Data from the University of Pittsburgh Medical Center (Pittsburgh, USA) indicate that the open approach to managing thrombosed haemodialysis grafts with venous anastomotic lesions is not associated with superior primary patency when compared to the hybrid approach. These findings also show that an existing stent in an arteriovenous graft (AVG) is a strong predictor of loss of patency and graft abandonment.
Catherine Go (Division of Vascular Surgery, University of Pittsburgh Medical Center, Pittsburgh, USA) presented the data at the 2019 Society for Vascular Surgery (SVS) Vascular Annual Meeting (VAM; 12–15 June, National Harbor, USA). She told delegates that AVG were “plagued by poor patency” and failure of an AVG leads to “increased cost and morbidity”, adding that thrombosis in an AVG most commonly occurred because of a venous anastomosis stenosis.
However, according to Go, the “jury is still out” about how to manage such thromboses. Two meta-analyses—published in 2002 and 2019—have reported the superiority of the open treatment over purely endovascular methods, but few studies have investigated the hybrid technique, which Go reported is commonly used at her centre (University of Pittsburgh Medical Center). She explained that this involves “open thrombectomy followed by endoluminal treatment (balloon angioplasty, cutting balloons, or stents/stent grafts)”.
“Based on historical data, we hypothesised that open revision for AVG venous anastomosis is associated with better patency when compared to hybrid therapies”, Go stated. Therefore, in the retrospective study, Go and colleagues compared patency outcomes for patients who underwent open repair for a first-time AVG thrombosis with those for patients who underwent the hybrid approach. The primary endpoints were primary patency (time from index thrombectomy to first reintervention or thrombosis) and secondary patency (time from index thrombectomy to graft abandonment). The investigators also reviewed technical success and primary-assisted patency (time from index thrombectomy to first rethrombosis).
Overall, 97 patients underwent intervention for occluded AVGs at the University of Pittsburgh Medical Center between 2014 and 2018. Of these, 34 underwent the open approach (73.5% patch angioplasty and 26.5% jump bypass) and 63 (22.2% stent graft, 7.9% cutting balloon, and 1.6% bare metal stent) underwent the hybrid approach. At 30 days, there were no significant differences in rate of the failure of the approach: 29.4% for open vs. 27% for hybrid (p=0.799). A multivariate analysis showed that a higher number of prior fistulograms and African-American race were both predictive of protection against failure. There were also no significant differences between groups in primary patency, primary-assisted patency, or secondary patency at either six months or 12 months.
Go reported that the number of prior fistulograms was again predictive of protection against primary patency failure (at six/12 months). Central occlusion and a hypercoagulable state were predictive of loss of primary-assisted patency. An existing venous anastomosis stent predicted both loss of primary patency and loss of secondary patency.
“AVG patency after thrombectomy is poor. Open versus hybrid treatment of AVG venous anastomosis stenosis are not associated with differences in patency. An existing stent at the venous anastomosis is a strong predictor of loss of patency and abandonment,” Go reported. She added that, given the association between an existing stent and graft abandonment, an operator should “start planning future access options” when using a stent to salvage a thrombosed AVG.