Use of apixaban and warfarin in patients undergoing invasive procedures: insights from ARISTOTLE

Abstract

Objective: The purpose of this study is to investigate rates of invasive procedures among patients in the ARISTOTLE (Apixaban for Reduction in Stroke and Other Thromboembolic Events in Atrial Fibrillation) trial as well as the risk of stroke and major bleeding following invasive procedures. The risk of stroke or systemic embolism (SSE) associated with stopping warfarin for invasive procedures is thought to be low. Little is known about apixaban use in patients undergoing invasive procedures. Methods: Using data from 18,201 patients in ARISTOTLE (median followup 1.8 years), this study described the most common invasive procedures, use of bridging therapy, and risk of stroke and major bleeding during the 30 days following these procedures. Procedures were classified as major if they required general anesthesia or were considered to pose a significant postoperative bleeding risk. Investigators classified procedures as emergent or nonemergent. Results: There were 11,417 invasive procedures performed in 6,162 patients: 477 (4.2%) were major and 10,940 (95.8%) nonmajor; 322 (2.8%) were emergent and 11,095 (97.2%) nonemergent. The most common procedures were dental extraction/oral surgery, colonoscopy, upper endoscopy, and ophthalmic surgery. In 4,082 procedures (35.8%), study drug was not stopped. Median time of study drug stop was 4 days before the procedure for both apixaban- and warfarin- treated patients. A second bridging anticoagulant, most commonly lowmolecular- weight heparin, was used in 1,335 procedures (11.7%). Of 5,660 events in the apixaban group, 0.43% were associated with SSE and 1.55% with major bleeding. Of 5,757 events in the warfarin group, 0.56% were associated with SSE and 1.80% with major bleeding. Conclusion: Invasive procedures are common in patients with atrial fibril- lation. Most procedures are nonmajor and nonemergent, and anticoagulation therapy is likely to be stopped periprocedure. Overall and among emergent procedures, rates of clinical events in the first 30 days post-procedure were low and comparable between patients receiving warfarin and apixaban.