Objective: To determine whether directly observed prostate-specific antigen (PSA) promoter diploid haplotype, either alone or in conjunction with androgen receptor (AR) genotype, is associated with prostate cancer risk.
Methods: We conducted a case-control study nested within the US population-based Cardiovascular Health Study cohort. Incident prostate cancers were identified by linkage to cancer registry records for the years 1989-2000. We genotyped 193 cases and 391 controls for the PSA -252 G/A and -158 G/A SNPs and the AR CAG microsatellite, and developed methods to directly determine proximal PSA promoter haplotypes. Exact logistic regression was used to estimate odds ratios and significance levels.
Results: No significant associations were observed between PSA diplotype and prostate cancer overall. Short (< 20) AR CAG repeat lengths were associated with modest increases in the risk of prostate cancer (OR, 1.46; 95% CI, 0.97-2.19; p = 0.071) that were significant for advanced disease (OR, 1.82; 95% CI, 1.02-3.26; p = 0.044). Men who possessed two copies of the PSA*2 (-252G/-158G) haplotype and short AR CAG repeat lengths had a 4-fold (95% CI, 1.05-20.75; exact p = 0.040) increased risk of prostate cancer, and a 7-fold (95% CI, 1.25-39.78; exact p = 0.026) increased risk of advanced disease.
Conclusions: We found evidence that the PSA*2*2 diplotype in combination with short AR CAG alleles increases a man's risk of developing prostate cancer. These findings support an etiologic role in prostate cancer of genetic interactions between polymorphisms that increase AR transactivation strength and those that alter the regulatory regions of target genes such as PSA that are responsive to androgen stimulation.