Abstract 1661: Fine-mapping of GWAS variants in BMP4, GREM1, CDH1, and RHPN2 in African American colorectal cancer patients

Background: Genome-wide association studies (GWAS) of colorectal cancer (CRC) have identified risk variants in candidate genes involved in TGF-ß signaling including BMP4, GREM1, CDH1, RHPN2 and SMAD7. We previously replicated the association with a 3′-UTR SNP in GREM1 (rs10318) in African Americans but did not find evidence of associations in the other genes. This finding could be explained by different genomic structure in African Americans. Therefore, we sought to fine-map these genes to identify CRC risk variants in African Americans who are at highest risk of CRC in the US. Methods: DNA from cases and controls was obtained from two US institutions. In total, we included DNA from 795 AA cases and 985 AA controls. TagSNPs with minor allele frequencies >5% and pairwise r2 >0.8 were identified in five genes using YRI Hapmap data. A total of 81 SNPs were genotyped in cases and controls using Sequenom MassARRAY. We calculated odds ratios and 95% confidence intervals using logistic regression controlling for ancestry, age, gender and multiple testing in PLINK. Results: Association analysis was completed for four genes: BMP4, GREM1, CDH1 and RHPN2. The top tagSNP associations for each gene are shown in the Table. None of the SNPs were significant after permutation testing. In GREM1, the strongest SNP associations were found at the 3′ end of the gene. In BMP4, we identified two SNPs (rs1957852 and rs7146040) that are in LD with rs4444235 previously identified in a European GWAS. In CDH1, rs12931189 is located in the same LD block as the strongest association signal (rs7199991) in a recent fine-mapping study in Europeans. Finally, we have no evidence for associations in RHPN2 in African Americans. Conclusion: These results provide preliminary evidence of associations in African American colorectal cancer patients in BMP4, GREM1 and CDH1 identified by GWAS in Europeans. Ongoing work is testing associations in SMAD7 and further fine-mapping by imputation and resequencing in these regions. Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 103rd Annual Meeting of the American Association for Cancer Research; 2012 Mar 31-Apr 4; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2012;72(8 Suppl):Abstract nr 1661. doi:1538-7445.AM2012-1661