Diffusion-Weighted MR Enterography to Monitor Bowel Inflammation after Medical Therapy in Crohn's Disease: A Prospective Longitudinal Study

OBJECTIVE To prospectively evaluate the performance of diffusion-weighted imaging (DWI) to monitor bowel inflammation after medical therapy for Crohn's disease (CD). MATERIALS AND METHODS Before and following 1-2 years of medical therapy, between October 2012 and May 2015, 18 randomly selected adult CD patients (male:female, 13:5; mean age ± SD, 25.8 ± 7.9 years at the time of enrollment) prospectively underwent MR enterography (MRE) including DWI (b = 900 s/mm2) and ileocolonoscopy. Thirty-seven prospectively defined index lesions (one contiguous endoscopy-confirmed inflamed area chosen from each inflamed anatomical bowel segment; 1-4 index lesions per patient; median, 2 lesions) were assessed on pre- and post-treatment MRE and endoscopy. Visual assessment of treatment responses on DWI in 4 categories including complete remission and reduced, unchanged or increased inflammation, and measurements of changes in apparent diffusion coefficient (ΔADC), i.e., pre-treatment-post-treatment, were performed by 2 independent readers. Endoscopic findings and CD MRI activity index (CDMI) obtained using conventional MRE served as reference standards. RESULTS ΔADC significantly differed between improved (i.e., complete remission and reduced inflammation) and unimproved (i.e., unchanged or increased inflammation) lesions: mean ± SD (× 10-3 mm2/s) of -0.65 ± 0.58 vs. 0.06 ± 0.15 for reader 1 (p = 0.022) and -0.68 ± 0.56 vs. 0.10 ± 0.26 for reader 2 (p = 0.025). DWI accuracy for diagnosing complete remission or improved inflammation ranged from 76% (28/37) to 84% (31/37). A significant negative correlation was noted between ΔADC and ΔCDMI for both readers with correlation coefficients of -0.438 and -0.461, respectively (p < 0.05). CONCLUSION DWI is potentially a feasible tool to monitor quantitatively and qualitatively bowel inflammation of CD after medical treatment.