P145 Phenotypic progression of Crohn’s disease : assessment of risk factors in a Tunisian population

Abstract Background Phenotype progression in patients with Crohn’s disease (CD) is common, but its risk factors vary between studies. The knowledge of these factors at diagnosis help to identify patients with high risk to progression. These patients should be given aggressive treatments given early in the disease course. The aim of our study was to evaluate the long-term behavioral course of Crohn's disease and to determine predictive factors for phenotypic progression. Methods We conducted a retrospective, descriptive study including patients with Crohn's disease that were managed at our department, over a 9-year period [January 2011 - January 2020]. Patients with less than 2 years of follow-up were excluded. We screened for factors associated with phenotypic progression during follow-up by a univariate study. By binary logistic regression, we sought for factors independently related to phenotypic progression, including variables that had a p <0.2 in univariate study. Results We included 341 patients with a sex ratio of 1.05. The median age at diagnosis of CD was 31 years. Smoking was observed in 36.4% of patients and 16.5% were alcohol users. Forty-four patients (12.9%) had a family history of Inflammatory Bowel Disease and 5.3% had associated autoimmune diseases. The mean BMI of the patients was 20.57 ± 4.5 kg/m². Seventy-six patients were at least overweighted at the time of diagnosis. The disease was classified as ileal, colonic and ileocolic in 150, 86 and 99 patients respectively. The involvement of the upper gastrointestinal (GI) tract was noted in 49 patients (14.36%) and 84 patients (24.6%) had perineal lesions mainly complex fistulas (17.3%). CD was initially classified as inflammatory, stricturing and fistulizing in 71%, 17% and 11% respectively. Phenotypic progression during follow-up was observed in 33 patients (9.6%), finally classifying the disease as inflammatory in 41%, stricturing in 22% and fistulizing in 35% of cases. In univariate study, the factors associated with phenotypic progression were represented by isolated ileal location (p=0.028, OR=2.28), the presence of perineal fistulas (p=0.006, OR=3.07) and hypoalbuminemia at diagnosis of less than 27.94g/L (p=0.035). Extensive ileal involvement and the presence of granuloma on histology were not associated with phenotypic progression. In multivariate study, only initial upper GI tract’s involvement was an independent predictor of phenotypic progression (p=0.024. OR=0.11, IC[0.18-0.75]). Conclusion Our study showed that upper GI tract’s involvement was the main predictor of phenotypic progression in CD.