Background: Fibrostenosis is a common complication of Crohn's disease (CD) occurring in 30% of patients. Although the pathophysiology of intestinal fibrosis is incompletely understood, evidence suggests a genetic contribution. Previous genetic association and candidate gene studies were based on clinical definitions which lack both sensitivity, specificity and have a high inter-observer disagreement. Additionally, the recent genotype-phenotype analysis by the IIBDGC did not consider the time to development of fibrostenotic disease. As the genetic risk may be more important in patients with early fibrostenosis, we aimed to identify novel genetic markers by focussing on early fibrostenotic disease.
Methods: In this multicenter, retrospective nested case-control study computed tomography (CT) and magnetic resonance imaging (MRI) from CD patients obtained between 2002 and 2016 were examined for the presence of ileal fibrostenotic disease. Patients with early fibrostenosis, defined as a the presence of bowel tickening with luminal narrowing and prestenotic dilatation on CT/MRI occurring within 5 years following diagnosis of ileal or ileocolic disease (Montreal L1 or L3), and with available Illumina Immunochip data were included. The control cohort consisted of inflammatory CD patients (Montreal L1 or L3) without arguments for fibrostenotic disease after minimum 10 years follow up. Allelic association was assessed using PLINK v1.07.
Results: In total 3024 CT or MRI scans of 2042 CD patients were screened. 112 patients were selected because of fibrostenosis occurring within 5 years of diagnosis. Of these, Immunochip data were available in 60 cases and 49 (82%) had confirmed stenosis on histopathology. 343 inflammatory CD controls with genotype data were included in the analysis. Of the 156.500 SNPs analysed, only rs35223850 in the MIS18BP1 gene passed genome-wide significance level for association with early fibrostenosis. The protein encoding MIS18BP1 is known to bind to the SP1 transcription factor which has been associated with cardiac, liver and kidney fibrosis. Five additional SNPs reached a statistically suggestive significance level of p<5×10–6, including rs116630177 in the IL23R gene, which has previously been associated with systemic sclerosis.
Conclusions: This carefully phenotyped study reveals an important role for genetic contribution to early development of fibrostenotic complications in CD. Our data suggest a role for MIS18BP1 and the SP1 transcription factor as well as the IL23 pathway in the pathogenesis of early intestinal fibrosis.