|Values are valid only on day of printing.|
Click CC to turn on closed captioning.
Published: October 2010Print Record of Viewing
Dr. Saenger discusses the challenges in identification of an acute myocardial infarction (AMI) and the use of troponin assays to afford earlier diagnosis and better risk stratification. She also explains the importance of precision and sensitivity in the troponin assay.
Presenter: Amy K. Saenger, PhD
Welcome to Mayo Medical Laboratories' Hot Topics. These presentations provide short discussions of current topics and may be helpful to you in your practice.
Our presenter for this program is Amy K. Saenger, PhD, Director of the Central Clinical Laboratory in the Department of Laboratory Medicine & Pathology at Mayo Clinic in Rochester, Minnesota. Dr. Saenger will discuss the challenges in identification of an acute myocardial infarction (AMI) and the use of troponin assays to afford earlier diagnosis and better risk stratification. She will also explain the importance of precision and sensitivity in the troponin assay.
One of the major challenges in cardiovascular laboratory medicine revolves around acute coronary syndromes. Although acute coronary syndrome constitutes a continuum, it is usually divided into non-ST elevation myocardial infarction (NSTEMI) and ST-elevation myocardial infarction (STEMI) based upon electrocardiogram changes at presentation.
Patients with unstable angina are also classified into the acute coronary syndrome definition and present with chest pain or other symptoms without electrocardiogram changes or evidence of myocardial necrosis. Annual statistics estimate there are 610,000 new and 325,000 recurrent acute myocardial infarctions with 6 million visits to emergency rooms across the United States. A smaller percentage (approximately 2% to 5%) of myocardial infarctions are missed in the emergency room.
Mortality rates for patients over 40 years of age are high; 20% within the first year following an MI and 30% to 40% within the next 5 years. Notably, the mortality rate is also higher for females than males.
The diagnostic challenge relies on clearly differentiating patients who do have had an acute myocardial infarction from those who have not and can be sent home. Furthermore, situations are often not black and white and decisions are often needed about what to do with those patients with slightly elevated troponins but aren’t changing (i.e., in the gray zone).
Presently most laboratories run a troponin assay, either troponin T or troponin I, and may also offer testing for CK-MB and/or myoglobin. Troponin demonstrates improved myocardial specificity and sensitivity over both CK-MB and myoglobin and obsoletes the utility of either marker.
Cardiac troponin is a structural biomarker of cardiac necrosis and is the best marker for definitive AMI diagnosis. Troponins appear in the serum relatively early after the onset of symptoms and remain abnormal for 4-10 days. Elevations of troponin T persist longer than troponin I because it is larger (at 37 kDa versus 24 kDa). Although worth noting is an increase in troponin after unstable angina or small myocardial infarction, although in both situations the troponin concentration would be above the 99th percentile.
In 2007, a universal definition of MI was established to improve the accuracy of MI diagnosis. Key to this universal definition was the use of a highly sensitive and specific biomarker, troponin, determined from serial blood samples. It requires detection of a rise and/or fall of cardiac biomarkers with at least 1 of those concentrations being above the 99th percentile of the assay and evidence of myocardial ischemia (being symptoms, ECG changes, pathological Q waves, or imaging evidence). The timing of samples remains critical and serial testing is recommended for interpretation.
Use of a serial sampling strategy allows for differentiation of an acute infarction versus a chronic troponin elevation. There is no absolute agreement about the timing of serial samples; cardiology guidelines recommend baseline and 6 hour samples, with a 12 hour sample drawn in patients with a high suspicion or risk of MI. The International Federation of Clinical Chemistry recommends 0, 4, 8, and 12 hour samples, although it is widely recognized that baseline and either 4 or 6 hour samples will be sufficient for rule in/rule out purposes depending upon the assay and cutoffs used. Biological variation, both short-term and long-term variation, may influence serial sampling as well. There is currently a lack of clear definition of the criteria which defines what a significant change really is.
It is also important to remember that an elevated concentration of troponin in the blood only signals myocardial damage has occurred and does not indicate the cause of the damage. A variety of conditions may induce myocardial damage besides ischemia and are listed in this table.
Based on evidence that even small amounts of the cardiac specific troponin reflect incremental risk and indicate myocardial injury, consensus documents recommend that the normal range of troponin be set at the 99th percentile of a normal healthy population. Furthermore, the recommendations from the IFCC working group on standardization of cardiac markers state the total imprecision should be < 10% at the 99th percentile. The logic being that failure of this goal could increase the risk of reporting misleading results that may prompt unnecessary confirmatory testing or lead to clinical inaction when inappropriately low concentrations are reported.
Because of this recommendation, manufacturers must now provide this information on the package insert and it has become clear that commercial troponin assays are unable to achieve this acceptable precision. The precision also varies by specimen type and platform within the same manufacturer, as the larger instruments generally have better analytical characteristics. Most healthy normal individuals are then below the limit of quantitation and the lack of precision limits the ability to detect and define significant elevations in troponin at the low end of the range.
This table shows the precision characteristics of the assay used here at Mayo, the 4th generation Roche troponin T assay. You can see the 10% coefficient of variation is slightly greater than the 99th percentile concentration.
It may be helpful to review the analytical definitions as they relate to troponin. The limit of blank is the highest apparent analyte concentration and is determined from analyte-free samples. The limit of detection is the lowest concentration of troponin that can be reliably differentiated from the limit of blank. Assays that have a lower limit of detection are considered more sensitive but not necessarily more precise. In other words, just because you can report down to a certain number does not mean that you should. The limit of quantitation is the lowest concentration that can be reliably detected and produce an acceptable precision that may be de a predefined goal for bias and imprecision. The limit of quantitation may be equivalent to the limit of detection or it could be at a much higher concentration. For troponin assays the limit of quantitation is the 10% CV concentration.
Definition of these analytical requirements places the responsibility on manufacturers to ensure that their assays have the necessary precision to permit use of the proposed cut-off (which is the 99th percentile). Presently, not all troponin assays perform equally well in clinical settings and commercially available assays cannot meet the 10% CV as noted in this table. However, manufacturers are responding and there are several assays either in development or awaiting FDA approval that meet the goals and are considered “high sensitivity” assays. Clinical laboratories should carefully consider the effect of imprecision on clinical decision making when implementing or choosing a new troponin assay.
For most tests, the normal range includes the mean +/- 2 standard deviations as a way of defining abnormal results. Given the critical importance of troponin testing in diagnosis and treatment the 99th percentile is used. There is a need to define reference values in a more precise manner. Populations used to generate normal values are often those most conveniently available and may not contain adequate numbers of patients to assess age and gender-related differences. Some suggest for troponin the population needs to be a healthy younger population, as chronic conditions often plague the older population. Many times, laboratories attempt to validate the range in the package insert for convenience and find they cannot statistically accept the range often due to slight differences in precision between platforms.
In the past there has been controversy over which biomarker is preferable, troponin I or T. There is no scientific evidence that either of these markers is superior to the other. Therefore, focus should be placed more on the analytic and clinical performance of the assay, rather than which troponin is being tested. There is no way to correlate results from a troponin I assay to troponin T assay and often even between troponin I assays themselves.
A major issue for troponin I assays is the lack of standardization among commercial assays and harmonization of troponin I is an ongoing area of effort. In terms of risk stratification, troponin T or I may be utilized but it is clear that it is predominantly the use of a high-sensitive assay that has benefit over contemporary assays. There is an increased prevalence of troponin T elevations in the setting of renal failure and these are not to be considered false positives but related to overall dysfunction in the cardiorenal system. Patients with a chronic “low level” elevation of troponin have a worse prognosis and increased mortality.
The question often arises if CK-MB is even needed anymore. It is well recognized and accepted that it does not provide any additional information over troponin even in the setting of suspected re-infarction (which is rare today), assessing infarct size, before or after percutaneous coronary intervention, or in end stage renal disease patients. The only time CK-MB should be ordered is if there are suspicious false-positive troponin T or I results (although these can usually be resolved with heterophile blocking tubes or dilution studies) or if troponin testing is simply unavailable. Medicare does not reimburse troponin and CK-MB testing performed simultaneously.
Point-of-care cardiac marker testing is also an area that is commonly debated. The National Academy of Clinical Biochemistry guidelines recommend the turnaround time for troponin should be less than 1 hour over 90% of the time. Ideally, the turnaround time would be less than 30 minutes and timing is defined from collection to reporting. If point of care testing is used the results should be quantitative and the analytical characteristics of the POC test should be identical to the central lab’s troponin assay. Currently there are no POC methods that have acceptable analytical sensitivity and it is often argued that the turnaround time is essentially sacrificed for a lower quality result.
While the use of POCT shows monetary benefits for rapid patient triage, decreased length of stay in the emergency department and improvement in turnaround time, the lack of precision, sensitivity, increased cost and no demonstration in improvement in patient outcomes essentially precludes their routine use and recommendation.
So, what about high-sensitive troponin assays? Remembering the definition of a high-sensitive assay being a total imprecision of less than 10% at the 99th percentile and some would propose also being able to quantitate over 50% of normal values below that 99th percentile.
What is all the hype about high-sensitive troponin assays? While there are no FDA-approved high-sensitive assays, there are many in development and being used for research use only. Their use has been shown to diagnose MI earlier, greater prediction of death or future MI, and an improvement in risk stratification. It should be noted that the improvement in sensitivity is at the expense of specificity.
Two recent papers in the New England Journal of Medicine reported results from large, multicenter evaluations of the diagnostic performance of several sensitive assays. The 2 studies were consistent in their conclusions, with an improved accuracy in AMI diagnosis. The cardiac troponin concentrations at patient presentation, as noted in this figure, were significantly higher in patients with a final diagnosis of AMI than in those who had a different diagnosis of unstable angina or other cardiac causes but not coronary artery disease. Diagnostic performance of the assays was similar in STEMI vs non-STEMI patients.
How sensitive does troponin testing really need to be at this point? Essentially troponin assays need to diagnose patients with AMI as early as possible and identify patients are risk of premature death from cardiovascular disease. To do this as accurately as possible the assays require an acceptable precision within the normal range.
At Mayo, our current cardiac biomarker panel uses the Roche 4th generation troponin T assay. It is performed in the Hospital Clinical Laboratory on the Roche e411 using plasma samples and in the Central Clinical Laboratory on the Roche Modular E170 using serum. When ordered, a baseline sample is drawn and subsequent samples are drawn 3 and 6 hours later. The normal range is defined as < or =0.01 ng/mL and anything above that is flagged as abnormal. CK-MB was removed from the biomarker panel a few years ago.
We further report a "delta" value with our panel; the delta is calculated between the 0-3 and 0-6 hour samples and is reported as significant or not significant. Analytically significant changes are defined based on the baseline troponin. If the baseline troponin is < or =0.2 ng/mL, a change of 0.03 ng/mL or greater would be considered significant. And, if the baseline troponin is greater than 0.20 ng/mL, a change of plus or minus 20% would be considered significant.
To conclude, the precision of troponin assays will continue to improve, allowing for earlier diagnosis of acute myocardial infarction and better risk stratification for our patients. Interpretation of a clinically significant change in troponin concentration will continue to be important, especially upon use of the new highly sensitive assays. Point of care testing for troponin has not achieved the same level of precision as highly automated methods and remains an area for improvement.
I will disclose grant/research funding from Roche Diagnostics.