The forearm arteriovenous graft between the brachial artery and the brachial vein as a reliable dialysis vascular access for patients with inadequate superficial veins.

OBJECTIVE: The arteriovenous fistula (AVF) is recommended as the preferred hemodialysis access. However, placing an AVF in all patients may result in poor access outcomes and increased central venous catheter (CVC) use because of increased comorbid conditions, age, and suboptimal vessels. In patients with inadequate superficial veins for AVFs, the use of the brachial veins for creation of forearm arteriovenous grafts (AVGs) has received limited attention. This retrospective study aimed to evaluate outcomes of forearm brachial-brachial AVGs (BB-AVGs) placed in patients with poor superficial veins.

METHODS: We identified 111 BB-AVGs created in 111 consecutive patients, using standard-walled polytetrafluoroethylene grafts, between January 2010 and December 2015. After excluding 6 patients (non-dialysis initiation, missing information, and death within 1 month), we included 105 patients from 21 dialysis centers. We analyzed primary failures, time to cannulation, patency, complications, and revisions. Patency rates were calculated by the Kaplan-Meier method. The incidence of complications and revisions was expressed as number of events per person-year.

RESULTS: A total of 105 patients (median age, 69 years) were followed up for a median time of 21.2 months (interquartile range, 9.2-36.5 months). Of the patients, 72.4% were on chronic hemodialysis and had previously undergone one or more access procedures. At the time of BB-AVG placement, prior accesses were 39 AVFs, 20 tunneled CVCs, and 17 AVGs. BB-AVG rates of primary failure and revision before cannulation were 7.6% and 5.7%, respectively. BB-AVGs were cannulated after a median time of 3.4 weeks (interquartile range, 2.8-4.1 weeks). Primary patency rates at 12, 24, and 36 months were 49.5%, 29.5%, and 19.5%. Secondary patency rates at 12, 24, and 36 months were 76.3%, 62.7%, and 54.6%. After cannulation, the incidence of complications and revisions was 1.054 and 0.649 per person-year, respectively. Most complications and interventions were due to thrombosis (0.527 per person-year) or stenosis (0.381 per person-year) and related interventions (0.490 per person-year). A minority of patients experienced AVG infections (0.052 per person-year), with only two requiring access removal.

CONCLUSIONS: In patients with poor superficial veins, the forearm BB-AVG is a reliable access because of low access-related morbidity and considerable long-term access survival. BB-AVG placement has the advantage of preserving proximal vessels. In these patients, such an approach can delay both rapid exhaustion of vascular sites and early recourse to CVC permanent use.