Safety and efficacy of hot avulsion as an adjunct to EMR (with videos)

Vinod Kumar, Heather Broadley, Douglas K. Rex

Research output: Contribution to journalArticlepeer-review

8 Scopus citations


Background and Aims: Excision of all visible neoplastic tissue is the goal of EMR of colorectal laterally spreading tumors. Flat and fibrotic tissue can resist snaring. Ablation of visible polyps is associated with high recurrence rates. Avulsion is a technique to continue resection when snaring fails. Methods: We retrospectively analyzed colonic EMRs of 564 consecutive referred polyps between 2015 and 2017. Hot avulsion was used when snaring was unsuccessful. Polyps treated with and without avulsion were compared. Results: Hot avulsion was used in 20.9% (n = 112) of all resected lesions. The recurrence rates on follow-up colonoscopy were 17.52% in the avulsion group versus 16.02% in the non-avulsion group (P =.76). Hot avulsion was associated with a trend toward higher rates of delayed hemorrhage (5.35% vs 2.58%; P =.15) and post-coagulation syndrome (1.8% vs 0.47%; P =.15), but polyps treated with any avulsion were larger than those in which no avulsion was used (P <.001). There were an insufficient number of adverse events to perform a multivariable analysis to test the effects of avulsion, size, and location on the risk of overall adverse events. Conclusions: Unlike previous reports of using argon plasma coagulation to treat visible polyps during EMR, hot avulsion of visible/fibrotic neoplasia was associated with similar EMR efficacy compared with cases that did not require hot avulsion. The safety profile of hot avulsion appears acceptable.

Original languageEnglish (US)
Pages (from-to)999-1004
Number of pages6
JournalGastrointestinal endoscopy
Issue number5
StatePublished - May 2019

ASJC Scopus subject areas

  • Radiology Nuclear Medicine and imaging
  • Gastroenterology

Fingerprint Dive into the research topics of 'Safety and efficacy of hot avulsion as an adjunct to EMR (with videos)'. Together they form a unique fingerprint.

Cite this