Predictors of arteriovenous graft patency after radiologic intervention in hemodialysis patients

Arteriovenous grafts in hemodialysis patients are prone to recurrent stenosis and thrombosis, requiring frequent radiologic and surgical interventions to optimize their long-term patency. Little is known about the factors that determine graft outcome after a radiologic intervention. The present study examined the clinical and radiologic predictors of intervention-free graft survival after elective angioplasty or thrombectomy. A prospective computerized database was used to determine the outcomes subsequent to all graft angioplasties (n = 330) and thrombectomies (n = 326) performed at the University of Alabama at Birmingham between April 1, 1996, and June 30, 1999. Primary graft survival rates after angioplasty and thrombectomy were 86% versus 43% at 1 month, 71% versus 30% at 3 months, 51% versus 19% at 6 months, and 28% versus 8% at 12 months, respectively. The median intervention-free graft survival time was substantially longer after angioplasty than thrombectomy (6.7 versus 0.6 months; P < 0.001). The superior outcome of angioplasty over thrombectomy was observed even for the subset of procedures with no residual stenosis (median survival, 6.9 versus 2.5 months; P < 0.001). The median graft survival was inversely related to the magnitude of residual stenosis for both elective angioplasty and thrombectomy. Median intervention-free graft survival after angioplasty was inversely related to the postangioplasty intragraft to systemic systolic pressure ratio (7.6, 6.9, and 5.6 months for ratios <0.4, 0.4 to 0.6, and >0.6, respectively; P < 0.001). Intervention-free graft survival after angioplasty or thrombectomy was not affected by graft location (forearm versus upper arm), number of stenotic sites, or presence of diabetes. In conclusion, graft survival is substantially longer after elective angioplasty than thrombectomy, even when the radiologic appearance after the procedure suggests complete resolution of the stenotic lesion. Moreover, the risk for requiring a subsequent graft intervention can be predicted from two simple radiologic measurements: grade of stenosis and intragraft to systemic systolic blood pressure ratio. These parameters may help determine the frequency of monitoring for recurrent stenosis in a given graft.