Background & aims: The endoscopic reference score (EREFS) is used to determine severity of 5 endoscopic findings: edema, rings, exudates, furrows, and strictures. Little is known about the relationship between EREFSs and histologic markers of disease activity in children with eosinophilic esophagitis (EoE). We aimed to determine whether the EREFS can be used to identify children with EoE and how it changes with treatment.
Methods: We performed a prospective study of consecutive children (ages 2-17 years) undergoing diagnostic or post-treatment endoscopy scored real-time with EREFS from December 2012 through 2016. Findings from 192 diagnostic endoscopies and 229 post-treatment endoscopies were evaluated, from 371 children. Incident EoE cases were diagnosed based on 2011 consensus guidelines. Patients were treated with either elimination diet or topical steroids. Subjects who underwent endoscopy for symptoms of esophageal dysfunction but had normal esophageal findings from histology analysis were used as controls. EREFS and receiver operating characteristic curves were determined for incident EoE cases (n = 77) vs controls (n = 115), patients with active EoE (n = 101) vs inactive EoE after treatment (n = 128), and paired pre- and post-treatment cases of EoE (n = 85). Component and composite scores were correlated with eosinophilia.
Results: Visual detection of more than 1 esophageal abnormality during the diagnostic endoscopy identified children with EoE with 89.6% sensitivity and 87.9% specificity. EREFS correlated with peak level of eosinophilia (P < .001) at all esophageal levels. Children who responded to therapy had mean EREFSs of 0.5 compared to 2.4 in non-responders. In comparing pre-treatment vs post-treatment data from 85 patients, we found a significant reduction in the composite EREFS (from 2.4 to 0.7) (P < .001) among patients who responded to treatment; 92% of responders had a reduced EREFSs after treatment. EREFSs identified children with EoE with an area under the curve value (AUC) of 0.93. EREFSs identified children with active EoE following treatment with an AUC of 0.81 before treatment and an AUC of 0.79 after treatment.
Conclusions: In a prospective study of children undergoing diagnostic or post-treatment endoscopy, we found the EREFS to accurately identify those with EoE. Children who responded to therapy had lower EREFS scores than non-responders. EREFSs can be used to measure outcomes of pediatric patients, in conjunction with histology findings, and assess treatments for children with EoE.
Keywords: Biomarker; Esophagus; Inflammation; Metric.
Copyright © 2018 AGA Institute. Published by Elsevier Inc. All rights reserved.