970 High Rate of Advanced Neoplasia At Surveillance in Patients With Nonpolypoid (Flat and Depressed) Colorectal Neoplasms: Longitudinal Matched Cohort Study

Abstract

970 High Rate of Advanced Neoplasia At Surveillance in Patients With Nonpolypoid (Flat and Depressed) Colorectal Neoplasms: Longitudinal Matched Cohort Study Sarah K. McGill*, Roy M. Soetikno, Robert V. Rouse, Hobart Lai, Tonya Kaltenbach Division of Gastroenterology, University of North Carolina, Chapel Hill, NC; Division of Gastroenterology, Veterans Affairs Palo Alto Health Care System, Stanford University School of Medicine, Palo Alto, CA; Division of Gastroenterology, Veterans Affairs Palo Alto Health Care System, Palo Alto, CA Background/Aims: Nonpolypoid colorectal neoplasms (NP-CRN) are more likely to harbor high-grade dysplasia or early cancer compared to those that are polypoid. However, it is unknown whether patients with NP-CRN have a propensity to develop advanced neoplasia at higher rates compared to patients with polypoid neoplasia alone. We aimed to determine the risk of advanced neoplasia at first surveillance colonoscopy in patients with at least one NP-CRN compared to patients with only polypoid neoplasia. Methods: We performed a longitudinal cohort study of patients undergoing elective colonoscopy from 2000-2004 found to have R1 NP-CRN, and then matched them in a 3:1 manner with control patients found to have one or more polypoid neoplasms in the same year. We extensively reviewed clinical, endoscopic and pathologic digital records, and followed surveillance colonoscopy data to 2014. We calculated the risk of advanced neoplasia at surveillance for our primary outcome, using the standard definition of advanced neoplasia: R1 tubular or sessile serrated adenomas R10 mm, tubulovillous adenoma, high-grade dysplasia, or invasive cancer. Advanced neoplasia diagnoses were reviewed by a GI pathologist. Results: We identified 4454 patients who underwent colonoscopy over the 5-year period, and studied 341 (7.7%; 95% CI: 6.9-8.5%) with R1 NP-CRN and 1025 matched controls with R1 polypoid neoplasm. Patients with NP-CRNs had a significantly higher rate of advanced neoplasia at surveillance colonoscopy than patients with polypoid neoplasia (16.1% vs. 8.8%, risk ratio 1.8, 95% CI 1.3-2.7, p!.001). The advanced neoplasms detected at surveillance colonoscopy in patients with NP-CRN compared to patients with polypoid neoplasms were more likely to be nonpolypoid (60% vs. 33%, RR 1.8), and located in the proximal colon (72% vs 54%, RR 1.3). Overall, flat neoplasms at surveillance were 5 times more likely to be advanced than polypoid neoplasms (RR 5.0, 95% CI 3.7-6.7, p!.001). Patients with NP-CRNs compared to controls also had higher rates of advanced neoplasia (63% vs. 26%, p!.001) and harbored more neoplasia (mean SD 3.2 2.9 vs. 2.1 1.6, p!.001) at baseline exam. One or more NP-CRN were found on surveillance in 17% of patients with NP-CRN at baseline, compared to only 7% of control patients (RR 2.3, 95% CI 1.6-3.3). Demographic characteristics of patients at baseline were similard mean age was 66 ( SD10) and most were men (98%) and tobacco users (64%). However, NSAID use was more common in patients with NP-CRN (54% vs 35%) and the indication for baseline colonoscopy was more likely to be screening in control patients (47% vs 41%). Conclusions: Patients with NP-CRN have a high rate of advanced neoplasia at baseline and surveillance examinations and moreover, a significantly greater risk for developing advanced neoplasia than patients with polypoid neoplasia alone.