PURPOSE: Fistulae between an arteriovenous hemodialysis graft (AVG) or fistula (AVF) and an adjacent vein are an unusual complication of hemodialysis access. Such fistulae may theoretically cause steal syndrome, extremity edema, or access dysfunction. We sought to use our experience and existing literature to develop a management algorithm for this access complication. METHODS: Twelve patients with AVG/AVF to adjacent vein fistulae found on fistulography were identified using a quality assurance database. Indications for fistulography, treatment rendered for both the fistulae and access stenosis, and outcome of treatment were determined. AVG/AVF to adjacent vein fistulae, when identified and considered to be significant, were treated with embolization. RESULTS: Five out of twelve patients had successful embolization of their AVG/AVF to adjacent vein fistulae. Reasons for treatment included partial thrombosis of the access to the level of the fistula (n=1), contribution to bleeding during dialysis (n=1), and concern for competing flow causing thrombosis (n=5). No recurrence was identified. Seven patients did not undergo embolization either because of failure to recognize the fistula (n=3) or determination that treatment was not indicated (n=4). Two untreated fistulae were found occluded at follow-up. Additional access treatment included angioplasty (n=11), covered stent (n=1), and mechanical thrombectomy (n=3). CONCLUSIONS: The significance of AVG/AVF to adjacent vein fistulae remains unclear; some resolve spontaneously, possibly related to PTA of outflow stenosis. Embolotherapy is an effective treatment for such fistulae when determined to be significant.
PURPOSE:Fistulae between an arteriovenous hemodialysis graft (AVG) or fistula (AVF) and an adjacent vein are an unusual complication of hemodialysis access. Such fistulae may theoretically cause steal syndrome, extremity edema, or access dysfunction. We sought to use our experience and existing literature to develop a management algorithm for this access complication. METHODS: Twelve patients with AVG/AVF to adjacent vein fistulae found on fistulography were identified using a quality assurance database. Indications for fistulography, treatment rendered for both the fistulae and access stenosis, and outcome of treatment were determined. AVG/AVF to adjacent vein fistulae, when identified and considered to be significant, were treated with embolization. RESULTS: Five out of twelve patients had successful embolization of their AVG/AVF to adjacent vein fistulae. Reasons for treatment included partial thrombosis of the access to the level of the fistula (n=1), contribution to bleeding during dialysis (n=1), and concern for competing flow causing thrombosis (n=5). No recurrence was identified. Seven patients did not undergo embolization either because of failure to recognize the fistula (n=3) or determination that treatment was not indicated (n=4). Two untreated fistulae were found occluded at follow-up. Additional access treatment included angioplasty (n=11), covered stent (n=1), and mechanical thrombectomy (n=3). CONCLUSIONS: The significance of AVG/AVF to adjacent vein fistulae remains unclear; some resolve spontaneously, possibly related to PTA of outflow stenosis. Embolotherapy is an effective treatment for such fistulae when determined to be significant.