Background
Use of arterial grafts for coronary artery bypass grafting (CABG) has been argued to improve the angiographic graft patency, and thereby, the patient outcome by fewer cardiac events such as reoccurrence of angina pectoris, repeat revascularization, myocardial infarction (MI) and even death. There is only limited and unclear evidence from randomized trials, but in large-scale retrospective studies it is clear that use of the internal thoracic artery (ITA) as a bypass graft to especially the left anterior descending (LAD) coronary artery leads to improved graft patency and survival and fewer cardiac events compared with use of the greater saphenous vein as a graft (SVG). Bilateral ITA (BITA) grafts seem to increase this benefit further. Because of sternal wound complications related to BITA harvest, the radial artery (RA) has been proposed and widely used in addition to the left ITA (LITA), but also together with BITAs when more grafts are necessesary. The question whether the RA has superior patency than the SVG is therefore central, but in the long term unanswered. Also, the safety and efficacy of a strategy of total arterial revascularization (TAR) using one or both ITA grafts together with the RA is poorly documented. Finally, health-related quality of life (HRQoL) seems to improve after CABG, but whether TAR has an influence is unclear.

Methods
In this trial including 161 patients randomized to TAR (LITA or BITA and RA) versus 170 patients randomized to conventional revascularization (CR) (LITA and SVGs) and operated with use of heart-lung machine, the primary outcome measures are overall graft patency (graft stenoses below 50%) defined as the patency index (number of patent grafts divided by the number of constructed grafts) and cardiac event-free survival (readmission for suspected or verified MI, unstable angina pectoris, arrhythmia, congestive heart failure, angiography or percutaneous or surgical revascularization) after 1 and 5 years. Secondary outcome measures are perioperative safety, reoccurrence of angina pectoris and HRQoL measured by means of Short Form-36 (SF-36) questionnaires preoperatively, after 3 months and 1 and 5 years.



Results
Our trial demontrates that TAR can be performed on low to medium risk patients without an increased number of complications in the first three postoperative months. With TAR, slightly fewer grafts were made compared with CR: 2.9 ±0.9 versus 3.2 ±0.9 (mean ± one standard deviation) (P=0.004), but clinical outcome was comparable.
Angiographic follow-up of 83% of all patients, showed equal overall mean patency index after 11 months of 87% (95% confidence interval [CI] 83-91%) for the TAR group patients versus 88% (95% CI 85-91%) for the CR patients (P=0.52). In 72% of patients of the TAR group and in 67% of patients of the CR group all grafts were patent (P=0.45). Likewise, the mean graft patency index for the RA in the TAR group and SVG in the CR group were comparable: 85% (95% CI 79-92%) versus 86% (95% CI 80-92%).
The cardiac event-free survival was comparable in the two groups after 11 months with a hazard ratio (HR) of 1.09 (95% CI 0.70-1.69) (P=0.70) for TAR compared to CR. There were 37 patients (23%) in the TAR group and 43 (25%) in the CR group who suffered a cardiac event (P=0.70).
HRQoL increased markedly following bypass surgery in both groups postoperatively after 3 and 11 months, however, there was a tendency towards better results in the TAR group. The prevalence of reocurring angina pectoris in the two groups and sequels related to harvest of graft material did not differ statistically.

Conclusion
Coronary bypass surgery using TAR can be performed safely with results comparable to that of CR for the first 11 months postoperatively. The main benefit from TAR is expected to occur at a later stage and later follow-up will help to clarify if results after TAR are comparable to or even superior to CR.