A common polymorphism (1245A>C) in the HSD3B1 gene is associated with increased de novo synthesis of androgens and worse outcomes in men treated with androgen-deprivation therapy for metastatic castration-sensitive prostate cancer. The objective of the study was to determine whether this polymorphism is associated with outcomes for metastatic castration-resistant prostate cancer (mCRPC) treated with abiraterone or enzalutamide.

A total of 547 patients treated with abiraterone or enzalutamide from two prospective cohorts were evaluated. The HSD3B1 genotype was determined by targeted sequencing and/or TaqMan single-nucleotide polymorphism genotyping. In cohort 1, patients were randomized to receive abiraterone + prednisone or enzalutamide. In cohort 2, patients received either agent according to investigator's choice. Prostate-specific antigen (PSA) response rate, time to PSA progression (TTPP), time to progression (TTP) and overall survival were determined. Associations between HSD3B1 genotypes and outcomes were evaluated via univariate Cox regression. Multivariable Cox model was used to determine the independent association of each covariate.

The HSD3B1 variant genotype (CC) was present in 15% of patients and was associated with worse TTP [hazard ratio (HR) 1.31, 95% confidence interval (CI) 1.02–1.67, P = 0.032] and PSA response rates (48% for CC versus 62% and 65% for AA and AC, respectively [P = 0.019]), with no significant difference in TTPP (HR 1.28, 95% CI 0.99–1.66, P = 0.064). The effect of genotype was similar for treatment with abiraterone or enzalutamide with a negative test for interaction for TTPP (P = 0.997) and TTP (P = 0.749). Multivariable analysis did not show a significant association between genotype and TTP or TTPP.

The HSD3B1 (CC) genotype was associated with shorter TTP and lower PSA response rate in patients with mCRPC treated with abiraterone or enzalutamide. However, the CC genotype did not provide prognostic information beyond that conferred by standard clinical variables, suggesting that it may not be a suitable stand-alone biomarker in mCRPC.