Androgen receptor expression and breast cancer mortality in a population-based prospective cohort

PURPOSE: The increase in clinical trials with androgen receptor (AR)-targeting drugs emphasizes the need of clarifying the role of AR expression in different breast cancer subtypes. AR confers good prognosis in estrogen receptor positive (ER+) breast cancer, but its role in ER-negative (ER−) breast cancer is unclear. The aim of this study was to elaborate on previous findings of a differential prognostic role for AR depending on ER status, using breast cancer mortality (BCM) as endpoint, in a population-based cohort from the Malmö Diet and Cancer Study. METHODS: Immunohistochemical AR expression was assessed in 910 women with invasive breast cancer diagnosed 1991–2010, supplemented with clinicopathological information, vital status, and cause of death, with the last follow-up in December 2014 (median 10 years). Survival analyses according to AR status and AR/ER combinations were performed. RESULTS: AR expression was available for 671 tumors. AR+ (n = 573, 85%) was associated with favorable established tumor markers and lower BCM in univariable analysis, especially during the first 5 years following diagnosis [HR 0.4; 95% confidence intervals (CI) 0.2–0.7]. Multivariable analysis for short-term follow-up indicated higher BCM among patients with AR+ER− tumors (HR 3.5; 95% CI 1.4–9.1) than other AR and ER combinations. CONCLUSIONS: AR expression added prognostic information to ER expression with respect to short-term prognosis. The worst prognosis was seen for patients with AR+/ER− tumors in short-term follow-up, supporting the pre-specified hypothesis. However, larger cohorts are needed for further characterization of the role of AR expression in ER− breast cancer.