Evidence of circadian disruption in patients with eosinophilic esophagitis

Abstract

RATIONALE: Eosinophilic esophagitis (EoE) is diagnosed and monitored by identification of esophageal eosinophilia via endoscopic biopsy. The endoscopic reference score (EREFS) is used to describe EoE by grading edema, rings, exudate, furrows, and strictures. This study aimed to predict esophageal eosinophilia with EREFS in pediatric patients. METHODS: This retrospective study stratified 137 endoscopies among 58 patients (ages 5 to 22) by the diagnosis of EoE, defined by a peak eosinophil count >15 eos/hpf. Component and composite EREFS were compared to histology using Fisher's exact and Chi-square tests. Generalized estimating equations were used to assess risk factors for EoE, estimate ROC curves, and determine sensitivity and specificity. RESULTS: On average, a 1-point increase in EREFS was associated with a 9.8-point increase in eosinophilic count (95% CI: 7.0, 12.6; p<0.001). The odds of having EoE were 31 times (95% CI: 6, 159) greater among patients with an EREFS > _4 compared with EREFS < _1. Utilizing exudate and furrows (reduced score) to discriminate among patients with EoE yielded an AUC of 0.88 for the ROC compared with 0.74 for EREFS (p50.15). A cut point of 1 for the reduced score had an 88% sensitivity and 86% specificity; while a cut point of 1 for EREFS yields a 91% sensitivity and 69% specificity. CONCLUSIONS: EREFS correlates with esophageal eosinophilia; however, a reduced score may be more accurate at identifying EoE. These findings suggest that esophageal appearance can screen for EoE, raising the possibility of using less invasive techniques, such as esophageal video capsule, for diagnosis or surveillance. Abstracts AB137 SUNDAY