Poster Session II – Poster of Distinction II - A212 PREDICTING SUSTAINED REMISSION AND MAXIMAL DISEASE SEVERITY IN PEDIATRIC CROHN’S DISEASE USING MACHINE LEARNING

Abstract Background Pediatric Crohn’s disease (CD) is a chronic inflammatory condition affecting the gastrointestinal tract. It displays more heterogeneous disease trajectories and treatment responses than adult-onset cases, posing significant management challenges. While patients following more severe trajectories may benefit from early aggressive treatments, no reliable objective method exists to identify which children will follow a severe trajectory at diagnosis. This prognostic gap leaves risk stratification dependent on subjective clinical judgment, potentially delaying interventions for high-risk patients. Early identification of severe trajectories could transform treatment decisions and improve outcomes through timely aggressive therapy. Aims This study aims to predict one-year sustained remission and maximal disease severity using machine learning models trained on baseline clinical and microbiome data in a nation-wide cohort of Canadian children with CD. Methods Using baseline clinical and microbiome data from the Canadian Children IBD Network inception cohort, we developed machine learning models to predict two first-year outcomes: 1) sustained remission vs non-sustained remission, defined as maintaining a post-remission Weighted Pediatric Crohn’s Disease Activity Index (wPCDAI)