SAN DIEGO—For resection of large polyps, cold endoscopic mucosal resection is known to be safer than hot EMR but carries a higher risk (≥25%) for recurrence at first surveillance colonoscopy. The risk for recurrence at the second surveillance exam and efficacy of resecting these lesions has not been established and was the aim of a multicenter study presented at DDW 2025 by Heiko Pohl, MD, of Dartmouth Hitchcock Medical Center, in Lebanon, N.H.

Dr. Pohl and his co-investigators conducted a follow-up analysis of 660 patients treated at 15 centers for nonpedunculated colorectal polyps measuring at least 20 mm, in an extension of a primary analysis reported at DDW last year (abstract Sp1251). The key finding in the 2024 report was a significantly higher rate of recurrence at six months in the cold snare group: 28% versus 14% in the intention-to-treat analysis (P<0.001) and 29% versus 12%, respectively, in the per-protocol analysis (P<0.001).

In the new analysis, the investigators found that recurrence rates at the second surveillance exam (SC2) were not significantly different between cold and hot EMR, and regardless of the initial treatment, these recurrences could be adequately managed (abstract Sa1691).

The investigators performed an interim analysis of all patients who completed the SC2. The primary end point was recurrence/residual rate of neoplastic polyps at the SC2, and secondary outcomes included proportion of polyps with no recurrence at initial surveillance but recurrence at the SC2. Treatment of recurrence was at the discretion of the endoscopist.

The SC2 was completed by 148 patients (48%) in the cold EMR group and 124 (41%) in the hot EMR group. Among these patients, recurrences were significantly more common at initial surveillance after cold EMR (36.3%) than hot EMR (16.9%; P<0.001). On the other hand, at the SC2, recurrences were not significantly different for cold EMR (12.7%) and hot EMR (8.9%; P=0.304) in the intention-to-treat analysis or in the per-protocol analysis (12.8% vs. 5.4%, respectively; P=0.08).

A ‘Surprising’ Finding
Among patients with recurrence at initial surveillance, complete removal of the lesion with no recurrence at the SC2 was more often achieved in the cold EMR group (80.7%) than in the hot EMR group (57.1%; P=0.035). The proportion of polyps with no recurrence at initial surveillance but a recurrence detected at the SC2 was 5.7% with cold snare and 1.6% with hot snare (P=0.127).

“It’s concerning that almost half of those who had a recurrence at the second surveillance after cold EMR had not been found to have a recurrence at first surveillance. … We missed those,” he said. He commented that this “surprisingly high” new SC2 recurrence rate after a supposedly “normal” initial surveillance finding after cold EMR suggests initial surveillance biopsies “should be taken, even in the absence of any visible recurrence.”

The median polyp size for recurrences was similar for both approaches, around 9 mm, and all recurrent lesions could be removed. Histology of recurrent polyps also was similar, and none of the lesions were cancerous.

The investigators concluded that the long-term efficacy of cold EMR of large nonpedunculated polyps appears comparable to hot EMR at the SC2, although incomplete SC2 follow-up data may limit the interpretation of the findings. Management of recurrence seems equally effective for cold and hot EMR.

—Caroline Helwick


Dr. Pohl reported no relevant financial disclosures.