By Michael H. Crawford, MD, Editor

SYNOPSIS: The results of a long-term follow-up study of matched pairs of patients undergoing surgical coronary artery revascularization showed that total arterial graft usage significantly reduced all-cause mortality vs. left internal mammary artery plus saphenous vein grafts.

SOURCES: Royse AG, Brennan AP, Ou-Young J, et al. 21-year survival of left internal mammary artery-radial artery-Y graft. J Am Coll Cardiol 2018:72:1332-1340.

Anyanwu AC, Adams DH. Total arterial revascularization for coronary artery bypass: A gold standard searching for evidence and
application. J Am Coll Cardiol 2018:72:1341-1345.

The concept of coronary artery bypass grafting (CABG) using only arterial grafts instead of veins is appealing. This is because of the excellent long-term results with the left internal mammary artery (LIMA) conduit and the observed 50% failure rate of saphenous vein grafts (SVGs) at 10 years. However, this practice has not caught on in most countries for several reasons. In Australia, radial artery grafts have been used instead of SVGs, either as a direct graft or a Y graft off the LIMA since 1997. Royse et al used a prosperity score-matching technique of 26 variables to compare the long-term survival of CABG using LIMA + SVG vs. total arterial revascularization (TAR). In a 1999 report from Australia, of 464 patients who underwent TAR, 346 were from the Royal Melbourne Hospital and had the LIMA + radial artery-Y graft (LIMA-RAY). Between 1996 and 2003, 534 patients from this hospital had LIMA + SVG. After prosperity-matching these two groups, there were 232 pairs. At the time the study ended, post-operative follow-up ranged from 13-21 years. All patients received statins post-operatively. The primary outcome was all-cause mortality.

Prior to matching, overall survival was 9.9 years in the LIMA-RAY group and 10.9 years in the LIMA + SVG group (P < 0.001). After matching, the mortality hazard ratio for LIMA + SVG vs. LIMA-RAY was 1.3 (95% confidence interval, 1.0-1.6; P = 0.038). When LIMA-RAY was compared to other methods of TAR in 5,800 patients with 332 matched pairs, no significant differences were found. The authors concluded that LIMA + SVG leads to lower long-term survival vs. TAR.


The results of this study and others have shown that TAR is superior to LIMA + SVG. Why haven’t cardiac surgeons embraced TAR? Currently, 90% of multivessel CABG procedures are of the LIMA + SVG variety. One reason is that if moderate lesions are bypassed with an arterial graft, the graft often fails due to competitive flow. Arterial grafts work best in > 90% stenoses. Another reason is that TAR takes longer (up to an hour or more in some cases). Yet, surgeons receive the same fee and lose time they could spend performing another operation. Also, longer surgeries can lead to higher morbidity. However, in the only large randomized trial of LIMA vs. bilateral IMA grafting (ART), morbidity (including wound infection) was not significantly different, even in diabetic patients. ART also showed no difference in outcomes, but it was flawed because in the LIMA group, arteries other than the LIMA could be used as second grafts.

Previous studies have revealed that revascularizing the LAD is the most important factor that determines long-term outcomes. These data discourage TAR; however, until now, there have been no very long-term data from TAR. This study shows a TAR survival advantage for 13-20 years. In addition to the long-term follow-up, the strengths of this study included a tight matching scheme of 26 variables. Detractors would argue that there are biases for which prosperity matching cannot account. For example, if a patient seems to be disadvantaged from noncardiovascular reasons such that long-term survival is not expected, the surgeon may choose SVG. Also, TAR requires a higher surgical skill level than SVG, so TAR patients may fare better for that reason alone. There were limitations to this study in addition to its retrospective, observational nature. There were no causality data concerning mortality. Further, the authors did not examine other major adverse cardiovascular events beyond all-cause mortality.

The authors of an accompanying editorial argued that the superiority of arterial grafts has been demonstrated by the world’s adoption of the LIMA to LAD bypass and prior research on TAR going back decades. They believe we should use TAR more and that health systems need to realize that the higher upfront costs will be mitigated by long-term cost savings due to fewer readmissions for ischemic events. The editorialists also suggested that we develop centers of excellence for CABG or at least employ one surgeon at each center who is dedicated to TAR. However, the editorialists acknowledged that in some cases SVGs are preferable, such as the patient with limited life expectancy where pain control is the main objective. Although studies of TAR revealed an overall 0.2% 30-day mortality, there may be patients for whom a longer operation would pose significant risks.

Given that there still are significant numbers of patients in whom CABG is preferable to a percutaneous intervention, we should push our surgeons to perform more TAR procedures.