Methods
Study Design and Patients
This was a prospective cohort study conducted at the University of North Carolina (UNC) between 2009 and 2011. Consecutive adult patients (aged 18–80 years) referred for routine outpatient esophagogastroduodenoscopy were recruited from the two UNC gastrointestinal (GI) procedure units. Patients were stratified by indication (dysphagia vs. other indications) and enrolled in an approximately 3:1 ratio of dysphagia vs. non-dysphagia indications, in order to enrich the study pool for patients with dysphagia. Subjects were excluded if they had a known (prevalent) diagnosis of EoE or a different eosinophilic gastrointestinal disorder, GI bleeding, active anticoagulation, known esophageal cancer, previous esophageal surgery, known esophageal varices, medical instability or multiple comorbidities precluding enrollment in the clinical opinion of the endoscopist, or inability to read or understand the consent form. Subjects provided informed consent and were enrolled before the endoscopy. This study was approved by the UNC Institutional Review Board.
Subjects with a new (incident) diagnosis of EoE met consensus guidelines. Specifically, cases were required to have at least one typical symptom of esophageal dysfunction (e.g., dysphagia, food impaction, heartburn, or feeding intolerance); at least 15 eosinophils per high-power field (eos/hpf) on esophageal biopsy persisting after an 8-week PPI trial (20–40 mg twice daily of any of the available agents, prescribed at the discretion of the clinician); and other causes of esophageal eosinophilia excluded. Although the majority of the EoE group included subjects who were PPI-naive on their index endoscopy (n=24), there were patients who were on high-dose PPI for at least 8 weeks at the time of their index endoscopy (n=17) who did not have pre-PPI endoscopy or histology data available.
Subjects with PPI-REE were required to have at least one typical symptom of esophageal dysfunction (e.g., dysphagia, food impaction, heartburn, or feeding intolerance); at least 15 eos/hpf on esophageal biopsy; improvement of esophageal eosinophilia to <15 eos/hpf after an 8-week PPI trial (20–40 mg twice daily of any of the available agents, prescribed at the discretion of the patient's clinician); and improvement of symptoms by self-report at the time of the repeat endoscopy. By definition, the PPI-REE group included only subjects who were PPI-naive at the time of their index endoscopy and required follow-up endoscopy after the PPI trial.
Clinical and Histological Data
Clinical data including demographics, symptoms, and the indications for endoscopy were recorded. At the time of enrollment, a blood sample was drawn for the peripheral eosinophil count (cells × 10/l) and total immunoglobulin E (IgE) level (kU/l). During endoscopy, all endoscopic findings were recorded using a standardized case report form. A total of five research protocol esophageal biopsies were obtained (two from the proximal, one from the mid, and two from the distal esophagus) to maximize the sensitivity of EoE diagnosis. Additional clinical biopsies were taken as needed at the discretion of the endoscopist. Esophageal biopsies were reviewed by the study pathologists to determine eosinophil counts according to our previously validated protocol. In brief, the slides were masked to the clinical case status and digitized. Using Aperio ImageScope (Aperio Technologies, Vista, CA), the maximum eosinophil density (eosinophils/mmeos/mm)) was determined after examination of five microscopic fields from each of the five biopsies. For purposes of comparison with previous studies, eosinophil density was then converted to eosinophil counts (eos/hpf) for an assumed hpf size of 0.24 mm, the size of an average field as reported in the literature. The eosinophil infiltration was further examined to determine whether it was patchy (localized eosinophilia ≥15 eos/hpf in only one hpf in the biopsy) or diffuse (eosinophilic inflammation seen in multiple hpfs) throughout the entire biopsy sample, as well as whether the eosinophil distribution throughout the mucosa was superficial only, basal only, or diffuse (throughout the epithelium). Notably, gastric and duodenal biopsies were also collected and examined to exclude co-existing eosinophilic gastritis or gastroenteritis.
Statistical Analysis
Characteristics of the study groups were summarized using descriptive statistics, and the proportion of EoE cases and PPI-REE subjects were calculated. Bivariate comparisons were made between patients with and without esophageal eosinophilia ≥15 eos/hpf and between EoE cases and PPI-REE subjects. Chi-square was used for categorical variables and t-tests were used for continuous variables. For the comparison between EoE and PPI-REE, baseline data from the PPI-naive visit were used. Thus, patients diagnosed with EoE on an index endoscopy while on high-dose PPI were excluded from these comparisons. Multivariable logistic regression was used to assess for factors that would independently distinguish PPI-REE and EoE. Based on the final sample size, different models were constructed with no more than four covariates.