Is there an optimal minimally invasive technique for left anterior descending coronary artery bypass?

For several decades, left internal thoracic artery (LITA) bypass grafting has been recognised as the gold standard for left anterior descending coronary artery (LAD) revascularization and its beneficial impact was demonstrated in conventional coronary artery bypass grafting (CABG). During the past decade, minimally invasive (MI) CABG, based on the lack of sternotomy approach, has been developed according to the evolution of technology and dedicated surgical tools, and it has been mainly performed in isolated LAD bypass grafting. Nowadays, MI-CABG brings together different surgical concepts: Port-Access surgery (PA-CABG) based on on-pump CABG with mini-thoracotomy and hand-sewn anastomosis, minimally invasive direct coronary artery bypass (MIDCAB) based on off-pump CABG with mini-thoracotomy and hand-sewn anastomosis, and totally endoscopic coronary artery bypass grafting (TECAB) based on on-pump or off-pump surgery with robotic assisted anastomosis. These MI-CABG techniques have been compared to conventional CABG according to observational research because of the lack of randomized trial: the expected clinical results must be at worst as good as conventional CABG; the observational results are at best the same [1–3]. In the same way, these different MI-CABG techniques have never been compared together. In this series, we report our experience in minimally invasive LAD grafting with a comparative analysis between PA-CABG, MIDCAB and off-pump TECAB, in order to answer the question: is there an optimal minimally invasive technique for isolated LAD grafting?

From January 1998 to December 2008, 160 eligible patients for elective LAD revascularization were referred to minimally invasive CABG surgery. There were 135 males and 25 females, mean age was 58 ± 11 years and Euroscore 1.7 ± 1.7. There were two surgical periods: from January1998 to September 2003, the intend-to-treat surgical procedure was either PA-CABG or MIDCAB depending on the patient's condition, and after the purchase of standard Da Vinci robotic system in September 2003, the intend-to-treat surgical procedure was either off-pump TECAB or robotic enhanced MIDCAB depending on the patient's condition. In this series, the patients were categorized into three groups according to the surgical procedure performed: PA-CABG (n = 48), MIDCAB (n = 53), TECAB (n = 59).

Our study confirms minimally invasive CABG, regardless the technique used, is safe with a 0.6% early mortality, and effective with a 98 ± 2% 5-year survival, a 93 ± 6% 5-year freedom from reintervention and a 85 ± 9% 5-year freedom from angina. The early patency of LITA to LAD (94%, 77/82) is comparable to those of conventional on-pump CABG (91%) according to IMAGE trial [5] or off-pump CABG (92%) according to randomized trial [6]. All procedures were performed without conversion from off-pump to on-pump procedure or to sternotomy approach, and all LAD bypass failures could be treated by stenting.

We have analysed MI-CABG results between three different techniques developed during the past decade. PA-CABG is the most sophisticated procedure involving on-pump surgery and endo-aortic clamping technique. In this group, we have observed very satisfactory early and late results, without any post-operative major event; symptom-based angiography demonstrated good graft and anastomosis patency in all cases. Results are comparable to those previously reported with this technique [7, 8].

MIDCAB is the less demanding procedure and has gained widespread acceptance according to excellent results provided [9, 10], which our series has confirmed. Only one case of reintervention occurred, related to early anastomosis bleeding. We have observed no differences in results between classical and robotic-enhanced MIDCAB. In this study, there was no difference in operative risks and mid-term results between PA-CABG and MIDCAB.

TECAB is controversial [2]; off-pump TECAB is the less invasive concept in LAD grafting, nevertheless results are not as good as expected. De Canniere [11] reported a 2.2% early mortality, a 92.1% early patency and a 4.1% reintervention rate at 30 days. In our series, early mortality was 1.7%, early patency was 93.2% and the reintervention rate before discharge was 6.8%. Two more patients underwent reintervention of LAD, 1 and 2 months postoperatively, after symptom-related angiography which showed LAD bypass dysfunction undetected by coronary CT scan before discharge. The actuarial freedom from angina and from reintervention were significantly lower in the TECAB group (Figure 1); it was directly related to a primary bypass failure which remains the main concern in the TECAB procedure. In our experience, modifications of the anastomosis technique allowed to improve the patency: after the occurrence of post-anastomotic dysfunction cases, distal LAD occlusion during anastomosis was abandoned and this type of failure disappeared; anastomotic dysfunction disappeared also when we changed from running suture to uclips suture which provided a 100% patency. Nevertheless, we have observed three cases of LITA occlusion with an opened LAD anastomosis: one seemed to be related to a twist of the graft, the two others remained unexplained. However, the rate of graft failure in the TECAB procedure is acceptable in comparison with classical coronary surgery; routine intraoperative completion angiography in classical CABG demonstrated that 7% of LAD-LITA grafts had a significant defect: 3% in the conduit and 4% at the distal anastomosis [12]. In our study, comparison of the patency between groups was not relevant because systematic assessment was not done in all groups; but there is no question regarding the end-point of LAD reintervention at 3 months (PAC-CAB, 0%; MIDCAB, 1.8%; TECAB, 10%; p = 0.01). Nevertheless, there was no difference in mortality and survival between the three groups.

In all, reoperation for bleeding was high, specially in the TECAB group, demonstrating hemostasis is difficult in a minimally invasive environment and more in a closed chest procedure. In this series, from patients with an intent-to-treat TECAB procedure (n = 78), 24% had a conversion to MIDCAB procedure and from patients who underwent a TECAB procedure (n = 58), 7% had a thoracotomy during reoperation for bleeding. Nevertheless, hospital stay was significantly shorter in TECAB group.

There is no evidence in published data that on-pump TECAB (or Port-Access TECAB) procedure provides better results. In the multicenter European trial [11], there was no difference in 6-month freedom from MACE between on-pump and off-pump TECAB procedures; in the on-pump TECAB multicenter US trial [13], 3-month freedom from reintervention or angiography failure was 91% versus 90% in our series. Better results could be expected with the fourth arm Da Vinci system with the advantage of the robotic endostabilizer.

Our study has its limitations. Patients were not randomized and they were referred to one of the three MI techniques according to the evolution of the MI surgical concept in our team, to their condition and the quality of the target vessel. Inclusion in the MIDCAB group of conversions from an intent-to-treat TECAB procedure is also open to criticism; it was reasonable because all conversions except one were decided before the anastomosis stage and were mainly related to the quality of the LAD. All these bias contributed to include in the MIDCAB group the "worst" cases regarding target vessel quality, which did not have any impact on results, as good as in PA-CABG group and better than in TECAB group. An intention to treat analysis would provide the same results. A systematic post-operative assessment of LAD bypass was performed only after TECAB procedure and patency comparison between groups was not relevant. In any case, the correlation between LAD bypass failure and recurrent angina is well known; in this study, all patients with angina recurrence underwent coronary angiography and comparison between groups was focused on reintervention events.

In conclusion, our study has confirmed minimally invasive CABG is safe and effective. If PA-CABG and MICAB provide results as good as conventional CABG, TECAB procedure is associated with a higher early rate of bypass failure and reintervention. Beyond the post-operative period, results are equivalent and stable regardless the surgical technique performed. According to these results, PA-CABG was abandoned considering its cost effectiveness [8] and patients for LAD grafting are referred either to robotic-enhanced MIDCAB or off-pump TECAB, mainly according to the quality of the target; but in any case of doubt or technical difficulty we don't hesitate to convert before the anastomosis stage, an intent-to-treat TECAB procedure to a MIDCAB procedure which remains the reference procedure for minimally invasive LAD grafting.