Background-Barrett's oesophagus carries a 30-fold to 40-fold increased
risk of oesophageal cancer. It is unknown whether endoscopic surveill
ance programmes reduce mortality from oesophageal cancer. Methods-A co
hort study was undertaken of all 166 patients in whom the diagnosis Ba
rrett's oesophagus had been established between 1973 and 1986. Results
-One hundred and fifty five of 166 patients could be traced (93%). Dur
ing a mean follow up of 9.3 years (amounting to 1440 patient years) ei
ght patients had developed oesophageal cancer at random intervals (one
case in 180 patient years). All but one of the tumours were diagnosed
at endoscopy for symptoms, three in the stage of carcinoma in situ. R
isk factors for the development of oesophageal cancer were extensive B
arrett's oesophagus exceeding 10 cm (p=0.02) and Barrett's ulcer at th
e time of intake (p=0.009). Seventy six patients were alive; three had
undergone surgery for oesophageal cancer and were without recurrence
respectively, 12.8 years, 12.1 years, and 7 months postoperatively. Se
venty nine patients had died; five of them had developed oesophageal c
ancer, but in only two cases this had been the cause of death (2.5%).
Conclusions-Oesophageal cancer is an uncommon cause of death in patien
ts with Barrett's oesophagus. The patients of this cohort would not ha
ve benefited from an endoscopic surveillance programme.