The natriuretic peptides are a family of ring shaped vasoactive hormones showing considerable sequence homology. Four natriuretic peptides have been described, named A-D type (see Figure 1). Of these, commercial immunoassays have been developed for B type natriuretic peptide (BNP) and the N terminus of the prohormone, N terminal pro BNP (NTproBNP). There has been an explosion of clinical studies on the role and clinical application of BNP/NTproBNP, which will be the subject of this article.
BNP was originally discovered in the porcine brain, where it was thought to be a neurotransmitter, hence its original name - brain natriuretic peptide. There is a much larger concentration in the ventricles of the heart, hence the current term, B type natriuretic peptide. There appears to be little storage of BNP in the ventricle, the main source. BNP is produced by direct synthesis in response to the degree of ventricular stretch, so acts as a marker of ventricular dilatation. The messenger RNA for proBNP is unstable so there is active regulation of BNP levels according to ventricular wall tension. The initial transcription product contains a leader sequence, which is cleaved to yield proBNP (see Figure 2). This is then cleaved either on secretion or after secretion (exactly when is not clear) to NTproBNP and BNP.
There are two receptors, natriuretic peptide receptor-A (NPR-A) and -B (NPR-B), which have an extracellular binding site, a transmembrane domain and an intracellular domain with a protein kinase-like region and a guanyl cyclase site. The intact ring structure is essential for receptor binding. Receptor binding results in generation of cyclic guanosine monophosphate (GMP) as second messenger. The half-life of BNP is short (20 minutes) and NTpBNP is 60 minutes under normal conditions. BNP binds to a third receptor, NPR-C, that clears it from the blood by binding, endocytosis and lysosomal degradation. The ring structure is also cleaved by a membrane-bound neutral endopeptidase, found in the kidneys and vascular tree, which inactivates the molecule.
Clinical Role of Natriuretic Peptides
The natriuretic peptides are relatively recently described although studies have indicated their potential clinical value. The advent of commercially available immunoassay systems has propelled BNP measurement from research applications to mainstream clinical practice. Current analytical techniques are robust and precise.1-3 Clinical studies have shown that there is little difference between BP or NTproBNP measurement for the majority of clinical situations. The major difference between the two molecules is the stability with NTproBNP, as would be expected, which is much more stable.4 The major clinical application of BNP/NTproBNP measurement is discussed in the following sections.
Detection of Ventricular Dysfunction (Cardiac Failure)
Cardiac failure is a major public health burden and is predicted to be the largest developing single health problem in the EU.5,6 The European Cardiac Society (ECS) has estimated that the prevalence is 0.2% to 4% of the population, some 10 million in the ECS area. Accurate diagnosis of cardiac failure is important - the morbidity and mortality of cardiac failure is high. In women, the mortality of cardiac failure exceeds that of most cancers apart from lung cancer. The median survival is from three to four years. There are effective therapies including diuretics, angiotensin converting enzyme (ACE) inhibitors, angiotensin 2 receptor blockers and beta-blockers such as carvedilol and bisoprolol. All of these agents are effective, improving quality of life and survival. Treatment with ACE inhibitors, to reduce inappropriate cardiac remodelling, is most effective when introduced early.
Clinical assessment of suspected heart failure (HF) has a sensitivity of 33% to 55%. The electrocardiogram (ECG) and chest radiograph have diagnostic sensitivities in the range of 55% to 65%. Detection of the underlying pathophysiology is by cardiac imaging, echocardiography, radionucleide ventriculography or cardiac magnetic resonance imaging (CMRI). Echocardiography is routinely available in most hospitals but there is often inadequate service provision and HF due to diastolic dysfunction (impaired cardiac relaxation and dilation during ventricular filling), or restrictive cardiomyopathy (as occurs in amyloidosis, sarcoidosis and other infiltrative conditions) may be difficult to assess by this technique. Radionucleide ventriculography requires radioactive isotopes and is not widely available. Similarly CMRI, although free from ionising radiation, requires expensive equipment and is not widely available.
Detection of Chronic Ventricular Dysfunction
Measurement of BNP/NTproBNP can be used for the initial assessment of patients prior to referral to secondary care. Initial studies of BNP measurement in referrals to HF clinics7,8 have been followed by studies of the diagnostic role in primary care. Measurement of NTpBNP by microtitre enzyme-linked immunosorbent assay (ELISA) compared with a clinical diagnosis was evaluated by receiver operator characteristic (ROC) curve analysis in different populations in primary care (see Table 1).9 The area under the curve (AUC) was consistently greater than 0.85 and the excellent negative predictive value of the test. This study has been confirmed subsequently in a study in 672 subjects in primary care in Copenhagen where again the AUC was 0.94.
More importantly, a NTproBNP below the median predicted mortality and admissions for HF. The role of BNP as an outcome predictor was explored in a Framingham offspring cohort study.10 Here, a BNP in the upper tertile was one of the most powerful predictors of cardiovascular events, death and HF.
A number of observational studies have examined the role of BNP measurement in monitoring treatment in congestive cardiac failure (CCF). The responses to the beta-blocker carvedilol11 and to the angiotensin 2 antagonist valsartan have been shown to be predicated by BNP levels.12,13 There has been one small randomised controlled trial (RCT) that compared therapeutic decision-making by clinical assessment of the patient state (n=36) with therapy adjustment based on BNP measurements (n=33).14 It was found that when BNP was used to titrate therapy, there was a significant improvement in outcome (19 compared with 54 events, BNP guided compared with clinical). Further studies are currently on-going in this area.
BNP measurements can be used for the diagnosis of acute cardiac failure. The Breathing Not Properly study enrolled 1,586 acute presentations with breathlessness.15 BNP was measured on admission using the Biosite triage. Patients were categorised into those with or without disease by an adjudicated final diagnosis of cardiac failure by two cardiologists who utilised all clinical investigations, except BNP. The diagnostic performance of BNP was then analysed by ROC curve analysis. The AUC was 0.91 (0.90-0.93). A subsequent prospective RCT compared acute admissions, randomised to BNP measurement on the same system, with usual care.16 Patients who had BNP measured had a shorter stay with fewer intensive treatment unit (ITU) admissions and at lower total cost with no difference in outcome.
BNP/NTpBNP in Acute Coronary Syndromes
Impaired myocardial perfusion occurs in acute coronary syndromes (ACS) and produces a zone of myocardial necrosis, hence troponin release. The infracted tissue is non-functional and so does not contract. In addition to the area of infarcted tissue, there is a surrounding area of viable, but non-functional, myocardium, referred to as 'stunned' myocardium. The non-functional myocardium can be directly visualised on echocardiography as a dyskinetic segment or wall motion abnormality. The non-functional myocardium produces impairment of myocardial function in proportion to its size and location.
This will produce ventricular dysfunction and BNP release. Using samples taken during the Orbofiban in Patients with Unstable Coronary Syndromes - Thrombolysis In Myocardial Infarction (OPUS TIMI) 16 trial (a study of the oral glycoprotein IIb/IIIa antagonist orbofiban) the ability of BNP (triage) to predict outcome was examined.17 The study enrolled patients admitted with ACS and measured BNP in the first 72 hours from admission.
The final population examined comprised 825 patients with sinustachycardia (ST) elevation acute myocardial infarction (AMI), 565 with non-ST elevation AMI, 1,133 with unstable angina and two with an unspecified final diagnosis with follow-up to ten months. There was no relationship between BNP level and time of symptom onset. BNP values were divided into quartiles and the BNP quartile predicted survival; the greater the elevation, the greater risk of a subsequent cardiac event. This was a clinical trial, so there is a potential for selection bias.
In a prospective single centre study with unselected admission of all suspected ACS patients without ST elevation on the ECG,18 samples were taken on admission and at six hours from admission NTpBNP (Roche) was measured. Patients were followed for a median of 40 months. NTpBNP values were divided into quartiles and again, NTpBNP level predicted outcome, with the highest quartile of NTpBNP having a relative risk of death 26.6 times that of the lowest quartile. In this study, it also made no difference whether the zero- or six-hour sample was used. This suggests that a single sample for BNP/NTpBNP taken at or around admission is sufficient for risk stratification.19 The investigators examined whether NTpBNP was an independent outcome predictor to cardiac troponin T (cTnT). A combination of NTpBNP and cTnT gave independent and complementary risk stratification. In a substudy of the OPUS TIMI 16 study, samples from 450 patients were combined with samples from 1,635 patients enrolled the Treat Angina with Aggrastat® and Determine Cost of Therapy with Invasive or Conservative Strategy (TACTICS) TIMI 18 study. This was a study of ACS patients without ST segment elevation on the ECG randomised to early revascularisation (in patient angioplasty or coronary artery bypass surgery) compared with conservative management. Samples were taken on enrolment, within the first 24 hours from symptom onset. BNP (triage), cardiac troponin I (Bayer Diagnostics) and sensitise C-reactive protein (CRP) (Dade-Behring) were measured.
The three markers were shown to be additive (hence independent) risk predictors. Further studies in AMI have compared BNP/NTproBNP measurement with ejection fraction20 and have shown that BNP/NTproBNP measurement provides further prognostic information. These studies have been confirmed across the range of ACS and BNP measurement has been shown to be superior to risk scoring and CRP as an outcome predictor.21-24
It can therefore be concluded that BNP/NTpBNP measurement offers additional risk stratification in patients admitted with ACS beyond that of existing markers. It is known that the major determinant of survival post-MI is ventricular ejection fraction. While cardiac troponin measurement will indicate the degree of myocyte damage, BNP/NTpBNP will assess ventricular function. Hence it is unsurprising that it is an outcome predictor; however, though comprehensive data from clinical trial datasets shows that BNP/NTproBNP is probably the best single prognostic marker, there have been no prospective studies in all acute chest pain populations and there are no data suggesting that BNP/NTproBNP measurement will alter management. The role in ACS therefore remains to be defined.
Conclusion
BNP/NTpBNP measurement is a powerful diagnostic and prognostic tool for detection of ventricular dysfunction. It is an ideal test for detection of cardiac failure in primary care, allowing cardiac failure to be definitively ruled out as a cause of dyspnoea. It should form part of the repertoire of all laboratories.