Technical factors affecting autogenous vein graft failure: Observations from a large multicenter trial

Abstract

The influence of operator-dependent variables on the outcomes of lower extremity bypass (LEB) surgery have primarily been reported in single-institution, retrospective studies. We utilized data from a prospective, multicenter trial to identify technical variables that were significantly associated with early and midterm results of autogenous LEB for limb salvage. The PREVENT III trial database includes 1404 North American patients with critical limb ischemia (CLI) who underwent LEB using excised autogenous vein. The study protocol excluded claudicants and in situ reconstructions. Technical factors analyzed included vein diameter, conduit type, graft length, vein orientation, location of proximal and distal anastomoses, and performance of completion imaging. Univariate analysis was used to determine the effect of these factors on 30 day and 1-year outcomes. Multivariate Cox regression models evaluated the influence of these factors while adjusting for age, sex, race, tobacco, diabetes, dialysis-dependency, previous index limb bypass, and study drug (edifoligide) administration. The primary outcomes were primary patency (PP), primary assisted patency (PAP), and secondary patency (SP) assessed by Kaplan-Meier method. Univariate analysis revealed that vein diameter <3.5 mm and composite graft type were significantly associated with early (30 day) graft failure. At 1 year, multivariate analysis revealed that patency rates were negatively associated with diameter <3.5 mm (PP, PAP, SP), non-great saphenous vein (GSV) type (PP, SP), and graft lengths >50 cm (PP only). Limb salvage and survival at 1 year were not significantly impacted by technical variables. Employing a prespecified trial definition of high-risk conduits (diameter <3mm or nonsingle segment GSV; 24% of entire cohort) revealed that use of such conduits was associated with a 2.1-fold increased risk of 30 day graft failure (P < .05), as well as reduced PP, PAP, and SP at 1 year. Use of a high-risk conduit was also associated with an increased index length of stay (mean 9.37 vs 8.71 days, P = .03) and a greater number of reinterventions (mean 0.67 vs 0.42, P < .0001) over the ensuing year. In this large, multicenter cohort of patients undergoing LEB for CLI, vein diameter and conduit type were the dominant technical determinants of early and late graft failure. High-risk conduits and longer grafts may benefit from aggressive postoperative graft surveillance.