BRCA1 and BRCA2 gene mutation carriers are at risk of developing breast cancer earlier than other patients. Breast cancer associated with this mutation has characteristic histopathological features: (i) the expression of estrogen and progesterone receptors is less frequently demonstrable, (ii) the grade of histopathological malignancy is higher and (iii) accumulation of p53 protein is observed more often than in sporadic cases of this malignancy [8,9]. Although these factors are usually associated with a poorer prognosis, their role in BRCA1 and BRCA2 mutation carriers is still controversial [10-15].

The role of tamoxifen in preventing the development of contralateral breast cancer in BRCA1 mutation carriers is not fully understood since it significantly reduces that risk despite low expression of ER [16]. The mechanism responsible for that has not been yet explained and estrogen receptor β may play a role here.

Estrogen receptor β (ERβ) was discovered in 1996 and was given its name in order to differentiate it from the previously known type of estrogen receptor (now named estrogen receptor α – ERα) [17,18]. The two estrogen receptors belong to a family of ligand-regulated transcription factors. They are transcripts of different genes sharing some structural similarities. When co-expressed, ERα and ERβ may form homo- or heterodimers upon binding specific ligands. As dimers, ERs are able to start transcription activity in two ways: through direct binding to specific regions of DNA, or through protein-protein interaction with other transcription factors. In the case of co-expression of both ERs, their roles may overlap. In certain situations, however, ERβ opposes the activity of ERα via the inhibition of ERα-mediated gene expression. These differences are also observed in the response to tamoxifen. This selective estrogen receptor modulator may work as a pure ER antagonist for ERβ, while it may have a partially agonistic effect for ERα [19]. In spite of increasing knowledge regarding the structure and in vitro activity of ERα and ERβ, their clinical role is still controversial and unclear [20].

Results of genetic tests were obtained from the Prophylactics and Epidemiology Center in Poznań. In the search for mutations in the BRCA1 gene, tests were performed on DNA isolated from peripheral blood lymphocytes using a commercially available kit. They were carried out for three most common mutations of BRCA1 gene affecting the Polish population (5382insC, C61G, 4153delA). In Poland those three mutations in BRCA1 account for 86% of all BRCA1 and BRCA2 mutations. In the search for 5382insC and 4153delA mutations, the ASA-PCR method was used. For the detection of C61G mutation, the RELP-PCR method was applied.

Histological and immunohistochemical tests were completed in the Department of Tumor Pathology of the Medical University in Poznań. From formalin fixed and paraffin embedded specimens, 4 μm sections were cut and mounted onto positively charged glass microscope slides (Superfrost Plus, Menzel-Glaser, Germany). The sections were deparaffinized, rehydrated and subjected to antigen retrieval in citrate buffer in a microwave oven. The slides were then incubated in 3% hydrogen peroxide for 10 minutes to block endogenous peroxidase activity.

The primary antibodies used for immunostaining were as follows:

1. for ERα – a monoclonal mouse antibody clone 1D5 (Dako, Glostrup, Denmark; code No M 7047) was used at 1:50 dilution and slides were incubated at room temperature for 1 h;

2. for ERβ – a polyclonal rabbit antibody in which immunogen corresponds to NH2-terminus of the human ERβ, and, according to the manufacturer, the sequence used is conserved in all known isoforms (Chemicon International, Temecula, CA; catalog No AB1410). This antibody was used at a dilution of 1:500 and overnight in an incubation chamber at 4°C;

3. for PgR – a monoclonal mouse antibody clone PgR636, (Dako, Glostrup, Denmark; code M 3569) was used at a dilution of 1:100 and slides were incubated at room temperature for 1 h. This antibody (according to the manufacturer) has been demonstrated by Western blot to react with both forms of the progesterone receptor: the PR-A and PR-B.

The antibody reactions were revealed using Dako EnVision™+System-HRP. Subsequently the slides were incubated in DAB chromogen for 5 minutes at room temperature. The sections were counterstained with hematoxylin, and coverslipped.

Immunostained slides were evaluated by two independent observers in a "blinded" fashion, using light microscopy. Tumors were considered to be expressing receptors if a positive reaction (regardless of its intensity) was identified in at least 10% of the cancer cells' nuclei.

HER-2 protein was identified by means of the HecepTest™ kit (Dako). Tumors were classified as 0, 1+, 2+ or 3+ on the basis of the spread and intensity of membranous staining in invasive portions of the cancer. Tumors showing staining reactions in class 3+ were considered to be over-expressing HER-2. In the case of the 2+ results obtained, suggestive of over-expression of the HER-2 protein, amplification of the c-erbB2 gene was tested by the application of fluorescent in-situ hybridization (FISH). Cases with an immunohistochemistry result of 2+ but without amplification of the gene, as tested by FISH, were not considered to be over-expressing HER2.

The average age of patients with mutations in the BRCA1gene was 45 years, and 57 years in the control group.

Invasive ductal carcinoma (no special type) was found to occur at the same frequency in both the BRCA1-positive group as well as the control group (87% and 85.8%, respectively). Other tumor types were much less frequent, although it is noteworthy that all 5 patients with "medullary like" carcinoma came from the BRCA1 mutation carriers group (10%).

The assessment of histological malignancy was performed according to the Bloom-Richardson classification and showed that patients carrying BRCA1 mutations developed grade 3 tumors more often and had less frequent metastases to axillary lymph nodes.

In the BRCA1-positive group, the rate of patients who developed bilateral breast cancer accounted for 27.1% (13 cases). In the control group, bilateral breast cancer was found in 2 patients only (1.7%).

The presence of estrogen receptor α was detected in the tumors of 7 (14.5%) patients with BRCA1gene mutations and in 69 patients (57.5%) in the control group (Fig. ​(Fig.11).

For progesterone receptors, the figures were 6 (12.5%) and 77 (64%), respectively. The expression of estrogen receptor β was detected in the tumors of 20 (42%) of the BRCA1-mutation patients and in 66 (55%) of the subjects in the control group (Table ​(Table2)2) (Fig. ​(Fig.22).

In the group with BRCA1 mutations, the most frequently found phenotype was ERα(-)/ERβ(-) observed in 52% of patients, followed by ERα(-)/ERβ(+) noticed in 33.3% cases. In the control group only 14.2% of the patients presented phenotype ERα(-)/ERβ(+). The most common phenotype observed in that group was ERα(+)/ERβ(+) – 40.8% (Table ​(Table33).

The over-expression of the HER-2 receptor in patients with mutated BRCA1 genes was observed only in 3 cases (6.2%). In the control group over-expression of HER-2 was found in 20 cases (16.7%).

Most of our control group patients were tested for the presence of BRCA1 mutations. Although those tests were not performed on all the controls, this cannot influence the overall findings of the study since in Poland BRCA1 mutations (unselected for age) occur only in about 3% of all breast cancer patients [23].

It has been pointed out in many studies that BRCA1 positive breast cancers rarely express ERα [8-12,20,21]. Further confirmation of that observation comes from recent studies made with the use of the cDNA microarray technique, showing that breast cancers may vary considerably in their molecular profile [24,25]. The largest group consists of tumors with a cellular profile matching that of the inner layers of the mammary glands, such as luminal cells (luminal types) and another group consists of tumors with basal cell profiles (basal types). These two groups differ in their expression of ERα. The luminal types express ERα, while the basal types usually do not. cDNA studies of breast cancers with underlying BRCA1 gene mutations show that the most common cancer type in this group is basal [26]. Moreover, basal type tumors show expression of receptors for epidermal growth factor (EGFR) more often. Generally, about half of the patients with BRCA1 gene mutations present the phenotype ER(-)/HER2(-)/EGFR(+) of the basal type of tumor, which is usually associated with poor prognosis [26-28]. On the other hand, there are data that suggest that the prognosis for BRCA1-positive patients is no worse than that for patients without mutations [10-14]. This discrepancy may suggest the influence of some incompletely explored prognostic and predictive factor(s).

Recent analyses suggest that ERβ may be an independent prognostic and predictive factor in the course of breast cancer [29-33]. However, its role has not yet been fully discovered. It is noteworthy that in our clinical material (in contrast to ERα) the expression of the ERβ in tumor tissues of patients with BRCA1 gene mutations is almost as frequent as in the non-hereditary breast cancer patients. It would be interesting to test whether frequent representation of ERβ in BRCA-positive tumors is actually responsible for the earlier reported activity of tamoxifen in the reduction of contralateral breast cancer risk.

Fan et al. showed that there is a link between estrogen receptors and the BRCA1 protein. In an in vitro study, BRCA1 protein proved to be one of the transcription regulators for active ERα. The transcription co-activator p300 plays an important role in this process, and its presence correlates with the ability of the BRCA1 protein to suppress the activation of ERα transcription [7].

Our study group displayed certain characteristics of hereditary breast cancer. One of these features was the young age of the patients. In this group, the average age of breast cancer diagnosis was 45 years, 12 years earlier than in the control group. The same average age at which the diagnosis was established, was reported for Spanish families and a similar age (43 years) in Markus's study of the population of Ashkenazi Jews [10]. Age is a well known factor influencing the expression of ERα. The fact that there was a 12-year difference in the average age between the BRCA1 gene mutations group and the control group must have some influence on the reported ERα expression in our study. Regarding ERβ expression, however, we believe that difference bears no material significance on the result. Our opinion is based on the fact that numerous studies have indicated that the expression of ERβ is not age-dependent.

In both groups, the most frequent histological diagnosis was that of ductal carcinoma (87% and 85.8%, respectively). Another common histological type of breast cancer found in the mutation carriers group was "medullary like" carcinoma (10%). In the control group, however, not a single case of that type of cancer was identified. A higher rate of "medullary like" carcinoma is typical of BRCA1 mutation carriers and this diagnosis is associated with a relatively good prognosis [34,35]. Consequently, it was believed that this diagnosis positively affected the prognosis for BRCA1-positive patients. Unfortunately, a recent study failed to confirm that finding [36].

Patients with hereditary disease have been found to develop bilateral breast cancer more often. Our study has confirmed this observation as it has been found that bilateral breast cancer affected 13 patients of the study group (27.1%).

The above characteristics of hereditary breast cancer have been frequently described in the literature and are consistent with the previously published data. Little, however, has been said regarding the expression of ERβ. This appears to have been mentioned only once in a letter to the Editor of the Journal of Clinical Oncology which reported on estrogen receptor-beta expression in hereditary breast cancer where positive staining for ERβ was detectable in 94% (15 of 16) of BRCA1 associated breast cancers [37].

In our study, ERβ expression was present in 42% of BRCA1 mutation carriers. This seems to be important as it may explain the protective effect of tamoxifen in the prevention of contralateral tumor development in BRCA1 mutation carriers.

Out of the hereditary cancer patients in our study group, 75% were found to be triple-negative, lacking ERα, PgR and HER-2 over-expression. Unfortunately, at the current state of the art, neither hormonal therapy, nor immunotherapy (trastuzumab) in an adjuvant setting, can be offered to those patients.

Gruvberger-Saal et al [38] reported that the expression of ERβ is an independent marker for favourable prognosis after adjuvant tamoxifen treatment in ERα negative breast cancer patients. Bearing in mind that as many as 44% of the triple-negative patients in our study expressed ERβ, we propose that hormone therapy might be used in the treatment of BRCA1 mutation carriers expressing the ERα(-)/PgR(-/HER2(-)/ERβ(+) phenotype.

ERα – estrogen receptor alpha

ERβ – estrogen receptor beta

PgR – progesterone receptor

BRCA1 – breast cancer associated gene

BRCA2 – breast cancer associated gene

PCR – polymerase chain reaction

HER-2 – human epidermal growth factor receptor-2

SERM – selective estrogen receptor modulator

IHC – immunohistochemistry

ASA-PCR – amplicon sequence analysis – polymerase chain reaction

RFLP-PCR – restriction fragment length polymorphism – polymerase chain reaction

FISH – fluorescent hybridization in situ

EGFR – epidermal growth factor receptor