We physicians are obsessed with classifying, sorting, and differentiating in a quest for never-ending precision. We gather all manner of “facts” from our patients. Sights, smells, reactions to pushing or pulling. We divine sounds with antiquated stethoscopes or peer underneath the skin with ultrasound. We subject them to tests of blood, urine, and fluids from any place our needles can reach.
All of this is to arrive at an exact diagnosis that is often frustrated by the secondary nature of the data. Our disappointment has driven us mad, but the promise of exactness from biomarkers leaves us giddy. We have convinced ourselves that these laboratory tests will provide us the dichotomous yes/no answers we tell ourselves that our patients demand, but is really more for us. Answers without all of the messiness.
Does the patient have pneumonia? Check procalcitonin. Want to know if he has a pulmonary embolism? How high is the D-dimer? Why is the patient short of breath? Better get a BNP level. Chest pain? Run a troponin level (three times every four hours).
Troponins have turned on us, however, and emergency physicians and cardiologists are quickly learning to hate their old friend. Troponin levels were such a good biomarker. Positive levels were a clear marker of acute myocardial infarction; negative values ruled out the diagnosis. Exactly what we wanted from a biomarker. As the clinical chemists devised ways to measure troponins at lower and lower levels, however, we started to find patients with positive troponin levels who were not having acute coronary syndrome (ACS). Deciphering the meaning of a positive troponin is becoming less clear, making our lives messy again.
Troponin is found as a complex (troponin I, troponin C, troponin T) in myocardial cells that displaces tropomyosin when activated by entering calcium, exposing binding sites on actin for myosin. In short, troponin forms the trigger mechanism that starts heart cells contracting, two fortunate facts that make it useful as a marker. First, there is a version of troponin I and T that are only present in myocytes. Second is that the cardiac form of troponin is about 50 percent different from the skeletal muscle version. These two facts together mean that we can specifically measure the cardiac version, and that any way we measure we know is coming from the heart, not skeletal muscle.
Intracellular troponin is mostly bound as a complex (94%) attached to tropomyosin with only 6% free in the cytosol. It must be released from the cell for us to measure troponin levels in the blood. This has traditionally been assumed only to occur when cells die and leak their contents, but several alternative mechanisms are likely. First, there is natural turnover of myocardial cells. About 40 percent of all myocytes will turn over (programmed cell death can occur as apoptosis or autophagy) during a person’s life. The troponin proteins wear out, and are degraded by proteasomes within the cell cytoplasm into fragments. These can be released and measured by the assays. Lastly, oxidative stress can increase cell wall permeability or cause formation of cellular blebs, allowing for the release of fragments and free cytosolic troponin. These mechanisms mean that even normal individuals have measurably low troponin levels, which we can now detect as our assays have become more sensitive.
Once troponin is released from the cell, its half-life (t½) is approximately 120 minutes before it is cleared renally. Troponin can continue to be measured for longer periods after an insult because there is continued released from the myofibrillar pool as myocytes continue to undergo degradation during necrosis.
The original troponin assay was developed in 1987. Unlike measurement of CK-MB, an enzymatic assay, troponin is a protein, and assays use a sandwich ELISA method. The capture antibody has binding sites directed at cardiac-specific epitopes on the troponin molecule. The detection antibody binds different epitopes to increase specificity. Determining which epitopes to target is dependent on the manufacturer, but no two antibody pairs can have 100% sensitivity because of their existing troponin fragments, post-translational modifications, oxidation, phosphorylation, and complexes.
So what defines a high-sensitivity test? The important point to remember is that we are talking about analytical sensitivity, not clinical sensitivity. The sensitivity of detecting troponin, not how useful that measurement is, answers our clinical questions, such as whether the patient is having a myocardial infarction. Two criteria have been established for assays to be labelled high-sensitivity. The first is that it must be able to measure a detectable value for 50% of healthy individuals. This is called the limit of detection. The second criterion is coefficient of variation (CV). This is the analytical impression at the 99% percentile. Most contemporary tests are about 20%, and high-sensitives are generally much better. The lower sensitivity is important, but the more precise CV may actually be more clinically useful because it makes troponin measurements easier to repeat, especially at the lower values. Table 1 lists generations of troponin assays, and Figure 1 provides a visualization of the assay criteria.
Manufacturers reengineered the detection antibody to meet the criteria of high-sensitivity assay. They also increase the sample volume, raise the ruthenium concentration, and optimize the buffer to reduce background noise.
A challenge that manufacturers face is defining what constitutes a healthy population and establishing the 99th percentile upper limit normal (ULN). The normal has age and sex variations. Some manufacturers obtain only a health questionnaire (ranging from simple to extensive), while others perform a physical exam, ECG, and echocardiogram. How the healthy population is defined can have significant effects on the normal limits established. There are no guidelines, but current recommendations suggest establishing one cohort of patients under age 30 and another over age 30 with a median age of 60-65. A detailed history should be taken, the blood should be measured, and BNP should be obtained to screen for cardiomyopathy. Of course, patients included should have an equal sex and racial mix. At least 400-500 individuals are needed to obtain a statistically valid sample.
Using the high-sensitivity assays in healthy populations, we have started to notice significant biologic variability. That is variation of troponin levels over a day and other cycles, which may become important when assessing serial measurements as part of evaluations.
After all of this work developing a more advanced troponin assay, we are at least left with something very useful. Well, maybe. Before we can answer the clinical usefulness of high-sensitivity troponins (PPV of hs-cTn measurements), we need to understand the definitions and mechanisms of myocardial injury and infarction. Myocardial injury involves cell death, and is indicated by positive troponins. Many mechanisms can cause cell death, however. The myocardial cells become ischemic if they are deprived of adequate oxygen supply, and an infarction can develop if that persists long enough.
But as we know, the clinical presentation of myocardial infarction can vary tremendously, and arriving at a diagnosis can be challenging. A consensus conference has determined diagnostic criteria for myocardial infarction. In the most recent version from 2007, there must be evidence of positive troponins that have a rise/fall pattern indicating acute process with at least one value above the 99th percentile ULN. This must be supported by the appropriate clinical context and either ECG, imaging, or angiography evidence.
Myocardial infarction can be subdivided into subtypes. Type 1 is what we typically think of with the term myocardial infarction — plaque rupture and formation of an intracoronary thrombus. This is an acute coronary syndrome. Type 2 myocardial infarction occurs when there is a temporary imbalance between the perfusion and cardiac oxygen demand. This leads to ischemia and can cause an infarction and necrosis without a plaque rupture or thrombus formation. By definition, this is not an acute coronary syndrome despite the presence of elevated troponins, so it is important to note that myocardial infarction is not synonymous with ACS. Types 3, 4, and 5 are less useful for emergency physicians.
Elevated troponins can also occur because of nonischemic mechanisms that cause direct myocardial damage. The troponins indicate cell death with necrosis, but ischemia was not the mechanism. Examples include cardiac contusion, myocarditis, malignancy, and sepsis.
Unfortunately, distinguishing between type 1 MI, type 2 MI, and nonischemic myocardial injury can be challenging, especially in coronary artery disease. Troponins were found to be more sensitive than CK-MB in those found to have myocardial necrosis, and there is a level-dependent increasing risk for adverse outcomes. The elevation itself, however, does not provide any information about etiology. It does indicate myocardial injury, but it fails as an absolute biomarker for acute myocardial infarction. Clinical context is very important, and the entirety of the clinical presentation (history, physical exam, ECG, imaging, and angiography) matters in determining the diagnosis. The dynamic rise and fall of troponins on serial measurement can help establish that there has been an acute myocardial injury, rather than a baseline troponin elevation as may be seen in renal failure. A diagnosis of type 2 MI should be considered only when there is clinical evidence of an acute imbalance. Nonischemic mechanisms should be evident from the history. Ultimately in cases where the diagnosis remains uncertain, advanced imaging such as cardiac MRI may be useful.