Arterial conduits (particularly the left internal mammary artery [LIMA]) have been shown to be superior to saphenous vein (SV) grafts in terms of long-term patency in patients with coronary artery disease (CAD) who are undergoing coronary artery bypass grafting (CABG). Although the LIMA is the arterial conduit of choice, patients who require more than one bypass must receive either a second arterial graft or a vein graft. While the radial artery is the most frequently used arterial graft in this setting—because it is the easiest to harvest—there is little data concerning its long-term graft patency.

The CABG arm of the prospective, randomized Veterans Administration (VA) Cooperative Study, also known as CSP-474, included 733 patients with stable CAD who were undergoing elective CABG with a LIMA and needed at least one other graft. Findings from CSP-474 were presented by Steven Goldman, MD, Tucson VA Hospital, Tucson, AZ.

Patients were randomized to either radial artery graft (RA; n=366) or SV graft (SV; n=367) to the best recipient vessel. Angiographic assessment was performed at one week and one year post-CABG (completed in 73%) in order to monitor disease progression. There was 89% power to detect a difference in the primary endpoint of angiographic patency at one year. The secondary endpoints included difference in selective graft patency (distal anastomosis to the left anterior descending, circumflex, or right coronary artery) between RA and SV at one year, high-grade disease (string sign) in the graft, and endoscopic harvesting. Other analyses included patency data on cardiopulmonary bypass pump versus “off pump,” cost analyses, and quality-of-life assessment at 3 months and one year.

There was no difference in angiographic patency at one year between RA and SV (89% for both). One-week patency rates were also similar (99% for RA vs 97% for SV). There was no difference between RA and SV in the secondary endpoint of selective graft patency at any target. More high-grade disease (defined as a string sign) was observed in the RA group (8%) compared with the SV group (1%; p<0.001), though these rates remained quite low. Endoscopic harvest of SV was associated with lower patency rates compared with traditional harvest of the SV (78% vs 91%; p=0.009), but there was no significant difference in RA patency rates (100% vs 89%) that were dependent on mode of harvest. Additionally, complication rates that were associated with RA and SV were low compared with similar cohorts. At one year, the rate of stroke was 2.0%, the rate of death was 2.0%, and the rate of MI was 1.0%. Operative mortality occurred in 0.7% of patients.

Use of cardiopulmonary bypass (ie, “on” versus “off pump”) did not appear to impact RA patency (89% for both). However, higher patency was observed in SV patients who were on pump (90% vs 78%). It is important to note that these data are based on a small number of participants (off-pump patients n=41 for RA and n=48 for SV) and may not be reflective of outcomes that are expected from a real-world population. Quality-of-life assessments for both study groups were comparable at 3 months and one year. While overall hospital costs were similar for the two groups, surgical costs were higher for RA compared with SV ($13,629 vs $12,484 for SV; p<0.001).

In patients with stable CAD who are undergoing CABG with a LIMA and are in need of at least one other graft, RA is not superior to SV in terms of patency at one year. The CSP-424 study is ongoing, and angiographic data will be evaluated during a planned 5-year angiographic follow-up.