Tu1102 EFFICACY AND SAFETY OF COLD ENDOSCOPIC MUCOSAL RESECTION OF LARGE COLONIC ADENOMAS

Abstract

Cold snare polypectomy is safe, and recent studies have shown a high level of efficacy for large sessile serrated lesions. In contrast, data on cold resection of large adenomas are limited, with previous studies utilizing high rates of adjuvant therapy (biopsy forceps or argon plasma coagulation). The aim of this study was to evaluate the safety and efficacy of a standardised cold EMR technique for large (≥20mm) colonic adenomas. We retrospectively analysed all consecutive large adenomas removed by cold EMR by the study endoscopists at a single academic endoscopy centre between June 2015-May 2019. High definition cap-fitted colonoscopy was performed with a standardized cold EMR technique1: a) submucosal injection of succinylated gelatin with methylene blue or indigo carmine; b) piecemeal cold snare resection with at least 3mm margin of normal mucosa; and c) targeted clip placement for persistent or focal ooze when early post-procedural anticoagulation was planned. Adjuvant therapy was never applied. All specimens were reviewed by a specialist GI pathologist. First surveillance colonoscopy (SC1) was performed at 3-6 months and SC2 at subsequent 6-12 months. Structured scar assessment involved white light and narrow band imaging examination. Flat bland scars were assessed endoscopically; non-bland scars were biopsied; suspected residual adenoma was treated by cold snare or cold biopsy avulsion followed by snare tip soft coagulation. 100 adenomas were removed in 93 patients over 51 months (mean age 68 years, 53 % female). Mean polyp size was 26mm (median 23mm, 20-70mm), 90% were located in the proximal colon and were predominantly Paris 0-IIa morphology (87%). Final histology was tubular/tubulovillous adenoma in 93 (58 low grade and 35 high grade dysplasia) and traditional serrated adenoma in 7. Unanticipated unifocal low risk submucosal invasive adenocarcinoma was present in a single case. A median of 2 clips were placed in 21 defects (23%). Adverse events occurred in 4 (4%); delayed bleeding 2 (2%) and hospitalisation for mild abdominal symptoms in 2 (2%). SC1 is complete for 95 (95%) lesions with surveillance of 3 lesions pending. No residual was detected in 88/95 (endoscopic 38, biopsy 50) at SC1. Residual neoplasia was identified in 7/95 (7.36%) and was treated endoscopically. All residual cases have completed SC2; 5 (71%) had no residual and 2 had further residual treated and are awaiting SC3. Overall SC2 is complete in 52 (52%) and residual was not detected in any additional cases. Selected large colonic adenomas can be safely removed by cold EMR with a low rate of residual neoplasia. Further study is required comparing cold and conventional EMR to further define which lesions and patients are best served by cold resection.