460 High Rates of Loss to Follow-Up After Esophageal Food Impaction in Suspected Eosinophilic Esophagitis: Time to Standardize Care

Abstract

INTRODUCTION: Esophageal food impaction (EFI) is often the first presentation for patients with eosinophilic esophagitis (EoE). Management of EFI when EoE is suspected lacks evidence-based guidelines. The goals of this study are to evaluate the biopsy rates, medical management, follow-up and recurrence of EFI in patients with suspected EoE. METHODS: Adult patients presenting with EFI between 2007-2017 were identified by ICD-9 (935.1) and ICD-10 (T18.1) codes for foreign body in the esophagus. Inclusion criteria required endoscopic intervention with documentation of suspected EoE: defined as endoscopic features of rings, furrows, edema or exudate. RESULTS: We identified 471 cases of EFI requiring endoscopic intervention, with 171 suspected to have EoE. The biopsy rate at time of EFI for patients with suspected EoE was 46%. When patients were not biopsied at time of EFI (n = 93), the majority were initiated on empiric therapy without histologic diagnosis: PPI (90%) and topical steroids (1%). Follow-up endoscopy was recommended in 97% of patients not biopsied at time of EFI and 32% (n = 27) were lost to follow-up. Empiric therapy after EFI did not impact the diagnostic yield at time of follow-up endoscopy (median: 36 days) as rates of diagnostic biopsy (≥15 eos/hpf) were similar compared to those biopsied at time of EFI (66% vs 65%, P = 0.612). Time to follow-up endoscopy did not impact diagnostic yield (P = 0.21). While on therapy, 16% of patients (n = 9) had low level esophageal eosinophilia (5-14 per hpf) on follow-up EGD (median: 29 days). Over the 11 year period, recurrent EFI occurred in 16 patients (3.4%) with a median time to recurrence of 1.65 years. CONCLUSION: EFI presents a unique opportunity to diagnose, treat and establish care for patients with EoE. Our study identified low rates of esophageal biopsy and high rates of loss to follow-up. Empiric therapy did not impact diagnostic yield on subsequent endoscopy with early follow-up. Low level eosinophilia while on therapy should be interpreted cautiously as this may represent partially treated disease. Although uncommon, EFI and recurrence are a disease defining event that create significant patient discomfort and require costly intervention. Our study highlights the need for standardize practice at the time of EFI when EoE is suspected including a structured system to streamline outpatient care, avoid loss to follow up and ultimately, prevent long term complications from this chronic disease.