B-type natriuretic peptide (BNP) and N-terminal prohormone brain natriuretic peptide (NT-proBNP) have been shown to be extremely helpful in the diagnosis and management of patients with heart failure (HF). These neurohormones are secreted from both the left and the right cardiac ventricle in response to ventricular volume expansion and pressure overload. BNP and NT-proBNP can be seen as quantitative markers of HF that summarise the extent of systolic and diastolic left ventricular dysfunction.
A wealth of research data from clinical studies together with five years of clinical experience with BNP allow us to provide clear recommendations and answer frequently encountered questions. With several additional indications still undergoing intense study, current evidence clearly supports the use of BNP and NT-proBNP in three clinical settings - patients with acute dyspnoea, prior to discharge in patients hospitalised with acute HF and the long-term management of patients with HF. This review will focus on the use in patients with acute dyspnoea as this indication has become an essential part of evidence-based medicine.
Research activities during the last decade have provided a wealth of clinical research data that have improved our understanding of BNP and NT-proBNP. This review will summarise clinical situations with sufficient evidence to support the use of BNP and NT-proBNP. For some indications, the question is no longer whether BNP and NT-proBNP should be used, but rather how best use can be made of them. This review will also provide detailed recommendations on the most appropriate cut-off values for clinical decisions.
Scope of the Clinical Problem
HF is a major public health problem; it is common, associated with high morbidity and mortality, extremely expensive and difficult to diagnose. Currently, there are nearly 1.5 million new cases of HF in Europe and North America every year. HF is characterised by frequently recurrent decompensation leading to worsening dyspnoea in many patients. Moreover, five years after the diagnosis of HF, 50% of HF patients will have died from the disease. HF is associated with high rates of hospitalisation in patients more than 65 years of age, and these hospitalisations contribute significantly to the enormous cost of the disease. It is estimated that in Europe the total cost of HF exceeds €50 billion every year.
HF is difficult to diagnose because symptoms are non-specific, and typical physical signs are present in less than half of all patients with HF. Our record in the diagnosis of HF is poor with less than 50% of patients being correctly identified during the initial consultation. Misdiagnosis of congestive heart failure (CHF) causes morbidity and increases time to discharge and treatment cost.
BNP and NT-proBNP - How Good are They?
BNP and NT-proBNP can be seen as quantitative markers of HF that summarise the extent of systolic and diastolic left ventricular dysfunction. In general, levels of BNP and NT-proBNP are directly related to the severity of HF symptoms and to the severity of the cardiac abnormality. BNP is a 32-amino acid polypeptide that is co-secreted with the inactive amino-terminal proBNP (NT-proBNP) from both the left and the right cardiac ventricle in response to ventricular volume expansion and pressure overload. Observational studies, including patients presenting with symptoms suggestive of HF, mainly dyspnoea, that validated BNP and NT-proBNP against a gold standard diagnosis of HF have convincingly demonstrated that BNP and NT-proBNP as single tests outperform all other variables available in the emergency department (ED). BNP and NT-proBNP seem to have similar test characteristics for the diagnosis of HF in patients presenting with acute dyspnoea. However, it is important to note that the actual cut-off values are different (see Figures 1 and 2).
The largest validating study, including more than 1,500 patients, examined the use of BNP in patients presenting with acute dyspnoea in the ED. At 100pg/ml, BNP had a sensitivity of 90% and specificity of 73% in the only multicentre study reported so far. In determining the correct diagnosis (HF versus no HF), adding BNP to clinical judgement would have enhanced diagnostic accuracy from 74% to 81%. In those participants with an intermediate (21-79%) clinical probability of HF, BNP at a cut-off of 100pg/ml correctly classified 74% of the cases. The areas under the receiver operating characteristic (ROC) curve were 0.86 for clinical judgement, 0.9 for BNP at a cut-off of 100pg/ml and 0.93 for the two in combination (p<0.0001 for all pair-wise comparisons).
These data led to the conclusion that the use of BNP increases the accuracy of the clinical evaluation in patients presenting with acute dyspnoea. Moreover, a randomised comparison of a strategy of making NT-proBNP results available to primary care physicians, in addition to the electrocardiogram (ECG), chest radiography and echocardiographic data, has reported a substantial increase in diagnostic accuracy for patients with new symptoms that might be caused by HF. The main impact of NT-proBNP measurement on diagnostic accuracy was the general practitioner correctly ruling out HF. In summary, BNP and NT-proBNP as single tests outperform all other variables available in the emergency diagnosis of HF. Moreover, when used in conjunction with other clinical information, BNP and NT-proBNP significantly increase diagnostic accuracy. However, there are certain pitfalls when HF may present with low BNP and NT-proBNP levels that need to be kept in mind. These include HF secondary to causes upstream from the left ventricle including mitral stenosis and acute mitral regurgitation.
The Use of BNP Improves Medical and Economic Outcome
Although the diagnostic potential of BNP in patients with acute dyspnoea was first reported more than 10 years ago, the assays available at that time had a turn-around time of several days. This severely limited its use in clinical practice. The development of a rapid fluorescence immunoassay (Biosite, San Diego, CA) allowed BNP levels to become available within 20-30 minutes. The central question remained whether the availability of a simple and rapid blood test that increases the diagnostic accuracy translates into improved patient management when used in clinical practice.
This important issue was addressed in the BNP for Acute Shortness of breath EvaLuation (BASEL) study. The median time from presentation to the emergency department (ED) to the initiation of the appropriate therapy according to the final discharge diagnosis was 90 minutes in the control group and 63 minutes in the BNP group (p=0.03). The use of BNP levels significantly reduced the need for hospital admission (75% compared with 85%) or intensive care (15% compared with 24%). Time to discharge was significantly reduced in the BNP group - median eight days compared with 11 days in the control group. Total cost of treatment was US$5,410 in the BNP group compared with US$7,264 in the control group, a significant reduction of 26%. These data support the conclusion that used in conjunction with other clinical information, rapid measurement of BNP in the ED improves medical and economic outcome.
BNP Should be Measured in All Patients Presenting with Acute Dyspnoea
The BASEL study included unselected consecutive patients presenting with acute dyspnoea. Recent data suggested that BNP levels are most useful in patients with an intermediate clinical probability of HF. Whether restricting BNP measurements to patients in this subgroup would yield similar medical and economic benefits to those observed in the BASEL study is unknown. Moreover, the approach used in the BASEL study has obvious logistical advantages. Delaying the venipuncture for BNP until the physician has collected all clinical data - and a chest X-ray on most occasions - to determine whether the individual patient has an intermediate clinical probability of HF would significantly increase the time to the correct diagnosis and accordingly the time to appropriate treatment in those patients who might benefit the most from BNP testing. As BNP testing is non-invasive, simple and cost-effective, measuring BNP directly at presentation, at the time of venipuncture for routine blood tests in all patients with acute dyspnoea, seems to be reasonable. Moreover, in addition to the diagnostic utility, BNP levels do provide valuable prognostic information in patients with HF. This prognostic information may have contributed to the improved outcomes of the patients in the BNP group.
What Cut-off Values Should be Used in Patients with Acute Dyspnoea?
In order to make best use of the diagnostic information of BNP and NT-proBNP levels, most experts agree on the fact that it is most appropriate to use two cut-off values - a lower one with a high negative predictive value to reliably exclude HF as the cause of acute dyspnoea and a second higher one with a high positive predictive value to 'rule in' HF as the cause of dyspnoea. As shown in Figure 1, for BNP 100pg/ml and 400pg/ml should be used. These cut-off values apply irrespective of age and sex. However, two clinical conditions require adjustment - kidney disease and obesity, as described in the figure.
About 75% of patients with acute dyspnoea will present with either low (<100pg/ml) or high (>400-500pg/ml) BNP levels. In these, the BNP level is extremely helpful and quickly leads to the correct diagnosis. In the other 25% of patients, the BNP level is in a grey zone. Although most patients with intermediate BNP levels do have mild HF, BNP is less helpful in this range due to considerable overlap with BNP levels in pulmonary embolism, pneumonia and other disorders.
The International Collaborative for NT-proBNP Study helped define the most appropriate cut-off values for NT-proBNP by pooling data from several single-centre studies that had suggested a wide range of cut-off values, with differences in baseline characteristics including age most likely responsible for this fact. As shown in Figure 2, 300pg/ml should be used to rule out HF. Depending on age, 450pg/ml, 900pg/ml or 1,800pg/ml should be used to rule in HF. Again, NT-proBNP levels below the lower cut-off or above the upper cut-off value are extremely helpful, whereas NT-proBNP levels in the grey zone are less helpful. Of note, obesity is also associated with significantly lower NT-proBNP levels.
In conclusion, our HF patients are in desperate need of better medical care. The introduction of BNP and NT-proBNP represents a major advance in the diagnosis and management of HF. The use of these quantitative markers of HF is cost-effective and allows us to improve medical and economic outcomes. We should take advantage of these simple tests to improve the management of our patients with HF and other conditions presenting with acute dyspnoea.