Endoscopic screening in the prevention of colorectal cancer.

Guidelines and recommendations regarding screening investigations and surveillance to prevent death from colorectal cancer have been released by several national medical societies and the World Health Ž Organization Winawer et al., 1995, 1997; Schmiegel . et al., 2000 . These guidelines are based on available clinical and statistical information at the time they were set up by expert panels. The strength of evidence for the recommendations given varies. There is no direct evidence from large, randomized, prospective trials for the effectiveness of endoscopic screening to prevent death from colorectal cancer. In average-risk subjects several case control studies reveal a reduction in mortality from colorectal cancer by screening sigmoidŽ oscopies Selby et al., 1992; Newcomb et al., 1992; . Muller and Sonnenberg, 1995 . In individuals at increased risk for the development of colorectal cancer it has been shown that preventive endoscopies may reduce cancer incidence by early detection and removal of precancerous lesions. Moreover, Ž cancers may be detected at an earlier stage Winawer . et al., 1993a; Thiis-Evensen et al., 1999 . The prevalence of adenomas of the colon increases with age. Adenomatous polyps can be found in almost a quarter of the population at the age of Ž . 50 years Williams et al., 1982 . These adenomas account for the development of most cancers of the colon and rectum. Also cancer incidence increases steeply over the age of 50. In the western world about 100 new cases of colorectal cancer per 100 000 inhabitants per year can be expected in people at the age of 58 and 200 new cases in people at the age Ž . of 65 Winawer et al., 1997 . This is the basis of the recommendation to start endoscopic screening in average-risk subjects at age 50. More than half of all neoplastic lesions of the colon are confined to the distal part in the reach of the flexible sigmoidoscope. Sigmoidoscopy is a relatively inexpensive endoscopic procedure with simple bowel preparation, involving very little discomfort, inconvenience and risk for the patient. But, visualizing only about one-third of the colon, it leaves proximal polyps and cancers undetected. The chance of right-sided neoplastic polyps is one in three in patients with adenomas or carcinomas Ž . found at sigmoidoscopy Grossman et al., 1989 . Therefore, a total colonoscopy is often performed in these cases, both to rule out the presence of additional polyps more proximally and for removal of the lesions. The reduction in mortality from rectosigmoid cancer achieved by preventive sigmoidoscopies is estimated to be about 60 70%. In the most comprehensive investigation on this issue Selby and co-workers analysed data from the Kaiser Permanente Medical