Treatment Outcomes in Male Breast Cancer: A Retrospective Analysis of 161 Patients

AIMS: Male breast cancer is a rare disease with limited evidence-based guidelines for treatment. This study aimed to identify demographic, pathological and clinical factors associated with its prognosis. MATERIALS AND METHODS: A retrospective review of 161 male breast cancer patients diagnosed at a single institution from 1987 to June 2017 was conducted. Patient demographics, disease characteristics, treatment and outcome were extracted and included in competing-risk analysis and the univariate Cox proportional hazard model for univariate analysis. Factors with P < 0.10 were included in multivariable analysis. RESULTS: The mean age at diagnosis was 67 years (standard deviation = 11.2) and the median follow-up duration was 5.3 years (range 0-25 years). There were 48 deaths, including 23 cancer-specific deaths. The actuarial median survival was 19.9 years. In multivariable analysis, factors associated with overall survival were size of tumours (hazard ratio 2.0; 95% confidence interval 1.4-2.7, P < 0.0001) and diagnosis of metastatic disease (hazard ratio 8.7; 95% confidence interval 1.9-40.6; P = 0.006). Of 138 patients without metastases at diagnoses, 11 had local-regional recurrence and 26 had distant metastases. In the multivariable model for local-regional recurrence, a more recent year of diagnosis was associated with reduced risk (hazard ratio 0.9, 95% confidence interval 0.8-1.0, P = 0.008), whereas more positive lymph nodes was associated with higher risk (hazard ratio 2.2, 95% confidence interval 1.2-4.0, P = 0.01). A higher risk of metastases was associated with more positive lymph nodes (hazard ratio 1.9; 95% confidence interval 1.1-3.3; P = 0.03) and tumour size (hazard ratio 1.8; 95% confidence interval 1.1-2.9; P = 0.01). A higher risk of any recurrence or metastases was associated with the number of positive nodes (hazard ratio 1.9; 95% confidence interval 1.2-3.0; P = 0.005) and tumour size (hazard ratio 1.6; 95% confidence interval 1.1-2.2; P = 0.01). CONCLUSION: In general, tumour size and more positive lymph nodes were associated with worse prognosis. Larger powered studies are needed to identify prognostic factors with smaller effect sizes.