Biomarkers in the emergency workup of chest pain: Uses, limitations, and future

HISTORY AND PHYSICAL EXAMINATION PROVIDE KEY INFORMATION

In a review, Heidenreich et al⁸ noted certain demographic characteristics associated with worse outcomes—ie, older age and male sex; a history of medical conditions such as diabetes, MI, and hypertension; and heart failure on presentation.

A careful assessment of chest pain and associated symptoms helps narrow the differential diagnosis. Features that increase the likelihood of a cardiac origin of chest pain are:

• Chest pain at the time of presentation (likelihood ratio [LR] = 2.0)
• Radiation of the pain to the right shoulder (LR = 2.9), the left arm (LR = 2.3), or both arms (LR = 7.1)
• Nausea or vomiting (LR = 1.9)
• Diaphoresis (LR = 2.0).¹⁷

The physical examination can detect highrisk features such as new murmurs, hypotension, diaphoresis, pulmonary edema, and rales. It is more specific than sensitive and is useful in identifying low-risk patients by targeting potential noncardiac causes of the patient’s symptoms.¹⁸

The efficacy of clinical assessment was studied in 2,271 patients with chest pain presenting to the emergency department.¹⁹ In this cohort, a low-risk group with a 30-day major cardiovascular event rate (death, MI, stroke, or revascularization) of 2.5% could be identified through the use of the US Agency for Health Care Policy and Research criteria.

Electrocardiography

ECG provides important diagnostic and prognostic information and independently predicts death or MI, even after adjustment for cardiac biomarker measurements,^20,21 making it pivotal in the evaluation.⁴ The key features on ECG that increase the probability of MI are:

• New ST-segment elevation (LR 5.7–53.9)
• New Q waves (LR 5.3–24.8).¹⁷

One study²⁰ found that while the troponin T level was a powerful independent marker in patients presenting with MI, its value for risk stratification was enhanced when it was combined with a standard measure such as ECG.²⁰ While more than 90% of patients with STsegment elevation had an adverse outcome, only 31.7% of those patients had an elevated troponin T level.

No component is sufficient by itself

Thus, in spite of the proliferation of cardiac diagnostic tests, the initial bedside assessment of chest pain remains paramount. In fact, in patients presenting to the emergency department with chest pain, low risk (ie, those with a < 5% probability of MI) may be identified by presenting symptoms, medical history, and ECG alone.¹⁹

Furthermore, although clinical assessment, ECG, and cardiac biomarker testing each provide incremental benefit in assessing chest pain, no component is sufficient by itself. Sanchis et al²² found that even in patients with a normal troponin I level, the risk remained high in the case of ST-segment depression, and that even without signs of ischemia, the probability of cardiac events was 16% when the chest pain score was 11 points or higher.²² Consequently, a normal troponin level, ECG, or any other predictor alone would not ensure a good prognosis.

BIOMARKERS INSTEAD OF STRESS TESTING?

The ACC/AHA guidelines for the diagnosis of patients with unstable angina and non-STsegment elevation MI say that stable patients at low risk with no evidence of ischemia on initial assessment can be admitted to a chest pain unit for observation with serial cardiac biomarkers and ECG.⁴ At the end of the observation period, those who have reassuring results on ECG and normal cardiac biomarker measurements undergo functional cardiac testing or stress testing, or both.⁴

Exercise treadmill testing is a cornerstone of confirmatory testing in an accelerated diagnostic protocol because it is readily available, safe, and easy to do.¹⁸ A low-risk result was shown to have a high negative predictive value,^23,24 so that the likelihood of an acute coronary syndrome is low enough for safe discharge.

However, the overall process is not ideal since it is time-consuming, generates additional costs, and can have false-positive results in patients who are otherwise deemed not to be at high risk. While some studies provided an optimistic view about discharging low-risk patients with negative biomarkers without stress testing,^7,25 others have discouraged omitting exercise treadmill testing from protocols.^22,26

Others have proposed combining a biomarker with an imaging study such as coronary computed tomographic (CT) angiography.²⁷ Normal findings on this study have been shown to have a negative predictive value of up to 100% for ruling out an acute coronary syndrome and the occurrence of major adverse cardiovascular events in the long term.^28,29 Furthermore, it allows more-inclusive assessments of chest pain and can exclude other life-threatening causes such as pulmonary embolism and aortic dissection (referred to as the “triple rule-out”).³⁰

However, 25% to 50% of patients presenting to the emergency department with chest pain may not be candidates for CT angiography because of obesity, contrast allergy, intolerance to beta-blockade, arrhythmia, renal insufficiency, or a history of coronary artery disease.¹⁸ Moreover, it may be more efficient and less costly to discharge some patients without coronary CT angiography³¹ with the help of novel biomarkers without routine additional testing. This may spare patients the additional radiation exposure from CT angiography or nuclear imaging.^27,32

New biomarkers may, it is hoped, better distinguish patients at low risk from those at high risk without resorting to stress testing. Several of these markers are moving toward mainstream clinical use. For a biomarker to be prognostically equivalent to stress testing, it must be able to tell us if the likelihood of an acute coronary syndrome is low enough for safe discharge—ie, it must have a significantly high negative predictive value. Also, it must be an independent predictor of adverse outcomes, particularly in patients deemed at low risk by initial low troponin measurements. Biomarkers that have shown promise in this regard include high-sensitivity troponin, brain-type natriuretic peptide (BNP), cystatin C, and ischemia-modified albumin.

HIGH-SENSITIVITY CARDIAC TROPONIN ASSAYS

Although we speak of “high-sensitivity troponin,” these new assays detect the same molecule as do traditional troponin assays. The difference is that high-sensitivity assays can detect and measure troponin at concentrations much lower than the traditional assays can. In fact, high-sensitivity troponin assays can detect and measure troponin at very low levels in almost all healthy people.

Studies have shown that the high-sensitivity assays have better analytical accuracy and sensitivity than older assays.¹²

Aldous et al³³ reported that, in patients who presented to the emergency department within 4 hours of the onset of chest pain, an elevation in troponin T on a high-sensitivity assay had a positive predictive value of 53.8% and a negative predictive value of 98.3%.

Weber et al³⁴ found the diagnostic value of the high-sensitivity troponin T assay to be superior to that of a contemporary troponin T assay (area under the receiver-operating-characteristics curve [AUC] of 0.949 vs 0.929). Even when the contemporary troponin T assay was negative, the high-sensitivity assay provided strong diagnostic information (AUC 0.81). Furthermore, the high-sensitivity assay provided superior independent prognostic power for death within 6 months.

Hochholzer et al³⁵ reported a prognostic accuracy for death significantly higher (AUC 0.79) than that of contemporary troponin T (AUC 0.69). A concentration of high-sensitivity troponin T above 14 ng/L improved the prediction of death (hazard ratio 2.60) but not of subsequent acute MI in patients with acute chest pain. Therefore, a negative high-sensitivity troponin T assay identifies patients with a good prognosis and who may be discharged without further testing if their clinical presentation and ECG are also reassuring.

Keller et al³⁶ compared the diagnostic performance of the high-sensitivity cardiac troponin I assay against 11 other biomarkers, including a contemporary cardiac troponin I assay. The contemporary troponin I and the high-sensitivity troponin I assays performed best. The high-sensitivity troponin I assay at admission had a sensitivity of 82.3% and a negative predictive value of 94.7% for ruling out acute MI, whereas the contemporary troponin I assay had a sensitivity of 79.4% and a negative predictive value of 94.0%.

Using levels obtained at 3 hours after admission, the sensitivity was 98.2% and the negative predictive value was 99.4% for both troponin I assays. Combining the 99th percentile cutoff at admission with the serial change in troponin concentration within 3 hours, the positive predictive value for ruling in acute MI for high-sensitivity cardiac troponin I increased from 75.1% at admission to 95.8% after 3 hours; for the contemporary assay, it increased from 80.9% at admission to 96.1%.³⁶

The authors concluded that performing either of the cardiac troponin I assays 3 hours after admission may help in ruling out MI early on, with a negative predictive value greater than 99%. Moreover, the relative change in concentration within the 3 hours after admission, combined with the 99th percentile diagnostic cutoff value on admission, improves specificity, allowing acute MI to be accurately ruled in.³⁶

Of note, though studies have confirmed that a measurement at 3 hours identifies most cases of MI early, they have not used the recommended maximal sensitivity interval for troponin measurements (6 hours or more).⁶

A proposed algorithm for diagnosing acute MI with a high-sensitivity assay

While high-sensitivity troponin T assays can improve the early diagnosis of acute MI, how best to use them is yet to be defined. They still lack specificity for acute coronary syndromes, with positive predictive values as low as 50%.³⁷

Reichlin et al³⁸ developed and validated an algorithm for rapidly ruling out or ruling in acute MI using a high-sensitivity cardiac troponin T assay, incorporating baseline values and absolute changes within the first hour. Using a baseline threshold of 12 ng/L or less and an absolute change of 3 ng/L or less, they found a sensitivity and negative predictive value of 100%, making these good criteria for ruling out acute MI.

Using a baseline threshold of 60 ng/L or greater and a change from baseline to 1 hour of at least 15 ng/L, the specificity was 97% and the positive predictive value was 84%, making these good criteria for ruling in acute MI.

Patients whose values were in between were classified as being in an “observationalzone group,” in which the prevalence of acute MI was 8%. The cumulative 30-day survival rate was 99.8% in patients in whom the test ruled out MI, 98.6% in the observational-zone patients, and 95.3% in patients in whom the test ruled in MI.³⁸ Using this simple algorithm allowed a safe rule-out as well as an accurate rule-in of acute MI within 1 hour in 77% of unselected patients with acute chest pain; thus, it may obviate the need for prolonged monitoring and serial measurements in three out of four patients.”

Newby³⁹ stated that such an algorithmic approach must be validated in a prospective study that assesses not only sensitivity, negative predictive value, specificity, and positive predictive value, but also the implications for clinical outcomes and the cost of widespread implementation.

In the meantime, clinicians must keep in mind that patient populations in clinical practice are less selected, the prevalence of MI may broadly vary, and confounding comorbidities such as heart failure and renal insufficiency are more common. Studies are also needed to verify whether other factors such as age, sex, and time from symptom onset should be considered.