4502 Colonoscopic surveillance of patients with a family history of colon cancer and past history of normal colonoscopy: is a 5-year interval between colonoscopies appropriate?

Abstract

Background: Patients with a family history of colorectal cancer (ie. diagnosis in a first degree relative) have an increased risk of developing adenomas and cancer. Experts recommend that these patients should have surveillance colonoscopy every 5 years after an initial normal colonoscopy. The aims of this trial are to: 1) determine the percentage of patients with adenomas on surveillance colonoscopy; 2) determine the percentage of patients with advanced adenomas (ie. villous features, adenoma size =10mm, presence of high grade dysplasia); and 3) assess risk factors for colonic adenomas in this population. Methods:In May 1999, consecutive patients with a family history of colorectal cancer who had a normal baseline colonoscopy 5 years earlier were offered a surveillance colonoscopy. On the day of their procedure, patients completed a questionnaire about risk factors potentially associated with adenoma formation including: age(>65 yo), male gender, African-American race, NSAID use (>2 doses/week), alcohol use (>2 ounces/day), and tobacco use (>20 pack-yrs). Multiple logistic regression was used to determine which risk factors were associated with adenomas. Results: To date, 38 patients have completed the trial. Demographic data: male:female ratio(59:41); mean age 57yo±8yrs; 98% Caucasian. Eleven percent(4/38) had advanced adenomas at surveillance colonoscopy, and 34% (13/38) had any adenomas present in the colon. Since only one African-American has entered the trial so far, race was excluded from multiple logistic regression analysis. The multiple logistic regression analysis did not identify any significant association between adenomas and risk factors. Age>65yo: OR=1.34(95%CI:0.2-9.7); male gender: OR=1.40(95%CI:0.3-6.8); alcohol use: OR=1.94(95%CI:0.1-57.5); and NSAID use:OR(95%CI:0.1-1.7). Among patients with advanced adenomas, 25%(1/4) had no adenomas in the left side of the colon (ie. distal to the splenic flexure). Among patients with any adenomas, 39% (5/13) had no adenomas in the left side of the colon. Conclusions: Expert recommendations that patients with a family history of colorectal cancer should have surveillance colonoscopy 5 years after an initial normal colonoscopy are supported by these preliminary results. Over 1/3 of adenomas would be missed if surveillance was performed with flexible sigmoidoscopy.