Establishing a Central Reader Pool for Magnetic Resonance Enterography Scoring in Crohn's Disease Clinical Trials

Background Magnetic resonance (MR) enterography is the gold‐standard cross‐sectional imaging test for assessing Crohn's disease activity, and there is increasing interest in its use for treatment response evaluation in clinical trials. Several validated scores have been developed for reporting in a standardised way, including the simplified Magnetic Resonance Index of Activity (MaRIAs) system and the Lemann Index. However, to date, no studies have assessed the reliability of these scores in a central group of readers. We present a protocol for establishing a validated central reader pool for MR enterography in Crohn's disease and describe its application to a multi‐centre clinical trial cohort. Methods Five consultant radiologists independently scored scans from the PROFILE trial using MaRIAs, MaRIAs‐E, and Lemann Index systems. The central reader pool development protocol encompassed independent scoring ( n = 8 scans), inter‐reader reliability assessment using intraclass correlation coefficient (ICC), radiology consensus meeting, and iterative rounds of scoring ( n = 18) and evaluation until satisfactory reliability was achieved. A further set of scans ( n = 53) were double‐read during the trial and reliability was assessed for validation. Results The central reader pool demonstrated good inter‐reader reliability following two rounds of scoring, which was maintained in the validation set, with ICC reaching 0.6 across all scoring systems. Conclusions This study establishes a robust protocol for the development and validation of a central reader pool for MR enterography scoring within a clinical trial context. This protocol is adaptable to other imaging modalities requiring expert interpretation against standardised scoring systems, ensuring consistent and reliable image assessment.