Decline in Esophageal Eosinophil Counts can be Substantial in Many Patients With Eosinophilic Esophagitis After Treatment With Proton Pump Inhibitors: 361

Introduction: Patients with esophageal eosinophilia are currently categorized into two distinct entities, namely eosinophilic esophagitis (EoE) or proton pump inhibitor (PPI) responsive esophageal eosinophilia (PPI-REE). This categorization depends on whether the eosinophil count is more or less than 15 eosinophils per high power field (15/HPF) after an 8-week trial of PPI. However, we have observed that some patients classified as having EoE can also have a decline in eosinophil count after a PPI trial, but to a level above the arbitrary cut-off of 15 eos/HPF. Methods: We analyzed prospectively collected data of a cohort of adult patients with esophageal eosinophilia (who attended the Department of Gastroenterology, Chapel Hill North Carolina). Results of histologic examinations were interrogated to determine treatment response to an 8-week trial of PPI (20-40 mg twice daily of any of the available agents). Pre-and post PPI peak eosinophil count (highest single reading in the upper, mid, or lower oesophagus), demographic data (age, gender, race), and comorbid atopy were noted. Results: 57 patients with PPI-REE and 81 cases of EoE were identified. Overall mean age (40.6 years), gender (61% male), presence of comorbid atopy (69%) did not differ between groups as described previously[1, 2]. The mean peak eosinophil count prior to PPI amongst those with EoE was 68.4 (range 23-204/HPF), and was 119.72 (range 23-666) after 8 weeks of PPI. 12/81(15%) with EoE sustained a >5 times (500%) increase in eosinophil count post PPI. The mean eosinophil count prior to PPI amongst those with PPI-REE was 65.60 (range 15-272/HPF), and was 6.6. (range 0-14) post treatment. 27 of 81 cases of EoE (33.33%) had a decline in eosinophil count, from a mean count of 97.4/HPF to 44.16/HPF, representing a 45% decrease. The demographic characteristics of patients with EoE who recorded a decline in eosinophil count with PPI did not differ compared to those with nil response (see table 1). Overall, of the 138 patients treated with PPI, 84 (60.8%) had a decrease in eosinophil count. Conclusion: Nearly 1/3 of patients with EoE previously determined to be PPI unresponsive had a 45% decline in eosinophil count (mean decrease of 53 eosinophils per HPF). While these patients did not have a large enough response to qualify as PPI-REE, it suggests that PPIs can be an important treatment even in patients with EoE. It remains necessary to correlate this histologic change with any improvements in symptoms or endoscopic findings.