Falling Through the Cracks on Rapid Troponin Testing Protocols

Transitioning high-sensitivity troponin testing protocols in the emergency department for suspected acute coronary syndrome can have some unintended consequences, studies showed.

In a prospective multicenter trial from Australia, patients randomized to a 0/1-hour high-sensitivity cardiac troponin-T (hs-cTnT) protocol unmasked to the care team had similar all-cause mortality and MI at 12 months as those whose early levels were masked and were treated based on the standard protocol of 0/3-hour testing (5.0% vs 3.8%, HR 1.32, 95% CI 0.95-1.83).

However, for those with an intermediate elevation in troponin of less than 29 ng/L — the only group for whom results differed between the two trial arms, with reporting to <5 ng/L in the 0/1-hour group vs ≤29 ng/L with standard care — the rapid protocol did appear to be a disadvantage.

The 0/1-hour unmasked group had a relative 60% higher risk of death or MI (3.7% vs 2.3% at 1 year) compared with the masked group, a difference with statistical significance both for interaction term and after adjustment for age, diabetes, prior cardiovascular events, and other factors, reported Derek Chew, MBBS, MPH, PhD, of Flinders University in Adelaide, South Australia, at the virtual American College of Cardiology (ACC) meeting.

Type 1 MI and unstable angina accounted for a large proportion of these events. Rehospitalization over the subsequent 12 months for non-coronary cardiovascular causes, like arrhythmia, was similar between groups and when stratified by early modest elevation in troponin.

In the overall cohort, though, the rapid protocol also reduced functional stress testing and increased coronary angiography and revascularization significantly.

“The observations of excess harm in the unmasked troponin group remain exploratory,” Chew told attendees at the late-breaking clinical trial session. “However, I would point out this represents 91% of the population this study enrolled, and there are really no significant differences in the baseline characteristics in the groups less than 29 [ng/L].”

Study Details

This study, published simultaneously in Circulation, took advantage of changing protocols across a network of four public hospital emergency departments participating in a registry to dig into the impact of what has been sold as an advantage of the high-sensitivity tests.

During this period when both the 0/1-hour and 0/3-hour results were recorded electronically, but the 1-hour results were not yet routinely provided to clinicians, Chew’s group randomized patients to have that information provided or not provided to their care team. They used electronic medical records and a statewide data linkage process to capture every patient’s data.

The study included 3,270 participants with 12-month follow-up after presenting with suspected acute coronary syndromes and no ECG evidence of coronary ischemia.

“It does look like knowing [about] modest elevations in troponin does drive up the use of coronary revascularization, and previous studies have shown that invasive strategies in those with low level troponin elevations does not reduce death and myocardial infarction,” Chew noted during an ACC press conference.

“Changes in practice associated with the implementation of this protocol may be associated with an increase in death and MI among those with newly identified troponin elevations,” the group concluded. “More sophisticated clinical decision-support approaches may improve outcomes and need to be prospectively validated in clinical practice.”

Generalizability

Chew said that U.S. emergency departments are largely using 0/3-hour protocols with the high-sensitivity tests. “The key question is how your emergency physician is then linked to subsequent testing,” he told reporters. “It depends on what you do with those patients with modest elevations,” and the 0/3-hour protocols actually have more of those patients.

“Certainly, clinicians are continuing to struggle about the application of high-sensitivity cardiac troponin T assays across a variety of clinical scenarios,” said ACC session study discussant Patrick O’Gara, MD, of Brigham and Women’s Hospital in Boston.

O’Gara cautioned, though, against taking away too much from this trial in which there were so relatively few events. There were 82 and 62 in the unmasked and masked groups overall, and 55 and 34, respectively, in those with initial troponin concentrations ≤29 ng/L.

The suggestion that downstream test ordering was to blame would be a “tough nut to crack,” O’Gara said. “Certainly when you have a population of patients in which 20% have previously established coronary artery disease — and in my anecdotal observations, and perhaps other panelists might share this — very little seems to trump the fear of missing something when patients come to the hospital with chest pain, even though it is conceived to be low risk on the basis of clinical markers alone.”

The question is whether coronary investigation can be uncoupled from revascularization for patients with very low troponin levels, Chew said. CT angiography might be the answer, he noted, pointing to its greater availability in the U.S. than Australia.

U.S. Studies

Two U.S.-based studies reported at the ACC meeting and simultaneously published in the Journal of the American College of Cardiology illustrated some of the challenges of implementing hs-cTnT.

One was an observational study comparing patients who presented when conventional cTnT assays were used and those presenting a year later after implementation of hs-cTnT.

“In full agreement with earlier studies, the use of hs-cTnT was associated with an increase in [acute] MI diagnosis, particularly type 2 MI, as well as myocardial injury, and a modest reduction in stress testing,” summarized Christian Mueller, MD, of University Hospital Basel in Switzerland, and colleagues in an accompanying editorial.

“However, also two rather surprising findings regarding resource use emerged,” they added. “Post-implementation there was only a minimal reduction in overall length of ED stay (LOS, median 4.3 vs. 4.2 h) and there was no overall difference in the proportion of patients directly discharged from the ED (54% vs 56%).”

The other study was a difference-in-differences analysis to quantify downstream cascades after implementation of the hs-cTnT assay among patients presenting with chest pain relative to patients presenting with other symptoms. It showed more net upfront tests with the more sensitive test but fewer stress tests, catheterizations, cardiology evaluations, and hospital admissions in chest pain patients compared with others.

The takeaways from Mueller’s group were that “the magnitude of the operational benefits observed with the implementation of hs-cTnT/I is reduced, if the full potential of these assays is not used” and that using formal multivariate risk scores likely further reduces the efficacy of the high-sensitivity assays compared with rapid protocols that no longer use those risk scores.

They recommended widespread adoption of the European Society of Cardiology’s 0/1-hour algorithm in North America.

“Further research as well as continuous interdisciplinary medical education is required to maximize the medical and economic value of hs-cTnT/I-testing to patients and institutions,” the editorialists concluded.