Comparison of The Efficacy of The Transposed Upper Arm Fistula And Upper Arm Prosthetic Graft: A Single Institutional Review of 368 Procedures

Karen Woo, MD, Cedars Sinai Medical Center, Department of Surgery, Division of Vascular Surgery, Cedars Sinai Medical Center, Los Angeles, CA; Alik Farber, MD, Los Angeles, CA; Som Kohanzadeh, MD, Los Angeles, CA; James Mirocha, PhD, Los Angeles, CA;

Purpose:
Dialysis access procedures constitute one of the most common operations performed by vascular surgeons. Despite a national initiative to promote the use of autogenous arteriovenous fistulas, some surgeons still preferentially place prosthetic grafts. We evaluated our experience with transposed upper arm arteriovenous fistulas (tAVF) and upper arm arteriovenous grafts (AVG) in order to compare patency rates and identify patient characteristics that would favor the use of one procedure over the other.
Methods:
A retrospective review was conducted of tAVF and AVG performed at our institution from 1998 to 2004. The tAVF group consisted of 127 basilic vein transpositions and 77 cephalic vein transpositions. We compared these with 164 AVG selected from a consecutive group of procedures chosen from the middle of our study period. tAVF was placed only for vein diameters ≥ 2.5 mm by duplex. Limb abandonment was defined as the time from access placement until the arm was deemed no longer functional for new access placement.
Results:
Mean follow up was 28 months. The patients in the tAVF group were significantly younger (mean age, 63 vs 67), fewer were African American (AA), and a higher proportion were male compared to the AVG group.
The primary patency rate for tAVF was significantly higher than for AVG: 57% vs 21% at 2 years (p<0.0001). The secondary patency rate for tAVF was also significantly higher than for AVG: 73% vs 35% at 2 years (p<0.001). Among the tAVF procedures, 17% required one or more revisions to maintain secondary patency, compared to 51% in the AVG group (p<0.0001). When a revision was necessary, a median of 1 revision in the tAVF group versus 3 in the AVG group were performed. Patients in the tAVF group had a significantly higher freedom from limb abandonment compared to those in the AVG group: 83% vs 55% at 5 years (p<0.0001).
Multivariate analysis revealed that AVG had a 64% higher risk of primary failure and a 51% higher risk of secondary failure compared to tAVF (p<0.0001). A history of previous upper arm access increased the risk of primary failure by 88% and the risk of secondary failure by 95% (p=0.0001). AA race was associated with an 89% increase (p=.007) and tAVF was associated with a 62% decrease (p=.0002) in the risk of limb abandonment.
Sub-group analysis of AA race, patients greater than 80 years old and patients with diabetes revealed that having a tAVF rather than an AVG significantly improved primary patency.
Conclusions:
These data reveal that tAVF have significantly higher patency rates than AVG, require significantly fewer revisions, and lead to higher freedom from limb abandonment. This effect extends to African Americans, patients older than 80 and diabetics. This data strongly supports the contention that as long as the patient is a candidate for an upper arm tAVF based on anatomical criteria, a tAVF should always be considered before an AVG.