Chest Pain Assessment: What Is Our Endgame?

Cardiac troponin (cTn) testing is crucial in assessing patients with chest pain and other acute anginal equivalent symptoms. The Fourth Universal Definition of Myocardial infarction (4th UDMI) recommends the use of troponin in identifying patients who benefit from targeted therapies, including early revascularization (1). Developments in immunoassays for cardiac troponin (I and T) over the past two decades have resulted in highly sensitive assays with improved ability to detect and quantify troponin concentrations. This has allowed clinicians to move away from using cTn as a dichotomous variable (where the 99th percentile of healthy individuals was used to define values as normal or increased), towards more complex approaches incorporating a range of cutoffs and delta values. This ensures that maximum clinical benefit can be obtained by each new generation of assay. Every cTn assay has different analytical characteristics (2), and clinical approaches need to be tailored to the analytical characteristics of each new assay. The work by Wildi et al. (3) and Tjora et al. (4) in this issue of Clinical Chemistry adds to our understanding of troponin assays and how they should be used in assessing for acute coronary syndromes (ACSs). In addition, we believe these articles can serve as a catalyst for an evaluation of our endgame. That is, what are we trying to achieve in the evaluation of patients with suspected ACS? The study by Wildi et al. defines metrics for the high sensitivity cardiac troponin (hs-cTnI) VITROS assay (VITROS®, Ortho Clinical Diagnostics) used over a 0- and 2-hour timeframe for early rule-in, and rule-out of non-ST segment elevation myocardial infarctions (NSTEMIs) (3). A priori targets were set for the analyses. For rule-out, targets were 99.5% for negative predictive value (NPV) and 99.0% for sensitivity. For rule-in, the target positive predictive value (PPV) was 70%. Major adverse cardiac events [cardiogenic shock, ventricular tachycardia, or higher degree atrioventricular-block requiring intervention, death, or new acute myocardial infarction (AMI) at 30 days] also were reported. In contrast, Tjora and colleagues delved into use of the well-established hs-cTnT assay (Roche Diagnostics) for identifying patients with non-ST segment elevation ACS (NSTE-ACS) during their index hospitalization (4). This included patients with either type 1 NSTEMI or unstable angina pectoris (UAP). Combined myocardial infarction (MI) and all-cause mortality (within 30 days) and urgent (24-hour revascularization) was a secondary endpoint. This is a novel approach and with no literature on acceptable miss rates for UAP, a target of ≥95% sensitivity for NSTE-ACS was set, and ≥99% for the secondary endpoint. There are several important differences in these studies worthy of consideration and debate. The 4th UDMI categorizes types of MI according to underlying pathological processes, associated with clinical, prognostic, and treatment strategies (1). Of greatest interest in the emergency setting is type 1 MI, associated with atherothrombotic coronary artery disease (CAD) and usually precipitated by atherosclerotic plaque rupture or erosion, where specific intervention clearly improves patient outcomes. By comparison type 2 MI, associated with myocardial ischemia in the context of oxygen supply/demand mismatch is frequently seen (5). Rates of type 2 MI in the emergency setting vary, and indeed there was a 2-fold difference in rates reported in these studies [3.7% (3) to 7% (4)]. These proportions are lower than those reported in some US-based studies where up to 75% of MIs are type 2. (5) In the emergency department (ED), the varying rates of type 2 MI are likely associated with troponin ordering practices, with more liberal ordering practices associated with detection of higher rates of myocardial injury and type 2 MI. Both studies in this edition of Clinical Chemistry included type 2 MI as part of the primary endpoint. From a clinical perspective it is worth considering the impact of this inclusion, with arguments for and against. Troponin is a strong prognostic marker, including for patients with type 2 MI. However, the cohort of patients with type 2 MI is heterogeneous, underlying pathophysiology is multifactorial, and no single therapy improves the outcome for patients with type 2 MI. An individualized phenotype-specific management approach is required (5). Consequently, reporting a combined endpoint may not be clinically relevant. In contrast, collective reporting of “MI” may be beneficial due to the difficulty in differentiating type 1 and type 2 MI in some emergency patients. Indeed, even after further detailed evaluation, for some there will remain disagreement on the diagnosis (5, 6). Without uniform selection of patients for troponin testing in the ED and a reduction in the large variations of the proportions of patients with type 2 MI, there is value in reporting the accuracy of assessment algorithms including patients with only type 1 NSTEMI, in addition to the combined MI cohorts. In a noteworthy departure from most prior ED studies Tjora et al., incorporate UAP within the primary endpoint (4). UAP is defined as myocardial ischemia at rest or on minimal exertion, without acute myocardial injury (7). In this study, overall rate of diagnosis of UAP (14.4%) was higher than that reported by Wildi et al. (9.9%) (3). UAP is a challenging diagnosis. In patients without a troponin increase, there must be objective findings in some other investigations supportive of myocardial ischemia (e.g., regional wall motion abnormalities on stress echocardiography). However, for a diagnosis of MI, the 4th UDMI criteria include symptoms of ischemia alone with increased troponin (1). By extension, those patients with typical anginal symptoms but no increased troponin value, may have UAP, but this assumption/extension is flawed. Symptoms of ischemia are diverse and varied, with many patients diagnosed with AMI reporting atypical symptoms (8, 9). The assumption that the cause of patients’ chest pain is myocardial ischemia even if subsequent investigations reveal inducible ischemia and/or the presence of coronary artery stenosis may be incorrect. However, without a universally agreed and objective method for diagnosing UAP, this assumption maintains clinical safety. More exploration of highly sensitive troponin assays and other biomarkers for diagnosis of UAP is needed. Small troponin changes (around 2 ng/L) have been reported in patients with ischemia on exercise stress testing and shown to be prognostically important (10, 11). Significant change is also seen in brief transient occlusive balloon inflation during angiography (12). The role of newer markers of myocardial injury such as myosin-binding protein C and glycosylated apolipoprotein J warrant more research (13, 14). Once well-defined clinical outcomes are determined, standardized evaluation would enable comparison of different assessment strategies. Several studies use a minimum NPV of 99.5% for strategies ruling-out AMI equating to a miss rate of 1 AMI in every 200 patients with a negative test (15). The NPV is clinically relevant, but these estimates directly depend on disease prevalence, and cannot be translated across settings. A strategy that rules out 50% of all patients [similar to that seen in Wildi et al. (3)] and achieves >99.5% NPV, would yield a sensitivity of 99% where the disease prevalence is 16%, but sensitivity of 95% if the prevalence was 5%. As noted above, incorporating both type 1 and type 2 MI as endpoints means that the prevalence of disease can differ widely, making it difficult to compare across studies. Sensitivity does not alter as much with prevalence and should be reported alongside NPV. There has been less research to identify a target sensitivity, but 99% has been used (3). This cutoff is pragmatic, with emergency physicians accepting of a miss rate of approximately 1 in 100 AMI (16). However, it should be noted that in cohorts with low disease prevalence, this cutoff may be difficult to achieve. Many troponin studies incorporate between 1000 to 2000 patients with one or two missed cases reducing sensitivity below the desired cutoff. Additional issues should be considered when determining a standardized cutoff. First, the use of a point estimate of sensitivity or NPV may be problematic as these are estimates based on a single sample and do not account for uncertainty. Uncertainty in estimates (quantified using confidence intervals) is large when the number of events is low. Identifying the acceptable lower bound confidence interval should be included when identifying a target sensitivity or NPV. Second, troponin is a continuous variable, with increasing values indicating increased risk. While defining a single cutoff is important, it may be useful for papers to report the predicted probability for AMI across a range of values (15). Such information could support shared decision-making incorporating patients' perspectives on acceptable disposition. Similar arguments about the metrics for rule-in MI in reporting of studies can be made. Together these papers extend knowledge about troponin assays used in suspected ACS assessment and highlight the need for clarity about relevant clinical endpoints. Consistency in the definition and inclusion of outcomes would aid comparison of different strategies. Secondly, internationally accepted standards for derivation and reporting of metrics for accelerated diagnostic algorithms of emergency patients with suspected ACS would allow greater understanding of biomarker strategies for clinical use. Finally, these studies highlight the crucial aspect of lab expertise in maintaining standards in assay use. Clinical use of low values and small changes in concentrations of troponin to determine management of patients requires precision and consistently accurate results. Collaboration of clinical and laboratory expertise is essential. cTn, cardiac troponin; ACS, acute coronary syndrome; NPV, negative predictive value; AMI, acute myocardial infarction; UAP, unstable angina pectoris; MI, myocardial infarction. All authors confirmed they have contributed to the intellectual content of this paper and have met the following 4 requirements: (a) significant contributions to the conception and design, acquisition of data, or analysis and interpretation of data; (b) drafting or revising the article for intellectual content; (c) final approval of the published article; and (d) agreement to be accountable for all aspects of the article thus ensuring that questions related to the accuracy or integrity of any part of the article are appropriately investigated and resolved. Upon manuscript submission, all authors completed the author disclosure form. Disclosures and/or potential conflicts of interest. W. Parsonage is a board member and chair of the Quality Standard Committee of the Cardiac Society of Australia and New Zealand. L. Cullen, Beckman Coulter, Abbott Diagnostics, Glycardial, and Siemens Healthineers. None declared. L. Cullen, Siemens, Beckman Coulter, and Abbott Diagnostics. L. Cullen, Siemens, Beckman Coulter, and Abbott Diagnostics; J. Greenslade, grants from Emergency Medicine Foundation, National Health and Medical Research Foundation, Heart Foundation of Australia, Advance Queensland, and Brisbane Diamantina Health Partners; W. Parsonage, Siemens, Beckman Coulter, and Abbott Diagnostics. None declared. None declared.