Background: CSP 380-Risk Factors for Large Colonic Adenomas, a screening colonoscopy study, enrolled 3,121 patients from 1994-97 at 13 VA sites. We report 10 year outcomes based on risk group after baseline colonoscopy.
Methods: Asymptomatic veterans age 50-75 who had not undergone colorectal cancer (CRC) screening in the prior 10 years underwent colonoscopy and were classified into 6 risk groups: no neoplasia, 1-2 small adenomas, 3-10 adenomas, > 10 adenomas, advanced adenoma (polyp ≥ 1 cm, villous histology, or high grade dysplasia), and CRC. Subjects were followed for 10 years until death or last colonoscopy. We report the proportions who developed advanced neoplasia (AN, defined as advanced adenoma or CRC) and CRC.
Results: Of the patients enrolled, 1,917 had at least one follow-up colonoscopy. Of these 1,917, 933 had no neoplasia at baseline; 4.0% developed AN, including 0.8% with CRC. At baseline, 560 patients had 1-2 small adenomas; 5.9% developed AN, including 0.9% with CRC. For those with 3-10 adenomas (134), 15.7% developed AN, including 0.7% with CRC. Two had > 10 adenomas; neither developed AN or CRC. At baseline, 265 patients had advanced adenoma; 19.2% developed AN, including 1.9% with CRC. Twenty-three patients had baseline CRC; 34.8% developed AN, including 21.7% with CRC. Patients with baseline CRC were at highest risk of developing AN. Those with advanced adenoma or 3-10 adenomas were at moderate risk. Those with no neoplasia or 1-2 small adenomas were at lowest risk. Except for those with baseline CRC, all risk groups showed a dramatic decline in CRC incidence after 3 years, with 10-year CRC risk similar to those with no neoplasia at baseline.
Conclusion: 1) Baseline colonoscopy result is a predictor of future risk for AN. 2) Those with baseline CRC remain at risk for recurrence. Intensive surveillance is warranted. 3) Those with 1-2 small adenomas or no adenomas at baseline have a low risk of AN after 2-3 years, and may not need surveillance. 4) Those with baseline advanced adenoma or > 3 adenomas show a dramatic decline in CRC risk after 3 years, suggesting that frequent surveillance may only be necessary in the first 5 years.