Natriuretic Peptides and Cardiovascular Events: Title and subTitle BreakMore Than a Stretch

In patients with known or suspected cardiovascular disease, biomarkers are increasingly being explored as indicators of disease presence, severity, or activity; prognosis; and therapeutic efficacy. A biomarker may be a by-product of the disease state and may also directly participate in its pathogenesis or modulation. The cardiac natriuretic peptides (secreted in response to myocardial stretch) promote vasodilation and natriuresis and are believed to mitigate adverse cardiovascular remodeling. Circulating levels of these peptides have proved useful in gauging the severity of heart failure and in assessing the differential diagnosis of acute dyspnea.¹ N-terminal fragment of the prohormone brain-type natriuretic peptide (NT-proBNP)² is an inactive cleavage product with much slower clearance than the biologically active BNP.³ Its plasma concentration, therefore, principally reflects myocardial secretion over a prolonged time.

Prior investigations have identified plasma BNP and NT-proBNP levels as independent predictors of mortality or cardiovascular events in populations with chronic heart failure,^{4 - 5} acute coronary syndromes,^{6 - 7} prior myocardial infarction,^{8 - 9} established vascular disease or elevated coronary risk,^{10 - 11} and in community-based samples.^{12 - 13} In this issue of JAMA, the study by Bibbins-Domingo and colleagues¹⁴ extends these observations within a population of patients with stable coronary artery disease (CAD) and further clarifies the utility of NT-proBNP level as a predictor of a variety of subsequent cardiovascular events. The authors observe that plasma NT-proBNP concentration remains an independent predictor of all-cause mortality, CAD mortality, stroke, and events related to heart failure, even after sequentially adjusting for known measures of cardiac risk, left ventricular remodeling and function, inducible ischemia, functional capacity, inflammation, myocyte necrosis, and symptoms. Furthermore, plasma NT-proBNP concentration approaches significance as an independent predictor of myocardial infarction. These findings challenge the conventional view of the natriuretic peptide levels as mere hemodynamic markers of worsening heart failure or left ventricular dysfunction. Furthermore, these findings provoke reconsideration of the value of these markers as diagnostic tools and potentially as triggers for therapeutic intervention.

How can a marker of myocardial stretch, measured at a single time point within a stable population, predict such a wide variety of cardiovascular events? Levels of NT-proBNP may reflect a variety of adverse factors, all potentially resulting in increased hormonal release from stretched myocardium. However, to the extent that levels denote myocardial stretch, it seems likely that their value in predicting cardiovascular events would diminish after accounting for numerous known biochemical, echocardiographic, and functional markers of myocardial injury, ventricular remodeling, and active or inducible ischemia. The value of NT-proBNP in predicting stroke, also observed for BNP in an analysis of the Framingham cohort,¹² could be linked to its role as a marker of cardiac dilation with a predisposition to atrial fibrillation. However, the preserved predictive value for stroke after correction for multiple covariates, including echocardiographic measures of left ventricular function, supports consideration of alternative mechanisms responsible for the linkage between NT-proBNP levels and vascular events, including stroke.

Along with myocardial stretch, a variety of other factors have been found to stimulate or enhance secretion of BNP in vitro, including myocardial ischemia¹⁵ and paracrine or endocrine agents such as endothelin A,¹⁶ angiotensin II,¹⁶ and tumor necrosis factor α.¹⁷ Cardiac fibroblasts, as well as cardiac myocytes, secrete BNP, and its secretion modulates fibrosis through induction of matrix metalloproteinases.¹⁷ These observations suggest that the predictive value of plasma NT-proBNP concentration, particularly in a stable population without clinical heart failure, results not merely from its relationship with acute myocardial stretch but to its interdependence with local and circulating factors that drive inflammation, fibrosis, and hypertrophy throughout the cardiovascular system.

In patients with essential hypertension, circulating levels of natriuretic peptides correlate with the magnitude and pattern of left ventricular hypertrophy.¹⁸ Assuming that the known correlation of left ventricular hypertrophy with subsequent cardiovascular events is derived at least in part through common pathways driving both left ventricular hypertrophy and vascular pathology, one hypothesis is that the greater frequency of adverse outcomes seen in patients with increased plasma levels of natriuretic peptides is linked to an association between these increased levels and the extent of vascular disease. In patient populations presenting with signs or symptoms of heart failure, the distribution of values for natriuretic peptides is considerably higher than levels in the population studied by Bibbins-Domingo et al.¹⁴ Levels at the higher end of this range likely denote more severe hemodynamic derangement. In contrast, within a patient population with stable CAD without overt heart failure, more modest increases in NT-proBNP may convey increased risk of subsequent cardiovascular events through an association between myocardial hypertrophy and fibrosis (driving the increases in NT-proBNP) and a greater burden of atherosclerotic and hypertensive vascular disease.

Bibbins-Domingo et al¹⁴ suggest that NT-proBNP may be used to screen for patients to be considered for therapeutic intervention. Before accepting this proposal, it is important to remember that use of a biomarker for triggering or guiding therapy requires more than the demonstration of its correlation with worse clinical outcomes. Blood pressure and low-density lipoprotein cholesterol represent markers that are linked directly to the pathology that drives adverse outcomes. Clinical trial demonstration of a therapeutic impact on either of these markers has consistently been associated with improvement in cardiovascular outcomes. Likewise, clinical trial evidence also should be required to demonstrate that matching a treatment strategy to an elevated natriuretic peptide level improves clinical outcomes.

Unlike low-density lipoprotein cholesterol and blood pressure, the natriuretic peptides predominantly mitigate disease rather than drive it. Nevertheless, as disease markers, assays for natriuretic peptides might represent instruments to aid decision making in initiation and guidance of treatment. Troughton et al¹⁹ performed a pilot investigation in which patients with symptomatic heart failure and reduced left ventricular ejection fraction (LVEF) were randomized to receive medical therapy guided by serial measurements of NT-proBNP levels or by standardized clinical assessment. During a median follow-up of 9.5 months, patients in the NT-proBNP group had fewer cardiovascular events (death, heart failure hospitalization, or heart failure decompensation). Attempts are under way to confirm these findings in a larger-scale investigation, with the primary end point of all-cause mortality.²⁰

Despite the predictive value of NT-proBNP, the clinical utility of screening patients with stable CAD is uncertain. Based on available clinical trial evidence, treatment with angiotensin-converting enzyme inhibitors, aspirin, and statins are already recommended in most of these patients. β-Blockers are recommended for those with prior myocardial infarction or with reduced LVEF based on echocardiography, a test routinely performed in most patients with known or suspected CAD.

Reaching beyond patients with known disease, Heidenreich et al²¹ modeled the cost-effectiveness of a strategy of screening asymptomatic populations using BNP levels, with elevated levels triggering echocardiography to identify patients with reduced LVEF. Assuming that treating such patients with angiotensin-converting enzyme inhibitors would reduce rates of hospitalization and death, Heidenreich et al²¹ determined that screening with BNP levels would be cost-effective for populations with a 1% prevalence of reduced LVEF (eg, men aged 60 years). Attractiveness of this approach is linked to assumptions regarding the receiver operating characteristic curves of the BNP assay for detecting left ventricular dysfunction, which have been found to be suboptimal within a community-based cohort.²² However, the independent value of plasma natriuretic peptide levels in predicting a variety of cardiovascular events within such cohorts^{12 - 13} broadens the potential justification for population screening, representing an opportunity for clinical strategies directed at preventing adverse outcomes.²³ These and the current findings by Bibbins-Domingo et al¹⁴ justify considering patients with elevated plasma levels of natriuretic peptides in the absence of clinical heart failure for the investigation of specific available or novel interventions designed to prevent progression of cardiovascular disease.

Strong evidence now supports the value of plasma levels of natriuretic peptides in predicting a variety of subsequent cardiovascular events in patients with CAD, independent of other known markers of disease severity and risk. Although the relationship between plasma NT-proBNP and cardiovascular risk may be driven partly through myocardial stretch, it is likely that circulating levels of natriuretic peptides are also influenced by left ventricular mass and fibrosis, myocardial ischemia, and systemic and local levels of neurohormones and inflammatory cytokines. Further research is warranted to determine whether plasma natriuretic peptide levels correlate with the degree and extent of vascular disease. Additional investigation is warranted to determine the value of such levels in driving or monitoring interventions in patients with stable atherosclerotic cardiovascular disease and those without overt disease, as is under way in patients with clinical heart failure.