The Role of High Sensitivity Troponin Pathways in the Emergency Department

Authors: Akilesh Honasoge, MD MA; Robert Brown, MD; Sharleen Yuan, MD PhD MA
Editors: Kami M. Hu, MD FAAEM and Kelly Maurelus, MD FAAEM
Originally Published: Common Sense May/June 2019

Clinical Question:
Can high sensitivity troponin meaningfully contribute to an acute coronary syndrome evaluation?

Introduction:
Traditional 4th-generation troponin assays are part of the standard cardiac evaluation when there is suspicion of acute coronary syndrome (ACS). Patients presenting to emergency departments with chest pain are typically risk stratified using a combination of historical risk factors, electrocardiogram (ECG) findings, troponin testing, and clinical suspicion. Patients are discharged from the emergency department if their risk is determined to be low while higher risk patients are usually admitted for further observation or testing. Recent data suggests, however, that high sensitivity troponin (hsTrop) testing could supplement or even replace current methods such as clinical risk scores when used in specific protocolized serial testing pathways. High sensitivity troponins first became available in the United States in 2017, but their use has not yet reached widespread acceptance. The test offers detection of cardiac troponin approximately 1,000 times more sensitive than standard 4th-generation cardiac troponin testing. With this higher sensitivity come questions regarding reliability and specificity in certain comorbidities such as chronic kidney disease and underlying coronary artery disease. Some studies also explore the use of computed tomography (CT) coronary angiography when used in conjunction with hsTrop. Here we explore a few studies that evaluate the role of hsTrop in the evaluation of potential ACS.

Chapman A, Hesse K, Andrews J et al. High-Sensitivity Cardiac Troponin I and Clinical Risk Scores in Patients with Suspected Acute Coronary Syndrome. Circulation. 2018;138:1654-65.
The authors of this study sought to determine whether the addition of hsTrop to previously developed clinical risk scores would alter their negative predictive value (NPV) for ruling out ACS in low-risk patients. They evaluated two hsTrop ACS rule-out protocols: the European Society of Cardiology 3-hour pathway (ESC) and the High-Sensitivity Troponin in the Evaluations of patients with Acute Coronary Syndrome (High- STEACS) pathway. For both pathways, myocardial infarction is ruled in if the hsTrop concentration rises above the 99th percentile.

The ESC pathway rules out ACS in patients with symptoms lasting longer than six hours and a non-ischemic ECG if the initial hsTrop level is below the 99th percentile. For patients with symptoms lasting less than six hours, a second hsTrop is collected three hours after presentation, and if it remains below the 99th percentile and changes by less than half the threshold for the 99th percentile, ACS is ruled out.

By comparison, the High-STEACS pathway rules out ACS in patients with symptoms lasting two or more hours, without ischemic changes on ECG, with a hsTrop < 5ng/L at presentation. If the patient is within two hours of symptom onset a second hsTrop is collected at three hours after presentation and if it remains below the 99th percentile and there is no change (Δ <3ng/L), ACS is ruled out. For patients with a presenting hsTrop > 5ng/L but below the 99th percentile, a re-test is performed three hours later and if there is no change (Δ <3ng/L) then the patient is considered ruled out for ACS. Of note, the 5ng/dL threshold to rule out ACS at presentation was derived and then validated using consecutive suspected ACS cases at hospitals in Scotland.

The authors compared the sensitivity and negative predictive value for ACS of these protocols both with and without the following additional clinical risk scores: Thrombolysis in Myocardial Infarction (TIMI), Global Registry of Acute Coronary Events (GRACE), Emergency Department Assessment of Chest Pain Score (EDACS), and the History, ECG, Age, Risk Factors, Troponin (HEART) score.

This was a prospective observational cohort study of 1,935 consecutive suspected ACS cases at a tertiary care center in Scotland. The primary outcome was the composite outcome of myocardial infarction (MI) due to ACS or death due to MI, dysrhythmia, or heart failure at either presentation or 30 days later. Diagnoses were adjudicated by 2 independent cardiologists. The authors achieved 100% follow up via individual patient calls as well as regional and national registries.

The primary outcome occurred in 276 patients (14.3%). The ESC 3-hour pathway NPV was 97.9% (95% CI: 97.1-98.6) but with a sensitivity of only 89.9% (95% CI: 86.3-93.4), corresponding to a pathway-based rule-out of 70% of patients, with 27 missed events. When combined with clinical decision rules there was a significant (p<0.001) improvement of the NPV (99% or greater in each case) but the proportion of patients characterized as “low risk” decreased dramatically to 25% with the addition of the HEART score, 42% with EDACS, 43% with TIMI, and 49% with GRACE.

The High-STEACS pathway ruled out a similar proportion of patients (65%) but with a better NPV (99.7% CI: 99.4-99.9) and sensitivity (98.7%, CI: 97.4-99.8), corresponding to 3 missed events. When combined with clinical decision tools there was no significant improvement in the NPV but the proportion of patients able to be categorized as low-risk fell significantly to 24.3% for the HEART score, 41% for EDACS, 44% for TIMI, and 47% for GRACE with p<0.001 for each.

Although the High-STEACS protocol outperformed the ESC pathway, there are important limitations to this data. The actual High-STEACS pathway was derived from the first 1,218 patients used for this study, which derived its patients from a single large academic center. Therefore, it is without broad multi-centered validation. The study also notes that there is a high coefficient of variance surrounding the reference limits of the hsTrop testing of about 12.6% across 33 devices. This study also requires replication to ensure the generalizability of using a 5ng/L cutoff for hsTrop.

The High-STEACS highly sensitive troponin pathway appears to offeran effective method of ACS evaluation that is not significantly improved when combined with clinical risk score such as the HEART score. However, hsTrop pathways require more studies with different study populations, further specification of reference limits, and actual data on the effects of utilizing said pathways before there can be widespread adoption.

Kraus D, von Jeinsen B, Tzikas S, et al. Cardiac Troponins for the Diagnosis of Acute Myocardial Infarction in Chronic Kidney Disease. J Am Heart Assoc. 2018;7:e008032.
Chronic kidney disease (CKD) is associated with an increased risk for myocardial infarction, but the utility of cardiac troponins in evaluating for ACS is hindered by the fact that patients with CKD often have elevations in troponin even in the absence of myocardial ischemia. In this study, the authors hypothesized that although the performance of high sensitivity troponins I and T (hereby referred to only as “hsTrop”) to diagnose ACS is confounded by the presence of CKD, dynamic changes in hsTrop may outperform static cutoffs in the diagnosis of NSTEMIs in CKD patients. The authors proposed an algorithm using hsTrop levels to increase specificity in CKD patients.

The authors included two patient cohorts including over 8,500 patients in their analysis. The first was a prospective cohort that enrolled 1,818 patients with suspected ACS. They excluded dialysis-dependent patients but included 280 with CKD. The second was a retrospective cohort of patients with hsTrop testing in clinically-suspected ACS. Dialysis patients were again excluded, and it was narrowed to 5,478 patients without CKD and 1,581 patients with CKD after exclusion criteria. Important exclusion criteria included dialysis dependence, pregnancy, recent surgery, IV drug abuse, and anemia. The authors used this data to calculate the appropriate initial and 3-hour hsTrop level that would optimize both positive predictive value (PPV) and negative predictive value (NPV) for ACS in patients with CKD. The final diagnosis of ACS was made by two independent cardiologists after thorough chart review.

For the prospective cohort, the average GFR of patients with CKD was 46 mL/min/1.73 m2, while in patients without CKD the GFR was 85 mL/min/1.73 m2. The prospective cohort resulted in an NSTEMI rate of 17% (without CKD: 15%, with CKD: 26%). The retrospective cohort had a similar 19% overall rate of NSTEMI (without CKD: 17%, with CKD: 28%). The high sensitivity troponin levels were overall higher in patients with CKD (measured in ng/L). In the prospective cohort, for example, the initial hsTrop on average was 5.[6] (CI: 3.0-18.2) in patients without CKD and 14.7 (CI: 6.5-76.9) in patients with CKD.

The authors derived optimized cutoffs for both absolute and relative changes for their hsTrop algorithms. The two main thresholds used were the 99th percentile and another “very high” level. This “very high” hsTrop level was calculated to reach an equivalent specificity for ACS in CKD patients when compared to an initial hsTrop in a patient without CKD based on the 99th percentile cutoff. This “very high” level maintains the sensitivity of hsTrop I while dropping the sensitivity of hsTrop T. Based on these cutoffs, the authors established three different algorithmic criteria upon which to rule ACS in or out, making the following statements:

If the initial hsTrop level at presentation falls below the 99th percentile with no more than a 2.8x increase in hsTrop after three hours, ACS in patients with CKD can be ruled out (NPV=98%, CI=94-100%). Any increases past this level should prompt at least inpatient observation.

If the initial hsTrop is above the 99th percentile and increases more than 2.8x for hsTrop I after three hours, ACS is ruled in (PPV=89%, CI=65- 99%). If they increase after three hours, but do not exceed the 2.8x cutoff, ACS is still possible, and the patient should be observed.

If the initial hsTrop level is above the very high level, ACS is ruled in without need for further troponin measurement in the emergency department (PPV=74%, CI not given).

Based on these cutoffs, when applied to the two cohorts being studied, the authors found the rule out approach described above had a sensitivity of 100% (CI=94-100%) for the diagnosis of ACS.

Limitations of the study include its use of initial serum creatinine at time of measurement of hsTrop rather than evaluation of the patient’s baseline creatinine clearance, which may damage the inherent validity of this study, as conditions like acute kidney injury offer a different pathophysiology and inherent ACS risk than chronic disease. The determination of final diagnosis of ACS was made by independent cardiologists in the prospective cohort, however the retrospective cohort established the final diagnosis of ACS from the patient’s evaluating provider. The study cohorts also excluded patients on dialysis, limiting the generalizability to that population.

Using a cutoff for hsTrop change, rather than absolute values, in the ACS evaluation of patients with CKD may have a place at some point in the future, however there is not enough data to make an adequate assessment on the validity of this approach at this time.

Ferencik M, Liu T, Mayrhofer T, et al. hs-Troponin I Followed by CT Angiography Improves Acute Coronary Syndrome Risk Stratification Accuracy and Work-Up in Acute Chest Pain Patients: Results from ROMICAT II Trial. JACC Cardiovasc Imaging. 2015;8(11):1272-81.
The ROMICAT II trial (Rule Out Myocardial Infarction/Ischemia using Computer Assisted Tomography) explored the effects of supplementing standard troponin testing with standard coronary computed tomography angiogram (CTA) to reduce hospital admissions and further testing. This study is a nested observational cohort of the larger ROMICAT II trial that evaluates the outcomes of combining hsTrop testing with advanced coronary CTA.

The study population included patients aged 40-75 years with at least two cardiac risk factors, in sinus rhythm, who presented to the ED with at least five minutes of chest pain or an anginal equivalent concerning for ACS. Notable exclusion criteria include a positive initial standard troponin, serum creatinine >1.5mg/dL, or an ECG with diagnostic ischemic changes such as ST elevations or depressions >1mm or T wave inversions >4 mm in two anatomically adjacent leads. This study narrowed the 501 members of the CTA arm of the ROMICAT II trial to those who also had hsTrop testing, which lead to a total study population of 160 patients.

The patients were divided into three main categories based on their initial hsTrop: low risk (below the level of detection, <0.5pg/mL), intermediate risk (0.5-0.49pg/mL), and high risk (above the 99th percentile, >0.49pg/mL). Coronary CTA findings stratified levels of coronary stenosis to define presence of coronary artery disease (CAD) defined as: no CAD (0%), non-obstructive CAD (1-49%), and obstructive CAD (>50%). Advanced evaluation of coronary CTA analyzed high-risk features of coronary artery plaque (remodeling index >1.1, plaque with low CT attenuation <30 HU, napkin-ring sign, and spotty calcium) independently of the level of underlying coronary artery stenosis. The eventual diagnosis of ACS was based on an independent committee review of the patient’s hospital or after-discharge course.

The overall presence of ACS in this cohort was 11.9%. In patients stratified by hsTrop level, 5.6% were deemed to be low risk and 7.5% were deemed to be high risk, with a 0% and 58.3% rate of diagnosed ACS respectively, independent of CTA testing. The 86.9% of patients with intermediate-risk hsTrop levels had an ACS rate of 8.6%. All patients with an eventual diagnosis of ACS had at least one high-risk feature on advanced CTA independent of the presence of obstructive CAD. No patient with an absolutely negative CTA (no CAD and no high-risk plaque) developed ACS, while patients with both obstructive CAD and high-risk plaque had an ACS rate of 69.2%.

The authors proposed an ACS evaluation algorithm that first stratifies patients based on initial hsTrop and associated rate of ACS (noted in parentheses): see Figure 1.

1. Patients with low-risk hsTrop levels (ACS=0%) would be deemed appropriate for ED discharge without a need for a coronary CTA.
2. Patients with intermediate-risk hsTrop levels (ACS=8.6%) would subsequently receive an advanced coronary CTA for further risk stratification. If no obstructive CAD and no high-risk plaques are found (ACS=0%), the patient is low risk. If either CAD or high-risk plaque is found, they are defined as intermediate risk (ACS=7.7%). If both are found the patient is deemed to be high risk (ACS=69.2%).
3. Patients with high-risk levels (ACS=58.3%) would go on to further inpatient evaluation without need for additional ED risk stratification with coronary CTA.

Figure 1: Combined hsTrop and Advanced Coronary CTA algorithm for ED evaluation of ACS derived from the ROMICAT II Trial[7]

Ferencik M, Liu T, Mayrhofer T, et al. hs-Troponin I Followed by CT Angiography Improves Acute Coronary Syndrome Risk Stratification Accuracy and Work-Up in Acute Chest Pain Patients: Results From ROMICAT II Trial. JACC Cardiovasc Imaging. 2015;8(11):1272-81.
The author’s overall conclusion is that use of hsTrop and advanced coronary CTA maintains the 100% (95% CI: 82.4-100.0%) sensitivity of ACS when compared to standard troponin and standard coronary CTA. However, it improves the specificity from 48.2% (95% CI: 39.7-56.8%) to 68.1% (95% CI: 59.7-75.7%).

This study is a small study (n=160) with no direct control or comparison group. Its generalizability is limited by excluding patients not in sinus rhythm and with chronic kidney disease, which may confound the hsTrop level determinations, even if it is sensible that physicians are not likely obtaining coronary CTAs in patients with CKD.

The use of sequential high sensitivity troponin and coronary CTA with assessment of advanced plaque features has potential to be an improved objective tool for stratifying ACS risk, potentially allowing discharge of patients who would otherwise fall into the intermediate risk category. Evaluation of high-risk plaque features on coronary CTA has the potential to change our approach to ischemia evaluation via stress testing and potentially even invasive angiography. Additional studies are required, however, to validate this proposal, and physicians will need to keep in mind the cohort of patients excluded due to limitations in the use of coronary CTA.

Shah ASV, Anand A, Strachan FE, et al. High-sensitivity troponin in the evaluation of patients with suspected acute coronary syndrome: a stepped-wedge, cluster-randomised controlled trial. Lancet. 2018; 392(10151):918-28.
The High-STEACS investigators sought to determine whether utilization of a hsTrop assay with a sex-specific threshold above the 99th percentile would decrease the rate of subsequent MI or cardiac death in patients with suspected ACS. A cluster-randomized, stepped-wedge trial across ten hospitals in Scotland over three years, the study included patients presenting to the ED with suspected ACS, collecting both a standard and high-sensitivity troponin at initial presentation and again at six or 12 hours after the onset of symptoms, at the attending physician’s discretion. The study was broken up into a six-month validation phase during which treating physicians were blinded to the hsTrop level, and an implementation phase during which the hsTrop level was known but the standard troponin level was masked to treating physicians.

The primary outcome was subsequent MI (type 1 or 4b) or cardiovascular death within a year of initial presentation to the hospital. Secondary outcomes were duration of hospital stay, type 1 or 4b MI, unplanned coronary revascularization, hospital admission for heart failure, ischemic stroke, major hemorrhage, unplanned hospital admission, and all-cause death. Diagnosis of MI was made by two physicians blinded to the study phase who determined that there was myocardial injury (hsTrop level above the 99th percentile) in the context of suspected ACS with consistent symptoms or signs on ECG or stent thrombosis at angiography.

A total of 48,282 patients were enrolled, with 10,360 (21%) having hsTrop concentrations above the 99th percentile and meeting criteria for myocardial injury at presentation. 1,771 (17%) patients had not been initially identified by the standard troponin assay and were reclassified based on the hsTrop. At one year, 2,586 (5%) patients overall had MI or cardiovascular death. Patients who demonstrated myocardial injury by troponin level were more likely to meet the primary outcome, although there was no difference in rate of primary outcome between implementation or validation phase. Of the 1,771 patients reclassified with myocardial injury by hsTrop, only a third were diagnosed with subsequent myocardial infarction, but these patients were more likely to undergo coronary angiography (11 vs 4%) without increase in percutaneous coronary intervention, be prescribed additional anti-platelet therapy, ace inhibitors, statins, and/or beta-blockers, and had a longer hospital stay. In patients without evidence of myocardial injury, the duration of hospital stay was decreased from a median of seven to four hours.

Some caveats to the study findings include its before-and-after study design, which may include inherent confounding from changes in practice or treating physicians at the various locations over time. Also, some argue that the benefits of additional medical therapy may not be seen within a year’s follow-up, leading to premature acceptance that there was no decrease in primary outcome. The authors also comment that they did not make adjustments to the threshold for renal dysfunction. Identification of higher risk patients with positive hsTrops did not translate to improved outcomes in this study, but the data leaves room to speculate on improved ability to discharge from the ED.

Conclusion:
The use of high sensitivity troponin pathways or protocols appear to circumvent some of the pitfalls of isolated hsTrop testing. The current data suggests a higher number of patients may be able to be discharged safely from the emergency department which may be able to lower health care costs, but the high sensitivity of the test puts physicians and patients at risk of increased testing, medication, and cost without apparent benefit. Research suggests that a result below the level of detection can effectively rule out ACS in emergency department visits regardless of underlying historical risk factors, however the small percentage of patients in which these levels occur limit the test’s utility in busy EDs, where confounding factors such as the presence of chronic kidney disease lead to an elevated level of high sensitivity troponin regardless of the patient’s underlying cardiac pathology. A variety of testing algorithms appear to suggest that serial testing in the emergency department with a high sensitivity troponin or use in conjunction with other cardiac testing such as advanced coronary CTA can lead to even better specificity than traditional methods while maintaining tremendous sensitivity for acute coronary syndrome. Serial testing algorithms may allow EM physicians to eliminate the subjectivity of currently accepted clinical risk stratification scores, such as the HEART score while combined use with coronary CTA may be able to recategorize intermediate risk patients to allow for discharge from the emergency department rather than admission. For the moment, however, the lack of an established laboratory standard, limited prospective data, and current ease of use and widespread acceptance of the HEART score limit the current implementation of the high sensitivity troponin.

Answer:
Current data on high sensitivity troponin pathways suggest they may be able to offer a more simplified and objective approach to ACS evaluation. However, due to the confounding factors of hsTrop, these pathways are not ready for general use and much more data is needed before widespread adoption can be considered.

References:

1. U.S. Food and Drug Administration. 510(k) Premarket Notification. https://www.accessdata.fda.gov/scripts/cdrh/cfdocs/cfPMN/pmn.cfm?ID=K162895. Accessed February 19, 2019.

2. Giannitsis E, Kurz K, Hallermayer K, Jarausch J, Jaffe AS, Katus HA. Analytical validation of a high-sensitivity cardiac troponin T assay. Clin Chem. 2010;56(2):254-61.
 
3. Roffi M, Patrono C, Collet JP, et al. 2015 ESC Guidelines for the management of acute coronary syndromes in patients presenting without persistent ST-segment elevation: Task Force for the Management of Acute Coronary Syndromes in Patients Presenting without Persistent ST-Segment Elevation of the European Society of Cardiology (ESC). Eur Heart J. 2016;37(3):267-315.

4. Chapman AR, Anand A, Boeddinghaus J, et al. Comparison of the Efficacy and Safety of Early Rule-Out Pathways for Acute Myocardial Infarction. Circulation. 2017;135(17):1586-1596.
 
5. Shah AS, Anand A, Sandoval Y, et al. High-sensitivity cardiac troponin I at presentation in patients with suspected acute coronary syndrome: a cohort study. Lancet. 2015;386(10012):2481-8.
 
6. U.S. National Library of Medicine. Multicenter Study to Rule Out Myocardial Infarction by Cardiac Computed Tomography (ROMICATII). https://clinicaltrials.gov/ct2/show/study/NCT01084239. Accessed February 19, 2019.
 
7. Ferencik M, Liu T, Mayrhofer T, et al. hs-Troponin I Followed by CT Angiography Improves Acute Coronary Syndrome Risk Stratification Accuracy and Work-Up in Acute Chest Pain Patients: Results From ROMICAT II Trial. JACC Cardiovasc Imaging. 2015;8(11):1272-81.
 
8. Thygesen K, Alpert JS, Jaffe AS, et al. Third universal definition of myocardial infarction. Circulation. 2012;126:2020-35