P388 Eliakim score performance among patients with active Crohn’s disease

Abstract

Abstract Background Pan-enteric capsule endoscopy [PEC] (PillCam Crohn’s, Medtronic, USA) is a useful tool for disease assessment and monitoring in Crohn’s disease (CD). We have previously demonstrated the reliability and accuracy of Eliakim score (ES) in the assessment of CD-activity among patients with quiescent disease. We aimed to examine the performance of ES among patients who experienced clinical flare, as well as its responsiveness and sensitivity to change during follow-up (≤52weeks). Methods Patients with CD who have been started on biologics were included. They were prospectively followed with clinical visits, biomarkers and PEC at baseline, and after 14 and 52 weeks. At each time-point Crohn’s disease activity index (CDAI), C-reactive protein (CRP) and fecal-calprotectin (FC) levels were collected, and Lewis score (LS) and ES were calculated (independently reviewed by two experienced readers). Inter-class classification (ICC) was used to asses for agreement between readers. Baseline correlations were obtained using the Spearman's correlation. Repeated-measures correlation (RMC) was calculated using the rmcorr package in R (version 3.3-1). William's test was used to assess difference between correlations. Results Seventy-four patients were included (median age of 31.0 [22.5-46.5] years, male–50%). 142 PEC procedures were performed (Baseline– 62, week 14– 58, week 52– 22). Inter-rater agreement between both readers was high for both the LS and ES (ICC of 0.872 [p<0.001] and 0.925 [<0.001], respectively). At baseline, the correlations between FC and ES (r=0.509 [p<0.001]) and FC and LS (r=0.467 [p<0.001]) were comparable (p=0.68). RMC between biomarkers and ES were higher than between biomarkers and LS (CRP: r=0.376 [p=0.005] vs. r=0.204 [p=0.138], FC: r=0.549 [p<0.001] vs. r=0.412 [p=0.003], for ES and LS, [p=0.034/0.021], respectively), while no difference was observed (p=0.77) with regard to CDAI score (r=0.381 [p=0.007] vs. r=0.346 [p=0.015]). Performing sub-group analysis restricted to procedures which were confined to the small bowel (n=88), RMC was numerically higher (p=0.12) between FC and ES (r=0.590 [p=0.001]) than between FC and LS (r=0.470 [p=0.010]). Conclusion ES is a reliable and accurate scoring system in assessing mucosal inflammation in patients with active CD, and might have a higher sensitivity to clinical/inflammatory biomarker changes over time compared to LS.