Eosinophilic peak counts in eosinophilic esophagitis : a retrospective study

Background : The histologic diagnosis of eosinophilic esophagitis (Eoe) is based on finding >15 eosinophils/high power field (HPF) on any level within the squamous epithelium of the esophagus. However, this criterion is based on a consensus statement, and controversy remains about the exact number of eosinophils/HPF needed to diagnose Eoe. We aimed to determine eosinophilic peak counts in esophageal, gastric, and duodenal biopsies from suspected Eoe patients, investigate the correlation between eosinophilic peak counts at different biopsy locations, and determine inter-observer and intra-observer reliability in reporting eosinophilic peak counts. Methods: We selected 103 suspected Eoe patients, who underwent an endoscopic procedure between June 1, 2010 and July 15, 2017. Eosinophilic peak counts in 1 HPF were obtained by a medical student and an experienced gastrointestinal pathologist. Results : Eosinophilic peak counts in suspected Eoe patients are highly variable (esophagus : IQR 66-178, median 110 ; stomach : IQR 2-10, median 3 ; duodenum : IQR 16-44, median 25). No significant correlation was found between eosinophilic peak counts at different biopsy locations. The inter-observer and intra-observer correlation for reporting eosinophilic peak counts was in the near- perfect range (? ranged from 0.93 to 0.99, P < 0.0001). Conclusions : Our data suggest that the accuracy of determining eosinophilic peak counts is not influenced by the pathologist's experience. Therefore, variability in reporting eosinophilic peak counts is unlikely to influence the diagnostic accuracy of Eoe. To further improve diagnostic accuracy, investigation of other histologic features observed in Eoe is needed. (Acta gastroenterol. belg., 2019, 82, 243-250).