Abstract
Our previous work suggested that there was no significant association between plasma steroid hormone levels and prostate cancer tumor grade at diagnosis. In this study, we systematically tested the hypothesis that inherited variations in the androgen and estrogen metabolic pathways may be associated with plasma levels of steroid hormones, or prostate cancer aggressiveness at diagnosis. Plasma hormone levels including total testosterone, total estradiol, and sex hormone-binding globulin were measured in a cohort of 508 patients identified with localized prostate cancer. D'Amico risk classification at diagnosis was also determined. A total of 143 single-nucleotide polymorphisms (SNPs) from 30 genes that are involved in androgen and estrogen metabolism were selected for analysis. The global association of genotypes with plasma hormone levels and prostate cancer aggressiveness (D'Amico risk classification) was statistically analyzed. Q values were estimated to account for multiple testing. We observed significant associations between plasma testosterone level and SNPs in HSD17B2 (rs1424151), HSD17B3 (rs9409407), and HSD17B1 (rs12602084), with P values of 0.002, 0.006, and 0.006, respectively. We also observed borderline significant associations between prostate aggressiveness at diagnosis and SNPs in AKR1C1 (rs11252845; P = 0.005), UGT2B15 (rs2045100; P = 0.007), and HSD17B12 (rs7932905; P = 0.008). No individual SNP was associated with both clinical variables. Genetic variants of genes in hormone metabolic pathways may influence plasma androgen levels or prostate cancer aggressiveness. However, it seems that the inherited variations affecting plasma hormone levels differ from those affecting disease aggressiveness.