OTU-022 Gwas of microscopic colitis in the UK biobank confirms association within the HLA region

Introduction The pathogenesis of microscopic colitis is poorly elucidated: it reportedly involves immune responses to luminal factors in genetically predisposed individuals. Previous genetic studies reported associations with HLA risk alleles in collagenous but not lymphocytic colitis. Non-steroidal anti-inflammatory drugs (NSAIDs), proton-pump inhibitors (PPIs) and selective serotonin reuptake inhibitors (SSRIs) are the most frequently cited environmental risk factors. We sought phenotypic and genetic associations with microscopic colitis in European patients enrolled in the UK Biobank. Methods We undertook a genome-wide association study of 483 cases of microscopic colitis defined by ICD code K52.8 (other specified noninfective gastroenteritis and colitis) and 4 50 616 controls. We tested 9527357 single nucleotide polymorphisms (SNPs) imputed using the haplotype reference consortium (HRC) reference panel. Association tests were performed using a linear mixed model (BOLT-LMM) including age, gender, study centre and chip as covariates. We also tested for associations with classical HLA alleles that were imputed using HLA*IMP 02. Results Participants with microscopic colitis were older (61.9 [56.2–65.4] vs. controls 58.6 [50.5–63.8], p=5e-15); more frequently female (65.6% [317/483] vs 54.2% [244531/450616], p=5e-07); more likely to smoke (14.7% [71/483] vs 10.4% [46792/450616], p=0.003) and were more often also diagnosed with coeliac disease (3.3% [16/483] and 0.4% [1991/450616], p=7e-10) than controls. In terms of drug factors, participants with microscopic colitis were more likely to have been exposed to proton-pump inhibitors (20.3% [98/483] cases vs 10.3% [46397/450156] controls, p=9e-11) than controls, but not aspirin/NSAIDs or SSRIs. We found a genome-wide significant association signal within the HLA region. The lead SNP was rs2596560 (OR 0.64, 95% CI: 0.57, 0.71, p=4e-9). Subsequent HLA imputation demonstrated that the signal was in linkage disequilibrium with the class I and II alleles that comprise the ancestral MHC 8.1 haplotype that has been linked with coeliac disease. Multivariable linear analyses showed that microscopic colitis (p=9e-08) and coeliac disease were independently associated with the rs2596560 SNP. There were no specific genetic associations seen in the subset of participants with microscopic colitis taking aspirin/NSAIDS, PPIs, or SSRIs. Conclusions We have confirmed a genome-wide significant association for microscopic colitis in the HLA region. Further studies are needed to understand the role of this locus in the pathogenesis of microscopic colitis and larger drug exposed cohorts will need to be identified to explore possible pharmacogenetic associations.