Bleeding and Thrombotic Outcomes Associated With Postoperative use of Direct Oral Anticoagulants After Open Peripheral Artery Bypass Procedures

Abstract

The widespread use of the direct oral anticoagulants (DOACs) for the treatment of atrial fibrillation and venous thromboembolism has inevitably resulted in peripheral bypass patients receiving therapeutic anticoagulation (AC) with DOACs postoperatively. No recommendations exist to guide whether DOACs can be safely used as an alternative to vitamin K antagonists (VKAs) for prophylaxis of high-risk bypass grafts. This study was undertaken to evaluate the bleeding and thrombotic outcomes of patients undergoing open peripheral bypass treated with DOACs compared with VKAs. We examined data from a large, statewide, multicenter quality improvement vascular surgery registry of all patients undergoing elective, urgent, or emergent peripheral bypass operations between January 2012 and December 2017. Outcomes included length of stay, transfusion, readmission, return to operating room, 1-year primary patency, and major amputation within 1 year. Statistical significance between treatment groups was determined using pairwise Wilcoxon rank-sum tests for continuous variables and pairwise χ2 or Fisher exact tests for categorical variables. Multivariate logistic regression was performed for bleeding complications and graft occlusion. Post hoc pairwise comparisons for treatment groups were calculated using Tukey’s P value adjustment. In total, 9682 patients met our inclusion criteria. Postoperatively 7605 patients received no AC, 1456 received a VKA, and 621 a DOAC. Patients receiving AC were older, had a higher body mass index, and were more likely to have preoperative anemia, hyperlipidemia, hypertension, congestive heart failure, coronary artery disease, and atrial fibrillation (all P < .05). Compared with patients treated with VKA, patients treated with DOACs were less likely to have valvular disease and chronic renal insufficiency and more likely to have atrial fibrillation (all P < .05). The length of stay (LOS) varied significantly among treatment groups, with the shortest length of stay in patients receiving no AC (median, 5 days; interquartile range [IQR], 3-9; P < .001), followed by DOACs (median, 6 days; IQR, 3-11; P < .001) and the longest LOS observed in patients treated with VKAs (median, 8 days; IQR, 5-13; P < .001). Perioperative transfusion was significantly higher in patients treated with VKAs (33.9%), compared with DOAC (28.7%; P = .023) and no anticoagulant (23.8%; P < .0001). Primary patency at 1 year did not differ among groups. Return to the operating room for graft occlusion (3.5% VKA, 3.4% DOAC vs 1% no AC; P < .001) and amputation (19.2% VKA, 17.1% DOAC, vs 11.9% no AC P < .001) was higher in patients receiving AC compared with those not receiving AC, regardless of type (VKA compared with DOAC). On multivariate analysis, procedure status (emergent: odds ratio, 2.31 [95% CI, 1.39-3.84; P = .001], urgent: odds ratio, 1.60 [95% CI, 1.04-2.45; P = .032]) were predictors of graft occlusion. No AC patients were less likely to experience graft occlusion (odds ratio, 0.40 [95% CI 0.27-0.59; P < .001]). Among a mixed group of patients undergoing lower extremity bypass, those treated with a DOAC postoperatively had a shorter LOS and were less likely to receive a transfusion in the perioperative setting without compromising graft patency or affecting amputation rate when compared with those treated with VKA. These data suggest the need for a prospective trial evaluating DOAC safety and efficacy in patients requiring AC for high-risk bypass grafts.