High-Sensitivity Troponin Assays May Help Cut Down on Unnecessary Stress Tests, Study Suggests

The real question, however, is whether surveillance of any type is needed in a population with stable CAD, Pamela Douglas says.

High-Sensitivity Troponin Assays May Help Cut Down on Unnecessary Stress Tests, Study Suggests

Among patients with stable CAD, very low levels of cardiac troponin I detected by high-sensitivity assays may help identify those who are unlikely to have inducible myocardial ischemia and thus reduce unnecessary stress testing, a new study indicates.

A troponin level below 2.5 pg/mL had a negative predictive value (NPV) of 90% for excluding patients with any inducible ischemia in a derivation cohort, lead authors Muhammad Hammadah, MD, and Jeong Hwan Kim, MD (Emory University School of Medicine, Atlanta, GA), and colleagues report in their paper published online November 5, 2018, ahead of print in Annals of Internal Medicine. The NPV was 97% when looking for ischemia affecting at least 10% of the myocardium, with similar findings in a separate validation cohort.

Moreover, no patients with a troponin level below 2.5 pg/mL in the derivation cohort had an MI or died from cardiovascular causes during a median follow-up of 3 years, compared with 7% of those with levels above the cutoff.

The findings could have implications for the use of stress testing, which is overused in patients with stable CAD, senior author Arshed Quyyumi, MD (Emory University School of Medicine), told TCTMD. Despite the fact that guidelines recommend against routine surveillance stress testing in the absence of new or progressive symptoms, stress testing remains common in this population. In fact, the investigators point out in their paper, a prior study suggested that of about 3.8 million stress tests performed in the United States each year, 30% of those with imaging and 14% of those without it were done inappropriately.

Previous research has shown that higher levels of troponin I detected by high-sensitivity assays are associated with inducible myocardial ischemia, and the idea put forth by the current study, Quyyumi said, is that unnecessary stress testing can be reduced by identifying patients with stable CAD who have low troponin levels.

Commenting for TCTMD, Pamela Douglas, MD (Duke University, Durham, NC), said this study doesn’t definitively show whether measuring troponin with high-sensitivity assays would be beneficial in this population, pointing out that there is not robust evidence that inducible ischemia on stress testing leads to events in the future.

“I think the real question that needs to be answered, which this study does not answer, is whether surveillance is better than no surveillance of any kind,” Douglas said, noting that prior work from her group has shown that surveillance leads to more revascularization but no reduction in MI or other events.

Those first two questions—whether inducible ischemia is the right endpoint and whether surveillance is needed at all—would need to be answered before moving onto the next question of which test, stress testing or measurement of high-sensitivity troponin, is better, she said.

Overall, Douglas was skeptical of the concept of measuring troponin to reduce unnecessary stress tests, a problem that has already been mitigated somewhat in recent years. “Then you’re doing unnecessary blood tests, which arguably is better [than doing a stress test]. It’s less expensive. It’s less of a hassle. But if it’s not necessary, it’s not necessary,” she said.

Ruling Out Inducible Ischemia

The analysis included 589 patients with stable CAD in a derivation cohort. In this cohort, the mean age was 63 years, and 76% of patients were men. Two-thirds of patients had abnormal findings on angiography, 55% had undergone PCI, 36% had had a prior MI, 32% had undergone CABG, and 13% had abnormal results on stress testing. Most patients were asymptomatic, with 28% reporting any chest pain in the last 30 days.

If it’s not necessary, it’s not necessary. Pamela Douglas

In all, 35% of patients had a reversible perfusion defect, indicating inducible myocardial ischemia, on myocardial perfusion imaging with technetium-99m-sestamibi single-photon emission CT during either treadmill or pharmacologic stress testing.

Median resting troponin level measured with the Architect STAT Troponin I assay (Abbott) within a week of the stress test was higher in those with inducible ischemia (5.4 vs 3.9 pg/mL).

A troponin cutoff of 2.5 pg/mL was chosen because it provided the highest NPV with a minimum sensitivity of 90%. Levels below this threshold were detected in 17% of the cohort. The performance of the 2.5-pg/mL cutoff in terms of NPV was relatively consistent across subgroups, although higher NPVs were seen among patients without a resting perfusion deficit.

In a separate validation cohort of 118 patients who had survived an MI in the past 6 months, 24% had inducible myocardial ischemia on a stress test. About one-quarter of the cohort had a troponin level below 2.5 pg/mL, and this excluded patients with inducible ischemia with an NPV of 88% for any ischemia and 94% for ischemia affected at least 10% of the myocardium, similar to the performance seen in the derivation cohort.

The investigators note that using that cutoff does not provide high specificity or positive predictive value. “Thus,” they write, “high-sensitivity cardiac troponin I levels are not useful for determining whether inducible myocardial ischemia is present, and if clinical suspicion exists, the patient should probably still have conventional stress testing.”

Guiding the Decision to Stress Test

Even though there are recommendations in the form of guidelines or appropriate use criteria to aid physicians in making decisions about cardiac stress testing, it remains overused in patients with stable CAD. Quyyumi said it’s often done because physicians may have questions during follow-up in patients who have undergone PCI or CABG about whether disease has progressed and may resort to performing stress testing even in the absence of symptoms.

“Thus,” the study authors say in their paper, “an unmet need exists for more effective decision support tools to help physicians identify patients with stable CAD who are likely to benefit from cardiac stress testing.”

Based on this study, they point out, a troponin I level below 2.5 pg/mL detected by a high-sensitivity assay picks up a patient subset that is unlikely to have inducible ischemia on stress testing or to have clinical events during follow-up.

“Certainly, below that level, despite having a history of coronary artery disease you are very safe in terms of adverse events,” Quyyumi said. Before using this approach in practice to help guide decisions about stress testing, however, the findings should be reproduced in a larger cohort, he added.

If the performance is confirmed, “then hopefully people will start to use this as a way of triaging people with stable coronary artery disease,” Quyyumi said.

But Douglas took different messages away from the study.

“In a nonacute setting in patients with known coronary disease, there’s actually a range of troponin values within the normal range that predict excess events, so there’s something that’s higher risk about these individuals and I think that prognostic information is actually pretty interesting and important,” she said.

“Similarly, within the normal range, there’s a prediction of inducible ischemia,” she continued. “Whether or not that translates into a need for additional investigation or intervention I think is still not known. And whether or not you can propose surveillance with high-sensitivity troponins as an effective strategy to reduce events is also not known.”

Todd Neale is the Associate News Editor for TCTMD and a Senior Medical Journalist. He got his start in journalism at …

Read Full Bio
  • Hammadah M, Kim JH, Tahhan AS, et al. Use of high-sensitivity cardiac troponin for the exclusion of inducible myocardial ischemia: a cohort study. Ann Intern Med. 2018;Epub ahead of print.

  • The study was supported by grants from the National Institutes of Health.
  • Hammadah, Kim, Quyyumi, and Douglas report no relevant conflicts of interest.