Abstract
Background
The heterogeneity of prostate cancer (PCa) poses a significant challenge to accurately predict those who will progress to lethal stage. While RNA signatures and somatic DNA alterations can be used to predict clinical outcomes in men with localized PCa, incorporating germline markers may further enhance early prediction of progression to incurable disease.Patients and methods
We examined germline markers in steroidogenic pathways in the multi-institutional prospective PROCURE cohort of 1926 patients with localized PCa. The association between 22 single nucleotide polymorphisms (SNPs) across 10 steroidogenic genes and 15 hormone levels, biochemical recurrence-free survival (BCRFS), development of metastasis, castration-resistance (CRPC) and PCa specific mortality (PCSM) was assessed along with their potential functional implications.Results
Thirteen markers were nominally associated with circulating hormone levels at prostatectomy. At a median follow-up (FU) of 9.8 years, variants in CYP17A1, SRD5A2, CYP19A1 and CYP1B1 showed nominal associations with BCRFS, while those located in SRD5A2, HSD17B2 and CYP1B1 were associated with lethal disease. Cumulative associations of these markers predicted time to metastasis/CRPC and fatal outcomes more accurately than individual SNP. Mechanistically, promoter variants in CYP17A1, SRD5A1, and SRD5A2 altered transcriptional activity in cell models, while a CYP1B1 variant was linked to reduced protein expression in primary tumors, likely due to enhanced proteasomal degradation.Conclusions
Our findings support an association between host genetic variation and circulating hormone levels at surgery, with potential clinical implications for disease progression. These markers represent a step toward the development of non-invasive strategies to refine inherited risk stratification for progression from localized to advanced prostate cancer.