Immunohistochemical characterization of subtypes of male breast carcinoma

Male breast cancer is an uncommon disease, accounting for approximately 1% of all breast cancer cases and less than 1% of all malignancies in men [1, 2]. The American Cancer Society estimates that 1,910 men will be diagnosed with breast cancer in the USA in 2009 and about 440 men will die from the disease [3]. Although breast carcinoma in both genders shares certain characteristics, there are notable differences in incidence, age distribution, prognosis and survival. The unfavorable overall prognosis in male breast cancer has been attributed to the older age and advanced tumor stage at the time of diagnosis [4].

Based on the recent DNA microarray studies on female breast cancer cases, distinct molecular subtypes of breast carcinoma were identified with different clinical outcomes [5]. Using an intrinsic set of 534 genes, Sorlie and colleagues [6] analyzed the expression profiles of 115 independent breast cancer specimens and categorized them into the following subtypes: luminal; human epidermal growth factor receptor 2 (HER2) over-expressing; normal breast-like; and basal-like. The tumor subtypes correlated well with clinical outcomes as measured by overall survival and distal metastasis, with the worst outcome observed in HER2 over-expressing and basal-like subtypes [5‐7].

The molecular subtypes of female breast cancer were originally identified by gene expression analysis using DNA microarrays. However, large-scale subtyping using gene expression profiling from formalin-fixed, paraffin-embedded samples is not currently feasible. Therefore, immunohistochemical (IHC) markers have been used as surrogates for DNA microarray in subtyping breast cancer. By using a panel of four antibodies including estrogen receptor (ER), HER1, HER2 and cytokeratin 5/6 (CK5/6), Nielsen and colleagues [8] characterized three IHC-defined subtypes: luminal (ER+, HER2-), HER2 (HER2+), and basal-like (ER-, HER2-, CK5/6+ or HER1+). Based on more recent gene expression studies, Carey and colleagues [9] updated IHC subtype definition as luminal A (ER+ and/or progesterone receptor (PR)+, HER2-), luminal B (ER+ and/or PR+, HER2+), HER2+/ER- (ER-, PR-, HER2+), basal-like (ER-, PR-, HER2-, CK5/6+), and unclassified (negative for all five markers).

In addition to these subtypes, the prognosis of breast cancer may be affected by expressions of two additional molecular markers. Epidermal growth factor receptor (EGFR) was reported to be over-expressed in aggressive female breast cancer [10‐13] and is currently being evaluated in clinical trials as a potential therapeutic target [14‐16]. It was also reported that transcription factor nuclear factor κB (NF-κB) was activated in breast cancer cells with over-expressed EGFR [17]. The elevation of NF-κB was speculated to be associated with unfavorable prognosis by promoting metastasis [18, 19], inhibiting apoptosis [20], and increasingly resistance to chemotherapy [21]. Increasing evidence has suggested that NF-κB is a potential target for the treatment of breast cancer and the prevention of metastasis [22, 23].

Despite these advances in female breast carcinoma, our understanding and treatment strategies for male breast cancer are limited and generally extrapolated from our knowledge of female breast cancer. In particular, the molecular subtypes of male breast cancer have not been studied although the subtypes were reportedly associated with both biological and clinical behaviors in female breast cancer. This article reports our attempt to subclassify male breast carcinomas based on the immunoprofile, and to evaluate the association of the subtypes with histologic type, nuclear grade, lymph node status, clinical stage, and expression of EGFR and NF-κB.

The clinicopathological features of breast carcinoma from 42 men are summarized in Table 2. The racial distribution was as follows: 28 Caucasians, seven African Americans and seven Hispanics. Ages of the patients ranged from 33 to 82 years with a mean age of 63 years. Patients with luminal B subtype tumor was slightly younger than those with luminal A subtype (mean age 58 vs. 64 years, respectively), but the difference was not statistically significant (P = 0.25). Most of the patients were diagnosed with stage I to II disease (81%, 34/42) and were treated surgically (95%, 40/42), either by lumpectomy or mastectomy. Many patients also received other treatments including hormonal therapy (88%, 36/42), chemotherapy (64%, 27/42, with 2/27 received pre-operatively), and radiation therapy (19%, 8/42). There were no significant differences identified between the subtypes of male breast cancer regarding patients' race, presence of distant metastasis, and clinical stage.

Although the molecular subtypes of breast cancer were originally identified by gene expression analysis using DNA microarrays, IHC markers have been used as surrogates for DNA microarray in subtyping breast cancer. Based on recent updated IHC subtype definitions [9], we identified two molecular subtypes of male breast carcinoma in our group as luminal A (83%) and luminal B (17%) subtypes. Basal-like, HER2+/ER-, and unclassified subtypes were not observed in this group. This distribution differs from large study series in female breast carcinoma [8, 9, 24, 25], which reported much lower frequencies of luminal A subtype (51 to 69%) but significantly higher frequencies of basal-like (12 to 21%) and HER2+/ER- (7 to 12%) subtypes. Our study also showed that male breast carcinomas in this group universally expressed ER (100%), a significantly higher frequency than that in female breast carcinomas (60 to 69%) [4, 9]. This result was in agreement with previous studies showing that male breast carcinoma has a higher percentage of ER positivity (81 to 100%) than female breast cancer or gynecomastia [4, 23, 26, 27]. The knowledge of immunophenotype-based subtyping of male breast carcinoma may be applied in comparison of treatment responses and prognosis in matched subtypes of female breast carcinoma.

The HER2/neu gene is a member of a gene family encoding transmembrane receptors for growth factors, including EGFR, HER2, HER3, and HER4. Approximately 25 to 30% of invasive female breast cancers over-express HER2 [28]. However, studies of HER2 over-expression in male breast cancer are limited with conflicting results. Pooled data demonstrated that HER2 over-expression detected by IHC was in a range of 2 to 56% with an average of 23% (reviewed in [29]). The inconsistency in percentage of positive HER2 immunostain may be due to differences in antibody preparation, scoring systems, and cut-off values used in these studies. A recent study of 99 cases of male breast cancer demonstrated that 15% of the tumors were 2+ (equivocal) or 3+ (positive) on HER2 immunostain and 11% were positive using FISH analysis [29]. In the current study, we detected 7 of 42 (16%) of male breast cancers with HER2 gene amplification determined using both IHC (3+) and FISH analysis. Although the early data tended to show higher percentages of positive HER2 immunostain, we believe that true HER2-positive cases in male breast cancer are probably fewer than those of female breast carcinoma when the current American Society of Clinical Oncology/College of American Pathologists guidelines for IHC with FISH analysis are applied.

Over-expression of HER2 is a well-known prognostic factor associated with poor survival in women with breast carcinoma [5, 28]. However, the data are limited in male breast cancer regarding the association between HER2 over-expression and survival. In an early analysis of 17 cases of invasive male breast cancer, HER2 over-expression was detected by IHC in seven (41%) patients who were associated with significantly shortened survival [30]. Nevertheless, in a recent study, male breast cancers failed to show an association between HER2 over-expression and lymph node metastasis, a known unfavorable prognostic factor [29]. In our study, the luminal B subtype tumors (phenotype HER2+/ER+ and/or PR+) were more frequently associated with high nuclear grade, an unfavorable factor in both women and men [31]. More frequently, patients with luminal B subtype tumors had nodal metastasis and an increased number of positive lymph nodes. Although these findings suggest that carcinoma with HER2 over-expression may be associated with unfavorable prognostic factors, the differences were statistically insignificant due to our small cohort and a relatively short period of follow-up. Obviously, a high percentage of ER expression in male breast carcinoma is inevitably associated with the rarity of HER2+/ER- subtype carcinoma, which reportedly has a worse prognosis than that of luminal B tumors (HER2+/ER+ or PR+) in female patients. This may raise questions on the correlation in prognosis of the breast cancer subtypes between males and females. It is unknown whether the paucity or absence of the unfavorable breast carcinoma subtypes in males, such as HER2+/ER-, basal-like, or unclassified subtypes, is associated with a favorable prognosis. It is equally important whether the same subtypes of breast carcinoma have similar responses to therapy in males compared with females. To answer these questions, further investigation by comparison of prognosis and therapy responses with matched subtypes of breast carcinoma in males and females is warranted.

Based on studies in female patients with breast cancer, EGFR was reportedly over-expressed in aggressive [10‐13] and metastatic breast carcinomas [32]. In breast carcinomas with EGFR over-expression, NF-κB was reportedly activated [17], which was also linked to unfavorable prognosis by promoting tumor metastasis and inhibiting apoptosis [18, 20]. The data on EGFR expression in male breast cancer are very limited with only a few earlier studies conducted more than a decade ago. The reported expression rate of EGFR in male breast carcinomas varied significantly from 8.5% to 76% [31, 33, 34], probably due to variations in antibody preparation, staining protocol, and interpretation criteria. In our study, 21% of the tumors expressed EGFR, which is in agreement with one of the previous studies (20%) [31]. Interestingly, we found stronger EGFR staining was present at the leading edge of tumor invasion in high-grade tumors of luminal B subtype. We speculate that the higher expression rates for EGFR and NF-κB in luminal B subtype is an important factor contributing to their aggressive behavior and poor clinical prognosis. However, although higher expression rates of EGFR and NF-κB were observed in the luminal B subtype than those in the luminal A subtype, the difference lacked statistical significance due to the small study group.

In summary, the current study attempts to characterize subtypes of male breast cancer by using IHC markers. It provides new information for future study on prognosis, drug trials, and clinical management of breast cancer in men. However, our data need to be interpreted with caution for several reasons. Firstly, the study was designed on the basis of published data from female breast cancer that allowed subclassification of these tumors according to their characteristic features in DNA microarray and expression profiles [5, 7]. DNA microarray data on male breast cancer are currently not available because of the rarity of this malignancy. Therefore, it is uncertain whether the gender differences present in these tumors regarding the genetic features and associated expression profiles. Secondly, the current study was unable to provide correlative data between the immunosubtypes of male breast cancer and their clinical behavior and survival information due to a relatively short period of follow-up. It is certainly critical in future studies to identify the subtypes of male breast carcinoma and their association with survival as well as the candidate markers with potential prognostic and predictive values. Finally, due to the rarity of the disease, the small number of cases collected in this group may affect validation of its conclusion and the statistical power of the observations. Additional studies with larger numbers of patients are needed to achieve sufficient statistical power.

Our study demonstrates that luminal A and luminal B subtypes are the major subtypes of male breast carcinoma. The immunophenotypical features of male breast cancer differ from those of its female counterpart. Luminal B tumors tend to have high nuclear grade and more frequent expression of EGFR and NF-κB.