A systematic description of the post-EMR defect to identify risk factors for clinically significant post-EMR bleeding in the colon.

ABSTRACT

BACKGROUND AND AIMS: Clinically significant post-EMR bleeding (CSPEB) is the most-frequent serious adverse event after EMR of large laterally spreading colonic lesions (LSLs). There is no proven prophylactic therapy, and it remains a significant drawback of EMR. We aimed to systematically describe and evaluate the features of the post-EMR mucosal defect (PED) and their relationship to CSPEB. METHODS: A prospective study of LSLs referred for EMR at a tertiary center was performed. PEDs without visible features were recorded as bland blue. Nonbland blue (NBB) PED features included size, number, and herniation of submucosal vessels and presence of submucosal hemorrhage, fibrosis, fat, and exposed muscle. NBB PEDs were analyzed for association with CSPEB, defined as bleeding occurring after completion of the procedure necessitating readmission or reintervention. RESULTS: From April 2012 to May 2017, 501 lesions in 501 patients were eligible for analysis. The frequency of CSPEB was 30 of 501 (6.0%). More than or equal to 3 visible vessels was a significant predictor of CSPEB (P = .016). None of the following showed a significant correlation with CSPEB presence of visible vessels, their diameter, herniation, or other nonvascular PED features. Submucosal vessels were more common in the left-sided colon segment (88.6% vs 78.3%, P = .004) and were significantly larger (20.8% vs 12.1% ≥1 mm, P = .037), more numerous (median 4 vessels [interquartile range, 2-7] vs 2 vessels [interquartile range, 1-4], P < .001), and more often herniated (32% vs 22.2%, P = .022). CONCLUSIONS: More than or equal to 3 visible vessels within the PED may be predictive for CSPEB and may define a target group for real-time prophylactic intervention. No other endoscopically visible features of the PEDs were predictive of CSPEB. (Clinical trial registration number NCT03117400.).