Systematic Review with Meta-analysis: Risk of Gastrointestinal Hemorrhage Post-polypectomy in Patients on Blood-Thinning Medications

Abstract

Introduction: Current society guidelines recommend performance of endoscopic procedures without interruption of antiplatelet therapy. For patients undergoing high-risk endoscopic procedures such as colonoscopy with polypectomy, temporary cessation of thienopyridine and anticoagulant therapy is recommended depending on the relative risk of thromboembolism and bleeding. The data regarding periprocedural management of these medications is sparse. Methods: We searched Pubmed, Scopus, Web of Science, Biosis, and Proceedings First from 1970 to 2014. PPB was defined as overt hemorrhage or >2g/dL drop in hemoglobin. Risk of PPB was calculated as odds ratios with 95% confidence intervals. Results: Of 1490 articles and abstracts identified, we included 3 papers and 1 abstract with patients on antiplatelet therapy, 2 abstract with patients taking clopidogrel therapy, 2 papers with patients on dual antiplatelet therapy, and 1 paper with patients on warfarin therapy. The rate of PPB was 4% (59/1489) in aspirin/NSAID users compared to 3% (104/3650, p=0.04) for non-users. The risk of immediate PPB was not increased (OR 1.1, 95% CI 0.7-1.9, p=0.7) and the risk of delayed PPB was increased with borderline significance (OR=2.0, 95% CI 1.0-4.0, p=0.05) but became non-significant when the study with small numbers of events was eliminated. Based on two abstracts reporting delayed PPB for patients taking clopidogrel therapy alone, clopidogrel use was associated with an increased risk of delayed PPB (OR 9.7,95% CI 3.1-30.8, p=0, I2=0). Two papers reported rates of PPB in patients on clopidogrel, 40% of whom were on concurrent ASA/NSAID therapy; there was no significant difference in PPB between clopidogrel users and nonusers (OR 1.56, 95% CI 0.76-3.19, p=0.2, I2=0%, p=0.6). There was, a statistically insignificant but borderline increased risk of delayed PPB in patients on dual antiplatelet therapy compared to aspirin/NSAID alone (OR = 2.3, 95% CI 0.9-6.1, p=0.08, I2=0). There were no trials comparing continued warfarin therapy to control, but the overall rate of bleeding reported in a single study in patients on continued warfarin at the time of polypectomy was higher than in the general population. (Immediate PPB rate of 23% for conventional polypectomy and 6% for cold snare). Conclusion: Usage of antiplatelet therapy does not increase risk of PPB. Clopidogrel and warfarin should be held when appropriate in the periprocedural period to prevent PPB. Further study in patients on combination therapy is warranted.Table 1: Post-Polypectomy Bleeding (PPB) In Patients on Aspirin or NSAID TherapyTable 2: Post-Polypectomy Bleeding (PPB) In Patients on Clopidogrel TherapyFigure 1Figure 2