Practical Breast Pathology: Yan Peng Ping Tang Editors

Practical Anatomic Pathology
Series Editors: Fan Lin · Ximing J. Yang
Yan Peng
Ping Tang Editors
Practical Breast
Pathology
Frequently Asked Questions
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

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 [1]. 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
© 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 [6]. 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 [11].
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 [9]. 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 [11]. expansion of the acini [14]. 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
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 [17]. 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 [1].
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
[18]. 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 [13]. However, there should be no true atypical micropapillary
structure for CCH [19]. 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 [20]. 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 [22].
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 [5].
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 [18]. 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 [18].
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%) [21].
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 [21]. 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
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 [23]. The
tubular carcinoma is poorly documented [21]. 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 [18]. be present with visible prominent nucleoli. Mitotic figures
may be seen [23].
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 [17]. 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
[3]. 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 [18]. 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 [19]. 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 [18].
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 [1].
29. Can FEA have mitosis?
27. How does one differentiate FEA from ADH?
Mitotic figures may be seen but are uncommon [18]. 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-
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%) [26].
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
[27]. 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 [21]. 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
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
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 ducts or <2 mm >2 ducts or >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 mm [21, 32–35, 46]. See Table 1.1. studies [47–51].
37. Can ADH have necrosis? Punctate and central? 40. What is grade 1 DCIS and what are its key histologic
features? (Fig. 1.6a–c)
ADH is usually associated with low-grade DCIS. Central
necrosis and punctate necrosis are usually associated with Ductal carcinoma in situ (DCIS) is a type of in situ breast
high-grade DCIS, but they can be rarely seen in ADH. The carcinoma in which tumor cells originate from ductal epi-
presence of necrosis or punctate necrosis does not change the thelium and are confined within ductal–lobular units with-
diagnosis of ADH [32–35]. out stromal invasion. The morphology of these tumor cells
ranges from monotonous with mild cytologic atypia to cells
38. Can ADH have mitosis? with marked nuclear pleomorphism. Tumor cells in grade 1
DCIS are monomorphic with only mild cytologic atypia.
Mitotic figures can be identified in ADH, but they are Nuclei are uniform size with inconspicuous nucleoli.
uncommon. There may be more mitoses in UDH compared Mitoses and necrosis are rare. It may be associated with
to ADH [32–35]. microcalcifications. To make a diagnosis of grade 1 DCIS,
we need to see that 1. qualitatively, tumor cells fill the entire
39. What types of breast lesions can ADH be associated ducts, without overlapping or streaming, or they form
with? punch-out spaces, cribriform, or micropapillary patterns;
and 2. these changes need to involve at least two ducts or
Atypical ductal hyperplasia (ADH) is not a type of breast involve 2 mm or above of the area. Architectural patterns of
cancer; instead, it is a marker indicating an increased risk for DCIS grades 1–3 can be solid, cribriform, micropapillary, or
developing breast cancer in the future. The risk of develop- comedo types [45, 46].
ing cancer includes both breasts but the risk is higher on the
ipsilateral breast. ADH is often associated with breast can- 41. What IHC markers can be helpful for the diagnosis
cer, most likely ductal carcinoma, either in situ or invasive. for grade 1 DCIS?
The diagnosis of ADH is most often treated by surgical exci-
sion or lumpectomy, and the diagnosis of ADH can be There are no IHC markers for making a diagnosis of grade 1
upgraded to DCIS or invasive carcinoma. The possibility of DCIS. However, if the differential diagnosis is usual ductal
finding carcinoma on surgery depends on the amount of hyperplasia, IHC of CK5/6 and ER can be of some help.
ADH identified on core biopsy and on associated radiologic Grade 1 DCIS like ADH is usually negative for CK5/6 and
findings. The more ADH identified on core needle biopsy, strong and diffusely positive for ER [11, 52].
losses at 16q, 17p, and 11q13, and gains at 1q. Loss of chro-
a
mosome 16 is the most common alteration [38–42].
43. How does one differentiate grade 1 DCIS from UDH?
The differences between DCIS and UDH are also mainly in

their morphology. Immunostains can be of some help in dif-
ficult cases. The morphologic differences between grade 1
DCIS and UDH include different cytologic and architectural
features. Cytologically, tumor cells appear uniform, monoto-
nous, and clonal with nearly fine chromatin pattern and
inconspicuous nucleoli. Mitoses are not common and necro-
sis, especially central necrosis, is rare. Architecturally, cells in
DCIS form solid nests or well-formed round or oval-shaped
spaces including cribriform, micropapillae, small tufts, or
b rigid Roman bridging. Cytologic features of UDH are differ-
ent from those of DCIS. UDH consists of heterologous cell
populations, and the nuclei of these cells are small and vari-
able in size and shape. These cells may have inconspicuous
nucleoli and have rare mitoses. They tend to be crowded and
haphazardly placed, with streaming and swirling. The growth
pattern of UDH can be solid, micropapillary, or fenestrated.
Fenestrated structures in UDH are ill formed, irregular
shaped, or slit like. The difference of IHC in grade 1 DCIS vs
UDH is the same as that of ADH vs UDH [11, 41, 42, 51].
44. How does one differentiate grade 1 DCIS from grade

2 DCIS?
Nuclear size and morphology: Tumor cells in grade 1 DCIS

c
are small and uniform (1–2x of normal breast epithelial
cells). They have small nuclei, normal chromatin pattern and
inconspicuous nucleoli. Mitoses and necrosis are very rare.
The features of tumor cells in grade 2 DCIS are in-between
grade 1 and grade 3 DCIS (>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
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)
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
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
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 (<1 mm) also present and one
Final Diagnosis: positive LN
• DCIS DDx:
Take-Home Messages: • IDC, multifocal, <1 mm to 5 mm
• Although attenuated myoepithelial cells most often • Then:
occur in high-grade DCIS, they can occur in low-grade • send to HER2 FISH, no invasive cancer
DCIS as well. Next Steps:
• Some myoepithelial marker is better than others; thus, • MEP present in the 5 mm focus; thus it is not IDC, but
use a panel of markers instend of single marker. DCIS with sclerotic changes
• Tumor markers repeated on the mets
Final Diagnosis:
Case 4: DCIS Versus IDC (Fig. 1.15a–e) • IDC, microinvasive (m), with positive node
Take-Home Messages:
History: A 40-year-old female with calcification on screen- • Cancerization of lobule with sclerotic changes can
ing mammogram underwent core needle biopsy mimic IDC
Histology: Extensive solid high-grade DCIS with central • Do not assume that IDC and DCIS in the same patient
necrosis; no prominent cancerization of the lobule present; are always morphologically different
a b
c d
e f
Fig. 1.14 Case 3: LG-DCIS with attenuated myoepithelial cells. A cially for p63. A high-power view (40X) confirms the low-grade nuclei
low-grade DCIS with cribriform and solid patterns (a), with low-power for this DCIS (d), while the calponin stains show complete presence
views of IHC analyses for myoepithelial markers calponin (b) and p63 (E-upper) and absence (E-lower) and partial presence (f) of staining
(c), which demonstrates the marked attenuated staining patterns, espe- patterns
22 X. Duan et al.
a b
c d
e f
Fig. 1.15 Case 4: DCIS vs IDC. Low-power view shows an area of staining clearly demonstrates the presence of myoepithelial cells around
small infiltrative glands with some larger glands with central necrosis at these glands, proving that this is an in situ lesion, i.e., a sclerotic appear-
the lower edge of the lesion, likely representing a mixed lesion with both ance of DCIS with cancerization of a lobule; a second look at (a) proved
IDC and DCIS components (a). Higher power shows that these small the lobular central appearance of this area (d). A separate area of classic
glands are angulated, with necrosis and associated with desmoplastic cancerization of the lobule by DCIS is shown for comparison (e, f)
stroma, all features suggesting an invasive process (b, c); however, p63
References 22. Dabbs DJ, Carter G, Fudge M, Peng Y, Swalsky P, Finkelstein

S. Molecular alterations in columnar cell lesions of the breast. Mod
Pathol. 2006;19:344–9.
1. Lakhani SR, Ellis IO, Schnitt SJ, Tan PH, van de Vijver 23. Yamashita Y, Ichihara S, Moritani S, et al. Dose flat epithelial
MJ. International Agency for Research on Cancer (IARC): WHO atypia has rounder nuclei than columnar cell change/hyperplasia?
classification of tumours of the breast, vol. 4. Lyon: IARC; 2012. A morphometric approach to columnar cell lesions of the breast.
2. Virnig BA, Tuttle TM, Shamliyan T, Kane RL. Ductal carcinoma in Virchows Arch. 2016;468:663–73.
situ of the breast: a systematic review of incidence, treatment, and 24. Go EM, Tsang JY, Ni YB, Yu AM, Mendoza P, Chan SK, et al.
outcomes. J Natl Cancer Inst. 2010;102(3):170–8. Relationship between columnar cell changes and low-grade car-
3. Nakhlis F, Morrow M. Ductal carcinoma in situ. Surg Clin North cinoma in situ of the breast – a cytogenetic study. Hum Pathol.
Am. 2003;83:821–39. 2012;43:1924–31.
4. Jacobs TW, Byrne C, Colditz G, Connolly JL, Schnitt SJ. Radial 25. Dialani V, Venkataraman S, Frieling G, Schnitt SJ, Mehta TS. Does
scars in benign breast-biopsy specimens and the risk of breast can- isolated flat epithelial atypia on vacuum-assisted breast core biopsy
cer. N Engl J Med. 1999;340:430–6. require surgical excision? Breast J. 2014;20:606–14.
5. Mooney KL, Bassett LW, Apple SK. Upgrade rates of high-risk 26. Haupt B, Schwartz MR, Xu Q, Ro JY. Columnar cell lesions:
breast lesions diagnosed on core needle biopsy: a single-institution a consensus study among pathology trainees. Hum Pathol.
experience and literature review. Mod Pathol. 2016;29:1471–84. 2010;41:895–901.
6. Owings DV, Hann L, Schnitt SJ. How thoroughly should needle 27. Racz JM, Carter JM, Degnim AC. Challenging atypical breast
localization breast biopsies be sampled for microscopic examina- lesions including flat epithelial Atypia, radial scar, and intraductal
tion? A prospective mammographic/pathologic correlative study. papilloma. Ann Surg Oncol. 2017;24:2842–7.
Am J Surg Pathol. 1990;14:578–83. 28. Rudin AV, Hoskin TL, Fahy A, Farrell AM, Nassar A, Ghosh K,
7. Johnson NB, Collins LC. Update on percutaneous biopsy of non- et al. Flat epithelial atypia on core biopsy and upgrade to can-
malignant breast lesions. Adv Anat Pathol. 2009;16:183–95. cer: a systematic review and meta-analysis. Ann Surg Oncol.
8. Landercasper J, Linebarger J. Contemporary breast imaging and 2017;24:3549–58.
concordance assessment. Surg Clin N Am. 2011;91:33–58. 29. McCroskey Z, Sneign N Herman CR, Miller RA, Venta LA, Ro JY,
9. The American Society of Breast Surgeons. Consensus guideline on et al. Flat epithelial atypia in directional vacuum-assisted biopsy of
concordance assessment of image-guided breast biopsies and man- breast microcalcifications: surgical excision may not be necessary.
agement of borderline or high-risk lesions. 2016, November. https:// Mod Pathol. 2018;31:1097–106.
www.breastsurgeons.org/new.../Concordance_and_High%20 30. Nakhlis F. How do we approach benign proliferative lesions? Curr
RiskLesions.pdf Oncol Rep. 2018;20:34.
10. Schnitt SJ, Collins L. Biopsy interpretation of the breast.
31. Calhoun BC, Sobel A, White RL, Gromet M, Flippo T, Sarantou
Philadelphia: Lippincott; 2012. T, et al. Management of flat epithelial atypia on breast core biopsy
11. Martinez AP, Cohen C, Hanley KZ, Li XB. Estrogen receptor and may be individualized based on correlation with imaging studies.
cytokeratin 5 are reliable markers to separate usual ductal hyper- Mod Pathol. 2015;28:670–6.
plasia from atypical ductal hyperplasia and low-grade ductal carci- 32. Simpson JF, Schnitt SJ, Visscher D, et al. Atypical ductal hyperpla-
noma in situ. Arch Pathol Lab Med. 2016;140:686–9. sia. In: Lakhani SR, Schnitt SJ, Tan PH, van de Vijver MJ, editors.
12. Kaneko M, Arihiro K, Takeshima Y, Fujii S, Inai K. Loss of hetero- WHO classification of tumors of breast. 4th ed. Lyon: IARC; 2012.
zygosity and microsatellite instability in epithelial hyperplasia of p. 88–9.
the breast. J Exp Ther Oncol. 2002;2:9–18. 33. Tavassli FA, Noeris HJ. A comparison of the results of long-term
13. Schnitt SJ. Benign breast disease and breast cancer risk: morphol- follow-up for atypical intraductal carcinoma of the breast. Cancer.
ogy and beyond. Am J Surg Pathol. 2003;27:836–41. 1990;65:518–29.
14. Shin SJ, Lal A, De Vries S, Suzuki J, Roy R, Hwang ES, et al. 34. Tavassli FA. Intraductal hyperplasia, ordinary and atypical. In:

Florid lobular carcinoma in situ: molecular profiling and compari- Tavassli FA, editor. Pathology of the breast. New York: Elsevier;
son to classic lobular carcinoma in situ and pleomorphic lobular 1992. p. 155–91.
carcinoma in situ. Hum Pathol. 2013;44:1998–2009. 35. Page DL, Rogers LW. Combined histologic and cytologic criteria
15. Bryan BB, Schnitt SJ, Collins LC. Ductal carcinoma in situ with for the diagnosis of mammary atypical ductal hyperplasia. Hum
basal-like phenotype: a possible precursor to invasive basal-like Pathol. 1992;23:1095–7.
breast cancer. Mod Pathol. 2006;19:617–21. 36. Otterbach F, Bànkfalvi À, Bergner S, Decker T, Krech R, Boecker
16. Steinman S, Wang J, Bourne P, Yang Q, Tang P. Expression of cyto- W. Cytokeratin 5/6 immunohistochemistry assists the differential
keratin markers, ER-alpha, PR, HER-2/neu, and EGFR in pure ductal diagnosis of atypical proliferations of the breast. Histopathology.
carcinoma in situ (DCIS) and DCIS with co-existing invasive ductal 2000;37:232–40.
carcinoma (IDC) of the breast. Ann Clin Lab Sci. 2007;37:127–33. 37. Wells JM, Liu Y, Ginter PS, Nguyen MT, Shin SJ. Elucidating
17. Tan PH, Ho BC, Selvariajan S, Yap WM, Hanby A. Pathological encounters of atypical ductal hyperplasia arising in gynaecomastia.
diagnosis of columnar cell lesions of the breast: are there issues of Histopathology. 2015;66:398–408.
reproducibility? J Clin Pathol. 2005;58:705–9. 38. Crissman JD, Visscher DW, Kubus J. Image cytophotometric DNA
18. Schnitt SJ, Collins LC. Biopsy interpretation of the breast, Biopsy analysis of atypical hyperplasias and intraductal carcinomas of the
interpretation series. 3rd ed. Philadelphia: Wolters Kluwer Health; breast. Arch Pathol Lab Med. 1990;114:1249–53.
2018. 39. O’Connell P, Pekkel V, Allred DC, Osborne CK, Clark GM, Allred
19. Collins LC. Precursor lesions of the low-grade breast neoplasia DC, et al. Analysis of loss of heterozygocity in 399 premalignant
pathway. Surg Pathol Clin. 2018;11:177–97. breast lesion at 15 genetic foci. J Natl Cancer Inst. 1998;90:697–703.
20. Simpson PT, Gale T, Reis-Filho JS, Jones C, Parry S, Sloane JP, 40. Ma XJ, Salunga R, Tuggle JT, Gaudet J, Enright E, McQuary P,
et al. Columnar cell lesions of the breast: the missing link in breast et al. Gene expression profiles of human breast cancer progression.
cancer progression? A morphological and molecular analysis. Am J Proc Natl Acad Sci USA. 2003;100:5974–9.
Surg Pathol. 2005;29:734–46. 41. Rubin E, Visscher DW, Alexander RW, Urist MM, Maddox

2 1. Hoda SA, Broji E, Koerner FC, Rosen PP. Rosen’s breast
WA. Proliferative disease and atypia in biopsies performed
pathology. 4th ed. Philadelphia: Lippincott Williams & for nonpalpable lesions detected mammographically. Cancer.
Wilkins; 2014. 1988;61:2077–82.
24 X. Duan et al.
42. Rosen PP. Ductal hyperplasia and intraductal hyperplasia. In: 53. Hall MJ, Reid JE, Wenstrup RJ. Prevalence of BRCA1 and BRCA
Rosen PP, editor. Breast pathology: diagnosis by needle core 2 mutations in women with breast carcinoma in situ and referred for
biopsy. Philadelphia: Lippincott Williams & Wikins; 1999. genetic testing. Cancer Prev Res. 2010;3:1579–85.
p. 89–92. 54. Simpson PT, Reis-Filho JS, Gale T, Lakhani SR. Molecular evolu-
43. Tavassoli FA, Rosai J, Holland R, et al. Intraductal proliferative tion of breast cancer. J Pathol. 2005;205:248–54.
lesions. In: Tavassoli FA, Devilee P, editors. Pathology and genetics 55. Abdel-Fatah T, Powe D, Hodi Z, Reis-Filho JS, Lee AH, Ellis
of tumours of the breast and female genital organs. Lyon: IARC IO. Morphologic and molecular evolutionary pathways of low
Press; 2003. p. 63–73. nuclear grade invasive breast cancers and their putative precur-
44. Yamaguchi R, Tanaka M, Tse GM, Yamaguchi M, Terasaki H, sor lesions: further evidence to support the concept of low nuclear
Akiba J, et al. Pure flat epithelial atypia is uncommon in subse- grade breast neoplasia family. Am J Surg Pathol. 2008;32:513–23.
quent breast excisions for atypical epithelial proliferation. Cancer 56. Buerger H, Mommers EC, Littman R, Simon R, Diallo R, Poremba
Sci. 2012;103:1580–5. C, et al. Ductal invasive G2 and G3 carcinomas of the breast are
45. Lavoue V, Roger CM, Poilblanc M, Proust N, Monghal-Verge C, the end stages of at least two different lines of genetic evolution. J
Sagan C, et al. Pure flat epithelial atypia (DIN 1a) on core nee- Pathol. 2001;194:165–70.
dle biopsy: study of 60 biopsies with follow-up surgical excision. 57. Rolance R, Gorman P, Harris W, Liebmann R, Barnes D, Hanby A,
Breast Cancer Res Treat. 2011;125:121–6. et al. Comparative genomic hybridization of breast tumors strati-
46. Schnitt SJ, Allred C, Britton P, et al. Ductal carcinoma in situ. In: fied by histological grade reveals new insights into the biological
Lakhani SR, Schnitt SJ, Tan PH, van de Vijver MJ, editors. WHO progression of breast cancer. Cancer Res. 1999;59:1433–6.
classifications of tumours of the breast. Lyon: IARC; 2012. p. 90–4. 58. Ashikari R, Park K, Huvos AG, Urban JA. Paget’s disease of the
47. Ely KA, Carter BA, Jensen RA, Simpson JF, Page DL. Core biopsy breast. Cancer. 1970;26:680–5.
of the breast with atypical ductal hyperplasia: a probabilistic 59. Berg JW, Hutter RV. Breast cancer. Cancer. 1995;75(1 Suppl):257–69.
approach to reporting. Am J Surg Pathol. 2001;25:1017–21. 60. Paget J. On the disease of the mammary areola preceding cancer of
48. Dupont WD, Page DL. Risk factors for breast cancer in women the mammary gland. St Bartholomew Hosp Rep. 1874;10:87–9.
with proliferative breast disease. N Engl J Med. 1985;312:146–51. 61. Dalberg K, Hellborg H, Wärnberg F. Paget’s disease of the

49. Jacobs TW, Connolly JL, Schnitt SJ. Nonmalignant lesions in
nipple in a population based cohort. Breast Cancer Res Treat.
breast core needle biopsies: to excise or not to excise? Am J Surg 2008;111:313–9.
Pathol. 2002;26:1095–110. 62. Chen CY, Sun LM, Anderson BO. Paget disease of the breast:
50. Deshaies I, Provencher L, Jacob S, Côté G, Robert J, Desbiens C, changing patterns of incidence, clinical presentation, and treatment
et al. Factors associated with upgrading to malignancy at surgery in the U.S. Cancer. 2006;107:1448–58.
of atypical ductal hyperplasia diagnosed on core biopsy. Breast. 63. Chaudary MA, Millis RR, Lane EB, Miller NA. Paget’s disease
2010;20:50–5. of the nipple: a ten-year review including clinical, pathological,
51. Margenthaler JA, Duke D, Monsees BS, Barton PT, Clark C, Dietz and immunohistochemical findings. Breast Cancer Res Treat.
JR. Correlation between core biopsy and excisional biopsy in breast 1986;8:139–46.
high-risk lesions. Am J Surg. 2006;192:534–7. 64. Mitchell S, Lachica R, Randall MB, Beech DJ. Paget’s disease
52. Niu F, Wang L, Zhang W, Lyu S, Niu Y. Value of CK5/6, CK14, of the breast areola mimicking cutaneous melanoma. Breast J.
ER and PR detection in differential diagnosis of intraductal 2006;12:233–6.
proliferative lesions of the breast. Zhonghua Zhong Liu Za Zhi. 65. Lloyd J, Flanagan AM. Mammary and extramammary Paget’s dis-
2015;37:749–52. ease. J Clin Pathol. 2000;53:742–9.
Invasive Ductal Carcinoma (NOS)
of the Breast 2
Xiaoxian Li, Zaibo Li, Xiaoyan Cui, and Yan Peng
List of Frequently Asked Questions those of primary breast carcinomas. Clinical information is
critical in these cases.
1. Should I worry about metastatic carcinoma to the Immunohistochemistry (IHC) studies can be helpful.
breast without in situ component? GATA3, mammaglobin, GCDFP-15, and estrogen receptor
(ER) are useful markers to identify breast carcinoma. It should
Metastatic carcinoma to the breast is always a concern when be noted that expression of all these markers is reduced par-
there is no ductal carcinoma in situ (DCIS) component iden- tially or completely in ER-negative breast carcinomas. Triple-
tified in the breast. Carcinoma from other origins metastasiz- negative breast carcinoma (TNBC) could lose all of these
ing to the breast is not common, and it happens in less than markers [4, 5]. On the other hand, a majority of these markers
2% of cases [1, 2]. The most common metastatic tumors to are not specific for primary breast carcinoma. GATA3 can
the breast are malignant melanoma and cancers from lung, express in urothelial carcinoma and many others [4, 6, 7].
ovary, stomach, and other gastrointestinal organs such as GCDFP-15 expression can be found in a subset of lung carci-
kidney, thyroid, and cervix. Prostate carcinoma metastasis to nomas, although it is considered as a relatively specific marker
the male breast was also reported [1, 2]. Metastatic tumor to for a breast origin [8]. ER expression is commonly seen in
the breast commonly forms a single mass lesion [2]. gynecological tumors [9]. Mammaglobin is a more sensitive
Occasionally, metastatic tumor to the breast can be the initial marker than GCDFP-15 for detecting breast carcinoma but
clinical presentation [3]. Some morphologic features may has lower specificity than GCDFP-15 [10–12]. IHC markers
provide diagnostic clues: for example, clear cell morphology commonly used to identify non-breast primary tumors can be
in metastatic renal cell carcinoma, papillary structure and expressed in breast carcinoma. Breast carcinoma can be occa-
psammoma bodies in metastatic ovarian serous carcinoma, sionally positive for TTF-1, S-100, and WT-1 [10, 13, 14].
and pigmentation in metastatic melanoma. However, these Breast carcinoma is generally negative for PAX-8 expression
morphological features may not be present or overlap with [15, 16], which is helpful to differentiate breast carcinoma
from thyroid, kidney, and Müllerian carcinomas. Although a
majority of the breast carcinomas are cytokeratin (CK)7 posi-
tive, up to 10% are negative for CK7 [17].
X. Li
Department of Pathology and Laboratory Medicine, In summary, metastatic tumors from other origins to the breast
Emory University, Atlanta, GA, USA are rare. Clinical information is always a key component to offer
e-mail: [email protected] accurate diagnosis. When in doubt, an initial IHC panel including
Z. Li · X. Cui GATA-3, ER, mammaglobin, TTF-1, and PAX-8 would be help-
Department of Pathology, ful. In an appropriate clinical setting, if an unknown primary tumor
The Ohio State University Wexner Medical Center,
is positive for GATA3 and ER and negative for TTF-1 and PAX-8,
Columbus, OH, USA
e-mail: [email protected]; [email protected] it is almost certain that it is a primary breast carcinoma. Sometimes,
distinguishing TNBC from poorly differentiated metastatic carci-
Y. Peng (*)
Department of Pathology, Clements University Hospital, noma to the breast can be challenging since TNBC can lose the
Dallas, TX, USA expression of all the breast carcinoma markers; clinical and radio-
Department of Pathology, graphic correlation and a broader IHC panel are essential to make
University of Texas Southwestern Medical Center, Dallas, TX, a correct diagnosis. IHC studies may have very limited value to
USA differentiate breast carcinoma from skin adnexal tumors; clinical

26 X. Li et al.
presentation may be helpful in such situations. Rarely, high-grade invasive carcinoma with neuroendocrine differentiation as “all
lymphoma and epithelioid sarcoma such as angiosarcoma can be tumors express neuroendocrine markers to a greater or a lesser
seen in the breast. degree” [24]. The 2012 WHO tumor classification indicates that
up to 30% of invasive carcinoma of no special type (NST)
2. Are all triple-negative breast cancers (TNBCs) basal- shows neuroendocrine differentiation (with any percentage of
like subtype in the molecular classification? tumor cells). However, the exact proportion of breast carcino-
mas with neuroendocrine differentiation is not clear due to lack
Triple-negative breast cancer (TNBC) is a clinical classifica- of studies. The neuroendocrine differentiation is more common
tion characterized by negative estrogen receptor (ER) and in solid papillary carcinoma and invasive mucinous carcinoma.
progesterone receptor (PR) expression, and the absence of Wei and coauthors studied 74 patients with primary mammary
HER2 protein overexpression and HER2 gene amplification. carcinoma with neuroendocrine differentiation using the crite-
Basal-like breast cancer is an intrinsic subtype based on gene rion by 2003 WHO tumor classification [25]. They found that
expression profile [18]. By gene expression profile studies, most of the mammary carcinomas with neuroendocrine features
breast cancers have at least four intrinsic subtypes including were ER positive (91.9%) or PR positive (68.9%). Only 2.7%
Luminal A, Luminal B, HER2 enriched, and basal like [18]. were HER2 positive. They showed that mammary carcinoma
Basal-like breast cancers have high expression of basal-type with neuroendocrine differentiation had worse recurrence-free
cytokeratins (CK5, 6, and 17) and are associated with a high survival and overall survival than invasive carcinoma of NST
frequency of p53 mutation. The majority of BRCA gene even when controlled for age, ethnicity, tumor stage, surgical
mutation–related breast cancers are basal-like subtype [19]. procedure, and HER2 status. However, such observation was
Recent studies identified another subtype by gene profile not seen in other studies [26]. Until more solid evidence from
study: claudin-low subtype [20]. Claudin-low breast cancers prospective studies are available, clinicians currently treat breast
are characterized by downregulation of tight junction pro- carcinoma with neuroendocrine features in the same fashion as
teins and high expression of genes associated with epithe- they treat breast carcinoma of NST. However, breast carcinoma
lial–mesenchymal transition (EMT) and breast cancer stem with neuroendocrine differentiation should be differentiated
cells [19, 20]. The claudin-low subtype includes the majority from metastatic neuroendocrine carcinoma of other origins.
of the spindle sarcomatoid metaplastic carcinoma and is
resistant to conventional chemotherapy. The majority of 4. What is microinvasive carcinoma?
basal-like and claudin-low subtypes are clinically TNBC but
the overlap is not 100%. Approximately 50–80% of TNBCs Microinvasive carcinoma is defined as invasive carcinoma that
are basal like, 10–30% are HER2 enriched, and a small num- measures no more than 1 mm in the greatest dimension. Tumor
ber of TNBCs are even luminal subtypes [19, 21]. Most with multiple foci of microinvasive carcinoma is still staged as
TNBCs with CK5/6 and epidermal growth factor receptor microinvasive carcinoma (pT1mi). The size of each focus
expression are basal-like subtype [22]. Therefore, there is a should not be added up. However, when two or more foci are
significant overlap between basal-like and TNBC subtypes, close, deeper H&E levels should be obtained to rule out the
but they are not completely inclusive of each other. So far, possibility of a larger invasive carcinoma, which may change
treatments are based on the status of ER, PR, and HER2 tumor pathology stage and patient managements. Microinvasion
expression determined by pathologists using immunohisto- is characterized by stromal infiltration of irregular tumor clus-
chemistry or fluorescence in situ hybridization (FISH) analy- ters or single tumor cells. Although the presence of myoepithe-
sis. Clinical trials are ongoing to tailor the best therapies for lial cells precludes the diagnosis of microinvasion, the absence
different subtypes of breast cancer including TNBC. of myoepithelial cells does not warrant a diagnosis of microin-
vasion. High-grade ductal carcinoma in situ (DCIS) with dis-
3. What is the significance of neuroendocrine differenti- torted architecture or tangential cut may result in small tumor
ation in invasive carcinoma? cell clusters without myoepithelial cells. These tumor clusters
generally have a regular and rounded contour. A diagnosis of
Primary mammary carcinoma with neuroendocrine features is suspicious for microinvasive carcinoma could be rendered
not uncommon. The 2003 World Health Organization (WHO) when it is difficult to give definite diagnosis, and surgeons will
Pathology and Genetics of Tumours of the Breast and Female generally treat the patient as positive for microinvasion. See
Genital Organs defines primary neuroendocrine carcinoma as Fig. 2.1a–c showing microinvasive carcinoma.
50% of the tumor cells showing neuroendocrine differentiation
with positive expression of either synaptophysin or chromo- 5. What is the prognosis of microinvasive carcinoma?
granin [23]. By this criterion, primary mammary neuroendo-
crine carcinoma is rare and represents <1% of breast carcinoma. Microinvasive carcinoma has excellent prognosis. Patients
However, this criterion was removed from the 2012 WHO with DCIS and associated microinvasion have similar recur-
Classification of Tumors of the Breast, which simply defines rence and overall survival as patients with DCIS only [27–30].
2 Invasive Ductal Carcinoma (NOS) of the Breast 27
Fig. 2.1 Microinvasive
carcinoma microinvasion is a
present at the periphery of
DCIS and the size is less than
1.0 mm (a); immunostains for
p40 (b) and SMMS (c) show
the presence of myoepithelial
cells around DCIS and the
absence of myoepithelial cells
around microinvasion
Compared to DCIS with single focus of microinvasion, invasive carcinomas have a lymph node metastasis [30–34].
DCIS with multifocal microinvasive carcinoma does not The majority of the lymph node metastases are either iso-
have an increased risk of axillary lymph node metastasis or lated tumor cells (ITCs) or micrometastasis. The benefit of
worse patient outcomes [30, 31]. sentinel lymph node biopsy in patients with microinvasive
carcinoma is unclear. There is no prospective clinical trial
6. Should a sentinel lymph node biopsy be performed in comparing the prognosis in patients with or without sentinel
patients with microinvasive carcinoma? lymph node biopsy. The rate of lymph node macrometastasis
is low in DCIS cases with only microinvasion. Pathologists
Sentinel lymph node biopsy in patients with microinvasive should extensively sample the specimen to avoid missing a
carcinoma is controversial. Approximately 3–15% of micro- large invasive carcinoma in such situations.
28 X. Li et al.
7. Should I order ER, PR, and HER2 on microinvasive microinvasive carcinoma have similar nodal metastasis rate
carcinoma? and prognosis [30, 31].
ER and PR should be ordered on ductal carcinoma in situ 10. Can estrogen receptor (ER)-positive breast cancer be
(DCIS) component in cases with DCIS and associated microin- a high Nottingham histologic grade?
vasion. However, HER2 testing is generally not ordered on
microinvasive carcinoma unless in clinical trials. DCIS with ER-positive breast cancers are generally low or intermediate
microinvasion has similar rates of recurrence and/or distant grade (Nottingham histologic grade 1 or 2). However,
metastasis as DCIS without microinvasion [27–30]. HER2 Nottingham grade 3 ER-positive breast cancer can be seen.
expression in the microinvasive carcinoma is not associated By molecular classification, ER-positive breast cancer
with recurrence [35]. There is no evidence to treat microinvasive includes Luminal A and Luminal B subtypes. The Luminal A
carcinoma with anti-HER2 therapies. Therefore, HER2 test is is generally ER+/HER2- with low proliferation, and the
not recommended on microinvasive carcinoma without lymph Luminal B is ER+/HER2+ or ER+/HER2- with high prolif-
node macrometastasis to avoid overtreatment. In cases of micro- eration index. Compared with the Luminal A breast cancer,
invasive carcinoma and lymph node macrometastasis, HER2 the Luminal B subtype is more aggressive and may benefit
test could be ordered on the nodal metastatic carcinoma. from chemotherapy.
8. Should microinvasive carcinoma be treated with 11. When should Oncotype DX® (Genomic Health,
chemotherapy? Redwood City, CA, USA) test be ordered?
DCIS with microinvasion has a similar prognosis as DCIS Oncotype DX recurrence score (RS) test measures mRNA
without microinvasion [27–30]. Chemotherapy is generally expression levels of 21 genes including 16 cancer related genes
not recommended in patients with microinvasive carcinoma. and 5 housekeeping genes from formalin-fixed paraffin-
However, physicians may treat patients with chemotherapy embedded (FFPE) tissues containing invasive breast cancer
when microinvasive carcinoma is associated with lymph [36]. The RS ranges from 0 to 100. The scores are categorized
node macrometastasis. into low (score < 18), intermediate (18–30), and high (>30). The
American Society of Clinical Oncology (ASCO) recommended
9. What is the prognosis of ductal carcinoma in situ that the RS may be used in ER/PR-positive, HER2-negative,
(DCIS) with multifocal microinvasive carcinoma? and lymph node–negative breast cancer [37]. The recommenda-
tion is based on several large prospective clinical trials [38–40].
The number of focus of microinvasive carcinoma does not ASCO does not recommend using the RS in patients with
adversely impact prognosis compared to single focus of ER-positive, HER2-negative, and lymph node–positive breast
microinvasive carcinoma. Matsen and colleagues reviewed cancer [37]. The RxPONDER trial (Rx for Positive Node,
414 cases with ductal carcinoma in situ (DCIS) with micro- Endocrine Responsive Breast Cancer) is accruing patients in an
invasion [31]. Of the 414 cases, 235 (57%) had one focus of attempt to set up a cutoff of the RS for potential benefits of che-
microinvasion and 179 (43%) had 2 or more foci of micro- motherapy in those patients with one to three positive lymph
invasion. Lymph node macrometastasis was found in 1.4% nodes. The ASCO also recommended not using the RS to guide
of the cases and micrometastasis in 6.3%. The frequency of treatments in patients with HER2-positive breast cancer. [37]
lymph node metastasis (micro or macro) was not different
between cases with 1 focus and 2 or more foci of microin- 12. What are the other tests in estrogen receptor-positive
vasion [31]. Shatat and colleagues examined 40 cases with breast cancers?
DCIS with microinvasive carcinoma. Twenty-eight of the
40 cases had 3 or less foci of microinvasive carcinoma, 5 Although Oncotype DX recurrence score (RS) test is the
had 4–9 foci, and 3 had 10 or more foci of microinvasive most widely used molecular test for ER-positive breast can-
carcinoma. A majority of the cases had high-grade cers, there are other tests available for these tumors, includ-
DCIS. Lymph node status was available in 35 cases: three ing MammaPrint (Agendia, Amsterdam, The Netherlands),
had isolated tumor cells (ITCs) in the lymph node (the PAM50, Breast Cancer Index® (Biotheranostics, San Diego,
patients had 3, 3, and 10 foci of microinvasion in the breast, CA, USA), Genomic grade index, EndoPredict® (Myriad,
respectively); 1 had micrometastasis (the patient had 4 foci Salt Lake City, UT, USA), IHC4 assay, and Magee
of microinvasion in the breast), and 1 had macrometastasis Equations™ [36, 37, 41]. ASCO recommends using some of
(the patient had 1 focus of microinvasion in the breast). these tests in ER-positive, HER2-negative, and lymph node–
These studies indicate that the status of nodal metastasis is negative breast cancers. [37] It should be noted that the
not associated with the number of foci of microinvasion in Magee equation has a very high concordance rate with the
the breast. Compared to patients with a single focus of RS [41] and it is available online: http://path.upmc.edu/onli-
microinvasive carcinoma, patients with multiple foci of neTools/MageeEquations.html.
13. Should Ki-67 be ordered in ER-positive breast

a
cancer?
ER-positive breast cancers are classified as Luminal A and

Luminal B subtypes by gene expression profiling. The
Luminal B breast cancer is characterized by high prolifera-
tive index. Patients with Luminal B type of ER-positive
breast cancer may benefit from chemotherapy. The 2011 St.
Gallen Breast Cancer Consensus Conference suggested
using 14% as a cutoff to separate Luminal A from Luminal B
subtype [42]. However, there is no universally accepted
Ki-67 cutoff because of the low reproducibility [43, 44].
Although Ki-67 is not recommended to be used as the sole
parameter to guide treatment for ER-positive, HER2-negative
breast cancer patients [37], as part of tumor profile, it may
provide additional information about tumor biology and b
tumor aggressiveness.
14. Where should we look for lymphovascular invasion

in invasive breast cancer?
Lymphovascular invasion in breast carcinoma is associated

with distant metastasis and survival independent of lymph
node metastasis [45–47]. It is not necessary to separate tumor
emboli in blood vessels or lymphatic channels. The lympho-
vascular invasion should be evaluated outside the invasive
carcinoma border and is usually found within 1–2 mm dis-
tance from the invasive tumor front. The tumor cells in the
vessels generally do not conform to the exact contour of the
vessels, and endothelial cells should be seen lining the lym-
phovascular spaces. The lymphovascular invasion should not c
be evaluated within the invasive carcinoma; the significance
of lymphovascular invasion within the carcinoma is not clear.
15. How do we measure the size of invasive carcinoma

arising from solid papillary carcinoma of the breast?
Solid papillary carcinoma is a well circumscribed tumor

with compact tumor cells. Delicate fibrovascular networks
are seen within the tumor. The tumor cells can have
neuroendocrine differentiation and mucin production. The
majority of the solid papillary carcinomas are ER positive,
PR positive, and HER2 negative [48, 49]. Solid papillary car-
cinoma can have myoepithelial cells at the periphery.
However, some very expanded tumors may lack myoepithe-
lial cells at the periphery [50]. It is thought that the myoepi- Fig. 2.2 Invasive carcinoma arising from solid papillary carcinoma.
Invasion is present at the periphery of solid papillary carcinoma (a);
thelial cells in a very expanded solid papillary carcinoma are immunostains for p40 (b) and SMMS (c) show the absence of myoepi-
too attenuated to be detected. Solid papillary carcinoma is thelial cells around both solid papillary carcinoma and invasive carci-
generally considered an in situ carcinoma although rare noma components
cases can have lymph node metastasis [50]. The invasive car-
cinoma arising from solid papillary carcinoma generally has
frank infiltrating growth pattern and stromal invasion. For papillary carcinoma even when the myoepithelial cells are
pathologic staging purpose, the size of invasive carcinoma absent. Overestimating the invasive carcinoma size may
should be measured on the invasive component only and result in false upstage of the tumor. See Fig. 2.2a–c showing
should not include the well-circumscribed portion of solid invasive carcinoma arising from solid papillary carcinoma.
30 X. Li et al.
16. H
ow do we evaluate tumor response to neoadjuvant • Residual disease may have scattered multiple tumor foci
therapy in invasive breast carcinoma? in a large tumor bed. It is recommended to document the
two dimensions of the largest cross section of the area
Triple-negative and HER2-positive breast cancers are more involved by residual invasive and average residual tumor
likely to be treated with neoadjuvant (pre-surgery) chemo- cellularity.
therapy (NACT). The purposes of the treatment are to down- • All surgically removed lymph nodes must be entirely
stage the tumor, pursue better cosmetic outcome, and, submitted.
importantly, evaluate tumor response to targeted therapy or • The number of positive lymph nodes, micro- and macro-
chemotherapy in breast resection specimens. Three parame- metastases, size of largest metastasis, and isolated tumor
ters are included to evaluate breast cancer response to the cells should be documented.
NACT: tumor bed size in two dimensions, residual invasive • Lymph nodes with treatment effect only without residual
carcinoma cellularity, and lymph node status. Of the three carcinoma should be regarded as negative lymph nodes.
parameters, lymph node status has the highest weight in pre- • Residual lymphovascular invasion should be documented
dicting cancer prognosis. There are four categories of resid- and is not classified as pCR.
ual cancer burden (RCB) indexed from 0 to III. RCB-0 is
also referred to as pathologic complete response (pCR), Repeat testing of ER, PR, and HER2 is not required but
which is defined as no residual invasive carcinoma in the may be helpful.
breast and lymph node at the time of surgery. Residual carci-
noma in situ in the breast is allowed in the pCR category. 18. How should breast carcinoma be staged after neoad-
Cancers in the RCB-I class have minimal amount of residual juvant therapy?
cancer burden in the breast and lymph nodes. Breast cancer
patients in both the pCR or the RCB-I category generally The eighth edition of the American Joint Cancer Committee
have good prognosis [51, 52]. More information about resid- (AJCC) cancer staging manual recommends staging residual
ual cancer burden can be found at: http://www3.mdander- breast cancer after NACT based on the largest focus when
son.org/app/medcalc/index.cfm?pagename=jsconvert3 multi-foci are seen [54]. The residual cancer burden (RCB)
method measures the largest dimension of the area involved
17. How do we measure residual cancer tumor bed in a by viable cancer cells [53]. The prognostic value of the RCB
surgical breast specimen post-NACT? tumor bed size has been validated in several studies [51, 52].
The RCB tumor bed evaluation is likely a better method to
Because the evaluation of breast cancer to NACT has been assess degree of tumor response to NACT, compared to the
widely used to predict patient prognosis and by the AJCC staging system. To avoid confusion stemming from the
U.S. Food and Drug Administration (FDA) to approve new two different methodologies, an e xplanation note in the pathol-
drugs, it is imperative to standardize the protocol for evalu- ogy report would be helpful for the patients and clinicians.
ation of breast specimens post-NACT. Recommendations
were given by an international multidisciplinary working 1 9. If there is no residual carcinoma identified in breast
group organized by the Breast Cancer International Group resection specimens post neoadjuvant therapy,
and the North American Breast Cancer Group (BIG- should the entire specimen be submitted?
NABCG) [53]. Recommendations are summarized as
follows: The international multidisciplinary working group organized
by the Breast Cancer International Group-North American
• Post-neoadjuvant systemic therapy specimen should be Breast Cancer Group (BIG-NABCG) recommends taking
correlated with pretreatment clinical and imaging find- five sections of the largest tumor bed in a slice. Sampling of
ings for pretreatment tumor size and location. A biopsy adjacent fibrotic areas with a total of 25 sections should suf-
clip placement is strongly recommended at the time of fice [53]. However, it is critical to correlate with breast imag-
diagnostic biopsy. ing studies to identify the location of the pretreatment cancer
• Systematic sampling of areas to include the largest cross and biopsy clips. When a tumor has a good response to
section of the tumor bed is preferable to exhaustive sam- NACT, it may be hard to grossly identify the tumor area, and
pling. Five blocks of the largest cross section of the tumor the biopsy clip can migrate from where it is initially placed.
bed, with a maximum of ~25 blocks, are sufficient. Therefore, it is also important to recognize the histologic
Mapping of sections taken is strongly recommended. findings and features of tumor bed on microscopic evalua-
• If multiple lesions are seen grossly, sections should be tion to ensure that the correct tumor bed area is sampled. The
taken from each lesion and additional sections should be tumor bed usually shows treatment effects including myxoid
taken in between these lesions to determine if they are changes, chronic inflammation, abundant foamy macro-
truly multiple lesions. phages, and hemosiderin deposition in the stroma. If these
tumor bed changes cannot be found, completely submitting tional chemotherapy [59], the spindle cell metaplastic carci-
the entire possible tumor bed area for further microscopic noma generally does not respond to chemotherapy and has a
evaluation merits consideration. worse prognosis when compared to other subtypes of triple-
negative breast cancer [60, 61].
20. How do we define pathologic complete response
(pCR)? 22. Is triple-negative breast cancer a homogenous dis-
ease group?
Currently in the United States, pathologic complete response
(pCR) is defined as no invasive carcinoma in the breast and The gene expression profiling studies classified breast cancer
no lymph node metastasis at the time of surgery after neoad- into Luminal A, Luminal B, HER2-enriched, and basal-like
juvant therapy [19, 51, 52]. Patients with pCR have good subtypes. The majority of the basal-like tumors have triple-
prognosis. If only residual carcinoma in situ is seen in the negative immunophenotype. Further studies discovered that
breast without any invasive component and the lymph nodes triple-negative breast cancer (TNBC is not a homogenous
are negative, the tumor is still classified as pCR. Occasionally, disease group. Initial studies identified 6 molecular subtypes
tumor emboli in lymphovascular spaces are the only finding of TNBC by gene profiling [62]. The six defined subtypes
in breast resection specimens after neoadjuvant therapy. were basal-like 1 and 2 (BL1 and BL2), immunomodulatory
Cheng and colleagues identified 6 patients with only lym- (IM), mesenchymal (M), mesenchymal stem like (MSL),
phovascular invasion after neoadjuvant therapy. Five of the 6 and luminal androgen receptor (LAR). By laser capture
patients developed distant metastasis [55]. Rabban and col- microdissection studies, the same group later found that the
leagues identified six patients with pure lymphovascular transcripts in the IM and MSL subtypes were contaminations
invasion in the breast without any residual invasive carci- from infiltrating lymphocytes and peritumoral stromal cells
noma. Five of the 6 patients also had lymph node metastasis [63]. Therefore, the subtypes of TNBC were refined from six
at the time of surgery, and four developed distant metastasis to four subtypes, referred to as TNBCtype-4 [63]. The LAR
[56]. Although the patient samples were small, these studies subtype has the intrinsic luminal gene signature and low pro-
suggested that patients with residual pure lymphovascular liferative rate with higher AR expression by immunohisto-
invasion post-NACT may have a worse prognosis than those chemistry (IHC) studies. The majority of BL1 and some BL2
without any residual carcinoma in the breast or lymphovas- breast cancers share similar gene signature with the intrinsic
cular space. Therefore, breast cancers with residual pure basal-like subtype and are associated with BRCA mutations.
lymphovascular invasion after NACT are currently not clas- The different TNBC subtypes had different clinical out-
sified as pCR [53]. comes [62, 64]. Masuda and coauthors showed that the BL1
subtype had the highest rate of pathological complete
21. What is claudin-low breast cancer? response (pCR) (52%) when treated with neoadjuvant che-
motherapy. The LAR had a low pCR rate (10%) but the best
Breast cancer is no longer regarded as a single disease. Gene overall survival. The M subtype had the worst survival [64].
expression studies have classified breast carcinoma into TNBC is a heterogeneous disease and has a complex cancer
intrinsic subtypes including Luminal A, Luminal B, HER2- biology. This concept helps understand that TNBC is diffi-
enriched, and basal-like subtypes [18]. The basal-like sub- cult to treat and has the worst outcome compared to other
type has a high expression of cytokeratins 5, 6, and 17, which subtypes of breast cancer. Developing effective targeted ther-
are usually expressed in the basal epithelial cells of normal apy for patients with TNBC is critical to improve their
tissues. The majority of the basal-like subtype tumors are survival.
triple-negative breast cancer. Recent studies identified a new
subtype of breast cancer by gene expression profiling, the 23. What is the significance of tumor infiltrating lym-
claudin-low subtype [20, 57]. The claudin-low breast cancer phocytes (TILs)?
is characterized by downregulation of tight junction proteins
including E-cadherin and some claudins, and is reportedly The immune system constantly scrutinizes abnormal cells
associated with poor survival. The claudin-low subtype has a and helps to control cancer initiation and development [65].
high expression of epithelial–mesenchymal transition There is mounting evidence showing the important role of
(EMT), immune response, and breast cancer stem cell– tumor infiltrating lymphocytes (TILs) in breast cancer prog-
related genes [20, 57]. Claudin-low tumors generally display nosis and response to therapies [66–69]. In addition to the
a triple-negative phenotype; however, only a small number innate surveillance role of the immune system, immunologi-
of triple-negative breast cancers are claudin-low tumors. cal response to control residual disease may be triggered by
This subtype of breast cancer includes a majority of the spin- radiotherapy or chemotherapy [70, 71]. It seems that TILs are
dle cell sarcomatoid metaplastic carcinoma [57, 58]. Similar associated with better therapy response, disease-free survival
to the breast cancer stem cells, which are resistant to conven- (DFS), and overall survival in triple-negative and HER2-
32 X. Li et al.
positive breast cancers but not in ER-positive cancer [66–69, tissue. Extranodal extension is measured by the largest
72–74]. High levels of TILs are significantly associated with dimension in the tissue outside the nodal capsule. Extranodal
high pathologic complete response (pCR) rate in both triple- extension is associated with increased tumor burden in the
negative and HER2-positive breast cancers in neoadjuvant axillary lymph nodes [75, 76]. Extranodal extension larger
therapy settings [69, 73, 74]. High TIL levels are also associ- than 2 mm is especially associated with increased axillary
ated with better disease-free survival and overall survival in lymph node metastasis. Therefore, the presence and size of
triple-negative and HER2-positive breast cancers in adjuvant extranodal extension should be reported. Reporting the pres-
therapy settings [67, 68, 72]. Although currently TILs are not ence and size of extranodal extension is especially important
recommended as a marker to prescribe or withhold chemo- for stage T1 and T2 tumors with <3 positive lymph nodes.
therapy in breast cancer treatment [37], it is conceivable that These tumors meet the criteria of the American College of
in the near future, evaluation of TILs may become an integral Surgeons Oncology Group (ACOSOG) Z0011 trial [75–77].
part of the pathologic breast cancer protocol checklist. Patients who meet the Z0011 trial criteria may undergo
lumpectomy and whole breast radiation therapy without
24. How do we score tumor infiltrating lymphocytes axillary dissection even with one or two positive axillary
(TILs)? lymph nodes. However, if these patients have an extranodal
extension (especially larger than 2 mm in size), axillary
With the emerging role of tumor infiltrating lymphocytes lymph node dissection may be needed to improve outcomes.
(TILs) in breast cancer response to therapy and prognosis, an [75, 76]
international TILs working group was convened and pro-
posed recommendations to standardize evaluation of TILs in 26. What is HER2 intratumoral heterogeneity (ITH)
breast cancer in clinical practice, translational research, and and what is the associated clinical significance?
clinical trials [72]. Recommendations are summarized as
follows: HER2 intratumoral heterogeneity (ITH) describes the coex-
istence of multiple tumor cell subpopulations with varying
• Stromal TIL evaluation is easier and more reproducible HER2 status within the same tumor, and has been found in
than intratumoral TIL assessment and therefore should be up to 40% of breast cancers [78–81]. Previously, HER2
used. genetic heterogeneity was categorized into either of two
• TIL evaluation is reported as % stromal TILs, which is the types based on heterogeneity distribution: clustered
area of stromal tissue occupied by mononuclear inflam- (regional) heterogeneity, which is defined as segregated
matory cells over the total stromal area. HER2 amplified tumor cells and non-amplified HER2 tumor
• TILs should be evaluated within the invasive carcinoma cell populations; mosaic (intermixed) heterogeneity, which
border but not outside the invasive carcinoma nor around is defined as comingled HER2 amplified and non-amplified
ductal carcinoma in situ and normal breast tissue tumor cells [82]. Studies have demonstrated a wide range of
components. prevalence of HER2 ITH in breast cancer, ranging from 5%
• Exclude areas with crush artifacts, necrosis, or biopsy site to 40% [83–87]. However, all these studies investigated
changes. HER2 ITH on the basis of only one method, either HER2
• All mononuclear cells including lymphocytes and plasma immunohistochemistry (IHC) or in situ hybridization (ISH)
cells should be scored, but polymorphonuclear leukocytes assay. Our recent studies have demonstrated that it is both
should be excluded. easier and more accurate to examine HER2 ITH by a novel
• Average TILs in the tumor area but not hot spots should gene protein assay (GPA), which combines HER2 IHC and
be used in the TIL evaluation. ISH on a single slide to allow identification of discordance
• One formalin fixed paraffin-embedded (FFPE) section between HER2 gene amplification and protein overexpres-
(4–5 μm thick, magnification x 200–400) per case is suf- sion [88, 89]. A recent study by Kurozumi and coauthors
ficient and full sections are preferred over biopsies. demonstrated that 17.2% (34/198) of HER2 IHC– and
• TIL evaluation should be assessed as a continuous HER2 ISH–negative cases showed ITH by GPA [90]. They
parameter. categorized the patterns of the HER2 protein and HER2
gene status analyzed by GPA into six types: (A) IHC & ISH
positive, (B) IHC positive & ISH negative, (C) IHC equivo-
25. What is the clinical significance of reporting extrano- cal & ISH positive, (D) IHC equivocal & ISH negative, (E)
dal extension? IHC negative & ISH positive, and (F) IHC & ISH negative.
Cases with at least two of the six types of HER2 gene and
Extranodal extension is defined as tumor cells invading protein status combinations were categorized as HER2 het-
through the lymph node capsule into the surrounding soft erogeneous [90].
HER2 ITH may contribute to inaccurate assessment of Additionally, several specific histologic types of breast
HER2 status, affecting treatment decisions, and thereby carcinoma including invasive tubular carcinoma (ITC),
patients’ clinical outcome. Reported findings in regard to invasive mucinous carcinoma (IMC), and classical invasive
this issue remain sparse. In one study, there was no signifi- lobular carcinoma (CILC) are low-grade tumors and are
cant difference in disease-free survival (DFS) in patients associated with a good prognosis. Almost all low-grade
with ITH in up to 30% of HER2 positive tumor cells as com- invasive breast carcinomas are ER-positive, PR-positive, and
pared to those without ITH [91]. Another study demonstrated HER2-negative and therefore may undergo Oncotype DX
that HER2 ITH was independently associated with inferior testing according to the National Comprehensive Cancer
DFS as compared with tumors with homogeneous HER2 Network (NCCN) guidelines. Studies including ours have
amplification. It has been reported that HER2 ITH might be indicated that these special types of tumors are unlikely to
associated with reduced DFS in primary and metastatic have a high RS score [97]. However, if clinicians do not feel
HER2-positive breast cancers treated with adjuvant trastu- comfortable making a management decision without an RS,
zumab [78, 92, 93]. Our recent data have demonstrated that the Magee equation may be an alternative method to stratify
HER2 ITH is associated with incomplete response to anti- these low-grade invasive breast carcinomas. If the Magee
HER2 neoadjuvant chemotherapy and this association is equation RS is less than 18, Oncotype DX testing is unlikely
independent of other factors [94]. to add any more information. It should be noted that a tumor
with prominent inflammation may have a false high Oncotype
27. What is Magee equation recurrence score (RS)? DX score.
Magee equations are derived by linear regression analysis 29. Should the tumor profile (ER, PR, and HER2) be
using several pathologic variables and semiquantitative repeated in breast specimens post-neoadjuvant
immunohistochemical results of ER, PR, HER-2, and Ki-67 therapy?
to calculate an RS that highly correlates with the Oncotype
DX RS and provides similar information as Oncotype DX Currently, there are no guidelines regarding whether residual
[41, 95]. Three Magee equations (#1, #2, and #3) using dif- tumor after neoadjuvant therapy should be retested for breast
ferent combinations of standard histopathologic variables cancer biomarkers. Studies have demonstrated wide ranges of
(Nottingham score (NS), ER, PR, HER-2, Ki-67, and tumor discrepancy of biomarker status [99–105]. The median fre-
size) have been previously reported to predict the Oncotype quencies of change reported in literatures are 13% for ER, 21%
DX RS, available at the website of University Pittsburgh for PR, and 12% for HER2 [99–105]. There are much higher
Medical Center Magee Women’s Hospital pathology depart- rates of changes from positive to negative in all three biomark-
ment: http://path.upmc.edu/onlineTools/mageeequations. ers than that from negative to positive. The average rates of
html. Many studies have demonstrated excellent concor- changes from negative to positive are 7% for ER, 7% for PR,
dance between the Magee RS and Oncotype DX RS and 3% for HER2 [99–105]. Some of the changes in biomarker
[96–98]. status are likely due to the receipt of neoadjuvant therapy (loss
of ER after hormone therapy or loss of HER2 after trastu-
28. Is it necessary to order Oncotype Dx on all ER- zumab); others may be caused by intratumoral heterogeneity,
positive early stage invasive carcinoma of the breast? different antibody clones, variability in pathologist’s interpreta-
tion, and/or specimen handling and processing.
The Magee equation uses standard histopathologic and A few studies have assessed clinical outcomes and man-
immunohistochemical variables and could identify a portion agement changes in patients with biomarker status change
of ER-positive breast cancer patients who may not need after neoadjuvant therapy, but the results are controversial
Oncotype DX for RS. The data from our recent study dem- [101, 104]. Some studies have suggested that repeat testing
onstrated cases with a Magee equation RS >30 or ≤11 can be may yield prognostic information [101, 104]; however, oth-
accurately classified in the same categories by the Oncotype ers have observed no survival difference in patients with the
DX test [96]. Therefore, these patients with either modified biomarker changes versus those without changes [104].
Magee equation RS < 11 or >30 are unlikely to benefit from Although the change of biomarkers from negative to positive
Oncotype DX test. Similar analysis has also been proposed may potentially impact clinical management, the observed
by a previous study using a screening algorithm with the difference is usually small (e.g., ER/PR negative-to-positive
combination of Magee equations, histologic criteria and bio- changes usually show low percentage of weak staining).
marker results to identify potential cases unlikely to benefit In summary, there are no guidelines and recommenda-
from the Oncotype DX [98]. Our data show up to 12% of tions for retesting breast biomarkers in residual tumors after
Oncotype DX eligible cases which may not need Oncotype neoadjuvant therapy. Currently, whether to retest the breast
DX. cancer biomarker is institution dependent.
34 X. Li et al.
References 19. Li X, Oprea-Ilies GM, Krishnamurti U. New developments in

breast cancer and their impact on daily practice in pathology. Arch
Pathol Lab Med. 2017;141:490–8.
1. Alva S, Shetty-Alva N. An update of tumor metastasis to the
20. Creighton CJ, Li X, Landis M, Dixon JM, Neumeister VM,
breast data. Arch Surg. 1999;134:450.
Sjolund A, et al. Residual breast cancers after conventional ther-
2. Lee AH. The histological diagnosis of metastases to the breast from
apy display mesenchymal as well as tumor-initiating features.
extramammary malignancies. J Clin Pathol. 2007;60:1333–41.
Proc Natl Acad Sci. 2009;106(33):13820–5.
3. Recine MA, Deavers MT, Middleton LP, Silva EG, Malpica
21. Prat A, Adamo B, Cheang MC, Anders CK, Carey LA,
A. Serous carcinoma of the ovary and peritoneum with metastases
Perou CM. Molecular characterization of basal-like and
to the breast and axillary lymph nodes: a potential pitfall. Am J
non-basal-like triple-negative breast cancer. Oncologist.
Surg Pathol. 2004;28:1646–51.
2013;18:123–33.
4. Davis DG, Siddiqui MT, Oprea-Ilies G, Stevens K, Osunkoya AO,
22. Rakha EA, Ellis IO. Triple-negative/basal-like breast cancer:
Cohen C, et al. GATA-3 and FOXA1 expression is useful to dif-
review. Pathology. 2009;41:40–7.
ferentiate breast carcinoma from other carcinomas. Hum Pathol.
23. Tavassoéli FA, Devilee P, editors. WHO classification of tumours:
2016;47:26–31.
pathology and genetics: tumours of the breast and female genital
5. Kandalaft PL, Simon RA, Isacson C, Gown AM. Comparative
organs. Lyon: IARC; 2003.
sensitivities and specificities of antibodies to breast markers
24. Lakhani SR, Ellis IO, Schnitt SJ, Tan PH, van de Vijver MJ, edi-
GCDFP-15, mammaglobin a, and different clones of antibodies
tors. WHO classification of tumours of the breast. Lyon: IARC;
to GATA-3: a study of 338 tumors using whole sections. Appl
2012.
Immunohistochem Mol Morphol. 2016;24:609–14.
25. Wei B, Ding T, Xing Y, Wei W, Tian Z, Tang F, et al.
6. Liu H, Shi J, Wilkerson ML, Lin F. Immunohistochemical evalu-
Invasive neuroendocrine carcinoma of the breast: a distinc-
ation of GATA3 expression in tumors and normal tissues: a useful
tive subtype of aggressive mammary carcinoma. Cancer.
immunomarker for breast and urothelial carcinomas. Am J Clin
2010;116:4463–73.
Pathol. 2012;138:57–64.
26. Lopez-Bonet E, Alonso-Ruano M, Barraza G, Vazquez-Martin A,
7. Miettinen M, McCue PA, Sarlomo-Rikala M, Rys J, Czapiewski
Bernadó L, Menendez JA. Solid neuroendocrine breast carcino-
P, Wazny K, et al. GATA3: a multispecific but potentially use-
mas: incidence, clinico-pathological features and immunohisto-
ful marker in surgical pathology: a systematic analysis of 2500
chemical profiling. Oncol Rep. 2008;20:1369–74.
epithelial and nonepithelial tumors. Am J Surg Pathol. 2014;38:
27. Lillemoe TJ, Tsai ML, Swenson KK, Susnik B, Krueger J, Harris
13–22.
K, et al. Clinicopathologic analysis of a large series of microinva-
8. Wells JM, Ginter PS, Liu Y, Chen Z, Narula N, Shin SJ. Evaluating
sive breast cancers. Breast J. 2018;24(4):574–9.
the utility of trefoil factor 1 as a mammary-specific immunostain
28. Sue GR, Lannin DR, Killelea B, Chagpar AB. Predictors of micro-
compared and in conjunction with GATA-3 and mammaglobin in
invasion and its prognostic role in ductal carcinoma in situ. Am J
the distinction between carcinoma of breast and lung. Am J Clin
Surg. 2013;206:478–81.
Pathol. 2015;144:444–51.
29. Parikh RR, Haffty BG, Lannin D, Moran MS. Ductal carcinoma
9. Yaziji H, Gown AM. Immunohistochemical analysis of gyneco-
in situ with microinvasion: prognostic implications, long-term
logic tumors. Int J Gynecol Pathol. 2001;20:64–78.
outcomes, and role of axillary evaluation. Int J Radiat Oncol Biol
10. Ni YB, Tsang JY, Shao MM, Chan SK, Tong J, Ka-Fai T, et al.
Phys. 2012;82:7–13.
TTF-1 expression in breast carcinoma: an unusual but real phe-
nomenon. Histopathology. 2014;64:504–11. 30. Shatat L, Gloyeske N, Madan R, O’Neil M, Tawfik O, Fan
11. Luo MH, Huang YH, Ni YB, Tsang JY, Chan SK, Shao F. Microinvasive breast carcinoma carries an excellent prog-
MM, et al. Expression of mammaglobin and gross cystic nosis regardless of the tumor characteristics. Hum Pathol.
disease fluid protein-15 in breast carcinomas. Hum Pathol. 2013;44:2684–9.
2013;44:1241–50. 31. Matsen CB, Hirsch A, Eaton A, Stempel M, Heerdt A, Van Zee
12. Bhargava R, Beriwal S, Dabbs DJ. Mammaglobin vs GCDFP-15: KJ, et al. Extent of microinvasion in ductal carcinoma in situ is not
an immunohistologic validation survey for sensitivity and speci- associated with sentinel lymph node metastases. Ann Surg Oncol.
ficity. Am J Clin Pathol. 2007;127:103–13. 2014;21:3330–5.
13. Matsushima S, Mori M, Adachi Y, Matsukuma A, Sugimachi 32. Gojon H, Fawunmi D, Valachis A. Sentinel lymph node biopsy in
K. S100 protein positive human breast carcinomas: an immuno- patients with microinvasive breast cancer: a systematic review and
histochemical study. J Surg Oncol. 1994;55:108–13. meta-analysis. Eur J Surg Oncol. 2014;40:5–11.
14. Domfeh AB, Carley AL, Striebel JM, Karabakhtsian RG, Florea 33. Hanna MG, Jaffer S, Bleiweiss IJ, Nayak A. Re-evaluating the
AV, McManus K, et al. WT1 immunoreactivity in breast carci- role of sentinel lymph node biopsy in microinvasive breast carci-
noma: selective expression in pure and mixed mucinous subtypes. noma. Mod Pathol. 2014;27:1489–98.
Mod Pathol. 2008;21:1217–23. 34. Lyons JM 3rd, Stempel M, Van Zee KJ, Cody HS 3rd. Axillary
15. Nonaka D, Chiriboga L, Soslow RA. Expression of pax8 as a use- node staging for microinvasive breast cancer: is it justified? Ann
ful marker in distinguishing ovarian carcinomas from mammary Surg Oncol. 2012;19:3416–21.
carcinomas. Am J Surg Pathol. 2008;32:1566–71. 35. Margalit DN, Sreedhara M, Chen YH, Catalano PJ, Nguyen
16. Gloyeske NC, Woodard AH, Elishaev E, Yu J, Clark BZ, Dabbs DJ, PL, Golshan M, et al. Microinvasive breast cancer: ER, PR, and
et al. Immunohistochemical profile of breast cancer with respect HER-2/neu status and clinical outcomes after breast-conserving
to estrogen receptor and HER2 status. Appl Immunohistochem therapy or mastectomy. Ann Surg Oncol. 2013;20:811–8.
Mol Morphol. 2015;23:202–8. 36. Gyorffy B, Hatzis C, Sanft T, Hofstatter E, Aktas B, Pusztai
17. Hou Y, Shen R, Chaudhary S, Tonkovich D, Li Z. Utility of dif- L. Multigene prognostic tests in breast cancer: past, present,
ferent immunostains for diagnosis of metastatic breast carcinomas future. Breast Cancer Res. 2015;17:11.
in both surgical and cytological specimens. Ann Diagn Pathol. 37. Harris LN, Ismaila N, McShane LM, Andre F, Collyar DE,
2017;30:21–7. Gonzalez-Angulo AM, et al. Use of biomarkers to guide decisions
18. Perou CM, Sorlie T, Eisen MB, van de Rijn M, Jeffrey SS, Rees on adjuvant systemic therapy for women with early-stage invasive
CA, et al. Molecular portraits of human breast tumours. Nature. breast cancer: American Society of Clinical Oncology clinical
2000;406:747–52. practice guideline. J Clin Oncol. 2016;34:1134–50.
38. Paik S, Shak S, Tang G, Kim C, Baker J, Cronin M, et al. A mul- cal outcome, even in node-negative patients. Am J Surg Pathol.
tigene assay to predict recurrence of tamoxifen-treated, node- 2017;41:1275–82.
negative breast cancer. N Engl J Med. 2004;351:2817–26. 56. Rabban JT, Glidden D, Kwan ML, Chen YY. Pure and predomi-
39. Paik S, Tang G, Shak S, Kim C, Baker J, Kim W, et al. Gene nantly pure intralymphatic breast carcinoma after neoadjuvant
expression and benefit of chemotherapy in women with node- chemotherapy: an unusual and adverse pattern of residual disease.
negative, estrogen receptor-positive breast cancer. J Clin Oncol. Am J Surg Pathol. 2009;33:256–63.
2006;24:3726–34. 57. Hennessy BT, Gonzalez-Angulo AM, Stemke-Hale K, Gilcrease
40. Dowsett M, Cuzick J, Wale C, Forbes J, Mallon EA, Salter J, MZ, Krishnamurthy S, Lee JS, et al. Characterization of a natu-
et al. Prediction of risk of distant recurrence using the 21-gene rally occurring breast cancer subset enriched in epithelial-to-
recurrence score in node-negative and node-positive postmeno- mesenchymal transition and stem cell characteristics. Cancer Res.
pausal patients with breast cancer treated with anastrozole or 2009;69:4116–24.
tamoxifen: a TransATAC study. J Clin Oncol. 2010;28:1829–34. 58. Weigelt B, Ng CK, Shen R, Popova T, Schizas M, Natrajan
41. Klein ME, Dabbs DJ, Shuai Y, Brufsky AM, Jankowitz R, Puhalla R, et al. Metaplastic breast carcinomas display genomic
SL, et al. Prediction of the Oncotype DX recurrence score: use of and transcriptomic heterogeneity [corrected]. Mod Pathol.
pathology-generated equations derived by linear regression analy- 2015;28:340–51.
sis. Mod Pathol. 2013;26:658–64. 59. Li X, Lewis MT, Huang J, Gutierrez C, Osborne CK, Wu MF,
42. Goldhirsch A, Wood WC, Coates AS, Gelber RD, Thürlimann B, et al. Intrinsic resistance of tumorigenic breast cancer cells to che-
Senn HJ, et al. Strategies for subtypes – dealing with the diversity motherapy. J Natl Cancer Inst. 2008;100:672–9.
of breast cancer: highlights of the St. Gallen international expert 60. Hennessy BT, Giordano S, Broglio K, Duan Z, Trent J, Buchholz
consensus on the primary therapy of early breast cancer 2011. Ann TA, et al. Biphasic metaplastic sarcomatoid carcinoma of the
Oncol. 2011;22:1736–47. breast. Ann Oncol. 2006;17:605–13.
43. Polley MY, Leung SC, McShane LM, Gao D, Hugh JC, 61. Lester TR, Hunt KK, Nayeemuddin KM, Bassett RL Jr, Gonzalez-
Mastropasqua MG, et al. An international Ki67 reproducibility Angulo AM, Feig BW, et al. Metaplastic sarcomatoid carcinoma
study. J Natl Cancer Inst. 2013;105:1897–906. of the breast appears more aggressive than other triple receptor-
44. Polley MY, Leung SC, Gao D, Mastropasqua MG, Zabaglo LA, negative breast cancers. Breast Cancer Res Treat. 2012;131:41–8.
Bartlett JM, et al. An international study to increase concordance 62. Lehmann BD, Bauer JA, Chen X, Sanders ME, Chakravarthy AB,
in Ki67 scoring. Mod Pathol. 2015;28:778–86. Shyr Y, et al. Identification of human triple-negative breast cancer
45. Rakha EA, Martin S, Lee AH, Morgan D, Pharoah PD, Hodi Z, subtypes and preclinical models for selection of targeted thera-
et al. The prognostic significance of lymphovascular invasion in pies. J Clin Invest. 2011;121:2750–67.
invasive breast carcinoma. Cancer. 2012;118:3670–80. 63. Lehmann BD, Jovanovic B, Chen X, Estrada MV, Johnson KN,
46. Lee AK, DeLellis RA, Silverman ML, Heatley GJ, Wolfe Shyr Y, et al. Refinement of triple-negative breast cancer molecu-
HJ. Prognostic significance of peritumoral lymphatic and blood lar subtypes: implications for neoadjuvant chemotherapy selec-
vessel invasion in node-negative carcinoma of the breast. J Clin tion. PLoS One. 2016;11(6):e0157368.
Oncol. 1990;8:1457–65. 64. Masuda H, Baggerly KA, Wang Y, Zhang Y, Gonzalez-Angulo
47. Song YJ, Shin SH, Cho JS, Park MH, Yoon JH, Jegal YJ. The role AM, Meric-Bernstam F, et al. Differential response to neoadju-
of lymphovascular invasion as a prognostic factor in patients with vant chemotherapy among 7 triple-negative breast cancer molecu-
lymph node-positive operable invasive breast cancer. J Breast lar subtypes. Clin Cancer Res. 2013;19:5533–40.
Cancer. 2011;14:198–203. 65. Demaria S, Pikarsky E, Karin M, Coussens LM, Chen YC,
48. Tan BY, Thike AA, Ellis IO, Tan PH. Clinicopathologic character- El-Omar EM, et al. Cancer and inflammation: promise for bio-
istics of solid papillary carcinoma of the breast. Am J Surg Pathol. logic therapy. J Immunother. 2010;33:335–51.
2016;40:1334–42. 66. Loi S, Sirtaine N, Piette F, Salgado R, Viale G, Van Eenoo F, et al.
49. Nassar H, Qureshi H, Adsay NV, Visscher D. Clinicopathologic Prognostic and predictive value of tumor-infiltrating lymphocytes
analysis of solid papillary carcinoma of the breast and associated in a phase III randomized adjuvant breast cancer trial in node-
invasive carcinomas. Am J Surg Pathol. 2006;30:501–7. positive breast cancer comparing the addition of docetaxel to
50. Nicolas MM, Wu Y, Middleton LP, Gilcrease MZ. Loss of myo- doxorubicin with doxorubicin-based chemotherapy: BIG 02-98. J
epithelium is variable in solid papillary carcinoma of the breast. Clin Oncol. 2013;31:860–7.
Histopathology. 2007;51:657–65. 67. Adams S, Gray RJ, Demaria S, Goldstein L, Perez EA, Shulman
51. Symmans WF, Peintinger F, Hatzis C, Rajan R, Kuerer H, Valero LN, et al. Prognostic value of tumor-infiltrating lymphocytes in
V, et al. Measurement of residual breast cancer burden to pre- triple-negative breast cancers from two phase III randomized
dict survival after neoadjuvant chemotherapy. J Clin Oncol. adjuvant breast cancer trials: ECOG 2197 and ECOG 1199. J Clin
2007;25:4414–22. Oncol. 2014;32:2959–66.
52. Symmans WF, Wei C, Gould R, Yu X, Zhang Y, Liu M, et al. 68. Krishnamurti U, Wetherilt CS, Yang J, Peng L, Li X. Tumor-
Long-term prognostic risk after neoadjuvant chemotherapy asso- infiltrating lymphocytes are significantly associated with bet-
ciated with residual cancer burden and breast cancer subtype. J ter overall survival and disease-free survival in triple-negative
Clin Oncol. 2017;35:1049–60. but not estrogen receptor-positive breast cancers. Hum Pathol.
53. Bossuyt V, Provenzano E, Symmans WF, Boughey JC, Coles C, 2017;64:7–12.
Curigliano G, et al. Recommendations for standardized pathologi- 69. Li XB, Krishnamurti U, Bhattarai S, Klimov S, Reid MD,
cal characterization of residual disease for neoadjuvant clinical O’Regan R, et al. Biomarkers predicting pathologic complete
trials of breast cancer by the BIG-NABCG collaboration. Ann response to neoadjuvant chemotherapy in breast cancer. Am J Clin
Oncol. 2015;26:1280–91. Pathol. 2016;145:871–8.
54. Amin MB, Edge S, Greene F, Byrd DR, Brookland RK, 70. Ma Y, Kepp O, Ghiringhelli F, Apetoh L, Aymeric L, Locher C,
Washington MK, et al. American joint committee on cancer et al. Chemotherapy and radiotherapy: cryptic anticancer vac-
(AJCC) cancer staging manual. 8th ed. New York: Springer; 2017. cines. Semin Immunol. 2010;22:113–24.
55. Cheng E, Ko D, Nguyen M, Moo TA, Andreopoulou E, Hoda 71. Demaria S, Volm MD, Shapiro RL, Yee HT, Oratz R, Formenti
SA, et al. Residual pure intralymphatic breast carcinoma SC, et al. Development of tumor-infiltrating lymphocytes in breast
following neoadjuvant chemotherapy is indicative of poor clini-
36 X. Li et al.
cancer after neoadjuvant paclitaxel chemotherapy. Clin Cancer 88. Hirschmann A, Lamb TA, Marchal G, Padilla M, Diebold
Res. 2001;7:3025–30. J. Simultaneous analysis of HER2 gene and protein on a single
72. Salgado R, Denkert C, Demaria S, Sirtaine N, Klauschen F, slide facilitates HER2 testing of breast and gastric carcinomas.
Pruneri G, et al. The evaluation of tumor-infiltrating lymphocytes Am J Clin Pathol. 2012;138:837–44.
(TILs) in breast cancer: recommendations by an international 89. Li Z, Dabbs DJ, Cooper KL, Bhargava R. Dual HER2 gene pro-
TILs working group 2014. Ann Oncol. 2015;26:259–71. tein assay: focused study of breast cancers with 2+ immunohisto-
73. Mao Y, Qu Q, Zhang Y, Liu J, Chen X, Shen K. The value of tumor chemical expression. Am J Clin Pathol. 2015;143:451–8.
infiltrating lymphocytes (TILs) for predicting response to neoad- 90. Kurozumi S, Padilla M, Kurosumi M, Matsumoto H, Inoue K,
juvant chemotherapy in breast cancer: a systematic review and Horiguchi J, et al. HER2 intratumoral heterogeneity analyses by
meta-analysis. PLoS One. 2014;9:e115103. concurrent HER2 gene and protein assessment for the prognosis
74. Denkert C, Loibl S, Noske A, Roller M, Müller BM, Komor M, of HER2 negative invasive breast cancer patients. Breast Cancer
et al. Tumor-associated lymphocytes as an independent predictor Res Treat. 2016;158:99–111.
of response to neoadjuvant chemotherapy in breast cancer. J Clin 91. Bartlett AI, Starcyznski J, Robson T, Maclellan A, Campbell FM,
Oncol. 2010;28:105–13. van de Velde CJ, et al. Heterogeneous HER2 gene amplification:
75. Gooch J, King TA, Eaton A, Dengel L, Stempel M, Corben AD, impact on patient outcome and a clinically relevant definition. Am
et al. The extent of extracapsular extension may influence the need J Clin Pathol. 2011;136:266–74.
for axillary lymph node dissection in patients with T1–T2 breast 92. Lee HJ, Seo AN, Kim EJ, Jang MH, Suh KJ, Ryu HS, et al. HER2
cancer. Ann Surg Oncol. 2014;21:2897–903. heterogeneity affects trastuzumab responses and survival in
76. Choi AH, Blount S, Perez MN, Chavez de Paz CE, Rodriguez SA, patients with HER2-positive metastatic breast cancer. Am J Clin
Surrusco M, et al. Size of extranodal extension on sentinel lymph Pathol. 2014;142:755–66.
node dissection in the American College of Surgeons Oncology 93. Lee HJ, Kim JY, Park SY, Park IA, Song IH, Yu JH, et al.
Group Z0011 trial era. JAMA Surg. 2015;150:1141–8. Clinicopathologic significance of the intratumoral heterogene-
77. Giuliano AE, Hunt KK, Ballman KV, Beitsch PD, Whitworth PW, ity of HER2 gene amplification in HER2-positive breast cancer
Blumencranz PW, et al. Axillary dissection vs no axillary dis- patients treated with adjuvant Trastuzumab. Am J Clin Pathol.
section in women with invasive breast cancer and sentinel node 2015;144:570–8.
metastasis: a randomized clinical trial. JAMA. 2011;305:569–75. 94. Hou Y, Nitta H, Wei L, Banks PM, Portier B, Parwani AV, et al.
78. Seol H, Lee HJ, Choi Y, Lee HE, Kim YJ, Kim JH, et al. HER2 intratumoral heterogeneity is independently associated
Intratumoral heterogeneity of HER2 gene amplification in with incomplete response to anti-HER2 neoadjuvant chemother-
breast cancer: its clinicopathological significance. Mod Pathol. apy in HER2-positive breast carcinoma. Breast Cancer Res Treat.
2012;25:938–48. 2017;166:447–57.
79. Brunelli M, Manfrin E, Martignoni G, Miller K, Remo A, 95. Flanagan MB, Dabbs DJ, Brufsky AM, Beriwal S, Bhargava
Reghellin D, et al. Genotypic intratumoral heterogeneity in breast R. Histopathologic variables predict Oncotype DX recurrence
carcinoma with HER2/neu amplification: evaluation according to score. Mod Pathol. 2008;21:1255–61.
ASCO/CAP criteria. Am J Clin Pathol. 2009;131:678–82. 96. Hou Y, Tozbikian G, Zynger DL, Li Z. Using the modified Magee
80. Striebel JM, Bhargava R, Horbinski C, Surti U, Dabbs DJ. The equation to identify patients unlikely to benefit from the 21-gene
equivocally amplified HER2 FISH result on breast core biopsy: recurrence score assay (Oncotype DX assay). Am J Clin Pathol.
indications for further sampling do affect patient management. 2017;147:541–8.
Am J Clin Pathol. 2008;129:383–90. 97. Hou Y, Zynger DL, Li X, Li Z. Comparison of Oncotype DX with
81. Lewis JT, Ketterling RP, Halling KC, Reynolds C, Jenkins RB, modified Magee equation recurrence scores in low-grade invasive
Visscher DW. Analysis of intratumoral heterogeneity and ampli- carcinoma of breast. Am J Clin Pathol. 2017;148:167–72.
fication status in breast carcinomas with equivocal (2+) HER-2 98. Turner BM, Skinner KA, Tang P, Jackson MC, Soukiazian N,
immunostaining. Am J Clin Pathol. 2005;124:273–81. Shayne M, et al. Use of modified Magee equations and histologic
82. Hanna WM, Ruschoff J, Bilous M, Coudry RA, Dowsett M, criteria to predict the Oncotype DX recurrence score. Mod Pathol.
Osamura RY, et al. HER2 in situ hybridization in breast cancer: 2015;28:921–31.
clinical implications of polysomy 17 and genetic heterogeneity. 99. Cockburn A, Yan J, Rahardja D, Euhus D, Peng Y, Fang Y, et al.
Mod Pathol. 2014;27:4–18. Modulatory effect of neoadjuvant chemotherapy on biomarkers
83. Ohlschlegel C, Zahel K, Kradolfer D, Hell M, Jochum W. HER2 expression; assessment by digital image analysis and relationship
genetic heterogeneity in breast carcinoma. J Clin Pathol. to residual cancer burden in patients with invasive breast cancer.
2011;64:1112–6. Hum Pathol. 2014;45:249–58.
84. Tubbs RR, Hicks DG, Cook J, Downs-Kelly E, Pettay J, Hartke 100. Kasami M, Uematsu T, Honda M, Yabuzaki T, Sanuki J, Uchida Y,
MB, et al. Fluorescence in situ hybridization (FISH) as primary et al. Comparison of estrogen receptor, progesterone receptor and
methodology for the assessment of HER2 status in adenocarci- her-2 status in breast cancer pre- and post-neoadjuvant chemo-
noma of the breast: a single institution experience. Diagn Mol therapy. Breast. 2008;17:523–7.
Pathol. 2007;16:207–10. 101. Tacca O, Penault-Llorca F, Abrial C, Mouret-Reynier MA,
85. Allison KH, Dintzis SM, Schmidt RA. Frequency of HER2 Raoelfils I, Durando X, et al. Changes in and prognostic value
heterogeneity by fluorescence in situ hybridization according to of hormone receptor status in a series of operable breast cancer
CAP expert panel recommendations: time for a new look at how patients treated with neoadjuvant chemotherapy. Oncologist.
to report heterogeneity. Am J Clin Pathol. 2011;136:864–71. 2007;12:636–43.
86. Chang MC, Malowany JI, Mazurkiewicz J, Wood M. ‘Genetic 102. Chen S, Chen CM, Yu KD, Zhou RJ, Shao ZM. Prognostic
heterogeneity’ in HER2/neu testing by fluorescence in situ value of a positive-to-negative change in hormone receptor
hybridization: a study of 2,522 cases. Mod Pathol. 2012;25:683–8. status after neoadjuvant chemotherapy in patients with
87. Murthy SS, Sandhya DG, Ahmed F, Goud KI, Dayal M, Suseela hormone receptor- positive breast cancer. Ann Surg Oncol.
K, et al. Assessment of HER2/Neu status by fluorescence in situ 2012;19:3002–11.
hybridization in immunohistochemistry-equivocal cases of inva- 103. Hirata T, Shimizu C, Yonemori K, Hirakawa A, Kouno T, Tamura
sive ductal carcinoma and aberrant signal patterns: a study at a K, et al. Change in the hormone receptor status following
tertiary cancer center. Indian J Pathol Microbiol. 2011;54:532–8. administration of neoadjuvant chemotherapy and its impact on the
long-term outcome in patients with primary breast cancer. Br J 105. Lee HC, Ko H, Seol H, Noh DY, Han W, Kim TY, et al. Expression
Cancer. 2009;101:1529–36. of immunohistochemical markers before and after neoadjuvant
104. Parinyanitikul N, Lei X, Chavez-MacGregor M, Liu S, Mittendorf chemotherapy in breast carcinoma, and their use as predictors of
EA, Litton JK, et al. Receptor status change from primary to resid- response. J Breast Cancer. 2013;16:395–403.
ual breast cancer after neoadjuvant chemotherapy and analysis of
survival outcomes. Clin Breast Cancer. 2015;15:153–60.
Invasive Carcinoma of the Breast:
Special Types 3
Zaibo Li, Xiaoyan Cui, Xiaoxian Li, and Yan Peng
List of Frequently Asked Questions 2. What is the most common tumor profile status of inva-
sive cribriform carcinoma?
1. What is invasive cribriform carcinoma and what are
its key diagnostic features? As a well-differentiated carcinoma, invasive cribriform car-
cinoma is usually estrogen receptor (ER) positive, progester-
Invasive cribriform carcinoma is characterized histologically one receptor (PR) positive, and human epidermal growth
by the cribriform growth pattern of the invasive carcinoma factor receptor 2 (HER2) negative. See Fig. 3.1a–g.
[1, 2]. The glands are morphologically similar to those of
cribriform-type ductal carcinoma in situ (DCIS), manifested 3. Does invasive cribriform carcinoma have a better
as fenestrated, rounded, or angulate infiltrating glands. The prognosis compared to other types of breast
tumor cells are usually low to intermediate nuclear grade. cancer?
Mucinous secretion is sometimes present within the lumens,
as well as microcalcifications. The surrounding stroma is Compared to invasive ductal carcinoma of no special type,
often fibroblastic, sometimes associated with osteoclast-like invasive cribriform carcinoma has a better and favorable
giant cells [3]. Pure invasive cribriform carcinoma has >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 (<50%) are also included in this category. nostic features?
If the minor component is of another morphological type
other than tubular pattern, they are regarded as being “mixed Tubular carcinoma is characterized by haphazardly arranged
type” [1, 2]. small tubules that closely resemble normal ductules. The
tubules are angulated, oval or round in shape, lined by a sin-
gle layer of epithelial cells with low-grade nuclear atypia and
enclosed in an open lumen. There is no consensus for
required proportion (75–100%) of tubule formation for the
diagnosis of tubular carcinoma. But practically, ≥90% is
Z. Li · X. Cui
Department of Pathology, The Ohio State University Wexner needed to render a diagnosis of pure tubular carcinoma.
Medical Center, Columbus, OH, USA When the tubular component involves <90% of the tumor, it
e-mail: [email protected]; [email protected] is referred to as a “mixed” tubular carcinoma or invasive duc-
X. Li tal carcinoma of no special type with tubular features.
Department of Pathology and Laboratory Medicine, Complex architecture, marked nuclear pleomorphism, and
Emory University, Atlanta, GA, USA high mitotic activity are contradictions for the diagnosis of
tubular carcinoma [4].
Y. Peng (*) Most tubular carcinomas are 2 cm or less in size. Under
Department of Pathology, University of Texas Southwestern
Medical Center, Dallas, TX, USA low-power examination, the stroma admixed with tubular
e-mail: [email protected] carcinoma is usually desmoplastic or fibroelastotic, different

40 Z. Li et al.
Fig. 3.1 Invasive cribriform

carcinoma. Carcinoma cells a
grow in cribriform pattern
with microcalcification (a).
Carcinoma cells show
low-grade nuclei at high
magnification (b). p40 (c),
SMMS (d), and CK5 (e)
immunostains show the
absence of myoepithelial cell
layers around carcinoma cells.
Carcinoma cells are strongly
and diffusely positive for ER
(f) and PR (g)
d
3 Invasive Carcinoma of the Breast: Special Types 41
e
from the surrounding benign breast stroma, providing a use- ated CCC, LCIS (classic type), and invasive tubular
ful clue for the diagnosis. carcinoma have been referred to as the “Rosen triad” [5].
Tubular carcinoma is frequently associated with colum-
nar cell lesions, ranging from columnar cell change (CCC), 5. What is the most common tumor profile status of
columnar cell hyperplasia (CCH), to CCC or CCH with tubular carcinoma?
atypia. DCIS and LCIS are also seen. DCIS arising in the
background of CCC often has a low nuclear grade and crib- Tubular carcinoma is usually ER positive, PR positive, and
riform or micropapillary architecture. The commonly associ- HER2 negative. See Fig. 3.2a–d.
42 Z. Li et al.
6. Does tubular carcinoma have the best prognosis in the outer surface of the micropapillae, indicating the
among all types of breast cancer? reversed polarity of the epithelium, similar to that in invasive
micropapillary carcinoma of the breast. The variant seems to
Tubular carcinoma has an excellent prognosis. Most studies have a more aggressive behavior than conventional pure
suggest that patients with tubular carcinoma have a longer mucinous carcinoma and has a higher frequency of lymph
disease-free survival than patients with invasive ductal carci- node metastasis.
noma of no specific type. In some studies, it is comparable to Radiographically, mucinous carcinoma usually mimics
that of age-matched set of women without breast cancer or fibroadenoma, a common benign breast tumor.
the general population [4, 6].
8. What is the most common tumor profile status of
7. What is mucinous carcinoma and what are its key mucinous carcinoma?
diagnostic features?
The majority of pure mucinous carcinomas are ER positive,
Mucinous carcinoma is characterized by clusters of tumor PR positive, and HER2 negative. Mucinous carcinomas pre-
cells floating in a pool of extracellular mucin. The relative dominantly express MUC2 and MUC6 [11]. See Fig. 3.3a, b.
proportion of mucin and tumor cells is variable. The diagno-
sis of pure mucinous carcinoma is reserved for at least 90% 9. Does mucinous carcinoma have a better prognosis
of the tumor showing mucinous component. Those in which compared to other types of breast cancer?
the mucinous component comprising 50–90% of the lesions
are regarded as “mixed” mucinous carcinoma. Invasive duc- Invasive mucinous carcinoma has a favorable prognosis
tal carcinomas with less than 50% of the mucinous compo- compared to breast cancer of no special type. The accumula-
nent are best referred to as having focal mucinous tion of extracellular mucin serves as a barrier to the spread of
differentiation. tumor cells. Major prognostic factors are similar to most
Pure mucinous carcinoma is uncommon and accounts for types of breast carcinoma. Nodal status is the most signifi-
about 2% of invasive breast carcinomas [4]. The mean age of cant prognostic factor; others include age at diagnosis, tumor
patients with invasive mucinous carcinoma is in general size, status of PR expression, and nuclear grade.
older (mean age is 71 years) than those with breast cancer of
no special type [7]. 10. What is invasive micropapillary carcinoma and what
Pure mucinous carcinoma grossly has a characteristic are its key diagnostic features?
gelatinous and glistening appearance on the cut surface due
to the presence of abundant extracellular mucin. Invasive micropapillary carcinoma is characterized by
Microscopically, the tumor cells form small clusters, large tumor cells forming micropapillae and tubuloalveolar or
sheets, or with papillary or cribriform configurations float- morule-like clusters without fibrovascular cores, sur-
ing in the pool of mucin. The tumor cells are usually low to rounded by clear stromal space. There is no universal crite-
intermediate nuclear grade. High nuclear grade is rare and rion to distinguish mixed and pure invasive micropapillary
should be emphasized in the diagnostic pathology report carcinoma. In practice, pure invasive micropapillary carci-
because the clinical course may be worse than usual pure noma refers to those with at least 75% of the tumor show-
mucinous carcinoma. The periphery of most tumors is char- ing micropapillary configuration. The cell clusters display
acterized by a pushing border due to their slow growth. reversed polarity with the luminal aspect of the cells pres-
When assessing the margin status, the presence of extracel- ent on the outer surface of the clusters close to the stroma.
lular mucin without tumor cells at the margin is considered This can be well demonstrated by epithelial membrane
positive. antigen (EMA) immunostaining the cell membrane facing
Based on the morphology, mucinous carcinoma has been toward the stroma. MUC1, like EMA, also stains the simi-
subclassified as type A and type B [8]. Overall, mucin is lar pattern. The nuclear grade of invasive micropapillary
more abundant in type A than in type B tumors, which show carcinoma is usually intermediate to high. The clear spaces
hypercellularity. Type B tumors also show frequent neuroen- around the tumor cells mimic lymphovascular invasion, but
docrine differentiation. Currently, no clinical implications they are not lined by endothelial cells. They are usually
have been noted in separating these subtypes, and the sub- attributed to artifacts during tissue processing. However,
typing is barely mentioned in routine diagnosis. invasive micropapillary carcinomas do have a higher fre-
A micropapillary variant of pure mucinous carcinoma has quency of lymphovascular invasion [12], and the tumor
been reported [9, 10]. The tumor cells form micropapillae. emboli in the lymphovascular spaces show similar micro-
Epithelial membrane antigen (EMA) immunostain is positive papillary morphology.
Fig. 3.2 Invasive tubular

carcinoma. (a) The majority a
of carcinoma cells grow in
tubules, which are angulated,
irregular, and infiltrating into
the surrounding stroma. (b)
magnification. (c) Positive for
ER (strong and diffuse). (d)
Positive for PR (variable)
d
44 Z. Li et al.
Fig. 3.3 Invasive mucinous

carcinoma. (a) Small clusters a
of carcinoma cells are floating
in a mucinous pool. (b)
magnification
11. What is the most common tumor profile status of 13. What is invasive apocrine carcinoma and what are its
invasive micropapillary carcinoma? key diagnostic features?
Most invasive micropapillary carcinomas are positive for ER Invasive apocrine carcinoma is composed of tumor cells with
and PR. HER2 protein is variably overexpressed in a fraction apocrine differentiation of tumor cells, characterized with
of tumors. See Fig. 3.4a, b. abundant densely eosinophilic, granular, or vacuolated cyto-
plasm, large nuclei, and often prominent nucleoli. Compared
12. D
oes invasive micropapillary carcinoma have an to benign apocrine cells, apocrine tumor cells demonstrate
increased risk for nodal metastasis and a worse prog- an increase in nuclear size, significant nuclear pleomor-
nosis compared to other types of breast cancer? phism, irregular nuclear membrane, hyperchromatic nuclei,
and one or more macronucleoli. Features of cytoplasm are
When compared with invasive ductal carcinoma of no special similar to those in the benign apocrine cells. Pure apocrine
type, invasive micropapillary carcinoma has a higher fre- carcinoma is reserved for a tumor consisting of almost all
quency of lymphovascular invasion and lymph node metasta- malignant apocrine cells. If only a portion (>10%) of the
sis, and more lymph nodes are involved [13]. Patients usually tumor consists of malignant apocrine cells, then it can be
have a significantly shorter disease-free survival (DFS) and considered as invasive carcinoma with apocrine differentia-
overall survival (OS) [14]. But when stratified for the number tion. Apocrine differentiation can be seen in up to 30% of all
of involved lymph nodes and other prognostic factors, patients breast cancers [16].
seem to have similar survival rates to those with non-micro- It has been reported that some benign cystic and papillary
papillary invasive ductal carcinoma [15]. Unlike other special- apocrine lesions show little or no detectable surrounding myo-
type breast carcinomas, the poor prognosis associated with epithelial cells [17]. Without cytological atypia, the absence of
this entity appears to be the same whether the micropapillary immunoreactive myoepithelial layer is not an absolute crite-
component is present focally or diffusely within a tumor [15]. rion for the diagnosis of invasive apocrine carcinoma.
Fig. 3.4 Invasive
micropapillary carcinoma. (a) a
Carcinoma cells grow in a
micropapillary pattern
without fibrovascular cores.
There are empty spaces
around the clusters of
carcinoma cells. (b)
variable grade nuclei at high
magnification
14. Is there any difference on the tumor profile in apo- similar pathologic parameters as those of non-apocrine
crine carcinoma compared to other types of breast breast carcinomas.
cancer?
16. What is mammary carcinoma with osteoclast-like
Most invasive apocrine carcinomas are negative for ER giant cells and what are its key diagnostic features?
and PR, but positive for androgen receptor (AR). Some
studies have regarded only tumors with apocrine mor- Carcinomas with osteoclast-like giant cells are characterized
phology and ER-negative, PR-negative, and AR-positive by the presence of multinucleated osteoclast-like giant cells
immunoprofile as pure apocrine carcinoma. About half of in the stroma. These cells are non-neoplastic, while the carci-
pure apocrine carcinomas are HER2 negative and the nomatous components can be a variety of histological types.
remaining HER2 positive [18]. Immunohistochemical Frequently, the carcinomatous components are invasive duc-
studies may be used to confirm the diagnostic impression tal carcinoma of no special type with a cribriform growth pat-
of apocrine differentiation but are not essential to estab- tern, but other histological types such as lobular, squamous,
lish the morphological diagnosis of apocrine carcinoma. papillary, mucinous, and metaplastic carcinoma have also
See Fig. 3.5a, b. been reported [19–22]. Grossly, the tumors display red-brown
GCDFP-15 (BRST-2) immunostain is positive in a high to dark-brown color, which is due to the presence of hemor-
percentage of invasive apocrine carcinoma [16]. rhage and hemosiderin-laden macrophages in the tumors.
Microscopically, the giant cells are variable in size as well as
15. D
oes invasive apocrine carcinoma carry a worse the number of nuclei. They are cytologically bland with no
prognosis than other types of breast cancer? mitotic activity. Hemorrhage in the stroma is commonly seen
in the tumor, which may be a clue for the presence of osteo-
The prognosis of invasive apocrine carcinoma is related to clast-giant cells under low-power examination. These giant
tumor grade, size, lymph node status, and tumor stage, cells may also be present in metastatic and recurrent tumors.
46 Z. Li et al.
Fig. 3.5 Invasive apocrine

carcinoma. (a) Carcinoma a
cells grow in solid nests with
minimal intervening stroma.
(b) Carcinoma cells show
intermediate to high nuclear
pleomorphism with abundant
eosinophilic cytoplasm at
high magnification
a b
Fig. 3.6 Invasive mammary carcinoma with osteoclast-like giant cells. (a) Carcinoma cells grow in solid nests intermixed with osteoclast-like
giant cells. (b) Osteoclast-like giant cells are large with abundant cytoplasm, multiple nuclei, and prominent nucleoli
17. What is the tumor profile status of mammary carci- nonspecific esterase, and lysozyme, and are negative for
noma with osteoclast-like giant cells? S100, actin, and keratin [23, 24]. See Fig. 3.6a, b. However,
the mechanism by which they are formed is still unknown.
The osteoclast-like giant cells in the carcinoma are of histio- The tumor profile status of the carcinoma depends on the
cytic lineage, which express CD68, acid phosphatase, histological type of its carcinomatous component.
18. W
hat is the prognosis of mammary carcinoma with degree of nuclear pleomorphism, prominent nucleoli, and a
osteoclast-like giant cells? brisk mitotic activity. Breast fibroglandular tissue should not
be present within the invasive carcinoma [26]. In the most
Lymph node metastases are seen in about one third of the recent 2012 edition of the WHO classification, the term of
cases, and the 5-year survival rate is around 70% [24]. The this entity was revised to invasive ductal carcinoma with
presence of osteoclast-like giant cells does not carry any medullary features (see Fig. 3.7a, b), which also includes
specific prognostic implications. Prognosis is related to the “atypical medullary carcinoma” referring to tumors that do
histologic and immunophenotypic features of the associated not fulfill all the diagnostic criteria [27].
carcinoma.
20. W
hat is the most common tumor profile and genomic
19. What is invasive ductal carcinoma with medullary abnormality of invasive ductal carcinoma with med-
features and what are its key diagnostic features? ullary features?
In the 2003 World Health Organization (WHO) Classification Invasive ductal carcinomas with medullary features are often
of Tumors of the Breast, medullary carcinoma was defined as triple-negative breast cancers with a basal-like phenotype
a “well circumscribed carcinoma composed of poorly dif- expressing CK5/6, CK14, and EGFR [28–31]. These tumors
ferentiated cells with scant stroma and prominent lymphoid are often associated with BRCA1 mutations (in up to 60% of
infiltration” [25]. The classical morphologic features include tumors), whereas less frequently associated with BRCA2
a well-circumscribed smooth rounded pushing border, a syn- mutations [27]. About 11% of patients showed
cytial growth pattern greater than 75% of the tumor (broad BRCA1germline mutations. In addition, invasive ductal car-
anastomosing sheets of tumor cells with indistinct cell bor- cinoma with medullary features shows more frequent
ders), diffuse lymphoplasmacytic infiltrates within the tumor, genomic instabilities, aneuploid or polyploid, and p53 muta-
and, at greater than 75% of the tumor periphery, a high tions than invasive ductal carcinoma NOS [27].
Fig. 3.7 Invasive ductal

carcinoma with medullary a
features. (a) Carcinoma with
well-circumscribed border
and prominent lymphocytic
infiltrates at the periphery.
(b) Syncytial growth pattern
of anaplastic tumor cells
admixed with
lymphoplasmacytic cells
b
48 Z. Li et al.
21. Does invasive ductal carcinoma with medullary fea- mucinous carcinoma or solid papillary carcinoma [46].
tures carry a better prognosis? Neuroendocrine differentiation can also be seen in invasive
lobular carcinoma, especially alveolar variant [43]. See
Invasive ductal carcinoma with medullary features has been Figs. 3.8a–h and 3.9a–f.
considered a distinctive subgroup of triple-negative carcino-
mas with a favorable prognosis despite its high-grade mor- 23. W
hat is the most common immunoprofile of invasive
phology. However, it is necessary to adhere to strict carcinoma with neuroendocrine features?
morphologic criteria for the diagnosis of this tumor in order
to predict its better prognosis [31–36]. The diagnosis of neuroendocrine tumor usually requires dem-
Recently, it has been reported that breast invasive carcino- onstrating the expression of neuroendocrine markers.
mas with prominent lymphocytic infiltrates (also called tumor Synaptophysin and chromogranin A are the most commonly
infiltrating lymphocytes, or TILs) have better prognosis and used neuroendocrine markers, with synaptophysin as the
response to neoadjuvant chemotherapy, especially in high- most sensitive and chromogranin A as the most specific
grade HER2-positive and triple-negative breast carcinoma immunohistochemical marker. Other neuroendocrine mark-
[37–39]. The relatively good outcome seen in patients with ers such as neuron-specific enolase (NSE) and CD56 may
this tumor may result from prominent lymphocytic infiltrates also be used, with less sensitivity and specificity.
rather than an inherently better prognosis. Therefore, most Neuroendocrine markers are usually diffusely positive in
breast pathologists prefer to diagnose invasive ductal carci- WD-NET and PD-NEC/SCC, while patchy and focal in IBC-
noma with medullary features as a basal-like triple-negative NED. There is only limited information available regarding
carcinoma with prominent lymphocytic infiltrates. the expression of biomarkers (tumor profile) in invasive carci-
nomas with neuroendocrine features. Available data suggest
22. What is invasive carcinoma with neuroendocrine fea- that these tumors are most commonly ER positive, PR posi-
tures and what are its key diagnostic features? tive, and HER2 negative [27]. ER and PR are positive in the
majority of WD-NETs and in greater than 50% of PD-NECs,
In the 2003 WHO classification, neuroendocrine carcinomas but variable in IBC-NEDs [43–47]. Similar to SCCs of other
of the breast were divided into solid neuroendocrine carci- sites, primary SCCs of the breast can show expression of thy-
noma, small cell/oat cell carcinoma, and large cell neuroen- roid transcription factor −1 (TTF-1) [47].
docrine carcinoma [25]. In the 2012 WHO classification, the
term of the tumor was revised to carcinomas with neuroen- 24. Do the neuroendocrine features of invasive carci-
docrine features, which was defined as carcinomas with neu- noma play a role in prognosis and treatment
roendocrine differentiation exhibiting morphology similar to decision?
that of neuroendocrine tumors of the lung and gastrointesti-
nal tract. No definitive threshold for neuroendocrine marker No specific guidelines exist for grading breast carcinomas
positivity was required [27]. with neuroendocrine features, and the 2012 WHO classifica-
Histologically, these tumors can be classified into three tion states that grading is unlikely to be clinically significant
categories: well-differentiated neuroendocrine tumor [27]. Currently carcinomas with neuroendocrine features of
(WD-NET), poorly differentiated neuroendocrine carci- the breast are staged, histologically graded, and treated simi-
noma/small cell carcinoma (PD-NEC/SCC), and invasive larly to invasive carcinomas of no special type. The use of
breast carcinoma with neuroendocrine differentiation (IBC- endocrine therapy or HER2 targeted therapy depends on the
NED). Morphologically, WD-NET consists of cellular solid status of the tumor’s ER, PR, and HER2 expressions [27].
expansile nests and trabeculae separated by delicate fibro- No consensus has been reached on the prognosis for this
vascular stroma, similar to NET from other sites. The tumor group of tumors. Although many studies demonstrate a poor
cells are usually spindled, plasmacytoid, or polygonal with prognosis for breast carcinomas with neuroendocrine fea-
abundant granular or clear vacuolated cytoplasm [40–42]. tures, the results are conflicting, likely due to varying inclu-
The nuclear features include classic smooth nuclear borders sion criteria [44, 47–50].
and salt-and-pepper chromatin seen in carcinoids of other
sites. PD-NEC/SCC is morphologically identical to its coun- 25. What is secretory carcinoma of the breast and what
terpart in other sites, consisting of densely packed hyper- are its key diagnostic features?
chromatic cells with scant cytoplasm and crushing artifact.
Mitotic activity and necrosis are common [43–47]. IBC- Secretory carcinoma is a rare, special type of invasive car-
NED can show variable morphology with only subtle cyto- cinoma with a solid, microcystic, and tubular architecture
logic/nuclear features of neuroendocrine differentiation. and large amounts of extracellular and intracellular secre-
Neuroendocrine differentiation has been demonstrated in up tions. Historically, secretory carcinoma was known as
to 30% of invasive ductal carcinomas, most commonly in “juvenile breast carcinoma” as it was originally identified
Fig. 3.8 Invasive ductal a

carcinoma with
neuroendocrine features.
The tumor is composed of
epithelial cells in trabecular
growth pattern
(a). Neuroendocrine nuclear
features are appreciated at
high magnification (b). The
tumor cells are positive for
synaptophysin (c),
chromogranin A (d), CK7 (e),
and ER (f). The tumor cells
are negative for PR (g) and
HER2 (0-1+) (h)
d
50 Z. Li et al.
e
h
Fig. 3.9 Small cell

carcinoma of the breast. The a
tumor is composed of nests of
malignant epithelial cells with
high nuclear to cytoplasmic
ratio and hyperchromatic
nuclei (a). Neuroendocrine
nuclear features and nuclear
molding (b). The tumor cells
are positive for chromogranin
A (c), GATA3 (very focal)
(d), CK7 (patchy) (e), and
TTF-1 (diffuse) (f)
d
52 Z. Li et al.
e
in young patients. However, it has been reported in patients can be seen together with invasive secretory carcinoma
in a wide range of age (3–87 years) and a median age of 25 [51–53]. See Fig. 3.10a, b.
[27, 51].
Secretory carcinomas are composed of well-circumscribed 26. What is the most common tumor profile status of
nodules with tumor cells growing in three patterns: solid, secretory carcinoma?
microcystic, and tubular patterns. The microcystic pattern
shows multiple small cysts resembling thyroid follicles. The Secretory carcinoma is significantly more common in
tubular pattern shows tubules with lumen containing secre- females and usually presents as a mobile, palpable lesion in
tions. Most tumors contain all three patterns with various the subareolar region. Radiological breast imaging shows a
combinations. Tumor cells are usually uniform with round or well-circumscribed mass with regular margins, which can be
angulated contour, mild nuclear atypia, and finely granular easily mistaken as a fibroadenoma in young patients.
or vacuolated cytoplasm containing dense eosinophilic Like adenoid cystic carcinoma, secretory carcinoma is
secretions. Signet ring cells can be present. Extracellular typically a low-grade triple-negative carcinoma with a
eosinophilic secretions are present within the lumens of basallike phenotype with expression of high-molecular-
tubules or microcysts. The eosinophilic secretions are posi- weight cytokeratins (CK5/6, 34E12, CK14, CK17),
tive for Periodic acid–Schiff (PAS), PAS diastase, and Alcian EGFR, and c-kit. Ki67 proliferative index is low (<15%).
blue. Ductal carcinoma in situ with similar secretory features The carcinoma cells are also positive for S100 (strong and
a b
Fig. 3.10 Secretory carcinoma of the breast. (a) The tumor is com- with cytologically bland tumor cells. (Courtesy of Dr. Shi Wei,
posed of irregular lobules of eosinophilic cells separated by band-like University of Alabama at Birmingham)
fibroconnective tissue. (b) Dense eosinophilic secretion is intermixed
diffuse) and mammaglobin but negative for GCDFP-15 nantly affects postmenopausal women with a median age of
[54, 55]. 60 years in contrast to triple-negative invasive ductal carci-
noma of no special type, which usually affects younger
27. Is secretory carcinoma associated with a better patients (<50 years) [27, 57, 58].
prognosis? Similar to ACC of the salivary gland, mammary ACC is
also composed of two populations of cells: glandular luminal
Secretory carcinoma usually manifests as an indolent, well- cells and basaloid cells, with three growth patterns: tubular,
circumscribed mobile lump with excellent prognosis. cribriform, and solid. The basaloid cells have myoepithelial
Axillary lymph node metastases may occur but rarely involve features. Eosinophilic hyaline or mucoid material may be
more than three lymph nodes. Secretory carcinoma should seen in the lumen of cribriform structures and tubules.
not be confused with invasive ductal carcinoma with apo- Carcinoma cells are usually small with scant cytoplasm and
crine features, which is more common and has a more vesicular nuclei without prominent nucleoli. The mitotic
aggressive behavior [54]. activity is low [59]. Nottingham histologic grading system is
also used for ACC of the breast. The solid variant of ACC is
28. What is genetic abnormality in secretory carcinoma? a high-grade variant with a more aggressive behavior. Tumor
cells in this variant are larger with moderate to marked
Secretory carcinoma is characterized with chromosomal nuclear pleomorphism and increased mitotic activity [60].
translocation t(12:15), resulting in the ETV6NTRK3 Mammary ACC is a triple-negative breast cancer with a
fusion gene. The ETV6 (TEL) oncogene encodes a tran- basallike phenotype. However, unlike most basal-like breast
scription factor involved in development. The same trans- cancers that are high grade with an aggressive clinical course,
location t(12:15) leading to ETV6NTRK3 fusion gene also mammary ACC except solid variant is usually low grade
occurs in congenital fibrosarcoma and mesoblastic with an indolent clinical course.
nephroma. FISH for the ETV6 break apart probe or ACC of the breast is morphologically similar to ACC of
RT-PCR for the ETV6NTRK6 fusion gene is a diagnostic the salivary gland. Recent studies reveal that both mammary
tool for these tumors [56]. and salivary gland ACCs share a recurrent translocation
t(6:9) which leads to the chimeric fusion gene MYBNFIB and
29. What is adenoid cystic carcinoma of the breast and may explain the phenotypic similarity [61, 62]. Clinical his-
what are its key diagnostic features? How does one tory is important to make a diagnosis of ACC of the breast
differentiate this entity from its counterpart in the instead of a metastasis from head/neck ACC.
head and neck? ACCs should be graded using the standard Nottingham
grading system with most exhibiting mild to moderate nuclear
Adenoid cystic carcinoma (ACC) of the breast, an ana- pleomorphism and low to moderate mitotic activity. As a
logue to its counterpart in the salivary gland, accounts for result, most are classified as histologic grade one or two
only about 0.1% of all breast carcinomas. ACC predomi- depending on the proportion of solid areas. See Fig. 3.11a–h.
54 Z. Li et al.
30. What is the immunohistochemical profile of adenoid ER or PR staining. Similar to ACC of the salivary gland, the
cystic carcinoma of the breast? basaloid cells of ACC of the breast are typically positive for
myoepithelial markers (p40, p63, smooth muscle myosin,
Mammary ACC cells are typically negative for ER, PR, and calponin, and S-100), basal cytokeratins (CK5 or CK5/6,
HER2 expressions; however, rarely they may exhibit weak CK14, and CK17), and epidermal growth factor receptor
Fig. 3.11 Adenoid cystic

carcinoma of the breast. a
Invasive cribriform nests of
carcinoma cells surrounded
by desmoplastic stroma (a).
Cribriform nests with
eosinophilic globular material
(b). P63 stains basaloid cells
of the tumor (c). Glandular
luminal cells of the tumor are
diffusely positive for c-Kit
(CD117) protein (d) and CK5
(e). Mucicarmine stains the
eosinophilic globular material
(f). The tumor cells are
negative for ER (g) and
PR (h)
c
d
f
56 Z. Li et al.
g
(EGFR) [63]. The glandular luminal cells are usually positive finding is confirmed with MYB RNA overexpression, which
for CK7, CK8/18, epithelial membrane antigen (EMA), car- can be demonstrated by in situ hybridization. Besides MYB
cinoembryonic antigen (CEA), and c-Kit (CD117) [64]. translocation, other genomic alterations in ACC of the breast
Interestingly, the CK5 or CK5/6 can be diffusely positive in include gains of 1p36.12-p35.3, 11p15.5, 12p13.31, 16p13.3,
the glandular luminal cells as well [27]. The proliferative and 19p13 and losses of 6p25.3-q26 and 9p11.1-q21.11 [69].
index labeled with Ki-67 is usually low but can be variable
depending on the variants or grading of the tumors. The 32. Is adenoid cystic carcinoma associated with a better
immunohistochemical profile of the mammary ACC is very prognosis?
similar to that of the basal-like triple-negative breast carci-
noma (TNBC); however, prognosis of mammary ACC is bet- ACC of the breast is usually indolent as a localized disease with
ter than that of basal-like TNBC [65–67]. Androgen receptor a low frequency of axillary lymph node metastasis (<8%) [71].
(AR) is negative in ACCs, but positive in around 30% of However, the solid variant of mammary ACCs has relatively
basal-like TNBCs [68]. higher incidence of the nodal metastases than classical ACCs,
which may indicate a more aggressive behavior [72]. Distant
31. What are the molecular features of adenoid cystic metastases may occasionally occur in patients with ACC of the
carcinoma of the breast? breast (<20%), most commonly to the lung or the bone [71, 73].
A breast-conserving surgical approach with or without
Similar to ACCs of the salivary gland, ACCs of the breast also radiotherapy is usually recommended for the treatment of
demonstrate recurrent t(6;9)(q22-23;p23-24) translocation ACC. Most studies have demonstrated an excellent clinical
with a MYB-NFIB gene fusion, resulting in an oncogenic fusion outcome with 10-year survival exceeding 90% after the treat-
protein with transcription factor function [61, 69, 70]. This ment. Patients with mammary ACC have a prolonged and
indolent clinical course even when they present with local are admixed with foci of squamous differentiation. The
recurrence or distant metastasis [74, 75]. lumens of the ducts are usually compressed. Eosinophilic
In ACCs of the salivary glands, MYB expression has been material or keratin may be present in the lumens. The tumor
associated with a better survival compared with MYB- cells show low-grade nuclear features. The tumor stroma can
negative ACCs [76]. However, the association of MYB be edematous or sclerotic and have variable spindle cells, but
expression with survival remains unknown in patients with the cellularity of the stroma around the epithelial nests is
ACC of the breast. often increased [79, 80]. See Fig. 3.12a–g.
33. What are the differential diagnoses for adenoid cys- 36. What are the differential diagnoses for low-grade
tic carcinoma of the breast? adenosquamous carcinoma of the breast?
The differential diagnosis of ACC includes other types of The differential diagnosis of low-grade adenosquamous car-
invasive breast carcinomas and intraductal lesions that have cinoma includes benign breast lesions, such as sclerosing
a cribriform growth pattern and collagenous spherulosis. adenosis, squamous metaplasia or syringomatous adenoma
Invasive cribriform carcinoma can be confused with ACC, of the nipple, and malignant lesions such as invasive tubular
but the cribriform carcinoma has only one cell type and has carcinoma. The absence of myoepithelial cells demonstrated
glandular lumina without the mucinous or basement mem- by immunostains (SMMS, p40 or p63) will help to exclude
brane material. In addition, most other types of breast carci- benign lesions [78]. The intramammary parenchymal loca-
nomas with a cribriform growth pattern are ER and PR tion of low-grade adenosquamous carcinoma is important to
positive and do not express p63 or c-kit. Collagenous spheru- differentiate it from syringomatous adenoma of the nipple.
losis, a benign breast lesion, can also be confused with ACC, Demonstrating squamous differentiation in low-grade ade-
especially as the p63 is expressed in both lesions. However, nosquamous carcinoma by careful sampling and histologic
ckit should not be expressed in collagenous spherulosis and examination is important to differentiate it from the invasive
can be helpful in the differential diagnosis. Another potential tubular carcinoma [80].
pitfall is with ACCs that have a predominantly nonclassical
growth pattern, such as the solid variant, which can be con- 37. What is fibromatosis-like metaplastic carcinoma and
fused with a higher-grade breast carcinoma. p63, EGFR, and what are its key diagnostic features?
c-kit may not be useful as these markers can be positive in
high-grade invasive ductal carcinoma. In such cases, the Fibromatosis-like spindle cell carcinoma (FLSCC) is a
FISH split-apart or fusion probes to detect the t(6;9) rear- recently described low-grade variant of metaplastic carci-
rangement and/or RTPCR for the MYBNFIB fusion gene noma with a favorable prognosis [81, 82]. FLSCC grossly
may be needed to establish the diagnosis of ACC. presents as a firm and white mass, and the cut surface shows
a fibrous, gray-white nodular parenchyma. Microscopically,
34. What are the current classification and subtypes of FLSCC shows the proliferation of cytologically bland, low-
metaplastic breast carcinoma? grade, spindled, fibroblast-like cells and stellate
myofibroblast-like cells, resembling fibromatosis. The cel-
Metaplastic carcinoma (MC) of the breast represents 0.25– lularity of proliferation of neoplastic cells is variable among
1% of all breast cancers diagnosed annually [27]. Based on FLSCCs. The neoplastic spindle cells show minimal nuclear
the 2012 World Health Organization classification of Tumors atypia and pale eosinophilic cytoplasm; the nuclei vary from
of the Breast, MC is classified based on the histological fea- thin, slender, spindled nuclei with tapered ends to more
tures of tumor cells: (1) purely epithelial components (low- plump, round to oval nuclei with discrete nucleoli. The
grade adenosquamous carcinoma, fibromatosis-like tumor border is usually infiltrative with broad, finger-like
metaplastic carcinoma, squamous cell carcinoma, and spin- projections into the surrounding tissue. Neoplastic squa-
dle cell carcinoma) and (2) mixed epithelial and mesenchy- mous or glandular epithelial elements may be present but
mal components (metaplastic carcinoma with mesenchymal should be less than 5% of the total tumor volume. FLSCC
differentiation and mixed metaplastic carcinoma) [27]. may also show collagenous stroma, similar to fibromatosis.
The presence of small, cohesive clusters of fusiform to
35. What is low-grade adenosquamous carcinoma of the polygonal epithelioid cells scattered among the spindle cells
breast and what are its key diagnostic features? is a defining and characteristic histologic feature for FLSCC
[83, 84]. See Fig. 3.13a–i.
Low-grade adenosquamous carcinoma is an uncommon vari- A panel of immunohistochemical stains is generally
ant of metaplastic carcinoma with a good prognosis [77, 78]. needed to demonstrate the epithelial origin of the spindle
Morphologically, round or comma-shaped infiltrating ducts cells in FLSCCs in order to differentiate it from other spin-
58 Z. Li et al.
a b
c d
e f
Fig. 3.12 Low-grade adenosquamous carcinoma of the breast. squamous and glandular differentiation (b). The tumor cells are dif-
Infiltrating solid glandular structures of carcinoma cells into surround- fusely positive for BerEP4 (c), CK5 (d), GATA3 (e), and p63 (f).
ing stroma (a). The tumor cells are low grade, bland looking with both SMMS stain shows loss of myoepithelial cells around the tumor (g)
nosed only after extensive sectioning and with negative

g
cytokeratin staining. Fibromatosis is a clonal proliferation
of benign-appearing fibroblasts and myofibroblasts with
an infiltrative growth pattern. The spindle cells in fibroma-
tosis are negative for cytokeratins but show positive stain-
ing of beta-catenin in the nuclei. Other differential
diagnoses include myofibroblastoma and pseudoangioma-
tous stromal hyperplasia (PASH). Myofibroblastoma is a
rare benign proliferation of myofibroblasts.
Myofibroblastomas were originally reported to occur more
frequently in males, but recent data suggest they are
equally frequent between males and females [87].
Histologically, myofibroblastomas are composed of bland-
appearing spindle cells in short haphazard fascicles sepa-
rated by collagen bands. Patchy perivascular chronic
Fig. 3.12 (continued) inflammatory infiltrates are characteristic findings. The
myofibroblast cells are positive for vimentin and variably
positive for desmin, CD34, smooth muscle actin, estrogen
receptor, progesterone receptor, and Bcl-2, but negative for
dle cell lesions of the breast. The cytokeratin immunohisto- cytokeratins [88, 89]. PASH is a benign lesion with
chemical stains can include antibodies against anastomosing empty, slit-like pseudovascular spaces lined
broad-spectrum cytokeratins (AE1/AE3 and pankeratin), by myofibroblasts (not endothelial cells) in a dense col-
basal cytokeratins (CK5, 34βE12, and CK14), and luminal lagenous stroma. Similar to those in the myofibroblastoma,
cytokeratins (CK7, CK19, and CAM 5.2). The spindle cell the spindle cells in the PASH are positive for vimentin and
component and small clusters of epithelioid cells usually variably positive for desmin, CD34, and smooth muscle
exhibit immunoreactivity for basal cytokeratins, but no to actin, but negative for cytokeratins [90].
focal immunoreactivity for luminal cytokeratins. It has
been suggested that the neoplastic spindle cells actually 39. What is squamous cell carcinoma of the breast and
demonstrate an immunoprofile more compatible with myo- what are its key diagnostic features?
epithelial differentiation with immunoreactivity of basal
cytokeratins (34βE12, CK14, and CK5) and myoepithelial Metaplastic squamous cell carcinomas can be pure or mixed
markers (smooth muscle actin, S100, p63 and p40). A study with other forms of invasive carcinoma. Pure squamous cell
has demonstrated that p63 was strongly positive in 87% of carcinomas in the breast are rare. More commonly, squamous
metaplastic carcinomas and was positive in all metaplastic differentiation is identified coexisting with invasive ductal car-
carcinomas with spindle cell and/or squamous differentia- cinomas and carcinomas with medullary features. Squamous
tion [85]. The neoplastic spindle cells of FLSCCs are nega- cell carcinomas usually present as cystic lesions with squa-
tive for smooth muscle myosin heavy chain (SMMHC) and mous lining cells showing variable atypia and nuclear pleomor-
epithelial membrane antigen (EMA); the proliferation phism. The tumor cells can show sheets, cords, or nests of
index of FLSCC is typically low with less than 5% of Ki-67 proliferation infiltrating into the stroma with a prominent stro-
staining [84, 86]. mal reaction and lymphocytic response [27]. See Fig. 3.14a–f.
38. What are the differential diagnoses for fibromatosis- 40. What are the differential diagnoses for squamous cell
like metaplastic carcinoma? carcinoma of the breast?
The main differential diagnoses for FLSCC include nodu- If the tumor is composed entirely of malignant squamous
lar fasciitis and fibromatosis. Nodular fasciitis is a benign cells, a metastasis from another site, especially skin,
proliferative lesion containing fibroblasts and myofibro- lung, or head/neck region, must be ruled out before mak-
blasts in myxoid stroma with prominent vasculature. The ing the diagnosis of mammary squamous cell carcinoma.
lesion is very rarely seen in the breast and should be diag- The other differential diagnosis is mucoepidermoid car-
60 Z. Li et al.
Fig. 3.13 Low-grade
fibromatosis-like spindle cell a
carcinoma (FLSCC). Broad
infiltrative projections of the
tumor extending into the
surrounding soft tissue (a).
The tumor is composed of
cytologically bland cells with
thin and spindled to round or
oval nuclei (b). The tumor
cells are positive for
cytokeratin AE1/AE3 (c),
MNF116 (d), and CK5 (e)
and positive for CK7 (f). The
tumor cells are focally
positive for GATA3 (g) and
negative for desmin (h) and
ER (i)
b
c
d
f
62 Z. Li et al.
g
cinoma (both low- and high-grade types), which usually the differential diagnosis of squamous cell carcinoma of
shows extracellular or intracellular mucin [91, 92]. the breast, including posttraumatic lobular squamous
Squamous metaplasia in the breast varies from syrin- metaplasia [93], mixed squamous–mucous cysts [94],
goma-like differentiation to inconspicuous foci in largely squamous metaplasia in gynecomastia [95], Zuska’s dis-
glandular lesions. Keratinizing cysts are uncommon, but ease (squamous metaplasia of lactiferous ducts), and
small osteocartilaginous foci can be seen [77]. Some infarction with squamous metaplasia of intraductal papil-
benign squamous lesions in the breast may also get into loma [96, 97].
41. What is spindle cell carcinoma (sarcomatoid carci- (storiform). The atypical spindle cells can range from bland
noma) of the breast and what are its key diagnostic appearing to highly pleomorphic. The cytoplasm can range
features? from elongated to plump spindle and the nuclei can range
from bland-looking to apparently pleomorphic. Mitotic rate
Spindle cell carcinomas of the breast can be pure or mixed can be variable among spindle cell carcinomas of the breast.
with other components, such as glandular, heterologous, or The spindle cells infiltrate into the surrounding stroma with
squamous elements [98, 99]. These tumors are composed entrapped benign ducts and lobules [100]. Nottingham
of atypical spindle cells in a growth pattern of long fasci- grading is not applicable to metaplastic spindle cell carci-
cles (herringbone or interwoven pattern) or short fascicles nomas [27].
Fig. 3.14 Metaplastic
squamous cell carcinoma of a
the breast. Invasive carcinoma
with both squamous
differentiation and ductal
differentiation with focal
necrosis and lymphocytic
response (a). Squamous
carcinoma cells show nuclear
pleomorphism and
keratinization (b). The tumor
cells are diffusely positive for
CK5 (c) and p40 (d) and
negative for ER (e) and PR (f)
c
64 Z. Li et al.
d
Spindle cell carcinomas of the breast can coexist with an muscle specific actin. Similar to other subtypes of metaplastic
epithelial component of invasive ductal carcinoma or ductal carcinoma, spindle cell carcinomas are generally negative for
carcinoma in situ. For any lesion with pure spindle cells, a sus- ER, PR, and HER2 [27, 100]. See Fig. 3.15a–i.
picion for metaplastic spindle cell carcinoma must be high so
that immunostains for epithelial differentiation should be per- 42. W
hat are the differential diagnoses for spindle cell
formed. A panel of cytokeratins is often necessary with a broad carcinomas of the breast?
spectrum of cytokeratins, including high- molecular-weight
cytokeratins (CK5/6 and 34βE12) and low-molecular-weight For spindle cell carcinomas of the breast, the main differential
cytokeratins. The neoplastic spindle cells usually express myo- diagnoses include malignant phyllodes tumor with prominent
epithelial markers such as p63, p40, smooth muscle actin, and spindle cell overgrowth, sarcomas (angiosarcoma, fibrosar-
Fig. 3.15 Spindle cell

carcinoma of the breast. The a
tumor is composed of atypical
spindle cells in a growth
pattern of long fascicles with
desmoplastic stromal reaction
(a). The atypical spindle cells
are mildly to moderately
pleomorphic and the
cytoplasm is mostly elongated
to plump spindly (b). The
tumor cells are diffusely
positive for cytokeratin AE1/
AE3 (c) and CAM5.2 (d) and
negative for CK7 (e). The
tumor cells are also diffusely
positive for p40 (f) and
negative for ER (g), PR (h),
b
and HER2 (i)
d
66 Z. Li et al.
e
h
i
coma, etc.), and benign spindle cell lesions, such as fibromato- 44. W
hat are the differential diagnoses for metaplastic
sis and PASH. Extensive sampling to identify malignant carcinomas with mesenchymal differentiation?
epithelial component is important, and epithelial immunohis-
tochemical markers such as cytokeratins and p63/p40 are The main differential diagnoses of metaplastic carcinoma
almost always necessary to make the diagnosis. The leaf-like with mesenchymal differentiation are sarcomas. Primary
architecture is characteristic of phyllodes tumor. The stromal breast sarcomas are exceedingly rare and most frequently
spindle cell proliferation in the phyllodes tumor is generally arise in association with a phyllodes tumor. To make a dis-
negative for cytokeratins and positive for CD34. The spindle tinction between these two entities, extensive sampling is
cells of fibromatosis usually show nuclear staining for beta- usually necessary to identify either malignant epithelial
catenin but negative staining for cytokeratins [27, 100]. component for diagnosis of metaplastic carcinoma with
mesenchymal differentiation or benign-appearing epithelial
43. What is metaplastic carcinomas with mesenchymal component and/or leaf-like architecture for phyllodes tumor
differentiation and what are its key diagnostic in cases with predominantly sarcomatous proliferation [27].
features?
45. What is the prognosis of most metaplastic carcino-
Metaplastic carcinomas with mesenchymal differentiation mas? Do all metaplastic carcinomas carry a bad
contain mesenchymal elements (cartilage, bone, rhabdoid, or prognosis?
a chondromyxoid matrix) admixed with carcinomatous com-
ponents [101]. The osseous and chondroid elements can Due to the heterogeneity of metaplastic carcinoma, the prog-
appear histologically benign or frankly malignant with an nosis largely depends upon the histologic features. Some
appearance of chondrosarcoma or osteosarcoma [101]. low-grade subtypes of metaplastic carcinomas such as low-
Extensive sampling may be necessary to identify epithelial grade adenosquamous carcinoma or fibromatosis-like meta-
components. At the same time, a broad panel of cytokeratins plastic carcinoma usually have a favorable prognosis with
may also be necessary to reveal the epithelial component only local recurrence and rare distant metastases, while oth-
when no apparent glandular component is present. ers (high-grade spindle cell carcinoma, metaplastic carci-
Metaplastic breast carcinomas with mesenchymal differen- noma with mesenchymal differentiation, or squamous cell
tiation originate from carcinomas that undergo sarcomatous carcinoma) have an aggressive clinical course with poor
transitions as a result of further genetic instability or muta- outcomes.
tions, and the identical clonality of the carcinomatous and In general, patients with metaplastic carcinoma have
mesenchymal components has been confirmed. The term larger tumors with negative hormone receptor status and
“matrix-producing carcinoma” was historically used for a less involvement of the regional lymph nodes [103].
subtype of metaplastic carcinomas with mesenchymal dif- However, even in the absence of lymph node metastasis,
ferentiation, which usually contains chondroid differentia- distant metastasis to the brain and lungs can occur [104,
tion or chondromyxoid matrix [102]. See Fig. 3.16a–c. 105]. The prognosis of fibromatosis-like metaplastic carci-
Similar to other subtypes of metaplastic carcinomas, meta- noma parallels that of fibromatosis, suggesting that wide
plastic carcinomas with mesenchymal differentiation are excision with clear margins or simple mastectomy without
also negative for ER, PR, and HER2 [27]. axillary lymph node dissections should be sufficient for
68 Z. Li et al.
Fig. 3.16 Metaplastic
carcinoma with mesenchymal a
differentiation. Metaplastic
carcinoma with
chondromyxoid matrix (a).
Metaplastic carcinoma with
chondromyxoid matrix
intermixed with malignant
epithelial cells (b).
Metaplastic carcinoma with
chondromyxoid matrix
intermixed with malignant
mesenchymal cells (c)
initial treatment of FLSCC; chemotherapy and radiation References

therapy may not be needed. However, the data are limited
and more studies are warranted. On the other hand, patients 1. Page DL, Dixon JM, Anderson TJ, Lee D, Stewart HJ. Invasive
cribriform carcinoma of the breast. Histopathology.
with high-grade metaplastic carcinomas usually have a 1983;7(4):525–36.
relatively poor prognosis and should be treated like 2. Venable JG, Schwartz AM, Silverberg SG. Infiltrating cribriform
Nottingham grade 3 invasive ductal carcinoma of the carcinoma of the breast: a distinctive clinicopathologic entity.
breast [106–108]. Hum Pathol. 1990;21:333–8.
3. Ng WK. Fine needle aspiration cytology of invasive cribriform tic” carcinoma with osteoclasts and osteoclast-like giant cells.
carcinoma of the breast with osteoclastlike giant cells: a case Histopathology. 1994;24:282–5.
report. Acta Cytol. 2001;45:593–8. 23. Viacava P, Naccarato AG, Nardini V, Bevilacqua G. Breast carci-
4. Diab SG, Clark GM, Osborne CK, Libby A, Allred DC, Elledge noma with osteoclast-like giant cells: immunohistochemical and
RM. Tumor characteristics and clinical outcome of tubular and ultrastructural study of a case and review of the literature. Tumori.
mucinous breast carcinomas. J Clin Oncol. 1999;17:1442–8. 1995;81:135–41.
5. Brandt SM, Young GQ, Hoda SA. The “Rosen Triad”: tubular car- 24. Wargotz ES, Norris HJ. Metaplastic carcinomas of the breast:
cinoma, lobular carcinoma in situ, and columnar cell lesions. Adv V. Metaplastic carcinoma with osteoclastic giant cells. Hum
Anat Pathol. 2008;15(3):140–6. Pathol. 1990;21(11):1142–50.
6. Rakha EA, Lee AH, Evans AJ, Menon S, Assad NY, Hodi Z, et al. 25. Tavassoéli FA, Devilee P. Pathology and genetics of tumours of
Tubular carcinoma of the breast: further evidence to support its the breast and female genital organs, World Health Organization
excellent prognosis. J Clin Oncol. 2010;28:99–104. classification of tumours. Lyon: IARC; 2003.
7. Di Saverio S, Gutierrez J, Avisar E. A retrospective review with 26. Foote FW Jr, Stewart FW. A histologic classification of carcinoma
long term follow up of 11,400 cases of pure mucinous breast car- of the breast. Surgery. 1946;19:74–99.
cinoma. Breast Cancer Res Treat. 2008;111:541–7. 27. Lakhani SR, Ellis IO, Schnitt SJ, Tan PH, van de Vijver MJ. WHO
8. Capella C, Eusebi V, Mann B, Azzopardi JG. Endocrine differ- classification of tumors of the breast, IARC WHO classification of
entiation in mucoid carcinoma of the breast. Histopathology. tumours. 4th ed. Lyon: IARC; 2012.
1980;4(6):613–30. 28. Ponsky JL, Gliga L, Reynolds S. Medullary carcinoma of the
9. Barbashina V, Corben AD, Akram M, Vallejo C, Tan LK. Mucinous breast: an association with negative hormonal receptors. J Surg
micropapillary carcinoma of the breast: an aggressive coun- Oncol. 1984;25:76–8.
terpart to conventional pure mucinous tumors. Hum Pathol. 29. Rosen PP, Menendez-Botet CJ, Nisselbaum JS, Urban JA, Miké V,
2013;44(8):1577–85. Fracchia A, et al. Pathological review of breast lesions analyzed
10. Ranade A, Batra R, Sandhu G, Chitale RA, Balderacchi for estrogen receptor protein. Cancer Res. 1975;35:3187–94.
J. Clinicopathological evaluation of 100 cases of mucinous carci- 30. Reiner A, Reiner G, Spona J, Schemper M, Holzner
noma of breast with emphasis on axillary staging and special ref- JH. Histopathologic characterization of human breast can-
erence to a micropapillary pattern. J Clin Pathol. 2010;63:1043–7. cer in correlation with estrogen receptor status: a comparison
11. Matsukita S, Nomoto M, Kitajima S, Tanaka S, Goto M, Irimura of immunocytochemical and biochemical analysis. Cancer.
T, et al. Expression of mucins (MUC1, MUC2, MUC5AC and 1988;61:1149–54.
MUC6) in mucinous carcinoma of the breast: comparison with 31. Jacquemier J, Padovani L, Rabayrol L, Lakhani SR, Penault-
invasive ductal carcinoma. Histopathology. 2003;42:26–36. Llorca F, Denoux Y, et al. Typical medullary breast carcinomas
12. Pettinato G, Manivel CJ, Panico L, Sparano L, Petrella G. Invasive have a basal/myoepithelial phenotype. J Pathol. 2005;207:260–8.
micropapillary carcinoma of the breast: clinicopathologic study 32. Rakha EA, Aleskandarany M, El-Sayed ME, Blamey RW, Elston
of 62 cases of a poorly recognized variant with highly aggressive CW, Ellis IO, et al. The prognostic significance of inflammation
behavior. Am J Clin Pathol. 2004;121:857–66. and medullary histological type in invasive carcinoma of the
13. Chen L, Fan Y, Lang RG, Guo XJ, Sun YL, Cui LF, et al. Breast breast. Eur J Cancer. 2009;45:1780–7.
carcinoma with micropapillary features: clinicopathologic 33. Marginean F, Rakha EA, Ho BC, Ellis IO, Lee AH. Histological
study and long term follow-up of 100 cases. Int J Surg Pathol. features of medullary carcinoma and prognosis in triple – nega-
2008;16:155–63. tive basal – like carcinomas of the breast. Mod Pathol. 2010
14. Paterakos M, Watkin WG, Edgerton SM, Moore DH 2nd, Thor Oct;23(10):1357–63.
AD. Invasive micropapillary carcinoma of the breast: a prognostic 34. Gaffey MJ, Mills SE, Frierson HF, Zarbo RJ, Boyd JC,
study. Hum Pathol. 1999;30:1459–63. Simpson JF, et al. Medullary carcinoma of the breast: interob-
15. Nassar H, Wallis T, Andea A, Dey J, Adsay V, Visscher server variability in histopathologic diagnosis. Mod Pathol.
D. Clinicopathologic analysis of invasive micropapillary differen- 1995;8:31–8.
tiation in breast carcinoma. Mod Pathol. 2001;14:836–41. 35. Lee AHS, Gillett CE, Ryder K, Fentiman IS, Miles DW, Millis
16. Eusebi V, Millis RR, Cattani MG, Bussolati G, Azzopardi RR. Different patterns of inflammation and prognosis in invasive
JG. Apocrine carcinoma of the breast. A morphologic and immu- carcinoma of the breast. Histopathology. 2006;48:692–701.
nocytochemical study. Am J Pathol. 1986;123:532–41. 36. Nielsen TO, Hsu FD, Jensen K, Cheang M, Karaca G, Hu Z,
17. Tramm T, Kim JY, Tavassoli FA. Diminished number or com- et al. Immunohistochemical and clinical characterization of the
plete loss of myoepithelial cells associated with metaplastic basal – like subtype of invasive breast carcinoma. Clin Cancer
and neoplastic apocrine lesions of the breast. Am J Surg Pathol. Res. 2004;10:5367–74.
2011;35(2):202–11. 37. Denkert C, Loibl S, Noske A, Roller M, Müller BM, Komor M,
18. Vranic S, Tawfik O, Palazzo J, Bilalovic N, Eyzaguirre E, Lee et al. Tumor-associated lymphocytes as an independent predictor
LM, et al. EGFR and HER-2/neu expression in invasive apocrine of response to neoadjuvant chemotherapy in breast cancer. J Clin
carcinoma of the breast. Mod Pathol. 2010;23(5):644–53. Oncol. 2010;28:105–13.
19. Agnantis NT, Rosen PP. Mammary carcinoma with osteoclast-like 38. Loi S, Sirtaine N, Piette F, Salgado R, Viale G, Van Eenoo F, et al.
giant cells. A study of eight cases with follow-up data. Am J Clin Prognostic and predictive value of tumor-infiltrating lymphocytes
Pathol. 1979;72:383–9. in a phase III randomized adjuvant breast cancer trial in node-
20. Iacocca MV, Maia DM. Bilateral infiltrating lobular carcinoma of positive breast cancer comparing the addition of docetaxel to
the breast with osteoclast-like giant cells. Breast J. 2001;7:60–5. doxorubicin with doxorubicin-based chemotherapy: BIG 02-98. J
21. Nielsen BB, Kiaer HW. Carcinoma of the breast with stromal mul- Clin Oncol. 2013;31:860–7.
tinucleated giant cells. Histopathology. 1985;9:183–93. 39. Adams S, Gray RJ, Demaria S, Goldstein L, Perez EA, Shulman
22. Herrington CS, Tarin D, Buley I, Athanasou N. Osteosarcomatous LN, et al. Prognostic value of tumor-infiltrating lymphocytes in
differentiation in carcinoma of the breast: a case of “metaplas- triple-negative breast cancers from two phase III randomized
70 Z. Li et al.
adjuvant breast cancer trials: ECOG 2197 and ECOG 1199. J Clin 60. Friedman BA, Oberman HA. Adenoid cystic carcinoma of the
Oncol. 2014;32:2959–66. breast. Am J Clin Pathol. 1970;54:1–14.
40. Cubilla AL, Woodruff JM. Primary carcinoid tumour of 61. Shin SJ, Rosen PP. Solid variant of mammary adenoid cystic car-
the breast: a report of eight patients. Am J Surg Pathol. cinoma with basaloid features: a study of nine cases. Am J Surg
1977;1(4):283–92. Pathol. 2002;26:413–20.
41. Azzopardi JG, Muretto P, Goddeeris P, Eusebi V, Lauweryns JM. 62. Persson M, Andrén Y, Mark J, Horlings HM, Persson F, Stenman
“Carcinoid” tumours of the breast: the morphological spectrum of G. Recurrent fusion of MYB and NFIB transcription factor genes
argyrophil carcinomas. Histopathology. 1982;6(5):549–69. in carcinomas of the breast and head and neck. Proc Natl Acad
42. Bussolati G, Gugliotta P, Sapino A, Eusebi V, Lloyd Sci. 2009;106:18740–4.
RV. Chromogranin-reactive endocrine cells in argyrophilic carci- 63. Arpino G, Clark GM, Mohsin S, Bardou VJ, Elledge RM. Adenoid
nomas (“carcinoids”) and normal tissue of the breast. Am J Pathol. cystic carcinoma of the breast: molecular markers, treatment, and
1985;120(2):186–92. clinical outcome. Cancer. 2002;94:2119–27.
43. Tang F, Wei B, Tian Z, Gilcrease MZ, Huo L, Albarracin CT, 64. Crisi GM, Marconi SA, Makari-Judson G, Goulart RA. Expression
et al. Invasive mammary carcinoma with neuroendocrine dif- of ckit in adenoid cystic carcinoma of the breast. Am J Clin Pathol.
ferentiation: histological features and diagnostic challenges. 2005;124:733–9.
Histopatholology. 2011;59(1):106–15. 65. Badve S, Dabbs DJ, Schnitt SJ, Baehner FL, Decker T, Eusebi
44. Sapino A, Papotti M, Righi L, Cassoni P, Chiusa L, Bussolati V, et al. Basal-like and triple-negative breast cancers: a critical
G. Clinical significance of neuroendocrine carcinoma of the review with an emphasis on the implications for pathologists and
breast. Ann Oncol. 2001;12(S2):S115–7. oncologists. Mod Pathol. 2011;24:157–67.
45. Park WM, Wu Y, Wei W, Yang WT. Primary neuroendocrine carci- 66. Wetterskog D, Lopez-Garcia MA, Lambros MB, A’Hern R, Geyer
noma of the breast: clinical, imaging and histologic features. AJR FC, Milanezi F, et al. Breast adenoid cystic carcinomas constitute
Am J Roentgenol. 2014;203(2):W221–30. a genomically distinct subgroup of triple-negative and basal-like
46. Wang J, Wei B, Albarracin CT, Hu J, Abraham SC, Wu Y. Invasive breast cancers. J Pathol. 2012;226:84–96.
neuroendocrine carcinoma of the breast: a population-based study 67. Thike AA, Cheok PY, Jara-Lazaro AR, Tan B, Tan P, Tan
from the surveillance, epidemiology and end results (SEER) data- PH. Triple-negative breast cancer: clinicopathological character-
base. BMC Cancer. 2014;14(1):147. istics and relationship with basal-like breast cancer. Mod Pathol.
47. Angarita FA, Rodriquez JL, Meek E, Sanchez JO, Tawil M, 2010;23:123–33.
Torregrosa L. Locally-advanced primary neuroendocrine carci- 68. Thike AA, Iqbal J, Cheok PY, Chong AP, Tse GM, Tan B, et al.
noma of the breast: case report and review of the literature. World Triple negative breast cancer: outcome correlation with immu-
J Surg Oncol. 2013;11(1):128. nohistochemical detection of basal markers. Am J Surg Pathol.
48. Shin SJ, DeLellis RA, Ying L, Rosen PP. Small cell carcinoma of 2010;34:956–64.
the breast: a clinicopathologic and immunohistochemical study of 69. Vranic S, Frkovic-Grazio S, Lamovec J, Serdarevic F, Gurjeva
nine patients. Am J Surg Pathol. 2000;24(9):1231–8. O, Palazzo J, et al. Adenoid cystic carcinomas of the breast
49. Kwon SY, Bae YK, Gu MJ, Choi JE, Kang SH, Lee SJ, et al. have low Topo IIα expression but frequently overexpress
Neuroendocrine differentiation correlates with hormone receptor EGFR protein without EGFR gene amplification. Hum Pathol.
expression and decreased survival in patients with invasive breast 2010;41:1617–23.
carcinoma. Histopathology. 2014;64(5):647–59. 70. Weigelt B, Horlings HM, Kreike B, Hayes MM, Hauptmann M,
50. Righi L, Sapino A, Marchio C, Papotti M, Bussolati Wessels LF, et al. Refinement of breast cancer classification by
G. Neuroendocrine differentiation in breast cancer: estab- molecular characterization of histological special types. J Pathol.
lished facts and unresolved problems. Semin Diagn Pathol. 2008;216:141–50.
2010;27(1):69–76. 71. Brill LB II, Kanner WA, Fehr A, Andrén Y, Moskaluk CA, Löning
51. Tavassoli FA, Norris HJ. Secretory carcinoma of the breast. T, et al. Analysis of MYB expression and MYB-NFIB gene
Cancer. 1980;45(9):2404–13. fusions in adenoid cystic carcinoma and other salivary neoplasms.
52. Akhtar M, Robinson C, Ashraf M, Godwin JT. Secretory carci- Mod Pathol. 2011;24:1169–76.
noma of the breast in adults. Cancer. 1983;51:2245–54. 72. Thompson K, Grabowski J, Saltzstein SL, Sadler GR, Blair
53. Rosen PP, Cranor ML. Secretory carcinoma of the breast. Arch SL. Adenoid cystic breast carcinoma: is axillary staging necessary
Pathol Lab Med. 1991;115:141–4. in all cases? Results from the California Cancer Registry. Breast J.
54. Krausz T, Jenkins D, Grontoft O, Pollock DJ, Azzopardi 2011;17:485–9.
JG. Secretory carcinoma of the breast in adults: emphasis on late 73. Fukuoka F, Hirokawa M, Shimizu M, Sadahira Y, Manabe T,
recurrence and metastasis. Histopathology. 1989;14(1):25–36. Kurebayashi J, et al. Basaloid type adenoid cystic carcinoma of
55. Lamovec J, Bracko M. Secretory carcinoma of the breast: light the breast. APMIS. 1999;107:762–6.
microscopical, immunohistochemical and flow cytometric study. 74. Ghabach B, Anderson WF, Curtis RE, Huycke MM, Lavigne JA,
Mod Pathol. 1994;7(4):475–9. Dores GM. Adenoid cystic carcinoma of the breast in the United
56. Tognon C, Knezevich SR, Huntsman D, Roskelley CD, Melnyk States (1977 to 2006): a population-based cohort study. Breast
N, Mathers JA, et al. Expression of the ETV6NTRK3 gene fusion Cancer Res. 2010;12:R54.
as a primary event in human secretory breast carcinoma. Cancer 75. Page DL. Adenoid cystic carcinoma of breast, a special histopath-
Cell. 2002;2(5):367–76. ologic type with excellent prognosis. Breast Cancer Res Treat.
57. Ro JY, Silva EG, Gallager HS. Adenoid cystic carcinoma of the 2005;93:189–90.
breast. Hum Pathol. 1987;18:1276–81. 76. Bell D, Roberts D, Karpowicz M, Hanna EY, Weber RS, El-Naggar
58. Anthony PP, James PD. Adenoid cystic carcinoma of the breast: AK. Clinical significance of Myb protein and downstream target
prevalence, diagnostic criteria, and histogenesis. J Clin Pathol. genes in salivary adenoid cystic carcinoma. Cancer Biol Ther.
1975;28:647–55. 2011;12:569–73.
59. Herzberg AJ, Bossen EH, Walther PJ. Adenoid cystic carcinoma 77. Rosen PP, Ernsberger D. Low-grade adenosquamous carcinoma:
of the breast metastatic to the kidney: a clinically symptomatic a variant of metaplastic mammary carcinoma. Am J Surg Pathol.
lesion requiring surgical management. Cancer. 1991;68:1015–20. 1987;11(5):351–8.
78. Van Hoeven KH, Drudis T, Cranor ML, Erlandson RA, Rosen unusual pseudosarcomatous metaplasia resembling squamous
PP. Low-grade adenosquamous carcinoma of the breast: a clini- (necrotizing) sialometaplasia of the salivary gland. Mod Pathol.
copathologic study of 32 cases with ultrastructural analysis. Am J 1988;1:385–90.
Surg Pathol. 1993;17(3):248–58. 94. Shousha S. An unusual cyst (of the breast). Histopathology.
79. Tan QT, Chuwa EW, Chew SH, Lim-Tan SK, Lim SH. Low-grade 1989;14:423–5.
adenosquamous carcinoma of the breast: a diagnostic and clinical 95. Gottfried MR. Extensive squamous metaplasia in gynecomastia.
challenge. Int J Surg. 2015;19:22–6. Arch Pathol Lab Med. 1986;110:971–3.
80. Ho BCS, Tan HW, Lee VKM, Tan PH. Preoperative and intra- 96. Zuska JJ, Crile G, Ayres W. Fistulas of lactiferous ducts. Am J
operative diagnosis of low-grade adenosquamous carcinoma Surg. 1951;81:312–7.
of the breast: potential diagnostic pitfalls. Histopathology. 97. Habif DV, Perzin KH, Lipton R, Lattes R. Subareolar abscess
2006;49:603–11. associated with squamous metaplasia of lactiferous ducts. Am J
81. Gobbi H, Simpson JF, Borowsky A, Jensen RA, Page DL. Metaplastic Surg. 1970;119:523–6.
breast tumors with a dominant fibromatosis-like phenotype have a 98. Raju GC, Wee A. Spindle cell carcinoma of the breast.
high risk of local recurrence. Cancer. 1999;85(10):2170–82. Histopathology. 1990;16:497–9.
82. Sneige N, Yaziji H, Mandavilli SR, Perez ER, Ordonez NG, Gown 99. Davis WG, Hennessy B, Babiera G, Hunt K, Valero V, Buchholz
AM, et al. Low-grade (fibromatosis-like) spindle cell carcinoma TA, et al. Metaplastic sarcomatoid carcinoma of the breast with
of the breast. Am J Surg Pathol. 2001;25(8):1009–16. absent or minimal overt invasive carcinomatous component: a
83. Rekhi B, Shet TM, Badwe RA, Chinoy RF. Fibromatosis-like car- misnomer. Am J Surg Pathol. 2005;29(11):1456–63.
cinoma – an unusual phenotype of a metaplastic breast tumor asso- 100. Carter MR, Hornick JL, Lester S, Fletcher CD. Spindle cell
ciated with a micropapilloma. World J Surg Oncol. 2007;27(5):24. (sarcomatoid) carcinoma of the breast: a clinicopathologic and
84. Dwyer JB, Clark BZ. Low-grade fibromatosis-like spindle cell car- immunohistochemical analysis of 29 cases. Am J Surg Pathol.
cinoma of the breast. Arch Pathol Lab Med. 2015;139(4):552–7. 2006;30(3):300–9.
85. Koker MM, Kleer CG. p63 expression in breast cancer: a highly 101. Wargotz ES, Norris HJ. Metaplastic carcinomas of the breast,
sensitive and specific marker of metaplastic carcinoma. Am J Surg I: matrix-producing carcinoma. Hum Pathol. 1989;20(7):
Pathol. 2004;28(11):1506–12. 628–35.
86. Dunne B, Lee AH, Pinder SE, Bell JA, Ellis IO. An immu- 102. Wargotz ES, Norris HJ. Metaplastic carcinomas of the breast, V:
nohistochemical study of metaplastic spindle cell carcinoma, metaplastic carcinoma with osteoclastic giant cells. Hum Pathol.
phyllodes tumor and fibromatosis of the breast. Hum Pathol. 1990;21(11):1142–50.
2003;34(10):1009–15. 103. Pezzi CM, Patel-Pareikh L, Cole K, Franco J, Klimberg VS, Bland
87. McMenamin ME, DeSchryver K, Fletcher CD. Fibrous lesions of K. Characteristics and treatment of metaplastic breast cancer:
the breast: a review. Int J Surg Pathol. 2000;8(2):99–108. analysis of 892 cases from the National Cancer Data Base. Ann
88. Rungta S, Kleer CG. Metaplastic carcinomas of the breast: diag- Surg Oncol. 2007;14(1):166–73.
nostic challenges and new translational insights. Arch Pathol Lab 104. Barnes PJ, Boutilier R, Chiasson D, Rayson S. Metaplastic breast
Med. 2012;136(8):896–900. carcinoma: clinical-pathologic characteristics and HER2/neu
89. Taccagni G, Rovere E, Masullo M, Christensen L. expression. Breast Cancer Res Treat. 2005;91(2):173–8.
Myofibrosarcoma of the breast: review of the literature on myo- 105. Lai HW, Tseng LM, Chang TW, Kuo YL, Hsieh CM, Chen ST,
fibroblastic tumors and criteria for defining myofibroblastic dif- et al. The prognostic significance of metaplastic carcinoma of the
ferentiation. Am J Surg Pathol. 1997;21(4):489–96. breast (MCB) – a case controlled comparison study with infiltrat-
90. Tse GM, Tan PH, Lui PC, Putti TC. Spindle cell lesions of the ing ductal carcinoma. Breast. 2013;22(5):968–73.
breast – the pathologic differential diagnosis. Breast Cancer Res 106. Shah DR, Tseng WH, Martinez SR. Treatment options for meta-
Treat. 2008;109(2):199–207. plastic breast cancer. ISRN Oncol. 2012;70:61–2.
91. Fisher ER, Palekar AS, Gregoria RM, Paulson JD. Mucoepidermoid 107. Luini A, Aguilar M, Gatti G, Fasani R, Botteri E, Brito JA, et al.
and squamous cell carcinomas of the breast with reference to Metaplastic carcinoma of the breast, an unusual disease with
squamous metaplasia and giant cell tumors. Am J Surg Pathol. worse prognosis: the experience of the European Institute of
1983;7:15–7. Oncology and review of the literature. Breast Cancer Res Treat.
92. Toikkanen S. Primary squamous cell carcinoma of the breast. 2007;101(3):349–53.
Cancer. 1981;48:1629–32. 108. Tse GM, Tan PH, Lui PC, Chaiwun B, Law BK. Metaplastic car-
93. Hurt MA, Diaz-Arias AA, Rosenholtz MJ, Havey AD, Stephenson cinoma of the breast: a clinicopathological review. J Clin Pathol.
HE Jr. Post-traumatic lobular squamous metaplasia of breast: an 2006;59(10):1079–83.
Lobular Breast Lesions
4
Megan L. Troxell, Yun An Chen, Jing Yu, Debra M. Ikeda,
and Kimberly H. Allison
Table 4.1 Clinical features of invasive lobular as compared to invasive

List of Frequently Asked Questions ductal carcinoma
Clinical feature Lobular Ductal
Invasive Lobular Carcinoma Focality Often multifocal Usually unifocal
Radiological Mass, architectural Mass, architectural
1. What are the differences in clinical, radiological, and imaging distortion, non-mass distortion, or
pathologic gross presentations, clinical management, enhancement (MRI), or calcifications.
occult Non-mass
and prognosis in invasive lobular carcinoma as com- enhancement (MRI)
pared to invasive ductal carcinoma? Gross Actual size of tumor Typically forms a
pathology often larger than what is dominant grossly
See Table 4.1. Also see questions 3 through 6 and question estimated on gross and evident mass
44 for further details. radiological examination
Management MRI may be performed Neoadjuvant
after biopsy to evaluate extent of chemotherapy may be
2. What is the incidence of invasive lobular carcinoma? disease; resect to given for tumor of high
negative margins grade, Her2+, or
• Invasive lobular is the second most common subtype of triple-negative receptor
status; resect to
female breast cancer, after ductal or so-called “no special negative margins
type” breast carcinoma. Distant Bone, GI, GYN ER+: Bone (and other
• Invasive lobular carcinoma accounts for 5–15% of inva- metastasis organs)
sive breast carcinomas based on current World Health ER-: Brain, lung, liver
Organization (WHO) estimates [1]; other sources cite
closer to 5% incidence [2].
• The incidence of invasive lobular carcinoma has been

estimated at 10.6/100,000 US women (based on registry
data from 1999 to 2004) [3]. The incidence of lobular
carcinoma was noted to decline over this interval, possi-
M. L. Troxell (*) · K. H. Allison
Stanford University School of Medicine, Department of Pathology, bly due to declining use of hormone replacement therapy
Stanford, CA, USA [3, 4].
e-mail: [email protected]; [email protected]
Y. A. Chen 3. How does invasive lobular carcinoma typically present
University of Washington, Department of Radiology, clinically?
Seattle, WA, USA
J. Yu Invasive lobular cancer may present in several ways:
Magee-Womens Hospital, University of Pittsburgh Medical Center,
Pittsburgh, PA, USA
e-mail: [email protected] • It is most commonly detected on screening imaging stud-
ies, although it can be imaging occult.
D. M. Ikeda
Stanford University School of Medicine, Stanford, CA, USA • It may come to attention as a clinically palpable breast
e-mail: [email protected] mass or masses. (It is commonly multifocal.)

74 M. L. Troxell et al.
• It may present as nipple retraction or, if advanced, with growth, invasive lobular carcinoma may be relatively more
breast retraction. However, Paget’s disease of the nipple is difficult to detect as compared to other subtypes of breast
usually caused by a ductal rather than lobular process. cancer [1, 5–7]. It may even be imaging occult on some
• If it has spread to lymph nodes, it may present with pal- studies [5–7].
pable, hard axillary lymph nodes.
• Occult lobular carcinoma can present with metastatic dis- • By mammography, invasive lobular carcinoma tends
ease to other organ sites, such as bone, abdominal organs, to present as some form of architectural finding, such
and serosal surfaces of the gynecologic and gastrointesti- as mass, spiculated mass, asymmetric density, or
nal tract. It is the most common type of breast cancer to architectural distortion (Figs. 4.1a–g and 4.2a–g) [1,
present with distant metastases from an occult primary. 5, 7]. Calcifications were noted in about 1/3 in one
study [5].
4. What are the typical features of invasive lobular carci- • The sensitivity of mammography for lobular carcinoma
noma on breast imaging (mammography, ultrasound, detection has been quoted as widely variable (34–92%) [6].
MRI)? • Digital breast tomosynthesis (DBT) is a newer variation
of X-ray mammography in which a series of angled digi-
Based on its growth pattern, interaction with surrounding tis- tal mammographic views are reconstructed to produce a
sue, paucity of microcalcifications, and often slow rate of three-dimensional view of the breast. It is said to enhance
a b c
Fig. 4.1 Invasive lobular carcinoma, breast imaging. A 47-year-old nonparallel orientation, with probable skin invasion, suspicious for can-
woman who noticed a lump in her lower left breast at the 5:30 position cer. (e) Axial contrast-enhanced breast MRI shows a 1.7 cm irregular
with associated skin dimpling. Craniocaudal (a) and mediolateral spiculated mass at the 5:00 position with skin tethering and skin inva-
oblique (b) full-field digital mammograms (FFDM) of a left breast pal- sion. The lesion is larger on MRI as compared to mammography.
pable mass. There is a two-view 8 mm focal asymmetry (arrow) with Kinetic analysis showed fast initial and washout delayed enhancement,
associated skin tethering. Note the linear scar marker in the upper outer suspicious for cancer. US-guided core biopsy with marker placement
left breast from a remote benign excisional biopsy. A skin marker above was performed, followed by lumpectomy. (f) Histologic sections of the
the nipple shows the superior extent of the areola. (c) FFDM spot com- lumpectomy specimen demonstrate an infiltrative breast cancer with
pression lateral medial mammogram clearly shows a round, irregular direct dermal invasion at low power; skin is at left (10×). (g) Higher
mass against the chest wall. (d) Real-time breast ultrasound shows an power demonstrates dyscohesive architecture and cytologic features of
irregular, spiculated hypoechoic mass with acoustic shadowing and invasive lobular carcinoma (200×). E-cadherin negative (not shown)
4 Lobular Breast Lesions 75
d e
f g
detection of architectural distortions and increase conspi- • The sensitivity of MRI for lobular carcinoma detection
cuity of invasive lobular carcinoma [8]. has been cited as 93–95% [6].
• By ultrasound, invasive lobular carcinoma is most often • Breast-specific gamma imaging (BSGI) utilizes a radio-
seen as a hypoechoic irregular area, or mass, with poorly tracer, often 99mTc sestamibi, to highlight metabolically
defined borders and posterior acoustic shadowing [1, 5, active sites and has shown high sensitivity in small studies
6]. One study reported mostly isoechoic lesions [6]. [5], yet it is not universally accepted.
However, small lesions may be relatively occult • Invasive lobular carcinoma is frequently undersized by
(Figs. 4.1a–g and 4.2a–g). each of these imaging modalities, as compared to final
• The sensitivity of ultrasound for lobular carcinoma detec- surgical pathology extent [8]. MRI tends to demonstrate
tion is about 68–98% [6]. the largest tumor extent yet may still underestimate in
• Magnetic resonance imaging (MRI, used with intrave- some cases.
nous gadolinium contrast) has the highest sensitivity • Breast imagers and pathologists should be wary of multi-
for lobular carcinoma detection. Invasive lobular carci- ple lesions in the setting of lobular differentiation. In
nomas may appear on MRI as mass-like lesions, often some institutions, patients diagnosed with lobular carci-
with spiculated margins, frequently with multiple noma undergo additional imaging, especially MRI, before
lesions. They also can present as enhancing areas surgical planning. Gross tissue sampling in pathology
(termed “non-mass enhancement”) and with the may also need to be more extensive to determine the
enhancement further described as focal, regional, duc- extent of a lobular carcinoma.
tal, clumped, segmental, or heterogeneous (Figs. 4.1a– • Additional breast imaging of lobular carcinoma is illus-
g and 4.2a–g) [5, 6]. trated with Cases 1, 4, and 5 at the end of this chapter.
a b c
d e
Fig. 4.2 Multifocal invasive lobular carcinoma, breast imaging. marker showing the location of the biopsied ILC. There also is linear
Woman with history of contralateral right breast invasive lobular carci- non-mass enhancement (NME) in the retroareolar region (double blue
noma (ILC), stage IIB (T2, N1) diagnosed 13 years earlier, treated with arrows), occult on mammography. (e) Computer-aided detection (CAD)
lumpectomy and radiation therapy, now with new imaging finding. MRI image shows fast initial and late washout kinetics (red color), in
Craniocaudal (a) and mediolateral oblique (b) digital synthetic mam- the NME (double blue arrow), suspicious for cancer. (f) Core biopsy of
mograms show a 5 mm ill-defined mass in the upper inner left breast, the mammographically detected mass demonstrates a densely cellular
posterior depth (circles). In (B) the mass appears just superior to a tumor composed of sheets of dyscohesive cells (40× magnification). (g)
prominent vessel. (c) Real-time breast ultrasound shows an irregular Core biopsy of the MRI-detected NME shows a somewhat less cellular
hypoechoic mass with an echogenic halo corresponding to the mam- infiltrative carcinoma (40× magnification). Both tumors were
mographic mass. US-guided core biopsy with marker placement was E-cadherin negative (not shown) with abundant pink-pale cytoplasm
performed, followed by breast MRI. (d) Axial contrast-enhanced breast and rather pleomorphic nuclei. Predictive markers from this case can be
MRI shows postlumpectomy change in the right breast; in the medial seen in Fig. 4.19
left breast there is a signal void (single red arrow) from the metallic
• f After inking and measuring the specimen, section it • g Submit sections to demonstrate largest extent of tumor
according to your protocol. Inspect the sections for tumor. and the status of all margins.
(See question 6.) • Do not simply submit “representative sections of tumor”;
• The College of American Pathologists (CAP) recom- a well-annotated gross submission map needs to be cre-
mends slicing at 0.4–0.5 cm intervals [11–13]. ated (Fig. 4.3).
5. What are the typical management steps after a 6. What are the gross pathologic features of invasive lob-
diagnosis of invasive lobular carcinoma on core ular carcinoma?
biopsy?
• Invasive lobular carcinoma may form a hard, gritty spicu-
The initial treatment of invasive lobular carcinoma shares lated mass on cut section.
many similarities with that of invasive breast carcinoma of • Many lobular carcinomas are difficult to identify grossly.
other types. Synthesis of visual inspection, palpation of the incised or
serially sectioned specimen, specimen X-ray, and clinical
• Based on clinical and radiologic features, imaging evalu- imaging findings may be needed to identify the carci-
ation of axillary lymph nodes may be performed. If suspi- noma, especially in a mastectomy specimen.
cious nodes are identified, they are usually sampled by • Identification of the prior biopsy site or deployed tissue
fine needle aspiration (FNA) or core biopsy. marker (“clip”) can be very helpful in locating the region
• Patients diagnosed with lobular carcinoma may undergo of the tumor. Prior biopsy site may appear as a small (or
additional imaging, especially MRI, before surgical rarely large) area of hemorrhage or fat necrosis.
planning.
• Surgical resection is often the first step in definitive treat- 7. How should a resected invasive lobular carcinoma be
ment (lumpectomy or mastectomy with axillary node sampled for histologic review?
sampling).
• A subset of patients may receive adjuvant therapy before Invasive cancers need to be sampled such that all data perti-
surgery (neoadjuvant) in an effort to shrink the cancer and nent to diagnosis, treatment, and staging are apparent on
facilitate less aggressive surgery. For lobular carcinoma, resulting histologic slides. As lobular cancers may often be
this may entail endocrine therapy (neoendocrine) for a larger than grossly or mammographically apparent, generous
prolonged period, or a standard course of cytotoxic che- but judicious sampling is recommended (Figs. 4.3 and 4.4).
motherapy. Estrogen receptor-positive invasive carcino-
mas with low proliferation fraction, including most • Correlate with imaging findings before grossing the spec-
lobular carcinomas, generally do not show robust response imen. Check tumor location, presumed size, focality, and
to cytotoxic chemotherapy [7, 9, 10]. Rates of pathologic lymph node status. If a specimen radiograph of a lumpec-
complete response (pCR) are quite low for classic inva- tomy was obtained, there may be comment on areas of
sive lobular carcinoma (6%) [7, 10]. margin concern.
Fig. 4.3 Grossing with Superior

cassette mapping. Top
diagrams slicing of a breast
specimen, after inking,
perpendicular to the longest
Lateral Medial
axis. Bottom: the slices are
laid out in order, with
orientation preserved, and the
specimen is X-rayed or
photographed. This is used to
create a map of tissue
sampling, with cassette
numbers superimposed. The
first and last sections should
be sectioned perpendicularly 1 2 3 4 5 6 7 8 9 10
(macroslices 8–10 not Inferior
shown). The tumor size for
staging would be derived
from composite of A5 + A6,
or the span across macroslices
2–5 (4 slices multiplied by
average slice thickness), and
should be correlated with size
by radiology and gross A1 A3 A5 A7 A9 A11 A13
assessment. Multifocal
carcinomas are sampled in
macrosection7, slides
A13 + A14 A2 A4 A6 A8 A10 A12 A14
1 2 3 4 5 6 7
• Mapping location of tissue cassettes directly onto an • Most often, the tumor dimension across multiple tissue
X-ray or photograph of the sliced specimen is most help- slices (e.g., medial-lateral) is the largest dimension.
ful in documenting the tumor sizes of complex specimens • The size of a cancer involving multiple slices can then be
(Figs. 4.3 and 4.4) [13]. calculated by multiplying the number of slices involved
• The largest contiguous histologic span of tumor will be by the slice thickness (0.4–0.5 cm as recommended
reported and used for staging purposes. An annotated map above) [13].
should ensure that size can be accurately calculated based • The average thickness of each slice can also be estimated
on the histologic findings in relation to the gross by dividing the overall specimen dimension by the total
sampling. number of slices [13].
• It is recommended to sample from sections beyond the • However, small errors in slice thickness estimates can
grossly obvious extent of tumor to ensure adequate size result in significant over- or underestimations if many
estimation. slices are involved, so calculated size should be recon-
• If the cancer only involves a single slide, the size can be ciled with imaging findings and gross exam.
measured on the single slide. However, most often it • If there is more than one tumor mass, sample in-between
involves multiple sections, and either a composite section them to determine whether they are connected by occult
or span across multiple sections will need to be used in invasive tumor.
size calculations. • Sample other areas of gross abnormality and/or imaging
• For a composite section of a single slice, the tumor can be density.
outlined with a dotting pen on each glass slide, and the • Carefully dissect axillary tissue (separate specimens or
adjacent slides juxtaposed to measure extent across the attached to mastectomy) for lymph nodes; submit all can-
composite. didate nodes.
Fig. 4.4 Radiograph of sectioned surgical specimen. Breast imaging radiography. Histologic sections demonstrated lobular carcinoma con-
studies demonstrated two masses, both biopsy-proven invasive lobular tinuously involving the entire span between markers (4 cm), along with
carcinoma: 1.3 × 1.1 cm at the ribbon-shaped marker (upper right) and a microinvasive satellite lesion in a separately submitted margin speci-
1.1 × 0.8 cm at the wing-shaped marker (bottom left). In the surgical men (not shown). The cassette map can be superimposed on a specimen
specimen. Masses were ill-defined by gross examination and specimen radiograph or photograph, as diagramed in Fig. 4.3
8. Does lobular carcinoma originate from breast nal” (or ER-positive) neoplasia pathway [15]. (See
lobules? question 35.)
• A putative precursor lesion, LCIS, was first identified in and 9. What are the characteristic cytologic features of clas-
named by its involvement of lobules. However, this is a sic invasive lobular carcinoma [1, 2, 9, 16, 17]?
misnomer, and the assumption that lobular carcinoma arises
in lobules is thought to be incorrect. (See question 55.) • Bland dyscohesive, small cells (Figs. 4.5a–f and 4.6a–i)
• Lobular and low-grade ductal carcinomas are both thought • Central or eccentric nuclei, often round, sometimes com-
to originate from the terminal duct lobular unit [14]. pressed or squared off by adjacent cells
• Invasive lobular carcinoma shares molecular similari- • Inconspicuous nucleoli
ties with the ductal carcinomas in the low-grade “lumi- • Pale cytoplasm, usually scant (even plasmacytoid)
• Occasionally intracytoplasmic mucin vacuoles liferative rates) and worse clinical outcomes [1, 2, 21]. (See
• Must be distinguished from plasma cells and/or question 13 for pleomorphic subtype.)
lymphocytes Architectural variants (Fig. 4.8a–f) (usually have typical
cytologic features as described in question 10):
1 0. What are the characteristic architectural fea-
tures of classic invasive lobular carcinoma [1, 2, • Solid: confluent sheetlike growth of invasive lobular cells,
9, 16, 17]? or large nests. May have higher mitotic rate or greater
nuclear pleomorphism (grade 2 nuclei) [1, 2, 15, 21]
• “Single file” cells infiltrating tissue (Figs. 4.5a–f and • Alveolar: globular arrangements of at least 20 cells [1, 2,
4.6a–i). 15, 21]
• Linearly arranged cords of cells, 1–2 cells thick. • Trabecular: cords (trabeculae) 2–3 cells thick [2, 21]
• Concentric cords around normal ducts, or lobules, impart- • Tubulolobular: admixed tubular growth pattern (see ques-
ing a so-called “targetoid” architecture. tion 23 for discussion of immunostaining) [1, 2, 15]
• Near absence of duct, gland, or tubule formation (although • Mixed: admixture of classic and variant pattern
these can rarely form).
• Relative lack of stromal desmoplasia. Cytologic variants (Fig. 4.9a–d) (usually have single file
• Density of lobular carcinoma cells varies greatly from architecture or can have any of the above patterns):
case to case.
• Associated ALH/LCIS, which often demonstrates a “pag- • Pleomorphic: Greater degree of cellular atypia and pleo-
etoid” growth pattern. morphism and higher mitotic rate than classic lobular car-
cinoma [1, 2, 9, 15, 20] (See question 13.):
11. What are type A, type B, and pleomorphic lobular –– Generally grade 3 nuclei or grade 2 nuclei with abun-
cells? dant mitoses.
–– Cells and nuclei are markedly larger than those classic
• Type A: classic lobular cells with small nuclei as described lobular carcinoma, nuclei more hyperchromatic, with
above (grade 1, Fig. 4.7a–c) [15–19]. prominent nucleoli.
• Type B: slightly larger cells with more variation in nuclear –– Recognizable lobular features include dyscohesion
and cell size (larger) and shape, often with paler chroma- and central to eccentric nuclei that are somewhat
tin and/or small nucleoli, more like grade 2 nuclei (twice round.
the size of lymphocyte nuclei, Fig. 4.7a–c) [15–19]. –– Despite the terminology, nuclear pleomorphism (cell
• Pleomorphic: lobular cells with grade 3 nuclei or grade to cell variation) is less than that of a high-grade ductal
2 nuclei with increased mitotic rate. Cells and nuclei (or no special type) breast carcinoma, and nuclei
are markedly larger than those of classic lobular carci- remain somewhat round in our experience [20]. (See
noma, with nuclei that are more hyperchromatic, with question 11.)
prominent nucleoli (Fig. 4.7a–c) [1, 2, 9, 15, 20]. (See –– May have apocrine features (abundant eosinophilic
question 13.) cytoplasm, GCDFP-15 and androgen receptor posi-
tive, molecular apocrine) [2, 16, 22].
12. What are the morphologic variants of invasive lobu- • Signet ring: Dyscohesive lobular cells with prominent
lar carcinoma? intracytoplasmic mucin vacuoles [1, 16]
• Histiocytoid: Dyscohesive lobular cells with abundant
The World Health Organization (WHO) classification pale finely granular cytoplasm [2, 9, 23]:
describes solid, alveolar, pleomorphic, tubulolobular, and –– Histiocytoid lobular may mimic reactive histiocytes,
mixed variants of lobular carcinoma (Figs. 4.8a–f and 4.9a– histiocytic neoplasm, or granular cell tumor and be
d) [1]. Most are architectural variants, whereas some are difficult to recognize as carcinoma on H&E sections
based on cytologic characteristics. The recognition of these [23].
variants can be helpful because they can mimic ductal (or no –– Many histiocytoid examples are androgen receptor
special type) carcinoma and may benefit from confirmatory (AR) positive, strongly GCDFP-15 positive, and
studies. (See questions 14, 23.) In addition, the pleomorphic, thought to have apocrine features [16, 22, 23].
solid, and alveolar subtypes have been associated with more • Signet ring (or histiocytoid) variants more likely to have
aggressive features (such as HER2 positivity or higher pro- pleomorphic cytologic features [15, 23]
a b
c d
e f
Fig. 4.5 Classic histology of invasive lobular carcinoma. (a) Densely tubules in edematous stroma. (d) Less cellular example with pale cyto-
cellular example of invasive lobular carcinoma with scant cytoplasm, plasm and dyscohesive single file cells in sclerotic stroma (a–d 400×).
slightly irregular nuclei and cord-trabecular architecture. (b) (e) Lobular carcinoma infiltrates as scattered cells within a fibrous
Dyscohesive invasive lobular carcinoma with cytoplasmic mucin vacu- septal area. Such foci are relatively occult radiologically (100× magni-
oles. Mucin droplets appears as a pink dot in intracytoplasmic lumens. fication). (f) Lobular carcinoma infiltrates around normal structures in a
(c) Invasive lobular carcinoma with single file cells and rudimentary “targetoid” pattern (200× magnification)
a b c
d e f
g h i
Fig. 4.6 Lobular carcinoma can be difficult to identify when paucicel- view reveals population of pale epithelioid cells in upper right and
lular, admixed with inflammation or granulomas. (a) Densely hyalin- lower center (200×). (f) Keratin stain confirms invasive carcinoma with
ized breast stroma with atrophic duct (top), prominent vessels and a few dispersed cells (100×); E-cadherin was negative (not shown). (g) This
linear collections of cells (100×). (b) Higher power demonstrates core biopsy demonstrated granulomatous inflammation (top), appar-
chains of dyscohesive epithelioid cells suspicious for lobular carcinoma ently with associated epithelioid histiocytes (bottom, 100×). (h) High-
(left, 200×). (c) Estrogen receptor (ER) staining is positive in epitheli- power view shows that the pink mononuclear cells have uniform round
oid cell clusters in a nearby but unmatched field (200×). While this nuclei, suspicious for carcinoma rather than epithelioid histiocytes. A
result is compatible with lobular carcinoma, stromal cells may also be vague granuloma is at lower right (400×). (i) Keratin staining reveals
ER positive (arrows); keratin staining is more specific and definitive. nearly confluent dyscohesive invasive lobular carcinoma cells between
(d) Breast core biopsy with inflammation (100×). (e) Higher-power granulomas. E-cadherin was negative (not shown) (100×)
a b c
Fig. 4.7 Cytologic heterogeneity of invasive lobular carcinoma. (a) size to lymphocytes at bottom right (“Type B,” 400×). (c) Pleomorphic
Infiltrating cells with small nuclei, no nuclei and scant cytoplasm lobular carcinoma with grade 3 vesicular nuclei. Nucleoli are apparent
(arrows, examples). Compare tumor cell size to normal breast glands but not especially large in this example. A few lymphocytes and stromal
(“Type A,” 400×). (b) Invasive lobular carcinoma with larger slightly cells are at far left (400×)
vesicular nuclei, and nuclei still with scant cytoplasm. Compare nuclear
a b
Fig. 4.8 Architectural variants of invasive lobular carcinoma. (a) lobular carcinoma with alveolar architecture upper left and trabecular
Densely cellular carcinoma demonstrating the classic single file archi- architecture lower right (200×). (e) Another example with alveolar
tecture (200×). (b) Solid variant characterized by sheets of tumor cells architecture (200×). (f) Rudimentary tubule formation in a lobular car-
(40×). (c) Higher power of the solid invasive lobular carcinoma shown cinoma (200×)
in b; these tumor cells are also rather pleomorphic (200×). (d) Invasive
c d
e f
13. Why is it important to recognize pleomorphic inva- • Pleomorphic lobular carcinoma shares more aggressive
sive lobular carcinoma? molecular features with high-grade ductal (or no special
type) breast carcinoma such as gains of chromosome
• Pleomorphic invasive lobular carcinoma has higher 8q24, 17q12, and 20q13, among others [2, 9, 15,
nuclear grade, higher mitotic rate, and thus higher overall 26–28].
grade than classic lobular carcinoma. Because of its high • Pleomorphic lobular carcinoma behaves more aggres-
grade, it can mimic a ductal carcinoma [1, 2, 9, 16]. sively with higher recurrence rate and lower 10-year sur-
• Higher-grade, more proliferative cancers may receive vival [1, 2, 7, 15, 22].
benefit from chemotherapy.
• Pleomorphic lobular carcinomas can have lower levels of 14. How is invasive lobular carcinoma distinguished
hormone receptor expression than classic invasive lobular from invasive ductal (or “no special type”)
carcinoma (and sometimes are ER and PR negative) [9, carcinoma?
22, 24–26].
• Pleomorphic lobular carcinomas have a HER2 positivity Invasive lobular carcinoma has characteristic architectural
rate more similar to ductal carcinomas (between 10% and and cytologic features, described in questions 9 and 10.
20%). When HER2 positive, patients should be offered Together, these are often quite distinctive from other types
HER2-targeted therapies (in addition to chemotherapy) of breast cancer, as listed in Table 4.2 and illustrated in
[15, 22, 24–26]. HER2 mutations have also been found in Fig. 4.10a–f. Variants of lobular carcinoma are listed in
pleomorphic lobular carcinoma [9, 24]. question 12.
a b
c d
Fig. 4.9 Cytologic variants of invasive lobular carcinoma. (a) large nuclei and nucleoli (compare to lymphocytes) and scant cyto-
Pleomorphic lobular carcinoma. Nuclei are very large as compared to plasm (400×). (c) Lobular carcinoma with signet ring cells; enlarged
normal breast epithelium and lymphocytes at left. Dyscohesive pleo- nuclei are distorted by mucin vacuoles (arrows, 400×). (d) Lobular car-
morphic lobular cells have abundant eosinophilic cytoplasm, suggestive cinoma with abundant pale foamy “histiocytoid” cytoplasm and at least
of apocrine features (400×). (b) Pleomorphic lobular carcinoma with grade 2 nuclei
Table 4.2 Histopathologic features of invasive lobular as compared to invasive ductal carcinomas
Feature Classic lobular Low-grade ductal/NST High-grade ductal/NST
Architecture Single file cells, no glands Well-formed glands (tubule Usually sheets, large nests
score 1–2)
Cell cohesion Dyscohesive Cohesive Usually cohesive
Cell polarization None, central or eccentric Columnar with polarized nuclei Variable
nuclei
Cell size Small cells, scant to Small to moderate size, often Large, variable cytoplasm
moderate cytoplasm more cytoplasm
Nuclei Grade 1–2, round regular Grade 1–2, regular, less round Grade 2–3, large, hyperchromatic, irregular
(see question 18) shape
Nucleoli Inconspicuous Inconspicuous Variable (often large)
E-cadherin (Table 4.3, Negative Positive along basolateral Positive along basolateral cell-cell
question 23) cell-cell membranes membranes, lost in small subset
p120 (Table 4.3, question 23) Cytoplasmic Positive along basolateral Positive along basolateral cell-cell membranes
cell-cell membranes
NST no special type
Invasive lobular carcinoma Invasive ductal, low grade Invasive ductal, high grade
a b c
d e f
Fig. 4.10 Comparison of classic invasive lobular carcinoma, low- can have highly variable morphology (40×). (d) Invasive lobular carci-
grade invasive ductal (no special type), and high-grade invasive ductal noma at higher power. Note single file cells, low-grade cytology, and
(no special type) carcinoma. (a) Low-power view of invasive lobular stromal desmoplasia in this case (200×). (e) Low-grade ductal carci-
carcinoma with single and single file cells and areas of variable density noma at higher power has low nuclear grade and well-formed tubules.
(40×). (b) Low power of low-grade invasive ductal carcinoma; note (f) High-grade ductal carcinoma with high nuclear/cytoplasmic ratio
prominent tubule formation (40×). (c) Low-power view of high-grade and abundant mitotic figures; high-grade ductal carcinoma can have
ductal carcinoma with large sheets of tumor, some with necrosis; these highly variable morphology
15. Are intracytoplasmic mucin droplets specific for lob- 16. Does invasive lobular carcinoma always have “single
ular carcinoma? file” architecture?
No, ductal (or no special type) breast carcinomas may also No, variant invasive lobular carcinoma includes solid, tra-
have intracytoplasmic mucin, although this feature is more becular, and alveolar architecture. (See question 12,
common in lobular carcinoma (Figs. 4.5a–f and 4.9a–d). Fig. 4.8a–f.)
a b c
Fig. 4.11 Grade 3 invasive ductal carcinoma with focal single cell lobular or pleomorphic lobular differentiation. (c) E-cadherin is strongly
architecture. (a) Ductal architecture is seen in upper portion, with single positive (membranous) throughout the tumor, consistent with morpho-
cell and cord-like infiltration at bottom (400×). (b) Nuclear contour and logic impression of invasive ductal carcinoma
hyperchromasia is typical of ductal carcinoma, without cytology of
17. Can invasive ductal (or “no special type”) carcinoma • Assign overall grade based on sum of points.
ever have “single file” architecture? –– Grade 1: 3–5 points
–– Grade 2: 6–7 points
Yes, invasive ductal carcinoma may also invade as “single –– Grade 3: 8–9 points
file” cells. Cytologic features with the addition of
immunohistochemistry when needed can help make this dis- Most classic lobular carcinomas receive scores of 5 or
tinction (Fig. 4.11a–c). (For immunohistochemistry, see 6 in the above schema, on the cusp of grade 1–grade 2 [9,
question 23.) 15]. Dyscohesion is a hallmark of lobular carcinoma, so
most are assigned 3 points for tubules; classic lobular car-
18. How is invasive lobular carcinoma graded? cinomas have a low mitotic rate (1 point). Thus, the
nuclear score largely influences the final grading.
Every invasive lobular carcinoma should be graded according to Pleomorphic lobular carcinomas have higher nuclear
the current WHO criteria, also termed Elston & Ellis, grade and higher mitotic rate and should calculate to
Nottingham, or modified Scarff-Bloom-Richardson grade [29]. grade 3.
Points are assigned for each of the following three features:
19. What are the diagnostic criteria for microinvasive
• Tubule and gland formation. lobular carcinoma?
–– 1: majority of carcinoma (>75%)
–– 2: moderate (10–75%) Microinvasive carcinoma is defined as invasive carcinoma
–– 3: little or no (<10%) measuring less than or equal to 1 mm (0.1 cm), regardless of
• Nuclear pleomorphism. subtype [30–32]. Small invasive lobular carcinomas can be
–– 1: small regular uniform cells very difficult to recognize histologically. An area appearing
–– 2: moderate increase in size and variability to have increased stromal cellularity should be examined at
–– 3: marked variation high power to rule out an invasive lobular carcinoma or other
• Mitotic rate: see microscope field diameter calibration in subtle findings. Small or low cellularity lobular carcinomas
WHO or CAP checklists. may mimic plasma cells or other inflammatory cells. Keratin
–– 1: low stains can be used to highlight infiltrating lobular cells and
–– 2: moderate confirm the diagnosis [9].
–– 3: high
20. What molecular mechanisms are responsible for the lobular breast cancer has lesser incidence and mortality than
characteristic architectural features of lobular gastric cancer [41]. Recently, hereditary lobular breast can-
carcinoma? cer without gastric cancer has been described [42].
Loss or dysfunction of the cell-cell adhesion molecule 22. What are the mechanisms of E-cadherin loss or
E-cadherin is characteristic of lobular carcinoma [1, 2, 9, downregulation in lobular carcinoma?
14–16, 27, 32–40].
There are several mechanisms of E-cadherin downregulation
• E-cadherin (Epithelial-cadherin) is a transmembrane pro- [1, 9, 14, 15, 27, 35, 43–45].
tein that forms calcium-dependent homodimers between
cadherin molecules on adjacent cells [14, 32–40]. • A frameshift mutation in the CDH1 gene (50–65% of
• E-cadherin is an integral part of adherens junctions and cases) resulting in E-cadherin protein truncation, often in
links to the cytoskeleton through catenins [14, 32–40]. combination with changes listed below.
–– alpha-catenin • Loss of heterozygosity of the other allele (>90% of cases);
–– beta-catenin 16q is frequently lost in low-grade invasive breast cancers.
–– gamma-catenin (also called plakoglobin) • Epigenetic silencing (promoter methylation), transcrip-
–– p120 catenin tional repression (less common).
• E-cadherin is encoded by the CDH1 gene.
• E-cadherin and catenins also influence cell polarization 23. How are E-cadherin and catenin immunohistochem-
and motility [14, 40]. istry utilized to discriminate lobular from ductal
differentiation?
21. Are there heritable mutations associated with risk of
lobular carcinoma? Lobular carcinoma is characterized by loss of functional
E-cadherin, and E-cadherin immunohistochemistry is a
Germline mutation in CDH1 (16q22.1) encoding E-cadherin helpful ancillary tool in differentiating invasive ductal and
is associated with hereditary diffuse gastric cancer and lobu- invasive lobular carcinoma (Figs. 4.12 and 4.13a–h,
lar carcinoma of the breast [35, 41, 42]. In these families, Table 4.3) [1, 2, 14–16, 27, 36, 39, 40].
Fig. 4.12 Immunohistoche Lobular Normal and Ductal

mical staining patterns in
ductal and lobular
differentiation. Top row:
E-cadherin and beta-catenin
are localized to basolateral E-cadherin &
cell membranes in normal and beta-catenin
ductal cells, but are negative
or weak in invasive lobular
carcinoma and ALH/
LCIS. Middle row: p120
catenin localizes to
basolateral cell membranes in
normal breast and ductal cells.
In lobular lesions, p120
catenin is relocalized to the
cytoplasm (cytoplasmic
granular), without membrane p120
staining. Bottom row:
expected staining with
E-cadherin (brown), p120
(red) dual stain
Dual
stain
Invasive lobular with normal Invasive ductal
a b
H&E
c d
E-cadherin
e f
P120 catenin
g h
Beta-catenin
Fig. 4.13 Immunophenotype of classic invasive lobular and invasive duc- catenin demonstrates granular cytoplasmic staining without membranous
tal carcinoma. Left column: invasive lobular carcinoma with focal normal accentuation. All three proteins have membranous expression in normal
acini (bottom center of each panel). Serial sections stained with (a) H&E, acini. Right column: invasive ductal carcinoma in serial sections including
(c) E-cadherin, (e) p120 catenin, (g) and beta-catenin (all 400×). In typical (b) H&E, (d) and cell membrane localization of E-cadherin, (f) p120
lobular carcinomas, E-cadherin and beta-catenin are negative, while p120 catenin, (h) and beta-catenin, all membranous (all 400×)
• E-cadherin is normally localized to basolateral cell mem- • Loss or dysfunction of E-cadherin results in loss or redis-
branes. Normal ducts or acini serve as a good comparative tribution of cytoplasmic catenins (beta-catenin, p120
control. catenin, respectively) (Fig. 4.12), also amenable to immu-
• Negative, fragmented, or very weak E-cadherin staining nohistochemistry, with p120 more widely published than
is generally seen in lobular carcinoma, with caveats below beta-catenin [1, 2, 14–16, 27, 37, 39, 40, 46–52].
[1, 2, 14–16, 27, 36, 39, 40]. • Caveats include:
–– The interpretation E-cadherin staining is not always
straightforward. Complete absence is not required for
Table 4.3 E-cadherin and catenin staining of lobular as compared to
ductal proliferations (also see Figs. 4.12, 4.13, 4.14, 4.15, 4.16, 4.17,
a lobular diagnosis. These patterns are also compatible
4.18, 4.34, and 4.36) with lobular differentiation [14, 16, 40, 51–53].
Immunostain Lobular Ductal
◦◦ Membranous E-cadherin staining that is markedly
E-cadherin Absent or weak Membrane weaker than normal internal control is aberrant,
p120 Cytoplasmic, granular, no membrane Membrane favoring lobular.
staining ◦◦ Partial or fragmented E-cadherin staining.
Beta- Absent or weak Membrane ◦◦ Cytoplasmic or Golgi pattern E-cadherin staining
catenin (Fig. 4.14a–d) [14, 40, 51–53].
Data from Refs. [1, 2, 14–16, 27, 36, 39, 40]
a b
c d
Fig. 4.14 Aberrant E-cadherin with cytoplasmic localization. (a) H&E p120 localization confirms E-cadherin dysfunction and lobular differ-
sections demonstrate invasive carcinoma with cords and alveolar archi- entiation. (d) In another focus of invasive lobular carcinoma from the
tecture. (b) E-cadherin is strongly positive in the carcinoma, but is even same case, cytoplasmic E-cadherin staining is weaker in the dyscohe-
throughout the cytoplasm, without membrane localization. This can be sive single file cells (upper right), and the pattern of staining contrasts
difficult to recognize in cases with scant cytoplasm. (c) Cytoplasmic with that of normal alveoli in the left portion of the field
–– A small fraction of lobular carcinomas may show posi- –– The interpretation of cytoplasmic and membranous
tive E-cadherin membrane expression (Fig. 4.15a–d) p120 staining is uncertain and should be considered
[14, 16, 40, 51, 52]. (See question 25.) equivocal.
◦◦ We generally perform E-cadherin staining first and –– Some commercial stain cocktails include E-cadherin
if results are unexpected add catenin (p120) stain- and p120 catenin (Fig. 4.12).
ing to resolve unusual cases. ◦◦ Staining conditions for multi-stains often represent
◦◦ Aberrant catenin staining (non-membranous) is a compromise and are not ideal for either antibody.
consistent with lobular carcinoma. ◦◦ Care must be taken with internal controls and inter-
–– Rarely, nonfunctional E-cadherin may be cytoplasmic pretation. An overly strong red stain may mask
or perinuclear (Golgi). This pattern supports lobular weak brown or vice versa.
carcinoma, and catenins will be redistributed. This
emphasizes the importance of the pattern of staining 24. Must E-cadherin immunostaining be performed in
(Fig. 4.14a–d). order to render a diagnosis of invasive lobular
–– Several studies affirmed that many tubulolobular carci- carcinoma?
nomas are E-cadherin positive (membranous) and are
thus best classified as invasive ductal carcinoma [54– No, it is not essential to demonstrate loss of E-cadherin in
56]. However, we have seen E-cadherin-negative diagnosing lobular carcinoma [57]. The diagnosis can be
examples of tubulolobular carcinoma and prefer to made based on typical architectural and cytologic features.
classify these as lobular (Fig. 4.16a–d). Immunostaining may be helpful if there is morphologic
a b
c d
Fig. 4.15 Invasive lobular carcinoma with membranous E-cadherin. E-cadherin is localized to cell membranes, although weak-moderate in
(a) This carcinoma has architectural and cytologic features of invasive intensity (400×). (d) p120 has cytoplasmic localization, indicating dys-
lobular carcinoma (200×). (b) Higher power demonstrates at least grade function of E-cadherin complex, supporting lobular differentiation in a
2 nuclei and the leading edge infiltrating adipose tissue (400×). (c) serial section (400×). Beta-catenin was also negative (not shown)
a b
c d
Fig. 4.16 Tubulolobular carcinoma. (a) High power demonstrates negative for E-cadherin (c) and have cytoplasmic p120 catenin (d), a
rudimentary tubules with lumens (400×). (b) Another area with more lobular pattern. Classic lobular architecture was seen in some areas of
well-formed tubules (200×); serial sections for immunostaining are this carcinoma (not shown)
doubt or mixed features (sheet or trabecular architecture, immunostain, loss of E-cadherin is neither necessary nor
nuclear irregularities, etc.) [14, 40]. sufficient for diagnosis of lobular carcinoma.
25. Can a diagnosis of lobular carcinoma ever be made if 27. How is mixed ductal lobular carcinoma defined?
E-cadherin is positive?
The designation mixed ductal lobular carcinoma has been
A small fraction of lobular carcinomas may maintain mem- used differently in different centers in different eras [1, 9, 16,
branous E-cadherin expression, without proper function 57, 59].
(15% in one study) [14, 16, 40, 51, 52]. Catenin staining
(especially p120, also beta-catenin) should be helpful in • The WHO monograph does not explicitly define this
resolving these cases (Fig. 4.15a–d). In our experience, the terminology, but discusses that the mixed terminol-
majority of carcinomas with lobular-like morphology and ogy may be used if the special type (lobular) makes
positive E-cadherin (membrane) also have membranous up 10–90% of the cancer recognized special type
catenin staining (ductal pattern). (such as lobular) in 10–49% of the tumor [1]. The
WHO definition differs for the in situ situation. (See
26. Can a diagnosis of ductal carcinoma ever be made if question 75.)
E-cadherin is negative? • We agree with Naidoo’s definition of mixed ductal and
lobular carcinoma as “A tumour is regarded as being of
A small fraction of high-grade ductal carcinomas may also mixed type, for example, ILC and ductal/no special type
lose E-cadherin (5–10%) [9, 32, 33, 58]. While a useful (IDC/NST), if unequivocal separate areas of both
orphological types are present and not if the lesion

m • Intermediate or indeterminate features may include
shows indeterminate features” (Fig. 4.17a–d) [57]. [16, 60]:
–– In one study using this definition, mixed tumors –– Ductal cytology with single file/cord architecture
showed metastatic patterns similar to invasive lobular –– Lobular cytology and single file/cord architecture with
carcinoma [59]. occasional tubule formation
–– In another study using different definitions, invasive ductal –– Lobular cytology and/or single file/cord architecture,
carcinoma with lobular features had clinical and biologic E-cadherin positive (Fig. 4.18a–d)
characteristics more similar to invasive lobular carcinoma –– Focal presence or absence of E-cadherin staining,
(ductal, E-cadherin positive, with areas of small cells dis- weak or equivocal E-cadherin/catenin staining
persed, in linear cords, or loose aggregates) [60]. –– Uncertain morphology without available E-cadherin
• If the carcinoma shows intermediate or indeterminate fea- stain (especially at core biopsy)
tures, we use diagnostic terminology and an explanatory –– For E-cadherin and other ancillary immunostaining
comment, such as: (see question 23)
–– Invasive mammary carcinoma • A recent molecular study categorized “mixed” histology
–– Invasive carcinoma with ductal and lobular features tumors as predominantly lobular or ductal based on
–– Invasive ductal carcinoma with lobular features mRNA expression profile [43].
–– Invasive ductal carcinoma with lobular growth pattern
a b
c d
Fig. 4.17 Mixed ductal and lobular carcinoma (WHO definition). (a) p120 catenin is cytoplasmic in lobular carcinoma (left), with weak
H&E demonstrating a single tumor with areas of single file and cord- membrane staining in the ductal component of this microscopic field;
like architecture (lobular, left) and glandular architecture (ductal, right), membrane staining was stronger in other ductal areas of the tumor. (d)
with concordant immunohistochemistry. (b) E-cadherin demonstrates Beta-catenin shows a pattern similar to E-cadherin. (all 100×)
membrane staining in ductal area and is negative in lobular (left). (c)
a b
c d
Fig. 4.18 Invasive ductal carcinoma with lobular growth pattern. (a) Beta-catenin is positive (membranous). The immunohistochemistry
High-power H&E-stained section demonstrating lobular cytology and supports ductal differentiation. This could be described as invasive duc-
single file or small clusters of cells without tubule formation. (b) tal carcinoma or invasive ductal carcinoma with lobular growth pattern
E-cadherin demonstrates membrane staining. (c) p120 also demon- or with lobular features (all 400×)
strates membrane localization, unlike the case shown in Fig. 4.15. (d)
28. How is invasive lobular carcinoma staged? ◦◦ T1mi: 0.1 cm or less (microinvasive)
◦◦ T1a: more than 0.1 up to 0.5 cm
Invasive lobular carcinoma is staged similarly to other breast ◦◦ T1b: more than 0.5 cm up to 1 cm
cancers, using the AJCC TNM system [1, 30, 31]. A simplified ◦◦ T1c: more than 1 cm up to 2 cm
version of pathologic TNM categories is listed below; please –– T2: Tumor more than 2 cm up to 5 cm in greatest
refer to the AJCC manual or CAP staging checklists for full dimension.
detail [30, 31]. The 8th edition was updated to incorporate data –– T3: Tumor more than 5 cm in greatest dimension.
on hormone receptor, Her2, proliferation rate, and Oncotype –– T4: Tumor of any size with direct extension to chest
DX recurrence scores into an overall prognostic stage in wall and/or skin (ulceration or skin nodules). See
extensive tables which can be found in the AJCC manual [30, AJCC manual for further details and substaging
31]. However, the anatomic TNM staging also still applies. [30, 31].
• pN – Nodes (regional) (see question 43)
• pT – Tumor (primary) –– NX: Cannot be assessed.
–– TX: Cannot be assessed. –– N0: No regional lymph node metastasis or only isolated
–– T0: No evidence of primary tumor. tumor cells (clusters of tumor cells not more than
–– Tis: Carcinoma in situ. 0.02 cm = 0.2 mm and less than 200 cells; see question 40).
–– T1: Tumor 2 cm or less in greatest dimension. ◦◦ pN0(i+): ITCs only
–– N1: Micrometastasis or metastasis in 1–3 axillary tumor on ink.” It specifically addressed lobular carcinoma
ipsilateral; see AJCC for internal mammary nodes in summary point 6: “Wider negative margins than no ink
[30, 31]. on tumor are not indicated for invasive lobular carcinoma
◦◦ N1mi: larger than 0.02 cm (0.2 mm) and/or more (ILC)” [61].
than 200 cells, but none larger than 0.2 cm (2 mm), However, in practice these guidelines are still considered
micrometastasis. controversial for cases with <1–2 mm margins that may not
◦◦ N1a: metastasis in 1–3 axillary lymph nodes, be receiving additional therapies. Multiple clinical, imaging,
including at least one larger than 0.2 cm (2 mm). and pathologic factors should be considered in the setting of
◦◦ N1b,c: internal mammary nodes involved; see multiple close margins in an invasive lobular carcinoma.
AJCC manual [30, 31].
–– N2: Metastasis in 4–9 axillary lymph nodes, including 31. What is the typical hormone receptor/Her2 profile
at least one larger than 0.2 cm (2 mm); see AJCC for and intrinsic type of classic invasive lobular
internal mammary nodes [30, 31]. carcinoma?
–– N3: Metastasis in ten or more axillary lymph nodes,
including at least one larger than 0.2 cm (2 mm); see Most classic invasive lobular carcinomas are estrogen recep-
AJCC for internal mammary nodes [30, 31]. tor (ER) positive and Her2 negative, consistent with the so-
• pM – distant Metastasis called “luminal” intrinsic type, as also demonstrated by gene
–– M0: No distant metastasis expression analysis [1, 2, 7, 15, 27, 45, 62–65]. Those with
–– M1: Distant metastasis low proliferation (often also strongly progesterone receptor
(PR) positive) are luminal A (see case 2), whereas those lob-
29. How should surgical margins be reported for inva- ular cancers with higher proliferation are luminal B (and
sive lobular carcinoma? may be low or negative PR) [1, 2, 7, 15, 27, 45, 62–65].
Reviews cite 95% ER positivity and 60–70% PR positivity
• Tumor resection specimens (lumpectomy, partial mastec- [1, 2, 7, 15, 21, 27, 45].
tomy, mastectomy) should be oriented and inked to pre-
serve margin status as described in question 7. 32. What is the typical hormone receptor/Her2 profile
• Ideally, the closest approach of invasive tumor to each sur- and intrinsic type of pleomorphic invasive lobular
gical margin should be reported (distance in cm or mm). carcinoma?
–– If tumor is well removed from margin (greater than
0.5 cm, greater than 1 cm), exact distance may not As compared to classic lobular carcinoma, pleomorphic lob-
need to be stated. ular is less likely to be ER positive [1] and more likely to be
–– If carcinoma is within 0.5 cm of margin, stating dis- Her2 positive (Fig. 4.19a–d) [15, 22, 24, 25]. (See question
tance of tumor to margin in millimeters is recom- 13 and case 6.) Thus, few pleomorphic lobular cancers would
mended, without further qualification or interpretation be of luminal A type, while some would be luminal B or
(without positive/negative). Her2.
–– Designation of a margin as positive or involved by car-
cinoma is reserved for cancer present at surgical mar- 33. Can the Oncotype DX assay be applied to invasive
gin (ink on carcinoma, carcinoma at cauterized lobular carcinoma?
specimen margin, distance to margin =0) [61].
–– If a margin is positive, the extent of margin involvement Yes, the Oncotype DX assay is being applied to invasive
should be described, best with a distance measurement. lobular carcinomas.
• In cases with a main lumpectomy specimen and sepa-
rately submitted final margins, practices may differ. • The Oncotype assay is a commercial RT-PCR assay
–– Some groups provide status of specimen margins offered by Genomic Health (Redwood City, CA, USA)
of the main specimen and along with diagnosis and for ER-positive, Her2-negative invasive carcinomas lim-
margins for each separately submitted specimens. ited to the breast or with low lymph node disease burden.
–– Other groups prefer to synthesize results of all separate • This assay evaluates mRNA expression levels of 16
specimens and report “final” margin status. genes along with 5 housekeeping genes; based on those
results, a “recurrence score” is mathematically calculated
30. What typically defines acceptable surgical margins which has been shown to have utility in predicting out-
for invasive lobular carcinoma? come with tamoxifen-alone treatment, in predicting ben-
efit of chemotherapy, and has been incorporated into
Recent consensus criteria have proposed that adequate NCCN treatment guidelines and the eighth edition AJCC
excision for invasive breast carcinoma of any type is “no staging [45, 66, 67].
a b
c d
Fig. 4.19 Hormone receptor studies in pleomorphic lobular carci- variable positive staining in LCIS (bottom) but is negative in the inva-
noma. (a) H&E stain shows invasive pleomorphic lobular carcinoma sive carcinoma. The positive nuclei are spindle stromal cell nuclei,
(top) and carcinoma in situ (bottom, 200×). (b) E-cadherin is negative which should not be overinterpreted as positivity in carcinoma (red
in both components, supporting lobular differentiation, despite the arrows, examples). (d) Her2/neu immunohistochemistry is negative (1+
residual intraductal spaces. The staining at the edge of the involved duct barely perceptible membrane staining, all 200×)
represents compressed myoepithelial cells. (c) Estrogen receptor shows
• Most initial studies were done on invasive ductal or no grade invasive ductal carcinoma. Differences include expres-
special type breast carcinoma. There are no outcome stud- sion of genes related to cell adhesion (E-cadherin program),
ies of lobular carcinoma, yet lobular carcinomas fre- actin cytoskeleton remodeling, and cell migration, with other
quently fall into this ER+ Her2- group. pathways differing by study [27, 43–45, 62–65]. Expression
• Recent small single-center studies demonstrate the vast profiling has also suggested subgroupings of lobular carcino-
majority of classic invasive carcinomas fall into the low- mas (immune related, and/or hormone related, reactive, pro-
and intermediate-risk group or have recurrence scores liferative) [15, 27, 43, 44, 71].
below 25, different in distribution than cohorts of ductal
carcinomas [45, 68–70]. 35. Are there recurrent genomic changes in invasive lob-
ular carcinoma (mutations, chromosomal gains/
34. How is the mRNA expression profile of classic lobu- losses)?
lar carcinoma different from low-grade ductal
carcinoma? Classic invasive lobular carcinoma has genomic features of
the “low-grade” breast carcinoma molecular pathway, with
Classic invasive lobular carcinomas generally belong to the the addition of E-cadherin loss. This often includes [1, 9, 15,
“luminal” intrinsic type of breast cancers, along with low- 27, 43–45, 71]:
• Gain of 1q, 16p 38. What are the features of lobular carcinoma after
• Loss of 16q (site of CDH1, E-cadherin gene) neoadjuvant or neoendocrine therapy?
Enriched in lobular Neoadjuvant therapy is cytotoxic chemotherapy before sur-

• PI3Kinase pathway dysregulation: gery. Neoendocrine therapy is anti-hormonal therapy such as
–– Activating mutations in PIK3CA (45%) tamoxifen or aromatase inhibitor prior to surgery
–– Activating mutations in AKT1 (4%) (Fig. 4.20a–d).
–– Inactivation of PTEN (4%)
• Truncating mutations of CDH1 • Proliferation rate generally decreases.
• Frequent mutation of FOXA1, TBX3 (inactivating) • Some tumors have minimal response and demonstrate the
• Increased copy number of ESR1 (estrogen receptor) same tumor cell density and tumor cell cytology after
• Paucity of TP53, GATA3 mutations, ERBB2 amplification chemotherapy.
(ILC may be slightly enriched in ERBB2 mutation.) • Tumors with good response have decreased cell density.
• Lobular carcinoma cells may become small and pyc-
36. What adjuvant therapy is typically given after resec- notic and closely resemble lymphocytes. Keratin stains
tion of invasive lobular carcinoma? may be needed to demonstrate residual tumor cells
(Fig. 4.20a–d) [9].
After resection of invasive lobular carcinoma to appropriate • Carcinomas with complete response have no residual
negative margins, most patients are recommended to receive: invasive carcinoma in the breast, no intralymphatic tumor,
and no carcinoma in lymph nodes. Areas of tumor may be
• Adjuvant hormonal therapy with tamoxifen or especially replaced by fibrous tissue in the breast and/or lymph
an aromatase inhibitor for 5 years or longer [7, 9, 15]. nodes.
• Adjuvant radiation therapy for patients who had breast- • It is essential to know pre-treatment status and correctly
conserving surgery (lumpectomy or partial mastectomy). identify and thoroughly sample the tumor bed in breast
• Radiation is considered in a subset of patients treated with cancer resection specimens.
mastectomy, such as those with large tumors, margin issues,
many positive lymph nodes, or other adverse features [7]. 39. Does invasive lobular carcinoma occur in men?
• Adjuvant cytotoxic chemotherapy: see question 37.
Yes, invasive lobular carcinoma occurs rarely in men, esti-
37. What is the role of cytotoxic chemotherapy in treatment mated at about 1% of male breast cancer, comprising a lower
of invasive lobular carcinoma (neoadjuvant, adjuvant)? proportion than in female breast cancer. A recent SEER
study catalogued 155 cases of male lobular breast cancer,
• Adjuvant cytotoxic chemotherapy may be considered, but with poorer 5- and 10-year unadjusted cancer-specific sur-
classic lobular carcinoma that is ER positive with low vival (both 82.9%) as compared to women with lobular
proliferation rate does not respond as well as higher-grade breast cancer (91.9% and 84.5%, respectively) [72]. This
breast carcinomas [2, 7]. study documented some treatment differences, as well as
• In the neoadjuvant setting (chemotherapy before surgery), more higher-grade lobular cancers, raising the possibility
the rate of pathologic complete response (pCR) for classic that registry data could have included some misclassified
invasive lobular carcinoma is low (6%) and lower than high-grade ductal cancers [72].
that of ductal carcinoma overall [7].
• Patients with multiple tumors have different responses to
neoadjuvant therapy based on tumor biology, for example: Metastatic Lobular Carcinoma
–– Excellent or even complete pathologic response of
high-grade “triple-negative” ductal carcinoma 40. How is lobular carcinoma recognized in lymph nodes
–– Minimal response in concurrent invasive lobular carci- (Figs. 4.21a–f, and 4.24a–d)?
noma (see Case 3)
• Cytotoxic chemotherapy should be considered for lobular Metastatic lobular carcinoma shares cytologic features
carcinomas with higher proliferation and more aggressive with primary lobular carcinoma and can be problematic
biologic profile (for instance, pleomorphic lobular to recognize in lymph nodes. Look for the following pat-
carcinoma). terns [2]:
a b
c d
Fig. 4.20 Invasive lobular carcinoma resected after endocrine therapy increased cellularity; epithelioid tumor cells with minimal cytoplasm
(neoendocrine therapy). (a) Obvious residual tumor cell clusters with are scattered among vascular or stromal cells (200×). (d) Keratin stain
perineural invasion (left, 200×). (b) Keratin stain of the same area can help identify the few small residual tumor cells (200×)
(200×). (c) Another area with dense collagenized stroma and minimal
• “Pinker” areas of lymph node may represent nodular or Immunostaining for keratin is helpful and sometimes essen-
diffuse involvement by metastatic lobular carcinoma. tial in confirming or refuting the presence of lobular carcinoma
Metastatic cells are slightly larger than the background in lymph nodes (Figs. 4.22a–f and 4.24a–d). While keratin stain-
lymphocytes with slightly more pale or pink cytoplasm ing of sentinel lymph nodes is no longer routinely performed in
(Figs. 4.21a–f and 4.24a–d). Eosinophilia may also be most centers, keratin staining of nodes in the setting of lobular
due to accompanying collagen fibrosis. carcinoma should be considered, especially when [73–76]:
• Small collections of lobular carcinoma in or near sinu-
soids. These may closely mimic sinusoidal histiocytes. • There is histologic difficulty in discerning histiocytes
Histiocytes have more filmy, foamy, less eosinophilic from potential lobular carcinoma.
cytoplasm as compared to lobular carcinoma and may • There is histologic concern for small clusters or dispersed
have more variable nuclear contours (bean-shaped, non- lobular carcinoma cells.
round). Lobular carcinomas have more uniform cytoplas- • It is important to note that isolated tumor cells are the most
mic density and monotonous nuclei (Figs. 4.22a–f and common additional finding when using immunohisto-
4.24a–d). Immunohistochemical staining may be essen- chemistry to detect metastatic lobular carcinoma; isolated
tial to diagnosis. tumor cells have minimal clinically significance currently
• Dispersed single cells. These may blend almost impercep- [73, 74, 77–79]. (See question 43.) Therefore, the routine
tibly with the surrounding nodal tissue and can be con- use of immunohistochemistry to detect isolated tumor
firmed by keratin staining (Figs. 4.23a–d and 4.24a–d). cells is not recommended.
a b c
d e f
Fig. 4.21 Lymph node involvement by lobular breast carcinoma. (a) higher-power examination reveals lobular carcinoma (not shown). (d)
At low power, this area of the node is unusually pink; there is an obvi- Subcapsular and septal infiltrate of dyscohesive atypical cells (200×).
ous interfollicular infiltrate of metastatic lobular carcinoma (100×). (b) (e) Keratin stain highlights extent of tumor in a serial section (200×). (f)
Keratin staining of a serial section (100×). There is more than 2 mm of Higher power demonstrates pleomorphic lobular cells, some with intra-
confluent involvement by tumor in this node, which is classified as a cytoplasmic mucin (400×). This focus is >200 cells and > 0.2 mm, but
(macro)metastasis. (c) This subcapsular metastasis is also pink but less than 2 mm, and is categorized as micrometastasis (mi)
blends in somewhat with the background lymph node (arrows, 200×);
• Keratin immunostains are recommended over other breast ogy, and location compatible with lobular carcinoma [74].
differentiation or histiocytic histochemical or immuno- Re-review the H&E sections looking for the cells high-
histochemical stains in this scenario [74]: lighted by immunohistochemistry.
–– While lobular carcinomas are ER positive, a subset of • Keratin immunostaining pitfalls in axillary lymph nodes
nodal lymphocytes, especially germinal center cells, include (Fig. 4.23a–d):
are also ER positive, such that ER is not a specific stain –– Dendritic-like cells with fine elongated processes com-
in the context of nodal metastasis [80]. monly stain for keratin (CAM5.2 more so than AE1/3).
• In interpreting keratin stains, a positive stain will have These cells may be light brown, but they do not have
strong dark staining in cells with size, nuclear morphol- the correct morphology of lobular carcinoma [74].
a b
c d
e f
Fig. 4.22 Lymph node involvement by lobular breast carcinoma. (a) In round nuclei and scant cytoplasm. (d) Red area shows histiocytes and
this “pink” node, metastatic carcinoma involved one portion (lower lymphocytes without carcinoma. Histiocytes have more abundant filmy
left), with abundant histiocytes at right (100×). The extent of tumor was or foamy cytoplasm as compared to carcinoma, somewhat exaggerated
very difficult to discern morphologically and is nicely illustrated in (b) in this field. Additional images of metastatic lobular carcinoma (e) and
with keratin stain (100×). Higher power from the boxed regions is histiocytes (f) from other areas of the same case
shown below. (c) Black box shows carcinoma with monotonous mostly
a b
c d
Fig. 4.23 Pitfalls of keratin immunostaining in axillary lymph nodes, vidual cells since they are not confluent (cells not touching one another).
with isolated tumor cells. (a) Subcapsular tumor cells can be identified This node is not included in the count of positive nodes. (c) A cluster of
by their pleomorphic nuclei, signet ring, and epithelioid features (black three keratin-positive metastatic lobular carcinoma cells is marked by
arrows, 200×). (b) They are highlighted in keratin stain of serial section. the black arrow. The keratin stain also weakly labels dendritic cells that
The red arrow highlights an extraneous keratin-positive squamous epi- have different morphology (red arrows, examples, 400×). (d) Numerous
thelial cell. Note the difference in morphology (larger, no nucleus, out clusters and single cells of strongly keratin staining metastatic lobular
of plane) and staining (pale) as compared to tumor cells in the same carcinoma are seen; the red arrow highlights a small cluster of weakly
field (200×). This node has <200 tumor cells and is classified as isolated staining plasma cells. This weak staining does not indicate metastatic
tumor cells (itc). The measurement should not include the span of indi- carcinoma (400×)
–– Plasma cells may weakly cross-react with keratin anti- zurin) blue dye and/or radiocolloid (usually Tc-99m sulfur
bodies. These may have morphology close to lobular colloid) is injected at the tumor site or dermis around the
carcinoma, but should not stain strongly [74]. areola [74, 81, 82]. After massage and time, the “hot” and/or
–– Large filmy polygonal usually anucleate cells out of “blue” lymph nodes are collected as the “first” nodes drain-
the plane represent squamous cells from handling of ing the breast tumor area and thus the most likely to harbor
the slides prior to staining, coverslipping. These do not metastatic tumor, if any [74, 81, 82].
have the morphology of lobular carcinoma.
42. Are sentinel lymph node frozen sections or touch
preps reliable for lobular carcinoma?
41. Is sentinel lymph node mapping reliable for lobular
carcinoma? Given the generally small nuclei, small cells, and often dys-
cohesive pattern in metastatic nodes as described in question
Yes, sentinel lymph node mapping is commonly used in 40, intraoperative identification of metastatic lobular carci-
invasive lobular carcinoma [74, 81, 82]. In brief, (lympha- noma by frozen section or touch preparation can be difficult.
a b
c d
Fig. 4.24 Metastatic lobular carcinoma, scattered cells. (a) Dispersed discrete cell boundaries and darker nuclei as compared to histiocytes
tumor cells are occult at low power, but (b) are highlighted by keratin (d) in this section. This lymph node was previously frozen, introducing
stain of a matched field. (c) Carcinoma cells can be discerned at high artifact
power in the area identified on keratin stain; they are smaller with more
• Several groups have reported lower sensitivity of frozen reported equivalent sensitivity for ductal and lobular car-
sections for lobular carcinoma [83], while others have cinoma [85, 86].
reported comparable sensitivity [84]. • Although lobular carcinoma can be easily missed intraop-
• Frozen sections should be evaluated with the same criteria eratively, it is more important not to “overcall” small
as permanent sections. (See question 40.) Intraoperative collections of histiocytes or other constituents as lobular
keratin staining is not available in most centers. carcinoma.
• We and others have found invasive lobular carcinoma par- • In the current era, axillary lymph node dissections are
ticularly difficult to identify in nodal “touch preps” or not performed for a major subset of patients even with
“imprint” cytology. Some groups disfavor cytologic anal- low axillary metastatic burden, especially if patients will
ysis of lobular carcinoma and preferentially or addition- be receiving adjuvant endocrine or radiation therapy
ally perform frozen sections [84]. Sensitivity of [73, 77–79]. Therefore, it is important to let the operat-
intraoperative cytology has been published as 41–71% as ing surgeon know when only rare metastatic cells are
compared to permanent sections with immunohistochem- identified on intraoperative touch preparations or frozen
istry [85, 86]. Granted, some specialized centers have sections.
43. How is metastatic lobular carcinoma measured and –– Widely dispersed single cells can be difficult to iden-
staged in lymph nodes? tify in bone marrow biopsies and may require keratin
stains, especially in the background of cellular marrow
The lymph nodes in breast cancer are categorized based on (Fig. 4.26a–d) [2].
size/quantity of nodal tumor deposit regardless of histologic • Involves lung less frequently than ductal, no special type
subtype of the primary. Tumor deposits are classified as macro- [1, 2, 7].
metastasis, micrometastasis, and isolated tumor cells [30, 31]. • Metastatic disease may rarely be the first presentation of
lobular breast carcinoma.
• It is important to section lymph nodes at 0.2 cm (2 mm)
intervals along either the long or short axis to facilitate 45. Is there a difference in late recurrence between duc-
detection of all macrometastasis [30, 31, 73]. tal “no special type” and lobular carcinoma?
• Each lymph node is classified based on its largest contigu-
ous focus of carcinoma (cells touching one another). For stage-matched invasive lobular vs. ductal or no special
Discontinuous foci are not added together, and areas of type invasive breast cancers, classic lobular tends to have:
fibrosis in between tumor cells do not count.
–– Macrometastasis: Tumor deposit larger than 0.2 cm • Better 5-year disease free and overall survival [2, 7]
(2 mm, Fig. 4.21b). • Higher rates of later recurrence (10 years) and worse
–– Micrometastasis (mi): Tumor deposit larger than longer-term survival [2, 7, 9]
0.02 cm (0.2 mm) but not larger than 0.2 cm (2 mm,
Fig. 4.21e). 46. What immunostains are helpful to diagnose meta-
–– Isolated tumor cells (ITC): Small clusters of tumor static lobular carcinoma (Figs. 4.25a–f, and 4.27a–f)?
cells not larger than 0.02 cm (0.2 mm, Fig. 4.23a–d).
–– If only dispersed cells are present, as commonly occurs Metastatic lobular carcinoma is usually positive for the fol-
with lobular carcinoma, more than 200 dispersed cells lowing immunomarkers:
in lymph node cross section qualify as micrometastasis
and less than 200 cells as isolated tumor cells. • GATA3 (positive in 96% of primary lobular carcinomas)
Pathologist judgment is needed. [88–90].
• The size of the largest nodal metastasis should be reported, • Mammaglobin (positive in 70% of primary lobular carci-
along with any extranodal extension. nomas) [89].
• The numbers of nodes with macro- and micrometastasis • GCDFP-15, gross cystic disease fluid protein-15 (positive
are added to determine pN stage. Nodes with isolated in 50% of primary lobular carcinomas) [89, 91, 92].
tumor cells are not added to the total for stage determina- • Estrogen receptor (ER).
tion [30, 31]. • Progesterone receptor (PR).
• A simplified version of pathologic N categories is listed in • Cytokeratin 7 (CK7).
question 28; please refer to the AJCC manual or CAP • FOXA1 (forkhead box protein A1) and PELP1 (proline-,
staging checklists for full detail [30, 31]. glutamate and leucine-rich protein 1) are transcription
factors with emerging data in identification of breast can-
44. What are the most common metastatic sites of inva- cers [93–95].
sive lobular carcinoma?
None of these are entirely specific for breast cancer
Lobular carcinoma has a propensity for metastatic involve- including lobular carcinoma, and the immunostain panel
ment of: should be tailored to the differential diagnosis and location,
for example:
• Abdominal cavity, including gastrointestinal tract (espe-
cially gastric), peritoneum, ovaries, and uterus (endome- • Lung:
trium) [1, 2, 7, 9, 87] (Fig. 4.25a–f). –– TTF-1, napsin generally negative in breast carcinoma,
• Meninges [1, 2, 87]. with 2.4% of breast carcinomas TTF-1 positive [95, 96]
• Skin [1, 7], periorbital area [9]. –– GATA3 positive in about 10% of lung adenocarcino-
• Bone, liver (equal to or more so than ductal) [1, 2, 9]. mas [88, 95]
a b
c d
e f
Fig. 4.25 Metastatic lobular breast cancer to colon. The patient was estrogen receptor (d), indicating metastatic breast cancer (all 200×).
diagnosed with and treated for a 6 cm invasive lobular carcinoma with Lobular carcinoma metastasis in a colon biopsy from another patient is
five positive axillary lymph nodes, 8 years prior. (a) Lamina propria is much more subtle. (e) The only focus shown on H&E (arrows), keratin
filled with cells with intermediate nuclear grade and scant cytoplasm positivity (f) confirms carcinoma (both 200×)
arranged in files. Immunostains positive for keratin (b), GATA3 (c), and
a b
c d
Fig. 4.26 Lobular carcinoma metastatic to bone marrow. (a) Obvious In this cellular marrow, metastatic tumor cells are very subtle. (d)
tumor infiltrate replacing marrow elements. (b) Lobular carcinoma with Keratin staining highlights metastatic carcinoma in a matched field.
mucin vacuoles (black arrows) with nearby osteoclasts (red arrows). (c) Sometimes the carcinoma can only be appreciated with keratin stains
–– ER positive in 10–20% of lung carcinomas, usually correctly score other immunostains against the ER+
weak, focal [92, 97] cellular background.
• Gastric: distinguish from primary diffuse gastric cancer –– ER is often positive in both and is not useful.
(many cases E-cadherin negative; see question 47 [98]): –– PAX8 useful gynecologic marker [103].
–– ER strongly positive in lobular carcinoma, rarely –– WT-1 stains ovarian carcinomas, but also 1–8% of
weakly positive in gastric [92, 99]. breast cancers [97, 104, 105].
–– GATA3 positive in 5% of gastric cancers (unspecified –– Mammaglobin positive in up to 40% of endometrial
type) [88] and 0/32 signet ring gastric carcinomas in a and ovarian carcinomas, mandating caution [106–
tissue microarray study [100]. 108]; smaller proportions stain with GCDFP-15
–– Mammaglobin and GCDFP-15 negative in signet ring (5–10%) [92, 108].
gastric cancer in a tissue microarray study [101]. –– Gynecologic carcinoma commonly CK7+
–– CDH17 and CDX2 may be positive in gastric [98, (non-discriminator).
102], no specific gastric carcinoma markers. • Urothelial: metastatic lobular carcinoma must be distin-
–– Gastric carcinoma commonly CK7+ (non-discriminator). guished from plasmacytoid or signet ring urothelial carci-
• Gynecologic: metastatic lobular breast cancer can insinu- nomas (Fig. 4.27a–f).
ate in cellular ovarian or endometrial stroma and be very –– GATA3 is frequently positive in urothelial carcinomas
hard to detect. and is not useful in this differential, including signet
–– Keratin staining may be necessary to demonstrate the ring and plasmacytoid variants [100, 101].
population of diffuse single cells within stroma and ◦◦ GATA3 also stains paraganglia [109].
a b c
d e f
Fig. 4.27 Metastatic lobular carcinoma to the bladder. (a) Dyscohesive only focally, as seen here. (e) Likewise, estrogen receptor is strongly
cells infiltrating around bundles of muscularis propria. The differential diffusely positive in the carcinoma cells. (f) In this field from the same
diagnosis includes plasmacytoid urothelial carcinoma, metastatic dif- case, estrogen receptor staining is positive in chains of carcinoma cells
fuse gastric cancer, as well as lobular carcinoma. The overlying urothe- at the bottom of the field, yet it also moderately labels smaller stromal
lium was normal (not shown). (b) Infiltrating cells are keratin positive nuclei in the superficial lamina propria at top. Other immunostains
and E-cadherin negative. (c) These stains and GATA3 do not discrimi- include CD45 negative in tumor cells and negative progesterone recep-
nate bladder and metastatic breast carcinoma. (d) GCDFP-15 is posi- tor and Her2 (immunohistochemistry score 0, not shown)
tive, supporting breast carcinoma. GCDFP-15 often stains breast cancer
–– ER is infrequently positive in such urothelial carcino- • Hematolymphoid:

mas and is a discriminator [100, 101]. –– Keratin immunostaining essential in demonstrating an
◦◦ PR stains 0–15% urothelial carcinomas [100, 101]. occult epithelial population amidst hematolymphoid
–– Mammaglobin is negative in urothelial carcinoma and cells (Fig. 4.26a–d)
is recommended [101]. • Some typical lymphoid markers stain epithelial neo-
–– GCDFP-15 stains 0–25% of these variant urothelial plasms and vice versa:
carcinomas suggesting caution [100, 101]. –– CD138 (syndecan, plasma cell marker) also stains
–– Urothelial carcinoma commonly CK7+ many carcinomas, including at least 40–60% of breast
(non-discriminator). carcinomas [110, 111].
–– GATA3 is positive in T-cells and related neoplasms 49. What are the typical features of ALH/LCIS on breast
[88]. imaging (mammography, ultrasound, MRI) (Fig.
• Soft tissue tumors may occasionally mimic 4.28a–c)?
–– Rhabdomyosarcoma: the dyscohesive cells with
eccentric pink cytoplasm may occasionally mimic lob- • Most ALH/LCIS is occult on imaging [123], with lobular
ular carcinoma. Rhabdomyosarcoma may rarely neoplasia identified incidentally in biopsies performed for
metastasize to the breast; patient age group rarely separate adjacent lesions (e.g., calcifications on mam-
overlaps. mography seen in association with columnar cell change,
◦◦ Rhabdomyosarcoma stains with desmin, myogenin, with incidental LCIS without calcification).
myoD1, and PAX7 [112, 113]. • Between 2% and 35% of ALH/LCIS may have an imag-
◦◦ Rhabdomyosarcoma should not stain with keratin ing correlate [123, 124].
or typical breast markers described above. • LCIS may appear as mammographic calcifications
–– The rare epithelioid hemangioendothelioma (EHE) (Fig. 4.28a–c). These may be of variable morphology,
and sclerosing fibrosarcoma may appear as small dys- including amorphous, fine, and heterogeneous, and are
cohesive pale cells in myxohyaline or fibrotic stroma. often grouped. Studies differ in the frequency of coarse or
Few examples of EHE may be keratin positive. pleomorphic calcifications [14, 123–125].
• Ultrasound findings are described as typically irregular,
47. Can E-cadherin be used to diagnose metastatic lobu- hypoechoic, avascular mass lesions, with posterior acous-
lar carcinoma? tic shadowing [123]. However, oval, circumscribed
masses without posterior shadowing have also been
No, E-cadherin loss is not specific for lobular breast carci- described [126].
noma. Dyscohesive carcinomas originating in other organ • MRI findings include mass lesions or non-mass enhance-
systems are also E-cadherin negative, notably: ment, with variable kinetics [123, 127].
• Subset of diffuse gastric cancer [98, 99, 114]. 50. What are the typical features of pleomorphic LCIS on
• Plasmacytoid urothelial carcinoma [101, 115]. breast imaging (mammography, ultrasound, MRI)?
• Subset of various high-grade carcinomas, especially “sig-
net ring” cell carcinomas [98]. As described below, pleomorphic LCIS is of higher grade
• Metastatic melanoma and other non-epithelial malignan- than ALH or classic LCIS, and thus imaging features may
cies that can mimic carcinoma. share more similarity to high-grade DCIS [128].
• E-cadherin could be helpful as part of a panel of stains, to
help confirm metastatic lobular differentiation, if other • In one study, over 80% of identified pleomorphic LCIS had
stains demonstrate breast origin, as described above. calcifications on mammography (Fig. 4.28a–c) [125, 128].
• MRI findings again include mass lesion, non-mass
enhancement, or essentially negative MRI [128].
Lobular Neoplasia In Situ (ALH/LCIS)
51. What are the typical management steps after a diag-
48. What is the incidence of atypical lobular hyperpla- nosis of ALH/LCIS on core biopsy?
sia/lobular carcinoma in situ (ALH/LCIS)?
Management of LCIS/ALH discovered on core biopsy is
• According to the World Health Organization, lobular neo- controversial, especially in the current era advocating against
plasia is encountered in 0.5–4% of “otherwise benign” “overtreatment.”
breast biopsies. Other sources report core biopsy inci-
dence of 1–1.5% for ALH/LCIS combined [116–118]. • Careful imaging correlation is recommended. Upgrade
• In two large contemporary studies of reduction mammo- rates to carcinoma are much lower if LCIS/ALH is “inci-
plasty specimens, ALH was found in 3–5% of specimens, dental” and it was another sampled lesion that explains
with LCIS in 0.7–1%, including one pleomorphic LCIS the imaging finding, as discussed in question 52 [116,
[119, 120]. 117, 129].
• In a SEER database analysis from 2000 to 2009, the inci- • Conservative excision is often recommended for patients
dence of LCIS was calculated as 2–2.75/100,000 women with classic LCIS on biopsy and should still be recom-
[121], reflecting an increase over prior SEER intervals mended for those with pleomorphic or florid LCIS on
[122]. biopsy [116, 128–132].
Fig. 4.28 LCIS with a b

calcifications. Calcifications
were seen on screening
mammography. (a) Post-
biopsy mammogram shows
residual calcifications (circle),
air in the biopsy site (above
circle), and an omega-shaped
post-biopsy marker. (b)
Real-time breast ultrasound
shows an irregular hypoechoic
region or ducts with internal
calcifications (arrow) in the
region of the mammographic
mass. (c) Histologic sections c
demonstrate calcification
(upper right) immediately
adjacent to pleomorphic LCIS
(200×), confirmed with
negative E-cadherin stain (not
shown). Additional radiologic
and histologic images from
this case can be seen in
Case 4, Figs. 4.48 and 4.49
• The management of ALH is even more controversial, family history along with imaging and pathologic fea-
given even lower upgrade rates [116, 129, 130]. tures. The patient is an important participant in individu-
• A recent consensus guideline statement from the alized decision-making.
American Society of Breast Surgeons (ASBS), published • Close clinical and imaging follow-up is necessary for
online, includes: non-excised lesions [132].
–– “we no longer advocate routine excision of ALH or
LCIS when the radiological and pathological diagno- 52. How often is invasive carcinoma diagnosed after
ses are concordant, and no other lesions requiring exci- finding ALH/LCIS on core biopsy?
sion are present” [132].
–– Regarding non-classic LCIS variants “these lesions, Rates of upgrade to invasive carcinoma, DCIS, or pleo-
and pleomorphic LCIS, in particular, should be treated morphic LCIS vary widely in the literature. In general fac-
with complete surgical excision, similar to DCIS” tors that influence upgrade rates include [117, 121,
[132]. LCIS variants are discussed below in questions 134–137]:
60 and 61.
• The 2017 National Comprehensive Cancer Network • Patient risk factors (personal or family history of breast
(NCCN) recommendations for LCIS are similar, with cancer)
excision recommended for non-concordant imaging and • Extent of imaging findings
risk-reducing therapy for LCIS with concordant imaging • Extent of pathology findings
[133].
• Expected risk of upgrade and potential outcomes should A recent single-center study with comprehensive review
be discussed with the patient, factoring in personal and of the literature is summarized in Table 4.4 [117].
53. What is long-term risk of invasive cancer after a the ALH/LCIS (ipsilateral), with contralateral cancer still
diagnosis of ALH? LCIS? prevalent; see Table 4.5 [137, 138, 143–147].
For over a half century, the implications of lobular neoplasia 55. Does ALH/LCIS originate in lobules?
as either a direct precursor of cancer or as a marker of
increased cancer risk (so-called risk lesion) have been a mat- • ALH/LCIS, as well as many other forms of ductal and
ter of considerable debate [116, 129]. Current WHO consen- columnar in situ proliferation, are believed to originate
sus recognizes lobular neoplasia as both a risk factor and a from the terminal duct lobular unit [14].
non-obligate precursor of breast cancer [116]. • The most recognizable location of ALH/LCIS and its
nomenclature may lead to the incorrect assumption that
• Compared to patients without these lesions, the relative lobular neoplasia arises from breast lobules.
risk of subsequent cancer is 3–5-fold for ALH and 6–12-
fold for LCIS [116, 129, 136]. 56. What are the characteristic cytologic and architec-
• Patients with ALH/LCIS have a 1–2% annual incidence tural features of ALH/classic LCIS?
of breast cancer, which is cumulative, even over decades
[130, 137–139]. Cytologic features are as follows (Figs. 4.29a–d and 4.30a–
• LCIS shares point mutations as well as genomic gains and d) [116, 129–131, 134, 139, 148, 149]:
losses with a significant fraction of synchronous invasive
lobular carcinomas [116, 130, 140–142]. • Small dyscohesive cells.
• Round monotonous nuclei.
54. Is future risk of invasive carcinoma local (unilateral) • Inconspicuous nucleoli.
or bilateral? • Pale cytoplasm usually scant, round, cuboidal, or polygo-
nal cells.
• Early and well-cited studies recognized a bilateral risk of • Non-polarized cells, though nuclei may be eccentric.
breast cancer, in alignment with the “risk lesion” concept • Occasionally intracytoplasmic mucin vacuoles.
[17, 143]. • Similar to invasive lobular carcinoma.
• Lobular neoplasia itself has a propensity for bilaterality [130]. • Variants include clear cell change, apocrine cytoplasm,
• More recent larger and longer-term studies have docu- and (rhabdo)myoid cytology [16].
mented a bias toward cancer arising in the same breast as
Architectural features are as follows (Figs. 4.29a–d and
Table 4.4 Upgrade rates of ALH and LCIS on core biopsy 4.30a–d) [116, 129–131, 134, 139, 148, 149]:
ALH % LCIS %
Upgrade rate incidental incidental • Cells fill and/or expand acini
on excision ALH % LCIS % only only • May appear dyscohesive or loosely cohesive
Single center 9% 24% 4% 7% • Most often do not form true luminal structures, unless a
Meta- 9% 18% Not done Not done
mixed population
analysis, (0–67%) (0–60%)
mean (range) • May extend along ductal structures in a “pagetoid” fash-
Data from ref. [117] ion (see question 65)
Table 4.5 Cancer incidence after ALH/LCIS diagnosis, selected studies

Number of invasive cancers/number Ipsilateral Contralateral
Study Year published of in situ lobular lesions cancer cancer
Rosen [143] 1978 36/99 (LCIS) 53% 44%
Bodian [144] 1996 (updates Haagensen cohort) 55/216 54% 46%
Page [145] 2003 44/237 (ALH) 72% 23%
King [138] 2015 168/1060 (LCIS) 63% 25%
Wong [137] 2017 (SEER) 1837/ 19,462 55% 45%
ILC only 368/1837 69% 31%
Coopey [147] 2012 106/784 untreated 61% 39%
Numbers may not total 100% given categorization of bilateral carcinomas
ALH atypical lobular hyperplasia, ILC invasive lobular carcinoma, LCIS lobular carcinoma in situ
a b
c d
Fig. 4.29 Classic cytologic and architectural features of lobular neo- space (400×). (c) A small tortuous ductule involved by ALH. A few
plasia in situ (ALH/LCIS). (a) A uniform population of cells with remnant luminal (ductal) epithelial cells are marked by the red arrow
eccentric round monotonous nuclei fills these spaces. Not all of the cells (400×). (d) This focus of LCIS, from the same patient, has somewhat
appear dyscohesive; cell-cell membranes are sometimes apparent. (b) larger more hyperchromatic nuclei
Two small ductal spaces filled by LCIS. Cells appear dyscohesive in top
a b c
Fig. 4.30 Lobular neoplasia (ALH) partially involving a terminal E-cadherin immunostaining is negative in lobular cells within acini at
duct lobular unit. (a) The ducts and acini at the top of the field are colo- the top of the field, with labeling of a few remnant ductal cells; myo-
nized and mildly expanded by lobular cells. Acini at bottom are unin- epithelial cells also stain such that E-cadherin outlines many of the
volved and appear smaller and less full (200×). (b) ALH involves most acini. E-cadherin shows membranous staining in normal cells at
of the acini in the field; red arrows show uninvolved acini. (c) bottom
57. How are ALH and LCIS distinguished from one involvement is not sufficient for a diagnosis of LCIS in
another? our opinion and that of other experts [130, 134].
Both ALH and LCIS are characterized by small uniform 58. What are the minimal diagnostic criteria for ALH?
lesional cells that show lobular phenotype and loss of cohe-
sion. They are defined and discriminated by extent • Based on the description of Page and Dupont, minimal
(Fig. 4.30a–d). The classic studies of Page and colleagues criteria for ALH include “clear round cells reminiscent of
used the following definition: ALH and LCIS are present within lobular acini” with one
of the following [150]:
• LCIS: “At least half of the acini in the lobular unit com- –– Acinar distention
pletely filled, distorted, and distended with characteristic –– Acinar distortion
cells….such that no intercellular lumina existed” [134]. –– Increase in number of cells
–– Various criteria for “distention” include acini with at • There is no entity of “lobular hyperplasia,” and this termi-
least eight cells across [14]. nology should not be used.
–– Other criteria include acinar diameter larger than
nearby uninvolved acini [129]. 59. What are type A and B cells?
• ALH: less than half of the acini in a lobule are filled and
distended. As per Page: “lesions that meet most of the In the 1970s, the Columbia group recognized subtle varia-
preceding criteria but fail to meet one of the criteria in tions in “classic” lobular neoplasia and categorized them
over 50% of the acini within a lobular unit” [134]. as follows (Fig. 4.31a–c). Although this group soon found
• There are no analogous criteria for lobular neoplasia the type A/B designation to be overly simplistic, and it
involving ductal structures. The presence of ductal remains of no known significance, it has been carried
Fig. 4.31 LCIS and normal a

breast. (a) H&E shows LCIS
at right and a terminal duct
lobular unit at left (100×). (b)
Dual staining for E-cadherin
(brown) and the myoepithelial
marker smooth muscle
myosin heavy chain (red) of a
section deeper in the same
block. In the normal duct
(left), luminal ductal cells
have E-cadherin membrane
staining, with underlying
myoepithelium in red. In the
LCIS (right), attenuated
ductal cells are highlighted in
brown, while smooth muscle
myosin heavy chain (red)
outlines the distended acini b
through into many textbooks. A parallel nomenclature has –– Large nuclei (>4 lymphocytes), pleomorphic with
been applied to invasive lobular carcinoma [14, 15, 17, 18, prominent nucleoli, nuclear membranes may be irreg-
116, 130, 148]: ular (grade 3 nuclei).
◦◦ The cell to cell variation and degree of nuclear
• Type A: small cells with small uniform nuclei, akin to membrane irregularity is generally less than high-
grade 1 nuclei. grade DCIS [20].
• Type B: slightly larger cells with more variation in nuclear –– Dyscohesive large cells with eccentric nuclei, often
and cell size (larger) and shape, often with paler chroma- with abundant cytoplasm.
tin and/or small nucleoli, more like grade 2 nuclei (twice –– Increased mitotic rate [130].
the size of a lymphocyte nuclei). –– Comedonecrosis, central calcification, and massive
• Both are considered “classic” ALH/LCIS. acinar distention may be seen.
–– No true intercellular lumens.
60. What are the morphologic variants of LCIS? –– May mimic DCIS (see question 69).
–– Variants include (usually pleomorphic):
• Pleomorphic LCIS: Lobular features in larger cells with ◦◦ Apocrine
high-grade nuclei (Figs. 4.31a–c and 4.32a–f) [18, 116, ◦◦ Signet ring
139, 149, 151, 152]: ◦◦ Histiocytoid
a b
c d
Fig. 4.32 ALH compared to LCIS. (a) These acini are colonized by lobular cells, but hardly distended (200×). (b) In LCIS, acini are filled and
markedly expanded by a similar population (200×). (c) ALH at higher power (400×). (d) LCIS at high power (400×)
• Florid LCIS: “Classic” LCIS cytology with (Fig. 4.32a–f) • The management of surgical margins in the setting of
[18, 116, 129, 149]: pleomorphic LCIS is more controversial and is discussed
–– Massive acinar or ductal distention, usually with little in question 82.
or no stroma between markedly distended acini • An explanatory comment in the pathology report is rec-
–– Central necrosis and associated calcifications ommended to explain the clinical significance of these
• Variants make up about 2–5% of LCIS [152]. variants.
61. Why is it important to recognize pleomorphic and 62. Can LCIS have comedonecrosis and calcifications
florid LCIS? (Figs. 4.28a–c and 4.32a–f)?
Although outcome data remain relatively sparse, pleomor- Classic LCIS is very rarely associated with comedonecrosis
phic and florid LCIS have cytologic, architectural, immuno- and calcifications. However, florid and pleomorphic LCIS
phenotypic, molecular, and clinical features of more often do have comedonecrosis and/or calcifications and may
aggressive in situ neoplasia as compared to classic LCIS mimic DCIS on mammography [18, 139, 149].
[116, 129, 151–156]. Pleomorphic and florid LCIS were
likely diagnosed DCIS in past decades and thus treated as 63. Is LCIS graded?
such; this management likely remains relevant despite the
reclassification. Grading of LCIS is not required by College of American
Pathologists (CAP) or American Joint Commission on
• Selected data from an excellent study summarizing Cancer (AJCC). However, it is important to recognize and
immunophenotypic and molecular features (comparative report LCIS with high-grade nuclei or aggressive features
genomic hybridization) are listed in Table 4.6 [155]. (pleomorphic LCIS, florid LCIS). (See questions 60–62.)
• Pleomorphic and florid LCIS on core biopsy is fre-
quently associated with invasive carcinoma and/or 64. How does ALH/LCIS translate to LIN nomenclature?
DCIS at excision [156]. Thus, excision is recom-
mended for pleomorphic, florid, or mass-forming in LIN designates “Lobular Intraepithelial Neoplasia” a nomen-
situ lobular lesions identified on core biopsy [116, clature championed by Tavassoli, Eusebi, and others. The
129–131, 139, 149]. LIN scheme eliminates the name “carcinoma” and acknowl-
–– Recent reviews of the literature calculated a 30% rate edges the qualitative and quantitative spectrum of such lobu-
for pleomorphic LCIS on core upgrading to invasive lar proliferations. This nomenclature has not been widely
carcinoma on excision and 8% upgrade to DCIS on adopted in the USA, but is more broadly applied in Europe.
excision [156, 157]. As defined by Tavassoli and colleagues [131, 160]:
–– Florid LCIS has been associated with concomitant
invasive lobular carcinoma in up to 70% of cases, • LIN1: “partial or complete replacement…by a prolifera-
based on small studies [116, 158, 159]. tion of generally uniform cells with poorly defined margins
which may fill, but do not distend, the acinar lumens (in
comparison to adjacent uninvolved acini).” Often loosely
cohesive proliferation, but with residual acinar lumens.
Table 4.6 Immunophenotypic and molecular features of pleomorphic –– Similar to ALH
lobular carcinoma
• LIN2: “more abundant proliferation of similar cells than
LCIS, LCIS, in LIN1, which fill and actually distend some or all acini,
LCIS, LCIS, pleomorphic pleomorphic
Parameter classic florid apocrine non-apocrine
but the acinar outlines remain distinct….with the persis-
Estrogen receptor 100% 92% 23% 100% tence of intervening lobular stroma; residual lumens may
+ persist in some acini.”
Her2+ 0% 18% 31% 0% –– Similar to LCIS
Ki-67 4.2% Not 13.9% 9.9% • LIN3: proliferation of cells similar to LIN1–2, “but there
done is a massive degree of acinar distension to the point that
Fraction genome 0.072 0.109 0.119 0.054
the acini appear almost confluent.” “When there is necro-
altered
Breakpoints 3.8 11.55 3.15 5.86
sis or the proliferating cells are completely of the pleo-
Amplifications, # 0.25 2.1 5.00 0.077 morphic or signet ring cell type, significant acinar
Data from Ref. [155] distension is not required.”
LCIS lobular carcinoma in situ –– Similar to florid and/or pleomorphic LCIS
65. How does LCIS/ALH involve ducts (Fig. 4.33a–e)? 66. Does Paget’s disease of the nipple arise in association
with LCIS?
Lobular neoplasia usually involves ducts in a so-called “pag-
etoid” fashion. LCIS rarely colonizes nipple ducts; in those cases, it could
theoretically extend to the squamous epithelium of the epi-
• LCIS/ALH undermines the ductal epithelium, colonizing dermal surface [163]. However, this is extremely rare and
the duct with layer(s) of neoplastic cells between the duc- should prompt further diagnostic confirmation (LCIS vs.
tal (luminal) epithelium and myoepithelial cells [14, 129, DCIS, Toker cells, melanocytes, etc.).
130, 139, 149, 150].
• The presence of lobular cells may alter the contour of the 67. How can pagetoid ductal involvement by ALH/LCIS
duct, resulting in characteristic architecture described as be distinguished from prominent myoepithelium?
budding, sawtooth, or cloverleaf [14, 129, 130, 139,
149, 150]. • Nuclei of myoepithelial cells are generally smaller and
–– The basement membrane remains intact. darker than ALH/LCIS [129].
–– The myoepithelial cell layer is generally continuous, • Epithelioid myoepithelial cells are most often present as a
although myoepithelial cells may be seen admixed single layer, whereas lobular neoplasia may be multilay-
with lobular cells [129, 130, 161, 162]. ered [129].
• More abundant LCIS/ALH may fill a small duct, resulting • Immunohistochemistry may be helpful in perplexing foci.
in a solid pattern and must be distinguished from low- (See questions 23 and 71.)
grade DCIS [130]. –– Myoepithelial cells:
• Duct extension of LCIS/ALH may even involve larger or ◦◦ Positive for p63 (nuclear). Usually positive for
lactiferous ducts. smooth muscle myosin heavy chain, calponin,
a b c
Fig. 4.33 Cytologic heterogeneity of lobular neoplasia in situ. (a) 400×). (c) Pleomorphic LCIS. Large dyscohesive cells with eccentric
Small dyscohesive cells with eccentric small nuclei with fine chromatin cytoplasm, large vesicular irregular nuclei. Note focal necrosis (central)
(“Type A,” 400×). (b) Larger dyscohesive cells with moderate amounts and a mitotic figure, lower (400×)
of pale cytoplasm, larger vesicular nuclei, and small nucleoli (“Type B,”
although the myoid staining may be diminished if • Pleomorphic LCIS with comedonecrosis and calcifica-
epithelioid [129]. Also positive for CD10, smooth tion may closely mimic high-grade DCIS, and the pos-
muscle actin (SMA), often D2–40. sibility of LCIS is important to consider. Immunostains
◦◦ Positive for E-cadherin, membranous p120 can be especially helpful with this differential. (See
–– ALH/LCIS: Table 4.3 and questions 23 and 71 for stain details and
◦◦ Negative for p63 (nuclear), smooth muscle myosin pitfalls.)
heavy chain, calponin –– Pleomorphic LCIS is E-cadherin negative (beta-
◦◦ Negative for E-cadherin, cytoplasmic granular p120 catenin negative, p120 cytoplasmic).
–– DCIS is E-cadherin positive in a membranous pattern
68. How is ALH/LCIS distinguished from low-grade (beta-catenin and p120 also membranous).
DCIS?
70. What molecular mechanisms are responsible for the
Distinguishing features include (Fig. 4.34a–h) [18, 130]: characteristic features of ALH/LCIS?
• ALH/LCIS lacks intraductal architecture such as cribri- Loss or dysfunction of the cell-cell adhesion molecule
form or punched out spaces (secondary lumens). E-cadherin is apparent in ALH and LCIS, as well as invasive
–– Partial involvement or LCIS “colonizing” usual hyper- lobular carcinoma. See Fig. 4.36a–e. E-cadherin function is
plasia, adenosis, or papillary lesions may leave resid- discussed in question 20 [14, 18, 32–40, 116, 129–131, 139,
ual architecture. 148, 149, 164, 165].
• ALH/LCIS is composed of non-polarized, dyscohesive
cells, in contrast to low-grade DCIS; DCIS has more dis- 71. How is E-cadherin and catenin immunohistochemis-
tinct cell membranes and cells that polarize around try utilized to discriminate lobular from ductal
neo-lumens. differentiation?
• ALH/LCIS more commonly demonstrates intracytoplas-
mic mucin vacuoles (although DCIS can as well). ALH/LCIS, like invasive lobular carcinoma, is character-
• ALH/LCIS may show pagetoid ductal involvement ized by loss of functional E-cadherin, and E-cadherin
(although DCIS can as well; see question 65). immunohistochemistry is a helpful ancillary tool in differ-
• Solid pattern low-grade in situ carcinoma can be particu- entiating in situ ductal and lobular neoplasia (Table 4.3,
larly challenging to determine if ductal or lobular. Subtle Figs. 4.12, 4.34a–h, 4.37a–c, and 4.38a, b) [14, 18, 32–40,
microacini and crisp/distinct cell borders favor a ductal 116, 129–131, 139, 148, 149, 164, 165]. Stain interpreta-
process. Immunostains can be helpful in distinguishing tion and caveats are thoroughly discussed in question 23.
lobular from ductal. (See questions 23 and 71 for stain Additional features unique to the in situ situation are
details and pitfalls.) below:
–– ALH/LCIS is E-cadherin negative (beta-catenin nega-
tive, p120 cytoplasmic). • In situ proliferations have surrounding or admixed myo-
–– DCIS is E-cadherin positive in a membranous pattern epithelial cells; these have intact membranous E-cadherin
(beta-catenin and p120 also membranous). and catenin and should not be misconstrued as lesional
cell staining [162].
69. How is pleomorphic LCIS distinguished from high- • A small population of ductal cells may be present together
grade DCIS? with a lobular proliferation; the ductal population would
be E-cadherin and catenin positive (membranous) and
Distinguishing features include (Fig. 4.35a–d) [130]: should be separately scored (Figs. 4.33a–e, 4.37a–c, and
4.38a, b).
• Pleomorphic LCIS lacks intraductal architecture such as
cribriform or punched out spaces (secondary lumens). 72. Must E-cadherin immunostaining be performed in
• Pleomorphic LCIS is composed of non-polarized, high- order to render a diagnosis of ALH/LCIS?
grade dyscohesive cells.
• Pleomorphic LCIS more commonly demonstrates intra- No, it is not essential to demonstrate loss of E-cadherin in
cytoplasmic mucin vacuoles. diagnosing lobular neoplasia [57]. The diagnosis can be
• Pleomorphic LCIS may be accompanied by classic ALH/ made based on typical architectural and cytologic features
LCIS and show pagetoid ductal involvement. [14, 40, 57, 116, 129–131, 139, 149].
a b
c d
e f
Fig. 4.34 Florid and pleomorphic LCIS. (a) Ductal structure floridly (d) Pleomorphic LCIS involves the right space; notice the degree of
distended by pleomorphic LCIS with central comedonecrosis and large nuclear enlargement and hyperchromasia compared to the LCIS in the
calcification (red arrow), E-cadherin negative (not shown, 100×). (b) center of the left space (400×). (e) This pleomorphic LCIS has associ-
Large and small calcifications (red arrows) associated with pleomor- ated microcalcifications (arrows, 200×). (f) Pleomorphic LCIS with
phic LCIS without necrosis. Some cells have intracytoplasmic lumens markedly enlarged nuclei, which remain somewhat round and uniform,
and signet ring nuclei (examples, black arrows, 200×). (c) Florid LCIS and abundant eosinophilic cytoplasm, possibly apocrine (400×)
with massive acinar distension, so as to be almost confluent (100×).
a b
c d
Fig. 4.35 Lobular neoplasia involving ducts. (a) Small duct colo- attenuated ductal cells (red arrows, examples, 400×). (d) Small duct
nized by ALH with “sawtooth” architecture (400×); typical lobular colonized by lobular neoplasia (400×). (e) Immunostained serial
cells are apparent, with some residual lumen (400×). (b) Larger duct section demonstrating E-cadherin-negative lobular cells between
with blunt buds due to involvement by ALH (200×). Neoplastic lob- E-cadherin-positive luminal ductal cells (brown stain) and smooth
ular cells insinuate between luminal ductal cells and myoepithe- muscle myosin heavy chain-positive myoepithelial cells (red stain;
lium/basement membrane. (c) In this small duct, the proliferation of vague brown E-cadherin positivity is also seen in myoepithelial
lobular cells protrudes inward, and each nodule is lined by a layer of cells, 400×)
73. Are there recurrent genetic changes in ALH/LCIS 74. Can ALH/LCIS and DCIS occur together?
(mutations, chromosomal gains/losses)?
Lobular neoplasia and DCIS can be seen within the same
Like invasive lobular carcinoma, LCIS very frequently con- specimen and may even coexist within the same acinar or
tains CDH1 (E-cadherin), as well as PIK3CA mutations [9, ductal space. This is seen as two discrete populations
14, 48, 116, 129–131, 139, 149, 166, 167]. (Fig. 4.39a–h) [40, 116]:
• ALH/LCIS are early lesions in the “low-grade” molecular • One with the cytologic and architectural features of ALH/
progression pathway, characterized by genomic changes LCIS.
including [9, 48, 140, 166–169]: • Another with different cytologic and/or architectural fea-
–– Loss of 16q (location of CDH1 gene) tures of DCIS (of any grade).
–– Gain of 1q • Immunohistochemistry can be helpful to highlight the
• LCIS shares point mutations as well as genomic gains and two distinct populations (Table 4.3).
losses with a significant fraction of synchronous invasive –– Lobular component: E-cadherin negative (beta-catenin
lobular carcinoma and DCIS, though less so with invasive negative, p120 cytoplasmic)
ductal carcinoma [116, 140–142, 170].
Lobular carcinoma in situ Ductal carcinoma in situ
a b
H&E
c d
E-cadherin
e f
P120 catenin
g h
Beta-catenin
Fig. 4.36 Morphology and immunophenotype of classic LCIS and membranous accentuation. Right column: DCIS in serial sections.
DCIS. Left column: LCIS with subtle residual spaces (arrows); LCIS is H&E (b) shows low-grade cells with some cytologic resemblance, but
usually solid. Serial sections stained with (a) H&E, (c) E-cadherin, (e) different architecture than LCIS, including multiple round sharply
p120 catenin, (g) beta-catenin (all 400x). In typical lobular carcinomas, punched out spaces. There is cell membrane localization of (d)
E-cadherin and beta-catenin are negative, while p120 catenin demon- E-cadherin, (f) p120 catenin, (h) beta-catenin. Also see diagrammatic
strates cytoplasmic staining, sometimes granular in texture, without Fig. 4.12
a b
c d
Fig. 4.37 Pleomorphic LCIS compared to DCIS. (a) Low-power view of a differential of high grade DCIS; E-cadherin was negative (not shown),
pleomorphic LCIS with ductal spaces markedly distended by a solid pro- confirming pleomorphic LCIS (200×). (c) Cribriform DCIS with comedo
liferation of malignant cells with comedonecrosis (100×). (b) Higher necrosis; this architecture would not be seen in LCIS (100×). (d) High
power of the left space shows high grade cells with vesicular eccentric grade DCIS with comedo necrosis has a rare intraductal lumen (red
nuclei and dyscohesion. The morphology favors pleomorphic LCIS with arrow) and different cytology as compared to pleomorphic LCIS (200×)
–– Ductal component: E-cadherin positive, membranous chemical features, a case cannot be definitively classified as
(beta-catenin and p120 membranous; see questions 23 DCIS or LCIS” [116].
and 71)
• It is prudent to make separate diagnoses of DCIS and • Recommendations for management of margins in this
LCIS (or ALH) with an explanatory comment. Alternative scenario are not provided.
nomenclature such as “mixed ductal and lobular lesion” • This is a very rare scenario, and additional data are
could be considered [116]. needed. It is unclear whether the above statement is
• Treatment should follow that of DCIS (unless pleomor- widely accepted, and this concept will likely evolve.
phic LCIS is the higher-grade lesion). • Bratthauer and colleagues have suggested the nomencla-
ture “Mammary Intraepithelial Neoplasia” (MIN) for this
75. How is “carcinoma in situ with mixed ductal and circumstance [160].
lobular features” defined? • This nomenclature is different than naming of intra-
ductal proliferations that contain discrete areas of
The 2012 WHO monograph suggests a diagnosis of “carci- E-cadherin-negative cytologically lobular cells and
noma in situ with mixed ductal and lobular features” when E-cadherin-positive cytologically ductal cells. (See
“after careful assessment of morphologic and immunohisto- question 74.)
a b c
d e
Fig. 4.38 Aberrant E-cadherin with cytoplasmic localization in LCIS. cytoplasmic E-cadherin stain. Remnant ductal epithelium lining the
(a) H&E sections demonstrate an in situ proliferation. (b) E-cadherin is lumen has a subtlety different cell membrane staining (arrow). (e) A
strongly positive in the CIS, but is not localized to the cell membranes. different E-cadherin antibody is negative in the lobular proliferation of
Instead, it is evenly localized through the cytoplasm. (c) Cytoplasmic the same case, with the remnant ductal cells showing obvious mem-
p120 localization confirms E-cadherin dysfunction and lobular differ- brane labeling
entiation (LCIS). (d) Another case of in situ lobular neoplasia with
Fig. 4.39 LCIS and DCIS together. (a) H&E shows intermediate- with serpiginous DCIS in a large duct, nodules of pale LCIS are seen at
grade DCIS in the top 2 spaces, with cribriform architecture. The space top and left (200×). (e) DCIS is E-cadherin positive (membranous),
at bottom left contains a cytologically and architecturally different while LCIS is negative (200×). (f) Beta-catenin shows the same stain-
population of dyscohesive cells with paler cytoplasm; focal DCIS is ing pattern as E-cadherin (200×). p120 was cytoplasmic in the LCIS,
seen at the periphery (200×). (b) E-cadherin staining is positive (mem- though weak (not shown). (g) In a different case, highly pleomorphic
branous) in DCIS and completely negative in LCIS (200×); beta-catenin squamoid DCIS is central and surrounded by pleomorphic LCIS char-
had the same staining pattern (not shown). (c) p120 catenin shows acterized by dyscohesive cells with scant cytoplasm (200×). (h) In a
membranous positivity in DCIS, but has cytoplasmic localization (non- different field, E-cadherin stains the central DCIS and is negative in
membranous) in LCIS (200×). (d) Another field from the same case peripheral LCIS (200×)
a b c
d e f
g h
76. Can LCIS/ALH occur with invasive ductal carci- may be helpful in ruling out invasive carcinoma if LCIS/
noma? DCIS with invasive lobular carcinoma? ALH involves sclerosing lesions such as sclerosing adenosis
or radial scar [14, 40, 116, 129].
An estimated 15–25% of DCIS are associated with invasive
lobular carcinoma, and likewise, LCIS may be associated with 78. Can LCIS/ALH involve collagenous spherulosis?
invasive ductal carcinoma (Fig. 4.40a–e) [116, 131, 170, 171].
Low-grade, estrogen receptor-positive DCIS and invasive ductal Lobular neoplasia has some propensity for involvement of
lesions tend to share genomic changes with synchronous lobular collagenous spherulosis (Fig. 4.41a–d) [40, 173, 174]. LCIS
processes [140, 171, 172], whereas high-grade ductal compo- was identified in 25% of collagenous spherulosis in one
nents are less likely to have common genomic alterations. series [173]. As such, it can be a difficult mimic of DCIS,
given the cribriform-like spaces [116, 129, 173, 174]. Unlike
77. Can LCIS/ALH involve proliferative breast lesions DCIS, collagenous spherulosis with LCIS:
such as sclerosing adenosis, papillary lesions?
• Contains eosinophilic basement membrane material
Lobular neoplasia can coexist with other breast lesions, within the “punched out” spaces.
seemingly “colonizing” the epithelium (Fig. 4.41a–d). The • Contains a mixed population of myoepithelial cells, duc-
same diagnostic criteria are applied. Myoepithelial stains tal epithelial cells, and luminal epithelial cells, with the
a b
c d e
Fig. 4.40 Ductal and lobular lesions in the same specimen. (a) Invasive shows invasive lobular carcinoma (black arrows) and DCIS (left). (c)
lobular carcinoma (right, black arrows) and low-grade DCIS involving a Invasive ductal carcinoma and ALH (black arrow, 40×). (d) Higher power
sclerosing lesion (right, 100×). (b) Higher power of another field also of invasive ductal carcinoma (200×). (e) Higher power of ALH (400×)
a b
c d
Fig. 4.41 LCIS in involving other proliferative lesions. (a) Lobular Immunostains would be prudent to rule out an invasive component (not
neoplasia colonizes a papillary lesion. The pale cells underlying the shown, 200×). (d) LCIS involving collagenous spherulosis. The
columnar epithelium are lobular (200×). (b) LCIS involving adenosis. punched out spaces are due to the underlying architecture of collage-
The border is circumscribed and slit-like architecture with remnant duc- nous spherulosis; the basement membrane component is not apparent in
tal epithelium is seen at left (100×). (c) A higher-power view of another this section. LCIS with cytoplasmic mucin is seen throughout, mixed
area from the same lesion with pale lobular cells throughout. with the myoepithelial component (400×)
myoepithelial cells surrounding the spaces; these can be three examples within phyllodes tumors and along with other
highlighted by immunohistochemistry (p63, calponin, cases of concurrent phyllodes and nearby LCIS [176].
smooth muscle myosin heavy chain) [129].
• Contains an E-cadherin-negative (p120 cytoplasmic) pop- 80. Is ALH/LCIS typically estrogen receptor positive?
ulation [129]. Care should be taken in stain interpretation,
as this will be a mixed population. Classic ALH/LCIS is almost invariably estrogen receptor
• May contain microcalcifications, in up to 40% [174]. positive [116, 149, 155, 177]. Florid or pleomorphic variants
• May rarely occur with pleomorphic LCIS [174]. of LCIS have lower rates of estrogen receptor positivity (see
• Collagenous spherulosis is an incidental finding involving question 61, Fig. 4.19a–d, Case 6) [129, 139, 149, 155, 177].
one or few ductal spaces [173, 174].
81. How should surgical margins be reported for classic
79. Can ALH/LCIS involve fibroepithelial lesions? ALH? LCIS? Pleomorphic LCIS?
Lobular neoplasia may involve the epithelial component of a In general, the relationship of classic ALH and classic LCIS
fibroepithelial lesion [129]. A single-center series docu- to surgical margin does not need to be reported [116]. Given
mented 18 cases of ALH/LCIS in fibroadenomas (17) or the higher-grade, more aggressive features of pleomorphic
phyllodes tumors (1) [175]. Sin and colleagues reported LCIS, and management considerations, documentation of
margin status is recommended for pleomorphic LCIS, florid Table 4.7 Anti-hormonal therapy decreases risk of subsequent cancer
in patients with ALH/LCIS, selected data
LCIS, or LCIS with comedonecrosis at margin [116, 139].
Selected ALH/ Subsequent cancer no Subsequent cancer with
LCIS studies chemoprevention chemoprevention
82. What typically defines acceptable surgical margins
Fisher 56% risk reduction
for classic ALH? LCIS? (NSABP-P1) tamoxifen
Pleomorphic LCIS? King (1980– 7% at 5 years 3% at 5 years
2009 MSKCC) 21% at 10 years 12% at 10 years
Traditionally, classic lobular neoplasia at surgical margin did Cuzick (IBIS-I LCIS: 28% 15-year LCIS: 27% tamoxifen
not warrant re-excision. This practice has been reaffirmed by trial) risk 15-year risk
LCIS: 17% 10-year LCIS: 5% AI 10-year
a recent evidence-based expert consensus guideline: risk risk
Cuzick (IBIS-II LCIS: 17% 10-year LCIS: 5% AI 10-year
• “Classic lobular carcinoma in situ (LCIS) at the margin is trial) risk risk
not an indication for re-excision” [61]. ALH: 14% 10-year ALH: 5% AI 10-year
risk risk
However, given the biologic features of pleomorphic Data from Ref. [181]
AI aromatase inhibitor, ALH atypical lobular hyperplasia, LCIS lobular
LCIS discussed previously (see question 61), there is greater carcinoma in situ
concern for recurrence and subsequent invasive cancer after
excision. Published rates vary between 0% and 50%, averag-
ing 10% in a meta-analysis with widely variable length of 84. Is radiation therapy indicated for classic LCIS/
follow-up, endocrine therapy, and margin status [152, 156, ALH? Pleomorphic LCIS?
178]. The management of margins remains uncertain if not
controversial. Radiation therapy is not currently recommended for diagno-
sis of classic LCIS/ALH (Fig. 4.42a, b) [129, 131, 133].
• The 2014 guideline document refrained from concrete While some patients with pleomorphic LCIS receive radia-
recommendation, awaiting additional data [61]. tion therapy [151, 152], there are no data to support this
• Some authors advocate for treatment similar to DCIS, in practice [NCCN]. Management considerations for LCIS ver-
other words excision to clear or 2 mm margins [157, 179]. sus DCIS and invasive carcinomas are summarized in
• The 2017 version of the NCCN guidelines stated in a Table 4.8.
footnote “Clinicians may consider complete excision
with negative margins for pleomorphic LCIS, but this 85. Does ALH/LCIS occur in men?
may lead to high mastectomy rate without proven clinical
benefit.” Yes, although rare, ALH/LCIS has been reported in men. A
• The WHO monograph and other groups advocate caution recent SEER study documented 16 cases of male LCIS, with
in recommending universal excision to negative margins 100% 5- and 10-year cancer-specific survival [72].
for pleomorphic LCIS [116, 156]. Multifactoral analysis
of extent of surgery anticipated, alternate risk manage-
ment (hormonal therapy if ER positive), pathology review
is recommended in such decisions [116]. Case Presentations
• WHO 2012 recommends against re-excising margins in
cases that contain pleomorphic LCIS, but have only clas- Case 1
sic LCIS at margin [116]. • History: A 76-year-old woman with history of left inva-
sive breast cancer treated with lumpectomy and radiation
83. Is adjuvant therapy indicated for classic LCIS/ALH? therapy 17 years previously, left excisional biopsy of
Pleomorphic LCIS? encapsulated papillary carcinoma/DCIS 5 years previ-
ously. She has a new right breast asymmetry on
Anti-hormonal agents (tamoxifen, raloxifene, and aromatase mammography.
inhibitors) have been shown to decrease the risk of subsequent • Imaging: See Fig. 4.43a–f. There was a 5 mm asymmetry
breast cancer in patients with LCIS and atypical hyperplasia, in the outer right breast on the craniocaudal mammo-
lesions which are typically estrogen receptor positive (Table 4.7) graphic view only. On craniocaudal tomosynthesis the
[129, 147, 180–182]. Small case series have documented the mass was more apparent. Breast ultrasound demonstrated
use of such agents in patients with pleomorphic LCIS, if estro- 9 mm irregular hypoechoic mass with size and location
gen receptor positive, but there are no outcome studies [128, corresponding to mammographic finding. US-guided
151, 152, 178]. Cytotoxic chemotherapy is not indicated. core biopsy with marker placement was performed.
a b
Fig. 4.42 Post-chemotherapy LCIS. (a) A ductal structure with paget- lar cells (black arrows), with membrane staining of the ductal popula-
oid involvement by LCIS, without discernable treatment effect. (b) tion (200×)
E-cadherin staining of a similar structure is negative in the buds of lobu-
Table 4.8 Management considerations for lobular as compared to ductal in situ and invasive lesions
Management after core biopsy Management of positive Radiation therapy in breast Adjuvant hormonal therapy
Lesion diagnosis surgical margin conservation (if ER+)
Classic LCIS Often excise (correlate) LCIS at margin OK No Consider
Pleomorphic Excise Consider re-excision Not standard Consider
LCIS
DCIS Excise Re-excise (typically to 2 mm Standard Standard
margins)
Invasive Excise (consider MRI for Re-excise (no cancer at ink) Standard Standard
lobular extent)
Invasive ductal Excise Re-excise (no cancer at ink) Standard Standard
Clinical, radiologic, and pathologic correlation is recommended in individualizing management
• Histologic Findings: See Figs. 4.6a–i and 4.44a–f. –– Keratin is the best diagnostic marker if immunostain-
Multiple tissue cores each demonstrated a sclerotic area ing is needed for diagnosis.
of varying cellularity. High-power examination showed –– Estrogen receptor is generally positive in classic infil-
small epithelioid cells with scant cytoplasm. trating lobular carcinoma, but also stains a subset of
• Differential Diagnosis: Invasive ductal carcinoma, inva- stromal cells and lymphocytes and is not specific.
sive lobular carcinoma, scar with inflammation, scar with
plump stromal cells Case 2
• Ancillary Studies: See Fig. 4.44a–f. The small cells were • History: A 59-year-old woman with family history of
keratin positive, E-cadherin negative, and estrogen recep- breast cancer (3 of 4 sisters in their 50s, both grandmoth-
tor positive (3+ >95%), although not necessary for diag- ers), with breast mass and separate area of skin dimpling.
nosis. The remainder of the invasive breast cancer Genetic testing was negative on a panel of multiple breast
prognostic panel showed PR positive (3+ 80%), Ki-67 cancer-associated genes.
5–10%, and Her2/neu negative (0) by immunohistochem- • Imaging: In the left breast at the 6:00 position, 6 cm from
istry, non-amplified by FISH. the nipple, there is a 1.7 cm irregular mass, overlying the
• Final Diagnosis: Invasive lobular carcinoma, grade 1 skin dimpling, with prior core biopsy diagnosed as inva-
• Take-Home Messages: sive lobular carcinoma. At the area of mass, 3:00 6 cm
–– Invasive lobular carcinoma can be relatively occult on from the nipple, a prior core biopsy demonstrated
imaging. LCIS. MRI was not performed. Both areas were excised
–– Paucicellular invasive lobular carcinoma can be subtle in separate lumpectomies, with sentinel lymph node
in areas of sclerosis. biopsy.
a b c d
e f
Fig. 4.43 Case 1. Right breast craniocaudal (a) and mediolateral with posterior acoustic shadowing corresponding in size and location to
oblique (b) synthetic digital mammograms show a 5 mm asymmetry in the mammographic finding. Ultrasound-guided core biopsy with
the outer right breast on the craniocaudal view only (arrow), not seen on marker placement. (f) After ultrasound-guided core biopsy of the mass,
the MLO view (circle). Craniocaudal tomosynthesis slice (c) shows the a post-biopsy tomosynthesis craniocaudal mammograms shows the
mass better (arrow), but the spot MLO tomosynthesis slice (d) shows marker (arrow), demonstrating that the mass identified by ultrasound
only architectural distortion (circle) and no mass in the upper breast. (e) corresponded to the tomosynthesis mass
Real-time breast ultrasound shows a 9 mm irregular hypoechoic mass
a b
c d
e f
Fig. 4.44 Case 1. Histologic sections of the core biopsy show invasive with the central area seen at medium power in (d). Also see Fig. 4.6a–c
lobular carcinoma with highly variable tumor cell density in sclerotic for additional images of the central more sclerotic portion. (e) Higher
stroma. (a) Low power of one of the core samples is densely cellular, power of a different relatively paucicellular area of the core biopsy
medium power of the central area (c) demonstrates dense areas of small shows that tumor cells are E-cadherin negative, with internal control at
dyscohesive cells, single and in small collections. (b) Low power of left (200×), and estrogen receptor (ER) positive (f) in a serial section
another core from the same specimen has widely scattered collections (200×)
of lobular cells in dense stroma against background fatty breast (right),
a b
c d
e f
Fig. 4.45 Case 2. Invasive lobular carcinoma. (a) One of the lumpec- tive markers are consistent with a “luminal A” phenotype (all 200×). (c)
tomy specimens contained a 2.4 cm invasive lobular carcinoma (40×). Estrogen receptor (ER) positive (3+ > 95%), as is ALH. (d) Progesterone
(b) Higher power shows single file infiltrating carcinoma and ALH in receptor (PR) positive (3+ >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
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
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
e f
g h
–– 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
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
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 (<5%), Her2
• Take-Home Messages: immunohistochemistry negative (1+), not amplified by
–– MRI may be helpful in assessing extent of lobular lesions. FISH. The subsequent resection contained a 2.7 cm focus of
–– LCIS on core biopsy may be associated with unsam- similar carcinoma, grade 1 (poor tubule formation (3), low
pled invasive carcinoma, especially if abundant or of nuclear grade (1), low mitotic rate (1), total of 5 points).
pleomorphic or florid type. Margins were positive, including separately submitted new
–– Invasive lobular carcinoma is graded with the WHO margin specimens, and were subsequently re-excised.
criteria. • Final Diagnosis: Invasive lobular carcinoma, grade 1,
–– Margins should be reported for pleomorphic LCIS, but 2.7 cm (pT2 N0), background of lobular neoplasia in situ
need not be reported for ALH or classic LCIS. • Take-Home Messages:
–– Management of pleomorphic LCIS margins is contro- –– Classic lobular carcinoma grows slowly and may be
versial, but many clinicians and patients chose to mammographically occult at first.
re-excise. –– Invasive lobular carcinoma is graded with the WHO
criteria.
Case 5 –– Invasive lobular carcinoma may be larger than mam-
• History: An 85-year-old female with history of left inva- mographically or grossly recognized.
sive ductal carcinoma (20 years ago) status post
lumpectomy Case 6
• Imaging: See Fig. 4.50a–h. A focal asymmetry in the • History: A 68-year-old woman with history of left breast
upper outer right breast was visualized on the most recent DCIS 15 years prior, now with right axillary adenopathy
study. In retrospect, there is a slowly developing focal on mammography, but no diagnostic findings in the right
asymmetry up to 2 years before, a pattern commonly seen breast. A right axillary lymph node FNA was performed
in ILC on mammography. A core biopsy was performed, and demonstrated metastatic carcinoma, consistent with
followed by surgical excision. breast primary.
a b e f
2 years prior
Current
c d g h
3 years prior
1 year prior
Fig. 4.50 Case 5. There is a focal asymmetry in the upper outer right breast (circle). (c, d) One year prior, there was asymmetry in the right
breast on the most recent study. In retrospect, there is a slowly develop- upper outer breast (circle) that had grown on the subsequent study. (e, f)
ing focal asymmetry up to 2 years before, a pattern commonly seen in Two years prior, there was asymmetry in the right upper outer breast
ILC on mammography. In the most recent study on craniocaudal (a) (circle) that was new since the year before (g, h)
and mediolateral (b), there is a focal asymmetry in the right upper outer
a b
Fig. 4.51 Case 5. Histologic sections from the core biopsy and surgi- invasive carcinoma was also difficult to determine at the time of sur-
cal specimen. (a) Core biopsy invasive lobular carcinoma in collage- gery; margins were positive for invasive lobular carcinoma (tumor at
nous stroma, with ALH at left (100×). (b) Higher power of single file ink, bottom, 40×). (d) Higher power of lobular carcinoma in the surgi-
cells in collagenous and adipocytic stroma (200×). (c) The extent of cal specimen (200×)
c d
• Imaging: Right breast MRI demonstrated mass with –– Residual invasive pleomorphic lobular carcinoma,
abnormal enhancement, and core biopsy was performed. 1.8 cm size (ypT1c pN0(i+)
• Histologic Findings: See Fig. 4.52a–d. Biopsy showed • Subsequent History: Four years later, the patient pre-
cords and small sheets of plump invasive carcinoma cells sented with a palpable right axillary mass; core biopsy is
with grade 2 nuclei. E-cadherin was negative with posi- shown in Fig. 4.54a–f. Both the morphology and predic-
tive internal control; however, given the pleomorphic tive markers were similar to the primary tumor, with min-
morphology along with ER negativity (predictive mark- imal ER staining, and Her2 FISH now scored as low
ers: ER negative (0), PR negative (0), Ki-67 20%, Her2 amplified. The morphology, together with the strong dif-
equivocal (2+ immunohistochemistry, Her2/CEP17 fuse androgen receptor staining, is suggestive of apocrine
ratio = 1.9, Her2/cell = 5)), a diagnosis of invasive mam- differentiation in this pleomorphic lobular carcinoma,
mary carcinoma was made, with classification deferred to consistent with recurrence.
the excisional specimen. • Take-Home Messages:
• The post-chemotherapy specimen is shown in Fig. 4.53a–f, –– Pleomorphic lobular carcinoma may have good
with only a small nodule of fairly high tumor cellularity response to chemotherapy.
remaining in the middle of a 1.8 cm tumor bed containing –– Pleomorphic lobular carcinoma is more often hormone
scattered individual tumor cells. Tumor cells had more receptor negative with higher proliferation than classic
abundant pale cytoplasm after chemotherapy and were lobular carcinoma.
again E-cadherin negative. One lymph node was largely –– Pleomorphic lobular carcinoma may behave more
replaced by fibrosis, consistent with an area of treated aggressively.
tumor, with a few residual isolated tumor cells. According –– A subset of pleomorphic lobular carcinomas, and pleo-
to the AJCC eighth edition, the span of fibrosis is not mea- morphic LCIS, show apocrine features.
sured in determination of nodal stage in the post-
chemotherapy setting, but rather the size of the largest Acknowledgments The authors would like to thank Norm Cyr for pro-
tumor deposit. viding expert assistance with figures.
• Final Diagnosis:
a b
c d
Fig. 4.52 Case 6. Diagnostic core biopsy. (a) Cords and small sheets noma with concern for pleomorphic lobular carcinoma and final clas-
of plump invasive carcinoma cells with grade 2 nuclei (200×). (b) sification deferred to resection. (c) Estrogen receptor negative with
E-cadherin negative with positive internal control (normal duct, upper, positive internal control (bottom left, 200×). (d) Her2/neu immunohis-
200×). This core biopsy was diagnosed as invasive mammary carci- tochemistry 2+, FISH equivocal (Her2/CEP17 ratio 1.9, Her2/cell = 5)
a b
c d
e f
Fig. 4.53 Case 6. Post-chemotherapy pathology. (a) There was one tive (not shown). (c) Axillary lymph node with treatment effect and
small nodule of residual invasive carcinoma 0.4 cm (100×), within a residual isolated tumor cells. At low power, the lymph node appears
1.8 cm span of scattered single cells in tumor bed (not shown). (b) High largely fibrotic (40×). Tumor cells are apparent at higher power on H&E
power shows eccentric nuclei and more abundant histiocytoid cyto- 200× (d), 400× (e) (arrows), and with keratin stain (f) (100×)
plasm as compared to original biopsy (400×), again E-cadherin nega-
a b
c d
e f
Fig. 4.54 Case 6. Recurrent carcinoma. (a) H&E section demonstrates chemistry was weak to focally moderate circumferential staining (2+,
sheets and cords of cells with grade 3 nuclei and moderate amounts of shown here, 400×); HER2 FISH studies were low amplified (Her2/
eosinophilic cytoplasm, infiltrating fat (200×). (b) E-cadherin is nega- CEP17 = 2.2, Her2/cell = 5). (f) Androgen receptor immunohistochem-
tive (400×) and p120 is cytoplasmic (c), supporting lobular differentia- istry was strong and diffuse (400×), consistent with a Her2+ pleomor-
tion (400×). (d) Estrogen receptor shows weak staining in only a subset phic apocrine lobular carcinoma
of nuclei; PR stains similarly (not shown). (e) Her2/neu immunohisto-
References 18. Brogi E, Murray MP, Corben AD. Lobular carcinoma, not only a
classic. Breast J. 2010;16(Suppl 1):S10–4.
19. Schnitt SJ, Collins L. Chapter 10, invasive breast cancer: invasive
1. Lakhani SR, Rakha E, Simpson PT. Invasive lobular carcinoma.
lobular carcinoma. In: Biopsy interpretation of the breast. 3rd ed.
In: Lakhani SR, Ellis IO, Schnitt SJ, Tan PH, van de Vijver MJ,
Philadelphia: Wolters Kluwer; 2018. p. 315–26.
editors. WHO classification of Tumours of the breast. Lyon:
20. Weidner N, Semple JP. Pleomorphic variant of invasive lobular
IARC; 2012. p. 40–5.
carcinoma of the breast. Hum Pathol. 1992;23(10):1167–71.
2. Hoda SA, Brogi E, Koerner FC, Rosen PP. Chapter 32, Invasive
21. Iorfida M, Maiorano E, Orvieto E, Maisonneuve P, Bottiglieri L,
lobular carcinoma. In: Rosen’s breast pathology. 4th ed.
Rotmensz N, et al. Invasive lobular breast cancer: subtypes and
Philadelphia: Wolters Kluwer; 2014. p. 855–92.
outcome. Breast Cancer Res Treat. 2012;133(2):713–23.
3. Eheman CR, Shaw KM, Ryerson AB, Miller JW, Ajani UA, White
22. Monhollen L, Morrison C, Ademuyiwa FO, Chandrasekhar R,
MC. The changing incidence of in situ and invasive ductal and
lobular breast carcinomas: United States, 1999–2004. Cancer Khoury T. Pleomorphic lobular carcinoma: a distinctive clinical and
Epidemiol Biomark Prev. 2009;18(6):1763–9. molecular breast cancer type. Histopathology. 2012;61(3):365–77.
4. Dossus L, Benusiglio PR. Lobular breast cancer: incidence 23. Tan PH, Harada O, Thike AA, Tse GM. Histiocytoid breast carci-
and genetic and non-genetic risk factors. Breast Cancer Res. noma: an enigmatic lobular entity. J Clin Pathol. 2011;64(8):654–9.
2015;17:37. 24. Costarelli L, Campagna D, Ascarelli A, Cavaliere F, Colavito MH,
5. Brem RF, Ioffe M, Rapelyea JA, Yost KG, Weigert JM, Bertrand Ponzani T, et al. Pleomorphic lobular carcinoma: is it more similar
ML, et al. Invasive lobular carcinoma: detection with mammogra- to a classic lobular cancer or to a high-grade ductal cancer? Breast
phy, sonography, MRI, and breast-specific gamma imaging. AJR Cancer (Dove Med Press). 2017;9:581–6.
Am J Roentgenol. 2009;192(2):379–83. 25. Lien HC, Chen YL, Juang YL, Jeng YM. Frequent alterations of
6. Savaridas SL, Bristow GD, Cox J. Invasive lobular Cancer of the HER2 through mutation, amplification, or overexpression in pleo-
breast: a pictorial essay of imaging findings on mammography, morphic lobular carcinoma of the breast. Breast Cancer Res Treat.
sonography, and magnetic resonance imaging. Can Assoc Radiol 2015;150(2):447–55.
J. 2016;67(3):263–76. 26. Simpson PT, Reis-Filho JS, Lambros MB, Jones C, Steele
7. Jacobs C, Clemons M, Addison C, Robertson S, Arnaout D, Mackay A, et al. Molecular profiling pleomorphic lobu-
A. Issues affecting the loco-regional and systemic Management lar carcinomas of the breast: evidence for a common molecu-
of Patients with invasive lobular carcinoma of the breast. Breast J. lar genetic pathway with classic lobular carcinomas. J Pathol.
2016;22(1):45–53. 2008;215(3):231–44.
8. Chamming’s F, Kao E, Aldis A, Ferré R, Omeroglu A, Reinhold 27. Yoder BJ, Wilkinson EJ, Massoll NA. Molecular and morphologic
C, et al. Imaging features and conspicuity of invasive lobu- distinctions between infiltrating ductal and lobular carcinoma of
lar carcinomas on digital breast tomosynthesis. Br J Radiol. the breast. Breast J. 2007;13(2):172–9.
2017;90(1073):20170128. 28. Dieci MV, Smutná V, Scott V, Yin G, Xu R, Vielh P, et al. Whole
9. Christgen M, Steinemann D, Kühnle E, Länger F, Gluz O, Harbeck exome sequencing of rare aggressive breast cancer histologies.
N, et al. Lobular breast cancer: clinical, molecular and morpho- Breast Cancer Res Treat. 2016;156(1):21–32.
logical characteristics. Pathol Res Pract. 2016;212(7):583–97. 29. Ellis IO, Simpson JF. Grading. In: Lakhani SR, Ellis IO, Schnitt
10. Purushotham A, Pinder S, Cariati M, Harries M, Goldhirsch SJ, Tan PH, van de Vijver MJ, editors. WHO classification of
A. Neoadjuvant chemotherapy: not the best option in estrogen Tumours of the breast. Lyon: IARC; 2012. p. 19–20.
receptor-positive, HER2-negative, invasive classical lobular car- 30. Lester S, Weaver D, Morrow M. Staging. In: Lakhani SR, Ellis IO,
cinoma of the breast? J Clin Oncol. 2010;28(22):3552–4. Schnitt SJ, Tan PH, van de Vijver MJ, editors. WHO classification
11. Lester SC, Connolly JL, Amin MB. College of American of Tumours of the breast. Lyon: IARC; 2012. p. 20–2.
Pathologists protocol for the reporting of ductal carcinoma in situ. 31. Hortobagyi GN, Connolly JL, D’Orsi CJ, Edge SB, Mittendorf
Arch Pathol Lab Med. 2009;133(1):13–4. ES, Rugo HA, et al. Breast. In: American joint committee on
12. Fitzgibbons PL, Bose S, Chen Y-Y, Connolly JL, de Baca ME, Cancer, AJCC Cancer staging manual. 8th ed. Chicago: Springer;
Edgerton M, et al. Protocol for the examination of specimens 2017. p. 589–628.
from patients with ductal carcinoma In Situ (DCIS) of the Breast. 32. Gamallo C, Palacios J, Suarez A, Pizarro A, Navarro P, Quintanilla
College of American Pathologists Cancer Protocol Template. M, et al. Correlation of E-cadherin expression with differentia-
http://www.cap.org/ShowProperty?nodePath=/UCMCon/ tion grade and histological type in breast carcinoma. Am J Pathol.
Contribution%20Folders/WebContent/pdf/cp-breast-dcis-18pro- 1993;142(4):987–93.
tocol-4100.pdf. Accessed 17 Feb 2018. 33. Moll R, Mitze M, Frixen UH, Birchmeier W. Differential loss of
13. Grin A, Horne G, Ennis M, O’Malley FP. Measuring extent of E-cadherin expression in infiltrating ductal and lobular breast car-
ductal carcinoma in situ in breast excision specimens: a compari- cinomas. Am J Pathol. 1993;143(6):1731–42.
son of 4 methods. Arch Pathol Lab Med. 2009;133(1):31–7. 34. Berx G, Cleton-Jansen AM, Nollet F, de Leeuw WJ, van de Vijver
14. Dabbs DJ, Schnitt SJ, Geyer FC, Weigelt B, Baehner FL, Decker M, Cornelisse C, et al. E-cadherin is a tumour/invasion suppres-
T, et al. Lobular neoplasia of the breast revisited with emphasis on sor gene mutated in human lobular breast cancers. EMBO J.
the role of E-cadherin immunohistochemistry. Am J Surg Pathol. 1995;14(24):6107–15.
2013;37(7):e1–11. 35. Berx G, Becker KF, Höfler H, van Roy F. Mutations of the human
15. McCart Reed AE, Kutasovic JR, Lakhani SR, Simpson E-cadherin (CDH1) gene. Hum Mutat. 1998;12(4):226–37.
PT. Invasive lobular carcinoma of the breast: morphology, bio- 36. De Leeuw WJ, Berx G, Vos CB, Peterse JL, Van de Vijver MJ,
markers and 'omics. Breast Cancer Res. 2015;17:12. Litvinov S, et al. Simultaneous loss of e-cadherin and catenins
16. Schnitt SJ, Collins L. Biopsy interpretation of the breast. 3rd ed. in invasive lobular breast cancer and lobular carcinoma in situ. J
Philadelphia: Wolters Kluwer; 2018. p. 315–26. Pathol. 1997;183:404–11.
17. Haagensen CD, Lane N, Lattes R, Bodian C. Lobular neopla- 37. Gonzalez MA, Pinder SE, Wencyk PM, Bell JA, Elston CW,
sia (so-called lobular carcinoma in situ) of the breast. Cancer. Nicholson RI, et al. An immunohistochemical examination of the
1978;42(2):737–69. expression of E-cadherin, alpha- and beta/gamma-catenins, and
alpha2- and beta1-integrins in invasive breast cancer. J Pathol. 56. Esposito NN, Chivukula M, Dabbs DJ. The ductal phenotypic
1999;187(5):523–9. expression of the E-cadherin/catenin complex in tubulolobular
38. Berx G, Van Roy F. The E-cadherin/catenin complex: an impor- carcinoma of the breast: an immunohistochemical and clinico-
tant gatekeeper in breast cancer tumorigenesis and malignant pro- pathologic study. Mod Pathol. 2007;20(1):130–8.
gression. Breast Cancer Res. 2001;3(5):289–93. 57. Naidoo K, Beardsley B, Carder PJ, Deb R, Fish D, Girling A, et al.
39. de Deus MR, Wludarski SC, Carvalho FM, Bacchi Accuracy of classification of invasive lobular carcinoma on needle
CE. Immunohistochemistry applied to the differential diagno- core biopsy of the breast. J Clin Pathol. 2016;69(12):1122–3.
sis between ductal and lobular carcinoma of the breast. Appl 58. Rakha EA, Abd El Rehim D, Pinder SE, Lewis SA, Ellis
Immunohistochem Mol Morphol. 2013;21(1):1–12. IO. E-cadherin expression in invasive non-lobular carcinoma
40. Canas-Marques R, Schnitt SJ. E-cadherin immunohistochem- of the breast and its prognostic significance. Histopathology.
istry in breast pathology: uses and pitfalls. Histopathology. 2005;46(6):685–93.
2016;68(1):57–69. 59. Rakha EA, Gill MS, El-Sayed ME, Khan MM, Hodi Z, Blamey
41. Hansford S, Kaurah P, Li-Chang H, Woo M, Senz J, Pinheiro H, RW, et al. The biological and clinical characteristics of breast car-
et al. Hereditary diffuse gastric Cancer syndrome: CDH1 muta- cinoma with mixed ductal and lobular morphology. Breast Cancer
tions and beyond. JAMA Oncol. 2015;1(1):23–32. Res Treat. 2009;114(2):243–50.
42. Corso G, Intra M, Trentin C, Veronesi P, Galimberti V. CDH1 60. Arps DP, Healy P, Zhao L, Kleer CG, Pang JC. Invasive ductal car-
germline mutations and hereditary lobular breast cancer. Familial cinoma with lobular features: a comparison study to invasive duc-
Cancer. 2016;15(2):215–9. tal and invasive lobular carcinomas of the breast. Breast Cancer
43. Ciriello G, Gatza ML, Beck AH, Wilkerson MD, Rhie SK, Pastore Res Treat. 2013;138(3):719–26.
A, et al. Comprehensive molecular portraits of invasive lobular 61. Moran MS, Schnitt SJ, Giuliano AE, Harris JR, Khan SA,
breast Cancer. Cell. 2015;163(2):506–19. Horton J, et al. Society of Surgical Oncology-American Society
44. Desmedt C, Zoppoli G, Gundem G, Pruneri G, Larsimont D, for Radiation Oncology consensus guideline on margins for
Fornili M, et al. Genomic characterization of primary invasive breast-conserving surgery with whole-breast irradiation in stages
lobular breast Cancer. J Clin Oncol. 2016;34(16):1872–81. I and II invasive breast cancer. Int J Radiat Oncol Biol Phys.
45. Desmedt C, Zoppoli G, Sotiriou C, Salgado R. Transcriptomic and 2014;88(3):553–64.
genomic features of invasive lobular breast cancer. Semin Cancer 62. Korkola JE, DeVries S, Fridlyand J, Hwang ES, Estep AL,
Biol. 2017;44:98–105. Chen YY, et al. Differentiation of lobular versus ductal breast
46. Karayiannakis AJ, Nakopoulou L, Gakiopoulou H, Keramopoulos carcinomas by expression microarray analysis. Cancer Res.
A, Davaris PS, Pignatelli M. Expression patterns of beta- 2003;63(21):7167–75.
catenin in in situ and invasive breast cancer. Eur J Surg Oncol. 63. Zhao H, Langerød A, Ji Y, Nowels KW, Nesland JM, Tibshirani R,
2001;27(1):31–6. et al. Different gene expression patterns in invasive lobular and duc-
47. Sarrio´ D, Perez-Mies B, Hardisson D, Moreno-Bueno G, Suárez tal carcinomas of the breast. Mol Biol Cell. 2004;15(6):2523–36.
A, Cano A, et al. Cytoplasmic localization of p120ctn and 64. Bertucci F, Orsetti B, Nègre V, Finetti P, Rougé C, Ahomadegbe
E-cadherin loss characterize lobular breast carcinoma from prein- JC, et al. Lobular and ductal carcinomas of the breast
vasive to metastatic lesions. Oncogene. 2004;23:3272–83. have distinct genomic and expression profiles. Oncogene.
48. Mastracci TL, Tjan S, Bane AL, O’Malley FP, Andrulis 2008;27(40):5359–72.
IL. E-cadherin alterations in atypical lobular hyperpla- 65. Weigelt B, Geyer FC, Natrajan R, Lopez-Garcia MA, Ahmad
sia and lobular carcinoma in situ of the breast. Mod Pathol. AS, Savage K, et al. The molecular underpinning of lobular his-
2005;18(6):741–51. tological growth pattern: a genome-wide transcriptomic analysis
49. Dabbs DJ, Bhargava R, Chivukula M. Lobular versus ductal breast of invasive lobular carcinomas and grade- and molecular subtype-
neoplasms: the diagnostic utility of p120 catenin. Am J Surg matched invasive ductal carcinomas of no special type. J Pathol.
Pathol. 2007;31(3):427–37. 2010;220(1):45–57.
50. Dabbs DJ, Kaplai M, Chivukula M, Kanbour A, Kanbour-Shakir 66. Paik S, Shak S, Tang G, Kim C, Baker J, Cronin M, et al. A mul-
A, Carter GJ. The spectrum of morphomolecular abnormalities tigene assay to predict recurrence of tamoxifen-treated, node-
of the E-cadherin/catenin complex in pleomorphic lobular car- negative primary breast cancer. N Engl J Med. 2004;351:2817–26.
cinoma of the breast. Appl Immunohistochem Mol Morphol. 67. Paik A, Tang G, Shuk S, Kim C, Baker J, Kim W, et al. Gene
2007;15(3):260–6. expression and benefit of chemotherapy in women with node-
51. Da Silva L, Parry S, Reid L, Keith P, Waddell N, Kossai M, et al. negative, estrogen receptor positive breast cancer. J Clin Oncol.
Aberrant expression of E-cadherin in lobular carcinomas of the 2006;24:3726–34.
breast. Am J Surg Pathol. 2008;32(5):773–83. 68. Conlon N, Ross DS, Howard J, Catalano JP, Dickler MN, Tan
52. Rakha EA, Patel A, Powe DG, Benhasouna A, Green AR, Lambros LK. Is there a role for Oncotype dx testing in invasive lobular
MB, et al. Clinical and biological significance of E-cadherin pro- carcinoma? Breast J. 2015;21(5):514–9.
tein expression in invasive lobular carcinoma of the breast. Am J 69. Tsai ML, Lillemoe TJ, Finkelstein MJ, Money JE, Susnik
Surg Pathol. 2010;34(10):1472–9. B, Grimm E, et al. Utility of Oncotype DX risk assessment in
53. Acs G, Lawton TJ, Rebbeck TR, LiVolsi VA, Zhang PJ. Differential patients with invasive lobular carcinoma. Clin Breast Cancer.
expression of E-cadherin in lobular and ductal neoplasms of the 2016 Feb;16(1):45–50.
breast and its biologic and diagnostic implications. Am J Clin 70. Felts JL, Zhu J, Han B, Smith SJ, Truica CI. An analysis of Oncotype
Pathol. 2001;115(1):85–98. DX recurrence scores and Clinicopathologic characteristics in
54. Wheeler DT, Tai LH, Bratthauer GL, Waldner DL, Tavassoli invasive lobular breast Cancer. Breast J. 2017;23(6):677–86.
FA. Tubulolobular carcinoma of the breast: an analysis of 27 cases 71. Michaut M, Chin SF, Majewski I, Severson TM, Bismeijer T, de
of a tumor with a hybrid morphology and immunoprofile. Am J Koning L, et al. Integration of genomic, transcriptomic and pro-
Surg Pathol. 2004;28(12):1587–93. teomic data identifies two biologically distinct subtypes of inva-
55. Kuroda H, Tamaru J, Takeuchi I, Ohnisi K, Sakamoto G, Adachi sive lobular breast cancer. Sci Rep. 2016;6:18517.
A, et al. Expression of E-cadherin, alpha-catenin, and beta- 72. Moten A, Obirieze A, Wilson LL. Characterizing lobular carci-
catenin in tubulolobular carcinoma of the breast. Virchows Arch. noma of the male breast using the SEER database. J Surg Res.
2006;448(4):500–5. 2013;185(2):e71–6.
73. Maguire A, Brogi E. Sentinel lymph nodes for breast carcinoma: 91. Raju U, Ma CK, Shaw A. Signet ring variant of lobular carcinoma
an update on current practice. Histopathology. 2016;68(1):152–67. of the breast: a clinicopathologic and immunohistochemical study.
74. Rosen PP. Chapter 44, pathologic examination of breast and lymph Mod Pathol. 1993;6(5):516–20.
node specimens, including sentinel lymph nodes. In: Rosen’s 92. Gown AM, Fulton RS, Kandalaft PL. Markers of metastatic carci-
breast pathology. 3rd ed. Philadelphia: Wolters Kluwer/Lippincott noma of breast origin. Histopathology. 2016;68(1):86–95.
Williams & Wilkins; 2009. p. 1034–102. 93. Habashy HO, Powe DG, Rakha EA, Ball G, Macmillan RD, Green
75. Cserni G, Bianchi S, Vezzosi V, Peterse H, Sapino A, Arisio R, AR, Ellis IO. The prognostic significance of PELP1 expression in
et al. The value of cytokeratin immunohistochemistry in the eval- invasive breast cancer with emphasis on the ER-positive luminal-
uation of axillary sentinel lymph nodes in patients with lobular like subtype. Breast Cancer Res Treat. 2010;120(3):603–12.
breast carcinoma. J Clin Pathol. 2006;59(5):518–22. 94. Davis DG, Siddiqui MT, Oprea-Ilies G, Stevens K, Osunkoya AO,
76. Patel A, D’Alfonso T, Cheng E, Hoda SA. Sentinel lymph nodes Cohen C, et al. GATA-3 and FOXA1 expression is useful to dif-
in classic invasive lobular carcinoma of the breast: cytokeratin ferentiate breast carcinoma from other carcinomas. Hum Pathol.
immunostain ensures detection, and precise determination of 2016;47(1):26–31.
extent, of involvement. Am J Surg Pathol. 2017;41(11):1499–505. 95. Peng Y, Butt YM, Chen B, Zhang X, Tang P. Update on
77. Giuliano AE, Hunt KK, Ballman KV, Beitsch PD, Whitworth PW, Immunohistochemical analysis in breast lesions. Arch Pathol Lab
Blumencranz PW, et al. Axillary dissection vs no axillary dissec- Med. 2017;141(8):1033–51.
tion in women with invasive breast cancer and sentinel node metas- 96. Robens J, Goldstein L, Gown AM, Schnitt SJ. Thyroid transcrip-
tasis: a randomized clinical trial. JAMA. 2011;305(6):569–75. tion factor-1 expression in breast carcinomas. Am J Surg Pathol.
78. Giuliano AE, Ballman KV, McCall L, Beitsch PD, Brennan MB, 2010;34:1881–5.
Kelemen PR, et al. Effect of axillary dissection vs no axillary dis- 97. Provenzano E, Byrne DJ, Russell PA, Wright GM, Generali D,
section on 10-year overall survival among women with invasive Fox SB. Differential expression of immunohistochemical mark-
breast Cancer and sentinel node metastasis: the ACOSOG Z0011 ers in primary lung and breast cancers enriched for triple-negative
(Alliance) randomized clinical trial. JAMA. 2017;318(10):918–26. tumours. Histopathology. 2016;68:367–77.
79. Weaver DL, Ashikaga T, Krag DN, Skelly JM, Anderson SJ, 98. Chu PG, Weiss LM. Immunohistochemical characterization of
Harlow SP, et al. Effect of occult metastases on survival in node- signet-ring cell carcinomas of the stomach, breast, and colon. Am
negative breast cancer. N Engl J Med. 2011;364(5):412–21. J Clin Pathol. 2004;121(6):884–92.
80. Sapino A, Cassoni P, Ferrero E, Bongiovanni M, Righi L, Fortunati 99. Mahmud N, Ford JM, Longacre TA, Parent R, Norton
N, et al. Estrogen receptor alpha is a novel marker expressed by JA. Metastatic lobular breast carcinoma mimicking primary signet
follicular dendritic cells in lymph nodes and tumor-associated ring adenocarcinoma in a patient with a suspected CDH1 muta-
lymphoid infiltrates. Am J Pathol. 2003;163(4):1313–20. tion. J Clin Oncol. 2015;33(4):e19–21.
81. Grube BJ, Hansen NM, Ye X, Giuliano AE. Tumor characteristics 100. Ellis CL, Chang AG, Cimino-Mathews A, Argani P, Youssef RF,
predictive of sentinel node metastases in 105 consecutive patients Kapur P, et al. GATA-3 immunohistochemistry in the differential
with invasive lobular carcinoma. Am J Surg. 2002;184(4):372–6. diagnosis of adenocarcinoma of the urinary bladder. Am J Surg
82. Classe JM, Loussouarn D, Campion L, Fiche M, Curtet C, Dravet Pathol. 2013;37(11):1756–60.
F, et al. Validation of axillary sentinel lymph node detection in the 101. Borhan WM, Cimino-Mathews AM, Montgomery EA, Epstein
staging of early lobular invasive breast carcinoma: a prospective JI. Immunohistochemical differentiation of Plasmacytoid
study. Cancer. 2004;100(5):935–41. Urothelial carcinoma from secondary carcinoma involvement of
83. Jensen AJ, Naik AM, Pommier RF, Vetto JT, Troxell ML. Factors the bladder. Am J Surg Pathol. 2017;41(11):1570–5.
influencing accuracy of axillary sentinel lymph node frozen sec- 102. Su MC, Yuan RH, Lin CY, Jeng YM. Cadherin-17 is a useful diag-
tion for breast cancer. Am J Surg. 2010;199(5):629–35. nostic marker for adenocarcinomas of the digestive system. Mod
84. Taras AR, Hendrickson NA, Pugliese MS, Lowe KA, Atwood Pathol. 2008;21(11):1379–86.
M, Beatty JD. Intraoperative evaluation of sentinel lymph 103. Laury AR, Perets R, Piao H, Krane JF, Barletta JA, French C, et al.
nodes in invasive lobular carcinoma of the breast. Am J Surg. A comprehensive analysis of PAX8 expression in human epithe-
2009;197(5):643–6; discussion 646–7 lial tumors. Am J Surg Pathol. 2011;35(6):816–26.
85. Creager AJ, Geisinger KR, Perrier ND, Shen P, Shaw JA, Young 104. Recine MA, Deavers MT, Middleton LP, Silva EG, Malpica
PR, Case D, Levine EA. Intraoperative imprint cytologic evalua- A. Serous carcinoma of the ovary and peritoneum with metastases
tion of sentinel lymph nodes for lobular carcinoma of the breast. to the breast and axillary lymph nodes: a potential pitfall. Am J
Ann Surg. 2004;239(1):61–6. Surg Pathol. 2004;28(12):1646–51.
86. Howard-McNatt M, Geisinger KR, Stewart JH 4th, Shen P, Levine 105. Nonaka D, Chiriboga L, Soslow RA. Expression of PAX8 as a
EA. Is intraoperative imprint cytology evaluation still feasible for useful marker in distinguishing ovarian carcinomas from mam-
the evaluation of sentinel lymph nodes for lobular carcinoma of mary carcinomas. Am J Surg Pathol. 2008;32:1566–71.
the breast? Ann Surg Oncol. 2012;19(3):929–34. 106. Bhargava R, Beriwal S, Dabbs DJ. Mammaglobin vs GCDFP-15:
87. Tavassoli FA, Eusebi V. Chapter8, major variants of carcinoma: an immunohistologic validation survey for sensitivity and speci-
invasive lobular carcinoma. In: Tumors of the mammary gland. ficity. Am J Clin Pathol. 2007;127:103–13.
Washington DC: ARP press; 2009. p. 156–68. 107. Onuma K, Dabbs DJ, Bhargava R. Mammaglobin expression in
88. Miettinen M, PA MC, Sarlomo-Rikala M, Rys J, Czapiewski the female genital tract: immunohistochemical analysis in benign
P, Wazny K, et al. GATA3: a multispecific but potentially use- and neoplastic endocervix and endometrium. Int J Gynecol Pathol.
ful marker in surgical pathology: a systematic analysis of 2008;27:418–25.
2500 epithelial and nonepithelial tumors. Am J Surg Pathol. 108. Kandalaft PL, Simon RA, Isacson C, Gown AM. Comparative
2014;38(1):13–22. sensitivities and specificities of antibodies to breast markers
89. Wendroth SM, Mentrikoski MJ, Wick MR. GATA3 expression GCDFP-15, mammaglobin a, and different clones of antibodies
in morphologic subtypes of breast carcinoma: a comparison with to GATA-3: a study of 338 tumors using whole sections. Appl
gross cystic disease fluid protein 15 and mammaglobin. Ann Immunohistochem Mol Morphol. 2016;24:609–14.
Diagn Pathol. 2015;19:6–9. 109. So JS, Epstein JI. GATA3 expression in paragangliomas: a pitfall
90. Asch-Kendrick R, Cimino-Mathews A. The role of GATA3 in potentially leading to misdiagnosis of urothelial carcinoma. Mod
breast carcinomas: a review. Hum Pathol. 2016;48:37–47. Pathol. 2013;26(10):1365–70.
110. Barbareschi M, Maisonneuve P, Aldovini D, Cangi MG, Pecciarini pathologic features and clinical management. Ann Surg Oncol.
L, Angelo Mauri F, et al. High syndecan-1 expression in breast 2015;22(13):4263–9.
carcinoma is related to an aggressive phenotype and to poorer 129. Hoda SA, Brogi E, Koerner FC, Rosen PP. Chapter 31 lobular
prognosis. Cancer. 2003;98(3):474–83. carcinoma in situ and atypical lobular hyperplasia. In: Rosen’s
111. Kambham N, Kong C, Longacre TA, Natkunam Y. Utility of syn- breast pathology. 4th ed. Philadelphia: Wolters Kluwer; 2014.
decan-1 (CD138) expression in the diagnosis of undifferentiated p. 797–854.
malignant neoplasms: a tissue microarray study of 1,754 cases. 130. Schnitt SJ, Collins L. Chapter 5, lobular carcinoma in situ and
Appl Immunohistochem Mol Morphol. 2005;13(4):304–10. atypical lobular hyperplasia. In: Biopsy interpretation of the
112. Cessna MH, Zhou H, Perkins SL, Tripp SR, Layfield L, Daines C, breast. 3rd ed. Philadelphia: Wolters Kluwer; 2018. p. 141–79.
et al. Are myogenin and myoD1 expression specific for rhabdo- 131. Tavassoli FA, Eusebi V. Chapter 3, lobular intraepithelial neopla-
myosarcoma? A study of 150 cases, with emphasis on spindle cell sia. In: Tumors of the mammary gland. Washington DC: ARP
mimics. Am J Surg Pathol. 2001;25(9):1150–7. Press; 2009. p. 53–66.
113. Charville GW, Varma S, Forgó E, Dumont SN, Zambrano E, Trent 132. American Society of Breast Surgeons. Consensus guideline on
JC, et al. PAX7 expression in Rhabdomyosarcoma, related soft concordance assessment of image-guided breast biopsies and man-
tissue tumors, and small round blue cell neoplasms. Am J Surg agement of borderline or high-risk lesions. https://www.breast-
Pathol. 2016;40(10):1305–15. surgeons.org/new_layout/about/statements/PDF_Statements/
114. Rogers WM, Dobo E, Norton JA, Van Dam J, Jeffrey RB, Concordance_and_High%20RiskLesions.pdf. Accessed 29 Oct
Huntsman DG, et al. Risk-reducing total gastrectomy for germline 2017.
mutations in E-cadherin (CDH1): pathologic findings with clini- 133. National Comprehensive Cancer Network. NCCN clinical prac-
cal implications. Am J Surg Pathol. 2008;32(6):799–809. tice guidelines in oncology: Breast cancer. https://www.nccn.
115. Fox MD, Xiao L, Zhang M, Kamat AM, Siefker-Radtke A, Zhang org/professionals/physician_gls/default.aspx Version 2.2017.
L, et al. Plasmacytoid Urothelial carcinoma of the urinary blad- Accessed 29 Oct 2017.
der: a Clinicopathologic and Immunohistochemical analysis of 49 134. Page DL, Kidd TE Jr, Dupont WD, Simpson JF, Rogers
cases. Am J Clin Pathol. 2017;147(5):500–6. LW. Lobular neoplasia of the breast: higher risk for subsequent
116. Lakhani SR, Schnitt SJ, O’Malley F, van de Vijver MJ, Simpson invasive cancer predicted by more extensive disease. Hum Pathol.
PT, Palcios J. Lobular neoplasia. In: Lakhani SR, Ellis IO, Schnitt 1991;22(12):1232–9.
SJ, Tan PH, van de Vijver MJ, editors. WHO classification of 135. Rendi MH, Dintzis SM, Lehman CD, Calhoun KE, Allison
Tumours of the breast. Lyon: IARC; 2012. p. 77–80. KH. Lobular in-situ neoplasia on breast core needle biopsy: imag-
117. Mooney KL, Bassett LW, Apple SK. Upgrade rates of high- ing indication and pathologic extent can identify which patients
risk breast lesions diagnosed on core needle biopsy: a single- require excisional biopsy. Ann Surg Oncol. 2012;19(3):914–21.
institution experience and literature review. Mod Pathol. 136. Susnik B, Day D, Abeln E, Bowman T, Krueger J, Swenson
2016;29(12):1471–84. KK, et al. Surgical outcomes of lobular neoplasia diagnosed in
118. Calhoun BC, Collins LC. Recommendations for excision follow- Core biopsy: prospective study of 316 cases. Clin Breast Cancer.
ing core needle biopsy of the breast: a contemporary evaluation of 2016;16(6):507–13.
the literature. Histopathology. 2016;68:138–51. 137. Wong SM, King T, Boileau JF, Barry WT, Golshan M. Population-
119. Clark CJ, Whang S, Paige KT. Incidence of precancerous lesions based analysis of breast Cancer incidence and survival outcomes
in breast reduction tissue: a pathologic review of 562 consecutive in women diagnosed with lobular carcinoma in situ. Ann Surg
patients. Plast Reconstr Surg. 2009;124(4):1033–9. Oncol. 2017;24(9):2509–17.
120. Ambaye AB, Goodwin AJ, MacLennan SE, Naud S, Weaver 138. King TA, Pilewskie M, Muhsen S, Patil S, Mautner SK, Park A,
DL. Recommendations for pathologic evaluation of reduction et al. Lobular carcinoma in situ: a 29-year longitudinal experi-
mammoplasty specimens: a prospective study with systematic tis- ence evaluating Clinicopathologic features and breast Cancer risk.
sue sampling. Arch Pathol Lab Med. 2017;141(11):1523–8. J Clin Oncol. 2015;33(33):3945–52.
121. Portschy PR, Marmor S, Nzara R, Virnig BA, Tuttle TM. Trends 139. Jorns J, Sabel MS, Pang JC. Lobular neoplasia: morphology and
in incidence and management of lobular carcinoma in situ: a management. Arch Pathol Lab Med. 2014;138(10):1344–9.
population-based analysis. Ann Surg Oncol. 2013;20(10):3240–6. 140. Andrade VP, Ostrovnaya I, Seshan VE, Morrogh M, Giri D,
122. Li CI, Anderson BO, Daling JR, Moe RE. Changing incidence of Olvera N, et al. Clonal relatedness between lobular carcinoma
lobular carcinoma in situ of the breast. Breast Cancer Res Treat. in situ and synchronous malignant lesions. Breast Cancer Res.
2002;75(3):259–68. 2012;14(4):R103.
123. Scoggins M, Krishnamurthy S, Santiago L, Yang W. Lobular car- 141. Sakr RA, Schizas M, Carniello JV, Ng CK, Piscuoglio S,
cinoma in situ of the breast: clinical, radiological, and pathologi- Giri D, et al. Targeted capture massively parallel sequenc-
cal correlation. Acad Radiol. 2013;20(4):463–70. ing analysis of LCIS and invasive lobular cancer: repertoire of
124. Amos B, Chetlen A, Williams N. Atypical lobular hyperplasia somatic genetic alterations and clonal relationships. Mol Oncol.
and lobular carcinoma in situ at core needle biopsy of the breast: 2016;10(2):360–70.
an incidental finding or are there characteristic imaging findings? 142. Begg CB, Ostrovnaya I, Carniello JV, Sakr RA, Giri D, Towers R,
Breast Dis. 2016;36(1):5–14. et al. Clonal relationships between lobular carcinoma in situ and
125. Maxwell AJ, Clements K, Dodwell DJ, Evans AJ, Francis A, other breast malignancies. Breast Cancer Res. 2016;18(1):66.
Hussain M, et al. The radiological features, diagnosis and man- 143. Rosen PP, Kosloff C, Lieberman PH, Adair F, Braun DW Jr.
agement of screen-detected lobular neoplasia of the breast: find- Lobular carcinoma in situ of the breast. Detailed analysis of 99
ings from the Sloane project. Breast. 2016;27:109–15. patients with average follow-up of 24 years. Am J Surg Pathol.
126. Choi BB, Kim SH, Park CS, Cha ES, Lee AW. Radiologic findings 1978;2(3):225–51.
of lobular carcinoma in situ: mammography and ultrasonography. 144. Bodian CA, Perzin KH, Lattes R. Lobular neoplasia. Long
J Clin Ultrasound. 2011;39(2):59–63. term risk of breast cancer and relation to other factors. Cancer.
127. Heller SL, Elias K, Gupta A, Greenwood HI, Mercado CL, Moy 1996;78(5):1024–34.
L. Outcome of high-risk lesions at MRI-guided 9-gauge vacuum- 145. Page DL, Schuyler PA, Dupont WD, Jensen RA, Plummer WD
assisted breast biopsy. AJR Am J Roentgenol. 2014;202(1):237–45. Jr, Simpson JF. Atypical lobular hyperplasia as a unilateral pre-
128. Flanagan MR, Rendi MH, Calhoun KE, Anderson BO, Javid dictor of breast cancer risk: a retrospective cohort study. Lancet.
SH. Pleomorphic lobular carcinoma in situ: radiologic- 2003;361(9352):125–9.
146. Collins LC, Baer HJ, Tamimi RM, Connolly JL, Colditz GA, 164. Vos CB, Cleton-Jansen AM, Berx G, de Leeuw WJ, ter Haar NT,
Schnitt SJ. Magnitude and laterality of breast cancer risk accord- van Roy F, et al. E-cadherin inactivation in lobular carcinoma in
ing to histologic type of atypical hyperplasia: results from the situ of the breast: an early event in tumorigenesis. Br J Cancer.
Nurses’ health study. Cancer. 2007;109(2):180–7. 1997;76(9):1131–3.
147. Coopey SB, Mazzola E, Buckley JM, Sharko J, Belli AK, Kim 165. Bratthauer GL, Moinfar F, Stamatakos MD, Mezzetti TP, Shekitka
EM, et al. The role of chemoprevention in modifying the risk of KM, Man YG, et al. Combined E-cadherin and high molecular
breast cancer in women with atypical breast lesions. Breast Cancer weight cytokeratin immunoprofile differentiates lobular, ductal,
Res Treat. 2012;136(3):627–33. and hybrid mammary intraepithelial neoplasias. Hum Pathol.
148. Simpson PT, Gale T, Fulford LG, Reis-Filho JS, Lakhani SR. The 2002;33(6):620–7.
diagnosis and management of pre-invasive breast disease: pathol- 166. Logan GJ, Dabbs DJ, Lucas PC, Jankowitz RC, Brown DD, Clark
ogy of atypical lobular hyperplasia and lobular carcinoma in situ. BZ, et al. Molecular drivers of lobular carcinoma in situ. Breast
Breast Cancer Res. 2003;5(5):258–62. Cancer Res. 2015;17:76.
149. O’Malley FP. Lobular neoplasia: morphology, biological poten- 167. Mastracci TL, Shadeo A, Colby SM, Tuck AB, O’Malley FP, Bull
tial and management in core biopsies. Mod Pathol. 2010;23(Suppl SB, et al. Genomic alterations in lobular neoplasia: a microarray
2):S14–25. comparative genomic hybridization signature for early neoplas-
150. Page DL, Dupont WD, Rogers LW. Ductal involvement by cells tic proliferation in the breast. Genes Chromosom Cancer. 2006
of atypical lobular hyperplasia in the breast: a long-term follow-up Nov;45(11):1007–17.
study of cancer risk. Hum Pathol. 1988;19(2):201–7. 168. Lu YJ, Osin P, Lakhani SR, Di Palma S, Gusterson BA, Shipley
151. Downs-Kelly E, Bell D, Perkins GH, Sneige N, Middleton JM. Comparative genomic hybridization analysis of lobular carci-
LP. Clinical implications of margin involvement by pleo- noma in situ and atypical lobular hyperplasia and potential roles
morphic lobular carcinoma in situ. Arch Pathol Lab Med. for gains and losses of genetic material in breast neoplasia. Cancer
2011;135(6):737–43. Res. 1998;58(20):4721–7.
152. Khoury T, Karabakhtsian RG, Mattson D, Yan L, Syriac S, 169. Hwang ES, Nyante SJ, Yi Chen Y, Moore D, DeVries S, Korkola
Habib F, et al. Pleomorphic lobular carcinoma in situ of the JE, et al. Clonality of lobular carcinoma in situ and synchronous
breast: clinicopathological review of 47 cases. Histopathology. invasive lobular carcinoma. Cancer. 2004;100(12):2562–72.
2014;64(7):981–93. 170. Buerger H, Simon R, Schäfer KL, Diallo R, Littmann R, Poremba
153. Chen YY, Hwang ES, Roy R, DeVries S, Anderson J, Wa C, C, et al. Genetic relation of lobular carcinoma in situ, ductal carci-
Fitzgibbons PL, et al. Genetic and phenotypic characteristics of noma in situ, and associated invasive carcinoma of the breast. Mol
pleomorphic lobular carcinoma in situ of the breast. Am J Surg Pathol. 2000;53(3):118–21.
Pathol. 2009;33(11):1683–94. 171. Tazaki E, Shishido-Hara Y, Mizutani N, Nomura S, Isaka

154. Boldt V, Stacher E, Halbwedl I, Popper H, Hultschig C, Moinfar H, Ito H, et al. Histopathological and clonal study of com-
F, Ullmann R, Tavassoli FA. Positioning of necrotic lobular bined lobular and ductal carcinoma of the breast. Pathol Int.
intraepithelial neoplasias (LIN, grade 3) within the sequence 2013;63(6):297–304.
of breast carcinoma progression. Genes Chromosom Cancer. 172. Ang DC, Warrick AL, Shilling A, Beadling C, Corless CL, Troxell
2010;49(5):463–70. ML. Frequent phosphatidylinositol-3-kinase mutations in prolif-
155. Shin SJ, Lal A, De Vries S, Suzuki J, Roy R, Hwang ES, Schnitt erative breast lesions. Mod Pathol. 2014;27(5):740–50.
SJ, Waldman FM, Chen YY. Florid lobular carcinoma in situ: 173. Resetkova E, Albarracin C, Sneige N. Collagenous spherulosis of
molecular profiling and comparison to classic lobular carcinoma breast: morphologic study of 59 cases and review of the literature.
in situ and pleomorphic lobular carcinoma in situ. Hum Pathol. Am J Surg Pathol. 2006;30(1):20–7.
2013;44(10):1998–2009. 174. Eisenberg RE, Hoda SA. Lobular carcinoma in situ with collag-
156. Fasola CE, Chen JJ, Jensen KC, Allison KH, Horst enous spherulosis: clinicopathologic characteristics of 38 cases.
KC. Characteristics and clinical outcomes of pleomorphic lobular Breast J. 2014;20(4):440–1.
carcinoma in situ of the breast. Breast J. 2018;24(1):66–9. 175. Singer B, Lin C-Y, West R. Fibroepithelial lesions of the breast
157. Pieri A, Harvey J, Bundred N. Pleomorphic lobular carcinoma in involved by atypical epithelial proliferations: a 12-year single
situ of the breast: can the evidence guide practice? World J Clin institution study. Mod Pathol. 2017;30(supple 2):71A.
Oncol. 2014;5(3):546–53. 176. Sin EI, Wong CY, Yong WS, Ong KW, Madhukumar P, Tan VK,
158. Fadare O, Dadmanesh F, Alvarado-Cabrero I, Snyder R, Stephen et al. Breast carcinoma and phyllodes tumour: a case series. J Clin
Mitchell J, Tot T, et al. Lobular intraepithelial neoplasia [lobular Pathol. 2016;69(4):364–9.
carcinoma in situ] with comedo-type necrosis: a clinicopathologic 177. Middleton LP, Perkins GH, Tucker SL, Sahin AA, Singletary
study of 18 cases. Am J Surg Pathol. 2006;30(11):1445–53. SE. Expression of ERalpha and ERbeta in lobular carcinoma in
159. Bagaria SP, Shamonki J, Kinnaird M, Ray PS, Giuliano
situ. Histopathology. 2007;50(7):875–80.
AE. The florid subtype of lobular carcinoma in situ: marker 178. De Brot M, Koslow Mautner S, Muhsen S, Andrade VP, Mamtani
or precursor for invasive lobular carcinoma? Ann Surg Oncol. A, et al. Pleomorphic lobular carcinoma in situ of the breast: a sin-
2011;18(7):1845–51. gle institution experience with clinical follow-up and centralized
160. Bratthauer GL, Tavassoli FA. Lobular intraepithelial neoplasia: pathology review. Breast Cancer Res Treat. 2017;165(2):411–20.
previously unexplored aspects assessed in 775 cases and their 179. Masannat YA, Bains SK, Pinder SE, Purushotham AD. Challenges
clinical implications. Virchows Arch. 2002;440(2):134–8. in the management of pleomorphic lobular carcinoma in situ of
161. Bussolati G, Botto Micca FB, Eusebi V, Betts CM. Myoepithelial the breast. Breast. 2013;22(2):194–6.
cells in lobular carcinoma in situ of the breast: a parallel immu- 180. Fisher B, Costantino JP, Wickerham DL, Redmond CK, Kavanah
nocytochemical and ultrastructural study. Ultrastruct Pathol. M, Cronin WM, et al. Tamoxifen for prevention of breast cancer:
1981;2(3):219–30. report of the National Surgical Adjuvant Breast and bowel project
162. Wang Y, Jindal S, Martel M, Wu Y, Schedin P, Troxell P-1 study. J Natl Cancer Inst. 1998;90(18):1371–88.
M. Myoepithelial cells in lobular carcinoma in situ: distribution 181. Cuzick J, Sestak I, Thorat MA. Impact of preventive therapy on
and immunophenotype. Hum Pathol. 2016;55:126–34. the risk of breast cancer among women with benign breast dis-
163. Sahoo S, Green I, Rosen PP. Bilateral Paget disease of the nipple ease. Breast. 2015;24(Suppl 2):S51–5.
associated with lobular carcinoma in situ. Arch Pathol Lab Med. 182. Bevers TB. Breast cancer risk reduction therapy: the low-hanging
2002;126(1):90–2. fruit. J Natl Compr Cancer Netw. 2015;13(4):376–8.
Papillary Lesions of the Breast
(IDP, IDPC, EPC, SPC) 5
Julia Y. Tsang, Ping Tang, and Gary M. Tse
from the same lactiferous duct opening. There are few, if

List of Frequently Asked Questions any, specific clinical features differentiating between benign
and malignant papillary lesions. Solid papilloma may be
1. What are papillary lesions [1]? more common in perimenopausal women and less common
with a palpable mass, whereas patients with papillary carci-
Papillary lesions encompass a broad spectrum of diseases, span-
noma are typically older and palpable masses are more
ning from benign intraductal papilloma (IDP) to malignant
frequent.
lesions, namely, intraductal papillary carcinoma (IDPC), encap-
sulated papillary carcinoma (EPC), and solid papillary carci- 3. What are the imaging findings of papillary lesions
noma (SPC). The characteristic pathologic features in all [2–6]?
papillary lesions are the intraductal epithelial proliferation
around fibrovascular cores arising from the ductal wall. As pap- The mammographic appearance of papilloma is usually a
illary lesions are derived from the ducts, a layer of intervening rounded or ovoid, well-circumscribed retroareolar mass,
myoepithelial cells is retained between the epithelial cell layer sometimes with ductal dilatation. Smaller lesions may be
and the fibrovascular cores, especially in the benign lesions. mammographically occult. Calcifications are uncommon.
However, it may be lost to various extents in atypical to malig- Ultrasonography (US) of papilloma typically shows solid
nant papillary lesions. The classification between different pap- mural nodule within a dilated duct. A cystic component may
illary lesions is dependent on the presence of cytological/ be seen, particularly in the larger, central lesions. It may be
architectural atypia and the presence of myoepithelial layers. hypervascular on color Doppler. Occasionally, a benign
The term “micropapillary” should not be confused with “papil- papilloma can possess speculated margins on imaging,
lary.” The former is composed of small, hollow, or morula-like mimicking a malignant disease. IDPC when detected in
clusters of cancer cells but not encasing true fibrovascular cores. imaging may show pleomorphic calcifications and architec-
tural distortion on mammography and ill-defined hypoechoic
2. What is the clinical presentation of papillary lesions? mass or calcifications on US. Other malignant papillary
lesions appear as oval circumscribed solitary or clustered
Clinically, papillary lesions may present with nipple dis-
masses that may be associated with microcalcifications on
charge, either blood stained or serosanguinous, originating
mammography. The lesions may be single or multiple, cir-
cumscribed solid or complex mixed cystic and solid masses
on US. They are vascular and tend to bleed centrally result-
J. Y. Tsang
ing in intracystic fluid-debris levels. Magnetic resonance
Department of Anatomical and Cellular Pathology, imaging (MRI) for papillary lesions may show variable
Chinese University of Hong Kong, Shatin, Hong Kong enhancement patterns. It has a high sensitivity, but low spec-
e-mail: [email protected] ificity, and is useful in establishing the extent and distribu-
P. Tang tion of multiple lesions. A nonparallel orientation, echogenic
Department of Pathology and Laboratory Medicine, halo, posterior acoustic enhancement, and associated calci-
Loyola University Medical Center, Maywood, IL, USA
e-mail: [email protected] fication are reported to be more frequent in malignant
G. M. Tse (*)
lesions. However, due to overlapping findings, imaging is
Department of Anatomical and Cellular Pathology, unreliable in differentiating between benign and malignant
Prince of Wales Hospital, Shatin, Hong Kong papillary lesions.

146 J. Y. Tsang et al.
4. What are the typical gross characteristics of papillary attenuated. The epithelial component may consist of one
lesions [7]? layer of cuboidal/columnar cells or show different degrees of
hyperplasia, particularly florid epithelial hyperplasia or
Papillary lesions have variable sizes, ranging from a few milli- metaplasia (apocrine, squamous). As the lesion is intraductal
meters to several centimeters. Palpable IDPs vary from soft to in origin, there is also an intact layer of myoepithelial cells
firm with dense sclerotic foci. Focal areas of hemorrhage and around the surrounding ductal wall. Identification of myo-
necrosis are common particularly in larger lesions. Careful gross epithelial cells in both compartments helps to confirm the
examination may show the lesions lying within dilated ducts. diagnosis. The fibrovascular cores are generally considered
There are no specific macroscopic features for IDPC. For EPC, a to be broad and fibrous (compared to IDPCs, which possess
friable mass within a cystic cavity may be appreciated. SPC may long and slender fibrovascular cores).
be observed as a whitish gray or yellowish brown fleshy firm or
soft nodular circumscribed mass on gross examination. 6. What are the benign changes that can be seen in IDP
[8, 9]? (Fig. 5.2a, b)
5. What are the histologic features of intraductal papil-
loma [1]? (Fig. 5.1a, b) The epithelial layer of IDP may show different degrees of
hyperplasia, particularly florid epithelial hyperplasia or meta-
IDPs are characterized by a cohesive but arborescent struc- plasia (apocrine, squamous). Apocrine change and squamous
ture composed of fibrovascular cores covered by myoepithe- metaplasia are mostly associated with areas of infraction.
lial cells and epithelial cells. The myoepithelial layer can be Very infrequently, mucinous clear cell changes and sebaceous
a b
Fig. 5.1 IDP. Intraductal papilloma with HE stain (a) and CK5/6 stain (b)
a b
Fig. 5.2 IDP with sclerotic changes. Intraductal papilloma with sclerosis, low-power view (a) and high-power view at the periphery with “infiltra-
tive glands” (b)
5 Papillary Lesions of the Breast (IDP, IDPC, EPC, SPC) 147
metaplasia may be seen. Other changes that can occur in IDPs 8. When is a diagnosis of IDP with atypia or DCIS made
include hemorrhage, infarction, and stromal fibrosis. [1, 11]?
Hemorrhage or infraction may occur secondary to a needling
procedure or due to torsion of fibrovascular cores. The The latest recommendation from WHO on the diagnosis for
changes can be very extensive so that the papillary architec- IDP with atypia is based on the extent of atypical epithelial
ture may be obscured, and the terms “sclerosed papillomas” proliferation. IDP with atypia is diagnosed when the atypical
or “ductal adenomas” are coined in these extreme cases. epithelial proliferation is <3 mm, while IDP with DCIS will
be diagnosed when this atypical population is greater than
7. What is atypical papilloma [1, 10, 11]? (Fig. 5.3a–d) 3 mm or high grade. (See also question 7.)
IDPs can display focal atypical epithelial proliferation com- 9. How can one differentiate IDP from IDPC [12–21]?
posed of a monotonous cell population. In earlier studies, (Fig. 5.4a–f)
atypical papilloma has been used to describe papilloma with
10 to <30% low-grade atypical epithelial hyperplasia, but this IDPC shows ducts and/or TDLU filled with slender, branch-
terminology is not in current use. The current recommended ing fibrovascular stalks covered by a monotonous neoplastic
terminology is IDP with atypical duct hyperplasia (ADH) or epithelial cell population. In contrast, IDP is composed of
ductal carcinoma in situ (DCIS). For the diagnosis, the WHO benign epithelial cells showing more heterogeneous mor-
Work Group recommends size-based criteria. A lesion with phology with variable degree of intraductal hyperplasia. The
low-grade atypical hyperplasia <3 mm in the greatest dimen- papillae in IDPC are typically thin and delicate with minimal
sion will be classified as papilloma with ADH. A diagnosis of stromal fibrosis when compared with the board blunt fronds
DCIS involving a papilloma will be made if the proliferation of IDP. The neoplastic epithelial cells in IDPC may form
is low grade and 3 mm or larger, or is high grade. micropapillary, cribriform, or solid structures obliterating
a b
c d
Fig. 5.3 IDP with atypia. Intraductal papilloma with ductal hyperplasia (a) and focally atypical with cribriform pattern (b). CK5/6 stain, low-
power view (c) and high-power view for the ADH area (d)
a b
c d
e f
Fig. 5.4 IDPC. Intraductal papillary carcinoma low power (a). High-power view shows low-grade monotonous tumor cells with cribriform pat-
tern (b). IHC stains for p63 (c), CK5 (d), calponin (e), and ER (f)
the spaces between papillary fronds. The complete or near mostly absent in the papillary processes, they may be present
complete (90%) absence of myoepithelial cells within the in attenuated form at the periphery of ducts. Nonetheless, the
intraluminal papillary fronds is one of the histologic features presence of myoepithelial cells, even extensively, does not
to distinguish IDPC from IDP. While myoepithelial cells are necessarily exclude the possibility of an IDPC.
Immunohistochemistry can assist in the differentiation of ferentiation is the presence of a thick fibrous capsule around
IDP and IDPC. Myoepithelial markers – such as p63, smooth the tumor in EPC, and the peripheral layer of myoepithelial
muscle actin, CD10, calponin, or high molecular weight cyto- cells may also be absent. In IDPC, a peripheral layer of myo-
keratins (HMWCK) – can be used. The absence of myoepithe- epithelial cells is almost always present, and a thick fibrous
lial markers along the papillary fronds and the presence of capsule is lacking. SPC has a distinct morphology compared
myoepithelial cells at the periphery of the involved ducts will to IDPC and EPC. The predominant architecture of SPC is
identify IDPC. Of note, some IDPCs may show a dimorphic solid, with the neoplastic epithelial cells filling up the inter-
cell population, with some of the neoplastic cells showing clear vening spaces between the fibrovascular cores within the
cytoplasm adjacent to the basement membrane, and these may lesion, resulting in a solid overall appearance. Cribriform or
be mistaken for myoepithelial cells. In these cases, staining of discrete papillary architectures are not apparent in low power.
myoepithelial marker will be particularly essential to illustrate The delicate underlying fibrovascular stromal network devoid
the absence of myoepithelial cells. Since in IDPCs the neoplas- of ME layer is typically discernible only at higher magnifica-
tic epithelial cells show clonal changes, they show diffuse and tion, and, even so, the fibrovascular cores tend to be fewer
strong epithelial expression of ER/PR and absence of epithelial than in IDPC and EPC. In SPC, the lesional neoplastic cells
HMWCK expression. Conversely, IDP will show positivity of may show streaming and occasionally spindled appearance
epithelial HMWCK and heterogeneous epithelial staining of which can be mistaken for intraductal papilloma with florid
ER/PR. As no single marker demonstrates entirely satisfactory ductal hyperplasia. Indeed, many SPCs display neuroendo-
performance in differentiating between these entities, variable crine differentiation, often have at least focal granular eosino-
marker combinations have been proposed to increase specific- philic cytoplasm, and are immunoreactive for neuroendocrine
ity and sensitivity. ER and MUC3 were shown to specifically markers. They may be associated with intracellular and extra-
identify IDPC, but CK5/6 and p63 were better suited for cellular mucin. As with EPC, myoepithelial cells may be lost
IDP. Other combinations that have been proposed to differenti- both inside and at the periphery of SPC.
ate between benign and malignant papillary lesions in general
include: CK5/6 and ER for atypical papillary lesions; CK5/6, 11. What are the different IHC myoepithelial staining
p63, and neuroendocrine markers for SPC; a CK5/CK8/18/p63 characteristics of IDPC, EPC, SPC, and DCIS involv-
cocktail; or a combination of HMWCK, namely, CK5/6, CK14, ing papilloma?
and 34βE12, for IDPC and EPC.
Papillary lesions can be differentiated by evaluation of myo-
10. How can one differentiate IDPC, EPC, and SPC epithelial markers staining in the two different compart-
[7, 22, 23]? (Figs. 5.5a, b and 5.6a–f) ments of the lesions, i.e., myoepithelial cells at the periphery
of the tumor, and around the fibrovascular cores within the
Compared to IDPC, EPC tends to show more extensive tumor. All papillary lesions (IDP and IDPC) are derived
involvement by the neoplastic epithelial cells, and the lesion from ducts, and myoepithelial cells are present at the periph-
is typically larger. The most important feature for their dif- ery. They are absent or markedly attenuated around the
a b
Fig. 5.5 EPC. Encapsulated papillary carcinoma (a) and ADH5 stains showing lack of myoepithelial cells both at the peripheral and fibrovascular
cores (b)
a b
c d
e f
Fig. 5.6 SPC. Solid papillary carcinoma (a, d) with myoepithelial marker p63 (b, e) and CK5 (c, f) IHC stains
fibrovascular cores within IDPC. In DCIS involving papil- 12. How is invasion defined in IDPC, EPC, and SPC
loma, myoepithelial cells can be found throughout the [7, 14, 17, 24–31]?
lesion. In EPC and SPC, there may be a lack of an outer
myoepithelial cell layer, and myoepithelial cells can be lost, IDPC is generally considered a carcinoma in situ, as evi-
attenuated, or retained around the fibrovascular core within denced by the presence of myoepithelial cells in the periph-
the lesion. See Table 5.1. ery of the ducts containing the IDPC. Occasionally, the
Table 5.1 ME marker staining on different papillary compartments

ME marker staining IDPC EPC SPC DCIS involving papilloma
Periphery Present Absent Absent/present Present in papilloma
Around fibrovascular cores Absent/scant Absent/attenuated/retained Absent/attenuated/retained Present in papilloma
presence of glandular entrapment within a hyalinized or X in 15–21% of the papillary carcinomas, but not in any of
fibrotic stroma and mechanical dislodgement during a previ- the IDPs. Also, IDPs with/without florid hyperplasia show
ous biopsy may mimic invasion. In addition to IDPC, these higher frequency in PIK3CA or AKT mutation compared to
may also occur in other types of papillary carcinomas. papillary carcinomas, suggesting a potentially divergent
To differentiate benign entrapment and true invasive foci, molecular pathway in pathogenesis.
identification of myoepithelial cells (albeit attenuated), the Genomic profiling of papillary carcinomas (including EPCs
bland morphology of the “infiltrative” epithelial cells, a hyalin- and SPCs) shows a similar pattern of gene copy number aber-
ized/altered stroma, and the general directional alignment rations as grade- and ER-matched infiltrating duct carcinomas,
between the epithelium and the fibrosis of the stroma are useful no special type (IDC-NSTs). However, papillary carcinomas
to confirm glandular entrapment rather than genuine invasion. display less genomic aberrations than IDC-NSTs and a higher
The lack of a myoepithelial cell layer in EPC and SPC has mutation rate in PI3KCA. The genomic profiles of the three
led to the postulation that these lesions represent a minimally morphologic papillary carcinomas (EPC, SPC, and IPC) are
invasive low-grade carcinoma rather than an in situ carci- remarkably similar. Regarding gene profiling, most papillary
noma. This postulation for EPC is further supported by evi- carcinomas are classified as luminal subtype, with rare basal-
dence of occasional lymph node and distant metastases. like subtype. In a study with limited cases, most EPCs and
Other evidence suggests EPC may represent an intermediate IDPCs are classified as luminal A cancers, while SPCs are clas-
form between DCIS and invasive duct carcinoma. EPCs with sified as luminal B cancers. Compared with grade- and
high-grade nuclear morphologic features and high mitotic ER-matched IDC-NSTs, papillary carcinomas show distinct
indexes are more often associated with an invasive compo- gene expression profiles with downregulation of proliferation-
nent. However, there are reports of continuous envelopment related, cell assembly and organization, cellular movement and
of basement membrane in EPC, arguing for an in situ rather migration genes, and overexpression of genes related to homeo-
than invasive carcinoma. So far, there is a lack of consensus stasis and angiogenesis, thus suggestive of a less invasive phe-
as to whether EPCs are invasive or in situ; thus, they are best notype. Apart from PI3KCA and TP53, highly recurrent gene
considered as in situ carcinomas to avoid overtreatment. fusions and recurrent mutations are unlikely to be involved in
In SPC, there may also be a lack of myoepithelial cells in papillary carcinomas.
the periphery of the tumor. In such cases, particularly when a
rounded configuration is maintained, the current thinking is 14. What are the further recommendations for manage-
to consider such as in situ carcinoma. SPC has a low but defi- ment of papillary lesion in biopsy? [38, 39]
nite risk of lymph node metastasis. Nevertheless, cases that
are purely “in situ” (without invasive component) have sig- The standard recommendations for patients with atypical/
nificantly better outcomes than other pure DCIS. Thus, most malignant papillary lesions are surgical consultation and
solid papillary carcinomas, when stringently defined and excision. However, for IDP the surgical management is more
excluding invasive component, represent carcinoma in situ controversial, as on the one hand there exists a small possi-
with favorable prognosis. bility of malignant upgrade in excision for core needle biopsy
In all papillary carcinomas (IDPC, EPC, and SPC), (CNB)-diagnosed benign papillary cases and, on the other
unequivocal invasion occurs when neoplastic elements infil- hand, excising all benign IDP may constitute overtreatment,
trate beyond the fibrous capsule, and the invasive component as a majority of CNB-diagnosed benign cases do not show
tends to show a geographic jigsaw pattern with ragged and any upgrade at excision. Some authors suggested surgical
irregular margin. The invasive component of SPC may show excision for cases with segmental abnormalities observed
mucinous tumor with neuroendocrine differentiation. radiologically because this may suggest coexisting DCIS. In
the absence of other indications for excision, observation
13. What are the underlying molecular changes associ- with close imaging follow-up has been suggested for radio-
ated with IDP and papillary carcinomas [32–37]? logically concordant IDP and incidental micropapillomas.
Both IDP and papillary carcinomas show LOH on chromo- 15. Papillary lesion in biopsy – upgrade in excision
some 16p13 in the TSC2 gene region, with similar frequen- [11, 17, 38, 40–43]
cies (60–63%), while LOH at locus 16q23 is limited to
papillary carcinomas. Comparing changes in chromosomes, A meta-analysis of 34 studies involving over 2000 patients
one study revealed alterations in chromosomes 3, 7, 17, and demonstrated that the pooled estimate for underestimation
was 15.7%. Other studies reported an upgrade rate for benign upgrade rates have also been associated with small core
papillary lesions on CNB diagnosis to atypia/malignancy in biopsies, with fewer cores being taken, when the pathologist
excision to range from 3% to 33%. Of note, most of the is not a subspecialty expert in breast pathology and when an
malignancies found in excision were low-grade DCIS rather immunohistochemical workup of the biopsy has not been
than invasive cancer. performed. Adequate sampling is a prerequisite in prevent-
The upgrade rate appears to be associated with tumor ing upgrade. Larger tissue samples significantly improved
topography. Up to 25% of DCIS associated with papilloma the predictive value of benign histology on CNB.
had demonstrable eccentric distribution after careful histo-
logic assessment and topographic reconstruction, and up to
83% of cases showed segmental abnormalities on radiology. Case Presentations
These constitute high-risk features for a small-volume CNB
and may result in sampling error, missing the malignant part of Case 1 Misdiagnosed Sclerotic IDP as IDC (Fig. 5.7a–d)
the lesions. Furthermore, by definition, underestimation of
DCIS involving papilloma in CNB is unavoidable. As the diag- History: A 45-year-old female with a mass lesion on screen-
nosis of DCIS involving papilloma is defined by the extent of ing mammogram, core needle biopsy, with a diagnosis of
the lesion (3 mm or more), in a small-volume CNB, sampling IDC NOS, histologic grade 1. You received this for review;
of only part of the atypical focus would mean sampling only a the patient is scheduled for surgery next Monday.
smaller focus, and then an atypical diagnosis would be ren- IHC: ER+, PR+, HER2 – (0), Ki67 <5%.
dered as the size criteria for DCIS may not be met. Higher DDx: IDC vs. sclerotic IP
a b
c d
Fig. 5.7 Case 1. IP with sclerosis. Sclerotic lesions in irregular border give a hint of a papillary lesion. Myoepithelial markers calponin (c) and
(a) and small infiltrative glands (b) can mimic invasive ductal carci- p63 (d) can help prove the noninvasive process
noma; however, the well-demarcated border and hyalinized stroma can
Next steps: History: A 50-year-old female with a breast mass, core

–– Call the surgeon to hold off on the for definitive diag- biopsy, diagnosed with atypical IDP. Surgical excision rec-
nosis or only do local resection without SLNB. ommended. You have received the surgical specimen.
–– Call the original lab for blocks to confirm your Histology: Solitary papillary lesion, with marked epithelial
suspicion. proliferation and sclerotic changes at the peripheral
–– IHC results: Myoepithelial cells were highlighted IHC: CK5 – positive for most of the cells, no area with com-
around the infiltrating glands. plete lack of it
Final diagnosis: sclerotic IDP P63 – only positive at the peripheral of the lesion
Take-home messages: ER – weakly positive in some cells
–– Be very cautious on core biopsy due to limited tissue Final diagnosis: Sclerotic IDP with florid ductal
sampled. hyperplasia
–– Be suspicious when seeing infiltrative glands within Take-home messages:
hyalinized stroma in a well-demarcated lesion. –– IDP with ADH has a specific diagnostic criteria, i.e.,
–– Be aware of clinical symptoms: near nipple, bloody <3 mm ADH within an IDP.
discharge, older patient. –– Sclerotic changes at the peripheral are often difficult to
differentiate from IDC; a panel of IHC is
recommended.
Case 2 Sclerotic IDP with Florid Ductal Hyperplasia –– Myoepithelial stains can be weak and attenuated.
(Fig. 5.8a–d)
a b
c d
Fig. 5.8 Case 2. IDP with florid ductal hyperplasia. Marked florid duc- for correct diagnosis, showing a nice mosaic staining pattern (c). Here,
tal hyperplasia in an IDP can be easily confused with SPC at both low- p63 is not as useful, as it only stains the peripheral myoepithelial cells
power (a) and high-power (b) views. IHC staining for CK5/6 is the key as it does for DCIS (d)
Case 3 SPC In Situ Lesion - Without Peripheral Case 4 SPC with Invasion - Estimation of Size
Myoepithelial Cells but with Prior Core Biopsy of the Invasion (Fig. 5.10a–d)
(Fig. 5.9a–e)
History: A 50-year-old female with a breast mass
History: A 50-year-old female with a mass lesion and core Histology:
needle biopsy –– Large nests of solid papillary lesions with focal irregu-
Histology: DCIS, solid type lar nests nearby.
IHC: ER+, PR+ –– The borders are largely smooth.
Surgical specimen, histology: –– No necrosis present.
–– Solid nests of tumor cells, low to intermediate grade. –– Few mitoses present.
–– One area shows smaller nest with irregular borders. IHC: ER+, PR+, HER2 (0) negative, Ki67 10%
–– Myoepithalial markers negative for the entire tumor. Next steps:
–– Most nests contain hidden fibrovascular cores. –– IHC for p63, calponin, and myosin are absent in all
Final diagnosis: SPC with biopsy site changes areas with tumor.
Take-home messages: Final diagnosis: Focal IDC associated with a large SPC
–– Neuroendocrine markers only positive in 50–60% of Take-home messages:
SPC, thus not good for diagnosis. –– SPC may or may not have the peripheral myoepithelial
–– Morphologic assessment and identification of fibro- cells; and is considered an in situ lesion.
vascular cores are most useful, but the latter may not –– True invasion is only diagnosed if the tumor cell
be obvious. clusters:
–– Glands trapped inside biopsy tract can be small, irreg- • Possess an irregular border.
ular, mimic IDC • Are not located within prior biopsy site.
• Lack myoepithelial cells.
a b
Fig. 5.9 Case 3. SPC with biopsy site changes. Roughly half of SPC laden macrophages, mimicking IDC (a); a higher-power view shows
lacks the peripheral myoepithelial cells, which makes the diagnosis of the smooth peripheral border of the tumor and confirms a diagnosis of
SPC with invasion very difficult purely based on morphology. A low- SPC without invasion (b). Although this tumor lacks myoepithelial
power view shows a very well-demarcated lesion, suggesting a SPC cells by ADH5 (c), calponin (d), and CK5 (e), it is still considered an
without invasion. The center of the lesion has a prior needle tract with SPC without invasion
smaller irregular and somewhat infiltrative glands and hemosiderin-
c d
a b
Fig. 5.10 Case 4. SPC with invasion. Although the larger tumor nests have smooth borders, the smaller ones in-between do not, suggesting inva-
sion (a); lack of p63 staining confirms the invasion (b) for the small tumor nests. Another area of this tumor with focal invasion is shown in (c, d)
c d
References 13. Tan PH, Aw MY, Yip G, Bay BH, Sii LH, Murugaya S, et al.
Cytokeratins in papillary lesions of the breast: is there a role in dis-
tinguishing intraductal papilloma from papillary ductal carcinoma
1. Lakhani SR, Ellis IO, Schnitt SJ, Tan PH, van de Vijver MJ, editors. in situ? Am J Surg Pathol. 2005;29(5):625–32.
World Health Organisation classification of tumors of the breast. 14. Collins LC, Carlo VP, Hwang H, Barry TS, Gown AM, Schnitt
4th ed. Lyon: IARC Press; 2012. SJ. Intracystic papillary carcinomas of the breast: a reevaluation
2. Eiada R, Chong J, Kulkarni S, Goldberg F, Muradali D. Papillary using a panel of myoepithelial cell markers. Am J Surg Pathol.
lesions of the breast: MRI, ultrasound, and mammographic appear- 2006;30(8):1002–7.
ances. AJR Am J Roentgenol. 2012;198(2):264–71. 15. Tse GM, Tan PH, Moriya T. The role of immunohistochemistry in
3. Muttarak M, Lerttumnongtum P, Chaiwun B, Peh WC. Spectrum the differential diagnosis of papillary lesions of the breast. J Clin
of papillary lesions of the breast: clinical, imaging, and pathologic Pathol. 2009;62(5):407–13.
correlation. AJR Am J Roentgenol. 2008;191(3):700–7. 16. Lefkowitz M, Lefkowitz W, Wargotz ES. Intraductal (intracystic)
4. Brookes MJ, Bourke AG. Radiological appearances of papillary papillary carcinoma of the breast and its variants: a clinicopatho-
breast lesions. Clin Radiol. 2008;63(11):1265–73. logical study of 77 cases. Hum Pathol. 1994;25(8):802–9.
5. Kim TH, Kang DK, Kim SY, Lee EJ, Jung YS, Yim H. Sonographic 17. Ni YB, Tse GM. Pathological criteria and practical issues in papillary
differentiation of benign and malignant papillary lesions of the lesions of the breast – a review. Histopathology. 2016;68(1):22–32.
breast. J Ultrasound Med. 2008;27(1):75–82. 18. Furuya C, Kawano H, Yamanouchi T, Oga A, Ueda J, Takahashi
6. Lam WW, Chu WC, Tang AP, Tse G, Ma TK. Role of radiologic M. Combined evaluation of CK5/6, ER, p63, and MUC3 for distin-
features in the management of papillary lesions of the breast. AJR guishing breast intraductal papilloma from ductal carcinoma in situ.
Am J Roentgenol. 2006;186(5):1322–7. Pathol Int. 2012;62(6):381–90.
7. Maluf HM, Koerner FC. Solid papillary carcinoma of the breast. A 19. Grin A, O’Malley FP, Mulligan AM. Cytokeratin 5 and estrogen
form of intraductal carcinoma with endocrine differentiation fre- receptor immunohistochemistry as a useful adjunct in identifying
quently associated with mucinous carcinoma. Am J Surg Pathol. atypical papillary lesions on breast needle core biopsy. Am J Surg
1995;19(11):1237–44. Pathol. 2009;33(11):1615–23.
8. Flint A, Oberman HA. Infarction and squamous metaplasia of intra- 20. Reisenbichler ES, Balmer NN, Adams AL, Pfeifer JD, Hameed
ductal papilloma: a benign breast lesion that may simulate carci- O. Luminal cytokeratin expression profiles of breast papillomas
noma. Hum Pathol. 1984;15(8):764–7. and papillary carcinomas and the utility of a cytokeratin 5/p63/
9. Jiao YF, Nakamura S, Oikawa T, Sugai T, Uesugi N. Sebaceous cytokeratin 8/18 antibody cocktail in their distinction. Mod Pathol.
gland metaplasia in intraductal papilloma of the breast. Virchows 2011;24(2):185–93.
Arch. 2001;438(5):505–8. 21. Tse GM, Ni YB, Tsang JY, Shao MM, Huang YH, Luo MH, et al.
10. MacGrogan G, Tavassoli FA. Central atypical papillomas of the Immunohistochemistry in the diagnosis of papillary lesions of the
breast: a clinicopathological study of 119 cases. Virchows Arch. breast. Histopathology. 2014;65(6):839–53.
2003;443(5):609–17. 22. Rabban JT, Koerner FC, Lerwill MF. Solid papillary ductal carci-
11. Page DL, Salhany KE, Jensen RA, Dupont WD. Subsequent breast noma in situ versus usual ductal hyperplasia in the breast: a poten-
carcinoma risk after biopsy with atypia in a breast papilloma. tially difficult distinction resolved by cytokeratin 5/6. Hum Pathol.
Cancer. 1996;78(2):258–66. 2006;37(7):787–93.
12. Moritani S, Ichihara S, Kushima R, Okabe H, Bamba M, Kobayashi 23. Nicolas MM, Wu Y, Middleton LP, Gilcrease MZ. Loss of myo-
TK, et al. Myoepithelial cells in solid variant of intraductal papillary epithelium is variable in solid papillary carcinoma of the breast.
carcinoma of the breast: a potential diagnostic pitfall and a pro- Histopathology. 2007;51(5):657–65.
posal of an immunohistochemical panel in the differential diagnosis 24. Nagi C, Bleiweiss I, Jaffer S. Epithelial displacement in breast
with intraductal papilloma with usual ductal hyperplasia. Virchows lesions: a papillary phenomenon. Arch Pathol Lab Med.
Arch. 2007;450(5):539–47. 2005;129(11):1465–9.
25. Hill CB, Yeh IT. Myoepithelial cell staining patterns of papillary 35. Troxell ML, Levine J, Beadling C, Warrick A, Dunlap J, Presnell
breast lesions: from intraductal papillomas to invasive papillary A, et al. High prevalence of PIK3CA/AKT pathway mutations in
carcinomas. Am J Clin Pathol. 2005;123(1):36–44. papillary neoplasms of the breast. Mod Pathol. 2010;23(1):27–37.
26. Mulligan AM, O’Malley FP. Metastatic potential of encapsulated 36. Duprez R, Wilkerson PM, Lacroix-Triki M, Lambros MB,

(intracystic) papillary carcinoma of the breast: a report of 2 cases MacKay A, A’Hern R, et al. Immunophenotypic and genomic
with axillary lymph node micrometastases. Int J Surg Pathol. characterization of papillary carcinomas of the breast. J Pathol.
2007;15(2):143–7. 2012;226(3):427–41.
27. Solorzano CC, Middleton LP, Hunt KK, Mirza N, Meric F, 37. Piscuoglio S, Ng CK, Martelotto LG, Eberle CA, Cowell CF,

Kuerer HM, et al. Treatment and outcome of patients with Natrajan R, et al. Integrative genomic and transcriptomic char-
intracystic papillary carcinoma of the breast. Am J Surg. acterization of papillary carcinomas of the breast. Mol Oncol.
2002;184(4):364–8. 2014;8(8):1588–602.
28. Rakha EA, Tun M, Junainah E, Ellis IO, Green A. Encapsulated 38. Moritani S, Ichihara S, Hasegawa M, Endo T, Oiwa M, Shiraiwa M,
papillary carcinoma of the breast: a study of invasion associated et al. Uniqueness of ductal carcinoma in situ of the breast concur-
markers. J Clin Pathol. 2012;65(8):710–4. rent with papilloma: implications from a detailed topographical and
29. Esposito NN, Dabbs DJ, Bhargava R. Are encapsulated papillary histopathological study of 50 cases treated by mastectomy and wide
carcinomas of the breast in situ or invasive? A basement membrane local excision. Histopathology. 2013;63(3):407–17.
study of 27 cases. Am J Clin Pathol. 2009;131(2):228–42. 39. Calhoun BC, Collins LC. Recommendations for excision following
30. Rakha EA, Gandhi N, Climent F, van Deurzen CH, Haider SA, core needle biopsy of the breast: a contemporary evaluation of the
Dunk L, et al. Encapsulated papillary carcinoma of the breast: literature. Histopathology. 2016;68(1):138–51.
an invasive tumor with excellent prognosis. Am J Surg Pathol. 40. Wen X, Cheng W. Nonmalignant breast papillary lesions at core-
2011;35(8):1093–103. needle biopsy: a meta-analysis of underestimation and influencing
31. Nassar H, Qureshi H, Adsay NV, Visscher D. Clinicopathologic factors. Ann Surg Oncol. 2013;20(1):94–101.
analysis of solid papillary carcinoma of the breast and associated 41. Armes JE, Galbraith C, Gray J, Taylor K. The outcome of papillary
invasive carcinomas. Am J Surg Pathol. 2006;30(4):501–7. lesions of the breast diagnosed by standard core needle biopsy within
32. Lininger RA, Park WS, Man YG, Pham T, MacGrogan G, Zhuang a BreastScreen Australia service. Pathology. 2017;49(3):267–70.
Z, et al. LOH at 16p13 is a novel chromosomal alteration detected 42. Koo JS, Han K, Kim MJ, Moon HJ, Kim EK, Park BW. Can
in benign and malignant microdissected papillary neoplasms of the additional immunohistochemistry staining replace the surgical
breast. Hum Pathol. 1998;29(10):1113–8. excision for the diagnosis of papillary breast lesions classified as
33. Di Cristofano C, Mrad K, Zavaglia K, Bertacca G, Aretini P,
benign on 14-gage core needle biopsy? Breast Cancer Res Treat.
Cipollini G, et al. Papillary lesions of the breast: a molecular pro- 2013;137(3):797–806.
gression? Breast Cancer Res Treat. 2005;90(1):71–6. 43. Shamonki J, Chung A, Huynh KT, Sim MS, Kinnaird M, Giuliano
34. Tsuda H, Takarabe T, Inazawa J, Hirohashi S. Detection of
A. Management of papillary lesions of the breast: can larger core
numerical alterations of chromosomes 3, 7, 17 and X in low-grade needle biopsy samples identify patients who may avoid surgical
Intracystic papillary tumors of the breast by multi-color fluores- excision? Ann Surg Oncol. 2013;20(13):4137–44.
cence in situ hybridization. Breast Cancer. 1997;4(4):247–52.
Fibroepithelial Lesions (Phyllodes
Tumor and Fibroadenoma) of the Breast 6
Julia Y. Tsang and Gary M. Tse
List of Frequently Asked Questions have overlapping histologic features with benign phyllodes
tumors (PTs).
Fibroadenoma: Morphological Variants An FA with prominent stromal myxoid changes throughout
is described as a myxoid FA. Myxoid FAs can be associated
1. What is the morphological spectrum of fibroade- with Carney syndrome, as such they are commonly multicen-
noma [1]? tric or bilateral while sporadic cases are more often solitary.
Hyalinized FAs may show prominent stromal sclerosis and
Fibroadenomas (FAs) are well-circumscribed biphasic hyalinization, with or without associated calcifications. These
lesions showing proliferation of stroma and epithelium aris- FAs tend to be distinctly hypocellular and are especially com-
ing from the terminal duct lobular unit. In general, the stro- mon in older women, presenting as long-standing static or
mal component may show focal or diffuse hypercellularity regressing breast masses. Popcorn calcifications can be seen in
(especially in young women <20 years), bizarre, multinucle- this type of FA.
ated giant cells, extensive myxoid changes, or hyalinization
with dystrophic calcification (popcorn calcification) and 2. What are the diagnostic criteria of complex fibroade-
rarely ossification (especially in postmenopausal women). nomas and the clinical implications [2–4]?
There are two distinct growth patterns: pericanalicular
and intracanalicular patterns, which are of no clinical signifi- Complex FAs contain one or more of the followings: cyst of
cance. In pericanalicular pattern, stromal cells proliferate or greater than 3 mm, sclerosing adenosis, papillary apocrine
around ducts or sometimes lobules in a circumferential fash- hyperplasia, or epithelial calcification. Among these, the
ion without epithelial compression. Open lumens can be presence of cysts and sclerosing adenosis are the most com-
observed. This pattern is more frequent in women in their mon. Proliferative epithelial changes may occur more often
second or third decades. In intracanalicular pattern, stromal adjacent to complex rather than noncomplex FAs. Complex
proliferation compresses ducts into clefts. Any FAs may FAs are associated with older age of patients and tend to be
show either or both patterns. smaller than noncomplex FAs. A minimally higher risk of
Apart from the growth pattern, the stromal composition of subsequent invasive cancers had been found in patients with
FA may show variations. The typical FA has even distribu- complex FA compared to those with noncomplex FA. Such
tion of epithelium and stroma which is consistent throughout risk remained elevated for decades after diagnosis.
the lesion. Prominent cellular stroma, either diffusely or
focally, may be encountered, especially for FA in younger 3. What are the criteria for juvenile fibroadenoma?
women. These FAs are called cellular FAs. Cellular FAs may
Juvenile FAs are characterized microscopically by mildly
J. Y. Tsang increased stromal cellularity with a fascicular stromal
Department of Anatomical and Cellular Pathology, arrangement and epithelial hyperplasia. The architecture is
Chinese University of Hong Kong, Shatin, Hong Kong
often pericanalicular rather than intracanalicular or mixed.
Usual ductal hyperplasia in the epithelium often features
G. M. Tse (*)
delicate micropapillary epithelial projections with epithelial
Department of Anatomical and Cellular Pathology,
Prince of Wales Hospital, Shatin, Hong Kong hyperplasia with bulbous tips, referred to as “gynecomastoid”-
e-mail: [email protected] like due to its resemblance to epithelial changes in

160 J. Y. Tsang and G. M. Tse
g ynecomastia. Stromal cellularity is typically evenly distrib- 7. What are the diagnostic criteria for benign phyllodes
uted throughout the lesion. However, mild intralesional het- tumor [12, 16–18]?
erogeneity may occasionally be observed. There is no
stromal cytologic atypia. Mitotic activity can be seen in some Benign PTs are well circumscribed and show mildly
adolescents but is mostly rare. increased stromal cellularity as compared with an FA with
no to mild nuclear atypia. Mitoses are rare, ranging from 0 to
4. How many mitoses are allowed in fibroadenoma [5–8]? 4 mitotic figures/10 HPF. Stromal overgrowth (defined as the
presence of stroma without epithelium in at least one low-
Stromal mitoses are mostly absent in the majority of common power field as observed with a × 4 microscope objective) is
FAs. However, widely scattered mitotic figures may be seen absent. The margins are usually well delimited and pushing,
in some cases, especially in young patients or during preg- although very small tumor buds may protrude into the sur-
nancy. Mitotic activity (up to 7 mitoses/10 HPF) has been rounding tissues. Such protrusive expansion may be left
described in juvenile FA. A diagnostic threshold for mitoses behind after surgical removal and is a source of local recur-
in FA has not been established. In general, less than 3 per 10 rence. Heterologous elements are absent. Benign lipoma-
HPF is applied in most studies. Nonetheless, a benign diagno- tous, cartilaginous, and osseous metaplasia have been
sis also requires support from other histologic and clinico- reported. Benign PTs are more likely to demonstrate pseudo-
pathologic features, such as a circumscribed border, lack of angiomatous stromal hyperplasia (PASH), multinucleated
cytologic atypia, stromal overgrowth, or necrosis. stromal giant cells, and epithelial hyperplasia. See Figs. 6.1,
6.2, and 6.3.
5. What are the pregnancy changes in fibroadenomas [9]?
8. What are the diagnostic criteria for malignant phyl-
During pregnancy, the high concentration of estrogen, pro- lodes tumor [16]?
gesterone, and prolactin promotes the growth of ducts, and
the formation of tubuloalveolar structures could lead to sig- Malignant PTs showed marked stromal cellularity and
nificant enlargement of an FA. The microscopic changes cor- nuclear atypia. Mitotic activity is usually abundant (at least
respond to the parenchymal hyperplasia seen in the 10 per 10 HPF). The tumor border is at least focally infiltra-
surrounding normal breast tissue. Secretory hyperplasia tive. Stromal overgrowth is often present. Owing to over-
sometimes occurs diffusely in FA during pregnancy, or a pre- growth of sarcomatous components, the epithelial component
existing FA may be unaltered. Epithelial proliferation may may only be identified after examining multiple sections
be extensive enough to obscure the original morphology or it with diligent sampling of the tumors. Malignant tumors are
may be almost or completely absent. Usually the distribution also diagnosed when malignant heterologous elements
of the change is very irregular, and large areas of the FA may (pleomorpgic liposarcoma, chondrosarcoma, and osteosar-
be apparently unmodified by the pregnancy-driven hyperpla- coma) are present even in the absence of other features.
sia. Multiple FAs in the same breast may have different
degrees of response. In general, unless the FAs are hyalin-
ized or of long standing, some degree of change is always
seen during pregnancy.
Phyllodes Tumor: Morphological Variants
6. What are the morphological variants in phyllodes

tumor [10–15]?
There are uncommon morphologic variants in phyllodes

tumors (PTs). The stromal component may show cartilagi-
nous, osteoid, or lipomatous metaplasia. The epithelial com-
ponent may show apocrine or squamous metaplasia.
Squamous metaplasia may occur in PTs of different grades,
Fig. 6.1 Benign phyllodes tumor showing expansion of the stroma,
with a reported incidence of 3.6–10%. Rarely, ductal or lobu- with a variability of the stromal cellularity, and condensation of the
lar carcinoma in situ or invasive carcinoma may arise in the stromal cells in a subepithelial location, whereas the intervening areas
epithelial component. retain a low cellularity. The ductal element may be tubular
6 Fibroepithelial Lesions (Phyllodes Tumor and Fibroadenoma) of the Breast 161
10. Can fibroadenomas or phyllodes tumors be associ-

ated with other breast diseases [4, 12, 14, 19–23]?
FA and PT may display a wide spectrum of proliferative

changes in their epithelium. Usual ductal hyperplasia was
reported in 43.7% of FAs, with most demonstrating a moder-
ate degree of hyperplasia. Atypical epithelial proliferations,
including atypical ductal or lobular hyperplasia and ductal or
lobular carcinoma in situ, have also been observed infre-
quently in FA. The rate of carcinoma associated with FA
ranges from 0.01% to 0.3% in screened populations. There is
a predominance of carcinomas in situ (over 95%) than inva-
sive carcinomas. Similar to FA, usual ductal hyperplasia is
also commonly seen in PT (up to 74%). Benign PTs showed
Fig. 6.2 Higher magnification of a benign phyllodes tumor showing more hyperplasia than borderline or malignant tumors. Other
increased stromal cellularity, and the stromal cells exhibit bland mor-
phological features devoid of significant atypism. The epithelial com- atypical epithelial proliferations may also be found rarely. In
ponent shows hyperplasia with stratification of the nuclei. There is a large series, there was 1.5% of PT with ADH and 0.03%
again no atypism each with DCIS, ALH, and LCIS. There are case reports
showing association of PT with invasive cancers.
Phyllodes Tumor: Core Needle Biopsy
11. Can we diagnose phyllodes tumor on needle core

biopsy [24–27]?
Diagnosis of PT on core needle biopsy (CNB) can be diffi-

cult. The false-negative rate ranged from 8% to 39% and
false-positive rate ranged from 0% to 17%, although a sensi-
tivity of 83% and a specificity of 92% were reported by oth-
ers. With increasing degree of malignancy, a definitive
diagnosis of PT on CNB can be made with more confidence.
In a reported series with PTs in excision, a diagnosis of PT in
the corresponding CNB was made in 44%, 67%, and 83% of
Fig. 6.3 Epithelial hyperplasia is common in all grades of phyllodes benign, borderline, and malignant PT, respectively. None of
tumors. In this example, there is prominent epithelial hyperplasia in
a benign phyllodes tumor. The hyperplastic epithelium shows typi-
the malignant PTs were diagnosed incorrectly. Only one case
cal florid epithelial hyperplasia changes, devoid of atypical of malignant PT was put in the less specific diagnosis of FEL
architecture and all others were correctly diagnosed in CNB.
12. How should we work up a core needle biopsy with

hypercellular and pleomorphic spindle cells based on
9. What are the diagnostic criteria for borderline phyl- H&E morphology?
lodes tumor [16]?
The main differential diagnoses for core needle biopsy with
Borderline PTs are diagnosed when the tumors possess some hypercellular and pleomorphic spindle cells are malignant
but not all features of malignancy. The classification is arbi- PT, metaplastic CA, and sarcoma. As PT and some metaplas-
trary, and borderline PTs have varied morphological appear- tic carcinoma possess a biphasic pattern, the presence of
ances, overlapping with malignant and benign PTs. In both malignant spindle cells and epithelial cells indicates
general, they show well-defined or focally invasive border, metaplastic carcinoma. Benign epithelial component points
frequent mitoses (5–9/10 HPF), moderate stromal cellularity, to a diagnosis of malignant PT, and a malignant epithelial
and mild to moderate stromal atypia. Stromal overgrowth is component (carcinoma, carcinoma in situ, squamous cell
usually absent or very focal. carcinoma) suggests metaplastic carcinoma. Sometimes the
epithelial component is much attenuated and requires exten- present for PT diagnosis on CNB, including stromal over-
sive sampling. The presence of single malignant spindle cell growth, increased stromal cellularity, mitotic activity, stro-
component indicates sarcoma or some metaplastic carci- mal fragmentation, adipose tissue infiltration or fat
noma. If the spindle cells exhibit cytokeratin expression or entrapment, stromal heterogeneity, and stromal cell pleo-
basal marker expression, the diagnosis should be in favor of morphism. The presence of any three or more of these fea-
metaplastic carcinoma. The presence of malignant heterolo- tures showed a sensitivity of PT detection of 85.2%.
gous elements is NOT useful for the differential diagnosis, as
all the lesions considered (malignant PT, metaplastic carci- 15. Can one diagnose phyllodes tumor in fine needle
noma, and sarcoma) may possess such heterologous element. aspiration cytology [31–38]?
(For more details, see the section on differential malignant
PT, questions 25 and 26.) PTs are characterized by heterogeneous distribution of stro-
mal cellularity, atypia, secondary epithelial component, and
13. What are the features in core needle biopsy that help frond-like architecture. In the absence of the architectural
to differentiate between phyllodes tumor and fibro- details and limited sampling in fine needle aspiration cytol-
adenoma [6, 8, 27–29]? ogy (FNAC), it is difficult to adopt universally accepted cyto-
logic criteria for accurate diagnosis of PT. A wide range of
Many studies identified increased stromal cellularity and/or accuracy rate (25–70%) was reported, and FA was the most
mitotic activity as useful for the differential diagnosis of PT common erroneous diagnosis. Occasionally, the presence of
and FA. Marked stromal cellularity is found in PTs, but FAs epithelial hyperplastic changes led to a misdiagnosis of carci-
have lower cellularity. Likewise, higher mitotic counts are noma. The cytologic differentiation of PT from FA is often
observed in PTs than FAs. However, in PT, there may be not possible. Some features, when present, may be more sug-
areas of variable stromal cellularity; in the low stromal cel- gestive of PT; these include cellular stromal fragments;
lularity area, there may be no histomorphologic difference enrichment of elongated spindled nuclei in over 30% of the
from FA. Hence, there may be broad overlap of these fea- sample; large (>1 mm), elongated, and wavy epithelial clus-
tures between PT and FA, and diagnosis based on these fea- ters (compared to smaller tubular or blunt branching clusters
tures solely is difficult. In addition, some studies also in FA); and increased stromal epithelial area ratio. Accuracy
suggested tissue fragmentation and adipose tissue within of FNAC depends on an adequate and representative sample.
lesional stroma in CNB of PT. However, one should note that
FA with prominent intracanalicular pattern may also frag-
ment with possible frond-like features in CNB; also FA Phyllodes Tumor Grading
rarely may show adipose metaplasia. Others also cited stro-
mal hypercellularity, stromal nuclear atypia, stromal over- 16. In general, what are the relative proportions of benign,
growth, the presence of >2 mitotic figures per 10 HPF, and borderline, and malignant phyllodes tumors [16, 39]?
the presence of pseudoangiomatous stromal hyperplasia
(PASH), to be predictive of PT over FA. Given the overlap- In general, benign, borderline, and malignant PTs comprise
ping histologic features between PT and FA, a combination 60–70%, 15–20%, and 10–20%, respectively, of all PTs.
of these histologic features could be helpful for diagnosing However, there is wide range of variation in the relative pro-
PT over FA, but no hard and fast rules can be laid down. portion of each grade in different series, indicating the cur-
rent practice of using a combination of histologic features in
14. How accurate is core needle biopsy in diagnosing PT grading, resulting in a large “gray zone” where the PTs
benign phyllodes tumor [8, 29, 30]? do not exhibit all features of benign or malignant PT and are
thus relegated to the borderline category.
Benign PT can usually be differentiated from FA with typical
features, such as absence of cellular stroma and mitotic 17. What are the histologic criteria for grading of phyl-
activity on CNB. However, for lesions showing atypical fea- lodes tumor [13, 16, 40, 41]?
tures, when a clear differentiation of benign PT and FA is not
possible, a generic diagnosis of fibroepithelial lesion is pru- Grading of PT is based on semiquantitative assessment of
dent. Several studies evaluated features in CNB to predict PT stromal cellularity, cellular pleomorphism, mitotic activity,
on excision. One study showed features including stromal margin/border, and stromal overgrowth, rather than any iso-
cellularity, stromal overgrowth, fragmentation, and adipose lated feature. The presence of malignant heterologous ele-
tissue within the lesion could correctly identify 90% of PTs, ment is also relevant and, if present, will lead to diagnosis of
but 31% of FAs were incorrectly diagnosed. Another malignant PT irrespective of other parameters. The histo-
approach took into account the number of histologic features logic grading of PT is subjective. The assessment of most of
Fig. 6.4 Phyllodes tumor of borderline malignancy showing expan- Fig. 6.5 Malignant phyllodes tumor showing very high stromal cellu-
sion of the stroma with spacing out of the epithelial component. The larity with crowding of stromal cells exhibiting moderate to high degree
stromal component shows high degree of stromal cellularity, but the of nuclear pleomorphism. Mitotic figure can be seen
degree of nuclear pleomorphism is mild to moderate. The margin is
infiltrative with seepage of irregular tongues of stromal cells into the
adjacent normal breast stroma
the criteria is into three grades; the separation from minimal/

mild to moderate and moderate to marked changes are not
clearly defined, in particular cellularity and atypia. It has
been suggested that mild hypercellularity is characterized by
a slight increase in stromal cells as compared with normal
perilobular stroma, with evenly spaced nuclei that are not
touching or overlapping. Marked stromal cellularity shows
confluent areas of densely overlapping nuclei, whereas mod-
erate stromal cellularity has findings that are intermediate,
with some overlapping stromal nuclei. Mild stromal atypia
shows nuclei with little variation in size, with smooth nuclear
contours. Moderate atypia shows some variation in nuclear Fig. 6.6 Overgrowth of the stroma is characteristic in borderline or
size, with wrinkled nuclear membranes, exceeding that in malignant phyllodes tumors. There is expansion of the stroma to the
mild atypia but less than that in marked atypia. Marked exclusion of the epithelial component
atypia shows marked variation in nuclear size, coarse chro-
matin, and irregular nuclear membranes with discernible inter-observer variation, and not all histologic parameters for
nucleoli. Furthermore, there is intralesional heterogeneity PT grading are actually relevant in predicting PT outcome.
within the PT. PT could harbor features that typify benign In fact, a recently developed predictive nomogram only
lesion in some areas and borderline and malignant features in included three histologic criteria (atypia, mitosis, and stro-
other foci. In such situation, the diagnosis should be based mal overgrowth) and surgical margin status for risk assess-
on the worse area. Lastly, there has not been any consensus ment of PT recurrence.
as to whether any criteria are more important than others.
See Figs. 6.4, 6.5, 6.6, and 6.7. 19. Use of a two-tier or three-tier grading system: which
one is more popular [16, 19, 43]?
18. What is the clinical relevance of grading phyllodes
tumor [42]? Earlier grading scheme with a two-tier system classified PT
as benign (low grade) and malignant (high grade). A three-
The grading of PT allows prediction of prognosis and clini- tier system grading PT into benign, borderline, and malig-
cal behavior. Benign PTs have potential for local recurrence nant has been used by most investigators and is adopted by
while borderline, and malignant PTs have the potential to WHO. It is suggested to be more clinically relevant as this
locally recur and to metastasize to distant organs, and the approach leads to greater certainty at the ends of the PT
risk is higher in malignant PT. While there is correlation spectrum (i.e., benign and malignant PT) (WHO
between grade and recurrence, grading of PT is subject to classification).
PT grading, overall stromal cellularity, stromal cell pleomor-

phism, and mitotic count in CNB show good correlation with
excisions. On the other hand, stromal architectural changes
are more variable. The certainty increases with increasing
degree of malignancy. Grading of PT in CNB could be prob-
lematic unless overtly malignant features are seen. A more
accurate assessment can only be made in the excision
specimens.
Differentials of Benign PT and FA
22. What are the differentiating features between benign

phyllodes tumor and fibroadenoma [7, 51]?
The features to differentiate benign PT from FA include a

pronounced intracanalicular growth pattern, the presence
of increased stromal cellularity along with well-developed
stromal fronds. In the absence of well-developed stromal
fronds, the presence of elongated, branching, and cleft-
Fig. 6.7 Another example of malignant phyllodes tumor showing like ductal spaces meandering through a cellular stroma,
high-grade sarcoma-like features, composing of poorly cohesive malig- giving a staghorn appearance, may be a histologic clue to
nant stromal cells with brisk mitotic activity the diagnosis of PT. In benign PT, the increased stromal
cellularity should be mostly present throughout the lesion,
with subepithelial accentuation (cambium layer). At
times, it may be difficult to distinguish benign PT from
20. Are there any biomarkers and genetic changes that cellular FA because of overlapping histologic features.
are useful for grading of phyllodes tumor [44–49]? Definitive differentiation may not be possible. Stromal
fronds typical of PT are not seen in cellular fibroadenoma,
Correlations of immunohistochemical (IHC) markers with but, if present, are focal and not well developed. Cellular
PT grading have been extensively studied. Some markers FA should not demonstrate significant nuclear overlap or
showed promise but have not been proven clinically; see areas of sheetlike stromal growth; the presence of these
question 28 for more details. Genetic changes in PT also cor- features favors PT.
related with increased malignancy. Benign PT showed few
chromosomal changes, compared to numerous chromosomal 23. Should we differentiate between fibroadenoma
imbalances in malignant and borderline PT. Chromosomal and benign phyllodes tumors in all cases [12, 16, 42,
changes, namely, a high number of copy number variations 52, 53]?
(gains of 7p and 8q, losses of 3q and 10, losses in 9p21.3)
and the presence of amplifications, especially involving The differentiation between FA and benign PT appears
EGFR, could be potential markers for malignant PT. IGF1R important because of their purported differences in treat-
gene amplification was also reported in malignant PT only ment and prognosis. A second surgical procedure may be
and could be another potential marker. Mutations in cancer considered to achieve negative margins for a benign PT
genes (e.g., p53, EGFR, RB1, etc.) have been found exclu- incompletely excised initially. FA shares with benign PT a
sively in borderline/malignant PT and may also be useful in similar recurrence rate (15–17%) if incompletely excised.
aiding of PT grading. However, due to the similar clinical outcomes, a differentia-
tion between the two may not be significant. Although some
21. In a biopsy report of phyllodes tumor, can one grade reports suggested that a minuscule proportion of benign PT
the phyllodes tumor [27, 29, 50]? may recur as malignant lesions, there is yet no identifiable
feature/marker to enable prediction of such recurrence.

In CNB reporting of PT, there exists up to 40% of over- Thus, when there is histologic ambiguity, a diagnosis of
grading (from benign to atypical) or under-grading (from fibroadenoma will be preferred over a benign PT to avoid
atypical to benign). Among the different parameters used in overtreatment.
24. In a biopsy, how should we report if we are not sure diagnosis of metaplastic carcinoma. However, such interpreta-
whether the lesion is a fibroadenoma or a phyllodes tion could be tempered with focal CK and p63 staining reported
tumor [54]? in both metaplastic carcinomas and PT. p40, postulated to be a
more specific marker for squamous differentiation, may also be
The term “benign fibroepithelial neoplasm” with the explana- expressed by the stromal cells in malignant PT. Nevertheless it
tion of the diagnostic difficulties as needed could be employed was considered more specific, albeit less sensitive than p63 for
in equivocal cases, in order to avoid overtreatment. This term, metaplastic carcinoma. It is important to note that not all CK
however, should be used sparingly, as it does not represent a markers are expressed in all metaplastic carcinoma. A CK
new diagnostic category. More definitive characterization can panel is required in the differential diagnosis of malignant spin-
be performed on the subsequent excision specimen. dle cell lesions. CD34 conversely is variably expressed in PT
but not in spindle cell metaplastic carcinoma. The expression
of CD34 was reported to be negatively correlated with adverse
Differentials of Malignant PT histologic features. Another differential diagnosis is fibromato-
sis, and IHC nuclear staining for β-catenin is routinely used.
25. What are the differential diagnoses for malignant However, aberrant nuclear expression of β-catenin has been
phyllodes tumor? How can one differentiate on H&E reported in up to 94% benign PT and less in higher-grade PT,
[41, 55]? compared to essentially all cases in fibromatosis, and 23% in
metaplastic carcinoma. Other markers, which are more fre-
The differential diagnoses for malignant PT mainly are spin- quently expressed in PT and which could be used as diagnostic
dle cell carcinoma, metaplastic carcinoma, and sarcomas adjuncts, include bcl-2 and CD117.
(primary or metastatic). The diagnosis depends on the pres-
ence of residual epithelial structures. Malignant PT show
frond-like epithelium with typically intracanalicular growth Biomarkers
pattern associated with the malignant spindle cell prolifera-
tion. However, stromal overgrowth can be prominent and the 27. Are there any IHC markers that help with the diag-
biphasic nature (benign epithelial component) may not be nosis of phyllodes tumors [28, 61, 65]?
readily appreciated, necessitating extensive sampling with
many sections. The stroma of a malignant PT may, on occa- For diagnosis of PT, IHC markers are mainly used for the
sion, show heterologous sarcomatous differentiation, includ- differentiation of PT from metaplastic carcinomas or sarco-
ing liposarcoma, rhabdomyosarcoma, leiomyosarcoma, mas. On IHC, PT and FA show similar IHC profiles and thus
angiosarcoma, chondrosarcoma, and osteosarcoma. A spin- are considered as a whole; the reported useful markers
dle cell metaplastic carcinoma contains varying proportion include CD34 and proliferative indices (Ki67 and topoisom-
of malignant epithelial components, which may be of squa- erase IIα). In normal breast, stromal fibroblasts express
mous, adenocarcinoma (ductal), or adenosquamous types. CD34 which becomes diminished in malignant disease. In
Metaplastic carcinoma can also be entirely devoid of frank FA, nearly all stromal cells express CD34 while patchy stain-
epithelial element or show similar heterologous mesenchy- ing is seen in PT. A lower proportion of cells shows CD34
mal differentiation, although liposarcomatous element is positivity in the malignant PT, compared to borderline and
unusual. The presence of DCIS adjacent to a malignant spin- benign PT. One of the discriminatory features of PT from FA
dle cell tumor greatly favors metaplastic carcinoma. Primary is the presence of mitotic activity. Thus proliferation indices
and metastatic sarcomas of the breast are rare. There is sig- correlate significantly with key histologic features of PT and
nificant overlap with malignant PT and metaplastic carci- may be discriminatory for PT and FA. IHC staining for pro-
noma, and differentiating them based on H&E staining is liferation indices in CNB was higher in those cases that diag-
difficult. A history of previous or metastatic sarcoma, imag- nosed as PT on the subsequent excision than those diagnosed
ing, and clinical correlation may be helpful. as FA. However, there are no established threshold and meth-
odology for its assessment in this setting. For the diagnosis
26. What are common IHC markers to differentiate of PT from spindle cell carcinoma, CD34, CK, and p63 are
malignant phyllodes tumors from other spindle cell useful. (Please refer to question 26 for details.)
lesions [56–64]?
28. Are there any IHC markers that help with the grad-
Spindle cell carcinoma (metaplastic carcinoma) expresses p63 ing of phyllodes tumors [46, 64, 66–78]?
and cytokeratins (CK) (AE1/3, CAM5.2), in particularly high
molecular weight cytokeratins (HMWCK) (34βE12, CK5/6, Biological markers that showed positive correlation with PT
CK14). Positive staining with any of these markers supports a grade, such as p53, Ki67, C-kit, CD10, and EGFR, may
potentially be useful for PT grading. Several studies showed unclear. The current data indicate a limited role for hormonal
p53 staining correlated positively with grade of PT. Moderate therapy in treatment of PTs. Thus, ER, PR, and HER2 testing
to strong nuclear staining of stromal cells was noted in in PT may not have clear clinical relevance and are not rou-
malignant PTs while no staining was noted in benign PTs. tinely tested in most centers.
p53 staining also correlated positively with stromal cellular-
ity and overgrowth. As mitosis is one of the criteria for PT 31. Are there genetic changes in fibroadenoma [89–97]?
grading, stromal proliferation index as measured by Ki67
staining is suggested for PT grading. A positive correlation Abnormal karyotypes have been identified in 20–30% of
of Ki67 staining was found with PT grade. Immunostaining FAs, usually with structural aberrations. However, there are
of Ki67 was reported as 5% and 15% for benign and malig- no consistent chromosomal alterations in FA within or across
nant PT, respectively. The stromal cells of PT express mem- studies. Genetic changes are reported to be polyclonal. A
brane tyrosine kinase receptor, c-kit. c-kit expression was lower level of nonsynonymous somatic mutations in FA than
found in 17% of benign, 24% of borderline, and 46% of breast cancer (a median of five mutations in FA compared to
malignant PTs. The overall rate of EGFR expression was 33 in breast cancers) has been demonstrated. Among these,
16.2%, 30.6%, and 56% for benign, borderline, and malig- MED12 somatic mutation has been reported in FA, with a
nant PTs, respectively. EGFR staining not only correlated frequency ranging from 47 to 62%. The mutation mostly
positively with PT grade but also with other markers impli- occurs at a hotspot of exon 2 in codon 44, although the aver-
cated for PT grading such as p53, Ki67, and c-kit. CD10 age allele frequency was around 14%.
(CALLA) expression in stromal cells of PT also positively
correlated with PT grade. A positive staining in 5.9% of 32. Are there genetic changes indicating that fibroade-
benign, 31.4% of borderline, and 50% of malignant PTs was noma and phyllodes tumor are distinct [44, 46–49,
reported, especially in the periductal zone with increased 85, 98–104]?
stromal cellularity and mitotic activity. Caution should be
taken for CD10 staining in malignant PT as CD10 is also PTs are genetically more advanced than FA and displayed
expressed in some high-grade sarcomas of the breast. more genetic changes. SNP arrays demonstrated rare LOH
in FA with fractional LOH rates ranging from 0 to 1.5%,
29. Are there any IHC markers that help with the pre- compared to an average of 9.4% for PT. Contrasting the lack
diction of recurrences and metastases in phyllodes of recurrent chromosomal aberrations in FA, consistent 1q
tumors [64, 66, 73, 78–85]? gain and 13q loss in PT was observed in a number of stud-
ies. One study showed these genomic alterations mainly
Despite the number of marker studies in PT, many showed occur in PT with higher grade; others showed no association
only an association with grade but not prognostic value in of 1q gain with grade or 1q gain in benign PT. These find-
multiple studies. p53 expression and Ki67 index were ings are not conclusive at this stage. In addition, regions of
reported in some studies to be significantly associated with amplification and homozygous deletion are observed in PT,
either disease-free and/or overall survivals. However, others particularly high-grade PT, but rarely in FA. Homozygous
showed no association with recurrence or clinical behavior. deletion at 9p21 accompanied by loss of p16 expression has
c-kit expression is more consistently correlated with poor been reported in high-grade PT. Recurrent EGFR amplifica-
clinical outcome. Cases with stromal c-kit expression are tions associated with malignant PTs have been reported in
more prone to recur, and a worse overall survival was noted. several studies. Other amplifications reported in PTs include
Other markers reported to have prognostic implications in MDM4, RAF1, TERT, and MET. More recently, with mas-
PT include six 1, Yap, and α-catenin; yet, these are mostly sively parallel sequencing, a higher mutational burden was
single studies and require further validation. shown to be displayed by PT (median non-silent somatic
mutation per case: 5 in FA vs 13 in PT). In addition to
30. Should we do ER, PR, and HER2 testing in malig- MED12 mutation which is also commonly found in FA, PT
nant phyllodes tumor [79, 86–88]? harbors other recurrent mutations. Recurrent mutations in
FLNA, SETD2, and KMT2D were found across the spec-
In malignant PT, the neoplastic stromal cells do not usually trum of PT while mutations in bona fide cancer genes (such
express ERα, PR, and HER2. Only ERβ expression has been as NF1, RB1, TP53, PIK3CA, and EGFR) were only
reported in these stromal cells. ERα and PR expression is detected in borderline and malignant PT. Furthermore, PT
confined to the epithelial compartment, but shows inverse may harbor TERT promoter mutations which are absent or
association to grade. Epithelial HER2 expression has been exceedingly rare in FA. TERT mutations have been detected
observed but there was no correlation with PT grade and in the majority of high-grade PTs and in a small subset of
prognosis. The practical significance of these observations is benign PTs.
33. Are there any potential targets in the genetic changes 35. What is the difference in outcome between fibroade-
of phyllodes tumor [46–48, 49, 75, 85, 102, 103, noma and phyllodes tumor [3, 2, 42, 52, 53,
105–112]? 120–123]?
Several potential targetable genetic changes have been found In general, FAs rarely recur, even complex FAs, and some
in PT. Recent NGS studies revealed hotspot mutations in PT may regress spontaneously. A study has shown that one-half
including AKT1 (E17K), ERBB2 (V777 L), ERBB3 (V104 L), of FAs followed clinically without excision disappeared on
NRAS (Q61K), and PIK3CA (H1047R). Some of these muta- imaging after 5 years. Several studies reported, however, a
tions have been shown to be predictors for response to spe- recurrence rate of 15–17% for FAs after excision, occurring
cific therapeutic agents in other malignancies. AKT inhibitor exclusively in FAs measuring more than 2 mm at initial diag-
AZD5363 induced partial responses in patients with breast nosis. Thus, FA and benign PT may have similar recurrence
and ovarian cancer with tumors containing AKT1E17K muta- rate. However, benign PT may recur as malignant PT. There
tions. Neratinib, an irreversible dual ERBB2/EGFR tyrosine are rare cases of benign PT showing metastasis.
kinase inhibitor, has been found to be active in breast cancers
with activating ERBB2 mutations. In addition, c-MET ampli-
fication, the major mechanism of constitutive c-MET activa- Case Presentations
tion in human cancers, was also reported in malignant PTs.
There is evidence showing responsiveness to Crizotinib in a Case 1
subset of esophagogastric adenocarcinoma with MET ampli- A 50-year-old woman has a fast-growing mass in her right
fication. Across different studies, the most frequently altered breast. No axillary lymph node enlargement was noted.
gene in PT with potential therapeutic relevance is EGFR. Incisional biopsy of the mass was performed, and the histol-
EGFR amplifications and overexpression have been reported ogy was reviewed.
in many studies. Its pathogenic missense mutations, namely, Highly pleomorphic malignant cells with no tubule for-
L62R, L62R, G63R, E84V, and V774 M, have been revealed mation are noted, but areas of necrosis are seen. Mitotic
in PT. Although the therapeutic relevance of these alterations account was very brisk, counting to 30 to 40 per 10 high-
remains to be investigated, these genetic changes could rep- power fields. No definite benign ductal and carcinoma com-
resent potential therapeutic target. The driver role of EGFR ponents are identified.
amplification and V774 M mutation has been shown in other What are the differential diagnoses? What is the recom-
cancers. The other mutations located at the site target by mended treatment? The differential diagnoses include spin-
cetuximab and panitumumab have also been identified to be dle cell carcinoma, sarcoma, and malignant phyllodes tumor:
new functional variants of EGFR.
1. To identify the presence of epithelial component or to
34. What is the outcome for phyllodes tumor [13, 16, 40, determine if it is not present, one needs to consider sar-
42, 113–119]? coma or spindle cell carcinoma (metaplastic carcinoma).
If benign epithelial component is present, a diagnosis of
Most PTs behave in a benign fashion without recurrence and malignant phyllodes tumor has to be considered. If the
metastasis following excision. The local recurrence rate is epithelial component is malignant (either glandular or
21%, with ranges of 10–17%, 14–25%, and 23–30% for squamous), then the diagnosis will be metaplastic
benign, borderline, and malignant PTs, respectively. Most carcinoma.
local recurrences occur within the first two years after diag- 2. If malignant heterologous element is identified, the diag-
nosis. It appears that Asian patients experienced a higher nosis can also be malignant phyllodes tumor, sarcoma, or
recurrence rate than non-Asians. The recurrences mostly metaplastic carcinoma; thus, the identification of malig-
mirror the microscopic pattern of the original tumor, but nant heterologous element is not useful in the diagnosis.
grade progression has been reported in 25–75% of cases.
Very rarely, PT (less than 2%) metastasizes to distant organs. In this case, no epithelial component is identified and no
It occurs essentially only in malignant PT. Metastases usu- malignant heterologous element is seen:
ally occur within 5 years of initial diagnosis. The most com-
mon sites are lung and pleura, but all organ sites can be 1. Additional sampling may be undertaken, and, in this case,
affected. PT tends to spread via the hematogenous rather a small area of benign epithelial component is seen.
than lymphatic route. Metastases in PT invariably indicate a 2. Immunohistochemistry can be done to characterize the
dismal prognosis with ensuing death. spindle cells.
Fig. 6.8 Staining for CK14, highlighting the staining in single cells in Fig. 6.9 Rare mitotic figure can be seen in a fibroadenoma, particu-
a focal manner in this case of malignant phyllodes tumor larly in a juvenile variant
3. AE1/3 will be positive in the spindle cells in spindle cell 1. The history of rapid growth, the mild increased stromal
carcinoma, but not usually in phyllodes tumor and seldom cellularity, and mitotic figures seem to suggest a diagno-
in sarcoma. In phyllodes tumor, AE1/3 and other cytokera- sis of phyllodes tumor.
tins (high molecular weight cytokeratins) have been 2. Immunohistochemistry is not helpful.
reported to be more likely positive in malignant than benign 3. The relatively young age is in favor of fibroadenoma and
phyllodes tumors and usually focally and patchy (Fig. 6.8). variants.
4. p63 will be positive in the spindle cells in spindle cell 4. Absence of areas of variable stromal cellularity favors a
carcinoma, but not usually in phyllodes tumor and seldom diagnosis of fibroadenoma, and the juvenile fibroade-
in sarcoma. In phyllodes tumor, p63 has been reported to noma may sometimes show mitotic activity (Fig. 6.9).
be more likely positive in malignant than benign phyl-
lodes tumors and usually focally and patchy. The diagnosis is a juvenile fibroadenoma.
The final diagnosis is malignant phyllodes tumor, as

benign epithelial component is identified, despite focal cyto- References
keratin positivity.
Recommended further treatment is complete excision. 1. Carney JA, Toorkey BC. Myxoid fibroadenoma and allied condi-
tions (myxomatosis) of the breast. A heritable disorder with special
There is no need for axillary sampling, as axillary lymph associations including cardiac and cutaneous myxomas. Am J Surg
node metastasis is uncommon. Pathol. 1991;15(8):713–21.
2. Dupont WD, Page DL, Parl FF, Vnencak-Jones CL, Plummer WD
Case 2 Jr, Rados MS, et al. Long-term risk of breast cancer in women with
fibroadenoma. N Engl J Med. 1994;331(1):10–5.
A 35-year-old woman presented with an asymptomatic 3. Sklair-Levy M, Sella T, Alweiss T, Craciun I, Libson E, Mally
breast mass. The mass has been rapidly growing for the past B. Incidence and management of complex fibroadenomas. AJR Am
2–3 months, and it now measures 3 cm. Physical examina- J Roentgenol. 2008;190(1):214–8.
tion shows a mobile firm mass not fixed to underlying tissue. 4. Kuijper A, Mommers EC, van der Wall E, van Diest
PJ. Histopathology of fibroadenoma of the breast. Am J Clin Pathol.
Imaging (USG) shows a lobulated mass that is wider than 2001;115(5):736–42.
tall, with posterior shadowing. Cleft-like spaces are seen. 5. Giri D. Recurrent challenges in the evaluation of fibroepithelial
A needle core biopsy was performed. The needle core lesions. Arch Pathol Lab Med. 2009;133(5):713–21.
biopsy shows a fibroepithelial lesion with fragmentation. 6. Jacobs TW, Chen YY, Guinee DG Jr, Holden JA, Cha I, Bauermeister
DE, et al. Fibroepithelial lesions with cellular stroma on breast core
There is prominent intracanalicular/phyllodal architectural needle biopsy: are there predictors of outcome on surgical excision?
pattern, with slightly increased stromal cellularity, and occa- Am J Clin Pathol. 2005;124(3):342–54.
sional mitotic figures can be seen. 7. Ross DS, Giri DD, Akram MM, Catalano JP, Olcese C, Van
What are the differential diagnoses? The differential diag- Zee KJ, et al. Fibroepithelial lesions in the breast of adoles-
cent females: a clinicopathological study of 54 cases. Breast J.
noses include fibroadenoma and benign phyllodes tumor: 2017;23(2):182–92.
8. Yasir S, Gamez R, Jenkins S, Visscher DW, Nassar A. Significant erative distinction between fibroadenoma and phyllodes tumour.
histologic features differentiating cellular fibroadenoma from phyl- Histopathology. 2010;56(4):489–500.
lodes tumor on core needle biopsy specimens. Am J Clin Pathol. 31. Jacklin RK, Ridgway PF, Ziprin P, Healy V, Hadjiminas D, Darzi
2014;142(3):362–9. A. Optimising preoperative diagnosis in phyllodes tumour of the
9. Raganoonan C, Fairbairn JK, Williams S, Hughes LE. Giant breast breast. J Clin Pathol. 2006;59(5):454–9.
tumours of adolescence. Aust N Z J Surg. 1987;57(4):243–7. 32. Dawson AE, Mulford DK, Sheils LA. The cytopathology of pro-
10. Grimes MM. Cystosarcoma phyllodes of the breast: histologic
liferative breast disease. Comparison with features of ductal carci-
features, flow cytometric analysis, and clinical correlations. Mod noma in situ. Am J Clin Pathol. 1995;103(4):438–42.
Pathol. 1992;5(3):232–9. 33. Jayaram G, Sthaneshwar P. Fine-needle aspiration cytology of phyl-
11. Salisbury JR, Singh LN. Apocrine metaplasia in phyllodes tumours lodes tumors. Diagn Cytopathol. 2002;26(4):222–7.
of the breast. Histopathology. 1986;10(11):1211. 34. Tse GM, Ma TK, Pang LM, Cheung H. Fine needle aspiration
12. Tan PH, Jayabaskar T, Chuah KL, Lee HY, Tan Y, Hilmy M, et al. cytologic features of mammary phyllodes tumors. Acta Cytol.
Phyllodes tumors of the breast: the role of pathologic parameters. 2002;46(5):855–63.
Am J Clin Pathol. 2005;123(4):529–40. 35. Bhattarai S, Kapila K, Verma K. Phyllodes tumor of the breast. A
13. Norris HJ, Taylor HB. Relationship of histologic features to behav- cytohistologic study of 80 cases. Acta Cytol. 2000;44(5):790–6.
ior of cystosarcoma phyllodes. Analysis of ninety-four cases. 36.
Dusenbery D, Frable WJ. Fine needle aspiration cytology
Cancer. 1967;20(12):2090–9. of phyllodes tumor. Potential diagnostic pitfalls. Acta Cytol.
14. Co M, Tse GM, Chen C, Wei J, Kwong A. Coexistence of ductal 1992;36(2):215–21.
carcinoma within mammary Phyllodes tumor: a review of 557 cases 37. Shabb NS. Phyllodes tumor. Fine needle aspiration cytology of
from a 20-year region-wide database in Hong Kong and southern eight cases. Acta Cytol. 1997;41(2):321–6.
China. Clin Breast Cancer. 2018;18(3):e421–5. 38. Tse GM, Tan PH. Diagnosing breast lesions by fine needle aspira-
15. Sugie T, Takeuchi E, Kunishima F, Yotsumoto F, Kono Y. A case of tion cytology or core biopsy: which is better? Breast Cancer Res
ductal carcinoma with squamous differentiation in malignant phyl- Treat. 2010;123(1):1–8.
lodes tumor. Breast Cancer. 2007;14(3):327–32. 39. Parker SJ, Harries SA. Phyllodes tumours. Postgrad Med J.

16. Lakhani SR, Ellis IO, Schnitee SJ, Tan PH, van de Vijver MJ, edi- 2001;77(909):428–35.
tors. World Health Organisation classification of tumors of the 40. Pietruszka M, Barnes L. Cystosarcoma phyllodes: a clinicopatho-
breast. 4th ed. Lyon: IARC Press; 2012. logic analysis of 42 cases. Cancer. 1978;41(5):1974–83.
17. Powell CM, Rosen PP. Adipose differentiation in cystosarcoma 41. Tan BY, Acs G, Apple SK, Badve S, Bleiweiss IJ, Brogi E, et al.
phyllodes. A study of 14 cases. Am J Surg Pathol. 1994;18(7):720–7. Phyllodes tumours of the breast: a consensus review. Histopathology.
18. Tse GM, Law BK, Chan KF, Mas TK. Multinucleated stromal giant 2016;68(1):5–21.
cells in mammary phyllodes tumours. Pathology. 2001;33(2):153–6. 42. Tan PH, Thike AA, Tan WJ, Thu MM, Busmanis I, Li H, et al.
19. McDivitt RW, Farrow JH, Stewart FW. Breast carcinoma arising in Phyllodes tumour network S: predicting clinical behaviour of breast
solitary fibroadenomas. Surg Gynecol Obstet. 1967;125(3):572–6. phyllodes tumours: a nomogram based on histological criteria and
20. Stafyla V, Kotsifopoulos N, Grigoriades K, Kassaras G, Sakorafas surgical margins. J Clin Pathol. 2012;65(1):69–76.
GH. Lobular carcinoma in situ of the breast within a fibroadenoma. 43. Hart WR, Bauer RC, Oberman HA. Cystosarcoma phyllodes. A
A case report. Gynecol Oncol. 2004;94(2):572–4. clinicopathologic study of twenty-six hypercellular periductal stro-
21. Buzanowski-Konakry K, Harrison EG Jr, Payne WS. Lobular
mal tumors of the breast. Am J Clin Pathol. 1978;70(2):211–6.
carcinoma arising in fibroadenoma of the breast. Cancer. 44. Lae M, Vincent-Salomon A, Savignoni A, Huon I, Freneaux

1975;35(2):450–6. P, Sigal-Zafrani B, et al. Phyllodes tumors of the breast segre-
22. Ozzello L, Gump FE. The management of patients with carcinomas gate in two groups according to genetic criteria. Mod Pathol.
in fibroadenomatous tumors of the breast. Surg Gynecol Obstet. 2007;20(4):435–44.
1985;160(2):99–104. 45. Lae M, La Rosa P, Mandel J, Reyal F, Hupe P, Terrier P, et al.
23. Abdul Aziz M, Sullivan F, Kerin MJ, Callagy G. Malignant phyl- Whole-genome profiling helps to classify phyllodes tumours of the
lodes tumour with liposarcomatous differentiation, invasive tubular breast. J Clin Pathol. 2016;69(12):1081–7.
carcinoma, and ductal and lobular carcinoma in situ: case report and 46. Tse GM, Lui PC, Vong JS, Lau KM, Putti TC, Karim R, et al.
review of the literature. Pathol Res Int. 2010;2010:501274. Increased epidermal growth factor receptor (EGFR) expression in
24. Dillon MF, Quinn CM, McDermott EW, O’Doherty A, O’Higgins malignant mammary phyllodes tumors. Breast Cancer Res Treat.
N, Hill AD. Needle core biopsy in the diagnosis of phyllodes neo- 2009;114(3):441–8.
plasm. Surgery. 2006;140(5):779–84. 47. Cani AK, Hovelson DH, McDaniel AS, Sadis S, Haller MJ, Yadati
25. Komenaka IK, El-Tamer M, Pile-Spellman E, Hibshoosh H. Core V, et al. Next-gen sequencing exposes frequent MED12 mutations
needle biopsy as a diagnostic tool to differentiate phyllodes tumor and actionable therapeutic targets in Phyllodes tumors. Mol Cancer
from fibroadenoma. Arch Surg. 2003;138(9):987–90. Res. 2015;13(4):613–9.
26. Bode MK, Rissanen T, Apaja-Sarkkinen M. Ultrasonography and 48. Piscuoglio S, Ng CK, Murray M, Burke KA, Edelweiss M, Geyer
core needle biopsy in the differential diagnosis of fibroadenoma and FC, et al. Massively parallel sequencing of phyllodes tumours of
tumor phyllodes. Acta Radiol. 2007;48(7):708–13. the breast reveals actionable mutations, and TERT promoter hotspot
27. Tsang AK, Chan SK, Lam CC, Lui PC, Chau HH, Tan PH, et al. mutations and TERT gene amplification as likely drivers of progres-
Phyllodes tumours of the breast – differentiating features in core sion. J Pathol. 2016;238(4):508–18.
needle biopsy. Histopathology. 2011;59(4):600–8. 49. Tan J, Ong CK, Lim WK, Ng CC, Thike AA, Ng LM, et al.
28. Jara-Lazaro AR, Akhilesh M, Thike AA, Lui PC, Tse GM, Tan Genomic landscapes of breast fibroepithelial tumors. Nat Genet.
PH. Predictors of phyllodes tumours on core biopsy specimens of 2015;47(11):1341–5.
fibroepithelial neoplasms. Histopathology. 2010;57(2):220–32. 50. Foxcroft LM, Evans EB, Porter AJ. Difficulties in the pre-operative
29. Lee AH, Hodi Z, Ellis IO, Elston CW. Histological features useful diagnosis of phyllodes tumours of the breast: a study of 84 cases.
in the distinction of phyllodes tumour and fibroadenoma on needle Breast. 2007;16(1):27–37.
core biopsy of the breast. Histopathology. 2007;51(3):336–44. 51. Tay TK, Chang KT, Thike AA, Tan PH. Paediatric fibroepi-

30. Morgan JM, Douglas-Jones AG, Gupta SK. Analysis of histo-
thelial lesions revisited: pathological insights. J Clin Pathol.
logical features in needle core biopsy of breast useful in preop- 2015;68(8):633–41.
52. Grady I, Gorsuch H, Wilburn-Bailey S. Long-term outcome of 71. Umekita Y, Yoshida H. Immunohistochemical study of MIB1

benign fibroadenomas treated by ultrasound-guided percutaneous expression in phyllodes tumor and fibroadenoma. Pathol Int.
excision. Breast J. 2008;14(3):275–8. 1999;49(9):807–10.
53. Organ CH Jr, Organ BC. Fibroadenoma of the female breast: a criti- 72. Sawyer EJ, Poulsom R, Hunt FT, Jeffery R, Elia G, Ellis IO, et al.
cal clinical assessment. J Natl Med Assoc. 1983;75(7):701–4. Malignant phyllodes tumours show stromal overexpression of
54. Tan PH, Ellis IO. Myoepithelial and epithelial-myoepithelial,
c-myc and c-kit. J Pathol. 2003;200(1):59–64.
mesenchymal and fibroepithelial breast lesions: updates from the 73. Tan WJ, Thike AA, Tan SY, Tse GM, Tan MH, Bay BH, et al.
WHO classification of Tumours of the breast 2012. J Clin Pathol. CD117 expression in breast phyllodes tumors correlates with
2013;66(6):465–70. adverse pathologic parameters and reduced survival. Mod Pathol.
55. Rakha EA, Tan PH, Shaaban A, Tse GM, Esteller FC, van Deurzen 2015;28(3):352–8.
CH, et al. Do primary mammary osteosarcoma and chondrosarcoma 74. Tse GM, Putti TC, Lui PC, Lo AW, Scolyer RA, Law BK, et al.
exist? A review of a large multi-institutional series of malignant Increased c-kit (CD117) expression in malignant mammary phyl-
matrix-producing breast tumours. Breast. 2013;22(1):13–8. lodes tumors. Mod Pathol. 2004;17(7):827–31.
56. Chia Y, Thike AA, Cheok PY, Yong-Zheng Chong L, Man-Kit Tse 75. Kersting C, Kuijper A, Schmidt H, Packeisen J, Liedtke C,

G, et al. Stromal keratin expression in phyllodes tumours of the Tidow N, et al. Amplifications of the epidermal growth fac-
breast: a comparison with other spindle cell breast lesions. J Clin tor receptor gene (egfr) are common in phyllodes tumors of the
Pathol. 2012;65(4):339–47. breast and are associated with tumor progression. Lab Investig.
57. Cimino-Mathews A, Sharma R, Illei PB, Vang R, Argani P. A sub- 2006;86(1):54–61.
set of malignant phyllodes tumors express p63 and p40: a diag- 76. Tsai WC, Jin JS, Yu JC, Sheu LF. CD10, actin, and vimentin expres-
nostic pitfall in breast core needle biopsies. Am J Surg Pathol. sion in breast phyllodes tumors correlates with tumor grades of the
2014;38(12):1689–96. WHO grading system. Int J Surg Pathol. 2006;14(2):127–31.
58. Dunne B, Lee AH, Pinder SE, Bell JA, Ellis IO. An immu-
77. Tse GM, Tsang AK, Putti TC, Scolyer RA, Lui PC, Law BK, et al.
nohistochemical study of metaplastic spindle cell carcinoma, Stromal CD10 expression in mammary fibroadenomas and phyl-
phyllodes tumor and fibromatosis of the breast. Hum Pathol. lodes tumours. J Clin Pathol. 2005;58(2):185–9.
2003;34(10):1009–15. 78. Leibl S, Moinfar F. Metaplastic sarcomatoid carcinoma of the

59. Tse GM, Tan PH, Chaiwun B, Putti TC, Lui PC, Tsang AK, et al. breast: not a misnomer. Am J Surg Pathol. 2006;30(8):1052–3.
p63 is useful in the diagnosis of mammary metaplastic carcinomas. author reply 1053–1055
Pathology. 2006;38(1):16–20. 79. Shpitz B, Bomstein Y, Sternberg A, Klein E, Tiomkin V, Kaufman
60. Lee AH. Recent developments in the histological diagnosis of spin- A, et al. Immunoreactivity of p53, Ki-67, and c-erbB-2 in phyllodes
dle cell carcinoma, fibromatosis and phyllodes tumour of the breast. tumors of the breast in correlation with clinical and morphologic
Histopathology. 2008;52(1):45–57. features. J Surg Oncol. 2002;79(2):86–92.
61. Moore T, Lee AH. Expression of CD34 and bcl-2 in phyllodes 80. Niezabitowski A, Lackowska B, Rys J, Kruczak A, Kowalska T,
tumours, fibroadenomas and spindle cell lesions of the breast. Mitus J, et al. Prognostic evaluation of proliferative activity and
Histopathology. 2001;38(1):62–7. DNA content in the phyllodes tumor of the breast: immunohisto-
62. Lacroix-Triki M, Geyer FC, Lambros MB, Savage K, Ellis IO, Lee chemical and flow cytometric study of 118 cases. Breast Cancer Res
AH, et al. Beta-catenin/Wnt signalling pathway in fibromatosis, Treat. 2001;65(1):77–85.
metaplastic carcinomas and phyllodes tumours of the breast. Mod 81. Yonemori K, Hasegawa T, Shimizu C, Shibata T, Matsumoto K,
Pathol. 2010;23(11):1438–48. Kouno T, et al. Correlation of p53 and MIB-1 expression with both
63. Noronha Y, Raza A, Hutchins B, Chase D, Garberoglio C, Chu P, the systemic recurrence and survival in cases of phyllodes tumors of
et al. CD34, CD117, and Ki-67 expression in phyllodes tumor of the breast. Pathol Res Pract. 2006;202(10):705–12.
the breast: an immunohistochemical study of 33 cases. Int J Surg 82. Kuijper A, van der Groep P, van der Wall E, van Diest PJ. Expression
Pathol. 2011;19(2):152–8. of hypoxia-inducible factor 1 alpha and its downstream tar-
64. Tan PH, Jayabaskar T, Yip G, Tan Y, Hilmy M, Selvarajan S, et al. gets in fibroepithelial tumors of the breast. Breast Cancer Res.
p53 and c-kit (CD117) protein expression as prognostic indicators 2005;7(5):R808–18.
in breast phyllodes tumors: a tissue microarray study. Mod Pathol. 83. Tsang JY, Mendoza P, Lam CC, Yu AM, Putti TC, Karim RZ,
2005;18(12):1527–34. et al. Involvement of α- and β-catenins and E-cadherin in the
65. Chauhan H, Abraham A, Phillips JR, Pringle JH, Walker RA, Jones development of mammary phyllodes tumours. Histopathology.
JL. There is more than one kind of myofibroblast: analysis of CD34 2012;61(4):667–74.
expression in benign, in situ, and invasive breast lesions. J Clin 84. Kim SK, Jung WH, Koo JS. Expression of yes-associated pro-
Pathol. 2003;56(4):271–6. tein (YAP) in breast phyllodes tumor. Int J Clin Exp Pathol.
66. Kleer CG, Giordano TJ, Braun T, Oberman HA. Pathologic, immu- 2014;7(9):5997–6005.
nohistochemical, and molecular features of benign and malignant 85. Tan WJ, Lai JC, Thike AA, Lim JC, Tan SY, Koh VC, et al. Novel
phyllodes tumors of the breast. Mod Pathol. 2001;14(3):185–90. genetic aberrations in breast phyllodes tumours: comparison
67. Korcheva VB, Levine J, Beadling C, Warrick A, Countryman G, between prognostically distinct groups. Breast Cancer Res Treat.
Olson NR, et al. Immunohistochemical and molecular markers in 2014;145(3):635–45.
breast phyllodes tumors. Appl Immunohistochem Mol Morphol. 86. Tse GM, Lee CS, Kung FY, Scolyer RA, Law BK, Lau TS,
2011;19(2):119–25. et al. Hormonal receptors expression in epithelial cells of mam-
68. Tse GM, Putti TC, Kung FY, Scolyer RA, Law BK, Lau TS, et al. mary phyllodes tumors correlates with pathologic grade of
Increased p53 protein expression in malignant mammary phyllodes the tumor: a multicenter study of 143 cases. Am J Clin Pathol.
tumors. Mod Pathol. 2002;15(7):734–40. 2002;118(4):522–6.
69. Karim RZ, Gerega SK, Yang YH, Spillane A, Carmalt H, Scolyer 87. Sapino A, Bosco M, Cassoni P, Castellano I, Arisio R, Cserni
RA, et al. p16 and pRb immunohistochemical expression increases G, et al. Estrogen receptor-beta is expressed in stromal cells of
with increasing tumour grade in mammary phyllodes tumours. fibroadenoma and phyllodes tumors of the breast. Mod Pathol.
Histopathology. 2010;56(7):868–75. 2006;19(4):599–606.
70. Kocova L, Skalova A, Fakan F, Rousarova M. Phyllodes tumour of 88. Carlson RW, Allred DC, Anderson BO, Burstein HJ, Carter WB,
the breast: immunohistochemical study of 37 tumours using MIB1 Edge SB, et al. Breast cancer: noninvasive and special situations. J
antibody. Pathol Res Pract. 1998;194(2):97–104. Natl Compr Cancer Netw. 2010;8(10):1182–207.
89. Calabrese G, Di Virgilio C, Cianchetti E, Guanciali Franchi P, 106. Ma CX, Suman V, Goetz MP, Northfelt D, Burkard ME,
Stuppia L, Parruti G, et al. Chromosome abnormalities in breast Ademuyiwa F, et al. A phase II trial of Neoadjuvant MK-2206,
fibroadenomas. Genes Chromosomes Cancer. 1991;3(3):202–4. an AKT inhibitor, with Anastrozole in clinical stage II or III
90. Dietrich CU, Pandis N, Teixeira MR, Bardi G, Gerdes AM, PIK3CA-mutant ER-positive and HER2-negative breast Cancer.
Andersen JA, et al. Chromosome abnormalities in benign hyper- Clin Cancer Res. 2017;23(22):6823–32.
proliferative disorders of epithelial and stromal breast tissue. Int J 107. Christensen JG, Burrows J, Salgia R. C-met as a target for human
Cancer. 1995;60(1):49–53. cancer and characterization of inhibitors for therapeutic interven-
91. Fletcher JA, Pinkus GS, Weidner N, Morton CC. Lineage- tion. Cancer Lett. 2005;225(1):1–26.
restricted clonality in biphasic solid tumors. Am J Pathol. 108. Lennerz JK, Kwak EL, Ackerman A, Michael M, Fox SB,
1991;138(5):1199–207. Bergethon K, et al. MET amplification identifies a small and
92. Stephenson CF, Davis RI, Moore GE, Sandberg AA. Cytogenetic aggressive subgroup of esophagogastric adenocarcinoma
and fluorescence in situ hybridization analysis of breast fibroad- with evidence of responsiveness to crizotinib. J Clin Oncol.
enomas. Cancer Genet Cytogenet. 1992;63(1):32–6. 2011;29(36):4803–10.
93. Noguchi S, Motomura K, Inaji H, Imaoka S, Koyama H. Clonal 109. Gatalica Z, Vranic S, Ghazalpour A, Xiu J, Ocal IT, McGill J,
analysis of fibroadenoma and phyllodes tumor of the breast. et al. Multiplatform molecular profiling identifies potentially tar-
Cancer Res. 1993;53(17):4071–4. getable biomarkers in malignant phyllodes tumors of the breast.
94. Vogelstein B, Papadopoulos N, Velculescu VE, Zhou S, Diaz Oncotarget. 2016;7(2):1707–16.
LA Jr, Kinzler KW. Cancer genome landscapes. Science. 110. Wu JY, Yu CJ, Chang YC, Yang CH, Shih JY, Yang

2013;339(6127):1546–58. PC. Effectiveness of tyrosine kinase inhibitors on “uncommon”
95. Lim WK, Ong CK, Tan J, Thike AA, Ng CC, Rajasegaran V, et al. epidermal growth factor receptor mutations of unknown clini-
Exome sequencing identifies highly recurrent MED12 somatic cal significance in non-small cell lung cancer. Clin Cancer Res.
mutations in breast fibroadenoma. Nat Genet. 2014;46(8):877–80. 2011;17(11):3812–21.
96. Mishima C, Kagara N, Tanei T, Naoi Y, Shimoda M, Shimomura A, 111. Dokala A, Thakur SS. Extracellular region of epidermal growth
et al. Mutational analysis of MED12 in fibroadenomas and phyl- factor receptor: a potential target for anti-EGFR drug discovery.
lodes tumors of the breast by means of targeted next-generation Oncogene. 2017;36(17):2337–44.
sequencing. Breast Cancer Res Treat. 2015;152(2):305–12. 112. Niu B, Scott AD, Sengupta S, Bailey MH, Batra P, Ning J, et al.
97. Yoshida M, Sekine S, Ogawa R, Yoshida H, Maeshima A, Kanai Protein-structure-guided discovery of functional mutations across
Y, et al. Frequent MED12 mutations in phyllodes tumours of the 19 cancer types. Nat Genet. 2016;48(8):827–37.
breast. Br J Cancer. 2015;112(10):1703–8. 113. Tan EY, Tan PH, Yong WS, Wong HB, Ho GH, Yeo AW, et al.
98. Wang ZC, Buraimoh A, Iglehart JD, Richardson AL. Genome- Recurrent phyllodes tumours of the breast: pathological features
wide analysis for loss of heterozygosity in primary and recur- and clinical implications. ANZ J Surg. 2006;76(6):476–80.
rent phyllodes tumor and fibroadenoma of breast using 114. Barrio AV, Clark BD, Goldberg JI, Hoque LW, Bernik SF,
single nucleotide polymorphism arrays. Breast Cancer Res Treat. Flynn LW, et al. Clinicopathologic features and long-term out-
2006;97(3):301–9. comes of 293 phyllodes tumors of the breast. Ann Surg Oncol.
99. Lv S, Niu Y, Wei L, Liu Q, Wang X, Chen Y. Chromosomal aber- 2007;14(10):2961–70.
rations and genetic relations in benign, borderline and malignant 115. Cohn-Cedermark G, Rutqvist LE, Rosendahl I, Silfversward
phyllodes tumors of the breast: a comparative genomic hybridiza- C. Prognostic factors in cystosarcoma phyllodes. A clinicopatho-
tion study. Breast Cancer Res Treat. 2008;112(3):411–8. logic study of 77 patients. Cancer. 1991;68(9):2017–22.
100. Lu YJ, Birdsall S, Osin P, Gusterson B, Shipley J. Phyllodes 116. Karim RZ, Gerega SK, Yang YH, Spillane A, Carmalt H,
tumors of the breast analyzed by comparative genomic hybrid- Scolyer RA, et al. Phyllodes tumours of the breast: a clinico-
ization and association of increased 1q copy number with stro- pathological analysis of 65 cases from a single institution. Breast.
mal overgrowth and recurrence. Genes Chromosomes Cancer. 2009;18(3):165–70.
1997;20(3):275–81. 117. de Roos WK, Kaye P, Dent DM. Factors leading to local recur-
101. Jones AM, Mitter R, Springall R, Graham T, Winter E, Gillett rence or death after surgical resection of phyllodes tumours of the
C, et al. Phyllodes tumour C: a comprehensive genetic profile breast. Br J Surg. 1999;86(3):396–9.
of phyllodes tumours of the breast detects important mutations, 118. Hawkins RE, Schofield JB, Fisher C, Wiltshaw E, McKinna
intra-tumoral genetic heterogeneity and new genetic changes on JA. The clinical and histologic criteria that predict metastases
recurrence. J Pathol. 2008;214(5):533–44. from cystosarcoma phyllodes. Cancer. 1992;69(1):141–7.
102. Agelopoulos K, Kersting C, Korsching E, Schmidt H, Kuijper A, 119. Goh CH, Lim YP, Su JW, Khoo KS, Thomas A, Sittampalam K,
August C, et al. Egfr amplification specific gene expression in et al. Cardiopulmonary thromboembolism of epithelioid angiosar-
phyllodes tumours of the breast. Cell Oncol. 2007;29(6):443–51. coma arising from malignant phyllodes tumour of the breast. J
103. Tsang JY, Go EM, Tse GM. Identification of clinically rel- Clin Pathol. 2014;67(5):450–4.
evant alterations in phyllodes tumor of the breast by amplicon- 120. Greenberg R, Skornick Y, Kaplan O. Management of breast fibro-
based next-generation sequencing. Breast Cancer Res Treat. adenomas. J Gen Intern Med. 1998;13(9):640–5.
2015;151(3):717–9. 121. Jayasinghe Y, Simmons PS. Fibroadenomas in adolescence. Curr
104. Yoshida M, Ogawa R, Yoshida H, Maeshima A, Kanai Y, Kinoshita Opin Obstet Gynecol. 2009;21(5):402–6.
T, et al. TERT promoter mutations are frequent and show associa- 122. Cant PJ, Madden MV, Coleman MG, Dent DM. Non-operative
tion with MED12 mutations in phyllodes tumors of the breast. Br management of breast masses diagnosed as fibroadenoma. Br J
J Cancer. 2015;113(8):1244–8. Surg. 1995;82(6):792–4.
105. Davies BR, Guan N, Logie A, Crafter C, Hanson L, Jacobs V, et al. 123. Reinfuss M, Mitus J, Duda K, Stelmach A, Rys J, Smolak K. The
Tumors with AKT1E17K mutations are rational targets for single treatment and prognosis of patients with phyllodes tumor of the
agent or combination therapy with AKT inhibitors. Mol Cancer breast: an analysis of 170 cases. Cancer. 1996;77(5):910–6.
Ther. 2015;14(11):2441–51.
Immunohistochemistry in Breast Cancer
7
Ping Tang, Marilyn M. Bui, and Yan Peng
List of Frequently Asked Questions 2. Why is ER/PR/HER2 testing called a companion test?
1. Why does ductal carcinoma in situ (DCIS) need only A companion test is a diagnostic test used to determine if a
ER and PR testing, while invasive breast carcinoma specific patient is applicable for a therapeutic drug. One of
needs ER, PR, and HER2 testing? the most common testing methods is by immunohistochem-
istry (IHC). These IHCs are not ordinary, because the results
Breast cancer is the most common cancer in women. have prognostic and predictive implications. For example,
Estrogen receptor (ER), progesterone receptor (PR), and ER- and PR-positive patients have better outcome and sig-
human epidermal growth factor receptor 2 (HER2) status not nificantly benefit from hormonal therapy. In contrast, HER2-
only provide prognostic information but also are critical pre- positive patients have poor outcome and significantly benefit
dictive markers for currently available anti-hormonal and from HER2-targeted therapy.
anti-HER2 therapies. DCIS patients are treated with hor- IHC studies of biomarkers ER, PR, and HER2 are com-
monal therapy; thus, knowing the ER and PR status is criti- monly used companion tests for breast cancer treatment.
cal. Although up to 40% of DCIS can be positive for HER2, These tests are typically performed on formalin-fixed
there is no evidence that DCIS will benefit from Herceptin paraffin-embedded tissue. Unlike other conventional IHC
(Genentech, San Francisco, CA, USA); that is why HER2 tests that serve solely as an adjunct to a diagnosis, cancer
testing is not needed for DCIS currently. It is the standard of biomarker testing by IHC evaluates tumor biology and
care that invasive breast carcinoma that are HER2 positive molecular pathways driving disease progression that can
are treated with HER2-targeted therapy. HER2 positivity can potentially be targeted for therapy. Unlike other conventional
be HER2 protein overexpressed and/or HER2 gene ampli- IHC tests, in which the score is typically binary—that is,
fied. Therefore, HER2 testing is required for all primary and positive or negative—the scoring of breast cancer biomark-
recurrent invasive breast cancers. ers is quantitative and more involved. Accurate breast cancer
biomarker testing is critically important for patient care.
However, 20% of ER, PR, and HER2 testing results world-
wide are inaccurate. Thus, one should not assume that all
breast cancer biomarker testing results are accurate, unless
P. Tang the laboratories performing IHC assays for breast cancer bio-
Department of Pathology and Laboratory Medicine, markers closely follow the updated quality control and qual-
Loyola University Medical Center, Maywood, IL, USA ity assurance measures outlined in published guidelines
recommended by the American Society of Clinical Oncology/
M. M. Bui College of American Pathologists (ASCO/CAP).
Department of Pathology, Moffitt Cancer Center, Tampa, FL, USA
3. How does one evaluate ER/PR testing, and what is the
Y. Peng (*)
cut-off for ER/PR positivity?
Department of Pathology, Clements University Hospital,
Dallas, TX, USA
ER and PR expression are routinely tested in breast cancers
Medical Center, Dallas, TX, USA as prognostic and predictive markers. These receptors are
e-mail: [email protected] expressed in about 75–80% and 65% of all breast tumors,

174 P. Tang et al.
a b
Fig. 7.1 (a, b) ER in breast carcinoma cells shows both nuclear and cytoplasmic staining (a) and only cytoplasmic staining (b)
respectively. The ASCO/CAP guidelines recommend that 5. How does one troubleshoot if ER/PR results are in
ER and PR should be considered positive if ≥1% of tumor question?
cells shows nuclear staining of any intensity [1]. The inten-
sity of ER and PR stains should be included in the pathol- Normal breast elements should show ER expression in
ogy report as weak (1+), moderate (2+), or strong (3+). The 10–20% or higher of epithelial cells, and this serves as posi-
evaluation of normal breast ductal epithelium as an internal tive internal controls for evaluation of hormone receptor sta-
positive control is an integral part of the IHC evaluation for tus. The presence of the positive internal controls indicates
ER and PR expression in breast cancer specimens. that the tissue is adequate for the evaluation, which is par-
Cytoplasmic staining should not be considered as positive ticularly important in ER- and PR-negative tumor cases. If
staining. See Fig. 7.1a, b. These guidelines also emphasize the internal normal breast epithelial cells are not stained
standardization and quality assurance that must be followed properly with ER and/or PR, the test should be repeated. If
to help ensure testing accuracy. Allred scoring and the absence of staining in normal breast epithelial cells per-
H-scoring are also two commonly used systems for ER and sists in a repeated test, the test should be signed out as inde-
PR reporting [2, 3]. terminate. The cold ischemic time and formalin fixation time
for each test should be documented as recommended by the
4. How does one correlate ER/PR results with clinico- ASCO/CAP guidelines, which are critical for troubleshoot-
pathological characteristics of breast cancer? ing for pre-analytical variables.
One slide with external positive controls is highly recom-
Expression of ER/PR plays a major role in tumor develop- mended for ER, PR, and HER2 tumor biomarker testing,
ment in hormonal receptor-positive tumors and drives dis- which helps troubleshooting issues at the analytical stage.
ease progression in these tumors; thus, ER/PR-positive Also, due to the marked intratumoral heterogeneity of breast
breast cancer is eligible for endocrine therapy. Clinically, cancer, a negative ER and PR result in a core biopsy speci-
ER/PR-expressing invasive breast cancers are usually better men should prompt a repeat test in the subsequent resection
differentiated and have a more indolent course and favorable specimen if the status of biomarker expression does not fit in
prognosis compared to ER/PR-negative tumors. However, the clinical pictures and pathological features of the tumor,
sometimes hormonal receptor-positive breast cancers can be for example, in the case of a negative ER and PR result in a
poorly differentiated and can have aggressive clinical behav- grade 1 invasive lobular carcinoma, classic type.
ior. There is a positive correlation between the likelihood of
tumor response to hormonal therapy and the levels of ER/PR 6 . What subtypes of breast cancer are usually positive
expression. However, on some occasions, tumors with very for ER?
low levels of ER/PR expression can show significant response
to hormonal therapy that is above that of entirely receptor- Low-grade invasive ductal carcinoma (IDC) NOS, classic
negative tumors. invasive lobular carcinoma, tubular carcinoma, and muci-
7 Immunohistochemistry in Breast Cancer 175
nous carcinoma are almost always strongly positive for ER 9. Is PR result important for treatment of breast
and PR. Of note, apocrine carcinoma can be ER positive or cancer?
negative. Using unsupervised clustering analysis, five intrin-
sic subtypes of breast cancer were identified: luminal A, Both ER and PR are a transcription factor. PR is largely regu-
luminal B, normal breast-like, HER2-enriched, and basal-like lated by ER [4] and to some degree by growth factors. Given
subtypes. Each is unique in incidence, patterns of recurrence that PR is regulated by an active ER pathway, PR is co-
and survival, and response to therapy. Among them, luminal- expressed with ER in most of luminal-type breast cancer
type breast cancers are largely ER/PR positive. cases. PR is expressed in 55–65% of invasive breast carcino-
mas. The loss of PR expression in ER-positive tumors is
7. Can low-grade (grade 1) invasive breast carcinoma be associated with a worse prognosis and decreased response to
ER/PR negative? anti-estrogen therapy [5]. Therefore, PR is considered as an
important prognostic marker for breast cancer patients, and
Yes, some special types of low-grade breast cancers can be routine PR testing is necessary in assisting endocrine treat-
ER/PR negative. Low-grade invasive ductal and lobular car- ment decision making.
cinomas are usually ER/PR positive; but rarely, when these Of note, both ER and PR can show immunophenotypic,
tumors demonstrate some degree of apocrine features, ER/ intratumoral heterogeneity; thus, obtaining an adequate sam-
PR can be negative. See Fig. 7.2a, b. Other low-grade inva- ple from a tumor tissue for their testing is very important.
sive carcinomas that are ER/PR negative include but not lim- See Fig. 7.3a, b.
ited to breast cancer resembling the tall cell variant of
papillary thyroid carcinoma, secretory carcinoma, adenoid 10. Do ER-negative/PR-positive tumors exist?
cystic carcinoma, acinic cell carcinoma, malignant adeno-
myoepithelioma, low-grade adenosquamous carcinoma, and Although it has long been controversial, recent studies
carcinoma arising from microglandular adenosis. revealed that ER-negative/PR-positive (ER-/PR+) tumors do
exist as a distinct subtype of breast cancer and have a reported
8. When should one communicate with clinicians regard- incidence of 1–4% [6, 7]. A study showed that no significant
ing ER/PR results? survival advantage was found between the ER+/PR- and
ER-/PR+ tumors; furthermore, a higher PR expression was
Direct communication with the clinician is critical in the era associated with a favorable relapse-free survival and disease-
of multidisciplinary approach in treating breast cancer. If the specific survival in patients with ER-/PR+ tumors [6].
status of ER/PR expression does not fit in with a patient’s
clinical and pathological pictures, such as low-grade invasive 11. Why use HER2 testing for breast cancer patients?
ductal carcinoma NOS or classic invasive lobular carcinoma
with negative ER/PR, communicating with the clinician The human epidermal growth factor receptor 2 (HER2) is a
directly is warranted for reconciliation. member of a family of transmembrane tyrosine kinase recep-
a b
Fig. 7.2 (a, b) Low-grade invasive lobular carcinoma with apocrine features (a) shows negative ER immunostain (b)
176 P. Tang et al.
a b
Fig. 7.3 (a, b) Different immunostaining patterns of intratumoral heterogeneity of PR (a and b)
tors that plays an important role in the regulation of cellular able HER2 testing to help ensure that the right patients
signaling that affects cell growth, differentiation, and sur- receive the right treatment is now more critical than ever.
vival. HER2 protein overexpression and HER2 gene amplifi-
cation are in 10–20% of invasive breast cancers and have an 12. How does one evaluate HER2 IHC testing based on
important bearing on prognosis, as HER2-positive breast the ASCO/CAP guidelines?
cancer is associated with an aggressive clinical course and
poor outcome. HER2 status of a breast cancer provides both In 2007, a Joint Expert Panel assembled by the American
prognostic and predictive information for clinicians, guiding Society of Clinical Oncology/College of American
their treatment planning [8]. Trastuzumab (Herceptin, Pathologists (ASCO/CAP) met to develop and publish
Genentech, San Francisco, CA, USA) is well known to be guidelines with the aim of improving the quality, consis-
used for treatment of Her2-amplified/overexpressing inva- tency, and reliability of HER2 testing in clinical samples
sive breast carcinoma. Since the approval of trastuzumab for from breast cancer patients [10]. According to the guide-
HER2-positive metastatic breast cancer in 1998, outcomes lines for HER2 testing in breast cancer, indicators for anti-
for patients diagnosed with this aggressive cancer have vastly HER2 therapy are HER2 protein overexpression by IHC
improved. However, trastuzumab has some serious side (score 3+) and HER2 gene amplification by in situ hybrid-
effects including severe headache, fast or pounding heart- ization (ISH). HER2 positivity can be seen in both invasive
beat, and easy bruising or bleeding. Pertuzumab (Perjeta, and in situ breast carcinoma components in tumor sam-
Genentech, San Francisco, CA, USA) is a new anti-HER2 ples; only positive HER2 result in invasive carcinoma por-
drug binding to a different domain of HER2 gene. It is tion should be reported and used for treatment decision
approved for use in combination with trastuzumab and che- making. See Fig. 7.4a, b.
motherapy for use prior to surgery (neoadjuvant treatment) The 2013 HER2 guideline recommended changes to the
in patients with HER2-positive locally advanced, inflamma- testing algorithm and pathologist interpretation criteria and
tory, or early-stage breast cancer (tumor is greater than 2 cm added new language on reflex and/or repeat testing when
in diameter or node positive) and use after surgery (adjuvant there is an apparent histopathologic discordance with the test
treatment) in patients with HER2-positive early breast can- result [11]. The 2013 guideline update also advocates inter-
cer that has a high likelihood of recurrence. preting the HER2 results in the context of the clinical and
Rate of pathologic complete response (pCR) in HER2- morphologic features of the patient’s breast cancer, and the
positive breast cancer after the neoadjuvant therapy can guideline further recommends that pathologists and oncolo-
reach greater than 60% [9]. gists should exercise clinical judgment with respect to which
Accurate determination of HER2 status is critical to guide patients will require additional testing before the HER2 sta-
anti-HER2 therapy for improving breast cancer outcomes tus can be assuredly determined [11]. In the 2018 guideline
and avoiding overtreatment. focused update, uncommon clinical scenarios that are of
Given the fact of continued expansion of options for tar- uncertain tumor biologic or clinical significance are
geting the HER2 pathway in breast cancer, accurate and reli- addressed [12].
a b
Fig. 7.4 (a, b) One tumor case shows both invasive ductal carcinoma and DCIS (a) and HER2 protein overexpression (score 3+) in both compo-
nents (b)
13. What are the main updates on HER2 testing in the • HER2/CEP17 ratio <2, average HER2 copy ≥4.0 and
2018 HER2 guideline recommended by the ASCO/ <6.0
CAP? In these situations, concomitant IHC review is
required to achieve the most accurate HER2
The 2013 guideline [11] indicated either the HER2/CEP17 interpretation.
ratio or HER2 copy number can be used for reporting HER2 4 . The expert panel recommends all single-probe ISH assays
FISH results: to be reviewed with concomitant IHC.
• Ratio ≥2.0 or HER2 copy number ≥6 shall be considered Compared to the 2013 guideline, the 2018 HER2 guideline
HER2 positive/amplified. includes more rigorous interpretation criteria for dual-probe in
• Ratio <2 and HER2 copy number <4 shall be considered situ hybridization (ISH) testing and requires concomitant IHC
HER2 negative. review to arrive at the most accurate HER2 status designation
• Ratio <2 and HER2 copy number ≥4.0 and <6.0 shall be (positive or negative); equivocal ISH cases no longer exist,
considered HER2 equivocal. which would better guide anti-HER2 therapy. IHC seems to
play a more important role in determining final HER2 status
Please note the minimal cell count required for HER2 compared to ISH; furthermore, HER2 protein overexpression
FISH analysis is 20 invasive tumor cells. appears to be more important for targeted therapy than HER2
In the 2018 guideline focused update [12], recommenda- gene amplification. In conjunction with the 2018 guideline
tions include the following: [12], we recommend the HER2 testing clinical practice in
invasive breast cancer as follows (Tables 7.1 and 7.2):
1. The revised definition of 2+ HER2 IHC (equivocal) is
weak to moderate complete membrane staining observed • Perform HER2 IHC first for all invasive carcinomas as a
in >10% invasive breast cancer cells. “screening” tool.
2. Instead of “must,” a new HER2 test “may be” ordered on • If IHC 3+ (Fig. 7.5a) = HER2 positive (no ISH), that is
the excision specimen based on specific clinical criteria, defined as circumferential membrane staining that is
if the initial core needle biopsy specimen of a primary complete and intense in more than 10% of tumor cells.
invasive breast cancer is negative. • If IHC 2+ (Fig. 7.5b) = HER2 equivocal, per the 2018
3. In less than 5% of the cases, dual-probe ISH assay may updated guideline [12], that is defined as tumor cells with
result in some uncommon scenarios: weak to moderate complete membrane staining observed
• HER2/CEP17 ratio ≥2.0, average HER2 copy <4.0 in greater than 10% of tumor cells; the HER2 IHC equiv-
signals per cell ocal result requires reflex ISH test (with or without
• HER2/CEP17 ratio <2.0, average HER2 copy ≥6.0 concomitant IHC review) to determine final HER2 status
signals per cell (positive or negative).
178 P. Tang et al.
Table 7.1 Suggested clinical practice on HER2 Testing in conjunction

with the 2018 Guideline. Starting with Her2 IHC as the “screening” tool a
IHC ISH HER2 status
3+ No Positive
2+ (equivocal) Reflex ISH Pending
0 or 1+ No Negative
Table 7.2 Suggested clinical practice on HER2 testing in conjunction

with the 2018 guideline for IHC 2+ cases requesting reflex ISH only
HER2/ Average
CEP17 HER2 copy
ratio number/cell Further work-up HER2 status ± comment
≥2.0 >4.0 No Positive
≥2.0 <4.0 Observer blinded Negative because of low
to previous copy number and lack of
results recounts protein overexpression.
ISH in at least 20 Evidence is limited on
cells = same the efficacy of HER2-
b
result targeted therapy
<2.0 ≥6.0 Recounts Positive
ISH = same
result
<2.0 ≥4.0 Recounts Negative because of
and < 6.0 ISH = same absence of protein
result overexpression. It is
uncertain that this group
of patients benefit from
HER2-targeted therapy
<2.0 <4.0 No Negative
• IHC 0 or 1+ (Fig. 7.5c) = HER2 negative (no ISH); the

IHC 0 is defined as no staining observed or membrane
staining that is incomplete, faint/barely perceptible, and
within ≤10% of tumor cells; the IHC 1+ is defined as c
incomplete membrane staining that is faint/barely percep-
tible and within >10% of tumor cells.
14. Are there invasive breast carcinoma cases with nega-

tive HER2 IHC and positive HER2 ISH?
Although it is rare, breast cancers showing HER2 gene

amplification but no HER2 protein overexpression have been
reported, particularly in high-grade invasive ductal carci-
noma [13]. See Fig. 7.6a, b.
15. What are key elements for HER2 IHC quality con-
trol and quality assurance?
Fig. 7.5 (a–c) Positive HER2 IHC 3+ (a), equivocal HER2 IHC 2+
(b), and negative HER2 IHC 1+ (c)
HER2 IHC quality control and quality assurance are critical
for accuracy of HER2 testing and should be conducted in
pathology laboratories on a regular basis. The key elements • Meet optimal tissue handling requirements (pre-
for the process include but not limited to the following: analytical factors) recommended by the ASCO/CAP (See
question 23).
• Use an FDA-approved IHC antibody, such as Ventana • Be aware that decalcification treatment for bone samples
(Oro Valley, AZ, USA) HER2 (4B5) or Genentech (San may affect Her2 IHC expression in bone metastatic breast
Francisco, CA, USA) HercepTest. cancer. If there is a discrepancy in Her2 IHC results
a b
Fig. 7.6 (a, b) Invasive carcinoma shows HER2 IHC negative (score +1) (a) and FISH positive (b)
between primary breast tumor and bone metastasis, a of the IHC, the entire tumor on a whole slide stained by IHC
diagnostic note with possible explanation should be in the can be reviewed to ensure invasive carcinoma areas are eval-
report. uated and to identify HER2 tumor heterogeneity immuno-
• Review and document external and internal controls with phenotypically. For FISH test, the key element is to identify
each test and each batch of tests. areas containing invasive tumor cells to count average copy
• Perform ongoing competency assessment and document number for HER2 gene and chromosome 17-centromere,
the actions taken as a part of the laboratory record. Per the respectively. However, the identification can be challenging
CAP recommendation and guideline, the labs should con- under the fluorescent microscope, because in situ carcinoma
duct an annual evaluation of interobserver variability and others can resemble invasive carcinoma. This is the
among pathologists who practice Her2 IHC major issue for FISH test. Thus, selecting tissue blocks with
interpretation. invasive tumor cells and marking the tumor areas for FISH
test are critical for getting correct results.
16. When should a reflex HER2 FISH testing be ordered? 18. What are chromosomal monosomy and chromo-
somal polysomy?
The HER2 guideline recommended reflex HER2 FISH test-
ing is only requested in HER2 IHC equivocal (2+) cases. Monosomy of chromosome 17 can make HER2/CEP17 ratio
However, pathologists shall take other clinical and patho- artificially high, and polysomy of chromosome 17 can make
logical information into consideration when making a deci- the ratio artificially low. Thus, interpretation of HER2 FISH
sion on ordering HER2 FISH test. For example, invasive results in these types of cases should be undertaken with cau-
breast micropapillary carcinoma cells often show U-shape, tion. For these situations, the 2018 HER2 guideline [12] rec-
weak, membranous HER2 staining; in the 2018 HER2 guide- ommended assessing IHC using sections from the same
line, it has been suggested to reflex a HER2 FISH test in such tissue sample used for FISH.
cases, especially when PR is low and Ki67 is high in the
tumor cells [12]. 19. How does one correlate HER2 result with clinico-
pathological characteristics of breast cancer?
17. What are the pros and cons of HER2 tests by IHC
and FISH? HER2-positive breast cancers are usually high-grade inva-
sive carcinomas, including but not limited to tumors with
Both HER2 IHC and FISH tests have their pros and cons, apocrine features, invasive micropapillary carcinoma, and
respectively. One of the major cons of the HER2 IHC is lack pleomorphic invasive lobular carcinoma. The levels of Ki67
of internal positive controls. In this situation, it is not certain expression in these tumors are usually greater than 15% but
if the tumor tissues have any immune reaction in a setting of less than 50–80% that are commonly seen in triple-negative
negative HER2 result. However, if HER2 IHC test runs breast cancer (TNBC) cases. If the results between HER2
together with ER/PR IHC and internal positive controls for and Ki67 are discordant, further investigation is warranted
ER/PR are present, this shall not be an issue. In terms of pros before releasing the pathology report.
180 P. Tang et al.
20. What are the most common causes for a discrepancy 4. Most importantly, the cold ischemic time and fixation
in HER2 IHC and FISH results? time must be documented in the pathology report for
reporting breast cancer biomarker results, which are sub-
While IHC evaluates the entire tumor tissues present on a jected to lab accreditation inspection.
whole slide for the status of HER2 protein overexpression, 5. Any exceptions on the handling must be mentioned in the
FISH usually only counts 20–40 invasive tumor cells for the pathology report.
status of HER2 gene amplification. Due to the inherent issues
of tumor heterogeneity in breast cancer, these two test meth-
ods may provide different results even though they are done 24. What is Ki67 testing?
in the same tissue block. Thus, it has been recommended to
use HER2 IHC slide as the guide map for FISH test to deter- Ki67 has been used as a proliferation marker. In general, high
mine which areas to count, rather than blindly counting. Ki67 expression is associated with poor differentiation and
prognosis, and low Ki67 expression is associated with better
21. How does one troubleshoot if HER2 IHC and FISH differentiation and prognosis for breast cancer. Ki67 expres-
results are discordant? How does one communicate sion level in breast cancer is prognostically and predictively
with clinicians on cases with HER2 discrepancy? important. The most widely used Ki67 testing involves an
IHC assessment of Ki67 antigen, also known as MKI67,
If HER2 results do not fit the clinical pictures, investigation which is a nuclear protein expressing in all cells except those
should be conducted for all possible variables: pre-analytic, in G0 phase.
analytic, and post-analytic factors. Repeat IHC and/or FISH
tests on the same block, increase the number of cells counted 25. Is Ki67 a useful marker for breast cancer, and when
for FISH analysis, or send the block to a reference lab for test- should it be tested?
ing: all these steps should be considered, and then appropriate
actions should be taken. If HER2 result discrepancy persists Although numerous data have shown that Ki67 is a useful
after repeating, the discrepancy should be clearly documented prognostic marker for breast cancer, it has not been recom-
in the pathology report. Direct communication with the clini- mended by the ASCO/CAP as a required biomarker for
cian on the results via the phone is highly recommended, as it breast cancer, like ER, PR, and HER2. However, Ki67 has
shall be a multidisciplinary decision on what is the most been routinely tested together with ER, PR, and HER2 for
appropriate management approach for this group of patients. breast cancer cases in many institutions at the request of cli-
nicians. Ki67 has also been included in many predictive
22. Can low-grade (grade 1) invasive breast carcinoma models for breast cancer, such as Magee Equations [16] pro-
be HER2 positive? posed by Klein and colleagues in 2013 and IHC4 scores [17]
proposed by Cuzick and colleagues in 2011. The PEPI score
Yes. Yu and colleagues [14] have shown that classic-type originally reported in 2008 [18] and recently modified in
grade 1 invasive lobular carcinomas can be rarely positive for 2017 [19] included Ki67 evaluation to better predict the
HER2, especially when they have apocrine or histiocytic fea- tumor response to neoadjuvant endocrine therapy in
tures or they are PR negative. ER is still expressed in these ER-positive breast cancer. Interestingly, it is frequently
HER2-positive low-grade invasive carcinomas, although the observed that Ki67 expression is higher in invasive carcino-
level of expression is significantly lower than that of HER2- mas without co-existing in situ carcinoma compared to those
negative low-grade tumors. with in situ carcinoma component. In cases with both inva-
sive and in-situ carcinoma components, Ki67 expression in
23. What are pre-analytical factors affecting breast can- the invasive carcinoma is usually higher than that in the in-
cer biomarker results? situ carcinoma. See Fig. 7.7a–d.
The ASCO/CAP guidelines on ER, PR, and HER2 testing 26. Why is Ki67 not recommended for routine testing for
[15] recommended optimal breast tissue handling, including breast cancer?
the following:
Per the ASCO Practice and Guidelines in 2016 [20], IHC for
1. Time from tissue acquisition to fixation should be as short Ki67 analysis lacks reproducibility across laboratories and,
as possible (less than 1 hour). therefore, cannot be consistently interpreted when performed
2. Tissue fixation in 10% neutral buffered formalin (NBF) in a broad range of laboratories. IHC for Ki67 is not recom-
with minimum fixation time of 6 hours and maximum mended for broad clinical use to determine whether a patient
fixation time 72 hours. should receive chemotherapy.
3. Core needle biopsy specimen is preferred for testing to At present, the enormous variation in analytical practice
avoid prolonged tissue fixation. markedly limits the value of Ki67 as a prognostic marker for
a b
c d
Fig. 7.7 (a–d) Ki67 expression in DCIS (a, b) and invasive ductal carcinoma (c, d) from the same tissue specimen, respectively
breast cancer. There are two major issues that need to be 28. How does one evaluate Ki67 testing manually and by
solved before Ki67 can be recommended for routine use: image analysis, respectively?
First is testing variability between laboratories. Thus, at the
St. Gallen International Expert Consensus meeting in 2015, In 2011, Dowsett and coworkers [23] proposed an interna-
experts suggested that each lab use its own cut-off for Ki67 tional working group recommendation for Ki67 evaluation.
[21]; for example, if 20% is the mean value for Ki67 for MIB1 antibody for Ki67 testing is recommended. Three
ER-positive tumors in a lab, then 30% is considered high random fields of a tumor section, 500–1000 tumor cells shall
Ki67 and 10% is low Ki67. Second is interpretation variabil- be counted and scored, using average across the section, not
ity of Ki67 scoring by manual assessment or via automated, edge or center, including cells stained with all levels of inten-
quantitative image analysis (QIA) system [22], as stromal sity. For image analysis, pathologists’ confirmation for tumor
cells, inflammatory cells, and sometimes even adipocytes areas scored is critical; it is important to avoid areas with stro-
stain Ki67. See Fig. 7.8a, b. mal cells and inflammatory cells, areas with prior biopsy site
changes, and areas of necrosis for scoring. Since the recom-
27. Is there any difference on Ki67 expression in breast mendation was published, numerous studies have investigated
cancers before and after chemotherapy or endocrine the ways to increase concordance in Ki67 scoring between
therapy? laboratories [24].
Generally speaking, in most cases with neoadjuvant chemo- 29. How does one correlate Ki67 result with clinicopath-
therapy, Ki67 expression levels are decreased in residual ological characteristics of breast cancer?
tumors. See Fig. 7.9a, b. Also, high Ki67 after endocrine
therapy is considered a poor prognostic factor for ER-positive Low-grade carcinoma usually has low Ki67 expression,
breast cancer by the modified PEPI score [19]. while high-grade carcinoma tends to have high Ki67.
182 P. Tang et al.
a b
Fig. 7.8 (a, b) High-grade invasive ductal carcinoma with prominent tumor-infiltrating lymphocytes (a), making assessment of Ki67 expression
in tumor cells difficult (b)
a b
Fig. 7.9 (a, b) High-grade invasive ductal carcinoma (a) with low Ki67 expression (b) post-neoadjuvant chemotherapy
However, apocrine carcinoma tends to have low Ki67 even case. Normal breast tissue (both epithelium and stroma)
when the tumor is high grade. 15–20% Ki67 has been used shows about 5% Ki67 labeling. Lack of internal Ki67
as a cut-off value for favorable vs. poor prognosis for inva- labeling indicates the tissue sample is suboptimal for
sive ductal carcinoma, largely because this is the cut-off evaluation of the test. Also, a proper Ki67 staining in dif-
value used to separate luminal B subtype breast cancer from ferent tissue samples serving as external positive controls
luminal A subtype [25]. Recently, Carbognin and coworkers is important for the assessment.
have proposed that 5% is a biologically meaningful cut-off
for Ki67 for invasive lobular carcinoma [26]. 31. Why use quantitative image analysis to score breast
cancer biomarker IHC testing?
30. How does one do troubleshooting if Ki67 result is in
question? The goal of quantitative image analysis (QIA) is to achieve
accurate, precise, and reproducible results of breast cancer
Ki67 is very sensitive with pre-analytic variables; thus, biomarker testing. Validation of an FDA-approved algorithm
the presence of adequate internal positive controls is the is an important step to ensure the algorithm delivers the
key to determine if this test works properly in a given intended result. The current guideline in validation is general
Fig. 7.10 The illustration shows this breast cancer sample is positive for ER in 97.63% cells
and non-specific. It is the medical director’s responsibility to 33. What percent of triple-negative breast cancer cases
establish the lab’s own standard operation procedure adher- are GATA3 positive?
ing to the CAP guideline. The parameters to consider vali-
dating include system, test, pathologist, and result. Using The sensitivity of GATA3 in triple-negative breast cancers
HER2 as an example, the gold standard for the validation is (TNBCs) is significantly lower than that in ER-positive
an alternate validated method including consensus scoring of breast cancers, and the sensitivity ranges from 43% to 66%
HER2 IHC, fluorescence in situ hybridization (FISH) result, [29–31]. Recent studies demonstrated that GATA3 is a much
bright-field chromogenic in situ hybridization (CISH) result, more sensitive marker for TNBC as compared with mam-
and previously validated QIA result. Histotechnologists are maglobin and GCDFP-15, which were commonly used
also critical team members in the delivery of accurate and breast markers previously. Of note, different clones of
reliable breast cancer biomarker quantitative results to the GATA3 antibodies have different sensitivities on breast duc-
clinical team for effective management of the patients. The tal epithelium. A study reported that GATA3-L (L50-823)
billing code for QIA is CPT 88381 which is different from was diagnostically more sensitive than GATA3-H (HG3-31)
that for manual reading (CPT 88380). An example of an in TNBC [30].
FDA-approved quantitative image analysis platform is illus- GATA3 is particularly useful in identifying metastatic
trated in Fig. 7.10. TNBCs as ER, PR, or HER2 by IHC cannot serve as markers
for the detection and other breast markers have a lower
32. What percent of ER-positive breast carcinomas are sensitivity.
GATA3 positive?
34. What non-breast tumors are also GATA3 positive?
GATA3, a zinc-binding transcription factor, is a novel,
sensitive, and relatively specific marker for primary breast According to a recent comprehensive immunohistochemis-
carcinomas. Some studies revealed that GATA3 is try (IHC) study on GATA3 [32], in non-breast epithelial neo-
expressed in up to 94% of ER-positive breast cancers [27, plasms, GATA3 was expressed in urothelial carcinomas
28]. Due to the high sensitivity and specificity of GATA3 (>90%), skin adnexal tumors (100%), mesotheliomas (58%),
for breast origin, currently GATA3 has been widely used chromophobe renal cell carcinomas (51%), and salivary
as a breast marker in routine clinical practice to confirm a gland (43%) and pancreatic (37%) ductal carcinomas,
breast primary tumor and to identify a metastatic carci- whereas frequency of expression in adenocarcinomas of the
noma of breast origin. lung, stomach, colon, endometrium, ovary, and prostate was
184 P. Tang et al.
<10%. Among mesenchymal and neuroectodermal tumors, types of TNBCs were identified using gene expression pro-
paragangliomas were usually positive, which differentiates filing; they were basal-like (BL1 and BL2) which is the
these tumors from epithelial neuroendocrine tumors. GATA3 major subtype, immunomodulatory, mesenchymal, mesen-
is a useful marker in characterization of not only mammary chymal stem-like, luminal androgen receptor (LAR), and
but also urothelial and renal tumors, mesotheliomas, para- unstable subtypes [38]. The AR-expressing TNBCs were
gangliomas, and other origins of tumors. defined as the LAR subtype [38], which may have a better
prognosis compared to other molecular subtypes of TNBC.
35. Can a TNBC be ruled out if GATA3 is negative in an Previous studies on the AR expression in TNBCs have
unknown primary immunohistochemistry work-up? shown that AR negativity is associated with a shorter disease-
free interval and overall survival compared to AR-positive
Since only up to around 70% of TNBC tumors are GATA3 TNBCs [40, 42–44]. Recently, a group reported that, among
positive, GATA3 negativity result in an unknown primary AR-positive TNBCs, cases with distant metastases (patho-
carcinoma cannot exclude the possibility of a metastatic logic staging: pM1) were significantly associated with a
TNBC. Clinical and radiographic correlation is essential to lower intratumoral expression of AR protein as compared to
accurately determine the primary site. cases without distant metastasis or only having loco-regional
disease (pathologic staging: pM0) [37]. The study also dem-
36. Can breast cancers express androgen receptor (AR)? onstrated that the AR expression inversely correlated with
If so, is there any difference on frequency of andro- Ki67. This result suggests that TNBCs with higher AR
gen receptor expression between ER-positive breast expression may be associated with a better prognosis, partly
cancer and TNBC? due to decreased tumor cell proliferation caused by the
increased anti-proliferative effect of androgen receptor stim-
The androgen receptor (AR) is widely expressed in breast ulation. These findings indicate AR could be used as a prog-
cancers. AR expression has prognostic implications in breast nostic marker for TNBCs.
cancers; higher AR expression levels have been associated
with higher expression of ER or PR, lower nuclear grade, 39. Is there any therapeutic value of AR expression for
and smaller tumor size with lower risk of recurrence and patients with TNBCs, particularly those with meta-
death [33, 34]. Significant differences in AR protein expres- static/recurrent TNBCs or chemotherapy-resistant
sion have been found in different molecular subtypes of TNBCs?
breast cancer. However, the role of AR in breast cancers
remains uncertain, particularly in ER-positive tumors. The AR is a clinical target for targeted therapy for treating
Enzalutamide, an AR inhibitor that impairs nuclear localiza- prostate cancer. Recent research indicates that it is an emerg-
tion of AR, was used to elucidate the role of AR in women ing hormonal target in breast cancer as well, with potential
with ER-positive and ER-negative breast cancers [35]. clinical benefit in both ER-positive and ER-negative tumors
[45]. The AR signaling pathway may represent a molecular
37. What is AR expression status in ER-positive breast driver that can be therapeutically targeted in AR-expressing
cancer? TNBCs by utilizing AR inhibitors. Several studies have
shown promising results for AR antagonists. One such study
AR is expressed in up to 90% of ERα-positive tumors. demonstrated that the LAR cell lines of TNBC were uniquely
Although multiple studies suggest that AR is associated with sensitive to bicalutamide, an AR antagonist [38]. In mouse
a favorable prognosis, AR overexpression and, in particular, models of TNBC with a low percentage of AR-positive
high AR/ER ratio seem to be involved in resistance to hor- tumor cells, an anti-androgen drug, enzalutamide, was sig-
monal treatment [36]. nificantly effective in reducing proliferation, growth, migra-
tion, and invasion of the cancer cells. This study also showed
38. What is AR expression status in TNBC? that only a mere 1% of TNBC cells in a tumor must be AR
positive to show benefit from AR-targeted therapies [46].
TNBCs that are by nature ER negative characteristically Anti-androgen therapy has been proposed as a targeted ther-
have a much lower frequency of AR expression compared to apy in women with TNBC. This may be of particular interest
ER-positive breast cancers [33, 34, 37, 38]. A recent study for those who do not respond to conventional chemotherapy
revealed that only 31% of TNBC patients expressed AR by or even as an addition to first-line therapy. Therefore, know-
IHC [37], which is similar to other studies reporting 25–35% ing the AR expression status of a TNBC tumor is necessary
positivity [33, 39–41]. for the selection of patients who may benefit from anti-
Recent studies by molecular analysis have demonstrated androgen therapy. Multiple ongoing clinical trials have
that TNBCs are a heterogeneous group of tumors. Six sub- shown promising preliminary results of AR-targeted thera-
pies against TNBC with a higher percentage of AR-positive 42. What IHC makers can be used to distinguish metastatic
cells [47]. We therefore propose testing for AR expression in p53+ p16+ TNBC to the gynecologic sites from high-
all metastatic/recurrent TNBCs and primary TNBCs that do grade serous carcinoma of the fallopian tube/ovary/peri-
not completely respond to chemotherapy in order to help toneum that is commonly p53 and p16 positive?
guide management decisions.
From a clinical practice standpoint, when dealing with an
40. What is p53 expression status in TNBC and the ER/PR/Her2-negative high-grade carcinoma presenting in
potential prognostic value of p53? the female reproductive tract in a patient with a history of
TNBC, the differential diagnosis should include metastatic
p53 acts to induce apoptosis and arrest the cell cycle in TNBC to the gynecologic site or a serous carcinoma present-
response to cellular stresses such as DNA damage. p53 is a ing as a second primary tumor. Conversely, metastatic high-
tumor suppressor protein, and its overexpression has been grade serous carcinoma to the breast or axillary lymph nodes
reported in the majority of TNBC tumors [48]. TNBCs can is not uncommon. However, the differential diagnosis can be
be divided into two subtypes based on their p53 expression problematic when these two different primary tumors have
status: p53-positive and p53-negative tumors [49]. From the an identical immunophenotypic profile, ER-/PR-/Her2−/
molecular perspective, TP53 gene mutations have been p53+/p16+, and similar morphologic features. For such
reported in the vast majority of TNBCs in recent publica- cases, we recommend using a differential IHC panel that
tions, and as such is a target of particular interest [48]. includes, but is not limited to, GATA3, PAX8, and WT-1.
A recent study demonstrated that p53 expression in Breast primary tumors are commonly GATA3+/PAX8-/
TNBCs with nodal metastasis was significantly higher than WT-1-, while serous carcinomas are GATA3-/PAX8+/
that in non-TNBCs with nodal metastasis [50]. It was WT-1+. Correlation with clinical and radiographic informa-
revealed that p53 expression had the strongest prognostic tion is essential for making a correct diagnosis.
significance in TNBC patients in a multivariate analysis [51]. Of note, both TNBC and high-grade serous carcinoma
It has also been demonstrated that TP53 mutations were can occur in patients with BRCA mutations, and both types
strongly predictive for relapse-free and overall survival in of carcinomas can co-overexpress p16 and p53. Making a
TNBC patients treated with adjuvant anthracycline-based distinction between these two primary tumors has important
chemotherapy regimen [52]. p53-positive TNBCs have prognostic and therapeutic implications.
worse overall and disease-free survival compared with
p53-negative tumors [48, 53]. Since p53 overexpression is 43. What is the potential predictive value of p16 expres-
associated with a poorer survival, it has been suggested that sion in TNBC?
p53 can be used as a prognostic marker for TNBC.
To better understand TNBC tumor biology as related to the
41. What is p16 expression status in TNBC? What is the Rb/p16 pathway, a group of researchers investigated the
diagnostic utility of p16 in metastatic TNBC? relationship of p16 expression with Ki67 in TNBC cases
[49]. Ki67 is widely used as a prognostic and predictive
p16 normally acts as a cyclin-dependent kinase (CDK) inhib- tumor biomarker for breast cancer, and a higher expression
itor by inactivating CDK4/6 and preventing the phosphoryla- of Ki67 is associated with a more aggressive clinical
tion of retinoblastoma (Rb). p16 is a tumor suppressor protein, behavior. They found that the mean value of Ki67 expres-
which plays an important role in cell cycle regulation. sion was high (>60%) in both p53-positive and p53-nega-
p16 by IHC is widely used as a surrogate marker for high- tive TNBC tumors regardless of the status of p16
risk HPV infection in diagnosing high-grade dysplasia and expression; there was a significant positive correlation
squamous cell carcinoma of the cervix and the oropharynx. between p16 and Ki67 only in p53-negative tumors, but
Recent studies demonstrated that a strong and diffuse p16 not in p53-positive tumors. These findings suggest that p16
expression was seen in the majority of TNBCs [54, 55]. A may play a role in the poor outcomes of p53-negative
case series showed that 75% of TNBC tumors overexpressed tumors that provide insights into the potential prognostic
p16 protein and the expression was significantly higher than value of p16 for TNBC. The p16-related cell cycle path-
that in paired normal breast ducts (Fig. 7.11a–d) [49]. While way might be potentially targetable for treatment of
there are very limited reports in the literature, this finding is TNBCs, particularly for p53-negative tumors. Another
especially important for the pathologist when attempting to study reported that TNBC patients whose tumors showed
distinguish metastatic TNBCs from p16-positive carcinomas strong p16 immunoreactivity had complete pathologic
originating from non-breast sites, such as the fallopian tube/ response and that decreasing response correlated with
ovary/peritoneum since the majority of high-grade serous decreasing expression of p16 [56]. This finding suggests
carcinomas arising from these locations show p16 that the status of p16 expression may have a predictive
overexpression. value for tumor response to chemotherapy.
186 P. Tang et al.
a b
c d
Fig. 7.11 (a–d) p16 expression in a TNBC tumor and paired normal tissue. TNBC shows high-grade ductal carcinoma morphology (a) and dif-
fuse, strong p16 immunostaining (b); paired normal breast tissue shows benign ducts (c) and negative p16 staining (d)
44. What is basal cytokeratin expression status in 45. What are potential prognostic and predictive values
TNBC? of basal cytokeratins in TNBC?
Basal cytokeratins include CK5/6, CK5, CK14, and CK17; Recent studies have demonstrated that, among the histo-
and they can express in TNBC tumors. logic subtypes of breast cancer, basal-like TNBCs have sig-
TNBCs can be divided into basal-like and non-basal-like nificantly worse outcomes [57, 61–63]. A group of
subtypes according to basal cytokeratin expression status, researchers reported that approximately one-third of basal-
with the majority falling into the basal-like subtype [57, 58]. like TNBCs developed distant metastasis [59]. They found
CK5/6 is expressed in around 62% of TNBCs [59]. However, that TNBCs with metastases had significantly higher expres-
a recent study reported CK5 positivity in 97% of TNBC sion of CK5/6 as compared to cases without metastases.
tumors and suggested that CK5 is more sensitive than Their findings suggest that high-level expression of CK5/6
CK5/6 in identifying the basal-like tumors [60]. may be predictive of metastatic disease in TNBC patients.
From a diagnostic standpoint, immunoreactivity of basal While some reports have demonstrated that TNBCs have a
cytokeratins can be helpful in diagnosing ER-/PR-/Her2- good response to an adjuvant anthracycline-based chemo-
poorly differentiated invasive carcinomas in the breast as therapy, it has been shown that patients with a basal-like
basal-like TNBCs, particularly in the core biopsy specimens phenotype of TNBC had a significantly poorer response to
where in situ lesions may be absent. the chemotherapy [64].
46. What cancer stem cell markers are enriched in primary sites [75–77]. Distinguishing metastatic carcinoma
triple-negative breast cancer? of lung origin from primary breast carcinoma has significant
prognostic and therapeutic implications. However, some-
Recent studies suggest that cancer stem cells play an impor- times making the distinction can be diagnostically challeng-
tant role in tumorigenesis and tumor biology of TNBC. Both ing due to the following reasons:
CD44+/CD24− and ALDH1+ breast cancer stem cells are
enriched in TNBC compared to non-TNBC tumors and may • No or insufficient clinical and radiographic information
contribute to the propensity of TNBC for chemotherapy available
resistance and tumor metastasis. It has been proposed that • Absence of in situ carcinoma of the breast, such as DCIS,
targeting cancer stem cells may be a promising, novel strat- particularly in core biopsy specimens
egy for treatment of TNBC patients [65]. • Morphologic similarity of both adenocarcinomas, espe-
cially in high-grade tumors
47. Can breast carcinomas be positive for TTF-1 and • Overlapping immunoprofile between these two primary
Napsin A that commonly express in adenocarcinoma tumors
of the lung?
Morphologically, identification of in situ carcinoma of the
It has been well documented that the majority of lung adenocar- breast helps make a diagnosis of primary breast carcinoma.
cinomas are TTF-1 (75%) and Napsin A (87%) positive [66, 67]. The majority of primary breast carcinomas express ER and/
Recently published studies showed that both the sensitive or PR by IHC, and some overexpress HER2 protein. GATA3
lung markers can express in a small portion of breast ductal is a recently recognized sensitive and specific maker for
carcinomas; TTF-1 was positive in 2.4% from focal and breast primary tumors. TTF-1 and Napsin A are sensitive and
weak to diffuse and strong [68] and Napsin A in 3% [66]. specific markers for lung adenocarcinomas. In general, an
IHC panel including TTF-1, Napsin A, GATA3, and ER
48. Can lung adenocarcinomas be positive for GATA3, a serves as a useful ancillary tool in the differential diagnosis
sensitive breast marker? between these two primary tumors in the majority of cases.
Lung adenocarcinomas are commonly TTF-1+/Napsin A+/
GATA3 expresses in more than 90% of ER-positive luminal- GATA3-, and breast carcinomas are GATA3+/ TTF-1-/
type breast carcinomas and up to around 70% of TNBCs [27, Napsin A-. A recent study reported that a simple IHC panel
31]. Currently, GATA3 is widely used as one of the most of GATA3 and TTF-1 correctly differentiated breast carci-
sensitive breast markers in routine clinical practice. However, noma from lung adenocarcinoma in 93% of cases in their
a recent study reported that GATA3 expressed in 8% of lung series [78].
adenocarcinomas [69]. However, a small portion of breast and lung carcinomas
have overlapping immunohistochemical expression on ER,
49. Can lung adenocarcinomas be positive for ER? Her2, GATA3, TTF-1, and Napsin A [66–68, 70–72], which
makes a distinction between these two primaries diagnosti-
Recent studies demonstrated that a subset of lung adenocar- cally challenging. Moreover, the presence of ER, Her2, or
cinomas express breast tumor biomarkers by IHC. ER posi- GATA3 immunoreactivity in a carcinoma cannot by itself be
tivity was identified in approximately 5–27% of lung used to exclude the possibility of a lung origin. On the other
adenocarcinomas, and most of ER-expressing lung tumors hand, TTF-1 and/or Napsin A expression cannot rule out a
are positive for TTF-1 [70–72]. primary breast cancer.
For diagnostically challenging cases—such as meta-
50. Can lung adenocarcinomas be positive for HER2 by static TTF-1+ and/or Napsin A+ breast carcinoma to the
IHC? lung vs. primary lung adenocarcinoma or metastatic ER+
and/or Her2+ lung adenocarcinoma to the breast vs. pri-
Her2 protein overexpression or gene amplification was found mary breast ductal carcinoma—comprehensive analysis
in primary lung adenocarcinomas [73, 74], although this with clinical history and radiographic finding and compari-
phenomenon is uncommon. son of morphology between primary and metastatic tumors
are essential to distinguish breast primary from lung origin
51. How does one distinguish breast primary tumor from [79]. Furthermore, comparison of primary breast cancer’s
lung adenocarcinoma despite the presence of overlap- tumor profile (extent and intensity of immunoreactivity)
ping immunohistochemistry between these two tumors? with secondary tumor’s profile is one of the important clues
to identify ER+ extramammary metastasis in the breast in
Incidence of metastatic extramammary carcinomas to the patients with a prior history of ER- and/or HER2-positive
breast is up to 2%, and the lung is one of the most common breast carcinomas [80].
188 P. Tang et al.
52. What ER-positive carcinomas from Müllerian origin tor conversion by immunohistochemistry in (non-bone) distant
can metastasize to the breast? breast cancer metastases does occur, and the conversion was
mainly from positive in the primary tumor to negative in the
Gynecologic cancers can metastasize to the breast and axil- metastases for ERα and PR, while HER2 conversion occurred
lary lymph nodes, although this is uncommon. The majority equally both ways. The conversion rates on ERα and PR are
of metastatic carcinomas from the fallopian tube/ovary/peri- 15.1% and 32.6%, respectively. The conversion is relatively
toneum are high-grade serous carcinomas, and 64% of those uncommon for ERα and HER2 and is more frequent for PR,
tumors are ER positive [81, 82]. Metastatic endometrial car- especially in brain, liver, and gastro-intestinal metastases [85].
cinoma to the breast is extremely rare, and there are only a We suggest taking the possibility of receptor conversion
few case reports available in the literature [83, 84]. into consideration while dealing with a discrepancy of hor-
In an appropriate clinical, radiographic, and morphologic mone receptor results between primary and metastatic/recur-
context, a limited IHC panel can help distinguish breast pri- rent breast cancers to accurately interpret biomarker results
mary from gynecologic origin: GATA3 and PAX8. Breast to guide optimal patient management.
tumors are usually GATA3+/PAX8-, and GYN tumors are
GATA3-/PAX8+.
Case Presentations
53. Can hormone receptor and Her2 status be different
between primary breast cancer and paired meta- Case 1. ER heterogeneity in breast cancer (Fig. 7.12a–d)
static tumor? • History: A 65-year-old female with a mass lesion on screen-
ing mammogram underwent a core needle biopsy
A group of researchers from the Netherlands studied receptor • Histology on the biopsy: Invasive ductal carcinoma, not
conversion for ERα, PR, and HER2 [85]. They found that recep- otherwise specified (NOS), histologic grade 2
a b
c d
Fig. 7.12 (a–d) Case 1. ER heterogeneity. Low power (a) and high power (b) view of the invasive carcinoma, partial positivity of ER immunos-
tain of this tumor (c), and positive internal controls on the same slide (d)
–– Discuss the results with the patient’s clinician.

a
–– Recommend repeating the testing on subsequent surgical
specimen.
• Final Diagnosis on the resection specimen: ER-positive,
PR-positive, HER2-negative invasive ductal carcinoma,
NOS
• Take-Home Messages:
–– Heterogeneity of intra-tumoral ER immunostain does
exist; further investigation is needed if result of biomark-
ers does not fit in the clinical/pathological features.
–– Triple-negative breast cancers are usually high grade
tumors with a high proliferative index (Ki67 > 50%),
however, there are some exceptions (see Question 7.)
Case 2. ER-negative/PR-positive breast cancer

b (Fig. 7.13a–c)
• History: A 50-year-old female with a breast mass that was
biopsied
• Histology on the biopsy: Invasive ductal carcinoma, NOS,
histologic grade 3, no necrosis present
• IHC on the biopsy: ER-negative, PR-positive (10%),
HER2-negative (1+), Ki67 30%
• Next Steps:
–– Report this tumor as ER-negative/PR-positive/HER2-
negative breast cancer.
–– Repeat ER and PR testing on the biopsy specimen.
–– Discuss the results with the patient’s clinician.
–– Recommend repeating the testing on subsequent surgi-
cal specimen.
• Final Diagnosis on the resection specimen: ER-negative/
c PR-positive/HER2-negative invasive ductal carcinoma, NOS
• Take-Home Messages: Up to 5% of breast cancers have been
reported as ER-negative/PR-positive breast cancers. Since
this group of tumor is rare and still controversial, many
pathologists would repeat the testing to confirm the results
before rendering final interpretation on the ER/PR results.
References
1. Hammond ME, Hayes DF, Dowestt M, Allred DC, Hagerty KL,
Badve S, et al. American Society of Clinical Oncology/College
of American Pathologists guideline recommendations for immu-
nohistochemical testing of estrogen and progesterone recep-
tors in breast cancer (unabridges version). Arch Pathol Lab Med.
2010;134(7):e48–72.
Fig. 7.13 (a–c) Case 2. ER-negative/PR-positive breast cancer. High- 2. Harvey JM, Clark GM, Osborne CK, Allred DC. Estrogen receptor
power view (a) of the invasive ductal carcinoma, with negative ER status by immunohistochemistry is superior to the ligand-binding
immunostain (b) and scattered positive PR immunostain (c). Both (b) assay for predicting response to adjuvant endocrine therapy in
and (c) show adequate positive internal controls adjacent to the tumor breast cancer. J Clin Oncol. 1999;17:1474–81.
3. McCarty KS, Miller LS, Cox EB, Konrath J, McCarty KS Sr.
• Immunohistochemistry (IHC) on the biopsy: ER-negative, Estrogen receptor analysis, correlation of biochemical and immu-
nohistochemical methods using monoclonal antireceptor antibod-
PR-negative, HER2-negative (0), Ki67 10% ies. Arch Pathol Lab Med. 1985;109(8):716–21.
• Next Steps: 4. Horwitz KB, Koseki Y, McGuire WL. Estrogen control of proges-
–– Report this tumor as triple-negative breast cancer. terone receptor in human breast cancer: role of estradiol and anties-
–– Repeat ER and PR testing on the biopsy specimen. trogen. Endocrinology. 1978;103(5):1742–51.
190 P. Tang et al.
5. Cui X, Schiff A, Arpino G, Osborne CK, Lee AV. Biology of pro- during and after neoadjuvant aromatase inhibitor treatment of breast
gesterone receptor loss in breast cancer and its implication for endo- cancer: results from the American College of Surgeons oncology
crine therapy. J Clin Oncol. 2005;23(30):7721–35. group Z1031 trial (Alliance). J Clin Oncol. 2017;35(10):1061–9.
6. Shen T, Brandwein-Gensler M, Hameed O, Siegal GP, Wei 20. Harris LN, Ismaila N, McShane LM, Andre F, Collyar DE,

S. Characterization of estrogen receptor-negative/progesterone Gonzalez-Angulo AM, et al. Use of biomarkers to guide decision
receptor positive breast cancer. Hum Pathol. 2015;46:1776–84. on adjuvant systemic therapy for women with early stage invasive
7. Schroth W, Buttner SWF, Goletz S, Goletz S, Faißt S, Brinkmann F, breast cancer: American Society of Clinical Oncology practice
et al. Clinical outcome and global gene expression data support the guideline. J Clin Oncol. 2016;34:1134–50.
existence of the estrogen receptor-negative/progesterone receptor- 21. Coates AS, Winer EP, Coldhirsch A, Gelber RD, Gnant M, Piccart-
positive invasive breast cancer phenotype. Breast Cancer Res Treat. Gebhart M, et al. Tailoring therapies – improving the management
2016;155:85–97. of early breast cancer: St Gallen international expert consensus
8. Slamon DJ, Leyland-Jones B, Shak S, Fuchs H, Paton V, Bajamonde on the primary therapy of early breast cancer 2015. Ann Oncol.
A, et al. Use of chemotherapy plus a monoclonal antibody against 2015;26:1533–46.
HER2 for metastatic breast cancer that overexpresses HER2. N 22. Han JS, Cao D, Molberg KH, Sarode VR, Rao R, Sutton LM,
Engl J Med. 2001;344(11):783–92. et al. Hormone receptor status rather than HER-2 status is signifi-
9. van Ramshorst MS, Loo CE, Groen EJ, Winter-Warnars GH, cantly associated with increased Ki67 and p53 expression in triple-
Wesseling J, van Duijnhoven F, et al. MRI predicts pathologic com- negative breast carcinomas, and high expression level of Ki67 but
plete response in HER2-positive breast cancer after neoadjuvant not p53 is significantly associated with axillary nodal metastasis in
chemotherapy. Breast Cancer Res Treat. 2017;164(1):99–106. triple-negative and high grade non-triple negative breast carcino-
10. Wolff AC, Hammond ME, Schwartz JN, Hagerty KL, Allred DC, mas. Am J Clin Pathol. 2011;135:230–7.
Cote RJ, et al. American Society of Clinical Oncology/College of 23. Dowsett M, Nielsen TO, A’Hern R, Bartlett J, Coombes RC, Cuzick
American Pathologists. American Society of Clinical Oncology/ J, et al. Assessment of Ki67 in breast cancer: recommendations
College of American Pathologists guideline recommendations for from the international Ki67 in breast cancer working group. J Nalt
human epidermal growth factor receptor 2 testing in breast cancer. Cancer Inst. 2011;103:1656–64.
Arch Pathol Lab Med. 2007;131(1):18–43. 24. Polley MY, Leung SCY, Gao D, Mastropasqua MG, Zabaglo LA,
11. Wolff AC, Hammond ME, Hicks DG, Dowsett M, McShane LM, Bartlett JM, et al. An international study to increase concordance in
Allison KH, et al. Recommendation for human epidermal growth Ki67 scoring. Mod Pathol. 2015;28:778–86.
factor receptor 2 testing in breast cancer: American Society of 25. Cheang MC, Chia SK, Voduc D, Gao D, Leung S, Snider J, et al.
Clinical Oncology/College of American Pathologists clinical prac- Ki67 index, HER2 status, and prognosis of patients with luminal B
tice guideline update. J Clin Oncol. 2013;31:3997–4013. breast cancer. J Natl Cancer Inst. 2009;101(10):736–50.
12. Wolff AC, Hammond ME, Allison KH, Harvey BE, Mangu PB, 26. Carbognin L, Sperduti S, Fabi A, Dieci MV, Kadrija D, Griguolo
Bartlett JMS, et al. Human epidermal growth factor receptor 2 G, et al. Prognostic impact of proliferation for resected early
testing in breast Cancer. American Society of Clinical Oncology/ stage “pure” invasive lobular carcinoma: cut-off analysis of
College of American Pathologists Clinical Practice Guideline Ki67 according to histology and clinical validation. Bresat.
Focused Update. Arch Pathol Lab Med. 2018;142:1364–82. 2017;35:21–6.
13. Dennis J, Parsa R, Chau D, Koduru P, Peng Y, Fang Y, et al. 27. Liu H, Shi J, Wilkerson ML, Lin F. Immunohistochemical evalu-
Quantification of human epidermal growth factor receptor 2 ation of GATA3 expression in tumors and normal tissues: a use-
immunohistochemistry using the Ventana image analysis sys- ful immunomarker for breast and urothelial carcinomas. Am J Clin
tem: correlation with gene amplification by fluorescence in Pathol. 2012 Jul;138(1):57–64.
situ hybridization: the importance of instrument validation for 28. Asch-Kendrick R, Cimino-Mathews A. The role of GATA3 in

achieving high (>95%) concordance rate. Am J Surg Pathol. breast carcinomas: a review. Hum Pathol. 2016 Feb;48:37–47.
2015;39(5):624–31. 29. Cimino-Mathews A, Subhawong AP, Illei PB, Sharma R, Halushka
14. Yu J, Dabbs DJ, Shaui Y, Niemeier LA, Bhargava R. Classical- MK, Vang R, et al. GATA3 expression in breast carcinoma: utility
type lobular carcinoma with HER2 overexpression: clinical, his- in triple-negative, sarcomatoid, and metastatic carcinomas. Hum
tologic and hormonal receptor characteristics. Am J Clin Pathol. Pathol. 2013 Jul;44(7):1341–9.
2011;136:88–97. 30. Krings G, Nystrom M, Mehdi I, Vohra P, Chen YY. Diagnostic util-
15. Wolff AC, Hammond ME, Hicks DG, Dowsett M, McShane LM, ity and sensitivities of GATA3 antibodies in triple-negative breast
Allison KH, et al. Recommendations for human epidermal growth cancer. Hum Pathol. 2014 Nov;45(11):2225–32.
factor receptor 2 testing in breast Cancer American society of clini- 31. Dang DN, Raj G, Sarode V, Molberg KH, Vadlamudi RK, Peng
cal oncology/college of American pathologists clinical practice Y. Significantly increased PELP1 protein expression in primary
guideline update. Arch Pathol Lab Med. 2014;138(2):241–56. and metastatic triple-negative breast carcinoma: comparison with
16. Klein ME, Dabbs DJ, Shuai Y, Brufsky AM, Jankowitz R, Puhalla GATA3 expression and PELP1’s potential role in triple-negative
SL, et al. Prediction of the Oncotype DX recurrence score: use of breast carcinoma. Hum Pathol. 2015 Dec;46(12):1829–35.
pathology-generated equations derived by linear regression analy- 32. Miettinen M, McCue PA, Sarlomo-Rikala M, Rys J, Czapiewski
sis. Mod Pathol. 2013 May;26:658–64. P, Wazny K, et al. Gata 3 – a multispecific but potentially use-
17. Cuzick J, Dowsett M, Pineda S, Wale C, Salter J, Quinn E, et al. ful marker in surgical pathology – a systematic analysis of 2500
Prognostic value of a combined estrogen receptor, progesterone epithelial and non-epithelial tumors. Am J Surg Pathol. 2014
receptor, Ki-67, and human epidermal growth factor receptor 2 Jan;38(1):13–22.
immunohistochemical score and comparison with the Genomic 33. Loibl S, Müller BM, von Minckwitz G, Schwabe M, Roller M,
Health recurrence score in early breast cancer. J Clin Oncol. Darb-Esfahani S, et al. Androgen receptor expression in primary
2011;29:4273–8. breast cancer and its predictive and prognostic value in patients
18. Ellis MJ, Tao Y, Luo J, A’Hern R, Evans DB, Bhatnagar AS, et al. treated with neoadjuvant chemotherapy. Breast Cancer Res Treat.
Outcome prediction for estrogen receptor positive breast cancer 2011 Nov;130(2):477–87.
based on postneoadjuvant endocrine therapy tumor characteristics. 34. Park S, Koo JS, Kim MS, Park HS, Lee JS, Lee JS, et al. Androgen
J Nalt Cancer Inst. 2008;100:1380–8. receptor expression is significantly associated with better outcomes
19. Ellis MJ, Suman VJ, Hoon J, Goncalves R, Sanati S, Creighton CJ, in estrogen receptor-positive breast cancers. Ann Oncol. 2011
et al. Ki67 proliferation index as a tool for chemotherapy decision Aug;22(8):1755–62.
35. Cochrane DR, Bernales S, Jacobsen BM, Cittelly DM, Howe EN, 54. Subhawong AP, Subhawong T, Nassar H, Kouprina N, Begum S,
D’Amato NC, et al. Role of the androgen receptor in breast can- Vang R, et al. Most basal-like breast carcinomas demonstrate the
cer and preclinical analysis of enzalutamide. Breast Cancer Res. same Rb−/p16+ immunophenotype as the HPV-related poorly dif-
2014;16(1):R7. ferentiated squamous cell carcinomas which they resemble mor-
36. Basile D, Cinausero M, Iacono D, Bonotto M, Vitale MG, Gerratana phologically. Am J Surg Pathol. 2009;33(2):163–75.
L, et al. Androgen receptor in estrogen receptor positive breast can- 55. Bohn OL, Fuertes-Camilo M, Navarro L, Saldivar J, Sanchez-Sosa
cer: beyond expression. Cancer Treat Rev. 2017;61:15–22. S. p16INK4a expression in basal-like breast carcinoma. Int J Clin
37. Sutton LM, Cao D, Sarode V, Molberg KH, Torgbe K, Haley B, et al. Exp Pathol. 2010;3(6):600–7.
Decreased androgen receptor expression is associated with distant 56. Arima Y, Hayashi N, Hayashi H, Sasaki M, Kai K, Sugihara E, et al.
metastases in patients with androgen receptor-expressing triple-neg- Loss of p16 expression is associated with the stem cell characteristics
ative breast carcinoma. Am J Clin Pathol. 2012 Oct;138(4):511–6. of surface markers and therapeutic resistance in estrogen receptor-
38. Lehmann BD, Bauer JA, Chen SME, Chakravarthy AB, Shyr Y, negative breast cancer. Int J Cancer. 2012;130(11):2568–79.
et al. Identification of human triple-negative breast cancer subtypes 57. Cheang MC, Voduc D, Bajdik C, Leung S, McKinney S, Chia SK,
and preclinical models for selection of targeted therapies. J Clin et al. Basal-like breast cancer defined by five biomarkers has supe-
Invest. 2011 Jul;121(7):2750–67. rior prognostic value than triple-negative phenotype. Clin Cancer
39. Hu R, Dawood S, Holmes MD, Collins LC, Schnitt SJ, Cole K, et al. Res. 2008;14(5):1368–76.
Androgen receptor expression and breast cancer survival in post- 58. Rakha E, Reis-Filho JS. Basal-like breast carcinoma: from expres-
menopausal women. Clin Cancer Res. 2011 Apr 1;17(7):1867–74. sion profiling to routine practice. Arch Pathol Lab Med. 2009
40. Rakha EA, El-Sayed ME, Green AR, Lee AH, Robertson JF, Ellis Jun;133(6):860–8.
IO. Prognostic markers in triple-negative breast cancer. Cancer. 59. Sutton LM, Han JS, Molberg KH, Sarode VR, Cao D, Rakheja
2007;109(1):25–32. D, et al. Intratumoral expression level of epidermal growth factor
41. Park S, Koo J, Park HS, Kim JH, Choi SY, Lee JH, et al. Expression receptor and cytokeratin 5/6 is significantly associated with nodal
of androgen receptors in primary breast cancer. AnnOncol. and distant metastases in patients with basal-like triple-negative
2010;21(3):488–92. breast carcinoma. Am J Clin Pathol. 2010;134(5):782–7.
42. Tang D, Xu S, Zhang Q, Zhao W. The expression and clinical sig- 60. Bhargava R, Beriwal S, McManus K, Dabbs DJ. CK5 is more sensi-
nificance of the androgen receptor and E-cadherin in triple-negative tive than CK5/6 in identifying the “basal-like” phenotype of breast
breast cancer. Med Oncol. 2012 Jun;29(2):526–33. carcinoma. Am J Clin Pathol. 2008;130:724–30.
43. Luo X, Shi YX, Li ZM, Jiang WQ. Expression and clinical signifi- 61. Foulkes WD, Smith IE, Reis-Filho JS. Triple-negative breast can-
cance of androgen receptor in triple negative breast cancer. Chin J cer. N Engl J Med. 2010;363(20):1938–48.
Cancer. 2010 Jun;29(6):585–90. 62. Nogi H, Kobayashi T, Suzuki M, Tabei I, Kawase K, Toriumi
44. He J, Peng R, Yuan Z, Wang S, Peng J, Lin G, et al. Prognostic Y, et al. EGFR as paradoxical predictor of chemosensitivity
value of androgen receptor expression in operable triple-negative and outcome among triple-negative breast cancer. Oncol Rep.
breast cancer: a retrospective analysis based on a tissue microarray. 2009;21(2):413–7.
Med Oncol. 2012 Jun;29(2):406–10. 63. Dawson SJ, Provenzano E, Caldas C. Triple negative breast cancers:
45. Iacopetta D, Rechoum Y, Fuqua SAW. The role of andro-
clinical and prognostic implications. Eur J Cancer. 2009;45(Suppl
gen receptor in breast Cancer. Drug Discov Today Dis Mech. 1):27–40.
2012;9(1–2):e19–27. 64. Tan DS, Marchió C, Jones RL, Savage K, Smith IE, Dowsett M,
46. Barton VN, D’Amato NC, Gordon MA, Lind HT, Spoelstra NS, et al. Triple negative breast cancer: molecular profiling and prog-
Babbs BL, et al. Multiple molecular subtypes of triple-negative nostic impact in adjuvant anthracycline-treated patients. Breast
breast cancer critically rely on androgen receptor and respond to Cancer Res Treat. 2008;111(1):27–44.
enzalutamide in vivo. Mol Cancer Ther. 2015;14(3):769–78. 65. O’Conor CJ, Chen T, González I, Cao D, Peng Y. Cancer stem cells
47. Gucalp A, Tolaney S, Isakoff SJ, Ingle JN, Liu MC, Carey LA, et al. in triple-negative breast cancer: a potential target and prognostic
Phase II trial of bicalutamide in patients with androgen receptor- marker. Biomark Med. 2018;12(7):813–20.
positive, estrogen receptor-negative metastatic breast Cancer. Clin 66. Turner BM, Cagle PT, Sainz IM, Fukuoka J, Sehn SS, Jagirda
Cancer Res. 2013;19(19):5505–12. J. Napsin a, a new marker for lung adenocarcinoma, is comple-
48. Dang D, Peng Y. Roles of p53 and p16 in triple-negative breast mentary and more sensitive and specific than thyroid transcription
Cancer. Breast Cancer Manage. 2013;2:537–44. factor 1 in the differential diagnosis of primary pulmonary carci-
49. Sugianto J, Sarode V, Peng Y. Ki-67 expression is increased in noma: evaluation of 1674 cases by tissue microarray. Arch Pathol
p16-expressing triple-negative breast carcinoma and correlates with Lab Med. 2012;136(2):163–71.
p16 only in p53-negative tumors. Hum Pathol. 2014 Apr;45(4):802–9. 67. Bishop JA, Sharma R, Illei PB. Napsin a and thyroid transcription
50. Han JS, Cao D, Molberg KH, Sarode VR, Rao R, Sutton LM, et al. factor-1 expression in carcinomas of the lung, breast, pancreas,
Hormone receptor status rather than HER2 status is significantly colon, kidney, thyroid, and malignant mesothelioma. Hum Pathol.
associated with increased Ki-67 and p53 expression in triple- 2010;41(1):20–5.
negative breast carcinomas, and high expression of Ki-67 but not 68. Robens J, Goldstein L, Gown AM, Schnitt SJ. Thyroid transcrip-
p53 is significantly associated with axillary nodal metastasis in tion Factor-1 expression in breast carcinomas. Am J Surg Pathol.
triple-negative and high-grade non-triple-negative breast carcino- 2010;34(12):1881–5.
mas. Am J Clin Pathol. 2011;135(2):230–7. 69. Miettinen M, McCue PA, Sarlomo-Rikala M, Rys J, Czapiewski
51. Lee DS, Kim SH, Suh YJ, Kim S, Kim HK, Shim BY. Clinical P, Wazny K, et al. GATA3: a multispecific but potentially useful
implication of p53 overexpression in breast cancer patients younger marker in surgical pathology: a systematic analysis of 2500 epithe-
than 50 years with a triple-negative subtype who undergo a modi- lial and nonepithelial tumors. Am J Surg Pathol. 2014;38(1):13–22.
fied radical mastectomy. Jpn J Clin Oncol. 2011;41(7):854–66. 70. Lau SK, Chu PG, Weiss LM. Immunohistochemical expres-

52. Hudis CA, Gianni L. Triple-negative breast cancer: an unmet medi- sion of estrogen receptor in pulmonary adenocarcinoma. Appl
cal need. Oncologist. 2011;16(Suppl 1):1–11. Immunohistochem Mol Morphol. 2006;14(1):83–7.
53. Biganzoli E, Coradini D, Ambrogi F, Garibaldi JM, Lisboa P, Soria 71.
Gomez-Fernandez C, Mejias A, Walker G, Nadji M.
D, et al. p53 status identifies two subgroups of triple-negative Immunohistochemical expression of estrogen receptor in adenocar-
breast cancers with distinct biological features. Jpn J Clin Oncol. cinomas of the lung: the antibody factor. Appl Immunohistochem
2011;41(2):172–9. Mol Morphol. 2010;18(2):137–41.
192 P. Tang et al.
72. Dabbs DJ, Landreneau RJ, Raab SS, Maley RH, Tung MY,
array study of 207 cases. Appl Immunohistochem Mol Morphol.
Silverman JF. Detection of estrogen receptor by immunohis- 2014;22:266–74.
tochemistry in pulmonary adenocarcinoma. Ann Thorac Surg. 79. Peng Y, Butt Y, Chen B, Zhang X, Tang P. Update on immu-
2002;73(2):403–5. nohistochemistry in breast lesions. Arch Pathol Lab Med.
73. Mar N, Vredenburgh JJ, Wasser JS. Targeting HER2 in
2017;141(8):1033–51.
the treatment of non-small cell lung cancer. Lung Cancer. 80. Saluja K, Peng Y. Metastatic ER positive lung adenocarcinoma to
2015;87(3):220–5. liver and breast mimicking recurrent breast carcinoma. Am J Clin
74. Yoshizawa A, Sumiyoshi S, Sonobe M, Kobayashi M, Uehara T, Pathol. 2015;144(suppl 2):A258.
Fujimoto M, et al. HER2 status in lung adenocarcinoma: a com- 81. Nofech-Mozes S, Khalifa MA, Ismiil N, Saad RS, Hanna WM,
parison of immunohistochemistry, fluorescence in situ hybrid- Covens A, et al. Immunophenotyping of serous carcinoma of the
ization (FISH), dual-ISH, and gene mutations. Lung Cancer. female genital tract. Mod Pathol. 2008;21(9):1147–55.
2014;85(3):373–8. 82. Recine MA, Deavers MT, Middleton LP, Silva EG, Malpica

75. Georgiannos SN, Chin J, Goode AW, Sheaff M. Secondary neo- A. Serous carcinoma of the ovary and peritoneum with metastases
plasms of the breast: a survey of the 20th century. Cancer. to the breast and axillary lymph nodes: a potential pitfall. Am J Surg
2001;92:2259–66. Pathol. 2004;28(12):1646–51.
76. Klingen TA, Klassen H, Aas H, Chen Y, Aksen LA. Secondary 83. Farghaly H. Metastatic endometrial endometrioid carcinoma with
breast cancer: a 5-year population-based study with review of the clear cell changes to the breast: a case report. Ann Diagn Pathol.
literature. APMIS. 2009;117(10):762–7. 2013;17(1):127–30.
77. Williams SA, Ehlers RA, Hunt KK, Yi M, Kuerer HM, Singletary 84. Li C1, Xia P, Tian T, Kou B, Nan K. Metastasis from endometrial
SE, et al. Metastases to the breast from nonbreast solid neoplasms. carcinoma to bilateral breasts presenting as inflammatory breast
Cancer. 2007;110:731–7. lesions. Eur J Gynaecol Oncol. 2011;32(5):563–6.
78. Kawaguchi KR, Lu FI, Kaplan R, Liu YF, Chadwick P, Chen Z, 85. Hoefnagel LD, van de Vijver MJ, van Slooten HJ, Wesseling P,
et al. In search of the ideal immunopanel to distinguish metastatic Wesseling J, Westenend PJ. Receptor conversion in distant breast
mammary carcinoma from primary lung carcinoma: a tissue micro- cancer metastases. Breast Cancer Res. 2010;12(5):R75.
Breast Cancer with Hereditary Cancer
Predisposition Syndromes 8
Roshni Rao, Caitlin B. Mauer, Margaret Chen-Seetoo,
and Yan Peng
List of Frequently Asked Questions the particular gene. High-penetrance genes, such as BRCA1
and BRCA2, are known to have earlier ages of onset. The
1. What percentage of breast cancers are associated with average age of onset of breast cancer in a woman with a
hereditary syndromes? BRCA1 mutation is between 35 and 55 years, while the aver-
age age of onset in a woman with a BRCA2 mutation is
At least 10% of breast cancers occur in patients with heredi- between 41 and 57 years [5]. More data are needed to deter-
tary mutations. Mutations in high-penetrance genes, such as mine the average age of onset of breast cancer in moderate-
BRCA1, BRCA2, PTEN, TP53, CDH1, and STK11, account penetrance genes, such as ATM, PALB2, or CHEK2.
for up to 25–50% of the hereditary breast cancers [1, 2].
Mutations in moderate-penetrance genes (CHEK2, ATM, and 3. What clinical presentations or indications in patients
PALB2) are thought to be as common as mutations in the with breast cancer should prompt genetic counseling
BRCA1 and BRCA2 genes [3]. Low-penetrance gene muta- and testing?
tions, such as mutations in BARD1 and RAD51D, have also
been identified to confer elevated breast cancer risks [4]. The current National Comprehensive Cancer Network
(NCCN) guidelines (version 2.2019) recommend a referral
2. What is the most common age of presentation of for genetic counseling for patients with personal or family
breast cancer in hereditary syndromes? histories of any of the following [6]:
The age of presentation of breast cancer in patients with a • Breast cancer diagnosed at or before age 50
hereditary breast cancer predisposition varies depending on • Two breast cancer primaries
• Ovarian cancer
• Triple-negative breast cancer diagnosed at or before
age 60
R. Rao • Pancreatic cancer
Division of Breast Surgery, Department of Surgery, Herbert Irving • Male breast cancer
Pavilion, Columbia University Vagelos College of Physicians and
• Metastatic prostate cancer
Surgeons, New York, NY, USA
e-mail: [email protected] • Ashkenazi Jewish ancestry with a diagnosis of breast,
ovarian, pancreatic, or high-grade (Gleason score ≥7)
C. B. Mauer
Department of Cancer Genetics, University of Texas Southwestern prostate cancer
Medical Center, Dallas, TX, USA • Three or more of the following cancers on the same side
e-mail: [email protected] of the family:
M. Chen-Seetoo –– Breast cancer
Department of Surgery, Herbert Irving Pavilion, Columbia University –– Breast cancer, sarcoma, adrenocortical cancer, brain
Vagelos College of Physicians and Surgeons, New York, NY, USA
tumor, leukemia (consider Li-Fraumeni syndrome)
–– Lobular breast cancer, diffuse gastric cancer (consider
Y. Peng (*)
Hereditary Diffuse Gastric Cancer syndrome)
Medical Center, Dallas, TX, USA –– Colon cancer, endometrial cancer, thyroid cancer, kid-
e-mail: [email protected] ney cancer, dermatologic manifestations, m
acrocephaly,

194 R. Rao et al.
hamartomatous polyps of the gastrointestinal tract (con- this phosphorylation that helps activate the repair of
sider Cowden syndrome) double-stranded breaks [12].
–– Breast cancer, gastrointestinal cancer or hamartoma-
tous polyps, pancreatic cancer, ovarian sex cord- 8. How is a founder mutation in BRCA1/BRCA2 carriers
stromal tumors, testicular Sertoli cell tumors, or defined?
childhood skin pigmentation (consider Peutz-Jeghers
syndrome) The concept of a founder effect is predicated on the fact that
• A known cancer gene mutation in the family new populations can be formed from a numerically smaller
group of individuals. Subsequently, only a small fraction of
4. What is the risk for patients with BRCA gene muta- the genetic variability from the original group will be
tions to develop breast cancer? expressed in the new population [13]. When a recurrent muta-
tion is identified within a specific geographic area, this vari-
BRCA1 and BRCA2 are considered high-penetrance genes, ant may have originated from a single mutation event and is
although there is considerable variation among specific pat- referred to as a founder mutation. Haplogroup analysis allows
terns of cancer in families. Meta-analyses have demonstrated for differentiation between a founder effect and independent
a lifetime risk (defined as age 70) of breast cancer to be 57% mutations. The Ashkenazi Jewish population in particular has
for women with a BRCA1 gene mutation and 49% for women three well-studied founder mutations in the BRCA genes [13].
with a BRCA2 gene mutation [7]. BRCA1 and BRCA2 founder mutations exist in many other
populations, including but not limited to the Icelandic,
5. What is the risk for patients with BRCA gene muta- Finnish, Dutch, Norwegian, French Canadian, Italian,
tions to develop ovarian cancer? Chinese, and Pakistani populations [14]. Ongoing research
among Latino [15], Middle Eastern [16], and Asian [17] pop-
Given the very large number of actual mutations identified in ulations reveals the importance of identifying these founder
the BRCA1 and BRCA2 genes, there is a broad range of ovar- mutations to more accurately stratify cancer risks in particu-
ian cancer risk estimates. Most contemporary series demon- lar ethnic populations [18].
strate a 40% lifetime risk of ovarian cancer in BRCA1
mutation carriers and an 11–17% lifetime risk in BRCA2 9. Do lifestyle modifications in BRCA1 and BRCA2
mutation carriers [8]. In addition, BRCA1 and BRCA2 muta- mutation carriers convey risk reduction?
tion carriers have a 1.3% lifetime risk of developing primary
peritoneal serous carcinoma [9]. Differences of cancer presentation among families reflect the
varying degrees of penetrance, suggesting that environmen-
6. On which chromosomes are the BRCA1 gene and the tal and behavioral factors may modify breast cancer risks in
BRCA2 gene located in humans? hereditary breast cancer. Among BRCA mutation carriers,
higher body weight increased breast cancer risk; and smok-
Mary-Claire King and colleagues were the first to identify ing, nulliparity, and oral contraceptive use were associated
the genetic susceptibility to breast cancer conferred by muta- with earlier disease onset [19, 20]. The effects of lifestyle
tions in the BRCA1 gene in 1994. This gene is located on the intervention on risk modification in hereditary breast cancer
long arm of chromosome 17 [10] and encodes a protein of are unknown. A multicenter trial is currently underway to
1863 amino acids. Shortly after Mary-Claire King’s discov- assess potential risk reduction benefits [21].
ery, in 1995, a research group led by Sir Michael Stratton at
the Institute of Cancer Research, United Kingdom, identified 10. What is the lifetime risk of developing breast cancer
the BRCA2 gene. The BRCA2 gene is located on the long after prophylactic mastectomy for BRCA1 and
arm of chromosome 13; it is larger than the BRCA1 gene as BRCA2 mutation carriers?
it encodes a protein of 3418 amino acids [10]. Both genes
function as tumor suppressor genes by mediating double- Multiple studies demonstrated prophylactic mastectomy in
stranded break repairs in genes through homologous recom- women with BRCA1 and BRCA2 mutations reduced the risk
bination [11]. of subsequent breast cancer by 90% or more [22].
7. Is the BRCA1 gene phosphorylated in response to cell 11. What are management recommendations for patients
damage? with BRCA1 and BRCA2 mutations in regard to their
ovarian cancer risks?
It was the initial recognition that the BRCA1 gene is
hyperphosphorylated in response to DNA damage that Despite significant advances in therapy, survival from ovarian
prompted its identification as a tumor suppressor gene. It is cancer continues to be poor, with 5-year survival in advanced
8 Breast Cancer with Hereditary Cancer Predisposition Syndromes 195
disease to be ~20%. Approximately 80% of patients are diag- any suspicious breast lesion should undergo needle biopsy
nosed at an advanced stage, primarily due to a lack of effective (if technically feasible) for tissue diagnosis [32]. Biopsy can
screening modalities. Nonetheless, secondary to the known be performed under palpation or image guidance.
high risk of ovarian cancer in BRCA mutation carriers, screen-
ing can be considered beginning at the age of 30 and continu- 14. What is the incidence of metachronous, contralateral
ing until prophylactic bilateral salpingo-oophorectomy (BSO) primary breast cancer in BRCA1 and BRCA2 muta-
is complete [23]. Screening techniques for ovarian cancer in tion carriers?
these patients include annual serum antigen to evaluate for
CA-125 levels and transvaginal ultrasound. Combined, these Compared to non-BRCA mutation carriers, the risk of a con-
continue to have low sensitivities of ~60% [24, 25]; currently, tralateral primary breast cancer for BRCA carriers is higher.
no proven alternatives for ovarian cancer screening have been Quantifying a specific risk for these patients is important so
validated. To help with risk reduction, a meta-analysis demon- informed choices regarding prophylactic surgery vs. appro-
strated that the use of oral contraceptives for 1 year in BRCA1 priate surveillance can be made. A large meta-analysis [33]
and BRCA2 mutation carriers can decrease the risk of ovarian revealed the cumulative risk of contralateral breast cancer
cancer by 33–80% and 58–63%, respectively [26]. Regardless, increases as a function of time from the initial cancer diagno-
the National Comprehensive Cancer Network (NCCN) and sis. In BRCA1 carriers, pooled risk estimates reached 33% at
the Society of Gynecologic Oncology (SGO) recommend 15 years. In contrast, for BRCA2 carriers, similar analysis
BRCA1 mutation carriers consider a prophylactic bilateral revealed only a 23% contralateral breast cancer risk at
salpingo-oophorectomy (BSO) between ages 35 and 40 or 15 years.
after childbearing is complete. Because the average age of
onset of ovarian cancer is slightly later in BRCA2 mutation 15. Do patterns of metastatic disease differ between
carriers, these women can consider prophylactic BSO between patients with BRCA1-associated breast cancers and
ages 40 and 45 or after completion of childbearing [6, 23]. breast cancer patients with no mutations?
Given that newer data suggest ovarian cancer may originate in
the fallopian tubes [27], current trials are being conducted to In a study comparing BRCA1 mutation carriers with triple-
determine if a salpingectomy followed by delayed oophorec- negative breast cancer to a group with sporadic triple-
tomy may reduce the incidence of ovarian cancer in BRCA negative breast cancer, overall survival and prognosis were
mutation carriers. not impacted by BRCA1 mutation carrier status [34]. In addi-
tion, patients developed metastases in a similar time frame
12. How does oophorectomy change survival in the treat- (median was ~20 months after diagnosis), with the most
ment of early-stage breast cancer in BRCA-positive common site of metastatic disease being the lungs in both
patients? groups, followed by liver metastases and then bone
metastases.
Several studies have attempted to determine the impact of
oophorectomy on mortality from breast cancer in BRCA muta- 16. Can endocrine therapy reduce breast cancer risk in
tion carriers. Of specific interest is whether oophorectomy is known BRCA1 mutation carriers?
helpful after a diagnosis of breast cancer, particularly given that
BRCA1 carriers have a predilection for triple-negative breast There are very limited data regarding the use of tamoxifen
cancers where endocrine manipulation and HER2 targeted ther- for breast cancer prevention in BRCA mutation carriers. The
apy would not be helpful. Retrospective cohort studies have only prospective study was the National Surgical Adjuvant
demonstrated a 70–80% [28, 29] reduction in mortality when Breast and Bowel Project P1 trial, which identified eight
oophorectomy is performed after a diagnosis of breast cancer in BRCA1 and eleven BRCA2 carriers. In contrast to BRCA2
BRCA mutation carriers. A multicenter cohort study of 2482 carriers whose breast cancers are estrogen receptor positive
BRCA1 and BRCA2 carriers by Domchek and colleagues [30] in 77% of cases, 75–80% of cancers in BRCA1 carriers are
evaluated breast cancer-specific mortality with oophorectomy estrogen receptor negative [22, 35]. The effectiveness of
and demonstrated an odds ratio of 0.35 (95% CI, 0.19–0.67). A tamoxifen in the prevention of breast cancer among BRCA1
similar analysis performed by Metcalfe and colleagues [31] carriers is currently unknown and warrants investigation.
confirmed this mortality benefit for BRCA1 carriers only.
17. Is a nipple-sparing mastectomy (NSM) for known
13. Should known deleterious gene mutation carriers with mutation carriers an appropriate management
a breast lesion on imaging undergo needle biopsy? strategy?
Both BRCA and non-BRCA mutation carriers, as well as Nipple-sparing mastectomy (NSM) is increasingly per-
other breast cancer predisposition mutation carriers, with formed for women with breast cancer, with no difference
196 R. Rao et al.
found in disease recurrence or survival compared with skin- 21. Are there any targeted therapies for breast cancer
sparing and total mastectomies [36]. Data examining the patients with hereditary cancer syndromes?
safety of NSM in BRCA mutation carriers are limited. Yao
and colleagues found that among 150 patients who under- Poly-ADP ribose polymerases (PARPs) are a group of
went NSM for prophylaxis and 51 cancer patients, only 3 of enzymes that aid in DNA repair. PARP inhibitors block this
the cancer patients (5.8%) had cancer in the nipple-areolar DNA repair mechanism and can result in cell death in
complex. Four cancer events were found with a mean follow- patients who already have DNA repair deficiencies, such as
up of 32.6 months, three in cancer patients and one in a risk individuals with a BRCA mutation. As such, PARP inhibitors
reduction patient, but none of the cancers were in the nipple- have been approved by the Food and Drug Administration
areolar complex. These findings suggested NSM in BRCA (FDA) to be used to treat breast cancer in BRCA-positive
carriers appears to be safe, but long-term studies are needed patients [40].
to assess its use [37].
22. Which genetic mutations are contraindications for
18. Is axillary staging utilizing sentinel node biopsy breast-conserving therapy in the setting of a breast
accurate in BRCA-associated breast cancers? cancer diagnosis?
Sentinel lymph node biopsy utilizes the known, orderly pat- Li-Fraumeni syndrome is caused by a germline mutation in
terns of breast lymphatic drainage to identify the initial the TP53 tumor suppressor gene. It is associated with soft tis-
lymph nodes that breast cancer would spread to if there were sue sarcoma, osteosarcoma, brain tumor, adrenocortical car-
lymph node metastases. The nodes are typically identified by cinoma, and premenopausal breast cancer. Breast-conserving
utilizing a combination of a radioactive tracer and a blue dye therapy is contraindicated in patients who carry TP53 gene
that allows visual confirmation of the sentinel node(s). In the mutations due to the concern for radiation-induced secondary
setting of breast cancer, surgical axillary staging with a sen- cancers in this specific patient population [41].
tinel node biopsy is appropriate regardless of gene mutation In addition to long-term, high-risk breast cancer surveil-
carrier status [38]. lance, the current NCCN guidelines for the management of
BRCA-associated breast cancer include the consideration of
19. How often is a clinically occult breast cancer identi- bilateral total mastectomy for risk reduction and treatment
fied on prophylactic mastectomy specimens? [6]. While breast-conserving therapy has been performed, a
large meta-analysis [42] that included ten studies specifically
Occult breast cancer is identified in 0–3% of BRCA mutation evaluating breast conservation in BRCA gene mutation carri-
carriers who undergo prophylactic mastectomy and 0.1– ers demonstrated higher rates of in-breast recurrence with
5.6% of non-BRCA mutation patients [37]. long-term follow-up. When median follow-up was at least
7 years, the in-breast recurrence rate was 23.7% in BRCA-
20. Is survival from genetic mutation-associated breast associated breast cancers versus only 15.9% in sporadic
cancers worse than that from sporadic breast breast cancers.
cancers?
23. What are the most common cancers associated with
Breast cancers associated with BRCA1 mutations are more Li-Fraumeni syndrome?
likely to be high grade and triple negative. BRCA2-positive
patients also have been identified to present with higher- Li-Fraumeni syndrome is a rare, inherited hereditary cancer
grade tumors. The largest meta-analysis [39] reveals that, disorder caused by mutations in a tumor suppressor gene
compared to patients with sporadic breast cancers, BRCA1- known as the TP53 gene. The TP53 gene is known as the
associated breast cancer patients have worse overall sur- “guardian of the genome,” and germline mutations in it can
vival and worse breast cancer-specific survival, resulting lead to the development of multiple cancers during an indi-
in a 30% higher risk of dying from breast cancer for vidual’s lifetime. Patients with germline TP53 mutations
patients who are BRCA1 mutation carriers [39]. In con- have nearly a 100% chance of cancer in their lifetime, with
trast, BRCA2-associated breast cancers do not demonstrate 50% of patients developing their first cancer before age 30
lower rates of overall survival, but do demonstrate worse [43]. The most common tumors seen in patients with
breast cancer-specific survival [39]. For BRCA-associated Li-Fraumeni syndrome include sarcomas (both soft tissue
triple-negative breast cancers, overall survival is better and osteosarcoma), premenopausal breast cancer, tumors of
when compared to sporadic cases of triple-negative breast the central nervous system, adrenocortical carcinomas, and
cancer [39]. myeloid leukemias. Patients with Li-Fraumeni syndrome are
thought to be sensitive to ionizing radiation, which may [55]. CDH1 gene mutations, causative of hereditary diffuse
induce secondary malignancies [44]. gastric cancer syndrome, increase the risks of lobular breast
cancer, diffuse gastric cancer, and colon cancer [56].
24. What are the mutations in which genes are associ-
ated with male breast cancer? 27. Is breast cancer associated with CHEK2 mutation? If
so, what are the most common subtypes of breast
Male breast cancer accounts for about 1% of all breast cancer cancer?
cases diagnosed in the United States annually; about 1/1000
(0.01%) of men will develop breast cancer in their lifetime [45]. The CHEK2 c.1100delC founder mutation occurs in 1–2%
Male carriers of a BRCA1 mutation are estimated to have a 1.2% of the European/North American population and is associ-
lifetime (defined as age 70) risk of breast cancer, while male ated with increased breast cancer risk [57]. This mutation is
BRCA2 mutation carriers are estimated to have a 6.8% lifetime one of the best studied and increases breast cancer risk by
risk of male breast cancer [7]. The CHEK2 c.1100delC mutation, twofold in women and tenfold in men [54]. The breast can-
a common European founder mutation, is suspected to increase cers most commonly seen in female CHEK2 mutation carri-
male breast cancer risk by tenfold in certain populations (e.g., ers are luminal type (estrogen receptor positive) and tend to
Finnish, Dutch) and may also confer an earlier average age of occur in the ducts [58–60].
onset of male breast cancer when compared to the general popu-
lation [46, 47]. Further studies are needed to determine if other 28. In addition to breast cancer, CHEK2 mutation carri-
CHEK2 mutations also confer a risk of male breast cancer. ers are at greater risk of what other types of
Mutations in the PALB2 gene may increase the risk of male cancers?
breast cancer by 6.6-fold, although larger population studies are
needed to more precisely define this risk [47]. To date, large stud- In addition to breast cancer, truncating CHEK2 mutations
ies have not shown a statistically significant increased risk of increase the risks of prostate cancer, renal cancer, colon can-
male breast cancer in PTEN, CDH1, or TP53 mutation carriers; cer, and thyroid cancer [54, 61]. Exact risks of the develop-
further analyses of these populations are warranted [47–49]. ment of these cancers are unknown at this time.
25. The diagnosis of a triple-negative breast cancer in a 29. What is the risk for patients with Cowden syndrome
young patient should prompt concern for which to develop breast cancer?
genetic mutations?
Cowden syndrome results from mutations in the tumor sup-
Triple-negative breast cancer makes up about 12–15% of all pressor gene, PTEN. Meta-analysis suggests that mutations
female breast cancer diagnoses and is more prevalent in in PTEN confer a 67–85% lifetime risk of breast cancer [62].
African American women than in Caucasian women [50, Additionally, women with PTEN mutations who have already
51]. Triple-negative breast cancer occurs in about 60% of had a diagnosis of breast cancer are at a 29% risk to develop
women with BRCA1 gene mutations and about 25% of a second primary breast cancer in the subsequent 10 years
women with BRCA2 gene mutations [52]. Other hereditary after diagnosis [63]. Given these risks, the current NCCN
breast cancer gene mutations, including PALB2, BARD1, guidelines recommend women with PTEN gene mutations
RAD51D, RAD51C, and BRIP1, have been identified in begin breast cancer screening (annual mammogram alternat-
patients with triple-negative breast cancer; but further studies ing every 6 months with breast MRI) beginning at age 30–35
are needed to clarify the association [53]. or 10 years younger than the earliest diagnosis of breast can-
cer in the family. Women with PTEN gene mutations should
26. What germline mutations should be of concern in a also be counseled on the consideration of prophylactic bilat-
patient who presents with synchronous breast and eral risk-reducing mastectomy [6].
colon cancers? Females breast cancer is the most common malignancy
observed in Cowden’s patients. Up to 75% demonstrate
Mutations in the CHEK2 gene increase the risks of both benign breast lesions such as intraductal papillomatosis,
female and male breast cancer and colon cancer. Mutations in fibroadenomas and others.
this gene (specifically the common founder mutation
c.1100delC) have also been associated with an increased risk 30. What are clinical manifestations of Cowden
of prostate, kidney, and thyroid cancers [54]. Additionally, syndrome?
mutations in the STK11 gene are associated with Peutz-Jeghers
syndrome and increase the risks of colon, breast, gastric, small In addition to the known predisposition to breast cancer,
bowel, pancreatic, gynecological, and testicular cancers, as PTEN mutations confer a 25–38% lifetime risk of follicu-
well as mucocutaneous staining most vivid during childhood lar thyroid cancer, a 21–28% lifetime risk of endometrial
198 R. Rao et al.
cancer, and a 2–34% lifetime risk of renal cancer. PTEN 33. What are polygenic risk scores, and how are they
mutations also increase the risk of hamartomatous/gan- used in the assessment of breast cancer risks?
glioneuromatous gastrointestinal polyps and colon cancer,
oral mucosal papillomatosis, macrocephaly, penile freck- Polygenic risk scores are the results of algorithms used to
ling, benign dermatological findings such as lipomas or compile genetic data, personal, and/or family risk factors
trichilemmomas, autism/developmental delay, and that may be associated with risk of specific disease. Polygenic
Lhermitte-Duclos disease (dysplastic gangliocytoma of risk scores may be clinically useful in helping to determine
the cerebellum) [62]. breast cancer risk management or in making lifestyle/behav-
ioral recommendations. However, current breast cancer
31. What are management and surveillance options for screening/management guidelines do not take polygenic risk
patients with Cowden syndrome? scores into account when making these recommendations
[6]. Historically, polygenic risk tools have been developed
Given the spectrum of cancers in known PTEN mutation car- based on data from primarily Caucasian populations; thus,
riers, risk-reducing management and surveillance recom- further research is needed to better refine the risk scores to be
mendations, per the current NCCN guidelines [6], include applicable across multiple demographics [69].
the following:
34. How has hereditary cancer predisposition testing
• Annual thyroid ultrasound beginning at the time of diag- changed over time?
nosis, as childhood thyroid cancers have been observed.
• Mammography alternating with breast MRI beginning at Historically, genetic testing consisted of single-gene
age 30–35 (or 10 years younger than the earliest diagnosis sequencing and exonic deletion/duplication analysis to ana-
of breast cancer in the family), consideration of prophy- lyze for mutations in disease-associated genes. In 2013, with
lactic bilateral mastectomy. the advent of next-generation sequencing technology, larger
• Awareness of abnormal menstrual/post-menopausal genetic panels became available to analyze for mutations in
bleeding as a sign of uterine cancer. Endometrial screen- multiple genes at one time. Currently, targeted panels focus-
ing with transvaginal ultrasound or endometrial sampling ing on particular cancers, such as breast cancer, are available,
has not been proven effective in patients with Cowden as are pan-cancer panels [70].
syndrome, but can be considered. Patients should con-
sider a prophylactic hysterectomy after the completion of 35. What types of test results can be received from
childbearing. genetic testing for breast cancer?
• Colonoscopy starting at the age of 35 (or 5–10 years
younger than the earliest colon cancer diagnosis in the Disease-causing mutations, deemed “pathogenic” or “delete-
family, if diagnosed at or before age 40), repeating every rious,” identified through genetic testing are classified as
5 years. positive results. When no disease-causing mutations are
• Consideration of renal ultrasound beginning at age 40, identified, this is considered a negative result. Variants of
repeating every 1–2 years. uncertain significance (VUSs) can also be identified through
• Consideration of dermatological exam, psychomotor genetic testing. VUSs are alterations in the DNA in which
exam, or brain MRI in patients presenting with not enough information is known to classify them as positive
symptoms. or negative. It is recommended that these VUSs are not used
to alter medical management until additional information is
learned [71].
32. What are the cancer risks associated with ATM gene
mutations? 36. Do all labs classify genetic variants that predispose to
cancer in the same way?
The ATM protein has multiple functions, including a central
role in DNA double-stranded break repair. Mutations in the The American College of Medical Genetics (ACMG) and the
ATM gene increase a woman’s lifetime risk to develop breast Association for Molecular Pathology (AMP) published
cancer to be between 28% and 50% [64, 65]. ATM gene guidelines in 2015 to help laboratories classify DNA varia-
mutations also increase the risk of prostate cancer in men tions [71]. However, these guidelines are subject to interpre-
[66, 67] and elevate the lifetime risks of pancreatic cancer in tation by independent laboratories, and thus variants may not
men and women [68]. be classified the same between labs. ClinVar is a public data-
base available upon which laboratories can contribute their vari- References
ant classifications to help unify classifications among labs [72].
1. Tung N, Battelli C, Allen B, Kaldate R, Bhatnagar S, Bowles K,
et al. Frequency of mutations in individuals with breast cancer
37. What is direct-to-consumer (DTC) testing, and how referred for BRCA1 and BRCA2 testing using next-generation
should it be used in patients with personal or family sequencing with a 25-gene panel. Cancer. 2015;121(1):25–33.
histories of breast cancer? 2. Shiovitz S, Korde LA. Genetics of breast cancer: a topic in evolu-
tion. Ann Oncol. 2015;26(7):1291–9.
3. Slavin TP, Maxwell KN, Lilyquist J, Vijai J, Neuhausen SL, Hart
Direct-to-consumer (DTC) testing is genetic testing that is
SN, et al. The contribution of pathogenic variants in breast can-
marketed toward the patient without the direct involvement cer susceptibility genes to familial breast cancer risk. NPJ Breast
of a medical provider. DTC testing can provide information Cancer. 2017;3:22.
such as ancestral background, phenotypic characteristics, or 4. Couch F, Shimelis H, Hu C, Hart SN, Polley EC, Na J, et al.
predilection to certain disease (e.g., Alzheimer’s, breast can- Associations between cancer predisposition testing panel genes and
breast cancer. JAMA Oncol. 2017;3(9):1190–6.
cer). Not all DTC companies are certified by CLIA or CAP 5. Madeleine MA. Optimal age to start preventative measures in
regulatory bodies; as such, clinicians need to be aware of the women with BRCA1/2 mutations or high familial breast cancer
quality of genetic test results from DTC testing. It is critical risk. Int J Cancer. 2012;133:156–64.
to assess the quality of genetic testing from DTC companies 6. Genetic/Familial High-Risk Assessment: Breast and Ovarian
Cancer 2018. National Comprehensive Cancer Network. Version
before using it to determine medical management [73, 74]. 2.2019; https://www.nccn.org/professionals/physician_gls/pdf/
genetics_screening.pdf
38. How is tumor sequencing of breast cancer different 7. Tai YC, Domchek S, Parmigiani G, Chen S. Breast cancer risk
from germline testing for breast cancer? among male BRCA1 and BRCA2 mutation carriers. JNCI: J Natl
Cancer Inst. 2007;99(23):1811–4.
8. Lee MV, Katabathina VS, Bowerson ML, Mityul MI, Shetty
Recent advancements in next-generation sequencing technol- AS, Elsayes KM, et al. BRCA-associated cancers: role of imag-
ogies allow for the sequencing of tumors to identify somatic ing in screening, diagnosis, and management. Radiographics.
mutations that may be driving the growth of the cancer. These 2017;37(4):1005–23.
9. Iavazzo C, Gkegkes ID, Vrachnis N. Primary peritoneal cancer in
mutations may be targetable by certain therapies, and thus BRCA carriers after prophylactic bilateral salpingo-oophorectomy.
somatic tumor testing can aid in cancer treatment [75]. Genetic J Turk Ger Gynecol Assoc. 2016;17(2):73–6.
testing for hereditary cancer predisposition syndromes evalu- 10. Brody LC, Biesecker BB. Breast cancer susceptibility genes.

ates for germline mutations, or mutations that an individual BRCA1 and BRCA2. Medicine (Baltimore). 1998;77(3):208–26.
11. Takaoka M, Miki Y. BRCA1 gene: function and deficiency. Int J
inherits. Germline mutations can be identified via somatic
Clin Oncol. 2018;23(1):36–44.
tumor testing; thus, clinicians ordering tumor testing should 12. Yoshida K, Miki Y. Role of BRCA1 and BRCA2 as regulators of
counsel their patients accordingly [76]. However, germline DNA repair, transcription, and cell cycle in response to DNA dam-
mutations can also be filtered out through somatic testing age. Cancer Sci. 2004;95(11):866–71.
reporting pipelines. Thus, individuals undergoing somatic 13. Ferla R, Calo V, Cascio S, Rinaldi G, Badalamenti G, Carreca I,
et al. Founder mutations in BRCA1 and BRCA2 genes. Ann Oncol.
tumor testing may still need a hereditary cancer work-up. As 2007;18(Suppl 6):vi93–8.
such, consideration of paired tumor and germline testing is 14. Ferla R, Calò V, Cascio S, Rinaldi G, Badalamenti G, Carreca I,
preferred as a means to avoid missed mutations [77]. Surmacz E, Colucci G, Bazan V, Russo A. Founder mutations in
BRCA1 and BRCA2 genes. Ann Oncol. 2007;18(6):vi93–8.
15. Ossa CA, Torres D. Founder and recurrent mutations in BRCA1
39. What happens if someone inherits two genetic muta- and BRCA2 genes in Latin American countries: state of the art and
tions in the same gene that predisposes to breast literature review. Oncologist. 2016;21(7):832–9.
cancer? 16. Bu R, Siraj AK, Al-Obaisi KA, Beg S, Al Hazmi M, Ajarim D, et al.
Identification of novel BRCA founder mutations in Middle Eastern
breast cancer patients using capture and Sanger sequencing analy-
Individuals who inherit two BRCA2 or two PALB2 gene muta- sis. Int J Cancer. 2016;139(5):1091–7.
tions (one from each parent) have significant genomic instabil- 17. Kwong A, Wong LP, Wong HN, Law FB, Ng EK, Tang YH, et al.
ity. As such, they have a rare, recessive condition known as A BRCA2 founder mutation and seven novel deleterious BRCA
Fanconi anemia that results in physical abnormalities and mutations in southern Chinese women with breast and ovarian can-
cer. Breast Cancer Res Treat. 2009;117(3):683–6.
bone marrow failure that frequently occurs in childhood/
18. Rao R, Rivers A, Rahimi A, Wooldridge R, Rao M, Leitch M, et al.
young adulthood [78]. Individuals who inherit mutations in Genetic Ancestry using Mitochondrial DNA in patients with Triple-
both alleles of the ATM gene mutation are known to have negative breast cancer (GAMiT study). Cancer. 2017;123(1):107–13.
ataxia-telangiectasia, a recessive syndrome that significantly 19. Manders P, Pijpe A, Hooning MJ, Kluijt I, Vasen HF, Hoogerbrugge
N, et al. Body weight and risk of breast cancer in BRCA1/2 muta-
affects childhood development, particularly motor control,
tion carriers. Breast Cancer Res Treat. 2011;126(1):193–202.
and greatly increases the risk for multiple types of cancer [79].
200 R. Rao et al.
20. Rieder V, Salama M, Glockner L, Muhr D, Berger A, Tea MK, 37. Yao K, Liederbach E, Tang R, Lei L, Czechura T, Sisco M, et al.
et al. Effect of lifestyle and reproductive factors on the onset of Nipple-sparing mastectomy in BRCA1/2 mutation carriers: an
breast cancer in female BRCA 1 and 2 mutation carriers. Mol Genet interim analysis and review of the literature. Ann Surg Oncol.
Genomic Med. 2016;4(2):172–7. 2015;22(2):370–6.
21. Kiechle M, Engel C, Berling A, Hebestreit K, Bischoff SC, Dukatz 38. Rao R, Euhus D, Mayo HG, Balch C. Axillary node interventions in
R, et al. Effects of lifestyle intervention in BRCA1/2 mutation car- breast cancer: a systematic review. JAMA. 2013;310(13):1385–94.
riers on nutrition, BMI, and physical fitness (LIBRE study): study 39. Baretta Z, Mocellin S, Goldin E, Olopade OI, Huo D. Effect of
protocol for a randomized controlled trial. Trials. 2016;17:368. BRCA germline mutations on breast cancer prognosis: a sys-
22. Hartmann LC, Lindor NM. The role of risk-reducing sur-
tematic review and meta-analysis. Medicine (Baltimore).
gery in hereditary breast and ovarian cancer. N Engl J Med. 2016;95(40):e4975.
2016;374(5):454–68. 40. Zimmer AS, Gillard M, Lipkowitz S, Lee JM. Update on

23. Society of Gynecology Oncology. Hereditary breast and ovarian PARP inhibitors in breast cancer. Curr Treat Options in Oncol.
cancer syndrome. ACOG Practice Bulletin. 2017; https://www.sgo. 2018;19(5):21.
org/wp-content/uploads/2012/09/PB-182.pdf 41. Schneider K, Zelley K, Nichols KE, Garber J. Li-Fraumeni syn-
24. Menon U, Gentry-Maharaj A, Hallett R, Ryan A, Burnell M, Sharma drome. In: Adam MP, Ardinger HH, Pagon RA, Wallace SE, Bean
A, et al. Sensitivity and specificity of multimodal and ultrasound LJH, Stephens K, et al., editors. GeneReviews®. Seattle: University
screening for ovarian cancer, and stage distribution of detected of Washington; 1993.
cancers: results of the prevalence screen of the UK Collaborative 42. Valachis A, Nearchou AD, Lind P. Surgical management of breast
Trial of Ovarian Cancer Screening (UKCTOCS). Lancet Oncol. cancer in BRCA-mutation carriers: a systematic review and meta-
2009;10(4):327–40. analysis. Breast Cancer Res Treat. 2014;144(3):443–55.
25. Buys SS, Partridge E, Black A, Johnson CC, Lamerato L, Isaacs C, 43. Mai PL, Best AF, Peters JA, DeCastro RM, Khincha PP, Loud JT,
et al. Effect of screening on ovarian cancer mortality: the Prostate, et al. Risks of first and subsequent cancers among TP53 mutation-
Lung, Colorectal and Ovarian (PLCO) cancer screening random- carriers in the NCI LFS cohort. Cancer. 2016;122(23):3673–81.
ized controlled trial. JAMA. 2011;305(22):2295–303. 44. Evans DG, Birch JM, Ramsden RT, Sharif S, Baser ME. Malignant
26. Friebel TM, Domchek SM, Rebbeck TR. Modifiers of cancer risk transformation and new primary tumours after therapeutic radia-
in BRCA1 and BRCA2 mutation carriers: systematic review and tion for benign disease: substantial risks in certain tumour prone
meta-analysis. J Natl Cancer Inst. 2014;106(6):dju091. syndromes. J Med Genet. 2006;43:289–94.
27. Crum CP, Drapkin R, Kindelberger D, Medeiros F, Miron A, Lee 45. Yalaza M, İnan A, Bozer M. Male breast cancer. J Breast Health.
Y. Lessons from BRCA: the tubal fimbria emerges as an origin for 2016;12(1):1–8.
pelvic serous cancer. Clin Med Res. 2007;5:35–44. 46. Zhang S, Phelan CM, Zhang P, Rousseau F, Ghadirian P, Robidoux
28. Huzarski T, Byrski T, Gronwald J, Gorski B, Domagala P, Cybulski A, et al. Frequency of the CHEK2 1100delC mutation among
C, et al. Ten-year survival in patients with BRCA1-negative and women with breast cancer: an international study. Cancer Res.
BRCA1-positive breast cancer. J Clin Oncol. 2013;31(26):3191–6. 2008;68(7):2154–7.
29. Valentini A, Lubinski J, Byrski T, Ghadirian P, Moller P, Lynch HT, 47. Pritzlaff M, Summerour P, McFarland R, Shuwei L, Reineke P,
et al. The impact of pregnancy on breast cancer survival in women Dolinsky JS, et al. Male breast cancer in a multi-gene panel testing
who carry a BRCA1 or BRCA2 mutation. Breast Cancer Res Treat. cohort: insights and unexpected results. Breast Cancer Res Treat.
2013;142(1):177–85. 2017;161:575.
30. Domchek SM, Friebel TM, Singer CF, Evans DG, Lynch HT, 48. Ferzoco RM, Ruddy KJ. The epidemiology of male breast cancer.
Isaacs C, et al. Association of risk-reducing surgery in BRCA1 or Curr Oncol Rep. 2016;18(1):1.
BRCA2 mutation carriers with cancer risk and mortality. JAMA. 49. Deb S, Lakhani SR, Ottini L, Fox SB. The cancer genetics and
2010;304(9):967–75. pathology of male breast cancer. Histopathology. 2016;68:110–8.
31. Metcalfe K, Lynch HT, Foulkes WD, Tung N, Kim-Sing C,
50. Foulkes WD, Smith IE, Reis-Filho JS. Triple-negative breast can-
Olopade OI, et al. Effect of oophorectomy on survival after breast cer. N Engl J Med. 2010;363(20):1938–48.
cancer in BRCA1 and BRCA2 mutation carriers. JAMA Oncol. 51. Plasilova ML, Hayse B, Killelea BK, Horowitz NR, Chagpar

2015;1(3):306–13. AB, Lannin DR. Features of triple-negative breast cancer: analy-
32. Warner E, Plewes DB, Hill KA, Causer PA, Zubovits JT, Jong RA, sis of 38.813 cases from the national cancer database. Medicine
et al. Surveillance of BRCA1 and BRCA2 mutation carriers with (Baltimore). 2016;95(35):e4614.
magnetic resonance imaging, ultrasound, mammography, and clini- 52. Atchley DP, Albarracin CT, Lopez A, Valero V, Amos CI, Gonzalez-
cal breast examination. JAMA. 2004;292(11):1317–25. Angulo AM, et al. Clinical and pathologic characteristics of
33. Molina-Montes E, Perez-Nevot B, Pollan M, Sanchez-Cantalejo E, patients with BRCA-positive and BRCA-negative breast cancer. J
Espin J, Sanchez MJ. Cumulative risk of second primary contra- Clin Oncol. 2008;26:4282–8.
lateral breast cancer in BRCA1/BRCA2 mutation carriers with a 53. Couch FJ, Hart SN, Sharma P, Toland AE, Wang X, Miron P, et al.
first breast cancer: a systematic review and meta-analysis. Breast. Inherited mutations in 17 breast cancer susceptibility genes among
2014;23(6):721–42. a large triple-negative breast cancer cohort unselected for family
34. Lee LJ, Alexander B, Schnitt SJ, Comander A, Gallagher B, Garber history of breast cancer. J Clin Oncol. 2015;33(4):304–11.
JE, et al. Clinical outcome of triple negative breast cancer in BRCA1 54. Apostolou P, Papasotiriou I. Current perspectives on CHEK2

mutation carriers and noncarriers. Cancer. 2011;117(14):3093–100. mutations in breast cancer. Breast Cancer (Dove Med Press).
35. King MC, Wieand S, Hale K, Lee M, Walsh T, Owens K, et al. 2017;9:331–5.
Tamoxifen and breast cancer incidence among women with inher- 55. Beggs AD, Latchford AR, Vasen HFA, Moslein G, Alonso A, Aretz
ited mutations in BRCA1 and BRCA2: National Surgical Adjuvant S, et al. Peutz–Jeghers syndrome: a systematic review and recom-
Breast and Bowel Project (NSABP-P1) Breast Cancer Prevention mendations for management. Gut. 2010;59:975–86.
Trial. JAMA. 2001;286(18):2251–6. 56. Fitzgerald RC, Hardwick R, Huntsman D on behalf of the

36. de Alcantara Filho P, Capko D, Barry JM, Morrow M, Pusic A, International Gastric Cancer Linkage Consortium, et al. Hereditary
Sacchini VS. Nipple-sparing mastectomy for breast cancer and diffuse gastric cancer: updated consensus guidelines for clini-
risk-reducing surgery: the Memorial Sloan-Kettering Cancer cal management and directions for future research. J Med Genet.
Center experience. Ann Surg Oncol. 2011;18(11):3117–22. 2010;47:436–44.
57. Meijers-Heijboer H, van den Ouweland A, Klijn J, Wasielewski M, 68. Roberts N, Jiao Y, Yu J, Kopelovich L, Petersen GM, Bondy ML,
de Snoo A, Oldenburg R, et al. Low-penetrance susceptibility to et al. ATM mutations in patients with hereditary pancreatic cancer.
breast cancer due to CHEK2(∗)1100delC in noncarriers of BRCA1 Cancer Discov. 2011;201(2):41–6.
or BRCA2 mutations. Nat Genet. 2002;31(1):55–9. 69. Torkamani A. The personal and clinical utility of polygenic risk
58. Cybulski C, Huzarski T, Byrski T, Gronwald J, Dębniak T, Jakubowska scores. Nat Rev Genet. 2018;19(9):581–90.
A, et al. Estrogen receptor status in CHEK2-positive breast cancers: 70. Mauer CB, Pirzadeh-Miller SM, Robinson LD, Euhus DM. The
implications for chemoprevention. Clin Genet. 2009;75:72–8. integration of next-generation sequencing panels in the clinical
59. Apostolou P, Fostira F, Mollaki V, Delimitsou A, Vlassi M,
cancer genetics practice: an institutional experience. Genet Med.
Pentheroudakis G, et al. Characterization and prevalence of two 2014;16:407–12.
novel CHEK2 large deletions in Greek breast cancer patients. J 71. Richards S, Aziz N, Bale S, Bick D, Das S, Gastier-Foster J, et al.
Hum Genet. 2018;63:877–86. Standards and guidelines for the interpretation of sequence vari-
60. Schmidt MK, Hogervorst F, van Hien R, Cornelissen S, Broeks ants: a joint consensus recommendation of the American College of
A, Adank MA, et al. Age- and tumor subtype–specific breast can- Medical Genetics and Genomics and the Association for Molecular
cer risk estimates for CHEK2∗1100delC carriers. J Clin Oncol. Pathology. Genet Med. 2015;17(5):405–24.
2016;34(23):2750–60. 72. Landrum M, Lee J, Riley G, Jang W, Rubinstein W, Church D, et al.
61. Cybulski C, Górski B, Huzarski T, Masojć B, Mierzejewski M, ClinVar. In: The NCBI handbook [Internet]. 2nd ed. Bethesda:
Debniak T, et al. CHEK2 is a multiorgan cancer susceptibility gene. National Center for Biotechnology Information (US); 2013.
Am J Hum Genet. 2004;75(6):1131–5. 73. Kaye J. The regulation of direct-to-consumer genetic tests. Hum
62. Mester J, Eng C. Cowden syndrome: recognizing and managing a not- Mol Genet. 2008;17(R2):R180–3.
so-rare hereditary cancer syndrome. J Surg Oncol. 2015;111:125–30. 74. Tandy-Connor S, Guiltinan J, Krempely K, LaDuca H, Reineke
63. Ngeow J, Stanuch K, Mester JL, Barnholtz-Sloan JS, Eng
P, Gutierrez S, et al. False-positive results released by direct-to-
C. Second malignant neoplasms in patients with Cowden syn- consumer genetic tests highlight the importance of clinical con-
drome with underlying germline PTEN mutations. J Clin Oncol. firmation testing for appropriate patient care. Genet Med. 2018;
2014;32(17):1818–24. https://doi.org/10.1038/gim.2018.38. [Epub ahead of print].
64. Ahmed M, Rahman N. ATM and breast cancer susceptibility. Hum 75. Spencer DH, Ley TJ. Sequencing of tumor DNA to guide cancer
Mutat. 2006;27(11):1122–8. risk assessment and therapy. JAMA. 2018;319(14):1497–8.
65. Bernstein JL, Teraoka S, Southey MC, Jenkins MA, Andrulis IL, 76. Raymond VM, Gray SW, Roychowdhury S, Joffe S, Chinnaiyan
Knight JA, et al. Population-based estimates of breast cancer risks AM, Parsons DW, et al. Germline findings in tumor-only sequenc-
associated with ATM gene variants c.7271T>G and c.1066-6T>G ing: points to consider for clinicians and laboratories. J Natl Cancer
(IVS10-6T>G) from the Breast Cancer Family Registry. Hum Inst. 2016;108(4):pii:djv351.
Mutat. 2006;27(11):1122–8. 77. Mandelker D. Toward concurrent testing for somatic and germline
66. Dombernowsky SL, Weischer M, Allin KH, Bojesen SE, Tybjaerg- variants in cancer patients. Clin Cancer Res. 2016;22(16):3987–8.
Hansen A, Nordestgaard BG. Risk of cancer by ATM missense muta- 78. Howlett NG, Taniguchi T, Olson S, Cox B, Waisfisz Q, De Die-
tions in the general population. J Clin Oncol. 2008;26(18):3057–62. Smulders C, et al. Biallelic inactivation of BRCA2 in Fanconi ane-
67. Meyer A, Wilhelm B, Dörk T, Bremer M, Baumann R, Karstens JH, mia. Science. 2002;297(5581):606–9.
et al. ATM missense variant P1054R predisposes to prostate cancer. 79. McKinnon PJ. ATM and ataxia telangiectasia. EMBO Rep.

Radiother Oncol. 2007;83(3):283–8. 2004;5(8):772–6.
Mesenchymal and Lymphoid Lesions
in the Breast 9
Xi Wang and Andrew G. Evans
List of Frequently Asked Questions spindle cell carcinoma (MSC) and malignant phyllodes (MP)
tumor which could be even more common tumors in the
1. How does one approach breast mesenchymal lesions? breast, could greatly mimic the mesenchymal lesions, mak-
ing the differential diagnosis difficult. Meanwhile, the post-
Histologically, the breast stroma could be divided into intra- treatment changes – such as fat necrosis, scarring, vascular
lobular and interlobular stroma [1]. Tumors, such as fibroad- proliferation, etc. – are very common in this location, given
enomas and phyllodes tumors, are considered as the that breast cancer is the most common malignancy in women
neoplasms arising from the intralobular stroma of the termi- and breast conservation is the common practice nowadays.
nal ductal-lobular unit (TDLU). They are virtually breast- These changes could be sometimes exuberant and impose
specific neoplasms and technically are not considered as diagnostic challenges for mesenchymal tumors. Furthermore,
breast mesenchymal lesions, even though they could share post-radiation sarcomas, especially angiosarcomas, and pos-
morphological similarities with other breast mesenchymal sible post-implant tumors have added to the complexity of
tumors. The interlobular stroma, on the other hand, may not breast mesenchymal lesions and should always be in the dif-
be breast specified; and the lesions/tumors arising in this ferential diagnosis in a patient with history of breast cancer
location are generally the same as the soft tissue lesions/ and treatments.
tumors in other anatomy locations, in terms of types of lesion As in the general soft tissue tumor classifications, breast
and morphology. One could appreciate this outlook from the mesenchymal tumors have been classified based on the types
2012 WHO classification of breast mesenchymal tumors of cell differentiation in the 2012 WHO classifications of
wherein pseudoangiomatous stromal hyperplasia (PASH) is tumors of the breast, such as fibroblastic/myofibroblastic
probably the only lesion relatively unique to the breast paren- lesions, lipocytic tumors, vascular lesions, peripheral nerve
chyma. Myofibroblastoma might be primarily a breast tumor. sheath tumors, muscular (including smooth muscle and skel-
However, the extramammary version does exist and is not etal muscle) differentiation lesions, and bone/cartilage
always along the anatomical “milk line.” matrix-forming lesions. However, in daily practice, the
Nevertheless, the approach to breast mesenchymal lesions growth pattern recognition might be more practical for initial
is unique, compared with the other organ systems. Non- differentiation, such as spindle cell vs. vascular vs. lipocytic
mesenchymal malignancies in the breast, such as metaplastic lesions. Giving the breast as the specific location, this chap-
ter will focus on the mesenchymal lesions and their mimics
relatively unique to the breast.
2. What are the entities that should be considered in the

differential diagnoses of spindle cell lesions in the
breast?
In lesions with spindle cell growth pattern, some authors pre-

X. Wang (*) · A. G. Evans fer the approach of dividing the lesions into bland-appearing
Department of Pathology and Laboratory Medicine,
and malignant-appearing lesions [2]. However, there might
University of Rochester Medical Center, Rochester, NY, USA
e-mail: [email protected]: be lesions that will fall in between these simplified
[email protected] groupings.

204 X. Wang and A. G. Evans
In principle, metaplastic spindle cell carcinoma (MSC)

should always be the number one on the list of differential
diagnoses for spindle cell lesions in the breast, regardless of
bland- or malignant-appearing morphology. Even though a
carcinoma, MSC could be lacking of epithelial component,
thus appearing as a stromal lesion [3]. This distinction is very
important, as the biology and clinical behavior of MSC might
be significantly different from other spindle cell lesions.
The bland-appearing spindle cell lesions are relatively
more common in the breast parenchyma, compared with
malignant-appearing tumors. They include previous proce-
dure site scar, nodular fasciitis, pseudoangiomatous stromal
hyperplasia, primary mammary fibromatosis, and mammary
myofibroblastoma. They are the tumors of fibroblast/myofi-
broblast proliferation and often with morphological and
immunohistochemical overlap. The other types of low-grade Fig. 9.1 The so-called poorly differentiated angiosarcoma with pre-
spindle cell lesions, not by any means particular to the breast, dominantly spindle cell appearance (100×)
but having been reported in the breast parenchyma, include
spindle cell lipoma, solitary fibrous tumor, leiomyoma, and
peripheral nerve sheath tumor. Benign and borderline phyl-
lodes tumor usually will not enter as a differential diagnosis
in this group, as the benign ductal epithelium will usually
present and associate with spindle cells in a distinct pattern
in phyllodes tumor.
The morphology of inflammatory myofibroblastic tumor
(IMT) could be variable. The myxoid vascular pattern or
scar-like pattern could appear bland, while the compact fas-
cicular spindle cell pattern could mimic malignancy
morphologically.
The list of differential diagnoses for malignant-appearing
spindle cell lesions, other than metaplastic spindle cell carci-
noma and malignant phyllodes tumor, is very short.
Angiosarcoma, primary or secondary, is the most common
sarcoma in the breast. It could appear as a spindle cell lesion,
when the vascular differentiation is not obvious (so-called Fig. 9.2 Malignant melanoma with spindle cell appearance (100×)
poorly differentiated). Other spindle cell sarcomas are very
rare in the breast, with leiomyosarcoma, synovial sarcoma, proliferation of fibroblasts and myofibroblasts. Typically, the
rhabdomyosarcoma, malignant peripheral nerve sheath lesional cells are elongated, bland, and uniform spindle cells,
tumor (MPNST), and others being reported. A metastasis to forming long, sweeping fascicles and infiltrating into the sur-
the breast from other anatomy locations should be ruled out rounding breast parenchyma. The diagnostic challenge is to
first, before the diagnosis of a breast primary sarcoma can be differentiate it, most importantly from fibromatosis- like
rendered. It is very important to rule out melanoma, primary metaplastic carcinoma (FLMC), as they share many morpho-
from breast skin or metastatic from other locations, in this logical similarities: (1) They are both infiltrative. (2) Both
setting, as it could assume spindle cell morphology and its are rich in collagen. (3) Tumor cellularity is similarly mod-
clinical significance could be completely different from est. (4) Cells are not as pleomorphic as other high-grade
other high-grade sarcomas (Figs. 9.1 and 9.2). tumors (Fig. 9.3a, b).
There are a few morphological features which could be
3. How does one differentiate mammary fibromatosis helpful in this regard. Firstly, FLMC could be associated
from fibromatosis-like metaplastic carcinoma with ductal carcinoma in situ (DCIS), or clusters of epithe-
(FLMC)? lioid cells could be identified in between the spindle cell
bundles, while fibromatosis will not have this association.
Fibromatosis is an uncommon neoplasm in the breast, with Secondly, the cells in fibromatosis are uniformly bland with
the incidence rate of around 0.2% in breast tumors. It is the pale nuclei and inconspicuous nucleoli, while the cells in
9 Mesenchymal and Lymphoid Lesions in the Breast 205
a b
Fig. 9.3 (a) Fibromatosis with infiltrative long sweeping fascicles (40×). (b) FLMC morphologically resembles fibromatosis (40×)
Beta-Catenin may not be helpful in this regard, as it could be

positive in FLMC [6, 7].
4. How does one differentiate nodular fasciitis from

metaplastic carcinoma?
Nodular fasciitis is a rare spindle cell lesion in the breast. It

consists of plump but uniform fibroblasts/myofibroblasts
forming irregular short bundles/fascicles in a myxoid back-
ground. The features such as focally infiltrative border, irreg-
ular spindle cell bundles/fascicles, active mitosis, and mixed
inflammatory cells make it often morphologically overlap
with metaplastic carcinoma, especially fibromatosis-like
metaplastic carcinoma (FLMC). There are case reports
showing that nodular fasciitis could be positive for cytokera-
Fig. 9.4 Epithelioid cell clusters in between the spindle cells in FLMC tin cocktail, which further complicates the issue [8].
(400×) However, the cells in nodular fasciitis are, even though
plump, uniform with pale-staining nuclei. There are often
numerous small vessels resembling granulation tissue. The
FLMC will at least have certain pleomorphism and hyper- mitotic activity could be brisk but not atypical, and p63 and
chromasia. One could appreciate this feature by referenc- high-molecular-weight cytokeratin should be negative. On
ing the nearby endothelial cells. The vasculature in the other hand, metaplastic spindle cell carcinoma, even the
fibromatosis is usually prominent, and the endothelial cells FLMC type, should have certain cellular pleomorphism,
are generally more prominent than the lesional cells, while such as hyperchromasia. It will usually have scattered clus-
in FLMC, the tumor cells are more prominent and pleomor- ters of epithelioid cells mixed with spindle cells. Atypical
phic compared with endothelial cells. Thirdly, lymphoid mitosis could be identified, and at least some of the epithelial
aggregates or even lymphoid follicles are often identified at markers in a panel including p63 and high-molecular-weight
the periphery of fibromatosis, while FLMC could have cytokeratin will be positive, especially for those epithelioid
some chronic inflammatory cells mixed with the spindle cells (Fig. 9.6a, b).
cells (Figs. 9.4 and 9.5a, b).
A panel of immuno markers will be very helpful. It could 5. How does one differentiate mammary myofibroblas-
not be emphasized more that a group of epithelial markers – toma from its mimics?
such as pan-cytokeratin, p63, and high-molecular-weight
cytokeratin – should be applied, as the FLMC could be only Mammary myofibroblastoma is a usually bland-appearing
focally positive or even negative for one of these markers [4, 5]. spindle cell neoplasm composed of fibroblasts/myofibro-
a b
Fig. 9.5 (a) The spindle cells in fibromatosis are less prominent than the adjacent endothelial cells (200×). (b) The spindle cells in FLMC are more
prominent than the adjacent endothelial cells (200×)
a b
Fig. 9.6 (a) Nodular fasciitis with short irregular spindle cell fascicles in a myxoid background. The cells are uniform with pale-staining nuclei
(40×). (b) Mitosis is easy to find in nodular fasciitis (200×)
blasts. It is normally circumscribed, with adipocytes as the chemical stain. Myofibroblastoma is variably positive for
tumor component, without breast glandular structure in the desmin, CD34, ER, PR, and AR, but negative for cytokeratin
tumor mass. The spindle cells are plump and form short fas- markers. The differential from spindle cell lipoma might be
cicles intersecting with each other and often interrupted by more academic, instead of any practical value, as they are in
the thick deeply eosinophilic collagen bands. Mitosis could the same histological spectrum and share the same genetic
be identified, but the rate is very low (<2 mitosis per 10 high- changes (a deletion of 13q14, which includes RB1 and
power fields (HPFs)). The classical myofibroblastoma usu- FOXO1A genes) [9]. Fibromatosis may enter as a differen-
ally will not pose diagnostic challenge. However, this is the tial diagnosis, especially for the infiltrative variant of myofi-
tumor with a wide spectrum of morphological variants, such broblastoma. However, fibromatosis usually has only modest
as cellular, epithelioid, infiltrative, lipomatous, collagenized/ cellularity, with broad bands of long fascicles, and a different
fibrous, and myxoid variants. A differential diagnosis from immunohistochemical staining panel. The myxoid variant of
metaplastic spindle cell carcinoma is always required and myofibroblastoma could share some morphological similari-
could be achieved through the appreciation of the general ties with nodular fasciitis. It is very rare and has only been
tumor features such as demarcation, no epithelial compo- reported in male patients [10]. The circumscription, low
nents, and low mitotic count, as well as the immunohisto- mitotic rate, and CD34 positivity could be helpful features in
Fig. 9.7 Myofibroblastoma with thick deeply eosinophilic collagen Fig. 9.8 Scar tissue radiating out from the center (40×)
bands (40×)
distinction. Myofibroblastoma could exhibit smooth muscle could be the first clue for scar tissue. However, this history
differentiation, which may complicate the diagnosis of leio- may not always be provided. Microscopically, compared
myoma in the breast. It is speculated that some of the reported with fibromatosis and other lesions, the spindle cell prolif-
“parenchymal leiomyoma” may represent the myofibroblas- eration in scar tissue is less cellular and denser with abundant
toma with smooth muscle differentiation. It appears that this collagen; the cells are usually stellate shaped, less organized,
variant could be negative for CD34 – the immunostain hall- not forming a long fascicular pattern, and radiating from the
marker of myofibroblastoma – but with the deletion of center of the lesion. Even though with irregular borders, scar
13q14, the molecular signature for myofibroblastoma tissue is usually locally restricted, not as destructively infil-
(Fig. 9.7) [11]. trative as fibromatosis. Further, depending on the age of the
scar tissue, there might be associated hemosiderin-laden
6. What are the breast mesenchymal lesions which could macrophages, foreign body-type granulomatous changes,
have smooth muscle differentiation? and fat necrosis. Beta-Catenin is generally negative, while it
is nuclear positive in approximately 80% of fibromatosis
Breast leiomyoma mostly arises from the skin or nipple/are- (Figs. 9.8 and 9.9a, b) [12].
olar complex, while breast parenchymal leiomyoma is rare.
Other benign tumors, which could have smooth muscle dif- 8. How does one differentiate inflammatory myofibro-
ferentiation, include myoid hamartoma, fibromatosis, fibroblastic tumor from spindle cell metaplastic
adenoma, benign phyllodes tumors, and myofibroblastoma. carcinoma?
Usually, these tumors will have their own characteristic fea-
tures, other than smooth muscle differentiation. Primary Inflammatory myofibroblastic tumor (IMT) is a distinct neo-
breast parenchymal leiomyosarcoma is even more rare, with plasm of myofibroblast/fibroblast spindle cell proliferation
about 50 cases being reported in the literature. mixed with prominent inflammatory cells. It can virtually
Leiomyosarcoma could also be a heterologous component of occur in any anatomy locations throughout the body, includ-
metaplastic carcinoma or malignant phyllodes tumor. ing the breast. The age for IMT in the breast ranges from 33
Extensive sampling and immunohistochemical staining to 76 years according to the case reports. Its morphological
should help the distinction. features can be variable among cases, depending on the cel-
lularity of spindle cells, collagenous background, and den-
7. How does one differentiate previous procedure site sity of inflammatory cells. The most important issue in breast
scar from fibromatosis and other lesions? pathology is to differentiate IMT from metaplastic spindle
cell carcinoma (MSC), as they do share some morphological
Previous procedure site scar could very much resemble similarities: (1) They both consist of spindle cells with vari-
mammary fibromatosis microscopically and sometimes enter able cellularity. (2) They both commonly have inflammatory
as the differential diagnosis of desmoplastic-like metaplastic cell infiltration. (3) Same as metaplastic carcinoma, osseous
carcinoma. Clinical history of previous surgery or trauma metaplasia can occasionally be seen in IMT. However, the
a b
Fig. 9.9 (a) Scar tissue is locally restricted (40×). (b) Fibromatosis destructively infiltrating in the breast parenchyma (40×)
a b
Fig. 9.10 (a, b) Inflammatory myofibroblastoma with spindle cell proliferation and inflammatory cell infiltration (40×, 20×)
presence of epithelial carcinoma – invasive or in situ, in or and high-molecular-weight cytokeratin will be necessary in
around the lesional spindle cells – will be the key feature to this regard (Fig. 9.10a, b).
recognize metaplastic carcinoma. The spindle cells in MSC
will usually show certain pleomorphism, mitotic activity, 9. How does one distinguish lipoma from normal adipose
and even necrosis, while these are not common features in tissue in the breast?
IMT. Immunohistochemically, 50–60% of IMT will be ALK
positive, while MSC usually will be at least focally positive Lipoma is a neoplasm consisting of mature adipocytes with
for one of the cytokeratin markers. Nevertheless, caution a thin fibrous capsule. The adipocytes in a lipoma are mor-
should be taken to differentiate IMT from MSC based on phologically not much different from the adipose tissue in
immunohistochemical stain only. As reported, about one- the breast parenchyma. The distinction between lipoma and
third of the IMT will be focally weakly positive for cytokera- localized overgrowth of fat in the breast, especially on a core
tin [13], while MSC could be only cytokeratin focally biopsy, if it is by any means necessary, could be impossible.
positive or, under rare circumstance, negative [4]. Also, ALK A demarcation by a thin fibrous membrane somewhere in the
stain is usually negative in IMTs in patients at an older age core could be the only hint for an intramammary lipoma.
[14], when the differential from breast carcinoma is even Fatty cores deprived of any breast glandular elements could
more pertinent. A panel of immune markers including p63 be the feature helping this distinction. The variant types of
lipoma, such as spindle cell lipoma and angiolipoma, usually be normal breast glandular tissue among the “liposarcoma”-
should not be a problem for diagnosis, as they will have their like fatty tissue. A true liposarcoma is usually a well-
own characteristic components to be distinguished from the demarcated mass often with a pushing border. It will not
fat in the breast parenchyma. “infiltrate” the breast parenchyma in such a manner. The
lipocytes in lipoatrophy are morphologically uniform in size,
10. What are the morphological mimics of liposarcoma and the nuclei are pale in color, unlike the variable sizes of
in the breast parenchyma? fat vacuoles and hyperchromatic nuclei of lipoblasts
(Figs. 9.11a, b and 9.12a).
Giving the breast as the specific location, the most important
differential diagnosis in this regard includes silicone granu- 12. How does one differentiate fat necrosis from
loma from the leaking of breast implant, fat necrosis due to liposarcoma?
trauma or previous surgery, and fat atrophy. These are the
lesions more often encountered in the breast parenchyma. Fat necrosis is a relatively common event in the breast paren-
Regular lipoma in the breast parenchyma or chest subcutane- chyma. The main etiology could be grouped as (1) trauma,
ous tissue could present with focal fat necrosis, which could which includes all kinds of surgical procedures, injections,
impose a diagnostic challenge. The variants of lipoma, such and incidental injuries; (2) radiation therapy; (3) fibrocystic
as pleomorphic lipoma/spindle cell lipoma and angiolipoma, changes, such as ductal ectasia; (4) breast infection; and (5)
could also mimic liposarcoma sometimes. anticoagulative treatment. These events could result in the
disruption of the oxygen supply to the fat cells and cause fat
11. How does one differentiate lipoatrophy from necrosis.
liposarcoma? Clinically, fat necrosis often presents as a palpable lump.
Sometimes it may show up on a screening mammogram as a
Breast lipoatrophy could occur in malnutrition associated mass. A core biopsy is usually conducted with a suspicion
with eating disorders such as anorexia nervosa or with for malignancy, mostly carcinoma.
chronic disease. Localized breast lipoatrophy could be the Fat necrosis is basically a non-suppurative inflammatory
result of pressure from the breast transplant or local trauma/ process. In the early stage, the disrupted fat will present as
injury to the breast parenchyma. The intracellular lipid is irregular fatty space with lipid-laden macrophages scattered
markedly reduced in lipoatrophy, so that the lipocytes are in between. This picture could be confused with an atypical
shrieked, which makes the nuclei much more prominent, lipomatous tumor/well-differentiated liposarcoma, espe-
superficially resembling the morphology of lipoblasts. cially in the thick sections when the nuclei of the macro-
However, there is no mass formation clinically, and histo- phages become somewhat hyperchromatic, mimicking
logically it is rather a diffuse process, instead of a localized atypical stromal cells. However, the macrophages in fat
neoplasm. No clear demarcation could be appreciated. The necrosis are with rounded nuclei and usually with faint stain-
lobular architecture of normal fat is maintained. There will ing in adequate sections. The fat vacuoles in these cells are
a b
Fig. 9.11 (a) Lipocytes in lipoatrophy with uniform nuclear size and fat vacuoles (200×). (b) Lobulation of lipoatrophy (40×)
inflammatory cells such as plasma cells, hemosiderin deposi-

a
tion, and even central cavity formation due to liquefaction. In
the later stage, fibrosis will gradually replace the fatty tissue,
with dystrophic calcification and sometimes squamous meta-
plasia. The whole process of fat necrosis could last for sev-
eral years (Fig. 9.12b).
13. How does one differentiate silicone granuloma from

liposarcoma?
Understandably, this differential has its significance in breast

pathology. Silicone granuloma, as a foreign body inflamma-
tory reaction, is associated with breast implants for postmas-
tectomy reconstruction or for breast augmentation.
Microscopically, the vacuolated macrophages could resem-
b ble lipoblasts. Liposarcoma could present as a primary breast
sarcoma or as a component of a malignant neoplasm, such as
malignant phyllodes tumor and metaplastic carcinoma.
Different types of breast primary liposarcomas have been
reported in the literature.
In silicone granuloma, the markedly distended macro-
phages become multivacuolated with different sizes when
the silicone is lost during the tissue process. The vacuoles
will distort the nuclei, resembling lipoblasts. However, the
lipoblast-like macrophages usually contain small, vague to
poorly delimited “bubbles” with small bland centrally
located nuclei, while the real lipoblasts are with sharply out-
lined, often large cytoplasmic vacuoles and indented hyper-
chromatic nuclei. The vacuolated macrophages in silicone
granuloma will be immunohistochemically positive for
c CD68 and lysozyme and negative for S-100, in contrast to
lipoblasts. In general, lipoblasts in a well-differentiated lipo-
sarcoma are usually few, difficult to find, and never as numer-
ous as the lipoblast-like macrophages in silicone granuloma.
More importantly, the background of silicone granuloma is
an inflammatory process, with fibrosis, chronic inflamma-
tion, and foreign body giant cells, which will not be an
appropriate background for true “lipoblasts.” Of course, clin-
ical history of breast implant would be the key (Fig. 9.12c).
14. What lesions in the breast will have chondroid matrix

formation?
It is well known that breast carcinoma could have chondroid

matrix formation, a type of the metaplastic carcinoma. It
Fig. 9.12 (a) Lipoblasts with hyperchromatic indented nuclei (400×).
(b) Macrophages in fat necrosis with paler and not indented nuclei
usually has a poorer prognosis compared with other types of
(400×). (c) Macrophages in silicone granuloma with multivacuoles and metaplastic carcinomas [15]. Chondrosarcoma could also be
centrally located small nuclei (400) one of the heterologous components of a metaplastic carci-
noma [16], so as in malignant phyllodes tumor. Mammary
multiple and small, without nuclear indentation. More hamartoma, fibroadenoma, and myofibroblastoma are the
importantly, the other changes will serve as an appropriate benign breast tumors which could have chondroid matrix
background to indicate a fat necrosis process. These changes formation. Breast de novo chondroma and chondrosarcoma
include multinucleated giant cells engulfing lipid, chronic have been reported [17–19], even though not common. Other
lesions, such as intraductal papilloma [20], sclerosing adeno- pling to identify the leaf-like structure and benign ductal
sis, and even microglandular adenosis, have been reported to epithelium in phyllodes tumor or the carcinomatous com-
have chondroid or chondromyxoid changes. ponent (with the help of cytokeratin stain) in the metaplas-
tic carcinoma will be the more definitive way to distinguish
15. How does one differentiate chondroid matrix from these tumors. This distinction may have clinical relevance,
myxoid changes? as chondrosarcoma could behave differently from meta-
plastic carcinoma. Other benign conditions with chondroid
Myxoid changes are commonly recognized in both benign component usually should not cause confusions with
and malignant neoplasms as well as non-neoplastic reactive chondrosarcoma.
lesions. They share the bluish appearance with the chondroid Breast primary chondrosarcoma, same as breast primary
matrix on the H&E-stained sections, sometimes overlapping osteosarcoma, is a diagnosis of exclusion, when other possi-
with each other (so-called chondromyxoid changes). In fact, bilities have been excluded. These other possibilities, other
both myxoid extracellular matrix and chondroid matrix than the abovementioned MC and MP tumor, include metas-
mainly consist of proteoglycans and collages, but with dif- tasis to the breast from other locations and primary tumor of
ferences in the types of collages and relative proportions of the adjacent bone (rib or sternum) involving the breast.
the constituent macromolecules. The interplay of various
macromolecules determines their texture and consistency. 17. What breast lesions will have osteoid matrix
Under the light microscope, the chondroid matrix is more formation?
finely textured and assembled with lacunae and round to oval
nucleated chondrocytes, while the myxoid changes are thin- Osteoid matrix formation or osseous metaplasia could be
ner, faint blue to clear, less organized, and resembling edem- identified in a variety of lesions in the breast parenchyma,
atous changes (Fig. 9.13a, b). including reactive/non-neoplastic lesions and benign or
malignant tumors. It can also be present idiopathically in the
16. How does one differentiate chondrosarcoma from breast without any associated lesions [21, 22]. The non-
chondroid matrix-forming tumors in the breast? neoplastic lesions reported with osseous metaplasia include
cholesterol granuloma, lipogranuloma, fasciitis ossificans,
Breast primary sarcoma is very rare, with a prevalence of implant capsules (saline or silicone implants), amyloid tumor
0.5% of all breast tumors. Primary chondrosarcoma is [23], and post-radiation changes. Fibroadenoma, pleomor-
even rarer. Approximately less than 20 cases have been phic adenoma, and papilloma are the common benign neo-
reported in the literature. The differential of chondrosar- plasms which could have osseous metaplasia as a degenerative
coma from malignant phyllodes (MP) tumor or metaplastic change. Primary osteosarcoma of the breast is extremely
carcinoma (MC) with chondrosarcomatous component rare, but has been reported, representing 12.5% of mammary
could be difficult. Immunohistochemical staining is not sarcomas. Metaplastic carcinoma and malignant phyllodes
helpful in differentiating the chondroid area of MP tumor tumor are the more common malignancies that could harbor
and MC from primary chondrosarcoma. Extensive sam- osteosarcomatous component.
a b
Fig. 9.13 (a) Chondroid matrix (100×). (b) Myxoid changes (100×)

18. What are the subtypes of osteosarcoma reported in fasciitis ossificans may sometimes cause confusion. It is a
the breast? bone-forming subtype of nodular fasciitis. It shares the same
features as the myositis ossificans, but is more superficially
Breast primary osteosarcoma is part of the extraskeletal located (fascia or tendon). It consists of fibroblast and myo-
osteosarcoma. It typically arises in the late adulthood. fibroblast proliferation with osteoid and/or trabecular bone
Approximately 150 cases of breast primary osteosarcoma formation. The zonal distribution pattern may not be obvious
have been reported in the literature. It has the similar mor- as in myositis ossificans. In the cellular area, it could resem-
phology as the conventional osteosarcomas of the bone or ble osteosarcoma. However, the spindle cells are usually not
other extraskeletal osteosarcomas. The subtypes of osteosar- pleomorphic, with no atypical mitosis. More importantly, the
coma that have been reported in the breast include osteoblas- woven bone trabeculae are lined with osteoblasts, indicating
tic [24, 25], fibroblastic, and chondroblastic osteosarcomas a reactive process (Fig. 9.14a, b).
[26], with the osteoblastic type as the major subtype reported.
Telangiectatic osteosarcoma has been reported in a case 20. What are the differential diagnoses of granular cell
associated with recurrent phyllodes tumor [27]. Small cell tumor in the breast?
and well-differentiated subtypes have not been reported in
the breast. Approximately 8.5% of granular cell tumors arise in the
As mentioned in the diagnosis of breast primary chondro- breast. Breast granular cell tumor shares the same morphol-
sarcoma, breast primary osteosarcoma is also a diagnosis of ogy with the granular cell tumor in other locations. The
exclusion. Interestingly, breast primary osteosarcomas have tumor cells are polygonal to slightly spindle shape, with
been reported to be associated with pre-existing ossified round to oval nuclei, distinct nucleoli, and abundant eosino-
fibroadenoma [28]. As with other sarcomas, metastasis is philic granular cytoplasm. The cells are arranged in cords,
usually hematogenous. However, axillary lymph node metas- nests, or sheets, in an infiltrative pattern, in between the
tasis was reported [29]. fibrous stroma (sometimes could be collagenous) and breast
lobules, with the tendency to grow into the terminal ductal
19. How does one differentiate osteosarcoma from oste- and lobular units. The cells in several breast lesions could
oid matrix-forming tumors in the breast? resemble the granular cell tumor. Breast primary carcino-
mas, such as invasive apocrine carcinoma and histiocytic-
Osteosarcoma by definition is the sarcomatous tumor cells type lobular carcinoma, could have abundant granular-like
producing osteoid matrix. The pleomorphic tumor cells are cytoplasm and are with an infiltrative growth pattern.
embedded in the lace-like or sheet osteoid matrix. It will usu- However, these carcinomas are usually associated with in
ally not be a problem to differentiate it from other reactive/ situ carcinomas, with pleomorphism, and will be cytokeratin
repair processes or benign tumors with osteometaplasia, positive. When the granular cell tumor presents as a breast
such as fibroadenoma. In these benign lesions, the typical skin lesion involving dermis/subcutaneous soft tissue, mela-
original lesional tissue usually can be discerned. However, noma could enter as a differential diagnosis. Again, an in situ
a b
Fig. 9.14 (a) Osteoblasts rimming the ossification in ossifying fasciitis (100×). (b) Osteoblastic osteosarcoma with pleomorphic tumor cells
forming bone (100×)
a b
Fig. 9.15 (a) Granular cell tumor infiltrating the dense stroma (200×). (b) Invasive lobular carcinoma with histiocytic features, resembling granu-
lar cell tumor (200×)
process or junctional activity could usually be identified in has a different histology. The secondary sarcoma is in the
melanoma. Other than S-100 positivity, melanoma cells will irradiated field. There is a prolonged latency period between
also be melan A and HMB45 positive. Histiocytes in a the two malignancies. A wide range of different types of
reactive inflammatory process could superficially resemble radiation-induced sarcoma have been reported, including
granular cells. However, the histiocytes will not be organized angiosarcoma, leiomyosarcoma, liposarcoma, fibrosarcoma,
as the granular cells, will have other inflammatory or reactive chondrosarcoma, and osteosarcoma [32–34]. These sarco-
changes, and will be alpha1-antitrypsin and alpha1- mas share the same morphological pattern as their sporadic
antichymotrypsin positive, even though the histiocytes and counterparts. More often, radiation-induced sarcoma lacks
granular cells could be both S-100 and CD68 positive. distinct differentiation and is classified as pleomorphic sar-
Metastatic disease, such as renal cell carcinoma, due to its coma (previously MFH).
clearing and sometimes eosinophilic cytoplasm, should be
considered, especially for the patient with appropriate his- 22. Is breast implant associated with increased risk of
tory. RCC is usually rich in vasculature, with the tumor cells mesenchymal neoplasm?
arranged in a sinusoidal pattern along the vessels and immu-
nohistochemical stain positive for cytokeratin and PAX8 The long-term safety of silicone implant has been a particu-
(Fig. 9.15a, b). lar concern, since it was first introduced in 1962. Experimental
Granular cell tumor of the breast is usually benign and studies did show its carcinogenicity in rodents. After a cer-
usually will not recur even with positive margin [30]. The tain latent period, solid silicone compound could elicit mes-
criteria for malignant granular cell tumor have been a debate.enchymal sarcoma at the implantation site in susceptible
Cellular pleomorphism, mitosis, and tumor necrosis are the rodents through the so-called solid-state carcinogenesis,
well-accepted features for malignancy [30]. with an incidence of approximately 29–40%. However, well-
designed epidemiologic studies have not proven an associa-
21. What are the lesions that could be induced by radiation between silicone implants and malignant solid tumors in
tion therapy in the breast? human breasts. Nevertheless, it is well documented that sili-
cone implants are associated with anaplastic large cell lym-
Breast parenchyma atrophic changes, fibrosis, dystrophic phoma. Implant-associated breast fibromatosis has been
calcification, and ossification are the common benign post- reported in more than 30 cases, with the mean interval time
radiation changes. Radiotherapy-induced neoplasms are usu- between implant placement and tumor occurrence of 3 years
ally sarcomas. Post-radiation lymphoma is also reported. [35]. There have been three case reports on silicone implant-
The latency of radiotherapy-induced sarcoma is usually associated breast stromal sarcomas. The first case was a
more than 10 years [31]. However, post-radiation cutaneous 55-year-old female who developed “malignant fibrous his-
angiosarcoma could have a shorter latency period even less tiocytoma” after receiving silicone injection augmentation
than 4 years. To be classified as radiotherapy-induced sar- mammoplasty 19 years ago [36]. The second case was a
coma, the proposed criteria are as follows: The initial tumor 69-year-old woman who was diagnosed with a high-grade
angiosarcoma after 35 years of silicone implantation [37]. frequently diagnosed in this location. It is commonly associ-
The third case was a 35-year-old-woman with Li-Fraumeni ated with the skin and infiltrating into the breast parenchyma.
syndrome who had bilateral mastectomy and implantation Atypical vascular lesion is another type of vascular prolifera-
for the treatment of her right breast triple-positive invasive tion associated with radiation exposure. It is considered by
ductal carcinoma. She developed bilateral sarcomas (not oth- many as a precursor lesion of secondary angiosarcoma, even
erwise specified) a year and half after the insertion of the though it may not be supported by others at this stage.
implants [38].
The speculation on the difference in propensity toward 24. How does one differentiate hemangioma from angio-
sarcomas with silicone implantation between humans and sarcoma in the breast?
rodents could be because of their contrasting genetic stabili-
ties [39]. Much evidence exists that rodents’ cells are less Hemangioma in general is demarcated vascular prolifera-
efficient in DNA repair and maintenance of DNA methylation, with a lobular architecture. Anastomosis could be pres-
tion. Therefore, it is understandable that the risk of malig- ent, but not prominent. It is usually lined by a single layer of
nant transformation could be much increased in the patients endothelial cells with occasional cell tufts, but without cyto-
with genetic instability syndromes, such as Li-Fraumeni logical atypia or mitosis. The Ki67 labeling index should be
syndrome. very low. In contrast, angiosarcoma has an infiltrative growth
pattern and is usually not lobulated. It could consist of well-
23. What are the vascular lesions in the breast? formed vascular channels with prominent anastomosis (mor-
phologically well differentiated) or be with solid growth
As in other anatomy sites, breast vascular lesions have pattern (morphologically poorly differentiated). The lining
benign vs. malignant versions. Benign vascular lesions endothelial cells are multi-layered, with cell tufts and pleo-
include different types of hemangiomas, namely, capillary, morphism, especially hyperchromasia. Mitosis is present,
cavernous, and venous hemangiomas, and angiomatosis. and Ki67 labeling index is high (Figs. 9.16a, b and 9.17a, b).
Perilobular hemangioma is a type of capillary hemangioma The most problematic differential diagnosis is between the
unique to the breast stroma. Even though named as perilobu- atypical vascular lesion and the angiosarcoma with well-
lar, it is not limited to this location and could be intralobular differentiated morphology. They are both post-radiation vas-
as well. Perilobular hemangioma is usually an incidental cular lesions with certain degrees of atypia. However, the
finding, identified microscopically in the specimens excised vascular channel for atypical vascular lesion is usually sim-
for other reasons. Other types of benign hemangiomas are pler, has no apparent anastomosis, is typically more superfi-
usually bigger in sizes and could be identified by palpation cial, and is not dissecting through dermis to subcutaneous soft
or mammogram. tissue, with only mild cellular pleomorphism and no mitosis;
Primary angiosarcoma in the breast is uncommon. It is and Ki67 labeling index is low. In recent years, immunohisto-
usually in the breast parenchyma of younger patients, chemical stain for MYC has been applied to this distinction,
whereas radiation-induced secondary angiosarcoma is more and the result seems to be convincing. The atypical vascular
a b
Fig. 9.16 (a) Well-demarcated hemangioma with lobulation formed by fibrous septa (40×). (b) Angiosarcoma with an infiltrative pattern (100×)
a b
Fig. 9.17 (a) The so-called well-differentiated angiosarcoma with well-formed lumens (100×). (b) The so-called poorly differentiated angiosar-
coma with solid growth pattern (100×)
lesion is usually MYC negative, while radiation- induced encapsulated lesion with clear circumference. No breast ter-
angiosarcoma is positive (Fig. 9.18a–c) [40–42]. minal ductal-lobular units should be identified in the tumor
Another differential diagnosis which could potentially be mass. The capillaries are more prominent at periphery. The
a problem is the angiomatosis vs. breast primary endothelial cells are bland and with no mitosis. Fibrin
angiosarcoma. Other than the general features mentioned thrombi are usually identified in the capillary lumen
above for benign and malignant vascular lesions, the vascu- (Fig. 9.20a, b).
lature in angiomatosis is, even though infiltrative, regularly
and evenly distributed throughout. It could replace the glan- 26. What is the definition of primary breast lymphoma
dular units of the breast parenchyma, but will not dissect or (PBL)?
destroy them, whereas angiosarcoma is irregularly infiltrat-
ing the breast parenchyma and could dissect though the ter- The definition of primary breast lymphoma (PBL) is
minal ductal-lobular units. somewhat controversial, as lymphomas by definition are
systemic diseases. However, owing to several distinct fea-
25. What are the other mimics of angiosarcoma in the tures and the predilection of certain lymphoma to present
breast? in the breast, much study has been devoted to the spec-
trum of lymphoma that can present primarily as breast
Pseudoangiomatous stromal hyperplasia (PASH) is a myofi- disease [43–54].
broblast proliferation of the breast stroma. It forms slit-like
spaces in a hyalinized stroma, with spindle cells lining the • The working definition(s) of PBL was proposed by
spaces, sometimes mimicking a structural variant of angio- Wiseman and Liao in 1972 and updated by Hugh and col-
sarcoma. Architecturally, even though PASH appears infil- leagues in 1990.
trative and could involve intralobular stroma, it does not • It includes a variety of lymphoma histologies that occur
disturb the epithelium in the lobules. The space is empty, and within the breast parenchyma or in close proximity to
the anastomoses are not as complex as angiosarcoma. The breast tissue.
lining spindle cells are usually discontinuous, flat, and bland. • Requires no antecedent diagnosis of lymphoma.
In the so-called atypical PASH, the cells could have some • Limited to cases with no extramammary disease other
atypia and even rare mitosis, but immunohistochemical stain than ipsilateral axillary lymph node involvement (although
will show that they are negative for endothelial markers some studies allow a slightly broader definition to include
(Fig. 9.19). regional lymph nodes, i.e., internal mammary and
Cellular angiolipoma could be another mimic of the supraclavicular).
angiosarcoma variant, when the angiosarcoma is composed • It comprises 0.5% of breast malignancies, ~1% of all non-
predominantly of spindle cells with a diffuse infiltrating pat- Hodgkin lymphomas (NHLs), and <3% of extranodal
tern, especially on a core biopsy. Cellular angiolipoma is an lymphomas.
b Fig. 9.19 Pseudoangiomatous stromal hyperplasia with bland spindle

cells lining the spaces (100×)
Fig. 9.18 (a) Post-radiation atypical vascular proliferation (400×). (b)

Post-radiation angiosarcoma with hyperchromatic cells lining the vas-
cular channels (400×). (c) Immunohistochemical stain for MYC in
post-radiation angiosarcoma (400×)
Fig. 9.20 (a) Cellular angiolipoma with bland endothelial cells (100×).

(b) Fibrin thrombi in the capillary lumen in angiolipoma (400×)
• Secondary involvement of the breast by systemic lym- Table 9.1 Benign lymphoid proliferations in the breast
phoma appears to be more common than primary disease. Condition Histologic finding
Lymphoma is thought to be the most common malignancy Lymphocytic B cell-predominant perivasculitis, lobulitis,
to involve the breast secondarily. mastitis/diabetic and ductitis (including lymphoepithelial
mastopathy lesions), surrounded by dense keloid-like
(auto-immune stromal fibrosis
27. How does lymphoma clinically present in the breast? associated)
Cutaneous lymphoid Cutaneous proliferation of B cells
Presentation of PBL can be similar to other more common hyperplasia organized as coalescing follicles with
primary breast tumors, including carcinoma. A constellation non-polarized germinal centers lacking
mantle zones and infiltrates of lymphoid
of signs and symptoms may be associated. Importantly, con- cells spreading into collagen, smooth
stitutional (a.k.a. “B-type”) symptoms (i.e., fever, night muscle, vessel walls, and nerve sheaths
sweats, fatigue, and weight loss) are relatively uncommon in Intramammary Intact nodal architecture with patent sinuses
PBL, being reported in only 4% of patients and more often lymph node with and B cell-rich polarized germinal centers,
reactive follicular containing tangible body macrophages and
occurring in those with breast involvement by disseminated hyperplasia surrounding polarized mantle zones
disease [54–60].
• Age distribution: The median age at presentation ranges

from 60 to 65 years in Western countries (reportedly 29. What types of benign proliferations or inflammatory
lower in East Asian countries, at 45–53 years), with a lesions can mimic lymphoma in the breast?
wide age distribution (teens thru 90s).
• Gender: Women are much more commonly affected than A few specific benign lymphoid proliferations or inflamma-
men. Cases among males have been anecdotally reported. tory breast lesions can mimic lymphoma in the breast, as well
• Bilateral breast involvement occurs in 5–11% of cases as generalized reactive lymph node hyperplasia of intramam-
and may occur more often during pregnancy or postpar- mary lymph nodes. The primary histologic differential is with
tum. Along with the strong female predilection, this pat- low-grade B cell NHL, as these malignant tumor cells most
tern suggests that hormone stimulation may promote closely resemble benign lymphocytes [64–69]. See Table 9.1.
tumor growth.
• Symptoms: Most PBLs present as painless masses 30. What are the most common types of primary lym-
(median tumor diameter 4 cm, up to 20 cm reported). phoma in the breast?
Right > left, for unknown reasons. Skin changes (nipple
retraction, color change, or discharge) are rare and more The most common form of PBL is diffuse large B cell lym-
likely to be found with T cell lymphoma. phoma (DLBCL), followed by marginal zone lymphoma
• Radiologic imaging. (See question 28.) (MZL), follicular lymphoma (FL), and Burkitt lymphoma
(BL). DLBCL presenting as PBL is not a distinct pathology
entity according to current WHO classification schemes
28. What radiographic features are most typical of lym- (unlike other anatomic DLBCL variants such as primary
phoma in the breast? cutaneous, CNS, intravascular, etc.) [48, 49, 54, 61, 70–74].
Table 9.2 outlines the relative proportion and salient his-
Up to 20% of patients with PBLs are detected by screening tologic and immunophenotypic features of the most common
mammography [61–63]. type of PBL. For more details, see additional sections on
each distinct entity (questions 34–38).
• Features that are more common in lymphoma than carci-
noma include larger tumor size and lack of the following: 31. What types of lymphoma occur more rarely, but are
spiculated appearance, calcification, or architectural tis- also known to occur as primary within the breast?
sue distortion.
• Typically solitary lesions, oval to round, circumscribed, The following entities each comprise ≤1% of lymphomas
or without distinct delineation. presenting within the breast: anaplastic large cell lymphoma
• Ultrasound may be more informative than mammography (ALCL), peripheral T cell lymphoma (PTCL), small lym-
for detecting infiltrating or small lymphoma tumors, phocytic lymphoma/chronic lymphocytic leukemia (SLL/
which appear as hypoechoic solid masses. CLL), lymphoplasmacytic lymphoma (LPL), mantle cell
• No radiographic features are definitively diagnostic, and lymphoma (MCL), and classic Hodgkin lymphoma (CHL).
biopsy is required. [48, 49, 54, 71, 73, 75] See Table 9.3.
Table 9.2 Features of common histologic subtypes of primary breast lymphoma

DLBCL MZL FL BL
PBL (%) 56–84% 9–28% 10–19% <6%
Cell size/ Large/pleomorphic, Small, mature Predominantly small with mature Intermediate in size,
morphology centroblastic, or monocytoid or cleaved nuclei (low grade), centroblastic high nuclear grade
immunoblastic features plasmacytoid (high grade)
CD20 + + + +
CD10 Variable − + +
BCL6 Variable − + +
BCL2 +/− + + −
Table 9.3 Immunohistochemical features of rare breast lymphoma subtypes

ALCL PTCL SLL/CLL LPL MCL CHL
Cell size/morphology
CD3 −/+ + − − − −
CD20 − − + + + −
CD30 ++ +/− − − − +
CD15 − − − − − +
CD5 − +/− + − + −
CD23 − − + − − −
Additional ALK +/− CD138+ Cyclin-D1+
32. Which types of lymphoma are associated with breast with median survival of 12 years. Treatments have often
implants? employed excision alone, and chemotherapy does not appear
to affect outcome in cases without tumoral involvement.
No definitive increase in risk of lymphoma among women with Presentation as a solid tumor mass appears to be the worst
breast implants has been found among several large epidemio- prognostic factor.
logic studies. Regardless, a distinct form of breast implant- Even more rarely, occasional and anecdotal case reports
associated anaplastic large cell lymphoma (BIA-ALCL) does of other NHLs have been reported in association with breast
occur [76–88], and is now considered a provisional entity by implants, including T cell lymphoma, primary effusion lym-
WHO diagnostic criteria. BIA-ALCL is an extremely rare phoma, follicular lymphoma, marginal zone lymphoma, and
malignancy, with an estimated incidence of 1 per 0.5–3 million lymphoplasmacytic lymphoma.
women with implants [77, 82]. While the etiology and patho-
genesis is unknown, an etiologic association between implants 33. How does one distinguish low-grade from various
and BIA-ALCL has at least been suggested. high-grade B cell lymphomas (HGBLs), and what is
Most patients present with involvement of a breast the clinical significance?
implant-associated seroma and infrequently with a breast
mass. Pathologic evidence of lymphoma may be limited to Low-grade B cell NHLs are generally indolent diseases.
the seroma fluid, or histologic evidence of infiltration within They do not constitute medical emergencies and are rou-
the pericapsular fibrotic tissue may also be seen if capsulec- tinely treated in the outpatient setting.
tomy is performed. Regional lymph nodes may also be
involved. • The major categories include FL, MZL, SLL/CLL, and
Morphologically, BIA-ALCL cells are typical of other LPL, with the first two being the only ones encountered
anaplastic lymphomas, with large pleomorphic features and frequently as primary diagnoses within the breast [90].
abundant cytoplasm, and usually include prototypical “hall- (See question 30.)
mark cells” (with horseshoe-shaped, reniform, or ring-like • Management of low-grade B cell NHL is highly variable
nuclear contours). Tumor cells express strong CD30, are and may include expectant “watch and wait” or active sur-
negative for ALK, and exhibit variable T cell antigen expres- veillance approach in low-stage non-symptomatic patients
sion [76, 84, 86]. or alternatively combined modalities including single-
A wide time interval has been reported between implant agent or combined chemotherapy, antibody treatment, or
placement and lymphoma diagnosis, averaging about radiation. While active surveillance is still indicated in
10 years. Clinical outcome has also generally been good, certain early-stage diseases among some asymptomatic
patients, more sophisticated clinical and genetic risk strat- Table 9.4 Histologic features distinguishing between low-grade and
high-grade B cell non-Hodgkin lymphoma
ification schemes (and the increasing availability of tar-
geted small molecular inhibitors) are changing treatment Low grade High grade
paradigms. Cell size Small with scant Medium to large, with
cytoplasm variable cytoplasm
Nuclear Round to slightly Round to highly
High-grade B cell NHLs present with variable disease contours irregular, may have pleomorphic, with
severity, from incidental detection to severely ill patients in cleaved or angulated prominent nuclear
near-critical condition. Although PBL is less likely to pres- contours membrane
ent as severe illness (as by definition unilateral disease is Chromatin Condensed, Vesicular, open, or finely
hyperchromatic, either dispersed chromatin,
early Ann Arbor stage IE [breast alone] or IIE [breast plus lacking nucleoli or with prominent nucleoli
ipsilateral lymph nodes]), some presentations of high-grade with small indistinct (single or multiple)
lymphoma constitute medical emergencies and/or require forms
immediate medical attention with inpatient management. Architecture Variable architecture, Diffuse replacement,
and associated from nodular to with abundant mitoses,
Therefore, such diagnoses should be communicated immedi-
findings sheet-like, depending apoptosis, and sometime
ately to providing clinicians according to guidelines for on subtype “starry sky” appearance
reporting urgent medical information or critical values [89]. (i.e., tangible body
macrophages)
• BL and particularly aggressive subtypes of DLBCL now Ki67 index Low, typically High, 90–100% in BL
10–20%, should not and ALL, 80–100% in
identified as “high-grade B cell lymphoma” (HGBL) may exceed 50–60% HGBL, >50% in
constitute medical emergencies when patients have con- DLBCL-NOS
comitant metabolic derangement (e.g., tumor lysis syn-
drome with incipient acute renal failure, attributable to
the rapid turnover and apoptosis of highly proliferative • Subclassification: When diagnosed in the breast, DLBCL-
tumor cells). (See question 34.) NOS should be subclassified similar to other nodal or
• Standard chemotherapy for BL (CODOX-M/IVAC) typi- extranodal presentations. The first subcategories of
cally necessitates inpatient admission. By comparison, DLBCL-NOS that have penetrated clinical practice are
standard therapy for DLBCL (R-CHOP) is routinely based on gene expression profiling studies of primary
administered outpatient, while more intensive regimens lymphomas. The so-called “cell-of-origin” (COO) classi-
that have been recently studied (e.g., R-EPOCH) may fication was developed by using gene expression profile
also necessitate inpatient admission for a subset of (GEP) patterns that are typical of benign B cells and
patients. applying them to malignant lymphoma. Among the pat-
• Although not counted among the typical forms of PBL, T terns that emerged were two groups which most closely
cell or B cell acute lymphoblastic leukemia/lymphoma resembled germinal center B cells (GCBs) and post-
(T- or B-ALL) is a medical urgency/emergency that germinal center B cells that are transitioning toward
necessitates inpatient admission for standard treatment memory or plasma cells. DLBCLs with GEPs that more
(Hyper-CVAD). See Table 9.4. closely match either of these two patterns have come to be
known as germinal center B cell (GCB) or activated B cell
34. How does one distinguish and subclassify diffuse (ABC) type, respectively. Then, a simplified algorithm of
large B cell lymphoma [91–95]? immunohistochemical stains was proposed by Hans and
colleagues in 2004 [95] to more readily classify GCB vs.
• Morphology: DLBCL-NOS (not otherwise specified) are ABC (a.k.a. non-germinal center type) DLBCL using
morphologically diverse tumors. Distinguishing features routine formalin-fixed paraffin-embedded tissue. This
include diffuse sheet-like infiltration of tissue by large- COO classification scheme (summarized below), which
sized cells (2–3 times the size of normal lymphocytes). corresponds to GEP patterns, is now routinely reported
Cytoplasm is moderate to abundant, with relatively low for de novo DLBCL, as recommended by the updated
nuclear-to-cytoplasmic ratio for lymphoma, and variably WHO classification system in 2016.
amphophilic to basophilic, with sometimes even clear • CD10 and BCL6 are both germinal center B cell markers,
appearance. Nuclei have irregular contours, with course- but BCL6 positivity alone is insufficient to qualify as
to-vesicular chromatin, and are prominent. The so-called GCB-type DLBCL, so MUM1 is needed to discriminate.
centroblastic (multiple nucleoli) vs. immunoblastic (sin- See Table 9.5.
gle central nucleoli) variants are commonly described.
Anaplastic variants, and even those with myxoid or fibril- Apart from the COO classification, routine evaluation of
lary stroma, have rarely been reported. additional cytogenetic and morphologic features is now indi-
Table 9.5 Cell-of-origin classification for DLBCL by Han’s algorithm tin is relatively fine or immature, and distinct nucleoli are
CD10 BCL6 MUM1 COO subtype typically multiple while small to intermediate in size.
+ n/a n/a GCB This tumor exhibits a characteristic “starry sky” appear-
− + − GCB ance from low power, owing to a very high proliferation
− + + ABC rate (≥95% as determined by Ki67 staining) and frequent
− − n/a ABC interspersed apoptotic cells engulfed by tingible body
n/a not applicable macrophages. Frequent mitoses are also present.
Data from Ref. [95] • Immunophenotype: BL cells are mature B cells, positive
for CD20, CD79a, and other markers routinely evaluated
cated to determine other risk factors, particularly those chro- by flow cytometry (e.g., CD19, CD22, with monotypic
mosomal rearrangements which identify an aggressive surface immunoglobulin light-chain expression, the latter
subtype of B cell NLH, commonly referred to as double- or being an important feature which helps distinguish it from
triple-hit lymphoma. These tumors are defined by the WHO the majority of B-ALL). Additional markers are similar to
as high-grade B cell lymphoma (HGBL) with MYC, BCL2, germinal center B cells, including CD10 and BCL6; but
and/or BCL6 rearrangements or HGBL-NOS when the mor- importantly BCL2 is negative in the vast majority of
phology is suggestive (i.e., high nuclear grade, appropriate cases, a feature which distinguishes BL from many
cytomorphology and architectural features, and nearly 100% DLBCLs or other high-grade B cell NHLs. TdT is also
Ki67 proliferation index), in the absence of these genetic negative. Epstein-Barr virus (EBV), best evaluated by
findings. EBV-encoded RNA (EBER) in situ hybridization, is only
seen in 20% of sporadic BL in Western countries and in
• FISH for the following markers is indicated: 30–40% of HIV-associated cases, although the vast
–– MYC (8q24) break apart: Alternatively MYC-IGH majority (98%) are positive in endemic cases in sub-
fusion, t(8;14), may be considered, but is less sensitive Saharan Africa.
owing to alternative MYC gene translocation/rear- • Genetics: Virtually all cases of BL have a translocation
rangements (e.g., kappa or lambda light chain on chro- between the long arm of chromosome 8, involving the
mosome 2 or 22, respectively). protooncogene MYC (8q24), and one of the three immu-
–– IGH-BCL2 fusion, t(14;18): Alternatively BCL2 noglobulin translocation partners. Translocation t(14;18),
(18q21) break apart may be used, but is less clinically involving the immunoglobulin heavy-chain (IGH) region
widely available and considered less susceptible to on chromosome 14, is the most common and can rou-
false negative due to lack of alternative common trans- tinely be tested by FISH for IGH-MYC fusion. Much less
location partners. commonly, the kappa light-chain locus on chromosome 2,
–– BCL6 (3q27) break apart or the lambda light-chain locus on chromosome 22, may
• IHC evaluation for MYC and BCL2 is also recom- alternatively be involved; and FISH using the so-called
mended: Some studies have suggested an independently MYC “break apart” probes can be employed to confirm a
poor prognosis for DLBCL co-expressing MYC and MYC gene rearrangement. Importantly, however, MYC
BCL2 protein (a.k.a. “double expressers”). gene rearrangement alone should not be used as a surro-
–– MYC IHC (≥40% scored as positive) gate for BL, as up to 10% of other aggressive high-grade
–– BCL2 IHC (≥70% scored as positive) B cell NHLs can harbor a conventional or non-
conventional MYC aberration. (See question 39.)
35. How does one distinguish Burkitt lymphoma?
Precise recognition of BL is critical due to the aggressive 36. What is the association between Burkitt lymphoma
nature of the disease, the often high clinical acuity of the and bilateral breast masses?
patients, and the high curability rate when appropriately
treated with distinct chemotherapy in advanced health-care Presentation of Burkitt lymphoma (BL) with bilateral breast
settings. involvement, often resulting in massive breast enlargement,
appears to be particularly associated with pregnancy and/or
• Morphology: BL is characterized by diffuse infiltration of lactation. (See also question 27.) Bilateral involvement is not
monomorphic medium-sized cells with a high nuclear-to- exclusive to BL; but since the 1960s, numerous case reports
cytoplasmic ratio, round nuclei, multiple nucleoli, and have described similar findings in women of child-bearing
basophilic cytoplasm that is often vacuolated. Nuclear age, usually occurring during pregnancy, postpartum, or in
contours are round to oval without cleaves or folds, a key association with lactation, leading to speculation that BL is a
feature in the distinction from DLBCL. Nuclear chroma- particularly hormonally responsive tumor, at least under
these conditions. In the setting of high-intensity chemother- CD10; thus, the presence of either of these should not be
apy, outcomes appear similar as reported with conventional used alone to exclude or overrule a diagnosis of MZL.
sporadic BL [96–104].
38. How does one distinguish follicular lymphoma?
37. How does one distinguish extranodal marginal zone
lymphoma? Follicular lymphoma (FL) is the most common indolent lym-
phoma and second most common B cell NHL outside of the
Extranodal marginal zone lymphoma (MZL) (a.k.a. mucosa- breast [90, 105]. It has a highly variable cytologic appear-
associated lymphoid tissue (MALT) lymphoma) is a low- ance, which is reflected in different histologic grades and
grade B cell NHL with distinct, but sometimes subtle, which partly explains the highly variable clinical outcome.
morphologic characteristics [90]. While the breast is a rela- FL has a natural rate of progression, with about 3% per year
tively uncommon site for organ of involvement by MZL/ undergoing “transformation” to DLBCL.
MALT lymphoma (only ~3% of total, after gastric, ocular,
pulmonary, cutaneous, and salivary gland), it is the second • Morphology: FL is defined by its nodular and follicular
most common subtype of all PBLs (after DLBCL). (See also architecture, in which neoplastic FL cells are intimately
question 34.) Although chronic inflammation, infection, and (but not exclusively) associated with intact follicular
antigenic stimulation have been tightly linked to MZL/ dendritic cell (FDC) networks, analogous to the architec-
MALT lymphoma pathogenesis at other sites (e.g., H. pylori ture exhibited by benign germinal centers. By compari-
in the stomach, Chlamydia psittaci in the eye, Borrelia burg- son, however, the FDC structures in FL are often
dorferi in the skin, Sjogren syndrome in the salivary gland, enlarged, expanded, and organized into a back-to-back
and Hashimoto disease in the thyroid), no such association pattern. Critical features which distinguish these neo-
has yet been identified in the breast. plastic follicles from benign follicles (or reactive follicu-
lar hyperplasia) are the lack of polarized mantle zones
• Morphology: MZL cells may resemble small lympho- surrounding the follicle, the lack of polarized light and
cytes, with plasmacytic differentiation in up to one-third dark zones within the FDC network, and the absence of
of cases. The lymphocytes are small-to-medium-sized tingible body macrophages. The neoplastic B cells con-
cells, with only slightly irregular nuclei, condensed chro- sist of two varieties: smaller centrocytes, with scant cyto-
matin, and indistinct nucleoli. Slightly more abundant plasm, irregular-to-cleaved nuclei, condensed chromatin,
pale cytoplasm may be seen, responsible for the so-called and indistinct nucleoli, and larger centroblasts with
“monocytoid” appearance of these cells. Large centro- round-to-oval nuclei (i.e., not cleaved), vesicular chro-
blast- or immunoblast-type cells are frequently present, matin, distinct nucleoli, and moderate amounts of cyto-
but should be scattered. Plasma cells may also be rare to plasm. Distinguishing grade 1 thru grade 3 FL is based
scattered, but rarely predominant so as to mimic plasma- on the absolute number of centroblasts per high-power
cytoma. Amyloid deposition may also be seen. Tissue field (HPF) averaged across ten neoplastic follicles.
infiltration may be diffuse and effacing or infiltrative Although clinically grade 1 FL and grade 2 FL are cur-
around existing structure and benign lymphoid follicles. rently treated almost identically, the following criteria
Colonization of intact lymphoid follicles by neoplastic are still routinely applied on morphologic evaluation:
MZL B cells may produce a subtle and misleading nodu- Grade 1 is defined as 0-5 centroblasts/HPF; grade 2 is
lar or follicular architecture in some cases. The so-called 6-15 centroblasts/HPF, and grade 3 is >15 centroblasts/
lymphoepithelial lesion, in which MZL cells infiltrate, HPF. Importantly, grade 3 is further subdivided based on
distort, and/or destroy local epithelial structures, is a help- the relative distribution of small centrocytes and large
ful diagnostic clue. centroblasts, with grade 3A containing an admixture of
• Immunophenotype: MZL expresses mature B cell anti- each and grade 3B consisting entirely of large centro-
gens CD19, CD20, and CD79a, with surface kappa or blast-like cells confined by FDC networks. Any distinct
lambda light-chain restriction (reliable among lympho- areas of diffuse large cell lymphoma should be separately
cytes usually only by flow cytometry) while typically diagnosed as concurrent DLBCL.
being negative for CD5, CD10, and BCL6. When tissue • Immunophenotype: FL is positive for mature B cell mark-
biopsy alone is available (i.e., no flow), in situ hybridiza- ers CD19 and CD20, with co-expression of germinal cen-
tion for immunoglobulin light-chain restriction (i.e., kappa ter markers CD10 and BCL6 and an aberrant expression
or lambda) among plasma cells may provide definitive evi- of BCL2 (attributable to oncogenic prototypical immuno-
dence of B cell clonality (capable of substituting for flow globulin heavy-chain (IGH)-BCL2 translocation).
or B cell gene rearrangements in some cases). Infrequent Negative markers include CD5 and typically CD43. Of
cases are CD5 positive and extremely rarely express note, in grade 3 FL, CD10 is occasionally lost. Staining
for FDC markers (e.g., CD21 and CD23) may aid in doc- immunohistochemical profile. CD45 may be used to
umenting follicular architecture in cases which are not screen for hematopoietic malignancies, and reliable B
overtly nodular. lymphoid markers that should be used include CD20 and
• Genetics: For routine clinical testing, confirmation of the CD79a. Rarely, poorly differentiated DLBCL may be
t(14;18)(q32;q21) translocation involving IGH-BCL2 negative for one or more of these, and secondary B cell-
(most commonly by employing FISH) is recommended specific markers (e.g., Pax5/BSAP, OCT-2, or BOB.1)
when differentiating among other low-grade B cell NHLs. must be employed to prove B cell histogenesis. Conversely,
IGH-BCL2 translocation cannot be used to distinguish FL IDC should be confirmed by broad-spectrum cytokeratin
form DLBCL. Furthermore, up to 10% of FL may lack immunoreactivity, and markers of breast tissue origin
this translocation, so negative FISH studies or lack of may be used for confirmation (e.g., mammaglobin,
BCL2 expression does not preclude the diagnosis if GCDFP-15, or GATA3). Importantly, ER is expressed in
definitive morphologic features are present. a subset of B cell lymphoma and should not be used to
distinguish this differential.
39. How does one distinguish lymphoma from triple-
negative breast carcinoma? 40. How does one distinguish lymphoma (particularly
anaplastic large cell lymphoma) from medullary
Mistaking PBL for invasive breast carcinoma is a critical pit- (anaplastic) breast carcinoma (MBC)?
fall to avoid. Given the prevalence of DLBCL within the
breast, this type of PBL may be most likely to mimic poorly Anaplastic large cell lymphoma (ALCL) is T cell lymphoma
differentiated, high nuclear grade, invasive ductal carcinoma that occurs in various forms, including systemic disease, pri-
(IDC). Most other varieties of B cell NHL have a more dis- mary cutaneous lymphoma, and a rare form of primary breast
tinctive and normal lymphoid appearance and are less likely implant-associated lymphoma. (See question 32.) Similar to
to be mistaken for breast carcinoma, with the notable excep- other anaplastic malignancies, poor histologic differentiation
tion of anaplastic large cell lymphoma (ALCL). (See ques- and limited antigen expression profile can make it difficult to
tion 40.) The pathologist’s impression of an aggressive IDC determine ALCL histogenesis.
may be reinforced by triple-negative staining for estrogen Of all the variants of epithelial breast cancer, medullary
receptor (ER), progesterone receptor (PR), and HER2 in (anaplastic) breast carcinoma (MBC) shares the most over-
PBL. Importantly, treatment modalities are entirely different lapping features with lymphoma.
for the two diseases. Routine surgical excision or mastec-
tomy is not indicated for PBL and may even be detrimental. • Morphology: Shared features of MBC and ALCL which
(See question 43.) may contribute to misdiagnosis include presentation as
well-circumscribed tumor mass, diffuse growth pattern,
• Morphology: As described in question 34, DLBCL is a lack of glandular differentiation, lack of mucin produc-
heterogeneous disease that may morphologically mimic tion, and absence of associated ductal carcinoma in situ
other primary cancers. These malignant lymphoma cells (DCIS). Cytologic features include numerous mitoses,
are large in cell size for lymphocytes (averaging 17–20 pleomorphic nuclei, large nucleoli, indistinct cell borders,
microns in diameter, 2–4 times the size of normal cells), and prominent intratumoral lymphoid; or lymphoplasma-
but in many cases, these are not substantially larger than cytic infiltration may be seen. Even bizarre features such
poorly differentiated carcinoma cells. Nuclear features as spindle cell morphology and pleomorphic giant cell
in DLBCL are also highly variable, with chromatin pat- formation can be seen in both.
tern and nucleoli that may mimic high-nuclear-grade • Immunophenotype: As with distinguishing DLBCL in the
breast carcinoma. Furthermore, cytoplasm may be abun- breast, immunohistochemical markers are frequently
dant, imparting an epithelioid appearance. Mitoses are essential to distinguish ALCL from MBC (Table 9.6).
also often readily identified. Histologic evidence of (See question 39.) Importantly, CD45 is variable and may
mucin or positive mucicarmine staining can help exclude be negative in ALCL. Thus, CD45 cannot be used alone to
DLBCL. Since the many cytologic features of DLBCL completely exclude lymphoma. Not only does ALCL lack
and IDC can be overlapping, often the most important B cell markers (i.e., CD20, CD79a), but up to 70% are
morphologic features can be found in tissue architec- also negative for CD3. Other pan-T cell markers (CD2,
ture. While DLBCL may infiltrate in single-cell or even CD5, and CD4) are more useful. CD30 is strongly and
nested patterns, distinct glandular architecture is never diffusely positive. CD43 and CD25 are also usually posi-
present. tive. Both ALK-positive and ALK-negative variants of
• Immunohistochemistry: In challenging cases, the most ALCL exist, so absence of ALK staining does not rule out
important distinction between lymphoma and IDC is the the diagnosis either.
Table 9.6 Immunohistochemical markers to differentiate poorly dif- Table 9.7 Immunohistochemical markers to distinguish myeloid and
ferentiated lymphoma from medullar breast carcinoma monocytic differentiation from B/T cell NHL
Marker DLBCL ALCL MBC Marker MS HS B-NHL T-NHL
CD45 + +/− − CD45 + (weak) +/− + +
CD20 + − − CD34 +/− − − −
CD79a + − − CD117 +/− − − −
CD3 − −/+ − MPO + − − −
ALK − +/− − Lysozyme −/+ +/− − −
CD30 − ++ − CD68 − + − −
CD43 −/+ +/− − CD163 − + − −
CD2/CD5 − +/− − CD20/CD79a − − + −
Cytokeratin − − + CD3 − − − +
ER −/+ − − CD2/CD5/CD7 − − − +/−
PR − − −
HER2 − − −
Nuclear chromatin appears finer than is typical of mature lym-
phoid malignancies, but sometimes vesicular chromatin pat-
41. What other hematopoietic malignancies besides lym- terns following formalin fixation obscure the characteristic
phoma can occur in the breast? nuclear details of these malignant blasts [113–115].
Histiocytic sarcoma (HS) is an often more well-
Although exceedingly rare, initial presentation of hemato- differentiated, but highly pleomorphic malignancy that
logic malignancies other than lymphoma can occur in the exhibits variable degrees of monocytic/histiocytic matura-
breast. Myeloid sarcoma (MS) (a.k.a. granulocytic sarcoma tion [108]. Cases range from overtly aggressive tumors with
or chloroma) refers to the extramedullary (i.e., outside of the sarcomatoid features, including abundant cytoplasm, indis-
bone marrow) presentation of acute myeloid leukemia tinct cell borders, round nuclei, and open chromatin, as well
(AML) and is the most commonly reported non-lymphoid as histiocytic giant cell formation, to more immature and
hematopoietic malignancy to involve the breast. Such “aleu- monotonous monocytoid tumors with smaller cells and more
kemic” presentations of AML can occur anywhere in the condensed nuclear chromatin.
body and do not appear to have any particular predilection Immunohistochemical (or cytochemical) staining is critical
for breast tissue. Genetic features, cellular phenotypes, and to demonstrate myeloid or monocytic differentiation.
morphology are all comparable to the bone marrow or circu- Myeloperoxidase (MPO) is a defining feature of myeloid malig-
lating forms of the disease [106–112]. nancies with the exception of monocytic tumors which more
Other hematopoietic tumors of non-lymphoid histogene- often display specific cell surface markers, as shown in Table 9.7.
sis that have sporadically been noted to occur primarily or
secondarily within the breast include histiocytic sarcoma 43. What is the role of mastectomy in primary breast
(HS) and Langerhans cell histiocytosis (LCH) [109]. lymphoma?
T or B cell acute lymphoblastic leukemia (T-/B-ALL) can
also present primarily or secondarily within the breast, with Mastectomy offers no benefit in the treatment of PBL; it pro-
or without peripheral blood involvement. Although these are vides neither survival benefit nor protection from recurrence. In
invariably systemic leukemias, current classification schemes fact, data suggest that surgical resection results in inferior local
call for designation as T or B lymphoblastic lymphoma when disease control and is associated with greater disease-specific
presenting as a solid tumor mass. and overall mortality. Accordingly, mastectomy or surgical
intervention is strongly contraindicated [46, 70, 71, 75, 116].
42. How does one distinguish myeloid or histiocytic sar- In instances where surgical excision does occur, attribut-
coma from lymphoma? able to misdiagnosis of breast carcinoma or other reasons,
chemo-immunotherapy and radiotherapy should be adminis-
The morphologic distinction of a myeloid sarcoma (MS) from tered as soon as adequate wound healing has occurred. (See
poorly differentiated lymphoma can be challenging. Many of questions 39 and 40.)
the diagnostic features of malignant “blasts” that distinguish
AML cytologically cannot be easily appreciated on the histo- 44. Is there a role for axillary/sentinel lymph node biopsy
logic section. Myeloid sarcoma/AML cells are intermediate to in evaluating breast lymphoma?
large in size, with round-to-irregular nuclei and moderate
amounts of cytoplasm, often eosinophilic to amphophilic, Sentinel lymph node biopsy and mapping, in the traditional
reflective of their oftentimes granular cytoplasmic content. surgical sense (utilizing radioactive colloid and/or blue dye
injection at the site of primary breast disease), is not indi- displays distinct extranodal tropism both at presentation
cated. Although some studies indicate that adjacent lymph and recurrence.
node involvement is a strong predictor of survival, this Recurrence in the ipsilateral breast presents typically
result has not been universally reproduced, and other con- within 3–5 years, but contralateral breast relapse can be
ventional risk stratification schemes are typically employed much delayed, occurring up to 13 years after initial presenta-
(e.g., Ann Arbor staging, International Prognostic Index tion. The frequency of CNS relapse (a high-risk feature with
(IPI), and Eastern Cooperative Oncology Group (ECOG) poor prognostic implications) appears to be greater in PBL
performance status) [116–118]. than in comparable limited-stage (IE/IIE) nodal DLBCL, but
As with all systemic lymphomas, adequate clinical staging with similar average time to recurrence (<2 years). Limited
typically involves whole-body positron emission tomography data suggest that outcomes with relapsed PB-DLBCL are
with computed tomography (PET-CT), with confirmatory poor. This is particularly true with CNS involvement, where
biopsy of peripheral lymph nodes performed only as clini- median overall survival averages less than a year.
cally indicated to exclude other causes of lymphadenopathy.
45. What are the most common patterns of relapse in Case Presentations
primary breast lymphoma?
Case 1 (Fig. 9.21a–d)
Relapse rates and patterns in PBL are most widely reported
for DLBCL, as other less common low-grade histologic Learning Objectives
forms (i.e., MZL, FL, etc.) have less extensive data available 1. To be aware of the association of the lesion with breast
and may behave similar to nodal disease [46, 49, 51, 75, implant
119–122]. 2. To be familiar with the morphological characteristics of
Primary breast DLBCL relapses predominantly at extra- this lesion
nodal sites. Recurrence within either the ipsilateral or con- 3. To be able to generate appropriate differential diagnosis
tralateral breast is most common (12–44%). Outside of the for this type of lesion
breast, the central nervous system (CNS) is the next most
common site (5–16%). Other extranodal sites do occur History
(including bone marrow, lung, skin, and gastrointestinal • A 35-year-old female with bilateral breast augmentation
tract), but are much less common; and interestingly these implants for 3 years presents with chest wall mass adja-
are involved more often with PB-DLBCL than is typical of cent to the implant, 2.5 cm in size clinically.
primary nodal disease. In this regard, PB-DLBCL is similar
to some other tissue-specific types of extranodal DLBCL Gross Examination
(e.g., primary testicular or primary CNS disease) in that it • Several cores of soft tissue
a b
Fig. 9.21 Case 1. (a) Low-power view of spindle cell lesion with mod- than endothelial cells (100×). (d) Nuclear positivity for beta-catenin by
est cellularity (20×). (b) Lesional cells which appear directly associated immunohistochemical stain (100×)
with thick fibrous capsule (40×). (c) Bland spindle cells, less prominent
c d
Histologic Findings History

1. Fragments of spindle cell lesion with modest cellularity, • A 19-year-old female patient presents with breast mass. A
tightly associated with a thick fibrous capsule. core biopsy and then subsequent lumpectomy were
2. The cells are less prominent than the endothelial cells; no performed.
mitosis is identified.
Gross Examination
Immunohistochemical Stain • An ill-defined, tan-white, lobulated mass, 2.6 cm in great-
1. Positive: Beta-catenin (nuclear), focally positive for
est dimension
SMA
2. Negative: Cytokeratin cocktail, p63, CK5, desmin, CD34 Histologic Findings
1. Hypercellular spindle cell neoplasm; cells are plump and
Differential Diagnosis hyperchromatic.
1. Metaplastic carcinoma 2. Focal tumor necrosis.
2. Nodular fasciitis 3. Focal cells with pink cytoplasm.
3. Fibromatosis
4. Solitary fibrous tumor Differential Diagnosis
5. Smooth muscle tumor 1. Rhabdomyosarcoma, primary or metastatic
2. Heterologous component of malignant phyllodes tumor
Final Diagnosis 3. Heterologous component of metaplastic carcinoma
• Fibromatosis associated with the capsule of implant 4. Other hypercellular spindle cell sarcomas, primary or met-
astatic: leiomyosarcoma, synovial sarcoma, and MPNST
Take-Home Messages 5. Melanoma
1. Fibromatosis could be associated with breast implant.
2. It is with modest cellularity and bland spindle cells; mito- Immunohistochemical Stain
sis is uncommon. 1. Positive: Myogenin, desmin
3. Beta-Catenin is nuclear positive. 2. Negative: Cytokeratin cocktail, EMA, SMA, S-100, CD34
Case 2 (Fig. 9.22a–d) Diagnosis

• Rhabdomyosarcoma, embryonal type, breast primary
Learning Objectives
1. To be able to generate a list of differential diagnoses for Take-Home Messages
hypercellular spindle cell tumor in the breast 1. Age could be a hint for the diagnosis.
2. To be familiar with the diagnostic characteristics of this 2. It will be very productive to carefully look for rhabdo-
tumor myoblasts under high-power view.
a b
c d
Fig. 9.22 Case 2. (a) Low-power view of hypercellular spindle cell plasm in many tumor cells (strap cells or tadpole cells) (100×). (d)
lesion with trapped-in fat vacuoles and focal tumor necrosis (20×). (b). Higher-power view (400×)
Spindle cells forming vague fascicles (40×). (c) Eccentric pink cyto-
Case 3 (Fig. 9.23a–d) Histologic Findings

1. Scattered vasculature with well-formed lumen, dissecting
Learning Objectives through dermis to subcutaneous soft tissue.
1. To know the morphological characteristics of this vascu- 2. The lining endothelial cells are hyperchromatic with

lar lesion enlarged nuclei.
2. To be able to differentiate it from other reactive vascular
lesions Differential Diagnosis
1. Normal vasculature
Clinical History 2. Post-radiation atypical vascular proliferation
• A 78-year-old female with 9-year history of left breast 3. Post-radiation angiosarcoma
cancer, S/P mastectomy, and radiation now presents with
skin lesion. Immunohistochemical Stain
1. Positive: ERG and MYC, with Ki67 labeling index of
Gross Examination 10% in endothelial cells
• Skin punch biopsy 2. Negative:
a b
c d
Fig. 9.23 Case 3. (a) Skin punch biopsy with scattered vasculature tissue (40×). (d) Immunohistochemical stain for MYC highlighting the
dissecting dermis (40×). (b) Large hyperchromatic cells lining the vas- hyperchromatic nuclei (400×)
cular channel (400×). (c) The same vasculature in subcutaneous soft
Final Diagnosis History

• Post-radiation angiosarcoma • A 69-year-old female with history of hypothyroidism and
Sjogren syndrome is found to have a 2.0 cm indetermi-
Take-Home Messages nate mass in the left upper outer quadrant on routine
1. Nuclear hyperchromasia and invasion to subcutaneous screening mammogram. Mammography report indicates
soft tissue are diagnostic features of this entity. bi-rads 4c, “probable invasive carcinoma.” Core needle
2. Immunohistochemical stain for C-MYC and Ki67 has biopsy was performed.
distinct value in helping the diagnosis.
Histologic Findings
Case 4 (Fig. 9.24a–h) • Infiltration of the sclerotic parenchyma by a diffuse
proliferation of highly atypical cells, moderate to large
Learning Objectives in cell size, with abundant intermixed small mature
1. To generate the differential diagnosis for lymphocyte rich lymphocytes containing hyperchromatic nuclei
lesions of the breast (Fig. 9.24a).
2. To become familiar with the immunohistochemical
• No evidence of ductal or lobular infiltration pattern.
workup to evaluate for lymphoma • Large atypical cells have irregular nuclei with open vesic-
3. To recognize the histologic features of the correct final ular chromatin, prominent centrally placed nucleoli, and
diagnosis
a b
c d
e f
Fig. 9.24 Case 4. (a) Medium-power view of diffuse mixed atypical matin, prominent central nucleoli, and moderate cytoplasm (600×). (c)
inflammatory infiltrate (200×). (b) High-power view of infiltrating cell CD20 (400×). (d) CD79a (400×). (e) CD10 (400×). (f) BCL6 (400×).
populations, with focus on large cells, irregular nuclei, vesicular chro- (g) MUM1 (400×). (h) Ki67 (400×)
g h
moderate amounts of amphophilic cytoplasm with indis- 2. CD20 expression may be weak to negative in

tinct cell borders (Fig. 9.24b). DLBCL. Secondary B cell markers (e.g., CD79a or Pax5)
are often needed.
Differential Diagnosis 3. Large cell size, marked nuclear pleomorphism, and interme-
• Lymphocytic mastitis diate Ki67 proliferation index in this case exclude low-grade
• Marginal zone lymphoma B cell NHL and are inconsistent with Burkitt lymphoma.
• Triple-negative invasive ductal carcinoma, high nuclear
grade Case 5 (Fig. 9.25a–e)
• Diffuse large B cell lymphoma
• Burkitt lymphoma Learning Objectives
1. To generate a differential diagnosis for infiltrating lym-
IHC and Other Ancillary Studies (Fig. 9.24c–h) phoid lesions of the breast
• Pancytokeratin negative 2. To become familiar with the immunohistochemical fea-
• ER, PR, Her2 negative tures that may distinguish a clonal B cell disorder from a
• CD20 weak/partial positive reactive one
• CD3 scattered positive on small cells 3. To become familiar with the gross and histologic features
• CD79a strong positive that distinguish teh correct final diagnosis
• CD10 negative
• BCL6 positive History
• MUM1 positive • An 89-year-old female presents with focal right breast
• Ki67: 80% asymmetry noted on screening mammography. No suspi-
• FISH: Negative for MYC, BCL2, or BCL6 rearrangement cious calcifications or dominant mass was noted in heter-
ogeneously dense breast tissue.
Final Diagnosis
• Diffuse large B cell lymphoma, activated B cell (ABC) Histologic Findings
type • Infiltration of the breast parenchyma by small lymphoid
cells, focally surrounding breast ductules (Fig. 9.25)
Take-Home Messages • Mostly small mature lymphoid cells surrounding residual
1. In the absence of definitive architectural features of inva- follicular structures, with predominantly round nuclei,
sive carcinoma, lymphoid markers are indicated to hyperchromatic condensed chromatin, and scant cyto-
exclude large cell, high-grade, or poorly differentiated plasm (Fig. 9.25b)
lymphoma, while cytokeratin is also recommended to • Focal areas, including residual areas of adiposity, with
rule in triple-negative breast carcinoma. prominent plasmacytic infiltration
a b
c d
Fig. 9.25 Case 5. (a) Medium-power view of diffuse lymphocytic (600×). (c) High-power view of extensive plasmacytic component
infiltrate, noting single breast ductile at the left (100×). (b) High-power (600×). (d) Lambda light-chain in situ hybridization among plasma
view of small cell lymphoid infiltrate with hyperchromatic nuclei cells (400×). (e) Kappa light-chain in situ hybridization (400×)
Differential Diagnosis 3. Documentation of B cell clonality by alternative meth-

• Diabetic mastopathy ods is critical in extranodal low-grade B cell NHL, par-
• Follicular hyperplasia ticularly with MZLs which frequently lack any other
• Follicular lymphoma, grade 1 specific aberrant pattern of B cell antigen expression.
• Marginal zone lymphoma 4. Given the likelihood that most non-lymphoid tissues are
• Diffuse large B cell lymphoma not available for flow cytometry evaluation, kappa vs.
lambda immunohistochemical or in situ hybridization
IHC and Other Ancillary Studies (Fig. 9.25d, e) staining is indicated, frequently only reliable among
• In situ hybridization kappa/lambda ratio among plasma plasma cells when performed by most routine laborato-
cells, <1:10 (kappa negative, lambda positive) ries. Therefore, alternative methods such as PCR for
• CD5 negative clonal immunoglobulin gene rearrangements are strongly
• CD10 negative recommended.
• PCR studies positive for clonal B cell population for
immunoglobulin light-chain analysis Case 6 (Fig. 9.26a–h)
Final Diagnosis Learning Objectives

• Marginal zone lymphoma 1. To generate the differential diagnosis for nodular lym-
phoid proliferations
Take-Home Messages 2. To become familiar with the immunohistochemical

1. Low-grade lymphoma may mimic benign lymphocytic approach to evaluating this differential
inflammation, both cytologically (i.e., hyperchromatic nuclei, 3. To become familiar with the gross and histologic features
indistinct nucleoli, and minimally irregular nuclei) and archi- that distinguish the correct final diagnosis
tecturally (with increased plasma cells and residual benign or
colonized lymphoid follicles present within the lesion). History
2. Marginal zone lymphoma, in particular, frequently pres- • A 49-year-old female noted a lump in her left breast, with
ents with clonally derived plasma cells as part of the mammographic evidence of a 0.9 cm mass in the left
malignant proliferation. inner quadrant.
a b
Fig. 9.26 Case 6. (a) Low-power view of nodular lymphoid prolifera- hyperchromatic nuclei within the follicular proliferation (400×). (d)
tion (12.5×). (b) Medium-power view of follicular organization (40×). CD20 (4×). (e) CD10 (40×). (f) BCL6 (40×). (g) BCL2 (40×). (h) Ki67
(c) High-power view of small irregular lymphoid cells with cleaved and (40×)
c d
e f
g h
Histologic Findings 4. Rungta S, Kleer CG. Metaplastic carcinomas of the breast: diagnos-

tic challenges and new translational insights. Arch Pathol Lab Med.
• Nodular proliferation of small lymphoid follicles. Some 2012;136(8):896–900.
follicles are back-to-back, while others are more widely 5. Tse GM, Tan PH, Chaiwun B, Putti TC, Lui PC, Tsang AK, et al.
spaced, all with attenuated mantle zones (Fig. 9.26a, b). p63 is useful in the diagnosis of mammary metaplastic carcinomas.
• Nodules contain small bland irregular lymphoid cells. Pathology. 2006;38(1):16–20.
6. Lee A. Recent developments in the histological diagnosis of spindle
Mitoses are rare, and large cells (centroblasts) are lacking cell carcinoma, fibromatosis and phyllodes tumour of the breast.
within proliferating nodules (Fig. 9.26c). Histopathology. 2008;52(1):45–57.
7. Lacroix-Triki M, Geyer FC, Lambros MB, Savage K, Ellis IO,
Differential Diagnosis Lee AH, et al. β-catenin/Wnt signalling pathway in fibromatosis,
metaplastic carcinomas and phyllodes tumours of the breast. Mod
• Follicular hyperplasia Pathol. 2010;23(11):1438.
• Follicular lymphoma, grade 1 8. Shiga M, Okamoto K, Matsumoto M, Maeda H, Dabanaka K,
• Marginal zone lymphoma Namikawa T, et al. Nodular fasciitis in the mesentery, a differen-
• Diffuse large B cell lymphoma tial diagnosis of peritoneal carcinomatosis. World J Gastroenterol:
WJG. 2014;20(5):1361.
9. Uehara K, Ikehara F, Shibuya R, Nakazato I, Oshiro M, Kiyuna M,
IHC and Other Ancillary Studies (Fig. 9.26d–h) et al. Molecular signature of tumors with monoallelic 13q14 dele-
• CD20 positive tion: a case series of spindle cell lipoma and genetically-related
• CD10 positive tumors demonstrating a link between FOXO1 status and p38 MAPK
pathway. Pathol Oncol Res. 2018;24(4):861–9.
• BCL6 positive 10. Schwartz CJ, Schandl CA, Morse J, Ralston J, Rapkiewicz A,

• BCL2 positive Darvishian F. Benign fibromyxoid lesion of the breast: a distinct
• Ki67 low, 10–20% within nodules entity from benign spindle cell tumors of the mammary stroma? Int
J Surg Pathol. 2018 Sep;26(6):488–93.
11. D’Alfonso TM, Subramaniyam S, Ginter PS, Mosquera JM,

Final Diagnosis
MacDonald TY, Noorzad Z, et al. Characterization of the leiomyo-
• Follicular lymphoma, WHO grade 1 (of 3) matous variant of myofibroblastoma: a rare subset distinct from other
smooth muscle tumors of the breast. Hum Pathol. 2016;58:54–61.
Take-Home Messages 12. Bhattacharya B, Dilworth HP, Iacobuzio-Donahue C, Ricci F, Weber
K, Furlong MA, et al. Nuclear β-catenin expression distinguishes
1. The primary differential for a nodular lymphoid prolifera-
deep fibromatosis from other benign and malignant fibroblastic and
tion (particularly in the breast) is reactive follicular hyper- myofibroblastic lesions. Am J Surg Pathol. 2005;29(5):653–9.
plasia, follicular lymphoma, and follicular colonization 13. Qiu X, Montgomery E, Sun B. Inflammatory myofibroblastic

by marginal zone lymphoma. tumor and low-grade myofibroblastic sarcoma: a comparative
2. Low-power architectural features are important to distin- study of clinicopathologic features and further observations on
the immunohistochemical profile of myofibroblasts. Hum Pathol.
guish benign from neoplastic follicles, but neoplastic fol- 2008;39(6):846–56.
licles may be variably and even widely spaced, without 14. Coffin CM, Hornick JL, Fletcher CD. Inflammatory myofibroblastic
back-to-back organization throughout. tumor: comparison of clinicopathologic, histologic, and immuno-
3. In low-grade neoplastic follicles, the cytologic atypia is histochemical features including ALK expression in atypical and
aggressive cases. Am J Surg Pathol. 2007;31(4):509–20.
less, and the rate of proliferation (i.e., Ki67) is lower than 15. Downs-Kelly E, Nayeemuddin KM, Albarracin C, Wu Y, Hunt

in benign reactive follicles where marked cellular pleo- KK, Gilcrease MZ. Matrix-producing carcinoma of the breast: an
morphism and high proliferation rates are physiologically aggressive subtype of metaplastic carcinoma. Am J Surg Pathol.
normal. 2009;33(4):534–41.
16. Bhosale S, Kshirsagar A, Sulhyan S, Jagtap S, Nikam Y. Metaplastic
4. Lymphoid follicles co-expressing germinal center B cell carcinoma with predominant chondrosarcoma of the right breast.
markers (CD20, CD10, and BCL6), along with aberrant Case Rep Oncol. 2010;3(2):277–81.
BCL2, are a hallmark of follicular lymphoma. 17. Amadu AM, Soro D, Marras V, Satta G, Crivelli P, Conti M, et al.
Primary breast chondrosarcoma: imaging and pathological findings.
Eur J Radiol Open. 2017;4:138–40.
18. Errarhay S, Fetohi M, Mahmoud S, Saadi H, Bouchikhi C, Banani
References A. Primary chondrosarcoma of the breast: a case presentation and
review of the literature. World J Surg Oncol. 2013;11(1):208.
1. Fleming J, Long E, Ginsburg E, Gerscovich D, Meltzer P, Vonderhaar 19. Mujtaba SS, Haroon S, Faridi N. Primary chondrosarcoma of breast.
B. Interlobular and intralobular mammary stroma: genotype may J Coll Physicians Surg Pak. 2013;23:754–5.
not reflect phenotype. BMC Cell Biol. 2008;9(1):46. 20. Agoumi M, Giambattista J, Hayes MM. Practical considerations in
2. Rakha EA, Aleskandarany MA, Lee AH, Ellis IO. An approach to breast papillary lesions: a review of the literature. Arch Pathol Lab
the diagnosis of spindle cell lesions of the breast. Histopathology. Med. 2016;140(8):770–90.
2016;68(1):33–44. 21. Joshi M, Remoundos DD, Ahmed F, Rees G, Cunnick G. An

3. Davis WG, Hennessy B, Babiera G, Hunt K, Valero V, Buchholz unusual breast lump: osseous metaplasia. BMJ Case Rep.
TA, et al. Metaplastic sarcomatoid carcinoma of the breast with 2013;2013:bcr2012008239.
absent or minimal overt invasive carcinomatous component: a mis- 22. Alyami H, Emad A-O, Harbi S, Bshait MB. Benign osseous meta-
nomer. Am J Surg Pathol. 2005;29(11):1456–63. plasia of the breast: case report. Int J Surg Case Rep. 2018;44:90–2.
23. Lynch LA, Moriarty AT. Localized primary amyloid tumor asso- 43. Aviv A, Tadmor T, Polliack A. Primary diffuse large B-cell lym-
ciated with osseous metaplasia presenting as bilateral breast phoma of the breast: looking at pathogenesis, clinical issues and
masses: cytologic and radiologic features. Diagn Cytopathol. therapeutic options. Ann Oncol. 2013;24(9):2236–44.
1993;9(5):570–5. 44. Caon J, Wai ES, Hart J, Alexander C, Truong PT, Sehn LH, et al.
24. Brustugun OT, Reed W, Poulsen JP, Bruland ØS. Primary osteosar- Treatment and outcomes of primary breast lymphoma. Clin Breast
coma of the breast. Acta Oncol. 2005;44(7):767–70. Cancer. 2012;12(6):412–9.
25. Guys N, Khan S, Kezlarian B, Shah BA. Primary osteosarcoma of 45. Hugh JC, Jackson FI, Hanson J, Poppema S. Primary breast lym-
the breast. Appl Radiol. 2017;46(12):28–9. phoma. An immunohistologic study of 20 new cases. Cancer.
26. Bahrami A, Resetkova E, Ro JY, Ibañez JD, Ayala AG. Primary 1990;66(12):2602–11.
osteosarcoma of the breast: report of 2 cases. Arch Pathol Lab Med. 46. Jeanneret-Sozzi W, Taghian A, Epelbaum R, Poortmans P, Zwahlen
2007;131(5):792–5. D, Amsler B, et al. Primary breast lymphoma: patient profile, out-
27. van Roggen JG, Zonderland H, Welvaart K, Peterse J, Hogendoorn come and prognostic factors. A multicentre Rare Cancer Network
P. Local recurrence of a phyllodes tumour of the breast presenting study. BMC Cancer. 2008;8:86.
with widespread differentiation to a telangiectatic osteosarcoma. J 47. Surov A, Holzhausen HJ, Wienke A, Schmidt J, Thomssen C,

Clin Pathol. 1998;51(9):706–8. Arnold D, et al. Primary and secondary breast lymphoma: preva-
28. Crevecoeur J, Jossa V, Gennigens C, Parmentier JC, Crevecoeur lence, clinical signs and radiological features. Br J Radiol.
A. Primary osteosarcoma of the breast: a case report. Clinical Case 2012;85(1014):e195–205.
Rep. 2016;4(1):62–6. 48. Talwalkar SS, Miranda RN, Valbuena JR, Routbort MJ, Martin
29. Momoi H, Wada Y, Sarumaru S, Tamaki N, Gomi T, Kanaya
AW, Medeiros LJ. Lymphomas involving the breast: a study of 106
S, et al. Primary osteosarcoma of the breast. Breast Cancer. cases comparing localized and disseminated neoplasms. Am J Surg
2004;11(4):396–400. Pathol. 2008;32(9):1299–309.
30. Papalas JA, Wylie JD, Dash RC. Recurrence risk and margin status 49. Validire P, Capovilla M, Asselain B, Kirova Y, Goudefroye R,

in granular cell tumors of the breast: a clinicopathologic study of 13 Plancher C, et al. Primary breast non-Hodgkin’s lymphoma: a large
patients. Arch Pathol Lab Med. 2011;135(7):890–5. single center study of initial characteristics, natural history, and
31. Fineberg S, Rosen PP. Cutaneous angiosarcoma and atypical vascu- prognostic factors. Am J Hematol. 2009;84(3):133–9.
lar lesions of the skin and breast after radiation therapy for breast 50. Wiseman C, Liao KT. Primary lymphoma of the breast. Cancer.
carcinoma. Am J Clin Pathol. 1994;102(6):757–63. 1972;29(6):1705–12.
32. Karlsson P, Holmberg E, Johansson K-A, Kindblom L-G, Carstensen 51. Yhim HY, Kang HJ, Choi YH, Kim SJ, Kim WS, Chae YS, et al.
J, Wallgren A. Soft tissue sarcoma after treatment for breast cancer. Clinical outcomes and prognostic factors in patients with breast dif-
Radiother Oncol. 1996;38(1):25–31. fuse large B cell lymphoma; Consortium for Improving Survival of
33. Kirova Y, Feuilhade F, Calitchi E, Otmezguine Y, Belembaogo E, Lymphoma (CISL) study. BMC Cancer. 2010;10:321.
Le JB. Radiation-induced sarcoma after breast cancer. Apropos 52. Domchek SM, Hecht JL, Fleming MD, Pinkus GS, Canellos

of 8 cases and review of the literature. Cancer Radiother. GP. Lymphomas of the breast: primary and secondary involvement.
1998;2(4):381–6. Cancer. 2002;94(1):6–13.
34. Kirova YM, Vilcoq JR, Asselain B, Sastre-Garau X, Fourquet
53.
Mattia AR, Ferry JA, Harris NL. Breast lymphoma. A
A. Radiation-induced sarcomas after radiotherapy for breast carci- B-cell spectrum including the low grade B-cell lymphoma
noma. Cancer. 2005;104(4):856–63. of mucosa associated lymphoid tissue. Am J Surg Pathol.
35. Alanis L, Roth R, Lerman N, Barroeta JE, Germaine P. Radiologic 1993;17(6):574–87.
images of an aggressive implant-associated fibromatosis of the 54. Cheah CY, Campbell BA, Seymour JF. Primary breast lymphoma.
breast and chest wall: case report and review of the literature. Cancer Treat Rev. 2014;40(8):900–8.
Radiology Case Rep. 2017;12(3):431–8. 55. Gualco G, Bacchi CE. B-cell and T-cell lymphomas of the breast:
36. Kobayashi S, Iwase H, Karamatsu S, Masaoka A, Nakamura T. A clinical – pathological features of 53 cases. Int J Surg Pathol.
case of stromal sarcoma of the breast occurring after augmentation 2008;16(4):407–13.
mammoplasty. Gan No Rinsho. 1988;34(4):467–72. 56. Gualco G, Chioato L, Harrington WJ Jr, Weiss LM, Bacchi

37. Smoll NR, Farhadieh RD, Ferguson R, Findlay MW, Hunter-Smith CE. Primary and secondary T-cell lymphomas of the breast: clinico-
DJ. High-grade angiosarcoma associated with ruptured breast pathologic features of 11 cases. Appl Immunohistochem Mol
implants. Plast Reconstr Surg Glob Open. 2013;1(1):e11–3. Morphol. 2009;17(4):301–6.
38. Bansal M, Vega S, Yeaney GA, Wang X. Bilateral chest wall sar- 57. Mpallas G, Simatos G, Tasidou A, Patra E, Galateros G, Lakiotis
comas associated with silicone implant capsules in a patient with G, et al. Primary breast lymphoma in a male patient. Breast.
Li-Fraumeni syndrome. J Case Rep Images Pathol. 2016;2:6–9. 2004;13(5):436–8.
39. Holliday R. Neoplastic transformation: the contrasting stability of 58. Murata T, Kuroda H, Nakahama T, Goshima H, Shiraishi T, Yatani
human and mouse cells. Cancer Surv. 1996;28:103–15. R. Primary non-Hodgkin malignant lymphoma of the male breast.
40. Guo T, Zhang L, Chang NE, Singer S, Maki RG, Antonescu
Jpn J Clin Oncol. 1996;26(4):243–7.
CR. Consistent MYC and FLT4 gene amplification in radiation- 59. Rathod J, Taori K, Disawal A, Gour P, Dhakate S, Mone R, et al.
induced angiosarcoma but not in other radiation-associated atypical A rare case of male primary breast lymphoma. J Breast Cancer.
vascular lesions. Genes Chromosom Cancer. 2011;50(1):25–33. 2011;14(4):333–6.
41.
Mentzel T, Schildhaus H, Palmedo G, Büttner R, Kutzner 60.
Kim SH, Ezekiel MP, Kim RY. Primary lymphoma of the
H. Postradiation cutaneous angiosarcoma after treatment of breast breast: breast mass as an initial symptom. Am J Clin Oncol.
carcinoma is characterized by MYC amplification in contrast to 1999;22(4):381–3.
atypical vascular lesions after radiotherapy and control cases: clini- 61. Sabate JM, Gomez A, Torrubia S, Camins A, Roson N, De Las
copathological, immunohistochemical and molecular analysis of 66 Heras P, et al. Lymphoma of the breast: clinical and radiologic
cases. Mod Pathol. 2012;25(1):75. features with pathologic correlation in 28 patients. Breast J.
42. Fraga-Guedes C, André S, Mastropasqua M, Botteri E, Toesca A, 2002;8(5):294–304.
Rocha R, et al. Angiosarcoma and atypical vascular lesions of the 62. Shim E, Song SE, Seo BK, Kim YS, Son GS. Lymphoma affect-
breast: diagnostic and prognostic role of MYC gene amplification and ing the breast: a pictorial review of multimodal imaging findings. J
protein expression. Breast Cancer Res Treat. 2015;151(1):131–40. Breast Cancer. 2013;16(3):254–65.
63. Lyou CY, Yang SK, Choe DH, Lee BH, Kim KH. Mammographic cell lymphoma: long-term follow-up of 60 patients. J Clin Oncol.
and sonographic findings of primary breast lymphoma. Clin 2014;32(2):114–20.
Imaging. 2007;31(4):234–8. 84. Roden AC, Macon WR, Keeney GL, Myers JL, Feldman AL, Dogan
64. Boudova L, Kazakov DV, Sima R, Vanecek T, Torlakovic E, Lamovec A. Seroma-associated primary anaplastic large-cell lymphoma adja-
J, et al. Cutaneous lymphoid hyperplasia and other lymphoid infil- cent to breast implants: an indolent T-cell lymphoproliferative dis-
trates of the breast nipple: a retrospective clinicopathologic study of order. Mod Pathol. 2008;21(4):455–63.
fifty-six patients. Am J Dermatopathol. 2005;27(5):375–86. 85. Rupani A, Frame JD, Kamel D. Lymphomas associated with

65. Ely KA, Tse G, Simpson JF, Clarfeld R, Page DL. Diabetic
breast implants: a review of the literature. Aesthet Surg J.
mastopathy. A clinicopathologic review. Am J Clin Pathol. 2015;35(5):533–44.
2000;113(4):541–5. 86. Taylor CR, Siddiqi IN, Brody GS. Anaplastic large cell lym-

66. Martin SJ, Duvic M. Treatment of cutaneous lymphoid hyper-
phoma occurring in association with breast implants: review of
plasia with the monoclonal anti-CD20 antibody rituximab. Clin pathologic and immunohistochemical features in 103 cases. Appl
Lymphoma Myeloma Leuk. 2011;11(3):286–8. Immunohistochem Mol Morphol. 2013;21(1):13–20.
67. Tomaszewski JE, Brooks JS, Hicks D, Livolsi VA. Diabetic mas- 87. Thompson PA, Lade S, Webster H, Ryan G, Prince HM. Effusion-
topathy: a distinctive clinicopathologic entity. Hum Pathol. associated anaplastic large cell lymphoma of the breast: time for it to
1992;23(7):780–6. be defined as a distinct clinico-pathological entity. Haematologica.
68. Valdez R, Thorson J, Finn WG, Schnitzer B, Kleer CG. Lymphocytic 2010;95(11):1977–9.
mastitis and diabetic mastopathy: a molecular, immunopheno- 88. Thompson PA, Prince HM. Breast implant-associated anaplastic
typic, and clinicopathologic evaluation of 11 cases. Mod Pathol. large cell lymphoma: a systematic review of the literature and mini-
2003;16(3):223–8. meta analysis. Curr Hematol Malig Rep. 2013;8(3):196–210.
69. Brogi E, Harris NL. Lymphomas of the breast: pathology and clini- 89. Fruchart C, Denoux Y, Chasle J, Peny AM, Boute V, Ollivier JM,
cal behavior. Semin Oncol. 1999;26(3):357–64. et al. High grade primary breast lymphoma: is it a different clinical
70. Radkani P, Joshi D, Paramo JC, Mesko TW. Primary breast lym- entity? Breast Cancer Res Treat. 2005;93(3):191–8.
phoma 30 years of experience with diagnosis and treatment at a 90. Martinelli G, Ryan G, Seymour JF, Nassi L, Steffanoni S, Alietti
single medical center. JAMA Surg. 2014;149(1):91–3. A, et al. Primary follicular and marginal-zone lymphoma of the
71. Uesato M, Miyazawa Y, Gunji Y, Ochiai T. Primary non-Hodgkin’s breast: clinical features, prognostic factors and outcome: a study
lymphoma of the breast: report of a case with special reference to 380 by the International Extranodal Lymphoma Study Group. Ann
cases in the Japanese literature. Breast Cancer. 2005;12(2):154–8. Oncol. 2009;20(12):1993–9.
72. Ganjoo K, Advani R, Mariappan MR, McMillan A, Horning S. Non- 91. Aviles A, Neri N, Nambo MJ. The role of genotype in 104 cases
Hodgkin lymphoma of the breast. Cancer. 2007;110(1):25–30. of diffuse large B-cell lymphoma primary of breast. Am J Clin
73. Guo HY, Zhao XM, Li J, Hu XC. Primary non-Hodgkin’s lym- Oncol. 2012;35(2):126–9.
phoma of the breast: eight-year follow-up experience. Int J Hematol. 92. Li D, Deng J, He H, Bu Y, Peng F, Tang X, et al. Primary breast
2008;87(5):491–7. diffuse large B-cell lymphoma shows an activated B-cell-like phe-
74.
Arber DA, Simpson JF, Weiss LM, Rappaport H. Non- notype. Ann Diagn Pathol. 2012;16(5):335–43.
Hodgkin’s lymphoma involving the breast. Am J Surg Pathol. 93. Niitsu N, Okamoto M, Nakamine H, Hirano M. Clinicopathologic
1994;18(3):288–95. features and treatment outcome of primary breast diffuse large
75. Ryan G, Martinelli G, Kuper-Hommel M, Tsang R, Pruneri G, Yuen B-cell lymphoma. Leuk Res. 2008;32(12):1837–41.
K, et al. Primary diffuse large B-cell lymphoma of the breast: prog- 94. Yoshida S, Nakamura N, Sasaki Y, Yoshida S, Yasuda M, Sagara H,
nostic factors and outcomes of a study by the International Extranodal et al. Primary breast diffuse large B-cell lymphoma shows a non-ger-
Lymphoma Study Group. Ann Oncol. 2008;19(2):233–41. minal center B-cell phenotype. Mod Pathol. 2005;18(3):398–405.
76. Aladily TN, Medeiros LJ, Amin MB, Haideri N, Ye D, Azevedo 95. Hans CP, Weisenburger DD, Greiner TC, Gascoyne RD, Delabie
SJ, et al. Anaplastic large cell lymphoma associated with breast J, Ott G, et al. Confirmation of the molecular classification of dif-
implants: a report of 13 cases. Am J Surg Pathol. 2012;36(7):1000–8. fuse large B-cell lymphoma by immunohistochemistry using a tis-
77. Brody GS, Deapen D, Taylor CR, Pinter-Brown L, House-Lightner sue microarray. Blood. 2004;103(1):275–82.
SR, Andersen JS, et al. Anaplastic large cell lymphoma occurring in 96. Armitage JO, Feagler JR, Skoog DP. Burkitt lymphoma dur-
women with breast implants: analysis of 173 cases. Plast Reconstr ing pregnancy with bilateral breast involvement. JAMA.
Surg. 2015;135(3):695–705. 1977;237(2):151.
78. de Jong D, Vasmel WL, de Boer JP, Verhave G, Barbe E, Casparie 97. Fahmy JL, Wood BP, Miller JH. Bilateral breast involve-
MK, et al. Anaplastic large-cell lymphoma in women with breast ment in a teenage girl with Burkitt lymphoma. Pediatr Radiol.
implants. JAMA. 2008;300(17):2030–5. 1995;25(1):56–7.
79. Largent J, Oefelein M, Kaplan HM, Okerson T, Boyle P. Risk of 98. Lingohr P, Eidt S, Rheinwalt KP. A 12-year-old girl presenting
lymphoma in women with breast implants: analysis of clinical stud- with bilateral gigantic Burkitt’s lymphoma of the breast. Arch
ies. Eur J Cancer Prev. 2012;21(3):274–80. Gynecol Obstet. 2009;279(5):743–6.
80. Lazzeri D, Agostini T, Bocci G, Giannotti G, Fanelli G, Naccarato 99. Savvari P, Matsouka C, Barbaroussi D, Christoulas D, Nikitas N,
AG, et al. ALK-1-negative anaplastic large cell lymphoma associ- Dimopoulos MA, et al. Burkitt’s lymphoma in pregnancy with
ated with breast implants: a new clinical entity. Clin Breast Cancer. bilateral breast involvement: case report with review of the litera-
2011;11(5):283–96. ture. Onkologie. 2010;33(8–9):461–4.
81.
Leberfinger AN, Behar BJ, Williams NC, Rakszawski KL, 100. Shepherd JJ, Wright DH. Burkitt’s tumour presenting as bilateral
Potochny JD, Mackay DR, et al. Breast implant-associated ana- swelling of the breast in women of child-bearing age. Br J Surg.
plastic large cell lymphoma: a systematic review. JAMA Surg. 1967;54(9):776–80.
2017;152(12):1161–8. 101. Zygogianni AG, Kokkakis J, Antypas C, Armpilia C, Kouloulias
82. Lipworth L, Tarone RE, McLaughlin JK. Breast implants and lym- V, Kouvaris JR. Bilateral primary breast Burkitt’s lymphoma.
phoma risk: a review of the epidemiologic evidence through 2008. Breast J. 2010;16(6):655–6.
Plast Reconstr Surg. 2009;123(3):790–3. 102. Jones DE, d’Avignon MB, Lawrence R, Latshaw RF. Burkitt’s
83. Miranda RN, Aladily TN, Prince HM, Kanagal-Shamanna R, de lymphoma: obstetric and gynecologic aspects. Obstet Gynecol.
Jong D, Fayad LE, et al. Breast implant-associated anaplastic large- 1980;56(4):533–6.
103. Horowitz NA, Benyamini N, Wohlfart K, Brenner B, Avivi and FLT3-mutated AML after allogeneic stem cell transplanta-
I. Reproductive organ involvement in non-Hodgkin lym- tion. J Clin Pathol. 2010;63(6):558–61.
phoma during pregnancy: a systematic review. Lancet Oncol. 114. D’Costa GF, Hastak MS, Patil YV. Granulocytic sarcoma

2013;14(7):e275–82. of breast: an aleukemic presentation. Indian J Med Sci.
104. Negahban S, Ahmadi N, Oryan A, Khojasteh HN, Aledavood A, 2007;61(3):152–5.
Soleimanpour H, et al. Primary bilateral Burkitt lymphoma of the 115. Delporte F, Voorhoopf LJ, Lodewyck T, De Paepe P. Primary
lactating breast: a case report and review of the literature. Mol granulocytic sarcoma of the breast: a case report and review of the
Diagn Ther. 2010;14(4):243–50. literature. Eur J Gynaecol Oncol. 2011;32(4):435–8.
105. Patron R, Miles EF. Stage IAE follicular lymphoma of the breast: 116. Jennings WC, Baker RS, Murray SS, Howard CA, Parker DE,
case report and review of the literature. Case Rep Oncol Med. Peabody LF, et al. Primary breast lymphoma: the role of mas-
2013;2013:597527. tectomy and the importance of lymph node status. Ann Surg.
106. Chatterjee D, Bal A, Das A, Ahluwalia J, Singh G. Extramedullary 2007;245(5):784–9.
myeloid sarcoma of bilateral breast as first manifestation of 117. Joks M, Mysliwiec K, Lewandowski K. Primary breast lym-
acute myeloid leukemia – a diagnostic challenge. Breast J. phoma - a review of the literature and report of three cases. Arch
2015;21(6):679–80. Med Sci. 2011;7(1):27–33.
107. Fu J, Luo J. Granulocytic sarcoma of the breast in acute myeloid 118. Zhao S, Zhang QY, Ma WJ, Zhang MH, Sun WZ, Li HB, et al.
leukemia: two case reports. Oncol Lett. 2014;7(1):145–7. Analysis of 31 cases of primary breast lymphoma: the effect of
108. Nangal JK, Kapoor A, Narayan S, Singhal MK, Beniwal S, nodal involvement and microvascular density. Clin Lymphoma
Kumar HS. A case of CD68 negative histiocytic sarcoma of Myeloma Leuk. 2011;11(1):33–7.
axilla masquerading as metastatic breast cancer. J Surg Case Rep. 119. Hosein PJ, Maragulia JC, Salzberg MP, Press OW, Habermann
2014;2014(7):pii TM, Vose JM, et al. A multicentre study of primary breast dif-
109. O’Kane D, Jenkinson H, Carson J. Langerhans cell histiocytosis fuse large B-cell lymphoma in the rituximab era. Br J Haematol.
associated with breast carcinoma successfully treated with topical 2014;165(3):358–63.
imiquimod. Clin Exp Dermatol. 2009;34(8):e829–32. 120. Wong WW, Schild SE, Halyard MY, Schomberg PJ. Primary non-
110. Stewart RL, Dell CM, Samayoa L. Myeloid sarcoma of the breast Hodgkin lymphoma of the breast: the Mayo Clinic experience. J
misdiagnosed as poorly differentiated mammary carcinoma with Surg Oncol. 2002;80(1):19–25.
lobular features. Breast J. 2015;21(2):192–3. 121. Yhim HY, Kim JS, Kang HJ, Kim SJ, Kim WS, Choi CW, et al.
111. Valbuena JR, Admirand JH, Gualco G, Medeiros LJ. Myeloid Matched-pair analysis comparing the outcomes of primary breast
sarcoma involving the breast. Arch Pathol Lab Med. and nodal diffuse large B-cell lymphoma in patients treated with
2005;129(1):32–8. rituximab plus chemotherapy. Int J Cancer. 2012;131(1):235–43.
112. Wu B, Li F, Zou S. MLL-AF9 rearrangement in myeloid sarcomas 122. Cheah CY, Herbert KE, O’Rourke K, Kennedy GA, George
involving the breast. Ann Hematol. 2014;93(4):709–10. A, Fedele PL, et al. A multicentre retrospective comparison of
113.
Choschzick M, Bacher U, Ayuk F, Lebeau central nervous system prophylaxis strategies among patients
A. Immunohistochemistry and molecular analyses in myeloid with high-risk diffuse large B-cell lymphoma. Br J Cancer.
sarcoma of the breast in a patient with relapse of NPM1-mutated 2014;111(6):1072–9.
Metastatic Cancer in the Breast
10
Bradley M. Turner
List of Frequently Asked Questions primary carcinoma in a known location [5, 13]; however,
in approximately 30% of patients with metastatic disease
1. What is the frequency of metastatic cancer to the to the breast, the metastases is the first sign of malignancy
breast? [2–5, 7, 8, 14]. In these cases, the occult primary tumor is
often misinterpreted as a primary breast malignancy [13].
The most common metastatic cancers in the breast are from The male-to-female ratio has been reported as anywhere
a contralateral primary breast carcinoma [1–7]. Contralateral from one male for every six to eleven females [1, 3, 5, 13].
primary breast metastases are excluded in most series dis- The age range for metastases to the breast from an extra-
cussing metastatic disease to the breast [2] and will not be mammary location has been reported from as young as
discussed further in this chapter. Metastases to the breast and 12 years of age to as old as 90 years of age [1, 3, 4]. The
axilla from extramammary locations are extremely uncom- most common tumor metastatic to the breast in the pediat-
mon. The frequency of metastases to the breast from extra- ric population is a rhabdomyosarcoma [1, 13, 15, 16], and
mammary locations varies depending on the clinical study, the proportion of metastases in children and young adults
with published reports between 0.2% and 3% [1, 2, 4, 5, 7– is significantly higher compared to older adults [2, 13, 15].
9]. Higher frequencies of 2–7% have been reported in post- Only a single study (the largest series found to date which
mortem studies [2, 6, 10]. addresses the topic) reports on ethnicity [3]. In this study
of 169 patients with metastases to the breast from an extra-
2. Other than frequency, what is the epidemiology of mammary location, 85.2% (n = 144) were Caucasian, 7.7%
metastatic cancer to the breast? (n = 13) were African–American, and 7.1% (n = 12) were
of Latino origin.
A review of the English-language literature suggests that
fewer than 750 cases of metastases to the breast from an 3. Why is it important to recognize a metastatic cancer
extramammary location have been reported [1–3, 9, 11]. The from a primary breast carcinoma?
first case of metastases to the breast from an extramammary
location may have been reported as early as 1855 [11], and It is critical to consider metastatic disease as a possibility
the first documented case in the peer-reviewed English litera- because the treatment options are different. Delays in diag-
ture was reported in 1903 [12]. nosis contribute to the poor prognosis associated with
In approximately 70% of patients with metastatic dis- most metastases to the breast from an extramammary loca-
ease to the breast who have a breast lump, there is a known tion. Patients with metastatic disease to the breast do not
usually benefit from mastectomy [13], so a correct diagno-
sis is crucial in avoiding unnecessary procedures. In par-
ticular, radical breast surgery and axillary nodal dissection
may not be appropriate in patients with systemic disease
[15]. Systemic chemotherapy options are vastly different
depending on the type of metastatic disease, and although
B. M. Turner (*)
most patients die within a year of diagnosis [2, 17–19],
Department of Pathology and Laboratory Medicine, University of
Rochester Medical Center/Highland Hospital, Rochester, NY, USA longer survival is well recognized if there is effective sys-
e-mail: [email protected] temic treatment [2, 13, 14, 18].

238 B. M. Turner
4. What is the differential diagnosis for metastatic can- Table 10.2 Frequency of extramammary metastasis to the breast by
cer to the breast? tumor typea
Type of metastatic diseasea % frequency % frequency (all
(n = 707) [1–4, 11, 15, 20] (carcinoma) metastasis)
The first consideration of a primary location for metastatic
Carcinoma (n = 398) 56
disease to the breast should be a primary breast carcinoma
Pulmonary 28 16
[1–7, 16]. Some authors have suggested that the rarity of Gynecological 28 15
metastatic disease from an extramammary location is due to Genitourinary 21 12
the characteristics of breast tissue [20–22]. When metastatic Gastrointestinal 16 9
disease from an extramammary location does occur, it has Head and Neck 7 4
been suggested that hormonal status may play a role, based Other carcinoma <1 <1
on the increased occurrence of breast metastasis in pubes- Melanoma 27
cent, lactating, and pregnant females [20, 23–25]. Other Neuroendocrine 10
hypotheses include a transfection phenomenon of the cancer Sarcoma 7
Other non-carcinoma <1
genome and the dissemination of stem cells that spread by
systemic, lymphatic, or transcoelomic migration [20, 26, Excluding hematopoietic tumors
a
27]. This latter hypothesis may offer some insight into the
second most common cause of metastatic disease to the carcinoma [1], neuroblastoma [15], and malignant fibrous
breast, hematopoietic neoplasms [16, 28, 29]. Because breast histiocytoma [1, 28].
lymphomas occur as a result of systemic involvement of the
lymphoid tissue, some authors propose exclusion of lympho- 5. Are there clinical differences in patients with meta-
mas under the heading of “metastasis in the breast” [28], and static cancer to the breast as opposed to a cancer of
most reviews and case series exclude hematopoietic disease primary breast origin?
when discussing metastatic disease to the breast from an
extramammary location. Excluding hematopoietic tumors, a A thorough history may reveal the possibility of a metastatic
review of the literature [1–4, 11, 15, 20] suggests that the origin in a patient diagnosed with breast carcinoma, as the
most common site of metastatic origin is the skin (Table 10.1), majority of patients with metastatic breast disease from
the most common classification of metastatic disease is a extramammary locations will have a history of extramam-
carcinoma (Table 10.2), and the most common type of meta- mary malignancy [15]. It is critical that any history of prior
static neoplasm is melanoma (Table 10.2). Discordance in carcinoma be provided to the pathologist in a patient diag-
prevalence in individual studies is likely due to study loca- nosed with breast carcinoma. In one series [1], the failure of
tion and disease prevalence. Many of the neuroendocrine the pathologist to recognize the metastatic nature of the
tumors arise from the lung, so a neuroendocrine carcinoma lesion resulted most often because the clinician failed to pro-
of the breast should prompt an evaluation for possible pul- vide history regarding a previous cancer. Most patients with
monary origin. Other less common sites and types of tumors primary breast carcinoma are asymptomatic, presenting after
that have been reported to be metastatic to the breast include an abnormal screening examination. In contrast, patients
non-melanoma skin (squamous cell carcinoma, Merkel cell with metastatic breast carcinoma may be more likely to pres-
[1]) and non-skin melanoma (typically ocular) [1, 4]. Other ent with a rapidly growing painless firm palpable breast mass
even more unusual sites and types of metastases have been [2, 4, 8, 17, 18, 23, 30–32], and the diagnosis is typically
reported, including the thymus [20], heart [20], mesotheli- made by physical exam [2, 3]. Diffuse skin involvement is
oma [4], tongue [1], choriocarcinoma [1, 15], adenoid cystic rare [2, 3]. The metastatic tumor is most frequently unilateral
and solitary [1, 3, 13] but may present as multiple and/or
bilateral lesions [1–3, 13]. Axillary and/or regional lymph-
Table 10.1 Frequency of extramammary metastasis to the breast by adenopathy may or may not be present; if present, suspicion
site of origina
of metastatic disease should be aroused in correlation with
Site of metastatic origina (n = 678) [1–4, 11, 15, 20] % frequency
other clinical, radiographic, and pathologic information [2,
Skin 28
16]. Suspicion should be particularly high in cases of an axil-
Pulmonary 24
Gynecological 17
lary tumor without any evidence of a primary breast carci-
Genitourinary 12 noma on clinical exam or radiographic imaging. Cases with
Gastrointestinal 10 metastatic disease beyond the axilla at presentation should
Soft tissue 5 also raise suspicion. In one series, the majority of metastatic
Head and neck 4 cases to the breast presented with widespread metastatic dis-
Other <1 ease, often at multiple sites, including the liver, bone, subcu-
Excluding hematopoietic tumors
a
taneous sites, and lymph nodes [15].
10 Metastatic Cancer in the Breast 239
6. Are there radiographic differences in patients with

metastatic cancer to the breast as opposed to a cancer
of primary breast origin?
Although there are no specific features to distinguish pri-

mary breast carcinoma and metastatic breast disease from an
extramammary location, imaging studies may be helpful in
correlation with clinical history and pathologic information.
The most common mammographic appearance is a rounded
mass with well-defined or slightly irregular margins [2, 8,
23, 33, 34]. The majority of metastases are solitary and uni-
lateral, although multiple and bilateral metastatic tumors do
occur [16]. Unlike primary breast carcinoma, mammo-
graphic spiculations are rare [2, 23, 35]. Most lesions present
in the upper outer quadrant [17, 20, 23, 35]. Calcifications
are rarely seen [3, 14, 33, 35]; however, if present and meta-
static disease is being considered, calcifications raise the sus- Fig. 10.1 Elastosis in breast carcinoma (20×). Note the clumps of elas-
picion for metastatic ovarian serous carcinoma [1–4, 14, 17, tic fibers (arrows), or elastosis, present associated with the invasive
19, 33, 35–37]. An ultrasound will often show a hypoechoic breast carcinoma. On hematoxylin and eosin staining, these clumps of
elastic fibers often have a distinct grayish to bluish hue, well demar-
mass, which may be heterogeneous or poorly defined [33].
cated from the intervening eosinophilic stroma. The presence of elasto-
Multiple lesions or well-circumscribed masses felt to be sis, which is common in primary breast carcinoma, would suggest a
radiographically benign, but with histology consistent with primary breast origin, as elastosis is rare in extramammary tumors
carcinoma, should raise suspicion for a metastatic disease metastatic to the breast
[15, 38]. Radiographically, the metastasis may be present in
the adipose tissue near the chest wall [16]. was notably absent in 87% (40/46) of metastatic tumors to
the breast in one report [1], and reports of metastatic hepato-
7. Are there histologic differences in patients with meta- cellular [20], gastric [20], and ovarian carcinomas [1–3] with
static cancer to the breast as opposed to a cancer of associated microcalcifications have all been previously
primary breast origin? reported in the literature. Additionally, the absence of estro-
gen receptor (ER) and progesterone receptor (PR) in a well-
Metastatic breast disease from an extramammary location differentiated carcinoma, and all triple-negative (ER/PR/
often has histologic features similar to primary breast carci- HER-2-negative) tumors should prompt consideration for
noma; however, histology that is atypical for a primary breast metastatic disease.
carcinoma – such as pigment, intranuclear inclusions, neuro-
endocrine features, pure squamous features, well-formed 8. How might a metastatic melanoma present differently
papillae, tall columnar cells with mucin, high-grade dysco- from primary breast carcinoma or a primary breast
hesive cells, or clear cell features – may be present and melanoma?
should prompt consideration for metastatic breast disease
from an extramammary location. A well-circumscribed A clinical history of melanoma should immediately raise sus-
lesion sharply demarcated from adjacent normal breast tis- picion for metastatic disease. A thorough skin examination
sue, in the absence of in situ carcinoma, should also prompt (including a thorough examination of the breast skin to rule
consideration for metastatic breast disease from an extrama- out the rare primary breast melanoma) should be performed
mmary location [1, 2, 13, 23, 39]. Elastosis, which is com- on any new cancer diagnosis, and breast cancer is no excep-
mon in primary breast carcinoma (Fig. 10.1), is rare in tion. Although uncommon, 4–5% of melanomas arise from
extramammary tumors [2, 13, 40]. non-cutaneous sites [41], typically the eye. As such, a thor-
It is likely that metastatic breast disease from an extrama- ough ophthalmologic examination should be considered if the
mmary location occurs in the setting of advanced disease and clinical presentation is suspicious for metastatic melanoma.
widespread systemic metastasis [3], and the presence of
extensive lymphovascular invasion and the absence of micro- 9. How might metastatic pulmonary carcinoma present
calcifications should raise suspicion that another primary differently from primary breast carcinoma?
focus might be present [13]. The absence of lymphovascular
invasion and the presence of microcalcifications should be A clinical history of pulmonary carcinoma should immedi-
viewed with caution, however, as lymphovascular invasion ately raise suspicion for a metastatic disease. A thorough
240 B. M. Turner
pulmonary examination should be performed on any new 13. How might a metastatic neuroendocrine tumor pres-
cancer diagnosis, and breast cancer is no exception. A cough ent differently from primary breast carcinoma or a
that does not go away or gets worse, coughing up blood or primary breast neuroendocrine tumor?
rust-colored sputum, hoarseness, shortness of breath,
decreased exercise intolerance, new onset of wheezing, Any neuroendocrine carcinoma in the breast should raise
recurrent cough, recurrent bronchitis, recurrent pneumonia, consideration for metastatic tumor to the breast, particularly
and chest pain that is often worse with deep breathing, from a pulmonary origin. A thorough pulmonary history and
coughing, or laughing may all be signs and symptoms of exam should be done, with consideration for radiographic
lung involvement. imaging. Symptoms of carcinoid syndrome (facial flushing,
severe diarrhea, wheezing, and tachycardia) should draw
10. How might a metastatic gynecological tract tumor suspicion to the possibility of metastatic disease, as the gas-
present differently from primary breast carcinoma? trointestinal tract (including colon, liver, small bowel, and
appendix) has been documented as an origin of metastatic
The most common type of metastatic gynecological tract neuroendocrine disease to the breast [1, 2, 11, 16, 23, 32].
tumor to the breast is ovarian serous papillary carcinoma [1, Neuroendocrine tumor has also been reported to have metas-
2, 16, 42–44]. If there is metastasis to the breast, it would be tasized to the breast from the cervix [1].
unusual for gynecologic malignancy not to be present clini-
cally [16]. A clinical history of a gynecologic tract tumor 14. How might a metastatic head and neck carcinomas
should immediately raise suspicion for metastatic disease. A present differently from primary breast carcinoma?
clinical history of vaginal bleeding, back or pelvic discom-
fort, or bowel obstruction may be present and might prompt The most common metastatic head and neck tumor reported
evaluation for a metastatic gynecologic tract tumor. to metastasize to the breast is thyroid carcinoma [1, 2, 4, 11,
20]. Submandibular gland (adenoid cystic and salivary duct
11. How might a metastatic genitourinary tract tumor carcinomas) [1, 45] and tongue (squamous cell carcinoma)
present differently from primary breast carcinoma? [1] origins have also been reported. A rapidly growing lump
in the neck, generalized neck swelling, pain in the front of
Metastatic cancer from the genitourinary tract to the breast is the neck (sometimes going up to the ears), hoarseness, trou-
most likely to arise from the kidney, bladder, or prostate (men) ble swallowing or breathing, or a chronic cough may all be
[1, 2, 4, 11, 15, 16, 20]. Physical examination may reveal a signs of head and neck involvement.
mass on the side or lower back in renal carcinoma. Blood in
the urine and low back pain (not caused by injury) should raise 15. How might a metastatic sarcoma present differently
suspicion for renal, bladder, or prostate involvement. Changes from primary breast carcinoma or a primary breast
in urinary habits (i.e., urgency, increased frequency, or frank sarcoma?
inability to urinate) should prompt consideration for bladder
or prostate involvement. Prostate cancer is much more com- Any sarcoma in the adult breast is likely a malignant phyl-
mon in men than breast cancer and should always be consid- lodes tumor, sarcoma arising in the post-irradiation breast,
ered in any breast cancer diagnosis in males. or other primary breast sarcoma. Still, any sarcoma in the
breast should raise consideration for metastatic tumor to the
12. How might a metastatic gastrointestinal tract tumor breast, as primary breast sarcomas are rare [1] (although
present differently from primary breast carcinoma? metastatic sarcoma is even less common [46]). Clinical his-
tory and exam are essential. If the sarcoma is present in a
Metastatic cancer from the gastrointestinal tract to the breast pediatric patient, the literature would support a metastatic
is most likely to arise from the stomach [4, 11, 15, 20], colon origin until proven otherwise, as the proportion of meta-
(including carcinoid) [1, 5, 20], and small bowel (including static tumor to the breast in children and young adults is
carcinoid and melanoma) [2, 11], although cases arising significantly higher [2, 13, 15].
from the pancreas [1, 4], esophagus [2], liver (including car-
cinoid) [1, 20], appendix (carcinoid) [2, 23], and biliary tract 16. How might a metastatic hematopoietic tumor pres-
[20] have also been reported. Patients that have a new breast ent differently from primary breast carcinoma or a
cancer diagnosis with abdominal complaints, difficulty swal- primary breast hematopoietic tumor?
lowing (particularly with esophageal carcinoma), jaundice
(biliary tract or liver), abdominal ascites, or blood in the Any diagnosis of lymphoma in the breast should raise con-
stool should raise consideration for gastrointestinal sideration for secondary involvement. Primary breast lym-
involvement. phoma is rare, accounting for less than 1% of all patients
with non-Hodgkin lymphoma and approximately 1.7% of all

patients with extralymphatic non-Hodgkin lymphoma [47].
A primary diagnosis is even less common in men [48, 49].
The clinical criteria for primary breast lymphoma have been
previously defined by Wiseman and Liao [50]. The absence
of mammary tissue in close association with lymphomatous
infiltrates after adequate pathologic evaluation, evidence of
disseminated lymphoma other than simultaneous ipsilateral
lymph node involvement, or any prior diagnosis of lym-
phoma would all suggest secondary involvement.
17. What are the typical gross characteristics of meta-

static cancer to the breast versus a cancer of primary
breast origin?
Metastatic cancer to the breast is typically diagnosed by core

needle biopsy and not excised, and therefore, gross examina- Fig. 10.2 Primary breast melanoma (40×). This patient presented with
tion is often not done. Even still, there are no specific gross a breast lesion without any prior history of melanoma. Melanoma is
characteristics that would raise suspicion for a metastatic present in the dermis (black arrow). The presence of an in situ compo-
nent (white arrow) supports a melanoma of primary breast origin
cancer to the breast, except for the black pigmentation that
might be present in a metastatic melanoma [16].
if there is keratinization, and adenocarcinoma. Although a
18. How do metastatic melanomas histologically differ distinctive histology may not be evident with an adeno-
from primary breast carcinoma or a primary breast carcinoma (Fig. 10.3a, b), morphologic clues such as an
melanoma? acinar growth pattern or mucin-secreting columnar cells
may be present [2].
The wide range of morphologic appearances in malignant
melanoma makes it particularly difficult to recognize sec- 20. How do metastatic gynecological tract carcinomas
ondary breast melanoma as a metastatic disease. Histologic histologically differ from primary breast
clues that might raise suspicion for a melanoma as opposed carcinoma?
to a carcinoma include high-grade dyscohesive epithelioid
cells (although lobular carcinoma can similarly present; see Serous carcinoma, the most common type of metastatic
Case 4 at the end of this chapter), intranuclear inclusions, gynecological tract tumor to the breast [1, 2, 16, 42–44], will
spindled cells, and cytoplasmic pigment. Clinical examina- often present with a papillary architecture, thus providing a
tion would then be of primary importance in differentiating histologic clue for the consideration of metastatic disease. A
primary from secondary disease, as histological differentia- solid growth pattern may, however, be the dominant histol-
tion between the two would be unlikely, unless there is an in ogy, making the diagnosis less evident (Fig. 10.4; Case 2 at
situ component (Fig. 10.2), which would support a primary the end of this chapter). Similar to breast carcinoma, serous
breast origin. carcinoma may present with calcifications [1, 2]; however,
the calcifications are typically larger and of the psammoma-
19. How do metastatic pulmonary carcinomas histologi- tous type [1, 3], and that histology should prompt consider-
cally differ from primary breast carcinoma? ation for a metastatic gynecological tract carcinoma.
Metastatic endometrial serous [15], endometrial endometri-
If a breast cancer patient presents with a neuroendocrine oid adenocarcinoma [1, 4, 15], cervical (includes a carci-
histology, consideration should be given for a pulmonary noid) [1, 11, 15], and choriocarcinoma [1, 15] have also been
primary origin. A review of the literature suggests that reported. These may have “unusual” histologic patterns such
anywhere from 30% to 50% of primary lung tumors that as well-differentiated glands with pseudostratified tall
metastasize to the breast have neuroendocrine features, columnar cells, possibly with mucin (endometrioid adeno-
either atypical carcinoid, small cell neuroendocrine, or carcinoma), a neuroendocrine pattern (carcinoid), or a bipha-
large cell neuroendocrine [1, 2, 4, 15]. Other presenta- sic syncytial growth pattern resembling multinucleated and
tions include squamous cell, which should also bring up mononucleated trophoblastic cells, such as might be seen in
consideration for a primary pulmonary origin particularly a choriocarcinoma.
242 B. M. Turner
a b
Fig. 10.3 (a) Primary breast carcinoma involving the dermis (20×). with high-grade nuclei present in the dermis of the breast. No clinical
(b) Metastatic lung adenocarcinoma involving the dermis of the breast history was given on the requisition form. A high degree of suspicion
(40×). Note the similar histology in (a and b), including squamoid nests would be necessary for consideration of metastatic disease in this case
Fig. 10.5 Metastatic renal cell carcinoma to the breast (10×). The clear
Fig. 10.4 Metastatic ovarian carcinoma involving the breast (10×).
cell morphology (arrow) should prompt consideration for metastatic
This tumor has varying morphology, including both solid and papillary
disease; however, focal clear cell change can also be seen in breast car-
architectures. The papillary architecture (white arrows) is classic for
cinoma, so clinical history is of vital importance. (Courtesy of Dr.
serous ovarian carcinoma; however, a solid growth pattern may be the
David Hicks, University of Rochester, Rochester, NY)
dominant histology, as in this case (black arrows). Note that no calcifi-
cations were evident on microscopic review
(likely to be more patchy, if present at all), should prompt

21. How do metastatic genitourinary carcinomas histo- consideration for a metastatic disease. Both bladder and
logically differ from primary breast carcinoma? prostate cancers have histology that overlaps with primary
breast carcinoma [2, 15]. A diffuse transitional cell pattern
Metastatic cancer from the kidney, bladder, or prostate (men) may signal the possibility of bladder cancer and should
may have distinctive histologic patterns, creating suspicion prompt consideration for a thorough evaluation of the clini-
for a metastatic disease. Conventional renal cell carcinoma is cal history. A metastatic bladder cancer will likely present
the most common renal malignancy likely to metastasize to with symptoms or history consistent with primary bladder
a wide range of sites [2, 51]. Diffuse clear cell change carcinoma. Prostate cancer may have low-grade morphology
(Fig. 10.5), unlikely to be seen in primary breast carcinoma with columnar cells containing a nucleolus [2, 16]. Any male
with breast cancer should be considered to have metastatic liver, small bowel, and appendix) and lung primaries [1, 2, 4,
disease until proven otherwise. 11, 16, 23, 29, 32]. Recognizing the characteristic neuroen-
docrine histologic features of mitotically active sheets of
22. How do metastatic gastrointestinal tract carcinomas cells with scant cytoplasm and speckled chromatin, lack of
histologically differ from primary breast prominent nucleoli, and the possibly associated crush artifact
carcinoma? or necrosis are the first steps in considering the presence of a
neuroendocrine tumor as opposed to a carcinoma (Figs. 10.7,
Metastatic cancer from the gastrointestinal tract may have an and 10.8). Histologic clues differentiating a primary and
intestinal or signet ring pattern (particularly stomach), caus- metastatic neuroendocrine tumor are not likely to be appar-
ing confusion with either ductal or lobular carcinoma, ent, with the exception of recognizing the presence of ductal
respectively. Columnar mucin–secreting cells favor a gastro-
intestinal origin [2] as opposed to primary breast ductal car-
cinoma (Fig. 10.6a, b). Signet ring cells with diffuse foamy
cytoplasm are more common in gastrointestinal carcinoma,
while signet ring cells with distinct vacuoles with a central
mucin dot would be more common in primary lobular carci-
noma [16]. A hepatoid pattern consisting of sheets and cords
of polygonal cells with abundant eosinophilic cytoplasm
might be a useful clue in considering a hepatic metastasis. A
thorough clinical history and exam would be particularly
critical in detecting a metastatic tumor from the pancreas,
esophagus, or biliary tract.
23. How do metastatic neuroendocrine tumors histologi-

cally differ from primary breast carcinoma or a pri-
mary breast neuroendocrine tumor?
Although neuroendocrine differentiation can be seen in up to

30% of invasive breast carcinomas (most commonly associ-
ated with mucinous and solid papillary carcinomas), primary Fig. 10.7 Metastatic neuroendocrine carcinoma (40×). Classic high-
grade neuroendocrine features are present including cohesive nests of
neuroendocrine tumors of the breast are at best rare [52], and
tumor cells with a high nuclear–cytoplasmic ratio, nuclear molding
their existence is controversial [16]. The most common (black arrows), and numerous mitoses (white arrows). The nuclear
metastases occur from gastrointestinal (including colon, chromatin is finely stippled, with inconspicuous nucleoli
a b
Fig. 10.6 (a) Primary breast carcinoma (20×). (b) Metastatic gastric adenocarcinoma involving the breast (20×). Both carcinomas are relatively
well differentiated. The mucin secreting cells (b, arrow) are a clue to the possibility of metastatic disease. Also see Fig. 10.20.
244 B. M. Turner
noma or classic phyllodes morphology (i.e., leaf-like areas)

is critical. The most common sarcoma metastatic to the
breast in adults is a uterine leiomyosarcoma [1]. A number of
other subtypes have been reported, including rhabdomyosar-
coma [1, 13, 15] (the most common pediatric sarcoma meta-
static to the breast), liposarcoma [1, 15], non-uterine sarcoma
[1, 2], Ewing sarcoma [1], malignant fibrous histiocytoma
(pleomorphic sarcoma) [1, 15], angiosarcoma [1, 15],
synovial sarcoma [1], dendritic cell sarcoma [1], and myxo-
fibrosarcoma [1].
26. How do metastatic hematopoietic tumors histologi-

cally differ from primary breast carcinoma or pri-
mary breast hematopoietic tumors?
As with sarcomas, the most common pitfall would be mis-

Fig. 10.8 Metastatic neuroendocrine carcinoma with crush artifact taking a hematopoietic tumor for a carcinoma, so recognition
(20×). Crush artifact (black arrow), which results in artificial elongation of characteristic morphology is again critical (Fig. 10.9). The
and distortion of cells with subsequent spillage of cytoplasmic contents most common type of hematopoietic tumors in the breast,
into the stroma, is present. Crush artifact in association with more clas-
primary or secondary, is a lymphoma, most often a non-
sic neuroendocrine morphology (white arrow) should prompt consider-
ation for a neuroendocrine tumor Hodgkin diffuse large B-cell lymphoma [2, 28, 47, 53]. Non-
Hodgkin T cell lymphoma, Hodgkin lymphoma, and Sézary
syndrome have also been reported [15, 46, 54]. These tumors
carcinoma in situ (DCIS), which would favor a primary are typically high grade. In non-Hodgkin lymphoma, the
breast origin. malignant cells are most commonly dyscohesive and centro-
blastic, less often immunoblastic. Hodgkin lymphoma may
24. How do metastatic head and neck carcinomas histo- offer histologic clues, including Reed–Sternberg cells with
logically differ from primary breast carcinoma? abundant basophilic or amphophilic cytoplasm and binucle-
ate or bilobed nucleus, with multilobate or large inclusion-
Thyroid carcinoma may have a variety of patterns including like eosinophilic nucleoli. A rich inflammatory background
papillary, follicular, insular (medullary), tall cell, diffuse may be present. Leukemia and multiple myeloma have also
sclerosing type, and papillary with Hashimoto’s thyroiditis. been reported in the breast [47]. Leukemia may present with
All of these patterns are unusual for primary breast carci- blasts, and myeloma may present with lobulocentric cells
noma and should raise suspicion for the possibility of a met- that have plasmacytic morphology, which should prompt
astatic focus. The presence of colloid may also be a useful consideration for a hematopoietic origin.
clue. Salivary duct carcinoma may present a challenge as
there is significant histological and immunophenotype over- 27. Can immunohistochemistry be helpful in distin-
lap. A diffuse invasive (i.e., lack of myoepithelial cells) guishing primary breast carcinoma from an extra-
“DCIS pattern” with associated necrosis should prompt an mammary malignancy metastatic to the breast?
evaluation of any abnormalities on the head and neck
examination. The use of a broad panel of antibodies can be helpful in dis-
tinguishing primary breast carcinoma from an extramam-
25. How do metastatic sarcomas histologically differ mary malignancy metastatic to the breast. Table 10.3 outlines
from primary breast carcinoma or a primary breast immunohistochemical panels that may be helpful in charac-
sarcoma? terizing a primary breast carcinoma from a metastatic tumor
to the breast. Breast carcinomas are typically cytokeratin
The most common pitfall would be mistaking a sarcoma for positive. Exceptions include CK20 and CK5, although posi-
a carcinoma, so recognition of characteristic morphology is tive staining for CK5 can be seen in the basal subtype, myo-
critical. A spindled cell or mesenchymal morphology should epithelial carcinoma, and sarcomatoid carcinoma
prompt consideration for a sarcoma. Differentiating a pri- (carcinosarcoma, spindle cell carcinoma, and metaplastic
mary from a secondary sarcoma is more challenging. carcinoma) [55]. p63 will be negative with rare exceptions,
Sarcoma in the breast is most commonly associated with a including myoepithelial carcinoma (occasionally) and sarco-
component of metaplastic carcinoma or phyllodes tumor, so matoid carcinoma (Fig. 10.10a, b) [55]. Gross cystic disease
a thorough sampling looking for areas of conventional carci- fluid protein fraction-15 (GCDFP-15) has a fairly high speci-
ficity (although low sensitivity) for breast carcinoma,

although it should be remembered that salivary duct carci-
noma will also stain positive for this marker. Mammaglobin
A may also be useful, although the sensitivity is variable and
the specificity for breast has not been firmly established [55].
Positive staining for ER and PR can be extremely helpful in
making a diagnosis of breast carcinoma, because with the
exception of gynecological (ovary and endometrial) carcino-
mas, most malignancies will be negative for ER and PR. It
must be remembered, however, that ER and rarely PR stain-
ing has been reported in carcinomas of the lung, stomach,
and thyroid [55]. GATA-3 is sensitive but not entirely spe-
cific for breast carcinoma, as it will often stain positive for
urothelial carcinoma, and has also been reported to have
more than infrequent positive staining in squamous cell car-
cinomas, mesotheliomas, salivary gland carcinomas, chorio-
Fig. 10.9 B-cell lymphoma involving the breast (20×). This tumor pre- carcinomas, chromophobe renal cell carcinomas, and
sented in the right axilla as a mass and was initially thought to be a pancreatic adenocarcinomas [56]. However, GATA-3 has
metastatic breast or uterine primary (a history of uterine carcinoma was infrequent labeling (<10%) in other common types of adeno-
given.) Note the “nests” of cells with epithelioid morphology. A carci-
carcinomas including those of pulmonary, gastrointestinal,
noma workup revealed negative immunohistochemistry for GATA-3
and PAX-8. Flow cytometry was negative (as it can be in 20–30% of and gynecologic origin [56]. A more detailed discussion of
lymphomas); however, a closer review of the histomorphology reveals the immunohistochemical profile of the more common extra-
the somewhat dyscohesive nature of these “nests” of epithelioid cells. mammary malignancies metastasizing to the breast can be
Immunohistochemistry for CD45 was strongly positive, supporting the
found in questions 28–36.
diagnosis of lymphoma. Also see Fig. 10.21
Table 10.3 Immunohistochemical profiles of breast carcinoma and the most common non-breast tumors metastasizing to the breast
Immunohistochemistry
Type of malignancy CK7 CK5 CK20 p63 GATA-3 TTF-1 CDX-2 CD45 S-100 Melan-A
Breast La Ub,d U U L U U U Oe U
Melanoma U U U U U U U U Lf L
Pulmonary
Adenocarcinoma L U U U U L U U U U
Squamous U L U L Oc U U U U U
Gynecological L U U U Ug U U U U U
Genitourinary
Urothelial L U L U L U U U U U
Chromophobe RCC L U U U L U U U U U
Clear cell RCC U U U U U U U U U U
Prostate U U U U U U U U U U
Gastrointestinal Uh U Li U U U L U U U
Neuroendocrine U U U U U Lj Lk U U U
Head and neck
Thyroid U U U U U L U U U U
Salivary duct L U U U L U U U L U
Sarcoma U U U U O U U U L U
Hematopoietic U U U U Ul U U L U U
a
L Likely positive
b
U Unlikely to be positive
c
O Occasionally positive
d
Likely weak scattered
e
Strong diffuse
f
May be positive in the basal subtype, myoepithelial carcinoma, and sarcomatoid carcinoma (carcinosarcoma, spindle cell carcinoma, and meta-
plastic carcinoma)
g
Expressed in a majority of trophoblastic (choriocarcinoma) and yolk sac tumors
h
Variable patterns in gastric carcinoma; occasionally will stain positive in pancreatic carcinoma
i
Variable patterns with gastric carcinoma; negative in hepatocellular carcinoma
j
Positive staining observed in neuroendocrine tumors of pulmonary origins
k
Positive staining observed in neuroendocrine tumors of gastrointestinal origins
l
Positive staining has been reported in Hodgkin lymphoma
246 B. M. Turner
a b
Fig. 10.10 (a) High-grade primary breast carcinoma with metaplastic cinoma, and metaplastic carcinoma). This high-grade tumor had
features (20×). (b) p63 immunohistochemistry in high-grade primary metaplastic features with scattered p63 staining in the invasive compo-
breast carcinoma with metaplastic features (20×). Although p63 will nent (b, black arrow). Note the adjacent DCIS, with the expected posi-
likely be negative in invasive breast carcinoma, focal positive staining tive p63 staining in the myoepithelial cell layer (white arrow in a and
can be seen in some high-grade invasive carcinomas, myoepithelial car- b). See also Fig. 10.22a, b
cinoma, and sarcomatoid carcinoma (carcinosarcoma, spindle cell car-
a b
Fig. 10.11 (a) Pleomorphic lobular primary breast carcinoma (40×). tumor (a) in comparison to the metastatic melanoma in Fig. 10.12a.
(b) S-100 immunohistochemistry of pleomorphic lobular primary S-100 staining (b), which will likely be more focal and weak in primary
breast carcinoma (40×). Note the similar high-grade appearance of this breast carcinoma, supports a diagnosis of a primary breast origin
28. How is metastatic melanoma immunophenotypically Case 4 at the end of this chapter), as opposed to more diffuse
differentiated from primary breast carcinoma or a stronger staining in melanoma (Fig. 10.12a, b, Case 4 at the
primary breast melanoma? end of this chapter). Melanoma can show aberrant expres-
sion for cytokeratins, particularly CAM 5.2, and epithelial
Expression of GATA-3, ER, PR, GCDFP-15, and mamma- membrane antigen (EMA) [2, 57], both of which are also
globin would favor a primary breast carcinoma. Breast carci- typically positive in breast carcinoma (Case 5 at the end of
nomas are not likely to stain for melanoma markers this chapter). As such, these markers should not be used if
(Melan-A, HMB-45, microopthalmia transcription factor), melanoma is being considered. CK7, which is typically posi-
with the exception of S-100, although S-100 will likely be tive in breast carcinoma and unlikely to be positive in mela-
more focal and weak in breast carcinoma (Fig. 10.11a, b, noma, would be the cytokeratin of choice. Clinical history
a b
Fig. 10.12 (a) Metastatic melanoma to the breast (40×). (b) S-100 dyscohesive epithelioid cells (a) with similar characteristics to a high-
immunohistochemistry of metastatic melanoma to the breast (40×). grade pleomorphic lobular carcinoma (Fig. 10.11a). Strong diffuse stain-
Compare (a) to Fig. 10.11a. Metastatic melanomas may have high-grade ing with S-100 (b) supports a diagnosis of metastatic melanoma
would be essential for differentiating a primary breast mela- staining may be seen in approximately 5% of gynecologic
noma from a metastatic melanoma. carcinomas (endometrial) [16] and in a majority of tropho-
blastic (choriocarcinoma) and yolk sac tumors [60]. Most
29. How is metastatic pulmonary carcinoma immuno- cytokeratins have a similar expression in breast and gyneco-
phenotypically differentiated from primary breast logical tract adenocarcinomas, although EMA has been
carcinoma? reported to have a different expression pattern in breast
micropapillary carcinoma (Fig. 10.15a, b; expression only
Expression of GATA-3, ER, PR, GCDFP-15, and mammaglo- on the outside of the papillary clusters) compared to serous
bin would favor a primary breast carcinoma. Less than 5% of carcinoma of the ovary (Fig. 10.16a, b; expression on the
breast carcinomas are positive for TTF-1 (Fig. 10.13a, b) or outside and in central spaces [2]; Case 2 at the end of this
Napsin-A (Fig. 10.13a, c) [57, 58]. TTF-1 (Fig. 10.14a, b) and chapter). PAX-8 has been found to be a useful marker, as it
Napsin-A (Fig. 10.14a, c) are positive in approximately 75% will show positive staining in most gynecologic cancers and
[2] and 85% [57, 58] of pulmonary adenocarcinomas, respec- has not been found to be positive in breast carcinoma [61–
tively. Primary pulmonary large cell carcinoma may be less 63]. WT-1 may be positive in up to 85% of ovarian serous
likely to express TTF-1 [59] and may be difficult to distin- carcinomas (Figs. 10.16a and 10.18, Case 2 at the end of this
guish from a poorly differential breast carcinoma. A primary chapter), but is rarely positive in breast carcinoma
squamous carcinoma of the lung would be difficult to differen- (Fig. 10.17a, b), although more than frequent focal weak
tiate from a primary breast squamous cell carcinoma, as both staining has been reported in mucinous breast carcinoma [1,
will stain positive for p63 and CK5, and possibly GATA-3 2, 16, 62–65]. ER and PR are expressed in both breast and
[60]. Primary squamous cell carcinoma of the breast is much gynecologic tumors of the ovary and endometrium, and are
less common, making the clinical history essential. Similar of limited value in differentiating a primary breast carcinoma
difficulty is experienced in differentiating primary breast small from a metastatic gynecological tract carcinoma (Fig. 10.18).
cell carcinoma from a metastasis. About 80% of small cell
carcinomas are positive for TTF-1, regardless of the primary 31. How is metastatic genitourinary tract carcinoma
location, again making the clinical history essential. GATA-3- immunophenotypically differentiated from primary
positive staining has also been reported in mesothelioma [60]. breast carcinoma?
30. How is metastatic gynecological tract carcinoma Expression of ER, PR, GCDFP-15, and mammaglobin would
immunophenotypically differentiated from primary favor a primary breast carcinoma. GATA-3 expression can be
breast carcinoma? seen in a majority of breast, urothelial, and chromophobe
renal cell carcinomas [56, 60], so these genitourinary tract
Expression of GATA-3, GCDFP-15, and mammaglobin carcinomas must be considered in any GATA-3-positive
would favor a primary breast carcinoma, although GATA-3 staining breast carcinoma with unusual morphology (i.e.,
248 B. M. Turner
a a
b b
c c
Fig. 10.13 (a) Primary breast carcinoma (20×). (b) TTF-1 immuno- Fig. 10.14 (a) Metastatic lung adenocarcinoma involving the breast
histochemistry of primary breast carcinoma (20×). Negative staining (20×). (b) TTF-1 immunohistochemistry of metastatic lung adenocarci-
for TTF-1 in primary breast carcinoma. Note the background cytoplas- noma involving the breast (20×). Note the positive nuclear TTF-1 stain-
mic blush. Do not mistake this for a positive TTF-1, which must show ing. (c) Napsin-A immunohistochemistry of metastatic lung
positive nuclear staining. See also Fig. 10.14b, c. Napsin-A immuno- adenocarcinoma involving the breast (20×). Note the diffuse strong
histochemistry of primary breast carcinoma (20×). (c) Negative stain- cytoplasmic and membranous staining with Napsin-A
ing for Napsin-A in primary breast carcinoma
a b
Fig. 10.15 (a) Primary breast carcinoma, micropapillary type (20×). (b) EMA immunohistochemistry of breast carcinoma, micropapillary type
(40×). Note the more distinct expression of EMA (b) on the outside of the papillary clusters. Compare to Fig. 10.16b
a b
Fig. 10.16 (a) Metastatic ovarian serous carcinoma involving the breast (20×). (b) EMA immunohistochemistry of metastatic ovarian serous
carcinoma involving the breast (20×). Note the strong EMA expression on the outside and in central spaces. Compare to Fig. 10.15b
a b
Fig. 10.17 (a) Primary breast carcinoma (20×). (b) WT-1 immunohistochemistry of primary breast carcinoma (20×). Negative staining for
WT-1 in primary breast carcinoma. Do not be fooled by the positive staining in the associated vascular structures
250 B. M. Turner
transitional, large polygonal cells with transparent or slightly cancer that express PSA [1, 16, 66–68], so both markers
reticulated cytoplasm with a prominent cell membrane). should be done if ruling out metastatic prostate carcinoma.
CK20 will likely be positive in urothelial carcinoma, and a Prostate carcinoma is CK7 and CK20 negative.
diffuse cytoplasmic staining reaction with Hale’s colloidal
iron will likely be seen with chromophobe renal cell carci- 32. How is metastatic gastrointestinal tract carcinoma
noma [55]. Clear cell renal carcinoma may express positive immunophenotypically differentiated from primary
staining for CD10, RCC antibody, and vimentin [2, 55], breast carcinoma?
which are unlikely to be expressed in a primary breast carci-
noma [2]. Prostate carcinoma will express both prostate- Expression of GATA-3, ER, PR, GCDFP-15, and mamma-
specific antigen (PSA) and prostatic acid phosphatase (PAP) globin would favor a primary breast carcinoma. CDX-2 and
in nearly 100% of tumors [1]. Although PAP is not expressed CK20 are not typically expressed in breast cancer
in breast carcinoma, there have been reports of male breast (Fig. 10.19a, b) but they are expressed in a majority of gas-
trointestinal carcinomas (Fig. 10.20) [2, 16, 69]. Colorectal
carcinoma is typically CK7 negative. A CK7-positive/CK20-
negative pattern can be seen in 25% of gastric carcinomas
[55], so additional immunohistochemistry should be done in
cases of suspected metastatic gastric carcinoma. CK17 stains
the majority of pancreatobiliary adenocarcinomas diffusely,
but only rarely stains breast carcinoma, and in those breast
carcinomas that do stain positive, staining usually only
occurs within the centers of solid high-grade breast carci-
noma nests [55]. Hepatocellular carcinoma is typically CK7
negative, CK20 negative, and Hepar-1 positive (negative in
breast carcinoma).
33. How is a metastatic neuroendocrine tumor immuno-

phenotypically differentiated from primary breast
carcinoma or a primary breast neuroendocrine
tumor?
Fig. 10.18 WT-1 immunohistochemistry of metastatic ovarian serous
carcinoma involving the breast (40×). Strong WT-1 expression support- Expression of GATA-3, ER, PR, GCDFP-15, and mamma-
ing a diagnosis of metastatic ovarian serous carcinoma globin would favor a primary breast carcinoma. ER, PR, and
a b
Fig. 10.19 (a) Primary breast carcinoma (10×). (b) CDX-2 immuno- blush may be seen; however, do not mistake this for a positive CDX-2,
histochemistry of primary breast carcinoma (10×). Negative staining which must show positive nuclear staining. Also see Figs. 10.20 and
for CDX-2 in primary breast carcinoma. A background cytoplasmic 10.22a, b
Fig. 10.20 CDX-2 immunohistochemistry of Fig. 10.6b: metastatic Fig. 10.21 CD45 immunohistochemistry of B-cell lymphoma involv-
gastric adenocarcinoma to the breast (20×). Positive nuclear staining for ing the breast (40×). Strong positive staining for CD45 supports the
CDX-2 in metastatic gastric adenocarcinoma involving the breast. See diagnosis of lymphoma. See also Fig. 10.9
also Fig. 10.19a, b
will stain positive in sarcomas; however, a metastatic sar-

GCDFP-15 are often expressed by breast neuroendocrine coma will have a similar immunophenotype as a primary
carcinomas [2]. Of note, PR is expressed in some pancreatic breast sarcoma. ER and PR may show scattered positive
endocrine tumors [2]. CK7 is typically negative in neuroen- staining in primary breast sarcomas, but should still prompt
docrine tumors but is positive in breast carcinoma. Positive a thorough search for a gynecological origin. GATA-3-
staining for CDX-2 and CK20 would favor a gastrointestinal positive staining has also been reported in sarcomas [60].
origin (both carcinoma and neuroendocrine tumor) [2]. Clinical history is most essential in differentiating a primary
Positive staining for TTF-1 would favor a pulmonary origin breast sarcoma from a metastatic sarcoma.
(both carcinoma and neuroendocrine tumor) [2]. The clinical
history would be essential if CDX-2, CK20, TTF-1, ER, PR, 36. How are metastatic hematopoietic tumors immuno-
and GCDFP-15 were negative. phenotypically differentiated from primary breast
carcinoma or a primary breast hematopoietic tumor?
34. How is metastatic head and neck carcinoma immun-
ophenotypically differentiated from primary breast Although GATA-3 staining has been reported in Hodgkin
carcinoma? lymphoma [71], the immunophenotype of hematopoietic
tumors is otherwise distinctly different from breast carci-
Expression of GATA-3, ER, PR, GCDFP-15, and mammaglo- noma. A CD45-positive tumor essentially rules out a breast
bin would favor a primary breast carcinoma; however, caution carcinoma (Fig. 10.21). Identifying light chain restriction in
must be used when suspicion arises for a metastatic salivary the neoplastic plasmacytic cells of myeloma can be helpful;
duct carcinoma, which can be positive for Her-2, GCDFP-15, however, clinical history and criteria [50] are most essential
and GATA-3 [45, 60]. Salivary duct carcinoma usually in further differentiating a primary breast hematopoietic
expresses androgen receptor (AR); however, AR can also be tumor from a metastatic hematopoietic tumor.
present in primary breast carcinoma, although in a lesser pro-
portion of cases [70]. TTF-1 and thyroglobulin can be helpful 37. Can molecular profiling be helpful in distinguishing
in distinguishing metastatic thyroid carcinoma (typically posi- primary breast carcinoma from an extramammary
tive) from primary breast carcinoma (typically negative). malignancy metastatic to the breast?
35. How is metastatic sarcoma immunophenotypically Although the diagnostic workup of any cancer of unknown
differentiated from primary breast carcinoma or primary strongly depends on clinical history, gross and mor-
primary breast sarcoma? phologic examination, as well as immunohistochemical phe-
notype, molecular profiling offers a promising diagnostic
Sarcoma will typically be negative for cytokeratin stains and technique to determine the tissue of origin in patients with
carcinomas will typically be negative for vimentin, which carcinoma of unknown primary site [72]. The clinical history
252 B. M. Turner
and presentation may bias the diagnostic workup and may other main cancer class with <5% probability, which has
influence the choice of immunohistochemistry panel. been ruled out [75, 76].
Immunohistochemistry interpretation is subject to interob- miRview mets2 was initially validated using a
server and intraobserver variability, and in 30% of the cases, 64-microarray assay on a reference database of 1282 primary
the immunohistochemical staining pattern does not result in and metastatic tumor samples of 42 tumor types, grouped
a conclusive diagnosis [73, 74]. into 12 different sites [73, 75], most of which have been
At least three molecular profiling tests are commercially reported to metastasize to the breast. These 12 sites include
available: (1) Tissue of Origin® (TOO) test (Cancer Genetics, breast, skin (including melanoma, squamous cell), pulmo-
Inc., Rutherford, NJ, USA) [74, 75], (2) bioTheranostics nary (non-small cell, neuroendocrine, and mesothelioma),
Cancer Type ID (CTID) (Biotheranostics, San Diego, CA, gynecological (ovary, cervix,), gastrointestinal (biliary,
USA) [75, 76], and (3) miRview® mets2 (Rosetta Genomics, esophageal, gastric, hepatocellular, neuroendocrine, pancre-
Princeton, NJ, USA) [73, 75]. The principle underlying these atic), genitourinary (kidney, prostate, bladder, testis [germ
molecular profiling tests is that different tissue types have cell]), head and neck (thyroid), soft tissue sarcoma (various
distinct RNA profiles. Each test was developed using gene subtypes) hematopoietic (non-Hodgkin lymphoma), thymus,
expression profiles [74, 75] from hundreds of different adrenal, and brain (various subtypes) [73]. Each test miRNA
tumors. For each test, subsets of discriminatory genes were profile is subjected to two classification algorithms, which
identified, and diagnostic algorithms were built for cancer assign a tissue of origin based on the normalized expression
classification. All three tests can use formalin-fixed, paraffin- of the 64 microRNAs to determine the tissue of origin [73,
embedded (FFPE) tissue or cytology specimens [73, 75, 76]. 75]. The two classification algorithms predict one of the 42
The cost is approximately $3000–$4000 per test [75]. tumor types, and predict one of the following seven tumor
The TOO test was initially validated using a 2000-gene classifications [73, 75]: (1) pulmonary (small-cell carcinoma
classification model on 462 metastatic, poorly differenti- or carcinoid); (2) gastrointestinal (adenocarcinoma of biliary
ated, or undifferentiated FFPE tumor specimens [74], tract or pancreas); (3) genitourinary (testicular germ cell
grouped into 15 classifications [74, 75] in 9 different sites, tumor, seminomatous or nonseminomatous); (4) genitouri-
all of which have been reported to metastasize to the nary (renal cell carcinoma); (5) head and neck (thyroid car-
breast. These 9 sites include breast, skin (melanoma), pul- cinoma, follicular or papillary); (6) soft tissue sarcoma
monary (non- small-cell lung), gynecological (ovarian), (Ewing sarcoma, chondrosarcoma, malignant fibrous histio-
genitourinary (bladder, kidney, prostate, testis [germ cell]), cytoma or fibrosarcoma, osteosarcoma, rhabdomyosarcoma,
gastrointestinal (colorectal, gastric, hepatocellular, pan- synovial sarcoma, liposarcoma); and (7) brain (astrocytic or
creatic), head and neck (thyroid), soft tissue sarcoma, and oligodendroglial tumor) [73]. The two classifier predictions
hematopoietic (non-Hodgkin lymphoma) [74, 75]. A simi- are then combined into a single predicted tissue of origin or
larity score (SS) is reported. The higher the SS, the more two different predictions [73]. If the two classifier predic-
likely the diagnosis is. As the SS falls, agreement declines tions agree with a high level of confidence, a single predicted
until a SS less than 5 rules out that tumor classification tissue of origin is reported. When the two classifier predic-
with >99% confidence [74, 75]. Most reports provide one tions have high degree of certainty regarding the tumor class
highly likely diagnosis and rule out at least 12 tumor clas- (e.g., sarcoma), but a low degree of certainty regarding the
sifications [75]. specific tumor type (e.g., which type of sarcoma), a single
CTID was initially validated using a 92-gene assay on a predicted tissue of origin is reported in addition to one of the
reference tumor database containing 2206 tumors [75, 76] of additional seven tumor classifications [73]. When two pre-
over 100 different histological subtypes [76], grouped into dictions are reported, they are ordered by the likelihood as
12 different sites, most of which have been reported to estimated by the positive predictive value of each of the
metastasize to the breast. These 12 sites include breast, skin answers. When both classifiers exhibit very low confidence
(including melanoma, squamous cell, basal cell, and Merkel in their result, the assay does not generate a result and reports
cell), pulmonary (non-small cell and neuroendocrine), gyne- that the microRNA expression pattern of the sample does not
cological (ovary, cervix, endometrium, germ cell), gastroin- match any of the expression patterns in the panel closely
testinal (biliary, esophageal, gastric, colon, small intestine, enough [73].
hepatocellular, neuroendocrine, pancreatic), genitourinary Overall, all three tests have a high specificity of ≥99%,
(kidney, prostate, bladder, testis[germ cell]), head and neck and their sensitivity in tumors of known origin ranges from
(thyroid, salivary gland), soft tissue sarcoma (various sub- 72% (CTID) to 95% (TOO) [75]. Of significance, sensitivity
types), hematopoietic (Hodgkin lymphoma and non-Hodgkin is often but not always lower in metastases than in primary
lymphoma), thymus, adrenal, and brain (various subtypes) tumors [75, 77], and profiling may not be helpful in necrotic
[76]. CTID reports one main cancer class (with its probabil- tumors [75]. Other molecular profiling approaches to identi-
ity), any main cancer class with >5% probability, and any fying cancers of unknown primary have been described.
Some are broad-based assays, classifying all likely tumors, significantly better survival was observed in patients who
while others are more specific, classifying tumor subsets. had no evidence of other disease at the time of diagnosis,
Like the previously discussed commercial tests, most of patients with neuroendocrine tumors, and patients who
these assays involve pre-specified gene, and have sensitivi- underwent surgical resection for the metastatic tumor. The
ties of 78–90% [75]. There is a clear and promising role for average interval between an extramammary tumor diagnosis
molecular profiling in the diagnostic workup of extramam- and development of metastatic breast disease is approxi-
mary tumors metastatic to the breast. The three tests dis- mately 2–3 years [4, 13, 39].
cussed all have the ability to diagnose the most common
origins of metastatic tumors to the breast; however, molecu-
lar profiling requires further validation including comparing Case Presentations
the performance of molecular profiling with immunohisto-
chemistry in independent tumor sets containing metastases Case 1
from both known primaries and cancers of unknown prima- An 88-year-old woman with a history of bilateral mastectomy
ries. The cost and logistics of integrating molecular profiling for bilateral invasive ductal carcinoma diagnosed 3 years prior
must also be considered relative to available therapies and and a history of high-grade urothelial carcinoma diagnosed
patient outcomes. 1 year prior presented with a left axillary tail mass confirmed
on ultrasound. A core biopsy was done. The history of high-
38. What is the prognosis for metastatic breast carci- grade urothelial carcinoma was provided to the pathologist.
noma from an extramammary location? The core biopsy found extensively necrotic poorly differenti-
ated carcinoma (Fig. 10.22a), suggestive of metastasis from
For patients with a known previous primary origin, the time urinary bladder, with positive immunohistochemistry for p63
from initial diagnosis to metastasis to the breast varies from (Fig. 10.22b). The tumor was additionally negative for AE1/
between 1 month and 22 years [1–3, 5, 13, 23, 30, 31], with AE3, cytokeratin 7, and ER (the previous breast carcinoma
most patients dying within a year of diagnosis [2, 17–19]. was ER positive). Given this immunophenotype and the
Longer intervals are more likely to be associated with malig- important provided history of urothelial carcinoma, a diagno-
nant melanoma and ovarian carcinoma [2, 13]. Prognosis for sis of metastatic urothelial carcinoma is supported. The patient
patients with metastases to the breast from an unknown died 3 months after the diagnosis of breast metastasis.
extramammary location is poor, with a median time having
been reported as 10–15 months in the two largest series Case 2
found to date [1, 3]. Other studies have shown similar out- A 47-year-old woman presented with a symptomatic cutane-
comes [1, 5, 18, 20]. In the largest series found to date, a ous lesion on the right breast. The patient had a history of
a b
Fig. 10.22 Case 1. (a) Metastatic urothelial carcinoma involving the can be rarely seen in high-grade breast carcinomas (Fig. 10.10b). The
breast (10×). The tumor has extensive necrosis with only focal areas of diffuse p63 staining in this case should prompt consideration for a met-
viable appearing tumor, making it difficult to assess the morphology. astatic focus. The tumor was negative for CK7 and AE1/AE3, support-
(b) p63 immunohistochemistry of metastatic urothelial carcinoma ing a primary urothelial metastasis
involving the breast (10×). The tumor was positive for p63. Focal p63
254 B. M. Turner
ovarian carcinoma diagnosed 2 years prior. The history of physician in the right axilla. The patient had overall felt well
ovarian carcinoma was provided to the pathologist. The over the past few months with no current symptoms, except
breast mass was excised and pathology showed a high-grade an occasional cough, which she attributed to postnasal drip.
tumor with both solid and papillary architectures (Figs. 10.4 The breast lump was verified mammographically and on
and 10.16a), suggestive of metastatic ovarian carcinoma. ultrasound. The axillary lymph node was verified on ultra-
The tumor was positive for WT-1 (Fig. 10.18), with strong sound. A core biopsy was done. The history of melanoma
EMA expression on the outside and in central spaces was provided to the pathologist. The core biopsy found inva-
(Fig. 10.16b), supporting metastatic ovarian carcinoma. The sive pleomorphic lobular carcinoma in the left breast
patient had widespread disease at the time of presentation (Fig. 10.11a), with negative immunohistochemistry for
and died 3 months after the diagnosis of breast metastasis. S-100 (Fig. 10.11b). The lymph node was biopsied and veri-
fied as melanoma (Fig. 10.12a), after consultation with der-
Case 3 matopathology and positive immunohistochemistry for
An 86-year-old man with a history of stage IV right lung S-100 (Fig. 10.12b). The patient underwent partial left mas-
adenocarcinoma diagnosed 2 years prior presented with a tectomy and left sentinel lymph node biopsy with adequate
right axillary mass diagnosed on PET scan. A core biopsy margins, and a right axillary dissection with metastatic mela-
was done. The history of stage IV lung adenocarcinoma was noma in one of the eleven lymph nodes. The patient was
provided to the pathologist. The core biopsy was reported as treated with radiation and briefly took hormone therapy. The
metastatic adenocarcinoma consistent with lung primary patient is still alive 5 years after the diagnosis of breast
(Fig. 10.14), supported by positive immunohistochemistry metastasis, with no evidence of recurrent breast cancer or
for TTF-1 (Fig. 10.14b) and Napsin-A (Fig. 10.14c). melanoma at the time of this writing.
Although the patient has widespread bone metastasis with
persistent malignant pleural effusion, the patient is still alive Case 5
seven months after the diagnosis of breast metastasis, at the An 82-year-old woman with a history of excised melanoma
time of this writing. three and a half years prior presented for a screening mam-
mogram, which identified new right axillary lymphadenopa-
Case 4 thy confirmed by ultrasound, which also confirmed a breast
A 68-year-old woman with a history of excised melanoma mass at 9:00 o’clock. The history of melanoma was not pro-
3 years prior presented to her physician with a left breast vided to the pathologist. The core biopsy showed predomi-
lump. A non-tender lymph node was also palpated by her nantly necrosis with a rim of viable tumor (Fig. 10.23a).
a b
Fig. 10.23 Case 5. (a) Metastatic melanoma to the breast (10×). (b) (20×). Melanoma can show aberrant expression for cytokeratins, particu-
“Keratin cocktail” immunohistochemistry of metastatic melanoma to the larly CAM 5.2, and epithelial membrane antigen (EMA). As such, these
breast (20×). Note the weak staining with a “keratin cocktail”. (c) S-100 markers should not be used if melanoma is being considered, and “keratin
immunohistochemistry of metastatic melanoma to the breast (20×). (d) cocktails” should also be avoided. Strong diffuse staining with S-100 (c)
Melan-A immunohistochemistry of metastatic melanoma to the breast and Melan-A (d) supports a diagnosis of metastatic melanoma
c d
Immunohistochemistry for a “keratin cocktail” was positive; 6. Abrams HL, Spiro R, Goldstein N. Metastases in carcinoma.
however, the pathologist astutely noted that it did not show Cancer. 1950;3:74–84.
7. McIntosh IH, Hooper AA, Millis RR, Greening WP. Metastatic car-
the diffuse strong staining pattern expected for a breast carcinoma within the breast. Clin Oncol. 1976;2:393–401.
cinoma (Fig. 10.23b). A review of the medical records 8. Alvarado CI, Carrera AM, Perez MD, Tavassoli FA. Metastases to
revealed a history of malignant melanoma. Suspicion for the breast. Eur J Surg Oncol. 2003;29:854–5.
metastasis led to an immunohistochemical workup which 9. Tempfer CB, Fizazi NE, Ergonenc H, Solass W. Metastasis of ovar-
ian cancer to the breast: a report of two cases and a review of the
was positive for S-100 (Fig. 10.23c) and Melan-A literature. Oncol Lett. 2016;11:4008–12.
(Fig. 10.23d), supporting a diagnosis of metastatic mela- 10. Sandison AT. Metastatic tumours in the breast. Br J Surg.

noma. The patient developed shortness of breath and a pleu- 1959;47:54–8.
ral effusion 2 months after the diagnosis of breast metastasis, 11. Alva S, Shetty-Alva N. An update of tumor metastasis to the breast
data. [letter]. Arch Surg. 1999;134:450.
was placed on hospice, and died within 9 months after the 12. Trevithick E. A case of chloroma with clinical history and account
diagnosis of breast metastasis. of post mortem appearances. Lancet. 1903;2:158–60.
13. Tavassoli FA, Eusebi V. Atlas of tumor pathology. 4th series, fascicle 10.
Take-Home Messages Tumors of the mammary gland. Washington, DC: American Registry
of Pathology/Armed Forces Institute of Pathology; 2009. Chapter18,
These case presentations highlight the importance of the Metastasis to the breast from nonmammary neoplasms: 391–5.
clinical history (Cases 1–4), the variations in outcomes after 14. McCrea ES, Johnston C, Haney PJ. Metastases to the breast. Am J
a diagnosis of metastatic carcinoma to the breast (Case 4 ver- Roentgenol. 1983;141:685–90.
sus Case 5), and the importance of a high suspicion for meta- 15. Wood B, Sterrett G, Frost F, Swarbrick N. Diagnosis of extrama-
mmary malignancy metastatic to the breast by fine needle biopsy.
static disease (Case 5). Pathology. 2008;40:345–51.
16. Lester SC, Hicks DG. Diagnostic pathology: breast. 2nd ed.

Philadelphia: Elsevier; 2016. Section 9. Other types of malignan-
References cies: metastasis to breast: 624–9.
17. Toombs BD, Kalisher L. Metastatic disease to the breast: clini-
cal, pathologic, and radiographic features. Am J Roentgenol.
1. DeLair DF, Corben AD, Catalano JP, Vallejo CE, Brogi E, Tan 1977;129:673–6.
LK. Non-mammary metastases to the breast and axilla: a study of 18. Chaignaud B, Hall TJ, Powers C, Subramony C, Scott-Conner
85 cases. Mod Pathol. 2013;26:343–9. CE. Diagnosis and natural history of extramammary tumors meta-
2. Lee AHS. The histological diagnosis of metastasis to the breast static to the breast. J Am Coll Surg. 1994;179:49–53.
from extramammary malignancies. J Clin Pathol. 2007;60:1333–41. 19. Vizcaíno I, Torregrosa A, Higueras V, Morote V, Cremades A,
3. Williams SA, Ehlers RA 2nd, Hunt KK, Yi M, Kuerer HM, Singletary Torres V, et al. Metastasis to the breast from extramammary malig-
SE, et al. Metastases to the breast from nonbreast solid neoplasms: nancies: a report of four cases and a review of literature. Eur Radiol.
presentation and determinants of survival. Cancer. 2007;110:731–7. 2001;11:1659–65.
4. Georgiannos SN, Chin J, Goode AW, Sheaff M. Secondary 20. Lee SK, Kim WW, Kim SH, Hur SM, Kim S, Choi JH, et al.
neoplasms of the breast: a survey of the 20th century. Cancer. Characteristics of metastasis in the breast from extramammary
2001;92:2259–66. malignancies. J Surg Oncol. 2010;101:137–40.
5. Hajdu SI, Urban JA. Cancers metastatic to the breast. Cancer. 21. Jochimsen PR, Brown RC. Metastatic melanoma in the breast mas-
1972;29:1691–6. querading as fibroadenoma. JAMA. 1976;236:2779–80.
256 B. M. Turner
22. Yeh CN, Lin CH, Chen MF. Clinical and ultrasonographic char- review of the literature about therapeutic management. Breast.
acteristics of breast metastases from extramammary malignancies. 2011;20:389–93.
Am Surg. 2004;70:287–90. 47. Domchek AM, Hecht JL, Fleming MD, Pinkus GS, Canellos

23. Vergier B, Trojani M, de Mascarel I, Coindre JM, Le Treut
GP. Lymphomas of the breast: primary and secondary involvement.
A. Metastases to the breast: differential diagnosis from primary Cancer. 2002;94:6–13.
breast carcinoma. J Surg Oncol. 1991;48:112–6. 48. Mattia AR, Ferry JA, Harris NL. Breast lymphoma. A B-cell spec-
24. Baranzelli MC, Granier AM, Aornillot M, Damaille MC. Five cases trum including the low grade B-cell lymphoma of mucosa associ-
of breast metastasis in children. Cytologic aspects. Arch Anat Cytol ated lymphoid tissue. Am J Surg Pathol. 1993;17:574–87.
Pathol. 1986;34:58–61. 49. Jeon HJ, Akagi T, Hoshida Y, Hayashi K, Yoshino T, Tanaka T,
25. Nayar M, Chandra M, Aggarwal R, Chander S. Carcinoma cervix et al. Primary non-Hodgkin malignant lymphoma of the breast.
presenting as a primary breast malignancy. Ind J Pathol Microbiol. An immunohistochemical study of seven patients and literature
1987;3:283–6. review of 152 patients with breast lymphoma in Japan. Cancer.
26. Mihai R, Christie-Brown J, Bristol J. Breast metastases from
1992;70:2451–9.
colorectal carcinoma. Breast. 2004;13:155–8. 50. Wiseman C, Liao KT. Primary lymphoma of the breast. Cancer.
27. Baum M, Colletta A. Breast cancer: a revolutionary concept. Breast 1972;29:1705–12.
Cancer. 1995;2:9–18. 51. Renshaw AA, Richie JP. Subtypes of renal cell carcinoma.

28. Canda AE, Sevine AI, Kocdor MA, Canda T, Balci P, Saydam S, Different onset and sites of metastatic disease. Am J Clin Pathol.
et al. Metastatic tumors in the breast: a report of 5 cases and review 1999;111:539–43.
of the literature. Clin Breast Cancer. 2007;7:638–43. 52. Rosen LE, Gattuso P. Neuroendocrine tumors of the breast. Arch
29. Rosen PP. Metastases in the breast from nonmammary malignant Pathol Lab Med. 2017;141:1577–81.
neoplasms. In: Rosen PP, editor. Rosen’s breast pathology. 2nd ed. 53. Hugh JC, Jackson FI, Hanson J, Poppema S. Primary breast lym-
Philadelphia: Lippincott Williams & Wilkins; 2001. phoma. An immunohistologic study of 20 new cases. Cancer.
30. Amichetti M, Perani B, Boi S. Metastases to the breast from extra- 1990;66:2602–11.
mammary malignancies. Oncology. 1990;47:257–60. 54. Balei P, Undar B, Yilmaz E, Seçil M, Ozsan H, Canda

31. Silverman EM, Oberman HA. Metastatic neoplasms in the breast. T. Bilateral breast involvement in Sézary syndrome. Eur Radiol.
Surg Gynecol Obstet. 1974;138:26–8. 2001;11:2468–71.
32. Nielsen M, Ørnvold K, Andersen JA, Kristensen PB, Lorentzen M, 55. Dabbs DJ. Diagnostic immunohistochemistry. 2nd ed. Philadelphia:
Ravn V, et al. Metastases to the breast from extramammary carcino- Churchill Livingstone; Elsevier; 2006.
mas. APMIS. 1981;89:251–6. 56. Asch-Kendrick R, Cimino-Mathews A. The role of GATA3 in

33. Lee SH, Park JM, Kook SH, Han BK, Moon WK. Metastatic breast carcinomas: a review. Hum Pathol. 2016;48:37–47.
tumors to the breast: mammographic and ultrasonographic find- 57. Banerjee SS, Harris M. Morphological and immunophenotypic vari-
ings. J Ultrasound Med. 2000;19:257–62. ations in malignant melanoma. Histopathology. 2000;36:387–402.
34. Sneige N, Zachariah S, Fanning TV, Dekmezian RH, Ordóñez 58. Turner BM, Cagle PT, Sainz IM, Fukuoka J, Shen SS, Jagirdar
NG. Fine needle aspiration cytology of metastatic neoplasms in the J. Napsin A, a new marker for lung adenocarcinoma, is comple-
breast. Am J Clin Pathol. 1989;92:27–35. mentary and more sensitive and specific than thyroid transcription
35. Bohman LG, Bassett LW, Gold RH, Voet R. Breast metastases from factor 1 in the differential diagnosis of primary pulmonary carci-
extramammary malignancies. Radiology. 1982;144:309–12. noma: evaluation of 1674 cases by tissue microarray. Arch Pathol
36. Paulus DD, Libshitz HI. Metastasis to the breast. Radiol Clin N Lab Med. 2012;136:163–71.
Am. 1982;20:561–8. 59. Kaufmann O, Dietel M. Thyroid transcription factor-1 is the supe-
37. Raptis S, Kanbour AI, Dusenberg D, Kanbour-Shakir A. Fine- rior immunohistochemical marker for pulmonary adenocarcinomas
needle aspiration cytology of metastatic ovarian carcinoma to the and large cell carcinomas compared to surfactant proteins A and
breast. Diagn Cytopathol. 1996;15:1–6. B. Histopathology. 2000;36:8–16.
38. Domanski H. Metastases to the breast from extramammary neo- 60. Miettinen M, McCue PA, Sarlomo-Rikala RJ, Czapiewski P, Wazny
plasms. A report of six cases with diagnosis by fine needle aspira- K, et al. GATA 3 – a multispecific but potentially useful marker in
tion cytology. Acta Cytol. 1996;40:1293–300. surgical pathology – a systematic analysis of 2500 epithelial and
39. Silverberg SG, Masood S. The breast. In: Silverberg SG, editor. non-epithelial tumors. Am J Surg Pathol. 2014;38:13–22.
Principles and practice of surgical pathology and cytopathology. 61. Laury AR, Perets R, Piao H, Krane JF, Barletta JA, French C, et al.
3rd ed. New York: Churchhill Livingstone; 1997. p. 660. A comprehensive analysis of PAX8 expression in human epithelial
40. Azzopardi JG. Problems in breast pathology. Major problems in tumors. Am J Surg Pathol. 2011;35:816–26.
pathology, vol. 11. Philadelphia: WB Saunders; 1979. 62. Nonaka D, Chiriboga L, Soslow RA. Expression of PAX8 as a use-
41. Husseini MR. Extracutaneous malignant melanomas. Cancer
ful marker in distinguishing ovarian carcinomas from mammary
Investig. 2008;26:516–34. carcinomas. Am J Surg Pathol. 2008;32:1566–71.
42. Moore DH, Wilson DK, Hurteau JA, Look KY, Stehman FB, Sutton 63. Espinosa I, Gallardo A, D’Angelo E, Mozos A, Lerma E, Prat
GP. Gynecologic cancers metastatic to the breast. J Am Coll Surg. J. Simultaneous carcinomas of the breast and ovary: utility of
1998;187:178–18. Pax-8, WT-1, and GATA3 for distinguishing independent primary
43. Recine MA, Deavers MT, Middleton LP, Silva EG, Malpica
tumors from metastases. Int J Gynecol Pathol. 2015;34:257–65.
A. Serous carcinoma of the ovary and peritoneum with metastases 64. Acs G, Pasha T, Zhang PJ. WT1 is differentially expressed in
to the breast and axillary lymph nodes: a potential pitfall. Am J Surg serous, endometrioid, clear cell, and mucinous carcinomas of the
Pathol. 2004;28:1646–51. peritoneum, fallopian tube, ovary, and endometrium. Int J Gynecol
44. Yamasaki H, Saw D, Zdanowitz J, Faltz LL. Ovarian carcinoma Pathol. 2004;23:110–8.
metastasis to the breast case report and review of the literature. Am 65. Domfeh AB, Carley AL, Striebel JM, Karabakhtsian RG, Florea
J Surg Pathol. 1993;17:193–7. AV, McManus K, et al. WT1 immunoreactivity in breast carcinoma:
45. Guo S, Yao J. Breast metastasis of salivary duct carcinoma in a selective expression in pure and mixed mucinous subtypes. Mod
patient: a case report. Int J Clin Exp Med. 2015;8:21765–9. Pathol. 2008;21:1217–23.
46. Fujita N, Kimura R, Yamamura J, Akazawa K, Kasugai T,
66. Gatalica Z, Norris BA, Kovatich AJ. Immunohistochemical local-
Tsukamoto F. Leiomyosarcoma of the breast: a case report and ization of prostate-specific antigen in ductal epithelium of male
breast. Potential diagnostic pitfall in patients with gynecomastia. 72. Hainsworth JD, Rubin MS, Spigel DR, Boccia RV, Raby S, Quinn
Appl Immunohistochem Mol Morphol. 2000;8:158–61. R, et al. Molecular gene expression profiling to predict the tissue
67. Kidwai N, Gong Y, Sun X, Deshpande CG, Yeldandi AV, Rao MS, of origin and direct site-specific therapy in patients with carcinoma
et al. Expression of androgen receptor and prostate-specific antigen of unknown primary site: a prospective trial of the Sarah Cannon
in male breast carcinoma. Breast Cancer Res. 2004;6:R18–23. Research Institute. J Clin Oncol. 2013;31:217–23.
68. Carder PJ, Speirs V, Ramsdale J, Lansdown MR. Expression
73. Meiri E, Mueller WC, Rosenwald S, Zepeniuk M, Klinke E,

of prostate specific antigen in male breast cancer. J Clin Pathol. Edmonston TB, et al. A second-generation MicroRNA-based assay
2005;58:69–71. for diagnosing tumor tissue origin. Oncologist. 2012;17:801–12.
69. O’Connell FP, Wang HH, Odze RD. Utility of immunohistochem- 74. Anderson GG, Weiss LM. Determining tissue of origin for meta-
istry in distinguishing primary adenocarcinomas from metastatic static cancers: meta-analysis and literature review of immunohis-
breast carcinomas in the gastrointestinal tract. Arch Pathol Lab tochemistry performance. Appl Immunohistochem Mol Morphol.
Med. 2005;129:338–47. 2010;18:3–8.
70. Zhu S, Schuerch C, Hunt J. Review and updates of immunohis- 75. Oien KA, Dennis JL. Diagnostic work-up of carcinoma of unknown
tochemistry in selected salivary gland and head and neck tumors. primary: from immunohistochemistry to molecular profiling. Ann
Arch Pathol Lab Med. 2015;139:55–66. Oncol. 2012;23(Suppl 10):x271–7.
71. Kezlarian B, Alhyari M, Venkataraman G, Karner K, Inamdar KV, 76. Pillai R, Deeter R, Rigl CT, Nystrom JS, Miller MH, Buturovic L,
Menon MP. GATA3 immunohistochemical staining in Hodgkin Henner WD. Validation and reproducibility of a microarray-based
lymphoma: diagnostic utility in differentiating classic Hodgkin gene expression test for tumor identification in formalin-fixed,
lymphoma from nodular lymphocyte predominant Hodgkin lym- paraffin-embedded specimens. J Mol Diagn. 2011;13:48–56.
phoma and other mimicking entities. Appl Immunohistochem Mol 77. Takei H, Monzon FA. Gene-expression assays and personalized
Morphol. 2017; https://doi.org/10.1097/PAI.0000000000000581. cancer care: tissue-of-origin test for cancer of unknown primary
[Epub ahead of print]. origin. Pers Med. 2011;8:429–36.
Index
A “pagetoid” fashion, 114

Activated B cell (ABC)-type, 226–229 pleomorphic LCIS, 107, 108
Adenoid cystic carcinoma (ACC) radiation therapy, 124
definition, 53, 54 recurrent genetic changes, 117
differential diagnosis, 57 reported in men, 124
features, 53, 54 risk lesion, 109
immunohistochemical profile, 54, 56 small duct, 114
molecular features, 56 small uniform lesional cells, 111
prognosis, 56, 57 surgical margin, 123, 124
Adjuvant cytotoxic chemotherapy, 97 terminal duct lobular unit, 109
Adjuvant hormonal therapy, 97, 125 Type A cells, 112
Adjuvant radiation therapy, 97 Type B cells, 112
American Joint Commission on Cancer (AJCC), 30, 94, 95, 113 ultrasound findings, 107
American Society of Clinical Oncology/College of American
Pathologists (ASCO/CAP), 173, 176–178
Anaplastic large cell lymphoma (ALCL), 217 B
histologic differentiation, 222 Benign fibroepithelial neoplasm, 165
immunohistochemical markers, 222, 223 Benign vascular lesions, 214
limited antigen expression profile, 222 Beta-catenin, 88
morphology, 222 Blunt duct adenosis, 19
Androgen receptor (AR), 184 Breast Cancer International Group-North American Breast Cancer
Apocrine carcinoma, 44–46 Group (BIG-NABCG), 30
Atypical ductal hyperplasia (ADH), 1, 7–9, 12, 147 Breast Imaging Reporting and Data System (BIRADS) category 4, 1
Atypical lobular hyperplasia/lobular carcinoma in situ Breast leiomyoma, 207
(ALH/LCIS), 1, 6 Breast mesenchymal lesion
anti-hormonal agents, 124 interlobular stroma, 203
architectural features, 109 non-mesenchymal malignancies, 203
cancer incidence, 109 PASH, 203
carcinoma in situ, 119 post-radiation sarcomas, 203
classic cytologic and architectural features, 110 soft tissue tumor classifications, 203
collagenous spherulosis, 122, 123 Breast specific gamma imaging (BSGI), 75
core biopsy, 107–109 Burkitt lymphoma (BL)
cytologic features, 109 clinical presentation, 220, 221
DCIS, 108 genetics, 220
E-cadherin, 110, 115 immunophenotype, 220
epithelioid myoepithelial cells, 114 morphology, 220
estrogen receptor positive, 123
fibroepithelial lesion, 123
immunohistochemistry, 114 C
invasive carcinoma, 108, 109 Carney syndrome, 159
invasive ductal carcinoma, 122 Cassette mapping, 78
LIN nomenclature, 113 Cell-of-origin (COO) classification, 219
lobular neoplasia and DCIS, 117, 119 Cellular angiolipoma, 215, 216
low grade DCIS, 115 Cellular FAs, 159
mammographic calcifications, 107 Chondroid matrix formation
minimal diagnostic criteria, 111 chondrosarcoma, 211
molecular mechanisms, 115 components, 210
MRI findings, 107 lesions, 211
myoepithelial cells, 114 myxoid changes, 211
myoepithelial stains, 122 prognosis, 210
nuclei of myoepithelial cells, 114 Chromogranin A, 48

https://doi.org/10.1007/978-3-030-16518-5
260 Index
CK5/6 staining, 16 ER-/PR+ tumors, 175

Classic Hodgkin lymphoma (CHL), 217 evaluation of, 173, 174
Classic invasive lobular carcinoma, 86, 96 external positive controls, 174
Classic lobular carcinomas, 87 formalin fixation time, 174
Claudin-low breast cancers, 26 formalin fixed paraffin embedded tissue, 173
Claudin-low tumors, 31 GATA3, 183, 184, 187
Coincidental tubular carcinoma, 7 gynecologic cancers, 188
College of American Pathologists (CAP), 113 heterogeneity, 188, 189
Columnar cell change (CCC), 1, 5, 41 image analysis, 181
Columnar cell hyperplasia (CCH), 1, 5, 41 Ki-67, 29
Columnar cell lesions (CCLs), 5 low grade invasive ductal and lobular carcinomas, 175
Core needle biopsy (CNB), phyllodes tumors, 2, 151, 152 luminal A subtype, 182
benign PT, 162 luminal type breast cancers, 175
diagnosis, 161 lung adenocarcinomas, 187
vs. fibroadenoma, 162 metastatic carcinoma, 187
FNAC, 162 metastatic/recurrent/chemotherapy-resistant TNBCs, 184, 185
hypercellular and pleomorphic spindle cells, 161, 162 Netherlands studied receptor conversion, 188
Cribriform carcinoma NOS, 28
definition, 39 Nottingham histologic grade, 28
features, 39 Oncotype DX test, 28
prognosis, 39 optimal breast tissue handling, 180
tumor profile status, 39, 40 p16 expression, 185
Cytokeratin 7 (CK7), 103 p53 expression, 185
Cytologic heterogeneity, 114 positive internal controls, 174
Cytotoxic chemotherapy, 97 pre-analytic variables, 182
prognostic and predictive implications, 173
QIA, 182, 183
D quality control and quality assurance, 173
Diffuse large B cell lymphoma (DLBCL), 219, 220 TNBC tumors, 184
immunohistochemistry, 222 treatment of, 175
morphology, 222 TTF-1 and Napsin A, 187
patient history, 226–229 unsupervised clustering analysis, 175
prevalence of, 222 Extranodal extension, 32
treatment modalities, 222 Extranodal marginal zone lymphoma (MZL), 221
Digital breast tomosynthesis (DBT), 74
Direct-to-consumer (DTC) testing, 199
Discordance, 2 F
Ductal carcinoma Fibroadenomas (FAs)
with medullary features, 47, 48 clinical outcome, 167
with neuroendocrine features, 48, 49, 51 definition, 159
Ductal carcinoma in situ (DCIS), 1, 7, 12, 28, 149, 151, 173 diagnostic criteria, 159
Dyscohesion, 87 differentiation, 164, 165
Dyscohesive invasive lobular carcinoma, 81 genetic changes, 166
juvenile FAs, 159, 160
mitoses, 160
E pregnancy changes, 160
E-cadherin, 5, 13, 88–90, 115, 128, 129 Fibroepithelial lesions
Encapsulated papillary carcinoma (EPC) case presentations, 167, 168
definition, 150, 151 fibroadenoma
differentiation, 149 clinical outcome, 167
IHC myoepithelial staining, 149, 151 definition, 159
Eosinophilia, 98 diagnostic criteria, 159
Epithelial HMWCK expression, 149 differentiation, 164, 165
Epithelial membrane antigen (EMA), 42 genetic changes, 166
Epithelioid hemangioendothelioma (EHE), 107 juvenile FAs, 159, 160
Estrogen receptor (ER), 11, 103 mitoses, 160
androgen receptor, 184 pregnancy changes, 160
apocrine carcinoma, 182 Fibromatosis, 204, 205
ASCO/CAP guidelines, 176 beta-catenin, 207, 208
basal cytokeratins, 186 clinical history, 207
cancer stem cells, 187 differential diagnosis, 207
clinicopathological characteristics, 174 IMT, 207, 208
cold ischemic time, 174 spindle cell proliferation, 207
DCIS patients, 173 Fibromatosis-like metaplastic carcinoma (FLMC), 204–206
direct communication, 175 Fibromatosis-like spindle cell carcinoma (FLSCC)
ER-/PR+ breast cancer, 189 definition, 57, 59, 60
Index 261
differential diagnoses, 59 penetrance gene mutations, 193

features, 57, 59, 60 polygenic risk scores, 198
Fine needle aspiration (FNA), 77 synchronous breast and colon cancers, 197
Fine needle aspiration cytology (FNAC), 162 triple negative breast cancer, 197
Flat epithelial atypia (FEA), 1, 5, 7 High-molecular-weight cytokeratins (HMW-CKs), 5
Florid ductal hyperplasia, 153 Histiocytic sarcoma (HS), 223
Florid lobular carcinoma in situ (FLCIS), 4, 113, 116 Human epidermal growth factor receptor 2 (HER2)
Follicular lymphoma (FL) ASCO/CAP guidelines, 177, 178
genetics, 222 chromosomal monosomy/polysomy, 179
immunophenotype, 221, 222 direct communication, 180
morphology, 221 external positive controls, 174
Forkhead-box protein A1 (FOXA1), 103 FISH testing, 179, 180
Full field digital mammograms (FFDM), 74 formalin fixed paraffin embedded tissue, 173
HER2 positive patients, 173
high grade invasive ductal carcinoma, 178
G Ki67 expression, 179–181
Genetic instability syndromes, 214 Netherlands studied receptor conversion, 188
Grade 2 DCIS, 2 optimal breast tissue handling, 180
Grade 3 invasive ductal carcinoma, 87 primary and recurrent invasive breast cancers, 173
Granular cell tumors, 212, 213 quality control and quality assurance, 178, 179
Gross-cystic disease fluid protein fraction-15 (GCDFP-15), 5, 244 transmembrane tyrosine kinase receptors, 175–176
type grade 1 invasive lobular carcinomas, 180
H
Hemangiomas, 214–216 I
Hematologic malignancies, 223 Immunohistochemical/cytochemical staining, 223
Hematolymphoid, 106, 107 Immunohistochemistry (IHC), see Estrogen receptor (ER);
Hereditary Diffuse Gastric Cancer Syndrome, 197 Progesterone receptor (PR)
Hereditary syndromes Implant-associated breast fibromatosis, 213
age of presentation, 193 Inflammatory myofibroblastic tumor (IMT), 207, 208
Ataxia Telangiectasia, 199 Intermediate grade DCIS, 2
ATM protein, 198 Intraductal papillary carcinoma (IDPC)
BRCA1 and BRCA2 gene mutation definition, 150, 151
breast cancer, 194 IDP differentiation, 147–149
chromosome 13, 194 IHC myoepithelial staining, 149, 151
environmental and behavioral factors, 194 Intraductal papilloma (IDP)
founder effect, 194 atypical papilloma, 147
gene phosphorylation, 194 benign changes, 146, 147
higher grade tumors, 196 diagnosis, 147
metastases, 195 florid ductal hyperplasia, 153
multi-center trial, 194 histologic features, 146
needle biopsy, 195 IDPC differentiation, 147–149
NSM, 196 molecular changes, 151
oophorectomy, 195 sclerosis, 152, 153
ovarian cancer risk estimates, 194 Intraductal proliferative disease
ovarian cancer screening, 195 ADH, 8, 9, 11, 12
PARPs, 196 diagnosis, 11
prophylactic mastectomy, 194, 196 FEA, 11
prophylactic surgery vs. appropriate surveillance, 195 molecular alteration of, 11
sentinel lymph node biopsy, 196 UDH, 11
tamoxifen, 195 architectural features, 2
triple breast cancers, 196 blunt duct adenosis, 6, 19, 20
CHEK2 mutations, 197 CCC, 6
ClinVar, 199 CCH, 5, 6
Cowden syndrome CCLs, 5
clinical manifestations, 198 central and punctate necrosis, 12
management and surveillance options, 198 central comedo type necrosis, 5
risk for, 197 clinical features, 1
DNA variations, 199 clinical implications, 1
DTC testing, 199 core biopsy specimen, 9
genetic counseling, 193, 194 cytologic features, 2, 15, 17
genetic testing, 198 DCIS, 6, 8, 9
genomic instability, 199 basal-like characteristics, 5
germline mutations, 199 grade 1, 11–16
Li-Fraumeni syndrome, 196, 197 grade 2, 2, 4, 5, 13, 17
male breast cancer, 197 grade 3, 15–17
262 Index
Intraductal proliferative disease (cont.) DCIS, 28

IDC, 20, 22 definition, 26, 27
UDH, 13 prognosis, 26
definition, 1 sentinel lymph node biopsy, 27
dermo-epidermal junction, 17 NACT
discordance, 2 AJCC staging, 30
ductal-lobular units, 12 international multidisciplinary, 30, 31
ER staining, 17 residual cancer tumor, 30
FDH, 18, 19 tumor response, 30
FEA, 7, 8 neuroendocrine differentiation, 26
FUDH, 19 pCR, 31
gross tissue findings, 1 solid papillary carcinoma, 29
high-molecular-weight cytokeratins, 2 take home messages, 133
IHC markers, 12, 17 TILs, 31, 32
imaging findings, 1 TNBC, 26, 31
immunohistochemical stains differentiate cells, 18 tumor profile, 33
intraluminal necrosis, 6 Invasive lobular carcinoma, 94, 95
LCIS, 6, 13 acceptable surgical margins, 95
LG-DCIS, 21 adjuvant cytotoxic chemotherapy, 97
mammographic calcifications, 9 adjuvant hormonal therapy, 97
microcalcifications, 12, 13, 17 adjuvant radiation therapy, 97
mitoses, 12 architectural features, 80
MPD, 17, 18 architectural variants of, 83
myoepithelial cells, 19, 20 breast lobules, 79
nuclear feature, 15 catenin staining, 88, 90–92
nucleoli and mitotic figures, 6 classic histology of, 81
PASD and CEA, 17 classic invasive lobular carcinomas, 95, 96
proliferating epithelial cells, 5 clinical features, 73
radial scar/radial sclerosing lesion, 5 clinical presentation, 73, 74
surgical specimen, 2 cytologic features, 79, 80
tumor markers, 17 cytologic heterogeneity of, 83
UDH, 2, 6 cytologic variants of, 85
ADH, 12 cytotoxic chemotherapy, 97
grade 1 DCIS, 12 E-cadherin, 88, 90–92
IHC, 4 germline mutation, 88
H&E stains, 3 gross pathologic feature, 77
FEA, 8, 12 gross tissue sampling, 75
Intraluminal necrosis, 6 histologic review, sample for, 77, 78
Intratumoral heterogeneity (ITH), 32, 33 histopathologic features, 85
Invasive breast cancer immunostaining, 91
ancillary studies, 125 incidence of, 73
diagnosis, 125 intracytoplasmic mucin, 86
histologic findings, 125 invasive ductal carcinoma, 84, 87
imaging, 124 LCIS/ALH, 122
take home messages, 125 lobular growth pattern, 94
Invasive breast carcinoma with neuroendocrine differentiation mammography, 74
(IBC-NED), 48 management steps, 77
Invasive ductal carcinoma (IDC), 87, 222 microinvasive carcinoma, 87
claudin-low subtype, 31 mitotic rate, 87
clinical features of, 73 mixed ductal lobular carcinoma, 92, 93
ER positive breast cancers, 28 morphologic variants, 80
Ki-67, 29 MRI, 75
Nottingham histologic grade, 28 neoadjuvant therapy, 97
Oncotype DX test, 28, 33 neoendocrine therapy, 97, 98
extranodal extension, 32 nuclear pleomorphism, 87
final diagnosis, 133 Oncotype DX assay, 95, 96
HER2 ITH, 32, 33 pleomorphic cells, 80
histologic findings, 133 pleomorphic invasive lobular carcinoma, 84
histopathologic features, 85 radiation, 97
history, 133 recurrent genomic changes, 96, 97
imaging, 133 ribbon shaped marker, 79
LCIS/ALH, 122 solid, trabecular and alveolar architecture, 86
lymphovascular invasion, 29 surgical margin, 95
metastatic carcinoma, 25, 26 tubule and gland formation, 87
microinvasive carcinoma Type A cells, 80
chemotherapy, 28 Type B cells, 80
Index 263
ultrasound, 75 FLSCC, 57, 59, 60

x-ray mammography, 74 with mesenchymal differentiation, 67, 68
Isolated tumor cells (ITC), 98, 103 prognosis, 67, 68
spindle cell carcinomas, 63–65, 67
squamous cell carcinomas, 59, 62
J Metaplastic spindle cell carcinoma (MSC), 204
Juvenile breast carcinoma, 48 Metastatic cancers
contralateral breast primary metastases, 237
differential diagnosis, 238
K epidemiology, 237
Keratin immunostaining, 87, 97, 99, 106 frequency of, 237
Ki-67 staining, 5 primary breast carcinoma, 237
adenocarcinoma, 241, 242
clinical differences, 238
L gastrointestinal tract, 240, 243, 250, 251
Lamina propria, 104 genitourinary carcinoma, 242, 243
Lipoma genitourinary tract, 240, 247, 250
adipocytes, 208 gross characteristics, 241
intramammary lipoma, 208 gynecological tract carcinoma, 240, 241, 247, 249, 250,
localized overgrowth, 208 253, 254
mimic liposarcoma hematopoietic tumor, 240, 241, 244, 245, 251
breast parenchyma, 209 histologic differences, 239
fat necrosis, 209, 210 immunohistochemical panels, 244–246
vs. lipoatrophy, 209, 210 malignant melanoma, 241
silicone granuloma, 210 metastatic head and neck carcinoma, 251
types, 208 metastatic head and neck tumor, 240, 244
Lobular carcinoma in situ (LCIS), 1 metastatic melanoma, 239, 241, 246, 247, 254, 255
calcifications, 108 molecular profiling tests, 251–253
colonizes nipple ducts, 114 neuroendocrine carcinoma, 240, 243, 244
comedonecrosis and calcifications, 113 neuroendocrine tumor, 250, 251
florid, 113 primary pulmonary, 247, 248, 254
grading of, 113 pulmonary carcinoma, 239, 240
pleomorphic, 112, 113, 115 radiographic differences, 239
post chemotherapy, 125 sarcoma, 240, 244, 251
Lobular Intraepithelial Neoplasia (LIN), 113 prognosis, 253
Lobular neoplasia, 117 urothelial carcinoma, 253
Loss of heterozygosity (LOH), 2, 5 Metastatic carcinoma, 25, 26
Low-molecular-weight cytokeratins (LMW-CKs), 5 Metastatic lobular carcinoma, 102
Lymph node, 100 axillary lymph node dissections, 102
Lymphoepithelial lesions, 221 E-cadherin, 107
Lymphoma frozen sections, 102
breast implants, 218 gastric carcinoma, 105
high grade B cell NHLs, 219 gynecologic carcinoma, 105
low grade B cell NHLs, 218, 219 hematolymphoid, 106, 107
lymphoid proliferations/inflammatory breast lesions, 217 immunomarkers, 103
types, 217, 218 isolated tumor cells, 103
Lymphoplasmacytic lymphoma (LPL), 217 lobular vs. ductal, 103
Lymphovascular invasion, 29 lung, 103
lymph nodes, 97–99, 101
macrometastasis, 103
M metastatic involvement, 103
Macrometastasis, 103 micrometastasis, 103
Magee equations, 33 rhabdomyosarcoma, 107
Mammaglobin, 25 touch preps, 102
Mammary carcinoma, 45–47 urothelial, 105, 106
Mammary Intraepithelial Neoplasia (MIN), 119 Microinvasive carcinoma, 87
Mammary myofibroblastoma, 205–207 chemotherapy, 28
Mammary Paget Disease (MPD), 17 DCIS, 28
Mantle cell lymphoma (MCL), 217 definition, 26, 27
Marginal zone lymphoma, 229–231, 233 prognosis, 26
Mesenchymal sarcoma, 213, 214 Micrometastasis, 103
Metaplastic carcinoma (MC) Micropapillary carcinoma, 42, 44, 45
classification, 57 Micropapillary DCIS, 2
definition, 57, 58 miRview mets2, 252
differential diagnosis, 57 Mixed ductal lobular carcinoma, 92, 93
features, 57, 58 Mucinous carcinoma, 42, 44
264 Index
Multifocal invasive lobular carcinoma, 76, 129 ER, PR and HER2 testing, 166
Myeloid sarcoma, 223 genetic changes, 166, 167
Myofibroblastoma, 59 IHC markers, 165, 166
Myxoid FA, 159 borderline PTs, 161
CNB
benign PT, 162
N diagnosis, 161
National Comprehensive Cancer Network (NCCN) guidelines, 193 vs. fibroadenoma, 162
Neoadjuvant chemotherapy (NACT) FNAC, 162
AJCC staging, 30 hypercellular and pleomorphic spindle cells, 161, 162
international multidisciplinary, 30, 31 components, 160
residual cancer tumor, 30 grading system
tumor response, 30 biomarkers and genetic changes, 164
Neuroendocrine differentiation, 26 clinical relevance, 163
Nipple sparing mastectomy (NSM), 196 CNB reporting, 164
No special type (NST), 26 histologic criteria, 162–164
Nodular fasciitis, 59, 205, 206 proportion, 162
Nonpalpable breast lesions, 2 malignant PT, 160, 165
Nottingham histologic grade, 28 Pleomorphic calcifications
Nuclear atypia, 9–11 final diagnosis, 133
histologic findings, 131
history, 131
O imaging, 131
Occult lobular carcinoma, 74 take home messages, 133
Oncotype DX recurrence score (RS) test, 28, 33 Pleomorphic invasive lobular carcinoma, 84
Osteoclast-like giant cells, 45–47 Pleomorphic LCIS, 112, 113, 115, 119
Osteoid matrix/osseous metaplasia Pleomorphic lobular carcinoma, 84, 85, 87, 96
axillary lymph node metastasis, 212 diagnosis, 135
diagnosis of exclusion, 212 histologic findings, 135
in late adulthood, 212 history, 133, 135
malignant phyllodes tumor, 211 imaging, 135
metaplastic carcinoma, 211 immunophenotypic and molecular features, 113
non-neoplastic lesions, 211 take home messages, 135
osteoblastic type, 212 Poly-ADP ribose polymerases (PARPs), 196
osteosarcoma, 212 Poorly-differentiated neuroendocrine carcinoma/small cell carcinoma
primary osteosarcoma, 211 (PD-NEC/SCC), 48
small cell and well-differentiated subtypes, 212 Primary breast lymphoma (PBL)
telangiectatic osteosarcoma, 212 clinical presentation, 217
definition, 215, 217
features, 217, 218
P mastectomy, 223
Paget’s disease, 17, 74 radiographic features, 217
Palpable breast masses, 2 recurrence, 224
final diagnosis, 130 relapse rates and patterns, 224
histologic findings, 130 types, 217
history, 129 Progesterone receptor (PR), 103
imaging, 129, 130 androgen receptor, 184
take home messages, 130 apocrine carcinoma, 182
Papillary lesions ASCO/CAP guidelines, 176
characteristics, 146 basal cytokeratins, 186
clinical presentation, 145 cancer stem cells, 187
CNB, 151, 152 clinicopathological characteristics, 174
definition, 145 cold ischemic time, 174
EPC (see Encapsulated papillary carcinoma) DCIS patients, 173
IDP (see Intraductal papilloma) direct communication, 175
IDPC (see Intraductal papillary carcinoma) ER-/PR+ breast cancer, 189
imaging findings, 145 ER-/PR+ tumors, 175
molecular changes, 151 evaluation of, 173, 174
SPC (see Solid papillary carcinoma) external positive controls, 174
Pathologic complete response (pCR), 31, 77, 97, 176 formalin fixation time, 174
Peripheral T cell lymphoma (PTCL), 217 formalin fixed paraffin embedded tissue, 173
Phyllodes tumors (PTs) GATA3, 183, 184, 187
association of, 161 gynecologic cancers, 188
benign PT, 160, 161, 164, 165 heterogeneity, 188, 189
biomarkers image analysis, 181
clinical outcome, 167 low grade invasive ductal and lobular carcinomas, 175
Index 265
Luminal A subtype, 182 imaging, 125

luminal type breast cancers, 175 take home messages, 129
lung adenocarcinomas, 187 Small invasive lobular carcinomas, 87
metastatic carcinoma, 187 Small lymphocytic lymphoma/chronic lymphocytic leukemia
metastatic/recurrent/chemotherapy-resistant TNBCs, 184, 185 (SLL/CLL), 217
Netherlands studied receptor conversion, 188 Solid papillary carcinoma (SPC), 29
optimal breast tissue handling, 180 biopsy site changes, 154
p16 expression, 185 definition, 150, 151
p53 expression, 185 differentiation, 149, 150
positive internal controls, 174 IHC myoepithelial staining, 149, 151
pre-analytic variables, 182 with invasion, 154, 155
prognostic and predictive implications, 173 Spindle cell carcinomas, 63–65, 67
QIA, 182, 183 Spindle cell tumors, 224–226
quality control and quality assurance, 173 Squamous cell carcinomas
TNBC tumors, 184 definition, 59, 63
treatment of, 175 differential diagnoses, 59, 62
TTF-1 and Napsin A, 187 features, 59, 63
unsupervised clustering analysis, 175 Squamous metaplasia, 160
Prostate-specific antigen (PSA), 250 Stereotactic core biopsy, 131
Prostatic acid phosphatase (PAP), 250 Subtle microacini, 115
Pseudoangiomatous stromal hyperplasia (PASH), 59, 203, 215, 216 Synaptophysin, 48
Q
Quantitative image analysis (QIA), 182, 183
T
T/B cell acute lymphoblastic leukemia (T/B-ALL), 223
R Terminal duct lobular units (TDLUs), 5, 11, 203
Radiation therapy, 97, 124, 125 Triple negative breast cancer (TNBC), 26, 31
Radiotherapy-induced neoplasms, 213 Tubular carcinoma, 39, 41–43
Radiotherapy-induced sarcoma, 213 Tumor infiltrating lymphocytes (TILs), 31, 32
Residual cancer burden (RCB) method, 30
Rhabdomyosarcoma, 107
Rosen Triad, 6 U
Usual ductal hyperplasia (UDH), 1
S
Sarcomatoid carcinoma, 63–65, 67 V
Sclerosis, 152, 153 Variants of uncertain significance (VUSs), 198
Secretory carcinoma, 48, 52, 53 Vascular lesion, 227
Sentinel lymph node biopsy, 27, 223, 224
Skin dimpling
diagnosis, 129 W
histologic findings, 128, 129 Well-differentiated neuroendocrine tumor (WD-NET), 48
history, 125 World Health Organization (WHO) classification, 8, 73, 80

Practical Breast Pathology: Yan Peng Ping Tang Editors

Uploaded by

Copyright:

Available Formats

Practical Breast Pathology: Yan Peng Ping Tang Editors

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Practical Breast Pathology: Yan Peng Ping Tang Editors

Uploaded by

Copyright:

Available Formats

Practical Anatomic Pathology

Series Editors: Fan Lin · Ximing J. Yang

More information about this series at http://www.springer.com/series/13808

Practical Breast Pathology

Practical Anatomic Pathology

© Springer Nature Switzerland AG 2019

Dallas, TX, USA Yan Peng, MD, PhD

Dallas, TX, USA Yan Peng, MD, PhD

1 Intraductal Proliferative Disease of the Breast . . . . . . . . . . . . . . . . . . . . . . . . . . 1

Kimberly H. Allison, MD Stanford University School of Medicine, Department of Pathology,

Julia Y. Tsang, PhD Department of Anatomical and Cellular Pathology, Chinese University

© Springer Nature Switzerland AG 2019 1

in CCLs with atypia, DCIS, and invasive carcinoma, with a

20. How does one differentiate CCC from UDH?

CCC and CCH are both featured by variably dilated TDLU

21. Can CCC have necrosis? Punctate and central?

Apical snouts and luminal secretions may be present, but

22. Can CCC have mitosis?

CCC has fine nuclear chromatin. Nucleoli and mitotic fig-

Table 1.1 Differences of UDH, FEA, ADH, and grade 1 DCIS

43. How does one differentiate grade 1 DCIS from UDH?

The differences between DCIS and UDH are also mainly in

44. How does one differentiate grade 1 DCIS from grade

Nuclear size and morphology: Tumor cells in grade 1 DCIS

45. How does one differentiate grade 1 DCIS from florid

Grade 1 DCIS cells are small, uniform, and cohesive, and

especially punctate necrosis. The presence of punctate necro-

Grade 3 DCIS is a high-grade in situ ductal carcinoma.

48. What IHC markers can be helpful for the diagnosis

The diagnosis of grade 3 DCIS is made by morphology:

49. What are the underlying molecular changes associ-

The molecular profile of grade 3 DCIS is different from

Table 1.2 Immunohistochemical stains differentiate cells in mela-

Next Steps: DDx: CCC vs G2 DCIS vs others

Case 2: Blunt Duct Adenosis mimic FEA (Fig. 1.13a–d)

References 22. Dabbs DJ, Carter G, Fudge M, Peng Y, Swalsky P, Finkelstein

© Springer Nature Switzerland AG 2019 25

13. Should Ki-67 be ordered in ER-positive breast

ER-positive breast cancers are classified as Luminal A and

14. Where should we look for lymphovascular invasion

Lymphovascular invasion in breast carcinoma is associated

15. How do we measure the size of invasive carcinoma

Solid papillary carcinoma is a well circumscribed tumor

References 19. Li X, Oprea-Ilies GM, Krishnamurti U. New developments in

© Springer Nature Switzerland AG 2019 39

Fig. 3.1 Invasive cribriform

Fig. 3.2 Invasive tubular

Fig. 3.3 Invasive mucinous

Fig. 3.5 Invasive apocrine

Fig. 3.7 Invasive ductal

Fig. 3.8 Invasive ductal a

Fig. 3.9 Small cell

Fig. 3.11 Adenoid cystic

nosed only after extensive sectioning and with negative

Fig. 3.15 Spindle cell

initial treatment of FLSCC; chemotherapy and radiation References

Table 4.1 Clinical features of invasive lobular as compared to invasive

• The incidence of invasive lobular carcinoma has been

© Springer Nature Switzerland AG 2019 73

Dallas, TX, USA Yan Peng, MD, PhD

Dallas, TX, USA Yan Peng, MD, PhD

1 Intraductal Proliferative Disease of the Breast . . . . . . . . . . . . . . . . . . . . . . . . . . 1

Fig. 4.12 Immunohistoche Lobular Normal and Ductal

orphological types are present and not if the lesion

Phyllodes Tumor: Morphological Variants

Phyllodes Tumor: Core Needle Biopsy

Differentials of Benign PT and FA