Impact of Renal Function in Patients with Multi-Vessel Coronary Disease (MVD) on Long Term Mortality Following Coronary Artery Bypass Grafting (CABG) Compared with Percutaneous Coronary Intervention (PCI)

Abstract

Background: Renal failure is an important predictor of mortality in patients with MVD who undergo coronary revascularisation. We sought to determine if the degree of renal impairment affects long-term mortality based on choice of revascularisation strategy. Method: We analysed 7841 patients with MVD undergoing either CABG (n = 6,739) or PCI (n = 1,102) between 2004 and 2008, enrolled in two large multi-centre Australian registries. Patients were assigned to three groups using stratified propensity matching based on their estimated glomerular filtration rate (eGFR) at baseline; ≥60 (CABG = 4674 vs. PCI = 839), 30–59 (CABG = 1799 vs. PCI = 226) and <30 ml/min/1.73 m2 (CABG = 266 vs. PCI = 37) respectively. Shock, myocardial infarction (MI) <24 h, previous CABG, valve surgery or PCI were exclusions. We compared Cox-proportional hazards-adjusted National Death Index-linked long-term mortality (mean 3.2 years). Results: In patients with eGFR30-59 and ≥60 ml/min/1.73 cm2, there were more women, octogenarians and recent MI in the PCI group and a higher prevalence of cerebrovascular disease, peripheral vascular disease, prior heart failure and MI in the CABG group. Observed long-term mortality rates (CABG vs. PCI) were 4.8% vs. 4.3% p = 0.50, 11.3% vs. 17.3% p = 0.009, 19.9% vs. 40.5% p = 0.005 in the three strata, respectively. Following adjustment, patients with eGFR ≥ 60 ml/min/1.73 cm2 had no significant difference in long-term mortality. However, with eGFR30-59 ml/min/1.73 cm2, PCI was an independent predictor of long-term mortality (HR1.55, 95%CI 1.07–2.25, p = 0.02). For eGFR < 30 ml/min/1.73 cm2, there was a trend towards higher mortality with PCI (HR1.80, 95%CI 0.95–3.41, p = 0.07). Conclusion: In this stratified, propensity-matched study there was a long-term mortality hazard associated with PCI for patients with eGFR30-59 ml/min/1.73 m2. The effect may be higher for eGFR < 30 ml/min/1.73 m2.