Prophylactic clip closure after right-colon endoscopic mucosal resection in a matched cohort of 470 patients exhibits subsite heterogeneity

Authors:

Abstract
Clinically significant post-endoscopic mucosal resection (EMR) bleeding (CSPEB) is the most common adverse event after EMR of large nonpedunculated colorectal polyps (LNPCPs). Randomized trials of prophylactic through-the-scope (TTS) clip closure have produced conflicting results, and it remains unclear whether the effect is uniform across right-colon subsites and whether complete apposition is required. We examined real-world outcomes of a universal right-colon clipping policy, with attention given to subsite heterogeneity and closure completeness. Consecutive patients who underwent EMR of right-sided LNPCPs ≥20 mm (January 2020–October 2025) with prophylactic TTS clip closure were identified from the prospectively maintained ACE (Australian Colonic Endoscopic resection) registry, and matched 1:1 by propensity score to a no-clip comparator from the same registry (pre-2020). The primary end point was CSPEB within 14 days; subsite heterogeneity and closure completeness were exploratory analyses. Propensity score matching yielded 235 matched pairs (n = 470), who were well balanced on all covariates. CSPEB occurred in 8/235 clipped (3.4%) versus 26/235 unclipped patients (11.1%; relative risk 0.31, 95%CI 0.14–0.66; number needed to treat 13; P = 0.002). Exploratory subsite analyses suggested benefit concentrated at the hepatic flexure (0% vs. 17.2%; P = 0.02) and ascending colon (2.0% vs. 10.1%; P = 0.03), with a nonsignificant difference at the cecum (8.2% vs. 13.7%; P = 0.43). Within the clipped cohort, CSPEB rates were similar with complete (5/143; 3.5%) and partial (3/92; 3.3%) closure (P > 0.99). In this single-center analysis, TTS clip closure was associated with lower CSPEB after right-colon EMR, especially at the hepatic flexure and ascending colon; partial apposition appeared as effective as complete closure. These hypothesis-generating observations warrant prospective evaluation.

This publication has 17 references indexed in Scilit: