Clinical effectiveness of open thrombectomy for thrombosed autogenous arteriovenous fistulas and grafts.

Objective: Arteriovenous (AV) fistulas are the preferred hemodialysis access for patients with end-stage renal disease, although multiple interventions are typically needed to maintain patency. When AV fistulas thrombose, however, there is debate as to whether open thrombectomy should be attempted, particularly for salvage of upper arm fistulas. This study was designed to evaluate outcomes after open thrombectomy of upper arm and forearm AV fistulas compared with AV grafts.

Methods: We identified all patients who underwent an open thrombectomy procedure for a thrombosed AV fistula or graft at a single academic medical center between January 2006 and March 2017. The specific type of AV fistula or graft was evaluated, as were the patients' demographics, comorbidities, medications, adjunctive procedures during thrombectomy, and secondary interventions. The primary outcome measures, postintervention primary patency and postintervention secondary patency, were analyzed using Kaplan-Meier curves and Cox regression models for risk adjustment.

Results: During the study period, 209 open thrombectomy procedures were performed in 139 patients; 73 (35%) were undertaken in AV fistulas and 136 (65%) in grafts. Patients with upper arm fistulas (n = 52; 54% brachiocephalic, 46% brachiobasilic) and forearm fistulas (n = 16) were more likely to be male but less likely to have cerebrovascular disease or ischemic heart disease and to be receiving anticoagulation therapy compared with graft patients. After thrombectomy, the majority of patients underwent dialysis successfully (70% upper arm fistulas, 56% forearm fistulas, 63% grafts; P > .05), and 1-year survival rates were similar in all three cohorts. Postintervention primary patency at 1 year was significantly higher for AV fistulas vs grafts (33% for upper arm fistulas and 25% for forearm fistulas vs 9% for grafts; P < .05), which was confirmed in multivariate analysis, where upper arm AV fistulas had a 46% lower risk of recurrent thrombosis or secondary intervention (hazard ratio, 0.56; 95% confidence interval, 0.35-0.85; P < .05). Postintervention secondary patency at 1 year was similar between AV fistulas and grafts (44% for upper arm fistulas vs 43% for forearm fistulas vs 31% for grafts; P = .16), but in multivariate analysis, upper arm fistulas were significantly less likely to fail (hazard ratio, 0.63; 95% confidence interval, 0.40-1.00; P = .05).

Conclusions: Our data suggest that AV fistula thrombectomy is successful in up to 70% of cases, with significantly improved risk-adjusted 1-year primary and secondary patency rates for upper arm fistulas compared with grafts. Whereas the risk of access failure is high after thrombectomy, efforts to salvage upper arm AV fistulas are effective in most patients and should be undertaken when feasible.