SS31. Comparable Patency of Open and Endovascular Treatment of Venous Anastomotic Lesions in Hemodialysis Grafts

Abstract

Arteriovenous graft (AVG) failures are typically associated with venous anastomotic stenosis. Most studies comparing outcomes of open vs endovascular thrombectomy have shown equipoise, but the inherent bias of these studies has been the fact that the venous anastomosis was managed with endovascular means even in the open surgical thrombectomy groups. There is little evidence to support balloon venoplasty or stenting as more durable compared with patch venoplasty or proximal bypass extension of the AVG. Consecutive patients with an AVG who underwent the first thrombectomy and a venous anastomosis intervention between January 2014 and July 2018 were included in our study. Patients’ demographics, previous access history, central vein patency, AVG anatomy, type of intervention, and follow-up data were recorded. Kaplan-Meier analysis was used to analyze time from thrombectomy to first reintervention (postintervention primary patency) and time to abandonment (postintervention secondary patency). Cox regression analysis was performed to evaluate predictors of failure. Our study included 105 patients (51 women; 64.2 ± 14.9 years old) with 42 forearm, 51 upper arm, and 12 lower extremity AVGs. There were 37 open revisions (26 patches, 11 proximal jump bypasses) and 68 endovascular interventions at the venous anastomosis (14 stent grafts, 5 cutting balloons); 6 of these were performed entirely percutaneously. Technique selection was based on the physician’s preference. There were no differences between open and endovascular groups in baseline demographics or graft anatomy, material, or age. Grafts undergoing open revision had an average of 1.3 ± 1.9 previous interventions, whereas those undergoing endovascular intervention had 0.46 ± 1.0 (P < .01). Postintervention primary patency for open and endovascular groups was 31.2% and 36.0% at 6 months and 21.0% and 14.0% at 12 months (P = .727). Secondary patency was 63.9% and 51.4% at 6 months for open and endovascular interventions, respectively, and 40.5% and 36.6% at 12 months (P = .677). The open group required an average of 3.0 procedures to maintain secondary patency to 12 months, whereas the endovascular group averaged 1.8 procedures (P = .198). The strongest predictor for primary failure was percutaneous thrombectomy (hazard ratio [HR], 4.18; P < .01), whereas the presence of an existing stent at the venous anastomosis (HR, 3.85; P < .05) and the age of the graft (HR, 0.973; P < .05) were predictive of abandonment. Open vs endovascular intervention was not predictive of failure or abandonment. Open and endovascular interventions at the venous anastomosis of AVGs lead to poor but comparable patency rates at 6 and 12 months.