Risk stratification in stable coronary artery disease

Abstract

The aim of the study was to compare N-terminal pro B-type natriuretic peptide (NT-proBNP), high-sensitivity C-reactive protein, and gamma-glutamyl transferase (γ-GT) with traditional risk markers for estimating prognosis in patients with stable coronary artery disease (CAD). Evaluation of mortality and a combined clinical endpoint (mortality, need for coronary revascularization, myocardial infarction, hospitalization for cardiac causes, or stroke) during an average 3.2-year follow-up in 394 consecutive patients (73% male patients, age: 67±9 years) with angiographically proven stable CAD. Univariate Kaplan-Meier survival rate analysis showed that traditional risk markers, apart from impaired renal function, three-vessel CAD, and a reduced left ventricular function at the time of coronary angiography, were not of prognostic relevance for prediction of outcome. NT-proBNP, high-sensitivity C-reactive protein, and gamma-glutamyl transferase were significant predictors of mortality; however, only NT-proBNP was a significant predictor of the combined endpoint. In age-adjusted and sex-adjusted multivariate Cox regression analysis, NT-proBNP was the strongest independent predictor of the combined endpoint (odds ratio 2.92, 95% confidence interval: 1.72-4.94, first vs. third tertile). All three laboratory parameters remained independent risk markers for mortality in multivariate analysis. NT-proBNP, however, revealed the highest odds ratio (5.23, 95% confidence interval: 1.17-23.23, first vs. third tertile). Concentrations greater than 356 ng/l predicted mortality with a sensitivity of 70%, a specificity of 71%, a positive likelihood ratio of 2.4, and a negative likelihood ratio of 0.42. In comparison with other tested novel biomarkers and traditional risk markers, NT-proBNP was the most predictive prognostic marker in multivariate analysis in patients with stable CAD.