Interobserver reproducibility in pathologist interpretation of columnar-lined esophagus.

Abstract

Confirmation of endoscopically suspected esophageal metaplasia (ESEM) requires histology, but confusion in the histological definition of columnar-lined esophagus (CLE) is a longstanding problem. The aim of this study is to evaluate interpathologist variability in the interpretation of CLE. Thirty pathologists were invited to review three ten-case sets of CLE biopsies. In the first set, the cases were provided with descriptive endoscopy only; in the second and the third sets, ESEM extent using Prague criteria was provided. Moreover, participants were required to refer to a diagnostic chart for evaluation of the third set. Agreement was statistically assessed using Randolph's free-marginal multirater kappa. While substantial agreement in recognizing columnar epithelium (K = 0.76) was recorded, the overall concordance in clinico-pathological diagnosis was low (K = 0.38). The overall concordance rate improved from the first (K = 0.27) to the second (K = 0.40) and third step (K = 0.46). Agreement was substantial when diagnosing Barrett's esophagus (BE) with intestinal metaplasia or inlet patch (K = 0.65 and K = 0.89), respectively, in the third step, while major problems in interpretation of CLE were observed when only cardia/cardia-oxyntic atrophic-type epithelium was present (K = 0.05-0.29). In conclusion, precise endoscopic description and the use of a diagnostic chart increased consistency in CLE interpretation of esophageal biopsies. Agreement was substantial for some diagnostic categories (BE with intestinal metaplasia and inlet patch) with a well-defined clinical profile. Interpretation of cases with cardia/cardia-oxyntic atrophic-type epithelium, with or without ESEM, was least consistent, which reflects lack of clarity of definition and results in variable management of this entity.