Following its introduction by Effler and Sabiston in 1965, SA-LMCA was abandoned because of the high mortality and surgical failure. The excellent report by Hichcock and colleagues revived this procedure. Dion, Sullivan, Villemot, and coworkers published innovative reports concerning the surgical angioplasty approach and technique. We used the anterior approach exclusively and divided the main pulmonary artery for better visualization in half of our cases. This technique always provided an excellent view extending from the LMCA ostium to the bifurcation.

The indications for SA-LMCA are controversial, particularly regarding the extent and location of the LMCA stenosis and the existence of calcification. Botman and colleagues excluded patients with visible calcification and disease extending to the LMCA bifurcation [7]. The involvement of the distal LMCA or bifurcation may make it more difficult to reconstruct the LMCA and cause disastrous consequences.

Age is another issue related to the indications and surgical risks of this procedure. In this study, 5 of the 6 patients over 65 years of age died (2 cardiac and 3 non-cardiac deaths). Elderly patients may have more calcified lesions in the LMCA. Additionally, they often tend to have coexisting valvular disease, peripheral coronary artery disease, and comorbidities. Dion and colleagues gradually broadened their indications to include more distal stenosis, visible calcifications on CAG, advanced age, and decreased left ventricle (LV) function [8]. However, they concluded that SA-LMCA should be carefully attempted in patients with LMCA calcification or patients over 60 years of age. Most surgeons, including the authors of this study, agree with the opinions expressed by Dion regarding these indications.

The onlay patch material is the most important issue regarding the prevention of acute thrombosis and late restenosis of the LMCA. The saphenous vein and autologous pericardium have been commonly used for SA- LMCA [9]. The saphenous vein is well matched in size and preserves the fibrinolytic properties of the endothelium. However, its elasticity may cause a tendency to dilate. Dion and colleagues suggested that the saphenous vein might be preferable to autologous pericardium due to its potential fibrinolytic activity [4]. Martinovic and colleagues used a saphenous vein as the patch material in 27 patients and reported one aneurysmal dilatation [10]. In the present study, we used bovine pericardium in the majority of our patients (17/24) because of easy handling. We believe it would be more difficult to tailor and sew a saphenous vein or internal thoracic artery patch to the LMCA compared with a pericardial patch because of thinness and weakness. Currently, both the saphenous vein and bovine pericardium have been widely used in carotid endarterectomy (CEA) with patch angioplasty. Several studies have documented the safety, efficacy and durability of bovine pericardium as a CEA patch. In a previous study that analyzed 456 CEA cases over a 10-year period, both the carotid clamping time and total operation time were shorter in the bovine pericardial patch group compared with the saphenous vein patch group because of the easy handling and suturing. The study also revealed a similar incidence of restenosis between the two patch materials (2.8% for bovine pericardium versus 3.4% for saphenous vein) and identified 4 patients who developed late aneurysmal dilatation in saphenous vein patches compared with no cases involving bovine pericardial patches [11]. Another study also reported a low incidence of restenosis (1.6%, 4/256) over 12 years in patients who underwent CEA with bovine pericardial patch angioplasty [12]. Bovine pericardium also provides the benefit of off-the-shelf availability and, has a reliable consistency and strength to allow a tight fitting closure, which yields less suture line bleeding and prevents aneurysmal dilatation [13]. However, bovine pericardium also has disadvantage of incurring calcification, degeneration and restenosis.

In this series, we observed no acute or early thrombosis after SA-LMCA using a pericardial patch, despite the absence of specific anticoagulation therapy except for aspirin. However, the late failure of a pericardial patch caused 8 target lesion related MACE (5 in the bovine pericardial patch group and 3 the autologous pericardial patch group). Both pericardial patches exhibited similar MACE rates (bovine pericardium, 5/16 [31.3%]; autologous pericardium, 3/7 [42.9%]), although these numbers were too small to evaluate for the presence of statistically significant differences. Nevertheless, 4 patients in the bovine pericardial group died eventually at 41, 60, 96, and 132 postoperative months, while 3 patients in the autologous pericardial group survived a catastrophe. Notwithstanding the higher number of fatal consequences in the bovine pericardial group, we did not implement repeat CAG for the patients who suffered sudden cardiac death. Unfortunately, we have no information about the conditions of the LMCA with regard to mortalities. As restenosis is frequently asymptomatic, several authors recommend frequent imaging studies including surveillance CAG to avoid catastrophic consequences for patients undergoing SA-LMCA, regardless of whether they have cardiac symptoms [14]. It is currently unclear whether restenosis is directly related to the patch material, surgical techniques, or LMCA stenosis disease process per se. More data are necessary to establish standard guidelines regarding patch materials because of the small number of reported cases and short follow-up periods studied. In addition to commonly used patches, Liska and colleagues proposed using a proximal segment of the internal thoracic artery, and Malyshev et al. introduced the pulmonary autograft patch. Both studies reported excellent early results. The proximal right internal thoracic artery is sizable, pliable, and sufficiently robust to reconstruct the LMCA, in contrast with the distal segment. The pulmonary artery shares a common embryological origin with the aorta and has similar endothelial properties [15,16]. Internal thoracic artery and pulmonary autograft patches may be ideal patch materials and are preferable to the saphenous vein and pericardium, provided that they exhibit excellent long-term outcomes.

The cumulative estimates in this study at 5, 10, and 15 years are summarized as follows: overall survival of 96, 87, and 73%; freedom from cardiac death of 96, 87 and 82%; and freedom from MACE of 91, 70, and 58%, respectively. In a previous study evaluating a 15-year follow-up after CABG from the Coronary Artery Surgery Study (CASS) registry, the overall survival was 90, 74, and 56% at 5, 10, and 15 years, respectively [17]. Furthermore, another study from the CASS registry including 630 cases with left main equivalent coronary artery disease treated with CABG demonstrated a cumulative survival of 88, 69, and 44% at 5, 10, and 15 years, respectively [18]. Sabik and colleagues evaluated 3,803 patients treated with CABG for LMCA stenosis and reported an overall survival of 83, 64, and 44% at 5, 10, and 15 years, respectively [19]. Taggart and colleagues reviewed several studies of CABG and PCI for LMCA stenosis and reported an in-hospital mortality of 2-3% and a 30-day mortality of 3–4.2% after CABG for LMCA stenosis [14]. Considering that CABG is a well-established procedure in contrast with SA-LMCA, the early and long-term outcomes of this study are remarkable despite the small number of cases and great controversy surrounding patch materials.

Notwithstanding our important findings, the present study had a few limitations such as a retrospective observational design with a small sample size and the lack of a control group. Because of the small number of patients included in this study, we could not perform a multivariate statistical analysis or draw appropriate conclusions regarding statistical significance.