Outcomes and Predictors of Sustained Remission After Drug Withdrawal in Pediatric Crohn Disease

Objectives: The primary aim of this study was to determine the proportion of pediatric Crohn disease (CD) subjects in sustained drug‐free remission 52 weeks after stopping pharmacological therapy. We also aimed to explore the effects of the Crohn Disease Exclusion Diet (CDED) and microbiome composition on remission. Methods: We performed a prospective study following 18 CD patients ages 13–21 years in deep clinical remission withdrawing from immunomodulator (n = 7) or anti‐TNFα (n = 11) monotherapy at two tertiary care centers. Stool for calprotectin and microbiome analyses was collected over 52 weeks. Participants followed either the CDED or free diet after drug withdrawal. The primary endpoint was sustained relapse‐free drug‐free remission (calprotectin <250 µg/g) at 52 weeks. Results: Seventeen participants were followed through 52 weeks with 11 (64.7%) in sustained remission. There was no improvement in remission among participants following the CDED (5/9; 55.6%), P = 0.63. By 104 weeks, only 8 (47.1 %) participants remained off immunosuppressive therapies. Analysis of shotgun metagenomic sequence data revealed that taxonomic and gene function abundance in the gut microbiome was relatively stable for participants in remission and relapse. However, a predictive model incorporating gut microbial gene pathway abundance for amino sugar/nucleotide sugar metabolism and galactose metabolism from baseline samples predicted relapse at 52 weeks with 80% accuracy. Conclusions: After withdrawal of immunomodulator or anti‐TNFα monotherapy among a small cohort of pediatric CD subjects in deep remission, nearly 65% sustained remission at 52 weeks. Baseline microbiome alterations predicted relapse. Large prospective studies are needed to better understand outcomes after treatment de‐escalation.