PTU-172 Oesophageal cancer risk: results of a prospective cohort with Barrett's oesophagus

IntroductionWidely varying rates of oesophageal adenocarcinoma (OAC) in patients with Barrett's oesophagus (BO) have been reported, with recent studies and meta-analyses suggesting a lower incidence, affecting the cost effectiveness of surveillance. However, advances in endoscopic therapy for dysplasia suggest surveillance should potentially be extended to more elderly patients. We have therefore examined long term outcomes in a BO cohort.MethodsBO patients with intestinal metaplasia from a prospectively maintained database (1982–2008) were analysed. Cancer registrations and causes of death from death certificates were obtained from the NHS information centre for health and social care and cross-referenced with local data. The incidence of OAC was calculated as events per 100 person years (% per year) of follow-up. Regression analysis was used to determine associations between the OAC development and age, sex, hiatus hernia, BO length, strictures and ulcers. Standardised mortality ratios were calculated using age adjusted indirect standardisation. Patients were subdivided into those suitable for surveillance and those deemed unfit due to age or comorbidity.Results713 (429 male, median age at diagnosis 64 years, range 30–92) BO patients were in the cohort. After a median of 11 (range 2–24) years of follow-up, 38 (27 male, median age 70 (48–90)) patients were diagnosed with OAC. The incidence of OAC was 0.5% per annum (p.a.). In patients considered suitable for surveillance, the incidence was 0.6% p.a. compared to 0.3% (p=0.06) in those not surveyed due to age or comorbidity. The rate of OAC in surveyed patients from 1982 to 1989 was 0.6% p.a., from 1990 to 1999 0.4% and from 2000 to 2008 0.5%. OAC was associated with increasing BO length (OR 1.11 (95% CI 1.01 to 1.13), p=0.03), but not with male sex (OR 1.66 (95% CI 0.8 to 3.4)), hiatus hernia (OR 1.31 (95% CI 0.68 to 2.57)), ulcer (OR 0.39, (95% CI 0.01 to 1.65)) or stricture (OR 0.97 (95% CI 0.37 to 3.1)). Standardised mortality ratios was elevated for the whole group at 181 (95% CI 162 to 181). Increasing age was associated with dying from OAC (OR 1.09 (95% CI 1.07 to 1.11) but not with the development of OAC (OR 0.99, (95% CI 0.96 to 1.01). 41% deaths in the cohort were from cardiorespiratory disease.ConclusionThe risk of OAC within this prospectively surveyed cohort was 0.5% per annum, which is higher than recent estimates. The rate of OAC in surveyed BO appears to have remained stable over the last 3