Assessment of small bowel mucosal healing by video capsule endoscopy for the prediction of short-term and long-term risk of Crohn's disease flare: a prospective cohort study

Background The optimal monitoring strategy for predicting disease course in Crohn's disease remains undefined. We aimed to evaluate the accuracy, safety, and tolerability of an intensive monitoring strategy designed to predict the future course of Crohn's disease in patients with quiescent disease. Methods In a prospective observational cohort study, we recruited patients older than 18 years with quiescent (for 3-24 months) Crohn's disease involving the small bowel with confirmed small bowel patency from three tertiary medical centres in Israel. Enrolled patients underwent baseline magnetic resonance enterography (MRE) and patency capsule, clinical or biomarker assessment every 3 months, and video capsule endoscopy (VCE) at baseline and every 6 months for 2 years or until a clinical flare (the primary outcome, defined as an increase in the Crohn's disease activity index score by 70 points or more) or disease worsening necessitating treatment intensification. We assessed the ability of the different Crohn's disease monitoring methods used to predict the occurrence of a flare during the 24-month follow-up period. Findings Of 90 screened patients, 29 were excluded (17 because of non-patent small bowel). Of the 61 patients enrolled between July 3, 2013, and Feb 1, 2015, 17 (28%) had a flare during the 24-month follow-up. No clinicodemographic parameter predicted future flare. A baseline VCE Lewis score of 350 or more identified patients with future flare (area under the curve [AUC] 0.79, 95% CI 0.66-0.88; p<0.0001; hazard ratio 10.7, 3.8-30.3). C-reactive protein at baseline had an AUC of 0.73 (0.6-0.84; p=0.0013) for predicting flare. The AUC of baseline faecal calprotectin for the prediction of flare occurring within 2 years was 0.62 (0.49-0.74; p=0.17), but progressively improved for shorter timespans and reached an AUC of 0.81 (0.76-0.85) for the prediction of flare occurring within 3 months. Of four MRE-based indices, only MRE global score correlated with 2-year flare risk (AUC 0.71, 0.58-0.82; p=0.024). During follow-up, a Lewis score increase of 383 points or more from baseline predicted imminent disease exacerbation within 6 months (AUC 0.79, 0.65-0.89; p=0.011). The safety and tolerability of the 231 VCEs ingested was excellent, with none being retained. Interpretation In patients with quiescent Crohn's disease involving the small bowel, faecal calprotectin predicts short-term flare risk, whereas VCE predicts both short-term and long-term risk of disease exacerbation. If corroborated by additional studies, protocols incorporating VCE could expand the scope of available methods for monitoring disease activity and predicting outcomes in small bowel Crohn's disease. Copyright (C) 2019 Elsevier Ltd. All rights reserved.