Patients and Definitions
A review was undertaken of all patients who underwent myocardial revascularization at our institution between January 2009 and December 2016. Of these patients, those who had intervention for TVD were selected for analysis.
We used the same definitions for stroke and myocardial infarction (long-term) as the SYNTAX trial (1). However, for periprocedural myocardial infarction (short-term outcome) we only included Type 1 MI (using the Third Universal Definition of Myocardial Infarction) that occurred on index hospital admission, which explains our low event rate (11). New-onset renal failure was defined as a non-dialysis dependent patient requiring one or more hemodialysis sessions post-procedure. Target vessel revascularization was defined as a repeat intervention for a prior stented lesion, either within the stent itself or within 5 mm of the stent, and/or a repeat procedure for a lesion that was previously surgically bypassed. We also included major adverse cardiovascular events (MACE) as a long-term clinical endpoint. Our definition of MACE denotes a composite end-point encompassing death, repeat revascularization and/or myocardial infarction. We did not include major adverse cardiovascular and cerebrovascular evens (MACCE) as a long-term clinical outcome. Instead we included stroke as a short-term clinical endpoint, as we felt that stroke 30 days post-revascularization is unlikely related to the revascularization method. For patients that underwent HCR, length of stay included the total hospital stay for both the MIDCAB and PCI. Our preferred method is to perform a MIDCAB-first approach, followed by interval PCI, typically within 4 to 6 weeks of surgery. However, select situations require a PCI-first approach. Among the PCI cohort in which revascularization was often performed on separate admissions, we defined length of stay as the days admitted for all hospitalizations combined. All clinical events were adjudicated by an independent clinical event committee (involving a cardiothoracic surgeon and interventional cardiologist).
The relative contraindications to HCR were the need for emergency revascularization, and/or severe pulmonary disease rendering the patient unable to tolerate single-lung ventilation. Patients with high BMI have been regarded by some investigators as being unsuitable for HCR; we did not withhold HCR from these patients, and, indeed, our experience with minimally-invasive direct coronary artery bypass (MIDCAB) in this patient population has been reported previously (12).
We analyzed propensity-matched patients who underwent HCR, CABG, or multi-vessel PCI for triple-vessel disease. Any patients who required concomitant non-coronary surgery, in addition to their revascularization procedure, were also excluded, as were patients who required emergency or salvage intervention, as well as patients who had prior cardiac and/or thoracic surgery.