Optimal timing of surveillance colonoscopy after removal of low-risk adenomas

BACKGROUND In addition to early detection of colorectal cancer (CRC), screening programs can prevent deaths through removal of precursor lesions, with regular surveillance colonoscopies recommended as a preventive strategy in populations at increased risk of developing CRC. The recent update of the Australian colonoscopy surveillance guidelines suggests that the interval can potentially be increased from 5 to 10 years following a low risk adenoma (LRA = 1-2 tubular adenomas <10mm in size). This recommendation is largely based on expert opinion as there are limited supporting Australian data. The aim of the current study was to determine the impact of delayed surveillance colonoscopy on the incidence of advanced neoplasia (AN) after a finding of LRA. METHODS A retrospective review was conducted on outcome data from patients at elevated risk for CRC (with either a personal history of neoplasia or a family history of CRC), who were enrolled in a surveillance colonoscopy program in the Southern Adelaide Local Health Network. Data from all colonoscopies performed between January 1999 and January 2014 were reviewed for diagnosis of LRA. The findings at subsequent good quality colonoscopies conducted 4-6 years later(classified as “on time” according to previous Australian guidelines) and >6 years later(classified as “delayed”) were reviewed for AN. AN was defined as CRC or high-risk adenoma(HRA= adenomas with size ≥10mm, villous change, high grade dysplasia, sessile serrated polyps, or 3 or more small tubular adenomas). Patient demographic data were also collected. Multivariate logistic regression analysis was performed to determine risk factors of advanced neoplasia at different time intervals after LRA. RESULTS There were 960 follow-up colonoscopies performed post-LRA. Of these, 664 (69.2%) occurred on time whilst 295 (30.7%)were delayed. A total of 195 cases of AN were found including 2 carcinomas. The incidence of AN with delayed colonoscopy was significantly greater (72/295, 24.4%) compared to on time colonoscopy (123/664, 18.5%), with increases in incidence evident with delays in surveillance (Figure). On multivariate analysis, after adjustment for age, gender, family history of CRC and previous adenoma findings, the time since removal of LRA was a significant predictor for AN (OR 1.24, 95% CI 1.09-1.42). CONCLUSION Increasing surveillance colonoscopy intervals beyond 6 years significantly increases the risk for developing an advanced neoplasia following removal of LRA. The potential outcomes of extending surveillance to 10 years are unclear and further studies are necessary to determine the effects of such a recommendation.