A study of the use of biomarkers to predict the risk of cardiovascular disease and death in an apparently healthy population has found that, even though some measurements are associated with future cardiovascular events, their usefulness for predicting risk in individuals may be limited. The report from the Framingham Heart Study appears in the Dec. 21 New England Journal of Medicine.
"We found that several contemporary biomarkers were associated with future cardiovascular disease or death, over and above what was indicated by established risk factors; but even in combination their utility for risk prediction was modest," says Thomas J. Wang, MD, of the Massachusetts General Hospital (MGH) Division of Cardiology, the report's lead author. "High biomarker levels can successfully identify groups of people at risk, but their ability to predict an individual person's risk - a goal of 'personalized medicine' - is still limited."
Many previous studies have identified potential biomarkers - laboratory measurements that may indicate a particular biological state - of cardiovascular risk. These include blood levels of C-reactive protein, B-type natriuretic peptide (BNP), and homocysteine, elevated levels of which have been associated with increased risk. However, few studies have looked at the use of multiple biomarkers, either to compare their usefulness or to evaluate the testing of several markers at once.
The current study was designed specifically to look at the ability of a multimarker testing approach to evaluate cardiovascular risk in a group of apparently healthy individuals. Participants were members of the Framingham Offspring Study, which follows a group of adult children of participants in the original Framingham Heart Study to evaluate risk factors for the development of cardiovascular disease. During participants' regular study visits between 1995 and 1998, they were tested for 10 potential biomarkers of cardiovascular risk, along with the usual history, physical examination and other assessments taken as part of the overall Offspring Study. Those known to have a prior heart attack or stroke were excluded from the biomarker study.
Among the 3,200 participants in the biomarker study, 169 experienced a major cardiovascular event - such as heart attack, stroke or cardiac death - during the study period of up to 10 years. The biomarkers that proved most useful in predicting future cardiovascular events were BNP and urinary albumin content. BNP appeared to be a stronger predictor of risk than C-reactive protein, possibly the best-known cardiovascular biomarker. However, even though those with high multimarker scores had twice the risk of a cardiovascular event as those with low scores, the information provided by the biomarkers only slightly improved predictions based on such conventional risk factors as hypertension, cholesterol levels and smoking.
"There has been a great deal of enthusiasm among cardiologists over the potential of biomarkers, but our findings suggest that we need to identify additional biomarkers to be able to predict individual risk in a useful fashion," says Wang. "Newer biological approaches, such as genomics and proteomics, may give us tools that can help identify these new biomarkers." Wang is an assistant professor of Medicine at Harvard Medical School.
"This important study shows that conventional risk factors have stood the test of time and are still good predictors of risk, while at the same time adding to the growing body of information on novel biomarkers," says Elizabeth G. Nabel, MD, director of the National Heart, Lung and Blood Institute (NHLBI), which sponsors the Framingham Heart Study. "Although there was only a modest improvement in risk assessment when new risk factors were added to conventional ones, it was enough to show the potential value of the newer markers. We look to further research to identify other biomarkers that are more predictive of risk, which could also have implications for the development of new treatments for cardiovascular disease."