Abstract
BACKGROUND: Single nucleotide polymorphisms (SNPs) located in the genes participating in the steroid hormone metabolism pathway may influence the outcomes of androgen deprivation therapy (ADT) in prostate cancer (PCa) patients, but findings on the Chinese population remain limited. METHODS: A multicentric cohort of 301 Chinese PCa patients receiving first-line ADT was enrolled. Germline SNPs located in 62 steroid hormone metabolism-related genes were analyzed for associations with time to ADT failure using multivariate Cox regression. Important expression quantitative trait loci (eQTLs) were discovered. RESULTS: Four SNPs were significantly associated with time to ADT failure: rs36119043 in AKR1D1 (hazard ratio, HR = 2.02, 95% confidence interval, 95% CI: 1.44-2.85, p = 5.72 × 10(-5)), rs151155810 in HSD17B12 (HR = 7.87, 95% CI: 2.78-22.30, p = 1.05 × 10(-4)), rs71179009 in SULT2B1 (HR = 2.16, 95% CI: 1.44-3.22, p = 1.85 × 10(-4)), rs28609134 in SRD5A3 (HR = 2.50, 95% CI: 1.51-4.15, p = 3.79 × 10(-4)). Potential causal eQTLs in the LD regions of these SNPs were identified, with significant impacts on AKR1D1, SULT2B1, and SRD5A3 expression in diverse tissues. A cumulative risk allele effect was observed: HR = 2.74 (95% CI: 1.86-4.03) under the dominant model and HR = 2.04 (95% CI: 1.63-2.55) under the additive model, with a median survival of 176 months (95% CI: N/A) in noncarrier patients vs. 92 months (95% CI: 65-N/A) in one risk locus-carriers and 55 months (95% CI: 26-N/A) in two risk loci-carriers. CONCLUSIONS: SNPs in the steroid hormone metabolism pathway can predict time to ADT failure in Chinese PCa patients, supporting their potential role for drug response and pharmacogenomic stratification.