Impact of Androgen Receptor Expression and the AR:ER Ratio on the Survival Outcomes in the Diverse Subgroups of Vietnamese Breast Cancer: A Single Institutional Retrospective Cohort Analysis.

The androgen receptor (AR) has recently emerged as a useful marker for the more favorable prognosis and better outcomes among women with estrogen receptor (ER) + ve breast cancer (BC) and the further refinement of BC subtype. Furthermore, AR expression in ER - ve tumors has a particular prognostic significance. Additionally, the ratio of nuclear AR to ER may critically have an influence on tumor biology and respond to endocrine therapy. To define the AR expression and AR:ER ratio, and explored their correlation with the clinicopathological features, prognosis, and survival outcomes in the various subclasses of invasive BC. The current study was conducted on 522 BC patients who had surgical operations, without neoadjuvant chemotherapy by applying a retrospective cohort analysis. The clinicopathological characteristics were recorded. Immunohistochemical staining was performed on AR, ER, PR, HER2, and Ki67. Expression of AR was paired into different immunophenotypes for analysis with clinicopathological features and survival. All BC patients' survival was analyzed using Kaplan-Meier and log-rank models. The presence of AR was detected in 65.3%. Positive AR, the ratio of AR:ER<2, luminal androgen receptor (LAR) + and AR + HER2 + immunophenotypes were significantly associated with better prognostic features. AR:ER<2 was observed in the prolonged overall survival (OS) and disease-free survival (DFS) (87.9 and 86.2%, respectively) compared to AR:ER≥2 (25.0% in both) ( < .001). In contrast, in HR + ve BCs, the AR expression was not significantly correlated with survival. The multivariate model revealed that the ratio of nuclear AR to ER remained as an independent prognostic variable. The AR expression had a distinct OS and DFS. The AR:ER ratio is an independent indicator for predicting the OS and DFS of BC patients in both univariate and multivariate analyses.