Risk factors for dysplasia in patients with Barrett's esophagus (BE): results from a multicenter consortium.

Studies show Barrett's esophagus prevalence increases with age, while mean length of Barrett's esophagus is unchanged. Few data are available about the relationship between age and length on the development of dysplasia. Our aim was to assess age and length as risk factors for dysplasia. Consecutive patients with Barrett's esophagus were enrolled in a multicenter study establishing a tissue bank of Barrett's esophagus patients 1994 and 1998. Demographics, length of Barrett's esophagus (centimeters), and histology were recorded. Risk factors for dysplasia were assessed, including patient age, gender, and length of Barrett's esophagus. Statistical analysis was performed comparing prevalence of dysplasia (which included the presence of any carcinoma and high- or low-grade dysplasia) to age and length. In all, 309 patients were studied [278 (90%) male and 31 (10%) female]: 5 had adenocarcinoma of the esophagus, 11 had high-grade dysplasia, and 29 had low-grade dysplasia. Patients with Barrett's esophagus without dysplasia were younger than those with dysplasia [62 +/- 0.8 years vs 67 +/- 1.7 years (mean +/- SEM, P = 0.02)]. The risk of dysplasia increased by 3.3%/yr of age. Mean length of Barrett's esophagus in patients with Barrett's alone vs dysplasia was 4.0 +/- 0.2 cm vs 5.4 +/- 0.4 cm (P = 0.003). Patients with Barrett's esophagus length > or = 3 cm had a significantly greater prevalence of dysplasia compared to length < 3 cm (23% vs 9%, P = 0.0001). The risk of dysplasia increased by 14%/cm of increased length. Multivariate analysis showed age and length to be independent risk factors. In conclusions: prevalence of dysplasia is strongly associated with age and length of Barrett's esophagus. These preliminary results can be used to develop a strategy for screening/surveillance based on age and length of Barrett's epithelium.