The role of histology in the diagnosis of non-erosive reflux disease: A systematic review and meta-analysis

Background and purpose: Non-erosive reflux disease (NERD) accounts for over half of all gastroesophageal reflux cases and is characterized by reflux symptoms with pathologic acid exposure on pH monitoring but no evidence of erosions on upper endoscopy. Ambulatory pH monitoring is limited by availability and patient tolerance. The utility of performing esophageal mucosal biopsies in diagnosing NERD is unclear. We conducted a systematic review and meta-analysis to determine the sensitivity of esophageal mucosal biopsies in diagnosing NERD.

Methods: Data were obtained from Embase and Ovid MEDLINE from inception to April 2021. Studies were included if esophageal mucosal biopsies were taken and analyzed using conventional histopathologic analysis in symptomatic NERD patients. Relevant data was including histologic abnormalities and location of the biopsy. Sensitivity and specificity were calculated against healthy controls or those with functional heartburn.

Results: The search yielded 2871 studies, of which 10 studies met our inclusion criteria and contained raw data. Histological abnormalities included histologic sum scores, papillary elongation, basal cell hyperplasia, and dilated intraepithelial spaces. When assessing for the presence of any abnormality, biopsies taken <3 cm from the lower esophageal sphincter (LES) had a pooled sensitivity of 0.71 (95% CI 0.64-0.77) and specificity of 0.64 (95% 0.54-0.73); however, analysis of individual histologic features such as the presence of eosinophils improved the sensitivity.

Conclusions: Although esophageal mucosal biopsies had poor sensitivity at diagnosing NERD, biopsies taken within 3 cm of the LES had higher sensitivity when pathologists reported upon eosinophils and dilated intraepithelial spaces.