Treatment with high-dose proton pump inhibitors helps distinguish eosinophilic esophagitis from noneosinophilic esophagitis

Abstract

Background: Eosinophilic esophagitis (EE) is a clinical entity that is recognized increasingly in children. The treatment of EE has been debated since its identification as a clinical entity separate from reflux esophagitis. We hypothesize that the treatment with a high-dose proton pump inhibitor (HDPPI) helps differentiate EE from noneosinophilic esophagitis (NEE).

Patients and methods: Retrospective review of 2221 patients who underwent esophagogastroduodenoscopy (EGD) with biopsies was undertaken. Sixty-nine patients had more than or equal to 15 eosinophils/high-power field (eos/HPF) in 1 or more esophageal levels. Of those, 36 were initially treated with HDPPI for 3 months followed by repeat EGD. Patients who demonstrated histologic response were classified as NEE. Patients with no histologic response were diagnosed as having EE and treated with HDPPI+swallowed fluticasone for 3 months followed by repeat EGD.

Results: Of the 36 patients, histologic response was seen in 14 (39%) after treatment with HDPPI; 95% confidence interval (0.23-0.54). Swallowed fluticasone was added to the treatment of the 22 patients who did not show histologic response to HDPPI alone. Of those, 15 patients underwent repeat endoscopies. Seven patients were lost to follow-up or did not have repeated EGDs. Histologic response was observed in 9 of 15 (60%) patients. Of the nonresponders (6 of 15), 5 of 6 (83%) self-reported noncompliance with the swallowed fluticasone. Patients with more than or equal to 15 eos/HPF at all 3 levels (25 of 36) were less likely to respond to HDPPI alone and more likely to be categorized as EE (18 of 25), P=or<0.043. Symptomatically, 28 of 36 patients reported resolution of symptoms after HDPPI therapy alone, P=or<0.0001, regardless of histology. Visual endoscopic findings during the first and second EGDs did not show any significance in differentiating EE from NEE, P=0.625 and P=0.2405, respectively.

Conclusions: The study demonstrates that HDPPI can be used to help differentiate EE from NEE histologically. Moreover, patients with more than or equal to 15 eos/HPF at all 3 levels are less likely to respond to HDPPI than patients with more than or equal to 15 eos/HPF at fewer than 3 levels. Therefore, having more than or equal to 15 eos/HPF at 1 or 2 biopsy levels does not necessarily establish the diagnosis of EE. Symptomatic response to HDPPI does not correlate with histologic findings. Clinical management guided by EGD with biopsy helps distinguish patients with EE from those with NEE.