Predictors of Early Graft Thrombosis During Index Admission for Lower Extremity Arterial Bypass

Abstract

Early graft thrombosis following lower extremity bypass (LEB) has long served as a quality indicator of surgical management. A granular understanding of the risk factors for graft thrombosis might help mitigate the potentially avoidable sequelae of this complication. Our study utilized Vascular Quality Initiative (VQI) data to identify predictor variables associated with graft thrombosis during index admission following LEB. We queried the VQI database for patients receiving unilateral infrainguinal LEB between 2003 and 2022. Patients undergoing concomitant PVI or suprainguinal bypass were excluded from our analysis. We formed six procedural groups: femoral-above knee popliteal bypass, femoral-below knee popliteal bypass, femoral-tibial bypass, femoral-ankle bypass, popliteal-tibial bypass, and popliteal-ankle bypass. Among these six groups, we utilized multivariable logistic regression modeling to identify technical and perioperative characteristics that predicted early graft thrombosis. Of the 54,504 LEB procedures that met inclusion criteria, 11,906 (21.8%) underwent femoral-above knee popliteal, 17,541 (32.2%) femoral-below knee popliteal, 19,224 (35.3%) femoral-tibial, 1414 (2.6%) femoral-ankle, 2661 (4.9%) popliteal-tibial, and 1758 (3.2%) popliteal-ankle bypass procedures. At the time of discharge, 1012 (1.9%) required intervention for a thrombotic intervention, and 814 (1.5%) were occluded. After adjusting for patient characteristics and comorbidities, we identified the following independent predictors for early graft thrombosis: female gender, distal target, symptomatic indication (rest pain and tissue loss), urgent/emergent presentation, and anticoagulation therapy, specifically warfarin and rivaroxaban. Conversely, the use of preoperative statin, aspirin and clopidogrel was protective against early graft thrombosis (Table). There were 9.5% of patients not receiving one or several of the protective medications identified (25.9% not receiving statins, 69.1% not receiving clopidogrel, and 27.6% not receiving aspirin). With regard to type of conduit used for LEB, native vein conduits demonstrated the lowest odds of graft thrombosis (Table, Figure). In this large national contemporary study of LEB, we have identified several factors including distal target, urgent/emergent presentation, preoperative warfarin or rivaroxaban therapy and use of nonautologous conduits as independent predictors of graft thrombosis. Conversely, preoperative statin, antiplatelet therapy and the use of native vein conduits were protective against graft thrombosis. Despite their beneficial effect, we identified that 9.5% of the LEB population are not receiving statin and/or antiplatelet therapies, which serves as a potential quality improvement initiative. Our findings suggest that utilization of native vein when possible, and promoting aggressive preoperative medical optimization including statin, aspirin and P2Y12 inhibitors may help avoid early graft occlusion in patients requiring lower extremity revascularization.TableMultivariable logistic regression model indicating technical and perioperative characteristics associated with early graft thrombosis after lower extremity arterial bypass surgeryPerioperative characteristicsOdds ratio (95% confidence interval)P valueAge0.99 (0.98-0.99)<.001Female sex1.32 (1.20-1.45)<.001White race1.10 (1.00-1.22).059Diabetes0.83 (0.75-0.92)<.001Dialysis1.02 (0.93-1.13).638Congestive heart failure1.03 (0.89-1.18).731Prior coronary artery bypass grafting0.87 (0.79-0.96).008Chronic obstructive pulmonary disease0.94 (0.83-1.05).266Smoking0.82 (0.73-0.93).002Aspirin0.73 (0.66-0.81)<.001Statin0.85 (0.77-0.95).003Baseline hemoglobin0.92 (0.90-0.95)<.001BMI1.00 (0.99-1.01).835Bypass type Femoral to above knee popliteal1 (reference)- Femoral to below knee popliteal1.62 (1.30-2.03)<.001 Femoral to tibial vessel3.18 (2.58-3.90)<.001 Femoral to ankle vessel4.41 (3.28-5.93)<.001 Popliteal to tibial vessel2.58 (1.98-3.37)<.001 Popliteal to ankle vessel3.15 (2.32-4.28)<.001Urgency Elective1 (reference)- Urgent1.74 (1.54-1.97)<.001 Emergent3.45 (2.75-4.34)<.001Conduit Native vein1 (reference)- Dacron1.09 (0.56-2.11).810 Polytetrafluoroethylene1.36 (1.20-1.55)<.001 Nonautologous biologic2.41 (1.95-2.97)<.001Indication Asymptomatic1 (reference)- Claudication1.02 (0.53-1.97).953 Rest Pain2.44 (1.28-4.66).007 Tissue loss2.18 (1.16-4.11).016 Acute ischemia3.39 (1.77-6.49)<.001 Not treated2.53 (1.34-4.77).004Per hour procedure time1.00 (1.00-1.00)<.001Preoperative P2Y12 antagonist therapy None1 (reference)- Clopidogrel0.81 (0.73-0.91)<.001 Prasugrel1.14 (0.58-2.24).697 Ticagrelor0.93 (0.49-1.75).818 Other0.81 (0.30-2.16).668 No, for medical reasons0.69 (0.50-0.96).027 Noncompliant1.35 (0.75-2.45).314Preoperative anticoagulation therapy None1 (reference)- Warfarin1.38 (1.16-1.65)<.001 Dabigatran1.06 (0.49-2.31).874 Rivaroxaban1.42 (1.15-1.75).001 Other1.54 (1.29-1.85)<.001 No, for medical reasons1.14 (0.74-1.74).548 Noncompliant2.35 (1.27-4.35).007Presented are odds ratios (95% confidence interval) from a multivariable model adjusting for age, sex, race, diabetes, dialysis, congestive heart failure, prior coronary artery bypass grafting, chronic obstructive pulmonary disease, smoking, use of aspirin, statin, anticoagulation, body mass index, urgency, indication, and total procedural time. Boldface entries indicate statistical significance. Open table in a new tab