Surgical Myocardial Revascularization with a Composite T-graft from the Left Internal Mammary Artery—Comparison of the Great Saphenous Vein with the Radial Artery

Abstract

Abstract Background Composite T-grafts between left internal mammary artery (LIMA) and radial artery (RA) are a common concept in complete arterial myocardial revascularization. The aim of the present study was to investigate whether the use of the great saphenous vein (SV) instead of RA leads to comparably good results in terms of outcome in this context. Methods Patients who underwent myocardial revascularization with a T-graft using RA or a segment of SV to the right coronary artery or circumflex artery between the beginning of 2014 and the end of 2019 at the Department of Cardiovascular Surgery, University Hospital Schleswig-Holstein, Campus Kiel were included. To minimize surgical variation, only patients were observed by a single senior surgeon in the department. Exclusion criteria were previous cardiac surgery, preoperative extracorporeal circulatory support, off-pump surgery, additional aortocoronary bypasses, and cardiac combination procedures. Results A total of 115 patients were studied. In 55 patients, the T-graft was placed between the LIMA and SV, and in 60 patients, the T-graft was placed between the LIMA and RA. Patients in the SV group were older (70.6 ± 7.8 vs. 58.5 ± 10.0 years; p < 0.001), suffered more frequently from non-ST elevation myocardial infarction (12.7 vs. 1.7%; p = 0.027), arterial hypertension (83.6 vs. 61.7%; p = 0.009), and atrial fibrillation (18.2 vs. 1.7%; p = 0.003). They were less likely to be active smokers (16.4 vs. 38.3%; p = 0.009) and less likely to have a history of variceal surgery (0 vs. 15.0%; p = 0.003). Calcification of the ascending aorta was also found more frequently in the saphenous group (18.2 vs. 3.3%, p = 0.009). Operative times and number of distal anastomoses did not differ significantly between the two groups. Postoperative deliriums (16.7 vs. 5.0%; p = 0.043) were observed more frequently in venous patients. Wound healing disorders of the leg (11.1 vs. 0%; p = 0.011) did only occur in SV group and wound infections of the arm only in the RA group. Complete follow-up was achieved in 74.8% of cases. Median follow-up was 60.3 (39.6; 73.2) months. Serious adverse cardiac–cerebral events (19.0 vs. 22.7%; p = 0.675) and mortality (14.5 vs. 6.7%; p = 0.167) did not differ significantly between the groups at follow-up. Myocardial infarction (0 vs. 2.5%; p = 1.000) and stroke (0 vs. 7.5%; p = 0.245) were observed exclusively in RA group. Percutaneous coronary intervention was required in single patients of RA group (0 vs. 15.0%; p = 0.028). No patient from either group underwent repeat coronary artery bypass grafting (CABG). The patients of SV group had angiographically competent grafts and open anastomoses. Graft failure was noted in a single patient in RA group, in which case both grafts and native coronary vessels were stented. Kaplan–Meier analysis revealed no significant survival disadvantage for SV group compared with RA group. Conclusion CABG with a composite T-graft between LIMA and a segment of SV may be comparable to bypass surgery with a composite T-graft between LIMA and RA. This might be true in terms of morbidity and mortality over an intermediate-term observation period. The results of our studies give rise to the hypothesis that the decision not to perform aortic bypass anastomosis may be more important than the choice of graft material.