New, Practical Guidance for Cardiac Troponin

Since cardiac troponin T and I (cTn) testing entered clinical practice in the early 1990s, these assays have become essential tools in assessing patients with suspected acute coronary syndrome (ACS). Yet as the assays improved analytically and research findings bolstered their utility in detecting myocardial necrosis even outside the setting of cardiac ischemia, confusion in the medical community about the proper use and interpretation of cTn elevations has never abated. Recognizing the lingering controversy, the American College of Cardiology Foundation recently convened a panel of experts to consider available evidence and develop a consensus document on practical clinical considerations in interpreting cTn elevations.

The goal of the document is to help clinicians of all specialties make sound use of this important biomarker, according to panel co-chair Robert Jesse, MD, PhD, principal deputy under secretary for health at the Veterans Health Administration in Washington, DC, and professor of medicine at Virginia Commonwealth University in Richmond. "There is great confusion about what troponin results mean. A decade or more ago, these assays were very sensitive to myocardial damage but they weren't terribly sensitive analytically," he said. "With sequential improvements in the assays, their analytical sensitivity to detect lower and lower concentrations of troponin was enhanced and we observed a number of instances in which there would be elevated troponins but the patient was not experiencing a classic myocardial infarction in the sense of plaque rupture, intracoronary thrombosis, or myocardial necrosis. That, coupled with the changing definition of myocardial infarction, caused a great deal of confusion as people adapted to the new terminology." The document was published in the December 2012 issue of the Journal of the American College of Cardiology (J Am Coll Cardiol 2012;60:2427–63).

Jesse went on to explain that over time, guidance on appropriate cutoffs for cTn has changed. The third universal definition of myocardial infarction (MI), published in 2012, called for a rising and/or falling pattern of cTn levels, with at least one cTn measurement exceeding the 99th percentile of a normal reference population with a coefficient of variation ≤10% as an element of diagnosing MI (J Am Coll Cardiol 2012;60:1581–98). "We've sequentially been ratcheting down the cutoffs and each time a better assay comes out there's sort of an equilibration period in which people have to come to grips with what the new assay means relative to the old ways they've been practicing," he added.

The Here and Now

Publication of the consensus document is timely, coming as it has before any fourth-generation, high-sensitivity or ultrasensitive assays with limits of detection 10 times lower than conventional assays have been approved for clinical use in the U.S. Though the panel discussed these assays and the impact they're expected to have when adopted in practice, the consensus document deals predominantly with third-generation assays commonly in use in the U.S. The April 2013 issue of CLN will feature an in-depth review of high-sensitivity assays.

AACC's 2013 president, Robert Christenson, PhD, DABCC, FACB, represented the organization on the panel along with representatives of the American College of Chest Physicians, American College of Emergency Physicians, American Heart Association, and Society for Cardiovascular Angiography and Interventions.

Christenson, an expert on cTn testing as well as a scholar of evidence-based medicine, elaborated on the niche the consensus document fills. "When there's not enough evidence to make clear statements, folks still need to make sense out of what evidence there is," he explained. "The idea is to inform clinicians, payers, and others of the prevailing opinion out there in evolving areas of practice and technology that are widely available or new to the practice community but just don't have a firm enough evidence base to really put out strict guidelines according to what would be acceptable from the evidence-based medicine world." Christenson is a professor of pathology and medical and research technology at the University of Maryland Medical School in Baltimore.

The Misuse of cTn

Members of the panel described their own experiences in dealing with the fall-out from inappropriate ordering or misinterpretation of cTn tests. The problem has many causes, including analytical issues, incorrect test ordering, and misunderstanding of cTn test results. "The exquisite sensitivity of the troponins now requires even more diagnostic interaction by the clinician to be certain what is really causing the elevation. It may not be a clinical MI of any description," said Joseph D. Babb, MD, professor of medicine at East Carolina University Brody School of Medicine in Greenville. "For emergency physicians in the triage business who might have a waiting room full of people to be seen, this biomarker becomes the trigger to refer to a cardiologist, to pass the patient on. Unfortunately—and this is well intended—it leads to a sequence of unnecessary testing that amplifies the cost of care, and that might have been solved with a little more careful thought at the beginning of the process."

Babb was quick to point out that emergency medicine physicians are not the only ones challenged by abnormal cTn results. "I see this problem even among cardiologists. They might, for example, have a patient with a little shortness of breath and an elevated troponin result. I'll get a call saying, 'we need to cath this patient.' My question in this circumstance is 'why,' and the answer often is, 'the patient is troponin-positive.' However, this elevation could be due to heart failure or many other things. So in my view, this guidance is needed across a spectrum of providers, including cardiologists," he explained.

Unlocking the Key to Clinical Context

Given the pervasive potential for misinterpreting cTn results, the panel emphasized the importance of considering cTn results in the context of each patient's clinical picture. "The single important confusion is that not all troponin elevation implies MI. Troponin elevation reflects myocardial necrosis, so it's specific for myocardial necrosis. However, all myocardial necrosis does not reflect MI. There's a proper context in which one evaluates myocardial necrosis, and troponin by itself is not sufficient," said Sanjay Kaul, MD, PhD, attending cardiologist at Cedars-Sinai Medical Center and a professor of medicine at the Geffen School of Medicine at the University of California-Los Angeles (UCLA).

An expert in statistical issues involving evidence-based medicine, Kaul noted the importance of considering the pretest probability of ACS along with other clinical parameters in discerning whether or not a rise in cTn signifies MI. "Bayes theorem for proper use of troponin says that if you have a patient with high pretest likelihood of acute coronary syndrome, even a modest increase in troponin will have increased specificity, meaning the post-test likelihood of having MI is going to be high," he explained. "However, if you have a very low pretest likelihood of having acute coronary syndrome as suggested by the clinical and electrocardiograph features, then even a troponin elevation does not signify that this is a case of MI." The panel proposed an algorithm for using pretest probabilities to interpret cTn elevations (See Online Extra).

cTn in Nonischemic Conditions

The committee also elucidated the many nonischemic causes of cTn release, of which individual clinicians have varying degrees of understanding and for which the level of evidence also varies (See Table, below). Some conditions, such as heart failure, pulmonary embolism, and chronic kidney disease have more or less defined roles in clinical management, like aiding in estimating survival, and determining appropriateness for aggressive therapy or the need for prolonged monitoring during hospitalization.

The case for cTn as a diagnostic and clinical management tool in nonischemic conditions perhaps is strongest when it comes to chemotherapy-associated cardiac toxicity. The panel noted that in patients receiving chemotherapy, cTn positivity at almost any level and at any time during chemotherapy signals "significantly increased" risk of permanent or severe diminution in left ventricular systolic function, and that the magnitude and frequency of cTn elevations correlate with cumulative drug dose from both current and prior regimens.

In other conditions, ranging from subarachnoid hemorrhage and autoimmune diseases to metabolic disorders and pregnancy, cTn elevations have been documented but their significance in practice is unclear, except as a sign of increased risk. "We know that any troponin elevation is an adverse prognostic factor. That applies to a rise or fall that would be a marker of MI and to elevations in other patients who ostensibly may not have any cardiac disease. In all of them it's very reasonable to say that troponin elevations are an adverse prognostic factor. That's become gradually and increasingly clear over the last few years, and is something the consensus document tried to point out," said Bernard Gersh, MB, ChB, DPhil, professor of medicine at the Mayo Clinic College of Medicine in Rochester, Minn.

Gersh and other panelists, including Charles McKay, MD, explained that more research is needed to understand the etiology of and develop responses to cTn release in many nonischemic conditions. "The issue is, whatever is driving some subtle myocardial damage needs to be addressed by risk factor control, underlying disease control, or more aggressive medical therapy. It's disconcerting because we don't understand the causes behind these nonischemic elevations, but it's also an incredible opportunity for research that might lead to interventions that mitigate this risk," he observed. McKay is professor of medicine at UCLA and a cardiologist at Harborview UCLA Medical Center.

Serial Testing: 2, 6, or 9 Hours?

The panel also commented on several issues about which there is conflicting or inconsistent evidence. One involves the best time interval for serial cTn measurements to diagnose MI. A smoldering pattern of elevated but relatively stable levels over a 6–9 hour period more likely can be attributed to chronic diseases such as heart failure or amyloidosis. However, a dynamic change of at least 20% over baseline within 3–6 hours is more consistent with MI. In an effort to improve emergency department throughput and decrease unnecessary treatments without compromising quality, researchers have been evaluating testing protocols as short as 2 hours. Committee co-chair L. Kristin Newby, MD, MHS, noted that while each approach has merits, clear evidence supporting one over another has not surfaced.

"There's pretty good evidence that if you sample out to six or nine hours and the patient does not have elevated levels in the clinical context consistent with ischemia, then you have not picked up myocardial necrosis. In this situation, the negative predictive value of the test result is pretty good," she said. "But whether you believe in a two-hour or six-hour delta, I don't think one has been proven to be better than the other." Newby is a professor of medicine at Duke University School of Medicine in Durham, N.C. and co-director of the Duke Clinical Research Institute's Biosignatures Advanced Biomarkers Group.

A POCT Challenge

The panel also cautioned about the use of point-of-care testing (POCT), particularly for serial cTn measurements. POCT "may be useful as a screening tool, but most [POC] assays are only semiquantitative. To confirm a rise and/or fall from an initially positive assay would require serial quantitative testing, and in general, high-quality quantitative assays are preferred," the committee wrote.

Christenson explained that the panel hoped to roil the waters with this statement. "It throws down the gauntlet to have point-of-care assays demonstrate the same characteristics as the quantitative tests we offer in the lab. There are a couple that have met that challenge. Otherwise, if the test is not as sensitive as troponin analyzed in the lab, patients could be sent home who might be at increased risk," he said. "That language is a bit challenging, and it's intended to be that way."

Analytical Concerns

Whether they realize it or not, analytical issues also play into physicians' interpretations of cTn results. Without a good grounding in the assay being used, they can draw incorrect conclusions, explained McKay. "Not all the assays report the concentration in a standard way. So instead of using ng/L across the board some report pg/mL. So a 0.3 in one hospital with one assay might be entirely different from a hospital down the road because of the different assays used. This can be confusing for clinicians," he said.

Given assay-to-assay differences, McKay suggested one of the most beneficial actions laboratorians can take is to include language on test reports explaining the test characteristics of their cTn assays. "State what the 99th percentile is, and what the coefficient of variation at the 99th percentile is so clinicians understand the variants around the assay and can plot a series of troponin values and understand what the pattern is," he said. "The statement also should add that a rising and/or falling pattern or declining values over three-to-six hours can be helpful in diagnosing myocardial necrosis and that other patterns with chronic elevation are related to non-acute coronary ischemia processes. In the latter case, consultation with cardiology may be indicated. That way clinicians will have serial numbers, clinical context, and the suggestion of expert consultation as warranted."

The National Academy of Clinical Biochemistry in 2007 published a laboratory medicine practice guideline on biomarkers of ACS which covered analytical issues in detail (Clin Chem 2007;53:552–74), and this topic has been the subject of studies and review articles published since then in Clinical Chemistry. Writing as it was for a physician audience, the consensus panel did not dwell extensively on analytical issues. However, it did call for a centrally maintained continuously updated database of assays to encourage improvements in assays, regulations to ensure standardization of cTn I assays to the National Institute of Standards and Technology reference material #2921, and creation of a healthy subject sample bank so that all cTn assay manufacturers could establish the 99th percentile of their assay against a common standard population of uniform size and clinical characteristics.

Laboratorians' Call to Action

All panel members with whom CLN spoke urged laboratorians to invest in educating clinicians about the cTn assays in use in their facilities, whether in one-on-one sessions, grand rounds, or through other means of communication.

"I hope this helps heighten laboratorians' awareness that outside the walls of the lab, people just don't understand these assays. To the extent that laboratorians are really engaged with their clinical teams, they can help educate physicians about the performance characteristics of the assays in use at their hospitals and how to interpret the results," said Newby. "This really is a call to action for laboratorians."

Christenson urged laboratorians to use the consensus document, written with clinicians in mind, to deepen their understanding of the diagnostic challenges physicians face. "This is going to help laboratorians think like clinicians. Our expertise involves ensuring that results are the best possible, accurate, and under control. This is what clinicians rely on us to do. But this document gives a boots-on-the-ground perspective of how physicians think about our test results, incorporating them with the pretest probability and clinical context to rule-in or rule-out disease."

Christenson and other panelists emphasized that now is the time, before the fourth-generation high-sensitivity assays enter clinical use in the U.S., to reach out to and nail down with clinicians the best use and interpretation of cTn results. As McKay observed, "remember, troponin and the issues surrounding it are fast-moving. There's a lot of development going on in this area, and we're about to hit a new era with the high-sensitivity assays. There's enough concern around the existing generation of assays that the high-sensitivity ones are going to put us over the edge if we don't get some convergence now around interpretation of troponin."