230. ENDOSCOPIC REFERENCE SCORE REMODELING SUBSCORE AND ESOPHAGEAL EOSINOPHIL COUNT DO NOT CORRELATE IN PLACEBO-TREATED EOSINOPHILIC ESOPHAGITIS PATIENTS FROM DUPILUMAB STUDIES

Abstract

Abstract Although eosinophils are the hallmark of eosinophilic esophagitis (EoE) with threshold levels required for diagnosis, their exact role in EoE is unclear. The EoE-Endoscopic Reference Score (EREFS) assesses endoscopic inflammatory (edema/exudate/furrows) and remodeling (rings/strictures) features. We assessed the correlations between peak esophageal intraepithelial eosinophil count (eos/high-power field [hpf]) and EREFS in placebo-treated EoE patients from two clinical trials: phase 2 proof-of-concept (POC; NCT02379052; n = 24) and part A of phase 3 TREET (Phase3-PartA; NCT03633617; n = 39). POC was a 12-week trial in adults and Phase3-PartA was a 24-week trial in adults and adolescents. Endoscopic examination of the esophagus was performed at baseline (BL) and end of treatment (EOT) and scored for inflammatory and remodeling subscores and total score using EREFS (higher scores indicate greater severity). Peak eos/hpf was measured from pinch biopsies of proximal, mid, and distal esophagus at BL and EOT. Pearson correlations were performed between total/proximal/mid/distal eos/hpf and EREFS total, subscores, and components at BL, EOT, and for change from BL at EOT. Some strong correlations (r > 0.5) were observed between total peak eos/hpf and EREFS at BL in both studies and at EOT in Phase3-Part A (Table). Moderate-to-strong correlations (r > 0.3) were observed between total peak eos/hpf and EREFS inflammation subscore and edema/exudates/furrows individual components (some P > 0.5) in both studies. Correlations of total peak eos/hpf with EREFS remodeling subscore and rings/stricture components were not significant (P = 0.07–0.94). Peak eos/hpf in the proximal, mid, or distal esophagus were more strongly associated with EREFS inflammation subscore (majority moderate-to-strong correlations; P = 0.001–0.66) than EREFS remodeling subscore (weak correlations, P = 0.22–0.99) at BL/EOT for both studies. In these EoE patients, moderate-to-strong correlations were observed between eos/hpf and endoscopic markers of inflammation. No strong correlations were observed between total, proximal, mid, or distal peak eos/hpf and endoscopic parameters of remodeling. These data support a role for intact eosinophils in inflammatory changes in the esophagus in EoE and demonstrate the need for both histology and endoscopic evaluation. Further study, with larger sample sizes, is required to verify these findings in other contexts.