Background and study aims: The European guidelines for quality assurance in colorectal cancer (CRC) screening have established high-risk (≥ 5 adenomas or an adenoma ≥ 20 mm) and intermediate-risk (3 - 4 adenomas or at least one adenoma 10 - 19 mm in size, or villous histology, or high grade dysplasia) groups with different endoscopic surveillance intervals. The aim of this study was to evaluate the difference in the incidence of advanced neoplasia (advanced adenoma or CRC) between the two risk groups. Patients and methods: This retrospective group study included patients meeting high- or intermediate-risk criteria for adenomas detected in CRC screening programs and the COLONPREV study before European guidelines were adopted in Spain (June 2011) with a 3-year surveillance recommendation according to Spanish guidelines. The primary outcome measure was the incidence of advanced neoplasia in patients undergoing surveillance. The secondary outcome measure was the CRC incidence. We used an adjusted proportional hazards regression model to control confounding variables. Results: The study included 5401 patients (3379 intermediate risk, 2022 high risk). Endoscopic surveillance was performed in 65.5 % of the patients (2.8 ± 1 years). The incidence of advanced neoplasia in the high- and intermediate-risk groups was 16.0 % (59.0 cases/1000 patient-years) and 12.3 % (41.2 cases/1000 patient-years), respectively. The CRC incidence was 0.5 % (1.4 cases/1000 patient-years) and 0.4 % (1 case/1000 patient-years), respectively. The advanced neoplasia and CRC attributable risk to the high risk group was of 3.7 % and 0.1 %, respectively. In the proportional hazards analysis, the risk of advanced neoplasia was greater in the high-risk group (hazard ratio [HR] 1.5, 95 % confidence interval [CI] 1.2 - 1.8), with no significant differences in the CRC incidence (HR 1.6, 95 %CI 0.6 - 3.8). Conclusions: Patients meeting high-risk criteria have a higher incidence of advanced neoplasia during endoscopic surveillance. No differences were found in the CRC incidence at a 3-year surveillance recommendation.
© Georg Thieme Verlag KG Stuttgart · New York.