Assessing CVD Risk in the Elderly
Study Supports Expanded Utility of NT-proBNP
By Bill Malone
Controversy has surrounded the extent to which N-terminal pro-B-type natriuretic peptide (NT-proBNP) provides additional guidance to clinicians beyond assessment of patients presenting with shortness of breath. A new study described in this issue of Strategies adds to previous research that NT-proBNP is associated with long-term cardiovascular disease (CVD) outcomes and also finds that the test independently predicts heart failure and CVD death in older adults.
At least 80% of cardiovascular deaths occur in the elderly, and complicating the matter further, subclinical CVD is common in this population as well. But clinicians have for decades faced a dilemma when it comes to predicting heart failure (HF) and cardiovascular death: even though Framingham and later studies identified parameters like lipoproteins, smoking, diabetes and other risk factors in CVD, for the elderly in whom heart failure is most prevalent, these risk factors lose much of their power.
Among other biomarkers, NT-proBNP has been proposed as a leading candidate for closing the gap in clinicians’ ability to predict HF, especially in older patients. Now a recent study from researchers at the University of Maryland School of Medicine adds weight to the argument for an expanded role for the protein, not only for independently predicting HF, but also for monitoring the dynamic changes in HF risk over time (J Am Coll Cardiol 2010;55:441-450).
The investigators measured NT-proBNP at baseline and 2 or 3 years later in 2,975 community-dwelling adults free of HF who were participants in the ongoing Cardiovascular Health Study (CHS). Funded by the National Heart, Lung and Blood Institute (NHLBI), CHS is a multi-center, prospective, observational study of risk factors for CVD in adults 65 years or older. The new study measured NT-proBNP in serum collected at baseline for the main and supplemental CHS trial cohorts, then performed a second measurement on serum collected 3 years later for the main CHS cohort and 2 years later for the supplemental cohort. Measurements were performed on the Elecsys 2010 system from Roche Diagnostics.
After adjusting for potential cofounders, the investigators found that NT-proBNP levels in the highest quintile of the cohort were independently associated with greater risks of HF and CVD death compared with the lowest quintile. In addition, participants with an initially low NT-proBNP level who had a >25% increase in the follow-up measurement were at higher risk of HF and CVD death compared with participants with continuous low levels. Similarly, those with high levels who showed a decrease upon the follow-up measurement had reduced risk. “The unique finding is that, even when adjusted for everything else a patient might have—risk factors like diabetes or coronary artery disease—a change in NT-proBNP portends a poor prognosis for these asymptomatic, elderly individuals,” said lead author Christopher deFilippi, MD, associate professor of medicine in the department of cardiology at the University of Maryland School of Medicine. “We also showed that, for the elderly, NT-proBNP provides prognostic information above and beyond all the traditional risk factors, something that has been controversial in the general population.”
HF is quickly becoming the predominant cardiac pathology for those 65 and older, noted deFilippi. Typically, once patients have developed HF, the outcomes are poor and the disease doesn’t respond well to established medical therapies. In 2005 the American Heart Association (AHA) developed guidelines for HF that include four stages. For stage A, an individual is asymptomatic but with risk factors and the absence of evidence of underlying structural heart disease. Stage B is asymptomatic but with evidence of structural disease, for example, left ventricular hypertrophy or depressed left ventricular function. State C is symptomatic heart failure, and with stage D, there are marked symptoms at rest that are refractory to treatment without special interventions. “We were intrigued by the concept that we might be able to use a biomarker to track the trajectory of change in stage,” said deFilippi. “We thought that, particularly if you could identify those stage A people who are on their way to stage B, and particularly stage C, that might be an opportunity for considering some earlier interventions.”
Significantly, deFilippi and his colleagues point out that in their study asymptomatic individuals with a significant rise in NT-proBNP went on to experience relatively high incidences of HF. Patients who started with measurements > 190 pg/mL and whose levels rose > 25% in the interval experienced a risk of HF that was about 8% per year. In contrast, those who had high levels initially and that then declined only experienced a risk slightly more than 2%—almost a four-fold difference in risk.
deFilippi proposed several reasons why NT-proBNP should turn out to be such a useful marker. It could be that NT-proBNP is driven in a large part by elevations in pressure within the heart, but other cardiovascular disease factors completely independent of pressure have been linked to elevations in NT-proBNP, like ischemic heart disease and cardiac fibrosis. “These are two factors that might be much more important in the elderly who are asymptomatic,” he explained. “Perhaps some people develop more cardiac fibrosis than others as they age, and higher NT-proBNP over time might show that kind of underlying pathology. Eventually, it reaches a point where the filling pressures in the heart begin to increase, NT-proBNP goes even higher, and the patient becomes symptomatic.”
In fact, deFilippi sees NT-proBNP as more of a direct measure of cardiac health, with the traditional risk factors as being more indirect. However, pushing NT-proBNP into use as a screening test or a prognostic marker leads to two important questions, he said. “We show that it’s a very robust predictor, and it’s probably worth checking twice, because even those people who start low, there is a substantial percentage who will ultimately rise by the time they have it drawn on a second occasion. So it is a good screening test, and it would identify people at risk, but the question is, what are you going to do about that? Is it going to be helpful to know that they’re going to develop heart failure?”
In the U.K., there is a precedence for expanded use of NT-proBNP. There, the National Health Service has officially made NT-proBNP available to assist primary care clinicians to refer patients more appropriately for echocardiograms. But the U.K. healthcare system has not gone so far as to make it a routine screening test.
Still, NT-proBNP has quite a few things going for it, deFilippi contended. “The test is easy to perform and not too expensive,” he said. “However, people are inundated with biomarkers and tests in general, even though many of us think this one has potential. It’s hard to say how these thing catch on. When troponin came along, there were some pretty skeptical people for six or seven years. Old habits die pretty hard.”