May 2009: Volume 36, Number 5
The BNP Debate
How Important is this Biomarker in Diagnosing Heart Failure?
By Gina Rollins
Heart failure is a significant health problem in the U.S., accounting for more than 1 million hospitalizations and about 3.4 million ambulatory care and emergency visits annually, at a cost of at least $37 billion. Despite its prevalence, the condition remains difficult to diagnosis, as most patients have several comorbidities, many of which share similar symptoms. In addition, heart failure involves a complex interaction between hemodynamic, neurohormonal, and cardiorenal responses, as well as other physiological factors. Perhaps for these reasons, a substantial proportion of patients referred to cardiologists by primary care physicians have been misdiagnosed initially, and up to 90% of emergency patients with heart failure are admitted to the hospital, versus only about 37% to 60% of those with chest pain. The need to make an accurate but rapid diagnosis is especially crucial in the over-taxed emergency setting.
As a result, much attention has been placed on the role of biomarkers, particularly the B-type natriuretic peptides BNP and NT-proBNP (collectively, BNP), in the diagnosis, disease staging, and risk of heart failure progression. While BNP stands out among a constellation of novel cardiac, metabolic, and inflammatory biomarkers that have been investigated in relation to heart failure, its role in distinguishing the condition is not definitive, and heart failure remains a clinically defined disease. Consequently, many clinicians find BNP helpful in diagnosing heart failure, at least in a sizeable subset of patients, while others have opined that it “looks like it helps but usually adds nothing” (Ann Emerg Med 2009;53:386–389).
Among those who find BNP measurements of benefit in diagnosing new-onset acute heart failure is Deborah Diercks, MD, MSC, professor of emergency medicine at University of California, Davis. “BNP is a useful diagnostic test for patients with intermediate risk who present with no prior diagnosis of heart failure, but with multiple risk factors for development of heart failure,” she said. “It presents in many different ways, so it’s a little bit of a challenge to diagnose. There’s a combination of things we can do to make the diagnosis easier, and BNP levels are one of them.”
A clinician with less enthusiasm about the utility of the BNP test is Donald M. Yealy, MD, professor and vice chair of emergency medicine at the University of Pittsburgh. “There are a few cases when you need it, but that’s invariably when the test is in the middle ground and of little help,” he noted. “In the end, there is no way around the truth that acute decompensated heart failure is a clinical diagnosis, not a laboratory one. BNP can only be used safely when deployed selectively and with a full grasp of its limits.”
Evolving State of Discovery
Uncertainties about the role of BNP relate to the evolving state of knowledge about BNP itself, analytic issues involving BNP assays, and the state of science about heart failure, according to researchers. BNP is in a family of natriuretic peptide hormones released by the cardiac muscle because of various stimuli, the foremost being an increase in wall tension in response to pressure and volume overload in both the atria and ventricles. The hormone has a strong diuretic effect, promotes vasodilatation, and facilitates cardiovascular remodeling and response to ischemia.
Initially a 134-amino acid peptide (preproBNP), BNP is further synthesized to form a 108-amino acid peptide called proBNP. ProBNP, in turn, is cleaved enzymatically to form the biologically active BNP, which is a 32-amino acid C-terminal peptide, and NT-proBNP, its 76-amino acid co-metabolite. Both BNP and NT-proBNP have been studied extensively, and for all practical purposes the two are considered of equal diagnostic value. “A lot of data supports the equivalence of both markers for diagnosis, although there is evidence that for prognosis there is more information to be gleaned from NT-proBNP,” noted Christopher deFilippi, MD, FACC, associate professor of medicine at the University of Maryland School of Medicine in Baltimore. Obesity, age, sex, and renal and thyroid function are known to affect levels of both markers.
Although much is known about the biomarkers, the next 5 years will bring much more understanding about their properties and how they can be used to guide therapy, according to Fred Apple, PhD, professor of laboratory medicine at the University of Minnesota School of Medicine, and medical director of clinical laboratories at Hennepin County Medical Center in Minneapolis. "There's still a lot to be learned in terms of what's really circulating in the blood," he said. For instance, there have been studies suggesting that BNP circulates in more than just the BNP (32-amino acid) and NT-proBNP forms. The IFCC committee on Standardization of Markers of Cardiac Damage, which Apple chairs, also has been examining the cross-reactivity between BNP and NT-proBNP assays. “We observed that BNP has very little cross-reactivity in NT-proBNP assays and vice versa, but proBNP does cross-react in both assays,” he explained. “We think the significance of this is that in the later stages of heart failure the processing of proBNP may be impaired. If we get to the point where we’re measuring both proBNP and BNP or proBNP and NT-proBNP in the respective assays, it might give us a better idea of what stage of heart failure the patient is in, and more importantly, how therapy should be initiated.” More evidence also is emerging about the biological variation of BNP. Apple also chaired the writing group for the analytic issues chapter of NACB’s Laboratory Medicine Practice Guidelines (LMPG) on acute coronary syndrome and heart failure.
Along with these developments, during the past few years there have been analytic improvements in BNP assays. “The REDHOT and BNP trials were done when the assay had a CV of about 20%, which is huge. Currently the CVs are around 4%, so that can't help but make the interpretation better,” observed Robert Fitzgerald, PhD, DABCC, associate clinical professor at the University of California San Diego and associate director of clinical chemistry for the VA Hospital in San Diego. (The Rapid Emergency Department Heart Failure Outpatient Trial (REDHOT) was published in 2004; the Breathing Not Properly (BNP) study was published in 2002.)
Heart Failure in Flux
Just as the science about BNP is evolving, there still is much to be learned about heart failure, particularly in the acute stage, according to Alan Storrow, MD, vice chairman for research and academic affairs in the department of emergency medicine at Vanderbilt University Medical Center in Nashville. “The diagnosis of heart failure is sometimes difficult and a little messier to make than with heart attack, where we have really good endpoints of what we call heart attack and what we don't," he explained. “Our ability to provide diagnostic endpoints for heart failure is many decades behind acute coronary syndrome, and our treatments are, too. They’re essentially the same as they were 25 years ago.”
The incidence of heart failure has increased with a growing elderly population and more people surviving heart attacks than in the past but subsequently developing heart failure, Storrow added. He noted that until recently there was more emphasis on understanding chronic rather than acute heart failure. “The treatment of chronic heart failure from an ambulatory setting has received a lot of attention, but the treatment of acute failure in people who come into the emergency department because they're decompensated or have new-onset heart failure has not been very well studied or defined,” Storrow said. “We're just now seeing in the last 8-to-10 years an emphasis on understanding what acute heart failure is and how to treat it.” Storrow was a committee member for NACB’s LPMG on ACS and heart failure and was a co-author of the Breathing Not Properly study.
Even as more is being learned about the biochemistry of BNP and the diagnosis and treatment of heart failure, guidelines of major professional organizations, including NACB, American College of Cardiology Foundation/American Heart Association, the Heart Failure Society of America, and the American College of Emergency Physicians, are united in their BNP-related recommendations. All note the utility of BNP measurements in discerning heart failure when the diagnosis isn’t clear based on clinical evaluation, the patient’s medical history, and other test results such as chest x-ray and ECG. For example, NACB’s LPMG states that BNP results can be used to rule out or confirm the diagnosis of heart failure in patients presenting with ambiguous signs and symptoms.
A 2009 focused update of ACCF/AHA guidelines affirmed the organizations’ 2005 recommendations that measurement of BNP can be useful in the evaluation of patients presenting in the urgent care setting in whom the clinical diagnosis of heart failure is uncertain. A 2006 evidence report of the Agency for Healthcare Research and Quality also found that in all settings, including the emergency department, measurement of BNP adds independent information beyond traditional diagnostic measures for heart failure.
So where does all this leave clinicians when it comes to evaluating emergency patients with dyspnea and other symptoms of heart failure? In many cases, BNP simply does not add to the diagnostic picture. If a physician is confident a patient has heart failure based on physical exam, history and other tests such as a chest x-ray, a BNP result probably will be of little utility. The same is true when there is little doubt the patient doesn't have heart failure. “Routine measurement of BNP is still not warranted,” said W.H. Wilson Tang, MD, assistant professor of medicine at the Cleveland Clinic. “Diagnosis and criteria for management was established on clinical grounds. You don’t necessarily need a BNP to pat yourself on the back and say, ‘congratulations, that’s what it is’.” Tang chaired the writing group for the chapter on clinical utilization of cardiac biomarker testing in heart failure for NACB’s LMPG on ACS and heart failure.
Still, at some institutions BNP may be part of routine testing in patients with shortness of breath or suspected heart failure. “There's a lot of variation in the use of the assay, from system-to-system and hospital-to-hospital. Some reflexively order it, and others use it more sparingly,” observed deFilippi. In the emergency setting in particular there may be a natural tendency to order the test to expedite suspected heart failure patients through the system. “In emergency medicine we don't have the luxury to wait until the sixth test comes back to decide if we need a seventh,” explained Judd Hollander, MD, professor and clinical research director of emergency medicine at the University of Pennsylvania in Philadelphia. “Maybe by the time you get all the results back you don't need the BNP. But in reality BNP results come back fast. So it may be confirming your clinical judgment, but we don't have the luxury of approaching patient care in a step-wise fashion.” Hollander was an investigator in the BNP and REDHOT studies.
deFilippi agreed that the bustling emergency setting is where BNP results can be particularly useful. “That makes a case for the biomarker. If you’re that busy, you might not be doing that great a physical exam, and patients are not great historians to put you in one direction or the other. That’s when the test can be pretty useful,” he said.
A 2007 CAP Q-Probe study with 119 participating institutions found that 40% of total BNP tests were performed on blood drawn in the emergency department, and that median turnaround time for non-point-of-care BNP testing was 40.5 minutes. A separate study found that both Massachusetts General Hospital and Brigham and Women's Hospital in Boston ran about 5,000 BNP tests per year, with one-third or less coming from the emergency department (Am J Clin Pathol 2005; 124:S122–128).
BNP/NT-pro BNP and Heart Failure
What’s the Evidence?
Many professional organizations have published guidelines on the use and interpretation of BNP in heart failure, including:
NACB Laboratory Medicine Practice Guidelines for Utilization of Biochemical Markers in Acute Coronary Syndromes and Heart Failure, Clin Chem 2007;53:2086–2096.
Heart Failure Society of America 2006 Comprehensive Heart Failure Practice Guideline, Journal of Cardiac Failure 2006;12:e1–e122 and Heart Failure Practice Guideline Website
2009 Focused Update: ACCF/AHA Guidelines for the Diagnosis and Management of Heart Failure in Adults, Circulation; DOI:10.1161/CIRCULATIONAHA.109.192064 and ACC/AHA 2005Guideline Update for the Diagnosis and Management of Chronic Heart Failure in the Adult, J Am Coll Cardiol 2005;46:1116–43.
Clinical Policy: Critical Issues in the Evaluation and Management of Adult Patients Presenting to the Emergency Department with Acute Heart Failure Syndromes, Ann Emerg Med 2007;49:627–669.
Testing for BNP and NT-proBNP in the Diagnosis and Prognosis of Heart Failure, Agency for Healthcare Research and Quality, AHRQ Publication No. 06-E014, September 2006, AHRQ Website
Stuck in the Middle
If BNP does not add to clinical decision making in patients with a high or low pretest probability of heart failure, it can be helpful in the workup of the one-third to one-half of patients at intermediate risk. Seminal studies of the two biomarkers established cutpoints that have been widely adopted but not endorsed specifically in practice guidelines. In the case of BNP, the Breathing Not Properly study found that in the one-third of subjects with an intermediate, 21% to 79% pre-test probability of heart failure, BNP at a cutoff of 100 pg/mL correctly classified 74% of the cases and only misclassified 7% of patients as not having heart failure when that turned out to be their final diagnosis (Circulation 2002;106:416–422). As an example, a clinical judgment of a 20% pre-test probability of heart failure, combined with a BNP measurement of 1000 pg/mL would result in an estimated 85% post-test probability of heart failure.
In the case of NT-proBNP, the PRIDE (ProBNP Investigation of Dyspnea in the Emergency Department) study found that cutpoints of >450 pg/mL for subjects less than age 50 and >900 pg/mL for those age 50 and older were highly sensitive and specific for heart failure, while a measurement of <300 pg/mL was an optimal rule-out threshold (Am J Cardiol 2005;95:948-54). Other studies have generally pointed to BNP’s utility in predicting outcomes and in reducing costs and/or utilization, although some recent data have been more neutral (JAMA 2009;301:383–392; Ann Intern Med 2009;150:365–371).
How these cutpoints are interpreted in practice influences clinicians’ perception of the utility of BNP results. Over time, physicians have come to better understand that intermediate results for patients with an intermediate pre-test probability of heart failure are not a panacea. “People have gotten more comfortable with the indecision range. When the BNP test first came out, it was interpreted as ‘well, it’s above normal, so let's call it heart failure’, but it’s now understood that that's not true,” noted Hollander.
For most “gray zone” intermediate-risk patients, Yealy does not think BNP testing brings much to the table. “If the pretest probability of heart failure is 40% and the BNP result raises it to 60%, those two numbers are meaningless in practice,” explained Yealy. “Do you think the average clinician can integrate those numbers while delivering care?”
However, others believe that even if a BNP result raises the post-test probability of heart failure only slightly, it can help direct care. “I think Dr. Yealy and I disagree on the magnitude of difference that would change your clinical practice,” said Diercks. “If the pretest probability of heart failure is 20% and BNP results increase the probability to 60% or 70%, it may be a small difference, but it would probably lead me to treat that patient as a heart failure case.”
The perspectives of both Yealy and Diercks are valid, according to Tang. “The latest data confirms the ongoing uncertainty about the use of BNP,” he said. “You can’t routinely rely on it, and yet it’s useful, so it’s kind of a catch-22.”
The key point is that if a BNP test is ordered, particular attention should be paid when the results don’t jibe with the clinician’s overall assessment. "When the doctor disagrees with the BNP result, they have to think long and hard about what they're doing because the data supports that in those discrepant cases, the BNP result is often right," said Hollander. Still, considering BNP results within a patient’s overall clinical circumstance has kept the biomarker from being discredited because of undue reliance on the test as a definitive diagnostic, like cardiac troponin in the case of MI. “We’ve learned over the past 10 years that even though there is a lot of argument, our biggest fear that people would falsely follow the BNP levels and administer drugs or treatments that might be harmful has been somewhat refuted. People are still using their clinical judgment,” Tang observed. “However, for clinicians who find measuring BNP or NT-proBNP useful in their practice, it can help them understand their patient's clinical status or alert them to potential changes—uses, as with any other blood test, that we have never required randomized controlled trials to demonstrate.”
As more data on BNP and heart failure accumulates, Tang recommended that laboratorians think carefully before switching from BNP assays to NT-proBNP assays and vice versa. Institutions that have done so have discovered that the change added to the confusion and controversy around interpreting BNP results, he indicated. Laboratorians also should continue to educate physicians about the assays, their analytic variations, and respective cutpoints. “One of the best messages they can get across is that BNP is like any other diagnostic test in that it is something that we should act on in some way. It should change our own suspicion of a diagnostic process or identify a high risk group, but it’s got to be something you’re going to use,” Diercks concluded.
BNP and NT-proBNP Assays
More information about commercially available BNP and NT-proBNP assays is available at the IFCC Website.
Dr. Tang has received research grant support from Abbott Laboratories.
Dr. Hollander has received research grant support from Biosite and spoken on behalf of Biosite.
Dr. Fitzgerald has received research grant support from Roche, Biosite, Abbott Laboratories, and Bayer Diagnostics.