Bilateral internal thoracic artery grafts for myocardial revascularization in insulin-dependent diabetic patients: time for wide clinical practice?

Abstract

In this issue of the European Journal of Cardiothoracic Surgery, Gatti et al. [1] analysed early and long-term outcomes of routine use of bilateral internal thoracic artery (BITA) grafts for left-sided myocardial revascularization in a consecutive cohort of 188 insulindependent diabetic patients. In addition to low in-hospital mortality, they reported that a higher incidence of deep sternal wound infection (DSWI), when compared with non-diabetic patients receiving BITA grafts, did not impact long-term outcome. While, these results are welcome in the cardiac surgical community, concern remains regarding the use of BITA grafting, possibly associated with an increased incidence of early complications, such as DSWI, particularly in diabetic patients. To stimulate change in clinical practice and encourage more widespread use of BITA grafting, in particular in diabetic patients, two questions should be addressed: (i) Is the use of BITA grafts important for all patients? (ii) Do all diabetic patients share a similar profile? Since the 1980s, it has been known that the internal thoracic artery (ITA) graft to the left anterior descending coronary artery significantly improves survival and is associated with lower incidence of late cardiac events and better quality of life over a 10-year period. These benefits result from greater long-term patency of ITA grafts when compared with saphenous vein grafts [2]. With increasing evidence, studies have shown that using two ITAs might be better than one [3]. However, not all patients benefit equally from BITA grafting, while the advantage at followup occurs at different time points for different patient subsets. Patients without abnormal left ventricular (LV) function or comorbidities are ideal patients for BITA grafting. For patients with LV dysfunction or non-cardiac risk factors, BITA grafting still carries better survival [4]. However, for patients having LV dysfunction and non-cardiac risk factors, the overall prognosis appears poor, and the benefit of BITA grafting to improve late survival is controversial [5, 6]. Overall, survival curves of patients with BITA grafting versus single ITA (SITA) grafting begin to diverge about 10 years after surgery, possibly related to saphenous vein graft progressive disease [3, 4]. The arterial revascularization trial randomized 3102 patients between BITA grafting and SITA grafting, with a primary outcome of 10-year survival [7]. Preliminary published data reported similar 30-day and 1-year outcomes in terms of death (1.2 and 2.4% for both groups), stroke, myocardial infarction and repeat revascularization. However, there was a significantly increased need for sternal wound reconstruction in the BITA group (1.9 vs 0.6% in the SITA group) [7]. Regarding the second question, it has been demonstrated that poorly controlled diabetes, but not well-controlled diabetes, significantly impairs endothelium-dependent and endotheliumindependent relaxation of the human peripheral microvasculature. These changes may contribute to the less favourable postoperative outcomes after coronary artery bypass grafting (CABG).. In contrast, well-controlled diabetes is associated with improved peripheral arteriolar function after cardiopulmonary bypass and cardiac surgery [8]. In this clinical scenario, the glycated haemoglobin A1c (HbA1c) measure proved to be a powerful predictor of in-hospital mortality and morbidity. Complication and DSWI rates were significantly higher in diabetic patients with HbA1c ≥7% when compared with patients with HbA1c 8.6%. In elective situations, it has been proposed that surgery should be delayed until adequate glycaemic control is achieved [9]. In the study of Gatti et al. [1], it is likely that the higher rate of sternal complications was at least partly due to the routine use of BITA grafts, without any preoperative selection of candidates for left-sided BITA grafting. Even so, DSWI was not a predictor of poor late survival. The 10-year non-parametric estimate of overall survival was 57.7% and it is likely that these patients will benefit from BITA grafting. With better glycaemic control and preoperative analysis of HbA1c, the incidence of DSWI can probably decrease and the overall results may improve in patients in the future.