Alan J Cameron1, Amindra S Arora. 1. Division of Gastroenterology and Hepatology, Mayo Clinic, Rochester 55905, USA.
Abstract
OBJECTIVES: Barrett's esophagus (BE) is associated with esophageal reflux. The development stage of BE is not well described. Epidemiological evidence indicates that the columnar epithelium in BE is acquired and reaches its full length rapidly. We tested the hypothesis that BE might result from direct replacement of erosions in reflux esophagitis (RE). METHODS: At endoscopy, we compared the length and distribution of esophageal erosions in 50 patients with RE with the length and distribution of columnar epithelium in 50 patients with BE. RESULTS: The median length of erosions in RE was 2 cm, less than the median length of columnar epithelium in BE, 5 cm (p < 0.001). Erosions in RE were usually multiple and scattered, involving the entire circumference of the esophagus in only 10% of cases, but circumferential involvement by columnar epithelium was found in 68% of BE cases (p < 0.001). Circumferential involvement, 3 cm or longer, was found in 0% of cases of RE versus 56% of BE cases (p < 0.001). Two patients without RE or BE had large areas of epithelial loss of uncertain etiology. CONCLUSIONS: The length and distribution of erosions in RE differ greatly from the length and distribution of columnar epithelium in BE. It is unlikely that BE arises directly from areas of esophagitis. We suggest that BE may develop after loss of a long segment of squamous epithelium, with columnar replacement in the presence of continuing acid reflux.
OBJECTIVES: Barrett's esophagus (BE) is associated with esophageal reflux. The development stage of BE is not well described. Epidemiological evidence indicates that the columnar epithelium in BE is acquired and reaches its full length rapidly. We tested the hypothesis that BE might result from direct replacement of erosions in reflux esophagitis (RE). METHODS: At endoscopy, we compared the length and distribution of esophageal erosions in 50 patients with RE with the length and distribution of columnar epithelium in 50 patients with BE. RESULTS: The median length of erosions in RE was 2 cm, less than the median length of columnar epithelium in BE, 5 cm (p < 0.001). Erosions in RE were usually multiple and scattered, involving the entire circumference of the esophagus in only 10% of cases, but circumferential involvement by columnar epithelium was found in 68% of BE cases (p < 0.001). Circumferential involvement, 3 cm or longer, was found in 0% of cases of RE versus 56% of BE cases (p < 0.001). Two patients without RE or BE had large areas of epithelial loss of uncertain etiology. CONCLUSIONS: The length and distribution of erosions in RE differ greatly from the length and distribution of columnar epithelium in BE. It is unlikely that BE arises directly from areas of esophagitis. We suggest that BE may develop after loss of a long segment of squamous epithelium, with columnar replacement in the presence of continuing acid reflux.
Authors: Seamus J Murphy; Lesley A Anderson; Brian T Johnston; Deirdre A Fitzpatrick; Peter Rg Watson; Pauline Monaghan; Liam J Murray Journal: World J Gastroenterol Date: 2005-12-14 Impact factor: 5.742