AbstractBackgroundEndoscopy is essential for measuring luminal disease activity in Crohn’s disease (CD). Current endoscopic indices for evaluating CD activity are limited: they are only modestly correlated with disease activity, contain items that are ambiguously defined and have poor inter-observer reliability, and lack validated definitions for clinically relevant change. We aimed to develop and internally and externally validate a reliable and responsive endoscopic CD activity index.MethodsIn this multi-part study, candidate endoscopic items for index development with face validity derived from a prior RAND/UCLA appropriateness exercise were assessed by blinded central readers using ileocolonoscopy videos from the phase 3b adalimumab EXTEND (NCT00348283) and phase 2 risankizumab (NCT02031276) trials. Acceptable item-level operating properties for further consideration as covariables in regression for model building included at least moderate inter-rater reliability (intraclass correlation coefficient [ICC]≥0.41), and medium responsiveness defined as the probability of a patient receiving active treatment having a better endoscopy score than a patient receiving placebo or win probability (WinP) ≥0.56. Model development and internal validation with bootstrapping utilised data from 112 paired baseline and post-treatment videos from the adalimumab trial. External model validation utilised data from 108 video pairs from the risankizumab trial.ResultsThe Endoscopic Activity Score for Evaluating CD (EASE-CD) is comprised of ulcer size, presence of deep ulcer(s), and percentage of ulcerated surface with each item evaluated and averaged across visualised bowel segments (Table 1). Total scores range from 0-100 (higher scores indicate greater endoscopic activity). EASE-CD demonstrated almost perfect inter-rater reliability (ICC 0.820 [95% CI: 0.662-0.906]) and a large degree of responsiveness (WinP 0.769 [95% CI: 0.658-0.851]). A 10-point reduction in EASE-CD had 87.8% specificity and 60.9% sensitivity for endoscopic response defined by the Simple Endoscopic Score for CD (SES-CD)/CD Endoscopic Index of Severity (CDEIS). Ulcer-free endoscopic remission results in EASE-CD=0.ConclusionEASE-CD is a novel, validated, continuous measure of endoscopic CD activity constructed using modern principles of index development, that is reliable, responsive to treatment, and linearly correlated with disease activity. EASE-CD does not include “narrowing” (which cannot be reliably evaluated between observers) or “affected area” (which is ambiguous in meaning, poorly reliable, and not independently associated with disease activity). EASE-CD is based only on ulcer-related items, the hallmark lesion in CD.