764. PREDICTING AGE-SPECIFIC PREVALENCE OF BARRETT'S OESOPHAGUS IN AUSTRALIAN AND UNITED STATES

Abstract Background Barrett’s esophagus is the precursor to esophageal adenocarcinoma (EAC). EAC detected from endoscopic surveillance programs accounts for <10% of all EAC, suggesting a large number of patients with Barrett’s esophagus are likely unaccounted for. Several studies have estimated the observed prevalence of Barrett’s esophagus to be approximately 1%, but this may be an underestimate. The aim of this study is to use population incidence of EAC and rate of progression from Barrett’s esophagus from large cohort studies to estimate a realistic prevalence of Barrett’s esophagus. Methods A simple tree cohort model was created for progression to EAC from birth to death (100 years) for American and Australian population. Lifetime risk of esophageal cancer and adenocarcinoma were endpoints. This was then used to back-calculate age-specific and overall prevalence of Barrett’s esophagus using an optimisation algorithm. Results Lifetime risk of esophageal cancer and adenocarcinoma in US Whites is 0.56% and 0.36% respectively, while in Australian population is slightly higher at 0.81% and 0.61% (range 0.57% - 0.65%). Estimated overall prevalence of Barrett’s esophagus is ~3% (±0.3%) and ~5.4% (±0.6%) in US White and Australian populations, while in males is 5.3% (±0.6%) and 7.4% (±0.3%), respectively. Female cohorts were seen to have lower chance of developing esophageal cancer, adenocarcinoma, and Barrett’s esophagus. Conclusion Cohort studies of Barrett’s esophagus are likely to underestimate its prevalence of this silent disease. Computational models based on progression from Barrett’s esophagus to EAC accounting for undetected disease reveal higher estimates, which can help in future screening options.