Gonadotropin-Releasing Hormone Agonists . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Androgens . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Progesterone Receptor Modulators . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Tamoxifen . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Taxanes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

406 406 407 408 408 409 409 409 409 411

Endometrial Epithelial Metaplasia (Epithelial Cytoplasmic Change) . . . . . . . . . . . . . Squamous Metaplasia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Mucinous Metaplasia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Ciliated (Tubal) Metaplasia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Clear Cell Metaplasia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Hobnail Cell Metaplasia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Eosinophilic (Oxyphilic, Oncocytic) Metaplasia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Papillary Syncytial Metaplasia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Arias-Stella Reaction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Papillary Proliferation of the Endometrium . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

411 412 413 414 415 415 415 416 416 416

Endometrial Mesenchymal Metaplasia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Smooth Muscle Metaplasia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Cartilaginous and Osseous Metaplasia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Glial Metaplasia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Adipose Metaplasia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Extramedullary Hematopoiesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

417 417 417 418 418 418

Endometrial Polyps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 418 Endometrial Polyp with Atypical Stromal Cells . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 421 Adenomyomatous Polyp . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 421

7

Benign Diseases of the Endometrium

377

Hyperplasia and Carcinoma Arising in an Endometrial Polyp . . . . . . . . . . . . . . . . . . . . . . . . . 421 Atypical Polypoid Adenomyoma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 422 Adenofibroma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 424 Effects of IUD . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 425 Effects of Mirena Coil . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 426 Radiation Effects on the Endometrium . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 427 Effects of Endometrial Ablation or Resection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 427 Effects of Curettage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 428 Asherman’s Syndrome . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 428 Postoperative Spindle Cell Nodule . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 429 Psammoma Bodies in the Endometrium . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 429 Emphysematous Endometritis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 429 Benign Endometrial Stromal Proliferations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 429 Benign Trophoblastic Lesions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 430 Intravascular Endometrium . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 430 Endometrial Autolysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 431 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 431

Embryology and Anatomy The endometrium and the myometrium are of mesodermal origin and are formed secondary to fusion of the müllerian (paramesonephric) ducts between the 8th and 9th postovulatory weeks. Likewise, the cervix is of müllerian origin. The squamous epithelial lining of the ectocervix and upper two-thirds of the vagina (müllerian vagina) are of vaginal müllerian origin, while the squamous epithelial lining of the lower third of the vagina (sinus vagina) develops from the urogenital sinus; the endocervical glandular epithelium has recently been shown to be of uterine müllerian origin (Fluhmann 1960; Fritsch et al. 2013). Until the 20th week of gestation, the endometrium consists of a single layer of columnar epithelium supported by a thick layer of fibroblastic stroma. After the 20th gestational week, the surface epithelium invaginates into the underlying stroma, forming glandular structures that extend toward the underlying myometrium. The uterus, which is made up of the uterine corpus and uterine cervix, measures approximately 4 cm in length at birth,

and in newborns, the cervix makes up the majority of the uterus. At this stage, the endometrium measures less than 0.5 mm in thickness, and the surface and glands are lined by a low columnar to cuboidal epithelium devoid of either proliferative or secretory activity, which resembles the inactive endometrium of postmenopausal women. During the prepubertal years, the endometrium remains inactive, and the cervix continues to comprise the major part of the uterus. In the reproductive years, the dimensions and weight of a normal uterus vary widely according to parity. In nulliparous women, the uterus measures approximately 8 cm in length, 5 cm in width at the level of the fundus, and 2.5 cm in thickness; most weigh between 40 and 100 g. Multigravid uteri are larger with increasing length and weight with increasing parity. The internal os, a fibromuscular junction, separates the muscular uterine corpus from the fibrous uterine cervix. The uterine corpus is divided into the fundus, body, and isthmus. The fundus is that part of the uterus above the orifices of the fallopian tubes, and the isthmus represents the lower uterine segment. The uterus is located between the rectum (posteriorly)

378

and the urinary bladder (anteriorly); it is supported by the round ligaments and the utero-ovarian ligaments and covered by the pelvic peritoneum. The endometrium during the reproductive period undergoes cyclical morphological changes (described in detail below), which are particularly evident in the superficial two-thirds, the so-called functionalis layer. Morphological alterations are minimal in the deeper one-third, the so-called basalis layer. In postmenopausal women, the endometrial morphology is similar to that in the prepubertal years (see section “Postmenopausal Endometrium”).

Vascular Anatomy The endometrium has an abundant vascular supply that originates from the radial arteries of the underlying myometrium. These arteries penetrate the endometrium at regular intervals and give rise to the basal arteries, which in turn divide into horizontal and vertical branches, the former providing the blood supply to the endometrial basalis and the latter to the overlying functionalis layer. The endometrial vessels in the functionalis layer are referred to as spiral arteries. Their development and arborization near the endometrial surface and their connections with the subsurface epithelial precapillary system, as well as extreme coiling during the menstrual cycle, are influenced by ovarian steroid hormones and prostaglandins. A differentiating feature between the endometrial and myometrial arteries is the absence of subendothelial elastic tissue in the endometrial arteries, except for those in the basal layer, and its presence in the myometrial arteries. Veins and lymphatics are closely associated with the endometrial arteries and glands, respectively. Uterine lymphatics drain from subserosal uterine plexuses to the pelvic and para-aortic lymph nodes.

R. R. Lastra et al.

(DES) (Kaufman et al. 1977), or imbalances in endogenous hormones associated with abnormal gonads and chromosomal defects. In utero exposure to DES is, of course, almost currently nonexistent. Genotypically normal females with normal gonads may also have müllerian duct abnormalities. These developmental aberrations, such as defects in the fusion of the müllerian ducts, are caused by errors in embryogenesis. The etiology of these developmental errors is mostly unknown, but hormonal imbalances or genetic abnormalities may be implicated. These disorders are frequently associated with malformations in the urinary system and the distal gastrointestinal tract. For practical purposes, müllerian duct abnormalities can be divided into two categories, namely, abnormalities of fusion and abnormalities caused by atresia.

Fusion Defects of the Müllerian Ducts Normally, the upper two-thirds of the vagina and the uterus are formed by fusion of the paired müllerian ducts. After fusion, the intervening wall degenerates, forming the endometrial cavity and the upper vaginal canal. Nonfusion of the müllerian ducts results in a bicornuate uterus. If the ducts fuse but the wall between the two lumens persists, an abnormal septate uterus results. Occasionally, a carcinoma develops in one cavity, and only the other normal cavity is sampled during the investigation of abnormal uterine bleeding. If the defect is minor or confined to the fundus, the uterus is referred to as arcuatus. If the full length of the uterus and the upper vagina is divided by a septum, the condition results in uterus didelphys with a partially double vagina. These congenital anomalies may result in infertility or spontaneous abortion and in some cases require surgical correction.

Atresia of the Müllerian Ducts and the Vagina

Congenital Defects Congenital abnormalities of the uterus are uncommon. They may be secondary to the in utero effects of exogenous hormones, such as diethylstilbestrol

Atresia of the müllerian ducts and the vagina may be partial or complete. The etiology of these conditions is obscure, although a genetic cause is suggested in families with multiple affected

7

Benign Diseases of the Endometrium

siblings. The pattern of inheritance may be autosomal recessive or dominant. In cases of bilateral müllerian duct atresia, the upper genital tract may consist of bilateral non-canalized muscular tissue located on the lateral pelvic walls. In Mayer–Rokitansky–Küster–Hauser syndrome, a severe defect characterized by müllerian and vaginal aplasia, patients may have urinary tract anomalies such as a pelvic kidney or anephria. Vertebral and other skeletal abnormalities may also be present, suggesting a more generalized morphogenetic abnormality. If only one of the müllerian ducts is involved, the affected side will show a portion of tubal fimbria and a small muscular mass at the pelvic sidewall. Occasionally, a rudimentary structure remains as an appendage attached to the unaffected side, giving rise to a uterus bicornis unicollis, which results in a pelvic mass and cyclic pelvic pain associated with menses. Patients with these conditions are endocrinologically normal with normal gonads. It has been postulated that activating mutations affecting the gene coding for antimüllerian hormone or its receptor may be related to the development of these syndromes (Lindenman et al. 1997). If the anomaly is associated with obstruction of the vagina and functional endometrial tissue is present, hydrocolpos may be present at birth, or patients may present with primary amenorrhea. A number of multiple malformation syndromes have been associated with müllerian or vaginal agenesis. Winter syndrome, a genetically inherited autosomal recessive disorder, is characterized by vaginal agenesis, renal agenesis, and middle ear anomalies (Winter et al. 1968). Management of patients with complete vaginal atresia requires surgery to create a neovagina. If the anomaly is isolated vaginal atresia, as most commonly occurs, the patient will usually be fertile if a normal uterus and fallopian tubes are present.

Normal Cyclical Endometrium In the reproductive years, the endometrium is characterized by cyclical proliferation, differentiation, and shedding in response to estrogen and progesterone secretion by the ovaries. Endometrial morphology, as a consequence, is

379

continually changing depending on the levels of estrogen and progesterone (Crum et al. 2003; Noyes et al. 1950). During the proliferative phase of the menstrual cycle, the endometrium has a relatively constant morphology, which does not differ significantly from day to day; as such, accurate dating is not possible in the proliferative phase. Following ovulation, the morphological appearances in the secretory phase have been considered relatively specific from day to day, such that it is possible to accurately date secretory phase endometrium to within 1 or 2 days. However, this view has been challenged with one study finding that traditional endometrial histological dating criteria are much less temporally distinct and discriminating than originally described (Murray et al. 2004). Presently, most endometrial biopsies are performed during the investigation of abnormal uterine bleeding, and it is relatively uncommon to be asked to date the endometrium, although formerly this was often requested in the investigation of infertility. The typical endometrial cycle is 28 days, although the length varies in individual women and between women. In general, the differences in cycle length are due to variation in the duration of the proliferative phase, the secretory phase usually being constant and lasting 14 days from the time of ovulation to the onset of menstruation. In the reproductive years, the endometrium is divided into two regions, namely, the superficial functionalis (stratum spongiosum) and the basalis (stratum basale). The former exhibits the greatest degree of hormonal responsiveness, while the latter is much more unresponsive, the morphology not varying greatly during the menstrual cycle; as such, a biopsy consisting entirely of basalis is not adequate for dating of the menstrual cycle. Usually the endometrial glands are regularly spaced and have a perpendicular arrangement from the basalis to the surface. The basalis abuts the myometrium and regenerates the functionalis following its shedding during menstruation. The basalis is composed of inactive appearing glands, cellular stroma, and spiral arteries that have thicker muscular walls than those in the functionalis. Accurate typing of the endometrium is, in general, not possible when polyps, endometritis, or other pathological lesions are present.

380

Proliferative Phase The onset of menstruation is the first day of the menstrual cycle. Following menstruation, which varies in length, the uterus is lined by shallow basal endometrium and the deeper part of the functionalis. The endometrium begins to proliferate on the third or fourth day of the cycle, and during the proliferative phase, it increases in thickness up to 4 or 5 mm. Between the 5th and 14th days of a typical 28-day cycle, there is glandular, stromal, and vascular growth, with the endometrium progressively increasing in depth until ovulation. The endometrial glands are uniform and widely and regularly spaced, and have a simple tubular architecture which can be appreciated on cross section (Fig. 1). An occasional mildly dilated gland is a normal feature and of no significance. Mitotic figures are easily identified within the glands, and the presence of mitotic activity should be confirmed before labeling an endometrium as proliferative in type. The glandular epithelium is composed of pseudostratified cuboidal or low columnar cells with a moderate amount of basophilic cytoplasm. The nuclei are oval or rounded, may contain small nucleoli, and are orientated perpendicular to the basement membrane. Estrogenic activity during the proliferative phase often results in focal ciliation of the surface epithelial cells; thus, surface ciliated cells are a feature of normal proliferative endometrium, and this does not indicate ciliated or tubal

Fig. 1 Proliferative endometrium. Widely spaced tubular glands with low columnar cells, exhibiting mitotic activity

R. R. Lastra et al.

metaplasia (see section “Endometrial Epithelial Metaplasia (Epithelial Cytoplasmic Change)”). Proliferative activity is maximal between the 8th and 10th days of the cycle, and at this stage, the glandular epithelium becomes more stratified and mitoses are more frequently seen. In the lateproliferative phase, the glands become progressively more convoluted and tortuous and appear more variable in size and shape; however, they remain tubular in configuration. Occasional subnuclear vacuoles may be seen. During the proliferative phase, the endometrial stroma is usually densely cellular, and the stromal cells are small and oval with hyperchromatic nuclei and indistinct cytoplasm and cell borders. Mitotic figures are present within the stroma, although less numerous than within the glands. Scanty thin-walled stromal blood vessels are present. Lymphoid aggregates resembling follicles can be seen within the stroma in the proliferative phase of the cycle. The degree of mitotic activity within both the glands and stroma decreases in the late-proliferative phase; simultaneously, early stromal edema develops.

Secretory Phase Secretory endometrium is characterized by glandular secretion, stromal maturation, and vascular differentiation occurring in response to progesterone produced by the postovulatory corpus luteum. The endometrium increases in thickness up to 7 or 8 mm. The secretory phase may be divided into three stages, namely, the early secretory phase (from the 2nd to 4th postovulatory day, days 16–18 of a normal 28-day cycle), the mid-secretory phase (from the 5th to 9th postovulatory day, days 19–23 of a normal 28-day cycle), and the late-secretory phase (from the 10th to 14th postovulatory day, days 24–28 of a normal 28-day cycle). These phases are continuous and not sharply demarcated, and some areas of the endometrium may appear at a slightly more advanced stage than others. The major morphological features that occur in the endometrium during the secretory phase are described in Table 1. There is an interval of 36–48 h between ovulation and the first

7

Benign Diseases of the Endometrium

381

Table 1 Morphological landmarks in secretory endometrium according to cycle day Cycle day 16

Phase Early secretory

17

Early secretory

18

Early secretory

19

Midsecretory

20–21

Midsecretory

22

Midsecretory

23

Midsecretory

24

Late secretory

25

Late secretory Late secretory

26

27–28

Late secretory

Morphological features Irregular basal vacuoles Pseudostratified nuclei Numerous mitoses present Even subnuclear vacuoles Aligned nuclei Occasional mitoses Supranuclear (luminal) vacuoles Pseudostratified nuclei Mitoses rare Straight glands Few remaining supranuclear vacuoles Intraluminal secretions Absent mitoses Angulated glands Prominent intraluminal secretions Absent mitoses Increasing stromal edema Peak stromal edema with “naked” stromal cells Centrally located inspissated luminal secretions Decreasing stromal edema Spiral arterioles prominent Earliest predecidual change around spiral arterioles Serrated glands with secretory exhaustion Minimal or absent stromal edema Prominent predecidual change around spiral arterioles with vascular groups bridged by predecidua Focal predecidua under surface epithelium Extensive predecidual change under surface epithelium Occasional mitoses reappear Mild stromal inflammatory infiltrate Diffuse predecidual change Increasing number of mitoses Prominent stromal inflammatory infiltrate Fibrin thrombi Foci of hemorrhage

morphologically recognizable signs of early secretory activity. In the early secretory phase, the endometrial glands still have a tubular appearance, and mitotic activity may be identified. The initial

Fig. 2 Early secretory phase endometrium. Tubular glands exhibit subnuclear vacuolation

morphological feature of ovulation is the appearance within the glandular epithelium of subnuclear vacuoles. These typically appear on the 16th day of the typical 28-day cycle, that is, the 2nd postovulatory day. Initially, subnuclear vacuoles are identified only within occasional cells and are irregularly distributed but are usually most obvious in the mid-zone of the functionalis. There is a progressive increase in the number and distribution of subnuclear vacuoles until they involve almost all cells within most glands in the functionalis (Fig. 2). Subnuclear vacuoles are maximal between the 17th and 18th day of the cycle (3rd and 4th postovulatory days). As stated, some areas of the endometrium may appear at a slightly more advanced stage than others, and in the early secretory phase, there may be an admixture of glands exhibiting proliferative and secretory activity; in fact, an individual gland may exhibit both mitotic activity and subnuclear vacuolation. It is generally assumed that ovulation has occurred when there are subnuclear vacuoles in at least 50% of the cells in at least 50% of the glands; scattered subnuclear vacuoles are not reliable evidence of ovulation and, as stated, may be seen in late-proliferative endometrium. In the early secretory phase, the stroma is indistinguishable from that of late-proliferative endometrium. Between the 19th and 23rd day of a typical 28-day cycle (the mid-secretory phase), the degree of glandular secretion increases. Cytoplasmic vacuoles become supranuclear, and secretions are seen

382

R. R. Lastra et al.

Fig. 3 Mid-secretory endometrium. The glands contain supranuclear vacuoles with secretions within the glandular lumina

within the glandular lumina (Fig. 3); it is important to realize that secretory material within the glandular lumina is not specific to secretory endometrium, but may also be seen in proliferative, hyperplastic, and malignant endometria. Mid-secretory glands are usually angular in shape, and mitotic activity is no longer apparent. The glands in the superficial layers of the functionalis tend to exhibit less secretory activity, and, as a result, superficial biopsies may produce a false impression of poorly developed secretory activity. Stromal edema progressively increases, is most obvious in days 22 and 23, and is most prominent in the mid-zone (Fig. 4). Spiral arteries become apparent. At this stage, the stromal cells surrounding spiral arteries acquire a more conspicuous eosinophilic cytoplasm; these cells are referred to as predecidual cells. In the late-secretory phase (days 24–28 of a typical 28-day cycle or 10th to 14th postovulatory days), there is typically diminution of glandular secretory activity (secretory exhaustion), and the glands become serrated. Predecidual stromal change increases, initially being most apparent in the cells surrounding the spiral arteries (Fig. 5). The predecidual change results in the formation of the so-called compact layer (stratum compactum) beneath the surface epithelium, the deeper layers of stroma exhibiting less predecidual change. Sometimes, the predecidual cells may have a spindle cell or even signet-ring morphology and may not be readily appreciated as predecidual cells. Occasional mitoses may

Fig. 4 Mid-secretory endometrium. There is marked stromal edema with “naked” stromal cells

Fig. 5 Late-secretory endometrium. There is predecidual change surrounding the spiral arteries and focally beneath surface epithelium

reappear in the predecidual stromal cells on day 26 or 27. A stromal infiltrate of granulated lymphocytes (described in detail later) is now obvious, and occasional neutrophils may appear in the premenstrual phase; the presence of granulated lymphocytes and neutrophils should not be misinterpreted as evidence of an endometritis.

7

Benign Diseases of the Endometrium

Fig. 6 Late-secretory endometrium. The glands may be closely packed superficially resembling a hyperplastic endometrium

In late-secretory endometrium, the glands may be closely packed (this impression can be exacerbated by tangential sectioning), and this can superficially resemble hyperplastic endometrium (Fig. 6); however, other features of a hyperplastic endometrium, such as mitotic activity, are absent. In some late-secretory endometria, the glands have a “hypersecretory” appearance, resembling Arias-Stella reaction. This should not, in itself, be interpreted as evidence of early pregnancy. In the immediate premenstrual days, apoptotic activity is seen within the glands, fibrin thrombi appear in the small blood vessels, and there is extravasation of red blood cells into the stroma.

Menstrual Phase Menstruation occurs after the 28th day of the normal cycle (the onset of menstruation is the first day of the menstrual cycle) and is characterized by glandular and stromal breakdown. Menstruation usually lasts for about 4 days. The endometrial glands are serrated and collapsed. Some of the glands remain vacuolated, imparting their secretory appearance. The stroma is condensed and collapsed, and the stromal cells aggregate into tightly packed balls (stromal blue balls) and separate from the glands (Fig. 7); the presence of tightly aggregated balls of stromal cells with hyperchromatic nuclei may be worrisome to the unwary. The predecidual appearance of the stromal cells is lost. Other features

383

Fig. 7 Menstrual phase endometrium. The stromal cells aggregate into “blue balls”

Fig. 8 Menstrual phase endometrium. Apoptotic bodies are present within the endometrial glands and stroma

include the presence of necrotic debris, neutrophil infiltration, interstitial hemorrhage, and fibrin deposition. Apoptotic bodies are identified within both the glands and the stroma (Fig. 8). When menstrual activity is well developed, little or no stromal tissue may remain, and the glands become closely packed, sometimes with a back-to-back appearance. To the inexperienced, this may result in consideration of a hyperplasia or carcinoma. Following breakdown, the endometrial glands assume a surface micropapillary architecture (Fig. 9). The term papillary syncytial metaplasia (discussed later) is used for this appearance, but this is an inaccurate term since this is not strictly speaking a metaplasia but rather a regenerative/degenerative process secondary to tissue breakdown. Mitotic figures may be present within the papillary proliferations. Papillary syncytial metaplasia is also seen following surface

384

Fig. 9 Papillary syncytial metaplasia. Following breakdown, the endometrial glands regenerate with a surface micropapillary architecture: neutrophils are often seen within the epithelium

breakdown associated with non-menstrual conditions. On occasion, the micropapillary architecture is particularly striking and, if associated with mitotic activity, raises the possibility of a serous carcinoma. The appreciation of the accompanying features of breakdown and immunohistochemical staining with p53 may be of value. Most serous carcinomas exhibit diffuse, intense, nuclear immunoreactivity with p53, while papillary syncytial metaplasia exhibits a wild-type staining pattern, which has been described as weak and heterogeneous (Quddus et al. 1999).

Lower Uterine Segment Endometrium Lower uterine segment or isthmic endometrium is poorly responsive to steroid hormones, and the morphology does not alter significantly during the menstrual cycle; as is the case with the endometrial basalis, lower uterine segment endometrium is not useful for dating the menstrual cycle. Lower uterine segment endometrium is composed of inactive poorly developed glands that are often ciliated. They are irregularly distributed, and some may be dilated. The stroma typically has a fibrous appearance, and the stromal cells are more elongate and “fibroblast-like” than in the corpus. Given these features, lower uterine segment endometrium may be mistaken for a polyp in a biopsy specimen. In the inferior part of the lower uterine segment, the

R. R. Lastra et al.

Fig. 10 Lower uterine segment endometrium. This is composed of a mixture of ciliated and mucinous glands within a fibrous stroma

glands merge with mucinous type glands from the upper endocervix, and the stroma becomes even more fibrous (Fig. 10).

Steroid Hormone, Steroid Hormone Receptor, and Immunopeptide Interactions in the Endometrial Cycle As detailed, the menstrual cycle in postmenarchal, premenopausal women follows a regular series of morphological and physiological events characterized by proliferation, secretory differentiation, shedding, and regeneration of the uterine lining. These alterations are controlled by the cyclical release of the steroid sex hormones estradiol (E2) and progesterone (P) from the ovaries; the endometrium is thus a highly sensitive indicator of the hypothalamic–pituitary–ovarian axis. Steroid hormone control of endometrial epithelial and stromal cells is mediated by estrogen receptors (ER) and progesterone receptors (PR). These steroid receptors are proteins concentrated in the nuclei of the endometrial epithelial and stromal cells that have high affinity to bind E2 and P. Because they are sex steroid hormone (ligand) specific, a particular receptor may display high affinity for a closely related class of hormones, and these classes may compete for available binding sites. For example, ER effectively binds not only E2 but also estrone (E1), as well as synthetic estrogens, such as DES.

7

Benign Diseases of the Endometrium

Although E2 plays a crucial role in the proliferation of endometrial cells in vivo, E2 alone is not able to induce the proliferation of endometrial cells in primary culture. It has been suggested that the mitogenic action of E2 is mediated indirectly via a paracrine effect by the polypeptide growth factor, epidermal growth factor (EGF). EGF promotes the transition of cells from the G0 to G1 phase of the cell cycle (Taketani and Mizuno 1991). Human endometrial cells possess EGF receptors and mRNA for EGF. EGF-like immunoreactivity is seen in both the endometrial epithelial and stromal cells, with higher concentrations in the epithelium than the stroma, and parallels the fluctuation of cyclic sex steroid hormones during the menstrual cycle. It appears that the EGF receptor content is regulated by ovarian E2 and P secretion (autocrine control). Indeed, EGF alone fails to influence cell proliferation, but in combination with E2, it increases the mean glandular but not stromal cell counts by more than 50% in vitro. The immunolocalization of EGF in normal human endometrium and the stimulation of epithelial cell proliferation in culture by EGF and E2 provide support for a role of EGF in endometrial growth. Similarly, endometrial stromal cells produce insulin-like growth factors 1 and 2 (IGF-1 and IGF-2) and high-affinity IGF binding proteins (IGFBP), which act on both epithelial cells and stromal cells via insulin-like growth factor receptors, stimulating proliferation, differentiation, and metabolic effect (Rutanen 1998). Estrogen stimulates IGF-1 gene expression in the endometrium and is presumed to mediate estrogen action, while IGF-2 gene expression is thought to be related to endometrial differentiation. IGFBP-1 is produced by decidualized endometrial stromal cells and inhibits the biological actions of IGF-1 and, as a result, inhibits the actions of estrogen. Therefore, clinical conditions characterized by the absence of IGFBP-1 may lead to unopposed IGF-1 action and subsequent endometrial proliferations and cancer development (Rutanen 1998). It has been shown that women with polycystic ovarian syndrome have increased endometrial expression of IGF-1 compared to controlled women, which may play a role in the higher incidence of endometrial carcinoma in this population (Shafiee et al. 2016).

385

The continuous dynamic remodeling of the endometrium results from a delicate balance of cellular proliferation and programmed cell death within specific subpopulations of stromal and epithelial cells, a process that is modulated by steroid hormones. A ladder pattern of DNA cleavage characteristic of apoptosis is seen in the latesecretory, menstrual, and early-proliferative phases (Shikone et al. 1997). Localization of apoptotic subpopulations using in situ assays for DNA breakage has shown that the majority of apoptotic cells represent glandular cells within the basalis and these cells increase in number throughout the secretory phase and peak during menses (Tabibzadeh et al. 1994). The apoptotic effects of steroid hormones are likely mediated through a complex network of inhibitors and initiators. Progestins have been shown to decrease endometrial secretion of the apoptosis inhibitor Bcl-2, a process that is reversed upon administration of antiprogestogenic agents (Critchley et al. 1999; Gompel et al. 1994). Progestins may also positively promote apoptosis by increasing levels of the apoptosis inducer gene BAK (Tao et al. 1998). The concentrations of ER and PR in the normal endometrium vary during the normal menstrual cycle according to fluctuating plasma levels of E2 and P. The highest values of ER (approximately 400 fmol/mg protein) and PR (approximately 1,000 fmol/mg protein) occur during the mid-proliferative phase (8th–10th day of the cycle) and correspond to rising plasma levels of E2. E2 promotes the synthesis of both ER and PR, whereas P inhibits the synthesis of ER. Monoclonal antibodies to ER allow the precise intracellular localization of ER by means of immunohistochemistry. Most ER is localized in the nuclei rather than the cytoplasm of endometrial epithelial and stromal cells. Endothelial cells fail to stain with ER antibodies. The concept of the mechanisms of sex steroid hormone–receptor action in target cells includes the following major steps: (1) circulating and unbound steroid hormone molecules are taken up from the cytoplasmic membrane, presumably by cytoplasmic receptors; (2) the hormone molecules enter the nucleus, which contains most (90–95%) of the cellular receptors; (3) the intranuclear hormone molecules induce conversion of the inactive

386

(nonfunctional) 4S form of receptor to the active (functional) 5S form of the receptor; (4) the hormonally activated 5S receptor binds to acceptor genes in the nucleus and influences gene expression by stimulating RNA polymerase and thus mRNA transcription; and (5) the newly formed mRNA is transported to the cytoplasm, where it is translated into proteins, including anabolic and catabolic enzymes, as well as new receptors (receptor replenishment). According to this concept, the most significant effect of sex hormones is intranuclear activation of receptors that in turn initiate a sequence of events, which results in alterations in the physiologic functions of target cells.

R. R. Lastra et al.

Fig. 11 ER in the endometrium. The normal endometrial glands and stroma are positive with ER

Immunohistochemistry of the Normal Endometrium The normal endometrial glands and stroma are ER and PR positive (Fig. 11). Endometrial glands are diffusely positive with the antiapoptotic protein Bcl-2 during the proliferative phase of the menstrual cycle. There is marked diminution in staining during the early and mid-secretory phases with reappearance during the late-secretory phase (Bozdogan et al. 2002; Gompel et al. 1994; Morsi et al. 2000). A minor population of endometrial epithelial cells exhibit nuclear immunoreactivity with p63; it has been speculated that these are reserve cells or basal cells and the origin of metaplastic endometrial epithelial cells (O’Connell et al. 2001). Endometrial glands are cytokeratin 7 positive and cytokeratin 20 negative. The normal endometrial stroma is CD10 positive (Fig. 12), Bcl-2 positive, and CD34 negative, in contrast to cervical stroma, which is largely CD10 and Bcl-2 negative and CD34 positive (Barroeta et al. 2007), although there may be immunoreactivity with CD10 of stromal cells surrounding endocervical glands (McCluggage et al. 2003a). The immunophenotype of lower uterine segment stroma overlaps with that of the endometrial and the cervical stroma. Normal endometrial stroma may be smooth muscle actin positive but is usually desmin negative. The

Fig. 12 CD10 in the endometrium. The normal endometrial stroma is CD10 positive

immunohistochemistry of endometrial hematopoietic cells is discussed in the next section.

Hematopoietic Cells Within the Endometrium The normal endometrium contains a variety of hematopoietic cells, the composition of which varies depending on the stage of the menstrual cycle and the menopausal status. Lymphocytes, including lymphoid aggregates and occasionally lymphoid follicles with germinal centers, are

7

Benign Diseases of the Endometrium

Fig. 13 Lymphoid aggregates in basal endometrium. Lymphoid aggregates are a normal phenomenon within the endometrial stroma

found at all stages of the menstrual cycle and in the postmenopausal endometrium. As mentioned above, in the normal menstrual cycle, lymphoid aggregates are most common in the proliferative phase. However, they are more commonly seen postmenopausally within the basal endometrium (Fig. 13); it is not clear whether this is due to lymphoid aggregates being more numerous postmenopausally or whether they are more obvious secondary to the glandular atrophy. Immunohistochemical studies have demonstrated that B lymphocytes (CD20 and CD79a positive) constitute approximately 1% of the normal lymphoid population of the endometrium and are present mainly in aggregates in the basalis and rarely as individual cells in the functionalis. T lymphocytes (CD3 positive) are more common and are present throughout the endometrial stroma, usually as individual cells, and are more numerous during the secretory phase (Bulmer et al. 1988; Disep et al. 2004). The distribution of B and T lymphoid cells is altered in endometritis (see section “Endometritis”). Granulated lymphocytes (CD56 positive) are present in large numbers in predecidualized endometrial stroma in the midand late-secretory phases (Bulmer et al. 1987, 1988); these were formerly designated endometrial stromal granulocytes and have a mononuclear or bilobed nucleus and abundant eosinophilic cytoplasm containing a variable number of granules. The function of granulated lymphocytes is not entirely clear, but they have the characteristics of natural killer cells and, as

387

well as being positive with CD56, are immunoreactive with T-cell markers. Neutrophils are present in small numbers throughout the menstrual cycle but only become morphologically evident in large numbers in association with the tissue breakdown and necrosis associated with menses. In contrast to lymphocytes and neutrophils, plasma cells are not considered to be a component of the normal endometrium, although the presence of an occasional plasma cell is allowable in an otherwise morphologically normal endometrium. Plasma cells are, of course, a feature of endometritis (see section “Endometritis”). Rare mast cells, demonstrable with toluidine blue or Giemsa staining, may be found in the endometrium, primarily within the basalis. Mast cells are also found normally in the myometrium, in endometrial polyps, and in leiomyomas, often in large numbers in the latter. The number of mast cells in the endometrium and myometrium tends to decrease with advancing age. Histiocytes are also seen in the normal endometrium (see section “Contaminants and Other Elements in Endometrial Biopsies”). Eosinophils are not a component of the normal endometrium. Rarely, foci of extramedullary hematopoiesis are present in the endometrium, usually in association with an underlying hematopoietic disorder or occasionally representing remnants of fetal tissue (Creagh et al. 1995; Valeri et al. 2002).

Gestational Endometrium Pregnancy is characterized by morphological changes involving the endometrial glands and stroma. The trophoblastic populations in an intrauterine gestation are covered in ▶ Chap. 20, “Gestational Trophoblastic Tumors and Related Tumorlike Lesions”. Early in pregnancy, the endometrium displays hypertrophic and hypersecretory features that have been referred to as “gestational hyperplasia.” The endometrium is characterized by (1) glandular ferning with epithelial and intraluminal secretions, (2) stromal edema and vascular congestion, and (3) decidual reaction of the stromal cells. The changes are similar to but quantitatively exaggerated

388

compared to those of nongestational endometrium at days 22–26 of the menstrual cycle. In the normal cycle, each of these alterations is prominent on a given day of the secretory phase, whereas during early pregnancy they occur simultaneously. The features described are not pathognomonic of early pregnancy but may occasionally be seen in other situations (see section “Arias-Stella Reaction”). Immediately after implantation of the blastocyst on the endometrial surface, changes begin to occur in the endometrial glands and the stroma, although the overall morphology of late-secretory endometrium is maintained for several days. The early changes include glandular serration and distension, increase in glandular secretions, stromal edema, and a stromal predecidual reaction. The morphological features during the first 2 weeks of gestation are subtle, but after approximately 15 days, they become more characteristic with the formation of decidual cells. Compared to predecidual cells, decidual cells are larger with prominent cell membranes and more abundant eosinophilic cytoplasm that may contain small vacuoles (Fig. 14). The nuclei of decidual cells are round to ovoid with finely dispersed chromatin and indistinct nucleoli. Stromal granulated lymphocytes are present during early pregnancy, and spiral arteries become more apparent; they have thicker walls than in nongestational secretory endometrium. Some of the spiral arteries display acute atherosis with concentric intimal proliferation of myofibroblasts and accumulation of foamy cells. As the pregnancy progresses, decidual cells become widespread with better defined cell borders and develop an epithelioid appearance. Small numbers of granulated lymphocytes remain throughout the gestation. Some of the glands become atrophic, while others have a hypersecretory appearance. Four to 8 weeks after implantation, the glands often exhibit, at least focally, the Arias-Stella reaction, which is a response to the presence of trophoblastic tissue in the uterus or at an ectopic site. Histologically, the Arias-Stella reaction is characterized by cellular stratification, secretory activity, vacuolated cytoplasm, and enlargement of the epithelial cell nuclei and cytoplasm (Fig. 15). The nuclei

R. R. Lastra et al.

Fig. 14 Gestational endometrium. The stroma is expanded and composed of decidualized cells with abundant eosinophilic cytoplasm

Fig. 15 Arias-Stella reaction in pregnancy. There is cellular stratification, vacuolated cytoplasm, and enlargement of the epithelial cell nuclei

may be enlarged up to three times normal and can exhibit considerable atypia and a hobnail appearance with bulging into the glandular lumina. Mitotic figures may be present, although these are rarely prominent, and there is a low MIB1 proliferation index. Atypical mitoses have rarely been described (Arias-Stella et al. 1994). The Arias-Stella reaction may be extensive, involving many glands, or focal with involvement of only a few glands or even part of a gland. The changes persist for at least 8 weeks following delivery. Several histological variants have been described, namely, atypical, early secretory pattern, hypersecretory pattern, regenerative pattern, and monstrous cell pattern (Arias-Stella 2002). However,

7

Benign Diseases of the Endometrium

there is considerable overlap between these patterns, and there is no value in attempting to subclassify Arias-Stella effect in the endometrium. The Arias-Stella reaction may also be seen in the glandular epithelium of the cervix and fallopian tube and involving endometriosis or vaginal adenosis (Nucci and Young 2004). Apart from the Arias-Stella reaction, the endometrial glands may undergo other changes in the presence of trophoblastic tissue. These include abundant clear glycogen-rich cytoplasm; this overlaps with the Arias-Stella reaction, but the nuclear enlargement of the latter is not present. Another pregnancy-related change is the presence of optically clear nuclei within the endometrial epithelium (Fig. 16) (Mazur et al. 1983). This may occur in association with the Arias-Stella reaction or independently. This appearance is due to the intranuclear accumulation of biotin and may simulate the ground-glass nuclei of herpes simplex virus infection (Yokoyama et al. 1993). However, the nuclei lack the Cowdry type A eosinophilic intranuclear inclusions and nuclear molding characteristic of herpes simplex virus infection. Due to the presence of biotin within these nuclei, a falsely positive immunohistochemical stain might result if immunohistochemistry is attempted in these cases, secondary to nonspecific binding of avidin to the nuclear biotin (Matias-Guiu et al. 1994).

Fig. 16 Optically clear nuclei in pregnancy. Endometrial glands in pregnancy may contain cells with optically clear nuclei

389

Localized endometrial glandular proliferations have also been described during pregnancy, usually in first trimester gestations of women in their fourth or fifth decades. These rare focal lesions are characterized by glandular expansion, nuclear stratification, a cribriform architecture, and intraluminal calcifications (Genest et al. 1995). There is mitotic activity, but the cytology is bland. These appear to be benign lesions based on uneventful follow-up and an unusual response to pregnancy. In early pregnancy, endometrial glands become strongly S100 positive (Nakamura et al. 1989). This immunoreactivity disappears after the 12th week of gestation. The reason for this S100 positivity is not clear. It is emphasized that the described endometrial morphological changes of pregnancy may be seen with both an intrauterine gestation and an ectopic pregnancy. Confirmation of an intrauterine gestation requires the presence of trophoblast, either in the form of chorionic villi or a placental site reaction (described in ▶ Chap. 20, “Gestational Trophoblastic Tumors and Related Tumorlike Lesions”). In the placental site, intermediate trophoblast infiltrates the decidua. On occasion, it may be difficult to distinguish between decidual cells and implantation site intermediate trophoblast, although intermediate trophoblastic cells are larger and more variable in size and shape, ranging from polygonal to spindle shaped. The nuclei are lobated, and hyperchromatic binucleate or multinucleate cells may be present. There may be prominent nucleoli and sharply defined cytoplasmic vacuoles. In contrast, the nuclei of decidualized stromal cells are uniform and round to oval with finely dispersed chromatin and inconspicuous nucleoli. In problematic cases, immunohistochemical staining may assist in distinguishing between intermediate trophoblastic cells at the placental site and decidualized stromal cells. Markers of value are discussed in ▶ Chap. 20, “Gestational Trophoblastic Tumors and Related Tumorlike Lesions,” but trophoblastic cells are immunoreactive with broad-spectrum cytokeratins, cytokeratin 7, human placental lactogen, HLA-G, inhibin, and mel-CAM (CD146), while decidua is negative (Kurman et al. 1984; McCluggage et al. 1998; O’Connor and Kurman 1988).

390

R. R. Lastra et al.

Endometrium Associated with Ectopic Pregnancy The morphology of gestational endometrium is discussed in the previous section. Endometrial changes also occur with an ectopic pregnancy. The features are variable, but by day 22–28 of the ectopic gestation, the endometrial glands usually exhibit secretory or hypersecretory features, sometimes with Arias-Stella reaction, at least focally. Occasionally, the endometrial glands are atrophic. If the ectopic trophoblast regresses, the endometrial glands have a variable appearance, ranging from proliferative to secretory to a picture identical to that seen in disordered proliferation or progestogen effect. The endometrial stroma in association with an ectopic gestation is usually decidualized and devoid of inflammation. Thickwalled spiral arteries are present. Trophoblastic tissue, in the form of chorionic villi or a placental site reaction, is absent.

Fig. 17 Atrophic endometrium. There are widely spaced small atrophic tubules within a fibrotic background

Postmenopausal Endometrium The age of menopause with cessation of ovulation and resultant diminution of hormone production by the ovaries is variable, but is usually around age 50. Postmenopausally, the endometrium becomes thin and atrophic, unless there is continuing estrogenic drive, either in the form of endogenous production or exogenous hormone use. When there is no estrogenic drive, the functionalis is absent and the endometrium is composed only of basalis, similar to the basalis of the reproductive years and of the premenarchal endometrium. The histological appearance of the postmenopausal endometrium is variable. The endometrium is usually thin, and this is appreciated in hysterectomy or endometrial resection specimens. The glands do not exhibit proliferative activity and vary from consisting entirely of small widely spaced atrophic tubules (Fig. 17) to cystically dilated glands throughout (so-called cystic atrophy or senile cystic atrophy) (Fig. 18). A mixture of small tubules and cystically dilated glands may occur. The cystic glands seen in atrophic

Fig. 18 Cystic atrophy. Cystically dilated glands are present within a fibrous stroma

endometria in hysterectomy or resection specimens may not be observed in endometrial biopsies because tissue fragmentation during the procedure disrupts the glands. Usually, tubular glands are more prominent in the years immediately following menopause, while cystic atrophy is more common in older women, but this is variable. The glandular cells have small, dark, regular nuclei that may be round, ovoid, or low columnar, and in cystic glands, the nuclei are often compressed and attenuated. Sometimes, there is a degree of nuclear pseudostratification, which may result in a false impression of proliferative activity. The cytoplasm is usually scanty. High-power examination is required to confirm an absence of mitotic activity, and this is especially so in the distinction

7

Benign Diseases of the Endometrium

between proliferative endometrium and atrophy with small tubular glands and between cystic forms of atrophy and simple hyperplasia. The stroma in postmenopausal endometria may be densely cellular and composed of ovoid- to spindle-shaped cells with scant cytoplasm, or has a more fibrous appearance than in the premenopausal endometrium. With advancing age, the stroma tends to become more hypocellular and fibrous. This may be the direct cause of the cystic change because of blockage of the glands. Lymphoid aggregates are often prominent, more so than in premenopausal endometria. As mentioned, it is not clear whether this is due to an actual increase in the number of aggregates or due to them being more obvious because of the glandular atrophy. Occasional mitotic figures may be seen in the glands of postmenopausal endometria with no proven source of estrogenic stimulation. This may occur in women in whom menopause appears gradually and also with uterine prolapse. The postmenopausal endometrium is hormone receptor positive, like that of premenopausal endometria, and retains the capacity to respond to estrogenic stimulation. The morphological features of postmenopausal endometria are similar to those of atrophic endometrium due to other causes, such as exogenous hormones, although there is often also stromal predecidualization in those patients taking progestogen-only compounds or combined preparations containing a progestogen. Atrophic endometrium may also occur in young patients with premature ovarian failure, either idiopathic or due to surgical removal, chemotherapy, or radiotherapy.

Endometrial Sampling Most endometrial samples are taken during the investigation of abnormal uterine bleeding in pre-, peri-, or postmenopausal women. Traditionally, most endometrial samples were obtained by cervical dilatation and curettage (D&C). It was believed that this had a therapeutic effect in some cases of abnormal uterine bleeding, as well as producing a sample for histological examination. Presently, most endometrial samples are

391

outpatient biopsies performed by pipelle or other methods of endometrial sampling. In contrast to the traditional D&C sample, pipelle biopsies do not require an anesthetic and are often performed in conjunction with ultrasound examination and/or hysteroscopy, both of which may identify focal lesions that could be missed by pipelle biopsy alone. The disadvantages of a pipelle biopsy are that often very scant tissue is obtained, especially in a postmenopausal woman with an atrophic endometrium, and focal lesions may be missed. Issues relating to adequacy of endometrial samples are discussed below. On occasions, a pipelle biopsy is followed by a D&C, for example, when there is a suspicion of malignancy and pipelle biopsy produces only a scant amount of tissue which is nondiagnostic, or when there are worrying features on the pipelle sample that are not diagnostic of malignancy. As stated earlier, endometrial sampling is currently not common in the investigation of infertility, but when this is the case, the timing of the biopsy is important; ideally, the biopsy should be taken in the mid-secretory phase between the 7th and 11th postovulatory days. Endometrial polypectomy may be undertaken, and in such instances, sampling of the non-polypoid endometrium should also be performed. Hysteroscopic endometrial resection may be performed under a variety of circumstances, such as in the management of menorrhagia due to large or multiple polyps, and in cases of submucosal leiomyomas. This procedure produces a specimen consisting of multiple chippings, similar to prostatic chips. Chips should be weighed (this should also be done when biopsy or D&C yields a significant amount of tissue, but may be impractical with very scant specimens). All endometrial specimens should be submitted in their entirety for histological examination. The exception is large endometrial polyps where representative sampling may be undertaken; it should be remembered that endometrial polyps may be involved by small carcinomas or by serous endometrial intraepithelial carcinoma (serous EIC), and, as a rule of thumb, at least one block per cm should be taken of large endometrial polyps. Endometrial chips should be orientated if possible, since poorly orientated specimens, especially

392

if tangentially sectioned, may result in a histological suspicion of adenomyosis; it is doubtful whether adenomyosis can be reliably diagnosed on a hysteroscopic endometrial resection sample, although this can be suspected on well-orientated specimens, especially when there is smooth muscle hypertrophy surrounding the islands of endometrial tissue within the myometrium or when there is endometrial tissue present on at least three sides of a fragment of myometrial tissue (Busca and Parra-Herran 2016). A related, although uncommon problem with endometrial resection specimens is the potential for overdiagnosis of myometrial invasion in cases of endometrial hyperplasia or adenocarcinoma; again, this is due to poor orientation and tangential sectioning. In evaluating any endometrial specimen, an adequate clinical history is important, including the age of the patient and the reason for biopsy. Knowledge of the menopausal status as well as the date of onset of the last menstrual period (LMP) and the length of the menstrual cycle in premenopausal women should be provided. In some cases of “postmenopausal bleeding,” the patient is not actually postmenopausal but rather perimenopausal with a prolonged interval between periods, resulting in the clinician and the patient assuming that the woman is postmenopausal. Many women with abnormal uterine bleeding have been prescribed exogenous hormones (especially progestins) before biopsy to control the bleeding, and this information is not always conveyed to the pathologist. Other women may be taking hormone replacement therapy (HRT) or contraceptive agents. These hormonal compounds may alter the morphological appearance of the endometrium, and a knowledge that these and other relevant medications (such as tamoxifen) are being taken is paramount to the pathologist.

Criteria for Adequacy of Endometrial Sample With the increasing trend to perform outpatient endometrial pipelle biopsies rather than formal curettage, pathologists are dealing with increasing numbers of endometrial specimens in which there

R. R. Lastra et al.

Fig. 19 Scanty endometrial biopsy. Endometrial pipelle biopsy in postmenopausal woman where the specimen consists entirely of superficial strips of atrophic endometrial glands without accompanying endometrial stroma

is scant, or even no endometrial, tissue, especially when the endometrium is atrophic. These specimens may consist entirely of superficial strips or wisps of atrophic glands (Fig. 19), with little or no stroma, admixed with cervical mucus, ectocervical or endocervical tissue, and tissue from the lower uterine segment. Paradoxically, it often takes the pathologist longer to examine such specimens, since no underlying architecture is present, and the tissue must be examined carefully under high power to look for mitotic activity, which is abnormal in a postmenopausal endometrium. In specimens such as this, it is controversial as to what constitutes an adequate or inadequate specimen. Designation of a biopsy as inadequate may be of importance since this can have management and medicolegal implications. For example, some clinicians routinely perform a repeat biopsy when an earlier sample has been reported as inadequate while others do not. A biopsy reported as inadequate may suggest to some that the clinician is at fault or has not undertaken the biopsy procedure correctly. While this may be the case in some instances, in most it is not. In published studies, inadequate rates of outpatient endometrial biopsies range from 4.8% to 33% (Antoni et al. 1997; Archer et al. 1991; Clark et al. 2001; Gordon and Westgate 1999; Machado et al. 2003), but in most of these studies, the criteria for adequacy are not clear. A recent study showed that the presence of more than 10 strips of endometrium appeared sufficient to exclude a malignant process, with a negative

7

Benign Diseases of the Endometrium

predictive value close to 100%; this negative predicative value dropped to 81% in cases showing less than 10 strips of endometrium (Sakhdari et al. 2016). It is also worth noting that studies have shown that with an atrophic endometrium and no focal lesion, minimal tissue is the norm with a pipelle biopsy, and there is little chance of missing significant pathology (Bakour et al. 2000). Although it is difficult to recommend precise criteria for adequacy, caution should be exercised before categorizing an endometrial biopsy as inadequate or insufficient. In the majority of cases, the presence of only scant tissue in an endometrial specimen is not a reason for a repeat biopsy, provided the endometrial cavity has been entered, and at least some endometrial tissue is present in the biopsy specimen to confirm this, although theoretically endometrial-type glands with or without stroma could be derived from tuboendometrial metaplasia or endometriosis within the cervix. It has been suggested that with an endometrial biopsy containing scant tissue for which the origin or differentiation cannot be determined, the term unassessable is more appropriate than inadequate or insufficient (Phillips and McCluggage 2005). In such cases, the gynecologist should correlate the biopsy results with the ultrasonographic and/or hysteroscopic findings. If, for example, there is a clinical suspicion of hyperplasia or malignancy, if there is recurrent postmenopausal bleeding, or if there are worrying ultrasonographic and/or hysteroscopic findings, then D&C should be performed. If the above investigations suggest an atrophic endometrium, rebiopsy is probably unnecessary.

Artifacts in Endometrial Biopsy Specimens There are several common artifacts in endometrial biopsy specimens that have received scant attention in the literature (McCluggage 2006). Occasionally, these may be misinterpreted as an endometrial hyperplasia or even a carcinoma if not appreciated to be artifactual. Telescoping is common and refers to the presence of glands within glands (Fig. 20). This artifact seems to be

393

Fig. 20 Telescoping (glands within glands). This is a common artifact in endometrial biopsy specimens

Fig. 21 Glandular molding. A common artifact in endometrial biopsies is glandular “molding.” There is tearing of the tissue around the glands, which is a clue to the artifactual nature

a result of mechanical disruption and “snap back” of the glands during biopsy, resulting in a form of intussusception. Artifactual crowding and compression of glands are also common and may result in consideration of a complex endometrial hyperplasia. With this artifact, the glands often become “molded” together, and there is commonly tearing of the tissue around the glands, which is a clue to the artifactual nature of the glandular crowding (Fig. 21). An artifact that is especially common with, but not exclusive to, outpatient biopsies is the presence of superficial strips of endometrial epithelium, sometimes accompanied by minimal stroma, with a pseudopapillary architecture (Fig. 22). This may result in

394

Fig. 22 Pseudopapillary endometrium. Endometrial biopsy composed of superficial strips of endometrial epithelium with a pseudopapillary architecture

consideration of a wide range of papillary lesions, benign and malignant, which occur in the endometrium. Such superficial strips of pseudopapillary epithelium, which are generally atrophic, should be examined carefully under high power to look for proliferative activity and nuclear atypia. Crushed endometrial glands and stroma may be extremely cellular and can cause concern. Extensive crush artifact is more likely to occur in biopsies from atrophic endometrium in postmenopausal patients. As with the examination of other tissues, crushed elements should not be viewed in isolation. Problems associated with poorly orientated hysteroscopic endometrial resection specimens have already been described. Hysteroscopic endometrial resection specimens might demonstrate vacuolation of the endometrial stromal cells secondary to cautery artifact, resulting in a signet-ring appearance, similar to the phenomenon of vacuolation sometimes observed in cervical stromal cells (McKenna and McCluggage 2008). Occasionally, vacuoles resembling adult fat might be seen in endometrial biopsy specimens, a finding termed pseudolipomatosis (Fig. 23). Since adipose tissue within an endometrial biopsy specimen is an alarming finding that would indicate iatrogenic uterine perforation (see section “Extrauterine Tissues in Endometrial Biopsy Specimens”), awareness of this finding is

R. R. Lastra et al.

Fig. 23 Pseudolipomatosis. Fatlike vacuoles can be distinguished from true adipose tissue by their variation in size and the absence of adipocyte nuclei or intervening capillaries

necessary to avoid overdiagnosing uterine perforation. In contrast to adipose tissue, which demonstrates relatively monotonous vacuoles with identifiable adipocyte nuclei and intervening capillaries, pseudolipomatosis shows variation in the size of the vacuoles and absence of both adipocyte nuclei and intervening capillaries (DeshmukhRane and Wu 2009). Pseudolipomatosis may be related to irritants used during sterilization or secondary to suction applied during the procedure.

Contaminants and Other Elements in Endometrial Biopsies Not uncommonly, fragments of tissue other than from the endometrium are present in endometrial biopsy or curettage specimens. Superficial myometrium is commonly seen, especially in vigorous curettage specimens and in postmenopausal women with an atrophic endometrial lining. It is very common to see cervical tissue as well as cervical mucus, often admixed with neutrophils, histiocytes, and giant cells (Fig. 24), in endometrial biopsy specimens. The cervical tissue usually takes the form of strips of endocervical glandular or squamous epithelium, sometimes with accompanying stroma. The squamous epithelium may be immature metaplastic in type. Usually the cervical origin is obvious, but occasionally this is not the

7

Benign Diseases of the Endometrium

Fig. 24 Cervical mucus in endometrial biopsy specimen. In many endometrial biopsies, mucus derived from the cervix is present, often admixed with neutrophils, histiocytes, and giant cells

case and diagnostic confusion may ensue. For example, if cervical glandular elements exhibiting microglandular hyperplasia are identified within an endometrial biopsy specimen, this may result in consideration of an endometrial hyperplasia or carcinoma, particularly in the postmenopausal setting. The confusion may be heightened by artifactual apposition such that it appears that the endometrial and cervical tissue are in continuity; assessment of whether the accompanying stroma is endometrial or cervical in type may assist in interpretation. Sometimes, dysplastic cervical squamous or glandular epithelium or tissue derived from a cervical neoplasm is present in an endometrial biopsy specimen. Fragments of fallopian tube epithelium may also be seen occasionally. Occasionally, aggregates or sheets of histiocytes may be seen in an endometrial biopsy specimen, either free floating or within the endometrial stroma. Small numbers of histiocytes are not uncommon and are usually inconspicuous, but when present in large aggregates, this may result in consideration of an epithelial or stromal neoplasm (Fig. 25). Recognition of the characteristic lobated, reniform, or coffee-bean nucleus of histiocytes assists, and immunohistochemical staining for histiocytic markers, such as CD68 or lysozyme, may be of value. The histiocytes are probably a reaction to debris within the endometrial cavity, and, when present in large numbers, it has been referred to as nodular histiocytic

395

Fig. 25 Histiocytes in endometrial biopsy. Aggregate of histiocytes in an endometrial biopsy composed of cells with a coffee-bean nucleus and abundant eosinophilic cytoplasm

hyperplasia (Fukunaga and Iwaki 2004; Kim et al. 2002). Occasionally, mitotic figures may be identified within the aggregates of histiocytes, and there may be prominent cell membranes. Rarely, the histiocytes have intracytoplasmic vacuoles and a signet-ring appearance (Iezzoni and Mills 2001), raising the possibility of a signet-ring carcinoma; staining for epithelial and histiocytic markers facilitates the diagnosis. Decidualized and predecidualized endometrial stromal cells may also contain intracytoplasmic vacuoles and simulate signet-ring cell carcinoma. Foamy histiocytes with abundant cytoplasm may also occur within the endometrium, either in association with an endometrial hyperplasia, carcinoma, or pyometra, or as a manifestation of xanthogranulomatous endometritis (see section “Endometritis”).

Extrauterine Tissues in Endometrial Biopsy Specimens Rarely, extrauterine tissues are present in an endometrial biopsy specimen, and this raises the possibility that uterine perforation has occurred either during the current biopsy procedure or some time previously and that a fistulous tract is present. The most common extrauterine tissue is adipose tissue. Although this may potentially be derived from a uterine lipoleiomyoma, lipoma, hamartomatous

396

lesion (McCluggage et al. 2000b), and other pathological lesions containing adipose tissue, or represent metaplastic adipose tissue within the endometrial stroma, in most instances the adipose tissue is derived from the pelvic soft tissues or omentum and indicates perforation. Occasionally, the adipose tissue is accompanied by fibrinous material and mesothelial cells; this is a reflection of underlying pelvic pathology with resultant mesothelial proliferation, which results in fixation of the uterus within the pelvis and makes perforation more likely. In some cases, the patient comes to no harm because of the perforation, presumably due to contraction of myometrial smooth muscle sealing off the defect, but in other instances, a pelvic and/or abdominal inflammatory process ensues and the patient becomes symptomatic. For this reason, the identification of adipose tissue in an endometrial biopsy specimen should prompt a phone call to the clinician. Rarely, other extrauterine tissues, such as intestinal mucosa, may be present in an endometrial biopsy specimen secondary to perforation.

Endometritis Endometritis is a histological diagnosis based upon the identification within the endometrium of an abnormal pattern of inflammatory infiltrate; as such, it must be distinguished from the normal hematopoietic component of the endometrium (see section “Hematopoietic Cells Within the Endometrium”). Most cases of endometritis occur in the reproductive years, but sometimes postmenopausal women are affected. Presentation is typically with abnormal vaginal bleeding, most commonly intermenstrual bleeding or menorrhagia. Endometritis may have both infective and noninfective etiologies. The endometrium is relatively resistant to ascending infection from the lower female genital tract because of the barrier created by the cervix and the cervical mucus. However, uncommonly, endometritis occurs secondary to ascending infection, and this is often a component of pelvic inflammatory disease with inflammation elsewhere in the genital tract.

R. R. Lastra et al.

Predisposing factors to endometritis include a recent pregnancy, the presence of an intrauterine device (IUD), cervical stenosis, and prior instrumentation. Endometritis may also accompany a pathological lesion within the uterus, such as an endometrial polyp, hyperplasia, carcinoma, or a leiomyoma. Usually, the morphological appearances are nonspecific, and, in the absence of an associated pathological lesion, an underlying cause cannot be determined, although occasionally the histological features suggest a particular etiology (see section “Specific Forms of Endometritis”). Endometritis has traditionally been divided into acute and chronic forms, but these constitute a continuum, and often there is an admixture of acute and chronic inflammatory cells. Endometritis may be focal or diffuse and can range from a subtle finding to a pronounced inflammatory reaction. Usually, the endometrial glands exhibit proliferative activity, and there may be mild glandular architectural distortion, in the form of occasional dilated glands. There is often associated surface breakdown with features identical to those seen in menstrual breakdown and breakdown due to non-menstrual causes. In some cases, an initial low-power clue to the diagnosis of endometritis is spindle cell alteration of the stroma (Fig. 26), although this feature is not specific and is not always present. In other cases, the stroma may be edematous. In acute endometritis, the predominant inflammatory cells are neutrophils, and collections of these may be seen within the glandular lumina, forming microabscesses (Fig. 27), or surrounding glands; neutrophils are often most easily seen just deep to the surface endometrium. In some cases, there is surface erosion with fibrinous debris and numerous acute inflammatory cells. By tradition, an unequivocal diagnosis of endometritis requires the presence of plasma cells (Fig. 28), since neutrophils are found normally in the endometrium just prior to and in association with menstruation, and lymphocytes, including lymphoid aggregates, are a normal component of the endometrial stroma. However, in acute forms of endometritis, plasma cells may be absent or few in number. Plasma cells are usually most numerous surrounding endometrial glands

7

Benign Diseases of the Endometrium

397

Fig. 26 Endometritis. In some cases of endometritis, the endometrial stroma has a spindle cell appearance Fig. 29 Xanthogranulomatous endometritis. Large numbers of foamy histiocytes are present within the endometrial stroma

Fig. 27 Acute endometritis. In acute endometritis, neutrophils are present, sometimes forming microabscesses within the glandular lumina

Fig. 28 Chronic endometritis. In cases of chronic endometritis, plasma cells are present within the endometrial stroma

and just deep to the surface epithelium. A form of endometritis without plasma cells has been described and termed focal necrotizing endometritis (Bennett et al. 1999). The histological features of this are of focal, patchy inflammation comprising lymphocytes and neutrophils, centered on individual glands. Due to the focal nature, this form of endometritis can be easily overlooked. The clinical significance of focal necrotizing endometritis, if any, is not currently known. Besides plasma cells, there are also increased numbers of lymphocytes in chronic endometritis, sometimes with prominent and unusually large lymphoid aggregates, and occasionally with the formation of lymphoid follicles. Other inflammatory cells, which may be a component of endometritis, include eosinophils and histiocytes. Usually, histiocytes are inconspicuous since they are admixed with other inflammatory cells, but occasionally large numbers of histiocytes, sometimes with abundant foamy cytoplasm, are present. When these are abundant, this is referred to as xanthogranulomatous endometritis (Fig. 29). Xanthogranulomatous endometritis may occur in association with an endometrial hyperplasia or carcinoma and secondary to cervical stenosis and obstruction. In endometritis, reactive and metaplastic changes may involve the endometrial surface and glandular epithelium. Squamous, ciliated,

398

eosinophilic, and other forms of epithelial metaplasia can occur, and there may be mild nuclear atypia with nuclear enlargement and prominent nucleoli. As stated, sometimes there are mild architectural changes with occasional dilated glands, but significant glandular crowding is not a feature of endometritis. As stated, there may be problems in identifying plasma cells when they are few in number, especially in suboptimally stained sections. Endometrial stromal cells may have a plasmacytoid appearance, especially predecidualized cells in the mid-and late-secretory phase, and unequivocal plasma cells with eccentric nuclei and a perinuclear hof should be present. Occasional plasma cells may be seen in an otherwise normal endometrium, and, in the absence of at least some of the other features of endometritis described, a rigorous search for plasma cells is not justified. Plasma cells may be present in the stroma of an endometrial polyp, and this is not classified as an endometritis unless these are also seen in the non-polypoid endometrium. In problematic cases, histochemical or immunohistochemical stains may be of value in identifying plasma cells. Histochemical stains include methyl green–pyronin, and immunohistochemistry using plasma cell markers such as VS38 or syndecan (CD138) has been described (Bayer-Garner and Korourian 2001; Bayer-Garner et al. 2004; Leong et al. 1997). However, the clinical utility of this is limited. Immunohistochemistry or in situ hybridization for kappa and lambda immunoglobulin light chains may also help demonstrate plasma cells, but this is not routinely performed (Euscher and Nuovo 2002). Immunohistochemical staining with B lymphoid markers (CD20 and CD79a) may also assist in distinguishing between the physiological endometrial lymphocytic infiltrate and the inflammatory infiltrate of endometritis. Normally, the vast majority of lymphoid cells within the endometrial stroma are T cells (CD3 positive) with B lymphocytes accounting for about 1% of all endometrial leucocytes (Marshall and Jones 1988). B lymphoid cells are largely confined to lymphoid aggregates within the endometrial basalis, with occasional individual cells in the functionalis. In most cases of chronic endometritis, there are increased numbers of B lymphoid cells, and these also have an

R. R. Lastra et al.

abnormal location being found outside lymphoid aggregates within the stroma and sometimes intraepithelially and within the glandular lumina (Disep et al. 2004). The number of T lymphocytes, histiocytes, and granulated lymphocytes in endometritis does not differ significantly from controls. It is emphasized that plasma cells do not usually stain with CD20 but are positive with CD79a.

Specific Forms of Endometritis Chlamydia Trachomatis Chlamydia trachomatis has been isolated from cases of both acute and chronic endometritis. However, it is unclear whether the organism is causative in such cases or an accompanying pathogen. Chlamydia trachomatis infection is relatively common in both the upper and lower female genital tracts and may be associated with pelvic inflammatory disease and infertility. Endometritis secondary to Chlamydia trachomatis has no specific histological features, although the inflammatory infiltrate may be intense and lymphoid follicles and large numbers of blasts may be seen (Paavonen et al. 1985). Stromal necrosis and reactive atypia of the endometrial epithelium may be present. Definitive diagnosis in most cases requires culture. However, Chlamydia trachomatis inclusion bodies have been identified within endometrial epithelial cells; these are extremely difficult to detect on hematoxylin and eosin-stained sections but can occasionally be recognized on Giemsa stain. Immunohistochemical staining may also be of value (Winkler et al. 1984), positivity being localized to the epithelial cells in the form of stippling within supranuclear intracytoplasmic vacuoles. Molecular investigations may also be of value in demonstrating Chlamydia trachomatis infection.

Cytomegalovirus Cytomegalovirus (CMV) endometritis is rare. It is most common in immunosuppressed patients, but occasionally occurs in women with no underlying immune disorder. Typical intranuclear and cytoplasmic CMV inclusions are seen, mainly in

7

Benign Diseases of the Endometrium

epithelial cells but occasionally in stromal or endothelial cells. The other histological features are nonspecific; granulomas have been described in occasional cases (Frank et al. 1992).

Herpes Simplex Virus Herpes simplex virus type II rarely results in an endometritis, usually secondary to ascending infection from the cervix and sometimes in women who are immunosuppressed. Typical herpes simplex virus inclusions are present within the glandular epithelium, and there are multinucleated cells with molded ground-glass nuclei. The other histological features are nonspecific, but patchy necrosis of the endometrial glands and stroma with an associated inflammatory infiltrate may occur (Duncan et al. 1989). Optically clear nuclei due to the accumulation of biotin, associated with the presence of trophoblast (see section “Gestational Endometrium”), may superficially resemble herpes simplex virus inclusions; while immunohistochemical staining may be of value in such instances, caution must be taken in their interpretation, as the biotin within these optically clear nuclei might result in a false-positive result.

399

1976). The bacterial colonies form the so-called actinomycotic granules (AMGs), referred to as sulfur granules because of their tan to yellow color on gross examination. It is important to recognize Actinomyces, since the organism may result in ascending infection with resultant tuboovarian abscess formation and pelvic inflammatory disease. Histologically, AMGs are usually seen on the endometrial surface or within the superficial stroma as non-refractile granules with thin basophilic radiating filaments and sometimes a dense more eosinophilic granular core (Fig. 30). They are Gram positive on Brown and Brenn stain (Fig. 31) and are highlighted with Gomori methenamine silver stain. Although the diagnosis can be strongly

Mycoplasma Rarely mycoplasma organisms, usually Ureaplasma urealyticum, result in endometritis. The inflammatory infiltrate is typically focal and has been termed “subacute focal endometritis” (Khatamee and Sommers 1989). The inflammatory cells comprise mainly lymphocytes and histiocytes with few neutrophils and plasma cells, and they tend to be concentrated beneath the surface epithelium, adjacent to the glands, or around the spiral arteries. Granulomas have rarely been described.

Fig. 30 AMGs composed of thin basophilic radiating filaments with a more dense eosinophilic granular core

Actinomyces The Gram-positive anaerobic bacterium Actinomyces may result in endometritis, often in association with a long-term IUD (Gupta et al.

Fig. 31 AMGs are Gram positive on Brown and Brenn stain

400

suspected on morphological examination, culture is recommended for confirmation since other Gram-positive filamentous bacteria may be found in the gynecological tract. Because of the potential complications, Actinomyces must be distinguished from pseudoactinomycotic radiate granules (pseudo-sulfur granules) (Pritt et al. 2006). These are noninfectious lesions, most commonly seen in association with an IUD, but sometimes in non-IUD users. They consist of thick, irregular, club-like peripheral projections without a dense central core (Fig. 32). An associated inflammatory response may be present. With Brown and Brenn stains, there is diffuse, intense nonspecific staining, while silver stains are negative. Pseudoactinomycotic radiate granules are probably more common than Actinomyces (O’Brien et al. 1981), and occasionally the two coexist. The former may also be seen in association with pelvic inflammatory disease. The noninfectious nature of pseudoactinomycotic radiate granules is supported by microbiological and histochemical studies, as well as ultrastructural analysis. Their exact composition and nature is unknown, but they may represent an unusual response to foreign bodies (Splendore–Hoeppli phenomenon). Their main significance is that they may be mistaken for AMGs.

R. R. Lastra et al.

Fungi and Parasites Fungi and parasites may rarely result in endometritis, more commonly in underdeveloped countries. Blastomycosis (Blastomyces dermatitidis) and coccidioidomycosis (Coccidioides immitis) may result in endometritis as part of a disseminated infection. Granulomas may be a component of the inflammatory infiltrate. There have been occasional reports of candidal and cryptococcal endometritis. Gomori methenamine silver and periodic acid–Schiff (PAS) stains are helpful in identifying these organisms. Schistosoma, Enterobius vermicularis, and Echinococcus granulosus are rare causes of endometritis in developed countries, but schistosomiasis is endemic in some parts of the world. Schistosomal endometritis may be mild or severe and is characterized by granulomatous inflammation with lymphocytes, plasma cells, eosinophils, and histiocytes, sometimes closely simulating a tubercle. The endometrial surface may be ulcerated and replaced by granulation tissue. Diagnosis is made by identifying the ova in tissue sections or in smears of vaginal secretions. Toxoplasmosis (Toxoplasma gondii) evokes a nonspecific inflammatory reaction in the endometrium. The microorganism can be identified by immunofluorescence.

Malakoplakia

Fig. 32 Pseudoactinomycotic radiate granules. Pseudoactinomycotic radiate granules with thick, irregular, clublike peripheral projections without a central dense core

Malakoplakia may involve several organs, most commonly the urinary bladder, and is characterized by the presence of sheets of foamy histiocytes (von Hansemann’s histiocytes) containing Michaelis–Gutmann bodies. These are small, round, laminated calcospherites, which are present in the cytoplasm of the histiocytes and in an extracellular location. They contain calcium and can be demonstrated by von Kossa stain. The histiocytes are often admixed with other inflammatory cells, including plasma cells and neutrophils. Occasional examples have been reported in the endometrium (Thomas et al. 1978). Malakoplakia is a result of an abnormal immune response to bacteria, most commonly Escherichia coli, which are retained within the

7

Benign Diseases of the Endometrium

phagolysosomes of the histiocytes but are not digested; the Michaelis–Gutmann bodies are the result of mineral encrustation of incompletely digested bacteria.

Lymphoma-Like Lesion The so-called lymphoma-like lesions are more common within the cervix, but have rarely been described in the endometrium (Young et al. 1985) (see ▶ Chap. 21, “Hematologic Neoplasms and Selected Tumorlike Lesions Involving the Female Reproductive Organs”). Histologically, they are characterized by dense aggregates of lymphoid cells, often with large numbers of blasts and a starry-sky appearance, forming a superficial band-like infiltrate, this only being appreciated on a hysterectomy or hysteroscopic endometrial resection specimen. Lymphoid follicles with germinal centers, which may be large and ill-defined, are typically present. Along with germinal centers and large numbers of blasts, a mixed inflammatory infiltrate is present with small lymphocytes, plasma cells, neutrophils, and histiocytes. Lymphomalike lesions represent an exaggerated form of chronic endometritis. The polymorphic nature of the infiltrate together with the presence of germinal centers and the superficial location of the inflammation (as stated, only appreciated on hysterectomy or endometrial resection specimens) help to distinguish lymphoma-like lesion from malignant lymphoma, as does the absence of a mass lesion grossly. Immunohistochemistry for kappa and lambda light chains or molecular investigations to demonstrate a polyclonal population may also be of value. Occasional cervical cases have been associated with Epstein–Barr virus infection (Young et al. 1985).

Endometrial Granulomas Granulomas within the endometrium are rare. Worldwide, the most common cause is tuberculosis, and, although rare in developed countries, granulomatous endometritis should be considered

401

as tuberculous in origin until proven otherwise. Tuberculosis of the endometrium usually occurs in premenopausal women and is rare after menopause. Caseous necrosis is characteristic of tuberculous granulomas, but due to the constant shedding associated with menstruation, endometrial granulomas in patients with tuberculosis are often noncaseating. Tubercle bacilli are seldom identified on Ziehl–Neelsen-stained sections, and culture should be undertaken in all cases in which histological examination raises the possibility of tuberculosis. Other infectious causes of granulomatous endometritis include various fungi, schistosomiasis, Enterobius vermicularis, and Toxoplasma gondii (see section “Fungi and Parasites”). As already discussed, granulomas are occasionally a feature of CMV and mycoplasma endometritis. Endometrial granulomas, especially when well circumscribed, also raise the possibility of sarcoidosis, and there have been rare reports of sarcoidosis involving the endometrium (Pearce and Nolan 1996). A granulomatous reaction to keratin may be seen in association with an endometrioid adenocarcinoma or atypical polypoid adenomyoma exhibiting squamous differentiation. Occasionally, such keratin granulomas may also be found on the surface of the ovaries, the fallopian tubes, or on the omentum or peritoneum; this is secondary to spread of keratin through the fallopian tubes and, in the absence of associated neoplastic cells, does not represent tumor spread. Foreign body granulomas in the endometrium may be secondary to talc or other substances and may be seen in association with an IUD. A palisading granuloma with fibrinoid material and a surrounding histiocytic and giant cell reaction, features resembling a rheumatoid nodule, may occur secondary to endometrial ablation; usually, the entire endometrium or much of the endometrium is affected, and well-circumscribed granulomas are not generally found (see section “Effects of Endometrial Ablation or Resection”). A similar picture may be seen secondary to endometrial resection. In rare instances, there is no obvious cause for endometrial granulomatous inflammation, so-called idiopathic granulomatous endometritis (Fig. 33).

402

Fig. 33 Granulomatous endometritis. A single granuloma is present within the endometrial stroma

Ligneous (Pseudomembranous) Endometritis Ligneous (pseudomembranous) endometritis is discussed here, although the term endometritis is not strictly appropriate since there is often little or no inflammatory infiltrate. Ligneous disease is an inherited autosomal recessive condition characterized by absent or low plasminogen levels, resulting in the accumulation and deposition of fibrin. The histological features in the endometrium are identical to those of other affected sites, most commonly the conjunctiva. Ligneous disease is rare in the female genital tract and most commonly affects the cervix, vulva, or vagina. Rare endometrial cases have been described (Karaer et al. 2007; Scurry et al. 1993). When the endometrium is affected, this may result in dysmenorrhea and infertility. Histology shows amorphous eosinophilic material, somewhat similar to amyloid, which represents fibrin (Fig. 34); a Congo red stain would be negative. There may be associated mild inflammation, including occasional multinucleated giant cells. The inflammatory infiltrate may be more severe if there is surface ulceration.

Dysfunctional Uterine Bleeding Dysfunctional uterine bleeding (DUB) is abnormal uterine bleeding in a premenopausal woman resulting from alterations in the normal cyclical

R. R. Lastra et al.

Fig. 34 Ligneous (pseudomembranous) endometritis. Band of amorphous eosinophilic material represents fibrin deposition

changes of the endometrium, without an underlying specific pathological cause such as endometritis, polyps, exogenous hormones, hyperplasia, or carcinoma. In many cases, DUB is probably secondary to endogenous hormone imbalance. There are several morphological alterations of the endometrium that are characteristic of DUB, the most common being those associated with anovulatory cycles or luteal phase defects. These can be regarded as estrogen-related and progesterone-related, respectively, and are discussed in the next sections. Often DUB is managed by hormonal therapy, and a biopsy is only performed when symptoms persist; the hormone therapy may result in modification of the morphology. Importantly, DUB is not a pathological diagnosis but rather a clinical term. A common, but not invariable feature of biopsies from patients with DUB is the presence of glandular and stromal breakdown. The features associated with this are not unique to DUB and are seen in menstrual endometrium and in bleeding associated with a variety of organic disorders. It is important to recognize the features of breakdown and to distinguish them from other pathological lesions. It is also important to realize that glandular and stromal breakdown is a nonspecific feature and that the intact endometrium must be assessed to evaluate the underlying abnormality. Glandular and stromal breakdown may also occur in an atrophic endometrium. The changes

7

Benign Diseases of the Endometrium

associated with glandular and stromal breakdown are described in the next paragraphs. In menstrual endometrium, the features of breakdown are diffuse and occur on a background of secretory endometrium. In contrast, in DUB the background endometrium is typically nonsecretory in type and breakdown is usually a focal phenomenon, resulting in a heterogeneous pattern with intact fragments of endometrium admixed with fragments exhibiting the features of breakdown. Furthermore, in menstrual endometrium the changes are acute and there are no features of chronic bleeding, such as hemosiderin deposition and accumulation of foam cells. The morphological features associated with glandular and stromal breakdown are summarized in Table 2. An early feature is the accumulation of nuclear (apoptotic) debris in the basal cytoplasm of the glandular cells (Fig. 35) (Stewart et al. 1999). The stromal cells collapse and aggregate into tight clusters, which are separated by lakes of blood. These clusters of stromal cells, sometime called “stromal blue balls,” (Fig. 36) may form small polypoid extrusions or become detached from the surrounding tissue. They are characterized by tightly packed cells with hyperchromatic nuclei and scanty cytoplasm admixed with apoptotic debris. They can exhibit nuclear molding and may raise the possibility of small cell carcinoma. However, they are often covered by an epithelial lining, which may be flat or may exhibit the features of papillary syncytial metaplasia (see below), and are associated with other features of

403

breakdown. Because of the stromal collapse, the endometrial glands become disrupted and crowded (Fig. 37). This glandular crowding may mimic hyperplasia or even adenocarcinoma, but recognition of the other features of breakdown facilitates the diagnosis. Fibrin thrombi are usually seen in small blood vessels, either in the spiral arteries or in superficial ectatic stromal vessels (Fig. 38). Another consistent feature of breakdown is papillary syncytial metaplasia (Fig. 39). Synonymous terms include eosinophilic syncytial change and surface syncytial change. The term papillary syncytial metaplasia is a misnomer since this is not a true metaplasia but has rather

Fig. 35 Glandular and stromal breakdown. An early feature is the accumulation of apoptotic debris in the cytoplasm of the glandular epithelial cells

Table 2 Features of endometrial breakdown Glandular changes Nuclear (apoptotic) debris in the basal cytoplasm of glandular cells Papillary syncytial metaplasia Glandular crowding Stromal changes Stromal collapse Aggregates of stromal cells Nuclear (apoptotic) debris in stroma Fibrin thrombi Hemosiderin pigment deposition Foam cell accumulation Fibrosis and hyalinization

Fig. 36 Endometrial breakdown. With breakdown, the stromal cells aggregate into stromal “blue balls”

404

Fig. 37 Endometrial breakdown. With breakdown, the endometrial glands become disrupted and crowded because of stromal collapse

R. R. Lastra et al.

Fig. 39 Chronic endometrial breakdown. With chronic breakdown, hemosiderin pigment may accumulate within histiocytes within the endometrial stroma

deposition, free within the stroma or within histiocytes (Fig. 39), and accumulation of foam cells, neither of which occurs in normal cyclical endometrium during the reproductive years. These features usually indicate chronic bleeding. The foam cells were initially thought to represent histiocytes, but it has been suggested that they may represent endometrial stromal cells, which become distended with lipid following erythrocyte breakdown (Fechner et al. 1979). Chronic bleeding may also occasionally result in focal stromal fibrosis and hyalinization. Fig. 38 Endometrial breakdown. With breakdown, fibrin thrombi are typically seen

either been considered a regenerative reaction to surface breakdown or a degenerative or regressive process (Shah and Mazur 2008). Papillary syncytial metaplasia is characterized histologically by syncytial sheets of epithelial cells with indistinct cell borders that form micropapillary structures, sometimes with the small glandular lumina. The syncytia are devoid of stromal support and lack fibrovascular stromal cores. The cells usually have eosinophilic cytoplasm, and there is often a neutrophilic infiltrate. There may be mild nuclear atypia, and sometimes mitoses are identified. Other features of breakdown, which are not present in all cases, include hemosiderin

Estrogen-Related DUB, Including Endometrium Associated with Anovulatory Cycles This endometrial morphology is most common in the perimenopausal years, where it is usually secondary to anovulatory cycles with resultant absence of development of the corpus luteum and decrease in progesterone secretion; in fact, this is sometimes referred to as anovulatory endometrium. The term persistent proliferative endometrium has also been used. The underlying causes are complex, but many cases may be a result of hypothalamic dysfunction. The developing follicles persist for a variable period of time and produce E2 before undergoing atresia,

7

Benign Diseases of the Endometrium

at which time estrogen withdrawal bleeding occurs. In other cases, estrogen breakthrough bleeding occurs when the persisting follicles produce E2 resulting in the proliferating endometrium becoming thicker and outgrowing its blood supply. The usual presentation is with perimenopausal bleeding, but younger women may also be affected, for example, perimenarchal adolescents in whom regular ovulatory cycles are not established, those with polycystic ovarian syndrome (Stein–Leventhal syndrome), or older women taking unopposed estrogens or with an increase in endogenous estrogens, for example, secondary to obesity. Anovulatory cycles may also occur sporadically throughout the reproductive years. The histological features involve the entire endometrial compartment and are those of proliferative endometrium but with superimposed breakdown, as described above (Mutter et al. 2007). The extent of breakdown is highly variable and ranges from minute focal changes to a widespread phenomenon involving most of the specimen. There are often foci of ciliated (tubal) or other types of epithelial metaplasia that are randomly dispersed. With chronic anovulatory cycles, there is abundant proliferative endometrium, and mild degrees of disorganization with dilated glands may occur. This results in a picture that is neither normal proliferative nor hyperplastic, which is referred to as disordered proliferative endometrium (Fig. 40). Occasional dilated glands within an otherwise normal proliferative endometrium do not warrant a diagnosis of disordered proliferation; in other words, there should be significant numbers of dilated glands and these should have a widespread distribution, although they are admixed with normal proliferative glands. In disordered proliferative endometrium, the normal gland to stroma ratio is largely maintained, although there may be focal mild glandular crowding and branching. In cases with significant numbers of dilated glands, the morphological appearances merge with and overlap with those of simple hyperplasia; in fact, disordered proliferative endometrium and simple hyperplasia almost certainly constitute a continuum, and it is likely that both represent a response to unopposed estrogens and

405

Fig. 40 Disordered proliferative endometrium. Occasional cystically dilated glands with abnormal glandular shapes are present within an otherwise typical proliferative endometrium

not a true premalignant lesion, although there is a small increased risk of the development of an endometrioid adenocarcinoma. In simple hyperplasia, there is usually more glandular dilatation, with a paucity of normal proliferative glands, and the glands exhibit more budding and branching. However, there is significant interobserver variability in the distinction between disordered proliferative endometrium and simple hyperplasia, and, as stated, these form part of a continuum without sharply defined borders. Disordered proliferative endometrium may occasionally be confused with a polyp because of the glandular architectural distortion and dilatation; however, the fibrous stroma and thick-walled stromal blood vessels characteristic of a polyp are absent. Cystic atrophy may also enter into the differential diagnosis, but in this there is an absence of mitotic activity and the cells lining the glands are attenuated. Menstrual shedding following ovulation often results in disordered proliferative endometrium reverting to normal.

Progesterone-Related DUB: Luteal Phase Defects Luteal phase defects (also known as inadequate luteal phase, secretory insufficiency, or inadequate secretory phase) are a relatively common

406

cause of DUB and also of ovulatory infertility. Although ovulation occurs, there is a relative or absolute insufficiency of progesterone secretion by the corpus luteum, which may either regress prematurely or fail to produce an adequate amount of progesterone. As a consequence, the luteal (secretory) phase does not develop appropriately, the secretory features in the endometrium being poorly developed. The underlying cause is unknown but is thought to be a result of hypothalamic or pituitary dysfunction, which results in decreased levels of follicle-stimulating hormone (FSH) and abnormal luteinizing hormone (LH) secretion. Because of inadequate progesterone secretion, there may be a lag in the histological date of the endometrium of at least 2 days compared to the actual postovulatory date. Other morphological features in some cases include discordance in development of the glands and stroma and different areas of the endometrium exhibiting marked variation in development; for example, some areas may exhibit early secretory activity while others show predecidual change. Alternatively, the glands may exhibit hypersecretory features while the stroma lacks predecidual change. Although the endometrium exhibits a secretory pattern, often this cannot be assigned to any day of the normal cycle. In addition to the features described, there is surface breakdown. The diagnosis can be made by a combination of repeated biopsies over several cycles and serial hormone measurements. In other cases, irregular shedding may be a result of a persistent corpus luteum with prolonged progesterone production. This is a poorly understood form of DUB, and, as such, the histological features are not well described.

Effects of Exogenous Hormonal Agents and Drugs There are a plethora of exogenous hormonal agents in widespread use for a variety of indications including contraception, alleviation of menopausal symptoms, management of organic lesions or DUB, treatment of infertility, management of endometrial hyperplasia or carcinoma, and endometrial prophylaxis in patients with

R. R. Lastra et al.

hyperestrogenic states or taking medications such as tamoxifen. The effects of the various hormonal agents on the endometrium are varied, although in many instances predictable, and depend on a number of factors, including the menopausal status of the patient, the exact composition of the hormonal preparation, and the dose and duration of administration. Not uncommonly, the endometrium is biopsied in patients taking exogenous hormones, for example, when abnormal bleeding occurs, when hormones do not correct suspected DUB, or when the effect of hormonal agents on the endometrium is assessed. Hysterectomy or endometrial resection may also be performed in patients taking exogenous hormones. Some nonhormonal agents may also result in endometrial morphological changes, although much less commonly than with hormonal compounds. In the following sections, the effects on the endometrium of the most common hormonal agents and of some nonhormonal medications are described. Full details regarding the preparations being taken are obviously of paramount importance to the pathologist in assessing the endometrium, but often the details are incomplete or not relayed to the pathologist at all. As such, the pathologist should always suspect the possibility of exogenous hormone use, especially when the endometrial morphology does not correspond to a normal cyclical or postmenopausal appearance.

Estrogen-Only HRT There are various synthetic preparations of estrogens that are largely given to perimenopausal or postmenopausal women to treat menopausal symptoms. In women with a uterus, estrogenonly HRT (unopposed estrogen) is contraindicated due to the risk of endometrial proliferative lesions, including hyperplasia and endometrioid adenocarcinoma, and, as such, the use of estrogen-only preparations is unusual in a woman with a uterus. The morphological features in the endometrium vary, but there may be proliferative activity, a picture identical to disordered proliferation, any type of endometrial hyperplasia,

7

Benign Diseases of the Endometrium

or an endometrioid adenocarcinoma (The Writing Group for the PEPI Trial 1996). There may be associated surface breakdown and epithelial cytoplasmic change, including squamous and ciliated metaplasia. The risk of malignancy increases with the dose and duration of therapy; those adenocarcinomas that develop are usually, but not always, low grade and early stage. Unopposed estrogens result in endometrial hyperplasia in approximately 20% of women following 1 year of treatment. The “postmenopausal estrogen/progestin intervention (PEPI) trial” concluded that women taking estrogens alone had a high incidence of simple (27.7%), complex (22.7%), and atypical (11.7%) hyperplasia; this was significantly higher than in those taking placebos. The reported risk ratio for endometrial carcinoma in women taking unopposed estrogens has ranged from 2.3 to 10 (Grady et al. 1995; Paganini-Hill et al. 1989); the risk persists for many years after estrogen treatment is discontinued (Brinton and Hoover 1993; Shapiro et al. 1985). Estrogen-only preparations may also result in proliferative changes and the development of premalignant and malignant lesions in endometriosis; as such, caution should be exercised before prescribing unopposed estrogens following hysterectomy in a woman with known endometriosis.

407

unexpected bleeding occurs; there is no correlation between bleeding patterns and endometrial histology. With sequential regimes, the endometrium may exhibit atrophy, secretory, or weak proliferative activity, the latter especially if the biopsy is taken during the period of estrogen therapy. If the endometrium is biopsied during the period of progestin therapy, there may be poorly developed secretory activity in the glands with cytoplasmic vacuoles and scant luminal secretions. Focal glandular and stromal breakdown may also be seen. Sequential regimes do not completely eliminate the risk of carcinoma associated with unopposed estrogen therapy; the prevalence of endometrial hyperplasia associated with sequential HRT is 5.4% and that of atypical hyperplasia 0.7% (Sturdee et al. 2000). It should be remembered that HRT is most commonly taken by postmenopausal women, and in this age group, there is a background incidence of endometrial hyperplasia and carcinoma. With the continuous combined regimes, the endometrium is usually atrophic (Fig. 41) or exhibits weak secretory activity, and, in many instances, biopsies yield scanty material. There is no increased risk of endometrial proliferative lesions with continuous combined HRT (Nand et al. 1998), and, in fact, these regimes may protect against the development of endometrial hyperplasia and carcinoma and normalize endometria that have exhibited

Combined Estrogen and Progestin HRT Because of the potential adverse effects of unopposed estrogens, in most women with a uterus, an estrogen is combined with a progestin for HRT. Estrogen and progestin combinations may be given sequentially or continuously. Sequential (cyclic) regimes are variable but usually employ daily estrogens for the first 21 days, 25 days, or the whole of the month, with daily progestins added for the last 10–13 days; these regimes result in a withdrawal bleed. Continuous combined regimes use both estrogen and progestin daily. With continuous regimes, breakthrough bleeding may occur during the first 6 months, but this bleeding then usually stops. Patients receiving sequential or continuous regimes may undergo biopsy as part of routine surveillance or when

Fig. 41 Endometrium associated with continuous combined hormone regimes. With continuous combined hormone regimes, the endometrium is usually atrophic

408

R. R. Lastra et al.

complex hyperplasia (Feeley and Wells 2001; Staland 1981; The Writing Group for the PEPI Trial 1996). However, data from the Women’s Health Initiative has shown an increased risk of breast and ovarian carcinoma in patients with long-term combined HRT (Anderson et al. 2003). Endometrial polyps are relatively common in women taking combined HRT and appear more common with sequential than continuous regimes (Feeley and Wells 2001).

Progestin-Only Compounds Various forms of synthetic analogues of progesterone, termed progestins, are in widespread use, either alone or in combination with an estrogen. Progestin-only hormonal compounds, taken either orally or systemically, are usually prescribed for abnormal uterine bleeding and result in suppression of ovulation and inhibition of endometrial growth. Progestins may also be given for the management of conditions such as endometriosis, for contraception, or for endometrial protection in patients taking tamoxifen. The effects of progestins on the endometrium are variable and depend on the degree of estrogen priming as well as the type of progestin and the dose and duration of therapy. They typically result in atrophy of the endometrial glands with predecidual change or decidualization (sometimes termed pseudodecidualization) of the stroma. The endometrial glands are most commonly small, tubular, and atrophic and lined by cuboidal cells with small round nuclei and scant cytoplasm, but sometimes exhibit poorly developed secretory activity. The stroma is usually expanded and composed of predecidualized or decidualized cells with abundant eosinophilic cytoplasm (Fig. 42) with infiltration by granulated lymphocytes; this may mimic endometritis, but plasma cells are absent. Marked decidualization is most common with high-dose progestins and may result in copious polypoid fragments of tissue being obtained at biopsy. These often exhibit superficial breakdown with associated neutrophil infiltration; again this may be mistaken for endometritis. On occasions, some of the polypoid tissue fragments are totally

Fig. 42 Endometrium associated with progestin-only compound. With progestin-only compounds, the stroma is typically expanded and composed of predecidualized cells

necrotic, probably due to the stromal expansion with outgrowth of the blood supply. The decidualized stroma may, on occasion, contain cells exhibiting variation in nuclear size and shape with nuclear hyperchromasia; this may result in an alarming appearance, which may be exacerbated by stromal myxoid change and cytoplasmic vacuolation, resulting in a signet-ring appearance and mimicry of signet-ring carcinoma. When the progestin dose is low, the stromal cells may not exhibit predecidual or decidual change. The morphological effects associated with the Mirena coil, a progestin-containing intrauterine device, are described below (see section “Effects of Intrauterine Device (IUD)”).

Progestin-Like Effect Without Exogenous Hormone Use Occasionally, the endometrium exhibits changes identical to those occurring with progestins, namely, atrophic glands and stromal decidualization, in a patient not taking exogenous hormones. This may occur in both premenopausal and postmenopausal women (Clement and Scully 1988), and the etiology is, in most cases, poorly understood. Most other cases are probably secondary to a persistent functioning corpus luteum or a luteinized unruptured follicle where a follicle develops without ovulation and persists with

7

Benign Diseases of the Endometrium

luteinization of the granulosa and theca cells that produce progesterone. Alternatively, the changes may be a result of local mechanical factors rather than a response to progesterone-like hormones; mechanical stimulation, including biopsy or the presence of an IUD, may rarely result in predecidual or decidual change in the endometrium. In very rare cases, the changes will be secondary to a progesterone secreting neoplasm in an ovary or elsewhere.

Gonadotropin-Releasing Hormone Agonists Gonadotropin-releasing hormone agonists (GnRH agonists), including buserelin acetate, goserelin acetate, and leuprolide acetate, are commonly used in the management of uterine leiomyomas and endometriosis. After initial stimulation of the pituitary gland with increased production of LH and FSH, further administration results in desensitization of the pituitary to GnRH and a subsequent decreased production of LH and FSH. This results in decreased estrogen production by the ovaries and a hypo-estrogenic state. As a consequence, there is shrinkage of uterine leiomyomas, thus alleviating symptoms and potentially allowing myomectomy rather than hysterectomy or vaginal hysterectomy rather than abdominal hysterectomy. The endometrium in patients taking GnRH agonists is typically atrophic with small tubular glands, or sometimes the glands exhibit weak proliferative activity. When GnRH agonists are used in conjunction with a progestogen, the endometrium may exhibit decidualization of the stroma.

Androgens Several androgens are in widespread use, for example, danazol and tibolone. The main indications for these preparations are the treatment of endometriosis, but androgens may also be used in the management of menorrhagia or endometrial hyperplasia, or as HRT. In the early stages of androgen therapy, the glands may exhibit weak

409

secretory activity, but with prolonged therapy, the endometrium becomes atrophic (Marchini et al. 1992).

Progesterone Receptor Modulators Progesterone receptor modulators (PRMs) are synthetic compounds that interact with the progesterone receptor to inhibit or stimulate a downstream hormonal response. Compounds with PR antagonist activity are used in contraception and in the management of uterine leiomyomas or endometriosis. The endometrial effects of PRMs have been described (Mutter et al. 2008). While in some cases the endometrium is inactive or has a normal cyclical appearance, in a subset of cases, there is asymmetry of stromal and epithelial growth, resulting in prominent cystically dilated glands with admixed estrogenic (mitotic) and progestogenic (secretory) activity. Vascular changes described include a chicken-wire vasculature and the presence of thick-walled and ectatic vessels. These novel changes are termed PRM-associated endometrial changes.

Tamoxifen Tamoxifen is a nonsteroidal triphenylethyl compound that is widely used as adjuvant therapy in the treatment of breast cancer. It is a selective estrogen receptor modulator (SERM) and prolongs overall and disease-free survival in breast cancer, reduces the likelihood of disease in the contralateral breast, and may reduce the risk of development of breast cancer in asymptomatic women with a strong family history. Results from the National Surgical Adjuvant Breast and Bowel Project Breast Cancer Prevention study demonstrated that 5 years of tamoxifen use at a dose of 20 mg/day reduced breast cancer risk in high-risk women by 49% (Fisher et al. 1998). The efficacy of tamoxifen in breast cancer is due to its antiestrogenic properties that are mediated by competitive binding to the estrogen

410

receptor. However, tamoxifen may also exert a weak estrogenic effect and act on the human endometrium (Ismail 1994, 1999; Seidman and Kurman 1999), and the effects of tamoxifen on the endometrium appear to depend on the menopausal status as well as on the dose and duration of tamoxifen usage. Transvaginal ultrasonography has shown that the endometrium of tamoxifen-treated postmenopausal patients is significantly thicker than that of age-matched controls of women not taking tamoxifen (Cheng et al. 1997), and benign postmenopausal endometria of patients treated with tamoxifen exhibit a higher Ki-67 proliferation index than controls of patients not taking tamoxifen (Hachisuga et al. 1999). Tamoxifen therapy may result in a spectrum of endometrial proliferative lesions, including polyps; simple, complex, and atypical hyperplasia; and adenocarcinomas. Various other malignancies have also been described in association with tamoxifen. Since the majority of women taking tamoxifen are postmenopausal, most of the information regarding the endometrial side effects is related to this age group. Nonneoplastic endometrium in postmenopausal patients taking tamoxifen is most commonly atrophic, while in other cases there is proliferative activity. The stroma is often fibrous, and, as a result, endometrial biopsies in patients taking tamoxifen are often very scanty. There may be glandular dilatation secondary to obstruction of the glands by stromal fibrosis. Stromal decidualization is usually secondary to simultaneous progestin administration (Cohen et al. 1996). One of the most characteristic and common endometrial lesions in women taking tamoxifen is polyps, which may be single or multiple and may occur on a background of hyperplasia, such that there is merging of polypoid and non-polypoid endometrium. There are no pathognomonic features of tamoxifen-associated endometrial polyps, but they tend to be larger than sporadic polyps. Periglandular stromal condensation, staghorn-shaped glands with intraglandular polypoid projections polarized along the long axis of the polyp, stromal edema and myxoid change, and epithelial metaplasias are all more common in

R. R. Lastra et al.

Fig. 43 Tamoxifen polyp. Staghorn-shaped glands with intraglandular polypoid projections, epithelial metaplastic changes, and focal stromal myxoid change are often present in tamoxifen polyps

tamoxifen-associated than sporadic polyps (Kennedy et al. 1999; Schlesinger et al. 1998), although all these features may be seen in sporadic polyps (Fig. 43). Cancers may develop in tamoxifenassociated polyps and can be of endometrioid or serous type. The presumed precursor lesion of serous carcinoma, serous EIC, may also involve tamoxifen-associated polyps (McCluggage et al. 2003b; Silva and Jenkins 1990). Rarely, a metastatic breast carcinoma, usually of lobular type, is identified within a tamoxifen-associated endometrial polyp (Houghton et al. 2003). As stated earlier, endometrial carcinomas may develop in association with tamoxifen. It is generally accepted that the risk of developing endometrial carcinoma in patients taking tamoxifen is two to three times higher than in an age-matched population of women not taking tamoxifen. It is probable that it is women who have been taking tamoxifen for a prolonged period of time and with a high cumulative dose who are at greatest risk. Since most endometrial cancers that develop in association with estrogenic stimulation are low-grade endometrioid in type, it might be expected that carcinomas associated with tamoxifen would have a similar profile. However, although some studies support this (Turbiner et al. 2008), it has also been suggested that high-

7

Benign Diseases of the Endometrium

grade endometrial cancers, including serous carcinomas and carcinosarcomas (in reality metaplastic carcinomas or carcinomas with sarcomatous differentiation) (McCluggage 2002), are more common in patients taking tamoxifen and that the development of these cancers is not secondary to estrogenic stimulation but due to other mechanisms such as the formation of DNA adducts (McCluggage et al. 2000c; Shibutani et al. 2000). Occasional uterine leiomyosarcomas, endometrial stromal sarcomas, adenofibromas, and adenosarcomas have been described in patients taking tamoxifen (Clement et al. 1996; Huang et al. 1996; McCluggage et al. 1996); it is possible that these complications arise simply by chance since tamoxifen is widely used. Adenomyosis has been shown to be more common in postmenopausal patients taking tamoxifen (Cohen et al. 1997), and this may exhibit unusual morphological features, such as stromal fibrosis, glandular dilatation, and epithelial metaplasias (McCluggage et al. 2000a). A rapid increase in size of uterine leiomyomas has been observed with tamoxifen therapy. Other SERMs, such as raloxifene, have a similar efficacy to tamoxifen in the management of breast cancer. In contrast to tamoxifen however, these seem to be pure estrogen antagonists, lacking the weak estrogen agonist effects of tamoxifen, and do not result in endometrial proliferative lesions (Delmas et al. 1997).

Taxanes Taxanes, such as paclitaxel, are commonly used chemotherapeutic agents as first-line treatment in ovarian, breast, and lung cancer. Taxanes act by the simultaneous promotion of tubulin assembly into microtubules and inhibition of microtubule disassembly. The morphological features in the endometrium have been rarely reported, and numerous mitoses in metaphase arrest with the formation of ring mitoses have been described (Irving et al. 2000). Similar morphological changes are seen in other organs, such as the gastrointestinal tract, in association with taxanes.

411

Endometrial Epithelial Metaplasia (Epithelial Cytoplasmic Change) Endometrial epithelial metaplasias are nonneoplastic epithelial alterations (metaplasias may coexist with endometrial hyperplasia or carcinoma but, in themselves, are nonneoplastic) in which the normal endometrial epithelium is replaced, focally or diffusely, by another type of differentiated epithelium. The variety of epithelial metaplasias encountered within the endometrium reflects the capacity of epithelium derived from the müllerian ducts to undergo differentiation into any other form of epithelium found in the müllerian system. The various epithelial metaplasias commonly coexist. It has been suggested that metaplasia is an inappropriate term for some of the alterations as, strictly speaking, the term metaplasia refers to the replacement of one type of epithelium by another that is not normally found in that organ (Hendrickson and Kempson 1980). For this reason, some authors use the term “epithelial cytoplasmic change” rather than metaplasia. Metaplasia usually involves nonsecretory endometrium and is often associated with hyperestrogenism. Metaplasias are common within endometrial polyps. Other associations include exogenous hormone therapy, especially but not exclusively unopposed estrogens, the presence of an IUD, chronic endometritis, and pyometria; the latter two conditions are particularly associated with squamous metaplasia. In some cases, there is no obvious underlying cause. It has been suggested that progestin therapy given for endometrial hyperplasia or endometrioid adenocarcinoma may result in various epithelial metaplasias within the malignant or premalignant lesion (Wheeler et al. 2007). Metaplasia by itself is not associated with clinical symptoms, but if there is an associated endometrial hyperplasia or carcinoma, there may be abnormal bleeding (McCluggage 2003). A particular problem with epithelial metaplasias is their tendency to be associated with endometrial hyperplasias (Carlson and Mutter 2008) or endometrioid adenocarcinoma. Squamous metaplasia and mucinous metaplasia are particularly

412

common in endometrioid adenocarcinomas. In such instances, it may be difficult, at the lower end of the spectrum, to distinguish between a metaplasia with mild glandular complexity and a hyperplasia with coexistent metaplasia. It is important to make this distinction, since metaplasia by itself has no premalignant potential, and it is recommended that similar criteria are employed to those that are used in the diagnosis of endometrial hyperplasia without metaplasia (see ▶ Chap. 8, “Precursors of Endometrial Carcinoma”). With some epithelial metaplasias, such as clear cell metaplasia and papillary syncytial metaplasia, the differential diagnosis may be between a metaplasia and a type 2 endometrial cancer or serous EIC. In such instances, immunohistochemistry may be of value in that serous EIC often exhibit diffuse, intense, nuclear p53 immunoreactivity while ER may show decreased expression. In contrast, most epithelial metaplasias are ER positive and exhibit a pattern of p53 immunoreactivity, which has been described as weak and heterogeneous (Quddus et al. 1999). A minor population of endometrial epithelial cells exhibit nuclear immunoreactivity with p63; it has been speculated that these are reserve cells or basal cells and the origin of the various epithelial metaplasias (O’Connell et al. 2001).

R. R. Lastra et al.

Typical squamous elements are characterized by sheets of cells exhibiting obvious squamous differentiation in the form of intercellular bridges, prominent cell membranes, or keratinization (Fig. 44). Sometimes there is a histiocytic and giant cell reaction to keratin. Typical squamous metaplasia rarely involves much of or all of the endometrial surface, such that the endometrial cavity is extensively lined by squamous epithelium, a condition known as ichthyosis uteri. This condition most commonly develops secondary to longstanding cervical obstruction or chronic inflammation. Usually, in ichthyosis uteri, the squamous epithelium is of normal appearance, but rarely, it may have the features of a condyloma acuminatum or intraepithelial neoplasia (Stastny et al. 1995). It may rarely extend to involve the fallopian tubes and ovaries. Rarely an invasive squamous carcinoma develops in this manner (Takeuchi et al. 2012). Squamous morules are morphologically distinct structures, which were named so because of their three-dimensional resemblance to mulberries (Dutra 1959). They are composed of rounded aggregates or syncytial sheets of cells that often fill the glandular lumina (Fig. 45). The constituent cells have central bland round, ovoid, or spindle shaped, evenly spaced nuclei, sometimes with small nucleoli. Some of these nuclei may contain

Squamous Metaplasia Squamous metaplasia is one of the most common forms of endometrial epithelial metaplasia. Although usually a focal finding, there may occasionally be widespread squamous metaplasia with obliteration of the glandular lumina, such that it is difficult to assess the underlying glandular component. This is especially common when the squamous metaplasia is of morular type (see below). Squamous metaplasia is common in endometrioid adenocarcinoma and in endometrial hyperplasias; these should be excluded by careful examination of the glandular elements. It may also be seen in endometrial polyps. There are two types of squamous metaplasia, namely, typical squamous metaplasia and morular metaplasia, although these sometimes coexist.

Fig. 44 Squamous metaplasia in the endometrium. Typical squamous metaplasia with obvious squamous differentiation in the form of prominent cell membranes

7

Benign Diseases of the Endometrium

Fig. 45 Squamous morules in the endometrium. Squamous morules are composed of rounded aggregates or syncytial sheets of cells filling the glandular lumina

optically clear biotin-rich inclusions. The cell borders are indistinct. Mitoses are rare or absent. There may be central necrosis. It is controversial whether morules actually exhibit squamous differentiation. Morphological features of overt squamous differentiation, such as keratinization, intercellular bridges, and prominent cell membranes, are typically absent in morules. Immunohistochemically, morules exhibit a somewhat different immunophenotype to that of typical squamous elements. Morules exhibit nuclear and cytoplasmic positivity with betacatenin (Fig. 46) (Brachtel et al. 2005; Saegusa and Okayasu 2001), while in typical squamous elements, the “normal” membranous pattern of immunoreactivity is maintained (Houghton et al. 2008). Endometrial proliferative lesions with morules often exhibit beta-catenin gene mutation, resulting in the above-mentioned nuclear and cytoplasmic immunoreactivity. Morules are usually ER and p63 negative (Chinen et al. 2004) and diffusely positive with CD10 (Chiarelli et al. 2006) and exhibit nuclear immunoreactivity with the intestinal transcription factor CDX2 (Fig. 47) (Houghton et al. 2008; Wani et al. 2008); it has been suggested that this is secondary to beta-catenin gene mutation. In contrast, typical squamous elements are usually positive with ER, p63, and CD10 and negative with CDX2. On the basis of the immunophenotype, it has been concluded that morules exhibit no firm immunohistochemical evidence of squamous differentiation, although

413

Fig. 46 Beta-catenin immunohistochemistry in squamous morules. Morules exhibit nuclear and cytoplasmic immunoreactivity with beta-catenin

Fig. 47 CDX2 immunohistochemistry in squamous morules. Morules exhibit nuclear immunoreactivity with the intestinal transcription factor CDX2

immature squamous features cannot be excluded (Houghton et al. 2008). It has been suggested that the term morular metaplasia is used instead of squamous morules.

Mucinous Metaplasia Mucinous metaplasia is a relatively uncommon form of endometrial epithelial metaplasia and is most commonly seen either in endometrial polyps or in association with a premalignant or malignant lesion. A diagnosis of mucinous metaplasia should be reserved for cases in which the

414

R. R. Lastra et al.

endometrial epithelial cells are replaced by cells with abundant intracytoplasmic mucin, the cells resembling endocervical cells (Fig. 48). Normal endometrial epithelial cells contain a little intracytoplasmic mucin, especially with a luminal distribution, and so abundant intracytoplasmic mucin is required to diagnose mucinous metaplasia. Rarely, intestinal metaplasia has been described in the endometrium where the mucinous epithelium contains goblet cells (Wells and Tiltman 1989). Enteric-type mucin may be demonstrable with mucin stains in cases without morphological evidence of intestinal metaplasia (McCluggage et al. 1995). In mucinous metaplasia without an associated premalignant or malignant glandular proliferation, there are often small micropapillary projections. The nuclei are small and uniform, and mitoses are rare or absent. An important point with a florid mucinous proliferation of the endometrium is that mucinous adenocarcinomas, even those exhibiting myometrial invasion, can be cytologically bland with little in the way of mitotic activity. As such, complex mucinous proliferations of the endometrium present a particular diagnostic problem, especially, but not exclusively, in biopsy material (Vang and Tavassoli 2003). Mucinous proliferations of the endometrium have been divided into three categories depending on the degree of architectural complexity and association with underlying adenocarcinoma (Nucci et al. 1999). With any architecturally complex mucinous proliferation in an endometrial biopsy, a diagnosis of well-

differentiated mucinous adenocarcinoma should be considered, and the term “complex endometrial mucinous proliferation” may be applied with a comment that there is a significant risk of a welldifferentiated mucinous adenocarcinoma in the uterus. Rarely, mucinous metaplasia in the endometrium is accompanied by mucinous lesions elsewhere in the female genital tract, for example, in the cervix and the ovary, perhaps as a manifestation of a field-change effect (Baird and Reddick 1991), or in the context of patients with Peutz–Jeghers syndrome (Tantipalakorn et al. 2009).

Fig. 48 Mucinous metaplasia in the endometrium. Focally the cells have abundant mucinous cytoplasm

Fig. 49 Ciliated metaplasia in the endometrium. The endometrial glands are lined by ciliated cells with abundant eosinophilic cytoplasm

Ciliated (Tubal) Metaplasia Ciliated epithelial cells are normal on the endometrial surface, especially in the proliferative phase of the menstrual cycle. A diagnosis of ciliated metaplasia should be made only when one or more endometrial glands contain ciliated cells, which may be interspersed among non-ciliated cells, or may be extensive and line most of the gland (Fig. 49). The nuclei are cytologically bland, may be rounded and mildly stratified, and contain small nucleoli; the ciliated cells often have abundant eosinophilic cytoplasm. Ciliated metaplasia is particularly associated with estrogenic stimulation. As with other types of epithelial metaplasia, ciliated cells may be found in nonneoplastic, hyperplastic, and malignant endometria. The presence or absence of hyperplasia or adenocarcinoma is evaluated by the usual parameters.

7

Benign Diseases of the Endometrium

415

Clear Cell Metaplasia Clear cell metaplasia is rare and is characterized by replacement of endometrial epithelial cells by cells with abundant clear cytoplasm (Fig. 50). This may be a feature of pregnancy when the features overlap with Arias-Stella reaction. Clear cell metaplasia may be misdiagnosed as clear cell carcinoma, especially on a biopsy specimen. Distinction is based on the bland nuclear features and the fact that in clear cell metaplasia, the endometrial glands maintain a normal architecture and distribution. Other features favoring clear cell metaplasia over clear cell carcinoma include the focal nature of the lesion, absence of a grossly visible tumor, absence of stromal invasion, and strong ER positivity. A significant number, but not all endometrial clear cell carcinomas, are ER negative.

Fig. 50 Clear cell metaplasia in the endometrium. The endometrial glands are replaced by cells with abundant clear cytoplasm

Hobnail Cell Metaplasia Hobnail cell metaplasia or change is rare and is characterized by the presence of cells with rounded apical blebs, which may involve the endometrial surface or protrude into the glandular lumina (Fig. 51). Hobnail cell metaplasia may be a reparative phenomenon following endometrial curettage or may be seen on the surface of a polyp. It may also occur in pregnancy. Hobnail cells are also a feature of some clear cell carcinomas, and this may enter into the differential diagnosis. Criteria useful in the distinction of hobnail cell metaplasia from clear cell carcinoma are similar to those used in distinction of the latter from clear cell metaplasia.

Fig. 51 Hobnail cell metaplasia in the endometrium. Hobnail cells are present on the surface of an endometrial polyp

Eosinophilic (Oxyphilic, Oncocytic) Metaplasia Eosinophilic or oxyphilic metaplasia is relatively common and is characterized by the presence of epithelial cells with abundant eosinophilic cytoplasm (Fig. 52). The cytoplasm may be granular, in which case the term oncocytic metaplasia has been used. The term pink cell metaplasia has also

Fig. 52 Eosinophilic metaplasia in the endometrium. The epithelial cells contain abundant eosinophilic cytoplasm

416

been used. Ultrastructurally, abundant cytoplasmic mitochondria may be present, as is characteristic of oncocytes in other organs. Ciliated metaplasia is often characterized by abundant eosinophilic cytoplasm and overlaps with eosinophilic metaplasia. The epithelial cells in eosinophilic metaplasia can exhibit a significant degree of nuclear atypia; this is analogous to the degenerative nuclear atypia that is common in oncocytic cells in other organs. The main differential diagnosis is the eosinophilic or oxyphilic variant of endometrioid adenocarcinoma. Distinction from adenocarcinoma is based on the absence of a grossly visible lesion and maintenance of the normal glandular architecture.

Papillary Syncytial Metaplasia The term papillary syncytial metaplasia is a misnomer since this does not actually represent a metaplasia but rather a degenerative or reparative phenomenon associated with surface breakdown, either menstrual or non-menstrual in type. However, since the term papillary syncytial metaplasia is in widespread use, it is discussed here. Synonymous terms include eosinophilic syncytial change and surface syncytial change. It has been suggested that papillary syncytial metaplasia is a degenerative or regressive phenomenon based on a low proliferation and mitotic index (Shah and Mazur 2008). Papillary syncytial metaplasia is common and is characterized by small syncytia or micropapillary proliferations of endometrial epithelial cells, which may contain the small glandular lumina and which are devoid of stromal support, lacking fibrovascular stromal cores (Fig. 39). The cells usually have eosinophilic cytoplasm, and there is often a neutrophilic infiltrate. There may be mild nuclear atypia, and in a minority of cases, mitoses are present. The distinction between papillary syncytial metaplasia and serous adenocarcinoma or serous EIC has been discussed earlier. Another important consideration is that foci similar to papillary syncytial metaplasia may occur on the surface of some endometrioid adenocarcinomas. The distinction between papillary syncytial metaplasia and papillary adenocarcinomas of endometrioid or serous type is facilitated by

R. R. Lastra et al.

recognition that papillary syncytial metaplasia is limited to the endometrial surface and is associated with other morphological features of breakdown such as apoptotic debris, neutrophils, and adjacent glandular and stromal breakdown.

Arias-Stella Reaction Arias-Stella reaction (Arias-Stella effect or change) has been discussed earlier and is almost always associated with pregnancy, either intrauterine or ectopic, or with trophoblastic disease. It rarely occurs secondary to hormone therapy, especially progestins; occasionally there is no obvious cause (Dhingra et al. 2007). The most important differential diagnosis is clear cell carcinoma, but the diagnosis of Arias-Stella reaction is usually straightforward if the patient is known to be pregnant and if other morphological features of pregnancy are present, such as decidualization of the stroma. Arias-Stella reaction involves preexisting endometrial glands without evidence of stromal infiltration and there is no mass lesion. Although there is nuclear enlargement and atypia, a low nuclear-to-cytoplasmic ratio is maintained.

Papillary Proliferation of the Endometrium The term hyperplastic papillary proliferation of the endometrium has been used for a lesion, usually occurring in postmenopausal women, characterized by the presence of papillae with fibrovascular stromal cores and variable degrees of branching and cellular tufting (Fig. 53). The papillae are lined by epithelial cells with bland nuclei (Lehman and Hart 2001). Although not strictly a metaplasia, the lesion is discussed here since epithelial metaplasias, most commonly mucinous, eosinophilic, or ciliated, are often also present. Sometimes the papillae are entirely intracystic (projecting into cystically dilated endometrial glands), while in other instances they involve the endometrial surface. Papillary proliferation is most commonly seen on the surface of an endometrial polyp, and, in some instances, the features are florid. There may be an association

7

Benign Diseases of the Endometrium

417

pathogenesis of the various mesenchymal metaplasias. They may result from metaplasia of the endometrial stroma or represent remnants of fetal tissue following abortion or instrumentation. They should be distinguished from heterologous mesenchymal components within a carcinosarcoma or another uterine neoplasm.

Smooth Muscle Metaplasia

Fig. 53 Papillary proliferation of the endometrium. Papillary projections lined by bland epithelial cells on the surface of an endometrial polyp

with hormonal preparations. A misdiagnosis of an adenocarcinoma of endometrioid or serous type is possible, especially if an underlying polyp is not present or not obvious. Awareness of this phenomenon and the realization that it often occurs in a polyp are clues to the diagnosis, although both endometrioid and serous adenocarcinomas may arise in and be confined to a polyp. The absence of nuclear atypia helps to exclude an adenocarcinoma. It has been considered that these papillary proliferations are a form of hyperplasia that is closely associated with epithelial metaplasia. These papillary proliferations can be subdivided into two separate categories, based on the degree of architectural complexity and proliferation (Ip et al. 2013). The first group, characterized by simple, short papillae without significant branching, is usually associated with a benign outcome, and the term “benign papillary proliferation of the endometrium” has been proposed to describe this finding. The second group, characterized by a more complex papillary proliferation, which is often more diffuse or multifocal, has an increased risk of recurrence or concurrent adenocarcinoma, and the term “complex papillary hyperplasia” is appropriate.

Endometrial Mesenchymal Metaplasia Various forms of mesenchymal metaplasia, all of which are rare, may involve the endometrial stroma. There are two theories regarding the

This is the most common mesenchymal metaplasia in the endometrium. Given the common embryonic origin of endometrial stromal and smooth muscle cells, it is thought that a multipotential cell exists in the uterus, which has the capacity to differentiate into endometrial stroma and smooth muscle. This is in keeping with the observation that hybrid endometrial stromal and smooth muscle neoplasms exist. It is not uncommon to find small foci of smooth muscle within the endometrial stroma, and these foci have sometimes been referred to as intraendometrial leiomyomas. It is probable that some intraendometrial smooth muscle nodules are a result of the irregular nature of the normal endometrial–myometrial junction, but others reflect the capacity of endometrial stroma to differentiate into smooth muscle.

Cartilaginous and Osseous Metaplasia Rarely, foci of benign cartilage or bone are found within endometrial biopsies, either within the endometrial stroma or free floating (Bahceci and Demirel 1996). In most cases, it is likely that this is of fetal origin, and this is especially likely if these tissues are found in the endometrium of a young woman with a past history of abortion. In other cases, the cartilage or bone is truly metaplastic. These tissues may also rarely be found within the stroma of an endometrial carcinoma. Benign cartilaginous or osseous metaplasia in an endometrial carcinoma should not be mistaken for the heterologous sarcomatous component of a carcinosarcoma. Rarely, endometrial ossification is associated with Asherman’s syndrome.

418

Glial Metaplasia The presence of glial tissue in the endometrium is extremely rare. In most cases, glial tissue (confirmed if necessary by positive immunohistochemical staining with glial fibrillary acidic protein (GFAP)) is a consequence of a previous abortion. Support for this may come from the identification of other elements such as cartilage or bone.

Adipose Metaplasia Metaplastic adipose tissue is rarely found within the endometrial stroma or within the stroma of an endometrial polyp. If identified in an endometrial biopsy or curettage specimen, the possibility of uterine perforation must be raised. Other explanations for the presence of adipose tissue in an endometrial biopsy specimen include derivation from a lipoma, a lipoleiomyoma, a uterine hamartomatous-like lesion containing adipose tissue (McCluggage et al. 2000b), or a carcinosarcoma.

Extramedullary Hematopoiesis Rarely, foci of extramedullary hematopoiesis are present in the endometrium, usually in association with an underlying hematopoietic disorder or occasionally representing remnants of fetal tissue (Creagh et al. 1995).

Endometrial Polyps Endometrial polyps are common and have been identified in between 2% and 23% of patients undergoing endometrial biopsy because of abnormal uterine bleeding (Schindler and Schmidt 1980). Polyps occur in pre- and postmenopausal women and are thought to be related in some way to hyperestrogenism, possibly originating as a localized hyperplasia of the endometrial basalis secondary to hormonal influences. There is an increased incidence of endometrial polyps with HRT, either estrogen-only HRT or combined

R. R. Lastra et al.

preparations. Tamoxifen is also associated with an increased risk of the development of endometrial polyps (see section “Tamoxifen”). Molecular studies have demonstrated that many endometrial polyps represent monoclonal endometrial stromal overgrowths (commonly with abnormalities of chromosome 6) with secondary induction of polyclonal benign glands through undefined stromal–epithelial interactions (Dal Cin et al. 1992; Fletcher et al. 1992). Polyps may be single or multiple, sessile or broad based, and pedunculated or attached to the endometrium by a slender stalk. They usually have a smooth surface, and small cysts may be seen on sectioning. They can arise anywhere in the endometrium, including the lower uterine segment, but are most common in the fundus. When large, they may fill the endometrial cavity and extend into the endocervical canal. The pathological diagnosis is generally straightforward if the gynecologist is aware of the presence of a polyp, has conveyed this information to the pathologist, and has removed the polyp intact. On occasion, the gynecologist believes that a polyp is present, but histological examination shows a cyclical endometrium, often secretory in type, reflecting the fact that an abundant secretory endometrium may have a polypoid appearance. In many cases, the gynecologist is not aware of the presence of a polyp, which is removed piecemeal with the result that in biopsy material, fragments derived from the polyp are admixed with fragments of non-polypoid endometrium, making the diagnosis difficult. In biopsies performed because of abnormal uterine bleeding, the pathologist should always consider the possibility of a polyp. Under low-power examination, the initial clue to the diagnosis is often the admixture of fragments of normal cyclical or atrophic endometrium and fragments that are morphologically different. The histological features of a polyp, not all of which are present in every case, include the following: 1. Polypoid pieces of tissue lined by epithelium on three sides. 2. Glands set in a stroma that is qualitatively different than the endometrial stroma in the

7

Benign Diseases of the Endometrium

419

Fig. 54 Endometrial polyp. Dilated glands are set in a fibrous stroma

The glands within a polyp are usually endometrioid in type, but not uncommonly exhibit metaplastic change, including ciliated, eosinophilic, mucinous, and squamous metaplasia. The epithelium may be atrophic, but often exhibits proliferative activity, even when the patient is postmenopausal and the surrounding endometrium is atrophic. The presence of proliferative activity in a polyp in a postmenopausal woman is of no clinical importance, although it is useful to comment in the pathology report on whether non-polypoid endometrium is also present and whether this exhibits proliferative activity. The stroma of a polyp is often more fibrous than that of the non-polypoid endometrium, but this is not invariable, and, in some polyps, the stroma is dense and cellular, resembling that of normal proliferative endometrium. As stated, collections of thick-walled stromal blood vessels are a characteristic feature of endometrial polyps, and ectatic thin-walled vessels are also sometimes seen. Some authors have divided endometrial polyps into different types, such as proliferative/hyperplastic (proliferative glands with or without glandular crowding), atrophic (atrophic glands), and functional (glands resembling those in the surrounding cyclical endometrium). However, these patterns often overlap, and assignment to a specific type may be difficult; moreover, there is no clinical significance attached to the different types. Some polyps originate at the junction of the upper endocervix and lower uterine segment

Fig. 55 Endometrial polyp. There may be a mild degree of glandular crowding within some endometrial polyps

Fig. 56 Endometrial polyp. Collections of thick-walled stromal blood vessels are a characteristic feature of endometrial polyps

non-polypoid fragments. The stroma is often, but not always, more fibrous than that in the non-polypoid fragments and is sometimes markedly hyalinized (Fig. 54). 3. Glandular architectural abnormality with dilated glands and sometimes mild glandular crowding (Fig. 55). 4. Glands that appear different to those in the surrounding endometrium; for example, the glands of the non-polypoid endometrium may be secretory in type while the glands within the polyp are atrophic or exhibit poorly developed secretory or proliferative activity. 5. Collections of thick-walled stromal blood vessels (Fig. 56).

420

and contain both endocervical and ciliated lower uterine segment type glands. The variety of morphological appearances affecting the epithelium or stroma that may be encountered in endometrial polyps can result in diagnostic difficulty. Papillary proliferations with fibrovascular cores (see section “Papillary Proliferation of Endometrium”) occasionally occur on the surface of an endometrial polyp or within cystically dilated glands. The epithelium on the surface of a polyp may exhibit a degree of atypia, often with degenerate-appearing nuclei (Fig. 57), and sometimes hobnail cell change. There may be focal surface glandular and stromal breakdown, and on occasions, polyps are extensively necrotic secondary to torsion or if they outgrow their blood supply (Fig. 58); vascular thrombosis may be seen in such cases. In some instances, this may result in the formation of a necrotic polypoid mass with only the surface epithelium or the ghost outlines of glands remaining. Variable amounts of stromal edema and occasionally myxoid change may be present as well as hemosiderin pigment and foamy histiocytes. Sex cord-like areas have rarely been described in the stroma of an endometrial polyp (De Quintal and De Angelo Andrade 2006). Some endometrial polyps contain bundles of smooth muscle within the stroma, often in close proximity to thick-walled blood vessels. This is usually a minor feature and of no significance. However, when the smooth muscle is prominent, the term adenomyomatous polyp has been used (see section “Adenomyomatous Polyp”). Stromal inflammatory cells, including plasma cells, may be present in endometrial polyps; this should not be interpreted as an endometritis unless plasma cells are also present in the non-polypoid endometrium. Stromal decidualization or pseudodecidualization may occur secondary to progestational compounds, but the degree of decidualization is typically less than in the surrounding endometrium. As discussed previously, there is an increased frequency of endometrial polyps in patients taking tamoxifen. The polyps may be single or multiple and may be large. There are no pathognomonic histological features of tamoxifen-associated endometrial polyps, but an increased incidence of

R. R. Lastra et al.

Fig. 57 Endometrial polyp. The epithelium on the surface of a polyp may exhibit a degree of nuclear atypia with a degenerate appearance

Fig. 58 Necrotic endometrial polyp. Necrosis has occurred secondary to torsion

epithelial metaplasias, periglandular stromal condensation, stromal edema, myxoid change, and staghorn-shaped glands with intraluminal polypoid projections, polarized along the long axis of the polyp, have been reported (Kennedy et al. 1999). As stated, diagnosing an endometrial polyp is generally straightforward when the polyp is large and removed intact. However, when small and fragmented, the diagnosis is more difficult. Lower uterine segment endometrium may be mistaken for a polyp because of the irregular glandular architecture and fibrous stroma. The spindle cell alteration of the stroma seen in some cases of endometritis may resemble the fibrous stroma of a polyp. However, other morphological features of a polyp are absent, and there is a plasma cell

7

Benign Diseases of the Endometrium

infiltrate within the stroma; it should be remembered, however, that plasma cells may occur within the stroma of an endometrial polyp. In cases of large polyps with a degree of stromal condensation and increased cellularity around the glands, the differential diagnosis may include an adenosarcoma. Adenosarcoma typically has a leaf-like or club-like architecture, with broad papillae lined by surface epithelium, and intraglandular stromal projections, the overall architecture resembling a phyllodes tumor of the breast. In contrast, endometrial polyps usually have a smooth outline. The stroma in adenosarcoma is usually more cellular than in a benign polyp with increased mitotic activity and a degree of nuclear atypia, especially immediately surrounding the glands. With multiple recurrent endometrial polyps, a diagnosis of adenosarcoma should be suspected since the morphological features may be subtle. Adenofibroma may also be considered, but this is rare and exhibits a similar low-power architecture to adenosarcoma. In polyps with a stromal smooth muscle component, an atypical polypoid adenomyoma may be considered. However, the stroma of atypical polypoid adenomyoma exhibits more extensive smooth muscle differentiation, the glandular architecture is more complicated, and there is often extensive squamous morule formation (see section “Atypical Polypoid Adenomyoma”). Polyps in which the glands are mildly crowded and exhibit proliferative activity may be confused with an endometrial hyperplasia. In such cases, the identification of normal background endometrium facilitates the diagnosis, since hyperplasia is usually a diffuse process.

421

Adenomyomatous Polyp As discussed, some endometrial polyps contain a minor component of stromal smooth muscle bundles, often in close proximity to thick-walled blood vessels. When the smooth muscle is prominent, the term adenomyomatous polyp has been used. This term and the term adenomyoma have also been used for a lesion in which endometrioid-type glands, sometimes with minor foci of ciliated, mucinous, or squamous metaplasia, are surrounded by endometrial stroma, which is in turn surrounded by smooth muscle (Fig. 59) (Gilks et al. 2000; Tahlan et al. 2006). These lesions, which may also be non-polypoid and located entirely within the myometrium, can be associated with underlying adenomyosis. They should not be confused with atypical polypoid adenomyoma; as mentioned above, the glands in adenomyomas are usually surrounded by endometrioid-type stroma, which is in turn surrounded by smooth muscle.

Hyperplasia and Carcinoma Arising in an Endometrial Polyp Occasionally, a hyperplasia or carcinoma arises within or involves an endometrial polyp (Carlson and Mutter 2008). It is recommended by some that a diagnosis of simple hyperplasia should not be made in a polyp, since a degree of glandular crowding and proliferative

Endometrial Polyp with Atypical Stromal Cells Rare endometrial polyps contain stromal cells with markedly atypical symplastic-like nuclei, resembling those seen in polypoid lesions elsewhere in the female genital tract, such as in fibroepithelial stromal polyps of the vulva and vagina (Tai and Tavassoli 2002). These atypical stromal cells are of no significance (see Fig. 71).

Fig. 59 Adenomyomatous polyp (adenomyoma). Endometrioid-type glands are surrounded by endometrialtype stroma, which is in turn surrounded by smooth muscle

422

activity within dilated glands are features of many endometrial polyps. Complex and atypical hyperplasia is diagnosed in the same way as in non-polypoid endometrium. The hyperplasia may be confined to the polyp, but also involves the non-polypoid background endometrium in approximately 50% of cases (Kelly et al. 2007; Morsi et al. 2000). Both endometrioid and serous carcinomas (and occasionally other malignancies) may arise within or involve an endometrial polyp and sometimes be confined to this. Serous carcinomas have a particular tendency to arise in or involve endometrial polyps, as does the precursor lesion (serous EIC) (Hui et al. 2005; Silva and Jenkins 1990). When serous EIC involves an endometrial polyp, there is partial replacement of the surface epithelium or sometimes the epithelium of the glands by cells with markedly atypical hyperchromatic nuclei, sometimes with prominent nucleoli, and prominent mitotic activity (Fig. 60). Immunohistochemical staining may assist in highlighting the serous proliferation, since the cells of serous EIC typically exhibit either diffuse, intense, nuclear immunoreactivity with p53 (Fig. 61) or complete absence of p53 expression, with a high MIB1 proliferation index; ER expression is usually decreased (see ▶ Chap. 8, “Precursors of Endometrial Carcinoma”). In contrast, the benign epithelium within the polyp exhibits a low MIB1 proliferation index and is ER positive, and only scattered nuclei are weakly positive with p53. p53 staining may reveal that the serous EIC is more extensive than is appreciated on initial morphological examination. Rarely, a carcinosarcoma arises in and is confined to an endometrial polyp. Occasional cases of metastatic carcinoma, especially breast lobular carcinoma, have been reported in endometrial polyps (Houghton et al. 2003).

Atypical Polypoid Adenomyoma Atypical polypoid adenomyoma is a biphasic polypoid lesion composed of endometrioid-type glands in a myomatous or fibromyomatous stroma

R. R. Lastra et al.

Fig. 60 Endometrial polyp with serous EIC. Serous EIC may arise on the surface of an endometrial polyp

Fig. 61 p53 immunohistochemistry in serous EIC. Diffuse, intense, nuclear p53 immunoreactivity of serous EIC on the surface of an endometrial polyp

(Mazur 1981; Young et al. 1986). Since the stroma may be fibromyomatous rather than overtly myomatous, some prefer the designation atypical polypoid adenomyofibroma (Longacre et al. 1996). Most patients are premenopausal or perimenopausal (average age 40 years) and present with abnormal uterine bleeding, usually in the form of menorrhagia. In some cases, the diagnosis is made during investigations for infertility. Occasional cases occur in postmenopausal women, and rare examples have been described in patients with Turner’s syndrome who have been prescribed unopposed estrogens (Clement and Young 1987). A single study has investigated molecular

7

Benign Diseases of the Endometrium

events in atypical polypoid adenomyoma and found MLH-1 promotor hypermethylation in some cases, a molecular alteration characteristic of some atypical hyperplasias and endometrioid adenocarcinomas (Ota et al. 2003). Atypical polypoid adenomyoma is most commonly located in the lower uterine segment, although some cases involve the fundus, uterine body, or endocervix. In most cases, the lesion has an obvious polypoid gross appearance, in the form of either a sessile or broad-based polyp, but sometimes the polypoid nature is not grossly obvious, especially in smaller lesions. The diagnosis may be made on endometrial biopsy, polypectomy, or at hysterectomy. Histology shows architecturally irregular endometrioidtype glands that may be widely separated and haphazardly arranged or somewhat crowded and arranged in groups, sometimes with a vaguely lobular pattern (Fig. 62). The endometrioid epithelium varies in appearance from cuboidal to low columnar to pseudostratified. The nuclei are usually round, sometimes with prominent nucleoli, and exhibit mild or, at the most, moderate cytological atypia. Occasional foci of ciliated or mucinous epithelium may be present. A characteristic histological feature that is present in most, but not all, cases is abundant squamous morule formation (Fig. 63); sometimes, the morules exhibit central necrosis. The glands are set in an abundant stroma, which varies from obviously smooth muscle in nature to fibromyomatous. Endometrial stroma is not present. The stromal cells are often arranged in short interlacing fascicles. Occasional mitotic figures may be identified within the stroma. The margin between the lesion and the underlying myometrium is usually rounded and well delineated, but occasionally there is merging with underlying adenomyosis. In some cases, there is significant glandular crowding with a back-to-back architecture and stromal exclusion, such that there are foci that are virtually indistinguishable from, and which are best regarded as, grade I endometrioid adenocarcinoma. The term atypical polypoid adenomyoma of low malignant potential has been used for lesions with marked architectural complexity (Longacre et al. 1996), but this term is not recommended.

423

Fig. 62 Atypical polypoid adenomyoma. Endometrioidtype glands are embedded within a myomatous stroma, displaying a vague lobular pattern

Fig. 63 Atypical polypoid adenomyoma. There is abundant squamous morule formation

Very rarely, there is underlying myometrial invasion, and/or an endometrioid adenocarcinoma is present in the surrounding endometrium. Atypical polypoid adenomyoma is generally a benign lesion, but there is a risk of recurrence if curettage or polypectomy is undertaken. In one series, 45% of cases treated by curettage or polypectomy recurred (Longacre et al. 1996). Given this risk of recurrence and the small but definite risk of transition to endometrioid adenocarcinoma, which was estimated at 8.8% in one meta-analysis (Heatley 2006), hysterectomy is the treatment of choice if the diagnosis is made on biopsy or polypectomy. In a woman who wishes to retain her uterus and in

424

whom a confident diagnosis of atypical polypoid adenomyoma has been made on biopsy or polypectomy, complete removal by curettage or polypectomy may be undertaken with close follow-up and imaging. Successful pregnancies have ensued in patients managed in this way. It has been suggested that recurrence is more likely in cases with marked architectural complexity. The most important differential diagnosis of atypical polypoid adenomyomas is an endometrioid adenocarcinoma exhibiting myometrial invasion or with a prominent desmoplastic stroma, an obviously important distinction since most atypical polypoid adenomyomas exhibit a benign behavior with a potential for conservative management. Recognition of the polypoid nature of the lesion assists in establishing the diagnosis. Marked cytological atypia favors a myoinvasive adenocarcinoma since in atypical polypoid adenomyomas, there is usually no more than mild to moderate cytological atypia. The stromal component of atypical polypoid adenomyoma grows in short interlacing fascicles, in contrast to the elongated fibers of the normal myometrium. In curettings or biopsy from an atypical polypoid adenomyoma, there are usually also fragments of normal background endometrium, and with an endometrioid adenocarcinoma, it would be unusual on biopsy to obtain only myoinvasive neoplasm without free tumor fragments. Immunohistochemistry is of little value in distinguishing between atypical polypoid adenomyoma and a myoinvasive endometrioid adenocarcinoma, since the stromal component of atypical polypoid adenomyoma and myometrium infiltrated by carcinoma are both desmin and smooth muscle actin positive (Soslow et al. 1996). It has been suggested that CD10 may be of value, since this is negative in the stromal component of atypical polypoid adenomyoma, while the myoinvasive glands of endometrioid adenocarcinoma are typically surrounded by CD10-positive stromal cells (Ohishi et al. 2008). The differential diagnosis can also include a benign endometrial polyp in which there may be a minor component of smooth muscle within the stroma. So-called typical adenomyomatous polyps or adenomyomas (see section “Adenomyomatous Polyps”) also occur and are composed of benign endometrioid-type glands in a myomatous stroma (Gilks et al. 2000; Tahlan

R. R. Lastra et al.

et al. 2006). Often, endometrial-type stroma surrounds the glands, and this in turn is surrounded by smooth muscle. Rarely, a carcinosarcoma enters into the differential diagnosis because of the admixture of epithelial and stromal elements. However, both the epithelial and mesenchymal components of a carcinosarcoma are obviously malignant and typically high grade.

Adenofibroma Adenofibroma is a mixed tumor of the endometrium (and rarely also of the cervix) consisting of a benign epithelial and a benign mesenchymal component, both of which are integral components of the neoplasm. Adenofibromas most commonly occur in postmenopausal women, but the age range is wide. The most common presenting symptom is abnormal vaginal bleeding. Occasional cases have arisen in patients taking tamoxifen (Huang et al. 1996). Adenofibroma is rare and must be distinguished from the more common adenosarcoma with a subtle malignant stromal component (Zaloudek and Norris 1981) (discussed in ▶ Chap. 10, “Mesenchymal Tumors of the Uterus”). Grossly, adenofibroma occupies the uterine cavity, typically arising as a broad-based polypoid mass. The cut surface may be spongy or overtly cystic. Bland epithelium that is usually of endometrioid type but which may be mucinous, ciliated, or even squamous covers broad or fine papillary stromal fronds (Fig. 64). Cells of fibroblastic type, and more rarely endometrial

Fig. 64 Adenofibroma of the endometrium. Adenofibroma containing benign glands with underlying bland fibrous stroma

7

Benign Diseases of the Endometrium

stroma or smooth muscle, make up the mesenchymal element. The stroma may be cellular or fibrous, and the constituent cells are cytologically bland without nuclear pleomorphism. Mitotic figures are usually absent and should not exceed 1 per 10 highpower fields; greater mitotic activity than this warrants a diagnosis of adenosarcoma, which is more common. Periglandular cuffing by stromal cells should also result in consideration of adenosarcoma. Occasionally, an adenocarcinoma arises within an adenofibroma, but this is probably a coincidental association (Venkatraman et al. 2003). While the adenofibroma is benign, hysterectomy is the most appropriate treatment. This is because adenosarcoma cannot be excluded in curettings or biopsy and also because recurrence of an adenofibroma treated by curettage or local excision may occur. It is for these reasons that some consider that adenofibroma does not exist, but rather is a well-differentiated form of adenosarcoma. At the heart of the controversy whether or not adenofibroma exists as an entity is the ability to distinguish it from adenosarcoma. The important features to be evaluated are the degree of mitotic activity in the stroma, the morphology of the stromal cells, and the presence of periglandular cuffing by stromal cells. Traditionally, a mitotic count greater than 1 per 10 high-power fields warrants a diagnosis of adenosarcoma, a diagnosis that should also be made if there is more than mild nuclear atypia or significant periglandular stromal cuffing. However, there can be significant morphological overlap between adenofibromatous endometrial polyps and adenosarcomas, and the presence of focal or poorly developed features can be allowed in a polypoid endometrial lesion, without the lesion behaving as an adenosarcoma (Howitt et al. 2015). Because of this considerable overlap, it is difficult to make a confident diagnosis of adenofibroma on curetted or avulsed material, and theoretically, an adenosarcoma cannot be excluded unless the whole tumor is available for examination. Thus, a hysterectomy is required to ensure that the tissue examined was not just the most benign area of an adenosarcoma. Cases have been reported in which repeated curettages have been carried out because the lesion was wrongly thought to be benign before a diagnosis of adenosarcoma has eventually been

425

made (Clement and Scully 1990). It is also important to differentiate between adenofibroma and a benign endometrial polyp since the latter does not require further treatment. While the distinction can be difficult, adenofibroma should be considered if the lesion has a papillary surface and a stroma that is cellular. The stroma of endometrial polyps tends to be more hyaline than that of adenofibroma, although there is considerable overlap.

Effects of IUD IUDs are in widespread use, mainly for contraceptive purposes. Older devices were usually composed entirely of plastic, while more modern devices are often composed of plastic with a copper coating. The effects of the Mirena coil, a progestin-containing IUD, are discussed in the next section (see section “Effects of Mirena Coil”). The histological features in the endometrium in association with an IUD are largely due to local mechanical effects. The surface endometrium may take on the configuration of the IUD due to a direct pressure effect. There may be surface micropapillary formations and focal reactive changes with nuclear enlargement, mild nuclear atypia, small nucleoli, and cytoplasmic vacuolation (Fig. 65). Micropapillary fragments of epithelium exhibiting these features in an endometrial biopsy may result in diagnostic difficulty,

Fig. 65 Reaction to IUD. Micropapillary formations are present, in keeping with a reaction to an IUD

426

and the clinical history of the presence of an IUD is helpful. The glandular epithelium may also exhibit epithelial cytoplasmic change, including squamous metaplasia, and there may be surface ulceration. The endometrial glands at the site of contact, and immediately surrounding this, may exhibit a different pattern of maturation from the rest of the endometrium. With long-term usage, the endometrium adjacent to the IUD may be fibrosed or simplified and composed of a monolayer. Rarely there is stromal microcalcification. Although the features may be subtle, a focal inflammatory infiltrate is commonly present consisting of neutrophils, lymphocytes, histiocytes, and plasma cells. Foreign body-type giant cells and granulomas may be a component of the inflammatory infiltrate, the severity of which may be related to the type of IUD and the duration of use. In most cases, the inflammation is superficial and largely confined to the site of contact, but in other instances it is more widespread. When confined to the IUD site, the inflammation is probably a consequence of irritation rather than infection, but when more widespread, it may be secondary to infection; in such cases, the inflammatory infiltrate, as well as being more widespread, is typically more intense. Infection appears more common with plastic devices than with coppercoated devices. Usually, a mixture of organisms is present. Long-term IUD use is associated with infection by the Gram-positive anaerobic bacterium Actinomyces (discussed earlier). A rare but serious complication of IUD use is uterine perforation or laceration, most commonly occurring at the time of insertion. The risk of perforation is greatest in the postpartum period, when the tissues are soft and expanded. Displacement of the IUD into the pelvis with an associated inflammatory reaction may ensue secondary to perforation. Occasionally, spontaneous expulsion may also occur.

R. R. Lastra et al.

licensed for the management of idiopathic menorrhagia and has been used to deliver progestin for endometrial protection in postmenopausal hormone replacement regimes. It has also been suggested that the Mirena coil can be used to prevent endometrial hyperplastic changes in patients taking tamoxifen or estrogen-only HRT. The histological features of the endometrium in association with the Mirena coil include a low-power polypoid architecture (Fig. 66), probably secondary to a direct mechanical effect. Ulceration, surface micropapillary proliferations, and reactive atypia of the surface endometrium may also be present. The endometrial glands are usually small, tubular, and atrophic, but occasionally exhibit weak secretory activity. There is stromal expansion and decidualization or pseudodecidualization with infiltration by granulated lymphocytes (Fig. 67) (Phillips et al. 2003). Other histological features found in some cases include stromal myxoid or mucinous change, hemosiderin pigment, and glandular metaplastic changes. Stromal necrosis, infarction, and microcalcifications are found in a small percentage of cases. In some cases, plasma cells are present within the stroma, indicating a coexistent chronic endometritis, secondary to the presence of the IUD. Stromal hyaline nodules have also been described (Fig. 68) (Hejmadi et al. 2007). There may be associated progestational effects in the cervix, including microglandular hyperplasia and stromal decidualization.

Effects of Mirena Coil The Mirena coil is a levonorgestrel (a progestin)releasing IUD that is in widespread use as a highly effective contraceptive. The Mirena coil is also

Fig. 66 Mirena coil-associated endometrium. There is a low-power polypoid architecture

7

Benign Diseases of the Endometrium

Fig. 67 Mirena coil-associated endometrium. There is stromal expansion and predecidualization. Hemosiderin pigment and inflammation are present within the stroma

Fig. 68 Mirena coil-associated endometrium. Stromal hyaline nodules may be seen in Mirena coil-associated endometrium

427

Fig. 69 Radiation effect on the endometrium. Glands are lined by cells with enlarged atypical nuclei

eosinophilic, clear, or vacuolated. The normal glandular architecture is typically maintained, or there may be glandular loss. The stroma can be fibrotic with associated vascular changes characteristic of radiation. The presence of cells with enlarged atypical nuclei can raise the possibility of serous EIC. Knowledge of the history of prior radiation is obviously paramount in establishing the diagnosis; a lack of mitotic activity and a low nuclear-to-cytoplasmic ratio are other diagnostic clues. p53 immunohistochemistry may assist, since serous EIC usually exhibits diffuse, intense, nuclear immunoreactivity, while in radiation effect the nuclei exhibit weak heterogeneous staining. Prior radiation is associated with an increased risk of subsequent development of uterine malignancies. These may be of any morphological type, but carcinosarcomas are proportionally overrepresented.

Radiation Effects on the Endometrium The endometrial morphology can be altered by the effects of radiation, which may have been administered many years earlier. Radiation effect in the endometrium is characterized by surface endometrial epithelium or glands lined by cells with enlarged atypical, sometimes bizarre, hyperchromatic nuclei (Fig. 69). The nuclear chromatin may be smudged and indistinct, or there may be prominent nucleoli. Hobnail cells may be present. There is usually abundant cytoplasm that can be

Effects of Endometrial Ablation or Resection Endometrial ablation is commonly undertaken as a nonsurgical procedure in the management of abnormal uterine bleeding, especially in premenopausal women where the suspicion of malignancy is low and preservation of fertility is not an issue. The object of endometrial ablation is destruction of the entire endometrium and

428

superficial myometrium using thermal coagulation (electrosurgical rollerball), laser vaporization, or resection. Most patients become amenorrheic or hypomenorrheic following the ablation, but some have continuous bleeding and/or pain, and repeat endometrial biopsy or sometimes hysterectomy is undertaken either soon after the procedure or some time later. The histological features depend on the time interval between the ablation and the subsequent biopsy or hysterectomy. In the early stages (up to 3 months), there may be complete or almost complete necrosis of the endometrium and superficial myometrium with replacement of this by a coagulum of necrotic fibrinoid material with a surrounding histiocytic and giant cell reaction (Fig. 70), the morphological features resembling that of a rheumatoid nodule or a necrotizing granulomatous process (Colgan et al. 1999; Ferryman et al. 1992). Spicules of thermally damaged tissue may be identified, and there is a variable degree of inflammation. Later, necrotic tissue is no longer present, but a giant cell and granulomatous reaction may persist, sometimes with pigment within giant cells. There is usually striking fibrosis, and there may be regeneration of the endometrium with the formation of a monolayer of simple cuboidal epithelium directly abutting the myometrium. In other cases, the endometrium is histologically relatively normal. Scarring may

R. R. Lastra et al.

result in obstruction and the subsequent development of hematometra or pyometra. The histological features following endometrial resection are similar (McCulloch et al. 1995). Usually, an endometrial biopsy is taken just prior to ablation being performed. Rarely this contains an unsuspected carcinoma, and hysterectomy is performed soon after the ablation. In such instances, the endometrial carcinoma may have undergone total or extensive necrosis, secondary to the ablation procedure.

Effects of Curettage Morphological changes may be seen in the endometrium secondary to recent curettage, especially if this has been vigorous. The changes are transient and will only be seen if repeat endometrial sampling or hysterectomy is undertaken a short time after curettage. There may be focal surface erosion with an associated mixed inflammatory infiltrate; in some cases, eosinophils are conspicuous (Miko et al. 1988). Following this, there is usually epithelial regeneration, sometimes with a micropapillary architecture, hobnail cell change, and mild reactive nuclear atypia. Usually, the changes are minor, result in no particular diagnostic difficulties, and return to normal within 10–14 days. A similar phenomenon has been described in the endocervix secondary to recent endometrial sampling and termed atypical reactive proliferation of the endocervix (Scott et al. 2006).

Asherman’s Syndrome

Fig. 70 Reaction to previous endometrial ablation. Endometrium is replaced by fibrinoid material with a surrounding histiocytic and giant cell reaction

Asherman’s syndrome is characterized by focal or diffuse endometrial fibrosis and loss of distinction between the functionalis and basalis. The endometrium may be composed of a monolayer of epithelium with underlying fibrous tissue, and adhesions may form across the cavity. The endometrial stroma is fibrosed and may be calcified and, on rare occasions, even ossified. The endometrial glands are typically sparse and inactive and may be cystically dilated. Asherman’s syndrome can

7

Benign Diseases of the Endometrium

result in infertility, amenorrhea, or hypomenorrhea. Pregnancies may be complicated by premature labor, placenta previa, or placenta accreta. In some, but not all, cases, there is a known cause for the Asherman’s syndrome, which may be secondary to previous surgery, curettage or ablation, infection, miscarriage, or retained products of conception. Hysteroscopic lysis of adhesions can result in successful pregnancies.

Postoperative Spindle Cell Nodule Postoperative spindle cell nodule represents an exaggerated reparative response at the site of previous surgery or biopsy and, in the female genital tract, is most common in the vagina. It has rarely been described in the endometrium (Clement 1988). Postoperative spindle cell nodule occurs within weeks or months of the inciting procedure and consists of a cellular proliferation of spindleshaped cells with admixed vascular channels and inflammatory cells. There may be abundant mitotic activity, raising the possibility of a sarcoma, usually a leiomyosarcoma. However, the high mitotic activity contrasts with the bland nuclear features and lack of cytological atypia. The history of a recent procedure is obviously paramount in establishing the diagnosis. The constituent cells may be positive with smooth muscle actin, desmin, and, on rare occasions, cytokeratins.

Psammoma Bodies in the Endometrium Calcified psammoma bodies may be seen in the endometrium in association with benign and malignant lesions and rarely in normal endometrium. They are present in up to one-third of uterine serous carcinomas, a lower frequency than in ovarian serous carcinomas (Hendrickson et al. 1982). More uncommonly, they are seen in other morphological types of endometrial malignancy, such as endometrioid carcinoma. Psammoma bodies are occasionally seen in normal endometria (usually atrophic or proliferative

429

in type), sometimes in association with hormonal preparations, and in benign lesions, most commonly endometrial polyps (Herbold and Magrane 1986; Truskinovsky et al. 2008). They are often located within the glandular lumina, in which case it is likely that they represent calcification of inspissated secretions. In other cases, they are situated within the stroma where they may be secondary to prior inflammation or the presence of an IUD. In the absence of a malignant lesion, the occurrence of psammoma bodies within the glandular lumina or endometrial stroma is not an indication for the evaluation of the upper female genital tract to exclude malignancy. However, free-floating psammoma bodies without attachment to tissue may be an indication of an extrauterine serous carcinoma.

Emphysematous Endometritis There have been occasional reports of emphysematous (pneumopolycystic) endometritis characterized by the presence of gas-filled cysts in the endometrial stroma (Perkins 1960; Val-Bernal et al. 2006). There may be simultaneous involvement of the cervix, or the condition may be confined to the endometrium. Histology shows empty cystic spaces of variable sizes and contour within the endometrial stroma lined by flattened stromal cells with occasional histiocytes and/or giant cells. Spontaneous resolution usually occurs. Emphysematous endometritis should be distinguished from sectioning artifact, dilated vascular spaces, and gas gangrene of the uterus, which is life-threatening and associated with tissue necrosis.

Benign Endometrial Stromal Proliferations Endometrial stromal neoplasms are discussed in ▶ Chap. 10, “Mesenchymal Tumors of the Uterus,” as is the differential diagnosis of fragments of tissue composed entirely of endometrial stroma in an endometrial biopsy. Occasional cases of multifocal microscopic benign endometrial

430

Fig. 71 Atypical endometrial stromal cells. Rarely, endometrial stromal cells with atypical symplastic-like nuclei are present in the normal endometrium and in some endometrial polyps. (Courtesy of Dr. Brigitte Ronnett)

R. R. Lastra et al.

Fig. 72 PSNP. A well-circumscribed lesion is composed of epithelioid cells with abundant cytoplasm

stromal proliferations confined to the endometrium without invasive growth have been reported (Stewart et al. 1998). These have been termed focal endometrial stromal hyperplasia and may mimic an endometrial stromal nodule or endometrial stromal sarcoma in biopsy samples. Rarely, markedly atypical stromal cells with a symplastic appearance are present within an otherwise normal endometrium and can also be found in some endometrial polyps (Fig. 71) (Usubutun et al. 2005).

Benign Trophoblastic Lesions Benign lesions of intermediate trophoblast, namely, placental site nodule or plaque (PSNP) and exaggerated placental site, are discussed in ▶ Chap. 20, “Gestational Trophoblastic Tumors and Related Tumorlike Lesions.” An endometrial PSNP may be identified in a biopsy, resection, or hysterectomy specimen many years following a pregnancy or abortion and rarely even in a postmenopausal woman. PSNP is characterized histologically by a well-circumscribed lesion composed of cells with degenerating large, often atypical, nuclei with an abundant eosinophilic cytoplasm (Fig. 72). The immunophenotype is discussed in detail later in ▶ Chap. 20, “Gestational Trophoblastic Tumors and Related Tumorlike Lesions.”

Fig. 73 Intravascular menstrual endometrium. Menstrual endometrium within myometrial blood vessels is occasionally seen and is of no significance

Intravascular Endometrium Occasionally in a hysterectomy specimen, menstrual endometrium is present within myometrial vascular channels (Fig. 73) (Banks et al. 1991). Rarely there is extensive vascular involvement and even invasion of parametrial vessels. Intravascular menstrual endometrium is of no significance other than it may be mistaken for neoplastic involvement of vascular channels. It should also be distinguished from intravascular foci of adenomyosis (Sahin et al. 1989).

7

Benign Diseases of the Endometrium

Endometrial Autolysis Hysterectomies are performed for a wide variety of benign and malignant conditions. Especially with large uteri, the formalin does not penetrate into the endometrial cavity, and, as a consequence, fixation is often not adequate and the endometrium may be markedly autolyzed. This not uncommonly results in problems in morphological assessment of the endometrium, which in some cases is so autolyzed that interpretation is impossible. This is important in all uteri but particularly so with endometrial neoplasms, where autolysis may result in problems in assessing tumor type and grade. Bisection of the uterus soon after surgery may improve fixation, but this often results in distortion of the specimen, making assessment of parameters such as the depth of myometrial invasion by tumor problematic. Packing the cavity with absorbable tissue paper following bisection and ensuring that the two halves of the uterus remain apposed minimizes the distortion. Uteri may also be injected with formalin using a needle and syringe directed alongside a probe, which is inserted through the external cervical os into the endometrial cavity (Houghton et al. 2004). This initiation results in significantly less endometrial autolysis.

References Anderson GL, Judd HL, Kaunitz AM et al (2003) Effects of estrogen plus progestin on gynecologic cancers and associated diagnostic procedures: the Women’s Health Initiative randomized trial. JAMA 290:1739–1748. https://doi.org/10.1001/jama.290.13.1739 Antoni J, Folch E, Costa J et al (1997) Comparison of cytospat and pipelle endometrial biopsy instruments. Eur J Obstet Gynecol Reprod Biol 72:57–61 Archer DF, McIntyre-Seltman K, Wilborn WW Jr et al (1991) Endometrial morphology in asymptomatic postmenopausal women. Am J Obstet Gynecol 165:317–320; discussion 320–312 Arias-Stella J (2002) The Arias-Stella reaction: facts and fancies four decades after. Adv Anat Pathol 9:12–23 Arias-Stella J Jr, Arias-Velasquez A, Arias-Stella J (1994) Normal and abnormal mitoses in the atypical endometrial change associated with chorionic tissue effect [corrected]. Am J Surg Pathol 18:694–701

431 Bahceci M, Demirel LC (1996) Osseous metaplasia of the endometrium: a rare cause of infertility and its hysteroscopic management. Hum Reprod 11:2537–2539 Baird DB, Reddick RL (1991) Extraovarian mucinous metaplasia in a patient with bilateral mucinous borderline ovarian tumors: a case report. Int J Gynecol Pathol 10:96–103 Bakour SH, Khan KS, Gupta JK (2000) Controlled analysis of factors associated with insufficient sample on outpatient endometrial biopsy. BJOG 107:1312–1314 Banks ER, Mills SE, Frierson HF Jr (1991) Uterine intravascular menstrual endometrium simulating malignancy. Am J Surg Pathol 15:407–412 Barroeta JE, Pasha TL, Acs G et al (2007) Immunoprofile of endocervical and endometrial stromal cells and its potential application in localization of tumor involvement. Int J Gynecol Pathol 26:76–82. https://doi.org/ 10.1097/01.pgp.0000225852.74868.0c Bayer-Garner IB, Korourian S (2001) Plasma cells in chronic endometritis are easily identified when stained with syndecan-1. Mod Pathol 14:877–879. https://doi. org/10.1038/modpathol.3880405 Bayer-Garner IB, Nickell JA, Korourian S (2004) Routine syndecan-1 immunohistochemistry aids in the diagnosis of chronic endometritis. Arch Pathol Lab Med 128:1000–1003. https://doi.org/10.1043/1543-2165 (2004)1282.0.co;2 Bennett AE, Rathore S, Rhatigan RM (1999) Focal necrotizing endometritis: a clinicopathologic study of 15 cases. Int J Gynecol Pathol 18:220–225 Bozdogan O, Atasoy P, Erekul S et al (2002) Apoptosisrelated proteins and steroid hormone receptors in normal, hyperplastic, and neoplastic endometrium. Int J Gynecol Pathol 21:375–382 Brachtel EF, Sanchez-Estevez C, Moreno-Bueno G et al (2005) Distinct molecular alterations in complex endometrial hyperplasia (CEH) with and without immature squamous metaplasia (squamous morules). Am J Surg Pathol 29:1322–1329 Brinton LA, Hoover RN (1993) Estrogen replacement therapy and endometrial cancer risk: a meta-analysis. Obstet Gynecol 81:265–271 Bulmer JN, Hollings D, Ritson A (1987) Immunocytochemical evidence that endometrial stromal granulocytes are granulated lymphocytes. J Pathol 153:281–288. https:// doi.org/10.1002/path.1711530313 Bulmer JN, Lunny DP, Hagin SV (1988) Immunohistochemical characterization of stromal leucocytes in nonpregnant human endometrium. Am J Reprod Immunol Microbiol 17:83–90 Busca A, Parra-Herran C (2016) The role of pathologic evaluation of endometrial ablation resections in predicting ablation failure and adenomyosis in hysterectomy. Pathol Res Pract 212:778–782. https://doi.org/ 10.1016/j.prp.2016.06.007 Carlson JW, Mutter GL (2008) Endometrial intraepithelial neoplasia is associated with polyps and frequently has metaplastic change. Histopathology 53:325–332. https://doi.org/10.1111/j.1365-2559.2008.03104.x

432 Cheng WF, Lin HH, Torng PL et al (1997) Comparison of endometrial changes among symptomatic tamoxifentreated and nontreated premenopausal and postmenopausal breast cancer patients. Gynecol Oncol 66:233–237. https://doi.org/10.1006/gyno.1997.4739 Chiarelli S, Buritica C, Litta P et al (2006) An immunohistochemical study of morules in endometrioid lesions of the female genital tract: CD10 is a characteristic marker of morular metaplasia. Clin Cancer Res 12:4251–4256. https://doi.org/10.1158/1078-0432.ccr-06-0398 Chinen K, Kamiyama K, Kinjo T et al (2004) Morules in endometrial carcinoma and benign endometrial lesions differ from squamous differentiation tissue and are not infected with human papillomavirus. J Clin Pathol 57:918–926. https://doi.org/10.1136/ jcp.2004.017996 Clark TJ, Mann CH, Shah N et al (2001) Accuracy of outpatient endometrial biopsy in the diagnosis of endometrial hyperplasia. Acta Obstet Gynecol Scand 80:784–793 Clement PB (1988) Postoperative spindle-cell nodule of the endometrium. Arch Pathol Lab Med 112:566–568 Clement PB, Scully RE (1988) Idiopathic postmenopausal decidual reaction of the endometrium. A clinicopathologic analysis of four cases. Int J Gynecol Pathol 7:152–161 Clement PB, Scully RE (1990) Mullerian adenosarcoma of the uterus: a clinicopathologic analysis of 100 cases with a review of the literature. Hum Pathol 21:363–381 Clement PB, Young RH (1987) Atypical polypoid adenomyoma of the uterus associated with Turner’s syndrome. A report of three cases, including a review of “estrogen-associated” endometrial neoplasms and neoplasms associated with Turner’s syndrome. Int J Gynecol Pathol 6:104–113 Clement PB, Oliva E, Young RH (1996) Mullerian adenosarcoma of the uterine corpus associated with tamoxifen therapy: a report of six cases and a review of tamoxifen-associated endometrial lesions. Int J Gynecol Pathol 15:222–229 Cohen I, Figer A, Altaras MM et al (1996) Common endometrial decidual reaction in postmenopausal breast cancer patients treated with tamoxifen and progestogens. Int J Gynecol Pathol 15:17–22 Cohen I, Beyth Y, Shapira J et al (1997) High frequency of adenomyosis in postmenopausal breast cancer patients treated with tamoxifen. Gynecol Obstet Investig 44:200–205 Colgan TJ, Shah R, Leyland N (1999) Post-hysteroscopic ablation reaction: a histopathologic study of the effects of electrosurgical ablation. Int J Gynecol Pathol 18:325–331 Creagh TM, Bain BJ, Evans DJ et al (1995) Endometrial extramedullary haemopoiesis. J Pathol 176:99–104. https://doi.org/10.1002/path.1711760114 Critchley HO, Tong S, Cameron ST et al (1999) Regulation of bcl-2 gene family members in human endometrium by antiprogestin administration in vivo. J Reprod Fertil 115:389–395

R. R. Lastra et al. Crum CP, Hornstein MD, Nucci MR et al (2003) Hertig and beyond: a systematic and practical approach to the endometrial biopsy. Adv Anat Pathol 10:301–318 Dal Cin P, De Wolf F, Klerckx P et al (1992) The 6p21 chromosome region is nonrandomly involved in endometrial polyps. Gynecol Oncol 46:393–396 De Quintal MM, De Angelo Andrade LA (2006) Endometrial polyp with sex cord-like pattern. Int J Gynecol Pathol 25:170–172. https://doi.org/10.1097/01. pgp.0000185409.31427.07 Delmas PD, Bjarnason NH, Mitlak BH et al (1997) Effects of raloxifene on bone mineral density, serum cholesterol concentrations, and uterine endometrium in postmenopausal women. N Engl J Med 337:1641–1647. https://doi.org/10.1056/nejm199712043372301 Deshmukh-Rane SA, Wu ML (2009) Pseudolipomatosis affects specimens from endometrial biopsies. Am J Clin Pathol 132:374–377. https://doi.org/10.1309/ ajcpb5vlbir7skds Dhingra N, Punia RS, Radotra A et al (2007) Arias-Stella reaction in upper genital tract in pregnant and non-pregnant women: a study of 120 randomly selected cases. Arch Gynecol Obstet 276:47–52. https://doi.org/ 10.1007/s00404-006-0297-x Disep B, Innes BA, Cochrane HR et al (2004) Immunohistochemical characterization of endometrial leucocytes in endometritis. Histopathology 45:625–632. https:// doi.org/10.1111/j.1365-2559.2004.02052.x Duncan DA, Varner RE, Mazur MT (1989) Uterine herpes virus infection with multifocal necrotizing endometritis. Hum Pathol 20:1021–1024 Dutra FR (1959) Intraglandular morules of the endometrium. Am J Clin Pathol 31:60–65 Euscher E, Nuovo GJ (2002) Detection of kappa- and lambda-expressing cells in the endometrium by in situ hybridization. Int J Gynecol Pathol 21:383–390 Fechner RE, Bossart MI, Spjut HJ (1979) Ultrastructure of endometrial stromal foam cells. Am J Clin Pathol 72:628–633 Feeley KM, Wells M (2001) Hormone replacement therapy and the endometrium. J Clin Pathol 54:435–440 Ferryman SR, Stephens M, Gough D (1992) Necrotising granulomatous endometritis following endometrial ablation therapy. Br J Obstet Gynaecol 99:928–930 Fisher B, Costantino JP, Wickerham DL et al (1998) Tamoxifen for prevention of breast cancer: report of the National Surgical Adjuvant Breast and Bowel Project P-1 Study. J Natl Cancer Inst 90:1371–1388 Fletcher JA, Pinkus JL, Lage JM et al (1992) Clonal 6p21 rearrangement is restricted to the mesenchymal component of an endometrial polyp. Genes Chromosomes Cancer 5:260–263 Fluhmann CF (1960) The developmental anatomy of the cervix uteri. Obstet Gynecol 15:62–69 Frank TS, Himebaugh KS, Wilson MD (1992) Granulomatous endometritis associated with histologically occult cytomegalovirus in a healthy patient. Am J Surg Pathol 16:716–720

7

Benign Diseases of the Endometrium

Fritsch H, Hoermann R, Bitsche M et al (2013) Development of epithelial and mesenchymal regionalization of the human fetal utero-vaginal anlagen. J Anat 222:462–472. https://doi.org/10.1111/joa.12029 Fukunaga M, Iwaki S (2004) Nodular histiocytic hyperplasia of the endometrium. Arch Pathol Lab Med 128:1032–1034. https://doi.org/10.1043/1543-2165 (2004)1282.0.co;2 Genest DR, Brodsky G, Lage JA (1995) Localized endometrial proliferations associated with pregnancy: clinical and histopathologic features of 11 cases. Hum Pathol 26:1233–1240 Gilks CB, Clement PB, Hart WR et al (2000) Uterine adenomyomas excluding atypical polypoid adenomyomas and adenomyomas of endocervical type: a clinicopathologic study of 30 cases of an underemphasized lesion that may cause diagnostic problems with brief consideration of adenomyomas of other female genital tract sites. Int J Gynecol Pathol 19:195–205 Gompel A, Sabourin JC, Martin A et al (1994) Bcl-2 expression in normal endometrium during the menstrual cycle. Am J Pathol 144:1195–1202 Gordon SJ, Westgate J (1999) The incidence and management of failed Pipelle sampling in a general outpatient clinic. Aust N Z J Obstet Gynaecol 39:115–118 Grady D, Gebretsadik T, Kerlikowske K et al (1995) Hormone replacement therapy and endometrial cancer risk: a meta-analysis. Obstet Gynecol 85:304–313. https:// doi.org/10.1016/0029-7844(94)00383-o Gupta PK, Hollander DH, Frost JK (1976) Actinomycetes in cervico-vaginal smears: an association with IUD usage. Acta Cytol 20:295–297 Hachisuga T, Hideshima T, Kawarabayashi T et al (1999) Expression of steroid receptors, Ki-67, and epidermal growth factor receptor in tamoxifen-treated endometrium. Int J Gynecol Pathol 18:297–303 Heatley MK (2006) Atypical polypoid adenomyoma: a systematic review of the English literature. Histopathology 48:609–610. https://doi.org/10.1111/j.13652559.2005.02315.x Hejmadi RK, Chaudhri S, Ganesan R et al (2007) Morphologic changes in the endometrium associated with the use of the mirena coil: a retrospective study of 106 cases. Int J Surg Pathol 15:148–154. https://doi. org/10.1177/1066896906299120 Hendrickson MR, Kempson RL (1980) Endometrial epithelial metaplasias: proliferations frequently misdiagnosed as adenocarcinoma. Report of 89 cases and proposed classification. Am J Surg Pathol 4:525–542 Hendrickson M, Ross J, Eifel P et al (1982) Uterine papillary serous carcinoma: a highly malignant form of endometrial adenocarcinoma. Am J Surg Pathol 6:93–108 Herbold DR, Magrane DM (1986) Calcifications of the benign endometrium. Arch Pathol Lab Med 110:666–669 Houghton JP, Ioffe OB, Silverberg SG et al (2003) Metastatic breast lobular carcinoma involving tamoxifenassociated endometrial polyps: report of two cases

433 and review of tamoxifen-associated polypoid uterine lesions. Mod Pathol 16:395–398. https://doi.org/ 10.1097/01.mp.0000062655.62606.86 Houghton JP, Roddy S, Carroll S et al (2004) A simple method for the prevention of endometrial autolysis in hysterectomy specimens. J Clin Pathol 57:332–333 Houghton O, Connolly LE, McCluggage WG (2008) Morules in endometrioid proliferations of the uterus and ovary consistently express the intestinal transcription factor CDX2. Histopathology 53:156–165. https:// doi.org/10.1111/j.1365-2559.2008.03083.x Howitt BE, Quade BJ, Nucci MR (2015) Uterine polyps with features overlapping with those of Mullerian adenosarcoma: a clinicopathologic analysis of 29 cases emphasizing their likely benign nature. Am J Surg Pathol 39:116–126. https://doi.org/10.1097/ pas.0000000000000303 Huang KT, Chen CA, Cheng WF et al (1996) Sonographic characteristics of adenofibroma of the endometrium following tamoxifen therapy for breast cancer: two case reports. Ultrasound Obstet Gynecol 7:363–366. https://doi.org/10.1046/j.1469-0705.1996.07050363.x Hui P, Kelly M, O’Malley DM et al (2005) Minimal uterine serous carcinoma: a clinicopathological study of 40 cases. Mod Pathol 18:75–82. https://doi.org/ 10.1038/modpathol.3800271 Iezzoni JC, Mills SE (2001) Nonneoplastic endometrial signet-ring cells. Vacuolated decidual cells and stromal histiocytes mimicking adenocarcinoma. Am J Clin Pathol 115:249–255. https://doi.org/10.1309/rl5hvc5d-9we2-x71h Ip PP, Irving JA, McCluggage WG et al (2013) Papillary proliferation of the endometrium: a clinicopathologic study of 59 cases of simple and complex papillae without cytologic atypia. Am J Surg Pathol 37:167–177. https://doi.org/10.1097/PAS.0b013e318272d428 Irving JA, McFarland DF, Stuart DS et al (2000) Mitotic arrest of endometrial epithelium after paclitaxel therapy for breast cancer. Int J Gynecol Pathol 19:395–397 Ismail SM (1994) Pathology of endometrium treated with tamoxifen. J Clin Pathol 47:827–833 Ismail SM (1999) Gynaecological effects of tamoxifen. J Clin Pathol 52:83–88 Karaer A, Mert I, Akinsu F et al (2007) Ligneous inflammation involving the female genital tract. J Obstet Gynaecol Res 33:581–584. https://doi.org/10.1111/ j.1447-0756.2007.00579.x Kaufman RH, Binder GL, Gray PM Jr et al (1977) Upper genital tract changes associated with exposure in utero to diethylstilbestrol. Am J Obstet Gynecol 128:51–59 Kelly P, Dobbs SP, McCluggage WG (2007) Endometrial hyperplasia involving endometrial polyps: report of a series and discussion of the significance in an endometrial biopsy specimen. BJOG 114:944–950. https://doi. org/10.1111/j.1471-0528.2007.01391.x Kennedy MM, Baigrie CF, Manek S (1999) Tamoxifen and the endometrium: review of 102 cases and comparison with HRT-related and non-HRT-related endometrial pathology. Int J Gynecol Pathol 18:130–137

434 Khatamee MA, Sommers SC (1989) Clinicopathologic diagnosis of mycoplasma endometritis. Int J Fertil 34:52–55 Kim KR, Lee YH, Ro JY (2002) Nodular histiocytic hyperplasia of the endometrium. Int J Gynecol Pathol 21:141–146 Kurman RJ, Young RH, Norris HJ et al (1984) Immunocytochemical 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 3:101–121 Lehman MB, Hart WR (2001) 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 25:1347–1354 Leong ASY, Vinyuvat S, Leong FJWM et al (1997) AntiCD38 and VS38 antibodies for the detection of plasma cells in the diagnosis of chronic endometritis. Appl Immunohistochem 5:189–193 Lindenman E, Shepard MK, Pescovitz OH (1997) Mullerian agenesis: an update. Obstet Gynecol 90:307–312. https://doi.org/10.1016/s0029-7844(97)00256-1 Longacre TA, Chung MH, Rouse RV et al (1996) Atypical polypoid adenomyofibromas (atypical polypoid adenomyomas) of the uterus. A clinicopathologic study of 55 cases. Am J Surg Pathol 20:1–20 Machado F, Moreno J, Carazo M et al (2003) Accuracy of endometrial biopsy with the Cornier pipelle for diagnosis of endometrial cancer and atypical hyperplasia. Eur J Gynaecol Oncol 24:279–281 Marchini M, Fedele L, Bianchi S et al (1992) Endometrial patterns during therapy with danazol or gestrinone for endometriosis: structural and ultrastructural study. Hum Pathol 23:51–56 Marshall RJ, Jones DB (1988) An immunohistochemical study of lymphoid tissue in human endometrium. Int J Gynecol Pathol 7:225–235 Matias-Guiu X, Pons C, Munoz J et al (1994) Biotincontaining intranuclear inclusions in tumor cells: possible cause for misinterpreting nuclear antigen immunostaining. Am J Clin Pathol 102:706–707 Mazur MT (1981) Atypical polypoid adenomyomas of the endometrium. Am J Surg Pathol 5:473–482 Mazur MT, Hendrickson MR, Kempson RL (1983) Optically clear nuclei. An alteration of endometrial epithelium in the presence of trophoblast. Am J Surg Pathol 7:415–423 McCluggage WG (2002) Malignant biphasic uterine tumours: carcinosarcomas or metaplastic carcinomas? J Clin Pathol 55:321–325 McCluggage WG (2003) Metaplasias in the female genital tract. In: Lowe D, Underwood J (eds) Recent advances in histopathology, vol 20. Royal Society of Medicine Press Ltd, London

R. R. Lastra et al. McCluggage WG (2006) My approach to the interpretation of endometrial biopsies and curettings. J Clin Pathol 59:801–812. https://doi.org/10.1136/jcp.2005.029702 McCluggage WG, Roberts N, Bharucha H (1995) Enteric differentiation in endometrial adenocarcinomas: a mucin histochemical study. Int J Gynecol Pathol 14:250–254 McCluggage WG, Varma M, Weir P et al (1996) Uterine leiomyosarcoma in patient receiving tamoxifen therapy. Acta Obstet Gynecol Scand 75:593–595 McCluggage WG, Ashe P, McBride H et al (1998) Localization of the cellular expression of inhibin in trophoblastic tissue. Histopathology 32:252–256 McCluggage WG, Desai V, Manek S (2000a) Tamoxifenassociated postmenopausal adenomyosis exhibits stromal fibrosis, glandular dilatation and epithelial metaplasias. Histopathology 37:340–346 McCluggage WG, Hamal P, Traub AI et al (2000b) Uterine adenolipoleiomyoma: a rare hamartomatous lesion. Int J Gynecol Pathol 19:183–185 McCluggage WG, Abdulkader M, Price JH et al (2000c) Uterine carcinosarcomas in patients receiving tamoxifen. A report of 19 cases. Int J Gynecol Cancer 10:280–284 McCluggage WG, Oliva E, Herrington CS et al (2003a) CD10 and calretinin staining of endocervical glandular lesions, endocervical stroma and endometrioid adenocarcinomas of the uterine corpus: CD10 positivity is characteristic of, but not specific for, mesonephric lesions and is not specific for endometrial stroma. Histopathology 43:144–150 McCluggage WG, Sumathi VP, McManus DT (2003b) Uterine serous carcinoma and endometrial intraepithelial carcinoma arising in endometrial polyps: report of 5 cases, including 2 associated with tamoxifen therapy. Hum Pathol 34:939–943 McCulloch TA, Wagner B, Duffy S et al (1995) The pathology of hysterectomy specimens following transcervical resection of the endometrium. Histopathology 27:541–547 McKenna M, McCluggage WG (2008) Signet ring cells of stromal derivation in the uterine cervix secondary to cauterisation: report of a previously undescribed phenomenon. J Clin Pathol 61:648–651. https://doi.org/ 10.1136/jcp.2007.054767 Miko TL, Lampe LG, Thomazy VA et al (1988) Eosinophilic endomyometritis associated with diagnostic curettage. Int J Gynecol Pathol 7:162–172 Morsi HM, Leers MP, Jager W et al (2000) The patterns of expression of an apoptosis-related CK18 neoepitope, the bcl-2 proto-oncogene, and the Ki67 proliferation marker in normal, hyperplastic, and malignant endometrium. Int J Gynecol Pathol 19:118–126 Murray MJ, Meyer WR, Zaino RJ et al (2004) A critical analysis of the accuracy, reproducibility, and clinical utility of histologic endometrial dating in fertile women. Fertil Steril 81:1333–1343. https://doi.org/ 10.1016/j.fertnstert.2003.11.030

7

Benign Diseases of the Endometrium

Mutter GL, Zaino RJ, Baak JP et al (2007) Benign endometrial hyperplasia sequence and endometrial intraepithelial neoplasia. Int J Gynecol Pathol 26:103–114. https://doi. org/10.1097/PGP.0b013e31802e4696 Mutter GL, Bergeron C, Deligdisch L et al (2008) The spectrum of endometrial pathology induced by progesterone receptor modulators. Mod Pathol 21:591–598. https://doi.org/10.1038/modpathol.2008.19 Nakamura Y, Moritsuka Y, Ohta Y et al (1989) S-100 protein in glands within decidua and cervical glands during early pregnancy. Hum Pathol 20:1204–1209 Nand SL, Webster MA, Baber R et al (1998) Bleeding pattern and endometrial changes during continuous combined hormone replacement therapy. The Ogen/ Provera Study Group. Obstet Gynecol 91:678–684 Noyes RW, Hertig AT, Rock J (1950) Dating the endometrial biopsy. Fertil Steril 1:3–25 Nucci MR, Young RH (2004) Arias-Stella reaction of the endocervix: a report of 18 cases with emphasis on its varied histology and differential diagnosis. Am J Surg Pathol 28:608–612 Nucci MR, Prasad CJ, Crum CP et al (1999) Mucinous endometrial epithelial proliferations: a morphologic spectrum of changes with diverse clinical significance. Mod Pathol 12:1137–1142 O’Brien PK, Roth-Moyo LA, Davis BA (1981) Pseudosulfur granules associated with intrauterine contraceptive devices. Am J Clin Pathol 75:822–825 O’Connell JT, Mutter GL, Cviko A et al (2001) Identification of a basal/reserve cell immunophenotype in benign and neoplastic endometrium: a study with the p53 homologue p63. Gynecol Oncol 80:30–36. https://doi. org/10.1006/gyno.2000.6026 O’Connor DM, Kurman RJ (1988) Intermediate trophoblast in uterine curettings in the diagnosis of ectopic pregnancy. Obstet Gynecol 72:665–670 Ohishi Y, Kaku T, Kobayashi H et al (2008) CD10 immunostaining distinguishes atypical polypoid adenomyofibroma (atypical polypoid adenomyoma) from endometrial carcinoma invading the myometrium. Hum Pathol 39:1446–1453. https://doi. org/10.1016/j.humpath.2008.02.006 Ota S, Catasus L, Matias-Guiu X et al (2003) Molecular pathology of atypical polypoid adenomyoma of the uterus. Hum Pathol 34:784–788 Paavonen J, Aine R, Teisala K et al (1985) Chlamydial endometritis. J Clin Pathol 38:726–732 Paganini-Hill A, Ross RK, Henderson BE (1989) Endometrial cancer and patterns of use of oestrogen replacement therapy: a cohort study. Br J Cancer 59:445–447 Pearce KF, Nolan TE (1996) Endometrial sarcoidosis as a cause of postmenopausal bleeding. A case report. J Reprod Med 41:878–880 Perkins MB (1960) Pneumopolycystic endometritis. Am J Obstet Gynecol 80:332–336 Phillips V, McCluggage WG (2005) Results of a questionnaire regarding criteria for adequacy of endometrial

435 biopsies. J Clin Pathol 58:417–419. https://doi.org/ 10.1136/jcp.2004.023564 Phillips V, Graham CT, Manek S et al (2003) The effects of the levonorgestrel intrauterine system (Mirena coil) on endometrial morphology. J Clin Pathol 56:305–307 Pritt B, Mount SL, Cooper K et al (2006) Pseudoactinomycotic radiate granules of the gynaecological tract: review of a diagnostic pitfall. J Clin Pathol 59:17–20. https://doi.org/10.1136/jcp.2005.028977 Quddus MR, Sung CJ, Zheng W et al (1999) p53 immunoreactivity in endometrial metaplasia with dysfunctional uterine bleeding. Histopathology 35:44–49 Rutanen EM (1998) Insulin-like growth factors in endometrial function. Gynecol Endocrinol 12:399–406 Saegusa M, Okayasu I (2001) Frequent nuclear betacatenin accumulation and associated mutations in endometrioid-type endometrial and ovarian carcinomas with squamous differentiation. J Pathol 194:59–67. https://doi.org/10.1002/path.856 Sahin AA, Silva EG, Landon G et al (1989) Endometrial tissue in myometrial vessels not associated with menstruation. Int J Gynecol Pathol 8:139–146 Sakhdari A, Moghaddam PA, Liu Y (2016) Endometrial samples from postmenopausal women: a proposal for adequacy criteria. Int J Gynecol Pathol 35:525–530. https://doi.org/10.1097/pgp.0000000000000279 Schindler AE, Schmidt G (1980) Post-menopausal bleeding: a study of more than 1000 cases. Maturitas 2:269–274 Schlesinger C, Kamoi S, Ascher SM et al (1998) Endometrial polyps: a comparison study of patients receiving tamoxifen with two control groups. Int J Gynecol Pathol 17:302–311 Scott M, Lyness RW, McCluggage WG (2006) Atypical reactive proliferation of endocervix: a common lesion associated with endometrial carcinoma and likely related to prior endometrial sampling. Mod Pathol 19:470–474. https://doi.org/10.1038/modpathol.3800556 Scurry J, Planner R, Fortune DW et al (1993) Ligneous (pseudomembranous) inflammation of the female genital tract. A report of two cases. J Reprod Med 38:407–412 Seidman JD, Kurman RJ (1999) Tamoxifen and the endometrium. Int J Gynecol Pathol 18:293–296 Shafiee MN, Seedhouse C, Mongan N et al (2016) Up-regulation of genes involved in the insulin signalling pathway (IGF1, PTEN and IGFBP1) in the endometrium may link polycystic ovarian syndrome and endometrial cancer. Mol Cell Endocrinol 424:94–101. https://doi.org/10.1016/j.mce.2016.01.019 Shah SS, Mazur MT (2008) Endometrial eosinophilic syncytial change related to breakdown: immunohistochemical evidence suggests a regressive process. Int J Gynecol Pathol 27:534–538. https://doi.org/10.1097/ PGP.0b013e31817323b3 Shapiro S, Kelly JP, Rosenberg L et al (1985) Risk of localized and widespread endometrial cancer in relation

436 to recent and discontinued use of conjugated estrogens. N Engl J Med 313:969–972. https://doi.org/10.1056/ nejm198510173131601 Shibutani S, Ravindernath A, Suzuki N et al (2000) Identification of tamoxifen-DNA adducts in the endometrium of women treated with tamoxifen. Carcinogenesis 21:1461–1467 Shikone T, Kokawa K, Yamoto M et al (1997) Apoptosis of human ovary and uterine endometrium during the menstrual cycle. Horm Res 48(Suppl 3):27–34 Silva EG, Jenkins R (1990) Serous carcinoma in endometrial polyps. Mod Pathol 3:120–128 Soslow RA, Chung MH, Rouse RV et al (1996) Atypical polypoid adenomyofibroma (APA) versus welldifferentiated endometrial carcinoma with prominent stromal matrix: an immunohistochemical study. Int J Gynecol Pathol 15:209–216 Staland B (1981) Continuous treatment with natural oestrogens and progestogens. A method to avoid endometrial stimulation. Maturitas 3:145–156 Stastny JF, Ben-Ezra J, Stewart JA et al (1995) Condyloma and cervical intraepithelial neoplasia of the endometrium. Gynecol Obstet Investig 39:277–280 Stewart CJ, Michie BA, Kennedy JH (1998) Focal endometrial stromal hyperplasia. Histopathology 33:75–79 Stewart CJ, Campbell-Brown M, Critchley HO et al (1999) Endometrial apoptosis in patients with dysfunctional uterine bleeding. Histopathology 34:99–105 Sturdee DW, Ulrich LG, Barlow DH et al (2000) The endometrial response to sequential and continuous combined oestrogen-progestogen replacement therapy. BJOG 107:1392–1400 Tabibzadeh S, Kong Q, Satyaswaroop P et al (1994) Distinct regional and menstrual cycle dependent distribution of apoptosis in human endometrium. Potential regulatory role of T cells and TNF-alpha. Endocr J 2:87–95 Tahlan A, Nanda A, Mohan H (2006) Uterine adenomyoma: a clinicopathologic review of 26 cases and a review of the literature. Int J Gynecol Pathol 25:361–365. https:// doi.org/10.1097/01.pgp.0000209570.08716.b3 Tai LH, Tavassoli FA (2002) Endometrial polyps with atypical (bizarre) stromal cells. Am J Surg Pathol 26:505–509 Taketani Y, Mizuno M (1991) Evidence for direct regulation of epidermal growth factor receptors by steroid hormones in human endometrial cells. Hum Reprod 6:1365–1369 Takeuchi K, Tsujino T, Yabuta M et al (2012) A case of primary squamous cell carcinoma of the endometrium associated with extensive “ichthyosis uteri”. Eur J Gynaecol Oncol 33:552–554 Tantipalakorn C, Khunamornpong S, Lertprasertsuke N et al (2009) Female genital tract tumors and

R. R. Lastra et al. gastrointestinal lesions in the Peutz-Jeghers syndrome. J Med Assoc Thail 92:1686–1690 Tao XJ, Sayegh RA, Tilly JL et al (1998) Elevated expression of the proapoptotic BCL-2 family member, BAK, in the human endometrium coincident with apoptosis during the secretory phase of the cycle. Fertil Steril 70:338–343 The Writing Group for the PEPI Trial (1996) Effects of hormone replacement therapy on endometrial histology in postmenopausal women. The Postmenopausal Estrogen/Progestin Interventions (PEPI) Trial. JAMA 275:370–375 Thomas W Jr, Sadeghieh B, Fresco R et al (1978) Malacoplakia of the endometrium, a probable cause of postmenopausal bleeding. Am J Clin Pathol 69:637–641 Truskinovsky AM, Gerscovich EO, Duffield CR et al (2008) Endometrial microcalcifications detected by ultrasonography: clinical associations, histopathology, and potential etiology. Int J Gynecol Pathol 27:61–67. https://doi.org/10.1097/pgp.0b013e31812e95cb Turbiner J, Moreno-Bueno G, Dahiya S et al (2008) Clinicopathological and molecular analysis of endometrial carcinoma associated with tamoxifen. Mod Pathol 21:925–936. https://doi.org/10.1038/modpathol.2008.49 Usubutun A, Karaman N, Ayhan A et al (2005) Atypical endometrial stromal cells related with a polypoid leiomyoma with bizarre nuclei: a case report. Int J Gynecol Pathol 24:352–354 Val-Bernal JF, Villoria F, Cagigal ML et al (2006) Pneumopolycystic endometritis. Am J Surg Pathol 30:258–261 Valeri RM, Ibrahim N, Sheaff MT (2002) Extramedullary hematopoiesis in the endometrium. Int J Gynecol Pathol 21:178–181 Vang R, Tavassoli FA (2003) Proliferative mucinous lesions of the endometrium: analysis of existing criteria for diagnosing carcinoma in biopsies and curettings. Int J Surg Pathol 11:261–270 Venkatraman L, Elliott H, Steele EK et al (2003) Serous carcinoma arising in an adenofibroma of the endometrium. Int J Gynecol Pathol 22:194–197 Wani Y, Notohara K, Saegusa M et al (2008) Aberrant Cdx2 expression in endometrial lesions with squamous differentiation: important role of Cdx2 in squamous morula formation. Hum Pathol 39:1072–1079. https:// doi.org/10.1016/j.humpath.2007.07.019 Wells M, Tiltman A (1989) Intestinal metaplasia of the endometrium. Histopathology 15:431–433 Wheeler DT, Bristow RE, Kurman RJ (2007) Histologic alterations in endometrial hyperplasia and welldifferentiated carcinoma treated with progestins. Am J Surg Pathol 31:988–998. https://doi.org/10.1097/ PAS.0b013e31802d68ce

7

Benign Diseases of the Endometrium

Winkler B, Reumann W, Mitao M et al (1984) Chlamydial endometritis. A histological and immunohistochemical analysis. Am J Surg Pathol 8:771–778 Winter JS, Kohn G, Mellman WJ et al (1968) A familial syndrome of renal, genital, and middle ear anomalies. J Pediatr 72:88–93 Yokoyama S, Kashima K, Inoue S et al (1993) Biotincontaining intranuclear inclusions in endometrial glands during gestation and puerperium. Am J Clin Pathol 99:13–17

437 Young RH, Harris NL, Scully RE (1985) Lymphoma-like lesions of the lower female genital tract: a report of 16 cases. Int J Gynecol Pathol 4:289–299 Young RH, Treger T, Scully RE (1986) Atypical polypoid adenomyoma of the uterus. A report of 27 cases. Am J Clin Pathol 86:139–145 Zaloudek CJ, Norris HJ (1981) Adenofibroma and adenosarcoma of the uterus: a clinicopathologic study of 35 cases. Cancer 48:354–366

8

Precursors of Endometrial Carcinoma Lora Hedrick Ellenson, Brigitte M. Ronnett, and Robert J. Kurman

Contents Endometrial Hyperplasia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Definition and Classification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Clinical Features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Pathologic Findings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Differential Diagnosis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Behavior . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Morphologic Changes Associated with Progestin Treatment . . . . . . . . . . . . . . . . . . . . . . . . . .

440 440 440 441 445 453 455 458

Endometrial Cellular Changes: Metaplasia, Cellular Differentiation . . . . . . . . . . . . . Definitions and Classification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Clinical Features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Pathologic Findings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Differential Diagnosis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Behavior . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

458 459 460 460 465 466 466

Serous Endometrial Intraepithelial Carcinoma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 466 Definition and Pathologic Findings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 466

L. Hedrick Ellenson (*) Department of Pathology and Laboratory Medicine, Division of Gynecologic Pathology, Weill Cornell Medical College and New York Presbyterian Hospital, New York, NY, USA e-mail: [email protected] B. M. Ronnett Department of Pathology, Division of Gynecologic Pathology, Johns Hopkins University School of Medicine, Baltimore, MD, USA e-mail: [email protected] R. J. Kurman Department of Gynecology, Obstetrics, Pathology and Oncology, Division of Gynecologic Pathology, Johns Hopkins University School of Medicine, Baltimore, MD, USA e-mail: [email protected] # Springer Nature Switzerland AG 2019 R. J. Kurman et al. (eds.), Blaustein’s Pathology of the Female Genital Tract, https://doi.org/10.1007/978-3-319-46334-6_8

439

440

L. Hedrick Ellenson et al. Differential Diagnosis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 468 Behavior and Treatment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 469 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 470

Endometrial hyperplasia (EH) often precedes the development of endometrioid carcinoma, the most common type of endometrial carcinoma. Obesity, anovulatory cycles, and exogenous hormones are associated with both endometrioid carcinoma and hyperplasia. In addition, the risk of EH is associated with increasing body mass index (BMI) and nulliparity (Epplein et al. 2008; Wise et al. 2016; Guraslan et al. 2016). All of these factors are thought to result in unopposed estrogen stimulation of the endometrium. The role of unopposed estrogen stimulation in the development of EH and carcinoma is further supported by studies demonstrating elevated serum estrogen levels in patients with endometrioid carcinoma (Brinton et al. 1992; Potischman et al. 1996). Atypical hyperplasia/endometrioid intraepithelial neoplasia (EIN) is considered the direct precursor to endometrioid carcinoma. However, other histologic types of endometrial carcinoma are less commonly associated with estrogenic stimulation (Sherman et al. 1997). Serous carcinoma is the prototypic endometrial carcinoma that is typically not related to estrogenic stimulation or hyperplasia. It usually arises in atrophic endometrium through a precursor lesion called serous endometrial intraepithelial carcinoma (SEIC). The following discussion summarizes current knowledge about these precursor lesions including their differential diagnosis, treatment, and relationship to endometrial carcinoma.

Endometrial Hyperplasia Definition and Classification EH is defined as a proliferation of glands of irregular size and shape with an associated increase in the gland/stroma ratio compared with proliferative endometrium. Although the process is often diffuse, it may also be focal.

The classification and terminology of EH has undergone several iterations over the past several decades. The modern-day classification was introduced in 1994 by the World Health Organization (WHO) and the International Society of Gynecologic Pathologists (ISGYP) and quickly gained widespread acceptance. The classification subdivided hyperplasia into four groups: simple hyperplasia, complex hyperplasia, simple atypical hyperplasia, and complex atypical hyperplasia. The 2003 WHO classification endorsed the same classification, but in 2014 the WHO made significant changes by simplifying the four-tier to a two-tier classification of hyperplasia without atypia and hyperplasia with atypia (atypical hyperplasia). In addition, a completely new term designated “endometrioid intraepithelial neoplasia (EIN)” was introduced and considered synonymous with atypical hyperplasia (Kurman et al. 2014). The pros and cons of both systems are discussed later in the chapter. Currently, both are being utilized in routine practice.

Clinical Features Patients with EH typically have abnormal bleeding. Occasionally, the lesion is detected fortuitously by endometrial biopsy performed during the course of an infertility workup or before the start of hormone replacement therapy in postmenopausal women. Hyperplasia develops because of unopposed estrogenic stimulation, and consequently most patients with hyperplasia have a history of either persistent anovulation or exogenous unopposed estrogen usage. Although anovulation occurs at menarche and in perimenopausal women, hyperplasia is not usually encountered in young women. Although hyperplasia has been reported in a patient at 16 years of age (Lee and Scully 1989). This may be because bleeding in menarchial women is seldom evaluated by an endometrial biopsy. During the reproductive

8

Precursors of Endometrial Carcinoma

years, hyperplasia is relatively uncommon, typically occurring in women with polycystic ovary syndrome (Stein–Leventhal syndrome). In the original description of this syndrome, these women were reported to be anovulatory, obese, infertile, and exhibit hirsutism, but many women with this disorder lack these features. Conversely, women who are obese but who do not have polycystic ovarian disease may have hyperplasia, presumably because of peripheral conversion of androstenedione to estrogen in adipose tissue. Diabetes mellitus and hypertension may occur in women with hyperplasia, but often these disorders are not present. Although hyperplasia or carcinoma should always be suspected in a postmenopausal woman with abnormal uterine bleeding, atrophy is the most common cause of bleeding in this age group. In one study of postmenopausal women with bleeding, 7% had endometrial cancer, 15% had various types of hyperplasia, and 56% had atrophy (Lidor et al. 1986). Typically, women with hyperplasia or carcinoma have moderate or heavy vaginal bleeding compared with women with atrophic endometria who present with spotting.

441

degrees of architectural abnormalities. In simple hyperplasia, the glands are cystically dilated, with occasional outpouchings surrounded by abundant cellular stroma (Fig. 1). In other instances, the glands are only minimally dilated but focally crowded (Fig. 2). Admixtures of the various patterns frequently occur (Fig. 3). The cells lining the glands are stratified and columnar with amphophilic cytoplasm. Mitotic activity is variable. With increasing degrees of architectural abnormality, glands become complex and branched with irregular outlines and papillary infoldings into the lumens. In addition, with increased proliferation glands become crowded,

Pathologic Findings Hyperplastic endometrium is not distinctive grossly. In hysterectomy specimens, hyperplasia may have a velvety, knobby surface of pale, spongy tissue with vague borders. Although diffuse thickening of the endometrium is common, hyperplasia can be focal and may simulate a polyp. The volume of tissue obtained in curettings is usually increased, but it may be quite variable and less than that obtained during the secretory phase of the normal cycle. A diagnosis of hyperplasia, therefore, depends on the histologic pattern and not on the volume of tissue.

Hyperplasia Without Atypia Hyperplasia is characterized by an increased gland/stroma ratio and a variety of abnormal architectural patterns. Glands typically vary in size and shape. Dilatation and outpouching of glandular epithelium characterize the lesser

Fig. 1 Hyperplasia without atypia (simple). Glands are only minimally crowded but are dilated and have outpouchings

Fig. 2 Hyperplasia without atypia (simple). Glands are mildly crowded and some are cystically dilated

442

L. Hedrick Ellenson et al.

Fig. 3 Hyperplasia without atypia (simple). Glands are mildly crowded and dilated, with some exhibiting outpouchings and simple branching

Fig. 5 Hyperplasia without atypia (complex). Dilated glands resemble those seen in simple hyperplasia but are sufficiently crowded for classification as complex hyperplasia

Fig. 4 Hyperplasia without atypia (complex). Glands are sufficiently crowded for classification as complex hyperplasia, despite the presence of some glands having only simple tubular profiles

Fig. 6 Hyperplasia without atypia. Nuclei are elongated, oriented perpendicular to the basement membrane and have even chromatin

compressing the intervening stroma, resulting in “back-to-back” glandular crowding. Thus, complex hyperplasia is composed of crowded glands with little intervening stroma (Kurman et al. 1985) (Figs. 4 and 5). Usually the glandular outlines are highly complex but at times are tubular, with or without dilatation (Figs. 4 and 5). Epithelial stratification can range from two to four layers, but some glands may exhibit little or no stratification. Mitotic activity is variable and is usually less than five mitotic figures per 10 high power fields. Even in highly complex hyperplasia with marked stratification, mitotic figures may be inconspicuous. In hyperplasia lacking atypia, the epithelial cells

contain oval, basally oriented bland nuclei with smooth, uniform contours resembling those in normal proliferative glands (Figs. 6, 7, 8, and 9). In simple hyperplasia, the stromal cells are more densely packed than in proliferative endometrium. The cells retain their spindle shape but are plump, with enlarged nuclei and indistinct cytoplasm. Mitotic activity in endometrial stromal cells is variable but may be increased. In complex hyperplasia, the stromal cells are spindle shaped and compressed by the glandular proliferation. In addition to densely packed stromal cells, clusters of foamy, lipid-laden cells may be present in the stroma of hyperplasia, atypical hyperplasia, and well-differentiated adenocarcinoma (Dawagne

8

Precursors of Endometrial Carcinoma

Fig. 7 Hyperplasia without atypia. Nuclei are elongated, oriented perpendicular to the basement membrane and have even chromatin

Fig. 8 Hyperplasia without atypia. Nuclei are elongated, oriented perpendicular to the basement membrane and are evenly hyperchromatic

Fig. 9 Hyperplasia without atypia. Nuclei are elongated, oriented perpendicular to the basement membrane and are evenly hyperchromatic

443

and Silverberg 1982; Silver and Sherman 1998). Foam cells have small pyknotic nuclei and cytoplasm that contain lipid droplets but no mucin. The foam cells have been shown to be histiocytes by immunohistochemistry (Silver and Sherman 1998). These histiocytic cells may also be observed in atrophic and nonneoplastic endometria. The isolated finding of histiocytes in cervicovaginal smears of asymptomatic postmenopausal women has not been associated with an increased likelihood of detecting EH or carcinoma (Hall et al. 1982). The presence of histiocytes alone in cervicovaginal smears from postmenopausal women with abnormal uterine bleeding also has not been shown to predict the presence of either endometrial hyperplasia or carcinoma. However, the finding of histiocytes containing phagocytosed acute inflammatory cells or normal endometrial cells in postmenopausal women with abnormal uterine bleeding has been associated with a threeto four-fold greater likelihood of coexistent endometrial carcinoma or hyperplasia (Nguyen et al. 1998).

Atypical Hyperplasia The most important feature in the evaluation of EH is the presence or absence of nuclear atypia. Cells with nuclear atypia are stratified and show loss of polarity and an increase in the nuclear/cytoplasmic ratio (Figs. 10, 11, 12, 13, 14, and 15). The nuclei are enlarged, irregular in size and shape, with coarse

Fig. 10 Atypical hyperplasia. Branching and tubular glands are crowded with very little intervening stroma

444

Fig. 11 Atypical hyperplasia. Nuclei are rounded and have vesicular chromatin

L. Hedrick Ellenson et al.

Fig. 14 Atypical hyperplasia. Nuclei vary from ovoid to rounded, have even chromatin with evident nucleoli and mitotic figures, and display some stratification and loss of polarity

Fig. 12 Atypical hyperplasia. Nuclei are rounded, have vesicular chromatin, and display stratification and loss of polarity

Fig. 15 Atypical hyperplasia. Nuclei are enlarged and rounded, have vesicular chromatin with fine granularity and evident nucleoli, and display stratification and loss of polarity

Fig. 13 Atypical hyperplasia. Nuclei are enlarged and rounded, have vesicular chromatin with fine granularity, and display stratification and loss of polarity

chromatin clumping, a thickened irregular nuclear membrane, and prominent nucleoli. Nuclei tend to be round as compared with the oval nuclei of proliferative endometrium and hyperplasia without atypia. As a result, the nuclei often have a cleared or vesicular appearance with condensation of the chromatin around the nuclear membrane. Nuclear atypia is variable, both qualitatively and quantitatively. Not all glands contain atypical cells, and in an individual gland some cells are atypical, and others are not. Rare atypical cells should be ignored, but if cellular atypia is evident without a diligent search, the diagnosis of atypical hyperplasia should be made. One of the main

8

Precursors of Endometrial Carcinoma

issues facing diagnostic surgical pathologists is the intra- and interobserver variability in the identification of atypia. Thus, grading atypia as mild, moderate, or severe is not recommended, as it is subjective and not reproducible. Furthermore, assessment of atypia is often problematic in the setting of metaplasia (see below). The nuclear enlargement, rounding, and vesicular change seen in tubal metaplasia, for example, can suggest cytologic atypia, but this metaplastic change is not interpreted as true atypia as it has not been shown to affect clinical outcome (Hendrickson and Kempson 1980). Moreover, metaplasia and atypia can occur together in the same glands, so if the nuclear changes are present in cells that are not metaplastic a diagnosis of atypical hyperplasia can be made even in the setting of metaplasia. (Fig. 16). Nuclear enlargement, pleomorphism, coarsening of chromatin, and presence of nucleoli are important features that constitute the morphologic features of cytologic atypia, but these changes can be subtle and as a result somewhat subjective. This problem is inherent in which ever classification system is used, atypical hyperplasia or EIN (see below). One of the most useful techniques in determining whether atypia is present is to compare the nuclear features of the glands in

Fig. 16 Hyperplasia with tubal metaplasia and atypia. Nuclei vary from ovoid to rounded, have finely granular chromatin, and display some stratification and loss of polarity. Ciliated cells are present, indicating tubal metaplasia. Some nuclear changes are related to the tubal metaplasia, but the loss of polarity and stratification are beyond those attributable to metaplasia (compare with Figs. 35 and 36)

445

question to adjacent normal appearing proliferative endometrium. If the nuclei in the glands in question appear more atypical than those in the proliferative glands then the lesion qualifies as atypical hyperplasia. If the nuclear features are similar to the adjacent proliferative glands, no matter how crowded the proliferative process is, the lesion qualifies as hyperplasia without atypia. If normal proliferative endometrium is not present in the tissue sample, the descriptive features described above must be employed to determine the presence of atypia. The architectural features of complex atypical hyperplasia are similar to its counterpart without cytologic atypia. In complex atypical hyperplasia, the glands almost invariably demonstrate marked structural complexity with irregular outlines and back-to-back crowding (Fig. 10). Epithelial stratification and mitotic activity are variable. Papillary infoldings also are seen. Some atypical hyperplasias have little stratification, and mitotic activity may be inconspicuous.

Differential Diagnosis Hyperplasia should be distinguished from disordered proliferative phase, polyps, ciliated cell change (tubal metaplasia), cystic atrophy, and endometrial glandular and stromal breakdown. Disordered proliferative phase is similar qualitatively to hyperplasia without atypia but is a focal lesion characterized by irregularly shaped and enlarged glands that are interspersed among normal proliferative glands (Fig. 17). The latter may be focally crowded. The key feature that distinguishes disordered proliferative phase from hyperplasia without atypia is the focal nature of the glandular abnormality in disordered proliferative phase. The fragments of endometrium containing the disordered glands should not have the appearance of a polyp. Hyperplastic endometrial polyps often contain areas of hyperplasia without atypia that are confined to one or just a few fragments of polypoid tissue. The polyp usually stands out as a large rounded tissue fragment in sharp contrast to the remainder of the uninvolved endometrium. Polyps typically have

446

Fig. 17 Disordered proliferative endometrium. Cystically dilated glands with outpouchings are admixed with small tubular proliferative glands

dense fibrous stroma and contain clusters of thickwalled blood vessels near the center of the fragment. The fragments often are covered on three sides by surface endometrium (see ▶ Chap. 7, “Benign Diseases of the Endometrium”). Endometrial glands with ciliated cell change are often found in association with simple and complex hyperplasia. When found with hyperplasia, the presence of ciliated cell change does not need to be specified. Ciliated glands are usually slightly dilated. When a few isolated glands show ciliated cell change in the absence of hyperplasia, a diagnosis of ciliated cell change (tubal metaplasia) is justified (see following). Ciliated cell change often contains scattered vesicular nuclei with occasional nucleoli that should not be mistaken as cytologic atypia when found in association with hyperplasia. Distinction of cystic atrophy from hyperplasia is seldom a problem in curettings because atrophic glands collapse as a result of the procedure. In hysterectomy specimens, glands are dilated and lined by a single layer of cells that are often flattened. Mitotic activity is not present. In contrast, in hyperplasia there is pseudo-stratification of columnar epithelial cells. Mitotic activity is variable but present. In endometrial glandular and stromal breakdown caused by estrogen withdrawal, proliferative-type glands appear back-to-back because of loss of intervening endometrial stroma. Glands are often

L. Hedrick Ellenson et al.

Fig. 18 Artifactual glandular crowding. Apposition of surface epithelium of fragments of proliferative endometrium simulates the cystically dilated glands of hyperplasia without atypia but is an artifact

fragmented, and apoptotic bodies are present. Clusters of stromal cells and fragmented glands surrounded by blood are consistent features (see ▶ Chap. 7, “Benign Diseases of the Endometrium”). In contrast, in hyperplasia the glandular outlines are more irregular and complex than the tubular, proliferative-type glands in breakdown. Furthermore, glandular fragmentation, apoptotic bodies, and rounded clusters of stromal cells, so-called “stromal blue balls,” are usually absent in hyperplasia. In addition, the tissue fragmentation and artifactual apposition of segments of surface epithelium encountered in biopsy and curettage specimens can complicate interpretation and should be considered prior to rendering a diagnosis of hyperplasia (Fig. 18). Atypical hyperplasia must be distinguished from an atypical polypoid adenomyoma and from well-differentiated adenocarcinoma. The atypical polypoid adenomyoma is composed of glands that show variable architectural complexity and some cytologic atypia (Fig. 19). Squamous differentiation in the form of squamous morules is an almost constant feature of the atypical polypoid adenomyoma. Characteristically, the glands in the atypical polypoid adenomyoma are surrounded by smooth muscle, in contrast with the dense proliferative stroma found in hyperplasia and the altered or desmoplastic stroma found in association with well-differentiated carcinoma.

8

Precursors of Endometrial Carcinoma

Fig. 19 Atypical polypoid adenomyoma. Crowded complex glands with squamous morules are surrounded by a fibromuscular stroma

Most endometrial carcinomas are readily identified. However, several studies highlight the difficulty in distinguishing some cases of welldifferentiated carcinoma from atypical hyperplasia on preoperative endometrial sampling by either biopsy or curettage. Specific histologic features can often be used to separate hyperplasia from well-differentiated carcinoma that reduce the subjectivity of the evaluation. In the presence of invasion, the endometrial stroma interacts directly with malignant cells, and the morphologic changes it undergoes can serve as a means of identifying carcinoma. The stromal and epithelial alterations associated with invasive carcinoma are referred to collectively as endometrial stromal invasion. There are three useful criteria, any of which identifies stromal invasion: (1) an irregular infiltration of glands associated with an altered fibroblastic stroma (desmoplastic response); (2) a confluent glandular pattern in which individual glands, uninterrupted by stroma, merge at times creating a cribriform pattern; and (3) an extensive papillary pattern. It should be noted that on occasion hyperplasia can display a papillary architecture, including the presence of fibrovascular cores, but in contrast to the papillary pattern of carcinoma, hyperplasia is characterized by bland cytology, absence of epithelial stratification, and a low level of mitotic activity (Lehman and Hart 2001), It has been reported that a process manifesting the features of invasion must be

447

Fig. 20 Atypical hyperplasia with foci of well differentiated endometrioid carcinoma. Small aggregates of fused glands consistent with early endometrioid carcinoma are architecturally distinct from the individual larger glands of associated atypical hyperplasia

Fig. 21 Atypical hyperplasia with foci of well differentiated endometrioid carcinoma. Small aggregates of glands demonstrate gland fusion and early cribriform growth

sufficiently extensive to involve half (2.1 mm) of a low-power field 4.2 mm in diameter to have value in predicting the presence of a biologically significant carcinoma in the uterus (Kurman and Norris 1982; Norris et al. 1983). This criterion, however, should not be applied too rigidly in view of the potential of missing a carcinoma in small samples. If unequivocal evidence of stromal invasion is present in an area measuring less than one-half a low-power field, a diagnosis of welldifferentiated carcinoma should be made (Figs. 20 and 21). The quantification criterion does not apply to moderately or poorly differentiated carcinomas.

448

Fig. 22 Well-differentiated endometrioid carcinoma. Crowded atypical glands with early glandular confluence are surrounded by eosinophilic spindled stromal cells which constitute a desmoplastic stromal reaction, indicating stromal invasion by carcinoma

Fig. 23 Well-differentiated endometrioid carcinoma. Stromal desmoplasia, a manifestation of endometrial stromal invasion by endometrioid carcinoma, is characterized by spindled cells with a fibroblastic appearance

The three criteria for the identification of stromal invasion are described in greater detail below. 1. The altered stroma that reflects invasion contains parallel, densely arranged fibroblasts with more fibrosis than normal endometrial stroma and disrupts the usual glandular pattern (Fig. 22). The stromal cells are more spindle shaped than are the stromal cells of proliferative endometrium, with more elongated nuclei. Collagen compresses the stromal cells so that they have an eosinophilic and wavy appearance (Fig. 23), compared with the basophilic naked-nucleus

L. Hedrick Ellenson et al.

Fig. 24 Well-differentiated endometrioid carcinoma. Glands are fused and interconnected in a confluent glandular pattern, without intervening stroma, indicating endometrial stromal invasion by carcinoma

appearance of stromal cells found in proliferative endometrium and hyperplasia. In specimens containing fragments of polyps with fibrous stroma, or specimens from the lower uterine segment, these features cannot be applied. The distinction of hyperplasia from carcinoma in these cases is based on the identification of a confluent pattern (see below). Atypical polypoid adenomyomas (see ▶ Chap. 7, “Benign Diseases of the Endometrium”) contain smooth muscle and may simulate myometrial invasion (see Fig. 19) (Mazur 1981). In contrast with the atypical polypoid adenomyoma, smooth muscle is rarely identified in curettings of welldifferentiated carcinoma even when there is deep myometrial invasion because only the exophytic portion of the tumor is removed in biopsies and curettings. 2. Confluent glandular aggregates without intervening stroma reflect stromal invasion (Figs. 24 and 25). Confluent patterns are characterized by glandular configurations in which individual glands are not surrounded by stroma. Instead, glands appear to merge into one another to form a complex labyrinth. Some proliferations are cribriform, resulting from proliferation and bridging of epithelium (Fig. 26). 3. Complex papillary patterns represent stromal invasion if multiple, branching, fibrous processes lined by epithelium are present

8

Precursors of Endometrial Carcinoma

Fig. 25 Well-differentiated endometrioid carcinoma. Complex branching glands create a confluent glandular/ labyrinthine pattern indicating endometrial stromal invasion by carcinoma

Fig. 26 Well-differentiated endometrioid carcinoma. Fused glands create a cribriform pattern indicating endometrial stromal invasion by carcinoma

(Figs. 27 and 28). At times, these may create a villoglandular pattern. Intraglandular epithelial papillations lacking a fibrovascular core do not qualify as a feature of invasion; such proliferation is often encountered in eosinophilic metaplasia within complex hyperplasia (see below). Delicate papillary structures (with or without fibrovascular cores), often accompanied by metaplastic changes (mucinous, eosinophilic cell) and often occurring in polyps, are a feature of simple and complex papillary hyperplasias (Figs. 29 and 30) (Lehman and Hart 2001). In these lesions, the cytology is bland and the epithelium overlying the fibrovascular cores is

449

Fig. 27 Well-differentiated endometrioid carcinoma. A papillary/villoglandular proliferation indicates carcinoma when the papillary/villous structures have an exophytic growth pattern and are not merely intraglandular epithelial papillae

Fig. 28 Well-differentiated endometrioid carcinoma. A papillary/villoglandular pattern indicates carcinoma when the papillary/villous structures are lined by endometrioid epithelium along the external aspect and have a central fibrovascular core

not stratified. The mitotic index and Ki-67 proliferation index is very low. In the past, the presence of masses of squamous epithelium replacing the endometrial stroma was considered a feature of invasion (Kurman and Norris 1982). Masses of squamous epithelium with minimal nuclear atypia that extensively replace the endometrium (over a 2-mm2 area) reflect stromal invasion only if they are associated with a desmoplastic response or a confluent glandular pattern.

450

L. Hedrick Ellenson et al. Table 1 Hysterectomy findings when atypical hyperplasiaa is present in curettings (89 patients) Finding Carcinoma Grade Well differentiated Moderately differentiated Poorly differentiated Myometrial invasion None Inner one third 1 mm or less 2–4 mm

Fig. 29 Papillary hyperplasia. Intraglandular papillae reflect a papillary variant of hyperplasia rather than carcinoma (compare with Fig. 28)

No. (%) 15 (17) 15 (100) 0 0 8 (53)b 7 (47)b 5 2

Adapted with permission from Kurman and Norris (1982) a A diagnosis of atypical hyperplasia based on cytological atypia in the absence of endometrial stromal invasion b The percentages refer to the proportion of carcinomas in the hysterectomy specimen Table 2 Hysterectomy findings when well-differentiated carcinomaa is present in curettings (115 patients) Finding Carcinoma Grade Well differentiated Moderately differentiated Poorly differentiated Myometrial invasion Inner one third Middle and outer third

Fig. 30 Papillary hyperplasia. Papillae contain fibrovascular cores; their intraglandular location indicates a hyperplasia rather than a papillary pattern of carcinoma

Increasing degrees of nuclear atypia, mitotic activity, and stratification of cells in curettings are associated with a higher frequency of carcinoma in the uterus but are of limited value because even a mild degree of these changes is associated with carcinoma in nearly one-third of cases (Lehman and Hart 2001). Even with mild atypia, low mitotic activity, and lesser degrees of stratification in curettings, 20% of residual carcinomas in the resected uterus are moderately or poorly differentiated, and 10% deeply invade the myometrium. These other features in curettings, although useful, therefore are not sufficiently accurate to predict whether a biologically significant lesion is present in the uterus. Unfortunately, assessing varying

No. (%) 58 (50) 38 (66)b 14 (24)b 6 (10)b 42 (72)b 28 (48)b 14 (24)b

Adapted with permission from Kurman and Norris (1982) a A diagnosis of well-differentiated carcinoma based on identification of endometrial stromal invasion b The percentages refer to the proportion of carcinomas in the hysterectomy specimen

degrees of nuclear atypia in this borderline group of lesions is subjective and not easily reproduced. In contrast, when stromal invasion is absent in curettings, carcinoma is found in the uterus in only 17% of cases, and the carcinomas are well differentiated and either confined to the endometrium or only superficially invasive (Table 1). If stromal invasion is present in curettings, residual carcinoma is found in the uterus in half; more than one-third of the carcinomas are moderately or poorly differentiated, and a fourth of them invade deeply into the myometrium (Table 2). A small proportion (7%) of patients with invasion in curettings will have extrauterine metastases at

8

Precursors of Endometrial Carcinoma

hysterectomy, and half with metastasis will die of tumor (Kurman and Norris 1982). Thus, the absence of stromal invasion provides the basis for distinguishing atypical hyperplasia from a biologically significant, well-differentiated carcinoma (Kurman and Norris 1982; King et al. 1984). A number of more recent studies have found higher frequencies of endometrial carcinoma (43–52%) in hysterectomy specimens following a diagnosis of atypical hyperplasia (Janicek and Rosenshein 1994; Widra et al. 1995; Trimble et al. 2006). Of the carcinomas detected in two of these studies, 43% were stage 1C or greater. In the most recent study, only 10.6% were stage 1C while 30.9% were stage 1B and the remaining were stage 1A. These studies included patients who had been diagnosed by either curettage or biopsy, but there were no significant differences in the frequencies with which carcinoma was detected at hysterectomy in those patients who received a curettage compared to those who had been biopsied. However, in one of these studies, the biopsy and curettage specimens were not reviewed to confirm that features of stromal invasion were absent in these specimens (Janicek and Rosenshein 1994). More recently, additional studies have demonstrated that clinically significant endometrial proliferations, that is, those that have a high likelihood of myometrial invasion, can be recognized when either sufficient architectural complexity or nuclear atypia, including prominence of nucleoli, is present (Longacre et al. 1995; McKenney and Longacre 2009). In addition, the strong association of a desmoplastic stromal response with a myoinvasive lesion was confirmed. The identification of stromal invasion is important because it is semiquantitative, therefore less subjective than other criteria, and it delineates a biologically significant lesion that has a much greater likelihood of metastasis than one in which invasion is absent. Experimental studies of neoplasms from the breast, colon, pancreas, and lung lend support to the division of endometrial proliferations into noninvasive and invasive forms based on the histologic alterations observed in the endometrial stroma. These studies demonstrate profound molecular and structural alterations in the

451

stroma adjacent to invasive as compared with noninvasive tumors. Invasive tumors can induce a conversion of stromal fibroblasts into myofibroblasts, which elaborate extracellular matrix components, such as type V collagen and proteoglycans, that are increased in desmoplasia and are readily observed by light microscopy using the criteria for stromal invasion as outlined. It has been shown that tumor cells produce growth factors such as plateletderived growth factor, epidermal-derived growth factor, and insulin-like growth factor, which may play a role in stimulating the growth of stromal cells surrounding tumors.

Reproducibility Studies and Adjunctive Techniques in the Classification of Endometrial Hyperplasias Several studies have addressed the reproducibility of the diagnosis of EH and its distinction from well-differentiated carcinoma (Kendall et al. 1998; Bergeron et al. 1999; Zaino et al. 2006). One study of 100 endometrial biopsy and curettage specimens ranging from proliferative endometrium to well-differentiated carcinoma found substantial interobserver agreement for diagnoses of hyperplasia without atypia and welldifferentiated carcinoma but only moderate agreement for the diagnosis of atypical hyperplasia (Kendall et al. 1998). Several histologic features, including nuclear enlargement, vesicular change, nuclear pleomorphism, chromatin irregularities, loss of polarity, nuclear rounding, and presence of nucleoli, were associated with a diagnosis of atypical hyperplasia by univariable logistic regression analysis. However, of the histologic features evaluated, the only feature that was associated with the distinction of atypical hyperplasia from hyperplasia without atypia in multivariable logistic regression analysis was the presence of nucleoli. The features that were associated with the distinction of carcinoma from atypical hyperplasia in both univariable and multivariable analysis included stromal alteration (stromal desmoplasia) and glandular confluence. A more recent study found similar values for intraobserver agreement and slightly lower values for interobserver agreement (Bergeron et al. 1999). In addition, the study

452

confirmed that the category of atypical hyperplasia has the lowest diagnostic reproducibility of the various categories. Similar histologic features were found to be useful for distinguishing the various diagnostic categories, although the utility of the presence of nucleoli for diagnosing atypical hyperplasia and of stromal alteration for diagnosing well-differentiated carcinoma were somewhat less, as evidenced by lower mean interobserver agreement values for these features. Thus, interobserver agreement is lowest for the diagnostic category of atypical hyperplasia, indicating that further refinement of the histologic criteria, enhanced endometrial sampling techniques, and novel objective analyses are required to improve the reproducibility of the diagnosis of atypical hyperplasia.

EIN In 2000, a new classification for EH was proposed based on histopathologic, molecular genetic changes and computerized morphometric analysis (Mutter et al. 2000; Mutter 2000). The system divides endometrial proliferative lesions into “EH,” a benign condition, and “EIN,” a true carcinoma precursor based on glandular architecture and genetic abnormalities that were interpreted as clonal and neoplastic. Proliferations that are polyclonal were regarded as a response to an abnormal hormonal environment – either unopposed estrogenic stimulation associated with anovulatory cycles or exogenous estrogenic stimulation – and were designated “hyperplasia.” Monoclonal lesions, on the other hand, were associated with an increased risk of progression to carcinoma and were originally designated “endometrial intraepithelial neoplasia” and subsequently changed to “endometrioid intraepithelial neoplasia (EIN)” in the 2014 WHO Classification. The rationale cited for this approach was that therapy for hyperplasia should be aimed at treating the suspected cause and symptoms, whereas monoclonal lesions should be removed or ablated. Because clonality cannot be performed on diagnostic specimens in most laboratories, it was proposed that the diagnosis of EIN be made when glandular crowding that

L. Hedrick Ellenson et al.

resulted in a volume percentage of stroma (VPS) was less than 55% (Mutter 2000). Ideally, this parameter would be assessed by morphometric analysis, which has been shown to separate hyperplasias, particularly those classified as nonatypical by light microscopy, into monoclonal and polyclonal lesions. However, to make the system applicable for routine practice, diagnostic parameters that could be determined on light microscopy were developed. The morphometric analysis to determine the relationship of glandular crowding to intervening stroma was referred to as a “D-score.” A “D-Score” or “multivariate discriminant score” focuses on three features: (1) VPS, (2) gland branching/convolution (outer surface density of glands), and (3) standard deviation of the shortest nuclear axis which is correlated with nuclear atypicality. This classification was used to predict the rate of progression to cancer. A “D-score” less or equal to 1 predicted a high rate of progression to endometrial carcinoma, whereas a score greater than 1 almost never indicated progression to cancer. Accordingly, the criteria for EIN are based on architecture (gland area that exceeds that of stroma, usually localized), cytological alterations (cytology in the lesion differs from the background), and lesion size (greater than 1 mm linear dimension) (Mutter 2000). Given that these criteria are essentially identical to those used to make the diagnosis of atypical hyperplasia the most recent WHO Classification system equated “atypical hyperplasia” with “EIN.” Although in many cases this is true, other lesions classified as EIN correspond to hyperplasia without atypia which can, therefore, lead to overdiagnosis and unnecessary treatment. Some studies have suggested improved reproducibility in the diagnosis of EIN versus atypical hyperplasia, but other studies have found similar reproducibility (Hecht et al. 2005; Ordi et al. 2014) Moreover, in a nested case-control study, the risk of progression to carcinoma was similar after either a diagnosis of EIN or atypical hyperplasia(Lacey et al. 2008). Although not included in the current WHO classification system, it is noteworthy that the European Working Group (EWG) developed

8

Precursors of Endometrial Carcinoma

another classification system for endometrial proliferative lesions. In this system, simple and complex hyperplasia without atypia are referred to as “hyperplasia,” while atypical hyperplasia and well-differentiated carcinomas are combined into a single category designated “endometrial neoplasia.” It was stated that this system showed greater reproducibility than the hyperplasia/atypical hyperplasia and the EIN system. This is not surprising as the EWG classification is a two-tier system and it was compared to the older 4 tier hyperplasia/atypical hyperplasia system (Bergeron et al. 1999). We are unaware of a carefully performed study in which all three systems (specifically the 2014 WHO Classification of hyperplasia/atypical hyperplasia and EIN) are compared head to head with the EWG Classification to substantiate this claim. Nonetheless, use of the term “endometrial neoplasia” in the EWG Classification has a completely different meaning than EIN in the WHO Classification which further compromises both the EIN and the “endometrial neoplasia” classifications.

Molecular Genetics and Immunohistochemistry Molecular genetic studies are mentioned only briefly here with a more detailed discussion in ▶ Chap. 9, “Endometrial Carcinoma.” There have been a number of molecular genetic alterations identified in atypical hyperplasia, including microsatellite instability and mutations in the PTEN tumor suppressor gene and the KRAS oncogene. Of note, these are among the most common molecular genetic alterations in endometrioid carcinoma. These findings support clinicopathologic and epidemiologic data indicating that atypical hyperplasia is the immediate precursor for endometrioid carcinoma. A study, comparing mutations in atypical hyperplasia and associated carcinomas, found shared and unique mutations in both components, suggesting a process of complex subclone evolution versus a linear accumulation of molecular events from hyperplasia to carcinoma (Russo et al. 2017). PTEN mutations are found at approximately the same frequency

453

in complex atypical hyperplasia and carcinoma and have rarely been described in hyperplasia without atypia (Hayes et al. 2006). In addition, KRAS mutations and microsatellite instability have been reported in atypical hyperplasia (Levine et al. 1998; Esteller et al. 1999; Enomoto et al. 1993). Only a relatively small number of hyperplasias without atypia have been analyzed for mutations in PTEN and KRAS and for microsatellite instability. Studies suggest that all of these alterations occur before the development of invasion, but it is not clear when in the progression of the disease they occur. Recent immunohistochemical studies have confirmed a decrease or loss of PTEN expression in 70% of atypical hyperplasia with some containing focal areas of ARID1A loss (Ayhan et al. 2015). Of note, the areas with loss of both PTEN and ARID1A showed an increase in the proliferation index as measured by Ki67 immunohistochemistry, suggesting that ARID1A prevents PTEN inactivation from promoting cell proliferation. At the present time, it is not recommended that DNA mismatch repair immunohistochemistry be utilized on atypical hyperplasia to screen for Lynch syndrome. Although a number of studies have been done with a wide variety of antibodies, none have proven adequately robust for the diagnosis of EH.

Behavior Many of the past studies designed to determine the outcome of women with EH did not consider cytologic and architectural features separately. This issue was addressed in a retrospective analysis of 170 patients with EH on curettings with a mean follow-up of 13.4 years in which hysterectomy was not performed before 1 year after the initial diagnosis. Various histologic features were evaluated, and cytologic and architectural abnormalities were analyzed independently in an effort to delineate the histologic features associated with an increased risk of progression to carcinoma. A third of the patients with both nonatypical and atypical hyperplasia were asymptomatic after the

454

L. Hedrick Ellenson et al.

diagnostic curettage and required no further treatment. Only 2 (2%) of 122 patients with hyperplasia lacking cytologic atypia, one with simple and one with complex hyperplasia, progressed to carcinoma. The two cases of simple hyperplasia that progressed to carcinoma first developed atypical hyperplasia. In contrast, 11 (23%) of the 48 women with atypical hyperplasia progressed to carcinoma (Table 3); 8% of patients with simple atypical hyperplasia and 29% of patients with complex atypical progressed to carcinoma (Table 4). The presence of glandular complexity and crowding superimposed on atypia, therefore, appears to place the patient at greater risk than does cytologic atypia alone. The differences in progression to carcinoma among the four subgroups, however, were not statistically significant. A subsequent nested-case-control study of 138 cases diagnosed with EH followed by carcinoma at least 1 year later with 241 controls (matched for age, date and follow-up duration and counter-matched for EH diagnosis) showed that women with both simple and complex EH without atypia had a 10% probability of developing carcinoma in contrast to a 40% probability for women with atypical hyperplasia, both simple and complex (Lacey et al. 2008). These studies show that cytologic atypia is the most useful feature in identifying a lesion that might progress to carcinoma and provided a robust rationale for the

adoption of a two-tiered classification system based solely on the presence of cytological atypia. The carcinomas that develop in patients with hyperplasia are relatively innocuous (Kurman et al. 1985; Gusberg and Kaplan 1963). In one study, the mean duration of progression of hyperplasia without atypia to carcinoma is nearly 10 years, and it takes a mean of 4 years to progress from atypical hyperplasia to clinically evident carcinoma (Kurman et al. 1985). In another study, the median interval from atypical hyperplasia to carcinoma was 6.7 years. It has been shown that 17–43% of women with atypical hyperplasia in curettings will have a welldifferentiated carcinoma in the uterus if a hysterectomy is performed within 1 month of the curettage (King et al. 1984; Tavassoli and Kraus 1978). Increasing degrees of nuclear atypia, mitotic activity, and stratification of cells in curettings are associated with a higher frequency of carcinoma in the uterus. With long-term follow-up, however, only 11–40% of women with atypical hyperplasia develop carcinoma if a hysterectomy is not done (Kurman et al. 1985; Gusberg and Kaplan 1963). Thus, the lesion designated as well-differentiated carcinoma usually remains stable for a long period of time. Reasons that may account for the relatively low rate of progression to carcinoma in untreated patients with atypical hyperplasia include a general tendency for the highest grade of atypical hyperplasia to be

Table 3 Follow-up of hyperplasia and atypical hyperplasia in 170 patients Type of hyperplasia Hyperplasia Atypical hyperplasia

No. of patients 122 48

Regressed No. (%) 97 (80) 29 (60)

Persisted No. (%) 23 (19) 8 (17)

Progressed to carcinoma No. (%) 2 (2) 11 (23)

Adapted with permission from Kurman et al. (1985)

Table 4 Follow-up of simple and complex hyperplasia and atypical hyperplasia in 170 patients Type of hyperplasia Simple Complex Simple atypical Complex atypical

No. of patients 93 29 13 35

Regressed No. (%) 74 (80) 23 (80) 9 (69) 20 (57)

Adapted with permission from Kurman et al. (1985)

Persisted No. 18 5 3 5

(%) (19) (17) (23) (14)

Progressed to carcinoma No. (%) 1 (1) 1 (3) 1 (8) 10 (29)

8

Precursors of Endometrial Carcinoma

selected for hysterectomy, leaving the lesser degree of atypia for conservative management. In addition, atypical hyperplasia may represent a heterogeneous group with different constellations of genetic or epigenetic abnormalities with varying propensities to develop carcinoma. And it is likely that patient factors (e.g., immune response, obesity, hormonal milieu) may also play an important role in determining progression to carcinoma. An additional study of the behavior of EH found that most cases of EH without atypia regressed spontaneously, whereas those with complex atypical hyperplasia were much more likely to persist (Terakawa et al. 1997). Another study confirmed the significance of cytologic atypia in predicting increased risk of associated endometrial carcinoma in hysterectomy specimens (Hunter et al. 1994). Because most atypical hyperplasias have complex architecture, it is complex atypical hyperplasia that is associated with a significant risk of persistence and progression to carcinoma. Hence, this lesion is regarded as a direct precursor of well-differentiated endometrioid carcinoma of the endometrium. However, hyperplasia is identified in a prior endometrial specimen or in the hysterectomy specimen in only 35–75% of women with endometrial carcinoma (Ayhan et al. 1991; Beckner et al. 1985; Bokhman 1983; Deligdisch and Cohen 1985; Gucer et al. 1998; Kaku et al. 1996). In those reports that specified the number of hyperplasias that were classified

455

as atypical, 14–36% of women with endometrial carcinoma had associated atypical hyperplasia (Gucer et al. 1998; Kaku et al. 1996). It is unclear whether failure to identify an associated atypical hyperplasia in all cases of endometrioid carcinoma reflects overgrowth of a preexisting hyperplasia by carcinoma or the development of carcinoma through an alternative pathway.

Management Management of patients with endometrial hyperplasia is based on clinical factors which include the desire to preserve fertility in young women and associated medical conditions that render older women at high risk for a surgical procedure in addition to the microscopic findings. (Kraus 1985)

Premenopausal Women (Less than 40 Years of Age) Most premenopausal women who present with abnormal bleeding have nonspecific hormonal disorders that are self-limited. These women are at low risk of having carcinoma (Table 5). In a study of 460 women 40 years of age and younger, 6 (1.3%) had “mild” hyperplasia (simple hyperplasia) but none had atypical hyperplasia or carcinoma (Kaminski and Stevens 1985). Therefore, most women in this age group with abnormal bleeding do not require an endometrial biopsy.

Table 5 Pertinent findings in endometrium in women with abnormal bleeding according to age

Finding in endometrial specimena Carcinoma Atypical hyperplasia Hyperplasia Atrophy Polyp Proliferative Secretory

Age Premenopausal 40 years (n = 5,460) No. (%) 0 (–) 0 (–) 6 (1) 7 (2) 6 (1) 139 (29) 241 (50)

Perimenopausal 40–55 years (n = 5,748) No. (%) 3 (0.4) 5 (0.7) 41 (6).0 51 (7).0 13 (2).0 273 (36) 287 (38)

Postmenopausal 55 years (n = 5,226) No. (%) 15 (7) NKb 34 (15) 127 (56) 19 (8) 31 (14) 0 (–)

NK not known a Not all the endometrial findings in the study by Kaminski and Stevens are listed and therefore percentages do not total 100% b A category of atypical hyperplasia was not specified in this study

456

L. Hedrick Ellenson et al.

Table 6 Subsequent pregnancies in “untreated” women with hyperplasia and atypical hyperplasia

Diagnosis Simple hyperplasia Complex hyperplasia Atypical hyperplasia (simple and complex)

No. of patients, 40 35

No. of patients who became pregnant 10 (29%)

Table 7 Hysterectomy findings according to the presence of atypical hyperplasia or well-differentiated adenocarcinoma in curettings in women under 40 years of age Curettings

No. of fullterm pregnancies 19

15

3 (20%)

4

24

3 (13%)

4

Adapted from Kurman et al. (1985)

Women with risk factors for endometrial cancer, such as polycystic ovarian syndrome or obesity, and women with persistent bleeding should have an endometrial biopsy performed. Recent studies suggest that a BMI of greater than 30 in premenopausal women resulted in a four- to sevenfold increase in atypical hyperplasia or cancer compared to women with a BMI less than 30. Thus, endometrial sampling is recommended for premenopausal women with abnormal bleeding and a BMI greater than 30 (Wise et al. 2016; Guraslan et al. 2016). If a diagnosis of hyperplasia (without atypia) is made, the patient can be treated conservatively because these lesions have an extremely low risk (1–2%) of progression to carcinoma. Because the transit time to carcinoma is approximately 10 years and hyperplasia without cytologic atypia first progresses through atypical hyperplasia before becoming carcinoma, follow-up and periodic endometrial biopsies suffice (Kurman et al. 1985). Conservative management of young women with simple hyperplasia and complex hyperplasia resulted in subsequent pregnancies in 29% and 20% of these women, respectively, in one study (Table 6) (Kurman et al. 1985). Women with atypical hyperplasia on an endometrial biopsy who wish to preserve their fertility should be treated with progestin suppression. In view of the very similar accuracy of endometrial biopsy and curettage and the low risk of an associated endometrial carcinoma in women younger than

Hysterectomy findings Carcinoma Grade 1 Grade 2 Myometrial invasion Endometrium only Inner one third Middle one third

Atypical hyperplasia (n = 517)

Welldifferentiated carcinoma (n = 535)

No. 2 (12%) 2 0

No. 13 (37%) 10 3

2 0 0

3 9 1

Adapted from Kurman and Norris (1982)

40 years of age, a curettage need not be performed to exclude carcinoma, but close follow-up and periodic endometrial biopsies are necessary. A conservative plan of management is justified because the risk of progression to carcinoma in young women is low and the carcinomas that do develop tend to be innocuous (Table 7); 20% of those women less than 40 years of age can subsequently become pregnant and have normal deliveries (Randall and Kurman 1997). A recent study of complex hyperplasia with and without atypia found that progestin therapy resulted in regression of complex atypical hyperplasia (Table 8) but that the majority of complex hyperplasias without atypia regressed with or without progestin therapy (Reed et al. 2009). This study also found that higher doses and longer duration of progestin therapy increase the likelihood of regression of complex atypical hyperplasia (Table 9). Conservative management also can be considered for women diagnosed with well-differentiated carcinoma. One study of progestin treatment of atypical hyperplasia and well-differentiated carcinoma in women under age 40 found that 75% of women with carcinoma and 95% with atypical hyperplasia had regression of their lesions (Randall and Kurman 1997). In addition, all patients were alive without evidence of progressive disease during the follow-up period. The median duration of progestin treatment necessary to effect regression was 9 months. In another study, 62% of women

8

Precursors of Endometrial Carcinoma

457

Table 8 Risk of persistence/progression of complex hyperplasia with and without atypia in relation to progestin therapy

Treatment No progestin (n = 20) Progestin (n = 95)

Complex hyperplasia without atypia (n = 115) Regression Persist/progress (n = 82) (n = 33) 14 (70%) 6 (30%) 68 (72%) 27 (28%)

Complex atypical hyperplasia (n = 70) Regression Persist/progress (n = 44) (n = 26) 6 (33%) 12 (67%) 38 (73%) 14 (27%)

Table 9 Risk of persistence/progression of complex hyperplasia with and without atypia in relation to duration of progestin therapy

Duration of Rx 3 months

Complex hyperplasia without atypia (n = 115) Regression Persist/progress (n = 82) (n = 33) 71% 30% 73% 28%

under 40 years treated with progestins alone for endometrial carcinoma responded to the hormonal therapy, although 23% of these later developed recurrent disease (Kim et al. 1997). Ninety percent of the patients were alive without evidence of disease during the follow-up period. The lower frequency of responders in this study may have been a result of the relatively short duration of the hormonal therapy. Thus, in premenopausal women, atypical hyperplasia and well-differentiated carcinoma can be regarded as a single clinicopathologic entity for management purposes. Nonetheless, pathologists should distinguish atypical hyperplasia from well-differentiated carcinoma since atypical hyperplasia is more likely to respond to progestin treatment. More recent studies suggest that levonorgestrel-releasing intrauterine systems may increase regression; however, clinical trials of hormone therapy are needed to establish standard therapeutic approaches (Chandra et al. 2016; Trimble et al. 2012). If conservative management is elected, magnetic resonance imaging (MRI) must be performed to exclude deep myometrial invasion or the presence of a coexisting ovarian neoplasm.

Perimenopausal Women (40–55 Years of Age) Abnormal bleeding in the perimenopausal age group can be managed in a similar fashion as in

Complex atypical hyperplasia (n = 70) Regression Persist/progress (n = 44) (n = 26) 62% 38% 87% 13%

younger women because perimenopausal women also are at low risk of having carcinoma (see Table 5). Most simple and complex hyperplasias in the 40- to 55-year-old age group are related to anovulation and are self-limited. Nonetheless, a biopsy is usually performed to exclude carcinoma. Patients with a diagnosis of atypical hyperplasia can be treated with progestins or a hysterectomy. Nearly 60% of atypical hyperplasias regress, but the likelihood of residual carcinoma in the uterus after a curettage increases with age. For patients in the 40- to 55-year age range, treatment should be individualized. Regression occurs frequently, and the risk of residual carcinoma is lower than in older women. Therefore, observation or suppression with progestins monitored by endometrial biopsies every 3 months suffices. If the lesion persists, a hysterectomy may have to be performed.

Postmenopausal Women (Over 55 Years of Age) Women in the postmenopausal age group who have abnormal bleeding have a significant risk of having either carcinoma or atypical hyperplasia (see Table 5). Accordingly, vaginal bleeding requires immediate evaluation with an endometrial biopsy. A diagnosis of hyperplasia or atypical hyperplasia should be evaluated with a fractional curettage. If the curettings demonstrate hyperplasia without

458

atypia, conservative management is an option because these types of hyperplasia are related to unopposed estrogenic stimulation, either from exogenous hormone treatment or because of peripheral conversion of androgens to estrogen in adipose tissue. Most (80%) hyperplasias treated with cyclic medroxyprogesterone acetate at 10 mg/day for 14 days regress; none progressed to carcinoma in a prospective study of 65 postmenopausal women (Ferenczy and Gelfand 1989). Conservative management, either observation only or treatment with medroxyprogesterone to produce a medical curettage, therefore, is adequate. Repeated episodes of irregular bleeding that are not responsive to hormone treatment require a hysterectomy. Hysterectomy is the treatment of choice for a diagnosis of atypical hyperplasia based on a curettage. In postmenopausal women with surgical risk factors that preclude a hysterectomy, continuous treatment with 20–40 mg/day megestrol acetate can be used effectively to avoid surgery. In a study of 70 treated women with complex hyperplasia (38 women) and atypical hyperplasia (32 women), surgery was avoided in 93% of patients. The hyperplasias (atypical and nonatypical) completely regressed in 85% after a mean follow-up of more than 5 years. None of the lesions progressed to carcinoma (Gal 1986). For postmenopausal women with hyperplasia or atypical hyperplasia who are receiving exogenous estrogen, termination of the estrogen usually suffices even for atypical hyperplasia, because these proliferations regress after the stimulus for their growth has been removed. Alternatively, the addition of cyclically or continuous administered medroxyprogesterone in women being treated with estrogen can be considered because the use of even low doses of progestins substantially reduces the risk of development of EH and carcinoma. Using a 7- to 14-day regimen of orally administered 10 mg medroxy-progesterone to postmenopausal women receiving estrogen, five endometrial carcinomas were detected in 5402 woman-years of continuous estrogen therapy (Greenblatt et al. 1982). This incidence is not greater than that of untreated postmenopausal women, in whom the expected incidence of endometrial cancer is 1–2 per 1000 woman-years, that is, 5.4–9.8 cases.

L. Hedrick Ellenson et al.

Morphologic Changes Associated with Progestin Treatment Treatment with progestins for women with complex atypical hyperplasi and well-differentiated carcinoma has become an accepted alternative to hysterectomy. Correct interpretation of the morphologic changes that result from the treatment is required for the appropriate management of women who chose this therapeutic approach. Unfortunately, there have been few studies to describe these changes. One study described a number of histologic changes including decreased gland-to-stroma ratio, decreased glandular cellularity, reduced to absent mitotic activity, loss of cytologic atypia and a variety of metaplastic changes (Figs. 31 and 32) (Wheeler et al. 2007). In addition, cribriform and papillary architectural changes may be induced by treatment and confused with disease progression. Importantly, the persistence of architectural abnormalities and/or cytologic atypia after 6 months of treatment were the only histologic features found to be associated with treatment failure. Based on these findings, a classification of progestin-treated lesions was proposed (Table 10). It is important that the pretreatment specimen be available for review for the pathologist to evaluate the response to progestin treatment and therefore provide information to the gynecologist that will assist in modifying or discontinuing therapy.

Endometrial Cellular Changes: Metaplasia, Cellular Differentiation In contrast to hyperplasia, which is a proliferative response to estrogenic stimulation, metaplasia represents cytoplasmic differentiation. The cytoplasmic alterations (metaplasia) are manifested by eosinophilic, ciliated cell (tubal), squamous, secretory/clear, and mucinous differentiation. Metaplasias develop most commonly in response to estrogenic and progestational stimulation, although these changes may develop in response to various other stimuli as well. Thus, the morphologic response of the endometrium to hormonal stimulation is complex and is reflected by

8

Precursors of Endometrial Carcinoma

Fig. 31 Complex hyperplasia with treatment (progestin) effect. Individual glands with complex profiles are interspersed with inactive glands within endometrial stroma (a).

459

The confluent glandular pattern of the prior endometrioid carcinoma is absent. Glands exhibit mucinous metaplasia (b)

Table 10 Classification of progestin-treated lesions of the endometrium Diagnosis Progestin-treated complex hyperplasia

Progestin-treated CAH Progestin-treated welldifferentiated carcinoma Fig. 32 Complex hyperplasia with treatment (progestin) effect. Complex glands exhibit prominent metaplastic changes (eosinophilic, mucinous, tubal) which are commonly seen, along with reduction of nuclear atypia, as a result of progestin therapy

a combination of architectural, nuclear, and cytoplasmic alterations. Although classifications separate hyperplasia and the various metaplasias, both are usually intimately associated and cannot always be separately classified.

Definitions and Classification Metaplasia is defined as replacement of one type of adult tissue by another type that is not normally found in that location. In the endometrium, most of

Histologic features No cytologic atypia with crowded, back-to-back glands and/or a confluent glandular pattern (cribriform and/or papillary) Cytologic atypia with crowded, back-to-back glands that lack a confluent glandular pattern Cytologic atypia with confluent glandular pattern (cribriform and/or papillary pattern)

Adapted with permission from Wheeler et al. (2007)

the changes that are designated as metaplasia represent a variety of cytoplasmic alterations or forms of differentiation that are not encountered in normal proliferative endometrium but do not qualify as true metaplasia. Accordingly, it has been suggested that a more appropriate term is change (Silverberg and Kurman 1992). Use of the term change also has the advantage of providing a descriptive designation without employing a specific mechanism of development. In this chapter, the terms metaplasia, change, and differentiation are used interchangeably. The various forms of cellular differentiation are typically focal when unaccompanied by hyperplasia but can be diffuse

460 Table 11 Classification of endometrial metaplasia Papillary syncytial Eosinophilic and ciliated Mucinous Hobnail Squamous Secretory Papillary proliferation

when hyperplasia is present. As previously noted, the endometrial epithelium can undergo a variety of cytoplasmic changes in response to different stimuli that can be observed in both benign and malignant conditions. A simplified classification of these is shown in Table 11. It is important to recognize the various cytoplasmic changes because they are benign and can be confused with hyperplasia. When hyperplasia and the cytoplasmic alterations coexist, as they often do, the hyperplasia should be classified, but it is not necessary to describe the cytoplasmic changes because they do not influence prognosis (see section “Behavior”).

Clinical Features The frequent association of the various endometrial cytoplasmic changes with hyperplasia probably results from a hyperestrogenic state. More than 70% of perimenopausal and postmenopausal women with metaplasia had received exogenous estrogen in one study (Hendrickson and Kempson 1980). In addition, most young women with metaplasia have clinical manifestations of persistent anovulation and primary infertility, features of polycystic ovarian syndrome (Hendrickson and Kempson 1980; Crum et al. 1981). Metaplasia also may occur in various benign conditions, including polyps, endometritis, trauma, and vitamin A deficiency (Hendrickson and Kempson 1980; Crum et al. 1981; Fluhmann 1954).

Pathologic Findings The various types of endometrial cytoplasmic changes have no distinctive gross features.

L. Hedrick Ellenson et al.

Papillary Syncytial Metaplasia On the endometrial surface, cells with eosinophilic cytoplasm typically merge into a syncytium that either can be flat or more commonly form papillary processes (Rorat and Wallach 1984). Typically, the papillary processes lack connective tissue support and contain small cystic spaces filled with polymorphonuclear leukocytes. This lesion has been referred to as surface syncytial change, papillary syncytial change, or papillary metaplasia (Hendrickson and Kempson 1980; Silverberg and Kurman 1992). We prefer the term eosinophilic syncytial change because the lesion is characteristically composed of eosinophilic cells forming a syncytium and can involve glands as well as the surface. Eosinophilic syncytial change is commonly associated with endometrial stromal breakdown or inflammation, suggesting that it is a degenerative or a reparative process (Zaman and Mazur 1993). A more recent study using immunohistochemical stains for proliferation further supports the regressive nature of this type of epithelial change (Shah and Mazur 2008). The nuclei within the syncytium are arranged haphazardly and piled up; they generally are small and bland but at times may be round and vesicular and display alterations in shape and chromatin texture. Mitotic figures are rare. Hyperchromatic nuclei with smudged chromatin and irregular nuclear membranes appear degenerated whereas enlarged, vesicular nuclei with a prominent nucleolus and smooth nuclear membranes appear reactive. These degenerative and reparative changes should not be interpreted as nuclear atypia. Eosinophilic and Ciliated Cell Metaplasia Eosinophilic change is the most common metaplasia (Figs. 33 and 34). Several types of eosinophilic cytoplasmic transformation occur, all of them innocuous. Ciliated cells, squamous cells, oncocytes, and papillary and surface syncytial change all may have eosinophilic cytoplasm. However, eosinophilic cells also occur in association with hyperplasia, particularly atypical hyperplasia. Glands may be partially or completely lined by eosinophilic cells. Eosinophilic cells that line glands can show considerable variation in shape.

8

Precursors of Endometrial Carcinoma

Fig. 33 Hyperplasia without atypia with eosinophilic metaplasia. Intraglandular epithelial tufts composed of bland cells with abundant eosinophilic cytoplasm are present within hyperplasia without atypia

Fig. 34 Atypical hyperplasia with eosinophilic metaplasia. Intraglandular epithelial clusters composed of rounded cells with abundant eosinophilic cytoplasm are present within an atypical hyperplasia; assessment of atypia is based on the appearance of the glandular epithelium which displays vesicular nuclei with prominent nucleoli and loss of polarity

They may be columnar when associated with atypical hyperplasia, rounded when associated with ciliated cells, or polygonal, forming pavementlike aggregates, when they merge with cells that show squamous differentiation. In hyperplastic lesions, aggregates of eosinophilic cells often form intraglandular papillary tufts and bridges, thus simulating carcinoma. Eosinophilic cells contain variable amounts of cytoplasm that at times can be partially vacuolated. The nuclei tend to be

461

round and somewhat stratified. In most instances, the nuclei are smaller, more uniform, and lack the irregular nuclear membrane, coarse chromatin, and nucleoli that characterize cells with true cytologic atypia. Occasionally, the nuclei can be enlarged and contain a single prominent nucleolus. Mitotic figures are rarely present. Eosinophilic change is often seen in combination with cliated cell change (tubal metaplasia). Cilia are not usually evident microscopically in proliferative endometrial glandular cells, although they may be observed on the endometrial surface (Masterton et al. 1975). Ciliated cells occasionally are observed in isolated glands in atrophic or inactive endometria or in polyps in the absence of hyperplasia. The presence of a significant number of ciliated glandular cells is referred to as ciliated cell change or tubal metaplasia because of the resemblance to the epithelium of the fallopian tube. The ciliated cells are often round and slightly enlarged, but the nuclear membranes are smooth and uniform and the chromatin is fine and evenly dispersed. There is no nuclear atypia. The ciliated cells may be interspersed singly or in small groups among nonciliated cells, or they may line a larger segment of a gland. Mitotic activity is limited to the adjacent nonciliated cells. Ciliated cell change may occur in glands in the absence of hyperplasia. Dilated venous sinusoids are also frequently present. All these changes reflect a mild degree of estrogenic stimulation. Ciliated cell change frequently accompanies simple, complex, or atypical hyperplasia (Figs. 35, 36, and 37).

Mucinous Metaplasia Mucinous change is characterized by mucinous epithelium resembling that of the endocervix cytologically, histochemically, and ultrastructurally (Demopoulos and Greco 1983). Although it is one of the least commonly encountered cytoplasmic alterations, it occurs more frequently than generally described. The mucinous epithelium tends to be distributed focally and is composed of tall columnar cells with bland, basally oriented nuclei and clear, slightly granular cytoplasm (Figs. 38 and 39). At times mucinous change is accompanied by a papillary proliferation. The papillary processes contain normal but

462

L. Hedrick Ellenson et al.

Fig. 35 Hyperplasia without atypia with ciliated metaplasia. Crowded glands have multiple cell types, including ciliated cells and those with rounded nuclei and cytoplasmic clearing (halos), similar to those seen in fallopian tube epithelium

Fig. 37 Atypical hyperplasia with ciliated metaplasia. Markedly crowded glands have some cells with rounded nuclei and cytoplasmic clearing indicating underlying tubal metaplasia, but the nuclear enlargement and loss of polarity are beyond that attributable to metaplasia and indicate atypia

Fig. 36 Hyperplasia without atypia with ciliated metaplasia. Crowded glands have some ciliated cells and cells with nuclear rounding and cytoplasmic clearing

Fig. 38 Hyperplasia without atypia with mucinous change. Crowded glands have abundant pale mucinous cytoplasm and basally situated small nuclei

compressed stromal cells and are lined by nonstratified columnar epithelium, which is mucinous in areas. Mitotic figures are rare. The cytoplasm is clear in hematoxylin and eosin (H&E) stains because it contains mucin, which is periodic acid–Schiff (PAS) positive and diastase resistant and stains with mucicarmine, toluidine blue, and alcian blue. In contrast to mucinous epithelium, the vacuolated cytoplasm of secretory endometrium contains glycogen. On rare occasion the mucinous epithelium may contain goblet cells and is referred to as intestinal metaplasia.

Mucinous differentiation can be seen in a spectrum of epithelial proliferations ranging from benign to malignant. In one study, the likelihood of finding carcinoma associated with mucinous proliferations of the endometrium varies according to the degrees of architectural complexity and cytologic atypia of the lesions (Nucci et al. 1999). Architecturally simple lesions with papillary projections into luminal spaces and no cytologic atypia were found to have carcinoma on follow-up only when the initial specimen also contained atypical hyperplasia; otherwise, none of these simple mucinous proliferations was

8

Precursors of Endometrial Carcinoma

Fig. 39 Hyperplasia without atypia with mucinous change. Crowded glands are lined by cells with pale mucinous cytoplasm and small round nuclei, with some forming papillary tufts

associated with carcinoma on follow-up. More complex proliferations with microglandular or cribriform patterns and minimal cytologic atypia, often presenting as endometrial surface lesions without coexistent atypical hyperplasia, were found to have well-differentiated noninvasive or minimally invasive carcinoma on follow-up in 65% of cases. Highly complex proliferations with glandular budding, cribriform growth, and branching of villous structures that also displayed moderate to severe cytologic atypia were invariably associated with carcinoma on follow-up (Fig. 40). Importantly, 80% of the study patients were over age 50. Thus, in perimenopausal and postmenopausal women with complex mucinous proliferations, including those with and without cytologic atypia, the risk of finding coexistent carcinoma is high.

Hobnail Metaplasia Hobnail change is characterized by luminal glandular epithelial cells with a nucleus that protrudes into the gland lumen. The cells are often eosinophilic so there is overlap with eosinophilic change, but hobnail metaplasia is recognized separately when there are an abundance of cells with protruding nuclei that produce a striking finding (Fig. 41). Unlike Arias-Stella reaction, the cells do not have prominent cytoplasmic clearing or cytologic atypia.

463

Fig. 40 Complex atypical mucinous proliferation. This term is used for limited specimens in which crowded, fused or cribriform glands with mucinous cytoplasmic change and atypical nuclei are concerning for at least complex atypical hyperplasia, but there is little or no associated stroma to diagnose stromal invasion. Such lesions are often associated with International Federation of Gynecology and Obstetrics (FIGO) grade 1 stage 1A endometrioid carcinoma on follow-up, as occurred in this case

Fig. 41 Hyperplasia without atypia with hobnail change. Crowded glands with cells containing hyperchromatic rounded nuclei that bulge into the gland lumen and some cytoplasmic clearing with vacuolization are reminiscent of the Arias-Stella reaction

Squamous Metaplasia (Squamous Differentiation) Squamous differentiation may occur in all forms of hyperplasia (Figs. 42, 43, and 44) as well as in carcinoma. It is especially common in the more atypical endometrial proliferations and is rare in normally cycling endometrium or in simple and complex hyperplasias. The squamous cells are usually cytologically bland. The degree of nuclear atypia, when present, generally parallels that of

464

Fig. 42 Hyperplasia without atypia with squamous metaplasia. Islands of squamous metaplastic epithelium are intimately admixed with nonatypical hyperplastic glands

Fig. 43 Hyperplasia without atypia with squamous metaplasia. Squamous morules are intimately admixed with nonatypical hyperplastic glands exhibiting tubal metaplasia as well

the glandular cells. Typically, the squamous cells have a moderate amount of eosinophilic cytoplasm and are surrounded by a well-defined cell membrane. Often they merge with eosinophilic cells that qualify as eosinophilic change. The squamous cells tend to be rounded or polygonal but may be spindle shaped, forming a circumscribed nest (squamous morule) within the gland lumen (Fig. 43). Morules reflect immature or incomplete squamous differentiation. The cells are smaller and the cytoplasm is less prominent than in more completely differentiated squamous cells. Central keratinization and necrosis rarely occur. Eventually, proliferation results in protrusion of the squamous cells into the

L. Hedrick Ellenson et al.

Fig. 44 Hyperplasia without atypia with squamous metaplasia. Squamous metaplastic epithelium with central necrosis forms a confluent mass intimately admixed with nonatypical hyperplastic glands but should not be interpreted as an indicator of stromal invasion when the glandular component does not manifest confluence or induce desmoplasia

lumen, leading to replacement of the lumen by nests of squamous cells and coalescence with neighboring glands undergoing the same process. Mitotic activity is rare. A recent study found that squamous morules in hyperplastic proliferations lack expression of the estrogen and progesterone receptors and demonstrate rare to undetectable Ki-67 proliferative activity when compared to the associated hyperplastic epithelium. However, the glandular and squamous components had identical PTEN mutations indicating that the squamous component is clonally related to the glandular component. The authors concluded that the squamous morules are inert elements of the proliferative lesions. Importantly, since they are often associated with complex atypical hyperplasia and endometrioid carcinoma, their presence on endometrial sampling, in the absence of an identifiable proliferative process, should result in close follow-up of the patient for the possibility of an under sampled or occult glandular lesion (Lin et al. 2009).

Secretory Change Secretory change is characterized by columnar cells with sub- or supranuclear vacuoles containing clear glycogenated cytoplasm resembling the glandular cells of early secretory endometrium. These cells also can be observed in nonneoplastic proliferative

8

Precursors of Endometrial Carcinoma

endometria but are seen more often in association with hyperplasia or carcinoma (Fig. 45). Rarely, the cells in secretory change can display hobnail morphology reminiscent of the Arias–Stella reaction (Fig. 41); such hobnail change, with or without the cytoplasmic vacuolization of secretory change (Fig. 46), should not be misinterpreted as endometrioid intraepithelial carcinoma (see below). At times secretory change can result from progestational stimulation, but often there is no such association. Columnar cells with secretory change may merge with polygonal-shaped clear cells and with

465

squamous cells containing clear glycogenated cytoplasm. The accumulation of glycogen can occur in the cytoplasm of a variety of cell types.

Papillary Proliferation This entity consists of papillary proliferations of fibrovascular stromal cores that are lined by benign endometrial epithelium. The papillae range from small, simple papillae to very complex, branching papillae. The epithelium is usually flattened, but may be tufted and often shows cytoplasmic change including mucinous metaplasia, eosoinophilic, ciliated, and hobnail changes. Squamous metaplasia has also been described. The entity is thought to be benign and should be distinguished from carcinoma with papillary growth patterns (Ip et al. 2013; Lehman and Hart 2001).

Differential Diagnosis

Fig. 45 Hyperplasia without atypia with secretory change. Crowded glands have elongated nuclei and discrete sub- and supranuclear vacuoles, reminiscent of day 18 secretory endometrium

Fig. 46 Hyperplasia with hobnail change. Crowded glands are partially lined by epithelial tufts containing cells with hyperchromatic rounded nuclei that bulge into the gland lumen. These tufts also share features with eosinophilic cell change

The most important aspect of the evaluation of the various metaplasias and cellular changes is not to confuse them with hyperplasia or carcinoma, which is best accomplished by evaluating the glandular architecture and cytological features. In hyperplasia, the glandular outlines are irregular and complex and there is stratification of the epithelium reflecting a proliferative process. In contrast, in the various cytoplasmic changes, the glandular outlines are regular and have a tubular configuration, although cystic dilatation and slight glandular irregularity occasionally can occur. Although the various cellular changes may be accompanied by slight nuclear enlargement, the cells lack the abnormal chromatin patterns that characterize the nuclei in atypical hyperplasia. At times the various cellular changes may look ominous and suggest carcinoma, but evidence of stromal invasion is lacking and therefore a diagnosis of carcinoma is not justified. For example, extensive squamous metaplasia may suggest a diagnosis of carcinoma but without a desmoplastic response or a confluent glandular pattern a diagnosis of carcinoma should not be made. Squamous and eosinophilic change associated with hyperplasia can fill and bridge gland lumens but lack a true confluent or cribriform

466

pattern. Mucinous change at times can form complex papillary processes, but the stroma of the papillae are composed of normal endometrial stroma and the epithelium lacks cytologic atypia.

Behavior Cytoplasmic changes, other than eosinophilic syncytial change, rarely occur in the absence of hyperplasia or carcinoma (Kaku et al. 1992). In the absence of hyperplasia, these changes (metaplasia) had no clinical significance in one study of 89 patients (Hendrickson and Kempson 1980). In a long-term follow-up study of endometrial hyperplasia, 5 of 11 patients with atypical hyperplasia and associated squamous metaplasia eventually developed carcinoma, indicating that atypical hyperplasia with squamous metaplasia has malignant potential (Kurman et al. 1985). Since the cytoplasmic changes by themselves have no prognostic significance, the importance of recognizing them lies in not confusing these benign processes with hyperplasia or carcinoma.

Management The management of endometrial cytoplasmic changes depends entirely on the nature of the associated proliferative process. If hyperplasia is present, it should be managed accordingly. Endometrial cytoplasmic changes without hyperplasia do not require treatment.

L. Hedrick Ellenson et al.

“carcinoma in situ” (CIS) (Spiegel 1995) and “uterine surface carcinoma” (Zheng et al. 1998), but we prefer the term SEIC because it can be associated with metastatic disease (see following), whereas the term CIS implies a lesion that does not have metastatic potential. In view of the association of SEIC with serous, as opposed to endometrioid carcinoma, it is reasonable to use the term serous EIC as has been proposed in the WHO classification. Furthermore, given its metastatic potential the latest WHO classification has included it as a type of carcinoma versus a precursor. SEIC is characterized by markedly atypical nuclei, identical to those of invasive serous carcinomas, lining the surfaces and glands of atrophic endometrium. The lesion can be very small and focal and is often present on the surface of a polyp (Figs. 47, 48, 49, 50, 51, 52, and 53) (Ambros et al. 1995; Sherman et al. 1992). SEIC often has a slightly papillary contour and some cells display hobnail morphology and smudged, hyperchromatic nuclei. The nuclei are enlarged and frequently display enlarged eosinophilic nucleoli. Numerous mitotic figures, including atypical ones, are present. On occasion, the abnormal proliferation involves only a portion of an endometrial gland (Fig. 49). More recently a lesion has been described, termed endometrial glandular dysplasia, which also exhibits cytologic atypia with serous features but lacks the marked atypia associated with SEIC (Zheng et al. 2004). It has been proposed that this lesion represents the precursor of SEIC and serous carcinoma.

Serous Endometrial Intraepithelial Carcinoma Definition and Pathologic Findings Serous carcinoma is the prototypic endometrial carcinoma that is usually not related to estrogenic stimulation and typically occurs in the setting of endometrial atrophy. Serous carcinoma is frequently associated with a putative precursor lesion, termed “serous endometrial intraepithelial carcinoma” (SEIC). The lesion also has been referred to as

Fig. 47 SEIC involving a polyp. The surface epithelium of the polyp (best seen in blunt papillary structures along upper left and middle surface) is lined by markedly atypical cells of SEIC (see Fig. 49)

8

Precursors of Endometrial Carcinoma

467

Fig. 48 SEIC involving a polyp. Higher magnification of an area of the polyp in Fig. 48 shows markedly atypical cells containing enlarged vesicular nuclei with prominent nucleoli, hobnail cells, and apoptotic bodies lining the surface and involving an underlying gland

Fig. 50 SEIC. Markedly atypical cells lining the surface epithelium have enlarged vesicular nuclei with prominent nucleoli and prominent hobnail morphology

Fig. 49 SEIC. Markedly atypical cells containing enlarged vesicular nuclei with prominent nucleoli, hobnail cells, and apoptotic bodies are lining the surface epithelium and partially involving an underlying gland

Fig. 51 SEIC. Markedly atypical cells lining endometrial glands have enlarged vesicular nuclei with prominent nucleoli, numerous mitotic figures and apoptotic bodies, and prominent hobnail morphology

Molecular Biology and Immunohistochemistry Molecular genetic evidence supports the concept that SEIC is a precursor lesion of serous carcinoma. Several studies have demonstrated immunohistochemical overexpression of p53 protein, loss of heterozygosity of chromosome 17p, and corresponding TP53 gene mutations in a high proportion of serous carcinomas and SEIC (Fig. 54) (Sherman et al. 1995; Tashiro et al. 1997). The finding of diffuse, intense staining for p53 is highly correlated with identification of

TP53 mutation in these cases. Lack of immunoreactivity for p53, however, does not exclude the presence of a mutation in TP53 because mutations have been detected in a small number of serous carcinomas that were nonreactive for p53 due to the formation of a truncated or unstable protein (Tashiro et al. 1997). Identical TP53 gene mutations have been found in SEIC and adjacent serous carcinoma in several cases. Examples of pure SEIC unassociated with serous carcinoma also have been shown to contain TP53 mutations. In addition, a case of pure SEIC has been shown to contain

468

Fig. 52 SEIC. Markedly atypical cells lining endometrial surface and underlying glands (upper, middle, and lower left) have enlarged vesicular nuclei with prominent nucleoli, distinct from the elongated nuclei in the normal glands (lower middle and middle right)

Fig. 53 SEIC. Markedly atypical cells lining endometrial glands have enlarged vesicular nuclei with prominent red nucleoli

L. Hedrick Ellenson et al.

TP53 mutation in the absence of loss of heterozygosity of chromosome 17p, suggesting that p53 mutation occurs early in the evolution of serous carcinoma (Tashiro et al. 1997). The finding of SEIC unassociated with invasive carcinoma and the presence of identical TP53 mutations in both lesions support the view that SEIC is the precursor lesion of serous carcinoma. As mentioned above, endometrial glandular dysplasia has been suggested as a precursor to SEIC, in part based on a study that has shown that these lesions show intense staining for p53, as well as TP53 mutations. Additional preliminary studies have suggested that endometrial glandular dysplasia is preceded by histologically normal lesions that demonstrate increased expression of p53 and TP53 mutations. These histologically normal appearing glands have been called “p53 signatures” because of the expression of p53 (Jarboe et al. 2009; Zhang et al. 2009). Presently, their relationship to serous carcinoma has not been definitely determined, but future studies will likely be done to further our understanding of their biologic significance in the pathogenesis of serous carcinoma. A recent study has found cyclin E amplification in 41% of SEIC, which is similar to the frequency found in endometrial serous carcinoma, suggesting that like TP53 mutations it is an early event in the pathogenesis of serous carcinoma (Kuhn et al. 2014).

Differential Diagnosis

Fig. 54 SEIC. Surface epithelium and underlying glands involved by SEIC are highlighted by diffuse/strong nuclear expression of p53; normal glands are negative

The distinction of extensive SEIC from early serous carcinoma has not been well defined. Crowded glands involved by SEIC within a polyp or within the endometrium should be classified as extensive SEIC when the proliferation lacks a confluent glandular pattern, demonstrates no evidence of stromal desmoplasia (stromal invasion), and is less than 1 cm in greatest dimension. When either glandular confluence or stromal invasion is present and the proliferation exceeds 1 cm in greatest dimension, the lesion qualifies as serous carcinoma. Lesions with glandular confluence or stromal invasion but measuring less than 1 cm can be subclassified as minimal

8

Precursors of Endometrial Carcinoma

uterine serous carcinoma (Figs. 55 and 56; see following). It is important to note, however, that metastatic serous carcinoma can be found in other sites in the genital tract and in the abdomen in the absence of demonstrable invasion in uteri with SEIC, indicating that SEIC is capable of metastasizing without first invading the stroma of the endometrium (Soslow et al. 2000; Baergen et al. 2001). SEIC must be distinguished from benign metaplastic endometrial lesions that can mimic the nuclear changes seen in SEIC, which include

Fig. 55 Extensive SEIC/minimal uterine serous carcinoma. An endometrial polyp involved by SEIC on its surface, as well as in the adjacent endometrium (left), contains crowded glands measuring less than 1 cm but verging on being confluent, suggesting early stromal invasion

469

eosinophilic cell change, hobnail change, and tubal metaplasia. At times eosinophilic cell change and hobnail change can display enlarged, smudged, hyperchromatic nuclei, but these nuclei usually have a degenerative appearance and typically lack the prominent nucleoli seen in SEIC. On occasion, however, the nuclei can appear more overtly atypical, with prominent nucleoli, suggesting SEIC (see Figs. 41 and 46). Tubal metaplasia typically displays enlarged, hyperchromatic nuclei, but these are admixed with other cell types, including ciliated cells and intercalated cells, and nucleoli are usually not prominent. Immunohistochemistry for Ki-67, a proliferation marker, is very useful for distinguishing SEIC from eosinophilic cell change and tubal metaplasia in that SEIC typically displays a very high proliferation index (virtually all the nuclei express Ki-67), whereas the metaplasias have very low proliferation indices. In addition, SEIC is usually diffusely and strongly positive for p53, whereas eosinophilic metaplasia is typically negative or occasionally displays weak or scattered moderate nuclear staining. Preliminary data based on a small number of cases indicate that tubal metaplasia and eosinophilic metaplasia do not strongly overexpress p53 (Quddus et al. 1999). Thus, the combination of Ki67 and p53 immunohistochemical stains are useful to distinguish SEIC from metaplasia when the distinction is difficult by morphologic assessment alone.

Behavior and Treatment

Fig. 56 Extensive SEIC/minimal uterine serous carcinoma. Immunohistochemical stain for p53 highlights the extent of the lesion in Fig. 55

There are limited data on the behavior of pure SEIC. One study found that patients with pure SEIC, and those with minimal uterine serous carcinoma (less than 1 cm of carcinoma in the endometrium) lacking myometrial or vascular invasion and no evidence of extrauterine disease, had an overall survival of 100% after a mean follow-up of 27 months (Wheeler et al. 2000). The majority of these patients received no treatment after hysterectomy. In addition, the few patients with involvement of endocervical glands by SEIC (stage IIA disease) were also alive without

470

evidence of disease at intervals ranging from 12 to 54 months. Similarly, in another study of stage IA serous carcinoma, 11 of 13 patients were alive without evidence of disease after a median follow-up of 38 months (Carcangiu et al. 1997). In contrast, patients with either SEIC or minimal serous carcinoma and evidence of extrauterine disease (even microscopic disease) all died of disease despite intensive chemotherapy (Wheeler et al. 2000). Accordingly, patients with a diagnosis of SEIC in an endometrial biopsy or curettage specimen should undergo careful surgical staging at the time of hysterectomy.

References Ambros RA et al (1995) Endometrial intraepithelial carcinoma: a distinctive lesion specifically associated with tumors displaying serous differentiation. Hum Pathol 26:1260–1267 Ayhan A, Yarali H, Ayhan A (1991) Endometrial carcinoma: a pathologic evaluation of 142 cases with and without associated endometrial hyperplasia. J Surg Oncol 46:182–184 Ayhan A et al (2015) Increased proliferation in atypical hyperplasia/endometrioid intraepithelial neoplasia of the endometrium with concurrent inactivation of ARID1A and PTEN tumour suppressors. J Pathol Clin Res 1:186–193 Baergen RN et al (2001) Early uterine serous carcinoma: clonal origin of extrauterine disease. Int J Gynecol Pathol 20:214–219 Beckner ME, Mori T, Silverberg SG (1985) Endometrial carcinoma: nontumor factors in prognosis. Int J Gynecol Pathol 4:131–145 Bergeron C et al (1999) A multicentric European study testing the reproducibility of the WHO classification of endometrial hyperplasia with a proposal of a simplified working classification for biopsy and curettage specimens. Am J Surg Pathol 23:1102–1108 Bokhman JV (1983) Two pathogenetic types of endometrial carcinoma. Gynecol Oncol 15:10–17 Brinton LA et al (1992) Reproductive, menstrual, and medical risk factors for endometrial cancer: results from a case-control study. Am J Obstet Gynecol 167:1317–1325 Carcangiu ML, Tan LK, Chambers JT (1997) Stage IA uterine serous carcinoma: a study of 13 cases. Am J Surg Pathol 21:1507–1514 Chandra V et al (2016) Therapeutic options for management of endometrial hyperplasia. J Gynecol Oncol 27:e8 Crum CP, Richart RM, Fenoglio CM (1981) Adenoacanthosis of endometrium: a clinicopathologic

L. Hedrick Ellenson et al. study in premenopausal women. Am J Surg Pathol 5:15–20 Dawagne MP, Silverberg SG (1982) Foam cells in endometrial carcinoma – a clinicopathologic study. Gynecol Oncol 13:67–75 Deligdisch L, Cohen CJ (1985) Histologic correlates and virulence implications of endometrial carcinoma associated with adenomatous hyperplasia. Cancer 56:1452–1455 Demopoulos RI, Greco MA (1983) Mucinous metaplasia of the endometrium: ultrastructural and histochemical characteristics. Int J Gynecol Pathol 1:383–390 Enomoto T et al (1993) Alterations of the p53 tumor suppressor gene and its association with activation of the c-K-ras-2 protooncogene in premalignant and malignant lesions of the human uterine endometrium. Cancer Res 53:1883–1888 Epplein M et al (2008) Risk of complex and atypical endometrial hyperplasia in relation to anthropometric measures and reproductive history. Am J Epidemiol 168:563–570; discussion 571–576 Esteller M et al (1999) MLH1 promoter hypermethylation is an early event in human endometrial tumorigenesis. Am J Pathol 155:1767–1772 Ferenczy A, Gelfand M (1989) The biologic significance of cytologic atypia in progestogen-treated endometrial hyperplasia. Am J Obstet Gynecol 160:126–131 Fluhmann CF (1954) Comparative studies of squamous metaplasia of the cervix uteri and endometrium. Am J Obstet Gynecol 68:1447–1463 Gal D (1986) Hormonal therapy for lesions of the endometrium. Semin Oncol 13:33–36 Greenblatt RB, Gambrell RD Jr, Stoddard LD (1982) The protective role of progesterone in the prevention of endometrial cancer. Pathol Res Pract 174:297–318 Gucer F et al (1998) Concomitant endometrial hyperplasia in patients with endometrial carcinoma. Gynecol Oncol 69:64–68 Guraslan H et al (2016) Could body mass index be an indicator for endometrial biopsy in premenopausal women with heavy menstrual bleeding? Arch Gynecol Obstet 294:395–402 Gusberg SB, Kaplan AL (1963) Precursors of corpus cancer. Iv. Adenomatous hyperplasia as stage O carcinoma of the endometrium. Am J Obstet Gynecol 87:662–678 Hall TE, Stapleton JJ, McCance JM (1982) The isolated finding of histiocytes in Papanicolaou smears from postmenopausal women. J Reprod Med 27:647–650 Hayes MP et al (2006) PIK3CA and PTEN mutations in uterine endometrioid carcinoma and complex atypical hyperplasia. Clin Cancer Res 12:5932–5935 Hecht JL et al (2005) Prediction of endometrial carcinoma by subjective endometrial intraepithelial neoplasia diagnosis. Mod Pathol 18:324–330 Hendrickson MR, Kempson RL (1980) Endometrial epithelial metaplasias: proliferations frequently misdiagnosed as adenocarcinoma. Report of 89 cases and proposed classification. Am J Surg Pathol 4:525–542

8

Precursors of Endometrial Carcinoma

Hunter JE et al (1994) The prognostic and therapeutic implications of cytologic atypia in patients with endometrial hyperplasia. Gynecol Oncol 55:66–71 Ip PP, Irving JA, McCluggage WG, Clement PB, Young RH (2013) Papillary proliferation of the endometrium: a clinicopathologic study of 59 cases of simple and complex papillae without cytologic atypia. Am J Surg Pathol. 37(2):167–77 Janicek MF, Rosenshein NB (1994) Invasive endometrial cancer in uteri resected for atypical endometrial hyperplasia. Gynecol Oncol 52:373–378 Jarboe EA et al (2009) Evidence for a latent precursor (p53 signature) that may precede serous endometrial intraepithelial carcinoma. Mod Pathol 22:345–350 Kaku T et al (1992) Endometrial metaplasia associated with endometrial carcinoma. Obstet Gynecol 80:812–816 Kaku T et al (1996) Endometrial carcinoma associated with hyperplasia. Gynecol Oncol 60:22–25 Kaminski PF, Stevens CW (1985) The value of endometrial sampling in abnormal uterine bleeding. Am J Gynecol Health II:33–36 Kendall BS et al (1998) Reproducibility of the diagnosis of endometrial hyperplasia, atypical hyperplasia, and well-differentiated carcinoma. Am J Surg Pathol 22:1012–1019 Kim YB et al (1997) Progestin alone as primary treatment of endometrial carcinoma in premenopausal women. Report of seven cases and review of the literature. Cancer 79:320–327 King A, Seraj IM, Wagner RJ (1984) Stromal invasion in endometrial adenocarcinoma. Am J Obstet Gynecol 149:10–14 Kraus FT (1985) High-risk and premalignant lesions of the endometrium. Am J Surg Pathol 9:31–40 Kuhn E, Bahadirli-Talbott A, Shih Ie M (2014) Frequent CCNE1 amplification in endometrial intraepithelial carcinoma and uterine serous carcinoma. Mod Pathol 27:1014–1019 Kurman RJ, Norris HJ (1982) Evaluation of criteria for distinguishing atypical endometrial hyperplasia from well-differentiated carcinoma. Cancer 49:2547–2559 Kurman RJ, Kaminski PF, Norris HJ (1985) The behavior of endometrial hyperplasia. A long-term study of “untreated” hyperplasia in 170 patients. Cancer 56:403–412 Kurman RJ et al (2014) WHO classification of tumors of the female reproductive organs. International Agency for Research on Cancer, Lyon Lacey JV Jr et al (2008) Risk of subsequent endometrial carcinoma associated with endometrial intraepithelial neoplasia classification of endometrial biopsies. Cancer 113:2073–2081 Lee KR, Scully RE (1989) 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 8:201–213 Lehman MB, Hart WR (2001) Simple and complex hyperplastic papillary proliferations of the endometrium: a

471 clinicopathologic study of nine cases of apparently localized papillary lesions with fibrovascular stromal cores and epithelial metaplasia. Am J Surg Pathol 25(11):1347–1354 Levine RL et al (1998) PTEN mutations and microsatellite instability in complex atypical hyperplasia, a precursor lesion to uterine endometrioid carcinoma. Cancer Res 58:3254–3258 Lidor A et al (1986) Histopathological findings in 226 women with post-menopausal uterine bleeding. Acta Obstet Gynecol Scand 65:41–43 Lin MC et al (2009) Squamous morules are functionally inert elements of premalignant endometrial neoplasia. Mod Pathol 22:167–174 Longacre TA et al (1995) Proposed criteria for the diagnosis of well-differentiated endometrial carcinoma. A diagnostic test for myoinvasion. Am J Surg Pathol 19:371–406 Masterton R, Armstrong EM, More IA (1975) The cyclical variation in the percentage of ciliated cells in the normal human endometrium. J Reprod Fertil 42:537–540 Mazur MT (1981) Atypical polypoid adenomyomas of the endometrium. Am J Surg Pathol 5:473–482 McKenney JK, Longacre TA (2009) Low-grade endometrial adenocarcinoma: a diagnostic algorithm for distinguishing atypical endometrial hyperplasia and other benign (and malignant) mimics. Adv Anat Pathol 16:1–22 Mutter GL (2000) Histopathology of genetically defined endometrial precancers. Int J Gynecol Pathol 19:301–309 Mutter GL et al (2000) Endometrial precancer diagnosis by histopathology, clonal analysis, and computerized morphometry. J Pathol 190:462–469 Nguyen TN et al (1998) Clinical significance of histiocytes in the detection of endometrial adenocarcinoma and hyperplasia. Diagn Cytopathol 19:89–93 Norris HJ, Tavassoli FA, Kurman RJ (1983) Endometrial hyperplasia and carcinoma. Diagnostic considerations. Am J Surg Pathol 7:839–847 Nucci MR et al (1999) Mucinous endometrial epithelial proliferations: a morphologic spectrum of changes with diverse clinical significance. Mod Pathol 12:1137–1142 Ordi J et al (2014) Reproducibility of current classifications of endometrial endometrioid glandular proliferations: further evidence supporting a simplified classification. Histopathology 64:284–292 Potischman N et al (1996) Case-control study of endogenous steroid hormones and endometrial cancer. J Natl Cancer Inst 88:1127–1135 Quddus MR et al (1999) p53 immunoreactivity in endometrial metaplasia with dysfunctional uterine bleeding. Histopathology 35:44–49 Randall TC, Kurman RJ (1997) Progestin treatment of atypical hyperplasia and well-differentiated carcinoma of the endometrium in women under age 40. Obstet Gynecol 90:434–440

472 Reed SD et al (2009) Progestin therapy of complex endometrial hyperplasia with and without atypia. Obstet Gynecol 113:655–662 Rorat E, Wallach RC (1984) Papillary metaplasia of the endometrium: clinical and histopathologic considerations. Obstet Gynecol 64:90S–92S Russo M et al (2017) Clonal evolution in paired endometrial intraepithelial neoplasia/atypical hyperplasia and endometrioid adenocarcinoma. Hum Pathol 67:69–77 Shah SS, Mazur MT (2008) Endometrial eosinophilic syncytial change related to breakdown: immunohistochemical evidence suggests a regressive process. Int J Gynecol Pathol 27:534–538 Sherman ME et al (1992) Uterine serous carcinoma. A morphologically diverse neoplasm with unifying clinicopathologic features. Am J Surg Pathol 16:600–610 Sherman ME, Bur ME, Kurman RJ (1995) p53 in endometrial cancer and its putative precursors: evidence for diverse pathways of tumorigenesis. Hum Pathol 26:1268–1274 Sherman ME et al (1997) Risk factors and hormone levels in patients with serous and endometrioid uterine carcinomas. Mod Pathol 10:963–968 Silver SA, Sherman ME (1998) Morphologic and immunophenotypic characterization of foam cells in endometrial lesions. Int J Gynecol Pathol 17:140–145 Silverberg SG, Kurman RJ (1992) Tumors of the uterine corpus and gestational trophoblastic disease. Atlas of tumor pathology, third series, fascicle 3. Armed Forces Institute of Pathology, Washington, DC Soslow RA, Pirog E, Isacson C (2000) Endometrial intraepithelial carcinoma with associated peritoneal carcinomatosis. Am J Surg Pathol 24:726–732 Spiegel GW (1995) Endometrial carcinoma in situ in postmenopausal women. Am J Surg Pathol 19:417–432 Tashiro H et al (1997) p53 gene mutations are common in uterine serous carcinoma and occur early in their pathogenesis. Am J Pathol 150:177–185 Tavassoli F, Kraus FT (1978) Endometrial lesions in uteri resected for atypical endometrial hyperplasia. Am J Clin Pathol 70:770–779

L. Hedrick Ellenson et al. Terakawa N et al (1997) The behavior of endometrial hyperplasia: a prospective study. Endometrial Hyperplasia Study Group. J Obstet Gynaecol Res 23:223–230 Trimble CL et al (2006) Concurrent endometrial carcinoma in women with a biopsy diagnosis of atypical endometrial hyperplasia: a Gynecologic Oncology Group study. Cancer 106:812–819 Trimble CL et al (2012) Management of endometrial precancers. Obstet Gynecol 120:1160–1175 Wheeler DT et al (2000) Minimal uterine serous carcinoma: diagnosis and clinicopathologic correlation. Am J Surg Pathol 24:797–806 Wheeler DT, Bristow RE, Kurman RJ (2007) Histologic alterations in endometrial hyperplasia and welldifferentiated carcinoma treated with progestins. Am J Surg Pathol 31:988–998 Widra EA et al (1995) Endometrial hyperplasia and the risk of carcinoma. Int J Gynecol Cancer 5:233–235 Wise MR et al (2016) Body mass index trumps age in decision for endometrial biopsy: cohort study of symptomatic premenopausal women. Am J Obstet Gynecol 215:598.e1–598.e8 Zaino RJ et al (2006) Reproducibility of the diagnosis of atypical endometrial hyperplasia: a Gynecologic Oncology Group study. Cancer 106:804–811 Zaman SS, Mazur MT (1993) Endometrial papillary syncytial change. A nonspecific alteration associated with active breakdown. Am J Clin Pathol 99:741–745 Zhang X et al (2009) Molecular identification of “latent precancers” for endometrial serous carcinoma in benign-appearing endometrium. Am J Pathol 174:2000–2006 Zheng W et al (1998) p53 immunostaining as a significant adjunct diagnostic method for uterine surface carcinoma: precursor of uterine papillary serous carcinoma. Am J Surg Pathol 22:1463–1473 Zheng W et al (2004) Endometrial glandular dysplasia: a newly defined precursor lesion of uterine papillary serous carcinoma. Part I: morphologic features. Int J Surg Pathol 12:207–223

9

Endometrial Carcinoma Lora Hedrick Ellenson, Brigitte M. Ronnett, Robert A. Soslow, Ricardo R. Lastra, and Robert J. Kurman

Contents Epidemiology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 474 Classification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 474 Etiology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 476 Hormonal Stimulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 476

L. Hedrick Ellenson (*) Department of Pathology and Laboratory Medicine, Division of Gynecologic Pathology, Weill Cornell Medical College and New York Presbyterian Hospital, New York, NY, USA e-mail: [email protected] B. M. Ronnett Department of Pathology, Division of Gynecologic Pathology, Johns Hopkins University School of Medicine, Baltimore, MD, USA e-mail: [email protected] R. A. Soslow Department of Pathology, Memorial Sloan-Kettering Cancer Center, New York, NY, USA Department of Pathology and Laboratory Medicine, Division of Gynecologic Pathology, Weill Cornell Medical College and New York Presbyterian Hospital, New York, NY, USA e-mail: [email protected] R. R. Lastra Department of Pathology, University of Chicago, Chicago, IL, USA e-mail: [email protected] R. J. Kurman Department of Gynecology, Obstetrics, Pathology and Oncology, Division of Gynecologic Pathology, Johns Hopkins University School of Medicine, Baltimore, MD, USA e-mail: [email protected] # Springer Nature Switzerland AG 2019 R. J. Kurman et al. (eds.), Blaustein’s Pathology of the Female Genital Tract, https://doi.org/10.1007/978-3-319-46334-6_9

473

474

L. Hedrick Ellenson et al. Constitutional Factors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Diet . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Molecular Genetics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Hereditary Syndromes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

477 477 477 478

Clinical and Pathologic Features of Specific Types of Carcinomas . . . . . . . . . . . . . . . . Endometrioid Carcinoma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Squamous Differentiation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Villoglandular Carcinoma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Secretory Carcinoma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Ciliated Carcinoma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Corded and Hyalinized Endometrioid Carcinoma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Mucinous Carcinoma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Serous Carcinoma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Clear Cell Carcinoma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Neuroendocrine Tumors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Mixed Cell Adenocarcinoma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Undifferentiated/Dedifferentiated Carcinoma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Carcinosarcoma (Malignant Mixed Mullerian Tumor (MMMT)) . . . . . . . . . . . . . . . . . . . . . .

480 480 495 497 498 500 500 500 503 510 513 514 514 516

Miscellaneous Epithelial Tumors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Squamous Cell Carcinoma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Glassy Cell Carcinoma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Yolk Sac Tumor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Giant Cell Carcinoma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Choriocarcinoma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Transitional Cell Carcinoma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Other Rare Variants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

521 521 521 522 522 522 522 523

Tumors Metastatic to the Endometrium . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 523 Ovarian Carcinoma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 523 Carcinomas from Extragenital Sites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 523 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 524

Epidemiology Endometrial carcinoma is the most common invasive neoplasm of the female reproductive tract and the fourth most frequently diagnosed cancer in women in the United States. In 2017, it was estimated there were 61,380 new cases and 10,920 deaths resulting from this neoplasm. Worldwide, approximately 319,600 cases are diagnosed each year, making endometrial carcinoma the sixth most common cancer in women (Torre et al. 2015). The incidence of endometrial cancer varies widely throughout the world. The highest rates occur in North America and Europe, whereas rates in developing countries are four to five times lower. The incidence of endometrial cancer in Japan has been steadily increasing since 1978 and is now the most common gynecologic malignancy in Japan (Yamagami et al. 2017). In 2012 in the United

States, the age-adjusted incidence rates of endometrial carcinoma in black and white women were similar, however since 1990 the incidence has been increasing in black women who have an 80% higher mortality rate than white women (Eheman et al. 2012; Jamison et al. 2013). Although the reason for this is not completely understood, it has been shown in a number of studies that an increase in aggressive histologic subtypes and advanced stage are contributing factors and that discrepancies in access to and quality of health care as well as genetics are also likely to play a role.

Classification Historically, endometrial carcinoma, like most other tumors, has been classified based on its light microscopic features on hematoxylin and eosin (H&E) stained tissue sections. Although molecular

9

Endometrial Carcinoma

investigation has expanded our understanding of the molecular basis of endometrial cancer, it has not, to date, been used to change the classification of endometrial carcinoma. Recent studies using high throughput analyses of DNA and RNA have suggested a molecular-based classification system. At the time of preparation of this chapter it has not been fully implemented, but many aspects of the classification system have been incorporated into diagnostic reports based on previous molecular studies. Given the speed at which molecular pathology is advancing, much of what is contained within this portion of the chapter must be viewed from the perspective of the time it was prepared. Since a landmark clinicopathological study in 1983, endometrial carcinoma has been broadly divided into two major categories, referred to as type I and type II (Bokhman 1983). It is important to recognize that these are not diagnostic categories but rather a framework for understanding the pathogenesis of endometrial carcinoma. As discussed below, factors associated with unopposed estrogenic stimulation, such as obesity and exogenous hormone use, as well as the presence of endometrial hyperplasia, are related to the development of the most common form of endometrial carcinoma, the endometrioid subtype, which is the prototype of type I carcinoma (Bokhman 1983). More recent studies have confirmed this association by demonstrating elevated serum estrogen levels in patients with endometrioid carcinoma (Lukanova et al. 2004). It also has been recognized that some forms of endometrial carcinoma appear to be largely unrelated to hormonal factors and hyperplasia (Sherman et al. 1997). Serous carcinoma is the most common form of endometrial carcinoma that is not usually related to estrogenic stimulation and is the prototypic type II carcinoma. Molecular genetic studies, as discussed below, have provided further support for the dualistic categorization by identifying significant molecular genetic differences between the two most common types, endometrioid and serous carcinoma. Most of the other histologic subtypes of endometrial carcinoma, with exceptions as discussed below, can be classified as variants of either type I or II on the basis of clinicopathologic, immunohistochemical, and molecular features. Thus, other low-grade

475

carcinomas, which are associated with endometrial hyperplasia and estrogenic stimulation, such as secretory, villoglandular, or low-grade endometrioid with squamous differentiation, are type I carcinomas. The exceptions to this are clear cell carcinoma and some International Federation of Gynecology and Obstetrics (FIGO) grade 3 endometrioid tumors. Over the years, clinicopathological studies of clear cell carcinoma have produced variable results. While most studies found that clear cell carcinomas were aggressive tumors, other studies demonstrated a more indolent behavior. Although most clear cell carcinomas have distinctive features, it has been recognized that in some instances the morphologic features are ambiguous. An early molecular study suggested that the genetic alterations of clear cell carcinoma were heterogenous with some sharing alterations with serous carcinoma, others with endometrioid tumors, and another group which did not overlap with either (An et al. 2004). More recent next-generation studies have confirmed the molecular heterogeneity of clear cell carcinoma (discussed in more detail in the section on “Clear Cell Carcinoma”) (DeLair et al. 2017). In addition, early molecular and clinicopathological studies suggested that grade 3 endometrioid tumors might be best categorized as type II tumors. This is an area under current investigation and at present it seems that, like clear cell carcinoma, they are a heterogenous group of tumors (Bosse et al. 2018). This categorization of endometrial carcinoma has been challenged by some based on the recent high-throughput molecular studies. However, the broad categorization takes into account additional etiologic factors of endometrial carcinoma that have not yet been thoroughly addressed with molecular analyses alone. In the future, high throughput RNA analyses, proteomics, and metabolomics integrated with DNA analysis will undoubtedly provide more comprehensive information about additional etiologic factors. Presently, however, the recent next-generation sequence analyses have suggested that endometrioid and serous endometrial carcinomas can be divided into four distinct molecular subtypes (Kandoth et al. 2013). Interestingly, the subtypes correspond to a large extent with type I and II tumors, as is discussed in detail below.

476 Table 1 Classification of endometrial carcinomaa Endometrioid adenocarcinoma Squamous differentiation Villoglandular Secretory Mucinous carcinoma Serous carcinoma Clear cell carcinoma Neuroendocrine tumors Low-grade neuroendocrine tumor Carcinoid tumor High-grade neuroendocrine carcinoma Small cell neuroendocrine carcinoma Large cell neuroendocrine carcinoma Mixed cell adenocarcinoma Undifferentiated carcinoma Dedifferentiated carcinoma a

Modified World Health Organization and International Society of Gynecological Pathologists Histologic Classification of Endometrial Carcinoma.

A modified version of the recent World Health Organization (WHO) and International Society of Gynecological Pathologists (ISGYP) classification of endometrial carcinoma is shown in Table 1.

Etiology Hormonal Stimulation The strong association between replacement estrogen therapy and the development of endometrial cancer was demonstrated in a number of casecontrol studies in the late 1970s that have been supported by more recent studies (Gray et al. 1977; Greenwald et al. 1977; Mack et al. 1976; McDonald et al. 1977; Shapiro et al. 1980; Smith et al. 1975; Ziel and Finkle 1975). A study of endogenous hormones and endometrial cancer demonstrated that the risk associated with elevated levels of unopposed estrogen varies according to menopausal status (Potischman et al. 1996). In particular, high estrone and albumin-bound estradiol levels were associated with increased risk in postmenopausal women, but high levels of total, free, and albumin-bound

L. Hedrick Ellenson et al.

estradiol were unrelated to increased risk in premenopausal women. In addition, high circulating levels of androstenedione were identified as a risk factor in both pre- and postmenopausal women. Factors that lower the risk of endometrial cancer include the addition of progestin to hormone replacement regimens, the use of oral contraceptives, and smoking (Jama 1987; Austin et al. 1993; Beral et al. 1999; Franks et al. 1987; Kaufman et al. 1980; Lesko et al. 1985; Pickar et al. 1998; Weir et al. 1994; Brinton and Felix 2014). It has been shown that women using unopposed estrogen for more than 2 years have a two- to three-fold increase in the risk of endometrial cancer, whereas women receiving progestins in conjunction with estrogen have no increased risk (Persson et al. 1989). One large case-control study demonstrated that the use of oral contraceptives for at least 1 year reduces the risk of endometrial carcinoma by 50% and that protection persists at least 15 years after discontinuation (Jama 1987). The risk of endometrial carcinoma may also be affected by polymorphisms in the estrogen receptor (ER) genes, however, the mechanism responsible is not currently understood (Ashton et al. 2009). Tamoxifen is a nonsteroidal compound that acts by competing with estrogen for ER. In reproductive age women it has an antiestrogenic effect, but in postmenopausal (hypoestrogenic) women it has weak estrogenic effects and as a result significantly increases the risk of endometrial cancer (Andersson et al. 1991; Boccardo et al. 1992; Cook et al. 1995; Fisher et al. 1994, 1998; Fornander et al. 1989; Katase et al. 1998; Ribeiro and Swindell 1992; Rutqvist et al. 1995; Ryden et al. 1992; van Leeuwen et al. 1994; Stewart 1992). In addition, some studies have reported a higher proportion of high-risk types of carcinomas in tamoxifen-treated women whereas others have found predominantly low-grade carcinomas (Fisher et al. 1998; Barakat et al. 1994; Silva et al. 1994; Curtis et al. 2004). Despite the risk of endometrial carcinoma, tamoxifen remains a mainstay of treatment for prevention of breast cancer recurrence. Conversely, although it has been recently reported that oral contraceptives increase the risk of breast

9

Endometrial Carcinoma

carcinoma, their beneficial effects in substantially reducing the risk of endometrial and ovarian cancer, not to mention their value as highly effective contraceptive agents, merit their continued use (Morch et al. 2017).

Constitutional Factors Obesity, like estrogen replacement therapy, is a well-defined risk factor for endometrial cancer, (Voskuil et al. 2007) with reported relative risks ranging from 2 to 10 (Parazzini et al. 1991; Parazzini et al. 1997; Onstad et al. 2016). The worldwide increase in obesity is thought to be related to the increase in endometrial carcinoma, as greater than half of the cases of endometrial carcinoma are associated with obesity. The risk can be explained by the increase of estrogens from aromatization of androgens to estrogens in adipose tissue and lower concentrations of sex hormonebinding globulins in obese women (Enriori and Reforzo-Membrives 1984). Diabetes is associated with an increased risk of endometrial cancer, ranging from 1.2 to 2.1, and this risk appears to be independent of other frequently associated variables such as obesity (Parazzini et al. 1991; Parazzini et al. 1997; Brinton et al. 1992). Other factors that have been associated with an increased risk of endometrial cancer include early age of menarche, later age of menopause, and nulliparity. The association with nulliparity appears to be primarily on the basis of infertility due to chronic anovulation in which unopposed estrogenic stimulation occurs (Brinton et al. 1992). The protective effect of pregnancy appears to be related and restricted to the first full-term pregnancy because abortions and increasing numbers of births do not influence the risk.

Diet Endometrial cancer risk is correlated with total caloric intake, total protein intake, and frequency of consumption of meat, eggs, milk, fats, and oils. These dietary factors, as well as decreased energy expenditure and physical exercise

477

associated with a sedentary lifestyle, are major determinants of obesity, which is an established risk factor. The independent contribution of specific dietary factors to endometrial cancer risk has not been clearly established (Parazzini et al. 1991; Levi et al. 1993). More recent studies suggest that activity decreases the risk of endometrial cancer independent of body weight (Voskuil et al. 2007).

Molecular Genetics Advances in the field of molecular biology and bioinformatics have provided novel information on the mutational landscape of endometrial carcinoma. This information is critical to our understanding of endometrial carcinoma and, thus, to improvements in diagnosis, patient management, and prevention. Although many of the common tumor suppressor genes, oncogenes, and mutator genes involved in the pathogenesis of endometrial carcinoma were identified prior to next-generation sequencing, the vast amount of data being generated and analyzed will not only affect our understanding of endometrial cancer but will undoubtedly have a significant impact on the classification of endometrial carcinoma. The most recent next-generation sequencing studies suggest that endometrioid and serous carcinoma can be classified into four major molecular subtypes. Studies on the more uncommon types of endometrial carcinoma are ongoing and will determine if they also fall into these molecular subtypes. The integration of these subgroups into diagnostic practice is currently under investigation. However, there are a number of molecular studies in routine use that place tumors into one of these molecular subtypes. Given the pace of personalized medicine, it is possible that in the not too distant future classification systems may become obsolete as each individual’s tumor will be classified based on its unique molecular alterations. Below, the four molecular subgroups will be described and the diagnostic aspects of the molecular studies will be presented in the subsequent sections on each specific WHO-classified tumor type.

478

Ultramutated Subtype This subtype of carcinoma is defined by an extremely high mutation rate, among the highest rates reported for any tumor type, with individual tumors typically demonstrating greater than 10,000 mutations. The high mutation rate is caused by mutations in the exonuclease domain of the POLE gene that encodes the central catalytic subunit of DNA polymerase epsilon. The mutations lead to a dysfunctional holoenzyme that results in lack of DNA repair during replication. Mutations in POLE have been described in approximately 5.6–6.5% of endometrial carcinomas (Kandoth et al. 2013; Billingsley et al. 2015). Although the majority of endometrial carcinomas exhibiting POLE mutations demonstrate endometrioid histology, the mutations have been found in undifferentiated and dedifferentiated carcinomas, and in tumors with ambiguous histology (Haruma et al. 2018) (Espinosa et al. 2017) (Hoang et al. 2017). The clinical importance of this molecular subtype stems from a number of studies that have shown an association with improved survival (Bosse et al. 2018; Kandoth et al. 2013). However, the majority of the studies have not shown statistically significant associations and one study did not show an association with favorable outcome (Billingsley et al. 2015). Given the lack of a clear association with reproducible histology, identification of tumors with POLE mutations requires molecular analysis and definitive recommendations for altering patient management await larger studies.

Hypermutated/Microsatellite Instability (MSI) Subtype This group of tumors is also characterized by an elevated mutation rate, albeit not as high as that of the ultramutated subtype. It was initially detected due to the presence of alterations in the length of microsatellite DNA sequences, hence the designation as the MSI subtype. The underlying molecular abnormality is loss of DNA mismatch repair (MMR), a post-replicative repair mechanism that predominately repairs mismatched base-pairs due to strand slippage in areas of nucleotide repeats. The loss of this repair

L. Hedrick Ellenson et al.

mechanism results in base-pair substitutions and small insertions and deletions that are characteristic of loss of DNA MMR. With the advent of high-throughput DNA sequence analysis, a number of target genes in endometrial cancer have been discovered due to the characteristic type of mutations. In addition, there are a number of genes, although not clearly a target of mismatch repair, commonly mutated in this subtype as is discussed below. This subtype consists of endometrioid carcinoma and is found in approximately 20–25% of sporadic cases. In addition, it is a molecular phenotype found in tumors arising in the setting of Lynch Syndrome. This is discussed in more detail in other sections of the chapter.

Copy Number Low/Microsatellite Stable (MSS) Subtype This subtype, as the name implies, lacks abnormalities in DNA MMR and significant copy number alterations. It does, however, have frequent mutations in PTEN, PIK3CA, PIK3R1, ARID1A, and CTNNB1 (beta-catenin) similar to the MSI subtype, although the frequency of CTNNB1 mutations is higher in this subtype. Like the MSI subtype, this group is composed entirely of endometrioid tumors. Copy Number High/Serous-Like Subtype One of the main characteristics of this subtype beyond the high level of copy number abnormalities is the extremely high frequency of TP53 mutations. These occur in over 90% of tumors in this subtype. As expected, this subtype consists predominately of serous tumors along with some grade 3 endometrioid carcinomas but also includes some clear cell carcinomas and carcinosarcomas.

Hereditary Syndromes Lynch Syndrome Lynch syndrome is the most common cause of familial endometrial carcinoma. It is due to germline transmission of defective DNA

9

Endometrial Carcinoma

MMR genes (MSH2, MLH1, MSH6, and PMS2) resulting in an autosomal dominant inheritance pattern. As described above, mutations in DNA MMR genes result in the molecular phenotype of MSI, which results in an increased rate of mutations in cancer-causing genes, thus predisposing affected individuals to the development of various cancers. Endometrial carcinoma is an integral part of Lynch syndrome, and women with endometrial carcinoma may be probands for affected families. Up to one-third of endometrioid carcinomas demonstrate abnormal DNA MMR protein expression (Modica et al. 2007; Vasen et al. 2004; Peiro et al. 2002; de Leeuw et al. 2000). This results from MLH1 promoter hypermethylation in most cases or mutation of MLH1, MSH2, MSH6, or PMS2 in the remaining. Mutation, but not loss of expression alone, of one of these genes indicates that the affected patient may be part of a Lynch syndrome kindred. Therefore, DNA MMR protein immunohistochemistry serves as a screen for Lynch syndrome; it is not a diagnostic test. For practical purposes, loss of expression of MSH2 and/or MSH6 or PMS2 is considered a surrogate for the presence of a somatic or germline (signifying Lynch syndrome) mutation involving one of the corresponding genes, whereas loss of expression of MLH1 and PMS2 is more likely associated with an epigenetic (promoter methylation of MLH1) etiology unassociated with Lynch syndrome. Presently, it is recommended that all newly diagnosed cases of endometrial carcinoma be screened for loss of DNA MMR using an immunohistochemical approach (Anagnostopoulos et al. 2017; Watkins et al. 2017; Mills and Longacre 2016). However, there remains some controversy on whether MSI analysis should accompany the immunohistochemical screening as some studies have shown that immunohistochemistry alone may miss a small fraction of cases with microsatellite instability (Mills and Longacre 2016). Carcinomas which demonstrate loss of expression of MLH1 and PMS2 are submitted for MLH1 methylation analysis. If methylation is identified, the MSI phenotype

479

is considered the result of a somatic alteration resulting in loss of MLH1. However, if methylation is absent or there is loss of MSH2 and/or MSH6 or PMS2 alone, the patients should be referred for a comprehensive genetic evaluation for Lynch syndrome. Loss of expression tends to occur in couplets (MLH1 with PMS2 and MSH2 with MSH6), although examples of isolated PMS2 loss (without MLH1) or MSH6 loss (without MSH2) are on record. Only complete loss of expression in the setting of a valid positive internal control is considered interpretable. Valid internal controls include non-neoplastic endometrial stroma and glands with reproducibly stained nuclei. Care should be taken to ensure that the lesion being assessed is carcinoma, not hyperplasia. It is also extremely important that the immunohistochemical methodology and interpretation of stains be performed using the strictest guidelines, as performing and interpreting the MLH1 stain, in particular, can be very problematic. Inappropriately interpreting an MLH1 stain as negative rather than as technically unsatisfactory in the absence of a valid positive internal control is a rather common occurrence.

Cowden Syndrome Cowden syndrome is an autosomal dominant disorder caused by mutations in the PTEN tumor suppressor gene and is defined by a number of benign conditions and an increase in the risk of malignancies of the breast, thyroid, and endometrium. The lifetime risk of endometrial carcinoma in women with Cowden syndrome is estimated to be between 5–10% versus 2.6% in the general population. The syndrome is recognized in approximately 1 in 200,000 individuals and the histologic type of endometrial carcinoma has not been described (Nelen et al. 1999). As discussed above, given the increase in lifetime risk, it is currently recommended that women with Cowden syndrome be screened for endometrial carcinoma with blind biopsies annually starting at 35–40 years of age or 5 years prior to the earliest diagnosis of endometrial carcinoma in the family and with annual endometrial ultrasound in postmenopausal women.

480

Clinical and Pathologic Features of Specific Types of Carcinomas Endometrioid Carcinoma Endometrioid carcinoma is the most common form of endometrial carcinoma, accounting for more than three-fourths of all cases. These tumors are referred to as endometrioid because they resemble proliferative-phase endometrium and to maintain consistency with the terminology used for describing tumors with the same histologic appearance in the cervix, ovary, and fallopian tube.

Clinical Features Patients with endometrioid carcinoma range in age from the second to the eighth decade, with a mean age of 59 years. Most women are postmenopausal, as the disease is relatively uncommon in young women. Only 1–8% of endometrial carcinomas occur in women under 40 years (Crissman et al. 1981; Dockerty et al. 1951; Gitsch et al. 1995a; Ross et al. 1983; Peterson 1968). A small number of cases have been reported in women under the age of 30 years, the youngest being 14 years with Cowden syndrome (Farhi et al. 1986; Lee and Scully 1989; Baker et al. 2013). In young women, the tumor is generally low grade and minimally invasive. In most series, the majority of patients have had clinical evidence of polycystic ovary syndrome (irregular menses, infertility, obesity, or hirsutism) but in some reports the patients lacked these features. Rarely, endometrioid carcinoma occurs during pregnancy (Hoffman et al. 1989). In pregnant women, endometrial carcinomas are nearly always low grade, superficially invasive or noninvasive, and have an excellent prognosis. The initial manifestation of endometrial carcinoma is typically abnormal vaginal bleeding, although rarely the patient is asymptomatic and the diagnosis is made fortuitously. In one study, 24 asymptomatic women with unsuspected endometrial carcinoma were detected among 8998 women dying of unrelated causes who were autopsied at the Yale–New Haven and Massachusetts General Hospitals (Horwitz et al. 1981). The

L. Hedrick Ellenson et al.

estimated rates of undetected endometrial carcinoma were 22 and 31 per 10,000, respectively. These rates were four to five times higher than the diagnosis of endometrial carcinoma recorded by the Connecticut State Tumor Registry, indicating that a number of endometrial carcinomas may be asymptomatic and are undetected during life. A number of studies have evaluated cytologic screening of endometrial cancer. The most recent studies support that the finding of atypical glandular cells on Pap test should trigger endocervical and endometrial sampling, especially in women over 50 years of age. One recent study of 554 women with endometrial carcinoma and a liquid-based Pap test within 36 months prior to the histologic diagnosis found that 38% had abnormal glandular cells on Pap and 6.2% had only benign endometrial cells in women 40 years or older. The detection of abnormal glandular cells correlated with tumor size, tumor type, higher FIGO stage, and lymph-vascular invasion (Serdy et al. 2016). However, the study did not comment on whether the women had vaginal bleeding (i.e., were symptomatic) at the time of the Pap test. In sum, the Pap test remains an insensitive method for the detection of endometrial carcinoma and cytologic detection methods in symptomatic women are of little value as women with abnormal vaginal bleeding are evaluated by either endometrial biopsy or curettage, which yields a more easily interpreted specimen. More recently, molecular analyses to detect endometrial and ovarian cancer have been applied to DNA isolated from thin-prep Pap samples and this is an active area of current research (Wang et al. 2018).

Gross Findings The gross appearance of endometrioid carcinoma is similar to the various other types of endometrial carcinoma with the possible exception of serous carcinoma or carcinosarcoma (see “Serous Carcinoma” and “Carcinosarcoma” (Malignant Mixed Mullerian Tumor). The endometrial surface is shaggy, glistening, and tan and may be focally hemorrhagic. Endometrioid carcinoma is almost uniformly exophytic even when deeply invasive. The neoplasm may be focal or diffuse but at times

9

Endometrial Carcinoma

may be composed of separate polypoid masses. Necrosis usually is not evident macroscopically in well-differentiated carcinomas but may be seen in poorly differentiated tumors, sometimes in association with ulcerated or firm areas. Myometrial invasion by carcinoma may result in enlargement of the uterus, but a small atrophic uterus may harbor carcinoma diffusely invading the myometrium. Myometrial invasion usually appears as well-demarcated, firm, gray-white tissue with linear extensions beneath an exophytic mass or as multiple, white nodules with yellow areas of necrosis within the uterine wall. However, some cases of well differentiated carcinoma may show extensive myometrial invasion in the absence of a grossly identifiable invasive component. Extension into the lower uterine segment is common, whereas involvement of the cervix occurs in approximately 20% of cases.

Microscopic Findings: Grading The grade of endometrioid carcinoma is determined by the microscopic appearance of the tumor. It is based on the architectural pattern, nuclear features, or both (Figs. 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, and 16). The architectural grade is determined by the extent to which the tumor is composed of solid masses of cells as compared with well-defined glands (Table 2) (Figs. 1, 3, 4, 5, 6, and 7, 10, 12, 13,

Fig. 1 Endometrioid carcinoma, FIGO grade 1 (architectural grade 1, nuclear grade 1). Well-differentiated endometrioid glands are interconnected in a confluent glandular pattern with surrounding desmoplastic stroma; these features indicate endometrial stromal invasion by carcinoma

481

15). In endometrioid carcinomas with squamous differentiation, it is important to exclude masses of squamous epithelium in determining the amount of solid growth. If the areas of squamous differentiation are non-keratinizing, it may not be possible to distinguish them from solid growth. It has been suggested that if the nuclear features in the solid areas are similar to those seen in the glandular component of the tumor they are best considered non-squamous, solid tumor growth. The nuclear grade is determined by the variation in nuclear size and shape, chromatin distribution, and size of the nucleoli. Grade 1 nuclei are oval, mildly enlarged, and have evenly dispersed chromatin (Figs. 2 and 8). Grade 3 nuclei are markedly enlarged and pleomorphic, with irregular, coarse chromatin and prominent eosinophilic nucleoli (Fig. 14). Grade 2 nuclei have features intermediate to grades 1 and 3 (Figs. 11 and 16). Mitotic activity is an independent histologic variable, but it is generally increased with increasing nuclear grade, as are abnormal mitotic figures. The most recent revision of the FIGO Staging System (Table 3) and the WHO Histopathologic Classification of uterine carcinoma recommend that tumors be graded using both architectural and nuclear criteria (Scully et al. 1994; Creasman 1989). The grade of tumors that are architecturally grade 1 or 2 should be increased by one grade in the presence of

Fig. 2 Endometrioid carcinoma, FIGO grade 1 (architectural grade 1, nuclear grade 1). Well-formed endometrioid glands have small, round to oval nuclei with uniform chromatin

482

L. Hedrick Ellenson et al.

Fig. 3 Endometrioid carcinoma, FIGO grade 1 (architectural grade 1, nuclear grade 1). Endometrioid glandular epithelium is interconnected in a confluent glandular fashion, a pattern indicating endometrial stromal invasion by carcinoma

Fig. 5 Endometrioid carcinoma, FIGO grade 1 (architectural grade 1, nuclear grade 2). Well-differentiated endometrioid glands exhibit cribriform growth, a pattern indicating endometrial stromal invasion by carcinoma

Fig. 4 Endometrioid carcinoma, FIGO grade 1 (architectural grade 1, nuclear grade 1). Well-differentiated endometrioid glands are back-to-back with foci of gland fusion. The latter is indicative of carcinoma

Fig. 6 Endometrioid carcinoma, FIGO grade 1 (architectural grade 1, nuclear grade 1). Back-to-back and fused glands are consistent with carcinoma

“notable” nuclear atypia (grade 3 nuclei) involving greater than 50% of the tumor (Zaino et al. 1995). For example, a tumor that is grade 2 by architecture but in which there is marked nuclear atypia (nuclear grade 3) should be upgraded to grade 3. Thus, tumors are graded primarily by their architecture, with the overall grade modified by the nuclear grade when there is discordance. Marked discordance between nuclear and architectural grade is unusual in endometrioid carcinoma and should raise suspicion that the tumor is a serous carcinoma (see “Serous Carcinoma”).

Given the importance of tumor grading in patient outcome, and consequently its significant role in clinical decision making, appropriate interobserver reproducibility is necessary in utilized grading schemes. Multiple studies have shown that the interobserver reproducibility of the FIGO grading method for endometrioid carcinomas based on architecture is acceptable, but have shown poor reproducibility when grading is based on nuclear features (Lax et al. 2000a; Nielsen et al. 1991). Nonetheless, it has been shown that upgrading architecturally grade 1 or 2 tumors based on nuclear features resulted in their reclassification into a higher grade category with similar risk of recurrence and

9

Endometrial Carcinoma

483

Fig. 7 Endometrioid carcinoma, FIGO grade 1 (architectural grade 1, nuclear grade 1). Back-to-back and fused well-differentiated glands have mucinous features

Fig. 9 Endometrioid carcinoma, FIGO grade 1 (architectural grade 1, nuclear grade 2). Nuclei are somewhat enlarged, rounded, and have granular to vesicular chromatin with occasional small nucleoli

Fig. 8 Endometrioid carcinoma, FIGO grade 1 (architectural grade 1, nuclear grade 1). Nuclei are round to oval with uniform chromatin

Fig. 10 Endometrioid carcinoma, FIGO grade 2 (architectural grade 2, nuclear grade 1). Well-formed glands are admixed with solid non-squamous nests of tumor, with the latter comprising more than 5% but less than 50% of the overall tumor

death, reinforcing the need for a uniform definition of nuclear atypia (Zaino et al. 1995). Marked differences in architectural grade can be seen within a tumor. It is not unusual to see wellformed glandular elements immediately adjacent to solid endometrioid areas. When a tumor displays this type of heterogeneity, the architectural grade should be based on the overall appearance. The heterogeneity in differentiation accounts for the differences in grade that can be observed between the endometrial curettings and the hysterectomy specimen. Discordance between the curettage and hysterectomy specimens occurs in 15–25% of cases (Daniel and Peters 1988; Larson et al. 1995; Obermair et al. 1999).

Myoinvasion Endometrial carcinoma may manifest different forms of myometrial invasion (Figs. 17 and 18). It can invade along a broad pushing front or it can infiltrate the myometrium diffusely as masses, cords, or clusters of cells and individual glands. When it invades along a broad front it may be difficult to determine whether invasion is, in fact, present unless it can be compared to the adjacent uninvolved endomyometrium. When the tumor diffusely invades the myometrium, the neoplastic glands usually elicit a reactive stromal response characterized by loose fibrous tissue accompanied

484

Fig. 11 Endometrioid carcinoma, FIGO grade 2 (architectural grade 2, nuclear grade 1). Glandular and solid areas have generally uniform small, round to oval nuclei with granular chromatin

Fig. 12 Endometrioid carcinoma, FIGO grade 2 (architectural grade 2, nuclear grade 2). Tumor is composed of intimately admixed glandular and solid non-squamous epithelium within an edematous and inflamed altered stroma

by a chronic inflammatory infiltrate that surrounds the glands. Occasionally, well-differentiated carcinomas may be deeply invasive with glands directly in contact with surrounding myometrium in the absence of a stromal response (diffusely infiltrative or adenoma malignum pattern of invasion) (Mai et al. 2002) (Longacre and Hendrickson 1999). In these cases, when myometrial invasion is superficial the presence of invasion can be identified if a haphazard glandular arrangement is present. Usually this pattern of invasion is found in deeply invasive tumors, however, and therefore recognizing myometrial

L. Hedrick Ellenson et al.

Fig. 13 Endometrioid carcinoma, FIGO grade 3 (architectural grade 3, nuclear grade 3). A few residual glandular structures are present within an otherwise solid non-squamous (>50%) tumor with areas of necrosis

Fig. 14 Endometrioid carcinoma, FIGO grade 3 (architectural grade 3, nuclear grade 3). Solid non-squamous tumor with foci of necrosis and rare residual glandular lumens displays notable nuclear atypia characterized by nuclear enlargement and pleomorphism with vesicular chromatin and prominent nucleoli

invasion is not a problem. Endometrioid carcinomas with the diffusely infiltrative pattern of invasion share the same prognostic indicators of clinically aggressive disease as those having the more conventional pattern of myometrial invasion (Longacre and Hendrickson 1999). An unusual form of myoinvasion has been described that consists of outpouching of neoplastic glands that become detached and may be lined by flattened epithelium sometimes appearing as microcysts which is associated with a fibromyxoid stromal reaction (Fig. 18). This type of invasion has been

9

Endometrial Carcinoma

485 Table 3 International Federation of Gynecology and Obstetrics Staging of Endometrial Cancer, 2009

Fig. 15 Endometrioid carcinoma, FIGO grade 3 (architectural grade 3, nuclear grade 2). Tumor is composed of solid non-squamous epithelium with areas of necrosis

IA

G123

IB

G123

II IIIA IIIB IIIC1 IIIC2 IVA

G123 G123 G123 G123 G123 G123

IVB

G123

Tumor limited to the endometrium or the inner half of myometrium Tumor invasion into the outer half of myometrium Tumor invades cervical stromaa Tumor invades serosa and/or adnexab Vaginal and/or parametrial invasion Metastases to pelvic lymph nodes Metastases to para-aortic lymph nodes Tumor invasion of bladder and/or bowel mucosa Distant metastases including intraabdominal and/or inguinal lymph nodes

G1, 5% or less of a non-squamous or nonmorular solid growth pattern; G2, 6–50% of a non-squamous or nonmorular solid growth pattern; G3, more than 50% of a non-squamous or nonmorular solid growth pattern Rules on staging: 1. Corpus cancer is now surgically staged. Those patients who do not undergo a surgical procedure should be staged according to the 1971 FIGO clinical staging. 2. Ideally, the thickness of the myometrium should be measured along with the depth of tumor invasion Notes on grading: 1. Notable nuclear atypia, inappropriate for the architectural grade, raises a grade 1 or grade 2 tumor by one. 2. In serous adenocarcinomas, clear cell adenocarcinomas, and squamous cell carcinomas, nuclear grading takes precedence. 3. Adenocarcinomas with squamous differentiation are graded according to the nuclear grade of the glandular component a Endocervical gland involvement should be considered Stage I b Positive peritoneal fluid cytology should be reported separately, but does not affect the stage

Fig. 16 Endometrioid carcinoma, FIGO grade 3 (architectural grade 3, nuclear grade 2). Nuclei are only modestly pleomorphic, with vesicular chromatin and numerous mitotic figures. Spaces consistent with residual gland lumens favor endometrioid rather than undifferentiated carcinoma and clear cytoplasmic change in the absence of any other characteristic features of clear cell carcinoma is insufficient to diagnose the latter

Table 2 Architectural grading of endometrial carcinoma Grade 1 Grade 2 Grade 3

No more than 5% of the tumor is composed of solid masses 6–50% of the tumor is composed of solid masses More than 50% of the tumor is composed of solid masses

Fig. 17 Endometrioid carcinoma, FIGO grade 1, myoinvasive. Myometrial invasion by carcinoma is characterized by islands of well-differentiated glands surrounded by smooth muscle

486

Fig. 18 Endometrioid carcinoma, FIGO grade 1, myoinvasive. Some endometrioid carcinomas lose their characteristic columnar endometrioid features and invade myometrium insidiously as attenuated and dilated glands, often intimately associated with an inflammatory reaction

Fig. 19 Endometrioid carcinoma involving adenomyosis. FIGO grade 1 carcinoma is present within an island of adenomyosis, identified by the residual benign endometrial glands and stroma along the periphery of the island. Uninvolved adenomyosis is also present

termed “microcystic, elongated, and fragmented (MELF)” (Murray et al. 2003). MELF invasion is associated with lymph-vascular invasion, but not with poor prognosis in a multivariate analysis (Euscher et al. 2013). It may be difficult to distinguish myometrial invasion from extension of the carcinoma into adenomyosis (Fig. 19). The distinction, however, is important because the presence of carcinoma in adenomyosis deeper than the maximum depth of true tumor invasion does not worsen the prognosis

L. Hedrick Ellenson et al.

Fig. 20 Endometrioid carcinoma, FIGO grade 1, noninvasive. Nests of carcinoma along an irregular endomyometrial junction suggest superficial myometrial invasion but preserved benign endometrial glands at the periphery of two nests indicate the tumor is still confined to the endometrium

(Hall et al. 1984; Hernandez and Woodruff 1980; Jacques and Lawrence 1990; Mittal and Barwick 1993). When the carcinoma is surrounded by endometrial stroma and residual benign glands are present in these foci, the diagnosis of carcinoma extending into adenomyosis is straightforward. At times, however, the distinction from myometrial invasion may be extremely difficult, particularly in older women in whom adenomyosis may have very minimal stroma as a result of fibrosis and atrophy. In these cases, it is necessary to evaluate additional features such as the presence of desmoplasia, surrounding edema and inflammation, and the shape of the glands (Jacques and Lawrence 1990). In contrast to carcinoma involving adenomyosis, true myometrial invasion is usually characterized by desmoplasia or loosening of the myometrium surrounding the glands. Often there is accompanying chronic inflammation and the glandular outline is jagged and irregular, as compared to carcinoma involving adenomyosis in which the glands have a smooth, rounded outline and desmoplasia and inflammation are lacking (Fig. 20). Since CD10 is normally expressed by endometrial stromal cells but not smooth muscle of the myometrium, it would seem that the presence of CD10 in the cells around an adenocarcinoma in the myometrium would indicate its presence in adenomyosis. Unfortunately, CD10 is

9

Endometrial Carcinoma

also often (52% of cases) expressed focally in the cells surrounding clusters of tumor in the myometrium of women in whom adenomyosis is absent, thus eliminating its utility (Nascimento et al. 2003; Srodon et al. 2003). A diagnosis of carcinoma involving adenomyosis should be made only when there is evidence of adenomyosis uninvolved by carcinoma or residual adenomyosis within foci involved by carcinoma in the uterus because some endometrioid carcinomas invade the myometrium without eliciting a stromal response. A recent study noted that adenocarcinoma involving adenomyosis frequently is associated with preceding estrogen use, low tumor grade, and an excellent prognosis (Mittal and Barwick 1993). Diagnosis of superficial myometrial invasion is often problematic due to irregularity of the normal endomyometrial junction, particularly in older women. The presence of residual non-neoplastic endometrial glands and stroma along the deep or peripheral aspect of rounded nests of carcinoma situated at the irregular endomyometrial junction is evidence that these nests are still within the endometrium proper and have not invaded superficial myometrium (Fig. 20). Fortunately, the 2009 FIGO staging eliminated the need to distinguish between superficially invasive and noninvasive tumors as they are both considered stage IA.

Differential Diagnosis The main problem in the differential diagnosis of low-grade endometrioid carcinoma is the distinction from atypical hyperplasia, atypical polypoid adenomyoma, hyperplasia with various types of cytoplasmic alterations (metaplasias), AriasStella reaction, and menstrual endometrium. The distinction from the first three conditions is discussed in ▶ Chap. 8, “Precursors of Endometrial Carcinoma.” At times, an extremely atypical Arias-Stella reaction may simulate adenocarcinoma. In the reproductive age group, Arias–Stella reaction is much more likely than carcinoma, especially if the clinical history indicates a recent pregnancy. Nonetheless, carcinoma can occur in young women and also in pregnancy. In contrast to a carcinoma, the Arias–Stella reaction tends to be multifocal and is admixed with secretory glands and decidua. The glands in the

487

Arias–Stella reaction may be complex and tortuous but lack confluent or papillary patterns. The stroma does not show a desmoplastic response. The nuclei in the glandular epithelium of the Arias–Stella reaction may be markedly enlarged, but the chromatin appears degenerated and smudged and mitotic figures are very unusual. Menstrual endometrium can be confused with adenocarcinoma because of the extensive tissue breakdown characterized by tissue fragmentation and hemorrhage. The pattern of stromal breakdown results in fragmented glands of varying size and compact clusters of stromal cells haphazardly mixed with blood, which can appear ominous. The glandular epithelium, however, is bland and shows evidence of secretory activity. Adjacent intact fragments of endometrium with associated predecidual change usually can be identified and assist in the differential diagnosis. Another problem in differential diagnosis is the distinction of primary endometrial carcinoma from endocervical adenocarcinoma. This is problematic because these carcinomas share morphologic features (endometrioid and mucinous differentiation). Distinction can be difficult even in hysterectomy specimens when the tumor involves both the lower uterine segment and endocervix, and precursor lesions are lacking or obscured by carcinoma. The distinction is important because surgical management of these tumors often differs (see “Immunohistochemical Findings” for further discussion of markers that assist in this distinction). A related problem is the distinction of a primary endometrial carcinoma from a metastasis from an extrauterine site, discussed in “Tumors Metastatic to the Endometrium.” A high-grade endometrioid carcinoma at times may be difficult to distinguish from a carcinosarcoma (discussed in that section below).

Immunohistochemical Findings Endometrioid carcinoma expresses PAX8, pan-cytokeratins, epithelial membrane antigen (EMA), and the glycoprotein associated markers CA125, Ber EP4, and B72.3, among others. Expression of carcinoembryonic antigen (CEA), which is uncommon, is almost always limited to apical membranes, although tumors showing extensive mucinous differentiation may express

488

this antigen more diffusely. Nearly all endometrioid carcinomas are cytokeratin 7 positive and cytokeratin 20 negative (Wang et al. 1995; Castrillon et al. 2002). Occasionally, endometrioid tumors in areas of mucinous and squamous morular metaplasia express CDX2 (Wani et al. 2008; Park et al. 2008). Unlike many other adenocarcinomas, endometrioid tumors frequently display strong staining for vimentin. Studies of the molecular pathogenesis of endometrioid carcinoma have led to a better understanding of its immunophenotype. The preponderance of FIGO grade 1 and 2 endometrioid carcinomas express ER and progesterone receptor (PR) and approximately one-half of FIGO grade 3 endometrioid carcinomas without serous, clear cell, or undifferentiated features are ER/PR positive (Koshiyama et al. 1993; Reid-Nicholson et al. 2006; Soslow et al. 2000a; Lax et al. 1998a; Vang et al. 2001; Darvishian et al. 2004). p53 overexpression resulting from TP53 mutation and accumulation of mutant p53 protein is extremely rare in FIGO grade 1 adenocarcinomas and only in a minority of FIGO grade 2 adenocarcinomas but is present in a significant number of FIGO grade 3 adenocarcinomas. However, when p53 staining is prominent, serous, clear cell, or undifferentiated tumors should be considered (Lax et al. 1998a; Darvishian et al. 2004; Tashiro et al. 1997a; Soslow et al. 1998; Lax et al. 2000b; Lax et al. 1998b; Sherman et al. 1995; Zheng et al. 1998). Overexpression is defined as diffuse and strong expression in more than 80% of tumor cell nuclei. This should be distinguished from low-level expression of p53 in less than 50% of tumor cell nuclei, which is commonly found in endometrioid adenocarcinomas. PTEN is frequently mutated in endometrioid carcinomas (Darvishian et al. 2004; Obata et al. 1998; Risinger et al. 1997; Simpkins et al. 1998; Tashiro et al. 1997b; Yokoyama et al. 2000; Bussaglia et al. 2000) and expression of this gene is sometimes silenced via hypermethylation of its promoter. Detecting loss of PTEN with immunohistochemistry, however, is challenging (Darvishian et al. 2004; Pallares et al. 2005). DNA MMR proteins are found to be lacking in tumor cell nuclei using immunohistochemistry in

L. Hedrick Ellenson et al.

one-fifth to one-third of endometrioid carcinomas (Modica et al. 2007; Vasen et al. 2004; Peiro et al. 2002; de Leeuw et al. 2000). In sporadic cases, this most often results from MLH1 promoter hypermethylation. Those that arise in the setting of Lynch syndrome are due to mutations of MSH6, MSH2, MLH1, or PMS2, listed in descending order of prevalence. Interpreting DNA MMR protein immunohistochemistry relies on complete loss of expression in the setting of a valid positive internal control (Fig. 21). Valid internal controls include non-neoplastic endometrial stroma and glands with reproducibly stained nuclei. Expression loss, when present, usually occurs in couplets (MLH1 with PMS2 and MSH2 with MSH6) due to the fact that these form protein–protein complexes and loss of one of the proteins leads to destabilization of the other protein in the complex. For the distinction of endometrial and endocervical adenocarcinomas, the most useful marker panel depends on which subtypes of endometrial and endocervical adenocarcinomas are being considered in the differential diagnosis. For the most common situation of distinguishing endometrial endometrioid carcinomas from highrisk HPV-related endocervical adenocarcinomas, a panel of immunohistochemical markers comprised of p16, ER, and PR has been shown to be useful (Yemelyanova et al. 2009a; Staebler et al. 2002; Missaoui et al. 2006; McCluggage and Jenkins 2003; Ansari-Lari et al. 2004). The vast majority of endocervical adenocarcinomas (~90%) are human papillomavirus (HPV)-related and exhibit diffuse/moderate-strong p16 expression due to complex molecular mechanisms by which high-risk HPV transforming proteins (E6, E7) interact with cell cycle regulatory proteins (p53, pRb) to generate a futile feedback loop resulting in p16 over-expression (see ▶ Chap. 6, “Carcinoma and Other Tumors of the Cervix” on Cervical Cancer). Interestingly, these HPV-related endocervical adenocarcinomas also often lack hormone receptor expression (ER and PR receptors) (Staebler et al. 2002; McCluggage et al. 2002; Yemelyanova et al. 2009b; Ronnett et al. 2008). In contrast, endometrial endometrioid carcinomas are considered

9

Endometrial Carcinoma

489

Fig. 21 Immunohistochemistry for DNA MMR proteins in endometrioid carcinoma. (a, b) Endometrioid carcinoma, FIGO grade 2 has predominantly glandular and focal solid growth. (c) Loss of expression of MLH1 in the tumor cells.

(d) Loss of expression of PMS2 in the tumors cells. (e) Retention of MSH2 expression in the tumor cells. (f) Retention of MSH6 expression (despite a small number of positive tumor cells, this is interpreted as retention)

etiologically unrelated to HPV. They have been shown to exhibit generally patchy p16 expression of variable intensity, with mean/median extent of expression in 30–40% of tumor cells across all FIGO grades and only rare tumors exhibiting diffuse/strong expression (Fig. 22) (Yemelyanova

et al. 2009a). This patchy pattern, even when extensive, is distinct from the completely diffuse expression characteristic of high-risk HPV-related endocervical adenocarcinomas. In addition, most endometrial endometrioid carcinomas, particularly FIGO grade 1 and 2 tumors but also many

490

Fig. 22 Endometrioid carcinoma, FIGO grade 1. Tumor exhibits patchy expression of p16

FIGO grade 3 tumors, express hormone receptors (Lax et al. 1998a; Staebler et al. 2002; McCluggage et al. 2002). In practice, p16 alone usually suffices for this particular differential diagnosis. If hormone receptor expression is assessed, we specifically recommend the use of PR in conjunction with ER based on our published (Yemelyanova et al. 2009a; Staebler et al. 2002; Ronnett et al. 2008) and unpublished experiences indicating that the subset of endocervical carcinomas that retains some expression of hormone receptors most often retains ER expression to some degree (often focal/weak-moderate) with loss of PR expression; thus, PR is often the more discriminatory marker of the two for this differential diagnosis. Of note, a subset (minority) of highrisk HPV-related endocervical adenocarcinomas will retain expression of both ER and PR, so retained hormone receptor expression should not necessarily be used to refute a diagnosis of highrisk HPV-related endocervical adenocarcinoma. Detection of high-risk HPV DNA or RNA is definitive for diagnosing high-risk HPV-related endocervical adenocarcinoma but it should be noted that in situ hybridization assays are not 100% sensitive.

Molecular Genetics Over the past three decades, a number of cancercausing genes have been analyzed in endometrial carcinoma. Recently, a number of studies have shown that the most frequently altered gene in

L. Hedrick Ellenson et al.

endometrioid carcinoma is the PTEN tumor suppressor gene, which is mutated in 30–54% of cases (Risinger et al. 1997; Tashiro et al. 1997b). PTEN is located on chromosome 10q23.3 and encodes a dual-specificity phosphatase (Li et al. 1997). The primary target is the lipid molecule phosphatidylinositol 3,4,5-triphosphate (PIP3) that is involved in a signal transduction pathway that regulates cell growth and apoptosis. The dephosphorylation of PIP3 counteracts the activity of a protein complex called PI3K (phosphoinositol 3 kinase) that leads to the conversion of PIP2 (phosphatidylinositol 4,5-diphosphate) to PIP3. Consequently, the inactivating mutations in PTEN result in increased levels of PIP3, which activates downstream molecules including phosphorylation of protein kinase B (AKT). AKT is a central regulator of numerous pathways involved in cell proliferation, cell growth, and apoptosis that are altered in cancer development. Although the specific consequences of PTEN mutation have not been completely elucidated in endometrial carcinoma development, it has been noted that the frequency of PTEN mutation is similar in all three grades of endometrioid carcinoma. In addition, it is mutated in approximately 20–48% of atypical and nonatypical hyperplasias (Levine et al. 1998; Maxwell et al. 1998). These findings suggest that inactivation of this gene is important early in the pathogenesis of endometrioid carcinoma. Genetic mouse models with germline heterozygous deletion of PTEN spontaneously develop endometrial hyperplasia in 100% of female mice with 20% of female mice showing progression to carcinoma supporting an early role of PTEN in endometrial tumorigenesis (Podsypanina et al. 1999). Furthermore, one epidemiologic study found that the presence of PTEN mutations in complex atypical hyperplasia did not predict progression to carcinoma (Lacey et al. 2008). In sum, these findings suggest that PTEN mutations may be central to the development of hyperplasia, but may not play a role in the transition to carcinoma. Of further interest, mutations in the PIK3CA oncogene, the catalytic subunit of PI3K, are common in endometrioid carcinoma (Oda et al. 2005; Hayes et al. 2006). Mutations in PIK3CA are

9

Endometrial Carcinoma

activating mutations and, like PTEN mutations, lead to activation of the PI3K pathway. The mutations are found in endometrioid carcinoma with and without PTEN mutations but are more common in tumors with PTEN mutations. An additional study showed that while PTEN mutations occur in a similar frequency in complex atypical hyperplasia and carcinoma, PIK3CA mutations are rare in complex atypical hyperplasia and occur in approximately 39% of carcinoma, and are present in all three tumor grades (Hayes et al. 2006). These studies suggest that inactivation of PTEN and activation of PIK3CA have different roles in the development of endometrioid carcinoma. While PTEN is important in the development of hyperplasia, mutations in PIK3CA may play a role in the transition of complex atypical hyperplasia to carcinoma. Although the biologic basis of this has not yet been elucidated, understanding PTEN and PIK3CA mutations and their roles in the development of endometrial hyperplasia and carcinoma may provide targets for therapeutic intervention before the development of invasive disease. ARID1A mutations are common in endometrioid carcinoma occurring in about 40% of tumors. The vast majority are truncation mutations resulting in loss of expression of the protein. ARID1A is a tumor suppressor gene and is a member of the SWI/SNF family that has both helicase and ATPase activity. This family of genes is thought to control transcription of specific genes through their role in regulating chromatin structure. The TP53 tumor suppressor gene has been extensively studied in endometrial cancer, as in other tumors. TP53 encodes a DNA-binding phosphoprotein that is involved in cell cycle control and apoptosis. Mutations in TP53 are found in approximately 10% of all endometrioid carcinomas, with most occurring in grade 3, and occasionally in grade 2 tumors. Overall, TP53 mutations occur in approximately 50% of grade 3 tumors, and they have rarely been identified in grade 1 tumors or endometrial hyperplasia (Lax et al. 2000b). This finding is consistent with a role for TP53 in the progression, but not the initiation, of endometrioid carcinoma.

491

As discussed in the section on Lynch syndrome, another common molecular alteration in endometrioid carcinoma is the molecular phenotype of MSI. MSI is defined as alterations in the length of short, repetitive DNA sequences, called microsatellites, in tumor DNA compared to DNA isolated from the same patient’s normal tissue. This molecular phenotype is detected in tumors that lack an intact DNA MMR system, a fundamental cellular mechanism for preventing DNA alterations that are created largely during DNA replication. In tumors that display MSI, the DNA MMR system has been inactivated either through mutation or “silencing” by promoter hypermethylation of one of the DNA MMR genes (Esteller et al. 1998). The consequence of inactivating the DNA MMR system is an increase in the rate at which mutations occur, a factor that clearly contributes to tumorigenesis. Microsatellite instability is found in tumors from patients affected by Lynch syndrome in which endometrial carcinoma is the most common noncolorectal malignancy, as discussed above in the section on Lynch syndrome (Eshleman and Markowitz 1995). MSI also is present in approximately 20–30% of sporadic endometrial cancers and can be found in complex atypical hyperplasias that are associated with cancers that demonstrate instability (Levine et al. 1998; Mutter et al. 1996; Duggan et al. 1994). It remains unclear exactly when in the development of endometrial neoplasia the DNA MMR system becomes inactivated. Further studies of endometrial hyperplasia are warranted to address this important biologic and potentially clinically relevant question. Recent studies indicate that MMR deficiency is associated with a more aggressive tumor phenotype, despite the fact that patients with MMR deficient tumors have a similar overall survival to patients with intact MMR tumors. The underlying mechanisms responsible for this seemingly antithetical finding is unclear at present but may be due, at least in part, to the immunological response to the production of neoantigens resulting from the hypermutable state. A recent study reports that patients with MMR deficient

492

tumors due to methylation of MLH1 had reduced recurrence free survival (Cosgrove et al. 2017). Clearly, future studies are necessary to better define the prognostic significance of MMR deficiency, recognizing that the cause of the deficiency, the specific mutational profile, and the patient’s immune response, as well as response to therapy may all play a role in determining tumor behavior. As mentioned above, next generation sequencing studies found that approximately 5–7% of endometrioid carcinomas have mutations in POLE, a component of DNA polymerase epsilon, that results in an ultramutated phenotype. Although there have been associations of POLE mutations with certain morphologic features, they are not robust enough to allow for recognition of this underlying molecular abnormality. Since POLE mutations are associated with a good prognosis and eligibility for immune checkpoint inhibitor treatment, these tumors may be important to identify. However, at the present time, assessment of POLE mutations is not routinely performed as it requires DNA sequence analysis. However, in cases with ambiguous or unclassifiable morphology and unusual immunohistochemical features, mutational analyses should be considered. A number of oncogenes have been studied in endometrioid carcinomas, but only a few are altered in a significant number of cases. Mutations in the KRAS proto-oncogene have been identified consistently in 10–30% of endometrial cancers in several studies (Lax et al. 2000b; Boyd and Risinger 1991; Enomoto et al. 1993). The mutations have been found in all grades of endometrioid carcinoma and have been reported in complex atypical hyperplasia, suggesting a relatively early role for KRAS mutations in this tumor type. KRAS encodes a guanine nucleotide-binding protein of 21 kDa that plays a role in the regulation of cell growth and differentiation by transducing signals from activated transmembrane receptors. In the mutant form, KRAS is constitutively “on” even in the absence of an activated receptor. Mutations in FGFR2 (fibroblast growth factor receptor 2) have been identified in 16% of endometrioid carcinomas and it has been shown that KRAS and FGFR2 mutations are mutually

L. Hedrick Ellenson et al.

exclusive (Byron et al. 2008; Pollock et al. 2007). CTNNB1 is the gene that encodes β-catenin, an integral component of the Wnt signaling pathway, involved in regulation and coordination of cell-cell adhesion and gene transcription. In endometrial carcinomas, mutations of CTNNB1 leading to nuclear overexpression of β-catenin can be seen in a subset of low-grade endometrioid carcinomas and significantly less so in either higher grade endometrioid carcinomas or endometrial carcinomas with nonendometrioid histologies (Scholten et al. 2003). However, CTNNB1 mutations and subsequent Wnt pathway activation are only present in a subset of low-grade endometrioid endometrial carcinomas. Multivariate analysis of prognostic factors in low-grade and early stage endometrioid carcinomas has shown that CTNNB1 mutations (in particular activating exon 3 mutations) are strongly associated with increased risk of recurrence and overall worse prognosis (Heckl et al. 2018; Kurnit et al. 2017; Liu et al. 2014). L1-cell adhesion molecule (L1CAM) is a transmembrane protein of the immunoglobulin family initially identified in the nervous system, which has been associated with invasive tumor growth and aggressive tumor behavior in various malignancies, by acting as a proangiogenic factor (Friedli et al. 2009; Kiefel et al. 2012; Kommoss et al. 2017; Raveh et al. 2009). In endometrial carcinoma, L1CAM appears to contribute to alterations in the Wnt signaling pathway and epithelialmesenchymal transition (EMT) (Kommoss et al. 2017; Colas et al. 2012; Heuberger and Birchmeier 2010). Studies have shown a significant association between L1CAM expression and worse clinical outcome in endometrioid histology, and have also established correlations between L1CAM expression and high risk factors such as non-endometrioid histology, vascular invasion, and high-grade tumors (Kommoss et al. 2017; Bosse et al. 2014; Dellinger et al. 2016; Geels et al. 2016; Smogeli et al. 2016; van der Putten et al. 2016; Zeimet et al. 2013). When looking at L1CAM expression in low grade, early stage endometrial endometrioid carcinomas, which would otherwise be considered as “low-risk” tumors, increased expression of L1CAM by

9

Endometrial Carcinoma

493

immunohistochemistry was associated with a significant decrease in 5-year survival (71.8% versus 100% in L1CAM negative tumors), suggesting a potential role of L1CAM in management algorithms for these patients (Kommoss et al. 2017). Other oncogenes that have been found to be overexpressed or amplified are EGFR, CMYC, HER-2/neu, BCL2, and CFMS (Borst et al. 1990; Hetzel et al. 1992; Leiserowitz et al. 1993; Taskin et al. 1997). Additional studies on these genes are needed to more definitively determine their role in endometrial cancer.

Behavior and Treatment Endometrioid adenocarcinoma spreads by lymphatic and vascular dissemination, direct extension to contiguous organs, and transperitoneal and transtubal seeding. Lymphatic metastasis is more common than hematogenous spread, but involvement of the lungs without metastasis to mediastinal lymph nodes suggests that hematogenous spread may occur early in the course of disease. Endometrial carcinoma tends to spread to the pelvic lymph nodes before involving paraaortic lymph nodes; however, rare examples of isolated para-aortic metastases have been reported. The relative frequency of metastasis to lymph node groups and various organs is shown in Tables 4 and 5, respectively. The standard treatment for endometrial carcinoma is hysterectomy and bilateral salpingooophorectomy. Over the years, preoperative or postoperative radiotherapy and chemotherapy have been used in addition to hysterectomy. The current approach is to treat all patients, when Table 4 Sites of lymph node metastasis from endometrial carcinomas at autopsy (From Hendrickson 1975) Lymph nodes Para-aortic Hypogastric External iliac Common iliac Obturator Sacral Mediastinal Inguinal Supraclavicular

Relative frequency (%) 64 61 48 40 37 22 18 16 12

feasible, by hysterectomy supplemented by surgical staging and to administer postoperative radiation to patients with poor prognostic factors that put them at high risk of recurrence. Postoperative estrogen replacement therapy has been advocated for patients with early stage disease and no significant poor prognostic factors (Creasman et al. 1986). One study showed that survival is not compromised in patients with low tumor grade (grades 1 and 2), less than 50% myometrial invasion, and no metastases to lymph nodes or other organs (Lee et al. 1990). Given the prognostic significance of pelvic and paraaortic lymph node status, these nodes should be sampled or dissected in patients when any of the following is present: greater than 50% myometrial invasion, grade 3 tumor, cervical involvement, extrauterine spread, serous, clear cell, or undifferentiated carcinoma, or palpably enlarged lymph nodes. In a Gynecologic Oncology Group (GOG) study, only a quarter of patients had these findings, but they accounted for the majority of patients with positive aortic lymph nodes (Morrow et al. 1991). Recently, evaluation of sentinel lymph nodes has become common, as several studies have suggested that it is both sensitive Table 5 Sites of metastasis from endometrial carcinoma at autopsy (From Hendrickson 1975) Organ site Lung Peritoneum and omentum Ovary Liver Bowel Vagina Bladder Vertebra Spleen Adrenal Ureter Brain or skull Vulva Breast Hand Femur Tibia Pubic bone Skin

Relative frequency (%) 41 39 34 29 29 25 23 20 14 14 8 5 4 4

Rare

494

and specific, and reduces the morbidity associated with lymph node dissection. However, the significance of micrometastases and isolated tumor cells on prognosis remains unclear (Holloway et al. 2017). Several studies have shown that the depth of myometrial invasion can be assessed by gross inspection and intraoperative frozen section (Noumoff et al. 1991; Shim et al. 1992; Egle et al. 2008). However, other studies have suggested that intraoperative assessments are not reproducible with discrepancies of up to 38% with the final diagnosis and that random sections in the absence of a gross lesion are not warranted (Desouki et al. 2017). Postoperatively, patients are classified as low, intermediate, or high risk based on surgical pathologic staging. Patients with grade 1 or 2 tumors that are confined to the endometrium or are minimally invasive are defined as low risk and require no further therapy. Patients with pelvic or paraaortic lymph node metastases, or involvement of the adnexa or intraperitoneal sites, are high risk and receive postoperative radiation (vaginal cuff, pelvis, paraaortic area, or whole abdominal). Radiation appears to be of benefit because the 5-year survival rate for women with positive aortic lymph nodes who were treated with postoperative radiation is nearly 40% (Morrow et al. 1991). Despite treatment with surgery and radiotherapy, 50% of stage III tumors recur. Half of these patients die with distant metastasis, although local control is also a major problem. About 4% of patients with endometrial carcinoma have stage IV disease. Spread to the lungs occurs in 36% of patients with stage IV disease. Patients who do not qualify as low or high risk are intermediate in risk. A decision as to whether or not these patients should receive postoperative radiation should be individualized because there are no conclusive data demonstrating a survival benefit for these patients treated with postoperative radiotherapy. Studies evaluating the use of adjuvant hormonal or cytotoxic chemotherapy have shown no improvement in survival over surgery and radiation, and consequently these methods currently are not recommended as standard treatment. In contrast, radiation, hormone, and cytotoxic chemotherapy are used for management of patients

L. Hedrick Ellenson et al.

with recurrent tumor; 50% of patients with isolated vaginal vault recurrence treated by irradiation are alive at 3 years (Podczaski et al. 1992).

Histologic Effects of Treatment Radiation The histologic changes in neoplastic tissues after intracavitary radiation are nonspecific and variable, showing minor to major alterations from their pre-irradiated state. Similarly, nonneoplastic endometrial or endocervical glands may be affected only minimally or show nuclear and cytoplasmic changes that are indistinguishable from those found in neoplastic cells. Because the cytologic changes in both neoplastic and nonneoplastic tissue are similar, identification of carcinoma depends largely on the recognition of histologic patterns and signs of invasion. Irradiated carcinoma generally retains a haphazard glandular pattern, but nonirradiated, nonneoplastic glands tend to maintain their normal architectural arrangement despite radiation effects in the endometrial stroma and myometrium. When radiation effect is evident, nuclei tend to be enlarged, highly pleomorphic, and hyperchromatic, with coarsely clumped chromatin. The cytoplasm often is granular and swollen. Vacuolation can be present in both the nucleus and the cytoplasm. The nuclear changes result from replication of DNA without cell division. Cytoplasmic vacuolation results from dilatation of various organelles and possible lysis caused by damaged lysosomal membranes. In some instances, radiation may enhance cellular differentiation. Occasionally, poorly differentiated carcinomas without squamous differentiation in the curettings may have nests of squamous epithelium in the resected uterus after radiation. It is in mitosis and the S phase of the cell cycle that a cell is most susceptible to radiation injury. Thus, the difference in radiosensitivity of tumor cells and benign cells is due largely to the increased mitotic activity of the neoplastic cells and the better reparative capacity of nonneoplastic cells. In view of the variable morphologic response to irradiation, it is often difficult to determine whether irradiated tumor cells are viable. On a practical basis, if tumor cells are

9

Endometrial Carcinoma

evident after irradiation, it should be assumed that some retain the capacity to persist however abnormal they appear. Radiation changes in the endometrial stroma and myometrium are greatest in the vicinity of the radiation source. The stromal cells are first converted to giant fibroblasts. Early vascular effects include damage to endothelial cells, resulting in thrombosis. The stroma undergoes progressive hyalinization, resulting in a collagenous scar. Elastic tissue often is fragmented and frayed, and blood vessels are thickened and sclerotic. Occasionally, changes similar to those found in atherosclerosis may be present in the intima of blood vessels. Foam cells occur in the intima, and myometrial cells may appear granular and swollen, especially in areas close to the radium source. Scarring, atrophy, and sclerosis of vessels characterize long-standing radiation effects. The endometrium is thin and easily traumatized, and small blood vessels in the stroma are thin walled and ectatic. Some blood vessels form plaques of lipid-filled clear cells in the media. Progestins Progestin-induced changes include secretory differentiation of glandular cells, mitotic arrest, conversion of spindle-shaped stromal cells to decidual cells, decrease in estrogen-related cellular changes such as ciliogenesis, and development or enlargement of squamous areas (Richart and Ferenczy 1974; Saegusa and Okayasu 1998). The earliest evidence of progestin effect is subnuclear vacuolization, observed within 2–3 days of treatment. The vacuoles are a manifestation of glycoprotein synthesis, which is followed by an apocrine-type secretion in which the apical portion of the cytoplasm of the cell is discharged into the gland lumen, with reduction in the size of the cell. Longer term therapy aimed at eliminating the disease, at least until patients can become pregnant, results in a number of morphologic changes that can predict response to therapy. These include decreased glandular-to-stroma ratio, decreased to absent mitotic activity, decreased glandular cellularity, loss of cytologic atypia, and a variety of cytoplasmic changes including mucinous, secretory, squamous, and eosinophilic metaplasia.

495

Persistent architectural abnormalities and/or cytologic atypia were predictive of treatment failure (Mentrikoski et al. 2012). Some architectural changes (cribriform and papillary patterns) induced by progestin treatment, are noteworthy as they may be confused with progression (Wheeler et al. 2007; Gunderson et al. 2014). Importantly, biopsies taken after the initiation of treatment require a comparison to the pre-treatment sample for correct interpretation and determination of treatment response.

Squamous Differentiation Many endometrioid adenocarcinomas contain squamous epithelium, but the amount of squamous epithelium can vary widely. In a wellsampled neoplasm, the squamous element should constitute at least 10% of a tumor to qualify as an adenocarcinoma with squamous differentiation. Endometrioid carcinomas with squamous epithelium should be classified simply as endometrioid carcinoma with squamous differentiation (not “with squamous metaplasia”) and graded on the basis of the glandular component as well, moderately, or poorly differentiated (grade 1, 2, or 3, respectively, per FIGO criteria). There are no differences in the clinical features of endometrioid carcinoma containing squamous epithelium and endometrioid carcinoma lacking squamous elements. Thus, there are no differences in the frequency of obesity, hypertension, diabetes, and nulliparity among the large series in which this has been analyzed (Alberhasky et al. 1982; Connelly et al. 1982).

Gross and Microscopic Findings These tumors have no distinctive gross findings. Low-grade tumors (grade 1) are composed of glandular and squamous elements but generally the glandular component predominates; the nests of squamous epithelium are confined to gland lumens. The squamous epithelium resembles metaplastic squamous cells of the cervical transformation zone. Frequently, nests of cells with a prominent oval to spindle cell appearance,

496

Fig. 23 Endometrioid carcinoma with squamous differentiation, FIGO grade 1. Well-differentiated endometrioid glands are intimately admixed with solid nests of low-grade squamous epithelium. Grading is based on the features of the glandular component alone

Fig. 24 Endometrioid carcinoma with focal squamous differentiation, FIGO grade 3. Carcinoma has areas of squamous differentiation and is classified as high-grade endometrioid based on the solid non-squamous epithelium (>50%) with focal residual endometrioid glandular differentiation

referred to as morules, are observed (Fig. 23). Intercellular bridges can be identified within the squamous epithelium, and keratin formation is common. The nuclei of the squamous cells are bland, uniform, and lack prominent nucleoli. Mitotic figures are rare. In higher-grade tumors, the squamous element is cytologically more atypical and is not confined to gland lumens but often extends out from the glands (Fig. 24). At times, the squamous cells have a spindle

L. Hedrick Ellenson et al.

appearance simulating a sarcoma. They may not be in direct continuity with the glandular epithelium, appearing in isolated nests within the myometrium or in vascular spaces. Keratinization and pearl formation occur to varying degrees. Generally, the glandular component predominates, but masses of epithelial cells that may represent poorly differentiated glandular or squamous cells can lie between glands. This epithelium should be considered glandular unless intercellular bridges are demonstrated or the cells have prominent eosinophilic cytoplasm, welldefined cytoplasmic borders, and a sheet-like proliferation without evidence of gland formation. Both the glandular and squamous components display grade 2 or 3 nuclear atypia, an increased nuclear cytoplasmic ratio, and increased mitotic activity. The glandular architecture usually is poorly differentiated. Tumors of intermediate differentiation are common. These neoplasms contain glandular and solid areas in which the squamous cells display a moderate degree of nuclear atypia, defying separation into a “benign” and “malignant” squamous category. A rare finding in patients with endometrioid carcinoma with squamous differentiation is the presence of keratin granulomas that may involve a wide variety of sites in the peritoneal cavity including the ovaries, tubes, omentum, and serosa of the uterus and bowel (Chen et al. 1978; Kim and Scully 1990; van der Horst and Evans 2008). Microscopically, these lesions consist of a central mass of keratin and necrotic squamous cells surrounded by a foreign body granulomatous reaction. A proliferation of mesothelial cells also may be present. The granulomas probably result from exfoliation of necrotic cells from the tumor, followed by transtubal spread and implantation on peritoneal surfaces. It is important to distinguish pure keratin granulomas from lesions with both viable-appearing tumor cells and keratin accompanied by a foreign body-type giant cell reaction because the former lesions have not been associated with an unfavorable prognosis. Thus, pure keratin granulomas should not be diagnosed as metastatic endometrioid carcinoma; the latter requires a viable-appearing carcinomatous component.

9

Endometrial Carcinoma

Differential Diagnosis The most common problem in the differential diagnosis of the low-grade tumors is with atypical hyperplasia showing squamous metaplasia. To distinguish between the two, the criteria for identifying endometrial stromal invasion should be employed (see ▶ Chap. 8, “Precursors of Endometrial Carcinoma”). At times, a low-grade tumor may be confused with a highgrade carcinoma because the masses of squamous epithelium are misconstrued as a solid proliferation of neoplastic cells. The nuclear grade is high in poorly differentiated carcinoma, however. Occasionally, squamous morules may be confused with granulomas, but the presence of foreign body giant cells and an inflammatory infiltrate helps identify the latter. For high-grade adenocarcinomas with squamous epithelium, the major problem in differential diagnosis in curettings is distinguishing a primary carcinoma of the endometrium from an adenosquamous carcinoma arising in the endocervix. In the cervix, the squamous component usually predominates, whereas in the endometrium the glandular component predominates. A profusion of cell types, especially mucinous or signet ring cells, is more characteristic of an endocervical neoplasm. (See above section on differential diagnosis). Behavior and Treatment As already described, when stratified according to stage, grade, and depth of myometrial invasion, there are few differences in the behavior of carcinomas with squamous epithelium compared with endometrioid carcinomas without squamous epithelium (Abeler and Kjorstad 1992; Zaino and Kurman 1988; Zaino et al. 1991). As occurs with endometrioid carcinomas, the low-grade carcinomas with squamous epithelium tend to be only superficially invasive and seldom invade vascular channels. In contrast, high-grade tumors have a high frequency of deep myometrial invasion, vascular space involvement, and pelvic and paraaortic lymph node metastasis. Metastasis of high-grade tumors occurs widely throughout the pelvis and abdomen, involving bowel, mesentery, liver,

497

kidney, spleen, and lymph nodes. Distant metastasis may involve the lungs, heart, skin, and bones. Nearly two-thirds of metastases contain both glandular and squamous elements, but pure adenocarcinoma or squamous carcinoma is encountered in 20% and 8%, respectively (Ng et al. 1973). Often it is the squamous component that is identified in vascular channels. Accordingly, the treatment for carcinomas with squamous differentiation is the same as that for endometrioid carcinomas without squamous differentiation of comparable stage.

Villoglandular Carcinoma Villoglandular carcinoma is a variant of endometrioid carcinoma that displays a papillary architecture in which the papillary fronds are composed of a delicate fibrovascular core covered by columnar cells that generally contain bland nuclei (Chen et al. 1985; Hendrickson et al. 1982). The median age is 61 years, similar to that of women with typical endometrioid carcinoma. In all other respects, women with these tumors are similar to patients with low-grade endometrioid carcinoma. The microscopic appearance of villoglandular carcinoma is characterized by thin, delicate fronds covered by stratified columnar epithelial cells with oval nuclei that generally display mild to moderate (grade 1 or 2) atypia (Figs. 25 and 26). Occasionally, more atypical (grade 3) nuclei may be observed. Mitotic activity is variable, and abnormal mitotic figures are rare (Chen et al. 1985). Myometrial invasion usually is superficial.

Differential Diagnosis The main consideration in the differential diagnosis is serous carcinoma because both villoglandular and serous carcinomas have a prominent papillary pattern. In contrast to serous carcinomas, villoglandular carcinomas have long delicate papillary fronds and are covered by columnar cells with only mild to moderate nuclear atypia. The cells look distinctly endometrioid with a smooth, luminal border. To have significance as a distinctive entity, the diagnosis is reserved for

498

Fig. 25 Endometrioid carcinoma, villoglandular type. Tumor has papillary architecture, which might lead to misclassification as serous carcinoma, but columnar epithelium with low-grade cytologic features (see Fig. 23) is consistent with endometrioid carcinoma

L. Hedrick Ellenson et al.

appearance. The nuclei of serous carcinomas are highly pleomorphic and atypical (grade 3). Cherry red macronucleoli typically are present and the cells have a hobnail appearance, often with smudged, hyperchromatic nuclei. It should be noted that considerable nuclear heterogeneity can be observed. If the possibility of serous carcinoma cannot be excluded based on morphologic findings, immunohistochemical analyses for p53, p16, ER, and PR will aid in the distinction. Villoglandular tumors show a wild-type p53 expression pattern, patchy p16 expression and the vast majority will show intense, diffuse expression of ER and PR. Further discussion of this immunohistochemical panel is presented in the serous carcinoma section. Villoglandular carcinomas are generally better differentiated than typical endometrioid carcinomas but are not significantly different with respect to depth of invasion or frequency of nodal metastases (Zaino et al. 1998). In addition, villoglandular carcinomas are frequently admixed with typical endometrioid carcinoma. In view of the frequent admixture of the two patterns and similar prognosis, villoglandular carcinoma is considered a variant of endometrioid carcinoma. Treatment is the same as for endometrioid carcinoma of comparable stage, grade, and depth of invasion.

Secretory Carcinoma Fig. 26 Endometrioid carcinoma, villoglandular type. Endometrioid differentiation is confirmed by the presence of columnar epithelium and low-grade cytologic features (elongated uniform nuclei)

tumors in which most of the neoplasm has a villoglandular appearance. In contrast to villoglandular carcinomas, serous carcinomas tend to have shorter, thick, densely fibrotic papillary fronds (Bartosch et al. 2011). The most important distinguishing feature is the cytologic appearance. The cells of serous carcinoma tend to be rounder, forming small papillary clusters that are detached from the papillary fronds, a finding that is often referred to as papillary tufts. As a consequence, the luminal border has a scalloped

Secretory carcinoma is a variant of typical endometrial carcinoma in which the majority of cells exhibit subnuclear or supranuclear cytoplasmic vacuoles resembling early secretory endometrium. An unusual pattern, it represents only 1–2% of endometrial carcinomas (Tobon and Watkins 1985). The age range is from 35 to 79 years, with a mean age of 55–58 (Tobon and Watkins 1985; Christopherson et al. 1982a). Most patients are postmenopausal and experience abnormal bleeding. This histologic subtype also may be seen after progestin treatment of an endometrioid carcinoma. In all other respects, including the association of obesity, hypertension, diabetes mellitus, and exogenous estrogen administration,

9

Endometrial Carcinoma

Fig. 27 Endometrioid carcinoma, secretory type, FIGO grade 1. Endometrioid-type epithelium displays prominent sub- and supranuclear vacuolization, reminiscent of day 18 secretory endometrium

499

Fig. 29 Endometrioid carcinoma, secretory type, FIGO grade 2. Endometrioid-type glands and solid non-squamous epithelium exhibit diffuse secretory differentiation manifested as clear cytoplasmic change. Tumor has uniform low-grade cytologic features and lacks any of the characteristic architectural patterns of clear cell carcinoma, supporting a diagnosis of secretory type endometrioid carcinoma

Fig. 28 Endometrioid carcinoma, secretory type, FIGO grade 1. Endometrioid-type glands with secretory differentiation are present within a desmoplastic stroma. Prominent subnuclear vacuolization is reminiscent of day 17 secretory endometrium

secretory carcinoma in young women typically shows a secretory pattern that is more advanced than 17 days, and a corpus luteum is found in most premenopausal patients when a hysterectomy and bilateral salpingo-oophorectomy are performed. Nonetheless, a relationship to progesterone stimulation is not always demonstrable. In fact, secretory carcinoma may occur spontaneously in postmenopausal women without exogenous or abnormal levels of progesterone. The secretory activity in the tumor may be transient because it has been observed in curettings but not in the later hysterectomy specimen (Christopherson et al. 1982a).

patients with secretory carcinoma are similar to women with endometrioid carcinoma. Microscopically, secretory carcinoma displays a well-differentiated glandular pattern and is composed of columnar cells, often unstratified, with subnuclear or supranuclear vacuolization closely resembling day 17–22 secretory endometrium (Figs. 27, 28, and 29) (Tobon and Watkins 1985). Usually the nuclei are grade 1. The secretory pattern may be focal or diffuse, and it is frequently admixed with typical endometrioid adenocarcinoma. The endometrium adjacent to

Differential Diagnosis It is important to distinguish secretory carcinoma from clear cell carcinoma in view of the excellent prognosis of the former and unfavorable prognosis of many of the latter. Although both tumors are composed of cells with clear, glycogen-rich cytoplasm, the histologic features are distinctive. At times a secretory carcinoma that has a predominantly glandular pattern can become solid and simulate clear cell carcinoma. The tumors are distinguished by their architectural and cytologic appearance. Secretory carcinoma displays a

500

glandular architecture like endometrioid carcinoma, is rarely papillary or cystic, and usually is not solid. The cells of secretory carcinoma are columnar, similar to those in endometrioid carcinoma, except that they have supranuclear or subnuclear vacuoles. In contrast, clear cell carcinoma often exhibits tubulocystic and/or papillary architecture but can be glandular. The cells usually have marked nuclear atypia (grade 3), with rounded cells having a variety of characteristic features including hobnail morphology, smudgy hyperchromasia, and prominent nucleoli (Bartosch et al. 2011). Cells with clear cytoplasm also may be seen in the squamous component of an endometrioid adenocarcinoma with squamous differentiation. The clear appearance of these cells is also due to the presence of glycogen. Clear squamous cells tend to be polygonal and usually merge with more typical squamous cells with abundant eosinophilic cytoplasm. The distinction of secretory carcinoma from atypical hyperplasia with secretory effect can be difficult and is based on the presence of stromal invasion in the carcinoma (see ▶ Chap. 8, “Precursors of Endometrial Carcinoma”). Treatment is the same as that for endometrioid carcinoma of the same stage and grade. Secretory carcinoma usually is low grade with a good prognosis (Tobon and Watkins 1985). Death from recurrent disease is rare (Christopherson et al. 1982a).

Ciliated Carcinoma Ciliated carcinoma is a very rare type of low-grade endometrioid carcinoma (Hendrickson and Kempson 1983). It does not need to be classified separately from endometrioid carcinoma; its only importance is to remind the pathologist that endometrial proliferations with cilia may still be carcinomas. Estrogen induces cilia formation in the normal endometrium. Despite the prevalence of estrogen use, ciliated carcinoma is an extremely rare carcinoma, and most endometrial proliferations in which cilia are observed represent hyperplasias associated with eosinophilic or ciliated change. Patients range in age from 42 to 79 years, are often postmenopausal, and present

L. Hedrick Ellenson et al.

with bleeding. Ciliated carcinoma has an association with exogenous estrogen treatment. Microscopically, ciliated carcinoma is well differentiated and often displays a cribriform pattern. The gland lumens in the cribriform areas are lined by cells with prominent eosinophilic cytoplasm and cilia. The nuclei of ciliated cells generally have an irregular nuclear membrane and display coarse nuclear chromatin with prominent nucleoli. In most cases, ciliated carcinoma is admixed with nonciliated endometrioid carcinoma and occasionally areas of mucinous carcinoma. Although some ciliated carcinomas are moderately differentiated and invade to the middle third of the myometrium, none of the patients has developed recurrence or died of disease. Thus, the presence of cilia in a bona fide carcinoma identifies a low-grade neoplasm.

Corded and Hyalinized Endometrioid Carcinoma This is an unusual variant of endometrioid carcinoma that is perhaps most important because its histologic features may suggest a carcinosarcoma. The defining feature of this tumor is the presence of cords of epithelial cells or spindle cells embedded in hyalinized stroma that may mimic a sarcomatous component (Fig. 30). However, unlike carcinosarcomas, this pattern is usually associated with typical low-grade endometrioid carcinoma, often with areas of squamous differentiation. In addition, endometrial hyperplasia is found in approximately half of the cases. On careful examination, the corded epithelial cells and the spindle cells have low-grade cytological atypia in distinction to carcinosarcomas that have high-grade atypia in both the epithelial and sarcomatous components (Murray et al. 2005).

Mucinous Carcinoma This uncommon type of endometrial carcinoma has an appearance similar to mucinous carcinoma of the endocervix (Czernobilsky et al. 1980; Tiltman 1980). Mucinous carcinoma represents the dominant cellular population in only 1–9%

9

Endometrial Carcinoma

Fig. 30 Endometrioid carcinoma, corded and hyalinized variant. (a) Low power magnification demonstrates a typical endometrioid glandular component admixed with a corded epithelioid component, both with low-grade

of endometrial carcinomas (Ross et al. 1983; Melhem and Tobon 1987). To qualify as a mucinous carcinoma, more than one-half the cell population of the tumor must contain periodic acid–Schiff- (PAS-) positive, diastase-resistant intracytoplasmic mucin. Judging from the few published cases, the clinical features of patients with mucinous carcinoma of the endometrium do not differ from those with endometrioid carcinoma. However, one study found that mucinous differentiation is a risk factor for nodal metastases but did not change overall survival (Musa et al. 2012). Patients range in age from 47 to 89 years and typically present with vaginal bleeding. In one study, more than 40% had a history of receiving exogenous estrogens (Melhem and Tobon 1987). Most patients present with stage I disease. These tumors do not have distinctive gross features. The most frequent architectural pattern is glandular, often in a villoglandular configuration (Figs. 31 and 32). The epithelial cells lining the glands and papillary processes tend to be uniform columnar cells with minimal stratification. Cribriform areas are unusual; cystically dilated glands filled with mucin and papillary fronds surrounded by extracellular lakes of mucin, containing neutrophils, are typical. Curiously, mucinous differentiation sometimes is associated with squamous differentiation. Nuclear atypia is mild to moderate,

501

nuclear features. (b) Higher magnification demonstrates the hyalinized background with embedded cords of epithelial cells adjacent to well-formed glands

Fig. 31 Mucinous carcinoma, FIGO grade 1. Confluent glands have abundant mucinous cytoplasm and small, basally situated nuclei

and mitotic activity is not prominent. Hyperplasia and mucinous metaplasia sometimes are present in the adjacent endometrium. One study reported that the carcinoma was present in a polyp in 27% of the cases (Melhem and Tobon 1987). The presence of intracytoplasmic mucin can be identified on H&E stains by its distinctive granular, foamy, or bubbly appearance and can be confirmed by PAS, mucicarmine, or alcian blue stains. The intracytoplasmic mucin is variable in both the distribution of mucinous cells in the tumor and in the location of the mucin within individual cells. Mucin may be diffusely present

502

Fig. 32 Mucinous carcinoma, FIGO grade 1. Tumor has extensive mucinous differentiation and confluent glandular and papillary growth; these architectural patterns allow for establishing a diagnosis of carcinoma

Fig. 33 Endometrioid carcinoma with mucinous differentiation, FIGO grade 1. Tumor exhibits a cribriform growth pattern and is comprised of glands with prominent mucinous cytoplasm as well as ones with a typical endometrioid appearance

in the cytoplasm, confined to the apical area, or show a combination of both patterns. Tumors dominated by typical endometrioid carcinoma with less than 50% of a mucinous component can be designated as endometrioid carcinomas with mucinous differentiation (Fig. 33).

Differential Diagnosis Endocervical epithelium merges with the endometrium in the lower uterine segment, so it is not surprising that the distinction of primary endocervical from endometrial mucinous carcinoma

L. Hedrick Ellenson et al.

in curettings can be difficult. There is no histochemical difference in the mucin at either site (Ross et al. 1983). The distinction of endocervical from endometrial adenocarcinomas has been discussed earlier (see differential diagnosis section for Endometrioid Carcinoma). The distinction of mucinous carcinoma of the endometrium from clear cell or secretory carcinoma is made on the basis of morphology and PAS and mucin stains. The cells in secretory carcinoma are clear (not granular or foamy) because of the presence of glycogen, which is PAS positive and is removed by diastase treatment. Mucin in these tumors is focal at most. Clear cell carcinoma is almost always papillary or solid in contrast to the glandular pattern of mucinous carcinoma. The cells in clear cell carcinoma tend to be polygonal rather than columnar and hobnail cells are almost invariably present, a cytologic feature that is absent in mucinous carcinoma. Rarely, a mucinous carcinoma or a mixed mucinous and endometrioid carcinoma may contain areas that simulate microglandular hyperplasia of the cervix (Young and Scully 1992; Zaloudek et al. 1997). Such foci are characterized by cells showing mucinous and eosinophilic change with microcystic spaces containing acute inflammatory cells. The patients are in their fifties and sixties, in contrast to women with microglandular hyperplasia, who are young. The complexity of the glandular pattern and the degree of cytologic atypia distinguish this type of carcinoma from microglandular hyperplasia.

Behavior and Treatment When stratified by stage, grade, and depth of myometrial invasion, mucinous tumors behave like endometrioid carcinomas (Ross et al. 1983). Mucinous carcinomas, however, tend to be low grade and minimally invasive and therefore as a group have an excellent prognosis. Treatment is the same as for endometrioid carcinoma. Because most of the tumors are stage I, low grade, and minimally invasive, total abdominal hysterectomy and bilateral salpingo-oophorectomy usually suffice.

9

Endometrial Carcinoma

Serous Carcinoma The existence of papillary patterns within endometrial carcinoma has been recognized since the turn of the century. In the past several decades, reports have described the morphologic similarity of serous carcinomas of the endometrium, which frequently display papillary architecture, to ovarian serous carcinomas and identified them as a highly aggressive type of endometrial carcinoma (Hendrickson et al. 1982; Christopherson et al. 1982b; Lauchlan 1981; Walker and Mills 1982). Although papillary architecture is a common finding in serous carcinoma, most other types of endometrial carcinoma can display papillary architecture but are usually not highly aggressive tumors. In addition, serous carcinomas can be predominantly glandular or solid without evident papillary growth. What distinguishes serous carcinoma from these other types is the uniformly marked cytologic atypia. Thus, the designation “serous carcinoma,” rather than “papillary serous carcinoma,” is preferred so that cell type rather than architecture is emphasized.

Clinical Features The prevalence of serous carcinomas reported from referral centers usually is about 10%; however, in a population-based study from Norway it was only 1% (Abeler and Kjorstad 1990). Patients with serous carcinoma range in age from 39 to 93 years but typically are postmenopausal and, in contrast to women with endometrioid carcinoma, are older (reported median and mean ages are in the late sixties). In addition, they are less likely to have received estrogen replacement therapy and are more likely to have abnormal cervical cytology. There are some data to suggest that women with this neoplasm are less likely to be obese and that a higher proportion of women are black (Dunton et al. 1991). In other respects, they appear similar. Gross Findings On gross examination, uteri containing these tumors often are small and atrophic. Generally, the tumor is exophytic and has a papillary appearance. Depth of invasion is difficult to assess on macroscopic examination. It is not unusual to find

503

a benign-appearing polyp containing the carcinoma in the hysterectomy specimen after a diagnosis of serous carcinoma or serous endometrial intraepithelial carcinoma (SEIC) has been made on a curetting, because these tumors frequently develop within a polyp (see chapter ▶ “Precursors of Endometrial Carcinoma”) (Carcangiu and Chambers 1992; Carcangiu et al. 1997; Sherman et al. 1992; Silva and Jenkins 1990; Soslow et al. 2000b; Wheeler et al. 2000).

Microscopic Findings As experience with serous carcinoma has increased, it has become apparent that this neoplasm demonstrates considerable diversity in its architectural features (Figs. 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, and 44). Although a papillary pattern typically predominates, glandular and solid patterns also occur (Darvishian et al. 2004; Hendrickson et al. 1982; Sherman et al. 1992; Lee and Belinson 1991). Serous carcinoma originally was described as having thick, short papillae, but subsequent studies have shown that thin papillae may be present in more than half of them. The cytologic features of these tumors also are quite varied. Polygonal cells with eosinophilic and clear cytoplasm often are seen, but hobnail cells are among the most frequently observed cells. Marked nuclear atypia is always present and is required for a tumor to qualify as serous carcinoma (Figs. 35, 37, 39, and 42). Thus, serous

Fig. 34 Serous carcinoma. Papillary tumor is lined by markedly atypical epithelium composed of cells with scalloped luminal borders, including hobnail-type cells

504

Fig. 35 Serous carcinoma. Papillary structures and detached epithelial clusters have markedly atypical cells, including hobnail-type cells

Fig. 36 Serous carcinoma. Papillary structure is lined by markedly atypical hobnail-type cells

Fig. 37 Serous carcinoma. Most nuclei are vesicular with prominent red nucleoli but some detached atypical cells have smudged hyperchromatic nuclei

L. Hedrick Ellenson et al.

Fig. 38 Serous carcinoma. Tumor is composed of papillae lined by epithelium having prominent scalloped luminal borders

Fig. 39 Serous carcinoma. Papillae are lined by cells with enlarged, vesicular nuclei with evident nucleoli. Several mitotic figures are present

Fig. 40 Serous carcinoma. Glands with prominent intraglandular papillary architecture infiltrate myometrium

9

Endometrial Carcinoma

Fig. 41 Serous carcinoma. Papillae are lined by markedly atypical epithelium

Fig. 42 Serous carcinoma. Some glandular epithelium has smooth luminal borders but papillary epithelial tufts and marked nuclear atypia, with numerous mitotic figures, are characteristic of serous carcinoma

carcinoma is defined by the discordance between its architecture, which appears well differentiated (papillary or glandular pattern), and its nuclear morphology, which is high grade (grade 3 nuclei) (Demopoulos et al. 1996). Areas containing clear cells do not preclude the diagnosis of serous carcinoma. Microscopically, the exophytic component of a serous carcinoma typically has a complex papillary architecture. The papillary fronds may be either short and densely fibrotic or thin and delicate. The cells covering the papillae and lining the glands form small papillary tufts, many of which

505

Fig. 43 Serous carcinoma. An area of endometrial intraepithelial carcinoma (serous intraepithelial carcinoma) is present within a background of atrophic endometrium. Markedly atypical epithelium replaces pre-existing endometrial glands

Fig. 44 Serous carcinoma. Detached papillary epithelial cell clusters are present within lymphatic spaces of endometrium

are detached and float freely in spaces between the papillae and in gland lumens. The cells are cuboidal or hobnail shaped and contain abundant granular eosinophilic or clear cytoplasm. The cells tend to be loosely cohesive. There may be considerable cytologic variability throughout the tumor, as many cells tend to show marked cytologic atypia manifested by nuclear pleomorphism, hyperchromasia, and macronucleoli whereas others are small and not so ominous in appearance. Multinucleated cells, giant nuclei, and bizarre forms occur in half the tumors. Lobulated nuclei with smudged chromatin also are frequently encountered. Mitotic activity usually is high and abnormal mitotic figures are easily

506

identified. Psammoma bodies are encountered in a third of cases. The invasive component of the neoplasm can show contiguous downgrowth of papillary processes, or solid masses or glands, the latter often have a gaping appearance. Nests of cells within vascular spaces are commonly found (Fig. 44). The adjacent endometrium in hysterectomy specimens with serous carcinoma is atrophic in almost all cases. Hyperplasia, generally without atypia, is present in less than 10% of the cases (Carcangiu and Chambers 1992; Sherman et al. 1992; Spiegel 1995). In nearly 90% of the cases, the surface endometrium adjacent to the carcinoma or at other sites away from the neoplasm is replaced by one or several layers of highly atypical cells that overlie atrophic endometrium and extend into normal glands. These cells are identical to those of the invasive carcinoma and at times form micropapillary processes. This lesion, which has been designated SEIC (Fig. 41), is discussed in detail in ▶ Chap. 8, “Precursors of Endometrial Carcinoma” (Spiegel 1995; Ambros et al. 1995). The intraepithelial carcinoma can extensively replace the surface endometrium and underlying glands without stromal invasion. The clinicopathologic features and distinction of extensive SEIC from early invasive serous carcinoma have been reported (Wheeler et al. 2000). It has been proposed that SEIC and serous carcinoma measuring 1 cm or less should be designated minimal uterine serous carcinoma because these lesions are difficult to distinguish and they behave in a similar fashion when confirmed as stage IA by meticulous surgical staging. It is important to recognize that patients whose uteri demonstrate only SEIC, without evidence of invasive serous carcinoma in the completely sampled endometrium, can have metastatic serous carcinoma in the ovary, peritoneum, or omentum, presumably as a result of exfoliation and implantation of the loosely cohesive tumor cells (Soslow et al. 2000b; Wheeler et al. 2000).

Differential Diagnosis Serous carcinoma must be distinguished from villoglandular carcinoma, which also has papillary architecture. Unlike serous carcinoma,

L. Hedrick Ellenson et al.

villoglandular carcinoma is characterized by the predominance of long, delicate papillary fronds that do not display papillary tufting. In addition, the cells are columnar, resembling cells in endometrioid carcinoma and lack high-grade nuclear atypia (see “Villoglandular Carcinoma”). A serous carcinoma with a prominent glandular pattern that lacks prominent papillary features may be confused with an endometrioid carcinoma. In this case, it is predominantly the nuclear morphology that aids in the distinction. The glands in an endometrioid carcinoma have a smooth luminal border and are lined by columnar cells with nuclei that are grade 1 or 2. Endometrioid carcinomas with grade 3 nuclei are almost always solid, not glandular. In contrast, the glands in a serous carcinoma are lined by cells with high-grade nuclei, some of which are hobnail shaped, thus imparting a scalloped luminal border to the glands. In addition, in most cases papillary tufts project or lie detached in the gland lumens. Immunohistochemical analysis can aid in the distinction of glandular serous carcinoma from endometrioid carcinoma. Several studies have demonstrated a very high frequency of strong, diffuse positivity for p53, or uncommonly a complete lack of staining, in serous carcinomas, and these patterns of staining are correlated with the presence of mutations in the TP53 gene (see above “Molecular Genetics of Endometrial Carcinoma”) (Tashiro et al. 1997a; Lax et al. 2000b; Sherman et al. 1995; Kovalev et al. 1998; Moll et al. 1996). In addition, most serous carcinomas demonstrate diffuse p16 expression which appears identical to the pattern observed in high-risk HPV-related endocervical adenocarcinoma but is unrelated to high-risk HPV (Yemelyanova et al. 2009a; Ansari-Lari et al. 2004). Serous carcinomas also have a relative lack of expression of ER and PR and very high proliferation indices as measured by immunohistochemical expression of Ki-67 compared to endometrioid carcinoma (Lax et al. 1998a; Carcangiu et al. 1990). In contrast, endometrioid carcinomas (particularly grade 1 and 2 tumors) frequently express hormone receptors and have lower proliferation indices (Lax et al. 1998a, b; Carcangiu et al. 1990). In addition, strong, diffuse

9

Endometrial Carcinoma

immunohistochemical expression of p53 is confined to a subset of grade 3 endometrioid carcinomas and is rarely encountered in lower-grade endometrioid carcinomas (Lax et al. 1998a, b). p16 expression in endometrioid carcinomas is typically patchy; even tumors with more extensive expression usually have interspersed negative patches or cells (usually one site IIIC Metastasis to pelvic and/or para-aortic lymph nodes IV IVA Tumor invades bladder and/or rectum IVB Distant metastasis (2) Mullerian adenosarcomaa Stage Definition I Tumor limited to uterus IA Tumor limited to endometrium/endocervix with no myometrial invasion IB Less than or equal to half myometrial invasion IC More than half myometrial invasion II Tumor extends to the pelvis IIA Adnexal involvement IIB Tumor extends to extrauterine pelvic tissue III Tumor invades abdominal tissues (not just protruding into the abdomen) IIIA One site IIIB > one site IIIC Metastasis to pelvic and/or para-aortic lymph nodes IV IVA Tumor invades bladder and/or rectum IVB Distant metastasis a

Note: Simultaneous tumors of the uterine corpus and ovary/pelvis in association with ovarian/pelvic endometriosis should be classified as independent primary tumors

epidemiology, pathology, molecular biology, cytogenetics, clinical course, and treatment of each of the named smooth muscle entities.

Evaluation of Smooth Muscle Neoplasms The most effective way to distinguish clinically benign from clinically malignant uterine smooth muscle neoplasms is through the use of

10

Mesenchymal Tumors of the Uterus

multivariate criteria, that is, criteria that consider several microscopic features as an ensemble (Bell et al. 1994; Longacre et al. 1997). These features include differentiated cell type within the smooth muscle group, presence and type of tumor necrosis, degree of cytologic atypia, mitotic index, and the relationship of the neoplasm to surrounding normal structures, including extrauterine sites.

Differentiated Cell Type The term usual smooth muscle differentiation denotes a pattern of differentiation recapitulating that of the constituent cells of the normal myometrium. Usual smooth muscle cells are elongated; possess distinct cell membranes; have readily apparent eosinophilic, sometimes fibrillar cytoplasm; and grow in a fascicular arrangement. Since malignant spindle cell tumors other than leiomyosarcoma arise in the uterus, desmin and caldesmon immunohistochemical stains are useful for ascertaining smooth muscle differentiation when a tumor’s appearance departs from the above description. These non-leiomyosarcoma spindle cell tumors include spindle cell variants of endometrial stromal sarcoma (Lewis et al. 2017; Oliva et al. 1999; Yilmaz et al. 2002), the sarcomatous component of adenosarcoma or carcinosarcoma, undifferentiated sarcoma (Kurihara et al. 2008), gastrointestinal stromal tumor, heterologous sarcoma (Fadare 2011), malignant solitary fibrous tumor (Baldi et al. 2013; Yang et al. 2017), and the extremely rare fibrosarcoma (Chiang et al. 2018). Epithelioid smooth muscle cells are round or polygonal and have eosinophilic to colorless cytoplasm. They may have perinuclear cytoplasmic vacuoles or there may be a perinuclear rim of eosinophilic cytoplasm while the rest of the cytoplasm is clear. When the cytoplasm is completely clear, the label “clear cell” is used. To distinguish perivascular epithelioid cell tumor (PEComa) from other myometrial epithelioid tumors, HMB45 and Melan-A stains are helpful (see section “PEComa and Related Lesions” at the end of this chapter). Other neoplasms in this category include metastatic (or locally invasive) carcinoma (cytokeratin positive), metastatic melanoma

539

(S-100 positive), placental site and epithelioid trophoblastic tumors (GATA-3 positive), and alveolar soft part sarcoma (ASPS) (HMB-45 negative with Xp11 translocation). Some endometrial stromal neoplasms may have an epithelioid appearance (Lee et al. 2012b). Myxoid smooth muscle proliferations feature widely spaced stellate cells with inapparent cytoplasm embedded in a myxoid matrix. Malignant myxoid smooth muscle neoplasms exhibit varying degrees of cytologic atypia and often have an appearance reminiscent of myxofibrosarcoma (myxoid malignant fibrous histiocytoma or myxofibrosarcoma) of the soft tissues. Tumors in this category include hydropic leiomyoma (edema fluid, not stromal mucin) (Clement et al. 1992), the vanishingly rare myxoid leiomyoma, myxoid change (Pugh et al. 2012), inflammatory myofibroblastic tumor (IMT) (ALK-rearranged) (Bennett et al. 2017a; Parra-Herran et al. 2015; Rabban et al. 2005), and fibromyxoid variants of endometrial stromal sarcoma (Lewis et al. 2017; Oliva et al. 1999; Yilmaz et al. 2002). Less common types of differentiation, such as fat and skeletal muscle, are discussed later (see section “Leiomyomas with Other Elements”).

Patterns of Necrosis The presence or absence and type of necrosis are powerful predictors of clinical behavior (Bell et al. 1994). Two patterns of necrosis in uterine smooth muscle tumors are diagnostically important: coagulative tumor cell necrosis and hyalinizing (or “infarction-type”) necrosis (Bell et al. 1994). Recognizing coagulative tumor cell necrosis is crucial because it is a key distinguishing feature of clinically malignant smooth muscle neoplasms. Coagulative tumor cell necrosis features an abrupt transition between necrotic and preserved cells (Fig. 1), where necrotic cells retain nuclear hematoxyphilia, and usually without associated inflammation or hemorrhage. The characteristic low-power microscopic pattern is one of blood vessels cuffed by viable cells surrounded by a sea of necrotic tumor (Bell et al. 1994; Clement 2000). In contrast, hyalinizing necrosis,

540

E. Oliva et al.

Fig. 1 Coagulative tumor cell necrosis. (a) Viable cells are present only around the blood vessel. Ghostlike outlines of necrotic atypical tumor cells can still be discerned

in the surrounding tissue. (b) Reticulin network is preserved. (c) Collagen deposition is not obvious (trichrome stain)

associated with damage following ischemia, has a distinctly zonal pattern with central necrosis, a more peripheral zone of granulation tissue, and, at the periphery, a variable amount of hyaline eosinophilic collagen interposed between the central degenerated region and peripheral preserved smooth muscle cells (Fig. 2). Hemorrhage is frequently present. When shadow cells or nuclei are discernible in the necrosis, there is little hyperchromasia or nuclear pleomorphism. One challenge in discerning coagulative tumor cell necrosis is its similarity to acute infarction (Fig. 3); both involve juxtaposition of preserved tumor adjacent to necrotic tumor. Coagulative tumor cell necrosis may be distinguished from acute ischemic necrosis by (1) the presence of hyperchromasia and nuclear pleomorphism in the “shadow cells” of the necrotic tumor and (2) (in the absence of these nuclear features) the absence of ongoing ischemia elsewhere in the

problematic smooth muscle neoplasm. The presence of viable, benign-appearing smooth muscle at the periphery of the devitalized tissue and hemorrhage argues strongly in favor of an early infarct, as does the absence of viable vessels within the infarct and sarcoma ghost cells. Unfortunately, there is a considerable degree of interobserver variation in the distinction between coagulative tumor cell necrosis and hyalinizing (postischemic) necrosis (Lim et al. 2013). Trichrome preparations are useful in detecting patchy foci of healed ischemic damage. A recent study showed that a combination of reticulin and trichrome histochemical stains and a Ki-67 immunostain helps discriminate these types of necrosis, although with imperfect accuracy. Most infarcts stain blue with trichrome, show reticulin loss, and, in some cases, display a rim of proliferation around the infarct that differs from more distant parts of the tumor. On the other hand, coagulative tumor cell necrosis usually

10

Mesenchymal Tumors of the Uterus

541

Fig. 2 Hyalinized necrosis. (a) An area of bland necrosis (N) is separated from viable spindle-shaped tumor cells (V) by a zone of hyalinized collagen (H). (b) Reticulin network is lost, in contrast to coagulative tumor cell

necrosis (Fig. 1b). (c) Hyalinizing necrosis. Collagen deposition is present (trichrome stain), unlike coagulative tumor cell necrosis (Fig. 1c)

but not always retains reticulin, but not trichrome staining, and exhibits less proliferation adjacent to the necrosis compared to viable portions of the tumor (Yang and Mutter 2015). Another pattern of necrosis that may be seen in ulcerated submucous leiomyomas features acute inflammatory cells and an associated zonal reparative process.

atypia is defined as nuclear hyperchromatism and pleomorphism that is obvious at scanning power (Fig. 4) (Bell et al. 1994). Neoplasms with this level of atypia often display enlarged and sometimes abnormal mitotic figures. Most commonly, moderate to severe atypia is present diffusely throughout the neoplasm, as in pleomorphic undifferentiated sarcoma (also termed malignant fibrous histiocytomas of the soft tissue), but it can, occasionally, be present only focally. In contrast, absent or mild atypia features uniform cells with no more than mild nuclear pleomorphism (Fig. 5), with fine to granular chromatin. The nuclei may be enlarged in comparison to those of the cells comprising the surrounding myometrium, but the enlargement is uniform throughout the tumor. More than one or two enlarged abnormal mitotic figures are sufficient to classify a tumor’s atypia as moderate to severe. Diffuse severe atypia may also manifest as uniformly enlarged, hyperchromatic cells (Fig. 6) that

Cytologic Atypia Several studies have demonstrated a relationship between cytologic atypia and clinical behavior in uterine smooth muscle neoplasms (Bell et al. 1994). The problem, as always, is defining “significant atypia” in a way that is reproducible and can be communicated to others. Bell et al. found that a two-tiered scheme of absent to mild atypia versus diffusely distributed moderate to severe atypia is reasonably reproducible, where moderate to severe

542

Fig. 3 Early infarct, mimicking coagulative tumor cell necrosis. Despite an abrupt transition from viable to necrotic tumor, the background lacks atypia and the necrotic focus does not contain hyperchromatic, atypical ghost cells

E. Oliva et al.

Fig. 6 Severe uniform atypia. Relatively uniform malignant cells exhibiting nuclear hyperchromasia and a high mitotic index

are difficult to discern as “atypical” on scanning magnification because nuclei do not appear pleomorphic; this pattern is analogous to that seen, for example, in monophasic synovial sarcoma of the soft tissues. Despite the absence of pleomorphism, this type of atypia must still be recognized; one key is to compare the constituent cells of the tumor to the surrounding normal myocytes to look for nucleomegaly and hyperchromasia of the neoplastic cells. A helpful adjunct in diagnosing leiomyosarcomas with this type of uniform severe atypia is the variably associated finding of infiltration of the surrounding myometrium.

Mitotic Index Fig. 4 Severe pleomorphic atypia. Nuclear pleomorphism against a background of diffuse severe atypia of spindled cells

Fig. 5 Diffuse mild atypia. Uniform mild atypia characterizes this infiltrating smooth muscle neoplasm

Mitotic index is expressed in terms of the number of definite mitotic figures per 10 high-power fields (Hilsenbeck and Allred 1992; van Diest et al. 1992). Whether intensive mitosis counting is required depends on whether significant cytological atypia or tumor cell necrosis is present. In the absence of these two features the precise mitotic index is of little importance below 20 mitotic figures per 10 high-power fields (MF/10 HPF). The mitotic index is determined by searching the slide at low magnification for the most mitotically active area, then counting the mitotic figures within that area at high magnification in five sets of 10 randomly chosen contiguous fields. Care must be taken not to count lymphocytes, karyorrhectic debris, precipitated hematoxylin, or mast cells as mitotic figures. Most reliable are mitotic figures in metaphase,

10

Mesenchymal Tumors of the Uterus

anaphase, or telophase. The reproducibility of mitotic counts may depend on consistent counting techniques (O’Leary and Steffes 1996) or on accurately defining a mitotic figure. Phosphohistone H3 staining can be used as an adjunct to traditional mitotic counting (Chow et al. 2017).

Relationship to Surrounding Normal Structures and Anatomic Distribution Another indicator of a smooth muscle neoplasm’s aggressiveness is its relationship to the surrounding myometrium and uterine vessels and whether it extends beyond the uterus. Infiltrative margins, intravascular growth, and extrauterine spread, although commonly encountered in uterine malignancies, are not, when seen as isolated findings, diagnostic of sarcoma. Some relatively rare benign or clinically low-grade smooth muscle proliferations mimic leiomyosarcoma by virtue of their relationship to normal uterine structures or their extrauterine extension.

Leiomyoma Leiomyomas are the most common uterine neoplasms (Payson et al. 2006; Vollenhoven 1998). They are noted clinically in 20–30% of women over 30 years of age, and are found in as many as 75% of uteri upon systematic searching (Baird et al. 2003; Cramer and Patel 1990; Payson et al. 2006). Leiomyomas are usually detected in middle-aged women and are uncommon in women less than 30 years of age; however, the youngest patient on record was 13 years old. Some leiomyomas apparently shrink after menopause, but their frequency does not decrease. Leiomyomas are more common in African American women than in white women (Kjerulff et al. 1996). The growth of leiomyomas is affected by the hormonal milieu (Andersen 1998; Marsh and Bulun 2006; Rackow and Arici 2006; Sozen and Arici 2006), as they contain estrogen receptors (ER) and progesterone receptors (PR) (Viville et al. 1997). Leiomyomas may grow larger during

543

estrogen therapy, and most become smaller when the patient is treated with a gonadotropin-releasing hormone (GnRH) agonist (Adamson 1992; Regidor et al. 1995; Shaw 1998; Stovall et al. 1991; Upadhyaya et al. 1990). Progestins, progesterone, hormone replacement therapy, clomiphene use, and pregnancy occasionally are associated with rapid leiomyoma growth and sometimes produce hemorrhagic degeneration (Sener et al. 1996).

Clinical Features The clinical presentation of leiomyomas depends on their size and location (Bukulmez and Doody 2006). Leiomyomas cause many signs and symptoms, the most common of which are pain, a sensation of pressure, and abnormal uterine bleeding. Even small leiomyomas, when submucosal, can cause bleeding due to compression of the overlying endometrium and compromise of its vascular supply. In some instances, infertility is attributed to the presence of leiomyomas. Large tumors can be detected during pelvic examination because they cause uterine enlargement or an irregular uterine contour. Some leiomyomas are pedunculated and protrude through the cervical os. On rare occasions, subserosal pedunculated leiomyomas undergo torsion, infarction, and separation from the uterus. Secondary infection of leiomyomas can result in fever, leukocytosis, and an elevated sedimentation rate. Among the complications of pregnancy ascribed to leiomyomas are spontaneous abortion, premature rupture of membranes, dystocia, inversion of the uterus, and postpartum hemorrhage.

Gross Findings Despite the variety of histologic subtypes of leiomyoma, many are grossly similar. Multiple leiomyomas are present in two-thirds of women with these neoplasms (Cramer and Patel 1990). Leiomyomas are spherical and firm and bulge above the surrounding myometrium. The cut surfaces are white to tan, with a whorled trabecular pattern (Fig. 7). Leiomyomas can be located

544

anywhere in the myometrium but are most commonly intramural. Submucosal leiomyomas compress the overlying endometrium and bulge into the endometrial cavity as they enlarge. Very rarely, some become attached to another pelvic structure (parasitic leiomyoma). The appearance of a leiomyoma is commonly altered by degenerative changes, such as hemorrhage, appearing as dark red areas, and necrosis, which can be recognized as sharply demarcated yellow areas. These features are more common in large leiomyomas and those of women who are pregnant or undergoing high-dose progestin therapy. Submucosal leiomyomas are frequently ulcerated and hemorrhagic. The hemorrhagic infarction damages smooth muscle, which is eventually replaced by firm white or translucent collagenous tissue. Cystic degeneration also occurs, and some leiomyomas become extensively calcified. The precise locations of leiomyomas may be determined by transvaginal ultrasound, which in complex cases is complemented by magnetic resonance imaging (Dueholm et al. 2002; Vitiello and McCarthy 2006).

E. Oliva et al.

Fig. 7 Enlarged uterus containing multiple leiomyomas. The leiomyomas have a whorled white-tan cut surface that bulges above the surrounding myometrium

Microscopic Findings Typical leiomyomas are composed of whorled, anastomosing fascicles of uniform fusiform smooth muscle cells. The spindle-shaped cells have indistinct borders and abundant fibrillar eosinophilic cytoplasm (Fig. 8). Nuclei are elongated with blunt or tapered ends and have finely dispersed chromatin and small nucleoli. Mitotic figures usually are infrequent. Most leiomyomas are more cellular than the surrounding myometrium; those that are not are identified by their nodular circumscription and by the disorderly arrangement of the smooth muscle fascicles within them, which are out of alignment with the surrounding myometrium. The degenerative changes mentioned above are also apparent microscopically. Hyaline fibrosis is present in more than 60%, particularly in postmenopausal women (Cramer et al. 1996). Edema is present in about 50% of leiomyomas, and, on occasion, marked hydropic change can

Fig. 8 Typical leiomyoma. The spindle-shaped tumor cells have cytologically bland, relatively uniform nuclei with fine chromatin and small nucleoli. The cytoplasm is abundant, eosinophilic, and fibrillar

mimic the appearance of a myxoid smooth muscle tumor or produce a pattern that can be confused with intravenous leiomyomatosis (IVL) (Clement et al. 1992; Coad et al. 1997). About 10% of leiomyomas contain significant areas of hemorrhage, which tend to be zonal and sharply demarcated, and cystic degeneration and microcalcification each occurs in about 4%. In addition to hemorrhage, edema, myxoid change, hypercellular foci, and cellular hypertrophy are also particularly frequent in the leiomyomas of women who are pregnant or taking progestins

10

Mesenchymal Tumors of the Uterus

(see following, section “Apoplectic Leiomyoma/ Hemorrhagic Cellular Leiomyoma”) (Bennett et al. 2016). Progestational agents are associated with a slight increase in mitotic activity but not to the level observed in a leiomyosarcoma. The margins of most leiomyomas are microscopically circumscribed, but some benign tumors interdigitate with the surrounding myometrium, occasionally extensively (see following, section “Dissecting Leiomyomas”). Submucous leiomyomas, particularly if they protrude into the endometrial cavity, may display extensive necrosis, often with acute inflammatory cells, unlike the necrosis common in leiomyosarcoma. The necrosis in these tumors is also distinguishable from that in malignant tumors by the inconspicuousness or absence of cell outlines. Not infrequently, areas of necrosis are accompanied by adjacent regions of increased mitotic activity, but these regions’ mitotic figures have normal morphology and tend not to be associated with significant nuclear atypia.

Immunohistochemistry Smooth muscle cells in the myometrium and within smooth muscle tumors retain staining with antibodies to muscle-specific actin, alpha-smooth muscle actin, desmin, and caldesmon (Eyden et al. 1992), and to a lesser degree with vimentin. ER are frequently expressed, as well as WT1 to a lesser degree; ER staining is used to determine whether extrauterine smooth muscle tumors are of gynecologic type (Lee et al. 2009). Leiomyomas frequently stain positive for cytokeratin, similar to the surrounding myometrium, the extent and intensity of reactivity depending on the antibodies used and the fixation of the specimen (Brown et al. 1987; Eyden et al. 1992; Gown et al. 1988). Epithelial membrane antigen (EMA) is usually negative in smooth muscle tumors.

Molecular Pathology Leiomyomas are a proliferation of a single clone of smooth muscle cells, as evidenced by nonrandom

545

inactivation of the X chromosome observable through glucose-6-phosphate dehydrogenase isoform expression and other techniques. Cytogenetic studies provide further evidence of these tumors’ clonal nature (Quade 1995). Leiomyomas’ cytogenetic abnormalities have recently become the focus of much research. While early cytogenetic alterations are considered by some investigators to be insufficient for tumor development, dysregulation of multiple signaling pathways may be transformative. Pathways that may contribute to the development of leiomyomas include steroids, growth factors, transforming growth factor-beta (TGF-β)/Smad, wingless-type (Wnt)/β-catenin, retinoic acid, and vitamin D, which converge in synergistic ways (Borahay et al. 2015). Approximately 40% of uterine leiomyomas have chromosomal abnormalities detectable by conventional cytogenetic analysis, including t (12;14)(q15;q23-24), rearrangements involving the short arm of chromosome 6, and interstitial deletions of the long arm of chromosome 7 (Lobel et al. 2006; Quade 1995; Sornberger et al. 1999). More recent work has established that, from a genomic standpoint, there are at least four mostly nonoverlapping categories of uterine leiomyomas, some of which have interesting clinicopathological correlates (Markowski et al. 2015; Mehine et al. 2013, 2014). Listed in roughly decreasing order of prevalence are mediator subcomplex 12 (MED12) point mutation or deletion with consequent upregulation of RAD51B (60–70% of cases) (Mehine et al. 2013; Quade et al. 2003; Schoenmakers et al. 1999); high mobility group protein AT-hook 2 gene (HMGA2) overexpression with RAD51B as an enhancer (10–20% of cases); fumarate hydratase (FH) inactivation; and COL4A6-COL4A5 deletion. HMGA2 overexpression is due to the t(12;14)(q15;q23-24) translocation or complex rearrangements involving the HMGA2 gene that occur in the setting of multiple interconnected rearrangements (also referred to as chromothripsis) (Markowski et al. 2015; Mehine et al. 2013). These complex chromosomal rearrangements tend to be found most often in leiomyomas lacking MED12 mutations (Mehine et al. 2014). Rare leiomyomas with both MED12

546

and HMGA2 abnormalities have been reported (Holzmann et al. 2015), while chromosome 7q deletions may occur with or without either MED12 and HMGA2 mutations (Markowski et al. 2012; Mehine et al. 2013). Chromosome 7 deletions may target yet another gene, CUX1 (Schoenmakers et al. 2013). Leiomyomas with HMGA2 rearrangements usually occur as single nodules, while leiomyomas with MED12 mutations present as multiple tumors; this has led to the hypothesis that MED12 mutations might be present in stemlike mesenchymal cells, which can proliferate in myometrium, leading to a “leiomyoma field effect.” MED12-mutated leiomyomas tend to be smaller than those with HMGA2 rearrangements (Markowski et al. 2015). Certain abnormalities cluster with specific leiomyoma subtypes. Namely, recurrent loss of 22q12.3-q13.1 is linked to IVL (Buza et al. 2014). Loss of 1p, often in combination with other aberrations, particularly loss of chromosomes 19 and/or 22, is associated with cellular leiomyoma (Christacos et al. 2006); 19q and 22q terminal deletions are often seen in benign metastasizing leiomyoma (Mehine et al. 2013); FH deficiency is linked to leiomyomas (Gross et al. 2004; Lehtonen et al. 2004; Mehine et al. 2014); and certain types of leiomyomas with bizarre nuclei, some with somatic abnormalities in the FH gene and rarely germline mutations as discussed subsequently (Bennett et al. 2017b; Gunnala et al. 2017).

Clinical Behavior and Treatment Most leiomyomas are asymptomatic, and only a minority requires treatment. Therapy is indicated only if leiomyomas are symptomatic, interfere with fertility, enlarge rapidly, or pose diagnostic problems (Bukulmez and Doody 2006; Ouyang et al. 2006; Wallach and Vlahos 2004). Sometimes they can be excised (myomectomy), but if they are large or multiple, a hysterectomy may be required. Treatment with leuprolide acetate or another gonadotropinreleasing hormone agonist (GnRHa), which

E. Oliva et al.

drastically lowers estrogen levels by causing pituitary desensitization, results in shrinkage of leiomyomas, a decrease in uterine volume, and alleviation of the patient’s symptoms (Marsh and Bulun 2006; Rackow and Arici 2006; Shaw 1998). The maximum effect is noted after 8–12 weeks, but the leiomyomas increase in size again with cessation of GnRH agonist therapy. Such therapy can be used before surgery to decrease uterine size (facilitating myomectomy or permitting vaginal rather than abdominal hysterectomy) and to reduce the risk of hemorrhage during surgery. Because GnRH agonists have unpleasant side effects such as hot flashes and have the potential to reduce bone mass loss and cause cardiovascular changes, alternatives have been sought. A more direct means of reducing estrogen is by GnRH antagonists; evidence suggests that they are effective and act rapidly without causing an initial flare in steroid levels, as is caused by GnRHa (Flierman et al. 2005). Finally, leiomyomas can be treated by uterine artery embolization, which leads to ischemia and tumor involution (Marshburn et al. 2006; Siskin et al. 2006; Spies et al. 2001). This method of treatment is of potential interest to pathologists, because (1) if a hysterectomy is subsequently necessary, there may be areas of ischemic necrosis in a leiomyoma that must be differentiated from the type of tumor cell necrosis seen in leiomyosarcoma, and (2) the embolic particles may cause confusion unless the pathologist can recognize them (Dundr et al. 2006; McCluggage et al. 2000; Weichert et al. 2005). Morcellation (a minimally invasive/laparoscopic technique that fragments myomas within the peritoneal cavity) of leiomyosarcomas is well known to lead to aggressive recurrence within the peritoneal cavity (discussed in detail later in this chapter), but it is perhaps less well publicized that morcellated leiomyomas may also recur intraperitoneally (Tulandi et al. 2016), though the clinical course in that case is usually not aggressive. Uncontained morcellation has also been suggested to lead to iatrogenic endometriosis and adenomyomatosis (Tulandi et al. 2016).

10

Mesenchymal Tumors of the Uterus

547

These findings prompted a number of gynecologic societies to issue recommendations regarding the appropriate use of morcellation for leiomyomas. The consensus is that the procedure can continue to be performed, particularly in women younger than 50 years and those who desire fertility. Current recommendations advise containing the morcellation within a bag that prevents peritoneal spillage and preoperative assessment for malignancy, the presence of which would discourage morcellation (Halaska et al. 2017; Hall et al. 2015). The problem, of course, is that a diagnosis of leiomyosarcoma is rarely evident preoperatively, even with the use of endometrial biopsy and imaging. However, qualitative magnetic resonance imaging has recently been reported to have significant discriminative power in distinguishing atypical leiomyomas and leiomyosarcomas (Lakhman et al. 2017). These recommendations then prompted a number of clinical studies that emphasized the rarity of morcellating a “fibroid” that turns out to be a leiomyosarcoma. Two studies, one of which included 10,119 patients, reported this to occur in 0.09% of such procedures (Seidman et al. 2012) (Cui and Wright 2016; Kho et al. 2016). Some practitioners still advocate for morcellation (Parker et al. 2016), especially since the procedure, when performed for leiomyomas in young women, is cost-effective and results in low rates of intra- and perioperative complications (Cui and Wright 2016).

Occasionally, a typical-appearing leiomyoma in a premenopausal woman will have 5 MF/10 HPF (Figs. 9a, b); these are designated mitotically active leiomyomas (Bell et al. 1994; Dgani et al. 1998; O’Connor and Norris 1990; Perrone and Dehner 1988; Prayson and Hart 1992). The number of mitotic figures is usually 5–9 MF/10 HPF, but occasional mitotically active leiomyomas with 10–20 MF/10 HPF have been reported. Some pathologists use 15 MF/10 HPF as the upper limit for a diagnosis of mitotically active leiomyoma and use the designation “smooth muscle tumor of uncertain malignant potential” (STUMP) when proliferative rates exceed that, whereas other pathologists use 20 MF/10 HPF as the upper limit (Bell et al. 1994). These tumors’ clinical evolution is benign, even if the neoplasm is treated by myomectomy. Their increased mitotic index may result from increased progesterone levels during the secretory phase of the menstrual cycle (Kawaguchi et al. 1989) or from progestin-only birth control (Tiltman 1985). The patient’s hormonal status may also contribute to the increased number of mitotic figures seen in mitotically active leiomyomas (Prayson and Hart 1992). This diagnosis must not be used for neoplasms that exhibit moderate to severe nuclear atypia, contain abnormal mitotic figures, or demonstrate zones of tumor cell necrosis, as these may be signs of malignancy.

Specific Subtypes of Leiomyoma

Cellular Leiomyoma

Most subtypes of leiomyoma are chiefly of interest in that they mimic malignancy in one or more respects. These subtypes are mitotically active leiomyoma, cellular leiomyoma, apoplectic leiomyoma (hemorrhagic cellular leiomyoma), fumarate hydratase-deficient leiomyoma, leiomyoma with bizarre nuclei, epithelioid leiomyoma, and myxoid leiomyoma. Other leiomyoma variants, vascular leiomyoma, leiomyoma with other elements, and leiomyomas with hematopoietic elements, are more curiosities than diagnostic problems.

A cellular leiomyoma is one in which the cellularity is “significantly” greater than the surrounding myometrium. Less than 5% of leiomyomas fall within this category. These tumors are described macroscopically as “fish flesh,” “rubbery,” and “soft,” whereas typical leiomyomas are firm (Oliva et al. 1995). Hypercellularity may suggest a diagnosis of leiomyosarcoma, but cellular leiomyoma lacks tumor cell necrosis, has few mitotic figures, and lacks the moderate to severe cytologic atypia seen in leiomyosarcoma. Some cellular leiomyomas

Mitotically Active Leiomyoma

548

Fig. 9 Mitotically active leiomyoma. (a) Two mitotic figures (center) in a non-hypercellular tumor without nuclear atypia. (b) High magnification view of the mitotic

display palisading of nuclei, reminiscent of that seen in the Verocay bodies of a neurilemoma, but their ultrastructural appearance is that of an ordinary leiomyoma (Gisser and Young 1977). Prolapse of submucosal leiomyomas may result in accentuated cellularity (McCluggage et al. 1999). Cellular leiomyomas composed of small cells with scanty cytoplasm can be difficult to distinguish from endometrial stromal tumors, especially in a “highly cellular leiomyoma” (Fig. 10) (Oliva et al. 1995). Features that help distinguish a cellular leiomyoma from a stromal tumor are the spindled shape of the cells, a fascicular growth pattern, and the absence of a plexiform vasculature. In a leiomyoma, the reticulin fibers tend to parallel the fascicles of cells, but they surround individual tumor cells in an endometrial stromal tumor. Additionally, Oliva et al. emphasized the presence of large thick-walled muscular vessels as features that serve to distinguish a highly cellular leiomyoma from a stromal proliferation (Oliva et al. 1995). Although some reports indicate that smooth muscle cells and stromal cells have immunophenotypic similarities, marked diffuse staining with muscle markers, particularly desmin (Fig. 11), is more suggestive of a smooth muscle tumor than of a stromal neoplasm (Oliva et al. 1995). Whether the lesion is invasive is best observed on a hysterectomy specimen. In the absence of myoinvasion or vascular invasion, the differential lies between two benign conditions: highly cellular leiomyoma and stromal nodule. When there is intravascular tumor,

E. Oliva et al.

figures in (a) surrounded by tumor cells of bland morphology with no necrosis

Fig. 10 Highly cellular leiomyoma. Cells are small and rounded and contain scanty cytoplasm

Fig. 11 Highly cellular leiomyoma. The tumor cells show strong cytoplasmic staining for desmin and immunoreactivity for smooth muscle actin and caldesmon but staining for CD10 was negative

10

Mesenchymal Tumors of the Uterus

the distinction between endometrial stromal and smooth muscle differentiation becomes clinically relevant as the differential diagnosis lies between stromal sarcoma (a clinically low-grade malignancy) and IVL (clinically benign unless there are cardiac or pulmonary complications). When a cellular mesenchymal proliferation is recovered in an endometrial sampling, care must be taken to determine whether the growth is benign or, much more rarely, a stromal sarcoma. Three issues need to be considered: (1) what differentiation does the proliferation exhibit (smooth muscle or endometrial stromal); (2) are the criteria of malignancy evaluable; and, finally, (3) are the criteria of malignancy met? Rarely, some uterine neoplasms appear to be composed of a mixture of stromal and smooth muscle cells (Bell et al. 1994; Oliva et al. 1998, 2007); these are currently categorized as endometrial stromal neoplasms, not “mixed endometrial stromal and smooth muscle neoplasms.” Regardless of the differentiation present, unequivocal diagnoses of “endometrial stromal sarcoma” on biopsy or curettage should be avoided since that would require a therapeutic hysterectomy. For older patients or young patients with no interest in having children, the issue is usually resolved by what amounts to a diagnostic hysterectomy. For premenopausal women wishing to retain fertility and older patients who are poor surgical candidates, diagnostic modalities such as hysteroscopy, imaging studies, repeat sampling, and immunohistochemistry should be considered, although conservative clinical management does not necessarily constitute the standard of care.

Apoplectic Leiomyoma (Hemorrhagic Leiomyoma) Hemorrhagic cellular leiomyoma, or “apoplectic” leiomyoma (Bennett et al. 2016; Oliva 2016), is a form of cellular leiomyoma that may occur in women who are taking oral contraceptives or who are either pregnant or postpartum, but many exceptions to these generalizations are on record (Myles and Hart

549

1985; Norris et al. 1988). Grossly, these tumors frequently exhibit hemorrhage (sometimes multifocal and stellate in shape), infarct-type necrosis, cyst formation, softening, and brownred discoloration (Fig. 12) (Bennett et al. 2016). Microscopically, the leiomyoma is cellular and contains patchy areas of hemorrhage and edema. Hemorrhagic areas are surrounded by a narrow zone resembling granulation tissue, around which the tumor cells show slightly increased number of mitotic figures. These organizing infarcts are usually easy to distinguish from the more ominous coagulative tumor cell necrosis, though early infarcts may appear similar, with abrupt transitions between viable and nonviable tissue. Nonetheless, the presence of hemorrhage and lack of significant nuclear atypia both within ghost cells and viable tumor cells should distinguish this type of infarct. In contrast to leiomyosarcoma, neither atypical mitotic figures nor significant cytologic atypia is present, and the neoplasm has a circumscribed, compressive margin.

FH-Deficient Leiomyoma Recognition of FH-deficient leiomyomas as a distinct class followed from the discovery that hereditary leiomyomatosis and renal cell carcinoma syndrome (HLRCC) is underpinned by germline mutations in FH (Stewart et al. 2008; Toro et al. 2003; Wei et al. 2006). Women in HLRCC-affected families often develop leiomyomas before age 40. In uterine tumors from these families, the smooth muscle cells contain viral inclusion-like eosinophilic macronucleoli with peri-nucleolar halos, similar to what is seen in syndromic renal cell carcinomas (Garg et al. 2011; Sanz-Ortega et al. 2013). Studies indicating that somatic FH abnormalities are more common than germline mutations also identified additional characteristic features (Miettinen et al. 2016; Reyes et al. 2014). These include round-to-oval nuclei in spindled smooth muscle cells with overtly fibrillary cytoplasm that sometimes also includes globoid eosinophilic bodies, the aforementioned

550

E. Oliva et al.

Fig. 12 Apoplectic leiomyoma. (a) Multiple foci of hemorrhage are visible on the cut surfaces of a myomectomy specimen. (b): “Zonation” phenomenon: Viable smooth muscle, granulation-like tissue, and hemorrhage. (c):

Smooth muscle surrounding this hemorrhagic infarct is mitotically active. Intact leiomyoma distant from the infarct showed a low mitotic index

nuclear characteristics, and a staghorn vasculature (Fig. 13). Nuclei are sometimes arranged in moreor-less linear rows, resulting in nuclear-dense and nuclear-free zones, somewhat reminiscent of “neurilemomatous leiomyoma.” These features are characteristic of FHdeficient leiomyoma, but they may not be specific to tumors carrying FH mutations (Alsolami et al. 2014; Miettinen et al. 2016; Reyes et al. 2014). Immunohistochemical stains can identify loss of FH protein (commercially available anti-FH antibodies) or the downstream metabolic changes associated with FH loss (antibodies against 2-succinyl cysteine, a protein modification caused by FH deficiency) (Bardella et al. 2011; Buelow et al. 2016; Joseph et al. 2015). These cells’ aberrations in the mitochondrial enzyme FH are thought to make them depend more heavily on glycolysis for their energy needs; this has been named the “Warburg effect” (Warburg 1956).

Although the Warburg effect is now usually thought of as an epiphenomenon or a downstream effect of oncogenes or tumor suppressor genes in other tumor types, in FH-deficient tumors, it likely has direct oncologic manifestations (Yang et al. 2012), perhaps through protein succination or fumarate accumulation. Patients with HLRCC may develop multiple cutaneous and uterine leiomyomas and an aggressive form of renal carcinoma that has been described as part of the spectrum of type II papillary renal cell carcinoma (Launonen et al. 2001; Tomlinson et al. 2002). As cutaneous and uterine leiomyomas may be sentinel disease manifestations, their recognition could enable early diagnosis of potentially lethal renal cell carcinomas. Cutaneous and uterine leiomyomas are present in 75% and 80% of HLRCC patients, respectively (Lehtonen et al. 2006; Sanz-Ortega et al. 2013; Toro et al. 2003; Wei et al. 2006), but in

10

Mesenchymal Tumors of the Uterus

551

Fig. 13 FH-deficient leiomyoma. (a) The tumor is more cellular than surrounding myometrium and contains fusiform cells with oval nuclei and a staghorn vasculature. Nuclear-free and nuclear-dense zones can be appreciated. (b) Intermediate power magnification emphasizing the

staghorn vasculature. (c) High power magnification reveals round nuclei with prominent eosinophilic nucleoli with perinuclear clearing and eosinophilic cytoplasmic globules. (d) Tumor cells are negative for FH, while expression is retained in normal constituents

40% of patients with cutaneous leiomyomas, the tumors’ presentation is subtle (Toro et al. 2003; Wei et al. 2006). Cutaneous leiomyomas typically present around 25 years of age, whereas uterine leiomyomas are characteristically multiple and occur at a mean age of 30 years. Renal cell carcinomas are present in only 20–30% of kindreds and occur at an average age of 46 years (Gardie et al. 2011; Merino et al. 2007). HLRCC is therefore an incompletely penetrant syndrome, with a wide spectrum of disease ranging from asymptomatic to lethal. Somatic FH abnormalities, which may affect as few as 2–3% of leiomyomas, are far more common than germline mutations. When deciding

which patients should be considered for germline genetic testing, one must account for not only the immunohistochemical and morphological features of the uterine leiomyoma but also the patient’s characteristics. Patients who have a large FH-deficient leiomyoma and present before 35–40 years of age are most likely to have a germline mutation, especially when cutaneous leiomyomas are also present.

Leiomyoma with Bizarre Nuclei Several different terms have been used to describe leiomyomas with bizarre nuclei, including

552

E. Oliva et al.

symplastic leiomyoma and atypical leiomyoma (Bennett et al. 2017b; Croce et al. 2014; Ly et al. 2013). Microscopically, leiomyoma with bizarre nuclei is distinguished by moderate to severe cytologic atypia, a feature shared with malignant uterine smooth muscle tumors, but in this case high mitotic activity (>10 MF/HPF) and tumor cell necrosis are absent (Bell et al. 1994; Downes and Hart 1997). The atypical cells may be distributed throughout the leiomyoma or they may be focal; they have enlarged hyperchromatic nuclei with prominent chromatin clumping and, often, smudging (Fig. 14), as well as large cytoplasmic pseudonuclear inclusions. Multinucleated tumor giant cells can be numerous and prompt the name “bizarre” or “symplastic” leiomyoma.

Mitotic counting in these tumors may be complicated by the frequent presence of bizarre-appearing karyorrhexis and pyknotic nuclei that mimic the appearance of highly atypical mitotic figures. Recognition of true mitotic figures can be facilitated with a phosphohistone-H3 immunohistochemical stain (Chow et al. 2017; Veras et al. 2009), but even when relying on this stain, only figures that resemble mitoses should be counted. When performing a traditional mitotic count using hematoxylin and eosin (H&E), it is often helpful to carefully assess areas of the tumor that are not overtly bizarre in appearance. If mitotic activity is readily apparent in these areas, the lesion in question may not be best diagnosed as leiomyoma with bizarre nuclei. Scrutiny of such areas may also reveal the presence of a population of monomorphic, enlarged and

Fig. 14 Leiomyoma with bizarre nuclei. (a) Atypical cells are distributed throughout. (b) Tumor cells have single or multiple pleomorphic nuclei with coarse, smudged

chromatin. (c) High power magnification reveals prominent eosinophilic nucleoli with perinuclear clearing, suggesting that this bizarre leiomyoma is FH deficient

10

Mesenchymal Tumors of the Uterus

hyperchromatic nuclei; a mitotic index of >10 MF/HPF in these areas would place the tumor in the leiomyosarcoma category. Because leiomyosarcomas can vary greatly in appearance and may contain areas that lack the typical features of hypercellularity, cytologic atypia, and increased mitotic activity, extensive sampling is important to rule out that diagnosis. Leiomyoma with bizarre nuclei is less common in postmenopausal women, so a careful search for other features of leiomyosarcoma is indicated when a smooth muscle tumor containing atypical cells is detected in an older woman. This is especially important in consideration of the possibility that leiomyoma with bizarre nuclei may represent a precursor to leiomyosarcoma (Mittal and Joutovsky 2007; Mittal et al. 2009). To detect such cases, one must liberally sample and carefully examine both pleomorphic portions of the tumor for increased mitoses and spindle cell areas for nuclear enlargement and hyperchromasia, elevated mitotic index, and coagulative tumor cell necrosis. The clinical course of smooth muscle neoplasms featuring mitotic indices of less than 10 MF/10 HPF that lack tumor cell necrosis and that exhibit diffuse moderate to severe atypia remains controversial. In one series of 24 such cases, all had a benign clinical course (Downes and Hart 1997), while another series of 43 cases, all of which had at least 2 years follow-up, included only a single malignancy (2%), though progression took place over several years, which is much slower than in leiomyosarcoma (Bell et al. 1994). The latter study included only uterine smooth muscle neoplasms with moderate to severe atypia, a mitotic index less than 10 MF/10 HPF, and without tumor cell necrosis. In a subsequent study of patients whose tumors had less than 4 MF/10 HPFs (Ly et al. 2013), recurrences were rare. Only one patient (130 g for nulliparous, >210 g for parity 1–3, and > 250 g for parity of 4 and above) only (Langlois 1970). Uterine weight increases with age and with increasing parity until menopause and decreases after.

Dissecting Leiomyoma Dissecting leiomyoma refers to a benign smooth muscle proliferation where the border consists of compressive tongues of smooth muscle that penetrate into the surrounding myometrium and, occasionally, into the broad ligament and pelvis (Roth and Reed 1999). This pattern of infiltration may also be seen in IVL (see following). When edema and congestion are prominent, a uterine dissecting leiomyoma with extrauterine extension may resemble placental tissue, hence the name cotyledonoid dissecting leiomyoma (Fukunaga and Ushigome 1998; Roth and Reed 2000; Roth et al. 1996) (Fig. 19).

IVL and Leiomyoma with Vascular Invasion IVL is a very rare smooth muscle tumor characterized by nodular masses of histologically benign smooth muscle cells growing within venous channels (Clement 1988; Cohen et al. 2007; Mulvany et al. 1994; Nogales et al. 1987; Norris and

557

Parmley 1975). These tumors affect women at a median age of 45 years; few patients with IVL are younger than 40 years. The condition is not associated with a history of infertility or decreased parity and affects all ethnicities equally. The main symptoms are abnormal bleeding and pelvic discomfort, and most patients present with a pelvic mass. Grossly, IVL is a complex coiled or nodular growth within the myometrium with convoluted, wormlike extensions into the uterine veins in the broad ligament or into other pelvic veins (Fig. 20). The growth extends into the vena cava in more than 10% of patients, and in some it reaches as far as the heart (Clement 1988; Cohen et al. 2007; Kokawa et al. 2002; Suginami et al. 1990). The wormlike masses vary from soft and spongy to rubbery and firm, and their color is pink-white or gray. Intravenous growth with an IVL-like pattern has been described in leiomyosarcoma (Coard and Fletcher 2002), so it is important to carefully assess any such growth for features that might signify malignancy (high mitotic rate, significant nuclear atypia, tumor cell necrosis). As mentioned previously, IVL may have a peculiar karyotype, showing recurrent loss of 22q12.3-q13.1 (Buza et al. 2014). Microscopically, IVL is found within venous channels lined by endothelium (Fig. 21). They have a highly variable histologic appearance, even within the same tumor. Some have a similar cellular composition to that of a

Fig. 19 Dissecting leiomyoma. (a) In broad ligament, (b) associated with IVL (Fig. 21)

558

Fig. 20 IVL. Brown and white plugs of intravascular tumor grow extensively in the myometrium

Fig. 21 IVL. A plug of smooth muscle tumor grows within a large vein in the myometrium

leiomyoma, but most contain prominent zones of fibrosis or hyalinization, sometimes making smooth muscle cells inconspicuous and difficult to identify. Cells within IVLs display the same range of smooth muscle differentiation as in a leiomyoma (Clement 1988). The intravenous growth is itself highly vascular (Fig. 22), and in some cases contains so many small and large blood vessels that it may resemble a vascular tumor. Cellular, atypical, epithelioid, and lipoleiomyomatous growth patterns have all been described; these have the same behavior and prognosis as ordinary IVL (Brescia et al. 1989; Clement 1988; Han et al. 1998). IVL can originate in vascular smooth muscle (Norris and Parmley 1975); in these cases, the tumor is predominantly or entirely intravascular, with many sites of attachment to the vein walls.

E. Oliva et al.

Fig. 22 IVL. Nearly the entire tumor lies within vascular spaces. The tumor is itself highly vascular and extensively hyalinized

Others develop by intravascular extension from a leiomyoma (Nogales et al. 1987; Norris and Parmley 1975), in which cases the bulk of the tumor is extravascular and sites of origin from a vein wall are not found. IVL is treated by total abdominal hysterectomy and bilateral salpingo-oophorectomy together with excision of any extrauterine extensions. The condition has a favorable prognosis even when tumors are incompletely excised (Mulvany et al. 1994), as pelvic recurrence is infrequent and usually amenable to surgical excision (Evans et al. 1981; Norris and Parmley 1975). Residual pelvic tumor may remain stable, but progressive growth is possible, especially in women whose treatment does not include bilateral salpingo-oophorectomy because IVL is hormonally dependent. Long-term survival is possible even after removal of plugs of tumor from the vena cava or right atrium or excision of nodules from the lung. In one case, leuprolide acetate induced tumor regression and rendered debulking surgery feasible in a patient with previously unresectable, widespread, retroperitoneal IVL (Tresukosol et al. 1995).

Benign Metastasizing Leiomyoma Benign metastasizing leiomyoma is a nebulous condition in which “metastatic” smooth muscle tumor deposits in extrauterine locations appear to be derived from a benign leiomyoma of the

10

Mesenchymal Tumors of the Uterus

uterus. These tumors most commonly affect the lung, followed by the lymph nodes and abdomen. One or multiple nodules of a low-grade smooth muscle tumor grow in an expansile pattern within the pulmonary parenchyma, often incorporating bronchioles (Fig. 23a, b). Almost all cases of benign metastasizing leiomyoma occur in women, most with a history of pelvic surgery. Because the primary neoplasm, typically removed years before the metastatic deposits are detected, often has been inadequately studied, reports of this condition are often difficult to assess. In some cases, the cytologic appearance, including mitotic counts, is not recorded for either the primary tumor or the alleged metastasis. A few examples may represent deportation metastases from IVL that reach the lungs, where they become implanted and grow as multiple intrapulmonary nodules of smooth muscle (Lee et al. 2008). Others may represent a multifocal smooth muscle proliferation involving the uterus and extrauterine sites (Cho et al. 1989). In the past, most examples of “benign metastasizing leiomyoma” have been interpreted as either a primary benign smooth muscle lesion of the lung in a woman with a history of uterine leiomyoma or pulmonary metastases from a morphologically noninformative smooth muscle neoplasm of the uterus (Bell et al. 1994; Cohen et al. 2007; Gal et al. 1989; Wolff et al. 1979). However, the findings of a cytogenetic study were consistent with a monoclonal origin of both uterine and

559

pulmonary tumors with the interpretation that the pulmonary tumors were metastatic (Tietze et al. 2000). Similarly, benign metastasizing leiomyoma has been reported to have a karyotype that partly overlaps with cellular leiomyoma (19q and 22q terminal deletion) (Mehine et al. 2013). Benign metastasizing leiomyoma may be hormonally dependent, as suggested by the expression of ER and PR in metastatic deposits (Jautzke et al. 1996) and the regression of tumors during pregnancy (Horstmann et al. 1977), after menopause, and after oophorectomy (Abu-Rustum et al. 1997). On radiography, these tumors usually appear as solitary or multiple pulmonary nodules, and on CT these nodules (or those in other organs) enhance homogeneously (Cohen et al. 2007).

Peritoneal Leiomyomas (“Parasitic” Leiomyomas) On rare occasions, leiomyomas have been reported to “detach” from their initial subserosal location and “attach” to some other pelvic site. This improbable event presumably occurs through a combination of infarction and inflammatory adhesions. A diagnosis of parasitic leiomyoma should be made with great caution because clinically malignant smooth muscle neoplasms arising in the retroperitoneum or gastrointestinal tract are notorious for being bland and having a low mitotic index. As mentioned

Fig. 23 Benign metastasizing leiomyoma. (a) A circumscribed smooth muscle tumor is present in the lung. (b) At high magnification, incorporated bronchioles are surrounded by cytologically bland smooth muscle cells

560

previously, ascertaining ER expression helps determine whether an extrauterine tumor is of gynecologic origin.

Disseminated Peritoneal Leiomyomatosis Disseminated peritoneal leiomyomatosis (DPL) is a rare condition characterized by the presence of multiple smooth muscle, myofibroblastic, and fibroblastic nodules on the peritoneal surfaces of the pelvic and abdominal cavities in women of reproductive age (Minassian et al. 1986; Tavassoli and Norris 1981). Most cases are associated with pregnancy, an estrinizing granulosa tumor, oral contraceptive use (Tavassoli and Norris 1981), or endometriosis (Clement 2007). The most common presentation is as an unexpected finding at the time of cesarean section. DPL appears as multiple, small, granular white or tan nodules on the pelvic and abdominal peritoneum, on the surfaces of the uterus, adnexa, intestines, and in the omentum (Fig. 24a, b). The nodules are distributed randomly, and most are less than 1 cm in diameter; this contrasts with metastatic leiomyosarcoma, in which the nodules tend to be fewer, larger, and invasive into adjacent tissues. Microscopically, DPL nodules consist of collagen, fibroblasts, myofibroblasts, smooth muscle cells, and, in pregnancy or the postpartum period, decidual cells (Fig. 25). Spindle cells usually

Fig. 24 DPL. Presents grossly as multiple or numerous small nodules of smooth muscle in the omentum (a) or on the peritoneum. A low-power photomicrograph illustrates

E. Oliva et al.

dominate, leading to potential confusion with metastatic sarcoma, but their different clinical presentations and cellular morphologies should clearly distinguish the two. Another key distinction is that mitotic figures are infrequent in DPL, and nuclear atypia and pleomorphism are minimal or absent. Most nodules are composed of smooth muscle and decidual cells, although some are mixtures of decidua and fibroblasts or myofibroblasts, as shown by electron microscopic studies (Goldberg et al. 1977; Nogales et al. 1978; Pieslor et al. 1979; Tavassoli and Norris 1981). DPL likely arises from a single transformation event, as indicated by a cytogenetic study. In each of the four patients, the same parental X chromosome was nonrandomly inactivated in all tumorlets (7–14 per patient), consistent with a metastatic unicentric neoplasm or, alternatively, selection for an X-linked allele in clonal multicentric lesions (Quade et al. 1997). In this regard, DPL more closely resembles IVL than typical uterine leiomyomas (Quade et al. 2002). Consistent with its association with hormonal stimuli, biochemical or immunohistochemical methods reveal ER and PR expression within the tumorlets (Due and Pickartz 1989). DPL generally regresses or remains static after removal of the hormonal stimulus (i.e., after delivery), so radical attempts at excision are unnecessary (Tavassoli and Norris 1981). In keeping with a hormonally dependent process, DPL may regress during therapy with a

multiple nodules of smooth muscle cells surrounded by omental fat (b)

10

Mesenchymal Tumors of the Uterus

561

leiomyoma. Patients can be managed with surgical resection with or without hormonal agents.

Leiomyosarcoma

Fig. 25 DPL. The peritoneal nodules consist of histologically bland spindle-shaped smooth muscle cells; mitotic figures are absent

GnRH agonist (Hales et al. 1992) and may enlarge again when the GnRH agonist is discontinued or if the patient becomes pregnant. In a few cases of DPL, leiomyosarcoma was diagnosed shortly after (Bekkers et al. 1999). These may represent a distinct entity, as several were distinguished from typical cases of DPL by lack of exposure to estrogen or associated uterine leiomyomas and by absence of ER and PR.

Abdominopelvic Implantation of Leiomyoma Following Morcellation Implantation of leiomyomas following morcellation, a rare event, has been described in the literature (Seidman et al. 2012; Tulandi et al. 2016). This condition is distinct from DPL because in this setting, implantation is iatrogenic, whereas tumor spread in DPL is likely to be hormonally driven (Hales et al. 1992; Tavassoli and Norris 1982). Another difference from DPL is that no long-term studies of abdominopelvic implantation of leiomyoma following morcellation have yet been performed, whereas the clinical presentation and course of DPL is well understood. Unfortunately, several cases of abdominopelvic implantation have been described in the literature under the term “DPL.” The clinical biology is currently thought to be benign provided that the histological features of the abdominopelvic tumors are those of

Leiomyosarcoma represents about 1.3% of uterine malignancies and more than 50% of uterine sarcomas, excluding carcinosarcoma (Abeler et al. 2009). Approximately 1 of every 800 smooth muscle tumors of the uterus is a leiomyosarcoma, but fewer than 1% of women with clinically suspected leiomyoma prove to have leiomyosarcoma (Leibsohn et al. 1990).

Clinical Features The median age of women with leiomyosarcoma is 50–55 years (Abeler et al. 2009; Giuntoli et al. 2003), nearly a decade older than women with leiomyomas, although the disease also affects women in the third decade of life. Leiomyosarcoma is more prevalent in African American women than in white women (Brooks et al. 2004) and has no relationship with gravidity or parity. The clinical presentation is nonspecific; patients present with abnormal vaginal bleeding, lower abdominal pain, or a pelvic or abdominal mass (Giuntoli et al. 2003). The average duration of symptoms before diagnosis is 5 months (Larson et al. 1990). Though a rapidly enlarging uterine smooth muscle neoplasm was previously thought to be indicative of leiomyosarcoma, the evidence does not bear out this dogma. In one study, only 1 of 371 women with a rapidly growing tumor proved to have a leiomyosarcoma (Parker et al. 1994). Unlike carcinosarcoma (malignant mixed müllerian tumor (MMMT)), leiomyosarcoma is seldom associated with a history of pelvic radiation.

Gross Findings Most leiomyosarcomas are intramural, and 50–75% are solitary masses (Schwartz et al. 1993). A higher proportion involves the cervix than is the case with leiomyoma. Leiomyosarcoma

562

E. Oliva et al.

averages 6–9 cm in diameter and is soft or fleshy with poorly defined margins (Abeler et al. 2009), and its cut surface is gray-yellow or pink, often with areas of necrosis and hemorrhage (Fig. 26; 09-1153). All of these characteristics help distinguish it from leiomyoma, which tends to be smaller, firmer, more clearly demarcated and is less likely to be hemorrhagic and necrotic (Table 3). On the other hand, most smooth muscle neoplasms that have a peculiar gross appearance are found to be benign and to exhibit some form of “degeneration,” usually ischemic. Benign and malignant smooth muscle neoplasms are not reliably distinguishable by most imaging modalities (Schwartz and Kelly 2006), except for qualitative magnetic resonance (MR) imaging with texture analysis (Lakhman et al. 2017).

Fig. 26 Typical leiomyosarcoma. A solitary neoplasm, softer than the usual leiomyoma, contains areas of necrosis and hemorrhage on the cut surface

Microscopic Findings in Conventional Leiomyosarcoma A typical leiomyosarcoma is composed of fascicles of spindle cells with abundant eosinophilic cytoplasm (Fig. 27), and frequently contains longitudinal cytoplasmic fibrils, best appreciated with a trichrome stain. The nuclei are fusiform, usually have rounded ends, and are hyperchromatic with coarse chromatin and prominent nucleoli (Fig. 28). These tumors often display marked cellular pleomorphism, especially in those that are poorly differentiated (Figs. 29 and 30), and 50% contain multinucleated cells. Leiomyosarcomas occasionally include giant cells resembling osteoclasts (Darby et al. 1975; Marshall et al. 1986; Patai et al. 2006) and, rarely, prominent xanthoma cells

Fig. 27 Leiomyosarcoma. The tumor cells are spindleshaped with eosinophilic cytoplasm. The nuclei are fusiform, hyperchromatic, and atypical, and there are many mitotic figures

Table 3 Comparison of the gross pathology of leiomyoma and leiomyosarcoma Leiomyoma Usually multiple Variable size, usually 3–5 cm Firm, whorled cut surface White Hemorrhage and necrosis (infarction type) infrequent

Leiomyosarcoma Usually solitary (50–75%) Large, usually 5–10 cm or larger Soft, fleshy cut surface Yellow or tan Hemorrhage and necrosis (coagulative tumor cell type) frequent

Fig. 28 Leiomyosarcoma. The tumor cells vary in size and shape. Several mitotic figures are present including an abnormal one

10

Mesenchymal Tumors of the Uterus

563 Table 4 Histologic criteria for the diagnosis of uterine smooth muscle tumors with standard smooth muscle differentiation Tumor cell necrosis Present

Present Present

Absent

Fig. 29 Leiomyosarcoma. The tumor cell nuclei are pleomorphic, with some giant nuclei and an atypical mitotic figure

Absent

Atypia Diffuse moderate to severe None to mild None to mild Diffuse moderate to severe Diffuse moderate to severe

Absent

None to mild

Absent

None to mild Focal moderate to severe Focal moderate to severe

Absent

Absent

MF/ 10 HPF Any level

Diagnosis Leiomyosarcoma

10 or more Less than 10 10 or more

Leiomyosarcoma

Less than 10

“Smooth muscle tumor with low risk of recurrence” STUMP Leiomyoma

Less than 5 5– 20 5 or more Less than 5

Smooth muscle tumor of LMP/STUMPa Leiomyosarcoma

Mitotically active leiomyoma STUMP (limited experience) Leiomyoma with bizarre nuclei

a

If infarcted/apoplectic leiomyoma is excluded LMP: low malignant potential; STUMP: smooth muscle tumor of uncertain malignant potential

Fig. 30 Leiomyosarcoma with anaplastic features and giant cells. The focus displayed is consistent with leiomyosarcoma, provided it is present in an otherwise typical leiomyosarcoma. Out of context, this focus could be diagnosed as undifferentiated uterine sarcoma, pleomorphic type

(Grayson et al. 1998). Many invade the surrounding myometrium, and 10–22% invade the vasculature, but a leiomyosarcoma with a circumscribed margin can give rise to metastases. The main criteria used to diagnose leiomyosarcoma of the uterus are the presence of nuclear atypia, a high mitotic index (>10 MF/10 HPF but typically >15 MF/10HPF) (Pelmus et al. 2009), and coagulative tumor cell necrosis, though the latter is not essential (Table 4). The differential diagnosis generally includes leiomyoma

variants, smooth muscle tumor of uncertain/low malignant potential, endometrial stromal sarcoma (particularly those variants containing spindle cells), IMT, undifferentiated sarcoma, the sarcomatous component of adenosarcoma or carcinosarcoma, malignant solitary fibrous tumor, and extension of gastrointestinal stromal tumor from the rectum. With the exception of carcinosarcoma (see ▶ Chap. 9, “Endometrial Carcinoma”) and gastrointestinal stromal tumor (which is CD117, DOG-1, and CD34-positive), these entities are discussed in detail later in this chapter. The remaining entities can be distinguished by morphology, immunophenotype, and assessment for gene fusions.

564

Microscopic Findings in Myxoid Leiomyosarcoma Myxoid leiomyosarcoma is a large, gelatinous neoplasm that usually appears circumscribed on gross examination (Kunzel et al. 1993; Schneider et al. 1995) and is becoming a diagnosis of exclusion given the wide differential diagnosis, presented below. Microscopically, the smooth muscle cells are usually widely separated by myxoid material (Fig. 31) (King et al. 1982). The characteristic low cellularity partly accounts for the low mitotic index in most myxoid leiomyosarcomas. Sometimes, however, the mitotic index is high and there is a high degree of atypia. In addition to the myxoid appearance, other microscopic features that help identify the tumor as a leiomyosarcoma include myometrial infiltration and vascular invasion. Infiltrative growth was key to recognition of an unusual myxoid leiomyosarcoma that arose within a leiomyoma (Mittal et al. 2000). Despite the low mitotic counts, myxoid leiomyosarcoma has the same unfavorable prognosis as typical leiomyosarcoma. Myxoid smooth muscle tumors of the uterus must be regarded with suspicion, and any myxoid smooth muscle tumor with significant nuclear atypia, regardless of the mitotic count or the presence or absence of necrosis, should be classified as a leiomyosarcoma. Proper diagnosis requires distinguishing the myxoid differentiation found in myxoid

Fig. 31 Myxoid leiomyosarcoma. (a) Abundant myxoid stroma widely separates bundles of smooth muscle cells, resulting in a hypocellular appearance. (b) The degree of

E. Oliva et al.

leiomyosarcomas from the vastly more prevalent hydropic changes seen in degenerating leiomyomas (Clement et al. 1992). In myxoid leiomyosarcoma, not only is the stroma myxoid but the cells are enlarged with hyperchromatic nuclei and pleomorphism is usually obvious. Myxoid leiomyosarcoma usually bears some resemblance to soft tissue myxofibrosarcoma, although the vasculature tends not to be as prominent and the degree of nuclear atypia is frequently lower in myxoid leiomyosarcoma. Myxoid leiomyosarcoma must be differentiated from focal myxoid change (Pugh et al. 2012), inflammatory fibromyxoid tumor (ALK-rearranged) (Bennett et al. 2017a; Parra-Herran et al. 2015; Rabban et al. 2005), the sarcomatous component of adenosarcoma, and myxoid/fibromyxoid variants of endometrial stromal sarcoma (Lewis et al. 2017; Oliva et al. 1999; Yilmaz et al. 2002), all discussed subsequently in this chapter.

Microscopic Findings in Epithelioid Leiomyosarcoma Epithelioid leiomyosarcomas are composed of round or polygonal cells and exhibit one of the patterns of epithelioid differentiation (leiomyoblastoma, clear cell, or plexiform) (Fig. 32). Like myxoid leiomyosarcoma, these tumors should be considered a diagnosis of exclusion based on similarity to many other tumors,

nuclear atypia can be deceptively bland, and because the tumor cells are widely separated by myxoid stroma, mitotic activity is low

10

Mesenchymal Tumors of the Uterus

565

to be clinically malignant. Unless the tumor is invasive or contains abnormal mitotic figures or areas of tumor cell necrosis, there are no good grounds for suspecting that it is a leiomyosarcoma until it announces itself by metastasizing.

Immunohistochemistry

Fig. 32 Epithelioid leiomyosarcoma of the “leiomyoblastoma” type. Tumor cells are polygonal with pale cytoplasm and atypical nuclei. Mitotic figures are not numerous, but coagulative tumor cell necrosis was extensive

discussed below. The leiomyoblastoma pattern is most common, although clear cell epithelioid leiomyosarcomas have also been reported (Prayson et al. 1997; Silva et al. 1995, 2004). These tumors are distinguished from other epithelioid smooth muscle tumors by the usual features of malignancy: significant nuclear atypia and either necrosis or > 5 MF/10 HPF qualify a tumor as leiomyosarcoma (Atkins et al. 2001; Clement 2000; Kempson and Hendrickson 2000; Kurman and Norris 1976; Moinfar et al. 2007; Prayson et al. 1997). The differential diagnosis of epithelioid leiomyosarcoma includes leiomyoma variants, smooth muscle tumor of uncertain/low malignant potential, PEComa, carcinoma, metastatic melanoma (S-100 positive), placental site and epithelioid trophoblastic tumors (GATA-3 positive), ASPS (HMB-45 negative with Xp11 translocation), and endometrial stromal tumors, as some have an epithelioid appearance (Lee et al. 2012b). With the exception of carcinoma (discussed in ▶ Chap. 9, “Endometrial Carcinoma”), placental site trophoblastic tumor (discussed in ▶ Chap. 20, “Gestational Trophoblastic Tumors and Related Tumorlike Lesions”), and melanoma, these entities are discussed in detail later in this chapter. Otherwise conventional smooth muscle tumors with low mitotic counts, on rare occasions, prove

Immunohistochemistry is generally not required for the diagnosis of leiomyosarcoma, but immunostains are occasionally necessary to differentiate it from other uterine malignancies such as undifferentiated endometrial sarcoma or sarcomatoid carcinoma. Immunostains using a variety of antibodies can confirm that extrauterine sarcoma deposits show smooth muscle differentiation and hence are compatible with metastatic leiomyosarcoma. Smooth muscle actin, desmin, and caldesmon are the best established markers, but calponin and smooth muscle myosin are also occasionally used for this purpose; all of these stain the cytoplasm. Because myofibroblasts can also stain with smooth muscle markers, particularly smooth muscle actin, positive staining with one of these markers is not conclusive evidence of smooth muscle differentiation. Distinction from endometrial stromal tumors requires close attention, as leiomyosarcoma often stains positive for the stromal marker CD10. In one study, eight of nine leiomyosarcomas were CD10-positive (low to moderate intensity, staining in 5–60% of tumor cells) (Oliva et al. 2002b). However, CD10 staining in smooth muscle tumors is generally weak and focal (McCluggage et al. 2001b; Toki et al. 2002). Myometrial smooth muscle cells are frequently immunoreactive for cytokeratin, so leiomyosarcomas occasionally show positive cytoplasmic staining for cytokeratin (Oliva et al. 2002b). Many investigators have proposed immunohistochemical stains that could distinguish between benign and malignant smooth muscle tumors, but markers such as p53 (Fig. 33), p16 (Fig. 34), mib-1, ER, and progesterone receptor are variably expressed in leiomyomas and leiomyosarcomas, limiting their

566

E. Oliva et al.

A summary of antibodies that can be used in the differential diagnosis of usual, myxoid, and epithelioid leiomyosarcomas is found in Tables 5, 6, and 7.

Molecular Pathology

Fig. 33 Positive staining for p53 in leiomyosarcoma

Fig. 34 Positive staining for p16 in leiomyosarcoma

utility in differential diagnosis (Amada et al. 1995; Atkins et al. 2008; Blom and Guerrieri 1999; Bodner et al. 2004; Bodner-Adler et al. 2005; Chen and Yang 2008; de Vos et al. 1994; Gannon et al. 2008; Hall et al. 1997; Layfield et al. 2000; Leiser et al. 2006; Leitao et al. 2004; Mittal and Demopoulos 2001; O’Neill et al. 2007; Watanabe and Suzuki 2006; Zhai et al. 1999). For example, although mutation of the p53 gene or positive staining for p53 (strong nuclear staining in >50% of tumor cell nuclei) is observed mainly in leiomyosarcomas, albeit in a minority (Gannon et al. 2008; O’Neill et al. 2007), leiomyomas with bizarre nuclei and smooth muscle tumors of low malignant potential (SM-LMP)/ STUMP have been reported to stain with the same frequency and intensity as leiomyosarcoma (Chen and Yang 2008). The greatest overlap in immunophenotype is between usual leiomyosarcoma and leiomyomas with bizarre nuclei.

Comparison of gene expression profiles in uterine leiomyosarcoma and normal myometrium has revealed differences in cell cycle regulation, DNA repair, and genomic integrity (Barlin et al. 2015). Gene expression profiling can also differentiate primary uterine leiomyosarcoma from leiomyosarcoma metastases (Davidson et al. 2014) and aggressive from more indolent tumors. In one study, unsupervised clustering of leiomyosarcomas identified two clades (genomic subgroups) that were reproducibly associated with significant differences in progression-free and overall survival (Barlin et al. 2015). Interpreting work performed by the Cancer Genome Atlas initiative (TCGA), van de Rijn and colleagues identified three genomic categories of leiomyosarcoma, one that was highly enriched for uterine tumors and two that were shared between uterine and soft tissue leiomyosarcomas (Guo et al. 2015). One of the latter subtypes was associated with poor prognosis. The subtypes differed significantly in expression levels for genes for which novel targeted therapies are being developed, suggesting that different leiomyosarcoma subtypes may respond differentially to targeted therapies (Guo et al. 2015). The most frequently mutated genes in uterine leiomyosarcoma are TP53 (in one-third of cases), alpha thalassemia/mental retardation syndrome X-linked (ATRX; in one-quarter of cases), and mediator complex subunit 12 (MED12; in one-fifth of cases) (Makinen et al. 2016). Loss of ATRX expression is associated with alternative lengthening of telomeres (ALT), leading to acquisition of an “ALT phenotype,” which may be therapeutically targeted (Makinen et al. 2016). The presence of MED12 mutations in both leiomyomas and leiomyosarcomas has led to speculation that some leiomyosarcomas may

10

Mesenchymal Tumors of the Uterus

567

Table 5 Immunohistochemical features of conventional (spindle) leiomyosarcoma and entities in the differential diagnosis LMS LMA-apo LMA-bizarre (1) LMA-bizarre (2) ESS-LG GIST SFT Sarcoma-het Sarcoma-undiff

Des ++ ++ ++ ++ /+

*

CD10 /+ /+ /+ /+ ++/ /+ +/ /+ /+

p53 /++ NA ++/ /+ ++ ++

MIB-1 Diffuse Geographic Low Variable Low Variable Variable Diffuse Diffuse

FH Intact Intact Lost Intact Intact NA NA NA NA

C-kit Variable NA NA NA Variable Diffuse

STAT6

++ Variable Variable

Des: desmin; FH: fumarate hydratase; LMS: leiomyosarcoma; LMA: leiomyoma; apo: apoplectic; ESS-LG: low-grade endometrial stromal sarcoma; GIST: gastrointestinal stromal tumor; SFT: deficient solitary fibrous tumor; het: heterologous; undiff: undifferentiated bizarre; NA: not analyzed in literature; LMA-bizarre (1): FH-deficicat bizarre leiomyoma; LMA-bizarre (2): p53 – mutated bizarte leiomyoma *Except in rhabdomyosarcoma

Table 6 Immunohistochemical features of myxoid leiomyosarcoma and entities in the differential diagnosis LMS-myx LMA-degen IMT ESS-myx ESS-BCOR Carcinoma-undiffa Sarcoma-undiff

Des +/ ++ +/ +/

SMA +/ ++ +/ +/ +/

ALK

CD10

++

+/ + +

/+

/+

Cycl D1

BCOR

Ker

/+ /+ +/

+/ +/ /+

+/

Des: desmin; SMA: smooth muscle actin; Ker: cytokeratin; LMS: leiomyosarcoma; myx: myxoid; LMA: leiomyoma; degen: degenerative; IMT: inflammatory myofibroblastic tumor; ESS: endometrial stromal sarcoma; undiff: undifferentiated a These carcinomas are also typified by loss of PAX-8 and e-cadherin expression, DNA mismatch repair deficiency, and abnormalities in chromatin remodeling (i.e., ARID1A, SMARCB1, or SMARCA4 expression loss)

Table 7 Immunohistochemical features of epithelioid leiomyosarcoma and entities in the differential diagnosis LMS-epi PEComa (1) PEComa (2) ESS-sex cord ESS-YWHAE (LG) ESS-YWHAE (HG) UTROSCT Carcinoma-undiff PSTT Sarcoma-undiff

Des ++/ +/ +/

HMB45 /+ + ++

TFE3

++

CD10 /+ /+ NA ++ ++

Inh

+

++

Cycl D1

++

Ker /+

+/ ++

+/

/+ +/ /+

+ /+

+/ +/ ++

hPL NA NA NA NA NA NA NA + NA

Des: desmin; Inh: inhibin; Ker: cytokeratin; hPL: human placental lactogen; LMS: leiomyosarcoma; PEComa (1): perivascular epithelioid cell tumor with epithelioid eosinophilic cells; PEComa (2): perivascular epithelioid cell tumor with clear cells, resembling Xp11-translocated renal cell carcinomas; ESS: endometrial stromal sarcoma; YWHAE: YWHAE translocation present; LG: low grade; HG: high grade; UTROSCT: uterine tumor resembling ovarian sex cord tumor; undiff: undifferentiated; PSTT: placental site trophoblastic tumor; NA: not analyzed in literature

568

arise from a preexisting leiomyoma (Bertsch et al. 2014; Makinen et al. 2016; Matsubara et al. 2013). Rare myxoid leiomyosarcomas harbor translocations involving PLAG1 (Arias-Stella et al. 2018). In a study that assessed clinical outcomes in uterine and extrauterine leiomyosarcomas by mutated genes (Yang et al. 2015), almost all TP53-mutated leiomyosarcomas were located in the uterus or retroperitoneum. ATRX mutations were associated with poor differentiation, the presence of tumor necrosis, and worse overall survival. In contrast to knowledge about gene mutations in leiomyosarcomas, the literature regarding epigenetic changes and posttranscriptional regulation of gene expression is limited. One study identified 94 micro-RNAs (miRNAs) that were significantly differentially expressed in endometrial stromal sarcoma and leiomyosarcoma, 18 of which were overexpressed in leiomyosarcoma, as well as a miRNA signature distinguishing primary from metastatic leiomyosarcoma (Ravid et al. 2016).

Clinical Behavior and Treatment A clinicopathologic portrait of leiomyosarcoma as currently defined is provided by a retrospective review of uterine leiomyosarcomas treated at the Mayo Clinic (Giuntoli et al. 2003). Disease was confined to the uterus in 68% of the 208 patients, while 6% had cervical involvement; approximately half of these had cervical involvement only. Nine percent were stage III and 20% stage IV according to the International Federation of Gynecology and Obstetrics (FIGO) system, meaning that they had invaded abdominal tissues (Table 2) (Prat 2009). Leiomyosarcoma is a highly malignant neoplasm with poor survival rates when tumors are classified using contemporary criteria. Overall 5-year survival rates in series using these criteria range from 15% to 35% (Blom et al. 1998; Larson et al. 1990; Pelmus et al. 2009); variation results from differences in how those criteria are interpreted and applied. When only stage I and II tumors are considered, the 5-year survival rate is 40–70% (Blom and Guerrieri 1999; Gadducci et al. 1996a; Larson et al. 1990; Mayerhofer et al. 1999;

E. Oliva et al.

Nola et al. 1996; Nordal et al. 1995; Pautier et al. 2000; Pelmus et al. 2009; Wolfson et al. 1994), and 3-year progression-free survival is about 30%, according to a Gynecologic Oncology Group (GOG) series of 59 cases (Major et al. 1993).

Prognostic Prediction and Clinical Outcomes The prognosis of leiomyosarcoma depends chiefly on its anatomical extent, often categorized as stage. For stage I tumors (confined to the uterus), some investigators have found the size of the neoplasm to be an important prognostic factor (Abeler et al. 2009). In one series, all patients with tumors larger than 5 cm died of disease, compared to only three of eight patients with tumors smaller than 5 cm (Evans et al. 1988). In another series of metastasizing leiomyosarcomas, only 20% were less than 5 cm (Jones and Norris 1995). Whether premenopausal status is linked with more favorable outcomes remains unclear. Mitotic index also appears to be a prognostic indicator; several series confirm this, including the large GOG study of early-stage leiomyosarcoma (Abeler et al. 2009; Gadducci et al. 1996a; Larson et al. 1990; Major et al. 1993; Pautier et al. 2000; Pelmus et al. 2009), while only one study found otherwise (Evans et al. 1988). PR expression may also be prognostically favorable in stage I leiomyosarcoma (Leitao et al. 2012), while myxoid or epithelioid differentiation and diffuse, severe nuclear atypia may be unfavorable (Wang et al. 2011). The nomogram created based on data from 270 patients treated at Memorial Sloan Kettering Cancer Center (MSKCC) weighs prognostic factors differently (Zivanovic et al. 2012). Median overall survival within this cohort was only 3.75 years. The nomogram’s power (https:// www.mskcc.org/nomograms/uterine) in predicting postresection 5-year overall survival was validated externally (Iasonos et al. 2013). These analyses indicate that the most important staging distinction is between uterus-confined disease and extrauterine disease; much less important is assessment of cervical involvement and tumor size. Listed in order of decreasing importance, the categorical (noncontinuous) predictive variables included “tumor grade” (which was not

10

Mesenchymal Tumors of the Uterus

clearly defined), loco-regional metastasis, distant metastasis, and cervical involvement. The continuous variables included mitotic index, tumor size (cm), and age at diagnosis. Notably absent from the list is FIGO stage; both FIGO and American Joint Commission on Cancer (AJCC) staging for leiomyosarcomas have been criticized (Raut et al. 2009; Zivanovic et al. 2009). Uncontained morcellation of a leiomyosarcoma, performed either manually or with a power morcellator, has a significantly negative impact on survival because of subsequent peritoneal spread. A preliminary study performed years before the current debate about the risks and benefits of morcellation demonstrated a statistically nonsignificant increase in the rate of pelvic dissemination following this procedure (Morice et al. 2003). Almost 10 years later, uncontained morcellation was conclusively shown to not only significantly increase the risk of abdomino-pelvic recurrence of leiomyosarcoma, as 64.3% of patients developed disseminated disease, but also to shorten survival (Park et al. 2011; Seidman et al. 2012). Peritoneal recurrence following morcellation is not limited to leiomyosarcomas, as it also occurs with SM-LMP/STUMPs, endometrial stromal sarcoma, and leiomyoma as well (Tulandi et al. 2016). Nonetheless, morcellation has only been conclusively shown to impair survival in leiomyosarcoma (Seidman et al. 2012). Patients with morcellated leiomyosarcomas should therefore be offered surgical reexploration to detect peritoneal dissemination (Oduyebo et al. 2014). As no universally agreed-upon grading scheme for leiomyosarcoma exists, pathologists should instead comment on maximum tumor diameter, mitotic index, the presence or absence of necrosis and its extent if present, the nature of the tumor periphery (invasive or circumscribed) presence of morcellation and the presence or absence of vascular space involvement.

Metastasis and Recurrence The frequency of lymph node metastasis varies from series to series but is substantially lower than that found in clinical stage I and II high-risk endometrial carcinomas (Giuntoli et al. 2003). In a GOG study, 2 of 59 patients (3%) had lymph

569

node involvement, 2 had adnexal involvement, and 1 had positive peritoneal cytology (Major et al. 1993). Another study of 108 patients from MSKCC, 14% of whom underwent lymph node sampling, found lymph node involvement in 8% of patients, all of whom had gross extrauterine disease and obviously enlarged lymph nodes (Leitao et al. 2003). Moreover, a high percentage of lymph node-negative patients experienced recurrence or died of disease. In view of this, lymph node sampling at initial surgery does not appear worthwhile in this disease. Nonetheless, as many as 44% of patients who die from leiomyosarcoma have lymph node metastasis at autopsy (Fleming et al. 1984; Rose et al. 1989). Leiomyosarcoma relapses occur both localregionally and hematogenously; the following locations were the sole site of relapse: vagina 22%, pelvis 19%, lung 22%, bone 9%, and retroperitoneum 12%. Relapse in both lung and pelvis occurred in 16% of patients (Giuntoli et al. 2003). In another study, the first recurrence was in the pelvis in 14% of cases and in the lung in 41% (Major et al. 1993).

Criticism/Value of Tumor Grading At least 75–90% of leiomyosarcomas defined by Stanford criteria are histologically high grade, and clinical outcome data indicate that such tumors are highly malignant. In a Mayo Clinic study (Giuntoli et al. 2003), leiomyosarcomas assigned grades 2, 3, or 4 had nearly identical survival curves. Similarly, a grading scheme designed for soft tissue neoplasms has been shown to have no prognostic significance for uterine sarcomas (Pautier et al. 2000; Pelmus et al. 2009). It is therefore inappropriate to assign either a numerical grade (i.e., grade 2 of 3 or 4) or a qualitative grade based on degree of differentiation (i.e., well differentiated) to uterine leiomyosarcoma; all leiomyosarcomas that meet Stanford criteria for leiomyosarcoma should be considered intrinsically high grade. Nonetheless, some uterine smooth muscle tumors have the potential to metastasize after long disease-free intervals and follow an indolent course. Such tumors appear to be a subset of

570

leiomyoma variants, sometimes termed “atypical leiomyomas.” In a Mayo Clinic study, 3 of 18 patients with leiomyoma variants died of disease, but between 6 and 11 years after diagnosis; in contrast, almost all patients whose lethal tumors met Stanford criteria for leiomyosarcoma died within 5 years of diagnosis (Giuntoli et al. 2007). The 18 patients with leiomyoma variants were originally considered to have “low-grade leiomyosarcoma,” but labeling all patients as such would potentially have left them vulnerable to unnecessary chemo- and radiotherapy (Giuntoli et al. 2003). Furthermore, there is no evidence that any type of adjuvant therapy changes the clinical course of potentially recurring atypical smooth muscle tumors that fail to meet Stanford criteria for leiomyosarcoma. Similarly, a study from MSKCC found that patients with recurrent atypical leiomyomas, some of which were leiomyomas with bizarre nuclei, experienced disease progression at a mean time of 12 years, and none died of disease (Veras et al. 2011). This study also highlighted the significant heterogeneity of the group of tumors historically considered to be “low-grade leiomyosarcomas,” which comprised usual leiomyosarcomas (the clinical outcomes of which were similar to usual leiomyosarcomas that were considered “high grade”), leiomyoma variants and endometrial stromal neoplasms with spindle cells. Whether it is reasonable to use the term “low-grade leiomyosarcoma” for recurrent, histologically low-grade smooth muscle tumors is discussed in the STUMP portion of this chapter.

Treatment For postmenopausal women, primary therapy for early-stage leiomyosarcoma is total abdominal hysterectomy and bilateral salpingo-oophorectomy. Whether oophorectomy is warranted in premenopausal women is more controversial. The ovaries are the site of metastatic disease in clinically low-stage leiomyosarcoma in only 2–3% of cases (Leitao et al. 2003; Major et al. 1993). While some studies indicate that oophorectomy does not influence outcomes (Gard et al. 1999; Giuntoli et al. 2003; Larson et al. 1990), another reported that patients with so-called low-grade smooth muscle neoplasms metastatic to the lung responded to

E. Oliva et al.

oophorectomy alone (Abu-Rustum et al. 1997). However, these patients’ leiomyosarcomas may not have met Stanford criteria, which suggest that patients with recurrent STUMPs and metastasizing leiomyomas may benefit from this procedure. Similarly, aromatase inhibitors are a suitable option for postmenopausal patients with ER/PR-positive, small volume, and/or slowly progressive disease, where neither resection nor cytotoxic chemotherapy is warranted (Hyman et al. 2014). Therefore, while there may be theoretical benefit to the hormonal ablation afforded by oophorectomy in patients whose tumors express ER and PR, there is no compelling need to perform oophorectomy to detect occult metastasis. The literature provides conflicting reports on the efficacy of radiotherapy and chemotherapy in the management of leiomyosarcoma (Gadducci et al. 1996a, 2008; Giuntoli and Bristow 2004; Giuntoli et al. 2003; Hensley et al. 2002; Mayerhofer et al. 1999; Sutton et al. 2005; Zivanovic et al. 2009). Adjuvant pelvic radiotherapy was previously used under the assumption that it might decrease the likelihood of pelvic recurrence. Postoperative pelvic radiation therapy is no longer advocated as a standard treatment since a prospective, randomized trial reported no benefit to overall or recurrence-free survival (O’Cearbhaill and Hensley 2010; Reed et al. 2008). Pelvic radiotherapy may still be considered in highly selected cases, such as for patients with local relapse or residual disease following surgery. For patients with resected stage I uterine leiomyosarcoma, neither radiation therapy nor chemotherapy has been conclusively demonstrated to be effective, and both are associated with significant toxicities. Thus, there is a strong argument in favor of “watching and waiting” (i.e., withholding therapy until disease progression) in early-stage cases (Hensley 2017; Littell et al. 2017). The combination of gemcitabine and docetaxel is considered a good first-line therapy in patients with disseminated disease (Hyman et al. 2014). Reported response rates are 27–53% with median progression-free survivals that range from 4.4 to 5.6 months (Hensley et al. 2002, 2008a, b); these results are disappointing for individual patients,

10

Mesenchymal Tumors of the Uterus

but they clearly outperform previous chemotherapeutic regimens. Among targeted agents, only the multi-tyrosine kinase inhibitor pazopanib has been approved for any soft tissue sarcoma, and the Aurora A kinase inhibitor is in trials based on strong preclinical evidence (Hyman et al. 2014). Hormonal agents are discussed above.

Smooth Muscle Tumor of Uncertain/ Low Malignant Potential “Smooth muscle tumor of uncertain malignant potential” (STUMP) is a term generally applied to smooth muscle tumors when there is uncertainty about diagnostic criteria for malignancy (i.e., is the necrosis of coagulative type?) (Fig. 35), the type of smooth muscle differentiation present (i.e., is the tumor truly epithelioid?), and malignant potential (i.e., scarcity of clinical outcomes data). The term may also be applied to tumors with a low risk of recurrence. Artificially separating the “uncertain” entities from the “possibly recurring” entities would create two

Fig. 35 STUMP. This atypical smooth muscle tumor contains several areas of necrosis that are indeterminate for coagulative tumor cell necrosis. The background resembles a leiomyoma with bizarre nuclei, with a mitotic index of 5–7 MF/10HPF. This zone of necrosis most closely resembles infarct-type necrosis, with loss of reticulin, but the proliferative rate using mib-1 was lowest in areas surrounding the infarct, a feature that has been suggested to support coagulative tumor cell necrosis

571

categories: STUMP and smooth muscle tumor of low malignant potential (SM-LMP). These terms are both controversial, although the “STUMP” term is widely understood by pathologists and clinicians. Furthermore, “lumpers” (as opposed to “splitters”) may argue that by expressing uncertainty about a given tumor’s malignant potential, one also implies an ill-defined risk of clinically malignant behavior. Regardless of the term used, it is very important to note that none of the following tumor types should be considered STUMPs or SM-LMPs despite the fact that some may recur locally, can be intravascular, and/or involve broad ligament, other pelvic soft tissues, peritoneum, and lung. These include mitotically active leiomyoma, cellular leiomyoma, apoplectic leiomyoma, leiomyoma with bizarre nuclei, diffuse leiomyomatosis, dissecting leiomyoma, metastasizing leiomyoma, IVL, DPL, IMT, endometrial stromal neoplasia with smooth muscle differentiation, and peritoneal implantation of morcellated leiomyoma. To further avoid contamination of the STUMP/SM-LMP category, one must be sure to extensively sample any smooth muscle tumor demonstrating at least one criterion for malignancy. Recurrence or metastasis of the aforementioned leiomyoma variants has historically been considered “benign” provided that the histology of the recurrence or metastasis retains the appearance of the preceding leiomyoma. These secondary tumors can be treated successfully by surgical excision, with or without hormonal agents. Some leiomyomas with bizarre nuclei and smooth muscle tumors described by the Stanford group as “atypical leiomyoma with low risk of recurrence” fall within the STUMP/SM-LMP category (Bertsch et al. 2014). Since most such tumors with mitotic indices of less than 5 MF/10 HPF follow a benign or at most locally recurring clinical course, it is reasonable to continue to categorize these tumors as leiomyoma variants provided they are very well sampled. Those with higher mitotic indices (5–10 MF/10HPF) may be classified as either SM-LMPs or as “atypical leiomyomas with recurring potential/low risk of recurrence.”

572

Therefore, after excluding leiomyoma variants and leiomyosarcoma, the STUMP category contains three clinical entities: (1) benign smooth muscle tumors that are difficult to diagnose; (2) potentially recurring smooth muscle tumors with indolent recurrence and/or long disease-free intervals (SM-LMPs); and (3) conventional leiomyosarcoma that is difficult to diagnose. No tools are yet able to distinguish between the first and second entities, but use of some of the immunohistochemical and microscopic approaches discussed subsequently may be diagnostically useful for the third set of tumors, though they have limitations. In addition to benign-appearing recurrences of leiomyomas and leiomyoma variants in extraperitoneal sites, some recurrences and metastases following a leiomyoma diagnosis may show evidence of morphologic evolution. Recurrent/metastatic leiomyomas that meet criteria for leiomyosarcoma should be categorized as such, but recurrent/metastatic tumors that are more cellular, more mitotically active, or are invasive beyond that of a typical leiomyoma, while falling short of a leiomyosarcoma diagnosis using Stanford criteria, should be categorized as STUMP or “SM-LMP.” The term “low-grade leiomyosarcoma” may be considered for such cases, especially since many clinicians will have difficulty understanding the concept of a recurrent/metastatic atypical smooth muscle neoplasm, but there is no firm precedent for this term in the gynecologic pathology literature. Regardless of the term used, it is crucial to convey that such recurrent histologically low-grade tumors do not constitute conventional uterine leiomyosarcoma, as they are treated by entirely different approaches.

Microscopic Findings Careful microscopic examination aids in distinguishing STUMP from usual leiomyosarcoma. This differential diagnosis is particularly problematic when there is uncertainty about the mitotic index, which would affect predictions of the tumor’s clinical behavior. This situation most

E. Oliva et al.

frequently occurs in smooth muscle neoplasms with standard smooth muscle differentiation that lack tumor cell necrosis, have moderate to severe atypia, and have a mitotic index that falls just short of 10 MF/10 HPF. The diagnosis of STUMP in this situation is treacherous and is potentially misleading, as classifying a tumor with diffusely distributed atypical nuclei and 9 MF/10 HPF as “smooth muscle tumor with low risk of recurrence” or STUMP may significantly underestimate the tumor’s metastatic potential. Excluding the diagnosis of conventional leiomyosarcoma by extremely thorough sampling to uncover a focus of high mitotic activity in such a tumor is critical.

Immunohistochemistry and Molecular Pathology Two strategies, immunohistochemistry and molecular analysis, including comparative genomic hybridization (CGH), have been proposed to separate STUMPs from usual leiomyosarcoma and to distinguish between “benign” and potentially malignant STUMPs immunohistochemistry and molecular analysis, including CGH. In addition to phosphohistone immunohistochemistry for accurate mitotic counting and prognostication (Chow et al. 2017; Veras et al. 2009), possibly useful markers include p53 and p16 (Ip et al. 2009; O’Neill et al. 2007). However, these should not be applied when the tumor has the appearance of a leiomyoma with bizarre nuclei, since many of these tumors harbor p53 mutations that lead to overexpression (Bennett et al. 2017b; Chen and Yang 2008; Ubago et al. 2016; Zhang et al. 2017). Similarities among leiomyomas with bizarre nuclei, STUMPs, and leiomyosarcoma include miRNA expression patterns and rates of PTEN deletions (Zhang et al. 2014). Although p53 and p16 would not distinguish between leiomyomas with bizarre nuclei and leiomyosarcoma, these markers might help to separate conventional leiomyosarcomas from STUMPs lacking “symplastic” features. Perhaps more informative is immunohistochemistry for

10

Mesenchymal Tumors of the Uterus

ATRX, DAXX, and MED12. Expression of ATRX and DAXX is known to be lost in a subset of leiomyosarcoma (Slatter et al. 2015) and was associated with death due to disease or recurrence in six of six patients with “aggressive” STUMPs and almost all aggressive leiomyosarcomas. These markers likely help determine which STUMPs are actually leiomyosarcomas, although expression of these markers remains unexamined in leiomyomas with bizarre nuclei. Retained expression of these markers is not diagnostically informative, however. Similarly, MED12 expression decreases on a gradient from leiomyoma to STUMP to leiomyosarcoma (Croce and Chibon 2015).While MED12 mutations are much more frequently found in usual and mitotically active leiomyomas, as stated previously, inhibition of MED12 expression may correlate with malignancy (Pérot et al. 2012). Finally, it has been proposed that CGH can be used to calculate a “genomic index” score that separates nonrecurring STUMPs from those with recurrences and unfavorable clinical outcomes (Croce et al. 2015). Similar to other approaches, this technique appears to uncover conventional leiomyosarcomas masquerading as STUMPs. As all reported recurrences were documented within 5 years of diagnosis, this technique may not be helpful in the identification of STUMPs prone to late recurrence.

Clinical Behavior and Treatment STUMP progresses slowly, if at all, and in the small proportion of cases in which it recurs, most affected patients survive. In the largest study on STUMPs to date, among 41 patients with a mean follow-up time of 45 months, recurrence was observed in 3 patients (7.3%; one as leiomyosarcoma and two as STUMP). Recurrence rates were similar for women who underwent myomectomy or hysterectomy. All three patients were alive and disease-free at a mean follow-up time of 121 months (Guntupalli et al. 2009). A prior series of 15 STUMP also found a high survival rate and slow progression when recurrence became evident (Peters et al.

573

1994). Although necrosis was not recorded, at least some of the tumors in the latter study would qualify as atypical leiomyomas using the Bell et al. criteria. Similarly, Ip and colleagues reported two recurrences out of 16 patients following a diagnosis of “atypical leiomyoma with limited experience,” “SM-LMP,” “atypical leiomyoma,” or “mitotically active leiomyoma, limited experience” (Ip et al. 2009). The two patients who experienced recurrence had been diagnosed with “atypical leiomyoma with limited experience” and both had diffuse immunoreactivity for p53 and p16. Finally, STUMPs may recur in intraperitoneal sites if they are subjected to morcellation (Bogani et al. 2016; Seidman et al. 2012). In summary, multiple studies since the introduction of the term STUMP have shown that the behavior of these tumors is not “uncertain.” The majority of them are benign and only a small percentage recur, which has led some investigators to prefer the designation “SM-LMP” or “smooth muscle tumor with low risk of recurrence” in order to avoid unnecessary adjuvant treatment and to reassure the patient that her fate is not “uncertain.”

PEComa and Related Lesions PEComa is a neoplasm composed of cells showing “perivascular epithelioid cell” (PEC) differentiation (Folpe et al. 2005; Vang and Kempson 2002), although non-neoplastic equivalents are not recognized. Vang and Kempson described two types of uterine PEComas: (1) clear-toeosinophilic tumor cells with variable staining for smooth muscle and melanocytic markers (Group B PEComa) (Martignoni et al. 2008) and (2) clear tumor cells arranged in nests with strong labeling for HMB-45, but significantly less staining for smooth muscle markers (Group A PEComa) (Vang and Kempson 2002). The PEC family of tumors also includes angiomyolipoma, lymphangiomyomatosis (LAM), and clear cell tumors of the lung and pancreas.

574

AML is a benign tumor that contains abnormal blood vessels, fat and spindled or epithelioid smooth muscle cells in varying proportions. LAM is generally a microscopic finding in the uterus and consists of dilated lymphovascular spaces associated with smooth muscle cells with flocculent eosinophilic cytoplasm. Pulmonary, uterine, and retroperitoneal nodal LAM may be associated with tuberous sclerosis. In such cases, LAM may be multifocal, invade the myometrium in a tongue-like fashion, and display a paucity of lymphatic vessels. Uterine LAM can also present in sporadic settings, where tongue-like growth is less prevalent and lymphatic vessels more prominent (Hayashi et al. 2011). Uterine PEComas differ significantly from those that arise in other locations. Among the peculiar attributes of group B uterine PEComas, (1) uterine PEComas more frequently meet criteria for malignancy; (2) malignant PEComas often harbor mutations matching those of uterine leiomyosarcoma and lack those typical of most PEComas (unpublished observations); and (3) hybrid smooth muscle/PEComas occur. Clinical Features and Gross Findings PEComas are neoplasms of adulthood, and when they involve the uterus, they typically present either as a mass or cause uterine bleeding. They may be associated with LAM and tuberous sclerosis. PEComas are usually solitary neoplasms ranging in size from 0.5 to 16.0 cm, although rarely multiple lesions are described. Microscopic Findings Microscopically, the low-power impression is of either a compressive or, less commonly, an infiltrative neoplasm with a nondistinctive texture and coloration. Sometimes there is a tongue-like pattern of invasion of the myometrium reminiscent of the type of invasion seen in low-grade endometrial stromal sarcoma (Fig. 36) (Vang and Kempson 2002). The pathology of gynecologic PEComas has been reviewed recently (Conlon et al. 2015). The more common Group B PEComas feature tumor cells that range from predominantly

E. Oliva et al.

Fig. 36 PEComa (group B). In this uterine PEComa, rounded protrusions of tumor infiltrate the myometrium in a pattern that is somewhat reminiscent of an endometrial stromal sarcoma

epithelioid to occasionally spindled with moderate to abundant clear-to-eosinophilic cytoplasm with well-defined cell borders (Fig. 37). The tumor cell cytoplasm sometimes has a distinctive granular or finely vacuolated bubbly appearance. The degree of nuclear atypia and the mitotic index are low and necrosis is uncommon, but tumors with significant atypia and frequent mitotic figures occur. A variant with abundant hyalinized stroma that can partially obscure the tumor cells has been reported (Hornick and Fletcher 2008). In addition to the classically described architectural and cytoplasmic characteristics, features that are typical, but not diagnostic of PEComa, include limited smooth muscle marker expression (particularly desmin), striking nuclear atypia with only very few mitotic figures, and metastasis to lymph nodes. The less common Group A PEComas exhibit a nested growth pattern characterized by clustered or alveolar aggregates of clear cells partially encircled by either thin-walled vessels or delicate collagenous stroma (Fig. 38) (Schoolmeester et al. 2015). About one-half of cases have mildly atypical nuclei while others show a range of nuclear atypia, including rare overtly pleomorphic cases. Almost all cases display myometrial invasion, whether permeative, infiltrative, or pushing. Occasional cases have necrosis, lymphovascular invasion, or scarce melanin pigmentation. These cases bear a striking resemblance, including

10

Mesenchymal Tumors of the Uterus

575

Fig. 37 PEComa (group B). (a) The tumor is composed of nests and sheets of epithelioid tumor cells with flocculant eosinophilic cytoplasm and vesicular nuclei with

coarse chromatin and conspicuous nucleoli. (b) Positive staining for HMB-45 in a uterine PEComa (group B). Staining is typically patchy and often of medium intensity

genomic and immunohistochemical features, to Xp11-translocated renal cell carcinomas (Argani et al. 2016).

kept in mind: (1) immunophenotype differs between group A and B PEComas, and (2) the immunophenotype of group B PEComas overlaps substantially with that of uterine smooth muscle tumors, particularly epithelioid smooth muscle tumors. Group A PEComas show diffuse cytoplasmic staining with HMB-45 and cathepsin K along with strong and diffuse staining of tumor cell nuclei with TFE3 (Fig. 38). They are negative or only focally weakly positive for Melan-A, microphthalmia transcription factor (MITF), SMA, desmin, and caldesmon, and negative for SOX10 (Schoolmeester et al. 2015). Strong and diffuse TFE3 expression is the result of chromosomal translocations involving the TFE3 transcription factor gene, which maps to the Xp11.2 locus. One of the TFE3 fusion partners is SFPQ/PSF (Agaram et al. 2015), but the full repertoire of fusion partners is not yet known. Strong and diffuse nuclear TFE3 staining usually correlates well with the presence of a TFE3 translocation identified with fluorescence in situ hybridization (FISH); any equivocal immunohistochemical result should be verified by FISH. When the differential diagnosis includes metastatic Xp11 translocation-associated renal cell carcinoma, PAX8 expression confirms that diagnosis, while a negative result would be more supportive of a group A PEComa (Argani et al. 2016). ASPS, which also features

Molecular Pathology Distinguishing between the two types of PEComas is important because only Group B PEComas, particularly those found in the soft tissues, frequently harbor mutations involving TSC2 and less frequently (approximately 25% of cases) TSC1 (Thway and Fisher 2015; van Slegtenhorst et al. 1997). These mutations lead to activation of the mammalian target of rapamycin (mTOR) pathway (Goncharova et al. 2002; Martin et al. 2004), which can be therapeutically targeted by mTOR inhibitors (Wagner et al. 2010). Malignant Group B uterine PEComas harbor TSC mutations less frequently than PEComas occurring elsewhere; a recent study found that only 2 of 15 malignant uterine PEComas were found to harbor a TSC2 mutation and none had a TSC1 mutation (unpublished observations). Group A PEComas lack such mutations (Agaram et al. 2015) and therefore would not respond to mTOR blockade. Instead, chromosomal fusions involving TFE3 at the Xp11.2 locus are characteristic of these tumors, as discussed below. Immunohistochemistry Immunostaining is essential to confirm the diagnosis of PEComa, although two provisos must be

576

E. Oliva et al.

Fig. 38 PEComa (group A; TFE3-translocated/Xp11 PEComa). (a) Alveolar and solid aggregates of epithelioid tumor cells with clear cytoplasm are surrounded by a prominent vasculature, reminiscent of renal cell carcinoma of clear cell type and certain translocation-associated renal

cell carcinomas. (b) Diffuse, strong staining for HMB-45 (compare to Fig. 37b). (c) Diffuse overexpression of TFE3 in tumor cell nuclei. The presence of a chromosomal translocation involving TFE3 can be confirmed by FISH

a TFE3 translocation involving Xp11.2, would not display diffuse HMB45, actin, or PAX8 staining. In group B PEComas, the most characteristic finding is patchy positive HMB-45 staining of tumor cell cytoplasm (Fig. 37). Other melanocytic markers such as Melan-A and MiTF are also often positive; S-100 can be positive but is more often negative (Folpe et al. 2005). Positive staining for smooth muscle markers, most commonly smooth muscle actin and sometimes caldesmon or desmin, is also typical (Fukunaga 2005; Schoolmeester et al. 2014). Cytokeratin, CD117, and CD34 are generally negative. Electron microscopic study of one uterine PEComa that showed

positive immunostaining for HMB-45 revealed pre-melanosomes in the tumor cells (Park et al. 2003). Despite the fact that these immunophenotypic characteristics are considered confirmatory of PEComa, many uterine smooth muscle tumors have nearly identical immunophenotypes (Fadare 2008; Oliva et al. 2006; Silva et al. 2004, 2005; Simpson and Albores-Saavedra 2007). Although it has been posited that co-expression of a muscle marker with two melanocyte-associated markers in a tumor that resembles PEComa is sufficient to make that diagnosis, this criterion has not been validated against a gold standard (Schoolmeester et al. 2014) such as TSC1 or 2 mutation.

10

Mesenchymal Tumors of the Uterus

HMB-45 is not a specific marker of PEComa; positive HMB-45 staining has been reported in conventional and epithelioid uterine leiomyosarcomas (Hurrell and McCluggage 2005; Silva et al. 2004) and in metastases from an epithelioid leiomyosarcoma (Silva et al. 2005). In at least one case, a conventional uterine leiomyosarcoma has metastasized to the lung in the form of a purely epithelioid tumor with strong HMB45 expression. Retrospective review of the uterine primary disclosed a minute focus of malignant epithelioid cells that expressed HMB45, unlike the preponderant conventional components (unpublished observations, RAS). Because criteria for malignancy in PEComa are more permissive than for uterine smooth muscle tumors and metastatic PEComa is treated using mTOR inhibitors unlike leiomyosarcoma which is treated with chemotherapy, every effort should be made to distinguish these two tumor types. The differential diagnosis includes endometrial stromal sarcoma (HMB-45 negative), metastatic melanoma (S100 positive), ASPS (smooth muscle marker and HMB45-negative) (Schoolmeester et al. 2017), and epithelioid smooth muscle neoplasms. As discussed previously, there is dispute about whether group B PEComas are a distinct type of neoplasm or a variant of a smooth muscle neoplasm. Along these lines, there has been some work to identify PEComa-specific biomarkers, such as beta catenin (Schoolmeester and Park 2015), cathepsin K (Rao et al. 2013), and CD1a (Ahrens and Folpe 2011), among others. Cathepsin K is also expressed in uterine smooth muscle tumors, limiting its diagnostic utility, and CD1a expression has been interpreted by some as an artifact. Until the controversy is resolved, it is best to classify a tumor with the morphologic and immunohistochemical features of leiomyosarcoma as such, regardless of whether it stains for HMB-45. Uterine tumors that very closely resemble PEComas that occur in extrauterine sites may be considered candidates for that diagnosis, but such diagnoses should be verified by testing for TSC1 or 2 mutations or activation of the mTOR pathway, particularly if the tumor is obviously malignant. Group B PEComas may exist as two types: those that arise de novo or

577

in association with “PEComatosis” and those that arise in the background of a uterine smooth muscle tumor. HMB-45 staining is also typical of other tumors in the PEComa family; those that are occasionally detected in the uterus include LAM and angiomyolipoma. In angiomyolipoma, a benign tumor, HMB-45 positive tumor cells are present in the fat and among the smooth muscle cells. In LAM, which is generally a microscopic finding (Gyure et al. 1995; Torres et al. 1995), some smooth muscle cells (Fig. 39) show positive staining for HMB-45. Clinical Behavior and Treatment Both benign and malignant variants of PEComa (based on the extent of tumor at diagnosis or clinical follow-up) have been reported (Dimmler et al. 2003; Greene et al. 2003). In one recent review, 44% of corpus cases were classified as malignant and 56% as benign (Fadare 2008). Features proposed as being predictive of an unfavorable outcome include large size (>5 cm), high cellularity, significant nuclear atypia, mitotic activity (>1 MF/10 HPF), coagulative tumor cell necrosis, invasive growth, and lymphovascular space invasion (Folpe et al. 2005). Other slight

Fig. 39 LAM involving the myometrium. The tumor cells are usually spindled with oval nuclei and flocculent eosinophilic cytoplasm. The growth is centered on dilated lymphovascular spaces, and intravascular growth is typically present, often at the periphery of the area of involvement. Sometimes, it mimics the appearance of intravascular endometrial stromal sarcoma

578

variations on this theme have been published (Conlon et al. 2015). As mentioned, mTOR inhibitors may be used for group B metastatic malignant PEComas, while treatment for group A PEComas has not been studied.

E. Oliva et al.

from malignant (low-grade endometrial stromal sarcoma); otherwise, they share histologic, immunohistochemical, and molecular features.

Endometrial Stromal Nodule and LowGrade Endometrial Stromal Sarcoma Endometrial Stromal Tumors Endometrial stromal tumors represent the second most common category of mesenchymal tumors of the uterus, but they are by far much less common than smooth muscle tumors and overall account for less than 10% of all uterine mesenchymal tumors and up to 25% of all uterine sarcomas (Abeler et al. 2009). In the latest WHO classification, they are divided into four categories: (a) endometrial stromal nodule, (b) low-grade endometrial stromal sarcoma, (c) high-grade endometrial stromal sarcoma, and (d) undifferentiated uterine sarcoma (Table 8). The high-grade category was reintroduced based on its distinct morphologic, immunohistochemical, and molecular characteristics (Oliva et al. 2014). Terminology for the last category was changed from “undifferentiated endometrial stromal sarcoma” to “undifferentiated uterine sarcoma” as some of these tumors may have a smooth muscle or other cell of origin. Among these categories, low-grade endometrial stromal sarcoma is by far the most common, but still only represents 20–30 MF/10 HPF). Cells are arranged in vague nests or as a diffuse growth associated with a delicate sinusoidal vasculature (Fig. 61). A focal pseudopapillary, cord-

E. Oliva et al.

like or rosette-like morphology has been reported (Amant et al. 2011; Lee et al. 2012b). Approximately 50% of these tumors display a second component characterized in most instances by a low-grade spindle morphology with or without associated myxoid background, as typically seen in low-grade fibromyxoid endometrial stromal sarcomas. This component is hypocellular and displays oval-to-spindle cells with bland cytologic features embedded in a delicate collagenous or myxoid background but with the arteriolar vasculature that characterizes typical endometrial stromal tumors (Oliva et al. 1999; Yilmaz et al. 2002). Rarely, YWHAE-FAM22 high-grade endometrial stromal sarcomas have evolved from a conventional low-grade endometrial stromal sarcoma at a metastatic site (Aisagbonhi et al. 2017). The high-grade component of these tumors is typically strongly and diffusely positive for cyclin D1 (>70% of cells) (Fig. 62a), and in contrast to low-grade tumors, negative for CD10 (Fig. 62b). ER and PR may be positive, but only minimally (Lee et al. 2012a). They also express BCOR and c-kit (without associated mutations) but are negative for DOG1 (Chiang et al. 2017a; Lee et al. 2014). BCOR is a more sensitive marker than cyclin D1 in the detection of these tumors (Chiang et al. 2017a). In contrast, the low-grade component expresses CD10, ER, and PR but no or very little cyclin D1. They may show focal and weak staining with IFITM1 (Busca et al. 2017; Parra-Herran et al. 2014). These tumors display t(10;17)(q22;p13) rearrangements associated with YWHAE-FAM22 (also known as YWHAENUTM2) fusions that can be detected by FISH or RT-PCR. FISH appears to be more sensitive, with a cutoff of at least 20–30% positive cells, though lower detection may still be indicative of this translocation (Croce et al. 2013; Kruse et al. 2014a). An identical fusion has been reported in clear cell sarcoma of the kidney (Punnett et al. 1989; Rakheja et al. 2004). This translocation has been rarely detected in endometrial stromal tumors with typical low-grade or variant morphologic features. Although this tumor has a very characteristic morphologic appearance, diagnostic

10

Mesenchymal Tumors of the Uterus

Fig. 61 YWHAE-FAM22 high-grade endometrial stromal sarcoma. (a) The tumor is hypercellular and characterized by a diffuse or nested growth of small uniform cells. (b) Cells are round with scant cytoplasm,

593

contain round to angulated nuclei without visible nucleoli, and display associated brisk mitotic activity. These tumors may contain a component of low-grade endometrial stromal sarcoma

Fig. 62 YWHAE-FAM22 high-grade endometrial stromal sarcoma. Tumor cells are diffusely positive for cyclin D1 (a) but negative for CD10 (b), in contrast to typical endometrial stromal tumors

considerations, especially in biopsy specimens, include undifferentiated/dedifferentiated carcinoma, epithelioid leiomyosarcoma, primitive neuroectodermal tumor (PNET), undifferentiated uterine sarcoma, gastrointestinal tumor (if the biopsy is of an extrauterine mass), or metastases. Undifferentiated/dedifferentiated carcinomas are composed of monomorphous noncohesive medium-sized round cells that may show strong and diffuse cyclin D1 staining and are typically negative for PAX8 and EMA, features that overlap with YWHAE-NUTM2 endometrial stromal

sarcoma (Shah and McCluggage 2015), especially if no associated low-grade carcinoma is present. Undifferentiated carcinomas have a striking diffuse growth and stain for keratin cocktail and keratin 8/18, and half display concurrent loss of MLH1 and PMS2 or loss of E-cadherin and CD44 (Ramalingam et al. 2016). Epithelioid leiomyosarcoma may display round cells and rarely can be cyclin D1 and BCOR positive (Chiang et al. 2017a; Lee et al. 2012a). However, they also commonly display an overtly malignant spindle cell component, as

594

well as positivity for smooth muscle markers including desmin and h-caldesmon, and are often keratin and EMA positive. PNETs are exceedingly rare in the uterus, but because they are highly cellular and may display rosettes, they may enter in the differential diagnosis of YWHAE-NUTM2 endometrial stromal sarcomas. Furthermore, the latter can, on occasion, express CD99, a common marker in PNETs. However, PNET cells lack cytoplasm and, if of central type, half express GFAP, while those of peripheral type are associated with EWSR1 rearrangement (Chiang et al. 2017a; Euscher et al. 2008). Rarer tumors that have morphologic and immunophenotypic features of Ewing sarcoma/peripheral PNET, but lack EWSR1 rearrangement, may enter in the differential diagnosis, although they have not been reported yet in the uterus; these harbor less common alterations affecting FUS, BCOR, CCNB3, CIC, or DUX4 (Hung et al. 2016). Undifferentiated sarcomas may rarely enter in the differential diagnosis as they may be extensively positive for cyclin D1. However, they also show extensive and strong CD10 staining and more importantly typically display marked pleomorphism as well as destructive myometrial invasion, in contrast to YWHAE-NUTM2 endometrial stromal sarcomas (Oliva et al. 2014; Sciallis et al. 2014). Epithelioid gastrointestinal stromal tumors and YWHAE-NUTM2 endometrial stromal sarcomas share expression of c-kit, but the latter typically involves primarily the uterus, lacks c-kit mutations, and does not express DOG1 (Miettinen and Lasota 2011; Novelli et al. 2010; Terada 2009). These tumors are detected at more advanced stages (stage II or III) when compared to low-grade endometrial stromal sarcomas, and patients more often develop recurrences, usually early, having an intermediate behavior between low-grade endometrial stromal sarcomas and undifferentiated uterine sarcomas (Kruse et al. 2014a; Lee et al. 2012b). Standard treatment is hysterectomy followed by chemotherapy with or without radiation. Chemotherapy with anthracycline-based drugs led to complete radiologic response in a small series (Hemming et al. 2017). YWHAE-NUTM2

E. Oliva et al.

endometrial stromal sarcoma may not be present in the primary tumor but in recurrences or metastases (Aisagbonhi et al. 2017). Although immunohistochemical or molecular tools may allow detection of low-grade tumors with YWHAE-NUTM2 fusion, it is unknown which tumors will evolve to a high-grade endometrial stromal sarcoma.

ZC3H7B-BCOR High-grade Endometrial Stromal Sarcoma ZC3H7B-BCOR endometrial stromal sarcoma is a newly described subtype of high-grade endometrial stromal sarcoma that closely mimics the appearance of myxoid leiomyosarcoma. Although its frequency is reported as low, it may in fact be higher, as in the past these tumors were frequently misdiagnosed as leiomyosarcoma (Hoang et al. 2017; Lewis et al. 2018; Marino-Enriquez et al. 2018). In the only series reported to date, typical tumors occur within a wide age range (28–71 years), but most frequently in the fifth decade. A subgroup characterized by BCOR internal tandem duplications (BCOR-ITD) occurs in younger patients (18–32 years) (Marino-Enriquez et al. 2018). Patients present with nonspecific symptoms including vaginal bleeding and/or pelvic mass, and not infrequently have extrauterine disease at initial diagnosis (Hoang et al. 2017; Lewis et al. 2018). On gross examination, these tumors are often large (mean 10 cm), polypoid, and centered in the endometrium, but can be myometrial-based. They have a solid, soft, fleshy-to-rubbery, tan-to-yellow-to-pink cut surface. On low-power microscopic examination, they typically show a tongue-like, broad front, or destructive (least common) pattern of invasion, or a combination thereof. Tumors tend to be uniformly cellular growing in haphazard fascicles of spindle cells without overt pleomorphism (Fig. 63). Cells have scant-to-relatively-abundant gray-to-eosinophilic cytoplasm and oval-to-spindle nuclei with inconspicuous nucleoli and evenly distributed chromatin. BCOR-ITD tumors often have a component

10

Mesenchymal Tumors of the Uterus

595

of round epithelioid cells admixed with the spindle cells. Mitotic activity often exceeds 15 MF/10 HPF. Vascularity, when prominent, is characterized by small arterioles without striking perivascular whorling of tumor cells but may include large and/or hemangiopericytoma-like vessels. The background stroma is either myxoid, including variably sized pools of basophilic material, or collagenous; collagen plaques are commonly seen (Fig. 63). Areas of necrosis and lymphovascular invasion are frequent (Hoang et al. 2017; Lewis et al. 2018). In contrast to YWHAE-NUTMT2 high-grade endometrial stromal sarcomas, these tumors are not

associated with a conventional or variant component of endometrial stromal sarcoma. The immunohistochemical profile of typical ZC3H7B-BCOR tumors and those with BCORITD closely overlaps, except for CD10 expression, which is typically positive, often with a diffuse and strong pattern of staining in typical ZC3H7B-BCOR tumors but negative or only focally positive staining in BCOR-ITD tumors (Hoang et al. 2017; Lewis et al. 2018; MarinoEnriquez et al. 2018). ER and PR are variably positive. Cyclin D1 is strong and diffuse in most tumors, and BCOR is also frequently positive (Fig. 64b), although staining ranges from weak

Fig. 63 ZC3H7B-BCOR high-grade endometrial stromal sarcoma. (a) The tumor may have a vague fascicular growth of tumor cells associated with prominent myxoid

background, mimicking a myxoid leiomyosarcoma. (b) Other areas contain spindle cells associated with collagen plaques, as seen in typical endometrial stromal tumors

Fig. 64 ZC3H7B-BCOR high-grade endometrial stromal sarcoma. The tumor is frequently positive for BCOR (a) and cyclin D1, and negative for caldesmon (b), in contrast to myxoid leiomyosarcomas

596

to strong. There tends to be concordant positivity between cyclin D1 and BCOR (Chiang et al. 2017a; Hoang et al. 2017; Lewis et al. 2018). Among smooth muscle markers, actin is more frequently expressed (typically focally), while desmin and caldesmon are almost always negative (Fig. 64a). These tumors are characterized by ZC3H7B-BCOR fusions that result from the t(X;22)(p11q13) rearrangement, which can be detected by FISH or PCR. A minority of tumors show internal tandem duplications (ITDs) involving exon 15 of BCOR (Hoang et al. 2017; Lewis et al. 2018; Marino-Enriquez et al. 2018). Myxoid leiomyosarcoma is the most common entity in the differential diagnosis of these highgrade endometrial stromal sarcomas and in the past were likely diagnosed as such. Distinguishing features of ZC3H7B-BCOR stromal sarcomas include lack of well-formed fascicles, cells with cigar-shaped nuclei, and expression of desmin and caldesmon. In one study, only 1 of 19 leiomyosarcomas was positive for BCOR (Chiang et al. 2017a), while in another study only 1 of 80 such tumors was diffusely positive for cyclin D1 (Lee et al. 2012a). Low-grade endometrial stromal sarcoma may enter in the differential diagnosis, as both share a tongue-like pattern of myometrial invasion as well as cells with uniform cytologic features, collagen bands, and in some instances CD10, ER, and PR positivity. However, BCOR-ITD and typical ZC3H7BBCOR high-grade endometrial stromal sarcomas lack the features reminiscent of proliferative phase endometrium, including the characteristic vasculature, and display brisk mitotic activity. Furthermore, low-grade endometrial stromal sarcomas express CD10, ER, and PR, but not BCOR or cyclin D1 (with rare exceptions; staining is only focal and weak) (Chiang et al. 2017a). Rarely, adenosarcoma with sarcomatous overgrowth and IMT may enter in the differential diagnosis of these tumors, as they can also display a myxoid background. However, the former typically has characteristic areas of low-grade müllerian adenosarcoma, and thus far has not been associated with BCOR genetic alterations (although immunohistochemical expression of BCOR can occur) (Howitt et al. 2015b), while the latter

E. Oliva et al.

contains an inflammatory infiltrate and overexpresses ALK as a result of genetic rearrangements. Finally, undifferentiated uterine sarcoma may be considered in the differential diagnosis, and in fact, two out of three BCORITD high-grade endometrial stromal sarcomas were originally diagnosed as undifferentiated uterine sarcomas (Marino-Enriquez et al. 2018). The latter is in general more pleomorphic, without associated collagen plaques, and typically has a destructive invasion pattern. Hysterectomy and bilateral salpingo-oophorectomy with adjuvant chemotherapy is standard treatment. Although experience with these tumors is limited, patients with ZC3H7B-BCOR highgrade endometrial stromal sarcomas have a prognosis that parallels that of patients with YWHAENUTM2 high-grade sarcomas, as both are associated with higher stage at presentation (including lymph node metastases) and frequent recurrences and metastases (Lewis et al. 2018; Marino-Enriquez et al. 2018).

Other High-Grade Endometrial Stromal Sarcomas Rarely, a high-grade pleomorphic or heterologous sarcoma may be seen in association with a low-grade endometrial stromal sarcoma, in which instance a diagnosis of high-grade endometrial stromal sarcoma can be rendered (Amant et al. 2006; Cheung et al. 1996; Kurihara et al. 2008; Malpica et al. 2006; McCluggage and Young 2008; Ohta et al. 2010; Sciallis et al. 2014); however, some investigators designate these tumors as dedifferentiated endometrial stromal sarcoma. Although many monomorphic undifferentiated uterine sarcomas are likely to represent high-grade endometrial stromal sarcoma (Sciallis et al. 2014), the current WHO maintains the “undifferentiated uterine sarcoma” nomenclature (Oliva et al. 2014). Patients often present with vaginal bleeding or a uterine mass and have extrauterine disease at the time of diagnosis. The highgrade component may show a permeative or destructive pattern of invasion. Cells may be pleomorphic or uniform, with epithelioid or spindle

10

Mesenchymal Tumors of the Uterus

morphology, variable amounts of cytoplasm, and hyperchromatic and irregular nuclei with prominent nucleoli and high mitotic index, including atypical forms. The cells grow in a diffuse manner and are associated with extensive areas of necrosis and, rarely, heterologous differentiation (rhabdomyosarcoma) (Kurihara et al. 2008; Sciallis et al. 2014). Areas of conventional low-grade endometrial stromal sarcoma are present in variable extent. By immunohistochemistry, CD10, ER, and PR are often negative in the high-grade component when pleomorphic but typically positive in the low-grade component (Malpica et al. 2006; Sciallis et al. 2014). Staining for cyclin D1 may be focally positive or there may be p53 overexpression (Jung et al. 2008; Kurihara et al. 2008, 2010; Ohta et al. 2010). They may express AE1/3 and CAM 5.2 (Rahimi et al. 2018). Fusion of JAZF1 and JJAZ1 genes has been identified in these tumors on rare occasions (Koontz et al. 2001; Kurihara et al. 2008).

Undifferentiated Uterine Sarcoma Undifferentiated uterine sarcoma is an extremely rare and heterogeneous category of malignant mesenchymal tumors, defined by the most recent WHO classification as neoplasms that arise in the endometrium or myometrium with high-grade cytologic features lacking any resemblance to proliferativephase endometrial stroma and with no specific differentiation. These tumors are difficult to classify, and thus, their diagnosis is one of exclusion after other possibilities, including high-grade endometrial stromal sarcoma, undifferentiated/ dedifferentiated carcinoma, adenosarcoma with sarcomatous overgrowth, carcinosarcoma (MMMT) with minimal epithelial component, or dedifferentiated leiomyosarcoma have been ruled out (Oliva et al. 2014). Before publication of the recent WHO classification, Kurihara and colleagues analyzed a group of undifferentiated endometrial sarcomas and divided them into uniform and pleomorphic types (Kurihara et al. 2008). In the former group, some tumors were ER- and PR-positive and had mutations in β-catenin but lacked p53 mutations and showed a component of

597

low-grade endometrial stromal sarcoma, while others were ER- and PR-negative and diffusely cyclin D1-positive. Thus, these authors recognized that most monomorphic undifferentiated sarcomas could be categorized into immunohistochemically defined entities that likely correspond to YWHAEand BCOR-rearranged high-grade endometrial stromal sarcomas. Tumors in the third, pleomorphic category were typically ER-, PR-, and β-catenin-negative but often expressed p53; this very small group of tumours likely represents true undifferentiated sarcomas. Undifferentiated uterine sarcomas typically occur in postmenopausal women that present with vaginal bleeding and signs and symptoms related to a rapidly growing mass and/or metastatic disease (Kurihara et al. 2008; Prat and Mbatani 2015; Tanner et al. 2012). On gross examination, these tumors are typically large fleshy masses that may involve the endometrium and/or myometrium, some being polypoid; they are often associated with extensive areas of hemorrhage and necrosis (Fig. 65). On microscopic examination, pleomorphic undifferentiated uterine sarcomas are characterized by a destructive pattern of invasion. Like other highgrade sarcomas (Fig. 66), the spindle or epithelioid tumor cells often grow in sheets or fascicles and display marked cytologic atypia, including multinucleation and brisk mitotic activity with atypical mitoses (Fig. 67). Necrosis (Fig. 68) and

Fig. 65 Undifferentiated uterine sarcoma. The tumor is polypoid, fills the uterine cavity, and displays a white, fleshy cut surface with extensive areas of necrosis

598

E. Oliva et al.

Fig. 66 Undifferentiated uterine sarcoma. The invasion pattern is destructive, in contrast to the tongue-like permeative invasion seen in endometrial stromal sarcoma

Fig. 69 Undifferentiated uterine sarcoma. These are very aggressive tumors that are frequently associated with lymphovascular invasion as revealed by a CD31 stain

Fig. 67 Undifferentiated uterine sarcoma. The tumor is composed of pleomorphic hyperchromatic cells including multinucleated cells. This is a diagnosis of exclusion

Fig. 70 Undifferentiated uterine sarcoma. The tumor cells may show extensive staining for CD10, but they are often negative for ER and PR

Fig. 68 Undifferentiated uterine sarcoma. These tumors often contain extensive areas of tumor cell necrosis

lymphovascular invasion are common (Fig. 69) (Bartosch et al. 2010; Evans 1982). These tumors may show variable staining for CD10 (Fig. 70), but pleomorphic sarcomas are typically ER- and PR-negative (Bartosch et al. 2010; Gremel et al. 2015; Kurihara et al. 2010). Cyclin D1 may be variably expressed. Pleomorphic undifferentiated sarcomas may also express p16, p53 (Gremel et al. 2015); other immunohistochemical markers including keratin, smooth muscle actin, or desmin may be positive, but only focally (Bartosch et al. 2010; Kurihara et al. 2008). Gene expression

10

Mesenchymal Tumors of the Uterus

studies have found that pleomorphic undifferentiated uterine sarcomas harbor more chromosomal alterations and complex karyotypes than typical endometrial stromal sarcomas. Data suggest that progression from typical endometrial stromal sarcomas to undifferentiated sarcomas is unlikely although not impossible (Flicker et al. 2015; Gil-Benso et al. 1999; Halbwedl et al. 2005; Micci et al. 2016). JAZF1-SUZ12 rearrangements have rarely been detected in undifferentiated uterine sarcomas (Koontz et al. 2001), particularly monomorphic sarcomas, but a recent study failed to identify this rearrangement (Jakate et al. 2013). Another study identified five copy number alterations encompassing cancer-related genes (EZR, CDH1, RB1, TP53, and PRKAR1A) accompanied by corresponding expression changes in undifferentiated uterine sarcomas, suggesting that they may have some impact on the development of these tumors (Choi et al. 2015). As this is a diagnosis of extensive sampling, immunohistochemistry and, if required, analysis for stromal sarcoma-related fusions are strongly recommended to exclude entities in the differential diagnosis. CD10 expression is insufficient evidence for a diagnosis of undifferentiated uterine sarcoma, as this marker is positive in most tumors (epithelial, mesenchymal, and mixed) that occur in the uterus. Very recently, a very small number of undifferentiated uterine sarcomas have been reported to display a rhabdoid morphology with loss of expression of SMARCA4 (Kolin et al. 2018). Overall, these tumors are associated with poor patient outcomes, even among those diagnosed with stage I tumors, despite aggressive treatment; a large number of patients have disease outside the uterus at the time of diagnosis (Evans 1982; Kurihara et al. 2008). Survivals for patients with pleomorphic undifferentiated sarcomas appear to have worse clinical outcomes compared to those with monomorphic tumors. Five-year survival rates of 70%, 43%, and 23% have been reported for localized, regional, and distant disease (American Cancer Society 2017). A

599

recent study stratified the prognosis of undifferentiated uterine sarcomas based on mitotic index, hormone receptor expression, and YWHAE-FAM22 translocation status. Tumors with >25 MF/10 HPF lacking ER, PR, and YWHAE-FAM22 translocation had poorer prognoses (Gremel et al. 2015).

Mixed Epithelial–Mesenchymal Tumors Mixed epithelial–mesenchymal tumors contain both epithelial and mesenchymal elements. The mixed epithelial–mesenchymal tumor group as listed in the 2014 WHO classification of tumors of the uterine corpus (Table 1) includes adenomyoma, atypical polypoid adenomyoma (APA), adenofibroma, adenosarcoma, and carcinosarcoma (also known MMMT) (Kurman et al. 2014). APA typically originates in the endometrium, and carcinosarcoma is considered to be a special type of sarcomatoid or metaplastic endometrial carcinoma, so in this book they are discussed in the chapters on benign diseases of the endometrium and endometrial carcinoma, respectively.

Adenomyosis and Adenomyoma Adenomyosis is a common condition, detected in 15–30% of hysterectomy specimens. It is characterized by the presence of endometrial glands and stroma within the myometrium. Adenomyomas are uncommon tumor-like masses composed of endometrial glands, endometrial stroma, and smooth muscle, with the latter generally predominating. They differ from adenomyosis mainly in that they are circumscribed nodular masses.

Clinical Features Patients are typically pre- or perimenopausal women who present with abnormal bleeding and dysmenorrhea (Gordts et al. 2018). Younger patients may present due to impaired

600

reproduction, and up to a third of patients with adenomyosis are asymptomatic. Symptoms tend to be more severe in women with deep myometrial involvement. The uterus is typically enlarged and may harbor other lesions associated with hyperestrinism, such as leiomyomas, pelvic endometriosis, and endometrial polyps. Adenomyosis is usually most extensive in the posterior wall, which may be thickened. A clinical diagnosis of adenomyosis can often be confirmed by imaging studies such as transvaginal ultrasonography or MRI.

Pathologic Findings On gross examination, the cut surface of the myometrium is trabeculated and contains hemorrhagic foci, but a distinct tumor nodule is not present. Small blood-filled cysts may be noted. Adenomyosis is a condition in which rounded or irregular foci of endometrial stroma and glands are present in the myometrium, usually surrounded by bundles of hypertrophic myometrial smooth muscle (Fig. 71). Cystic or blood-filled endometrial glands are sometimes present in foci of adenomyosis, accounting for the small blood-filled cysts that are seen on gross examination in some cases. The lower border of the endometrium is irregular and dips into the superficial myometrium. To avoid misclassifying a normal histologic finding as adenomyosis, the diagnosis is made only when the distance between the lower border of the endometrium and the adenomyosis exceeds the diameter of a 100x microscopic field (about 2 mm), an admittedly arbitrary measurement (Bergeron et al. 2006). Adenomyosis exhibits a varied functional response to ovarian hormones. Proliferative glands and stroma generally are observed in the first half of the menstrual cycle. Adenomyosis may not respond to physiologic levels of progesterone, and secretory changes frequently are absent or incomplete during the second half of the cycle. Variants of adenomyosis can suggest a malignant tumor. In one of these, endometrial tissue protrudes into myometrial vessels, simulating vascular invasion by a neoplasm, such as an endometrial stromal sarcoma (Fig. 72a). Intravascular

E. Oliva et al.

Fig. 71 Adenomyosis. The periphery of this focus of adenomyosis containing endometrial hyperplasia is relatively lacking in endometrial stroma. This focus can be recognized as adenomyosis because of its lobular shape and the circumferential smooth muscle hypertrophy that surrounds it

endometrial tissue, consisting of endometrial stroma and glands or endometrial stroma alone, can be found in 5–12% of uteri with adenomyosis (Meenakshi and McCluggage 2010; Sahin et al. 1989). Intravascular adenomyosis usually appears to originate in the perivascular region and push into the vessel lumen; immunostains reveal the intravascular adenomyosis to be covered by a layer of CD31-positive endothelial cells (Fig. 72b). Cases of IVL may also contain endometrial glands and stroma and have been designated “intravascular adenomyomatosis” (Hirschowitz et al. 2013). Other problematic variants of adenomyosis include those where either the glandular or stromal component is altered or sparse. A low power clue that these represent adenomyosis is the circumferential muscular hypertrophy that surrounds foci of adenomyosis and the maintenance of a lobular appearance without surrounding stromal desmoplasia. In the gland-poor variant that tends to occur in elderly women, sometimes designated “adenomyosis with sparse glands,” glands are few in number and some adenomyotic foci consist mainly or exclusively of endometrial stromal cells (Goldblum et al. 1995). Careful evaluation reveals that these variants of adenomyosis lack features of malignancy such as mitotic figures in the stromal cells and invasion into the surrounding myometrium and that they are almost always

10

Mesenchymal Tumors of the Uterus

Fig. 72 Intravascular endometrial tissue in a patient with adenomyosis. (a) In this example no glands are present, raising the possibility of endometrial stromal sarcoma. However, there was no mass and no myoinvasive stromal tumor was present. Typical foci of adenomyosis were widely present in the region. (b) An immunostain for

Fig. 73 Adenomyosis with sparse glands. Nearby foci of adenomyosis contained both glands and stroma, but this one consists of stroma only. The center is less cellular and appears pale. The periphery is more cellular and therefore is more darkly stained

accompanied by foci of typical adenomyosis. The gland-poor variant can be suspected at low power examination, as the stroma is frequently less cellular centrally than at the periphery, resulting in a focus of adenomyosis with a pale center and a darkly stained peripheral zone (Fig. 73). The stromal component can also be less conspicuous than usual and undergo modifications that make it difficult to recognize. It can be atrophic and fibrotic and may resemble the stroma of an atrophic endometrial polyp, with an eosinophilic, fibrillary appearance and loss of the monotonous, blue ovoid cells more typically associated with

601

CD31 reveals that the vein is lined by a layer of endothelial cells. The intravascular endometrial tissue is completely covered by a layer of CD31-positive endothelial cells, suggesting that it is extravascular and has protruded into the vein lumen, pushing the endothelium over it

endometrial stroma. The eosinophilic, fibrillary appearance may superficially resemble myometrium, but high-power examination can reveal a clear demarcation between the surrounding, hypertrophic, well-organized bundles of myometrium and the disorganized, thin fibrils of altered endometrial stroma. Most examples of adenomyosis contain endometrial stroma with a CD10-positive immunophenotype. The eosinophilic, fibrillary stroma found in atrophic adenomyosis frequently expresses CD10 only weakly or focally, so absent CD10 staining does not entirely exclude the presence of endometrial stroma and adenomyosis. An adenomyoma is a circumscribed mass composed of smooth muscle, endometrial glands, and endometrial stroma (Gilks et al. 2000; Tahlan et al. 2006). The average patient age is 40–49, and the most common presentation is with abnormal bleeding. Leiomyomas may also be present in the uterus. Grossly, an adenomyoma may be a rounded, sometimes cystic, tan mass within the myometrium, or it may involve or originate in the endometrium and grow as a polyp. About 2% of endometrial polyps are adenomyomas. Microscopically, endometrial glands and endometrial stroma are present within the mass, which typically consists mainly of smooth muscle. The

602

smooth muscle component of an adenomyoma generally consists of typical spindle-shaped smooth muscle cells, but epithelioid smooth muscle differentiation also occasionally occurs (Kenny and McCluggage 2014). Adenomyomas also occur in the cervix, where they are more likely to be asymptomatic (Casey and McCluggage 2015; Gilks et al. 1996). They can grow as polyps that sometimes prolapse through the external os or as mural nodules. Cervical adenomyomas are composed of smooth muscle usually admixed with glands lined by columnar mucinous endocervical-type epithelium. The glands not infrequently have at least a focal lobular arrangement. Some also contain a component of tubal-type epithelium or endometrial-type glands and stroma. The admixture of endocervical-type glands and smooth muscle can raise the possibility of a minimal deviation adenocarcinoma (“adenoma malignum”). Unlike minimal deviation adenocarcinoma, a cervical adenomyoma is circumscribed or polypoid, does not infiltrate surrounding tissues, and does not display atypia, mitotic activity, or stromal reaction. Immunohistochemical stains for ER are generally positive in the glandular cells of an adenomyoma, while ER is generally negative in minimal deviation adenocarcinoma. A rare variant of an adenomyomatous polyp, the APA, has atypical hyperplastic glands that usually contain foci of squamous metaplasia (Longacre et al. 1996; Heatley 2006). APA is discussed in detail in the chapter on benign endometrial conditions (▶ Chap. 7, “Benign Diseases of the Endometrium”). Various types of malignant neoplasms, such as variants of endometrial adenocarcinoma (Abushahin et al. 2011; Koike et al. 2013; Koshiyama et al. 2002) and adenosarcoma (Elshafie et al. 2013), have been reported to rarely arise in adenomyosis or in an adenomyoma.

E. Oliva et al.

and Norris 1981). It is composed of an admixture of histologically benign epithelial and mesenchymal elements.

Clinical Features and Gross Findings Women with adenofibromas tend to be elderly. The median age is 68 years, and most patients are peri- or postmenopausal. Despite a predilection for the elderly, adenofibroma occurs in women of all ages, from less than 20 years to more than 80 years. There is no known association with race, nor does adenofibroma have the epidemiologic features of endometrial carcinoma. Abnormal vaginal bleeding is the most frequent complaint. Less common findings include abdominal pain, abdominal enlargement, or a polypoid tumor projecting from the cervix. Some patients have a history of prior removal of polyps. Adenofibroma is a lobulated polypoid tumor that can arise anywhere in the uterus or in the cervix. It varies from soft to firm and is tan or brown. About 50% of adenofibromas contain small cysts that give the cut surface a spongy or mucoid appearance. The tumor ranges from 2 to 20 cm in maximum diameter, with a median of 7 cm. A large adenofibroma may fill the endometrial cavity and enlarge the uterus. Microscopic Findings Adenofibroma is composed of a mixture of histologically bland epithelium and mesenchyme that originates in the endometrium or cervix. Broad

Adenofibroma First described in the cervix (Abell 1971), adenofibroma is a benign neoplasm that more typically occurs in the endometrium (Zaloudek

Fig. 74 Adenofibroma. Cleft-like glands and polypoid stromal projections are lined by benign epithelium in adenofibroma

10

Mesenchymal Tumors of the Uterus

papillary or polypoid fronds covered by epithelium project from the surface of the neoplasm and extend into cystic spaces within it (Fig. 74). Columnar or cuboidal epithelial cells, most often of endometrioid type, line cysts and cleft-like spaces. A mixture of various types of epithelia, including endocervical, tubal, and squamous, often occurs within the same neoplasm. The epithelium can be hyperplastic and stratified, but in this case the possibility that the tumor might be an APA should be carefully considered. Endometrioid and serous carcinoma have both been reported to involve adenofibromas; in such cases, the behavior is determined by the carcinoma and the patient should be treated accordingly (Miller and McClure 1992; Venkatraman et al. 2003). The mesenchymal component is usually fibrous, consisting of fibroblasts and collagen (Fig. 75), but mixtures of endometrial stromal cells and fibroblasts are present in some neoplasms. The cellularity of the stroma is generally low and there is no periglandular condensation of stromal cells. The mesenchymal cells exhibit no nuclear atypia or mitotic activity. Rarely, histologically benign heterologous elements such as fat or skeletal muscle are present (Akbulut et al. 2008; Horie et al. 1995; Sinkre

Fig. 75 Adenofibroma. The epithelium in an adenofibroma is benign, and in this example it is of endometrioid type. The stromal cells are benign and have pale oval or fusiform nuclei and ill-defined cell borders. No mitotic figures, hypercellular stroma, atypia or mitotic figures are present

603

et al. 2000b). Adenofibromas are usually confined to the endometrium or cervical mucosa and do not invade the underlying myometrium or cervical stroma (Fig. 76). Two unique tumors that were classified as adenofibromas invaded the myometrium, and, in one case, myometrial veins (Clement and Scully 1990a). These tumors might well be viewed as adenosarcomas using current diagnostic criteria, since the stroma was moderately cellular and some mitotic activity was present in it; in general, any tumor classified as adenosarcoma for practical purposes is more likely to be an adenosarcoma than an adenofibroma.

Differential Diagnosis Adenofibromas and benign endometrial or endocervical polyps can be difficult to distinguish. A papillary configuration or a vaguely “phyllodeslike” pattern in which polypoid stromal cores covered by benign epithelium project from the surface or into cystically dilated glands favors an adenofibroma. The stromal component of an adenofibroma tends to be more fibrous and more uniform than the stroma of a typical polyp. The most important differential diagnosis is with adenosarcoma because adenofibroma and adenosarcoma have a somewhat similar appearance at low magnification. In the past this

Fig. 76 Adenofibroma. A typical superficial tumor that is limited to the endometrium and does not invade the myometrium. No mitotic figures, hypercellular stroma, atypia or mitotic figures are present

604

differential diagnosis rested on finding hypercellular or atypical stroma with more than 4 MF/10 HPF in an adenosarcoma (Kaku et al. 1992; Zaloudek and Norris 1981). Over time the diagnostic criteria for adenosarcoma have been broadened to the point that some authors now doubt the existence of adenofibromas (Gallardo and Prat 2009) or at least doubt that they can be diagnosed in any type of specimen other than a hysterectomy (McCluggage 2016). We think that adenofibroma can be diagnosed, but only upon examination of the entire tumor to rule out the presence of abnormal areas indicative of adenosarcoma; this usually requires a hysterectomy. Features that are indicative of adenosarcoma include hypercellular periglandular stroma, stromal cell atypia, and virtually any detectable mitotic activity (Kurman et al. 2014). A practical approach is to classify any tumor with the appropriate architecture, a cellular or atypical stroma, and more than a rare mitotic figure as an adenosarcoma. Occasionally, a problematic adenofibromatous tumor occurs in a young woman where conservation of fertility is an important consideration. Such tumors can be designated as “atypical adenofibromatous tumors,” but the pathology report should contain a note cautioning that a recurrence could have fully developed features of an adenosarcoma.

Clinical Behavior and Treatment Hysterectomy is the preferred treatment for an adenofibroma because the neoplasm may recur if it is incompletely curetted or excised. Hysterectomy ensures complete removal and also permits the thorough sampling needed to exclude an adenosarcoma. Conservative therapy, such as hysteroscopy with repeat curettage or targeted resection, can be considered in situations in which hysterectomy is not the first choice of treatment, such as in a young woman who wishes to preserve her fertility. Adenofibroma is benign and no tumor-related deaths have been reported. Importantly, the clinical behavior of tumors with unusual gross or microscopic features is unclear and pathologists should be cautious about diagnosing such tumors as adenofibromas.

E. Oliva et al.

Adenosarcoma Initially reported by Clement and Scully (1974), adenosarcoma is a biphasic tumor with benign epithelial elements and a sarcomatous stroma (McCluggage 2010). It comprises 5–6% of uterine sarcomas (Abeler et al. 2009) and most often occurs in the endometrium, although it can also arise in the cervix (Jones and Lefkowitz 1995) and in extrauterine pelvic locations such as the fallopian tube, ovary, and paraovarian tissues. Rarely, synchronous tumors occur in the uterus and an extrauterine site, such as the ovary.

Clinical Features Adenosarcoma occurs in women of all ages. The median age is 50–59 years, with a range of 15–90 years. Extrauterine adenosarcoma occurs in younger women and is more aggressive than its uterine counterpart. Adenosarcoma is not associated with obesity or hypertension. A few patients have a history of prior pelvic radiation, and occasional patients are diabetic. A few cases of adenosarcoma have been reported in women who have been treated for breast cancer with tamoxifen. Some patients have recurrent cervical or endometrial polyps and give a history of one or more prior polypectomies. The most common presenting symptom is abnormal vaginal bleeding. Vaginal discharge, pain, nonspecific urinary symptoms, a palpable pelvic mass, and a tumor protruding from the

Fig. 77 Adenosarcoma. A polypoid tumor arises in the endometrium and fills the endometrial cavity

10

Mesenchymal Tumors of the Uterus

605

cervix are other common signs and symptoms. Most patients have stage I tumors at the time of diagnosis.

Gross Findings Adenosarcoma most often arises in the endometrium and fills the uterine cavity, often resulting in an enlarged uterus. Rare tumors grow as nodules in the myometrium, presumably arising in adenomyosis. Adenosarcoma arises in the cervix in 5–10% of cases. Adenosarcoma is usually polypoid and averages 5–6 cm in maximum dimension (Fig. 77), although it occasionally grows as multiple papillary or polypoid masses. It can be either soft or firm. The cut surface is tan, brown, or gray, and zones of hemorrhage and necrosis are observed in about 25% of adenosarcomas. Small cysts are present in most tumors. Microscopic Findings Tubular glands and cleft-like spaces are distributed throughout the tumor, and papillary stromal fronds covered by epithelium project from the surface and into cysts (Fig. 78), resulting in a phyllodes tumorlike appearance (Clement and Scully 1990b; Zaloudek and Norris 1981). Glands are often present along with stroma in areas of myometrial invasion, which are observed in 15–52% of adenosarcomas. The surface and glandular epithelium most often resembles inactive or proliferative endometrial epithelium. Many other types of epithelium also occur in adenosarcomas, including secretory, mucinous, squamous, and clear cell. The epithelium typically is cytologically bland, but hyperplastic and even atypical hyperplastic epithelium is occasionally noted. Small foci of low-grade endometrioid adenocarcinoma can rarely be present in an adenosarcoma, and endometrioid adenocarcinoma also occasionally occurs in the endometrium adjacent to the adenosarcoma (Clement and Scully 1990b). If the adenocarcinoma is serous carcinoma or some other high-grade type of carcinoma, the tumor is best diagnosed as carcinosarcoma rather than adenosarcoma. The mesenchymal component of an adenosarcoma is generally a low-grade homologous sarcoma such as low-grade endometrial stromal sarcoma or a fibroblastic/myofibroblastic

Fig. 78 Adenosarcoma. Papillary stromal fronds are lined by benign epithelium. The stroma is hypercellular, and the cellularity is greatest beneath the epithelium

Fig. 79 Adenosarcoma. The stroma of an adenosarcoma is more cellular than that of an adenofibroma, especially in the vicinity of the epithelial component. The stromal cells usually resemble endometrial stromal cells or fibroblasts. The nuclei can be relatively uniform but, as in this case, atypia and pleomorphism can be conspicuous

Fig. 80 Adenosarcoma. Periglandular stromal hypercellularity is a characteristic of adenosarcoma

606

sarcoma resembling the fibroblastic variant of low-grade endometrial stromal sarcoma (Fig. 79) (Clement and Scully 1990b; Gallardo and Prat 2009; Soslow et al. 2008). Smooth muscle is present in some tumors and can be conspicuous. Stromal hypercellularity is a characteristic feature of adenosarcoma and hypercellular stromal cuffs around glands (Fig. 80) or band-like hypercellular zones beneath the surface are present at least focally in almost every case. The degree of mesenchymal cell nuclear atypia is variable but is mild to moderate in most tumors. Mitotic figures are readily identified in most tumors and generally number  2–4/10 HPF. Mitotic figures tend to be most numerous in the cellular stromal cuffs around the glands. Neoplasms with the morphologic features of adenosarcoma (cellular stroma, periglandular cuffing, stromal cell atypia, and, in some cases, myometrial invasion), but in which mitotic activity is inconspicuous, can recur or metastasize. Therefore, a neoplasm with the typical appearance of an adenosarcoma in which atypical, hypercellular stroma is condensed around the epithelial elements should be diagnosed as an adenosarcoma even if there are only 1–2 MF/10 HPF. Adenosarcomas often contain bland areas indistinguishable from an adenofibroma, so extensive microscopic study may be required to identify a sarcomatous component. Trabecular, insular, or tubular arrangements of plump epithelial-like cells, some having abundant foamy cytoplasm, are present in about 5% of adenosarcomas (Fig. 81) (Clement

Fig. 81 Adenosarcoma. Sex cord-like trabeculae or tubules are occasionally present in the stroma of an adenosarcoma

E. Oliva et al.

and Scully 1989; Gallardo and Prat 2009; Hirschfield et al. 1986). These structures, which are designated as sex cord-like elements, resemble the sex cord-like structures commonly seen in endometrial stromal tumors. Occasionally, overgrowth of the sex cord-like elements dominates the histologic picture. Tumors with overgrowth of sex cord-like elements appear to have the same prognosis as typical adenosarcomas, and this histologic pattern should not be viewed as a high-grade sarcomatous pattern or as sarcomatous stromal overgrowth (Stolnicu et al. 2016). The mesenchymal component of an adenosarcoma is the significant component of the tumor in terms of histogenesis (Piscuoglio et al. 2016) and prognosis. Features of the mesenchymal component that may have prognostic significance and should be noted in the pathology report include its grade and mitotic activity, and the presence or absence of myometrial invasion, heterologous mesenchymal elements, and sarcomatous overgrowth. In general, a high-grade sarcomatous component means that the tumor cells are markedly atypical, similar to those in pure high-grade sarcomas of the uterus and soft tissue. The nuclei in such tumors have been characterized as showing 3 + atypia on a scale of 0–3+ (Clement and Scully 1990b; Zaloudek and Norris 1981), or as showing nuclear atypia and pleomorphism identifiable at

Fig. 82 Adenosarcoma with sarcomatous overgrowth. A high-grade spindle cell sarcoma has developed in an adenosarcoma and comprises more than 25% of the tumor volume

10

Mesenchymal Tumors of the Uterus

Fig. 83 Adenosarcoma with rhabdomyosarcoma in zones of sarcomatous overgrowth. Rhabdomyoblasts are round or spindle-shaped and have prominent eosinophilic cytoplasm. It is sometimes possible to identify cross striations, but these days immunohistochemistry is typically used to confirm the presence of rhabdomyoblasts

low magnification (Gallardo and Prat 2009; Hodgson et al. 2017). Sarcomatous overgrowth, reported to be present in 33–50% of cases, is said to occur when the sarcomatous component of the tumor occupies 25% or more of the total tumor volume (Bernard et al. 2013; Carroll et al. 2014; Gallardo and Prat 2009; Kaku et al. 1992). In these areas, epithelial elements are absent, and the mesenchymal component is typically of high grade, with increased cellularity and mitotic activity and greater nuclear atypia (Fig. 82) compared with the background adenosarcoma, although occasionally the grade is the same as that of the background (Clement 1989). The sarcoma can be stromal sarcoma, fibrosarcoma, or leiomyosarcoma, or a mixture of elements. Heterologous elements, particularly rhabdomyosarcoma (Fig. 83), may occur in and be limited to the zone of sarcomatous overgrowth. The zones of pure sarcomatous growth can be present in or constitute the entire myoinvasive component of an adenosarcoma. Lymphovascular invasion, which is rare in adenosarcoma, is most often found in zones of sarcomatous overgrowth. Heterologous mesenchymal elements are present in 20–25% of adenosarcomas. Striated muscle, which typically resembles embryonal rhabdomyosarcoma, is the most common heterologous element, but cartilage, fat, and other

607

elements are occasionally observed. Rhabdomyosarcoma is characterized by the presence of round-to-spindled tumor cells with atypical hyperchromatic nuclei and variable amounts of eosinophilic cytoplasm. Cytoplasmic cross striations can occasionally be identified in rhabdomyoblasts on H&E-stained slides, but immunohistochemical staining for desmin and myogenin is now more widely used to identify rhabdomyoblastic differentiation. Rhabdomyosarcoma can be present in adenosarcomas with and without sarcomatous overgrowth, although it is more commonly present in tumors with overgrowth. In one study, the presence of rhabdomyosarcoma was associated with myoinvasion and lower overall survival (Mentrikoski et al. 2015).

Immunohistochemistry and Molecular Pathology The epithelial component of adenosarcoma is keratin-positive and usually stains for ER and PR. The mesenchymal component often resembles endometrial stromal sarcoma, so it is not surprising that adenosarcoma and endometrial stromal sarcoma share many immunophenotypic features. The mesenchymal cells in adenosarcoma typically show cytoplasmic staining for CD10 and nuclear staining for ER and PR and for WT-1 (Amant et al. 2004; Soslow et al. 2008). Staining is often most conspicuous in the periglandular stromal cuffs where the cell density is greatest. Staining for CD10 and hormone receptors is often weaker or lost in areas of high-grade sarcomatous overgrowth. Increased levels of nuclear staining for the proliferation marker Ki-67 (MIB-1) are typically present in hypercellular periglandular zones and in areas of sarcomatous stromal overgrowth (Gallardo and Prat 2009). Distinctive periglandular cuffs of Ki-67-positive mesenchymal cells can be helpful in the diagnosis of adenocarcinoma; these are not seen in some entities in the differential diagnosis such as endometrial polyps and APA (Aggarwal et al. 2012). Aberrant staining (diffuse strong positive staining in >80% of tumor cell nuclei or complete loss of staining) for p53 is sometimes noted particularly in cases of

608

high-grade adenosarcoma which are most likely to have p53 mutations (Hodgson et al. 2017). Focal staining for keratin, sometimes with a dot-like pattern, is occasionally noted in the mesenchymal component, and patchy weak staining for smooth muscle actin and/or desmin is present in many adenosarcomas. Areas of smooth muscle and rhabdomyosarcomatous differentiation show strong positive cytoplasmic staining for desmin, and rhabdomyosarcoma shows positive nuclear staining for myogenin. A number of cytogenetic and molecular studies on adenosarcoma have been reported in recent years. In general, these have found alterations in adenosarcomas, but abnormalities have only been identified in around half or fewer of the cases studied. In one study, cytogenetic abnormalities were identified in 45% of adenosarcomas. Aneuploidy with many translocations was observed in two cases, and less complex abnormalities were found in seven, including chromosome 8 alterations such as rearrangements of 8q13 or extra copies of chromosome 8 (Howitt et al. 2016). Molecular studies have revealed occasional mutations in genes such as ATRX, FGFR2, KMT2C, and DICER1. Some studies have found TP53 mutations to be infrequent (Howitt et al. 2015b), but one focused on high-grade adenosarcomas found them to be frequent and accompanied by aberrant immunohistochemical staining (Hodgson et al. 2017). In general, copy number variations have been more prominent than specific mutations and have included amplifications of MDM2, CDK4, HMGA2, and TERT, among others (Hodgson et al. 2017; Howitt et al. 2015b; Lee et al. 2016; Piscuoglio et al. 2016).

Differential Diagnosis The differential diagnosis includes benign entities such as endometrial and endocervical polyps and adenofibroma, as well as various malignant tumors, including endometrial stromal sarcomas, other uterine sarcomas, and, in young patients with cervical tumors, botryoid rhabdomyosarcoma. Many patients with adenosarcoma have a history of prior removal of “polyps.” In general, polyps are smaller than adenosarcomas. Microscopically, the stroma of benign polyps tends to

E. Oliva et al.

be fibrous, with few if any mitotic figures. Polyps often have conspicuous central blood vessels. Problems arise when large polyps have more cellular zones with endometrial-type stroma, where mitotic activity can overlap with the lower end of the range seen in adenosarcoma (Hattab et al. 1999). However, polyps lack the characteristic architecture of an adenosarcoma, do not have periglandular stromal hypercellularity, usually lack nuclear atypia, and do not display mitotic activity at the level commonly present in adenosarcoma (>4 MF/10 HPF). Rare atypical polyps exhibit some features of an adenosarcoma, such as abnormal architecture, increased periglandular cellularity, or increased mitotic activity, in the stromal cells, but in polyps the features are focal or incompletely developed, the polyps are small (10% of the tumor) of epithelial-like structures that appear similar to an ovarian sex cord–stromal tumor. Tumors of this type are commonly referred to as endometrial stromal tumors with sex cord-like elements, or ESTSCLE, and are discussed in the section of this chapter that deals with endometrial stromal tumors. The second variant, the type II tumors, consists predominantly or exclusively of sex cordlike elements, and tumors of this type are referred to as UTROSCT. Tumors of the latter type are discussed in this section. Clinical Features and Gross Findings Uterine tumors with sex cord-like elements occur in middle-aged women; the average age is around 50 (Blake et al. 2014). The main symptom is abnormal bleeding or pelvic pain. Most patients have an enlarged uterus or a palpable uterine mass. UTROSCTs are intramural or submucosal nodules surrounded by myometrium or polypoid tumors that grow into the endometrial cavity. They are yellow or tan and have a circumscribed or slightly irregular periphery. The average diameter is 6–7 cm. Microscopic Findings Microscopically, most are circumscribed, but examples with infiltrative margins and, rarely, vascular invasion have been reported. The tumor cells form plexiform cords, trabeculae, and nests, and may line well-formed tubules with lumens (Fig. 84). Glomeruloid formations or tubules with a retiform appearance are occasionally present. In some tumors, retiform tubules dominate

Fig. 84 UTROSCT. The low columnar tumor cells grow in a tubular pattern with scanty stroma

Fig. 85 UTROSCT. This tumor contains tubules lined by low columnar cells and nests of polygonal cells with abundant foamy cytoplasm

the histologic picture; designation of these as retiform RUTROSCT has been proposed (Nogales et al. 2009). The tumor cells have uniform small bland nuclei with inconspicuous nucleoli, and mitotic figures are rare, with 2 MF/10 HPF regarded as “significant mitotic activity” (Moore and McCluggage 2017). The cytoplasm varies from scant to moderate and is typically eosinophilic, although occasional tumors contain cells with abundant pale foamy cytoplasm. The cells are spindled, cuboidal, or columnar in shape. Columnar sertoliform cells, polygonal cells with

10

Mesenchymal Tumors of the Uterus

611

eosinophilic or foamy cytoplasm (Fig. 85), or cells resembling granulosa cells are present in some tumors. The stroma ranges from endometrial-like to hyaline or fibrous and smooth muscle is present in some UTROSCT. Stroma accounts for less than 50% of the tumor and is often scanty. The histogenesis of UTROSCT is unclear; an origin from endometrial stroma or uncommitted cells in the uterus has been proposed. Immunohistochemistry and Molecular Pathology Diverse immunohistochemical results have been described in tumors of this type, but the sex cordlike structures are usually immunoreactive for vimentin; cytokeratin; sex cord markers including calretinin (Fig. 86), inhibin (Fig. 87), CD 99, MelanA, CD56, and WT-1; and, often, smooth muscle actin or desmin (de Leval et al. 2010; Hurrell and McCluggage 2007; Irving et al. 2006). Variable staining has been reported for FOXL2, including strong nuclear staining in 1 of 15 tumors and weak or moderate staining in 5 additional tumors in one study (Chiang et al. 2015) and strong positive staining in 2 of 19 tumors as well as weak or moderate staining in 8 additional tumors in another (Croce et al. 2016). Importantly, however, endometrial stromal cells typically show weak to moderate nuclear staining for FOXL2. Steroidogenic factor-1 (SF-1) has also been shown to stain UTROSCT, being strongly positive in 1 of 19 tumors and showing weak or moderate staining in another 10, with no staining in endometrial stromal cells (Croce et al. 2016). Similar findings were reported in a smaller number of cases in another series, where it was noted that none of the tumors likely to mimic a UTROSCT showed staining for SF-1 (Stewart et al. 2016b). Immunostains for EMA have been reported as negative in most tumors, but weak to moderate staining was reported in one study of four cases (Hurrell and McCluggage 2007). Positive staining for ER and PR is often present. Positive staining for two or more markers of sex cord differentiation is seen in most UTROSCTs, with calretinin being the marker most likely to be positive (Irving et al. 2006; Pradhan and Mohanty 2013).

Fig. 86 UTROSCT. Sex cord-like tubules show strong positive staining for calretinin in this UTROSCT

Fig. 87 UTROSCT. Cord-like arrangements of cells, many with foamy cytoplasm, show modest but definite staining for inhibin in this UTROSCT

Molecular studies indicate that these tumors do not harbor JAZF1-SUZ12 gene fusions, nor do they show PHF1 rearrangements, indicating that they are unlikely to be endometrial stromal neoplasms (Staats et al. 2009). Neither FOXL2 nor DICER1 mutations have been identified in any UTROSCT tested for these mutations, though some of the tested tumors showed limited immunohistochemical staining for FOXL2 (Chiang et al. 2015; Croce et al. 2016). Differential Diagnosis The differential diagnosis includes an ESTSCLE, adenosarcoma, endometrioid adenocarcinoma with

612

a sex cord-like pattern, epithelioid smooth muscle tumor, and, less likely, a carcinosarcoma. ESTSCLE almost invariably contain recognizable areas of typical endometrial stromal tumor, and they are more likely than a UTROSCT to invade the myometrium or grow into blood vessels. The endometrial stromal areas usually show strong positive staining for CD10, ESTSCLE tends to show less staining for sex cord-stromal markers than UTROSCT, and FISH or molecular testing may show rearrangements of JAZF1 or PHF1, which are not present in UTROSCTs. As discussed above, adenosarcoma is a biphasic neoplasm in which benign epithelium is admixed with malignant mesenchyme. Differentiating between a UTROSCT and an adenosarcoma is usually straightforward, but some adenosarcomas contain sex cord-like elements. Rarely, these sex cord-like elements are so extensive that they dominate the histologic appearance of the tumor (Stolnicu et al. 2016), and such a neoplasm could be difficult to differentiate from a UTROSCT. The location of the tumor helps with this differential diagnosis as adenosarcoma generally originates in the endometrium and grows into the endometrial cavity, while UTROSCT is typically located within the myometrium. Nevertheless, there is some overlap; to make the correct diagnosis, adequate sampling of the specimen is necessary to identify areas of the tumor that display the characteristic morphology of an adenosarcoma. Endometrioid adenocarcinomas is a sex cordlike growth pattern almost always contain gland forming elements, would be negative for sex cordassociated markers and tend to express EMA, unlike UTROSCT. Epithelioid smooth muscle tumors may have a corded and trabecular architecture that superficially resembles sex cords. In contrast to UTROSCT, these tumors tend to stain differently smooth muscle marker while they lack expansion of sex cord-associated markers. Finally, a carcinosarcoma could potentially enter the differential diagnosis, but the immunophenotype together with the absence of high-grade carcinomatous and sarcomatous components differentiates these rare tumors from a carcinosarcoma.

E. Oliva et al.

Clinical Behavior and Treatment The clinical behavior of these tumors is difficult to assess because in some studies a clear distinction has not been made between ESTSCLE and UTROSCT. Where the distinction has been made, UTROSCT have typically been reported to generally have a benign clinical evolution, although rare examples have exhibited more aggressive behavior, extending beyond the uterus or metastasizing. In a recent report of 34 cases, including 32 drawn from a pathology consultation practice, it was found that 8 patients (23.5%) developed metastases and 3 (8.8%) died (Moore and McCluggage 2017). Tumor necrosis and significant mitotic activity (2 MF/10 HPF) were the main factors associated with an adverse outcome. UTROSCT appears to have more malignant potential than has been appreciated up to now. Nevertheless, the literature contains reports of cases in which conservative uterus sparing surgery was successfully used in a young woman to conserve fertility (Hillard et al. 2004).

IMT IMT is an uncommon uterine spindle cell tumor that typically has a prominent myxoid stroma that contains variable numbers of chronic inflammatory cells. It was initially considered to be an inflammatory pseudotumor (Gilks et al. 1987), but the documentation of the clonal nature of some IMT and the demonstration of chromosomal rearrangements involving the ALK locus indicate that these are best viewed as neoplasms. Clinical Features and Gross Findings IMT occurs in all age groups, from children to postmenopausal women. The average patient age is around 40. The presentation is with abnormal bleeding or symptoms related to the presence of a mass, such as pain or pressure, but some patients have constitutional symptoms including fever, weight loss, and fatigue. Occasionally, the tumor is an incidental finding at cesarean section or at surgery performed for another condition.

10

Mesenchymal Tumors of the Uterus

Grossly, the tumors range up to 12 cm in maximum diameter. The average diameter is 5–7 cm. They are firm or soft, and the cut surfaces are tan or white and often described as mucoid or gelatinous. Microscopic Findings Microscopically, the myofibroblastic tumor cells are spindle shaped, stellate, or epithelioid and have pale eosinophilic cytoplasm (Fig. 88). Their nuclei are granular or vesicular, and they may have prominent nucleoli. Nuclear atypia is mild or moderate in most cases, but marked atypia is seen in some tumors. Mitotic activity is variable and mostly ranges from 2 to 5 MF/10 HPF. Occasional tumors exhibit greater mitotic activity, sometimes exceeding 10 MF/10 HPF. Atypical mitotic figures are generally not present. A lymphoplasmacytic infiltrate is invariably present in these tumors. It ranges from mild and visible only at higher magnification to marked and diffuse, visible at low magnification. Three basic histologic patterns have been described: a myxoid pattern with lymphocytes and plasma cells scattered among the tumor cells, a compact cellular pattern in which the spindle cells are arranged in fascicles mimicking a smooth muscle tumor, and a hyalinized or collagenous pattern (Rabban et al. 2005). Myxoid changes and inflammatory infiltrates can be present only focally. Mixtures of the three main patterns are common in individual

Fig. 88 IMT. The tumor is a histologically low-grade spindle cell proliferation of largely mitotically inactive spindle cells set in a myxoid and inflammatory background. (Courtesy of Joseph Rabban, M.D.)

613

tumors. An infiltrative tumor border is almost always present. Infiltrative growth can take the form of finger-like projections of tumor cells into the surrounding myometrium, an irregular zigzag or sawtooth tumor periphery, or clusters of cells or single cells that infiltrate the myometrium. Immunohistochemistry and Molecular Pathology Immunohistochemical stains for smooth muscle actin, desmin, and, usually, CD10 show moderate or marked staining in the tumor cell cytoplasm. Caldesmon is less likely to stain. Immunostains for ER and PR are generally but not invariably positive. IMT shows cytoplasmic staining for ALK in 90% or more of cases (Bennett et al. 2017a; Fuehrer et al. 2012; Parra-Herran et al. 2015; Rabban et al. 2005). The staining is variable in intensity and distribution but is often strong and diffuse. ALK staining is usually coarsely granular, although it can also be homogeneous and can show membrane or perinuclear accentuation. A low threshold for performing ALK immunostains appears warranted to avoid misclassifying IMTs as smooth muscle tumors; any uterine tumor with any degree of myxoid changes or any lymphoplasmacytic infiltration should be considered for staining (Pickett et al. 2017). FISH testing reveals ALK rearrangements in at least 70–80% of tumors (Parra-Herran et al. 2015), and various fusion partners have been detected by molecular analysis (Bennett et al. 2017a; Haimes et al. 2017). Fusion partners that have been detected include THBS1, IGFBP5, DES, FN1, SEC31, and TIMP3; THBS1 and IGFBP5 appear to be the most common fusion partners. IGFBP5, DES, and FN1 are located on the same chromosome as ALK and are thought to arise via chromosomal inversions. FISH analysis consequently may not reveal ALK rearrangements when the fusion partner is one of these genes. A negative FISH result should therefore not exclude a diagnosis of IMT if the histologic appearance of the tumor suggests the diagnosis and immunohistochemistry for ALK is positive (Haimes et al. 2017). ROS1 rearrangements have been documented in extrauterine IMT, but to date none have been reported in uterine IMT.

614

Differential Diagnosis Cases of IMT have undoubtedly been misdiagnosed as other tumor types in the past. Tumors with which IMT is easily confused are smooth muscle tumors, including leiomyomas, leiomyosarcomas and STUMP (Pickett et al. 2017). Myxoid leiomyosarcoma is a particular problem, along with other myxoid mesenchymal tumors of the uterus such as myxoid variants of endometrial stromal sarcoma and, occasionally, solitary fibrous tumors (SFT) (Busca and ParraHerran 2017). Myxoid leiomyosarcoma is an uncommon variant of leiomyosarcoma that is characterized by abundant myxoid stroma that is present in at least 50% of the tumor. The degree of nuclear atypia and mitotic activity varies, but some myxoid leiomyosarcomas display less nuclear atypia and mitotic activity than a conventional leiomyosarcoma, and the malignant nature of the tumor is recognized by the combination of the myxoid nature of the tumor and invasive growth into the surrounding myometrium. The morphologic appearance of myxoid leiomyosarcoma overlaps with that of IMT; performing immunohistochemistry, and possibly FISH testing, for ALK rearrangement is essential to differentiate between these tumor types. Immunohistochemical staining for p53 and p16 may also be helpful, as mutations in TP53 and CDKN2A occur in about 50% of myxoid leiomyosarcomas, and result in aberrant staining patterns for p53 and complete loss of staining for p16 (Schaefer et al. 2017). Diffuse strong staining for p16 in the absence of any mutation is also seen in some myxoid leiomyosarcomas. In contrast, abnormalities in p53 and p16 staining are uncommon in IMT. Thus, immunohistochemical staining for ALK, p53, and p16 can assist with the differential diagnosis between an IMT and a myxoid leiomyosarcoma. Benign myxoid leiomyomas are rare and pose less of a diagnostic problem; absence of staining for ALK provides support for a diagnosis of a myxoid leiomyoma rather than an IMT. Both low- and high-grade variants of endometrial stromal sarcoma (ESS) can have myxoid

E. Oliva et al.

features. Low-grade variants of endometrial stromal sarcoma have abundant myxoid stroma, but they show the same patterns of myometrial and vascular invasion that are seen in more typical examples of low-grade ESS, a similar capillary vascular pattern is seen, and the tumor cells resemble endometrial stromal cells or fibroblasts (Oliva et al. 1999). Positive staining for CD10 is typically present in ESS, but this is frequently observed in IMT as well, so staining for ALK is required to differentiate these two tumor types. High-grade endometrial stromal sarcomas associated with ZC3H7B-BCOR gene fusions or ITD of BCOR also have a myxoid appearance and can mimic an IMT or myxoid leiomyosarcoma (Hoang et al. 2017; Lewis et al. 2018; Marino-Enriquez et al. 2018). These have permeative, tongue-like or pushing invasion of the myometrium, variable nuclear atypia, usually frequent mitotic figures, and variable amounts of myxoid stroma. They show positive staining for cyclin D1 and may show staining for CD10 and BCOR; they can be differentiated from an IMT by a lack of staining for ALK. Staining for BCOR does not appear to be a completely reliable way to identify these tumors, since many of them show weak, focal, or absent staining, and BCOR also stains high-grade endometrial stromal sarcomas with YWHAE-NUTM2, indicating that molecular testing is required to confirm the diagnosis of a BCOR-related endometrial stromal sarcoma. Finally, SFT can have myxoid stroma and occasionally involve the uterus. SFT are cellular spindle cell tumors with a patternless arrangement and prominent blood vessels. They show positive nuclear staining for STAT6 and cytoplasmic staining for CD34. STAT 6 and CD34 tend to be negative in IMT, although in one study a single case of IMT showed positive staining (Yang et al. 2018). SFT is ALK negative, so the morphologic appearance and immunophenotype generally readily differentiate these two types of tumor.

10

Mesenchymal Tumors of the Uterus

Clinical Behavior and Treatment Most IMTs are clinically benign, but a significant minority, 20–30%, show extrauterine spread at diagnosis or recur, usually within the pelvis or abdomen, and tumors with extrauterine spread can behave aggressively and cause the death of the patient. Features that are associated with aggressive behavior include older patient age, large tumor size, lymphovascular space invasion, tumor cell necrosis, and high mitotic activity, frequently in excess of 10 MF/10 HPF (Bennett et al. 2017a). The ALK rearrangements present in these tumors theoretically make them amenable to targeted therapy, and responses to crizotinib have been reported (Pickett et al. 2017; Subbiah et al. 2015).

Heterologous and Homologous Sarcomas Other than Leiomyosarcoma and Endometrial Stromal Sarcoma The tumors in this category are high-grade sarcomas that often resemble the mesenchymal component of a carcinosarcoma. Most pleomorphic homologous sarcomas arise in the endometrium and consist of round or spindled cells with variable amounts of cytoplasm and pleomorphic atypical nuclei. These are a type of undifferentiated endometrial sarcoma and are discussed in the section on undifferentiated uterine sarcoma. Some may result from sarcomatous stromal overgrowth in an adenosarcoma or carcinosarcoma or by dedifferentiation of a low-grade endometrial stromal sarcoma. Although most undifferentiated sarcomas are of endometrial origin, a few appear to arise in the myometrium, either from nonspecific mesenchymal elements or by dedifferentiation of a leiomyosarcoma fibromatous sarcoma has recently been described with NTRK fusions (Chiang et al. 2018). Pure heterologous sarcomas occasionally arise in the uterus. Some are assumed to represent complete heterologous stromal overgrowth in an adenosarcoma or carcinosarcoma. Rhabdomyosarcoma and angiosarcoma are the most common heterologous uterine sarcomas, but

615

chondrosarcoma, osteosarcoma, liposarcoma, and tumors containing mixtures of heterologous elements also occur (Fadare 2011). Histologically benign heterotopic bone, cartilage, and fat are occasionally found in the uterus, and rare benign tumors contain one or more of these elements (Roth and Taylor 1966). They should not be mistaken for heterologous sarcomas or carcinosarcomas, in which the mesenchymal elements are histologically malignant.

Rhabdomyosarcoma Rhabdomyosarcoma is a malignant neoplasm that displays skeletal muscle differentiation. It is the most common soft tissue tumor in children and approximately 20% of pediatric rhabdomyosarcoma originate in the genital tract. In adults, rhabdomyosarcoma of the female genital tract is a rare tumor that can involve either the cervix or the body of the uterus. Rhabdomyosarcoma of the uterus and cervix is categorized into three major variants with significant differences in clinical behavior and prognosis. The most common and generally most favorable variant is embryonal rhabdomyosarcoma, which includes botryoid and anaplastic histologic subtypes. The other two types of rhabdomyosarcoma, alveolar and pleomorphic rhabdomyosarcomas, are less common and usually have a significantly less favorable clinical outcome.

Clinical Features In adults, rhabdomyosarcomas can occur at any age, including young to middle-aged women with predominantly cervical tumors and older women in the fifth to seventh decade, mainly with tumors of the uterine corpus. The most common presenting symptom is abnormal vaginal bleeding. The majority of women, approximately 75%, present with local or regional disease. The primary site is the cervix in approximately one-half of women, making it the most common site of origin of genital tract rhabdomyosarcoma in adults; only 20% of tumors are uterine.

616

Children and young adults with cervical or uterine rhabdomyosarcomas present with vaginal bleeding or with a polypoid tumor that protrudes from the vagina. Children with cervical rhabdomyosarcoma have an average age of 12 years (Dehner et al. 2012), but identical tumors occur in young and middle-aged women (Daya and Scully 1988; Li et al. 2013). Cervical rhabdomyosarcoma occurs in patients with the DICER1 syndrome in up to 20% of cases (Stewart et al. 2016a). Other tumors that occur in patients with the syndrome include pleuropulmonary blastoma, cystic nephroma, multinodular goiter, and Sertoli-Leydig cell tumor of the ovary. A diagnosis of cervical rhabdomyosarcoma in a patient of any age, but especially in young women, should prompt investigation of the patient and her family to determine whether the patient has the syndrome, which is caused by a germline mutation of the DICER1 gene.

Gross Findings Cervical rhabdomyosarcomas are mainly polypoid gray-tan or red tumors 1.5–5 cm in maximum diameter (Dehner et al. 2012). Botryoid rhabdomyosarcoma, the most common type in the cervix, is often described as resembling a cluster of grapes. Uterine rhabdomyosarcoma tends to be a polypoid endometrial tumor that grows into the uterine cavity and invades the myometrium. Some rhabdomyosarcomas are nodular tumors located entirely within the myometrium. In one recent series, the average tumor size at presentation was 11.7 cm (Pinto et al. 2018). Microscopic Findings Rhabdomyosarcomas of the cervix are more common than those of the uterine corpus, and most are embryonal rhabdomyosarcomas of the botryoid subtype. Botryoid rhabdomyosarcomas are polypoid tumors that have a densely cellular zone of primitive cells beneath the surface epithelium (the “cambium layer”) (Fig. 89) (Daya and Scully 1988; Dehner et al. 2012; Li et al. 2013). The substance of the polyps is generally myxoid or edematous, with varying cellularity. It often contains hyperchromatic cellular nodules and zones of hemorrhage. The tumor cells range from undifferentiated small round cells with hyperchromatic nuclei and scanty

E. Oliva et al.

Fig. 89 Botryoid rhabdomyosarcoma. These tumors are typically composed of grape-like tumors with a paucicellular matrix that appears myxoid or edematous. Condensation of primitive cells beneath the overlying epithelium and alongside any entrapped epithelium (so-called cambium layer) is common. Phyllodes-like architecture and intraglandular stromal papillae, seen in adenosarcoma, are lacking

cytoplasm (“small round blue cells”) to strapshaped cells with eosinophilic cytoplasm. Cells with cross striations are typically difficult to identify. Foci of immature appearing cartilage are admixed with the rhabdomyoblasts in a significant minority of cases. Non-polypoid and infiltrative portions of the tumors are usually histologically indistinguishable from embryonal rhabdomyosarcomas of the usual (non-botryoid) type, with cellular zones that alternate with paucicellular zones with a myxoid or edematous matrix. An unusual botryoid rhabdomyosarcoma that contained areas with a more pleomorphic pattern has been reported (Houghton and McCluggage 2007). Rhabdomyosarcomas of the uterine body in older patients are often high-grade pleomorphic sarcomas composed of round, polygonal, or spindle-shaped cells admixed with rhabdomyoblasts (Ordi et al. 1997; Pinto et al. 2018). The rhabdomyoblasts range from round cells with prominent perinuclear rims of eosinophilic cytoplasm to spindle- or tadpole-shaped cells with fibrillar eosinophilic cytoplasm. In the female genital tract, alveolar rhabdomyosarcoma occurs most often in the vulva but also occurs in the uterine body and cervix (Fukunaga 2011; Pinto et al. 2018). The tumor cells tend to be larger than those in embryonal rhabdomyosarcoma. In some tumors round or irregular spaces are surrounded by tumor cells,

10

Mesenchymal Tumors of the Uterus

resulting in an alveolar appearance. Tumor cells often seem to cling to the conspicuous fibrovascular septae that traverse sheets of tumor cells. In other alveolar rhabdomyosarcomas, tumor cells form a solid mass with no alveolar spaces. Regardless of the growth pattern, the tumor cells have a distinctive cytologic appearance with round nuclei, larger than those in embryonal rhabdomyosarcoma, and scanty cytoplasm. Multinucleated tumor cells are commonly present. Round and spindled cells with brightly eosinophilic cytoplasm are present in most cases and are important clues to the diagnosis. Tumor cells with cross striations can generally be identified.

Immunohistochemistry and Molecular Pathology The typical immunophenotype includes expression of desmin, muscle-specific actin, myogenin, and MyoD1 (Fig. 90). Staining for these markers is present in greater than 90% of rhabdomyosarcomas (Dehner et al. 2012; Li et al. 2013; Pinto et al. 2018). Desmin and musclespecific actin are not specific for rhabdomyosarcoma, as positive staining is also found in tissues demonstrating smooth muscle and myofibroblastic differentiation. Myogenin and Myo-D1 are nuclear regulatory proteins that are expressed early in skeletal muscle differentiation. Myogenin is the more widely used of the two. In general, the expression of these markers is negatively correlated with differentiation; nuclear staining is widespread in tumors with

Fig. 90 Botryoid rhabdomyosarcoma. Myogenin expression in tumor cell nuclei confirms rhabdomyoblastic differentiation

617

numerous differentiated rhabdomyoblasts, but only scattered myogenin positive cells are found in rhabdomyosarcomas composed predominantly of undifferentiated cells with few histologically recognizable rhabdomyoblasts. PAX7 has been proposed as an additional useful marker of rhabdomyosarcoma; it appears most likely to stain embryonal and pleomorphic rhabdomyosarcomas, including some cases that fail to stain for myogenin (Charville et al. 2016). Expression of myogenin and Myo-D1 is common to all tumors demonstrating skeletal muscle differentiation, which means that tumors other than pure rhabdomyosarcoma (e.g., adenosarcoma or carcinosarcoma containing rhabdomyoblasts) must be excluded before using myogenin or Myo-D1 immunoreactivity to confirm a diagnosis of rhabdomyosarcoma. Rhabdomyosarcoma may rarely express markers that are more commonly expressed in its histologic mimics, including staining for CD99, cytokeratin, S100 or WT1. Occasional tumors show co-expression of neuroendocrine markers such as chromogranin and synaptophysin along with desmin and myogenin. Immunohistochemistry may help with the subclassification of rhabdomyosarcoma. Staining for myogenin is likely to be strong and diffuse in alveolar rhabdomyosarcoma; staining is weaker and more focal in embryonal rhabdomyosarcoma (Heerema-McKenney et al. 2008; Morotti et al. 2006). Positive nuclear staining for PAX-5 is reported to be present in about two-thirds of alveolar rhabdomyosarcomas, but staining tends to be absent in embryonal rhabdomyosarcoma (Sullivan et al. 2009). In a limited number of cases with genetic correlation, staining occurred only in tumors that had one of the characteristic translocations, but there was no correlation with a specific translocation. A majority of alveolar rhabdomyosarcomas exhibit a clonal chromosomal translocation, either t(2;13)(q35;q14), resulting in a PAX3-FOX01 fusion, or t(1;13) (p36;q14), resulting in a PAX7FOX01 fusion. These translocations can be identified using FISH and demonstration of a translocation can support a diagnosis of alveolar rhabdomyosarcoma (Rivasi et al. 2008). There

618

appears to be a survival difference between patients whose tumors have the PAX3-FOX01 fusion and those whose tumors have the PAX7FOX01 fusion, with the former having a significantly worse prognosis, at least when metastatic disease develops (Sorensen et al. 2002).

Differential Diagnosis Myxoid leiomyosarcoma may resemble embryonal rhabdomyosarcoma and pleomorphic leiomyosarcoma can appear similar to pleomorphic rhabdomyosarcoma. Embryonal rhabdomyosarcoma, when spindled and growing in fascicles, frequently has a subtle moth-eaten appearance that results from a heterogeneous admixture of cells, some containing densely eosinophilic cytoplasm and others clear or amphophilic cytoplasm (Fig. 91). Round rhabdomyoblasts with bright red cytoplasm are often haphazardly intermixed. The low power appearance of leiomyosarcoma, in contrast, is generally more uniform. Most embryonal rhabdomyosarcomas with an infiltrative, spindle cell appearance underlie a botryoid tumor and/or are clearly epitheliotropic. Leiomyosarcomas, in contrast, are usually more deeply seated lesions. Immunohistochemistry is useful for distinguishing rhabdomyosarcoma from leiomyosarcoma. Diffuse desmin expression is present in both, but only rhabdomyosarcoma shows staining for myogenin.

Fig. 91 Rhabdomyosarcomas with spindle cell morphology may mimic leiomyosarcoma. As compared to leiomyosarcomas, rhabdomyosarcomas more frequently lack an eosinophilic appearance at low power and instead demonstrate a subtle moth-eaten look owing to the admixture of round and spindled rhabdomyoblasts

E. Oliva et al.

Caldesmon can be positive in leiomyosarcoma but is generally negative in rhabdomyosarcoma. Adenosarcoma and embryonal rhabdomyosarcoma both show polypoid growth and stromal condensation beneath epithelium, but botryoid rhabdomyosarcoma more typically contains conspicuous myxoid stroma and has a sprinkling of small cellular aggregates of dark blue, primitive and mitotically active cells in a paucicellular background. These features can sometimes be appreciated macroscopically, such that gross inspection of a glass slide can suggest the correct diagnosis. In contrast to adenosarcoma, rhabdomyosarcoma does not exhibit phyllodes-like growth or intraglandular stromal papillae. Adenosarcomas with stromal overgrowth frequently contain rhabdomyoblastic foci and these are the ones most likely to be misdiagnosed as a rhabdomyosarcoma. Stromal-predominant carcinosarcoma is excluded by a careful search for a malignant epithelial component. The presence of any type of carcinoma indicates that the tumor is a carcinosarcoma. Some genital pleomorphic rhabdomyosarcomas represent carcinosarcomas in which the epithelial component is overgrown by the sarcomatous mesenchymal component. Pleomorphic undifferentiated sarcoma can resemble rhabdomyosarcoma, as it is composed of mitotically active atypical round or spindle cells. However, rhabdomyoblasts with eosinophilic cytoplasm are not present, and staining for myoid markers such as desmin and myogenin is negative. Undifferentiated and high-grade stromal sarcomas can express CD10, but this marker may not be helpful in the differential diagnosis with pleomorphic rhabdomyosarcoma since some of those also express CD10 (Fadare et al. 2010). Undifferentiated carcinoma also enters the differential diagnosis as it consists of medium-sized round cells with scanty cytoplasm with no obvious glandular differentiation. Undifferentiated carcinoma may show areas of cellular cohesion, and the tumor cells show at least focal staining for keratin or EMA. Staining for markers of myoid differentiation is absent. Also, many undifferentiated carcinomas are associated with a component of endometrioid adenocarcinoma

10

Mesenchymal Tumors of the Uterus

somewhere in the tumor, a combination that is referred to as a dedifferentiated carcinoma.

Clinical Behavior and Treatment Most adult patients are treated surgically, with or without chemotherapy and radiation therapy. In a retrospective review of genital tract rhabdomyosarcoma in adults, the median time to progression was only 9 months and the median disease-specific survival was 21 months; the 5-year disease-specific survival was only 29% (Ferguson et al. 2007). Neither age nor stage correlated with survival. These patients were not offered pediatric therapeutic protocols, which perhaps resulted in unanticipated poor survivals. In this study, embryonal rhabdomyosarcomas appeared to have better survivals compared to other rhabdomyosarcoma subtypes. Other series reporting predominantly pleomorphic rhabdomyosarcomas in older women have also documented poor survivals (Fadare et al. 2010; Ordi et al. 1997; Pinto et al. 2018). Women with cervical rhabdomyosarcoma have a longer time to progression than women with disease at other gynecologic sites, and women with an embryonal rhabdomyosarcoma have improved progression free survivals compared to those with nonembryonal types of rhabdomyosarcoma (Kriseman et al. 2012; Nasioudis et al. 2017a). In one recent series of botryoid embryonal rhabdomyosarcomas in women having an average age of 44 years, 5 of 7 patients were alive with no evidence of disease (Li et al. 2013). Children, teenagers, and some young adults with cervical botryoid rhabdomyosarcomas are generally treated by limited excisions such as polypectomies or LEEP excisions to establish the diagnosis, followed by chemotherapy and in some instances radiation. Primary surgical management using fertility sparing techniques is also an option in young women (Bouchard-Fortier et al. 2016). The poor survival statistics for adults with genital tract rhabdomyosarcoma contrasts with the favorable survival rates in children and young women with cervical embryonal rhabdomyosarcomas, most of who appear to be cured by limited surgery and chemotherapy (Daya and Scully 1988; Dehner et al. 2012). It is uncertain

619

whether contrasting clinical outcomes are attributable to intrinsically different biological attributes and/or differences in therapy.

ASPS ASPS is uncommon in the female genital tract, but it occasionally occurs in the vagina, cervix, or uterus. Uterine tumors have been described in the endometrium, lower uterine segment, and myometrium (Kasashima et al. 2007; Nielsen et al. 1995; Radig et al. 1998; Schoolmeester et al. 2017). The average patient age is about 30 years, and most patients present because of abnormal bleeding. ASPS is composed of cells with abundant clear-to-eosinophilic cytoplasm that grow in solid nests or, when there is loss of cellular cohesion, an alveolar pattern (Fig. 92). The tumor cell cytoplasm is filled with granules and crystals that are periodic acid–Schiff (PAS)-positive and diastase resistant. Nuclei are usually round with prominent nucleoli. Mitotic activity is low and mitotic figures can be difficult to identify. The fibrovascular framework that supports the nests and alveoli can be conspicuous. The tumors tend to involve the cervical mucosa or endometrium. ASPS that arise in the soft tissues frequently show vascular invasion, but this is uncommon in gynecologic cases. ASPS are characterized by a chromosomal translocation, t(x;17)(p11;q25), in which the

Fig. 92 ASPS. The tumor is formed of nested aggregates of epithelioid tumor cells with granular or crystalline cytoplasm

620

TFE3 transcription factor gene on chromosome Xp11 is fused to ASPSCR1 (ASPL) on chromosome17q25 (Ladanyi et al. 2001). An immunohistochemical stain for the TFE3 protein, which is a nuclear antigen, can be used as a diagnostic adjunct to recognize tumors with this translocation (or other abnormalities involving TFE3) (Argani et al. 2001; Argani et al. 2003; Kasashima et al. 2007; Roma et al. 2005). TFE3 immunostains should only be interpreted as positive when staining is diffuse and strong in tumor cell nuclei. The presence of the translocation can be confirmed by FISH breakapart probe testing for TFE3, by dual color probe FISH testing for the ASPSCR1-TFE3 fusion, and by molecular testing to identify the fusion transcripts (Jabbour et al. 2014; Schoolmeester et al. 2017). ASPS are mainly negative for markers that stain smooth muscle tumors and PEComas, but rare cases have been reported to show strong but focal staining for HMB45 (Schoolmeester et al. 2017). The differential diagnosis of ASPS includes adenocarcinoma, epithelioid smooth muscle tumors, PEComa, metastatic melanoma, and a UTROSCT. Adenocarcinoma, epithelioid smooth muscle tumor, metastatic melanoma, and UTROSCT all have a different immunophenotype than ASPS, but the immunohistochemical features of PEComa and ASPS can sometimes be similar. This is because some PEComas express TFE3 (Folpe et al. 2005; Schoolmeester et al. 2015). Any appreciable desmin staining or staining with markers such as HMB-45, MITF, or Melan-A would support PEComa over ASPS. Gynecologic ASPS has a relatively good prognosis compared to soft tissue ASPS, but the number of cases and the length of follow-up reported are insufficient to draw definitive conclusions. In one series of 9 patients, 1 patient died of tumor and the other 8 were alive with no evidence of tumor 9 months to 17 years after diagnosis (Nielsen et al. 1995). In another series of 10 patients, 4 were alive with no evidence of tumor after short follow-up and the rest were recent cases or lost to follow-up (Schoolmeester et al. 2017). Recently, there has been interest in using MET inhibitors to

E. Oliva et al.

treat ASPS, since the fusion present in ASPS activates MET signaling, and in blocking the VEGF signaling pathway.

PNET PNET only rarely arises in the uterus. It can occur at any age, but most are found in postmenopausal women. The median age in the largest series was 58 years, and most patients had stage III or stage IV tumors (Euscher et al. 2008). In another large series the average age of women with uterine PNETs was 51 years (Chiang et al. 2017b). The usual clinical presentation is with abnormal vaginal bleeding. PNET is a soft, fleshy, gray or white polypoid mass that originates in the endometrium and invades the myometrium. Microscopically, PNET is composed of small cells with round to oval hyperchromatic nuclei and scanty cytoplasm (Fig. 93). Mitotic figures are usually numerous. Evidence of neuroectodermal differentiation includes the presence of an eosinophilic fibrillary background or the formation of rosettes or pseudo-rosettes. Immunostains are generally positive for one or more neural or neuroendocrine marker and staining

Fig. 93 PNET of the uterus. The tumor is composed of small cells with hyperchromatic nuclei, coarse chromatin, and scanty cytoplasm, arranged in nests and trabeculae. Rosettes or pseudorosettes are seen in some cases. This case was immunoreactive for CD56, chromogranin, and neurofilament

10

Mesenchymal Tumors of the Uterus

for keratin is generally absent. The most useful immunostains are keratin, which is generally negative excluding a carcinoma, and synaptophysin, glial fibrillary acidic protein (GFAP), and neurofilament, which are usually positive. Staining for other neuroectodermal markers such as chromogranin, neuron-specific enolase, and CD56 is more variable but occasionally one or more are positive. Many uterine PNET, especially those with a t(11;22), express CD99 in a diffuse, membranous pattern, as well as FLI-1. Positive staining for CD99 and FLI-1 is not proof that a tumor is a peripheral type PNET, since central type PNET without evidence of a translocation frequently show CD99 staining (Chiang et al. 2017b; Euscher et al. 2008). Uterine PNET appear to fall into two categories. Some have a chromosomal translocation, usually t(11;22) resulting in a fusion between the EWS and FLI1 genes. These demonstrate histologic, immunohistochemical, and biologic similarities to Ewing sarcoma (Blattner et al. 2007; Varghese et al. 2006). Other uterine PNET, including all tested uterine tumors in the two largest reported series, lacked evidence of an EWSR1 rearrangement and are thus more akin to embryonal tumors of the central nervous system, which were previously called central PNET (Chiang et al. 2017b; Euscher et al. 2008). Mixtures of PNET and other tumor types are occasionally seen. PNET has been reported in association with various types of sarcoma, with carcinosarcoma and adenosarcoma and with endometrioid adenocarcinoma (Quddus et al. 2009; Sinkre et al. 2000a). When associated with carcinosarcoma or adenosarcoma PNET is viewed by some as a form of heterologous differentiation. Too few patients have been studied to define the clinical behavior and most appropriate treatment for PNET of the uterus. Women with stage I neoplasms can be cured, but more advanced tumors are frequently fatal. In one series 7 of 13 patients with follow-up died of tumor and 6 were alive with no evidence of disease (Euscher et al. 2008), while in another series 2 of 4 patients with follow-up died (Chiang et al. 2017b).

621

Miscellaneous Mesenchymal Tumors and Conditions Adenomatoid Tumor Adenomatoid tumors are distinctive benign genital tract neoplasms of mesothelial origin that occur in both men and women (Nogales et al. 2002). In women they occur in the uterus, the fallopian tube, and the ovary.

Clinical Features Adenomatoid tumors typically occur in women of reproductive age; the median age is 42 years. There is no evidence that they impair fertility, and they are usually incidental findings in uteri removed for other causes. Adenomatoid tumors are reportedly found in about 1% of hysterectomy specimens, although in a systematic prospective study adenomatoid tumors were found in 5% of hysterectomies and 5% of uterus-preserving tumor excisions, suggesting that they may be more common than is generally appreciated (Nakayama et al. 2013). Adenomatoid tumors are typically thought to be small leiomyomas and, except for rare large cystic tumors, no specific symptoms have been attributed to them (Nogales et al. 2002). Adenomatoid tumors are benign. Pathologic Findings Adenomatoid tumors are usually located subserosally in the cornual myometrium. They are typically small, measuring 0.5–1 cm in diameter, but some are larger and giant, and cystic adenomatoid tumors have been reported. Adenomatoid tumors are round and rubbery and are often thought to be leiomyomas. The cut surfaces are gray or tan and may have a spongy appearance due to the presence of uniform small cysts. Microscopically, adenomatoid tumors tend to be circumscribed, although rare diffuse variants have been described. They consist of tubules and cords of varying size and shape that are lined by flat or cuboidal epithelial cells (Fig. 94). Strands of cytoplasm, so-called thread-like bridging strands, cross the lumens are distinctive finding that is

622

E. Oliva et al.

Adenomatoid tumors have recently been shown to be clonal proliferations based on nonrandom x-chromosome inactivation (Wang et al. 2016). Our group recently identified TRAF7 mutations in all tested examples of adenomatoid tumors of the male and female genital tracts, providing a genetic basis for the tumor (Goode et al. 2018).

Vascular Tumors Fig. 94 Adenomatoid tumor within hypertrophic myometrium. Variably sized tubules are lined by flattened or cuboidal mesothelial cells

almost invariably seen in adenomatoid tumors (Sangoi et al. 2009). Collagen, elastic tissue, and smooth muscle surround the epithelial elements. The smooth muscle may predominate such that the tumor appears at first glance to be a leiomyoma or lipoleiomyoma. The cuboidal epithelial cells have cytologically bland, eccentric, round nuclei and abundant pale cytoplasm. The cytoplasm is often vacuolated, sometimes to the extent that some tumor cells resemble signet-ring cells. The growth of the epithelial cells between smooth muscle bundles and the presence of signet-ring-like cells may raise the suspicion of metastatic adenocarcinoma. Nuclear atypia, however, is absent or minimal, mitotic figures are infrequent, and stains for mucin are negative. When the cells lining the tubules are flattened, an adenomatoid tumor may resemble a hemangioma or lymphangioma. However, the lumens do not contain blood, and immunostains for such vascular markers as factor VIII-related antigen and CD 31 are negative. Ultrastructural and immunohistochemical studies reveal that the epithelial cells in adenomatoid tumors have a mesothelial phenotype. Immunostains are positive for cytokeratin and vimentin and for such mesothelial cell-associated antigens as calretinin, WT1, and D2-40 (Sangoi et al. 2009). Stains for adenocarcinoma-associated antigens such as CD 15, CEA, B72.3, and Ber-EP4 are usually negative, and GATA-3 is negative in uterine adenomatoid tumors (Ronaghy et al. 2018), as is PAX8 (Wachter et al. 2011). Lymphoid aggregates are present in some adenomatoid tumors.

Hemangiomas of the uterus, like those at other sites, are composed of neoplastic vessels lined by flat or cuboidal endothelial cells. The endothelial cells have bland nuclei and mitotic figures are rare or, most typically, absent. Uterine hemangiomas can be subclassified as capillary, cavernous, or venous, depending on the appearance of the vessels (Lotgering et al. 1989; Weissman et al. 1993). Rare histologic types of hemangiomas can occur in the uterus, such as Kaposiform hemangioendothelioma and glomeruloid hemangiomas (Giner et al. 2012; Zhang et al. 2012). The subtypes do not differ clinically, except that rare variants like the glomeruloid hemangioma may be a marker for other disease states (polyneuropathy, organomegaly, endocrinopathy, monoclonal gammopathy, and skin changes (POEMS) syndrome). Capillary hexmangiomas of the cervix are the most common vascular tumors of the uterus. Most occur in women of reproductive age, and expression of ER and PR in the endothelial cells and stroma suggests that hormonal stimulation may play a part in their growth (Busca and ParraHerran 2016). The average diameter is about 2 cm. Women with cervical hemangiomas often experience abnormal bleeding and pain. Hemangiomas of the corpus are uncommon (Chou and Chang 2012), and they vary considerably in size. Large hemangiomas can extend through the full thickness of the myometrium and can result in severe bleeding that requires a hysterectomy. Arteriovenous malformations can occur in the uterus (Fleming et al. 1989; Majmudar et al. 1998). They are differentiated from venous hemangiomas by the presence of thick-walled vessels of both arterial and venous types. Histologic distinction between a hemangioma and a vascular malformation can be difficult but is not critical because their clinical features are similar.

10

Mesenchymal Tumors of the Uterus

Fig. 95 Angiosarcoma of the uterus. The growth is partly solid but vascular lumens are readily identified. The malignant cells are cuboidal or polygonal and have atypical hyperchromatic nuclei. Numerous mitotic figures were visible at higher magnification. A benign gland is surrounded by tumor (lower center)

A few examples of angiosarcoma of the uterus have been reported (Cardinale et al. 2008; Liu et al. 2016; Schammel and Tavassoli 1998). Angiosarcoma is a large, hemorrhagic, and often extensively necrotic tumor that grows in the myometrium. It consists of anastomosing vascular channels that are lined by atypical cuboidal or “tombstone”-shaped endothelial cells (Fig. 95). Many mitotic figures are usually present. Some high-grade angiosarcomas consist partly or completely of solid sheets of difficult-to-recognize epithelioid endothelial cells. When these cells predominate, the nature of the tumor can be determined by identifying characteristic foci of vascular growth, often at the periphery of the tumor, and by positive immunohistochemical stains for markers of vascular differentiation such as factor VIII-related antigen, CD 31, or ERG. Angiosarcoma extensively invades and replaces the myometrium and has a poor prognosis, with 10 of 15 reported cases dead of disease at a mean of 13 months (Roma et al. 2017). In one review, surgery followed by adjuvant radiotherapy seemed to provide the best treatment results (Kruse et al. 2014b).

Lymphoma Lymphomas involving the corpus and cervix are discussed in ▶ Chap. 21, “Hematologic

623

Neoplasms and Selected Tumorlike Lesions Involving the Female Reproductive Organs.” Accordingly, what follows is brief overview of these disorders. Lymphoma rarely occurs with initial signs or symptoms suggestive of a uterine tumor, but when it does, the cervix is involved more often than is the endometrium (Harris and Scully 1984; Nasioudis et al. 2017b). Most patients are older than 20 years and present with an abdominal or pelvic mass, abnormal vaginal bleeding, a vaginal discharge, or pelvic discomfort. Diffuse large cell lymphomas of B-cell type are most common (Frey et al. 2006; Kosari et al. 2005; Vang et al. 2000). An 80–90% survival rate has been reported for women with localized lymphomas of the uterus and vagina (Ahmad et al. 2014; Vang et al. 2000). The differential diagnosis includes a leiomyoma with a heavy lymphocytic infiltrate and an inflammatory lymphoma-like lesion (pseudolymphoma). Rare leiomyomas contain a heavy lymphocytic infiltrate; however, these are circumscribed tumors containing recognizable areas of residual smooth muscle tumor. Additionally, the lymphocytic infiltrate consists of a mixture of cell types (Botsis et al. 2005; Ferry et al. 1989). Inflammatory “pseudolymphomas” mainly involve the cervical or endometrial surface or are just beneath it, whereas lymphoma is larger and more deeply situated (Geyer et al. 2010; Ma et al. 2007; Young et al. 1985). These lesions consist of a polymorphous inflammatory infiltrate with a mixture of B and T cells and no immunoglobulin light chain restrictions. However, four of nine analyzed cases had clonal rearrangements of the immunoglobulin heavy chain. None of the patients had evidence of a lymphoma on staging or on follow-up, and the authors concluded that in this setting, the clonal immunoglobulin heavy chain rearrangement was insufficient evidence for a diagnosis of lymphoma (Geyer et al. 2010). Uterine involvement as a manifestation of leukemia is very rare (Garcia et al. 2006; Oliva et al. 1997). Inflammatory processes contain a heterogeneous population of lymphoid cells in contrast to the more monomorphic population seen in most lymphomas, and they are polyclonal. Intravascular lymphocytic accumulations have been described in association with severe chronic

624

cervicitis. Small and medium-sized non-atypical lymphocytes accumulate in lymphatic channels in these cases and express a mixture of T and B cell phenotypes indicative of a benign condition (Karpathiou et al. 2018). The differential diagnosis of lymphomas and leukemias also includes neoplastic entities, including small cell carcinoma, undifferentiated carcinoma, and IMT, which are discussed elsewhere in this text.

References Abeler VM, Nenodovic M (2011) Diagnostic immunohistochemistry in uterine sarcomas: a study of 397 cases. Int J Gynecol Pathol 30:236–243. https://doi.org/ 10.1097/PGP.0b013e318200caff Abeler VM, Royne O, Thoresen S, Danielsen HE, Nesland JM, Kristensen GB (2009) Uterine sarcomas in Norway. A histopathological and prognostic survey of a total population from 1970 to 2000 including 419 patients. Histopathology 54:355–364. https://doi. org/10.1111/j.1365-2559.2009.03231.x Abell MR (1971) Papillary adenofibroma of the uterine cervix. Am J Obstet Gynecol 110:990–993 Abrams J, Talcott J, Corson JM (1989) Pulmonary metastases in patients with low-grade endometrial stromal sarcoma. Clinicopathologic findings with immunohistochemical characterization. Am J Surg Pathol 13:133–140 Abu-Rustum NR, Curtin JP, Burt M, Jones WB (1997) Regression of uterine low-grade smooth-muscle tumors metastatic to the lung after oophorectomy. Obstet Gynecol 89:850–852 Abushahin N, Zhang T, Chiang S, Zhang X, Hatch K, Zheng W (2011) Serous endometrial intraepithelial carcinoma arising in adenomyosis: a report of 5 cases. Int J Gynecol Pathol 30:271–281. https://doi.org/10.1097/ PGP.0b013e318200868e Adamson GD (1992) Treatment of uterine fibroids: current findings with gonadotropin-releasing hormone agonists. Am J Obstet Gynecol 166:746–751 Adany R, Fodor F, Molnar P, Ablin RJ, Muszbek L (1990) Increased density of histiocytes in uterine leiomyomas. Int J Gynecol Pathol 9:137–144 Agaram NP et al (2015) Dichotomy of genetic abnormalities in PEComas with therapeutic implications. Am J Surg Pathol 39:813–825. https://doi.org/10.1097/ pas.0000000000000389 Aggarwal N, Bhargava R, Elishaev E (2012) Uterine adenosarcomas: diagnostic use of the proliferation marker ki-67 as an adjunct to morphologic diagnosis. Int J Gynecol Pathol 31:447–452. https://doi.org/ 10.1097/PGP.0b013e318249285b

E. Oliva et al. Agoff SN, Grieco VS, Garcia R, Gown AM (2001) Immunohistochemical distinction of endometrial stromal sarcoma and cellular leiomyoma. Appl Immunohistochem Mol Morphol 9:164–169 Ahmad AK et al (2014) Institutional review of primary non-hodgkin lymphoma of the female genital tract: a 33-year experience. Int J Gynecol Cancer 24:1250–1255. https://doi.org/10.1097/igc.000000000 0000201 Ahrens WA, Folpe AL (2011) CD1a immunopositivity in perivascular epithelioid cell neoplasms: true expression or technical artifact? A streptavidin-biotin and polymer-based detection system immunohistochemical study of perivascular epithelioid cell neoplasms and their morphologic mimics. Hum Pathol 42:369–374. https://doi.org/10.1016/j.humpath.2010.09.001 Aisagbonhi O, Harrison B, Zhao L, Osgood R, Chebib I, Oliva E (2017) YWHAE rearrangement in a purely conventional low-grade endometrial stromal sarcoma that transformed over time to high-grade sarcoma: importance of molecular testing. Int J Gynecol Pathol. https://doi.org/10.1097/PGP.0000000000000451 Akahira J et al (2006) Prognoses and prognostic factors of carcinosarcoma, endometrial stromal sarcoma and uterine leiomyosarcoma: a comparison with uterine endometrial adenocarcinoma. Oncology 71:333–340. https://doi.org/10.1159/000107107 Akbulut M, Zekioglu O, Terek MC, Ozdemir N (2008) Lipoadenofibroma of the endometrium: a rare variant of benign mullerian mixed tumor. Arch Gynecol Obstet 278:283–286. https://doi.org/10.1007/ s00404-008-0572-0 Albrektsen G, Heuch I, Wik E, Salvesen HB (2009) Prognostic impact of parity in 493 uterine sarcoma patients. Int J Gynecol Cancer 19:1062–1067. https://doi.org/10.1111/ IGC.0b013e3181aaa950. 00009577-200908000-00013 [pii] Ali RH et al (2014) Molecular characterization of a population-based series of endometrial stromal sarcomas in Kuwait. Hum Pathol 45:2453–2462. https://doi. org/10.1016/j.humpath.2014.08.012. S0046-8177(14) 00356-6 [pii] Alsolami S et al (2014) Current morphologic criteria perform poorly in identifying hereditary leiomyomatosis and renal cell carcinoma syndrome-associated uterine leiomyomas. Int J Gynecol Pathol 33:560–567. https:// doi.org/10.1097/pgp.0000000000000091 Amada S, Nakano H, Tsuneyoshi M (1995) Leiomyosarcoma versus bizarre and cellular leiomyomas of the uterus: a comparative study based on the MIB-1 and proliferating cell nuclear antigen indices, p53 expression, DNA flow cytometry, and muscle specific actins. Int J Gynecol Pathol 14:134–142 Amant F et al (2004) Immunohistochemical determination of estrogen and progesterone receptor positivity in uterine adenosarcoma. Gynecol Oncol 93:680–685. https:// doi.org/10.1016/j.ygyno.2004.03.021. S009082580400 1921 [pii]

10

Mesenchymal Tumors of the Uterus

Amant F, Woestenborghs H, Vandenbroucke V, Berteloot P, Neven P, Moerman P, Vergote I (2006) Transition of endometrial stromal sarcoma into high-grade sarcoma. Gynecol Oncol 103:1137–1140. https://doi.org/ 10.1016/j.ygyno.2006.07.013 Amant F et al (2007) Clinical study investigating the role of lymphadenectomy, surgical castration and adjuvant hormonal treatment in endometrial stromal sarcoma. Br J Cancer 97:1194–1199. https://doi.org/10.1038/sj. bjc.6603986 Amant F, Tousseyn T, Coenegrachts L, Decloedt J, Moerman P, Debiec-Rychter M (2011) Case report of a poorly differentiated uterine tumour with t(10;17) translocation and neuroectodermal phenotype. Anticancer Res 31:2367–2371. 31/6/2367 [pii] American Cancer Society (2017) Survival rates of uterine sarcoma by stage. Cited 5 Jun 2017. www.cancer.org/ cancer/uterine-sarcoma/detection-diagnosis-staging/ survival-rates.html Andersen J (1998) Factors in fibroid growth. Baillieres Clin Obstet Gynaecol 12:225–243 Argani P et al (2001) Primary renal neoplasms with the ASPL-TFE3 gene fusion of alveolar soft part sarcoma: a distinctive tumor entity previously included among renal cell carcinomas of children and adolescents. Am J Pathol 159:179–192. https://doi.org/10.1016/s00029440(10)61684-7 Argani P, Lal P, Hutchinson B, Lui MY, Reuter VE, Ladanyi M (2003) Aberrant nuclear immunoreactivity for TFE3 in neoplasms with TFE3 gene fusions: a sensitive and specific immunohistochemical assay. Am J Surg Pathol 27:750–761 Argani P, Zhong M, Reuter VE, Fallon JT, Epstein JI, Netto GJ, Antonescu CR (2016) TFE3-fusion variant analysis defines specific clinicopathologic associations among Xp11 translocation cancers. Am J Surg Pathol 40:723–737. https://doi.org/10.1097/ pas.0000000000000631 Arias-Stella JA, Lewis N, Benayed R, Oliva E, Young R, Hoang L, Lee C-H, Jungbluth A, Soslow R, Antonescu C, Ladanyi M, Chiang S (2018) Novel PLAG1 gene rearrangement distinguishes uterine myxoid leiomyosarcoma from other uterine myxoid mesenchymal tumors. Paper presented at the United States and Canadian Academy of Pathology (USCAP) 107th annual meeting, Vancouver Atkins KA, Bell ME, Kempson RL, Hendrickson MR (2001) Epithelioid smooth muscle neoplasms of the uterus. Mod Pathol 14:132A Atkins KA, Arronte N, Darus CJ, Rice LW (2008) The use of p16 in enhancing the histologic classification of uterine smooth muscle tumors. Am J Surg Pathol 32:98–102. https://doi.org/10.1097/PAS.0b013e3181574d1e Attygalle AD et al (2017) An unusual case of YWHAENUTM2A/B endometrial stromal sarcoma with confinement to the endometrium and lack of high-grade morphology. Int J Gynecol Pathol 36:165–171. https:// doi.org/10.1097/PGP.0000000000000286

625 Aubry MC, Myers JL, Colby TV, Leslie KO, Tazelaar HD (2002) Endometrial stromal sarcoma metastatic to the lung: a detailed analysis of 16 patients. Am J Surg Pathol 26:440–449 Baird DD, Dunson DB, Hill MC, Cousins D, Schectman JM (2003) High cumulative incidence of uterine leiomyoma in black and white women: ultrasound evidence. Am J Obstet Gynecol 188:100–107 Baker P, Oliva E (2007) Endometrial stromal tumours of the uterus: a practical approach using conventional morphology and ancillary techniques. J Clin Pathol 60:235–243 Baker RJ, Hildebrandt RH, Rouse RV, Hendrickson MR, Longacre TA (1999) Inhibin and CD99 (MIC2) expression in uterine stromal neoplasms with sex-cord-like elements. Hum Pathol 30:671–679 Baker PM, Moch H, Oliva E (2005) Unusual morphologic features of endometrial stromal tumors: a report of 2 cases. Am J Surg Pathol 29:1394–1398. 00000478200510000-00018 [pii] Baldi GG et al (2013) Solitary fibrous tumor of all sites: outcome of late recurrences in 14 patients. Clin Sarcoma Res 3:4. https://doi.org/10.1186/20453329-3-4 Balleine RL, Earls PJ, Webster LR, Mote PA, deFazio A, Harnett PR, Clarke CL (2004) Expression of progesterone receptor A and B isoforms in low-grade endometrial stromal sarcoma. Int J Gynecol Pathol 23:138–144. 00004347-200404000-00008 [pii] Bardella C et al (2011) Aberrant succination of proteins in fumarate hydratase-deficient mice and HLRCC patients is a robust biomarker of mutation status. J Pathol 225:4–11. https://doi.org/10.1002/path.2932 Barlin JN et al (2015) Molecular subtypes of uterine leiomyosarcoma and correlation with clinical outcome. Neoplasia 17:183–189. https://doi.org/10.1016/j. neo.2014.12.007 Barlow JF et al (1983) Myxoid tumor of the uterus and right atrial myxomas. S D J Med 36:9–13 Barney B, Tward JD, Skidmore T, Gaffney DK (2009) Does radiotherapy or lymphadenectomy improve survival in endometrial stromal sarcoma? Int J Gynecol Cancer 19:1232–1238. https://doi.org/10.1111/IGC.0b013e3181 b33c9a. 00009577-200910000-00015 [pii] Bartosch C, Exposito MI, Lopes JM (2010) Low-grade endometrial stromal sarcoma and undifferentiated endometrial sarcoma: a comparative analysis emphasizing the importance of distinguishing between these two groups. Int J Surg Pathol 18:286–291. https://doi. org/10.1177/1066896909337600 Beer TW, Buchanan R, Buckley CH (1995) Uterine stromal sarcoma following tamoxifen treatment. J Clin Pathol 48:596 Bekkers RL, Willemsen WN, Schijf CP, Massuger LF, Bulten J, Merkus JM (1999) Leiomyomatosis peritonealis disseminata: does malignant transformation occur? A literature review. Gynecol Oncol 75:158–163. https://doi.org/10.1006/gyno.1999.5490

626 Bell SW, Kempson RL, Hendrickson MR (1994) Problematic uterine smooth muscle neoplasms. A clinicopathologic study of 213 cases. Am J Surg Pathol 18:535–558 Bennett JA, Lamb C, Young RH (2016) Apoplectic leiomyomas: a morphologic analysis of 100 cases highlighting unusual features. Am J Surg Pathol 40:563–568. https://doi.org/10.1097/pas.0000000000 000569 Bennett JA, Nardi V, Rouzbahman M, Morales-Oyarvide V, Nielsen GP, Oliva E (2017a) Inflammatory myofibroblastic tumor of the uterus: a clinicopathological, immunohistochemical, and molecular analysis of 13 cases highlighting their broad morphologic spectrum. Mod Pathol 30:1489–1503. https://doi.org/ 10.1038/modpathol.2017.69 Bennett JA et al (2017b) Leiomyoma with bizarre nuclei: a morphological, immunohistochemical and molecular analysis of 31 cases. Mod Pathol 30:1476–1488. https://doi.org/10.1038/modpathol.2017.56 Bergeron C, Amant F, Ferenczy A (2006) Pathology and physiopathology of adenomyosis. Best Pract Res Clin Obstet Gynaecol 20:511–521 Bernard B et al (2013) Uterine adenosarcomas: a dualinstitution update on staging, prognosis and survival. Gynecol Oncol 131:634–639. https://doi.org/10.1016/ j.ygyno.2013.09.011 Bertsch E et al (2014) MED12 and HMGA2 mutations: two independent genetic events in uterine leiomyoma and leiomyosarcoma. Mod Pathol 27:1144–1153. https://doi.org/10.1038/modpathol.2013.243 Binder SW, Nieberg RK, Cheng L, al-Jitawi S (1991) Histologic and immunohistochemical analysis of nine endometrial stromal tumors: an unexpected high frequency of keratin protein positivity. Int J Gynecol Pathol 10:191–197 Blake EA, Sheridan TB, Wang KL, Takiuchi T, Kodama M, Sawada K, Matsuo K (2014) Clinical characteristics and outcomes of uterine tumors resembling ovarian sex-cord tumors (UTROSCT): a systematic review of literature. Eur J Obstet Gynecol Reprod Biol 181:163–170. https://doi.org/10.1016/j.ejogrb. 2014.07.050 Blattner JM, Gable P, Quigley MM, McHale MT (2007) Primitive neuroectodermal tumor of the uterus. Gynecol Oncol 106:419–422 Blom R, Guerrieri C (1999) Adenosarcoma of the uterus: a clinicopathologic, DNA flow cytometric, p53 and mdm-2 analysis of 11 cases. Int J Gynecol Cancer 9:37–43 Blom R, Guerrieri C, Stal O, Malmstrom H, Simonsen E (1998) Leiomyosarcoma of the uterus: a clinicopathologic, DNA flow cytometric, p53, and mdm-2 analysis of 49 cases. Gynecol Oncol 68:54–61 Bodner K et al (2001) Prognostic parameters in endometrial stromal sarcoma: a clinicopathologic study in 31 patients. Gynecol Oncol 81:160–165. https://doi. org/10.1006/gyno.2001.6152. S0090-8258(01)961529 [pii]

E. Oliva et al. Bodner K, Bodner-Adler B, Kimberger O, Czerwenka K, Mayerhofer K (2004) Estrogen and progesterone receptor expression in patients with uterine smooth muscle tumors. Fertil Steril 81:1062–1066. https://doi.org/ 10.1016/j.fertnstert.2003.08.043 Bodner-Adler B, Bodner K, Czerwenka K, Kimberger O, Leodolter S, Mayerhofer K (2005) Expression of p16 protein in patients with uterine smooth muscle tumors: an immunohistochemical analysis. Gynecol Oncol 96:62–66. https://doi.org/10.1016/j.ygyno.2004.09.026 Bogani G et al (2016) Morcellator’s port-site metastasis of a uterine smooth muscle tumor of uncertain malignant potential after minimally invasive myomectomy. J Minim Invasive Gynecol 23:647–649. https://doi.org/ 10.1016/j.jmig.2016.01.021 Borahay MA, Al-Hendy A, Kilic GS, Boehning D (2015) Signaling pathways in leiomyoma: understanding pathobiology and implications for therapy. Mol Med 21: 242–256. https://doi.org/10.2119/molmed.2014.00053 Botsis D, Koliopoulos C, Kondi-Pafitis A, Creatsas G (2005) Frequency, histological, and immunohistochemical properties of massive inflammatory lymphocytic infiltration of leiomyomas of the uterus: an entity causing diagnostic difficulties. Int J Gynecol Pathol 24:326–329 Bouchard-Fortier G, Kim RH, Allen L, Gupta A, May T (2016) Fertility-sparing surgery for the management of young women with embryonal rhabdomyosarcoma of the cervix: a case series. Gynecol Oncol Rep 18:4–7. https://doi.org/10.1016/j.gore.2016.08.004 Brescia RJ, Tazelaar HD, Hobbs J, Miller AW (1989) Intravascular lipoleiomyomatosis: a report of two cases. Hum Pathol 20:252–256 Brooks JJ, Wells GB, Yeh IT, LiVolsi VA (1992) Bizarre epithelioid lipoleiomyoma of the uterus. Int J Gynecol Pathol 11:144–149 Brooks SE, Zhan M, Cote T, Baquet CR (2004) Surveillance, epidemiology, and end results analysis of 2677 cases of uterine sarcoma 1989–1999. Gynecol Oncol 93:204–208. https://doi.org/10.1016/j.ygyno.2003.12.029 Brown DC, Theaker JM, Banks PM, Gatter KC, Mason DY (1987) Cytokeratin expression in smooth muscle and smooth muscle tumours. Histopathology 11:477–486 Buelow B, Cohen J, Nagymanyoki Z, Frizzell N, Joseph NM, McCalmont T, Garg K (2016) Immunohistochemistry for 2-Succinocysteine (2SC) and fumarate hydratase (FH) in cutaneous leiomyomas may aid in identification of patients with HLRCC (Hereditary leiomyomatosis and renal cell carcinoma syndrome). Am J Surg Pathol 40:982–988. https://doi.org/10.1097/ pas.0000000000000626 Bukulmez O, Doody KJ (2006) Clinical features of myomas. Obstet Gynecol Clin N Am 33:69–84. https://doi. org/10.1016/j.ogc.2005.12.002 Burch DM, Tavassoli FA (2011) Myxoid leiomyosarcoma of the uterus. Histopathology 59:1144–1155. https:// doi.org/10.1111/j.1365-2559.2011.04053.x

10

Mesenchymal Tumors of the Uterus

Busca A, Parra-Herran C (2016) Hemangiomas of the uterine cervix: association with abnormal bleeding and pain in young women and hormone receptor expression. Report of four cases and review of the literature. Pathol Res Pract 212:532–538. https://doi. org/10.1016/j.prp.2016.03.003 Busca A, Parra-Herran C (2017) Myxoid mesenchymal tumors of the uterus: an update on classification, definitions, and differential diagnosis. Adv Anat Pathol 24:354–361. https://doi.org/10.1097/pap. 0000000000000164 Busca A, Gulavita P, Parra-Herran C, Islam S (2017) IFITM1 outperforms CD10 in differentiating lowgrade endometrial stromal sarcomas from smooth muscle neoplasms of the uterus. Int J Gynecol Pathol. https://doi.org/10.1097/PGP.0000000000000424 Buza N, Xu F, Wu W, Carr RJ, Li P, Hui P (2014) Recurrent chromosomal aberrations in intravenous leiomyomatosis of the uterus: high-resolution array comparative genomic hybridization study. Hum Pathol 45:1885–1892. https:// doi.org/10.1016/j.humpath.2014.05.010 Cardinale L, Mirra M, Galli C, Goldblum JR, Pizzolitto S, Falconieri G (2008) Angiosarcoma of the uterus: report of 2 new cases with deviant clinicopathologic features and review of the literature. Ann Diagn Pathol 12:217–221. https://doi.org/10.1016/j.anndiagpath.2007.08.001. S1092-9134(07)00123-2 [pii] Carroll A et al (2014) Uterine adenosarcoma: an analysis on management, outcomes, and risk factors for recurrence. Gynecol Oncol 135:455–461. https://doi.org/ 10.1016/j.ygyno.2014.10.022 Casey S, McCluggage WG (2015) Adenomyomas of the uterine cervix: report of a cohort including endocervical and novel variants [corrected]. Histopathology 66:420–429. https://doi.org/10.1111/his.12546 Chan JK, Kawar NM, Shin JY, Osann K, Chen LM, Powell CB, Kapp DS (2008) Endometrial stromal sarcoma: a population-based analysis. Br J Cancer 99:1210–1215. https://doi.org/10.1038/sj.bjc.6604527 Chang V, Aikawa M, Druet R (1977) Uterine leiomyoblastoma: ultrastructural and cytological studies. Cancer 39:1563–1569 Chang KL, Crabtree GS, Lim-Tan SK, Kempson RL, Hendrickson MR (1990) Primary uterine endometrial stromal neoplasms. A clinicopathologic study of 117 cases. Am J Surg Pathol 14:415–438 Charville GW et al (2016) PAX7 expression in rhabdomyosarcoma, related soft tissue tumors, and small round blue cell neoplasms. Am J Surg Pathol 40:1305–1315. https:// doi.org/10.1097/pas.0000000000000717 Chauveinc L et al (1999) Uterine sarcomas: the Curie Institut experience. Prognosis factors and adjuvant treatments. Gynecol Oncol 72:232–237. https://doi.org/10.1006/ gyno.1998.5251. S0090-8258(98)95251-9 [pii] Chen KT (1999) Uterine leiomyohibernoma. Int J Gynecol Pathol 18:96–97 Chen L, Yang B (2008) Immunohistochemical analysis of p16, p53, and Ki-67 expression in uterine smooth

627 muscle tumors. Int J Gynecol Pathol 27:326–332. https://doi.org/10.1097/PGP.0b013e31815ea7f5 Cheung AN, Ng WF, Chung LP, Khoo US (1996) Mixed low grade and high grade endometrial stromal sarcoma of uterus: differences on immunohistochemistry and chromosome in situ hybridisation. J Clin Pathol 49:604–607 Chew I, Oliva E (2010) Endometrial stromal sarcomas: a review of potential prognostic factors. Adv Anat Pathol 17:113–121. https://doi.org/10.1097/PAP.0b01 3e3181cfb7c2. 00125480-201003000-00003 [pii] Chiang S et al (2011) Frequency of known gene rearrangements in endometrial stromal tumors. Am J Surg Pathol 35:1364–1372. https://doi.org/10.1097/ PAS.0b013e3182262743 Chiang S et al (2015) FOXL2 mutation is absent in uterine tumors resembling ovarian sex cord tumors. Am J Surg Pathol 39:618–623. https://doi.org/ 10.1097/pas.0000000000000367 Chiang S et al (2017a) BCOR is a robust diagnostic immunohistochemical marker of genetically diverse high-grade endometrial stromal sarcoma, including tumors exhibiting variant morphology. Mod Pathol 30:1251–1261. https:// doi.org/10.1038/modpathol.2017.42 Chiang S et al (2017b) Primitive neuroectodermal tumors of the female genital tract: a morphologic, immunohistochemical, and molecular study of 19 cases. Am J Surg Pathol 41:761–772. https://doi. org/10.1097/pas.0000000000000831 Chiang S et al (2018) NTRK fusions define a novel uterine sarcoma subtype with features of fibrosarcoma. Am J Surg Pathol 42:791–798. https://doi.org/10.1097/ pas.0000000000001055 Cho KR, Woodruff JD, Epstein JI (1989) Leiomyoma of the uterus with multiple extrauterine smooth muscle tumors: a case report suggesting multifocal origin. Hum Pathol 20:80–83 Choi YJ et al (2015) Genomic landscape of endometrial stromal sarcoma of uterus. Oncotarget 6:33319–33328. https://doi.org/10.18632/oncotarget.5384 Chou WY, Chang HW (2012) Uterine hemangioma: a rare pathologic entity. Arch Pathol Lab Med 136:567–571. https://doi.org/10.5858/arpa.2011-0078-RS Chow KL et al (2017) The mitosis-specific marker phosphohistone-H3 (PHH3) is an independent prognosticator in uterine smooth muscle tumours: an outcome-based study. Histopathology 70:746–755. https://doi.org/10.1111/his.13124 Christacos NC, Quade BJ, Dal Cin P, Morton CC (2006) Uterine leiomyomata with deletions of Ip represent a distinct cytogenetic subgroup associated with unusual histologic features. Genes Chromosomes Cancer 45:304–312. https://doi.org/10.1002/gcc.20291 Chu P, Arber DA (2000) Paraffin-section detection of CD10 in 505 nonhematopoietic neoplasms. Frequent expression in renal cell carcinoma and endometrial stromal sarcoma. Am J Clin Pathol 113:374–382. https://doi.org/10.1309/8vav-j2fu-8cu9-ek18

628 Chu PG, Arber DA, Weiss LM, Chang KL (2001) Utility of CD10 in distinguishing between endometrial stromal sarcoma and uterine smooth muscle tumors: an immunohistochemical comparison of 34 cases. Mod Pathol 14:465–471. https://doi.org/10.1038/modpathol.3880335 Chu MC, Mor G, Lim C, Zheng W, Parkash V, Schwartz PE (2003) Low-grade endometrial stromal sarcoma: hormonal aspects. Gynecol Oncol 90:170–176. S0090825803002580 [pii] Clement PB (1988) Intravenous leiomyomatosis of the uterus. Pathol Annu 23 Pt 2:153–183 Clement PB (1989) Mullerian adenosarcomas of the uterus with sarcomatous overgrowth. A clinicopathological analysis of 10 cases. Am J Surg Pathol 13:28–38 Clement PB (2000) The pathology of uterine smooth muscle tumors and mixed endometrial stromal-smooth muscle tumors: a selective review with emphasis on recent advances. Int J Gynecol Pathol 19:39–55 Clement PB (2007) The pathology of endometriosis: a survey of the many faces of a common disease emphasizing diagnostic pitfalls and unusual and newly appreciated aspects. Adv Anat Pathol 14:241–260. https:// doi.org/10.1097/PAP.0b013e3180ca7d7b Clement PB, Scully RE (1974) Mullerian adenosarcoma of the uterus. A clinicopathologic analysis of ten cases of a distinctive type of mullerian mixed tumor. Cancer 34:1138–1149 Clement PB, Scully RE (1976) Uterine tumors resembling ovarian sex-cord tumors. A clinicopathologic analysis of fourteen cases. Am J Clin Pathol 66:512–525 Clement PB, Scully RE (1989) Mullerian adenosarcomas of the uterus with sex cord-like elements. A clinicopathologic analysis of eight cases. Am J Clin Pathol 91:664–672 Clement PB, Scully RE (1990a) Mullerian adenofibroma of the uterus with invasion of myometrium and pelvic veins. Int J Gynecol Pathol 9:363–371 Clement PB, Scully RE (1990b) Mullerian adenosarcoma of the uterus: a clinicopathologic analysis of 100 cases with a review of the literature. Hum Pathol 21:363–381 Clement PB, Scully RE (1992) 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 11:163–173 Clement PB, Young RH, Scully RE (1988) Intravenous leiomyomatosis of the uterus. A clinicopathological analysis of 16 cases with unusual histologic features. Am J Surg Pathol 12:932–945 Clement PB, Young RH, Scully RE (1992) Diffuse, perinodular, and other patterns of hydropic degeneration within and adjacent to uterine leiomyomas. Problems in differential diagnosis. Am J Surg Pathol 16:26–32 Coad JE, Sulaiman RA, Das K, Staley N (1997) Perinodular hydropic degeneration of a uterine leiomyoma: a diagnostic challenge. Hum Pathol 28:249–251 Coard KC, Fletcher HM (2002) Leiomyosarcoma of the uterus with a florid intravascular component

E. Oliva et al. (“intravenous leiomyosarcomatosis”). Int J Gynecol Pathol 21:182–185 Cohen DT, Oliva E, Hahn PF, Fuller AF Jr, Lee SI (2007) Uterine smooth-muscle tumors with unusual growth patterns: imaging with pathologic correlation. AJR Am J Roentgenol 188:246–255. https://doi.org/ 10.2214/ajr.05.1070 Conlon N, Soslow RA, Murali R (2015) Perivascular epithelioid tumours (PEComas) of the gynaecological tract. J Clin Pathol 68:418–426. https://doi.org/ 10.1136/jclinpath-2015-202945 Cramer SF, Patel A (1990) The frequency of uterine leiomyomas. Am J Clin Pathol 94:435–438 Cramer SF, Horiszny J, Patel A, Sigrist S (1996) The relation of fibrous degeneration to menopausal status in small uterine leiomyomas with evidence for postmenopausal origin of seedling myomas. Mod Pathol 9:774–780 Croce S, Chibon F (2015) MED12 and uterine smooth muscle oncogenesis: state of the art and perspectives. Eur J Cancer 51:1603–1610. https://doi.org/10.1016/j. ejca.2015.04.023 Croce S et al (2013) YWHAE rearrangement identified by FISH and RT-PCR in endometrial stromal sarcomas: genetic and pathological correlations. Mod Pathol 26:1390–1400. https://doi.org/10.1038/ modpathol.2013.69 Croce S, Young RH, Oliva E (2014) Uterine leiomyomas with bizarre nuclei: a clinicopathologic study of 59 cases. Am J Surg Pathol 38:1330–1339. https:// doi.org/10.1097/pas.0000000000000249 Croce S et al (2015) Uterine smooth muscle tumor analysis by comparative genomic hybridization: a useful diagnostic tool in challenging lesions. Mod Pathol 28:1001–1010. https://doi.org/10.1038/modpathol.2015.3 Croce S, de Kock L, Boshari T, Hostein I, Velasco V, Foulkes WD, McCluggage WG (2016) Uterine tumor resembling ovarian sex cord tumor (UTROSCT) commonly exhibits positivity with sex cord markers FOXL2 and SF-1 but lacks FOXL2 and DICER1 mutations. Int J Gynecol Pathol 35:301–308. https://doi.org/ 10.1097/pgp.0000000000000240 Crow J, Wilkins M, Howe S, More L, Helliwell P (1991) Mast cells in the female genital tract. Int J Gynecol Pathol 10:230–237 Cui RR, Wright JD (2016) Risk of occult uterine sarcoma in presumed uterine fibroids. Clin Obstet Gynecol 59:103–118. https://doi.org/10.1097/grf. 0000000000000163 D’Angelo E, Prat J (2010) Uterine sarcomas: a review. Gynecol Oncol 116:131–139. https://doi.org/10.1016/ j.ygyno.2009.09.023. S0090-8258(09)00722-7 [pii] D’Angelo E, Ali RH, Espinosa I, Lee CH, Huntsman DG, Gilks B, Prat J (2013) Endometrial stromal sarcomas with sex cord differentiation are associated with PHF1 rearrangement. Am J Surg Pathol 37:514–521. https:// doi.org/10.1097/PAS.0b013e318272c612 Dal Cin P, Aly MS, De Wever I, Moerman P, Van Den Berghe H (1992) Endometrial stromal sarcoma t(7;17) (p15-21;q12-21) is a nonrandom chromosome change.

10

Mesenchymal Tumors of the Uterus

Cancer Genet Cytogenet 63:43–46. 0165-4608(92) 90062-D [pii] Darby AJ, Papadaki L, Beilby JO (1975) An unusual leiomyosarcoma of the uterus containing osteoclastlike giant cells. Cancer 36:495–504 Davidson B et al (2014) Gene expression signatures of primary and metastatic uterine leiomyosarcoma. Hum Pathol 45:691–700. https://doi.org/10.1016/j. humpath.2013.11.003 Daya DA, Scully RE (1988) Sarcoma botryoides of the uterine cervix in young women: a clinicopathological study of 13 cases. Gynecol Oncol 29:290–304 De Fusco PA, Gaffey TA, Malkasian GD Jr, Long HJ, Cha SS (1989) Endometrial stromal sarcoma: review of Mayo Clinic experience, 1945–1980. Gynecol Oncol 35:8–14 de Leval L, Waltregny D, Boniver J, Young RH, Castronovo V, Oliva E (2006) Use of histone deacetylase 8 (HDAC8), a new marker of smooth muscle differentiation, in the classification of mesenchymal tumors of the uterus. Am J Surg Pathol 30:319–327. https://doi.org/10.1097/01.pas.0000188029.63706.31. 00000478-200603000-00004 [pii] de Leval L, Lim GS, Waltregny D, Oliva E (2010) Diverse phenotypic profile of uterine tumors resembling ovarian sex cord tumors: an immunohistochemical study of 12 cases. Am J Surg Pathol 34:1749–1761. https://doi. org/10.1097/PAS.0b013e3181f8120c de Vos S, Wilczynski SP, Fleischhacker M, Koeffler P (1994) p53 alterations in uterine leiomyosarcomas versus leiomyomas. Gynecol Oncol 54:205–208. https:// doi.org/10.1006/gyno.1994.1194 Dehner LP, Jarzembowski JA, Hill DA (2012) Embryonal rhabdomyosarcoma of the uterine cervix: a report of 14 cases and a discussion of its unusual clinicopathological associations. Mod Pathol 25:602–614. https:// doi.org/10.1038/modpathol.2011.185 Devaney K, Tavassoli FA (1991) Immunohistochemistry as a diagnostic aid in the interpretation of unusual mesenchymal tumors of the uterus. Mod Pathol 4:225–231 Dewaele B et al (2014) Identification of a novel, recurrent MBTD1-CXorf67 fusion in low-grade endometrial stromal sarcoma. Int J Cancer 134:1112–1122. https:// doi.org/10.1002/ijc.28440 Dgani R et al (1998) Clinical-pathological study of uterine leiomyomas with high mitotic activity. Acta Obstet Gynecol Scand 77:74–77 Dharkar DD, Kraft JR, Gangadharam D (1981) Uterine lipomas. Arch Pathol Lab Med 105:43–45 Dimmler A, Seitz G, Hohenberger W, Kirchner T, Faller G (2003) Late pulmonary metastasis in uterine PEComa. J Clin Pathol 56:627–628 Dionigi A, Oliva E, Clement PB, Young RH (2002) Endometrial stromal nodules and endometrial stromal tumors with limited infiltration: a clinicopathologic study of 50 cases. Am J Surg Pathol 26:567–581 Dos Santos LA et al (2011) Incidence of lymph node and adnexal metastasis in endometrial stromal sarcoma.

629 Gynecol Oncol 121:319–322. https://doi.org/10.1016/ j.ygyno.2010.12.363. S0090-8258(10)01270-9 [pii] Downes KA, Hart WR (1997) Bizarre leiomyomas of the uterus: a comprehensive pathologic study of 24 cases with long-term follow-up. Am J Surg Pathol 21:1261–1270 Due W, Pickartz H (1989) Immunohistologic detection of estrogen and progesterone receptors in disseminated peritoneal leiomyomatosis. Int J Gynecol Pathol 8:46–53 Dueholm M, Lundorf E, Hansen ES, Ledertoug S, Olesen F (2002) Accuracy of magnetic resonance imaging and transvaginal ultrasonography in the diagnosis, mapping, and measurement of uterine myomas. Am J Obstet Gynecol 186:409–415 Dundr P, Mara M, Maskova J, Fucikova Z, Povysil C, Tvrdik D (2006) Pathological findings of uterine leiomyomas and adenomyosis following uterine artery embolization. Pathol Res Pract 202:721–729. https:// doi.org/10.1016/j.prp.2006.07.001 Eddy GL, Mazur MT (1997) Endolymphatic stromal myosis associated with tamoxifen use. Gynecol Oncol 64:262–264. https://doi.org/10.1006/gyno.1996.4552. S0090-8258(96)94552-7 [pii] Eichhorn JH, Young RH, Clement PB (1996) Sertoliform endometrial adenocarcinoma: a study of four cases. Int J Gynecol Pathol 15:119–126 Elshafie M, Rahimi S, Ganesan R, Hirschowitz L (2013) Mullerian adenosarcoma arising in a subserosal adenomyoma. Int J Surg Pathol 21:186–189. https:// doi.org/10.1177/1066896912453852 Euscher ED, Deavers MT, Lopez-Terrada D, Lazar AJ, Silva EG, Malpica A (2008) Uterine tumors with neuroectodermal differentiation: a series of 17 cases and review of the literature. Am J Surg Pathol 32:219–228. https://doi.org/10.1097/PAS.0b013e318093e421 Evans HL (1982) Endometrial stromal sarcoma and poorly differentiated endometrial sarcoma. Cancer 50: 2170–2182 Evans AT 3rd, Symmonds RE, Gaffey TA (1981) Recurrent pelvic intravenous leiomyomatosis. Obstet Gynecol 57:260–264 Evans HL, Chawla SP, Simpson C, Finn KP (1988) Smooth muscle neoplasms of the uterus other than ordinary leiomyoma. A study of 46 cases, with emphasis on diagnostic criteria and prognostic factors. Cancer 62:2239–2247 Eyden BP, Hale RJ, Richmond I, Buckley CH (1992) Cytoskeletal filaments in the smooth muscle cells of uterine leiomyomata and myometrium: an ultrastructural and immunohistochemical analysis. Virchows Arch A Pathol Anat Histopathol 420:51–58 Fadare O (2008) Perivascular epithelioid cell tumor (PEComa) of the uterus: an outcome-based clinicopathologic analysis of 41 reported cases. Adv Anat Pathol 15:63–75. https://doi.org/10.1097/ PAP.0b013e31816613b0 Fadare O (2011) Heterologous and rare homologous sarcomas of the uterine corpus: a clinicopathologic review.

630 Adv Anat Pathol 18:60–74. https://doi.org/10.1097/ PAP.0b013e3182026be7 Fadare O, McCalip B, Mariappan MR, Hileeto D, Parkash V (2005) An endometrial stromal tumor with osteoclast-like giant cells. Ann Diagn Pathol 9:160–165. S109291340500033X [pii] Fadare O, Bonvicino A, Martel M, Renshaw IL, Azodi M, Parkash V (2010) Pleomorphic rhabdomyosarcoma of the uterine corpus: a clinicopathologic study of 4 cases and a review of the literature. Int J Gynecol Pathol 29:122–134. https://doi.org/10.1097/PGP.0b013e3181bc98c0. 00004347-201003000-00005 [pii] Farhood AI, Abrams J (1991) Immunohistochemistry of endometrial stromal sarcoma. Hum Pathol 22: 224–230 Fekete PS, Vellios F (1984) The clinical and histologic spectrum of endometrial stromal neoplasms: a report of 41 cases. Int J Gynecol Pathol 3:198–212 Felix AS et al (2013) The etiology of uterine sarcomas: a pooled analysis of the epidemiology of endometrial cancer consortium. Br J Cancer 108:727–734. https:// doi.org/10.1038/bjc.2013.2 Feng W, Hua K, Malpica A, Zhou X, Baak JP (2013a) Stages I to II WHO 2003-defined low-grade endometrial stromal sarcoma: how much primary therapy is needed and how little is enough? Int J Gynecol Cancer 23:488–493. https://doi.org/10.1097/IGC.0b013e318247aa14. 00009577-201303000-00015 [pii] Feng W, Malpica A, Robboy SJ, Gudlaugsson E, Hua K, Zhou X, Baak JP (2013b) Prognostic value of the diagnostic criteria distinguishing endometrial stromal sarcoma, low grade from undifferentiated endometrial sarcoma, 2 entities within the invasive endometrial stromal neoplasia family. Int J Gynecol Pathol 32:299–306. https://doi.org/10.1097/PGP.0b013e318229adfb Feng W et al (2013c) Can proliferation biomarkers reliably predict recurrence in World Health Organization 2003 defined endometrial stromal sarcoma, low grade? PLoS One 8:e75899. https://doi.org/10.1371/journal. pone.0075899. PONE-D-13-18388 [pii] Feng W, Malpica A, Yinhua Y, Janssen E, Gudlaugsson E, Zhou X, Baak JP (2013d) Diagnostic and prognostic morphometric features in WHO 2003 invasive endometrial stromal tumours. Histopathology 62:688–694. https://doi.org/10.1111/j.1365-2559.2011.04120.x Ferguson SE, Gerald W, Barakat RR, Chi DS, Soslow RA (2007) Clinicopathologic features of rhabdomyosarcoma of gynecologic origin in adults. Am J Surg Pathol 31:382–389 Ferry JA, Harris NL, Scully RE (1989) Uterine leiomyomas with lymphoid infiltration simulating lymphoma: a report of seven cases. Int J Gynecol Pathol 8:263–270 Fitko R, Brainer J, Schink JC, August CZ (1990) Endometrial stromal sarcoma with rhabdoid differentiation. Int J Gynecol Pathol 9:379–382 Fleming WP, Peters WA, Kumar NB, Morley GW (1984) Autopsy findings in patients with uterine sarcoma. Gynecol Oncol 19:168–172

E. Oliva et al. Fleming H, Ostor AG, Pickel H, Fortune DW (1989) Arteriovenous malformations of the uterus. Obstet Gynecol 73:209–214 Flicker K, Smolle E, Haybaeck J, Moinfar F (2015) Genomic characterization of endometrial stromal sarcomas with array comparative genomic hybridization. Exp Mol Pathol 98:367–374. https://doi.org/10.1016/j. yexmp.2015.03.009. S0014-4800(15)00044-1 [pii] Flierman PA, Oberye JJ, van der Hulst VP, de Blok S (2005) Rapid reduction of leiomyoma volume during treatment with the GnRH antagonist ganirelix. BJOG 112:638–642. https://doi.org/10.1111/j.1471-0528.200 4.00504.x Folpe AL, Mentzel T, Lehr HA, Fisher C, Balzer BL, Weiss SW (2005) Perivascular epithelioid cell neoplasms of soft tissue and gynecologic origin: a clinicopathologic study of 26 cases and review of the literature. Am J Surg Pathol 29:1558–1575 Fornelli A, Pasquinelli G, Eusebi V (1999) Leiomyoma of the uterus showing skeletal muscle differentiation: a case report. Hum Pathol 30:356–359 Franquemont DW, Frierson HF Jr, Mills SE (1991) An immunohistochemical study of normal endometrial stroma and endometrial stromal neoplasms. Evidence for smooth muscle differentiation. Am J Surg Pathol 15:861–870 Frey NV et al (2006) Primary lymphomas of the cervix and uterus: the University of Pennsylvania’s experience and a review of the literature. Leuk Lymphoma 47:1894–1901. https://doi.org/10.1080/104281906006 87653. U8R528Q43L276436 [pii] Fuehrer NE, Keeney GL, Ketterling RP, Knudson RA, Bell DA (2012) ALK-1 protein expression and ALK gene rearrangements aid in the diagnosis of inflammatory myofibroblastic tumors of the female genital tract. Arch Pathol Lab Med 136:623–626. https://doi.org/ 10.5858/arpa.2011-0341-OA Fukunaga M (2005) Perivascular epithelioid cell tumor of the uterus: report of four cases. Int J Gynecol Pathol 24:341–346 Fukunaga M (2011) Pure alveolar rhabdomyosarcoma of the uterine corpus. Pathol Int 61:377–381. https://doi. org/10.1111/j.1440-1827.2011.02672.x Fukunaga M, Endo Y (1996) Pelvic bone involvement in low-grade endometrial stromal sarcoma with ovarian sex cord-like differentiation. Histopathology 29:391–393 Fukunaga M, Ushigome S (1998) Dissecting leiomyoma of the uterus with extrauterine extension. Histopathology 32:160–164 Fukunaga M, Miyazawa Y, Ushigome S (1997) Endometrial low-grade stromal sarcoma with ovarian sex cordlike differentiation: report of two cases with an immunohistochemical and flow cytometric study. Pathol Int 47:412–415 Gadducci A et al (1996a) Uterine leiomyosarcoma: analysis of treatment failures and survival. Gynecol Oncol 62:25–32. https://doi.org/10.1006/gyno.1996.0185 Gadducci A et al (1996b) Endometrial stromal sarcoma: analysis of treatment failures and survival. Gynecol

10

Mesenchymal Tumors of the Uterus

Oncol 63:247–253. https://doi.org/10.1006/gyno.1996. 0314. S0090-8258(96)90314-5 [pii] Gadducci A, Cosio S, Romanini A, Genazzani AR (2008) The management of patients with uterine sarcoma: a debated clinical challenge. Crit Rev Oncol Hematol 65:129–142. https://doi.org/10.1016/j. critrevonc.2007.06.011 Gal AA, Brooks JS, Pietra GG (1989) Leiomyomatous neoplasms of the lung: a clinical, histologic, and immunohistochemical study. Mod Pathol 2:209–216 Gallardo A, Prat J (2009) Mullerian adenosarcoma: a clinicopathologic and immunohistochemical study of 55 cases challenging the existence of adenofibroma. Am J Surg Pathol 33:278–288. https://doi.org/ 10.1097/PAS.0b013e318181a80d Gannon BR, Manduch M, Childs TJ (2008) Differential Immunoreactivity of p16 in leiomyosarcomas and leiomyoma variants. Int J Gynecol Pathol 27:68–73. https://doi.org/10.1097/pgp.0b013e3180ca954f Garcia MG, Deavers MT, Knoblock RJ, Chen W, Tsimberidou AM, Manning JT Jr, Medeiros LJ (2006) Myeloid sarcoma involving the gynecologic tract: a report of 11 cases and review of the literature. Am J Clin Pathol 125:783–790 Gard GB, Mulvany NJ, Quinn MA (1999) Management of uterine leiomyosarcoma in Australia. Aust N Z J Obstet Gynaecol 39:93–98 Gardie B et al (2011) Novel FH mutations in families with hereditary leiomyomatosis and renal cell cancer (HLRCC) and patients with isolated type 2 papillary renal cell carcinoma. J Med Genet 48:226–234. https:// doi.org/10.1136/jmg.2010.085068 Garg K, Tickoo SK, Soslow RA, Reuter VE (2011) Morphologic features of uterine leiomyomas associated with hereditary leiomyomatosis and renal cell carcinoma syndrome: a case report. Am J Surg Pathol 35:1235–1237. https://doi.org/10.1097/PAS.0 b013e318223ca01 Gersell DJ, Fulling KH (1989) Localized neurofibromatosis of the female genitourinary tract. Am J Surg Pathol 13:873–878 Geyer JT, Ferry JA, Harris NL, Young RH, Longtine JA, Zukerberg LR (2010) Florid reactive lymphoid hyperplasia of the lower female genital tract (lymphoma-like lesion): a benign condition that frequently harbors clonal immunoglobulin heavy chain gene rearrangements. Am J Surg Pathol 34:161–168. https://doi.org/10.1097/PAS.0b013e3181cc4f12 Gil-Benso R, Lopez-Gines C, Navarro S, Carda C, Llombart-Bosch A (1999) Endometrial stromal sarcomas: immunohistochemical, electron microscopical and cytogenetic findings in two cases. Virchows Arch 434:307–314 Gilks CB, Taylor GP, Clement PB (1987) Inflammatory pseudotumor of the uterus. Int J Gynecol Pathol 6:275–286 Gilks CB, Young RH, Clement PB, Hart WR, Scully RE (1996) Adenomyomas of the uterine cervix of of endocervical type: a report of ten cases of a benign cervical

631 tumor that may be confused with adenoma malignum [corrected]. Mod Pathol 9:220–224 Gilks CB, Clement PB, Hart WR, Young RH (2000) Uterine adenomyomas excluding atypical polypoid adenomyomas and adenomyomas of endocervical type: a clinicopathologic study of 30 cases of an underemphasized lesion that may cause diagnostic problems with brief consideration of adenomyomas of other female genital tract sites. Int J Gynecol Pathol 19:195–205 Giner F, Compan A, Monteagudo C (2012) Uterine glomeruloid hemangioma in a patient without POEMS syndrome. Int J Surg Pathol 20:407–410. https://doi.org/10.1177/1066896911429297 Gisser SD, Young I (1977) Neurilemoma-like uterine myomas: an ultrastructural reaffirmation of their non-Schwannian nature. Am J Obstet Gynecol 129:389–392 Giuntoli RL 2nd, Bristow RE (2004) Uterine leiomyosarcoma: present management. Curr Opin Oncol 16:324–327 Giuntoli RL 2nd, Metzinger DS, DiMarco CS, Cha SS, Sloan JA, Keeney GL, Gostout BS (2003) Retrospective review of 208 patients with leiomyosarcoma of the uterus: prognostic indicators, surgical management, and adjuvant therapy. Gynecol Oncol 89:460–469 Giuntoli RL 2nd, Gostout BS, DiMarco CS, Metzinger DS, Keeney GL (2007) Diagnostic criteria for uterine smooth muscle tumors: leiomyoma variants associated with malignant behavior. J Reprod Med 52:1001–1010 Goldberg MF, Hurt WG, Frable WJ (1977) Leiomyomatosis peritonealis disseminata. Report of a case and review of the literature. Obstet Gynecol 49:46–52 Goldblum JR, Clement PB, Hart WR (1995) Adenomyosis with sparse glands: a potential mimic of low-grade endometrial stromal sarcoma. Am J Clin Pathol 103:218–223 Gomez-Laencina AM, Martinez Diaz F, Izquierdo Sanjuanes B, Vicente Sanchez EM, Fernandez Salmeron R, Meseguer Pena F (2012) Localized neurofibromatosis of the female genital system: a case report and review of the literature. J Obstet Gynaecol Res 38:953–956. https://doi.org/10.1111/j.1447-0756. 2011.01816.x Goncharova EA et al (2002) Tuberin regulates p70 S6 kinase activation and ribosomal protein S6 phosphorylation: a role for the TSC2 tumor suppressor gene in pulmonary lymphangioleiomyomatosis (LAM). J Biol Chem 277:30958–30967. https://doi.org/10.1074/jbc. M202678200 Goode B et al (2018) Adenomatoid tumors of the male and female genital tract are defined by TRAF7 mutations that drive aberrant NF-kB pathway activation. Mod Pathol 31:660–673. https://doi.org/10.1038/modpath ol.2017.153 Gordts S, Grimbizis G, Campo R (2018) Symptoms and classification of uterine adenomyosis, including the place of hysteroscopy in diagnosis. Fertil Steril

632 109:380–388.e381. https://doi.org/10.1016/j.fertnstert. 2018.01.006 Gown AM, Boyd HC, Chang Y, Ferguson M, Reichler B, Tippens D (1988) Smooth muscle cells can express cytokeratins of “simple” epithelium. Immunocytochemical and biochemical studies in vitro and in vivo. Am J Pathol 132:223–232 Grayson W, Fourie J, Tiltman AJ (1998) Xanthomatous leiomyosarcoma of the uterine cervix. Int J Gynecol Pathol 17:89–90 Greene LA, Mount SL, Schned AR, Cooper K (2003) Recurrent perivascular epithelioid cell tumor of the uterus (PEComa): an immunohistochemical study and review of the literature. Gynecol Oncol 90:677–681 Gremel G et al (2015) A prognosis based classification of undifferentiated uterine sarcomas: identification of mitotic index, hormone receptors and YWHAEFAM22 translocation status as predictors of survival. Int J Cancer 136:1608–1618. https://doi.org/10.1002/ ijc.29141 Grignon DJ, Carey MR, Kirk ME, Robinson ML (1987) Diffuse uterine leiomyomatosis: a case study with pregnancy complicated by intrapartum hemorrhage. Obstet Gynecol 69:477–480 Gross KL et al (2004) Involvement of fumarate hydratase in nonsyndromic uterine leiomyomas: genetic linkage analysis and FISH studies. Genes Chromosomes Cancer 41:183–190. https://doi.org/10.1002/gcc.20079 Gunnala V et al (2017) Novel fumarate hydratase mutation in siblings with early onset uterine leiomyomas and hereditary leiomyomatosis and renal cell cancer syndrome. Int J Gynecol Pathol. https://doi.org/10.1097/ pgp.0000000000000423 Guntupalli SR, Ramirez PT, Anderson ML, Milam MR, Bodurka DC, Malpica A (2009) Uterine smooth muscle tumor of uncertain malignant potential: a retrospective analysis. Gynecol Oncol 113:324–326. https://doi.org/ 10.1016/j.ygyno.2009.02.020 Guo X et al (2015) Clinically relevant molecular subtypes in leiomyosarcoma. Clin Cancer Res 21:3501–3511. https://doi.org/10.1158/1078-0432.ccr-14-3141 Gyure KA, Hart WR, Kennedy AW (1995) Lymphangiomyomatosis of the uterus associated with tuberous sclerosis and malignant neoplasia of the female genital tract: a report of two cases. Int J Gynecol Pathol 14:344–351 Haimes JD et al (2017) Uterine inflammatory myofibroblastic tumors frequently harbor ALK fusions with IGFBP5 and THBS1. Am J Surg Pathol 41:773–780. https://doi.org/10.1097/pas.00000000000 00801 Halaska MJ, Haidopoulos D, Guyon F, Morice P, Zapardiel I, Kesic V (2017) European Society of gynecological oncology statement on fibroid and uterine morcellation. Int J Gynecol Cancer 27:189–192. https://doi.org/10.1097/igc.0000000000000911 Halbwedl I et al (2005) Chromosomal alterations in low-grade endometrial stromal sarcoma and undifferentiated endometrial sarcoma as detected by

E. Oliva et al. comparative genomic hybridization. Gynecol Oncol 97:582–587. https://doi.org/10.1016/j.ygyno.2005.01. 002. S0090-8258(05)00022-3 [pii] Hales HA, Peterson CM, Jones KP, Quinn JD (1992) Leiomyomatosis peritonealis disseminata treated with a gonadotropin-releasing hormone agonist. A case report. Am J Obstet Gynecol 167:515–516 Hall KL et al (1997) Analysis of Ki-ras, p53, and MDM2 genes in uterine leiomyomas and leiomyosarcomas. Gynecol Oncol 65:330–335. https://doi.org/10.1006/ gyno.1997.4653 Hall T, Lee SI, Boruta DM, Goodman A (2015) Medical device safety and surgical dissemination of unrecognized uterine malignancy: morcellation in minimally invasive gynecologic surgery. Oncologist 20:1274–1282. https://doi.org/10.1634/theoncologist. 2015-0061 Han HS, Park IA, Kim SH, Lee HP (1998) The clear cell variant of epithelioid intravenous leiomyomatosis of the uterus: report of a case. Pathol Int 48:892–896 Harris NL, Scully RE (1984) Malignant lymphoma and granulocytic sarcoma of the uterus and vagina. A clinicopathologic analysis of 27 cases. Cancer 53:2530–2545 Hart WR, Yoonessi M (1977) Endometrial stromatosis of the uterus. Obstet Gynecol 49:393–403 Hattab EM, Allam-Nandyala P, Rhatigan RM (1999) The stromal component of large endometrial polyps. Int J Gynecol Pathol 18:332–337 Hayashi T et al (2011) Prevalence of uterine and adnexal involvement in pulmonary lymphangioleiomyomatosis: a clinicopathologic study of 10 patients. Am J Surg Pathol 35:1776–1785. https://doi.org/ 10.1097/PAS.0b013e318235edbd Heatley MK (2006) Atypical polypoid adenomyoma: a systematic review of the English literature. Histopathology 48:609–610 Heerema-McKenney A et al (2008) Diffuse myogenin expression by immunohistochemistry is an independent marker of poor survival in pediatric rhabdomyosarcoma: a tissue microarray study of 71 primary tumors including correlation with molecular phenotype. Am J Surg Pathol 32:1513–1522 Hemming ML, Wagner AJ, Nucci MR, Chiang S, Wang L, Hensley ML, George S (2017) YWHAE-rearranged high-grade endometrial stromal sarcoma: two-center case series and response to chemotherapy. Gynecol Oncol 145:531–535. https://doi.org/10.1016/j.ygyno. 2017.03.021. S0090-8258(17)30237-8 [pii] Hendrickson MR, Tavassoli F, Kempson RL, McCluggage G, Hailer U, Kubik-Huch RA (2003) Mesenchymal tumours and related lesions. In: Tavassoli F, Devilee P (eds) World Health Organization classification of tumours. Pathology and genetics of tumors of breast and female genital organs. IARC Press, Lyon, pp 233–244 Hennig Y, Caselitz J, Bartnitzke S, Bullerdiek J (1997) A third case of a low-grade endometrial stromal sarcoma with a t(7;17)(p14 approximately 21;q11.2 approximately 21). Cancer Genet Cytogenet 98:84–86

10

Mesenchymal Tumors of the Uterus

Hensley ML (2017) Difficult choices in stage I uterine leiomyosarcoma- it’s okay to “stand there”. Gynecol Oncol 147:1–2. https://doi.org/10.1016/j. ygyno.2017.09.004 Hensley ML et al (2002) Gemcitabine and docetaxel in patients with unresectable leiomyosarcoma: results of a phase II trial. J Clin Oncol 20:2824–2831. https://doi. org/10.1200/jco.2002.11.050 Hensley ML, Blessing JA, DeGeest K, Abulafia O, Rose PG, Homesley HD (2008a) Fixed-dose rate gemcitabine plus docetaxel as second-line therapy for metastatic uterine leiomyosarcoma: a Gynecologic Oncology Group phase II study. Gynecol Oncol 109:323–328. https:// doi.org/10.1016/j.ygyno.2008.02.024 Hensley ML, Blessing JA, Mannel R, Rose PG (2008b) Fixed-dose rate gemcitabine plus docetaxel as firstline therapy for metastatic uterine leiomyosarcoma: a Gynecologic Oncology Group phase II trial. Gynecol Oncol 109:329–334. https://doi.org/10. 1016/j.ygyno.2008.03.010 Hillard JB, Malpica A, Ramirez PT (2004) Conservative management of a uterine tumor resembling an ovarian sex cord-stromal tumor. Gynecol Oncol 92: 347–352 Hilsenbeck SG, Allred DC (1992) Improved methods of estimating mitotic activity in solid tumors. Hum Pathol 23:601–602 Hirschfield L, Kahn LB, Chen S, Winkler B, Rosenberg S (1986) Mullerian adenosarcoma with ovarian sex cordlike differentiation. A light- and electron-microscopic study. Cancer 57:1197–1200 Hirschowitz L, Mayall FG, Ganesan R, McCluggage WG (2013) Intravascular adenomyomatosis: expanding the morphologic spectrum of intravascular leiomyomatosis. Am J Surg Pathol 37:1395–1400. https://doi.org/10.1 097/PAS.0b013e31828b2c99. 00000478-20130900000013 [pii] Hoang LN et al (2017) Novel high-grade endometrial stromal sarcoma: a morphologic mimicker of myxoid leiomyosarcoma. Am J Surg Pathol 41:12–24. https:// doi.org/10.1097/pas.0000000000000721 Hoang L, Chiang S, Lee CH (2018) Endometrial stromal sarcomas and related neoplasms: new developments and diagnostic considerations. Pathology 50:162–177. https://doi.org/10.1016/j.pathol.2017.11.086. S00313025(17)30393-8 [pii] Hodgson A, Amemiya Y, Seth A, Djordjevic B, Parra-Herran C (2017) High-grade mullerian adenosarcoma: genomic and clinicopathologic characterization of a distinct neoplasm with prevalent TP53 pathway alterations and aggressive behavior. Am J Surg Pathol 41:1513–1522. https://doi.org/10.1097/pas.0000000000000907 Holzmann C, Markowski DN, Bartnitzke S, Koczan D, Helmke BM, Bullerdiek J (2015) A rare coincidence of different types of driver mutations among uterine leiomyomas (UL). Mol Cytogenet 8:76. https://doi.org/ 10.1186/s13039-015-0177-9 Horie Y, Ikawa S, Kadowaki K, Minagawa Y, Kigawa J, Terakawa N (1995) Lipoadenofibroma of the uterine

633 corpus: report of a new variant of adenofibroma (benign mÅllerian mixed tumor). Arch Pathol Lab Med 119:274–276 Hornick JL, Fletcher CD (2008) Sclerosing PEComa: clinicopathologic analysis of a distinctive variant with a predilection for the retroperitoneum. Am J Surg Pathol 32:493–501. https://doi.org/10.1097/ PAS.0b013e318161dc34 Horstmann JP, Pietra GG, Harman JA, Cole NG, Grinspan S (1977) Spontaneous regression of pulmonary leiomyomas during pregnancy. Cancer 39:314–321 Houghton JP, McCluggage WG (2007) Embryonal rhabdomyosarcoma of the cervix with focal pleomorphic areas. J Clin Pathol 60:88–89 Howitt BE, Quade BJ, Nucci MR (2015a) Uterine polyps with features overlapping with those of mullerian adenosarcoma: a clinicopathologic analysis of 29 cases emphasizing their likely benign nature. Am J Surg Pathol 39:116–126. https://doi.org/10.1097/ pas.0000000000000303 Howitt BE et al (2015b) Targeted genomic analysis of mullerian adenosarcoma. J Pathol 235:37–49. https:// doi.org/10.1002/path.4442 Howitt BE, Dal Cin P, Nucci MR, Quade BJ (2016) Involvement of chromosome 8 in mullerian adenosarcoma. Int J Gynecol Pathol. https://doi.org/ 10.1097/PGP.0000000000000287 Hrzenjak A (2016) JAZF1/SUZ12 gene fusion in endometrial stromal sarcomas. Orphanet J Rare Dis 11:15. https://doi.org/10.1186/s13023-016-0400-8 Huang HY, Ladanyi M, Soslow RA (2004) Molecular detection of JAZF1-JJAZ1 gene fusion in endometrial stromal neoplasms with classic and variant histology: evidence for genetic heterogeneity. Am J Surg Pathol 28:224–232 Hung YP, Fletcher CD, Hornick JL (2016) Evaluation of NKX2-2 expression in round cell sarcomas and other tumors with EWSR1 rearrangement: imperfect specificity for Ewing sarcoma. Mod Pathol 29:370–380. https://doi.org/10.1038/modpathol.2016.31 Hurrell DP, McCluggage WG (2005) Uterine leiomyosarcoma with HMB45+ clear cell areas: report of two cases. Histopathology 47:540–542 Hurrell DP, McCluggage WG (2007) Uterine tumour resembling ovarian sex cord tumour is an immunohistochemically polyphenotypic neoplasm which exhibits coexpression of epithelial, myoid and sex cord markers. J Clin Pathol 60:1148–1154. https://doi.org/10.1136/ jcp.2006.044842 Hyde KE, Geisinger KR, Marshall RB, Jones TL (1989) The clear-cell variant of uterine epithelioid leiomyoma. An immunohistologic and ultrastructural study. Arch Pathol Lab Med 113:551–553 Hyman DM, Grisham RN, Hensley ML (2014) Management of advanced uterine leiomyosarcoma. Curr Opin Oncol 26:422–427. https://doi.org/10.1097/cco. 0000000000000094 Iasonos A et al (2013) External validation of a prognostic nomogram for overall survival in women with uterine

634 leiomyosarcoma. Cancer 119:1816–1822. https://doi. org/10.1002/cncr.27971 Ip PP, Cheung AN, Clement PB (2009) Uterine smooth muscle tumors of uncertain malignant potential (STUMP): a clinicopathologic analysis of 16 cases. Am J Surg Pathol 33:992–1005. https://doi.org/ 10.1097/PAS.0b013e3181a02d1c Irving JA, Carinelli S, Prat J (2006) Uterine tumors resembling ovarian sex cord tumors are polyphenotypic neoplasms with true sex cord differentiation. Mod Pathol 19:17–24. https://doi.org/10.1038/modpathol. 3800475 Ito H, Sasaki N, Miyagawa K, Tahara E (1986) Bizarre leiomyoblastoma of the cervix uteri. Immunohistochemical and ultrastructural study. Acta Pathol Jpn 36:1737–1745 Jabbour MN, Seoud M, Al-Ahmadie H, Abdul-Karim FW, Zaatari GS (2014) ASPL-TFE3 translocation in vulvovaginal alveolar soft part sarcoma. Int J Gynecol Pathol 33:263–267. https://doi.org/ 10.1097/PGP.0b013e318290407c Jakate K et al (2013) Endometrial sarcomas: an immunohistochemical and JAZF1 re-arrangement study in low-grade and undifferentiated tumors. Mod Pathol 26:95–105. https://doi.org/10.1038/modpathol.2012.136 Jautzke G, Muller-Ruchholtz E, Thalmann U (1996) Immunohistological detection of estrogen and progesterone receptors in multiple and well differentiated leiomyomatous lung tumors in women with uterine leiomyomas (so-called benign metastasizing leiomyomas). A report on 5 cases. Pathol Res Pract 192:215–223. https://doi.org/10.1016/s0344-0338(96) 80224-x Jones MW, Lefkowitz M (1995) Adenosarcoma of the uterine cervix: a clinicopathological study of 12 cases. Int J Gynecol Pathol 14:223–229 Jones MW, Norris HJ (1995) Clinicopathologic study of 28 uterine leiomyosarcomas with metastasis. Int J Gynecol Pathol 14:243–249 Joseph NM, Solomon DA, Frizzell N, Rabban JT, Zaloudek C, Garg K (2015) Morphology and immunohistochemistry for 2SC and FH aid in detection of fumarate hydratase gene aberrations in uterine leiomyomas from young patients. Am J Surg Pathol 39:1529–1539. https://doi.org/10.1097/pas.00000000 00000520 Jung CK, Jung JH, Lee A, Lee YS, Choi YJ, Yoon SK, Lee KY (2008) Diagnostic use of nuclear beta-catenin expression for the assessment of endometrial stromal tumors. Mod Pathol 21:756–763. https://doi.org/ 10.1038/modpathol.2008.53 Kaku T, Silverberg SG, Major FJ, Miller A, Fetter B, Brady MF (1992) Adenosarcoma of the uterus: a Gynecologic Oncology Group clinicopathologic study of 31 cases. Int J Gynecol Pathol 11:75–88 Kaminski PF, Tavassoli FA (1984) Plexiform tumorlet: a clinical and pathologic study of 15 cases with ultrastructural observations. Int J Gynecol Pathol 3:124–134

E. Oliva et al. Karpathiou G, Dal P, Chauleur C, Peoc’h M (2018) Intravascular lymphocytic accumulation in the uterine cervix: a frequent finding. Int J Gynecol Pathol 37:229–232. https://doi.org/10.1097/pgp.0000000000000407 Kasashima S, Kobayashi M, Yamada M, Oda Y (2003) Myxoid endometrial stromal sarcoma of the uterus. Pathol Int 53:637–641. 1525 [pii] Kasashima S, Minato H, Kobayashi M, Ueda Y, Oda Y, Hashimoto S, Inoue M (2007) Alveolar soft part sarcoma of the endometrium with expression of CD10 and hormone receptors. APMIS 115:861–865. https://doi. org/10.1111/j.1600-0463.2007.apm_635.x Kawaguchi K, Fujii S, Konishi I, Nanbu Y, Nonogaki H, Mori T (1989) Mitotic activity in uterine leiomyomas during the menstrual cycle. Am J Obstet Gynecol 160:637–641 Keel SB, Clement PB, Prat J, Young RH (1998) Malignant schwannoma of the uterine cervix: a study of three cases. Int J Gynecol Pathol 17:223–230 Kempson RL, Hendrickson MR (2000) Smooth muscle, endometrial stromal, and mixed mullerian tumors of the uterus. Mod Pathol 13:328–342. https://doi.org/ 10.1038/modpathol.3880055 Kenny SL, McCluggage WG (2014) Adenomyomatous polyp of the endometrium with prominent epithelioid smooth muscle differentiation: report of two cases of a hitherto undescribed lesion. Int J Surg Pathol 22:358–363. https://doi.org/10.1177/1066896913499630 Kho KA, Lin K, Hechanova M, Richardson DL (2016) Risk of occult uterine sarcoma in women undergoing hysterectomy for benign indications. Obstet Gynecol 127:468–473. https://doi.org/10.1097/aog. 0000000000001242 Kibar Y, Aydin A, Deniz H, Balat O, Cebesoy B, Al-Nafussi A (2008) A rare case of low-grade endometrial stromal sarcoma with myxoid differentiation and atypical bizarre cells. Eur J Gynaecol Oncol 29: 397–398 Kim YH, Cho H, Kyeom-Kim H, Kim I (1996) Uterine endometrial stromal sarcoma with rhabdoid and smooth muscle differentiation. J Korean Med Sci 11:88–93 Kim HS, Yoon G, Jung YY, Lee YY, Song SY (2015) Fibromyxoid variant of endometrial stromal sarcoma with atypical bizarre nuclei. Int J Clin Exp Pathol 8:3316–3321 King ME, Dickersin GR, Scully RE (1982) Myxoid leiomyosarcoma of the uterus. A report of six cases. Am J Surg Pathol 6:589–598 Kjerulff KH, Langenberg P, Seidman JD, Stolley PD, Guzinski GM (1996) Uterine leiomyomas. Racial differences in severity, symptoms and age at diagnosis. J Reprod Med 41:483–490 Koike N, Tsunemi T, Uekuri C, Akasaka J, Ito F, Shigemitsu A, Kobayashi H (2013) Pathogenesis and malignant transformation of adenomyosis (review). Oncol Rep 29:861–867. https://doi.org/10.3892/ or.2012.2184 Koivisto-Korander R, Butzow R, Koivisto A, Leminen A (2008) Clinical outcome and prognostic factors in

10

Mesenchymal Tumors of the Uterus

100 cases of uterine sarcoma: experience in Helsinki University Central Hospital 1990–2001. Gynecol Oncol 111:74–81 Kokawa K, Yamoto M, Yata C, Mabuchi Y, Umesaki N (2002) Postmenopausal intravenous leiomyomatosis with high levels of estradiol and estrogen receptor. Obstet Gynecol 100:1124–1126 Kolin DL et al (2018) SMARCA4-deficient undifferentiated uterine sarcoma (malignant rhabdoid tumor of the uterus): a clinicopathologic entity distinct from undifferentiated carcinoma. Mod Pathol. https:// doi.org/10.1038/s41379-018-0049-z Koontz JI et al (2001) Frequent fusion of the JAZF1 and JJAZ1 genes in endometrial stromal tumors. Proc Natl Acad Sci U S A 98:6348–6353. https://doi.org/ 10.1073/pnas.101132598. 98/11/6348. [pii] Kosari F, Daneshbod Y, Parwaresch R, Krams M, Wacker HH (2005) Lymphomas of the female genital tract: a study of 186 cases and review of the literature. Am J Surg Pathol 29:1512–1520 Koshiyama M et al (2002) Adenocarcinomas arising from uterine adenomyosis: a report of four cases. Int J Gynecol Pathol 21:239–245 Kriseman ML, Wang WL, Sullinger J, Schmeler KM, Ramirez PT, Herzog CE, Frumovitz M (2012) Rhabdomyosarcoma of the cervix in adult women and younger patients. Gynecol Oncol 126:351–356. https://doi. org/10.1016/j.ygyno.2012.05.008 Krivak TC, Seidman JD, McBroom JW, MacKoul PJ, Aye LM, Rose GS (2001) Uterine adenosarcoma with sarcomatous overgrowth versus uterine carcinosarcoma: comparison of treatment and survival. Gynecol Oncol 83:89–94 Kruse AJ, Croce S, Kruitwagen RF, Riedl RG, Slangen BF, Van Gorp T, Van de Vijver KK (2014a) Aggressive behavior and poor prognosis of endometrial stromal sarcomas with YWHAE-FAM22 rearrangement indicate the clinical importance to recognize this subset. Int J Gynecol Cancer 24:1616–1622. https://doi.org/ 10.1097/IGC.0000000000000278 Kruse AJ, Sep S, Slangen BF, Vandevijver NM, Van Gorp T, Kruitwagen RF, Van de Vijver KK (2014b) Angiosarcomas of primary gynecologic origin: a clinicopathologic review and quantitative analysis of survival. Int J Gynecol Cancer 24:4–12. https://doi.org/ 10.1097/igc.0000000000000020 Kunzel KE, Mills NZ, Muderspach LI, d’Ablaing G 3rd (1993) Myxoid leiomyosarcoma of the uterus. Gynecol Oncol 48:277–280. https://doi.org/10.1006/ gyno.1993.1048 Kurihara S et al (2008) Endometrial stromal sarcomas and related high-grade sarcomas: immunohistochemical and molecular genetic study of 31 cases. Am J Surg Pathol 32:1228–1238. https://doi.org/10.1097/ PAS.0b013e31816a3b42 Kurihara S, Oda Y, Ohishi Y, Kaneki E, Kobayashi H, Wake N, Tsuneyoshi M (2010) Coincident expression of beta-catenin and cyclin D1 in endometrial stromal tumors and related high-grade sarcomas.

635 Mod Pathol 23:225–234. https://doi.org/10.1038/ modpathol.2009.162 Kurman RJ, Norris HJ (1976) Mesenchymal tumors of the uterus. VI. Epithelioid smooth muscle tumors including leiomyoblastoma and clear-cell leiomyoma: a clinical and pathologic analysis of 26 cases. Cancer 37: 1853–1865 Kurman RJ, Carcangiu ML, Herrington CS, Young RH (2014) WHO classification of tumours of female reproductive organs. World Health Organization classification of tumours, 4th edn. IARC Press, Lyon Ladanyi M et al (2001) The der(17)t(X;17)(p11;q25) of human alveolar soft part sarcoma fuses the TFE3 transcription factor gene to ASPL, a novel gene at 17q25. Oncogene 20:48–57. https://doi.org/10.1038/sj.onc. 1204074 Lai FM, Wong FW, Allen PW (1991) Diffuse uterine leiomyomatosis with hemorrhage. Arch Pathol Lab Med 115:834–837 Lakhman Y et al (2017) Differentiation of uterine leiomyosarcoma from atypical leiomyoma: diagnostic accuracy of qualitative MR imaging features and feasibility of texture analysis. Eur Radiol 27:2903–2915. https://doi.org/10.1007/s00330-016-4623-9 Langlois PL (1970) The size of the normal uterus. J Reprod Med 4:220–228 Larson B, Silfversward C, Nilsson B, Pettersson F (1990) Prognostic factors in uterine leiomyosarcoma. A clinical and histopathological study of 143 cases. The Radiumhemmet series 1936–1981. Acta Oncol 29: 185–191 Launonen V et al (2001) Inherited susceptibility to uterine leiomyomas and renal cell cancer. Proc Natl Acad Sci U S A 98:3387–3392. https://doi.org/ 10.1073/pnas.051633798 Layfield LJ, Liu K, Dodge R, Barsky SH (2000) Uterine smooth muscle tumors: utility of classification by proliferation, ploidy, and prognostic markers versus traditional histopathology. Arch Pathol Lab Med 124:221–227. https://doi.org/10.1043/0003-9985(2000)1242.0.co;2 Leath CA 3rd et al (2007) A multi-institutional review of outcomes of endometrial stromal sarcoma. Gynecol Oncol 105:630–634. https://doi.org/10.1016/ j.ygyno.2007.01.031. S0090-8258(07)00056-X [pii] Lee HJ, Choi J, Kim KR (2008) Pulmonary benign metastasizing leiomyoma associated with intravenous leiomyomatosis of the uterus: clinical behavior and genomic changes supporting a transportation theory. Int J Gynecol Pathol 27:340–345. https://doi.org/ 10.1097/PGP.0b013e3181656dab Lee CH, Turbin DA, Sung YC, Espinosa I, Montgomery K, van de Rijn M, Gilks CB (2009) A panel of antibodies to determine site of origin and malignancy in smooth muscle tumors. Mod Pathol 22:1519–1531. https://doi. org/10.1038/modpathol.2009.122 Lee CH et al (2012a) Cyclin D1 as a diagnostic immunomarker for endometrial stromal sarcoma with YWHAE-FAM22 rearrangement. Am J Surg Pathol 3

636 6:1562–1570. https://doi.org/10.1097/PAS.0b013e3 1825fa931. 00000478-201210000-00018 [pii] Lee CH et al (2012b) The clinicopathologic features of YWHAE-FAM22 endometrial stromal sarcomas: a histologically high-grade and clinically aggressive tumor. Am J Surg Pathol 36:641–653. https://doi.org/10.1097/ PAS.0b013e31824a7b1a Lee CH et al (2012c) 14-3-3 fusion oncogenes in high-grade endometrial stromal sarcoma. Proc Natl Acad Sci U S A 109:929–934. https://doi.org/10.1073/pnas.1115528109 Lee CH et al (2014) Frequent expression of KIT in endometrial stromal sarcoma with YWHAE genetic rearrangement. Mod Pathol 27:751–757. https://doi. org/10.1038/modpathol.2013.199 Lee JC et al (2016) Genomewide copy number analysis of mullerian adenosarcoma identified chromosomal instability in the aggressive subgroup. Mod Pathol 29: 1070–1082. https://doi.org/10.1038/modpathol.2016.99 Lehtonen R et al (2004) Biallelic inactivation of fumarate hydratase (FH) occurs in nonsyndromic uterine leiomyomas but is rare in other tumors. Am J Pathol 164:17–22. https://doi.org/10.1016/s0002-9440(10) 63091-x Lehtonen HJ et al (2006) Increased risk of cancer in patients with fumarate hydratase germline mutation. J Med Genet 43:523–526. https://doi.org/10.1136/ jmg.2005.036400 Leibsohn S, d’Ablaing G, Mishell DR Jr, Schlaerth JB (1990) Leiomyosarcoma in a series of hysterectomies performed for presumed uterine leiomyomas. Am J Obstet Gynecol 162:968–974; discussion 974-966 Leiser AL, Anderson SE, Nonaka D, Chuai S, Olshen AB, Chi DS, Soslow RA (2006) Apoptotic and cell cycle regulatory markers in uterine leiomyosarcoma. Gynecol Oncol 101:86–91. https://doi.org/10.1016/j. ygyno.2005.09.055 Leitao MM Jr et al (2012) Immunohistochemical expression of estrogen and progesterone receptors and outcomes in patients with newly diagnosed uterine leiomyosarcoma. Gynecol Oncol 124:558–562. https://doi.org/10.1016/j.ygyno.2011.11.009 Leitao MM, Sonoda Y, Brennan MF, Barakat RR, Chi DS (2003) Incidence of lymph node and ovarian metastases in leiomyosarcoma of the uterus. Gynecol Oncol 91:209–212 Leitao MM et al (2004) Tissue microarray immunohistochemical expression of estrogen, progesterone, and androgen receptors in uterine leiomyomata and leiomyosarcoma. Cancer 101:1455–1462. https://doi. org/10.1002/cncr.20521 Lewis N et al (2017) ZC3H7B-BCOR high-grade endometrial stromal sarcomas: a report of 17 cases of a newly defined entity. Mod Pathol. https://doi.org/10.1038/ modpathol.2017.162 Lewis N et al (2018) ZC3H7B-BCOR high-grade endometrial stromal sarcomas: a report of 17 cases of a newly defined entity. Mod Pathol 31:674–684. https://doi.org/ 10.1038/modpathol.2017.162

E. Oliva et al. Li N, Wu LY, Zhang HT, An JS, Li XG, Ma SK (2008) Treatment options in stage I endometrial stromal sarcoma: a retrospective analysis of 53 cases. Gynecol Oncol 108:306–311. https://doi.org/10.1016/ j.ygyno.2007.10.023. S0090-8258(07)00863-3 [pii] Li RF, Gupta M, McCluggage WG, Ronnett BM (2013) Embryonal rhabdomyosarcoma (botryoid type) of the uterine corpus and cervix in adult women: report of a case series and review of the literature. Am J Surg Pathol 37:344–355. https://doi.org/10.1097/ PAS.0b013e31826e0271 Liang SX, Patel K, Pearl M, Liu J, Zheng W, Tornos C (2007) Sertoliform endometrioid carcinoma of the endometrium with dual immunophenotypes for epithelial membrane antigen and inhibin alpha: case report and literature review. Int J Gynecol Pathol 26:291–297. https://doi.org/10.1097/01.pgp.0000236948.67087.56. 00004347-200707000-00017 [pii] Lifschitz-Mercer B, Czernobilsky B, Dgani R, DallenbachHellweg G, Moll R, Franke WW (1987) Immunocytochemical study of an endometrial diffuse clear cell stromal sarcoma and other endometrial stromal sarcomas. Cancer 59:1494–1499 Lillemoe TJ, Perrone T, Norris HJ, Dehner LP (1991) Myogenous phenotype of epithelial-like areas in endometrial stromal sarcomas. Arch Pathol Lab Med 115:215–219 Lim D, Alvarez T, Nucci MR, Gilks B, Longacre T, Soslow RA, Oliva E (2013) Interobserver variability in the interpretation of tumor cell necrosis in uterine leiomyosarcoma. Am J Surg Pathol 37:650–658. https://doi.org/10.1097/PAS.0b013e3182851162 Littell RD et al (2017) Adjuvant gemcitabine-docetaxel chemotherapy for stage I uterine leiomyosarcoma: trends and survival outcomes. Gynecol Oncol 147:11–17. https://doi.org/10.1016/j.ygyno.2017.07.122 Liu Y, Guo S, Wang L, Suzuki S, Sugimura H, Li Y (2016) Uterine angiosarcoma: a case report and literature review. Int J Gynecol Pathol 35:264–268. https://doi. org/10.1097/pgp.0000000000000219 Lloreta J, Prat J (1992) Endometrial stromal nodule with smooth and skeletal muscle components simulating stromal sarcoma. Int J Gynecol Pathol 11:293–298 Lloreta J, Prat J (1993) Ultrastructure of an endometrial stromal nodule with skeletal muscle. Ultrastruct Pathol 17:405–410 Lobel MK, Somasundaram P, Morton CC (2006) The genetic heterogeneity of uterine leiomyomata. Obstet Gynecol Clin N Am 33:13–39. https://doi.org/10.1016/ j.ogc.2005.12.006 Loddenkemper C, Mechsner S, Foss HD, Dallenbach FE, Anagnostopoulos I, Ebert AD, Stein H (2003) Use of oxytocin receptor expression in distinguishing between uterine smooth muscle tumors and endometrial stromal sarcoma. Am J Surg Pathol 27:1458–1462 Longacre TA, Chung MH, Rouse RV, Hendrickson MR (1996) Atypical polypoid adenomyofibromas (atypical polypoid adenomyomas) of the uterus – a clinicopathologic study of 55 cases. Am J Surg Pathol 20:1–20

10

Mesenchymal Tumors of the Uterus

Longacre TA, Hendrickson MR, Kempson RL (1997) Predicting clinical outcome for uterine smooth muscle neoplasms with a reasonable degree of certainty. Adv Anat Pathol 4:95–104 Lotgering FK, Pijpers L, van Eijck J, Wallenburg HC (1989) Pregnancy in a patient with diffuse cavernous hemangioma of the uterus. Am J Obstet Gynecol 160:628–630 Ly A, Mills AM, McKenney JK, Balzer BL, Kempson RL, Hendrickson MR, Longacre TA (2013) Atypical leiomyomas of the uterus: a clinicopathologic study of 51 cases. Am J Surg Pathol 37:643–649. https://doi.org/10.1097/PAS.0b013e31 82893f36 Ma J, Shi QL, Zhou XJ, Meng K, Chen JY, Huang WB (2007) Lymphoma-like lesion of the uterine cervix: report of 12 cases of a rare entity. Int J Gynecol Pathol 26:194–198 Machida H, Nathenson MJ, Takiuchi T, Adams CL, Garcia-Sayre J, Matsuo K (2017) Significance of lymph node metastasis on survival of women with uterine adenosarcoma. Gynecol Oncol 144:524–530. https://doi.org/10.1016/j.ygyno.2017.01.012 Majmudar B, Ghanee N, Horowitz IR, Graham D (1998) Uterine arteriovenous malformation necessitating hysterectomy with bilateral salpingo-oophorectomy in a young pregnant patient. Arch Pathol Lab Med 122: 842–845 Major FJ et al (1993) Prognostic factors in early-stage uterine sarcoma. A Gynecologic Oncology Group study. Cancer 71:1702–1709 Makinen N et al (2016) Exome sequencing of uterine leiomyosarcomas identifies frequent mutations in TP53, ATRX, and MED12. PLoS Genet 12:e1005850. https://doi.org/10.1371/journal.pgen.1005850 Malpica A, Deavers MT, Silva EG (2006) High grade sarcoma in endometrial stromal sarcoma: dedifferentiated endometrial stromal sarcoma [Abstract]. Modern Pathol 19:188A Maluf HM, Gersell DJ (1994) Uterine leiomyomas with high content of mast cells. Arch Pathol Lab Med 118:712–714 Mansi JL, Ramachandra S, Wiltshaw E, Fisher C (1990) Endometrial stromal sarcomas. Gynecol Oncol 36:113–118 Marino-Enriquez A et al (2018) BCOR internal tandem duplication in high-grade uterine sarcomas. Am J Surg Pathol 42:335–341. https://doi.org/10.1097/ pas.0000000000000993 Markowski DN, Bartnitzke S, Loning T, Drieschner N, Helmke BM, Bullerdiek J (2012) MED12 mutations in uterine fibroids – their relationship to cytogenetic subgroups. Int J Cancer 131:1528–1536. https://doi. org/10.1002/ijc.27424 Markowski DN, Holzmann C, Bullerdiek J (2015) Genetic alterations in uterine fibroids – a new direction for pharmacological intervention? Expert Opin Ther Targets 19:1485–1494. https://doi.org/10.1517/ 14728222.2015.1075510

637 Marsh EE, Bulun SE (2006) Steroid hormones and leiomyomas. Obstet Gynecol Clin N Am 33:59–67. https://doi.org/10.1016/j.ogc.2005.12.001 Marshall RJ, Braye SG, Jones DB (1986) Leiomyosarcoma of the uterus with giant cells resembling osteoclasts. Int J Gynecol Pathol 5:260–268 Marshburn PB, Matthews ML, Hurst BS (2006) Uterine artery embolization as a treatment option for uterine myomas. Obstet Gynecol Clin N Am 33:125–144. https://doi.org/10.1016/j.ogc.2005.12.009 Martignoni G, Pea M, Reghellin D, Zamboni G, Bonetti F (2008) PEComas: the past, the present and the future. Virchows Arch 452:119–132. https://doi.org/10.1007/ s00428-007-0509-1 Martin KA, Rzucidlo EM, Merenick BL, Fingar DC, Brown DJ, Wagner RJ, Powell RJ (2004) The mTOR/p70 S6K1 pathway regulates vascular smooth muscle cell differentiation. Am J Physiol Cell Physiol 286:C507–C517. https://doi.org/10.1152/ajpcell.002 01.2003 Martin-Reay DG, Christ ML, LaPata RE (1991) Uterine leiomyoma with skeletal muscle differentiation. Report of a case. Am J Clin Pathol 96:344–347 Matsubara A et al (2013) Prevalence of MED12 mutations in uterine and extrauterine smooth muscle tumours. Histopathology 62:657–661. https://doi.org/10.1111/ his.12039 Mayerhofer K et al (1999) Leiomyosarcoma of the uterus: a clinicopathologic multicenter study of 71 cases. Gynecol Oncol 74:196–201. https://doi.org/10.1006/ gyno.1999.5436 Mazur MT, Kraus FT (1980) Histogenesis of morphologic variations in tumors of the uterine wall. Am J Surg Pathol 4:59–74 Mazur MT, Priest JB (1986) Clear cell leiomyoma (leiomyoblastoma) of the uterus: ultrastructural observations. Ultrastruct Pathol 10:249–255 McCluggage WG (2010) Mullerian adenosarcoma of the female genital tract. Adv Anat Pathol 17:122–129. https://doi.org/10.1097/PAP.0b013e3181cfe732 McCluggage WG (2016) A practical approach to the diagnosis of mixed epithelial and mesenchymal tumours of the uterus. Mod Pathol 29(Suppl 1):S78–S91. https:// doi.org/10.1038/modpathol.2015.137 McCluggage WG, Young RH (2008) Endometrial stromal sarcomas with true papillae and pseudopapillae. Int J Gynecol Pathol 27:555–561. https://doi.org/10.1097/ PGP.0b013e31817a82f9 McCluggage WG, Date A, Bharucha H, Toner PG (1996) Endometrial stromal sarcoma with sex cord-like areas and focal rhabdoid differentiation. Histopathology 29:369–374 McCluggage WG, Alderdice JM, Walsh MY (1999) Polypoid uterine lesions mimicking endometrial stromal sarcoma. J Clin Pathol 52:543–546 McCluggage WG, Ellis PK, McClure N, Walker WJ, Jackson PA, Manek S (2000) Pathologic features of uterine leiomyomas following uterine artery embolization. Int J Gynecol Pathol 19:342–347

638 McCluggage WG, Cromie AJ, Bryson C, Traub AI (2001a) Uterine endometrial stromal sarcoma with smooth muscle and glandular differentiation. J Clin Pathol 54: 481–483 McCluggage WG, Sumathi VP, Maxwell P (2001b) CD10 is a sensitive and diagnostically useful immunohistochemical marker of normal endometrial stroma and of endometrial stromal neoplasms. Histopathology 39:273–278 McCluggage WG, Ganesan R, Herrington CS (2009) Endometrial stromal sarcomas with extensive endometrioid glandular differentiation: report of a series with emphasis on the potential for misdiagnosis and discussion of the differential diagnosis. Histopathology 54:365–373. https://doi.org/10.1111/j.1365-2559.200 9.03230.x. HIS3230 [pii] McDonald AG, Dal Cin P, Ganguly A, Campbell S, Imai Y, Rosenberg AE, Oliva E (2011) Liposarcoma arising in uterine lipoleiomyoma: a report of 3 cases and review of the literature. Am J Surg Pathol 35:221–227. https:// doi.org/10.1097/PAS.0b013e31820414f7 Meenakshi M, McCluggage WG (2010) Vascular involvement in adenomyosis: report of a large series of a common phenomenon with observations on the pathogenesis of adenomyosis. Int J Gynecol Pathol 29:117–121. https://doi.org/10.1097/PGP.0b013e3181b7018d. 00004347-201003000-00004 [pii] Mehine M et al (2013) Characterization of uterine leiomyomas by whole-genome sequencing. N Engl J Med 369:43–53. https://doi.org/10.1056/NEJ Moa1302736 Mehine M, Makinen N, Heinonen HR, Aaltonen LA, Vahteristo P (2014) Genomics of uterine leiomyomas: insights from high-throughput sequencing. Fertil Steril 102:621–629. https://doi.org/10.1016/j.fertnstert.2014. 06.050 Mentrikoski M, Lastra R, Kurman RJ (2015) Mullerian adenosarcomas with rhabdomyoblastic differentiation are associated with an aggressive clinical course. Mod Pathol 28:298A Meredith RF, Eisert DR, Kaka Z, Hodgson SE, Johnston GA Jr, Boutselis JG (1986) An excess of uterine sarcomas after pelvic irradiation. Cancer 58:2003–2007 Merino MJ, Torres-Cabala C, Pinto P, Linehan WM (2007) The morphologic spectrum of kidney tumors in hereditary leiomyomatosis and renal cell carcinoma (HLRCC) syndrome. Am J Surg Pathol 31:1578–1585. https://doi. org/10.1097/PAS.0b013e31804375b8 Micci F, Walter CU, Teixeira MR, Panagopoulos I, Bjerkehagen B, Saeter G, Heim S (2003) Cytogenetic and molecular genetic analyses of endometrial stromal sarcoma: nonrandom involvement of chromosome arms 6p and 7p and confirmation of JAZF1/JJAZ1 gene fusion in t(7;17). Cancer Genet Cytogenet 144:119–124. S0165460803000256 [pii] Micci F, Panagopoulos I, Bjerkehagen B, Heim S (2006) Consistent rearrangement of chromosomal band 6p21 with generation of fusion genes JAZF1/PHF1 and EPC1/PHF1 in endometrial stromal sarcoma. Cancer

E. Oliva et al. Res 66:107–112. https://doi.org/10.1158/0008-5472. CAN-05-2485. 66/1/107 [pii] Micci F, Gorunova L, Gatius S, Matias-Guiu X, Davidson B, Heim S, Panagopoulos I (2014) MEAF6/ PHF1 is a recurrent gene fusion in endometrial stromal sarcoma. Cancer Lett 347:75–78. https://doi.org/ 10.1016/j.canlet.2014.01.030 Micci F et al (2016) Cytogenetic and molecular profile of endometrial stromal sarcoma. Genes Chromosomes Cancer 55:834–846. https://doi.org/10.1002/gcc.22380 Micci F, Brunetti M, Dal Cin P, Nucci MR, Gorunova L, Heim S, Panagopoulos I (2017) Fusion of the genes BRD8 and PHF1 in endometrial stromal sarcoma. Genes Chromosomes Cancer 56:841–845. https://doi. org/10.1002/gcc.22485 Miettinen M, Lasota J (2011) Histopathology of gastrointestinal stromal tumor. J Surg Oncol 104:865–873. https://doi.org/10.1002/jso.21945 Miettinen M, Felisiak-Golabek A, Wasag B, Chmara M, Wang Z, Butzow R, Lasota J (2016) Fumarase-deficient uterine leiomyomas: an immunohistochemical, molecular genetic, and clinicopathologic study of 86 cases. Am J Surg Pathol 40:1661–1669. https://doi.org/ 10.1097/pas.0000000000000703 Mikami Y, Hata S, Kiyokawa T, Manabe T (2002) Expression of CD10 in malignant mullerian mixed tumors and adenosarcomas: an immunohistochemical study. Mod Pathol 15:923–930. https://doi.org/10.1097/01. MP.0000026058.33869.DB Miller KN, McClure SP (1992) Papillary adenofibroma of the uterus: report of a case involved by adenocarcinoma and review of the literature. Am J Clin Pathol 97:806–809 Mills AM, Karamchandani JR, Vogel H, Longacre TA (2011) Endocervical fibroblastic malignant peripheral nerve sheath tumor (neurofibrosarcoma): report of a novel entity possibly related to endocervical CD34 fibrocytes. Am J Surg Pathol 35:404–412. https://doi.org/10.1097/PAS.0b013e31 8208f72e. 00000478-201103000-00011 [pii] Minassian SS, Frangipane W, Polin JI, Ellis M (1986) Leiomyomatosis peritonealis disseminata. A case report and literature review. J Reprod Med 31: 997–1000 Mittal K, Demopoulos RI (2001) MIB-1 (Ki-67), p53, estrogen receptor, and progesterone receptor expression in uterine smooth muscle tumors. Hum Pathol 32:984–987. https://doi.org/10.1053/hupa.2001.27113 Mittal K, Joutovsky A (2007) Areas with benign morphologic and immunohistochemical features are associated with some uterine leiomyosarcomas. Gynecol Oncol 104:362–365. https://doi.org/ 10.1016/j.ygyno.2006.08.034 Mittal K, Popiolek D, Demopoulos RI (2000) Uterine myxoid leiomyosarcoma within a leiomyoma. Hum Pathol 31:398–400 Mittal KR et al (2009) Molecular and immunohistochemical evidence for the origin of uterine leiomyosarcomas from associated leiomyoma and symplastic

10

Mesenchymal Tumors of the Uterus

leiomyoma-like areas. Mod Pathol 22:1303–1311. https://doi.org/10.1038/modpathol.2009.96 Moinfar F, Regitnig P, Tabrizi AD, Denk H, Tavassoli FA (2004) Expression of androgen receptors in benign and malignant endometrial stromal neoplasms. Virchows Arch 444:410–414. https://doi.org/10.1007/s00428004-0981-9 Moinfar F, Azodi M, Tavassoli FA (2007) Uterine sarcomas. Pathology 39:55–71. https://doi.org/10.1080/ 00313020601136146 Moore M, McCluggage WG (2017) Uterine tumour resembling ovarian sex cord tumour: first report of a large series with follow-up. Histopathology 71:751–759. https://doi.org/10.1111/his.13296 Morice P et al (2003) Prognostic value of initial surgical procedure for patients with uterine sarcoma: analysis of 123 patients. Eur J Gynaecol Oncol 24:237–240 Morotti RA et al (2006) An immunohistochemical algorithm to facilitate diagnosis and subtyping of rhabdomyosarcoma: the Children’s Oncology Group experience. Am J Surg Pathol 30:962–968 Mulvany NJ, Slavin JL, Ostor AG, Fortune DW (1994) Intravenous leiomyomatosis of the uterus: a clinicopathologic study of 22 cases. Int J Gynecol Pathol 13:1–9 Mulvany NJ, Ostor AG, Ross I (1995) Diffuse leiomyomatosis of the uterus. Histopathology 27: 175–179 Myles JL, Hart WR (1985) Apoplectic leiomyomas of the uterus. A clinicopathologic study of five distinctive hemorrhagic leiomyomas associated with oral contraceptive usage. Am J Surg Pathol 9:798–805 Nakayama H, Teramoto H, Teramoto M (2013) True incidence of uterine adenomatoid tumors. Biomed Rep 1:352–354. https://doi.org/10.3892/br.2013.72 Nasioudis D, Alevizakos M, Chapman-Davis E, Witkin SS, Holcomb K (2017a) Rhabdomyosarcoma of the lower female genital tract: an analysis of 144 cases. Arch Gynecol Obstet 296:327–334. https://doi.org/ 10.1007/s00404-017-4438-1 Nasioudis D, Kampaktsis PN, Frey M, Witkin SS, Holcomb K (2017b) Primary lymphoma of the female genital tract: an analysis of 697 cases. Gynecol Oncol 145:305–309. https://doi.org/10.1016/j.ygyno.2017.02.043 Navarro D et al (1992) Endometrial stromal sarcoma expression of estrogen receptors, progesterone receptors and estrogen-induced srp27 (24K) suggests hormone responsiveness. J Steroid Biochem Mol Biol 41:589–596 Ng TL et al (2005) Nuclear beta-catenin in mesenchymal tumors. Mod Pathol 18:68–74. https://doi.org/10.1038/ modpathol.3800272 Nielsen GP, Oliva E, Young RH, Rosenberg AE, Dickersin GR, Scully RE (1995) Alveolar soft-part sarcoma of the female genital tract: a report of nine cases and review of the literature. Int J Gynecol Pathol 14:283–292 Nogales FF Jr, Matilla A, Carrascal E (1978) Leiomyomatosis peritonealis disseminata. An ultrastructural study. Am J Clin Pathol 69:452–457

639 Nogales FF et al (1987) Uterine intravascular leiomyomatosis: an update and report of seven cases. Int J Gynecol Pathol 6:331–339 Nogales FF et al (2002) Adenomatoid tumors of the uterus: an analysis of 60 cases. Int J Gynecol Pathol 21:34–40 Nogales FF, Stolnicu S, Harilal KR, Mooney E, GarciaGalvis OF (2009) Retiform uterine tumours resembling ovarian sex cord tumours. A comparative immunohistochemical study with retiform structures of the female genital tract. Histopathology 54:471–477. https://doi. org/10.1111/j.1365-2559.2009.03244.x. HIS3244 [pii] Nola M et al (1996) Prognostic parameters for survival of patients with malignant mesenchymal tumors of the uterus. Cancer 78:2543–2550 Nordal RR, Kristensen GB, Kaern J, Stenwig AE, Pettersen EO, Trope CG (1995) The prognostic significance of stage, tumor size, cellular atypia and DNA ploidy in uterine leiomyosarcoma. Acta Oncol 34: 797–802 Nordal RR, Kristensen GB, Kaern J, Stenwig AE, Pettersen EO, Trope CG (1996) The prognostic significance of surgery, tumor size, malignancy grade, menopausal status, and DNA ploidy in endometrial stromal sarcoma. Gynecol Oncol 62:254–259. https://doi.org/10.1006/ gyno.1996.0224. S0090-8258(96)90224-3 [pii] Norris HJ, Parmley T (1975) Mesenchymal tumors of the uterus. V. Intravenous leiomyomatosis. A clinical and pathologic study of 14 cases. Cancer 36:2164–2178 Norris HJ, Taylor HB (1966) Mesenchymal tumors of the uterus. I. A clinical and pathological study of 53 endometrial stromal tumors. Cancer 19:755–766 Norris HJ, Hilliard GD, Irey NS (1988) Hemorrhagic cellular leiomyomas (“apoplectic leiomyoma”) of the uterus associated with pregnancy and oral contraceptives. Int J Gynecol Pathol 7:212–224 Novelli M et al (2010) DOG1 and CD117 are the antibodies of choice in the diagnosis of gastrointestinal stromal tumours. Histopathology 57:259–270. https:// doi.org/10.1111/j.1365-2559.2010.03624.x. HIS3624 [pii] Nucci MR (2016) Practical issues related to uterine pathology: endometrial stromal tumors. Mod Pathol 29(Suppl 1):S92–S103. https://doi.org/10.1038/mod pathol.2015.140 Nucci MR, O’Connell JT, Huettner PC, Cviko A, Sun D, Quade BJ (2001) h-Caldesmon expression effectively distinguishes endometrial stromal tumors from uterine smooth muscle tumors. Am J Surg Pathol 25:455–463 Nucci MR, Harburger D, Koontz J, Dal Cin P, Sklar J (2007) Molecular analysis of the JAZF1-JJAZ1 gene fusion by RT-PCR and fluorescence in situ hybridization in endometrial stromal neoplasms. Am J Surg Pathol 31:65–70. https://doi.org/10.1097/01.pas.0000213327.86992. d1. 00000478-200701000-00007 [pii] Nunez-Alonso C, Battifora HA (1979) Plexiform tumors of the uterus: ultrastructural study. Cancer 44:1707–1714 O’Cearbhaill R, Hensley ML (2010) Optimal management of uterine leiomyosarcoma. Expert Rev Anticancer Ther 10:153–169. https://doi.org/10.1586/era.09.187

640 O’Connor DM, Norris HJ (1990) Mitotically active leiomyomas of the uterus. Hum Pathol 21:223–227 O’Leary TJ, Steffes MW (1996) Can you count on the mitotic index? Hum Pathol 27:147–151 O’Neill CJ, McBride HA, Connolly LE, McCluggage WG (2007) Uterine leiomyosarcomas are characterized by high p16, p53 and MIB1 expression in comparison with usual leiomyomas, leiomyoma variants and smooth muscle tumours of uncertain malignant potential. Histopathology 50:851–858. https://doi.org/10.1111/ j.1365-2559.2007.02699.x Oduyebo T et al (2014) The value of re-exploration in patients with inadvertently morcellated uterine sarcoma. Gynecol Oncol 132:360–365. https://doi.org/ 10.1016/j.ygyno.2013.11.024 Ohta Y, Suzuki T, Omatsu M, Hamatani S, Shiokawa A, Kushima M, Ota H (2010) Transition from low-grade endometrial stromal sarcoma to high-grade endometrial stromal sarcoma. Int J Gynecol Pathol 29:374–377. https://doi.org/10.1097/PGP.0b013e3181cef14b. 00004347-201007000-00013 [pii] Oliva E (2014) Cellular mesenchymal tumors of the uterus: a review emphasizing recent observations. Int J Gynecol Pathol 33:374–384. https://doi.org/10.1097/ PGP.0000000000000141 Oliva E (2016) Practical issues in uterine pathology from banal to bewildering: the remarkable spectrum of smooth muscle neoplasia. Mod Pathol 29(Suppl 1):S104–S120. https://doi.org/10.1038/modpathol.2015.139 Oliva E, Young RH, Clement PB, Bhan AK, Scully RE (1995) Cellular benign mesenchymal tumors of the uterus. A comparative morphologic and immunohistochemical analysis of 33 highly cellular leiomyomas and six endometrial stromal nodules, two frequently confused tumors. Am J Surg Pathol 19:757–768 Oliva E, Ferry JA, Young RH, Prat J, Srigley JR, Scully RE (1997) Granulocytic sarcoma of the female genital tract: a clinicopathologic study of 11 cases. Am J Surg Pathol 21:1156–1165 Oliva E, Clement PB, Young RH, Scully RE (1998) Mixed endometrial stromal and smooth muscle tumors of the uterus: a clinicopathologic study of 15 cases. Am J Surg Pathol 22:997–1005 Oliva E, Young RH, Clement PB, Scully RE (1999) Myxoid and fibrous endometrial stromal tumors of the uterus: a report of 10 cases. Int J Gynecol Pathol 18:310–319 Oliva E, Clement PB, Young RH (2000) Endometrial stromal tumors: an update on a group of tumors with a protean phenotype. Adv Anat Pathol 7:257–281 Oliva E, Clement PB, Young RH (2002a) Epithelioid endometrial and endometrioid stromal tumors: a report of four cases emphasizing their distinction from epithelioid smooth muscle tumors and other oxyphilic uterine and extrauterine tumors. Int J Gynecol Pathol 21:48–55 Oliva E, Young RH, Amin MB, Clement PB (2002b) An immunohistochemical analysis of endometrial stromal and smooth muscle tumors of the uterus: a study of 54 cases emphasizing the importance of using a panel

E. Oliva et al. because of overlap in immunoreactivity for individual antibodies. Am J Surg Pathol 26:403–412 Oliva EW, Wang WL, Branton P, Logani S, Zannoni GF, Linkov I, Asher M, Soslow RA (2006) Expression of melanocytic (“PEComa”) markers in smooth muscle tumors of the uterus: an immunohistochemcial analyisis of 86 cases. Paper presented at the United States and Canadian Academy of Pathology, Atlanta Oliva E, de Leval L, Soslow RA, Herens C (2007) High frequency of JAZF1-JJAZ1 gene fusion in endometrial stromal tumors with smooth muscle differentiation by interphase FISH detection. Am J Surg Pathol 31:1277–1284. https://doi.org/10.1097/ PAS.0b013e318031f012 Oliva E, Carinelli SG, Carcangiu ML (2014) Mesenchymal tumours. Chapter 5, Tumours of the uterine corpus. In: Kurman R, Carcangiu ML, Herrington CS, Young RH (eds) WHO classification of tumours of female reproductive organs. IARC Press, Lyon, pp 135–145 Ordi J, Stamatakos MD, Tavassoli FA (1997) Pure pleomorphic rhabdomyosarcomas of the uterus. Int J Gynecol Pathol 16:369–377 Orii A, Mori A, Zhai YL, Toki T, Nikaido T, Fujii S (1998) Mast cells in smooth muscle tumors of the uterus. Int J Gynecol Pathol 17:336–342 Ouyang DW, Economy KE, Norwitz ER (2006) Obstetric complications of fibroids. Obstet Gynecol Clin N Am 33:153–169. https://doi.org/10.1016/j.ogc.2005. 12.010 Panagopoulos I et al (2012) Novel fusion of MYST/Esa1associated factor 6 and PHF1 in endometrial stromal sarcoma. PLoS One 7:e39354. https://doi.org/10.1371/ journal.pone.0039354. PONE-D-12-09483 [pii] Panagopoulos I et al (2013) Fusion of the ZC3H7B and BCOR genes in endometrial stromal sarcomas carrying an X;22-translocation. Genes Chromosomes Cancer 52:610–618. https://doi.org/10.1002/gcc.22057 Park SH, Ro JY, Kim HS, Lee ES (2003) Perivascular epithelioid cell tumor of the uterus: immunohistochemical, ultrastructural and molecular study. Pathol Int 53:800–805 Park JY, Kim DY, Suh DS, Kim JH, Kim YM, Kim YT, Nam JH (2008) Prognostic factors and treatment outcomes of patients with uterine sarcoma: analysis of 127 patients at a single institution, 1989–2007. J Cancer Res Clin Oncol 134:1277–1287. https://doi.org/ 10.1007/s00432-008-0422-2 Park JY, Park SK, Kim DY, Kim JH, Kim YM, Kim YT, Nam JH (2011) The impact of tumor morcellation during surgery on the prognosis of patients with apparently early uterine leiomyosarcoma. Gynecol Oncol 122:255–259. https://doi.org/10.1016/j.ygyno.2011. 04.021 Park JY, Sung CO, Jang SJ, Song SY, Han JH, Kim KR (2013) Pulmonary metastatic nodules of uterine low-grade endometrial stromal sarcoma: histopathological and immunohistochemical analysis of 10 cases. Histopathology 63:833–840. https://doi.org/ 10.1111/his.12232

10

Mesenchymal Tumors of the Uterus

Parker WH, Fu YS, Berek JS (1994) Uterine sarcoma in patients operated on for presumed leiomyoma and rapidly growing leiomyoma. Obstet Gynecol 83:414–418 Parker WH, Kaunitz AM, Pritts EA, Olive DL, Chalas E, Clarke-Pearson DL, Berek JS (2016) U.S. Food and Drug Administration’s guidance regarding morcellation of leiomyomas: well-intentioned, but is it harmful for women? Obstet Gynecol 127:18–22. https://doi.org/10.1097/aog.0000000000001157 Parra-Herran CE, Yuan L, Nucci MR, Quade BJ (2014) Targeted development of specific biomarkers of endometrial stromal cell differentiation using bioinformatics: the IFITM1 model. Mod Pathol 27:569–579. https://doi.org/10.1038/modpathol.2013.123 Parra-Herran C, Quick CM, Howitt BE, Dal Cin P, Quade BJ, Nucci MR (2015) Inflammatory myofibroblastic tumor of the uterus: clinical and pathologic review of 10 cases including a subset with aggressive clinical course. Am J Surg Pathol 39:157–168. https://doi.org/ 10.1097/pas.0000000000000330 Parra-Herran C, Schoolmeester JK, Yuan L, Dal Cin P, Fletcher CD, Quade BJ, Nucci MR (2016) Myxoid leiomyosarcoma of the uterus: a clinicopathologic analysis of 30 cases and review of the literature with reappraisal of its distinction from other uterine myxoid mesenchymal neoplasms. Am J Surg Pathol 40:285–301. https://doi. org/10.1097/PAS.0000000000000593. 00000478201603000-00001 [pii] Patai K, Illyes G, Varbiro S, Gidai J, Kosa L, Vajo Z (2006) Uterine leiomyosarcoma with osteoclast like giant cells and long standing systemic symptoms. Gynecol Oncol 102:403–405. https://doi.org/10.1016/j.ygyno. 2006.02.030 Pautier P et al (2000) Analysis of clinicopathologic prognostic factors for 157 uterine sarcomas and evaluation of a grading score validated for soft tissue sarcoma. Cancer 88:1425–1431 Pauwels P, Dal Cin P, Van de Moosdijk CN, Vrints L, Sciot R, Van den Berghe H (1996) Cytogenetics revealing the diagnosis in a metastatic endometrial stromal sarcoma. Histopathology 29:84–87 Payson M, Leppert P, Segars J (2006) Epidemiology of myomas. Obstet Gynecol Clin N Am 33:1–11. https:// doi.org/10.1016/j.ogc.2005.12.004 Pedeutour F, Quade BJ, Sornberger K, Tallini G, Ligon AH, Weremowicz S, Morton CC (2000) Dysregulation of HMGIC in a uterine lipoleiomyoma with a complex rearrangement including chromosomes 7, 12, and 14. Genes Chromosomes Cancer 27:209–215 Pelmus M et al (2009) Prognostic factors in early-stage leiomyosarcoma of the uterus. Int J Gynecol Cancer 1 9:385–390. https://doi.org/10.1111/IGC.0b013e3181 a1bfbc Pérot G et al (2012) MED12 alterations in both human benign and malignant uterine soft tissue tumors. PLoS One 7:e40015. https://doi.org/10.1371/journal. pone.0040015 Perrone T, Dehner LP (1988) Prognostically favorable “mitotically active” smooth-muscle tumors of the

641 uterus. A clinicopathologic study of ten cases. Am J Surg Pathol 12:1–8 Peters WA 3rd, Howard DR, Andersen WA, Figge DC (1994) Uterine smooth-muscle tumors of uncertain malignant potential. Obstet Gynecol 83:1015–1020 Pickett JL et al (2017) Inflammatory myofibroblastic tumors of the female genital tract are under-recognized: a low threshold for ALK immunohistochemistry is required. Am J Surg Pathol 41:1433–1442. https://doi. org/10.1097/pas.0000000000000909 Pieslor PC, Orenstein JM, Hogan DL, Breslow A (1979) Ultrastructure of myofibroblasts and decidualized cells in leiomyomatosis peritonealis disseminata. Am J Clin Pathol 72:875–882 Pinto A, Kahn RM, Rosenberg AE, Slomovitz B, Quick CM, Whisman MK, Huang M (2018) Uterine rhabdomyosarcoma in adults. Hum Pathol. https://doi.org/ 10.1016/j.humpath.2018.01.007 Piscuoglio S et al (2016) Uterine adenosarcomas are mesenchymal neoplasms. J Pathol 238:381–388. https:// doi.org/10.1002/path.4675 Portugal R, Oliva E (2009) Calretinin: diagnostic utility in the female genital tract. Adv Anat Pathol 16:118–124. https://doi.org/10.1097/PAP.0b013e31819923ce. 00125480-200903000-00005 [pii] Pounder DJ (1982) Fatty tumours of the uterus. J Clin Pathol 35:1380–1383 Pradhan D, Mohanty SK (2013) Uterine tumors resembling ovarian sex cord tumors. Arch Pathol Lab Med 137:1832–1836. https://doi.org/10.5858/arpa. 2012-0634-RS Prat J (2009) FIGO staging for uterine sarcomas. Int J Gynaecol Obstet 104:177–178. https://doi.org/10. 1016/j.ijgo.2008.12.008 Prat J, Mbatani (2015) Uterine sarcomas. Int J Gynaecol Obstet 131(Suppl 2):S105–S110. https://doi.org/ 10.1016/j.ijgo.2015.06.006. S0020-7292(15)00377-X [pii] Prayson RA, Hart WR (1992) Mitotically active leiomyomas of the uterus. Am J Clin Pathol 97:14–20 Prayson RA, Goldblum JR, Hart WR (1997) Epithelioid smooth-muscle tumors of the uterus: a clinicopathologic study of 18 patients. Am J Surg Pathol 21:383–391 Press MF, Scully RE (1985) Endometrial “sarcomas” complicating ovarian thecoma, polycystic ovarian disease and estrogen therapy. Gynecol Oncol 21:135–154. 0090-8258(85)90246-X [pii] Pugh A, McCluggage WG, Hirschowitz L (2012) Multifocal uterine myxoid change: a newly recognized association with neurofibromatosis type 1. Int J Gynecol Pathol 31:580–583. https://doi.org/ 10.1097/PGP.0b013e31824d359a Punnett HH, Halligan GE, Zaeri N, Karmazin N (1989) Translocation 10;17 in clear cell sarcoma of the kidney. A first report. Cancer Genet Cytogenet 41:123–128. 0165-4608(89)90116-7 [pii] Quade BJ (1995) Pathology, cytogenetics and molecular biology of uterine leiomyomas and other smooth muscle lesions. Curr Opin Obstet Gynecol 7:35–42

642 Quade BJ, McLachlin CM, Soto-Wright V, Zuckerman J, Mutter GL, Morton CC (1997) Disseminated peritoneal leiomyomatosis. Clonality analysis by X chromosome inactivation and cytogenetics of a clinically benign smooth muscle proliferation. Am J Pathol 150: 2153–2166 Quade BJ, Dal Cin P, Neskey DM, Weremowicz S, Morton CC (2002) Intravenous leiomyomatosis: molecular and cytogenetic analysis of a case. Mod Pathol 15:351–356. https://doi.org/10.1038/modpathol.3880529 Quade BJ, Weremowicz S, Neskey DM, Vanni R, Ladd C, Dal Cin P, Morton CC (2003) Fusion transcripts involving HMGA2 are not a common molecular mechanism in uterine leiomyomata with rearrangements in 12q15. Cancer Res 63:1351–1358 Quddus MR, Rashid L, Sung CJ, Steinhoff MM, Cunxian Z, Lawrence WD (2009) Ewing’s sarcoma/ peripheral primitive neuroectodermal tumor (ES/PNET) differentiation in endometrial serous carcinomas. Reprod Sci 16:591–595. https://doi.org/10. 1177/1933719109332824 Rabban JT, Zaloudek CJ, Shekitka KM, Tavassoli FA (2005) Inflammatory myofibroblastic tumor of the uterus: a clinicopathologic study of 6 cases emphasizing distinction from aggressive mesenchymal tumors. Am J Surg Pathol 29:1348–1355 Rackow BW, Arici A (2006) Options for medical treatment of myomas. Obstet Gynecol Clin N Am 33:97–113. https://doi.org/10.1016/j.ogc.2005.12.014 Radig K, Buhtz P, Roessner A (1998) Alveolar soft part sarcoma of the uterine corpus. Report of two cases and review of the literature. Pathol Res Pract 194:59–63 Rahimi S, Akaev I, Marani C, Chopra M, Yeoh CC (2018) Immunohistochemical expression of different subtypes of cytokeratins by endometrial stromal sarcoma. Appl Immunohistochem Mol Morphol. https://doi.org/ 10.1097/PAI.0000000000000642 Rakheja D, Weinberg AG, Tomlinson GE, Partridge K, Schneider NR (2004) Translocation (10;17)(q22;p13): a recurring translocation in clear cell sarcoma of kidney. Cancer Genet Cytogenet 154:175–179. https://doi. org/10.1016/j.cancergencyto.2004.02.024. S01654608(04)00111-6 [pii] Ramalingam P, Masand RP, Euscher ED, Malpica A (2016) Undifferentiated carcinoma of the endometrium: an expanded immunohistochemical analysis including PAX-8 and basal-like carcinoma surrogate markers. Int J Gynecol Pathol 35:410–418. https://doi.org/ 10.1097/PGP.0000000000000248 Rao Q et al (2013) Cathepsin K expression in a wide spectrum of perivascular epithelioid cell neoplasms (PEComas): a clinicopathological study emphasizing extrarenal PEComas. Histopathology 62:642–650. https://doi.org/10.1111/his.12059 Raut CP et al (2009) Predictive value of FIGO and AJCC staging systems in patients with uterine leiomyosarcoma. Eur J Cancer 45:2818–2824. https:// doi.org/10.1016/j.ejca.2009.06.030

E. Oliva et al. Ravid Y, Formanski M, Smith Y, Reich R, Davidson B (2016) Uterine leiomyosarcoma and endometrial stromal sarcoma have unique miRNA signatures. Gynecol Oncol 140:512–517. https://doi.org/10. 1016/j.ygyno.2016.01.001 Reed NS et al (2008) Phase III randomised study to evaluate the role of adjuvant pelvic radiotherapy in the treatment of uterine sarcomas stages I and II: an European Organisation for Research and Treatment of Cancer Gynaecological Cancer Group Study (protocol 55874). Eur J Cancer 44:808–818. https://doi.org/ 10.1016/j.ejca.2008.01.019 Regidor PA, Schmidt M, Callies R, Kato K, Schindler AE (1995) Estrogen and progesterone receptor content of GnRH analogue pretreated and untreated uterine leiomyomata. Eur J Obstet Gynecol Reprod Biol 63:69–73 Reich O, Regauer S (2004) Aromatase expression in low-grade endometrial stromal sarcomas: an immunohistochemical study. Mod Pathol 17:104–108. https:// doi.org/10.1038/sj.modpathol.3800031 Reich O, Regauer S (2007) Hormonal therapy of endometrial stromal sarcoma. Curr Opin Oncol 19:347–352. https://doi.org/10.1097/CCO.0b013e3281a7ef3a. 00001622-200707000-00011 [pii] Reich O, Regauer S, Urdl W, Lahousen M, Winter R (2000) Expression of oestrogen and progesterone receptors in low-grade endometrial stromal sarcomas. Br J Cancer 82:1030–1034. https://doi.org/10.1054/ bjoc.1999.1038. S0007092099910388 [pii] Reyes C, Karamurzin Y, Frizzell N, Garg K, Nonaka D, Chen YB, Soslow RA (2014) Uterine smooth muscle tumors with features suggesting fumarate hydratase aberration: detailed morphologic analysis and correlation with S-(2-succino)-cysteine immunohistochemistry. Mod Pathol 27:1020–1027. https://doi.org/ 10.1038/modpathol.2013.215 Rivasi F, Botticelli L, Bettelli SR, Masellis G (2008) Alveolar rhabdomyosarcoma of the uterine cervix. A case report confirmed by FKHR break-apart rearrangement using a fluorescence in situ hybridization probe on paraffin-embedded tissues. Int J Gynecol Pathol 27:442–446 Rizeq MN, van de Rijn M, Hendrickson MR, Rouse RV (1994) A comparative immunohistochemical study of uterine smooth muscle neoplasms with emphasis on the epithelioid variant. Hum Pathol 25:671–677 Roma AA, Yang B, Senior ME, Goldblum JR (2005) TFE3 immunoreactivity in alveolar soft part sarcoma of the uterine cervix: case report. Int J Gynecol Pathol 24:131–135 Roma AA, Allende D, Fadare O, Forscher C, Rutgers JK (2017) On uterine angiosarcomas: 2 additional cases. Int J Gynecol Pathol 36:369–371. https://doi.org/ 10.1097/pgp.0000000000000352 Ronaghy A, Xiao GQ, Santagada E, Hasanovic A, Unger P (2018) Expression of GATA-3 in testicular and gynecologic mesothelial neoplastic and non-neoplastic

10

Mesenchymal Tumors of the Uterus

tissues. Int J Gynecol Pathol 37:284–289. https://doi. org/10.1097/pgp.0000000000000403 Rose PG, Piver MS, Tsukada Y, Lau T (1989) Patterns of metastasis in uterine sarcoma. An autopsy study. Cancer 63:935–938 Rosty C, Genestie C, Blondon J, Le Charpentier Y (1998) Endometrial stromal tumor associated with rhabdoid phenotype and and zones of “sex cord-like” differentiation. Ann Pathol 18:133–136. MDOI-AP-07-199818-2-0242-6498-101019-ART57 [pii] Roth LM, Reed RJ (1999) Dissecting leiomyomas of the uterus other than cotyledonoid dissecting leiomyomas: a report of eight cases. Am J Surg Pathol 23:1032–1039 Roth LM, Reed RJ (2000) Cotyledonoid leiomyoma of the uterus: report of a case. Int J Gynecol Pathol 19:272–275 Roth E, Taylor HB (1966) Heterotopic cartilage in the uterus. Obstet Gynecol 27:838–844 Roth LM, Reed RJ, Sternberg WH (1996) Cotyledonoid dissecting leiomyoma of the uterus. The Sternberg tumor. Am J Surg Pathol 20:1455–1461 Roy M et al (2017) Androgen receptor expression in endometrial stromal sarcoma: correlation with clinicopathologic features. Int J Gynecol Pathol 36:420–427. https://doi.org/10.1097/PGP.0000000000000353 Rush DS, Tan J, Baergen RN, Soslow RA (2001) h-Caldesmon, a novel smooth muscle-specific antibody, distinguishes between cellular leiomyoma and endometrial stromal sarcoma. Am J Surg Pathol 25: 253–258 Rushing RS et al (2003) Uterine sarcomas express KIT protein but lack mutation(s) in exon 11 or 17 of c-KIT. Gynecol Oncol 91:9–14 Sahin AA, Silva EG, Landon G, Ordonez NG, Gershenson DM (1989) Endometrial tissue in myometrial vessels not associated with menstruation. Int J Gynecol Pathol 8:139–146 Sangoi AR, McKenney JK, Schwartz EJ, Rouse RV, Longacre TA (2009) Adenomatoid tumors of the female and male genital tracts: a clinicopathological and immunohistochemical study of 44 cases. Mod Pathol 22:1228–1235. https://doi.org/10.1038/ modpathol.2009.90 Sanz-Ortega J, Vocke C, Stratton P, Linehan WM, Merino MJ (2013) Morphologic and molecular characteristics of uterine leiomyomas in hereditary leiomyomatosis and renal cancer (HLRCC) syndrome. Am J Surg Pathol 37:74–80. https://doi.org/10.1097/PAS.0b013e31825ec16f Satoh Y, Ishikawa Y, Miyoshi T, Mukai H, Okumura S, Nakagawa K (2003) Pulmonary metastases from a low-grade endometrial stromal sarcoma confirmed by chromosome aberration and fluorescence in-situ hybridization approaches: a case of recurrence 13 years after hysterectomy. Virchows Arch 442:173–178. https://doi.org/10.1007/s00428002-0731-9 Schaefer IM, Hornick JL, Sholl LM, Quade BJ, Nucci MR, Parra-Herran C (2017) Abnormal p53 and p16 staining patterns distinguish uterine leiomyosarcoma from

643 inflammatory myofibroblastic tumour. Histopathology 70:1138–1146. https://doi.org/10.1111/his.13176 Schammel DP, Tavassoli FA (1998) Uterine angiosarcomas: a morphologic and immunohistochemical study of four cases. Am J Surg Pathol 22:246–250 Schammel DP, Silver SA, Tavassoli FA (1999) Combined endometrial stromal/smooth muscle neoplasms of the uterus. A clinicopathologic study of 38 cases [abstract]. Mod Pathol 12:124A Schmid C, Beham A, Kratochvil P (1990) Haematopoiesis in a degenerating uterine leiomyoma. Arch Gynecol Obstet 248:81–86 Schneider D, Halperin R, Segal M, Maymon R, Bukovsky I (1995) Myxoid leiomyosarcoma of the uterus with unusual malignant histologic pattern – a case report. Gynecol Oncol 59:156–158. https://doi.org/10.1006/ gyno.1995.1284 Schoenmakers EF, Huysmans C, Van de Ven WJ (1999) Allelic knockout of novel splice variants of human recombination repair gene RAD51B in t(12;14) uterine leiomyomas. Cancer Res 59:19–23 Schoenmakers EF et al (2013) Identification of CUX1 as the recurrent chromosomal band 7q22 target gene in human uterine leiomyoma. Genes Chromosomes Cancer 52:11–23. https://doi.org/10.1002/gcc.22001 Schoolmeester JK, Park KJ (2015) Incidental nodal lymphangioleiomyomatosis is not a harbinger of pulmonary lymphangioleiomyomatosis: a study of 19 cases with evaluation of diagnostic immunohistochemistry. Am J Surg Pathol 39:1404–1410. https:// doi.org/10.1097/pas.0000000000000470 Schoolmeester JK, Howitt BE, Hirsch MS, Dal Cin P, Quade BJ, Nucci MR (2014) Perivascular epithelioid cell neoplasm (PEComa) of the gynecologic tract: clinicopathologic and immunohistochemical characterization of 16 cases. Am J Surg Pathol 38:176–188. https:// doi.org/10.1097/pas.0000000000000133 Schoolmeester JK et al (2015) TFE3 translocationassociated perivascular epithelioid cell neoplasm (PEComa) of the gynecologic tract: morphology, immunophenotype, differential diagnosis. Am J Surg Pathol 39:394–404. https://doi.org/10.1097/ pas.0000000000000349 Schoolmeester JK, Stamatakos MD, Moyer AM, Park KJ, Fairbairn M, Fader AN (2016) Pleomorphic Liposarcoma arising in a lipoleiomyosarcoma of the uterus: report of a case with genetic profiling by a next generation sequencing panel. Int J Gynecol Pathol 35:321–326. https://doi. org/10.1097/pgp.0000000000000241 Schoolmeester JK, Carlson J, Keeney GL, Fritchie KJ, Oliva E, Young RH, Nucci MR (2017) Alveolar soft part sarcoma of the female genital tract: a morphologic, immunohistochemical, and molecular cytogenetic study of 10 cases with emphasis on its distinction from morphologic mimics. Am J Surg Pathol 41:622–632. https://doi.org/10.1097/pas. 0000000000000796

644 Schwartz PE, Kelly MG (2006) Malignant transformation of myomas: myth or reality? Obstet Gynecol Clin N Am 33:183–198, xii. https://doi.org/10.1016/j. ogc.2005.12.003 Schwartz LB, Diamond MP, Schwartz PE (1993) Leiomyosarcomas: clinical presentation. Am J Obstet Gynecol 168:180–183 Sciallis AP, Bedroske PP, Schoolmeester JK, Sukov WR, Keeney GL, Hodge JC, Bell DA (2014) High-grade endometrial stromal sarcomas: a clinicopathologic study of a group of tumors with heterogenous morphologic and genetic features. Am J Surg Pathol 38:1161–1172. https:// doi.org/10.1097/PAS.0000000000000256. 00000478201409000-00001 [pii] Seidman JD, Thomas RM (1993) Multiple plexiform tumorlets of the uterus. Arch Pathol Lab Med 117:1255–1256 Seidman MA, Oduyebo T, Muto MG, Crum CP, Nucci MR, Quade BJ (2012) Peritoneal dissemination complicating morcellation of uterine mesenchymal neoplasms. PLoS One 7:e50058. https://doi.org/10.1371/ journal.pone.0050058 Sener AB, Seckin NC, Ozmen S, Gokmen O, Dogu N, Ekici E (1996) The effects of hormone replacement therapy on uterine fibroids in postmenopausal women. Fertil Steril 65:354–357 Shah R, McCluggage WG (2009) Symplastic atypia in neoplastic and non-neoplastic endometrial stroma: report of 3 cases with a review of atypical symplastic cells within the female genital tract. Int J Gynecol Pathol 28:334–337. https://doi.org/ 10.1097/PGP.0b013e3181999450 Shah VI, McCluggage WG (2015) Cyclin D1 does not distinguish YWHAE-NUTM2 high-grade endometrial stromal sarcoma from undifferentiated endometrial carcinoma. Am J Surg Pathol 39:722–724. https://doi.org/ 10.1097/PAS.0000000000000427. 00000478201505000-00019 [pii] Shah JP, Bryant CS, Kumar S, Ali-Fehmi R, Malone JM Jr, Morris RT (2008) Lymphadenectomy and ovarian preservation in low-grade endometrial stromal sarcoma. Obstet Gynecol 112:1102–1108. https://doi.org/ 10.1097/AOG.0b013e31818aa89a. 112/5/1102 [pii] Shaw RW (1998) Gonadotrophin hormone-releasing hormone analogue treatment of fibroids. Baillieres Clin Obstet Gynaecol 12:245–268 Signorelli M, Fruscio R, Dell’Anna T, Buda A, Giuliani D, Ceppi L, Milani R (2010) Lymphadenectomy in uterine low-grade endometrial stromal sarcoma: an analysis of 19 cases and a literature review. Int J Gynecol Cancer 20:1363–13 66. https://doi.org/10.1111/IGC.0b013e3181efd861. 00009577-201011000-00015 [pii] Silva EG, Tornos C, Ordonez NG, Morris M (1995) Uterine leiomyosarcoma with clear cell areas. Int J Gynecol Pathol 14:174–178 Silva EG, Deavers MT, Bodurka DC, Malpica A (2004) Uterine epithelioid leiomyosarcomas with clear cells:

E. Oliva et al. reactivity with HMB-45 and the concept of PEComa. Am J Surg Pathol 28:244–249 Silva EG, Bodurka DC, Scouros MA, Ayala A (2005) A uterine leiomyosarcoma that became positive for HMB45 in the metastasis. Ann Diagn Pathol 9:43–45 Simpson KW, Albores-Saavedra J (2007) HMB-45 reactivity in conventional uterine leiomyosarcomas. Am J Surg Pathol 31:95–98. https://doi.org/10.1097/01. pas.0000213346.57391.70 Sinkre P, Albores-Saavedra J, Miller DS, Copeland LJ, Hameed A (2000a) Endometrial endometrioid carcinomas associated with Ewing sarcoma/peripheral primitive neuroectodermal tumor. Int J Gynecol Pathol 19:127–132 Sinkre P, Miller DS, Milchgrub S, Hameed A (2000b) Adenomyofibroma of the endometrium with skeletal muscle differentiation. Int J Gynecol Pathol 19:280–283 Siskin GP et al (2006) A prospective multicenter comparative study between myomectomy and uterine artery embolization with polyvinyl alcohol microspheres: long-term clinical outcomes in patients with symptomatic uterine fibroids. J Vasc Interv Radiol 17:1287–1295. https://doi.org/10.1097/01. rvi.0000231953.91787.af Slatter TL et al (2015) Loss of ATRX and DAXX expression identifies poor prognosis for smooth muscle tumours of uncertain malignant potential and early stage uterine leiomyosarcoma. J Pathol Clin Res 1:95–105. https://doi.org/10.1002/cjp2.11 Sorensen PHB et al (2002) PAX3-FKHR and PAX7FKHR gene fusions are prognostic indicators in alveolar rhabdomyosarcoma: a report from the Children’s Oncology Group. J Clin Oncol 20:2672–2679 Sornberger KS et al (1999) Expression of HMGIY in three uterine leiomyomata with complex rearrangements of chromosome 6. Cancer Genet Cytogenet 114:9–16 Soslow RA, Ali A, Oliva E (2008) Mullerian adenosarcomas: an immunophenotypic analysis of 35 cases. Am J Surg Pathol 32:1013–1021 Sozen I, Arici A (2006) Cellular biology of myomas: interaction of sex steroids with cytokines and growth factors. Obstet Gynecol Clin N Am 33:41–58. https:// doi.org/10.1016/j.ogc.2005.12.005 Spies JB, Ascher SA, Roth AR, Kim J, Levy EB, GomezJorge J (2001) Uterine artery embolization for leiomyomata. Obstet Gynecol 98:29–34 Staats PN et al (2009) Uterine tumors resembling ovarian sex cord tumors (UTROSCT) lack the JAZF1-JJAZ1 translocation frequently seen in endometrial stromal tumors. Am J Surg Pathol 33:1206–1212. https://doi. org/10.1097/PAS.1200b1013e3181a1207b1209cf Stemme S, Ghaderi M, Carlson JW (2014) Diagnosis of endometrial stromal tumors: a clinicopathologic study of 25 biopsy specimens with identification of problematic areas. Am J Clin Pathol 141:133–139. https://doi. org/10.1309/AJCPXD0TPYSNVI8I. 141/1/133 [pii] Stewart L et al (2008) Association of germline mutations in the fumarate hydratase gene and uterine fibroids in

10

Mesenchymal Tumors of the Uterus

women with hereditary leiomyomatosis and renal cell cancer. Arch Dermatol 144:1584–1592. https://doi.org/ 10.1001/archdermatol.2008.517 Stewart CJ, Leung YC, Murch A, Peverall J (2014) Evaluation of fluorescence in-situ hybridization in monomorphic endometrial stromal neoplasms and their histological mimics: a review of 49 cases. Histopathology 65:473–482. https://doi.org/10.1111/his.12406 Stewart CJ, Charles A, Foulkes WD (2016a) Gynecologic manifestations of the DICER1 syndrome. Surg Pathol Clin 9:227–241. https://doi.org/10.1016/j.path. 2016.01.002 Stewart CJ, Crook M, Tan A (2016b) SF1 immunohistochemistry is useful in differentiating uterine tumours resembling sex cord-stromal tumours from potential histological mimics. Pathology 48:434–440. https:// doi.org/10.1016/j.pathol.2016.03.013 Stolnicu S, Molnar C, Barsan I, Boros M, Nogales FF, Soslow RA (2016) The impact on survival of an extensive sex cord-like component in mullerian adenosarcomas: a study comprising 6 cases. Int J Gynecol Pathol 35:147–152. https://doi.org/10.1097/ pgp.0000000000000231 Stovall TG, Ling FW, Henry LC, Woodruff MR (1991) A randomized trial evaluating leuprolide acetate before hysterectomy as treatment for leiomyomas. Am J Obstet Gynecol 164:1420–1423; discussion 1423-1425 Subbiah V et al (2015) STUMP un “stumped”: anti-tumor response to anaplastic lymphoma kinase (ALK) inhibitor based targeted therapy in uterine inflammatory myofibroblastic tumor with myxoid features harboring DCTN1-ALK fusion. J Hematol Oncol 8:66. https:// doi.org/10.1186/s13045-015-0160-2 Suginami H, Kaura R, Ochi H, Matsuura S (1990) Intravenous leiomyomatosis with cardiac extension: successful surgical management and histopathologic study. Obstet Gynecol 76:527–529 Sullivan LM, Atkins KA, LeGallo RD (2009) PAX immunoreactivity identifies alveolar rhabdomyosarcoma. Am J Surg Pathol 33:775–780. https://doi.org/ 10.1097/PAS.0b013e318191614f Sumathi VP, Al-Hussaini M, Connolly LE, Fullerton L, McCluggage WG (2004) Endometrial stromal neoplasms are immunoreactive with WT-1 antibody. Int J Gynecol Pathol 23:241–247. 00004347-20040700000007 [pii] Sutton G, Blessing J, Hanjani P, Kramer P (2005) Phase II evaluation of liposomal doxorubicin (Doxil) in recurrent or advanced leiomyosarcoma of the uterus: a Gynecologic Oncology Group study. Gynecol Oncol 96:749–752. https://doi.org/10.1016/j.ygyno. 2004.11.036 Tahlan A, Nanda A, Mohan H (2006) Uterine adenomyoma: a clinicopathologic review of 26 cases and a review of the literature. Int J Gynecol Pathol 25:361–365 Tai LH, Tavassoli FA (2002) Endometrial polyps with atypical (bizarre) stromal cells. Am J Surg Pathol 26:505–509

645 Tanimoto A, Sasaguri T, Arima N, Hashimoto H, Hamada T, Sasaguri Y (1996) Endometrial stromal sarcoma of the uterus with rhabdoid features. Pathol Int 46:231–237 Tanner EJ, Garg K, Leitao MM Jr, Soslow RA, Hensley ML (2012) High grade undifferentiated uterine sarcoma: surgery, treatment, and survival outcomes. Gynecol Oncol 127:27–31. https://doi.org/10.1016/j. ygyno.2012.06.030. S0090-8258(12)00481-7 [pii] Tanner EJ, Toussaint T, Leitao MM Jr, Hensley ML, Soslow RA, Gardner GJ, Jewell EL (2013) Management of uterine adenosarcomas with and without sarcomatous overgrowth. Gynecol Oncol 129:140–144. https://doi.org/10.1016/j.ygyno.2012.12.036 Tavassoli FA, Norris HJ (1981) Mesenchymal tumours of the uterus. VII. A clinicopathological study of 60 endometrial stromal nodules. Histopathology 5:1–10 Tavassoli FA, Norris HJ (1982) Peritoneal leiomyomatosis (leiomyomatosis peritonealis disseminata): a clinicopathologic study of 20 cases with ultrastructural observations. Int J Gynecol Pathol 1:59–74 Tawfik O, Rao D, Nothnick WB, Graham A, Mau B, Fan F (2014) Transgelin, a novel marker of smooth muscle differentiation, effectively distinguishes endometrial stromal tumors from uterine smooth muscle tumors. Int J Gynecol Obstet Reprod Med Res 1:26–31 Terada T (2009) Gastrointestinal stromal tumor of the uterus: a case report with genetic analyses of c-kit and PDGFRA genes. Int J Gynecol Pathol 28:29–34. https://doi.org/10.1097/PGP.0b013e3181808000 Thway K, Fisher C (2015) PEComa: morphology and genetics of a complex tumor family. Ann Diagn Pathol 19:359–368. https://doi.org/10.1016/j. anndiagpath.2015.06.003 Tietze L, Gunther K, Horbe A, Pawlik C, Klosterhalfen B, Handt S, Merkelbach-Bruse S (2000) Benign metastasizing leiomyoma: a cytogenetically balanced but clonal disease. Hum Pathol 31:126–128 Tiltman AJ (1985) The effect of progestins on the mitotic activity of uterine fibromyomas. Int J Gynecol Pathol 4:89–96 Toki T, Shimizu M, Takagi Y, Ashida T, Konishi I (2002) CD10 is a marker for normal and neoplastic endometrial stromal cells. Int J Gynecol Pathol 21:41–47 Tomlinson IP et al (2002) Germline mutations in FH predispose to dominantly inherited uterine fibroids, skin leiomyomata and papillary renal cell cancer. Nat Genet 30:406–410. https://doi.org/10.1038/ng849 Toro JR et al (2003) Mutations in the fumarate hydratase gene cause hereditary leiomyomatosis and renal cell cancer in families in North America. Am J Hum Genet 73:95–106. https://doi.org/10.1086/376435 Torres VE et al (1995) Extrapulmonary lymphangioleiomyomatosis and lymphangiomatous cysts in tuberous sclerosis complex. Mayo Clin Proc 70:641–648. https://doi.org/10.1016/s0025-6196(11)63915-3 Tresukosol D, Kudelka AP, Malpica A, Varma DG, Edwards CL, Kavanagh JJ (1995) Leuprolide acetate

646 and intravascular leiomyomatosis. Obstet Gynecol 86:688–692 Tulandi T, Leung A, Jan N (2016) Nonmalignant sequelae of unconfined morcellation at laparoscopic hysterectomy or myomectomy. J Minim Invasive Gynecol 23:331–337. https://doi.org/10.1016/j.jmig.2016.01.017 Ubago JM, Zhang Q, Kim JJ, Kong B, Wei JJ (2016) Two subtypes of atypical leiomyoma: clinical, histologic, and molecular analysis. Am J Surg Pathol 40:923–933. https://doi.org/10.1097/pas.0000000000000646 Upadhyaya NB, Doody MC, Googe PB (1990) Histopathological changes in leiomyomata treated with leuprolide acetate. Fertil Steril 54:811–814 van Diest PJ et al (1992) Reproducibility of mitosis counting in 2,469 breast cancer specimens: results from the Multicenter Morphometric Mammary Carcinoma Project. Hum Pathol 23:603–607 van Slegtenhorst M et al (1997) Identification of the tuberous sclerosis gene TSC1 on chromosome 9q34. Science 277:805–808 Vang R, Kempson RL (2002) Perivascular epithelioid cell tumor (‘PEComa’) of the uterus: a subset of HMB-45positive epithelioid mesenchymal neoplasms with an uncertain relationship to pure smooth muscle tumors. Am J Surg Pathol 26:1–13 Vang R, Medeiros LJ, Ha CS, Deavers M (2000) Non-Hodgkin’s lymphomas involving the uterus: a clinicopathologic analysis of 26 cases. Mod Pathol 13:19–28. https://doi.org/10.1038/modpathol.3880005 Varghese L, Arnesen M, Boente M (2006) Primitive neuroectodermal tumor of the uterus: a case report and review of literature. Int J Gynecol Pathol 25:373–377 Venkatraman L, Elliott H, Steele EK, McClelland HR, McCluggage WG (2003) Serous carcinoma arising in an adenofibroma of the endometrium. Int J Gynecol Pathol 22:194–197 Veras E, Malpica A, Deavers MT, Silva EG (2009) Mitosisspecific marker phospho-histone H3 in the assessment of mitotic index in uterine smooth muscle tumors: a pilot study. Int J Gynecol Pathol 28:316–321. https:// doi.org/10.1097/PGP.0b013e318193df97 Veras E, Zivanovic O, Jacks L, Chiappetta D, Hensley M, Soslow R (2011) “Low-grade leiomyosarcoma” and late-recurring smooth muscle tumors of the uterus: a heterogenous collection of frequently misdiagnosed tumors associated with an overall favorable prognosis relative to conventional uterine leiomyosarcomas. Am J Surg Pathol 35:1626–1637. https://doi.org/10.1097/ PAS.0b013e31822b44d2 Vitiello D, McCarthy S (2006) Diagnostic imaging of myomas. Obstet Gynecol Clin N Am 33:85–95. https://doi.org/10.1016/j.ogc.2005.12.013 Viville B, Charnock-Jones DS, Sharkey AM, Wetzka B, Smith SK (1997) Distribution of the A and B forms of the progesterone receptor messenger ribonucleic acid and protein in uterine leiomyomata and adjacent myometrium. Hum Reprod 12:815–822

E. Oliva et al. Vollenhoven B (1998) 1 Introduction: the epidemiology of uterine leiomyomas. Baillieres Clin Obstet Gynaecol 12:169–176. https://doi.org/10.1016/S0950-3552(98) 80059-X Wachter DL, Wunsch PH, Hartmann A, Agaimy A (2011) Adenomatoid tumors of the female and male genital tract. A comparative clinicopathologic and immunohistochemical analysis of 47 cases emphasizing their site-specific morphologic diversity. Virchows Arch 458:593–602. https://doi.org/10.1007/s00428-011-1054-5 Wagner AJ et al (2010) Clinical activity of mTOR inhibition with sirolimus in malignant perivascular epithelioid cell tumors: targeting the pathogenic activation of mTORC1 in tumors. J Clin Oncol 28:835–840. https:// doi.org/10.1200/jco.2009.25.2981 Wallach EE, Vlahos NF (2004) Uterine myomas: an overview of development, clinical features, and management. Obstet Gynecol 104:393–406. https://doi.org/ 10.1097/01.aog.0000136079.62513.39 Wang X, Kumar D, Seidman JD (2006) Uterine lipoleiomyomas: a clinicopathologic study of 50 cases. Int J Gynecol Pathol 25:239–242. https:// doi.org/10.1097/01.pgp.0000192273.66931.29 Wang WL et al (2011) Histopathologic prognostic factors in stage I leiomyosarcoma of the uterus: a detailed analysis of 27 cases. Am J Surg Pathol 35:522–529. https://doi.org/10.1097/PAS.0b013e31820ca624 Wang W, Zhu H, Wang J, Wang S, Wang D, Zhao J, Zhu H (2016) Clonality assessment of adenomatoid tumor supports its neoplastic nature. Hum Pathol 48:88–94. https://doi.org/10.1016/j.humpath.2015.09.032 Warburg O (1956) On the origin of cancer cells. Science 123:309–314 Watanabe K, Suzuki T (2006) Uterine leiomyoma versus leiomyosarcoma: a new attempt at differential diagnosis based on their cellular characteristics. Histopathology 48:563–568. https://doi.org/10.1111/j.1365-2559. 2006.02368.x Wei MH et al (2006) Novel mutations in FH and expansion of the spectrum of phenotypes expressed in families with hereditary leiomyomatosis and renal cell cancer. J Med Genet 43:18–27. https://doi.org/10.1136/ jmg.2005.033506 Weichert W, Denkert C, Gauruder-Burmester A, Kurzeja R, Hamm B, Dietel M, Kroencke TJ (2005) Uterine arterial embolization with tris-acryl gelatin microspheres: a histopathologic evaluation. Am J Surg Pathol 29:955–961 Weissman A, Talmon R, Jakobi P (1993) Cavernous hemangioma of the uterus in a pregnant woman. Obstet Gynecol 81:825–827 Wolff M, Silva F, Kaye G (1979) Pulmonary metastases (with admixed epithelial elements) from smooth muscle neoplasms. Report of nine cases, including three males. Am J Surg Pathol 3:325–342 Wolfson AH et al (1994) A multivariate analysis of clinicopathologic factors for predicting outcome in uterine sarcomas. Gynecol Oncol 52:56–62

10

Mesenchymal Tumors of the Uterus

Wu TI et al (2013) Clinicopathologic parameters and immunohistochemical study of endometrial stromal sarcomas. Int J Gynecol Pathol 32:482–492. https:// doi.org/10.1097/PGP.0b013e3182729131 Xue WC, Cheung AN (2011) Endometrial stromal sarcoma of uterus. Best Pract Res Clin Obstet Gynaecol 25:719–732. https://doi.org/10.1016/j.bpobgyn.2011.07. 004. S1521-6934(11)00108-8 [pii] Yamadori I, Kobayashi S, Ogino T, Ohmori M, Tanaka H, Jimbo T (1993) Uterine leiomyoma with a focus of fatty and cartilaginous differentiation. Acta Obstet Gynecol Scand 72:307–309 Yang EJ, Mutter GL (2015) Biomarker resolution of uterine smooth muscle tumor necrosis as benign vs malignant. Mod Pathol 28:830–835. https://doi.org/10.1038/ modpathol.2015.35 Yang Y et al (2012) A novel fumarate hydratase-deficient HLRCC kidney cancer cell line, UOK268: a model of the Warburg effect in cancer. Cancer Genet 205:377–390. https://doi.org/10.1016/j.cancergen.2012.05.001 Yang CYet al (2015) Targeted next-generation sequencing of cancer genes identified frequent TP53 and ATRX mutations in leiomyosarcoma. Am J Transl Res 7:2072–2081 Yang EJ, Howitt BE, Fletcher CD, Nucci MR (2017) Solitary fibrous tumor of the female genital tract: a clinicopathologic analysis of 25 cases. Histopathology. https://doi.org/10.1111/his.13430 Yang EJ, Howitt BE, Fletcher CDM, Nucci MR (2018) Solitary fibrous tumour of the female genital tract: a clinicopathological analysis of 25 cases. Histopathology 72:749–759. https://doi.org/10.1111/his.13430 Yilmaz A, Rush DS, Soslow RA (2002) Endometrial stromal sarcomas with unusual histologic features: a report of 24 primary and metastatic tumors emphasizing fibroblastic and smooth muscle differentiation. Am J Surg Pathol 26:1142–1150 Yoon A et al (2014) Prognostic factors and outcomes in endometrial stromal sarcoma with the 2009 FIGO staging system: a multicenter review of 114 cases. Gynecol Oncol 132:70–75. https://doi.org/10.1016/j. ygyno.2013.10.029

647 Young RH, Scully RE (1990) Sarcomas metastatic to the ovary: a report of 21 cases. Int J Gynecol Pathol 9:231–252 Young RH, Prat J, Scully RE (1984) Endometrioid stromal sarcomas of the ovary. A clinicopathologic analysis of 23 cases. Cancer 53:1143–1155 Young RH, Harris NL, Scully RE (1985) Lymphoma-like lesions of the lower female genital tract: a report of 16 cases. Int J Gynecol Pathol 4:289–299 Zaloudek CJ, Norris HJ (1981) Adenofibroma and adenosarcoma of the uterus: a clinicopathologic study of 35 cases. Cancer 48:354–366 Zamecnik M, Michal M (1998) Endometrial stromal nodule with retiform sex-cord-like differentiation. Pathol Res Pract 194:449–453 Zhai YL, Kobayashi Y, Mori A, Orii A, Nikaido T, Konishi I, Fujii S (1999) Expression of steroid receptors, Ki-67, and p53 in uterine leiomyosarcomas. Int J Gynecol Pathol 18:20–28 Zhang H, Luo J, Feng X (2012) Kaposiform hemangioendothelioma in the uterine cervix of a 5-year girl. Fetal Pediatr Pathol 31:273–277. https://doi.org/10.3109/ 15513815.2012.659382 Zhang Q et al (2014) Molecular analyses of 6 different types of uterine smooth muscle tumors: emphasis in atypical leiomyoma. Cancer 120:3165–3177. https:// doi.org/10.1002/cncr.28900 Zhang Q et al (2017) Fumarate hydratase mutations and alterations in leiomyoma with bizarre nuclei. Int J Gynecol Pathol. https://doi.org/10.1097/pgp. 0000000000000447 Zivanovic O et al (2009) Stage-specific outcomes of patients with uterine leiomyosarcoma: a comparison of the International Federation of Gynecology and Obstetrics and American Joint Committee on Cancer staging systems. J Clin Oncol 27:2066–2072. https:// doi.org/10.1200/jco.2008.19.8366 Zivanovic O et al (2012) A nomogram to predict postresection 5-year overall survival for patients with uterine leiomyosarcoma. Cancer 118:660–669. https://doi. org/10.1002/cncr.26333

Diseases of the Fallopian Tube and Paratubal Region

11

Russell Vang

Contents Normal Fallopian Tube and Gross Examination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Embryology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Gross Anatomy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Histology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Gross Examination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

650 650 652 654 656

Nonneoplastic Lesions of Fallopian Tube . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Metaplasia, Hyperplasia, and Other Epithelial Changes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Endometriosis and Endosalpingiosis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Salpingitis Isthmica Nodosa (SIN) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Ectopic Pregnancy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Polyps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Infertility . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Issues Related to Sterilization Procedures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Salpingitis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Torsion, Prolapse, and Intussusception . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Congenital Anomalies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Vasculitis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

658 658 660 661 662 664 664 665 665 673 673 674

Neoplasms of the Fallopian Tube . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 674 Benign Neoplasms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 674 Malignant Neoplasms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 678 Gestational Trophoblastic Disease of the Fallopian Tube . . . . . . . . . . . . . . . . . . . . . . . . . . . 699 Paratubal Lesions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Adrenal Rests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Paratubal Cysts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Wolffian Adnexal Tumor (“Female Adnexal Tumor of Probable Wolffian Origin”; FATWO) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

700 700 700 701

R. Vang (*) Department of Pathology, Division of Gynecologic Pathology, The Johns Hopkins Medical Institutions, Baltimore, MD, USA e-mail: [email protected] # Springer Nature Switzerland AG 2019 R. J. Kurman et al. (eds.), Blaustein’s Pathology of the Female Genital Tract, https://doi.org/10.1007/978-3-319-46334-6_11

649

650

R. Vang Papillary Cystadenoma Associated with Von Hippel-Lindau Disease . . . . . . . . . . . . . . . . . 702 Other Paratubal/Para-ovarian and Pelvic Ligament Lesions . . . . . . . . . . . . . . . . . . . . . . . . . . . 703 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 704

The Italian physician and anatomist Gabriele Falloppio provided the first detailed and accurate description of the oviducts in humans in 1561 A.D. and designated it the Uteri Tuba (Graham 1951). This organ was eventually named after him. Since that time, a wide variety of nonneoplastic and neoplastic diseases of the fallopian tube has become recognized, but it is only recently that the pathogenesis of fallopian tube carcinoma is beginning to be understood. Surgical specimens removed specifically for lesions of the fallopian tube are much less common than specimens from other sites in the gynecologic tract; nonetheless, the fallopian tube is frequently examined by the surgical pathologist because it accompanies specimens which were removed for lesions of other gynecologic organs, and the tube plays an important role in reproduction, including problems related to infertility. The majority of fallopian tube lesions examined by the surgical pathologist are nonneoplastic. Benign and malignant tumors of the fallopian tube are uncommon, but, as discussed below, early carcinomas of the fimbriated end of the fallopian tube are becoming more frequently recognized because of complete examination of all fallopian tube tissue submitted as part of prophylactic bilateral salpingo-oophorectomy specimens or major resections for ovarian carcinoma. Also, intraepithelial carcinomas are detected as incidental findings in a small subset of routine specimens when the fallopian tubes are completely submitted for histologic examination. This chapter provides a detailed discussion of normal fallopian tube (embryology, gross anatomy, and histology) and gross examination, nonneoplastic lesions, benign and malignant tumors, and gestational trophoblastic disease of the fallopian tube. Paratubal/paraovarian and pelvic ligament lesions are presented as well.

Normal Fallopian Tube and Gross Examination Embryology Regardless of genetic sex, the paired müllerian (paramesonephric) ducts develop on the anterolateral surface of the paired urogenital ridges in both females and males beginning in the sixth week of embryonic life (O’Rahilly 1983, 1989; Robboy et al. 1982). At the cranial end of the urogenital ridge, the peritoneum gives rise to a population of epithelial cells which segregate from the peritoneal layer (Guioli et al. 2007). This new population proliferates and forms the longitudinally oriented müllerian ducts (Guioli et al. 2007; Orvis and Behringer 2007). The mesenchyme surrounding the luminal epithelial layer of the müllerian duct is also derived from the peritoneum. Cranially, the ducts open into the peritoneal cavity. Each of the paired ducts grows caudally in the urogenital ridge immediately lateral to and using the wolffian (mesonephric) duct as a guide. Spatially lateral to the cranial aspect of the wolffian ducts, the müllerian ducts then ventrally cross the wolffian ducts. The longitudinally oriented and caudal portions of the müllerian ducts now lie medial to the wolffian ducts as they enter the pelvis. The caudal ends of the müllerian ducts abut on the posterior wall of the urogenital sinus immediately between the two wolffian ducts. In the eighth week of embryonic life, these caudal ends of the paired müllerian ducts fuse with each other but are still separated by a septum (Figs. 1 and 2). All these developments occur in both female and male fetuses and are completed before the testis (if the embryo is male) begins to secrete müllerian inhibiting substance (MIS), also known as anti-müllerian hormone (In the absence of MIS, the müllerian ducts develop passively to form the fallopian tubes, uterus, and vaginal wall. Likewise, in the absence

11

Diseases of the Fallopian Tube and Paratubal Region

a

651

Abdominal ostium of uterine tube

b

Fimbriae Paroöphoron

Cortical cords of ovary

Epoöphoron

Corpus uteri

Mesonephros Cervix

Vagina Mesonephric duct

Gärtner’s cyst

Uterine canal Paramesonephric tubercle

Fig. 1 (a) Diagram of ventral aspect of coronal section through female embryo at the end of the eighth week. The arrangement of the müllerian (red) and wolffian (mesonephric) [blue] ducts is shown. The cranial portion of the müllerian duct is lateral to the wolffian duct. The former grows in a caudal direction and crosses ventral to the latter and is in a medial position at the caudal end. The caudal

a

ends of the müllerian ducts fuse, which eventually form the uterus. (b) The developed fallopian tube with accompanying wolffian remnants. The red and blue structures correspond to their precursors in (a). (From Sadler TW. Langman’s Medical Embryology, 6th Edition. Baltimore: Williams & Wilkins; 1990:Fig. 15–22. Printed with permission from Lippincott Williams & Wilkins)

b

c

Urogenital ridge Ovary

Paramesonephric duct

Fig. 2 Diagram of transverse section of female embryo. (a) through (c) show progressively lower levels through urogenital ridge. The müllerian (paramesonephric) ducts (orange) eventually fuse and are then located medial to the wolffian ducts (blue). The fusion of the müllerian

Mesonephirc duct

Broad ligament of Fused uterus paramesonephric ducts

ducts creates a transverse fold which becomes the broad ligament (c). (From Sadler TW. Langman’s Medical Embryology, 6th Edition. Baltimore: Williams & Wilkins; 1990:Fig. 15–23. Printed with permission from Lippincott Williams & Wilkins)

652

of testosterone, the wolffian ducts regress.) In the development of the female, the first two parts of the müllerian duct (the cranial longitudinal segment which opens into the peritoneal cavity and the transverse portion which crosses the wolffian duct) form the fallopian tube. The cranial-most aspect of the first part forms the fimbriated end, and the caudal-most segment of the müllerian duct (the fused portion) forms the uterus. During the growth of the second portion of the müllerian duct (the transverse segment which crosses the wolffian duct), the urogenital ridges form a transverse pelvic fold. After the fusion of the caudal segment of the müllerian duct, the transverse pelvic fold extends laterally from the fused müllerian duct toward the pelvic sidewall (Fig. 2). This fold forms the broad ligament, to which the fallopian tube is attached. The lumen of the fallopian tube is initially oval to round and lined by immature columnar epithelium, but the mucosa forms plicae at week 14. In week 16, the fallopian tube begins an active growth phase and starts to coil. Smooth muscle appears in the walls of the genital canal between 18 and 20 weeks. The fallopian tube muscular wall develops only around the müllerian duct, so that the wolffian duct remnants are external to the true wall of the canal. From the 22nd to 36th weeks, there is an increase in the growth and coiling of the fallopian tube at a rate of approximately 3 mm/week (Hunter 1930). Fimbriae do not develop until the 20th week, at which time only 3–4 are present in each fallopian tube (Sulak et al. 2005). The fimbriae increase in number throughout gestation, and at term, 6–8 are present in each tube. The number continues to increase after birth. Important genes in the embryologic development of the müllerian duct include the Wnt family, Lim1, Pax2, and Emx2 (Yin and Ma 2005). In addition, the Hox family of genes (Hox in mice, HOX in humans) is particularly essential for development of this anatomic structure (Du and Taylor 2004; Taylor et al. 1997; Yin and Ma 2005). The Hox family represents four clusters of genes (Hoxa through Hoxd) which encode transcription factors that direct embryogenesis. Their main function is to control patterning and

R. Vang

positional identity along a developing axis, such as the hindbrain, axial skeleton, and limbs. One of the anatomic sites controlled by the Hoxa cluster during embryogenesis is the müllerian duct, in which each Hoxa gene controls the morphogenesis of different segments along the developing axis of the müllerian duct. Hoxa-9 through Hoxa-13 are sequentially located in tandem with one another in the same region of the chromosome. It has been shown in mice that the physical order of the Hoxa genes on the chromosome corresponds to the same spatial order of the different segments of the developing müllerian duct (i.e., Hoxa-9 is expressed in the fallopian tube, Hoxa10 and Hoxa-11 in the uterine corpus, Hoxa-11 in the uterine cervix, and Hoxa-13 in the upper vagina). This same spatial organization of Hoxa genes with their respective derivatives of the different segments of the müllerian duct is also maintained in humans (Taylor et al. 1997). Thus, interaction between HOXA-9 and presumably several other non-HOX genes determines the proper development of the human fallopian tube.

Gross Anatomy The fallopian tube is located anterior to the ovary. The tube extends medially from the area of its corresponding ovary to its origin in the posterosuperior aspect of the uterine fundus. In an adult during the reproductive years, its length is usually between 9 and 12 cm. The tube at the ovarian end opens to the peritoneal cavity and is composed of about 25 finger-like extensions of the tube – the fimbriae. The fallopian tube consists of five main segments. From medial to lateral, they are the intramural (interstitial) portion, isthmus, ampulla, infundibulum, and fimbriated end (Fig. 3). The fimbriae attach to the expanded end of the tube, the infundibulum, which is about 1 cm long and 1 cm in diameter. The infundibulum lies within a few millimeters of the lateral or tubal end of the ovary. It narrows gradually to about 4 mm in diameter and merges medially with the ampullary portion of the tube, which extends about 6 cm, passing anteriorly around the ovary. At a point characterized by relative thickening of the

11

Diseases of the Fallopian Tube and Paratubal Region

653

Fig. 3 Posterior aspect (upper) and coronal section (lower) of fallopian tube including anatomic relationships with adjacent structures. All five segments of the fallopian tube (intramural segment, isthmus, ampulla, infundibulum,

and fimbriated end) are illustrated. (From Netter FH. Atlas of Human Anatomy. West Caldwell: CIBA-GEIGY Corporation; 1989:Plate 350. Printed with permission from Elsevier, Inc. All rights reserved)

muscular wall along with a smaller diameter compared with the ampulla, the isthmic portion begins and extends about 2 cm toward the uterus. Within the myometrium, the tube extends as a 1 cm-long intramural segment until it joins the extension of the endometrial cavity at the uterotubal junction. Throughout its extrauterine course, the tube lays in a peritoneal fold along the superior margin of the broad ligament – the mesosalpinx. The arterial blood supply has a dual origin from branches of the ovarian and uterine arteries. Tubal branches of the uterine artery pass in the mesosalpinx laterally from the cornu of the uterus to anastomose with tubal branches of the ovarian

artery. Venous drainage parallels the arterial supply via anastomosing tubal branches of uterine and ovarian veins, also located in the mesosalpinx. The arterial and venous distributions for lateral portions of the tube are supplied by the ovarian vessels whereas the uterine vessels supply the medial portions of the tube. Drainage from the ovarian veins is to the inferior vena cava on the right and renal vein on the left. Drainage from the uterine plexus is to the internal iliac vein. Tubal lymphatics typically drain into ovarian and uterine vessels. The former and latter drain into the para-aortic and internal iliac lymph nodes, respectively.

654

The nerve supply of the tube is both sympathetic and parasympathetic. Sympathetic fibers from T10 through L2 synapse in the celiac, aortic, renal, inferior mesenteric, cervicovaginal, and possibly presacral plexuses. Sensory pain fibers pass along with the sympathetic nerves to the spinal cord at the level of T10-T12. Parasympathetic fibers from the vagus nerve supply the lateral portion of the tube via postganglionic fibers from the ovarian plexus whereas the medial portion is innervated via S2-S4 parasympathetic fibers synapsing in the pelvic plexuses.

Histology A mucosa, wall of smooth muscle (muscularis or myosalpinx), and serosa constitute the three layers of the fallopian tube. The mucosal layer lies directly on the muscularis. It consists of a luminal epithelial lining and a scanty underlying lamina propria containing vessels and spindle or oval mesenchymal cells. Although the lamina propria may be small in area, this is the site of decidua in 5–12% of postpartum tubes (Fig. 4), and decidua may be seen in 80% of tubes removed for ectopic pregnancy (Green and Kott 1989). The stroma of the plicae of the fallopian tube tends to be more fibrotic in the postmenopausal years. The mucosa increases significantly in its gross structural complexity as the lumen enlarges from the uterine to the ovarian ends. The interstitial/intramural portion contains a mostly flat mucosa with minimal undulation. It is lined by endometrium in the most proximal portion at the junction of the endometrial cavity and tubal ostium. Farther away from the tubal ostium, the mucosa of the interstitial/intramural portion is lined by an epithelial lining that is more typical of distal portions of the tube but with lesser numbers of ciliated cells. The isthmus shows slightly greater undulation than is seen in the interstitial/intramural portion and contains a limited number of blunted plicae (Fig. 5). In the ampulla, the plicae are frond-like and delicate, and both secondary and tertiary branches may be appreciated (Fig. 6). The infundibular and fimbriated end plical patterns are similar to that of the ampulla except that the plicae of the

R. Vang

Fig. 4 Decidua. The plicae are expanded due to decidual change within the lamina propria. The decidual cells have cytologic features similar to those of endometrial decidua. Scattered lymphocytes are present in the background

Fig. 5 Isthmus. The appearance is similar to the interstitial (intramural) segment except that the mucosa shows slightly more undulation with a limited number of blunted plicae

fimbriated end are essentially exophytic and have no underlying smooth muscle wall. A distinct fimbria, the fimbria ovarica, runs from the tubal ostium to one pole of the ovary and is involved in ovum pick-up, in which there appears to be a realignment of fimbriae in their relationship to the ovary itself. However, it should be noted that histologic characterization of this structure is very limited (Okamura et al. 1977). The epithelium of the mucosa is composed of a single layer of cells, or it may be pseudostratified. It predominantly consists of ciliated and secretory

11

Diseases of the Fallopian Tube and Paratubal Region

Fig. 6 Ampulla. In comparison with the interstitial (intramural) segment and isthmus, the ampulla has a greater diameter of the entire cross section of the tube, greater diameter of the lumen, and thinner muscularis. The plical architecture of the ampulla is more complex than that of the interstitial (intramural) segment and isthmus

Fig. 7 Mucosal epithelium of the fallopian tube. The epithelium contains a mixture of ciliated (arrow) and secretory (arrowhead) cells

cells; the latter are more numerous (Fig. 7). A third cell, the intercalated (“peg”) cell, is thought to exist. Some believe that this is a variant of the secretory cell, but it is not reliably identified on hematoxylin and eosin (H&E) sections. Ciliated cells are more abundant in the lateral portions of the tube, and the secretory cells are more numerous in the medial portions. The ciliated cell is columnar or round and has a mild to moderate amount of eosinophilic or clear cytoplasm. The nucleus is oval to round, and the chromatin is moderately granular and slightly basophilic. Ultrastructurally, each of the cilia is composed

655

of a central pair of microtubules that is surrounded by nine outer doublet microtubules. In Kartagener’s syndrome, where cilia are scanty and structurally/functionally defective, fertility is impaired but not abolished (Halbert et al. 1997). The reader is referred elsewhere for additional details regarding the structure and physiology of fallopian tube cilia (Lyons et al. 2006). The secretory cell also is columnar and approximately the same height as the ciliated cell but often narrower with scant eosinophilic cytoplasm. Its nucleus is columnar, and it is thinner and slightly darker than the nucleus of the ciliated cell. Immunohistochemically, the normal mucosal epithelium of the fallopian tube frequently and diffusely expresses WT-1, estrogen receptor (ER), and progesterone receptor (PR). The Ki-67 labeling index is typically 20 cigarettes/day), pelvic inflammatory disease, multiple spontaneous abortions (3), increasing age (>40 years), prior medically induced abortion, infertility (>1 year), multiple sexual partners (>5), and previous IUD use (Bouyer et al. 2003).

Clinical Features Currently, ectopic pregnancy accounts for 1–2% of clinically known pregnancies (ACOG 2008; Farquhar 2005; Van Den Eeden et al. 2005). Simultaneous ectopic and intrauterine implantations (heterotopic pregnancy) used to occur in 1 in 30,000 pregnancies decades ago; however, the frequency now can be as high as 0.75–1% of pregnancies after undergoing assisted reproductive technology (Habana et al. 2000; Marcus et al. 1995). The classic presentation of ectopic pregnancy includes amenorrhea with subsequent vaginal bleeding and/or abdominal pain. Tubal rupture is associated with intra-abdominal hemorrhage. The frequency of left- versus right-sided ectopic tubal pregnancies is similar, but they are slightly more common on the right (Breen 1970; Brenner et al. 1980). Rare cases are bilateral. Serial serum beta-human chorionic gonadotropin (β-hCG) measurements and transvaginal ultrasonography are important parts of the clinical evaluation. Management typically consists of either surgery (salpingectomy or salpingostomy) or medical therapy (methotrexate). Incomplete removal of trophoblastic tissue may result in persistent ectopic pregnancy, which occurs in 2–11% and 4–20% of cases after laparotomy with salpingostomy and laparoscopic salpingostomy (Farquhar 2005; Fylstra 1998). These figures are similar to the frequency of failure with systemic methotrexate therapy. In some cases, persistent ectopic pregnancy may be a result of spillage of gestational tissue from disruption/morcellation of the specimen after salpingostomy or salpingectomy. In such instances, the lesional

11

Diseases of the Fallopian Tube and Paratubal Region

663

Fig. 18 Ectopic pregnancy. External view of dilated fallopian tube containing an ectopic pregnancy. The cross section will show a hemorrhagic cut surface

tissue may be found as nodules/implants on pelvic, omental, or uterine serosal surfaces (Cataldo et al. 1990; Doss et al. 1998).

Pathologic Features The unruptured tubal pregnancy is characterized grossly by a somewhat irregular elongated dilatation of the tube, with a blue discoloration caused by hematosalpinx (Fig. 18). Within the tube, most ectopic pregnancies are found in the ampulla (~80%), isthmus (12%), and fimbriae (5%) (Breen 1970). Nearly two-thirds of cases contain a grossly or microscopically identifiable embryo. Chorionic villi usually are found in the blood-filled and dilated tubal lumen and, in 75% of cases, appear viable. Implantation is deeper and more apt to be associated with a viable pregnancy when the placentation occurs on the mesosalpingeal side of the tube as compared with the anti-mesosalpingeal side (Kemp et al. 1999). The extra-villous intermediate trophoblast of the conceptus penetrates deep in the tubal wall. On occasion, this proliferation may exhibit diffuse sheet-like architecture, which can raise concern for a gestational trophoblastic neoplasm or hydatidiform mole (Burton et al. 2001; Sebire et al. 2005); however, this appearance is within the range of proliferation that can be encountered in ectopic pregnancies (Fig. 19). Perhaps because of the limited ability of the endosalpingeal stroma to undergo decidualization, and analogous to a placenta increta, the chorionic villi invade muscularis and then serosa (Pauerstein et al. 1986). Another major difference compared with uterine implantation is the failure of tubal

Fig. 19 Exuberant intermediate trophoblast proliferation in an ectopic pregnancy. The sheet-like architecture and atypical epithelioid cells should not be mistaken for a gestational trophoblastic neoplasm or hydatidiform mole. Chorionic villi are present on the left side of the photograph

trophoblast to differentiate into chorion frondosum and chorion laeve (Randall et al. 1987), but a gestational sac can be seen (Pauerstein et al. 1986). Vascular changes in mid-sized tubal arteries adjacent to ectopic pregnancies are similar to those found in the vessels near uterine implantations, with invasion by intermediate trophoblast, proliferation of the vascular intima, and accumulation of foam cells in the intima. Chronic salpingitis is found in nearly half the patients, with a reported range of 29–88% (Green and Kott 1989). SIN may also be present. AriasStella reaction can be seen in the fallopian tube mucosa (Milchgrub and Sandstad 1991). Clear cell hyperplasia has been reported (Tziortziotis et al. 1997).

664

The clinicopathologic features of extratubal ectopic pregnancy vary according to the site (Oliver et al. 2007). Cornual or interstitial pregnancies may expand up to about 12 weeks, when rupture may lacerate one of the uterine arteries as well as the entire side of the uterus. Cervical ectopic pregnancy presents with bleeding similar to an incomplete abortion. Because of the nature of the cervical tissue underlying placental implantation, control of bleeding may be difficult. Ovarian pregnancy is clinically similar to tubal pregnancy, including frequent preoperative rupture. More than half the patients in one series had a history of previous reproductive tract disease or infertility (Grimes et al. 1983). Macroscopic examination typically reveals a hemorrhagic mass replacing the ovary. Pathologic criteria for ovarian pregnancy have been proposed: (1) the tube must be intact and separate from the ovary; (2) the gestational sac must occupy the normal position of the ovary; (3) the gestational sac must be connected to the uterus by the utero-ovarian ligament; and (4) ovarian tissue must be demonstrated within the wall of the sac (Grimes et al. 1983). Pathologic documentation of ovarian tissue within the pregnancy may be difficult or impossible if treatment consists of conservative resection or if the pregnancy has extensively replaced the ovarian tissue.

Sequelae The natural history of tubal ectopic pregnancy includes spontaneous expulsion from the fimbriated end (tubal abortion), as well as embryonal death and involution of the conceptus. Typically, however, continued growth of the trophoblast leads to increasing dilatation and weakening of the muscularis, with rupture at about the eighth week. About 25% of tubal pregnancies have ruptured by the time of diagnosis (Falcone et al. 1998). Hemorrhage due to rupture may be massive, and ectopic pregnancy is a major cause of maternal mortality during pregnancy. Rare ectopic pregnancies have proceeded to term with fetal viability. In grossly normal fallopian tubes, chorionic villi or placental site nodules may be found, indicative of a prior unsuspected ectopic pregnancy (Jacques et al. 1997; Nayar et al. 1996). A subset

R. Vang

of tubal pregnancies forms a mass that, with involution of trophoblast and reestablishment of the menstrual cycle, may present problems in differential diagnosis. This convoluted, blood-filled tube (including organization, variable inflammation, and adhesions), often with involved ipsilateral ovary, may simulate a tumor or an endometrioma. Most, but not all, patients will have detectable serum β-hCG. Extensive microscopic sampling of this so-called chronic ectopic pregnancy may be required to demonstrate trophoblastic tissue, which may consist of nonviable chorionic villi (Ugur et al. 1996). In more advanced pregnancy, death of the fetus with retention in the extrauterine location may be followed by calcification of the fetus (lithopedion) or both membranes and fetus (lithokelyphopedion).

Polyps Polyps of the fallopian tube have been classified as tumors in other textbooks on gynecologic pathology; however, they are included in the nonneoplastic section of this chapter for conceptual purposes. They are found in approximately 1–3% of women undergoing hysterosalpingography for infertility and may cause proximal tubal occlusion (David et al. 1981; Fernstrom and Lagerlof 1964). Their causality of and relationship with infertility has been debated in the literature. They are typically small and preferentially occur in the intramural segment of the fallopian tube, particularly at the tubal ostium. Polyps are frequently bilateral. Although they are removed for microscopic examination only rarely, histologically they are of endometrioid type and resemble endometrial polyps (Lisa et al. 1954). Given that endometriosis of the tubal mucosa is found in some patients, it is possible that tubal polyps might represent a microscopic form of polypoid endometriosis.

Infertility Most of the diseases discussed in this chapter may result in sufficient anatomic distortion to cause tubal infertility. In contrast, purely physiologic

11

Diseases of the Fallopian Tube and Paratubal Region

tubal dysfunction is not well defined but may be illustrated by the immotile cilia of Kartagener’s syndrome that can lead to reduced fertility – only 3 of 12 women in one series succeeded in becoming pregnant (Afzelius and Eliasson 1983). Paratubal or fimbrial adhesions secondary to endometriosis, prior pelvic inflammatory disease, or appendicitis may interfere with normal tubal motility and ovum pick-up. For a detailed discussion of the pathophysiology of fallopian tube cilia in various diseases in relation to infertility, the reader is referred elsewhere (Lyons et al. 2006). Obliterative fibrosis (possibly secondary to inflammation within the uterus) or polyps at the uterine tubal ostium may lead to obstruction at the uterotubal junction (Fortier and Haney 1985; Lee et al. 1997).

Issues Related to Sterilization Procedures Sterilization by interference with tubal function involves procedures designed to damage or obstruct the mucosa or lumen of the fallopian tube by surgical removal of a segment of the tube (bilateral partial salpingectomy), rings/ clamps, electrocoagulation, intratubal chemical methods (e.g., silicone plugs and methylcyanoacrylate), or intratubal mechanical devices (e.g., Essure System) (Clark et al. 2017; Donnez et al. 1979; Stock 1983). Tubal resection should be confirmed by histologic demonstration of a cross section of the fallopian tube including the entire lumen. On occasion, histologic sections may only show fibromuscular/fibrovascular tissue without fallopian tube mucosa. Such cases may represent pelvic ligaments or vessels which were clinically mistaken for a fallopian tube. In order to completely evaluate those cases, it may be necessary to cut deeper levels from the paraffin block, including the possibility of reorienting and reembedding the tissue in the block so as to entirely cut through it and find a fallopian tube lumen. The above protocol also applies to cases in which a fallopian tube is definitely present histologically but for which a complete cross section of the lumen is not seen on the H&E slide.

665

In women that were initially treated by surgical resection, spontaneous reanastomosis or fistula formation, which may lead to fertilization and ectopic or intrauterine pregnancy, are common mechanisms of sterilization failure (Soderstrom 1985). To identify the cause of failure of tubal sterilization procedures, careful gross examination of the specimen, occasionally specimen salpingography, longitudinal orientation of the tubal segment in the paraffin block, and meticulous sectioning techniques may be necessary.

Salpingitis Salpingitis consists of three major types: acute, chronic, and granulomatous/histiocytic. On occasion, some cases will have mixed features, but this section is arranged according to the predominant histologic appearance.

Acute Salpingitis Acute salpingitis is a purulent inflammatory process usually secondary to the passage of bacteria from the cervix and uterine cavity into the tubal lumen (Lareau and Beigi 2008; McCormack 1994). It is the pathologic correlate of the clinical entity, pelvic inflammatory disease. It typically occurs in young, sexually active, and reproductive-age women. Important risk factors include patterns of sexual behavior and contraceptive use. Grossly, the fallopian tube is enlarged and edematous (Fig. 20). The serosa is erythematous and may be covered with fibrinopurulent exudates. Pus may also fill the lumen. Histologically, the fallopian tube lumen, mucosa, and wall contain neutrophils, fibrinous debris, and ulceration (Fig. 21). Edema and lymphocytes may be present as well. Mucosal hyperplasia and distortion can be seen. The histologic appearance varies according to the severity and phase of the disease. The appearance may also vary somewhat based on the specific causative microbial agent, as discussed below. Significant fallopian tubespecific sequelae include infertility and ectopic pregnancy. It is not clear if organisms are carried upward by sperm or trichomonads as vectors or whether

666

R. Vang

Fig. 20 Gross appearance of fallopian tube containing acute salpingitis. The tube is enlarged with a dilated lumen and erythematous mucosa. The wall is edematous and thickened. Other examples may contain pus in the lumen

Fig. 21 Acute salpingitis. (a) The mucosa shows distorted architecture and abundant inflammation, including pus in the overlying lumen. (b) The inflammatory

component is mixed but mostly composed of abundant neutrophils

some form of passive transport is in effect (Keith et al. 1984). The bacteria implicated in acute salpingitis appear to be from two sources: sexual transmission and lower genital tract flora. Although Neisseria gonorrhoeae and Chlamydia trachomatis have been considered the most common causative organisms, meticulous bacteriologic studies indicate that most cases are polymicrobial and that anaerobic bacteria, especially Bacteroides species and peptostreptococci, frequently are present, as well as aerobes such as Escherichia coli. The presence of serum antibodies against gonococcal pili in some of these women, however, suggests that gonococci may

initiate the process, only to be supplanted by anaerobes. The role of mycoplasmas in acute salpingitis is controversial, and viruses do not appear to be etiologic. However, Herpes simplex virus infection involving the mucosa of a prolapsed tube with mixed acute and chronic inflammation has been reported (Lefrancq et al. 1999). The gonococcus gains access to the tube most readily at the time of menstruation. This situation corresponds to the typical clinical presentation in which the onset of acute pain occurs a few days after menses. The onset of non-gonococcal, non-chlamydial acute salpingitis is not, however, clearly related to the recent onset of menses

11

Diseases of the Fallopian Tube and Paratubal Region

(Sweet et al. 1986). Elegant in vitro studies by Ward et al. (1974) have clarified the likely initial steps in gonococcal infection, and the molecular mechanisms involved have been reviewed elsewhere (Edwards and Apicella 2004; Lyons et al. 2006). N. gonorrhoeae perfused through the lumen of cultured whole tubes attach only to non-ciliated cells. Within 3 h, microvilli from the cells appear to embrace the gonococci and adhere to them. The bacteria then penetrate both the cells and intercellular junctions, with cell lysis and sloughing. Adjacent ciliated cells are also destroyed but are not invaded directly. Gonococcal lipopolysaccharide and gonococcal-induced tumor necrosis factor-alpha and various other cytokines cause much of the epithelial damage (Maisey et al. 2003; McGee et al. 1999), and the degree of pathogenicity likely depends on the bacterial as well as the host genome (Arvidson et al. 1999). After cell lysis, the bacteria penetrate the subepithelial connective tissue. In vivo, this process is considerably modified by the host response. N. gonorrhoeae spreads via the epithelial surface and thus causes mucosal damage. A brisk diapedesis of granulocytes from capillaries into the mucosa and lumen occurs, and there is vascular engorgement and edema of all tubal layers. As the lumen fills with granulocytes and cellular debris, and as the tube distends, pus may be seen dripping from the fimbriated end in patients undergoing laparoscopy. In severe cases, transudation of plasma proteins results in a fibrinous exudate on the serosal surface, which is erythematous because of vascular dilatation. The cell necrosis, distension of the tube, and focal peritonitis account for the symptoms of abdominal and pelvic pain. Over time, repeated infections result in recurrent symptoms as well as the anatomic changes of chronic salpingitis, discussed below. Chlamydia trachomatis is cultured frequently from the cervix, uterus, and fallopian tubes in women with acute salpingitis (Kristensen et al. 1985; Mardh et al. 1977). It is thought that the damage of the fallopian tube by Chlamydia is due to the 60 kDa chlamydial heat shock protein (hsp60), as well as other various cytokines (Hafner 2015; Lyons et al. 2006). The histologic

667

appearance of tubes removed during the acute or subacute phase of chlamydial salpingitis is virtually identical to that caused by the gonococcus. There is an initial transmural and mucosal infiltration of neutrophils with an intraluminal exudate. Subsequently, there is a lymphoplasmacytic response with variable numbers of residual granulocytes. Chlamydial inclusion bodies have been identified within the epithelial cells (Winkler et al. 1985). On occasion, the lymphofollicular response may be so florid as to suggest lymphoma (Wallace and Hart 1991). As a result of acute salpingitis (usually in the context of gonococcal or chlamydial disease), fibrinous adhesions develop between the fallopian tube serosa and surrounding peritoneal surfaces. Peritoneal inflammation may be widespread, and thin so-called violin-string adhesions may form between the liver and anterior abdominal wall as part of the Fitz-Hugh-Curtis syndrome. In severe cases of acute salpingitis (as well as with chronic or granulomatous salpingitis) with involvement of the ovary regardless of the specific microorganism, both the ovary and tube are attached to one another by adhesions and create a mass in the form of a tubo-ovarian abscess (Fig. 22). Tubo-ovarian abscesses can be unilateral or bilateral. Histologically, the inflammatory component may be predominantly composed of

Fig. 22 Gross features of tubo-ovarian abscess. In this case, bilateral tubo-ovarian abscesses are composed of fibroinflammatory masses, and the ovary and fallopian tube on each side are attached to one another by adhesions. The cut surface will show distorted fibrous tissue with edema, hemorrhage, and necrosis

668

neutrophils or contain a mixture of neutrophils, lymphocytes, histiocytes, and plasma cells. Widespread necrosis is common. The background fallopian tube and ovarian parenchyma will be markedly distorted, and abundant fibrous and edematous stroma may be present. Although N. gonorrhoeae and C. trachomatis are common bacterial causes of acute salpingitis, they are isolated in culture only rarely from tubo-ovarian abscesses. Both aerobic and anaerobic cultures of any tubo-ovarian abscess should be obtained in the operating room or laboratory. Prior treatment with antibiotics possibly may eliminate culturable organisms, but anaerobes are commonly isolated. E. coli, Bacteroides fragilis and other Bacteroides species, Peptostreptococcus, Peptococcus, and aerobic streptococci are the most commonly found organisms (Landers and Sweet 1983). Typically, these infections are polymicrobial. Another organism that can result in tuboovarian abscess is Actinomyces israelii, which is part of the indigenous female genital tract flora (Persson and Holmberg 1984). Actinomycotic infections of the tube are associated with intrauterine contraceptive devices (IUDs) (see ▶ Chap. 7, “Benign Diseases of the Endometrium”) (Dische et al. 1974). Anaerobic culture is necessary to permit growth of Actinomyces israelii. Microscopically, fragments of grampositive filaments and sulfur granules may be recognized within pus. Pseudoactinomycotic radiate granules should not be mistaken for the sulfur granules of actinomycosis (Bhagavan et al. 1982; Pritt et al. 2006). An asymptomatic form of acute salpingitis (“physiologic salpingitis”) is seen in tubes removed during postpartum ligation. Beginning about 5 h after delivery and present up to 7–10 days later, a small number of acute or mixed acute and chronic inflammatory cells are found in the mucosa or lumen of 10% or more of specimens (Fig. 23). Attempts to culture aerobic or anaerobic bacteria have been almost uniformly unsuccessful. The process may be secondary to the trauma of delivery or intrauterine tissue necrosis and is apparently of no clinical significance.

R. Vang

Fig. 23 Physiologic “salpingitis.” A mild amount of acute inflammation is present within the vascular spaces in the mucosa. Other cases may have mild acute inflammation within the mucosal epithelium or lamina propria

Chronic Salpingitis As a result of acute salpingitis, the proximity of the ovary to the fimbriae allows multiple tuboovarian adhesions to form, which may also cause occlusion of the tubal ostium. If the fimbriae close before the ovary is involved as part of a tuboovarian abscess, the inflamed and dilated tube can form a pyosalpinx full of acute and chronic inflammatory cells. When acute salpingitis resolves, the acute and most of the chronic inflammatory cells gradually disappear, and the patient is left with either a severely scarred tube in the form of chronic salpingitis or a hydrosalpinx. C. trachomatis DNA has even been detected in fallopian tubes in a subset of cases that contained only chronic salpingitis (Hinton et al. 2000). Thus, the finding of chronic salpingitis may imply previous pelvic inflammatory disease in some patients. In chronic salpingitis, the mucosal plicae are often adherent to one another secondary to surface fibrin deposition from acute salpingitis. This may be focal or extensive. If it is severe enough, the bases of the fimbriae may coalesce in the center with the fimbriae radiating outward, or the tips of the fimbriae may adhere blocking the lumen and causing a blunted end – the so-called clubbed tube (Fig. 24). Healing and organization in the non-fimbriated portions of the tube also lead to permanent bridging between folds. Classically,

11

Diseases of the Fallopian Tube and Paratubal Region

Fig. 24 So-called clubbed tube. The fimbriated end is closed because of fimbrial adhesions, creating a blunted end

669

communication usually can be demonstrated between dilated and non-dilated portions of the tube, the etiology of the dilatation can be obscure. The dilated tube can become cystic and filled with serous fluid, and the wall is generally white, thin, and translucent with occasional fibrous adhesions on the external surface. The muscle wall is either thin and atrophic or replaced by fibrous tissue. Most of the epithelial lining consists of low-cuboidal cells, but an occasional plica may persist with columnar epithelium containing histologically normal ciliated and secretory cells (Fig. 28). A few lymphocytes may be found in the wall of the hydrosalpinx but are more commonly absent.

Granulomatous/Histiocytic Salpingitis and Foreign Bodies Granulomatous and histiocytic inflammation of the fallopian tube may result from infection by a number of different organisms or be induced by a variety of noninfectious processes, some of which include tissue reactions due to microscopic foreign bodies.

Fig. 25 So-called follicular salpingitis. The plicae are adherent to one another, creating follicle-like spaces in the setting of chronic salpingitis

this results in so-called follicular salpingitis (Fig. 25); however, that term is a misnomer as it suggests a pattern of inflammation characterized by lymphoid follicles. In chronic salpingitis, plicae may retain much of their size and shape, but plasma cells, lymphocytes, or both are still present in the mucosa (Fig. 26). Often the height of the folds appears reduced, and the plicae may become blunted and have fibrous stroma. Therefore, the once orderly pattern of the mucosa becomes distorted. The mucosa may also be hyperplastic. Hydrosalpinx is characterized by obliteration of the fimbriated end and dilation of the tube, usually the ampullary and infundibular portions (Fig. 27). If the ovary is first involved by tuboovarian adhesions, the ovary may be compressed by the dilated tube. Because a luminal

Pseudoxanthomatous/ Xanthogranulomatous Salpingitis Pseudoxanthomatous salpingitis (variably referred to as “pigmentosis tubae”) is characterized by lipofuscin- and hemosiderin-laden macrophages within the lamina propria of the mucosa of the fallopian tube, including distension of the plicae, and is associated with endometriosis (Fig. 29) (Clement et al. 1988; Furuya et al. 2002; Herrera et al. 1983; Munichor et al. 1997; Seidman et al. 1993; Seidman and Woodburn 2015). The tubes may be enlarged and edematous, with the mucosa having a dark brown polypoid gross appearance. Despite the association with endometriosis, this process also might result from salpingitis with associated bleeding (Clement et al. 1988; Seidman et al. 1993). It has been suggested that pseudoxanthomatous salpingitis should be distinguished from xanthogranulomatous salpingitis because of the latter’s association with pelvic inflammatory disease and lack of association with endometriosis (Furuya et al. 2002). In contrast with pseudoxanthomatous

670

R. Vang

Fig. 26 Chronic salpingitis. (a) Fibrotic and blunted plicae. (b) The distorted plicae show the lamina propria filled with lymphocytes and plasma cells

Fig. 27 Gross features of hydrosalpinx. The fallopian tube is massively dilated, producing a large cystic mass

Fig. 29 Pseudoxanthomatous salpingitis. The plicae are expanded and distorted due to sheets of histiocytes with eosinophilic cytoplasm in the lamina propria. This should not be mistaken for decidualization

salpingitis, xanthogranulomatous salpingitis has mucosa which is usually grossly yellow and purulent, macrophages which are foamy (as opposed to the dark brown macrophages in pseudoxanthomatous salpingitis), and other types of inflammatory cells, including multinucleated giant cells (Furuya et al. 2002; Ladefoged and Lorentzen 1988). A potentially related lesion that has been described in the tube is xanthelasma (Chetty et al. 2003).

Fig. 28 Hydrosalpinx. Most of the dilated fallopian tube contains a thin wall and atrophic mucosa. Residual small plicae are present. Note smooth muscle within the wall

Tuberculous Salpingitis Mycobacterium tuberculosis historically has been the predominant etiologic agent of

11

Diseases of the Fallopian Tube and Paratubal Region

granulomatous salpingitis. The frequency of tuberculous salpingitis in women studied for tubal causes of infertility ranges from much less than 1% in the United States to nearly 40% in India (Parikh et al. 1997). Twenty percent of women who die of tuberculosis have tubal involvement (Schaefer 1970). Primary infection of the genitalia, as by coitus with a partner with genitourinary tuberculosis, is extremely rare. Secondary spread via a hematogenous route, usually from a primary pulmonary infection, is the usual route of infection. For unclear reasons, the blood-borne organism prefers the tubes rather than other parts of the female genital tract. The primary pulmonary lesion may not be radiologically evident, but extrapulmonary involvement of the peritoneum, kidneys, or other sites may be present. Lymphatic spread from primary intestinal tuberculosis or direct spread from the urinary bladder or gastrointestinal tract may occur. Tubal involvement is usually bilateral. Although the earliest pathologic lesions are microscopic, with advancing disease the tube increases in diameter and may become nodular, mimicking SIN. In the more common adhesive form of the disease, multiple dense adhesions may form between the tube and ovary, and the fimbriae and ostium may be obliterated (Haines 1958). With the exudative form of disease, progressive distension mimics bacterial pyosalpinx. Hematosalpinx, hydrosalpinx, tuboovarian abscesses, or a so-called frozen pelvis may be found late in the disease process (Parikh et al. 1997). In either form, serosal tubercles may be present. In early disease, microscopic lesions are mucosal-based with a typical granulomatous reaction of epithelioid histiocytes and lymphocytes arranged in a nodular configuration. Multinucleated giant cells are often seen, and focal or massive central caseation may be present (Fig. 30). Immunosuppression can modify cellular immunity to a point where granulomas fail to form, and with this clinical information, the mere finding of acute and chronic inflammatory cells should lead to consideration of staining for acidfast organisms. From the mucosa, extension to the muscularis and serosa may occur. As the

671

Fig. 30 Granulomatous salpingitis due to tuberculosis. Noncaseating granulomas are present in the mucosa. Note calcifications and multinucleated giant cells. Other cases may contain caseating granulomas

tubercles enlarge and coalesce, they may erode through the mucosa and discharge their contents into the tubal lumen, and the tube may then become dilated. The mucosal inflammatory reaction leads to progressive scarring, with plical distortion and agglutination. Calcification can occur in areas of fibrosis. Because tubercles may not be present in a given section, the presence of caseation, fibrosis, or calcification in a tube may be the only histologic finding that suggests the need for more thorough evaluation. Notable mucosal distortion may result in hyperplasia mimicking carcinoma. There are several complications of tuberculous salpingitis. Alteration in function is expected, and sterility is almost universal because of the common bilaterality of the disease. Because of repeated seeding of the endometrium from the infected tubes, mycobacterial culture and the histologic finding of endometrial granulomas on curettage are diagnostically useful (see ▶ Chap. 7, “Benign Diseases of the Endometrium”). Parasitic Salpingitis Pinworm (Oxyuriasis)

The pinworm, Enterobius vermicularis, may migrate up the female genital tract, embed in the tube, and cause an inflammatory reaction. The tube can be involved with the ovary as a tubo-

672

R. Vang

may involve the female genital tract, including the adnexae. Cysticercosis (Taenia solium) also has been described in the tube (Abraham et al. 1982). Other rare parasites that have been reported in the fallopian tube include Entamoeba histolytica (amebiasis). Liesegang rings should not be mistaken for parasites (Clement et al. 1989).

Fig. 31 Granulomatous salpingitis due to pinworm. This caseating granuloma contains abundant eosinophils. A pinworm egg (arrow) is present within the necrotic zone

ovarian abscess, or a fibrous nodular area may be present. Acute and chronic inflammatory cells may be found together with eosinophils and portions of a gravid female worm. Ova can be released into the tissue, where they provoke a granulomatous reaction (Fig. 31), but identification of ova can be obscured by calcification and granulomas. Schistosomiasis (Bilharziasis)

Although tubal schistosomiasis may be one of the more common causes of granulomatous salpingitis worldwide, it is rare in the United States. In Africa, the fallopian tube is involved by schistosomiasis in 22% of all infected women (Gelfand et al. 1971). The ova of Schistosoma haematobium are most common, but Schistosoma mansoni eggs may be present in some women. Gross findings appear to be related to fibrosis surrounding the ova, producing a nodular or fibrotic tube. Histologically, the inflammatory reaction is typically granulomatous and may contain eosinophils, neutrophils, plasma cells, lymphocytes, and macrophages, including multinucleated giant cells. Some granulomas may be present within the stroma of the plicae and produce plical expansion. Ectopic pregnancy has been reported as occurring synchronously with tubal schistosomiasis in some patients. Other Parasites

Where the condition is common, hydatid disease secondary to Echinococcus granulosus infection

Fungal Salpingitis Other fungi which rarely can cause tubo-ovarian abscesses or granulomatous salpingitis include Blastomyces dermatitidis, Coccidioides immitis, Candida, and Aspergillus. These may be secondary to hematogenous spread or disseminated disease. Sarcoidosis Sarcoidosis of the tube is rare (Boakye et al. 1997). Histologically, noncaseating granulomas may be seen in the mucosa. Culture, histochemical stains, and clinical information are necessary to exclude other granulomatous diseases. Crohn’s Disease Crohn’s disease of the ileum, colon, or appendix may secondarily involve the tube and ovary to produce a granulomatous salpingitis and tuboovarian abscess. Noncaseating granulomas can involve the entire thickness of the tubal muscularis as well as the mucosa. The mucosa may exhibit hyperplasia with reactive atypia (Brooks and Wheeler 1977). Fistulas from bowel to tube also can occur. Other Types of Granulomatous/Histiocytic Lesions and Foreign Bodies Malakoplakia has rarely been reported in the fallopian tube. Some vasculitides have a granulomatous pattern (see “Vasculitis” section below) (Bell et al. 1986). Foreign body granulomas due to starch and talc, as well as pulse granulomas (Rhee and Wu 2006), can occur in the tube. In order to detect some foreign bodies, granulomatous or histiocytic reactions should be examined under polarized light. Extruded keratin from endometrioid carcinomas with squamous differentiation of the endometrium or ovary can produce keratin granulomas on the serosa or

11

Diseases of the Fallopian Tube and Paratubal Region

fimbriated end of the tube, as well as within the tubal lumen (Kim and Scully 1990). It should be noted, however, that not all foreign bodies produce a significant granulomatous or histiocytic response. Gelatin microsphere embolization particles (Embospheres ®/EmboGold ®) used in uterine artery embolization for treatment of uterine leiomyomas may sometimes be found in the fallopian tubes or ovaries because of the vascular anastomosis between the uterine and ovarian arteries (Kim et al. 2007). In the fallopian tube, these particles are typically found within arterial lumens in the outer wall of the tube or para-tubal locations. They usually elicit only a mild lymphocytic response with rare multinucleated giant cells as opposed to the marked multinucleated giant cell reaction with a granulomatous or histiocytic component seen with other foreign bodies.

Torsion, Prolapse, and Intussusception Among the various anatomic displacements of the tube, torsion is the most common. The usual predisposing factor is cystic enlargement of the ipsilateral ovary. A benign ovarian cyst or tumor is present in the majority of patients, but in a minority, a malignant ovarian tumor is the cause. Paraovarian cysts also are associated with torsion. Tubal enlargement secondary to hydrosalpinx or pyosalpinx, or previous gynecologic surgery (especially sterilization), are additional etiologies, but torsion may occur in the absence of apparent adnexal disease (Bernardus et al. 1984). The typical patient is in the reproductive years, occasionally pregnant, and complains of the sudden onset of lower abdominal pain. At operation, the adnexa on one side is twisted, usually once or twice. Venous outflow is compromised early, and the resulting congestion may lead to arterial compression. The adnexa often is enlarged, edematous, and dark and shows hemorrhagic infarction. If surgical intervention is prompt, the tube may be preserved. Asymptomatic or undiagnosed torsion can occur. Tubal prolapse into the vagina may occur rarely as a complication of vaginal or abdominal hysterectomy (see ▶ Chap. 3, “Diseases of the Vagina”) (Ouldamer et al. 2013; Ramin et al.

673

1999). Clinically, this is characterized by dyspareunia, vaginal bleeding/discharge, or abdominal pain beginning a few days to several years after hysterectomy. However, some women may be asymptomatic. On clinical examination, an excrescence is seen in the vaginal vault, suggestive of granulation tissue or carcinoma. Fimbriae may be apparent grossly. Severe acute and chronic inflammation can be present microscopically, and papillary architecture or pseudogland formation by the tubal epithelium can mimic adenocarcinoma. Due to admixed granulation tissue, it may be difficult to recognize the lesional tissue as a distorted segment of fallopian tube (Song et al. 2005); however, close scrutiny should reveal papillae lined by benign tubal epithelium. Depending on the specific differential diagnosis, immunohistochemical staining for WT-1, ER/PR, CK7, CK20, p16, and/or Ki-67 may be of help. Intussusception of the tube is rare. In one case, a para-ovarian cyst was engulfed by the end of the tube, and the fimbriated end was pulled into the ampulla (Adams 1969).

Congenital Anomalies Structural congenital anomalies of the fallopian tube are rare. Diethylstilbestrol (DES) use during pregnancy was discontinued decades ago, but surgical specimens from patients that were born during the DES era may still be examined today. In utero exposure to DES produces shortened, sacculated, and convoluted fallopian tubes. The fimbriae are constricted, and the os is pinpoint (DeCherney et al. 1981). The mucosa may be absent, or when it is present, the plicae do not develop (Robboy et al. 1982). Tubal duplication and accessory fallopian tubes are uncommon, but the latter occur more frequently (Beyth and Kopolovic 1982; Coddington et al. 1990; Daw 1973; Gardner et al. 1948). Absence of various segments of the fallopian tube (also variably referred to as atresia, hypoplasia, or interruption), absence of the ampullary muscularis, and complete absence of the tube have been described. These may be unilateral or bilateral, and they can occur with or without uterine

674

anomalies, such as unicornuate or bicornuate uterus (reviewed in Nawroth et al. (2006)).

Vasculitis The fallopian tube can be involved by vasculitis as part of localized or systemic disease, and involvement of the tube is less frequent compared with other sites within the gynecologic tract. Vasculitis involving the female genital tract can be of polyarteritis nodosa or giant cell arteritis types, but the former is more common (Ganesan et al. 2000; Hernandez-Rodriguez et al. 2009; Hoppe et al. 2007). Either one or multiple arterioles/small arteries may be involved. Clinical correlation is needed to determine whether or not vasculitis is localized to the gynecologic tract; however, it is uncommon for patients with vasculitis involving the female genital tract to have either a previous diagnosis of systemic disease or subsequent development of systemic disease (Ganesan et al. 2000; Hoppe et al. 2007).

Neoplasms of the Fallopian Tube The neoplasms arising in the fallopian tube include benign and malignant types, but malignant tumors are more common than benign ones; however, both are infrequent. They are commonly mistaken for chronic salpingitis or pyosalpinx, both preoperatively and during the operative procedure itself. Many benign tumors are small enough to be incidental findings at laparotomy. The World Health Organization (WHO) Classification of fallopian tube tumors is listed in Table 1 (Crum et al. 2014). Most of these are nonspecific histologic types since the same ones can be seen in other gynecologic sites, especially the ovary.

Benign Neoplasms Adenomatoid Tumor Adenomatoid tumor (benign mesothelioma) is the most frequent type of benign tubal tumor. Previously reported lymphangiomas probably

R. Vang Table 1 2014 WHO classification of tumors of the fallopian tube Epithelial Benign Papilloma Serous adenofibroma Precursor lesion Serous tubal intraepithelial carcinoma Borderline tumor Serous borderline tumor/atypical proliferative serous tumor Malignant Low-grade serous carcinoma High-grade serous carcinoma Endometrioid carcinoma Undifferentiated carcinoma Others Mucinous carcinoma Transitional cell carcinoma Clear cell carcinoma Mixed epithelial-mesenchymal Adenosarcoma Carcinosarcoma (malignant müllerian mixed tumor; MMMT) Mesenchymal Leiomyoma Leiomyosarcoma Others Mesothelial Adenomatoid tumor Germ cell Teratoma Mature Immature Lymphoid and myeloid Lymphomas Myeloid neoplasms

represent examples of this entity. They are usually only 1–2 cm in diameter, appearing as a nodule beneath the tubal serosa, and are yellow or whitegray on cross section. Rare cases are bilateral. Similar lesions may be found in the uterus, cul-de-sac, or ovary (see ▶ Chap. 10, “Mesenchymal Tumors of the Uterus”). Their origin is presumed to be from the cells of the serosal mesothelium. A fortuitous section may demonstrate a connection between serosa and tumor, but usually the serosa covers the lesion.

11

Diseases of the Fallopian Tube and Paratubal Region

675

Microscopically, the tumor may be large enough to displace the tubal lumen eccentrically and may grow into the supporting stroma of the luminal folds in an infiltrating manner. Multiple, small, slit-like, or ovoid tubules proliferate through the muscular wall of the fallopian tube; however, the stroma may be fibrous or hyalinized (Fig. 32). Foci of chronic inflammation can also be present. The tubules are lined by a single layer of low-cuboidal or flattened epithelial-like cells which contain abundant eosinophilic cytoplasm with variably sized vacuoles and round and bland nuclei. Mitotic figures are rare. The tubules may be empty or contain pale fluid. Infarction may occur in adenomatoid tumor. When this occurs, and when marked, there is the potential for confusion with other lesions, such as

adenocarcinoma, because of the obscured junction between adenomatoid tumor and nonneoplastic tissue, pseudoinfiltration, solid patterns in viable tumor, and reactive atypia (Skinnider and Young 2004). Histochemical studies have shown alcian bluepositive, hyaluronidase-digestible material in the cells and spaces; however, this material may be absent after routine processing for histologic sections. No significant glycogen or intracellular epithelial mucin is present, as might be found in a tumor of müllerian origin. Immunohistochemically, tumors express mesothelial markers (WT-1, calretinin, CK5/6, D2-40) and are usually negative for epithelial-specific markers (Ber-EP4, B72.3, MOC-31, ER, PR) (Sangoi et al. 2009; Wachter et al. 2011). Electron microscopic studies

Fig. 32 Adenomatoid tumor. (a) Nodular configuration in the wall of the fallopian tube. (b) Diffuse proliferation of tubules which infiltrate the myosalpinx. Foci of chronic inflammation are also present. (c) The tubules are lined by

a single layer of flat eosinophilic cells with cytoplasmic vacuoles and bland nuclei. The lumens of the tubules are empty. An intra-lumenal thin cytoplasmic strand, which can be focally seen in some cases, is present (arrow)

676

also support a mesothelial origin for these lesions. Clinically, they are asymptomatic, and recurrences after adequate excision are rare. The most important lesion to be considered in the differential diagnosis, particularly at the time of frozen section, is metastatic signet ring cell carcinoma. Clinical data such as prior history of a carcinoma (some primary tumors may be occult, and such a history will not be known), extra-fallopian tube tumor seen intraoperatively (especially if multifocal), bilaterality, and histologic presence of a combination of glands, papillae, and solid sheets of tumor help favor a diagnosis of carcinoma. Nuclear atypia and mitotic activity should raise suspicion for carcinoma, but some signet ring cell carcinomas may lack these features. A source of confusion may be the presence of the alcian bluepositive, hyaluronidase-digestible material within the lumens of the tubules in adenomatoid tumor. At the time of intraoperative consultation, this material resembles epithelial-type mucin on H&E frozen section slides; however, it is lost during routine processing and, as a result, is not present in permanent H&E sections. After frozen section analysis, immunohistochemical stains specific for mesothelial and epithelial markers will help aid distinction.

Epithelial Tumors Papilloma is rare (Gisser 1986). It is composed of an intraluminal mass with an “adenomatous” and very complex papillary proliferation. At low-power, the proliferation resembles an exaggerated pattern of tubal mucosa with fine stromal fibrovascular cores, and the quantity of papillae is much greater than in the normal fallopian tube (Fig. 33). On higher power, the epithelium resembles normal fallopian tube mucosa, including the presence of ciliated and secretory cells. The nuclei are bland, and mitotic activity is not seen. In our anecdotal experience, fallopian tube papilloma diffusely expresses ER and WT-1, and the Ki-67 proliferation index is low. Because of the complex papillary proliferation, this tumor may be confused with atypical proliferative (borderline) serous tumor, low-grade serous carcinoma, or a villoglandular variant of endometrioid carcinoma. Although very complex, the orderly degree of papillary branching of papilloma is in contrast to

R. Vang

Fig. 33 Fallopian tube papilloma. Abundant papillae with complex branching resemble the plicae of normal fallopian tube. Closer magnification will show the same type of epithelium seen in normal fallopian tube

a greater degree of complexity and hierarchical papillary branching with cellular stratification and tufting of atypical proliferative (borderline) serous tumor. The fine micropapillary tufting, associated psammoma bodies, and stromal invasion of low-grade serous carcinoma are not seen in papilloma. At low-power, papilloma can resemble a villoglandular endometrioid carcinoma, but closer examination at high-power shows endosalpingeal cell types and an absence of endometrioid differentiation. Furthermore, squamous metaplasia, endometriosis, and foci with solid growth would favor endometrioid carcinoma. Metaplastic papillary tumor is rare and found as an incidental finding in the lumen of the fallopian tube during the postpartum period (Bartnik et al. 1989; Keeney and Thrasher 1988; Saffos et al. 1980). It is of microscopic size and composed of broad papillae lined by stratified and tufted epithelium with cells showing abundant eosinophilic cytoplasm (Fig. 34). The nuclei do not exhibit malignant features. It is not clear whether this lesion is a papillary metaplastic proliferation or small atypical proliferative (borderline) serous tumor associated with pregnancy. Regardless, the behavior appears benign. Cystadenomas have been reported but are rare. Although the topic of whether or not atypical proliferative (borderline) tumors are benign has been debated in the literature, these tumors are

11

Diseases of the Fallopian Tube and Paratubal Region

Fig. 34 Metaplastic papillary tumor. The tumor is small and characteristically located within the lumen of the fallopian tube. It contains a limited number of mediumsized papillae. Closer magnification will show papillae lined by columnar to cuboidal cells that have abundant, dense, and eosinophilic cytoplasm with nuclei that are bland or have, at most, mild atypia. Only a limited degree of stratification is present

included in the benign neoplasm section of this chapter for simplicity. Rare atypical proliferative (borderline) serous, endometrioid, and clear cell tumors of the fallopian tube have been reported. The literature is too limited to predict their outcome; however, behavior similar to their ovarian counterparts would be anticipated. Atypical proliferative (borderline) mucinous tumors have been described, but such cases should be rigorously evaluated to exclude the likely possibility that they represent secondary involvement of the fallopian tubes from a non-ovarian site.

Leiomyoma and Adenomyoma Leiomyoma is the most common mesenchymal tumor of the fallopian tube; however, these are much less common than uterine leiomyomas. They are usually small and grossly and microscopically similar to those found in the uterus and other gynecologic sites, and they can undergo similar degenerative changes. Rarely, benign glands and smooth muscle may be so intimately involved in a tumor that a diagnosis of adenomyoma may be warranted; however, these may arguably represent endometriosis with smooth muscle metaplasia (“endomyometriosis”) or the so-called uterine-like mass.

677

Fig. 35 Adenofibroma. Note the biphasic architecture with cellular fibromatous stroma and round glands lined by a single layer of bland cuboidal epithelium. Other cases may be predominantly composed of round and blunted papillae with cleft-like architecture or have an abundant component of tubules

Other Benign Mesenchymal and Mixed Epithelial–Mesenchymal Tumors Although adenofibromas producing clinical masses are uncommon, those of microscopic size are not infrequent. In one consecutive series of fallopian tubes unassociated with tubo-ovarian malignancy or inflammatory disorders, and in which all tubal tissue was submitted for histologic examination, adenofibromas were found in 30% of cases (Bossuyt et al. 2008). In that study, the majority were < 0.3 cm in size, and only a small subset were > 1 cm. All arose in the fimbria. Some may be synchronously associated with an ovarian adenofibroma. Histologically, fallopian tube adenofibromas resemble their ovarian counterparts, with admixed epithelial and mesenchymal components (Fig. 35). The epithelial component may contain papillary clefting and is frequently cystic or composed of small round tubules. Most tumors are of serous histologic type, but a minority are of endometrioid type (Alvarado-Cabrero et al. 1997). The mesenchymal component is hypercellular, densely fibrotic, or hyalinized. In lesions 1 cm in greatest dimension IIIA2: Microscopic extrapelvic (above the pelvic brim) peritoneal involvement with or without positive retroperitoneal lymph nodes IIIB: Macroscopic peritoneal metastasis beyond the pelvis up to 2 cm in greatest dimension, with or without metastasis to the retroperitoneal lymph nodes IIIC: Macroscopic peritoneal metastasis beyond the pelvis >2 cm in greatest dimension, with or without metastasis to the retroperitoneal lymph nodes (includes extension of tumor to capsule of liver and spleen without parenchymal involvement of either organ) Distant metastasis, including pleural effusion with positive cytology; liver or splenic parenchymal metastasis; metastases to extra-abdominal organs (including inguinal lymph nodes and lymph nodes outside the abdominal cavity); and transmural involvement of intestine IVA: Pleural effusion with positive cytology IVB: Liver or splenic parenchymal metastasis; metastases to extra-abdominal organs (including inguinal lymph nodes and lymph nodes outside the abdominal cavity); and transmural involvement of intestine

It has been suggested that intraluminal masses without invasion qualify as neither stage 0 (old staging system) nor stage IA (Alvarado-Cabrero et al. 1999). Also, because of observed differences in prognosis for stage I fallopian tube carcinomas with different depths of invasion into the wall (similar to other abdominal/pelvic organs with a muscular wall), it has been recommended that the FIGO stage should be modified since such cases are not appropriately represented by the current version of the FIGO staging system. AlvaradoCabrero et al. have proposed that stage I cases should be divided into substages IA-0 (intraluminal masses without invasion into lamina propria), IA-1 (invasion into lamina propria but not muscularis), and IA-2 (invasion of muscularis) (Alvarado-Cabrero et al. 1999). As it

has also been suggested that carcinomas in the fimbriated end without invasion have a worse prognosis than carcinomas invading the wall of the tube because of direct access to the peritoneal cavity, it has been proposed that the FIGO system should be modified since the former are not represented by the current version (AlvaradoCabrero et al. 1997, 1999). This proposed stage would be designated I(F). The majority of patients with symptomatic disease have advanced stage disease at presentation (stage >I). In the largest clinicopathologic study using hospital-based cases by Baekelandt et al., the distribution of stage was: stage 0 (6%) [old staging system], stage I (27%), stage II (22%), stage III (35%), and stage IV (12%) (Baekelandt et al. 2000). These results are similar to those of

11

Diseases of the Fallopian Tube and Paratubal Region

other large hospital-based or population-based studies which have found the percentage of stage I cases to be 30–56% (Heintz et al. 2006; Hellstrom et al. 1994; Rosen et al. 1999; Stewart et al. 2007). The true stage distribution of all fallopian tube carcinomas in the general population is difficult to determine because of the existence of two patient populations which are usually not included together in the same studies – patients with asymptomatic tumors (i.e., occult tubal carcinomas in prophylactic bilateral salpingooophorectomy specimens from women with an increased genetic risk for carcinoma) and those with symptomatic tumors (i.e., women with bulky tumors and advanced stage disease who may not be suspected of having an increased genetic risk for carcinoma). Intraoperative and Gross Features Bilaterality is infrequent (3–13% of cases) (Alvarado-Cabrero et al. 1999, 2013, Baekelandt et al. 2000; Hellstrom et al. 1994; Stewart et al. 2007). The average tumor size is 4–5 cm (range, 0.2–10 cm) (Alvarado-Cabrero et al. 1999, 2013). The fallopian tube is dilated in slightly over one-half of cases, which intraoperatively can be mistaken for a hydrosalpinx, hematosalpinx, or pyosalpinx (Alvarado-Cabrero et al. 1999). The tumor can appear as one or more yellow to tan nodules or a mass that fills the lumen (Fig. 37). Hemorrhage or necrosis is frequent. Most tumors are within the tubal portion (usually the distal two-thirds), but a small percentage is located in the fimbriated end. Histologic Features Intraepithelial Carcinoma

Noninvasive carcinomas of the fallopian tube have been considered “carcinoma in situ” in the past. With recognition of early carcinomas without invasion of underlying fallopian tube stroma in prophylactic bilateral salpingo-oophorectomy specimens, as well as their detection as incidental findings in routine specimens, the term STIC has emerged over the last several years. Given the

681

Fig. 37 Gross features of fallopian tube carcinoma. The cut surface is slightly heterogeneous, nodular, irregular, and yellow-tan. For comparison, the structure at the upper left is the ovary uninvolved by tumor

ability of STIC to spread beyond the fallopian tube without invasion of underlying stroma (see below), the term carcinoma in situ should be abandoned because it implies that there is no potential for metastasis. Essentially all intraepithelial carcinomas of the fallopian tube are of high-grade serous type, and the lesional cells of STIC show secretory cell differentiation (Lee et al. 2007). However, rare endometrioid intraepithelial carcinomas have been reported (Jarboe et al. 2008). Histologically, STIC is the earliest morphologically recognizable form of tubal carcinoma. It is characterized by absence of invasion of underlying fallopian tube stroma and the presence of cytologic abnormalities that result in the fallopian tube epithelium appearing darker than adjacent normal epithelium at low-power magnification (Fig. 38). In cases with invasive carcinoma in the same tube, STIC may be found directly adjacent to invasion. STIC can occur as a single focus or multifocally. It preferentially occurs in the distal or fimbriated end of the fallopian tube, and some investigators have also suggested that STIC frequently occurs at or near the tuboperitoneal junction (see section on “Normal Histology”) (Seidman 2015). The lesional epithelium is typically flat, but some degree of stratification may be seen. The luminal border may be straight or exhibit variable amounts of irregular contours and hobnail morphology. With increased

682

R. Vang

Fig. 38 STIC. (a) STIC is composed mostly of a flat proliferation of cells. However, at low-power magnification, STIC is visible because of the thicker and darker

epithelium (arrowheads) compared with adjacent normal tubal epithelium (arrow). (b) Abrupt transition between STIC and normal tubal epithelium

Fig. 39 STIC. (a) The lesional cells can be hyperchromatic with high nuclear-to-cytoplasmic ratios, stratification, and loss of polarity (upper half of photograph), or they may show enlarged and round nuclei with vesicular

chromatin, nucleoli, and mitotic figures (lower half of photograph). (b) Comparison between normal tubal epithelium (upper half of photograph) and STIC (lower half of photograph)

stratification, small tufts of detached cells can be found within the tubal lumen. At high-power magnification, the lesional cells lack cilia and show variable combinations of nuclear enlargement, increased nuclear-to-cytoplasmic ratios, hyperchromasia or irregular chromatin distribution, loss of polarity, prominent nuclei, and mitotic figures (Fig. 39). Nuclei may be oval or columnar but are frequently round. Before diagnosing an incidental lesion as STIC in a routine specimen, it is necessary to submit all remaining fallopian tube tissue for histologic

examination, as invasive carcinoma can be of small size. Invasive Carcinoma

Some otherwise conventional-appearing highgrade serous carcinomas within the lumen of the fallopian tube may be small without enlargement of the fallopian tube and lack invasion of the underlying tubal stroma. Such morphology in occasional cases can overlap with the upper morphologic limit of STIC, especially when the latter shows an increased degree of stratification. In

11

Diseases of the Fallopian Tube and Paratubal Region

683

Fig. 40 Invasive high-grade serous carcinoma. (a) Complex papillae with stratified epithelium producing irregular slit-like spaces and small epithelial tufts. (b) The nuclei are high-grade with abundant mitotic figures. Note

enlarged nuclei with vesicular chromatin and nucleoli. In other cases, the nuclei may be hyperchromatic rather than vesicular

such cases, the diagnostic threshold between STIC vs. a small intraluminal high-grade serous carcinoma without invasion of underlying fallopian tube stroma is subjective and varies between authors (Jarboe et al. 2008; Young 2007). However, some authors have suggested that intraluminal masses without invasion qualify as neither stage 0 (old staging system) nor stage IA (see section “FIGO Stage” above) (AlvaradoCabrero et al. 1999). Some invasive carcinomas may be of small volume. As these may not be clinically evident, and in order to assess the possibility of the fallopian tube being a primary site of occult disease, it is important to submit all remaining fallopian tube tissue from grossly unremarkable tubes for histologic examination in women presenting with metastatic high-grade adenocarcinoma of unknown primary site. The histologic types and appearances of invasive fallopian tube carcinoma are similar to its ovarian counterparts. In the largest clinicopathologic study using hospital-based cases, the distribution of histologic types was: serous (80%), adenocarcinoma, NOS (10%), endometrioid (7%), clear cell (2%), mucinous (2%), and mixed serous-mucinous (1%) (Baekelandt et al. 2000). Most fallopian tube carcinomas are poorly differentiated, and well-differentiated tumors are very uncommon. Fallopian tube carcinomas are graded

in a fashion analogous to those in the ovary (i.e., low- vs. high-grade for serous, FIGO uterine criteria for endometrioid, etc.). The majority, if not all, tubal serous carcinomas are histologically indistinguishable from high-grade serous carcinomas of the ovary and include broad papillae with epithelial stratification, irregular slit-like spaces with micropapillary tufting, invasion by solid nests of variable size or sheets of tumor cells, necrosis, and psammoma bodies (Fig. 40) (Alvarado-Cabrero et al. 1999). The nuclei are high-grade, characterized by nuclear enlargement, hyperchromasia or irregular chromatin distribution with prominent nucleoli, irregular nuclear membranes, and abundant mitotic figures. STIC may be found adjacent to invasive carcinoma. Most endometrioid carcinomas are grade 2 or 3, but some are grade 1 (Navani et al. 1996; Rabczynski and Ziolkowski 1999). They may resemble conventional endometrioid carcinomas as seen in the endometrium, including squamous differentiation and villoglandular architecture, but oxyphilic types, sex cord-like appearances, and spindled epithelial cells may be seen. In some cases, associated endometriosis is present. Benign stromal osseous metaplasia can be seen in a minority of cases. Almost one-half of cases have an appearance resembling female adnexal tumor of wolffian origin (FATWO)-like type

684

(Daya et al. 1992; Navani et al. 1996). Independent primary endometrioid carcinomas can synchronously arise in the fallopian tube and uterus (Culton et al. 2006). Clear cell and transitional cell carcinomas resemble those seen in the ovary (AlvaradoCabrero et al. 1999; Koshiyama et al. 1994). Other rare histologic types include undifferentiated, small cell neuroendocrine, lymphoepitheliomalike, mixed serous-transitional cell, squamous cell, adenosquamous, hepatoid, glassy cell, and giant cell carcinomas (Alvarado-Cabrero et al. 1999; Aoyama et al. 1996; Cheung et al. 1994a; Herbold et al. 1988). Mucinous carcinomas with a synchronous endocervical adenocarcinoma have been described, but such cases should be rigorously evaluated to exclude the likely possibility that they represent secondary involvement of the fallopian tubes from the endocervix. Unlike the

R. Vang

ovary, low-grade serous carcinomas typically are not seen in the fallopian tube. Immunohistochemical Features Although some cases of STIC may be diagnosed solely on histologic features without the need for immunohistochemistry, immunostains are helpful in establishing the diagnosis in problematic cases. STIC diffusely and strongly expresses p53 (Fig. 41) (pattern associated with TP53 missense mutations) in the majority of cases while a minority of cases have complete loss of expression (pattern associated with TP53 null mutations) (Kuhn et al. 2012c). The Ki-67 proliferation index is usually elevated (Fig. 41). In one study, the mean Ki-67 proliferation index was 72% (range: 40–95%) (Jarboe et al. 2008) while in another the mean index and range were 36% and 12–71%, respectively (Kuhn et al. 2012a). Also,

Fig. 41 Immunohistochemical features of STIC. (a) H&E. (b) Diffuse expression of p53. (c) Elevated Ki-67 proliferation index. (d) Diffuse expression of p16

11

Diseases of the Fallopian Tube and Paratubal Region

p16, which is diffusely and strongly expressed in endometrial serous carcinomas and many ovarian high-grade serous carcinomas, can be expressed with a similar pattern in STIC; however, as this occurs in only a subset of cases, it is not a reliable diagnostic marker for this lesion. Laminin has been reported as being overexpressed in STIC with diffuse and intense cytoplasmic staining (see section on “Pathogenesis, Including Molecular Features” of “Serous Carcinoma”) (Kuhn et al. 2012b). Invasive high-grade serous carcinomas usually show diffuse expression of WT-1. It should be noted, however, that expression is not restricted to invasive carcinomas, as normal fallopian tube epithelium and STIC also show WT-1 expression. Expression of ER and PR is variable. p53 expression, similar to STIC, will have an abnormal pattern, either diffuse or null. Immunohistochemical data are limited for endometrioid carcinomas, but in our experience, WT-1 and p53 can be diffusely expressed in some cases. Diagnostic Criteria and Classification for STIC and Intraepithelial Atypias: Algorithmic Approach Based on Morphology and Immunohistochemistry Diagnosis of STIC based solely on histologic features has been shown to lack high interobserver agreement and, therefore, not be highly reproducible (Carlson et al. 2010; Vang et al. 2012; Visvanathan et al. 2011). Accordingly, a diagnostic algorithm utilizing a combination of morphology and immunohistochemistry for p53 and Ki-67 (Fig. 42) has been proposed for classifying epithelial atypia of the fallopian tube, which results in a more reproducible diagnosis of STIC (Vang et al. 2012, 2013, Visvanathan et al. 2011). As there has been no standardized criteria for this spectrum of lesions, this algorithm allows for classification of STIC, atypical lesions intermediate between STIC and p53 signature (serous tubal intraepithelial lesion [STIL]), and p53 signature. Also, this type of algorithmic approach facilitates consistency of terminology between practicing pathologists, as well as clinicians, and more uniform classification of lesions among investigators at different academic centers.

685

According to the algorithm, first the morphology of atypical foci is classified as unequivocal for STIC, suspicious for STIC, or not suspicious for STIC utilizing a variable combination of histologic features described above. Then, p53 expression is scored as abnormal (diffuse or null pattern, with moderate to strong intensity) or normal. Of note, a “wild-type”/normal pattern will have occasional cells with usually mild intensity. In contrast, a “null” pattern demonstrates complete loss of expression provided an adequate positive internal control with a “wild-type”/normal pattern is present. The Ki-67 proliferation index is scored as either low (I. Most of these invasive tumors are high-grade serous carcinomas, but a minority are of endometrioid type. Also, see section above “Treatment and Prognosis: Invasive Carcinoma” and section below “Behavior of STIC (BRCA Germline MutationAssociated and Sporadic Cases).” Occult Disease in Routine Specimens in Women Without Known Genetic Risk (Presumably Sporadic Cases) Occult fallopian tube carcinomas are not restricted to the setting of risk-reducing salpingo-oophorectomy specimens (women at increased risk for hereditary ovarian carcinoma). On occasion, occult sporadic carcinomas can be detected in specimens performed for other surgical indications, such as uterine leiomyomas (Gilks et al. 2015; Morrison et al. 2015). When occult sporadic carcinoma is found in this scenario, STIC is almost always present, either in isolation or with associated invasive high-grade serous carcinoma. When the latter component is identified, it is usually found in the fallopian tube but sometimes the ovary. The invasive carcinomas are typically small (often 1 cm). While cases with invasive carcinoma can be FIGO stage I, staging has revealed advanced disease in some patients. With regard to the frequency of incidentally finding STIC in routine specimens, a number of large studies have shown that sporadic STIC would be expected to be present in 10 cm3) (Balen et al. 2003). The typical gross or ultrasonographic ovarian morphology, however, is unnecessary to make the diagnosis of PCOS and, in the absence of the typical clinical findings, is not diagnostic of the syndrome (Differential Diagnosis) (Taylor 1998).

Microscopic Findings The superficial cortex is fibrotic and hypocellular, resembling a capsule (Fig. 30a), and may contain prominent thick-walled blood vessels (Hughesdon 1982). Tongues of similarly fibrotic stroma may extend from the superficial cortex into the deeper cortex and medulla. The cysts are atretic cystic follicles and have an inner lining of several layers of nonluteinized, focally exfoliated granulosa cells. There is a prominent outer layer of luteinized theca interna cells (Fig. 30b), giving rise to the term follicular HT, but other studies have found that cystic follicles in women with PCOS differ from those in normal women only in their increased number (Lunde et al. 1988). Maturing follicles up to midantral stage and atretic follicles exhibiting prominent luteinization of the theca interna may be twice as

J. A. Irving and P. B. Clement

numerous as in normal ovaries. Primordial follicles are normal in number and appearance (Hughesdon 1982). As noted, stigmata of prior ovulation are typically absent, but corpora lutea have been described in as many as 30% of otherwise typical cases of PCOS. The deeper cortical and medullary stroma may have as much as a fivefold increase in volume. The stroma contains luteinized stromal cells in 80% of cases and, less commonly, foci of smooth muscle (Hughesdon 1982). Nests of ovarian hilus (Leydig) cells may be more numerous in patients with PCOS than in age-matched controls.

Pathophysiology The pathophysiology of PCOS is complex, and the initiating factor(s) is (are) not yet completely understood (Taylor 1998; Dumesic et al. 2015). A cardinal finding is an elevation of the serum level of LH or an elevated LH:FSH ratio; of note, documentation of abnormal gonadotropin levels is not required for the diagnosis of PCOS in routine clinical practice, since their release is pulsatile and concentrations vary throughout the menstrual cycle (Ehrmann 2005). LH stimulates the follicular theca interna cells to produce androstenedione, which is converted peripherally, primarily within adipose tissue, to estrone (E1), and to a lesser extent, testosterone. More importantly, ovarian follicular production of testosterone is also increased, leading to the small increases in serum testosterone concentrations that are present in PCOS. Estradiol (E2) levels remain normal or

Fig. 30 Polycystic ovary disease. (a) Multiple cystic follicles lie beneath the superficially fibrotic cortex. (b) A cystic follicle is lined by nonluteinized granulosa cells and an outer, thicker layer of luteinized theca interna cells

12

Nonneoplastic Lesions of the Ovary

low normal, resulting in an elevated E1/E2 ratio. Elevated E1 levels, and in some patients an increased secretion of inhibin, a nonsteroidal peptide produced by granulosa cells (Tanabe et al. 1983), inhibit secretion of FSH. An elevated LH: FSH ratio is thus a characteristic finding in PCOS. Ovarian estrogen production in PCOS is markedly diminished, probably a result of inactivity of the FSH-dependent aromatase system within the granulosa cells. Inadequate intrafollicular estrogen synthesis, increased intrafollicular androgens, and an elevated LH:FSH ratio result in cessation of follicle growth at the midantral stage, anovulation, and sclerocystic ovaries. A number of interlinked factors potentially play a role in initiating or perpetuating PCOS. According to Taylor (1998), “there are correlations between gonadotropin secretion, insulin secretion, and androgen secretion across the spectrum of patients with PCOS such that it remains impossible to determine the primary etiologic factor in the vast majority of patients.” Obesity-related factors including hyperestronemia due to conversion of androstenedione to E1 and hyperinsulinemia play a role in the pathogenesis of PCOS, and the increase in obesity in the United States has been postulated as an explanation for the reported increased incidence of PCOS (Ehrmann 2005). An estimated 20–30% of patients with PCOS have adrenal androgen excess, as manifested by an elevated androgen dehydroepiandrosterone sulfate (DHES) and abnormal adrenal androgen responses to adrenocorticotropic hormone (ACTH) (Rodin et al. 1994; Yildiz and Azziz 2007). The increased adrenal androgens can lead to hyperestronemia and consequently an elevated LH:FSH ratio. Although some investigators believe that the adrenal abnormalities (which may include late-onset congenital adrenal hyperplasia) are a primary disturbance, others have concluded that they are secondary to the hormonal milieu of PCOS (Carmina et al. 1995). Insulin resistance and hyperinsulinemia are present in most obese and nonobese women with PCOS, although these features tend to be more severe in the former group (Nestler et al. 1989). Despite peripheral insulin resistance, ovarian tissues remain responsive to insulin in women with

739

PCOS. Insulin appears to amplify LH action, enhancing production of estradiol and progesterone from the follicular granulosa cells and possibly contributing to the arrest of follicle growth (Granks et al. 1999). As discussed later (HAIRAN syndrome), insulin and insulin-like growth factor stimulate proliferation of ovarian stromal cells and their production of androgen. Hyperinsulinemia thus can increase circulating androgen levels (and by peripheral conversion, estrone) in patients with PCOS. The resultant hyperandrogenism may in turn increase insulin resistance. Hyperprolactinemia is present in approximately 15–25% of patients with PCOS, but recent evidence indicates that they are distinct entities, with galactorrhea a feature of hyperprolactinemia but not PCOS; in one study, elevated plasma prolactin levels were attributable to pituitary adenomas and drug-related effect in 19% and 23% of cases, respectively (Ehrmann 2005; Filho et al. 2007).

Differential Diagnosis PCOS is a clinicopathologic syndrome, and the finding of polycystic ovaries with little or no evidence of prior ovulation does not warrant the diagnosis per se in the absence of the usual clinical findings. Polycystic ovaries that resemble those of PCOS are seen occasionally in prepubertal children and in otherwise normal girls during the first few years after the onset of puberty. Similarly, ultrasonographic studies have revealed that ovulating women with minor evidence of hyperandrogenism, but without menstrual irregularity, can have polycystic ovaries similar to those of patients with overt clinical manifestations except that the ovaries also contain corpora lutea and albicantia (Polson et al. 1988). Thus, the boundary between the clinical syndrome of PCOS and normality is not clear cut. Although PCOS does not cause a problem in differential diagnosis with a neoplasm for the pathologist, the clinical manifestations may suggest the possibility of an androgenic or estrogenic ovarian tumor, particularly in the exceptional cases in which the disease coexists with a nonfunctioning ovarian neoplasm. In some cases, in

740

which the associated ovarian tumor is capable of function, such as a Sertoli–Leydig cell tumor, it may be difficult to determine which lesion was responsible for the endocrine manifestations. The differential diagnosis includes a wide variety of other disorders that result in abnormal gonadotropin release, chronic anovulation, and sclerocystic ovaries. Sclerocystic ovaries are a nonspecific morphologic expression of chronic anovulation in the premenopausal patient and can accompany (1) adrenal lesions such as Cushing’s syndrome, congenital adrenal hyperplasia (most commonly 21-hydroxylase or 11-beta-hydroxylase deficiency), and virilizing adrenal tumors; (2) primary hypothalamic–pituitary disorders; and (3) ovarian lesions that produce excessive quantities of estrogens or androgens, including sex cordstromal tumors, steroid cell tumors, and nonneoplastic lesions such as Leydig cell hyperplasia and stromal hyperthecosis (HT). As previously noted, the latter overlaps both clinically and pathologically with PCOS, and the two disorders may represent opposite poles of a single disease spectrum. Sclerocystic ovaries have also been described in patients with autoimmune oophoritis (Bannatyne et al. 1990), after long-term use of oral contraceptives, in association with periovarian adhesions, and after long-term androgen therapy in female to male transsexuals (Pache et al. 1991). An association between a PCOS-like syndrome and the use of the antiepileptic drug valproate has been found (Isojärvi et al. 1993).

Stromal Hyperplasia and Stromal Hyperthecosis Normal, spindle-shaped stromal cells, which have scant cytoplasm and resemble fibroblasts, are typically arranged in whorls or a storiform pattern within a dense reticulum network and a variable amount of collagen, which is most abundant in the superficial cortex. Stromal cells are typically immunoreactive for CD56, WT-1, vimentin, and estrogen and progesterone receptors (Al-Timimi et al. 1985; He et al. 2008; McCluggage et al. 2007). In many women of late-reproductive and postmenopausal age, there

J. A. Irving and P. B. Clement

is a decrease in stromal volume and cellularity, with an increase in collagen. A variety of other cells may be found within the ovarian stroma, most of which are probably derived from the spindle-shaped stromal cells, including luteinized stromal cells, enzymatically active stromal cells, decidual cells, endometrial stromal-type cells (stromal endometriosis, “ovarian stromatosis”) (▶ Chap. 13, “Diseases of the Peritoneum”), smooth muscle cells, fat cells, stromal Leydig cells, and rare cells of neuroendocrine or APUD (amino precursor uptake and decarboxylation) type (Clement 2007). Stromal hyperplasia (SH) is characterized by varying degrees of proliferation of the ovarian stromal cells. Stromal hyperthecosis (HT) refers to the presence of luteinized cells within the stroma at a distance from the follicles; it is usually accompanied by at least a moderate degree of SH.

Clinical Features The clinical manifestations are variable. Moderate to severe SH is most commonly encountered in women in their sixth and seventh decades and has been documented in more than one third of autopsied patients in this age group (Snowden et al. 1989). In a recent study, SH was identified in 24% of grossly normal ovaries (Seidman and Krishnan 2016) Similar degrees of SH are found less commonly in patients in the eighth decade, suggesting that it may be a reversible process. A strong negative association with parity was found in one study (Snowden et al. 1989). SH of moderate to severe degree may be found in women with disorders associated with androgenic or estrogenic manifestations including endometrial carcinoma, obesity, hypertension, and glucose intolerance, but these findings are less frequent and less obtrusive than in HT. HT is most frequent in patients in the sixth to ninth decades. Some familial cases of HT have been reported (Judd et al. 1973). The process has been documented in one third of autopsied patients over the age of 55, and exhaustive microscopic sampling may reveal that rare luteinized stromal cells are even more common in this age group (Clement 2007). In postmenopausal women, HT is usually mild and of doubtful

12

Nonneoplastic Lesions of the Ovary

clinical significance, but rare cases associated with postmenopausal hirsutism have been reported (Damodaran et al. 2011). Clinically florid examples of HT are more common in patients in the younger reproductive age group, although rare cases occur in adolescents and postmenopausal patients (Madeido et al. 1985). The findings include marked virilization, obesity, hypertension, hyperinsulinemia, and decreased glucose tolerance. In addition, a small subset of women with HT (or occasionally PCOS) have the HAIR-AN syndrome (see following). The clinical picture of HT typically evolves gradually, but occasionally there is an abrupt onset, potentially suggesting the presence of an androgenic tumor, especially if the process is unilateral and associated with ovarian enlargement. Less commonly, the clinical findings are more characteristic of PCOS. In some patients with HT, especially postmenopausal women, estrogenic findings predominate and may include endometrial hyperplasia or even well-differentiated adenocarcinoma (Madeido et al. 1985; Nagamani et al. 1988; Sasano et al. 1989; Snowden et al. 1989). Conversely, women with endometrial hyperplasia or carcinoma in some studies have had a high frequency of HT on microscopic examination of their ovaries (Nagamani et al. 1988; Sasano et al. 1989). These findings are corroborated in a recent retrospective study of 238 women over the age of 60 who had undergone hysterectomy and bilateral oophorectomy, with HT detected in 23.9% of those with atrophic endometrium, and in approximately half of women with endometrial polyp, hyperplasia, or adenocarcinoma (Zhang et al. 2017). HT-related virilization has been present in two cases of placental site trophoblastic tumor, and in one case it was the presenting manifestation of the tumor (Nagamani et al. 1990; Nagelberg and Rosen 1985).

Gross Findings SH and HT are almost invariably bilateral, and the ovaries range from normal size to 8 cm in maximum dimension, thus potentially mimicking an ovarian neoplasm (Madeido et al. 1985). The sectioned surfaces are usually solid, firm, homogenous, and white to yellow (Fig. 31). In cases of

741

Fig. 31 Stromal hyperplasia and stromal hyperthecosis. Solid, homogeneous, yellow sectioned surface

nodular HT, multiple yellow nodules may be appreciable. In premenopausal women, sclerocystic changes similar to those seen in PCOS may also be present (Hughesdon 1982). HT rarely coexists with a neoplasm, usually a stromal luteoma, which may also have hormonesecreting potential (Scully 1964).

Microscopic Findings In both SH and HT, a variable degree of nodular or diffuse cortical and medullary proliferation of ovarian stromal cells is present (Fig. 32a). A mild degree of SH cannot be reliably distinguished from the normal appearance. Follicular derivatives may lie within the hyperplastic stroma but may be rare or absent in advanced cases (Fig. 32b). The stromal cells in SH are plumper than normal postmenopausal ovarian stromal cells and have oval to fusiform, vesicular nuclei, and, frequently, cytoplasmic lipid. The luteinized stromal cells of HT are more common in the medulla but may also be present in the cortex. They appear as single cells, small nests, or nodules of polygonal cells with abundant eosinophilic to vacuolated cytoplasm, containing variable amounts of lipid (Fig. 33). The round nucleus of the luteinized cells typically has a central small nucleolus. As noted, in premenopausal women including those with the HAIR-AN syndrome, sclerocystic changes characteristic of PCOS are also commonly present. In cases of HT accompanying the HAIR-AN

742

J. A. Irving and P. B. Clement

Fig. 32 Stromal hyperplasia. (a) A diffuse proliferation of ovarian stromal cells within the cortex and medulla is seen. (b) Note absence of follicular derivatives

Fig. 33 Stromal hyperthecosis. (a) A nest of luteinized stromal cells is present within the ovarian stroma. (b) Calretinin immunostain highlights luteinized stromal cells

syndrome (see following), edema and fibrosis of the ovarian stroma, rather than SH, are frequently a prominent change (Massachusetts General Hospital Case Records 1988). Other ovarian findings occasionally encountered in HT include an increased number of atretic follicles, small stromal nodules of metaplastic smooth muscle, hilus cell hyperplasia, hilus cell tumors, stromal luteomas, and thecomas (Roth and Sternberg 1973; Scully 1981; Sternberg and Roth 1973; Zhang et al. 1982). SH in the absence of HT has also been associated with thecomas (Zhang et al. 1982). Some cases of HT may be associated with massive ovarian edema.

Histochemical analyses have shown that nonluteinized and luteinized stromal cells and cells transitional in appearance between the two have oxidative activity important in steroid hormone production (Scully and Cohen 1964). The luteinized cells were immunoreactive for cytochrome P-450-17-alpha, which catalyzes androgen synthesis, in approximately 50% of the cases of SH in one series (Sasano et al. 1989). Luteinized stromal cells are also immunoreactive for inhibin, calretinin (Fig. 33), testosterone, estradiol, and FSH (Madeido et al. 1985; McCluggage and Maxwell 2001; Nagamani et al. 1988; Pelkey et al. 1998).

12

Nonneoplastic Lesions of the Ovary

Pathophysiology In vitro and in vivo studies have shown that ovaries with SH secrete more androstenedione, estrone, and estradiol than normal ovaries (Dennefors et al. 1980). Similar studies using ovarian tissue from patients with HT (Nagamani et al. 1992) as well as in vivo studies in these patients have shown, respectively, markedly increased production rates and serum levels of ovarian testosterone, dihydrotestosterone, and androstenedione, usually in the male range (Madeido et al. 1985). As already noted, immunohistochemical staining for various enzymes involved in the conversion of cholesterol to steroid hormones in cases of HT has been consistent with androgen synthesis not only in the luteinized stromal cells characteristic of the disorder but also in the adjacent spindle-shaped stromal cells (Sasano et al. 1989). As in PCOS, the predominant estrogen in patients with SH and HT is estrone, derived predominantly from peripheral aromatization of ovarian androgens, resulting in an increased estrone/estradiol ratio (Sasano et al. 1989). Unlike patients with PCOS, most premenopausal patients with HT have normal gonadotropin levels (Judd et al. 1973). That gonadotropins may play a role in SH and HT, however, is suggested by (1) elevated LH levels in occasional premenopausal women with HT and most postmenopausal patients with SH and HT; (2) immunoreactivity for FSH and LH receptors within ovarian stromal cells (Nakano et al. 1989); (3) in vitro incubation studies showing that FSH and LH stimulate proliferation of the ovarian stroma of pre- and postmenopausal women (Snowden et al. 1989); (4) studies showing that androgen production by the ovarian stromal cells in patients with and without HT is enhanced by LH (Dennefors et al. 1980; Nagamani et al. 1992); (5) the often prominent stromal luteinization during pregnancy; (6) cases of symptomatic HT complicating pregnancy and trophoblastic disease (Nagamani et al. 1990; Nagelberg and Rosen 1985); and (7) the increase in pituitary amphophils in some cases of severe HT (Madeido et al. 1985). As noted, insulin resistance and hyperinsulinemia occur in as many as 90% of

743

patients with HT and likely play a role in the pathogenesis of the stromal luteinization in these patients (HAIR-AN syndrome) (Nagamani et al. 1986).

Differential Diagnosis In contrast to a fibroma, SH is almost always bilateral and is characterized by cells with smaller nuclei, scanty collagen, and nodules that commonly coalesce. The lesion is distinguished from a low-grade endometrioid stromal sarcoma by the spindle shape of its cells and by an absence of mitotic figures and regularly distributed arterioles. The differential diagnosis of HT includes other nonneoplastic and neoplastic solid proliferations of luteinized cells, most of which are also virilizing. The nonneoplastic category includes pregnancy luteoma and Leydig cell hyperplasia (discussed elsewhere in this chapter) and the neoplasms include luteinized thecoma and steroid cell tumors (▶ Chap. 15, “Sex Cord-Stromal, Steroid Cell, and Other Ovarian Tumors with Endocrine, Paraendocrine, and Paraneoplastic Manifestations”). These neoplasms, in contrast to HT, are almost always unilateral and typically form distinct tumors or nodules appreciable on gross examination. Luteinized stromal cells, histologically similar to those present in HT, may also be encountered within the nonneoplastic stroma of a variety of benign and malignant ovarian tumors, including primary surface epithelial and germ cell tumors as well as metastatic tumors, that is, “tumors with functioning stroma” (▶ Chap. 15, “Sex Cord-Stromal, Steroid Cell, and Other Ovarian Tumors with Endocrine, Paraendocrine, and Paraneoplastic Manifestations”). Clinical Behavior and Treatment In contrast to patients with PCOS, those with HT usually exhibit little or no response to clomiphene treatment (Judd et al. 1973). Many patients require bilateral oophorectomy to halt progressive virilization. Such treatment may also result in disappearance of hypertension and abnormalities in glucose tolerance. More recently, successful treatment of HT has been achieved with gonadotropin-releasing hormone (GnRH) agonists (Vollaard et al. 2011).

744

HAIR-AN Syndrome In addition to the common occurrence of insulin resistance and hyperinsulinemia in patients with HT, some patients with HT have the HAIR-AN syndrome, a syndrome estimated to occur in as many as 5% of all women with hyperandrogenism (Barbieri and Ryan 1983; Barbieri et al. 1988; Dunaif et al. 1985; Massachuessetts General Hospital Case Records 1982). The syndrome consists of hyperandrogenism (HA), typically of early, sometimes premenarcheal, onset; insulin resistance (IR); and acanthosis nigricans (AN) (Barbieri and Ryan 1983). Striking degrees of masculinization are present in some patients with the HAIR-AN syndrome and may be disproportionate to the degree of hyperandrogenism (Dunaif et al. 1985). Some patients have a normal glucose tolerance, whereas others have symptomatic diabetes (Kahn et al. 1976). The syndrome has been most frequently described in patients with PCOS, although it appears likely that most, if not all, such patients also have HT (Dunaif et al. 1985; Massachuessetts General Hospital Case Records 1988). Unusual histologic findings in patients with HT and the HAIR-AN syndrome have included prominent follicular atresia, large numbers of degenerating oocytes, medullary stromal fibrosis, and numerous small nests of granulosa cells forming Call–Exner bodies (Massachuessetts General Hospital Case Records 1982). Dermoid cysts and stromal luteoma have been rarely described in patients with the HAIR-AN syndrome in association with HT, sclerocystic ovaries, or both. The typical laboratory findings include hyperinsulinemia and increased production rates and elevated serum levels of testosterone and androstenedione (Dunaif et al. 1985). In some patients, the severity of the insulin resistance is proportional to the testosterone elevation. Proposed mechanisms of insulin resistance have included a decreased number or functional capacity of insulin receptors, which may be associated with obesity, or in other cases, genetic alterations in the structure of the receptors (type A); antiinsulin receptor antibodies that decrease insulin receptor affinity for insulin and which are often associated

J. A. Irving and P. B. Clement

with autoimmune diseases (type B); and postreceptor defects in insulin action or clearance (type C) (Dunaif et al. 1985; Kahn et al. 1976). It has been postulated that the primary defect in the HAIR-AN syndrome is insulin resistance leading to hyperinsulinemia and the other findings in the syndrome. Thus, any cause of insulin resistance leading to hyperinsulinemia can produce the HAIR-AN syndrome. The hyperandrogenism itself may increase the severity of the insulin resistance, and thus a self-perpetuating cycle that increases in severity may result (Barbieri and Ryan 1983). The acanthosis nigricans is probably an epiphenomenon secondary to the hyperandrogenism, hyperinsulinemia, or both. Bilateral oophorectomy in patients with the HAIR-AN syndrome decreases hyperandrogenism but usually does not ameliorate insulin resistance (Massachuessetts General Hospital Case Records 1982; Nagamani et al. 1986). Gonadotropin suppression with oral contraceptives has been successful in decreasing ovarian androgen production in some patients. Marked improvement of acanthosis nigricans may follow correction of the hyperandrogenism.

Massive Edema and Ovarian Fibromatosis Tumor-like enlargement of one, or occasionally, both ovaries secondary to an accumulation of edema fluid within the ovarian stroma is referred to as massive ovarian edema. Over 100 cases of this disorder have been reported (Nogales et al. 1996; Roth et al. 1979; Young and Scully 1984). A rarer lesion designated ovarian fibromatosis (Young and Scully 1984), characterized by diffuse ovarian fibrosis, is closely related to massive edema and is therefore considered in this section.

Clinical Features Patients with massive edema are typically young, with a mean age of 21 years (range, 6–37 years), and have abdominal or pelvic pain, menstrual irregularities, and abdominal distension. The pain may be of several years duration or have a sudden onset and mimic the pain of acute

12

Nonneoplastic Lesions of the Ovary

appendicitis. Androgenic manifestations are present in approximately 20% of patients and are nearly always associated with the presence of luteinized stromal cells. Of these, two thirds are masculinized and the rest exhibit only hirsutism (Young and Scully 1984). Serum testosterone has been elevated in some cases. Rare patients have had estrogenic manifestations, manifested by isosexual pseudoprecocity (Nogales et al. 1996; Roth et al. 1979; Young and Scully 1984). Pelvic examination typically reveals a palpable adnexal mass, which in 70% of cases has been right-sided. Abdominal exploration reveals unilateral involvement in 90% of cases, and in approximately half the patients, partial or complete torsion of the involved ovary. In one patient, the contralateral ovary had a twisted pedicle and was infarcted. Intraperitoneal fluid is not usually present, although rare patients have had an associated Meigs’ syndrome. Rare cases of ovarian edema secondary to lymphatic permeation by metastatic cervical carcinoma have also been reported (Krasević et al. 2004). Patients with ovarian fibromatosis have ranged in age from 13 to 39 years, with an average of 25 years (Young and Scully 1984). Clinical manifestations include menstrual abnormalities or amenorrhea, abdominal pain, and rarely hirsutism or virilization. The majority of patients have a palpable adnexal mass. Occasionally, the ovarian enlargement is an incidental finding late in pregnancy or during cesarean section. In some cases, the involved ovary was found twisted on its pedicle at the time of operation. The endocrine manifestations, including, in several cases, infertility, disappear after oophorectomy, indicating that the lesion produces steroid hormones. A case of unilateral massive ovarian edema arising in association with a large leiomyoma of the ipsilateral broad ligament has also been reported (Harrison et al. 2014)

Gross Findings The involved ovary in massive edema is enlarged, soft, and fluctuant, ranging from 5.5 to 35 cm in maximum dimension (mean, 11.5 cm). The heaviest ovary weighed 2,400 g (Young and Scully 1984). The ovary has a shiny, white, smooth

745

Fig. 34 Massive ovarian edema. The ovary is enlarged and markedly edematous

exterior as a result of a white and fibrotic superficial cortex and a sectioned surface that is edematous or gelatinous and exudes watery fluid (Fig. 34). Occasional superficial FCs may be present. The ipsilateral fallopian tube may also be edematous. In ovarian fibromatosis, there is usually complete or almost complete ovarian involvement by a fibromatous process (Young and Scully 1984). In 20% of cases, the process is bilateral. The ovaries are 8–14 cm in maximum dimension with white and typically smooth or lobulated external surfaces. The cut surfaces are firm, white to gray, and solid except for the presence of cystic follicles in one third of cases (Fig. 35).

Microscopic Findings The striking finding on low magnification in massive edema is marked, diffuse, stromal edema that separates and sometimes involves the follicular structures but typically spares the superficial cortex (Fig. 36). The latter is usually thickened and fibrotic. Higher magnification reveals spindleshaped ovarian stromal cells separated by abundant pale-staining fluid that focally may impart a microcystic appearance. In nonedematous areas, the stroma has the appearance of normal stroma, hyperplastic stroma, or ovarian fibromatosis (Young and Scully 1984). In approximately 40% of cases, foci of luteinized cells are present (Fig. 37). Associated nonspecific findings include vascular and lymphatic dilatation within the ovary

746

J. A. Irving and P. B. Clement

Fig. 35 Bilateral ovarian fibromatosis. Convoluted external surface with patchy areas of white, fibrous tissue (right ovary)

Fig. 36 Massive ovarian edema. The edematous ovarian stroma separates several corpora fibrosa

Fig. 37 Massive ovarian edema. Luteinized stromal cells are present

and occasionally the mesosalpinx, focal necrosis, extravasated erythrocytes, hemosiderin-laden macrophages, and mast cells (Roth et al. 1979; Young and Scully 1984). The contralateral ovary is normal in more than 75% of cases; in the rest, it is enlarged and edematous or is nonedematous but altered by stromal HT or sclerocystic changes. Ovarian fibromatosis is characterized by a fibromatoid proliferation of collagen-producing spindle cells (Fig. 38) that typically surrounds normal follicular structures and thickens the superficial cortex (Fig. 39) (Young and Scully 1984). In most cases of fibromatosis, the process is diffuse but it may be localized, and occasionally it is confined to or predominantly involves the cortex (“cortical fibromatosis”). The process varies from moderately cellular fascicles of spindle cells with a focal storiform pattern to relatively acellular bands of dense collagen. Small foci of uninvolved ovarian stroma are usually present. In rare cases, luteinized cells are seen within the lesion or the adjacent nonfibrotic stroma. Minor foci of stromal edema and microscopic foci of sex cord elements within the fibromatous tissue, alone or in combination, have been encountered in occasional cases (Young and Scully 1984).

Pathogenesis The pathogenesis of massive edema is thought to be intermittent torsion of the ovary on its pedicle,

12

Nonneoplastic Lesions of the Ovary

Fig. 38 Ovarian fibromatosis. Fibrotic spindle cell proliferation invests follicular structures within the cortex

Fig. 39 Ovarian fibromatosis. Fibrotic ovarian stroma surrounds a preantral follicle

causing partial obstruction of venous and lymphatic drainage. Torsion is observed in half the cases of massive edema, and a few cases of massive edema have been reported in association with obstruction of ovarian lymphatics secondary to metastatic carcinoma within pelvic and paraaortic lymph nodes (Krasević et al. 2004; Young and Scully 1984). Luteinization of the ovarian stromal cells is considered a secondary phenomenon. In at least some cases, massive edema likely occurs in an ovary with an underlying stromal proliferation, either fibromatosis or stromal HT, that enlarges the ovary, promoting torsion with subsequent edema (Young and Scully 1984). This interpretation is supported by the clinical similarities and pathologic overlap between

747

massive edema and ovarian fibromatosis. Young and Scully suggest that massive edema is simply ovarian fibromatosis following torsion and accumulation of edema fluid (Young and Scully 1984). Similarly, some examples of massive edema in which luteinized stromal cells are present in the same ovary and in the contralateral, edematous, or nonedematous ovary may represent cases of stromal HT in which one or both ovaries have undergone torsion. Rather than accepting that fibromatosis is a precursor of massive edema, Russell and Farnsworth (1997) hypothesized that the fibromatoses described by Young and Scully represent the “burned-out” stage of a reactive fibroblastic proliferation that at one end of the spectrum is represented by massive edema and at the other by a variety of highly cellular fibroblastic tumorlike lesions.

Differential Diagnosis The differential diagnosis of massive edema includes ovarian neoplasms that may exhibit an edematous or myxoid appearance, most commonly fibroma, but also sclerosing stromal tumor, Krukenberg tumor, luteinized thecoma associated with sclerosing peritonitis, and the rare ovarian myxoma. Recognition of massive edema is therefore of great importance to prevent unnecessary oophorectomy in a young female. Fibromatosis also may be confused with a fibroma, or if sex cord-like nests are prominent, a Brenner tumor. Massive edema and fibromatosis are distinguished from a neoplasm by the presence of follicular derivatives visible on both macroscopic and microscopic examination. A neoplasm may be surrounded by a rim of normal ovarian tissue in contrast to massive edema and fibromatosis, which usually diffusely involve the ovarian tissue. Additionally, ovarian fibromas occur in an older age group and are hormonally inactive, and Krukenberg tumors, in contrast to massive edema and fibromatosis, are characterized by signet-ring cells. Finally, the sex cord-like nests in fibromatosis are distinguishable from those of a Brenner tumor in their number, shape, and cell type.

748

J. A. Irving and P. B. Clement

Clinical Behavior and Treatment Although most of the reported patients with massive edema have been successfully treated by oophorectomy, the condition should be managed conservatively, especially if the patient is young, because there is a strong likelihood that the condition will resolve. After an intraoperative frozen section of a wedge biopsy to exclude a neoplasm, an ovarian suspension procedure should be performed, with fixation of the involved ovary.

Pregnancy Luteoma

Fig. 40 Pregnancy luteoma. Multiple, solid, circumscribed, reddish-brown, and hemorrhagic nodules replace the normal parenchyma

Pregnancy luteoma is a distinctive, nonneoplastic lesion of pregnancy, characterized by solid proliferations of luteinized cells, resulting in tumorlike ovarian enlargement that regresses during the puerperium (Norris and Taylor 1967; Burandt and Young 2014).

is made by excisional biopsy and frozen section examination of one nodule.

Clinical Features The patients are usually in their third or fourth decades and are multiparous in 80% of the cases, and a similar proportion are black. Most patients are asymptomatic, and the ovarian enlargement is discovered incidentally at or near term cesarean section or postpartum tubal ligation. Rarely, a pelvic mass is palpable or obstructs the birth canal. In approximately 25% of cases, hirsutism or virilization appears or worsens during the latter half of pregnancy. Seventy percent of female infants born to virilized mothers are born with clitoromegaly and labial fusion. Plasma testosterone and other androgens may reach levels 70 times normal in virilized patients; increased values have also been demonstrated in nonvirilized patients (Nagamani et al. 1982). Androgen levels in the infants may be increased but are usually lower than maternal levels or normal (Nagamani et al. 1982). Regression of the luteomas usually begins within days after delivery and is complete within several weeks. Simultaneously, elevated androgen levels decrease rapidly, usually normalizing within 2 weeks postpartum. In rare cases, pregnancy luteomas occur in consecutive pregnancies. The diagnosis

Gross Findings Pregnancy luteomas are solid, fleshy, circumscribed, red to brown, and occasionally black or yellow nodules ranging from microscopic up to 15–20 cm in diameter (median, 6.6 cm) (Fig. 40) (Norris and Taylor 1967; Burandt and Young 2014). Hemorrhagic foci are common. The lesions are multiple in almost half the cases and bilateral in at least one third. A separate corpus luteum of pregnancy may also be visible. Examination of the ovaries days to weeks postpartum reveals brown puckered scars. Microscopic Findings The lesions are composed of sharply circumscribed, diffuse rounded masses of cells (Fig. 41) that are also occasionally arranged in a trabecular or follicular pattern, the latter associated with spaces containing colloid-like material (Fig. 42a). The cells are intermediate in size between the luteinized granulosa cells and luteinized theca cells of adjacent follicles and have abundant eosinophilic cytoplasm that contains little or no stainable lipid and central nuclei (Fig. 42b). The nuclei may vary slightly in size and are hyperchromatic; nucleoli may be prominent. Mitotic figures may range up to seven per ten high-power fields, with an average of two or three,

12

Nonneoplastic Lesions of the Ovary

and may be atypical (Norris and Taylor 1967). Less common features include focal balloon-like cytoplasmic degeneration and colloid droplets similar to those seen in the corpus luteum of pregnancy. The stroma is scanty, and reticulin fibrils surround groups of cells. Examination of lesions removed postpartum shows shrunken aggregates of degenerating lipid-filled luteoma cells with pyknotic nuclei, infiltration by lymphocytes, and fibrosis.

Pathogenesis Pregnancy luteomas most likely arise from hCG-induced proliferations of luteinized ovarian stromal cells. Some authors, however, have favored origin from luteinized follicular granulosa and theca cells (Norris and Taylor 1967). The exclusive occurrence of the lesion in pregnancy

Fig. 41 Pregnancy luteoma. Solid, circumscribed nodule of luteinized cells replaces the normal parenchyma

749

suggests a role for hCG in its pathogenesis, and augmentation of steroidogenesis by pregnancy luteomas in response to hCG, both in vitro and in vivo, supports this interpretation. However, the rarity of pregnancy luteomas in association with GTD, which is typically accompanied by very high levels of hCG, and the almost exclusive recognition of the lesions during the third trimester when hCG levels are lower than earlier in pregnancy, indicate that hCG is not the only factor in their development. The occasional history of hirsutism, sometimes familial, antedating the pregnancy suggests that a preexistent endocrinopathy, such as stromal HT or PCOS, may predispose to the development of the lesion in some patients.

Differential Diagnosis When pregnancy luteomas are multiple, intraoperative inspection may suggest nodules of metastatic tumor. Such a diagnosis can usually be excluded by frozen section examination of one of the nodules, but the distinction may be difficult if the patient has a history or clinical evidence of an oxyphilic malignant tumor such as a malignant melanoma. When the luteoma is a single nodule, the microscopic differential diagnosis includes a number of lesions composed of luteinized cells occurring during pregnancy. However, the typical gross appearance of the pregnancy luteoma readily distinguishes it from a large solitary luteinized FC of pregnancy and puerperium, HL, and corpus luteum of pregnancy. Solid primary neoplasms composed

Fig. 42 Pregnancy luteoma. (a) Follicular pattern; (b) solid growth pattern of polygonal luteinized cells

750

partially or entirely of luteinized cells such as granulosa tumors, thecomas, and steroid cell tumors may occur during pregnancy and enter the differential diagnosis (Burandt and Young 2014). Such tumors are almost always unilateral and solitary compared to the more frequent bilaterality and multinodularity of the pregnancy luteoma. The partly luteinized group, that is, luteinized granulosa cell tumors and luteinized thecomas, contain typical nonluteinized foci and usually have denser reticulum patterns and more abundant intracellular lipid than seen in pregnancy luteoma. Entirely luteinized tumors belonging to the steroid cell category may closely resemble pregnancy luteoma histologically. Features favoring a steroid cell neoplasm include a dense reticulum pattern, intracellular lipid, lipochrome pigment, and in Leydig cell tumors, a hilar location, and the presence of Reinke crystals (▶ Chap. 15, “Sex Cord-Stromal, Steroid Cell, and Other Ovarian Tumors with Endocrine, Paraendocrine, and Paraneoplastic Manifestations”). Differentiation of a solitary pregnancy luteoma from a lipidpoor steroid cell tumor may be impossible, but such a lesion in a pregnant woman is generally considered a pregnancy luteoma until proven otherwise.

J. A. Irving and P. B. Clement

Granulosa Cell Proliferations of Pregnancy

Clinical Behavior and Treatment Because the pregnancy luteoma is a benign, selflimited condition, no treatment is required.

General and Pathologic Features Granulosa cell proliferations that simulate small neoplasms have been encountered as incidental findings in the ovaries of pregnant women (Clement et al. 1988). The older literature documented the presence of similar lesions in the ovaries of nonpregnant women, and we have encountered them in the ovary of a newborn that also contained a corpus luteum. The lesions in pregnant women are usually multiple and lie within atretic follicles, which are typically enveloped by a thick layer of luteinized theca cells. The granulosa cells may be arranged in solid, insular, microfollicular (Fig. 43), or trabecular patterns, mimicking similar patterns in clinically evident granulosa cell tumors. In one case, a solid tubular pattern was identical to that seen in some Sertoli cell tumors. The granulosa cells typically contain scanty cytoplasm and grooved nuclei, resembling the cells of the adult-type granulosa cell tumor and in the case with a sertoliform pattern, the cells contained moderate amounts of finely vacuolated cytoplasm, suggesting the presence of lipid. In one case, there were large nodules of luteinized granulosa cells with variably sized, round, nongrooved nuclei, resembling pregnancy luteomas except for their obvious origin in granulosa cells and the larger size of their cells.

Fig. 43 Granulosa cell proliferations of pregnancy mimicking a small granulosa cell tumor. The proliferations are within the center of an atretic follicle and are surrounded by the luteinized cells of the theca interna. (a) The

proliferating cells form an irregular island containing Call-Exner formations. (b) In a different case, the granulosa cells form small nests and contain nuclear grooves

12

Nonneoplastic Lesions of the Ovary

Differential Diagnosis The differential diagnosis in most of the cases is with a small granulosa or Sertoli cell tumor. Although similar proliferations have been previously interpreted as small tumors, the frequency of the lesions during pregnancy suggests an unusual nonneoplastic response to the hormonal milieu, possibly to the FSH-like property of hCG. The microscopic size of the lesions, their multifocality, and their confinement to atretic follicles support this interpretation.

751

Leydig cells typically occur in the ovarian hilus, where they are also referred to as hilus cells and can be found in virtually all adult ovaries, typically intermingled with nonmyelinated nerves. Rarely, Leydig cells occur in nonhilar locations, either within the ovarian stroma or in extraovarian sites such as the lamina propria or adventitia of the fallopian tube (Honoré and O’Hara 1979). Stromal (nonhilar) Leydig cell hyperplasia, in which Reinke crystal-containing Leydig cells are present within the ovarian stroma, is much less common than hilar Leydig cell hyperplasia. One report documented bilateral stromal Leydig cell hyperplasia in a postmenopausal woman who presented with virilization; hilus cell hyperplasia was absent (Taylor et al. 2000).

More recently, a postmenopausal woman with hirsutism, elevated serum testosterone, and bilateral Leydig cell hyperplasia (ovarian location not otherwise specified), was reported, with symptomatic improvement and normalization of testosterone levels on follow-up (Hofland et al. 2013). Stromal Leydig cells are likely the origin of the rare Leydig cell tumors encountered within the ovarian stroma (▶ Chap. 15, “Sex Cord-Stromal, Steroid Cell, and Other Ovarian Tumors with Endocrine, Paraendocrine, and Paraneoplastic Manifestations”). In one such case, there was also bilateral hilar Leydig cell hyperplasia and bilateral hilar Leydig cell tumors (Sternberg and Roth 1973). Stromal Leydig cells have also been rarely encountered within the nonneoplastic stroma of a variety of ovarian neoplasms and cysts, including mucinous and serous cystadenomas, Brenner tumors, struma ovarii, and strumal carcinoid tumors (Rutgers and Scully 1986). Hilar Leydig cell hyperplasia is difficult to define because hilus cell nests are typically widely separated and cannot be quantitated adequately without sectioning both ovaries extensively. Also, hilus cell proliferation can occur physiologically as a result of elevated hCG or LH levels, such as during pregnancy and after the menopause (Fig. 44) (Clement 2007). In such cases, the proliferation is often mild and

Fig. 44 Hilar Leydig cell hyperplasia in a postmenopausal woman. (a) Nodular proliferation of hilar Leydig cells; (b) the hilus cells have abundant eosinophilic

cytoplasm. No crystals are seen in this field, but spherical hyaline globules are present (lower right), likely representing crystal precursors

Leydig Cell Hyperplasia

752

generally not accompanied by a clinical endocrine disturbance, although such proliferations may account for at least some of the hirustism that is frequently observed during pregnancy. Severe degrees of hyperplasia, often associated with virilization, may occur in both pregnant and nonpregnant women. In some cases, elevated serum testosterone levels have been documented (Hofland et al. 2013). Hilus cell hyperplasia is characterized by an increased number of cells in a nodular or, less commonly, a diffuse arrangement, increased cell size, the presence of mitotic figures, cellular and nuclear pleomorphism, hyperchromasia, and multinucleation; crystals of Reinke may or may not be apparent by light microscopy (Figs. 44 and 45) (Sternberg and Roth 1973). Hilar Leydig cell hyperplasia may be associated with other ovarian lesions, including SH, stromal HT, stromal Leydig cell hyperplasia, rete cysts (Miscellaneous Lesions), and hilus cell neoplasia (Sternberg and Roth 1973). One case of hilus cell hyperplasia has been associated with the resistant ovary syndrome and other cases have been associated with gonadal dysgenesis (Judd et al. 1970); in both disorders, LH levels are elevated. From a pathologic point of view, the distinction between a large hyperplastic nodule of hilus cells and a hilus cell tumor is arbitrary; we diagnose neoplasia when the nodule is more than 1 cm in diameter.

Fig. 45 Hilar Leydig cell hyperplasia. Some nuclei are enlarged and hyperchromatic. Many crystals are seen in this field

J. A. Irving and P. B. Clement

Ovarian “Tumor” of the Adrenogenital Syndrome One ovarian example of this lesion has been reported (Al-Ahmadie et al. 2001). A 36-yearold woman with congenital adrenal hyperplasia presented with an abrupt aggravation of her virilizing symptoms, relieved by bilateral salpingo-oophorectomy. Ovarian or paraovarian soft brown masses were present in both adnexae, which on microscopic examination were identical to the testicular tumor of the adrenogenital syndrome.

Ovarian Stromal Metaplasias Including Decidual Reaction The ovarian stromal cell has the potential to differentiate, presumably by a process of metaplasia, into a variety of other mesenchymal cell types, most commonly decidua, but rarely smooth muscle, fat, and bone.

Ovarian Decidual Reaction An ectopic decidual reaction may be encountered within the ovarian stroma as an isolated finding or as part of a more widespread decidual transformation of the subperitoneal pelvic mesenchyme (▶ Chap. 13, “Diseases of the Peritoneum”) (Boss et al. 1965). As in other sites of the secondary müllerian system, an ovarian decidual reaction usually represents a response of the indigenous stromal cells to the hormonal milieu of pregnancy. Ectopic decidua may be seen as early as the 9th week of gestation and is present in almost all ovaries at term. Less commonly, the decidua is associated with trophoblastic disease, in patients treated with progestagens, in the vicinity of a corpus luteum, and in association with hormonally active neoplastic and nonneoplastic lesions of the ovaries and adrenal glands (Boss et al. 1965; Ober et al. 1957). Prior ovarian radiation may be a predisposing factor by increasing the sensitivity of the stromal cells to hormonal stimulation (Ober

12

Nonneoplastic Lesions of the Ovary

et al. 1957). Foci of ectopic decidua occasionally may occur within the ovaries of pre- and postmenopausal women with no obvious cause (Ober et al. 1957). The decidual foci are usually seen only microscopically but in some cases, may be visible on macroscopic examination as variably sized, soft, tan to hemorrhagic nodules, or patches. The decidual cells typically occur singly, as small nodules, or confluent sheets in the superficial cortical stroma and the ovarian surface, often within periovarian adhesions (Fig. 46). The cells are indistinguishable from eutopic decidua on light microscopic and ultrastructural examination. Smooth muscle cells, probably derived from submesothelial fibroblasts or the ovarian stroma, may be admixed. A rich vascular

753

network of distended capillaries and a sprinkling of lymphocytes are typically found within the decidual foci. Focal nuclear pleomorphism and hyperchromasia, sometimes in association with hemorrhagic necrosis, should not be misinterpreted as evidence of a malignant tumor. Occasionally, ectopic decidual cells may have vacuoles and eccentric nuclei simulating signetring cells (Clement et al. 1989). The bland appearance of most of the nuclei, the absence of mitotic figures, the periodic acid–Schiff (PAS) negativity of the vacuolar contents, and the association with pregnancy should facilitate the correct diagnosis. Ectopic decidua in postpartum patients may undergo hyalinization.

Rarer Ovarian Stromal Metaplasias and Calcification

Fig. 46 Ectopic decidua. Small nodules of decidual cells involving ovarian surface adhesions

Foci of metaplastic smooth muscle (Fig. 47a) may be rarely encountered in the ovarian stroma of otherwise normal ovaries, within hyperplastic ovarian stroma (as in stromal HT or polycystic ovaries), or within the walls of nonneoplastic or neoplastic cysts (Hughesdon 1982; Scully 1981; Seidman and Krishnan 2016). Foci of mature fat have been described as a rare incidental histologic finding within the superficial ovarian stroma in obese women (Fig. 47b) (Honoré and O’Hara 1980; Seidman and Krishnan 2016). Heterotopic bone formation in the ovary in

Fig. 47 Ovarian stromal metaplasia. (a) Foci of smooth muscle are separated by ovarian stromal cells. (b) A focus of adipose tissue lies within the ovarian stroma, surrounded by epithelial inclusion glands

754

Fig. 48 Idiopathic ovarian calcification. Numerous spherical, laminated, and calcific foci without accompanying epithelial cells occupy the ovarian stroma. Both ovaries were stony hard on gross examination

the absence of an ovarian neoplasm is also unusual, typically occurring within periovarian adhesions or the walls of endometriotic cysts but rarely within otherwise normal ovaries (Shipton and Meares 1965). In one case, extensive idiopathic calcification resulted in a stony hard consistency of both ovaries, which were of normal size (Clement and Cooney 1992). Microscopic examination showed numerous spherical, laminated, calcific foci without accompanying epithelial cells (Fig. 48). This process must be distinguished from an atypical proliferative serous tumor/serous borderline tumor or carcinoma with confluent psammoma bodies, in which at least occasional neoplastic epithelial cells should be identified, and from a “burned-out” gonadoblastoma replaced by laminated calcified masses. In cases of the latter type, the patient almost always has evidence of abnormal gonadal development and Y-chromosome material in her karyotype as well as residual typical gonadoblastoma in the same or contralateral gonad.

J. A. Irving and P. B. Clement

van Karseren and Schoemaker 1999). POF is uncommon, affecting an estimated 1% of women under the age of 40 (Kalu and Panay 2008) and accounting for only 4–10% of patients with secondary amenorrhea (Russell et al. 1982). The ovarian failure is usually permanent, but occasionally it is reversible, at least temporarily, as manifested by subsequent ovulation and even conception (Rebar et al. 1982; van Karseren and Schoemaker 1999). Patients with POF typically have a 46XX karyotype, normal secondary sexual characteristics, and secondary amenorrhea, although rarely prepubertal ovarian failure causes primary amenorrhea or oligomenorrhea and incompletely developed secondary sexual features. POF, therefore, probably represents a continuum in which individuals may be affected at any age before the expected age of menopause (Rebar et al. 1982). In contrast to patients with POF, patients with gonadal dysgenesis usually have an abnormal karyotype, streak gonads or abdominal testes, primary amenorrhea, ambiguous internal and external genitalia, and somatic abnormalities. The absence or decline in follicular activity in patients with POF typically results in low serum estrogen levels, often accompanied by estrogen withdrawal symptoms. Because of the failure of negative feedback, the low estrogen levels lead to elevated levels of pituitary gonadotropins, a feature that differentiates POF from central causes of amenorrhea related to hypothalamic or pituitary dysfunction. Although three histologic patterns have been recognized, specifically premature follicular depletion (true premature menopause), resistant ovary syndrome, and autoimmune oophoritis, it is not known with certainty if each represents a distinct disorder or a nonspecific morphologic manifestation of a number of different disorders (Russell et al. 1982).

Disorders of Ovarian Failure Premature Follicular Depletion Premature ovarian failure (POF) or premature menopause is a result of a variety of disorders that lead to the onset of amenorrhea and infertility before the age of 35 years, or according to others, 40 years (Rebar et al. 1982; Russell et al. 1982;

This disorder is characterized by ovaries that are typically small on gross inspection, resembling streak gonads. On microscopic examination, there is premature follicular depletion, with the

12

Nonneoplastic Lesions of the Ovary

ovaries resembling normal peri- or postmenopausal ovaries with complete, or nearly complete, absence of primordial and developing follicles (Russell et al. 1982). Follicles in varying stages of atresia and stigmata of prior ovulation are typically present, excluding a streak gonad. Postulated pathogenetic mechanisms include a decreased number of ovarian germ cells at birth, acceleration of normal follicular atresia, or prepubertal or postpubertal destruction of germ cells. With respect to the last, there is strong evidence, including the presence of antiovarian antibodies, autoimmune disorders, or both, implicating immune factors in a substantial proportion of women with POF. As some or all these cases likely represent an end stage of autoimmune oophoritis, they are considered further in that section (see page 756) (Russell et al. 1982). Additionally, postnatal destruction of germ cells may be caused by cytotoxic drugs or radiation (Ovarian Changes Secondary to Cytotoxic Drugs and Radiation) and mumps oophoritis (Viral Infections). Because mumps oophoritis is probably clinically occult in the majority of cases, it may be a more frequent cause of premature menopause, including familial cases, than is generally suspected. The occurrence of familial cases of POF, in a pattern consistent with an autosomal dominant mode of inheritance (Mattison et al. 1984), implicates genetic factors in some cases; familial POF may account for 12–28% of cases of this disorder (Kalu and Panay 2008). Occasional patients with an otherwise typical presentation have had chromosomal abnormalities, usually 47XXX, pure or mosaic, but occasionally 45XO/46XX (Villanueva and Rebar 1983). In some familial cases, the affected women were 46XX but had an interstitial deletion of the long arm of the X chromosome (Krauss et al. 1987). A proportion of these patients have fragile X syndrome, with a premutation in the FMR1 gene found in 14% of cases (Sherman 2000). Some authors, however, exclude cases with chromosomal abnormalities from the category of POF so that it includes only “chromosomally competent” patients. A detailed discussion of genetic alterations in POF is provided in a recent comprehensive review (Rossetti et al. 2017).

755

The presence of galactosemia in some patients with POF suggests that it may play a pathogenetic role. Approximately two thirds of females with galactosemia in one study had POF (Kaufman et al. 1981). In many such patients, dietary treatment of galactosemia had been delayed. One galactosemic patient with POF had the pattern of the resistant ovary syndrome (see following) on ovarian biopsy, but in a series of galactosemic patients who were not biopsied, some patients had severely atrophic ovaries, suggesting premature follicular depletion (Escobar et al. 1982). Similarly, experimental studies indicate that galactose or its metabolites may interfere with normal prenatal oogenesis. Depletion of primordial follicles has been described in several women with ataxia telangiectasia, which may be related to their severe immunosuppression or to their athymic state (Friedman et al. 1984).

Resistant Ovary Syndrome This rare syndrome, also known as Savage syndrome, is found in approximately 20% of patients with POF and characterized by primary or secondary amenorrhea, endogenous hypergonadotropinemia, and resistance to exogenous gonadotropins, often in massive doses (Massachussetts General Hospital Case Records 1986; Russell et al. 1982). The resistance to endogenous and exogenous gonadotropins may be relative or absolute, episodic, or chronic. The ovaries typically have a normal prepubertal or adult appearance on macroscopic inspection. Microscopic examination reveals an appropriate number of normal-appearing primordial follicles but a complete, or nearly complete, absence of developing follicles. Atretic follicles and stigmata of prior ovulation may be present. In occasional patients, the space normally occupied by the ovum in some of the atretic follicles contains calcified material. In another case, numerous abnormal preantral follicles were found that contained multiple nodules of basement membrane material (Massachussetts General Hospital Case Records 1986). A histologic pattern similar to that in the resistant ovary syndrome occurs in

756

morbid obesity, Cushing’s syndrome, and hypogonadotropic ovarian failure secondary to hypothalamic–pituitary dysfunction. The pathogenesis of this disorder is not yet established, but a possible deficiency of FSH and LH receptors within the ovary, the presence of antibodies to these receptors, or a postreceptor defect have been postulated. An IgG-like substance that alters FSH receptors and thereby impairs binding of this hormone was present in the serum of several patients with associated myasthenia gravis (Escobar et al. 1982). In another patient with the resistant ovary syndrome, lupus erythematosus appeared while the ovarian failure was evolving, and a serum antibody specific for the FSH receptor was found (Massachusetts General Hospital Case Records 1986). In another study, circulating autoimmune antibodies to thyroglobulin and smooth muscle were found in some patients (Russell et al. 1982). As noted earlier, one patient with the resistant ovary syndrome had galactosemia (Russell et al. 1982).

Autoimmune Oophoritis Clinical and Pathogenetic Features Approximately 25 cases of autoimmune oophoritis have been documented pathologically (Bannatyne et al. 1990; Jacob and Koc 2015; Lonsdale et al. 1991; Russell et al. 1982; Sedmak et al. 1987). The patients, who have ranged in age from 17 to 48 years (mean, 31), typically present with oligomenorrhea or amenorrhea, or symptoms relating to multiple follicular cysts, including pelvic pain, adnexal torsion, or estrogenic manifestations, such as abnormal bleeding, and in one case endometrial adenocarcinoma (Bannatyne et al. 1990; Lonsdale et al. 1991). Most patients have steroid cell antibodies in their sera; Addison’s disease, Hashimoto’s thyroiditis, Sjogren’s disease, and myasthenia gravis have been additionally present in some cases (Bannatyne et al. 1990; Çakır et al. 2011; Jacob and Koc 2015; Euthymiopoulou et al. 2007). The Addison’s disease may arise at the same time or subsequent to the ovarian failure and is associated at least in some cases with a lymphocytic

J. A. Irving and P. B. Clement

adrenalitis. The steroid cell antibodies, which are rare in the general population, belong to a group of antibodies reactive with a range of antigens in steroid-producing cells. They are typically reactive against adrenal cortex, but in some cases also to theca interna, corpus luteum, thyroid epithelium and thyroglobulin, parathyroid cells, gastric parietal cells, and thymocytes, alone or in combination (Damewood et al. 1986; Russell et al. 1982). In one 18 year-old myasthenia gravis patient, following thymectomy, there was clinical resolution of amenorrhea secondary to autoimmune oophoritis (Çakır et al. 2011). There is also evidence supporting a role for cell-mediated immune mechanisms in the pathogenesis of autoimmune oophoritis. Studies have shown expression of major histocompatibility class II antigens by granulosa cells in autoimmune oophoritis, a phenomenon inducible by interferon gamma, a product of activated T cells (Hill et al. 1990). Additionally, there have been reports of occasional patients with POF, including some with documented autoimmune oophoritis, in whom menses and ovulation resumed after administration of corticosteroids (Cowchock et al. 1988). All the foregoing observations suggest that a complex immune process with an interplay of humoral and cellular mechanisms is involved in the pathogenesis of autoimmune oophoritis (Sedmak et al. 1987). Autoimmune oophoritis is almost certainly more common than the small number of histologically documented cases would suggest. In two studies of women with POF who were not biopsied or in whom a biopsy revealed an afollicular pattern, two thirds (Pekonen et al. 1986) and 90% (Mignot et al. 1989) of patients, respectively, had some evidence of autoimmune phenomena with immunologic testing. In a third study, some women with POF had a decreased ratio of inducer/helper lymphocytes to suppressor/cytotoxic lymphocytes as well as a decreased concentration of serum IgA, suggesting a mild suppression of immune competence (Friedman et al. 1984). Similarly, as many as one half of patients with POF in some series have or subsequently develop one or more associated autoimmune disorders; an average figure calculated from

12

Nonneoplastic Lesions of the Ovary

the literature is 20% (LaBarbera et al. 1988). Addison’s disease or thyroid disease (Hashimoto’s thyroiditis, Grave’s disease) is the most common of these disorders and is typically accompanied or preceded by the presence of steroid cell antibodies (Coulam 1983; Reato et al. 2011). Conversely, as many as 25% of patients with idiopathic Addison’s disease may have POF, the latter usually preceding the former by several, but occasionally many, years (LaBarbera et al. 1988). Other autoimmune diseases that occur less commonly in these patients include rheumatoid arthritis, hypoparathyroidism, myasthenia gravis, diabetes mellitus, atrophic gastritis, pernicious anemia, hemolytic anemia, idiopathic thrombocytopenia purpura, alopecia, vitiligo, and sicca syndrome (Coulam 1983; LaBarbera et al. 1988). POF occurs frequently in patients with two or more such diseases (polyglandular endocrinopathy) (LaBarbera et al. 1988). Additionally, a subgroup of patients with POF have chronic mucocutaneous candidiasis or chronic vaginal candidiasis, suggesting a defect in T-cell function, possibly secondary to circulating antibodies against T lymphocytes demonstrable in some of these patients (Mathur et al. 1980). Patients with these two types of candidiasis also frequently have anti-Candida, antithymocyte, and antiovarian antibodies, suggesting a shared antigen on these cells (Mathur et al. 1980). A partial genetic basis is suggested by a family history of an autoimmune disease in approximately one fifth of patients who have both POF

757

and an autoimmune disease. Similarly, a prevalence of certain human leukocyte antigens (HLA) has been found in some patients with autoimmune endocrine disease (Walfish et al. 1983). The foregoing findings suggest that a substantial proportion of patients with the histologic pattern of premature follicular depletion likely represent an end stage of autoimmune oophoritis that is no longer recognizable on histologic examination.

Gross Findings On gross examination, the ovaries may be small or normal in size, but in one third of cases one or both are enlarged by multiple follicular cysts, potentially simulating cystic neoplasms (Bannatyne et al. 1990; Lonsdale et al. 1991). The cysts are more common in the earlier phases of the disease and are likely caused by elevated gonadotropin levels. Small ovaries presumably reflect a late- or end stage of the disorder, after complete destruction of the follicles. Microscopic Findings The cardinal feature of autoimmune oophoritis on microscopic examination is a folliculotropic lymphoid infiltrate that affects developing follicles with a theca layer, corpora lutea, and atretic follicles (Fig. 49). The intensity of the infiltrate increases with the degree of follicular maturation. The theca interna layer is typically more intensely infiltrated than the granulosa layer and may be focally destroyed; the granulosa layer may be focally disrupted with sloughing of its cells. The

Fig. 49 Autoimmune oophoritis. (a) A maturing follicle is infiltrated by mononuclear inflammatory cells. (b) Corpus luteum partially destructed by intense lymphoplasmacytic infiltrate

758

inflammatory infiltrate consists predominantly of lymphocytes and plasma cells, but eosinophils, histiocytes, and, rarely, sarcoid-like granulomas are also present and may predominate (Bannatyne et al. 1990). Primordial follicles are typically present but uninvolved. Additionally, perivascular and perineural lymphoid infiltrates may be found in the hilus, and in some such cases, there has been an absence of Leydig cells, suggesting destruction of the latter by the inflammatory process (Gloor and Hurlimann 1984). Nonspecific findings have included the presence of abnormal “dysplastic” follicles, FCs (as noted earlier), and superficial cortical fibrosis (Bannatyne et al. 1990). Immunophenotyping of the inflammatory infiltrate has revealed, variously, B and T lymphocytes, polyclonal plasma cells, macrophages, and natural killer cells (Gloor and Hurlimann 1984; Lonsdale et al. 1991). A predominance of CD4and HLA-DR-positive T-helper cells has also been demonstrated (Bats et al. 2008).

Vascular Lesions Ovarian Hemorrhage Rupture of a normal corpus luteum or a CLC may occasionally result in hemorrhage and, rarely, fatal hemoperitoneum. Although hemorrhage may occur in otherwise normal women, it is observed more often in women receiving anticoagulant therapy (Fig. 50) (Hallatt et al. 1984). The

J. A. Irving and P. B. Clement

right ovary is the source of the hemorrhage in almost two thirds of patients, and the clinical manifestations frequently resemble acute appendicitis (Hallatt et al. 1984).

Ovarian Torsion and Infarction Ovarian or adnexal torsion is most frequently a complication of an underlying ovarian lesion, usually a nonneoplastic cyst, abscess, or benign tumor but occasionally a malignant neoplasm (Hibbard 1985). Torsion of a normal ovary occurs rarely, especially in infants or children but also in adults. Bilateral adnexal torsion, synchronous or asynchronous, has been reported. The patients present with clinical findings similar to those of acute appendicitis or with recurrent episodes of abdominal pain; occasionally, an adnexal mass is palpable. Laparotomy reveals a swollen, hemorrhagic, and in some cases, infarcted, tuboovarian mass twisted on its pedicle. Where possible, detorsion is advocated, rather than oophorectomy, to preserve ovarian function, and some authors suggest that in such cases, ovarian fixation may minimize the potential risk of re-torsion, especially during pregnancy (Hyttel et al. 2015). In rare cases, the torsion and infarction may be asymptomatic, and autoamputation can result in a mass, which is occasionally calcified, lying free in the peritoneal cavity or attached to adjacent structures. The differential diagnosis in such cases, as noted earlier (Congenital Lesions and Ectopic Tissues), is with congenital unilateral absence of the ovary and tube. Torsion of the fallopian tube may rarely occur in isolation (Athanasias et al. 2013). It is crucial to examine thoroughly any hemorrhagic infarcted ovarian mass to exclude a neoplasm. A search should be made for viable foci at the periphery of the lesion, and the necrotic tissue should be scrutinized for shadows of neoplastic cells.

Ovarian Vein Thrombophlebitis Fig. 50 Ovarian hemorrhage. Ovary from a patient being treated with anticoagulation therapy is replaced by a large hematoma

Ovarian vein thrombosis or thrombophlebitis is an uncommon but potentially fatal disorder that most often occurs postpartum but may follow pelvic

12

Nonneoplastic Lesions of the Ovary

operations or pelvic trauma or complicate other pelvic disorders such as pelvic inflammatory disease (Witlin and Sibai 1995). Patients usually present with fever and lower abdominal pain and an abdominal mass, almost always on the right side. In one report, a nonpuerperal patient had left ovarian vein thrombophlebitis associated with a large ipsilateral uterine leiomyoma (Haynes et al. 2014). The clinical picture may simulate acute appendicitis or pyelonephritis. The marked rightsided predominance in the puerperal cases is explained on the basis of retrograde venous flow in the left ovarian vein during the puerperium, protecting that side from bacterial spread from the uterus. Sonography, computed tomography (CT), and magnetic resonance imaging studies may be useful in establishing the diagnosis preoperatively, thus avoiding unnecessary laparotomy, as the mainstays of treatment are anticoagulation and antibiotic therapy (Dessole et al. 2003). If surgery is undertaken, the involved ovarian vein will be markedly enlarged and the thrombus usually extends to the inferior vena cava on the right or to the renal vein on the left. Rarely, one or both of the latter structures are also thrombosed. There is marked edema and inflammation of the surrounding retroperitoneal tissues. The ipsilateral ovary is usually congested but not infarcted, although asymptomatic bilateral ovarian infarction in a postpartum patient secondary to massive pelvic venous thrombosis has been reported. Some cases may be associated with the ovarian vein syndrome (Rare Vascular Lesions).

Rare Vascular Lesions Giant cell arteritis can rarely involve the female genital tract, including the ovaries (Bell et al. 1986; Marrogi et al. 1991). The patients are almost invariably postmenopausal, and some patients with ovarian involvement have had systemic manifestations, such as a history of polymyalgia rheumatica or temporal arteritis, an elevated erythrocyte sedimentation rate (ESR), or both. Ovarian vasculitis has also been rarely reported in adolescence with lupus (Pereira et al. 2009). Less commonly, the arteritis is an incidental microscopic finding in an

759

Fig. 51 Vasculitis of polyarteritis nodosa type in a postmenopausal woman. Three ovarian hilar vessels are involved in this field. Polyarteritis also involved the uterine cervix in this case

asymptomatic patient. Treatment is probably unnecessary in this group of patients, but they should be followed carefully, including repeated determinations of the ESR (Bell et al. 1986; Marrogi et al. 1991). Rare examples of vasculitis of polyarteritis nodosa type involving the ovaries or ovarian hilus have also been reported (Fig. 51). Rarely, the finding is a reflection of systemic involvement, but usually polyarteritis in the female genital tract (typically the cervix) is an isolated finding without systemic manifestations on follow-up (Francke et al. 1998; Ganesan et al. 2000). A rare cause of retroperitoneal hemorrhage is rupture of an ovarian artery or vein, typically during pregnancy or the puerperium (Ginsburg et al. 1987). In some cases, the rupture represents a complication of an aneurysm of the ovarian artery. Varicosities of the ovarian vein, almost always on the right side, may occur in pregnant or parous women and cause ipsilateral ureteric compression and pyelonephritis, constituting the so-called ovarian vein syndrome. Ovarian arteriovenous fistulas have been reported as a rare complication of gynecologic surgery.

Ovarian Pregnancy Up to 1–3% of all ectopic pregnancies are ovarian (Hallatt 1982; Ito et al. 2003). The diagnosis of ovarian pregnancy should be restricted to cases in

760

Fig. 52 Ovarian pregnancy. A nodule of hemorrhagic placental tissue protrudes from the ovarian surface (lower right)

which there is no involvement of the fallopian tube. There is an increased frequency of ovarian pregnancy in patients with an intrauterine contraceptive device or who have had fertility treatments (Joseph and Irvine 2012). The typical clinical presentation is severe pain with hemoperitoneum, and at laparotomy and on gross pathologic examination, the enlarged hemorrhagic ovary may mimic a hemorrhagic neoplasm (Fig. 52). In a minority of the cases, gross identification of an embryo is indicative of the diagnosis, and in other cases, microscopic examination is diagnostic. One placental site nodule of the ovary has been described in a 61-year-old woman, presumably representing the remnants of an old ovarian pregnancy (Al-Hussaini et al. 2002). Distinction between an ovarian pregnancy and the very rare examples of primary ovarian gestational trophoblastic disease (▶ Chap. 20, “Gestational Trophoblastic Tumors and Related Tumorlike Lesions”) is made by applying criteria similar to those used in the uterus.

Ovarian Changes Secondary to Metabolic Diseases Amyloidosis may rarely involve the ovaries, usually as an incidental microscopic finding in patients with systemic amyloidosis (Fig. 53) (Copeland et al. 1985). There are several reports of tumor-like ovarian enlargement secondary to

J. A. Irving and P. B. Clement

Fig. 53 Ovarian amyloidosis. Amyloid involves the ovarian stroma and vessels

systemic or presumably localized amyloidosis (Mount et al. 2002). Rare cases of ovarian enlargement secondary to involvement by systemic storage disorders (lipidoses, mucopolysaccharidoses) have been reported (Dincsoy et al. 1984). In such cases, the stored material is typically within macrophages, allowing histologic distinction from a steroid cell tumor or foci of fat within the ovarian stroma. In contrast to frequent testicular involvement in hemochromatosis, in which hemosiderin is typically seen within walls of testicular blood vessels, pathologic changes in the ovary secondary to this disorder appear to be rare or nonexistent.

Ovarian Changes Secondary to Cytotoxic Drugs and Radiation Cytotoxic Drugs Cytotoxic drugs may cause a variety of histologic changes in the ovaries of prepubertal and postpubertal patients, including focal or diffuse cortical fibrosis, impaired follicular maturation, and a reduction or depletion in follicle numbers (Cohen 2008; Himelstein-Braw et al. 1977). Some studies have shown a direct correlation between the severity of these changes and the duration of the chemotherapy, the number of drugs, and malnourishment of the patients. These morphologic findings are consistent with

12

Nonneoplastic Lesions of the Ovary

clinical observations of diminished ovarian endocrine function or ovarian failure in some of these patients. The risk of ovarian failure appears to be greater in patients who received higher doses and in whom treatment is begun after the age of 25 (Cohen 2008; Letourneau et al. 2012. In rare cases, the ovarian failure has reversed after cessation of the therapy.

Radiation The ovary is among the most radiosensitive of organs. Relatively low doses of radiation (500–600 R) to the ovaries cause complete or nearly complete disappearance of primordial and developing follicles, fibrosis of the ovarian stroma, and vascular sclerosis in more than 90% of patients (Fig. 54) (Cohen 2008; HimelsteinBraw et al. 1977). Follow-up studies of both children and adults who received pelvic radiation have shown that ovarian failure occurs in the majority of such patients (Himelstein-Braw et al. 1977; Stillman et al. 1981). The ovarian stroma appears to be more radioresistant than the follicles and may continue to secrete androgens after radiation (Janson et al. 1981). Fertility-preservation technologies, an evolving field that includes ovarian transposition, may benefit select patients undergoing gonadotoxic therapy (Practice Committee of American Society for Reproductive Medicine 2013).

Fig. 54 Radiation changes. Blood vessels are hyalinized, and the ovarian stroma is fibrotic

761

Miscellaneous Lesions Ovarian Remnant Syndrome Ectopic, accessory, and supernumerary ovary (Congenital Lesions and Ectopic Tissues) should be distinguished from examples of the ORS, which is also known as the ovarian implant syndrome (Kaminski et al. 1990; Lafferty et al. 1996; Magtibay et al. 2005; Pettit and Lee 1988). The ORS should in turn be distinguished from the residual ovary syndrome, in which pelvic symptoms originate from ovaries preserved at the time of hysterectomy (Lafferty et al. 1996). Patients with ORS have a history of a presumably total bilateral oophorectomy but present with findings related to the presence of residual ovarian tissue. The oophorectomy in such cases was often complicated by dense adhesions that are usually caused by endometriosis, PID, a previous pelvic operation, inflammatory bowel disease, or combinations thereof. Clues to the diagnosis in patients who have had bilateral oophorectomy are premenopausal FSH, LH, and estradiol levels, an absence of menopausal symptoms, and a lack of atrophic changes on cervicovaginal smears. Within weeks to months, but occasionally years, after the oophorectomy, women with ORS usually present with chronic or cyclic pelvic pain and, in about half the cases, a palpable pelvic mass. Ultrasonography or CT scanning may aid preoperative detection of nonpalpable symptomatic remnants (Pettit and Lee 1988), and stimulation of the residual ovarian tissue with clomiphene citrate therapy can facilitate their intraoperative localization (Kaminski et al. 1990). Reoperation typically reveals a 3to 4-cm cystic mass, covered by dense adhesions on the pelvic side wall, or less commonly, the mesentery; bilateral ovarian remnants rarely have been encountered. Obstruction or compression of the ureter, the colon, the small intestine, or the bladder may occur (Pettit and Lee 1988). Pathologic examination usually reveals one or several follicular or CLCs within a remnant of ovarian tissue surrounded by chronically inflamed fibrous tissue (Fig. 55). Less common

762

Fig. 55 ORS. A CLC is surrounded by congested fibroadipose tissue

findings have included endometriosis, benign neoplasms, or normal ovarian tissue. Excision of the remnants may be difficult and require multiple operations.

Rete Cysts The rete ovarii, the ovarian analogue of the rete testis, which is present in the hilus of all ovaries, consists of a network of anastomosing branching tubules with intraluminal polypoid projections, lined by an epithelium that varies from flat to cuboidal to columnar. Solid cords of similar cells may also be seen. The rete is surrounded by a cuff of spindle cell stroma morphologically similar to, but discontinuous from, the ovarian stroma. The rete lies adjacent to and may communicate with mesonephric tubules within the mesovarium. The rete epithelium may undergo transitional metaplasia. Occasional hilar cysts originate from the rete, and small tumor-like proliferations of the rete have been referred to as rete adenomas. Rete cysts are typically located in the ovarian hilus, and in one series, the cysts had a mean diameter of 8.7 cm (range, 1–24 cm) (Rutgers and Scully 1988). Most are unilocular, although occasionally they are multilocular. Rete cysts typically are lined by a single layer of nonciliated epithelium that varies from flat to cuboidal to columnar. In

J. A. Irving and P. B. Clement

Fig. 56 Rete cyst. The lining is composed of a single layer of flattened cells which line crevices. Note smooth muscle and nests of Leydig cells in the wall

addition to their hilar location, clues to the origin of the cysts are an irregular contour of their inner surface with small crevice-like outpouchings and a wall that often contains bundles of smooth muscle and hyperplastic hilus cells (Fig. 56). A lesion interpreted as adenomatous hyperplasia of the rete ovarii has been reported, which formed a 3-mm nodule in the ovarian hilus of a 43-year-old woman. The lesion consisted of a poorly circumscribed proliferation of benign appearing tubules with scanty fibromuscular stroma, merging with the normal rete (Heatley 2000).

Artifacts and Normal Findings The granulosa cells of normal follicles can be artifactually introduced into tissue spaces or vascular channels during sectioning (Figs. 57 and 58). This finding, especially when the displaced cells are shrunken or crushed, is occasionally misinterpreted as small cell carcinoma (McCluggage and Young 2004). Awareness of this artifact, the bland nuclear features of the cells, and their similarity to cells lining nearby follicles are helpful clues to the correct diagnosis. Granulosa cells that appear to be deposited on the surface of the ovary secondary to follicle rupture may be misinterpreted as mesothelial cells and when numerous, may even suggest the possible diagnosis of a mesothelioma.

12

Nonneoplastic Lesions of the Ovary

763

mucicarminophilic histiocytosis and infarcted appendix epiploica.

References

Fig. 57 Artifactual displacement of granulosa cells. Nests of granulosa cells occupy an artifactual space in the ovarian stroma (bottom). Note adjacent cystic follicle (top) with detached granulosa cells similar to those in the stroma

Fig. 58 Displaced luteinized granulosa cells in ovarian lymphatics

Immunohistochemical staining for inhibin may confirm their presence in difficult cases. Normal findings that may be misinterpreted as neoplastic include the occasionally highly mitotic granulosa cells and the theca externa cells of the normal developing follicle. Similarly, the corpus luteum of late pregnancy and the puerperium may contain numerous calcific deposits; we have encountered one case in which their presence in a patient with a history of a serous borderline tumor was misinterpreted as recurrent tumor. Lesions that can involve the ovary, but which are more appropriately discussed in ▶ Chap. 13, “Diseases of the Peritoneum,” include

Abrao MS, Dias JA Jr, Podgaec S et al (2006) Bowel endometriosis and schistosomiasis: a rare but possible association. Fertil Steril 85:1060 Adler E, Mhawech-Fauceglia P, Gayther SA et al (2015) PAX8 expression in ovarian surface epithelial cells. Hum Pathol 46:948–956 Akhan SE, Dogan Y, Akhan S et al (2008) Pelvic actinomycosis mimicking ovarian malignancy: three cases. Eur J Gynaecol Oncol 29:294–297 Al-Ahmadie HA, Stanek J, Liu J et al (2001) Ovarian “tumor” of the adrenogenital syndrome. The first reported case. Am J Surg Pathol 25:1443–1450 Al-Hussaini M, Lioe TF, McCluggage WG (2002) Placental site nodule of the ovary. Lett Histopathol 41:471–472 Al-Timimi A, Buckley CH, Fox H (1985) An immunohistochemical study of the incidence and significance of sex steroid hormone binding sites in normal and neoplastic human ovarian tissue. Int J Gynecol Pathol 4:24–41 Almenoff IA (1966) Splenic-gonadal fusion. N Y State J Med 66:1679–1691 Alujevic A, Andelinovic S, Forempoher G (1995) Splenicgonadal fusion of the continuous type in an adult female. Pathol Int 45:871–874 Athanasias P, Doumouchtsis K, Malick R et al (2013) Isolated fallopian tube torsion: a rare variant of a common entity with successful laparoscopic detorsion. J Obstet Gynaecol 33:318 Auersperg N (2013) The stem-cell profile of ovarian surface epithelium is reproduced in the oviductal fimbriae, with increased stem-cell marker density in distal parts of the fimbriae. Int J Gynecol Pathol 32:444–453 Balen AH, Laven JS, Tan SL et al (2003) Ultrasound assessment of the polycystic ovary: international consensus definitions. Hum Reprod Update 9:505–514 Banet N, Kurman RJ (2015) Two types of ovarian cortical inclusion cysts: proposed origin and possible role in ovarian serous carcinogenesis. Int J Gynecol Pathol 34:3–8 Bannatyne P, Russell P, Shearman RP (1990) Autoimmune oophoritis: a clinicopathologic assessment of 12 cases. Int J Gynecol Pathol 9:191–207 Barbieri RL, Ryan KJ (1983) Hyperandrogenism, insulin resistance, and acanthosis nigricans syndrome: a common endocrinopathy with distinct pathophysiologic features. Am J Obstet Gynecol 147:90–101 Barbieri RL, Smith S, Ryan KJ (1988) The role of hyperinsulinemia in the pathogenesis of ovarian hyperandrogenism. Fertil Steril 50:197–212

764 Bats AS, Barbarino PM, Bene MC et al (2008) Local lymphocytic and epithelial activation in a case of autoimmune oophoritis. Fertil Steril 90:849.e5–849.e8 Bell DA, Mondschein M, Scully RE (1986) Giant cell arteritis of the female genital tract. A report of three cases. Am J Surg Pathol 10:696–701 Ben-Rafael Z, Bider D, Menashe Y et al (1990) Follicular and luteal cysts after treatment with gonadotropinreleasing hormone analogue for in vitro fertilization. Fertil Steril 53:1091–1094 Berger NG, Repke JT, Woodruff JD (1984) Markedly elevated serum testosterone in pregnancy without fetal virilization. Obstet Gynecol 63:260–262 Blaustein A (1981) Surface cells and inclusion cysts in fetal ovaries. Gynecol Oncol 12:222–233 Blaustein A, Kaganowicz A, Wells J (1982a) Tumor markers in inclusion cysts of the ovary. Cancer (Phila) 49:722–726 Blaustein A, Kantius M, Kaganowicz A et al (1982b) Inclusions in ovaries of females aged day 1–30 years. Int J Gynecol Pathol 1:145–153 Bonar BE, Rabson AS (1957) Gynecologic aspects of leprosy. Obstet Gynecol 9:33–43 Boss JH, Scully RE, Wegner KH et al (1965) Structural variations in the adult ovary: clinical significance. Obstet Gynecol 25:747–764 Bowen NJ, Walker LD, Matyunina LV et al (2009) Gene expression profiling supports the hypothesis that human ovarian surface epithelial are multipotent and capable of serving as ovarian cancer initiating cells. BMC Med Genom 2:71 Boyle DP, McCluggage WG (2009) Combined actinomycotic and pseudoactinomycotic radiate granules in the female genital tract: description of a series of cases. J Clin Pathol 62:1123–1126 Bradshaw KD, Santos-Ramos R, Rawlins SC et al (1986) Endocrine studies in a pregnancy complicated by ovarian theca lutein cysts and hyperreactio luteinalis. Obstet Gynecol 67:66S–69S Burandt E, Young RH (2014) Pregnancy luteoma: a study of 20 cases on the occasion of the 50th anniversary of its description by Dr. William H. Sternberg, with an emphasis on the common presence of follicle-like spaces and their diagnostic implications. Am J Surg Pathol 38:239–244 Burkman R, Schlesselman S, McCaffrey L et al (1982) The relationship of genital tract actinomycetes and the development of pelvic inflammatory disease. Am J Obstet Gynecol 143:585–589 Bylund DJ, Nanfro JJ, Marsh WL Jr (1986) Coccidioidomycosis of the female genital tract. Arch Pathol Lab Med 110:232–235 Caillouette JC, Koehler AL (1987) Phasic contraceptive pills and functional ovarian cysts. Am J Obstet Gynecol 156:1538–1542 Çakır ED, Özdemir Ö, Eren E et al (2011) Resolution of autoimmune oophoritis after thymectomy in a myasthenia gravis patient. J Clin Res Pediatr Endocrinol 3:212–215

J. A. Irving and P. B. Clement Carmina E, Gonzalez F, Chang L et al (1995) Reassessment of adrenal androgen secretion in women with polycystic ovary syndrome. Obstet Gynecol 85:971–976 Cavoretto P, Giorgione V, Sigismondi C et al (2014) Hyperreactio luteinalis: timely diagnosis minimizes the risk of oophorectomy and alerts clinicians to the associated risk of placental insufficiency. Eur J Obstet Gynecol Reprod Biol 176:10–16 Chen ZJ, Zhao H, He L (2011) Genome-wide association study identifies susceptibility loci for polycystic ovary syndrome on chromosome 2p16.3, 2p21 and 9q33.3. Nat Genet 43:55–59 Chou SC, Wang JS, Tseng HH (2002) Malacoplakia of the ovary, fallopian tube and uterus: a case associated with diabetes mellitus. Pathol Int 52:789–793 Clement PB (2007) Ovary. In: Mills SE (ed) Histology for pathologists. Lippincott, Williams, and Wilkins, Philadelphia, pp 1663–1694 Clement PB, Cooney TP (1992) Idiopathic multifocal calcification of the ovarian stroma. Arch Pathol Lab Med 116:204–205 Clement PB, Scully RE (1980) Large solitary luteinized follicle cyst of pregnancy and puerperium: a clinicopathological analysis of eight cases. Am J Surg Pathol 4:431–438 Clement PB, Young RH, Scully RE (1988) Ovarian granulosa cell proliferations of pregnancy. A report of nine cases. Hum Pathol 19:657–662 Clement PB, Young RH, Scully RE (1989) Nontrophoblastic pathology of the female genital tract and peritoneum associated with pregnancy. Semin Diagn Pathol 6:372–406 Cohen LE (2008) Cancer treatment and the ovary: the effects of chemotherapy and radiation. Ann N Y Acad Sci 1135:123–125 Cohen I, Figer A, Tepper R et al (1999) Ovarian overstimulation and cystic formation in premenopausal tamoxifen exposure: comparison between tamoxifentreated and nontreated breast cancer patients. Gynecol Oncol 72:202–207 Copeland W Jr, Hawley PC, Teteris NJ (1985) Gynecologic amyloidosis. Am J Obstet Gynecol 153:555–556 Coulam CB (1983) The prevalence of autoimmune disorders among patients with primary ovarian failure. Am J Reprod Immunol 4:63–66 Coulam CB, Annegers JF, Kranz JS (1983) Chronic anovulation syndrome and associated neoplasia. Obstet Gynecol 61:403–407 Cowchock FS, McCabe JL, Montgomery BB (1988) Pregnancy after corticosteroid administration in premature ovarian failure (polyglandular endocrinopathy syndrome). Am J Obstet Gynecol 158:118–119 Cramer DW, Welch WR, Cassells S et al (1983) Mumps, menarche, menopause, and ovarian cancer. Am J Obstet Gynecol 147:1–6 Damewood MD, Zacur HA, Hoffman GJ et al (1986) Circulating antiovarian antibodies in premature ovarian failure. Obstet Gynecol 68:850–854

12

Nonneoplastic Lesions of the Ovary

Damodaran S, Mahmood T, Nawroz I et al (2011) Ovarian stromal hyperplasia and hyperthecosis: an unusual case of post-menopausal hirsutism. Eur J Obstet Gynecol Reprod Biol 158:114–115 DastranjTabrizi A, MostafaGharabaghi P, SheikhzadehHesari F et al (2016) Impact and mechanistic role of oral contraceptive pills on the number and epithelial type of ovarian cortical inclusion cysts; a clinicopathology and immunohistochemical study. Diagn Pathol 22(11):30 Deligdisch L, Einstein AJ, Guera D et al (1995) Ovarian dysplasia in epithelial inclusion cysts. A morphometric approach using neural networks. Cancer (Phila) 76:1027–1034 Dennefors BL, Janson PO, Knutson F et al (1980) Steroid production and responsiveness to gonadotropin in isolated stromal tissue of human postmenopausal ovaries. Am J Obstet Gynecol 136:997–1002 Dessole S, Capobianco G, Arru A et al (2003) Postpartum ovarian vein thrombosis: an unpredictable event: two case reports and review of the literature. Arch Gynecol Obstet 267:242–246 Dincsoy HP, Rolfes DB, McGraw CA et al (1984) Cholesterol ester storage disease and mesenteric lipodystrophy. Am J Clin Pathol 81:263–269 Draper R, Walker EA, Bujalska IJ et al (2003) Mutations in the genes encoding 11beta-hydroxysteroid dehydrogenase type 1 and hexose-6-phosphate dehydrogenase interact to cause cortisone reductase deficiency. Nat Genet 34:434–439 Drapkin R, Crum CP, Hecht JL (2004) Expression of candidate tumor markers in ovarian carcinoma and benign ovary: evidence for a link between epithelial phenotype and neoplasia. Hum Pathol 35:1014–1021 Dumesic DA, Oberfield SE, Stener-Victorin E et al (2015) Scientific statement on the diagnostic criteria, epidemiology, pathophysiology, and molecular genetics of polycystic ovary syndrome. Endocr Rev 36:487–525 Dunaif A, Hoffman AR, Scully RE et al (1985) Clinical, biochemical, and ovarian morphologic features in women with acanthosis nigricans and masculinization. Obstet Gynecol 66:545–552 Edwards-Silva RN, Han CS, Hoang Y et al (2008) Spontaneous ovarian hyperstimulation in a naturally conceived pregnancy with uncontrolled hypothyroidism. Obstet Gynecol 111:498–501 Ehrmann DA (2005) Polycystic ovary syndrome. N Engl J Med 352:1223–1236 Elahi AA, Nigam A, Pujani M et al (2015) Xanthogranulomatous oophoritis mimicking malignancy: a diagnostic dilemma. BMJ Case Rep 25. https://doi.org/ 10.1136/bcr-2015-210642 Elmore RG, Li AJ (2007) Peritoneal tuberculosis mimicking advanced-stage epithelial ovarian cancer. Obstet Gynecol 110:1417–1419 Escobar ME, Cigorraga SB, Chiauzzi VA et al (1982) Development of gonadotropin resistant ovary syndrome in myasthenia gravis: suggestion of similar autoimmune mechanisms. Acta Endocrinol 99:431–436

765 Euthymiopoulou K, Aletras AJ, Ravazoula P et al (2007) Antiovarian antibodies in primary Sjogren’s syndrome. Rheumatol Int 27:1149–1155 Filho RB, Domingues L, Naves L (2007) Polycystic ovary syndrome and hyperprolactinemia are distinct entities. Gynecol Endocrinol 23:267–272 Francke M, Mihaescu A, Chaubert P (1998) Isolated necrotizing arteritis of the female genital tract: a clinicopathologic and immunohistochemical study of 11 cases. Int J Gynecol Pathol 17:193–200 Friedman CI, Neff J, Kim MH (1984) Immunologic parameters in premature follicular depletion: T and B lymphocytes, T-cell subpopulations, cutaneous reactivity, and serum immunoglobulin concentrations. Diagn Immunol 2:48–52 Ganesan R, Ferryman SR, Meier L, Rollason TP (2000) Vasculitis of the female genital tract and its clinicopathologic correlation: a study of 46 cases with follow-up. Int J Gynecol Pathol 19:258–265 Gao HW, Lin CK, Yu CP et al (2002) Oxidized cellulose (Surgicel™) granuloma mimicking a primary ovarian tumor. Int J Gynecol Pathol 21:422–423 Gilks CB, Clement PB (1987) Colo-ovarian fistula: a report of two cases. Obstet Gynecol 69:533–537 Ginsburg KA, Valdes C, Schnider G (1987) Spontaneous utero-ovarian vessel rupture during pregnancy: three case reports and a review of the literature. Obstet Gynecol 69:474–476 Gloor E, Hurlimann J (1984) Autoimmune oophoritis. Am J Clin Pathol 81:105–109 Granks S, Gilling-Smith C, Watson H, Willis D (1999) Insulin action in the normal and polycystic ovary. Endocrinol Metab Clin N Am 28:361–378 Guarino A, Di Benedetto L, Giovanale V et al (2017) Hyperandrogenism in a postmenopausal woman: a rare case of ectopic adrenal cortical gland. Gynecol Endocrinol 33:185–187 Gupta R, Verma S, Bansal K et al (2016) Mature teratomas in a supernumerary ovary in a child: report of a first case. J Pediatr Adolesc Gynecol 29:e5–e7. https://doi. org/10.1016/j.jpag.2015.07.005 Hahn-Pedersen J, Munkholm Larsen P (1984) Supernumerary ovary. Acta Obstet Gynecol Scand 63: 365–366 Hallatt JG (1982) Primary ovarian pregnancy: a report of twenty-five cases. Am J Obstet Gynecol 143:55–60 Hallatt JG, Steele CH Jr, Snyder M (1984) Ruptured corpus luteum with hemoperitoneum: a study of 173 surgical cases. Am J Obstet Gynecol 149:5–9 Haning RV Jr, Strawn EY, Nolten WE (1985) Pathophysiology of the ovarian hyperstimulation syndrome. Obstet Gynecol 66:220–224 Harrison BT, Berg RE, Mittal K (2014) Massive ovarian edema associated with a broad ligament leiomyoma: a case report and review. Int J Gynecol Pathol 33:418–412 Haynes MC, Lu BY, Winkel AF (2014) Ovarian vein thrombophlebitis related to large uterine myoma. Obstet Gynecol 123:450–453

766 He H, Luthringer DJ, Hui P et al (2008) Expression of CD56 and WT1 in ovarian stroma and ovarian stromal tumors. Am J Surg Pathol 32:884–890 Heatley MK (2000) Adenomatous hyperplasia of the rete ovarii. Letter Histopathol 36:383–384 Hibbard LT (1985) Adnexal torsion. Am J Obstet Gynecol 152:456–461 Hill JA, Welch WR, Faris HM et al (1990) Induction of class II major histocompatibility complex antigen expression in human granulosa cells by interferon gamma: a potential mechanism contributing to autoimmune ovarian failure. Am J Obstet Gynecol 162:534–540 Himelstein-Braw R, Peters H, Faber M (1977) Influence of irradiation and chemotherapy on the ovaries of children with abdominal tumours. Br J Cancer 36:269–275 Hofland M, Cosyns S, De Sutter P et al (2013) Leydig cell hyperplasia and Leydig cell tumour in postmenopausal women: report of two cases. Gynecol Endocrinol 29:213–215 Honoré LH, O’Hara KE (1979) Ovarian hilus cell heterotopia. Obstet Gynecol 53:461–464 Honoré LH, O’Hara KE (1980) Subcapsular adipocytic infiltration of the human ovary: a clinicopathological study of eight cases. Eur J Obstet Gynaecol Reprod Biol 10:13–20 Hosfield EM, Rabban JT, Chen L et al (2008) Squamous metaplasia of the ovarian surface epithelium and subsurface fibrosis: distinctive pathologic findings in the ovaries and fallopian tubes of patients of peritoneal dialysis. Int J Gynecol Pathol 27:465–474 Hughesdon PE (1976) The endometrial identity of benign stromatosis of the ovary and its relation to other forms of endometriosis. J Pathol 119:201–209 Hughesdon PE (1982) Morphology and morphogenesis of the Stein-Leventhal ovary and of so-called “hyperthecosis”. Obstet Gynecol Surv 37:59–77 Hyttel TE, Bak GS, Larsen SB et al (2015) Re-torsion of the ovaries. Acta Obstet Gynecol Scand 94:236–244 Isojärvi JIT, Laatikainen TJ, Pakarinen AJ et al (1993) Polycystic ovaries and hyperandrogenism in women taking valproate for epilepsy. N Engl J Med 329:1383–1388 Ito H, Ishihara A, Koita H et al (2003) Ovarian pregnancy: report of four cases and review of the literature. Pathol Int 53:806–809 Jacob S, Koc M (2015) Autoimmune oophoritis: a rarely encountered ovarian lesion. Indian J Pathol Microbiol 58:249–251 Janson PO, Jansson I, Skryten A et al (1981) Ovarian endocrine function in young women undergoing radiotherapy for carcinoma of the cervix. Gynecol Oncol 11:218–223 Joseph RJ, Irvine LM (2012) Ovarian ectopic pregnancy: aetiology, diagnosis, and challenges in surgical management. J Obstet Gynaecol 32:472–474 Judd HL, Scully RE, Atkins L et al (1970) Pure gonadal dysgenesis with progressive hirsutism. Demonstration

J. A. Irving and P. B. Clement of testosterone production by gonadal streaks. N Engl J Med 282:881–885 Judd HL, Scully RE, Herbst AL et al (1973) Familial hyperthecosis: comparison of endocrinologic and histologic findings with polycystic ovarian disease. Am J Obstet Gynecol 117:976–982 Kahn CR, Flier JS, Bar RS et al (1976) The syndromes of insulin resistance and acanthosis nigricans: insulinreceptor disorders in man. N Engl J Med 294:739–745 Kaiser UB (2003) The pathogenesis of the ovarian hyperstimulation syndrome. N Engl J Med 349:729–732 Kalu E, Panay N (2008) Spontaneous premature ovarian failure: management challenges. Gynecol Endocrinol 24:273–279 Kaminski PF, Sorosky JI, Mandell MJ et al (1990) Clomiphene citrate stimulation as an adjunct in locating ovarian tissue in ovarian remnant syndrome. Obstet Gynecol 76:924–926 Kaufman DW, Shapiro S, Rosenberg L et al (1980) Intrauterine contraceptive device use and pelvic inflammatory disease. Am J Obstet Gynecol 136:159–162 Kaufman FR, Kogut MD, Donnell GN et al (1981) Hypergonadotropic hypogonadism in female patients with galactosemia. N Engl J Med 304:994–998 Kim JJ, Kim D, Yim JY et al (2017) Polycystic ovary syndrome with hyperandrogenism as a risk factor for non-obese non-alcoholic fatty liver disease. Aliment Pharmacol Ther 45:1403–1412 Klempner LB, Giglio PG, Niebles A (1987) Malacoplakia of the ovary. Obstet Gynecol 69:537–540 Knochenhauer ES, Key TJ, Kahsar-Miller M et al (1998) Prevalence of the polycystic ovary syndrome in unselected black and white women in the Southeastern United States: a prospective study. J Clin Endocrinol Metab 83:3078–3082 Kostelnik FV, Fremount HN (1976) Mycotic tubo-ovarian abscess associated with the intrauterine device. Am J Obstet Gynecol 125:272–274 Krasević M, Haller H, Rupcić S et al (2004) Massive edema of the ovary: a report of two cases due to lymphatic permeation by metastatic carcinoma from the uterine cervix. Gynecol Oncol 93:564–567 Krauss CM, Turksoy N, Atkins L et al (1987) Familial premature ovarian failure due to an interstitial deletion of the long arm of the X chromosome. N Engl J Med 317:125–131 LaBarbera AR, Miller MM, Ober C et al (1988) Autoimmune etiology in premature ovarian failure. Am J Reprod Immunol Microbiol 16:115–122 Ladefoged C, Lorentzen M (1988) Xanthogranulomatous inflammation of the female genital tract. Histopathology (Oxf) 13:541–551 Lafferty HW, Angioli R, Rudolph J et al (1996) Ovarian remnant syndrome: experience at Jackson Memorial Hospital, University of Miami, 1985 through 1993. Am J Obstet Gynecol 174:641–645 Lai C, Hsueh S, Lin C et al (1992) Human papillomavirus in benign and malignant ovarian and endometrial tissues. Int J Gynecol Pathol 11:210–215

12

Nonneoplastic Lesions of the Ovary

Landers DV, Sweet RL (1985) Current trends in the diagnosis and treatment of tuboovarian abscess. Am J Obstet Gynecol 151:1098–1110 Latza U, Niedobitek G, Schwarting R et al (1990) Ber-EP4: new monoclonal antibody which distinguishes epithelia from mesothelia. J Clin Pathol 43:213–219 Lee AT, Zane LT (2007) Dermatologic manifestations of polycystic ovary syndrome. Am J Clin Dermatol 8:201–219 Letourneau JM, Ebbel EE, Katz PP et al (2012) Acute ovarian failure underestimates age-specific reproductive impairment for young women undergoing chemotherapy for cancer. Cancer 118:1933–1939 Liapi C, Evain-Brion D (1987) Diagnosis of ovarian follicular cysts from birth to puberty: a report of twenty cases. Acta Pediatr Scand 76:91–96 Lindsay AN, Voorhess ML, MacGillivray MH (1983) Multicystic ovaries in primary hypothyroidism. Obstet Gynecol 61:433 Lomme M, Kostadinov S, Zhang C (2011) Large solitary luteinized follicle cyst of pregnancy and puerperium: report of two cases. Diagn Pathol 6:3 Lonsdale RN, Roberts PF, Trowell JE (1991) Autoimmune oophoritis associated with polycystic ovaries. Histopathology (Oxf) 19:77–81 Lunde O, Hoel PS, Sandvik L (1988) Ovarian morphology in patients with polycystic ovaries and in an age-matched reference material. Gynecol Obstet Investig 25:192–201 Luzzatto R, Brucker N (1981) Benign inclusion cysts of the ovary associated with psammoma bodies in vaginal smears. Acta Cytol 25:282–284 Madeido G, Tieu TM, Aiman J (1985) Atypical ovarian hyperthecosis in a virilized postmenopausal woman. Am J Clin Pathol 83:101–107 Magtibay PM, Nyholm JL, Hernandez JL et al (2005) Ovarian remnant syndrome. Am J Obstet Gynecol 193:2062–2066 Mahmood K (1975) Granulomatous oophoritis due to Schistosoma mansoni. Am J Obstet Gynecol 123:919–920 Manolitsas TP, Lanham SA, Hitchcock A et al (1998) Synchronous ovarian and cervical squamous intraepithelial neoplasia: an analysis of HPV status. Gynecol Oncol 70:428–431 March WA, Moore VM, Willson KJ et al (2010) The prevalence of polycystic ovary syndrome in a community sample assessed under contrasting diagnostic criteria. Hum Reprod 25:544–551 Marrogi AJ, Gersell DJ, Kraus FT (1991) Localized asymptomatic giant cell arteritis of the female genital tract. Int J Gynecol Pathol 10:51–58 Massachusetts General Hospital Case Records (1982) Case 25–1982. Ovarian stromal hyperthecosis. Acanthosis nigricans. N Engl J Med 306:1537–1544 Massachusetts General Hospital Case Records (1986) Case 46–1986. Resistant-ovary syndrome, with hyalinization of preantral follicles. N Engl J Med 315:1336–1344

767 Massachusetts General Hospital Case Records (1988) Case 22–1988. Ovarian stromal hyperthecosis, with virilization, insulin resistance, and acanthosis nigricans. N Engl J Med 318:1449–1457 Mathur S, Melchers JT III, Ades EW et al (1980) Antiovarian and anti-lymphocyte antibodies in patients with chronic vaginal candidiasis. J Reprod Immunol 2:247–262 Matsubara Y, Fujioka T, Ikeda T et al (2009) Periodic size changes in a supernumerary ovary with associated corpus luteal cyst. J Obstet Gynaecol Res 35:180–182 Mattison DR, Evans MI, Schwimmer WB et al (1984) Familial premature ovarian failure. Am J Hum Genet 36:1341–1348 McCluggage WG, Allen DC (1997) Ovarian granulomas: a report of 32 cases. J Clin Pathol 50:324–327 McCluggage WG, Maxwell P (2001) Immunohistochemical staining for calretinin is useful in the diagnosis of ovarian sex cord-stromal tumours. Histopathology 38:403–408 McCluggage WG, Young RH (2004) Non-neoplastic granulosa cells within ovarian vascular channels: a rare potential diagnostic pitfall. J Clin Pathol 57:151–154 McCluggage WG, McKenna M, McBride HA (2007) CD56 is a sensitive and diagnostically useful immunohistochemical marker of ovarian sex cord-stromal tumors. Int J Gynecol Pathol 26:322–327 McClure N, Healy DL, Rogers PA (1994) Vascular endothelial growth factor as capillary permeability agent in ovarian hyperstimulation. Lancet 344:235–236 Meneses MF, Ostrowski ML (1989) Female splenicgonadal fusion of the discontinuous type. Hum Pathol 20:486–488 Mignot MH, Schoemaker J, Kleingeld M et al (1989) Premature ovarian failure. I: the association with autoimmunity. Eur J Obstet Gynecol Reprod Biol 30:59–66 Mittal KR, Zeleniuch-Jacquotte A, Cooper JL et al (1993) Contralateral ovary in unilateral ovarian carcinoma: a search for preneoplastic lesions. Int J Gynecol Pathol 12:59–63 Montz FJ, Schlaerth JB, Morrow CP (1988) The natural history of theca lutein cysts. Obstet Gynecol 72:247–251 Mount SL, Eltabbakh GH, Hardin NJ (2002) Beta-2 microglobulin amyloidosis presenting as bilateral ovarian masses. A case report and review of the literature. Am J Surg Pathol 26:130–133 Muller-Holzner E, Ruth NR, Abfalter E et al (1995) IUD-associated pelvic actinomycosis: a report of five cases. Int J Gynecol Pathol 14:70–74 Murray JJ, Clark CA, Lands RH et al (1985) Reactivation blastomycosis presenting as a tuboovarian abscess. Obstet Gynecol 64:828–830 Na K, Park SY, Kim HS (2017) Clinicopathological characteristics of primary ovarian adenomyoma: a singleinstitutional experience. Anticancer Res 37: 2565–2574

768 Nagamani M, Gomez LG, Garza J (1982) In vivo steroid studies in luteoma of pregnancy. Obstet Gynecol 59:105S–111S Nagamani M, Van Dinh T, Kelver ME (1986) Hyperinsulinemia in hyperthecosis of the ovaries. Am J Obstet Gynecol 154:384–389 Nagamani M, Hannigan EV, Van Dinh T et al (1988) Hyperinsulinemia and stromal luteinization of the ovaries in postmenopausal women with endometrial cancer. J Clin Endocrinol Metab 67:144–148 Nagamani H, Kaspar HG, Van Dinh T et al (1990) Hyperthecosis of the ovaries in a woman with a placental site trophoblastic tumor. Obstet Gynecol 76:931–935 Nagamani M, Stuart CA, Doherty MG (1992) Increased steroid production by the ovarian stromal tissue of postmenopausal women with endometrial cancer. J Clin Endocrinol Metab 74:172–176 Nagelberg SB, Rosen SW (1985) Clinical and laboratory investigation of a virilized woman with placental-site trophoblastic tumor. Obstet Gynecol 65:527–534 Nakano R, Shima K, Yamoto M et al (1989) Binding sites for gonadotropins in human postmenopausal ovaries. Obstet Gynecol 73:196–200 Nestler JE, Clore JN, Blackard WG (1989) The central role of obesity (hyperinsulinemia) in the pathogenesis of the polycystic ovary syndrome. Am J Obstet Gynecol 161:1095–1097 Nogales FF, Silverberg SG (1976) Epidermoid cysts of the ovary: a report of five cases with histogenetic considerations and ultrastructural findings. Am J Obstet Gynecol 124:523–528 Nogales FF, Martin-Sances L, Mendoza-Garcia E et al (1996) Massive ovarian edema. Histopathology (Oxf) 28:229–234 Nogales-Ortiz F, Taracon I, Nogales FF (1979) The pathology of female genital tract tuberculosis. Obstet Gynecol 53:422–428 Nomelini RS, Oliveira LJ, Jammal MP et al (2013) Serous papillary cystadenocarcinoma in supernumerary ovary. J Obstet Gynaecol 33:324 Norman RJ, Dewailly D, Legro RS et al (2007) Polycystic ovary syndrome. Lancet 370:685–697 Norris HJ, Taylor HB (1967) Nodular theca-lutein hyperplasia of pregnancy (so-called “pregnancy luteoma”). A clinical and pathologic study of 15 cases. Am J Clin Pathol 47:557–566 Ober WB, Grady HG, Schoenbucher AK (1957) Ectopic ovarian decidua without pregnancy. Am J Pathol 33:199–217 Osathanondh R, Berkowitz RS, de Cholnoky C et al (1986) Hormonal measurements in patients with theca lutein cysts and gestational trophoblastic disease. J Reprod Med 31:179–183 Pache TD, Chadha S, Goorens LJG et al (1991) Ovarian morphology in long-term androgen-treated female to male transsexuals A human model for the study of polycystic ovary syndrome? Histopathology (Oxf) 19:445–452

J. A. Irving and P. B. Clement Parveen AS, Elliott H, Howells R (2004) Sarcoidosis of the ovary. J Obstet Gynaecol 24:465 Peer E, Peretz BA, Makler A et al (1981) Bilateral adnexal agenesis with an ectopic ovary: case report and review of the literature. Eur J Obstet Gynecol Reprod Biol 12:37–42 Pekonen F, Seigberg R, Makinen T et al (1986) Immunological disturbances in patients with premature ovarian failure. Clin Endocrinol 25:1–6 Pelkey TJ, Frierson HF Jr, Mills SE et al (1998) The diagnostic utility of inhibin staining in ovarian neoplasms. Int J Gynecol Pathol 17:97–105 Pereira RM, de Carvalho JF, de Medeiros AC et al (2009) Ovarian necrotizing vasculitis in a patient with lupus. Lupus 18:1313–1315 Pettit PD, Lee RA (1988) Ovarian remnant syndrome: diagnostic dilemma and surgical challenge. Obstet Gynecol 71:580–583 Planner RS, Abell DA, Barbaro CA et al (1982) Massive enlargement of the ovaries after evacuation of hydatidiform moles. Aust NZ J Obstet Gynaecol 22:96–100 Plentl AA, Friedman EA (1971) Lymphatic system of the female genitalia. Saunders, Philadelphia Podgajski M, Kukura V, Duic Z et al (2007) Ascites, high CA-125 and chronic pelvic pain in an unusual clinical manifestation of Enterobius vermicularis ovarian and sigmoid colon granuloma. Eur J Gynaecol Oncol 28:513–515 Polson DW, Wadsworth J, Adams J et al (1988) Polycystic ovaries – a common finding in normal women. Lancet 1:870–872 Practice Committee of American Society for Reproductive Medicine. Fertility preservation in patients undergoing gonadotoxic therapy or gonadectomy (2013) Fertil Steril 100:1214–1223 Pritt B, Mount SL, Cooper K et al (2006) Pseudoactinomycotic radiate granules of the gynaecological tract: review of a diagnostic pitfall. J Clin Pathol 59:17–20 Pryse-Davies J (1974) The development, structure and function of the female pelvic organs in childhood. Clin Obstet Gynaecol 1:483–508 Putschar WGJ, Manion WC (1956) Splenic-gonadal fusion. Am J Pathol 32:15–33 Reato G, Morlin L, Chen S et al (2011) Premature ovarian failure in patients with autoimmune Addison’s disease: clinical, genetic, and immunological evaluation. J Clin Endocrinol Metab 96:E1255–E1261 Rebar RW, Erickson GF, Yen SSC (1982) Idiopathic premature ovarian failure: clinical and endocrine characteristics. Fertil Steril 37:35–41 Reeves G (1971) Specific stroma in the cortex and medulla of the ovary. Cell types and vascular supply in relation to follicular apparatus and ovulation. Obstet Gynecol 37:832–844 Resta L, Scordari MD, Colucci GA et al (1989) Morphological changes of the ovarian surface epithelium in

12

Nonneoplastic Lesions of the Ovary

ovarian polycystic disease or endometrial carcinoma and a control group. Eur J Gynaecol Oncol 10:39–41 Rodin A, Thakkar H, Taylor N et al (1994) Hyperandrogenism in polycystic ovary syndrome. Evidence of dysregulation of 11-beta-hydroxysteroid dehydrogenase. N Engl J Med 330:460–465 Rossetti R, Ferrari I, Bonomi M et al (2017) Genetics of primary ovarian insufficiency. Clin Genet 91:183–198 Roth LM, Sternberg WH (1973) Ovarian stromal tumors containing Leydig cells. II. Pure Leydig cell tumor, non-hilar type. Cancer (Phila) 32:952–960 Roth LM, Deaton RL, Sternberg WH (1979) Massive ovarian edema. A clinicopathologic study of five cases including ultrastructural observations and review of the literature. Am J Surg Pathol 3:11–21 Russell P, Farnsworth A (1997) Massive edema and fibromatosis. In: Russell P, Farnsworth A (eds) Surgical pathology of the ovaries. Churchill Livingstone, New York, pp 147–154 Russell P, Bannatyne P, Shearman RP et al (1982) Premature hypergonadotropic ovarian failure: clinicopathological study of 19 cases. Int J Gynecol Pathol 1:185–201 Rutgers JL, Scully RE (1986) Functioning ovarian tumors with peripheral steroid cell proliferation: a report of twenty-four cases. Int J Gynecol Pathol 5:319–337 Rutgers JL, Scully RE (1988) Cysts (cystadenomas) and tumors of the rete ovarii. Int J Gynecol Pathol 7:330–342 Sasano H, Fukunaga M, Rojas M et al (1989) Hyperthecosis of the ovary. Clinicopathologic study of 19 cases with immunohistochemical analysis of steroidogenic enzymes. Int J Gynecol Pathol 8:311–320 Schildkraut J, Schwingl PJ, Bastos E et al (1996) Epithelial ovarian cancer risk among women with polycystic ovary syndrome. Obstet Gynecol 88:554–559 Schindlbeck C, Dziura D, Mylonas I (2014) Diagnosis of pelvic inflammatory disease (PID): intra-operative findings and comparison of vaginal and intraabdominal cultures. Arch Gynecol Obstet 289:1263–1269 Schmidt WA (1982) IUDs, inflammation, and infection: assessment after two decades of IUD use. Hum Pathol 13:878–881 Scully RE (1964) Stromal luteoma of the ovary. Cancer (Phila) 17:769–778 Scully RE (1981) Smooth-muscle differentiation in genital tract disorders [Editorial]. Arch Pathol Lab Med 105:505–507 Scully RE (1986) Ovary. In: Henson DE, AlboresSaavedra J (eds) The pathology of incipient neoplasia. Saunders, Philadelphia, pp 279–293 Scully RE, Cohen RB (1964) Oxidative-enzyme activity in normal and pathologic human ovaries. Obstet Gynecol 24:667–681 Sedin G, Bergquist C, Lindgren PG (1985) Ovarian hyperstimulation syndrome in preterm infants. Pediatr Res 19:548–551

769 Sedmak DD, Hart WR, Tubbs RR (1987) Autoimmune oophoritis: a histopathologic study of involved ovaries with immunologic characterization of the mononuclear cell infiltrate. Int J Gynecol Pathol 6:73–81 Seidman JD, Krishnan J (2016) Non-neoplastic conditions of the ovaries in grossly normal adnexa: a clinicopathologic study of 403 completely embedded cases. Int J Gynecol Pathol 35:544–548 Shawker TH, Hubbard VS, Reichert CM et al (1983) Cystic ovaries in cystic fibrosis: an ultrasound and autopsy study. J Ultrasound Med 2:439–444 Sherman SL (2000) Permature ovarian failure in the fragile X syndrome. Am J Med Genet 97:189–194 Shimizu M, Toki T, Takagi Y et al (2000) Immunohistochemical detection of the Wilm’s tumor gene (WT-1) in epithelial ovarian tumors. Int J Gynecol Pathol 19:158–163 Shipton EA, Meares SD (1965) Heterotopic bone formation in the ovary. Aust N Z J Obstet Gynaecol 5:100–102 Simstein NL (1981) Colo-tubo-ovarian fistula as complication of pelvic inflammatory disease. South J Med 74:512–513 Smith CET, Toplis PJ, Nogales FF (1999) Ovarian prostatic tissue originating from hilar mesonephric rests. Am J Surg Pathol 23:232–236 Smits G, Olatunbosun O, Delbaere A (2003) Ovarian hyperstimulation syndrome due to a mutation in the follicle-stimulating hormone receptor. N Engl J Med 349:760–766 Snowden JA, Harkin PJR, Thornton JG et al (1989) Morphometric assessment of ovarian stromal proliferation: a clinicopathological study. Histopathology (Oxf) 14:369 Sternberg WH, Roth LM (1973) Ovarian stromal tumors containing Leydig cells. I. Stromal-Leydig cell tumor and non-neoplastic transformation of ovarian stroma to Leydig cells. Cancer (Phila) 32:940–951 Stillman RJ, Schinfeld JS, Schiff I et al (1981) Ovarian failure in long-term survivors of childhood malignancy. Am J Obstet Gynecol 139:62–66 Strickler RC, Kelly RW, Askin FB (1984) Postmenopausal ovarian follicle cyst: an unusual cause of estrogen excess. Int J Gynecol Pathol 3:318–322 Symonds DA, Driscoll SG (1973) An adrenal cortical rest within the fetal ovary: report of a case. Am J Clin Pathol 60:562–564 Tanabe K, Gagliano P, Channing CP et al (1983) Levels of inhibin-F activity and steroids in human follicular fluid from normal women and women with polycystic ovarian disease. J Clin Endocrinol Metab 57:24–31 Tatum ET, Beattie JF Jr, Bryson L (1996) Postoperative carbon pigment granuloma. A report of eight cases involving the ovary. Hum Pathol 27:1008–1011 Taylor AE (1998) Polycystic ovary syndrome. Endocrinol Metab Clin N Am 27:877–902 Taylor HC, Pillay I, Setrakian S (2000) Diffuse stromal Leydig cell hyperplasia: a unique cause of

770 postmenopausal hyperandrogenism and virilization. Mayo Clin Proc 75:288–292 The Rotterdam ESHRE/ASRM-Sponsored PCOS consensus workshop group (2004) Revised 2003 consensus on diagnostic criteria and long-term health risks related to polycystic ovary syndrome. Hum Reprod 19:41–47 Tressera F, Grases PJ, Labastida R et al (1998) Histological features of the contralateral ovary in patients with unilateral ovarian cancer: a case control study. Gynecol Oncol 71:437–441 van Karseren YM, Schoemaker J (1999) Premature ovarian failure: a systematic review on therapeutic interventions to restore ovarian function and achieve pregnancy. Hum Reprod Update 5:483–492 Villanueva AL, Rebar RW (1983) Triple-X syndrome and premature ovarian failure. Obstet Gynecol 62:70S–73S Vollaard ES, van Beek AP, Verburg FA et al (2011) Gonadotropin-releasing hormone agonist treatment in postmenopausal women with hyperandrogenism of ovarian origin. J Clin Endocrinol Metab 96:1197–1201 Walfish PG, Gottesman IS, Shewchuk AB et al (1983) Association of premature ovarian failure with HLA antigens. Tissue Antigens 21:168–169 Weinstein D, Rabinowitz R, Malach D et al (1983) Ovarian hemorrhage in women with Von Willebrand’s disease. A report of two cases. J Reprod Med 28:500–502 Wierman ME, Beardsworth DE, Mansfield MJ et al (1985) Puberty without gonadotropins. A unique mechanism of sexual development. N Engl J Med 312:65–72 Williams DJ, Connor P, Ironside JW (1990) Pre-menopausal cytomegalovirus oophoritis. Histopathology (Oxf) 16:405–407

J. A. Irving and P. B. Clement Witlin AG, Sibai BM (1995) Postpartum ovarian vein thrombosis after vaginal delivery: a report of 11 cases. Obstet Gynecol 85:775–780 Wojcik EM, Naylor B (1992) “Collagen balls” in peritoneal washings. Acta Cytol 36:466–470 Wood JR, Nelson VL, Ho C (2003) The molecular phenotype of polycystic ovary syndrome (PCOS) theca cells and new candidate PCOS genes defined by microarray analysis. J Biol Chem 278:26380–26390 Yildiz BO, Azziz R (2007) The adrenal and polycystic ovary syndrome. Rev Endocr Metab Disord 8: 331–342 Young RH, Scully RE (1984) Fibromatosis and massive edema of the ovary, possibly related entities: a report of 14 cases of fibromatosis and 11 cases of massive edema. Int J Gynecol Pathol 3:153–178 Young RH, Prat J, Scully RE (1980) Epidermoid cyst of the ovary. A report of three cases with comments on histogenesis. Am J Clin Pathol 73:272–276 Zergeroğlu S, Küçükah T, Koç Ö (2004) Primary ovarian echinococcosis. Arch Gynecol Obstet 270:285–286 Zhang J, Young RH, Arseneau J (1982) Ovarian stromal tumors containing lutein or Leydig cells (luteinized thecomas and stromal Leydig cell tumors): a clinicopathological analysis of fifty cases. Int J Gynecol Pathol 1:270–285 Zhang C, Sung CJ, Quddus MR et al (2017) Association of ovarian hyperthecosis with endometrial polyp, endometrial hyperplasia, and endometrioid adenocarcinoma in postmenopausal women: a clinicopathological study of 238 cases. Hum Pathol 59:120–124

Diseases of the Peritoneum

13

Julie A. Irving and Philip B. Clement

Contents Inflammatory Lesions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Acute Peritonitis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Granulomatous Peritonitis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Nongranulomatous Histiocytic Lesions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Peritoneal Fibrosis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Rare Types of Peritonitis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

772 772 773 775 776 778

Tumor-Like Lesions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Mesothelial Hyperplasia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Peritoneal Inclusion Cysts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Splenosis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Trophoblastic Implants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Peritoneal Keratin Granulomas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Infarcted Appendix Epiploica . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

778 778 780 782 782 782 783

Mesothelial Neoplasms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Adenomatoid Tumor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Well-Differentiated Papillary Mesothelioma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Malignant Mesothelioma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

783 783 783 784

J. A. Irving (*) Department of Laboratory Medicine, Pathology, and Medical Genetics, Royal Jubilee Hospital, Victoria, BC, Canada e-mail: [email protected] P. B. Clement Department of Pathology, Vancouver General Hospital, Vancouver, BC, Canada e-mail: [email protected] # Springer Nature Switzerland AG 2019 R. J. Kurman et al. (eds.), Blaustein’s Pathology of the Female Genital Tract, https://doi.org/10.1007/978-3-319-46334-6_13

771

772

J. A. Irving and P. B. Clement Miscellaneous Primary Tumors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Intra-abdominal Desmoplastic Small Round Cell Tumor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Solitary Fibrous Tumor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Inflammatory Myofibroblastic Tumor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Calcifying Fibrous Tumor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Omental–Mesenteric Myxoid Hamartoma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Sarcomas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Gestational Trophoblastic Disease . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Metastatic Tumors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

788 788 790 790 791 791 791 791 791

Lesions of the Secondary Müllerian System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Endometriosis in Usual Sites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Cervical and Vaginal Endometriosis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Tubal Endometriosis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Intestinal Endometriosis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Urinary Tract Endometriosis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Cutaneous Endometriosis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Inguinal Endometriosis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Endometriosis of the Lymph Nodes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Pleuropulmonary Endometriosis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Soft Tissue and Skeletal Endometriosis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Upper Abdominal Endometriosis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Endometriosis of the Nervous System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Endometriosis in Males . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Neoplasms Arising from Endometriosis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Peritoneal Endometrioid Lesions Other Than Endometriosis . . . . . . . . . . . . . . . . . . . . . . . . . . Peritoneal Serous Lesions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Endocervicosis (Including Müllerianosis) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Extraovarian Mucinous Tumors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Peritoneal Transitional, Squamous, Clear Cell, and Non-Epithelial Lesions . . . . . . . . . . Peritoneal Decidual Reaction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Diffuse Peritoneal Leiomyomatosis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Benign Intranodal Glands of Müllerian Type . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Intranodal Ectopic Decidua . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Intranodal Leiomyomatosis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

792 792 807 809 809 811 812 814 814 814 816 816 816 816 816 820 820 824 824 825 826 826 827 828 829

References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 829

This chapter considers the wide range of nonneoplastic and neoplastic lesions that involve the peritoneum, and in some cases the retroperitoneal lymph nodes, of females. The first half of the chapter covers inflammatory lesions, tumor-like lesions (including mesothelial hyperplasia), mesothelial neoplasms, miscellaneous primary tumors, and metastatic tumors. The final half of the chapter is devoted to a large group of lesions that exhibit müllerian differentiation on microscopic examination and share a potential origin from the secondary müllerian system, the prototypical example of which is endometriosis.

Inflammatory Lesions Acute Peritonitis Acute diffuse peritonitis, characterized by a serosal fibrinopurulent exudate, is most commonly associated with a perforated viscus and is usually bacterial or chemical (bile or gastric or pancreatic juice) in origin. The lipases in pancreatic juice typically produce fat necrosis. Spontaneous bacterial peritonitis occurs most often in children and in adults who are immunocompromised or have cirrhosis of the liver (Weinstein et al. 1978), with proton pump inhibitor use a potential risk factor in

13

Diseases of the Peritoneum

cirrhotic patients (Min et al. 2014). Rare infectious causes of acute peritonitis include Candida (Bayer et al. 1976), Actinomycetes, and amoebas (Kapoor et al. 1972). Recurrent attacks of acute peritonitis are an almost constant feature of familial Mediterranean fever (recurrent polyserositis) (Sohar et al. 1967). Localized acute peritonitis may be associated with infection (or infarction) of specific organs, as in pelvic inflammatory disease.

Granulomatous Peritonitis A variety of infectious and noninfectious agents can cause granulomatous peritonitis. The peritoneum may be studded with nodules, which can mimic disseminated tumor at operation. The diagnosis rests on the histologic, and, in some cases microbiologic, identification of the causative agent.

Infectious Tuberculous peritonitis, which is increasing in frequency, particularly among immunosuppressed patients, may be secondary to spread from a focus within the abdominopelvic cavity or be a manifestation of miliary spread (Koc et al. 2006). The granulomas are characterized by caseous necrosis and Langhans type giant cells; Mycobacteria may be demonstrated by acid-fast stains or immunofluorescence methods. Rarely, granulomatous peritonitis is a complication of fungal infections, including histoplasmosis, coccidioidomycosis, and cryptococcosis, and parasitic infestations, including schistosomiasis, oxyuriasis, echinococcosis, ascariasis, and strongyloidiasis. Noninfectious Foreign material, typically recognizable on histologic examination, can elicit a granulomatous reaction on the peritoneum. Starch granules from surgical gloves, douche fluid, and lubricants typically incite a granulomatous and fibrosing peritonitis; in occasional cases, the inflammatory reaction may be of tuberculoid type with caseous

773

necrosis (Nissim et al. 1981). The periodic acid–Schiff- (PAS) positive starch granules exhibit a characteristic Maltese cross configuration under polarized light. Talc was once an important cause of granulomatous and fibrosing peritonitis because of its use as a lubricant on surgical gloves, and talcinduced peritonitis has also been described in drug abusers. Other iatrogenic causes of granulomatous peritonitis include cellulose and cotton fibers from surgical pads and drapes, microcrystalline collagen hemostat (Avitene) (Park et al. 1981), and oily materials such as hysterosalpingographic contrast medium, which can be associated with a lipogranulomatous reaction. In one described case, a foreign body reaction to Surgicel™ resulted in a pelvic mass that mimicked recurrent ovarian cancer (Deger et al. 1995). Contamination of the peritoneal cavity by bowel contents, including vegetable matter, food-derived starch, and barium sulfate, can produce a peritoneal foreign body reaction. Sebaceous material and keratin from ruptured dermoid cysts typically evoke an intense granulomatous, lipogranulomatous, and fibrosing peritoneal inflammatory reaction that may mimic a neoplasm at operation (Fig. 1). We have also observed a case of florid granulomatous peritonitis that was detected 3 months after resection of a ruptured ovarian intestinal-type mucinous atypical proliferative/borderline tumor (Fig. 2).

Fig. 1 Lipogranulomatous peritonitis due to ruptured ovarian dermoid cyst

774

Fig. 2 Granulomatous peritonitis due to ruptured ovarian intestinal-type mucinous atypical proliferative/borderline tumor

Granulomatous inflammation to keratin derived from uterine and ovarian endometrioid carcinomas with squamous differentiation is discussed later in section “Tumor-Like Lesions.” Spillage of amniotic fluid at Cesarean section, with its content of vernix caseosa (keratin, squames, sebum, and lanugo hair) and meconium (bile, pancreatic, and intestinal secretions), produces a granulomatous peritonitis (George et al. 1995). Meconium peritonitis caused by bowel perforation in utero can also be a problem in newborn infants. In contrast to vernix caseosa peritonitis, calcification rather than granulomatous inflammation dominates the microscopic picture, which in some cases is associated with striking radiographic findings. In boys, the process may involve the tunica vaginalis and result in a tumor-like scrotal mass (Forouhar 1982). Rare cases of meconium peritonitis are associated with disseminated intravascular spread of the meconium. Granulomatous peritonitis has also been described secondary to Crohn’s disease, sarcoidosis, and Whipple’s disease. Necrotizing peritoneal granulomas have been described following diathermy ablation of endometriosis (Fig. 3) (Clarke and Simpson 1990). Necrotic pseudoxanthomatous nodules of endometriosis, can resemble necrotic granulomas, are described on page 805. Granulomatous inflammation can occur in the peritoneum secondary to spillage of bile or

J. A. Irving and P. B. Clement

Fig. 3 Peritoneal granuloma secondary to diathermy ablation for endometriosis. Histiocytes surround a necrotic center containing wisps of brown pigment

Fig. 4 Gallstone peritonitis. Gallstones are embedded in the pelvic peritoneum. Note the rim of foreign body giant cells and surrounding fibrosis. Laparoscopic cholecystectomy had been performed 18 months earlier

gallstones during laparoscopic cholecystectomy, with subsequent implantation of gallstones on peritoneal surfaces, including the ovaries (“ovarian cholelithiasis”; see also ▶ Chap. 12, “Nonneoplastic Lesions of the Ovary,” Fig. 12) (Vadlamudi et al. 1997). The embedded gallstones may cause abdominal pain, be associated with a foreign body granulomatous reaction and fibrosis, or act as a nidus for infection; the reddish-brown, pigmented lesions visible at laparoscopy can mimic endometriosis (Merchant et al. 2000). Cholesterol crystals and bile pigment may be identifiable within the foreign body giant cells (Fig. 4).

13

Diseases of the Peritoneum

Positive Fouchet’s staining for bile and awareness of the history of a previous cholecystectomy will facilitate the correct diagnosis.

Nongranulomatous Histiocytic Lesions The peritoneum can be occasionally involved by histiocytic infiltrates rather than discrete granulomas. Ceroid- and lipid-rich histiocytes involving the peritoneum and omentum can be secondary to endometriosis (Clement et al. 1988) or can occur in association with a peritoneal decidual reaction (White and Chan 1994). Peritoneal lesions consisting of pigment-laden histiocytes have been referred to as peritoneal melanosis. Reported cases of melanosis have usually been associated with ovarian dermoid cysts, sometimes with preoperative rupture (Jawarski et al. 2001); association with a serous cystadenoma has also been reported. At laparotomy, focal or diffuse, tan or black, and peritoneal staining or similarly pigmented tumor-like nodules are encountered within the pelvis and in the omentum. Some of the cysts within the ovarian tumors exhibit pigmentation of their contents and lining. On histologic examination, the ovarian and peritoneal pigmentation consists of pigment-laden histiocytes within a fibrous stroma (Fig. 5). In at least three of the reported cases and in a fourth case we have encountered, gastric mucosa was prominent within an otherwise typical dermoid cyst. No obvious source for the pigment could be identified

Fig. 5 Peritoneal melanosis. Histiocytes are laden with brown-black pigment

775

in most of the cases; Jaworski et al. (2001) demonstrated the pigment to have neither the histochemical features of melanin nor hemosiderin, but was rich in iron, and postulated that the pigment was derived from hemorrhage secondary to peptic ulceration of the gastric mucosa. These cases of benign peritoneal melanosis should obviously be distinguished from metastatic malignant melanoma, a distinction that is straightforward because of the bland nuclear features and absence of mitotic figures in the pigmented histiocytes. In one case, peritoneal melanosis was coexistent with malignant melanoma metastatic to the omentum (Lim et al. 2012). Nonpigmented histiocytes can occasionally occur as a nonspecific peritoneal inflammatory response in the form of nodular, plaque-like, or more extensive aggregates that may appear as small, grossly visible peritoneal nodules at operation, or more commonly as a microscopic finding. Histologically, the aggregates are composed of a monotonous population of histiocytes with moderate amounts of pale eosinophilic cytoplasm; the nuclei may be reniform and/or contain a groove, reminiscent of Langerhans-type histiocytes (Fig. 6). We are aware of one such case from a patient with a granulosa cell tumor in which the histiocytes were initially misinterpreted microscopically as metastatic granulosa cell tumor. Admixed mesothelial cells may also be present (“nodular mesothelial/histiocytic hyperplasia”) (Michal et al. 2016). Diffuse histiocytic proliferation of the pelvic peritoneum associated with endocervicosis has also been reported (Ruffolo and Suster 1993). In these cases, immunohistochemical staining for CD68, calretinin, and cytokeratin can aid the distinction of histiocytes from mesothelial cells (Fig. 6). Extensive histiocytic peritonitis may also rarely be associated with peritoneal malignancies, including serous carcinoma (Fig. 7) (Lv et al. 2012). Mucicarminophilic histiocytosis is characterized by histiocytes that contain polyvinylpyrrolidone (PVP), a substance that has been used as a blood substitute (Kuo and Hsueh 1984). These cells can be found in many sites, both within and outside the female reproductive organs, including the ovary, the pelvic lymph nodes, and the

776

J. A. Irving and P. B. Clement

Fig. 6 Nodule of histiocytes involving the peritoneal surface. (a) The cells have moderate, pale eosinophilic cytoplasm, some with reniform nuclei. (b) Positive immunostaining of histiocytes with CD68

Fig. 7 Florid histiocytic peritonitis associated with highgrade serous carcinoma in a primary debulking specimen (no prior chemotherapy)

Fig. 8 Mucicarminophilic histiocytosis. Note the multiple vacuolated histiocytes, some with a signet-ring cell appearance

Peritoneal Fibrosis omentum. The histiocytes have vacuolated basophilic to lavender cytoplasm and an eccentric nucleus, an appearance that may suggest the diagnosis of signet-ring cell adenocarcinoma (Fig. 8). The histiocytes are mucicarminophilic, but, in contrast to neoplastic signet-ring cells, are PAS negative; a variety of other stains are also helpful in the differential diagnosis (Kuo and Hsueh 1984). Peritoneal collections of mucicarmine-positive histiocytes have also been described in association with topical administration of oxidized regenerated cellulose, a hemostatic agent (Tang et al. 2009). The cytoplasm of these cells is PAS positive, diastase resistant, CD68 positive, and S-100 and cytokeratin negative.

Reactive peritoneal fibrosis, often accompanied by fibrous adhesions, is a common sequela of prior peritoneal inflammation and a frequent complication of a surgical procedure. The fibrosis can on occasion take the form of well-circumscribed fibrous nodules. Some reactive peritoneal fibrous lesions may contain spindle cells that are immunoreactive for vimentin, smooth muscle actin, and cytokeratin, referred to as multipotential subserosal cells (Bolen et al. 1986). It has also been postulated that under pathologic conditions, mesothelial cells undergo transition to myofibroblasts, resulting in fibrous peritoneal adhesions (Sandoval et al. 2016). Rarely, reactive fibrous proliferations of the peritoneum can form tumor-

13

Diseases of the Peritoneum

like nodules, in contrast to the more widespread peritoneal thickening of sclerosing peritonitis. In one case of an ovarian mucinous cystadenocarcinoma, several nodules composed of moderately cellular fascicles of benign-appearing spindle cells resembling fibroblasts and myofibroblasts that contained occasional mitotic figures were found in the cul-de-sac and serosal aspect of the tumor. Some of the spindle cells had the immunoprofile of the multipotential subserosal cells noted earlier. We refer to these lesions as peritoneal fibrous nodules (Clement 1995). Localized hyaline plaques are a common incidental finding on the splenic capsule and are probably related to splenic congestion (Wanless and Bernier 1983). Nonspecific fibrous thickening of the peritoneum may be seen in patients with hepatic cirrhosis and ascites. The designation sclerosing peritonitis has been applied to a clinically significant, potentially fatal lesion that represents a reactive hyperplasia of the submesothelial mesenchymal cells to a variety of stimuli. The first description, by Concato, was that of pearly white thickening of the visceral peritoneum, either as discrete plaques or continuous sheets involving the hepatic, splenic, and diaphragmatic peritoneum. The process often encases the small bowel (“abdominal cocoon”), causing bowel obstruction. Sclerosing peritonitis occurs in an idiopathic form, which most frequently, but not invariably, affects adolescent girls in tropical countries (Foo et al. 1978). Known causes include practolol therapy, chronic ambulatory peritoneal dialysis, the use of a peritoneovenous (LeVeen) shunt, bacterial or mycobacterial infection, sarcoidosis, the carcinoid syndrome, familial Mediterranean fever, and fibrogenic foreign materials as seen in drug users. Additionally, sclerosing peritonitis has an enigmatic association with luteinized thecomas of the ovary (Fig. 9; see ▶ Chap. 15, “Sex CordStromal, Steroid Cell, and Other Ovarian Tumors with Endocrine, Paraendocrine, and Paraneoplastic Manifestations”) (Clement et al. 1993; Staats et al. 2008). Some patients with sclerosing peritonitis have been successfully treated utilizing antiestrogens and/or GnRH agonists. Sclerosing peritonitis should be distinguished from the rarer

777

Fig. 9 Sclerosing peritonitis associated with bilateral luteinized thecomas of the ovary. Omentum shows surface involvement by cellular, reactive fibrous tissue

“peritoneal encapsulation,” a congenital malformation in which an accessory peritoneal membrane encases loops of small bowel in a saclike structure. The latter condition is largely asymptomatic and is usually found incidentally at laparotomy or autopsy. Confusion arises when the two terms are used interchangeably or even together, as in “encapsulating peritonitis.” Reactive nodular fibrous pseudotumor is a term that has been applied to single or multiple lesions ranging up to 6 cm involving the gastrointestinal tract or mesentery in adults; some have been associated with bowel wall infiltration, but all have had a benign clinical course (Yantiss et al. 2003). Microscopically, the lesions are composed of a low to moderate cellular proliferation of fibroblasts, collagen, and mononuclear inflammatory cells that are usually sparse. The fibroblastic cells show variable immunoreactivity for vimentin, CD117, muscle-specific actin, smooth muscle actin, and desmin, with negative staining for CD34 and ALK-1. Sclerosing mesenteritis (mesenteric panniculitis, mesenteric lipodystrophy) usually occurs as a localized mass in the small bowel mesentery and is characterized by variable fibrosis, inflammation, and fat necrosis (Emory et al. 1997). Sclerosing mesenteritis may also rarely develop in IgG4-related disease, characterized by numerous IgG4-positive plasma cells and obliterative phlebitis (Minato et al. 2012).

778

In occasional cases, it may be difficult to differentiate between markedly reactive peritoneal fibrosis and a desmoplastic mesothelioma lacking frankly sarcomatoid areas, particularly in a small biopsy specimen. These tumors, however, are very rare in the peritoneal cavity, especially in women. Features favoring a diagnosis of mesothelioma include nuclear atypia, necrosis, organized patterns of collagen deposition (fascicular, storiform), and infiltration of adjacent tissues (Mangano et al. 1998).

Rare Types of Peritonitis Eosinophilic peritonitis is seen rarely in cases of eosinophilic gastroenteritis and the hypereosinophilic syndrome (Adams and Mainz 1977). Isolated cases of eosinophilic ascites have been associated with childhood atopy, peritoneal dialysis, vasculitis, lymphoma or metastatic carcinoma, and ruptured hydatid cysts (Adams and Mainz 1977). Rare cases of peritonitis may be secondary to peritoneal involvement by collagen vascular diseases, including systemic lupus erythematosus and Degos disease.

Tumor-Like Lesions Mesothelial Hyperplasia Hyperplasia of mesothelial cells is a common response to inflammation (including pelvic inflammatory disease) and chronic effusions (Figs. 10 and 11). Hyperplastic lesions may be noted at operation as solitary or multiple small nodules, but more commonly are incidental findings on microscopic examination (Churg et al. 2006). Mesothelial hyperplasia often involves the adnexal areas in cases of chronic salpingitis and endometriosis and is occasionally encountered, particularly in the omentum, in association with ovarian tumors (Clement and Young 1993). Mesothelial hyperplasia can also occur within the superficial ovarian stroma overlying an atypical proliferative/borderline epithelial tumor and in such cases can be misinterpreted as invasive

J. A. Irving and P. B. Clement

tumor (Fig. 12) (Clement and Young 1993). Mesothelial hyperplasia may be confined to a hernia sac and in such cases may be caused by trauma or incarceration (Rosai and Dehner 1975). Hyperplastic mesothelial cells occasionally are an incidental microscopic finding within the pelvic and intra-abdominal lymph nodes and in such cases are usually associated with mesothelial hyperplasia of the peritoneum (Fig. 13) (Clement et al. 1996a). The mesothelial cells may be misinterpreted as metastatic tumor, particularly in a woman with a known primary pelvic tumor. The appearance of the cells on routine stains suggests the correct diagnosis and can be confirmed by histochemical and immunohistochemical staining (see following). In florid examples, solid, trabecular, tubular, papillary, or tubulopapillary patterns (Figs. 10 and 11) and limited degrees of extension of the mesothelial cells into the underlying reactive fibrous tissues or the walls of ovarian tumors, endometriotic cysts, and peritoneal inclusion cysts (see below) may be seen (“mural mesothelial proliferation”). Incorporation into the ovarian tissue and true lymphovascular space involvement have also been described (Oparka et al. 2011). The cells are often focally disposed in linear, sometimes parallel, thin layers, separated by fibrin or fibrous tissue (Fig. 12). The mesothelial cells may have cytoplasmic vacuoles containing acid mucin (predominantly hyaluronic acid) or, less commonly, exhibit marked cytoplasmic clearing.

Fig. 10 Mesothelial hyperplasia with a papillary pattern

13

Diseases of the Peritoneum

779

Fig. 11 Mesothelial hyperplasia. (a) Nodular pattern. (b) Tubular pattern

Fig. 12 Mesothelial hyperplasia within the superficial ovarian stroma. An underlying atypical proliferative/borderline epithelial tumor was present

Fig. 13 Hyperplastic mesothelial cells within the subcapsular sinus of a pelvic lymph node

Superimposed deciduoid change (“deciduoid mesothelial hyperplasia”), especially in nodal mesothelial cells, can be a pitfall in the distinction from metastatic tumor (Stewart 2013). Mild to

moderate nuclear pleomorphism, mitotic figures, and occasional multinucleated cells may be seen. Psammoma bodies are encountered in occasional cases, and, rarely, eosinophilic strapshaped cells resembling rhabdomyoblasts have been described. Variable numbers of admixed histiocytes may also be present (nodular mesothelial/histiocytic hyperplasia) (Michal et al. 2016). The major differential diagnosis is with peritoneal malignant mesothelioma (PMM). The presence of grossly visible nodules, necrosis, conspicuous large cytoplasmic vacuoles, marked nuclear pleomorphism, and deep infiltration favor PMM over mesothelial hyperplasia (Churg et al. 2006). Some of these features, however, such as marked nuclear atypia, are not always present or may be present only focally within a PMM. Immunostains may facilitate the differential diagnosis. Immunoreactivity for p53, epithelial membrane antigen (EMA), insulin-like growth factor 2 messenger RNA-binding protein (IMP)-3, glucose transporter (GLUT)-1, XIAP, and high expression of EZH2 are characteristics of the cells of PMMs but not hyperplastic mesothelial cells; reactive mesothelial cells, in contrast to PMMs, tend to be desmin-positive (Attanoos et al. 2003; Chang et al. 2014; Shen et al. 2009; Shi et al. 2011; Shinozaki-Ushiku et al. 2017). Loss of BAP1 immunoreactivity, particularly in combination with homozygous deletion of 9p21 and p16 by fluorescent in situ hybridization, is highly characteristic of malignant mesothelioma,

780

whereas these alterations have not been observed in reactive mesothelial proliferations (Churg et al. 2016; Cigognetti et al. 2015; Ito et al. 2015; Kawai et al. 2016; Sheffield et al. 2015; Shinozaki-Ushiku et al. 2017). Despite these differential features, in occasional cases the distinction between a hyperplastic and malignant mesothelial lesion may be difficult or impossible, particularly in a biopsy specimen. If the lesion in question is a PMM, follow-up usually reveals its nature within several months because of its typical rapid growth. In contrast, an atypical mesothelial proliferation occasionally persists for years without an apparent cause. An apparently benign, otherwise typical mesothelial proliferation, however, occasionally precedes the appearance of a PMM (Churg et al. 2006). Some cases of “atypical mesothelial hyperplasia” evolving into PMM, however, likely represent PMM ab initio (Padmanabhan et al. 2003). The differential diagnosis of mesothelial hyperplasia also includes atypical proliferative/ borderline serous tumors of primary peritoneal or ovarian origin. Grossly visible ovarian or peritoneal tumor, columnar cells with or without cilia, the presence of intracellular or extracellular neutral mucin, and numerous psammoma bodies all favor a serous tumor. Immunohistochemical markers for epithelial differentiation (see section on Malignant Mesothelioma) may also be of value in the differential diagnosis.

J. A. Irving and P. B. Clement

contents that vary from yellow and watery to gelatinous. Although most of these unilocular mesothelial cysts are probably reactive in origin, some of those located in the mesocolon, mesentery of the small intestine, retroperitoneum, and splenic capsule may be developmental (Ross et al. 1989). Multilocular peritoneal inclusion cysts (MPICs) may form large bulky masses (Fig. 14); these lesions have also been referred to as benign cystic mesotheliomas, inflammatory cysts of the peritoneum, or postoperative peritoneal cysts. MPICs are usually associated with clinical manifestations, most commonly lower abdominal pain, a palpable mass, or both. They are usually adherent to the pelvic organs and may simulate a cystic ovarian tumor on clinical examination, at laparotomy (McFadden and Clement 1986), or even on pathologic examination; the upper abdominal cavity, the retroperitoneum, or the hernia sacs may also be involved (Ross et al. 1989). One cutaneous MPIC, involving the umbilical skin (without an associated hernia sac), has been reported (Konstantiniva et al. 2013). Unlike the smaller unilocular cysts, the septa and walls of MPICs may contain considerable amounts of fibrous tissue. Their contents may resemble those of the unilocular cysts or be serosanguineous or bloody. On microscopic examination, MPICs are typically lined by a single layer of flat to cuboidal,

Peritoneal Inclusion Cysts Peritoneal inclusion cysts typically occur in the peritoneal cavity of women in the reproductive age group, but may develop over a wide age range (McFadden and Clement 1986; Ross et al. 1989; Veldhuis et al. 2013). Rarely, they occur in males and in the pleural cavity. Unilocular peritoneal inclusion cysts are usually incidental findings at laparotomy in the form of single or multiple, small, thin-walled, translucent, unilocular cysts that may be attached or lie free in the peritoneal cavity. Occasionally, they may involve the round ligament simulating an inguinal hernia (Harper et al. 1986). The cysts have a smooth lining and

Fig. 14 Peritoneal inclusion cyst. Multilocular cystic mass consists of thin-walled cysts with a smooth lining

13

Diseases of the Peritoneum

Fig. 15 Multilocular peritoneal inclusion cysts. Cystic spaces are lined by a single layer of flat mesothelial cells and are separated by thin fibrous septa

Fig. 16 Multilocular peritoneal inclusion cysts. Cystic spaces are lined by cells with mild reactive nuclear atypia

occasionally hobnail-shaped, mesothelial cells with generally bland nuclear features, although a degree of reactive atypia is not infrequent (Figs. 15 and 16). The lining cells occasionally form small papillae and cribriform patterns or undergo squamous metaplasia. In some cases, mural proliferations of typical or atypical mesothelial cells arranged singly, as gland-like structures or nests (Figs. 17 and 18) (McFadden and Clement 1986), or in patterns resembling those in adenomatoid tumors may be encountered. Occasional vacuolated mesothelial cells in the stroma may simulate signet-ring cells (Ross et al. 1989). The septa typically consist of a loose, fibrovascular connective tissue with a sparse inflammatory infiltrate. In some cases, marked acute and chronic

781

Fig. 17 Multilocular peritoneal inclusion cysts with mural mesothelial proliferation. Cord-like arrangements within a reactive fibrous stroma create an infiltrative pattern

Fig. 18 Multilocular peritoneal inclusion cysts with mural mesothelial proliferation. High-power view showing benign-appearing mesothelial cells forming small nests and lining small tubules

inflammation, abundant fibrin, broad bands of granulation and fibrous tissue, and evidence of recent and remote hemorrhage are present in the cyst walls. The mesothelial cells are typically immunoreactive for calretinin, and some cases are positive for estrogen receptors (ER), progesterone receptors (PR), or both (Sawh et al. 2003). A history of a prior abdominal operation (most common), pelvic inflammatory disease, endometriosis, inflammatory bowel disease, radiotherapy, or abdominal trauma, or combinations thereof, was present in 70% and 84% of patients in two series (Ross et al. 1989; Veldhuis et al. 2013), suggesting a role for inflammation in the pathogenesis of the cysts. An inflammatory

782

pathogenesis is also supported by the occurrence of cases in which the dividing line between florid adhesions associated with inflammation and a MPIC may be difficult. With rare exceptions, there has been no association with asbestos exposure. Follow-up examinations have not disclosed malignant behavior in cases that we consider MPICs, but in as many as one half of these, the lesions have recurred from months to many years postoperatively (Ross et al. 1989). It is likely, however, that at least some of these “recurrences” are the result of newly formed postoperative adhesions. Some patients have responded favorably to treatment with GnRH agonists, tamoxifen, or oral contraceptives (Sawh et al. 2003; Yokoyama et al. 2014). For these reasons, we prefer the designation MPIC to benign cystic mesothelioma for such lesions, until there is convincing evidence for their neoplastic nature. MPICs are confused most often with multilocular cystic lymphangiomas. In contrast to MPICs, the latter typically occur in children, more frequently in boys. In addition, they are usually extrapelvic, being almost always localized to the mesentery of the small intestine, omentum, mesocolon, or retroperitoneum. Their contents may be chylous, and on histologic examination lymphoid aggregates and smooth muscle, which are rare findings in MPICs, are typically present within their walls. In problematic cases, immunohistochemical stains are useful in distinguishing endothelial from mesothelial cells. Another lesion that merits consideration in the differential diagnosis of MPICs is the rare multicystic adenomatoid tumor. In contrast to MPICs, the latter typically involve the myometrium, contain foci of typical adenomatoid tumor, and lack prominent numbers of inflammatory cells. A detailed discussion of other lesions in the differential diagnosis of MPICs has been presented elsewhere (Ross et al. 1989).

Splenosis Splenosis, which results from the implantation of splenic tissue, is typically an incidental finding at laparotomy or autopsy months to years after

J. A. Irving and P. B. Clement

splenectomy for traumatic splenic rupture (Carr and Turk 1992). A few to innumerable, red-blue, peritoneal nodules, ranging from punctate to 7 cm in diameter, are scattered widely throughout the abdominal cavity and, less commonly, over the pelvic cavity. The intraoperative appearance may mimic endometriosis, benign or malignant vascular tumors, or metastatic cancer.

Trophoblastic Implants Implants of trophoblast on the pelvic or omental peritoneum may complicate the operative treatment of tubal pregnancy (Bucella et al. 2009; Doss et al. 1998). The implants are more likely to occur in cases managed by laparoscopy (1.9% of cases) than those managed by laparotomy (0.6% of cases) and are more likely to occur after salpingotomy than salpingectomy. The clinical presentation in such cases includes an initial decline in the serum human chorionic gonadotropin (hCG) level after removal of the ectopic pregnancy, followed by a rising level, abdominal pain, and in some cases intra-abdominal hemorrhage. Microscopic examination of the implants reveals viable trophoblastic tissue that may include chorionic villi. Some lesions resemble a placental site nodule or plaque (Fig. 19).

Peritoneal Keratin Granulomas Peritoneal granulomas that form in response to implants of keratin derived from neoplasms of the female reproductive tract may be confused with metastatic tumor (Kim and Scully 1990). The tumors are most commonly endometrioid carcinomas with squamous differentiation originating in the endometrium or ovary, or, rarely, squamous cell carcinomas of the cervix or atypical polypoid adenomyomas of the uterus. The granulomas consist of laminated deposits of keratin, sometimes with ghost squamous cells, surrounded by foreign body giant cells and fibrous tissue (see Fig. 20; see also ▶ Chap. 12, “Nonneoplastic Lesions of the Ovary,” Figs. 9 and 10). Follow-up data on these patients suggest that the granulomas have no

13

Diseases of the Peritoneum

783

Fig. 19 Placental site plaque of the peritoneum Fig. 21 Infarcted appendix epiploica

calcified zone in which infarcted adipose tissue is usually recognizable (Fig. 21).

Mesothelial Neoplasms Adenomatoid Tumor

Fig. 20 Peritoneal keratin granuloma

prognostic significance, although they should be thoroughly sampled by the gynecologist and carefully examined microscopically to exclude the presence of viable tumor. The differential diagnosis includes peritoneal granulomas in response to keratin derived from other sources, as discussed earlier in this chapter.

Infarcted Appendix Epiploica Appendices epiploicae may undergo torsion and infarction (Vuong et al. 1990). Subsequent calcification can result in a hard tumor-like mass that may be found attached or loose in the peritoneal cavity. In the late stages, these structures are typically composed of layers of hyalinized connective tissue surrounding a central necrotic and

This benign tumor of mesothelial origin, adenomatoid tumor, rarely arises from extragenital peritoneum, such as the omentum or mesentery, but is much more commonly encountered within the fallopian tube and myometrium (see ▶ Chaps. 10, “Mesenchymal Tumors of the Uterus,” and ▶ 11, “Diseases of the Fallopian Tube and Paratubal Region”) and, in the male, the epididymis.

Well-Differentiated Papillary Mesothelioma Well-differentiated papillary mesotheliomas (WDPMs) of the peritoneum are uncommon lesions (Chen et al. 2013; Churg et al. 2014; Daya and McCaughey 1990; Goldblum and Hart 1995; Malpica et al. 2012). Eighty percent of the cases have occurred in women, who are usually of reproductive age; occasional patients are postmenopausal. WDPMs are usually an incidental finding at operation, but rare cases have been associated with abdominal pain or ascites.

784

Occasional patients, including two who were sisters, have had possible exposure to asbestos, but the association may be incidental (Daya and McCaughey 1990). At laparotomy and on gross examination, WDPMs may be solitary but are usually multiple and appear as gray to white, firm, papillary, or nodular lesions measuring less than 2 cm in diameter. The omental and pelvic peritoneum are typically involved; several examples have also been encountered on the gastric, intestinal, or mesenteric peritoneum. Microscopic examination reveals fibrous papillae covered by a single layer of flattened to cuboidal mesothelial cells (Fig. 22) with occasional basal vacuoles; the nuclear features are bland, and mitotic figures are rare or absent. Uncommon patterns include tubulopapillary, adenomatoid-like, branching cords, or solid sheets. The stroma of some tumors may be extensively fibrotic. Multinucleated stromal giant cells and psammoma bodies are encountered in occasional cases. When multiple lesions are present, they should each be sampled histologically as lesions with the appearance of a WDPM may rarely be associated with others that have the appearance of malignant mesothelioma and progressive disease (Goldblum and Hart 1995). The diagnosis of WDPM should be strictly reserved for tumors with bland nuclear features and no evidence of invasion. With the exception of one case that appeared to evolve into a diffuse malignant mesothelioma,

Fig. 22 Well-differentiated papillary mesothelioma. Fibrous papillae are lined by a single layer of uniform, flat to cuboidal, mesothelial cells

J. A. Irving and P. B. Clement

follow-up studies suggest that most WDPMs are associated with benign or indolent behavior. Recurrences are rare; in one study, only 1 of 26 patients developed a recurrence of WDPM, after an interval of 46.5 months from the original diagnosis (Malpica et al. 2012). Occasional examples, however, have persisted for as many as 29 years (Daya and McCaughey 1990). Several patients with WDPM have died, although the adjuvant therapy used in such cases possibly was a contributory factor (Daya and McCaughey 1990). In a recent study of 20 WDPMs with invasive foci, most occurred in the female peritoneum and were often multifocal (Churg et al. 2014). Five cases tested for p16 deletion were negative, but two thirds had abnormal karyotypes. Recurrences developed in eight patients (40%), including one patient who died of disseminated disease (but without histologic confirmation of the recurrent tumor). Thus, WDPMs, when multifocal or with invasive foci, warrant clinical follow-up.

Malignant Mesothelioma Clinical Features Peritoneal malignant mesotheliomas (PMMs) are much less common than similar tumors in the pleural cavity and account for only 10–20% of all mesotheliomas (Baker et al. 2005; Goldblum and Hart 1995; Kerrigan et al. 2002). These tumors are particularly rare in women, in whom most malignant papillary neoplasms of the peritoneum are extraovarian papillary serous carcinomas (see section “Lesions of the Secondary Müllerian System”). Historically, most PMMs occurred in middleaged to elderly males, but a recent study found an equal male to female ratio (Liu et al. 2014); occasional PMMs occur in young adults or children. The patients typically present with nonspecific manifestations, including abdominal discomfort and distension, digestive disturbances, and weight loss. Ascites is present in the majority of cases, and cytologic examination of the ascitic fluid may be diagnostic of PMM in some cases. The diagnosis, however, usually requires laparotomy or laparoscopy and biopsy. PMMs may rarely

13

Diseases of the Peritoneum

present within a hernia or hydrocele sac; as a retroperitoneal, umbilical, intestinal, or pelvic tumor; or as cervical or inguinal lymphadenopathy (Sussman and Rosai 1990). Rarely there is prominent ovarian involvement, the intraoperative appearance mimicking that of a primary ovarian tumor with peritoneal spread (Clement et al. 1996b). More than 80% of the patients in one large series had a history of asbestos exposure, but most of them were identified because of an occupational exposure to asbestos. In contrast, two series of PMMs in women found no association with a history of asbestos exposure (Goldblum and Hart 1995; Kerrigan et al. 2002). Asbestos fibers, however, have been identified with special techniques in some of these women (Heller et al. 1999). Aside from asbestos, radiation, chronic inflammation, organic chemicals, and nonasbestos mineral fibers may be etiologic agents in some cases. Most males with PMMs reported in the literature survived less than 2 years after diagnosis, although there have been occasional long-term survivors. A study of PMMs in women (Kerrigan et al. 2002), however, found that 40% of the patients survived longer than 4 years. Increasing nuclear and nucleolar size has been shown to correlate with shorter survival in epithelial tumors (Ceruto et al. 2006). The histopathological subtype (see below) is of prognostic significance, as biphasic PMMs are associated with a much shorter survival than pure epithelial tumors

785

(Ceruto et al. 2006); deciduoid mesotheliomas are usually rapidly fatal (Shia et al. 2002). PMMs with low WT-1 expression (25% positive tumor cells), loss of p16 expression, homozygous deletion of p16/CDKN2A, and hemizygous loss of the neurofibromatosis type 2 gene have also been associated with unfavorable prognosis (Krasinskas et al. 2010; Scattone et al. 2012; Singhi et al. 2016). A number of favorable prognostic factors have been identified, including an age less than 60 years, low nuclear grade, low mitotic count, an absence of deep invasion, and low genomic copy number aberrations (Chirac et al. 2016; Feldman et al. 2003; Krasinskas et al. 2016; Liu et al. 2014; Nonaka et al. 2005). A two-tier grading system for epithelioid PMMs, utilizing a combined score for nuclear atypia and mitotic count, has shown that patients with low-grade tumors have a longer overall survival than those with high-grade tumors (Valente et al. 2016). Current therapeutic regimens of cytoreductive surgery (optimally debulked with minimal or no residual disease) and hyperthermic intraperitoneal chemotherapy have obtained some improvement in long-term survival rates (Alexander et al. 2013; Lee et al. 2013).

Pathologic Findings At laparotomy, the visceral and parietal peritoneum are diffusely thickened or extensively involved by nodules and plaques (Fig. 23). The viscera are often encased by tumor (Fig. 24) and may be invaded, although local invasion and

Fig. 23 Peritoneal malignant mesothelioma with (a) extensive omental nodules and (b) plaque-like encroachment of small bowel mesentery

786

Fig. 24 Peritoneal malignant mesothelioma. The tumor encases loops of bowel. (Courtesy of J. Prat, M.D., Barcelona, Spain)

J. A. Irving and P. B. Clement

Fig. 26 Peritoneal malignant mesothelioma. Tubulopapillary pattern, with prominent involvement of ovarian surface

Fig. 25 Peritoneal malignant mesothelioma. Papillary pattern

Fig. 27 Peritoneal malignant mesothelioma. Tumor cells are arranged as small tubules and nests

metastases to the lymph nodes, liver, lungs, and pleura are less frequent than in association with carcinomas with comparable degrees of peritoneal involvement. Significant degrees of invasion or metastatic involvement of the abdominal viscera, however, may be encountered at autopsy, such as transmural invasion of bowel wall or massive replacement of the pancreas. Some tumors incite a striking desmoplastic reaction. As noted earlier, rare PMMs form localized solitary masses. The typical histologic features (Figs. 25, 26, 27, and 28) are identical to malignant mesotheliomas involving the pleura. Most tumors are composed of epithelial cells arranged in tubulopapillary and solid patterns; areas of necrosis may be present. There is usually evidence of

invasion of subperitoneal tissues, such as the omentum. As already noted, intra-abdominal lymph nodes may be involved. The tumor cells usually retain some resemblance to mesothelial cells, with a cuboidal shape and eosinophilic cytoplasm. Usually there are mild to moderate degrees of nuclear atypicality and variably prominent nucleoli. Mitotic figures usually are present but are not numerous. Rare tumors with an exclusively solid pattern of polygonal cells with abundant eosinophilic cytoplasm and prominent nucleoli (“deciduoid” PMMs) (Fig. 29), with one exception, have arisen in the peritoneum (Shia et al. 2002). Two thirds of such tumors have occurred in females, some of whom were adolescents or young adults. In deciduoid

13

Diseases of the Peritoneum

Fig. 28 Peritoneal malignant mesothelioma. The cells are cuboidal or polygonal, with eosinophilic cytoplasm and moderate nuclear atypia

Fig. 29 Peritoneal malignant mesothelioma. Solid growth pattern composed of cells with abundant eosinophilic cytoplasm (deciduoid PMM)

787

mesotheliomas, the presence of high-grade features (marked nuclear pleomorphism, severe atypia, loss of cellular cohesion, and >5 mitoses/ 10 HPF) was associated with shorter mean survival and thus merits notation in the pathology report (Ordonez 2012a). Pleomorphic mesothelioma may also rarely occur in the peritoneum (Ordonez 2012b). Biphasic and sarcomatoid PMMs occur, but are less common than their pleural counterparts, accounting for only 5 of 75 PMMs in one study (Baker et al. 2005; Pavlisko and Roggli 2015). Psammoma bodies are present in approximately one third of cases, but are usually less prominent than in serous neoplasms. Occasional tumors contain a prominent inflammatory infiltrate, such as a dense lymphocytic infiltrate with lymphoid follicles, granulomas, and large numbers of foamy lipid-rich histiocytes (Kitazawa et al. 1984). Some tumors consisting predominantly of cells with clear cytoplasm, rich in glycogen or occasionally lipid, have also been reported (Ordonez 2005a). PMMs may also exhibit areas with signet-ring cell features (Fig. 30) (Ordonez 2013a). The immunohistochemical (see next section) and ultrastructural features of PMMs are similar to their pleural counterparts.

Differential Diagnosis The differential diagnoses of PMM with atypical mesothelial hyperplasia (see section “Mesothelial Hyperplasia”) and of desmoplastic PMM versus reactive fibrosis (see section “Peritoneal

Fig. 30 Peritoneal malignant mesothelioma. (a) The cells are discohesive, with variable cytoplasmic vacuolation, some resembling signet-ring cells (lower left). (b) Diffuse positive calretinin immunohistochemical staining

788

Fibrosis”) have been previously discussed. Rarely, PMM may form multiloculated cysts, but, in contrast to multilocular peritoneal inclusion cyst, are at least focally lined by markedly atypical mesothelial cells, and areas of conventional PMM may be disclosed with thorough sampling. A frequently problematic lesion in the differential diagnosis is adenocarcinoma with diffuse peritoneal involvement, including metastatic adenocarcinomas (see section “Metastatic Tumors”) and adenocarcinomas of primary peritoneal origin, of which the majority are papillary serous carcinomas morphologically identical to those arising in the fallopian tube or ovary (see section “Lesions of the Secondary Müllerian System”). Features favoring a diagnosis of PMM include a prominent tubulopapillary pattern, polygonal cells with moderate amounts of eosinophilic cytoplasm, only mild to moderate nuclear atypia, a paucity of mitotic figures, and the presence of acid mucin (alcianophilic material) rather than neutral (PASD) mucin. Most PMMs are immunoreactive for cytokeratin 5/6, WT-1, and calretinin (Fig. 30b) and usually lack immunoreactivity for a variety of “epithelial” antigens, including claudin-4, carcinoembryonic antigen (CEA), B72.3, Leu-M1 (CD 15), MOC-31, and Ber-EP4. In addition, positive immunoreactivity for PAX-8, PAX-2, and ER favors serous carcinoma; positive staining with calretinin, cytokeratin 5/6, podoplanin, and D2–40, as well as loss of BAP-1, favors PMM, but these markers are less discriminatory as they may be positive in a minor proportion of serous carcinomas (Andrici et al. 2016; Barneston et al. 2006; Chapel et al. 2017; Comin et al. 2007; Joseph et al. 2017; Ordonez 2005b, 2013b). One study found that an h-caldesmon+/calretinin+/ER–/Ber-EP4– immunophenotype strongly favors PMM over serous carcinoma (Comin et al. 2007). However, no single immunohistochemical stain is diagnostic in the separation of PMM from adenocarcinoma, and the results of a panel of antibodies should be interpreted in conjunction with the hematoxylin and eosin (H&E) and mucin stains. “Deciduoid” PMMs must be distinguished from an ectopic decidual reaction involving the peritoneum. Prominent nucleoli, often brisk

J. A. Irving and P. B. Clement

mitotic activity, and cytokeratin immunoreactivity in the deciduoid tumors exclude an ectopic decidual reaction. One study (Lin et al. 1996) reported peritoneal epithelioid hemangioendotheliomas or epithelioid angiosarcomas that have mimicked PMM. Features that suggested the diagnosis of PMM in some of the cases included epithelioid cells in a tubulopapillary pattern and the presence of reactive or neoplastic spindle cells resulting in a focal biphasic pattern. Variable degrees of vascular differentiation and immunoreactivity of the neoplastic cells for endothelial antigens (and negative or weak cytokeratin staining) excluded the diagnosis of PMM. Perivascular epithelioid cell tumor (PEComa) can rarely arise in the mesentery (see ▶ Chap. 10, “Mesenchymal Tumors of the Uterus”), and diffuse peritoneal involvement may mimic mesothelioma (Folpe et al. 2005; Salviato et al. 2006).

Miscellaneous Primary Tumors Intra-abdominal Desmoplastic Small Round Cell Tumor Clinical Features Intra-abdominal Desmoplastic Small Round Cell Tumor is a rare tumor of uncertain histogenesis, but it may ultimately prove to be a primitive tumor of mesothelial origin (“mesothelioblastoma”) (Lae et al. 2002; Ordi et al. 1998; Ordonez 1998a, b; Young et al. 1992). Although most of the tumors are intraabdominal, similar tumors have also been described in the pleura and rarely at a distance from a mesothelium-lined surface (parotid gland, tentorium, and hand). DSRCTs exhibit a reciprocal translocation [t(11;22) (p13;q12)], resulting in fusion of the EWS1 gene on chromosome 22 and the Wilms’ tumor suppressor gene (WT1) on chromosome 11 that appears to be unique for this tumor (Ordi et al. 1998). This fusion results in the expression of the EWS/WT1 chimeric transcript detectable by reverse transcriptase polymerase chain reaction (PCR). The EWS/ERG fusion gene characteristic of Ewing’s

13

Diseases of the Peritoneum

Fig. 31 Intra-abdominal Desmoplastic Small Round Cell Tumor. The cellular nests of tumor are sharply circumscribed and separated by a fibrous stroma. Focal necrosis of the tumor is seen

sarcoma/peripheral neuroectodermal tumors has been found in rare DSRCTs, suggesting some overlap between the two groups of tumors. DSRCTs have a strong male predilection (M:F ratio, 4:1) and are most common in adolescents and young adults (range, 5–76 years) who usually have abdominal distension, pain, and a palpable abdominal, pelvic, or scrotal mass, sometimes in association with ascites. Some patients have had an elevated serum level of CA-125 or neuronspecific enolase (NSE). Laparotomy typically discloses variably sized but usually large, intraabdominal masses associated with smaller peritoneal “implants” of similar appearance. The tumor is sometimes confined to the pelvis, and prominent involvement of the tunica vaginalis or the ovaries may mimic a primary testicular or ovarian tumor (Young et al. 1992). The retroperitoneum is involved in some cases. One tumor appeared to originate within the liver. After initial treatment (debulking and postoperative chemotherapy, irradiation, or both), there may be an initial response, but more than 90% of patients die of tumor progression. Recent studies have advocated complete cytoreductive surgery and hyperthermic intraperitoneal chemotherapy to optimized local disease control (Msika et al. 2010). The bulk of the tumor tends to remain within the peritoneal cavity, although extraabdominal metastases occur in some patients.

789

Fig. 32 Intra-abdominal Desmoplastic Small Round Cell Tumor. The tumor cells have scant cytoplasm and malignant nuclear features

Pathologic Findings On gross examination, the tumors, which may reach 40 cm in maximal dimension, have smooth or bosselated outer surfaces and firm to hard, graywhite, focally myxoid, and necrotic sectioned surfaces. Direct invasion of intra-abdominal or pelvic viscera may occur. Microscopic examination reveals sharply circumscribed aggregates of small epithelioid cells delimited by a cellular desmoplastic stroma (Fig. 31). The aggregates vary from tiny clusters (or even single cells) to rounded or irregularly shaped islands. Other common features include rounded rosette-like or gland-like spaces, peripheral palisading of basaloid cells in some of the nests, and central necrosis with or without calcification. The tumor cells are typically uniform with scanty cytoplasm and indistinct cell borders (Fig. 32), although tumor cells with eosinophilic cytoplasmic “inclusions” and an eccentric nucleus, resulting in a rhabdoid appearance, are frequently also present. Small- to medium-sized, round, oval, or spindle-shaped hyperchromatic nuclei have clumped chromatin and nucleoli that are usually inapparent. Mitotic figures and single necrotic cells are numerous. Architectural features noted in a minority of cases, which can occasionally predominate and lead to diagnostic problems, include tubules, glands (sometimes with luminal mucin), cysts, papillae, anastomosing trabeculae, cords of cells mimicking lobular breast carcinoma, adenoid

790

cystic-like foci, and only a sparse desmoplastic stroma. Cytologic features noted in a minority of cases, which can occasionally predominate, include spindle cells; cells with abundant eosinophilic or clear cytoplasm, which may create a biphasic pattern; signet-ring-like cells; and cells with marked nuclear pleomorphism which may include bizarre nuclei (Ordonez 1998a). Invasion of vascular spaces, especially lymphatics, is a common feature. Lymph nodes are occasionally involved by tumor.

Immunohistochemical and Ultrastructural Findings The usual immunoreactivity for epithelial (low molecular weight cytokeratins, EMA), neural/ neuroendocrine (NSE, CD57/Leu-7), and muscle (desmin) markers, as well as vimentin, suggests divergent differentiation. Desmin and vimentin immunoreactivity is typically paranuclear and globular and is particularly intense in the rhabdoid cells. Immunoreactivity for a wide variety of other antigens has been present in a variable proportion of cases, including most with nuclear staining for the C-terminal of Wilms’ tumor protein (WT1) (Lae et al. 2002; Ordonez 1998b; Zhang et al. 2003). Ultrastructural variability suggests a range of differentiation, with cell junctions of various types, paranuclear intermediate cytoplasmic filaments, and basal lamina surrounding the nests of tumor (Ordonez 1998b). Differential Diagnosis The typical age of the patient, the absence of an extraperitoneal primary tumor, the distribution of the tumor, and its typical microscopic features and immunoprofile facilitate the distinction from other malignant small blue cell tumors in most cases. Distinction of DSRCT from blastemalpredominant Wilms’ tumor may be problematic, as the former may show atypical staining patterns due to full-length or variant transcripts (Murphy et al. 2008) and the latter can exhibit paranuclear desmin and cytokeratin positivity (Arnold et al. 2014). Cyclin D1 immunohistochemistry may be helpful to discriminate DSCRT (negative or 50% tumor cells) (Magro et al. 2017). Identification of the unique reciprocal translocation is diagnostic and may be essential in problem cases.

Solitary Fibrous Tumor Although once referred to as fibrous mesotheliomas, these tumors are now designated solitary fibrous tumors and are believed to originate from submesothelial fibroblasts (Brunnemann et al. 1999; Young et al. 1990). The clinical and pathologic features are similar to their much more common pleural counterparts, including immunoreactivity for CD34 and lack of immunoreactivity for cytokeratin, an immunoprofile that is useful in distinguishing these tumors from desmoplastic mesotheliomas (Brunnemann et al. 1999). Typical tumors are clinically benign. One peritoneal solitary fibrous tumor that was focally sarcomatous was clinically malignant (Fukunaga et al. 1996).

Inflammatory Myofibroblastic Tumor Day et al. reviewed the features of seven cases of abdominal “inflammatory pseudotumor” (Day et al. 1986), a lesion that has also been referred to as plasma cell granuloma or, more recently, inflammatory myofibroblastic tumor (Pettinato et al. 1990). Various anatomical locations have been reported, but most tumors arise in the lung, mesentery, omentum, or retroperitoneum. The abdominal lesions are typically encountered in patients younger than 20 years of age, often in the first decade, who present with a mass, fever, growth failure or weight loss, hypochromic anemia, thrombocytosis, and polyclonal hypergammaglobulinemia. Laparotomy typically reveals a solid mesenteric mass that on microscopic examination consists of myofibroblastic spindle cells, mature plasma cells, and small lymphocytes. The spindle cells often show positive cytoplasmic immunoreactivity for ALK-1, with associated chromosomal translocations detected in approximately 50% of cases. Inflammatory

13

Diseases of the Peritoneum

myofibroblastic tumors are regarded as neoplasms of low-grade or intermediate biologic behavior, which can be associated with favorable outcome, but have a tendency for local recurrence and generally a low risk of distant metastasis. Coffin et al. (2007) recently reported that abdominopelvic tumors had a higher rate of recurrence relative to other anatomical sites and that ALK-negative tumors were more likely to be associated with distant metastases.

Calcifying Fibrous Tumor The rare lesion known as calcifying fibrous tumor, initially considered a pseudotumor, is likely neoplastic, with a predilection for children and young adults but which can occur over a wide age range and in a variety of anatomical sites including the subcutaneous or deep soft tissues and the pleura (Nascimento et al. 2002; Sigel et al. 2001). In the peritoneum, the calcifying fibrous tumor is usually an incidental finding involving the visceral peritoneum of the small intestine or stomach. The tumors are often small (less than 5 cm) but can be larger and sometimes multiple. Microscopically, they are hypocellular, composed of bland spindle cells, hyalinized collagen, a chronic lymphoplasmacytic inflammatory infiltrate, and psammomatous or dystrophic calcifications. The spindle cells are typically CD34-positive and ALK-negative, the latter regarded as evidence that these lesions are distinct from the inflammatory myofibroblastic tumor; rare cells may show positive staining with muscle actin and desmin (Nascimento et al. 2002; Sigel et al. 2001). A recent review of 157 patients reported a 10% recurrence rate and no patient deaths (Chorti et al. 2016).

Omental–Mesenteric Myxoid Hamartoma The omental–mesenteric myxoid hamartoma designation was applied by Gonzalez-Crussi et al. (1983) to a lesion in infants characterized by multiple omental and mesenteric nodules composed

791

of plump mesenchymal cells in a myxoid, vascularized stroma. The diagnosis of the referring pathologists was usually that of some type of sarcoma, but the follow-up was uneventful. The lesions may be hamartomatous or a variant of inflammatory myofibroblastic tumor.

Sarcomas The majority of intra-abdominal sarcomas are of non-peritoneal origin and arise in the retroperitoneum or gastrointestinal tract; they include leiomyosarcomas, liposarcomas, and gastrointestinal stromal tumors and are not discussed further here. Rarely, malignant vascular tumors may arise from the peritoneum (epithelioid hemangioendothelioma, epithelioid angiosarcoma) and are briefly discussed above in the differential diagnosis with malignant mesothelioma (Lin et al. 1996).

Gestational Trophoblastic Disease Rarely, gestational trophoblastic disease (including placental site trophoblastic tumor, hydatidiform mole, and choriocarcinoma) may arise in the peritoneum, presumably secondary to an intra-abdominal pregnancy.

Metastatic Tumors Peritoneal involvement by metastatic tumor is typically a result of seeding from a primary tumor arising within the abdomen or pelvis, most commonly the fallopian tube or ovary. Peritoneal serous tumors in which the ovaries are normal or only minimally involved may arise directly from the fallopian tube or peritoneum (see section “Lesions of the Secondary Müllerian System”) or rarely are metastatic from a serous carcinoma of the endometrium. Other tumors that may be associated with peritoneal seeding include carcinomas of the breast and gastrointestinal tract, especially the colon and stomach, and the pancreas. In such cases, the metastatic tumor may take the form of signet-ring cells widely scattered

792

J. A. Irving and P. B. Clement

Fig. 33 Poorly differentiated adenocarcinoma with signet-ring cells involving the peritoneum. (a) Deceptively bland, malignant cells infiltrate the omental fat lobules,

with an associated desmoplastic stromal reaction. (b) High-power view of signet-ring cells

in a fibrous stroma (Fig. 33). Occasionally, the signet-ring cells can have relatively bland nuclear features, resulting in a deceptively benign appearance.

epithelium. Displacement of coelomic epithelium and subcoelomic mesenchyme during embryonic development could account for the presence of identical lesions within the pelvic and abdominal lymph nodes. The origin of many of these lesions, however, is not known with certainty, and other proposed histogenetic mechanisms are discussed where appropriate. Lesions of the secondary müllerian system include those containing endometrioid, serous, and mucinous epithelium, simulating normal or neoplastic endometrial, tubal, and endocervical epithelium. The metaplastic potential of the pelvic peritoneum also includes differentiation toward cells of transitional (urothelial) type, exemplified most commonly by Walthard nests. Proliferation of the subjacent mesenchyme may accompany epithelial differentiation of the mesothelium or may give rise to a variety of pure mesenchymal lesions composed of endometrial stromal-type cells, decidua, or smooth muscle.

Pseudomyxoma Peritonei Pseudomyxoma peritonei, which is a clinical term referring to the presence of masses of jelly-like mucus in the pelvis and often the abdomen, is usually a result of peritoneal spread from a typically low-grade mucinous neoplasm, usually originating within the appendix or, less commonly, from a primary tumor elsewhere in the gastrointestinal tract. Ovarian involvement is common in such cases, and this topic is discussed in detail in ▶ Chaps. 14, “Epithelial Tumors of the Ovary,” and ▶ 18, “Metastatic Tumors of the Ovary.”

Lesions of the Secondary Müllerian System These peritoneal lesions are characterized by müllerian differentiation on microscopic examination and share an origin from the so-called secondary müllerian system, that is, the pelvic and lower abdominal mesothelium and the subjacent mesenchyme of females (Lauchlan 1972). The müllerian potential of this layer is consistent with its close embryonic relation to the müllerian ducts that arise by invagination of the coelomic

Endometriosis in Usual Sites Endometriosis is defined as the presence of endometrial tissue outside the endometrium and myometrium. Usually both epithelium and stroma are seen, but occasionally the diagnosis of endometriosis can be made when only one component is present, as discussed below.

13

Diseases of the Peritoneum

Etiology and Pathogenesis Two theories have been proposed for the pathogenesis of endometriosis: (1) metastases of endometrial tissue to its ectopic location (metastatic theory) and (2) metaplastic development of endometrial tissue at the ectopic site (metaplastic theory). The metastatic theory explains the majority of cases, but a metaplastic origin likely accounts for occasional cases in which metastatic spread of endometrial tissue is unlikely or impossible (see following). Metastatic Theory Sampson (1927) proposed that endometriosis was caused by reflux of endometrial tissue through the fallopian tubes by a process of retrograde menstruation, with subsequent implantation and growth on peritoneal surfaces. Implantation of menstrual endometrium has also been proposed to explain endometriosis within surgical scars, on traumatized cervical and vaginal mucosa, and within perineal and vulvar scars following vaginal delivery. Passage of refluxed menstrual endometrium from the peritoneal cavity through diaphragmatic defects, diaphragmatic lymphatics, or both may explain pleural endometriosis. Observations supporting the menstrual implantation hypothesis include the following: (1) endometriotic lesions are most common in areas closest to the tubal ostia and occur in a distribution that appears dependent on gravity and uterine position (Ishimaru and Masuzaki 1991); (2) lateral predisposition of ovarian endometriomas, which are more commonly left-sided than right-sided, is a phenomenon attributed to reduced flow of peritoneal fluid due to the presence of the sigmoid colon in the left pelvis (Sznurkowski and Emerich 2008); (3) retrograde menstruation through the fallopian tubes is a common physiologic process, occurring in 90% of menstruating women with patent tubes (Halme et al. 1984); (4) endometriosis is more common in women with early menarche, heavy menstrual flow, long menstrual flow (greater than 7 days), and frequent menses (cycle less than 27 days); (5) breastfeeding has an inverse association with risk of endometriosis (at least partially attributable to postpartum amenorrhea) (Farland et al. 2017); (6) menstrual

793

endometrium is viable, capable of growth in tissue culture and after subcutaneous or intrapelvic injection (D’Hooghe et al. 1995); (7) endometriosis is more frequent in females with congenital obstruction to menstrual flow (Olive and Henderson 1987); and (8) endometriosis may follow uteropelvic or utero-abdominal wall fistulas in experimental animals and humans. Although endometriosis in some scars may be a result of menstrual implantation, endometriosis within scars after uterine operations may be secondary to intraoperative implantation of endometrial tissue (Chatterjee 1980; Steck and Helwig 1966). Supporting this theory is the greater frequency of scar endometriosis after abdominal hysterotomy than after Cesarean section in some studies, consistent with the greater viability of transplanted early-pregnancy endometrium compared to late-pregnancy endometrium. Also, the occurrence of endometriosis within an episiotomy scar is much higher if uterine curettage is performed immediately after delivery than in patients without postdelivery curettage (Paull and Tedeschi 1972). The presence of endometriosis in distant sites (e.g., lungs, extremities, and brain) is most easily explained by hematogenous spread from the uterus. Similarly, endometriosis within lymph nodes is likely a result of lymphatic spread. Evidence supporting the origin of endometriosis from lymphatic or hematogenous spread includes (1) the presence of normal endometrial tissue within endothelium-lined spaces as an incidental histologic finding within the myometrium, most often associated with adenomyosis; (2) the presence of intraluminal vascular involvement in rare endometriotic lesions; (3) the presence of intravascular or perivascular trophoblastic tissue and “decidua” as an incidental microscopic finding within the lungs of pregnant patients (Jelihovsky and Grant 1968); (4) the occurrence of pulmonary endometriosis almost exclusively in women who have had prior uterine operations that could predispose to the embolization of endometrial tissue; (5) the experimental production of pulmonary endometriosis by intravenous injection of endometrial tissue in rabbits; and (6) the observations that tumor cells, blood, dye, and radiographic

794

material can migrate from the pelvis to the umbilicus by retrograde lymphatic flow. Perineural spread may be an alternate mechanism to account for rare cases of endometriosis that involve the nervous system (Siquara de Sousa et al. 2015). Metaplastic Theory The origin of pelvic endometriosis by a process of metaplasia from the pelvic peritoneum is consistent with the putative müllerian potential of this tissue, which, as noted earlier, has been referred to as the secondary müllerian system (Lauchlan 1972). Evidence for the metaplastic theory includes (1) the demonstration of endometriosis in subjects in whom metastasis of normally situated endometrium could not occur or is highly unlikely, such as those with Turner’s syndrome and pure gonadal dysgenesis who are amenorrheic and have hypoplastic uteri (Peress et al. 1982), and in males, (2) the experimental induction of peritoneal endometriosis adjacent to millipore filters that contain endometrial tissue but that prevent cellular transfer, (3) the observation that autologous endometrial implants in rabbits degenerate but are associated with the subsequent development of endometriosis in adjacent tissues, and (4) the juxtaposition of endometriosis with other putative metaplastic lesions of the peritoneum, such as diffuse peritoneal leiomyomatosis (Guarch et al. 2001). Other Etiologic Factors Endometriosis is an idiopathic disease in most patients, and why only a minority of females are affected despite the common occurrence of retrograde menstruation is unknown. The unique microenvironment of the pelvic peritoneum and the pathogenetic mechanisms that may facilitate survival, implantation, and proliferation of endometriotic tissue have garnered recent attention. These include altered cellular and immune response to peritoneal injury and inflammation, with aberrant activation of macrophages and promotion of angiogenesis, and increased expression of key extracellular matrix proteoglycans in the endometriotic peritoneum (Capobianco et al. 2017; Tani et al. 2016). Some potential etiologic factors have been discussed (congenital

J. A. Irving and P. B. Clement

obstruction, iatrogenic implantation); others are summarized in the following section. Familial and Genetic Factors

Several studies concluded that the prevalence of endometriosis is greater in mothers and sisters of women with endometriosis than in the mothers and sisters of their husbands (Lamb et al. 1986; Simpson et al. 1980). Lamb et al. (1986) calculated the overall risk for first-degree relatives to be 4.9%. Genetic studies suggest a polygenic mode of inheritance (influenced by several different genes) or one that is multifactorial (a result of interaction between genetic and environmental factors). In opposition to the foregoing, Houston et al. (1988) concluded that there were methodologic flaws in these studies and that an inherited tendency to endometriosis has not yet been substantiated. Recently, genetic markers that may impart an increased risk of endometriosis have been identified by genome-wide association studies (Fung et al. 2015). Molecular genetic analysis has elucidated a number of intriguing theories regarding the pathogenesis of endometriosis (Bulun 2009). By microarray analysis, Wu et al. (2006) have shown that in patients with endometriosis, the ectopic and eutopic endometria have different gene expression profiles. Putative endometrial progenitor/stem cells, which are thought to reside in the basalis endometrium and possibly in the bone marrow, have been characterized by in vitro and in vivo assays (Sasson and Taylor 2008). It has been shown that patients with endometriosis shed significantly more basalis layer during menstruation compared with normal controls, supporting the hypothesis that endometriotic implants develop from endometrial progenitor/ stem cells, which are in turn derived from retrograde menstrual flow, known to be higher in women with endometriosis (Halme et al. 1984; Leyendecker 2002; Sasson and Taylor 2008). Hormonal Factors

Because endometriosis occurs almost exclusively in women of reproductive age, hormonal factors may play an etiologic role. The rare examples of endometriosis in phenotypic females with gonadal

13

Diseases of the Peritoneum

dysgenesis and in males have usually been associated with the use of exogenous estrogens (Martin and Hauck 1985; Peress et al. 1982). Similarly, smoking and exercise, which are inversely correlated with endogenous estrogen levels, appear to be protective factors for the development of endometriosis. In a large epidemiologic study, factors associated with an increased risk of endometriosis included low body weight, alcohol use, and certain menstrual characteristics (early menarche, short cycle length, and heavy menstrual cycles) (Matalliotakis et al. 2008). It has been suggested that the progestational milieu of pregnancy may inhibit the development of endometriosis. Many studies have indicated that endometriosis is more likely to occur in women who have delayed pregnancy and is less common in multiparous women (Redwine 1987). Similarly, in some studies, patients with endometriosis are much less likely to have used oral contraceptives than similar patients without endometriosis. Some studies have found an increased frequency of the luteinized unruptured follicle syndrome (LUFS) in patients with endometriosis. In normal women, the ruptured corpus luteum releases its progesterone-rich fluid into the peritoneal cavity. It has been postulated that this fluid may inhibit implantation and growth of refluxed endometrial fragments at the time of menstruation (Koninckx et al. 1980). In patients with LUFS, a corpus luteum is formed, but rupture and fluid release do not occur, resulting in lowered lutealphase levels of progesterone in the peritoneal fluid (Koninckx et al. 1980). This local hormonal imbalance may be critical in allowing endometrial cells to implant on the peritoneum. Other studies, however, have shown no difference in the lutealphase peritoneal fluid hormone values in women with and without endometriosis. Immune Factors

One study has demonstrated a reduced T lymphocyte-mediated cytotoxicity to autologous endometrial cells and a decreased lymphocyte stimulation response to autologous endometrial antigens in patients with endometriosis (Steele et al. 1984). The degree of depressed cellular

795

immunity was directly proportional to the severity of the disease. The authors of this study suggested that certain cell-mediated immune mechanisms that may be operative in limiting the growth of endometriotic tissue may be impaired in patients with endometriosis. A recent study found a significant decrease in activated regulatory T cells from endometrioma and eutopic endometrium, in comparison to endometrium from women without endometriosis, providing additional evidence for a dysregulated immune response in the pathogenesis of endometriosis (Tanaka et al. 2017). Other authors have suggested that the growth of endometriotic implants may be stimulated by activated macrophages. ER positivity has been demonstrated in endometriotic macrophages, and, in murine models, estradiol stimulated macrophage–nerve interactions (Greaves et al. 2015). It has been shown that local production of interleukin 4, a cytokine involved in the Th2 immune response, induces proliferation of endometriotic stromal cells (OuYang et al. 2008). Cyclooxygenase-2, which is involved in the biosynthesis of prostaglandin E2, is highly expressed in ectopic endometria relative to eutopic endometria and is thought to play a promotory role in the development of endometriosis (Banu et al. 2008).

Clinical Features Epidemiologic Factors The highest risk of the disease has traditionally been considered to be in the upper socioeconomic levels of developed societies, especially among women who delay pregnancy, although, according to Houston, these associations have not been proven statistically (1988). Although endometriosis was once considered to be more common in Caucasians, studies showing a similar frequency of the disease in Asians and Africans cast doubt on this view. The true prevalence of endometriosis is unknown as many patients are asymptomatic; estimates for the prevalence of the disease in women of reproductive age are 10–15%. Prevalence figures, however, have varied widely, depending on the population studied and the

796

method of diagnosis (clinical, operative, or pathologic). Similarly, a study of the incidence rates of pelvic endometriosis in white females of reproductive age in Rochester, Minnesota (USA), found that the overall incidence of the disease more than doubled (from 108.8 to 246.9 cases per 100,000 person-years) as the definition of a case was extended from histologically confirmed cases to clinically and surgically diagnosed cases (Houston et al. 1987). More than 80% of affected patients are in the reproductive age group. In one study, the age-specific incidence rates increased in successive age groups through age 44 and then declined for women 45–49 years (Houston et al. 1987). Less than 5% of cases occur in postmenopausal women, and in these patients the disease is frequently not diagnosed premenopausally (Kempers et al. 1960). Endometriosis can be clinically significant in this age group, with 20–30% of affected patients requiring operative management (Kempers et al. 1960; Punnonen et al. 1980). In some postmenopausal patients with endometriosis, an association with obesity and endometrial carcinoma has been noted, suggesting that hyperestrinism may play a role, but in other series, a majority of patients have had no obvious exogenous or endogenous source of estrogen (Kempers et al. 1960). Chronic endometritis has been shown to be more prevalent in women with endometriosis (38.5% of patients vs. 14.1% without endometriosis) (Cicinelli et al. 2017). Almost 10% of patients with endometriosis are adolescents (Chatman and Ward 1982). Endometriosis was found at laparoscopy in approximately 50% of teenage patients with dysmenorrhea or chronic pelvic pain in three studies (Chatman and Ward 1982). In some studies, adolescents with endometriosis have a particularly high frequency of a congenital obstruction to menstrual flow. Symptoms and Signs The recurrent cyclic menstrual, inflammatory, and fibrotic changes within the endometriotic lesions are likely responsible for most of the symptomatology of endometriosis, although there is often no direct relationship between the extent of the disease and the severity of the symptoms

J. A. Irving and P. B. Clement

(Chatman and Ward 1982). An exception to the foregoing applies to women with deeply infiltrating endometriosis, a clinical term used for patients with deep pelvic pain, usually in the form of severe dyspareunia and dysmenorrhea, which is often associated with rectovaginal lesions or involvement of the bowel, ureter, or bladder (Cornillie et al. 1990). One study has shown significant reduction in painful symptoms obtained with complete surgical excision of deep lesions (Chopin et al. 2005). Hormonal responsiveness of the lesions as judged histologically also does not correlate with symptoms, and microscopic examination of symptomatic endometriosis in postmenopausal patients typically reveals atrophic changes (Kempers et al. 1960). Age generally does not appear to affect disease severity in most studies (Houston et al. 1988). An exception to the foregoing is one study in which women in the age group of 26–52 years had less extensive disease than women 16–25 years of age (Redwine 1987). A higher frequency of nulliparity in the younger women appeared to account for part of this difference (Houston et al. 1988). Another study found that endometriosis in postmenopausal patients was morphologically less extensive and less active in appearance relative to endometriosis in premenopausal women, but that the endometriotic foci retained the same immunoprofile by ER and PR immunostaining (Cumiskey et al. 2008). The typical symptoms that are attributed to pelvic endometriosis are acquired dysmenorrhea; lower abdominal, pelvic, and back pain; dyspareunia; irregular bleeding; and infertility. Infertility is present in up to 30% of women with endometriosis, although the putative association between mild endometriosis and infertility has been challenged and remains controversial. The subject of endometriosis-related infertility has been reviewed elsewhere (Gupta et al. 2008) and is not considered in detail here. Potential pathogenetic factors include tubal factors (adhesions, luminal obstruction), ovarian factors (anovulation, luteal-phase dysfunction, LUFS), immune factors (antiendometrial antibodies), peritoneal factors (increased prostaglandins, increased macrophages), and an increased risk of spontaneous abortion.

13

Diseases of the Peritoneum

Pelvic examination may reveal tender nodules in the cul-de-sac and uterosacral ligaments; tender, semifixed, cystic ovaries; and a fixed, retroverted uterus. The rectovaginal septum may also be tender and indurated. The endometriotic lesions frequently enlarge and become more painful during menses. The clinical manifestations also vary according to the site of the endometriosis, as is discussed later in this chapter. As the clinical manifestations of endometriosis are frequently nonspecific, vary widely between patients, and may be absent in a high proportion of patients, a definitive diagnosis requires direct visualization by laparoscopy (or laparotomy) and, ideally, biopsy. Hormonal suppression and surgical ablation remain the commonly employed therapeutic modalities, and while treatment of endometriosis is not discussed here, it is important to note that with advances in the understanding of the pathogenesis of endometriosis, future methods may include molecular-targeted drug therapies such as cyclooxygenase inhibitors and immunomodulators in an attempt to minimize the need for surgical intervention (Bulun 2009; Gupta et al. 2008). Laparoscopic Findings A number of studies have stressed that endometriotic foci, especially early ones, are frequently nonpigmented and may have a wide variety of laparoscopic appearances, including clear, white, and red lesions (Jansen and Russell 1986; Martin et al. 1989). Sequential laparoscopic examinations indicate that nonpigmented endometriotic implants eventually evolve into the typical pigmented lesions (Jansen and Russell 1986). Even in patients with laparoscopically typical disease, biopsy may yield only nondiagnostic tissue, and, thus, in the opinion of some authors, diagnosis and treatment should not always depend on microscopic confirmation (Chatman and Zbela 1987). Other authors have found that 25% of laparoscopically atypical lesions prove to be endometriosis on histological examination and therefore advocate that all lesions suggestive of endometriosis, both typical and atypical, should be excised if eradication is the surgical objective (Albee et al. 2008). In another study, only 50% of

797

all laparoscopic biopsies from clinically suspicious foci were proven microscopically to be endometriosis (Walter et al. 2001). Laparoscopically detectable defects or “pockets” involving the pelvic peritoneum are frequently associated with, and likely caused by, endometriosis. In one study, 80% of women with pelvic peritoneal defects had endometriosis, and, in another, the endometriotic foci were often located along the edges of the defects. Conversely, 18–28% of women with endometriosis had peritoneal defects (Redwine 1989). Serum Markers Levels of serum CA-125 may be elevated in patients with endometriosis, and concentrations correlate with both the severity and the clinical course of the disease (Santulli et al. 2015). The serum test has low sensitivity, however, and is not appropriate for general screening purposes. In contrast, CA-125 levels have acceptable sensitivities and very high specificities in populations with a relatively high prevalence of the disease and are useful in monitoring response to treatment. A wide range of blood biomarkers have been investigated to potentially aid in the nonsurgical diagnosis of endometriosis, including antiendometrial antibodies, but in a recent large meta-analysis, none were found to be of sufficient sensitivity and specificity to support routine clinical application (Nisenblat et al. 2016). Effects of Pregnancy Although rare cases of endometriosis undergo permanent regression during pregnancy, the ameliorative effect of pregnancy noted in many cases of endometriosis is only temporary. The behavior of endometriosis during pregnancy is extremely variable among different patients and between one pregnancy and another in the same patient. During pregnancy, visible endometriotic lesions frequently undergo initial enlargement, with occasional ulceration and bleeding, followed by shrinkage. In most sites, there is a decrease in the associated pain. A rare complication of endometriosis during pregnancy is intrapartum or postpartum rupture of

798

J. A. Irving and P. B. Clement

the lesion, most probably caused by a softening of the lesion secondary to stromal decidualization, pressure from the expanding uterus, or both. Rupture occurs most frequently in the ovaries or bowel, typically resulting in perforation and an acute abdomen. Rarely, hemoperitoneum, sometimes fatal, is caused by hemorrhage from decidualized endometriotic lesions at term. Rare Complications Massive, sometimes serosanguineous, ascites can occur in patients with pelvic endometriosis; a right pleural effusion is also present in one third of such patients (Muneyyirci-Delale et al. 1998). If one or both ovaries are involved, the operative findings may simulate those of an ovarian carcinoma. The pathogenesis of the ascites is not clear. Possible sources include production by endometriotic cysts, irritated peritoneal mesothelial cells, or the ovarian serosa (Meigs-like syndrome). Other rare complications include hemorrhage from an endometriotic focus and spontaneous rupture of ovarian endometriotic cysts, resulting in an acute abdomen.

Gross Features of Peritoneal and Ovarian Endometriosis The most common anatomical sites of endometriosis are the ovaries; the uterosacral, broad, or round ligaments; the rectovaginal septum and cul-de-sac; and the serosa of the uterus, fallopian tubes, or other pelvic organs (see Table 1). Less common sites include the serosa of the large

bowel, small bowel, and appendix, the mucosa of the female genital tract, the skin, the urinary tract, and the pelvic lymph nodes. These sites and other additional rare sites of involvement are discussed separately below. Depending on their duration and their superficial or deep location in relation to the peritoneal surface, endometriotic foci may appear as punctate, red, blue, brown, or white spots or patches with either a slightly raised or a puckered surface (Fig. 34). Ecchymotic or brown areas have sometimes been described as “powder burns.” The endometriotic foci are frequently associated with dense fibrous adhesions. The lesions may form nodules or cysts or both. Rarely, endometriosis can take the form of polypoid masses that project from the serosal surfaces, into the lumens of endometriotic cysts, or from the mucosa of the bowel (Fig. 35) (Stewart and Bharat 2016) or bladder. In some of these cases, there is a history of exogenous estrogen use, and hyperplastic changes are found on microscopic examination (Parker et al. 2004). This appearance, which we refer to as polypoid endometriosis, can simulate a malignant tumor on clinical, intraoperative, or pathologic examination (Mostoufizadeh and Scully 1980; Parker et al. 2004). Endometriotic cysts (endometriomas) most commonly involve the ovaries, where they can partially or almost completely replace the normal tissue; bilateral involvement occurs in one third to one half of the cases (Egger and Weigmann 1982). The cysts rarely exceed 15 cm in diameter; larger

Table 1 Sites of endometriosis Common Less common Ovaries Large bowel, small bowel, appendix Uterine ligaments (uterosacral, round, broad) Mucosa of the cervix, vagina, and fallopian tubes Rectovaginal septum Skin (scars, umbilicus, vulva, perineum, inguinal region) Cul-de-sac Ureter, bladder Peritoneum of the uterus, tubes, rectosigmoid, ureter, bladder

Omentum, pelvic lymph nodes Inguinal (noncutaneous)

Rare Lungs, pleura Soft tissues, breast Bone Upper abdominal peritoneum Stomach, pancreas, liver Urethra, kidney, prostate, paratesticle Sciatic nerve, subarachnoid space, brain

13

Diseases of the Peritoneum

Fig. 34 Endometriosis of the ovary. Multiple, hemorrhagic lesions involve the ovarian surface. (Courtesy of R.E. Scully, M.D., Boston, MA)

Fig. 35 Polypoid endometriosis. A polypoid mass projects from the mucosa of the large bowel

examples are more likely to harbor a neoplasm. Endometriotic cysts are commonly covered by dense fibrous adhesions, which may result in fixation to adjacent structures. The cyst walls are usually thick and fibrotic, with a smooth or shaggy, brown to yellow lining (Fig. 36). The cyst contents typically consist of altered, semifluid or inspissated, chocolate-colored material; rarely, the cyst is filled with watery fluid. Any solid areas in the cyst wall or intraluminal polypoid projections should be sampled histologically to exclude a neoplasm originating in the cyst (see page 818).

Typical Microscopic Findings Many of the problems and pitfalls encountered in the histological diagnosis of endometriosis have

799

Fig. 36 Endometriotic cyst of the ovary. The cyst has been opened to reveal a focally hemorrhagic lining. Multiple hemorrhagic lesions also involve the uterine serosal surface

Fig. 37 Endometriosis of cul-de-sac. Cystic endometrial glands with a cuff of endometrial stroma are surrounded by fibrous and adipose tissue

been addressed in detail in a comprehensive review (Clement 2007). The typical appearance in reproductive age women, in whom the disease is usually diagnosed, is of one or more glands lined by endometrioid epithelium, surrounded by a mantle of densely packed small fusiform cells with scanty cytoplasm and bland cytology, typical of nonneoplastic endometrial stromal cells (Figs. 37 and 38). Small blood vessels, which may be engorged, are present and sometimes draw attention to the lesion on low-power examination. When seen in the ovary, the most common site encountered by the surgical pathologist, endometriosis varies from simple to microscopically dilated glands (Fig. 39) to grossly recognizable endometriotic cysts. This spectrum is seen at

800

Fig. 38 Endometriosis of cul-de-sac (higher magnification of Fig. 37). Endometriotic glands are lined by inactive epithelium and surrounded by a thin rim of endometrial stroma

Fig. 39 Subtle endometriosis involving the ovarian surface. (a) The periglandular endometriotic stroma is only focal and less cellular than usual. (b) The periglandular endometriotic stroma is obscured by hemorrhage. In both examples, failure to recognize the endometriotic stroma

J. A. Irving and P. B. Clement

extraovarian sites, although striking cysts are less common to rare, depending on the site. Endometriosis may occur anywhere in the ovary but is most common in the cortex. Sometimes it is very superficial and may occur on the surface as small nodules and irregularly shaped aggregates, or even have a plaque-like configuration (Fig. 39). Surface endometriosis is typically associated with fibrous tissue and inflammatory cells, and, if prominent and of significant duration, there may be conspicuous adhesions. Glands, which can sometimes be cystic, may hang off the surface of the ovary, tethered to it by the associated stroma and fibrous tissue. Endometriotic glands in the cortex of the ovaries of perimenopausal or

could result in the diagnosis of endometriosis being missed and the endometriotic glands being misinterpreted as epithelial inclusion glands. (c) Endometrial stromal cells highlighted by positive CD10 immunoreactivity

13

Diseases of the Peritoneum

postmenopausal women, or glands that are atrophic for any reason, may be mistaken for inclusion glands and cysts if the often subtle, sometimes barely perceptible, cuffs of stroma are overlooked or obscured by hemorrhage or histiocytes (see Fig. 39). Immunostaining for CD10 can facilitate the recognition of the stromal cells, particularly when sparse and when glandular epithelium is minimal or absent (Figs. 39c and 40) (Sumathi and McCluggage 2002). At the time of menstruation, hemorrhage may occur within the stroma and glandular lumens of endometriotic foci, as well as a secondary inflammatory response consisting predominantly of a diffuse infiltration of histiocytes. The histiocytes

Fig. 40 Endometriosis in a postmenopausal woman. The glands are cystic, atrophic, and separated by a fibrous stroma

Fig. 41 Lining of ovarian endometriotic cyst. (a) The lining consists of cystically dilated endometrial glands and numerous pigment-laden histiocytes within the

801

typically convert the extravasated red blood cells into glycolipid and granular brown pigment, becoming so-called pseudoxanthoma cells (Figs. 41 and 42) that can replace most or all the endometriotic stroma (Clement et al. 1988). Most of the pigment is ceroid (lipofuscin, hemofuscin), and hemosiderin is typically present to a much lesser extent (Clement et al. 1988). The amount of pigment in an endometriotic lesion appears to increase with its age, and early lesions are frequently nonpigmented (Jansen and Russell 1986). Variable numbers of lymphocytes and smaller numbers of other inflammatory cells may be present. Large numbers of neutrophils with microabscess formation should raise the possibility of secondary bacterial infection (Schmidt et al. 1981). As already mentioned, a common manifestation of ovarian endometriosis is striking cystification resulting in an endometriotic cyst. The epithelial and stromal lining of an endometriotic cyst frequently becomes attenuated, and the former may be reduced to a single layer of cuboidal cells that may retain some endometrial characteristics but which are often devoid of specific features. In such circumstances, recognition of the cyst as endometriotic may only be possible if a rim of subjacent endometrial stroma persists. Commonly, the cyst lining of the endometrial epithelium and stroma is totally lost and replaced by granulation tissue, dense fibrous tissue containing fibroblasts with particularly small nuclei, and variable

subjacent stroma. (b) Endometriotic surface epithelial lining with underlying, lightly pigmented histiocytes

802

numbers of pseudoxanthoma cells (presumptive endometriosis) (Fig. 42). In some “old” endometriotic cysts, ossification, calcification, and old luminal blood clot can produce striking gross and microscopic appearances. The epithelial cells lining the endometriotic cysts are often focally large and cuboidal with abundant eosinophilic cytoplasm and large atypical nuclei (Fig. 43) (Clement 2007; Seidman 1996). The significance of such nuclear atypia is unclear. Although it may be reactive, cells with these features may merge with clear cell adenocarcinomas and endocervical-like mucinous (seromucinous) atypical proliferative/ borderline tumors (EMBLTs) (see section “Neoplasms Arising from Endometriosis”) (Fukunaga

J. A. Irving and P. B. Clement

et al. 1997; Rutgers and Scully 1988a, b). When this atypia is an isolated finding in an endometriotic cyst, the follow-up is typically uneventful (Seidman 1996), but in one study, a patient with atypical endometriosis had a subsequent diagnosis of extraovarian endometrioid carcinoma (Fukunaga et al. 1997). Endometriosis that involves smooth muscle in the uterine ligaments or the walls of hollow viscera differs significantly in its appearance from that of endometriosis in the ovaries and the peritoneal surfaces. In the former, there is typically a striking proliferation of the indigenous smooth muscle, often resulting in a firm, solid, tumorlike mass. The appearance is similar to that of adenomyosis with secondary striking myometrial hypertrophy.

Unusual Microscopic Findings

Fig. 42 Lining of the ovarian endometriotic cyst. In this field, the lining consists only of fibrotic granulation tissue and pigment-laden histiocytes (presumptive endometriosis)

Fig. 43 Lining of the ovarian endometriotic cyst. The epithelial cells show notable nuclear atypia

Metaplastic Glandular Changes Metaplastic changes similar to those occurring in eutopic endometrial glands have been described in endometriotic glands (Fukunaga and Ushigome 1998). These changes include ciliated, eosinophilic, hobnail, and, rarely, squamous and mucinous metaplasia (Fig. 44); the latter may be characterized by the presence of endocervicaltype cells or, less often, goblet cells. In one study of ovarian endometriosis (Fukunaga and Ushigome 1998), there was a significant association between the presence of metaplasia in the endometriosis and a synchronous ovarian epithelial cancer. Additionally, all four endocervical-

Fig. 44 Mucinous metaplasia in endometriosis

13

Diseases of the Peritoneum

803

Fig. 45 Pregnancy-induced changes within the endometriosis. (a) The endometriotic gland is atrophic, and the stroma exhibits marked decidual transformation. (b) The endometriotic glands exhibit the Arias–Stella reaction

Fig. 46 Hyperplasia within the endometriosis. Endometriotic glands exhibit architectural and cytologic atypia. Endometrioid carcinoma was found elsewhere in the specimen (see Fig. 63)

like mucinous (seromucinous) atypical proliferative/borderline tumors (EMBLTs) in the same study were associated with foci of ovarian endometriosis that exhibited both mucinous metaplasia and hyperplasia. In some cases of endometriosis, the distinction between papillary mucinous metaplasia and an early EMBLT may be arbitrary. The specific circumstance of intestinal metaplasia involving appendiceal endometriosis is discussed below (see section “Intestinal Endometriosis”). Unusual Hormonal Changes Endometriotic tissue usually exhibits striking progestational changes (Fig. 45) during pregnancy or

progestin therapy. In such cases, examination reveals a decidual reaction with atrophy of the endometrial glands, which are small and lined by cuboidal or flattened epithelial cells (Fig. 5a). In pregnancy the glands can rarely exhibit the Arias–Stella reaction (Fig. 45b), optically clear nuclei, or both. Necrosis of the decidual cells, foci of marked stromal edema, and infiltration by lymphocytes are additional findings in patients receiving progestational agents. Inactive or atrophic changes similar to those that are seen typically in the endometriotic foci of postmenopausal patients may be present in premenopausal patients treated with hormones (Nisolle-Pochet et al. 1988). Additionally, endometriotic foci often disappear or are replaced by fibrous tissue after danazol therapy. Irregular cystic glandular dilatation of tubal endometriosis, with concurrent eutopic endometrial changes secondary to PR modulator treatment for uterine leiomyomas, has also been reported (Bateman et al. 2017). Hyperplastic Glandular Changes A variety of hyperplastic and atypically hyperplastic changes similar to those occurring in the endometrium have been described in endometriotic glands, sometimes related to an endogenous or exogenous estrogenic stimulus (Fig. 46) (Fukunaga et al. 1997; Sampson 1927; Yantiss et al. 2000) or tamoxifen therapy (McCluggage et al. 2000; Schlesinger and Silverberg 1999). Hyperplastic changes are particularly common in

804

J. A. Irving and P. B. Clement

cases of polypoid endometriosis (Parker et al. 2004). It is logical to conclude that such atypical changes have a malignant potential similar to those in the endometrium, and, indeed, rare cases of hyperplastic endometriosis have preceded the development of an adenocarcinoma in the same area or have coexisted with carcinoma in the same specimen (Fig. 46; see also Fig. 63, later in this chapter) (LaGrenade and Silverberg 1988). Stromal Changes The endometriotic stroma may also undergo metaplasia, typically smooth muscle metaplasia, which is encountered most often within the walls of ovarian endometriotic cysts but occasionally elsewhere (Fig. 47) (Fredericks et al. 2005; Scully 1981). Extensive amounts of smooth muscle within the endometriotic stroma can result in “endomyometriosis” or uterus-like masses, which have been described within an obturator lymph node, the ovary, the small bowel, the broad ligament, and the lumbosacral region and, in males, in the scrotum (Pai et al. 1998; Rahilly and Al-Nafusi 1991; Young and Scully 1986). In some cases, a uterus-like mass in the region of the ovary may possibly represent a congenital malformation rather than an unusual manifestation of endometriosis (Pueblitz-Peredo et al. 1985). In one case, multifocal endometriosis with marked nodular smooth muscle metaplasia involved the pelvic sidewall (Kim et al. 2015a). Occasional cases of endometriosis can elicit a striking periglandular myxoid (Clement et al. 1994) (Fig. 48) or elastotic (Clement and Young 2000) response (Fig. 49), which in both situations can focally obliterate the endometriotic stroma. In rare cases, extensive myxoid change in endometriosis was misinterpreted as pseudomyxoma peritonei and/or metastatic adenocarcinoma, one at the time of frozen section (Clement et al. 1994; Hameed et al. 1996; Tang et al. 2010). Anatomical location and hormonal factors appear to be predisposing factors, as there is a propensity for myxoid change to occur in endometriosis of the skin or superficial soft tissues, and also during pregnancy or the puerperium; the latter situation may be further confounded by the presence of decidual change of the stromal cells (Clement 2007).

Fig. 47 Endometriosis with smooth muscle metaplasia. Endometriotic glands and stroma surrounded by extensive, metaplastic smooth muscle

Fig. 48 Endometriosis with prominent myxoid stroma. A small endometriotic gland with a periglandular rim of endometriotic stroma is surrounded by loose fibrous tissue and pools of acellular mucin. This appearance was misinterpreted as pseudomyxoma peritonei on frozen section examination

Fig. 49 Endometriosis with prominent elastotic stroma. Large masses of elastic tissue replace the normal endometriotic stroma (elastic tissue stain)

13

Diseases of the Peritoneum

Stromal Endometriosis Some cases of endometriosis are characterized by an absence or rarity of glands, so-called stromal endometriosis (Boyle and McCluggage 2009; Clement and Young 2000; Clement et al. 1990); the same term was used in the older literature to refer to what is now designated low-grade endometrial stromal sarcoma (ESS). Stromal endometriosis is most commonly encountered in the ovary, where it is typically an incidental microscopic finding within the ovarian stroma (“benign stromatosis”). There is usually no associated pelvic endometriosis, and the process likely represents a metaplastic response of the ovarian stromal cells. A disproportionate number of cases of stromal endometriosis are seen within the superficial stroma of the uterine cervix (see page 807–808) (Clement et al. 1990). Endometriosis involving the pelvic peritoneum can take the form of multiple small nodules of endometriotic stroma in which endometriotic glands are absent or rare, a finding referred to as micronodular stromal endometriosis (Boyle and McCluggage 2009; Clement and Young 2000) (Fig. 50). As noted above, immunostaining for CD10 may be of assistance in confirming the presence of endometriotic stromal cells, but this marker is less useful in cases of stromal endometriosis involving the cervix, as normal cervical stroma may be strongly CD10-positive (McCluggage et al. 2003). Rarely, endometriotic stromal cells may show foci of symplastic-type atypia (Shah and McCluggage 2009).

805

Necrotic Pseudoxanthomatous Nodules Occasionally, ovarian and extraovarian endometriosis take the form of “necrotic pseudoxanthomatous nodules,” which typically occur in postmenopausal women (Clement et al. 1988). Multiple nodules can be attached to the peritoneum or, less commonly, lie free in the peritoneal cavity. When associated with enlargement of one or both ovaries, the intraoperative findings can mimic those of carcinoma with peritoneal spread. The nodules are characterized by a central zone of necrosis surrounded by pseudoxanthoma cells, often in a palisaded arrangement, hyalinized fibrous tissue, or both (Fig. 51). Typical endometriotic glands and stroma are sparse or absent within the nodules and their immediate vicinity, but foci of recognizable endometriosis are usually present in the ovaries. The typical postmenopausal age group of the patient and the appearance of the nodules suggest that they represent end-stage or burned-out foci of endometriosis that should be distinguished from other necrotic peritoneal and ovarian granulomas, as well as necrotic tumor, on histologic examination. Rare Miscellaneous Findings Rare examples of endometriosis have been encountered in intimate association with foci of peritoneal leiomyomatosis, glial implants of ovarian teratomas, and nodules of splenosis. Perineural and vascular invasion can occur rarely in otherwise typical, benign endometriotic lesions,

Fig. 50 Micronodular stromal endometriosis involving the appendiceal serosa. (a) Two stromal nodules (arrows) are evident at far left and far right. (b) High-power view of one nodule

806

J. A. Irving and P. B. Clement

Fig. 53 Schistosomiasis endometriotic cyst

Fig. 51 Necrotic pseudoxanthomatous nodule of the endometriosis. A central area of necrosis is surrounded by pseudoxanthoma cells and an outer zone of fibrous tissue

Fig. 52 Liesegang rings in an endometriotic cyst

findings that may incorrectly suggest the diagnosis of malignancy (Roth 1973). Liesegang rings are eosinophilic, acellular, ringlike structures composed of periodic precipitation zones from colloidal solutions that are supersaturated in vitro or in vivo. They are typically encountered within necrotic, inflamed, or fibrotic tissues and have been found on microscopic examination within endometriotic cysts (Fig. 52) (Perrotta et al. 1998). These structures have been confused with, and should be distinguished from, parasites and foreign material on histologic examination (Clement et al. 1989). Indeed, on rare occasion, schistosomiasis ova within an endometriotic cyst may be encountered (Fig. 53) (Abrao et al. 2006).

ova

in

ovarian

Microscopic examination of the fallopian tubes in patients with endometriosis has revealed nonspecific chronic salpingitis in as many as one third of cases (Czernobilsky and Silverstein 1978). A less common lesion, so-called pseudoxanthomatous salpingitis or pseudoxanthomatous salpingiosis, characterized by infiltration of the tubal mucosa by pseudoxanthoma cells, is almost always associated with pelvic endometriosis (Czernobilsky and Silverstein 1978; Clement et al. 1988).

Ultrastructural, Histochemical, and Steroid Receptor Studies Endometriotic glands typically exhibit ultrastructural features that represent a response, but an incomplete one, to the prevailing hormonal milieu of the particular phase of the menstrual cycle. In contrast to eutopic endometrial glands, it is usually not possible to date the glands precisely within the secretory phase because of marked interglandular and intraglandular variability. Ultrastructural examination of endometriotic tissue following danazol treatment shows either arrest of the endometriotic glandular epithelium in the early proliferative phase or disorganization of the epithelial cells with atrophic changes. ER and PR are present in the endometriotic glands and stroma but usually in lower concentrations than in eutopic endometrium (Bur et al. 1987). In a variable number of cases, one or both receptors are absent. Moreover, the normal variation in the quantity of both receptors exhibited by eutopic endometrium during the menstrual cycle is diminished or absent within foci of endometriosis (Lessey et al. 1989). Differences in receptor

13

Diseases of the Peritoneum

concentrations between eutopic endometrium and endometriotic epithelium in response to danazol have also been noted. No correlation has been found between receptor levels and severity of symptoms. In summary, the findings of these studies are consistent with the incomplete and variable hormonal response of endometriotic foci observed on microscopic examination. They indicate a greater degree of autonomy of endometriotic tissue from the mechanisms controlling eutopic endometrium and may explain the failure of hormonal therapy in some patients (Metzger et al. 1991).

Differential Diagnosis Endometriosis may be accompanied by, and should be distinguished from, endosalpingiosis, which is characterized by glands lined by benign tubal-type epithelium, unassociated with endometrial stroma or the usual histiocytic inflammatory reaction of endometriosis (see section “Endosalpingiosis”). A misdiagnosis of endosalpingiosis or, if in the ovary, an epithelial inclusion gland (see ▶ Chap. 12, “Nonneoplastic Lesions of the Ovary”) is likely when the endometriotic stroma is sparse or obscured by hemorrhage (see Fig. 39). Necrotic pseudoxanthomatous nodules should be distinguished from other ovarian and peritoneal necrotic nodules, such as infectious granulomas and isolated palisading granulomas of the ovary (see ▶ Chap. 12, “Nonneoplastic Lesions of the Ovary”), and, as noted earlier, peritoneal granulomas related to diathermy (Clarke and Simpson 1990). Such lesions, in addition to having characteristic features, lack the numerous pseudoxanthoma cells that are typical of endometriotic lesions. Rare low-grade ESSs contain numerous benign-appearing or atypical endometrial glands, to the extent that confusion with endometriosis may occur (Clement and Scully 1992). Indeed, it is likely that at least some cases referred to as aggressive endometriosis are examples of ESS with prominent glandular differentiation. These tumors, however, in contrast to typical endometriosis, contain foci of more typical ESS devoid of glands, and, in some cases, prominent mitotic

807

activity of the stromal cells, sex cord-like elements, and prominent vascular invasion. A diagnosis of adenosarcoma was initially considered in some cases of polypoid endometriosis. Adenosarcomas, in contrast to polypoid endometriosis, are characterized by a stromal component that usually exhibits dense periglandular cellularity, atypia (albeit mild in many cases), intraglandular papillae, and increased mitotic activity.

Cervical and Vaginal Endometriosis Superficial endometriosis of the uterine cervix is more common than is generally appreciated (Baker et al. 1999; Clement et al. 1990; Gardner 1966). The predilection for sites of trauma and the usual absence of associated pelvic endometriosis suggest implantation as the most likely pathogenetic mechanism. The condition may be an incidental finding in an asymptomatic patient or be associated with premenstrual or postcoital spotting or menorrhagia. The solitary or multiple lesions typically involve the ectocervix; endocervical lesions have been described only rarely. The endometriotic foci appear as friable, ecchymotic streaks, patches, nodules, or cysts measuring from 1 mm to 2 cm in diameter. Rare lesions have been puckered secondary to fibrosis within the lesion, or papillary, simulating a carcinoma. In patients who have had a recent cone biopsy or extensive cautery, the entire transformation zone may be involved (Ismail 1991). Before menses, the lesions typically enlarge and change from bright red to blue; during menses they may rupture, leaving an irregular ulcer. Because a punch biopsy may yield nondiagnostic tissue due to the size of the lesion (which is frequently small), tissue crushing, and fragmentation, aspiration cytology may be useful in establishing the diagnosis. Cervical endometriosis may be the source of abnormal gland cells identified on cervicovaginal smears (Szyfelbein et al. 2004). On histologic examination, the endometriotic focus is usually confined to the superficial lamina propria (Fig. 54). The diagnosis can be missed when the endometriotic stromal component is

808

J. A. Irving and P. B. Clement

Fig. 55 Superficial endometriosis of the uterine cervix. The endometriotic glands show cellular stratification and mitotic figures. If the scanty endometriotic stroma and histiocytes had not been appreciated, these glands may have been misinterpreted as endocervical glandular adenocarcinoma in situ

Fig. 54 Superficial cervical endometriosis. Endometriotic glands and surrounding stroma lie beneath the squamous epithelium

sparse or obscured by edema, hemorrhage, or inflammatory cells (Baker et al. 1999). In such cases, the endometriotic glands, particularly when they show atypia or mitotic activity, can be misinterpreted as endocervical glandular dysplasia, adenocarcinoma in situ, or even invasive adenocarcinoma (Fig. 55). As previously noted, only endometrial stroma (stromal endometriosis) is found in occasional cases of superficial cervical endometriosis, even after serial sectioning (Fig. 56) (Clement et al. 1990). In contrast to superficial cervical endometriosis, deep cervical endometriosis is usually an extension of cul-de-sac involvement in association with more widespread pelvic endometriosis. It may be palpable as deep, firm nodules or cysts in the posterior wall of the cervix (Gardner 1966). The diagnosis is made by biopsy or pathologic examination of the hysterectomy specimen. The differential diagnosis includes downgrowth of adenomyosis from the uterine corpus. Superficial vaginal endometriosis, which typically involves the vault, is rarer than cervical

Fig. 56 Stromal endometriosis of uterine cervix. A cellular sheet of hemorrhagic, endometriotic stroma lies below the exocervical squamous epithelium

endometriosis but is similar to the latter macroscopically, both in its predilection for involving sites of prior trauma and in its lack of associated pelvic endometriosis (Gardner 1966). Deep vaginal endometriosis is more common, is typically associated with pelvic endometriosis, and appears as nodular or polypoid masses involving the posterior vaginal fornix (Fig. 57) (Gardner 1966). The differential diagnosis of vaginal endometriosis, particularly of the superficial type, includes vaginal adenosis of the tuboendometrial variety; the latter, however, lacks endometrial stroma and the characteristic inflammatory response of endometriosis. Endometriosis of the vulva is discussed

13

Diseases of the Peritoneum

Fig. 57 Polypoid endometriosis of the vagina

in a subsequent section (see section “Cutaneous Endometriosis”). In a study of vaginal endometrioid adenocarcinoma, a strong association with vaginal endometriosis was found, with the latter present in 14 of 18 cases (Staats et al. 2007). As the vagina is a common site for recurrence of endometrial adenocarcinomas, identification of endometriosis is an important observation in establishing a vaginal origin (Clement 2007; Staats et al. 2007).

Tubal Endometriosis The term “tubal endometriosis” has been applied to at least three different unrelated lesions of the fallopian tube. The most common type is serosal or subserosal endometriosis, typically associated with endometriosis elsewhere in the pelvis; the myosalpinx is usually not involved. Endometrial tissue may extend directly from the uterine cornu and replace the mucosa of the interstitial and isthmic portions of the tube in as many as 25% and 10% of women in the general population, respectively (Clement 2007). This finding is considered to represent a normal morphologic variation, although in some cases the ectopic endometrial tissue may give rise to intratubal polyps (David et al. 1981). In occasional cases, the endometrial tissue may occlude the tubal lumen, that is, intraluminal endometriosis (“endometrial colonization”) (Fig. 58); involvement may be bilateral. Intraluminal

809

Fig. 58 Endometriosis (colonization) of the fallopian tube. The tubal lumen is occluded by endometrial glands and stroma. Spaces at the junction of the endometrial tissue and myosalpinx represent dilated lymphatic channels

endometriosis is typically unassociated with endometriosis elsewhere. The disorder accounts for 15–20% of tubal-related infertility; it may also be associated with tubal pregnancy. The third type of endometriosis involving the fallopian tube has been designated postsalpingectomy endometriosis. It occurs in the tip of the proximal tubal stump, typically 1–4 years following tubal ligation (Rock et al. 1981). It is closely related to, and may be associated with, salpingitis isthmica nodosa. The lesion is analogous to uterine adenomyosis, consisting of endometrial glands and stroma extending from the endosalpinx into the myosalpinx and frequently to the serosal surface. Hysterosalpingography or India ink injection of the specimen may show tuboperitoneal fistulous tracts; postligation pregnancies are a rare complication. Postsalpingectomy endometriosis has been documented in 20–50% of tubes examined following ligation. The frequency of this complication is increased with the electrocautery method of ligation, with short proximal stumps, and with increasing postligation intervals.

Intestinal Endometriosis Intestinal involvement has been documented in as many as 37% of patients with endometriosis undergoing laparotomy (Williams and Pratt 1977), although the average frequency appears

810

to be approximately 12%. In the majority of such cases, the involvement is confined to the serosa or subserosa and is unassociated with clinical manifestations referable to the intestinal tract. In contrast, from 0.7% to 2.5% of patients with endometriosis require bowel resection for symptomatic lesions (Prystowsky et al. 1988). In some series, as many as half the patients with symptomatic intestinal endometriosis have no extraintestinal involvement; the endometriotic nature of the intestinal lesions in such cases is more likely to be unrecognized preoperatively or at the time of laparotomy. Misdiagnosis is also common in postmenopausal patients because of a decreased index of suspicion, even though the intestine is one of the more common sites of clinically significant endometriosis in this age group (Kempers et al. 1960). As many as 7% of patients with symptomatic intestinal endometriosis are postmenopausal. Intestinal sites of involvement include, in descending order of frequency, the rectum and sigmoid, the appendix, the terminal ileum, the cecum, and other parts of the large and small bowel, including Meckel’s diverticulum (Yantiss et al. 2001). In one large study (Prystowsky et al. 1988), 15% of patients had more than one site of involvement. The presenting symptoms include, alone or in combination, acute or chronic abdominal pain, diarrhea, constipation, hematochezia, and decrease in stool caliber. Although the frequently catamenial nature of the symptoms may suggest the correct diagnosis, the clinical presentation can mimic acute appendicitis, bowel obstruction due to adhesions or a hernia, a neoplasm, or even inflammatory bowel disease. Endoscopic and radiographic studies typically demonstrate an extramucosal stenosing lesion; endoscopic biopsies are usually of no diagnostic value. Endometriosis of the rectosigmoid area is usually a solitary lesion, involving a segment several centimeters in length, whereas ileal involvement is frequently multifocal and may involve segments of bowel up to 45 cm in length (Yantiss et al. 2001). On gross examination, the segment of bowel is indurated and often angulated by a poorly defined, usually noncircumferential mass;

J. A. Irving and P. B. Clement

the serosal surface may be puckered and covered by adhesions. Sectioning typically reveals a firm, gray-white, solid, mural mass, the bulk of which represents markedly thickened muscularis propria; the latter often has a radiating fanlike appearance. Small cystic spaces containing altered blood may be seen but are uncommon. In contrast to a primary adenocarcinoma, the overlying mucosa is usually intact, despite the high frequency of symptomatic bleeding in some series of patients. However, rare cases of polypoid endometriosis have involved the intestinal mucosa; such lesions can grossly mimic an adenocarcinoma (see Fig. 35) (Jiang et al. 2013; Parker et al. 2004). On microscopic examination of symptomatic intestinal endometriosis, islands of endometriotic tissue are typically scattered throughout the hyperplastic muscularis propria, with or without involvement of other layers (Fig. 59) (Jiang et al. 2013; Yantiss et al. 2001). In one recent study, 15 of 103 patients who underwent bowel resection for deep colorectal endometriosis were found to have microscopic foci of endometriosis present at

Fig. 59 Colonic endometriosis. A nest of endometriotic glands and stroma lie in the muscularis propria

13

Diseases of the Peritoneum

one or both resection margins; this finding had no impact on clinical symptoms after 1 year of postoperative follow-up (Roman et al. 2016). Endometriosis involving the appendix and cecum has a predilection to undergo transition to intestinaltype epithelium, thought to represent intestinal metaplasia or colonization of endometriosis; some cases, particularly when extensive and associated with mucocele and extravasated mucin, may prompt consideration of a ruptured low-grade appendiceal mucinous neoplasm (Fig. 60) (Kim et al. 2013; Misradji et al. 2014). A complication of intestinal endometriosis is perforation, which is usually associated with pregnancy; a marked decidual reaction is typically seen with the endometriotic stroma in such cases. Other complications include volvulus, intussusception, acute appendicitis, appendiceal mucocele, intramural hematoma, and the development of a malignant neoplasm (see below) (Mostoufizadeh and Scully 1980; Yantiss et al. 2000).

Urinary Tract Endometriosis Urinary tract involvement has been documented at laparotomy from 16% to 20% of patients with endometriosis (Redwine 1987; Williams and Pratt 1977). In most of these cases, the endometriosis is found on the serosa of the urinary bladder or that overlying the ureter and is without local clinical manifestations. Similarly, high-volume

Fig. 60 Appendiceal endometriosis with replacement by intestinal-type mucinous epithelium

811

intravenous urography has demonstrated subtle, clinically insignificant abnormalities in 15% of women with proven pelvic endometriosis before therapy. In contrast, only 0.5–1% of patients with endometriosis have clinically significant urinary tract involvement; approximately 30% of such patients ultimately require nephrectomy for a hydronephrotic or nonfunctioning kidney. Most reported cases of urinary tract endometriosis have involved the ureters or urinary bladder (with approximately equal frequency), although in one large recent study, 95% and 14% of patients had ureteral and bladder involvement, respectively (Knabben et al. 2015). The kidneys and urethra are involved much less commonly. Urinary tract involvement is usually associated with endometriosis elsewhere in the pelvis, although the symptoms relating to the urinary tract may be the initial or sole manifestations of the disease in such patients (Stanley et al. 1965). In some series, however, as many as half the patients with ureteral involvement have disease restricted to the ureter and the adjacent uterosacral ligament (Kane and Drouin 1985). Patients with renal endometriosis typically do not have endometriosis elsewhere, suggesting an embolic, likely blood-borne, origin. One third to one half of the affected patients are over 40 years of age, and almost 5% of the patients are postmenopausal, some of whom had received estrogen replacement therapy. A preoperative diagnosis may be suspected by the catamenial nature of the symptoms, which include suprapubic or flank pain, frequency, urgency, dysuria, and hematuria; chills and fever secondary to a urinary tract infection have been the presenting symptoms in occasional cases. A tender suprapubic or flank mass may be palpable. Many patients, however, particularly those with ureteric involvement, have nonspecific manifestations or present with a silent obstructive uropathy, occasionally complicated by hypertension, renal failure (in cases of bilateral involvement), or both (Kane and Drouin 1985; Stanley et al. 1965). In patients with bladder involvement, urography may reveal a filling defect; a stricture in the lower ureter with hydroureter and hydronephrosis or a nonfunctioning kidney is the typical urographic finding in those with ureteral

812

involvement. All seven patients with ureteral endometriosis in one study had hydroureter, in most cases accompanied by hydronephrosis, two with superimposed pyelonephritis (Al-Khawaja et al. 2008). Endoscopy may confirm vesical or even ureteral involvement, and the lesions may exhibit catamenial enlargement, darkening, and bleeding. Endoscopy and biopsy, however, are often nondiagnostic (Stanley et al. 1965). Symptomatic endometriosis of the bladder is usually a result of mural involvement, and the lesions are typically located on the trigone, the floor of the bladder, or low on the posterior wall (Stanley et al. 1965). Involvement is rarely confined to the lateral walls, the dome, or the ureterovesical junction. Gross examination typically reveals a solitary, blue, red, gray, or brown multicystic mass that thickens the wall and sometimes projects into the bladder lumen; the lesions have ranged from several millimeters to 14 cm in diameter. The mucosa is usually intact, but occasionally may be ulcerated and bleeding, particularly during menses. Histologic examination reveals fibrosis and proliferation of the muscularis around the foci of endometriosis; the lamina propria was also involved in 60% of the cases in one study (Stanley et al. 1965). Obstruction of both ureteric orifices, vesicocolic fistula, and malignant transformation have been rare complications. With rare exceptions, endometriosis of the ureter is confined to its lower one third, usually involving a segment less than 2 cm in length that lies 2–5 cm from the ureterovesical junction; involvement has been bilateral in approximately 10% of the cases (Al-Khawaja et al. 2008; Stanley et al. 1965). With unilateral ureteric disease, there is predilection for involvement of the left ureter, as identified in 6 of 7 cases in one study (Al-Khawaja et al. 2008), and in 54 of 69 cases in another (Knabben et al. 2015). Ureteral endometriosis has been traditionally divided into extrinsic and intrinsic forms, although this distinction has not been possible in many of the reported cases because the affected segment of the ureter was not removed for microscopic examination. Also, it is likely that at least some intrinsic cases were initially of extrinsic type. In the latter, endometriosis of the uterosacral ligament or ureteral

J. A. Irving and P. B. Clement

Fig. 61 Polypoid endometriotic nodule protruding into the lumen of the ureter

adventitia causes ureteral luminal narrowing by compression, fibrosis, or both; in some such cases, there is transmural scarring of the ureter. Intrinsic involvement is characterized by endometriotic tissue within a typically hyperplastic and fibrotic muscularis; in some cases, the lamina propria is also involved. Mucosal involvement rarely takes the form of a polypoid mass projecting into the lumen (Fig. 61). On gross examination, endometriosis of the kidney is typically a solitary, well-circumscribed, hemorrhagic, solid and cystic mass that focally replaces the renal parenchyma; the lesions in the ten reported cases measured from 1.5 to 13 cm in diameter. In occasional cases, polypoid masses have projected into the renal pelvis. Foci of the smooth muscle have been found admixed with the endometriotic tissue on microscopic examination in some of the cases. Only rare cases of urethral endometriosis have been described, usually involving a urethral diverticulum (Chowdhry et al. 2004).

Cutaneous Endometriosis The majority of the reported cases of cutaneous endometriosis have occurred within surgical scars (Chatterjee 1980; Horton et al. 2008; Kazakov et al. 2007; Minaglia et al. 2007; Steck and Helwig 1966) and rarely within needle tracts or associated with ventriculo- or lumboperitoneal

13

Diseases of the Peritoneum

shunts (Healey and McCluggage 2012); the remainder are spontaneous. Both types are associated with pelvic endometriosis in only a minority of cases (Chatterjee 1980; Steck and Helwig 1966). Because scar-related endometriosis typically occurs after operations on the uterus or fallopian tubes, the site most commonly involved is the lower abdominal wall; the umbilicus is involved less commonly. Similarly, most cases of endometriosis of the lower vagina, vulva, Bartholin’s gland, perineum, and perianal region involve areas of obstetric or surgical trauma, most commonly episiotomy scars (Chatterjee 1980; Gardner 1966; Paull and Tedeschi 1972; Steck and Helwig 1966). The overall frequency of post-Cesarean scar endometriosis was 0.08% in one study, and the authors hypothesized that an increased risk of incisional endometriomas may result from failure to close the parietal and visceral peritoneum with sutures (Minaglia et al. 2007). Scar-related cases occur less commonly after nongynecologic procedures, such as an appendectomy or inguinal hernia repair (Steck and Helwig 1966). Spontaneous cutaneous endometriosis typically involves the umbilicus and, less commonly, the inguinal and perianal regions (Steck and Helwig 1966). The most common symptoms are those relating to a cutaneous mass or nodule that, in the scarrelated cases, appears weeks to years following surgery (Horton et al. 2008; Steck and Helwig 1966); the average postoperative interval from the time of Cesarean section in the study cited above was 3.2 years (Minaglia et al. 2007). In a more recent study of 65 patients with abdominal wall endometriosis, the time from initial surgery (usually Cesarean section) to presentation ranged from 1 to 32 (median 7) years (Ecker et al. 2014). A catamenial increase in size and tenderness, and occasionally bleeding from the lesion, suggest the diagnosis. Patients with perianal lesions may have involvement of the external sphincter producing anorectal pain and irritation simulating an anal fistula, abscess, or thrombosed hemorrhoid. Umbilical endometriosis may simulate an umbilical hernia on physical examination. The lesions occasionally recur following excision; the recurrence rate of 445 cases of abdominal wall

813

endometriosis after surgical excision was 4.3% (Horton et al. 2008). On clinical examination, the lesions are firm, solitary nodules, varying up to 6–12 cm in diameter, and pink to brown to blue-black depending on the age of the lesion and the depth within the skin. The cut surface of the scar-related lesions is typically gray-white, with or without focal areas of recent or old hemorrhage (Chatterjee 1980). On microscopic examination, the endometriosis may involve the dermis (Fig. 62), the subcutis, or both (Steck and Helwig 1966) and, in occasional cases, underlying skeletal muscle. Of note, four cases of reactive skeletal muscle regeneration in association with abdominal wall endometriosis have been described, characterized by a tumor-like proliferation of myoblast-like cells, localized around islands of endometriosis (Colella et al. 2010). Metaplastic glandular and stromal changes may be present, similar to those observed in endometriotic lesions elsewhere, most commonly tubal metaplasia, in addition to oxyphilic, hobnail, mucinous, deciduoid, and papillary syncytial

Fig. 62 Cutaneous endometriosis. Endometriotic foci are present within the dermis

814

metaplasia (Kazakov et al. 2007). In one study, which evaluated 71 cases of cutaneous and superficial soft tissue endometriosis, smooth muscle metaplasia was found in one third of cases, and 25% of lesions showed reactive epithelial atypia (Kazakov et al. 2007). There is typically no continuity between the cutaneous and peritoneal lesions in patients with associated pelvic endometriosis. The association of abdominal scar-related endometriosis and episiotomy scar-related endometriosis with uterine operations and episiotomies, respectively, suggests implantation of endometrial tissue as the most likely pathogenesis. The risk of implantation appears to be much higher after hysterotomy than after Cesarean section or vaginal delivery, suggesting that the decidua of late pregnancy has a reduced ability to implant. When curettage is performed immediately after vaginal term delivery, however, the frequency of endometriosis in the episiotomy scar becomes much higher (Paull and Tedeschi 1972). In nonpregnant patients, implantation of endometrium during endometrial curettage or spontaneous implantation of menstrual endometrium has also been implicated in occasional cases of scar-related endometriosis. Lymphatics have been demonstrated between the pelvis and umbilicus that may explain cases of spontaneous endometriosis in the latter site. Rare cases of clear cell carcinoma arising in abdominal scar-associated endometriosis have been reported (Shalin et al. 2012).

Inguinal Endometriosis Noncutaneous, non-nodal inguinal endometriosis, secondary to involvement of the extraperitoneal portion of the round ligament, occurs in less than 1% of patients with endometriosis (Candiani et al. 1991). The usual presentation is that of a painful, typically right-sided, hernia-like inguinal mass, with catamenial exacerbation in some cases. In approximately one third of the reported cases, an inguinal hernia may also be present. The lesion can impinge on the pubic tubercle and mimic arthritis, bursitis, or tendinitis. Rarely, endometriosis in the inguinal region has also been described

J. A. Irving and P. B. Clement

in inguinal or femoral hernia sacs or the canal of Nuck (Quagliarello et al. 1985). In the largest available series of inguinal endometriosis, the age range was 20–53 (mean 35) years, and the correct diagnosis was suspected preoperatively in 31% of patients; 5 of the 42 patients had a prior history of endometriosis (Mourra et al. 2015). No malignant transformation was found in any of the cases, and one patient developed a recurrence 3 years postoperatively. All four patients who underwent concurrent laparoscopy were found to have ovarian endometriomas. The authors noted that inguinal endometriosis was an incidental finding in 20% of cases and advocated histopathologic examination of hernia sac tissue in women (Mourra et al. 2015).

Endometriosis of the Lymph Nodes Lymph node involvement by endometriosis is uncommon, and many examples reported as such, particularly in the older literature, are lymph nodes involved by benign müllerian (usually endosalpingiotic) glands devoid of an endometrial stromal component. The involved lymph nodes may be visibly or palpably enlarged at operation. On microscopic examination, in contrast to glandular inclusions, endometriotic foci are characterized by a more central location within the node, an endometrial stromal component, and the frequent presence of erythrocytes and pseudoxanthoma cells. Endosalpingiosis and endometriosis may coexist, however, in the same lymph node. As in other sites, decidual transformation of the endometriotic stroma has been encountered during pregnancy. One case of decidualized intranodal endometriosis has been reported in a postmenopausal woman on hormonal replacement therapy (Kim et al. 2015a). As previously noted, one case of intranodal endomyometriosis has been reported.

Pleuropulmonary Endometriosis Pathologically documented cases of endometriosis involving the lungs or pleura are rare. Some

13

Diseases of the Peritoneum

reported examples interpreted as pulmonary endometriosis have taken the form of microscopic foci of “decidua” found at autopsy in pregnant or recently pregnant women. Most such lesions would likely be interpreted by current criteria as foci of embolic intermediate trophoblast, although one case of bona fide deciduosis of the lung has been documented (Flieder et al. 1998). Many cases of purported pleuropulmonary endometriosis have been diagnosed solely on the basis of clinical manifestations or in conjunction with nonspecific histologic or cytologic findings. Coverage here is based on the 38 pathologically documented cases of thoracic endometriosis in the literature, 21 of which were pleural and 17 of which were parenchymal (Flieder et al. 1998), and from a recent retrospective study of 18 patients with histologically confirmed, thoracic endometriosis-related pneumothorax (Ghigna et al. 2015). The affected patients are usually in the reproductive age group, although rare patients are postmenopausal. The clinical manifestations of pleural endometriosis usually differ from those associated with parenchymal involvement. In the former, the characteristic presentation is one of recurrent catamenial shortness of breath related to catamenial pneumothorax, typically rightsided. In the aforementioned study, 18 (7.3%) of 246 women who underwent surgery for spontaneous pneumothorax were found to have thoracic endometriosis (Ghigna et al. 2015). Less common presentations include recurrent right-sided, typically hemorrhagic effusions, hemoptysis, or catamenial pain. Chest X-rays usually reveal a pneumothorax or, occasionally, a hemothorax, a pleural effusion, or a pleural lesion. Coexistent intra-abdominal endometriosis has been demonstrated in approximately one third of cases, although in another one third of cases, its presence or absence was not confirmed. In contrast, patients with parenchymal endometriosis typically present with catamenial hemoptysis or blood-tinged cough; other patients are asymptomatic and the lesion is an incidental radiographic finding. Chest X-ray typically shows a nodule, infiltrates, or opacification of an entire lobe (Flieder et al. 1998). Only one patient has had documented

815

peritoneal endometriosis, although in most patients the peritoneum has not been visualized. The majority of patients with parenchymal endometriosis have had prior uterine operations. Pleural endometriosis is almost invariably confined to the right side; one case with bilateral involvement has been reported. The lesions are typically multiple and dark red or blue nodules or cysts on the diaphragmatic pleura; parietal, visceral, and pericardial pleural surfaces are also affected less commonly. Associated pathologic changes have included diaphragmatic fenestrations in 50% of the cases and occasionally pleural blebs. In half of the pneumothorax-associated cases, the diaphragmatic and pleural lesions were composed of endometriotic stroma only, which was often scant, and recognition was facilitated by immunohistochemical staining for hormone receptors and CD10 (Ghigna et al. 2015). Parenchymal endometriotic lesions are typically solitary, tan to gray, focally hemorrhagic nodules or thin-walled cysts measuring up to 6 cm in diameter. Several lesions have been subpleural or have involved the bronchial walls and lumina. Parenchymal lesions lack the almost exclusively right-sided location of pleural endometriosis; one case had a bilateral miliary distribution. In additional contrast to pleural lesions, associated diaphragmatic fenestrations have not been described. The clinicopathologic differences between pleural endometriosis and parenchymal endometriosis of the lung suggest that they differ in their histogenesis. The distribution of the parenchymal lesions and their strong association with prior uterine trauma strongly suggest an embolic origin. In contrast, most if not all pleural lesions are likely a result of passage of endometriotic tissue from the peritoneal cavity through diaphragmatic defects or diaphragmatic lymphatics, consistent with the right-sided predominance of both structures. The catamenial pneumothorax in these patients, and in those with catamenial pneumothorax unassociated with pleural endometriosis, may be related to the diaphragmatic defects that allow the passage of air from the peritoneal into the pleural cavity. The escape of air from defects in the visceral pleura produced by the endometriotic lesions or from preexistent blebs is another

816

possible explanation for the pneumothorax in these patients. It has been suggested that prostaglandins produced by eutopic endometrium or endometriotic tissue at the time of the menses may predispose to alveolar rupture.

Soft Tissue and Skeletal Endometriosis Rarely, typical endometriomas have occurred in skeletal muscle or deep soft tissues in distant sites. The presentation is usually that of a mass associated with catamenial pain, tenderness, and enlargement. The involved sites have included the trapezius, extensor carpi radialis, thumb, biceps femoris, thigh, and the knee. A unique endometrioma occurred in the breast of a patient with a 2-year history of catamenial bloody nipple discharge (Moloshok and Ivanko 1984). Rare pelvic endometriotic cysts have eroded lumbar vertebrae, causing catamenial lumbar pain.

Upper Abdominal Endometriosis Endometriotic implants may occasionally occur on the omentum; omental endometriosis was only one eighth as common as omental endosalpingiosis in one study (Zinsser and Wheeler 1982). Rarely, endometriotic implants may involve the peritoneal surfaces of the liver or the diaphragm. As with pleural diaphragmatic involvement, implants on the peritoneal side of the diaphragm have occasionally been associated with diaphragmatic defects and catamenial pneumothorax. Rare endometriomas of the epigastrium, the tail of the pancreas, and the liver parenchyma have been reported.

Endometriosis of the Nervous System Based on a recent comprehensive literature review of 378 cases of neural involvement in endometriosis, 97% involved the peripheral nervous system, most frequently the sacral plexus and sciatic nerve, of which the vast majority presented with catamenial sciatica (Siquara de Sousa et al. 2015).

J. A. Irving and P. B. Clement

Some cases have been associated with a visible peritoneal evagination attached to the involved portion of the nerve (“pocket sign”). Thirteen cases of central nervous system involvement have been reported, most involving the conus medullaris or cauda equina, and associated with catamenial back pain or lower extremity weakness and paresthesia. Both patients with cerebral (frontal or parietal lobe) endometriomas and one patient with gait disturbance from cerebellar involvement presented with headache; one developed subarachnoid hemorrhage and the other a generalized seizure (Siquara de Sousa et al. 2015).

Endometriosis in Males Rare examples of endometriosis have been described in men receiving long-term estrogen therapy for prostatic carcinoma. With the exception of one case involving the abdominal wall (Martin and Hauck 1985), the sites of involvement have been confined to the genitourinary tract, specifically the urinary bladder, prostate, and paratesticular region (Young and Scully 1986). The two paratesticular lesions were endomyometriotic in composition.

Neoplasms Arising from Endometriosis One study evaluating consecutive cases of endometriosis found that a malignant tumor was associated with ovarian and pelvic endometriosis in 4% and 10% of cases, respectively (Stern et al. 2001). However, exact frequencies of malignancy arising from pelvic endometriosis in the general population are not known, as some tumors likely overgrow and obliterate the endometriotic foci from which they arose (Mostoufizadeh and Scully 1980). Coexistence of endometriosis and a müllerian-type tumor is not definitive evidence that the tumor has arisen from the endometriosis, unless merging of the two lesions is histologically identified. In most cases, the term “endometriosisassociated” tumor is preferable. For stage I epithelial ovarian cancers, up to 30% have associated ovarian endometriosis, with an even higher

13

Diseases of the Peritoneum

frequency for endometrioid and clear cell carcinomas. Studies to determine putative precursor lesions in such cases have shown that hyperplastic changes (“atypical endometriosis,” discussed above), similar to those that arise in eutopic endometrium, may occur in endometriotic lesions. Such morphologic findings may be observed in the setting of endogenous or exogenous estrogenic stimuli or tamoxifen therapy (see Fig. 46). Atypical ovarian endometriosis was found in approximately 60% of endometriosis-associated carcinomas, but in only 2% of cases of ovarian endometriosis not associated with carcinoma (Fukunaga et al. 1997). Some endometriotic lesions, including atypical endometriosis, and the synchronous carcinoma share similar molecular genetic alterations, including phosphatase and tensin homolog (PTEN), PIK3CA, and AT-rich interaction domain 1A (ARID1A) gene mutations, loss of heterozygosity (LOH), and overexpression of p53 (Akahane et al. 2007; Ayhan et al. 2012; Matsumoto et al. 2015; Sato et al. 2000). Of note, in a recent intriguing study, somatic mutations were detected in glandular epithelium from deep infiltrating endometriosis, without associated malignancy, in 19 of 24 (79%) cases, and of these, 5 harbored cancer driver mutations (including ARID1A, PIK3CA, KRAS, and PPP2R1A) (Anglesio et al. 2017). Molecular alterations in endometriosisassociated neoplasms have been reviewed in detail elsewhere (Lu et al. 2015; Maeda and Shih 2013; Wei et al. 2011) and are briefly summarized here. Immunohistochemical loss of ARID1A, a tumor suppressor gene, has been identified in tumor cells and contiguous endometriotic epithelium in two thirds of ovarian endometrioid and clear cell carcinoma cases (Ayhan et al. 2012). In addition, ARID1A mutations have been identified in 46% and 30% of endometrioid and clear cell carcinomas, respectively, in correlation with loss of BAF250a expression (Wiegand et al. 2010). The latter finding was also demonstrated in clear cell carcinoma and adjacent atypical endometriosis, with concurrent upregulation of hepatocyte nuclear factor-1β and loss of ER and PR (Kato et al. 2006; Xiao et al. 2012). Other stepwise changes that have been identified in endometriotic

817

epithelium and contiguous clear cell carcinoma include overexpression of Skp2, a cell cycle regulator, and elevation of Ki67 proliferative index (Yamamoto et al. 2010). Stepwise decreases in levels of LINE-1 methylation, expression of DNA mismatch repair (MMR) proteins, and microsatellite instability have been observed in endometriosis and the associated ovarian carcinoma (Fuseya et al. 2012; Senthong et al. 2014). Furthermore, one study (Lu et al. 2012) has proposed that selective screening for Lynch syndrome may be justified, as loss of MMR protein expression was also found in 10% of patients with endometriosis-associated ovarian carcinomas. Mutations in exon 3 of the β-catenin (CTNNB1) gene have been found in 60% and 73% of ovarian endometrioid carcinomas and associated atypical endometriosis, respectively, whereas PIK3CA mutations were detected in approximately one third of ovarian endometrioid and clear cell carcinomas (Matsumoto et al. 2015). From the foregoing, it is apparent that endometriosis-associated ovarian clear cell and endometrioid carcinoma share at least some molecular genetic alterations, but a mutually exclusive, histotype-specific genetic profile has not yet been elucidated. It has recently been proposed that ovarian endometrioid carcinoma may arise from a secretory cell precursor, whereas those of clear cell type may be derived from ciliated cells, based on highly differential tumor expression of secretory and ciliated cell markers, shared with eutopic and ectopic endometrium (Cochrane et al. 2017). It has been shown that women with carcinomas arising in endometriosis tend to be younger (and premenopausal), obese, and have a history of unopposed estrogens, in comparison to women with uncomplicated endometriosis (Zanetta et al. 2000). Furthermore, endometriosis-associated tumors are more often lower grade and lower stage; some studies have demonstrated a better prognosis than similar tumors without associated endometriosis (Erzen et al. 2001), but others have found no significant survival difference independent of stage (Noli et al. 2013). Approximately 75% of tumors complicating endometriosis arise within the ovary. The most common extraovarian

818

J. A. Irving and P. B. Clement

site is the rectovaginal septum; less frequent sites include the vagina, colon and rectum (Yantiss et al. 2000), urinary bladder, and other sites in the pelvis and abdomen. In some cases, there is a history of prolonged unopposed estrogen replacement therapy (Yantiss et al. 2000). As previously noted, hyperplastic and metaplastic changes within the endometriosis may precede or be found synchronously with the neoplasm. Tumors arising in endometriosis in unusual sites are more likely to be misdiagnosed than similar tumors arising in ovarian endometriosis, such as an endometrioid adenocarcinoma arising in colonic endometriosis being mistaken for a primary colonic adenocarcinoma (see below), an error that could result in inappropriate staging and treatment (Yantiss et al. 2000). Endometrioid carcinoma (Fig. 63) is the most common tumor arising within ovarian endometriosis, accounting for almost 75% of such cases. Direct origin of endometrioid carcinoma from endometriotic tissue has been demonstrated in as many as 24% of cases in some series (Mostoufizadeh and Scully 1980). At least 90% of the carcinomas arising from extraovarian endometriosis have been of endometrioid type (Mostoufizadeh and Scully 1980). Rarely, endometrioid tumors arising in ovarian and extraovarian endometriosis may exhibit a benign or borderline adenofibromatous pattern (Yantiss et al. 2000). Endometrioid carcinoma is the most

common tumor to arise in intestinal endometriosis; the majority of endometriosis-associated intestinal tumors occur in the rectosigmoid colon, with most of the remainder in the ileum and the cecum (Clement 2007; Petersen et al. 2002; Slavin et al. 2000). The following features are in favor of endometrioid adenocarcinoma over a primary colonic adenocarcinoma: atypical gross features, the presence of endometriosis, an absence of mucosal involvement, lower grade nuclei than expected for colonic adenocarcinoma, squamous differentiation, an absence of dirty necrosis, and a cytokeratin 7+/cytokeratin 20-/ CDX2 immunoprofile (Clement 2007; Kelly et al. 2008; Slavin et al. 2000). Other tumors arising in intestinal endometriosis include endometrioid stromal sarcoma, müllerian adenosarcoma, carcinosarcoma, clear cell carcinoma, squamous cell carcinoma, and mixed germ cell tumor (Clement 2007). Clear cell carcinoma (Figs. 64 and 65) is the second most common tumor originating in endometriosis, accounting for approximately 15% of such cases. In most studies, the frequency of endometriosis coexisting with clear cell carcinoma of the ovary is even higher than with endometrioid carcinoma (Wei et al. 2011). A few examples of clear cell carcinoma arising within extraovarian endometriosis have also been described (Ahn and Scully 1991; Hitti et al. 1990). One study, which utilized the PCR

Fig. 63 Endometrioid carcinoma arising within the endometriosis. Benign endometriotic glands (left) with adjacent carcinomatous glands (right)

Fig. 64 Clear cell carcinoma arising within an endometriotic cyst. Fleshy pale tumor nodules protrude into the cyst lumen

13

Diseases of the Peritoneum

Fig. 65 Clear cell carcinoma arising within an endometriotic cyst. (a) Low-power view showing nodules of clear cell carcinoma protruding into the cyst lumen. (b)

and LOH analysis using laser-microdissected tumor tissue, has indicated that ovarian clear cell adenofibroma may be a clonal precursor for clear cell carcinoma, shown by high concordance rates of allelic patterns between clear cell carcinoma and benign and borderline clear cell adenofibromatous components, with 95% of cases showing an identical LOH pattern (Yamamoto et al. 2008). Patients with endometriosisassociated clear cell carcinomas of the ovary have also recently been found to have improved progression-free and overall survival rates in comparison to those without endometriosis (Orezzoli et al. 2008). Ovarian and extraovarian epithelial tumors of other types arising from endometriosis include endometrioid adenofibromas and atypical proliferative/borderline tumors of endocervical-like mucinous (seromucinous) and mixed cell types, as well as serous atypical proliferative/borderline tumors and squamous cell carcinomas (Naresh et al. 1991; Rutgers and Scully 1988a, b; Yantiss et al. 2000). Endometrioid stromal sarcomas (ESSs), carcinosarcomas (malignant mixed müllerian tumors), and adenosarcomas (both typical and with sarcomatous overgrowth) (Fig. 66) account for approximately 10% and 20% of tumors arising in ovarian and extraovarian endometriosis, respectively (Clement and Scully 1978; Yantiss et al. 2000; Young et al. 1984). Approximately one quarter of tumors arising in colonic

819

High-power view of a different case showing clear cell carcinoma with background pigmented histiocytes

Fig. 66 Müllerian adenosarcoma arising in ovarian endometriosis. Low-grade sarcomatous stroma forms periglandular cuffs and intraglandular papillae

endometriosis are adenosarcomas (Slavin et al. 2000; Yantiss et al. 2001). In one study, 60% of ESSs apparently arising within the ovary were associated with ovarian endometriosis (Young et al. 1984). In one large study, features of extrauterine ESS, including unusual location and atypical histologic features, contributed to misdiagnosis in one quarter of cases (Masand et al. 2013). In the same study, follow-up was available in 53 patients, and recurrences developed in almost two thirds; 15 patients were alive with disease, and 9 died of disease. Of six cases of primary extrauterine ESS, only one harbored a JAZF1–JJAZ1 fusion transcript, suggesting

820

that this genetic aberration occurs less commonly than in ESS (Amador-Ortiz et al. 2011). Rare examples of yolk sac tumor have arisen in association with endometriosis (Rutgers et al. 1987), and in one unique case, a sex cord tumor with annular tubules was intimately associated with endometriosis of the tubal serosa (Griffith and Carcangiu 1991).

Peritoneal Endometrioid Lesions Other Than Endometriosis Benign glands lined by endometrial epithelium (but lacking endometrial stroma) with the peritoneal distribution of endosalpingiosis occasionally occur (Lauchlan 1972); some may represent foci of endometriosis in which the stromal component has undergone atrophy. Benign endometrioid peritoneal “implants” lacking an endometrial stromal component have also been reported in association with an atypical proliferative/borderline ovarian endometrioid tumor (Russell 1979). The peritoneal lesions were interpreted as having arisen directly from the peritoneum. A variety of extrauterine, extraovarian, pelvic, or retroperitoneal neoplasms of endometrioid type occur in the absence of demonstrable endometriosis. These tumors have generally been considered to arise directly from the mesothelium or submesothelial stroma, or possibly from foci of endometriosis that have been obliterated by the tumor. They have included endometrioid cystadenofibroma and cystadenocarcinoma, endometrioid stromal sarcoma, homologous and heterologous types of carcinosarcoma (malignant mixed Mullerian tumor), and Mullerian adenosarcoma.

Peritoneal Serous Lesions Serous lesions of the peritoneum include those that are nonneoplastic (endosalpingiosis) and neoplastic, which are morphologically analogous to their ovarian counterparts.

J. A. Irving and P. B. Clement

Endosalpingiosis Clinical Findings Endosalpingiosis typically refers to the presence of benign glands lined by tubal-type epithelium involving the peritoneum and subperitoneal tissues; the term may also be used to refer to similar glands within retroperitoneal lymph nodes (see section “Benign Intranodal Glands of Müllerian Type”). This disorder occurs almost exclusively in females, typically during their reproductive years, with a mean age of 29.7 years in one study (Zinsser and Wheeler 1982), although occasional cases have been described in postmenopausal women. Endosalpingiosis is almost always an incidental finding at either the time of operation or more commonly on microscopic examination. In a retrospective study, endosalpingiosis was found in 12.5% of surgically removed omenta, but this figure doubled when omenta were examined more thoroughly in a prospective study by these same investigators (Zinsser and Wheeler 1982). Endosalpingiosis may be detected as multiple fine pelvic calcifications on X-ray examination or as psammoma bodies within cul-de-sac fluid, peritoneal washings (Sidaway and Silverberg 1987), the lumen of the fallopian tube, or cervical Papanicolaou smears (Kern 1991). An origin from the secondary müllerian system is favored by most investigators, but the association of endosalpingiosis with chronic salpingitis implicates implantation of sloughed tubal epithelium as a possible histogenetic mechanism in some cases (Zinsser and Wheeler 1982). A similar association with serous atypical proliferative/borderline tumors suggests that some endosalpingiotic foci may represent tumor implants that have undergone maturation (Vang et al. 2013). Intralymphatic spread of tubal epithelial cells has also been proposed (Russell et al. 2016). Endosalpingiosis in the absence of residual tumor at the time of secondlook laparotomy in patients with ovarian epithelial neoplasms does not justify additional treatment (Copeland et al. 1988). Pathologic Findings Endosalpingiosis is most commonly encountered on the pelvic peritoneum covering the uterus,

13

Diseases of the Peritoneum

821

fallopian tubes, ovaries, and cul-de-sac (Zinsser and Wheeler 1982). Less frequent sites include the pelvic parietal peritoneum, omentum, bladder and bowel serosa, para-aortic area, and skin, including laparotomy scars. Endosalpingiosis is usually inapparent at the time of operation or on gross inspection of the involved tissues but may be visible as multiple, punctate (1–2 mm), white to yellow, opaque or translucent, fluid-filled cysts, which impart a vesicular or granular appearance to the involved surface; rarely larger cysts may be seen (Clement and Young 1999). Rare examples of cystic endosalpingiosis have involved the wall of the uterus, resulting in grossly apparent transmural cysts (Clement and Young 1999). Microscopic examination reveals multiple, simple glands, often cystically dilated and lined by a single layer of epithelium resembling that of the normal fallopian tube (Figs. 67 and 68). The glands are frequently surrounded by a loose or dense connective tissue stroma that may contain a sparse mononuclear inflammatory cell infiltrate. The glands may exhibit irregular contours, crowding, and intraluminal stromal papillae. The three cell types of the normal fallopian tube epithelium are found in varying numbers: pale ciliated cells, secretory cells, and dark rodlike, intercalated, or “peg” cells. The cells have prominent luminal margins, distinct borders, and basal nuclei. Focal cellular pseudostratification may be present. The nuclei have fine chromatin and delicate nuclear membranes and typically lack

significant atypia or mitotic activity. Psammoma bodies are frequently present within the lumens or in the adjacent stroma, and, in occasional cases, numerous psammoma bodies are embedded in subserosal connective tissue. Perineural infiltration can be a rare finding (Satgunaseelan et al. 2016). Endosalpingiotic glands can rarely extend into the underlying tissues, such as the wall of the appendix or, as noted earlier, the uterus (Clement and Young 1999). Endosalpingiotic epithelium exhibits positive immunohistochemical staining for PAX8, WT-1, as well as ER and PR (Carney et al. 2014; Esselen et al. 2014). The term atypical endosalpingiosis has been applied to endosalpingiotic lesions in which there is cellular stratification, including cellular buds, cribriform patterns, and varying degrees of cellular atypia, occurring in the absence of a serous atypical proliferative/borderline tumor (SBT). Such lesions may also merge histologically with peritoneal SBT (see next section). Bell and Scully (1990) use the latter term if the “lesions composed of tubal-type epithelium exhibit papillarity, tufting, or detachment of cell clusters.. .. even when they arise on a background of endosalpingiosis.” Endosalpingiotic glands should be differentiated from mesonephric remnants, which are common incidental microscopic findings in the region of the fallopian tube. Mesonephric tubules are typically located more deeply than endosalpingiosis and characteristically have a collar of smooth muscle under the epithelial lining,

Fig. 67 Endosalpingiosis. Complex glandular structure lies beneath the uterine serosa. Glands are lined by a single layer of benign endosalpingeal epithelium

Fig. 68 Endosalpingiosis. Glands within the omentum are lined by benign endosalpingeal epithelium

822

which is typically a single layer of nonciliated, low columnar to cuboidal cells. Rare extraovarian atypical proliferative/borderline and malignant serous tumors have been shown to arise from endosalpingiosis (Carrick et al. 2003; McCoubrey et al. 2005).

Peritoneal Serous Tumors The full spectrum of serous neoplasms arising within the ovary may also arise directly from the extraovarian peritoneum. These tumors are considered only briefly here because their clinicopathologic features closely resemble those of their tubal and ovarian counterparts. Primary peritoneal serous atypical proliferative/borderline tumors are usually associated with widespread extraovarian peritoneal involvement and normalsized ovaries that are free of disease or which have only very minimal surface involvement (Bell and Scully 1990; Biscotti and Hart 1992). The most common presenting features in patients with these tumors, who are typically under the age of 35 years (range, 16–67), are infertility and chronic pelvic or abdominal pain. Many cases, however, are discovered incidentally at laparotomy for other conditions. At operation, focal or diffuse miliary granules, fibrous adhesions, or both involve the pelvic peritoneum and omentum and, less commonly, the abdominal peritoneum.

J. A. Irving and P. B. Clement

Microscopic examination reveals superficial tumor that resembles noninvasive epithelial or desmoplastic implants of borderline serous tumors of ovarian origin. Coexistent endosalpingiosis has been found in 85% of cases. The prognosis of peritoneal SBT is favorable; approximately 85% of patients have had no clinically persistent or progressive disease on follow-up. In rare cases, transformation to an invasive low-grade peritoneal serous carcinoma (LGPSC) may occur, although, in a proportion of these, the invasive tumor may have been present but not sampled at the time of the initial operation. Although most primary peritoneal serous carcinomas are high grade (see following paragraph), some have low-grade nuclear features and are distinguished from peritoneal SBT by the presence of invasion. LGPSCs resemble invasive implants of SBT (Fig. 69) (Weir et al. 1998). Some may have a micropapillary pattern (Elmore et al. 2000). They lack high-grade nuclear atypia, invade tissue or lymphovascular spaces or both, and have appreciable solid epithelial proliferation. Peritoneal psammocarcinomas (Gilks et al. 1990; Weir et al. 1998) are a subtype of LGPSCs with psammoma bodies in most of the tumor nests and absent or rare solid epithelial proliferation (see ▶ Chap. 14, “Epithelial Tumors of the Ovary”); lymphatic invasion is often conspicuous. The average ages of patients in one study were

Fig. 69 Low-grade primary papillary serous carcinoma of the peritoneum. (a) The tumor is infiltrating the omental fat. (b) The papillae are lined by tumor cells resembling serous borderline neoplasia. Note the psammoma bodies

13

Diseases of the Peritoneum

57 years (LGPSC of usual type) and 40 years (peritoneal psammocarcinomas) (Weir et al. 1998). Features present in both tumors are usually abdominal pain, mass, or both, but approximately 40% are incidental findings. Operative and gross findings vary from nodules to adhesions to a dominant mass. In the short-term, outcomes for LGPSCs and peritoneal psammocarcinomas are favorable. In two related studies of patients with stage II–IV ovarian and peritoneal low-grade serous carcinoma who underwent primary cytoreductive surgery and platinum-based chemotherapy, the median progression-free survival was 28.1 months, the overall median survival was 104.7 (range, 75.1–134.2) months, and hormonal maintenance therapy was associated with longer progression-free survival than observation alone (Gersheron et al. 2015, 2017). LGPSC has also been associated with a lower risk of progression and death from disease than those of primary ovarian origin (Gersheron et al. 2015, 2017). These tumors should be distinguished from primary peritoneal SBT and noninvasive implants of ovarian atypical proliferative/borderline serous tumors, which have similar features but lack invasion. Adequate sampling is necessary to identify invasion, with highest yields of invasive foci in the omentum. In equivocal cases which show features that are of concern for LGPSC, a descriptive designation of low-grade serous proliferation may be considered, and follow-up with additional sampling may be necessary to establish or refute a definitive diagnosis of LGPSC. Typical peritoneal serous carcinomas have highgrade nuclear features (Ben-Baruch et al. 1996; Truong et al. 1990) with an intraoperative appearance that of widespread peritoneal tumor associated with ovaries of normal size, mimicking a diffuse malignant mesothelioma or peritoneal carcinomatosis associated with an unknown primary tumor. In some series, the patients have had an average age that is a decade older than patients with similar tumors of ovarian origin. Some tumors have occurred in women who had had bilateral oophorectomy performed as prophylactic treatment for BRCA-related familial ovarian cancer (Casey et al. 2005).The risk of peritoneal serous carcinoma in BRCA1 mutation carriers is about 4% at 20 years

823

after prophylactic salpingo-oophorectomy (Casey et al. 2005; Finch et al. 2006). Primary peritoneal serous carcinomas resemble their tubal and ovarian counterparts on microscopic (Fig. 70) and immunohistochemical examination, including positivity for PAX-8, WT-1, and mutational pattern p53 (diffuse positive or null staining) (Euscher et al. 2005; Kobel et al. 2011); their distinction from malignant mesothelioma has been previously discussed. Using recently proposed criteria for assignment of primary site in pelvic high-grade serous carcinomas, tumors with coexistent serous tubal intraepithelial carcinoma, which is usually fimbrial, are now regarded to be of primary fallopian tube origin. Only rare cases are likely considered to be of primary peritoneal origin, with requisite normal or benign findings in the fallopian tubes, ovaries, and endometrium after complete histologic examination (McCluggage et al. 2015a; Seidman et al. 2011; Singh et al. 2016) (see ▶ Chaps. 9, “Endometrial Carcinoma,” and ▶ 11, “Diseases of the Fallopian Tube and Paratubal Region”). The outcome of peritoneal and tubo-ovarian high-grade serous carcinoma has historically been similar; with refined criteria for primary site assignment, primary peritoneal tumors with nodal metastases had a better prognosis than ovarian primaries with secondary peritoneal and nodal spread (Bakkar et al. 2014). Rare extraovarian serous tumors take the form of localized, typically cystic masses, usually within

Fig. 70 Primary peritoneal high-grade serous carcinoma

824

the broad ligament and less commonly within the retroperitoneum. Serous papillary cystadenomas and adenofibromas, SBT, and serous carcinomas have been described in these sites (Aslani et al. 1988; Aslani and Scully 1989; Ulbright et al. 1983).

Endocervicosis (Including Müllerianosis) Benign glands of endocervical type involving the peritoneum, so-called endocervicosis, are rare, but examples involving the posterior uterine serosa, cul-de-sac, vaginal apex, outer wall of the uterine cervix, and the urinary bladder have been documented (Clement and Young 1992; Lauchlan 1972; Martinka et al. 1999; Nazeer et al. 1996; Young and Clement 1996, 2000). In the last site, the lesions usually formed tumor-like masses that involved the posterior wall or posterior dome of the bladder in women of reproductive age. On microscopic examination, benign endocervical-type glands were located predominantly within the smooth muscle of the muscularis propria (Fig. 71) (Clement and Young 1992; Nazeer et al. 1996). In several cases, the infiltrative pattern of the glands, mild epithelial atypia, and a reactive periglandular stroma, alone or in combination, resulted in an initial misdiagnosis of well-differentiated adenocarcinoma. In such cases, the absence of a mucosal-based tumor and

J. A. Irving and P. B. Clement

no more than mild atypia facilitate the diagnosis of endocervicosis. The differential diagnosis also includes müllerianosis, a term applied to lesions composed of an admixture of müllerian glandular epithelia (tubal, endocervical, and endometrioid), sometimes with foci of endometriotic stroma. Examples of müllerianosis have been reported in the urinary bladder, mesosalpinx, and inguinal lymph nodes (Lim et al. 2003; Young and Scully 1986). In exceptional cases, malignant transformation may occur, as evidenced by one case of adenocarcinoma arising in endocervicosis of the urinary bladder (Nakaguro et al. 2016).

Extraovarian Mucinous Tumors Ovarian-type mucinous neoplasms, in the absence of a primary tumor within the ovary, have been described in extraovarian sites, typically in the retroperitoneum (Fig. 72) (de Peralta et al. 1994; Lauchlan 1972; Roma and Malpica 2009); a single case has been described in the inguinal region (Sun et al. 1979). These tumors form large cystic masses that on histologic examination resemble ovarian mucinous cystadenomas, atypical proliferative/borderline tumors, or cystadenocarcinomas (Fig. 73); some tumors contain ovarian-type stroma in their walls. Mural nodules, usually consisting of anaplastic carcinoma, similar to those in ovarian mucinous tumors, have been present in several cases (Mikami et al. 2003). Although it is possible that some of

Fig. 71 Endocervicosis of urinary bladder. (a) Benign endocervical-type glands lie within the muscularis propria. (b) High-power view of glands with cytologically bland, endocervical-type mucinous epithelium

13

Diseases of the Peritoneum

825

Fig. 72 Retroperitoneal mucinous tumor. The specimen has been opened to reveal multiple locules with mucinous contents. (Courtesy of R.E. Scully, M.D., Boston, MA)

Fig. 74 Walthard nests. Multiple small cysts cover the serosa of the fallopian tube and mesosalpinx

Fig. 73 Retroperitoneal mucinous cystadenoma

Fig. 75 Walthard nests on the fallopian tube serosa. A nest (far left) and multiple cysts are formed or lined by benign transitional-type cells

these tumors originate within a supernumerary ovary, the great rarity of the latter, the absence of follicles or their derivatives within the ovarian-like stroma, and the rare occurrence of similar tumors in males strongly support a peritoneal origin. In the largest available study of primary retroperitoneal mucinous tumors, both patients who died of the disease had tumors consisting of carcinoma or sarcoma, but in the absence of these features, the clinical outcome has been favorable (Roma and Malpica 2009).

Peritoneal Transitional, Squamous, Clear Cell, and Non-Epithelial Lesions Nests of transitional (urothelial) epithelium referred to as Walthard nests are commonly present on the pelvic peritoneum in women of all ages, typically

involving the serosal surfaces of the fallopian tubes (Figs. 74 and 75), mesosalpinx, and mesovarium (Bransilver et al. 1974). Walthard nests are uncommon on the ovarian surface but may be seen in the hilus, probably originating from the peritoneum of the mesovarium; they are most common on the tubal serosa (see ▶ Chap. 11, “Diseases of the Fallopian Tube and Paratubal Region”). The nests are immunoreactive for GATA3 and p63 and are typically negative for PAX2 and PAX8 (Esheba et al. 2009; Roma and Masand 2014). Rare extraovarian Brenner tumors have been encountered, most commonly in the broad ligament. In contrast to Walthard nests, squamous metaplasia of the peritoneum is rare; it is usually an incidental microscopic finding, but in one case, it resulted in tiny, but grossly visible, nodules (Mourra et al. 2004). Squamous metaplasia of

826

ovarian and fallopian tube surfaces, secondary to peritoneal dialysis, has also been described (Hosfield et al. 2008). Five clear cell carcinomas of apparent peritoneal origin have been reported, without evidence of endometriosis. Two were abdominopelvic, one was a localized sigmoid mesocolonic mass, one involved the anterior abdominal wall and ileal serosa, and one diffusely involved the peritoneum (Evans et al. 1990; Insabato et al. 2015; Lee et al. 1991; Takano et al. 2009).Rare extragonadal yolk sac tumor and unclassifiable malignant sex cord–stromal tumors of peritoneal origin have also been described (Ravishankar et al. 2017; Shah and McCluggage 2017). Six cases of extraovarian sex cord proliferations have been reported, in the absence of a sex cord–stromal neoplasm, characterized by microscopic involvement of the fallopian tube, paraovarian tissue, pelvic sidewall, and appendiceal serosa, and thought to represent a nonneoplastic proliferation of embryonic remnants (McCluggage et al. 2015b).

Peritoneal Decidual Reaction Clinical and Operative Findings An ectopic decidual reaction similar to that seen in the lamina propria of the fallopian tube, cervix, and vagina may also be seen within the submesothelial stroma of the peritoneal cavity. Frequent sites of ectopic decidua include the submesothelial stroma of the fallopian tubes, uterus and uterine ligaments, appendix and omentum, and within pelvic adhesions. Rare sites have included the serosal surfaces of the diaphragm, liver, spleen, and renal pelvis. Submesothelial decidua is typically an incidental microscopic finding, but florid lesions may be visible at the time of Cesarean section or postpartum tubal ligation as multiple, gray to white, focally hemorrhagic nodules or plaques studding the peritoneal surfaces and simulating metastatic malignancy (Adhikari and Shen 2013). Several cases have been associated with massive, occasionally fatal, intraperitoneal hemorrhage during the third trimester, labor, or the puerperium. Appendiceal deciduosis may also mimic acute appendicitis during pregnancy (Chai and

J. A. Irving and P. B. Clement

Wijesuriya 2016). Other rare clinical presentations include hydronephrosis and hematuria secondary to renal pelvic involvement.

Microscopic Findings Microscopic examination discloses submesothelial decidual cells disposed individually or arranged in nodules or plaques (Fig. 76). Smooth muscle cells, probably derived from submesothelial myofibroblasts, may be admixed. The decidual foci are typically vascular and contain a sprinkling of lymphocytes. Focal hemorrhagic necrosis and varying degrees of nuclear pleomorphism and hyperchromasia of the decidual cells may suggest a tumor such as a deciduoid malignant mesothelioma (Shia et al. 2002), but their bland appearance and mitotic inactivity militate against such a diagnosis. We have seen several cases of an omental decidual reaction in which most of the decidual cells exhibited striking vacuolization with basophilic mucin and an eccentric location of the nucleus. The appearance of the cells raised the possibility of metastatic signet-ring cell carcinoma, but in contrast to the cells of the latter, the vacuoles within the decidual cells contain acid rather than neutral mucin, and their cytoplasm lacks immunoreactivity for cytokeratin.

Diffuse Peritoneal Leiomyomatosis Diffuse peritoneal leiomyomatosis is a rare disorder characterized by the presence of multiple

Fig. 76 Ectopic decidua beneath the pelvic peritoneum. Note the prominent cytoplasmic vacuoles

13

Diseases of the Peritoneum

submesothelial nodules of cytologically benign smooth muscle, frequently associated with uterine leiomyomas and, rarely, ovarian leiomyomas. The nodules are generally considered to arise from multipotential submesothelial mesenchymal cells. This disorder is discussed elsewhere (see ▶ Chap. 10, “Mesenchymal Tumors of the Uterus”).

Benign Intranodal Glands of Müllerian Type Clinical Features Benign glands of müllerian type are most commonly encountered within the pelvic and paraaortic lymph nodes of females (Horn and Bilek 1995; Kheir et al. 1981) and less often in inguinal and femoral lymph nodes. Because these glands are almost always incidental microscopic findings in lymph nodes removed in cases of pelvic carcinoma, their reported frequency, which has varied from 2% to 41%, depends on the number of lymph nodes removed and the extent of the histologic sampling. Almost all the patients have been adults, although rare examples have been reported in children. In males, the presence of similar glands has been recorded rarely within lymph nodes in the pelvis, abdomen, and mediastinum (Gallan and Antic 2016). Although typically without clinical or intraoperative manifestations, rare examples of lymph nodes containing mülleriantype glands have been associated with a falsepositive lymphangiogram, ureteral obstruction secondary to lymph node enlargement, or visible enlargement at the time of operation. In a number of patients, intranodal glandular inclusions have been accompanied by endosalpingiosis of the peritoneum, salpingitis isthmica nodosa, or acute and chronic salpingitis (Kheir et al. 1981). Other patients have had coexistent ovarian serous tumors, which have been benign, atypical proliferative/borderline tumors, or carcinomas (Prade et al. 1995). Pathologic Findings On gross examination, the glands are usually not apparent, although rarely they are recognizable as cysts measuring up to a few millimeters in

827

diameter. The glands are typically located in the periphery of the node, most commonly within its capsule or between the lymphoid follicles in the superficial cortex (Fig. 77); rarely, they lie free within the subcapsular sinuses (Kheir et al. 1981). In florid cases, they can be diffusely distributed throughout the lymph node. Intraglandular or periglandular psammoma bodies are commonly present. Intranodal glands may be surrounded by a thin rim of fibrous tissue or abut directly on the surrounding lymphoid cells. The glands may be round and cystically dilated or exhibit an irregular contour as a result of infolding. They are most commonly lined by a single layer of cuboidal to columnar tubal-type epithelium, with an admixture of ciliated, secretory, and intercalated cell types (Fig. 78). With

Fig. 77 Endosalpingiotic glands within the pelvic lymph node. The glands are located within and immediately beneath the node capsule as well as deeper within the node

Fig. 78 Endosalpingiotic glands within the pelvic lymph node. The glands are lined by benign cells of multiple types, including ciliated cells

828

Fig. 79 Atypical endosalpingiotic glands within the pelvic lymph node. Some of the glands exhibit luminal obliteration by cells growing in solid and cribriform patterns

special stains, mucin can be demonstrated in the apical portion of the secretory cells and within the gland spaces. The cells have a benign appearance with regular, basally oriented or pseudostratified, oval to round nuclei, fine nuclear chromatin, and occasional small nucleoli; mitotic figures are typically absent. In rare cases, the cells can exhibit varying degrees of atypia and stratification; the latter can produce an intraglandular cribriform pattern or luminal obliteration by sheets of cells (Fig. 79). These cases of atypical endosalpingiosis may rarely be the origin of intranodal serous neoplasms (see following). Examples of intranodal glandular inclusions lined by benign endometrioid epithelium, mucinous epithelium of endocervical or goblet cell type, or metaplastic squamous epithelium have been reported (Lauchlan 1972; Mills 1983).

Differential Diagnosis In most cases the distinction between glandular inclusions and metastatic adenocarcinoma is not difficult unless a primary ovarian SBT is present, in which case the distinction may be difficult or impossible. Features favoring a benign diagnosis include a capsular or interfollicular location of the glands, lining cells of multiple types including ciliated forms, a lack of significant cellular atypia and mitotic activity, and an absence of a

J. A. Irving and P. B. Clement

desmoplastic stromal reaction. Complicating the differential diagnosis is the very rare development of atypical proliferative/borderline or frankly malignant change in müllerian glandular inclusions in the lymph nodes. This diagnosis is suggested in cases in which the intranodal neoplasm merges with the foci of atypical endosalpingiosis. Awareness that ER-positive endosalpingiosis can occur in axillary lymph nodes will avoid the potential misdiagnosis of metastatic breast carcinoma (Corben et al. 2010). Positivity for PAX8, WT-1, and hormone receptors has also been demonstrated in rare cases of nodal endosalpingiosis identified in men who underwent pelvic lymphadenectomy for prostatic or urothelial carcinoma (Gallan and Antic 2016). Intranodal nests of benign squamous epithelium should not be mistaken for metastatic squamous cell carcinoma. Features favoring a benign diagnosis include bland cytologic features, lack of mitotic activity, and, in some cases, an origin within benign glands.

Intranodal Ectopic Decidua Ectopic decidua unassociated with endometriosis has been described as a rare, incidental microscopic finding in para-aortic and pelvic lymph nodes, usually removed as part of a radical hysterectomy for carcinoma of the cervix in pregnant patients (Ashraf et al. 1984; Mills 1983; Wu et al. 2005). A subserosal ectopic decidual reaction may be present elsewhere in the pelvis. In some cases, the decidual tissue has been recognized on careful macroscopic examination as tiny, gray, subcapsular nodules. On microscopic examination, the decidual nests typically occupy the subcapsular sinus and superficial cortex (Fig. 80), although more central parts of the lymph node may also be involved. The cells appear benign, but may contain occasional bizarre, hyperchromatic nuclei, mimicking metastatic squamous cell carcinoma. The absence of mitotic activity, keratinization, and stromal desmoplasia facilitate the diagnosis. Metastatic squamous cell carcinoma, however, may be present in the same node.

13

Diseases of the Peritoneum

Fig. 80 Ectopic decidua within the pelvic lymph node. The nodal architecture is focally replaced by sheets of decidualized cells

Intranodal Leiomyomatosis Rare cases of lymph node involvement by mitotically inactive, cytologically benign smooth muscle have been described (Fig. 81) (Cramer et al. 1980; Hsu et al. 1981). Most patients have had concurrent typical uterine leiomyomas or, less commonly, diffuse peritoneal leiomyomatosis (Hsu et al. 1981) or similar nodules within the lungs (Cramer et al. 1980). In pregnant patients, the process may merge with intranodal decidua (Hsu et al. 1981). The finding in most cases likely is secondary to lymphatic spread from uterine leiomyomas (“benign metastasizing leiomyoma”; see ▶ Chap. 10, “Mesenchymal Tumors of the Uterus”), but in some cases the intranodal smooth muscle may arise from entrapped subcoelomic mesenchyme or myofibroblastic organization of intranodal decidua. The presence of benignappearing smooth muscle in a lymph node should also bring into consideration the diagnosis of lymphangioleiomyomatosis (LAM). This disorder is usually, but not invariably, associated with pulmonary involvement. However, in a recent study of 19 women with incidental nodal LAM, detected in pelvic and/or para-aortic lymph nodes (removed predominantly for surgical staging of ovarian or uterine carcinomas), none were associated with pulmonary LAM. The authors suggested that size of nodal LAM is a prognostic factor, as all 18 patients with small nodal lesions (less than 10 mm) had no recurrences, whereas the

829

Fig. 81 Benign smooth muscle within the pelvic lymph node

single patient with bulky nodal LAM developed persistent local nodal disease (Schoolmeester and Park 2015). Diffusely positive, cytoplasmic immunohistochemical staining for beta-catenin was present in all 18 cases tested in this study, supporting the use of this marker to facilitate the diagnosis; HMB45 is also typically positive but may be focal (Schoolmeester and Park 2015). Benign intranodal smooth muscle should also be distinguished from metastatic well-differentiated leiomyosarcoma of uterine origin. Patients with the latter usually have a large uterine mass, and on histologic examination the intranodal tumor is cellular and exhibits evidence of cellular atypicality and mitotic activity.

References Abrao MS, Dias JA Jr, Podgaec S et al (2006) Bowel endometriosis and schistosomiasis: a rare but possible association. Fertil Steril 85:1060.e1–1060.e2 Adams HW, Mainz DL (1977) Eosinophilic ascites. A case report and review of the literature. Dig Dis 22:40–42 Adhikari LJ, Shen R (2013) Florid diffuse peritoneal deciduosis mimicking carcinomatosis in a primigravida patient: a case report and review of the literature. Int J Clin Exp Pathol 6:2615–2619 Ahn GH, Scully RE (1991) Clear cell carcinoma of the inguinal region arising from endometriosis. Cancer (Phila) 67:116–120 Akahane T, Sekizawa A, Purwosunu Y et al (2007) The role of p53 mutation in the carcinomas arising from endometriosis. Int J Gynecol Pathol 26:345–351 Albee RB Jr, Sinervo K, Fisher DT (2008) Laparoscopic excision of lesions suggestive of endometriosis or

830 otherwise atypical in appearance: relationship between visual findings and final histologic diagnosis. J Minim Invasive Gynecol 15:32–37 Alexander HR Jr, Bartlett DL, Pingpank JF et al (2013) Treatment factors associated with long-term survival after cytoreductive surgery and regional chemotherapy for patients with malignant peritoneal mesothelioma. Surgery 153:779–786 Al-Khawaja M, Tan PH, MacLennan GT et al (2008) Ureteral endometriosis: clinicopathological and immunohistochemical study of 7 cases. Hum Pathol 39: 954–959 Amador-Ortiz C, Roma AA, Huettner PC et al (2011) JAZF1 and JJAZ1 gene fusion in primary extrauterine endometrial stromal sarcoma. Hum Pathol 42:939–946 Andrici J, Jung J, Sheen A et al (2016) Loss of BAP1 expression is very rare in peritoneal and gynecologic serous adenocarcinomas and can be useful in the differential diagnosis with abdominal mesothelioma. Hum Pathol 51:9–15 Anglesio MS, Papadopoulos N, Ayhan A et al (2017) Cancer-associated mutations in endometriosis without cancer. N Engl J Med 376:1835–1848 Arnold MA, Schoenfield L, Limketkai RB et al (2014) Diagnostic pitfalls of differentiating desmoplastic small round cell tumor (DSRCT) from Wilms tumor (WT). Am J Surg Pathol 38:1220–1226 Ashraf M, Boyd CB, Beresford WA (1984) Ectopic decidual reaction in para-aortic and pelvic lymph nodes in the presence of cervical squamous cell carcinoma during pregnancy. J Surg Oncol 26:6–8 Aslani M, Scully RE (1989) Primary carcinoma of the broad ligament. Report of four cases and review of the literature. Cancer (Phila) 64:1540–1545 Aslani M, Ahn G, Scully RE (1988) Serous papillary cystadenoma of borderline malignancy of broad ligament. Int J Gynecol Pathol 7:131–138 Attanoos RL, Griffin A, Gibbs AR (2003) The use of immunohistochemistry in distinguishing reactive from neoplastic mesothelium. A novel use for desmin and comparative evaluation with epithelial membrane antigen, p53, platelet-derived growth factorreceptor, P-glycoprotein and bcl-2. Histopathology 43:231–238 Ayhan A, Mao T, Tamer S et al (2012) Loss of ARID1A expression is an early molecular event in tumor progression from ovarian endometriotic cyst to clear cell and endometrioid carcinoma. Mod Pathol 22: 1310–1315 Baker PM, Clement PB, Bell DA et al (1999) Superficial endometriosis of the uterine cervix: a report of 20 cases of a process that may be confused with endocervical glandular dysplasia or adenocarcinoma in situ. Int J Gynecol Pathol 18:198–205 Baker PM, Clement PB, Young RH (2005) Malignant peritoneal mesotheliomas in women: a study of 75 cases with emphasis on their morphologic spectrum and differential diagnosis. Am J Clin Pathol 123: 724–737

J. A. Irving and P. B. Clement Bakkar R, Gershenson D, Silva E (2014) Stage IIIC ovarian cancer: a heterogeneous group of patients with different prognosis. Int J Gynecol Pathol 33:302–308 Banu SK, Lee J, Speights VO Jr et al (2008) Cyclooxygenase-2 regulates survival, migration, and invasion of human endometriotic cells through multiple mechanisms. Endocrinology 149:1180–1189 Barnetson RJ, Burnett RA, Downie I et al (2006) Immunohistochemical analysis of peritoneal mesothelioma and primary and secondary serous carcinoma of the peritoneum. Antibodies to estrogen and progesterone receptors are useful. Am J Clin Pathol 125:67–76 Bateman J, Bougie O, Singh S et al (2017) Histomorphological changes in endometriosis in a patient treated with ulipristal: a case report. Pathol Res Pract 213:79–81 Bayer AS, Blumenkrantz MJ, Montgomerie JZ et al (1976) Candida peritonitis. Report of 22 cases and review of the English literature. Am J Med 61:832–840 Bell DA, Scully RE (1990) Serous borderline tumors of the peritoneum. Am J Surg Pathol 14:230–239 Ben-Baruch G, Sivan E, Moran O et al (1996) Primary peritoneal serous papillary carcinoma: a study of 25 cases and comparison with stage III–IV ovarian papillary serous carcinoma. Gynecol Oncol 60:393–396 Biscotti CV, Hart WR (1992) Peritoneal serous micropapillomatosis of low malignant potential (serous borderline tumors of the peritoneum). A clinicopathologic study of 17 cases. Am J Surg Pathol 16:467–475 Bolen JW, Hammar SP, McNutt MA (1986) Reactive and neoplastic serosal tissue. A light microscopic, ultrastructural and immunohistochemical study. Am J Surg Pathol 10:34–47 Boyle DP, McCluggage WG (2009) Peritoneal stromal endometriosis: a detailed morphological analysis of a large series of cases of a common and under-recognized form of endometriosis. J Clin Pathol 62:530–533 Bransilver BR, Ferenczy A, Richart RM (1974) Brenner tumors and Walthard cell nests. Arch Pathol Lab Med 98:76–86 Brunnemann RG, Ro JY, Ordonez NG et al (1999) Extrapleural solitary fibrous tumor: a clinicopathologic study of 24 cases. Mod Pathol 12:1034–1042 Bucella D, Buxant F, Anaf V et al (2009) Omental trophoblastic implants after surgical management of ectopic pregnancy. Arch Gynecol Obstet 280:115–117 Bulun SE (2009) Mechanisms of disease: endometriosis. N Engl J Med 360:268–279 Bur ME, Greene GL, Press MF (1987) Estrogen receptor localization in formalin-fixed, paraffin-embedded endometrium and endometriotic tissues. Int J Gynecol Pathol 6:140–151 Candiani GB, Vercellini P, Fedele L et al (1991) Inguinal endometriosis: pathogenetic and clinical implications. Obstet Gynecol 78:191–194 Capobianco A, Cottone L, Monno A et al (2017) The peritoneum: healing, immunity, and diseases. J Pathol 243:137–147

13

Diseases of the Peritoneum

Carney E, Cimino-Mathews A, Argani C et al (2014) A subset of nondescript axillary lymph node inclusions have the immunophenotype of endosalpingiosis. Am J Surg Pathol 38:1612–1617 Carr NJ, Turk EP (1992) The histological features of splenosis. Histopathology (Oxf) 21:549–553 Carrick KS, Milvenan JS, Albores-Saavedra J (2003) Serous tumor of low malignant potential arising in inguinal endosalpingiosis. Int J Gynecol Pathol 22:412–415 Casey MJ, Synder C, Bewtra C et al (2005) Intraabdominal carcinomatosis after prophylactic oophorectomy in women of hereditary breast ovarian cancer syndrome kindreds associated with BRCA1 and BRCA2 mutations. Gynecol Oncol 97:457–467 Ceruto CA, Brun EA, Chang D et al (2006) Prognostic significance of histomorphologic parameters in diffuse malignant peritoneal mesothelioma. Arch Pathol Lab Med 130:1653–1661 Chai D, Wijesuriya R (2016) Deciduosis of the appendix: diagnostic dilemma continues despite MRI evidence. Ann R Coll Surg Engl 98:e200–e202 Chang S, Oh MH, Ji SY et al (2014) Practical utility of insulin-like growth factor II mRNA-binding protein 3, glucose transporter 1, and epithelial membrane antigen for distinguishing malignant mesotheliomas from benign mesothelial proliferations. Pathol Int 64:607–612 Chapel DB, Husain AN, Krausz T et al (2017) PAX8 expression in a subset of malignant peritoneal mesotheliomas and benign mesothelium has diagnostic implications in the differential diagnosis of ovarian serous carcinoma. Am J Surg Pathol 41:1675–1682. [Epub ahead of print] Chatman DL, Ward AB (1982) Endometriosis in adolescents. J Reprod Med 27:156–160 Chatman DL, Zbella EA (1987) Biopsy in laparoscopically diagnosed endometriosis. J Reprod Med 32:855–857 Chatterjee SK (1980) Scar endometriosis: a clinicopathologic study of 17 cases. Obstet Gynecol 56:81–84 Chen X, Sheng W, Wang J (2013) Well-differentiated papillary mesothelioma: a clinicopathologic and immunohistochemical study of 18 cases with additional observation. Histopathology 62:805–813 Chirac P, Maillet D, Leprêtre F et al (2016) Genomic copy number alterations in 33 malignant peritoneal mesothelioma (sic) analyzed by comparative genomic hybridization array. Hum Pathol 55:72–82 Chopin N, Vieira M, Borghese B et al (2005) Operative management of deeply infiltrating endometriosis: results on pelvic pain symptoms according to a surgical classification. J Minim Invasive Gynecol 12: 106–112 Chorti A, Papavramidis TS, Michalopoulos A (2016) Calcifying fibrous tumor: review of 157 patients reported in international literature. Medicine 95:e3690 Chowdhry AA, Miller FH, Hammer RA (2004) Endometriosis presenting as a urethral diverticulum: a case report. J Reprod Med 49:321–323

831 Churg A, Cagle PT, Roggli VL (2006) Tumors of the serosal membranes, Atlas of tumor pathology, ser IV. Armed Forces Institute of Pathology, Washington, DC Churg A, Allen T, Borczuk AC et al (2014) Welldifferentiated papillary mesothelioma with invasive foci. Am J Surg Pathol 38:990–998 Churg A, Sheffield BS, Galateau-Salle F (2016) New markers for separating benign from malignant mesothelial proliferations. Arch Pathol Lab Med 140:318–321 Cicinelli E, Trojano G, Mastromauro M et al (2017) Higher prevalence of chronic endometritis in women with endometriosis: a possible etiopathogenetic link. Fertil Steril 108:289–295 Cigognetti M, Lonardi S, Fisogni S et al (2015) BAP1 (BRCA1-associated protein 1) is a highly specific marker for differentiating mesothelioma from reactive mesothelial proliferations. Mod Pathol 28:1043–1057 Clarke TJ, Simpson RHW (1990) Necrotizing granulomas of peritoneum following diathermy ablation of endometriosis. Histopathology (Oxf) 16:400–402 Clement PB (1995) Reactive tumor-like lesions of the peritoneum (Editorial). Am J Clin Pathol 103:673–676 Clement PB (2007) The pathology of endometriosis: a survey of the many faces of a common disease emphasizing diagnostic pitfalls and unusual and newly appreciated aspects. Adv Anat Pathol 14:241–260 Clement PB, Scully RE (1978) Extrauterine mesodermal (müllerian) adenosarcoma. A clinicopathologic analysis of five cases. Am J Clin Pathol 69:276–283 Clement PB, Scully RE (1992) 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 11:163–173 Clement PB, Young RH (1992) Endocervicosis of the urinary bladder: a report of six cases of a benign müllerian lesion that may mimic adenocarcinoma. Am J Surg Pathol 16:533–542 Clement PB, Young RH (1993) Florid mesothelial hyperplasia associated with ovarian tumors: a possible source of error in tumor diagnosis and staging. Int J Gynecol Pathol 12:51–58 Clement PB, Young RH (1999) Tumor-like manifestations of florid cystic endosalpingiosis: a report of four cases including the first reported cases of mural endosalpingiosis of the uterus. Am J Surg Pathol 23:166–175 Clement PB, Young RH (2000) Two previously unemphasized features of endometriosis: micronodular stromal endometriosis and endometriosis with stromal elastosis. Int J Surg Pathol 8:223–227 Clement PB, Young RH, Scully RE (1988) Necrotic pseudoxanthomatous nodules of the ovary and peritoneum in endometriosis. Am J Surg Pathol 12:390–397 Clement PB, Young RH, Scully RE (1989) Liesegang rings in endometriosis. A report of three cases. Int J Gynecol Pathol 8:271–276

832 Clement PB, Young RH, Scully RE (1990) Stromal endometriosis of the uterine cervix. A variant of endometriosis that may simulate a sarcoma. Am J Surg Pathol 14:449–455 Clement PB, Young RH, Hanna W et al (1993) Sclerosing peritonitis associated with luteinized thecomas of the ovary: a clinicopathological analysis of six cases. Am J Surg Pathol 18:1–13 Clement PB, Granai CO, Young RH et al (1994) Endometriosis with myxoid change: a case simulating pseudomyxoma peritonei. Am J Surg Pathol 18:849–853 Clement PB, Young RH, Oliva E et al (1996a) Hyperplastic mesothelial cells within abdominal lymph nodes: a mimic of metastatic ovarian carcinoma and serous borderline tumor. A report of two cases associated with ovarian neoplasms. Mod Pathol 9:879–886 Clement PB, Young RH, Scully RE (1996b) Malignant mesotheliomas presenting as ovarian masses. Am J Surg Pathol 20:1067–1080 Cochrane DR, Tessier-Cloutier B, Lawrence KM et al (2017) Clear cell and endometrioid carcinomas: are their differences attributable to distinct cells of origin? J Pathol 243:26–36 Coffin CM, Hornick JL, Fletcher CD (2007) Inflammatory myofibroblastic tumor: comparison of clinicopathologic, histologic, and immunohistochemical features including ALK expression in atypical and aggressive cases. Am J Surg Pathol 31:509–520 Colella R, Mameli MG, Bellezza G et al (2010) Endometriosis-associated skeletal muscle regeneration: a hitherto undescribed entity and a potential diagnostic pitfall. Am J Surg Pathol 34:10–17 Comin CE, Saieva C, Messerini (2007) h-caldesmon, calretinin, estrogen receptor, and Ber-EP4: a useful combination of immunohistochemical markers for differentiating epithelioid peritoneal mesothelioma from serous papillary carcinoma of the ovary. Am J Surg Pathol 31:1139–1148 Copeland LJ, Silva EG, Gershenson DM et al (1988) The significance of müllerian inclusions found at secondlook laparotomy in patients with epithelial ovarian neoplasms. Obstet Gynecol 71:763–770 Corben AD, Nehhozina T, Garg K et al (2010) Endosalpingiosis in axillary lymph nodes: a possible pitfall in the staging of patients with breast carcinoma. Am J Surg Pathol 34:1211–1216 Cornillie FJ, Oosterlynck D, Lauweryns JM et al (1990) Deeply infiltrating pelvic endometriosis: histology and clinical significance. Fertil Steril 53:978–983 Cramer SF, Meyer JS, Kraner JF et al (1980) Metastasizing leiomyoma of the uterus. S-phase fraction, estrogen receptor, and ultrastructure. Cancer (Phila) 45:932–937 Cumiskey J, Whyte P, Kelehan P et al (2008) A detailed morphologic and immunohistochemical comparison of pre- and postmenopausal endometriosis. J Clin Pathol 61:455–459 Czernobilsky B, Silverstein A (1978) Salpingitis and ovarian endometriosis. Fertil Steril 30:45–49

J. A. Irving and P. B. Clement D’Hooghe T, Bambra CS, Raeymaekers BM et al (1995) Intrapelvic injection of menstrual endometrium causes endometriosis in baboons (Papio cynocephalus and Papio anubis). Am J Obstet Gynecol 173:125–134 David MP, Ben-Zwi D, Langer L (1981) Tubal intramural polyps and their relationship to infertility. Fertil Steril 35:526–531 Day DL, Sane S, Dehner LP (1986) Inflammatory pseudotumor of the mesentery and small intestine. Pediatr Radiol 16:210–215 Daya D, McCaughey WTE (1990) Well-differentiated papillary mesothelioma of the peritoneum. A clinicopathologic study of 22 cases. Cancer (Phila) 65:292–296, 185–195 de Peralta MN, Delahoussaye PM, Tornos CS et al (1994) Benign retroperitoneal cysts of müllerian type: a clinicopathologic study of three cases and review of the literature. Int J Gynecol Pathol 13:273–278 Deger RB, LiVolsi VA, Noumoff JS (1995) Foreign body reaction (gossypiboma) masking as recurrent ovarian cancer. Gynecol Oncol 56:94–96 Doss BJ, Jacques SM, Qureshi F et al (1998) Extratubal secondary trophoblastic implants: clinicopathologic correlation and review of the literature. Hum Pathol 29:184–187 Ecker AM, Donnellan NM, Shepherd JP et al (2014) Abdominal wall endometriosis: 12 years of experience at a large academic institution. Am J Obstet Gynecol 211:363.e1–363.e5 Egger H, Weigmann P (1982) Clinical and surgical aspects of ovarian endometriotic cysts. Arch Gynecol 233:37–45 Elmore LW, Sherman ME, Seidman JD et al (2000) p53 expression and mutational status of primary peritoneal micropapillary serous carcinoma (abstract). Mod Pathol 13:124A Emory TS, Monihan JM, Carr NJ et al (1997) Sclerosing mesenteritis, mesenteric panniculitis and mesenteric lipodystrophy: a single entity? Am J Surg Pathol 21:392–398 Erzen M, Rakar S, Klancar B et al (2001) Endometriosisassociated ovarian carcinoma: an entity distinct from other ovarian carcinomas as suggested by a nested casecontrol study. Gynecol Oncol 83:100–108 Esheba GE, Longacre TA, Atkins KA et al (2009) Expression of the urothelial differentiation markers GATA3 and placental S100 (S100P) in female genital tract transitional cell proliferations. Am J Surg Pathol 33:347–453 Esselen KM, Ng S, Hua Y et al (2014) Endosalpingiosis as it relates to tubal, ovarian and serous neoplastic tissues: an immunohistochemical study of tubal and mullerian antigens. Gynecol Oncol 132:316–321 Euscher ED, Malpica A, Deavers MT et al (2005) Differential expression of WT-1 in serous carcinomas in the peritoneum with or without associated serous carcinoma in endometrial polyps. Am J Surg Pathol 29:1074–1078 Evans H, Yates WA, Palmer WE et al (1990) Clear cell carcinoma of the sigmoid mesocolon: a tumor of the

13

Diseases of the Peritoneum

secondary müllerian system. Am J Obstet Gynecol 162:161–163 Farland LV, Eliassen AH, Tamimi RM et al (2017) History of breast feeding and risk of incident endometriosis: prospective cohort study. BMJ 358:j3778 Feldman AL, Libutti SK, Pingpank JF et al (2003) Analysis of factors associated with outcome in patients with malignant peritoneal mesothelioma undergoing surgical debulking and intraperitoneal chemotherapy. J Clin Oncol 24:4560–4567 Finch A, Beiner M, Lubinski J et al (2006) Salpingooophorectomy and the risk of ovarian, fallopian tube, and peritoneal cancers in women with a BRCA1 or BRCA2 mutation. JAMA 296:185–192 Flieder DB, Moran CA, Travis WD et al (1998) Pleuropulmonary endometriosis and pulmonary ectopic deciduosis: a clinicopathologic and immunohistochemical study of 10 cases with emphasis on diagnostic pitfalls. Hum Pathol 29:1495–1503 Folpe AL, Mentzel T, Lehr HA et al (2005) Perivascular epithelioid cell neoplasms of soft tissue and gynecologic origin: a clinicopathologic study of 26 cases and review of the literature. Am J Surg Pathol 29:1558–1575 Foo KT, Ng KC, Rauff A et al (1978) Unusual small intestinal obstruction in adolescent girls: the abdominal cocoon. Br J Surg 65:427–430 Forouhar F (1982) Meconium peritonitis. Pathology, evolution, and diagnosis. Am J Clin Pathol 78:208–213 Fredericks S, Russell P, Cooper M et al (2005) Smooth muscle in the female pelvic peritoneum: a clinicopathological analysis of 31 women. Pathology 37:14–21 Fukunaga M, Ushigome S (1998) Epithelial metaplastic changes in ovarian endometriosis. Mod Pathol 11:784–788 Fukunaga M, Naganuma H, Ushigome S et al (1996) Malignant solitary fibrous tumour of the peritoneum. Histopathology (Oxf) 28:463–466 Fukunaga M, Nomura K, Ishikawa E et al (1997) Ovarian atypical endometriosis: its close association with malignant epithelial tumors. Histopathology (Oxf) 30:249–255 Fung JN, Rogers PA, Montgomery GW (2015) Identifying the biological basis of GWAS hits for endometriosis. Biol Reprod 92:87 Fuseya C, Horiuchi A, Hayashi A et al (2012) Involvement of pelvic inflammation-related mismatch repair abnormalities and microsatellite instability in the malignant transformation of ovarian endometriosis. Hum Pathol 43:1964–1972 Gallan AJ, Antic T (2016) Benign müllerian glandular inclusions in men undergoing pelvic lymph node dissection. Hum Pathol 57:136–139 Gardner HL (1966) Cervical and vaginal endometriosis. Clin Obstet Gynecol 9:358–372 George E, Leyser S, Zimmer HL et al (1995) Vernix caseosa peritonitis: an infrequent complication of cesarean section with distinctive histopathologic features. Am J Clin Pathol 103:681–684

833 Gershenson DM, Bodurka DC, Lu KH et al (2015) Impact of age and primary disease site on outcome in women with low-grade serous carcinoma of the ovary or peritoneum: results of a large single-institution registry of a rare tumor. J Clin Oncol 33:2675–2682 Gershenson DM, Bodurka DC, Coleman RL et al (2017) Hormonal maintenance therapy for women with low-grade serous cancer of the ovary or peritoneum. J Clin Oncol 35:1103–1111 Ghigna MR, Mercier O, Mussot S et al (2015) Thoracic endometriosis: clinicopathologic updates and issues about 18 cases from a tertiary referring center. Ann Diagn Pathol 19:320–325 Gilks CB, Bell DA, Scully RE (1990) Serous psammocarcinoma of the ovary and peritoneum. Int J Gynecol Pathol 9:110–121 Goldblum J, Hart WR (1995) Localized and diffuse mesotheliomas of the genital tract and peritoneum in women. A clinicopathological study of nineteen true mesothelial neoplasms, other than adenomatoid tumors, multicystic mesotheliomas and localized fibrous tumors. Am J Surg Pathol 19:1124–1137 Gonzalez-Crussi F, deMello DE, Sotelo-Avila C (1983) Omental-mesenteric myxoid hamartomas. Am J Surg Pathol 7:567–578 Greaves E, Temp J, Esnal-Zufiurre A et al (2015) Estradiol is a critical mediator of macrophage-nerve cross talk in peritoneal endometriosis. Am J Pathol 185: 2286–2297 Griffith LM, Carcangiu M (1991) Sex cord tumor with annular tubules associated with endometriosis of the fallopian tube. Am J Clin Pathol 96:259–262 Guarch R, Puras A, Ceres R et al (2001) Ovarian endometriosis and clear cell carcinoma, leiomyomatosis peritonealis disseminata, and endometrial adenocarcinoma: an unusual, pathogenetically related association. Int J Gynecol Pathol 20:267–270 Gupta S, Goldberg JM, Aziz N et al (2008) Pathogenic mechanisms in endometriosis-associated infertility. Fertil Steril 90:247–257 Halme J, Hammond MG, Hulka JF et al (1984) Retrograde menstruation in healthy women and in patients with endometriosis. Obstet Gynecol 64:151–154 Hameed A, Jafri N, Copeland LJ et al (1996) Endometriosis with myxoid change simulating mucinous adenocarcinoma and pseudomyxoma peritonei. Gynecol Oncol 62:317–319 Harper GB Jr, Awbrey BJ, Thomas CG Jr et al (1986) Mesothelial cysts of the round ligament simulating inguinal hernia. Report of four cases and review of the literature. Am J Surg 151:515–517 Healey EG, McCluggage WG (2012) Abdominal wall endometriosis associated with ventriculoperitoneal and lumboperitoneal shunts: a report of 2 cases of an extremely rare phenomenon. Int J Surg Pathol 20: 301–304 Heller DS, Gordon RE, Clement PB et al (1999) Presence of asbestos in peritoneal mesotheliomas in women. Int J Gynecol Cancer 9:452–455

834 Hitti IF, Glasberg SS, Lubicz S (1990) Clear cell carcinoma arising in extraovarian endometriosis: report of three cases and review of the literature. Gynecol Oncol 39:314–320 Horn L-C, Bilek K (1995) Frequency and histogenesis of pelvic retroperitoneal lymph node inclusions of the female genital tract. An immunohistochemical study of 34 cases. Pathol Res Pract 191:991–996 Horton JD, Dezee KJ, Ahnfeldt EP et al (2008) Abdominal wall endometriosis: a surgeon’s perspective and review of 445 cases. Am J Surg 196:207–212 Hosfield EM, Rabban JT, Chen L et al (2008) Squamous metaplasia of the ovarian surface epithelium and subsurface fibrosis: distinctive pathologic findings in the ovaries and fallopian tubes of patients on peritoneal dialysis. Int J Gynecol Pathol 27:465–474 Houston DE, Noller KL, Melton J et al (1987) Incidence of pelvic endometriosis in Rochester, Minnesota, 1970–1979. Am J Epidemiol 125:959–969 Houston DE, Noller KL, Melton JIII et al (1988) The epidemiology of pelvic endometriosis. Clin Obstet Gynecol 31:787–800 Hsu YK, Rosenshein NB, Parmley TH et al (1981) Leiomyomatosis in pelvic lymph nodes. Obstet Gynecol 57:91S–93S Insabato L, Natella V, Somma A et al (2015) Primary peritoneal clear cell carcinoma versus ovarian carcinoma versus malignant transformation of endometriosis: a vexing issue. Int J Surg Pathol 23:211–216 Ishimaru T, Masuzaki H (1991) Peritoneal endometriosis: endometrial tissue implantation as its primary etiologic mechanism. Am J Obstet Gynecol 165:210–214 Ismail SM (1991) Cone biopsy causes cervical endometriosis and tubo-endometrioid metaplasia. Histopathology (Oxf) 18:107–114 Ito T, Hamasaki M, Matsumoto S et al (2015) p16/CDKN2A FISH in differentiation of diffuse malignant peritoneal mesothelioma from mesothelial hyperplasia and epithelial ovarian cancer. Am J Clin Pathol 143:830–838 Jansen RPS, Russell P (1986) Nonpigmented endometriosis: clinical, laparoscopic, and pathologic definition. Am J Obstet Gynecol 155:1154–1159 Jaworski RC, Boadle R, Greg J et al (2001) Peritoneal “melanosis” associated with a ruptured ovarian dermoid cyst: report of a case with electron-probe energy dispersive x-ray analysis. Int J Gynecol Pathol 20:386–389 Jelihovsky T, Grant AF (1968) Endometriosis of the lung: a case report and brief review of the literature. Thorax 23:434–437 Jiang W, Roma AA, Lai K et al (2013) Endometriosis involving the mucosa of the intestinal tract: a clinicopathologic study of 15 cases. Mod Pathol 26:1270–1278 Joseph NM, Chen YY, Nasr A et al (2017) Genomic profiling of malignant peritoneal mesothelioma reveals recurrent alterations in epigenetic regulatory genes BAP1, SETD2, and DDX3X. Mod Pathol 30:246–254

J. A. Irving and P. B. Clement Kane C, Drouin P (1985) Obstructive uropathy associated with endometriosis. Am J Obstet Gynecol 151:207–211 Kapoor OP, Nathwani BN, Joshi VR (1972) Amoebic peritonitis. A study of 73 cases. J Trop Med Hyg 75:11–15 Kato N, Sasou S, Motoyama T (2006) Expression of hepatocyte nuclear factor-1 beta (HNF-1 beta) in clear cell tumors and endometriosis of the ovary. Mod Pathol 19:83–89 Kawai T, Tominaga S, Hiroi S et al (2016) Peritoneal malignant mesothelioma (PMM), and primary peritoneal serous carcinoma (PPSC) and reactive mesothelial hyperplasia (RMH) of the peritoneum. Immunohistochemical and fluorescence in situ hybridization (FISH) analyses. J Clin Pathol 69:706–712 Kazakov DV, Ondic O, Zamecnik M et al (2007) Morphological variations of scar-related and spontaneous endometriosis of the skin and superficial soft tissue: a study of 71 cases with emphasis on atypical features and types of müllerian differentiations. J Am Acad Dermatol 57:134–146 Kelly P, McCluggage WG, Gardiner KR et al (2008) Intestinal endometriosis morphologically mimicking colonic adenocarcinoma. Histopathology 52:510–514 Kempers RD, Dockerty MB, Hunt AB et al (1960) Significant postmenopausal endometriosis. Surg Gynecol Obstet 111:348–356 Kern SB (1991) Prevalence of psammoma bodies in Papanicolaou-stained cervicovaginal smears. Acta Cytol 35:81–88 Kerrigan SAJ, Turnnir RT, Clement PB et al (2002) Diffuse malignant epithelial mesotheliomas of the peritoneum in women: a clinicopathologic study of 25 cases. Cancer 94:378–385 Kheir SM, Mann WJ, Wilkerson JA (1981) Glandular inclusions in lymph nodes. The problem of extensive involvement and relationship to salpingitis. Am J Surg Pathol 5:353–359 Kim K, Scully RE (1990) Peritoneal keratin granulomas with carcinomas of endometrium and ovary and atypical polypoid adenomyoma of endometrium. A clinicopathological analysis of 22 cases. Am J Surg Pathol 14:925–932 Kim J, Russell P, Arendse M et al (2013) Endometriosis in appendix and adjacent caecum with intestinal gland differentiation. Lett Pathol 45:513–515 Kim HS, Yoon G, Ha SY et al (2015a) Nodular smooth muscle metaplasia in multiple peritoneal endometriosis. Int J Clin Exp Pathol 8:3370–3373 Kim HS, Yoon G, Kim BG et al (2015b) Decidualization of intranodal endometriosis in a postmenopausal woman. Int J Clin Exp Pathol 8:1025–1030 Kitazawa M, Kaneko H, Toshima M et al (1984) Malignant peritoneal mesothelioma with massive foamy cells. Acta Pathol Jpn 34:687–692 Knabben L, Imboden S, Fellman B et al (2015) Urinary tract endometriosis in patients with deep infiltrating endometriosis: prevalence, symptoms, management,

13

Diseases of the Peritoneum

and proposal for a new clinical classification. Fertil Steril 103:147–152 Köbel M, Turbin D, Kalloger SE (2011) Biomarker expression in pelvic high-grade serous carcinoma: comparison of ovarian and omental sites. Int J Gynecol Pathol 30:366–371 Koc S, Beydilli G, Tulunay G et al (2006) Peritoneal tuberculosis mimicking advanced ovarian cancer: a retrospective review of 22 cases. Gynecol Oncol 103:565–569 Koninckx PR, Ide P, Vandenbroucke W et al (1980) New aspects of the pathophysiology of endometriosis and associated infertility. J Reprod Med 24:257–260 Konstantinova AM, Michal M, Kacerovska D et al (2013) Multilocular mesothelial proliferation involving the skin of the umbilicus. Am J Dermatopathol 35:856–858 Krasinskas AM, Bartlett DL, Cieply K et al (2010) CDKN2A and MTAP deletions in peritoneal mesotheliomas are correlated with loss of p16 protein expression and poor survival. Mod Pathol 23:531–538 Krasinskas AM, Borczuk AC, Hartman DJ et al (2016) Prognostic significance of morphologic growth patterns and mitotic index of epithelioid malignant peritoneal mesothelioma. Histopathology 68:729–737 Kuo T, Hsueh S (1984) Mucicarminophilic histiocytosis. A polyvinylpyrrolidone (PVP) storage disease simulating signet-ring cell carcinoma. Am J Surg Pathol 8:419–428 Lae ME, Roche PC, Jin L et al (2002) Desmoplastic small round cell tumor. A clinicopathological, immunohistochemical, and molecular study of 32 tumors. Am J Surg Pathol 26:823–835 LaGrenade A, Silverberg SG (1988) Ovarian tumors associated with atypical endometriosis. Hum Pathol 19:1080–1084 Lamb K, Hoffmann RG, Nichols TR (1986) Family trait analysis: a case-control study of 43 women with endometriosis and their best friends. Am J Obstet Gynecol 154:596–601 Lauchlan SC (1972) The secondary müllerian system. Obstet Gynecol Surv 27:133–146 Lee KR, Verma U, Belinson J (1991) Primary clear cell carcinoma of the peritoneum. Gynecol Oncol 41:259–262 Lee M, Alexander HR, Burke A (2013) Diffuse mesothelioma of the peritoneum: a pathological study of 64 tumours treated with cytoreductive therapy. Pathology 45:464–473 Lessey BA, Metzger DA, Haney AF et al (1989) Immunohistochemical analysis of estrogen and progesterone receptors in endometriosis: comparison with normal endometrium during the menstrual cycle and the effect of medical therapy. Fertil Steril 51:409–415 Leyendecker G (2002) Endometriosis results from the dislocation of basalis endometrium. Hum Reprod 17:2736 Lim S, Kim JY, Park K et al (2003) Mullerianosis of the mesosalpinx: a case report. Int J Gynecol Pathol 22:209–212

835 Lim CS, Thompson JF, McKenzie PR et al (2012) Peritoneal melanosis associated with metastatic melanoma involving the omentum. Pathology 44:255–257 Lin BT-Y, Colby T, Gown AM et al (1996) Malignant vascular tumors of the serous membranes mimicking mesothelioma. A report of 14 cases. Am J Surg Pathol 20:1431–1439 Liu S, Staats P, Lee M et al (2014) Diffuse mesothelioma of the peritoneum: correlation between histological and clinical parameters and survival. Pathology 46:604–609 Lu F-I, Gilks CB, Mulligan A et al (2012) Prevalence of loss of expression of DNA mismatch repair proteins in primary ovarian tumors. Int J Gynecol Pathol 31:524–531 Lu Y, Cuellar-Partida G, Painter JN et al (2015) Shared genetics underlying epidemiological association between endometriosis and ovarian cancer. Hum Mol Genet 24:5955–5964 Lv Y, Li P, Zheng J et al (2012) Nodular histiocytic aggregates in the greater omentum in patients with ovarian cancer. Int J Surg Pathol 20:178–184 Maeda D, Shih I (2013) Pathogenesis and the role of ARID1A mutation in endometriosis-related ovarian neoplasms. Adv Anat Pathol 20:45–52 Magro G, Salvatorelli L, Alaggio R et al (2017) Diagnostic utility of cyclin D1 in the diagnosis of small round blue cell tumors in children and adolescents. Hum Pathol 60:58–65 Malpica A, Sant’Ambrogio S, Deavers MT et al (2012) Well-differentiated papillary mesothelioma of the female peritoneum: a clinicopathologic study of 26 cases. Am J Surg Pathol 36:117–127 Mangano WE, Cagle PT, Churg A et al (1998) The diagnosis of desmoplastic malignant mesothelioma and its distinction from fibrous pleurisy. A histologic and immunohistochemical analysis of 31 cases including p53 immunostaining. Am J Clin Pathol 110:191–199 Martin JD Jr, Hauck AE (1985) Endometriosis in the male. Am Surg 51:426–430 Martin DC, Hubert GD, Vander Zwaag R et al (1989) Laparoscopic appearances of peritoneal endometriosis. Fertil Steril 51:63–67 Martinka M, Allaire C, Clement PB (1999) Endocervicosis presenting as a painful vaginal mass: a case report. Int J Gynecol Pathol 18:274–276 Masand RP, Euscher ED, Deavers MT et al (2013) Endometrioid stromal sarcoma: a clinicopathologic study of 63 cases. Am J Surg Pathol 37:1635–1647 Matalliotakis TM, Cakmak H, Fragouli YG et al (2008) Epidemiological characteristics in women with and without endometriosis in the Yale series. Arch Gynecol Obstet 277:389–393 Matsumoto T, Yamazaki M, Takahashi H et al (2015) Distinct β-catenin and PIK3CA mutation profiles in endometriosis-associated ovarian endometrioid and clear cell carcinomas. Am J Clin Pathol 144:452–463 McCluggage WG, Bryson C, Lamki H et al (2000) Benign, borderline, and malignant endometrioid neoplasia

836 arising in endometriosis in association with tamoxifen therapy. Int J Gynecol Pathol 19:276–279 McCluggage WG, Oliva E, Herrington CS et al (2003) CD10 and calretinin staining of endocervical glandular lesions, endocervical stroma and endometrioid adenocarcinomas of the uterine corpus: CD10 positivity is characteristic of, but not specific for, mesonephric lesions and is not specific for endometrial stroma. Histopathology 43:144–150 McCluggage WG, Judge MJ, Clarke BA et al (2015a) Data set for reporting of ovary, fallopian tube and primary peritoneal carcinoma: recommendations from the International Collaboration on Cancer Reporting (ICCR). Mod Pathol 28:1101–1122 McCluggage WG, Stewart CJ, Iacobelli J et al (2015b) Microscopic extraovarian sex cord proliferations: an undescribed phenomenon. Histopathology 66:555–564 McCoubrey A, Houghton O, McCallion K et al (2005) Serous adenocarcinoma of the sigmoid mesentery arising in cystic endosalpingiosis. J Clin Pathol 58:1221–1223 McFadden DE, Clement PB (1986) Peritoneal inclusion cysts with mural mesothelial proliferation. A clinicopathological analysis of six cases. Am J Surg Pathol 10:844–854 Merchant SH, Haghir S, Gordon GB (2000) Granulomatous peritonitis after laparoscopic cholecystectomy mimicking pelvic endometriosis. Obstet Gynecol 96:830–831 Metzger DA, Lessey BA, Soper JT et al (1991) Hormoneresistant endometriosis following total abdominal hysterectomy and bilateral salpingo-oophorectomy: correlation with histology and steroid receptor content. Obstet Gynecol 78:946–950 Michal M, Kazakov DV, Dundr P et al (2016) Histiocytosis with raisinoid nuclei: a unifying concept for lesions reported under different names as nodular mesothelial/histiocytic hyperplasia, mesothelial/monocytic incidental cardiac excrescence, intralymphatic histiocytosis, and others. A report of 50 cases. Am J Surg Pathol 40:1507–1516 Mikami M, Tei C, Takehara K et al (2003) Retroperitoneal primary mucinous adenocarcinoma with a mural nodule of anaplastic tumor: a case report and literature review. Int J Gynecol Pathol 22:205–208 Mills SE (1983) Decidua and squamous metaplasia in abdominopelvic lymph nodes. Int J Gynecol Pathol 2:209–215 Min YW, Lim KS, Min BH et al (2014) Proton pump inhibitor use significantly increases the risk of spontaneous bacterial peritonitis in 1965 patients with cirrhosis and ascites: a propensity score matched cohort study. Aliment Pharmacol Ther 40:695–704 Minaglia S, Mishell DR Jr, Ballard CA (2007) Incisional endometriomas after cesarean section: a case series. J Reprod Med 52:630–634 Minato H, Shimizu J, Arano Y et al (2012) IgG4-related sclerosing mesenteritis: a rare mesenteric disease of unknown etiology. Pathol Int 62:281–286

J. A. Irving and P. B. Clement Misdraji J, Lauwers GY, Irving JA et al (2014) Appendiceal or cecal endometriosis with intestinal metaplasia. Am J Surg Pathol 38:698–705 Moloshok AA, Ivanko AI (1984) Endometriosis of the breast (an observation). Vopr Onkol 30:88–89 Mostoufizadeh M, Scully RE (1980) Malignant tumors arising in endometriosis. Clin Obstet Gynecol 23:951–963 Mourra N, Nion I, Parc R et al (2004) Squamous metaplasia of the peritoneum: a potential diagnostic pitfall. Histopathology 44:621–622 Mourra N, Cortez A, Bennis M et al (2015) The groin: an unusual location of endometriosis – a multiinstitutional clinicopathologic study. J Clin Pathol 68:579–581 Msika S, Gruden E, Sarnacki S et al (2010) Cytoreductive surgery associated to hyperthermic intraperitoneal chemoperfusion for desmoplastic round small cell tumor with peritoneal carcinomatosis in young patients. J Pediatr Surg 45:1617–1621 Muneyyirci-Delale O, Neil G, Serur E et al (1998) Endometriosis with massive ascites. Obstet Gynecol 69:42–46 Murphy AJ, Bishop K, Pereira C et al (2008) A new molecular variant of desmoplastic small round cell tumor: significance of WT1 immunostaining in this entity. Hum Pathol 39:1763–1770 Nakaguro M, Tsuzuki T, Shimada S (2016) Adenocarcinoma arising in urinary bladder endocervicosis. Pathol Int 66:108–113 Naresh KN, Ahuja VK, Rao CR et al (1991) Squamous cell carcinoma arising in endometriosis of the ovary. J Clin Pathol 44:958–959 Nascimento AF, Ruiz R, Hornick JL et al (2002) Calcifying fibrous ‘pseudotumor’: clinicopathologic study of 15 cases and analysis of its relationship to inflammatory myofibroblastic tumor. Int J Surg Pathol 10:189–196 Nazeer T, Ro JY, Tornos C et al (1996) Endocervical type glands in urinary bladder: a clinicopathologic study of six cases. Hum Pathol 27:816–820 Nisenblat V, Bossuyt PM, Shaikh R et al (2016) Blood biomarkers for the non-invasive diagnosis of endometriosis. Cochrane Database Syst Rev 5:CD012179 Nisolle-Pochet M, Casanas-Roux F, Donnez J (1988) Histologic study of ovarian endometriosis after hormonal therapy. Fertil Steril 49:423 Nissim F, Ashkenazy M, Borenstein R et al (1981) Tuberculoid cornstarch granulomas with caseous necrosis. Arch Pathol Lab Med 105:86–88 Noli S, Cipriani S, Scarfone G et al (2013) Long term survival of ovarian endometriosis associated clear cell and endometrioid ovarian cancers. Int J Gynecol Cancer 23:244–248 Nonaka D, Kasamura S, Baratti D et al (2005) Diffuse malignant mesothelioma of the peritoneum. A clinicopathologic study of 35 patients treated locoregionally at a single institution. Cancer 104:2181–2188 Olive DL, Henderson DY (1987) Endometriosis and müllerian anomalies. Obstet Gynecol 69:412–415

13

Diseases of the Peritoneum

Oparka R, McCluggage WG, Herrington CS (2011) Peritoneal mesothelial hyperplasia associated with gynaecological disease: a potential diagnostic pitfall that is commonly associated with endometriosis. J Clin Pathol 64:313–318 Ordi J, de Alava E, Torné A et al (1998) Intraabdominal desmoplastic small round cell tumor with EWS/ERG fusion transcript. Am J Surg Pathol 22:1026–1032 Ordonez NG (1998a) Desmoplastic small round cell tumor. I: a histopathologic study of 39 cases with emphasis on unusual histologic patterns. Am J Surg Pathol 22:1303–1313 Ordonez NG (1998b) Desmoplastic small round cell tumor. II: an ultrastructural and immunohistochemical study with emphasis on new histochemical markers. Am J Surg Pathol 22:1303–1313 Ordóñez NG (2005a) Mesothelioma with clear cell features: an ultrastructural and immunohistochemical study of 20 cases. Hum Pathol 36:465–473 Ordóñez NG (2005b) Value of estrogen and progesterone receptor immunostaining in distinguishing between peritoneal mesotheliomas and serous carcinomas. Hum Pathol 36:1163–1167 Ordóñez NG (2012a) Deciduoid mesothelioma: report of 21 cases with review of the literature. Mod Pathol 25:1481–1495 Ordóñez NG (2012b) Pleomorphic mesothelioma: report of 10 cases. Mod Pathol 25:1011–1022 Ordóñez NG (2013a) Mesothelioma with signet-ring cell features: report of 23 cases. Mod Pathol 26:370–384 Ordóñez NG (2013b) Value of PAX8, PAX2, caludin-4, and h-caldesmon immunostaining in distinguishing peritoneal epithelioid mesotheliomas from serous carcinomas. Mod Pathol 26:553–562 Orezzoli JP, Russell AH, Oliva E et al (2008) Prognostic implication of endometriosis in clear cell carcinoma of the ovary. Gynecol Oncol 110:336–344 OuYang Z, Hirota Y, Osuga Y et al (2008) Interleukin-4 stimulates proliferation of endometriotic stromal cells. Am J Pathol 173:463–469 Padmanabhan V, Mount SL, Eltabbakh GH (2003) Peritoneal atypical mesothelial proliferation with progression to invasive mesothelioma: a case report and review of the literature. Pathology 35:260–263 Pai SA, Desai SB, Borges AM (1998) Uterus-like masses of the ovary associated with breast cancer and raised serum CA 125. Am J Surg Pathol 22:333–337 Park SA, Giannattasio C, Tancer ML (1981) Foreign body reaction to the intraperitoneal use of avitene. Obstet Gynecol 58:664–668 Parker RL, Dadmanesh F, Young RH et al (2004) Polypoid endometriosis: a clinicopathologic analysis of 24 cases and a review of the literature. Am J Surg Pathol 28:285–297 Paull T, Tedeschi LG (1972) Perineal endometriosis at the site of episiotomy scar. Obstet Gynecol 40:28–34 Pavlisko EN, Roggli VL (2015) Sarcomatoid peritoneal mesothelioma. Am J Surg Pathol 39:1568–1575

837 Peress MR, Sosnowski JR, Mathur RS et al (1982) Pelvic endometriosis and Turner’s syndrome. Am J Obstet Gynecol 144:474–476 Perrotta PL, Ginsburg FW, Siderides CI et al (1998) Liesegang rings and endometriosis. Int J Gynecol Pathol 17:358–362 Petersen VC, Underwood JCE, Wells M et al (2002) Primary endometrioid adenocarcinoma of the large intestine arising in colorectal endometriosis. Histopathology 40:171–176 Pettinato G, Manivel JC, De Rosa N et al (1990) Inflammatory myofibroblastic tumor (plasma cell granuloma). Clinicopathologic study of 20 cases with immunohistochemical and ultrastructural observations. Am J Clin Pathol 94:538–546 Prade M, Spatz A, Bentley R et al (1995) Borderline and malignant serous tumor arising in pelvic lymph nodes: evidence of origin in benign glandular inclusions. Int J Gynecol Pathol 14:87–91 Prystowsky JB, Stryker SJ, Ujiki GT et al (1988) Gastrointestinal endometriosis. Incidence and indications for resection. Arch Surg 123:855–858 Pueblitz-Peredo S, Luevano-Flores E, Rincon-Taracena R et al (1985) Uteruslike mass of the ovary: endomyometriosis or congenital malformation? A case with a discussion of histogenesis. Arch Pathol Lab Med 109:361–364 Punnonen R, Klemi PJ, Nikkanen V (1980) Postmenopausal endometriosis. Eur J Obstet Gynecol Reprod Biol 11:195–200 Quagliarello J, Coppa G, Bigelow B (1985) Isolated endometriosis in an inguinal hernia. Am J Obstet Gynecol 152:688–689 Rahilly MA, Al-Nafusi A (1991) Uterus-like mass of the ovary associated with endometrioid carcinoma. Histopathology (Oxf) 18:549–551 Ravishankar S, Malpica A, Ramalingam P (2017) Yolk sac tumor in extragonadal pelvic sites: still a diagnostic challenge. Am J Surg Pathol 41:1–11 Redwine DB (1987) The distribution of endometriosis in the pelvis by age groups and fertility. Fertil Steril 47:173–175 Redwine DB (1989) Peritoneal pockets and endometriosis. Confirmation of an important relationship, with further observations. J Reprod Med 34:270–272 Rock JA, Parmley TH, King TM et al (1981) Endometriosis and the development of tuboperitoneal fistulas after tubal ligation. Fertil Steril 35:16–20 Roma AA, Malpica A (2009) Primary retroperitoneal mucinous tumors. A clinicopathologic study of 18 cases. Am J Surg Pathol 33:526–533 Roma AA, Masand RP (2014) Ovarian Brenner tumors and Walthard nests: a histologic and immunohistochemical study. Hum Pathol 45:2417–2422 Roman H, Hennetier C, Darwish B et al (2016) Bowel occult microscopic endometriosis in resection margins in deep colorectal endometriosis specimens has no impact on short-term postoperative outcomes. Fertil Steril 105:423–429

838 Rosai J, Dehner LP (1975) Nodular mesothelial hyperplasia in hernia sacs. A benign reactive condition stimulating a neoplastic process. Cancer (Phila) 35:165–175 Ross MJ, Welch WR, Scully RE (1989) Multilocular peritoneal inclusion cysts (so-called cystic mesotheliomas). Cancer (Phila) 64:1336–1346 Roth LM (1973) Endometriosis with perineural involvement. Am J Clin Pathol 59:807–809 Ruffolo R, Suster S (1993) Diffuse histiocytic proliferation mimicking mesothelial hyperplasia in endocervicosis of the female pelvic peritoneum. Int J Surg Pathol 1:101–106 Russell P (1979) The pathological assessment of ovarian neoplasms. II. The proliferating “epithelial” tumours. Pathology 11:251–282 Russell P, van der Griend R, Anderson L et al (2016) Evidence for lymphatic pathogenesis of endosalpingiosis. Lett Pathol 48:72–75 Rutgers JL, Scully RE (1988a) Ovarian müllerian mucinous papillary cystadenomas of borderline malignancy. A clinicopathological analysis. Cancer (Phila) 61:340–348 Rutgers JL, Scully RE (1988b) Ovarian mixed-epithelial papillary cystadenomas of borderline malignancy of mullerian type. A clinicopathological analysis. Cancer (Phila) 61:546–554 Rutgers JL, Young RH, Scully RE (1987) Ovarian yolk sac tumor arising from an endometrioid carcinoma. Hum Pathol 18:1296–1299 Salviato T, Altavill G, Busatto G et al (2006) Diffuse intraabdominal clear cell myomelanocytic tumor: report of an unusual presentation of “PEComatosis” simulating peritoneal mesothelioma. Ann Diagn Pathol 10:352–356 Sampson JA (1927) Peritoneal endometriosis due to the menstrual dissemination of endometrial tissue into the peritoneal cavity. Am J Obstet Gynecol 14:422–469 Sandoval P, Jiménez-Heffernan JA, Guerra-Azcona G et al (2016) Mesothelial-to-mesenchymal transition in the pathogenesis of post-surgical peritoneal adhesions. J Pathol 239:48–59 Santulli P, Streuli I, Melonio I et al (2015) Increased serum cancer antigen-125 is a marker for severity of deep endometriosis. J Minim Invasive Gynecol 22:275–284 Sasson IE, Taylor HS (2008) Stem cells and the pathogenesis of endometriosis. Ann N YAcad Sci 1127:106–115 Satgunaseelan L, Russell P, Phan-Thien K et al (2016) Perineural space infiltration by endosalpingiosis. Lett Pathol 48:76–78 Sato N, Tsunoda H, Nishida M et al (2000) Loss of heterozygosity on 10q.23.3 and mutation of the tumor suppressor gene PTEN in benign endometrial cysts of the ovary: possible sequence progression from benign endometrial cyst to endometrioid carcinoma and clear cell carcinoma of the ovary. Cancer Res 60:7052–7056 Sawh RN, Malpica A, Deavers MT et al (2003) Benign cystic mesothelioma of the peritoneum: a clinicopathologic study of 17 cases and immunohistochemical

J. A. Irving and P. B. Clement analysis of estrogen and progesterone receptor status. Hum Pathol 34:369–374 Scattone A, Serio G, Marzullo A et al (2012) High Wilms’ tumor gene (WT1) expression and low mitotic count are independent predictors of survival in diffuse peritoneal mesothelioma. Histopathology 60:472–481 Schlesinger C, Silverberg SG (1999) Tamoxifenassociated polyps (basalomas) arising in multiple endometriotic foci: a case report and review of the literature. Gynecol Oncol 73:305–311 Schmidt CL, Demopoulos RI, Weiss G (1981) Infected endometriotic cysts: clinical characterization and pathogenesis. Fertil Steril 36:27–30 Schoolmeester JK, Park KJ (2015) Incidental nodal lymphangioleiomyomatosis is not a harbinger of pulmonary lymphangioleiomyomatosis: a study of 19 cases with evaluation of diagnostic immunohistochemistry. Am J Surg Pathol 39:1404–1410 Scully RE (1981) Smooth-muscle differentiation in genital tract disorders (Editorial). Arch Pathol Lab Med 105:505–507 Seidman JD (1996) Prognostic importance of hyperplasia and atypia in endometriosis. Int J Gynecol Pathol 15:1–9 Seidman JD, Zhao P, Yemelyanova A (2011) “Primary peritoneal” high-grade serous carcinoma is very likely metastatic from serous tubal intraepithelial carcinoma: assessing the new paradigm of ovarian and pelvic serous carcinogenesis and its implications for screening for ovarian cancer. Gynecol Oncol 120:470–473 Senthong A, Kitkumthorn N, Rattanatanyong P et al (2014) Differences in LINE-1 methylation between endometriotic ovarian cyst and endometriosisassociated ovarian cancer. Int J Gynecol Cancer 24:36–42 Shah R, McCluggage WG (2009) Symplastic atypia in neoplastic and non-neoplastic endometrial stroma: report of 3 cases with a review of atypical symplastic cells within the female genital tract. Int J Gynecol Pathol 28:334–337 Shah R, McCluggage WG (2017) Unclassifiable malignant extraovarian sex cord-stromal tumors: report of 3 cases and review of extraovarian sex cord-stromal tumors. Int J Gynecol Pathol 36:438–446 Shalin SC, Haws AL, Carter DG et al (2012) Clear cell adenocarcinoma arising from endometriosis in abdominal wall cesarean section scar: a case report and review of the literature. J Cutan Pathol 39:1035–1041 Sheffield BS, Hwang HC, Lee AF et al (2015) BAP1 immunohistochemistry and p16 FISH to separate benign from malignant mesothelial proliferations. Am J Surg Pathol 39(7):77–82 Shen J, Pinkus GS, Deshpande V, Cibas ES (2009) Usefulness of EMA, GLUT-1, and XIAP for the cytologic diagnosis of malignant mesothelioma in body cavity fluids. Am J Clin Pathol 131:516–523 Shi M, Fraire AE, Chu P et al (2011) Oncofetal protein IMP3, a new diagnostic biomarker to distinguish

13

Diseases of the Peritoneum

malignant mesothelioma from reactive mesothelial proliferation. Am J Surg Pathol 35:878–882 Shia J, Erlandson R, Klimstra DS (2002) Deciduoid mesothelioma: a report of 5 cases and literature review. Ultrastruct Pathol 26:355–363 Shinozaki-Ushiku A, Ushiku T, Morita S et al (2017) Diagnostic utility of BAP1 and EZH2 expression in malignant mesothelioma. Histopathology 70:722–733 Sidaway MK, Silverberg SG (1987) Endosalpingiosis in female peritoneal washings: a diagnostic pitfall. Int J Gynecol Pathol 6:340–346 Sigel JE, Smith TA, Reith JD et al (2001) Immunohistochemical analysis of anaplastic lymphoma kinase expression in deep soft tissue calcifying fibrous pseudotumor: evidence of a late sclerosing stage of inflammatory myofibroblastic tumor? Ann Diagn Pathol 5:10–14 Simpson JL, Elias S, Malinak LR et al (1980) Heritable aspects of endometriosis. I. Genetic studies. Am J Obstet Gynecol 137:327–331 Singh N, Gilks CB, Hirschowitz L et al (2016) Primary site assignment in tubo-ovarian high-grade serous carcinoma: consensus statement on unifying practice worldwide. Gynecol Oncol 141:195–198 Singhi AD, Krasinskas AM, Choudry HA et al (2016) The prognostic significance of BAP1, NF2, and CDKN2A in malignant peritoneal mesothelioma. Mod Pathol 29:14–24 Siquara de Sousa AC, Capek S, Amrami KK et al (2015) Neural involvement in endometriosis: review of anatomic distribution and mechanisms. Clin Anat 28:1029–1038 Slavin RE, Krum R, Van Dinh T (2000) Endometriosisassociated intestinal tumors: a clinical and pathological study of 6 cases with a review of the literature. Hum Pathol 31:456–463 Sohar E, Gafni J, Pras M et al (1967) Familial Mediterranean fever. A survey of 470 cases and review of the literature. Am J Med 43:227–253 Staats PN, Clement PB, Young RH (2007) Primary endometrioid adenocarcinoma of the vagina: a clinicopathologic study of 18 cases. Am J Surg Pathol 31:1490–1501 Staats PN, McCluggage WG, Clement PB et al (2008) Luteinized thecomas (thecomatosis) of the type typically associated with sclerosing peritonitis: a clinical, histopathologic, and immunohistochemical analysis of 27 cases. Am J Surg Pathol 32:1273–1290 Stanley KE, Utz DC, Dockerty MB (1965) Clinically significant endometriosis of the urinary tract. Surg Gynecol Obstet 120:491–498 Steck WD, Helwig EB (1966) Cutaneous endometriosis. Clin Obstet Gynecol 9:373–383 Steele RW, Dmowski WP, Marmer DJ (1984) Immunologic aspects of human endometriosis. Am J Reprod Immunol 6:33–36 Stern RC, Dash R, Bentley RC et al (2001) Malignancy in endometriosis: frequency and comparison of ovarian

839 and extraovarian types. Int J Gynecol Pathol 20:133–139 Stewart CJ (2013) Deciduoid mesothelial hyperplasia of the pelvic peritoneum. Histopathology 63:598–600 Stewart CJR, Bharat C (2016) Clinicopathological and immunohistological features of polypoid endometriosis. Histopathology 68:398–404 Sumathi VP, McCluggage WG (2002) CD10 is useful in demonstrating endometrial stroma at ectopic sites and confirming a diagnosis of endometriosis. J Clin Pathol 55:391–392 Sun CJ, Toker C, Masi JD et al (1979) Primary low grade adenocarcinoma occurring in the inguinal region. Cancer (Phila) 44:340–345 Sussman J, Rosai J (1990) Lymph node metastasis as the initial manifestation of malignant mesothelioma. Report of six cases. Am J Surg Pathol 14:818–828 Sznurkowski JJ, Emerich J (2008) Endometriomas are more frequent on the left side. Acta Obstet Gynecol Scand 87:104–106 Szyfelbein WM, Baker PM, Bell DA (2004) Superficial endometriosis of the cervix: a source of abnormal glandular cells on cervicovaginal smears. Diagn Cytopathol 30:88–91 Takano M, Yoshikawa T, Kato M et al (2009) Primary clear cell carcinoma of the peritoneum: report of two cases and a review of the literature. Eur J Gynaecol Oncol 30:575–578 Tanaka Y, Mori T, Ito F et al (2017) Exacerbation of endometriosis due to regulatory T-cell dysfunction. J Clin Endocrinol Metab 102:3206–3217 Tang KW, Lamaro V, Jaworski R (2009) Peritoneal histiocytic reaction associated with oxidised regenerated cellulose, a form of mucicarminophilic histiocytosis. Pathology 41:598–600 Tang K, Lyons S, Valmadre S et al (2010) Endometriosis with myxoid change mimicking pseudomyxoma peritonei. Lett Pathol 42:95–97 Tani H, Sato Y, Ueda M et al (2016) Role of versican in the pathogenesis of peritoneal endometriosis. J Clin Endocrinol Metab 101:4349–4356 Truong LD, Maccato ML, Awalt H et al (1990) Serous surface carcinoma of the peritoneum: a clinicopathologic study of 22 cases. Hum Pathol 21:99–110 Ulbright TM, Morley DJ, Roth LM et al (1983) Papillary serous carcinoma of the retroperitoneum. Am J Clin Pathol 79:633–637 Vadlamudi G, Graebe R, Khoo M et al (1997) Gallstones implanting in the ovary. A complication of laparoscopic cholecystectomy. Arch Pathol Lab Med 121:155–158 Valente K, Blackham AU, Levine E et al (2016) A histomorphologic grading system that predicts overall survival in diffuse malignant peritoneal mesothelioma with epithelioid subtype. Am J Surg Pathol 40:1243–1248 Vang R, Shih IM, Kurman RJ (2013) Fallopian tube precursors of ovarian low- and high-grade serous neoplasms. Histopathology 62:44–58

840 Veldhuis WB, Akin O, Goldman D et al (2013) Peritoneal inclusion cysts: clinical characteristics and imaging features. Eur Radiol 23:1167–1174 Vuong PN, Guyot H, Moulin G et al (1990) Pseudotumoral organization of a twisted epiploic fringe or ‘hard-boiled egg’ in the peritoneal cavity. Arch Pathol Lab Med 114:531–533 Walter AJ, Hentz JG, Magtibay PM et al (2001) Endometriosis: correlation between histologic and visual findings at laparoscopy. Am J Obstet Gynecol 184:1407–1413 Wanless IR, Bernier V (1983) Fibrous thickening of the splenic capsule. Arch Pathol Lab Med 107:595–599 Wei J, William J, Bulun S (2011) Endometriosis and ovarian cancer: a review of clinical, pathologic, and molecular aspects. Int J Gynecol Pathol 30:553–568 Weinstein MP, Iannini PB, Stratton CW et al (1978) Spontaneous bacterial peritonitis. A review of 28 cases with emphasis on improved survival and factors influencing prognosis. Am J Med 64:592–598 Weir M, Bell DA, Young RH (1998) Grade 1 peritoneal serous carcinomas. A report of 14 cases and comparison with 7 peritoneal serous psammocarcinomas and 19 peritoneal serous borderline tumors. Am J Surg Pathol 22:849–862 White J, Chan Y-F (1994) Lipofuscinosis peritonei associated with pregnancy-related ectopic decidua. Histopathology (Oxf) 25:83–85 Wiegand KC, Shah SP, Al-Agha O et al (2010) ARID1A mutations in endometriosis-associated ovarian carcinomas. New Engl J Med 363:1532–1543 Williams TJ, Pratt JH (1977) Endometriosis in 1,000 consecutive celiotomies: incidence and management. Am J Obstet Gynecol 129:245–250 Wu DC, Hirschowit S, Natarajan S (2005) Ectopic decidua of pelvic lymph nodes: a potential diagnostic pitfall. Arch Pathol Lab Med 129:e117–e120 Wu Y, Strawn E, Basir Z et al (2006) Genomic alterations in ectopic and eutopic endometria of women with endometriosis. Gynecol Obstet Investig 62:148–159 Xiao W, Awadallah A, Xin W (2012) Loss of ARID1A/ BAF250a expression in ovarian endometriosis and clear cell carcinoma. Int J Clin Exp Pathol 5:642–650 Yamamoto S, Tsuda H, Takano M et al (2008) Clear-cell adenofibroma can be a clonal precursor for clear-cell adenocarcinoma of the ovary: a possible alternative ovarian clear-cell carcinogenic pathway. J Pathol 216(1):103–110 Yamamoto S, Tsuda H, Miyai K et al (2010) Cumulative alterations of p27Kip1-related cell-cycle regulators in

J. A. Irving and P. B. Clement the development of endometriosis-associated ovarian clear cell adenocarcinoma. Histopathology 56:740–749 Yantiss RK, Clement PB, Young RH (2000) Neoplastic and pre-neoplastic changes in gastrointestinal endometriosis: a study of 17 cases. Am J Surg Pathol 24:513–524 Yantiss RK, Clement PB, Young RH (2001) Endometriosis of the intestinal tract. A study of 44 cases of a disease that may cause diverse challenges in clinical and pathological evaluation. Am J Surg Pathol 25:445–454 Yantiss RK, Nielsen GP, Lauwers GY et al (2003) Reactive nodular fibrous pseudotumor of the gastrointestinal tract and mesentery. Am J Surg Pathol 27:532–540 Yokoyama N, Yasuda R, Ichida K et al (2014) Recurrent peritoneal inclusion cysts successfully treated with oral contraceptives: a report of two cases. Clin Exp Obstet Gynecol 41:83–86 Young RH, Clement PB (1996) Müllerianosis of the urinary bladder. Mod Pathol 9:731–737 Young RH, Clement PB (2000) Endocervicosis involving the uterine cervix: a report of four cases of a benign process that may be confused with deeply invasive endocervical adenocarcinoma. Int J Gynecol Pathol 19:322–328 Young RH, Scully RE (1986) Testicular and paratesticular tumors and tumor-like lesions of ovarian common epithelial and müllerian types. A report of four cases and review of the literature. Am J Clin Pathol 86:146–152 Young RH, Prat J, Scully RE (1984) Endometrioid stromal sarcomas of the ovary. A clinicopathologic analysis of 23 cases. Cancer (Phila) 53:1143–1155 Young RH, Clement PB, McCaughey WTE (1990) Solitary fibrous tumors (“fibrous mesotheliomas”) of the peritoneum: a report of three cases. Arch Pathol Lab Med 114:493–495 Young RH, Eichhorn JH, Dickersin GR et al (1992) Ovarian involvement by the intra-abdominal desmoplastic small round cell tumor with divergent differentiation: a report of three cases. Hum Pathol 23:454–464 Zanetta GM, Webb MJ, Li H et al (2000) Hyperestrogenism: a relevant risk factor for the development of cancer from endometriosis. Gynecol Oncol 79:18–22 Zhang PJ, Goldblum JR, Pawel BR et al (2003) Immunophenotype of desmoplastic small round cell tumors as detected in cases with EWS/WT1 gene fusion product. Mod Pathol 16:229–235 Zinsser KR, Wheeler JE (1982) Endosalpingiosis in the omentum. A study of autopsy and surgical material. Am J Surg Pathol 6:109–117

Epithelial Tumors of the Ovary

14

Jeffrey D. Seidman, Brigitte M. Ronnett, Ie-Ming Shih, Kathleen R. Cho, and Robert J. Kurman

Contents Epidemiology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 842 Geographic Distribution, Incidence, and Mortality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 842 Etiology and Risk Factors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 844 Morphologic and Molecular Pathogenesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 846 A New Model of Ovarian Cancer Pathogenesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 846 Putative Histopathologic Precursor Lesions of Ovarian Cancer . . . . . . . . . . . . . . . . . . . . . . . 850

J. D. Seidman (*) Center for Devices and Radiological Health, Office of In Vitro Diagnostics and Radiological Health, Food and Drug Administration, Silver Spring, MD, USA e-mail: [email protected]; [email protected] B. M. Ronnett Department of Pathology, Division of Gynecologic Pathology, Johns Hopkins University School of Medicine, Baltimore, MD, USA e-mail: [email protected] I.-M. Shih Gynecologic Pathology Laboratory in the Department of Gynecology and Obstetrics, Johns Hopkins University School of Medicine, Baltimore, MD, USA e-mail: [email protected] K. R. Cho Department of Pathology, University of Michigan Medical School, Ann Arbor, MI, USA e-mail: [email protected] R. J. Kurman Department of Gynecology, Obstetrics, Pathology and Oncology, Division of Gynecologic Pathology, Johns Hopkins University School of Medicine, Baltimore, MD, USA e-mail: [email protected] # Springer Nature Switzerland AG 2019 R. J. Kurman et al. (eds.), Blaustein’s Pathology of the Female Genital Tract, https://doi.org/10.1007/978-3-319-46334-6_14

841

842

J. D. Seidman et al. Familial (Hereditary) Ovarian Cancer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 856 BRCA1, BRCA2, and Other Genes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 856 Clinicopathologic Features of Familial Ovarian Cancers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 858 GEMMS of Ovarian Cancer and Translational Applications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 858 Mouse Models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 858 Translational Applications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 859 Screening, Early Diagnosis, and Prevention . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Screening Tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Early Diagnosis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Prevention . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

859 859 861 861

Prognostic Factors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Cell Type and Histologic Grade . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Other Prognostic Factors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Stage, Patterns of Spread, and Survival . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

862 862 863 863

Cytopathology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 868 Fine-Needle Aspiration of Ovarian Cysts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 868 Peritoneal Fluid Cytology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 869 Treatment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Surgery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Chemotherapy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Targeted Therapy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Other Therapeutic Modalities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

870 870 870 871 871

Pathology of Ovarian Epithelial Neoplasms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 872 The College of American Pathologists (CAP), WHO, International Collaboration of Cancer Reporting (ICCR), and FIGO Recommendations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 873 Serous Tumors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 874 Mucinous Tumors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 906 Endometrioid Tumors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 923 Clear Cell Tumors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 935 Brenner (Transitional Cell) Tumors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 943 Squamous Tumors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 948 Mixed Epithelial Tumors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 948 Undifferentiated Carcinomas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 949 Carcinosarcoma (Malignant Mixed Mesodermal/Mullerian Tumor) . . . . . . . . . . . . . . . . . . . 949 Sarcomas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 950 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 951

Epidemiology Geographic Distribution, Incidence, and Mortality Worldwide, ovarian cancer is the seventh most common cancer in women and the eighth most common cause of cancer death. There are about 239,000 new cases and 152,000 deaths annually (American Cancer Society 2015). In the Western hemisphere, ovarian cancer accounts for 4% of cancer in women and is the most frequent cause of death due to gynecological cancer. In American

women, ovarian cancer represents 3% of all new cancers (SEER 2017). It ranks 11th in incidence and 5th in mortality and accounts for 5% of cancer deaths. It is estimated that in the USA in 2017, there were 22,440 new ovarian cancer cases and 14,080 deaths, making it the most lethal gynecologic malignancy (Siegel et al. 2017). Approximately 1.3% of American women will develop ovarian cancer in their lifetime, or 12 new cases per 100,000. In general, the disease is more common in industrialized countries where parity is lower, but there are notable exceptions such as Japan which has a low parity and low rate of

14

Epithelial Tumors of the Ovary

ovarian cancer. The lifetime risk varies widely from 0.45% in Japan to 1.7% in Sweden. Annual incidence rates of ovarian cancer are lower in developing countries as compared to developed countries, averaging 5.0 and 9.1 per 100,000, respectively. Similarly, mortality rates average 3.1 and 5.0 per 100,000, respectively (American Cancer Society 2015). Denmark and other Scandinavian countries have among the highest annual incidence rates at greater than 16 per 100,000 women. Interestingly, as compared to all other common cancers, ovarian cancer varies the least in age-standardized incidence rates across registry populations worldwide (Bray et al. 2015). In the USA, incidence and mortality rates have been declining by 1.9% and 2.2% per year, respectively, from 2004 to 2013 (SEER 2017). Over longer periods, the incidence has been relatively stable worldwide from the 1970s to the 2000s. There have been, however, small but significant increases in Eastern and Southern Europe and Asia and decreases in Northern Europe and North America (Coburn et al. 2017). In addition, the incidence of ovarian carcinoma appears to have decreased as peritoneal and tubal carcinomas have increased, reflecting a recent change in classification. Migration studies have shown that ovarian cancer rates approach those of the place of immigration rather than the place of emigration, suggesting a significant environmental component to ovarian cancer risk. Ovarian cancer rates vary among different ethnic groups. White women have higher rates than black and Asian women. In the USA, incidence rates vary from 9.3 per 100,000 in Asians to 12.5 per 100,000 in whites, while mortality rates range from 4.5 per 100,000 in Asians to 7.8 per 100,000 in whites. Also in the USA, African-American women have an incidence three-fourths that of white women, with a 17% lower mortality rate (SEER 2017). In Israel, Jewish women have an eightfold higher risk compared to non-Jewish women. This is in part due to the high frequency of BRCA mutations (2.5%) in Ashkenazi Jewish women. There are global geographic differences in the incidence of the different histologic types (Sung et al. 2014). For example, clear cell carcinoma (CCC) is more common in Asians,

843

particularly in Japanese, and in Asian Americans as compared to Caucasians in the USA (Anglesio et al. 2011; Fuh et al. 2015; Yamamoto et al. 2011). Interestingly, Surveillance, Epidemiology, and End Results (SEER) data indicate that the incidence of CCC among Asian Americans is increasing (Park et al. 2017). Endometrioid carcinomas also appear to be relatively more common in Asia, whereas serous carcinomas comprise a lower proportion of ovarian cancers (Coburn et al. 2017). For 2005–2009, the overall 5-year survival for all stages (including all histologic types) was 40% or greater in North America and in many Asian and European countries. Most other countries have 5-year overall survival in the 30–40% range (Allemani et al. 2015). Population-based data on ovarian cancer incidence often do not accurately reflect a woman’s actual risk because such data do not correct for those who have undergone bilateral salpingooophorectomy (BSO), a commonly used procedure in the USA. Though BSO is not 100% protective (see section “Screening, Early Diagnosis, and Prevention,” later in this chapter), for practical purposes, women who have undergone BSO have a risk close to zero, and accordingly, populationbased data underestimate the risk of women who have not had BSO. A population-based study from Kentucky estimated that corrected age-adjusted rates of ovarian cancer that considered the protective effect of BSO were one-third to two-thirds higher than the rates from the standard population (Baldwin et al. 2017). In other words, populationbased incidence rates, since they do not apply to those who have had BSO, underestimate the risk for the remaining women. Epidemiologic studies of ovarian cancer rely on accurate tumor classification. Variable and changing proportions of high-grade serous carcinomas (HGSCs) being classified as of tubal and peritoneal origin as compared to the historical practice of classifying nearly all extrauterine HGSCs as of ovarian origin have undoubtedly influenced epidemiologic studies. It has also become apparent that a large proportion of mucinous carcinomas involving the ovaries are metastatic from elsewhere, and accordingly older studies and studies without centralized pathology

844

review are not as reliable (see section “Invasive Mucinous Carcinomas,” later in this chapter). The large volume of literature on atypical proliferative (borderline) tumors belies their low prevalence. Large institutions encounter only about three atypical proliferative serous tumors (APSTs) per year in-house. Population-based incidence data on borderline tumors are scant; in the USA, the annual incidence is 2.5 per 100,000 (1.5 per 100,000 for serous borderline tumors (SBTs) in white women) and in Sweden, 6.6 per 100,000 (all borderline tumors) (Mink et al. 2002).

Etiology and Risk Factors Age Ovarian cancer rates increase with age. In the USA, 5% of new cases occur in women under age 35, while 70% occur over age 54. The annual incidence steadily increases from less than 3 per 100,000 in women under age 30 and plateaus at 48 per 100,000 in the 80–84-year age group. The median age is about 63 years at diagnosis, and the median age at death is 70 years. The mean age varies substantially among subgroups, a reflection of differences in hereditary syndromes and differences in the pathogenesis of different types of ovarian carcinoma (Table 1) (see section “Morphologic and Molecular Pathogenesis,” later in this chapter).

J. D. Seidman et al.

Many studies have shown that age is an independent prognostic factor, but these data are difficult to evaluate as there are multiple interacting factors involved. To some extent, age-related prognostic differences are largely explained by higher proportions of low-grade, low-stage, optimally debulked, and Type I tumors (see sections “Morphologic and Molecular Pathogenesis” and “Prognostic Factors,” later in this chapter) in younger patients, who are generally healthier than older patients. The latter are more likely to have a lower performance status and more comorbid conditions and are less likely to respond to chemotherapy. Older women are also treated less aggressively than younger women; in the SEER database, over 40% of American women with ovarian cancer older than 85 years received only palliative treatment. The risk of death in the presence of comorbid conditions is 30–40% higher than those without such conditions. Accordingly, younger women with invasive ovarian cancer generally have more favorable features and thus a better prognosis, even when stratified by stage.

Reproductive Factors, Ovulation, and Hormonal Factors Reproductive factors are well documented to affect ovarian cancer risk. Increased parity is a well-established protective factor. The parityassociated risk reduction wanes with age and disTable 1 Mean age in subgroups of ovarian cancer (years). appears over age 75 (McGuire et al. 2016). The (Modified from 6th ed Table 1) relative effects of most other risk factors do not Lynch syndrome-associated carcinoma 43 appear to differ by parity (Bodelon et al. 2013). Noninvasive low-grade serous carcinoma 43–45 Oral contraceptive pills (OCPs) decrease ovarian Endometrioid carcinoma arising in 50 cancer risk to a similar extent in high-risk and endometriosis Endometrioid carcinoma associated with 50 average-risk women (Cibula et al. 2011b; uterine endometrioid carcinoma Moorman et al. 2013). Many studies report a FIGO stage I carcinoma 53 50% risk reduction for ovarian cancer in women Clear cell carcinoma 50–53 who have been on OCPs. In a large meta-analysis, BRCA1-associated carcinoma 54 it was shown that the longer women used OCPs, Low-grade serous carcinoma 56 the greater the reduction of risk ( p < 0.0001). Endometrioid carcinoma 55–58 Moreover, the reduction in risk persisted for BRCA2-associated carcinoma 59 more than 30 years after stopping OCP use (ColFIGO stage III high-grade serous carcinoma 62–64 laborative Group on Epidemiological Studies of Carcinosarcoma 64–66 Ovarian Cancer 2008).

14

Epithelial Tumors of the Ovary

Early menarche and late menopause are also significant risk factors. Pregnancies appear to be more strongly protective for endometrioid and CCC as compared to serous carcinoma. High socioeconomic status is associated with an increased ovarian cancer risk and lower fertility. A population-based study from Norway found that nulliparous women treated with fertility drugs, particularly clomiphene citrate, have an increased risk; the risk was not significantly increased for parous women (Reigstad et al. 2017). Several meta-analyses of hormone replacement therapy and ovarian cancer risk have shown a slightly elevated risk with odds ratios of 1.1–1.3. Since a 2002 report confirming the increased risk with menopausal hormonal therapy, the age-standardized ovarian cancer incidence in women over age 50 declined by 2.4% per year as compared to 0.8% per year before 2002 (Yang et al. 2013). The protective effect of increased parity and oral contraceptive use applies only to non-mucinous tumors. Surgically induced protective factors include hysterectomy (Jordan et al. 2013), tubal ligation, and BSO. The mechanisms for risk reduction with hysterectomy and tubal ligation are unclear, although both can prevent passage of endometrial tissue via retrograde menstruation which is one of the proposed mechanisms for the development of endometriosis, and endometriosis is a precursor of some ovarian cancers (see sections “Putative Histopathologic Precursor Lesions of Ovarian Cancer” and “Endometriosis,” later in this chapter). This mechanism is further supported by the finding that tubal ligation is more strongly protective against clear cell and endometrioid carcinomas (50% or more reduction in frequency) as compared to serous carcinoma (about 20% reduction) (Cibula et al. 2011a; Sieh et al. 2013). In addition, hysterectomy and tubal ligation reduce or prevent potential environmental carcinogens from entering the peritoneal cavity and thereby contacting tubal and ovarian tissue. Opportunistic salpingectomy greatly reduces the risk of ovarian cancer as demonstrated in a recent study comparing women who had a previous salpingectomy (HR = 0.65, 95% CI = 0.52–0.81) to the unexposed population (Falconer et al. 2015),

845

suggesting that this procedure alone can substantially reduce ovarian cancer risk. The most commonly cited mechanism for the etiology of ovarian cancer is the incessant ovulation hypothesis which proposes that ovulation traumatizes the ovarian surface epithelium (OSE) and thus stimulates proliferation, creating a milieu that predisposes to malignant transformation. This hypothesis had been supported by epidemiologic observations correlating a decreased number of lifetime ovulations, as occurs with increased parity or OCP use, and the risk of ovarian cancer (Yang et al. 2016). There are, however, mechanisms other than trauma to the OSE that can explain the correlation between ovarian cancer risk and ovulation. Recent studies implicate the fallopian tube as the site of origin of most HGSCs (see section “Morphologic and Molecular Pathogenesis,” later in this chapter), and it has been noted that women who incessantly ovulate also menstruate more frequently, suggesting that exposure of the fimbria to blood from retrograde menstruation could lead to iron-induced oxidative stress (Vercellini et al. 2011). Old blood (hemosiderin and/or pseudoxanthoma cells) is seen in 5% of grossly normal fallopian tubes (Seidman et al. 2016) and in 20% of tubes of women with advanced-stage extrauterine HGSC (Seidman 2013). Follicular fluid also contains reactive oxygen species which are potential carcinogens and may thereby increase risk from incessant ovulation (Bahar-Shany et al. 2014).

Inflammation Inflammation is associated with a wide range of diseases including cardiovascular disease, autoimmune diseases, osteoarthritis, inflammatory bowel disease, and cancer. It has been suggested that inflammation, with concomitant rapid cellular proliferation, oxidative stress, and elevation of cytokines and growth factors, may lead to DNA damage that could result in the development of carcinoma. Inflammation incited by ovulationinduced damage to the ovarian surface and pelvic inflammatory disease (PID) have been the most commonly proposed sources by which inflammation could lead to the development of ovarian cancer. The reported association of PID with

846

ovarian carcinoma risk is inconsistent; this risk may be significant only for serous carcinomas, with a hazard ratio of 1.19 in a population-based Danish study, but was not significant in a pooled analysis of 13 case-control studies (Rasmussen et al. 2017a, b). SBTs are more consistently and strongly associated with PID as well as infertility (Rasmussen et al. 2017c; Seidman et al. 2002). It would be expected that hysterectomy and bilateral tubal ligation, by preventing infectious agents from reaching the ovary, would reduce the risk of ovarian cancer; most adequately powered studies have demonstrated a reduction in risk, but a few have shown no effect (Vitonis et al. 2011). Similarly, aspirin and nonsteroidal anti-inflammatory drugs (NSAIDs) should reduce the risk of ovarian cancer if inflammation played an important role, but here again studies have not been consistent (Merritt et al. 2008; Trabert et al. 2014). Further observational and experimental data will be necessary to confirm whether inflammation plays a significant role in ovarian carcinogenesis.

Other Risk Factors Other potential risk factors have been studied, but associations with ovarian cancer risk have been weak, inconclusive, or negative. These include age at birth of first child, breastfeeding, diet (Crane et al. 2014), smoking, certain types of viral infections in childhood, and ionizing radiation. A recent systematic review showed that body mass index (BMI), weight, and height are all positively associated with ovarian cancer risk (Aune et al. 2015). A pooled analysis of 39 studies showed that a genetically predicted higher BMI was associated with increased risk only for non-HGSCs (Dixon et al. 2016). Diabetes mellitus also appears to increase risk even after controlling for BMI (Lee et al. 2013a). Smoking appears to be associated only with an increased risk of mucinous tumors (Licaj et al. 2017). Some epidemiologic studies suggest an association between genital use of talc powders and increased risk of ovarian cancer, but the evidence is not consistent. A recent meta-analysis evaluating 24 case-control studies and 3 cohort studies, including 302,705 women with ovarian cancer, showed a relative risk (RR) for ever use of genital

J. D. Seidman et al.

talc and ovarian cancer of 1.22 [95% confidence interval (CI): 1.13–1.30]. The RR for case-control studies was 1.26 (95% CI: 1.17–1.35) and for cohort studies, 1.02 (95% CI: 0.85–1.20). Serous carcinoma was the only histologic type for which a significant association was detected (RR: 1.24; 95% CI: 1.15–1.34). Although statistically significant, these are weak associations. The investigators concluded that several aspects of their study, including the heterogeneity of results between case-control and cohort studies, and the lack of a dose-response with duration and frequency of use did not support a causal role of talc exposure and ovarian cancer (Berge et al. 2018).

Morphologic and Molecular Pathogenesis The biological manifestations of ovarian carcinoma can be divided into two broad phases: malignant transformation and peritoneal dissemination. Until recently, some investigators believed that benign, “borderline,” and malignant ovarian tumors reflect sequential steps in malignant transformation, regardless of their cell type (i.e., serous, mucinous, endometrioid, or clear cell). Accordingly, many studies in the cell biology and molecular biology literature juxtaposed these entities in an attempt to elucidate the events in ovarian carcinogenesis and often considered the different histologic types of ovarian tumors essentially interchangeable. Further, the traditional view of peritoneal dissemination was that carcinoma begins in the ovary, undergoes progressive dedifferentiation from a well to a poorly differentiated tumor, and then spreads through the peritoneal cavity before metastasizing to distant sites. These time-honored views now do not appear to be valid for the majority of ovarian cancers as described below.

A New Model of Ovarian Cancer Pathogenesis A new model of ovarian carcinogenesis has been developed that correlates the molecular and

14

Epithelial Tumors of the Ovary

morphologic features of all the histologic types of borderline tumors and invasive carcinomas. This model challenges many of the timehonored concepts of ovarian neoplasia and has important implications for prevention, early detection, and treatment. In addition, recent studies have provided evidence that the origin of ovarian cancer, traditionally regarded as the OSE, is more likely the fallopian tube and other non-ovarian tissues such as endometriosis. Moreover, the view that ovarian cancer begins in the ovary and spreads systematically to the pelvis, abdomen, and then distant sites and that it progresses from well to poorly differentiated does not appear to be valid. This section will broadly summarize these new developments, while the individual sections on the different cell types will go into greater detail. The new model divides surface epithelial tumors into two broad categories: Type I and Type II, based on their clinicopathologic features and characteristic molecular genetic changes. Type I and Type II refer to tumorigenic pathways and are not histopathologic diagnostic terms (Table 2). This classification system will

847

undoubtedly continue to evolve, particularly as the molecular changes that characterize the less common types of ovarian carcinomas and the distinction between morphologically overlapping categories (e.g., high-grade serous versus highgrade endometrioid) become more clearly delineated Cho and Shih 2009. Type I tumors are low-grade, relatively indolent neoplasms that arise from well-characterized precursor lesions (atypical proliferative [borderline] tumors and endometriosis) and usually present as large stage I neoplasms. This group includes low-grade serous (invasive micropapillary serous carcinoma (MPSC)) and low-grade endometrioid, mucinous, seromucinous, and CCC. The latter, although it exhibits most of the features of the Type I tumors, such as association with a well-established precursor (endometriosis) and frequent presentation in stage I, is typically high grade unlike the other Type I tumors. Nonetheless, molecular genetic data show a greater similarity of CCC to Type I as compared to Type II tumors. Type I tumors often harbor somatic mutations of genes encoding protein kinases including

Table 2 Type I versus Type II ovarian carcinoma: precursors and molecular features Most frequent mutations

Chromosomal instabilitya

KRAS, BRAF

Low

CTNNB1, PTEN

Low

PIK3CA

Low

KRAS

Low

?p53 signature, STIC U

TP53 TP53

High High

U STIC

U TP53

Probably highb U

Common precursors Type I tumors Low-grade serous carcinoma (LGSC) Low-grade endometrioid CA Clear cell carcinoma Mucinous carcinoma Type II tumors High-grade serous carcinoma High-grade endometrioid carcinoma Undifferentiated carcinomab Carcinosarcoma

?PTH, APST, noninvasive LGSC Endometriosis/ endometriotic cyst Endometriosis/ endometriotic cyst APMT

? Proposed but needs confirmation PTH papillary tubal hyperplasia, APST atypical proliferative serous tumor, LGSC low-grade serous carcinoma, APMT atypical proliferative mucinous tumor, U unknown, STIC serous tubal intraepithelial carcinoma a Low versus high chromosomal instability refers to comparison between low-grade and high-grade carcinomas within the same histologic type b Most undifferentiated carcinomas are probably HGSC, SET variant, but very few cases have been studied

848

J. D. Seidman et al.

KRAS, BRAF, PIK3CA, and ERBB2, as well as other signaling molecules including PTEN and CTNNB1 (β-catenin). The atypical proliferative

Re me trogra nst d rua e tion

CCC EMC

EM/CC atypical proliferative tumor Endometriosis

Fig. 1 Schematic of the proposed origin of type I tumors. Retrograde menstruation leads to the development of endometriosis and atypical proliferative tumors which can give rise to endometrioid and CCC Fig. 2 Schematic of proposed mechanism for the origin of ovarian surface epithelial inclusions from fimbrial epithelium. Epithelial cells from fimbria implant on site of rupture on the ovary where ovulation occurred. These epithelial cells can conceivably invaginate to form a cortical inclusion cyst (CIC)

or borderline serous and mucinous tumors in turn appear to develop from cystadenomas, while the atypical proliferative endometrioid and clear cell tumors arise from endometriosis, typically endometriotic cysts (endometriomas) (Fig. 1). In contrast, Type II tumors, of which the vast majority are HGSCs, are aggressive, high-grade neoplasms from the outset; in the past they have been said to arise “de novo.” Recent data, however, suggest that HGSCs arise from intraepithelial carcinomas, the majority of which have been detected in the tubal fimbriae (Fig. 2). For practical purposes, TP53 mutations are found in virtually all HGSCs (Vang et al. 2016). Interestingly, the few examples of stage I HGSCs that have been studied have been shown to harbor mutations of TP53. Therefore, it appears that conventional HGSC even in its earliest stage of development resembles advanced-stage serous carcinoma at both the morphologic and molecular level. Of further interest has been the detection of TP53 mutations in serous tubal intraepithelial carcinoma (STIC) and in a recently described putative precursor lesion of tubal intraepithelial

Fimbria

Ovulation site Ovary

14

Epithelial Tumors of the Ovary

carcinoma designated as “p53 signature” (Lee et al. 2007). The latter lesion appears morphologically normal but overexpresses p53 and can harbor a TP53 mutation. These findings highlight the importance of this mutation in the early development of HGSC which has important clinical implications (see section “Screening, Early Diagnosis, and Prevention”). The anatomical progression of ovarian carcinoma is beginning to be more clearly understood. It had been assumed that carcinoma originates in the ovary, is confined to the ovary for a period of time, and then disseminates to the pelvis, followed by the abdominal cavity before spreading to distant sites. This view underlies the basis of the International Federation of Gynecology and Obstetrics (FIGO) staging system in which tumors confined to the ovary are stage I; those involving the pelvic organs, stage II; with involvement of abdominal organs, stage III; and when spread occurs to distant sites, stage IV (Table 3). There are, however, significant problems with this assumption, which now appears to be accurate for only a minority of ovarian carcinomas (Type I tumors). Clinicopathologic comparison of stage I with stage III carcinomas shows that stage I tumors are predominantly Type I and non-serous while stage III tumors are predominantly Type II (Yemelyanova et al. 2008a). It is important to recognize that Type II tumors account for the vast majority (85–90%) of ovarian cancer deaths (Seidman et al. 2015; Temkin et al. 2017). Most Type II tumors are HGSCs and at the time of diagnosis are widely disseminated throughout the peritoneum (FIGO stages III and IV) with the largest volume of tumor often outside the ovaries. Serous carcinoma and its variants (peritoneal serous carcinoma, carcinosarcoma, undifferentiated carcinoma, and mixed carcinomas with a high-grade serous component) account for 87% of cases of carcinomatosis from ovarian carcinoma and accordingly the majority of ovarian cancer deaths (Table 4). Like other cancers, ovarian carcinomas arise through a multi-step process in which clonal selection acts on cells with somatic mutations and altered gene expression to allow outgrowth of progeny with increasingly aggressive growth properties. The genes mutated in cancer are not selected randomly, but frequently encode proteins

849 Table 3 2012 FIGO staging of ovarian carcinoma Stage I: tumor confined to ovaries or fallopian tube(s) (T1N0M0) IA: tumor limited to 1 ovary (capsule intact) or fallopian tube; no tumor on ovarian or fallopian tube surface; no malignant cells in ascites or peritoneal washings (T1aN0M0) IB: tumor limited to both ovaries (capsules intact) or fallopian tubes; no tumor on ovarian or fallopian tube surface; no malignant cells in ascites or peritoneal washings (T1bN0M0) IC: tumor limited to 1 or both ovaries or fallopian tubes, with any of the following: IC1: surgical spill (T1c1N0M0) IC2: capsule ruptured before surgery or tumor on ovarian or fallopian tube surface (T1c2N0M0) IC3: malignant cells in ascites or peritoneal washings (T1c3N0M0) Stage II: tumor involves 1 or both ovaries or fallopian tubes with pelvic extension (below pelvic brim) or primary peritoneal cancer (T2N0M0) IIA: extension and/or implants on uterus and/or fallopian tubes and/or ovaries (T2aN0M0) IIB: extension to other pelvic intraperitoneal tissues (T2bN0M0) Stage III: tumor involves 1 or both ovaries or fallopian tubes, or primary peritoneal cancer, with cytologically or histologically confirmed spread to the peritoneum outside the pelvis and/or metastasis to retroperitoneal lymph nodes (T1/T2 N1M0) IIIA1: positive retroperitoneal lymph nodes only (cytologically or histologically proven): IIIA1(i): metastasis up to 10 mm in greatest dimension IIIA1(ii): metastasis more than 10 mm in greatest dimension IIIA2: microscopic extrapelvic (above the pelvic brim) peritoneal involvement with or without positive retroperitoneal lymph nodes (T3a2N0/N1M0) IIIB: macroscopic peritoneal metastasis beyond the pelvis up to 2 cm in greatest dimension, with or without metastasis to retroperitoneal lymph nodes (T3bN0/N1M0) IIIC: macroscopic peritoneal metastasis beyond the pelvis more than 2 cm in greatest dimension, with or without metastasis to retroperitoneal lymph nodes (includes extension of tumor to capsule of liver and spleen without parenchymal involvement of either organ) (T3cN0/N1M0) Stage IV: distant metastasis excluding peritoneal metastases (any T, any N, M1) IVA: pleural effusion with positive cytology IVB: parenchymal metastases and metastases to extraabdominal organs (including inguinal lymph nodes and lymph nodes outside of the abdominal cavity) From Prat et al. (2015)

850

J. D. Seidman et al.

Table 4 Distribution of ovarian carcinoma by cell type and FIGO stage HG serous LG serous Endometrioid Clear cell Carcinosarcoma Transitional Mucinous Seromucinous Mixed Undifferentiated Squamous Totals (%)

Stage I 11 2 21 25 1 6 13 4 4 0 0 87 (15.5)

Stage II 18 0 8 8 5 0 1 1 3 1 0 45 (8.0)

Stage III 248 22 10 10 21 0 1 0 4 1 0 317 (56.3)

Stage IV 94 5 2 3 7 0 0 0 2 0 1 114 (20.2)

Totals (%) 371 (65.9) 29 (5.2) 41 (7.3) 46 (8.2) 34 (6.0) 6 (1.1) 15 (2.7) 5 (0.9) 13 (2.3) 2 (0.4) 1 (0.2) 563

This table includes only invasive carcinomas, and also includes tubal and peritoneal primary tumors. Consecutive cases over a 13-year period from a large community hospital HG high grade, LG low grade

that function in highly conserved signaling pathways. Over the past several years, a number of studies have evaluated ovarian epithelial tumors for molecular genetic alterations such as point mutations, gene amplifications, deletions, and translocations. Although a detailed description of the genetic alterations identified to date in ovarian epithelial tumors is beyond the scope of this chapter, some useful themes have emerged. First, while few if any changes appear to be unique to ovarian cancer, studies have shown that certain alterations appear to be particularly characteristic of specific histologic types of ovarian carcinomas. Second, for carcinomas with serous or endometrioid differentiation, specific genetic alterations distinguish low-grade from high-grade carcinomas. Third, identification of shared genetic alterations in some histologic types of ovarian carcinomas and their putative precursor lesions has provided useful insights into the pathogenetic pathways leading to ovarian cancer. Details of the molecular biology of the different cell types of ovarian carcinoma are reviewed in their respective sections later in this chapter. With the emerging view that both HGSC and low-grade serous carcinoma (LGSC) are of tubal origin, and the wellestablished view that endometrioid and CCCs are derived from implantation of endometriotic tissue, it can be argued that the only true primary ovarian carcinomas are germ cell tumors and sex

cord-stromal tumors analogous to those in the testis (Kuhn et al. 2012a).

Putative Histopathologic Precursor Lesions of Ovarian Cancer Surface Epithelial Inclusions (Cortical Inclusion Cysts (CICs)) and Dysplasia The search for precursors of ovarian cancer has been ongoing for at least four decades. Initially, investigators focused on the OSE which for many years was thought to be a distinctive cell type (Auersperg 2013). The general notion was that the OSE invaginated into the underlying stroma to form simple glands and cysts lined by a single layer of flat or nondescript cuboidal epithelium identical to the OSE or, more often, lined by tubal-type epithelium. These small glands and cysts are termed “cortical inclusion cysts (CICs)” (formerly called germinal inclusion cysts). They are morphologically identical to inclusions outside of the ovary where they are classified as “endosalpingiosis.” In the ovary, CICs are more common in older women (see also ▶ Chap. 12, “Nonneoplastic Lesions of the Ovary”). Historically, they had been considered to arise after postovulatory repair of the damaged ovarian surface; however the evidence for this is weak.

14

Epithelial Tumors of the Ovary

Many investigators currently believe that CICs arise from implantation of fimbrial epithelium which is normally in close apposition with the ovarian surface (Banet and Kurman 2015) (Fig. 2). CICs are lined by ciliated and secretory cells identical to those in the fallopian tube and in addition express PAX8 like the fallopian tube (unlike the OSE which expresses calretinin, a mesothelial marker, and generally does not express PAX8). Another intriguing finding in CICs is the presence of CD45(+) leukocytes that are regularly observed in the fallopian tube (Ardighieri et al. 2014a). Over the years it was hypothesized that CICs underwent malignant transformation to ovarian cancer, but arguments have been advanced to demonstrate the weakness of this hypothesis. These have been summarized by Dubeau (1999). First, carcinomas generally resemble the epithelial cells in the organ from which they are derived. The ovary is composed of germ cells and hormonally active stromal cells which do not resemble the various histologic subtypes of ovarian cancer. The latter resemble tissues of mullerian origin (serous tumors resemble tubal epithelium, and endometrioid and clear cell tumors are akin to endometrial tissue). Second, the ovaries are not of mullerian origin. Third, the OSE, as noted earlier, closely resembles mesothelium, as confirmed by expression profile studies showing a similarity of various subtypes of ovarian tumors to the mullerian-derived normal tissues and not to OSE (Marquez et al. 2005). Taking this argument to its logical conclusion, one would expect that if ovarian tumors were derived from the OSE, they should resemble mesotheliomas, not mullerianrelated carcinomas. Initially investigators focused on the OSE near carcinomas to define the putative entity of “ovarian dysplasia,” and a few reported atypical cellular and nuclear features that appear more frequently in OSE near or contralateral to carcinomas, and in ovaries from women at increased risk of ovarian cancer, in comparison to control ovaries. Efforts to reproduce these findings were unsuccessful. Based on molecular genetic studies on human tissue and genetically engineered mouse models (GEMMs) (see below), this cell type is no longer

851

considered by most experts to be a credible source of ovarian carcinomas. Some studies on prophylactic oophorectomy specimens in high-risk women and normalappearing ovaries contralateral to stage I carcinomas have shown a higher number of these CICs in comparison to controls, but others have not confirmed these findings. These studies have also evaluated a variety of other features including cortical invaginations (clefts) and surface papillomatosis, and some have found the latter two to be more common in cancer-prone ovaries than in controls, but these findings have not been confirmed by other investigators. Many of these studies have significant drawbacks and are not strictly comparable to one another (Seidman and Wang 2007). Most recently, studies utilizing GEMMs have provided further evidence implicating mullerian epithelia such as the fallopian tube epithelium (FTE) rather than the OSE as the site of origin of ovarian carcinomas. GEMMs of various types of ovarian cancer have been generated by altering tumor suppressor genes and oncogenes that are frequently mutated in each type of human ovarian cancer in the mouse OSE or FTE. For example, deletion of Brca1, Trp53, and Rb1 in the mouse FTE yields tumors resembling HGSC, while activating canonical Wnt signaling and PI3K/AKT signaling via Apc and Pten deletion yields tumors resembling endometrioid carcinoma. When the same genetic defects (i.e., Apc and Pten deletion) were induced in the FTE vs. the OSE of otherwise genetically identical mice, tumors developing in the GEMMs derived from FTE transformation more closely resembled human ovarian endometrioid carcinomas with respect to their morphology, global gene expression, and biological behavior than tumors derived from OSE transformation (Wu et al. 2016; Zhai et al. 2017). Although the tubal fimbriae are now regarded as the main source of HGSC involving the ovary, it is conceivable that serous carcinomas may also develop from endosalpingiosis. It has been reported that 85% of so-called primary peritoneal serous tumors were associated with

852

endosalpingiosis (Irving and Clement 2011). Other precursor lesions include atypical proliferative (borderline) serous tumors for LGSCs, p53 signatures, and serous tubal intraepithelial carcinomas (STICs) for invasive high-grade serous carcinoma and endometriosis/endometriotic cysts for endometrioid, clear cell, and seromucinous carcinomas.

Endometriosis Endometriosis is common and found in about 10% of reproductive-age women, but the risk of malignant transformation in an individual patient is very low. Nonetheless, endometriosis is associated with up to 20% of ovarian cancers and is acknowledged to be the precursor of most endometrioid, clear cell, and seromucinous carcinomas. Endometriosis is found incidentally in 11–18% of women with HGSC (Ritterhouse et al. 2016; Seidman 2013) but is unlikely to be histogenetically related in such cases. A pooled analysis of 13 case-control studies with 7,911 ovarian cancer patients and 13,226 controls showed that self-reported endometriosis is associated with an increased risk of endometrioid and CCCs and also demonstrated an increased risk of LGSC, while there was no association with HGSC, mucinous carcinoma, or borderline tumors (Pearce et al. 2012). As there is no obvious biologically plausible mechanism to account for the association of endometriosis with LGSC, this interesting finding requires confirmation. Endometriosis-associated carcinomas have the same prognosis as other ovarian carcinomas of the same histologic types when stratified by stage and other established prognostic factors (Kim et al. 2014). The view that endometriosis is an ovarian cancer precursor is supported by our understanding of endometrial cancer precursors. Atypical endometrial hyperplasia in the uterus is a welldefined precursor of endometrial adenocarcinoma (see ▶ Chap. 8, “Precursors of Endometrial Carcinoma”), and changes like this lesion are occasionally observed in endometriosis (Seidman 1996). In addition, atypical changes are also seen in endometriosis near endometrioid adenocarcinomas of the ovary and even more frequently in

J. D. Seidman et al.

association with CCC. On occasion, the full morphologic spectrum from endometriosis with hyperplasia to atypical hyperplasia and welldifferentiated endometrioid adenocarcinoma is observed. The absolute risk of neoplastic transformation appears to be low. Further support for the premalignant potential of endometriosis comes from molecular biological studies which have demonstrated several types of molecular alterations also observed in endometriosisassociated ovarian cancers. These include somatic mutations in several cancer driver genes [including ARID1A, KRAS, PPP2R1A, and PIK3CA, all of which are frequently mutated in ovarian clear cell and endometrioid carcinomas (Lin et al. 2017)], LOH (loss of heterozygosity) at the PTEN locus, microsatellite instability (MSI), and chromosomal aberrations including trisomies and monosomies. A significant proportion of endometriotic lesions, therefore, appears to have the potential to undergo additional genetic changes and malignant transformation. Nonetheless, it has been estimated that carcinoma develops in only 0.3–3% of cases of endometriosis, and 3% is very likely an overestimate since many cases of endometriosis never come to biopsy. The true figure is probably closer to 1% or lower, although population-based data are unavailable. Since extraovarian endometriosis is not uncommon, the rarity of carcinomas arising in extraovarian endometriosis suggests that ovarian endometriosis is significantly more likely to undergo malignant transformation than extraovarian endometriosis. As metastatic tumors to the ovary also have a propensity for substantial growth at this site, factors in the ovarian microenvironment conducive to malignant transformation and tumor growth may play an important role and have been under active investigation. A recent immunohistochemical study described the presence of activated stromal cells immediately adjacent to tumors in the ovary. These cells were shown to express enzymes involved in steroid synthesis, suggesting that the tumors in the ovary were stimulated to grow by these activated stromal cells (Blanco et al. 2017). As noted earlier, the relative proportion of endometrioid and CCCs that are associated

14

Epithelial Tumors of the Ovary

853

with endometriosis is much higher than that for the other cell types. The strong protective effect of tubal ligation against endometriosisassociated cancers supports endometriosis as a precursor (see section “Reproductive Factors, Ovulation, and Hormonal Factors,” earlier in this chapter). Further evidence that endometriosis is a precursor of ovarian carcinoma is provided by large studies from Sweden and Japan. In a Swedish study of over 20,000 women hospitalized with endometriosis, after a mean follow-up of 11.4 years, the risk of ovarian cancer in comparison to the control population was 1.9. Patients with a long-standing history of endometriosis (10 years or longer) had a RR of 4.2 (Brinton et al. 1997). In a study in Japan of 6398 women with endometriosis, the standardized incidence ratio was 9.0 after 17 years of follow-up, with a RR of 13.2 in women diagnosed after age 50 (Kobayashi et al. 2007). In the latter study, the mean age at diagnosis of ovarian carcinoma was 51, a reflection of the significantly younger age of women with endometriosis-associated cancers (see Table 1 and section “Endometrioid Carcinoma Arising in Endometriosis,” later in this chapter).

Benign and Atypical Proliferative (Borderline) Neoplasms Although the natural history of benign ovarian tumors cannot be observed since they must be Fig. 3 Schematic of the pathogenesis of LGSC (Type I tumor). (Modified from Singer et al. 2005)

completely removed for accurate diagnosis, the observation of morphologically benign areas within carcinomas and recent molecular evidence provide strong support that borderline tumors are precursors of low-grade serous, endometrioid, and mucinous carcinomas (Type I tumors in the dualistic model). KRAS and BRAF mutations are largely confined to LGSCs and APSTs and suggest that APSTs are likely precursors of LGSCs, but not HGSC (Fig. 3). KRAS and BRAF mutations are lacking in isolated serous cystadenomas, putative precursors of APSTs. However, identical KRAS or BRAF mutations were detected in APSTs and adjacent cystadenoma epithelium in serous cystadenomas associated with small APSTs. These findings suggest mutations of KRAS and BRAF are early events associated with serous tumor initiation and that a small subset of serous cystadenomas which acquire KRAS or BRAF mutations may progress to APST. TP53 mutations are very uncommon in LGSCs and APSTs.

Papillary Tubal Hyperplasia (PTH) PTH is characterized by small, rounded clusters of tubal epithelial cells and small papillae, with or without associated psammoma bodies that are present within the tubal lumen, within the tubal mucosa and the lamina propria of the tube. Histologically, PTH

Development of low grade (micropapillary) serous carcinoma

APST

MPSC

SBT Cystadenoma KRAS/BRAF/ERBB2

mutations

Low-grade carcinoma Loss of 1p36 loss of CDKN2A/B

854

bears a close resemblance to an APST. Although frequently associated with APSTs, it can occur in the absence of an APST. Sometimes it is found in cases in which there are noninvasive implants, identical to those associated with an APST or with endosalpingiosis but in which an APST is not present. It has been speculated that the small papillae and clusters of cells from the fallopian tube implant on ovarian and peritoneal surfaces to produce a constellation of low-grade ovarian and extraovarian low-grade serous proliferations (APST, noninvasive epithelial implants, and endosalpingiosis). PTH is frequently associated with active chronic salpingitis. In other cases in which active salpingitis is not present, evidence of prior salpingitis, specifically destruction and blunting of tubal plicae, is observed. It has been proposed that PTH is induced by chronic inflammation and can be a precursor of a wide variety of low-grade ovarian and extraovarian serous proliferations (Kurman et al. 2011; Seidman et al. 2002).

STIC and p53 Signature There is now persuasive evidence indicating that the origin of most HGSCs is the tubal fimbriae. (STIC and related tubal precursor lesions are further discussed in ▶ Chap. 11, “Diseases of the Fallopian Tube and Paratubal Region.”) The fimbriae normally extend over, and are in close contact with, the ovarian surface. Epithelial atypia, “carcinoma in situ” (STIC), and small invasive high-grade serous tubal carcinomas have been found in prophylactic specimens from women with BRCA mutations (see below). Meticulous examination of risk-reducing salpingo-oophorectomy (RRSO) specimens has disclosed occult intraepithelial or small invasive carcinomas in about 3%, but this figure varies widely, 0–12% among 16 series including 1750 patients. This wide range could reflect differing thresholds for the diagnosis of tubal carcinoma, particularly STIC. In addition, there are several potential sources of bias in these types of studies. The most reliable estimate is based on the findings of GOG199, a prospective study of 966 high-risk women, which found STIC or invasive carcinoma in 2.6% (Sherman et al. 2014). Further, it now

J. D. Seidman et al.

appears likely that apparent primary peritoneal carcinomas after RRSO reflect undetected microscopic, occult tubal carcinomas that had spread beyond the tube prior to the surgical procedure Leonhardt et al. 2011. In light of the above findings, the search for precursors of Type II ovarian carcinomas has shifted to the fallopian tubes where STICs have been identified in association with a high proportion of HGSCs. A recent systematic review of 10 series including over 600 pelvic HGSCs found 13–53% had concurrent STIC (mean 37%, 95% CI 27–48%) (Chen et al. 2017). The wide range in frequencies is probably due to a variety of types of bias in these studies including lack of blinding, variability in inclusion of postchemotherapy specimens, and the method of diagnosis (i.e., inconsistent use of immunohistochemistry and deeper levels). In addition, even when the tubes are embedded in their entirety, the examination of 1 or a few 5 μm-thick sections in blocks containing tissue 1–3 mm thick actually leaves over 99% of the tissue unexamined (Sherman et al. 2012; Visvanathan et al. submitted). It is important to note that the morphology of intraepithelial and invasive carcinoma in the fallopian tube may be mimicked by metastatic carcinoma (Rabban et al. 2015; Stewart et al. 2012). Further, limited data from molecular analyses suggest that intraepithelial carcinoma may on occasion represent bonafide metastatic disease rather than a primary tubal lesion (McDaniel et al. 2015; Eckert et al. 2016; Singh and Cho 2017). Also of note, STIC may persist after neoadjuvant chemotherapy (Colon and Carlson 2014). STIC is occasionally discovered incidentally in tubal specimens not removed for ovarian cancer. Several reports suggest an association with atypical endometrial hyperplasia and endometrioidtype endometrial adenocarcinoma, but as these tumors have low-grade cytologic features, a relationship with STIC appears unlikely (Chay et al. 2016; Gilks et al. 2015; Morrison et al. 2015; Rabban et al. 2014; Seidman et al. 2016). There is a well-documented association with endometrial serous carcinoma (see sections “High-Grade Serous Carcinoma (HGSC) of Tubal, Ovarian or

14

Epithelial Tumors of the Ovary

Peritoneal Origin” and “Differential Diagnosis,” later in this chapter). STICs contain TP53 mutations and are cytologically malignant but are confined to the tubal epithelium. Another lesion characterized by a short stretch (minimum of 12 tubal secretory cells) of morphologically normal tubal epithelium that is immunohistochemically positive for p53 and has a Ki-67 proliferation index higher than normal tubal epithelium but lower than that of a STIC has been designated as a “p53 signature.” p53 signatures are associated with STICs and ovarian HGSCs but are also found in the general population. TP53 mutations have been found in some p53 signatures. These findings have led to the proposal by some investigators that the p53 signature is the precursor of a subset of ovarian HGSCs (Cass et al. 2014; Crum et al. 2013). In a recent study using next-generation sequencing platforms to examine whole exome sequences and chromosomal alterations of fallopian tube lesions (p53 signatures, STICs, and invasive carcinomas), ovarian tumors, and metastases in matched tumor and normal specimens from the same individual, it was found that all had the identical TP53 mutation that were present in the p53 signature and the STIC. Moreover, compared to the STIC, the ovarian tumors and metastases had acquired additional somatic mutations indicating an evolutionary relationship between the STIC and the ovarian tumors and metastases to other sites in which the STIC harbored the ancestral clone. In some cases, a unique sequence change was present in the STIC and the metastases (omentum or appendix) and not the ovarian tumor, suggesting that the ancestral clone in the STIC had the capacity to spread to other organs and bypassed the ovaries (Labidi-Galy et al. 2017). This helps to explain why nearly all HGSCs present in advanced stage. Data suggesting that p53 signatures are precursors of STICs come from a multi-institutional study of 479 women at high risk for ovarian carcinoma who underwent a RRSO with comprehensive sampling (sectioning and entirely embedding the fimbriated end (SEE-FIM)) of the fallopian tubes. In this study, p53 signatures were detected in 22% of cases compared to STICs (4%) and serous tubal

855

intraepithelial lesions (STILs) (5%). The latter are defined as intraepithelial tubal lesions in which there is nuclear atypia but to a degree that is insufficient for a diagnosis of a STIC. Moreover, p53 signatures were present in all portions of the tube, although most were detected in the fimbria. Those in the fimbria occurred in significantly older women (54 years) compared to p53 signatures in other parts of the tube (45 years). These findings suggest that it is the fimbrial p53 signatures that are the precursor lesions as opposed to those occurring in other parts of the tube (Visvanathan et al. submitted). It has been shown that even when the fallopian tubes in women with ovarian cancer have been comprehensively evaluated, a STIC is found in at most 60% of cases. This raises the question of the source of these carcinomas, although as noted earlier, one possible explanation is that standard histological sections leave over 99% of the tubal tissue, where STIC may be present, unexamined. It is plausible that a STIC cannot be identified in some cases because a large carcinoma involving the ovary has overgrown the fallopian tube, thereby obscuring the site of origin. This is supported by a recent study comparing HGSCs in which a STIC was found to those in which a STIC was not detected. In that study there were no differences in copy number alterations, messenger RNA sequencing, and microRNA profiling between the two groups (Ducie et al. 2017).

The Tubal-Peritoneal Junction (TPJ) It is well known that junctional regions between different types of epithelia are hotspots for carcinogenesis. The prime examples are the cervical transformation zone, anorectal and gastroesophageal junctions, but there are others. In the adnexal region, mesothelium, OSE, and tubal epithelium are in close proximity. The microanatomy and corresponding histology of this region is not completely understood and has only recently been more carefully studied. It has been suggested that the adnexal region encompassing the OSE, peritoneum, and tubal epithelium be considered a unit of common embryological derivation with certain areas at high risk of neoplasia (Auersperg et al. 2008).

856

J. D. Seidman et al.

Fig. 5 Histology of the TPJ displaying mesothelium toward lower right and tubal epithelium toward top left. The junction is seen at the arrow

Fig. 4 Schematic diagram of the fimbrial region showing the TPJ which appears as a tortuous blue line which reflects the transition from tubal epithelium to peritoneal mesothelium

Recently the junction between the pelvic peritoneum and the tubal fimbrial epithelium where the peritoneal cavity communicates with the lumen of the fallopian tube has been characterized and designated as the TPJ (Seidman et al. 2011a) (Figs. 4 and 5). Foci of transitional metaplasia are frequently present at this site; in the past these have been termed “Walthard nests” or “-rests.” As these foci are generally observed in the absence of inflammation or other associated lesions, they appear to be a normal finding. It has been hypothesized that transitional metaplasia at the TPJ plays a role in the origin of both Brenner and mucinous tumors (see sections “Brenner (Transitional Cell) Tumors” and “Mucinous Tumors,” later in this chapter). A recent series showed that STIC was present at a mean of 1.8 mm from the TPJ among 81 women with advanced-stage, high-grade extrauterine serous carcinoma, and in occasional cases, STIC was found precisely at this junction (Seidman 2015).

These findings were confirmed by Schmoeckel and associates who found STIC at a mean of 1.3 mm from the TPJ (Schmoeckel et al. 2017). The precise role of this junction, if any, in the pathogenesis of these tumors, deserves further investigation. Interestingly, studies in mice suggest that the junctional region contains a cancerprone stem cell niche (Flesken-Nikitin 2013). In contrast to the clearly defined tubal-peritoneal junction, the putative “ovarian-tubal junction” remains to be defined.

Familial (Hereditary) Ovarian Cancer Two major types of genetic predisposition to ovarian cancer have been identified: hereditary breast/ ovarian cancer (HBOC) and hereditary non-polyposis colorectal cancer (HNPCC, Lynch syndrome). Notably, the increased frequency of ovarian cancer in relatives of women with ovarian cancer is not paralleled by an increase among relatives of women with borderline or mucinous ovarian tumors.

BRCA1, BRCA2, and Other Genes In North America, about 10–15% of ovarian carcinomas arise in the setting of highly penetrant, autosomal dominant genetic predisposition, but

14

Epithelial Tumors of the Ovary

recent studies suggest that it is closer to 24% (Walsh et al. 2011). It is well known that BRCA mutations are significantly higher in the Ashkenazi Jewish population, and a recent study from China found that 28.5% of ovarian cancer patients had a BRCA mutation (Wu et al. 2017). HBOC is associated with germline mutations of either BRCA1 or BRCA2. The risk of developing ovarian carcinoma by age 80 years for BRCA1 and BRCA2 carriers is estimated to be 44% and 17%, respectively (Kuchenbaecker et al. 2017). The BRCA1 and BRCA2 genes map to chromosome 17q and 13q, respectively, and encode proteins that play important roles in DNA repair, cell cycle checkpoint control, protein ubiquitinization, and chromatin remodeling. Germline mutations of these genes include small deletions, insertions, point mutations, and gene rearrangements that typically lead to prematurely truncated protein products. Mutations are widely distributed throughout both genes. Studies suggest the site of the mutation within these genes may correlate with risk for ovarian cancer. For example, mutations between nucleotides 2401 and 4190 in BRCA1 increase the risk of ovarian cancer but reduce the risk of breast cancer. Mutations involving nucleotides 4075–6503 of BRCA2 exon 11 are also associated with increased risk of ovarian cancer, and hence, this region has been referred to as the “ovarian cancer cluster region.” In up to half of HGSCs, BRCA1 and/or BRCA2 are mutated or are functionally inactivated through allelic deletions and/or silencing due to promoter hypermethylation. Germline or somatic mutations in homologous recombination repair (HRR) genes (BRCA1/2 and others) positively impact overall survival and platinum-responsiveness (Pennington et al. 2014). However, the effect on progression-free survival appears more pronounced than that on overall survival (Rudaitis et al. 2014). Further, the survival benefit appears to attenuate with longer follow-up such that there may be no significant survival difference between germline BRCA mutation-positive and mutation-negative patients at 10 years (Kotsopoulos et al. 2016). Increased risk for developing ovarian cancer has also been observed in patients affected by Lynch syndrome, and this accounts for about 2%

857

of ovarian cancers. This autosomal dominant syndrome is associated with elevated risk of colon carcinoma without polyposis, endometrial carcinoma, and less frequently, carcinomas of other organs, including the ovary. HNPCC is caused by mutations in genes that encode proteins involved in DNA mismatch repair (MMR). Over 70% of individuals with Lynch syndrome harbor germline mutations of hMSH2 or hMLH1, with mutations of hPMS1, hPMS2, and hMSH6 being less common. Lynch syndrome-associated ovarian carcinomas (LSAOC) arise at an earlier age than sporadic cases and tend to be relatively low stage (FIGO stage I or II) at diagnosis. Recent data indicate that clear cell and endometrioid ovarian cancers are overrepresented in ovarian cancer patients with Lynch syndrome (Chui et al. 2013; Vierkoetter et al. 2014). This cell type predilection explains the low stage and young age distribution for LSAOCs (Ketabi et al. 2011). More patients with ovarian carcinoma carry cancer-predisposing mutations and in more genes than previously appreciated. In a study of 360 women with primary ovarian, peritoneal, or fallopian tube carcinoma not selected for age or family history, using targeted capture and massively parallel genomic sequencing for germline mutations in 21 tumor suppressor genes, it was found that 24% carried germline loss-of-function mutations: 18% in BRCA1 or BRCA2 and 6% in BARD1, BRIP1, CHEK2, MRE11A, MSH6, NBN, PALB2, RAD50, RAD51C, or TP53. Six of these genes were not previously implicated in inherited ovarian carcinoma. Of women with inherited mutations, over 30% had no family history of breast or ovarian carcinoma, and over 35% were 60 years or older. Based on these findings, the investigators proposed comprehensive genetic testing for inherited carcinoma for all women with ovarian, peritoneal, or fallopian tube carcinoma, regardless of age or family history (Walsh et al. 2011). Recent literature has highlighted a variety of other potential genetic markers for increased risk of ovarian cancer. As the cost of whole genome sequencing has greatly declined, such data are accumulating rapidly, and some laboratories have begun to market such tests in the USA. Data on

858

many of these variants are conflicting, however, and the magnitude of the putative increased risk is, in many cases, marginal and accordingly of minimal or no clinical significance. The standardized incidence ratios (SIRs) for some of these markers are in the 1.1–1.5 range. In comparison, the SIRs for BRCA1 and BRCA2 are 49.6 [95% CI, 40.0–61.5] and 13.7 [95% CI, 9.1–20.7], respectively (Kuchenbaecker et al. 2017). In one recent example, the KRAS variant rs61764370 was initially reported to elevate ovarian cancer risk; subsequently the Ovarian Cancer Association Consortium was unable to validate this finding (Ovarian Cancer Association Consortium, Breast Cancer Association Consortium, and Consortium of Modifiers of BRCA1 and BRCA2 2016).

Clinicopathologic Features of Familial Ovarian Cancers Ovarian cancer in women with familial ovarian cancer occurs at a younger age than sporadic ovarian cancer. For BRCA1- and BRCA2-associated ovarian cancer, the age at diagnosis is about 54 and 59 years, respectively, and for Lynch syndrome, 43 years (Table 1). Nearly all BRCA mutation-associated tumors are high-stage HGSCs. Chui and associates studied 20 women with germline mutation-confirmed LSAOCs. They found 90% of cases to contain an endometrioid carcinoma component (14 pure, 4 mixed) and two (10%) pure CCCs. Tumor-infiltrating lymphocytes were prominent in only two cases (Chui et al. 2014). Another series of 53 LSAOC, though limited by lack of pathologic review, also found a predominance of endometrioid carcinomas (Ryan et al. 2017). Accordingly, Lynch syndrome screening is now recommended for all women with a diagnosis of clear cell or endometrioid ovarian carcinoma (Rambau et al. 2016; Singh and Gilks 2017). Metastatic foci of HGSC may display a variety of patterns. A recent study showed that tumors with germline or somatic BRCA1 or BRCA2 abnormalities were more likely to display a pushing pattern or infiltrative micropapillary

J. D. Seidman et al.

metastases, while those without such abnormalities showed infiltrative metastases with papillary, glandular, and rarely cribriform and micropapillary features (Reyes et al. 2014). These metastatic features appeared more reliable than features of the primary tumor for predicting BRCA mutations. Limited data suggest that a predominantly pushing metastatic pattern is associated with a better prognosis as compared to infiltrating micropapillary metastases (Hussein et al. 2016), but whether this is independent of BRCA status, a favorable prognostic feature in the short term, is not clear. The 5-year survival for germline BRCA-mutation-associated cases appears significantly better than that for sporadic cases (Bolton et al. 2012). A study with longer follow-up found that this difference seems to disappear by 10 years (Kotsopoulos et al. 2016). Whether this will change with treatment with poly (ADP-ribose) polymerase (PARP) inhibitors remains to be studied.

GEMMS of Ovarian Cancer and Translational Applications Mouse Models GEMMs of each major subtype of ovarian cancer will undoubtedly prove useful for improving knowledge of ovarian cancer biology and for testing new strategies for prevention, early detection, and treatment of ovarian cancer. Historically, most animal models of ovarian cancer were based on xenografting human ovarian cancer cells into immunodeficient mice. Though much can be learned from xenografts, especially those that are derived from tumor tissue directly transferred from the patient into the mouse (i.e., patientderived xenografts – PDXs), they also have limitations, including incomplete recapitulation of tumor-host interactions and inability to replicate early stages of tumor development. Recently described GEMMs of ovarian cancer appear to overcome some weaknesses of the xenograft models as tumors arise orthotopically in immunologically intact animals, and many closely mimic

14

Epithelial Tumors of the Ovary

the morphology and biological behavior of their human ovarian cancer counterparts. Given the morphologic and molecular heterogeneity of ovarian carcinomas described above, a goal for researchers in the field is the development of mouse models that recapitulate each major histologic type of ovarian cancer. The majority of ovarian cancer GEMMs reported to date have been summarized in a recent report (National Academies of Sciences, Engineering and Medicine 2016). GEMMs of high-grade and low-grade serous, endometrioid, and CCCs have been developed, most based on transformation of the mouse OSE following tissue-specific induction of genetic alterations characteristic of each human tumor type. Now that many, if not most, ovarian cancers are believed to arise from the fallopian tube fimbriae, more recent models have focused on transformation of the mouse oviductal epithelium (equivalent to human FTE). Investigators have shown that genetic alterations characteristic of human HGSCs (e.g., deletion of various combinations of Trp53, Brca1/2, Rb1, Pten, Nf1) also lead to STICs and HGSClike tumors in the mouse oviduct (Kim et al. 2012; Perets et al. 2013; Zhai et al. 2017).

Translational Applications Mouse models that faithfully recapitulate their human tumor counterparts may also prove useful for preclinical testing of novel strategies to prevent, detect, and treat ovarian cancer. For example, GEMMs can be used to explore the mechanisms by which many of the factors associated with altered ovarian cancer risk described earlier in this chapter exert their effects. Understanding why high parity and oral contraceptive use are protective, and why the cancers typically manifest in menopausal rather than younger women, could help researchers develop new strategies with which to prevent ovarian cancer, or at least slow its progression. Mice with early-stage disease can be used to test novel approaches to improve early diagnosis of ovarian cancer (e.g., detection of tumor-specific DNA in the circulation or in Pap smear-type samples as described below). Finally, the models can be used for preclinical

859

testing of novel therapeutics for ovarian cancer. Given the overwhelming number of potential new drugs that can be tested, alone and in combination with conventional therapies, GEMMs may prove useful for identifying those drugs and drug combinations with greatest likelihood of success in humans.

Screening, Early Diagnosis, and Prevention Screening Tests The relatively low prevalence of ovarian cancer in the general population makes the development of an effective screening test extremely challenging. An understanding of the new model of ovarian cancer pathogenesis has important implications for screening. As the majority of HGSCs and its variants, which cause most ovarian cancer deaths (85–90%), arise in the fallopian tube, all screening trials, which have been based at least in part on pelvic ultrasound, are predicated on an incorrect model of carcinogenesis (Bodelon et al. 2014). These trials target enlarged ovaries which, when they contain carcinoma, generally reflect Type I tumors which typically present in low stages and have a favorable prognosis (see section “Morphologic and Molecular Pathogenesis,” earlier in this chapter). In contrast to Type II tumors, only about 20% of ovarian carcinomas are Type I tumors. These appear to grow slowly from benign cystadenomas, atypical proliferative tumors, and endometriomas, which become quite large before undergoing malignant transformation. As a result, most patients with these tumors are diagnosed with large pelvic masses on pelvic examination without the need of ultrasound before the neoplasms have undergone peritoneal dissemination. These tumors are associated with a generally favorable prognosis since they are diagnosed in early stage, and accordingly, there is little justification to target them for screening. In contrast, a test for the early detection of HGSC, which is responsible for the vast majority of deaths from ovarian cancer and is typically not

860

diagnosed before it is in advanced stage, is more likely to have the potential to reduce mortality from this disease. The putative precursor of invasive HGSC is STIC, which is microscopic and therefore too small to be detected by ultrasound. In addition, serum CA125 lacks sufficient sensitivity and specificity to effectively detect ovarian cancer at a curable stage. Screening trials have generally identified low-stage non-serous tumors, borderline tumors, nonepithelial tumors, and advanced-stage HGSC, with the identification of stage I HGSC an exceedingly rare event, even in series focusing on high-risk women (Skates et al. 2017). Most trials have not reported mortality data sufficient to determine whether the screening was successful. Two large clinical trials of ovarian cancer screening, with mortality data, have now been reported. The Prostate, Lung, Colon, and Ovarian (PLCO) Trial randomized 78,216 women to annual screening with CA125 and ultrasound versus usual care (Buys et al. 2011). After a median of over 12 years (recently updated to 14.7 years) (Pinsky et al. 2016), no significant mortality reduction was identified (RR 1.06). In fact, there were more deaths in the screening arm (not statistically significant). A recent analysis of the ovarian cancer cell types diagnosed in PLCO demonstrated a possible stage-shift to lower stages for type I tumors but not for type II tumors (Temkin et al. 2017), further supporting the lack of efficacy of screening for reducing ovarian cancer mortality. The UK Collaborative Trial of Ovarian Cancer Screening (UKCTOCS) is one of the largest clinical trials ever conducted (Jacobs et al. 2016). Over 200,000 women were followed for a median of over 11 years. Over 100,000 women were randomized to no screening, 50,623 to ultrasound alone, and 50,624 to multimodal screening which incorporated analysis of CA125 levels over time using a proprietary algorithm (risk of ovarian cancer algorithm – ROCA), with referral to ultrasound under defined conditions. No significant reduction in mortality due to ovarian cancer was demonstrated. In addition to failing to demonstrate a benefit, all ovarian cancer screening trials have demonstrated significant harm to enrolled

J. D. Seidman et al.

patients. Indeed, many women have undergone surgery, and some have developed serious complications, while only a small minority have had ovarian cancer. In fact, in 2016 the US Food and Drug Administration (FDA) issued a safety communication indicating that there are no safe and effective ovarian cancer screening tests and advising women and clinicians not to use ROCA, which was being marketed based on the UKCTOCS results, because of potential harms and insufficient evidence of any benefit (Food and Drug Administration 2016). In 2018, the US Preventive Services Task Force reaffirmed their previous recommendations against ovarian cancer screening (Henderson et al. 2018) The fact that many ovarian cancer patients in retrospect have had symptoms, sometimes chronic, prior to diagnosis, has led to the development of a so-called symptom index for early diagnosis. Unfortunately, a symptom index is very unlikely to be effective for several reasons. First, the early symptoms of ovarian cancer are too nonspecific and very common. Second, any symptoms referable to ovarian cancer generally indicate that, at least for HGSC, disease has progressed to an incurable stage. Based on the current understanding of pathogenesis, the curable phase of ovarian cancer is STIC which is asymptomatic. The problems with such an approach are highlighted by a recent systematic review of five studies using the “Ovarian Cancer Symptom Index,” which showed an estimated positive predictive value of less than 1% (Ebell et al. 2016). In the past few years, a variety of innovative approaches to identifying early lesions in fallopian tubes have been investigated using molecular techniques. The sources of these specimens include brush specimens from the cervix (Bakkum-Gamez and Dowdy 2014), tissue isolated from tampons (Erickson et al. 2014), ex vivo washings of the fimbriae (Dobrinski et al. 2014), and uterine cavity lavage (Maritschnegg et al. 2015). Recent advances in technology have also enabled the detection of circulating cell-free DNA (cfDNA) in circulating blood. Most cfDNA in the circulation is derived from ruptured nonmalignant cells that are of germline origin, but in addition, a small amount

14

Epithelial Tumors of the Ovary

of circulating tumor DNA (ctDNA), resulting from tumor cell apoptosis and necrosis, is present. Newly developed novel genomic and bioinformatic approaches have facilitated highly sensitive molecular assays that can detect tumor-specific alterations from ctDNA (Kinde et al. 2011).

Early Diagnosis Early diagnosis has been demonstrated to reduce mortality for cervical, breast, and colon carcinoma, but the same cannot be said for ovarian cancer. Although this has not been directly demonstrated, a retrospective case-control study found that among 1318 women with ovarian cancer who presented to a medical practitioner after the onset of symptoms, a delay in diagnosis of greater than 12 months (time from symptom onset to diagnosis) did not change survival as compared to those who were diagnosed within 1 month of symptoms (Nagle et al. 2011). The hazard ratio was 0.94, which was not statistically significant. A prospective study of 1442 woman with ovarian cancer provides informative data for the value of early diagnosis of ovarian cancer recurrence. In this cohort, 529 women who experienced elevated CA125 were subsequently randomized to immediate treatment or delayed treatment at the time of clinical relapse (Rustin et al. 2010). The median survival among women with delayed treatment was 27.1 months as compared to 25.7 months among those treated upfront (not statistically significant). Cervical, breast, and colon carcinomas progress in a stepwise fashion in which the early manifestations of the disease occur when the lesion has not yet become invasive or is only superficially invasive and confined to a local site that can be surgically resected before the cancer has spread. In contrast, at the time of the earliest symptoms of ovarian cancer or an elevated CA125 level, the tumor has already spread to distant sites and is in an advanced stage. The reason for this is that the earliest “ovarian” cancer is a STIC which, despite its small size, is an established cancer capable of exfoliating malignant cells that can spread to distant sites in the

861

absence of invasion. For early detection of ovarian cancer to be effective, the carcinoma must be detected when it is microscopic and asymptomatic, i.e., STIC. The difference in the pattern and time course of ovarian cancer progression results from the direct progression of STIC, which is noninvasive and microscopic, to stage III ovarian carcinoma, without a gradual progression of a locally invasive tumor through the intermediate steps that are well characterized for other cancers.

Prevention The Society of Gynecologic Oncology’s five recommendations for the prevention of ovarian cancer are oral contraceptive use, tubal sterilization, RRSO in women at high risk, genetic counseling for those at high risk, and salpingectomy after completion of childbearing (Walker et al. 2015). Among these, RRSO is the most frequently utilized for women at high risk. In a multiinstitutional study evaluating the frequency of unsuspected neoplasia of women at high risk undergoing RRSO, it was found that unsuspected neoplasia was present in 5% (Conner et al. 2014). When there was no evidence of disease in the tubes after surgery, the recurrence rate was 4–5% which is a four- to ninefold-greater frequency than in the general population. If the resected tubes contained a STIC only, there was an 11% recurrence rate which occurred 4 years after surgery, and if invasion was identified in the tubes, there was a 17% recurrence rate. A literature review of BRCA1 and BRCA2 mutation carriers who developed peritoneal carcinomatosis after RRSO found a median interval of 54 months between RRSO and carcinomatosis. In this series, older age at RRSO, BRCA1 (as compared to BRCA2) mutation, and presence of STIC were more strongly associated with recurrence as compared to an unselected single institution cohort of mutation carriers undergoing RRSO (Harmsen et al. 2018). Although RRSO significantly reduces the risk of ovarian and breast cancer, removal of the ovaries has a detrimental effect on overall survival as shown in a Nurses’ Health Study of 30,000 women followed for 28 years that compared

862

BSO to ovarian conservation in women undergoing hysterectomy for benign uterine disease (Parker et al. 2013). In that study, removal of the ovaries resulted in a 23% increase in coronary artery disease mortality, 29% increase in lung cancer mortality, and 49% increase in colorectal cancer mortality, with a 13% increase in mortality from all causes. Accordingly, some investigators are now advocating that for women who are not at high risk, instead of RRSO, salpingectomy with conservation of the ovaries be performed and that perhaps an oophorectomy be performed after menopause. For young women who wish to have a permanent form of contraception, salpingectomy as opposed to bilateral tubal ligation is being recommended as the fimbria are the leading source of ovarian carcinoma. Salpingectomy appears to be safe (Song et al. 2017), and it has been shown that bilateral salpingectomy alone reduced the risk of ovarian cancer by 61% at 10 years (Falconer et al. 2015). The risk may not be completely eliminated by salpingectomy due to portions of fimbrial tissue left on the ovarian surface after bilateral salpingectomy (Ayres et al. 2017; Gan et al. 2017) or to carcinoma developing from endosalpingiosis in the peritoneum. Finally, multiple epidemiologic studies have shown that the use of OCPs can reduce ovarian cancer risk by as much as 50%. Increased parity, which reduces the number of lifetime ovulatory cycles, also substantially reduces risk. These findings had been interpreted as consistent with the “incessant ovulation” hypothesis of ovarian carcinogenesis as described earlier. Accordingly, it has been argued that young women should choose OCPs as the preferred method of contraception as it significantly reduces the risk of ovarian cancer while providing optimal contraception. Indeed, in an epidemiologic meta-analysis of 100,000 women, it was found that the longer women used OCPs, the greater was the reduction of risk ( p < 0.0001) and that the reduction in risk persisted over 30 years after OCP use ceased (Collaborative Group on Epidemiological Studies of Ovarian Cancer 2008). There are limited data on modifiable diet and lifestyle factors. In the Nurse’s Health Study of

J. D. Seidman et al.

over 80,000 women, there was a modest inverse association between caffeine intake and ovarian cancer risk in women not using hormones, while alcohol and smoking had no effect. However, smoking increased the risk of mucinous carcinoma (Tworoger et al. 2008). Another large review found that smoking doubles the risk of mucinous carcinoma (Jordan et al. 2007); however, as noted earlier, data on mucinous tumors prior to the mid- to late 1990s are unreliable because it is now recognized that most mucinous carcinomas involving the ovary are metastatic (see sections “Epidemiology,” earlier in this chapter, and “Mucinous Tumors,” later in this chapter).

Prognostic Factors The only universally accepted prognostic factors for patients with ovarian cancer are FIGO stage and, in stage IIIC and IV patients, volume of residual disease after surgical staging (with or without debulking). Age is a strong prognostic factor in many studies but may not be independent, as discussed earlier (see sections “Etiology and Risk Factors” and “Age,” earlier in this chapter). Other factors that may be important but about which there is continued debate include cell type, histologic grade, and tumor rupture.

Cell Type and Histologic Grade The prognostic importance of cell type and grade is complex because these features are interdependent. It is important to recognize that three of the five major subtypes of epithelial ovarian cancer have a definitionally assigned grade, so that only mucinous and endometrioid carcinomas need to be graded once the cell type is assigned. The reproducibility of assigning cell type to ovarian carcinomas has improved over the years as the types have become better defined through more specific histologic criteria that have been correlated and validated with both immunohistochemistry and molecular studies (Kobel et al. 2014; Seidman et al. 2015).

14

Epithelial Tumors of the Ovary

Histologic grade is reported to be a significant prognostic factor in many studies and in literature reviews. However, most of these data are not interpretable because of the number of different grading systems, the failure to specify which system was used, the lack of uniform pathologic review, the failure to correct for cell type and stage, and the fact that grade is among the most poorly reproducible observations among pathologists. Further, most reported studies predate recent refinements in cell type assignment and grading. A recent analysis of ovarian carcinoma grade agreement between two gynecological pathologists and the grade reported in NCIs SEER database demonstrated fair agreement regardless of the grading system used. The authors concluded that recorded grade in SEER is probably not reliable enough for epidemiologic studies (Matsuno et al. 2013). In stage-stratified comparisons, available data suggest that, in comparison to HGSC, endometrioid carcinomas and LGSCs have a better prognosis, and carcinosarcomas have a poorer outcome. Stage I clear cell and mucinous carcinomas do as well as other comprehensively staged patients in stage I (Kobel et al. 2010b). Mucinous and clear cell carcinomas may do worse in advanced stage, but data are limited by the rarity of advanced-stage mucinous carcinomas and diagnostic reproducibility issues between clear cell and HGSCs (Mackay et al. 2010; Zaino et al. 2011). Further discussion by cell type is found in the respective sections later in this chapter. At present, grading of ovarian carcinoma is clinically important only for stage IA, IB, and IIA patients because chemotherapy can be withheld for low-grade tumors in view of their excellent prognosis when untreated. Stage IA and IB grade 1 carcinomas have survival rates as high as 97%. However, stage I ovarian cancer is a heterogeneous group with only a small minority being serous carcinomas. There are insufficient data available to determine whether the binary grading system for serous carcinoma produces a grade that has reliable treatment implications for non-serous tumors (further discussion of grade by cell type is found in the respective sections later in this chapter).

863

Other Prognostic Factors Surgical expertise is recognized as a prognostic factor in ovarian cancer. Patients treated in hospitals treating larger numbers of ovarian cancers, and those operated on by gynecological oncologists, have superior survival rates as compared to those operated on in low volume hospitals and by general surgeons and other nonspecialists (Bristow et al. 2014). Similarly, the use of neoadjuvant chemotherapy is associated with improved survival in high volume hospitals using this regimen as compared to hospitals that infrequently use neoadjuvant chemotherapy (Barber et al. 2017).

Stage, Patterns of Spread, and Survival FIGO updated the staging of gynecological cancers in 2012 (Table 3). Notable changes from the previous edition include the elimination of stage IIC and the reassignment of metastases to inguinal lymph nodes from stage IIIC to IVB. FIGO stage is a very powerful predictor of outcome in ovarian cancer that most other putative prognostic factors are of little importance in comparison. Volume of residual disease after staging and debulking is also a very useful prognostic factor in stages IIIC and IV, which constitute the vast majority of patients with HGSCs. Comprehensive staging as described later in this chapter (sections “Treatment” and “Surgery”), which is needed in most patients to avoid the historical problem of understaging, will generally be sufficient surgical treatment for patients in stages I and II. Patients with advanced-stage disease often require debulking, or cytoreductive surgery, as well. Although cure is uncommon, cytoreductive surgery improves survival. Other benefits include improved patient comfort, reduction in the adverse metabolic consequences due to tumor including enhanced ability to maintain nutrition, enhanced ability to tolerate chemotherapy, and enhanced responsiveness of residual tumor to chemotherapy. Aggressive primary cytoreductive surgery has a low morbidity and

864

mortality rate and is supported by an improved survival rate in multiple studies. The median survival is about 4–5 years or longer after optimal cytoreduction as compared to about 1–2 years after suboptimal cytoreduction. In optimally debulked patients, removal of nodal tissue from at least one pelvic or para-aortic node site improves progression-free and overall survival by about 2 and 4 months, respectively (Rungruang et al. 2017). Further, survival after recurrence also appears improved for those without suspected nodal disease who underwent lymphadenectomy (Paik et al. 2016). Many patients with bulky disease cannot be optimally cytoreduced at primary surgery and require neoadjuvant chemotherapy followed by interval debulking. Surgicopathologic stage can usually be deduced solely from the histopathologic and cytologic findings. It has been argued that dense adhesions of an apparent stage I tumor are associated with a prognosis similar to that of a more advanced-stage tumor. One study, however, found that pathologic stage I tumors upstaged to stage II based on dense adhesions did not have a worse prognosis (Seidman et al. 2010a), although this remains common practice in some centers, and FIGO staging guidelines are unclear on this point. This and other problems in assigning stage are summarized in Table 5. The stage distribution of ovarian cancer varies by histologic type (Table 4) (Kobel et al. 2010a). The highest proportion of FIGO stage I cases is found among Type I tumors. CCCs are stage I in about 50% of cases. The vast majority of mucinous carcinomas are stage I. Endometrioid carcinomas are stage I in 50% or more. In fact, most stage I mucinous, endometrioid, and CCCs are confined to one ovary (stage IA). Bilateral involvement (stage IB) is found in only 1–5% of stage I cases. In contrast, only 3% or fewer of HGSC are diagnosed in stage I. LGSC, a Type I tumor, is also rarely diagnosed in stage I, although noninvasive LGSC is somewhat more often diagnosed in stage I. In our series (Table 4), only 15% of ovarian cancers were stage I, and one-third of these were not comprehensively

J. D. Seidman et al. Table 5 Problems and pitfalls in staging ovarian carcinoma. (Modified from 6th ed. Table 8) Problem Does STIC without invasion qualify for tubal involvement? What constitutes ovarian surface involvement?

Solution Yes

Tumor cells directly exposed to the peritoneal cavity, generally manifested as exophytic surface papillae Does slight leakage warrant Yes substage C? Does rupture of a benign No guidelines or data are component of a available heterogeneous tumor warrant substage C? Does sigmoid colon II. The sigmoid colon is a involvement indicate stage pelvic structure II or III? Does invasion of skeletal No. This is stage III muscle of diaphragm, posterior abdominal wall (psoas muscles), or anterior abdominal wall (rectus muscles) warrant stage IV? Does invasion of the Invasion through to the diaphragm indicate stage pleural surface of the IV? diaphragm or into the parietal pleura warrants stage IV Does abdominal wall Invasion through the involvement indicate stage anterior rectus sheath, or III or IV? into subcutis or skin, warrants stage IV Does splenic parenchymal IV (clarified in 2012 FIGO involvement indicate stage update) III or IV? Does dense adherence to Unresolved. Practices vary extraovarian structures and FIGO guidelines are warrant upstaging in the not clear absence of histologic confirmation of extraovarian disease? Are chest CT and other No extra-abdominal evaluations for apparent stage III patients needed to rule out stage IV?

staged. Five-year survival rates by stage are shown in Table 6. Stage I ovarian cancer is confined to the ovaries and peritoneal fluid or washings. Tumor rupture or

14

Epithelial Tumors of the Ovary

865

Table 6 5-year survival by stage (%), in four continents FIGO stage I II III IV

FIGOa 86 70 34 18.6

USA (SEER)b 92.5 73.0 28.9h 28.9h

Koreac 91.4 75.6 45.7 20.4

Netherlandsd 81.2 60 24.5 11.7

Australiae 88 65 27 12

Swedenf 71 67 31 17

UKg 90.0 42.8 18.6 3.5

a

Heintz et al. (2006) SEER (2017) c Chung et al. (2007) d Vernooij et al. (2008) e National Health and Medical Research Council, Australia (2008) f Dahm-Kahler et al. (2017) g Cancer Research UK (2017) h Categorized as “distant” disease b

tumor cells in peritoneal washings or ascitic fluid warrant a stage of IC. Ovarian surface involvement by tumor is also considered to reflect stage IC disease. We consider ovarian surface involvement to be present only when tumor cells are exposed to the peritoneal cavity (Seidman et al. 2004). Thus, surface involvement is characterized by exophytic papillary tumor on the surface of the ovary or on the outer surface of a cystic neoplasm replacing the ovary (Figs. 6 and 7). Assessment of surface involvement requires careful gross examination and cooperation between the surgeon and the pathologist. Poor prognostic factors in stage I have historically been grade 3, clear cell type, and IC substage (including rupture). However, this is controversial. Several large recent studies have shown that stage IC tumors do not have a poorer outcome than IA, nor do clear cell histology or rupture portend a poor outcome. When comprehensive staging procedures have been performed, HGSC rarely occurs in stage I; when it does, the prognosis is much worse than other cell types in stage I (Bamias et al. 2011; Karamurzin et al. 2013; Kobel et al. 2010b; Morency et al. 2016; Seidman et al. 2010b). Stage I CCC is more frequently substage IC as compared to the other cell types because of an increased risk of intraoperative rupture (Anglesio et al. 2011). Based on recent studies (Higashi et al. 2011; Kumar et al. 2014), it has been suggested that when CCC or endometrioid carcinoma is confined to the ovary at the time of surgery, the likelihood of occult extraovarian disease is extremely low, and therefore, comprehensive surgical staging can be omitted (Tolcher et al. 2015).

Fig. 6 LGSC. The outer surface displays exophytic papillary excrescences reflecting ovarian surface involvement

Fig. 7 LGSC. Sectioning reveals papillary excrescences on both inner and outer surfaces (same tumor as Fig. 6)

866

In contrast to the other cell types, HGSC that appears to be confined to the ovary are very often upstaged due to the presence of occult disease. The explanation for the discrepancy between CCC and endometrioid carcinomas compared to HGSC lies in the differences in pathogenesis. HGSC involving the ovary results from implantation of malignant cells that are derived from a STIC usually located in the tubal fimbriae. Malignant cells implant both on the ovary and at other sites in the pelvis and abdomen and continue to evolve as HGSC. This accounts for the typical presentation of HGSC in advanced stage and its highly aggressive behavior. In contrast, the pathogenesis of CCC and endometrioid carcinoma begins with implantation of benign endometrial tissue on the surface of the ovary resulting in the development of an endometriotic cyst. Progression to carcinoma results from transformation of the endometrial tissue lining the cyst to atypical endometriosis and then to either endometrioid carcinoma or CCC. The carcinoma is confined to the lining of the cyst and therefore can remain in a relatively dormant stage for a long period of time. Until there is invasion into the cyst wall, which allows for vascular or lymphatic transport of malignant cells or rupture of the cyst with spillage of malignant cells into the peritoneal cavity, these carcinomas remain confined and can be successfully treated simply by oophorectomy. Stage II ovarian carcinoma is a small and heterogeneous group and comprises 5–8% of ovarian cancers (Table 4) (Pennington et al. 2014). Stage II is defined as extension or metastasis to extraovarian pelvic organs. As such, it includes examples of direct extension to the tubes and pelvic sidewall, as well as metastatic seeding of the pelvic peritoneum, and therefore may include curable tumors that have directly extended to adjacent organs but have not yet metastasized, as well as tumors that have seeded the pelvic peritoneum by metastasis and therefore have a poor prognosis. Substage IIC was eliminated in the 2012 FIGO staging update. As noted earlier, some “pathologic stage I” tumors are considered “surgical stage II” by many surgeons.

J. D. Seidman et al.

Ovarian cancer, typically HGSC, most commonly presents in stage III, and the vast majority of these (84%) are stage IIIC. These tumors characteristically spread along peritoneal surfaces involving both pelvic and abdominal peritoneum. Less often, there is a solitary dominant adnexal tumor that directly invades bowel or other abdominal structures without the usual features of peritoneal studding. Ascites is found in two-thirds and its presence correlates with suboptimal cytoreduction and a higher number of positive lymph nodes. Metastases to retroperitoneal (para-aortic and precaval) or pelvic lymph nodes are found in the majority of patients who undergo node sampling or dissection and in up to 78% of advanced-stage patients. Inguinal lymph nodes are less commonly involved, and now qualify for stage IVB, although SEER data indicate that survival of those with stage IV based on inguinal nodes is similar to those with pelvic/para-aortic node metastases (stage IIIC) (Nasioudis et al. 2017a). Apparent early-stage patients have nodal metastases in about 14% of cases. This varies by cell type, as one-fourth to one-third of apparent early-stage HGSCs have positive nodes, clear cell carcinomas in 5–14%, and endometrioid carcinomas in 6.5%. In contrast mucinous carcinomas are nearly always node negative (Kleppe et al. 2011; Mahdi et al. 2013; Powless et al. 2011). Retroperitoneal nodal metastasis indicates stage IIIA1 in the absence of peritoneal metastasis. Nodal metastasis without peritoneal metastasis is relatively uncommon and seems to portend a better outcome as compared to stage IIIC patients with bulky abdominal disease, even those who have been optimally cytoreduced. Accordingly, in the updated 2012 FIGO staging, nodal disease without peritoneal disease is stage IIIA1 rather than IIIC as previously classified (Table 3). Bowel involvement is grossly evident in 72% of stage III patients, and 40% require bowel resection. Bowel involvement is manifested as serosal and subserosal involvement which is frequently extensive, involves multiple segments, and often forms large tumorous masses. The rectosigmoid is the most common site and is involved in 80% of those with bowel involvement. Focal invasion of the outer layer of muscularis propria is not uncommon. Invasion into the

14

Epithelial Tumors of the Ovary

submucosa or transmural invasion with mucosal ulceration occurs much less often. Volume of residual disease is an important prognostic factor in most studies, but this applies only to stages IIIC and IV, as stages IIIA and IIIB by definition have low volume disease (Table 3). Criteria for optimal versus suboptimal debulking rely on the size of the largest nodule remaining. Most gynecologic oncologists use 0, 0.5, or 1.0 cm as the threshold, with 1.0 cm the most widely used. Most studies show that the complete elimination of macroscopic disease (i.e., a zero threshold) is associated with the most favorable prognosis. Stage IV is defined as distant metastasis and includes patients with parenchymal liver metastases and extra-abdominal metastases. Twelve to 21% of patients present in stage IV. The median survival for stage IV is just under 2 years, and the overall 5-year survival is 19%, but can be as high as 39% in optimally debulked patients (Ataseven et al. 2016a). The liver and lungs/pleurae are the most common metastatic sites, and anterior abdominal wall metastasis including periumbilical skin and subcutis is frequent. A solitary distant metastasis is a favorable prognostic factor as compared to multiple metastases. Lung and pleural metastases have been reported in up to 45% of ovarian cancer patients, and respiratory failure is one of the most common clinical causes of death. During the course of the disease, one-third of patients develop pleural effusions, and three-quarters of these contain malignant cells on cytopathological examination. Among patients with pulmonary involvement, malignant pleural effusion is three times as common as parenchymal lung metastasis. The 5-year survival had been 6% after pulmonary metastasis; however this has improved to 17–36% as surgical resection of isolated metastases has been widely performed. Hepatic metastases have been found in half of ovarian cancer patients at autopsy. The median survival of stage IV patients with liver metastases is about 1 year. With surgical resection of metastases, the survival improves to a median of about 2 years. Anterior abdominal wall metastases frequently involve periumbilical subcutis and skin.

867

Extraperitoneal abdominal wall involvement warranting the stage IV designation occurs when tumor has invaded through the anterior rectus sheath (Table 5). Patients whose only site of distant metastasis is the abdominal wall (stage IVB) have a prognosis similar to that for stage IIIC patients. Abdominal wall involvement may manifest as a port site metastasis after laparoscopy (Ataseven et al. 2016b). Splenic parenchymal metastases are found in 20% of patients at autopsy. The FIGO 2012 staging update (Table 3) clarified that splenic parenchymal metastasis warrants stage IV; however it remains unclear whether splenic involvement is an adverse prognostic factor if stage IV is based solely on this criterion. Among ovarian cancer patients who undergo splenectomy as part of the debulking procedure, only about half have splenic metastases. Diaphragmatic metastases are usually confined to the peritoneal surface. On occasion, tumor will invade into or through the diaphragm, and the depth of invasion or extent of involvement may necessitate partial diaphragm resection. In a study of 36 full-thickness diaphragm resections for ovarian cancer, tumor was found invading pleura in 19% and muscle in nearly half. Of note, half of these patients with pleural involvement had no visible pleural disease intraoperatively (Majd et al. 2016). As survival rates have improved in the taxane era, previously rare sites of distant metastasis such as bone and brain are being diagnosed more often. Brain metastases are present in 0.1% of patients at presentation and are found in up to 6% of patients at autopsy. CNS recurrence is clinically manifested as carcinomatous meningitis in 4–6% of patients, and the median survival is less than 5 months. Among patients who develop brain parenchymal involvement after diagnosis, the median time after diagnosis is 20–30 months, and the subsequent survival is approximately 1 year (Cormio et al. 2011; Marchetti et al. 2016). Bone metastases occur in 1–2% of patients during the course of the disease and are found in up to 15% at autopsy. The median survival after bone metastasis is diagnosed is 4 months. Interestingly, among women with a history of ovarian

868

carcinoma who present with a breast or axillary mass, one-third have metastatic ovarian carcinoma and two-thirds have a primary breast carcinoma (Karam et al. 2009).

Cytopathology Ovarian epithelial neoplasms are usually evaluated in two types of cytopathological specimens: fine-needle aspirates (FNA) of ovarian cysts and peritoneal fluids (obtained by peritoneal washing or by aspiration of ascitic fluid). Intraoperative smears or imprints of ovarian tumors can also be useful adjuncts to, or replacements for, frozen section examination (Azami et al. 2018). Once disseminated, ovarian cancers may also be examined in FNA specimens or effusions from sites of distant metastasis. Rarely, the presence of psammoma bodies in a Pap smear will be the first sign of primary or recurrent ovarian serous carcinoma, more commonly when associated with atypical glandular cells or in older women with symptoms and signs suggesting malignancy.

Fine-Needle Aspiration of Ovarian Cysts FNA may be useful in patients who appear to have inoperable ovarian cancer or who cannot undergo surgery for other reasons. Unsatisfactory specimens from ovarian cyst aspirates are common and limit the usefulness of the procedure. In a recent study of 300 ovarian cyst FNAs with concurrent or subsequent definitive diagnosis on surgical pathology specimens, the sensitivity and specificity were 54% and 100%, respectively (Zhou et al. 2018). Accordingly, false positives are not an issue; however there is a high false negative rate. FNA specimens from most carcinomas are cellular and contain cytologically malignant cells, but accurate subclassification is often difficult and may be impossible based solely on cytologic material. Features useful in subclassification of epithelial neoplasms include psammoma bodies and papillary structures which suggest serous

J. D. Seidman et al.

differentiation, elongated cells, and focal squamous features which suggest endometrioid differentiation and cells with abundant clear, frothy, cytoplasm, and prominent enlarged nucleoli which suggest clear cell differentiation. Columnar cells with large cytoplasmic vacuoles containing basophilic cytoplasmic mucin suggest mucinous differentiation and, if malignant nuclear features are present, raises the question of metastatic carcinoma to the ovaries (see section “Invasive Mucinous Carcinomas,” later in this chapter, and ▶ Chap. 18, “Metastatic Tumors of the Ovary”). FNA specimens from serous carcinomas are usually very cellular and display malignant cells, singly and in clusters, with nuclear enlargement, hyperchromasia, irregular chromatin clumping, and prominent nucleoli. Bizarre tumor giant cells are common. Mucinous carcinomas yield mucus and high cellularity with single cells, clusters, and syncytial fragments displaying pleomorphism, coarse chromatin, prominent nucleoli, and vacuolated cytoplasm. Exclusion of metastatic mucinous carcinoma is generally not possible in cytologic material, but if sufficient cell block material is available for immunoperoxidase stains, the differential diagnosis can be narrowed. Cytologic material from endometrioid carcinomas displays scanty, more granular cytoplasm as compared to HGSCs, with nuclear crowding, and microacini. Squamous differentiation may be present. CCCs display cells with abundant pale vacuolated cytoplasm, nuclear pleomorphism, and macronucleoli. FNA of benign epithelial neoplasms usually produces a paucicellular specimen. Most of the material consists of macrophages and lymphocytes with few epithelial cells. The background is generally clean, unless torsion or necrosis has occurred. Benign serous tumors may display cohesive sheets of uniform cells with round to oval nuclei, moderate amounts of cytoplasm with well-defined cell borders, and occasionally cilia. The nuclei have finely granular chromatin and small nucleoli. Tumors with a fibromatous component also may display spindled stromal cells without atypia. Mucinous cystadenomas display tall columnar cells with basal nuclei without atypia and occasionally signet ring-like cells.

14

Epithelial Tumors of the Ovary

Atypical proliferative (borderline) epithelial tumors of all types may display a degree of cytologic atypia that overlaps with invasive welldifferentiated carcinoma, and therefore this distinction requires tissue examination in all cases.

Peritoneal Fluid Cytology Peritoneal fluid may be obtained by paracentesis. This is performed with recurrent ovarian cancer, often for symptomatic relief of recurrent ascites, but not often for primary diagnosis. Cytological samples of peritoneal fluid are routinely obtained during staging procedures for ovarian cancer. Washings are performed, or ascites aspirated upon entering the peritoneal cavity, prior to surgical manipulation that could dislodge tumor cells. Cytological findings are important in substaging FIGO stage I ovarian cancer; malignant cells in peritoneal washings or ascites warrant assignment to stage IC. As noted earlier, the 2012 FIGO staging update eliminated stage IIC. Cytology is more sensitive in detecting ovarian carcinoma in ascites than in peritoneal washes, as well as in patients with peritoneal metastases measuring greater than 0.5 cm as compared to those with smaller volume disease. Malignant cells are more often present in ascites as compared to washings, and their presence correlates positively with volume of ascites, serous histology, stage, and positive lymph nodes. About two-thirds of patients with HGSC and ascites have positive cytology at primary staging (Allen et al. 2017). The cytologic features of tumor cells generally resemble those in FNA specimens (see above) but may be more degenerated. Some cell types display distinctive features such as hyaline matrix deposits in CCCs (Damiani et al. 2016). A few studies have shown that patients with stage III disease with positive cytology have a poorer prognosis than stage III patients with negative cytology. The prognosis of patients with stage IC ovarian cancer based on positive peritoneal cytology is poorer than for stages IA and IB. Peritoneal lavage is often performed at the time of RRSO in high-risk women. Occult carcinomas

869

can be identified in these cytology specimens. Rarely, malignant cells have been reported in washings at RRSO without an identifiable carcinoma by histology, and positive cytology specimens may rarely lead to the discovery of earlystage tubal carcinomas (Agoff et al. 2002). The main component of benign peritoneal washings is mesothelial cells arranged singly and in sheets. With Papanicolaou stain, mesothelial cells appear as round or polygonal cells with dense, cyanophilic cytoplasm and centrally placed, round nuclei with smooth contours and finely granular chromatin. Degenerate and reactive mesothelial cells often display fine or course cytoplasmic vacuolation and a lightly stained perinuclear zone. The presence of epithelial cells in peritoneal fluid samples from patients with atypical proliferative (“borderline”) epithelial ovarian tumors has been highlighted in the past as a problematic area of cytopathology. The presence of invasion cannot be assessed in cytologic specimens of low-grade serous tumors. Cytologic findings do not add prognostic information for APST/ SBT. The small proportion of patients who have bona fide carcinomas (patients with invasive LGSCs) in stage I can be substaged based on the presence or absence of epithelial cells resembling the primary ovarian tumor in the cytology specimens. This situation is rarely encountered since most of these latter tumors are in advanced stage. The most important pitfall in the examination of peritoneal cytology specimens in women involves benign epithelial proliferations (see ▶ Chap. 13, “Diseases of the Peritoneum”). Women with or without cancer can have endometriosis and/or endosalpingiosis involving peritoneal surfaces. These lesions often shed epithelial fragments into peritoneal washings or ascites; in addition, benign FTE, particularly if salpingitis is present, and benign eutopic endometrial tissue via expulsion through the fallopian tubes may also be shed into the fluid. If the cells in the fluid are not obviously malignant, comparison of the cytologic features of the epithelium in the fluid with those of the tissue sections is essential in arriving at the correct

870

diagnosis. It is also important to be aware of the cytological abnormalities that may be caused by intraperitoneal chemotherapy as well as radiation, as they may mimic malignancy. In the future molecular genetic analysis of peritoneal fluid specimens may provide an even more valuable technique for tumor characterization.

Treatment Surgery Initial surgical management of ovarian cancer includes staging which is aimed at defining the extent of disease and debulking which is aimed at reducing tumor burden. In addition to removing the primary tumor intact, total abdominal hysterectomy, BSO, and omentectomy are performed. Random peritoneal biopsies including the diaphragm and para-aortic and pelvic lymph node sampling are performed to exclude occult metastases in apparent stage I disease, though the yield is low in the absence of grossly evident disease, particularly for apparent stage IA non-serous tumors in whom the need for comprehensive staging has been questioned (Mahdi et al. 2013; Mueller et al. 2016). Fertility-sparing surgery can be performed in selected young patients. Comprehensive staging will generally be sufficient surgical treatment for patients in stages I and II. Patients with advanced-stage disease often require debulking, or cytoreductive surgery, as well. Cytoreductive surgery, both primary (at presentation) and secondary (after recurrence), prolongs survival and progression-free interval. Second-look laparotomy has no demonstrated effect on survival and is no longer used. The median survival after first recurrence followed by secondary cytoreduction is 2–4 years in optimally cytoreduced patients and about 1 year in suboptimally cytoreduced patients. Patients with a longer disease-free interval (greater than 1–2 years) and those with limited sites of disease derive the most benefit from secondary cytoreduction (Berek et al. 2015).

J. D. Seidman et al.

Chemotherapy The survival for patients with stage IA and IB, low-grade tumors is better than 90%, and there is no demonstrable benefit from adjuvant chemotherapy for this small defined group. Stage IC is generally regarded as a negative prognostic indicator, but some recent studies have not confirmed this. It is possible that the inability to demonstrate a prognostic difference between stages IA and IC in some studies is due to a beneficial effect of chemotherapy on stage IC patients. A recent comprehensive review of randomized trials found that adjuvant platinumbased chemotherapy improves survival in early-stage (FIGO stage I/IIa) ovarian cancer, but that it remains uncertain whether women with low- and intermediate-risk early-stage disease will benefit as much from adjuvant chemotherapy as those with high-risk disease (Lawrie et al. 2015). A recent SEER analysis of stage I endometrioid and CCCs demonstrated a favorable effect of adjuvant chemotherapy only for stage IC grade 3 endometrioid carcinomas (Oseledchyk et al. 2017). Of note, SEER data are limited by the lack of centralized pathology review. For low-stage ovarian cancer with prognostic features that are generally considered adverse (high grade and stages IC and IIA), it therefore now appears that adjuvant chemotherapy is superior to no treatment, with the possible exception of CCC. Therefore, it is likely that patients with high-grade, low-stage endometrioid and mucinous carcinomas derive some benefit from this approach. Although there certainly exists a group of low-stage patients that are cured without any further treatment, identification of this group has been difficult as studies of low-stage patients have evaluated a wide variety of features, and many of the results are conflicting. Patients with advanced-stage (FIGO IIB, III, and IV) disease benefit from platinum and paclitaxel (platinum-based) chemotherapy. Platinum-based chemotherapy has resulted in a significant improvement in response rate, response duration, time to progression, and overall survival. The median time to first recurrence is 16 months. Long-term survival

14

Epithelial Tumors of the Ovary

(greater than 5 years) can be achieved with platinum-based combination chemotherapy in nearly half of patients with stage III small-volume residual disease. Recent data on platinum-based intraperitoneal chemotherapy for first-line treatment of optimally debulked stage III disease have shown about 20% improvement in both progression-free and overall survival, with the median survival increasing by about 12 months. The advantages of the intraperitoneal route include higher drug concentrations, prolonged tumor exposure, and reduced systemic toxicity. Upfront debulking, though the mainstay of primary treatment in the USA, may be deferred due to unresectable disease or inability of the patient to tolerate the procedure. These patients generally undergo neoadjuvant chemotherapy to reduce the extent of disease or improve performance status, followed by interval debulking surgery. Neoadjuvant chemotherapy followed by interval debulking has been widely used in Europe for many years. Data suggest that survival with this approach is noninferior to upfront debulking, with lower surgical morbidity and mortality; however questions remain due to selection bias in published trials as well as differences in surgical expertise and potential alterations in tumor biology (Fotopolou et al. 2017b). It is important to highlight here that the diagnosis of histologic type in this setting is based on a small biopsy or cytology specimen and that there are limitations to this approach. Hoang and associates analyzed 30 cases based on such limited material and found that core needle biopsy with or without immunohistochemistry had 85% accuracy, while cytology alone was accurate in only 52% (Hoang et al. 2015). In recurrent ovarian cancer, platinum with or without a taxane can be of value. Paclitaxel is active in many patients whose tumors are platinum resistant. A diagnosis of STIC without invasive carcinoma is quite uncommon and usually occurs in the setting of RRSO. Although data are quite limited, and there is no consensus on how to manage these women, there is no evidence that adjuvant chemotherapy provides any benefit (Chay et al. 2016; Conner et al. 2014; Powell et al. 2013; Powell 2014).

871

Targeted Therapy Olaparib, a PARP (poly (ADP-ribose) polymerase) inhibitor, received accelerated FDA approval in 2014 for fourth-line treatment of recurrent ovarian cancer in women with germline BRCA mutations. With this targeted therapy, PARP is trapped at sites of DNA damage and results in the formation of a PARP-DNA complex causing cell death in tumors that are deficient in HRR pathways (Ledermann et al. 2016). More recently, olaparib was approved for maintenance treatment of recurrent platinumsensitive disease. Other useful PARP inhibitors include rucaparib, which was approved in 2016 for third-line treatment of BRCA-mutated ovarian cancers, and niraparib (Mirza et al. 2016), which was approved in 2017 for maintenance treatment of recurrent platinum-sensitive disease. The vascular endothelial growth factor (VEGF) inhibitor, bevacizumab, has activity in ovarian cancer and is FDA-approved for recurrent disease. In vitro drug resistance assays are sometimes used to guide therapy, more often with recurrent tumor, but data are limited. Finally, immune checkpoint inhibitors such as agents targeting the PD-1/PD-L1 pathway of immune regulation may provide new opportunities for treatment of ovarian cancer. Such agents have shown substantial efficacy for carcinomas of the lung and several other sites, and in some settings, clinical response correlates with immunohistochemical expression of PD-L1 in tumor cells and tumor-infiltrating immune cells. A recent study found that PD-L1 expression in ovarian carcinomas is predominantly confined to tumorassociated macrophages (Webb et al. 2016). Several clinical trials of these agents in ovarian cancer are currently in progress.

Other Therapeutic Modalities Hormonal therapy may in some cases be a therapeutic option with relatively low toxicity. A recent meta-analysis found that hormonal therapy (tamoxifen, aromatase inhibitors, progestins, LHRH analogs, and others) was associated with a 41% clinical benefit rate. However, this review

872

included randomized trials judged to be of low quality and that did not distinguish cell types or first-line versus recurrent therapy and did not consider other important patient and tumor characteristics (Paleari et al. 2017). An earlier systematic review found that tamoxifen is associated with a 10% objective response and a 32% disease stabilization rate (Berek et al. 2015). Of all the cell types, endometrioid carcinoma has the greatest potential for response to hormonal therapy. Radiation may be an alternative therapeutic modality in patients who cannot tolerate surgery or chemotherapy. Long-term follow-up of patients who received consolidation radiation therapy and those treated with radiation and melphalan in the 1970s has shown long-term survival rates comparable to those currently treated with platinumbased chemotherapy; however radiation toxicity rates are high (Petit et al. 2007). Pelvic irradiation may be of benefit to patients with recurrences confined to the pelvis. Hyperthermic intraperitoneal chemotherapy (HIPEC) combined with aggressive cytoreduction is used in some centers, but there are no prospective randomized trials demonstrating equivalence, let alone superior efficacy, as compared to platinum-based chemotherapy (Harter et al. 2017). One recent randomized trial suggests that there may be a survival benefit to HIPEC during interval cytoreduction (van Driel et al. 2018).

Pathology of Ovarian Epithelial Neoplasms The 2014 World Health Organization (WHO) classification of surface epithelial tumors is summarized in Table 7, and the distribution of ovarian surface epithelial tumors is shown in Table 8. Ovarian epithelial tumors comprise about half of all ovarian tumors and account for about 40% of benign tumors and nearly 90% of malignant tumors. The apparent cell type distribution has changed significantly in the past two decades for several reasons. First, metastatic mucinous carcinomas have been recognized and properly categorized

J. D. Seidman et al. Table 7 2014 WHO classification of ovarian epithelial tumors Serous tumors Benign Serous cystadenoma Serous adenofibroma Serous surface papilloma Borderline SBT/APST SBT-micropapillary variant/noninvasive LGSC Malignant Low-grade serous carcinoma High-grade serous carcinoma Mucinous tumors Benign Mucinous cystadenoma Mucinous adenofibroma Borderline Mucinous borderline tumor (MBT)/APMT Malignant Mucinous carcinoma Endometrioid tumors Benign Endometriotic cyst Endometrioid cystadenoma Endometrioid adenofibroma Borderline Endometrioid borderline tumor/atypical proliferative endometrioid tumor (APET) Malignant Endometrioid carcinoma Clear cell tumors Benign Clear cell cystadenoma Clear cell adenofibroma Borderline Clear cell borderline tumor/atypical proliferative clear cell tumor (APCCT) Malignant Clear cell carcinoma Brenner tumors Benign Brenner tumor Borderline Borderline Brenner tumor/atypical proliferative Brenner tumor Malignant Malignant Brenner tumor (continued)

14

Epithelial Tumors of the Ovary

Table 7 (continued) Seromucinous tumors Benign Seromucinous cystadenoma Seromucinous adenofibroma Borderline Seromucinous borderline tumor/atypical proliferative seromucinous tumor (APSMT) Malignant Seromucinous carcinoma Undifferentiated carcinoma From Kurman et al. (2014)

(Bruls et al. 2015). Accordingly, primary invasive mucinous carcinoma is quite uncommon, comprising 2–3% of ovarian carcinomas and less than 1% of advanced-stage carcinomas in the USA (Table 4) (Zaino et al. 2011; Pennington et al. 2014), although they may be somewhat more common in Asia. Second, correlative molecular studies have led to reclassification of several histologic types, notably endometrioid and CCCs. In one series originally classified in 2002 by an expert gynecologic pathologist, nearly half of the cases were reclassified after using 2014 WHO criteria (Kommoss et al. 2016). Carcinosarcoma, previously regarded as a rare primary ovarian tumor, now appears to comprise 6% of ovarian carcinomas in the USA. Accurate pathologic classification is predicated on examination of histological sections that are representative of the entire neoplasm. Intratumoral heterogeneity is a common phenomenon in many carcinomas, and ovarian carcinomas are no exception. Type I tumors display a wide variety of patterns and degrees of epithelial proliferation in different areas. Although extensive sampling is not generally needed for invasive carcinomas, noninvasive tumors need careful gross examination and directed sampling of papillary, solid, and any unusual areas to make certain that invasive foci are not overlooked. This is particularly important for the atypical proliferative (borderline) endometrioid, mucinous, clear cell, and low-grade serous cell types.

873

The College of American Pathologists (CAP), WHO, International Collaboration of Cancer Reporting (ICCR), and FIGO Recommendations CAP has issued guidelines for reporting ovarian cancer which were updated in 2016 (Gilks et al. 2016). Support for the 2014 WHO classification of tumors is indicated. Complete reporting with the CAP cancer checklist requires assessment of ovarian surface involvement and tumor rupture for stage I. It is therefore important that the pathologist communicate with the surgeon or review the operative report to provide these data. For advanced-stage tumors, the size of the largest peritoneal nodule needs to be assessed, and although this is often evident from the gross pathologic examination, the surgeon’s input may be required since the tumor may be incompletely resected or resected in a piecemeal fashion. In apparent stage III tumors, clinical or pathologic information may be needed to determine whether distant metastases have been diagnosed (stage IV). Other required elements of the checklist include specimen type, procedure, specimen integrity, primary site, tumor size, histologic type, lymph node numbers and status, extent of involvement of other organs, status of peritoneal and pleural fluid specimens, and pTNM stage. Grade is required for mucinous and endometrioid carcinomas (as noted earlier, HGSC, LGSC, and CCC contain the grade in their definitions); endometrioid carcinomas should be graded using the FIGO system for uterine endometrioid carcinomas (see section “Endometrioid Tumors,” later in this chapter). FIGO stage is recommended but not required. There is an optional chemotherapy response score for patients who have received neoadjuvant chemotherapy. CAP recognizes that LGSC and HGSC are separate tumor types and includes a category of mixed carcinomas. A likely tubal origin of many HGSCs is recognized, as is the unified entity of HGSC of the ovary/fallopian tube/peritoneum. Support for a recent proposal for primary site designation is indicated (Singh et al. 2014). CAP also recognizes that “invasive implants” are invasive LGSCs.

874

J. D. Seidman et al.

Table 8 Distribution of ovarian epithelial tumors by cell type Serous Endometrioid Clear cell Mucinous Seromucinous Transitional Mixed Undifferentiated Carcinosarcoma Squamous Totals

Benign (%) 48.6% 0.8 0 7.6 1.8 9.9 0.6c – – 1.3 70.6

Atypical proliferative (%)a 1.8% 0.2 0.2 1.0 0.3 0.2 0 – – – 3.7

Malignant (%)b 17.8% 1.9 2.2 0.8 0.2 0.3 0.7 0.1 1.6 0.1 25.7

Totals (%) 68.2% 2.9 2.4 9.4 2.3 10.4 1.3 0.1 1.6 1.4 100%

Data are based on 1304 cases from a large community hospital over a 12-year period Also referred to as “borderline” tumors and includes those with intraepithelial carcinoma and/or microinvasion b Includes tubal and peritoneal primaries c 0.5% serous-endometrioid and 0.1% mucinous-endometrioid a

The 2014 WHO Classification (Kurman et al. 2014) contains several notable changes from the 2003 edition. First, the fallopian tube is recognized as the likely source of a large majority of HGSCs. Second, HGSC and LGSC are recognized as distinctive tumor types rather than different grades of the same tumor. Invasive peritoneal implants associated with SBT/APST are equated with invasive LGSC. The micropapillary variant of SBT is equated with noninvasive LGSC. WHO does not recommend a category of mixed carcinoma. Finally, seromucinous tumors are placed in their own category, separate from the majority of mucinous tumors which are of intestinal type (see section “Mucinous Tumors,” later in this chapter). The ICCR recommendations for data reporting are generally in line with the CAP recommendations. ICCR is also aligned with WHO and CAP in recognition of a likely tubal origin and equating invasive implants with invasive LGSC. They also accept the “atypical proliferative” and “noninvasive LGSC” terminology (McCluggage et al. 2015). None of the three organizations recommends the “low malignant potential” terminology. Like CAP, ICCR also supports a recent proposal for primary site designation (Singh et al. 2014), the use of which results in the fallopian tube being considered the primary site in the vast majority of extrauterine HGSC. All three organizations cite the SEE-FIM protocol for complete sampling of the fallopian tubes. CAP

recommends submitting the fimbriae in total only for patients with HGSC with no gross tubal involvement. All three also accept the FIGO grading system for endometrioid carcinoma and accept all CCCs (as well as undifferentiated carcinoma and carcinosarcoma) as high grade. FIGO concurs with the recommendations to group serous carcinomas of ovarian, tubal, and peritoneal origin together, recommending the term “serous carcinoma” rather than “mullerian carcinomas” or “pelvic serous carcinomas.” FIGO also recommends the binary grading system for serous carcinomas and the FIGO grading system for non-serous carcinomas (see section “Endometrioid adenocarcinoma,” later in this chapter). However, FIGO does not clearly recommend the classification of invasive peritoneal implants associated with borderline tumors as invasive carcinomas and their statements on this topic are ambiguous (Berek et al. 2015).

Serous Tumors The classification of serous neoplasms of the ovary has been mired in controversy due to the historically widely misunderstood intermediate category between benign and malignant. The terminology for the borderline group has been greatly clarified as noted in the respective sections below.

14

Epithelial Tumors of the Ovary

The modifiers “cystadeno,” “papillary,” “surface,” “fibro,” and “adeno” for serous carcinomas (i.e., papillary serous cystadenocarcinoma, a term commonly used historically) should not be used, as these create multiple different names for the most common type of ovarian cancer. Because of this multiplicity of names, the International Classification of Diseases for Oncology contains at least four different codes for synonyms for serous carcinoma. This leads to confusion in cancer registries as well as difficulties in the interpretation of populationbased data.

Serous Cystadenoma and Adenofibroma Benign serous tumors include cystadenomas, adenofibromas, cystadenofibromas, and surface papillomas; the general term serous cystadenoma can be used to refer to all of them. These tumors are common and account for two-thirds of benign ovarian epithelial tumors and most ovarian serous tumors. They occur in adults of all ages, with reported mean ages varying widely from 40 to 60 years. The symptoms and signs associated with large tumors are nonspecific and most commonly include pelvic pain, discomfort, or an asymptomatic pelvic mass. Tumors measuring 1–3 cm are usually incidental findings. Bilaterality rates are variable depending upon both the thoroughness with which an apparently uninvolved ovary is examined and the threshold for diagnosis of a small serous neoplasm. Accordingly, 12–23% of cystadenomas are bilateral. Benign serous tumors are equally distributed among unilocular cysts, multilocular cysts, and cystadenofibromas. They are composed of cysts filled with clear watery (serous) fluid or thin mucoid material. Occasionally, they contain thicker mucus-like material more typical of mucinous neoplasms. The external surfaces of the cysts are smooth and glistening. Occasionally, papillary excrescences are found on the external surface of the cyst. The tumors vary widely in size up to 30 cm, with a mean of 5–8 cm. The lining of the cyst is either flat or may have a varying number of coarse papillary projections. Such papillary

875

excrescences rarely cover the entire inner surface of the cyst. Cystadenofibromas are solid neoplasms composed of tough, rubbery tissue with interspersed glandular spaces. Normal sized ovaries often have small surface papillary projections with a fibrotic stromal component resembling a microscopic adenofibroma or cystadenofibroma; these have been termed surface papillomas and when multiple surface papillomatosis (see ▶ Chap. 12, “Nonneoplastic Lesions of the Ovary”). In addition, surface epithelial inclusions may become cystically dilated (CICs). It is common practice to diagnose a serous neoplasm only if the lesion is greater than 1 cm in diameter. This is arbitrary and does not distinguish a true neoplastic growth from simple serous cysts or nonneoplastic hyperplasias of the ovarian cortex. The vast majority of lesions classified as serous cystadenoma display a serous epithelial lining lacking proliferation. Only a small minority have any significant epithelial proliferation. Accordingly, most are not true neoplasms, but rather represent cystically dilated inclusions Cystadenomas are lined by pseudostratified, tubal-type epithelium, with the characteristic elongated (secretory cell) and rounded (ciliated cell) nuclei (Fig. 8). A single layer of flattened to cuboidal cells with uniform basal nuclei is also common. Although the cells may produce mucin

Fig. 8 Serous cystadenoma. A single layer of serous epithelium characterized by ciliated, secretory, and intercalated cells as in the normal fallopian tube

876

which is secreted into the cystic spaces, they do not contain basophilic cytoplasmic vacuoles or granules characteristic of mucinous neoplasms. In large cysts, the epithelium often becomes attenuated. Mitoses and atypia are generally absent. Psammoma bodies are often present in the stroma. There is a broad spectrum of epithelial proliferation in benign serous tumors which is manifested by variation in the prominence and complexity of the papillae, from a simple, single layer, and blunt papillae to focal epithelial stratification and detachment of cell clusters approaching the degree of proliferation seen in APSTs. Identification of these features in 10% of the tumor separates serous cystadenoma from APST. If these features are focal (3:1 variation in nuclear size and shape). Low-grade tumors have a mitotic index less than 12 per 10 HPF, and usually much lower (less than 5 per 10 HPF), while high-grade tumors usually display greater than 12 per 10 HPF. This binary grading system effectively separates typical HGSC from LGSC with 5-year survival rates of about 35% and over 50%, respectively, for FIGO stage III. However, it is possible that patient age and comorbidity have as much to do with this survival difference as grade. Women with HGSC are a decade older and consequently have more comorbid conditions and a shorter life expectancy as compared to those with LGSC. In a population-based study of ovarian cancer in Denmark, ovarian cancer patients with severe comorbidity were about a decade older, and within each stage grouping, patients with more comorbidity had increased mortality. Postoperative mortality (3 low-power (4 objective) magnification fields

which is endorsed by the WHO. In this system, it is the immature neuroepithelium that is graded. The grade is based on the aggregate amount of immature neuroepithelium on any single slide (Norris et al. 1976; O’Connor and Norris 1994) (Table 2). Metastases/implants are considered grade 0 when no immature tissue is present, regardless of the grade of the ovarian tumor. In particular, a specific type of grade 0 implant composed entirely of mature glial tissue (“gliomatosis peritonei”) is associated with an excellent outcome (Liang et al. 2015). The recommended treatment for patients with grade 1 (low-grade) immature teratoma confined to one ovary is unilateral salpingo-oophorectomy and careful follow-up (National Comprehensive Cancer Network 2017). This treatment is curative in nearly all cases (Bonazzi et al. 1994; Peccatori et al. 1995). For grade 2 and 3 (high-grade) tumors, adjuvant chemotherapy is administered following unilateral salpingo-oophorectomy, resulting in complete cure in the majority of patients (Bonazzi et al. 1994). More extensive surgery is necessary if the tumor extends beyond the ovary. The currently recommended combination therapy is the BEP regimen (National Comprehensive Cancer Network 2017). Therapy with vincristine, dactinomycin, and cyclophosphamide (VAC) had been the treatment of choice (Norris et al. 1976) because the results obtained were considered to be similar to those with vinblastine, bleomycin, and cisplatin (VBP) or BEP regimens and the latter are more toxic. However, there is evidence that the recurrence rate with BEP is less than with VAC regimen, and for patients with metastatic disease, the cisplatin-containing

regimens are the treatment of choice, especially the BEP regimen as it is less toxic than the VBP regimen (Bonazzi et al. 1994; Gershenson 1993; Peccatori et al. 1995). In a collaborative study of immature teratoma (ovarian, testicular, and extragonadal) in children (Heifetz et al. 1998), it was demonstrated that pure immature teratomas in this population have a very good prognosis. The authors concluded that the presence of microscopic foci of yolk sac tumor rather than the grade of immature teratoma was the only valid predictor of recurrence; however, it should be noted that ovarian cases were not separately analyzed for correlation between grade and outcome.

Mature Solid Teratoma General Features Mature solid teratoma is an uncommon ovarian teratoma. The age at presentation is similar to that of immature solid teratoma, the tumor occurring mainly in children and young adults (Peterson 1956; Woodruff et al. 1968). Most solid ovarian teratomas are composed at least partly of immature tissues and therefore are considered to be malignant. The occasional cases of solid ovarian teratoma composed entirely of mature tissues have usually been included in this group and thus misinterpreted as malignant. As the presence of immature neural elements immediately excludes the tumor from this group, it is very important to recognize the mature tissue as such, as by definition only tumors lacking immature neuroectoderm may be diagnosed as mature solid teratoma. Gross Features The tumors are usually large, do not exhibit any specific gross features, and show an appearance similar to immature solid teratomas. They grow slowly in comparison with immature solid teratoma, but because they are usually discovered after they have reached a considerable size, this feature is of little help in diagnosis. In all reported cases of mature solid teratoma, the tumor has been unilateral (Peterson 1956; Wisniewski and Deppisch 1973; Woodruff et al. 1968).

16

Germ Cell Tumors of the Ovary

Microscopic Features Mature solid teratoma is composed of mature tissues derived from the three germ layers. Rigid diagnostic criteria must be used, and the examination and sampling of the tumor must be thorough, because inclusion of cases with immature neuroectodermal elements changes the prognosis of this neoplasm (Peterson 1956; Wisniewski and Deppisch 1973; Woodruff et al. 1968). The tissues derived from the three germ layers are arranged in an orderly manner resembling the much more common mature cystic teratoma, except that the neoplasm is solid or at least predominantly solid. Neurogenic elements, which are among the most common tissues present in this tumor, often pose diagnostic problems because they may not be recognized as mature. Occasionally, mature solid teratoma may be associated with peritoneal implants composed entirely of mature glial tissue (gliomatosis) (Fig. 31). Despite extensive peritoneal disease and irrespective of the mode of therapy employed, the prognosis is excellent (Robboy and Scully 1970; Roth and Talerman 2006; Scully et al. 1998). The presence of peritoneal implants composed entirely of mature glial tissue may also be observed occasionally in patients with immature solid teratoma and with mature cystic teratoma. The presence of these implants does not affect the prognosis (Heifetz et al. 1998; Nielsen et al. 1985; Robboy and Scully 1970; Roth and Talerman 2006; Scully et al. 1998).

Fig. 31 Mature glial implant. Mature glial elements on omentum (gliomatosis peritonei)

1079

Clinical Behavior and Treatment Because the tumor is unilateral, oophorectomy or unilateral adnexectomy is the treatment of choice, resulting in a complete cure (Peterson 1956; Wisniewski and Deppisch 1973; Woodruff et al. 1968).

Mature Cystic Teratoma General Features Mature cystic teratoma of the ovary, or dermoid cyst, has been known since antiquity. The tumor is composed of well-differentiated derivatives of the three germ layers – ectoderm, mesoderm, and endoderm – with ectodermal elements predominating. In its pure form, mature cystic teratoma is always benign, but occasionally it may undergo malignant change in one of its elements. It may also form a part of a mixed germ cell tumor. Clinical Features Mature cystic teratoma is the most common type of ovarian teratoma and the most common type of ovarian germ cell neoplasm. It occurs relatively frequently and comprises approximately 20% of all ovarian neoplasms (Peterson et al. 1955; Roth and Talerman 2006; Scully et al. 1998). Mature cystic teratoma occurs most commonly during the reproductive years, but, unlike other germ cell tumors of the ovary, it has a wider age distribution and may be encountered at any age from infancy to old age (Ayhan et al. 2000; Caruso et al. 1971; Peterson et al. 1955). In some series, more than 25% of cases have been observed in postmenopausal women (Malkasian et al. 1967). It has also been encountered in newborns. Mature cystic teratoma is often discovered as an incidental finding on physical examination, radiologic examination, or during abdominal surgery performed for other indications. Patients present with abdominal pain (47.6%), abdominal mass or swelling (15.4%), and abnormal uterine bleeding (15.1%) (Peterson et al. 1955). The abdominal pain is usually constant, slight, or moderate but, in a number of cases, may be severe and acute because of torsion or rupture of the tumor. In children and young adults, the tumors tend to be more easily mobile and therefore are more frequently affected by torsion.

1080

K. P. Maniar and R. Vang

Abnormal uterine bleeding and its cessation after excision of the tumor suggest hormone synthesis by the tumor, but histologic examination has failed to reveal any explanation for the endocrine function (Malkasian et al. 1967). Slightly decreased fertility has been observed in patients with mature cystic teratoma, but in most cases there is no satisfactory explanation. In 10% of cases, the tumor is diagnosed during pregnancy (Ayhan et al. 2000; Caruso et al. 1971). Mature cystic teratoma has been diagnosed radiologically because of the presence of teeth, bone, and cartilage (Malkasian et al. 1967; Peterson et al. 1955). Cytogenetic Features Mature cystic teratomas are diploid, have a normal 46,XX karyotype, and are believed to originate from germ cells after the first meiotic division; thus, they are in contrast to mature teratomas of the postpubertal testis, which are malignant, aneuploid with complex cytogenetic abnormalities including 12p amplification, and considered to originate from other forms of germ cell tumor (Atkin 1973; Linder and Power 1970; Ulbright 2005). Of note, ovarian mature cystic teratomas and prepubertal testicular mature teratomas are similar in that both are diploid, have normal karyotypes, and are benign (Ulbright 2005). Previous studies (Nomura et al. 1983; Parrington et al. 1984) using banding techniques demonstrated diverse modes of origin of mature cystic teratoma. Although most ovarian mature cystic teratomas originate from germ cells after the first meiotic division, some originate before this event (Nomura et al. 1983; Parrington et al. 1984). This distinction also applies to immature ovarian teratomas, which tend to be aneuploid, resembling their testicular counterparts. Gross Features Mature cystic teratoma does not have a predilection for either ovary; 8–15% of cases are bilateral (Peterson et al. 1955; Roth and Talerman 2006; Scully et al. 1998). The tumor varies from very small (0.5 cm) to large (measuring more than 40 cm) and can weigh up to several kilograms. Approximately 60% of mature cystic teratomas measure from 5 to 10 cm, and more than 90%

Fig. 32 Mature cystic teratoma. Well-encapsulated tumor containing hair

measure less than 15 cm (Peterson et al. 1955). The tumor is round, ovoid, or globular, with a smooth, gray–white, and glistening surface (Fig. 32). It is usually freely mobile but occasionally may form adhesions to surrounding structures, especially if there has been leakage. On palpation, the tumor is soft and fluctuant, with firm or hard areas; this is usually observed immediately after its removal, because at room temperature the tumor tends to solidify. The contents of the tumor are liquid at temperatures above 34  C and become solid at temperatures below 25  C (Blackwell et al. 1946). The cut surface of the tumor reveals a cavity filled with fatty material and hair surrounded by a firm capsule of varying thickness. The fatty material is similar to normal sebum. The tumor is usually unilocular but may be multilocular. Several tumors may be present in the same ovary. Arising from the cyst wall and projecting into the cavity is a protuberance that may vary from a small nodule to a rounded elevated mass. It is usually single but may be multiple and is frequently solid but may be partly cystic. This protuberance has been variously termed dermoid mamilla, dermoid protuberance, Rokitansky protuberance, embryonic node, or dermoid nipple. The hair present in the tumor arises from this protuberance, and when bone or teeth are present (Fig. 33), they tend to be located within this area, which is composed of a variety of different tissues and is one of the sites that should always be carefully sampled. Mature cystic teratomas

16

Germ Cell Tumors of the Ovary

1081

Fig. 33 Mature cystic teratoma. Left: Well-encapsulated spherical cystic mass. Arrows point to teeth. Right: The tumor was diagnosed via pelvic radiograph. A row of teeth is seen at arrow tip. (Courtesy of A. Blaustein, M.D.)

contain macroscopically recognizable and wellformed teeth in 31% of cases (Blackwell et al. 1946). Phalanges, long and other bones, parts of the rib cage, loops of intestine, and even fetus-like structures are occasionally encountered (Abbott et al. 1984; Weldon-Linne and Rushovich 1983). These have been classified as fetiform teratomas or homunculi (Abbott et al. 1984; Weldon-Linne and Rushovich 1983). Microscopic Features The outer side of the cyst wall is composed of ovarian stroma that may often be hyalinized, making its recognition difficult. The inner cavity of the cyst is lined mainly by skin, and in small tumors cutaneous structures may form the entire lining. The skin is composed of keratinized squamous epithelium and usually contains abundant sebaceous and eccrine glands associated with fat (Fig. 34). In some tumors, a lipogranulomatous, fat necrosis-like, sievelike, or pneumatosis cystoides-like pattern may be prominent. Hair and other dermal appendages are usually present. Occasionally, the cyst wall may be lined by bronchial or gastrointestinal epithelium or epithelium of columnar or cuboidal type (Fig. 35). The squamous epithelium may be present only in the region of the dermoid protuberance. Sometimes there may be loss of the lining epithelium caused by desquamation, and this may be associated with a foreign body giant cell reaction. The latter may be seen in

Fig. 34 Mature cystic teratoma. The lining of the cyst is composed of skin with its appendages

Fig. 35 Mature cystic teratoma. Mature cystic teratoma lined by well-differentiated mature respiratory epithelium. Mature adnexal structures are seen beneath the lining

1082

other parts of the tumor as a reaction to the contents of the tumor. Foreign body giant cell reactions may also be seen when the contents of the tumor are spilled, leading to the formation of adhesions. The area around the dermoid protuberance may contain a large variety of tissues derived from the three germ layers. Ectodermal tissue, represented by squamous epithelium and other skin derivatives, is usually most abundant. Brain tissue, glia, neural tissue, retina, choroid plexus, and ganglia may also be encountered. In occasional cases, the glial tissue may be highly cellular, making the distinction between gliosis and a low-grade glial tumor, such as astrocytoma or oligodendroglioma, arising in a teratoma very difficult. Criteria have not been established for this distinction in the ovary. Furthermore, in view of the rarity of this problem in teratomas, the precise behavior of such lesions is not known. Mesodermal tissue is represented by bone, cartilage, smooth muscle, and fibrous and fatty tissue. Endodermal tissue is represented by gastrointestinal and bronchial epithelium and glands, thyroid, and salivary gland tissue. In a careful study of 100 cases, ectodermal structures were found in 100%, mesodermal in 93%, and endodermal in 71% of cases (Blackwell et al. 1946). Rare tissues, such as prostate, have been reported (Halabi et al. 2002). The various tissues present in mature cystic teratoma show an orderly organoid arrangement forming cutaneous, bronchial, and gastrointestinal tissues, as well as bone and other structures. Although these tissues may be scattered diffusely, they do not exhibit the disorderly haphazard arrangement that is observed in immature teratoma. With the exception of thyroid tissue, the presence of endocrine tissue is distinctly uncommon in mature cystic teratoma, but pituitary, adrenal, and parathyroid tissues have been documented. Occasionally functioning endocrine tissue, forming an adenoma, may be found in a mature cystic teratoma (Roth and Talerman 2006; Scully et al. 1998). Mature cystic teratoma must be differentiated from the rare cases of fetus in fetu, considered most likely to be caused by an inclusion of a monozygotic diamniotic twin (Brand et al. 2004). Fetus in fetu can be distinguished

K. P. Maniar and R. Vang

from a teratoma by its location in the retroperitoneal space, presence of vertebral organization with formation of limb buds, and a welldeveloped organ system. Fetus in fetu shows better organization than the most differentiated teratomas. Like mature cystic teratoma, fetus in fetu is a benign lesion (Brand et al. 2004). Clinical Behavior and Treatment Mature cystic teratoma of the ovary may be associated with various complications. In view of the fact that in many of these cases the condition is amenable to cure, their recognition is of considerable importance. These complications include (1) torsion, (2) rupture, (3) infection, (4) hemolytic anemia, (5) paraneoplastic encephalitis, and (6) development of malignancy. Torsion is the most frequent complication (Ayhan et al. 2000; Caruso et al. 1971; Pantoja et al. 1975a; Peterson et al. 1955), occurring in 16.1% of cases in one large series (Peterson et al. 1955). This complication tends to be more common during pregnancy and puerperium (Malkasian et al. 1967; Peterson et al. 1955). Mature cystic teratoma is said to comprise from 22% to 40% of ovarian tumors in pregnancy, and from 0.8% to 12.8% of reported cases of mature cystic teratoma have occurred during pregnancy (Caruso et al. 1971; Peterson et al. 1955). The fact that these tumors, when they occur during pregnancy, are more liable to be associated with torsion is of considerable importance. Torsion is also more common in children and younger patients (Pantoja et al. 1975a; Peterson et al. 1955). The patients usually have severe acute abdominal pain, and the condition is an acute abdominal emergency. Excision of the affected ovary or salpingo-oophorectomy is the treatment of choice. Torsion tends to predispose to rupture of the tumor. Rupture of mature cystic teratoma is an uncommon complication, occurring in approximately 1% of cases (Malkasian et al. 1967; Peterson et al. 1955). The immediate result of the rupture may be shock or hemorrhage, especially during pregnancy or labor, but the prognosis even in these cases is usually favorable. Rupture of the tumor into the peritoneal cavity may be followed by chemical peritonitis caused by the spillage of

16

Germ Cell Tumors of the Ovary

tumor contents. It produces a marked granulomatous reaction and leads to the formation of dense adhesions throughout the peritoneal cavity. Rupture of the tumor occasionally may be followed by the development of glial implants on the peritoneum. This condition occurs when the tumor contains mature neuroglial elements, and spillage leads to deposition of numerous small nodules composed of mature glia in the peritoneal cavity. Despite the wide dissemination of these deposits throughout the peritoneal cavity, the prognosis is favorable, and simple surgical excision of the primary tumor is considered to be adequate therapy (Robboy and Scully 1970). Mature cystic teratoma may rupture not only into the peritoneal cavity but also into adjacent organs, usually the bladder or the rectum. Several such cases have been reported (Dandia 1967). Infection is an uncommon complication and occurs in approximately 1% of cases (Malkasian et al. 1967). The infecting organism is usually a coliform, but Salmonella infection causing typhoid fever has also been reported (Hingorani et al. 1963). Autoimmune hemolytic anemia has been noted occasionally in patients with teratoma of the ovary, mainly mature cystic teratoma. A small number of mature cystic teratomas and other cystic ovarian tumors associated with this complication have been reported (Bernstein et al. 1974; Payne et al. 1981). The patients have symptoms and signs of progressive anemia, which may be moderate or severe; it is accompanied by reticulocytosis, spherocytosis, and increased osmotic fragility. Normoblasts may be present in the peripheral blood. The indirect serum bilirubin is elevated, and the direct antiglobulin test (Coombs test) is positive, indicating the presence of autoantibodies that react with the patient’s red blood cells. The platelets are normal in number. The spleen may be palpable but is only slightly enlarged. Steroids are only transiently effective in treating the disease, and splenectomy has no effect on the progress of the disease (Bernstein et al. 1974; Payne et al. 1981). Excision of the ovarian tumor leads to the permanent disappearance of the anemia (Bernstein et al. 1974; Dawson et al. 1971; Payne et al. 1981). The following possible

1083

pathogenetic mechanisms have been suggested (Bernstein et al. 1974) (1) Presence in the tumor of substances that are antigenically different from the host and that stimulate the production of antibodies by the host, which cross-react with the patient’s own red blood cells; (2) Antibody production by the tumor directed specifically against the host’s red blood cells resembling the graft-versus-host reaction; and (3) Coating of red blood cells with products secreted by the tumor, resulting in changed red blood cell antigenicity. In view of this, pelvic and radiologic examination is indicated in a young woman with autoimmune hemolytic anemia that does not respond to steroid treatment, as it may help detect an ovarian teratoma and prevent an unnecessary splenectomy (Payne et al. 1981). A syndrome of paraneoplastic limbic encephalitis has been reported in multiple cases of both mature and immature teratoma and has been demonstrated to be secondary to antibodies against subunits of the N-methyl-D-aspartate (NMDA) receptor; these subunits are expressed in the neural and/or squamous tissue of the associated teratomas (Clark et al. 2014; Dalmau et al. 2007; Gultekin et al. 2000; Vitaliani et al. 2005). The syndrome is characterized by acute psychiatric symptoms, seizures, memory deficits, altered level of consciousness, central hypoventilation, and inflammatory abnormalities in the cerebrospinal fluid (Vitaliani et al. 2005), and the responsible antibodies can be isolated from the serum and the cerebrospinal fluid (Dalmau et al. 2007). Encephalitis-related tumors show an increased intratumoral lymphoid infiltrate associated with mature neuroglial elements, including reactive germinal centers, diffuse lymphoplasmacytic infiltrates, and degenerative neuronal changes (Dabner et al. 2012). Treatment consists of tumor resection and immunotherapy, which is effective in the majority of cases (Dalmau et al. 2007). The treatment of choice for an uncomplicated mature cystic teratoma in young patients is excision of the cyst with conservation of part of the ovary if possible. This treatment usually results in a complete cure. Local recurrences after conservative treatment for mature cystic teratoma are uncommon and occur in less than 1% of cases.

1084

Mature Cystic Teratoma (Dermoid Cyst) with Malignant Transformation General Features Malignant transformation is an uncommon complication of mature cystic teratoma. It occurs in approximately 2% of cases (Ayhan et al. 2000; Krumerman and Chung 1977; Malkasian et al. 1967; Park et al. 2008; Peterson 1957; Roth and Talerman 2006; Scully et al. 1998; Stamp and McConnell 1983), although in one report the frequency was almost 4% (Pantoja et al. 1975b). The age of patients with this complication as reported in the literature ranges from 19 to 88 years (Peterson 1957), but this tumor usually is observed in postmenopausal patients (Krumerman and Chung 1977; Malkasian et al. 1967; Pantoja et al. 1975b; Peterson 1957; Stamp and McConnell 1983). The most common presenting symptom is abdominal pain (Chen et al. 2008). While this tumor cannot always be readily clinically differentiated from an uncomplicated mature cystic teratoma, features such as larger tumor size (>10 cm), older patient age (>50 years), and elevated tumor antigens, in particular CA125 and squamous cell carcinoma (SCC) antigen, have been demonstrated to be predictive of malignancy (Chen et al. 2008; Dos Santos et al. 2007; Hackethal et al. 2008). Certain imaging findings on Doppler ultrasound, CT, and MRI have also been shown to have predictive value for malignancy (Chiang et al. 2011; Dos Santos et al. 2007). Sometimes, the tumor may be an incidental finding. Gross Features The tumor is frequently larger than the average mature cystic teratoma (Caruso et al. 1971; Krumerman and Chung 1977; Scully et al. 1998; Stamp and McConnell 1983); it may exhibit a more solid appearance (Fig. 36), but differentiation usually cannot be made on gross examination. Malignant transformation in mature cystic teratoma tends to occur in patients with unilateral tumors (Peterson 1957; Peterson et al. 1955).

K. P. Maniar and R. Vang

Microscopic Features The most common tumor component to undergo malignant transformation is squamous epithelium, with formation of a typical squamous cell carcinoma (Figs. 37 and 38) (Caruso et al. 1971; Hirakawa et al. 1989; Krumerman and Chung 1977; Pantoja et al. 1975b; Peterson 1957; Stamp and McConnell 1983). Any of the tissues present in a mature cystic teratoma may undergo malignant transformation, and a variety of malignant tumors have been reported, including mucinous carcinoma, carcinoid tumor, thyroid carcinoma, basal cell carcinoma, sebaceous carcinoma, malignant melanoma, leiomyosarcoma, chondrosarcoma, and angiosarcoma (Figs. 39 and 40) (Gupta et al. 2004; McCluggage et al. 2006;

Fig. 36 SCC (left) arising in and overgrowing mature cystic teratoma

Fig. 37 SCC arising in mature cystic teratoma. The tumor has an infiltrating pattern showing numerous keratin pearls

16

Germ Cell Tumors of the Ovary

1085

Fig. 38 SCC arising in a mature cystic teratoma. The tumor displays a pushing pattern Fig. 40 Invasive mucinous carcinoma arising within mature cystic teratoma. Note extensive glandular component, slightly haphazard glandular arrangement, and prominent pseudomyxoma ovarii. The diagnosis of invasion in examples such as this can be very difficult, particularly with regard to distinction from non-carcinomatous tumors that have pseudomyxoma ovarii coupled with a slightly less extensive glandular proliferation and density of glands

Fig. 39 Melanoma arising in mature cystic teratoma. Note nested pattern within squamous epithelium

McKenney et al. 2008; Peterson 1957; Ueda et al. 1993; Vang et al. 2007; Venizelos et al. 2009). The malignant element invades other parts of the tumor and its wall, which may perforate. The malignant component may overgrow the remaining part of the mature cystic teratoma and pose diagnostic problems. Clinical Behavior and Treatment The mode of spread of the malignant tumor differs from that observed in other tumors of germ cell origin. The tumor spreads by direct invasion and peritoneal implantation and generally does not metastasize to the lymph nodes (Krumerman and Chung 1977; Peterson 1957). Extensive local invasion and absence of lymph node involvement

usually is observed at laparotomy (Pantoja et al. 1975b; Stamp and McConnell 1983). Hematogenous dissemination is uncommon. The prognosis of patients with mature cystic teratoma with malignant transformation is unfavorable (Krumerman and Chung 1977; Pantoja et al. 1975b; Peterson 1957; Stamp and McConnell 1983); only 15–48.4% of patients survive 5 years (Chen et al. 2008; Krumerman and Chung 1977; Peterson 1957). Better prognosis has been reported when the malignant element is a SCC confined to the ovary and is excised without spillage of the contents. In such cases, the reported 5-year survival is 75.7% (Chen et al. 2008). Treatment is hysterectomy and bilateral adnexectomy (Krumerman and Chung 1977; Stamp and McConnell 1983). Because the tumors are usually unilateral, in cases where there is no penetration of the capsule and no involvement of the adjacent structures, a more conservative surgical procedure may be just as effective. However, because malignant transformation of a mature cystic teratoma usually occurs in postmenopausal women, total hysterectomy and bilateral salpingo-

1086

oophorectomy is performed. If the tumor has spread beyond the confines of the ovary and there is involvement of the adjacent structures, a more radical procedure with resection of the tumor and the involved structures or viscera is advocated (Pantoja et al. 1975b). For SCC adjuvant chemotherapy has been demonstrated to improve survival, in particular for higher-stage patients (Chen et al. 2008). Some authors have reported a benefit with whole pelvic radiation, while others have not confirmed the utility of this treatment modality (Chen et al. 2008; Chiang et al. 2011; Dos Santos et al. 2007; Hackethal et al. 2008). Poor prognostic factors include higher stage at presentation, older patient age, larger tumor size, and positive serum tumor markers (Chen et al. 2008).

Mucinous Tumors Arising in Mature Cystic Teratoma General Features In general, the vast majority of primary ovarian mucinous tumors are of epithelial origin, but a subset is of germ cell origin. In the cases studied, 2–11% of ovarian mature cystic teratomas are associated with a mucinous tumor, and many of the latter in this setting are likely of germ cell origin. Further evidence for this concept is provided by the occurrence of occasional teratomatous tumors composed mainly of mucinous (intestinal-type) epithelium of endodermal derivation and only a small amount of other teratomatous elements, as well as the presence of intestinal-type mucinous epithelium as the only other tissue element present in association with some cases of struma ovarii and strumal carcinoid. In 21% of cases, the epithelium lining mucinous tumors of the ovary contains argyrophil and argentaffin granules, and Paneth cells are also present in a considerable number of cases. These findings are considered to be a strong argument in favor of the derivation of at least some mucinous ovarian tumors from intestinal-type epithelium of teratomatous (germ cell) origin. A number of molecular studies have provided evidence that at least a subset of mucinous tumors associated with teratomas are of germ cell origin (Fujii et al. 2014; Kerr et al. 2013; Magi-Galluzi et al. 2001; Snir et al. 2016; Wang et al. 2015).

K. P. Maniar and R. Vang

Microscopic Features Primary ovarian mucinous tumors associated with a mature cystic teratoma show a spectrum of histologic appearances (McKenney et al. 2008; Vang et al. 2007). At the lower end of the spectrum, tumors display a cystadenomatous pattern. Proliferative tumors with architectural complexity and epithelial stratification resemble atypical proliferative (borderline) mucinous tumor of ovarian epithelial origin or low-grade adenomatous mucinous neoplasm of the appendix. Pseudomyxoma ovarii can be seen with some cystadenomatous and proliferative neoplasms. Compared to tumors without pseudomyxoma ovarii, neoplasms with pseudomyxoma ovarii more closely resemble lower gastrointestinal tract adenomatous tumors and tend to have hypermucinous columnar epithelium and abundant goblet cells. Other tumors may show goblet cell carcinoid-like morphology. At the upper end of the spectrum, the carcinomatous neoplasms may be of glandular (Fig. 40) or signet ring cell type. Pseudomyxoma ovarii can also be associated with goblet cell carcinoid-like tumors or carcinoma. Immunohistochemical Features and Differential Diagnosis Immunohistochemical stains for CK7 and CK20 show variable coordinate expression profiles (Vang et al. 2007). Tumors without pseudomyxoma ovarii and having cystadenomatous or proliferative patterns show a variety of CK7/CK20 profiles, including a CK7 diffuse/ CK20 variable pattern (a pattern frequently seen in ovarian epithelial tumors). Those with pseudomyxoma ovarii and having cystadenomatous, proliferative, or goblet cell carcinoid-like patterns characteristically display a CK7( )/CK20 diffuse or CK7 focal/CK20 diffuse profile (patterns typical of lower gastrointestinal tract tumors). A minority of these tumors histologically and immunohistochemically resembling lower gastrointestinal tract adenomatous tumors can have the clinical syndrome of pseudomyxoma peritonei without a tumor in the appendix. Parenthetically, it should be emphasized that although nearly all cases of pseudomyxoma peritonei are of appendiceal origin, rare cases are of primary

16

Germ Cell Tumors of the Ovary

ovarian origin due to an appendiceal-type mucinous tumor arising within a mature cystic teratoma. The carcinomas can have variable CK7/CK20 profiles, but some will show a CK7( )/CK20 diffuse or CK7 focal/CK20 diffuse pattern. Ovarian mucinous tumors of germ cell origin with histologic and immunohistochemical features typical of primary lower gastrointestinal tract tumors can be misclassified as metastatic or secondary tumors involving the ovary. Thus, it is important to search for focal teratomatous components in such ovarian mucinous tumors in order to suggest a possible primary ovarian origin. Nonetheless, when problematic mucinous neoplasms in ovarian mature cystic teratomas histologically and immunohistochemically resemble lower gastrointestinal tract tumors, extensive sampling of the gross specimen and further clinical evaluation to exclude the rare possibility of a similar primary mucinous tumor in the appendix or colorectal region as part of a tumor-to-tumor metastasis (e.g., a primary lower gastrointestinal tract tumor with a metastasis to a coexisting ovarian teratoma) are recommended. Primary ovarian mucinous tumors arising in a teratoma, which histologically and immunohistochemically resemble lower gastrointestinal tumors, are considered to be of germ cell origin. Tumors that are histologically and immunohistochemically analogous to ovarian epithelial mucinous cystadenoma or atypical proliferative (borderline) mucinous tumor may have developed in the same ovary containing a teratoma as an independent tumor; however, it should also be considered that some of those mucinous tumors could be of germ cell origin, having potentially arisen from upper gastrointestinal/pancreaticobiliary or sinonasal tissue in a teratoma, which would have histologic and immunohistochemical features similar to mucinous tumors of ovarian epithelial origin. Nomenclature For primary ovarian mucinous tumors histologically and immunohistochemically resembling lower gastrointestinal tract tumors, descriptive terminology that parallels the nomenclature for tumors in lower gastrointestinal sites (e.g., low-grade mucinous neoplasm for ovarian tumors

1087

histologically and immunohistochemically analogous to those of the appendix) is preferred, considering that (a) terms such as borderline tumor and atypical proliferative tumor are used for primary epithelial tumors of the ovary, (b) these mucinous tumors are of germ cell rather than epithelial origin, and (c) they resemble their counterparts in the lower gastrointestinal tract. Clinical Behavior and Treatment Data on the behavior of mucinous ovarian tumors of germ cell origin are limited, but in two series patients with cystadenomatous and proliferative/ low malignant potential tumors on follow-up remained well and disease-free (McKenney et al. 2008; Vang et al. 2007). In the same series, mucinous carcinomas showed variable outcomes but exhibited the potential for aggressive behavior. However, one report (Ueda et al. 1993) documented a case of a patient with an adenocarcinoma arising in a mature cystic teratoma who has survived for more than 15 years.

Struma Ovarii General Features Thyroid tissue is a relatively frequent constituent of mature cystic teratoma and has been demonstrated in 5–20% of cases. Struma ovarii is considered a one-sided development of a teratoma in which the thyroid tissue has overgrown all other tissues or one in which only the thyroid tissue has developed. The term struma ovarii should be reserved for tumors composed either entirely or predominantly of thyroid tissue (Roth and Talerman 2006; Roth and Talerman 2007; Scully et al. 1998). Clinical Features Struma ovarii is uncommon; it comprises nearly 3% of ovarian teratomas. The age distribution of patients with struma ovarii is generally the same as that of patients with mature cystic teratoma and ranges from 6 to 74 years. Most patients are in the reproductive years (Roth and Talerman 2006; Roth and Talerman 2007; Woodruff et al. 1966). There are usually no specific symptoms; the clinical findings are similar to those observed in

1088

K. P. Maniar and R. Vang

patients with mature cystic teratoma. The only differences are that in some cases struma ovarii is associated with enlargement of the thyroid gland and in other cases there is clinical evidence that the struma ovarii is responsible for the development of thyrotoxicosis, although this has not been confirmed preoperatively by laboratory tests (Smith 1946; Woodruff et al. 1966). The ectopic thyroid tissue present within struma ovarii, therefore, may be subject to the same physiologic and pathologic changes as the eutopic thyroid gland (Smith 1946). Gross Features Struma ovarii is usually unilateral but is often associated with mature cystic teratoma and rarely with a cystadenoma in the contralateral ovary (Roth and Talerman 2006; Roth and Talerman 2007; Scully et al. 1998; Woodruff et al. 1966). In some cases, the teratoma present in the contralateral ovary also contained thyroid tissue. Struma ovarii varies in size, but usually measures less than 10 cm. The surface is usually smooth, and, before sectioning, the tumor resembles a mature cystic teratoma. Occasionally, adhesions may be present. The cut surface of the tumor may be composed entirely of light tan, glistening thyroid tissue. Hemorrhage, necrosis, and foci of fibrosis may be present. Solid tumors with small amounts of colloid appear less glistening and more fleshy. Some tumors may be cystic (Szyfelbein et al. 1994). Microscopic and Immunohistochemical Features The tumor is composed of mature thyroid tissue consisting of follicles of various sizes, lined by a single layer of columnar or flattened epithelium (Fig. 41). The follicles contain eosinophilic, PAS-positive colloid. The intensity of the staining may vary. There may be considerable variation in the size of the thyroid follicles, which may be large, containing a large amount of colloid, or may be small. Thyroglobulin and TTF-1 can be identified in the epithelial cells by immunohistochemistry (Fig. 42). Occasionally, the lining of the follicles may be columnar, containing small papillary projections not unlike those seen in a

Fig. 41 Struma ovarii. The tumor is composed of normal thyroid tissue

Fig. 42 Struma ovarii showing diffuse expression of TTF-1

hyperactive thyroid gland. Sometimes the appearance may resemble a nodular adenomatous goiter. Adenoma-like lesions may also be observed. When the tumor exhibits markedly crowded follicles without features of malignancy, the designation “proliferative struma ovarii” has been suggested (Fig. 43) (Devaney et al. 1993). Struma ovarii showing appearances suggestive of Hashimoto’s thyroiditis has also been reported (Erez et al. 1965). Rarely, the tumor may show clear cell or cystic patterns (Szyfelbein et al. 1994; Szyfelbein et al. 1995). Clinical Behavior and Treatment Most cases of struma ovarii are benign and can be treated by excision of the ovary or by unilateral

16

Germ Cell Tumors of the Ovary

Fig. 43 Proliferative struma ovarii. The follicles are markedly crowded, but architectural features diagnostic of follicular thyroid carcinoma (invasion of cortex with growth on ovarian serosa, vascular invasion) are not seen

1089

mature thyroid tissue. The condition is thought by some to be benign and is termed benign strumosis (Karseladze and Kulinitch 1994), although some authorities believe that it represents a very well-differentiated follicular carcinoma (Rose et al. 1974; Roth and Karseladze 2008). This condition is only rarely associated with untoward side effects, which are mainly caused by the formation of adhesions. As most patients with this rare condition have been treated by excision of the tumor deposits and by administration of radioactive iodine (131I) with successful outcomes, or have been lost to follow-up after varying periods of time with no ill effects, it is not possible to resolve this controversy. However, rare patients with strumosis have died of disease (Rose et al. 1974).

Malignancy Arising in Struma Ovarii salpingo-oophorectomy. In a small number of cases, there are complications, the most important being the development of malignancy (Devaney et al. 1993; Garg et al. 2009; Rose et al. 1974; Roth and Talerman 2006; Roth and Talerman 2007). In one series (Devaney et al. 1993), proliferative struma ovarii was benign; however, in another series 26% of cases were clinically malignant, but the entire cohort of proliferative struma ovarii had good overall survival (98% survival at 10 years, 92% at 25 years) (Robboy et al. 2009). However, due to the lack of clarity as to which morphologic features predict malignancy, some authors have advocated regarding all proliferative struma as having malignant potential (Robboy et al. 2009; Shaco-Levy et al. 2012). Another complication is the presence of ascites or ascites associated with pleural effusion producing a pseudo-Meigs syndrome (Kawahara 1963). Ascites may be found in 17% of cases of struma ovarii, and its presence does not indicate that the tumor is malignant (Smith 1946). The cause of the ascites and pleural effusion has not been fully elucidated. In most reported cases, excision of the tumor led to complete remission. Occasionally, struma ovarii may be associated with extra-ovarian extension caused either by rupture of the tumor or by local spread. In such cases, the peritoneal cavity contains tumor deposits, which may be numerous and are composed of

Clinical Features Malignant change in struma ovarii, in which the tumor shows histologic/cytologic features of thyroid carcinoma (“malignant struma ovarii”), is uncommon (3% of all cases of struma ovarii in one study (Robboy et al. 2009)). A number of reported cases of malignant struma ovarii were actually examples of strumal carcinoid. Patients with malignant struma ovarii are relatively young. The mean ages of women with the different types of thyroid carcinoma arising in struma ovarii range from 38 to 46 years (Roth et al. 2008). Most patients with papillary thyroid carcinoma and typical follicular carcinoma present with stage I disease. Microscopic Features More than a hundred well-documented cases have been reported (Garg et al. 2009; Robboy et al. 2009; Roth et al. 2008). Microscopically, most of the tumors were of the papillary type (Fig. 44), including its follicular variant, followed by follicular carcinoma. A few tumors showed features of poorly differentiated carcinoma. BRAF mutations and RET/PTC rearrangements, as seen in papillary thyroid carcinoma of the eutopic thyroid gland, have been identified in papillary thyroid carcinoma arising in struma ovarii (Boutross-Tadross et al. 2007; Schmidt et al. 2007).

1090

K. P. Maniar and R. Vang

Fig. 44 Papillary thyroid carcinoma arising in struma ovarii. (a) Low power. Struma ovarii is present at the bottom, and carcinoma is present at the top. (b) Prominent

papillary architecture. (c) Nuclear features characteristic of papillary thyroid carcinoma

In a number of reported cases, the diagnosis was based on the histology of the tumor, and there were no metastases or other features of malignancy (Devaney et al. 1993; Roth et al. 2008; Roth and Talerman 2007). Whether malignant struma ovarii should be diagnosed based on the same criteria used for tumors in the eutopic thyroid gland is unclear. Application of the eutopic thyroid diagnostic criteria for the follicular variant of papillary thyroid carcinoma (FVPTC) in struma ovarii (optically clear nuclei, overlapping nuclei, nuclear pseudoinclusions, nuclear grooves (Robboy et al. 2009)) may be difficult given the recognized interobserver variability in diagnosis for tumors in eutopic thyroid locations. It should be noted that even ovarian tumors with subtle atypical nuclear features, which are suggestive of but not fully diagnostic for the FVPTC, still have the ability to metastasize (Garg et al. 2009) and

that the usual nuclear features distinguishing benign from malignant thyroid tissue (clearing, grooves, pseudoinclusions, and overlap) are not necessarily reliable in distinguishing cases of clinically benign vs. malignant struma (Robboy et al. 2009; Shaco-Levy et al. 2012). Therefore, it has been recommended that whenever an ovarian follicular thyroid-type tumor with cytologic features borderline for the FVPTC is encountered, it should be reviewed by at least two surgical pathologists with experience in thyroid pathology in order to prevent under- or overdiagnosis (Garg et al. 2009). Additionally, it is not yet clear if and how the evolving terminology for the FVPTC, many cases of which are now termed “noninvasive follicular thyroid neoplasm with papillary-like nuclear features” (NIFTP) based on their indolent clinical behavior (Nikiforov et al. 2016), should be applied to struma ovarii.

16

Germ Cell Tumors of the Ovary

Application of eutopic thyroid criteria for the diagnosis of follicular carcinoma in struma ovarii (tumor invasion of cortex with growth on ovarian serosa and vascular invasion (Robboy et al. 2009)) is also problematic, particularly because the ovary lacks a true capsule; additionally, vascular invasion has not been found to be a good predictor of clinically benign vs. malignant struma ovarii (Robboy et al. 2009; Shaco-Levy et al. 2012). Furthermore, ovarian tumors that histologically qualify as struma ovarii but which have an extra-ovarian metastasis or recurrence that histologically resembles nonneoplastic thyroid tissue have been designated “highly differentiated follicular carcinoma” (Roth and Karseladze 2008). Thus, this form of follicular carcinoma can only be diagnosed when an extra-ovarian lesion is identified. It should be noted, however, that this entity is controversial and considered “peritoneal strumosis” by others. However, tumor deposits of obviously malignant papillary thyroid carcinoma and typical follicular carcinoma should be diagnosed as such. For any ovarian tumor featuring a problematic proliferation of thyroid tissue, extensive sampling is strongly suggested. In rare and unusual cases (especially those with bilateral tumors), it may be necessary to recommend further clinical evaluation in order to exclude the possibility of a metastasis from the thyroid gland with secondary involvement of the ovary. Clinical Behavior and Treatment See above section on Struma Ovarii for behavior of proliferative struma ovarii. Metastases of malignant struma ovarii are uncommon (Garg et al. 2009; Roth et al. 2008; Shaco-Levy et al. 2012). In addition to peritoneal involvement, other routes of spread are via the lymphatics to the para-aortic and other lymph nodes and via the bloodstream to the lungs and bones (Roth et al. 2008). The disease is treatable with good outcome in most cases, including those with metastatic disease. Only 7%, 14%, and 0% of patients with papillary thyroid carcinoma, typical follicular carcinoma, and “highly differentiated follicular carcinoma,” respectively, in struma ovarii died of disease (Roth et al. 2008). A recent analysis of

1091

the SEER database also demonstrated excellent disease-specific survival, with an overall survival of 96.7%, 94.3%, and 84.9% at 5, 10, and 20 years, respectively (Goffredo et al. 2015). In one study, of all histologically malignant thyroidtype tumors in struma ovarii, 81% and 60% of patients were alive at 10 years and 25 years, respectively (Robboy et al. 2009). In that study, malignant struma ovarii was assessed for pathologic features that might predict aggressive behavior, and the size of the strumal component correlated with malignant outcome. It was also demonstrated that abundant peritoneal fluid, numerous adhesions, and ovarian serosal defects were more common in clinically malignant tumors. In general, however, it is not possible to reliably determine which cases will develop progressive disease. Treatment of malignant struma ovarii should consist of at least oophorectomy but may also include thyroidectomy, radioactive iodine (131I), and follow-up with serum thyroglobulin measurement. Long-term follow-up is important as metastases can occur decades later.

Carcinoid Carcinoid tumors of the ovary may be primary or metastatic. Primary carcinoids are subdivided into four categories: (1) insular, (2) trabecular, (3) strumal, and (4) mucinous. In the 2014 WHO classification, carcinoids are also referred to as “well-differentiated neuroendocrine tumor, grade 1” (Prat et al. 2014). Mixed types also occur (composed of any combination of the pure types). The latter are uncommon and often associated with a mature cystic teratoma. Of the metastatic carcinoids, the insular carcinoid tumor is the most common, followed by the trabecular and mucinous types. Metastatic carcinoid tumors of the ovary are discussed in ▶ Chap. 18, “Metastatic Tumors of the Ovary.” Insular Carcinoid General Features

Insular carcinoid tumor, considered to be of midgut derivation, is the most common type of primary ovarian carcinoid tumor. It usually arises in

1092

association with gastrointestinal or respiratory epithelium present in a mature cystic teratoma. It may also be observed within a solid teratoma, a mucinous tumor, in association with a Sertoli–Leydig cell tumor (Young et al. 1982), or may occur in a pure form (Robboy et al. 1975; Soga et al. 2000; Talerman 1984). The latter is considered to arise either as a one-sided development of a teratoma or from enterochromaffin cells present within the ovary. The former is much more likely. Approximately 40% of ovarian insular carcinoids occur in pure form; the remaining 60% are combined (Scully et al. 1998; Soga et al. 2000). Clinical Features

More than 200 cases of primary ovarian insular carcinoid tumors have been reported (Davis et al. 1996; Robboy et al. 1975; Scully et al. 1998; Soga et al. 2000). The age of patients ranges from 31 to 83 years, but most patients are either postmenopausal or perimenopausal (Davis et al. 1996; Robboy et al. 1975; Soga et al. 2000). One third of the reported cases have been associated with the typical carcinoid syndrome, despite the absence of metastases (Davis et al. 1996; Robboy et al. 1975; Soga et al. 2000). This is in contrast to intestinal carcinoids, which are associated with the syndrome only when there is metastatic spread to the liver. The reason for this difference is that the blood flow from the ovary goes directly into the systemic circulation and does not pass through the liver, which inactivates the serotonin produced by the tumor. The presence or absence of symptoms of carcinoid syndrome is also dependent on the number of secreting tumor cells. Functioning ovarian carcinoid tumors, with only one exception, have all measured approximately 10 cm in greatest dimension, whereas intestinal carcinoids are usually smaller. Thus, there is a good correlation between the size of the tumor and the presence of carcinoid syndrome. The excision of the tumor is associated with rapid remission of the symptoms, disappearance of 5-hydroxyindoleacetic acid (5-HIAA) from the urine (Robboy et al. 1975), and marked decrease of serum serotonin. Determination of serum serotonin and urinary 5-HIAA may be used to monitor disease activity and response to

K. P. Maniar and R. Vang

therapy. If the tumor is nonfunctioning, there is no specific presentation. Gross Features

The tumor shows similar appearances to those of mature cystic teratoma, within which it is usually found. The same applies if the tumor is associated with a solid teratoma or a mucinous tumor. If the carcinoid is not associated with other tissue elements, the tumor is solid. The carcinoid may vary from microscopic to 20 cm in greatest dimension and is solid and homogeneous. Its color may vary from light brown to yellow or pale gray. Primary ovarian carcinoids are practically always unilateral, although they may be associated with a mature cystic teratoma in the contralateral ovary. Microscopic Features

Primary ovarian insular carcinoid usually shows the typical appearance associated with midgut carcinoids (Robboy et al. 1975). The tumor is composed of collections of small acini and solid nests of uniform polygonal cells with ample amounts of cytoplasm and round or oval, centrally located hyperchromatic nuclei (Fig. 45). Mitotic activity is low. The cytoplasm is basophilic or amphophilic and may contain red, brown, or orange argentaffin or argyrophil granules, which are demonstrated in the majority of cases of primary ovarian carcinoid (Robboy et al. 1975). Ultrastructurally, the cells of the ovarian insular carcinoid show similar appearances to those of insular carcinoid tumors from other locations (Soga et al. 2000) and show abundant neurosecretory granules, which exhibit marked variation in size and shape, being round, oval, or elongated. Immunohistochemical Features and Differential Diagnosis

Demonstration of immunohistochemical expression of chromogranin and synaptophysin (Fig. 46) further supports the diagnosis and has become the method of choice to confirm the diagnosis. These two markers frequently show diffuse and strong staining (Zhao et al. 2007). Serotonin may be demonstrated within the cytoplasm of the tumor cells by immunohistochemical techniques (Sporrong et al. 1982). Occasionally, other

16

Germ Cell Tumors of the Ovary

Fig. 45 Primary insular carcinoid tumor of the ovary. Note the bright pink staining of the tumor cells due to the presence of neurosecretory granules

Fig. 46 Primary insular carcinoid tumor of the ovary. The tumor shows diffuse expression of chromogranin. Note that the compartments of the cells showing chromogranin expression are the same compartments containing bright pink cytoplasmic granules in Fig. 45

neurohormonal polypeptides may also be demonstrated within the cytoplasm of the tumor cells, but their finding is much less frequent than in trabecular or strumal carcinoids (Sporrong et al. 1982). The connective tissue surrounding the tumor nests is frequently dense and hyalinized as a result of the fibrogenic effect of the serotonin produced by the tumor. If immunohistochemistry is needed because a carcinoma is in the differential diagnosis, one should be aware that carcinoid frequently expresses pan-cytokeratin and low molecular weight cytokeratin (CK8/CK18), which

1093

occasionally may be diffuse. CK7 and EMA are more specific in that they are frequently expressed in carcinoma and expressed uncommonly in carcinoid (Zhao et al. 2007). In carcinoids positive for CK7 and EMA, expression tends to be of limited extent. Also, ER and PR are usually negative in carcinoid and positive in endometrioid carcinomas and can be added to an immunohistochemical panel (Zhao et al. 2007). Primary insular carcinoid of the ovary must be differentiated from metastatic insular carcinoid in the ovary, which is usually of gastrointestinal origin. Metastatic carcinoid nearly always affects both ovaries (Robboy et al. 1974), unlike primary ovarian carcinoid, which is unilateral (Robboy et al. 1975). Macroscopically, the metastatic carcinoid is composed of tumor nodules, whereas primary ovarian carcinoid forms a single homogeneous mass. The presence of other teratomatous elements associated with an ovarian carcinoid confirms that it is primary (Robboy et al. 1975; Soga et al. 2000). Immunohistochemical studies are generally not helpful for this distinction, as metastatic midgut carcinoids and insular ovarian carcinoids demonstrate similar staining profiles, including CDX2 positivity (Rabban et al. 2009). Primary ovarian carcinoid sometimes may be confused with Brenner tumor, but the appearances of the cell nests and the grooved coffee bean nuclei of the cells of Brenner tumor militate against the diagnosis of a carcinoid, whereas the typical small acinar pattern and expression of chromogranin A favor diagnosis as a carcinoid. Confusion with granulosa cell tumor may also arise because Call–Exner bodies may be mistaken for carcinoid acini, but the cells of the carcinoid tumor usually show an acinar pattern and contain more cytoplasm and argentaffin granules (Robboy et al. 1975). Cystic areas that may be present in a granulosa cell tumor are nearly always absent in a carcinoid. The presence of nuclear grooves typical of adult granulosa cell tumor further supports this diagnosis. Inhibin and calretinin are typically negative in carcinoid (Zhao et al. 2007). Occasionally, ovarian carcinoid may be confused with a Krukenberg tumor, but the latter is usually bilateral and larger. The cells of Krukenberg tumor tend to merge with the stroma, are larger, and show

1094

greater pleomorphism, at least focal signet ring appearance, and more brisk mitotic activity. An acinar pattern is less evident. Demonstration of chromogranin and synaptophysin expression supports a diagnosis of carcinoid.

K. P. Maniar and R. Vang

Soga et al. 2000), but in a later study of four cases of trabecular carcinoid, two tumors were pure and not associated with teratomatous elements (Talerman and Evans 1982). Clinical Features

Clinical Behavior and Treatment

Although insular carcinoid tumors of the ovary are considered to be malignant, they are slow growing and only occasionally associated with metastases. Metastases have been observed in 11 patients, 7 of whom died with metastatic disease (Davis et al. 1996; Robboy et al. 1975); this includes six patients with metastatic disease from a series of nine cases of insular carcinoid tumors of the ovary collected over 40 years by Davis et al. (1996). This series suggested that metastatic disease may be more frequent than generally reported. Although the malignant potential of insular carcinoid should not be minimized, it is considered that the series of Davis et al. (1996) was somewhat selective for metastasizing tumors and lethal outcome. In occasional patients, features of carcinoid syndrome, such as tricuspid incompetence resulting in right-sided heart failure, may progress after the excision of the tumor and lead to the death of the patient, as has been observed in two cases (Robboy et al. 1975). In most patients with the carcinoid syndrome, the symptoms and signs of the syndrome observed preoperatively disappear or regress during the postoperative period (Robboy et al. 1975). Because nearly all patients with this tumor are postmenopausal or perimenopausal, bilateral salpingooophorectomy and hysterectomy is the treatment of choice. Surgical excision of foci of extra-ovarian spread if present is indicated. There is little experience with chemotherapy. Estimation of serum serotonin and 5-HIAA in the urine may be used to monitor the progress of the disease. Trabecular Carcinoid General Features

Trabecular carcinoid includes carcinoid tumors of hindgut or foregut derivation. Primary trabecular or ribbon carcinoid usually arises in association with teratomatous elements (Robboy et al. 1977;

Trabecular carcinoid is rare. Patient age varies from 24 to 74 years, with most patients being postmenopausal (Davis et al. 1996; Robboy et al. 1977; Soga et al. 2000; Talerman and Evans 1982). Trabecular carcinoid is a slowly growing neoplasm that can reach a large size. None of the known cases have been associated with the carcinoid syndrome. In three patients whose urine was examined immediately after surgery, 5-HIAA was normal (Robboy et al. 1977). Gross Features

The appearance of trabecular carcinoid depends on whether the tumor is associated with teratomatous elements. When associated with teratoma, the appearance is similar to that of a mature cystic teratoma. When the tumor is pure, it is a solid, firm to hard, round or ovoid mass with a smooth outline and a tan to yellow cut surface. Reported cases have always been unilateral (Robboy et al. 1977; Soga et al. 2000; Talerman and Evans 1982), but occasionally have been associated with mature cystic teratoma in the opposite ovary (Robboy et al. 1977; Soga et al. 2000). In the reported cases, the tumors measured from 4 to 25 cm in greatest dimension (Robboy et al. 1977; Soga et al. 2000; Talerman and Evans 1982). Microscopic Features

The tumor is composed of long, usually wavy ribbons, cords, or parallel trabeculae surrounded by fibromatous connective tissue stroma that is usually dense (Fig. 47). The ribbons, cords, or trabeculae are composed of cells that form usually one but sometimes two cell layers. The nuclei are elongated or ovoid and contain finely dispersed chromatin. The cytoplasm is abundant and often contains orange to red–brown granules, which usually stain with argyrophil and argentaffin stains. Ultrastructurally, the neurosecretory granules are round or oval and show slight variation in

16

Germ Cell Tumors of the Ovary

1095

distinguished from a Sertoli–Leydig cell tumor showing a cord-like pattern. In contrast to a Sertoli–Leydig tumor, trabecular carcinoid lacks tubules. Immunohistochemical expression of chromogranin and synaptophysin and lack of staining for inhibin and calretinin confirm the diagnosis of carcinoid. Clinical Behavior and Treatment

Fig. 47 Primary trabecular carcinoid tumor of the ovary. The tumor is composed of long ramifying cords of tumor cells surrounded by dense fibrous stroma

size (Serratoni and Robboy 1975; Talerman and Okagaki 1985), thus differing from those seen in insular carcinoids. Immunohistochemical Features and Differential Diagnosis

See section on Insular Carcinoid for details of immunohistochemistry. Immunohistochemical staining demonstrates a much wider range of neurohormonal polypeptides than insular carcinoids; these include serotonin, pancreatic polypeptide, glucagon, enkephalin, gastrin, vasoactive intestinal polypeptide, and calcitonin (Sporrong et al. 1982). Primary trabecular carcinoid must be distinguished from metastatic trabecular carcinoid, which is usually bilateral and frequently associated with metastases elsewhere. The presence of teratomatous elements, which are found frequently in the primary lesion, helps distinguish a primary from a metastatic lesion. Immunohistochemistry is not useful in this distinction, as metastatic hindgut or foregut carcinoids and ovarian trabecular carcinoids demonstrate a similar staining pattern, including CDX2 negativity (Rabban et al. 2009). Trabecular carcinoid sometimes may exhibit an insular pattern in foci, but unless this is a major component the tumor need not be classified as a mixed carcinoid. The presence of thyroid follicles indicates that the tumor is a struma ovarii and carcinoid (strumal carcinoid), and their presence must be excluded before a diagnosis of trabecular carcinoid is made. Occasionally, trabecular carcinoid must be

The prognosis of patients with trabecular carcinoid of the ovary is favorable because these tumors are not associated with metastases (Davis et al. 1996; Robboy et al. 1977; Talerman and Evans 1982). In one case, a peritoneal implant was found 2 years after bilateral salpingooophorectomy and hysterectomy (Robboy et al. 1977). The optimal treatment is the excision of the affected adnexa, which results in a complete cure, but follow-up of the patient is advisable. Mucinous Carcinoid (Goblet Cell Carcinoid) General Features

Mucinous carcinoid is a variant of carcinoid tumor that has been encountered mainly in the vermiform appendix (Klein 1974; Subbuswamy et al. 1974; Warkel et al. 1978) and occasionally has been observed in the ovary (Alenghat et al. 1986; Baker et al. 2001; Soga et al. 2000). However, it should be noted that at least some of the tumors described as primary Krukenberg tumors of the ovary may have been examples of this entity. Also, it should be emphasized that before establishing a diagnosis of primary ovarian mucinous carcinoid, a metastatic appendiceal carcinoma with goblet cell carcinoid-like features must be excluded. A number of the latter have been reported (Hristov et al. 2007), and several more are known to the authors. It is likely that a proportion of “primary ovarian” mucinous carcinoids reported in the literature really represent metastases from an occult appendiceal primary. Clinical Features

The age of patients ranges from 14 to 74 years. Mucinous carcinoid is usually observed in pure form but may be seen in association with mature cystic teratoma. The tumor is unilateral but may

1096

be associated with metastases in the contralateral ovary (Alenghat et al. 1986; Baker et al. 2001; Soga et al. 2000). Gross Features

Macroscopically, the tumor is usually of considerable size, ranging from 4 to 30 cm, and most of the tumors have been more than 8 cm in greatest dimension. The tumor is gray–yellow, firm, and usually solid but may contain cystic areas (Alenghat et al. 1986; Baker et al. 2001; Soga et al. 2000). Similar appearances are encountered when the tumor forms part of a mature cystic teratoma. Microscopic Features

Microscopically, mucinous carcinoid is composed of numerous small glands or acini with very small lumina lined by uniform columnar or cuboidal epithelium. The cells contain small round or oval nuclei or appear as goblet cells distended with mucin (Fig. 48). Some cells may be disrupted by excessive distension with mucin, which may result in the formation of small pools of mucin within the glands or even in the obliteration of the gland with pools of mucinous material within the connective tissue. The glands are surrounded by connective tissue, which may vary from loose and edematous to dense fibrous or hyalinized. Some of the glands or acini may be larger and occasionally may be cystic; this represents the typical or

K. P. Maniar and R. Vang

classical pattern of mucinous carcinoid. In some areas, the tumor cells tend to invade the surrounding connective tissue, often assuming a signet ring appearance. The tumor cells may form large solid aggregates and show a less uniform appearance and more atypical features, with large hyperchromatic nuclei and brisk mitotic activity. In some tumors, such appearances may predominate. This second pattern resembles Krukenberg tumor and is described as atypical or Krukenberg tumorlike pattern. Sometimes mucinous carcinoid showing either the typical or atypical pattern or both merges with intestinal-like adenocarcinoma showing numerous neuroendocrine cells; this is regarded as a third pattern present in this tumor. Some mucinous carcinoids may be admixed with other types of carcinoid tumor, such as insular or trabecular, forming a mixed carcinoid tumor. Thus, primary ovarian mucinous carcinoid tumors can be divided into four histologic types. In a study of 17 mucinous carcinoid tumors, 6 were of the typical type, 4 of the atypical or Krukenberg tumorlike type, 5 were admixed with intestinal-like adenocarcinoma, and 2 were of the mixed type (Baker et al. 2001). The cytoplasm of the tumor cells may exhibit orange–red granules and may even be bright red. Argyrophil and argentaffin granules are always present and, in some tumors, may be abundant (Alenghat et al. 1986; Baker et al. 2001; Soga et al. 2000). Ultrastructurally, neurosecretory granules are present in some cells and absent in others. The tumor cells may contain both neurosecretory granules and mucinous material. The neuroendocrine nature of the tumor cells is further confirmed using immunohistochemical stains. Immunohistochemical Features and Differential Diagnosis

Fig. 48 Primary mucinous carcinoid tumor. The tumor is composed of numerous small glands and acini with imperceptible or very small lumina. Numerous goblet cells distended with mucin are present

The tumor cells react positively with chromogranin A. Using immunohistochemical techniques, some of the tumor cells have been shown to contain serotonin and gastrin, and both substances may be present within the same tumor cell. Other neurohormonal polypeptides such as pancreatic polypeptide and prolactin also have been detected in the tumor cells, but the range is narrower than that observed in trabecular carcinoids. CEA and low

16

Germ Cell Tumors of the Ovary

1097

between these two entities may be difficult, especially if the mucinous carcinoid assumes a predominantly Krukenberg-like pattern or if the Krukenberg tumor contains numerous argentaffin and argyrophil granules. The presence of these granules as well as of the neurosecretory granules observed ultrastructurally cannot be used for differentiation between these two entities. Involvement of both ovaries and the presence of primary extraovarian signet ring or mucinous adenocarcinoma are indicative of Krukenberg tumor. Clinical Behavior and Treatment Fig. 49 Primary mucinous carcinoid tumor. The tumor cells demonstrate diffuse strong membranous expression of CK20

molecular weight cytokeratin also can be demonstrated within the cytoplasm of the tumor cells. These tumors are expected to show a CK7( )/ CK20 diffuse profile (Vang et al. 2007) (Fig. 49). Primary mucinous carcinoid tumor of the ovary must be differentiated from metastasis, including secondary involvement by an appendiceal carcinoma with goblet cell carcinoid-like features (Hristov et al. 2007; Soga et al. 2000). Metastatic mucinous carcinoid, in common with other types of carcinoid metastatic to the ovary, is nearly always bilateral and instead of forming a single tumor mass shows multiple scattered tumor deposits within ovarian tissue. Depending on their size, these deposits may form tumor nodules observed macroscopically or may be detectable only microscopically. Histologically, they may have appearances indistinguishable from a primary tumor. The presence of teratomatous elements supports primary ovarian origin. Mucinous carcinoid must be distinguished from other mucinous tumors of the ovary, particularly, a mucinous carcinoma with goblet cell carcinoid-like features arising in a teratoma. That is especially so when the carcinoid tumor is composed of large acini, shows increased mucin production, and exhibits a pleomorphic pattern. Occasionally, confusion may arise with welldifferentiated endometrioid tumors of the ovary, which may resemble mucinous tumors. Mucinous carcinoid must also be distinguished from a Krukenberg tumor. The differentiation

Primary mucinous carcinoid of the ovary behaves in a somewhat more aggressive manner than other types of primary ovarian carcinoid tumors (Alenghat et al. 1986; Baker et al. 2001; Soga et al. 2000), similar to the behavior of mucinous carcinoid tumors of the vermiform appendix (Klein 1974; Subbuswamy et al. 1974; Warkel et al. 1978). The tumor tends to spread mainly via the lymphatics, and metastases may be present at the time of initial laparotomy. Patients who do not exhibit metastatic disease at the time of diagnosis have a much better prognosis compared to those who have metastases, however small, at the time of diagnosis (Alenghat et al. 1986; Baker et al. 2001; Soga et al. 2000). A series of 17 patients has been reported (Baker et al. 2001). Six patients with mucinous carcinoid of the typical or classical type had tumors confined to the ovary, and all five with available follow-up (27–147 months) were well and disease-free after excision of the tumor. Three patients with mucinous carcinoid of the atypical type had tumors confined to the ovary and were well and disease-free after excision of the tumor (followup, 36–168 months). Of the eight women with carcinoma arising in mucinous carcinoid, six were stage I, one was stage II, and one was stage III. Seven had available follow-up. Two patients died of disease at 9 and 12 months, respectively. Four were alive and well with 48–120 months of follow-up, and a fifth patient died of other causes at 36 months. The treatment is surgical, depending on the extent of the disease. In postmenopausal women, patients with involvement of the contralateral ovary

1098

K. P. Maniar and R. Vang

and patients who do not want to retain fertility, hysterectomy, bilateral salpingo-oophorectomy, and omentectomy, as well as excision of all the tumor deposits, are indicated. Para-aortic lymph node dissection may be necessary because metastatic tumor deposits may be present. Surgery may be followed by combination chemotherapy, including 5-fluorouracil, although the efficacy of this mode of therapy is not proven. Premenopausal patients with tumors localized to the ovary may be treated by unilateral salpingooophorectomy with careful follow-up. Strumal Carcinoid General Features

Strumal carcinoid is an uncommon ovarian tumor composed of thyroid tissue intimately admixed with carcinoid tumor, showing a ribbonlike or cord-like pattern. Other teratomatous elements are also present in most of the tumors (Robboy and Scully 1980). Tumors showing the histologic pattern of what is now regarded as strumal carcinoid were historically interpreted as carcinoma arising in struma ovarii, although the resemblance to a carcinoid tumor had been noted in some cases. Clinical Features

More than 60 cases have been reported (Robboy and Scully 1980; Snyder and Tavassoli 1986; Soga et al. 2000), and there are probably as many unreported cases. The age distribution is similar to struma ovarii, ranging from 21 to 77 years. The tumor is usually not associated with any specific clinical findings. In one reported case, it was associated with virilization. Like hindgut carcinoids and unlike the primary ovarian insular carcinoid, strumal carcinoid is not, as a rule, associated with the carcinoid syndrome (Robboy and Scully 1980; Snyder and Tavassoli 1986; Soga et al. 2000) although this association has been described in a single case (Ulbright 2005). Gross Features

Macroscopically, this tumor, if pure, may be similar to struma ovarii or carcinoid. If the tumor is a part of a teratoma, it manifests as a yellow nodule within the teratoma (Robboy and Scully 1980; Snyder and Tavassoli 1986).

Fig. 50 Strumal carcinoid. The carcinoid forms long narrow cords and ribbons (right) merging with thyroid follicles (left)

Microscopic and Immunohistochemical Features

Microscopically, strumal carcinoid is composed of thyroid follicles containing colloid that merge with ribbons of neoplastic cells usually set in dense fibrous tissue stroma similar to trabecular carcinoid (Fig. 50). The thyroid follicles are often small at the junction between the two types of tissue. The carcinoid is usually composed of long, winding, or straight ribbons of columnar cells with elongated hyperchromatic nuclei. It may also be composed of small islands of tumor cells surrounded by dense fibrous tissue stroma. Low mitotic activity is present in the carcinoid part of the lesion. Argyrophil and argentaffin granules are identified in the carcinoid cells (Robboy and Scully 1980; Snyder and Tavassoli 1986; Stagno et al. 1987), as well as in some cells lining the thyroid follicles, both histochemically and immunohistochemically. Chromogranin and synaptophysin are expressed in the carcinoid component. TTF-1 and CK7 are usually expressed in the strumal component with no expression in the carcinoid component (Rabban et al. 2009); however, we have seen cases with focal TTF-1 expression in the carcinoid component. Ultrastructural examination demonstrates neurosecretory granules in the carcinoid component and in some of the thyroid follicular cells (Snyder and Tavassoli 1986; Stagno et al. 1987). In two tumors, amyloid deposits were identified and were verified

16

Germ Cell Tumors of the Ovary

both immunohistochemically and ultrastructurally (Arhelger and Kelly 1974; Dayal et al. 1979). Some investigators consider that strumal carcinoid is a carcinoid tumor and that the thyroid tissue only resembles thyroid (Hart and Regezi 1978). Other investigators have conclusively demonstrated thyroglobulin and TTF-1 within the thyroid component of the tumor, thus indicating its thyroid nature (Rabban et al. 2009; Robboy and Scully 1980; Snyder and Tavassoli 1986). It is, therefore, considered that in verified cases of strumal carcinoid, the tumor consists of thyroid tissue intimately admixed with a carcinoid. Strumal carcinoid should be distinguished from carcinoma of the thyroid arising in struma ovarii, with which it has often been confused. The carcinoma has the typical appearances observed in the eutopic thyroid and usually exhibits a follicular or papillary pattern. Clinical Behavior and Treatment

Strumal carcinoid has been associated with metastases in only one reported case, and even in this case the patient was apparently cured by a combination of surgery and radiation therapy (Robboy and Scully 1980). All other cases have followed a benign course (Robboy and Scully 1980; Snyder and Tavassoli 1986).

Monodermal Teratomas with Neuroectodermal Differentiation Cysts lined entirely by mature glial (Ulirsch and Goldman 1982) or ependymal tissue (Tiltman 1985) have been described in the ovary. More importantly, however, neuroectodermal tumors may also develop in the ovary and are considered monodermal teratomas with neuroectodermal differentiation (Chiang et al. 2017; Liang et al. 2016). Kleinman et al. have reported a series of 25 cases of primary neuroectodermal tumors of the ovary (Kleinman et al. 1993). The average age was 23 years (range 6–69 years). The tumors were cystic and/or solid with an average size of 14 cm (range 4–20 cm). They consisted of three histologic types: differentiated (ependymoma (Fig. 51)), primitive (medulloepithelioma, ependymoblastoma, neuroblastoma, and medulloblastoma), and anaplastic (glioblastoma). Some were associated with a mature cystic teratoma. Occasional patients may present with advanced-stage disease. It is noteworthy that

1099

Fig. 51 Monodermal teratoma with ependymoma. Perivascular pseudorosettes are present

ependymomas may contain papillary areas mimicking serous tumors or gland-like structures resembling endometrioid tumors. Survival was dependent on stage; however, the differentiated group showed superior outcome compared with the other two categories of tumors, and deaths with stage I tumors were seen in the anaplastic group. A recent study of central nervous system (CNS)type tumors and tumorlike proliferations arising in the gynecologic tract and pelvis characterized a wide spectrum of neuroepithelial tumors in the ovary, often associated with teratoma, with disease recurrences observed for embryonal type tumors (Murdock et al. 2018). Diagnostic evaluation employing morphologic criteria, immunohistochemical markers, and ancillary molecular analysis used for primary CNS tumors is recommended for these tumors, with classification per the WHO system for primary CNS tumors advocated.

Monodermal Teratoma Composed of Vascular Tissue Another type of monodermal teratoma is represented by neoplasms composed entirely or predominantly of immature vascular tissue. These occur in children and young adults, and the patients present with symptoms and signs suggestive of an ovarian tumor. The tumors may vary in size, are smooth, soft, solid, and gray–pink, and may be hemorrhagic. Microscopically, they consist of collections of small vascular spaces lined by immature endothelial cells and surrounded by connective tissue, which varies from loose and

1100

edematous to dense and fibrous. The lining of the vascular spaces may be multilayered, and the endothelial cells may form small projections bulging into the lumen. Small collections of endothelial cells, some forming abortive lumina and some devoid of a lumen, are also seen within the connective tissue and may predominate. The endothelial cells show a considerable degree of cellular and nuclear pleomorphism, and mitotic activity is usually evident. Occasionally, hematopoietic activity may be seen within some of the vascular spaces. When these tumors contain small teratomatous foci, their nature is more readily apparent, but when they occur in pure form, especially when the endothelial cells form a more solid pattern with fewer obvious vascular spaces, the nature of the lesion is more difficult to recognize. Occasionally, these tumors may be composed of immature pericytes and resemble a hemangiopericytoma. Further sampling of the tumor, which may reveal a more typical vascular pattern, and immunohistochemical stains for CD31 and CD34 as well as for factor VIII may be helpful in reaching the correct diagnosis. This distinction is important because monodermal teratomas composed of immature vascular tissue or with a predominant vascular component behave on the whole in a less aggressive manner compared with high-grade immature teratomas and angiosarcomas of the ovary, with which they tend to be confused. As in most immature teratomas, the grade of the tumor is an important prognostic feature. Baker et al. have reported a series of teratomas containing prominent benign vascular proliferations associated with neural tissue (Baker et al. 2002). The underlying tumors were either mature cystic teratoma, immature teratoma, or a mixed germ cell tumor. The vascular proliferations consisted of long, thin-walled, and curved vessels or rounded cellular aggregates with glomeruloid patterns. Small vessels could be seen arranged in whorls, and focally, a trabecular pattern was present.

Monodermal Teratoma with Sebaceous Differentiation Sebaceous tumors resulting from one-sided development of a teratoma or arising in mature cystic teratomas with other components are rare (Chumas and Scully 1991). The reported cases

K. P. Maniar and R. Vang

showed an age range from 31 to 79 years at presentation, but the majority of patients were older than 49 years. All presented with lower abdominal enlargement. The ovarian tumors found at laparotomy were large, ranging from 10 to 35 cm. Some of the patients were found to have a mature cystic teratoma in the contralateral ovary. The tumors were mainly cystic. Partly solid yellow and tan masses protruded into the cysts. The latter contained necrotic or cheesy material (Chumas and Scully 1991). Microscopically, the tumors consisted of sebaceous adenoma, basal cell carcinoma with sebaceous differentiation, and sebaceous carcinoma (Chumas and Scully 1991). The adenomas were all composed of nodules or lobules of proliferating normal sebaceous cells showing various degrees of maturity, with mature cells predominating. The basal cell carcinomas with sebaceous differentiation were composed of masses or nests of malignant basal cells containing collections of mature sebaceous cells. The sebaceous carcinoma was composed of sebaceous cells showing marked cellular and nuclear pleomorphism growing in an infiltrative pattern. The tumor cells had the typical appearance of sebaceous cells. Lipid stains were strongly positive in all the tumors, confirming the diagnosis (Chumas and Scully 1991). The patients were treated either by excision of the affected adnexa or by hysterectomy and bilateral salpingo-oophorectomy (Chumas and Scully 1991). The outcome was favorable; only one tumor was known to have recurred. The tumor, a basal cell carcinoma with sebaceous differentiation, recurred in the pelvis; further follow-up was not available. One patient had, in addition to the sebaceous adenoma, a SCC arising in the same ovary and died as a result of disseminated disease 1 year after the diagnosis. All the other patients were well and disease-free for periods ranging from 1.5 to 6 years postoperatively. The patient with sebaceous carcinoma was well and diseasefree 6 years after diagnosis (Chumas and Scully 1991). Two other reported cases of sebaceous carcinoma treated with surgical resection also described good patient outcomes, with no recurrence at 19 months in one case and 32 months in the other (Moghaddam et al. 2013; Venizelos et al. 2009).

16

Germ Cell Tumors of the Ovary

Other Types of Monodermal Teratoma Some mucinous tumors of germ cell origin arising in mature cystic teratomas may overgrow the background teratomatous component and appear to be entirely composed of a mucinous tumor (see section on “Mucinous Tumors Arising in Mature Cystic Teratoma”) (Vang et al. 2007). This phenomenon is similar to the situation in which thyroid tissue in a pure struma ovarii or carcinoid has developed in a pure form or has overgrown all the other tissues. At least a subset of pure mucinous tumors of the ovary are thought to derive from teratomas and therefore represent a type of monodermal teratoma. Other rare examples of monodermal teratomatous neoplasms observed in the ovary include the epidermoid cyst, which is lined by epidermis without appendages; the melanotic tumor, resembling the retinal anlage tumor (King et al. 1985); and the possible benign cystic counterpart of the latter (Anderson and McDicken 1971). Monodermal teratomatous origin of some malignant connective tissue tumors is difficult to prove because of the occurrence of connective tissue neoplasms derived from normal ovarian tissue. Monodermal teratomatous origin of tumors derived from ectodermal or endodermal tissues is more easily acceptable, and there may be as yet undescribed tumors of this type.

Mixed Germ Cell Tumors Mixed germ cell tumors are tumors composed of more than one neoplastic germ cell element, such as dysgerminoma combined with teratoma, yolk sac tumor, choriocarcinoma, embryonal carcinoma, or polyembryoma, as well as any other possible combination of these tumor types (Fig. 52). Some tumors may contain all these neoplastic germ cell elements. A few studies indicate a greater frequency of mixed germ cell tumors (8–19% of all malignant germ cell tumors) (Gershenson et al. 1984; Kurman and Norris 1976c; Pedowitz et al. 1955) as compared with earlier reports. This finding is a result of a more detailed examination of the tumors and a better recognition of the fact that germ cell tumors may be composed of histologically different neoplastic elements occurring in combination. This

1101

Fig. 52 Malignant mixed germ cell tumor. The tumor is composed of poorly differentiated neuroepithelial elements and yolk sac tumor (right)

group includes only neoplasms composed entirely of neoplastic germ cell elements and does not include gonadoblastoma and mixed germ cell–sex cord–stromal tumor, which in addition to germ cells contains sex cord–stromal derivatives as an integral component. The relatively frequent finding of different neoplastic germ cell elements in gonadal tumors of germ cell origin is considered to be a strong argument in favor of the common histogenesis of this group of neoplasms. The various tumor elements present in these tumors may be intimately admixed or may form separate areas adjacent to each other and separated by fibrous septa. Although many ovarian tumors belonging to this group are classified according to the predominant element present, it is emphasized that all areas of varying appearance should be sampled carefully. All the neoplastic germ cell elements observed within the tumor, however small, should be reported and described and, if possible, their size or relative proportion estimated. This is important because the behavior and treatment of these neoplasms vary considerably, and the presence of a small area composed of a more malignant element may alter the therapeutic approach and prognosis. This is especially true in children (Heifetz et al. 1998) where most immature teratomas behave in a nonaggressive manner; however, the presence of small foci of yolk sac tumor in an immature teratoma is associated with aggressive behavior. Even though the presence of very small foci of yolk sac tumor or high-grade immature teratoma may not alter the behavior of a mixed germ cell

1102

tumor largely composed of less aggressive components, the presence of larger amounts of more malignant elements within a tumor is usually associated with a more aggressive behavior. Before the introduction of effective combination chemotherapy, the presence of yolk sac tumor or other more aggressive elements was associated with an unsatisfactory response to therapy and poor prognosis (Asadourian and Taylor 1969; Kurman and Norris 1976c; Talerman et al. 1973). The clinical course in most patients with tumors composed of yolk sac tumor associated with dysgerminoma or other germ cell elements usually does not differ materially from that observed in patients with pure yolk sac tumor (Gershenson 1993; Peccatori et al. 1995). The different response to treatment and the different behavior of some cases of dysgerminoma described in the past may have been a result of the presence of other germ cell elements that were not identified.

Clinical and Pathologic Features of Tumors Composed of Germ Cells and Sex Cord–Stromal Derivatives

K. P. Maniar and R. Vang

designated gonadoblastoma because it appeared to recapitulate the development of the gonads and because it occurred in individuals with abnormal sexual development and in gonads of uncertain nature (Scully 1970a). It was subsequently demonstrated that both patients were sex chromatin negative (indicative of an XY karyotype). The neoplastic nature of gonadoblastoma has been questioned because some lesions are very small and may undergo complete regression by hyalinization and calcification. Furthermore, when malignancy supervenes, it manifests itself as germ cell neoplasia despite the fact that gonadoblastoma is composed of two or three different cell types. When the tumor has metastasized, gonadoblastoma as such has never been observed in the metastases. Nevertheless, gonadoblastoma shows exactly the same pattern in the very small lesions as in the large ones, including mitotic activity in the germ cell element and early overgrowth by dysgerminoma. The association with dysgerminoma is seen in 50% of cases and with other more malignant germ cell neoplasms in an additional 10% (Scully 1970a). In view of this, the concept that gonadoblastoma represents an in situ germ cell malignancy is considered to be justified.

Gonadoblastoma General Features In 1953, Scully (1953) described two patients with a distinctive gonadal tumor, which he designated gonadoblastoma. The tumor was composed of germ cells and sex cord–stromal derivatives, resembling immature granulosa and Sertoli cells. One of the tumors also contained stromal elements indistinguishable from lutein or Leydig cells. Both tumors occurred in phenotypic females who showed abnormal sexual development. The older patient, who was postpubertal, showed virilization, and it was postulated that the tumor was capable of steroid hormone secretion. The tumors were located at the site of normal ovaries, but normal ovarian tissue was not discernible, and the exact nature of the gonads in which the tumors had originated could not be determined. Both patients had bilateral tumors that were partly overgrown by dysgerminoma. The tumor was

Genetic and Molecular Features Gonadoblastoma occurs almost entirely in patients with pure or mixed gonadal dysgenesis or in male pseudohermaphrodites. Occasional patients are of short stature and may have other stigmata of Turner syndrome (Brant et al. 2006; Schellhas 1974b; Shah et al. 1988). Nearly all patients with gonadoblastoma whose karyotype was recorded (96%) were found to have a Y chromosome (Schellhas 1974b), with the most frequent karyotype being 46,XY (found in half of cases), followed by 45,X/46,XY mosaicism (found in a quarter of cases) (Schellhas 1974b). A small subset of patients had a 46,XX karyotype (Bergher De Bacalao and Dominguez 1969; Erhan et al. 1992; Nakashima et al. 1989; Obata et al. 1995; Talerman et al. 1990; Zhao et al. 2000), some of whom were fertile (Bergher De Bacalao and Dominguez 1969; Erhan et al. 1992; Nakashima et al. 1989; Talerman et al. 1990; Zhao

16

Germ Cell Tumors of the Ovary

et al. 2000). The remainder showed many different forms of mosaicism, including 45,X/46,XX mosaicism (Schellhas 1974b; Scully 1953). Six patients with morphologic abnormalities of the Y chromosome were reported. The similarity between the distribution of the karyotypes in patients with gonadoblastoma and patients with dysgerminoma and gonadal dysgenesis is striking. In one review, 24 of 25 patients with dysgerminoma and gonadal dysgenesis had a Y chromosome. The karyotype was 46,XY in 60%, followed by 45,X/46,XY in 24%, and the remainder showed various forms of mosaicism (Schellhas 1974b). One patient had 45,X monosomy and Turner syndrome. All other patients with features of Turner syndrome had various forms of mosaicism containing a Y chromosome. Gonadal dysgenesis and dysgerminoma have also been reported in a female with a 46,XX karyotype who had no evidence of Y chromosomal DNA (Letterie and Page 1995). Gonadoblastoma was not detected in the affected gonad, but the authors suggested that it was probably overgrown by the dysgerminoma. The reported cases suggest that gonadoblastoma and gonadal dysgenesis may infrequently occur in patients who do not have Y chromosomal DNA. However, some authors have postulated that such reports represent undetected mosaicism, with the dysgenetic gonads containing cells with at least some Y chromosome material (Ulbright and Young 2014). Family history of gonadal dysgenesis has been noted in at least ten reports of patients with gonadoblastoma (Allard et al. 1972; Anderson and Carlson 1975; Boczkowski et al. 1972; Talerman 1971). Evidence of gonadal dysgenesis affecting three generations of the family of a patient with gonadoblastoma was obtained in two instances (Allard et al. 1972; Bartlett et al. 1968). Gonadoblastoma has been reported in one pair of twins (Frasier et al. 1964) and in four pairs of siblings (Allard et al. 1972; Anderson and Carlson 1975); Boczkowski et al. 1972; Talerman 1971). All these patients had 46,XY karyotypes. It has been postulated that the mode of inheritance is either an X-linked recessive gene or an autosomal sex-linked mutant gene (Bartlett et al. 1968; Schellhas 1974a, b). The TSPY gene, located on the GBY locus of the

1103

Y chromosome, has been postulated to play a role in the pathogenesis of gonadoblastoma (Hertel et al. 2010; Lau et al. 2009).

Endocrine Features The association of gonadoblastoma with certain endocrine abnormalities was noted in one of the two cases first reported (Scully 1953). In view of the fact that gonadoblastoma occurs almost entirely in patients with gonadal dysgenesis, the defective gonadal development present in these patients should not be confused with the presence of endocrine effects that are associated with the tumor. Although the virilization produced by the tumor may regress after excision, there is no further gonadal development, and the gonadal abnormalities remain. Although the exact source of the steroid hormone production was not originally known, the interstitial cells resembling Leydig or lutein cells were considered to be the most likely source of the androgens (Scully 1953). Further observations have shown that the presence of Leydig or lutein-like cells is not always associated with the presence of virilization, although they are encountered more frequently in tumors from virilized phenotypic female patients than in those from non-virilized patients. The possibility that the tumor may secrete estrogens, as evidenced by complaints of hot flushes and other menopausal symptoms after excision of the tumor, has also been noted (Scully 1970a). Originally the evidence of hormone secretion was mainly clinical, usually evidenced by virilization occurring after puberty and manifesting itself as masculine body contour, hirsutism, and clitoromegaly. Slight elevation of the urinary 17-ketosteroid excretion was noted in some cases. The gonadotropins, when measured, were usually elevated. Subsequently, it has been shown that gonadoblastoma is capable of producing testosterone and estrogens from progesterone in vitro (Anderson and Carlson 1975; Rose et al. 1974). Evidence of testosterone secretion in vivo in patients with gonadal dysgenesis has been presented (Judd et al. 1970). Androgen and estrogen formation from progesterone in vitro has been demonstrated in a streak gonad that did not contain any Leydig and lutein cells microscopically,

1104

but which, based on the description, may have contained a small burnt-out gonadoblastoma (Mackay et al. 1974). Although in vitro testosterone formation has been ascribed to the Leydig or lutein-like cells present in gonadoblastoma (Rose et al. 1974), the demonstration of steroid production by a streak gonad that did not contain Leydig or lutein cells indicates that the nondescript stromal tissue also has the capability of steroid synthesis (Mackay et al. 1974). Despite the advances in the understanding of the hormonal aspects of gonadoblastoma and dysgenetic gonads, a number of questions remain, the most important being why some patients become virilized and others do not. Although there is an approximate relationship between virilization and the presence of Leydig or luteinlike cells in the tumor, this relationship is not constant. It may be that the amount of steroid secretion is inadequate to produce virilization in some cases because of a small cell mass. Another possible explanation is that the steroid metabolic pathways may be different and that some gonadoblastomas may produce hormones that are metabolically nonfunctioning, whereas other gonadoblastomas produce metabolically active steroids.

Clinical Features The exact prevalence of gonadoblastoma is not known, but it is certainly uncommon. Gonadoblastoma is reported to occur in approximately one third of patients with 45,X/46,XY mosaicism (Coyle et al. 2016; Zelaya et al. 2015) and in approximately 5% of all phenotypically female patients with disorders of sex development (Jiang et al. 2016). Gonadoblastoma is usually seen in young patients, occurring most frequently in the second decade and somewhat less frequently in the third and first decades, in that order. With a few exceptions, all the reported cases occurred in patients under 30 years of age. Patients with gonadoblastoma usually have primary amenorrhea, virilization, or developmental abnormalities of the genitalia. The discovery of gonadoblastoma is made in the course of investigating these conditions. Another mode of presentation is the presence of a gonadal tumor. The gonadoblastoma forms part of the tumor in these

K. P. Maniar and R. Vang

cases and is discovered on histologic examination. Most patients with gonadoblastoma (80%) are phenotypic females, and the remainder are phenotypic males with cryptorchidism, hypospadias, and female internal secondary sex organs. Among the phenotypic females, 60% are virilized and the remainder are normal in appearance (Scully 1970a). Most of the phenotypic female patients exhibit abnormal genital development, and breast development is often diminished even among the non-virilized females. Although primary amenorrhea is a common presenting symptom among phenotypic females with gonadoblastoma, a few patients have episodes of spontaneous cyclical bleeding, and occasional patients menstruate normally (Scully 1970a). The virilization present in phenotypic female patients with gonadoblastoma usually does not regress after excision of the tumor, although this has been seen in occasional cases, and in a few additional cases there was partial regression. Although most patients have gonadal dysgenesis, gonadoblastoma has been described in patients who have had normal pregnancies. These include patients with a 46,XX karyotype (Bergher De Bacalao and Dominguez 1969; Erhan et al. 1992; Nakashima et al. 1989; Zhao et al. 2000) and true hermaphrodites (Talerman et al. 1990). At least eight true hermaphrodites with gonadoblastoma have been described. Of these, two had a 46,XX karyotype (McDonough et al. 1976; Talerman et al. 1981), four had 46,XY (Damjanov and Klauber 1980; Park et al. 1972; Quigley et al. 1981; Szokol et al. 1977), and two had 46,XX/46,XY mosaicism (Radharrishnan et al. 1978; Talerman et al. 1990). Gonadoblastoma has also been diagnosed in five males with normally descended testes (Hughesdon and Kumarasamy 1970; Talerman and Dlemarre 1975), some of whom fathered children subsequent to the excision of the testis bearing the lesion.

Gross Features Gonadoblastoma has been found more often in the right gonad than in the left and has been bilateral in 40% of cases or higher (Scully 1970a; Talerman and Roth 2007). Although some tumors are

16

Germ Cell Tumors of the Ovary

recognized on gross examination, in a number of cases the lesion is detected only on histologic examination; this may be the case with bilateral tumors, only one of which may be recognized macroscopically. In most cases, the gonad of origin is indeterminate because it is overgrown by the tumor. When the nature of the gonad can be identified, it is usually a streak or a testis. The contralateral gonad in these cases may be a streak or a testis, and the former is more likely to harbor a gonadoblastoma (Scully 1970a; Talerman and Roth 2007). Occasionally gonadoblastoma has been found in otherwise normal ovaries (Nakashima et al. 1989; Nomura et al. 1999; Pratt-Thomas and Cooper 1976). Pure gonadoblastoma varies from a microscopic lesion to 8 cm, with most tumors measuring a few centimeters. When gonadoblastoma becomes overgrown by dysgerminoma or other malignant germ cell elements, much larger tumors may be observed (Scully 1970a; Talerman 1974; Talerman and Roth 2007). The macroscopic appearance of the tumor varies to some extent according to the presence of hyalinization and calcification, as well as overgrowth by dysgerminoma. Gonadoblastoma is a solid tumor with a smooth or slightly lobulated surface. It varies from soft and fleshy to firm and hard. It is speckled with calcific granules and may be almost completely calcified. Calcification has been recognized on gross examination in 45% of cases, and in more than 20%, it has been detected radiologically (Scully 1970a). The tumor varies from gray or yellow to brown, and on cross section it appears to be somewhat granular (Fig. 53). Although the external sex organs in patients with gonadoblastoma present a wide variety of appearances ranging from normal to completely ambiguous, the secondary internal sex organs almost always include a uterus, which is hypoplastic in most cases, and two or occasionally one normal fallopian tube; this is also seen in the phenotypic males. Male secondary internal sex organs, such as the epididymis, vas deferens, and prostate, are found occasionally in the virilized phenotypic females and are always found in the phenotypic male pseudohermaphrodites (Scully 1970a).

1105

Fig. 53 Bilateral gonadoblastomas with dysgerminoma. The outer surface is smooth, and the cut surface is solid, granular, and yellow–brown. The white nodule in the lower pole of the tumor on the right is a dysgerminoma. (Published with kind permission of the late Dr. Robert E. Scully, Boston, MA)

Microscopic and Immunohistochemical Features Gonadoblastoma is composed of collections of cellular nests surrounded by connective tissue stroma (Figs. 54, 55, and 56). The nests are solid, usually small, and oval or round, but occasionally may be larger and elongated. The cellular nests contain a mixture of germ cells and sex cord derivatives resembling immature Sertoli and granulosa cells (Fig. 55). The germ cells are large and round, with pale or slightly granular cytoplasm and large round vesicular nuclei, often with prominent nucleoli showing histologic and ultrastructural appearances and histochemical and immunohistochemical staining patterns (CD117[+], OCT-4[+], and SALL4[+] (Cao et al. 2009)) similar to the germ cells of dysgerminoma or seminoma. The sex cord cells show immunohistochemical expression of inhibin and SF-1 and have also been demonstrated to consistently express FOXL2 (a marker of granulosa cells) with only focal or weak expression of SOX9 (a marker of Sertoli cells) (BuellGutbrod et al. 2011; Hersmus et al. 2008; Kao et al. 2014). The germ cells show mitotic activity, which may be marked in some cases. They are intimately admixed with immature sex cord cells, which are smaller, round or oval, epithelioid cells with dark,

1106

Fig. 54 Gonadoblastoma. Cellular nests surrounded by connective tissue stroma. Note foci of calcification

Fig. 55 Gonadoblastoma. A nest composed of large germ cells is intimately admixed with smaller sex cord derivatives. Hyaline Call–Exner-like bodies also are seen

oval, or slightly elongated carrot-shaped nuclei. Mitotic activity is not seen in these cells. The immature sex cord cells are arranged within the cell nests in three typical patterns: (1) along the periphery of the nests in a coronal pattern, sometimes with a palisading arrangement, (2) surrounding individual or collections of germ cells in the same way as the follicular epithelium surrounds the ovum of the primary follicle, or (3) surrounding small round spaces containing amorphous hyaline, eosinophilic, and PAS-positive material, resembling Call–Exner bodies. The connective tissue stroma surrounding the cellular nests frequently contains collections of cells indistinguishable from Leydig cells or

K. P. Maniar and R. Vang

Fig. 56 Microscopic gonadoblastoma arising in a normal ovary. It should be noted that microscopic gonadoblastoma-like foci can also be found in a subset of normal fetuses and newborns (Safneck and deSa 1986; Scully et al. 1998)

luteinized stromal cells. There is considerable variation in the number of these cells from case to case; in some cases they are numerous, in others they are identified with difficulty, or they may be absent. Although in many cases the cells are indistinguishable from Leydig cells and may contain lipochrome granules, Reinke crystals, which are specifically diagnostic of Leydig cells, have never been identified in their cytoplasm. The Leydig or lutein-like cells are identified in 66% of cases, and they are present nearly twice as frequently in older patients as in those 15 years of age or younger (Scully 1970a). The presence of Leydig or lutein-like cells is not necessary for the diagnosis of gonadoblastoma. The connective tissue stroma surrounding the cellular nests may be scanty or abundant and may vary from dense and hyalinized to cellular, resembling ovarian stroma. These latter appearances are more common in tumors that either have arisen in or are suspected to originate in a gonadal streak (Scully 1970a). Occasionally, the stroma may be loose and edematous. The basic composition of gonadoblastoma, consisting of the two cell types present within the cellular nests and with the Leydig or luteinlike cells present in the stroma, has been confirmed by electron microscopy (Garvin et al.

16

Germ Cell Tumors of the Ovary

1976; Ishida et al. 1976; Mackay et al. 1974). Although there is agreement concerning the nature of the germ cells, the nature of the sex cord–stromal cells is in dispute. They are considered by some to be Sertoli cells or granulosa cells or their precursors (Mackay et al. 1974), whereas others consider them to be primitive sex cord–stromal cells and are unable to differentiate them further (Scully 1970a; Talerman 1980). The latter view is more widely accepted. Recent immunohistochemical studies demonstrating more consistent and extensive expression of FOXL2 compared to SOX9 suggest differentiation toward granulosa cells; however, the presence of some co-expression lends support to the idea of incompletely differentiated sex cord elements with hybrid features (Kao et al. 2014; Ulbright and Young 2014). The nature of the amorphous, hyaline, and eosinophilic material forming Call–Exner-like bodies also was a matter of dispute. It was considered to be either of basement membrane origin (Ishida et al. 1976; Mackay et al. 1974) or composed of fibrillar material formed by the stromal cells before they undergo fragmentation and cell death. The former view is supported by most investigators. The basic histologic appearance of gonadoblastoma may be altered by three processes: hyalinization, calcification, and overgrowth by dysgerminoma (Scully 1970a; Talerman 1980; Talerman and Roth 2007). Hyalinization takes place by coalescence of the hyaline Call–Exner-like bodies within the nests and of the basement membrane-like band of similar material present around the nests. The hyaline material replaces the tumor cells, and the whole nest may be replaced. Calcification is a common feature (Fig. 54) and is seen microscopically in 81% of cases; it usually begins in the Call–Exner-like bodies with formation of small calcific spherules that are frequently laminated, resembling psammoma bodies. The process continues with enlargement and fusion of the calcified bodies and calcification of the hyalinized material, resulting in formation of a calcified mass embracing the whole nest. The process may extend to the stroma, which may also undergo hyalinization and calcification. In such

1107

cases, tumor cells become very scarce or absent, and the presence of smooth, rounded, and calcified masses may be the only evidence that gonadoblastoma was present (Fig. 56). Although this finding is not considered to be diagnostic of gonadoblastoma, it has been called a burnt-out gonadoblastoma (Scully 1970a; Talerman 1980; Talerman and Roth 2007) and is a strong argument in favor of the diagnosis, indicating that a careful search should be made for more viable areas of the tumor. Gonadoblastoma is frequently overgrown by dysgerminoma, as is seen in 50% of cases (Scully 1970a; Talerman 1980; Talerman and Roth 2007). The overgrowth may vary from the presence of a small collection of malignant germ cells in the stroma outside the gonadoblastoma nests to massive overgrowth of the whole tumor, in which occasional nests of gonadoblastoma may be seen. The dysgerminoma in these cases shows the typical appearances of pure dysgerminoma or seminoma – histologically, histochemically, immunohistochemically, and ultrastructurally. It should be noted that when gonadoblastoma becomes overgrown by dysgerminoma, the germ cell component present within the gonadoblastoma nests shows marked proliferative activity and overgrows the sex cord elements. When gonadoblastoma undergoes regressive changes, they manifest first as a decrease in germ cells. Gonadoblastoma may also be associated with, and overgrown by, other more malignant germ cell neoplasms, such as immature teratoma, yolk sac tumor, embryonal carcinoma, and choriocarcinoma, as occurs in 10% of cases (Scully 1970a; Talerman 1980; Talerman and Roth 2007). A gonadoblastoma overgrown by dysgerminoma and containing a proliferation of sex cord elements resembling a Sertoli cell tumor has been reported in the dysgenetic gonad of a 19-year-old phenotypic female with 46,XY karyotype (Nomura et al. 1999). Although it has been postulated that gonadoblastoma may coexist with mixed germ cell–sex cord–stromal tumor, the two cases describing such an association (Bhathena et al. 1985; Cholafranceschi and Massi 1995) were in reality typical gonadoblastomas and not combined tumors.

1108

Dissecting Gonadoblastoma and Undifferentiated Gonadal Tissue (UGT) In 2006, Cools and colleagues introduced the concept of UGT, which they defined as an admixture of germ cells and sex cord cells arranged in cords rather than tubules or follicles and/or with an unorganized pattern in a background of fibrous stroma (Cools et al. 2006). This tissue was found adjacent to gonadoblastoma in the majority of cases studied (67%), and it was postulated to represent a precursor to gonadoblastoma (Cools et al. 2006). The proposed model of gonadoblastoma tumorigenesis begins with abnormal development of Sertoli cells due to SOX9 deficiency, possibly due to upstream mutations. In the absence of normal Sertoli cells, the germ cells fail to mature, as evidenced by the expression of fetal germ cell markers such as OCT-3/4. The immature germ cells may then transform into the neoplastic germ cells of gonadoblastoma and may eventually form organized nests with the sex cord cells, resulting in the appearance of classic gonadoblastoma (Cools et al. 2006; Ulbright 2014). Dr. Scully recognized an entity demonstrating overlapping features with UGT, which he termed “dissecting gonadoblastoma” and which has been described in detail by his colleagues in recent reviews (Kao et al. 2016; Ulbright and Young 2014). Dissecting gonadoblastoma contains the same cellular elements as classic gonadoblastoma, but displays an infiltrative or a diffuse pattern. Like UGT, dissecting gonadoblastoma is frequently found in gonads with classic gonadoblastoma, lending support to the idea that it represents a precursor lesion (Kao et al. 2016). Almost all cases described occurred in phenotypic females, with karyotypes of 46,XY, 45,XO/46,XY, or 46, XX, in order of decreasing frequency (Kao et al. 2016). The clinical presentation is similar to that of classic gonadoblastoma. Microscopically, a few different patterns are described, and these frequently coexist. The solid/expansile pattern, occurring in 68% of cases, consists of large nests of germ cells with a minor component of sex cord cells, often with intervening fibrovascular septa. The classic round deposits of basement membrane material are less conspicuous in this pattern. The

K. P. Maniar and R. Vang

anastomosing pattern, found in 63% of cases, consists of small interconnected nests of germ cells with surrounding sex cord cells, often with conspicuous deposits of basement membrane material and with cellular background stroma. Finally, the cord-like pattern, present in 58% of cases, demonstrates irregularly distributed cords of germ cells and small nests of sex cord cells, also within a cellular stroma. Calcifications are rare in the first two patterns and absent in the cord-like pattern (Kao et al. 2016). The germ cells demonstrate a variable appearance, ranging from spermatogonialike to germinoma-like; only the latter expressed OCT-3/4. The sex cord cells are small, with dark ovoid or angulated nuclei and inconspicuous nucleoli, and stain similarly to the sex cord cells in classic gonadoblastoma, with positivity for inhibin, SF-1, and FOXL2 and only focal/weak expression of SOX9 (Kao et al. 2016). The cord-like and anastomosing patterns of dissecting gonadoblastoma are essentially identical to UGT, and the high frequency of associated gonadoblastoma supports that these lesions are precursors to classic gonadoblastoma. The solid/ expansile pattern, in contrast, has been suggested as the immediate precursor to invasive dysgerminoma, with classic gonadoblastoma representing an intermediate lesion (Kao et al. 2016).

Differential Diagnosis Gonadoblastoma, because of its distinctive histologic appearance and its cellular composition, cannot be easily confused with any of the wellrecognized gonadal neoplasms. Gonadoblastoma may be confused with the mixed germ cell–sex cord–stromal tumor (Talerman 1971; Talerman 1972a), which shares with gonadoblastoma the unique distinction of being composed of germ cells and sex cord–stromal derivatives. The mixed germ cell–sex cord–stromal tumor shows a less uniform appearance, absence of a nest-like pattern, absence of calcification and hyalinization, more pronounced proliferative activity involving also the sex cord–stromal derivatives, the tendency to occur in normal gonads, and other genetic, endocrine, and somatic differences. Another lesion resembling gonadoblastoma is the ovarian sex cord tumor with annular tubules

16

Germ Cell Tumors of the Ovary

1109

(SCTAT) (Scully 1970b), which is frequently found in patients with Peutz–Jeghers syndrome. This lesion, which is also frequently bilateral, is composed of tubules lined by Sertoli and granulosa-like cells; contains similar round, eosinophilic, and hyaline Call–Exner-like bodies; and tends to calcify in the same manner as gonadoblastoma. The basic difference from gonadoblastoma is the absence of germ cells (see ▶ Chap. 15, “Sex Cord-Stromal, Steroid Cell, and Other Ovarian Tumors with Endocrine, Paraendocrine, and Paraneoplastic Manifestations”). Sertoli cell nodules with intratubular germ cell neoplasia may also mimic gonadoblastoma; however, the sex cord cells in intratubular germ cell neoplasia demonstrate full Sertoli cell differentiation, with positivity for SOX9 and no expression of FOXL2 (Hersmus et al. 2008; Kao et al. 2014). Immunohistochemistry for these two markers can, therefore, assist in the distinction from gonadoblastoma. The solid/expansile pattern of dissecting gonadoblastoma may be mistaken for dysgerminoma, given the presence in both of large nests and sheets of germ cells separated by fibrovascular septa. However, the presence of sex cord cells indicates that the tumor is a gonadoblastoma, and immunohistochemistry for SF-1 can assist in highlighting these cells in difficult cases (Kao et al. 2016).

different when gonadoblastoma is associated with more malignant germ cell neoplasms, such as embryonal carcinoma, yolk sac tumor, choriocarcinoma, and immature teratoma. In the past, none of these patients survived longer than 18 months (Talerman 1974). Subsequently, the administration of combination chemotherapy used successfully in the treatment of malignant germ cell tumors has markedly improved the prognosis, which with adequate treatment is now favorable. Because gonadoblastoma occurs almost entirely in patients with dysgenetic gonads, which are not capable of normal function, and as the gonadoblastoma may act as a source from which malignant germ cell neoplasms may originate (Schellhas 1974b), there is general agreement that excision of the gonads is the treatment of choice (Schellhas 1974a; Scully 1970a; Talerman 1980; Talerman and Roth 2007). This consensus applies not only to a contralateral gonad that appears to be abnormal but also, in most cases, to a normalappearing gonad. Routine resection of the uterus or other müllerian duct derivatives, which may rarely give rise to malignancy, is not recommended in asymptomatic patients with intersex disorders (Hughes et al. 2006; Mouriquand et al. 2014).

Clinical Behavior and Treatment The prognosis of patients with pure gonadoblastoma is excellent, provided the tumor and the contralateral gonad, which may be harboring an undetectable gonadoblastoma, are excised. When gonadoblastoma is associated with dysgerminoma, the prognosis is still very good. Metastases tend to occur later and more infrequently than in dysgerminoma that is not associated with a gonadoblastoma. All patients with gonadoblastoma and dysgerminoma with known follow-up, including the occasional cases with metastases (Hart and Burkons 1979; Schellhas et al. 1971; Scully 1970a), are alive and well after treatment, with the exception of two patients who died of disseminated dysgerminoma (Hart and Burkons 1979; Teter 1970). The prognosis is

General Features The descriptive term mixed germ cell–sex cord–stromal tumor originally was intended to embrace all the tumors composed of these cell types, including gonadoblastoma. In view of the fact that the latter term is now so well established, the term mixed germ cell–sex cord–stromal tumor should be reserved for tumors composed of these cell types that exhibit distinctive histologic appearances differing from those of gonadoblastoma (Talerman 1972a, b). The WHO classification designates these tumors as “mixed germ cell-sex cord-stromal tumor, unclassified.”

Mixed Germ Cell–Sex Cord–Stromal Tumor

Genetic and Molecular Features Nearly all female patients with this neoplasm have had genotype and karyotype determinations and

1110

have been found to have the normal female chromosome complement of 46,XX. All the patients with this tumor showed normal somatosexual development. There is no evidence that patients with this tumor have chromosomal abnormalities or gonadal dysgenesis, except for a single patient reported to have monosomy 22 (Speleman et al. 1997). In a small number of tumors studied, a subset of neoplasms showed amplification of chromosome 12p, and no mutations of the c-kit or PDGFRA genes were identified (Michal et al. 2006).

Endocrine Features Most patients with mixed germ cell–sex cord–stromal tumor do not exhibit any clinical manifestations of endocrine abnormalities. In most cases, tests of hormonal function have not been performed preoperatively. In cases in which tests have been performed postoperatively, function has been found to be normal. In one case, the patient, an 8-year-old girl, exhibited signs of precocious pseudopuberty manifesting as mammary development and menstrual bleeding for 3 years before the discovery of a large ovarian tumor (Talerman and van der Harten 1977). There was increased urinary estrogen excretion. After excision of the ovarian tumor, the uterine bleeding ceased and the urinary estrogens became normal (Talerman and van der Harten 1977). Another similar case involved a 4-year-old girl with a 46, XX karyotype who presented with precocious puberty and elevation of estradiol, progesterone, testosterone, and androstenedione (Metwalley et al. 2012). She was found to have a mixed germ cell–sex cord–stromal tumor and yolk sac tumor and had resolution of symptoms and hormone levels following excision of the tumor. Isosexual precocious pseudopuberty has been seen in ten other patients in the first decade, including four infants less than 1 year of age, who exhibited mammary development and vaginal bleeding (Lacson et al. 1988; Metwalley et al. 2012; Michal et al. 2006; Zuntova et al. 1992). The urinary estrogens were elevated, and vaginal smears showed estrogen effect. After excision of the tumor, there was a complete return to normality. There was no evidence of virilization in any of the patients. These findings indicate that female

K. P. Maniar and R. Vang

patients with this neoplasm either do not have any associated endocrine abnormalities or, if these are present, that they are manifested as feminization. One reported patient with a mixed germ cell–sex cord–stromal tumor excised at the age of 10 years (Talerman 1972a) developed normally and commenced menstruating at age 15. She was well and disease-free 12 years after excision of the tumor.

Clinical Features These neoplasms are rare. A number of adequately documented cases have been reported (Arroyo et al. 1998; Lacson et al. 1988; Michal et al. 2006), but it is likely that some cases may not have been recognized and have been included with tumors of germ cell origin or with sex cord–stromal tumors. Tumors of this type have been observed more frequently in normal phenotypic female patients but have also been encountered in normal adult males. Most of the reported cases in females were encountered in children in the first decade. More than a dozen cases occurred in infants less than 1 year of age (Talerman 1972b, 1980). In three cases, the tumor occurred in women aged 26, 31, and 43 years, who had normal pregnancies (Talerman 1980). In the ovary, the tumor is most common in the first decade, followed by the second and third, and is uncommon thereafter. Therefore, the age distribution of patients with this neoplasm differs from that of patients with gonadoblastoma (Talerman 1980). Gross Features The tumors encountered have been relatively large, varying from 7.5 to 18 cm and weighing from 100 to 1050 g. The tumor was found to be unilateral in all except two patients, and the contralateral gonad has always been described as a normal ovary. In some cases in which excision or biopsy was performed, this was confirmed on microscopic examination. The tumor is usually round or ovoid, firm in consistency, and surrounded by a smooth, slightly glistening, gray or gray–yellow capsule. In most cases, the tumor is solid (Michal et al. 2006; Talerman 1972a, b; Talerman and Roth 2007), but in some cases it is partly cystic (Talerman and van der Harten 1977). The cut surface of the

16

Germ Cell Tumors of the Ovary

1111

tumor is uniformly gray, pink, or yellow to pale brown. Neither calcified areas nor foci of necrosis have been observed on gross examination. The fallopian tubes and the uterus have always been found to be normal. There have been no abnormalities affecting the external genitalia.

Microscopic and Immunohistochemical Features The tumor is composed of germ cells and sex cord derivatives, intimately admixed with each other. The tumor cells form four different histologic patterns (Arroyo et al. 1998; Michal et al. 2006; Talerman 1972a, b, 1980; Talerman and van der Harten 1977). One is composed of long, narrow ramifying cords or trabeculae (Figs. 57 and 58), which in places expand to form wider columns and larger round or oval cellular aggregates surrounded by connective tissue stroma. The second consists of tubular structures devoid of a lumen and surrounded by a fine connective tissue network. In some places, the tubular pattern is less obvious, and the tumor forms small clusters or larger round or oval cellular masses surrounded by connective tissue stroma. The latter varies in amount and appearance and tends to be more abundant in tumors showing mainly the cordlike or trabecular pattern, whereas the tubular variety tends to be more cellular and contains less connective tissue. The stroma may vary from loose and edematous to dense fibrous and hyalinized. The former is seen more often where the cord-like pattern is most prominent, whereas the latter surrounds the larger cellular aggregates. The third pattern consists of scattered collections of germ cells surrounded by sex cord elements that may be very abundant. The germ cells admixed with sex cord derivatives may also be scattered individually and in small groups within connective tissue stroma. Sometimes there may be a suggestion of an insular pattern with islands of various sizes surrounded by fine fibrovascular stroma coalescing and forming aggregates or occasionally being separated by large amounts of connective tissue and forming a more pronounced insular pattern. Admixture with all these patterns is often seen. The typical nest-like pattern present in gonadoblastoma is not observed. In only one case were a few small

Fig. 57 Mixed germ cell–sex cord–stromal tumor. The neoplasm is composed of large cellular aggregates and more slender cords

Fig. 58 Mixed germ cell–sex cord–stromal tumor. Note large germ cells surrounded by sex cord–stromal cells

collections of Leydig or lutein-like cells observed (Talerman 1972a), but in all the remaining cases, these cells were not identified. The fourth more recently encountered pattern (Arroyo et al. 1998; Michal et al. 2006) shows similar appearances to the SCTAT (Scully et al. 1998; Scully 1970b), but differs from the latter by the presence of germ cells within the tumor (Figs. 59 and 60). The germ cells show an appearance similar to those in the other three patterns, including mitotic activity. The sex cord elements are similar to those observed in typical sex cord tumors with annular tubules. The two cellular elements present in the tumor, the germ cells and the sex cord derivatives, are intimately admixed. The sex cord derivatives are arranged peripherally in a single file, forming long rows at the periphery of the cords or peripherally

1112

Fig. 59 Mixed germ cell–sex cord–stromal tumor with SCTAT-like pattern. Note the marked similarity to the SCTAT but differing from it by the presence of germ cells

Fig. 60 Mixed germ cell–sex cord–stromal tumor with SCTAT-like pattern. Typical cytologic features. Compare with Figs. 55 and 58

lining the tubular structures, as well as surrounding individual or groups of germ cells within the small clusters or larger aggregates. The sex cord derivatives generally tend to resemble Sertoli cells more than granulosa cells. They show variable degrees of mitotic activity. The germ cells resemble those observed in dysgerminoma and gonadoblastoma in all respects, including ultrastructural, histochemical, and immunohistochemical (CD117[+] and OCT-4[+]) features. In some cases, a number of the germ cells present in this tumor appear more mature than the germ cells observed in gonadoblastoma or dysgerminoma

K. P. Maniar and R. Vang

and tend to resemble primordial germ cells. In view of this, it is possible that they may represent a later stage in the maturation of the germ cell than that seen in gonadoblastoma or dysgerminoma. The germ cells show brisk mitotic activity. The tumor does not show hyalinization, calcification, or the regressive changes observed in gonadoblastoma and appears to be actively proliferative. There is some variation in the cellular content in some parts of the tumor; in some areas there is a preponderance of germ cells, whereas in others the sex cord derivatives predominate. However, the intimate admixture of these two cell types is seen everywhere. Most tumors show a solid pattern, although occasional small clefts lined by sex cord elements may be present. In some tumors, cystic spaces of varying sizes either lined by sex cord derivatives or flattened epithelial-like cells or devoid of lining may be observed (Talerman and van der Harten 1977; Tavassoli 1983; Tokuoka et al. 1985); they closely resemble the cystic spaces observed in some retiform Sertoli–Leydig cell tumors (Talerman 1987; Talerman and Haije 1985; Young and Scully 1983) or cystic sex cord–stromal tumors. In occasional tumors, this pattern may be pronounced and may suggest that the tumor contains epithelial cells in addition to germ cells and sex cord derivatives (Tavassoli 1983). It is considered, however, that these cells are in fact sex cord derivatives and that the tumor exhibits a retiform and/or cystic pattern similar to that seen in some pure sex cord tumors. Normal ovarian tissue, as evidenced by the presence of normal ovarian stroma and at least some primordial follicles, has been identified in all cases, including a case in which it could not be identified in the original sections available (Talerman 1972a). In a number of cases, graafian follicles also are present (Talerman 1972b; Talerman and van der Harten 1977). In other cases, tumor deposits are found very close to the surface of the ovary, obliterating primordial and graafian follicles.

Differential Diagnosis Histologically, this tumor is most likely to be confused with gonadoblastoma. In contrast to

16

Germ Cell Tumors of the Ovary

gonadoblastoma, this tumor lacks the typical nestlike pattern, has greater proliferative activity of both the germ cell and sex cord component, and lacks calcification, hyalinization, and in most cases Leydig or lutein-like cells. Macroscopically, the tumors are larger. The gonad of origin is a normal ovary, and there is no evidence of gonadal dysgenesis or any somatosexual abnormalities. The patients have a normal female 46,XX karyotype. There is no evidence of virilization, and if there are signs of abnormal endocrine activity, they manifest as feminization. Occasionally, if the germ cells are relatively scanty, the tumor may be confused with pure sex cord–stromal tumors of the ovary, but the presence of germ cells should alert the observer to the true identity of the tumor. If the sex cord derivatives are few in number, are missed, or are disregarded, the tumor may be misclassified as a germ cell tumor, but the presence of sex cord elements intimately admixed with the germ cells should indicate its true identity. The presence of prominent clefts and cystic spaces, especially when the latter contain papillary projections, may cause confusion with Sertoli–Leydig cell tumors showing a retiform pattern or even with serous papillary tumors. The presence of germ cells admixed with sex cord derivatives indicates that the tumor is a mixed germ cell–sex cord–stromal tumor, irrespective of its pattern.

Clinical Behavior and Treatment The prognosis of patients with mixed germ cell–sex cord–stromal tumor of the ovary occurring in pure form is favorable. In the great majority of known cases when the tumor was confined to the ovary and not associated with other malignant neoplastic germ cell elements, there has been no recurrence or metastases after excision of the affected adnexa. The patients are well and disease-free for periods varying from 1 to 15 years (Michal et al. 2006; Talerman 1980; Talerman and Roth 2007). Accordingly, after a unilateral salpingo-oophorectomy, careful examination of the abdominal cavity is recommended. If the contralateral ovary shows signs of abnormality, biopsy is advisable. After this procedure, the patient should have chromosome studies. If the

1113

karyotype is 46,XX and if no other abnormalities are detected, further therapy is not necessary, although careful long-term follow-up is essential. One well-documented case of metastasizing mixed germ cell–sex cord–stromal tumor occurring in a 5-year-old girl has been reported (Lacson et al. 1988). The metastases were found in the paraaortic lymph nodes and in the peritoneal cavity. The patient was well and disease-free 2 years after excision of the affected adnexa, para-aortic lymphadenectomy, excision of peritoneal metastases, and a course of cisplatin-based combination chemotherapy (Lacson et al. 1988). Another case of metastasizing mixed germ cell–sex cord–stromal tumor showing an unusual SCTATlike pattern and occurring in a 30-year-old woman has been reported (Arroyo et al. 1998). Three years after excision of a right-sided ovarian tumor, a large tumor mass was noted in the region of the uterine fundus. The mass and a number of peritoneal implants were excised together with the uterus and the left ovary. The primary and the metastatic tumors showed identical appearances. The left ovary was normal. The patient was well and disease-free 1 year after excision of the metastases and administration of combination chemotherapy. In three patients in their 20s, one in her early 30s, and one in her early 40s, the mixed germ cell–sex cord–stromal tumor was associated with dysgerminoma. There was no evidence of metastases. The patients were well and disease-free from 2 to 7 years after unilateral adnexectomy and radiation therapy (Talerman 1980). In five children, aged 4–16 years, the tumor was overgrown by other malignant germ cell elements, including choriocarcinoma and yolk sac tumor. In three of these cases, the tumor metastasized and resulted in the death of the patient. The metastases were composed of the malignant germ cell elements. Two patients treated with cisplatinbased chemotherapy were alive and well at follow-up intervals of 5 years and 8 months (Metwalley et al. 2012). When the tumor is associated with malignant germ cell elements, the patient should be treated with the appropriate combination chemotherapy used in treatment of non-dysgerminomatous malignant germ cell tumors, in addition to excision of the affected

1114

adnexa. This is of particular concern in postmenarchal women, in whom there is an increased possibility that the tumor may not present in pure form but be associated with other neoplastic germ cell elements.

References A. L. Husaini H, Soudy H, El Din Darwish A, Ahmed M, Eltigani A, A L Mubarak M, Sabaa AA, Edesa W, A L-Tweigeri T, Al-Badawi IA (2012) Pure dysgerminoma of the ovary: a single institutional experience of 65 patients. Med Oncol 29:2944–2948 Abbott TM, Hermann WJ, Scully RE (1984) Ovarian fetiform teratoma (homunculus) in a 9-year-old girl. Int J Gynecol Pathol 2:392–402 Abdulkader MM, Yousef MM, Abdelhadi MK, Amr SS, Alabsi ES, Al-Abbadi MA (2013) Microscopic dysgerminoma associated with anti-Ma2 paraneoplastic encephalitis in a patient with gonadal dysgenesis. Int J Gynecol Pathol 32:277–282 Abiko K, Mandai M, Hamanishi J, Matsumura N, Baba T, Horiuchi A, Mikami Y, Yoshioka S, Wakasa T, Shiozawa T, Konishi I (2010) Oct4 expression in immature teratoma of the ovary: relevance to histologic grade and degree of differentiation. Am J Surg Pathol 34:1842–1848 Al-Thubaiti I, Al-Hayek K, Binfalah M (2013) Anti-Maassociated encephalitis due to dysgerminoma in a woman with Swyer syndrome. Neurology 80:1439–1440 Alenghat E, Okagaki T, Talerman A (1986) Primary mucinous carcinoid tumor of the ovary. Cancer (Phila) 58:777–783 Ali L, Bambach BJ, Kozielski R, Wagner H, Lele S, Mhawech-Fauceglia P (2010) Ovarian yolk sac tumor associated with major fibrosarcoma component in a 13-year-old girl. Int J Gynecol Pathol 29:252–255 Allard S, Cadotte M, Boivin Y (1972) Dysgenesie gonadique pure familiale et gonadoblastome. Union Med Can 101:448–452 Anderson CT Jr, Carlson IH (1975) Elevated plasma testosterone and gonadal tumors in two 46XY “sisters”. Arch Pathol 99:360–363 Anderson MC, McDicken IW (1971) Melanotic cyst of the ovary. J Obstet Gynaecol Br Commonw 78:1047–1049 Arafah MA, Tulbah AM, Al-Husaini H, Al-Sabban M, Al-Shankiti H, Al-Badawi IA (2015) Extrauterine epithelioid trophoblastic tumor arising in the ovary with multiple metastases: a case report. Int J Surg Pathol 23:339–344 Arhelger RB, Kelly B (1974) Strumal carcinoid. Report of a case with electron microscopical observations. Arch Pathol 97:323–325 Arroyo JG, Harris W, Laden SA (1998) Recurrent mixed germ cell-sex cord-stromal tumor of the ovary in an adult. Int J Gynecol Pathol 17:281–283

K. P. Maniar and R. Vang Arroyo MR, Podda A, Cao D, Rodriguez MM (2009) Placental site trophoblastic tumor in the ovary of a young child with isosexual precocious puberty. Pediatr Dev Pathol 12:73–76 Asadourian LA, Taylor HB (1969) Dysgerminoma. An analysis of 105 cases. Obstet Gynecol 33:370–379 Atkin NB (1973) High chromosome numbers of seminomata and malignant teratomata of the testis: a review of data on 103 tumours. Br J Cancer 28:275–279 Atkin NB, Baker MC (1983) i(12p): specific chromosomal marker in seminoma and malignant teratoma of the testis? Cancer Genet Cytogenet 10:199–204 Atkin NB, Baker MC (1987) Abnormal chromosomes including small metacentrics in 14 ovarian cancers. Cancer Genet Cytogenet 26:355–361 Ayhan A, Bukulmez O, Genc C, Karamursel BS, Ayhan A (2000) Mature cystic teratomas of the ovary: case series from one institution over 34 years. Eur J Obstet Gynecol 88:153–157 Baergen RN, Rutgers J, Young RH (2003) Extrauterine lesions of intermediate trophoblast. Int J Gynecol Pathol 22:362–367 Bagg HJ (1936) Experimental production of teratoma testis in the fowl. Am J Cancer 26:69–84 Bai S, Wei S, Ziober A, Yao Y, Bing Z (2013) SALL4 and SF-1 are sensitive and specific markers for distinguishing granulosa cell tumors from yolk sac tumors. Int J Surg Pathol 21:121–125 Baker PM, Oliva E, Young RH, Talerman A, Scully RE (2001) Ovarian mucinous carcinoids including some with a carcinomatous component. Am J Surg Pathol 25:557–568 Baker PM, Rosai J, Young RH (2002) Ovarian teratomas with florid benign vascular proliferation: a distinctive finding associated with the neural component of teratomas that may be confused with a vascular neoplasm. Int J Gynecol Pathol 21:16–21 Bartlett DJ, Grant JK, Pugh MA, Aherne W (1968) A familial feminizing syndrome. A family showing intersex characteristics with xy chromosomes in three female members. J Obstet Gynaecol Br Commonw 75:199–210 Beck JS, Fulmer HF, Lee ST (1969) Solid malignant ovarian teratoma with “embryoid bodies” and trophoblastic differentiation. J Pathol 99:67–73 Bergher De Bacalao EB, Dominguez I (1969) Unilateral gonadoblastoma in a pregnant woman. Am J Obstet Gynecol 105:1279–1281 Bergstrand CG, Czar B (1956) Demonstration of a new protein fraction in serum from the human fetus. Scand J Lab Invest 8:174–174 Berney DM, Shamash J, Pieroni K, Oliver RT (2001) Loss of CD30 expression in metastatic embryonal carcinoma: the effects of chemotherapy? Histopathology 39:382–385 Bernstein D, Naor S, Rikover M, Menahem H (1974) Hemolytic anemia related to ovarian tumor. Obstet Gynecol 43:276–280 Bhathena D, Haning RV, Shapiro S, Hafez GR (1985) Coexistence of a gonadoblastoma and mixed germ

16

Germ Cell Tumors of the Ovary

cell-sex cord stroma tumor. Pathol Res Pract 180:203–206 Bing Z, Pasha T, Wang LP, Zhang PJ (2012) Malignant mixed mullerian tumor: an immunohistochemical study. Patholog Res Int 2012:569609 Björkholm E, Gyftodimos A, Lundell M, Silfverswärd C (1990) Dysgerminoma. The Radiumhemmet series 1927-1984. Cancer (Phila) 65:38–44 Blackwell WJ, Dockerty MB, Masson JC, Mussey RD (1946) Dermoid cysts of the ovary: clinical and pathological significance. Am J Obstet Gynecol 1:415 Boczkowski K, Teter J, Sternadel Z (1972) Sibship occurrence of XY gonadal dysgenesis with dysgerminoma. Am J Obstet Gynecol 113:952–955 Bonazzi C, Peccatori F, Colombo N, Lucchini V, Cantu MG, Mangioni C (1994) Pure ovarian immature teratoma, a unique and curable disease: 10 years’ experience of 32 prospectively treated patients. Obstet Gynecol 84:598–604 Boutross-Tadross O, Saleh R, Asa SL (2007) Follicular variant papillary thyroid carcinoma arising in struma ovarii. Endocr Pathol 18:182–186 Brand A, Alves MC, Saraiva C, Loío P, Goulão J, Malta J, Palminha JM, Martins M (2004) Fetus in fetu – diagnostic criteria and differential diagnosis – a case report and literature review. J Pediatr Surg 39:616–618 Brant WO, Rajimwale A, Lovell MA, Travers SH, Furness PD, Sorensen M, Oottamasathien S, Koyle MA (2006) Gonadoblastoma and Turner syndrome. J Urol 175:1858–1860 Breen JL, Neubecker RD (1967) Ovarian malignancy in children with special reference to the germ-cell tumors. Ann N Y Acad Sci 142:658–674 Buell-Gutbrod R, Ivanovic M, Montag A, Lengyel E, Fadare O, Gwin K (2011) FOXL2 and SOX9 distinguish the lineage of the sex cord-stromal cells in gonadoblastomas. Pediatr Dev Pathol 14:391–395 Cao D, Guo S, Allan RW, Molberg KH, Peng Y (2009) SALL4 is a novel sensitive and specific marker of ovarian primitive germ cell tumors and is particularly useful in distinguishing yolk sac tumor from clear cell carcinoma. Am J Surg Pathol 33:894–904 Carleton RL, Friedman NB, Bomze EJ (1953) Experimental teratomas of the testis. Cancer (Phila) 6:464–473 Caruso PA, Marsh MR, Minkowitz S, Karten G (1971) An intense clinicopathologic study of 305 teratomas of the ovary. Cancer (Phila) 27:343–348 Chang MC, Vargas SO, Hornick JL, Hirsch MS, Crum CP, Nucci MR (2009) Embryonic stem cell transcription factors and D2-40 (podoplanin) as diagnostic immunohistochemical markers in ovarian germ cell tumors. Int J Gynecol Pathol 28:347–355 Chen RJ, Chen KY, Chang TC, Sheu BC, Chow SN, Huang SC (2008) Prognosis and treatment of squamous cell carcinoma from a mature cystic teratoma of the ovary. J Formos Med Assoc 107:857–868 Chen E, Fletcher CD, Nucci MR (2010) Meningothelial proliferations in mature cystic teratoma of the ovary: evidence for the common presence of cranially derived tissues paralleling anterior embryonic plate

1115 development. An analysis of 25 consecutive cases. Am J Surg Pathol 34:1014–1018 Cheng L, Thomas A, Roth LM, Zheng W, Michael H, Karim FWA (2004) OCT4: a novel biomarker for dysgerminoma of the ovary. Am J Surg Pathol 28:1341–1346 Cheng L, Zhang S, Talerman A, Roth LM (2010) Morphologic, immunohistochemical, and fluorescence in situ hybridization study of ovarian embryonal carcinoma with comparison to solid variant of yolk sac tumor and immature teratoma. Hum Pathol 41:716–723 Cheng L, Roth LM, Zhang S, Wang M, Morton MJ, Zheng W, Abdul Karim FW, Montironi R, LopezBeltran A (2011) KIT gene mutation and amplification in dysgerminoma of the ovary. Cancer 117:2096–2103 Chiang AJ, La V, Peng J, Yu KJ, Teng NN (2011) Squamous cell carcinoma arising from mature cystic teratoma of the ovary. Int J Gynecol Cancer 21:466–474 Chiang S, Snuderl M, Kojiro-Sanada S, Quer Pi-Sunyer A, Daya D, Hayashi T, Bosincu L, Ogawa F, Rosenberg AE, Horn LC, Wang L, Iafrate AJ, Oliva E (2017) Primitive Neuroectodermal tumors of the female genital tract: a morphologic, Immunohistochemical, and molecular study of 19 cases. Am J Surg Pathol 41:761–772 Cholafranceschi M, Massi D (1995) Gonadoblastoma with coexistent features of mixed germ cell-sex cord stroma tumor: a case report. Tumori 81:215–218 Chumas JC, Scully RE (1991) Sebaceous tumors arising in ovarian dermoid cysts. Int J Gynecol Pathol 10:356–363 Clark RM, Lynch MP, Kolp R, Zukerberg LR, Growdon WB, Rueda BR (2014) The N-methyl-D-aspartate receptor, a precursor to N-methyl-D-aspartate receptor encephalitis, is found in the squamous tissue of ovarian teratomas. Int J Gynecol Pathol 33:598–606 Clement PB, Young RH, Scully RE (1987) Endometrioidlike variant of ovarian yolk sac tumor. A clinicopathological analysis of eight cases. Am J Surg Pathol 11:767–778 Cohen MB, Friend DS, Molnar JJ, Talerman A (1987) Gonadal endodermal sinus (yolk sac) tumor with pure intestinal differentiation: a new histologic type. Pathol Res Pract 182:609–616 Condous G, Thomas J, Okaro E, Bourne T (2003) Placental site trophoblastic tumor masquerading as an ovarian ectopic pregnancy. Ultrasound Obstet Gynecol 21:504–506 Cools M, Stoop H, Kersemaekers AM, Drop SL, Wolffenbuttel KP, Bourguignon JP, SlowikowskaHilczer J, Kula K, Faradz SM, Oosterhuis JW, Looijenga LH (2006) Gonadoblastoma arising in undifferentiated gonadal tissue within dysgenetic gonads. J Clin Endocrinol Metab 91:2404–2413 Cossu-Rocca P, Jones TD, Roth LM, Eble JN, Zheng W, Karim FW, Cheng L (2006a) Cytokeratin and CD30 expression in dysgerminoma. Hum Pathol 37:1015–1021 Cossu-Rocca P, Zhang S, Roth LM, Eble JN, Zheng W, Karim FW, Michael H, Emerson RE, Jones TD, Hattab EM, Cheng L (2006b) Chromosome 12p abnormalities

1116 in dysgerminoma of the ovary: a FISH analysis. Mod Pathol 19:611–615 Coyle D, Kutasy B, Han Suyin K, Antao B, Lynch SA, McDermott MB, O’Connell SM, Quinn F (2016) Gonadoblastoma in patients with 45,X/46,XY mosaicism: a 16-year experience. J Pediatr Urol 12:283 e1–283 e7 Dabner M, McCluggage WG, Bundell C, Carr A, Leung Y, Sharma R, Stewart CJ (2012) Ovarian teratoma associated with anti-N-methyl D-aspartate receptor encephalitis: a report of 5 cases documenting prominent intratumoral lymphoid infiltrates. Int J Gynecol Pathol 31:429–437 Dalmau J, Tuzun E, Wu HY, Masjuan J, Rossi JE, Voloschin A, Baehring JM, Shimazaki H, Koide R, King D, Mason W, Sansing LH, Dichter MA, Rosenfeld MR, Lynch DR (2007) Paraneoplastic antiN-methyl-D-aspartate receptor encephalitis associated with ovarian teratoma. Ann Neurol 61:25–36 Damjanov I, Klauber G (1980) Microscopic gonadoblastoma in dysgenetic gonad of an infant: an ultrastructural study. Urology 15:605–609 Dandia SD (1967) Rectovesical fistula following an ovarian dermoid with recurrent vesical calculus: a case report. J Urol 97:85–87 Davis KP, Hartmann LK, Keeney GL, Shapiro H (1996) Primary ovarian carcinoid tumors. Gynecol Oncol 61:259–265 Dawson MA, Talbert W, Yarbro JW (1971) Hemolytic anemia associated with an ovarian tumor. Am J Med 50:552–556 Dayal Y, Tashjian AH Jr, Wolfe HJ (1979) Immunocytochemical localization of calcitonin-producing cells in a strumal carcinoid with amyloid stroma. Cancer (Phila) 43:1331–1338 de la Motte Rouge T, Pautier P, Genestie C, Rey A, Gouy S, Leary A, Haie-Meder C, Kerbrat P, Culine S, Fizazi K, Lhomme C (2016) Prognostic significance of an early decline in serum alpha-fetoprotein during chemotherapy for ovarian yolk sac tumors. Gynecol Oncol 142:452–457 Devaney K, Snyder R, Norris HJ, Tavassoli FA (1993) Proliferative and histologically malignant struma ovarii: a clinicopathologic study of 54 cases. Int J Gynecol Pathol 12:333–343 Dietl J, Horny HP, Ruck P, Kaiserling E (1993) Dysgerminoma of the ovary. An immunohistochemical study of tumor-infiltrating lymphoreticular cells and tumor cells. Cancer (Phila) 71:2562–2568 Dixon F, Moore R (1952) Tumors of the male sex organs. Atlas of tumor pathology, sect VIII, fasc 31b, 32. Armed Forces Institute of Pathology, Washington, DC Djordjevic B, Euscher ED, Malpica A (2007) Growing teratoma syndrome of the ovary: review of literature and first report of a carcinoid tumor arising in a growing teratoma of the ovary. Am J Surg Pathol 31:1913–1918 Dos Santos L, Mok E, Iasonos A, Park K, Soslow RA, Aghajanian C, Alektiar K, Barakat RR, Abu-Rustum NR (2007) Squamous cell carcinoma arising in mature cystic teratoma of the ovary: a case series and review of the literature. Gynecol Oncol 105:321–324

K. P. Maniar and R. Vang Erez SE, Richart RM, Shettles LB (1965) Hashimoto’s disease in a benign cystic teratoma of the ovary. Am J Obstet Gynecol 92:273–274 Erhan Y, Toprak AS, Ozdemir N, Tiras B (1992) Gonadoblastoma and fertility. J Clin Pathol 45:828–829 Esheba GE, Pate LL, Longacre TA (2008) Oncofetal protein glypican-3 distinguishes yolk sac tumor from clear cell carcinoma of the ovary. Am J Surg Pathol 32:600–607 Evans R (1957) Developmental stages of embryo-like bodies in teratoma testis. J Clin Pathol 10:321 Evans KN, Taylor H, Zehnder D, Kilby MD, Bulmer JN, Shah F, Adams JS, Hewison M (2004) Increased expression of 25-hydroxyvitamin D-1alpha-hydroxylase in dysgerminomas: a novel form of humoral hypercalcemia of malignancy. Am J Pathol 165:807–813 Exman P, Takahashi TK, Gattas GF, Cantagalli VD, Anton C, Nalesso F, Diz Mdel P (2013) Primary ovary choriocarcinoma: individual DNA polymorphic analysis as a strategy to confirm diagnosis and treatment. Rare Tumors 5:89–92 Fadare O, Zhao C, Khabele D, Parkash V, Quick CM, Gwin K, Desouki MM (2015) Comparative analysis of Napsin A, alpha-methylacyl-coenzyme A racemase (AMACR, P504S), and hepatocyte nuclear factor 1 beta as diagnostic markers of ovarian clear cell carcinoma: an immunohistochemical study of 279 ovarian tumours. Pathology 47:105–111 Frasier SD, Bashore RA, Mosier HD (1964) Gonadoblastoma associated with pure gonadal dysgenesis in monozygous twins. J Pediatr 64:740–745 Friedman NB (1951) The comparative morphogenesis of extragenital and gonadal teratoid tumors. Cancer (Phila) 4:265–276 Friedman NB, Moore RA (1946) Tumors of the testis; a report on 922 cases. Mil Surg 99:573–593 Fujii S, Konishi I, Suzuki A, Okamura H, Okazaki T, Mori T (1985) Analysis of serum lactic dehydrogenase levels and its isoenzymes in ovarian dysgerminoma. Gynecol Oncol 22:65–72 Fujii K, Yamashita Y, Yamamoto T, Takahashi K, Hashimoto K, Miyata T, Kawai K, Kikkawa F, Toyokuni S, Nagasaka T (2014) Ovarian mucinous tumors arising from mature cystic teratomas – a molecular genetic approach for understanding the cellular origin. Hum Pathol 45:717–724 Futagami M, Yokoyama Y, Shigeto T, Mizunuma H (2010) A case of recurrent yolk sac tumor as spindle cell sarcoma of the abdominal wall. Eur J Gynaecol Oncol 31:690–693 Gallager HS, Lewis RP (1973) Sequential gonadoblastoma and choriocarcinoma. Obstet Gynecol 41:123–128 Garg K, Soslow RA, Rivera M, Tuttle MR, Ghossein RA (2009) Histologically bland “extremely well differentiated” thyroid carcinomas arising in struma ovarii can recur and metastasize. Int J Gynecol Pathol 28:222–230 Garvin AJ, Pratt-Thomas HR, Spector M, Spicer SS, Williamson HO (1976) Gonadoblastoma: histologic, ultrastructural, and histochemical observations in five cases. Am J Obstet Gynecol 125:459–471

16

Germ Cell Tumors of the Ovary

Gershenson DM (1993) Update on malignant ovarian germ cell tumors. Cancer (Phila) 71:1581–1590 Gershenson DM, Del Junco G, Herson J, Rutledge FN (1983) Endodermal sinus tumor of the ovary: the M. D. Anderson experience. Obstet Gynecol 61:194–202 Gershenson DM, Del Junco G, Copeland LJ, Rutledge FN (1984) Mixed germ cell tumors of the ovary. Obstet Gynecol 64:200–206 Gillman J (1948) The development of the gonads in man with consideration of the role of fetal endocrines and the histogenesis of ovarian tumors. Contrib Embryol 32:83 Gitlin D, Perricelli A, Gitlin GM (1972) Synthesis of -fetoprotein by liver, yolk sac, and gastrointestinal tract of the human conceptus. Cancer Res 32:979–982 Goffredo P, Sawka AM, Pura J, Adam MA, Roman SA, Sosa JA (2015) Malignant struma ovarii: a populationlevel analysis of a large series of 68 patients. Thyroid 25:211–215 Gonzalez-Crussi F, Roth LM (1976) The human yolk sac and yolk sac carcinoma. An ultrastructural study. Hum Pathol 7:675–691 Gonzalez-Roibon N, Katz B, Chaux A, Sharma R, Munari E, Faraj SF, Illei PB, Torbenson M, Netto GJ (2013) Immunohistochemical expression of SALL4 in hepatocellular carcinoma, a potential pitfall in the differential diagnosis of yolk sac tumors. Hum Pathol 44:1293–1299 Gordon A, Lipton D, Woodruff JD (1981) Dysgerminoma: a review of 158 cases from the Emil Novak Ovarian Tumor Registry. Obstet Gynecol 58:497–504 Gultekin SH, Rosenfeld MR, Voltz R, Eichen J, Posner JB, Dalmau J (2000) Paraneoplastic limbic encephalitis: neurological symptoms, immunological findings and tumour association in 50 patients. Brain 123(Pt 7):1481–1494 Guo YL, Zhang YL, Zhu JQ (2015) Prognostic value of serum alpha-fetoprotein in ovarian yolk sac tumors: a systematic review and meta-analysis. Mol Clin Oncol 3:125–132 Gupta D, Deavers MT, Silva EG, Malpica A (2004) Malignant melanoma involving the ovary: a clinicopathologic and immunohistochemical study of 23 cases. Am J Surg Pathol 28:771–780 Hackethal A, Brueggmann D, Bohlmann MK, Franke FE, Tinneberg HR, Munstedt K (2008) Squamous-cell carcinoma in mature cystic teratoma of the ovary: systematic review and analysis of published data. Lancet Oncol 9:1173–1180 Halabi M, Oliva E, Mazal PR, Breitenecker G, Young RH (2002) Prostatic tissue in mature cystic teratomas of the ovary: a report of four cases, including one with features of prostatic adenocarcinoma, and cytogenetic studies. Int J Gynecol Pathol 21:261–267 Hart WR, Burkons DM (1979) Germ cell neoplasms arising in gonadoblastomas. Cancer (Phila) 43:669–678 Hart WR, Regezi JA (1978) Strumal carcinoid of the ovary: ultrastructural observations and long-term follow-up study. Am J Clin Pathol 69:356–359 Heifetz SA, Cushing B, Giller R, Shuster JJ, Stolar CJ, Vinocur CD, Hawkins EP (1998) Immature teratomas in children: pathologic considerations: a report from the

1117 combined Pediatric Oncology Group/Children’s Cancer Group. Am J Surg Pathol 22:1115–1124 Hersmus R, Kalfa N, de Leeuw B, Stoop H, Oosterhuis JW, de Krijger R, Wolffenbuttel KP, Drop SL, Veitia RA, Fellous M, Jaubert F, Looijenga LH (2008) FOXL2 and SOX9 as parameters of female and male gonadal differentiation in patients with various forms of disorders of sex development (DSD). J Pathol 215:31–38 Hersmus R, Stoop H, van de Geijn GJ, Eini R, Biermann K, Oosterhuis JW, Dhooge C, Schneider DT, Meijssen IC, Dinjens WN, Dubbink HJ, Drop SL, Looijenga LH (2012) Prevalence of c-KIT mutations in gonadoblastoma and dysgerminomas of patients with disorders of sex development (DSD) and ovarian dysgerminomas. PLoS One 7:e43952 Hertel JD, Huettner PC, Dehner LP, Pfeifer JD (2010) The chromosome Y-linked testis-specific protein locus TSPY1 is characteristically present in gonadoblastoma. Hum Pathol 41:1544–1549 Hingorani V, Narula RK, Bhalla S (1963) Salmonella typhi infection in an ovarian dermoid. Obstet Gynecol 22:118–120 Hirabayashi K, Yasuda M, Osamura RY, Hirasawa T, Murakami M (2006) Ovarian nongestational choriocarcinoma mixed with various epithelial malignancies in association with endometriosis. Gynecol Oncol 102:111–117 Hirakawa T, Tsuneyoshi M, Enjoji M (1989) Squamous cell carcinoma arising in mature cystic teratoma of the ovary. Clinicopathologic and topographic analysis. Am J Surg Pathol 13:397–405 Hirata Y, Yanaihara N, Yanagida S, Fukui K, Iwadate K, Kiyokawa T, Tanaka T (2012) Molecular genetic analysis of nongestational choriocarcinoma in a postmenopausal woman: a case report and literature review. Int J Gynecol Pathol 31:364–368 Hoei-Hansen CE, Kraggerud SM, Abeler VM, Kaern J, Rajpert-De Meyts E, Lothe RA (2007) Ovarian dysgerminomas are characterised by frequent KIT mutations and abundant expression of pluripotency markers. Mol Cancer 6:12 Howitt BE, Magers MJ, Rice KR, Cole CD, Ulbright TM (2015) Many postchemotherapy sarcomatous tumors in patients with testicular germ cell tumors are sarcomatoid yolk sac tumors: a study of 33 cases. Am J Surg Pathol 39:251–259 Hristov AC, Young RH, Vang R, Yemelyanova AV, Seidman JD, Ronnett BM (2007) Ovarian metastases of appendiceal tumors with goblet cell carcinoidlike and signet ring cell patterns: a report of 30 cases. Am J Surg Pathol 31:1502–1511 Hu YJ, Ip PP, Chan KK, Tam KF, Ngan HY (2010) Ovarian clear cell carcinoma with choriocarcinomatous differentiation: report of a rare and aggressive tumor. Int J Gynecol Pathol 29:539–545 Hughes IA, Houk C, Ahmed SF, Lee PA, Lawson Wilkins Pediatric Endocrine Society/European Society for Paediatric Endocrinology Consensus G (2006) Consensus statement on management of intersex disorders. J Pediatr Urol 2:148–162

1118 Hughesdon PE, Kumarasamy T (1970) Mixed germ cell tumours (gonadoblastomas) in normal and dysgenetic gonads. Virch Arch (Pathol Anat) 349:258–280 Iczkowski KA, Butler SL, Shanks JH, Hossain D, Schall A, Meiers I, Zhou M, Torkko KC, Kim SJ, MacLennan GT (2008) Trials of new germ cell immunohistochemical stains in 93 extragonadal and metastatic germ cell tumors. Hum Pathol 39:275–281 Ishida T, Tagatz GE, Okagaki T (1976) Gonadoblastoma: ultrastructural evidence for testicular origin. Cancer (Phila) 37:1770–1781 Jacobs AJ, Newland JR, Green RK (1982) Pure choriocarcinoma of the ovary. Obstet Gynecol Surv 37:603–609 Jacobsen G, Talerman A (1989) Atlas of germ cell tumors. Munksgaard, Copenhagen Jiang JF, Xue W, Deng Y, Tian QJ, Sun AJ (2016) Gonadal malignancy in 202 female patients with disorders of sex development containing Y-chromosome material. Gynecol Endocrinol 32:338–341 Jiao LZ, Xiang Y, Feng FZ, Wan XR, Zhao J, Cui QC, Yang XY (2010) Clinical analysis of 21 cases of nongestational ovarian choriocarcinoma. Int J Gynecol Cancer 20:299–302 Judd HL, Scully RE, Atkins L, Neer RM, Kliman B (1970) Pure gonadal dysgenesis with progressive hirsutism. N Engl J Med 282:881–885 Kao CS, Idrees MT, Young RH, Ulbright TM (2012) Solid pattern yolk sac tumor: a morphologic and immunohistochemical study of 52 cases. Am J Surg Pathol 36:360–367 Kao CS, Ulbright TM, Idrees MT (2014) Gonadoblastoma: an immunohistochemical study and comparison to Sertoli cell nodule with intratubular germ cell neoplasia, with pathogenetic implications. Histopathology 65:861–867 Kao CS, Idrees MT, Young RH, Ulbright TM (2016) “Dissecting gonadoblastoma” of Scully: a morphologic variant that often mimics germinoma. Am J Surg Pathol 40:1417–1423 Karseladze AI, Kulinitch SI (1994) Peritoneal strumosis. Pathol Res Pract 190:1082–1085 discussion 1086-8 Kawahara H (1963) Struma ovarii with ascites and hydrothorax. Am J Obstet Gynecol 85:85–89 Kazancigil T, Laquer W, Ladewig P (1940) Papilloendothelioma of the ovary; report of three cases and discussion of Schiller’s “mesonephroma ovarii”. Am J Cancer 40:199 Kerr SE, Flotte AB, McFalls MJ, Vrana JA, Halling KC, Bell DA (2013) Matching maternal isodisomy in mucinous carcinomas and associated ovarian teratomas provides evidence of germ cell derivation for some mucinous ovarian tumors. Am J Surg Pathol 37:1229–1235 King ME, Micha JP, Allen SL, Mouradian JA, Chaganti RS (1985) Immature teratoma of the ovary with predominant malignant retinal anlage component. A parthenogenically derived tumor. Am J Surg Pathol 9:221–231 King ME, Hubbell MJ, Talerman A (1991) Mixed germ cell tumor of the ovary with a prominent polyembryoma component. Int J Gynecol Pathol 10:88–95 Klein H (1974) Mucinous carcinoid tumor of the vermiform appendix. Cancer (Phila) 33:770

K. P. Maniar and R. Vang Kleinman GM, Young RH, Scully RE (1993) Primary neuroectodermal tumors of the ovary. A report of 25 cases. Am J Surg Pathol 17:764–778 Kommoss F, Franklin WA, Talerman A (1989) Estrogen and progesterone receptors in endodermal sinus (yolk sac) tumor. Evaluation of immunocytochemical and biochemical methods. J Reprod Med 34:943–945 Kommoss F, Bibbo M, Talerman A (1990) Nuclear deoxyribonucleic acid content (ploidy) of endodermal sinus (yolk sac) tumor. Lab Investig 62:223–231 Kong B, Tian YJ, Zhu WW, Qin YJ (2009) A pure nongestational ovarian choriocarcinoma in a 10-year-old girl: case report and literature review. J Obstet Gynaecol Res 35:574–578 Kraggerud SM, Szymanska J, Abeler VM, Kaern J, Eknaes M, Heim S, Teixeira MR, Trope CG, Peltomaki P, Lothe RA (2000) DNA copy number changes in malignant ovarian germ cell tumors. Cancer Res 60:3025–3030 Kraggerud SM, Hoei-Hansen CE, Alagaratnam S, Skotheim RI, Abeler VM, Rajpert-De Meyts E, Lothe RA (2013) Molecular characteristics of malignant ovarian germ cell tumors and comparison with testicular counterparts: implications for pathogenesis. Endocr Rev 34:339–376 Krumerman MS, Chung A (1977) Squamous carcinoma arising in benign cystic teratoma of the ovary: a report of four cases and review of the literature. Cancer (Phila) 39:1237–1242 Kumar S, Shah JP, Bryant CS, Imudia AN, Cote ML, Ali-Fehmi R, Malone JM Jr, Morris RT (2008) The prevalence and prognostic impact of lymph node metastasis in malignant germ cell tumors of the ovary. Gynecol Oncol 110:125–132 Kurman RJ, Norris HJ (1976a) Embryonal carcinoma of the ovary: a clinicopathologic entity distinct from endodermal sinus tumor resembling embryonal carcinoma of the adult testis. Cancer (Phila) 38:2420–2433 Kurman RJ, Norris HJ (1976b) Endodermal sinus tumor of the ovary: a clinical and pathologic analysis of 71 cases. Cancer (Phila) 38:2404–2419 Kurman RJ, Norris HJ (1976c) Malignant mixed germ cell tumors of the ovary. A clinical and pathologic analysis of 30 cases. Obstet Gynecol 48:579–589 Kurman RJ, Carcangiu ML, Herrington CS, Young RH (eds) (2014) WHO classification of tumours of female reproductive organs. International Agency for Research on Cancer, Lyon Lacson AG, Gillis DA, Shawwa A (1988) Malignant mixed germ-cell-sex cord-stromal tumors of the ovary associated with isosexual precocious puberty. Cancer (Phila) 61:2122–2133 Lau YF, Li Y, Kido T (2009) Gonadoblastoma locus and the TSPY gene on the human Y chromosome. Birth Defects Res C Embryo Today 87:114–122 Letterie GS, Page DC (1995) Dysgerminoma and gonadal dysgenesis in a 46,XX female with no evidence of Y chromosomal DNA. Gynecol Oncol 57:423–425 Liang L, Zhang Y, Malpica A, Ramalingam P, Euscher ED, Fuller GN, Liu J (2015) Gliomatosis peritonei: a

16

Germ Cell Tumors of the Ovary

clinicopathologic and immunohistochemical study of 21 cases. Mod Pathol 28:1613–1620 Liang L, Olar A, Niu N, Jiang Y, Cheng W, Bian XW, Yang W, Zhang J, Yemelyanova A, Malpica A, Zhang Z, Fuller GN, Liu J (2016) Primary glial and neuronal tumors of the ovary or peritoneum: a clinicopathologic study of 11 cases. Am J Surg Pathol 40:847–856 Linder D, Power J (1970) Further evidence for postmeiotic origin of teratomas in the human female. Ann Hum Genet 34:21–30 Linder D, McCaw BK, Hecht F (1975) Parthenogenic origin of benign ovarian teratomas. N Engl J Med 292:63–66 Lu Y, Yang J, Cao D, Huang H, Wu M, You Y, Chen J, Lang J, Shen K (2014) Role of neoadjuvant chemotherapy in the management of advanced ovarian yolk sac tumor. Gynecol Oncol 134:78–83 Lynn JA, Varon HH, Kingsley WB, Martin JH (1967) Ultrastructural and biochemical studies of estrogen secretory capacity of a “nonfunctional” ovarian neoplasm (dysgerminoma). Am J Pathol 51:639–661 Mackay AM, Pettigrew N, Symington T, Neville AM (1974) Tumors of dysgenetic gonads (gonadoblastoma): ultrastructural and steroidogenic aspects. Cancer (Phila) 34:1108–1125 Maeda D, Ota S, Takazawa Y, Aburatani H, Nakagawa S, Yano T, Taketani Y, Kodama T, Fukayama M (2009) Glypican-3 expression in clear cell adenocarcinoma of the ovary. Mod Pathol 22:824–832 Magi-Galluzi C, O’Connell J, Neffen E et al (2001) Are mucinous cystadenomas of the ovary derived from germ cells? A genetic analysis (Abstract). Mod Pathol 14:140A (abstract # 818) Malkasian GD Jr, Dockerty MB, Symmonds RE (1967) Benign cystic teratomas. Obstet Gynecol 29:719–725 Malkasian GD, Symmonds RE, Dockerty MB (1965) Malignant ovarian teratomas. Report of 31 cases. Obstet Gynecol 25:810–814 Mangili G, Sigismondi C, Lorusso D, Cormio G, Scollo P, Vigano R, Gamucci T, Candiani M, Pignata S (2011) Is surgical restaging indicated in apparent stage IA pure ovarian dysgerminoma? The MITO group retrospective experience. Gynecol Oncol 121:280–284 Marin-Padilla M (1965) Origin, nature and significance of the “embryoids” of human teratomas. Virchows Arch (Pathol Anat) 340:105–121 Marrubini G (1949) Primary chorionepithelioma of the ovary; report of two cases. Acta Obstet Gynecol Scand 28:251–284 Mazur MT, Talbot WH Jr, Talerman A (1988) Endodermal sinus tumor and mucinous cystadenofibroma of the ovary. Occurrence in an 82-year-old woman. Cancer (Phila) 62:2011–2015 McCluggage WG, Bissonnette JP, Young RH (2006) Primary malignant melanoma of the ovary: a report of 9 definite or probable cases with emphasis on their morphologic diversity and mimicry of other primary and secondary ovarian neoplasms. Int J Gynecol Pathol 25:321–329 McDonough PG, Byrd JR, Tho PT, Otken L (1976) Gonadoblastoma in a true hermaphrodite with a 46, XX karyotype. Obstet Gynecol 47:355–358

1119 McKay D, Hertig AT, Adams EC, Danziger S (1953) Histochemical observations on the germ cells of human embryos. Anat Rec 117:201–219 McKenney JK, Soslow RA, Longacre TA (2008) Ovarian mature teratomas with mucinous epithelial neoplasms: morphologic heterogeneity and association with pseudomyxoma peritonei. Am J Surg Pathol 32:645–655 McNamee T, Damato S, McCluggage WG (2016) Yolk sac tumours of the female genital tract in older adults derive commonly from somatic epithelial neoplasms: somatically derived yolk sac tumours. Histopathology 69:739–751 Merard R, Ganesan R, Hirschowitz L (2015) Growing teratoma syndrome: a report of 2 cases and review of the literature. Int J Gynecol Pathol 34:465–472 Metwalley KA, Elsers DA, Farghaly HS, Abdel-Lateif H, Abdel-Kader M (2012) Precocious puberty secondary to a mixed germ cell-sex cord-stromal tumor associated with an ovarian yolk sac tumor: a case report. J Med Case Rep 6:162 Michal M, Vanecek T, Sima R, Mukensnabl P, Hes O, Kazakov DV, Matoska J, Zuntova A, Dvorak V, Talerman A (2006) Mixed germ cell sex cord–stromal tumors of the testis and ovary. Morphological, immunohistochemical, and molecular genetic study of seven cases. Virchows Arch 448:612–622 Miettinen M, Virtanen I, Talerman A (1985a) Intermediate filament proteins in human testis and testicular germcell tumors. Am J Pathol 120:402–410 Miettinen M, Wahlstrom T, Virtanen I, Talerman A, Astengo-Osuna C (1985b) Cellular differentiation in ovarian sex-cord-stromal and germ-cell tumors studied with antibodies to intermediate-filament proteins. Am J Surg Pathol 9:640–651 Miettinen M, McCue PA, Sarlomo-Rikala M, Rys J, Czapiewski P, Wazny K, Langfort R, Waloszczyk P, Biernat W, Lasota J, Wang Z (2014a) GATA3: a multispecific but potentially useful marker in surgical pathology: a systematic analysis of 2500 epithelial and nonepithelial tumors. Am J Surg Pathol 38:13–22 Miettinen M, Wang Z, McCue PA, Sarlomo-Rikala M, Rys J, Biernat W, Lasota J, Lee YS (2014b) SALL4 expression in germ cell and non-germ cell tumors: a systematic immunohistochemical study of 3215 cases. Am J Surg Pathol 38:410–420 Moghaddam Y, Lindsay R, Tolhurst J, Millan D, Siddiqui N (2013) A case of sebaceous carcinoma arising in a benign cystic teratoma of the ovary and review of the literature. Scott Med J 58:e18–e22 Mouriquand P, Caldamone A, Malone P, Frank JD, Hoebeke P (2014) The ESPU/SPU standpoint on the surgical management of Disorders of Sex Development (DSD). J Pediatr Urol 10:8–10 Mueller CW, Topkins P, Lapp WA (1950) Dysgerminoma of the ovary; an analysis of 427 cases. Am J Obstet Gynecol 60:153–159 Murdock T, Orr B, Allen S, Ibrahim J, Sharma R, Ronnett BM, Rodriguez F (2018) Central nervous system-type neuroepithelial tumors and tumor-like proliferations developing in the gynecologic tract and pelvis:

1120 clinicopathologic analysis of 23 cases. Am J Surg Pathol 42:1429–1444 Nakashima N, Nagasaka T, Fukata S, Oiwa N, Nara Y, Fukatsu T, Takeuchi J (1989) Ovarian gonadoblastoma with dysgerminoma in a woman with two normal children. Hum Pathol 20:814–816 National Comprehensive Cancer Network (2017) Ovarian cancer (Version 3.2017). National Comprehensive Cancer Network. Available: https://www.nccn.org/pro fessionals/physician_gls/pdf/ovarian.pdf. Accessed 5 Oct 2017 Nielsen SN, Scheithauer BW, Gaffey TA (1985) Gliomatosis peritonei. Cancer (Phila) 56:2499–2503 Nikiforov YE, Seethala RR, Tallini G, Baloch ZW, Basolo F, Thompson LD, Barletta JA, Wenig BM, Al Ghuzlan A, Kakudo K, Giordano TJ, Alves VA, Khanafshar E, Asa SL, El-Naggar AK, Gooding WE, Hodak SP, Lloyd RV, Maytal G, Mete O, Nikiforova MN, Nose V, Papotti M, Poller DN, Sadow PM, Tischler AS, Tuttle RM, Wall KB, LiVolsi VA, Randolph GW, Ghossein RA (2016) Nomenclature revision for encapsulated follicular variant of papillary thyroid carcinoma: a paradigm shift to reduce overtreatment of indolent tumors. JAMA Oncol 2:1023–1029 Nogales FF, Bergeron C, Carvia RE, Alvaro T, Fulwood HR (1996) Ovarian endometrioid tumors with yolk sac tumor component, an unusual form of ovarian neoplasm. Analysis of six cases. Am J Surg Pathol 20:1056–1066 Nogales FF, Preda O, Nicolae A (2012) Yolk sac tumours revisited. A review of their many faces and names. Histopathology 60:1023–1033 Nogales FF, Dulcey I, Preda O (2014) Germ cell tumors of the ovary: an update. Arch Pathol Lab Med 138:351–362 Nogales-Fernandez F, Silverberg SG, Bloustein PA, Martinez-Hernandez A, Pierce GB (1977) Yolk sac carcinoma (endodermal sinus tumor): ultrastructure and histogenesis of gonadal and extragonadal tumors in comparison with normal human yolk sac. Cancer (Phila) 39:1462–1474 Nomura K, Ohama K, Okamoto E, Fujiwara A (1983) Cytogenetic studies of multiple ovarian dermoid cysts in a single host. Acta Obstet Gynecol Jpn 35:1938–1944 Nomura K, Matsui T, Aizawa S (1999) Gonadoblastoma with proliferation resembling Sertoli cell tumor. Int J Gynecol Pathol 18:91–93 Norgaard-Pedersen B, Albrechtsen R, Teilum G (1975) Serum alpha-foetoprotein as a marker for endodermal sinus tumour (yolk sac tumour) or a vitelline component of “teratocarcinoma”. Acta Pathol Microbiol Scand A 83:573–589 Norris HJ, Jensen RD (1972) Relative frequency of ovarian neoplasms in children and adolescents. Cancer (Phila) 30:713–719 Norris HJ, Zirkin HJ, Benson WL (1976) Immature (malignant) teratoma of the ovary: a clinical and pathologic study of 58 cases. Cancer (Phila) 37:2359–2372 O’Connor DM, Norris HJ (1994) The influence of grade on the outcome of stage I ovarian immature (malignant) teratomas and the reproducibility of grading. Int J Gynecol Pathol 13:283–289

K. P. Maniar and R. Vang Obata NH, Nakashima N, Kawai M, Kikkawa F, Mamba S, Tomoda Y (1995) Gonadoblastoma with dysgerminoma in one ovary and gonadoblastoma with dysgerminoma and yolk sac tumor in the contralateral ovary in a girl with 46XX karyotype. Gynecol Oncol 58:124–128 Oliva E, Andrada E, Pezzica E, Prat J (1993) Ovarian carcinomas with choriocarcinomatous differentiation. Cancer 72:2441–2446 Ozturk E, Ugur MG, Cebesoy FB, Aydin A, Sever T, Balat O (2010) Good prognosis for primary ovarian pure nongestational choriocarcinoma using the EMA/CO regime. Eur J Gynaecol Oncol 31:123–125 Pantoja E, Noy MA, Axtmayer RW, Colon FE, Pelegrina I (1975a) Ovarian dermoids and their complications. Comprehensive historical review. Obstet Gynecol Surv 30:1–20 Pantoja E, Rodriguez-Ibanez I, Axtmayer RW, Noy MA, Pelegrina I (1975b) Complications of dermoid tumors of the ovary. Obstet Gynecol 45:89–94 Park IJ, Pyeatte JC, Jones HW Jr, Woodruff JD (1972) Gonadoblastoma in a true hermaphrodite with 46, XY genotype. Obstet Gynecol 40:466–472 Park JY, Kim DY, Kim JH, Kim YM, Kim YT, Nam JH (2008) Malignant transformation of mature cystic teratoma of the ovary: experience at a single institution. Eur J Obstet Gynecol Reprod Biol 141:173–178 Parkash V, Carcangiu ML (1995) Transformation of ovarian dysgerminoma to yolk sac tumor: evidence for a histogenetic continuum. Mod Pathol 8:881–887 Parrington JM, West LF, Povey S (1984) The origin of ovarian teratomas. J Med Genet 21:4–12 Payne D, Muss HB, Homesley HD, Jobson VW, Baird FG (1981) Autoimmune hemolytic anemia and ovarian dermoid cysts: case report and review of the literature. Cancer (Phila) 48:721–724 Peccatori F, Bonazzi C, Chiari S, Landoni F, Colombo N, Mangioni C (1995) Surgical management of malignant ovarian germ-cell tumors: 10 years’ experience of 129 patients. Obstet Gynecol 86:367–372 Pedowitz P, Felmus LB, Grayzel DM (1955) Dysgerminoma of the ovary; prognosis and treatment. Am J Obstet Gynecol 70:1284–1297 Peterson WF (1956) Solid, histologically benign teratomas of the ovary; A report of four cases and review of the literature. Am J Obstet Gynecol 72:1094–1102 Peterson WF (1957) Malignant degeneration of benign cystic teratomas of the overy; a collective review of the literature. Obstet Gynecol Surv 12:793–830 Peterson WF, Prevost EC, Edmunds FT, Hundley JM Jr, Morris FK (1955) Benign cystic teratomas of the ovary; A clinico-statistical study of 1,007 cases with a review of the literature. Am J Obstet Gynecol 70:368–382 Peyron A (1939) Faits nouveaux relatifs l’origine et l’histogenese des embryomes. Bull Assoc Fr Cancer 28:658 Pierce GB, Dixon FJ Jr (1959a) Testicular teratomas. I. Demonstration of teratogenesis by metamorphosis of multipotential cells. Cancer (Phila) 12:573–583 Pierce GB, Dixon FJ Jr (1959b) Testicular teratomas. II. Teratocarcinoma as an ascitic tumor. Cancer (Phila) 12:584–589

16

Germ Cell Tumors of the Ovary

Pierce GB Jr, Midgley AR Jr (1963) The origin and function of human syncytiotrophoblastic giant cells. Am J Pathol 43:153–173 Pierce GB Jr, Verney EL (1961) An in vitro and in vivo study of differentiation in teratocarcinomas. Cancer (Phila) 14:1017–1029 Pierce GB Jr, Midgley AR Jr, Ram JS, Feldman JD (1962) Pariental yolk sac carcinoma: clue to the histogenesis of Riechert’s membrane of the mouse embryo. Am J Pathol 41:549–566 Pierce GB Jr, Midgley AR Jr, Beals TF (1964) An ultrastructural study of differentiation and maturation of trophoblast of the monkey. Lab Investig 13:451–464 Pitman MB, Triratanachat S, Young RH, Oliva E (2004) Hepatocyte paraffin 1 antibody does not distinguish primary ovarian tumors with hepatoid differentiation from metastatic hepatocellular carcinoma. Int J Gynecol Pathol 23:58–64 Poulos C, Cheng L, Zhang S, Gersell DJ, Ulbright TM (2006) Analysis of ovarian teratomas for isochromosome 12p: evidence supporting a dual histogenetic pathway for teratomatous elements. Mod Pathol 19:766–771 Prat J, Bhan AK, Dickersin GR, Robboy SJ, Scully RE (1982) Hepatoid yolk sac tumor of the ovary (endodermal sinus tumor with hepatoid differentiation): a light microscopic, ultrastructural and immunohistochemical study of seven cases. Cancer (Phila) 50:2355–2368 Prat J, Cao D, Carinelli S, Nogales FF, Vang R, Zaloudek CJ (2014) Monodermal teratomas and somatic-type tumours arising from a dermoid cyst. In: Kurman RJ, Carcangiu M, Herrington CS, Young RH (eds) WHO classification of tumours of female reproductive organs, 4th edn. IARC Press, Lyons, pp 63–66 Pratt-Thomas HR, Cooper JM (1976) Gonadoblastoma with tubal pregnancy. Am J Clin Pathol 65:121–125 Quigley MM, Vaughn TC, Hammond CB, Haney AF (1981) Production of testosterone and estrogen in vitro by gonadal tissue from a 46,XY true hermaphrodite with gonadal failure and gonadoblastoma. Obstet Gynecol 58:253–259 Rabban JT, Lerwill MF, McCluggage WG, Grenert JP, Zaloudek CJ (2009) Primary ovarian carcinoid tumors may express CDX-2: a potential pitfall in distinction from metastatic intestinal carcinoid tumors involving the ovary. Int J Gynecol Pathol 28:41–48 Radharrishnan S, Sivaraman L, Natarajan PS (1978) True hermaphrodite with multiple gonadal neoplasms: report of a case with cytogenetic study. Cancer 42:2726–2732 Ramalingam P, Malpica A, Silva EG, Gershenson DM, Liu JL, Deavers MT (2004) The use of cytokeratin 7 and EMA in differentiating ovarian yolk sac tumors from endometrioid and clear cell carcinomas. Am J Surg Pathol 28:1499–1505 Rashad MN, Fathalla MF, Kerr MG (1966) Sex chromatin and chromosome analysis in ovarian teratomas. Am J Obstet Gynecol 96:461–465 Riopel MA, Spellerberg A, Griffin CA, Perlman EJ (1998) Genetic analysis of ovarian germ cell tumors by comparative genomic hybridization. Cancer Res 58:3105–3110 Rittiluechai K, Wilcox R, Lisle J, Everett E, Wallace HJ 3rd, Verschraegen CF (2014) Prognosis of hepatoid

1121 yolk sac tumor in women: what’s up, Doc? Eur J Obstet Gynecol Reprod Biol 175:25–29 Robboy SJ, Scully RE (1970) Ovarian teratoma with glial implants on the peritoneum. An analysis of 12 cases. Hum Pathol 1:643–653 Robboy SJ, Scully RE (1980) Strumal carcinoid of the ovary: an analysis of 50 cases of a distinctive tumor composed of thyroid tissue and carcinoid. Cancer (Phila) 46:2019–2034 Robboy SJ, Scully RE, Norris HJ (1974) Carcinoid metastatic to the ovary. A clinocopathologic analysis of 35 cases. Cancer (Phila) 33:798–811 Robboy SJ, Norris HJ, Scully RE (1975) Insular carcinoid primary in the ovary. A clinicopathologic analysis of 48 cases. Cancer (Phila) 36:404–418 Robboy SJ, Scully RE, Norris HJ (1977) Primary trabecular carcinoid of the ovary. Obstet Gynecol 49:202–207 Robboy SJ, Shaco-Levy R, Peng RY, Snyder MJ, Donahue J, Bentley RC, Bean S, Krigman HR, Roth LM, Young RH (2009) Malignant struma ovarii: an analysis of 88 cases, including 27 with extraovarian spread. Int J Gynecol Pathol 28:405–422 Rose LI, Underwood RH, Williams GH, Pinkus GS (1974) Pure gonadal dysgenesis. Studies of in vitro androgen metabolism. Am J Med 57:957–961 Roth LM, Karseladze AI (2008) Highly differentiated follicular carcinoma arising from struma ovarii: a report of 3 cases, a review of the literature, and a reassessment of so-called peritoneal strumosis. Int J Gynecol Pathol 27:213–222 Roth LM, Talerman A (2006) Recent advances in the pathology and classification of ovarian germ cell tumors. Int J Gynecol Pathol 25:305–320 Roth LM, Talerman A (2007) The enigma of struma ovarii. Pathology 39:139–146 Roth LM, Miller AW 3rd, Talerman A (2008) Typical thyroid-type carcinoma arising in struma ovarii: a report of 4 cases and review of the literature. Int J Gynecol Pathol 27:496–506 Roth LM, Talerman A, Levy T, Sukmanov O, Czernobilsky B (2011) Ovarian yolk sac tumors in older women arising from epithelial ovarian tumors or with no detectable epithelial component. Int J Gynecol Pathol 30:442–451 Rutgers JL, Scully RE (1986) Functioning ovarian tumors with peripheral steroid cell proliferation: a report of twenty-four cases. Int J Gynecol Pathol 5:319–337 Rutgers JL, Young RH, Scully RE (1987) Ovarian yolk sac tumor arising from an endometrioid carcinoma. Hum Pathol 18:1296–1299 Safneck JR, deSa DJ (1986) Structures mimicking sex cord-stromal tumours and gonadoblastomas in the ovaries of normal infants and children. Histopathology 10:909–920 Santesson L (1947) Clinical and pathological survey of ovarian tumours treated at radiumhemmet. Acta Radiol (Stockh) 28:644–668 Schellhas HF (1974a) Malignant potential of the dysgenetic gonad. II. Obstet Gynecol 44:455–462 Schellhas HF (1974b) Malignant potential of the dysgenetic gonad. Part 1. Obstet Gynecol 44:298–309

1122 Schellhas HF, Trujillo JM, Rutledge FN, Cork A (1971) Germ cell tumors associated with XY gonadal dysgenesis. Am J Obstet Gynecol 109:1197–1204 Schiller W (1939) Mesonephroma ovarii. Am J Cancer 35:1 Schmidt J, Derr V, Heinrich MC, Crum CP, Fletcher JA, Corless CL, Nose V (2007) BRAF in papillary thyroid carcinoma of ovary (struma ovarii). Am J Surg Pathol 31:1337–1343 Schmitz E (1925) Malignant endothelioma of perithelioma type in the ovary. Am J Obstet Gynecol 9:247 Schuldt M, Rubio A, Preda O, Nogales FF (2016) GATA binding protein 3 expression is present in primitive patterns of yolk sac tumours but is not expressed by differentiated variants. Histopathology 68:613–615 Schwartz PE, Morris JM (1988) Serum lactic dehydrogenase: a tumor marker for dysgerminoma. Obstet Gynecol 72:511–515 Scully RE (1953) Gonadoblastoma; a gonadal tumor related to the dysgerminoma (seminoma) and capable of sex-hormone production. Cancer (Phila) 6:455–463 Scully RE (1970a) Gonadoblastoma. A review of 74 cases. Cancer (Phila) 25:1340–1356 Scully RE (1970b) Sex cord tumor with annular tubules a distinctive ovarian tumor of the Peutz-Jeghers syndrome. Cancer (Phila) 25:1107–1121 Scully R, Young R, Clement P (1998) Tumors of the ovary, maldeveloped gonads and broad ligament. Armed forces Institute of Pathology. Washingon, DC Serratoni FT, Robboy SJ (1975) Ultrastructure of primary and metastatic ovarian carcinoids: analysis of 11 cases. Cancer (Phila) 36:157–160 Sever M, Jones TD, Roth LM, Karim FW, Zheng W, Michael H, Hattab EM, Emerson RE, Baldridge LA, Cheng L (2005) Expression of CD117 (c-kit) receptor in dysgerminoma of the ovary: diagnostic and therapeutic implications. Mod Pathol 18:1411–1416 Shaco-Levy R, Peng RY, Snyder MJ, Osmond GW, Veras E, Bean SM, Bentley RC, Robboy SJ (2012) Malignant struma ovarii: a blinded study of 86 cases assessing which histologic features correlate with aggressive clinical behavior. Arch Pathol Lab Med 136:172–178 Shah KD, Kaffe S, Gilbert F, Dolgin S, Gertner M (1988) Unilateral microscopic gonadoblastoma in a prepubertal Turner mosaic with Y chromosome material identified by restriction fragment analysis. Am J Clin Pathol 90:622–627 Shigematsu T, Kamura T, Airnia T et al (2000) DNA polymorphism analysis of a pure nongestational choriocarcinoma of the ovary: case report. Eur J Gynecol Oncol 21:153–154 Shirai T, Itoh T, Yoshiki T, Noro T, Tomino Y (1976) Immunofluorescent demonstration of alpha- fetoprotein and other plasma proteins in yolk sac tumor. Cancer (Phila) 38:1661–1667 Shojaei H, Hong H, Redline RW (2016) High-level expression of divergent endodermal lineage markers in gonadal and extra-gonadal yolk sac tumors. Mod Pathol 29:1278–1288 Sigismondi C, Scollo P, Ferrandina G, Candiani M, Angioli R, Vigano R, Scarfone G, Mangili G (2015) Management of bilateral malignant ovarian germ cell

K. P. Maniar and R. Vang tumors: a MITO-9 retrospective study. Int J Gynecol Cancer 25:203–207 Simard LC (1957) Polyembryonic embryoma of the ovary of parthenogenetic origin. Cancer (Phila) 10:215–223 Smith FC (1946) Pathology and physiology of struma ovarii. Arch Surg 53:603–626 Snir OL, Buza N, Hui P (2016) Mucinous epithelial tumours arising from ovarian mature teratomas: a tissue genotyping study. Histopathology 69:383–392 Snir OL, DeJoseph M, Wong S, Buza N, Hui P (2017) Frequent homozygosity in both mature and immature ovarian teratomas: a shared genetic basis of tumorigenesis. Mod Pathol 30:1467–1475 Snyder RR, Tavassoli FA (1986) Ovarian strumal carcinoid: immunohistochemical, ultrastructural, and clinicopathologic observations. Int J Gynecol Pathol 5:187–201 Soga J, Osaka M, Yakuwa Y (2000) Carcinoids of the ovary: an analysis of 329 reported cases. J Exp Clin Cancer Res 19:271–280 Speleman F, Dermaut B, De Potter CR, Van Gele M, Van Roy N, De Paepe A, Laureys G (1997) Monosomy 22 in a mixed germ cell-sex cord-stromal tumor of the ovary. Genes Chromosomes Cancer 19: 192–194 Sporrong B, Falkmer S, Robboy S et al (1982) Neurohormonal peptides in ovarian carcinoids. An immunohistochemical study of 81 primary carcinoids and of intraovarian metastases from six midgut carcinoids. Cancer (Phila) 49:68–74 Stagno PA, Petras RE, Hart WR (1987) Strumal carcinoids of the ovary. An immunohistologic and ultrastructural study. Arch Pathol Lab Med 111:440–446 Stamp GW, McConnell EM (1983) Malignancy arising in cystic ovarian teratomas. A report of 24 cases. Br J Obstet Gynaecol 90:671–675 Stevens LC (1959) Embryology of testicular teratomas in strain 129 mice. J Natl Cancer Inst 23:1249–1295 Stevens LC (1960) Embryonic potency of embryoid bodies derived from a transplantable testicular teratoma of the mouse. Dev Biol 2:285–297 Stevens LC (1962) The biology of teratomas including evidence indicating their origin form primordial germ cells. Annee Biol 1:585–610 Subbuswamy SG, Gibbs NM, Ross CF, Morson BC (1974) Goblet cell carcinoid of the appendix. Cancer (Phila) 34:338–344 Szokol M, Kondrai G, Papp Z (1977) Gonadal malignancy and 46, XY karyotype in a true hermaphrodite. Obstet Gynecol 49:358–360 Szyfelbein WM, Young RH, Scully RE (1994) Cystic struma ovarii: a frequently unrecognized tumor. A report of 20 cases. Am J Surg Pathol 18:785–788 Szyfelbein WM, Young RH, Scully RE (1995) Struma ovarii simulating ovarian tumors of other types. A report of 30 cases. Am J Surg Pathol 19:21–29 Takeda A, Ishizuka T, Goto T, Goto S, Ohta M, Tomoda Y, Hoshino M (1982) Polyembryoma of ovary producing alpha-fetoprotein and HCG: immunoperoxidase and electron microscopic study. Cancer (Phila) 49: 1878–1889

16

Germ Cell Tumors of the Ovary

Talerman A (1971) Gonadoblastoma and dysgerminoma in two siblings with dysgenetic gonads. Obstet Gynecol 38:416–426 Talerman A (1972a) A distinctive gonadal neoplasm related to gonadoblastoma. Cancer (Phila) 30:1219–1224 Talerman A (1972b) A mixed germ cell-sex cord stroma tumor of the ovary in a normal female infant. Obstet Gynecol 40:473–478 Talerman A (1974) Gonadoblastoma associated with embryonal carcinoma. Obstet Gynecol 43:138–142 Talerman A (1975) The incidence of yolk sac tumor (endodermal sinus tumor) elements in germ cell tumors of the testis in adults. Cancer (Phila) 36:211–215 Talerman A (1980) The pathology of gonadal neoplasms composed of germ cells and sex cord stroma derivatives. Pathol Res Pract 170:24–38 Talerman A (1984) Carcinoid tumors of the ovary. J Cancer Res Clin Oncol 107:125–135 Talerman A (1987) Ovarian Sertoli-Leydig cell tumor (androblastoma) with retiform pattern. A clinicopathologic study. Cancer (Phila) 60:3056–3064 Talerman A, Dlemarre JF (1975) Gonadoblastoma associated with embryonal carcinoma in an anatomically normal man. J Urol 113:355–359 Talerman A, Evans MI (1982) Primary trabecular carcinoid tumor of the ovary. Cancer (Phila) 50:1403–1407 Talerman A, Haije WG (1974) Alpha-fetoprotein and germ cell tumors: a possible role of yolk sac tumor in production of alpha-fetoprotein. Cancer (Phila) 34:1722–1726 Talerman A, Haije WG (1985) Ovarian Sertoli cell tumor with retiform and heterologous elements. Am J Surg Pathol 9:459–460 Talerman A, Okagaki T (1985) Ultrastructural features of primary trabecular carcinoid tumor of the ovary. Int J Gynecol Pathol 4:153–160 Talerman A, Roth LM (2007) Recent advances in the pathology and classification of gonadal neoplasms composed of germ cells and sex cord derivatives. Int J Gynecol Pathol 26:313–321 Talerman A, van der Harten JJ (1977) Mixed germ cell-sex cord stroma tumor of the ovary associated with isosexual precocious puberty in a normal girl. Cancer (Phila) 40:889–894 Talerman A, Huyzinga WT, Kuipers T (1973) Dysgerminoma. Clinocopathologic study of 22 cases. Obstet Gynecol 41:137–147 Talerman A, Haije WG, Baggerman L (1977a) Alpha-1 antitrypsin (AAT) and alphafoetoprotein (AFP) in sera of patients with germ-cell neoplasms: value as tumour markers in patients with endodermal sinus tumour (yolk sac tumour). Int J Cancer 19:741–746 Talerman A, van der Pompe WB, Haije WG, Baggerman L, Boekestein-Tjahjadi HM (1977b) Alpha-foetoprotein and carcinoembryonic antigen in germ cell neoplasms. Br J Cancer 35:288–291 Talerman A, Haije WG, Baggerman L (1978) Serum alphafetoprotein (AFP) in diagnosis and management of endodermal sinus (yolk sac) tumor and mixed germ cell tumor of the ovary. Cancer (Phila) 41:272–278 Talerman A, Haije WG, Baggerman L (1980) Serum alphafetoprotein (AFP) in patients with germ cell

1123 tumors of the gonads and extragonadal sites: correlation between endodermal sinus (yolk sac) tumor and raised serum AFP. Cancer (Phila) 46:380–385 Talerman A, Jarabak J, Amarose AP (1981) Gonadoblastoma and dysgerminoma in a true hermaphrodite with a 46,XX karyotype. Am J Obstet Gynecol 140:475–477 Talerman A, Verp MS, Senekjian E, Gilewski T, Vogelzang N (1990) True hermaphrodite with bilateral ovotestes, bilateral gonadoblastomas and dysgerminomas, 46, XX/46,XY karyotype, and a successful pregnancy. Cancer (Phila) 66:2668–2672 Talukdar S, Kumar S, Bhatla N, Mathur S, Thulkar S, Kumar L (2014) Neo-adjuvant chemotherapy in the treatment of advanced malignant germ cell tumors of ovary. Gynecol Oncol 132:28–32 Tavassoli FA (1983) A combined germ cell-gonadal stromalepithelial tumor of the ovary. Am J Surg Pathol 7:73–84 Teilum G (1944) Homologous tumours in ovary and testis: contribution to classification of gonadal tumours. Acta Obstet Gynecol Scand 24:480 Teilum G (1946) Gonocytoma; homologous ovarian and testicular tumours. 1. With discussion of “mesonephroma ovarii” (Schiller: Am J Cancer 1939). Acta Pathol Microbiol Scand 23:242 Teilum G (1950) “Mesonephroma ovarii” (Schiller) extraembryonic mesoblastoma of germ cell origin in ovary and testis. Acta Pathol Microbiol Scand 27:249 Teilum G (1959) Endodermal sinus tumors of the ovary and testis. Comparative morphogenesis of the so-called mesoephroma ovarii (Schiller) and extraembryonic (yolk sac-allantoic) structures of the rat’s placenta. Cancer (Phila) 12:1092–1105 Teilum G (1965) Classification of endodermal sinus tumour (mesoblatoma vitellinum) and so-called “embryonal carcinoma” of the ovary. Acta Pathol Microbiol Scand 64:407–429 Teter J (1970) Prognosis, malignancy, and curability of the germ-cell tumor occurring in dysgenetic gonads. Am J Obstet Gynecol 108:894–900 Tiltman AJ (1985) Ependymal cyst of the ovary. A case report. S Afr Med J 68:424–425 Tokuoka S, Aoki Y, Hayashi Y, Yokoyama T, Ishii T (1985) A mixed germ cell-sex cord-stromal tumor of the ovary with retiform tubular structure: a case report. Int J Gynecol Pathol 4:161–170 Tsuchida Y, Kaneko M, Yokomori K, Saito S, Urano Y, Endo Y, Asaka T, Takeuchi T (1978) Alphafetoprotein, prealbumin, albumin, alpha-1-antitrypsin and transferrin as diagnostic and therapeutic markers for endodermal sinus tumors. J Pediatr Surg 13:25–29 Tsujioka H, Hamada H, Miyakawa T, Hachisuga T, Kawarabayashi T (2003) A pure nongestational choriocarcinoma of the ovary diagnosed with DNA polymorphism analysis. Gynecol Oncol 89:540–542 Tsuura Y, Hiraki H, Watanabe K, Igarashi S, Shimamura K, Fukuda T, Suzuki T, Seito T (1994) Preferential localization of c-kit product in tissue mast cells, basal cells of skin, epithelial cells of breast, small cell lung carcinoma and seminoma/dysgerminoma in human:

1124 immunohistochemical study on formalin-fixed, paraffin-embedded tissues. Virchows Arch 424:135–141 Ueda G, Fujita M, Ogawa H, Sawada M, Inoue M, Tanizawa O (1993) Adenocarcinoma in a benign cystic teratoma of the ovary: report of a case with a long survival period. Gynecol Oncol 48:259–263 Ulbright TM (2005) Germ cell tumors of the gonads: a selective review emphasizing problems in differential diagnosis, newly appreciated, and controversial issues. Mod Pathol 18(Suppl 2):S61–S79 Ulbright TM (2014) Gonadoblastoma and hepatoid and endometrioid-like yolk sac tumor: an update. Int J Gynecol Pathol 33:365–373 Ulbright TM, Young RH (2014) Gonadoblastoma and selected other aspects of gonadal pathology in young patients with disorders of sex development. Semin Diagn Pathol 31:427–440 Ulirsch RC, Goldman RL (1982) An unusual teratoma of the ovary: neurogenic cyst with lactating breast tissue. Obstet Gynecol 60:400–402 Vance RP, Geisinger KR (1985) Pure nongestational choriocarcinoma of the ovary. Report of a case. Cancer (Phila) 56:2321–2325 Vang R, Gown AM, Zhao C, Barry TS, Isacson C, Richardson MS, Ronnett BM (2007) Ovarian mucinous tumors associated with mature cystic teratomas: morphologic and immunohistochemical analysis identifies a subset of potential teratomatous origin that shares features of lower gastrointestinal tract mucinous tumors more commonly encountered as secondary tumors in the ovary. Am J Surg Pathol 31:854–869 Venizelos ID, Tatsiou ZA, Roussos D, Karagiannis V (2009) A case of sebaceous carcinoma arising within a benign ovarian cystic teratoma. Onkologie 32:353–355 Vicus D, Beiner ME, Klachook S, Le LW, Laframboise S, Mackay H (2010) Pure dysgerminoma of the ovary 35 years on: a single institutional experience. Gynecol Oncol 117:23–26 Vitaliani R, Mason W, Ances B, Zwerdling T, Jiang Z, Dalmau J (2005) Paraneoplastic encephalitis, psychiatric symptoms, and hypoventilation in ovarian teratoma. Ann Neurol 58:594–604 Wang WC, Lai YC (2016) Genetic analysis results of mature cystic teratomas of the ovary in Taiwan disagree with the previous origin theory of this tumor. Hum Pathol 52:128–135 Wang Y, Schwartz LE, Anderson D, Lin MT, Haley L, Wu RC, Vang R, Shih Ie M, Kurman RJ (2015) Molecular analysis of ovarian mucinous carcinoma reveals different cell of origins. Oncotarget 6:22949–22958 Warkel RL, Cooper PH, Helwig EB (1978) Adenocarcinoid, a mucin-producing carcinoid tumor of the appendix: a study of 39 cases. Cancer (Phila) 42:2781–2793 Weinberg LE, Lurain JR, Singh DK, Schink JC (2011) Survival and reproductive outcomes in women treated for malignant ovarian germ cell tumors. Gynecol Oncol 121:285–289 Weldon-Linne CM, Rushovich AM (1983) Benign ovarian cystic teratomas with homunculi. Obstet Gynecol 61:88S–94S

K. P. Maniar and R. Vang Williamson HO, Underwood PB Jr, Kreutner A Jr, Rogers JF, Mathur RS, Pratt-Thomas HR (1976) Gonadoblastoma: clinicopathologic correlation in six patients. Am J Obstet Gynecol 126:579–585 Wisniewski M, Deppisch LM (1973) Solid teratomas of the ovary. Cancer (Phila) 32:440–446 Witschi E (1948) Migration of the germ cells of human embryos from the yolk sac to the primitive gonadal folds. Contrib Embryol 32:67 Woodruff JD, Rauh JT, Markley RL (1966) Ovarian struma. Obstet Gynecol 27:194–201 Woodruff JD, Protos P, Peterson WF (1968) Ovarian teratomas. Relationship of histologic and ontogenic factors to prognosis. Am J Obstet Gynecol 102:702–715 Yanai-Inbar I, Scully RE (1987) Relation of ovarian dermoid cysts and immature teratomas: an analysis of 350 cases of immature teratoma and 10 cases of dermoid cyst with microscopic foci of immature tissue. Int J Gynecol Pathol 6:203–212 Young RH, Scully RE (1983) Ovarian Sertoli-Leydig cell tumors with a retiform pattern: a problem in histopathologic diagnosis. A report of 25 cases. Am J Surg Pathol 7:755–771 Young RH, Prat J, Scully RE (1982) Ovarian SertoliLeydig cell tumors with heterologous elements. I. Gastrointestinal epithelium and carcinoid: a clinicopathologic analysis of thirty-six cases. Cancer (Phila) 50:2448–2456 Young RH, Stall JN, Sevestre H (2016) The polyembryoma: one of the most intriguing human neoplasms, with comments on the investigator who brought it to light, Albert Peyron. Int J Gynecol Pathol 35:93–105 Zaloudek CJ, Tavassoli FA, Norris HJ (1981) Dysgerminoma with syncytiotrophoblastic giant cells. A histologically and clinically distinctive subtype of dysgerminoma. Am J Surg Pathol 5:361–367 Zelaya G, Lopez Marti JM, Marino R, Garcia de Davila MT, Gallego MS (2015) Gonadoblastoma in patients with Ullrich-Turner syndrome. Pediatr Dev Pathol 18:117–121 Zhao S, Kato N, Endoh Y, Jin Z, Ajioka Y, Motoyama T (2000) Ovarian gonadoblastoma with mixed germ cell tumor in a woman with 46, XX karyotype and successful pregnancies. Pathol Int 50:332–335 Zhao C, Bratthauer GL, Barner R, Vang R (2007) Comparative analysis of alternative and traditional immunohistochemical markers for the distinction of ovarian sertoli cell tumor from endometrioid tumors and carcinoid tumor: a study of 160 cases. Am J Surg Pathol 31: 255–266 Zuntova A, Motlik K, Horejsi J, Eckschlager T (1992) Mixed germ cell-sex cord stromal tumor with heterologous structures. Int J Gynecol Pathol 11: 227–233 Zynger DL, McCallum JC, Luan C, Chou PM, Yang XJ (2010) Glypican 3 has a higher sensitivity than alphafetoprotein for testicular and ovarian yolk sac tumour: immunohistochemical investigation with analysis of histological growth patterns. Histopathology 56: 750–757

Nonspecific Tumors of the Ovary, Including Mesenchymal Tumors

17

Lauren E. Schwartz and Russell Vang

Contents Mesenchymal Tumors Nonspecific to the Ovary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Low–Grade Endometrioid Stromal Sarcoma (LGESS) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Myxoma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Undifferentiated Sarcoma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1126 1127 1128 1129

Tumors of Muscle Differentiation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Leiomyoma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Leiomyosarcoma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Rhabdomyoma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Rhabdomyosarcoma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Myofibroblastoma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1129 1129 1130 1131 1131 1133

Tumors of Vascular and Lymphatic Differentiation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Hemangioma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Angiosarcoma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Lymphangioma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1133 1133 1135 1136

Tumors of Cartilage Differentiation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1136 Chondroma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1136 Chondrosarcoma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1136 Tumors of Bone Differentiation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1137 Osteoma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1137 Osteosarcoma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1137 Tumors of Neural Differentiation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1137 Neurofibroma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1137

L. E. Schwartz (*) Department of Pathology and Laboratory Medicine, Perelman School of Medicine at the University of Pennsylvania, Philadelphia, PA, USA e-mail: [email protected] R. Vang Department of Pathology, Division of Gynecologic Pathology, The Johns Hopkins Medical Institutions, Baltimore, MD, USA e-mail: [email protected] # Springer Nature Switzerland AG 2019 R. J. Kurman et al. (eds.), Blaustein’s Pathology of the Female Genital Tract, https://doi.org/10.1007/978-3-319-46334-6_17

1125

1126

L. E. Schwartz and R. Vang Schwannoma (Neurilemmoma) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Malignant Peripheral Nerve Sheath Tumor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Paraganglioma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Ganglioneuroma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1138 1138 1138 1138

Tumors of Adipose Tissue Differentiation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1139 Tumors of Mesothelial Differentiation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1139 Adenomatoid Tumor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1139 Peritoneal Malignant Mesothelioma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1140 Wolffian Tumor (Formally Known as Female Adnexal Tumor of Probable Wolffian Origin (FATWO)) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1140 Lesions of the Rete Ovarii . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1141 Primary Ovarian Tumors of Uncertain Histogenesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Hepatoid Carcinoma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Small Cell Carcinoma, Pulmonary Type (Neuroendocrine Type) . . . . . . . . . . . . . . . . . . . . Neuroendocrine Carcinoma, Non-Small Cell Type . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Salivary Gland-Like Carcinomas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Nephroblastoma (Wilms Tumor) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1142 1142 1144 1145 1146 1147

References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1147

The tumors discussed in this chapter comprise a heterogeneous group of neoplasms, many of which are not specific to the ovary. Most are uncommon in this location, occurring much more frequently in other parts of the body. Consequently, whenever they are encountered in the ovary, these tumors pose difficult problems in diagnosis, histogenesis, behavior, and therapy for the pathologist and clinician. These neoplasms must be differentiated from primary ovarian neoplasms containing mesenchymal tissue, as well as from metastatic and disseminated neoplasms affecting the ovary. Thus, mesenchymal neoplasms nonspecific to the ovary must be differentiated primarily from teratomatous neoplasms containing large amounts of mature or immature mesenchymal elements and from carcinosarcoma (malignant mixed mullerian tumor), which are composed of different malignant mesenchymal elements in addition to their malignant epithelial components. Tumors of teratomatous origin containing mesenchymal tissue are described in Chap. 16, “Germ Cell Tumors of the Ovary,” and carcinosarcomas, endometrioid stromal sarcomas, and adenosarcoma are discussed in Chap. 14, “Epithelial Tumors of the Ovary.”

In addition to the mesenchymal neoplasms nonspecific to the ovary, adenomatoid tumor, which is of mesothelial origin, Wolffian tumors, ovarian neoplasms of neural origin, hepatoid carcinoma of the ovary, small cell carcinoma of the ovary (pulmonary type), and tumors of salivary gland type are included. Although small cell carcinoma of hypercalcemic type is listed in the miscellaneous ovarian tumor category in the 2014 World Health Organization (WHO) classification (Kurman et al. 2014), that tumor is discussed in Chap. 15, “Sex Cord-Stromal, Steroid Cell, and Other Ovarian Tumors with Endocrine, Paraendocrine, and Paraneoplastic Manifestations.” Lastly, solid pseudopapillary neoplasm of the ovary (Deshpande et al. 2010) is also listed in the miscellaneous ovarian tumor category in the 2014 WHO classification (Kurman et al. 2014) but will not be discussed.

Mesenchymal Tumors Nonspecific to the Ovary Mesenchymal neoplasms that arise in the ovary are rare and thought to arise from connective tissue found in the ovary, rather than thought to

17

Nonspecific Tumors of the Ovary, Including Mesenchymal Tumors

be of teratomatous or surface epithelial–stromal (müllerian) origin. Teratomatous origin or origin from surface epithelial-stromal (mullerian) tumors cannot be excluded in a number of cases. The neoplasms discussed here are composed of a single neoplastic mesenchymal element, either benign or malignant, in contrast to teratomatous tumors or carcinosarcomas, which are usually composed of a number of tissue elements. Some issues of classification and histogenesis may not be resolvable in view of the possibility of a monomorphic teratoma. Thus, although some of these neoplasms can be shown to originate directly from ovarian tissue, a considerable number of cases are of indeterminate histogenesis and origin. Mesenchymal neoplasms of the ovary can be benign or malignant, just like their counterparts that arise in other areas of the body, and are classified on the basis of the tissue of origin.

1127

(Figs. 1 and 2). Mitotic figures (MF) are variable and the vasculature seen is similar to that described in uterine low-grade endometrial stroma sarcomas (see ▶ Chap. 10, “Mesenchymal Tumors of the Uterus”). In some cases sex-cord differentiation and smooth muscle metaplasia have been identified. In the majority of reported cases, intermixed endometriosis has been identified, suggesting the tumor arises from endometriosis rather than ovarian stroma. Immunohistochemically, these tumors tend to exhibit strong and diffuse positivity for CD10. Similar to the uterine counterpart, JAZF1-

Low–Grade Endometrioid Stromal Sarcoma (LGESS) LGESS primary to the ovary is a rare neoplasm with less than 100 cases reported (Oliva et al. 2014; Young et al. 1984; Chang et al. 1993; Masand et al. 2013). The patients are of variable age with the majority of tumors occurring during the fifth and sixth decades (Oliva et al. 2014; Chang et al. 1993). Clinically patients tend to present with non-specific symptoms including abdominal distension and/or pain. The neoplasms have been noted to predominately be unilateral, with some cases being bilateral in nature. Macroscopically the tumors range in size with a mean size close to 10 cm (Oliva et al. 2014) Macroscopically the lesions vary from predominately cystic to predominately solid. The cut surface is typically tan-yellow with areas of hemorrhage and/or necrosis. Microscopically the tumor appears similar to its uterine counterpart and consists of sheets of small, closely packed cells resembling proliferative phase endometrial stroma

Fig. 1 Low-grade endometrioid stromal sarcoma (low power)

Fig. 2 Low-grade endometrioid stromal sarcoma (high power)

1128

SUZ12 gene fusions and PHF1 have been detected in some of these tumors (AmadorOrtiz et al. 2011; Chiang et al. 2011). The differential diagnosis of these lesions consists of metastatic lesions, sex cord-stromal tumors and other less common entities. Endometrioid stromal sarcomas are for the most part considered primary to the ovary if the uterus is uninvolved after extensive evaluation, the largest mass is in the ovary, and presentation is similar to ovarian cancer (Oliva et al. 2014). Besides metastatic endometrial stromal sarcoma from the uterus, metastatic gastrointestinal stromal tumors (GIST) should enter the differential diagnosis, as these tumors tend to occur in similar aged patients. Histologically, metastatic GISTs are more likely to show nuclear palisading and the absence of the typical vasculature seen in LGESS. Further, immunohistochemical stains should help to evaluate this differential as GISTs should show positivity for c-kit, DOG-1 and/or CD34.

Myxoma Primary myxoma of the ovary is a very rare neoplasm: only a very small number of cases have been reported in the literature (Brady et al. 1987; Eichhorn and Scully 1991; Scully et al. 1998; Roth et al. 2013). The reported patients range in age from 12–80 years, with only one being regarded as post-menopausal. All reported cases are unilateral (Brady et al. 1987; Eichhorn and Scully 1991; Scully et al. 1998; Roth et al. 2013). Macroscopically, the tumors range from 3.5 to 22 cm in greatest dimension (Eichhorn and Scully 1991; Scully et al. 1998; Roth et al. 2013; Dutz and Stout 1961). The tumors are encapsulated, gray-white, and soft; on cut section, they are partly cystic. Solid areas are slimy and mucinous, whereas the cystic spaces contain a viscous, glassy, gelatinous material. Microscopically, the tumors show the typical appearance of myxomas seen in other locations. They are composed of loose myxomatous stroma

L. E. Schwartz and R. Vang

within which there are scattered stellate or spindle-shaped cells, some of which contain hyperchromatic nuclei. There is no nuclear pleomorphism, and mitotic activity is absent. The tumors vary from poorly vascularized, containing only a few capillary blood vessels and showing absence of plexiform vessels, to tumors with prominent capillary vessels within the tumor, and larger vessels with muscular walls at its periphery. The myxomatous stroma stains positively with alcian-blue stain and contains a network of fine reticulum fibers. Stains for fat were negative. In some areas, fibrosis can be present. There are no other connective tissue elements, and the tumors have a homogeneous appearance. Myxoma is immunoreactive for vimentin and focally for actin, but negative for desmin, inhibin, cytokeratins, vascular markers, S-100, and neurofilaments (Eichhorn and Scully 1991; Roth et al. 2013; Costa et al. 1993). In a relatively recent review of ovarian myxomas, three cases were noted to have occurred within ovaries also containing additional sex cord-stromal tumors (Roth et al. 2013). Two cases showed a sclerosing stromal tumor, whereas the other showed a luteinized theca cell tumor. This led the authors to speculate as to whether these ovarian myxomas arose from the other stromal lesions, suggesting two distinct pathways for the development of ovarian myxoma, one from ovarian myxoid connective tissue and the other from other sex cord-stromal neoplasms (Roth et al. 2013). The authors advocate for mentioning a myxomatous component if it is distinct and measures >1 cm in greatest (Roth et al. 2013). More research is needed to further investigate the various origins of these tumors. Although myxoma is considered a benign neoplasm, because of its viscous nature it is difficult to excise it completely, and recurrences are not uncommon unless the entire adnexa bearing the tumor is excised. All the known patients treated by unilateral adnexectomy, in which the diagnosis of myxoma was confirmed, and for whom there is follow-up information are free of disease after 1–21 years (Roth et al. 2013; Costa et al. 1993).

17

Nonspecific Tumors of the Ovary, Including Mesenchymal Tumors

The differential diagnosis for an ovarian myxoma is broad. Most importantly, myxomas must be distinguished from sarcomas, including myxoid liposarcoma and embryonal rhabdomyosarcoma. Myxoid liposarcoma which contains fat, is more vascularized, and shows lipoblasts at least in some areas. Embryonal rhabdomyosarcoma shows less of a uniform appearance, displays greater cellular and nuclear pleomorphism, and contains rhabdomyoblasts. In addition, embryonal rhabdomyosarcoma shows immunohistochemical staining for musclespecific actin, desmin, and myogenin. Even focal atypia should raise suspicious for a low grade sarcoma and the differential should be investigated. The differential diagnosis for myxoma also includes various benign entities. These include fibroma with myxoid degeneration, which contains normal fibrous tissue in some areas, and massive edema of the ovary (see ▶ Chap. 12, “Nonneoplastic Lesions of the Ovary”) (Young and Scully 1984; Kalstone et al. 1969). The patients with massive edema usually are younger, and the lesion shows entrapment of follicular derivatives, which is not observed in ovarian myxoma. Lastly, myxoma must be differentiated from mucinous cystadenomas and carcinomas, either primary or metastatic, which contain epithelial cells, show absence of stellate and spindle-shaped cells, and may show glandular differentiation. Cytokeratin stains should help with this diagnosis.

Undifferentiated Sarcoma Some ovarian tumors are poorly differentiated, and although a diagnosis of sarcoma can be made, the tumor does not exhibit further differentiation beyond showing its mesenchymal origin. Careful and extensive gross sampling and histologic examination in such cases are helpful and may result in finding better-differentiated areas, which will yield a more accurate diagnosis. Immunohistochemical investigations may be very helpful in accurately detecting the tissue of

1129

origin and should be undertaken in all such cases. In some cases, a more precise diagnosis cannot be made despite very extensive investigations.

Tumors of Muscle Differentiation Leiomyoma Primary leiomyoma of the ovary is uncommon (Scully et al. 1998; Doss et al. 1999; Kandalaft and Esteban 1992; Lerwill et al. 2004; Prayson and Hart 1992). A small number of cases are on record, but it is likely that many cases are not reported, especially when the tumor is small and is discovered incidentally. Primary ovarian leiomyoma probably originates from smooth muscle present in the walls of blood vessels in the cortical stroma, in the corpus luteum, and in the ovarian ligaments at their point of attachment to the ovary; its precise histogenesis is uncertain, however. This tumor usually is found in menopausal and postmenopausal women, but sometimes occurs in young women. The age of patients ranges from 3 to 65 years. Clinically, many patients are asymptomatic, and the tumor is discovered incidentally. When symptoms are present, they are related to the presence of an adnexal mass, often accompanied by abdominal swelling and pain. The latter may be acute because of torsion. Ascites is rare, and hydrothorax has not been reported. The uterus usually contains leiomyomas. Ovarian leiomyoma is usually unilateral, although a single case of large bilateral ovarian leiomyomas occurring in a 21-year-old woman has been reported (Kandalaft and Esteban 1992). Macroscopically, the tumors are solid, firm, and round or oval masses having a smooth surface. On cut section they have a white or gray-white solid whorled surface. Coagulative type necrosis should not be identified. Microscopically, the tumor shows typical appearances of a leiomyoma, as observed in the uterus, the tumor being composed of smooth muscle cells that are uniformly spindle-shaped or elongated and contain elongated

1130

L. E. Schwartz and R. Vang

tendency to diagnose leiomyoma as a fibroma, but use of immunohistochemical stains for this distinction should be done with caution since fibromas can also show expression of muscle markers (Costa et al. 1993; Tiltman and Haffajee 1999). The treatment is excision of the affected adnexa.

Leiomyosarcoma

Fig. 3 Primary leiomyoma of the ovary. The tumor shows an appearance similar to the much more common uterine leiomyoma

blunt-ended or cigar-shaped nuclei (Fig. 3). Palisading of the nuclei may be present and may be prominent. Mitotic activity is absent or very low, and cellular and nuclear pleomorphism is not a feature. Most leiomyomas are of the typical type, but cellular, mitotically active, myxoid, epithelioid and other subtypes can be seen. The tumor cells form bundles intersected by fibrous septa that may be wide and show marked hyalinization. Other degenerative changes seen in uterine leiomyomas also may be present. Occasionally a leiomyoma may show an epithelioid pattern, which may cause some diagnostic problems. Immunohistochemical expression of smooth muscle actin, muscle-specific actin, desmin, ER, and PR can be seen. A well-documented case of a large ovarian lipoleiomyoma occurring in a 63-year-old woman has been reported (Mira 1991). The tumor replaced nearly the entire ovary. The adipose tissue was found replacing and dissecting the smooth muscle within the tumor. There was no associated uterine leiomyomatosis. Primary ovarian leiomyoma must be differentiated from pedunculated subserosal (parasitic) uterine leiomyoma, which has lost its attachment and instead has become attached to the ovary, from which it draws its blood supply. Leiomyoma also must be differentiated from ovarian fibroma, as the latter is much more common. There is a

Primary leiomyosarcoma of the ovary is very rare. These tumors usually are found in postmenopausal women, but sometimes may be seen in younger women (Lerwill et al. 2004; Vijaya Kumar et al. 2015; Balaton et al. 1987). The tumors usually are large and solid, and patients have symptoms and signs related to the presence of an abdominal or pelvic mass. The tumors are gray-yellow, soft, fleshy, and frequently associated with hemorrhage and necrosis. Microscopically, they differ from a leiomyoma by the presence of a variable combination of mitotic activity, cellular and nuclear pleomorphism, and necrosis (Figs. 4 and 5). It has been proposed that leiomyosarcoma should be diagnosed when 2 of the following histologic features are present: significant nuclear atypia, mitotic index 10 MF/10 high-power fields, and tumor cell necrosis; it has also been suggested that a smooth muscle tumor with nuclear atypia can qualify as a leiomyosarcoma if the mitotic index is 5 MF/10 high-power fields (HPF), even if tumor cell necrosis is not present (Lerwill et al. 2004). The diagnosis of “smooth muscle tumor of uncertain malignant potential” can be used for tumors with histologic features intermediate between leiomyoma and leiomyosarcoma. Most leiomyosarcomas are of the conventional type, but occasional ovarian leiomyosarcomas may be of the myxoid or epithelioid type. It is important to recognize these unusual variants, which are similar to their counterparts in the uterus. Primary leiomyosarcoma of the ovary metastasizes via the bloodstream; the prognosis is generally unfavorable. Primary leiomyosarcoma of the ovary must be distinguished from carcinosarcomas containing

17

Nonspecific Tumors of the Ovary, Including Mesenchymal Tumors

1131

cystadenoma in a 48 year old has been reported (Huang et al. 2005).

Rhabdomyosarcoma

Fig. 4 Primary leiomyosarcoma of the ovary. Low magnification. Note the similarity to the leiomyoma seen in Fig. 3

Fig. 5 Primary leiomyosarcoma of the ovary. High magnification. Note the marked cellular and nuclear pleomorphism

a prominent leiomyosarcomatous component. Primary leiomyosarcoma also should be distinguished from immature teratoma with a prominent leiomyomatous tissue component. It also must be distinguished from metastatic leiomyosarcoma of uterine or other origin, as well as from poorly differentiated sarcomas and carcinosarcomas, both primary and metastatic to the ovary.

Rhabdomyoma No well-documented case of pure ovarian rhabdomyoma has been recorded. A case of mural nodules of rhabdomyoma within a serous

Primary rhabdomyosarcoma of the ovary is uncommon. A small number of cases have been reported in the literature. A careful review of the literature shows that some cases, such as the frequently quoted case reported by Sandison (1955), were not pure rhabdomyosarcomas but rather examples of carcinosarcoma or teratomas with a marked rhabdomyoblastic component. Therefore, before a diagnosis of primary ovarian rhabdomyosarcoma can be made, the tumor must be sampled carefully and extensively to exclude the presence of other neoplastic elements, the presence of which would preclude a diagnosis of a pure rhabdomyosarcoma of the ovary. The diagnosis of an embryonal rhabdomyosarcoma in a young patient should raise consideration for DICER1 syndrome as these tumors may be a manifestation of this syndrome (Stewart et al. 2016; de Kock et al. 2015). The histogenesis of primary rhabdomyosarcoma of the ovary is uncertain. These tumors may originate from the connective tissue of the ovary, as a one-sided development of a teratoma, as a result of malignant transformation of a mature cystic teratoma with the malignant element overgrowing the tumor, or as a one-sided development of a carcinosarcoma. The age of patients with ovarian rhabdomyosarcoma ranges from 2.5 to 84 years. The small number of cases makes it impossible to state whether there is a predilection for any particular age group, but, as with rhabdomyosarcomas occurring in other locations, the pleomorphic type occurs in older patients, whereas the embryonal and alveolar types occur in young women and children (Chan et al. 1989). Patients with ovarian rhabdomyosarcoma usually have symptoms associated with the presence of a large, usually rapidly growing, abdominal mass, often associated with hemorrhagic ascites. Metastases frequently are seen at presentation.

1132

Macroscopically, the tumors are unilateral, but metastatic involvement of the contralateral ovary may be present. The tumors usually are large, exceeding 10 cm in diameter. They are solid, soft, fleshy, and gray-pink to yellow-tan, with areas of hemorrhage and necrosis that may be prominent. Microscopically, the tumors may be of the embryonal (including the botryoid type), alveolar, or pleomorphic type and contain variable numbers of rhabdomyoblasts (Figs. 6, 7, and 8). Tumors composed of the former types occur in children and young adults, whereas those of the pleomorphic type are observed in older women. The diagnosis of pleomorphic rhabdomyosarcoma should not present undue difficulty, because of the presence of at least some typical rhabdomyoblasts showing crossstriations. In cases of embryonal rhabdomyosarcoma, the diagnosis is much more difficult because the tumor cells are poorly differentiated, making rhabdomyoblastic differentiation discernible only with difficulty. Furthermore, it is necessary to recognize the distinctive alveolar or botryoid patterns, which also may not be easy. The embryonal rhabdomyosarcoma is composed of small round primitive cells having a narrow rim of cytoplasm. They are poorly differentiated rhabdomyoblasts in various stages of differentiation. Therefore, the lesion is difficult to distinguish from poorly differentiated small cell carcinoma, lymphoma/leukemia, or even neuroblastoma (Scully et al. 1998; Nielsen et al. 1998; Nunez et al. 1983). Among the small round cells are scattered occasional rhabdomyoblasts, which are better-differentiated large cells with bright eosinophilic cytoplasm and eccentric nuclei (Fig. 8). Presence of these cells and their recognition may provide a diagnostic clue. Occasionally, large, more typical rhabdomyoblasts may be seen. The presence of cross-striations is not necessary for diagnosis, but the cells comprising the tumor may be wellenough differentiated to exhibit cross-striations (Fig. 7). Demonstration of Z bands or their precursors by electron microscopy is helpful in making the diagnosis. Immunohistochemical

L. E. Schwartz and R. Vang

Fig. 6 Primary rhabdomyosarcoma of the ovary. This portion of the tumor contains a significant spindle cell component

Fig. 7 Primary rhabdomyosarcoma of the ovary. Distinctive cross-striations are seen (arrow)

Fig. 8 Primary rhabdomyosarcoma of the ovary. The tumor contains abundant rhabdomyoblasts, which have an ample amount of bright eosinophilic cytoplasm and eccentric nuclei

17

Nonspecific Tumors of the Ovary, Including Mesenchymal Tumors

1133

Myofibroblastoma

Fig. 9 Primary rhabdomyosarcoma of the ovary showing diffuse expression of myogenin

demonstration of myoglobin, desmin, musclespecific actin, and myogenin are helpful in this respect (Fig. 9). The tumor is frequently affected by edema, hemorrhage, and necrosis, making the diagnosis even more difficult. Therefore, thorough examination and sampling of the tumor are essential to make the correct diagnosis. The tumor may be more common than has been hitherto believed, but because of its poor differentiation, it may have been either assigned to the group of undifferentiated ovarian tumors or misdiagnosed. In some cases, the tumor infiltrated the bone marrow and was originally diagnosed as leukemia (Nunez et al. 1983). It is therefore emphasized that embryonal rhabdomyosarcoma must be considered in the differential diagnosis of undifferentiated small round cell tumor of the ovary in a young patient. The presence of other neoplastic elements always must be excluded when making this diagnosis. The importance of making the correct diagnosis is not only academic but practical, in view of the advances that have been made in the therapy of embryonal rhabdomyosarcoma during the past few decades. In the past, the prognosis was poor, and in most reported cases, the patients died of extensive metastatic disease within 1 year of diagnosis. Recently, patients with embryonal rhabdomyosarcoma, some of whom had metastases, are well and disease-free after surgery, chemotherapy, and radiotherapy.

A well-documented case of an ovarian myofibroblastoma has been reported (Rhoades et al. 1999). A 22-year-old woman who was involved in an automobile accident was found to have an enlarged right ovary but refused laparotomy. Over the next 3 years the mass gradually increased in size, and laparotomy was performed. A 9  8.5  6 cm, right-sided ovarian tumor weighing 215 g and adherent to the right fallopian tube and omentum was found and excised. The tumor was solid, white-tan, and on sectioning revealed whorled areas and focal calcification. Microscopically, it was composed of uniform bland-looking spindle cells arranged haphazardly in fascicles separated by bands of hyalinized collagen. In some areas there was increased vascularity. There was no atypia or mitotic activity. The tumor cells showed vimentin, smooth muscle actin, and muscle-specific actin positivity. There was no immunoreactivity with desmin and cytokeratin. The patient was well and diseasefree 21 months after treatment. Myofibroblastoma is a benign lesion, and complete excision results in cure.

Tumors of Vascular and Lymphatic Differentiation Hemangioma Hemangiomas, cavernous type, capillary type and anatomosing type, are found only occasionally in the ovary, with less than 100 cases reported in the literature (Ziari and Alizadeh 2016; Lawhead et al. 1985; Talerman 1967; Dundr et al. 2017) Rare patients with this lesion present with clinical findings suspicious for a malignant process (Schoolmeester et al. 2015). The origin of ovarian hemangioma, in common with hemangioma in general, is a matter of controversy; it is considered either a hamartomatous malformation or a true neoplasm. It is likely that both modes of origin are responsible for their formation. The reported age of patients with ovarian hemangioma ranges from 4 months to 81 years (Ziari and Alizadeh

1134

2016), and does not show a predominance in any decade. In most patients, ovarian hemangioma has been noted as an incidental finding at operation or autopsy (Talerman 1967). In a few cases, the lesion was large and the patient had abdominal enlargement because of the presence of an ovarian mass (Gehrig et al. 2000; Mann and Metrick 1961; Mc and Trumbull 1955) or had acute abdominal pain associated with torsion of the tumor (Mann and Metrick 1961; Shaffer and Cancelmo 1939). In some cases, there was ascites, which resolved following removal of the lesion (Gehrig et al. 2000; Mc and Trumbull 1955; Savargaonkar et al. 1994). The lesions usually are unilateral, although in four patients they were bilateral (Talerman 1967). Ovarian hemangiomas have been noted in patients with generalized hemangiomatosis (Lawhead et al. 1985) and in patients with hemangiomas in other parts of the genital tract (Lawhead et al. 1985; Talerman 1967). Further, ovarian hemangiomas have also been linked to various syndromes including Kasabach-Merritt syndrome and pseudo-Meigs syndrome and have also been seen in patients with elevated CA-125 (Schoolmeester et al. 2015). Macroscopically, the lesions are small, red or purple, round or oval nodules, measuring from a few millimeters to 24 cm in diameter. On cut section, they usually are spongy and show a honeycomb appearance. Although they have been found in different parts of the ovary, the medulla and the hilar region appear to be the most common sites (Talerman 1967). Microscopically, ovarian hemangioma is of the cavernous, mixed capillary-cavernous type or anastomosing type. Generally, hemangiomas consist of collections of vascular spaces, which may vary in size but usually are small, lined by a single layer of endothelial cells, and usually contain red blood cells in their lumens (Fig. 10). The anastomosing hemangioma, which has been recently described in the ovary, consists of a non-lobular vascular proliferation of capillary sized vessels which are usually intermixed with a larger vessel (Dundr et al. 2017; Kryvenko et al. 2011). Occasionally, in hemangiomas thrombosis may be seen. In a few reported cases, the hemangioma was associated with the

L. E. Schwartz and R. Vang

Fig. 10 Hemangioma of the ovary. The tumor is composed of numerous small blood vessels, some of which contain red blood cells in their lumens

presence of luteinized cells in the stroma of the lesion. In one such case, there was evidence of hormonal function (Savargaonkar et al. 1994). Hemangioma must be differentiated from proliferations of dilated blood vessels, frequently seen in the hilar region of the ovary. Although a very small hemangioma may not be easily distinguished from such vascular proliferations, the hemangioma usually forms a nodule or a small mass. The presence of a circumscribed nodule composed of vascular spaces tends to distinguish hemangioma from vascular proliferations, which usually are smaller and more diffuse. The presence of numerous blood cells within the vascular spaces and the absence of pale eosinophilic homogeneous material usually distinguish hemangioma from the less common lymphangioma, but immunohistochemical stains including CD31, CD34, ERG, and FLI-1 can also be used (Schoolmeester et al. 2015). Hemangioma also must be distinguished from teratoma with a prominent vascular component. In such cases, careful sampling will detect other teratomatous elements, the presence of which distinguishes the lesion from a hemangioma. The differential diagnosis for hemangiomas of the ovary, especially anastomosing hemangiomas, also includes well differentiated angiosarcoma and Kaposi sarcoma. Unlike these two malignant entities, hemangiomas lack cytologic atypia and tend to be small and somewhat lobulated (Kryvenko et al. 2011). The treatment of choice is oophorectomy.

17

Nonspecific Tumors of the Ovary, Including Mesenchymal Tumors

1135

Angiosarcoma Angiosarcoma is a very rare ovarian neoplasm (Scully et al. 1998; Nielsen et al. 1997; Nucci et al. 1998; Kruse et al. 2014; Yaqoob et al. 2014). In some reported cases, the angiosarcoma had arisen within a mature cystic teratoma or may have been associated with an immature teratoma. Such cases are considered as germ cell tumors and are excluded from consideration here. We have also encountered an angiosarcoma as a mural nodule in an ovarian atypical proliferative (borderline) mucinous tumor and this rare type is also excluded from this discussion. The age of the patients with angiosarcoma varied from 19 to 77 years. The tumor usually is unilateral, but bilateral tumors have been recorded. The histogenesis of primary ovarian angiosarcoma is uncertain. It may originate from the vascular tissue present in the ovary, as a one-sided development of a teratoma, or from a teratoma in which the vascular component has overgrown the other parts of the tumor. Patients usually have symptoms related to the presence of a lower abdominal mass, which may be associated with torsion and rupture of the tumor and hemorrhage. Macroscopically, the tumors usually are large, blue-brown, hemorrhagic, soft, and friable. They may be confined to the ovary, but often are associated with invasion of the surrounding structures. Microscopically, they are composed of vascular spaces of varying size and appearance, lined by endothelial cells that usually are large, showing an atypical appearance, bizarre nuclei, and mitotic activity (Figs. 11 and 12). In some areas, the tumor may contain a considerable amount of connective tissue interspersed between the vascular spaces. Fine papillary projections lined by atypical endothelial cells may be seen and are prominent. Some tumors are composed of small closely packed spaces lined by atypical cells with a suggestion of a solid pattern (Nucci et al. 1998). Angiosarcoma of the ovary must be distinguished from immature teratomatous neoplasms with a prominent vascular component. The

Fig. 11 Primary angiosarcoma of ovary. The tumor contains closely packed vessels forming a solid spindle cell pattern (top). More typical vascular pattern is seen below

Fig. 12 Primary angiosarcoma of ovary. The tumor is composed of dilated vascular spaces lined by enlarged hyperchromatic cells

presence of other neoplastic germ cell elements distinguishes teratoma from primary angiosarcoma. Immunohistochemical stains for CD31, CD34, and ERG are useful in confirming the diagnosis of angiosarcoma when the tumor is poorly differentiated, especially when showing a solid pattern. The tumor invades locally and metastasizes via the bloodstream. Prognosis is poor, especially in patients who have metastases at the time of presentation. When the tumor is confined to the ovary, the prognosis is better and a few survivors have been reported. Recent studies have shown some improved survival with adjuvant chemotherapy (Kruse et al. 2014).

1136

Lymphangioma Lymphangioma of the ovary is very rare (Singer et al. 2010; Radhouane et al. 2016). The tumors tend to be small and found incidentally. They are slow growing and often remain asymptomatic for a long time. The tumor is usually unilateral but bilateral lesions have been reported. Macroscopically, the tumor is small with a smooth, gray surface. On cut section, it is yellow, honeycombed, and composed of numerous small cystic spaces exuding clear yellow fluid. Microscopically, lymphangioma of the ovary is composed of closely packed, thin-walled vascular spaces lined by flattened endothelial cells and containing pale, homogeneous eosinophilic fluid (Fig. 13). Lymphocytes may be seen within the vascular spaces. The histogenesis is a matter of controversy. Some investigators consider these lesions as malformations and some as neoplasms. Both modes of histogenesis are likely. Lymphangioma is differentiated from a teratoma with a prominent vascular component by the absence of other germ cell elements. Lymphangioma also must be distinguished from hemangioma and an adenomatoid tumor that contains thin-walled, vessel-like spaces. In contrast to hemangioma, lymphangioma does not contain blood cells in the vascular spaces and shows D2–40 staining. Adenomatoid tumor has solid

Fig. 13 Primary ovarian lymphangioma. The tumor is composed of large and closely packed thin-walled lymphatic spaces lined by flattened endothelial cells and contains pale eosinophilic fluid

L. E. Schwartz and R. Vang

areas, and the cells lining the vessel-like spaces show immunohistochemical expression of cytokeratin and calretinin.

Tumors of Cartilage Differentiation Chondroma Only a few reports of ovarian chondroma are available, and documentation in most cases is unsatisfactory. One well-documented case considered to originate from the ovarian stroma has been reported (Nogales 1982). The tumor, which measured 4  3  3 cm and was composed entirely of mature cartilage, was found incidentally. Although chondroma may originate from the connective tissue of the ovary by a process of metaplasia, it is more likely that most ovarian tumors described as chondroma were either fibromas showing cartilaginous metaplasia or teratomas having a prominent cartilaginous component.

Chondrosarcoma Pure chondrosarcoma of the ovary (Figs. 14 and 15) is rare. In one report (Talerman et al. 1981), a 61-year-old woman had an abdominal mass that on extensive microscopic examination proved to

Fig. 14 Primary magnification

ovarian

chondrosarcoma.

Low

17

Nonspecific Tumors of the Ovary, Including Mesenchymal Tumors

1137

Osteosarcoma

Fig. 15 Primary ovarian chondrosarcoma. Marked nuclear atypia

be a pure, well-differentiated chondrosarcoma. The patient was well and disease-free 6 years after unilateral oophorectomy. The histogenesis of this tumor is uncertain, but the age of the patient and the histologic appearances of the tumor point to an origin in a dermoid cyst with malignant transformation and overgrowth by the malignant cartilaginous component (Talerman et al. 1981). Well-documented cases of mature cystic teratoma (dermoid cyst) with malignant transformation of the cartilaginous element have been reported (Yasunaga et al. 2011; Climie and Heath 1968). In malignant ovarian tumors with cartilaginous differentiation, additional sampling is recommended in order to exclude a teratomatous background or a carcinosarcoma.

Few cases of pure extraskeletal osteosarcoma of the ovary have been reported. Patients range in age from 24 to 76 years (Lacoste et al. 2015). The tumor is frequently associated with extensive metastatic disease. Survival is poor. In one case metastatic tumor deposits affecting the abdominal cavity were excised at operation, and the patient was treated with triple chemotherapy consisting of cyclophosphamide, mitomycin C, and bleomycin (Hirakawa et al. 1988). The tumor recurred, and cisplatin and doxorubicin were added to the chemotherapeutic regimen. In spite of this, the tumor progressed and the patient died 8 months after diagnosis. Currently, only two patients have been reported as long term survivors (greater than 3 years disease free). Both received doxorubicin and cisplatin after complete surgical resection (Lacoste et al. 2015). Histologically, the tumors show typical appearances of osteosarcoma occurring in the skeleton. Although it was believed that the tumors originated directly from ovarian stroma, their histogenesis is uncertain. Occasional cases of osteosarcoma originating in ovarian teratoma have been recorded (Stowe and Watt 1952), but such cases should not be confused with pure ovarian osteosarcoma or with cases of carcinosarcoma with a prominent osteosarcomatous component.

Tumors of Neural Differentiation Tumors of Bone Differentiation Osteoma Few documented examples of osteoma occurring in the ovary exist. Although an origin from ovarian stroma is possible, most such lesions probably were examples of osseous metaplasia occurring in fibromas or leiomyomas, or possibly examples of metaplasia or heterotopia and not neoplasia occurring in the connective tissue of the ovary. Teratomatous origin is also possible. The lesions usually are small, but may be large, and are histologically composed of dense cortical bone.

Ovarian tumors originating from neural tissue are rare. The presenting symptoms usually are related to the presence of an intraabdominal mass. The tumors are solid and usually are small. The histogenesis is uncertain and probably is similar to that of other mesenchymal tumors of the ovary.

Neurofibroma Several cases of neurofibroma of the ovary have been reported in patients with generalized neurofibromatosis (von Recklinghausen’s disease)

1138

(Hegg and Flint 1990; Smith 1931; Protopapas et al. 2011). Histologically, the tumors resembled neurofibroma occurring elsewhere.

L. E. Schwartz and R. Vang

metastases, and the patient was well and diseasefree 1 year after diagnosis (Dover 1950). A malignant epithelioid schwannoma of the ovary has been reported (Laszlo et al. 2006).

Schwannoma (Neurilemmoma) Paraganglioma Three cases of ovarian schwannoma have been reported (Meyer 1943; Mishura 1963). In one case, the tumor was large (Mishura 1963). The tumors were solid, and the patients were well and disease-free after the excision of the tumor. Histologically, the tumors resembled schwannoma occurring in other locations.

Malignant Peripheral Nerve Sheath Tumor One case of “malignant neurilemmoma” (malignant schwannoma) of the ovary has been reported (Stone et al. 1986). The affected patient, a 71-year-old nulliparous woman, was admitted for evaluation of lower abdominal enlargement and pain. There were no stigmata of generalized neurofibromatosis. At laparotomy, a 15 cm, firm, somewhat hemorrhagic tumor was found arising from the left ovary. There were numerous tumor deposits involving the peritoneal cavity. A debulking procedure was performed, and the ovarian tumor was excised together with the omentum. Histologic and ultrastructural examinations revealed that the tumor was a malignant neurilemmoma. After surgery, the patient was treated with combination chemotherapy consisting of doxorubicin and cyclophosphamide, but the disease progressed and she died 5 months after surgery of extensive intraabdominal metastatic disease (Stone et al. 1986). One case of “neurofibrosarcoma” occurring in a 38-year-old woman with generalized neurofibromatosis (von Recklinghausen’s disease) has been described (Dover 1950). The tumor was an incidental finding and had replaced the ovary. It was solid, and histologically showed the typical appearance of a neurofibrosarcoma with a moderate degree of nuclear pleomorphism and mitotic activity. There was no evidence of

Paraganglioma (extra-adrenal pheochromocytoma) of the ovary is rare with very few cases reported in the literature. Several patients presented with hypertension. The lesions usually follow a benign course (Schuldt et al. 2015). In a report of three cases, patients ranged in age from 22 to 68 years (McCluggage and Young 2006); however a younger patient aged 15 was described in a separate report (Fawcett and Kimbell 1971). Tumors are characterized by the “zellballen” growth pattern seen in primary non-ovarian paragangliomas. With immunohistochemical stains, tumors are positive for neuroendocrine markers and negative for cytokeratin. Sustentacular cells can show expression of S-100. It should be noted that inhibin expression has been observed, which may create diagnostic confusion if a sex cord–stromal tumor, especially a Sertoli cell tumor, is in the differential diagnosis.

Ganglioneuroma A single case of ovarian ganglioneuroma occurring in a 4-year-old girl has been reported (Schmeisser and Anderson 1938). The child had abdominal enlargement. The tumor was solid, weighing 200 g, and replaced nearly the whole ovary. Histologically, the tumor was composed of well-differentiated ganglion cells. There was a recurrence after the excision of the tumor. True ganglioneuroma must be differentiated from teratomas showing prominence of ganglion cells and from proliferations of ganglion cells occasionally seen in the hilar region of the ovary; the latter are nonneoplastic and probably hamartomatous in nature. However, ganglioneuroma has been reported as arising from a teratoma (Coy et al. 2018).

17

Nonspecific Tumors of the Ovary, Including Mesenchymal Tumors

1139

Tumors of Adipose Tissue Differentiation Collections of adipose cells forming islands of fatty tissue that are not encapsulated are seen occasionally within ovarian tissue and are attributed to metaplasia of connective tissue of the ovary. These collections have been described as adipose prosoplasia. Benign adipose tissue seen in the ovary may be part of a teratoma with a prominent adipose tissue component. Pure benign fatty tumors in the ovary are very rare. A single case of a pure lipoma has been reported (Zwiesler et al. 2008). Malignant adipose tissue may be part of a carcinosarcoma with a prominent liposarcomatous component, or it may represent metastases from a liposarcoma occurring at another location. Rare myxoid liposarcomas have been reported in the ovary (Liang et al. 2015; Tirabosco et al. 2010).

Tumors of Mesothelial Differentiation Adenomatoid Tumor Adenomatoid tumor, which in females is found most frequently in the fallopian tubes and broad ligament and occasionally in the uterus near the serosal surface, is found only rarely in the ovary (see ▶ Chaps. 10, “Mesenchymal Tumors of the Uterus,” and ▶ 11, “Diseases of the Fallopian Tube and Paratubal Region”). Although its histogenesis was long disputed, it is now considered to be of mesothelial origin, as is supported by morphologic, histochemical, immunohistochemical, and ultrastructural observations. Adenomatoid tumors are benign and, therefore, are considered a benign mesothelioma. Few cases of ovarian adenomatoid tumor have been recorded, most of which occurred in patients in the third and fourth decades (Scully et al. 1998; Young et al. 1991; Phillips et al. 2007). The lesions, which are small, round or oval, and 0.5–3 cm in diameter, usually are found in the hilus of the ovary as incidental findings. In two cases the tumors were larger, measuring 6 and 8 cm in the longest diameter, respectively, and were symptomatic.

Fig. 16 Adenomatoid tumor. The tumor has numerous clefts and small round spaces lined by a single layer of flattened cells

Histologically, the tumors show similar appearances to adenomatoid tumors occurring in other locations and are composed of clefts and spaces lined by cuboidal, low columnar, or flattened epithelial-like cells (Fig. 16) and of solid aggregates of similar cells surrounded by connective tissue that varies from loose and edematous to dense and hyalinized. The epithelial-like cells may exhibit marked vacuolation. An oxyphilic variant has been described (Phillips et al. 2007). They exhibit positive staining with alcian blue, which is digestible with hyaluronidase, and similarly staining material is present in the clefts and spaces. Occasionally, the cells may show weak periodic acid -Schiff (PAS) staining. The tumor cells show strong positive immunohistochemical staining for low molecular weight cytokeratin, WT-1, calretinin, and D2-40. ER, PR, and Ber-EP4 are negative. Ultrastructural observations support the mesothelial origin of this lesion and show an abundance of microvilli, bundles of cytoplasmic filaments, tight junctional complexes, and intercellular spaces. The lesion is benign, and its excision results in a complete cure. The differential diagnosis for adenomatoid tumor in the ovary can be broad. Adenomatoid tumor may be confused with yolk sac tumor (YST) because the clefts and spaces may resemble the microcystic pattern of YST, but the nuclear appearances are totally different. The nuclei of adenomatoid tumor are bland and generally

1140

small and flattened, differing from the larger round and ovoid vesicular nuclei of YST, which exhibit brisk mitotic activity. The absence of other patterns associated with YST helps to distinguish adenomatoid tumor from YST. Lymphangioma may simulate adenomatoid tumor. Immunohistochemical studies can be helpful in differentiating between these two entities. Lymphangioma is low molecular weight cytokeratin negative, whereas adenomatoid tumor is strongly positive. Vascular markers such as factor VIII, CD34, and CD31 show negative reactions with adenomatoid tumor and positive staining with lymphangioma. Calretinin and WT-1 are also expressed in adenomatoid tumor but negative in lymphvascular tumors.

Peritoneal Malignant Mesothelioma Occasionally, peritoneal mesothelioma may involve the surface of the ovary (see ▶ Chap. 13, “Diseases of the Peritoneum”). When the tumor affects the ovary, confusion with primary ovarian neoplasms (serous borderline tumor with implants, low-grade serous carcinoma, and high-grade serous carcinoma) or benign conditions may occur (Talerman et al. 1985). The involvement of the ovary may be very extensive, and the presentation is that of a primary ovarian neoplasm. In one series of nine malignant peritoneal mesotheliomas presenting as ovarian masses, two tumors were considered as primary ovarian malignant mesotheliomas because the tumors were confined to the ovary (Scully et al. 1998; Clement et al. 1996). The histologic and immunohistochemical features, and distribution of the lesion are helpful in making the correct diagnosis (Scully et al. 1998; Talerman et al. 1985; Clement et al. 1996; Bollinger et al. 1989; Ordonez 1998; Vang and Ronnett 2009). Most patients with malignant peritoneal mesothelioma are middle-aged or elderly adults. Very rarely, it may occur in children (Talerman et al. 1985). The exact association with asbestos exposure is unclear. Additionally, malignant mesothelioma should be distinguished from well-differentiated papillary mesothelioma, which also can involve the ovary.

L. E. Schwartz and R. Vang

Wolffian Tumor (Formally Known as Female Adnexal Tumor of Probable Wolffian Origin (FATWO)) In the original report describing tumors of this type (Kariminejad and Scully 1973), all the tumors were located within the leaves of the broad ligament or were attached to it or to the fallopian tube; this also applied to subsequent reports dealing with this entity. Subsequently, a small number of ovarian tumors of probable wolffian origin were reported (DevouassouxShisheboran et al. 1999; Hughesdon 1982; Young and Scully 1983), indicating that tumors of this type also occur in the ovary. In the 2014 WHO Classification, female adnexal tumors of probable Wolffian origin are now classified as Wolffian tumors (Kurman 2014). The age of most patients ranges from 28 to 79 years. Some patients have abdominal enlargement, and in other patients, the tumor is found on physical examination (Hughesdon 1982; Young and Scully 1983). All the tumors are unilateral. In most cases, they are confined to the ovary, but in one case metastatic deposits in the abdominal cavity were reported. In the latter case, the tumor contained foci of undifferentiated carcinoma (Young and Scully 1983). Most tumors range in size from 2 to 20 cm in the largest diameter. They are smooth and often lobulated and are either solid or solid and cystic. The cysts vary in size and may range up to 11 cm (Young and Scully 1983). Microscopically, the tumor is composed of relatively uniform epithelial cells that line cysts and tubules, sometimes forming a sieve-like pattern. The tumor cells also may form closely packed tubules, grow in a diffuse pattern, or fill tubules or tubular spaces (Fig. 17). They have uniform round or oval nuclei, and there is low mitotic activity. The tumor cells do not contain mucin but occasionally may contain glycogen. The amount of intervening connective tissue varies from imperceptible to considerable, forming fibrous bands separating the islands of tumor cells and producing a lobular pattern (Young and Scully 1983). In two patients in whom the tumors were associated with aggressive behavior, there

17

Nonspecific Tumors of the Ovary, Including Mesenchymal Tumors

1141

papillary pattern, and intraluminal and intracellular mucin.

Lesions of the Rete Ovarii

Fig. 17 Ovarian tumor of probable wolffian origin. The tumor is composed of closely packed tubules and clefts with adjacent solid patterns containing spindle cells

was brisk mitotic activity with 10 or more mitoses (MF)/10 HPF, and in one of these patients, there was nuclear pleomorphism. Details of immunohistochemistry can be found in ▶ Chap. 11, “Diseases of the Fallopian Tube and Paratubal Region.” Two patients have subsequently developed metastases (Young and Scully 1983). Eight patients were known to be alive and diseasefree from 1 to 15 years postoperatively, and one was lost to follow-up (Young and Scully 1983), indicating that in most cases the tumor is not associated with an aggressive course. It is also of note that there is a good correlation between the mitotic activity and the behavior of this neoplasm. Wolffian tumors may be confused with sex cord–stromal tumors, especially various types of Sertoli–Leydig cell tumors and surface epithelial–stromal tumors (see ▶ Chap. 15, “Sex Cord-Stromal, Steroid Cell, and Other Ovarian Tumors with Endocrine, Paraendocrine, and Paraneoplastic Manifestations”). The presence of the typical features of this tumor described here and the absence of the various patterns observed in Sertoli–Leydig cell tumors differentiate it from the latter. Wolffian tumors with a prominent spindle cell component may mimic cellular fibroma (Fanghong et al. 2008). The Wolffian tumor is distinguished from the various surface epithelial–stromal tumors of the ovary by the absence of cellular and nuclear pleomorphism,

Rete cysts/cystadenomas are uncommon, although probably more common than realized. The average age of patients with tumors of the rete ovarii is 59 years (range, 23–80 years). Patients can present with abdominal discomfort, pelvic pressure, virilization, postmenopausal bleeding, and/or hirsutism (Rutgers and Scully 1988). Most tumors are unilateral, and the average size is 9 cm. They may be uni- or multicystic, and the internal lining is usually smooth. The rete lesions range in histologic type and include cyst, cystadenoma, adenoma, adenomatous hyperplasia, and adenocarcinoma (Rutgers and Scully 1988; Heatley 2000; Nogales et al. 1997). Most are cysts/cystadenomas. Rete lesions are either found within the ovarian hilus or are anatomically related to the rete ovarii. The distinction between cyst and cystadenoma is arbitrary, but an upper limit of 1 cm for cysts has been proposed. The epithelial lining of the rete cyst/cystadenoma is simple, non-ciliated, and bland and characterized by irregular crevices (Fig. 18). The periphery of rete adenoma is well circumscribed, and the tumor contains crowded tubules and papillae. The tubules and papillae have a simple layer of bland epithelial cells. Adenomatous hyperplasia is histologically similar to adenoma; however, the former is not well circumscribed. Rete adenocarcinoma is rare and ill-defined. Only one well-documented case of adenocarcinoma of rete ovarii occurring in a 52-year-old woman with abdominal enlargement and ascites has been reported (Rutgers and Scully 1988). The patient had bilateral, partly solid and partly cystic tumors without specific macroscopic features. The tumor showed a predominant pattern of branching tubules and cysts containing simple papillae with fibrovascular or hyalinized cores. Some cysts contained eosinophilic material. Focally the tumor showed a solid tubular pattern. The cells lining the tubules and papillae were

1142

L. E. Schwartz and R. Vang

Wolffian tumors and sex cord-stromal tumors; however, morphology is of the utmost importance especially with relation to lesions of mullerian origin (Goyal et al. 2016), and the immunophenotype of Wolffian tumors can be non-specific in some cases.

Primary Ovarian Tumors of Uncertain Histogenesis Hepatoid Carcinoma Fig. 18 Rete cyst with cuboidal epithelium

cuboidal, nonciliated, and atypical. Focally, they were multilayered and stratified. There was brisk mitotic activity. Adenocarcinoma of rete ovarii can only be diagnosed if the tumor is small enough to appreciate a hilar location and is composed of collections of slit-like retiform tubules and cysts containing papillae lined by cells similar to those of normal rete ovarii (Scully et al. 1998; Rutgers and Scully 1988). Rete cysts/cystadenomas are benign. The data on adenoma and adenocarcinoma is limited; thus, the behavior of these lesions is uncertain. The most common lesion to histologically mimic rete cystadenoma is serous cystadenoma, but this is not a clinically important distinction. Rete cystadenoma is favored based on location in the hilus or anatomic connection with the rete ovarii, cyst lining with crevice-like contours, absence of ciliated cells, and smooth muscle and hilus cells within the cyst wall. Confusion may occur with retiform Sertoli–Leydig cell tumor, but the latter is likely to show other patterns of Sertoli–Leydig cell tumor and stain positively for inhibin and SF-1. Some small serous carcinomas may also resemble adenocarcinoma of the rete ovarii, but they tend to be found in a cortical location, generally do not show the fine slit-like papillae, and exhibit much greater nuclear pleomorphism. Immunohistochemically, rete ovarii are positive for PAX-8, while negative for PAX-2 and GATA-3. SF-1 is noted to show weak diffuse nuclear staining. These stains may help to distinguish rete ovarii lesions from

In 1987, Ishikura and Scully (1987) described five cases of ovarian carcinoma with hepatoid features, three of them primary and two probably primary. Since then, the number of cases reported in the literature still remains less than 50. The age of reported patients ranges from 35 to 78 years (Randolph et al. 2015) and, thus, differs considerably from patients with YST with a hepatoid pattern, which is usually seen in children, adolescents, and young women, although yolk sac tumor can uncommonly arise in postmenopausal women. The age range as well as the histologic appearances of the tumors showed considerable similarity to gastric carcinomas with hepatic features described some years earlier (Ishikura et al. 1986). Unlike YST with a hepatoid pattern, which may be pure, mixed with other YST patterns, or combined with other germ cell tumors, hepatoid carcinoma of the ovary occurs in pure form, although occasionally it is associated with serous adenocarcinoma or other types of ovarian surface epithelial–stromal tumors (Scurry et al. 1996, 1998; Pitman et al. 2004; Tochigi et al. 2003). Hepatoid carcinoma of the ovary (Ishikura and Scully 1987; Pitman et al. 2004; Tochigi et al. 2003)], like YST with hepatoid pattern and gastric adenocarcinoma with hepatoid features (Ishikura et al. 1986), is associated with alpha-fetoprotein (AFP) secretion, and AFP can be demonstrated within the tumor cells by immunohistochemical techniques. In several cases, high levels of serum AFP were noted, and serum AFP was used to monitor the disease activity. Serum CA-125 also appears to be increased in patients (Randolph et al. 2015).

17

Nonspecific Tumors of the Ovary, Including Mesenchymal Tumors

1143

Fig. 19 Hepatoid carcinoma of the ovary. Note the close resemblance to hepatocellular carcinoma

Fig. 20 Hepatoid carcinoma of the ovary showing expression of AFP

Clinically, patients present with symptoms and signs related to the presence of an adnexal mass. Abdominal enlargement, which may be associated with pain, malaise, and weight loss, is the main presenting sign (Ishikura and Scully 1987). In several cases the tumor was bilateral at presentation (Randolph et al. 2015). Hepatoid carcinomas of the ovary are large and are associated with metastatic tumor deposits within the abdominal cavity (stage III) in most cases (Ishikura and Scully 1987). Histologically, the tumor shows a close resemblance to hepatocellular carcinoma (Fig. 19), and is composed of solid sheets or aggregates of uniform cells with moderate or abundant eosinophilic cytoplasm, distinct cell borders, and centrally located nuclei with prominent nucleoli (Ishikura and Scully 1987; Pitman et al. 2004; Tochigi et al. 2003). Mitotic activity generally is brisk, and abnormal forms are seen. In some parts of the tumor there may be a considerable degree of nuclear pleomorphism, and multinucleated giant cells may be seen (Ishikura and Scully 1987). PAS-positive diastase-resistant hyaline globules may be seen, and glycogen can be demonstrated within the cytoplasm of the tumor cells. Histologic patterns seen in germ cell tumors or surface epithelial–stromal tumors are not detectable when the tumor is seen in pure form (Scully et al. 1998; Ishikura and Scully 1987). Immunohistochemical studies demonstrate the presence of AFP and hepatocyte paraffin

1 (Pitman et al. 2004) in a considerable number of tumor cells (Fig. 20). In addition, the tumor cells are immunoreactive for albumin, alpha-1antitrypsin, and alpha-1-antichymotrypsin. Focal positive immunostaining for carcinoembryonic antigen (CEA) also is seen (Ishikura and Scully 1987). Hepatoid carcinoma of the ovary is a highly malignant neoplasm (Scully et al. 1998; Ishikura and Scully 1987). The histogenesis of hepatoid carcinoma of the ovary has not been established. Unlike YST with hepatoid pattern, it is thought not to be of germ cell origin, as it occurs in older patients, is not associated with other neoplastic germ cell elements, and is not found in patients with gonadal dysgenesis. Because of the age distribution and the occasional association with ovarian surface epithelial–stromal tumors, it is likely that it is a metaplastic tumor and represents a variant of a surface epithelial–stromal tumor (Scurry et al. 1996, 1998; Ishikura and Scully 1987; Randolph et al. 2015; Pitman et al. 2004; Tochigi et al. 2003). Hepatoid carcinoma of the ovary must be distinguished from YST with hepatoid pattern. It can be distinguished clinically by its occurrence in older, usually postmenopausal patients (although some YST also arise in postmenopausal women), and by its presentation in a more advanced clinical stage, usually stage III. Histologically, hepatoid carcinoma shows a greater degree of cellular and nuclear

1144

L. E. Schwartz and R. Vang

Eichhorn et al. (1992) have reported 11 primary ovarian tumors that resembled small cell carcinoma of the lung and differed both clinically and histologically from primary small cell carcinoma of the ovary of hypercalcemic type (Young et al. 1994). The age of the patients ranged from 28 to 85 years (Eichhorn et al. 1992). Most patients presented with abdominal enlargement. Six of

the tumors were unilateral, and five were bilateral. Spread beyond the ovary was noted in seven tumors. None of the patients had distant metastases at presentation (Eichhorn et al. 1992). The tumors measured from 4.5 to 26 cm in the greatest dimension; they were mostly solid with a variable minor cystic component (Eichhorn et al. 1992). Histologically, the tumor is composed of small to medium-sized round to spindle-shaped cells with scanty cytoplasm, hyperchromatic nuclei, and inconspicuous nucleoli forming sheets, large aggregates, and closely packed nests (Fig. 21). Sometimes an insular or a trabecular pattern was seen (Eichhorn et al. 1992). In four tumors a component of endometrioid carcinoma was present, one tumor showed focal squamous differentiation, two tumors were associated with Brenner tumor, and one contained a cyst lined by atypical mucinous cells (Young and Scully 1984). In two of six tumors, argyrophil granules were demonstrated. In nine cases, immunohistochemical studies were performed that demonstrated positive staining for cytokeratin in six cases, epithelial membrane antigen (EMA) in five, and chromogranin in two. All nine tumors were vimentin-negative. In a small number of cases evaluated, perinuclear dot-like staining for CK20 and variable TTF-1 expression have been observed (Carlson et al. 2007; Rund and Fischer 2006). Flow cytometric studies performed on eight tumors showed that five tumors were

Fig. 21 Small cell carcinoma, pulmonary type. (a) Solid growth pattern containing abundant geographic necrosis. (b) Small to medium-sized round cells with

hyperchromatic nuclei, absence of nucleoli, scant cytoplasm, and numerous apoptotic bodies and MF. (Case courtesy of Dr. Robert H. Young, Boston, MA)

pleomorphism, and tumor giant cells are much more frequently seen. In most cases demonstration of positive immunocytochemical staining for AFP in the tumor cells and elevated levels of serum AFP differentiate hepatoid carcinoma from other ovarian tumors such as undifferentiated adenocarcinomas, endometrioid adenocarcinomas with marked squamous differentiation, and steroid cell tumors (Ishikura and Scully 1987). One case of hepatoid carcinoma without AFP staining has been reported (Sung et al. 2013). Primary hepatoid carcinoma of the ovary also must be differentiated from hepatocellular carcinoma metastatic to the ovary (Young et al. 1992). Although the latter is uncommon, this possibility must be carefully excluded before the diagnosis of primary hepatoid carcinoma of the ovary is made.

Small Cell Carcinoma, Pulmonary Type (Neuroendocrine Type)

17

Nonspecific Tumors of the Ovary, Including Mesenchymal Tumors

aneuploid and three were diploid (Eichhorn et al. 1992). It is important to note that tumors can be diagnosed in the absence of neuroendocrine maker positivity if morphology is typical. The tumors were aggressive, and of the nine patients with known follow-up, five died of the disease 1–13 months after diagnosis, one died after an unknown interval, and two had recurrent disease 6 and 8 months after surgery. Five of the patients with stage III tumors and two with stage I tumors were treated with combination chemotherapy, which included cisplatin in all cases and doxorubicin in most cases; one of these treated patients was alive at 7.5 years (Eichhorn et al. 1992). Aggressive treatment with agents effective in treating small cell pulmonary carcinoma appears to be the treatment of choice. Primary ovarian small cell carcinoma (pulmonary type) must be distinguished from pulmonary small cell carcinoma metastatic to the ovary, which shows both clinical and pathologic differences (Young and Scully 1985; Irving and Young 2005). It also must be differentiated from primary ovarian small cell carcinoma of hypercalcemic type (Scully et al. 1998; Young et al. 1994) (see ▶ Chap. 15, “Sex Cord-Stromal, Steroid Cell, and Other Ovarian Tumors with Endocrine, Paraendocrine, and Paraneoplastic Manifestations”). The patients with primary ovarian small cell carcinoma of pulmonary type are older. The tumor is seen either in perimenopausal or postmenopausal women (Scully et al. 1998; Eichhorn et al. 1992). Hypercalcemia is absent. The tumors are bilateral in 45% of cases, whereas in the hypercalcemic type, bilaterality is seen only rarely (1% of cases) (Eichhorn et al. 1992). Histologically, the cells of primary ovarian small cell carcinoma of pulmonary type differ from those of the hypercalcemic type in having finely dispersed chromatin and inconspicuous nucleoli, whereas the latter is composed of cells with nuclei showing clumped chromatin and prominent nucleoli, as well as showing the presence of larger cells with abundant eosinophilic cytoplasm in 40% of cases (Eichhorn et al. 1992). Follicle-like spaces are frequently seen in the hypercalcemic type and are virtually absent in the pulmonary type of small cell carcinoma

1145

(Eichhorn et al. 1992). It is important to recognize that small cell carcinoma of pulmonary type is a neuroendocrine carcinoma whereas the hypercalcemic type is not. Immunohistochemically, the pulmonary type usually but not always expresses chromogranin and synaptophysin while the hypercalcemic type does not express these markers. Conversely, the hypercalcemic type shows complete loss of BRG1. Although data are limited, it would be expected that the pulmonary type retains expression of this marker. Endometrioid and Brenner tumor components are present in more than half the small cell carcinomas with pulmonary differentiation and are absent in the hypercalcemic type. The former also tend to be more frequently aneuploid (Eichhorn et al. 1992) Although the histogenesis of the primary ovarian small cell carcinoma of pulmonary type has not been established, the frequent association with endometrioid and Brenner tumors points toward a surface epithelial–stromal origin, as is supported further by the age range of the patients (Eichhorn et al. 1992).

Neuroendocrine Carcinoma, Non-Small Cell Type A small number of ovarian tumors composed of solid sheets, nests, cords, trabeculae, or solid islands of cells showing neuroendocrine differentiation have been reported (Scully et al. 1998; Eichhorn et al. 1996; Veras et al. 2007). These tumors can be associated with either a surface epithelial-stromal or possibly germ cell component (Collins et al. 1991; Agarwal et al. 2016). The age of reported patients ranges from 22 to 77 years. Some tumors were stage I, but in spite of this the prognosis was poor. Several cases presented with advanced stage disease, and the tumors demonstrated aggressive behavior, in common with neuroendocrine tumors occurring in other sites. Histologically, the neuroendocrine component of the tumor consists of solid islands or cords of medium to large epithelial cells with variable amount of cytoplasm and large nuclei, some of which have prominent nucleoli. Mitotic activity

1146

is variable but frequently high. The cellular islands and cords are surrounded only by a small amount of connective tissue. Immunohistochemical stains are frequently positive for pancytokeratin, CK7, CK20, chromogranin, and synaptophysin (Veras et al. 2007) Other neurohormonal polypeptides are also detected in some of the tumors (Eichhorn et al. 1996). The neuroendocrine component of the tumor may resemble insular carcinoid tumor of the ovary, but the cells are usually larger and show much greater degree of cellular and nuclear pleomorphism. The presence of the surface epithelial–stromal component also helps to differentiate between these two entities. The distinction between them is very important because the prognosis of patients with neuroendocrine carcinoma is by far worse than that of patients with carcinoid tumors. The size of the tumor cells, proliferative index, and the strong positive immunohistochemical reactions distinguish this tumor from the ovarian carcinoma of the small cell pulmonary type. Another tumor in the differential is adult granulosa cell tumor. Immunohistochemistry should help to resolve this differential as neuroendocrine carcinoma should be negative for inhibin and calretinin (Agarwal et al. 2016), as well as SF-1. The presence of the surface epithelial–stromal component confirming the ovarian origin differentiates this tumor from metastatic small cell tumors to the ovary (Eichhorn et al. 1996).

Salivary Gland-Like Carcinomas Ovarian tumors resembling salivary gland carcinomas are rare, but a series of 12 tumors has been reported (Eichhorn and Scully 1995). The tumors often resemble adenoid cystic carcinoma. Most of the tumors also exhibite a minor component of surface epithelial–stromal neoplasia. The latter can be of various histologic types and include serous, endometrioid, and clear cell carcinomas. The affected patients are elderly, with nearly all in the seventh or eighth decade. Most of the tumors were associated with extensive metastatic disease

L. E. Schwartz and R. Vang

Fig. 22 Ovarian carcinoma resembling adenoid cystic carcinoma. (Case courtesy of Dr. Robert H. Young, Boston, MA)

and the prognosis was poor, except for one case where the tumor was in pure form and another in which the associated surface epithelial–stromal component was of the serous borderline type. Histologically, the tumors show architectural patterns seen in adenoid cystic carcinoma of the salivary glands (Fig. 22). The tumor cells resemble myoepithelial cells, although this has not been confirmed immunohistochemically because in the majority of cases the cells did not stain positively for actin and S-100 protein (Scully et al. 1998; Eichhorn and Scully 1995). Histogenetically, these tumors are probably of surface epithelial–stromal origin because they are usually associated with a surface epithelial–stromal component or recurred as an adenoid cystic carcinoma in a patient for whom the original tumor was an endometrioid carcinoma. Not uncommonly, the tumors show basaloid or ameloblastomatous features (Scully et al. 1998; Eichhorn and Scully 1995). The age distribution for tumors with these features is wide, ranging from 19 to 65 years. Most of the tumors were confined to the ovary (stage IA), and the prognosis was excellent after excision of the tumor, although some of the follow-up periods were relatively short, varying from 16 to 71 months. Histologically, the tumors show either a basaloid or ameloblastomatous pattern (Fig. 23). Focal squamous and glandular differentiation can be seen, as can a minor endometrioid carcinoma component

17

Nonspecific Tumors of the Ovary, Including Mesenchymal Tumors

1147

in differentiating between these two entities. Demonstration of positive inhibin and SF-1 staining using immunohistochemistry further supports the diagnosis of Sertoli–Leydig cell tumor.

References

Fig. 23 Ovarian carcinoma with basaloid appearance. Note peripheral palisading at interface between neoplastic epithelium and fibrovascular cores. (Case courtesy of Dr. Robert H. Young, Boston, MA)

(Scully et al. 1998; Eichhorn and Scully 1995). The histogenesis of this neoplasm is uncertain, but surface epithelial–stromal origin, particularly endometrioid carcinoma, appears to be most likely.

Nephroblastoma (Wilms Tumor) Less than 10 cases of ovarian nephroblastoma have been reported. Patients range in age from 1 year to 36 years. Pure nephroblastomas of the ovary as well as nephroblastomas arising in conjunction with an ovarian teratoma have been documented (Alexander et al. 2017). The tumors show typical features of well-differentiated nephroblastoma with glomeruloid formations, small tubules, and prominent blastema (Alexander et al. 2017; Sahin and Benda 1988). Although the tumors were described as primary ovarian nephroblastomas, their histogenetic origin is uncertain. It is thought that these tumors either arise from mesonephric remnants or teratomas (Alexander et al. 2017). Occasionally, retiform Sertoli–Leydig cell tumors, because of the presence of tubules and papillary pattern resembling glomeruloid formations, have been misdiagnosed as ovarian nephroblastomas. Careful sectioning and examination of the tumor for the presence of other patterns associated with Sertoli–Leydig cell tumors and absence of renal blastema are helpful

Agarwal L, Gupta B, Jain A (2016) Pure large cell neuroendocrine carcinoma of the ovary with metastasis to cervix: a rare case report and review of literature. J Clin Diagn Res 10(9):ED01–ED03 Alexander VM et al (2017) Wilms’ tumor of the ovary. Gynecol Oncol Rep 19:18–21 Amador-Ortiz C et al (2011) JAZF1 and JJAZ1 gene fusion in primary extrauterine endometrial stromal sarcoma. Hum Pathol 42(7):939–946 Balaton A et al (1987) Primary leiomyosarcoma of the ovary: a histological and immunocytochemical study. Gynecol Oncol 28(1):116–120 Bollinger DJ et al (1989) Peritoneal malignant mesothelioma versus serous papillary adenocarcinoma. A histochemical and immunohistochemical comparison. Am J Surg Pathol 13(8):659–670 Brady K et al (1987) Ovarian myxoma. Am J Obstet Gynecol 156(5):1240–1242 Carlson JW et al (2007) Biomarker-assisted diagnosis of ovarian, cervical and pulmonary small cell carcinomas: the role of TTF-1, WT-1 and HPV analysis. Histopathology 51(3):305–312 Chan YF, Leung CS, Ma L (1989) Primary embryonal rhabdomyosarcoma of the ovary in a 4-year-old girl. Histopathology 15(3):309–311 Chang KL et al (1993) Primary extrauterine endometrial stromal neoplasms: a clinicopathologic study of 20 cases and a review of the literature. Int J Gynecol Pathol 12(4):282–296 Chiang S et al (2011) Frequency of known gene rearrangements in endometrial stromal tumors. Am J Surg Pathol 35(9):1364–1372 Clement PB, Young RH, Scully RE (1996) Malignant mesotheliomas presenting as ovarian masses. A report of nine cases, including two primary ovarian mesotheliomas. Am J Surg Pathol 20(9):1067–1080 Climie AR, Heath LP (1968) Malignant degeneration of benign cystic teratomas of the ovary. Review of the literature and report of a chondrosarcoma and carcinoid tumor. Cancer 22(4):824–832 Collins RJ et al (1991) Primary mixed neuroendocrine and mucinous carcinoma of the ovary. Arch Gynecol Obstet 248(3):139–143 Costa MJ et al (1993) Histologic and immunohistochemical evidence for considering ovarian myxoma as a variant of the thecoma-fibroma group of ovarian stromal tumors. Arch Pathol Lab Med 117(8):802–808

1148 Coy S et al (2018) De novo tumors of teratoma: ganglioneuroma arising from a mature cystic teratoma of the ovary. Int J Gynecol Pathol 37(3):296–300. https://doi.org/10.1097/PGP.0000000000000409 de Kock L et al (2015) Ovarian embryonal rhabdomyosarcoma is a rare manifestation of the DICER1 syndrome. Hum Pathol 46(6):917–922 Deshpande V, Oliva E, Young RH (2010) Solid pseudopapillary neoplasm of the ovary: a report of 3 primary ovarian tumors resembling those of the pancreas. Am J Surg Pathol 34(10):1514–1520. Devouassoux-Shisheboran M, Silver SA, Tavassoli FA (1999) Wolffian adnexal tumor, so-called female adnexal tumor of probable Wolffian origin (FATWO): immunohistochemical evidence in support of a Wolffian origin. Hum Pathol 30(7):856–863 Doss BJ et al (1999) Ovarian leiomyomas: clinicopathologic features in fifteen cases. Int J Gynecol Pathol 18(1):63–68 Dover H (1950) Neurofibrosarcoma of the ovary associated with neurofibromatosis. Can Med Assoc J 63(5):488–490 Dundr P et al (2017) Anastomosing hemangioma of the ovary: a clinicopathological study of six cases with stromal luteinization. Pathol Oncol Res 23(4):717–722 Dutz W, Stout AP (1961) The myxoma in childhood. Cancer 14:629–635 Eichhorn JH, Scully RE (1991) Ovarian myxoma: clinicopathologic and immunocytologic analysis of five cases and a review of the literature. Int J Gynecol Pathol 10(2):156–169 Eichhorn JH, Scully RE (1995) “Adenoid cystic” and basaloid carcinomas of the ovary: evidence for a surface epithelial lineage. A report of 12 cases. Mod Pathol 8(7):731–740 Eichhorn JH, Young RH, Scully RE (1992) Primary ovarian small cell carcinoma of pulmonary type. A clinicopathologic, immunohistologic, and flow cytometric analysis of 11 cases. Am J Surg Pathol 16(10):926–938 Eichhorn JH et al (1996) Ovarian neuroendocrine carcinomas of non-small-cell type associated with surface epithelial adenocarcinomas. A study of five cases and review of the literature. Int J Gynecol Pathol 15(4):303–314 Fanghong L, Szallasi A, Young RH (2008) Wolffian tumor of the ovary with a prominent spindle cell component: report of a case with brief discussion of unusual problems in differential diagnosis, and literature review. Int J Surg Pathol 16(2):222–225 Fawcett FJ, Kimbell NK (1971) Phaeochromocytoma of the ovary. J Obstet Gynaecol Br Commonw 78(5):458–459 Gehrig PA, Fowler WC Jr, Lininger RA (2000) Ovarian capillary hemangioma presenting as an adnexal mass with massive ascites and elevated CA-125. Gynecol Oncol 76(1):130–132 Goyal A, Masand RP, Roma AA (2016) Value of PAX-8 and SF-1 immunohistochemistry in the distinction between

L. E. Schwartz and R. Vang female adnexal tumor of probable wolffian origin and its mimics. Int J Gynecol Pathol 35(2):167–175 Heatley MK (2000) Adenomatous hyperplasia of the rete ovarii. Histopathology 36(4):383–384 Hegg CA, Flint A (1990) Neurofibroma of the ovary. Gynecol Oncol 37(3):437–438 Hirakawa T et al (1988) Ovarian sarcoma with histologic features of telangiectatic osteosarcoma of the bone. Am J Surg Pathol 12(7):567–572 Huang TY, Chen JT, Ho WL (2005) Ovarian serous cystadenoma with mural nodules of genital rhabdomyoma. Hum Pathol 36(4):433–435 Hughesdon PE (1982) Ovarian tumours of Wolffian or allied nature: their place in ovarian oncology. J Clin Pathol 35(5):526–535 Irving JA, Young RH (2005) Lung carcinoma metastatic to the ovary: a clinicopathologic study of 32 cases emphasizing their morphologic spectrum and problems in differential diagnosis. Am J Surg Pathol 29(8): 997–1006 Ishikura H, Scully RE (1987) Hepatoid carcinoma of the ovary. A newly described tumor. Cancer 60(11): 2775–2784 Ishikura H et al (1986) Hepatoid adenocarcinomas of the stomach. An analysis of seven cases. Cancer 58(1):119–126 Kalstone CE, Jaffe RB, Abell MR (1969) Massive edema of the ovary simulating fibroma. Obstet Gynecol 34(4):564–571 Kandalaft PL, Esteban JM (1992) Bilateral massive ovarian leiomyomata in a young woman: a case report with review of the literature. Mod Pathol 5(5):586–589 Kariminejad MH, Scully RE (1973) Female adnexal tumor of probable Wolffian origin. A distinctive pathologic entity. Cancer 31(3):671–677 Kruse AJ et al (2014) Angiosarcomas of primary gynecologic origin: a clinicopathologic review and quantitative analysis of survival. Int J Gynecol Cancer 24(1):4–12 Kryvenko ON et al (2011) Anastomosing hemangioma of the genitourinary system: eight cases in the kidney and ovary with immunohistochemical and ultrastructural analysis. Am J Clin Pathol 136(3): 450–457 Lacoste C et al (2015) Primary osteosarcoma of the ovary. Gynecol Obstet Fertil 43(7–8):555–556 Laszlo A, Ivaskevics K, Sapi Z (2006) Malignant epithelioid ovarian schwannoma: a case report. Int J Gynecol Cancer 16(Suppl 1):360–362 Lawhead RA, Copeland LJ, Edwards CL (1985) Bilateral ovarian hemangiomas associated with diffuse abdominopelvic hemangiomatosis. Obstet Gynecol 65(4):597–599 Lerwill MF et al (2004) Smooth muscle tumors of the ovary: a clinicopathologic study of 54 cases emphasizing prognostic criteria, histologic variants, and differential diagnosis. Am J Surg Pathol 28(11): 1436–1451

17

Nonspecific Tumors of the Ovary, Including Mesenchymal Tumors

Liang SX et al (2015) Primary myxoid liposarcoma of the ovary in a postpartum female: a case report and review of literature. Int J Gynecol Pathol 34(3):298–302 Mann LS, Metrick S (1961) Hemangioma of the ovary. Report of a case. J Int Coll Surg 36:500–502 Masand RP et al (2013) Endometrioid stromal sarcoma: a clinicopathologic study of 63 cases. Am J Surg Pathol 37(11):1635–1647 Mc BR, Trumbull M (1955) Hemangioma of the ovary with ascites. Miss Doct 32(10):271–274 McCluggage WG, Young RH (2006) Paraganglioma of the ovary: report of three cases of a rare ovarian neoplasm, including two exhibiting inhibin positivity. Am J Surg Pathol 30(5):600–605 Meyer R (1943) Nerve tumors of the female genitals and pelvis. Arch Pathol 36:437–464 Mira JL (1991) Lipoleiomyoma of the ovary: report of a case and review of the English literature. Int J Gynecol Pathol 10(2):198–202 Mishura VI (1963) Report of large benign tumor – report of three cases. Vopr Onkol 9:103 Nielsen GP et al (1997) Primary angiosarcoma of the ovary: a report of seven cases and review of the literature. Int J Gynecol Pathol 16(4):378–382 Nielsen GP et al (1998) Primary ovarian rhabdomyosarcoma: a report of 13 cases. Int J Gynecol Pathol 17(2):113–119 Nogales FF (1982) Primary chondroma of the ovary. Histopathology 6:376 Nogales FF et al (1997) Adenomas of the rete ovarii. Hum Pathol 28(12):1428–1433 Nucci MR et al (1998) Angiosarcoma of the ovary: clinicopathologic and immunohistochemical analysis of four cases with a broad morphologic spectrum. Am J Surg Pathol 22(5):620–630 Nunez C et al (1983) Ovarian rhabdomyosarcoma presenting as leukemia. Case report. Cancer 52(2):297–300 Oliva E, Egger JF, Young RH (2014) Primary endometrioid stromal sarcoma of the ovary: a clinicopathologic study of 27 cases with morphologic and behavioral features similar to those of uterine low-grade endometrial stromal sarcoma. Am J Surg Pathol 38(3):305–315 Ordonez NG (1998) Role of immunohistochemistry in distinguishing epithelial peritoneal mesotheliomas from peritoneal and ovarian serous carcinomas. Am J Surg Pathol 22(10):1203–1214 Phillips V, McCluggage WG, Young RH (2007) Oxyphilic adenomatoid tumor of the ovary: a case report with discussion of the differential diagnosis of ovarian tumors with vacuoles and related spaces. Int J Gynecol Pathol 26(1):16–20 Pitman MB et al (2004) Hepatocyte paraffin 1 antibody does not distinguish primary ovarian tumors with hepatoid differentiation from metastatic hepatocellular carcinoma. Int J Gynecol Pathol 23(1):58–64 Prayson RA, Hart WR (1992) Primary smooth-muscle tumors of the ovary. A clinicopathologic study of four

1149

leiomyomas and two mitotically active leiomyomas. Arch Pathol Lab Med 116(10):1068–1071 Protopapas A et al (2011) Ovarian neurofibroma: a rare visceral occurrence of type 1 neurofibromatosis and an unusual cause of chronic pelvic pain. J Minim Invasive Gynecol 18(4):520–524 Kurman RJ, Carcangiu ML, Simon Herrington C, Young RH (2014) WHO classification of tumours of female reproductive organs, 4th edn. IARC Press, Lyon Radhouane A, Mayada S, Khaled N (2016) Lymphangioma of the ovary: etiology and management. Eur J Obstet Gynecol Reprod Biol 203:342–343 Randolph LK et al (2015) Hepatoid carcinoma of the ovary: a case report and review of the literature. Gynecol Oncol Rep 13:64–67 Rhoades CP, McMahon JT, Goldblum JR (1999) Myofibroblastoma of the ovary: report of a case. Mod Pathol 12(9):907–911 Roth LM, Gaba AR, Cheng L (2013) The pathogenesis of ovarian myxoma: a neoplasm sometimes arising from other ovarian stromal tumors. Int J Gynecol Pathol 32(4):368–378 Rund CR, Fischer EG (2006) Perinuclear dot-like cytokeratin 20 staining in small cell neuroendocrine carcinoma of the ovary (pulmonary-type). Appl Immunohistochem Mol Morphol 14(2):244–248 Rutgers JL, Scully RE (1988) Cysts (cystadenomas) and tumors of the rete ovarii. Int J Gynecol Pathol 7(4):330–342 Sahin A, Benda JA (1988) Primary ovarian Wilms’ tumor. Cancer 61(7):1460–1463 Sandison AT (1955) Rhabdomyosarcoma of the ovary. J Pathol Bacteriol 70(2):433–438 Savargaonkar PR et al (1994) Ovarian haemangiomas and stromal luteinization. Histopathology 25(2):185–188 Schmeisser HC, Anderson W (1938) Ganglioneuroma of the ovary. J Am Med Assoc 111:2005–2007 Schoolmeester JK et al (2015) Ovarian hemangiomas do not harbor EWSR1 rearrangements: clinicopathologic characterization of 10 cases. Int J Gynecol Pathol 34(5):437–444 Schuldt M et al (2015) Ovarian paraganglioma. Int J Surg Pathol 23(2):130–133 Scully RE, Young RH, Clement PB (1998) Tumors of the ovary, maldeveloped gonads, fallopian tube, and broad ligament, 3rd ser, vol 23. Armed Forces Institute of Pathology, Washington, DC Scurry JP, Brown RW, Jobling T (1996) Combined ovarian serous papillary and hepatoid carcinoma. Gynecol Oncol 63(1):138–142 Shaffer MD, Cancelmo JJ (1939) Cavernous hemangioma of the ovary in a girl twelve years of age. Am J Obstet Gynecol 38:722–723 Singer T et al (2010) Rare case of ovarian cystic lymphangioma. J Minim Invasive Gynecol 17(1):97–99 Smith FR (1931) Neurofibroma of the ovary associated with Recklinghausen’s disease. Am J Cancer 15:859–862

1150 Stewart CJ, Charles A, Foulkes WD (2016) Gynecologic manifestations of the DICER1 syndrome. Surg Pathol Clin 9(2):227–241 Stone GC et al (1986) Malignant schwannoma of the ovary. Report of a case. Cancer 58(7):1575–1582 Stowe LM, Watt JY (1952) Osteogenic sarcoma of the ovary. Am J Obstet Gynecol 64(2):422–426 Sung JH et al (2013) Hepatoid carcinoma of the ovary without staining for alpha-fetoprotein. Obstet Gynecol Sci 56(1):41–44 Talerman A (1967) Hemangiomas of the ovary and the uterine cervix. Obstet Gynecol 30(1):108–113 Talerman A, Auerbach WM, van Meurs AJ (1981) Primary chondrosarcoma of the ovary. Histopathology 5(3):319–324 Talerman A et al (1985) Diffuse malignant peritoneal mesothelioma in a 13-year-old girl. Report of a case and review of the literature. Am J Surg Pathol 9(1):73–80 Tiltman AJ, Haffajee Z (1999) Sclerosing stromal tumors, thecomas, and fibromas of the ovary: an immunohistochemical profile. Int J Gynecol Pathol 18(3):254–258 Tirabosco R et al (2010) Primary myxoid liposarcoma of the ovary in an adolescent girl: a case report. Int J Gynecol Pathol 29(3):256–259 Tochigi N et al (2003) Hepatoid carcinoma of the ovary: a report of three cases admixed with a common surface epithelial carcinoma. Int J Gynecol Pathol 22(3):266–271 Vang R, Ronnett B (2009) Metastatic and miscellaneous primary tumors of the ovary. In: Oliva E, Nucci MR (eds) Gynecologic pathology. Elsevier, Philadelphia, pp 539–613 Veras E et al (2007) Ovarian nonsmall cell neuroendocrine carcinoma: a clinicopathologic and immunohistochemical study of 11 cases. Am J Surg Pathol 31(5):774–782 Vijaya Kumar J et al (2015) A rare presentation of primary leiomyosarcoma of ovary in a young woman. Ecancermedicalscience 9:524

L. E. Schwartz and R. Vang Yaqoob N et al (2014) Ovarian angiosarcoma: a case report and review of the literature. J Med Case Rep 8:47 Yasunaga M et al (2011) Dedifferentiated chondrosarcoma arising in a mature cystic teratoma of the ovary: a case report and review of the literature. Int J Gynecol Pathol 30(4):391–394 Young RH, Scully RE (1983) Ovarian tumors of probable wolffian origin. A report of 11 cases. Am J Surg Pathol 7(2):125–135 Young RH, Scully RE (1984) Fibromatosis and massive edema of the ovary, possibly related entities: a report of 14 cases of fibromatosis and 11 cases of massive edema. Int J Gynecol Pathol 3(2):153–178 Young RH, Scully RE (1985) Ovarian metastases from cancer of the lung: problems in interpretation– a report of seven cases. Gynecol Oncol 21(3): 337–350 Young RH, Prat J, Scully RE (1984) Endometrioid stromal sarcomas of the ovary. A clinicopathologic analysis of 23 cases. Cancer 53(5):1143–1155 Young RH, Silva EG, Scully RE (1991) Ovarian and juxtaovarian adenomatoid tumors: a report of six cases. Int J Gynecol Pathol 10(4):364–371 Young RH et al (1992) Hepatocellular carcinoma metastatic to the ovary: a report of three cases discovered during life with discussion of the differential diagnosis of hepatoid tumors of the ovary. Hum Pathol 23(5):574–580 Young RH, Oliva E, Scully RE (1994) Small cell carcinoma of the ovary, hypercalcemic type. A clinicopathological analysis of 150 cases. Am J Surg Pathol 18(11):1102–1116 Ziari K, Alizadeh K (2016) Ovarian hemangioma: a rare case report and review of the literature. Iran J Pathol 11(1):61–65 Zwiesler D et al (2008) A case report of an ovarian lipoma. South Med J 101(2):205–207

Metastatic Tumors of the Ovary

18

Melinda F. Lerwill and Robert H. Young

Contents General Principles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1152 Extragenital Tumors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Carcinoma of the Stomach . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Intestinal Carcinoma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Tumors of the Appendix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Carcinoid Tumors and Neuroendocrine Carcinomas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Tumors of the Pancreas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Tumors of the Gallbladder and Extrahepatic Bile Ducts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Tumors of the Liver . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Breast Carcinoma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Renal Tumors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Tumors of the Urinary Bladder, Ureter, and Urethra . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Adrenal Gland Tumors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Malignant Melanoma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Pulmonary and Mediastinal Tumors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Extragenital Sarcomas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Miscellaneous Rare Ovarian Metastases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1157 1157 1167 1172 1177 1181 1186 1187 1190 1195 1197 1198 1198 1200 1203 1205

Female Genital Tract Tumors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Tubal Carcinoma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Endometrial Carcinoma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Cervical Carcinoma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Other Uterine Tumors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Vulvar and Vaginal Tumors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1205 1205 1206 1206 1210 1213

Peritoneal Tumors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1213 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1215

M. F. Lerwill (*) Department of Pathology, Massachusetts General Hospital and Harvard Medical School, Boston, MA, USA e-mail: [email protected] R. H. Young Anatomic Pathology, James Homer Wright Pathology Laboratories, Massachusetts General Hospital, Harvard Medical School, Boston, MA, USA e-mail: [email protected]

Tumors that spread to the ovary are important because, while commonest as an autopsy finding, they are not rare in surgical pathology specimens and their misinterpretation may have significant adverse consequences for the patient. The spread may be from adjacent sites by direct local extension or from distant extragenital sites (Young

# Springer Nature Switzerland AG 2019 R. J. Kurman et al. (eds.), Blaustein’s Pathology of the Female Genital Tract, https://doi.org/10.1007/978-3-319-46334-6_18

1151

1152

2006, 2007). The latter tumors are truly metastatic, whereas the designation “metastatic” is sometimes not used for those that are secondary from local sites. For simplicity this discussion includes spread to the ovary from all sites. We first review general principles of clinical, gross, and microscopic evaluation that aid the pathologist in arriving at the correct diagnosis of a metastatic tumor, also highlighting some general pitfalls that are encountered. The subsequent discussion is site, or organ, specific with the exception that a few tumors that may originate at more than one site (Krukenberg tumors, carcinoids, gastrointestinal stromal tumors) are considered under those headings. We divide our consideration into three basic categories: (1) spread from extragenital sites (the most significant practical issue), (2) spread from other sites in the genital tract, and (3) involvement by peritoneal tumors. Hematopoietic tumors are covered in a separate chapter.

General Principles Recognition of the metastatic nature of an ovarian tumor depends on several factors: (1) an awareness of the frequency with which metastases occur and simulate a variety of primary tumors; (2) a thorough clinical history, which in some cases may require the pathologist to prompt the clinician to explore it in greater detail than first done; (3) when indicated a thorough clinical and operative search by the surgeon for a primary tumor outside the ovary; (4) a careful evaluation of the gross and routine microscopic features of the ovarian tumor by the pathologist, including in some cases reinspection of the gross specimen and submission of additional sections; and (5) judicious use of conventional special stains and immunohistochemistry. The diagnosis of a metastatic tumor is often missed by the pathologist because the existence of a concurrent or prior tumor in another organ is either not known or disregarded. The surgical and pathologic findings from previous operations should be reviewed if there is any possibility that they could be related to the ovarian tumor being evaluated. In some cases a search for an

M. F. Lerwill and R. H. Young

extraovarian primary tumor must be conducted postoperatively, based on the pathologist’s suspicion that the ovarian tumor is metastatic. Even if an extraovarian primary tumor is not detected, a diagnosis of a metastasis to the ovary must be strongly considered if the distribution of disease is atypical for primary ovarian cancer or if pathologic examination is highly suggestive of metastasis. For example, the presence of pulmonary or hepatic metastases in the absence of extensive peritoneal disease would be an unusual pattern of spread for an ovarian cancer, but not for certain other tumors that are prone to metastasize to the ovary. The mere presence of tumor outside the ovaries should lead to the serious consideration of a metastasis in certain situations. For example, if a well-differentiated ovarian mucinous tumor is associated with extensive mucinous adenocarcinoma in the omentum and on the peritoneal surfaces, the possibility of spread to the ovary, particularly from the pancreas, biliary tract, or appendix (if pseudomyxoma peritonei is present) should be entertained. Additionally, certain tumors, such as Sertoli cell tumors or primary carcinoid tumors, which most often show benign behavior, should be diagnosed with caution in cases in which there is also extraovarian tumor. In cases of these types, the putative Sertoli cell tumor may prove to be a metastatic tumor that is mimicking it, such as tubular Krukenberg tumor (Bullon et al. 1981), and the carcinoid tumor probably is metastatic rather than primary. An association of an ovarian tumor with clinical or pathologic evidence of excess estrogens, androgens, or progesterone does not exclude the diagnosis of a metastatic tumor, which may have a functioning stroma (Scully and Richardson 1961) (see ▶ Chap. 15, “Sex Cord-Stromal, Steroid Cell, and Other Ovarian Tumors with Endocrine, Paraendocrine, and Paraneoplastic Manifestations”). For various reasons, it is difficult to establish accurately the frequency of metastatic tumors. Some studies have been based on autopsy findings, others on surgical specimens, and still others on both. In addition, some series have included clinically silent metastases such as breast carcinoma found in prophylactic or therapeutic oophorectomy specimens and small metastases detected incidentally during operations for gastric or

18

Metastatic Tumors of the Ovary

intestinal carcinoma. In contrast, other series have been restricted to metastatic tumors that presented clinically as pelvic or abdominal masses. Finally, some studies have included as metastases ovarian carcinomas associated with uterine cancers of similar histologic type, but in some cases the ovarian tumors are independent primary tumors (Ulbright and Roth 1985a; Zaino et al. 1984). The frequency of metastases to the ovary also varies from one country to another because of wide differences in the prevalence of the various cancers that are associated with high rates of ovarian spread. For example, metastatic carcinoma accounts for approximately 40% of ovarian cancers in Japan where gastric carcinoma is common but is far less common in Africa where this form of cancer is relatively rare. The frequency of metastases also has varied greatly in series in which differences in the prevalence of the primary tumors do not adequately explain the discrepant results. Such variations may be related, in part, to the frequency and thoroughness of microscopic examination of the ovaries, because gross inspection may not reveal evidence of involvement in one-third to one-half the cases. The figure for the frequency of ovarian metastases that is most meaningful is one that expresses the probability that a clinically identified ovarian neoplasm is metastatic; this figure is on the order of 5–10%. The age distribution of patients with ovarian metastases depends to a great extent on that of the corresponding primary tumors, but for each of the most common types (intestinal, gastric, and breast), the average age of patients with ovarian involvement is significantly lower than that of those without ovarian spread, suggesting that the richly vascularized ovaries of young women are more receptive to metastases than those of older patients. Tumors spread to the ovary by several routes. Spread from distant sites is mainly via blood vessels and lymphatics. The frequent association of ovarian metastases with other blood-borne metastases, the common finding of tumor within ovarian blood vessels on microscopic examination in cases of metastasis, and the higher frequency of ovarian metastases in young patients support the importance of hematogenous spread.

1153

Transcoelomic dissemination with surface implantation is an important route by which intraabdominal cancers spread to the ovary, as supported by the common association of ovarian involvement with generalized peritoneal spread. Foci of metastatic carcinoma are often found on the surface of the ovary or superficially within the cortex, supporting implantation as the mechanism. Direct spread is an important pathway for carcinomas of the fallopian tube and uterus, for mesotheliomas, and for occasional colonic carcinomas and retroperitoneal sarcomas. Another mechanism of spread of genital tract carcinomas is through the lumen of the fallopian tube and onto the surface of the ovary; this route is taken most often by carcinomas of the uterine corpus (Creasman and Lukeman 1972) but likely also accounts for some cases of spread from the uterine cervix (Pins et al. 1997; Ronnett et al. 2008). The gross features of tumors metastatic to the ovary vary greatly, and they may mimic a variety of primary tumors. Because of the relatively high frequency with which metastases are bilateral (two-thirds to three-quarters of the cases), the possibility of metastasis should especially be considered when evaluating bilateral tumors (Fig. 1) other than serous and undifferentiated carcinomas, which also are commonly bilateral. Endometrioid and mucinous carcinomas, in contrast, are bilateral in less than 15% of cases, and

Fig. 1 Metastatic appendiceal adenocarcinoma. The tumor is bilateral, and the smaller ovary shows several discrete nodules. In the other ovary, the nodules have become confluent, but ill-defined nodularity is still evident

1154

bilateral tumors with endometrioid-like or mucinous features merit more serious consideration for possibly being metastatic (Young and Scully 1991a). However, many metastatic tumors are unilateral, and, if the microscopic features of a tumor suggest metastasis, unilaterality should not be considered a significant argument against it. Almost 10% of bilateral ovarian cancers presenting as adnexal masses prove to be metastatic on careful evaluation. In the relatively common problem of determining whether a mucinous tumor is primary or metastatic, an assessment of laterality and size can be a clue. Bilateral tumors of any size or unilateral tumors under 10 cm are likely metastatic, compared to unilateral tumors over 10 cm, which are usually primary (Seidman et al. 2003). This general guideline may be helpful, particularly in the intraoperative setting; however, there are many exceptions (Stewart et al. 2005; Khunamornpong et al. 2006a), especially in cases of colorectal and endocervical primaries (Yemelyanova et al. 2008). It is also of note that metastatic tumors involving the ovary are often large and may dwarf a much smaller primary tumor. Two other gross findings that are suggestive, but not pathognomonic, of metastasis are the presence of multiple tumor nodules (Fig. 1) and tumor on the surface of the ovary (Fig. 2), sometimes without significant involvement of the underlying parenchyma. As examples of the only suggestive nature of these findings, we note the well-known surface location of some serous carcinomas and, in other cases, the multinodularity of some serous and undifferentiated carcinomas. The microscopic features of these carcinomas generally are not problematic, however. It should also be noted that some endometrioid carcinomas arise from foci of endometriosis in the superficial cortex or on the ovarian surface and accordingly may project off the ovary. An association with endometriosis, which is easily overlooked, can be crucial in supporting an ovarian origin for such a tumor. A gross feature of some metastatic tumors that should not be regarded as establishing the primary nature of the tumor is the presence of cysts (Fig. 3). These are often large and occasionally thin-walled, and they can occur even when there is an absence of cysts in the primary neoplasm.

M. F. Lerwill and R. H. Young

Fig. 2 Metastatic cecal adenocarcinoma. Foci of mucoid tumor are seen on the ovarian surface

Fig. 3 Metastatic cecal adenocarcinoma. Sectioned surface of tumor in prior figure showing a mainly cystic neoplasm grossly compatible with a primary neoplasm

The microscopic appearance of a metastatic tumor obviously varies with the appearance of the primary neoplasm. In addition to the specific features of various primary tumors, the microscopic correlates of the findings noted in the prior paragraph, namely, implants on the surface of the

18

Metastatic Tumors of the Ovary

1155

Fig. 4 Metastatic pancreatic adenocarcinoma. Typical surface implant in the form of a nodular protrusion composed of infiltrating small glands and prominent stroma. Note maturation of the underlying cystic component. (Reproduced with permission from Young 2006)

Fig. 6 Heterogeneous nodular growth. Three separate distinct patterns of growth are evident in this case of metastatic colon cancer: a dilated gland at the top, conventional adenocarcinoma at the right, and small gland adenocarcinoma in a prominent desmoplastic stroma at the bottom left. (Courtesy of Dr. Kenneth R. Lee). (Reproduced with permission from Young 2006)

Fig. 5 Metastatic malignant melanoma. Two discrete nodules of tumor are seen

ovary (Fig. 4) and multinodularity (Fig. 5), suggest metastasis in many cases but with the same caveats as above. These findings have particular weight in cases with mucinous or endometrioid-like

morphologies, as well as in those with any of a diverse number of unusual microscopic appearances. Surface implants typically are focal, often projecting above the surface of the adjacent cortex, and the tumor is frequently embedded in desmoplastic or sometimes hyalinized fibrous tissue (Fig. 4). A conspicuous stromal reaction of the type just noted is also often seen in more central regions of metastatic tumors (Fig. 6) and, particularly if multifocal, is on average much more common in metastatic tumors than in primary neoplasia. Metastatic tumors more often envelop preexisting normal ovarian structures than primary tumors (Fig. 7). Another feature more typical of metastasis is what has been referred to as heterogeneous or nodular invasive growth (Fig. 6). This

1156

Fig. 7 Metastatic colon carcinoma surrounding a corpus albicans

terminology is intended to reflect the varied appearance that results when scattered foci of obviously invasive growth, often in a desmoplastic stroma, are present within a background neoplasm that has a more leisurely growth pattern and sometimes even deceptively benign features. Infiltrating moderate- to high-grade cancer with destructive invasion in a primary mucinous tumor may have a similar appearance, but it is generally not as multifocal nor as striking as in many cases of metastasis. Although certainly not diagnostic, this pattern should cause metastasis to be entertained. The mere nature of the neoplasia may be important. For example, primary mucinous carcinomas of the ovary only rarely have a colloid morphology, whereas this is a well-known pattern of colonic carcinoma; when this morphology is seen in the ovary (Fig. 8), metastasis should be excluded before the tumor is accepted as a primary.

M. F. Lerwill and R. H. Young

Fig. 8 Metastatic colloid adenocarcinoma. This pattern of mucinous carcinoma is uncommonly primary in the ovary, and accordingly such a picture should cause concern for a metastasis. (Reproduced with permission from Young 2006)

The epithelium of most metastatic carcinomas in the ovary is clearly malignant, but a treacherous aspect is the propensity of some tumors, particularly mucinous, to differentiate and result in a borderline-like or even a cystadenoma-like appearance. This so-called maturation phenomenon (Young and Scully 2001) may even result in flattened epithelium that appears benign, and if separated by bland stroma, an adenofibroma or cystadenofibroma may also be mimicked (Fig. 9). Another confusing microscopic feature of some metastatic tumors is the presence of cysts, some of which simulate follicles. These follicle-like spaces (Fig. 10) may be encountered in a variety of metastatic tumors including gastric and intestinal carcinomas, carcinoids, small cell carcinomas from various sites, and malignant melanomas. A wide variety of other patterns

18

Metastatic Tumors of the Ovary

1157

Fig. 9 Metastatic cholangiocarcinoma. The epithelium has undergone marked maturation. The presence of a background cellular stroma results in mimicry of a cystadenofibroma Fig. 10 Follicle-like spaces in case of metastatic malignant melanoma. (Reproduced with permission from Young 2006)

and cell types in metastatic tumors suggest diverse possible primary sites, as discussed in detail and presented in tabular form elsewhere (Young and Scully 2001). Lymphatic or blood vessel invasion, sometimes particularly striking in the hilus, strongly suggests metastasis (Fig. 11). Immunohistochemistry, selectively applied based on a differential diagnosis generated by routine microscopic features, may aid in certain cases (Baker and Oliva 2004; McCluggage and Wilkinson 2005; McCluggage and Young 2005; McCluggage 2012) but is uncommonly diagnostic on its own. Even after the most thorough evaluation, it is sometimes impossible for the pathologist to be certain whether a neoplasm is primary or metastatic, but on the basis of the morphology, one may suggest the most likely possible extraovarian primary sites to direct clinical evaluation. Table 1 presents a comparison of

various features of primary and metastatic mucinous tumors in the ovary but is also broadly applicable, in great part, in other tumor types.

Extragenital Tumors We begin with tumors of the gastrointestinal tract (except sarcomas, considered later) and associated structures, which cause the majority of diagnostic problems, and first discuss the Krukenberg tumor, one of the most well-known cancers involving the ovary (Young 2006).

Carcinoma of the Stomach (i) Metastatic Tumors with Signet-Ring Cells (Krukenberg Tumor).

1158

Fig. 11 Prominent lymphatic involvement by metastatic carcinoma. (Reproduced with permission from Young 2006)

The great majority of metastatic gastric carcinomas to the ovary are Krukenberg tumors, defined as metastatic tumors characterized by the presence of mucin-filled signet-ring cells accounting for at least 10% of the tumor. The source of Krukenberg tumors in the great majority of reported cases is a gastric carcinoma, usually arising in the pylorus. Carcinomas of the large intestine, appendix, and breast are the next most common primary sites; the gallbladder, biliary tract, pancreas, cervix, and urinary bladder are rare sources of these tumors. Saphir (1951) demonstrated in an autopsy study that signet-ring cell carcinomas of various organs are associated more often with ovarian metastasis than carcinomas of other histologic types by a ratio of about 4:1. Subsequent studies have supported his observation: gastric signet-ring cell carcinomas metastasize to the ovary much more often than

M. F. Lerwill and R. H. Young

intestinal-type carcinomas of the stomach (Lerwill and Young 2006), and signet-ring cell carcinomas of the colon also metastasize to the ovaries more frequently than conventional colonic adenocarcinomas (Amorn and Knight 1978). The frequency of the Krukenberg tumor varies with that of gastric carcinoma in the population analyzed. In countries such as Japan, with a high prevalence of gastric carcinoma and a low prevalence of primary ovarian carcinoma, the Krukenberg tumor accounts for a large proportion of all ovarian cancers (Yakushiji et al. 1987). The average age of patients with Krukenberg tumors is about 45 years. One-quarter to almost one-half the patients are under 40 years, and only slightly more than 10% of them are over 60 years of age. This age distribution is related in part to the disproportionate frequency of gastric signet-ring cell carcinomas in young women as well as the greater vascularity of the ovary in young women. In one study, 10% of women 35 years or younger with this tumor had ovarian metastases at presentation (Tso et al. 1987). Almost 90% of patients with Krukenberg tumors have symptoms related to ovarian involvement, usually abdominal pain and swelling; occasionally, there is abnormal uterine bleeding and rarely, particularly during pregnancy, overt signs of excess hormone production such as virilization. The remainder of the patients have gastrointestinal or miscellaneous symptoms related to spread of the cancer to other sites such as lungs or bone or are asymptomatic. A history of prior carcinoma of the stomach or, less often, another organ can be obtained in 20–30% of the cases. The interval between the diagnosis of a gastric carcinoma and the subsequent discovery of ovarian involvement usually is 6 months or less, but periods up to 12 years have been reported (Hale 1968). In most cases, the diagnosis of the gastric carcinoma is made preoperatively, during the operation for the ovarian metastasis or within a few months thereafter. Not infrequently, the primary tumor is too small to be detected at operation, and radiographic examination of the upper gastrointestinal tract also may fail to reveal evidence of a tumor even after the diagnosis of Krukenberg tumor has been

18

Metastatic Tumors of the Ovary

1159

Table 1 Comparison of helpful clinicopathologic features in distinguishing metastatic from primary mucinous cystic tumors Feature History of, or clinical evidence of, extraovarian primary Extraovarian disease Bilaterality Size >15 cm Gross tumor on surface of ovary Microscopic surface implants or surface mucin Heterogeneous nodular invasive growth Colloid pattern Mucin granulomas Vascular invasion Single cell growth Müllerian nature of epithelium Association with teratoma, endometriosis, adenofibroma, Brenner tumor++++ So-called mural nodules

Metastatic Usual++ Common Common Uncommon Occasional Common Common Occasional Uncommon Occasional Occasional Rare Rare Absent

Primary Rare +++ Rare Rare Common Rare Rare Rare Rare Common Rare Uncommon Occasional Occasional Occasional

Modified from Lee and Young (2003) ++ In some cases the primary tumor may initially be occult and require clinical evaluation to detect +++ The possibility that a patient may have independent primary tumors always exists, particularly as primary mucinous tumors of the ovary and certain tumors that may mimic them when they spread to the ovary are common ++++As the four listed lesions are common, by happenstance they might be present in an ovary involved by metastasis. Also it should be noted that maturation in some metastatic tumors may impart a focal adenofibroma-like appearance

established. Rarely, the gastric carcinoma may not be detected until 5 or more years after discovery of the ovarian metastatic tumor. Primary carcinomas, particularly those arising in the breast and stomach, may be very small, requiring exhaustive sectioning to detect them, despite the presence of metastases in some cases. It is possible that tiny primary tumors were missed in these or other organs in the reported autopsied cases of “primary” Krukenberg tumors. Ulbright and Roth (1985b) cited a case in which a primary tumor in the stomach was detected only after microscopic slides prepared from 200 blocks had been examined. Almost all the patients die within a year of the diagnosis of ovarian metastasis, but a rare patient has survived, apparently free of tumor, for as long as 6 years after gastrectomy and bilateral oophorectomy (Holtz and Hart 1982). Such a result, even though exceptional, justifies removal of both the stomach and the ovarian metastases for possible cure in cases in which the tumor appears limited to those organs. It also is prudent for the surgeon to remove the ovaries routinely in menopausal and postmenopausal women who have a

gastric resection for carcinoma so as to prevent the later complication of ovarian metastasis and avoid another operation. Gross Findings Krukenberg tumors typically form round to oval, firm, white masses that may be bosselated and may attain a large size. The surfaces are generally devoid of appreciable adhesions. The sectioned surfaces usually are tan or white (Fig. 12), but areas of purple, red, or brown discoloration and extensive hemorrhage also are encountered. The appearance may be relatively uniform, or, in other cases, ill-defined nodularity or even discrete nodules may be seen. The consistency is characteristically firm, but fleshy, gelatinous, or spongy areas are common. Sometimes the periphery is distinctly different in appearance from the central region, the latter often being softer than the periphery (Fig. 13). Occasionally, the gross presentation is atypical with large, thin-walled cysts containing mucinous or watery fluid, separated by relatively small amounts of solid tissue. Both ovaries are involved in 80% or more of the cases.

1160

M. F. Lerwill and R. H. Young

Fig. 12 Krukenberg tumor. The sectioned surface is more or less uniform and tan

Fig. 14 Krukenberg tumor. Numerous signet-ring cells are present within a cellular stroma

Fig. 13 Krukenberg tumor. The central region in this case differed in appearance from the periphery, a feature occasionally seen

Microscopic Findings The histologic appearance of these tumors is much more varied than the most emphasized morphology in the literature, namely, that of signet-ring cells in a cellular stroma (Fig. 14). This picture, which is of historical interest because it resulted in the initial confusion with fibrosarcoma, is a prominent finding in only a minority of cases. We first consider the low-power features and then aspects related to signet-ring cells (and other cell types), other epithelial elements, and stroma in turn.

Growth as distinct nodules, or at least vague nodularity, is typically conspicuous on low-power (Fig. 15), but the nodules and intervening stroma may be quite variable in appearance, particularly when other features (see below) are present (Figs. 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, and 32). The nodules are often separated by edematous stroma, and a picture of densely cellular pseudolobules is often seen (Fig. 16). The nodules themselves are generally composed of jumbled admixtures of signet-ring cells, indifferent cells, glands, cysts, and background stroma. Similar admixtures of epithelial and stromal elements are seen in tumors that have a more diffuse arrangement. Sometimes a striking low-power feature is a greater cellularity at the periphery, with aggregates of tumor cells, stroma, or both ramifying from the periphery into central more edematous regions (Fig. 17). Residual ovarian structures may be present in the midst of the tumor (Fig. 18).

18

Metastatic Tumors of the Ovary

1161

Fig. 16 Krukenberg tumor. Pseudolobular pattern Fig. 15 Krukenberg tumor. Typical alternating hypercellular and hypocellular regions

The signet-ring cells vary greatly in amount and, accordingly, prominence. Occasionally vast numbers are present, and a “sea” of signet-ring cells is immediately striking (Fig. 19). Conversely, it is not rare for these cells to be relatively inconspicuous, at least on initial low-power evaluation and even sometimes on high-power scrutiny. The arrangement of the signet-ring cells (Fig. 20) to one another and to other epithelial elements is equally variable. They may grow diffusely, in somewhat orderly clusters (Fig. 21), in pseudotubular formations, or in a totally random fashion within the stroma or between glands and cysts. The individual signet-ring cells are generally of relatively similar size and usually have pale to basophilic cytoplasm, which compresses the nucleus to the periphery. This results in the nuclei often having rather deceptively bland cytologic characteristics. Occasionally the cytoplasm is eosinophilic, sometimes densely so (Fig. 22).

The cytoplasm may have a bull’s-eye appearance, containing a large vacuole with a central eosinophilic body. Other neoplastic cells may be present that are mucinous but not of signet-ring cell morphology, and it is not rare for a component of the tumor to have nondescript mucin-free cells; such regions may be prominent (Fig. 23). Rarely, cells with clear cytoplasm are present (Fig. 24), and exceptionally, one can even see squamous or transitional-like cells. Other epithelial elements, specifically glands and cysts, are present in most Krukenberg tumors (Kiyokawa et al. 2006; Young 2006). In many cases, these can be as striking a finding as the signet-ring cells. The glands are usually small and often impart a microcystic appearance, but a spectrum is encountered up through mediumsized glands to large cysts. The glands can appear indifferent or have a striking intestinal-type appearance, but the pseudoendometrioid morphology typical of non-signet-ring cell intestinal carcinoma (see below) is rarely present, and dirty

1162

M. F. Lerwill and R. H. Young

Fig. 17 Krukenberg tumor. The outer region of the tumor (top) contrasts with the more central region that shows marked edema and clusters of tumor cells

Fig. 18 Krukenberg tumor. An entrapped follicle is evident

necrosis is also uncommon. The glands and cysts may have attenuated lining cells or, uncommonly, columnar mucinous cells. Small round tubules, larger hollow tubules, and solid pseudotubular formations with or without signet-ring cells may all be seen (Fig. 27) and, when prominent, account for the term “tubular Krukenberg tumor.” Mucin stains may highlight signet-ring cells in these and other cases (Fig. 28). Many tumors have nondescript patterns of carcinoma, such as masses, nests, cords (Fig. 29), and individual cells, at least focally. Follicle-like spaces may be prominent (Fig. 30). In our experience, the stroma is more often edematous than cellular, and a highly cellular “sarcoma-like” picture is uncommon. Indeed, densely cellular regions are more often due to a conspicuous content of small hyperchromatic epithelial cells with scant cytoplasm than to a cellular stromal proliferation. Mucin in the stroma is

sometimes prominent and may contain signetring cells, form acellular mucin pools, or be separated by wispy collagen producing a pattern that has been referred to as feathery degeneration (Fig. 31). When glands are set in a relatively regular fashion on the background stroma, there may be a deceptively orderly architecture, albeit usually only focally. If the stroma is cellular in such instances, a superficial resemblance to an adenofibroma may result; this can be particularly treacherous at the time of frozen section examination. If only a bland “fibroma-like” appearance of the stroma is sampled for frozen section, a misdiagnosis of fibroma may result (Fig. 32). Occasionally, these fibromatous regions have a storiform appearance. The Krukenberg tumor is one of the ovarian tumors most often associated with stromal lutein cells (Fig. 20). They are often present in at least small numbers and may be prominent, particularly if the patient is pregnant

18

Metastatic Tumors of the Ovary

1163

Fig. 20 Krukenberg tumor. Typical signet-ring cells are intermingled with luteinized stromal cells Fig. 19 Krukenberg tumor. Numerous signet-ring cells

as is sometimes the case given the relative youth of many patients. As with metastases in general, blood vessel and lymphatic invasion are common, generally identified at the periphery or in the hilar aspect of the neoplasm. In one unique case of a gastric primary, the ovarian metastases showed yolk sac differentiation, none being identified in the primary tumor (Zamecnik et al. 2008). In two other unusual cases, metastatic signet-ring cell carcinoma in the ovaries originated in the cervix, a judgment based significantly on identification of human papillomavirus by in situ hybridization (Veras et al. 2009). Differential Diagnosis The Krukenberg tumor may resemble a fibroma or any other type of solid ovarian tumor on gross examination. Its appearance also occasionally may be deceptive on frozen section, as noted above, or low-power examination, but it should be readily diagnosable on high-power microscopic

examination, especially with the aid of mucin stains. A frequent misdiagnosis is a Sertoli–Leydig cell tumor, particularly when a prominent tubular component and luteinization of the stroma are encountered (Bullon et al. 1981); signet-ring cells, however, are not a feature of Sertoli–Leydig cell tumors except for occasional tumors of heterologous type (see ▶ Chap. 15, “Sex Cord-Stromal, Steroid Cell, and Other Ovarian Tumors with Endocrine, Paraendocrine, and Paraneoplastic Manifestations”). The sclerosing stromal tumor may contain cells resembling signet-ring cells as well as a proliferating fibroblastic component, but such cells contain lipid rather than mucin. The rare signet-ring stromal tumor also may enter the differential diagnosis, but the signet-ring cells in that tumor also fail to react with mucin stains. Surface epithelial tumors generally cause fewer problems in differential diagnosis than sex cordstromal tumors. Cells with clear cytoplasm may raise the issue of clear cell carcinoma, but the clear cells in the latter contain glycogen; mucin, when present, is typically luminal and

1164

M. F. Lerwill and R. H. Young

extracellular. In rare cases, clear cell carcinomas may have focal signet-ring cells, but the presence of other characteristic features, such as a typical tubulocystic pattern or papillae, permits their identification. Primary mucinous tumors are rarely difficult to distinguish from Krukenberg tumors, as the latter have so many features pointing to a metastatic process. It should be noted that rare mucinous tumors can contain signet-ring cells, but their similarity to Krukenberg tumors is otherwise minimal (McCluggage and Young 2008). Occasional serous and even some endometrioid and undifferentiated carcinomas can have signet-ring cells (Che et al. 2001), but the many differences between these tumors and Krukenberg tumors are such that resolving the differential diagnosis should be straightforward. Mucinous carcinoid tumors that contain large numbers of signet-ring cells are distinguished from Krukenberg tumors by their additional Fig. 21 Krukenberg tumor. Signet-ring cells aggregate together in a rather orderly manner within uniform spaces

Fig. 22 Krukenberg tumor. Many signet-ring cells in this illustration have dense eosinophilic cytoplasm

Fig. 23 Krukenberg tumor. Epithelial cells without a conspicuous mucinous content grow as cords and clusters separated by luteinized stromal cells, imparting a superficial resemblance to Sertoli–Leydig cell tumor

18

Metastatic Tumors of the Ovary

Fig. 24 Krukenberg tumor. Cords and clusters of cells with appreciable clear cytoplasm are present in a moderately cellular fibrous stroma. The appearance in this field is non-specific

component of carcinoid, the presence of which can be confirmed by special stains. Infrequently, a mesothelial neoplasm may be a diagnostic consideration. The vacuoles of the rare adenomatoid tumor that involves the ovary may be misconstrued as signet-ring cells, but many differences, including even the much less ominous gross appearance of the adenomatoid tumor, should help (Phillips et al. 2006). Adenomatoidlike foci may be seen in some malignant mesotheliomas (Baker et al. 2005), such that one involving the ovary could cause a Krukenberg tumor to be entertained in the differential diagnosis; however, the tubulopapillary patterns of mesothelioma are generally distinctive, and these tumors lack the many above described features seen in most Krukenberg tumors. The rare nonneoplastic lesion, mucicarminophilic histiocytosis, which is caused by medical use of substances containing polyvinylpyrrolidone

1165

Fig. 25 Krukenberg tumor. Glandular differentiation

or oxidized regenerated cellulose, is characterized by signet-ring-like cells and may involve numerous tissues and organs including the ovaries (Kuo and Hsueh 1984; Kershisnik et al. 1995). Both types of material are positive for mucicarmine. Polyvinylpyrrolidone is periodic acid-Schiff and Alcian Blue negative, whereas oxidized regenerated cellulose can be positive for both. Immunohistochemistry is definitive, as the cells are negative for cytokeratin and positive for histiocytic markers such as CD68. (ii) Metastatic Intestinal-Type Adenocarcinoma of the Stomach. Only a small number of cases of this type are documented (Lerwill and Young 2006). The limited information available suggests that these patients are somewhat older than the usual patient with a Krukenberg tumor. The endocrine manifestations seen with some cases of the latter have not been a feature of the intestinal-type cancers.

1166

Fig. 26 Krukenberg tumor. Small microcysts of the type seen in many cases

Gross Findings The tumors, which may be bilateral or unilateral, are typically solid and cystic and large, resembling metastatic colon cancer rather than the usual Krukenberg tumor. Microscopic Findings The tumors are typically formed of medium-sized tubular glands, resulting in a pseudoendometrioid pattern (Fig. 33) as seen with metastatic colon cancer. Other familiar features of the latter such as dirty necrosis may also be encountered. Findings common to many cases of metastatic disease in the ovary, such as prominent stromal edema (Fig. 34) and notable morphologic variability within a small zone of tumor (Fig. 35), are seen, as are non-specific patterns of growth such as cords. Less often the picture is that of a mucinous neoplasm. A very minor component (definitionally 90% of primary and metastatic ductal and lobular carcinomas of the breast, including many

1194

M. F. Lerwill and R. H. Young

Fig. 88 Metastatic breast carcinoma. Growth in cords typical of metastatic lobular carcinoma

Fig. 89 Metastatic breast carcinoma. This tumor had a signet-ring cell component qualifying it as a Krukenberg tumor

triple-negative carcinomas (Byrne et al. 2017; Miettinen et al. 2014). In contrast, less than 10% of ovarian surface epithelial carcinomas show positivity for GATA3, and when present, staining is usually weaker and more focal than in breast cancers. It is important to remember, however, that GATA3 is positive in a wide range of other neoplasms, including >90% of transitional cell carcinomas (Miettinen et al. 2014), so its utility depends on the specific differential diagnosis being considered. Gross cystic disease fluid protein-15 (GCDFP-15) (Fig. 90) is another commonly utilized marker of breast origin (Monteagudo et al. 1991). Between 40% and 70% of breast carcinomas metastatic to the ovary have been reported to be positive for GCDFP-15, whereas primary ovarian carcinomas are only rarely positive (80% of serous and transitional cell carcinomas of the ovary, but it is only positive in approximately 3% of unselected breast carcinomas (Bombonati and Lerwill 2012). However, approximately 50% of pure and mixed mucinous

18

Metastatic Tumors of the Ovary

1195

Fig. 90 Metastatic carcinoma to a serous neoplasm. Confirmation that the infiltrating carcinoma was metastatic carcinoma from the breast rather than a component of the

primary serous neoplasm was aided by positive immunohistochemical staining of the breast cancer for gross cystic disease fluid protein-15 (right)

carcinomas and 11% of micropapillary carcinomas of the breast can show positivity for WT-1, which on occasion is diffuse and strong (Bombonati and Lerwill 2012; Lee et al. 2007). Cytokeratin 7/20 profiles do not aid in the distinction of metastatic breast from primary ovarian adenocarcinoma, as both are typically CK7+/CK20–. Finally, in cases of ovarian involvement by the intraabdominal desmoplastic small round cell tumor (Young et al. 1992b), the diagnosis of metastatic breast cancer may be suggested in areas. However, these patients usually are in their teens, when breast cancer is rare, and other more characteristic foci of the tumor and its typical immunohistochemical profile facilitate the interpretation. We have seen one case in which this tumor involved the breast, confusing the picture, but the breast involvement suggested metastasis, and the tumor exhibited the characteristic immunohistochemical staining of the desmoplastic small round cell tumor.

Renal Tumors Renal cell carcinoma rarely spreads to the ovary, with only a small number of ovarian metastases reported in detail (Insabato et al. 2003; Liang et al. 2016; Vara et al. 1998; Spencer et al. 1993; Young and Hart 1992). The ovarian tumor may be discovered first, leading to an initial misdiagnosis of a primary ovarian clear cell carcinoma. The renal tumors usually are detected within a short period of time in these patients, but in one case the renal primary was not detected until 8 years later (Young and Hart 1992). Gross Findings The ovarian tumors, only a minority of which are bilateral, are often large (average, 12.5 cm in greatest dimension) and are either solid or solid and cystic, with one cystic tumor being unilocular and containing a 2.5 cm solid nodule in one area. The solid components of the tumors are either uniformly or focally yellow to orange (Fig. 92).

1196

Fig. 91 Metastatic breast carcinoma. Immunohistochemical reactivity for mammaglobin

M. F. Lerwill and R. H. Young

Fig. 93 Metastatic renal cell carcinoma, clear cell type. The primary tumor in this case was not diagnosed until 8 years after the removal of the metastasis. Note the sinusoidal vascular pattern

prominent sinusoidal vascular pattern was almost always present (Fig. 93).

Fig. 92 Metastatic renal cell carcinoma. The tumor is orange. (Courtesy of Dr. Mahul B. Amin)

Microscopic Findings With one possible exception, the reported renal tumors were well-differentiated clear cell adenocarcinomas; microscopic examination showed a relatively uniform picture of diffuse sheets of clear cells or tubules lined by similar cells and containing eosinophilic material or blood; a

Differential Diagnosis It is helpful in differential diagnosis that primary clear cell carcinoma of the ovary has a tubulocystic and papillary component, hobnail cells, and intraluminal mucin in the great majority of cases. Hobnail cells and conspicuous mucin production, in contrast, are exceptional in renal cell carcinomas. In addition, the typical sinusoidal vascular framework of renal cell carcinoma is not a feature of ovarian clear cell carcinoma. In cases of pure clear cell carcinoma of the ovary without hobnail cells or mucin secretion, radiologic evaluation of the kidney may rarely be necessary to exclude a renal cell carcinoma. A panel of antibodies may aid in evaluation: ovarian clear cell carcinomas are usually positive for CK7 and mesothelin and negative for CD10 and renal cell carcinoma marker (RCCma), whereas renal clear cell carcinomas

18

Metastatic Tumors of the Ovary

1197

series of this neoplasm. In one remarkable case, a patient with a rhabdoid tumor of the kidney presented with an ovarian metastasis, initially misinterpreted as a granulosa cell tumor, the primary renal tumor being undiscovered until autopsy (Young et al. 1993a).

Tumors of the Urinary Bladder, Ureter, and Urethra

Fig. 94 Metastatic renal cell carcinoma. This tumor, which had an unusual oxyphilic morphology, was immunoreactive for CD10

often demonstrate the opposite immunoprofile (CK7 /mesothelin-/CD10+/RCCma+) (Fig. 94) (Cameron et al. 2003; Leroy et al. 2007; Ohta et al. 2005). Although PAX2 may be a somewhat more sensitive marker than RCCma for metastatic renal cell carcinoma, it should be noted that about 40% of ovarian clear cell carcinomas also express this protein (Gokden et al. 2008). PAX8 is expressed in 90% of renal cell carcinomas and therefore does not aid in differentiating ovarian surface epithelial carcinoma from metastatic renal cell carcinoma (Laury et al. 2011). Renal transitional cell tumors rarely spread to the ovary, but exceptionally a patient with a renal pelvic tumor of this type has an ovarian metastasis at the time of presentation (Hsiu et al. 1991; Oliva et al. 1993). One report documents a renal pelvic tumor with glandular differentiation and signetring cells that was associated with bilateral Krukenberg tumors (Irving et al. 2006). Ovarian metastases from Wilms’ tumor of the kidney are rare, and no examples are present in several large

Tumors from these sites uncommonly metastasize to the ovaries, although recent evidence suggests that the uncommon plasmacytoid bladder cancer may have some propensity for ovarian spread (Ricardo-Gonzalez et al. 2012). Rare signet-ring cell carcinomas metastatic from the bladder have had the appearance of a Krukenberg tumor (Young and Scully 1988a). A small number of urachal adenocarcinomas have formed metastatic mucinous cystic tumors in the ovary (Ohira et al. 2003; Young 1995). Only isolated examples of ovarian metastasis of ureteral or urethral cancer have been reported in the literature. In many cases of possible transitional cell carcinoma metastatic to the ovary, it is difficult to distinguish between a metastatic tumor and a borderline or malignant Brenner tumor or independent primary transitional cell carcinoma of the ovary (Soslow et al. 1996; Young and Scully 1988a). In almost all borderline or malignant Brenner tumors, however, foci of typical benign Brenner tumor also can be found, and the presence of associated benign mucinous elements also favors the diagnosis of a Brenner tumor. The extent of invasion of the primary extraovarian tumor and the general features of metastatic involvement of the ovary all have to be considered in the evaluation of these cases. The metastatic transitional cell carcinoma in the series of Ulbright et al. (1984) was cystic, exemplifying the great propensity of ovarian metastases from various sites to undergo cystic change. Transitional cell carcinomas of the bladder are often positive for CK20, uroplakin III, and thrombomodulin; they are typically negative for WT-1 (Logani et al. 2003). Primary ovarian transitional cell carcinomas, on the other hand, are usually

1198

negative for CK20, uncommonly express uroplakin III and thrombomodulin, and are frequently positive for WT-1. This difference in immunoprofile supports that the shared histologic features of the two tumors do not indicate a shared histogenesis and that primary ovarian transitional cell carcinomas are variants of surface epithelial neoplasia. Interestingly, Brenner tumors show a high frequency of uroplakin III and thrombomodulin expression, suggesting true urothelial differentiation in these tumors (Logani et al. 2003).

M. F. Lerwill and R. H. Young

arisen in the choroid or elsewhere. Occasionally ovarian involvement is clinically symptomatic, as exemplified by many of the 52 cases in 3 relatively large series of melanomas metastatic to the ovary (Fitzgibbons et al. 1987; Gupta et al. 2004; Young and Scully 1991b). The patients in these reports had an average age of 38 years and two were teenagers. The usual presentation is abdominal swelling or pain, but a history of melanoma is usually existent, albeit not always immediately made known to the pathologist. Approximately 80% of the patients have metastatic tumor outside the ovary, usually within the pelvis and upper abdomen.

Adrenal Gland Tumors Neuroblastoma spreads to the ovary more frequently than other tumors of the adrenal gland. 25% to 50% of females with neuroblastoma have ovarian involvement at autopsy (Meyer et al. 1979). Clinically significant metastases during life are rare but documented (Sty et al. 1980, Young et al. 1993a). Rarely neuroblastoma is primary in the ovary, and such tumors must be distinguished from metastatic neuroblastomas. The unilaterality of the primary tumors, their occasional association with a teratoma, and the absence of a known primary tumor elsewhere are helpful in the differential diagnosis in individual cases. The prominent fibrillary background of neuroblastoma and the presence of pseudorosettes should aid in the distinction of metastatic neuroblastoma from other metastatic small cell tumors; immunohistochemical staining also may help in a case in which routine stains are not diagnostic. Metastases of adrenal cortical carcinomas to the ovary are rarely found even at autopsy. One non-autopsy case resulted in the broad differential of a malignant oxyphilic ovarian tumor (Kurek et al. 2001). Pheochromocytomas spread to the ovary even less commonly.

Gross Findings The ovarian tumors average 10 cm in diameter; about 30% are noted to be black (Fig. 95) or brown. About 80% have a minor cystic component. Rarely, a tumor is mostly cystic.

Malignant Melanoma Autopsies of patients who died of malignant melanoma have revealed ovarian involvement in about 20% of the cases. Most of the tumors have originated in the skin, but occasional examples have

Fig. 95 Metastatic malignant melanoma. The tumor has a bosselated external surface and is black

18

Metastatic Tumors of the Ovary

Fig. 96 Metastatic malignant melanoma. The tumor cells have appreciable eosinophilic cytoplasm

Microscopic Findings On low-power examination, a feature suggesting the metastatic nature in a number of the cases is growth of the tumor in the form of multiple nodules. The most common microscopic appearance is that of large cells with abundant eosinophilic cytoplasm (Fig. 96). Occasional tumors are characterized by small cells with scanty cytoplasm (Fig. 97) and a minority by spindle cells; admixtures of these cells types may be encountered. Follicle-like spaces are seen in approximately 40% of the cases (Fig. 10). A helpful diagnostic feature of many metastatic melanomas is the presence of discrete rounded aggregates with a nevoid appearance (Fig. 98). Prominent nucleoli are seen in many cases, and cytoplasmic pseudoinclusions are present in many nuclei in about 25%. The presence of melanin pigment is an obvious clue to the nature of the tumor in these cases, but melanin was inconspicuous or absent in approximately half the cases in the reported series. Miscellaneous potentially confusing findings have

1199

Fig. 97 Metastatic malignant melanoma. The picture is that of a small cell malignant tumor

been clear cells, rhabdoid cells, growth as cords, and a myxoid stroma. Differential Diagnosis Metastatic melanoma must be distinguished from the rare primary melanoma (McCluggage et al. 2006) that usually arises in the wall of a dermoid cyst, which is sometimes accompanied by junctional activity beneath the squamous lining of the cyst or is associated with another teratomatous component such as a struma ovarii. Because recognition of teratomatous elements is important in establishing the primary nature of a melanoma, the pathologist should sample the specimen extensively. In cases of apparently pure ovarian melanoma without obvious evidence of a primary tumor elsewhere, a meticulous search for an occult primary tumor should be conducted. If there is no evidence of a primary tumor elsewhere, it is possible that a primary cutaneous melanoma that has regressed was the source of the ovarian tumor. In these cases, bilaterality or growth of the ovarian tumor in the form of multiple nodules

1200

strongly suggests metastasis even in the absence of a known primary tumor. In some cases, removal of a primary melanoma may be remote and possibly not considered relevant by the patient or known by the clinician. Metastatic melanoma, particularly if it is amelanotic, may resemble closely a lipid-poor steroid cell tumor or, if it is found during pregnancy, a pregnancy luteoma. Melanin can be misinterpreted as lipochrome pigment, the presence of which may be a feature of steroid cell tumors and impart a dark green-brown or almost black color to the neoplastic tissue. The presence of follicle-like spaces in metastatic melanomas has resulted in their confusion with small cell carcinomas of the hypercalcemic type (when the cells are small) as well as juvenile granulosa cell tumors (when the cells have conspicuous eosinophilic cytoplasm). Rarely, surface epithelial neoplasms, particularly undifferentiated carcinoma and transitional cell carcinoma, may be reasonable considerations in the differential diagnosis, but surface epithelial neoplasms will often have at least minor obvious epithelial characteristics that rule out melanoma. In a young person, dysgerminoma is rarely entertained, but a variety of architectural and cytologic differences should resolve this issue. In all the aforementioned situations, the diagnosis of metastatic melanoma to the ovary may be confirmed by the immunohistochemical positive staining for markers such as S-100, HMB-45, MART-1, SOX10, and MITF and negative staining for keratin and other antigens characteristic of other neoplasms that may be in the differential diagnosis.

Pulmonary and Mediastinal Tumors Only approximately 5% of women with lung cancer have ovarian metastases at autopsy, and the surgical pathologist uncommonly encounters an ovarian tumor of this type. Exceptionally, an ovarian metastasis either precedes the discovery of a pulmonary tumor or is found simultaneously. Salient features of this phenomenon based on a series of 32 cases are summarized (Irving and Young 2005).

M. F. Lerwill and R. H. Young

Fig. 98 Metastatic malignant melanoma. Nests of cells have a nevoid appearance

These tumors have occurred at an average age of 47 years. A history of lung cancer is known at presentation in slightly more than half the cases. In most of the remainder, the two tumors are discovered essentially synchronously, but in almost 20% the ovarian manifestations antedate recognition of the pulmonary primary, sometimes by as much as 2 years. Gross Findings The tumors have been bilateral in only about one-third of the cases, and they average about 10 cm in maximum dimension. No unique gross characteristics have been evident in the cases encountered to date, but some have features, such as striking multinodularity (Fig. 99), in keeping with a metastatic neoplasm. Rarely a primary surface epithelial neoplasm is mimicked (Fig. 100). Microscopic Findings The greatest number of cases, about 44%, is small cell carcinomas (Figs. 101 and 102), the remaining neoplasms being largely split between

18

Metastatic Tumors of the Ovary

1201

Fig. 99 Metastatic bronchioloalveolar carcinoma of the lung. Note the multiple nodules. (Courtesy of Dr. Jaime Prat)

Fig. 101 Metastatic small cell carcinoma from the lung. Trabecular pattern

Fig. 100 Metastatic small cell carcinoma from the lung. The solid and cystic sectioned surface mimics a surface epithelial neoplasm

adenocarcinoma and large cell carcinoma with a ratio of about 2:1. Rare more differentiated neuroendocrine neoplasms falling in the atypical carcinoid-neuroendocrine carcinoma group have spread to the ovary (Fig. 103). Perhaps surprisingly, spread of squamous cell carcinoma, although documented, is exceptionally uncommon. The morphologic features of the tumors in the ovary are similar to those encountered in the lung except for features relevant to metastatic ovarian disease, specifically surface involvement, nodularity, and vessel space invasion. Rare examples represent Krukenberg tumors (Giordano et al. 2017)

Differential Diagnosis When a patient has pulmonary and ovarian neoplasms, it can be difficult to decide which tumor is primary. When the histologic features are typical of a lung carcinoma, a pulmonary origin can be assumed with rare exceptions. Small cell carcinomas of pulmonary type may be primary in the ovary, but there is typically no pulmonary involvement in such cases, facilitating the diagnosis of an ovarian primary. The focal presence of a surface epithelial tumor is also sometimes helpful in excluding a metastasis. In the absence of such a finding and in the presence of tumor in the lung, it may be impossible to decide whether an ovarian small cell carcinoma of pulmonary type is primary or metastatic. Gland differentiation may be seen in some cases of metastatic small cell carcinoma (Fig. 102). This should be borne in mind in the differential diagnosis with a primary ovarian small cell carcinoma of pulmonary type, which is sometimes associated with endometrioid carcinoma; the latter may be suggested by the glands of what in fact is a metastatic small cell

1202

Fig. 102 Metastatic small cell carcinoma from the lung. There is focal gland differentiation

M. F. Lerwill and R. H. Young

carcinoma. This exemplifies the diagnostic problems that can be hard to resolve in rare cases. The metastatic adenocarcinomas generally have non-specific glandular features, which, should a pulmonary neoplasm be known to exist, will aid in distinction from independent primary ovarian adenocarcinoma. Unfortunately, given the spectrum of primary ovarian neoplasia, particularly of endometrioid type, recognizing a tumor as metastatic in the absence of known pulmonary neoplasia may be difficult or impossible. The large cell carcinomas may have a broad differential diagnosis of oxyphilic tumors of the ovary. The utility of thyroid transcription factor-1 (TTF-1) in the recognition of metastatic lung carcinoma depends on the morphologic subtype under consideration. Approximately 75% of pulmonary adenocarcinomas are positive for TTF-1, whereas only 40% of large cell carcinomas are positive and the majority of squamous cell carcinomas are negative (DiLoreto et al. 1997; Hecht et al. 2001; Kaufmann and Dietel 2000; Reis-Filho et al. 2000). TTF-1 may be also expressed in a subset of ovarian carcinomas, including notably up to 37%

Fig. 103 Metastatic neuroendocrine carcinoma from the lung (left) with immunoreactivity for chromogranin (right)

18

Metastatic Tumors of the Ovary

of serous carcinomas (Kubba et al. 2008; Zhang et al. 2009). Its expression in primary ovarian tumors is usually focal but is occasionally diffuse. A negative reaction for TTF-1 does not exclude a lung primary, and although diffuse positivity may suggest a pulmonary origin, it is not independently diagnostic of such. Napsin A is often used in conjunction with TTF-1 as a marker of pulmonary adenocarcinoma. It is positive in 80% to 90% of lung adenocarcinomas but is usually negative in squamous cell, large cell, and small cell carcinomas (Bombonati and Lerwill 2012). Among ovarian carcinomas, it is frequently positive in clear cell carcinoma but only rarely positive in other surface epithelial subtypes (Kandalaft et al. 2014). Therefore its ability to discriminate between lung and ovarian origin depends upon the specific tumor types being considered in the differential diagnosis. Cytokeratin 7/20 profiles do not aid in the distinction of metastatic pulmonary from primary ovarian adenocarcinoma, as both are typically CK7+/CK20–. Metastatic small cell carcinomas in the ovary may originate in sites other than the lung. Three such tumors were primary in the mediastinum, apparently of thymic origin, and had ovarian metastases at the time of presentation (Eichhorn et al. 1993). Small cell carcinomas from various sites express TTF-1, and thus TTF-1 positivity is not specific for pulmonary origin among tumors of this type (Agoff et al. 2000). One of the two ovarian small cell carcinomas of pulmonary type was found to be positive for TTF-1 in one study (Carlson et al. 2007). One neuroblastoma primary in the posterior mediastinum metastasized to the ovary168. Thymomas have involved the ovary rarely (Martin-Hernandez et al. 2015).

1203

some appreciable diagnostic problems (Irving et al. 2005). The five patients described were all adults. In three, the primary (two in the small bowel and one in the mesentery) and metastatic tumors were discovered synchronously. In one, however, ovarian masses were discovered 18 months before the gastric primary, and in the fifth case, ovarian spread was found 27 years after a primary small bowel tumor had been resected. The ovarian tumors had no specific gross features but were often sizeable. Microscopic examination, as expected given the known morphology of this tumor, typically showed a low-grade spindle cell neoplasm (Fig. 104), but a number of features such as the presence of signet-ring-like cells and palisading (Fig. 105) complicated the appearance. The differential diagnosis may be broad and include cellular and typical fibromas, smooth muscle tumors, and other primary soft tissue-type tumors. In cases in which any of these is a consideration but there is an atypical feature, such as bilaterality or extraovarian

Extragenital Sarcomas Extragenital sarcomas, whether from the viscera or the soft tissues, uncommonly metastasize to the ovary except in late stages of the disease, and the interpretation is usually obvious. An exception to this is the gastrointestinal stromal tumor, a small series of which metastatic to the ovary pointed out

Fig. 104 Metastatic gastrointestinal stromal tumor. The picture may be confused with cellular fibroma

1204

M. F. Lerwill and R. H. Young

Fig. 105 Metastatic gastrointestinal stromal tumor. Palisading is seen

Fig. 106 Metastatic gastrointestinal stromal tumor. Immunoreactivity for c-kit

disease, immunohistochemistry for c-kit (Fig. 106) or other markers such as DOG1 and succinate dehydrogenase B is appropriate to rule out the possibility of a gastrointestinal stromal neoplasm (Gill et al. 2010; Miettinen et al. 2009). In one study, eleven rhabdomyosarcomas metastatic to the ovary were diagnosed in patients 6 to 27 years of age (Young and Scully 1989). Six tumors were alveolar rhabdomyosarcomas, three embryonal, one mixed embryonal and alveolar rhabdomyosarcoma, and one of unstated subtype. In most of the cases, the ovarian spread was a late manifestation of disease. The ovarian tumors were symptomatic in only two patients, in whom the ovarian involvement was detected within a few weeks of discovery of a soft tissue mass by the patient. The ovarian tumors were bilateral in two cases. In cases of embryonal rhabdomyosarcoma metastatic to the ovary, the diagnosis of rhabdomyosarcoma usually is evident because of the presence of strap cells, and the tumor must be distinguished from a primary embryonal rhabdomyosarcoma, which is the most common subtype

of primary malignant striated muscle tumor of the ovary. Metastatic alveolar rhabdomyosarcoma more commonly raises the differential of other primary and metastatic small cell tumors of the ovary in young women. A combination of clinical findings, thorough sampling, and immunohistochemistry to varying degrees in individual cases will help resolve what can be a very challenging area. Two other cases in which the ovary was involved by rhabdomyosarcoma have occurred in patients with a clinical picture that simulated acute leukemia (Hayashi et al. 1988). In one study of 21 metastatic sarcomas to the ovary other than rhabdomyosarcoma, 10 were extragenital in origin, and all were clinically significant (Young and Scully 1990b). These tumors have included a miscellaneous group of soft tissue sarcomas primary at a variety of sites. Rare cases of hemangiosarcoma that metastasized to the ovaries have been documented as have a few cases of Ewing’s sarcoma. The latter are noteworthy as they may be part of the broad differential diagnosis of a small cell malignant tumor of the ovary.

18

Metastatic Tumors of the Ovary

Miscellaneous Rare Ovarian Metastases Metastases to the ovary other than those already discussed are of great rarity and generally only of relevance to autopsy pathology. Carcinomas of the thyroid only exceptionally spread to the ovary even in autopsy series, but a few cases documented during life exist. In one case (Young et al. 1994), a 29-year-old woman had a 17 cm right ovarian tumor 12 years after undergoing a partial thyroidectomy for follicular carcinoma. The tumor also had spread to the brain and one adrenal gland by the time the ovarian tumor was discovered. Initial consideration was given to the diagnosis of a malignant struma ovarii in this case because of the interval since the thyroid tumor and also because it was only the existence of the ovarian tumor that prompted review of the thyroid neoplasm and its reinterpretation as carcinoma, a diagnosis not made initially. In another case (Brogioni et al. 2007), a 38-year-old woman with papillary carcinoma and local lymph node spread returned 7 years later with bilateral cystic ovarian metastases. In another case of papillary carcinoma, ovarian metastasis was diagnosed after 10 years (Logani et al. 2001). A review of the literature on parathyroid carcinoma has not disclosed any examples of ovarian metastasis. Rare examples of head and neck carcinoma metastatic to the ovary are documented, and we have seen one case in which the primary tumor was an undifferentiated carcinoma of the ethmoid sinus. Salivary gland tumors also spread to the ovary with extreme rarity. We have seen a case of a young woman who had an adenoid cystic carcinoma of the parotid gland excised at the age of 12 years followed by local recurrence, lung metastasis, and bilateral symptomatic ovarian metastases 11 years after presentation. Longacre and colleagues (1996) described a case of a 30-year-old woman with an adenoid cystic carcinoma of the submandibular gland who had a 10 cm left ovarian metastasis, followed by a smaller tumor in the opposite ovary 10 years later. These cases emphasize that a history of neoplasia of any type, even relatively remote, may be relevant in the evaluation of an unusual ovarian tumor. Esophageal cancer rarely spreads

1205

to the ovary. A case of metastatic esophageal adenocarcinoma to the ovary is briefly mentioned in one series (Riopel et al. 1999).119 There are only two reports to our knowledge in which ovarian spread of tumors of the central nervous system and cranium is mentioned. One was a case of metastatic meningioma, and the other was a metastatic medulloblastoma in a 4-year-old girl, in whose ovary “a cleft near the hilum was full of tumor cells” (Young et al. 1993a). Tumors of the skin, other than malignant melanoma, rarely spread to the ovary; clinically significant spread of Merkel cell tumor is documented (Eichhorn et al. 1993). One chordoma has metastasized to the ovary (Zukerberg and Young 1990). There are sporadic reports of a metastatic tumor involving an ovary that contains a primary ovarian neoplasm, and there seem to be no unique features of these happenstance events, a few examples of which we have seen (Fig. 90).

Female Genital Tract Tumors Tubal Carcinoma Discussion of this area has been made more challenging by recent evidence suggesting that many (in the opinion of some, most) ovarian serous carcinomas are of tubal origin (Kindelberger et al. 2007). This is discussed in greater detail elsewhere in this text. We personally believe most serous carcinomas dominant in the ovary arise there and categorize them primarily on where the greatest tumor bulk is located and the pattern of disease involvement. It should be emphasized that surface growth within the tube mimicking in situ neoplasia may be seen as a result of implantation from an ovarian carcinoma and does not necessarily indicate a tubal primary. Because of the great rarity of primary mucinous and clear cell carcinomas of the fallopian tube, a tumor of either of these cell types involving both organs is almost certainly primary in the ovary. When the morphology is endometrioid, an ovarian primary is still more likely although endometrioid carcinoma of the fallopian tube is much more common than either clear cell or mucinous.

1206

Endometrial Carcinoma Ovarian involvement in cases in which a diagnosis of endometrial carcinoma has been made has been reported in 34–40% of autopsy cases (Beck and Latour 1963; Bunker 1959) and 5–15% of hysterectomy and bilateral salpingooophorectomy specimens. Conversely, in approximately one-third of the cases in which a diagnosis of endometrioid carcinoma of the ovary has been made, an endometrial carcinoma also has been found. When the uterine corpus and the ovary are both involved by carcinomas, the question arises whether both cancers are primary or one is metastatic from the other (Eifel et al. 1982; Ulbright and Roth 1985; Zaino et al. 1984). If the endometrial carcinoma extends deeply into the myometrium with lymphatic or vascular invasion, if tumor is present in the lumen of the fallopian tube, or if tumor is on the ovarian surface or within its lymphatics or blood vessels, it is usually reasonable to conclude that the ovarian involvement is secondary. On the other hand, if lymphatic or hematogenous spread is absent, if the corpus carcinoma is small and limited to the endometrium or superficial myometrium, and if it has a background of atypical hyperplasia and the ovarian tumor has a background of endometriosis, the tumors probably are independent primaries. Criteria that are helpful in the determination of primary versus metastatic concomitant ovarian and endometrial carcinomas are presented in Table 2. Although synchronous ovarian and uterine tumors are of endometrioid type in most cases, occasionally they are of similar but other cell types, and rarely the histologic type of tumor is different in the two organs (Eifel et al. 1982). In some cases of synchronous involvement, it is impossible to establish the site of origin even after consideration of the features just described. Most synchronous, organ-confined ovarian and corpus carcinomas have been considered independent primary tumors. The favorable prognosis associated with such tumors supports the hypothesis that they are independent primaries. Recent molecular data, however, shows evidence of clonality in many cases of concurrent endometrioid carcinomas of the corpus and ovary,

M. F. Lerwill and R. H. Young

suggesting that they represent spread from one site to another (Anglesio et al. 2016; Schultheis et al. 2016); unique biological and/or microenvironmental factors may contribute to the lack of broader dissemination and the favorable prognosis. Despite this new perspective, synchronous organ-confined ovarian and uterine endometrioid carcinomas should continue to be clinically managed as early stage disease due to their generally high survival rates (Gilks and Kommoss 2018). Rarely, an adenocarcinoma of the uterine corpus with squamous differentiation deposits keratin or degenerated mature squamous cells on the serosal surface of one or both ovaries (Kim and Scully 1990). These are typically associated with a foreign body giant cell response. If no viableappearing tumor cells can be identified in these deposits on careful sampling, this finding does not appear to worsen the prognosis even when the keratin granulomas are also found elsewhere on the peritoneum.

Cervical Carcinoma There has been considerable recent interest in this topic (Elishaev et al. 2005; Ronnett et al. 2008; Young and Scully 1988b), stimulated by observations suggesting that it is more common in cases of adenocarcinoma than in cases of squamous cell carcinoma and raising the question of whether ovarian conservation is justified in patients with cervical adenocarcinoma. Occasional patients with cervical carcinomas of diverse types have clinically significant ovarian metastases (Young et al. 1993b). In an autopsy series by Tabata et al. (1987), ovarian metastases were detected in 104 of 597 (17%) cases of squamous cell carcinoma and in 22 of 77 (28.6%) cases of adenocarcinoma. The frequency of ovarian metastases of squamous cell carcinoma in their series is much higher than the 3% frequency in the prior literature. Ovarian metastases discovered during life are much rarer. In their series of 318 patients with stage IA cervical carcinoma treated by hysterectomy with ovarian preservation, Tabata et al. (1987) found no examples of subsequent ovarian metastasis during follow-up periods that exceeded 5 years in more

18

Metastatic Tumors of the Ovary

1207

Table 2 Criteria for interpretation of nature of concomitant uterine corpus and ovarian carcinomas Corpus primary Ovarian metastasis Direct extension to ovary from large corpus tumor Deep myometrial invasion from endometrium Lymphatic or blood vessel invasion in corpus, ovary, or both Atypical hyperplasia of endometrium frequent Tumor present in fallopian tube

Tumor predominant on surface of ovary

Usually no endometriosis in ovary Histological types uniform and consistent with corpus primary

Ovarian primary Corpus metastasis Direct extension to corpus from large ovarian primary

Ovarian primary Corpus primary No direct extension of either tumor

Ovarian metastasis Corpus metastasis Usually no direct extension of tumors

Myometrial invasion from serosal surface

Myometrial invasion usually absent or superficial

Lymphatic or blood vessel invasion in corpus, ovary, or both

No lymphatic or blood vessel invasion

Atypical hyperplasia of endometrium usually absent Tumor present on peritoneal surfaces and sometimes in fallopian tube Tumor predominant within ovary

Atypical hyperplasia of endometrium frequent

Tumor characteristically in endometrial stroma Lymphatic or blood vessel invasion frequent in ovary and corpus Atypical hyperplasia of endometrium absent Tumor usually evident outside female genital tract

Endometriosis sometimes present in ovary Histological types uniform and consistent with ovarian primary

Endometriosis sometimes present in ovary Histological types uniform or dissimilar

Usually both tumors confined to primary sites or have spread minimally Tumor predominant within ovary and endometrium

than half the cases. In cases of stage IB, II, and III carcinoma in their series, there were no ovarian metastases in 278 cases of squamous cell carcinoma in contrast to 6 ovarian metastases of 48 (12.5%) cases of adenocarcinoma. In two large series of patients with stage IB or higher disease, the incidence of ovarian metastases ranged from 0.5% to 1.5% (Shimada et al. 2006; Toki et al. 1991). Ovarian metastases were found in approximately 5% of adenocarcinomas and < 1% of squamous cell carcinomas. Shimada et al. (2006) identified ovarian metastases in 3.72% of stage IB, 5.26% of stage IIA, and 9.85% of stage IIB adenocarcinomas and 0.22%, 0.75%, and 2.17%, respectively, of squamous cell carcinomas. Spread

Uncertain primary Massive involvement of both organs or conflicting findings listed in first four columns Myometrial invasion may be present

Ovarian tumor usually bilateral Ovarian surface involvement frequent Endometriosis absent Type of tumor inconsistent with or unusual for either organ

to the ovaries is often associated with extension of cancer into the uterine corpus (Reyes et al. 2015; Ronnett et al. 2008; Tabata et al. 1987; Toki et al. 1991) and mucosal involvement of the fallopian tube (Reyes et al. 2015).

Adenocarcinomas Ovarian metastasis of endocervical adenocarcinoma of the usual type is encountered most often, likely due to its greater incidence, but the rarer gastric-type adenocarcinoma also exhibits ovarian spread with some frequency (Karamurzin et al. 2015). The primary cervical tumors may only be microinvasive or even have features that objectively make recognition of definite invasion

1208

questionable. Contiguous spread to the corpus may play a role by heightening the chance of transtubal spread to the ovarian surface (Ronnett et al. 2008). The ovarian tumors are bilateral in only about one-third of the cases and are often over 10 cm. Given that usual-type endocervical adenocarcinoma has a somewhat distinctive histologic appearance, it is not surprising that this morphology is also seen in the ovaries (Figs. 107, 108, and 109). Approaching the topic from the ovarian perspective, the picture is more often pseudoendometrioid than truly mucinous in nature, and yet the morphology is not typical of classic endometrioid glandular neoplasia. As noted by Ronnett et al. (2008), the tumors often have a hybrid appearance with low-power endometrioid-like features but with apical mucin appreciable on higher power. The nuclei are typically hyperchromatic and elongated (Fig. 109), more atypical than seen in true endometrioid carcinomas with a similar degree of gland differentiation. Apically located mitotic figures are usually

Fig. 107 Metastatic endocervical adenocarcinoma of the usual type. There is a pseudoendometrioid picture

M. F. Lerwill and R. H. Young

prominent, and apoptotic cells are often numerous. p16 immunohistochemical staining (Fig. 108) supports a secondary nature but is not specific. Should p16 be positive, more definitive evidence can be obtained if necessary by evaluating the ovarian tumor for human papillomavirus using in situ hybridization or polymerase chain reaction. Cases have been documented in which these results have prompted evaluation of a clinically non-suspicious cervix and subsequent identification of the occult primary neoplasm (Ronnett et al. 2008). In a series from the Armed Forces Institute of Pathology, as many as 10% of mucinous adenocarcinomas of the cervix were reported to metastasize to the ovary (Fig. 110) (Kaminski and Norris 1984), although the referral nature of that material may have introduced some bias; occasional other examples have been reported in detail (Young and Scully 1988b) or included in series of metastatic mucinous carcinomas in the ovary (Riopel et al. 1999). Eight of the 23 (35%) gastric-type endocervical adenocarcinomas showed ovarian metastases in one study (Karmurzin et al. 2015).

Squamous Cell Carcinomas The ovarian metastasis may be discovered synchronously at the time of treatment for the cervical cancer or occur up to 10 years after treatment of the cervical primary tumor. In one case, the primary cervical tumor was not discovered until autopsy 7 months after the patient had been treated for a squamous cell carcinoma involving the left ovary (Young et al. 1993b). The cervical tumor in that case was invasive only to 3.8 mm, whereas in the most cases, there is deep infiltration of the cervical wall with frequent extrauterine extension of the neoplasm (Young et al. 1993b). Clinically significant ovarian metastases have ranged from 5 to 17 cm in greatest dimension. They can be solid, solid and cystic, or cystic. Microscopic examination shows the typical features of squamous cell carcinoma except that many of the tumors have striking cystification within the squamous nests. In one case, a cervical squamous cell carcinoma that was invasive to only 1.2 mm extended in an in situ manner to involve the endometrium and involved the surface

18

Metastatic Tumors of the Ovary

1209

Fig. 108 Metastatic endocervical adenocarcinoma of the usual type. Companion staining with hematoxylin and eosin (left) and for p16 (right)

and inclusion glands and cysts within one ovary, tumor cells presumably having spread there via the tubal lumen. In one remarkable case, a cervical squamous cell carcinoma in situ was associated with contiguous spread to the endometrium, fallopian tubes, and ovaries, extensively replacing the endometrial and tubal epithelium and focally invading the wall of the tubes and parenchyma of both ovaries (Pins et al. 1997). The differentiation of metastatic squamous cell carcinoma from the cervix from primary squamous cell carcinoma of the ovary usually has been aided by the knowledge of the presence of a cervical tumor, but in some cases the identification of the ovarian tumor antedates knowledge of the cervical primary (Young et al. 1993b). Before the diagnosis of primary squamous cell carcinoma of the ovary is made, the possibility of spread from a cervical tumor should be considered unless obvious features of primary neoplasia are immediately obvious. As most squamous cell carcinomas of the ovary arise on the background of a

preexistent neoplasm such as a dermoid or endometriotic cyst, thorough sampling to identify such a component may be crucial in determining the primary nature of the neoplasm. Although the evidence strongly points to the ovarian tumors being metastatic when both organs have been involved by squamous cell carcinoma, the rare association of squamous cell carcinoma of the ovary with squamous cell carcinoma in situ of the cervix leaves open the possibility of independent primary neoplasms in some cases. Although the differential in these cases is generally between primary and metastatic squamous cell carcinoma, in some cases metastatic squamous cell carcinoma undergoes cystic degeneration, and as squamous and transitional cell types are closely related, a resemblance to primary transitional cell carcinoma may result (Fig. 111).

Other Carcinomas Two adenosquamous carcinomas and two glassy cell carcinomas with ovarian metastases have

1210

Fig. 109 Metastatic endocervical adenocarcinoma of the usual type. Typical cytologic features

been reported (Young et al. 1993b). Both metastatic adenosquamous carcinomas were discovered at the same time as the cervical primary tumors. The ovarian tumors were bilateral in both cases, and the cervical tumors were deeply invasive with extracervical extension, findings facilitating the diagnosis. In one of the cases of glassy cell carcinoma, the ovarian involvement was a microscopic finding; in the other the ovarian involvement was grossly evident, but the ovary was not enlarged. Rare cases of ovarian metastases of cervical small cell carcinomas or mixed tumors with a component of adenocarcinoma and small cell carcinoma or high-grade neuroendocrine tumor have been reported. In one report of four such cases, the patients were from 23 to 34 years of age, and the ovarian spread had manifest clinically (Young et al. 1993b). One patient had evidence of carcinoid syndrome. The cervical tumors and ovarian tumors were synchronous findings in two cases; in the other two, the ovarian tumors were discovered 10 months and 3 years after the cervical tumors.

M. F. Lerwill and R. H. Young

Fig. 110 Metastatic endocervical adenocarcinoma, mucinous type. Highly differentiated glands of the type seen in adenoma malignum

Another report documents two cases of mixed endocervical adenocarcinoma and high-grade neuroendocrine carcinoma in which the metastases to the ovaries were comprised of the adenocarcinoma component only (Ramalingam et al. 2012). In one unusual case, a cervical transitional cell carcinoma metastasized to the ovary and was detected before the cervical tumor (Young et al. 1993b). The metastatic tumor in the ovary in this case was a large cystic mass that was indistinguishable microscopically from a primary transitional cell carcinoma of the ovary but was associated with prominent vascular space invasion, suggesting its metastatic nature.

Other Uterine Tumors The most important of these by far is metastatic endometrial stromal sarcoma because it has some proclivity for ovarian spread and can cause a wide array of problems in differential diagnosis. This

18

Metastatic Tumors of the Ovary

1211

Fig. 112 Metastatic endometrial stromal sarcoma. Peculiar nodular growth

Fig. 111 Metastatic squamous cell carcinoma from cervix. The lining of cysts by thick undulating bands of neoplastic cells without overt squamous differentiation imparts a resemblance to a primary transitional cell carcinoma of the ovary

sarcoma metastasizes to the ovary more frequently than leiomyosarcoma. In a series of 11 uterine sarcomas that metastasized to the ovary (none of which were autopsy findings), 8 were endometrial stromal sarcomas (Young and Scully 1990b). The patients ranged from 33 to 79 (average, 50) years of age; 5 of them were less than 50 years old. The ovarian metastases accounted for the clinical presentation in three of the patients. In two patients, the primary uterine tumors were not discovered until 7 months and 10 months after bilateral ovarian tumors had been resected. In four of the other cases, the ovarian and uterine tumors were found synchronously, and in the remaining two, the ovarian metastases occurred 4 to 9 years after the uterine neoplasms had been discovered. The ovarian tumors were bilateral in six cases and ranged up to 17 cm in greatest dimension. These tumors usually are solid (Fig. 112) or solid and cystic but rarely are cystic.

Fig. 113 Metastatic endometrial stromal sarcoma. The tumor has a diffuse pattern

The major problem with the interpretation of these tumors on microscopic examination is that, in the ovary, the tongue-like pattern of infiltration characteristic of this neoplasm often is inconspicuous. A diffuse pattern is common (Fig. 113).

1212

M. F. Lerwill and R. H. Young

Fig. 114 Metastatic endometrial stromal sarcoma. Typical neoplasia at the top and unusual fibroma-like morphology at the bottom

Fig. 115 Metastatic endometrial stromal sarcoma. Hyaline plaques are conspicuous and may cause the diagnosis of thecoma to be entertained

Other diagnostic problems result from the presence in some of the tumors of large fibromatous areas (Fig. 114) and hyaline plaques (Fig. 115). Yu et al. (1986) described the case of a 24-year-old woman with a metastatic endometrial stromal sarcoma that was misinterpreted initially as a thecoma for this reason. In other cases with a diffuse pattern, the characteristic small arteries resembling the spiral arteries of the endometrium are inconspicuous, resulting in a resemblance to a diffuse granulosa cell tumor. Confusion with sex cord–stromal tumors may be heightened by the occasional presence of areas of sex cord-like differentiation in metastatic endometrial stromal sarcomas. However, high-power examination does not show the typical nuclear features of granulosa cell tumors, and careful examination usually demonstrates, at least focally, the typical vascular pattern of endometrial stromal tumors (Fig. 116).

Bilaterality also is far more common in metastatic endometrial stromal sarcoma as is the presence of extraovarian tumor. Metastatic endometrial stromal sarcomas in the ovary must be distinguished from primary endometrioid stromal sarcomas (Oliva et al. 2014; Young et al. 1984). An association of the tumor with endometriosis is evidence for an ovarian origin. Bilaterality favors metastasis, although it is possible that some bilateral tumors may represent independent primaries. Leiomyosarcomas of the uterus with ovarian metastases probably are more common than the rare reports in the literature suggest, particularly in patients with widespread disease. The three leiomyosarcomas with ovarian metastases in one series occurred in patients 35, 44, and 49 years of age (Young and Scully 1990b). In the first patient, a large ovarian metastatic tumor became

18

Metastatic Tumors of the Ovary

1213

Fig. 116 Metastatic endometrial stromal sarcoma. Typical arterioles are present (same tumor as in Fig. 113)

Fig. 117 Ovarian involvement by malignant mesothelioma of peritoneum. Note typical tubulopapillary pattern

symptomatic 14 months after hysterectomy. In the second patient, the ovarian metastatic tumor occurred in the setting of widespread disease. In the third case, the ovarian involvement was only microscopic. Ovarian spread of malignant mixed Müllerian tumors is common but not usually a diagnostic problem. Ovarian involvement in cases of Müllerian adenosarcoma of the uterus is uncommon and in some cases may be an independent primary, particularly if it is associated with endometriosis. Gestational choriocarcinoma of the uterus may spread to the ovary but in the light of known uterine disease is not a diagnostic problem (Acosta-Sison 1958). If such an ovarian tumor in a woman of childbearing age is not clearly metastatic from a uterine or tubal choriocarcinoma, thorough sampling may be needed to demonstrate the presence or absence of teratomatous elements. If such elements are not found, it may be difficult or impossible to differentiate between a primary choriocarcinoma of the ovary of either gestational or germ cell origin and a metastatic tumor from a

choriocarcinoma of the uterus that has regressed. Invasive hydatidiform mole also has been documented to spread to the ovary, as have rare cases of placental site trophoblastic tumor (AbdulHafeez et al. 1987; Milingos et al. 2007).

Vulvar and Vaginal Tumors Vulvar and vaginal carcinomas rarely exhibit ovarian spread. Occasional vaginal clear cell adenocarcinomas have metastasized to the ovary, in most cases associated with extensive pelvic spread.

Peritoneal Tumors Although ovarian involvement in cases of peritoneal serous carcinoma may be secondary in most of the cases, this subject is generally not included in discussions of secondary tumors of the ovary because the ovarian involvement is only one part

1214

of widespread peritoneal disease. More pertinent to our interest here are cases of significant ovarian involvement in cases of malignant mesothelioma which may engender a broader array of problems in differential diagnosis (Baker et al. 2005). In one series of peritoneal mesotheliomas, ovarian involvement was common (Goldblum and Hart 1995), and this phenomenon was the focus of another report that included seven cases of peritoneal malignant mesothelioma that presented clinically as “ovarian cancer” (Clement et al. 1996). The differential diagnosis in these cases is primarily with an ovarian surface epithelial carcinoma, particularly serous carcinoma. Although there is some overlap, in typical cases the tubulopapillary (Fig. 117) and diffuse patterns of mesothelioma, and the characteristic cuboidal to rounded cells with abundant eosinophilic cytoplasm (Fig. 118), produce a picture that is distinctly different from that of serous carcinoma. Psammoma bodies, although occasionally seen in mesotheliomas, are rarely numerous and if present in significant

Fig. 118 Ovarian involvement by malignant mesothelioma of peritoneum. Note typical eosinophilic cytoplasm and relatively bland cytology

M. F. Lerwill and R. H. Young

number favor a serous neoplasm. Immunohistochemical staining for a panel of markers such as calretinin, estrogen receptor, MOC-31, and Ber-EP4 can be helpful in difficult cases (Ordóñez 2006). Of note, PAX8 is positive in a minority of peritoneal mesotheliomas (6–18%) and is thus not specific for serous carcinoma when entertaining this differential diagnosis (Chapel et al. 2017; Tandon et al. 2018). The broad spectrum of morphology of peritoneal mesothelioma may result in other diverse but generally rare issues in differential diagnosis, should a case be dominated by ovarian involvement. This topic has been reviewed in detail by Baker et al. (2005), and only two issues that have struck us as being perhaps more realistic than most are briefly noted here. A superficial resemblance to clear cell carcinoma may be imparted in some cases by the tubular and papillary patterns of mesothelioma, but the latter is generally less cytologically atypical, and when the spectrum of morphology is evaluated, the two neoplasms should be distinguishable. A malignant mixed mesodermal tumor may be entertained

Fig. 119 Ovarian involvement by intraabdominal desmoplastic small round cell tumor. Nested pattern

18

Metastatic Tumors of the Ovary

1215

shows large nodules, smaller nests, and clusters composed predominantly of small cells with hyperchromatic nuclei and scanty cytoplasm surrounded by a prominent desmoplastic stroma (Figs. 119 and 120). The neoplasms exhibit the characteristic immunohistochemical staining profile, with many of the tumor cells staining for cytokeratin, epithelial membrane antigen, desmin, and vimentin. The differential diagnosis in these cases is extensive and is primarily related to a number of small cell tumors that may involve the ovary, either primarily or secondarily in young females, as presented in detail elsewhere (Young and Scully 1990). Rarely other problems in diagnosis may be encountered related to unusual findings such as focal tubular differentiation or cells with more appreciable eosinophilic cytoplasm than is typical. Generally these are only focal findings in neoplasms dominated by more distinctive morphology.

References Fig. 120 Ovarian involvement by intraabdominal desmoplastic small round cell tumor. In this illustration, desmoplastic stroma is conspicuous and, in the setting of widespread abdominal disease, is a strong clue to the diagnosis, particularly in a young female

when a mesothelioma has a spindle cell morphology or a prominent myxoid stroma, but the mesothelial cells almost always are less pleomorphic than the high-grade carcinomatous component of a malignant mixed mesodermal tumor. Another peritoneal tumor that may have ovarian manifestations as a major component of the clinical presentation and present as “ovarian cancer” is the intraabdominal desmoplastic small round cell tumor with divergent differentiation. A small number of these tumors with ovarian involvement at presentation have been described in patients in the second and third decades (Elhajj et al. 2002; Parker et al. 2002; Young et al. 1992b; Zaloudek et al. 1995). In some cases the ovarian tumor initially was thought to be the primary neoplasm. In all the cases, there was extensive extraovarian tumor at the time of presentation. The ovarian involvement is usually bilateral. Microscopic examination of the ovarian tumors

Abdul-Hafeez M, Akhtar M, Aqeel HSB, Kidess EA (1987) Placental site trophoblastic tumor: report of a case with review of literature. Ann Saudi Med 7:340–344 Acosta-Sison H (1958) The relative frequency of various anatomic sites as the point of first metastasis in 32 cases of chorionepithelioma. Am J Obstet Gynecol 75:1149–1152 Agoff SN, Lamps LW, Philip AT, Amin MB, Schmidt RA, True LD, Folpe AL (2000) Thyroid transcription factor-1 is expressed in extrapulmonary small cell carcinomas but not in other extrapulmonary neuroendocrine tumors. Mod Pathol 13:238–242 Alvarado-Cabrero I, Rodríguez-Gómez A, CastelanPedraza J, Valencia-Cedillo R (2013) Metastatic ovarian tumors. A clinicopathologic study of 140 cases. Anal Quant Cytopathol Histopathol 35:241–248 Amorn Y, Knight WA (1978) Primary linitis plastica of the colon: report of two cases and review of the literature. Cancer (Phila) 41:2420–2425 Anglesio MS, Wang YK, Maassen M, Horlings HM, Bashashati A, Senz J, Mackenzie R, Grewal DS, Li-Chang H, Karnezis AN, Sheffield BS, McConechy MK, Kommoss F, Taran FA, Staebler A, Shah SP, Wallwiener D, Brucker S, Gilks CB, Kommoss S, Huntsman DG (2016) Synchronous endometrial and ovarian carcinomas: evidence of clonality. J Natl Cancer Inst 108:djv428. https://doi.org/10.1093/jnci/djv428 Baker PM, Oliva E (2004) Immunohistochemistry as a tool in the differential diagnosis of ovarian tumors: an update. Int J Gynecol Pathol 24:39–55

1216 Baker PM, Clement PB, Young RH (2005) Malignant peritoneal mesothelioma in women. A study of 75 cases with emphasis on their morphologic spectrum and differential diagnosis. Am J Clin Pathol 123:724–737 Beck RP, Latour JPA (1963) Necropsy reports on 36 cases of endometrial carcinoma. Am J Obstet Gynecol 85:307–311 Bennett JA, Young RH, Chuang AY, Lerwill MF (2017) Ovarian metastases of breast cancers with signet ring cells: a report of 17 cases including 14 Krukenberg tumors. Int J Gynecol Pathol (in press). https://doi. org/10.1097/PGP.0000000000000462 Berg KB, Schaeffer DF (2017) SATB2 as an immunohistochemical marker for colorectal adenocarcinoma. A concise review of benefits and pitfalls. Arch Pathol Lab Med 141:1428–1433 Birnkrant A, Sampson J, Sugarbaker PH (1986) Ovarian metastasis from colorectal cancer. Dis Colon Rectum 29:767–771 Bombonati A, Lerwill MF (2012) Metastases to and from the breast. Surg Pathol Clin 5:179–747 Brogioni S, Viacava P, Tomisti L, Martino E, Macchia E (2007) A special case of bilateral metastases in a woman with papillary carcinoma of the thyroid. Exp Clin Endocrinol Diabetes 115:397–400 Brown BL, Scharifker DA, Gordon R, Deppe GG, Cohen CJ (1980) Bronchial carcinoid tumor with ovarian metastasis. A light microscopic and ultrastructural study. Cancer 46:543–546 Bullon A, Arseneau J, Prat J, Young RH, Scully RE (1981) Tubular Krukenberg tumor. A problem in histopathologic diagnosis. Am J Surg Pathol 5:225–232 Bunker ML (1959) The terminal findings in endometrial carcinoma. Am J Obstet Gynecol 77:530–538 Byrne DJ, Deb S, Takano EA, Fox SB (2017) GATA3 expression in triple-negative breast cancers. Histopathol 71:63–71 Cameron RI, Ashe P, O’Rourke DM, Foster H, McCluggage WG (2003) A panel of immunohistochemical stains assists in the distinction between ovarian and renal clear cell carcinoma. Int J Gynecol Pathol 22:272–276 Cao D, Ji H, Ronnett BM (2004) Expression of mesothelin, fascin, and prostate stem cell antigen in primary ovarian mucinous tumors and their utility in differentiating primary ovarian mucinous tumors from metastatic pancreatic mucinous carcinomas of the ovary. Int J Gynecol Pathol 24:67–72 Carlson JW, Nucci MR, Brodsky J, Crum CP, Hirsch MS (2007) Biomarker-assisted diagnosis of ovarian, cervical and pulmonary small cell carcinomas: the role of TTF-1, WT-1 and HPV analysis. Histopathology 51:305–312 Caron P, Roche H, Gorguet B, Martel P, Bennet A, Carton M (1990) Mammary ovarian metastases with stromal cell hyperplasia and postmenopausal virilization. Cancer (Phila) 66:1221–1224 Chapel DB, Husain AN, Krausz T, McGregor SM (2017) PAX8 expression in a subset of malignant peritoneal

M. F. Lerwill and R. H. Young mesotheliomas and benign mesothelium has diagnostic implications in the differential diagnosis of ovarian serous carcinoma. Am J Surg Pathol 41:1675–1682 Che M, Tornos C, Deavers MT (2001) Ovarian mixedepithelial carcinomas with a microcystic pattern and signet-ring cells. Int J Gynecol Pathol 20:323–328 Chen ZME, Wang HL (2004) Alteration of cytokeratin 7 and cytokeratin 20 expression profile is uniquely associated with tumorigenesis of primary adenocarcinoma of the small intestine. Am J Surg Pathol 28:1352–1359 Clement PB, Young RH, Scully RE (1996) Malignant mesotheliomas presenting as ovarian masses. A report of nine cases, including two primary ovarian mesotheliomas. Am J Surg Pathol 20:1067–1080 Creasman WT, Lukeman J (1972) Role of the fallopian tube in dissemination of malignant cells in corpus cancer. Cancer (Phila) 20:456–457 DeCostanzo DC, Elias JM, Chumas JC (1997) Necrosis in 84 ovarian carcinomas: a morphologic study of primary versus metastatic colonic cancer with a selective immunohistochemical analysis of cytokeratin subtypes and carcinoembryonic antigen. Int J Gynecol Pathol 16:245–249 Desouki MM, Lioyd J, Xu H, Cao D, Barner R, Zhao C (2013) CDX2 may be a useful marker to distinguish primary ovarian carcinoid from gastrointestinal metastatic carcinoids to the ovary. Hum Pathol 44:2536–2541 DiLoreto C, Di Lauro V, Puglisi F et al (1997) Immunocytochemical expression of tissue specific transcription factor-1 in lung carcinoma. J Clin Pathol 50:30–32 Eichhorn JH, Young RH, Scully RE (1993) Non-pulmonary small cell carcinomas of extragenital origin metastatic to the ovary: a report of seven cases. Cancer (Phila) 71:177–186 Eifel P, Hendrickson M, Ross J, Ballon S, Martinez A, Kempson R (1982) Simultaneous presentation of carcinoma involving the ovary and the uterine corpus. Cancer (Phila) 50:163–170 Elhajj M, Mazurka J, Daya D (2002) Desmoplastic small round cell tumor presenting in the ovaries: report of a case and review of the literature. Int J Gynecol Cancer 12:760–763 Elishaev E, Gilks CB, Miller D, Srodon M, Kurman RJ, Ronnett BM (2005) Synchronous and metachronous endocervical and ovarian neoplasms. Evidence supporting interpretation of the ovarian neoplasms as metastatic endocervical adenocarcinomas simulating primary ovarian surface epithelial neoplasms. Am J Surg Pathol 29:281–294 Fitzgibbons PL, Martin SE, Simmons TJ (1987) Malignant melanoma metastatic to the ovary. Am J Surg Pathol 11:959–964 Gagnon Y, Tetu B (1989) Ovarian metastases of breast carcinoma. A clinicopathologic study of 59 cases. Cancer (Phila) 64:892–898 Gilks CB, Kommoss F (2018) Synchronous tumors of the female reproductive tract. Pathol 50:214–221

18

Metastatic Tumors of the Ovary

Gill AJ, Chou A, Vilain R, Clarkson A, Lui M, Jin R, Tobias V, Samra J, Goldstein D, Smith C, Sioson L, Parker N, Smith RC, Sywak M, Sidhu SB, Wyatt JM, Robinson BG, Eckstein RP, Benn DE, Clifton-Bligh RJ (2010) Immunohistochemistry for SDHB divides gastrointestinal stromal tumors (GISTs) into 2 distinct types. Am J Surg Pathol 34:636–644 Giordano G, Cruz Viruel N, Silini EM, Nogales FF (2017) Adenocarcinoma of the lung metastatic to the ovary with a signet ring cell component. Int J Surg Pathol 25:365–367 Gokden N, Gokden M, Phan DC, McKenney JK (2008) The utility of PAX-2 in distinguishing metastatic clear cell renal cell carcinoma from its morphologic mimics. An immunohistochemical study with comparison to renal cell carcinoma marker. Am J Surg Pathol 32:1462–1467 Goldblum J, Hart WR (1995) Localized and diffuse mesotheliomas of the genital tract and peritoneum in women. A clinicopathological study of nineteen true mesothelial neoplasms, other than adenomatoid tumors, multicystic mesotheliomas and localized fibrous tumors. Am J Surg Pathol 19:1124–1137 Graffner HOL, Alm POA, Oscarson JEA (1983) Prophylactic oophorectomy in colorectal carcinoma. Am J Surg 146:233–235 Green LK, Silva EG (1989) Hepatoblastoma in an adult with metastasis to the ovaries. Am J Clin Pathol 92:110–115 Gupta D, Deavers MT, Silva EG et al (2004) Malignant melanoma involving the ovary. A clinicopathologic and immunohistochemical study of 23 cases. Am J Surg Pathol 28:771–780 Hagemann IS, Pfeifer JD, Cao D (2013) Mammaglobin expression in gynecologic adenocarcinomas. Hum Pathol 44:628–635 Hale RW (1968) Krukenberg tumor of the ovaries. A review of 81 records. Obstet Gynecol 32:221–225 Harris M, Howell A, Chrissohou M, Swindell RIC, Hudson M, Sellwood RA (1984) A comparison of the metastatic pattern of infiltrating lobular carcinoma and infiltrating duct carcinoma of the breast. Br J Cancer 50:23–30 Hayashi Y, Kikuchi F, Oka T et al (1988) Rhabdomyosarcoma with bone marrow metastasis simulating acute leukemia. Report of two cases. Acta Pathol Jpn 38:789–798 Hecht JL, Pinkus JL, Weinstein LJ et al (2001) The value of thyroid transcription factor-1 in cytologic preparations as a marker for metastatic adenocarcinoma of lung origin. Am J Clin Pathol 116:483–488 Holtz F, Hart WR (1982) Krukenberg tumors of the ovary. A clinicopathologic analysis of 27 cases. Cancer (Phila) 50:2438–2447 Hopping RA, Dockerty MB, Masson JC (1942) Carcinoid tumor of the appendix. Report of a case in which extensive intra-abdominal metastases occurred, including involvement of the right ovary. Arch Surg 45:613–622

1217 Hristov AC, Young RH, Vang R, Yemelyanova AV, Seidman JD, Ronnett BM (2007) Ovarian metastases of appendiceal tumors with goblet cell carcinoid-like and signet ring cell patterns: a report of 30 cases. Am J Surg Pathol 31:1502–1511 Hsiu J-G, Kemp GM, Singer GA, Rawls WH, Siddiky MA (1991) Transitional cell carcinoma of the renal pelvis with ovarian metastasis. Gynecol Oncol 41:178–181 Insabato L, DeRosa G, Franco R et al (2003) Ovarian metastasis from renal cell carcinoma: a report of three cases. Int J Surg Pathol 11:309–312 Irving JA, Young RH (2005) Lung carcinoma metastatic of the ovary. A clinicopathologic study of 32 cases emphasizing their morphologic spectrum and problems in differential diagnosis. Am J Surg Pathol 29:997–1006 Irving JA, Lerwill MF, Young RH (2005) Gastrointestinal stromal tumors metastatic to the ovary. A report of five cases. Am J Surg Pathol 29:920–926 Irving JA, Vasques DR, McGuinness TB, Young RH (2006) Krukenberg tumor of renal pelvic origin: report of a case with selected comments on ovarian tumors metastatic from the urinary tract. Int J Gynecol Pathol 25:147–150 Ji H, Isacson C, Seidman JD, Kurman RJ, Ronnett BM (2002) Cytokeratins 7 and 20, Dpc4, and MUC5AC in the distinction of metastatic mucinous carcinomas in the ovary from primary ovarian mucinous tumors: Dpc4 assists in identifying metastatic pancreatic carcinomas. Int J Gynecol Pathol 21:391–400 Judson K, McCormick C, Vang R, Yemelyanova AV, Wu LSF, Bristow RE, Ronnett BM (2008) Women with undiagnosed colorectal adenocarcinomas presenting with ovarian metastases: Clinicopathologic features and comparison with women having known colorectal adenocarcinomas and ovarian involvement. Int J Gynecol Pathol 27:182–190 Kaminski PF, Norris HJ (1984) Coexistence of ovarian neoplasms and endocervical adenocarcinoma. Obstet Gynecol 64:553–556 Kandalaft PL, Gown AM, Isacson C (2014) The lungrestricted marker napsin a is highly expressed in clear cell carcinomas of the ovary. Am J Clin Pathol 142:830–836 Karamurzin YS, Kiyokawa T, Parkash V, Jotwani AR, Patel P, Pike MC, Soslow RA, Park KJ (2015) Gastric-type endocervical adenocarcinoma: an aggressive tumor with unusual metastatic patterns and poor prognosis. Am J Surg Pathol 39:1449–1457 Kaufmann O, Dietel M (2000) Thyroid transcription factor-1 is the superior immunohistochemical marker for pulmonary adenocarcinomas and large cell carcinomas compared to surfactant proteins a and B. Histopathology 36:8–16 Kershisnik MM, Ro JY, Cannon GH, Ordóñez NG, Ayala AG, Silva EG (1995) Histiocytic reaction in pelvic peritoneum associated with oxidized regenerated cellulose. Am J Clin Pathol 103:27–31 Khunamornpong S, Siriaunkgul S, Chunduan A (1999) Metastatic hepatocellular carcinoma of the ovary. Int J Gynecol Obstet 64:189–191

1218 Khunamornpong S, Suprasert PS, Pojchamarnwiputh S, Chiangmai WN, Settakorn J, Siriaunkgul S (2006a) Primary and metastatic mucinous adenocarcinomas of the ovary: evaluation of the diagnostic approach using tumor size and laterality. Gynecol Oncol 101:152–157 Khunamornpong S, Siriaunkgul S, Suprasert P, Pojchamarnwiputh S, Chiangmai WN, Young RH (2007) Intrahepatic cholangiocarcinoma metastatic to the ovary: a report of 16 cases of an underemphasized form of secondary tumor of the ovary that may mimic primary neoplasia. Am J Surg Pathol 31:1788–1799 Khunamornpong S, Lerwill MF, Siriaunkgul S, Suprasert P, Pojchamarnwiputh S, Chiangmai WN, Young RH (2008) Carcinoma of extrahepatic bile ducts and gallbladder metastatic to the ovary: a report of 16 cases. Int J Gynecol Pathol 27:366–379 Kim K-R, Scully RE (1990) Peritoneal keratin granulomas with carcinomas of endometrium and ovary and atypical polypoid adenomyoma of endometrium. A clinicopathological analysis of 22 cases. Am J Surg Pathol 14:925–932 Kim KR, Lee HI, Lee SK, Ro JY, Robboy SJ (2007) Is stromal microinvasion in primary mucinous ovarian tumors with “mucin granuloma” true invasion? Am J Surg Pathol 31:546–554 Kim CJ, Baruch-Oren T, Lin F, Fan XS, Yang XJ, Wang HL (2016) Value of SATB2 immunostaining in the distinction between small intestinal and colorectal adenocarcinomas. J Clin Pathol 69:1046–1050 Kindelberger DW, Lee Y, Miron A, Hirsch MS, Feltmate C, Medeiros F, Callahan MJ, Garner EO, Gordon RW, Birch C, Berkowitz RS, Muto MG, Crum CP (2007) Intraepithelial carcinoma of the fimbria and pelvic serous carcinoma: evidence for a causal relationship. Am J Surg Pathol 31:161–169 Kiyokawa T, Young RH, Scully RE (2006) Krukenberg tumors of the ovary. A clinicopathologic analysis of 120 cases with emphasis on their variable pathologic manifestations. Am J Surg Pathol 31:277–299 Kubba LA, McCluggage WG, Liu J, Malpica A, Euscher ED, Silva EG, Deavers MT (2008) Thyroid transcription factor-1 expression in ovarian epithelial neoplasms. Mod Pathol 21:485–490 Kuo T-T, Hsueh S (1984) Mucicarminophilic histiocytosis. A polyvinylpyrrolidone (PVP) storage disease simulating signet-ring cell carcinoma. Am J Surg Pathol 8:419–428 Kurek R, Von Knoblock R, Feek U, Heidenreich A, Hofmann R (2001) Local recurrence of an oncocytic adrenocortical carcinoma with ovary metastasis. J Urol 166:985 Lash RH, Hart WR (1987) Intestinal adenocarcinomas metastatic to the ovaries. A clinicopathological evaluation of 22 cases. Am J Surg Pathol 11:114–121 Laury AR, Peret R, Piao H, Krane JF, Barletta JA, French C, Chirieac LR, Lis R, Loda M, Hornick JL, Drapkin R, Hirsch MS (2011) A comprehensive analysis of PAX8 expression in human epithelial tumors. Am J Surg Pathol 35:816–826

M. F. Lerwill and R. H. Young Lee KR, Young RH (2003) The distinction between primary and metastatic mucinous carcinomas of the ovary. Am J Surg Pathol 2:281–292 Lee MJ, Lee HS, Kim WH, Choi Y, Yang M (2003) Expressions of mucins and cytokeratins in primary carcinomas of the digestive system. Mod Pathol 16:403–410 Lee AH, Paish EC, Marchio C, Sapino A, Schmitt FC, Ellis IO, Reis-Filho JS (2007) The expression of Wilms’ tumour-1 and Ca125 in invasive micropapillary carcinoma of the breast. Histopathology 51:824–828 Lee S-J, Bae J-H, Lee A-W, Tong S-Y, Park Y-G, Park J-S (2009) Clinical characteristics of metastatic tumors to the ovaries. J Korean Med Sci 24:114–119 Leroy X, Farine MO, Buob D, Wacrenier A, Copin MC (2007) Diagnostic value of cytokeratin 7, CD10 and mesothelin in distinguishing ovarian clear cell carcinoma from metastasis of renal clear cell carcinoma. Histopathology 51:846–876 Lerwill MF, Young RH (2006) Ovarian metastases of intestinal-type gastric carcinoma. A clinicopathologic study of 4 cases with contrasting features to those of the Krukenberg tumor. Am J Surg Pathol 31:1382–1388 Lewis MR, Deavers MT, Silva EG, Malpica A (2006) Ovarian involvement by metastatic colorectal adenocarcinoma. Still a diagnostic challenge. Am J Surg Pathol 30:177–184 Liang L, Huang H, Dadhania V, Zhang J, Zhang M, Liu J (2016) Renal cell carcinoma metastatic to the ovary or fallopian tube: a clinicopathological study of 9 cases. Hum Pathol 51:96–102 Logani S, Baloch ZW, Snyder PJ, Weinstein R, LiVolvi VA (2001) Cystic ovarian metastasis from papillary thyroid carcinoma: a case report. Thyroid 11:1073–1075 Logani S, Oliva E, Amin MB, Folpe AL, Cohen C, Young RH (2003) Immunoprofile of ovarian tumors with putative transitional cell (urothelial) differentiation using novel urothelial markers. Histogenic and diagnostic implications. Am J Surg Pathol 27:1434–1441 Logani S, Oliva E, Arnell PM, Amin MB, Young RH (2005) Use of novel immunohistochemical markers expressed in colonic adenocarcinoma to distinguish primary ovarian tumors from metastatic colorectal carcinoma. Mod Pathol 18:19–25 Longacre TA, O’Hanlan K, Hendrickson MR (1996) Adenoid cystic carcinoma of the submandibular gland with symptomatic ovarian metastases. Int J Gynecol Pathol 15:349–355 MacKeigan JM, Ferguson JA (1979) Prophylactic oophorectomy and colorectal cancer in premenopausal patients. Dis Colon Rectum 22:401–405 Martin-Hernández R, Mollejo Villanueva M, Navarro Sánchez M, Carabias López E (2015) Ovarian metastasis of a thymoma: report of a case and literature review. Int J Gynecol Pathol 34:374–378 McCluggage WG (2008) Immunohistochemical markers as a diagnostic aid in ovarian pathology. Diagn Histopathol 14:335–351

18

Metastatic Tumors of the Ovary

McCluggage WG (2012) Immunohistochemistry in the distinction between primary and metastatic ovarian mucinous neoplasms. J Clin Pathol 65:596–600 McCluggage WG, Wilkinson N (2005) Metastatic neoplasms involving the ovary: a review with an emphasis on morphological and immunohistochemical features. Histopathology 47:231–247 McCluggage WG, Young RH (2005) Immunohistochemistry as a diagnostic aid in the evaluation of ovarian tumors. Semin Diagn Pathol 22:3–32 McCluggage WG, Young RH (2008) Primary ovarian mucinous tumors with signet ring cells: report of three cases with discussion of so-called primary Krukenberg tumor. Am J Surg Pathol 32:1373–1379 McCluggage WG, Bissonnette JP, Young RH (2006) Primary malignant melanoma of the ovary: a report of 9 definite or probable cases with emphasis on their morphologic diversity and mimicry of other primary and secondary ovarian neoplasms. Int J Gynecol Pathol 25:321–329 McKenney JK, Soslow RA, Longacre TA (2008) Ovarian mature teratomas with mucinous epithelial neoplasms: morphologic heterogeneity and association with pseudomyxoma peritonei. Am J Surg Pathol 32:645–655 Meriden Z, Yemelyanova AV, Vang R, Ronnett BM (2011) Ovarian metastases of pancreaticobiliary tract adenocarcinomas: analysis of 35 cases, with emphasis on the ability of metastases to simulate primary ovarian mucinous tumors. Am J Surg Pathol 35:276–288 Merino MJ, Edmonds P, LiVolsi V (1985) Appendiceal carcinoma metastatic to the ovaries and mimicking primary ovarian tumors. Int J Gynecol Pathol 4:110–120 Mete O, Yeğen G, Güllüoğlu MG, Kapran Y, Klöppel G (2011) An unusual clinical presentation of pancreatic solid pseudopapillary tumor with ovarian metastases: a diagnostic dilemma. Int J Surg Pathol 19:342–345 Meyer WH, Yu GW, Milvenan ES, Jeffs RD, Kaizer H, Leventhal BG (1979) Ovarian involvement in neuroblastoma. Med Pediatr Oncol 7:49–54 Miettinen M, Wang ZF, Lasota J (2009) DOG1 antibody in the differential diagnosis of gastrointestinal stromal tumors: a study of 1840 cases. Am J Surg Pathol 33:1401–1408 Miettinen M, McCue PA, Sarlomo-Rikala M, Rys J, Czapiewski P, Wazny K, Langfot R, Waloszczyk P, Biernat W, Lasota J, Wang Z (2014) GATA3: a multispecific but potentially useful marker in surgical pathology: a systematic analysis of 2500 epithelial and nonepithelial tumors. Am J Surg Pathol 38:13–22 Milingos D, Doumplis D, Savage P, Seckl M, Lindsay I, Smith JR (2007) Placental site trophoblastic tumor with an ovarian metastasis. Int J Gynecol Cancer 17:925–927 Misdraji J, Yantiss RK, Graeme-Cook FM, Balis UJ, Young RH (2003) Appendiceal mucinous neoplasms. A clinicopathologic analysis of 107 cases. Am J Surg Pathol 27:1089–1103

1219 Mitsushita J, Netsu S, Suzuki K, Nokubi M, Tanaka A (2017) Metastatic ovarian tumors originating from a small bowel adenocarcinoma - a case report and brief literature review. Int J Gynecol Pathol 36:253–260 Moh M, Krings G, Ates D, Aysal A, Kim GE, Rabban JT (2016) SATB2 expression distinguishes ovarian metastases of colorectal and appendiceal origin from primary ovarian tumors of mucinous or endometrioid type. Am J Surg Pathol 40:419–432 Monteagudo C, Merino MJ, Laporte N, Neumann RD (1991) Value of gross cystic disease fluid protein-15 in distinguishing metastatic breast carcinomas among poorly differentiated neoplasms involving the ovary. Hum Pathol 22:368–372 Nonaka D, Chiriboga L, Soslow RA (2008) Expression of Pax8 as a useful marker in distinguishing ovarian carcinomas from mammary carcinomas. Am J Surg Pathol 32:1566–1571 Oberg KC, Wells K, Seraj IM et al (2002) ACTH-secreting islet cell tumor of the pancreas presenting as bilateral ovarian tumors and Cushing’s syndrome. Int J Gynecol Pathol 21:276–280 Ohira S, Shiohara S, Itoh K et al (2003) Urachal adenocarcinoma metastatic to the ovaries: case report and literature review. Int J Gynecol Pathol 22:189–193 Ohta Y, Suzuki T, Shiokawa A, Mitsuya T, Ota H (2005) Expression of CD10 and cytokeratins in ovarian and renal clear cell carcinoma. Int J Gynecol Pathol 24:239–245 Oliva E, Musulen E, Prat J, Young RH (1993) Transitional cell carcinoma of the renal pelvis with symptomatic ovarian metastases. Int J Surg Pathol 2:231–236 Oliva E, Egger JF, Young RH (2014) Primary endometrioid stromal sarcoma of the ovary: a clinicopathologic study of 27 cases with morphologic and behavioral features similar to those of uterine low-grade endometrial stromal sarcoma. Am J Surg Pathol 38:305–315 Ordóñez NG (2006) Value of immunohistochemistry in distinguishing peritoneal mesothelioma from serous carcinoma of the ovary and peritoneum. A review and update. Adv Anat Pathol 13:16–25 Parker LP, Duong JL, Wharton JT, Malpica A, Silva EG, Deavers MT (2002) Desmoplastic small round cell tumor: report of a case presenting as a primary ovarian neoplasm. Eur J Gynaec Oncol 23:199–202 Perez Montiel D, Arispe Angulo K, Cantú-de León D, Bornstein Quevedo L, Chanona Vilchis J, Herrera Montalvo L (2015) The value of SATB2 in the differential diagnosis of intestinal-type mucinous tumors of the ovary: primary vs metastatic. Ann Diagn Pathol 19:249–252 Petru E, Pickel H, Heydarfadai M, Lahousen M, Haas J, Schaider H, Tamussino K (1992) Nongenital cancers metastatic to the ovary. Gynecol Oncol 44:83–86 Phillips V, McCluggage WG, Young RH (2006) Oxyphilic adenomatoid tumor of the ovary: a case report with discussion of the differential diagnosis of ovarian tumors with vacuoles and related spaces. Int J Gynecol Pathol 26:16–20

1220 Pins MR, Young RH, Crum CP, Leach IH, Scully RE (1997) Cervical squamous carcinoma in situ with superficial extension to corpus and tubes and invasion of tubes and ovaries. Int J Gynecol Pathol 16:272–278 Pitluk H, Poticha SM (1983) Carcinoma of the colon and rectum in patients less than 40 years of age. Surg Gynecol Obstet 157:335–337 Pitman MB, Triratanachat S, Young RH, Oliva E (2003) Hepatocyte paraffin 1 antibody does not distinguish primary ovarian tumors with hepatoid differentiation from metastatic hepatocellular carcinoma. Int J Gynecol Pathol 23:58–64 Rabban JT, Lerwill MF, McCluggage WG, Grenert JP, Zaloudek CJ (2009) Primary ovarian carcinoid tumors may express CDX-2: a potential pitfall in distinction from metastatic intestinal carcinoid tumors involving the ovary. Int J Gynecol Pathol 28:41–48 Ramalingam P, Malpica A, Deavers MT (2012) Mixed endocervical adenocarcinoma and high-grade neuroendocrine carcinoma of the cervix with ovarian metastasis of the former component: a report of 2 cases. Int J Gynecol Pathol 31:490–496 Reid MD, Bastruk O, Shaib WL, Xue Y, Balci S, Choi HJ, Akkas G, Memmis B, Robinson BS, El-Rayes BF, Staley CA, Staley CA, Winer JH, Russell MC, Knight JH, Goodman M, Krasinskas AM, Adsay V (2016) Adenocarcinoma ex-goblet cell carcinoid (appendiceal-type crypt cell adenocarcinoma) is a morphologically distinct entity with highly aggressive behavior and frequent association with peritoneal/ intra-abdominal dissemination: an analysis of 77 cases. Mod Pathol 29:1243–1253 Reis-Filho JS, Carrilho C, Valenti C et al (2000) Is TTF1 a good immunohistochemical marker to distinguish primary from metastatic lung adenocarcinomas? Pathol Res Pract 196:835–840 Reyes C, Murali R, Park KJ (2015) Secondary involvement of the adnexa and uterine corpus by carcinomas of the uterine cervix: a detailed morphologic description. Int J Gynecol Pathol 34:551–563 Ricardo-Gonzalez RR, Nguyen M, Gokden N, Sangoi AR, Presti JC, McKenney JK (2012) Plasmacytoid carcinoma of the bladder: a urothelial carcinoma variant with a predilection for intraperitoneal spread. J Urol 187:852–855 Riopel MA, Ronnett BM, Kurman RJ (1999) Evaluation of diagnostic criteria and behavior of ovarian intestinaltype mucinous tumors. Atypical proliferative (borderline) tumors and intraepithelial, microinvasive, invasive, and metastatic carcinomas. Am J Surg Pathol 23:617–635 Robboy SJ, Scully RE, Norris HJ (1974) Carcinoid metastatic to the ovary. A clinicopathologic analysis of 35 cases. Cancer (Phila) 33:798–811 Ronnett BM, Kurman RJ, Zahn CM, Shmookler BM, Jablonski KA, Kass ME (1995a) Pseudomyxoma peritonei in women. A clinicopathologic analysis of 30 cases with emphasis on site of origin, prognosis, and relationship to ovarian mucinous tumors of low malignant potential. Hum Pathol 26:509–524

M. F. Lerwill and R. H. Young Ronnett BM, Zahn CM, Kurman RJ, Kass ME, Sugarbaker PH, Shmookler BM (1995b) Disseminated peritoneal adenomucinosis and peritoneal mucinous carcinomatosis: a clinicopathologic analysis of 109 cases with emphasis on distinguishing pathologic features, site of origin, prognosis, and relationship to “pseudomyxoma peritonei”. Am J Surg Pathol 19:1390–1408 Ronnett BM, Kurman RJ, Shmookler BM, Sugarbaker PH, Young RH (1997) The morphologic spectrum of ovarian metastases of appendiceal adenocarcinomas. A clinicopathologic and immunohistochemical analysis of tumors often misinterpreted as primary ovarian tumors or metastatic tumors from other gastrointestinal sites. Am J Surg Pathol 2:1144–1155 Ronnett BM, Yemelyanova AV, Vang R, Gilks CB, Miller D, Gravitt PE, Kurman RJ (2008) Endocervical adenocarcinomas with ovarian metastases. Analysis of 29 cases with emphasis on minimally invasive cervical tumors and the ability of the metastases to simulate primary ovarian neoplasms. Am J Surg Pathol 32:1835–1853 Saphir O (1951) Signet-ring cell carcinoma. Mil Surg 109:360–369 Schultheis AM, Ng CK, De Filippo MR, Piscuoglio S, Macedo GS, Gatius S, Perez Mies B, Soslow RA, Lim RS, Viale A, Huberman KH, Palacios JC, ReisFilho JS, Matias-Guiu, Weigelt B (2016) Massively parallel sequencing-based clonality analysis of synchronous endometrioid endometrial and ovarian carcinomas. J Natl Cancer Inst 108:djv427. https://doi.org/ 10.1093/jnci/djv427 Scully RE, Richardson GS (1961) Luteinization of the stroma of metastatic cancer involving the ovary and its endocrine significance. Cancer (Phila) 14:827–840 Seidman JD, Kurman RJ, Ronnett BM (2003) Incidence in routine practice with a new approach to improve intraoperative diagnosis. Primary and metastatic mucinous adenocarcinomas in the ovaries. Am J Surg Pathol 27:985–993 Shimada M, Kigawa J, Nishimura R, Yamaguchi S, Kuzuya K, Nakanishi T, Suzuki M, Kita T, Iwasaka T, Terakawa N (2006) Ovarian metastasis in carcinoma of the uterine cervix. Gynecol Oncol 101:234–237 Soslow RA, Rouse RV, Hendrickson MR, Silva EG, Longacre TA (1996) Transitional cell neoplasms of the ovary and urinary bladder: a comparative immunohistochemical analysis. Int J Gynecol Pathol 15:257–265 Spencer JR, Eriksen B, Garnett JE (1993) Metastatic renal tumor presenting as ovarian clear cell carcinoma. Urology 41:582–584 Stewart CJR, Brennan BA, Hammond IG et al (2005) Accuracy of frozen section in distinguishing primary ovarian neoplasia from tumors metastatic to the ovary. Int J Gynecol Pathol 24:356–362 Stewart CJ, Ardakani NM, Doherty DA, Young RH (2014) An evaluation of the morphologic features of low-grade mucinous neoplasms of the appendix metastatic in the ovary, and comparison with primary ovarian mucinous tumors. Int J Gynecol Pathol 33:1–10

18

Metastatic Tumors of the Ovary

Strosberg J, Nasir A, Cragun J, Gardner N, Kvols L (2007) Metastatic carcinoid tumor to the ovary: a clinicopathologic analysis of seventeen cases. Gynecol Oncol 106:65–68 Sty JR, Kun LE, Casper JT (1980) Bone scintigraphy in neuroblastoma with ovarian metastasis. Wis Med J 79:28–29 Szych C, Staebler A, Connolly DC, Wu R, Cho KR, Ronnett BM (1999) Molecular genetic evidence supporting the clonality and appendiceal origin of pseudomyxoma peritonei in women. Am J Pathol 154:1849–1855 Tabata M, Ichinoe K, Sakuragi N, Shina Y, Yamaguchi T, Mabuchi Y (1987) Incidence of ovarian metastasis in patients with cancer of the uterine cervix. Gynecol Oncol 28:255–261 Tandon RT, Jimenez-Cortez Y, Taub R, Borczuk AC (2018) Immunohistochemistry in peritoneal mesothelioma. A single-center experience of 244 cases. Arch Pathol Lab Med 142:236–242 Tang LH, Shia J, Soslow RA, Dhall D, Wong WD, O’Reilly E, Qin J, Paty P, Weiser MR, Guillem J, Temple L, Sobin LH, Klimstra DS (2008) Pathologic classification and clinical behavior of the spectrum of goblet cell carcinoid tumors of the appendix. Am J Surg Pathol 32:1429–1443 Toki N, Tsukamoto N, Kaku T et al (1991) Microscopic ovarian metastasis of the uterine cervical cancer. Gynecol Oncol 41:46–51 Tornos C, Soslow R, Chen S et al (2005) Expression of WT1, CA 125, and GCDFP-15 as useful markers in the differential diagnosis of primary ovarian carcinomas versus metastatic breast cancer to the ovary. Am J Surg Pathol 29:1482–1489 Tso PL, Bringaze WL III, Dauterive AH, Correa P, Cohn I Jr (1987) Gastric carcinoma in the young. Cancer (Phila) 59:1362–1365 Ulbright TM, Roth LM (1985a) Metastatic and independent cancers of the endometrium and ovary: a clinicopathologic study of 34 cases. Hum Pathol 16:28–34 Ulbright TM, Roth LM (1985b) Secondary tumors of the ovary. In: Roth LM, Czernobilsky B (eds) Tumors and tumor-like conditions of the ovary, Contemporary issues in surgical pathology, vol 6. ChurchillLivingstone, New York, pp 129–152 Ulbright TM, Roth LM, Stehman FB (1984) Secondary ovarian neoplasia. A clinicopathologic study of 35 cases. Cancer (Phila) 53:1164–1174 Vakiani E, Young RH, Carcangiu ML, Klimstra DS (2008) Acinar cell carcinoma of the pancreas metastatic to the ovary. A report of 4 cases. Am J Surg Pathol 32:1540–1545 Vang R, Gown AM, Barry TS, Wheeler DT, Yemelyanova A, Seidman JD, Ronnett BM (2006a) Cytokeratins 7 and 20 in primary and secondary mucinous tumors of the ovary: analysis of coordinate immunohistochemical expression profiles and staining distribution in 179 cases. Am J Surg Pathol 30:1130–1139

1221 Vang R, Gown AM, Wu LSF, Barry TS, Wheeler DT, Yemelyanova A, Seidman JD, Ronnett BM (2006b) Immunohistochemical expression of CDX2 in primary ovarian mucinous tumors and metastatic mucinous carcinomas involving the ovary: comparison with CK20 and correlation with coordinate expression of CK7. Mod Pathol 19:1421–1428 Vang R, Gown AM, Zhao C, Barry TS, Isacson C, Richardson MS, Ronnett BM (2007) Ovarian mucinous tumors associated with mature cystic teratomas. Morphologic and immunohistochemical analysis identifies a subset of potential teratomatous origin that shares features of lower gastrointestinal tract mucinous tumors more commonly encountered as secondary tumors in the ovary. Am J Surg Pathol 31:854–869 Vara A, Madrigal B, Veiga M, Diaz A, Garcia J, Calvo J (1998) Bilateral ovarian metastases from renal clear cell carcinoma. Acta Oncol (Stockh) 37:379–380 Veras E, Srodon M, Neijstrom ES, Ronnett BM (2009) Metastatic HPV-related cervical adenocarcinomas presenting with thromboembolic events (trousseau syndrome): clinicopathologic characteristics of two cases. Int J Gynecol Pathol 28:134–139 Wick MR, Lillemoe TJ, Copland GT, Swanson PE, Manivel JC, Kiang DT (1989) Gross cystic disease fluid protein15 as a marker for breast cancer: immunohistochemical analysis of 690 human neoplasms and comparison with alpha-lactalbumin. Hum Pathol 20:281–287 Yakushiji M, Tazaki T, Nishimura H, Kato T (1987) Krukenberg tumors of the ovary: a clinicopathologic analysis of 112 cases. Acta Obstet Gynaecol Jpn 39:479–485 Yang C, Sun L, Zhang L, Zhou L, Zhao M, Peng Y, Niu D, Li Z, Huang X, Kang Q, Jia L, Lai J, Cao D (2018) Diagnostic utility of SATB2 in metastatic Krukenberg tumors of the ovary. An immunohistochemical study of 70 cases with comparison to CDX2, CK7, CK20, chromogranin, and synaptophysin. Am J Surg Pathol 42:160–171 Yemelyanova AV, Vang R, Judson K, Wu LSF, Ronnett BM (2008) Distinction of primary and metastatic mucinous tumors involving the ovary: analysis of size and laterality data by primary site with reevaluation of an algorithm for tumor classification. Am J Surg Pathol 32:128–138 Young RH (1995) Urachal adenocarcinoma metastatic to the ovary simulating primary mucinous cystadenocarcinoma of the ovary: report of a case. Virchows Arch 426:529–532 Young RH (2004) Pseudomyxoma peritonei and selected other aspects of the spread of appendiceal neoplasms. Semin Diagn Pathol 21:134–150 Young RH (2006) From Krukenberg to today: the ever present problems posed by metastatic tumors in the ovary. Part I. Historical perspective, general principles, mucinous tumors including the Krukenberg tumor. Adv Anat Pathol 13:205–227 Young RH (2007) From Krukenberg to today: the ever present problems posed by metastatic tumors in the ovary. Part II. Adv Anat Pathol 14:149–177

1222 Young RH, Hart WR (1989) Metastases from carcinomas of the pancreas simulating primary mucinous tumors of the ovary: a report of seven cases. Am J Surg Pathol 13:748–756 Young RH, Hart WR (1992) Renal cell carcinoma metastatic to the ovary: a report of three cases emphasizing possible confusion with ovarian clear cell adenocarcinoma. Int J Gynecol Pathol 11:96–104 Young RH, Hart WR (1998) Metastatic intestinal carcinomas simulating primary ovarian clear cell carcinoma and secretory endometrioid carcinoma. A clinicopathologic and immunohistochemical study of five cases. Am J Surg Pathol 22:805–815 Young RH, Scully RE (1985) Ovarian metastases from cancer of the lung: problems in interpretation – a report of seven cases. Gynecol Oncol 21:337–350 Young RH, Scully RE (1988a) Urothelial and ovarian carcinomas of identical cell types: problems in interpretation. A report of three cases and review of the literature. Int J Gynecol Pathol 7:197–211 Young RH, Scully RE (1988b) Mucinous tumors of the ovary associated with mucinous adenocarcinomas of the cervix. A clinicopathological analysis of 16 cases. Int J Gynecol Pathol 7:99–111 Young RH, Scully RE (1989) Alveolar rhabdomyosarcoma metastatic to the ovary. A report of two cases and discussion of the differential diagnosis of small cell malignant tumors of the ovary Cancer (Phila) 64:899–904 Young RH, Scully RE (1990a) Ovarian metastases from carcinoma of the gallbladder and extrahepatic bile ducts simulating primary tumors of the ovary: a report of six cases. Int J Gynecol Pathol 9:60–72 Young RH, Scully RE (1990b) Sarcomas metastatic to the ovary. A report of 21 cases. Int J Gynecol Pathol 9:231–252 Young RH, Scully RE (1991a) Metastatic tumors in the ovary: a problem-oriented approach and review of the recent literature. Semin Diagn Pathol 8:250–276 Young RH, Scully RE (1991b) Malignant melanoma metastatic to the ovary: a clinicopathologic analysis of 20 cases. Am J Surg Pathol 15:849–860 Young RH, Scully RE (2001) Differential diagnosis of ovarian tumors based primarily on their patterns and cell types. Semin Diagn Pathol 18:161–235 Young RH, Carey RW, Robboy SJ (1981) Breast carcinoma masquerading as a primary ovarian neoplasm. Cancer (Phila) 48:210–212 Young RH, Prat J, Scully RE (1984) Endometrial stromal sarcomas of the ovary. A clinicopathologic analysis of 23 cases. Cancer (Phila) 53:1143–1155

M. F. Lerwill and R. H. Young Young RH, Gilks CB, Scully RE (1991) Mucinous tumors of the appendix associated with mucinous tumors of the ovary and pseudomyxoma peritonei: a clinicopathological analysis of 22 cases supporting an origin in the appendix. Am J Surg Pathol 15: 415–429 Young RH, Gersell DJ, Clement PB, Scully RE (1992a) Hepatocellular carcinoma metastatic to the ovary: a report of three cases discovered during life with discussion of the differential diagnosis of hepatoid tumors of the ovary. Hum Pathol 23:574–580 Young RH, Eichhorn JH, Dickersin GR, Scully RE (1992b) Ovarian involvement by the intra-abdominal desmoplastic small round cell tumor with divergent differentiation. A report of three cases. Hum Pathol 23:454–464 Young RH, Kozakewich HPW, Scully RE (1993a) Metastatic ovarian tumors in children: a report of 14 cases and review of the literature. Int J Gynecol Pathol 12:8–19 Young RH, Gersell DJ, Roth LM, Scully RE (1993b) Ovarian metastases from cervical carcinomas other than pure adenocarcinomas: a report of 12 cases. Cancer (Phila) 71:407–418 Young RH, Jackson A, Wells M (1994) Ovarian metastasis from thyroid carcinoma twelve years after partial thyroidectomy mimicking struma ovarii. Report of a case. Int J Gynecol Pathol 13:181–185 Yu TJ, Iwasaki I, Horie H, Tamaru J, Takahashi A (1986) Endolymphatic stromal myosis of the uterus with metastasis to ovary and recurrence in vagina. Acta Pathol Jpn 36:301–308 Zaino RJ, Unger ER, Whitney C (1984) Synchronous carcinomas of the uterine corpus and ovary. Gynecol Oncol 19:329–335 Zaloudek C, Miller TR, Stern JL (1995) Desmoplastic small cell tumor of the ovary: a unique polyphenotypic tumor with an unfavorable prognosis. Int J Gynecol Oncol 14:260–265 Zamecnik M, Voltr L, Stuk J, Chlumska A (2008) Krukenberg tumor with yolk sac tumor differentiation. Int J Gynecol Pathol 27:223–228 Zhang PJ, Gao HG, Pasha TL, Litzky L, LiVolsi V (2009) TTF-1 expression in ovarian and uterine epithelial neoplasia and its potential significance, an immunohistochemical assessment with multiple monoclonal antibodies and different secondary detection systems. Int J Gynecol Pathol 28:10–18 Zukerberg LR, Young RH (1990) Chordoma metastatic to the ovary: report of a case. Arch Path Lab Med 114:208–210

Diseases of the Placenta

19

Rebecca N. Baergen, Deborah J. Gersell, and Frederick T. Kraus

Contents Normal Anatomy and Development . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1225 Abnormal Placentation and Villous Development . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Aberrant Shapes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Extrachorial Placenta . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Placenta Accreta, Increta, and Percreta . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Placental Mesenchymal Dysplasia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1228 1228 1229 1230 1232

Multiple Pregnancy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Twin Gestation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Complications of Multiple Pregnancy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Higher Multiple Births . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1233 1233 1239 1247

Placental Inflammatory Lesions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Ascending Infection and ACA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Subacute Chorioamnionitis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Chronic Villitis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Other Patterns of Placental Inflammation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1248 1248 1254 1254 1260

Maternal and Fetal Vascular Malperfusion (FVM) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1261 Maternal Circulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1261 Fetal Circulation and FVM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1270

R. N. Baergen (*) Department of Pathology and Laboratory Medicine, Weill Cornell Medicine, New York-Presbyterian Hospital, New York, NY, USA e-mail: [email protected] D. J. Gersell Department of Laboratory Medicine, St. John’s Mercy Medical Center, St. Louis, MO, USA e-mail: [email protected] F. T. Kraus Perinatal Biology Laboratory, Department of Obstetrics and Gynecology, Washington University School of Medicine, St. Louis, MO, USA e-mail: [email protected] # Springer Nature Switzerland AG 2019 R. J. Kurman et al. (eds.), Blaustein’s Pathology of the Female Genital Tract, https://doi.org/10.1007/978-3-319-46334-6_19

1223

1224

R. N. Baergen et al. Fetal Membranes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Squamous Metaplasia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Amnion Nodosum . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Amniotic Bands . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Meconium . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Gastroschisis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1273 1273 1274 1275 1276 1277

Umbilical Cord . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Normal Anatomy and Embryonic Development . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Embryonic Remnants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . “Cord Accidents” . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Cord Length . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Knots . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Hypercoiling and Hypocoiling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Stricture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Cord Diameter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Rupture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Insertion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Umbilical Vessels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1277 1277 1278 1278 1279 1280 1281 1282 1282 1283 1283 1285

Clinical Syndromes and Their Pathologic Correlates in the Placenta . . . . . . . . . . . . Preeclampsia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Essential Hypertension . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Diabetes Mellitus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . PTB, PTL, and Preterm Rupture of Membranes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Post-term Pregnancy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Fetal Growth Restriction, Intrauterine Growth Restriction (IUGR) . . . . . . . . . . . . . . . . . . NE, CP, and “Birth Asphyxia” . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Fetal and Placental Hydrops . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Nucleated Red Blood Cells in the Fetal Circulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Thrombophilias . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Acute Fatty Liver of Pregnancy (AFLP) and HELLP Syndrome . . . . . . . . . . . . . . . . . . . . . Sickle-Cell Trait/Disease and Other Hemoglobinopathies . . . . . . . . . . . . . . . . . . . . . . . . . . . . Storage Disorders . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1288 1288 1288 1288 1289 1289 1289 1290 1290 1291 1292 1293 1293 1293

Abortion, Stillbirth, and Intrauterine Fetal Death . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1293 Early Abortion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1294 Late Abortion, Stillbirth, and Intrauterine Fetal Death . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1295 Nontrophoblastic Tumors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Chorangioma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Hepatocellular Adenoma and Adrenocortical Nodules . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Other Placental “Tumors” . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Placental Metastases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1295 1295 1296 1297 1297

Examination of the Placenta . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Fetal Membranes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Umbilical Cord . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Placenta . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Sections . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Special Aspects of Multiple Pregnancy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Special Techniques . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1298 1298 1298 1298 1299 1299 1299

References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1299

The placenta is crucial for fetal growth and survival, performing the most important functions of many somatic organs before birth. Thus, pathologic processes interfering with placental

function may result in abnormalities of fetal growth or development, malformation, or stillbirth, and there is increasing recognition that some long-term (especially neurologic)

19

Diseases of the Placenta

1225

disabilities can be traced to injury occurring before birth. The purpose of this chapter is to describe clinically important placental lesions and to emphasize the context in which these lesions are directly or indirectly important to the fetus, the mother, or both.

Normal Anatomy and Development The monograph by Boyd and Hamilton (Boyd and Hamilton 1970) provides a detailed description and exquisite illustrations of the various stages of human implantation. The ovum is fertilized in the fallopian tube and develops rapidly, reaching the endometrial cavity as a blastocyst. At this stage, the outer cell layer of the blastocyst has differentiated into trophoblast and cells in the inner cell mass from which the embryo will develop. The trophoblast attaches to and penetrates the endometrium on the 6th to 7th postovulatory day, and by the 10th to 11th postovulatory day, the blastocyst is totally embedded in endometrial stroma which will have reestablished continuity over the penetration defect. The trophoblast grows rapidly and circumferentially, invading maternal blood vessels forming blood-filled spaces (lacunae) which separate the trophoblast into trabecular columns (Fig. 1), with an outer syncytiotrophoblastic layer oriented radially around central solid cores of cytotrophoblast (CT). As the extraembryonic mesenchyme penetrates the cytotrophoblastic cores, small blood vessels form within it, and these eventually connect with each other and with those forming independently in the allantois of the body stalk (chorioallantoic placentation), establishing the fetoplacental circulation by the fifth to sixth week. A shell of solid trophoblast remains deep to the stem villi, anchoring them to the basal plate and continuing to grow and expand the placenta and the intervillous space (Fig. 2). Successful implantation requires a series of complex, coordinated interactions between maternal tissue and the trophoblast. The trophoblast consists of several morphologically and functionally distinct cell types, each with characteristic

Fig. 1 Implantation at 13 days. Trophoblast has differentiated into inner (CT) and outer (syncytiotrophoblast) layers. Focally, the CT has proliferated to form projections, the forerunners of the primary villi. The embryonic disk is located near the center. (Reprinted courtesy of Department of Embryology, Davis Division, Carnegie Institute of Washington)

Fig. 2 Secondary villi. Mesenchyme has penetrated the trophoblastic cores. The trophoblastic shell is deep toward the endometrium (top of the figure)

1226

anatomic distribution. CT, the germinative, mitotically active component of trophoblast, is present as a layer of uniform cells with single, round–oval nuclei, clear cytoplasm, and distinct cell borders directly overlying the stromal core of the villus. The syncytiotrophoblast (ST) overlies the CT and is the terminally differentiated component of trophoblast responsible for transport functions, protection, and pregnancy-specific protein and hormone production. Its abundant, often vacuolated cytoplasm is amphophilic with multiple small dark nuclei and a distinct brush border. Intermediate or extravillous trophoblast (IT) is present in the anchoring cell columns but is most prevalent in extravillous sites including the implantation site, chorionic plate, extraplacental membranes, cells islands, and placental septae. IT developing from the trophoblastic shell invades the endometrium and myometrium at the implantation site. Subpopulations of IT in the villi (villous IT), implantation site, and membranes (chorionic IT) are morphologically and immunohistochemically distinct (Shih and Kurman 2004). IT nuclei are irregular and hyperchromatic with coarsely granular chromatin. Most cells are mononucleate, although multinucleate forms occur. IT may be round, polyhedral, or spindle shaped depending on the location with abundant eosinophilic, amphophilic, or clear cytoplasm. The cytologic features are generally sufficiently distinctive to identify IT, but their intermingling with the decidua at the implantation

R. N. Baergen et al.

site is so intimate that it may be difficult to characterize any particular cell as maternal or fetal by conventional light microscopy. When findings are equivocal (as in some abortion specimens or in uterine specimens where placenta accreta is suspected), immunohistochemical stains for keratin help distinguish IT (keratin positive) from the decidua (keratin negative). The IT that infiltrates the decidua and myometrium at the implantation site is responsible for remarkable physiologic structural modifications in the spiral arteries called physiologic conversion. In the early weeks of pregnancy, IT invades the decidual segments of the spiral arteries, forming intraluminal plugs. Later, between the 12th and 20th weeks of pregnancy, endovascular IT extends from the decidual into the myometrial segments of the spiral arteries. Eventually, IT and fibrinoid completely replace the endothelium and the muscular and elastic tissue of the media (Fig. 3), and the IT takes on an endothelial phenotype. Altered by this process, the spiral arteries undergo progressive distension, eventuating in large funnel-shaped channels that augment blood flow to the implantation site. Dissolution of the muscular media results in fixed vascular dilatation unresponsive to vasculospastic influences. Initially, villi surround the entire chorionic cavity, but as the chorion prolapses into the endometrial cavity, the villi oriented toward the uterine cavity undergo progressive atrophy to form the

Fig. 3 Normal spiral arteriole remodeling. Intraluminal IT (left) later invades and replaces the vascular media along with fibrinoid matrix (right)

19

Diseases of the Placenta

Fig. 4 Formation of chorion laeve. As the chorion prolapses into the endometrial cavity, the villi on the intracavitary aspect atrophy to form the fetal membranes

Fig. 5 Chorion laeve. Atrophied villous remnants remain in the membranes at term

smooth chorion (chorion laeve) or fetal membranes (Fig. 4). These atrophic villi may still be apparent in sections of membranes in the mature placenta (Fig. 5). The villi on the endometrial aspect of the chorion continue to proliferate, forming the definitive placenta (chorion frondosum). Departure from the usual pattern of villous growth and atrophy is thought to result in some of the aberrant placental shapes described below as well as aberrant cord insertions. Continued growth and enlargement of the chorion results in eventual obliteration of the uterine cavity through fusion of the decidua capsularis and the decidua vera of the opposite uterine wall, usually around 20 weeks. In time, the chorionic cavity is obliterated by progressive expansion of the amnion. Irregular folds of the basal plate are

1227

drawn into the intervillous space by the relatively slow growth of anchoring villi to form the placental septae, appearing at about 3 months. IT is prominent in the septae. The septae partition the maternal surface incompletely and irregularly into 15 to 20 divisions that have no known physiologic significance. During growth and maturation of the placenta, five villous types have been detailed in the work of Kaufmann and others (Benirschke et al. 2006, 2012). Mesenchymal villi are a primitive, transient stage in villous development. The loose stroma of mesenchymal villi is abundant with numerous Hofbauer cells, central small vessels, and an orderly surface bilayer of CT and ST. Mesenchymal villi predominate in early pregnancy but may be found in small numbers even at term as they are the precursors of the other villous types. According to Kaufmann, mesenchymal villi begin to develop into immature intermediate villi around 7–8 weeks. Immature intermediate villi are defined by abundant loose reticular stroma with prominent stromal channels which occasionally contain Hofbauer cells (placental macrophages), features that may lead to erroneous interpretation as villous edema (Fig. 6). Immature intermediate villi predominate through the second trimester, but small clusters persist in the center of lobules at term (normally 0–5% volumetrically). Immature intermediate villi gradually develop into stem villi as their vessels acquire a distinct muscular media with progressively prominent adventitia and fibrous stroma. Stem villi support the villous tree and transport blood but do not participate significantly in oxygen or nutrient exchange. Stem villi comprise 20–25% of villi in a normal-term placenta, with highest concentration centrally beneath the chorionic plate. Beginning in the third trimester, newly formed mesenchymal villi develop into mature intermediate villi. Mature intermediate villi are long and slender with roughly the same diameter as terminal villi and numerous small vessels and capillaries, comprising less than 50% of the villus. Roughly one fourth of the villous volume at term is made up of mature intermediate villi. Terminal villi are the final ramifications of the villous tree produced along the surfaces of mature

1228

Fig. 6 Immature intermediate villi. These are characterized by reticular stroma and stromal channels containing Hofbauer cells

Fig. 7 Third-trimester terminal villi, vasculosyncytial membranes. Fetal capillaries protrude beneath thinned ST cytoplasm between knots. The endothelial and syncytiotrophoblastic basement membranes fuse forming vasculosyncytial membranes

intermediate villi during the third trimester. Terminal villi have sinusoidally dilated capillaries (by definition occupying more than 50% of the villous stroma) that bulge beneath overlying attenuated trophoblastic cover. Fusion of endothelial and trophoblastic basement membranes results in the formation of vasculosyncytial membranes (Fig. 7). In these areas, where the maternal and fetal circulations are most closely approximated, gas and nutrient exchange occurs. Normally, terminal villi make up more than 40% of the villous volume at term.

R. N. Baergen et al.

With villous maturation, the barrier between maternal and fetal circulations is reduced by the thinning of ST and cytotrophoblast, decrease in mean villous diameter, and apposition of fetal capillaries to the villous surface. The villi in any placenta are not completely homogeneous. The peripheral villi and those beneath the chorionic plate tend to be smaller with more collagenous stroma and a thicker trophoblastic basement membrane and are less well perfused than the villi located centrally in the fetal lobule. These regional differences should always be considered when judging placental maturity, which is best assessed in standardized sections from central placental zones. The placenta has two circulations maternal and fetal. Maternal blood is delivered to the intervillous space through maternal, uteroplacental vessels in the basal plate. Maternal blood flows toward the chorionic plate, disperses laterally, percolates around the villi, and exits through venous outlets concentrated peripherally in the placental floor. Deoxygenated fetal blood reaches the placenta through the two umbilical arteries that branch and divide in stem villi until they ultimately terminate in the complex capillary network of the terminal villi. Oxygenated and fortified fetal blood returns via venous tributaries to the umbilical vein. The placental circulation normally receives about 55% of the fetal cardiac output. Lacking autonomic innervation, it responds only to local factors such as pressure and flow.

Abnormal Placentation and Villous Development Aberrant Shapes The pattern of villous atrophy and proliferation that occurs during placental development resulting in ultimate placental shape and configuration is thought to be determined by maternal blood flow. This is the concept of trophotropism (Benirschke et al. 2012) which simply states that the placenta grows and expands in areas of good nutrition and atrophy in areas of poor nutrition. This can result in portions of the placenta being

19

Diseases of the Placenta

“left behind” to for an accessory or succenturiate lobe or velamentous cord insertion (vide infra). The succenturiate lobe is the most common shape variation (3–6% of placentas) in which usually one but occasionally multiple discrete masses of placental tissue are separated from the main placenta by fetal membranes (Fig. 8). The umbilical cord usually inserts into the main placental mass, and the fetal vessels supplying the accessory lobe traverse membranes unsupported by underlying villous parenchyma. If these membranous vessels are traumatized during delivery, severe fetal hemorrhage may result. Thrombosis of membranous vessels also may occur. By their very nature, succenturiate lobes have a tendency to atrophy or infarct but otherwise show no specific histologic changes. The bilobate placenta is a variant in which there are two relatively equal-sized placental lobes separated by fetal membranes. The umbilical cord often inserts centrally between the lobes. Other anomalous shapes are uncommon. Placenta membranacea is a large, thin placenta with functional chorionic villi covering the entire gestational sac. In this condition, differential atrophy of the chorion laeve and proliferation of the chorion frondosum do not occur. The placental parenchyma may vary in thickness, but only exceptionally is there a dominant area resembling a placental disk. Placenta membranacea occurs normally in some animal species but is extremely rare in humans. When they do occur, they are

1229

complicated by antepartum bleeding and abnormal placental adherence (placenta accreta) relating to the obligate placenta previa that accompanies this form of placentation. Nearly all cases are associated with preterm delivery and high fetal mortality. The annular or cylindrical ring-shaped placenta is very rare. In fenestrate placentas, focal absence of the villous parenchyma may result in a through-and-through hole or the chorionic plate and occurs most commonly due to implantation over a leiomyoma or other uterine defects.

Extrachorial Placenta Extrachorial placentation is a common gross structural deviation in which the extraplacental membranes do not insert in the margin of the chorionic but rather inward toward the central portion of the placenta. This leaves a bare rim of placental tissue (extrachorial portion) that is not covered by fetal membranes (Fig. 9). Fetal vessels appear to terminate at the margin of the chorionic plate but actually continue their course peripherally in the deeper villous tissue. Etiology and pathogenesis. There have been many theories attempting to explain the etiology and pathogenesis of extrachorial placentation including abnormally deep implantation, abnormally superficial implantation, and premature fixation of the disk membrane boundary. Recently,

Fetal vessels Extrachorial placenta

Fig. 8 Accessory (succenturiate) lobe. A discrete mass of placental tissue is separated from the main disk by fetal membranes containing velamentous, unsupported fetal vessels

Fig. 9 Diagram of an extrachorial placenta, fetal side. Vessels appear to terminate at the margin of the chorionic plate but continue peripherally in the extrachorial portion. (After Fox 1997)

1230

R. N. Baergen et al.

Fig. 11 Extrachorial placenta. The membrane transition in this extrachorial placenta is flat (circummarginate insertion) Fig. 10 Extrachorial placenta. The fetal membranes do not extend to the peripheral margin of the disk, leaving a ring of placental tissue extending beyond the chorionic plate. This membrane ring is rolled and reflected centrally (circumvallate insertion)

recurrent marginal hemorrhage resulting in circumvallation has been documented on serial ultrasound images, providing evidence that recurrent hemorrhage at the disk margin elevates and displaces the membrane insertion site centrally (Redline et al. 1999). This may represent the pathologic correlate of recurrent and persistent maternal bleeding throughout pregnancy (chronic abruption). Amniotic fluid loss also likely contributes to this process. If the extrachorial implantation is extreme, rupture of the membranes may occur early and lead to an extramembranous pregnancy. Pathology. Historically, extrachorial placentas have been subdivided into circummarginate and circumvallate types based on the nature of the transition from the chorionic plate to the fetal membranes (Benirschke et al. 2006, 2012). In circumvallate placentas, there is a distinct fibrin ring often reflected centrally, folded, and rolled back upon itself (Fig. 10). Variable amounts of fibrin and recent and old blood clot are often found in the reflected membrane fold, with focal or, in severe cases, diffuse hemosiderin deposits in the chorionic plate. A flat transition from the chorionic plate to the membranes without reflection or prominent fibrin accumulation has been

termed a circummarginate placenta (Fig. 11). Either condition may be partial or complete, and they frequently merge imperceptibly with one another. Multiple authors have suggested that the term circummarginate should be abandoned and that all extrachorial placentas should be considered part of the spectrum of circumvallation (Kraus et al. 2004). Clinical behavior. Estimates of the frequency of extrachorial placentation vary widely presumably because the terms are not used uniformly. It is more common in multigravidas. The clinical sequelae seem to parallel the amount of associated hemorrhage and the extent of circumvallation or circummargination. Mild or focal circumvallation without hemorrhage is clinically insignificant. More severe cases with chronic marginal abruption and oligohydramnios are associated with antepartum bleeding, preterm labor (PTL), intrauterine growth restriction (IUGR), and long-term neurologic impairment (Redline and O’Riordan 2000). Circumvallate placentation may recur in successive pregnancies.

Placenta Accreta, Increta, and Percreta Placenta accreta, increta, and percreta are defined as abnormal adherence of the placenta to the uterine wall so that placental separation does not occur after delivery of the newborn, and the

19

Diseases of the Placenta

1231

underlying etiology is lack of decidua. The degree of abnormal adherence/invasion is variable; placental villi may adhere to (placenta accreta) or invade the myometrium (placenta increta), sometimes penetrating through the serosa (placenta percreta.) In practice though, almost all placenta accretas show areas of invasion and thus are truly placenta incretas. Small focal placenta accretas do occur. Etiology. The pathologic basis of this condition is the absence of the decidua. Decidua normally regulates trophoblastic invasion during gestation and allows the placenta to separate from the myometrium as there is a shearing force through the decidual layer when the uterus contracts after delivery of the fetus. Placenta accreta/increta also characterizes ectopic (cornual, tubal, or cervical) pregnancies as these aberrant locations do not develop proper decidua (Benirschke et al. 2006, 2012). Clinical features. The frequency of placenta accreta is difficult to determine. Reported figures have varied widely from 1 in 540 to 1 in 70,000 pregnancies. Fox emphasized the particular tendency of placenta accreta to occur in multigravida and obstetrically elderly women (Fox and Sebire 2007). A number of predisposing factors have

been linked to this condition, the two most significant being placenta previa (placental implantation in the lower uterine segment near or overlying the cervical os) and previous cesarean section. In some reports, as many as 64% of placenta accretas have been associated with placenta previa, due to the poor decidualization that often occurs in the cervix. Other risk factors include prior uterine instrumentation or intrauterine infection, previous manual removal of the placenta, uterine structural defects (leiomyomas, septae), and nonfundic implantation. The common endpoint in all these conditions is a deficiency in, or absence of, the decidua. In less than 10% of cases, no risk factors are identified. Pathology. The diagnosis is usually obvious in a hysterectomy performed at the time of delivery. The placenta is most often implanted in the lower uterine segment or cervix, often anteriorly in the region of prior cesarean section (Fig. 12). The myometrium is variably but often markedly thinned. The uterus and placenta may be disrupted by attempts to remove the adherent placenta at the time of delivery. Microscopically, the cardinal feature is partial or complete absence of the decidua basalis. Placental villi adhere directly

Fig. 12 Placenta percreta. The placenta at the left implanted over the cervical os (upper left) and penetrated the serosa. The placenta at the right also penetrated the

serosa in the anterior lower uterine segment at the site of three prior cesarean sections

1232

R. N. Baergen et al.

Fig. 13 Placenta increta. The placenta has invaded the myometrium almost to the serosa. Villi invade the myometrium without intervening decidua

to, or invade into, the myometrium (Fig. 13), penetrating through the uterine serosa in some instances. The villi typically do not adhere directly to the myometrium but rather are enmeshed in fibrin and extravillous trophoblast. The key feature for diagnosis is that between the villi and the myometrium, no decidua exists. The diagnosis can also be established on placental exam. Microscopic sections of grossly fragmented, low-lying, or manually removed placentas may show a thin layer of myometrial fibers adherent to the placenta without intervening decidua. Occasional smooth muscle cells in a basal plate containing decidua are not diagnostic of accreta. The diagnosis in a postpartum curettage calls attention to the potential for further bleeding or infection if all the placental tissue is not removed. Clinical behavior and treatment. Placenta accreta is compatible with normal fetal growth and development. It is often diagnosed by imaging and so usually anticipated; however, in some cases, it may not be suspected until the placenta fails to separate in the third stage of labor. Postpartum bleeding may be lifethreatening, requiring immediate hysterectomy, or bleeding may be delayed. Antepartum bleeding and premature labor are common due to the high frequency of associated placenta previa. Uterine rupture may occur at any stage of pregnancy or during labor. Maternal and fetal mortality is now uncommon.

Fig. 14 Mesenchymal dysplasia. The gross abnormalities in this placenta (enlargement and cystic villi) with mesenchymal dysplasia were identified on antenatal ultrasound. The infant was normal

Placental Mesenchymal Dysplasia Placentas with mesenchymal dysplasia are usually very large, often over 1,000 g, and have a distinctive gross appearance. The chorionic plate vessels are aneurysmally dilated and often thrombosed. Enlarged cystic stem villi are recognizable grossly (Fig. 14). Microscopically, the stem villi are large and edematous with cyst formation and prominent thick-walled muscular stem vessels (Fig. 15). Terminal villi are usually normal but may show distal villous immaturity and chorangiosis. They may be confused grossly and microscopically with a partial hydatidiform

19

Diseases of the Placenta

1233

Twin Gestation

Fig. 15 Mesenchymal dysplasia. The stem villi are greatly enlarged with abundant, focally degenerated cystic stroma and large thick muscular vessels

mole but show no trophoblastic hyperplasia. In addition, this rare lesion is often diagnosed as partial mole on ultrasound. Over one half of reported cases have been associated with Beckwith–Wiedemann syndrome. Mesenchymal dysplasia is associated with IUGR, intrauterine fetal demise (IUFD), and neonatal death (Jauniaux et al. 1997; Moscoso et al. 1991; Pham et al. 2006). Females are disproportionally affected.

Multiple Pregnancy Multiple gestations are common and becoming more so with assisted reproductive techniques (ART). In the United States in 2000, ART accounted for 14% of all twin births, and 45% of ART births were twins (Reynolds et al. 2003). Multiples are associated with a disproportionate share of complications including higher rates of morbidity, mortality, low birth weight, anomalous development, and malformation than singletons, and this is particularly true of monozygotic twins. Careful pathologic examination of the placenta(s) can provide important insight into problems peculiar to multiples, and pathologists must be aware of the special considerations required in the examination of placentas from multiple births.

Zygosity Definition. Twins may arise from the fertilization of two separate ova (dizygous or fraternal twins) or from the division of a single fertilized ovum (monozygous or identical twins). Monozygous twins are genetically and usually phenotypically identical, but dizygous twins are genetically dissimilar like singleton siblings. Rare variants of dizygous twinning result when ova are fertilized by sperm from different sources (superfecundation) or when ova ovulate and are fertilized at different times resulting in twins of disparate developmental ages (superfetation). A rare variant of monozygous twins, monozygous heterokaryotic twins, has different karyotypes and sometimes even different sex (Baldwin 1994). These twins are thought to result from chromosome nondisjunction, most commonly involving the sex chromosomes, but on occasion the autosomes as well. A third type of twinning, polar body or dispermic monovular twinning, presumably results from fertilization of an oocyte and a polar body. Bieber described a triploid acardiac/ diploid twin gestation resulting from fertilization of an oocyte and its first polar body (Bieber et al. 1981). Frequency and etiology. Twins occur in about 1 in 80 Caucasian pregnancies in the United States, and approximately 30% of these are monozygous. The frequency of monozygous twinning is relatively constant worldwide (about 3.5 in 1,000 pregnancies), but there are marked geographic differences in total twinning rate reflecting excess dizygous twinning in certain populations and families who have a genetic predisposition for high FSH (follicle-stimulating hormone) and polyovulation. Dizygous twinning is also age-related, increasing with maternal age until 35 years, again probably reflecting a role for increasing FSH levels with age. The cause of monozygous twinning is unknown. Monozygous twinning is increased in pregnancies following ART (Menezo and Sakkas 2002). ART has changed the epidemiology of multiple pregnancy. It is as yet unclear if or how this technology may

1234

affect developmental events (Chan et al. 2007). So far, pathologic placental lesions do not appear to be increased in ART multiples (Sato and Benirschke 2006a).

Placentation The object of placental examination in multiple gestation is the same as in singleton gestation – to identify pathologic processes affecting placental function. The gross and microscopic manifestations of these processes are identical in the placentas of singletons and multiples, although some lesions are particularly pertinent to the problems experienced by multiples. Comparison of the relative distribution of lesions in the placental territories and correlation with the size and condition of the fetuses is an important aspect of placental assessment in multiple gestation. Features unique to multiples include (1) the type of placentation, (2) the relationships of the disks and fetal membranes, and (3) the pattern and degree of vascular anastomoses.

R. N. Baergen et al.

Chorionicity, Amnionicity, and Placental Sharing Placentas in twin gestation are either monochorionic or dichorionic. Virtually all dizygous twins have dichorionic placentas (diamniotic dichorionic – DiDi). In double ovulation, each blastocyst generates a placenta. If these implant in close proximity, varying degrees of fusion may result (DiDi fused); otherwise, they are entirely separate. Monozygous twins may show any type of placentation depending on when division occurs (Fig. 16). If the single fertilized ovum divides very early, before differentiation of the chorion (first 5–6 days), the situation is analogous to dizygous twinning; two placentas develop that may be separate or fused (25% of monozygous twins). If splitting occurs after formation of the chorion but before formation of the amnion (eighth day after fertilization), there will be a single placenta with two amniotic sacs (diamniotic monochorionic – DiMo), occurring in about 75% of monozygous twins. A split after formation of the amnion (between the 8th and

Fig. 16 Diagrammatic representation of twin placentation. (After Fox 1997 used with permission of the American Registry of Pathology/Armed Forces Institute of Pathology)

19

Diseases of the Placenta

1235

Fig. 18 DiDi-fused placentas. The two placentas are fused but discrete. The dividing membranes are thick, and the fetal vessels do not cross the line of fusion. (Used with permission of the American Registry of Pathology/Armed Forces Institute of Pathology)

Fig. 17 Conjoined twins. (Used with permission of the American Registry of Pathology/Armed Forces Institute of Pathology)

15th days) will result in one placenta with one amniotic cavity (monoamniotic monochorionic – MoMo) and, still later splitting, in conjoined twins (Fig. 17). Practically speaking, twins with monochorionic placentas are monozygous with very rare exceptions (monovular dispermic and one case of dizygous monochorionic twins) (Souter et al. 2003). Twins with dichorionic placentas, whether separate or fused, may be either dizygous or monozygous. Obviously, different fetal sex establishes a dizygous relation, but further investigation (blood group analysis, human leukocyte antigen (HLA) typing, and DNA analysis) is required to determine zygosity in like-sex dichorionic twins. Establishment of placentation type is important, not only as an initial step in determining zygosity but primarily because it has an important relationship to the increased morbidity and

mortality that occur in multiple gestations. Twins with monochoriotic placentas have much higher mortality rates than those with dichorionic placentas, and monoamniotic placentation is associated with a fetal mortality rate as high as 50%. Placentation type is usually reliably established on gross examination. Two entirely separate placentas are obviously dichorionic (DiDi separate), each requiring routine examination. Frequently the disks are separate but the membranes are fused. The membranes of one twin almost invariably overlap the disk of the co-twin (irregular chorionic fusion). Thus, care must be taken when separating the placental disks of fused placentas, to divide them along the vascular equator and not at the dividing membranes. When the blastocysts implant close to one another, the two placental disks fuse to form an apparently single disk with a dividing septum (DiDi fused). The septum in dichorionic fused placentas is relatively thick and opaque due to the presence of chorionic tissue between the two amniotic layers and atrophied vessels which can be seen as white streaks in the membranes (Fig. 18). Because each embryo is developing within its own chorion, any fusion between them will contain chorionic tissue. The chorionic tissue in the septum is continuous with the underlying placenta remaining firmly attached to the fetal surface. If the amnions are separated, chorionic tissue will remain as a ridge at the base of the septum, which can be seen on

1236

R. N. Baergen et al.

Fig. 21 DiMo septum. The septum is composed of two directly apposed amnions Fig. 19 DiMo twin placenta. The dividing membranes are thin and filmy. Note velamentous insertion of the cord into the dividing membranes. (Used with permission of the American Registry of Pathology/Armed Forces Institute of Pathology)

Fig. 20 DiDi septum. A central layer of two chorions intervenes between amnions

prenatal ultrasound as the “twin peak” sign. In contrast, the septum in a DiMo placenta is thin and translucent, composed of two directly apposed layers of amnion (Fig. 19). This septum, devoid of chorion, is easily detached from the placental surface. Histologic examination of the membranous septum confirms the gross impression. This is most easily accomplished in a section of rolled septal membranes (Figs. 20 and 21), although identical information may be obtained from a section of the T zone where the septum meets the fetal surface. The latter method is

particularly useful when the septal membrane has been torn or otherwise distorted and cannot be rolled but is notoriously difficult to cut by pathologist and histotechnologist alike. The distribution of the fetal vessels on the chorionic plate is equally helpful in distinguishing DiDi fused from DiMo placentas. In dichorionic placentas, the chorionic plate vessels of each twin stop at the true vascular equator and as such define the vascular equator of the two placentas. In DiDifused placentas, fetal chorionic vessels approach but do not cross the area of fusion (Fig. 18). In DiMo placentas, portions of the same placenta are shared by both fetuses, and the two vascular districts are intimately intermingled (Fig. 22). Surface anastomoses between the two twins can be seen on the fetal surface by visible inspection and identification of unpaired vessels, i.e., an artery or vein which branches out and does not have a corresponding vein or artery. The position of the membranous septum in the DiMo placenta is independent of, and does not usually conform to, the fetal vascular districts. Rarely, a monochorionic placenta presents as two apparently distinct disks, superficially resembling dichorionic separate placentas. The nature of the septum and the intermingling of the vascular districts with anastomosis identify these as monochorionic. The appearance of two separate “disks” is due to involution or atrophy of intervening villous tissue related to factors at the implantation site affecting maternal blood supply.

19

Diseases of the Placenta

Fig. 22 DiMo twin placenta. The fetal surface shows anastomoses. (Used with permission of the American Registry of Pathology/Armed Forces Institute of Pathology)

MoMo placentas are very uncommon, accounting for less than 1% of twin gestations. In MoMo gestations, both twins develop within a single amniotic sac. Before considering a monochorionic placenta to be monoamniotic, the amnion between the cord insertions should be complete and continuous. Because the amnion commonly separates from the underlying chorion and is often nearly completely detached at delivery, more apparently single sacs are artifactual than truly monoamniotic, a so-called pseudomonochorionic placenta. Although membrane patterns are established early and persist throughout pregnancy, intragestational disruption of diamniotic septae has been reported. This suggests that at least some monoamniotic placentas may have been diamniotic originally. The umbilical cords in monoamniotic gestations are typically closely inserted, usually within 6 cm of each other (Fig. 23). Rarely there is partial cord fusion. Large-vessel anastomoses are frequent but not invariable between the closely inserted cords. Cord entanglement is common and a significant cause of morbidity and mortality (Fig. 24). The relative proportion of the chorionic surface populated by vessels from each twin and the relative placental mass serving each twin are pertinent observations in multiple gestations. Fused placentas can be divided and weighed. Estimates of the placental mass serving monochorionic twins are based on the chorionic vasculature. The pattern of venous return is a good indicator of placental supply. Not all discordantly grown

1237

Fig. 23 Cord insertion, MoMo. Close cord insertion is typical of MoMo placentation

Fig. 24 Cord entanglement in MoMo. This cord entanglement resulted in IUFD at 25 weeks

twins have asymmetric placentas, and not all twins with asymmetric placentas are discordantly grown, but these associations are not uncommon. Vascular Anastomoses An important feature of monochorionic placentas is the presence of vascular communications between the two fetal circulations. The fetal circulation is established when small vessels developing independently within the villous mesenchyme connect with larger vessels in the chorionic plate and body stalk. When twins share the same placenta, the potential exists for the two developing circulations to merge in a number of different ways. For example, the

1238

capillary bed supplied by an artery from one twin might establish continuity with a vein returning to its co-twin resulting in a parenchymal arteriovenous (A–V) anastomosis. In contrast, the two circulations in dichorionic twins are established independently explaining why vascular communications in DiDifused placentas are absent with rare exception. It is generally agreed that vascular communication is invariable in monochorionic placentas, although the number, size, and type of anastomoses are highly variable. Anastomoses between large chorionic plate vessels are very common. The majority of large-vessel anastomoses are between arteries (A–A); vein to vein (V–V) anastomoses are less common. Of greater physiologic significance are the arteriovenous (A–V) anastomoses that occur between capillaries deep within shared villous parenchyma. Vascular anastomoses may be compound with surface (large vessel) and parenchymal (villous capillary) connections involving the same vessels (Fig. 25). A–A anastomoses alone (20–28%) and A–A with A–V anastomoses (25–40%) are most common. A–V anastomoses alone, unmodified by concomitant large-vessel anastomoses, are estimated to occur in 11–20% of cases. Large-vessel anastomoses are easily identified grossly (Fig. 22) by the finding of unpaired vessels on the fetal surface, such as an artery that does not have a paired vein extending back to the umbilical cord (Fig. 26), keeping in mind that arteries cross over veins. The size, type (arteries cross over veins), and number of anastomoses should be recorded. A–V anastomoses are more important physiologically but are more difficult to identify (Fig. 27). Since injection studies are qualitative and do not necessarily reflect the physiologic significance of an anastomosis or the overall balance of blood flow in vivo, they are rarely performed. Nevertheless, their documentation can be essential if vascular anastomosis is to be invoked as a cause of, or contributor to, discordant fetal growth. Umbilical Cord in Multiple Pregnancy Anomalies of the umbilical cord are much more common in multiple than in singleton gestations. The incidence of velamentous cord insertion in twins has been reported to be up to nine times higher than in singletons. The frequency of

R. N. Baergen et al.

Fig. 25 Types of vascular anastomoses in monochorionic placentas. Large-vessel anastomoses (artery–artery and vein–vein) are easily identifiable on the chorionic plate. Arteriovenous anastomoses occur between capillaries in shared placental lobules. The presence of an unpaired artery penetrating the chorionic plate in the vicinity of an unpaired vein from its co-twin suggests this possibility. Superficial (large vessel) and capillary anastomoses may involve the same vessels (Used with permission of the American Registry of Pathology/Armed Forces Institute of Pathology)

Fig. 26 Vascular anastomosis. This large anastomosis has been highlighted by air injection. (Used with permission of the American Registry of Pathology/Armed Forces Institute of Pathology)

anomalous insertion (marginal and velamentous) increases with proximity of the twins (dichorionic separate < dichorionic fused < monochorionic). Anomalous cord insertion in multiples has the same significance as in singletons. In addition,

19

Diseases of the Placenta

1239

Fig. 27 Arteriovenous anastomosis. An unpaired artery from the umbilical cord barely visible at the bottom penetrates the chorionic plate adjacent to an unpaired dilated vein from the co-twin at the right. (Used with permission of the American Registry of Pathology/Armed Forces Institute of Pathology)

there is some evidence that velamentous cord insertion is a factor that influences twin–twin transfusion and associated early delivery. Single umbilical artery (SUA) is more frequent in multiple gestations, occurring in approximately 3% as compared to 0.53% of singletons. Most twins are discordant for SUA, although concordance is greater in monozygous than dizygous twins. The length of the umbilical cord is on average 7.6 cm shorter in twins than singletons, and the incidence of hypocoiled cords is also increased. Twins that share an amniotic cavity (monoamniotic) are at an increased risk for cord entanglements. Cord entanglements of all types occur in monochorionic twins with a reported frequency between 53% and 71%, and they may be remarkably complex (Fig. 28). Such entanglements may result not only in cord compression but in mistaken cord transection when one twin’s cord is looped around its co-twin at delivery. Cord entanglement as a factor in fetal mortality is most common prior to 24 weeks while there is still sufficient room for fetal movement. After 30–32 weeks and in higher multiples, cord entanglement is less common as opportunity for movement decreases. Cord entanglements may result in chronic vascular compromise, or they may not become a problem until the forces of labor result in acute vascular obstruction.

Fig. 28 Complex cord entanglements in MoMo twins (a and b)

Complications of Multiple Pregnancy Twin–Twin Transfusion Syndrome (TTTS) Vascular communications in monochorionic placentas create the potential for blood flow between the twins. When there is a net flow of blood from one twin to the other, the clinical manifestations depend on the size, number, and type of vascular communications. Capillarysized arteriovenous (A–V) anastomoses involve the transfusion of small amounts of blood over long periods of time (chronic transfusion). A large volume of blood can be transferred rapidly through large-caliber chorionic plate vessels at the time of labor and delivery (acute transfusion) with demise of one twin. Not infrequently, both occur – an acute transfusion is superimposed on a chronic long-term process (acute on chronic transfusion). Definition and etiology. TTTS or chronic twin transfusion is an important cause of perinatal mortality in monochorionic twins. Schatz proposed the view that the chronic TTTS results when

1240

there is unbalanced flow of blood from one twin (the donor) to its co-twin (the recipient) through A–V anastomoses deep within shared placental lobules (the “third circulation”) (Benirschke et al. 2006). Chronic unidirectional diversion of blood results in anemia, relative deprivation, oligohydramnios, and growth restriction of the donor as compared to the larger, polycythemic recipient with polyhydramnios. Hypervolemia and hypertension in the recipient result in increased urine output and polyhydramnios, while hypotension and hypovolemia in the donor lead to oliguria, oligohydramnios, and decreased movement (the “stuck” twin.) Either twin may be hydropic, reflecting cardiac dysfunction and congestive heart failure in the recipient and profound anemia in the donor. Although vascular anastomoses are a necessary precondition for the development of the TTTS, additional factors beyond shifts of blood (differential protein, atriopeptin, growth factor concentration, or colloid osmotic pressure) may contribute. Velamentous cord insertion is significantly more common in monochorionic gestations with the TTTS. It has been proposed that velamentous cords are more easily compressed, resulting in decreased umbilical vein flow and enhanced flow through A–V anastomoses. Others have speculated that polyhydramnios may contribute to umbilical and chorionic vessel compression providing the therapeutic rationale for amniocentesis. The clinical definition of the TTTS is complicated because twins commonly show asymmetric growth for reasons other than a chronic twin–twin transfusion (maternal, fetal, umbilical cord, or placental factors). A hemoglobin concentration difference > 5 g/100 mL and twin weight difference of 15–20% are considered definitive criteria by some, although similar weight and hemoglobin discrepancies are just as common in dichorionic twins without an anatomic basis for transfusion (Redline et al. 2001). A relatively selective criterion is a recipient heart weight that is two to four times that of the donor. This feature is generally not present in any other process resulting in twin weight–growth discordance. The disparity in

R. N. Baergen et al.

heart size is due not only to a generalized growth differential but also to the increased cardiac work load associated with the hypervolemia, hypertension, and hyperviscosity experienced by the recipient twin. Heart size discrepancy may be the first manifestation of the chronic TTTS, evident on ultrasound as early as 10 weeks. Difference in overall fetal size is usually apparent only later in gestation. Discrepant amniotic fluid volumes, hydrops, and a “stuck twin” are other ultrasound findings very suggestive of the TTTS and in fact are used by clinicians in the diagnosis of TTTS. Quintero (Kontopoulos et al. 2016) has recently developed a clinical definition and staging system of TTTS used for clinical evaluation and treatment. Frequency. Recent studies indicate close correspondence between the number of cases with well-documented twin–twin transfusion and the number of monochorionic twin placentas with A–V anastomoses without concomitant largevessel anastomoses (Benirschke and Masliah 2001). This occurs in 9–20% of monochorionic gestations. Large superficial vascular anastomoses appear to mitigate against the effects of A–V transfusion probably by allowing transfer of blood back to the donor from the recipient, thus equilibrating the two circulations. In the absence of superficial anastomoses, arteriovenous transfusions are uncompensated, resulting in marked growth discordance, hydramnios, early delivery, and high perinatal mortality. The TTTS is not a significant cause of fetal mortality in monoamniotic twins presumably because the majority of MoMo twins have large-vessel anastomoses that prevent preferential shunting of blood from one twin to the other. For unexplained reasons, the TTTS is more common in females. Gross and microscopic pathology. After delivery, there is a marked discrepancy in the size and appearance of the infants and their corresponding placental territories. The donor twin is smaller, pale, and anemic; the recipient is heavier, edematous, plethoric, and polycythemic (Fig. 29). Either twin may be hydropic. The organs of the recipient are larger and heavier than those of the donor (Fig. 30). The recipient’s heart especially is comparatively enlarged, with

19

Diseases of the Placenta

Fig. 29 TTTS. The donor twin (left) is smaller and pale. The recipient twin (right) is larger and plethoric. (Used with permission of the American Registry of Pathology/ Armed Forces Institute of Pathology)

myocardial hypertrophy involving all chambers. Smooth muscle mass is increased in the media of pulmonary and systemic arteries and arterioles. Pulmonary arterial calcification has been described in the recipient twin. The donor heart is usually small, and arterial muscle mass is decreased. Glomeruli are enlarged, up to twice normal size in the recipient twin, and they are either normal or small in the donor. The donor’s placental territory may be large, bulky, and pale reflecting fetal anemia (Fig. 31). The villi are large and edematous with numerous Hofbauer cells and capillaries containing nucleated red blood cells (Fig. 32). Amnion nodosum may be found when there is associated oligohydramnios. The recipient’s placental territory is generally smaller, firm, and deep red. The

1241

Fig. 30 TTTS. The organs of the recipient twin are larger and heavier (top). Congestion of the smaller donor organs (bottom) and pallor of the larger recipient organs (top) are a reversal of the expected pattern, caused by acute transfusion of the recipient twin to the donor twin after intrauterine death of the donor. (Used with permission of the American Registry of Pathology/Armed Forces Institute of Pathology)

villi are appropriately mature with dilated and intensely congested vessels (Fig. 33). Clinical features. Typically, the chronic TTTS is clinically manifest in the second trimester with acute hydramnios and growth discrepancy between the twins, but it may be suspected earlier based on the ultrasound demonstration of different amniotic fluid volumes. The consequences of the TTTS are grave. Mortality rates are as high as 70–100% depending on the gestational age at diagnosis and delivery. The earlier the syndrome is manifest, the more likely the outcome will be fatal. When the condition develops in the second trimester, it is often

1242

Fig. 31 TTTS. The donor’s placental territory (left) is larger and pale reflecting anemia. The recipient’s territory is smaller and congested. (Used with permission of the American Registry of Pathology/Armed Forces Institute of Pathology)

Fig. 32 TTTS. The donor’s villi are large, relatively immature, and edematous with numerous immature erythroid precursors

R. N. Baergen et al.

associated with PTL and death of one or both fetuses or significant morbidity if the neonates survive. Both twins are at great risk. The recipient twin is subject to cardiac failure, hemolytic jaundice, kernicterus, and thrombosis due to hemoconcentration. The donor twin may be severely anemic or hypoglycemic. Both twins are likely to suffer from complications of prematurity and heart failure. Multiorgan necrotic lesions including white matter infarcts, leukoencephalopathy, hydranencephaly, porencephaly, intestinal atresia, renal cortical necrosis, and aplasia cutis (Fig. 34) may occur in either or both twins. These lesions are attributed to the altered hemodynamics, transitory cardiovascular compromise, and hypoperfusion associated with complex fluctuating placental vascular connections. Attempts to alter TTTS have included treatment with indomethacin to reduce fetal urine output and polyhydramnios, digoxin to treat congestive heart failure, decompressive amniocentesis to prolong pregnancy and affect fetal blood flow, division of intervening membranes, selective feticide, and laser ablation of anastomoses. The latter approach requires accurate prenatal mapping of physiologically significant arteriovenous anastomoses and their successful ablation in vivo. Increasingly sophisticated Doppler studies are used to tailor treatment strategies to individual vascular patterns and to assess the results of intervention. Pathologic evaluation of laser sites

Fig. 33 TTTS. The recipient’s villi are appropriately mature but congested (left). The contrast between the donor and recipient placental territories is dramatic (right)

19

Diseases of the Placenta

Fig. 34 TTTS. Both twins may have necrotic lesions including aplasia cutis as seen here in the donor twin. (Used with permission of the American Registry of Pathology/Armed Forces Institute of Pathology)

and mapping of residual anastomoses play an important part in this treatment strategy. The enthusiasm for intervention must be tempered by the knowledge that the two fetal circulations are connected and that any manipulation will potentially affect both fetuses. Given the complexity and uniqueness of the vascular anatomy, it is difficult to predict the consequences. Acute Transfusion If one twin dies in utero, regardless whether chronic twin transfusion or TTTS is present, the surviving twin may exsanguinate into the suddenly relaxed circulatory system of the dead co-twin (Fig. 30). Acute hypotension and/or blood loss after co-twin death is considered a likely explanation for necrotic lesions in the surviving twin, a view supported by Doppler studies showing dramatic umbilical flow velocity changes in the surviving twin. It has been estimated that neurologic sequelae occur in as many as 27% of surviving twins. Cerebral lesions have been demonstrated sonographically in twins at birth and in utero, developing rapidly in some cases. These observations are important in attempting to explain the high incidence of cerebral palsy (CP) in twins, reportedly five times the incidence in singletons and affecting primarily monozygous twins. Recognition that lesions predictive of CP are of prenatal onset is important not only in medicolegal considerations but also when

1243

contemplating the possible consequences of intentional fetal reduction for the surviving twin. Plethora of one twin and pallor of the other do not always signify a chronic transfusion but may reflect acute shifts of blood through large superficial vascular anastomoses. A large blood volume may shift rapidly through large-vessel connections. Simple acute transfusions are usually diagnosed at birth. The donor is pale and the recipient plethoric, but there is no growth differential, and hemoglobin and hematocrit levels are equal in initial assessments. Acute on Chronic Transfusion When superimposition of acute transfusion on established chronic transfusion occurs, there is reversal in the expected pattern with plethora of the smaller, original donor twin and pallor of the larger, recipient twin (Fig. 30). Whether this pattern develops after the donor dies and the recipient’s blood drains into the donor’s flaccid vascular tree, or the acute transfusion occurs first, overwhelming the capacity of the donor’s heart causing death is unknown. Both mechanisms may be operative. On occasion, an acute transfusion and/or fetomaternal hemorrhage may occur in twins with discordant growth for reasons unrelated to chronic transfusion. While this scenario may mimic an acute on chronic transfusion, the lack of evidence for adaptation to chronic transfusion (no heart size discrepancy, polyhydramnios, or oligohydramnios) may help exclude chronic transfusion as a relevant contributing factor.

Asymmetric Growth Established growth standards indicate that the growth curves for dichorionic and monochorionic twins approximate those of singletons prior to 30–34 weeks (Ananth et al. 1998; Kraus et al. 2004). Twins weigh progressively less than singletons as pregnancy advances. When twin growth is discordant, the larger twin approximates the growth of an age-matched singleton, and the growth rate of the smaller twin slows and may gradually decline into the range of small for gestational age (SGA). In dichorionic twins, growth discordancy is usually manifest around 25 weeks,

1244

but in monochorionic twins, the onset is more variable, commencing in some cases as early as 18–20 weeks. Twin birth weight discordance is strongly associated with preterm birth (PTB), perinatal death, and postnatal morbidity (Cooperstock et al. 2000). The majority of twins with discordant growth are dichorionic, and even among monochorionic twins, TTTS is not the most common cause of growth discrepancy. Even when chronic transfusion occurs, other factors may contribute as much or more to discordant growth. Cord anomalies including velamentous insertion and SUA are associated with decreased fetal weight, and both are much more frequent in twin gestations. In one recent study, peripheral cord insertion (velamentous, marginal, and markedly eccentric) was the strongest predictor of discordant growth and SGA in dichorionic and monochorionic twins (Redline et al. 2001). Discordant birth weight is also associated with placental mass; smaller babies have smaller placentas. Poor placentation with asymmetrical trophotropism and abnormal cord insertion, asymmetric placental volume, and decreased placental vascularization are major causative factors in IUGR and twin growth discordance (Bruner et al. 1998). Whether placental and fetal growth asymmetries reflect a generalized problem in the conceptus or result from a primary placental problem is unclear. While discordant parenchymal lesions present a possible explanation for growth asymmetry, most placental disease processes involve twin placentas to similar degree; concordance for most findings is higher than might be expected based on their prevalence in singleton placentas. In one recent study, the only lesion significantly associated with discordant SGA twins was the finding of avascular villi (FAV) (Redline et al. 2001). This lesion, indicative of fetal vascular occlusion (fetal thrombotic vasculopathy) and associated with neurologic impairment in term infants, might contribute to the increased morbidity and mortality in twins. The incidence of fetal vessel thrombosis is higher in monochorionic than dichorionic placentas, correlating with peripheral cord insertion, vascular cushions, and IUGR (Sato and Benirschke 2006a).

R. N. Baergen et al.

Vanishing Twin Twin gestations are commonly converted to singletons with the embryonic/early fetal death of one twin (vanishing twin). Of twins diagnosed before 10 weeks, about 70% are born as singletons. When a twin dies spontaneously early in the first trimester, it is often difficult or impossible to identify any residue. A flattened mass of atrophic placenta may remain as a vaguely thickened area in the membranes of the surviving co-twin. Microscopic examination may identify the membranous septum and establish chorionicity. Rarely there is an embryonic remnant, invariably severely degenerated (Fig. 35). Electively reduced embryonic remnants are more easily identified (Fig. 36). Fetus Papyraceus A twin dying in the second trimester may be retained, progressively compressed, and flattened by the growth of the co-twin. Depending on the length of retention, this flattened dead twin, a fetus papyraceus, is variably altered (Fig. 37). Some are still easily recognizable, and others resemble amorphous necrotic material. Specimen radiographs may reveal skeletal remnants. It has been estimated that the creation of a fetus papyraceus takes about 10 weeks. The degree of maceration and autolysis of both the

Fig. 35 Early twin demise. This growth-disorganized tiny twin was delivered at term with its normal co-twin. The abnormal morphology suggests that this early twin death was the result of an abnormal karyotype

19

Diseases of the Placenta

1245

surviving the in utero death of a co-twin. The risk of serious cerebral damage in surviving twins is reportedly around 20% and may be even higher in certain subsets. Monochorionic twins appear to be at greatest risk (Pharaoh and Adi 2000), and much of the morbidity can be attributed to the vascular anastomoses between these twins. Survivors of co-twin death also have a higher postnatal mortality rate.

Fig. 36 Multiple pregnancy with elective reduction. Reduced fetuses are usually about the same size and relatively easily identified in the membranes. (Used with permission of the American Registry of Pathology/Armed Forces Institute of Pathology)

Fig. 37 Fetus papyraceus

fetus and placental tissue generally precludes establishment of the cause of death, although anomalous cord insertion and TTTS have been implicated in many cases. Fetus papyraceus may occur in both monochorionic and dichorionic gestations. The amniotic cavity of the fetus papyraceus is gradually compressed as the amniotic fluid is resorbed. The corresponding placenta is generally pale, thinned, and atrophied with avascular villi enmeshed in fibrin. Survivors of Co-twin Demise Abnormalities including intestinal atresia, skin defects, amputations, gastroschisis, and especially brain damage have been reported in twins

Duplication Abnormalities: Conjoined Twins Occasional monozygous twins show marked discrepancies in size and configuration or varying degrees of incomplete separation. These asymmetric and incomplete duplications include acardiac and parasitic twins and conjoined fetuses. A parasitic twin is a variably developed fetiform mass attached to its co-twin either internally or externally (Fig. 38). Conjoined twins retain their overall symmetry but are either incompletely separated or possibly secondarily fused during development. Neither parasitic nor conjoined twins show specific placental abnormalities. Acardiac Twin (Chorangiopagus Parasiticus) Acardiac twinning is the commonest form of asymmetric duplication anomaly, occurring in 1% of monochorionic twins. The acardius is a grossly malformed, often bizarre fetus of variable size, appearance, and degree of organogenesis. No two are alike. Some are amorphous, shapeless masses resembling teratomas (Fig. 39), and others are remarkably well developed. Commonly, acardiac fetuses have relatively well-developed lower bodies including the legs and perineal structures, a trunk into which the umbilical cord inserts, and a rounded, dome-like upper body (Fig. 40). A single body cavity may contain abdominal viscera, but thoracic structures and the heart are typically absent, and the upper body is very poorly formed. Cardiac remnants or a deformed heart may be identified on occasion. Organ development is highly variable; some acardiacs may demonstrate absence of most organs, while in others, the organs may be well

1246

R. N. Baergen et al.

Fig. 38 Parasitic twin. This parasitic twin (a and b) was removed from the abdominal cavity of its co-twin. (Used with permission of the American Registry of Pathology/Armed Forces Institute of Pathology)

Fig. 39 Acardiac fetus. This amorphous acardiac fetus had a SUA. (Used with permission of the American Registry of Pathology/Armed Forces Institute of Pathology)

Fig. 40 Acardiac fetus. A dome-shaped upper body with lower extremities is a relatively common configuration in acardiac fetuses. (Used with permission of the American Registry of Pathology/Armed Forces Institute of Pathology)

developed. Acardiacs may be hydropic, and some are larger than the co-twin. The essential feature common to all acardiac fetuses is their circulation, which is maintained entirely by the co-twin (the “pump” twin) (Benirschke et al. 2012). Blood from the normal “pump” twin reaches the acardiac through an artery to artery anastomosis, flows through the acardiac in reverse course, and then returns to the normal twin through a vein-to-vein anastomosis (twin reversed arterial perfusion – TRAP). The majority of acardiacs have a SUA. The specific vascular communications between the acardiac and pump twins are large, occurring at the level of the umbilical cord or chorionic plate. There are no placental parenchymal vascular connections, and therefore the acardiac is analogous to a conjoined twin and not to the TTTS. The histologic features of the placenta in acardiac twinning have not been described in detail. Villous immaturity and recent and remote thrombosis of umbilical and chorionic plate veins have been reported (Steffensen et al. 2008). Acardiacs may occur in diamniotic or monoamniotic monochorionic placentas, although monoamniotic placentation is especially prevalent. Acardiacs are greatly overrepresented in triplet and higher multiple gestations. The pump twin is at risk for cardiovascular overload. The extra work involved in circulating blood through the acardiac may result in

19

Diseases of the Placenta

1247

cardiomegaly and high-output failure associated with hydrops and hydramnios. Pump twins also tend to be growth impaired perhaps because the blood diverted away from the placenta to the acardiac returns deoxygenated. Many pump twins have congenital anomalies. Classification of the acardiac anomaly based on prognostic factors including the cardiovascular condition of the pump twin and the size of the acardiac twin has been proposed (Wong and Sepulveda 2005). Strategies similar to those employed in management of the TTTS have been successful in ameliorating the effects of cardiac overload in the pump twin (Tan and Sepulveda 2003).

Higher Multiple Births Examination of placentas from higher multiple births extends the observations applied to twins. Chorionicity, membrane relationships, cord insertion, chorionic vascularity, potential vascular anastomoses, and relative placental volumes are assessed as in twins. Patterns of placentation vary depending on zygosity. Triplets,

for example, may be monochorionic, dichorionic, or trichorionic with variable amnionicity (Figs. 41 and 42). Multichorionic placentas are usually fused as space in the uterus is limited. Combinations of monozygous and polyzygous multiples are common. With increasing use of assisted reproductive techniques, triplets and even quadruplets or quintuplets are no longer unusual in many institutions (Fig. 43). Higher multiples experience the same complications as twins – prematurity, low birth weight, congenital anomalies, and increased perinatal morbidity/mortality – all progressively problematic with increasing numbers. Acardiacs are much more frequent in higher multiples as compared to twins. Cord entanglement is less common in monoamniotic higher-order multiples presumably due to limited opportunity for movement in the increasingly crowded uterus. While the outlook in higher multiple births has improved, complications are common, and the outcome is often poor. Consequently, selective fetal reduction has been advocated, and there is now a substantial body of accumulated experience. Remnants of reduced fetuses are regularly identified when carefully sought.

Higher multiple birth triplets Monochorionic monozygous Triamniotic monoamniotic

Diamniotic monoamniotic

Monochoriotic monoamniotic

Dichorionic mono or dizygous Triamniotic dichorionic

Trichorionic mono, di or trizygous

Diamniotic dichorionic Placenta Chorion Amnion

Fig. 41 Triplet placentation. (After Fox 1997)

1248

Fig. 42 Triplet placentation. Tri-Tri triplets (left). DiMo triplets (right). Although the placental territory at the left is discrete, the fetal vascular pattern confirms

R. N. Baergen et al.

monochorionicity. The middle and right triplets are monoamniotic

from the maternal bloodstream. Passage through an infected birth canal is an important mode of fetal infection, but the placenta is not involved in this circumstance. The consequences of intrauterine infection and inflammation include abortion, stillbirth, PTB, fetal malformation, active postnatal infection, and long-term sequelae, frequently neurologic. At worst, affected children may be severely handicapped, retarded, blind, or deaf. The social, financial, and emotional burden posed by the support of these children is enormous. Fig. 43 Quintuplet placentas

Ascending Infection and ACA

Placental Inflammatory Lesions Placental inflammation is a common and important histologic finding. The character of a given inflammatory infiltrate and its pattern of involvement reflect, in general, its etiology and determine its clinical significance (Redline 2002). Placental inflammatory infiltrates can be divided into two major categories: (1) acute chorioamnionitis (ACA), which is invariably infectious due to microorganisms, usually bacteria that ascend from the vagina or cervix (ascending infection), and (2) villitis – inflammation, usually chronic, of the villous parenchyma, which is usually idiopathic, but rarely infectious from viruses, some bacteria, or protozoa that reach the placenta homogeneously

Frequency, etiology, and pathogenesis. ACA is the most common form of placental infection and inflammation in humans (Redline 2007a). Histologic ACA is found in approximately 10–15% of term and in up to 50–70% of preterm placentas. This striking inverse relationship between ACA and gestational age emphasizes the strong association between ACA and PTB. The term “chorioamnionitis” when used by clinicians to describe a symptom complex suggestive of intrauterine infection (maternal fever, uterine tenderness, or elevated white blood cell count) correlates poorly with histologic chorioamnionitis. The latter, however, is diagnostic of an intrauterine infection. While historically controversial, it is now clear that ACA is a maternal and fetal acute

19

Diseases of the Placenta

inflammatory response to infectious agents that have gained access to the gestational sac. Microorganisms, usually bacteria, have been cultured from fetal membranes, amniotic fluid, cord blood, and fetal tissues in a high percentage of cases (Zhang et al. 1985) and have been demonstrated by PCR in culture-negative cases (Gardella et al. 2004). ACA can be produced experimentally in animals by the intraamniotic injection of bacteria but not by sterile exogenous irritants. Most cases of ACA result from the ascent of vaginal or cervical bacteria into the uterus. Other rare sources of intrauterine infection include endometritis or contiguous spread of organisms from the fallopian tubes, bladder, appendix, or intestines. Recent epidemiologic and experimental data have linked periodontal disease due to Fusobacterium nucleatum to PTL and have suggested that these microorganisms may reach the decidua hematogenously in some cases (Han et al. 2004; Hill 1998). In the setting of an intrauterine infection, ACA almost always precedes membrane rupture and in fact is the likely cause of rupture in those cases. That being said, if there is spontaneous membrane rupture without a preexisting infection, the risk of subsequent infection is slightly increased. Outcome studies indicate that ACA with intact membranes is a distinct entity with a greater risk to the developing fetus than ACA following membrane rupture, although the histologic changes are identical (Redline 2004b). Susceptibility to ACA is

1249

likely strongly related to factors that facilitate access of microorganisms to the intrauterine environment as well as the virulence of, and host response to, a given infectious agent. Pathology: maternal response. In most cases of ACA, the fetal membranes are macroscopically normal. In cases of particularly severe, longstanding infection, they may be opaque, friable, or foul smelling. Histologically there is a maternal and then later a fetal inflammatory response to amniotic infection that progress sequentially and predictably as originally described by Blanc (Blanc 1959). The initial response is a localized response by the maternal circulation wherein neutrophils migrate from the membranous decidua into the membranes at the bacterial entry point, the cervical os. Once bacteria have gained access to the amniotic fluid, there is migration of maternal neutrophils from the intervillous space as well, lining the underside of the chorionic plate. Neutrophils accumulate first in the subchorionic fibrin (Fig. 44), migrating progressively through the connective tissue layer of the chorion and amnion (Fig. 45). With time, the neutrophils undergo apoptosis and karyorrhexis followed by amniotic epithelial necrosis (necrotizing chorioamnionitis) (Fig. 46). The sequential changes and disease progression (stage) in the maternal inflammatory response are constant. A clinically relevant scheme for staging and grading the maternal inflammatory response has been consolidated by

Fig. 44 ACA, maternal response, early. Maternal neutrophils migrate out of the intervillous space aggregating in the subchorionic fibrin (left) and from decidual vessels infiltrating the extraplacental membranes (right)

1250

the Amsterdam group (Table 1) (Khong et al. 2016). The time course of progression in the maternal inflammatory response has been estimated based on personal observation and clinical correlation as well as experimental data. According to Redline, the initial maternal inflammatory response occurs within 6–12 h of infection. Involvement of amniotic connective tissue (ACA) most probably develops over 12–36 h after which the initial wave of neutrophils undergoes apoptosis and karyorrhexis (36–48 h) (Redline 2006a). Whether and how the timing of the maternal response may be altered by maternal factors

Fig. 45 ACA, maternal response, intermediate. Maternal neutrophils migrate into the connective tissue of the chorion and amnion

R. N. Baergen et al.

such as preexisting maternal antibodies or the characteristics of the specific causative organism is not understood. The intensity (grade) of the maternal inflammatory response is more difficult to quantify. Grading based on the number of neutrophils in the most inflamed area of the chorionic plate has been suggested.

Ascending Infection Fetal inflammatory response: Depending on the gestational age and status of the fetal immune system, the fetus may also respond to amniotic infection. A fetal leukocytic response is often absent in gestations less than 19–20 weeks and in fetuses less than 500 g. The first manifestation of a fetal inflammatory reaction is the migration of fetal neutrophils into the umbilical vein (umbilical phlebitis) and/or chorionic plate vessels (chorionic vasculitis). The migration of inflammatory cells is crescent shaped, oriented toward the source of infection in the amniotic cavity (Fig. 47). In later stages, fetal neutrophils migrate from the umbilical arteries (umbilical arteritis) and into Wharton’s jelly. With increasing duration, the neutrophils undergo necrosis forming necrotic bands around the umbilical vessels (necrotizing/subacute necrotizing funisitis) that may undergo calcification or vascularization (Fig. 48). Umbilical cord inflammation is often segmental, sometimes identified in only one of multiple sections, often from the fetal end.

Fig. 46 ACA, maternal response (severe). Diffuse intense inflammation with karyorrhexis of neutrophils and amniotic epithelial necrosis (left) and chorionic microabscesses (right)

19

Diseases of the Placenta

Table 1 Staging of ACA (Modified from Khong et al. 2016) Maternal response Stage 1 early Acute subchorionitis or chorionitis Stage ACA 2 intermediate Stage 3 late Necrotizing chorioamnionitis Grade severe Confluent inflammation with or without subchorionic microabscesses Fetal response Stage 1 early Umbilical phlebitis or chorionic vasculitis Stage Umbilical phlebitis and arteritis 2 intermediate Stage 3 late Necrotizing funisitis Grade severe Near confluent inflammation with attenuation of vascular smooth muscle

Fig. 47 ACA, fetal response (early). Fetal neutrophils migrate into the chorionic plate vessels (chorionic vasculitis)

Progression of the fetal inflammatory response has also been staged (Table 1), although the rate of progression of the fetal inflammatory response is more variable than the maternal response, depending significantly on gestational age. The intensity (grade) of the fetal inflammatory response has important associations with adverse fetal outcome. Scattered neutrophils in the chorionic plate or umbilical vessels are considered to be mild to moderate (grade 1). Intense chorionic vasculitis (severe/grade 2) is characterized by near-confluent neutrophilic infiltration of vessels accompanied by attenuation or disarray of the vascular smooth muscle media (Fig. 49). Inflamed vessels may be thrombosed.

1251

Amniotic infection, then, is a unique situation in which two individuals, mother and fetus, respond to the same infectious insult. The acute inflammatory infiltrate in chorioamnionitis is characteristically confined to the fetal membranes and umbilical cord. The villous parenchyma is not involved unless fetal or maternal septicemia results secondarily in villitis. In this event, bacteria (usually Escherichia coli, group B streptococci) are present in villous capillaries, often with accumulation of neutrophils beneath the trophoblastic basement membrane (Fig. 50). Acute villitis and intervillous abscesses are most commonly due to Listeria monocytogenes. Although the majority of leukocytes in the fetal membranes are maternal, fluorescence in situ hybridization studies have demonstrated that the majority of neutrophils in amniotic fluid and fetal lung are fetal, emphasizing the important contributions of both maternal and fetal immune systems. The character of the inflammatory infiltrate is usually not specific enough to identify a particular offending agent. In fact, it is relatively unusual to find bacteria in histologic sections even when they have been demonstrated on smears of the amnion. Notable exceptions to this include infections with group B streptococci in which colonies of the organism are frequently found without difficulty even in the no or minimal histologic evidence of chorioamnionitis. Fusobacterium species may also be visible on conventional hematoxylin and eosin (H&E) stains as long (at least 15 μm), faintly basophilic wavy organisms often associated with very severe inflammation and necrosis in the membranes. Fusobacteria may be demonstrated with silver stains but are only faintly gramnegative on tissue Gram stain. In rare instances of Candida infection, tiny white fungal colonies 2–3 mm in size may be seen grossly on the amniotic surface of the umbilical cord. Yeast and hyphal forms Candida may also be identified on conventional H&E stains, characteristically in small, superficial, crescentic microabscesses under the amniotic surface of the umbilical cord (peripheral funisitis) (Fig. 51). The organisms are easily seen with stains for organisms such as periodic acid–Schiff (PAS) or Grocott’s methenamine silver (GMS).

1252

R. N. Baergen et al.

Fig. 48 ACA, fetal response (late). Necrotizing funisitis. Prominent crescentic band of necrotic leukocytes around umbilical vessels can be appreciated grossly

Fig. 49 ACA, fetal response (severe). Fetal neutrophilic infiltration of chorionic vessels with medial disarray and in this case thrombosis

Fig. 50 Acute villitis. Bacterial colonies are prominent in villous capillaries, and neutrophils have accumulated beneath the trophoblast in the peripheral villous stroma

Clinical significance. Although there are maternal hazards associated with chorioamnionitis (maternal sepsis), the principal clinical impact of chorioamnionitis is its potential adverse effects on the fetus. PTB. It is estimated that approximately 70% of perinatal mortality and nearly half of longterm neurologic morbidity can be attributed to PTB (Esplin 2006). Multiple factors contribute to PTB, but chorioamnionitis is implicated in a high percentage of cases. Intrauterine infection as a cause of PTB is usually asymptomatic until labor begins or the membranes rupture prematurely, and therefore, early diagnosis is difficult. Many recent strategies have been aimed at the identification of women at risk and the development of therapeutic interventions targeting infection and the inflammatory response to reduce spontaneous PTB and its associated mortality and long-term morbidity. Data accumulating from experimental and human studies are clarifying how bacterial infection and chorioamnionitis result in spontaneous PTB. The inflammatory response to bacterial invasion results in the production of cytokines that initiate prostaglandin synthesis, resulting in uterine contractions. Certain bacterial species commonly associated with chorioamnionitis are high in phospholipase A2, which releases the prostaglandin precursor arachidonic acid from membrane phospholipids. The inflammatory response also results

19

Diseases of the Placenta

1253

Fig. 51 Candida peripheral funisitis. Small yellow foci on the umbilical cord correspond to superficial microabscesses

in increased synthesis of metalloproteases that are thought to remodel and soften cervical collagen and to degrade the extracellular matrix of the fetal membranes, leading to membrane rupture (Kumar et al. 2006). Approximately 30% of cases of ACA at less than 25 weeks are associated with retroplacental hematoma that occurs due to decidual bleeding from inflammation. This may lead to a “chronic abruption” eventuating in delivery and even death in many (Redline 2004b). Fetal inflammatory response syndrome (FIRS). Traditional thinking has attributed many of the neonatal complications of chorioamnionitis-induced PTB to prematurity, not to infection per se. Evidence is accumulating that pathologic lesions predictive of long-term morbidity (periventricular leukomalacia, intraventricular hemorrhage) are initiated in utero by the fetal response to placental infection (FIRS). Most data would not implicate infectious organisms as a direct cause of tissue damage but rather mediators of the fetal inflammatory reaction, principally cytokines (IL1, IL6, IL8, TNF-α) in the genesis of fetal lesions, especially white matter damage predictive of CP (Gomez et al. 1998; Leviton et al. 1999). Cytokines may cause white matter damage directly via a direct toxic effect or indirectly by activating endothelium and microglia resulting in thrombosis or increased vascular permeability. In preterm infants, cytokines also interfere with maturation of oligodendrocytes. Whether endotoxins or exotoxins cross the

placenta or blood–brain barrier to directly injure the central nervous system (CNS) is unknown. The only histologic finding directly correlated with CNS injury in ACA is the severity of the fetal inflammatory response and associated fetal thrombosis. Umbilical arteritis is associated with higher levels of circulating fetal cytokines than inflammation of the umbilical vein alone (Kim et al. 2001; Rogers et al. 2002). Intense chorionic vasculitis has been associated with increased risk of CP in term and preterm infants (Redline 2005; Redline et al. 1998b). The concomitant presence of mural thrombi in inflamed chorionic plate vessels is an additional risk factor for neurologic impairment in very low birth weight (VLBW) infants (Redline et al. 1998b). Neonatal infection. ACA implies that the fetus has been exposed to infection but not necessarily that the fetus is infected. The exposed fetus may be infected through the fetal skin, eyes, nose, or ear canals or by inspiring or swallowing infected amniotic fluid. Infants whose placentas show ACA are at increased risk to develop sepsis and die in the neonatal period. Although neither mother nor neonate is overtly ill in most cases, neonatal infection is an important cause of perinatal death, and most serious neonatal infections are associated with ACA. Fetal outcome is determined by a number of factors including gestational age and causative organism. Group B β-hemolytic streptococcus, E. coli, and Haemophilus influenzae are the most common causes of significant neonatal infection. Placental features associated with increased risk of neonatal infection include severe maternal and

1254

fetal inflammatory responses (Keenan et al. 1977; Zhang et al. 1985).

Subacute Chorioamnionitis Subacute chorioamnionitis is characterized by a mixed inflammatory infiltrate of mononuclear cells and degenerating neutrophils, with associated necrosis. This pattern may result from ongoing low-grade or repetitive infection and has been associated with chronic lung disease in a subgroup of VLBW infants with concomitant amniotic necrosis. The maternal history may include repetitive bouts of vaginal bleeding (Ohyama et al. 2002).

R. N. Baergen et al.

villitides is an inflammatory infiltrate in the villi, but its character and the nature of associated changes are variable from case to case. The inflammatory infiltrate is usually chronic, composed of lymphocytes, histiocytes, and plasma cells in variable proportion (Fig. 52), but occasionally it may be granulomatous, and rarely neutrophils are prominent. Necrosis, dystrophic calcification, and stromal hemosiderin deposition may be present in some cases. The inflammatory cells are generally concentrated within the villi but may also extend into the surrounding intervillous space, a feature commonly associated with intervillous fibrin deposition and agglutination of contiguous inflamed villi (Fig. 53). Occasionally

Chronic Villitis Etiology. Chronic villitis, or chronic inflammation of the villous parenchyma, is of two different etiologies. The vast majority of villitides are villitis of unknown etiology or VUE, which is thought to be an immune-mediated response of the mother to fetal antigens in the placenta (host vs. graft reaction), a view supported by recent investigations identifying the inflammatory cells as CD8-positive maternal T lymphocytes (Myerson and Parkin 2006; Redline 2007b). Very rarely, villitis may result from a hematogenous infection in which infectious agents, usually viruses but also some bacteria and protozoa, reach the placenta through the maternal blood (hematogenous infection). In contrast to chorioamnionitis, a local infection, placental involvement in hematogenous infection is just one manifestation of maternal systemic disease. Although catastrophic fetoplacental infections caused by hematogenously acquired agents have been well documented, they are very infrequent. Pathology. Villitis is almost always discovered incidentally on microscopic examination. Typically, there is neither clinical suspicion of nor gross pathologic clue to the underlying inflammatory process. There may be nonspecific findings – the placenta may be small or large and edematous – but only rarely are inflammatory foci identified grossly. The essential feature common to all

Fig. 52 VUE. Villi infiltrated by lymphocytes and histiocytes

Fig. 53 VUE. Inflammation involving the perivillous space with agglutination of adjacent inflamed villi

19

Diseases of the Placenta

an intervillous component (intervillositis) may predominate. There are usually scattered foci of inflamed villi randomly distributed, but in some cases, they may be concentrated in the basal regions of the placenta. The degree of villous involvement varies greatly from case to case and may be graded semiquantitatively as outlined by the Amsterdam group (Khong et al. 2016) as low grade and high grade. Low grade is defined as inflammation affected fewer than ten contiguous villi in any one focus, with more than one focus required for diagnosis. High grade is defined as multiple more with at least one showing involvements or more than ten contiguous villi (Khong et al. 2016). They can be further characterized by distribution with low grade being either focal or multifocal and high grade being further classified as patchy or diffuse.

VUE More than 95% of villitides are of unknown etiology. Although the extent and severity of VUE vary, it is low grade in the majority of cases (60–75%). The villous inflammatory infiltrate is typically composed of macrophages and lymphocytes, but giant cells may also be present and are not indicative of infection. An active component with perivillous neutrophils may be present and when fulminant may be accompanied by intervillositis and fibrin deposition with villous necrosis (Fig. 54). Since this pattern may be seen in some infectious villitides (gram-negative bacteria, nonsyphilitic spirochetes), special stains (gram, silver stains) may be considered if there is clinical suspicion of infection. The inflammatory infiltrate is often confined to terminal villi but may also involve stem villi. Inflammatory obliteration of stem villous vessels with downstream avascular villi (VUE with fetal obliterative vasculopathy) has been associated with neurologic impairment (Fig. 55) (Kraus et al. 2004). Chronic deciduitis and chronic chorioamnionitis may accompany VUE in some cases. Villitis involving basal/periseptal villi (basal villitis) is frequently present and represents a subgroup of VUE often associated with plasma cell deciduitis. Clinical significance. VUE is a common lesion involving 5–15% of third-trimester placentas

1255

Fig. 54 Chronic villitis with intervillositis and villous necrosis. The villous stroma and intervillous space are infiltrated by neutrophils and chronic inflammatory cells. Fetal stem villous vasculitis, villous necrosis, and diffuse perivillous fibrin deposition result in grossly visible necrotic foci. Obliteration of stem vessels results in distal villous changes, here FAV

Fig. 55 VUE with obliterative fetal vasculopathy. Vasculitis and obliteration of stem vessels result in avascular villi

(Redline 2007b). When mild, VUE is usually clinically silent and the infant unaffected. More severe, extensive VUE is associated with IUGR, IUFD, CP, and other forms of neurologic impairment (Redline and O’Riordan 2000). These complications correlate directly with the severity of the villitis. Whether this contributes to perinatal morbidity or mortality is unclear. When fatal, VUE is generally massive, suggesting that impaired placental function may be a mechanism of injury. VUE may recur (10–25% when

1256

R. N. Baergen et al.

diffuse) and is associated with total reproductive failure. IUGR, IUFD, and preterm delivery are especially likely in cases of recurrent VUE.

Infectious Villitides Rarely the morphologic features of a villitis suggest an infectious cause, although differences between the infectious villitides and VUE may be quite subtle. In general, infectious villitides tend to be more severe and diffuse with more villous necrosis and may be associated chronic chorioamnionitis. Cytomegalovirus (CMV) Pathology. The placenta may be normal, small (in cases of IUGR), or large and edematous. Histologically the villi may exhibit any or all of a wide spectrum of changes related in part to gestational age. A plasmacytic villous infiltrate and stromal hemosiderin, often deposited around the remnants of occluded vessels, are characteristic (Fig. 56). Foci of stromal necrosis, calcification, and avascular villi may be found. The diagnostic cytopathic viral changes, large eosinophilic intranuclear and smaller basophilic cytoplasmic inclusions, may be present in endothelial cells, Hofbauer cells, or trophoblast (Fig. 57). They are most numerous and easily found in early, severe infections; as gestation proceeds, viral inclusions are typically scarce. When inclusions are not visualized on conventional H&E stains, CMV infection may be confirmed by immunohistochemistry, polymerase chain reaction (PCR), or in situ hybridization. Clinical significance. CMV is the commonest identified infectious cause of villitis. The prevalence of congenital infection ranges from 0.2% to 2.2% of all live births. Transmission to the fetus may occur in utero, at birth, or postnatally. Intrauterine infection may result from either primary or recurrent maternal infection. Congenital infections resulting from reactivation of latent virus are less likely to produce fetal damage and late sequelae than those resulting from primary maternal infection. At the present time, CMV is the most commonly recognized infectious cause of developmental impairments. Late complications including mental retardation, chorioretinitis,

Fig. 56 CMV placentitis. Plasmacytic villous infiltrate

Fig. 57 CMV inclusions

seizures, and especially neurosensory hearing loss are most common among survivors of symptomatic congenital infection but may also occur later in children with no manifestations at birth. Treponema pallidum Pathology. Infected placentas tend to be large and bulky. Histologically, the villi are large and relatively immature but not markedly edematous. The villous stroma is cellular as a result of the prominence of Hofbauer cells, and a lymphoplasmacytic infiltrate with subtrophoblastic neutrophils and microabscesses may be seen focally. Rarely, the inflammatory reaction has granulomatous features. Subendothelial and perivascular fibrosis resulting in luminal narrowing, recanalization, or occlusion of villous vessels is particularly characteristic. Marked decidual plasmacytic vasculitis

19

Diseases of the Placenta

1257

and a chronic inflammatory infiltrate in the fetal membranes and umbilical cord have been noted in some cases. Necrotizing funisitis is a frequent finding but is not universal in, or specific for, congenital syphilis. The histologic changes in the placenta and umbilical cord are characteristic but not diagnostic. Definitive diagnosis depends on identification of spirochetes, most easily demonstrated in the umbilical cord whether inflamed or not. Listeria monocytogenes Pathology. The placenta is usually grossly normal but may contain minute yellow-white necrotic foci or rarely larger abscesses or infarcts (Fig. 58). Unlike most other villitides, neutrophils predominate in Listeria villitis. Acute villitis is so characteristic of Listeria infection that its presence requires that infection be ruled out, preferably with a tissue Gram stain. Rarely, acute villitis and intervillous abscesses may be caused by maternal sepsis. Villi are commonly enmeshed in intervillous acute inflammation and fibrin, which, when extensive, may result in villous necrosis and intervillous abscesses. Neutrophils in the villi are often localized to a rim, often between the trophoblast and villous stroma (Fig. 59). Listeria villitis may or may not be accompanied by chorioamnionitis or funisitis. Listeria monocytogenes is a small gram-positive motile bacillus with rounded ends that may be demonstrated with tissue Gram stain (e.g., Brown–Hopps), Warthin–Starry stain, or Dieterle stain. Immunohistochemical stains have also been utilized to establish the diagnosis and to exclude other infections. Clinical significance. L. monocytogenes is a significant cause of intrauterine infection, spontaneous abortion, prematurity, and neonatal sepsis, morbidity, and death. Perinatal listeriosis takes one of two forms. In the “early type,” congenital infection results in devastating neonatal sepsis (granulomatosis infantisepticum) in which microabscesses similar to those in the placenta are disseminated in fetal organs. The “late type” of perinatal listeriosis, presumably acquired during birth, presents as meningitis in the second or third week of life. L. monocytogenes usually does not cause serious disease in adults, although gravidas

Fig. 58 Listeria placentitis. Yellow-white necrotic areas representative of intervillous abscesses may be grossly apparent in some cases

Fig. 59 Listeria placentitis. Neutrophils aggregate in the intervillous space. Entrapped villi undergo necrosis forming abscesses. Neutrophils accumulate peripherally in the villous stroma beneath the trophoblast

may experience a flu-like syndrome and fever. The most common mode of infection is the ingestion of contaminated food, often a milk product. A brief hematogenous phase is followed by fecal shedding until immunity is established. Placental infection can result from hematogenous dissemination or via the ascending route. The particular predisposition for significant listerial infection in pregnant women and the fetus appears to be related to local factors at the maternofetal interface that compromise an effective immune response (Redline 1988). Toxoplasma gondii Pathology. The placenta infected with Toxoplasma gondii may be grossly normal or large and edematous. Microscopically, the changes are

1258

highly variable, ranging from a subtle low-grade lymphocytic villous infiltrate to a destructive process associated with necrosis. True granulomas with central necrosis, palisaded histiocytes, and Langerhans giant cells may also be present. Nodular accumulations of histiocytes beneath the trophoblast or extending into the intervillous space, decidual plasmacytic infiltrates and vasculitis, and chronic inflammatory infiltrates in the fetal membranes and umbilical cord have been described. However, the most common presentation is the presence of the encysted form of the organism in the fetal membranes, chorionic plate, umbilical cord, or villi (Fig. 60) without appreciable inflammation. They are often found just under the amniotic epithelium. Toxoplasma cysts are typically unassociated with inflammation, but once ruptured, the tachyzoites incite an intense inflammatory reaction and necrosis. Identification of tachyzoites, very difficult on H&E stains, is aided by immunohistochemistry, immunofluorescence, or PCR. Clinical significance. Congenital infection appears to result mainly from primary maternal infection acquired early in pregnancy, usually by the ingestion of oocytes in undercooked meat or by contact with cat feces. During the parasitemic stage of maternal infection, the organism is transmitted to the placenta and fetus. The risk of fetal infection in these circumstances is about 50%. The likelihood of fetal transmission increases with gestational age, although the

Fig. 60 Toxoplasma villitis. A toxoplasma cyst is present in this chronically inflamed villus

R. N. Baergen et al.

severity of infection is greatest when the infection is acquired in the first trimester. The clinical spectrum of fetal involvement ranges from severe damage to the CNS and eyes to completely asymptomatic infection, recognized only by the development of chorioretinitis after months or years of follow-up. Prenatal diagnosis (via fetal blood sampling or culture) and antibiotic treatment seem to reduce the frequency of congenital infection. In the presence of maternal antibodies from past infections, fetal lesions, with rare exception, do not occur. Parvovirus B19 Pathology. Pathologic changes in the placenta reflect fetal anemia. Parvovirus B19 preferentially infects actively replicating cells, especially erythroblasts, which are then destroyed. Grossly, the placentas are large, pale, and friable and may be hydropic particularly when the infection is associated with fetal hydrops. Microscopically, there is a uniform pattern of relative villous immaturity and edema. Diagnostic intranuclear eosinophilic inclusions with peripheral chromatin condensation are present in erythroblasts in the villous vessels (Fig. 61). In situ hybridization and immunohistochemistry are somewhat more sensitive than conventional microscopy in identifying infected cells, or PCR may be used to confirm the diagnosis. Unlike most other congenital hematogenous infections, there is no villous inflammation.

Fig. 61 Parvovirus infection. Erythrocytes in the villous capillaries show central nuclear eosinophilic inclusions with peripheral chromatin condensation

19

Diseases of the Placenta

Clinical significance. Human parvovirus B19 is the agent of “fifth disease,” or erythema infectiosum, a mild, acute exanthematous disease of children. In adults, most infections are asymptomatic, although self-limited polyarthropathy is common especially in women, and aplastic crises occur in individuals with chronic hemolytic anemias. To date, the most commonly recognized consequences of fetal parvovirus infection are nonimmune hydrops (which may resolve spontaneously) and abortion. Most abortions occur between the 10th and 28th weeks of pregnancy, but the risk of fetal loss is low, estimated to be less than 10%. Neonatal anemia has been observed in a few infants infected in the third trimester, and malformations reminiscent of ocular rubella embryopathy have been reported very rarely. Herpes Simplex Virus Disseminated herpes simplex virus infection is an important cause of devastating disease and death in the newborn. Intrapartum infection of the fetus is the most common, although ascending and transplacental dissemination has been described. Villous necrosis and agglutination, lymphocytic villitis, and fibrinoid necrosis of villous vessels have been documented in cases of hematogenous infection. Acute necrotizing and chronic lymphoplasmacytic chorioamnionitis, amniotic viral inclusions, and funisitis have been described in cases of ascending infection. An increased frequency of spontaneous abortion and congenital malformations has been reported in patients with primary infection in the first 20 weeks of pregnancy. Varicella Zoster Varicella infection in pregnancy is uncommon in the United States because the majority of women of childbearing age (95%) are immune. In congenital infection, the placenta may show small, grossly visible necrotic foci and villous necrosis, vascular occlusion, lymphoplasmacytic infiltrates, and granulomas with giant cells. Viral inclusions have been reported in villi and decidua. The spectrum of fetal manifestations is wide, ranging from completely asymptomatic babies to those with perinatal varicella/zoster or full-blown embryopathy, the latter occurring in less than 5% of fetuses infected in the

1259

first trimester. Reactivation of maternal zoster during pregnancy does not appear to be associated with severe fetal sequelae. Rubella Congenital rubella infection is now rare due to the effectiveness of immunization programs. The placental findings have been well documented during previous rubella epidemics, mainly in first- and second-trimester abortions but in a few term placentas as well. Human Immunodeficiency Virus (HIV) HIV may be transmitted to the fetus transplacentally, at the time of delivery, or after birth (through breastfeeding). In utero transmission can be confirmed by the detection of virus in infants by PCR or coculture within 48 h of birth. Several factors (maternal, fetal, obstetric, and virologic) affect maternal–infant viral transmission, and timing of transmission may determine the subsequent course of infection in the infant. Antiretroviral therapy given to the mother before and during delivery and to the infant after delivery has reduced vertical transmission substantially. No histopathologic lesions directly attributable to HIV have been described in the placenta. Specifically, there have been no reports of villitis. Placentas from seropositive mothers have demonstrated an increased incidence of ACA. Zika Virus Zika virus is a single-stranded RNA arbovirus in the Flaviviridae family, genus Flavivirus initially isolated in Africa. It is transmitted to humans by mosquitoes, and in the last few years, outbreaks in other parts of the world have been reported. Studies in animal models and human tissue culture and epidemiologic studies have shown that the virus crosses the placenta and can infect fetal tissues, preferentially the fetal brain (Platt and Miner 2017). A broad range of abnormalities in the fetus have been described including growth restriction, miscarriage or stillbirth, perinatal death, and microcephaly (Chibueze et al. 2017) with a reported incidence of the latter up to 2.3%. Although further study is needed to ascertain the full spectrum of placental

1260

R. N. Baergen et al.

pathologic findings, initial studies have shown that infection leads to enlarged, immature-appearing villi with increased Hofbauer cells but no significant inflammatory infiltrate (Kaplan et al. 2016; Rosenberg et al. 2017).

idiopathic, encountered most often in first-trimester spontaneous abortions but also seen in the second and third trimesters. It is associated with recurrent spontaneous abortion, IUGR, IUFD, and a high overall perinatality mortality rate.

Other Organisms Massive chronic histiocytic intervillositis may occur in malaria (Ordi et al. 1998). Recurrent villitis has been described in association with nonsyphilitic spirochetes (Abramowsky et al. 1991). The placental lesions associated with hematogenous spread of other organisms are detailed elsewhere (Fox and Sebire 2007).

Chronic Deciduitis Chronic deciduitis has been defined as either the presence of plasma cells or diffuse chronic inflammation (with or without plasma cells) in the decidua basalis. The chronic inflammatory response may be directed against maternal or fetal antigens or microorganisms. Chronic deciduitis is often associated with VUE at term and with ACA in preterm placentas, but it may also occur as an isolated finding. In the absence of associated findings, it is of limited clinical significance.

Other Patterns of Placental Inflammation Chronic Histiocytic Intervillositis Chronic histiocytic intervillositis is the diffuse uniform infiltration of the intervillous space by monocyte–macrophages accompanied by variable perivillous fibrin deposition (Fig. 62). Excluded from this category are cases with a predominance of villitis or a polymorphous intervillous inflammation. Placentas are often SGA. The histologic features are very similar to those in placental malaria, distinguished from it by the absence of malarial pigment. Chronic histiocytic intervillositis is

Fig. 62 Chronic histiocytic intervillositis. There is diffuse infiltration of the intervillous space by monocytemacrophages

Chronic Chorioamnionitis Rarely, an inflammatory infiltrate occurring in the same distribution as ACA is composed of chronic inflammatory cells (Fig. 63). Typically, small mature lymphocytes predominate, but plasma cells, histiocytes, and rarely large lymphoid cells and immunoblasts are admixed (Jacques and Qureshi 1998). In some cases, the chronic inflammatory cells may be accompanied by a minor component of neutrophils, which may be either entirely distinct from, or intimately admixed with, the chronic inflammatory component. The

Fig. 63 Chronic chorioamnionitis. Small lymphocytes infiltrate the amnion and chorion

19

Diseases of the Placenta

inflammatory infiltrate in chronic chorioamnionitis is commonly focal and typically mild. It is generally confined to the membranes, although involvement of the chorionic plate may occur. Rarely, the large fetal vessels of the chorionic plate and umbilical cord are also chronically inflamed. Chronic chorioamnionitis is often accompanied by chronic villitis and rarely by chronic or subacute necrotizing funisitis. A specific infectious etiology is not identified in the great majority of cases, although rarely chronic chorioamnionitis may occur in association with rubella, herpes simplex, T. pallidum, or T. gondii infection. Importantly, chronic chorioamnionitis is not related to acute or subacute chorioamnionitis.

Eosinophilic/T-Cell Vasculitis Eosinophilic/T-cell vasculitis is a chronic inflammatory infiltrate composed of fetal eosinophils and lymphocytes involving chorionic plate and large stem villous vessels (Fig. 64). Typically, the infiltrate is focal rather than the diffuse involvement of the chorionic plate vessels in acute chorioamnionitis. It often involves a single vessel. In addition, the infiltrate faces the placenta, or the maternal side of the vessel, in contrast to the fetal inflammatory response in ascending infection where the inflammatory cells migrate toward the amniotic cavity. It has been associated with thrombosis in the fetal circulation and chronic villitis,

1261

but there are no specific clinical associations (Fraser and Wright 2002; Jacques et al. 2011).

Maternal and Fetal Vascular Malperfusion (FVM) The placenta is a vascular organ with separate fetal and maternal circulations. The integrity of both circulations is essential to placental function. Clots, hematomas, and other pathologic processes affecting the vessels and spaces in and around the placenta may cause both fetal and maternal injury depending on their size, location, and extent. Measurement of size and evaluation of extent are important in assessing the significance of individual lesions. There is, however, no absolute “cutoff” beyond which a bad outcome is inevitable or below which a good outcome is assured. The normal placenta has considerable functional reserve which has been variably estimated to be between 10% and 20% of the placental volume. The potential for injury depends not only on the nature and extent of the pathology but also on the amount and functional status of the uninvolved parenchyma. Thus, even a relatively small lesion in a compromised or small placenta may have greater clinical significance than a larger lesion in an otherwise normal placenta. A balance between factors that favor coagulation and fibrinolysis is necessary for homeostasis; pregnancy itself shifts that balance toward thrombosis. Within the maternal intervillous space, clotting inhibitors such as annexin V at the syncytiotrophoblastic surface normally inhibit clotting. Thrombogenic factors may overcome this, resulting in thrombi and infarcts (Rand 2000; Rayne and Kraus 1993). Within the fetal circulation, stasis, local injury from inflammation or compression, or thrombophilic factors may result in clotting.

Maternal Circulation Fig. 64 Eosinophilic T-cell vasculitis. Eosinophils and small lymphocytes invade the muscular wall of this chorionic plate vessel

In relation to placental pathophysiology, the maternal circulation includes the uterine arteries and their branches, the spiral arterioles (including

1262

R. N. Baergen et al.

Decidual Arteriopathy (Vasculopathy), Lack of Physiologic Conversion, Acute Atherosis, and Maternal Vascular Malperfusion Etiology. When normal physiologic trophoblastic remodeling of the spiral arterioles does not evolve properly, arterial smooth muscle persists, the lumina fail to expand, and uteroplacental blood flow is reduced. The maternal blood pressure rises, the maternal vascular endothelium is injured, and the maternal syndrome of preeclampsia develops. Pathology. Non-remodeled spiral arterioles with persistent musculoelastic media (lack of physiologic conversion) may be identified in the basal plate (Fig. 65). Non-remodeled spiral arteries are predisposed to develop a distinctive lesion, acute atherosis, characterized by necrosis and dense eosinophilia of the vessel wall with large foamy macrophages and inflammation (Fig. 66). The lumens may be partly or completely occluded by thrombus. Acute atherosis involves muscularized maternal arteries in the basal plate and membranous decidua. The latter is often the best place to find acute atherosis (Fig. 67). In

addition, fibrinoid necrosis, vasculitis, and thrombosis may also occur, and all these lesions are encompassed by the term decidual vasculopathy or decidual arteriopathy (Khong et al. 2016). Immature IT in the superficial basal plate and trophoblastic giant cells in the deep basal plate are increased (Redline and Patterson 1995). These changes together with the characteristic vascular alterations have been referred to as superficial implantation (Kraus et al. 2004). Mural hypertrophy of arterioles is a less common pattern of decidual vasculopathy that occurs more commonly in women whose preeclampsia is complicated by diabetes mellitus (Barth et al. 1996). The muscular walls of the affected decidual vessels are thickened, and there is marked luminal narrowing (Fig. 67). Mural hypertrophy is diagnosed when the mean wall diameter is greater than 30% of the overall vessel diameter (Redline et al. 2004b). Chronic vasculitis and particularly perivasculitis can also occur as well as persistence of endovascular trophoblast in the third trimester (Khong et al. 2016). The most direct adverse effect of decidual vasculopathy is the reduction of maternal blood flow into the intervillous space (maternal vascular malperfusion) leading to reduced placental growth and weight (Khong et al. 2016). Microscopic changes include increased and prominent syncytiotrophoblastic knots (Fig. 68). At term, approximately 30% of the terminal villi should show

Fig. 65 Incomplete vascular remodeling. This basal plate arteriole retains its muscular wall

Fig. 66 Acute atherosis. The spiral arterioles show fibrinoid necrosis and accumulation of foamy macrophages. The largest artery is thrombosed, and there is infarction of the overlying placenta

the converted arterioles – the uteroplacental vessels), the maternal intervillous space, and the venous drainage of the uterus. Many pathologic processes affecting the maternal circulation can be traced to abnormal implantation and vascular remodeling.

19

Diseases of the Placenta

1263

Fig. 67 Fibrinoid necrosis (left) and mural hypertrophy of membrane arterioles (right). In mural hypertrophy, the vessel wall is greater than one third of the total vessel diameter

Fig. 68 Maternal malperfusion. Increased syncytial knots. Syncytiotrophoblastic nuclei clustered in the terminal (distal) villi

Fig. 69 Maternal malperfusion. Villous agglutination. Clusters of degenerating terminal (distal) villi are adherent to one another

syncytial knots, with lesser numbers earlier in gestation (Loukeris and Baergen 2010). In preeclampsia, the nuclei of the pathologic knots undergo ischemic necrosis (“aponecrosis”) (Huppertz et al. 2006), releasing membrane fragments, DNA, and proteins into the maternal circulation. This circulating cellular debris may contribute to maternal endothelial injury in preeclampsia (Huppertz et al. 2006). The villi are also smaller than expected for the gestational age, referred to as accelerated villous maturation. Adhesions between syncytial knots undergoing aponecrosis may result in the agglutination of small villous clusters that commonly show karyorrhexis (villous agglutination) (Fig. 69). With more severe underperfusion, intervillous fibrin is increased, first around the stem villi and then around distal villi. Fibrin accumulation in the intervillous space may be the result of decreased maternal perfusion and stasis or abnormal coagulation or both. When reduced maternal blood flow is severe and long-standing, the villi develop abnormally. The distal villi are decreased in number and slender with reduced branching, and terminal villi are very small with increased syncytial knots. This constellation of features is referred to as distal villous hypoplasia (synonym: terminal villous deficiency) (Fig. 70) (Benirschke et al. 2006; Khong et al. 2016). To be significant, these changes must be demonstrated centrally in the mid and basal zones of the placenta because similar changes occur

1264

R. N. Baergen et al.

syndrome (Salafia and Cowchock 1997), thrombophilias, and occasionally in the absence of any identifiable underlying maternal disease. Similar changes in the spiral arterioles with constrictions, dilations, hyalinization, and thrombi occur in the placenta when thrombotic thrombocytopenic purpura develops in pregnancy (Jamshed et al. 2007).

Fig. 70 Maternal malperfusion. Distal villous hypoplasia. Chronic severe maternal malperfusion results in distal villi that are thin and non-branched when cut longitudinally and very small when cut in cross section. Syncytial knots are prominent

Fig. 71 Maternal malperfusion. Infarcts of varying ages. Recent infarct is red

normally in the less perfused peripheral and subchorionic regions (Wyatt et al. 2005). Hypoxiarelated gene expression in trophoblast occurs in these areas even in the normal placenta (Wyatt et al. 2005). Localized areas of severe ischemia result in infarcts that may be large and numerous in cases of severe preeclampsia (Fig. 71). Longstanding maternal vascular malperfusion may cause fetal volume depletion, decreased extracellular fluid, and an abnormally thin cord. Clinical significance. Maternal vascular compromise with placental malperfusion is one of the major causes of IUGR. Severe maternal vascular malperfusion may result in IUFD and is associated with CP in VLBW infants (Redline et al. 2007). Identical pathologic changes occur in the placentas of patients with systemic lupus erythematosus (Magid et al. 1998), scleroderma (Doss et al. 1998), the antiphospholipid antibody

Infarct Etiology. An infarct in the placenta, as in any other organ, is an area of ischemic necrosis resulting from obstruction of blood supply. The spiral arterioles that supply the placenta may be narrowed by acute atherosis, occluded by thrombi, or disrupted by a retroplacental hematoma. Extensive placental infarcts have been reported in association with thrombophilic states (Arias et al. 1998; Dizon-Townson et al. 1997; Khong 1999). Pathology. On macroscopic examination, infarcts are wedge-shaped areas of induration often located at the placental margin. Fresh infarcts are difficult to see; they differ little in color but are firmer and drier than the surrounding placenta. Older infarcts grow progressively firmer and change from red to brown and then tan, yellow, or white (Fig. 71). Microscopically, the earliest change is collapse of the intervillous space. The villi are crowded, separated by only a thin layer of fibrin, and trophoblastic nuclei cluster together forming knots. The ST, vascular endothelium, and villous stroma undergo progressive degeneration with loss of nuclear staining and smudging until eventually only crowded, ghostlike villous outlines remain (Fig. 72). The mummified infarcted villi are not removed by macrophages or replaced by fibrous tissue as occurs in other organs. A mild acute inflammatory response may occur at the margin of an infarct. Clinical behavior. Infarcts occur in about 25% of otherwise normal-term placentas. The finding of a small peripheral infarct in an otherwise normal placenta is of no clinical significance. Multiple or large (>3 cm) infarcts, central infarcts, or infarcts in the first or second trimester are indicative of significant, underlying maternal vascular disease, most often preeclampsia. Extensive

19

Diseases of the Placenta

1265

Fig. 72 Maternal malperfusion. Infarcts. The intervillous space is collapsed with necrotic villi undergoing progressive necrosis eventuating in ghostlike villous remnants

placental infarction is associated with fetal hypoxia, IUGR, IUFD, and neurologic injury. These ill effects on the fetus are not simply the result of the destruction of villous tissue but reflect the superimposition of infarction on a placenta already compromised by low maternal blood flow.

Retroplacental Hematoma Definition. A retroplacental hematoma is a clot located in the decidua between the placental floor and the myometrium. It represents the pathologic lesion indicative of premature separation of the placenta from the uterus and corresponds to the clinical diagnosis of placental abruption that is made by the obstetrician These two terms should not be used interchangeably, a practice which leads to confusion between a clinical diagnosis with grave implications and a pathologically demonstrated clot whose impact depends primarily on size. Frequency. Retroplacental hematomas are found in approximately 4.5% of placentas examined pathologically (Fox and Sebire 2007). Based on data from the National Hospital Discharge Summary, symptomatic abruption occurs in 1% of singleton deliveries. The incidence appears to be increasing (Ananth et al. 2005). Etiology. The pathogenesis of retroplacental hematoma is likely due to bleeding from a decidual artery followed by dissection of the enlarging clot. An occluded ischemic artery that undergoes reperfusion may rupture, leading to a retroplacental hematoma. Retroplacental hematomas are increased threefold in women with

Fig. 73 Retroplacental hematoma. The placenta is compressed and infarcted over this large retroplacental hematoma

preeclampsia, presumably due to the related vascular pathology. Other associations include PTL, chorioamnionitis, anemia, smoking, cocaine abuse, trauma, diabetes mellitus, and short umbilical cord. Traumatic separation of the attached placenta followed by bleeding from a disrupted vessel probably occurs, for instance, following an automobile accident. Pathology. When confined by peripherally attached placenta, a retroplacental hematoma distorts and indents the overlying placental parenchyma (Fig. 73). Separated from its blood supply, the overlying placenta infarcts. The characteristic depression and overlying infarct are easily recognized even when the clot itself has become detached during delivery. Larger clots may dissect into the basal region of the placenta. Older retroplacental hematomas may be much more subtle, forming only a thin, inconspicuous layer of red-brown clot beneath an infarct. When

1266

retroplacental hematomas extend to the placental margin, the blood may be evacuated without causing any indentation. Very recent extensive placental separation is typically associated with little, if any, gross or histologic change in the placenta. These large hematomas generally result in immediate fetal distress, necessitating emergent delivery. A large fresh clot behind a “floating” detached placenta observed at the time of cesarean section may be the only objective sign of an acute retroplacental hemorrhage. Obstetricians should document this observation and submit the clot along with the placenta. Microscopic examination of the clot may offer supportive evidence when clumps of decidua associated with strands of fibrin are demonstrated. Retroplacental hematomas consist of stratified red cells and fibrin, the proportion of fibrin increasing as the lesion ages and the red cells degenerate. The time course of these evolutionary changes is unknown. The adjacent decidua may be necrotic. Macrophages containing hemosiderin are often present around older clots and sometimes involve the membranes. Organization may occur where the clot interfaces with the uterus, but not where it interfaces with the placenta. The placenta overlying the hematoma is often infarcted. Villous edema and villous stromal hemorrhage are secondary placental abnormalities indicating that the retroplacental hematoma has adversely affected the fetus (Fig. 74). Intraplacental hematomas associated

R. N. Baergen et al.

with malperfusion also may occur by a similar mechanism. Clinical significance. The clinical significance of a retroplacental hematoma is related primarily to its size. Retroplacental clots block the passage of blood from the spiral arteriole into the intervillous space. Small clots may have little effect, since adjacent spiral arterioles may suffice to keep the overlying placental villi functional. The larger the clot and infarct, the more likely it will exceed the functional reserve capacity of the placenta, which may be already marginal in a background of chronic uteroplacental ischemia. Older chronic lesions associated with hemosiderosis in the membranes indicate the possibility of chronic reduction of placental function. Extensive accumulations of blood lead to the full clinical picture of abruption – pain and shock in the mother and severe acute hypoxia in the fetus.

Marginal Hematoma Definition. A marginal hematoma occurs where the lateral margin of the placental disk joins the fetal membranes. Etiology. Acute marginal hematomas are thought to result from the rupture of uteroplacental veins at the margin of a low-lying placenta. Chronic recurrent venous hemorrhage at the disk margin (chronic marginal hematoma) elevates and centrally displaces the membrane insertion site, resulting in circumvallation.

Fig. 74 Villous stromal hemorrhage (left). Villous stromal hemorrhage is associated with acute retroplacental hematoma

19

Diseases of the Placenta

Pathology. Grossly, an acute marginal hematoma forms a crescent-shaped clot at the lateral margin of the placenta. On cut section, the clot is triangular with the apex at the junction of the membranous and villous chorion. Microscopically, a recent marginal clot usually lies entirely outside the placental disk but may occasionally involve the intervillous space. With this exception, the presence of an acute marginal hematoma has no effect on the adjacent villi. Chronic marginal hematomas are tan-brown clots usually occurring in the region of circumvallate membrane insertion and associated with membrane hemosiderin deposition (Fig. 75). Clinical significance. An acute marginal hematoma may be associated with antepartum bleeding, in some cases followed by the onset of labor, but it does not have any untoward effect on the fetus. Chronic marginal hematomas and circumvallation are often associated with antenatal bleeding. In this context, retroplacental hematomas and abruption are also more frequent.

Intervillous Thrombus Definition, frequency, and etiology. Intervillous thrombi are clots in the intervillous space. They are common, found in 36–48% of placentas. Intervillous thrombi are thought, at least in some cases, to be initiated by fetal bleeding into the intervillous space through ruptured vasculosyncytial membranes. Small amounts of fetal blood can be demonstrated, although most of the clot is composed of maternal blood (Kaplan et al. 1982). Factors cited as potential causes of villous damage resulting in fetomaternal hemorrhage include trauma, amniocentesis, and external version. Basilar intervillous hematomas may have a different etiology. Pathology. Intervillous thrombi are usually angular lesions that may occur anywhere in the intervillous space but are most common midway between the chorionic and basal plates. They begin as red, fluid, or semifluid blood and become progressively laminated and depigmented with age. They may be single, although multiple lesions are common. Most are 1–3 cm in diameter. Microscopically, the thrombi consist of layered red cells and fibrin, the proportion of fibrin

1267

Fig. 75 Chronic marginal hematoma. Old brown clot is prominent at the margin of this circumvallate placenta

Fig. 76 Intervillous thrombus. A typical lesion composed of laminated fibrin and red cells

increasing as the lesion ages (Fig. 76). Villi displaced to the margins of the clot may be infarcted and/or avascular. Clinical significance. An intervillous thrombosis is significant in that it marks a site of hemorrhage from the fetal into the maternal circulation. Intervillous hematomas occur at the bleeding site. These are common and usually do not indicate that a significant fetal blood loss has occurred. Occasionally, a larger fetomaternal hemorrhage results in fetal anemia and fetoplacental hydrops if it occurs slowly and chronically, or sudden unexpected death if severe and acute. Severe fetal morbidity and mortality result from

1268

hemorrhages of 150 mL or more, but lesser amounts can be significant if chronic bleeding has occurred episodically. It has been suggested that the risk is greater when the mother and fetus have compatible ABO blood types, perhaps because of deficient clot formation at the site of bleeding (Ziska et al. 2008). There is often no clinical evidence of fetal distress or injury, even in cases resulting in IUFD. The amount of fetomaternal hemorrhage can be quantitated using the Kleihauer–Betke test. Flow cytometry may be more sensitive (Davis et al. 1998; Fernandes et al. 2007). Correlation between the severity of fetomaternal hemorrhage and size and number of intervillous thrombi is variable. In many cases of severe fetomaternal hemorrhage, there may be no thrombus, and a site of bleeding may not be identified. Small amounts of fetal blood leak into the maternal circulation in virtually every pregnancy. Fetomaternal hemorrhage may lead to isoimmunization of Rh-negative mothers resulting in hemolytic anemia and immune hydrops in the fetus.

Massive Subchorial Hematoma (Breus’ Mole) The massive subchorial hematoma has been defined as coagulated blood, at least 1 cm in thickness, separating the chorionic plate from the underlying placenta over much of its area (Fig. 77).

Fig. 77 Massive subchorial hematoma. Thick fresh clot separates the chorionic plate from the underlying placental parenchyma

R. N. Baergen et al.

These are generally relatively fresh, red hematomas that distort the chorionic plate and protrude as nodular masses into the amniotic cavity. They may dissect into the chorionic plate or extend into the intervillous space, sometimes as far as the basal plate. This is a rare lesion, the incidence estimated to be 0.53 per 1,000 deliveries in one large study (Shanklin and Scott 1975). The etiology and pathogenesis are unknown. Most authors agree that the hematomas are maternal. Massive subchorial hematomas are usually acute because catastrophic reduction in placental function results in abortion or IUFD. Rarely babies are live-born.

Maternal Floor Infarct (MFI): Massive Perivillous Fibrin Deposition (MPVFD) Definition. Although some authors separate these two lesions characterized by perivillous fibrin deposition, they are most likely variations of the same process. The original name of maternal floor infarction is a misnomer as they are not truly infarcts and often involve the placenta diffusely, not just the maternal floor. When classified by criteria proposed by Katzman and Genest, 44% of lesions did not conform to any of the categories (Katzman and Genest 2002). Thus, massive perivillous fibrin deposition is the current and preferred term. Etiology. Currently two types of fibrinoid or fibrin-like material are identified in the placenta (Frank et al. 1994; Vernof et al. 1992). Matrixtype fibrinoid is composed of oncofetal fibronectin, collagen IV, laminin, and tenascin and contains little or no fibrin. It is consistently associated with, and apparently produced by, extravillous trophoblast. This type of fibrinoid and associated trophoblast surrounds and encases the villi in MPVF. Fibrin-type fibrinoid has immunohistochemical features of blood clot product and lacks the cellular trophoblast component. This form of fibrin accumulates in maternal vascular underperfusion associated with decidual vasculopathy and the syndrome of preeclampsia. The fact that the intervillous space is involved would seem to implicate a maternal circulatory abnormality. Isolated reports have identified a group of associations including antiphospholipid antibody syndrome (Sebire et al. 2002),

19

Diseases of the Placenta

1269

Fig. 78 MPVFD. Here, the maternal surface is diffusely discolored and firm (left). The basal villi are entrapped in fibrinoid (right)

polymyositis (Al-Adnani et al. 2008; Hung et al. 2005), long-chain L-3-hydroxyacyl-coA dehydrogenase deficiency (Matern et al. 2001; Rakheja et al. 2002; Griffin et al. 2012), chronic intervillositis (Weber et al. 2006), and thrombophilias (Arias et al. 1998; Katz et al. 2002). Two separate reports, however, document involvement of only one of dichorionic twins, suggesting a role for the fetus (Gupta et al. 2004; Redline et al. 2003). Whether the fibrinoid material that encases the terminal villi in these disorders reflects coagulation secondary to stasis of maternal blood in the intervillous space or aberrant secretion of matrix by villous trophoblast is unknown. Pathology. On macroscopic examination, wide strands and interlacing clumps of firm, pale tan or gray material accumulate as a thickened rind in the basal portion and ramify throughout the placenta in a net-like pattern (Fig. 78). The deposition of fibrinoid may sometimes be concentrated on the basal plate but is more commonly diffuse (Fig. 79). Microscopically, the villi are widely separated and encased by pink amorphous fibrinoid containing variable numbers of mononuclear extravillous trophoblast (Fig. 80). The syncytiotrophoblast and capillary endothelium disappear from entrapped villi, but the villous stroma and villous outlines persist. Clinical significance. The reported incidence ranges from 0.5 to 5 per 1,000 deliveries (Redline et al. 2003). First-trimester spontaneous abortions, sometimes recurrent, and mid- and late-trimester IUFD are common when the process is massive.

Fig. 79 Massive perivillous fibrin deposits. Firm, white strands of fibrin are distributed throughout the placenta including the maternal surface

Fig. 80 MPFVD Villi are separated by and enmeshed in dense perivillous fibrinoid containing IT

Survivors are often delivered preterm with IUGR or develop long-term neurologic impairment (Adams-Chapman et al. 2002). MPVFD has a high recurrence rate in subsequent pregnancies. Villi entrapped in fibrinoid are isolated and nonfunctional and eventually are infarcted. Redline

1270

and Patterson observed that perivillous fibrin deposits that entrapped more than 20% of villi in the central basal portion of the placenta (thought to be the primary region of gas and nutrient exchange) were significantly associated with an IUGR and low placental weight (Redline and Patterson 1994). Milder cases of perivillous fibrin deposition enveloping 5–20% of terminal villi showed a lesser degree of similar complications.

Fetal Circulation and FVM The fetal circulation of the placenta begins with blood propelled by the fetal heart from the branches of the fetal iliac arteries to the paired umbilical arteries. The umbilical arteries pass branch on the chorionic plate and progressively rebranch in smaller stem villi ultimately reaching the villous capillaries where oxygen and nutrient exchange occurs across vasculosyncytial membranes. The fortified blood flows through progressively larger veins in the stem villi and chorionic plate, reaching the umbilical vein, which passes back through the umbilical cord to drain into the sinus venosus of the fetus.

FVM: Fetal Vascular Thrombosis Definition. Clots that form in the fetal circulation obstruct blood flow to the villi, rendering them nonfunctional for the transfer of oxygen and nutrients from the mother to the fetus. This is referred to as fetal vascular malperfusion by the Amsterdam group (Khong et al. 2016). Fetal thrombotic vasculopathy is a term often used to reflect the changes in stem vessels and villi that follow occlusion and cessation of blood flow. These include villous stromal–vascular karyorrhexis, avascular villi (Salafia 1997; Salafia et al. 1997), and hemorrhagic endovasculopathy (previously hemorrhagic endovasculitis) (Sander 1980; Sander et al. 2002). Related lesions or terms with similar connotations include fetal stem artery thrombosis (Fox 1966), fibrinous vasculosis (Scott 1983), and intimal fibrin cushion (DeSa 1973). Etiology. The rules of Virchow’s triad apply to thromboses in the placenta. Stasis, vascular

R. N. Baergen et al.

injury, and a hypercoagulable state, especially in combination, may result in clots in the fetal circulation. The most common cause of stasis is chronic partial or recurrent intermittent umbilical cord compression resulting from cord entanglements or anatomic factors (excessive length, hypercoiling, etc.). Stasis may also result from compression of velamentous vessels or in the meandering vascular spaces of chorangiomas or mesenchymal dysplasia. Vascular injury may be due to a fetal inflammatory response to infection, exposure to cytokines or meconium, or vascular inflammation in VUE (VUE with obliterative vasculopathy). Hypercoagulable states occur in inherited (factor V Leiden, protein S or C deficiency, etc.) and acquired thrombophilias (antiphospholipidemias, lupus erythematosus). The occurrence of a thrombophilic state in the mother or newborn does not by itself predict vascular lesions in the placenta or fetus, but it may represent a risk factor in association with other conditions (Ariel et al. 2004). Pathology. Thrombi in large vessels of the chorionic plate or umbilical cord are often visible on gross inspection (Fig. 81). Villi in the distribution of a large thrombosed vessel form a pale triangular area often with the same consistency as the surrounding parenchyma (Fig. 82). These are representative of foci of avascular villi. Older lesions may be firmer, gray-white, and better delimited. Both thrombi and villous lesions are

Fig. 81 Chorionic vessel thrombus. This large chorionic plate vessel is occluded by a hard white thrombus

19

Diseases of the Placenta

Fig. 82 Fetal vascular obstruction. When a large fetal vessel is occluded, the downstream villous changes are visible as a pale wedge-shaped area of avascular villi. This is one of a few findings that are more easily seen in the fixed placenta

Fig. 83 Fetal vascular obstruction. Red cell extravasation and septation in fetal stem vessels

Fig. 84 Fetal vascular obstruction, villous stromal–vascular karyorrhexis. Fragmented red cells are extravasated, and there is nuclear debris in and around terminal villous capillaries. These represent early changes in villi downstream from an occluded vessel

1271

Fig. 85 Fetal vascular obstruction, avascular villi. Avascular villi reflect fetal vessel occlusion, contrasting with the functional villi in this placenta from a live-born baby

subtle, sometimes better estimated after formalin fixation. In the majority of cases, the occluded vessels are small, associated with small clusters of distal avascular villi that are identifiable only microscopically. The evolution of large-vessel thrombi is similar to that in other sites except organization does not occur. Recent thrombi expand the vessel, are attached, and may have a layered appearance. The endothelium disappears, and spindle-shaped cells invade the thrombus producing a multilocular pattern (septation) (Fig. 83). The erythrocytes fragment and blend into adjacent connective tissue. These “hemorrhagic” lesions are what have been called hemorrhagic endovasculopathy, or hemorrhagic endovasculitis as originally described by Sander (1980). Smooth muscle may disappear, and the wall may calcify, indicating chronicity. The villi distal to an occluded vessel show a sequence of distinctive alterations. Early changes include karyorrhexis of intravascular, endothelial, and villous stromal cells with destruction of capillaries and extravasation of red cells (Fig. 84). The stroma may be mineralized. Eventually, the villi become hyalinized with bland, dense, acellular stroma ( foci of avascular villi – FAV) (Fig. 85). The surrounding ST is preserved and often shows increased knotting. The thrombosed vessel resulting in these distal villous

1272

alterations may or may not be apparent depending on the plane of sectioning. The same sequence of changes occurs diffusely throughout the entire placenta after IUFD. The diffuse versus focal nature helps distinguish involutional from pathologic vascular alterations. Clinical significance. FVM is associated with neonatal encephalopathy (NE) (McDonald et al. 2004), CP (Redline 2005), IUFD, and fetal and neonatal thromboocclusive disease (Kraus and Acheen 1999; Redline and Pappin 1995). FVM represents a major placental pathologic finding in IUGR and discordant twin growth (Redline et al. 2001). The negative effects of FVM occur in at least two ways. First, the presence of thrombi in fetal circulation of the placenta indicates the potential for thrombi to occur elsewhere either as a result of generalized activation of the fetal coagulation system or direct embolism. Thrombi or emboli and infarcts in the fetal brain, kidney, lung, and liver causing severe perinatal liver disease have been reported (Burke and Tannenberg 1995; Dahms et al. 2002; Kraus and Acheen 1999). While there is definite connection when clots in the placenta and fetus occur together, the finding of FVM in the placenta alone is not predictive. The prevalence of systemic thrombi or infarcts among newborns with placental FVM is low (Leistra-Leistra et al. 2004). A second form of injury results from reduction of placental reserve caused by loss of a significant fraction of the placental circulatory bed. The occurrence of multiple lesions in the same placenta compounds the prospects for injury (Redline et al. 2004a; Redline and O’Riordan 2000; Viscardi and Sun 2001).

Intimal Fibrin Cushions DeSa described intimal fibrin cushions composed of intramural fibrin deposits and proliferating fibroblasts forming nonocclusive intraluminal protrusions, some calcified, in chorionic veins (Fig. 86) (DeSa 1973). The finding was attributed to local injury from elevated venous pressure. Clinical associations included low birth weight, abruption, maternal hypertension, and intrapartum hypoxia.

R. N. Baergen et al.

Fig. 86 Intimal fibrin cushion. Fibrin is deposited in the wall of large chorionic plate vessel. Mural calcification is consistent with a long-standing process

Emphasizing associated wall edema, Scott et al. described a similar lesion, fibrinous vasculosis, associated with stillbirth and other adverse outcomes (Scott 1983). Currently these large-vessel lesions are classified as intimal fibrin cushions and are thought to be pressurerelated changes affecting vessels between the actual site of fetal vascular obstruction and the terminal villi (Redline et al. 2004a). They are often associated with other evidence of fetal vascular occlusion in distal villi including avascular villi and villous stromal–vascular karyorrhexis and are the most common thrombotic lesions seen in the placenta. Calcification in the cushion indicates long-standing obstruction and thrombosis.

Fetal Vascular Narrowing and Increased Umbilical Vascular Resistance Doppler velocimetry studies have identified a group of growth-restricted fetuses with increased resistance to fetal blood flow. In general, these placentas are small with distal villous hypoplasia. Small fetal stem arteries in such cases have been described as narrowed with thickened walls (Giles et al. 1985), an observation supported by morphometric studies (Fok et al. 1990; Mitra et al. 2000). When severe and in the setting of a strong supporting clinical history, these changes may be noted, but they overlap significantly with postdelivery arterial vasospasm.

19

Diseases of the Placenta

Chorangiosis and Chorangiomatosis Although chorangiosis and chorangiomatosis are not lesions of FVM, they are lesions involving the fetal vasculature and so are conveniently described in this section. Villous chorangiosis is a form of villous hypervascularity in which villi have expanded outlines and contain increased capillaries. Groups of more than 10 terminal villi with more than 10 capillaries (actually more than 15 are usually present) involving several areas per microscopic section are definitional (Fig. 87) (Altshuler 1984). The capillaries are centrally located with a thin basement membrane, and pericytes are absent. This condition is distinguished from congestion in which capillaries are prominent but normally distributed and not numerically increased. Although chorangiosis has been reported in up to 7% of placentas, strict adherence to criteria results in a significantly lower incidence. The changes are common in diabetics. The occurrence of chorangiosis in pregnancies at high altitude suggests that it is an adaptive response to hypoxia (Soma et al. 1996). While frequent in cases with abnormal outcome and hypoxia, chorangiosis has not yet been shown to be an independent causative factor (Ogino and Redline 2000). In diffuse, multifocal chorangiomatosis, villous capillaries are also numerically increased, but unlike chorangiosis, they are accompanied by pericytes, involve stem villi, and usually occur in immature placentas of less than

1273

32 weeks. The proliferation of capillaries permeates the villous tree rather than involving only the terminal villi. The clinical significance of this recently described and uncommon lesion has not been fully defined (Ogino and Redline 2000).

Fetal Membranes The fetal membranes and contiguous placenta form a sac containing amniotic fluid in which the fetus grows and develops. The fetal membranes provide a crucial barrier against infection, contain the amniotic fluid, and have metabolic functions including the modulation of events related to the onset of labor. The fetal membranes examined in their natural anatomic configuration after reconstructing the gestational sac as it exists in utero permit assessment of size, completeness, membrane insertion, and point of rupture. The distance from the point of membrane rupture to the edge of the placental disk reflects the site of uterine implantation; the lower the implantation, the closer the membrane rupture site is to the disk. A configuration indicative of low implantation should prompt consideration of associated conditions such as marginal hematoma, marginal placenta previa, or placenta accreta. Microscopic sections of the fetal membranes include fetal amnion and chorion and maternal decidua. The amnion is composed of amniotic epithelium and a connective tissue layer separated from the surrounding chorion by a loose spongy layer. The chorion is also composed of a connective tissue layer and a cellular layer containing chorionictype IT and atrophied villi. The most peripheral layer is maternal decidua capsularis (Fig. 5).

Squamous Metaplasia

Fig. 87 Chorangiosis. Small, central villous capillaries are greatly increased

Foci of squamous metaplasia are slightly elevated, sometimes targetoid, pearly white macules that tend to be most numerous at the site of cord insertion (Fig. 88). Although they are generally no more than a few millimeters, rarely they may form larger plaques. Histologically, foci of squamous epithelium, with or without keratinization,

1274

have a sharp transition with the surrounding amniotic epithelium. Squamous metaplasia has no clinical significance, as it is merely a maturation of the amniotic epithelium which is contiguous over the fetal membranes, umbilical cord, and fetal skin. It is important to distinguish from amnion nodosum, which it superficially resembles grossly.

Amnion Nodosum Definition and pathology. Amnion nodosum is a rare condition in which the amniotic surface is studded with small (1–5 mm), irregular, yellowish elevated nodules (Fig. 89). These nodules are generally concentrated on the chorionic plate, particularly around the insertion of the umbilical cord, although they may occur anywhere on the

Fig. 88 Squamous metaplasia. Squamous metaplasia is visible as flattened white plaques

R. N. Baergen et al.

amniotic surface. The nodules are composed of amorphous, eosinophilic material containing cells and hair fragments (Fig. 89). The amniotic epithelium is generally absent beneath the nodules. Etiology and clinical features. Amnion nodosum is associated with severe oligohydramnios. As the amniotic epithelium depends on the amniotic fluid for nourishment, the lack of fluid causes death of the epithelium leading to denuded basement membrane. The particulate debris and cellular elements from the fetal epidermis, oral cavity, and urinary and gastrointestinal tracts and the amnion itself are abnormally concentrated when amniotic fluid is scant and are deposited on the sticky basement membrane. The cause of the oligohydramnios varies. In many cases, a fetal urinary tract abnormality (renal agenesis or urinary tract obstruction) is responsible for diminished fetal urine resulting in decreased amniotic fluid. Oligohydramnios may occur in association with IUGR, but this rarely causes amnion nodosum. However, the donor twin in the TTTS often shows amnion nodosum. Long-standing amniorrhea is less likely to result in amnion nodosum than decreased amniotic fluid production, presumably because in the former, the cellular and particulate debris are lost along with the amniotic fluid. Amnion nodosum is a reliable indicator of oligohydramnios that should prompt an investigation for fetal abnormalities known to accompany it, including urinary tract anomalies and pulmonary hypoplasia (Benirschke et al. 2012).

Fig. 89 Amnion nodosum. Irregular elevated amniotic nodules (left) correspond to nodular deposits composed of degenerating cell fragments and hair embedded in amorphous granular material (right)

19

Diseases of the Placenta

Amniotic Bands Definition. Disruption of the amnion early in gestation with subsequent separation of the amnion and chorion results in amniotic fragmentation and shredding and leads to the formation of thin fibrous strands. These may encircle fetal limbs, digits, neck, and umbilical cord, causing characteristic constriction, amputation, and syndactyly (Fig. 90). The sequelae are highly variable depending on which fetal part becomes entangled. Some babies with characteristic amniotic band defects have additional structural anomalies, usually severe, including major limb deficiencies, body wall or open cranial defects, short umbilical cord, club feet, or internal abnormalities. The wide spectrum of anomalies in this condition has been variously referred to as the amniotic band syndrome, amniotic band disruption complex, early amnion rupture sequence, limb–body wall complex, and amnion adhesion malformation syndrome.

1275

Pathology. Gross identification of the bands and strings may be difficult. The surface of the placenta may be dull and slightly roughened as it often lacks amniotic epithelium. Microscopically, the bands usually consist of amniotic connective tissue, but occasionally amniotic epithelium is recognizable. In cases associated with body wall or open cranial defects, the amnion may be continuous with the fetal skin at the site of the defect. Broad adhesions may occur between the placenta and fetus. Etiology. Multiple theories have been proposed to explain the abnormalities in this syndrome. Torpin strongly espoused the concept that amniotic bands cause the structural defects (Torpin 1965). The widely variable nature of the defects has been attributed to the timing of amniotic rupture; early rupture is thought to result in fetal compression, tethering, or swallowing of bands, resulting in severe multisystem defects, whereas constriction and amputation defects in limbs and digits have been attributed to amniotic rupture

Fig. 90 Amniotic bands. Amniotic bands entangle limbs (a), digits (b), and the umbilical cord (c)

1276

later in pregnancy. Malformations are thought to result when a band interferes with the normal sequence of embryonic development. Kalousek maintains that the extent of amniotic bands is as important as the developmental stage at which they occur in determining of the pattern of fetal involvement (Kalousek and Bamforth 1988). The etiology of amniotic rupture is unknown. Amniotic bands have been noted rarely after trauma, in amniocentesis, and in women with connective tissue disorders. Others have postulated that the amniotic bands are secondary and that the more severe fetal abnormalities are the result of vascular disruption or a primary embryologic defect. Clinical significance. The recognition of amnion rupture and its consequences is important in counseling parents because the risk of recurrence is negligible. Unless accompanied by typical constriction/amputation lesions, major craniofacial or body wall defects may be difficult to diagnose. An important clue is the variety and asymmetry of the fetal defects, which are unlike the pattern of any heritable syndrome. No two cases are alike.

Meconium Definition and frequency. Meconium, the intestinal contents of the fetus, is commonly passed into the amniotic fluid, particularly at or near term. The reported prevalence ranges from 7% to 25%. Meconium passage is especially common in post-term placentas, present in up to 31%. It is unlikely to be present before 30 weeks. Pathology. Meconium can often be identified on gross examination. Meconium passed shortly before delivery may be recognizable as such as it is present on the placental surfaces but does not stain the membranes, fetal surface, or cord. The membranes become progressively green-stained and slimy, with longer exposure, dark and edematous, and eventually dull, muddy brown (Fig. 91). Microscopically, the amniotic epithelium exposed to meconium shows degenerative changes including heaping, stratification, and eventually nuclear pyknosis and necrosis. There may be marked edema of the spongy layer. With time,

R. N. Baergen et al.

Fig. 91 Meconium-stained placenta

meconium is engulfed by macrophages in the amnion, chorion, and decidua (Fig. 92). In vitro studies have shown variable time frames for meconium discharge ranging from 1 to 3 h to 24 to 48 h (Miller et al. 1985; Funai et al. 2009). How closely these observations approximate events in vivo is unknown. Meconium is usually scant in the umbilical cord due to the paucity of macrophages. As meconium passes through the membranes and cord, it eventually reaches the large fetal vessels where the toxic effect of its constituents can lead to apoptotic cell death of medial smooth muscle cells at the periphery of umbilical or chorionic plate vessels, meconium-associated myonecrosis (Fig. 93). Meconium must be distinguished from hemosiderin, generally a larger, more refractile, and yellowish crystalline granule. Iron stain is helpful when an ambiguous pigment is encountered. Lipochrome and nonhemosiderin, nonmeconium pigments of unknown composition, have been identified in a variety of diverse clinical situations. It has been hypothesized that these pigments may represent metabolites of remotely passed meconium. Clinical significance. Meconium staining has long been perceived as an indicator of perinatal morbidity. Meconium passage has been significantly associated with parameters of fetal distress including low Apgar scores, umbilical artery pH of 7.0 or less, respiratory distress, seizures in the first 24 h, and the need for delivery room resuscitation. Neonatal morbidity of all kinds has been significantly associated with meconium stained as compared to clear amniotic fluid (Kraus et al.

19

Diseases of the Placenta

Fig. 92 Meconium reaction. Amniotic epithelium is heaped and stratified with macrophages containing meconium in the subjacent stroma

1277

streptococcus is enhanced by the presence of meconium, and meconium has been shown to inhibit neutrophil function in vitro. Meconium may induce injury directly (in amniotic epithelium and umbilical and chorionic plate vessels) and has also been demonstrated to cause vasoconstriction, a potential cause of ischemia (Altshuler and Hyde 1989; Sienko and Altshuler 1999). There is some evidence that meconium may interfere with surfactant function and, in high enough concentrations, have a direct toxic effect on type II pneumocytes, possibly contributing to the meconium aspiration syndrome (Cleary and Wiswell 1998). The role of meconium as the primary factor in the meconium aspiration syndrome is controversial. Autopsy studies indicate that in most cases, meconium aspiration syndrome is of prenatal origin, especially in relation to intrauterine infection and chronic hypoxia (Ghidini and Spong 2001).

Gastroschisis

Fig. 93 Meconium-induced myonecrosis. The outer smooth muscle in the wall of this umbilical artery has undergone apoptosis in response to prolonged meconium exposure. The nuclei are pyknotic and the cytoplasm is rounded, dense, and eosinophilic

2004). Equally clear is that many infants, especially at term or post-term, pass meconium as a reflection of physiologic maturity, usually unassociated with significant problems and many fetuses are clearly in distress or even suffer demise without discharging meconium. The role of meconium as the primary factor in perinatal injury is, therefore, controversial. The most important consideration is the circumstance under which meconium is passed. Meconium staining is often superimposed on other significant pathologies, especially chorioamnionitis that may represent the major injurious factors. Meconium may, however, potentiate the effects of these underlying pathologies. For example, the growth of group B

Gastroschisis is a defect in the paraumbilical abdominal wall through which bowel protrudes. This is distinguished from the more common omphalocele in which bowel protrudes into a saccular defect at the cord insertion but remains enclosed in peritoneum and amnion. Gastroschisis is characterized by extensive, fine, uniform vacuolization of amniotic epithelial cells (Fig. 94). Ultrastructural studies confirm that the vacuoles contain lipid, but the origin of the lipid is obscure (Benirschke et al. 2012). These amniotic changes are not present in omphalocele.

Umbilical Cord Normal Anatomy and Embryonic Development Early in gestation, the blastocyst is filled with a loose meshwork of extraembryonic mesoderm that cavitates centrally to form the chorionic cavity. The embryonic structures are connected to the trophoblastic shell by a bridge of

1278

R. N. Baergen et al.

Embryonic Remnants

Fig. 94 Gastroschisis. The amniotic epithelium in gastroschisis shows fine vacuolization. The epithelial stratification in this case is the result of meconium passage

extraembryonic mesoderm, the connecting stalk, and forerunner of the umbilical cord. The yolk sac and the allantois protrude into the connecting stalk. As the amnion enlarges, the embryo prolapses into the amniotic cavity, progressively lengthening the connecting stalk. The allantoic vessels connect with vessels developing independently in the villi to establish the fetoplacental (chorioallantoic) circulation. The normal umbilical cord contains two arteries and one vein suspended in Wharton’s jelly, a loosely structured myxoid tissue covered by firmly attached amnion. Wharton’s jelly is derived from the extraembryonic mesenchyme and consists of myofibroblasts, abundant ground substance, and water. The combination of loose gel and contractile cells helps maintain turgor and protect the vessels against compression. The umbilical cord is supplied by oxygen and nutrients from the umbilical vessels. No other vessels or lymphatics are found in the normal umbilical cord. Most umbilical arteries are either fused or connected via an anastomosis (Hyrtl’s anastomosis) generally within 1.5 cm of the placental insertion site. This connection is important to equalize flow and distribute blood uniformly to the placenta. The normal umbilical cord is spiraled, usually counterclockwise or to the left (counterclockwise/clockwise = 7:1), and the average number of coils is 0.2 coil/cm. The spiral is established early in the first trimester as demonstrated sonographically.

Embryonic remnants dating back to the formation of the connecting stalk and umbilical cord are common microscopic findings. The presence of embryonic remnants has not been correlated with congenital anomalies, maternal age, race, gravidity, or gestational age at delivery. Remnants of the allantoic duct are frequent (about 15%) in the proximal portion of the cord. They are lined by flat or cuboidal cells reminiscent of transitional epithelium, with or without a lumen (Fig. 95). Allantoic remnants are located between the umbilical arteries. Rarely, they are large enough to expand the cord, or they may remain patent predisposing to urinary leakage from the cord stump. Traces of the omphalomesenteric duct, which connects the fetal ileum and the yolk sac in the early embryo, are infrequent, occurring in 1.5% of umbilical cords. These remnants are usually discontinuous, located peripherally, and lined by columnar cells resembling intestinal epithelium (Fig. 95). Omphalomesenteric remnants sometimes have a muscular wall, occasionally containing ganglion cells or liver, pancreatic, gastric, or small intestinal mucosa. Vitelline vessels may accompany omphalomesenteric remnants, or they may occur in isolation. These are usually paired but sometimes clustered, lined by endothelium lacking a muscular coat. Omphalomesenteric remnants are of little clinical significance. They are rarely associated with Meckel’s diverticulum, small intestinal atresia, or intestinal protrusion into the cord that may be inadvertently clamped or cut. Cystic omphalomesenteric remnants are rare, more common in males (M:F = 4:1). The yolk sac remnant is commonly visible as a small flattened, white nodule between the amnion and chorion composed of amorphous basophilic material histologically.

“Cord Accidents” The umbilical cord is a crucial lifeline between the fetus and placenta. Cessation of or diminished umbilical blood flow can result in severe fetal

19

Diseases of the Placenta

1279

Fig. 95 Allantoic (left) and omphalomesenteric remnants (right)

compromise or death. Umbilical flow may be compromised by mechanical factors (compression), vessel damage (trauma, inflammation, or meconium), or thrombosis. Cord abnormalities with the potential for obstruction of blood flow – abnormal coiling, stricture, abnormal length, true knots, entanglements, prolapse, and velamentous insertion – have been associated with increased risk of IUFD, IUGR, and neurologic injury (Baergen 2007; Baergen et al. 2001). Many of these conditions are related; true knots and excessive coiling commonly occur in long cords, and stricture almost always occurs in a hypercoiled cord. Chronic partial or intermittent flow obstruction may be evidenced by umbilical, chorionic, or stem vessel dilatation and thrombosis, intimal fibrin cushions, and/or villous alterations reflecting fetal vascular occlusion (FAV/villous stromal–vascular karyorrhexis) (Parast et al. 2008). However, actual thrombosis in the umbilical cord is a rare phenomenon. Based on these findings, non-acute cord compression has been implicated in over half of unexplained fetal deaths. Acute flow obstruction may supervene when a knot or entanglement tightens during delivery or there is cord prolapse. Doppler studies have confirmed that cord obstruction and compression impede venous return. Increasingly sophisticated imaging techniques provide opportunities for better assessment of the relationship between pathologic findings and significant alterations in blood flow.

Cord Length Umbilical cord length is an important parameter most accurately documented in the delivery room before the cord shrinks or cord segments are removed for other studies. Measurement of cord length in the hours after delivery reveals shrinkage of the length by at least several centimeters. Standards for cord length relative to gestational age have been established (Kraus et al. 2004). The mean cord length at term is about 55–60 cm. Cord length reflects factors that influence its growth – mainly tensile forces related to fetal activity and intrauterine conditions affecting fetal movement – although a genetic component to the determination of umbilical cord length also likely exists (Baergen et al. 2001). Umbilical cord growth slows during the last trimester as room for fetal movement declines, although some cord growth occurs normally until term. Conditions restricting fetal mobility – amniotic bands, oligohydramnios, anomalies such as skeletal dysplasia and crowding (multiple pregnancy) – are often associated with relatively short cords. Infants with Down’s syndrome have short, untwisted cords. Naeye has correlated short cord with subsequent motor and mental impairment (Naeye 1985). Extremes of cord length are associated with potentially adverse outcomes with both short and long cords being associated with an increased risk

1280

of neurologic impairment (Baergen 2007). Some consideration of relative as well as absolute cord length is appropriate. For example, a long cord with extensive looping may function as a relatively short cord.

Short Cord Definition and frequency: Gardiner’s calculation that a normal vertex delivery requires a minimum cord length of 32 cm provides a common definition of an abnormally short cord (Gardiner 1922). Using this definition, between 0.4% and 0.9% of umbilical cords are abnormally short. Berg and Rayburn found that 2% of cords are less than 35 cm (Berg and Rayburn 1995). Clinical significance: Unduly short cords have been linked to fetal distress in some cases, although blood pH and base deficit values in short cords are reportedly the same as in cords of normal length (Berg and Rayburn 1995). In the absence of fetal anomalies, short cords have been associated with low Apgar scores, neonatal hypotonia, and the need for resuscitation. Short cords may be associated with rupture or hemorrhage, delayed second stage of labor, abruption, subamniotic hemorrhage, and uterine inversion. At the extreme, there may be complete or nearcomplete cord absence (acordia) characteristically associated with fetal anterior abdominal wall defects that are directly attached to the placenta. Long Cord/Entanglements/Prolapse Definition and frequency. Excessively long cords have been variously defined as greater than 70 (Baergen et al. 2001), 80 (Berg and Rayburn 1995), or 100 cm (Kraus et al. 2004). Cord entanglements encircling the fetal neck, body, and extremities are common, occurring in about 23% of deliveries, but they are much more common in excessively long cords. Umbilical cord prolapse is an obstetric emergency defined as presentation of the cord in advance of the presenting part, occurring in 0.25–0.5% of deliveries. Clinical significance. Long cords have been associated with IUGR, IUFD, brain imaging abnormalities, and poor neurologic outcomes (Baergen 2007; Baergen et al. 2001). Histological

R. N. Baergen et al.

evidence consistent with venous obstruction has been described in the placenta. Abnormally long cords are also associated with excess knotting, hypercoiling, entanglements, and prolapse. Most cord entanglements do not have adverse outcomes, but some may result in cord compression. Tight entanglements have been associated with low Apgar scores and stillbirths. Babies with sonographically documented nuchal cords have higher rates of cesarean section and NICU admission. Nuchal cords have been demonstrated as a cause of IUGR, indicating that the deleterious effect is long term in some cases (Soernes 1995). Tight nuchal cords restricting venous return may result in neonatal anemia or even hypovolemic shock. CP has been linked to tight nuchal cords at delivery (Nelson and Grether 1998). Cord prolapse has a perinatal mortality rate of 20% (Lin 2006). Pathologic changes. Constriction of the umbilical cord and the encircled fetal part may be dramatic in some cord entanglements. Cord compression may be associated with edema, venous congestion, hemorrhage, and thrombosis of umbilical or large chorionic plate vessels and/or villous abnormalities reflecting fetal vascular occlusion.

Knots Frequency and etiology. Between 0.35% and 0.5% of umbilical cords contain true knots. True knots (Fig. 96) should be distinguished from false knots, which are focal vascular redundancies caused by differential growth of the umbilical vessels (Fig. 97). True knots are thought to be related to fetal movement and are increased in long cords, male fetuses, monoamniotic twins, and multigravidae and in association with excess amniotic fluid. They occur in abortions as well as in the third trimester indicating that they probably develop early in pregnancy when there is ample opportunity for movement. Pathology. Umbilical cord knots should be assessed for evidence of chronicity, tightness, and circulatory compromise. In long-standing tight knots, there are grooving and loss of

19

Diseases of the Placenta

Fig. 96 True knots. Long cord with two true knots. (Used with permission of the American Registry of Pathology/ Armed Forces Institute of Pathology)

1281

Fig. 98 Tight knot resulting in IUFD. (Used with permission of the American Registry of Pathology/Armed Forces Institute of Pathology)

Hypercoiling and Hypocoiling

Fig. 97 False knot

Wharton’s jelly with constriction of blood vessels, changes that persist when the knot is untied. Thrombosis of umbilical or chorionic plate vessels, sometimes calcified, and villous alterations reflecting fetal vascular occlusion indicate chronic vascular obstruction. Acutely tightened knots may be associated with venous distention distal to the knot, edema, and villous vascular congestion. Clinical significance. True knots are associated with an overall perinatal mortality rate of 8–11%, attributable to their potential for fetal circulatory obstruction (Fig. 98). Either an acutely tightened or long-standing knot may be responsible for intrauterine or intrapartum fetal death (Hershkovitz et al. 2001). False knots are generally of no clinical significance.

Definition and frequency. The normal umbilical cord averages about 0.2 coils/cm (coiling index = no. of coils per cord length). Hypercoiled cords are generally considered to be those in which the coiling index is >0.3 coils/cm, and hypocoiled cords are those in which the coiling index is 10 cm) from the placental insertion site because the umbilical arteries frequently fuse close to the placenta (Hyrtl anastomoses). Etiology. Whether SUA is due to primary aplasia or secondary atrophy has long been debated, but it is likely that both occur and that primary aplasia is more likely to be associated with other fetal anomalies. When specifically sought, histologic evidence of vascular remnants is demonstrable in some single-artery cords. Pathology. Muscular or elastic remnants of an atrophied vessel are identified in some cases. Single-artery cords are commonly associated with velamentous insertion (as high as 12%) and circumvallation. On occasion, two umbilical arteries are present but are of markedly different sizes. A difference of at least 1 mm in umbilical artery diameter as established sonographically is considered discordance (Fig. 105) (Raio and Ghezzi 1998). Marked umbilical artery discordance may be associated with fetal anomalies similar to those encountered with SUA. Clinical Significance Congenital malformations. There is a welldocumented association between SUA and fetal malformations, but there is no particular organ or specific abnormality characterizing this association. Any organ system may be affected, and malformations are frequently multiple.

Fig. 105 Discordant umbilical arteries

Congenital malformations are most numerous and most severe in stillborn and aborted babies and those dying in the neonatal period. Infants with single-artery cords but no detectable abnormalities at birth who survive the neonatal period are unlikely to have other significant abnormalities detected subsequently. SUA is almost invariable in sirenomelia and acardiac fetuses. Whether SUA plays a role in the development of congenital malformations or is just another manifestation of them is unclear. Perinatal mortality. The perinatal mortality rate of infants with SUA is greatly increased (11–41%). This is attributable to associated major malformations in most instances, although otherwise normal infants with SUA have an increased perinatal mortality rate as well. The sonographic demonstration of decreased Wharton’s jelly in single-artery cords may contribute to cord vulnerability. Low birth weight. SUA is associated with low birth weight even when infants with malformations are excluded from analysis.

Miscellaneous Vascular Abnormalities Segmental Thinning Qureshi and Jacques reported segmental thinning of umbilical vessels with virtual absence of the media in 1.5% of consecutively examined

1286

placentas (Qureshi and Jacques 1994). The umbilical vein was involved in the majority of cases, although on occasion, one or both arteries exhibited identical changes. Both superficial and medial aspects of the vessels exhibited wall deficiencies. Segmental vascular thinning was accompanied by fetal malformations in a significant number of cases, and there was a high incidence of fetal distress. Meconium-Associated Myonecrosis Long-standing meconium exposure may induce muscle necrosis in umbilical and chorionic plate vessels. Myonecrosis usually involves the superficial aspects of the arteries closest to the surface. The necrotic cells are rounded with dense eosinophilic-smudged cytoplasm and pyknotic or absent nuclei (Fig. 93) (King et al. 2004). Meconium-associated vascular necrosis has been linked to CP (Redline and O’Riordan 2000). In addition to vascular damage, meconium has been shown to cause vasoconstriction of umbilical vessels in vitro (Altshuler and Hyde 1989).

Ulceration Linear ulceration of Wharton’s jelly with vascular necrosis, aneurysmal dilatation, and rupture may result in intraamniotic hemorrhage, profound fetal anemia, and fetal death in utero (Fig. 106). These cord anomalies have been described in association with fetal intestinal atresia.

Fig. 106 Ulceration of Wharton’s jelly. Necrosis of Wharton’s jelly with thinning and rupture of umbilical vessels resulted in IUFD in this baby with duodenal atresia

R. N. Baergen et al.

Thrombosis Definition and frequency. Thrombosis of umbilical vessels is quite rare but when it occurs, it may be occlusive or nonocclusive and is usually associated with similar changes in chorionic plate or stem villous vessels. Thrombosis of umbilical cord vessels has been reported to occur in 1 in 1,300 deliveries, 1 in 938 perinatal autopsies, and 1 in 250 high-risk pregnancies (Sato and Benirschke 2006b). Etiology. Thrombi may be associated with cord compression, abnormal coiling, knots, stricture, hematoma, inflammation, anomalous insertion, amniotic bands, or entanglements. Other factors such as thrombophilic states could act synergistically to precipitate thrombosis. In many instances, the etiology is obscure. Pathology. Thrombi more commonly involve the umbilical vein alone (71%) with a lesser frequency of combined vein and artery thrombosis (18%) or arterial thrombosis alone (11%). If chronic, thrombi may show vessel calcification (Fig. 107). The chorionic plate and stem villous vessels may be similarly affected. Clinical significance. Fetal morbidity and mortality are high, particularly with occlusion of both umbilical arteries. Thrombi may embolize to the fetus causing infarcts in fetal organs, or they may result in the loss of a significant fraction of the fetoplacental circulatory bed. Thrombi have also been associated with CP, IUGR, and IUFD.

(Used with permission of the American Registry of Pathology/Armed Forces Institute of Pathology)

19

Diseases of the Placenta

1287

Fig. 107 Thrombosed umbilical artery. This thrombosed umbilical artery, cause undetermined, is visible as a dark spiral. (Used with permission of the American Registry of Pathology/Armed Forces Institute of Pathology)

Hematoma Definition and frequency. Umbilical cord hematomas are accumulations of blood in Wharton’s jelly. They are uncommon. Etiology. In the great majority of cases, an obvious etiology is not apparent. Rarely, hemorrhage has a demonstrable origin from an umbilical vein or artery, and origin from omphalomesenteric vessels has been proposed. Rupture of a varix, traumatic damage at the time of amniocentesis or percutaneous umbilical cord sampling, or inflammation or structural anomalies of a vessel wall have been suggested as possible mechanisms. In the majority of cases, hematomas of the cord are iatrogenic, caused by clamping of the cord after delivery. Thus, care must be taken in looking for adjacent clamp marks in cases with umbilical cord hematoma. Pathology. Most umbilical cord hematomas present as red-purple fusiform swellings (Fig. 108). They are generally confined to the cord, but on occasion, they may rupture into the amniotic cavity. Small collections of fresh blood usually reflect cord blood sampling or traction at the time of delivery. Clinical significance. A perinatal mortality rate in the range of 40–50% has been reported in association with umbilical cord hematoma. Fetal death or severe neurologic injury may occur due to blood loss or umbilical vascular compression with circulatory compromise. Hemangioma Definition. Hemangiomas are benign vascular tumors composed of proliferating vessels sometimes associated with marked myxoid degeneration of Wharton’s jelly (angiomyxoma).

Fig. 108 Localized cord hematoma. (Used with permission of the American Registry of Pathology/Armed Forces Institute of Pathology)

Fig. 109 Umbilical cord hemangioma. (Used with permission of the American Registry of Pathology/Armed Forces Institute of Pathology)

Pathology. Hemangiomas present as fusiform swelling of the umbilical cord usually at the placental end. They can attain substantial size (up to 900 g in one report) (Fig. 109). Microscopically, these benign tumors show features similar to benign hemangiomas at other body sites. Clinical significance. Hemangiomas with myxoid degeneration (angiomyxoma) can be associated with hemorrhage, increased alpha fetoprotein levels, and rarely nonimmune fetal hydrops, presumably due to high-output cardiac failure.

1288

Aneurysm Umbilical cord aneurysms are rare. They may be so large as to compress adjacent vessels resulting in fetal death. Umbilical artery aneurysm has been reported in trisomy 18 (Sepulveda et al. 2003).

Clinical Syndromes and Their Pathologic Correlates in the Placenta Although placentas often come to the pathologist with limited clinical data supplied by the obstetrician, here follow definitions of common clinical conditions that lead to submission of a placenta for pathologic examination together with the relevant pathologic lesions.

Preeclampsia Definition. Gestational hypertensive, pregnancyinduced hypertension and preeclampsia are clinical diagnoses associated with lesions of maternal vascular malperfusion in the placenta. A previously normotensive pregnant woman whose blood pressure reaches 140/90 mm Hg or greater after 20 weeks gestation on at least two occasions has pregnancy-induced hypertension. The addition of proteinuria (1+ or greater on urine dipstick confirmed by a 24 h collection containing 300 mg of protein) means that preeclampsia has occurred. Severe preeclampsia indicates the addition of one or more of the following: blood pressure over 160 mm Hg systolic or 110 mm Hg diastolic, proteinuria greater than 5 g/24 h or 3+ or greater on two dipstick evaluations at least 4 h apart, headache, visual disturbances or epigastric or right upper quadrant pain, oliguria, thrombocytopenia or liver enzyme elevation, fetal growth restriction, and/or pulmonary edema. The HELLP syndrome is a form of severe preeclampsia with a specific triad of findings including hemolysis, elevated liver enzymes, and low platelet count. Seizures signify the onset of eclampsia. Delivery of the placenta is often the only effective medical intervention. Etiology. The spiral arteries that supply maternal blood to the placenta must expand to accommodate the growing placentofetal unit. This is

R. N. Baergen et al.

normally accomplished by trophoblastic remodeling of spiral arteries. A failure of this process results in decidual vasculopathy, reduced maternal flow, placental ischemia, and the maternal syndrome of preeclampsia (Lim et al. 1997). In preeclampsia, there is generalized maternal endothelial dysfunction thought to be caused by soluble factors produced by ischemic trophoblast. Endoglin and sFlt-1 are two factors that have been implicated. Both appear prior to disease onset and cause endothelial dysfunction and hypertension in the mother. These factors act by neutralizing the angiogenic and vasodilatory effects of vascular endothelial growth factor and placental growth factor produced by the placenta (Mutter and Karumanchi 2008). When endoglin and sFlt-1 are overexpressed in an experimental rat model, the result is severe proteinuria, hypertension, and growth restriction as well as thrombocytopenia and liver dysfunction (Venkatesha et al. 2006). Pathology. Correlation between placental lesions and severity of maternal symptoms is poor, but correlation with fetal outcome is good. Two types of abnormal placentas are associated with preeclampsia. The first and most common is the small placenta with decidual vasculopathy, the pathologic changes characteristic of maternal vascular malperfusion and a thin umbilical cord. Less common is the large placenta associated with a heterogeneous group of conditions including diabetes mellitus, placental hydrops, multiple gestation, and hydatidiform mole.

Essential Hypertension Patients with preexisting hypertension not complicated by superimposed preeclampsia may have similar but less pronounced abnormalities in the placenta. Arteriosclerotic changes in the uterine or intramyometrial arteries may be sufficient to produce some degree of placental ischemia.

Diabetes Mellitus Placentas associated with uncomplicated maternal diabetes mellitus vary. In about half of cases, they appear normal, grossly, and microscopically.

19

Diseases of the Placenta

1289

ACA is most common, occurring in 61% of premature placentas (Hansen et al. 2000a, b). ACA is more common in earlier PTB. Lesions resulting from impaired blood flow are more common in later PTB (Hecht et al. 2008). Premature labor and delivery often are associated with decidual hemorrhage and placental abruption.

Post-term Pregnancy Fig. 110 Distal villous immaturity. Increased numbers of large distal villi with increased capillaries

There is some correlation between placental findings, metabolic control, and clinical severity. When they are abnormal, diabetic placentas tend to be large, heavy, and congested, with large and edematous umbilical cords. Villi in the larger placentas often show villous hypervascularity or even chorangiosis, distal villous immaturity, large terminal villi with increased central capillaries, and the presence of nucleated red blood cells in fetal vessels (Fig. 110). Placental, fetal, and neonatal thrombi are more frequent, a manifestation of the thrombophilia associated with diabetes. When complicated by hypertension or preeclampsia, the placenta may be small with evidence of low maternal flow.

PTB, PTL, and Preterm Rupture of Membranes A term birth takes place between 37 and 42 weeks of gestation. The most severe sequelae for the newborn occur before the 34th week. The category of “VLBW” infants refers to live-born infants weighing between 500 and 1,500 g. These are at greatest risk of neonatal death, neurologic injury, and pulmonary immaturity. PTB occurs in 12–13% of all pregnancies in the United States. The result is a 40-fold increase in perinatal morbidity and mortality. Most PTBs are due to ACA or preeclampsia. The two pathologic subgroups show very little overlap (Arias et al. 1993; Hansen et al. 2000b).

A gestation longer than 42 completed weeks (294 days) is considered post-term. Perinatal mortality is increased. The postmature newborn has a characteristically slender, elongated body and limbs; wrinkled, often meconium-stained skin; and long fingernail. Meconium staining is more common in post-term placentas, but otherwise there are no distinctive or definitive pathologic changes in the placenta (Benirschke et al. 2006). Contrary to previous beliefs, the placenta does not become senescent after it reaches term but could continue to function adequately if the fetus did not deliver.

Fetal Growth Restriction, Intrauterine Growth Restriction (IUGR) Prenatal evaluation of fetal growth is based on ultrasound measurements of abdominal circumference, head circumference, biparietal diameter, and femur length (among others), which may also be used to estimate fetal weight (Resnik 2002). IUGR is a significant risk factor for poor outcome and stillbirth (Turan et al. 2008). IUGR is symmetric when head and abdominal measurements are decreased proportionately and asymmetric when there is a greater decrease in abdominal girth. When IUGR is identified, Doppler studies of blood flow through the umbilical cord, middle cerebral artery, or other sites may be obtained. Absent or reduced umbilical artery blood flow is a significant adverse finding, which may begin early and persist for weeks until delivery (Rigano et al. 2001). Infants whose birth weight lies below an expected percentile for gestational age (which may be arbitrarily set at the third, fifth, or tenth

1290

percentile) are classified as SGA and qualify as growth-restricted. Maternal factors that contribute to IUGR include preeclampsia, hypertension, diabetes mellitus, thrombophilias, extreme malnutrition, chronic renal disease, tobacco and other drug abuse, and poor obstetric history in general. Fetal factors include chromosomal anomalies (including confined placental mosaicism), congenital malformations, and multiple gestation (WilkinsHaug et al. 2006). Placental lesions associated with IUGR include (1) vascular lesions that reduce maternal blood flow (decidual vasculopathy or chronic abruption), (2) vascular lesions that reduce fetal blood flow (FTV, large chorangiomas, or umbilical cord abnormalities), and (3) lesions that greatly reduce the amount of functional placenta (extensive chronic villitis, MPF, MFI, or multiple infarcts) (Sebire and Sepulveda 2008).

NE, CP, and “Birth Asphyxia” Historically, the occurrence of abnormal neurologic findings in the first week of life was assumed to have followed a period of asphyxia (severe hypoxia with metabolic acidosis) during labor and delivery. This concept led to such clinical terms as “birth asphyxia” and “hypoxic–ischemic encephalopathy.” While episodes of pure hypoxia do occur (acute abruption is a good example), other factors – intrauterine infection and other acute and chronic placental lesions – are much more common. The transient clinical state, characterized by hypotonia, apnea, coma, and seizures, is now called neonatal encephalopathy. The incidence varies from 6 to 8 per 1,000 births. CP, by contrast, is a chronic, nonprogressive neurologic disorder, most often a form of spastic diplegia, hemiplegia, or quadriplegia, with an incidence of 2 per 1,000 births. It is usually not diagnosed much before age 2 years, certainly not at the time of delivery. Approximately 50% of cases develop after full-term birth following an apparently normal gestation. The remainder are divided between VLBW infants (12 weeks). In general, embryonic/fetal factors, primarily chromosomal abnormalities, are responsible for most early abortions. Late spontaneous abortion/stillbirth is more likely to be associated with placental and maternal factors.

Early Abortion The embryonic stage of development extends to the eighth gestational week. Early abortion includes losses up to the 12th week. More than half of early spontaneous abortions have demonstrable chromosomal anomalies, and most of these are eliminated in the embryonic period. Trisomies, triploidy, and monosomy X are most common (Lash et al. 2008). In spontaneously passed tissue, the chorionic cavity, intact or disrupted, may be identified. The chorionic cavity is frequently empty but may contain remnants of cord or embryo. Embryos frequently exhibit generalized abnormal development (growth disorganization), pointing to a high probability of a chromosomal anomaly (Fig. 112). Localized developmental defects (facial fusion defects, ocular anomalies, limb bud deformities, neural tube defects, and cervical edema) are also commonly genetically

R. N. Baergen et al.

Fig. 112 Spontaneous abortion. The abnormal embryo in this early abortion is closely correlated with an abnormal karyotype

determined, although specific morphologic defects are often difficult to identify in the early embryo. These are more easily evaluated in the larger previable fetus. The finding of a well-developed, normal embryo suggests the possibility of a maternal causative factor (infection, inflammation, immune rejection, or coagulopathy), although 20–25% of morphologically normal embryos are also karyotypically abnormal, usually triploid. Accurate assessment of embryo morphology requires familiarity with normal developmental stages, which are well illustrated in the comprehensive work of Kalousek et al. (1990). Gross assessment of the embryo and/or placenta is, at best, suboptimal in the dissociated tissues of curetted specimens. The placenta is commonly retained in the uterus for a variable period after fetal death, and villi in abortion specimens often show alterations reflecting fetal death. Regardless of cause, fetal death leads to progressive involution of fetal vessels diffusely throughout the placenta. Involution begins with intravascular karyorrhexis (6 h), followed by septation of stem vessels (48 h), and ends with complete obliteration of vessels and villous stromal hyalinization (when extensive >2 weeks) (Dr and Singer 1992; Genest 1992; Genest et al. 1992). This sequence of changes is identical to that which occurs locally in villi distal to an occluded fetal vessel.

19

Diseases of the Placenta

The villi of abortions with abnormal karyotypes often appear abnormal, but the patterns are variable and nonspecific. Dysmorphic features suggestive of a chromosomal abnormality – villous enlargement with myxoid stroma, irregular villous outlines, multiple trophoblastic invaginations, and trophoblastic inclusions – are common and may be more marked in karyotypically abnormal abortions but are found in abortions with normal karyotypes as well. Complete and partial moles are the only morphologic entities with defined karyotypic abnormalities that can be diagnosed with any degree of confidence, although their distinction is becoming more difficult with earlier evacuation. Rarely, significant trophoblastic hyperplasia without molar change may occur in association with chromosomal abnormalities (especially trisomies 7, 15, 21, and 22) (Redline et al. 1998a). These cases are best managed with a follow-up serum human chorionic gonadotropin (HCG) titer to assure return to baseline. Karyotyping may be clinically very useful in the evaluation of couples distressed by repeated spontaneous abortion. Trisomies can be identified by fluorescence in situ hybridization (FISH) in formalin-fixed tissue. Occasionally, karyotypic abnormalities are confined to the placenta. In confined placental mosaicism, mosaicism is expressed in the placenta but not the fetus (Kalousek and Barrett 1994). These placentas show no histologic abnormalities; their identification requires genetic evaluation of villous stroma and trophoblast and the fetus.

1295

Fetomaternal hemorrhage is often overlooked as a cause of unexpected stillbirth. A Kleihauer–Betke test should be performed in all such cases. When the fetus is normally formed and the placental findings seem trivial, confined placental mosaicism is a consideration. Recurrent pregnancy loss may be associated with thrombophilias, maternal vascular obstructive lesions, and VUE when these conditions are severe and of early onset. Massive perivillous fibrin deposition and chronic histiocytic intervillositis are rare lesions, but they almost always lead to recurrent reproductive loss in affected individuals (Redline 2007b; Waters and Ashikaga 2006).

Nontrophoblastic Tumors Chorangioma Chorangiomas are placental hemangiomas arising in stem villi. They occur in about 0.5–1% of carefully examined placentas and are usually small, entirely intraplacental, and may be difficult to appreciate, especially in the unfixed specimen. Large tumors distorting either the chorionic or basal plate are rare. Exceptionally a chorangioma is attached to the placenta by a thin pedicle. Chorangiomas are usually solitary (Fig. 113), but they may be multiple (Fig. 114). They may be brown, yellow, tan, red, or white and are usually firm and well demarcated from

Late Abortion, Stillbirth, and Intrauterine Fetal Death Chromosomal abnormalities still occur but are uncommon after 20 weeks. Important considerations in explaining late abortion, stillbirth, and IUFD include intrauterine infections and inflammatory processes, maternal and/or fetal circulatory compromise, destructive placental lesions, and cord accidents. The gross and microscopic features of these conditions have been described throughout this chapter. Involutional changes that invariably occur following fetal death will be superimposed to greater or lesser extent.

Fig. 113 Chorangioma

1296

the surrounding parenchyma. They commonly bulge from the fetal surface and are often located peripherally. Most chorangiomas are composed of capillary-sized blood vessels supported by inconspicuous, loose stroma. Occasionally, they may be more cellular or show prominent myxoid change, hyalinization, necrosis, or calcification. Mitotic figures and nuclear atypicality have been reported in some chorangiomas, but these have not behaved aggressively. Trophoblastic hyperplasia may also be prominent (Ogino and Redline 2000). Localized groups of large stem villi with similar alterations that permeate through the villous tree but without nodular expansion have been

R. N. Baergen et al.

termed localized chorangiomatosis (Ogino and Redline 2000). The association with preeclampsia and highaltitude pregnancy suggests that decreased oxygen tension may play a role in their development. Although the majority of chorangiomas are of no clinical significance, various complications have been reported, and these are directly related to the size of the lesion. Hydramnios and premature delivery are the most significant. Fetal cardiomegaly, congestive heart failure, and hydrops have been attributed to the increased workload in shunting blood through a large chorangioma. When significant amounts of blood are directed through the chorangioma away from functional villi, chronic hypoxia may lead to intrauterine growth retardation. Complications including fetal–maternal hemorrhage, fetal anemia, and thrombocytopenia reflect sequestration or destruction of cellular elements as they traverse the chorangioma. Skin angiomas have been reported in a few babies with placental chorangiomas.

Hepatocellular Adenoma and Adrenocortical Nodules

Fig. 114 Multiple placental chorangiomas

Rarely small nodules of hepatocytes occur in the placenta. Termed hepatocellular adenoma, these nodules may contain hematopoietic foci but lack bile ducts and central veins (Fig. 115). Their immunoprofile is typical of hepatocytes. Tiny

Fig. 115 Hepatic adenoma. This discrete collection of hepatocytes showed extramedullary hematopoiesis

19

Diseases of the Placenta

1297

nodules of adrenal cortical cells have been considered heterotopias. The pathogenesis of these unusual findings is not understood.

Other Placental “Tumors” Despite intraplacental location and infiltrative border, careful molecular analysis of one intraplacental leiomyoma documented its maternal origin. Teratomas have been reported to occur rarely between the amnion and chorion on the chorionic plate and in the umbilical cord. Although teratomas have been distinguished from acardiac fetuses by lack of an umbilical cord and disorganization of the component tissues, this distinction is a matter of dispute.

Fig. 116 Metastatic breast cancer. The metastatic tumor is visible as small nodules. (Used with permission of the American Registry of Pathology/Armed Forces Institute of Pathology)

Placental Metastases Either maternal or fetal neoplasms may metastasize to the placenta, but they are very rare. Malignant melanoma is the most common maternal tumor to metastasize to the placenta despite the fact that other tumors are more common in this age group. Other maternal neoplasms metastasizing to the placenta have included leukemia, lymphoma, breast carcinoma, and lung carcinoma. Malignant melanoma, leukemia, and lymphoma have been reported to metastasize to the fetus transplacentally (Kraus et al. 2004; Baergen et al. 1997). Placentas harboring maternal metastases are often normal on gross inspection, although metastatic tumor deposits may be apparent grossly (Fig. 116). Tumor cells are usually confined to the intervillous space (Fig. 117). Villous or fetal vascular invasion is very uncommon and, even when present, does not correlate well with fetal spread. Dissemination of a congenital fetal tumor to the placenta is also very rare. Neuroblastoma is most common, but rare cases of fetal leukemia, lymphoma, hepatoblastoma, sarcoma, and sacrococcygeal teratoma have metastasized to the placenta (Fig. 118). Disseminated

Fig. 117 Metastatic breast carcinoma. Metastatic tumor cells are confined to the intervillous space

Fig. 118 Metastatic fetal melanoma. This congenital scalp melanoma metastasized to the placenta

histiocytosis involving the vessels of the umbilical cord has been documented. The placenta involved by a metastatic fetal tumor is consistently hydropic, pale, and bulky with tumor cells

1298

distending the fetal vessels. Pigmented nevus cells may involve the villi in association with giant nevi, but these are not considered to be malignant or metastatic.

Examination of the Placenta A gross examination of all placentas should be performed and recorded at the time of delivery by the clinician. Some information – length of umbilical cord, number of umbilical vessels, location of placenta in utero, retroplacental hematoma, ruptured membranous vessels, completeness of membranes, or maternal surface – may be of immediate relevance and is best documented by the practitioner. Cultures for microorganisms and tissue for cytogenetics are best obtained at this time. Placentas are most commonly submitted to the pathologist when either the mother or infant is abnormal; the course of pregnancy, labor, or delivery is complicated; or a gross placental abnormality is noted or for the evaluation of multiple pregnancy. Guidelines for submission of placentas for pathologic exam have been established (Kraus et al. 2004). A request for pathologic examination of the placenta should include the reason for submission with specific questions as well as a summary of the clinical history pertinent to that evaluation – the mother’s gravidity, parity, details concerning previous pregnancies, underlying maternal disease, antepartum course, labor and delivery, and the infant’s gestational age, weight, and Apgar scores. Placentas not submitted to pathology can be retained in the event that a complete placental exam should become necessary. A good placental exam can be performed on a fresh or fixed placenta. In our institutions, the placenta is examined fresh to establish weight and pertinent measurements, document gross lesions, and roll the membranes. After fixation, the cut surfaces are examined, and sections are submitted. Important abnormalities are unlikely to be overlooked when the placenta is approached systematically as follows.

R. N. Baergen et al.

Fetal Membranes Point of rupture: Measure the distance between the point of rupture and the closest placental margin. This provides information relating to the site of implantation (low-lying or placenta previa). Completeness: By attempting to reconstruct the amniotic sac, an estimation may be made as to whether membranes have been retained in utero. Insertion: The membranes may insert at the margin of the placenta or central to the placental margin (circumvallate). Note the extent (complete, partial) and associated changes. Color and opacity: Opaque or cloudy membranes suggest infection. Green/brown staining suggests meconium or hemosiderin deposition. Roll the membranes to include the point of rupture and the placental margin. Remove the membranes.

Umbilical Cord Length: Measure all pieces. The most accurate assessment of cord length is obtained in the delivery room. Umbilical cord length decreases by as much as 7 cm within a few hours of delivery. Diameter: Measure greatest and least. Insertion: Measure the distance between the insertion of the umbilical cord and closest placental margin. Note the mode and site of insertion and the condition of all membranous vessels. Vessels: Normally 3. Torsion: Degree of twist can be expressed as number of coils per 10 cm. Other: Strictures, edema, thrombosis, hematomas, ulceration, and knots. Remove the cord flush with the placental surface.

Placenta Measurements: Measure the diameter and thickness. Weight: Weigh after fetal membranes, umbilical cord, and large clots have been removed.

19

Diseases of the Placenta

Formalin fixation adds about 10% to placental weight. Shape: Note anomalous shapes. Fetal surface: Evaluate color (blue, green, brown) and chorionic vessels. Arteries cross over veins. Maternal surface: The maternal surface should be complete. Missing fragments suggest retained placental tissue. Evaluate all clots for site, size, and related placental alterations. Breadloaf the placenta at 1 cm intervals. Cut surface: Placental color reflects fetal blood content. Describe, measure, and locate focal abnormalities. Assess percentage when large lesions or significant proportions of the placenta are altered.

Sections Two sections of the umbilical cord and two sections of the membrane roll can be included in one block. Do not section the cord within 2 cm of the insertion site – this may give a false impression of SUA. We routinely submit at least three sections of placenta from grossly normal central placental parenchyma. An additional block with multiple sections of maternal surface can facilitate identification of abnormal maternal vessels. Additional sections to demonstrate pathologic alterations are submitted as appropriate. Selection of microscopic sections should be tailored to the specific clinical scenario. For example, additional sections of the fetal chorionic plate vessels optimize evaluation of fetal vessels for thrombosis when there is the possibility of cord complications.

Special Aspects of Multiple Pregnancy The delivering physician should identify the umbilical cords, but if not, they can be identified arbitrarily or by distinguishing characteristics (cord length, insertion site, etc.). Maintain the separate identity of each placenta grossly and microscopically. Placentation type: Two entirely separate placentas are obviously dichorionic, requiring routine

1299

examination. To distinguish between DiMo- and DiDi-fused placentas, assess the dividing membranes and fetal vascular pattern. In DiDi-fused placentas, the septum is opaque, and the fetal vessels do not cross the line of fusion. In DiMo placentas, the septum is thin, and the two vascular districts are intermingled. Placental mass: Quantify the placental mass belonging to each infant. DiDi-fused placentas can be separated along the line of fusion and weighed separately. Note that the placental territories of monochorionic infants correspond best with their chorionic vascular distributions and do not correspond to the location of the dividing membranes. Vascular anastomoses: Assess monochorionic placentas for vascular anastomoses. Large chorionic vessel anastomoses are easily identified grossly. Deep arteriovenous anastomoses cannot be seen but can be seen as an unpaired artery or vein.

Special Techniques Cytogenetics: Placental tissue may be easier to grow and karyotype than other fetal tissues. Chorionic villous sampling early in pregnancy is being used for prenatal diagnosis. Flow cytometry may be useful in the evaluation of molar pregnancies. Genetic questions regarding specific chromosomes may be addressed by FISH. Culture: This is usually done by the obstetrician. Cultures may be accomplished by swabs of the amniotic surface or subchorionic fibrin after searing the amnion. Immunoperoxidase stains: This technique may be useful in the identification of infectious agents and the delineation of cell types (i.e., CD68 in chronic histiocytic intervillositis).

References Abramowsky C, Beyer-Patterson P et al (1991) Nonsyphilitic spirochetosis in second-trimester fetuses. Pediatr Pathol 11:827–838 Adams-Chapman I, Vaucher YE et al (2002) Maternal floor infarction of the placenta: association with central

1300 nervous system injury and adverse neurodevelopmental outcome. J Perinatol 22:236–241 Al-Adnani M, Kiho L et al (2008) Recurrent massive perivillous fibrin deposition associated with polymyositis: a case report and review of the literature. Pediatr Dev Pathol 11:226–229 Altshuler G (1984) Chorangiosis. An important placental sign of neonatal morbidity and mortality. Arch Pathol Lab Med 108:71–74 Altshuler G, Hyde S (1989) Meconium-induced vasocontraction: a potential cause of cerebral and other fetal hypoperfusion and of poor pregnancy outcome. J Child Neurol 4:137–142 Ananth CV, Vintzileos AM et al (1998) Standards of birth weight in twin gestations stratified by placental chorionicity. Obstet Gynecol 91:917–924 Ananth CV, Oyelese Y et al (2005) Placental abruption in the United States, 1979 through 2001: temporal trends and potential determinants. Am J Obstet Gynecol 192:191–198 Arias F, Rodriguez L et al (1993) Maternal placental vasculopathy and infection: two distinct subgroups among patients with preterm labor and preterm ruptured membranes. Am J Obstet Gynecol 168:585–591 Arias F, Romero R et al (1998) Thrombophilia: a mechanism of disease in women with adverse pregnancy outcome and thrombotic lesions in the placenta. J Matern Fetal Med 7:277–286 Ariel IB, Anteby E et al (2004) Placental pathology in fetal thrombophilia. Hum Pathol 35:729–733 Badawi N, Felix JF et al (2005) Cerebral palsy following term newborn encephalopathy. Dev Med Child Neurol 47:293–298 Baergen RN (2007) Cord abnormalities, structural lesions, and cord accidents. Semin Diag Pathol 24:23–32 Baergen RN, Johnson D et al (1997) Maternal melanoma metastatic to the placenta: case report and review of the literature. Arch Pathol Lab Med 121:508–511 Baergen RN, Malicki D et al (2001) Morbidity mortality, and placental pathology in excessively long umbilical cords; retrospective study. Pediatr Dev Pathol 4:144–153 Baldwin VJ (1994) Pathology of multiple pregnancy. Springer, New York Barth WH Jr, Genest DR et al (1996) Uterine arcuate artery Doppler and decidual microvascular pathology in pregnancies complicated by type I diabetes mellitus. Ultrasound Obstet Gynecol 8:98–103 Baschat AA, Gungor S et al (2007) Nucleated red blood cell counts in the first week of life: a critical appraisal of relationships with perinatal outcome in preterm growth restricted neonates. Am J Obstet Gynecol 197(3):286. e1–286.e8 Benirschke K, Masliah E (2001) The placenta in multiple pregnancy: outstanding issues. Reprod Fertil Dev 13(7–8):615–622 Benirschke K, Kaufmann P et al (2006) Pathology of the human placenta, 5th edn. Springer, New York

R. N. Baergen et al. Benirschke K, Burton GJ et al (2012) Pathology of the human placenta, 6th edn. Springer, New York Berg TG, Rayburn WF (1995) Umbilical cord length and acid-base balance at delivery. J Reprod Med 40:1–12 Bieber FR, Nance WE et al (1981) Genetic studies of an acardiac monster: evidence of polar body twinning in man. Science 213:775–777 Blanc WA (1959) Amnionic infection syndrome. Pathogenesis, morphology, and significance in circumnatal mortality. Clin Obstet Gynaecol 2:705–734 Bloomfield RD, Suarez JR et al (1978) The placenta: a diagnostic tool in sickle cell disorders. J Natl Med Assoc 70:87–88 Boyd JD, Hamilton WJ (1970) The human placenta. Heffer, Cambridge Bruner JP, Anderson TL et al (1998) Placental pathophysiology of the twin: oligohydramnios-polyhydramnios sequence and the twin-twin transfusion syndrome. Placenta 19:81–86 Burke CJ, Tannenberg AE (1995) Prenatal brain damage and placental infarction-an autopsy study. Dev Med Child Neurol 37:555–562 Chan OT, Mannino FL et al (2007) A retrospective analysis of placentas from twin pregnancies derived from assisted reproductive technology. Twin Res Hum Genet 10(2):385–393 Chasen ST, Baergen RN (1999) Necrotizing funisitis with intrapartum umbilical cord rupture. J Perinatol 19(4):325–326 Chibueze EC, Tirado V et al (2017) Zika virus infection in pregnancy: a systematic review of disease course and complications. Reprod Health 14(1):28 Cleary GM, Wiswell TE (1998) Meconium-stained amniotic fluid and the meconium aspiration syndrome. An update. Pediatr Clin North Am 45:511–529 Cooperstock MS, Tummara R et al (2000) Twin birth weight discordance and risk of preterm birth. Am J Obstet Gynecol 183:63–67 Curtin WM, Shehata BM et al (2002) The feasibility of using histologic placental sections to predict newborn nucleated red blood cell counts. Obstet Gynecol 110:305–310 Dahms B, Boyd T et al (2002) Severe perinatal liver disease associated with fetal thrombotic vasculopathy in the placenta. Pediatr Dev Pathol 5:80–85 Davis BH, Olsen S et al (1998) Detection of fetal red cells in fetomaternal hemorrhage using a fetal hemoglobin monoclonal antibody by flow cytometry. Transfusion 38:749–756 De Laat MWM, van Alderen ED et al (2007) The umbilical coiling index in complicated pregnancy. Eur J Obstet Gynecol Reprod Biol 130:66–72 Decastel M, Leborgne-Samuel Y (1999) Morphological features of the human umbilical vein in normal, sickle cell trait, and sickle cell disease pregnancies. Hum Pathol 30(1):13–20 DeSa DJ (1973) Intimal cushions of foetal placental veins. J Pathol 110:347–352

19

Diseases of the Placenta

Di Naro E, Ghezzi F et al (2001) Umbilical vein blood flow in fetuses with normal and lean umbilical cord. Ultrasound Obstet Gynecol 17:224–228 Dizon-Townson DS, Meline L et al (1997) Fetal carriers of the factor V Leiden mutation are prone to miscarriage and placental infarction. Am J Obstet Gynecol 177:402–405 Doss BJ, Jacques SM et al (1998) Maternal scleroderma: placenta findings and perinatal outcome. Hum Pathol 28:1524–1530 Dr G, Singer DB (1992) Estimating the time of death in stillborn fetuses: III. External fetal examination; a study of 86 stillborns. Obstet Gynecol 80:593–600 Esplin MS (2006) Preterm birth: a review of genetic factors and future directions for genetic study. Obstet Gynecol Surv 61:800–806 Fernandes BJ, von Dadelszen P et al (2007) Flow cytometric assessment of feto-maternal hemorrhage; a comparison with Betke-Kleihauer. Prenat Diagn 27:641–643 Fok RY, Pavlova Z et al (1990) The correlation of arterial lesions with umbilical artery Doppler velocimetry in the placentas of small-for-dates pregnancies. Obstet Gynecol 75:578–583 Fox H (1966) Thrombosis of the foetal stem arteries in the human placenta. J Obstet Gynaecol Br Commonw 73:961–965 Fox H (1997) Pathology of the placenta, 2nd edn. Saunders, Philadelphia Fox H, Sebire NJ (2007) Pathology of the placenta, 3rd edn. Saunders, Philadelphia Frank HG, Malekzadeh F et al (1994) Immunohistochemistry of two different types of placental fibrinoid. Acta Anat 150:55–68 Fraser RB, Wright JR (2002) Eosinophilic/T-cell chorionic vasculitis. Pedatr Dev Pathol 5:350–355 Funai EF, Labowsky AT et al (2009) Timing of fetal meconium absorption by amniotic macrophages. Am J Perinatol 26(1):93–97 Gardella C, Riley DE et al (2004) Identification and sequencing of bacterial rDNAs in culture-negative amniotic fluid from women in premature labor. Am J Perinatol 21:319–323 Gardiner JP (1922) The umbilical cord: normal length; length in cord complications; etiology and frequency of coiling. Surg Gynecol Obstet 34:252–256 Genest DR (1992) Estimating the time of death in stillborn fetuses: II. Histologic evaluation of the placenta; a study of 71 stillborns. Obstet Gynecol 80:585–592 Genest DR, Williams MA et al (1992) Estimating the time of death in stillborn fetuses: I. Histologic evaluation of fetal organs; an autopsy study of 150 stillborns. Obstet Gynecol 80:575–584 Ghidini A, Spong CY (2001) Severe meconium aspiration syndrome is not caused by aspiration of meconium. Am J Obstet Gynecol 185:931–938 Giles WB, Trudinger BJ et al (1985) Fetal umbilical artery flow velocity waveforms and placental

1301 resistance: pathological correlation. Br J Obstet Gynaecol 92:31–38 Gomez R, Romero R et al (1998) The fetal inflammatory response syndrome. Am J Obstet Gynecol 179: 194–202 Greer IA (1999) Thrombosis in pregnancy: maternal and fetal issues. Lancet 353:1258–1265 Griffin AC, Strauss AW et al (2012) Mutations in longchain 3-hydroxyacyl coenzyme a dehydrogenase are associated with placental maternal floor infarction/ massive perivillous fibrin deposition. Pediatr Dev Pathol 15(5):368–374 Gupta N, Sebire NJ et al (2004) Massive perivillous fibrin deposition associated with discordant fetal growth in a dichorionic twin pregnancy. J Obstet Gynaecol 24:579–580 Han YW, Redline RW et al (2004) Fusobacterium nucleatum induces premature and term stillbirths in pregnant mice: implication of oral bacteria in preterm birth. Infect Immun 72:2272–2279 Hansen AR, Collins MH et al (2000a) Very low birthweight infant’s placenta and its relation to pregnancy and fetal characteristics. Pediatr Dev Pathol 3:419–430 Hansen AR, Collins MH et al (2000b) Very low birthweight placenta; clustering of morphologic characteristics. Pediatr Dev Pathol 3:431–438 Hashimoto K, Clapp JF (2003) The effect of nuchal cord on amniotic fluid and cord blood erythropoietin at delivery. J Soc Gynecol Investig 10:406–411 Hecht JL, Allred EN et al (2008) Histologic characteristics of singleton placentas delivered before the 28th week of gestation. Pathology 40:372–376 Heller DS, Rush D et al (2003) Subchorionic hematoma associated with thrombophilia: report of three cases. Pediatr Dev Pathol 6:261–264 Hermansen MC (2001) Nucleated red blood cells in the fetus and newborn. Arch Dis Child Fetal Neonatal Ed 84:F211–F215 Hershkovitz R, Silberstein T et al (2001) Risk factors associated with true knots of the umbilical cord. Eur J Obstet Gyneco Biol 93:36–39 Hill GB (1998) Preterm birth: associations with genital and possibly oral microflora. Ann Periodontal 3:222–232 Hung NA, Jackson C et al (2005) Pregnancy-related polymyositis and massive perivillous fibrin deposition in the placenta: are they pathogenetically related? Arthr Rhematism 55:154–156 Huppertz B, Kadyrov M et al (2006) Apoptosis and its role in the trophoblast. Am J Obstet Gynecol 195:29–39 Ibdah JA, Bennett MJ et al (1999) A fetal fatty acid oxidation disorder as a cause of liver disease in pregnant women. N Engl J Med 340:1723–1731 Jacques SM, Qureshi F (1998) Chronic chorioamnionitis: a clinicopathologic and immunohistochemical study. Hum Pathol 29:1457–1461

1302 Jacques SM, Qureshi F et al (2011) Eosinoophilia/T-cell chorionic vasculitis: a Clinicopathologic and immunohistochemical study of 51 cases. Pediatr Dev Pathol 14(3):198–205 Jamshed S, Kouides P et al (2007) Pathology of thrombotic thrombocytopenic purpura in the placenta, with emphasis on the “snowman sign”. Pediatr Dev Pathol 10:455–462 Jauniaux E, Nicolaides KH et al (1997) Perinatal features associated with placental mesenchymal dysplasia. Placenta 18:701–706 Kalousek DK, Bamforth S (1988) Amnion rupture sequence in previable fetuses. Am J Med Genet 31:63–73 Kalousek DK, Barrett I (1994) Confined placental mosaicism and stillbirth. Pediatr Pathol 14:151–159 Kalousek DK, Fitch N et al (1990) Pathology of the human embryo and previable fetus. An atlas. Springer, New York Kaplan C, Blanc WA et al (1982) Identification of erythrocytes in intervillous thrombi. A study using immunoperoxidase identification of hemoglobins. Hum Pathol 13:554–557 Kaplan CG, Covinsky MH et al (2016) Letter to the editor. Pediatr Dev Pathol 19:258 Katz VL, DiTomasso J et al (2002) Activated protein C resistance associated with maternal floor infarction treated with low-molecular-weight heparin. Am J Perinatol 19:273–277 Katzman PJ, Genest DR (2002) Maternal floor infarction and massive perivillous fibrin deposition: histological definitions, association with intrauterine fetal growth restriction, and risk of recurrence. Pediatr Dev Pathol 5:159–164 Keenan WJ, Steichen JJ et al (1977) Placental pathology compared with clinical outcome. Am J Dis Child 131:1224–1227 Keogh JM, Badawi N (2006) The origins of cerebral palsy. Curr Opin Neurol 19:129–134 Khong TY (1999) The placenta in maternal hyperhomocysteinaemia. Br J Obstet Gynaecol 106:2733–2738 Khong TY, Mooney EE et al (2016) Sampling and definitions of placenta lesion; Amsterdam placental workshop group consensus statement. Arch Pathol Lab Med 140:698–713 Kim CJ, Yoon BH et al (2001) Umbilical arteritis and phlebitis mark different stages of the fetal inflammatory response. Am J Obstet Gynecol 185(2):496–500 King EL, Redline RW et al (2004) Myocytes of chorionic vessels from placentas with meconium-associated vascular necrosis exhibit apoptotic markers. Hum Pathol 35(4):412–417 Kontopoulos E, Chmait RH et al (2016) Twin-to-twin transfusion syndrome: definition, staging and ultrasound assessment. Twin Res Hum Genet 19(3):175–183 Kraus FT, Acheen VI (1999) Fetal thrombotic vasculopathy in the placenta: cerebral thrombi and infarcts, coagulopathies, and cerebral palsy. Hum Pathol 30:759–769

R. N. Baergen et al. Kraus FT, Redline RW et al (2004) Placental pathology. ARP/AFIP, Washington, DC Kumar D, Fung W et al (2006) Proinflammatory cytokines found in amniotic fluid induce collagen remodeling, apoptosis, and biophysical weakening of cultured human fetal membranes. Biol Rep 74 (1):29–34 Lash GB, Quenby S et al (2008) Gestational diseases – a workshop report. Placenta 22:S92–S94 Leistra-Leistra MJ, Timmer A et al (2004) Fetal thrombotic vasculopathy in the placenta: a thrombophilic connection between pregnancy complications and neonatal thrombosis? Placenta 25(Suppl A): S102–S105 Leviton A, Paneth N et al (1999) Maternal infection, fetal inflammatory response, and brain damage in very low birth weight infants. Developmental epidemiology network investigators. Pediatr Res 46(5):566–575 Lim LK, Zhou Y et al (1997) Human cytotrophoblast differentiation/invasion is abnormal in pre-eclampsia. Am J Pathol 151:1809–1818 Lim FT, Scherjon SA et al (2000) Association of stress during delivery with increased numbers of nucleated cells and hematologic progenitor cells in umbilical cord blood. Am J Obstet Gynecol 183:1144–1151 Lin MG (2006) Umbilical cord prolapse. Obstet Gynecol Surv 61(4):269–277 Loukeris K, Baergen RN (2010) Syncytial knots as a reflection of placental maturity: reference values for 20 to 40 weeks gestational age. Pediatr Dev Pathol 13(4):305–309 Machin GA, Ackerman J et al (2000) Abnormal umbilical cord coiling is associated with adverse perinatal outcomes. Pediatr Dev Pathol 3:462–471 Magid MS, Kaplan C et al (1998) Placental pathology in systemic lupus erythematosus: a prospective study. Am J Obstet Gynecol 179:226–236 Many A, Schreiber L et al (2001) Pathologic features of the placenta in women with severe pregnancy complications and thrombophilia. Obstet Gynecol 98:1041–1044 Many A, Elad R et al (2002) Third trimester unexplained intrauterine fetal death is associated with inherited thrombophilia. Obstet Gynecol 99:684–687 Matern D, Shehata BM et al (2001) Placental floor infarction complicating the pregnancy of a fetus with longchain 3-hydroxyacyl-CoA dehydrogenase (LCHAD) deficiency. Mol Genet Metab 72:265–268 McDonald DG, Kelehan P et al (2004) Placental fetal thrombotic vasculopathy is associated with neonatal encephalopathy. Hum Pathol 35:875–880 Menezo YJ, Sakkas D (2002) Monozygotic twining: is it related to apoptosis in the embryo? Hum Reprod 17:247–248 Miller PW, Coen RW et al (1985) Dating the time interval from meconium passage to birth. Obstet Gynecol 66:459–462 Mitra SC, Seshan SV et al (2000) Placental vessel morphometry in growth retardation and increased resistance of the umbilical artery Doppler flow. J Matern Fetal Med 9:282–286

19

Diseases of the Placenta

Moscoso G, Jauniaux E et al (1991) Placental vascular anomaly with diffuse mesenchymal stem villous hyperplasia. A new clinico-pathological entity? Pathol Res Pract 187:324–328 Mutter WP, Karumanchi SA (2008) Molecular mechanisms of preeclampsia. Microvasc Res 75:1–8 Myerson D, Parkin RK (2006) The pathogenesis of villitis of unknown etiology: analysis with a new conjoint immunohistochemistry-in situ hybridization procedure to identify specific maternal and fetal cells. Pediatr Dev Pathol 9(4):257–265 Naeye RL (1985) Umbilical cord length: clinical significance. J Pediatr 107:278–281 Nelson KB, Grether KJ (1998) Potentially asphyxiating conditions and spastic cerebral palsy in infants of normal birth weight. Am J Obstet Gynecol 179:507–513 Nelson KB, Dambrosia JM et al (1998) Neonatal cytokines and coagulation factors in children with cerebral palsy. Ann Neurol 44:665–675 Nelson KB, Dambrosia JM et al (2005) Genetic polymorphisms and cerebral palsy in very preterm infants. Pediatr Res 57:494–499 Ogino S, Redline RW (2000) Villous capillary lesions of the placenta: distinctions between chorangioma, chorangiomatosis, and chorangiosis. Hum Pathol 31:945–954 Ohyama M, Itani Yet al (2002) Re-evaluation of chorioamnionitis and funisitis with a special reference to subacute chorioamnionitis. Hum Pathol 33:183–190 Ordi J, Ismail MR et al (1998) Massive chronic intervillositis of the placenta associated with malaria infection. Am J Surg Pathol 22:1006–1011 Parast MM, Crum CP et al (2008) Placental histological criteria for umbilical flow restriction in unexplained stillbirth. Hum Pathol 39:948–953 Patel D, Dawson M et al (1989) Umbilical cord circumference at birth. Am J Dis Child 143:638–639 Peng HW, Levitin-Smith M et al (2006) Umbilical cord stricture and overcoiling are common causes of fetal demise. Pediatr Dev Pathol 9:14–19 Pham T, Steele J et al (2006) Placental mesenchymal dysplasia is associated with high rates of intrauterine growth restriction and fetal demise. A report of 11 new cases and review of the literature. Am J Clin Pathol 126:67–78 Pharaoh P, Adi Y (2000) Consequences of in-utero death in twin pregnancy. Lancet 355:1597–1602 Platt DJ, Miner JJ (2017) Consequences of congenital Zika virus infection. Curr Opin Virol 27:1–7 Qureshi F, Jacques SM (1994) Marked segmental thinning of the umbilical cord vessels. Arch Pathol Lab Med 118:826–830 Raio L, Ghezzi F (1998) The clinical significance of antenatal detection of discordant umbilical arteries. Obstet Gynecol 91:86–91 Raio L, Ghezzi F et al (1999) Prenatal diagnosis of a lean umbilical cord: a simple marker for the fetus at risk of being small for gestational age at birth. Ultrasound Obstet Gynecol 13:176–180 Rakheja D, Bennett MJ et al (2002) Long-chain L-3hydroxyacyl-coenzyme a dehydrogenase deficiency: a

1303 molecular and biochemical review. Lab Investig 82:815–824 Rand JH (2000) Antiphospholipid antibody-mediated disruption of the annexin-V antithrombotic shield: a thrombogenic mechanism for the antiphospholipid syndrome. J Autoimmun 15:107–111 Rayne SC, Kraus FT (1993) Placental thrombi and other vascular lesions. Classification, morphology, and clinical correlations. Pathol Res Pract 189:2–17 Redline RW (1988) Specific defects in the anti-listerial immune response in discrete regions of the murine uterus and placenta account for susceptibility to infection. J Immunol 140:3947–3955 Redline RW (2002) Clinically and biologically relevant patterns of placental inflammation. Pediatr Dev Pathol 5:326–328 Redline RW (2004a) Clinical and pathological umbilical cord abnormalities in fetal thrombotic vasculopathy. Hum Pathol 35:1494–1498 Redline RW (2004b) Placental inflammation. Semin Neonatol 9:265–274 Redline RW (2005) Severe fetal placental vascular lesions in term infants with neurologic impairment. Am J Obstet Gynecol 192:452–457 Redline RW (2006a) Inflammatory responses in the placenta and umbilical cord. Semin Fetal Neonatal Med 11:296–301 Redline RW (2006b) Thrombophilia and placental pathology. Clin Obstet Gynecol 49(4):885–894 Redline RW (2006c) Placental pathology and cerebral palsy. Clin Perinatol 33:503–516 Redline RW (2007a) Infections and other inflammatory conditions. Semin Diagn Pathol 24:5–13 Redline RW (2007b) Villitis of unknown etiology: noninfectious chronic villitis in the placenta. Hum Pathol 38(10):1439–1446 Redline RW (2008) Elevated circulating fetal nucleated red blood cells and placental pathology in term infants who develop cerebral palsy. Hum Pathol 39:1378–1384 Redline RW, O’Riordan MA (2000) Placental lesions associated with cerebral palsy and neurologic impairment following term birth. Arch Pathol Lab Med 124:1785–1791 Redline RW, Pappin A (1995) Fetal thrombotic vasculopathy: the clinical significance of extensive avascular villi. Hum Pathol 26:80–85 Redline RW, Patterson P (1994) Patterns of placental injury: correlation with gestational age, placental weight, and clinical diagnosis. Arch Pathol Lab Med 118:698–701 Redline RW, Patterson P (1995) Preeclampsia is associated with an excess of proliferative immature intermediate trophoblast. Hum Pathol 26:594–600 Redline RW, Hassold T et al (1998a) Determinants of villous trophoblastic hyperplasia in spontaneous abortions. Mod Pathol 11(8):762–768 Redline RW, Wilson-Costello D et al (1998b) Placental lesions associated with neurologic impairment and cerebral palsy in very low birth weight infants. Arch Pathol Lab Med 122:1091–1098

1304 Redline RW, Wilson-Costello D et al (1999) Chronic peripheral separation of placenta. The significance of diffuse chorioamniotic hemosiderosis. Am J Clin Pathol 111:804–810 Redline RW, Dinesh S et al (2001) Placental lesions associated with abnormal growth in twins. Pediatr Dev Pathol 4:473–481 Redline RW, Jiang JG et al (2003) Discordancy for maternal floor infarction in dizygotic twin placentas. Hum Pathol 34:822–824 Redline RW, Ariel IB et al (2004a) Fetal vascular obstructive lesions: nosology and reproducibility of placental reaction patterns. Pediatr Dev Pathol 7:443–452 Redline RW, Boyd T et al (2004b) Maternal vascular underperfusion: nosology and reproducibility of placental reaction patterns. Pediatr Dev Pathol 7:237–249 Redline RW, Minich N et al (2007) Placental lesions as predictors of cerebral palsy and abnormal neurocognitive function in extremely low birth weight infants (39 Abortion

2

4 5 cm

Spleen, kidney 1–4

GI tract, liver 4–8

4

Term

Single drug

>12 >105

Brain >8

Two or more drugs

hCG, human chorionic gonadotropin; GI tract, gastrointestinal tract a The total score for a patient is obtained by adding the individual scores for each prognostic factor. Total score 8, high risk b Interval time (months) between end of antecedent pregnancy and start of chemotherapy

rate compared with involvement of other visceral organs. Development of CNS metastases during treatment confers an even worse prognosis. Patients with hepatic metastases also have a poor prognosis, but multiagent chemotherapy appears to increase the survival rate. Choriocarcinoma diagnosed after a term gestation has a worse prognosis than choriocarcinoma diagnosed after a hydatidiform mole.

Pathogenesis of GTD The pathogenesis of gestational trophoblastic disease (GTD) is largely unknown as few molecular studies have been performed (Shih and Kurman 2002). This is in part due to the relative rarity of GTD and the lack of appropriate experimental

1323

models. The most well-studied gestational trophoblastic lesion is hydatidiform mole and, to a lesser extent, choriocarcinoma (Li et al. 2002). Development of a hydatidiform mole appears to be associated with an excess of paternal haploid set of chromosomes. The higher the ratio of paternal/ maternal chromosomes, the greater the molar change. CHMs show a 2:0 paternal/maternal ratio, whereas PHMs show a 2:1 ratio. This hypothesis is best supported by the following two studies. First, a mouse model in which molecularly engineered mice that were either androgenetic or gynogenetic (parthenogenetic) were created by microtransfer of male or female pronuclei into enucleated oocytes. Androgenetic embryos transplanted to a foster mother developed a bulky, hypertrophic placenta similar to CHMs in humans, whereas the gynogenetic embryos developed only a small placenta with a secondarily stunted embryo. Another study using the mouse embryonic fibroblasts as the cell model demonstrates that paternal and maternal genomes confer opposite effects on proliferation, cell-cycle progression, senescence, and tumor formation (Hernandez et al. 2003). In comparison with biparental fibroblasts, androgenetic fibroblasts whose genomic complement is exclusively from paternal origin exhibit highly proliferative activity, spontaneous transformation, and formation of tumors at low passage number, while the gynogenetic fibroblasts with exclusive maternal genomic complement show decreased proliferation and increased senescence. The molecular mechanism underlying this observation is the differential expression of imprinted genes. For example, the maternally expressed and paternally imprinted genes such as p57kip2 decrease cellular proliferation. In contrast, the paternally expressed growth factor, Igf2, is essential for the long-term proliferation of all genotypes (Hernandez et al. 2003). Therefore, it is likely that a similar mechanism is also involved in the pathogenesis of complete and PHMs. In rare familial/recurrent hydatidiform moles which account for 2% of all molar cases, mutations have been detected in the maternal gene, NALP7 which plays a role in inflammation and apoptosis (Murdoch et al. 2006; Slim and Mehio 2007).

1324

I.-M. Shih et al.

As previously discussed (Table 1), there are at least three distinctive types of gestational trophoblastic neoplasia (GTN) including the most common type, choriocarcinoma, and the less common ones, PSTT and ETT. Molecular analysis on GTN is largely based on the identification and characterization of trophoblastic markers in various types of GTN and the reference of their unique gene expression patterns to different trophoblastic subpopulations in normal early placentas (reviewed in reference of (Shih 2007a)). The main conclusion from those studies is that following neoplastic transformation of trophoblastic stem cells, presumably the CT, specific differentiation programs dictate the type of trophoblastic tumor that develops (Fig. 8). These patterns of differentiation in GTN recapitulate the stages of early placental development. For example, choriocarcinoma is composed of variable amounts of neoplastic CT, ST, and intermediate (extravillous) trophoblast and resembles the previllous blastocyst which is composed of a similar mixture of trophoblastic subpopulations. On the other hand, the neoplastic CT in PSTT differentiates mainly into intermediate (extravillous) trophoblastic cells in an implantation site, whereas the neoplastic CT in ETT differentiates into chorionic-type intermediate (extravillous) trophoblastic cells in the chorion laeve. This model suggests that choriocarcinoma is the most primitive trophoblastic tumor whereas PSTT and ETT are relatively more differentiated. Furthermore, it explains the

Fig. 8 Trophoblastic neoplasms can be related to the different subpopulations of intermediate trophoblastic cells. Exaggerated placental site and PSTT are related to the differentiation of implantation site IT whereas placental site nodule and ETT are related to chorionic-type intermediate trophoblast

histologic mixture of choriocarcinoma and PSTT and/or ETT in some GTNs.

Clinicopathological Features, Behavior, and Treatment of Hydatidiform Moles A hydatidiform mole is an abnormal placenta characterized by enlarged, edematous, and vesicular chorionic villi accompanied by villous trophoblastic hyperplasia. There are two types of hydatidiform mole, the CHM and the PHM, which are genetically distinct, as described above. While CHMs and PHMs are usually readily distinguished on histologic examination when morphologic features are well developed, studies have demonstrated that distinction of hydatidiform moles from non-molar specimens and subtyping of hydatidiform moles are subject to diagnostic variability (Vang et al. 2012; Gupta et al. 2012). The routine use of ultrasound evaluation in pregnancy has made diagnosis more challenging because hydatidiform moles are now diagnosed and evacuated much earlier in gestation, often in the first trimester. As a result, the classical features of complete and PHMs, which in the past were based on examination of specimens obtained in the second trimester, are not as apparent, making the histopathologic diagnosis more difficult (Garner et al. 2007). In addition, a variety of other genetic abnormalities such as trisomy and

Cytotrophoblast Syncytiotrophoblast Normal trophoblast EVT (IT) at chorion laeve

EVT (IT) at implantation site Neoplastic transformation

Choriocarcinoma Trophoblastic neoplasia

PSTT

ETT

20

Gestational Trophoblastic Tumors and Related Tumorlike Lesions

monosomy may be associated with abnormal placentas that display minor degrees of hydropic change and trophoblastic proliferation but are not hydatidiform moles. Recent studies have demonstrated the utility of ancillary diagnostic techniques for refined diagnosis of molar specimens (Banet et al. 2013; Bifulco et al. 2008; McConnell et al. 2009a). The ratio of CHMs versus PHMs varies among studies and most reports show that CHMs outnumber PHMs (Czernobilsky et al. 1982; Bifulco et al. 2008; McConnell et al. 2009a; Olsen et al. 1999; Jacobs et al. 1982; Szulman and Surti 1982; Golfier et al. 2007). Clinical, pathologic, and cytogenetic differences separate the two forms of hydatidiform mole, yet all molar pregnancies have the potential for persistent GTD (Altieri et al. 2003), although the risk is much higher for CHM than PHMs.

CHM CHMs are characterized by hydropic swelling of most villi and a variable degree of trophoblastic proliferation and atypia. Fetal tissue usually is not present. CHMs are purely androgenetic conceptions (~85% homozygous/monospermic) and most often diploid (46,XX) but some are tetraploid (Banet et al. 2013).

Clinical Features As discussed above, CHMs are being diagnosed now at an earlier gestational age than in the past (8.5–12 weeks vs. 16–18 weeks) because of the routine use of sonography in pregnancy. Pelvic ultrasonic examination discloses a diagnostic snowstorm pattern but high-resolution sonography typically reveals a complex intrauterine mass with many small cysts. This pattern, especially when associated with a markedly elevated beta-hCG level, is clinically diagnostic of a molar pregnancy. Consequently, a CHM now rarely presents with the classic signs and symptoms such as excessive uterine size, hyperemesis, theca lutein ovarian cysts, hyperthyroidism, or preeclampsia. Most patients present with vaginal bleeding or are discovered by sonography. While a presumptive diagnosis of an incomplete or missed abortion is usually made,

1325

serum beta-hCG levels greater than 100,000 mIU/mL should prompt the physician to consider the diagnosis of a molar pregnancy. In the past, excessive uterine enlargement for the gestational age occurred in approximately two-thirds of patients. Occasionally, the initial clinical manifestation is sudden passage of molar vesicles. Preeclampsia (pregnancy-induced hypertension with edema and proteinuria) occurs in up to one-fourth of patients with CHM. In contrast to non-molar gestations in which preeclampsia occurs typically in the last trimester, in molar gestations, preeclampsia occurs in the first trimester. Thus, early onset of preeclampsia, especially when coupled with excessive uterine enlargement, suggests the presence of a molar pregnancy. Additional clinical signs of established molar pregnancy include hyperemesis gravidarum, hyperthyroidism, pulmonary embolization of trophoblast, and massive ovarian enlargement due to benign theca lutein cysts (hyperreactio luteinalis) (Garner et al. 2007). Usually with CHMs the beta-hCG titer is markedly elevated. Although these clinical signs and symptoms permit the diagnosis of a molar pregnancy before evacuation, the clinical presentation is quite variable. Up to 80% of cases are first diagnosed by histological study of spontaneously passed or curetted tissue. Hydatidiform moles also can be found unexpectedly in elective abortion specimens of asymptomatic patients and may be rarely detected in the fallopian tube and ovary. A primary mole involving the adnexa should be discriminated from the hydropic change that is frequent in an aborting ectopic pregnancy and from invasive mole with extension into the broad ligament.

Gross Findings In typical cases, massively enlarged, edematous villi give the characteristic grape-like appearance to the placenta (Fig. 9). However, the specimen volume of contemporary CHMs is significantly less than in the past and in very early CHMs grossly hydropic change may be absent. In advanced hydatidiform mole, which is now rarely encountered, swollen villi may range from a few millimeters to as large as 3.0 cm in diameter but usually average about 1.5 cm. Rarely, fetal

1326

Fig. 9 CHM. (a) A hysterectomy specimen shows an enlarged uterus with molar tissue protruding into the uterine cavity. (Reprint from reference of (Shih 2009) with permission from Churchill Lvingstone Elsevier) (b) In a well-developed complete mole, the hydropic villi range from a few millimeters to more than 1 cm in diameter

development may occur in CHM. After suction curettage, molar villi may collapse, and a large amount of bloody tissue may obscure the edematous villi, especially if a hydatidiform mole is extracted early in pregnancy when villous enlargement is less striking. In this instance, there may be no gross evidence of molar change. Histological evaluation of the tissue adherent to the gauze that collects suctioned uterine contents is necessary to establish the diagnosis. Immersing the gross tissue in saline or formalin can resuspend collapsed villi.

Microscopic Findings CHMs have two key features: trophoblastic proliferation and villous edema. Many villi display

I.-M. Shih et al.

central cistern formation characterized by a prominent central space that is entirely acellular (Fig. 10). Smaller villi usually are present but these, too, are edematous. The villous stroma has a distinctive appearance with a pale bluegray appearance having widely separated spindle cells beneath the villous surface, and edematous central cistern. Villous stroma of a CHM contains numerous but inconspicuous CD34 positive blood vessels. All hydatidiform moles display some degree of haphazard trophoblastic proliferation on the villous surface. This trophoblastic proliferation in CHMs is circumferential around the villus (Fig. 11). Columns and streamers of cells composed of a mixture of CT, ST, and villous IT project randomly from the villous surface (Fig. 11). The amount of proliferative trophoblast in CHMs varies greatly. It may be marked, affecting most villi, or it may be subtle and only focally present, emphasizing the need for thorough sampling. Large sheets of trophoblast including villous IT that appear to be unattached to villi also may be present. These result from tangential sectioning, or they represent detached fragments of the trophoblast from the implantation site. Trophoblastic islands, structures seen in the normal early placentas, are rarely present in CHMs. The trophoblast of a CHM always displays cytologic atypia. At times it may be as marked as in choriocarcinoma (Fig. 11, see below). In addition, a linear distribution of atypical intermediate trophoblast overlying fibrin is a feature that occurs in the implantation site of molar pregnancies as opposed to a non-molar abortus. As noted previously, the typical morphologic features of a CHM may be absent or very subtle in molar specimens evacuated in the first trimester (Fig. 12). These CHMs are referred to as “early CHMs” or “very early CHMs”; they do not have the typical clinical presentation or classic sonographic appearance. The histologic features of an early CHM are: (1) redundant or polypoid bulbous terminal villi; (2) hypercellular villous stroma with primitive stellate cells; (3) a labyrinthine network of villous stromal canaliculi; (4) focal

20

Gestational Trophoblastic Tumors and Related Tumorlike Lesions

1327

Fig. 10 (a, b) CHM. Hydropically enlarged villi have circumferential trophoblastic hyperplasia, cisterns, and trophoblastic inclusions. (c) CHM. Villous CT and stromal

cells are negative for p57 (with internal positive control in intermediate trophoblastic cells) and genotyping demonstrated a purely androgenetic conception

Fig. 11 (a) CHM. Florid circumferential trophoblastic hyperplasia surrounds hydropically enlarged molar villus.

(b) CHM. Hyperplastic molar trophoblast is an admixture of CT, ST, and IT

hyperplasia of CT and ST on both villi and the undersurface of the chorionic plate; (5) atypical trophoblast lining the villi and in the implantation site (Keep et al. 1996; Mosher et al.

1998); and (6) increased villous stromal apoptosis (Kim et al. 2006). The implantation sites associated with CHMs almost always show features of an exaggerated

1328

I.-M. Shih et al.

Fig. 12 (a, b) Early CHM. Bulbous “cauliflower-like” villous structures have some circumferential trophoblastic hyperplasia, slightly cellular myxoid stroma, small canalicular vascular structures, and karyorrhectic nuclear debris

in the stroma. (c) Early CHM. Villous CT and stromal cells are negative for p57 (with internal positive control in intermediate trophoblastic cells) and genotyping demonstrated a purely androgenetic conception

placental site, with the intermediate trophoblastic cells having more atypical features than encountered in non-molar exaggerated placental sites (Shih and Kurman 1998a) (Figs. 13 and 14). Molar and non-molar exaggerated placental sites share expression of markers of implantation site IT, but molar-type exaggerated placental sites have some variably increased proliferative activity as assessed with Ki-67, whereas non-molar exaggerated placental sites have no proliferative activity (Fig. 14c) (Shih and Kurman 1998a; Montes et al. 1996). Interestingly, as compared with the normal placentas, CHMs exhibit a higher level of apoptosis in CT, indicating a complex but delicate regulation of this cell population in CHMs.

exaggeration of a physiologic process occurring in normal pregnancy. The most serious complication after molar evacuation is persistent or metastatic GTD and the risk of developing choriocarcinoma. Postmolar trophoblastic disease may represent a persistent hydatidiform mole in the uterine cavity or it may be an invasive hydatidiform mole or a choriocarcinoma. Persistent GTD occurs in approximately 17–20% of women who have undergone evacuation (Seckl et al. 2000) and in 3–5% of women who have undergone hysterectomy (Genest 2001). The risk of development of choriocarcinoma following a CHM is about 2–5% in the United States (Elston 1995; Coukos et al. 1999) and 13% in Japan (Matsui et al. 1996). Prediction of the behavior of CHMs based on biomarkers is notoriously difficult and the capricious nature of molar pregnancies should never be underestimated. Some studies have found that higher mRNA and protein expression levels of Nanog are associated with worse clinical outcome in hydatidiform moles (increased the risk of developing persistent GTD), but other studies have failed to identify reliable clinical or morphologic features or immunohistochemical markers for predicting prognosis (Siu et al. 2008; Shih and Kuo 2008). Thus, follow-up with serial

Differential Diagnosis See below section on PHM.

Behavior and Treatment In advanced cases patients may present with severe respiratory distress immediately after uterine evacuation of a hydatidiform mole. This phenomenon usually is attributed to massive deportation of trophoblast to the lungs, an

20

Gestational Trophoblastic Tumors and Related Tumorlike Lesions

1329

plateauing beta-hCG titers for 2–4 weeks, rising beta-hCG titers, persistent uterine disease such as abnormal bleeding, or evidence of metastases. Initial therapy for persistent disease is usually methotrexate, dactinomycin, or a combination of the two.

PHM A PHM has an intimate admixture of two populations of villi: enlarged, edematous villi and normal-sized villi which may be fibrotic. PHMs may have a grossly identifiable embryo or fetus with congenital anomalies. PHMs are almost always diandric triploid conceptions (~99% heterozygous/dispermic), but rare triandric tetraploid forms exist (Banet et al. 2013).

Fig. 13 Molar-type exaggerated placental (implantation) site (associated with early CHM [not shown]). Atypical, hyperchromatic intermediate trophoblastic cells infiltrate the endomyometrium and a vascular structure in which the muscular wall is replaced with fibrinoid material

serum beta-hCG titers using sensitive assays and imaging analysis remains the mainstay in managing patients with CHM. After treatment with chemotherapy there is minimal increase in the risk of spontaneous abortion or congenital anomalies and many patients have successful term pregnancies. In general, patients with molar pregnancy, after appropriate treatment, can be reassured that they can anticipate a normal future reproductive outcome (Berkowitz et al. 1998). A chest radiograph before treatment and 4 weeks later should be performed to exclude metastases. Following evacuation of a hydatidiform mole, careful beta-hCG monitoring is mandatory since it is the most reliable and sensitive method for the early detection of persistent GTD. The titers of beta-hCG should fall to normal between 10 and 170 days after evacuation of a hydatidiform mole, and most patients will have normal titers by 60 days post-evacuation. Persistent GTD after molar pregnancy is heralded by

Clinical Features Patients with PHMs may have signs and symptoms like those seen in CHMs, but usually this is less likely. Uterine size is generally small for dates. Enlargement greater than that expected for the gestational age is unusual. More often patients appear to have a missed abortion, and vaginal bleeding is the main presenting symptom. Fortytwo percent of patients with PHMs are at risk of preeclampsia which tends to occur later than with CHMs but it can be equally severe (Jauniaux 1999). Serum beta-hCG levels often are in the low or normal range for gestational age. Only a few patients with PHMs show markedly elevated beta-hCG titers such as those seen with CHMs. Gross Findings The volume of tissue is generally small, less than 100 or 200 ml. The villi may be grossly evident and recognizable as molar, yet are smaller than those found in a CHM. For early PHMs, these gross features may not be apparent (Fig. 15). In some cases, a fetus or fetal membranes is present. When a fetus is found, it often shows gross congenital anomalies. Microscopic Findings PHMs show features in some villi that are similar to those seen in CHMs, but the molar change is focal (Fig. 16). There should be a mixture of

1330

I.-M. Shih et al.

Fig. 14 (a) Molar-type exaggerated placental site (associated with CHM) in hysterectomy specimen. Atypical molar trophoblast lines the surface of the implantation site (a) and IT associated with fibrinoid material infiltrates the endomyometrium and involves blood vessels

(b). (c) Molar-type exaggerated placental site. Double immunostain demonstrates that HLA-G-positive intermediate trophoblastic cells (red) have some Ki-67 proliferative activity (brown nuclei) [compare with figures for EPS non-molar]

Fig. 15 Partial hydatidiform mole. Hydropic villi mixed with smaller, “normal-appearing” villi

edematous villi and small, relatively normalsized villi. Central cisterns are less conspicuous than in CHMs. Smaller villi usually show stromal fibrosis like that seen in missed abortions (Fig. 17). Trophoblastic hyperplasia is less marked than in CHM. Generally, it is focal and shows little, if any, atypia, consisting of small, haphazard tufts of trophoblast, often ST emanating from the surface of some of the abnormal villi. Another feature commonly encountered in PHMs is a scalloped outline of the enlarged villi, yielding a pattern of trophoblastic invaginations into the villous stroma. When the invaginations do not show continuity with the surface trophoblast, they appear as inclusions within the stroma (Figs. 16 and 17). Invaginations are not exclusive for PHMs and may, on occasion, be found in other conditions including CHM and non-molar hydropic abortus. PHMs are usually associated with the presence of a fetus (usually abnormally formed) or its amnion, in contrast to the absence of fetal structures in most CHMs. Fetal demise with subsequent degeneration of fetal structures may make identification of fetal tissue difficult. A subtle clue is the presence of a functioning villous circulation containing nucleated red cells, a feature that

20

Gestational Trophoblastic Tumors and Related Tumorlike Lesions

1331

Fig. 16 (a, b) PHM. Irregularly shaped enlarged villi with scalloped contours, trophoblastic inclusions, and only mild trophoblastic hyperplasia are admixed with smaller villi.

(c) PHM. Villous CT and stromal cells are positive for p57 and genotyping demonstrated diandric triploidy

Fig. 17 (a, b) PHM. Irregularly shaped enlarged villi with scalloped contours, a few small trophoblastic inclusions, and some circumferential trophoblastic hyperplasia are

admixed with smaller villi. (c) PHM. Villous CT and stromal cells are positive for p57 and genotyping demonstrated diandric triploidy

requires fetal development. In contrast, the embryo associated with a CHM usually dies before organogenesis and, therefore, fetal structures are not present in the specimen and fetal erythrocytes are not present within placental vessels. One concern in the diagnosis of an apparent PHM is that the specimen represents a twin gestation with a non-molar fetus and a CHM. Such twin

pregnancies do occur (Lage et al. 1992) but are an infrequent occurrence relative to singleton gestations of a PHM. In summary, a PHM is optimally diagnosed when the following four microscopic features are present: (1) two populations of villi (one hydropic and one “normal”); (2) enlarged villi with central cavitation; (3) irregular villi with geographic, scalloped borders with

1332

I.-M. Shih et al.

trophoblast inclusions; and (4) minimal trophoblast hyperplasia (if present, usually focal and involving ST) (Genest 2001; Lee et al. 2003).

Differential Diagnosis Distinguishing features of CHMs and PHMs are presented in Table 6. The differential diagnosis of hydatidiform moles includes a variety of non-molar entities which can exhibit some features suggestive of a molar pregnancy. These include products of conception specimens with abnormal villous morphology, early non-molar specimens with prominent trophoblastic hyperplasia, hydropic abortuses, and androgenetic/ biparental mosaic conceptions. Abnormal villous morphology is a term used to label specimens in which villi have some dysmorphic features suggestive of a hydatidiform mole, usually a PHM but sometimes an early CHM, but lack fully developed diagnostic features of either type. In some cases, these changes are associated with other (non-molar) genetic abnormalities such as trisomy. PHMs usually display at least three of the following histologic features: two discrete populations of villi, circumferential mild Table 6 Pathologic features and behavior of complete and PHMs Feature Karyotype Embryo/fetus Villous outline Hydropic swelling

Trophoblastic proliferation Trophoblastic atypia Implantation site P57 (kip2) staining Behavior

Complete 46,XX, 46,XY Absent Rounded

Partial 69,XXY, 69,XXX Present Scalloped

Marked; cisterns present All villi involved Circumferential Variable, may be marked Often present

Less pronounced and focal Cisterns less prominent Villous fibrosis Focal and minimal

Exaggerated

Negative

Normal or occasionally exaggerated Positive

17–20% develop pGTD

35, hCG >1000 mIU/mL, depth of invasion (Piura et al. 2007; Baergen et al. 2006), and p53 immunoreactivity (Nagai et al. 2007). Some studies have shown that a high mitotic count is an adverse prognostic indicator (Hoekstra et al. 2004; Baergen et al. 2006). In our experience, malignant PSTTs as compared with benign cases generally are composed of larger masses and sheets of cells, many with clear instead of amphophilic cytoplasm. They have more extensive necrosis, prominent nuclear atypia, and higher mitotic activity (Fig. 35). Our findings suggest that the Ki-67 proliferation index may be a reliable indicator of prognosis as the index is usually higher than 50% in malignant tumors.

ETT The term “epithelioid trophoblastic tumor” (ETT) was introduced to describe an unusual type of

I.-M. Shih et al.

Fig. 35 Malignant PSTT. As in a benign PSTT, malignant PSTT grows confluently, and the cells are large with abundant eosinophilic cytoplasm. However, the tumor cells in a malignant PSTT exhibit a higher level of nuclear atypica

trophoblastic tumor that is distinct from PSTT and choriocarcinoma with features resembling a carcinoma. The tumor was originally termed “atypical choriocarcinoma” and was described in the lungs of patients with antecedent choriocarcinoma following intensive chemotherapy (Mazur 1989; Jones et al. 1993). Similar lesions referred to as “multiple nodules of IT” were subsequently reported in the uteri of patients following evacuation of hydatidiform moles (Silva et al. 1993). Subsequently, similar tumors were observed in the uterus without a history of antecedent GTD (Shih and Kurman 1998b). The recognition of ETT as a distinctive form of trophoblastic disease is in part due to its rarity and because many of the morphologic features of ETT are more reminiscent of a carcinoma than of a trophoblastic tumor (Shih and Kurman 1998b). The trophoblastic nature of this tumor has

20

Gestational Trophoblastic Tumors and Related Tumorlike Lesions

1353

been confirmed using molecular approaches (Oldt et al. 2002; Singer et al. 2002). Based on morphologic, ultrastructural, and immunohistochemical studies, ETT appears to develop from neoplastic transformation of cytotrophoblastic cells that differentiate toward chorionic-type intermediate trophoblastic cells (Shih 2007a)

Clinical Features Based on an analysis of 52 published cases, Palmer et al. reported that 67% of patients present with abnormal vaginal bleeding, 36% had evidence of antecedent molar pregnancy, and 35% presented with metastases (Palmer et al. 2008). Mean age at diagnosis is 38 years, and the mean interval between the preceding gestation and the diagnosis of ETT is 76 months. hCG levels are usually low (90%) in the viable epithelial cells. In contrast, the vascular/chorangiosis component is negative for cytokeratin and HSD3B1 but is positive for vimentin

of viable epithelial tumor cells. At low magnification, the necrotic areas may occupy most of the tumor. The epithelial cells are highly proliferative with frequent mitotic figures. They express lowmolecular-weight cytokeratin (such as cytokeratin18), beta-hCG, and HSD3B1, supporting their trophoblastic origin (Mao et al. 2008). The vascular component is similar to chorangiosis or a chorangioma with numerous intercalated vascular channels in the villous stroma. The pathogenesis in the development of

chorangiocarcinoma is undetermined. It has been proposed that the lesion may represent a chorangioma with associated trophoblastic hyperplasia, a true trophoblastic neoplasm with a reactive chorangiosis response, a reactive lesion of trophoblastic cells and villous vascular channels, or a collision tumor of a chorangioma and a choriocarcinoma. The clinical experience with chorangiocarcinoma is very limited, but the available data from the reported cases reveal that patients have an uneventful postpartum course

1368

I.-M. Shih et al.

and do not develop persistent GTD (Faes et al. 2012; Huang et al. 2015).

Immunohistochemistry Approach for Differential Diagnosis Diagnosis of trophoblastic tumors and tumorlike lesions is usually straightforward based on clinical and morphological features (Tables 7 and 8). However, the diagnosis can be challenging in certain cases, especially in small biopsy and curettage specimens (Narita et al. 2003). To assist in differential diagnosis, we have developed an algorithmic approach, termed “trophogram,” using a three-tiered stepwise immunohistochemical staining procedure (Shih 2007a; Mao et al. 2008) (Fig. 56). The antibodies utilized in the algorithm are all

commercially available and are listed in Table 9. The first tier in the algorithm discriminates a trophoblastic versus a non-trophoblastic lesion. The second tier determines if the lesion is related to implantation site IT (i.e., exaggerated placental site and PSTT), to chorionic-type intermediate trophoblastic cells (i.e., placental site nodule and ETT) or a choriocarcinoma. The third tier distinguishes a benign tumorlike lesion versus a trophoblastic neoplasm (Fig. 56). An example of how the algorithm can be used in the diagnosis of a PSTT is illustrated in Fig. 57. Using the algorithm, immunoreactivity of HSD3B1 and low-molecular-weight cytokeratin (e.g., cytokeratin 8 or 18) should first be assessed to determine if the lesion in question is trophoblastic. Diffuse and intense HSD3B1 and lowmolecular-weight cytokeratin immunoreactivity strongly suggests trophoblast origin (Mao et al.

Trophoblastic lesions? HSD3B +++ (diffuse) LMW cytokeratin +++ (diffuse)

Trophoblastic lesions

p63 – hPL +++

Lesions of implantation site intermediate trophoblast

Ki-67 < 1%

EPS

Ki-67 > 10%

PSTT

beta-hCG positive syncytiotrophoblast

p63 +++ hPL +/–

Lesions of chorionic type intermediate trophoblast

Choriocarcinoma

Ki-67 > 12% cyclin E ++

Ki-67 < 8% cyclin E– PSN

Fig. 56 Trophogram is an immunohistochemistry algorithm designed to assist differential diagnosis of trophoblastic tumors and tumorlike lesions. It is a three-tier sequential staining procedure that utilizes a panel of commercially available antibodies listed in Table 9. The first tier discriminates a trophoblastic versus a non-trophoblastic lesion. The second determines if the lesion is a choriocarcinoma, a

ETT

lesion related to implantation site IT or a lesion related to chorionic-type intermediate trophoblastic cells. The third tier distinguishes a benign tumorlike lesion versus a true trophoblastic neoplasm. EPS exaggerated placental site, PSTT placental site trophoblastic tumor, PSN placental site nodule, ETT epithelioid trophoblastic tumor

20

Gestational Trophoblastic Tumors and Related Tumorlike Lesions

1369

Table 9 Commercial antibodies used in the trophogram Marker HSD3B1 Cytokeratin p63 hPL β-hCG Cyclin E Ki-67

Antibody source Abnova (3C11-D4a) Ventana (CAM5.2) Neomarker (4A4) Dako (polyclonal) Dako (polyclonal) Zymed (HE12) Dako (MIB-1)

Application Distinguish GTD from most non-GTD Exclude mimickers of mesenchymal origin Distinguish ETT versus PSTT Distinguish PSTT versus ETT Highlight ST in choriocarcinoma Distinguish ETT versus PSN Distinguish ETT versus PSN; PSTT versus EPS

GTD gestational trophoblastic disease, ETT epithelioid trophoblastic tumor, PSTT placental site trophoblastic tumor, PSN placental site nodule, EPS exaggerated placental site, ST syncytiotrophoblast a Clone or name for the monoclonal antibody which may be obtained from other commercial sources

Trophoblastic lesion? HSD 3B +++ (diffuse) LMW cytokeratin +++ (diffuse)

Trophoblastic lesion p63 – hPL +++

Lesions of implantation site intermediate trophoblast Ki-67 < 1%

EPS

Ki-67 > 10%

beta-hCG positive syncytiotrophoblast

p63 +++ hPL –/+

Lesions of chorionic type intermediate trophoblast Choriocarcinoma

Ki-67 < 8% cyclin E –

PSTT

Fig. 57 An example how the algorithm is used to diagnose a PSTT. Hematoxylin-and-eosin section shows a tumor composed of large, atypical cells forming confluent masses and cords. The differential diagnosis includes PSTT, exaggerated placental site, ETT, smooth muscle neoplasm, and poorly differentiated carcinoma. Immunohistochemistry was performed to demonstrate that the tumor cells are diffusely positive for HSD3B1, a marker

PSN

Ki-67 > 12% cyclin E ++

ETT

specific to a trophoblastic origin. The trophoblastic lesion was then stained with hPL followed by Ki-67 and is diffusely positive for hPL, indicating that it is a lesion derived from the implantation site intermediate trophoblastic cells. The Ki-67 labeling index is approximately 15%, confirming its neoplastic nature. Based on the immunostaining findings, the tumor is compatible with a PSTT

1370

2008). It should be noted that HSD3B1 must be used in conjunction with low-molecular-weight cytokeratins because low-molecular-weight cytokeratins are expressed in a wide variety of adenocarcinomas. Thus, cytokeratin should be used to further support the diagnosis of a trophoblastic lesion in HSD3B1-positive cases. GATA3 can be included as part of an immunohistochemical panel particularly when the HSD3B1 antibody is either not available or yields ambiguous results because GATA-3 is frequently expressed in normal trophoblast and trophoblastic lesions (Banet et al. 2015). Once the diagnosis of a trophoblastic lesion is established, choriocarcinoma should be considered because it is the most common trophoblastic neoplasm. The presence of beta-hCG-positive ST supports the diagnosis of a choriocarcinoma. In choriocarcinomas where the syncytiotrophoblastic component is attenuated, beta-hCG positive ST tends to lie between or envelop masses of mononucleate trophoblastic cells. ST in choriocarcinoma should not be mistaken for beta-hCG positive multinucleated intermediate trophoblastic cells in PSTT and ETT. The multinucleated intermediate trophoblastic cells are usually polygonal or round with centrally located nuclei whereas ST assumes a labyrinthlike growth pattern with dark cytoplasm and elongated and evenly distributed nuclei. Furthermore, choriocarcinoma has very high proliferative activity and accordingly the Ki-67 proliferation index in mononucleate tumor cells in choriocarcinoma (Ki-67 index usually >40%) is significantly higher than that in most PSTTs and ETTs (Ki-67 index usually 50% of lesional cells) and p63 is either negative or very focally positive in PSTT and exaggerated placental site. In contrast, hPL is either negative or focally positive in ETTs and placental site nodules, but p63 is diffusely positive (usually >50% of lesional cells) (Shih and Kurman 2004; Lee et al. 2007; Zhang et al. 2009). Of note, the antibody to detect p63 should be the one that recognizes all the p63 isoforms. The Ki-67 proliferation index is then used to distinguish an ETT from a placental site nodule and a PSTT from an exaggerated placental site. The Ki-67 indices are near zero (10 cm and not infrequently >20 cm. Small tumors under 5 cm make up 5 cm in size (Carney et al. 1986). They contain abundant hyaluronic acid and a sparse to moderately cellular population of stellate-shaped and spindle cells with pale eosinophilic cytoplasm and mildly pleomorphic nuclei with “smudgy” chromatin and occasional cytoplasmic-nuclear invaginations (Figs. 5 and 6). Focal multinucleation is common. The myxoid matrix disrupts the mesenchymal

Fig. 5 Superficial angiomyxoma (cutaneous myxoma). Note the multinodular growth pattern, abundant myxoid matrix, and presence of paucicellular cleft-like spaces

element, forming paucicellular pools and cleftlike spaces, which oftentimes separate tumor nodules from native tissues. The degree of vascularity is variable, but large-caliber vessels as seen in aggressive angiomyxoma are typically absent. Scattered inflammatory cells (e.g., polymorphonuclear leukocytes, lymphocytes, and plasma cells) are a frequent finding. Adnexal epithelial inclusions, sometimes with cystic change and peripheral buds of basaloid epithelium, are occasionally encountered (Allen et al. 1988; Carney et al. 1986; Fetsch et al. 1997). The neoplastic cells are vimentin and CD34 positive (Fetsch et al. 1997). Some weak cytoplasmic reactivity for S100 protein can be seen, and there may be limited actin expression. This entity lacks nuclear reactivity for estrogen and progesterone receptor proteins (Fetsch et al. 1997) and HMGA2 (McCluggage et al. 2010). Negative results are also obtained for glial fibrillary acidic protein and desmin.

1412

J. F. Fetsch and W. B. Laskin

Fig. 6 Superficial angiomyxoma. (a) Tumor nodules often have peripheral cleft-like spaces filled with hyaluronic acid; (b) cystic epithelial inclusions may be

present; (c) tumor cells frequently have smudgy chromatin, and multinucleation and cytoplasmic-nuclear invaginations are common; and (d) tumor cells often express CD34

Differential Diagnosis Superficial angiomyxoma has similarity to aggressive angiomyxoma only in name. It is easily distinguished from the latter by its superficial location, small size, abundant alcian blue(pH 2.5) positive hyaluronic acid content, different tumor cell morphologies, and immunoprofile.

abnormalities), evaluation for the Carney complex is recommended.

Behavior and Treatment Superficial angiomyxoma has a documented recurrence rate of >30%, so complete excision with attention to margins is advisable. There are no reports of metastases. When this process is multifocal, encountered in childhood, or evident in a patient with other unusual clinical findings (e.g., prominent facial skin pigmentation, cardiac symptoms, or endocrine

Angiomyofibroblastoma Clinical Features Angiomyofibroblastoma is a benign soft tissue tumor that primarily occurs in the subcutaneous tissue of the vulva. Rare examples have involved the vagina, usually near the introitus (Fletcher et al. 1992; Fukunaga et al. 1997; Laskin et al. 1997; Magro et al. 2014). Affected individuals are frequently in the fourth through seventh decades of life, with a peak incidence in the fifth decade (Fukunaga et al. 1997; Laskin et al. 1997; Nielsen and Young 2001). Patients usually present with a

22

Soft Tissue Lesions Involving Female Reproductive Organs

1413

Fig. 7 Angiomyofibroblastoma. Medium- and high-power views of a conventional (a, b) and lipomatous (c, d) angiomyofibroblastoma. Note a tendency for epithelioid tumor cells to aggregate around the vasculature

relatively small, painless, subcutaneous mass. Rare examples may be pedunculated (Sims et al. 2012).

Pathologic Findings Gross examination generally reveals a welldemarcated mass under 5 cm in size with a soft to rubbery consistency, pinkish tan to yellow color, and mucoid or myxedematous cut surface. Areas grossly consistent with fat may be present. Microscopic features include relatively sharp margination, low to moderate cellularity, and abundant vasculature that consists predominantly of rather uniformly distributed small capillarysized vessels and veins, sometimes with focal hyalinization (Fig. 7a, b). The cellularity often varies from one region to the next, and some areas may contain abundant myxedematous,

loosely collagenized stroma. A polymorphic population of epithelioid, spindle, and multinucleated tumor cells with low-level nuclear atypia and a low mitotic index is typically present. Some epithelioid cells may have abundant eosinophilic cytoplasm, imparting a plasmacytoid appearance. There is a tendency for tumor cells to exhibit focal corded, nested, and trabecular growth patterns and for the cells to aggregate around the vasculature. Adipose tissue is present in >50% of the AFIP cases, and in a small percentage, this is a prominent or even dominant feature, as documented in the lipomatous variant of angiomyofibroblastoma (Fig. 7c, d) (Cao et al. 2005; Laskin et al. 1997). Large-caliber vessels and medium-sized nerve segments, as seen in aggressive angiomyxoma, are typically absent.

1414

J. F. Fetsch and W. B. Laskin

Fig. 8 Superficial (cervicovaginal/vulvovaginal) myofibroblastoma (SCVMF). Note the close proximity of the tumor to the mucosa (a). The lesional cells tend to be small and delicate and have epithelioid or spindle

morphology. Growth patterns vary from sievelike (b) to random, nonpatterned (c) to focally fascicular (d). Immunoreactivity for desmin is depicted in figure (e)

Angiomyofibroblastomas are immunoreactive for vimentin, and they usually express desmin (Fletcher et al. 1992; Fukunaga et al. 1997; Laskin et al. 1997; Nielsen et al. 1996a) and estrogen and progesterone receptor proteins (Laskin et al. 1997). There is variable, but generally minimal, reactivity for actins. CD34 expression is notably uncommon in conventional examples, but it is encountered with some frequency in the lipomatous variant (Cao et al. 2005; Laskin et al. 1997). Some vaginal tumors with both desmin and CD34 expression that have been reported as angiomyofibroblastomas are, in our opinion, more consistent with SCVMF (see Differential Diagnosis below) (Ganesan et al. 2005; Laskin et al. 2001; Nielsen et al. 1996a). S100 protein expression and nuclear immunoreactivity for HMGA2 are absent (Nucci et al. 2003; McCluggage et al. 2010).

Differential Diagnosis The differential diagnosis for angiomyofibroblastoma revolves primarily around aggressive angiomyxoma, cellular angiofibroma, SCVMF and mammary-type myofibroblastoma (MMF). Aggressive angiomyxomas typically present as large deep-seated masses with a “pushing” infiltrative border. They incorporate a variety of regional structures, including large vessels and nerves that are sometimes bordered by loosely organized collections of spindle myoid (smooth muscle-like) cells. The small neoplastic cells of aggressive angiomyxoma are generally more monomorphic and uniformly distributed than the cells of angiomyofibroblastoma. Immunostaining for nuclear HMGA2 expression and Fluorescence In Situ Hybridization (FISH) analysis for HMGA2 gene rearrangement can occasionally be helpful, as these are present in some cases of

22

Soft Tissue Lesions Involving Female Reproductive Organs

aggressive angiomyxoma but are absent in angiomyofibroblastomas (McCluggage et al. 2010). Although it has been stated in the literature that there are tumors with hybrid features of aggressive angiomyxoma and angiomyofibroblastoma (Granter et al. 1997), we have yet to encounter a resection specimen where distinction between the two was not possible, and molecular genetic analysis lends support to the view that these are two distinct biologic entities (Medeiros et al. 2007; Chen et al. 2012; Schoolmeester and Fritchie 2015). Angiomyofibroblastoma and cellular angiofibroma have some overlapping features, as both often present as relatively small superficial masses with high vascularity, and both may be dominated by spindle cells and contain fat. However, cellular angiofibroma lacks a prominent epithelioid tumor cell population with a tendency to aggregate around the vessels, it often has more notable perivascular hyalinization, and it can have more pronounced mitotic activity. Immunohistochemically, cellular angiofibroma is often CD34 positive, unlike conventional angiomyofibroblastoma, and it is much less likely to have desmin expression. Additionally, cellular angiofibroma frequently has monoallelic loss of RB1/13q14 with loss of RB1 immunoexpression, a feature shared with SCVMF, MMF, and spindle cell lipoma, but one that is not present in angiomyofibroblastoma (Chen et al. 2012; Flucke et al. 2011; Maggiani et al. 2007; Magro et al. 2012b, 2014; Schoolmeester and Fritchie 2015). SCVMF of the lower female genital tract can affect the uterine cervix, vagina, or vulva (Ganesan et al. 2005; Laskin et al. 2001; Magro et al. 2012a). An infrequent association with tamoxifen exposure has been noted (Ganesan et al. 2005). This process is usually more superficially located than angiomyofibroblastoma, and it is commonly excised with the overlying mucosa or skin (Fig. 8a). It is generally moderately cellular and features a proliferation of spindle, stellate-shaped, and small epithelioid cells with relatively scant cytoplasm. These cells are arranged in reticular (or sievelike), random, vague storiform and short fascicular growth patterns within finely

1415

collagenous, myxedematous, or, occasionally, hyalinized matrix (Fig. 8b–d). Focal multinucleation and mild atypia may be present, but mitotic activity is characteristically low. The lesion cells are estrogen and progesterone receptor positive, and they commonly co-express desmin and CD34. Focal smooth muscle actin expression is detected in a minority of cases. This tumor also shows loss of RB1 immunoexpression. MMF, an additional member of the RB1-deficient tumor family, features spindle cells admixed with thick ropy collagen bundles, hyalinized collagen, or myxoid matrix. Infrequent tumor cells may have epithelioid morphology, and some examples have focal nuclear hyperchromasia or multinucleation, often with a so-called degenerative or symplastic-type atypia. Fat cells are often present in the background. Most cases are immunoreactive for CD34, desmin, and hormone receptor proteins, and most show loss of RB1 immunoexpression. This tumor is closely related to cellular angiofibroma and has much more morphologic overlap with this entity than angiomyofibroblastoma (Howitt and Fletcher 2016).

Behavior and Treatment Angiomyofibroblastoma is typically cured by simple local excision. There is a single unique case report of an angiomyofibroblastoma with sarcomatous transformation (angiomyofibrosarcoma) in an elderly female (Nielsen et al. 1997a). The sarcomatous component resembled myxofibrosarcoma (myxoid malignant fibrous histiocytoma).

Cellular Angiofibroma Clinical Features Cellular angiofibroma (Iwasa and Fletcher 2004b; Nielsen and Young 2001; Nucci et al. 1997) (angiomyofibroblastoma-like tumor) (Laskin et al. 1998) is a recently described benign superficial soft tissue tumor of the vulvoperineal and inguinal regions. Vaginal examples have been

1416

reported, but in our experience, these are usually located at the introitus. The process generally affects individuals in the fourth through eighth decades of life, with a median age in the 40s (Iwasa and Fletcher 2004b). Most lesions are painless.

Pathologic Findings Gross examination usually reveals a small (often