POPULATION-BASED STUDY OF THE TREATMENT AND PROGNOSIS OF CARCINOMA OFTHE RECTUM

Background Few population-based studies address the issue of treatment of carcinoma of the rectum (15 cm or less from the anal verge) both f rom surgical and epidemiological aspects. Methods Some 827 patients we re analysed in the cancer registry of the Cote-d'Or (Burgundy, France) from 1976 to 1990 (493 931 inhabitants). Results Resection for cure i ncreased from 57.2 per cent before 1981 to 77.0 per cent after 1985 (P <0.001), and the proportion of Dukes A and B cases from 35.8 to 52.5 p er cent (P<0.001). Among patients resected for cure, continence-preser ving resections were performed more frequently during the 1986-1990 pe riod (48.0 per cent) than during the two previous 5-year periods (20.0 per cent; P<0.001), more often in women, in the upper half of the amp ulla and for tumours of less than 45 mm. The operative mortality rate after curative surgery decreased from 13.9 to 3.7 per cent (P<0.001) b etween the first and the last period whereas the 5-year crude survival rate rose from 25.8 to 42.6 per cent (P<.0001). Age, stage of disease and period of diagnosis were independent prognostic factors of death in a relative survival model. Conclusion This study indicates that sig nificant advances have been achieved at a population level in the trea tment of rectal cancer in terms of diagnosis, continence-preserving pr ocedures and survival.