Association of genetic variations of selenoprotein genes, plasma selenium levels, and prostate cancer aggressiveness at diagnosis

Background: Genetic variations in some of the selenoprotein genes, alone or together with an individual's selenium status, may influence risk or progression of prostate cancer. We investigated the impact of genetic variants of selenoproteins on plasma selenium levels and cancer aggressiveness at diagnosis in men with localized prostate cancer (PCa).

Methods: The study cohort comprised 722 patients seen at Dana-Farber Cancer Institute who had localized/locally advanced PCa (i.e., stage T3 or less, N0, and M0) from 1994 to 2001. Fifty-five tagging single nucleotide polymorphisms (SNPs) from six selenoprotein genes (TXNRD1, TXNRD2, SEP15, GPX3, SELENBP1, and SEPP1) were analyzed. Logistic regression is used to examine associations of genotypes and plasma selenium levels with risk of aggressive disease, defined as D'Amico intermediate/high risk categories. Step down permutation was applied to adjust for multiple comparisons.

Results: Three hundred and forty-eight patients (48%) had aggressive disease at diagnosis. Two SNPs were associated with cancer aggressiveness at diagnosis (unadjusted P = 0.017 and 0.018, respectively). The odds ratio for aggressive disease in patients carrying TXNRD2 rs1005873-AG/GG genotypes or SELENBP1 rs10788804-AG/AA genotypes was 1.54 (95% CI = 1.08, 2.20) and 1.45 (95% CI = 1.07, 1.98), respectively, compared to TXNRD2 rs1005873-AA or SELENBP1 rs10788804-GG carriers. Four SNPs in TXNRD2 (rs1005873, rs13054371, rs3788310, and rs9606174) and the rs230820 in SEPP1 were associated with plasma selenium levels (unadjusted P < 0.05). Permutation adjusted P-values were not statistically significant for all these comparisons at the cut-off point of 0.05.

Conclusion: We identified polymorphisms in selenoproteins that may influence the plasma selenium levels and may be associated with the risk of presenting with aggressive PCa in men with localized or locally advanced PCa. These results should be validated in other independent datasets.