Background: The aim of this present prospective study was to investigate the accuracy of cardiac markers for the prediction of subsequent cardiac events (cardiac death, acute myocardial infarction and recurrent ischemia requiring coronary revascularization).
Methods: Fibrinogen, cardiac troponin T, troponin I, creatine phosphokinase myocardial fraction, C-reactive protein and myoglobin at baseline and after 6 h were measured on 154 patients (109 male, 63+/-11 years) with chest pain. Receiver operator characteristic analyses were performed to determine cut-off points of cardiac markers in prediction of adverse events.
Results: The following cut-off values for prediction of cardiac events were calculated: troponin I at baseline 0.3 ng/ml (predictive accuracy=0.870), troponin I at 6 h 0.50 ng/ml (p.a.=0.909); troponin T at baseline 0.05 ng/ml (p.a.=0.643), troponin T at 6 h 0.05 ng/ml (p.a.=0.612), creatine phosphokinase myocardial fraction at baseline 2.0 ng/ml (p.a.=0.721), creatine phosphokinase myocardial fraction at 6 h 2.5 ng/ml (p.a.=0.734), myoglobin at baseline 23 ng/ml (p.a.=0.623), myoglobin at 6 h 26 ng/ml (p.a.=0.617), C-reactive protein at baseline 0.31 mg/dl (p.a.=0.662), C-reactive protein at 6 h 0.55 mg/dl (p.a.=0.682), and fibrinogen at baseline 360 mg/dl (p.a.=0.701). The combination of baseline troponin I with different parameters resulted in a higher sensitivity of up to 98%, with a similar predictive accuracy, but a lower specificity. Additive measurements of cardiac troponin I at 6 h to baseline cardiac troponin T and I proved to be the best combination for prediction of subsequent cardiac events.
Conclusions: Changes in cut-off levels of cardiac markers and inflammatory parameters results in a high accuracy of risk stratification in patients with chest pains. Combination of these measurements might further help in the identification of patients who would benefit from early coronary revascularization.