Cover Story | Classified: Controversial and Highly Sensitive New assays are raising issuesready for a plague of 'troponinitis?'
By Debra L. Beck
There is a new condition on the horizon that hurts the brain just thinking about it: troponinitis. It's not an inflammation of troponin... but it is fun when scientists get creative with scientific jargon. The term describes a scenario where a high-sensitivity (hs) cardiac troponin immunoassay detects leaking troponin above the 99th percentile of a reference population in a patient not experiencing an acute coronary syndrome (ACS). Translation: when you can measure almost any elevation in troponin, then almost anything can be the problem.
Detectable troponin could be due to myocarditis, Takotsubo cardiomyopathy, heart failure, hypertensive urgency, trauma, cardioversion or cardiopulmonary resuscitation, atrial fibrillation with rapid ventricular response, chronic kidney disease, sepsis, pulmonary embolismor any of several dozen possibilities. Maybe two of the biggest headaches: procedure-related troponin elevations (due to percutaneous coronary intervention, for example, where procedural myocardial infarction [MI] rates may uptick dramatically with hs troponin assays compared to CK-MB assessment) or troponin triggered by increased demand on a stable lesion. When stable coronary artery disease starts to look like ACS to the untrained eye, referral becomes as common as a breeze.
Oh, and under the category "let's-not-even-go-there," there are troponin "elevations" due to analytical causes that are assay-based (due to poor performance or calibration errors) or sample-based (due to the susceptibility of troponin assays to interfering substances, such as heterophile antibodies and rheumatoid factor).
In a lab, it is one thing, but many fear that these assays, once they arrive in US emergency departments (EDs), will create rampant troponinitis, overwhelming inpatient cardiology services with wrongly diagnosed patients or "just-in-case" or "CYA" referrals.
On the other hand, these assays offer tremendous advantages in assessing chest pain and, if used properly, may save lives, reduce ED overcrowding, and spare scant health care dollars. Appropriate use, however, requires a keen understanding of their utility and potential challenges, matched to clinical acumen. Lacking these, they are just "tools for fools."
"I'm very hopeful that the high-sensitivity troponin assays will actually clean up some of the issues with chest pain evaluation in the ED," said L. Kristin Newby, MD, in an interview with CardioSource WorldNews: Interventions. Dr. Newby, a cardiologist at Duke University Medical Center, Durham, North Carolina, was the co-chair of the writing committee for the ACCF 2012 Expert Consensus Document on Practical Clinical Considerations in the Interpretation of Troponin Elevations.1 "I think that they will put more certainty around the rule-out of MI, even if we're not being really careful about how we use them and we use them too often," she said.
While the hs assays are not yet commercially available in the United States, a handful obtained CE mark in Europe: the Roche Elecsys hs cardiac troponin T (hs-cTnT) and Elecsys hs-cTnT STAT, the Abbott ARCHITECT STAT hs cardiac troponin I (hs-cTnI), and the Trinity Meritas hs-cTnI assay. "As soon as the assay was available we took the chance and bought it," said Alexander Ghanem, MD, PhD, in an interview. Dr. Ghanem presented a session at EuroPCR 2014 on the rapid rule-in and rule-out of MI using hs troponin. His expertise arises from his experience as an interventional cardiologist and the head of the cardiology section of the multidisciplinary ED at the University Hospital Bonn, Germany. They see more than 40,000 patients in the ED each year and have had an hs-cTnI assay for more than 2 years.
Over time, he said, his team learned that hs troponin-positive patients may be at risk for something, even if it was not ACS. "In the beginning you give everyone a cath, but after half a year when you've seen many rule-outs, you start thinking about adding in better risk stratification. As soon as we started doing this, we saw the outcomes were good."
One of the most common presenting complaints in the ED, chest pain accounts for up to 20 million ED visits annually in Europe and the United States. Only 10-20% of patients seen for chest pain, however, are diagnosed with MI during their visit. Another 2% or so are actually having an ACS but are inadvertently discharged from the ED.
The timely rule-out of the majority of non-ACS chest pain patients is a health care priorityboth to save time and resources. Because the new hs troponin assays can detect small increases in troponin in the serum several hours before the older assays, hs troponin testing has been proposed as a means to limit long ED stays.
Shorter stays (usually) equal overall cost savings. The hs-cTnI assay used at University Hospital Bonn costs 7 Euros per test (about US$9.50) and is the biggest laboratory and biomarker cost they have; the last generation of assays cost between 2 and 3 Euros each. But safer patient rule-out erases that cost difference and more. "If patients leave the hospital within 2 hours, it costs a lot less than if you have to keep them for several hours, so, in the end, we find that it is cheaper," Dr. Ghanem explained. Although they haven't formally measured it, he suspects their non-ST-elevation myocardial infarction (NSTEMI) patients are probably staying at least a day or two less in hospital because of the new assay and they get to the cath lab earlier.
At ACC.14, Nadia Bandstein, MD, and colleagues at the Karolinska University Hospital, Stockholm, Sweden, presented a late-breaking clinical trial that assessed the use of hs-cTnT to quickly rule out ACS in chest pain presenters with normal electrocardiograms (ECGs).2 They retrospectively assessed a large cohort of 14,636 patients who presented with a principal complaint of chest pain. All patients had at least one hs-cTnT analyzed using the Roche Elecsys 2010 immunoassay and an ECG without signs of ischemia.
In 61% of patients, hs-cTnT was undetectable (<5 ng/L) at the time of their initial test and ECG testing showed no significant ischemic changes. An additional 21% had readings between 5 and 14 ng/L, detectable but below the 99th percentile and therefore not considered "positive," and the remaining 18% had hs-cTnT above the 99th percentile.
During 30-day follow-up, only 39 of 8,907 patients with undetectable hs-cTnT suffered an MI (0.44%)15 (0.17%) of whom had no ischemic ECG changes when first tested in the ED. Eleven were ultimately diagnosed with NSTEMI and four with STEMI. There were two deaths in this group within 30 days.
A first hs-cTnT level of <5 ng/L in combination with no signs of ischemia on ECG had a negative predictive value for MI of 99.8% and an absolute risk of 0.17%. For death, the negative predictive value was 100% and the absolute risk was 0.02%.
According to senior author Martin J. Holzmann, MD, PhD, the "simple strategy" may prevent up to 25% of hospital admissions for chest pain. In their JACC paper, Dr. Holzmann and colleagues noted that 11 of the 15 patients diagnosed with MI, but with undetectable hs-cTnT on presentation combined with no signs of ischemia on the initial ECG, had a first measurement done less than 2 hours after symptom onset. "If a second hs-cTnT level would have been obtained 3 to 4 hours after onset of symptoms, most likely it would have been significantly elevated in most of these patients," he added.
"In the Bandstein study, the risk of future events was well under 1%, which is right in the framework where I think most cardiologists and most ED physicians would tolerate sending people home," said Dr. Newby.
The Karolinska study supports previous work by Than et al., who showed that an accelerated diagnostic protocol with hs troponin (as the only biomarker studied), in conjunction with ECG and the TIMI risk score, could identify a large group of patients presenting with chest pain who are suitable for safe early discharge.3 This trial, called ADAPT, tested hs-cTnI at 0 and 2 hours and found a sensitivity of 99.7%, a negative predictive value of 99.7%, and a specificity of 23.4%. Only 2% of patients deemed negative for ACS underwent procedural intervention. The protocol has subsequently been externally validated.4,5
To be sure, everyone agrees that all of these patients identified as low-risk and released from the ED require further outpatient investigation, including stress testing in most cases. Dr. Newby spoke for many when she noted the hs assays' greatest use is to define a population at a low enough risk that nothing else needs to be done in the ED.
Serial Testing Needed, but Faster
Remember the mnemonic taught in medical schoolSPIN and SNOUT? If a sign, test, or symptom is highly specific, a positive result tends to rule in the diagnosis (SPIN). Conversely, a highly sensitive test, when negative, rules out the disease (SNOUT). Except when it doesn't.
As Dr. Ghanem pointed out in his EuroPCR session, a negative troponin does not always rule out ischemia or infarction. Sometimes the patient presents so early that there has been inadequate leakage of troponin into the serum, and some patients with unstable angina will not produce positive troponin, but remain at risk for significant morbidity and mortality.
So, an hs troponin test stratifies a patient into a lower risk of ischemia or infarction, but over-reliance on an alternative diagnosis can be dangerous, as shown by Campbell et al. when 2.7% of patients with a "clear cut" alternative diagnosis and a TIMI risk score of only 0 or 1 had an infarction after 30 days.6
The nearly 3% risk of adverse cardiac events within 30 days underscores Dr. Ghanem's argument that clinicians should check troponin. Because of the possibility of a false negative, perhaps from too early testing, Dr. Ghanem's hospital still relies on serial testing. In the beginning, they tested hs troponin at 0, 3, and 6 hours in the ED.
However, after gaining more experience and more data, a second test after 3 hours offered 100% detection of MI,7 so they switched to testing only at 0 and 3 hours.
They don't rely on a single troponin assay despite patient claims that chest pain has been gone for 6 hours, because it could be the pain's improved but still present. "There are 5% false negatives with the test when you rely on only one value, and we never want to see a patient discharged who collapses on his way to his car and has to be resuscitated. But if the first test is negative, we don't wait 6 hours anymore," Dr. Ghanem said. He added that in very low-risk patients, they will sometimes only wait 1 hour to re-test.
Beyond their ability to rule out ACS, cardiac troponins are the gold standard of biomarkers to diagnose myocardial injury. In response to ischemia, troponins start to rise within about 2 to 3 hours and can remain elevated for 2 weeks or more. Elevations of cTnT persist longer than cTnI because of its larger molecular weight (37 kDa vs. 24 kDa).
While STEMIs can be readily and quickly diagnosed using an ECG with a prognostic assist from troponins, the diagnosis of NSTEMI remains heavily reliant on troponin measurement. Several research groups have shown that hs troponin measurements on admission offer higher diagnostic accuracy than conventional troponin measurements.7,8
Other studies have compared the effectiveness of hs-cTnT and hs-cTnI for the early diagnosis of acute MI, and found them to be similar.9 The patient presenting with clear clinical signs of ischemia and ST-segment changes is a no-brainer, and the one with a low pre-test likelihood of ACS and undetectable hs troponin doesn't constitute much of a challenge either.
Then there is the middle ground...
The middle ground gets shaky because a positive troponin (>99th percentile) does not diagnose MI due to its low specificity. Indeed, with the hs assays, troponins are detectable in up to 90% of the general population, albeit not above the 99th percentile, but certainly more non-ACS patients will test positive than with the older assays.
As Dr. Newby points out, neither the current generation or the new hs assays will ever be etiologically specific as they measure troponin from cardiac muscle. "As we move towards having these assays we have to understand that they will, by definition, detect troponin in normal patients and in patients who aren't necessarily normal, but aren't having a myocardial infarction. If we can't get our heads around that, these assays will be difficult to manage."
The universal definition of acute MI does not rest on elevated troponin, but rather on whether one sees a characteristic rise and/or fall of cardiac biomarkers of necrosis along with symptoms of ischemia, ECG abnormalities, or delayed diagnosis by nuclear imaging, angiography, or autopsy.
For all their wins, troponin assays should really be considered timekeepers rather than game-changers when it comes to diagnosing MI, said Dr. Newby. "The rules aren't going to change. What will change is that because the assays are so precise and so sensitive, we're not going to have to wait 4 to 8 hours to look for that characteristic rise or fallwe'll be able to see it in a very short period of time."
Dr. Ghanem stressed that clinicians still need to keep their "brains turned on" or the assay becomes a "tool for a fool" if test results are not put into proper context. For example, even a troponin-negative patient who still presents with a high-test probability or who describes a certain clinical course should head to the cath lab.
Although Dr. Ghanem conceded that troponinitis was an issue they had to deal with initially, after 2 years, they are "now very happy to have this very sensitive tool" because they can catch and treat at-risk patients very early, even when their risk is very low.
In Bonn they implemented standardized checklists to ensure that the clinical context is not overwhelmed by the hs troponin results. Thus, if a patient presents with a leading symptom such as angina or dyspnea, the applicable checklist will include questions about other scores, such as TIMI risk or the Geneva score for pulmonary embolism. By adding everything together and determining clinical pretest probability for the differential diagnosis, the context for the troponin result becomes clearer.
Dr. Ghanem, who will soon join the Asklepios Klinik St. Georg in Hamburg, Germany, doesn't really see the lower specificity of the new assays as a downside: false positives for MI in troponinpositive patients signal an alarm"a blinking red light"indicating something else, such as sepsis or stroke. "It's not really false positive, you just have the wrong diagnosis," he said. He added that although the catheterization volume at his hospital might have risen slightly when they first started using the hs assay, their volumes have actually remained stable or even dropped due to the increasing amounts of structural disease they are seeing.
One Size Never Fits All
Why aren't these assays available in the United States? When asked by CSWN: Interventions, Roche Diagnostics North America would only comment that, "Although Roche does not currently have a high-sensitivity troponin assay available in the US market, we are preparing a submission for eventual review by the US Food and Drug Administration. Our timeline for submission is not finalized."
There have been concerns voiced over the reference limits, which are largely derived from small studies of presumably healthy individuals. Inaccurate reference limits will substantially increase the potential for over-diagnosis of MI, and there have been calls for specific age, gender, and racial cutoffs for hs troponins.
When the upper reference limit of 14 ng/L for the hs-cTnT assay was recently applied in three large independent population-based cohorts, Gore et al. "unequivocally" demonstrated that "in a very large and well-characterized population, the 99th percentile values for the hs-cTnT assay are greater in men and rise notably with increasing age in both men and women."10,11
In an attempt to rectify this, the authors proposed revised cutoff values, albeit using retrospectively collected data. Dr. Newby suspects that the FDA may also be waiting for the new assays to be prospectively validated, a standard she agrees with given that assays are not held to prospectively proving outcomes in the biomarker world. "All they have to show is that they measure something, but we haven't prospectively shown that practicing based on a cut point actually works."
In most of the recent studies, many done in Europe, the assays have been tested on populations that were collected prospectively, but analyzed retrospectively.
Although that creates a model, Dr. Newby stressed, the final step of prospective validation remains unfulfilled. "They're exciting data, but what I haven't seen is a good study that takes the algorithm and employs it prospectively and then looks at outcomes related to management decisions using that algorithm." And until then, we need to remain vigilant against a troponinitis epidemic.
- Newby L, Jesse RL, Babb JD, et al. J Am Coll Cardiol. 2012;60:2427-63.
- Bandstein N, Ljung R, Johansson M, Holzmann MJ. J Am Coll Cardiol. 2014 March 25. [Epub ahead of print]
- Than M, Cullen L, Aldous S, et al. J Am Coll Cardiol. 2012;59:2091-8.
- Six AJ, Cullen L, Backus BE, et al. Crit Pathw Cardiol. 2013;12:121-6.
- Cullen L, Greenslade JH, Than M, et al. Am J Emerg Med. 2014;32:129-34.
- Campbell CF, Chang AM, Sease KL, et al. Am J Emerg Med. 2009;27:37-42.
- Keller T, Zeller T, Peetz D, et al. N Engl J Med. 2009;361:868-77.
- Reichlin T, Hochholzer W, Bassetti S, et al. N Engl J Med. 2009;361:858-67.
- Gimenez MR, Twerenbold R, Reichlin T, et al. Eur Heart J. 2014 May 19. [Epub ahead of print]
- Gore MO, Seliger SL, Defilippo CR, et al. J Am Coll Cardiol. 2014;63:1441-8.
- Kramer CM. J Am Coll Cardiol. 2014;63:1449-50.
< Back to Listings