Preferential expression of NY-BR-1 and GATA-3 in male breast cancer

In metastatic cancer, the knowledge of the site of origin is a key factor in patients’ management. Indeed, most of the therapies available rely on tissue-specific features and tumoral molecular characterization. On the other hand, the identification of the site of origin of a newly discovered metastatic adenocarcinoma, especially in case of poorly differentiated cancers, can represent a diagnostic dilemma. To identify the site of the primary tumor, several immunohistochemical approaches have been developed (Dennis et al. 2005; Rubin et al. 2001); ideally, each is aiming to reach the best specificity and sensitivity with the use of an ideal number of marker combination.

Male breast cancer is an uncommon disease, accounting for < 1% of all breast tumors. Due to its rarity and to the absence of a population-based screening, it is characterized by a consistent diagnostic delay and a more advanced stage at presentation than in female breast cancer. It is reported that from 7 to 30% of patients present with stage IV disease (Sanguinetti et al. 2016; Harlan et al. 2010; Fentiman et al. 2006) and 11% of these patients seek medical help merely for symptoms or evidences of extramammary dissemination.

NY-BR-1 is a novel marker of differentiation of the female mammary gland. The expression of this protein has been almost solely detected in breast tissue, at every level of cancer progression: 60–100% of primary lesions show positive staining. Metastasis are positive in approximately 50% of cases, with a concordance > 88–91% with the primary tumor (Jäger et al. 2007; Varga et al. 2006). Other markers of mammary differentiation such as GATA-3, mammaglobin, and BRST-2 (or gross cystic disease fluid protein 15) have been successfully employed in female breast cancer. These markers exhibit prognostic and therapeutic implications, represent important tools for detecting primary breast cancer and metastasis, and monitor lymph nodes during and after surgery (Wang et al. 2009; Kandalaft et al. 2016). Yet, none of these markers has been sufficiently investigated in male breast cancer.

The aim of this study was to analyze the expression of female breast cancer established markers NY-BR-1, GATA-3, mammaglobin, and BRST-2 in a cohort of male breast cancer in primary and metastatic lesions.

As male breast cancer is often diagnosed at an advanced stage, breast origin must be taken into account in case of newly discovered metastatic adenocarcinomas. In this study, we could show that immunohistochemical profile of the so-called breast markers can be reliably applied also to male breast cancer both in the primary and in the metastatic settings.

In this study, we observed a high sensitivity for all known female breast markers in a cohort of male breast cancer and in their corresponding axillary and/or distant metastases. Among the markers, NY-BR-1 and GATA-3 showed the highest expression.

NY-BR-1 is frequently expressed in normal breast, female breast cancer, and metastasis, whilst no significant expression has been detected in any other normal or neoplastic tissue (Varga et al. 2006). The highest sensitivity rate for NY-BR-1 was observed in normal glandular breast tissue and in ductal carcinoma in situ in females (Varga et al. 2006). Invasive female breast carcinomas were shown to express NY-BR-1 depending on histological grading, being mostly expressed in G1 and G2 carcinomas (Varga et al. 2006). Therefore, NY-BR-1 is considered both as a breast-specific marker and as a differentiation marker as well (Varga et al. 2006). Since our cohort comprised a significant number of grade 1 and 2 primary tumors (25/30) and metastases came mainly from neoplasms of the same grade of differentiation (5/8), positive cases for NY-BR-1 were likely to represent the vast majority in this study. As NY-BR-1 was shown to be inversely related to breast differentiation, problems may arise in clinical practice when considering distant metastasis from a poorly differentiated carcinoma. Assumed the peculiar specificity of NY-BR-1, it is to emphasize that high-grade lesions may lose the expression of the protein more frequently than reported in the present study. Nevertheless, the concordance highlighted between metastases and matched primary tumors (100%) encourages the application of this marker also in metastatic setting.

GATA-3 is a protein linked to estrogen receptor pathway. It is not surprising that we found a high prevalence of positive cases in a group with 96.6% of ER + tumors. Male breast cancer is expected to be ER + in a higher percentage of case rather than female breast cancer (Doebar et al. 2017).

The role of GATA-3 has been extensively investigated in female breast pathology.

Sangoi et al. compared the sensitivity of GATA-3 to mammaglobin and GCDFP-15 (BRST-2), and found that GATA-3 showed superior sensitivity in both metastatic carcinoma and matched primary tumor. Moreover, IHC for GATA-3 harbored consistently fewer problems of interpretation, thanks to less background staining, than the other two markers (Sangoi et al. 2016).

In our subset of carcinomas, we observed the same tendency. The sensitivity of mammaglobin vs GATA-3 in labelling male breast cancer was 73.3 vs 100%, and BRST-2 vs GATA-3 was 60 vs 100%, respectively. Expression in metastases showed similar levels of concordance: 75% for both GATA-3 and mammaglobin, and 62.5% for BRST-2.

Although GATA-3 has been reported to be highly sensitive for breast tumors, it is not entirely specific. Commonly, GATA-3 expression characterizes tumors arising from skin adnexa, urothelium, salivary glands, and pancreatic ducts, whereas adenocarcinomas of lung, stomach, colon, and prostate showed positivity to a lesser extent (Miettinen et al. 2014).

To avoid clinical concerns arising from the lack of specificity of GATA-3 in the setting of metastatic adenocarcinoma, we carried further analysis on the same cohort with the novel marker NY-BR-1, which showed a high concordance between GATA-3 and NY-BR-1.

Mammaglobin, is a protein of unknown function, whose expression increases tenfold in breast cancer (Watson and Fleming 1996). It has not been detected in any other type of cancer, whereas in small amounts only in normal breast tissue.

The discovery of its specificity brought to employ this protein for labelling breast neoplasms. The expression varies between different subtypes: high in luminal and Her2 positive tumors and rather low in triple negative and basal-like tumors. Mammaglobin raised sensitivity and specificity in distinguishing between cutaneous metastases of breast carcinomas and sweat gland carcinomas (Lewis et al. 2011; Diel et al. 1996). Other studies conducted with RT-PCR analysis reported mammaglobin utility in monitoring lymph nodes, predicting the risk of metastases, and characterizing pleural effusions in breast cancer patients (Ciampa et al. 2004).

However, given its high specificity, IHC for mammaglobin in our study showed a sensitivity of 73.3%. These data make this marker only partially suitable for labelling male lesions, if adopted on its own.

BRST-2 (also known as gross cystic disease fluid protein GCDFP-15) is a breast marker expressed in up to 73% of female breast cancer (Fritzsche et al. 2007). His detection in specimens is clearly influenced by the histological and molecular subtype: sensitivity is high in tumors with apocrine and lobular signet ring features, as well as androgen receptor positivity which is also related to BRST-2 expression. On the other hand, medullary histological phenotype and triple negative intrinsic subtype showed lower rates of BRST-2 positivity.

Similar to GATA-3, this marker does not exclusively label breast cancer: it is expressed by apocrine glands, including those in the breast, the tracheobronchial tree, sweat gland carcinomas, and salivary gland carcinomas (Wick et al. 1998). Finally, an important overlap is observed in 5–6% of lung adenocarcinomas. In our cohort, we found that 60% of male breast cancer scored positive for BRST-2, enhancing a substantial lack of sensitivity.

Either in female or in male breast cancer, BRST-2 and mammaglobin may show a clinical value only when combined with other markers. In our study, we could show that NY-BR-1 possesses greater sensitivity than both the aforementioned markers.

In the setting of metastases of unknown origin, there are established immunohistochemical panels to apply. In case clinical history, complete physical examination, routine laboratory tests, imaging, and radio-metabolic techniques do not come to any conclusive results, the detection of tumor specific antigens or specific immunohistochemical stains can be also helpful.

For a subset of tumors, one marker alone may show the sufficient diagnostic accuracy.

In appropriate clinical context, the prostate-specific antigen (PSA) for prostate cancer, PAX8 for ovarian cancer, alphafetoprotein for liver cancer, and TTF1 for lung cancer can be reliably used as a single marker (Dennis et al. 2005; Rubin et al. 2001; Woodard et al. 2011).

If one marker alone does not possess sufficient specificity nor sensitivity for the suspected neoplasms, a combination of IHC stains may provide additional support. An initial approach with cytokeratin subtypes as CK7 and CK20 gives a first hint at the likely site of origin for carcinomas as to gastrointestinal or gynecological origin. These tests can be completed with CDX2 and CK19 for upper gastrointestinal tract, biliary ducts and pancreas, thyroglobulin for thyroid adenocarcinoma, NY-BR-1 and GATA-3 for female breast cancer, Mel-A, S100 and HMB45 for melanoma, and vimentin and Mel-A for kidney and adrenal gland tumors.

In our study we could show that NY-BR-1 and GATA-3 are highly sensitive breast markers in primary and metastatic male breast cancer. As already demonstrated in females, both NY-BR-1 and GATA-3 expression in male breast cancer outperforms the use of mammaglobin and BRST-2 alone. Using either NY-BR-1 or GATA-3, rather than mammaglobin and BRST-2 may help confirming breast origin both in primary and in metastatic carcinoma in males.