Objectives: Barrett's mucosa is considered as a mosaic of three epithelial types but little is known about the topography of intestinal metaplasia in columnar lined esophagus. The aims of the study were to determine the prevalence of intestinal metaplasia within long and short segments of columnar lined esophagus and to analyze the distribution of the intestinal metaplasia within long segments of Barrett's esophagus.
Patients and methods: The study was performed on the initial endoscopy carried out among 112 patients enrolled in an endoscopic surveillance program. Seventy-two patients with columnar mucosa extending more than 3 cm into the esophagus (group I) and 40 patients with a short segment of columnar mucosa (group II) had multiple biopsies according to a standardized protocol. 1163 biopsies were analyzed on the whole: 949 biopsies in group I and 214 biopsies in group II.
Results: Intestinal metaplasia was identified among 650 (68.5%) and 50 (23.4%) biopsies in groups I and II respectively (P<10-7). The proportion of biopsies with foci of intestinal metaplasia increased significantly with the length of the columnar mucosa. The diagnosis of Barrett's esophagus was confirmed in 100% of the patients in group I and in 45% of the patients in group II. In long segments of Barrett's esophagus, intestinal metaplasia was more frequently observed in the 2 upper thirds of the columnar mucosa that in the lower third (P<10-7). Detailed mapping of the distribution of epithelial types within the columnar lined esophagus identified three patterns of distribution of intestinal metaplasia within long segments of Barrett's esophagus: unifocal, multifocal and diffuse, in 5%, 56% and 39% of the patients respectively. Dysplasia was present in 15% of patients with long segments of Barrett's esophagus and 11% of patients with short segments (NS).
Conclusion: The distribution of intestinal metaplasia within columnar lined esophagus is heterogeneous and three distinct patterns can be identified: unifocal, multifocal and diffuse. Considering the risk of sampling error, the current recommendations concerning the biopsy protocols are mandatory until the validation of new techniques such as chromoendoscopy or magnifying endoscopy.