Losartan Treatment Reduces Esophageal Eosinophilic Inflammation in a Subset of Eosinophilic Esophagitis.

ABSTRACT

BACKGROUND: Eosinophilic esophagitis (EoE) is a chronic, food antigen-driven esophageal disorder. Connective tissue disorders (CTDs) and esophageal connective tissue alterations are associated with EoE. Therefore, angiotensin II type 1 receptor blockade with losartan, an accepted CTD treatment, is a potential EoE treatment. OBJECTIVE: We evaluated losartan's effects on esophageal pathology, symptoms, and safety in patients with EoE with and without a CTD in an open-label, non-placebo controlled multisite study. METHODS: Fifteen participants with EoE, aged 5 to 23 years, underwent treatment with per-protocol titrated doses of losartan in an open-label, 16-week pilot trial. Losartan was added to standard of care therapy and 14 patients completed the study. Eosinophil counts served as the primary end point, whereas we also assessed the EoE Histology Scoring System, Endoscopic Reference Scores, EoE Diagnostic Panel, and patient-reported outcomes. RESULTS: Esophageal eosinophilia was not reduced after losartan. The peak eosinophil count was not reduced for the proximal (median [interquartile range] -3 [-22 to 3]; P = .49) and distal esophagus (median [interquartile range] -18 [-39 to -1]; P = .23). There were no differences in losartan response in EoE with or without CTD (n = 7 and 8, respectively). Regardless, in a small subset of four participants esophageal eosinophilia was resolved with a concomitant reduction in EoE Histology Scoring System score and Endoscopic Reference Score. Across all subjects, the Pediatric EoE Symptom Score, Pediatric Quality of Life Inventory EoE Module, and EoE Diagnostic Panel improved after losartan (P < .05). CONCLUSIONS: Losartan treatment was associated with improved patient-reported outcome scores and EoE Diagnostic Panel biomarkers although without a reduction in esophageal eosinophilia overall. A subset of patients demonstrated improved histopathologic and endoscopic features that could not be tied to a specific feature predicting response to treatment.