Eosinophilic gastritis: histopathological characterization and quantification of the normal gastric eosinophil content

摘要

There is limited information about normal eosinophil counts in the gastric mucosa. The purpose of this study was to evaluate the histopathology of 60 patients whose biopsies showed increased eosinophils in the gastric mucosa. We also investigated the eosinophil content in gastric biopsies from normal controls (matched for age, sex, and zip code), from patients with Helicobacter pylori gastritis, and patients with Crohn's disease. Eosinophils were counted in five random high-power fields (HPFs) and reported in eosinophils/mm2. Involvement of the muscularis mucosae or submucosa, sheets of eosinophils, and infiltration of the gastric epithelium were also evaluated. The median eosinophil count in the study patients was 539 eosinophils/mm2; mean±SD=653±418 eosinophils/mm2; range 127–2108. Sheets of eosinophils were seen in 38 patients, 27 showed involvement of the muscularis mucosae or submucosa. There were 7 patients without epithelial infiltration by eosinophils, whereas 34 were tallied as rare and 19 were scored as abundant. No study patient had no evidence of H. pylori. The mean eosinophil count for the 135 normal controls was 15.5±16.8 SD eosinophils/mm2 (range 0–110); in the 93 controls with H. pylori gastritis the mean eosinophil count was 25±32.6 SD eosinophils/mm2 (range 0–219); and for the 53 controls with Crohn's disease it was 31.4±44.4 SD eosinophils/mm2 (range 0–203). There were no significant differences between the counts in biopsies from the antrum and corpus, and no significant variations by age, geographic location, or season. This study confirms that, in the United States population, the normal gastric eosinophilic counts are usually 2. We recommend ‘histological eosinophilic gastritis’ for the diagnosis of gastric biopsies that show an average density ≥127 eosinophils/mm2 (or ≥30 eosinophils per HPF) in at least five HPFs in the absence of known associated causes of eosinophilia.