2007 Abstracts: Surgical and Endovascular Revision of Infrainginal Vein Bypass Grafts: Analysis of Midterm Outcomes from a Multi-Center Prospective Trial

Background: Data supporting the utility of percutaneous treatment to maintain vein graft patency has been limited to a collection of single-institution, retrospective analyses. Using the prospective, multi-institutional PREVENT III database, we seek to define the outcomes for percutaneous versus surgical vein bypass graft revision and define the predictors for success or failure of these interventions.
Methods: A nested cohort study of 1404 patients in the PREVENT III trial, who underwent infrainguinal vein bypass grafting for critical limb ischemia, was performed to identify those patients who underwent either open surgical (SURG) or endovascular (ENDO) graft revision. All patients in PREVENT III were followed for one year from initial bypass surgery. Freedom from graft re-intervention, occlusion, amputation, or death was modeled as endpoints from the time of initial SURG or ENDO revision.
Results: 179 SURG and 134 ENDO primary revisions were performed, with a mean follow-up after revision of 229 and 212 days, respectively. Conduit and lesion characteristics were similar for each group (Table 1). However, the interval time following graft placement significantly influenced the method of repair (Table 2), with SURG revisions most prevalent in the early post-operative period.
Univariate analysis (Figure 1) demonstrated a significant survival benefit for SURG as compared to ENDO revisions (P=0.03). Further, subgroup analysis revealed this benefit to be most profound in the minority of thrombosed grafts undergoing salvage (P=0.006). For revisions performed to treat graft stenosis, early outcomes were similar, with a trend favoring SURG in the 6 to 12 month timeframe.
Disadvantaged conduits (<3mm diameter, arm or spliced veins) were at higher risk for additional graft-related events (P=0.02). Independent multivariate testing of the demographic variables demonstrated an increased incidence of a subsequent re-intervention when the first SURG revision was in the graft body (OR=2.2, CI 95%=1.2-4.1). No dominant factors were identified in the ENDO group.
Although 80% (SURG) and 64% (ENDO) of the revised grafts required no further intervention, ENDO revisions required significantly more re-interventions to maintain patency (Figure 2). Mean hospital lengths of stay (SURG=2.1 days, ENDO=1.7 days) and quality of life at completion of the study (VascuQOL: SURG=4.72, ENDO=4.76) were similar between the groups.
Conclusions: Open surgical or percutaneous repair of patent vein grafts yield similar intermediate outcomes, however at the expense of an increased number of repeat revisions in the endovascular group. In contrast, thrombosed grafts undergoing revision demonstrated marked differences between these repair options, with percutaneous interventions exhibiting substantially inferior results.