Cold snare endoscopic mucosal resection (EMR) was safer than hot snare EMR for large nonpedunculated colorectal polyps but carried a higher risk of residual or recurrent adenoma or neoplasia, according to an individual patient data meta-analysis published in Gut.
In the pooled analysis, major adverse events occurred in 1.7% of resections with cold snare EMR vs. 5.9% with hot EMR. Residual or recurrent adenoma or neoplasia at first follow-up occurred in 26.5% vs. 12.8%.
The authors wrote that the growing use of cold snare techniques for large adenomas appears to reflect mainly an expectation of lower bleeding risk, “while potentially underestimating the clinical relevance of recurrence, including malignant recurrence.”
The investigators conducted a post hoc individual patient data meta-analysis of four randomized controlled trials. The trials were conducted at 50 centers in Germany, Australia, Spain, the United States, and Canada. The analysis included 1,509 nonpedunculated polyps 20 mm or larger in 1,423 patients. Mean patient age was 66.1 years, and mean lesion diameter was 31.3 mm.
There were two primary outcomes. The first was major adverse events, defined as deep mural injury or perforation and/or delayed bleeding, assessed 14 to 30 days after resection. The second was residual or recurrent adenoma or neoplasia at the first follow-up endoscopy four to six months after resection, confirmed by biopsy or retreatment specimen. Follow-up data were available for 1,315 polyps (87.3%).
Safety outcomes
After adjustment, cold snare EMR was the only factor independently associated with a lower risk of major adverse events.
The safety difference was also evident across individual outcomes. Clinically significant postendoscopic bleeding occurred in 1.7% of patients treated with cold snare EMR and 3.5% of those treated with hot EMR. Deep mural injury or perforation occurred in no patients in the cold snare EMR group compared with 2.6% in the hot EMR group.
Recurrence by lesion size
Lesion diameter was the only subgroup characteristic that significantly modified the difference in recurrence between cold snare EMR and hot EMR.
Among lesions measuring 20 to 29 mm, residual or recurrent neoplasia occurred in 16.9% after cold snare EMR compared with 11.7% after hot EMR. The gap widened as lesion size increased.
For lesions measuring 30 to 39 mm, recurrence was 27.4% with cold snare EMR and 10.3% with hot EMR. Among lesions 40 mm or larger, recurrence reached 46.3% after cold snare EMR compared with 19.1% after hot EMR.
The authors attributed the declining efficacy of cold snare EMR with increasing lesion size to several possible factors. These included more fragmented resection, incomplete mucosal excision, reduced visibility from capillary bleeding, and the absence of cautery at the resection margins. They suggested limiting cold EMR to lesions no larger than 4 cm.
Recurrence by histology
By histology, the difference in recurrence was smallest for sessile serrated lesions: 15.0% after cold snare EMR vs. 8.1% after hot EMR, a difference that did not reach statistical significance. The authors attributed that gap mainly to one trial, in which recurrence was 34.8% vs. 5.0%. Rates were comparable in the other two trials that included these lesions.
The difference was larger for adenomas. Among adenomas with low-grade dysplasia, recurrence was 27.5% with cold snare EMR compared with 12.9% with hot EMR. Among those with high-grade dysplasia, recurrence was 40.8% vs. 18.0%.
“Current evidence excludes cold EMR as an adequate technique for advanced adenomas,” the authors wrote. They suggested that sessile serrated lesions and selected smaller adenomas without high-grade dysplasia may be suitable for cold snare EMR when combined with adjunctive measures.
In multivariable analysis, thermal margin ablation was the only factor associated with lower recurrence. According to the study, whether it and other adjunctive measures can reduce recurrence after cold snare EMR is under investigation.
Patient risk factors
The authors also pointed to patient factors that might aid decision-making. Clinically significant postendoscopic bleeding was three times as common among patients receiving anticoagulant or antiplatelet therapy as among those who were not, the authors reported. It was highest among patients with American Society of Anesthesiologists (ASA) class III status. ASA III was also the only positive predictor of major adverse events in multivariable analysis.
Cold snare EMR may be appropriate for selected patients receiving anticoagulant or antiplatelet therapy and for those with multimorbidity, the authors wrote, though further studies are needed.
Limitations and next steps
The authors noted several limitations. The four trials differed in inclusion criteria, randomization strategy, and technical approach, including the use of margin ablation and clip closure. Prophylactic clip closure was more common after hot EMR (47.8% vs. 13.6%), which the authors said could have influenced results in favor of hot EMR. One trial did not include sessile serrated lesions, and follow-up was limited to one endoscopic assessment within the first year.
According to the study, that short follow-up does not adequately reflect later colorectal cancer risk or the total number of procedures needed to eradicate residual or recurrent lesions. Recurrences are often technically more difficult to treat because of fibrosis and scarring, the authors noted, and may require repeat EMR or salvage endoscopic submucosal dissection.
“Recurrence should therefore be regarded not merely as a surrogate endpoint,” the authors wrote. They said repeated treatment can add procedural burden, healthcare use, patient anxiety, and cumulative complication risk.
“The key question remains whether the improved safety of cold snare outweighs its lower efficacy,” the authors wrote. They called for studies of whether technical modifications or adjunctive measures could make cold snare EMR more effective.
Ingo Steinbrück, MD, reported lecture fees and travel grants from Olympus and Falk Pharma. Several coauthors reported consulting relationships, research support, honoraria, travel support and/or ownership interests involving Olympus, Boston Scientific, Medtronic, Fujifilm, ERBE, Pentax and other companies. The study was funded by the Gastroenterology Foundation, Küsnacht, Switzerland.