P-191 YI Lack of Association of Genetic Markers with Progression of Crohnʼs Disease

While progress has been made in the identification of Crohn's disease (CD) susceptibility loci, efforts to identify a genetic basis for disease progression have been less fruitful. The specific aim of this study was to build upon the major genetic advances made in IBD and to apply genome wide technologies toward predicting disease progression in CD. 1495 CD cases from 3 large IBD centers were reviewed by experienced phenotypers. Clinical and demographic details were collected, focusing on the time to first disease progression. Cases were defined as having either an indolent or progressive phenotype. Those who developed stricturing or penetrating disease within 5 years of diagnosis or who required abdominal surgery within 5 years or who required a biologic within 5 years of diagnosis were classified as having progressive disease. Genotyping was performed on 3 different platforms. Imputation was then conducted on the genotypic data using 1000 Genomes Phase 1 Integrated (March 2012) imputation panel by IMPUTE V2 after per chip QC. All 3 datasets were merged and minor allele frequencies <0.01 were then removed. Genome wide association (GWA) analysis was carried out on 3 clinical outcomes: (1) time to disease progression from inflammatory (B1) to stricturing (B2) or penetrating disease (B3), (2) time to first abdominal surgery, and (3) a binary analysis of indolent versus progressive disease. Cox-proportional hazards models and logistic regression models were used for the analyses. P < 10−8 was considered the threshold for genome wide significance. Time to disease progression. The median time to disease progression from B1 to B2/B3 behavior was 16.3 years (95% CI 13.8–18.9). Six hundred sixty-two cases were included, after QC and exclusion of any cases with B2/B3 behavior at baseline (n = 450), in the GWA analysis to determine any genetic variation associated with the time to disease progression. No SNP association reached genome wide significance in this survival analysis. Next we performed a targeted analysis to determine the association between time to disease progression and known IBD susceptibility SNPs. Six susceptibility SNPs were found to be associated with the time to disease progression, (P < 0.05). (Table1). Time to abdominal surgery Median time to abdominal surgery was 10.5 years (95% CI 9.5–11.4). 1360 cases were included after QC. Once again no SNP association reached genome wide significance; however, 8 known IBD susceptibility SNPs were found to be associated with the time to abdominal surgery, (P < 0.05). (Table1). Indolent versus Progressive disease. Seven hundred one cases were defined as having progressive disease and 562 indolent disease. Regression modeling was performed, analyzing the association between SNPs and the binary outcome. No SNP association reached genome wide significance. Four IBD susceptibility SNPs were associated with progressive disease and 1 with indolent disease, P < 0.05. (Table1). Our GWAS failed to show any SNP-phenotype association reaching genome wide significance. It is likely that multiple variables affect disease progression, with genetic factors potentially having only a small effect size. Very large studies would be required to identify them, however, such markers may not be clinically useful for predicting disease progression.