Tu1128 Meta-Analysis and Systematic Review: Topical Steroids vs Placebo and/or PPIs for the Treatment of Eosinophilic Esophagitis (EoE)

Abstract

Introduction: Current ACG guidelines define EoE as patients experiencing symptoms of esophageal dysfunction, > 15 eos/hpf on esophageal biopsy, and those that do not respond to a 2 month trial of proton pump inhibitor, after ruling out esophageal eosinophilia. Controversy surrounds the clinical and histological response to steroids, especially in adult patients with EoE and whether steroids or PPIs are better for initial treatment. Methods: Primary outcomes included complete and partial response (clinical and histological) to topical steroids vs placebo and PPIs vs steroids. Subgroup analyses included: adults, children, budesonide and fluticasone alone, as well as adverse events. We searched MEDLINE and Cochrane Central Register of Controlled Trials (CENTRAL) (The Cochrane Library) from inception until 11/15/2014. For quality assurance purposes throughout two investigators were involved. Data were pooled under a random effects model using intention to treat protocol. The systematic review was performed as per the standards of Cochrane collaboration. Results: Six studies on topical steroids vs placebo were pooled that included 240 EoE patients, and 2 studies pooling 72 patients assessing PPIs compared to topical steroids. There was statistically significant benefit of using topical steroids vs placebo, for the outcomes of complete and partial histological response, as well as complete clinical response. Whereas histological response held an 10-fold advantage over placebo, there was a much smaller clinical response with only a 2-fold advantage over placebo RR 10.4(4.4,24.5;p=<0.00001) and RR 2.1(1.3,3.5;p=0.003), respectively. On subgroup analysis, adults and pediatric patients had similar statistically significant benefits of topical steroids over placebo for complete histological remission RR 7.30(2.4,21.8);p=0.0004, and RR 12.7(3.7, 43.1); p<0.0001. Similarly, budesonide and fluticasone offer a statistically significant advantages over placebo for both complete and partial histological response. Candidiasis appeared to occur more frequently in the placebo group, but did not reach statistical significance RR 0.93(0.84,1.03;p=0.14). Although both partial clinical remission and complete histological response favored PPIs over topical steroid, this did not reach significance. Risk of biases in these studies was minimal. Conclusions: Topical corticosteroids appear to be a safe and effective treatment compared to placebo in both adults and children for complete clinical and histological response. However, the histological response was much more powerful(10vs 2-fold) when compared to the clinical response. There does not appear to be an advantage of either PPIs or steroid for achieving clinical or histological remission. Further studies need to investigate the explanation for the relatively poor clinical response across all groups in the steroid arm.