Alterations in Adenoma Characteristics Over Time: More Patients with >2 Small Adenomas and Not Associated With Metachronous Advanced Adenoma: Fellows-in-Training Award (Colorectal Cancer Prevention Category): Presidential Poster Award

Introduction: Adenoma size, number and pathology are used to stratify risk for metachronous advanced adenoma (> 9mm or villous features or high-grade dysplasia). Post-polypectomy colonoscopy in high risk (HR:> 2 adenomas or advanced adenoma) and low risk (LR:1-2, < 10 mm, tubular adenomas) patients is 3 vs 5-10 years respectively. High-definition (HD) endoscopy enhances diminutive polyp detection. We hypothesized that patients having colonoscopy in a contemporary era of HD colonoscopy are more likely to be HR due to > 2 small non-advanced adenomas vs a historic cohort (HC) before HD colonoscopy, and that > 2 small non-advanced adenomas in the contemporary cohort (CC) would not predict metachronous advanced adenoma. We (1) compared adenoma features in the HC (< 2006) and CC (≥ 2006) and (2) analyzed association of baseline adenoma features by time cohort on the risk of metachronous advanced adenoma. Methods: Patients having their 1st colonoscopy at age ≥ 50 years were identified through natural language processing. Exclusions included history of colorectal cancer or surgery, inflammatory bowel disease, incomplete colonoscopy or inadequate bowel preparation. Patients with follow-up colonoscopy > 24 months from baseline exam were included for aim 2. Baseline adenoma findings between CC and HC were compared by univariate analysis. Multivariate regression assessed impact of time cohort on baseline and post-polypectomy adenoma findings, adjusting for age, gender, baseline HR features. Results: 255,074 colonoscopies were performed between 1990 and 2015. 9,797 patients with baseline examination < 2006 and 59,617 ≥ 2006 were eligible for analysis. At first colonoscopy, CC patients were younger, more likely to be HR, and HR due to >2 small tubular adenomas, and adenomas > 9mm(Table1). Adjusted analysis confirmed CC patients were more likely than HC patients to be HR [OR 1.5, CI(1.4,1.7)], have > 2 adenomas [OR 2.3, CI(1.9, 2.7)] and > 2 small tubular adenomas without advanced pathology [OR 2.1, CI(1.6,2.6)] (Table2). Overall and in CC, baseline HR features of advanced pathology and size > 9mm predicted metachronous advanced adenoma, but not > 2 non-advanced adenomas compared to LR patients(Table3). Conclusion: In an era of HD colonoscopy, patients are more likely to be classified as HR, driven by > 2 adenomas especially non advanced adenomas. > 2 non-advanced adenomas do not predict risk of metachronous advanced adenoma in an era of HD colonoscopy.264_A Figure 1 No Caption available.264_B Figure 2 No Caption available.264_C Figure 3 No Caption available.