Abstract
BACKGROUND: Coronary artery bypass grafting is the standard of care for patients with obstructive left main (LM) coronary disease. In poor surgical candidates, high-risk percutaneous coronary artery intervention (PCI) is an alternative. METHODS: We investigated a retrospective cohort of patients who underwent LM PCI from January 2010 to March 2014 (n = 89). Obstructive LM disease was defined as 50% angiographic obstruction of luminal flow, and the primary endpoint was inhospital mortality. Ventricular assist device (VAD) was defined as the use of either intra-aortic balloon pump (IABP) or Impella 2.5 devices before, during, or following PCI. RESULTS: A total of 89 patients with LM PCI were divided into those with (n = 39) and without (n = 50) VAD support. The former group was further divided into those with support from either Impella 2.5 (n = 28) or IABP (n = 11). Age, race, and gender did not differ between patients who received unassisted LM-PCI from those with VAD support (P = 0.142, 1.0, and 0.776, respectively). The angiographic stenosis of atherosclerotic lesions in LM, proximal left anterior descending artery, and other native/surgical coronary vessels was similar between the groups. The duration of hospitalization was significantly longer for patients with VAD support compared to those without (7.19 ± 6.89 vs. 2.78 ± 3.39, P < 0.001). The incidence of cardiogenic shock and inhospital mortality was significantly higher in the VAD group (P = 0.009 and 0.001, respectively). Overall, inhospital mortality was 9% (8/89). The IABP and Impella 2.5 groups had mortality proportions of 46% (5/11) and 11% (3/28), respectively; P = 0.028. For all patients, inhospital mortality was higher for those with versus without cardiogenic shock (56% or 5/9 vs. 4% or 3/80; P < 0.001), and for those with versus without left ventricular systolic function <40% (17% vs. 2%;P < 0.025). CONCLUSION: In a selected group of patients with LM disease, unsupported PCI appears to be a feasible and safe procedure. In high-risk patients, the use of Impella 2.5 appears to be superior to IABP in LM PCI resulting in favorable short-term outcomes.