Development and Validation of a Simplified Magnetic Resonance Index of Activity for Crohn's Disease.

Abstract

The magnetic resonance index of activity (MARIA) for Crohn's disease (CD) is used to assess the activity of luminal CD. However, it has a number of practical limitations. We aimed to develop and validate a simplified MARIA to more easily and quickly assess CD activity and response to therapy. We performed a retrospective analysis of magnetic resonance imaging data from 98 participants in 2 studies. We used logistic regression analysis to identify magnetic resonance imaging parameters independently associated with CD endoscopic index of severity (CDEIS) scores (the reference standard). We validated the responsiveness and reliability of the simplified MARIA in an independent cohort of 37 patients who underwent magnetic resonance imaging and endoscopy before and after a therapeutic intervention. Logistic regression analysis showed that dichotomous qualitative assessment of wall thickening (>3 mm), presence of mural edema, perienteric fat stranding, and ulcers were independently associated with CDEIS scores; we used these factors to create a simplified MARIA. Simplified MARIA scores greater than 1 identified segments with active CD with 90% sensitivity and 81% specificity (area under the curve 0.91; 95% confidence interval 0.88-0.94). Simplified MARIA scores of 2 or more detected severe lesions (ulcers) with 85% sensitivity and 92% specificity (area under the curve 0.94; 95% confidence interval 0.91-0.96). For each patient, there was a high level of correlation between simplified MARIA scores and CDEIS scores (r= 0.83) and simplified MARIA scores and original MARIA scores (and r= 0.93) (P < .001). The simplified MARIA score accurately detected changes in lesion severity in response to therapy and was as reliable as endoscopy for the assessment of mucosal healing. We developed and validated a simplified MARIA for easier and faster assessment of CD activity and severity. This index identifies patients with a response to therapy with a high level of accuracy. These findings require confirmation in independent, multireader studies.