High-sensitivity cardiac troponin assays for risk stratification and for the diagnosis of acute myocardial infarction

Abstract

The measurement of cardiac troponin (cTn) plays a central role in the diagnosis of acute myocardial infarction (AMI). The universal definition of AMI requires the detection of a rise and/or fall of a cardiac biomarker above the 99th percentile of a reference population together with evidence of ischaemia (clinical symptoms, electrocardiogram changes or imaging evidence for loss of viable myocardium). The development of cTn assays with increased analytical sensitivity was requested by the profession and the manufacturers are in the process of delivering. The highsensitivity cTn (hs-cTn) assays by definition measure with adequate precision cTn in the majority of normal healthy subjects. While this increases their diagnostic sensitivity for the detection of myocardial insults including AMI, the increase in the proportion of patients without AMI but with a positive cTn has led to a reduction in diagnostic specificity for AMI, to the extent that the diagnostic accuracy of hs-cTnT for AMI is no superior to that of secondgeneration sensitive cTnI assays. Christ et al. point out, ‘the aim of disposition decisions in the emergency department is not to identify patients with AMI or risk of death alone, but to identify patients at increased risk for adverse outcome, who may benefit from hospitalization and/or tailored treatment strategies.’ Aldous et al. had shown in the long-term follow-up arm of their study that hs-cTnT was in fact superior to sensitive cTnI assay in risk stratification for long-term outcomes. Patients who were not diagnosed with AMI at the index presentation were found to have a significantly increased risk of major adverse events (cardiac death, AMI or revascularization) over the subsequent two years if an hs-cTnT result was recorded between the limit of detection and the 99th percentile, independent of age. These findings raise a number of questions for the physician confronted with a patient with detectable hs-cTn below the diagnostic cut-off point for AMI. What further investigations should be performed? Should this be as an inpatient or outpatient? If evidence of obstructive coronary disease is found, what is the benefit of interventional strategies over medical management? What is the additional value in terms of outcome prediction of an hs-cTn result when considered alongside known long-term risk factors such as blood pressure, diabetes, tobacco smoking and cholesterol? What are the economic implications of the implied significant increase in diagnostic testing? Finally, does responsibility for initiating this proposed investigative pathway lie with the emergency doctor or the primary care physician, and to what extent is the process dependent on the patient engaging with follow-up and risk modification strategies? The major advantage of hs-cTn is to allow the earlier exclusion of AMI in emergency department patients presenting with chest pain. Diagnostic sensitivity approaching 100% may be achieved by measuring a second sample within three hours of presentation. However, the lower specificity may present a number of additional problems if these tests are used unselectively in patients with low clinical likelihood of acute coronary syndrome, or inpatients with multiple co-morbidities. Christ et al. have rightly pointed out that ‘the correct working hypothesis and deduced therapeutic management strategies can only be made using all available information including clinical signs and symptoms of the patient, results of 12-lead electrocardiogram, and levels and dynamics of cTn concentrations to calculate post-test probabilities’. The universal definition of AMI does include the detection of a rise and/or fall of cTn in blood, but the optimum cut-off point for delta troponin is unclear, and may not be the same for cTnI and cTnT or even for different cTnI assays. In addition, the 99th percentile cut-off point would depend on the reference population used to determine it. For example, the 99th percentile upper limit for the Roche hs-cTnT assay was derived from a relatively young population. Serum cTn concentrations increase with age, and although any detectable hs-cTn denotes an increased long-term risk of cardiovascular events and mortality, the use of a 99th percentile cut-off derived from a young population for the diagnosis of AMI in an older patient may lead to loss of specificity and diagnostic accuracy for the latter diagnosis. Gender differences may also be relevant in determining the AMI cut-off, and the need to use a genderand age-matched healthy reference population for this purpose has been emphasized. There are other factors which are independently associated with hs-cTnT, such as level of kidney function and left ventricular hypertrophy. Although these may not apply to a healthy