A comparison of brachial artery-brachial vein arteriovenous fistulas with arteriovenous grafts in patients with poor superficial venous anatomy

Objective The autogenous arteriovenous fistula (AVF) has been shown to be superior to the arteriovenous graft (AVG) with respect to cost, complications, and primary patency. Therefore, the National Kidney Foundation Disease Outcomes Quality Initiative guidelines recommend reserving AVGs for patients who do not have adequate superficial venous anatomy to support AVF placement. The brachial artery-brachial vein arteriovenous fistula (BVAVF) has emerged as an autologous last-effort alternative. However, there are limited data comparing BVAVFs and AVGs in patients who are otherwise not candidates for a traditional AVF. Methods Patients who received a BVAVF from July 2009 to July 2014 were compared with those who received an AVG during the same period. At our institution, BVAVF and AVG are only performed in patients with poor superficial venous anatomy. Patient demographic data, operative details, and subsequent follow-up were collected. BVAVFs were performed with a two-stage approach, with initial arteriovenous anastomosis, followed by delayed superficialization or transposition. Our primary outcome measure was primary functional assisted patency at 1 year. Patients lost to follow-up were excluded. A subgroup analysis was also performed for patients in whom the BVAVF or the AVG was their first hemodialysis access surgery. Results During the study period, 29 patients underwent BVAVF and 32 underwent AVG. There were no differences in age, gender, or presence of diabetes between the two groups. The median days to cannulation from the initial operation were 141 (interquartile range, 94-214) in the BVAVF group and 29 (interquartile range, 14-33) in the AVG group (P < .001). Fewer patients required interventions to maintain or re-establish patency in the BVAVF group than in the AVG group (10% v. 44%; P < .01). The 1-year primary patency was greater for BVAVF (62% vs 25%; P < .01); however, there was no difference in the functional assisted primary patency rates at 1 year (45% vs 25%; P = .1). Subgroup analysis demonstrated greater 1-year primary functional assisted primary patency (52% vs 19%; P < .05) in patients without prior access surgery. Conclusions The BVAVF is a viable alternative to the AVG in patients with inadequate superficial venous anatomy, especially in access-naïve patients. The decision to perform BVAVF must be weighed against the delay in functional maturation expected compared with AVG.