Endoscopic characteristics influencing postpolypectomy bleeding in 1147 consecutive pedunculated colonic polyps: a multicenter retrospective study.

BACKGROUND AND AIMS: Postpolypectomy bleeding is the most common adverse event with pedunculated polyps. We clarified the endoscopic characteristics influencing postpolypectomy bleeding for pedunculated colonic polyps.

METHODS: We reviewed clinical data for 1147 pedunculated colonic polyps removed by polypectomy in 5 Japanese institutions. Pedunculated polyps were defined as polyps with a stalk length ≥5 mm. Analyzed clinical data were age, sex, polyp location/size, stalk length/width, prophylactic clipping or endoloop before polypectomy, injecting the stalk, closing the polypectomy site, antithrombotic agent use, and endoscopist experience. Postpolypectomy bleeding was classified as immediate bleeding or delayed bleeding.

RESULTS: Immediate and delayed bleeding was observed in 8.5% (97/1147) and 2% (23/1147) of polypectomies, respectively. Comparing immediate bleeding with nonbleeding, multivariate analysis showed that stalk width ≥6 mm (odds ratio [OR], 1.9; 95% confidence interval [CI], 1.1-3.4) was a significant risk factor for immediate bleeding. For polyp size ≥15 mm, prophylactic endoloop use (OR, .17; 95% CI, .04-.72) was a significant inhibiting factor. Comparing delayed bleeding with nonbleeding, multivariate analysis showed that prophylactic clipping before polypectomy (OR, 4.2; 95% CI, 1.3-13) and injecting the stalk (OR, 4.0; 95% CI, 1.4-12) were significant risk factors for delayed bleeding.

CONCLUSIONS: The increased risk for delayed bleeding with injecting the stalk and prophylactic clipping before polypectomy suggests that simple resection with coagulation mode is a suitable strategy in endoscopic resection of pedunculated polyps. Moreover, prophylactic endoloop use was highly likely to inhibit immediate bleeding with polyp size ≥15 mm.