Abstract 4606: Fine mapping of 64 prostate cancer GWAS regions identifies multiple novel association signals

Abstract

Abstract Genome-wide association studies (GWAS) have identified 100 common prostate cancer (PrCa) susceptibility loci to date. We performed comprehensive fine-mapping of 64 GWAS regions using genotyping and imputation to a 1000 Genomes reference panel for 25,779 PrCa cases and 26,218 controls of European ancestry from the PRACTICAL Consortium and two UK GWAS studies. In order to identify independent variants associated to PrCa, SNPs significant at P ≤10-4 were included in a stepwise logistic regression (SLR). Where the initial SLR identified multiple independent SNPs in a region, we re-analysed the region conditioning on the top hit. The adjusted results were subsequently trimmed using a P-value cut-off of ≤10-5 and a second SLR performed to identify independently significant index SNPs. We observed a single independent signal at 39 regions, with a novel, more significantly associated index SNP at 35 of these. Amongst these, we confirmed association in the European population for 2 loci previously reported in Asian GWAS. At 16 regions there was evidence for multiple independent signals, 14 of these contain newly identified additional significant associations. Functional annotation using data from ENCODE filtered for PrCa cell lines showed enrichment for overlap with bio-features within the fine-mapped SNP set and eQTL analysis identified novel candidate genes regulated by SNPs discovered in this study. Furthermore, we observed a 7% (from 32%-to 39%) improvement in the estimated proportion of familial relative risk explained through these refined and newly identified genetic variants. This study demonstrates the utility of fine-mapping, in silico functional annotation and eQTL approaches to narrow down the number of candidate functional variants. In addition, since a greater proportion of GWAS loci contained multiple independent risk variants than previously appreciated; this may explain a proportion of the missing heritability of complex diseases. Citation Format: Zsofia Kote-Jarai, Ali Amin Al Olama, Tokhir Dadaev, Dennis Hazelett, Qiuyan Li, Daniel Leongamornlert, Ed Saunders, Matthew Feedman, David Conti, Douglas Easton, Gerhard Coetzee, Rosalind Eeles, The PRACTICAL Consortium. Fine mapping of 64 prostate cancer GWAS regions identifies multiple novel association signals. [abstract]. In: Proceedings of the 106th Annual Meeting of the American Association for Cancer Research; 2015 Apr 18-22; Philadelphia, PA. Philadelphia (PA): AACR; Cancer Res 2015;75(15 Suppl):Abstract nr 4606. doi:10.1158/1538-7445.AM2015-4606