To the Editor: Dr Melander and colleagues1 studied the use of novel and conventional biomarkers for prediction of cardiovascular events. Although they concluded that the gains of selected biomarkers over conventional risk factors are minimal, an important strategy has not yet been fully explored.
When assessing the value of a biomarker within a well-calibrated system, the C statistic of the receiver operating characteristic curve is used to evaluate its potential incremental benefit over other risk factors or panels. In the study by Melander et al, as well as in past studies,2 the results of this approach have been disappointing because of the very modest improvement compared with most traditional risk markers. The more correlated a novel marker is with traditional risk factors, the less its addition may improve discrimination.3
Rosenstein RS, Parra D. Biomarkers for Prediction of Cardiovascular Events. JAMA. 2009;302(19):2089-2090. doi:10.1001/jama.2009.1637