A novel risk-assessment tool uses troponin levels at serial high-sensitivity (hs-troponin) assays to rule in or rule out myocardial infarction (MI) in the emergency department (ED) based on changing levels of the biomarker, without relying on fixed cut-point values, a new report says.
For patients ruled out for MI but still with elevated hs-troponins indicating some long-term risk, the calculator can help single out those who would benefit from further noninvasive cardiovascular testing, suggests the analysis published June 26 in the New England Journal of Medicine.
There have been concerns in some quarters that the exquisite sensitivity of hs-troponin assays might lead to an excess of referrals for coronary angiography in patients with suspected MI, observed the senior author, Stefan Blankenberg, MD, University Heart Center, Hamburg, Germany. The assays often detect troponin elevations due to other causes.
But the new assessment tool, now available online, “takes advantage of the high-sensitivity nature of the tests” by predicting risk based on acute dynamic changes in troponin levels according to the time intervals separating the two measurements, Blankenberg told theheart.org | Medscape Cardiology.
By interpreting hs-troponin concentrations that way, rather than as “positive” or “negative” for MI, “you could end up with fewer coronary angiographies,” he said.
Dynamic hs-Tn Changes
“This is a fundamental change in how to handle troponin diagnostics in suspected myocardial infarction,” Blankenberg said. “You don’t act at a specific troponin cut-off. You act according to its dynamics and its positive and negative predictive value. You also get information about 30-day mortality.”
The risk calculator, derived from and validated in cohorts from 15 prospective studies, indicates positive and negative predictive values (PPV and NPV) for MI based on the rate of change in hs-troponin concentrations from the first to a second blood draw, without regard to specific levels of either hs-troponin I or hs-troponin T.
The current report provides algorithms based on a range of studies and populations, and allows different centers to tailor the risk prediction to their own criteria, such as preferred PPV and NPV values, observed James de Lemos, MD, University of Texas Southwestern Medical Center, Dallas.
That may be an advance over similar algorithms based on more limited data, but the principles behind the risk assessment tool “have already been laid out by the individual studies that were combined for it,” said de Lemos, who isn’t associated with the current analysis.
The group’s new calculator, he said, “allows customization of the protocol, which has pluses and minuses. The plus is, if you want better safety, or you want more efficiency, you can customize it.”
Still, the ability to have varied protocols for MI rule-out based on the calculator “could add confusion,” he said, and “create more room for misinterpretation.”
So while the concepts behind it aren’t new, the international collaboration that combined the study cohorts for derivation and validation, and its flexibility in considering different hs-troponin thresholds, are strengths, de Lemos said.
For anyone who had doubts about the individual studies going into the calculator, the “consistency of the data across the studies and their combination here,” similar to a meta-analysis, should “seal the deal.” “It should provide a potentially higher level of confidence for clinical application,” he said.
The report also estimates the composite risk of MI or death from any cause over 1 and 2 years according to hs-troponin concentration at the first blood draw for patients in whom acute MI is ruled out.
In patients with any hs-troponin elevation, “The exclusion of myocardial infarction immediately turns into a question of long-term prognosis. With higher troponin there is a very bad prognosis over years,” Blankenberg said.
Clinicians should investigate such patients further for possible causes of their troponin elevation, such as hypertension, renal failure, or a cardiomyopathy. They need not undergo coronary angiography, at least not routinely, Blankenberg said. But they probably should get echocardiography or cardiac magnetic resonance (CMR) imaging, and perhaps be considered for preventive therapies.
The diagnostic calculator was developed within the Calculation of Myocardial Infarction Risk Probabilities to Manage Patients with Suspicion of Myocardial Infarction (COMPASS-MI) program. Its numbers are derived from more than 22,000 patients compiled from 15 previously studied cohorts of patients with suspected acute MI in 13 countries who underwent serial hs-troponin I or hs-troponin T testing in the ED.
They included 9604 patients contributing to the derivation data set, the report notes, and 13,047 patients in the risk-assessment tool’s validation data set.
All patients underwent troponin testing at presentation and again at an “early” time point of 45 to 120 minutes later, or “late” at > 120 to 210 minutes after the initial test. The hs-troponin assays, used on frozen samples or during the course of care during the studies, used the Architect platform for troponin I (Abbott) and the Elecsys platform (Roche) for troponin T.
Across the entire population of > 22,000 patients, 15.3% were ultimately determined to be positive for acute MI.
Patients at low, intermediate, or high risk for acute MI were identified according to absolute changes in hs-troponin values, using individually selected cut points and by whether the second hs-troponin test was performed early or late after the first.
A patient was low risk for MI if both hs-troponin concentrations were below the preselected cut-off concentrations; a patient was high risk if either the first or second hs-troponin reading was equal to or greater than the preselected cut-off value.
For long-term risk estimates in patients ruled out for acute MI, 7682 matched pairs were created from 8345 patients in the suspected acute MI cohort and 71,150 persons from the population-based Biomarker for Cardiovascular Risk Assessment in Europe (BiomarCaRE) study. Median clinical follow-up for the two overall cohorts were 730 days and 8.0 years, respectively.
Matching was according to age, sex, presence or absence of hypertension, presence or absence of diabetes, smoking status, and presence or absence of dyslipidemia, the report says.
Patients in the acute cohort who had been ruled out for acute MI and had an hs-troponin I exceeding 10-14 ng/L, for example, had a risk of death or MI of 4.8% at 1 year and 8.1% at 2 years, compared with 1.4% and 3.4%, respectively, in the matched population-based cohort.
“These data suggest that the concentration of hs-troponin may be useful as a risk-prediction biomarker as well as a diagnostic test,” the group writes.
The diagnostic risk calculator based on serial blood draws as described in the report is “definitely validated” and ready for use now, Blankenberg said.
Clinicians in Europe are years ahead of the those in the United States with respect to the latest generations of hs-troponin assays. But even in the US, clinicians should be using and applying the principles on which the calculator is based, said de Lemos.
“Anybody that doesn’t use this approach isn’t taking advantage of the assays. Focusing on rapid changes over short periods of time is the way to capitalize on their advantages,” he said.
“So some version of it is the right approach. For US centers that are aiming to implement them, with the purpose of reducing ED overcrowding and shortening the rule-out time, this the way to go, there’s no question.”
N Engl J Med. 2019;380:2529-2540. Report
Blankenberg has disclosed nonfinancial support from Abbott Diagnostics, grants and personal fees from Abbott Diagnostics, and personal fees from Siemens, Thermo Fisher, and Roche Diagnostics. Disclosures for the other authors are available at nejm.org. de Lemos has disclosed receiving grant and consulting income from Roche Diagnostics and Abbott Diagnostics; and consulting for Ortho Clinical Diagnostics.