Evaluation of Polygenic Risk Scores for Prediction of Prostate Cancer in Korean Men

Front Oncol. 2020 Oct 22;10:583625. doi: 10.3389/fonc.2020.583625. eCollection 2020.

Abstract

Aims: The purpose of this study is to evaluate an aggregate influence of prostate cancer (PCa) susceptibility variants on the development of PCa in Korean men by using the polygenic risk score (PRS) approach.

Methods: An analysis of 1,001 cases of PCa and 2,641 controls was performed to: (i) identify potential PCa-related risk loci in Koreans and (ii) validate the cumulative association between these loci and PCa using the PRS. Subgroup analyses based on risk stratification were conducted to better characterize the potential correlation to key PCa-related clinical outcomes (e.g., Gleason score, prostate-specific antigen levels). The results were replicated using 514 cases of PCa and 548 controls from an independent cohort.

Results: Genome-wide association analysis from our discovery cohort revealed 11 candidate single-nucleotide polymorphisms (SNPs) associated with PCa showing statistical significance of p < 5.0 × 10-5. Seven variants were located at 8q24.21 (rs1016343, rs16901979, and rs13252298 in PRNCR1; rs4242384, rs7837688, and rs1447295 in CASC8; and rs1512268 in NKX3). Two variants located within HNF1B (rs7501939 and rs4430796) had a significant negative association with PCa risk [odds ratio (OR) = 0.717 and 0.747, p = 6.42 × 10-7 and 3.67 × 10-6, respectively]. Of the six independent SNPs that remained after linkage disequilibrium (LD) pruning, the top four SNPs best predicted PCa risk with an area under the receiver operating characteristic curve (AUC) of 0.637 (95% CI: 0.582-0.692). Those with top 25% polygenic risk had a 4.2-fold increased risk of developing PCa compared with those with low risk.

Conclusion: Eleven PCa risk variants in Korean men were identified; PRSs of a subset of these variants could help predict PCa susceptibility.

Keywords: Genomics; Koreans; polygenic risk score; prediction; prostate cancer.