Author + information
- ↵*Reprint requests and correspondence: Dr. Elliott M. Antman, Cardiovascular Division, Brigham and Women’s Hospital, 75 Francis Street, Boston, Massachusetts 02115
Current nomenclature categorizes patients with ischemic discomfort into those who present with ST elevation on the 12-lead electrocardiogram (ECG) versus those who do not present with ST elevation (1). Patients presenting with ST-segment elevation are, of course, easy to recognize using the ECG—a simple, inexpensive tool that is available in all emergency departments. The majority of such patients will develop a Q-wave on the ECG and will ultimately be diagnosed as having sustained a Q-wave myocardial infarction (MI). Patients who present without ST-segment elevation are experiencing either unstable angina (UA) or a non–ST-segment elevation MI (NSTEMI) (1). Most patients who present with NSTEMI do not develop a Q-wave on the ECG. The distinction between UA and NSTEMI is made based on the presence or absence of a cardiac marker in the blood. When such a cardiac marker is detected in a patient’s blood, the patient is ultimately diagnosed as having a non–Q-wave MI.
Biochemical markers of myocyte necrosis are useful not only for making the diagnosis of an MI but also for estimating prognosis. Of interest are the macromolecules that diffuse out of necrosing myocytes as membrane integrity is lost. Once outside the myocyte, the macromolecules are cleared from the interstitium by cardiac lymphatics (1). Eventually, when the capacity of the lymphatics to clear the macromolecules is exceeded, the markers become detectable in the peripheral circulation. Both clinical chemistry laboratory assays and bedside assays are now available to measure several biochemical cardiac markers, notably myoglobin, the MB fraction of creatine kinase (CK-MB) and the cardiac-specific troponins T and I (cTnT and cTnI).
Although myoglobin and CK-MB are familiar to physicians and provide reasonable sensitivity for detection of MI, they lack specificity in the setting of skeletal muscle disease or injury. Hence, there is intense interest in cTnT and cTnI as markers that are found in high concentration in the myocardium and are released with a stoichiometric relationship proportional to the amount of myocardial injury (1,2). Monoclonal–antibody-based assays are available for cTnT and cTnI that capitalize on the fact that the amino acid sequences of the cardiac and skeletal muscle forms of troponins T and I are sufficiently dissimilar. After much debate and discussion, expert panels have declared cTnT and cTnI to be the preferred biomarkers for detection of myocardial damage (1,3).
Myocardial necrosis (i.e., MI) is said to be present if the maximal concentration of cTnT or cTnI exceeds the decision limit (99th percentile of the values for a reference control group) on at least one occasion during the first 24 h after the index clinical event (3). It is important to note that patients may have an episode of microinfarction where cTnT or cTnI measurements exceed the decision limit and yet CK-MB even by mass assay remains in the normal range (4). It is estimated that about one-third of patients presenting without ST elevation who would previously have been diagnosed as experiencing UA on the basis of normal CK-MB levels are now diagnosed as experiencing NSTEMI because of detectable troponin levels (1).
Is it clinically important to be able to diagnose episodes of microinfarction that are below the “radar screen” provided by CK-MB? Let us first look at the answer to this question as one might view a black and white photograph—that is, a dichotomous analysis of troponin tests with the results declared as positive (above the decision limit) or negative (below the decision limit). Heidenreich and Hlatky pooled the results of several trials and reported that the relative risk for troponin-positive patients (compared with troponin-negative patients) was 3.9 for mortality and 3.8 for death or nonfatal recurrent MI (1). An example of a black and white dichotomous analysis is shown in Figure 1, which emphasizes the significantly increased risk of mortality at 42 days among cTnI-positive patients presenting without ST elevation even in the subset whose CK-MB levels were normal (5). The adverse prognostic significance of a positive troponin test has been demonstrated by multiple clinical investigators, across multiple trials, involving multiple patient groups from different countries, indicating that the observation is a robust one (6).
Why do patients with a positive troponin test have a worse prognosis? Part of the answer may simply be that a positive troponin is indicative of myocardial necrosis. Because it is well established that left ventricular (LV) function is a pivotal determinant of prognosis, it is plausible that loss of functioning myocardium is associated with a worse outcome. Several investigators have reported dichotomous analyses showing that patients who present with UA/NSTEMI and a positive troponin test are significantly more likely to have high-risk angiographic anatomy of the culprit lesion (7,8). For example, troponin-positive patients are more likely to have visible thrombus in the culprit vessel, complex lesions (Type B2/C), and worse Thrombolysis In Myocardial Infarction (TIMI) flow grade. It has been argued that these anatomic findings put troponin-positive patients at risk for downstream embolization of atherothrombotic debris, occluding the microvasculature and causing foci of microinfarction—hence the release of troponin. This line of reasoning leads to the oft-repeated recommendation by several investigators that patients with a positive troponin test should be treated aggressively with antithrombotic therapy, with particular emphasis on the intravenous (IV) glycoprotein IIb/IIIa inhibitors (2). It has also been argued that troponin-positive patients should be selected for referral for an early invasive strategy (2).
As with most issues in clinical medicine, the situation is more complex than suggested by simple qualitative, dichotomous analyses (9). In Figure 2, we have moved from a black and white view to a color view of the picture (5). The troponin story is now shown with the rich mosaic provided by a quantitative analysis of the cTnI measurements. Clearly, not all troponin-positive patients are at the same level of risk. Note that there is a highly significant gradient of increasing risk of mortality with increasing troponin levels.
Is there any evidence that there is a gradient of high-risk coronary anatomy with rising troponin levels? In this issue of the Journal, Lindahl et al. (10)from the Fragmin during Instability in Coronary Artery Disease (FRISC-II) study group provide important angiographic and clinical findings that have profound therapeutic implications. They report that at presentation in the FRISC-II study, patients with increasing cTnT levels tended to have progressively greater delays from the onset of symptoms to blood sampling, were significantly more likely to present with ST depression or abnormal Q-waves on their ECG and were progressively more likely to have depressed LV function on echocardiography (ejection fraction <45%). In the subset who were randomized to an early invasive strategy, angiography did not show any correlation between the severity of underlying coronary artery disease and the cTnT level but did show a progressive increase in the odds of visible thrombus and a progressive decrease in the odds of TIMI flow grade 3 with increasing cTnT levels.
Complementing the angiographic findings from the early invasive strategy cohort are the interesting clinical findings in the early noninvasive cohort (10). The 12-month mortality was numerically lowest in the cTnT-negative group and was slightly higher in the cTnT-positive patients, but there was no clear trend toward increasing mortality with increasing troponin levels when analyzed by tertiles of positive cTnT. However, a U-shaped relationship was observed between troponin measurements and the rate of MI through one year; MI occurred in 5.5% of cTnT-negative patients (<0.01 ng/ml), 17.5% of patients in the first tertile of positive cTnT results (0.01 to 0.17 ng/ml), 16.2% in the second tertile (0.18 to 0.63 ng/ml), but only 8.4% in the third tertile (>0.63 ng/ml); p < 0.001. A similar U-shaped relationship was found with referral for revascularization: the rates through one year were 38.8%, 51.9%, 46.1% and 34.3%, respectively, in the same four groups. Interestingly, in a multivariate analysis that adjusted for baseline clinical characteristics, the odds ratio for MI through 12 months actually decreased progressively with increasing levels of cTnT.
Were these findings really unexpected? I do not think so. Careful inspection of a report from the Platelet Receptor Inhibition for Ischemic Syndrome Management in Patients Limited to very Unstable Signs and symptoms (PRISM-PLUS) investigators showed a similar U-shaped relationship. Patients with either undetectable or only slightly elevated levels of cTnT or cTnI showed no benefit of tirofiban with respect to prevention of death or nonfatal MI by 30 days (11). Those with intermediate elevations of either cTnT or cTnI had the maximum benefit, whereas those with higher biomarker levels showed progressively less benefit compared with the intermediately elevated group (see Figs. 1 and 4 of Heeschen et al. ).
How are we to interpret all this information, and what are the implications for clinical practice? I believe we should not be taking black and white pictures of our patients by the simplistic description of troponin test results as either positive or negative. I hypothesize that taking a color photograph by a quantitative assessment of troponin measurements will prove to be a more efficient and cost-saving approach to care of our patients presenting without ST-segment elevation (Fig. 3). Those patients with negative tests (below the decision limit) or just barely elevated quantitative results are at low risk. They are unlikely to benefit from IV glycoprotein IIb/IIIa inhibitors. In general, they are managed equally well with either an early conservative or early invasive strategy (depending on patient and physician preference). Multiple clinical factors other than the results of biomarkers should be included in the decision about referral for revascularization (12). Those patients with intermediately elevated troponin levels have not yet lost substantial amounts of myocardium and are excellent candidates for more effective antithrombotic therapy (e.g., enoxaparin in place of unfractionated heparin) and prompt referral for an early invasive approach supported by IV glycoprotein IIb/IIIa inhibitors in the catheterization laboratory (Fig. 3). Those patients with the highest levels of troponin have already lost substantial amounts of myocardium. It is not so clear that aggressive antithrombotic therapy with IV glycoprotein IIb/IIIa inhibitors will be helpful in such patients. Here our focus should be on early diagnostic coronary arteriography. Some patients may have a completed left circumflex infarction masquerading as NSTEMI, a point raised by Lindahl et al. (10). Others may have severe multivessel disease or other high-risk coronary anatomy in association with depressed LV function (as suggested by the high troponin levels); this group is likely to be best served by referral for coronary artery bypass surgery.
Because of the uncertainty about the best course of action for patients with the highest levels of troponin, I would not favor the “upstream” use of IV glycoprotein IIb/IIIa inhibitors in such patients (Fig. 3). Instead, I would wait for the results of the diagnostic catheterization to determine whether revascularization is needed, and, if so, whether referral for surgery is best (without IV glycoprotein IIb/IIIa inhibitors) or whether a percutaneous intervention supported by IIb/IIIa inhibitors is preferred.
Finally, the approach proposed above, stratified by troponin levels, needs testing and potential refinement as we add more “color” to the clinical picture. A good place to start would be an analysis of the benefits of the early invasive strategy in FRISC-II, stratified by quantitative troponin measurements. It is hoped that such an important analysis will also be forthcoming from the FRISC-II investigators.
↵∗ Editorials published in the Journal of the American College of Cardiologyreflect the views of the authors and do not necessarily represent the views of JACCor the American College of Cardiology.
- American College of Cardiology
- Braunwald E,
- Antman E.M,
- Beasley J.W,
- et al.
- Antman E,
- Bassand J.P,
- Klein W,
- et al.
- Heeschen C,
- van Den Brand M.J,
- Hamm C.W,
- Simoons M.L
- ↵Lindahl B, Diderholm E, Lagerqvist B, Venge P, Wallentin L, and the FRISC-II Investigators. Mechanisms behind the prognostic value of troponin T in unstable coronary artery disease: a FRISC-II substudy. J Am Coll Cardiol 2001;38:979–86.
- Heeschen C,
- Hamm C.W,
- Goldmann B,
- Deu A,
- Langenbrink L,
- White H.D