A majority of patients with Barrett's oesophagus are unlikely to benefit from endoscopic cancer surveillance

Background Endoscopic cancer surveillance has been advocated for patients with Barrett's oesophagus. However, only a small minority of patients dies from adenocarcinoma in Barrett's oesophagus. It has been calculated that endoscopic cancer surveillance will only add to the quality of life of individuals in whom the incidence of adenocarcinoma in Barrett's oesophagus is greater than 1/200 patient-years. Objective To determine the proportion of a consecutive cohort of patients, in whom Barrett's oesophagus was diagnosed over a 5-year period, likely to benefit from endoscopic cancer surveillance. Methods All patients who had died during the observation period or were over 75 years old and those with diseases likely to impair survival were excluded. Next, all patients in whom the risk of developing adenocarcinoma in Barrett's oesophagus fell below 1/200 patient-years were excluded (including all women, all men under the age of 60 and all men with Barrett's oesophagus of < 3 cm in length). Patients with dysplasia of any degree and/or presence of an ulcer or stricture in Barrett's oesophagus were reinstated. Results Of 335 adult patients diagnosed with Barrett's oesophagus but without adenocarcinoma or high-grade dysplasia, 75 had died from unrelated causes, 47 had other diseases limiting survival and 59 were over 75 years old. After exclusion of all women, all men with Barrett's oesophagus of < 3 cm in length and all men under 60 years old, 15 patients were left. However, 32 were reinstated because of risk factors and another five because of insufficient data, resulting in 52 of the original 335 patients (15.5%) being eligible for endoscopic cancer surveillance. Conclusion This study suggests that less than 20% of patients identified with Barrett's oesophagus at routine endoscopy would benefit from endoscopic cancer surveillance. Prospective surveillance programmes should be limited to patients with an increased cancer risk and a good health profile. Eur J Gastroenterol Hepatol 13:639-645 (C) 2001 Lippincott Williams & Wilkins.