High-sensitivity cardiac troponin for risk prediction in patients with and without coronary heart disease

Background: In stable patients with unknown coronary anatomy, higher levels of cardiac troponin are associated with an increased risk of cardiovascular events. It was supposed that this association might be explained by the ability of cardiac troponin to detect minor myocardial necrosis which might be caused by subclinical coronary atherosclerosis. Thus, this analysis tested if the predictive value of high-sensitivity troponin T (hsTnT) in stable patients is dependent of the presence or absence of angiographically documented coronary heart disease. Methods: Stable patients undergoing elective coronary angiography were enrolled (n = 2046). HsTnT was determined before diagnostic procedures. The patients were followed for up to seven years. Primary endpoint was all-cause mortality or non-fatal myocardial infarction. All endpoints were adjudicated by independent physicians. Results were adjusted to a clinical model including independent clinical predictors of the primary endpoint. Results: Out of the 2046 patients enrolled, 1236 (60%) had a diagnosis of obstructive coronary heart disease. HsTnT predicted independently the primary endpoint (adjusted HR 1.33, 95%-CI 1.21-1.46, P < 0.001). The use of hsTnT in addition to the clinical model significantly improved discrimination (c-statistic: 0.751 to 0.773, P < 0.001) as well as reclassification of the primary endpoint (NRI = 0.362, P < 0.001). This significant improvement persisted across various subsets and was independent of the presence of clinically detectable coronary heart disease and other variables. Conclusion: The use of hsTnT in addition to clinical variables significantly improves discrimination and reclassification of patients with respect to all-cause mortality or non-fatal myocardial infarction irrespective of the presence of clinically detectable coronary heart disease. (C) 2014 Elsevier Ireland Ltd. All rights reserved.