Abstract 4840: TGFBR1 and TP53 SNPs interactions associated with colorectal cancer risk: Analysis of metabolic pathways using a Random Forest approach.

Abstract

Abstract Background: We recently found the association of african ancestry with colorrectal cancer risk in Colombian population. Currently due to sample size limitations we can not perform further analysis (ej.GWAS) to explore this finding. Pathway analysis has been proposed as a complementary approach to single SNP analyses in GWAS, specially in hypothesis driven analysis. AIM: To evaluate the association of SNP included in the SNP500 cancer with colorectal cancer risk in colombian population. Methods: We recruited 264 controls and 203 colorectal cancer cases from 5 Colombian cities.Participants gave written informed consent, donated blood samples and answered guided-questionnaires. SNP analysis was done using and high reliable comertial plataform the Illumina Cancer Panel This panel includes 1492 SNP related to different cancer metabolic pathways and were drawn from the SNP500 cancer database. After QC procedures we grouped these SNP according to the pathway maps reported in the KEGG database. Using a Random Forest Analysis we evaluate the discriminative capacity between cancer case and control of each SNP in each pathway separately. Age sex and african ancestry where considered covariates for this analysis. Those SNPs showing a better o similar performance to african ancestry were further analyzed in a conditional logistic regression model, where all known risk factors for colorectal cancer were included. Results: 1192 SNPs and 418 individuals pass through the QC protocol. SNP were grouped in 16 different pathways. Random Forest analysis showed that only 1 pathway, the hsa05210 (i.e colorrectal cancer) contained 2 SNPs, rs868 and rs2909430, that outperformed the discriminatory capacity of african ancestry between cases and controls. Conditional logistic regression analysis showed an interaction between both polymorphisms (p<0.001). Compared against carriers of the mayor alleles of these two SNPs, we found a protective effect against colorectal cancer risk among carriers of only one of the variants (OR 0.04 CI95% 0.01-0.25 and OR 0.26 CI95% 0.12-.053 for rs868 and rs2909430 respectively). In contrast we found a positive association with cancer risk among carriers of both variants (OR 12.5 CI95% 3.15-49-98). African ancestry possitive association remained significant in the multiple regression analysis (p<0.01). Rs868 and rs2909430 are located in TP53 and TGFBR1 genes. Other Polymorphisms in this genes has shown conflicting results in previous analysis on colorectal cancer risk. Conclusion: We report for the first time an epistatic effect between two polymorphism located in genes involved in the colorectal cancer metabolic pathway in an admixed population after controlling for population structure and known risk factors. These results may help to explain the diverging results observed in studies of the association of these genes with colorectal cancer risk Citation Format: Gustavo Hernandez-Suarez, MariaCarolina Sanabria, Martha Serrano, Jovanny Zabaleta, Albert Tenesa. TGFBR1 and TP53 SNPs interactions associated with colorectal cancer risk: Analysis of metabolic pathways using a Random Forest approach. [abstract]. In: Proceedings of the 104th Annual Meeting of the American Association for Cancer Research; 2013 Apr 6-10; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2013;73(8 Suppl):Abstract nr 4840. doi:10.1158/1538-7445.AM2013-4840 Note: This abstract was not presented at the AACR Annual Meeting 2013 because the presenter was unable to attend.