Risk Factors Associated with Incomplete Endoscopic Mucosal Resection of Colonic Polyps

Abstract

Introduction: Risk factors for incomplete endoscopic mucosal resection (EMR) of large colonic polyps are uncertain. We aimed to characterize the rates and risk factors of incomplete resections (IR) at a single US center. We hypothesize that polyp size > 40mm predicts IR and resections initiated by endoscopists specializing in colonic EMR protect against IR. Methods: Using an electronic endoscopic database, we identified patients who underwent EMR with at least one follow up colonoscopy from 2006-2015 at our institution. Patients with IBD were excluded. The primary outcome (IR) was defined as cases which the provider noted residual polyp at the end of the first resection attempt. Polyp characteristics and the endoscopist first performing EMR (specialist (SE) vs. non-specialist endoscopist) were noted. Univariate and multivariate analysis was used to identify factors associated with IR. Clinical outcomes were recorded. Results: 258 patients with 289 polyps underwent EMR with at least one follow up colonoscopy. 39/258 (15%) had IR. Age, gender, polyp location, and follow up were similar between complete resections (CR) and IR (Table 1). SE as the sole EMR provider achieved CR in fewer colonoscopies than non-SE± subsequent SE referral (mean 2.4 vs. 3, p< 0.0001). On univariate and multivariate logistic regressions, polyp > 40 mm (odds ratio [OR] = 2.3; 95% confidence interval [CI] = 1.1-14.9), piecemeal removal (OR = 7.5, CI 1.8-31.9), and difficulty lifting the polyp (OR = 3.1, CI 1.1-8.9) were associated with IR. Resections performed by SE alone were protective against IR (OR 0.31, CI 0.15-0.65). Based on our regression model, the probability of IR in piecemeal resections of polyps>40 mm with and without difficulty lifting (and started by a non-SE) is 88% and 56%,respectively. When SE are the sole EMR provider, the probability of IR drops to 41% and 11%, respectively, (Figure 1). Of the IR patients, 12 (28%) were referred for surgery, 2 of which were for adenocarcinoma and 1 of which were for high grade dysplasia. 16 (41%) achieved CR after a mean of 2 more colonoscopies (range 1-4), 7(18%) are actively undergoing further EMR, and 4 were lost to follow-up.Figure 1Table 1: Demographics, endoscopic findings and pathology of patients with complete and incomplete resections.Conclusion: In this single center US experience, IR of large colon polyps occurred infrequently. Polyps > 40mm, piecemeal resections and difficulty lifting were risk factors for IR. Primary resections by specialist endoscopists reduce the risk of IR and the number of procedures to achieve CR.