The predictive value of s-cystatin C for mortality after coronary artery bypass surgery.

Abstract

To evaluate serum creatinine (s-creatinine) and serum cystatin C (s-cystatin C) levels and estimated glomerular filtration rate (eGFR) at different time points as predictors for mortality in patients undergoing coronary artery bypass grafting (CABG). A total of 1638 patients undergoing elective CABG were studied prospectively over a median follow-up of 3.5 years (range, 2.0-5.0 years). Renal function was assessed by a comparison of s-creatinine, s-cystatin C values measured preoperatively and at the lowest postoperative level of renal function. The eGFR was estimated by different formulas: Modification of Diet in Renal Disease, the 2009 Chronic Kidney Disease Epidemiology (CDK-EPI) for s-creatinine, the 2012 CKD-EPI formula for s-cystatin C, the 2012 CKD-EPI formula for s-cystatin C and s-creatinine in combination, and the Caucasian Asian, Pediatric, and Adult subjects formula for s-cystatin C. Cox proportional hazards model analysis and C-statistics were used to evaluate independent predictors of mortality and to assess the predictive ability of the different renal function measures. The 30-day mortality was 0.8%. Overall survival was 96.1% ± 0.4% at 2 years and 90.0% ± 1.2% at 5 years. Preoperative s-cystatin C showed greater predictive power than s-creatinine for overall mortality (area under the curve, 0.794 vs 0.653). Preoperative s-cystatin C (hazard ratio [HR], 1.65; 95% confidence interval [CI], 1.36-1.99) and eGFR based on s-cystatin C (HR, 0.96; 95% CI, 0.95-0.98), were both independent predictors of mortality. The unadjusted HR for mortality comparing the lowest preoperative cystatin C quintile (Q1) with Q4-Q5 were as follows: Q1 versus Q5, HR, 2.0; 95% CI, 1.6-2.5 (P < .001); Q1 versus Q4, HR, 1.6; 95% CI, 1.2-2.2 (P = .005). The s-cystatin C level and s-cystatin C-based eGFR measured preoperatively are strong predictors for mortality after elective CABG.