Objectives: Eosinophilic esophagitis (EoE) is characterized clinically by dysphagia, chest pain, and food impaction, and morphologically by increased numbers of intraepithelial eosinophils and marked basal hyperplasia of the squamous mucosa. The consensus criteria for a diagnosis of EoE include the presence of ≥15 eosinophils/HPF in biopsies from both proximal and distal esophagus in the absence of other causes of esophageal eosinophilia, and the lack of clinical response to proton pump inhibitor therapy. Because of the variability in the distribution of intraepithelial eosinophils among biopsy fragments and the lack of standardized biopsy practices, we sought to determine the optimal number of esophageal biopsies from the mid and distal esophagus needed to reach the minimum morphologic criteria of ≥15 eosinophils/HPF.
Methods: From 5 January 2009 to 26 September 2011, 771 patients were diagnosed with EoE at our institution. From that patient population, 102 sequential cases were chosen for further study, all of whom had biopsies taken from the mid and distal esophagus. Cases with only gastric mucosa present and biopsies taken from patients with a previous diagnosis of EoE were excluded. The original H&E-stained slides were reviewed, and the number of biopsy fragments containing squamous mucosa was recorded. By using a × 40 objective and × 10 oculars (field diameter=0.52 mm, field area=0.21 mm(2)), the number of eosinophils per high power field (EOS/HPF) in up to three HPFs was counted in each biopsy fragment.
Results: The EOS/HPF were counted in 1,342 biopsy fragments. The number of biopsy fragments obtained from the mid esophagus ranged from 1 to 20 (mean 7; median 7) and those obtained from the distal esophagus ranged from 1 to 18 (mean 6; median 5). There was no significant difference between the mean number of EOS/HPF from the mid (26) and lower (25) esophagus or between the mean peak number of EOS/HPF from the mid (69.1) and lower (60.4) esophagus. The probability of one, four, five, and six biopsy fragments containing >15 EOS/HPF was 0.63, 0.98, 0.99, and >0.99, respectively.
Conclusions: From these data, at least four biopsy fragments should be submitted from the mid and/or proximal esophagus to optimize the chances of a positive diagnosis of EoE in populations not known to have undergone previous proton pump inhibitor therapy. However, the yield is not increased beyond six biopsy fragments. In order to morphologically exclude a diagnosis of reflux esophagitis as the cause of intraepithelial eosinophilia, distal esophageal biopsies, if obtained, must be accompanied by more proximal biopsies (i.e., mid esophagus or higher).