Evaluating MR enterography and intestinal ultrasound for ileal Crohn's disease activity: a multicentre agreement...

We aimed to establish the level of agreement between signs of Crohn's disease activity common to MR Enterography (MRE) and Intestinal Ultrasound (IUS).

Retrospective review of the METRIC multicentre prospective cohort trial (ISRCTN03982913) of newly diagnosed or suspected relapsed Crohn's disease undergoing contemporaneous MRE and IUS. Intermodality agreement for prespecified disease variables (10 MRE, 14 IUS) from the terminal ileum was assessed using Bland-Altman plots and percentage agreement against a multidisciplinary consensus panel reference standard. Imputation was used to account for disease not identified on either modality.

284 patients were included, median age 36 years (interquartile range, IQR: 17-68), 154/284 (54%) female. In the full cohort, intermodality 95% limits of agreement for terminal ileal mural thickness and disease length were -4.8 to 6.7 mm and -22.8 to 23.6 cm, respectively, and in the 158 patients with disease identified on both modalities, -4.0 to 6.3 mm and -27.5 to 29.1 cm, respectively. Increased mural T2 signal demonstrated the highest concordance with IUS submucosal layer thickening (91%, 141/155). Increased perimural T2 signal showed limited agreement with various IUS mesenteric observations (32-48%). There was 95% (147/155) agreement for global judgement of segmental disease severity and 74% (116/157) for obstruction.

While intermodality agreement for disease severity and obstruction is reasonable in patients with disease identified on both modalities, clinically important disagreement for mural thickness and disease length is frequent. Findings suggest caution in comparing individual parameters in isolation with potentially clinically significant intermodality disagreement.

Question MR Enterography and Intestinal Ultrasound are used interchangeably in Crohn's disease, but comparison between modalities is challenging. Findings We found reasonable intermodality agreement for global segmental disease activity but poor agreement for mural thickness, disease length, and most modality-specific activity parameters. Relevance statement This limited agreement is clinically significant with the potential to impact management decisions. Findings suggest caution in comparing individual disease parameters in isolation between MR Enterography and Intestinal Ultrasound. Instead, a holistic judgement of activity is more reliable.