Discussion
Exceptional results of the NT SVGs have been demonstrated when used in on-pump CABG 3-5 . To our knowledge, no randomized study on patency rates of NT SVGs in off-pump CABG has been reported. This is a sub-study of a large international randomized multicenter trial 9 that was designed to compare the clinical outcomes in high risk patients operated with either on- or off-pump CABG. Graft patency of the conduits was not investigated in this study. As we routinely use NT SVGs at our department, all of our patients included in this trial received one or two NT SVGs in addition to the LITA graft. Thus, the aim of this study was to evaluate the patency of the NT SVGs in both on- and off-pump CABG.
There was no superiority of on-pump CABG in terms of graft patency, 91.3% vs 84.7% for on- and off-pump respectively, p=0.35. The patency was excellent at five years for both the LITA and NT SV grafts when used to bypass the LAD territory, 29/30 (96.7%) for the LITA vs 32/32 (100%) for NT SVGs independent of the surgical technique used. This is in agreement with our previous results 13. On the other hand, the patency rates of NT SVGs to the right coronary targets were clearly less favorable, 80% in on-pump vs 62.5% in off-pump surgery. Further analysis showed that 24.7% of the PDAs in this study had a diameter around 1 mm, potentially jeopardizing the patency of SVGs attached to these vessels. These results differed from earlier results of a randomized trial with on-pump CABG in which the patency rate of NT SVGs anastomosed to the PDA at 1.5, 8.5 and 16 years were 100%, 95% and 94% respectively 3-5.
The ROOBY and DOORS trials 14, 15 reported similar results with lower patency of grafts used to bypass the right coronary region in off-pump CABG. ROOBY showed 75.6% patency at one year while DOORS reported 72% patency at six months postoperatively. This was also true for both SV and arterial grafts used in a composite formation to bypass the right coronary artery (RCA). Hwang et al. reported lower patency of SV and arterial grafts to the RCA at one year postoperatively, although the patency was similar between the different conduits, 84.9% vs 82.9 for SV and arterial grafts respectively16. There are several possible explanations for this discrepancy in the patency rates of grafts anastomosed to different coronary territories between on- and off-pump CABG. Difficulties in surgical access, a demanding technique requiring more surgical experience and smaller target vessels could be the most acceptable justifications.
Kim et al. reported improved NT SVG patency in off-pump CABG in the context of composite grafts one year postoperatively17. The patency rate in the NT group was significantly higher than the minimal manipulation SVG group before and after propensity score matching, before, 97.4% vs 92.4%, p  = 0.024; after, 97.3% vs 92.6%, p  = 0.051. Hence, the NT SVG may improve the results and simplify a complex off-pump procedure especially when it comes to sequential grafts where the surrounding tissue in NT SVGs remains intact and acts as a biological, external stent protecting the long sequential grafts from kinking and reducing the potential for technical error. The number of grafts and distal anastomoses was higher in on-pump than in the off-pump group (3.8 vs 3.1). This corresponds with several previous studies 18 19. Although previous studies have shown no differences in graft patency between hand-sewn proximal anastomoses with the clamp-less Heartstring device vs partial clamp 20, we chose to use the clamp-less technique avoiding excessive manipulation of the aorta.
The NT SVG was used to substitute the LITA in low-grade stenosis, in cases where the LITA was surgically damaged, and in patients with multiple co-morbidities. This is in accordance with a previous study on a similar group of patients. Hence, the patency rate of NT SVGs to the LAD was 98% at a meantime of six years 13. The patients who received a planned SVG to the LAD were older, hade a lower left ventricular ejection fraction, higher risk scores, a low grade stenosis in the LAD and underwent concomitant procedures. The obvious limitation of our study is the small number of patients. Being a sub-study, the power estimation was not calculated to investigate the current issue.