Biomarkers for prediction of cardiovascular events.

To the Editor: Wang et al. (Dec. 21 issue)1 suggest that novel biomarkers in aggregate have little influence on the prediction of first cardiovascular events or death. We wonder whether there were sufficient analyses to warrant this conclusion. As described, there were only 68 “major” cardiovascular events in women and 101 in men. In the crucial multimarker analyses, a quarter of these events were eliminated owing to the obligatory inclusion of the urinary albumin-to-creatinine ratio in all the multimarker scores. Would these scores have been more useful if urine biomarkers had been excluded and 100% of the events included? Second, it is surprising that a study of risk prediction would include prevalent cardiovascular disease, yet analyses that include stroke would not consider atrial fibrillation. Finally, the inclusion of heart failure and coronary insufficiency as “major” cardiovascular events is problematic, since they are not included in the more restrictive definition of “hard” events (myocardial infarction and death from coronary causes) outlined by the third report of the Adult Treatment Panel (ATP III) of the National Cholesterol Education Program. The ATP III definition is currently recommended to determine lipid treatment goals and aspirin use.2 The study by Wang et al. would be more useful for clinical practice if analyses were also performed according to the ATP III definition. Kiran Musunuru, M.D., Ph.D. Roger S. Blumenthal, M.D. Ciccarone Center Baltimore, MD 21287 [email protected]