PWE-071 The natural history of Barrett's oesophagus: the view from a UK district general hospital: Abstract PWE-071

Abstract

Introduction All patients diagnosed with gastro-oesophageal reflux disease (GERD) in our DGH over the last 3 decades have been systematically followed up. We present the outcome of BE cohort diagnosed 1977–2006 and followed until 31.10.2009, set against the numbers of newly diagnosed GERD. Methods BE was diagnosed visually and confirmed by histology in about 90%. Follow-up (FU) was by surveillance endoscopy+biopsy in the fit or clinically (clinic visit/telephone survey) with endoscopy reserved for alarm symptoms in the others. The GI database closed for non-BE GERD 31.12.2001 but continues for BE. Results BE and GERD 1977-2001 Comparison 1. Incidence (Abstract 071): Both GERD and BE rose in successive 5-year periods. 2. Demography: BE patients were a decade older than reflux patients (mean age at presentation 62 vs 52 years); both had a slight male preponderance (BE 62%, GERD 55%). 3. Complications: Presentation with complications was more common in BE patients (haemorrhage and/or anaemia 10% vs 5%, stricture 10% vs 2%). 4. Mortality. BE 316/752 (42%), GERD 2799/11 622 (24%), mean FU 6 years, range 0–31. BE 1977-2009 Natural history. 1. Numbers continued to rise, 364 diagnosed in the final 5 years (2002-06), totalling 1116. 2. Oesophageal adenocarcinoma (EAC). 53 who presented with or developed EAC within 1 year are excluded as prevalent cancers. Outcome: The remaining 1063 BE had total FU of 6885 patient-years (mean 6.5 years range 0–31) during which 42 (4%) developed EAC, that is, 1 tumour per 164 patient-years of FU. The mean time from diagnosis of BE to EAC was 9 years (range 13 months–24.4 years) occurring at a mean age of 71 years (range 50–86). 3. Mortality: 320/1063 (30%) died during follow-up, 27 (2.5%) from EAC, 54 (5%) from other tumours, the remaining 239 (22%) mainly from cardio-vascular and respiratory causes. 4. Surveillance group, development of EAC, potentially curative resection and outcome. Endoscopic FU: EAC n=26; resection 12; 6/12 died from tumour. Clinical FU: EAC n=11; resection 4; all 4 died from tumour. No FU: EAC n=5; resection 1, still alive. Conclusions (1). During BE follow-up 4% developed EAC. (2) 8% of BE deaths were from EAC but the vast majority from other causes, perhaps reflecting their older age. (3) Surveillance. The rising prevalence of BE and increasing use of endoscopic therapy for early cancer in those unfit for resection has nevertheless strengthened the case for surveillance.