Barrett's esophagus: a retrospective analysis of 13 years surveillance.

BACKGROUND AND AIMS: The incidence of esophageal adenocarcinoma has increased significantly. Barrett's esophagus (BE), a known precursor, has a high prevalence but only few patients with this condition progress to malignancy--surveillance and screening programs are controversial and lack proven efficacy. This retrospective analysis reviews the 13-year outcome for patients entered into a surveillance program.

METHODS: Data from patients with histologically proven Barrett's esophagus (1992-2003) that participated in a surveillance program were identified and analyzed retrospectively until 2005.

RESULTS: 404/536 patients had Barrett's esophagus confirmed histologically of which 212 (53%) were followed in a surveillance program (mean 3.95 years per patient). This resulted in 749 gastroscopies (3.5/patient). Histologically, Barrett's mucosa was seen in 54%, low-grade dysplasia in 18%, ulcerations in 9%, high-grade dysplasia in 2%. No metaplasia was seen in 13%, no biopsy was obtained in 3%. Nine of 212 patients (4.3%) under surveillance developed esophageal cancer; two presented with symptoms, requiring gastroscopy outside the surveillance program (1/2 was operated successfully, one had advanced disease). In seven asymptomatic patients, cancer was detected on routine endoscopy; curative esophagectomy was performed in six. All patients who developed cancer were male and all but one patient had dysplasia or ulcerations on index endoscopy.

CONCLUSION: During 13 years of Barrett's surveillance, 88% of all adenocarcinoma occurred in a subset of only 11% patients. To stratify surveillance for Barrett's esophagus, programs could focus on male patients with dysplasia or ulcerations on index endoscopy. However, the cost-effectiveness of this remains unproven.