Background The objective of the study was to evaluate the implications of androgen receptor (AR) in breast cancers. Patients and methods We investigated immunohistochemical AR expression from the tissue microarrays of 931 patients between 1999 and 2005, and analyzed demographics and outcomes using uni-/multivariate analyses. Tumors with ≥10% nuclear-stained cells were considered positive for AR. Results AR was expressed in 58.1% of patients. AR was significantly related to older age at diagnosis, smaller size, well-differentiated tumors, higher positivity of hormone receptors, non-triple-negative breast cancers (non-TNBCs), and lower proliferative index. In estrogen receptor (ER)-negative tumors, AR was distinctively associated with human epidermal growth factor receptor type 2 (HER2) overexpression. With a mean follow-up of 72.7 months, AR was positively related to survival in ER-positive but not in ER-negative tumors. In Cox’s models, AR was an independent prognostic factor for disease-free survival in ER-positive cancers. Interestingly, molecular apocrine tumors (ER negative and AR positive) with HER2 positive status showed trends of poorer outcome, but AR had no impact on survival in patients with TNBC. Conclusions AR is significantly associated with favorable features in breast cancers and related to better outcomes in ER-positive not in ER-negative tumors. These results suggest that AR could be an additional marker for endocrine responsiveness in ER-positive cancers and a candidate for therapeutic targeting of ER-negative tumors.
