College of American Pathologists
Printable Version

  Feature Story


cap today

Getting a jump on cardiovascular disease

Treating with nesiritide

March 2002
William Check, PhD

In the race to recognize cardiovascular disease, two new lab tests are breaking away. One is already out, and the other is in the final phases of investigation.

Measurement of brain (B-type) natriuretic peptide, or BNP, improves recognition of congestive heart failure in patients presenting to the emergency department with shortness of breath. "We believe that all patients presenting to the emergency room with shortness of breath in which congestive heart failure might be a cause should have this test performed," says Alan Maisel, MD, professor of medicine at the University of California, San Diego, and director of the coronary care unit and of heart failure research at the San Diego VA Health Care System. His research was instrumental in demonstrating the importance of BNP in congestive heart failure.

Pathologist Tai-Po Tschang, MD, medical director at St. Agnes Medical Center, Fresno, Calif., says the BNP test has had a major impact in the few months his hospital has offered it. "Emergency room physicians and intensivists have experienced some situations where it surprised them that their patients’ shortness of breath was due to congestive heart failure rather than COPD[chronic obstructive pulmonary disease] because BNP was elevated. That opened their eyes and convinced them of the value of this test."

BNP assay is being explored also for possible use in patients with acute coronary syndromes. "Our ability to predict death or congestive heart failure [with BNP] is profound," says James de Lemos, MD, assistant professor of medicine and director of the coronary care unit at University of Texas Southwestern Medical School in Dallas. BNP is not a diagnostic marker for acute coronary syndromes, he says; rather, it is a risk predictor that may differentiate subsets of patients.

Still in clinical trials is the second new test, ischemia-modified albumin (IMA, formerly called albumin cobalt binding). Increased IMA levels have been shown to characterize myocardial ischemia, the earliest stage of coronary insufficiency. "The ideal place for IMA will be to pick up ischemic injury earlier than any marker we currently have, even myoglobin," says Fred Apple, PhD, medical director of clinical laboratories at Hennepin County Medical Center in Minneapolis and professor of laboratory medicine and pathology at the University of Minnesota School of Medicine. "We want a marker that, when it is negative, clearly points away from ischemic injury." Dr. Apple sees IMA as potentially complementary to cardiac troponin, assuming results of ongoing trials are positive and the assay receives approval.

Others are equally clear about what is needed.

"If there is one thing we ED physicians and cardiologists have been searching for," says cardiologist Kristin Newby, MD, associate professor of medicine at Duke University Medical Center, "it is to have a marker that would identify a patient who is having an acute coronary syndrome, in its very early stages. Right now we don’t have anything that can do that." Measurement of IMA promises to fill that niche.

BNP for congestive heart failure

Originally called brain natriuretic peptide because the brain was the first source identified, BNP was found also to be secreted by the cardiac ventricles under conditions of volume expansion and pressure overload. It causes vasodilatation and natriuresis. "BNP is a measure of the stretching of the myocardial wall," explains Alan H. B. Wu, PhD, director of clinical chemistry and toxicology in the Department of Pathology and Laboratory Medicine at Hartford (Conn.) Hospital. Patients with CHF have enlarged hearts from sustained catecholamine neurostimulation and remodeling. "In CHF, the heart goes from elliptical to spherical, which is a bad geometry for pumping blood," Dr. Wu says. BNP levels reflect this abnormality.

In a pilot study Dr. Maisel performed, the negative predictive value (NPV) of a BNP level <80 pg/mL was 98 percent, while the positive predictive value (PPV) of BNP >100 pg/mL was >90 percent (J Am Coll Cardiol. 2001;37:379-385). These findings were verified in a recently reported multinational study, according to Dr. Maisel. "In multivariate analysis, BNP added strongly to any other tools a physician had available," he says. "BNP levels proved more accurate in diagnosing CHF than did Framingham criteria."

At Hartford Hospital, Dr. Wu reports that clinicians find BNP testing useful for diagnosing CHF. "In patients presenting with dyspnea, BNP can distinguish CHF as a cause of symptoms from non-CHF causes such as pneumonia and chronic obstructive pulmonary disease," he says. BNP is elevated in both diastolic and systolic dysfunction, he notes, whereas the echocardiogram is abnormal only with systolic dysfunction.

For patients known to have CHF, BNP levels can help with staging. "BNP levels correlate well with New York Heart Association classification," Dr. Wu says. BNP can monitor therapy with an ACE inhibitor as well. Because of its short half-life, declines in BNP are expected in one or two days. "If BNP goes up, that might suggest the medication is not working," Dr. Wu says.

Dr. Tschang introduced BNP testing on Sept. 17, 2001. "In the beginning we got only a few requests per week," he says. "Now we are doing 10 to 13 per day." He considers BNP a useful test. "Before, physicians never had a handle on that kind of patient," he says. It was difficult to tell whether shortness of breath was due to chronic obstructive pulmonary disease or heart problems. No gold standard existed. Diagnosis was based on clinical impression, EKG, and chest X-ray, which were often not definitive. "It is great from our point of view," Dr. Tschang says. "We are finally able to provide something that helps with diagnosis."

Through Jan. 31, 2002, Dr. Tschang’s laboratory had done 518 BNP tests, of which 77.2 percent were elevated. "This is a pretty high positive rate for a clinical test," Dr. Tschang says. "It may mean that previously CHF was underdiagnosed."

In the future, BNP levels may help screen for left ventricular dysfunction, Dr. Maisel suggests, perhaps eliminating the need for echocardiograms in some patients. "If BNP is very low, maybe <20 pg/mL, ventricular function is typically normal by echo," he says, "even in patients with clinical symptoms." These patients’ left ventricular dysfunction could be due to diabetes or drugs that affect cardiac function, such as anthracycline chemotherapeutic agents or Herceptin. At the present time this is not an approved use of BNP, Dr. Maisel cautions.

Other possible uses of BNP now being studied are to gauge prognosis and to evaluate the long-term effects of treatment. In a study just published, Dr. Maisel and colleagues tracked patients who came to the ED with shortness of breath. "Over six months BNP levels were highly predictive of outcome—cardiac death or CHF readmission—even when the emergency room physician did not diagnose CHF," Dr. Maisel says. And pilot data have shown that patients whose BNP levels decrease during in-hospital treatment are less likely to be readmitted within 30 days of discharge.

The current BNP test (Triage, Biosite Diagnostics) is a whole-blood immunoassay that takes two drops of blood and yields a result in 15 to 20 minutes. It is performed on a portable meter with reagent cartridges. Although the Triage reader is designed for point-of-care use, Dr. Tschang uses it in the central laboratory. "The cost is very high," he says. "We would welcome a BNP assay that could be done on an automated platform."

Other companies have licensed the right to develop a BNP assay from Scios Inc., says Dr. Wu, including Abbott Diagnostics and Bayer. Roche is working on an assay for N-terminal pro-BNP, an inactive metabolite of BNP.

BNP for acute coronary syndrome

Dr. de Lemos’ work tested whether BNP has prognostic value in ACS—unstable angina and myocardial infarction with or without ST-segment elevation. "Many patients with ACS have underlying left ventricular dysfunction," he reasoned. But, he adds, some data suggest that BNP levels may be released in the setting of myocardial injury in proportion to deprivation of blood flow. Whatever the mechanism, Dr. de Lemos says, "There is a need for better markers in patients with ACS. Cardiac troponin helps identify at-risk patients, but hemodynamic status also contributes to risk."

Using frozen blood specimens from a large trial that had evaluated aggressive therapy for ACS patients (OPUS-TIMI 16), Dr. de Lemos and colleagues found that BNP predicted mortality more strongly than any other factor, including age and troponin level (N Engl J Med. 2001;345:1014- 1021). "Even when adjusted for all other known predictors of risk, BNP remained highly predictive of mortality," Dr. de Lemos says. BNP levels were proportional to ensuing mortality even in subgroups where BNP was not previously thought to be useful—low-risk patients with no evidence of myocardial necrosis and negative troponin values.

"Clearly BNP is measuring something different from traditional markers of how much muscle died," Dr. de Lemos says.

Whereas troponin identifies patients at risk for MI or recurrent unstable angina, BNP is better for predicting congestive heart failure, Dr. de Lemos notes. It doesn’t identify patients with "angry" plaque—inflammatory plaque with residual thrombus; rather, it identifies patients with pump dysfunction. "We might eventually be able to choose selective therapy based on marker risk profiles," Dr. de Lemos speculates. Right now, however, "We don’t yet know what to do with this information." The next step is to devise and test selective strategies to manage acute coronary syndrome patients with elevated BNP.

IMA for ischemia

Ischemia-modified albumin refers to an alteration of albumin binding that occurs during cardiac ischemia. Oxygen deficit leads to changes, such as acidosis and anaerobic metabolism, that generate free radicals, Dr. Wu explains. Free radicals activate enzymes that modify the N-terminal sequence of albumin so that it does not bind metals well. In the commercial IMA assay (Ischemia Technologies), cobalt is added to a blood sample to measure this change.

The format of the IMAassay to be first approved will probably be for the clinical chemistry laboratory, says Robert Christenson, PhD, professor of pathology and director of rapid response, clinical chemistry, and point-of-care services at the University of Maryland School of Medicine, Baltimore. However, he notes, "Most likely the company is working on a point-of-care method, which is really where the test’s biggest value would be for rapid triage of patients suspected of having ischemia." Dr. Apple agrees with the need for near-patient testing. "If IMA becomes a portable assay," he says, "it could even be used in an ambulance on the way to the hospital."

Early studies suggested that IMA is a potent early marker of ischemia, as distinct from troponin or CK-MB, which are products of myocardial necrosis. "In ACS, the first step is ischemia, which is characterized by reduced oxygen delivery to the myocardium and is caused by plaque rupture or platelet thrombus formation," Dr. Wu says. When vessels become occluded, necrosis occurs, but it takes several hours for enzymes to appear in the blood. That is why IMA can be positive well before troponin or CK-MB.

Dr. Wu was an investigator in a four-center study showing that a negative IMA test at presentation to the ED had a 96 percent NPV for a negative troponin result six hours later (Clin Chem. 2001;47:464-470). "No laboratory test is going to be the only criterion for deciding diagnosis or triage," he says. "But this test can certainly help an ED physician rule out ischemia as a cause of chest pain." If EKG and clinical signs support this finding, then the physician may be more confident in discharging the patient.

When results were plotted as an ROC curve, the optimal value was determined to be 75 units/mL, says Dr. Christenson, lead author of the four-center study. At this value, NPV was 96 percent and PPV was 33 percent, suggesting that the chief utility of IMA would be as a rule-out test.

Cardiologist Christopher de Filippi, MD, associate professor of medicine at the University of Maryland School of Medicine, Baltimore, does see IMA "more as a test of exclusion." If IMA’s negative predictive value continues to be upheld, patients with atypical symptoms, a normal EKG, and a normal IMA level might not need further workup. "We might not even draw a troponin or order a stress test," Dr. de Filippi says. If IMA is positive, on the other hand, the clinician could move to the algorithm’s next step.

Dr. Christenson notes that median time to presentation in the study was within three hours of onset of signs and symptoms. Median presentation time across the country is about two hours. Says Dr. Apple, also a co-investigator on this study, "At our hospital we are seeing a median time to presentation for MI patients of about three hours." These results should generalize, then, to routine clinical practice.

Future trials will need to clarify whether IMA is specific for cardiac ischemia, rather than other tissues. "Even if it is not 100 percent specific to the heart, it can still be a powerful tool to triage patients to cardiac care," Dr. Apple says. The early marker most often used now is myoglobin, he notes, which has only 60-70 percent specificity. "If the specificity of IMA is in that ballpark, we will be in great shape, since IMA turns positive even earlier than myoglobin. And it detects ischemic injury, not necrosis," he says.

Precisely because IMA and troponin measure different myocardial states, it is still unknown how closely IMA should be expected to correlate with troponin, Dr. de Filippi notes. How should we interpret the 33 percent PPV of a positive IMA for a positive troponin? Myocardial damage is a continuum in which ischemia precedes injury. A true marker of ischemia should detect a subset of patients who don’t yet have injury and may not develop frank myocardial damage, that is, patients who have spontaneously reversible ischemia. Such patients may never develop a positive troponin test.

"If this is true," Dr. de Filippi says, "then IMA may allow us to carve out a much wider breadth of patients who might have ischemic heart disease than we could detect with troponin alone." This would be analogous to the way that mild elevations of troponin in the absence of elevated CK-MB were found to identify patients with so-called micronecrosis, who had worse prognoses than those with a totally negative troponin.

Physicians may need to be educated to appreciate this situation, Dr. Wu suggests. "We want a test that increases as early as possible, before myocardial necrosis takes place," he says. "But the clinician may think that a positive ischemia test means the patient will have an MI and then get disappointed when troponin does not increase." Primary care physicians will need to understand that a patient can have an ischemic event that doesn’t progress to full-blown myocardial necrosis but still carries increased risk for future cardiac episodes.

Dr. de Filippi cites another potential benefit of a positive IMA test in the setting of a negative troponin assay and a normal EKG. "Often what we most want to know is whether this patient’s chest pain has a cardiac or non-cardiac etiology," he says. For patients presenting with chest pain for the first time, it may be the first opportunity to diagnose coronary disease. "So even if the patient’s short-term prognosis is excellent, we can begin steps to prevent future cardiac events," he says.

IMA’s unknowns

Crucial questions remain to be answered about IMA, and trials are underway to address some of them. "We are still pretty early in the development of this test," Dr. Wu says. "Outcome studies need to be performed to see if individuals ruled out by an IMA assay at presentation end up having cardiac events hours or days or weeks later."

He calls it "premature" to suggest that therapeutic options could be recommended on the basis of ischemic testing. No data exist to show that outcomes can be modified. "But we could postulate down the road that a patient positive for ischemia might be given a glycoprotein IIb/IIIa inhibitor or even percutaneous coronary intervention [PCI]," Dr. Wu says.

In addition to defining the reliability of a negative IMA result, it will be important to evaluate the prognostic significance of a positive IMA with a negative troponin assay. "Individuals positive for ischemia have a greater likelihood to develop full-blown MI," says Dr. Wu, "but a patient can also have an ischemic event that resolves spontaneously before substantial damage occurs." Are such patients at greater risk of future cardiac events?

A first step is being taken toward answering one of these questions—how sensitively IMA detects ischemia—in a multicenter trial whose outcome will form part of the Food and Drug Administration submission for the IMA assay. Dr. Wu has already presented in abstract form data obtained for this study at Hartford Hospital (Clin Chem. 2001;47:A205 abs 672). Patients at risk for unstable angina who were undergoing elective myocardial nuclear perfusion imaging with an exercise stress test had blood drawn for IMA assays just before starting exercise (a baseline value) and one hour after peak exercise. Dr. Wu calls the perfusion scan "probably the closest technique to a gold standard for detecting myocardial ischemia." In this study, the sensitivity of IMA for cardiac ischemia was 92.3 percent. Its specificity was 86.5 percent; that is, stress testing did not show a perfusion defect in 13.5 percent of the patients in whom IMA was positive. Whether these are false positive IMA results or false negative perfusion scans is difficult to tell.

The University of Maryland is also one of the sites in the multicenter trial. In the larger trial IMA results are correlated with the nuclear stress test and other clinical data. "There is no real gold standard for ischemia," Dr. de Filippi reiterates. Specificity and sensitivity of nuclear stress testing is no better than 80-85 percent. So an expert panel reviews all data and its decision is correlated with IMA results, though a perfusion defect remains the primary standard for ischemia.

The IMA assay’s performance is being assessed simultaneously in a multisite study of 700 patients presenting to the ED with chest pain. Resting nuclear perfusion imaging will be used as evidence of acute ischemia in this trial and correlated with a rise in IMA, again using the judgment of an expert panel as the final criterion. "Probably this will be one of the most important studies," Dr. de Filippi says, "since it will evaluate the performance of the test in a community-based population." Testing in this study is now finished, he estimates, and data analysis is beginning.

Duke University’s Dr. Newby is a member of the study’s expert adjudication and advisory panel. "This study is trying to establish IMA as a viable early marker of ischemia in real-world clinical practice," she says. Such a marker is sorely needed, in her view, because current markers, troponin or myoglobin, turn positive only after heart cells are already dead. "We want to be able to identify a patient who is having ischemia—low blood flow to the heart muscle—before heart muscle cells have died," she emphasizes. "Then we can ask whether we can identify ischemia early enough to treat and prevent heart muscle damage. That is ideally what we as clinicians want."

Echoing other sources, Dr. Newby says, "The hard thing is that there is no true gold standard with which to compare the IMA test." Right now ischemia is defined by a combination of clinical symptoms, EKG, and nuclear perfusion imaging. Dr. Newby sees her role as adjudicator on the panel of experts as looking at clinical and nuclear data and asking whether she thinks a particular patient has ischemia or not. Panel members will make their decisions in a blinded fashion, not knowing results of IMA testing. Correlations between IMA results and the panel’s judgments will determine whether the test is working in a clinical setting.

"You would want to see a fairly high correlation between a positive IMA result and what we would clinically define as an ischemic syndrome," Dr. Newby says. But she raises a key question: If IMA is positive and other markers are negative, is that a false positive for IMA or a false negative for clinical criteria? "The test may be more sensitive than clinical factors," she says.

A good outcome in the evaluation studies would lead to testing the prognostic and therapeutic implications of IMA test results in a clinical setting, as was done with troponin. "Does an IMA result have a clinical impact?" Dr. Newby asks. "Can we change outcome based on early treatment?" Objectives would be to lessen the degree of myocardial necrosis or prevent it altogether, and to shorten time in the ED or coronary care unit, thereby reducing costs.

What kind of treatment would be appropriate for a patient with early ischemia?

"If I knew early on that this patient had ischemia," Dr. Newby says, "and I could say with some certainty that this chest pain is atypical but really ischemic, I might put them on low-molecular-weight heparin earlier than I otherwise would."

Another possible strategy worth testing in clinical trials would be starting glycoprotein IIb/IIIa inhibitors early. Should all patients with a positive ischemia test be taken to the catheterization lab early? Or scheduled for early elective angioplasty or stenting? All these are options worth considering. "We could certainly change the level of observation to which we assign that patient in the hospital," Dr. Newby says.

Even if the IMA test turns out to be accurate for detecting ischemia, it won’t be used in all acute coronary syndrome patients, Dr. Christenson points out. "Patients with a clear indication for thrombolytic therapy or angioplasty won’t need an ischemic marker," he says. Patients having a diagnostic EKG, such as those with ST-segment elevation, would fall into this category. "Individuals with unstable angina or non-ST-segment elevation MI—who make up the majority of ACS patients—will really benefit from this test," he says.

IMA is the marker that has the "most promise" now as a test for ischemia, Dr. Wu says. "Other markers have come and gone—glycogen phosphorylase bb, fatty acid binding protein, p-selectin—none have come to fruition." Most often the stumbling block is specificity—the test doesn’t indicate myocardial ischemia. "This one seems to have gotten further along in development," Dr. Wu says. "Both negative predictive value for rule-out and sensitivity for rule-in appear to be good."

No test will ever be the sole criterion for initiating intervention, Dr. Wu notes. Intervention must be determined in the context of multiple indicators. Eventually, perhaps ST-segment deviation indicative of an ischemic event coupled with a positive ischemic marker might lead to early intervention.

"Maybe half of the angioplasties that we do today are elective," Dr. Wu says. "They are done to reduce future risks based on a variety of clinical factors. Maybe IMA will become part of that list of things to be positive before someone recommends PCI."

William Check is a medical writer in Wilmette, Ill.