BNP-Guided Therapy for Heart Failure

Clinicians continue to search for that one test, the one biomarker that will help them diagnose, prognosticate, and treat specific syndromes. Heart failure is undoubtedly one of those syndromes. All physicians who treat patients with heart failure seek a robust marker for this syndrome, and some may wonder, with the mounting evidence on brain natriuretic peptide (BNP), is such a biomarker finally here?

The current status of BNP in heart failure has been the result of an important historic journey. In 1981, de Bold et al¹ injected myocardial homogenates into nondiuretic rats. The atrial muscle extract increased sodium and chloride excretion 30-fold, along with an impressive increase in urine volume. By 1985, this same group had identified specific granules that secreted the peptide, now known as atrial natriuretic factor (ANF), and noted that the substance also had a hypotensive effect and an inhibitory action on renin and aldosterone secretion.² Thus, the heart was behaving as an endocrine organ. The investigators noted that the inability of the kidney to excrete sodium in chronic heart failure could be related to ANF and that the peptide might hold promise in the therapy for both hypertension and heart failure. Francis et al³ subsequently described the elevation of ANF early in heart failure and surprisingly in patients with asymptomatic left ventricular dysfunction. Several years later in 1988, Sudoh et al,⁴ from Japan, isolated an ANF-like peptide from mammalian brain tissue that had similar properties to ANF but was distinct in its amino acid sequence, hence the name brain natriuretic peptide. These investigators suggested that in human disease, both ANF and BNP might perhaps have a dual mechanistic action in sodium and volume homeostasis.

Since these discoveries, a tremendous amount of work has ensued to further the understanding of the family of natriuretic peptides, which is made up of at least 4 distinct entities, each with its own biological effect. N-terminal pro-BNP (76 amino acids) and BNP (32 amino acids) are the result of cleavage of the prohormone BNP.⁵ Prohormone BNP is secreted by the ventricles in response to hemodynamic stress, including dilatation and wall tension.⁶ Brain natriuretic peptide causes decreases in blood pressure by vasodilation, promotes diuresis and natriuresis, and reduces sympathetic nervous system activity as well as the activities of the renin-angiotensin system. Levels of BNP are increased in acute heart failure and in other conditions, such as pulmonary embolism, pulmonary hypertension, and older age.⁷ Brain natriuretic peptide is cleared by the kidney, so BNP levels also may be elevated in patients with renal dysfunction.

The ability to measure BNP as a rapid point-of-care test, and N-terminal pro-BNP in a laboratory assay, have stimulated interest in using this peptide as a marker of heart failure presence and severity. Studies consistently have established that BNP can be used in the emergency setting to differentiate causes of dyspnea^{8 - 9} and can help predict adverse outcomes including readmission when used at discharge following a hospitalization for heart failure.^{10 - 11} Levels of BNP at presentation with acute heart failure have a linear relationship to in-hospital mortality.¹² Therefore, BNP testing has taken its place among studies performed both in the emergency setting and at the time of hospital discharge and has been used as a prognostic tool to predict readmission and mortality.

What has been more contentious is BNP-guided monitoring of therapy to improve symptoms and ultimately change outcomes. There are current arguments in the cardiology community, and specifically in the heart failure community, about the utility and outcomes of treating elevated BNP levels. The variability and ranges of normal BNP levels by age, sex, and renal function, among other parameters, are poorly understood by the practicing physician. Whether to obtain point-of-care BNP testing or laboratory assay of N-terminal pro-BNP also is poorly understood in the clinical setting. Serial BNP levels are often ordered for patients hospitalized with heart failure, but without a definitive plan of action for high values. This leads to wasteful tests and increased costs. Poor transitions of care and lack of communication contribute to inpatient information not being transmitted to the outpatient clinician, thus rendering this important laboratory value unavailable for follow-up care. Clinicians need to know if and how to respond to elevated levels of BNP to improve heart failure symptoms and signs, and ultimately to improve survival and reduce morbidity.

The first step to address the question of appropriate use of BNP in the clinical setting was taken by Troughton et al¹³ in 69 patients primarily with New York Heart Association class II heart failure. Using N-terminal BNP levels (target level <200 pg/mL) to guide therapy, the authors reported a significant reduction in heart failure events including cardiovascular death plus hospitalization or heart failure decompensation. Subsequently, the larger multicenter Systolic Heart Failure Treatment Supported by BNP (STARS-BNP) trial¹⁴ reported that among patients who already were appropriately treated according to guideline-recommended heart failure therapy¹⁵ and also were randomized to receive BNP-guided strategy (target N-terminal BNP level <100 pg/mL), further increases in doses of angiotensin-converting enzyme inhibitors and β-blockers significantly reduced the combined end point of hospitalizations and death related to heart failure. The end point benefit was mainly driven by a reduction in heart failure hospitalizations. Preliminary data from 2 additional studies (pilot trial of BNP-guided therapy in patients with advanced heart failure [STARBRITE]¹⁶ and the BNP Assisted Treatment to Lessen Serial Cardiac Readmissions and Death [Battlescarred]¹⁷ ) also suggested improved outcome with BNP-guided therapy, although Battlescarred revealed benefit only in those younger than 75 years.

In this issue of JAMA, Pfisterer et al¹⁸ and the Trial of Intensified vs Standard Medical Therapy in Elderly Patients With Congestive Heart Failure (TIME-CHF) investigators have taken the next important step in a well-designed trial examining N-terminal BNP–guided vs symptom-guided heart failure therapy. The authors enrolled 499 patients aged 60 years or older (mean age, approximately 77 years, including 289 patients >75 years), the most rapidly increasing group of patients with heart failure, and expanded the end point to survival free of all-cause hospitalizations plus quality of life. The study strategy was symptom-guided therapy (as recommended by clinical practice guidelines) with a target of reduction in New York Heart Association class to II or less vs N-terminal BNP–guided therapy to a BNP level of less than 2 times the upper limit of normal and New York Heart Association class of II or less. The primary end point of 18-month survival free of all-cause hospitalization was not significantly different between the 2 groups (41% for symptom-guided therapy and 40% for N-terminal BNP–guided therapy; hazard ratio, 0.91 [95% confidence interval, 0.72-1.14]; P = .39) and included similar improvements in quality of life in both groups.

In secondary analyses, the N-terminal BNP–guided group experienced fewer hospitalizations for heart failure, a finding consistent with previous reports. However, a strong interaction with age was present and patients older than 75 years did not experience the same benefits of reduction in hospitalization for heart failure and in fact, had more adverse effects from uptitration. Pfisterer et al propose that N-terminal BNP–guided therapy may be harmful in those aged 75 years or older. There were, however, baseline differences in the older group that may identify a different pathophysiology and may have included patients with preserved systolic function heart failure—a different entity, indeed. Patients in the age group older than 75 years were more likely to be women and to have hypertension, higher ejection fraction, more comorbidities including atrial fibrillation, elevated creatinine levels, and prior stroke. Many of these comorbidities exist in patients with preserved systolic function heart failure.

Thus, there are lessons to be learned from the TIME-CHF study in the context of the 4 other guided trials. The desire to find the perfect biomarker that will point the way to better outcomes needs to be balanced by its application in the right population. In a well-treated group of patients with heart failure, as in the TIME-CHF study, there was no significant reduction in N-terminal BNP levels between the 2 strategy groups and yet in the N-terminal BNP–guided therapy group, medical therapy was further intensified to reach a target with subsequent reductions in hospitalizations for heart failure, an effect limited to those younger than 75 years. Medical therapy, therefore, can usually be further optimized and uptitrated even in the absence of worsening symptoms—an important clinical point. However, upward push of therapeutic doses may be only applicable to younger patients. Older patients may have a different syndrome of systolic dysfunction mixed with diastolic dysfunction. Therapy by symptoms alone may be insufficient to reduce hospitalization for heart failure, although for patients in the TIME-CHF study, quality of life improved. Using N-terminal BNP–guided therapy in addition to clinically based judgment and application of recommended doses of evidence-based care may have limited value in older patients despite its well-established diagnostic and prognostic value.

Thus, the totality of available information from the 5 trials is that the strategy of using N-terminal BNP–guided therapy appears safe in patients younger than 75 years (ie, no excess of hypotension, renal failure, or hyperkalemia) and some data suggest a modest reduction in mortality for some patients. Future trials need to prospectively analyze biomarker-guided therapy in a larger number of patients to ensure adequate power to definitively answer the question of whether this strategy reduces clinically important events including mortality, and should include sufficient numbers of older patients and women, as well as a detailed prespecified protocol for therapy choices and uptitration schedules. The detailed clinical plan should consider reduction of diuretics if no pulmonary or systemic congestion is present and include a flexible diuretic regimen as recommended in current practice guidelines.¹⁹ A detailed plan of approach to BNP levels will tend to minimize individual physician behavior as a confounder. Therefore, the TIME-CHF study places N-terminal BNP–guided therapy into perspective and introduces important caveats in the use of BNP levels in clinical practice.

In conclusion, the time course of heart failure therapy is gradual, composed of uptitration of medications, reassessment of patient symptoms and signs, clinician persistence and patience, and obtaining BNP levels. There are no easy answers and no simple solutions in the search for a single biomarker for diagnosis, prognosis, and treatment of heart failure. While the BNP level may prove to be a useful tool for guiding therapy, it may be the method of reduction of BNP levels that matters most in improving outcomes for patients with heart failure.