Bilateral internal thoracic artery use: will another retrospective study ever strengthen the prospect?

Abstract

Since the introduction of coronary artery bypass grafting (CABG) 50 years ago [1], the conduit choice to bypass coronary artery stenosis continues to be a debate [2]. Already in the early years, CABG was performed using internal thoracic arteries (ITAs). However, it took several decades before evidence supported the use of ITAs, when Loop et al. showed that the left ITA (LITA) to the left anterior descending (LAD) artery prolonged long-term survival when compared with a saphenous vein graft [3]. Several research groups went even further by claiming that bilateral ITA (BITA) grafting was superior to the use of a single ITA (SITA). Although these initial studies were merely retrospective, the use of the LITA to the LAD became widely accepted with >90% of patients receiving this construction in contemporary CABG practice [4]. The fate of BITA grafting has been less fortunate. Over the past 25 years, a staggering number of retrospective and prospective observational studies have published long-term results of BITA versus SITA use. The majority of studies concluded that BITA grafting either reduced mortality or showed to have comparable results with SITA grafting. In a recent meta-analysis that exclusively included studies with long-term follow-up of at least a mean of 9 years (n = 9 studies comprising 15 583 patients with nearly 200 000 patient-years of follow-up), BITA versus SITA grafting significantly reduced mortality with a hazard ratio of 0.79, 95% confidence interval (CI) 0.75–0.84 [5]. Despite these encouraging results, BITA use remains remarkably low. Among patients who underwent CABG in the SYNTAX trial and registry, BITA grafting was performed in 10% in the USA and 25% in Europe [6]. Another ‘real-world’ analysis of 541 368 patients receiving CABG in the USA showed that the rate of BITA use increased over consecutive years but remained low during the entire study period: 3.6% in 2002 to 4.5% in 2005 [7]. The reason for the disappointing adoption of BITA use is multifactorial. The choice to use both ITA conduits depends strongly on the motivation of the surgeon. Surgeons in favour of BITA grafting cite evidence of observational studies that reported improved long-term survival, while those surgeons not particularly in favour of BITA grafting refer to the fact that there is currently no randomized clinical trial with long enough follow-up to confirm that BITA is indeed better than SITA. Moreover, surgeons may be reticent to using both ITA grafts because of the higher complexity of the procedure and the increased risk of sternal wound complications found particularly in diabetic, obese women. Besides patient characteristics like diabetes, body mass index and gender, age is a decisive factor [7], as the benefit of BITA over SITA only occurs after prolonged follow-up. A life expectancy of at least 10 years appears to be needed to show significant benefit of BITA use. In this regard, the current issue of the European Journal of CardioThoracic Surgery includes a retrospective study from Pettinari et al. in which they performed a propensity-matched analysis comparing BITA with SITA in elderly patients >70 years of age, to assess whether the benefit of BITA grafting can be extended to elderly patients [8]. The authors showed that patients who underwent BITA grafting, when compared with patients who received only a SITA, were younger, and less often had peripheral vascular disease, diabetes, chronic obstructive pulmonary disease or a recent myocardial infarction, all factors known to be associated with impaired survival during follow-up. The authors therefore were correct in performing a propensity-matched analysis after which no differences in the BITA and SITA groups existed. In terms of outcomes, BITA grafting showed to have superior survival after 3 months, 1 year and 10 years in the unmatched groups. After matching and as anticipated due to the differences in baseline characteristics, the difference between BITA and SITA was smaller, with BITA grafting only being superior after 10 years of follow-up (multivariate hazard ratio of 0.78, 95% CI 0.64–0.96; P = 0.04) [8]. The authors have to be congratulated for achieving such good results with BITA grafting. However, several significant limitations of the current study should be acknowledged, putting these results in the correct perspective. Of the 1328 and 2168 patients who underwent BITA and SITA grafting, respectively, only 892 propensity-matched pairs could be achieved. The fact that two-thirds of patients could not be matched indicates that these groups were significantly different. Although there were no longer any significant differences between the groups after matching, the authors were unable to match patients on the refinement of the surgical technique; the inclusion was during 1972–2006, with BITA being performed in the later decades, during which the CABG procedure changed considerably with improved outcomes. Furthermore, clinical variables as well as clinical judgement, which