an immunohistochemical panel to differentiate

An Immunohistochemical Panel to DifferentiateMetastatic Breast Carcinoma to Skin From Primary Sweat

Gland Carcinomas With a Review of the Literature

Marian Rollins-Raval, MD, MPH; Mamatha Chivukula, MD; George C. Tseng, ScD; Drazen Jukic, MD, PhD; David J. Dabbs, MD

N Context.—Approximately 25% of patients with breastcancer develop cutaneous metastases. Sweat gland carci-nomas (SGCs) account for about 0.05% of all cutaneousneoplasms. Cutaneous metastases of breast carcinoma(CMBCs) (especially the ductal type) can be difficult todistinguish from SGCs. Treatment and prognoses for these2 types of tumors differ radically, making accuratehistologic diagnosis crucial. Although a few studies attemptto differentiate these entities employing immunohisto-chemical (IHC) studies (some of which we review here), todate, no panel of IHC stains exists, to our knowledge, todistinguish these entities.

Objective.—To devise a panel of IHC stains to distin-guish CMBC from SGC.

Design.—Twelve cases of ductal CMBCs (11 nototherwise specified type, and 1 basal phenotype), 11 casesof SGCs (5 eccrine carcinomas, 3 porocarcinomas, and 3microcystic adnexal carcinomas), 2 benign sweat gland

neoplasm cases, and 2 primary breast cancer cases wereretrieved and analyzed with the following IHC panel:mammaglobin, gross cystic disease fluid protein (GCDFP)15, p63, basal cytokeratins (CK5, CK14, and CK17),androgen receptor, and PAX5.

Results.—The p63 was only weakly expressed in 1 of 12CMBC cases (8.3%), whereas it was strongly expressed in10 of 11 SGC cases (90.9%) (P , .001). Basal cytokeratinsdemonstrated a similar immunoprofile in the SGC group,with 10 of 11 cases (90.9%) expressing all 3 markers, and avariable immunoprofile in the CMBC group with 0%(CK14) (P , .001) to 16.7% (2 of 12 cases; CK5 and CK17)(P , .001) expression. Mammaglobin was expressed in 8 of12 cases (66.7%) of CMBC.

Conclusions.—Together, these 5 IHC stains were com-bined to make a panel that was 100% sensitive and 91%specific in distinguishing between CMBC and SGC.

(Arch Pathol Lab Med. 2011;135:975–983)

Approximately 25% of patients with breast cancer maydevelop cutaneous metastases.1,2 Sweat gland carci-

nomas (SGCs) account for 0.05% of all cutaneousneoplasms.3 Clinically, the presentation of these 2 entitiesis often distinct. Ductal cutaneous metastases of breastcarcinoma (CMBCs) present as multiple lesions in patientswith a previous diagnosis of primary breast carcinoma(PBC), whereas SGC presents as a single cutaneous lesionin patients with no known history of breast cancer.However, CMBCs can be difficult to distinguish fromSGCs histologically, and complex clinical circumstancescan further complicate a diagnosis. Several recent case

reports continue to demonstrate this potential diagnosticpitfall.4,5 We recently received 2 diagnostically challengingcases presenting similar challenges.

Accepted for publication October 12, 2010.From the Department of Pathology, University of Pittsburgh Medical

Center, Pittsburgh, Pennsylvania (Drs Rollins-Raval, Chivukula, Jukic,and Dabbs); and the Department of Biostatistics, University of Pittsburgh(Dr Tseng). Dr Jukic is now with the Division of Dermatopathology,Department of Pathology and Laboratory Medicine, James A. HaleyVeterans’ Hospital, Tampa, Florida.

The authors have no relevant financial interest in the products orcompanies described in this article.

Presented in part at the annual meeting of the United States andCanadian Academy of Pathology, Denver, Colorado, March 5, 2008.

Reprints: Marian Rollins-Raval, MD, MPH, Department of Pathology,University of Pittsburgh Medical Center, A711 Scaife Hall, 3550 TerraceSt, Pittsburgh, PA 15261 (e-mail: [email protected]).

Figure 1. Eccrine carcinoma ex spiradenocylindroma with insetshowing the eccrine carcinoma ex spiradenocylindroma at higherpower (hematoxylin-eosin, original magnifications 3100 and 3400[inset]).

Original Articles

Arch Pathol Lab Med—Vol 135, August 2011 IHC Panel of CMBC Versus SGC—Rollins-Raval et al 975

Case 1

A 75-year-old woman with a history of bilateralmastectomies for infiltrating ductal carcinoma of bothbreasts, colectomy for carcinoma of the rectum, andhysterectomy with bilateral salpingo-oophorectomy forbenign disease presented with skin lesions on her back

7 years after her bilateral mastectomies. Although themorphology of the PBC and the skin lesions was similar,immunohistochemical (IHC) studies were performed onall lesions. For the right PBC, the tumor cells demonstrat-ed strong positivity for cytokeratin (CK) 7, estrogenreceptor (ER), and ERBB2 (formerly HER2/neu) and focalpositivity for progesterone receptor (PR). For the left PBC,

Figure 2. Infiltrating carcinoma shown with hematoxylin-eosin stain (A), as well as immunohistochemical stains for p63 (B), AE1/AE3 (C), CK5/6(D), epithelial membrane antigen (E), and CK7 (F) (original magnifications 3200).

976 Arch Pathol Lab Med—Vol 135, August 2011 IHC Panel of CMBC Versus SGC—Rollins-Raval et al

the tumor cells were strongly positive for CK7, ER, and PR.Both PBCs were negative for CK903, S100, cross cystic fluidprotein–15 (GCDFP-15), and CK20. On hematoxylin-eosinstain, the skin lesion, which demonstrated areas of tightintermingling between carcinoma and a spiradenocylin-droma, was more compatible with a malignant neoplasmarising in the background of a benign mixed tumor ratherthan metastasis to the benign tumor (Figure 1). In thebenign component, CK7 was positive only in the internalmature cells and negative in basal cells, ERBB2 demon-strated weak and less than 1+ positivity, and CK903 wasstrongly positive. CK20, ER, PR, and S100 were all negativein the tumor cells. In the malignant component, CK903 wasstrongly positive, whereas CK7 and S100 were focallypositive. CK20, GCDFP-15, ER, PR, and ERBB2 were allnegative. Although these findings supported the diag-nosis of a primary eccrine carcinoma (EC) arising in aspiradenocylindroma, the possibility of a CMBC to thebenign neoplasm could not be ruled out. Molecular studiesperformed on the paraffin-embedded tissues showedmultiple allelic losses in the PBC and no genetic alterationsin the EC. The patient subsequently developed additionalmetastases to lymph nodes and additional skin sites fromthe EC. Although a definitive diagnosis was rendered inthis case, a simplified IHC panel would have benefitedboth the pathologist and patient.

Case 2

An 84-year-old woman with a history of bilateral PBC20 years earlier presented with a skin lesion of the rightbreast. The morphology of the previous PBC was notconsistent with the new skin lesion, which showed aninfiltrating carcinoma (Figure 2, A). Numerous IHC stainswere performed with the following results: the tumor cellswere positive for p63, AE1/AE3, CK5/6, and epithelialmembrane antigen, were focally positive for CK7 (shown

in Figures 2, B through F, respectively), and were negativefor S100, ER, PR, CK20, GCDFP-15, and mammaglobin.Based on this information alone, it was unclear whetherthe lesion was a CMBC or an SGC, making the devel-opment of a treatment plan difficult.

As the treatment and prognosis for CMBC and SGCdiffer significantly, developing a panel of IHC markers todifferentiate these 2 entities would be of high clinicalvalue. Numerous studies have evaluated these entitieseither individually6–15 or comparatively using various(IHC) stains,16–23 but there has been no study, to ourknowledge, to determine the most useful IHC panel fordifferentiating CMBC from SGC.

Several IHC stains have shown promise in their abilityto differentiate CMBC and SGC. Mammaglobin, a proteinfrequently overexpressed in breast cancer,11 has not beenthoroughly examined in either CMBC or SGC. GCDFP-15,a glycoprotein originally isolated in human breast grosscystic fluid, although present in most primary andmetastatic breast cancers, has also been reported to beexpressed in a few cases of SGC.16,24 Androgen receptor(AR), although often found in high-grade invasive breastcancer, has also been identified in some SGCs.7–9,13

The basal cytokeratin stains (CK5, CK14, and CK17),recently shown to be expressed in the basal phenotype ofbreast carcinoma, were not present in most other types ofPBC.25,26 These stains have also been shown to be presentin a few SGC cases.14,19,20 However, with the exception ofCMBC with a basal phenotype, these markers are morelikely to be seen in SGC; p63, a homologue of the p53 geneand expressed primarily by basal and myoepithelial cellsof the skin, would be more likely to be seen in cases of SGCthan in cases of CMBC.10,21,22

In addition, 1 of the authors (M.R.-R.) noticed a strongcytoplasmic and/or membranous staining with PAX5, aB-cell–specific activator protein of the basal layer ofhealthy epithelium, in an unrelated research study anddecided to use the antibody in the current study,hypothesizing that the staining pattern might be similarin SGC. The PAX5 expression has been previously studiedin B lymphocytes, in most small cell carcinomas, and inMerkel cell carcinomas.27,28

The goal of this study was to construct a panel of IHCstains that would be highly sensitive and specific indistinguishing these 2 morphologically similar entities.

MATERIALS AND METHODS

Literature Review

A MEDLINE (US National Library of Medicine, Bethesda,Maryland) search was performed from 1995 to 2009 to identify

Table 1. Demographics of Cutaneous MetastaticBreast Carcinoma (CMBC) and Sweat Gland Carcinoma

(SGC) Cases

Case Group No. of Cases % FemaleMean Age

(Age Range), y

PBC 2 100 68 (53–83)CMBC 12 100 57 (40–80)a

BAN 2 50 48.5 (34–63)SGC 11 58.3 73.6 (49–93)

Abbreviations: BAN, benign adnexal neoplasm; PBC, primary breastcarcinoma.a Age at presentation of metastatic lesion, not primary breast lesion.

Table 2. Immunohistochemistry Antibody Information

Antibodya Clone Dilution Vendor Location Pretreatment

GCDFP 23A3 1:25 Vector Labs Burlingame, California CC1b mild and standardAR AR441 1:100 Dako North America Carpinteria, California CC1 mild and standardMammaglobin 304-1A5 and 31A5 Predilute Zeta Corporation Sierra Madre, California CC1 mildp63 4A4 1:200 Neomarkers Inc Fremont, California CC1 mild and standardPAX5 24 1:25 BD Biosciences San Jose, California CC1 mild and standardCK5 XM26 1:25 Novocastra

Laboratories LtdNewcastle Upon Tyne,

United KingdomCC1 standard

CK14 LL002 Predilute Ventana MedicalSystems Inc

Tucson, Arizona CC1 standard

CK17 E3 1:20 Dako North America Carpinteria, California CC1 mild and protease 3a All immunostains were performed on the BenchMark XT using the iView/DAB detection system (Ventana Medical Systems, Tucson, Arizona).b Ventana Medical Systems, Tucson, Arizona.


studies similar to our own, comparing primary SGC to CMBC (orin 1 study, to PBC) using IHC stains. We selected a panel of 8 IHCstains to investigate their staining patterns in CMBC and SGC.We did not repeat IHC studies that appeared in previous studiesto be less useful.

Case Identification

A retrospective search for CMBC and SGC cases using a NaturalLanguage Search was performed in the University of PittsburghMedical Center (Pittsburgh, Pennsylvania) CoPath Plus database(Cerner DHT, Inc, Waltham, Massachusetts) for the 9 years from 1998to 2007. Based on the results of the database search, 27 cases wereretrieved from the University of Pittsburgh Medical Center archivesfor the study: 12 cases of ductal CMBC (44%), which included 11cases with morphologies not otherwise specified and 1 case of basalphenotype; 11 cases of SGC (41%), which included 5 cases of EC, 3cases of porocarcinoma (PC), and 3 cases of microcystic adnexalcarcinoma (MAC); and 4 additional, randomly selected cases (15%),which included 2 primary cutaneous adnexal benign neoplasms (aporoma and an apocrine adenoma) and 2 cases of PBC.

Demographics

The patients in PBC and metastatic breast cancer groups were allwomen, whereas the SGC group consisted of both men and women.The mean ages for patients with ductal CMBC and SGC were57 years and 73.6 years, respectively (Table 1). The difference in ageswas significantly different. Although all the patients with CMBC (12of 12; 100%) had documented PBCs, only 1 (case 1, described above)of the 11 patients with SGC (9%) had a history of PBC.

Immunohistochemistry

Hematoxylin-eosin slides were reviewed; a representative tumorblock was selected; sections were obtained on formalin-fixed,

paraffin-embedded tissues; and samples were analyzed with thefollowing IHC stains: mammaglobin, GCDFP-15, p63, CK5, CK14,CK17, AR, and PAX5.

All IHC stain analysis was performed on the VentanaBenchMark, XT (Ventana Medical Systems, Inc, Tucson, Arizona)using the iView DAB Detection System (Ventana Medical) withcommercially available antibodies according to standard proto-cols (Table 2). All negative and positive controls demonstratedappropriate immunolabeling.

The IHC stain results were semiquantitated as follows: AR,PAX5, and p63 were nuclear stains, and strong nuclear positivitywas considered positive staining. The intensity of immunostain-ing was graded as negative (0), weak (1), moderate (2), andstrong (3), and the proportion of positive staining cells wasrecorded as 0% through 5% (1), 6% through 20% (2), 21% through80% (3), and greater than 80% (4).

Cytoplasmic staining was considered positive for GCDFP-15and mammaglobin, and the intensity of immunostaining wasgraded as weak, moderate, or strong.

Basal cytokeratins—CK5, CK14, and CK17—were scored aspositive if any cytoplasmic or membranous staining wasobserved in the tumor cells.

A strong cytoplasmic and/or membranous staining patternwas considered positive for PAX5.

Statistical Analysis

A statistical analysis of the positive and negative results wasperformed with R statistical software 2.10.1 (R Project forStatistical Computing, Wien, Austria). The SGCs were allcompiled into one group for the statistical analysis. Comparisonsof the IHC staining profiles in the CMBC and SGC groups wereperformed using a z test to calculate the P values for eachIHC stain. Subsequently, a simple computer algorithm wascreated to examine the effectiveness of a combination of

Table 3. Comparative Review of Additional Similar Studies and the Immunohistochemical Results

Study Wallace et al,16 1995 Wick et al,17 1998 Busam et al,18 1999 Plumb et al,19 2004

IHC Marker MAC CMBC ECa PBCb MAC PC EC CMBC MAC CMBCc

ER 2/8 1/15 9/27 31/59 2/8 0/4 2/13 10/30PR 5/8 15/15 1/4 27/59 2/8 0/4 2/13 8/30c-erbB-2(HER2/neu) 9/27 12/59ARGCDFP 1/6 7/15 5/27 41/59CEA 21/27 5/59EGF-R 5/8 4/4 11/13 5/30CK5/6 3/3 2/17CK5CK7CK14CK17CK20p63 (nuclear)S100 12/27 27/59PodoplaninPAX5 (C/M)g

Mammaglobin

Abbreviations: 1A, 1 apocrine case; 1N, 1 not otherwise specified case; AR, androgen receptor; CEA, carcinoembryonic antigen; CK, cytokeratin; C/M, cytoplasmic/membranous; CMBC, cutaneous metastatic breast carcinoma (ductal); EC, eccrine carcinoma; EGF-R, epidermal growth factorreceptor; ER, estrogen receptor; ES, ex spiradenocylindroma; GCDFP, gross cystic disease fluid protein; IHC, immunohistochemistry; MAC,microcystic adnexal carcinoma; PBC, primary breast carcinoma; PC, porocarcinoma; PR, progesterone receptor.a In this study, EC refers to 23 eccrine carcinomas, as well as 4 apocrine lesions in the same category that were only separated by their staining for

GCDFP-15, where 1 of 23 cases of eccrine carcinoma was positive whereas 4 of 4 cases of the apocrine lesions were positive for GCDFP.b These authors looked at primary breast carcinomas as opposed to cutaneous metastases.c There were 6 cases with weak staining, considered negative, for CK5/6.d Of the 3, 1 was mucinous and stained only 1+ for p63 and was thus counted as negative here; the other 2 were strongly positive.e There were 5 cases: 2N, 1A, 2ES.f Two of 3 were graded as 1+.g PAX5 showed a strong C/M staining pattern noted previously in normal epithelium which is reported here.


the near statistically significant and statistically significantbiomarkers.

RESULTS

Literature Review

Wallace et al16 investigated the IHC staining character-istics of 15 cases (from 12 patients) of CMBC andcompared them to a series of primary eccrine tumors,including 8 MACs. The authors16 concluded that usingstandard IHC markers, such as ER, PR and GCDFP-15,would not reliably distinguish these 2 populations.

Wick et al17 examined 59 cases of ductal PBCs andcompared them with 27 cases of ductal SGC, which werefurther described as demonstrating eccrine morphology(23 cases; 85%) and apocrine differentiation (4 cases; 15%).Using pancytokeratins, carcinoembryonic antigen, S100protein, GCDFP-15, ER, PR, and c-erbB-2 protein (ERBB2)IHC stains, their findings17 concluded that the infrequencyof GCDFP-15 in eccrine sweat gland tumors as well as thepaucity of carcinoembryonic antigen in breast carcinomascould be useful in predefined differential diagnosticsettings involving these 2 entities with the appropriateclinicopathologic information provided. We included thisstudy in our review because of its similarity to the currentstudy, even though it focused on distinguishing PBC fromSGC, rather than distinguishing CMBC from SGC.

Busam et al18 studied 30 cases of CMBC compared with42 primary SGC cases for their expression of ER, PR, andepidermal growth factor receptor (EGFR). Several addi-tional histologic types of SGC were examined in thisstudy,18 including apocrine, hidradenocarcinoma, mucin-ous, and basaloid carcinomas, which were not included inour study. Only the 3 types of tumors included in ourstudy—MAC, PC, and EC—are presented in Table 3.Their results18 suggested that the expression of EGFR maybe diagnostically helpful in differentiating these 2 groupsof tumors, whereas ER and PR continued to show nosignificant difference between the 2 groups.

Plumb et al19 used CK5/6 to differentiate primarycutaneous adnexal neoplasms, including 3 MACs, fromcutaneous metastatic lesions, including 17 CMBCs. Only 2of 17 CMBC cases (12%) displayed strong positive staining

with CK5/6, with 6 cases (35%) having weak positivestaining, whereas all 3 MACs (100%) expressed strongpositive staining. This difference, although not evaluatedspecifically in the study,19 revealed the diagnostic poten-tial of CK5/6 in differentiating these lesions.

Qureshi et al20 also examined the diagnostic potential ofseveral IHC stains, including p63, CK5/6, CK7, and CK20,to differentiate metastatic carcinomas from primarycutaneous adnexal neoplasms. Several benign adnexalneoplasms as well as malignant neoplasms, such ashidradenocarcinoma, adenoid cystic carcinoma, seba-ceous carcinoma, digital papillary adenocarcinoma, syr-ingomatous carcinoma, and mucinous carcinoma, wereincluded in their study.20 Only their examinations of PCsand CMBCs were included in Table 3. Three cases of PCwere all strongly positive for CK5/6 and p63, whereasCK7 demonstrated positivity in less that 10% of cells for 2cases, and CK20 was completely negative in all cases. The6 CMBC cases were diffusely positive for CK7 and werenegative for CK5/6, p63, and CK20. This study20 demon-strated the diagnostic potential of CK5/6 and p63 indifferentiating these 2 entities.

Ivan et al22 assessed the utility of p63 antibody fordifferentiating primary cutaneous adnexal neoplasms andadenocarcinoma metastatic to the skin. In addition toseveral benign adnexal tumors, the authors22 analyzed 4MACs and 3 ECs (1 of the latter with mucinousdifferentiation) as well as a case of hidradenocarcinomaand 2 cases of trichilemmal carcinomas. Of the MACs andECs, only the ECs with mucinous differentiation werenegative for nuclear positivity of p63, and none of theCMBCs demonstrated nuclear positivity, confirming theusefulness of this marker in distinguishing SGC fromCMBC.

Finally, Liang et al23 investigated the use of podoplaninto differentiate metastases to skin from various organsites, including the breast, from primary skin adnexalcarcinomas. In their study,23 the authors examined 11cases of metastatic breast cancer to the skin, all of which(100%) were completely negative for podoplanin. Theyalso examined a total of 40 primary skin adnexalcarcinomas, only 2 of which (5%; a case of adenoid cystic

Study Qureshi et al,20 2004 Ivan et al,22 2005 Liang et al,23 2007 This Study

IHC Marker PC CMBC MAC ECd CMBC PC CMBC MAC PC ECe CMBC

ERPRc-erbB-2(HER2/neu)AR 0/3 2/3 0/5 5/12GCDFP 0/3 0/3 1/5 (ES) 5/12CEAEGF-RCK5/6 3/3 0/6CK5 3/3 3/3 4/5 2/12CK7 2/3f 6/6CK14 3/3 3/3 4/5 0/12CK17 3/3 3/3 4/5 2/12CK20 0/3 0/6p63 (nuclear) 3/3 0/6 4/4 2/3 0/12 3/3 3/3 4/5 1/12S100Podoplanin 6/6 0/11PAX5 (C/M)g 3/3 2/3 2/5 (1A + 2N) 0/12Mammaglobin 0/3 0/3 2/5 (ES) 7/12

Table 3. Extended


and a case of poorly differentiated adnexal carcinoma)were negative for podoplanin. However, because thedistinction between SGC and metastatic adenocarcinomasmay be equivocal in many cases, the only SGCs includedin their study were the 6 PCs, which were all (100%)positive for podoplanin. The authors23 suggest thatadditional studies may be necessary in the future toevaluate more SGCs for podoplanin. From that limitedstudy,23 podoplanin appeared to show promise in distin-guishing SGC from CMBC.

In our study, the only 2 SGCs that were positive for ARwere 2 of the 3 PCs (67% of the PCs; 18% of all SGCs).Although most adnexal carcinomas were negative forGCDFP-15, only 1 of the 2 ECs (50%), ex spiradenocylin-droma,29 was positive. That same carcinoma was positivefor mammaglobin as well, while staining negative for allother stains. The second EC ex spiradenocylindroma waspositive for mammaglobin, while also staining positive forp63, CK5, CK14, and CK17.

Although most of the CMBC cases in this study werenegative for the basal CK markers, 3 of the 12 cases (25%)were positive for either or both CK5 and CK17. One casethat was positive for both of these immunostains wasmorphologically of a basal phenotype.

Immunohistochemical Analysis

Table 4 provides a summary of IHC profile stainingpattern in CMBC and SGC. The p63 was only weaklyexpressed in 8.3% (1 of 12) of the CMBC cases, whereas itwas strongly expressed in 90.9% (10 of 11) of the SGC cases(P , .001). All 3 basal CKs were expressed in 90.9% (10 of11) of the SGC cases. In comparison, CMBC casesdemonstrated a staining profile of 0% (0 of 12) for CK14and 16.7% (2 of 12) for CK5 and CK17. One case of CMBC(8.3%) expressed both CK and CK17, whereas 2 more casesof CMBC (16.7%) were positive for either CK5 or CK17. Thedifference in basal CK staining among the SGC cases andthe CMBC cases was statistically significant for all 3 stains (P, .001). Mammaglobin expression was seen in 66.7% (8 of12) of the CMBC cases, compared with 18.2% (2 of 11) of theSGC group (P 5 .06), which was marginally statisticallysignificant. The difference in expression for both AR orGCDFP-15 in the 2 groups was not statistically significant.

The appearance of a typical CMBC is demonstrated inFigure 3, A, whereas a typical SGC (a PC) is demonstratedin Figure 3, B. Using these 2 cases as examples, the mostprominent staining pattern for each group is representedin Figure 4. The CMBCs were generally positive formammaglobin and negative for p63, CK5, CK14, and

Table 4. Summary of Immunohistochemistry (IHC) Profile Staining Pattern in Cutaneous Metastatic Breast Carcinoma(CMBC) and Sweat Gland Carcinoma (SGC)

IHC Staina

Mammaglobin,% (No.) p63, % (No.) CK5, % (No.) CK14, % (No.) CK17, % (No.)

PAX5 (C/M),b

% (No.)

CMBC, N 5 12 66.7 (8) 8.3 (1) 16.7 (2) 0 (0) 16.7 (2) 0 (0)SGC, N 5 11 18.2 (2) 90.9 (10) 90.9 (10) 90.9 (10) 90.9 (10) 54.5 (6)P value .06c ,.001 ,.001 ,.001 ,.001 .01

Abbreviation: CK, cytokeratin; C/M, cytoplasmic/membranous.a The differences in staining between CMBC and SGC with androgen receptor (P 5 .25) and gross cystic disease fluid protein–15 (P 5 .19) were not

statistically significant.b PAX5 did not show nuclear staining in either tumor but did show strong C/M staining, which is reported here.c Marginally significant.

Figure 3. Examples of ductal cutaneousmetastatic breast carcinoma (A) and sweatgland carcinoma (porocarcinoma) (B) (hema-toxylin-eosin, original magnifications 3100[A and B] and 3400 [insets]).


CK17 (Figure 4, A through E, respectively), whereas SGCs(represented in the figure by a PC) were generallynegative for mammaglobin and positive for p63, CK5,CK14, and CK17 (Figure 4, F though J, respectively).

None of the CMBC or SGC tumor cells (0%) demon-strated nuclear staining for PAX5 (as shown in the rightinset of Figure 5, B). However, 54.5% (5 of 11) of the casesof SGC expressed a distinct cytoplasmic and/or membra-nous staining pattern (Figure 5, B). Figure 5, A, demon-strates the faint cytoplasmic blush seen in ductal CMBC

(left inset), as well as the benign basal layer epitheliumshowing strong cytoplasmic/membranous staining (rightinset).

Only 5 out of the 8 stains (63%) examined by acceptedcriteria demonstrate statistically significant, or nearstatistically significant, results. Incorporating those 5IHC markers into a sum score system, we constructed apanel to predict the disease represented in each case. Theconditions set were based on the assumption that breastcancer was usually expected to demonstrate the following

Figure 4. Immunohistochemical panel applied to ductal cutaneous metastatic breast carcinoma: mammaglobin (A), p63 (B), CK5 (C), CK14 (D),and CK17 (E); and to sweat gland carcinoma (porocarcinoma): mammaglobin (F), p63 (G), CK5 (H), CK14 (I), and CK17 (J) (original magnifications3100 [A through J] and 3400 [insets]).

Figure 5. PAX5 cytoplasmic/membranousstaining. A, Ductal cutaneous metastaticbreast carcinoma showing only faint cyto-plasmic blush (left inset) and a benign basallayer of epithelium showing strong cytoplas-mic/membranous staining (right inset). B,Sweat gland carcinoma (porocarcinoma)showing diffuse cytoplasmic/membranouspattern (left inset) with scattered B lympho-cytes showing the classic nuclear pattern(right inset) (original magnifications 3100 [Aand B] and 3400 [insets]).


IHC staining profile: mammaglobin+, p632, CK52, CK142,CK172.

For each condition that was not met, 1 point was added.If the score was less than 3 of 5 (0, 1, or 2; ,60%), the casewas defined as CMBC; if it was greater than or equal to 3 of5 (3, 4, or 5; $60%), it was defined as SGC. Using this sumscore system with these conditions, 12 of 12 patients withCMBC (100% sensitivity) were correctly identified as were10 of 11 patients with SGC (91% specificity) (Table 5).

COMMENT

After reviewing several studies that also attempted todifferentiate breast cancer and SGC, we attempted toidentify the most specific antibodies to differentiate these2 neoplasms. The ER, PR, CK7, and CK20 stains were noteffective in differentiating these entities. The GCDFP-15,carcinoembryonic antigen, EGFR, CK5/6, podoplanin,and p63 stains all showed potential based on previousstudies. We further investigated 3 of these 6 IHC stains(GCDFP-15, CK5/6, and p63). In addition, we incorporat-ed 5 additional IHC stains (CK14, CK17, AR, mammaglo-bin, and PAX5), which had not been previously employed,to our knowledge, in differentiating these lesions. Ourlimited panel did not include carcinoembryonic antigen,EGFR, or podoplanin. Carcinoembryonic antigen andEGFR had not shown as much promise in the studiesreviewed as other IHC stains we wished to include. Thepromising utility of podoplanin (published after thecompletion of our study) was unknown during ourinvestigation.

Our study demonstrated a sustained potential of CK5/6(or CK5 in our study) and p63 in distinguishing CMBCfrom SGC. The GCDFP-15 stain did not reveal astatistically significant difference between the staining

patterns of CMBC and SGC. Ultimately, combiningmammaglobin, p63, CK5, CK14, and CK17, we construct-ed the IHC panel described above that consistentlydifferentiates CMBC from SGC in our cases.

Numerous difficulties have hindered researchers in theidentification of a clinically useful IHC panel to distin-guish these entities: the paucity of material, varyingmorphologic appearances of the entities, and differencesin IHC staining techniques across laboratories are only afew. At the outset of this study, the number of casesidentified was few and reflected the rarity of theseneoplasms. In addition, there was considerable heteroge-neity among the groups of tumors examined. Within theclassifications of CMBC and SGC, rare subtypes existedthat were challenging to evaluate. For example, only onecase of basal-phenotype CMBC was included in our study.The IHC staining pattern of that CMBC subtype includedknown positivity for the basal cytokeratins (CK5, CK14,and CK17).26 Although that one case expressed positivityfor both CK5 and CK17, it still fulfilled the criteria of thepanel for classification of the neoplasm as a CMBC.Another rare entity with possible confounding IHCstaining would be metaplastic breast carcinoma. Althoughits morphologic characteristics can be quite distinct, it hasbeen reported to be positive for p63.10 The staining patternof that entity for basal CK has not been extensivelyevaluated. Both of these subtypes of breast cancer warrantfurther investigation of their unique IHC staining pat-terns.

For SGC, 2 cases of EC ex spiradenocylindroma wereincluded. These extremely rare neoplasms demonstratedareas of apocrine differentiation. Undoubtedly, furtherstudy of a larger cohort of SGC cases with apocrinedifferentiation would be desirable. However, of these 2 EC

Table 5. Immunohistochemistry Condition Panela

Case ID Diagnosis MGB+ p632 CK52 CK142 CK172 Score Groupb

CMBC-1 Ductal, NOS type 1 0 0 0 0 1 1CMBC-2 Ductal, NOS type 1 0 0 0 0 1 1CMBC-3 Ductal, NOS type 0 0 0 0 0 0 1CMBC-4 Ductal, NOS type 0 0 0 0 0 0 1CMBC-5 Ductal, NOS type 0 0 0 0 0 0 1CMBC-6 Ductal, basal phenotype 0 0 1 0 1 2 1CMBC-7 Ductal, NOS type 1 0 0 0 0 1 1CMBC-8 Ductal, NOS type 0 0 0 0 0 0 1CMBC-9 Ductal, NOS type 0 0 0 0 0 0 1CMBC-10 Ductal, NOS type 1 1 0 0 0 2 1CMBC-11 Ductal, NOS type 0 0 0 0 1 1 1CBMC-12 Ductal, NOS type 0 0 1 0 0 1 1SGC-1 Eccrine carcinoma 1 1 1 1 1 5 2SGC-2 Eccrine carcinoma, ex

spiradenocylindroma0 0 0 0 0 0 1

SGC-3 Porocarcinoma 1 1 1 1 1 5 2SGC-4 Porocarcinoma 1 1 1 1 1 5 2SGC-5 Microcystic adnexal carcinoma 1 1 1 1 1 5 2SGC-6 Eccrine carcinoma 1 1 1 1 1 5 2SGC-7 Porocarcinoma 1 1 1 1 1 5 2SGC-8 Microcystic adnexal carcinoma 1 1 1 1 1 5 2SGC-9 Microcystic adnexal carcinoma 1 1 1 1 1 5 2SGC-10 Eccrine carcinoma 1 1 1 1 1 5 2SGC-11 Eccrine carcinoma, ex

spiradenocylindroma0 1 1 1 1 4 2

Abbreviations: CK, cytokeratin; CMBC, cutaneous metastatic breast carcinoma; ID, identification; MGB, mammaglobin; NOS, not otherwisespecified; SGC, sweat gland carcinoma.a Condition is the presumed result for metastatic breast carcinoma given the significant difference between CMBC and SGC for each of the 5

immunohistochemistry stains (see Table 4).b Group 1, ,3 conditions not met 5 CMBC; group 2, $3 conditions not met 5 SGC.


ex spiradenocylindroma, only 1 did not fulfill enoughconditions to be defined as an SGC by our immunopanel(see Table 5, SGC-2). The other EC ex spiradenocylin-droma (case 1, described above) fulfilled the criteria of theIHC panel and was correctly identified as an SGC (seeTable 5, SGC-11). The 5-stain IHC panel would havesignificantly reduced the diagnostic difficulty initiallyencountered by the pathologists involved with that case.Although these rare subtypes of CMBC and SGC haveslightly different IHC staining patterns compared withmost cases within these categories, the IHC panel correctlyclassified the entities in all but one case.

Applying the conditions of the IHC panel to case 2(described above) also pointed to a more definitivediagnosis. Mammaglobin was negative, whereas p63and CK5/6 were positive, which fulfilled 3 of theconditions of the IHC panel and identified the lesion asan SGC. The case pathologist commented that themorphology of this new lesion was not completelyconsistent with primary breast lesion but agreed thatbecause breast carcinomas may progress into poorlydifferentiated forms over time, a breast carcinoma couldnot be completely ruled out. With the addition of the IHCpanel to the morphologic examination and clinicopatho-logic information, the case pathologist was more confidentin favoring a diagnosis of SGC. The IHC panel was usefulin distinguishing CMBC from SGC in cases with classicmorphologies as well as those with unique characteristicsthat yielded broader differential diagnoses.

In addition to organizing a diagnostically useful IHCpanel, we present other interesting findings. The basal-phenotype CMBC cases have the potential to metastasizeto the skin, apart from other known metastatic sites, suchas brain and bone. Furthermore, the percentage ofmammaglobin expression in CMBC appeared similar toits previously reported expression in PBC.11 This findingmay indicate preservation of this marker from PBC to themetastases.

Also, the novel interpretation of PAX5 in a cytoplas-mic and/or membranous staining pattern may providea specific marker for tumors of adnexal origin. Theknown role of PAX5 as a transcription factor wasfunctionally consistent with its commonly describednuclear localization in B lymphocytes. However, thecytoplasmic and/or membranous staining pattern wasstriking and raises the possibility of alternative func-tions in cellular pathways. This staining pattern war-rants further investigation.

CONCLUSIONS

We formed a highly sensitive and specific IHC panel,composed of mammaglobin, p63, and 3 basal cytokeratins,with sufficient power to aid in the differentiation betweenCMBC and SGC. We recommend the use of this panel todifferentiate most cases of these 2 entities in routineclinical practice.

We thank Cary Sipos, HT (ASCP) and Kim McManus, HT(ASCP) for their technical assistance. We would also like tothank Jay S. Raval, MD (Department of Pathology, University ofPittsburgh Medical Center) and Darice Y. Wong, PhD (Depart-ment of Bioengineering, University of California, Los Angeles)for thoughtful discussions and critical reviews of earlier versionsof this manuscript. Financial support for this article wasprovided through the University of Pittsburgh Medical Center’sDepartment of Pathology.

References

1. Krathen RA, Orengo IF, Rosen T. Cutaneous metastasis: a meta-analysis ofdata. South Med J. 2003;96(2):164–167.

2. Spencer PS, Helm TN. Skin metastases in cancer patients. Cutis. 1987;39(2):119–121.

3. Cooper PH. Carcinomas of sweat glands. Pathol Annu. 1987;22(pt 1):83–124.

4. Cangelosi JJ, Nash JW, Prieto VG, Ivan D. Cutaneous adnexal tumor with anunusual presentation—discussion of a potential diagnostic pitfall. Am JDermatopathol. 2009;31(3):278–281.

5. McLean SR, Shousha S, Francis N, et al. Metastatic ductal eccrineadenocarcinoma masquerading as an invasive ductal carcinoma of the malebreast. J Cutan Pathol. 2007;34(12):934–938.

6. Bayer-Garner IB, Smoller B. Androgen receptors: a marker to increasesensitivity for identifying breast cancer in skin metastasis of unknown primarysite. Mod Pathol. 2000;13(2):119–122.

7. Riva C, Dainese E, Caprara G, et al. Immunohistochemical study ofandrogen receptors in breast carcinoma: evidence of their frequent expression inlobular carcinoma. Virchows Arch. 2005;447(4):695–700.

8. Narita D, Raica M, Suciu C, Cimpean A, Anghel A. Immunohistochemicalexpression of androgen receptor and prostate-specific antigen in breast cancer.Folia Histochem Cytobiol. 2006;44(3):165–172.

9. Moinfar F, Okcu M, Tsybrovskyy O, et al. Androgen receptors frequently areexpressed in breast carcinomas: potential relevance to new therapeutic strategies.Cancer. 2003;98(4):703–711.

10. Koker MM, Kleer CG. p63 expression in breast cancer: a highly sensitiveand specific marker of metaplastic carcinoma. Am J Surg Pathol. 2004;28(11):1506–1512.

11. Bhargava R, Beriwal S, Dabbs DJ. Mammaglobin vs GCDFP-15: animmunohistologic validation survey for sensitivity and specificity. Am J ClinPathol. 2007;127(1):103–113.

12. Smith KJ, Williams J, Corbett D, Skelton H. Microcystic adnexalcarcinoma: an immunohistochemical study including markers of proliferationand apoptosis. Am J Surg Pathol. 2001;25(4):464–471.

13. Kariya Y, Moriya T, Suzuki T, et al. Sex steroid hormone receptors inhuman skin appendage and its neoplasms. Endocr J. 2005;52(3):317–325.

14. Hoang MP, Dresser KA, Kapur P, High WA, Mahalingam M. Microcysticadnexal carcinoma: an immunohistochemical reappraisal. Mod Pathol. 2008;21(2):178–185.

15. Hiatt KM, Pillow JL, Smoller BR. Her-2 expression in cutaneous eccrineand apocrine neoplasms. Mod Pathol. 2004;17(1):28–32.

16. Wallace ML, Longacre TA, Smoller BR. Estrogen and progesteronereceptors and anti-gross cystic disease fluid protein 15 (BRST-2) fail to distinguishmetastatic breast carcinoma from eccrine neoplasms. Mod Pathol. 1995;8(9):897–901.

17. Wick MR, Ockner DM, Mills SE, Ritter JH, Swanson PE. Homologouscarcinomas of the breasts, skin, and salivary glands: a histologic andimmunohistochemical comparison of ductal mammary carcinoma, ductal sweatgland carcinoma, and salivary duct carcinoma. Am J Clin Pathol. 1998;109(1):75–84.

18. Busam KJ, Tan LK, Granter SR, et al. Epidermal growth factor, estrogen,and progesterone receptor expression in primary sweat gland carcinomas andprimary and metastatic mammary carcinomas. Mod Pathol. 1999;12(8):786–793.

19. Plumb SJ, Argenyi ZB, Stone MS, De Young BR. Cytokeratin 5/6immunostaining in cutaneous adnexal neoplasms and metastatic adenocarcino-ma. Am J Dermatopathol. 2004;26(6):447–451.

20. Qureshi HS, Ormsby AH, Lee MW, Zarbo RJ, Ma CK. The diagnostic utilityof p63, CK5/6, CK 7, and CK 20 in distinguishing primary cutaneous adnexalneoplasms from metastatic carcinomas. J Cutan Pathol. 2004;31(2):145–152.

21. Ivan D, Nash JW, Prieto VG, et al. Use of p63 expression in distinguishingprimary and metastatic cutaneous adnexal neoplasms from metastatic adenocar-cinoma to skin. J Cutan Pathol. 2007;34(6):474–480.

22. Ivan D, Hafeez Diwan A, Prieto VG. Expression of p63 in primarycutaneous adnexal neoplasms and adenocarcinoma metastatic to the skin. ModPathol. 2005;18(1):137–142.

23. Liang H, Wu H, Giorgadze TA, et al. Podoplanin is a highly sensitive andspecific marker to distinguish primary skin adnexal carcinomas from adenocar-cinomas metastatic to skin. Am J Surg Pathol. 2007;31(2):304–310.

24. Ormsby AH, Snow JL, Su WP, Goellner JR. Diagnostic immunohisto-chemistry of cutaneous metastatic breast carcinoma: a statistical analysis of theutility of gross cystic disease fluid protein-15 and estrogen receptor protein. J AmAcad Dermatol. 1995;32(5, pt 1):711–716.

25. Bhargava R, Beriwal S, McManus K, Dabbs DJ. CK5 is more sensitive thanCK5/6 in identifying the ‘‘basal-like’’ phenotype of breast carcinoma. Am J ClinPathol. 2008;130(5):724–730.

26. Dabbs DJ, Chivukula M, Carter G, Bhargava R. Basal phenotype of ductalcarcinoma in situ: recognition and immunohistologic profile. Mod Pathol. 2006;19(11):1506–1511.

27. Feldman AL, Dogan A. Diagnostic uses of Pax5 immunohistochemistry.Adv Anat Pathol. 2007;14(5):323–334.

28. Mhawech-Fauceglia P, Saxena R, Zhang S, et al. Pax-5 immunoexpressionin various types of benign and malignant tumours: a high-throughput tissuemicroarray analysis. J Clin Pathol. 2007;60(6):709–714.

29. Carlsten JR, Lewis MD, Saddler K, et al. Spiradenocylindrocarcinoma: amalignant hybrid tumor. J Cutan Pathol. 2005;32(2):166–171.


an immunohistochemical panel to differentiate

Health & Medicine

ihc panel of cmbc

sgc cases

ductal carcinoma

panel of ihc stains

negativebreast carcinoma

accurate cmbc cases

sweat gland carcinoma

cases of ductal cmbcs