Management of Arteriovenous Fistula Side Branches: Ligation or Coil Embolization

Objective: Failure of maturation of arteriovenous fistulae (AVF) remains an ongoing concern for dialysis access. One etiology is the presence of side branches that divert flow from the main AVF channel. This study aims to evaluate the outcomes of endovascular and open surgical interventions for AVF side branches in the setting of maturation failure. Methods: A retrospective review of all patients within a 10-year period with primary radio cephalic and brachiocephalic AVF was undertaken, and 380 cases of maturation failure related to branch diversion were selected for the study. Fifty-four percent and 48% of the AVF in the ENDO and OPEN groups respectively have concomitant stenosis further along in the flow path that required intervention by balloon angioplasty at the same time as a side branch intervention. All patients underwent duplex imaging or a fistulogram before intervention. Indications were low flow (<600 mL/min) or failure to increase in size (<6 mm diameter) in all cases. Interventions were divided into endovascular (coil embolization; ENDO) and surgical (branch ligation; OPEN) interventions. Outcomes of maturation (successful progression to hemodialysis (HD)), re-intervention, and functional dialysis (continuous HD for three consecutive months) were examined. Results: From January 2008 to December 2018, 187 patients (49 boolean AND of all cases with side branches; 65% female, age of 57 +/- 18 years; mean +/- SD) with poorly maturing radiocephalic (70%) and brachiocephalic AVF (30%) underwent intervention due to the presence of accessory venous branches only. Indications were failure to mature in 54% and low flow in 46%. The average time to intervention due to failure to mature was 5 +/- 4 weeks (mean +/- SD) after primary access placement. Eighty-one had coil embolization and 106 had open branch ligation. Technical success was 90% in ENDO and 100% in OPEN. Technical ENDO failures had a secondary open branch ligation but were considered failures for analysis. Repeat interventions by balloon-assisted maturation were required in 45% of all the cases with no difference between ENDO and OPEN. Recannulation of the ENDO branches occurred in 10% of the cases requiring repeat intervention. Sixty one percent of isolated endovascular (n = 49) and 64% of isolated open (n = 68) matured to successful cannulation (P = 0.84). Median functional dialysis durations remained equivalent between ENDO (2.6 years) and OPEN (2.8 years) groups (P = 0.12). Conclusion: There is an improved maturation rate following the ENDO group compared to OPEN interventions while both ENDO and OPEN modalities demonstrated similar long-term functionality