Comparative effectiveness of anticoagulation on midterm infrainguinal bypass graft patency

Abstract

Objective: Therapeutic anticoagulation (AC) is used clinically for prolongation of infrainguinal bypass patency, but evidence for the efficacy of this practice is conflicting. The objective of our study was to determine the association of AC with bypass graft primary patency.

Methods: Clinical and comorbid data of patients undergoing infrainguinal bypass grafts to a below-knee target with at least 1 year of follow-up performed from 2003 to 2015 were obtained from the Society for Vascular Surgery Vascular Quality Initiative. Inverse propensity of treatment-weighted Cox regression was used to assess the effect of AC on patency in the total cohort while adjusting for clinical, operative, and comorbid differences between treatment groups. Subgroup analyses of distal targets and conduit type were performed. Perioperative complications were analyzed using propensity-weighted logistic regression.

Results: We identified 7612 bypass grafts with intact 1-year follow-up information from 2003 to 2015. The mean age was 67.5 ± 11.2 years; 30.5% (n = 2320) were female, and 28.6% (n = 2165) were discharged on therapeutic AC. The anticoagulated group had a higher rate of tibial, ankle, and pedal targets (52.1% [n = 1127] vs 47.6% [n = 2269]; P < .001), had a greater use of non-single-segment vein conduits (44.3% [n = 951] vs 26.5% [n = 1426]; P < .001), and was more likely to have had a previous ipsilateral bypass (27.2% [n = 589] vs 14.7% [n = 794]; P < .001) or stent (25.4% [n = 550] vs 20.9% [n = 1130]; P < .001). Estimated unadjusted primary patency was 70.8% ± 0.6% at 1 year and lower for anticoagulated bypasses (66.9% ± 1.2% vs 72.4% ± 0.7%; P < .001). Propensity-weighted analysis showed no significant association of AC with primary patency in the overall cohort (hazard ratio [HR], 0.98; 95% confidence interval [CI], 0.86-1.11; P = .8) but demonstrated a trend toward improvement of primary patency in those with a non-single-segment vein conduit to a below-knee popliteal target (HR, 0.85; 95% CI, 0.80-1.02; P = .09). AC was associated with significantly improved secondary patency in those with prosthetic bypass grafts (HR, 0.77; 95% CI, 0.62-0.96; P = .02) or prosthetic bypasses to an infrapopliteal target (HR, 0.72; 95% CI, 0.54-0.97; P = .02). Odds of postoperative wound complications were significantly higher in those receiving AC (odds ratio, 1.33; 95% CI, 1.11-1.61; P = .002).

Conclusions: This study does not demonstrate a significant impact of therapeutic AC on primary patency for infrainguinal bypass grafts. Treatment with AC may benefit secondary patency in those with a prosthetic bypass, especially to an infrapopliteal target, but at an increased risk of postoperative wound complications.

Figure 1. Adjusted Primary Patency Outcomes

Adjusted primary patency outcomes by primary, secondary, and exploratory analyses. HR: hazard ratio. NSV-BKP: non-single-segment vein to a below-knee popliteal target subgroup.

Figure 2. Adjusted Secondary Patency Outcomes

Adjusted secondary patency outcomes. HR: hazard ratio.