Background & aims: As the risk of esophageal adenocarcinoma (EAC) and mortality in patients with Barrett's esophagus (BE) are important determinants of the potential yield and cost-effectiveness of BE surveillance, clarification of these factors is essential. We therefore performed a systematic review and meta-analysis to determine the incidence of EAC and mortality due to EAC in BE under surveillance.
Methods: Databases were searched for relevant cohort studies in English language that reported EAC risk and mortality due to EAC in BE. Studies had to include patients with histologically proven BE, documented follow-up, and histologically proven EAC on surveillance. A random effects model was used with assessment of heterogeneity by the I(2)-statistic and of publication bias by Begg's and Egger's tests.
Results: Fifty-one studies were included in the main analysis. The overall mean age of BE patients was 61 years; the mean overall proportion of males was 64%. The pooled estimate for EAC incidence was 6.3/1000 person-years of follow-up (95% confidence interval, 4.7-8.4) with considerable heterogeneity (P < .001; I(2) = 79%). Nineteen studies reported data on mortality due to EAC. The pooled incidence of fatal EAC was 3.0/1000 person-years of follow-up (95% confidence interval, 2.2-3.9) with no evidence for heterogeneity (P = .4; I(2) = 7%). No evidence of publication bias was found.
Conclusions: Patients with BE are at low risk of malignant progression and predominantly die due to causes other than EAC. This undermines the cost-effectiveness of BE surveillance and supports the search for valid risk stratification tools to identify the minority of patients that are likely to benefit from surveillance.
Copyright 2010 AGA Institute. Published by Elsevier Inc. All rights reserved.