乳腺癌中雄激素受体的表达及其预后意义：一项系统综述和荟萃分析

Background: Molecular classification of breast cancer subtypes is based on the expression of estrogen, progesterone and HER2 receptors. The androgen receptor has emerged as a biomarker however, its prognostic role remains incompletely elucidated. Objective: to determine androgen receptor expression value as a biomarker and its potential impact on the clinical course of breast cancer. Methods: We conducted a systematic review and meta-analysis in accordance with the PRISMA statement, protocol was registered in PROSPERO (CRD420251166750). PubMed, Embase, and Web of Science were searched from inception to May 2025. We included analytical observational studies that enrolled histologically confirmed breast cancer, of any molecular subtype, and evaluated tumor androgen receptor expression by immunohistochemistry using a predefined positivity cut-off. Definitions of androgen receptor positivity were accepted as reported in each study, given the limited number of studies per outcome, which precluded formal subgroup analyses by cut-off. Studies were eligible if they reported hazard ratios (HRs) with 95% CIs for survival outcomes. Two reviewers independently assessed study eligibility, extracted data, and evaluated risk of bias using the Newcastle-Ottawa Scale; disagreements were resolved by consensus. Given the between-study heterogeneity, hazard ratios were pooled using an inverse-variance random-effects model, and heterogeneity was quantified with the I² statistic. Results: From a total of 966 records identified, 118 full texts were assessed after removing 228 duplicates, including 15 studies for qualitative synthesis. Seven of these reported hazard ratios for the predefined outcomes and were included in the meta-analysis. For overall survival in patients with triple-negative breast cancer and androgen receptor expression, the pooled HR was 0.60 (95% CI 0.29-1.21; I² = 85.8%, p = 0.0009), showing a non-significant trend toward better survival in AR-positive tumors. For breast cancer-specific survival in AR-positive breast cancer, the pooled HR was 0.55 (95% CI 0.20-1.49; I² = 88.2%, p for heterogeneity < 0.0001), also indicating a non-significant trend toward reduced breast cancer mortality. Conclusions: Although our meta-analysis did not demonstrate a statistically significant association between androgen receptor expression and overall or breast cancer-specific survival, these findings are relevant because they highlight substantial heterogeneity across studies. Differences in how androgen receptor status was assessed, along with variability in the definition and reporting of prognostic outcomes, likely contribute to the inconsistent estimates. Developing standardized criteria for positivity and harmonized outcome reporting will be crucial to clarify the true prognostic role of androgen receptor in breast cancer.