MOTILITY: Small Bowel Motility Quantified by Cine MRI As a Predictor of Longer Term Response in Patients with Crohn’s Disease Commencing Biological Therapy

Background Small bowel Crohn’s disease is increasingly treated with biological therapies. Although many patients do not achieve response or remission, this is unpredictable. Motility magnetic resonance imaging reliably measures Crohn’s disease inflammation, but its predictive ability for longer-term response to biologic therapies is unknown. Objectives Primary: to determine if motility magnetic resonance imaging at weeks 12–30 is superior to C-reactive protein for the prediction of response or remission at 1 year in patients commencing biologics for small bowel Crohn’s disease. Secondary: to determine if motility magnetic resonance imaging changes can predict patient quality of life at 1 year; to determine if diffusion-weighted magnetic resonance imaging is predictive of response or remission; to compare the predictive ability of motility magnetic resonance imaging with that of faecal calprotectin; to investigate the effect of body composition on response or remission; to determine the relationship between plasma gut peptide and cytokine levels, patient symptoms and bowel dysmotility. Design and methods Prospective multicentre cohort study. Setting and participants Thirteen National Health Service hospitals with established inflammatory bowel disease and magnetic resonance imaging practices. Participants with active non-stricturing small bowel Crohn’s disease requiring biological therapy. Interventions Motility magnetic resonance imaging (test under evaluation) and C-reactive protein (comparator). Faecal calprotectin as secondary comparator. Diffusion-weighted magnetic resonance imaging and plasma gut peptide/cytokine level measurement in a subset. Main outcome measures Comparative accuracy of motility magnetic resonance imaging, C-reactive protein and faecal calprotectin to predict response or remission at 1 year, defined using a combination of clinical factors and structural magnetic resonance enterography scores, and measures of quality of life at 1 year, using EuroQol-5 Dimensions, five-level version and disease-specific patient-reported outcome measures. Motility magnetic resonance imaging interobserver variability. Post hoc confirmatory analysis of motility magnetic resonance imaging correlation with magnetic resonance enterography-defined disease activity. Results Two hundred and nineteen participants were screened, of whom 199 were treated. Ninety-three participants were excluded prior to their first visit, with a further 20 lost to follow-up subsequently. Eighty-six participants completed all assessments. Stable or improved motility magnetic resonance imaging post induction was more sensitive than normalisation of C-reactive protein for 1-year response or remission (motility magnetic resonance imaging: 71.0%, 95% confidence interval 52.0 to 85.8; C-reactive protein: 45.2%, 95% confidence interval 27.3% to 64.0%; p = 0.008) but less specific (motility magnetic resonance imaging: 30.9%, 95% confidence interval 19.1 to 44.8; C-reactive protein: 67.3%, 95% confidence interval 53.3% to 79.3%; p < 0.001). There was no significant difference in area under the receiver operating characteristic curve (motility magnetic resonance imaging: 0.48; C-reactive protein: 0.53, p = 0.65). Similar results were obtained for faecal calprotectin. None of motility magnetic resonance imaging, C-reactive protein or faecal calprotectin predicted patient quality of life at 1 year. Diffusion-weighted magnetic resonance imaging was also not predictive of response or remission. Interobserver variability of motility magnetic resonance imaging quantified by intraclass correlation coefficients was 0.75–0.83. We confirmed prior work that motility magnetic resonance imaging is negatively correlated with small bowel Crohn’s disease activity (correlation coefficient [rho]= −0.34). Baseline magnetic resonance imaging-measured body composition was not associated with 1-year response or remission. Patient symptoms, gut peptide and cytokine levels and overall small bowel dysmotility were not consistently correlated. Limitations Loss to follow-up was high, mainly due to the coronavirus disease discovered in 2019 (COVID-19) pandemic. Response assessment using ileocolonoscopy was not possible. Conclusions Although improved motility magnetic resonance imaging is more sensitive than C-reactive protein and faecal calprotectin to predict response or remission at 1 year, it is less specific. None of these factors predict patient quality of life. Motility magnetic resonance imaging measurements are reliable between readers, and it remains a marker of disease activity at a given time point. Future work Impact of motility magnetic resonance imaging on therapeutic management decisions and clinician confidence in Crohn’s disease. Utility of motility magnetic resonance imaging in phenotyping stricturing Crohn’s disease. Funding This synopsis presents independent research funded by the National Institute for Health and Care Research (NIHR) Efficacy and Mechanism Evaluation programme as award number 14/201/16. Plain language summary Crohn’s disease is a condition that causes inflammation of the bowels, and this inflammation can eventually lead to irreversible bowel damage. Powerful treatments that suppress inflammation called biologic drugs are now available, but it is not possible to give these to all patients because of side effects, costs and patient inconvenience. Also, they do not always work, but we cannot predict this reliably. In this study, we tested if a novel scanning technique using magnetic resonance imaging that measures bowel motion (called motility magnetic resonance imaging) can help predict which patients are likely to benefit in the longer term from these biologic drugs. The study recruited 86 patients from 13 National Health Service centres, all of whom had motility magnetic resonance imaging before and after starting their biologic treatment. After 1 year, we checked to see if the biologics had worked for each patient. We tested how good motility magnetic resonance imaging was at predicting whether or not the biologics were going to work, and compared this with other simple blood and faeces tests. We also tested some other magnetic resonance imaging techniques called diffusion-weighted imaging and body composition measurements. We found that none of the tests (motility magnetic resonance imaging, blood tests or faeces tests) were better than chance at predicting if biologics were going to work or not. Diffusion-weighted imaging and body composition measurements were not useful either. We did confirm some previous work that motility magnetic resonance imaging is good at telling us if the bowel is inflamed or not, and we also found that many different doctors (not just experts) can measure it reliably. The study means that motility magnetic resonance imaging and diffusion-weighted imaging are not useful ways to predict which patients are likely to benefit from biologic drugs. It can still be used to measure if the bowel is inflamed or not.