Search

Background & Aims:: The contribution of duodeno-gastroesophageal reflux to the development of Barrett's esophagus has remained an interesting but controversial topic. The present study assessed the risk for Barrett's esophagus after partial gastrectomy. Methods:: The data of outpatients from a medicine and gastroenterology clinic who underwent upper gastrointestinal endoscopy for any reason were analyzed in a case-control study. A case population of 650 patients with short-segment and 366 patients with long-segment Barrett's esophagus was compared in a multivariate logistic regression to a control population of 3047 subjects without Barrett's esophagus or other types of gastroesophageal reflux disease. Results:: In the case population, 25 (4%) patients with short-segment and 15 (4%) patients with long-segment Barrett's esophagus presented with a history of gastric surgery compared with 162 (5%) patients in the control population, yielding an adjusted odds ratio of 0.89 with a 95% confidence interval of 0.54-1.46 for short-segment and an adjusted odds ratio of 0.71 (0.30-1.72) for long-segment Barrett's esophagus. Similar results were obtained in separate analyses of 64 patients with Billroth-1 gastrectomy, 105 patients with Billroth-2 gastrectomy, and 33 patients with vagotomy and pyloroplasty for both short- and long-segment Barrett's esophagus. Caucasian ethnicity, the presence of hiatus hernia, and alcohol consumption were all associated with elevated risks for Barrett's esophagus. Conclusions:: Gastric surgery for benign peptic ulcer disease is not a risk factor for either short- or long-segment Barrett's esophagus. This lack of association between gastric surgery and Barrett's esophagus suggests that reflux of bile without acid is not sufficient to damage the esophageal mucosa.

Background: Heterotopic gastric-type mucosa occurs as a flat island or islands of red mucosa in the proximal third of the esophagus where it gives rise to the ''cervical inlet patch'' (CIP). The aim of the present study was to delineate the clinical epidemiology of the CIP, especially its possible relationship to Barrett's esophagus. Methods: A case-control study compared 53 case subjects with CIP and 4882 control subjects without CIP. In a multivariate logistic regression, the presence of CIP was chosen as the outcome variable, whereas demographic characteristics, social habits, and presence of other endoscopic diagnoses served as predictor variables. Results: The prevalence of CIP was 1.1%. Its presence was associated with hiatal hernia (odds ratio 2.26: 95% CI [1.12, 4.56]) gastric ulcer (2.93: 95% CI [1.34, 6.40]) and Barrett's esophagus (4.41: 95% CI [2.31, 8.41]). Conclusions: The coincidence of the cervical inlet patch and Barrett's esophagus could suggest a shared embryonic etiology. (Gastrointest Endosc 2001;53:717-21.)