Abstract 185: Comparative Efficacy of Revascularization Procedures for Severe CAD in Patients with Chronic Kidney Disease.

Introduction: Patients with Chronic Kidney Disease (CKD) are at increased risk for cardiovascular mortality. However, there are few comparative efficacy studies of coronary revascularization procedures in this high-risk population. Hypothesis: There is an association between treatment strategy, estimated glomerular filtration rate (eGFR), and survival among patients with CKD and coronary artery disease (CAD). Methods: We examined patients in the Duke Databank for Cardiovascular Diseases (N=8604) who had significant CAD by cardiac catheterization from 2003-2009. Patients were excluded if they had prior revascularization or if pre-procedure serum creatinine was missing. Patients were stratified by estimated pre-cath estimated glomerular filtration rate (eGFR) into 5 groups: normal (eGFR ≥ 60 ml/min/1.73m 2 ), mild CKD (eGFR 45-59 ml/min/1.73m 2 ), moderate CKD (eGFR 30-44 ml/min/1.73m 2 ), severe CKD (eGFR <30 ml/min/1.73m 2 ) and ESRD (dialysis). Multi-variable Cox Proportional Hazards models were used to determine the association between CKD severity and mortality according to four possible treatment strategies: coronary artery bypass-graft surgery (CABG), PCI with drug eluting stent (DES), PCI with bare metal stent (BMS), or medical management (MED). In addition, we tested for interactions between treatment strategy and eGFR groups. Results: The sample was mostly male (64.9%, 5591/8604), median age was 62 y (IQR 54-71), and 28.8% (2476/8604) had diabetes. Median follow up was 4.0 y (IQR 2.0-6.1). After adjusting for 17 covariates, baseline eGFR was significantly associated with mortality. For all treatment strategies, there was an inverse association between mortality and each 10ml/min drop in baseline kidney function: MED HR = 1.185 (1.146-1.225), BMS HR = 1.246 (1.190-1.303), DES HR = 1.294 (1.244-1.347), and CABG HR = 1.262 (1.204-1.323). Multivariable models comparing treatment strategies demonstrated differential effects by kidney function (Table; interaction for treatment effect across eGFR groups, p=0.0053). In contrast to BMS, DES and CABG were associated with similar mortality reductions among groups with lower eGFR. Conclusions: There are strong associations between eGFR and increased mortality among patients with documented CAD, however, these vary significantly by revascularization strategy. These data suggest that a randomized trial of revascularization strategies in patients with CKD may be warranted to guide clinical practice.