Risk of Postpolypectomy Bleeding With Uninterrupted Clopidogrel Therapy in an Industry-Independent, Double-Blind, Randomized Trial.

BACKGROUND & AIMS:Guidelines recommend withholding clopidogrel 7 days before polypectomy to decrease bleeding risk, but these were written based on limited evidence. We investigated whether uninterrupted clopidogrel therapy increases the risk of delayed postpolypectomy bleeding in patients undergoing colonoscopy. METHODS:We identified patients receiving clopidogrel for cardiovascular disease undergoing elective colonoscopies in Hong Kong from February 28, 2012 through April 11, 2018. Eligible patients were instructed to stop taking clopidogrel 7 days before colonoscopy. Then, they were randomly assigned to groups given clopidogrel (75 mg) or placebo daily until the morning of colonoscopy. All patients resumed their usual prescriptions of clopidogrel after colonoscopy. The primary end point was delayed postpolypectomy bleeding that required hospitalization or intervention up to 30 days after colonoscopy. Secondary end points were immediate postpolypectomy bleeding and serious cardio-thrombotic events for as long as 6 months after colonoscopy, according to Antithrombotic Trialists' criteria. All events were adjudicated by an independent masked committee. RESULTS:In total, 387 patients underwent colonoscopy and 216 required polypectomies (106 patients in the clopidogrel group and 110 patients in the placebo group). The cumulative incidence of delayed postpolypectomy bleeding was 3.8% (95% confidence interval 1.4-9.7) in the clopidogrel group and 3.6% (95% confidence interval 1.4-9.4) in the placebo group (P = .945 by log-rank test). There were no significant differences in immediate postpolypectomy bleeding (8.5% vs 5.5%; P = .380) and cardio-thrombotic events (1.5% vs 2%; P = .713). CONCLUSIONS:In a randomized controlled trial of clopidogrel users undergoing colonoscopy, a slightly larger proportion of patients continuing clopidogrel developed delayed and immediate postpolypectomy bleeding, although this difference was not statistically significant. ClinicalTrials.gov, number NCT01806090.