After Removal of Adenomatous Polyps

Options for surveillance in patients who have had adenomatous polyps removed are colonoscopy and barium enema. Performance of these tests is judged by the proportion of adenomatous polyps detected, their ability to distinguish clinically important cancers and polyps from other lesions, and their effectiveness in reducing colorectal cancer incidence and mortality. The performance of these tests has been described previously in this report.

The best evidence of effectiveness of surveillance following identification of polyps or cancer is for colonoscopy. In the National Polyp Study, a cohort of 1,418 patients who had undergone complete colonoscopy and removal of one or more adenomatous polyps from the colon or rectum was followed for an average of 5.9 years with periodic colonoscopy and in some cases barium enema (Winawer, Zauber, Ho, et al., 1993). After adjusting for age, sex, and polyp size, rates of cancer were 76 to 90 percent lower than expected from comparison with three reference groups who had not undergone surveillance. It is not possible to distinguish the benefits of surveillance from those of the initial colonoscopy and clearance of the bowel. If the first surveillance colonoscopy was negative, subsequent examinations were unlikely to reveal further adenomatous polyps. The study used reference groups as controls, with the assumption that patients undergoing polypectomy would have experienced the same incidence of cancer as the reference population if they had not undergone polypectomy. However, the differences are so large that it is unlikely that bias could entirely account for them.

The National Polyp Study provides evidence for surveillance frequency. Patients who had undergone prior polypectomy were randomized to undergo surveillance colonoscopy either twice (at 1 and 3 years) or only once (at 3 years) after polypectomy (Winawer, Zauber, Ho, et al., 1993). The two groups showed no difference in the proportion of detected adenomatous polyps with advanced pathology (3.3 percent in both groups), suggesting that the first followup surveillance colonoscopy after polypectomy can be deferred for at least 3 years.

There have been no studies of surveillance after polypectomy using barium enema. Similarly, there are no published studies comparing surveillance with colonoscopy and barium enema. There also is no direct evidence of when surveillance may be stopped.

Patients with Inflammatory Bowel Disease

Surveillance for colorectal cancer in patients with inflammatory bowel disease (IBD) aims primarily to detect moderate to severe dysplasia and early cancers rather than polyps. FOBTs are not helpful because colorectal bleeding is a prominent feature of IBD. Sigmoidoscopy does not examine all areas of the bowel likely to harbor cancers. Barium enema does not identify flat infiltrating cancers or dysplasia - for these, diagnosis relies on biopsy and pathologic examination. Thus, surveillance is usually performed with colonoscopy and biopsies of suspicious lesions.

Evidence is insufficient to support use of colonoscopy in surveillance of patients with IBD in terms of reducing colorectal cancer mortality. In a study of 401 patients with ulcerative colitis, all of whom were offered surveillance, most cancers detected during regular surveillance were at an early stage (Lennard-Jones, Melville, Morson, et al., 1990). A recent decision analysis suggests that both surveillance colonoscopy and prophylactic colectomy increase life expectancy in patients with ulcerative colitis, and that both are better than no surveillance (Provenzale, Kowdley, Arora, and Wong, 1995).