Reports

This report is based on medical evidence presented at sanctioned medical congress, from peer reviewed literature or opinion provided by a qualified healthcare practitioner. The consumption of the information contained within this report is intended for qualified Canadian healthcare practitioners only.

Canadian Cardiovascular Congress 2007

Quebec City, Quebec / October 20-24, 2007

Heart failure (HF) is typically diagnosed based on clinical evaluation, including history, physical examination, chest X-ray, and sometimes a left ventricular function assessment. However, various symptoms of HF, such as dyspnea, edema and fatigue, may mimic those of other illnesses. Greater speed and accuracy of HF diagnosis in primary and emergency care settings has the potential to expedite and improve both early and longer-term management of the condition. Several studies have now determined that measurement of B-type or brain natriuretic peptide (BNP) and/or its amino-terminal fragment NT-proBNP are complementary to clinical evaluation in the diagnosis and management of HF. Released when cardiac myocytes are stretched excessively, these peptides reach elevated levels in the blood in the setting of HF and other conditions in which cardiac ventricles are under strain. Their release leads to decreases in cardiac output and blood volume.

Biomarker Measurement Recommended in Guidelines

In its current guidelines on HF (Arnold et al. Can J Cardiol 2006;22(1):23-45, 2007;23(1):21-45), the Canadian Cardiovascular Society (CCS) recommends that one of the two biomarkers should be measured to help confirm or rule out heart failure in patients in whom the clinical diagnosis is in doubt. Their measurement may also help establish prognosis in patients with confirmed HF (for example, the concentration of NT-proBNP in the blood has been shown to be significantly related to New York Heart Association HF class) and help guide therapeutic decisions, although the evidence for these applications is considered less robust, commented Dr. Gordon Moe, Associate Professor of Medicine, University of Toronto, Ontario, and Director, Heart Failure Program and Biomarker Laboratory, St. Michael’s Hospital. “For diagnostic purposes... the data are strong. In someone in whom the diagnosis is obvious, I don’t think it’s helpful. That said, if the patient already has a history of HF, it’s a good prognostic marker but the problem is we’re not sure how to deal with that information: we know the patients with higher measurements do not do well, but what to do about it? That’s still a dilemma. A third indication is following patients in terms of therapy. The data are not very clear-cut yet—that’s why we don’t give it a very strong recommendation—but that’s something to be considered as well.”

Clinicians measuring BNP or NT-proBNP as an adjunct to diagnosis can be guided by the cutpoints in the CCS guidelines, which emanate from the recent PRIDE study (Januzzi et al. Am J Cardiol 2005;95:948-54) and take into account the effect of age on NT-proBNP, Dr. Moe added. For example, in patients aged between 50 and 75 years, HF is considered very likely if the patient’s NT-proBNP is >900 pg/mL, while the cutpoint is >1800 pg/mL in more elderly individuals.

IMPROVE-CHF Findings

The findings of the IMPROVE-CHF study (Improved Management of Congestive Heart Failure) (Circulation 2007;115(24):3103-10) contributed to the CCS recommendations on the use of BNP or NT-proBNP, commented Dr. Moe. In this study, investigators evaluated the clinical and economic value of adding NT-proBNP measurement to clinical assessment of patients presenting to emergency rooms (ERs) with dyspnea of suspected cardiac origin. Five hundred patients (mean age was approximately 70 years) were included in the analysis; about 35% had or reported a prior history of HF or left ventricular dysfunction. They were randomly allocated to NT-proBNP-guided care (n=246) or usual care (n=254). Physicians caring for the first group were given the NT-proBNP findings and information on their interpretation. “We also did a second NT-proBNP assay on patients who were admitted to hospital and those results were also given to the doctor [to guide management post-discharge],” Dr. Moe specified.

The IMPROVE-CHF results confirmed that NT-proBNP levels have strong predictive value in patients with dyspnea: they are significantly higher in patients with HF than in those without (median 3697 vs. 212 pg/mL, P<0.00001). The study’s key findings were that management guided by NT-proBNP measurement leads to time and resource savings in the first 60 days after the patient’s presentation. Notably, individuals whose biomarker levels were known spent 21% less time in the ER (5.6 vs. 6.3 hours, P=0.031) and underwent fewer tests for investigation of dyspnea. In the 60 days following their entry into the study, direct inpatient and outpatient medical costs for patients whose NT-proBNP levels was measured were 15% lower (US$5180 vs. $6129, P=0.0232). Over the same period, the number of patients who required rehospitalization for HF was 35% lower in the NT-proBNP group than in the usual care group. According to Dr. Moe, there are several possible reasons for this finding, including more targeted use of HF therapies in patients with a definitive diagnosis.

More Analysis on Outcomes

Dr. Moe and colleagues presented additional analysis from the IMPROVE-CHF study here at the 2007 Canadian Cardiovascular Congress. In one paper (abstract 628), they indicate that the patients whose therapy was guided by NT-proBNP had shorter length of both initial hospital stay (4.6 vs. 6.1 days, P=0.014) and subsequent hospital stays (1.3 vs. 2.5 days, P<0.004). They also had about 50% fewer readmissions (24 vs. 50, P=0.029). Dr. Moe commented, “Interestingly enough, NT-proBNP did not predict rehospitalization, which is in contrast to our previous studies. But then, our follow-up was only out to 60 days following randomization, so that may be too soon [to see an effect].” In this analysis, older age was the principal predictor of rehospitalization or death. Body mass index >25 to 30 kg/m2 was predictive of fewer events, Dr. Moe added.

In a separate presentation, Dr. Moe described the finding that anemia strongly predicted 60-day rehospitalization in the IMPROVE-CHF study. “This is observed consistently with the patient with chronic HF, but in acute HF, it has not been reported before,” he remarked. There is a strong relationship between anemia and NTproBNP, he added. “The higher the NT-proBNP, the lower the hemoglobin.”

Wider Role for These Biomarkers?

If the results of IMPROVE-CHF can be extrapolated to national health care, the use of NT-proBNP would mean savings of approximately $210 to 345 million annually. Unfortunately, Canadian hospitals and physicians have been slow to employ these biomarkers on a routine basis, Dr. Moe commented. He added that while the IMPROVE-CHF study focused on ER presentation, many elderly patients with dyspnea present to general or family physicians’ offices and that the test might prove highly useful in such primary care settings. “GPs see a lot of patients who have shortness of breath and in whom a diagnosis is unclear. This test could be extremely helpful.”

Questions and Answers

The following section is based on discussions with Dr. Malcolm Arnold, Professor of Medicine, University of Western Ontario, and Director, Research Affairs, Division of Cardiology, London Health Sciences Centre; Dr. Anique Ducharme, Associate Professor of Medicine, Université de Montréal, Quebec, and Director, Heart Failure Clinic, Montreal Heart Institute; and Dr. Jonathan Howlett, Associate Professor of Medicine, Dalhousie University, and Medical Director, Cardiac Transplant and Heart Function Clinic, Queen Elizabeth II Health Sciences Centre, Halifax, Nova Scotia, during the scientific sessions.

Q: How important is the use of biomarkers in the diagnosis of HF?

Dr. Arnold: HF symptoms can occur in normal people at times and can occur with other medical disorders such as lung disease and anemia. In many cases, a physician can with reasonable certainty make or exclude a diagnosis of HF on the basis of the clinical presentation, examination, associated illnesses and a chest X-ray. However, if the diagnosis is in doubt—and as the patient’s treatment plan and long-term outcomes depend on the correct diagnosis—it is important to be sure. In this instance, the measurement of natriuretic peptides can be very valuable.

Dr. Ducharme: Nowadays, we are relying more than ever on blood tests in cardiology, notably troponins for acute coronary syndromes and BNP or NT-proBNP for acute HF. We’ve known since 2001 that BNP or NT-proBNP could be used as an adjunct to clinical judgment.

Dr. Howlett: The use of biomarkers represents a significant advance. Clinical judgment alone, even in the best of hands and with routine diagnostic testing (such as echocardiogram and X-ray), is only about 80% correct. A strategy where BNP or NT-proBNP is utilized can increase this accuracy to over 90% as well as reduce time to diagnosis and time in the ER. This leads to better, faster patient care and overall less cost to the hospital. Q: Do you have access to BNP/NT-proBNP measurement? If so, what advantages have you noticed?

Dr. Arnold: Unfortunately, our institution has not yet provided BNP or NT-proBNP as an available test but we hope they will be available soon!

Dr. Ducharme: We have had access to the tests for almost five years. I was surprised to hear in discussions at the CCC that they are not widely available elsewhere. In our institution, we see many patients who had a myocardial infarction and used to be smokers. Sometimes it’s difficult to determine whether they have decompensated chronic obstructive pulmonary disease or HF. It’s an expensive test but if you can rely on it to avoid further testing such as echocardiogram, it’s certainly cost-effective.

Dr. Howlett: We do have access to NT-proBNP. In many cases, knowledge of the result of this test has assisted in decisions to increase or decrease diuretic and/or vasoactive medication use, as well as to determine that the cause of shortness of breath was non-cardiac. In the past we either had to perform a “trial of diuresis” and wait for several days to see if there was improvement, or to perform further diagnostic tests.

Q: HF guidelines indicate clinicians may test for BNP or NT-proBNP. Why measure one as opposed to the other ?

Dr. Arnold: They give similar results. NT-proBNP has a longer half-life than BNP. The diagnostic cutpoints vary for each test and NT-proBNP levels need to be adjusted for age while BNP does not. BNP is also available at the bedside as a point of care test. I think the physician can rely on both.

Dr. Ducharme: The only distinction I can make is in terms of biological half-life.NT-proBNP has a longer half-life so it better reflects what happened in the last six hours or so. For example, if a patient has not been feeling well for several days, but he takes his diuretic before coming to the hospital, his BNP level comes down within about two hours of his taking the diuretic. The NT-proBNP level remains elevated for a longer period. On the other hand, in the ER or the intensive care unit, when you want to know how the patient is responding to therapy, perhaps the BNP would be better.

Dr. Howlett: Studies to date have not shown a clear superiority of one over the other. They each have their advantages and disadvantages, which must be known to the user in order to prevent inappropriate use.

Q: Given the findings of IMPROVE-CHF and similar studies, should BNP/NT-proBNP testing be performed routinely in patients with suspected HF?

Dr. Arnold: IMPROVE-CHF was able to show that prompt confirmation of a HF diagnosis with this blood test does improve patient care. Time can be wasted in doing nonspecific tests or waiting to see if there is a response to bronchodilators. I think BNP or NT-proBNP should be more widely available in all acute care settings where HF patients might come for help.

Dr. Ducharme: I have been saying it for six years! IMPROVE-CHF was the first to show that in our health care system, we could improve decision-making and shorten both the ER and hospital stay by using the NT-proBNP level. That said, in the ER, I think you don’t need it for full-blown pulmonary edema. And there is also some overlap, so that in the guidelines we discuss its use in ruling in and ruling out HF.

Dr. Howlett: The IMPROVE-CHF study was performed in a real-world setting in which there was no previous experience with the test. Despite this, ER physician knowledge of the NT-proBNP result led to greater diagnostic accuracy, less time in ER, and less total health care cost per patient. Given these results, I cannot understand why a health care institution would not want to have this test available. Q: Dr. Moe has suggested that Canadian hospitals have been slow to start using the NT-proBNP test used in IMPROVE-CHF. Could you comment?

Dr. Arnold: I agree. I think there is a concern that the test will be used too often. Yet Dr. Moe’s study showed cost savings within our health care system. The IMPROVE-CHF study will be critical in helping hospitals and laboratories decide that it is in their and patients’ best interests to make the test available.

Dr. Ducharme: There is concern that ER physicians may over-rely on the test and be less confident in their clinical evaluation. Also, there is the issue of cost. But if the BNP or NT-proBNP is normal and the patient clearly does not have HF, you avoid additional expensive tests such as echocardiogram or multiple gate acquisition scan. I compare this situation to when troponin testing was introduced in the early 1990s. I think nobody would want to go back to when the troponins were not available.

Dr. Howlett: This is true and a frustrating situation. Once physicians learn the data, they want this test. But adding blood tests to clinical care means someone has to pay. The current administrative setup in Canadian hospitals lends itself to a “silo” budget where one department is reluctant to add costs because the savings may accrue in a different department (such as diagnostic testing or radiology, for example). Best practices demand a solution to these challenges.

Q: Are there ways to ensure success in the introduction/use of a test such as BNP/NT-proBNP?

Dr. Arnold: To allay concerns about overuse or abuse of the test, I think each hospital or laboratory can easily draw up appropriate criteria that reflect the best evidence from clinical trials.

Dr. Ducharme: We want to make sure the ER physicians don’t stop relying on their clinical judgment. But in the hospitals in which it’s been implanted in the US there has not been a shift toward discharging sick patients. What has been observed is an increase in cardiology consults in the ER for mildly elevated BNP or NT-proBNP usually because of renal disease. A Calgary group has looked at patients with low BNP levels and it was very interesting to see that probably, if you employ a higher cut-off point, there may be fewer inappropriate cardiology consults. But this is still to be confirmed.

Dr. Howlett: Successful introduction and use of any new test requires a rationale and data for the test, a physician championing its use, administrative support, a high level of agreement regarding cost or cost sharing, a robust education and in-service program to prevent inappropriate use, and finally a surveillance process to ensure quality care.

Q: Do you also believe the test is relevant to and feasible for use in primary care?

Dr. Arnold: As primary care physicians see most of these patients first, the opportunity to make an earlier accurate diagnosis could have an enormous positive impact on future health care utilization and cost. As in the hospital setting, appropriate criteria can be established to ensure optimal use.

Dr. Ducharme: BNP and NT-proBNP are not markers of HF but of wall stretch so if the patient’s condition is very well controlled, sometimes the test result may be normal. In addition, we have little data on outpatients. So we have to be careful. Certainly I don’t think it makes sense for screening.

Dr. Howlett: Evidence is accumulating that this test can add significantly to quality care in the primary care setting. Proper use requires effective education. At this time I would prefer to focus on the ER where the evidence is best.