Thermal ablation of the defect margin after endoscopic mucosal resection (EMR-T) is associated with reduced recurrence in the treatment of large (≥20-mm) nonpedunculated colorectal polyps (LNPCPs), according to a prospective international cohort study.
Residual or recurrent adenomas (RRAs) are found during 15%-20% of first surveillance endoscopies. EMR-T was previously shown in a randomized trial to be effective at reducing adenoma recurrence during surveillance endoscopy (relative risk, 0.3; P < .01).
The U.S. Multi-Society Task Force currently recommends EMR-T for LNPCPs, but real-world effectiveness remains unknown, wrote Mayenaaz Sidhu, MBBS, of the department of gastroenterology and hepatology at Westmead Hospital in Sydney and colleagues in Gastroenterology. Therefore, they undertook an international, multicenter, prospective trial to evaluate the technique in the real world.
The researchers analyzed data from consecutive patients who were referred for treatment of LNPCPs at six tertiary centers. Between May 2016 and August 2020, the study included 1,049 LNPCPs from 1,049 patients. The mean age was 67.3 years, and the median lesion size was 35 mm. Of LNPCPs, 58.7% were tubulovillous adenomas. EMR was technically successful in 98.9% of cases. Overall, 19.1% of cases required an auxiliary modality to completely remove polypoid tissue; most often this was cold avulsion with adjuvant snare-tip soft coagulation (44.4%).
Complete EMR-T was achieved in 95.4% cases. Reasons for failure included extensive post-EMR defect (n = 29), unstable colonoscope position or difficult access (n = 14), and intraprocedural adverse events (n = 5).
Of 803 patients eligible for surveillance colonoscopy, 94% underwent the procedure at a median interval of 6 months. Overall, RRAs were found in 3% of cases. Among lesions with complete EMR-T, 1.4% (10 of 707) had RRAs at first surveillance colonoscopy versus 27.1% (13 of 48) with incomplete EMR-T (P < .001). In cases with incomplete EMR-T, lesions were larger (median size, 42.50 mm vs. 37.60 mm; P = .03), there was longer procedure time (mean, 60.2 vs. 35.0 minutes; P = .01), and there was a greater likelihood of referral for surgery (8.3% vs. 3.0%; P = .04).
Intraprocedural bleeding occurred in 6% of cases, and endoscopic hemostasis was achieved in all. Clinically significant post-EMR bleeding occurred in 6.8% of cases, 59.2% of which were managed conservatively, and the remainder were evaluated endoscopically. Bleeding was controlled in every case.
Unlike RRA risk scores that use size, morphology, site, and access score, EMR-T can be used proactively to reduce RRA frequency. It is believed to work by thermally ablating microscopic tissue at the margin. The adverse events reported in the current study were similar to a systematic review and meta-analysis.
“These findings clearly support and exceed those of a recent randomized trial for EMR-T in the colorectum. They likely reflect refinements in the performance of EMR-T over time, due to greater technical experience and enhanced confidence in its safety. At its inception, the approach to EMR-T may have been timid, however, as experience grew and the safety of EMR-T became evident, a meticulous approach to uniform and complete thermal ablation of the defect margin became the standard of care,” the authors wrote.
They added that EMR-T has been shown to benefit in complex LNPCPs, including those that have undergone previous excision attempts and those involving the anorectal junction. The procedure has no added cost, since many endoscopists can readily use snare-tip soft coagulation to manage bleeding events.
“Thermal ablation of the defect margin should be viewed as an essential component of high-quality EMR for LNPCPs, consistent with recent recommendations by the U.S. Multi-Society Task Force on Colorectal Cancer,” the authors wrote.
The study was funded by the Cancer Institute of New South Wales, the Gallipoli Medical Research Foundation, and the University of British Columbia. One author reported research support for Olympus, Cook Medical, and Boston Scientific, but the remaining authors disclosed no conflicts.