Risk Factors for Metachronous Advanced Colorectal Neoplasia in a Cohort of Adenoma Patients: Advanced Morphology and Multiplicity

Abstract

Background Surveillance of adenoma patients aims to prevent colorectal cancer (CRC) by removing recurrent adenomas. Adenoma removal and subsequent surveillance can reduce CRC incidence by 76-90%. Colonoscopy is however scarce, expensive and potentially harmful. To ensure efficient use of resources, surveillance colonoscopy should be targeted at patients who will benefit most from the procedure. Current surveillance guidelines use advanced morphology or multiplicity as criteria for surveillance interval. However, none of the guidelines have separate recommendations for patients with both multiple and advanced adenomas.AimTo assess the relative risks of advanced andmultiple (≥3) adenomas separately and combined on metachronous advanced colorectal neoplasia in a representative cohort of adenoma patients. Methods We collected prospective data on adenoma patients from 10 hospitals throughout the Netherlands, using a nationwide histopathology registry to select newly diagnosed adenoma patients from 1988 to 2002. Patients with CRC history or CRC at index colonoscopy, hereditary cancer syndromes or IBD were excluded. Electronic medical records were reviewed until December 1, 2008 for follow-up. Index colonoscopy was defined as colonoscopy with first adenoma diagnosis. Presence of advanced (≥10 mm, a villous histology or high-grade dysplasia) or multiple (≥3) adenomas and the combination at index colonoscopy were considered as potential risk factors for metachronous advanced colorectal neoplasia (advanced adenoma or CRC) at first follow-up endoscopy. To assess hazard ratios (HR) for the relative risk we performed a Cox-regression analysis, adjusted for age and gender. Results 3,041 adenoma patients (55% male, mean age 61 yrs (range 40 88)) were analyzed, of whom 1,351 (44%) patients had advanced adenomas at index endoscopy, and 161 (6%) ≥3 non-advanced adenomas. Median interval (interquartile range) to first surveillance endoscopy was 21 months (12-39); 15 months (11-35) for patients with advanced and/or ≥3 adenomas, and 27 months (13-45) for patients with 1-2 non-advanced adenomas at index endoscopy (p<0.01). At follow-up, 831 patients had any colorectal neoplasia (adenoma or CRC), of whom 182 patients had advanced colorectal neoplasia, including 26 CRC cases. Relative risks for metachronous advanced colorectal neoplasia are given in Table 1. Conclusion Advanced adenomas and ≥3 adenomas at index colonoscopy are equally important risk factors for metachronous advanced colorectal neoplasia, resulting in a 3-fold increased risk of developing advanced colorectal neoplasia during follow-up. However, having both risk factors results in a 6-fold increased risk. The results suggest that advanced morphology and multiplicity should be used to tailor surveillance guidelines with a separate recommendation for adenoma patients that have both these risk factors. Table 1. Relative risk of advanced and multiple (≥3) adenomas on metachronous advanced colorectal neoplasia