Matched-pair analysis of patients with female and male breast cancer: a comparative analysis

In contrast to the breast cancer in women, male breast carcinoma (MBC) is rare, accounting for less than 1% of all cases of breast carcinoma with an incidence of 1 in 100.000 men [1, 2]. In the United States up to 1.500 primary male breast cancer patients per year were identified [3]. Data from the SEER Program of the National Cancer Institute (NCI) showed a slight increase over time from 1.0 per 100,000 in the late 1970s to about 1.2 per 100,000 in 2000-2004 [4]. Age-standardized rates for male breast cancer revealed that the incidence peaked in the year 2000 at 1.24 per 100.000 for men whereas the peak in women was observed in 1999 at 165 per 100.000 women [1]. Because of its rarity, most information about this disease has been obtained from small, single-institutional or retrospective studies or by extrapolation from breast cancer trials in women [5]. Its phenotypic alterations are not well studied, and therapy is mainly based on experiences with female breast cancer [6].

Although both diseases share similarities, there are notable differences reported [7]. Breast cancer in men occurs with higher stage, possibly because of delayed breast cancer detection [2, 8]. Male breast cancers also are typically more low grade and frequently hormone receptor--positive [7]. The relatively unfavorable outcome in male breast cancer has been attributed to more advanced local tumor stage and high incidence of lymph node invasion at the time of diagnosis [9‐11]. It has been postulated that close proximity to skin and nipple facilitates early invasion of lymph vessels leading to earlier regional and distant metastasis.

Gender-comparative survival studies revealed divergent data on patients outcome [11‐16]. This might depend on the lacking adjustment for prognostic relevant parameters in comparative analyses between male and female cohorts. However, specific informations about disease outcome in men has emerged great attention since tailored therapy is of incremental importance in breast cancer.

Several groups have recently reported comparative analyses for male versus female breast cancers, but these studies failed to be adjusted for a variety of prognostic relevant parameters in breast cancer like age, tumor size, nodal stage, tumor grade hormone receptors and HER2 expression [10‐16].

To our knowledge this is the first comprehensive gender-specific survival analysis considering all established tumor and patients characteristics established in female breast cancer in order to elucidate potential differences in both disease-free and overall survival.

Consecutively in the years 1995 to 2007 113 men were diagnosed with breast cancer in the administrative district of Chemnitz in the State of Saxony, Germany. All data of these patients was obtained from the district's two cancer registers located in the cities of Chemnitz and Zwickau which together oversee a total of about 1.5 million inhabitants. Primary surgery was performed exclusively in community or district hospitals. Adjuvant treatment and follow-up care were carried out additionally in outpatient departments. With the approval of the institutional review boards data regarding patients' age, histology, TNM stage, tumor grade, date of diagnosis, date of metastasizing and date of death were recorded. In addition, detailed information was gathered on estrogen, progesterone and HER2 receptor expression, type of primary surgery, as well as on chemotherapy, endocrine treatment and radiation therapy in the adjuvant and palliative setting. The study was approved by the Ethic Committee of the University of Bonn, Germany. Patient data were collected retrospectively and blinded (name and date of birth). Therefore and in accordance with the Ethic Committee patient consent was not required for the inclusion into the study. For 108 male breast cancer patients one matching woman could be chosen from a total of 13.333 female breast cancer patients. The matching process was based on eight features relevant for breast cancer prognosis i.e. year of diagnosis (within ± 5 year) age, tumor stage, nodal stage, tumor grade, estrogen and progesterone receptor expression, as well as HER2 expression. Five men had to be excluded from the current study because no women with breast cancer could be matched. If more than one female patient was eligible the best match was chosen by random selection. The matching procedure was conducted blinded without any information about patients outcome.

To ensure the comparability of the male with the female patients the chi-square test was applied for every matching criterion. Overall survival (OS) and disease-free survival estimations were calculated according to Kaplan and Meier. Disease-free survival (DFS) was defined as the time period from diagnosis to death of recurrence or death whichever occurred first. Overall survival was considered as the time period from first diagnosis until death. The effect of gender on survival was estimated by hazard ratio (HR) and its 95% confidence interval (95% CI). P values < 0.05 were considered statistically significant. All statistical analyses were performed using SPSS version 17.0.

In the past decade it has emerged great attention to obtain evidences about the clinical outcome of male patients with breast cancer [17‐19]. This is due to its rising incidence and the persistent lack of established treatment guidelines. Actually, treatment strategies are derived from female breast cancer [7]. Without evidence-based data to support female-to-male extrapolation, epidemiologic comparisons become an alternative source of information.

A couple of features in male breast cancer are of particular interest. Male breast cancer was more like late-onset than female breast cancer with more than 90% of male patients aged 50 years and older. It is well known that breast cancer in men is diagnosed 5 to 10 years later than in women [20] which also might account for the less aggressive adjuvant treatment modalities given to male patients in this study.

Furthermore, more than 90% of male breast carcinomas were hormone receptor positive whereas less than 10% overexpressed HER2. This is in line with previous findings [21, 22] and supports the hypothesis that the luminal like molecular subtype is predominant in men [20, 23, 24]. As opposed to female breast cancer the progesterone receptor expression in men was an independent prognostic factor in this study.

Male patients had a significant proportion of advanced tumor stages (pT2-4 > 60%) and more than 40% showed nodal involvement [25]. This is at least in part caused by a delay of diagnosis in men of more than six months in average [26‐28]. There is little public awareness of breast carcinoma in men and public education regarding the existence of male breast carcinoma. No recommendations for self-examination or examination of the male breast by physicians exist. No guidelines recommend screening mammography at any age for men because of the rarity of the disease.

In male breast cancer adjuvant treatment was administered less frequently. Almost one third of men received no adjuvant treatment as compared to 10% in the matched-pair female group. Only 32% of male patients had adjuvant chemotherapy (54% in women) and less than 50% adjuvant endocrine treatment (65% in women). Nevertheless, our comparative study revealed no survival differences between male and female breast cancer patients and gives evidence that gender is no predictor for neither disease-free nor overall survival. These findings are remarkable considering the different systemic treatment given to male patients as compared to the matched female patients. Due to exact matching procedures with respect to all established clinical and pathological prognostic factors (Table 1), a gender specific bias of prognostic features seems unlikely.

In addition, considering clinically relevant subgroups no significant overall survival differences could be observed between men and women. In our study men with pT1 tumors had significantly worse DFS compared to females. Those male patients with pT1 tumors who have relapsed so far showed no poor prognostic features e.g. all of them were hormone receptor positive. Furthermore, all of them received adjuvant tamoxifen. Therefore, no specific predictors for metastasis in this low risk group could be identified. Potential explanations for a comparatively worse outcome of males with early breast cancer are obviously. There is no evidence that tamoxifen is highly effective in primary male breast cancer patient. Data supporting response to treatment are mainly retrospective or from small trials [29]. There are major concerns regarding patients compliance and tamoxifen metabolic activity which might have negative impact on the efficacy. Clinical response to tamoxifen depends on the biotransformation via the cytochrome P450 isoenzyme CYP2D6 isoform, to the active metabolite endoxifen. CYP2D6 activity can be reduced in men both by genetic variation or concurrent use of drug inhibitors, which can significantly reduce endoxifen plasma concentrations [30]. However, tamoxifen remains the standard therapy in endocrine responsive male breast cancer and clincial and translational studies on this field are needed.

Our cohort study has several limitations and selection bias regarding the match pair analysis could not be excluded. Beside its retrospective character the number of male patients is small compared to female studies. Therefore analyses especially of patients subgroups has limited clinical significancy.

Several other groups compared male and female breast cancer features and prognosis [10, 11, 15, 31‐33] with divergent results. Marchal and colleagues found similar DFS but worse OS for male as compared to female breast cancer patients in a retrospective study [11]. The authors explained the OS difference with a higher amount of comorbidities in men. However, the study was based on matched-pair analyses of only 58 male breast carcinomas and three matching criteria (age, year of diagnosis, stage). Tumor differentiation, hormone receptor status and HER2 positivity were not considered. This might reflect an incomplete matching procedure and could better explain the divergent findings to our results.

Our retrospective study showed no disease and overall survival differences between male and female breast cancer patients matched for year of diagnosis, age, tumor size, nodal stage, tumor grade, estrogen receptor, progesterone receptor and HER2 expression. This is of great importance since male patients receive obviously less aggressive adjuvant treatment.comparted their female matched pairs. A significantly reduced DFS in male with pT1 tumors was found. In multivariate analyses only early stage and progesterone receptor positvity were statistically significant related to improved overall survival in male breast cancer.