Prognostic value of androgen receptor (AR) in breast cancer: A systematic review.

e12762 Background: The androgen receptor (AR) is a potential therapeutic target for breast cancer (BC). Nonetheless, the utility of AR is being called into question due to inconsistent data about its prognostic value. This meta-analysis aimed to investigate AR expression and its impact on overall survival (OS) and disease-free survival (DFS). Methods: PubMed, Medline, Web of Science, Scopus, Cochrane Library, Embase, Google Scholar, and CINAHL databases were searched to identify relevant studies that assessed the relationship between AR expression and prognosis in patients diagnosed with BC. The hazard ratios (HR) and confidence intervals (CI) of OS and DFS were pooled and displayed as forest plots. Additionally, Cochran’s Q and I2 statistics were used to assess heterogeneity. Funnel plots and Egger’s tests were performed to evaluate potential publication bias. Finally, subgroup analysis was conducted to assess the influence of the AR cutoff on OS and DFS outcomes. Results: Of the 1465 articles identified, 28 met the inclusion criteria and were included in this meta-analysis. AR positivity defined by IHC was associated with improved OS (HR, 0.904; 95% CI, 0.848–0.964; P = 0.002) and DFS (HR = 0.810; 95% CI = 0.738–0.890; P = 0.000) in all BC patients. Furthermore, we noticed that the prognostic value of AR depends on the BC subtype. Indeed, neither the triple negative breast cancer (TNBC) subtype (HR = 0.919; 95% CI = 0.786-1.075; p = 0.290) nor the unrelated BC subtype (HR = 0.936; 95% CI = 0.845-1.037; p = 0.208) revealed AR-related benefits for OS, while only the ER-negative subtype did not reveal AR-related benefits for DFS (HR = 1.154; 95% CI = 0.867-1.536; p = 0.327). However, only the high cutoff improved OS in ER-negative (HR = 1.480; 95% CI = 1.024-2.139; P = 0.037) and TNBC (HR = 0.546; 95% CI = 0.335-0.891; = 0.016) subtypes. Interestingly, there was no statistical evidence of publication bias. Conclusions: Our research demonstrates that AR expression is a reliable predictor of good prognosis for patients with BC, but it depends on the BC subtype. The differential predictive significance of AR expression in various BC subtypes has been discussed in more detail, and additional research is required to corroborate these results. It is clear that androgens and AR affect the biology of BC in a variety of ways. As a therapeutic target, AR should be extensively investigated in BC.