Esophageal Eosinophilia and Gastric Mucosal Pathology: Is There a Link?

Abstract

Purpose: We have previously reported an inverse relationship between eosinophilic esophagitis (EoE) and H. pylori infection. The present study tests the hypothesis that other gastric conditions likely related to H. pylori infection might also be less common in EoE patients, and that conditions unrelated to H. pylori infection would be independent of esophageal eosinophilia. Methods: We used the Miraca Life Sciences database to extract histopathologic, demographic, and clinical information from all patients with esophageal and gastric biopsies obtained between 1.2008 and 6.2012; those with upper GI cancer or esophageal surgery were excluded. Patients were stratified in 5 groups based on the numbers of eosinophils per high-power field (eos/HPF) in their esophageal squamous mucosa (<15, the controls; 15 to 45; 46 to 60; 61 to 75; and > 75). The prevalence of common gastric histopathologic diagnoses was determined according to the updated Sydney System: normal mucosa, H. pylori gastritis, chronic inactive gastritis (CIG), reactive gastropathy (RG), intestinal metaplasia (IM), and atrophic gastritis (AIG). Results: Of the 212,982 unique patients with both esophageal and gastric biopsies, 204,525 had <15 eos/HPF in the esophageal epithelium (median age 57 years [yrs]; 68.5% F), and 8,457 had ≥15 eos/HPF (median age 45 y; 37.8% F). A normal gastric mucosa was significantly more common in patients with EoE than in patients with <15 eos/HPF (23% vs. 18%; odds ratio [OR] 1.35; 95% CI 1.28 - 1.42; p<.0001; see Figure 1). H. pylori gastritis and, as expected, its associated conditions (CIG and IM), were inversely related to EoE. Reactive gastropathy, damage likely caused by NSAIDs or bile reflux, had an inverse relationship with EoE (12% vs. 16%; OR 0.70; 95% CI 0.66 - 0.75; p<.0001). AIG, an autoimmune condition whose etiologic relationship with H. pylori is uncertain, had the strongest inverse association with EoE (0.4% versus 0.1%; OR 0.36; 95% CI 0.20 - 0.65; p<.001).FigureConclusion: This study confirms the inverse relationship between EoE and H. pylori and its sequelae. Conditions that are clearly unrelated (RG) or uncertainly linked to H. pylori (AIG) were also less likely in patients with EoE. While selection bias and age may play a role in these findings, the possibility that the reduced acid reflux (due to hypo- or achlorhydria) in patients with AIG may limit eosinophilic infiltration in the esophageal mucosa may warrant further investigation.