Barrett's surveillance identifies patients with early esophageal adenocarcinoma

Abstract

Background: Barrett's surveillance for dysplasia is recommended, but few studies have documented the benefit of endoscopic surveillance for dysplasia or cancer.

Objectives: Using a retrospective study design, we aim to demonstrate the impact of a Barrett's surveillance program on the stage of esophageal adenocarcinoma and identify factors for progression of metaplasia to cancer.

Subjects: The Institutional Review Board at Veterans Affairs Connecticut Healthcare approved the study. We report a retrospective review of a prospectively followed Barrett's cohort in a surveillance program and compared their outcome with patients with a new diagnosis of esophageal adenocarcinoma, identified at the same center between 1999 and 2005.

Results: There were 248 patients with Barrett's esophagus entered into a surveillance program from 1999 to 2005. During the surveillance period of 987 patient-years, 5 (0.5% patient-year) patients developed esophageal adenocarcinoma. During the same period, 46 patients were diagnosed with new-onset esophageal adenocarcinoma outside of our surveillance program. Only 5% of these patients had a history of gastroesophageal reflux disease. There were 248 patients who underwent a mean number of 2.7+/-1.7 upper endoscopic procedures, with 26 (10%) patients developing dysplasia. Compared with nonsurveillance, more patients had early stage of cancer in the surveillance group (P <.001). All 5 patients with cancer diagnosed from Barrett's esophagus surveillance endoscopy were alive, compared with 20 of 46 (43%) patients with cancer diagnosed outside of the surveillance program. The length of Barrett's segment >3 cm was found to be associated with development of dysplasia, P=.004 (odds ratio 1.2; 95% confidence interval, 1.07-1.34).

Conclusion: Patients with Barrett's esophagus undergoing endoscopic surveillance benefit from early-stage cancer diagnosis. Progression to adenocarcinoma is low, but long-segment and high-grade dysplasias have an increased risk of cancer. A significant number of patients with newly diagnosed esophageal adenocarcinoma do not complain of gastroesophageal reflux disease and are therefore not investigated for Barrett's esophagus nor entered into surveillance. Patients and physicians can use this information in making a decision about surveillance.