Optimizing endoscopic resection for challenging large non-pedunculated colorectal polyps

Abstract

Endoscopic mucosal resection (EMR) has revolutionized the management of large (≥ 20mm) non-pedunculated colorectal polyps. However, there are still challenges in performing EMR. Our aim was to assess EMR outcomes for challenging LNPCPs using the Australian Colonic Endoscopic Resection (ACE) study, a prospective multicenter observational cohort.

No difference in technical success or recurrence were identified between LNPCPs at the anorectal junction (ARJ-LNPCPs) and large non-pedunculated rectal polyps. No recurrence was identified at first surveillance colonoscopy (SC1; 0.0% vs. 25.0%; p=0.002) amongst 30 ARJ-LNPCPs treated by EMR with margin thermal ablation (EMR-T) vs. those that did not.

Comparing a universal EMR algorithm (UEA) and a selective resection algorithm (SRA), significant differences in cancer after EMR (SRA 1 (1.0%) vs. UEA 35 (12.1%); p = 0.001), and curative oncologic resection (SRA 7 (33.3%) vs. UEA 2 (5.7%); p = 0.010) were identified.

Significant differences in resection duration (35 minutes vs. 25 minutes; p<0.001) technical success (93.0% vs. 96.6%; p=0.026) and use of adjuvant modalities (46.2% vs. 7.6%; p<0.001), were identified between previously attempted LNPCPs (PA-LNPCPs) and naïve LNPCPs. Recurrence was not identified in 65 PA-LNPCPs which underwent EMR-T at SC1 vs. 9 (18.0%; p<0.001) which did not.

Significant deep mural injury (S-DMI) occurred in 101 cases (2.7%) which underwent EMR. Successful defect closure was achieved in 98 (97.0%). No difference in technical success (94 (93.1%) vs. 3316 (91.7%) p = 0.62) or SC1 recurrence (12 (20.0%) vs. 363 (13.6%); p = 0.15) were identified between LNPCPs with and without S-DMI.

Significant differences in sensitivity (90.9% vs. 52.7%), specificity (96.3% vs. 93.7%) and SMIC miss rate (0.6% vs. 5.9%) between flat and nodular LNPCPs were identified (all p < 0.027). Multiple logistic regression identified nodular morphology (OR 7.2; 95%CI 2.8-18.9) as a predictor of missed cancer.