Predictors of Non-Response to Topical Steroids Treatment in Eosinophilic Esophagitis

Abstract

Purpose: Topical steroids are used as first-line treatment agents for eosinophilic esophagitis (EoE). Randomized controlled studies have demonstrated only modest efficacy of these agents for inducing histologic remission. The aim of this study was to examine whether there were any characteristics that could predict non-response to treatment with topical steroids. Methods: Using data from two prospective studies and an EoE registry, children (< 18 years) and adults with EoE, as defined by consensus guidelines, were included in the analysis. All patients were treated with an 8-week course of either swallowed fluticasone or viscous budesonide. Responders were defined as achieving <5 eos/hpf in both the mid-proximal and distal esophageal biopsies and non-responders as having > 5 eos/hpf in proximal and/or distal esophagus. Demographic, clinical, endoscopic, and histologic features were examined. Results: The study cohort included 75 EoE patients: median age was 33 years (range 2-64 years), 71% adults, 84% male, and 76% Caucasian. Children had a higher response rate to topical steroids compared to adults (60% vs. 33%; p=0.047). Response rate was similar between males and females (33% vs. 58%, p=0.101) and between the two types of steroids (38% fluticasone vs. 36% budesonide, p=0.823). With regards to clinical symptoms, responders were more likely to present with food impaction vs. non-responders (43% vs. 21%, p=0.047); however. dysphagia (86% vs. 85%; p=0.943) and heartburn (29% vs 21%, p=0.474) were similar between responders and non-responders. With regards to pretreatment endoscopy, responders were less likely to have furrows compared to nonresponders (64% vs 87%, p=0.019).Other endoscopic features were similar between responders and non-responders, rings (54% vs 62%, p=0.489) and white plaques (21% vs. 36%, p=0.181). Nonresponders were more likely to undergo dilation compared to responders (43% vs 18%, p=0.028). Peak proximal (49±50 vs 55±48, p=0.606) and distal eosinophil (71±81 vs 82±86, p=0.578) counts were similar between responders and non-responders. In multivariate logistic regression, the independent predictors of non-response to topical steroids were adult age (OR 5.13 p=0.048), without food impaction (OR 0.116 p=0.005), furrows (OR 8.24 p=0.006) and dilation (OR 6.30 p=0.023). Conclusion: Non-responders to topic steroids are more likely to be adults, with longitudinal furrows on endoscopy, who undergo dilation and are less likely to present with food impactions. Consideration should be given to either increasing dosage of steroids or using an alternate form of treatment, such as dietary elimination, in these patients.