VALIDATION OF THE SIMPLIFIED ENDOSCOPIC MUCOSAL ASSESSMENT OF CROHN’S DISEASE (SEMA-CD) A NOVEL ENDPOINT TO ASSESS ENDOSCOPIC IMPROVEMENT WITH REAL WORLD DATA

Abstract

Abstract BACKGROUND To demonstrate treatment efficacy in Crohn’s disease (CD), regulatory authorities require that trials include an endoscopic remission endpoint. However, standardized endoscopic assessment of mucosal improvement, such as the simple endoscopic score of CD (SES-CD), is not typically recorded by clinicians in practice or outside of clinical trials. A previous study demonstrated the Simplified Endoscopic Mucosal Assessment of CD (SEMA-CD) correlates strongly with SES-CD and that SEMA-CD is a score that can be easily used by clinicians in routine practice (Adler et al., 2021). The objective of this current study was to validate the SEMA-CD scored colonoscopy videos in additional populations and to examine the reliability of SEMA-CD over a range of mucosal disease severity in pediatric and adult participants and to evaluate sensitivity to change over time. METHODS We compared SEMA-CD and SES-CD scores in 110 patients with existing pre- and post-treatment colonoscopy videos from two ustekinumab clinical trials (UNISTAR, pediatric, n=36; SEAVUE, adult, n=74; Figure 1). Videos were separately scored with the SEMA-CD and SES-CD by different central readers, blinded to each other’s scores and to clinical history. Spearman’s rank correlation coefficient was used to evaluate correlation between scores, under different settings by study population (pediatric, adult), disease severity, and video quality. Inter- and intra-rater reliability were assessed with interclass correlation coefficient (ICC). Readers also rated ease of SEMA-CD scoring on a 7-point Likert scale. RESULTS SEMA-CD was highly correlated (rho [95% CI]) with the SES-CD, 0.89 (0.86, 0.92; Figure 2). Pre-to post-treatment changes in SEMA-CD scores compared with SES-CD scores were highly correlated 0.84 (0.77, 0.89). The positive relationship remained strong between scores when comparing total scores in pediatric 0.94 (0.90, 0.96) and adult 0.86 (0.80, 0.89) participants; across SES-CD disease severity categories (inactive, mild, moderate, severe; rho range 0.69-0.85); and in “optimal” 0.90 (0.86, 0.93) and “less than optimal” quality videos 0.88 (0.83, 0.91). Intra- and inter-reader reliability (ICC [95%CI]) were high, 0.93 (0.88, 0.96) and 0.89 (0.85, 0.91), respectively. SEMA-CD was rated as the same as, or easier than SES-CD 99.6% of the time. CONCLUSION This study demonstrated the SEMA-CD is feasible, reliable, reproducible, and sensitive to change in both adult and pediatric patients with CD. SEMA-CD should serve as an easy score for clinicians to record mucosal improvement and provide quality real world evidence for research using data from clinical registries.