Practical Breast Pathology Frequently Asked Questions PDF

Practical Anatomic Pathology Series Editors: Fan Lin · Ximing J. Yang Yan Peng Ping Tang Editors Practical Breast Pathology Frequently Asked Questions Practical Anatomic Pathology Series Editors Fan Lin Geisinger Health System Danville, PA, USA Ximing J. Yang Feinberg School of Medicine Northwes...

Practical Anatomic Pathology Series Editors: Fan Lin · Ximing J. Yang Yan Peng Ping Tang Editors Practical Breast Pathology Frequently Asked Questions Practical Anatomic Pathology Series Editors Fan Lin Geisinger Health System Danville, PA, USA Ximing J. Yang Feinberg School of Medicine Northwestern University Chicago, IL, USA The proposed Book Series will be designed to provide a comprehensive, practical and state-of-­ the art review and update of the major issues and challenges specific to each subspecialty field of surgical pathology in a question and answer (Q&A) format. Making an accurate diagnosis especially from a limited sample can be quite challenging, yet crucial to patient care. The proposed Book Series, using the most current and evidence-based resources, will 1) focus on frequently asked questions in surgical pathology in day-to-day practice; 2) provide quick, accurate, terse, and useful answers to many practical questions encountered in daily practice; 3) emphasize the importance of a triple test (clinical, radiologic, and histologic correlation); 4) delineate how to appropriately utilize immunohistochemistry, in situ hybridization and molecular tests; and 5) minimize any potential diagnostic pitfalls in surgical pathology. These books will also include highly practical presentations of typical case scenarios seen in an anatomic pathology laboratory. These will be in the form of case presentations with step-by-­ step expert analysis. Sample cases would include common but challenging situations, such as evaluation of well-differentiated malignant tumors vs. benign/reactive lesions; distinction of two benign entities; sub-classification of a malignant tumor; identification of newly described tumor and non-tumor entities; workup of a tumor of unknown origin; and implementation of best practice in immunohistochemistry and molecular testing in a difficult case. The Q&A format will be well accepted, especially by junior pathologists, for several reasons: 1) this is the most practical and effective way to deliver information to a new generation of pathologists accustomed to using the Internet as a resource and, therefore, comfortable and familiar with a Q&A learning environment; 2) it’s impossible to memorialize and digest massive amounts of new information about new entities, new and revised classifications, molecular pathology, diagnostic IHC, and the therapeutic implications of each entity by reading large textbooks; 3) sub-specialization is a very popular practice model highly demanded by many clinicians; and 4) time is very precious for a practicing pathologist because of increasing workloads in recent years following U.S. health care reforms. This Book Series will meet all of the above expectations. These books will be written by established and recognized experts in their specialty fields and will provide a unique and valuable resource in the field of surgical pathology, both for those currently in training and for those already in clinical practice at various skill levels. It does not seek to duplicate or completely replace other large standard textbooks; rather, it will be a new, comprehensive yet concise and practical resource on these timely and critical topics. More information about this series at http://www.springer.com/series/13808 Yan Peng Ping Tang Editors Practical Breast Pathology Frequently Asked Questions Editors Yan Peng, MD, PhD Ping Tang, MD, PhD Department of Pathology Department of Pathology and Laboratory UT Southwestern Medical Center Medicine Dallas, TX Loyola University Medical Center USA Maywood, IL USA Practical Anatomic Pathology ISBN 978-3-030-16517-8    ISBN 978-3-030-16518-5 (eBook) https://doi.org/10.1007/978-3-030-16518-5 © 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, expressed 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 To my husband, Zhongyue Jiang, PhD, and to our beloved daughter, Sharon, for their love, inspiration and continuous support” Yan Peng To my husband, Bruce Xu, and two daughters, Sara and Lora, for their unconditional love and support Ping Tang Preface With recent understanding of breast tumor biology and rapid advances in breast cancer treat- ment, timely updates on diagnostic surgical pathology of breast disease become keenly neces- sary and important. The purpose of this book is to provide a practical, evidence-based, up-to-date, problem-solving guide to frequently encountered diagnostic problems, challenges, and controversies in breast pathology and associated molecular pathology practice, with emphasis on addressing diagnostic issues that have significant impact on clinical management. This textbook is based on a PubMed (US National Library of Medicine, Bethesda, Maryland) literature review and the editors’ and chapter authors’ personal experiences. It consists of ten chapters with an abundance of color photomicrographs and images. The book is organized in a question-and-answer format accompanied by case presentations. We hope this format will facilitate finding answers to frequently asked questions in breast pathology. The book also discusses genetic alterations and molecular abnormalities in breast cancer and commonly encountered interpretation dilemmas regarding immunohistochemistry in breast cancer and metastatic cancer to the breast, with a focus on prognostic and predictive tumor biomarkers. In addition, the book covers some uncommon, diagnostically challenging breast lesions. We hope that practicing pathologists and pathologists-in-training will find this book helpful for efficiently solving diagnostic problems in their daily practice in breast pathology. Dallas, TX, USA Yan Peng, MD, PhD Maywood, IL, USA Ping Tang, MD, PhD vii Acknowledgments We are grateful to all the contributing chapter authors for working with us on this book and for making this collaboration a memorable and rewarding experience. We are honored to have been invited to contribute to this volume in the Practical Anatomic Pathology series, which is edited by Dr. Fan Lin and Dr. Ximing J. Yang and published by Springer. We greatly appreciate our outstanding mentors—Dr. David Dabbs, Dr. Steven Hajdu, and Dr. Daryl Carter—for their teaching and guidance during our breast pathology training. Dallas, TX, USA Yan Peng, MD, PhD Maywood, IL, USA Ping Tang, MD, PhD ix Contents 1 Intraductal Proliferative Disease of the Breast..........................   1 Xiuzhen Duan, Yihong Wang, Hua Guo, and Ping Tang 2 Invasive Ductal Carcinoma (NOS) of the Breast.........................  25 Xiaoxian Li, Zaibo Li, Xiaoyan Cui, and Yan Peng 3 Invasive Carcinoma of the Breast: Special Types........................ 39 Zaibo Li, Xiaoyan Cui, Xiaoxian Li, and Yan Peng 4 Lobular Breast Lesions..............................................  73 Megan L. Troxell, Yun An Chen, Jing Yu, Debra M. Ikeda, and Kimberly H. Allison 5 Papillary Lesions of the Breast (IDP, IDPC, EPC, SPC)................... 145 Julia Y. Tsang, Ping Tang, and Gary M. Tse 6 Fibroepithelial Lesions (Phyllodes Tumor and Fibroadenoma) of the Breast....................................................... 159 Julia Y. Tsang and Gary M. Tse 7 Immunohistochemistry in Breast Cancer............................... 173 Ping Tang, Marilyn M. Bui, and Yan Peng 8 Breast Cancer with Hereditary Cancer Predisposition Syndromes.......... 193 Roshni Rao, Caitlin B. Mauer, Margaret Chen-Seetoo, and Yan Peng 9 Mesenchymal and Lymphoid Lesions in the Breast...................... 203 Xi Wang and Andrew G. Evans 10 Metastatic Cancer in the Breast...................................... 237 Bradley M. Turner Index................................................................. 259 xi Contributors Kimberly H. Allison, MD Stanford University School of Medicine, Department of Pathology, Stanford, CA, USA Marilyn M. Bui, MD, PhD Department of Pathology, Moffitt Cancer Center, Tampa, FL, USA University of South Florida, Department of Pathology, Tampa, FL, USA Yun An Chen, MD University of Washington, Department of Radiology, Seattle, WA, USA Margaret Chen-Seetoo, MD Department of Surgery, Herbert Irving Pavilion, Columbia University Vagelos College of Physicians and Surgeons, New York, NY, USA Xiaoyan Cui, MD, PhD Department of Pathology, The Ohio State University Wexner Medical Center, Columbus, OH, USA Xiuzhen Duan, MD, PhD Department of Pathology and Laboratory Medicine, Loyola University Medical Center, Maywood, IL, USA Andrew G. Evans, MD, PhD Department of Pathology and Laboratory Medicine, University of Rochester Medical Center, Rochester, NY, USA Hua Guo, MD, MS Department of Pathology and Cell Biology, Columbia University Irving Medical Center, New York, NY, USA Debra M. Ikeda, MD, FACR, FSBI Stanford University School of Medicine, Stanford, CA, USA Xiaoxian Li, MD, PhD Department of Pathology and Laboratory Medicine, Emory University, Atlanta, GA, USA Zaibo Li, MD, PhD Department of Pathology, The Ohio State University Wexner Medical Center, Columbus, OH, USA Caitlin B. Mauer, MA, MS, CGC Department of Cancer Genetics, University of Texas Southwestern Medical Center, Dallas, TX, USA Yan Peng, MD, PhD Department of Pathology, Clements University Hospital, Dallas, TX, USA Department of Pathology, University of Texas Southwestern Medical Center, Dallas, TX, USA Roshni Rao, MD, FACS Division of Breast Surgery, Department of Surgery, Herbert Irving Pavilion, Columbia University Vagelos College of Physicians and Surgeons, New York, NY, USA Ping Tang, MD, PhD Department of Pathology and Laboratory Medicine, Loyola University Medical Center, Maywood, IL, USA Megan L. Troxell, MD, PhD Stanford University School of Medicine, Department of Pathology, Stanford, CA, USA xiii xiv Contributors Julia Y. Tsang, PhD Department of Anatomical and Cellular Pathology, Chinese University of Hong Kong, Shatin, Hong Kong Gary M. Tse, MBBS, FRCPC Department of Anatomical and Cellular Pathology, Prince of Wales Hospital, Shatin, Hong Kong Bradley M. Turner, MD, MPH, MHA Department of Pathology and Laboratory Medicine, University of Rochester Medical Center/Highland Hospital, Rochester, NY, USA Xi Wang, MD Department of Pathology and Laboratory Medicine, University of Rochester Medical Center, Rochester, NY, USA Yihong Wang, MD, PhD Department of Pathology and Laboratory Medicine, Rhode Island Hospital/Lifespan Medical Center, Warren Alpert Medical School of Brown University, Providence, RI, USA Jing Yu, MD, PhD Magee-Womens Hospital, University of Pittsburgh Medical Center, Pittsburgh, PA, USA  Intraductal Proliferative Disease of the Breast 1 Xiuzhen Duan, Yihong Wang, Hua Guo, and Ping Tang List of Frequently Asked Questions 3. What is the imaging finding of intraductal prolifera- tive disease? 1. What is intraductal proliferative disease? Imaging findings are not specific. The most frequent imaging Intraductal proliferative diseases are diverse groups of prolif- findings are microcalcification, architecture distortion, or erations typically originating in and confined to the terminal mass/nodule. Those lesions are often assessed as a Breast ductal–lobular unit (TDLU). They are associated with an Imaging Reporting and Data System (BIRADS) category 4, increased risk of subsequent development of breast cancer of which is an indication for biopsy [2, 3]. different magnitudes. They often include usual ductal hyperplasia (UDH); columnar cell lesions (CCLs): columnar 4. What are the clinical implications of intraductal pro- cell change (CCC), columnar cell hyperplasia (CCH), and liferative diseases? Why is it important to differentiate flat epithelial atypia (FEA); atypical ductal hyperplasia them from one another? (ADH); and ductal carcinoma in situ (DCIS). Atypical lobu- lar hyperplasia (ALH) and lobular carcinoma in situ (LCIS) Lesions of low-grade breast neoplastic pathways are a group are also part of this group, and they will be discussed of high-risk lesions including columnar cell lesions, ALH/ separately. LCIS, ADH, and low-grade DCIS. They often coexist and share morphologic, immunophenotypic, and genetic charac- 2. What is the clinical presentation of intraductal prolif- teristics. Patients with ADH and low-grade DCIS were erative disease? found to have a four- to ten-fold increased risk of breast cancer over the general population. Notably, these high-risk No clinical features are specifically correlated with this lesions are not only markers of future carcinoma but also group of diseases, because most of these lesions are micro- indicators of concurrent carcinoma missed due to biopsy scopic in size and not palpable, with the exception of some sampling. The practical implication for pathologists is to DCIS that can present as a mass lesion. correctly identify these lesions and promptly search for other coexisting lesions. Although the management of lesions such as FEA is currently debatable, core needle Drs. Duan, Wang, and Guo contributed equally to this chapter. biopsy (CNB) diagnosis of ADH (management by surgical X. Duan ∙ P. Tang (*) Department of Pathology and Laboratory Medicine, excision) is done in most institutions. A multidisciplinary Loyola University Medical Center, Maywood, IL, USA management with radiology–pathology correlation is advo- e-mail: [email protected]; [email protected] cated [4, 5]. Y. Wang Department of Pathology and Laboratory Medicine, Rhode Island 5. What are the typical gross characteristics of intra- Hospital/Lifespan Medical Center, Warren Alpert Medical School ductal proliferative disease? of Brown University, Providence, RI, USA e-mail: [email protected] There are no specific gross tissue findings associated with H. Guo Department of Pathology and Cell Biology, UDH, CCC, FEA, or ADH except for rare cases involving Columbia University Irving Medical Center, New York, NY, USA e-mail: [email protected] © Springer Nature Switzerland AG 2019 1 Y. Peng, P. Tang (eds.), Practical Breast Pathology, Practical Anatomic Pathology, https://doi.org/10.1007/978-3-030-16518-5_1 2 X. Duan et al. mass lesions such as radial scar, sclerosing adenosis, or 9. What is UDH and what are its key histologic features? papilloma. (Fig. 1.1a–f) 6. What is the most effective way to sample the surgical UDH is characterized by a solid or fenestrated proliferation specimen? of epithelial cells that often show streaming growth, particu- larly in the center of involved spaces [1, 10]. The cytologic These lesions are often nonpalpable and excised by guided features are heterogeneous epithelial cell proliferation; vari- wire localization. It is recommended to submit the entire ability in size, shape, and orientation; poorly defined cell specimen for microscopic evaluation if the specimen can fit borders; and variation in size, shape, and placement of into 20 cassettes. If the specimen is larger, the most effective nuclei. The architectural features are fenestrated, micropap- way to sample the surgical specimen is to sample the area/ illary (gynecomastoid type), and solid growth patterns; for- lesion of interest (a clip or biopsy site). Serially slice the mation of irregularly shaped oval, angulated, or slit-like specimen and submit the entire area of lesion with focus on spaces; and streaming or swirling appearance. the relationship with surrounding breast tissue and surgical margins. If the patient has a prior biopsy of ADH or DCIS 10. What IHC markers can be helpful for the diagnosis and there is no grossly visible lesion, the entire specimen for UDH? (Fig. 1.2a–f) should be mapped and examined in order to estimate the extent of DCIS if present. Immunohistochemistry mosaic patterns of expression of high-molecular-weight cytokeratins such as CK5/6 and vari- 7. What is the clinical–pathologic correlation? What is able expression of estrogen receptor (ER) are the most help- considered concordance? ful markers for differential diagnosis of UDH from ADH/LG DCIS and columnar cell lesions. Core needle biopsy (CNB) is the preferred, initial, minimally invasive diagnostic procedure for nonpalpable breast lesions 11. What are the underlying molecular changes associ- or palpable breast masses. Concordance assessment of the ated with UDH? histologic, imaging, and clinical findings determines further management. For this reason, CNB findings require correla- Molecular genetic studies have identified loss of heterozy- tion with imaging and clinical findings to determine concor- gosity (LOH) at several loci in UDH; however, there are no dance, and to either exclude the diagnosis of a malignancy consistent genetic alterations for UDH. Epidemiologic stud- by further histological evaluation or to establish a formal ies have shown that UDH is associated with a slightly plan of follow-up through risk-based, shared decision-mak- increased general breast cancer risk. Currently, UDH is not ing with the patient. viewed as a precursor lesion of DCIS [12, 13]. If CNB was performed for mammographic calcifica- tions, then radiographic and microscopic confirmation of 12. How does one differentiate UDH from grade 2 DCIS? calcifications in the specimen should be documented; oth- erwise, further efforts to identify and excise them are indi- Grade 2 DCIS or intermediate-grade DCIS refers to a group cated. If imaging reveals features suspicious for of DCIS that cannot be assigned readily to the high– or low– malignancy – such as a spiculated or irregular mass or nuclear grade categories. The nuclei are moderate pleomor- architectural distortion – and histology reveals a lesion – phic, less than in high-grade type, but lack the uniformity such as invasive carcinoma, fat necrosis, or radial scleros- and typically are larger than those seen in the low-grade type. ing lesion – it is considered concordant after radiology and Histologically, the micropapillary and solid pattern of clinical correlation [7–9]. DCIS can be difficult to distinguish from the gynecomastoid and solid pattern of UDH. Oftentimes, the cytologic and archi- 8. What is considered discordance? What is the next step tectural features are of little help because variation in cell size for a non-concordant case? and shape and even secondary irregular lumina may be found in both lesions. Micropapillary DCIS is characterized by bul- Discordance refers to the situation in which a breast CNB bous tips in which the nuclei are enlarged throughout the demonstrates benign histology, while the clinical or imaging micropapillations. In contrast, the micropapilla of gynecomas- findings are suspicious for malignancy. If there is discor- toid UDH have a broad base with a narrower, pinched tip and dance between imaging and pathology, histological evalua- the nuclei are larger at the base and smaller at the tip. In those tion is still needed. This can be accomplished either by repeat cases, careful morphologic observation is required, and, when CNB or surgical excision. needed, demonstration of CK5/6 mosaic pattern of immuno- 1 Intraductal Proliferative Disease of the Breast 3 a b c d e f Fig. 1.1 UDH H&E stains. Intraductal proliferation of a mixed population of cells (a); streaming pattern of proliferation (b); with focal central necrosis (c); with mitosis (d); involved in papilloma (e); and involved in gynecomastia (f) 4 X. Duan et al. a b c d e f Fig. 1.2 UDH and IHC for CK5/6. Proliferation into the lumen (a), with cribriform-like pattern (c) and with solid proliferation (e); and their cor- responding mosaic pattern of ER staining (b, d, f) histochemistry is indicative of UDH whereas grade 2 DCIS hesion, low to intermediate nuclear grade, and massive shows a distinct CK5/6 negative expression. expansion of the acini. The differential diagnosis LCIS variants such as florid lobular carcinoma in situ between solid pattern of UDH and FLCIS can sometimes be (FLCIS) refers to a proliferative lesion that exhibits cell dys- difficult. Histologically, the solid pattern of UDH usually 1 Intraductal Proliferative Disease of the Breast 5 contains some fenestration at the periphery that may be the ally secretions and/or microcalcifications inside the lumen. clue to the nature of the epithelial proliferation, whereas loss CCLs are increasingly being encountered in breast biopsies of polarity, loss of cell–cell cohesion, intracytoplasmic vacu- with association of microcalcifications which are detected on oles, and the absence of microacini should raise concern for mammographic screening. CCLs have been described lobular neoplasia. Immunohistochemistry loss of E-cadherin and classified under a variety of names by different authors. can be used to confirm the presence of lobular neoplasia. The World Health Organization (WHO) Working Group on the Pathology and Genetics of Tumors of the Breast currently 13. How does one differentiate UDH from basal-like categorizes them as columnar cell change (CCC), columnar DCIS? cell hyperplasia (CCH), and flat epithelial atypia (FEA) refer- ring to previously called CCC or CCH with atypia. A small proportion of high-grade DCIS exhibits basal-­like The involved acini of CCC usually have irregular con- characteristics with a triple negative phenotype, expression tours and are lined by one or two layers of columnar epithe- of basal cytokeratins (such as CK5/6, CK14, CK17, etc.), lial cells with uniform, ovoid to elongated nuclei regularly and/or EGFR compared to non-triple-negative high-grade oriented perpendicular to the basement membrane, with DCIS. Both UDH and basal-like DCIS can express high-­ evenly dispersed fine chromatin and inconspicuous nucleoli. molecular cytokeratins; however, UDH has a unique mosaic CCH also has variably dilated acini with epithelium lining pattern of expression with variable ER, while basal-like showing similar cytologic features to CCC but is composed DCIS is negative for ER [15, 16]. of cellular stratification or tufting more than two cell layers thick. Crowding or overlapping of the nuclei in the prolifera- 14. Can UDH have necrosis? Punctate and central? tive foci may give the impression of nuclear hyperchromasia. Therefore, low-power view often exhibits a distinct UDH can have focal necrosis, but rarely has central eye-catching blue color of the columnar arrangement. necrosis. However, there should be no true atypical micropapillary structure for CCH. Sometimes, exaggerated apical cyto- 15. Can UDH have mitosis? plasmic snouts may give a hobnailed appearance. Yes, but this is not common. 18. What IHC markers can be helpful for the diagnosis of CCC/CCH/FEA? 16. What types of breast lesions can UDH be associated with? The full spectrum of columnar cell lesions – including CCC, CCH, and FEA – shares similar immunophenotypic features UDH can associate with any types of breast lesions. It is with low-grade DCIS. The lining epithelium expresses low- often found against a background of normal tissue, adenosis, molecular-weight cytokeratins (LMW-CKs) such as CK7, or in association with a group of benign intraductal prolifera- CK8, CK18, and CK19, and broad-spectrum cytokeratin tive breast diseases such as columnar cell lesions, ADH, and AE1/AE3, but is negative for high-molecular-­weight cyto- DCIS. It may be observed in the context of benign tumor– keratins (HMW-CKs) such as 34βE12 (CK903), CK5 or forming lesions such as radial scar/radial sclerosing lesion, CK5/6, and CK14. CCLs are typically strongly positive or an intraductal papilloma. Florid UDH in an intraductal for E-cadherin. A majority of cells exhibit intense and dif- papilloma often shows a syncytial or streaming pattern with fuse immunoreactivity to estrogen receptor (ER) and proges- secondary, slit-like lumina or fenestrations. The proliferating terone receptor (PR), but negative for HER2 [18, 21]. They epithelial cells are typically heterogeneous and overlapping also often express gross cystic disease fluid protein-15 or irregularly spaced, and have indistinct cell borders and (GCDFP-15) and Bcl-2. Proliferation rate as indicated by variably sized nuclei with frequent nuclear grooves and Ki-67 staining is low. pseudonuclear inclusions. 19. What are the underlying molecular changes associ- 17. What is CCC and what are its key histologic ated with CCC? features? By microdissection approach, low level of allelic imbalance Columnar cell lesions (CCLs) of the breast are a spectrum of and recurrent 16q loss have been demonstrated in CCLs. benign to atypical entities, which are characterized by No mutational changes are found in simple CCLs without enlarged terminal duct lobular units (TDLUs) with variably atypia [7, 8]. Some examples of CCH have shown to exhibit dilated acini lined by single or multilayered columnar-shaped loss of heterozygosity (LOH) at chromosome 9q and 10q. epithelial cells, which may have apical snouts. There are usu- Studies showed progressive accumulation of allelic damage 6 X. Duan et al. in CCLs with atypia, DCIS, and invasive carcinoma, with a a fractional mutation percentage increasing progressively from CCH through invasive carcinoma. Allelic loss damage appeared to preferentially target loci at 9q, 10q, 19p, 16q, 17p, and 17q [20, 22]. Recurrent changes were identified as loss on 16q, 17p, and X and gains on 15q, 16p, and 19. These findings raise the possibility that CCL may be the pre- cursor lesion for low grade DCIS and low grade IDC. Additional investigations are needed to further characterize the features of CCLs. 20. How does one differentiate CCC from UDH? (Fig. 1.3a–c) CCC and CCH are both featured by variably dilated TDLU acini containing secretory material. From low-power view, it b may be mistaken for microcysts from fibrocystic changes, which are typically lined by attenuated, cuboidal, or apocrine epithelium. Occasionally, CCLs may show more prolif- erative changes: especially when areas of cellular stratification or tufting from CCH become broad and multiple layered, then the lesion may mimic UDH. The gynecomastoid pattern of micropapillations in UDH may have similar features to the tufts seen in CCH, including broad-­based papillae and narrow- pinched tip. The lining cells of CCH always have pronounced ovoid to elongated nuclei regularly oriented perpendicular to the basement membrane, with apical snouts. CK5/6 is strongly positive in mixed population of UDH and negative in CCH. ER shows heterogeneously low positive in UDH, while showing intense and diffuse positivity in CCC. However, spending effort to make a distinction between CCH and UDH is “of no c diagnostic importance” according to some authors. 21. Can CCC have necrosis? Punctate and central? Apical snouts and luminal secretions may be present, but usually not prominent or exaggerated in CCC. In CCH, api- cal snouts or luminal secretions can become exaggerated and prominent. True intraluminal necrosis is very rare in CCC. When present, high-power view to exclude possible atypia is necessary. 22. Can CCC have mitosis? CCC has fine nuclear chromatin. Nucleoli and mitotic fig- Fig. 1.3 UDH – blunt duct adenosis. Dilated acinar structures in ures are exceedingly rare or absent. The Ki67 index was sig- TDUL (a) with larger plump cells without polarity (b) can be overlap- nificantly lower in CCC (mean, 0.1%) than in normal TDLUs ping and retain CK5/6 stain (c) (mean, 2.4%). 23. What types of breast lesions can CCC be associated “Rosen Triad.” The observation is that patients with tubular with? carcinoma of the breast often had foci of CCC distributed in surrounding tissue or sometimes even merging with the car- Atypical lobular hyperplasia (ALH) or LCIS frequently cinomatous lesions. LCIS may sometimes be present accompany columnar cell abnormalities, and tubular carci- also. CCLs are increasing being encountered in breast biop- noma may also be present, which composes a triad termed sies with association of microcalcifications detected on 1 Intraductal Proliferative Disease of the Breast 7 mammographic screening. Although we should be aware of the native epithelial cells by one to several layers of rela- the possibility of coincidental tubular carcinoma, it is not tively round or oval monotonous cells which have loss of demonstrated in most women, and the risk of subsequent polarity and increased nuclear/cytoplasmic ratio. The tubular carcinoma is poorly documented. Follow-up nuclei are evenly distributed and align the basement mem- studies suggested that CCC is associated with a mild (~ 1.5 brane, and the cytological atypia resembles the cells diag- fold) increase in breast cancer risk. However, this increased nostic for G1-DCIS. The nuclear chromatin may be slightly risk is not clearly independent of the risk associated with clumping, irregular, and vacuolated, and margination may the concurrent proliferative disease such as UDH. be present with visible prominent nucleoli. Mitotic figures may be seen. 24. What is FEA and what are its key histologic features? Since the interobserver and intraobserver agreement is (Fig. 1.4a–f) poor for the diagnosis of FEA, careful investigation of all CCLs at medium to high magnification is necessary to detect Flat epithelial atypia (FEA) is a neoplastic alteration of the any cytological atypia. More importantly, recognition of enlarged dilated TDLUs, characterized by replacement of true ADH or even low-­grade DCIS is crucial. a b c d Fig. 1.4 CCC and FEA. CCC with TDLU with dilated acinar spaces and columnar cells perpendicular to BM (a). EFA with round cells lost polar- ity and lack prominent secondary structures (b, c). Negative for CK5/6 (d), negative for p63 (e), and uniformly positive for ER (f) 8 X. Duan et al. e f Fig. 1.4 (continued) 25. How does one differentiate UDH from FEA? lations,” complex architectural patterns should not be seen. The complex architectural patterns which indicate ADH In contrast to mixed population and overlapping nuclei fea- include secondary architecture such as rigid Roman bridges, tures of UDH, FEA is characterized by epithelial cells that bars and arcades, cellular tufts with well-developed, club- are more cuboidal and rounded monomorphic nuclei with shaped micropapillations, or cribriforming growth pattern loss of polarization, which resemble those seen in low-­grade with cell polarization around the lumens. Whether to DCIS. The gynecomastoid pattern of micropapillations in diagnose ADH or low-grade DCIS will depend on the details UDH may have similar features to the tufts seen in CCH, (quality and quantity) of the architectural and cytological including broad-based papillae and narrow-pinched tip. atypia. However, in the majority of times with multiple epithelial When a flat lesion shows high-grade lining epithelium with cell layers, the lining epithelium in FEA remains flat with no pleomorphic and marked nuclear atypia, even with a single architectural atypia. cell layer, a clinging-type high-grade DCIS diagnosis should Immunohistochemical stain for CK5/6 and ER shows dis- be rendered. For such a case, other histological features of tinct difference between UDH and FEA. CK5/6 is strongly high-grade DCIS are usually easy to find. Rarely, columnar positive in mixed population of UDH and negative in FEA. cell lesion with intermediate-grade cytological atypia without ER shows heterogeneously low positive in UDH, while complex architectural patterns may be encountered in breast appearing intense and showing diffuse positivity in FEA. specimens. Currently, there is no consensus on how to best classify the lesion. A diagnosis of FEA accompanied by a 26. What are the underlying molecular changes associ- comment indicating that the degree of nuclear atypia is greater ated with FEA? than that typically seen in FEA is recommended. Some genetic studies have indicated that FEA is a clonal 28. Can FEA have necrosis? Punctate and central? lesion and shares genetic alterations with low-grade DCIS and tubular carcinoma, such as LOH at loci chromosome Apical snouts and luminal secretions may be present in FEA, 16q, allelic loss, or damage to 9q, 10q, 17p, and 17q [22, 24]. occasionally becoming exaggerated and prominent with a However, the available follow-up studies of patients with hobnailed appearance. Intraluminal punctate necrosis or FEA demonstrated an extremely low risk of subsequent pro- apoptosis can be present. However, central comedonecrosis gression to invasive breast cancer when present as an isolated is extremely rare. When central necrosis is identified, the lesion [18, 25]. Thus, World Health Organization (WHO) degree of epithelial atypia needs to be carefully investigated Working Group on the Pathology and Genetics of Tumors of to rule out clinging growth pattern of high-grade DCIS or Breast recommended that FEA should not be treated as undersampled DCIS. equivalent to ADH or ALH. 29. Can FEA have mitosis? 27. How does one differentiate FEA from ADH? Mitotic figures may be seen but are uncommon. Studies Although the low-grade atypical lining epithelial cells in have shown that Ki-67 proliferative index was significantly FEA may “form mounds, tufts, or short, abortive micropapil- higher in FEA (8.2%) than in CCLs without atypia and nor- 1 Intraductal Proliferative Disease of the Breast 9 mal TDLUs, similar to that of low-grade DCIS (8.9%), but should be subject to radiologic–pathologic correlation to was significantly lower than the proliferation rate in interme- determine the need for surgical excision [18, 30, 31]. diate- to high-grade DCIS (25.4%). 31. What is ADH and what are its key histologic fea- 30. What types of breast lesions can be associated with tures? (Fig. 1.5a–i) FEA? Atypical ductal hyperplasia (ADH), by definition, is a prolif- FEA is reported to occur in 0.7–12.2% of percutaneous eration of uniform and monomorphic epithelial cells, evenly breast biopsies obtained for mammographic calcifications distributed as solid nests or well-formed architectures within. FEA significantly has been often detected in conjunc- terminal-duct lobular units. In another words, ADH is a pro- tion with a lesion of higher concern, including ALH, LCIS, liferative epithelial lesion, with some but not all architectural ADH, and low-grade DCIS, and has been associated with tubu- and cytologic features of low-grade ductal carcinoma in situ. lar carcinomas. The presence of FEA should trigger a The key histologic features of ADH include both cytologic careful search for areas with diagnostic features of these and architectural alterations. These lesions are usually small lesions, especially invasive tubular carcinomas. However, and may be multicentric like low-grade DCIS. The cytologic the upgrade rates to carcinoma among the current follow-up features of ADH at least partially resemble those of low- studies varied from 0 to 42%. FEA identified at the surgical grade DCIS. Cells in ADH are relatively small and uniform resection margins is not an indication for additional surgery in shape and size. They appear monomorphic and clonal with [25, 27–29]. The approach to the diagnosis of FEA on a core nearly normal chromatin pattern and may be mixed with biopsy specimen has been controversial; such a situation cells of usual-type ductal epithelial hyperplasia. Nuclear a b c Fig. 1.5 ADH. Partially involved glands (a), associated with calcium (b), and irregular spaces (c). The ADH cells can be overlapping (d), look like a mixed population (f), and streaming (h), but they are uniformly lacking of stains for CK5 (e, g, i) 10 X. Duan et al. d e f g h i Fig. 1.5 (continued) 1 Intraductal Proliferative Disease of the Breast 11 atypia is mild and mitosis is uncommon. Necrosis – espe- nests or well-formed round- or oval-shaped spaces including cially central necrosis – is usually not associated with cribriform, micropapillae, small tufts, or rigid Roman bridg- ADH. Architecturally, ADH can be solid: evenly distributed ing, similar to that in DCIS. Cytologic features of UDH are uniform cells fill the lumen of terminal ducts and lobules different from those in ADH. UDH consists of heterologous without overlapping or streaming. They may also form clear- cell populations, and the nuclei of these cells are small and cut round or oval shaped spaces, such as cribriform pattern variable in size and shape. They may have inconspicuous with polarized lumen, micropapillae, small tufts, or ridged nucleoli and have rare mitoses. Cells in UDH tend to be Roman bridging, such as in DCIS [32–35]. crowded and haphazardly placed, with streaming and swirl- ing. The growth pattern of UDH can be solid, micropapillary, 32. What IHC markers can be helpful for the diagnosis or fenestrated. Fenestrated structures in UDH are ill formed, for ADH? irregular shaped, or slit like. With regard to immunohistochemistry, estrogen receptor The diagnosis of ADH is mainly based on morphology of (ER) is usually diffusely and strongly positive and CK5/6 is cytologic and architectural patterns. There are no immuno- negative or scattered cells positive in ADH. On the other logic markers for the diagnosis of ADH. That being said, hand, hyperplastic cells of UDH are also positive for ER, but some IHC markers can be helpful if the differential diagno- the staining pattern is not diffuse and the staining intensity is ses are benign ductal hyperplasia and ADH. Estrogen recep- variable. Cells in UDH usually show patchy or mosaic stain- tor (ER) is usually diffusely and strongly positive, and CK5/6 ing pattern for CK5/6 [32–35, 41, 42]. is negative or scattered cells positive in ADH. On the other hand, hyperplastic cells of UDH are also positive for ER but 35. How does one differentiate ADH from FEA? the staining pattern is not diffuse and the staining intensity is variable. Cells in UDH usually show patchy or mosaic stain- These two lesions are different in both architecture and ing pattern for CK5/6 [11, 36, 37]. cytology but do share some similarities. Architecturally, the terminal ductal lobular units (TDLU) are usually dilated, 33. What are the underlying molecular changes associ- without formation of cribriform, micropapillary or other ated with ADH? types of structure in FEA, while ADH can form different architectures including solid growth pattern, Roman bridg- There are very few studies on the molecular alteration of ing, blunt micropapillae, or cribriform formation. pure ADH. The diagnosis of ADH is often associated with Cytologically, these two lesions share some cytologic simi- low-grade DCIS and/or invasive ductal carcinoma, and the larity in that cells in both lesions are low grade with mild morphology of these lesions is similar in the same specimen. nuclear atypia. However, cells lining the dilated ducts in Most studies on molecular alterations of ADH are based on FEA are usually single layered and may be crowded; the tissue with combine ADH and DCIS or invasive carcinoma. morphology of these cells is enlarged cuboidal shaped or These studies found genomic similarity and nearly identical rounded with columnar configuration while cells in ADH are molecular alteration and chromosome imbalances in ADH, more uniform and can form a solid nest. There are no IHC and associated DCIS or invasive carcinoma in the same spec- markers to differentiate ADH from FEA. Cells in both lesions imens. Reported molecular changes include aneuploidy and are positive for ER, usually uniform and strong. They are LOH in at least one focus with loss of chromosome 16 being negative for CK5/6 [32–35, 43–45]. the most common. Other molecular changes include loss at 17p and 11q13, and gains at 1q [38–40]. 36. How does one differentiate ADH from grade 1 DCIS? 34. How does one differentiate ADH from UDH? ADH resembles grade 1 DCIS both in cytology and archi- tecture; the differences of these two lesions are both quali- The differences of ADH and UDH are mainly in their mor- tative and quantitative, both in cytology and in architectures. phology, and immunostains can be of some help in difficult Qualitatively, to make a diagnosis of low-grade DCIS, the cases. The morphologic differences between ADH and UDH lesion needs to fulfill all features of ductal carcinoma in include different cytologic and architectural features. The cyto- situ, that is, cytologically, atypical tumor cells are uniform, logic features of ADH at least partially resemble those of low- and, monomorphic, no streaming, or overlapping. grade DCIS. Cells in ADH are small, uniform, and monomorphic Architecturally, these cells may form a solid pattern or with nearly normal chromatin pattern and may be mixed with sharply defined cribriform, micropapillae, or rigid Roman cells of usual-type ductal epithelial hyperplasia. Mitoses and bridging. These atypical cells and well-formed architec- necrosis are rare. Architecturally, cells in ADH are evenly dis- tures need to fill the entire duct. The other criteria are quan- tributed without overlapping or streaming. They also form solid titative, which requires the size of lesional tissue fulfilling 12 X. Duan et al. Table 1.1 Differences of UDH, FEA, ADH, and grade 1 DCIS UDH FEA ADH Grade 1 DCIS Cytology Multiple types of mixed Enlarged cuboidal shaped or Uniform and monomorphic Uniform and monomorphic cell populations; no rounded cells with columnar cells; fine chromatin, cells, fine chromatin, nuclear atypia, normal configuration, fine chromatin inconspicuous nucleoli, rare inconspicuous nucleoli, rare pattern chromatin, mitosis pattern and rare mitosis, mitosis, may be mixed with mitosis can be seen maybe crowded normal cells Architecture Slit-like, streaming, and Dilated acinar structure in Complex secondary Complex secondary structures overlapping arrangement terminal ductal lobular unit structures such as rigid bars, such as rigid bars, bulbous with single layer lining cells bulbous micropapillae, micropapillae, round punctate that have loss of polarity round punctate space, and space, and rigid Roman rigid Roman bridging bridging Size No size limit, can be rare No size limit 2 mm or extensive Immunostains Variable positive for ER, Diffuse and strong positive for Diffuse and strong positive Diffuse and strong positive mosaic pattern CK5/6 ER and negative for CK5/6 for ER and negative for for ER and negative for CK5/6 CK5/6 Risk of cancer Slight Mild increase Moderate High risk the above criteria either involving two ducts or measuring the higher chance of cancer identified on the surgical speci- at least 2 mm in length. The diagnosis of ADH is made men. The rate of cancer identified on surgical specimen var- when ducts are only partially involved by carcinoma cells, ied by a large range; however, estimated upgrading of only one duct is fully involved, or the involved area is carcinoma on surgical specimen is 30% based on two large 2.5x of normal breast epithelial cells). They are not as uniform as that of grade 1 DCIS. Tumor cells in grade 2 DCIS show mild to moderate variation in cell shape and size. Chromatin is variably coarse and may have prominent nucleoli. 45. How does one differentiate grade 1 DCIS from florid LCIS? (Figs. 1.7a–e, 1.8a–c, 1.9a–d) Grade 1 DCIS cells are small, uniform, and cohesive, and Fig. 1.6 G1 DCIS. Low-grade and monotonous neoplastic cells form solid nests, cribriform, micropapillary, and other archi- >2 mm or 2 spaces (a), negative for CK5/6 (b), and uniformly positive tectures. Florid LCIS consists of classic lobular cells that are for ER (c) small, uniform, and noncohesive. These cells present in marked distended TDLU and occasionally are associated 42. What are the underlying molecular changes associ- with microcalcifications and central necrosis. Frequently, ated with grade 1 DCIS? they are associated with invasive lobular carcinoma. When suspicious, a confirmation IHC for lack of E-cadherin stain- Molecular changes associated with grade 1 DCIS are similar ing is recommended for diagnosis. Occasionally, low-grade to those of ADH, including aneuploidy, LOH, chromosomal DCIS can mimic LCIS (involves TDLU, termed c­ ancerization 14 X. Duan et al. a b c d e Fig. 1.7 G1 DCIS and florid LCIS. GI DCIS, solid and cribriform patterns and with central necrosis (a, b). FLCIS with classic lobular cells and central necrosis (c, d) and negative for E-cadherin (e) 1 Intraductal Proliferative Disease of the Breast 15 especially punctate necrosis. The presence of punctate necro- a sis does not change the diagnosis of grade 1 DCIS. Mitoses are rare in grade 1 DCIS; however, it does present. Grade 1 DCIS can associate with invasive ductal breast cancer, low grade [21, 46]. 47. What is grade 3 DCIS and what are its key histologic features? (Fig. 1.10a, b)? Grade 3 DCIS is a high-grade in situ ductal carcinoma. Tumor cells are large with pleomorphic nuclei, coarse clumped chromatin, and prominent nucleoli. Mitoses are common and often atypical. Grade 3 DCIS may present as solid pattern, cribriform, or micropapillary types with necro- sis and calcifications; it is often associated with marked peri- b ductal desmoplastic changes, which is not a feature of lower-grade DCIS. Comedo-type necrosis associated with necrotic debris and calcifications are common. Key features of grade 3 DCIS are high-grade pleomorphic nuclei (>2.5× of normal breast epithelial cells). The size of grade 3 DCIS is usually large, more than 5 mm; however, the diagnosis does not have a size limit [21, 46]. 48. What IHC markers can be helpful for the diagnosis of grade 3 DCIS? The diagnosis of grade 3 DCIS is made by morphology: high-grade pleomorphic nuclear features associated with central necrosis and calcifications. These tumors are often negative for ER and PR and positive for Her-2. Sometimes, c the tumor cells are positive for CK5/6 and EGFR, defining the basal-like subtype of DCIS. No definitive IHC markers are helpful in the diagnosis. 49. What are the underlying molecular changes associ- ated with grade 3 DCIS? The molecular profile of grade 3 DCIS is different from grade 1 DCIS both quantitatively and qualitatively and is similar to that of high-grade invasive ductal carcinoma. In general, 24.5% of patients with grade 3 DCIS carry germline mutation of BRCA1 and BRCA2. They tend to be ER and PR negative and Her-2 positive, and have a p53 mutation. Grade 3 DCIS is usually aneuploid with complex genetic Fig. 1.8 G1 DCIS mimics LCIS in TDLU. G1 DCIS involving TDLU patterns. It frequently has a loss of 11q, 14q, 8q, and 13q and (a, b) and showing strong positive membrane staining for E-cadherin (c) gains of 17q, 5p, and 8q. It also has amplifications of 17q12, 17q22–24, 6q22, 8q22, 11q13, and 20q13 [53–56]. of the lobules), and DCIS and LCIS could coexist in the same 50. How does one differentiate grade 3 DCIS from ADH? ductal space [21, 46]. Cytologic and nuclear features: Cells in ADH are uniform 46. Can grade 1 DCIS have necrosis? without variation in size and shape. Nuclei are small with nearly normal pattern chromatin, inconspicuous nucleoli, and Grade 1 DCIS usually does not have necrosis, especially very infrequent mitoses. On the other hand, tumor cells in central necrosis; however, it occasionally can have necrosis, grade 3 are large, pleomorphic, and hyperchromatic, with 16 X. Duan et al. a b c d Fig. 1.9 (a–d) G1 DCIS admix with LCIS. DCIS and LCIS involving single glandular spaces (HE and E-cadherin) a b Fig. 1.10 (a, b) G3 DCIS with central necrosis and positive CK5/6 staining 1 Intraductal Proliferative Disease of the Breast 17 prominent nucleoli and brisk mitoses. Necrosis, especially cen- Paget’s disease of the breast most often occurs in post- tral necrosis associated with microcalcifications, is common. menopausal women. A large majority of patients (>95%) Tumor markers: ADH is usually diffusely and strongly with MPD have underlying breast cancers, either DCIS or positive for ER and PR while grade 3 DCIS is usually nega- invasive ductal carcinoma. Symptoms of Paget’s disease tive for ER and PR and positive for Her-2 [21, 32–35, 46]. include eczematous rash, flaky or scaly skin lesion, and nip- ple discharge with straw colored or bloody fluid. These 51. How does one differentiate grade 3 DCIS from UDH? symptoms usually affect the nipple and areola and later spread to the remaining breast. Diagnosis is established by Cytologic features: In UDH, lesion cells are heterologous punch biopsy. mixed cell populations; nuclei are small and variable in size Histology of MPD: Glandular tumor cells present within and shape. They may have inconspicuous nucleoli and rare the epidermis of the nipple and areola and may extend to mitoses. Cells tend to be haphazardly placed and crowded adjacent skin. These tumor cells (Paget cells) are usually with streaming and swirling. The growth pattern can be large and round, with abundant pale cytoplasm. They have solid, micropapillary, or fenestrated. Fenestrated structures large nuclei with large and prominent nucleoli. Paget cells in UDH are ill formed, irregular shaped, or slit like. Tumor can present as isolated single cells, small clusters, or packed cells in grade 3 are large, pleomorphic, and hyperchromatic, clusters. Involved epidermis is thickened with papillomatosis with prominent nucleoli and brisk mitoses. Necrosis, espe- and enlargement of the interpapillary ridges. Hyperkeratosis, cially central necrosis associated with microcalcifications, parakeratosis, and marked inflammation may also occur and is common. The growth pattern of grade 3 DCIS can be may be associated with surface ulcer [57–60]. solid, irregular shaped, cribriform, or micropapillary, often associated with central punctate necrosis and calcification. 54. What IHC markers can be helpful for the diagnosis IHC: ER staining is of variable intensity and of mosaic of MPD? pattern staining for CK5/6 in UDH, while most grade 3 DCIS is negative for ER and negative or positive for CK5/6 [21, 41, The immunohistological pattern of Paget cells is the same as 42, 46]. the underlying breast cancer. Paget cells are usually positive for low-molecular-weight cytokeratin, such as cytokeratin 7 52. How does one differentiate grade 3 DCIS from grade and CAM 5.2. They are also positive for immunostaining of 2 DCIS? CEA and for special stain of PAS and diastase resistance. MPD is most often associated with high-grade carcinoma, Grade 3 DCIS and grade 2 DCIS are separated by their nuclear and most MPDs are also positive for Her-2 and less than half and cytoplasmic features and architectures. Tumor cells in of them express ER and PR. By contrast, surrounding squa- grade 3 are large, often more than three times that of RBC; mous cells are positive for CK5/6 and p63 and negative for nuclei are pleomorphic and hyperchromatic, with course low-molecular-weight keratin, CEA and PASD [58, 61, 62]. clumped chromatin, prominent nucleoli, and brisk mitoses. On the other hand, tumor cells in grade 2 DCIS show mild to 55. How does one differentiate MPD from nipple moderate variation in size and shape. They have variable melanoma? course chromatin, inconspicuous nucleoli, and rare mitoses. Central punctate necrosis with necrotic debris and calcifica- To differentiate MPD from melanoma can be difficult clini- tions are common in grade 3 DCIS, while these changes are cally and might have to be done through histopathology. not as common in grade 2 DCIS. Additionally, the diagnosis The morphology of MPD and nipple melanoma is similar of grade 3 DCIS does not have a size requirement. Simply put, in that tumor cells can present within the epidermis, and features of tumor cells in grade 2 DCIS are in-between grade melanin can be present in both tumors; however, the pattern 1 DCIS and grade 3 DCIS but closer to grade 1 DCIS [21, 46]. of distribution of tumor cells within the epidermis is differ- ent. Tumor cells of melanoma tend to form nests along the 53. What is mammary Paget’s disease and what are its dermo-epidermal junction with extension into papillary key histologic features? (Fig. 1.11a–d) dermis whereas tumor cells of MPD are usually distributed more diffusely. Tumor cells in MPD may form acini which Mammary Paget’s disease (MPD) is a rare type of breast cancer is not seen in melanoma. Immunohistochemical stains can involving the nipple and surrounding areola, in which glandu- be of help to differentiate these entities in difficult cases. lar tumor cells are located within squamous epithelium. It Cells in melanoma are positive for melanocytic markers accounts for 1–3% of all primary breast tumors. It was first and negative for markers of cytokeratin, PASD and CEA, described by British doctor Sir James Paget in 1874, who noted while Paget cells are negative for melanocytic markers but a relationship between a chronic eczematous ­disease of the positive for markers of cytokeratin, PASD and CEA nipple and areola and intraductal and invasive breast cancer. (Table 1.2) [63–65]. 18 X. Duan et al. a b c d Fig. 1.11 Mammary Paget disease (MPD). MPD grossly involved in nipple (a). The large tumor cells scattered in epidermis (b) are positive for CK7 (c) and HER2 (d) Table 1.2 Immunohistochemical stains differentiate cells in mela- noma, Paget cells, and squamous cells 57. How does one stage MPD? Melanocytic markers (HMB45, Low-molecular-­ PASD MPD itself is an intraepithelial carcinoma in the great S100, Mela A, and weight CK (CK7 CK5/6 and majority of cases. The staging of MPD depends on its Sox10 and Cam 5.2) and p63 CEA underlying carcinoma since >95% of MPD have underly- Paget’s N P N P disease ing carcinoma. Melanoma P N N N Squamous N N P N cells Case Presentations 56. What types of breast lesions can be associated with Case 1: FDH with Necrosis (Fig. 1.12a–d) MPD? History: A 40-year-old female with calcification on screen- MPD is associated with either DCIS or IDC in the great ing mammogram underwent core needle biopsy majority of cases. More than 95% of MPD is associated with Histology: Intraductal proliferation with central necrosis; underlying breast cancer, invasive or in situ carcinoma. cells are relatively uniform and not overlapping Immunostaining pattern and molecular changes of MPD are IHC: ER: positive 70%, PR: positive 60%, P63: peripheral stains the same as the underlying breast cancer. DDx: Consider DCIS, G2 1 Intraductal Proliferative Disease of the Breast 19 a b c d Fig. 1.12 Case 1: FUDH with necrosis. High-power view (40×) shows and a 10× view shows that the FUDH is part of intraductal papillary central necrosis with enlarged epithelial cells, mimic grade 2 DCIS (a); lesion (c). The mosaic staining pattern for CK5/6 supports the diagnosis lower-power view (20×) shows the cells with features of florid ductal of FUDH (d) hyperplasia (disorganizing and overlapping with irregular spaces) (b), Next Steps: DDx: CCC vs G2 DCIS vs others CK5/6: positive with mosaic pattern Next Steps: ER/PR staining pattern: positive with variable intensity CK5 Final Diagnosis: Final Diagnosis: FDH with necrosis Blunt duct adenosis Take-Home Messages: Take-Home Messages: Central necrosis can occur in UDH Overlapping with no clear cytoplasmic membrane CK5, not p63 is a best marker for UDH Myoepithelial layer usually prominent Mixed population identified by CK5 Case 2: Blunt Duct Adenosis mimic FEA (Fig. 1.13a–d) Case 3: G1-DCIS with Attenuated Myoepithelial Cells History: A 40-year-old female with calcification on screen- (Fig. 1.14a–f) ing mammogram underwent core needle biopsy Histology: Lobular centric proliferation with enlarged acinar History: A 40-year-old female with mass on screening mam- structures; cells are enlarged and without polarity; apocrine mogram underwent core needle biopsy snouts can be seen; no secondary structures Histology: DCIS, low-grade and cribriform pattern 20 X. Duan et al. a b c d Fig. 1.13 Case 2: Blunt duct adenosis vs FEA. This is a lobular centric highlighted by IHC analysis for CK5/6, which is different from CCC epithelial proliferation with all features of UDH, such as enlarged epi- that is lacking CK5/6 positive luminal cells (c). A low-power view dem- thelial cells that are overlapping, with different orientations and without onstrates the lobular centric architecture (d) prominent cell borders (a, b); the mixed population of epithelium is IHC: ER and PR: strongly positive focal areas (>1 mm) with small angulated glands in a desmo- Next Steps: plastic background, with necrosis, and suspicious of inva- Surgical resection, IHC for myoepithelial cells sion; areas of microinvasion (90% prognosis [1, 2]. of the tumor with this cribriform morphology. Areas of tubu- lar growth pattern are commonly seen, and those with minor  4. What is tubular carcinoma and what are its key diag- tubular component (95%). (e) Low Ki-67. (f) Negative Her2 (0). the small ductule at top center (200×). Serial sections stained for predic- E-cadherin staining was not performed Histologic Findings: See Figs. 4.45a–f and 4.46a–d. The ER+, PR+, Her2 negative, and Ki-67 low (5–10%). 6:00 lumpectomy contained a 2.4 cm invasive carcinoma, Multiple sub-2 mm satellite foci of invasive carcinoma which was densely cellular with grade 2 nuclei, without were also present. Margins were focally positive. The tubule formation. E-cadherin staining was not performed. 3:00 lumpectomy specimen contained multifocal invasive 4 Lobular Breast Lesions 129 a b c d Fig. 4.46 Case 2. Multifocal invasive lobular carcinoma. In addition to upper right (40×). (b) Keratin stain highlights the infiltrating lobular the 2.4 cm invasive lobular carcinoma shown in Case 2, Fig. 4.45, there carcinoma (40×). (c) A second separate focus of invasive tumor in the were multiple separate subcentimeter carcinomas, two of which are same histologic slide has more sclerotic stroma and is less cellular. It shown here, similar to the (pink) satellite masses diagramed in Fig. 4.3. does not stand out as well on H&E. ALH involves the central duct and (a) Low-power view of a discrete infiltrating mass lesion with desmo- the peripheral terminal duct lobular units. (40×). (d) Keratin stain again plastic stroma. The prominent ducts in the center demonstrate mild duc- highlights lobular carcinoma tal hyperplasia with a focus of inflammation below. LCIS is seen at lobular carcinoma, including multiple sub-1 cm tumors, –– Classic invasive lobular carcinoma is often ER+ PR+ along with LCIS and ALH. Predictive markers were per- Her2 negative. formed on one section containing two discrete tumors, –– Lobular carcinoma can rarely be hereditary (CDH1 and results were identical to the larger tumor. Margins or partners), although genetic testing was negative were focally positive and were later re-excised. Isolated here. tumor cells were present in one of the three lymph nodes (not shown). Case 3 Final Diagnosis: Multifocal invasive lobular carcinoma, History: A 53-year-old woman with multiple palpable grade 2. AJCC stage: pmT2 pN0(i+) breast masses Take-Home Messages: Imaging: Spiculated right breast mass at 12:00, 2 cm size; –– Invasive lobular carcinoma is often multifocal; MRI adjacent mass at 11:00, 1 cm size; and lobulated right can be helpful in this setting. breast mass at 9:00, 1.3 cm size. Right axillary ultrasound –– E-cadherin immunostaining is not necessary in mak- identified two suspicious lymph nodes with thickened cor- ing a diagnosis of lobular carcinoma. tices. After core biopsy, the patient received 130 M. L. Troxell et al. c­ hemotherapy: dose dense AC-T (anthracycline/­ leaving only fibrotic tumor bed in the breast and no evi- cyclophosphamide ­followed by taxane) with Her2-targeted dence of metastatic ductal carcinoma in lymph nodes. agents, namely, trastuzumab and pertuzumab. Further However, a 3.6 cm invasive lobular carcinoma remained imaging demonstrated near resolution of the 9:00 mass, after chemotherapy, with somewhat diminished tumor but minimal change in the 12:00 masses. Right mastec- cellularity, but with lymph node metastases that were not tomy was performed and grossly a 2.1 cm mass was iden- sampled on prior biopsy. tified at 12:00, with a 1.5 cm irregular mass at 9:00. Final Diagnosis: Residual invasive lobular carcinoma, Histologic Findings: See Fig. 4.47a–h. Note the differ- 3.6 cm with metastatic lobular carcinoma in one of three ence in cytologic, architectural, and immunophenotypes sentinel nodes (ypT2 pN1). Complete pathologic response (E-­cadherin, ER, PR, Her2) of the two concurrent tumors of invasive ductal carcinoma at the time of core biopsy. After chemotherapy, the grade Take-Home Messages: 3 ductal Her2-positive tumor had completely resolved, a b c d Fig. 4.47 Case 3. Concurrent invasive ductal and lobular carcinoma, markers: ER positive (2–3+ 80%), PR positive (2–3 + 80%), Her2/neu treated with chemotherapy. (a) Core biopsy of 9:00 mass, demonstrat- 1+ IHC, and Her2 FISH negative. (e) Tumor bed from site of invasive ing grade 3 invasive ductal carcinoma, with necrosis (upper right), and ductal carcinoma demonstrates complete pathologic response (40×); (b) lymph node metastasis with identical ductal histology. This carci- lymph nodes were negative for ductal carcinoma (not shown). (f) At the noma was E-cadherin positive (not shown), weakly positive for ER (1+ 12:00 site, there is residual invasive lobular carcinoma, perhaps of 30%), PR negative, and Her2 equivocal by immunohistochemistry and lower cellularity than seen on core biopsy (200×). (g) Lymph nodes amplified by FISH (Her2/CEP17 = 8, Her2/cell = 22). (c) Core biopsy contain residual metastatic lobular carcinoma (100×), seen at higher of ipsilateral 12:00 mass with grade 2 invasive lobular carcinoma in power (h) (200×), but there was no evidence of ductal carcinoma in sclerotic stroma. (d) This carcinoma is E-cadherin negative. Predictive nodes or breast 4 Lobular Breast Lesions 131 e f g h Fig. 4.47 (continued) –– Multiple simultaneous tumors can have disparate biol- performed, sampling calcifications, followed by wire ogy. Predictive markers may reveal differences. localized lumpectomy. The biopsy marker, residual calci- –– Classic lobular carcinoma responds poorly to cyto- fications, and localizing wires can be seen in the speci- toxic chemotherapy, whereas high-grade carcinomas men radiograph. Grossly, there was a 2 cm mass that was often respond better. 0.8 cm from the closest margin. –– Her2-negative tumors respond poorly to Her2-targeted Histologic Findings: See Fig. 4.49a–c. The core biopsy therapy. contained classic, florid, and pleomorphic LCIS with cal- cifications, without invasive carcinoma. The excisional Case 4 specimen revealed a 0.2 cm focus of invasive lobular car- History: A 49-year-old premenopausal G1P1 woman with cinoma, grade 2 (poor tubule formation (3), intermediate a maternal aunt with breast cancer with suspicious calci- nuclear grade (2), low mitotic rate (1), total of 6). fications on mammography Predictive marker immunohistochemistry was performed Imaging: See Fig. 4.48a–f. Suspicious 1.2 cm group of and reported on the invasive component: ER positive (3+ punctate and fine pleomorphic calcifications in the upper >95%), PR positive (3+ >95%), Ki-67 < 5%, Her2 2+ outer left breast, posterior depth. MRI was performed and immunohistochemistry, non-amplified by FISH. There showed at least three enhancing masses in the outer left was a background of extensive classic, florid, and pleo- breast in a clumped non-mass enhancement pattern cor- morphic LCIS. The pleomorphic LCIS was at one margin responding to, but larger than, the suspicious calcifica- and within 0.1 cm of another. Re-excision was performed, tions on mammography. Stereotactic core biopsy was containing further classic LCIS. Sentinel lymph node 132 M. L. Troxell et al. a b c d e f Fig. 4.48 Case 4. Craniocaudal (a) and lateral medial (b) magnifica- calcifications on mammography. (d) Computer-aided detection (CAD) tion full-field digital mammograms show a suspicious 1.2 cm group of MRI images show fast initial and late washout kinetics (red color), sus- punctate and fine pleomorphic calcifications in the upper outer left picious for cancer. (e) Core biopsy specimen radiography shows the breast, posterior depth (circles). Ultrasound from this can be seen in targeted calcifications (arrow). Post-biopsy mammogram from this case Fig. 4.28. (c) Images from axial contrast-enhanced breast MRI shows is shown in Fig. 4.28, including large calcifications. (f) Subsequent marked background parenchymal enhancement and at least 3–4 enhanc- specimen radiograph from the time of surgery shows the two localiza- ing masses (arrows) in the outer left breast in a clumped non-mass tion wires, omega marker (arrow) and residual calcifications (circle) enhancement pattern corresponding to, but larger than, the suspicious 4 Lobular Breast Lesions 133 a b c Fig. 4.49 Case 4. Histologic sections from the core biopsy and subse- men (40×). (b) Pleomorphic LCIS at top, and at bottom invasive lobular quent excision demonstrated LCIS: classic, florid, and pleomorphic carcinoma with cords and small nests of infiltrating tumor (100×). (c) (also see Fig. 4.28), with a small focus of invasive lobular carcinoma Serial section with negative E-cadherin immunostain (100×) apparent only on excision. (a) Florid LCIS from the excisional speci- biopsy was performed at the time of second surgery and Histologic Findings: See Fig. 4.51a–d. The core biopsy dem- nodes were negative. onstrates collagenous stroma with a haphazard infiltrate of Final Diagnosis: Invasive lobular carcinoma, grade 2, small dyscohesive epithelial cells, characteristic of invasive 0.2 cm (pT1a pN0), extensive classic, florid, and pleo- lobular carcinoma. Predictive marker results: ER positive morphic LCIS (3+ >95%), PR positive (2+ 30%), Ki-67 low (