초록/요약

Background: An optimal surveillance program is important to prevent advanced colorectal neoplasm. In this context, we have evaluated the cumulative risk of high-risk adenoma (HRA) or colorectal cancer (CRC) according to surveillance interval time after polypectomy. In addition, we assessed risk factors for late surveillance to determine whether late surveillance can impact the risk of subsequent advanced colorectal neoplasm. Methods: This was a multicenter retrospective cohort study involving 3562 subjects who had undergone removal of at least one adenoma at the index colonoscopy and who subsequently underwent a surveillance colonoscopy. The subjects were classified into an early, appropriate or late group depending on the timing of the surveillance colonoscopy, performed using modified U.S. guidelines. Results: With 3% of the study population with LRA and HRA at the index colonoscopy going on to develop HRA or CRC, the estimated surveillance intervals calculated would be 6.3 [95% confidence interval (CI) 5.42–7.10] years and 3.1 (95% CI 2.61–4.45) years, respectively. The predictors of late surveillance were female gender [odd ratio (OR) 1.21; 95% CI 1.04–1.40], having undergone the procedure in small-to-medium-sized cities (OR 1.92; 95% CI 1.36–2.72) and HRA at index colonoscopy (OR 1.37; 95% CI 1.19–1.59). The risk factors for subsequent HRA or CRC were late surveillance (OR 1.34; 95% CI 1.03–1.74), male gender (OR 2.13; 95% CI 1.54–2.95), having undergone the procedure in small-to-medium-sized cities (OR 1.63; 95% CI 1.11–2.40) and HRA at index colonoscopy (OR 2.60; 95% CI 2.04–3.33). Conclusions: Women, having undergone the procedure in small-to-medium-sized cities and the presence of an HRA at the index colonoscopy were found to be independent risk factors for late surveillance colonoscopy. Late surveillance is significantly predictive of subsequent HRA or CRC. © 2016 Japanese Society of Gastroenterology