Background: Two prostate cancer genome-wide scans in populations of European ancestry identified several genetic variants that are strongly associated with prostate cancer risk. The effect of these risk variants and their cumulative effect in other populations are unknown.
Methods: We evaluated the association of 23 risk single-nucleotide polymorphisms (SNPs) with prostate cancer risk and clinical covariates (Gleason score, tumor aggressiveness, and age at diagnosis) in men of Japanese ancestry (311 case subjects and 1035 control subjects) using unconditional logistic regression. We also used logistic regression to test the association between increasing numbers of independently associated risk alleles and the risk of prostate cancer, prostate cancer aggressiveness, and age at diagnosis. All statistical tests were two-sided.
Results: Seven of the 23 SNPs (five independent loci) were associated with prostate cancer risk (P values ranged from .0084 to 2.3 x 10(-8) and effect sizes [estimated as odds ratios, ORs] ranged from 1.35 to 1.82). None of the seven SNPs was associated with Gleason score or aggressive disease. rs6983561 and rs4430796 were associated with age at diagnosis (Ps = .0188 and .0339, respectively). Men with six or more risk alleles (27% of case patients and 11% of control subjects) had a higher risk of prostate cancer than men with two or fewer risk alleles (7% of case patients and 20% of control subjects) (OR = 6.22, P = 1.5 x 10(-12)).
Conclusions: These results highlight the critical importance of considering ancestry in understanding how risk alleles influence disease and suggest that risk estimates and variants differ across populations. It is important to perform studies in multiple ancestral populations so that the composite genetic architecture of prostate cancer can be rigorously addressed.