Coronary CT Angiography in Acute Chest Pain is a Cost Effective Risk Stratification Strategy

Background and Significance

Standard-of-care risk-stratification algorithm (SOC):

In the past 20 years since Goldman et al4 described a clinical algorithm to predict MI in ED patients, the clinical ability to decrease the false-negative rate for myocardial events has not improved. Hence, the emergency physician (EP) is compelled to admit to the hospital the majority of patients who present with acute chest pain for further observation and investigation due to inconclusive evidence of ACS or MI during the index ED visit; false-negatives are still sent home with CAD; and false-positives are admitted without CAD; accounting for a significant consumption of resources every year in the US1.

The current state of the art for EPs includes clinical data, electrocardiograms (ECGs) and cardiac biomarkers5 (Fig. 1). The limitations of using the Goldman prediction rule and the ECG is that they are insensitive indicators of myocardial injury in patients with MI4-9. The sensitivity and specificity of cardiac biomarkers is proportional to the time from onset of chest pain. Cardiac troponin begins to rise within 3-4 hours after the onset of myocardial injury and may remain increased for up to 4-7 days for cTnI and 10-14 days for cTnT10. There is enthusiasm for myoglobin as an early marker; however, myoglobin is non-specific11-16 In ED based trials; cardiac biomarkers have performed well but have not changed the cost and admission rate (false-positives). In 1999, McCord et al9 published a single center prospective cohort study that examined point-of-care cardiac biomarkers during the first 90 minutes and if they could help to exclude AMI in the ED chest pain patient. The post-hoc analysis and resultant negative predictive value (NPV) was 98 to 99% at time 0 minutes, 90 minutes, and 3-hours with various combinations of the three biomarkers including Myoglobin, Troponin I and CK-MB. These results demonstrated a consistency of negative predictive values with both combinations of CK-MB with myoglobin equal to Troponin with myoglobin. In 2004, Fesmire, et al17 reported in another prospective ED cohort study that a 2-hour delta CK-MB level outperforms myoglobin level in the early identification and exclusion of acute MI and can effectively risk-stratify patients for 30-day adverse outcomes. In both of these ED based trials, despite the impressive post-hoc NPV in the McCord et al trial, the cost and admission rate (false-positives) remains the same since the rule-out MI process requires admission and 24 hours of serial cardiac enzyme testing.

Patients presenting to the ED with acute chest pain undergo the SOC risk-stratification algorithm to determine the etiology of the complaint (Figure 1). The SOC risk-stratification initially includes the clinical assessment and an ECG. If the ECG reveals an ST elevation MI (STEMI) the patient is treated immediately in the cardiac catheterization laboratory. If the ECG is normal or indeterminate, and there are clinical risk factors, the patient will undergo cardiac biomarker testing. If the cardiac biomarkers are negative, the patient will then be admitted to the hospital for further testing. If the cardiac biomarkers are positive, the patient will be admitted to the cardiac care unit (CCU) for further testing. If the clinical risk factors are absent and the ECG is normal, the patient may be discharged without further testing4-8, however this is practiced in a small percentage (11%) of patients9. This practice risk-stratification algorithm continues to be the SOC. However, the combination of the Goldman predictor risk algorithm (ECG and symptoms) with a Troponin I (cTnI) < 0.3 was not able to identify a group of chest pain patients at <1% risk for the composite outcome of death, AMI or revascularization18. Further demonstrating the need for integration of additional testing that will decrease the false-positive and false-negative rates.

Adding nuclear perfusion cardiac testing to the SOC risk-stratification in the ED:

Investigators have studied adding exercise stress testing and echocardiograms in the ED in the SOC risk-stratification of acute chest pain. Both the exercise stress test and the echocardiogram can be enhanced with the addition of nuclear perfusion that can define recent reduction in arterial flow to specific segments of the myocardium. Due to the known weakness of exercise testing 19 and the cost of adding nuclear perfusion enhancement to both the exercise stress test and the echocardiogram, neither study findings have informed the standard-of-care risk-stratification algorithm20-22. Furthermore, no apriori clinical research had performed a cost-analysis to confirm the obvious excessive expense of adding nuclear study in chest pain patients. However, stress testing is considered a worthy test to include in the risk-stratification algorithm because it is capable of detecting flow limiting stenosis, though does not establish the presence or absence of coronary artery disease.

Adding non-invasive cardiac imaging: Coronary CT angiography (CTA) to the SOC (CTA + SOC) risk-stratification in the ED:

Appropriate decision-making in the low to intermediate-risk subgroup of patients presenting with acute chest pain requires new tools. We postulate that perhaps the incorporation of coronary CTA can, by providing an anatomical window that defines the amount of atherosclerotic burden within the coronary tree, advance the risk-stratification algorithm in low to intermediate-risk acute chest pain in a cost-effective manner. In recent years, there has been much interest in noninvasive imaging techniques that allow direct visualization and assessment of the coronary arteries (Image A). The coronary tree is challenging to image due to the continuous motion of the coronary arteries. This motion necessitates both high spatial and temporal resolution. Recent advances in computed tomography (CT) techniques have made imaging of the coronary arteries possible. The introduction of electron-beam CT (EBCT) in 1984 first allowed for ECG-synchronized imaging of the heart. In 1998, the first multiple detector spiral CT (MDCT) system with 4 detector rows was introduced, allowing for greater volume of coverage with each rotation of the CT gantry than EBCT. 23 Since that time, gantry rotation times have decreased, resulting in improved temporal resolution. An increase in number of detectors has allowed for sub-millimeter spatial resolution. At this time, 64 slice MDCT (Image A) is considered the "state of the art" technology for performance of coronary CTA, although the next generation of scanners, including dual source and 256 slice MDCT are currently under evaluation. Several small studies in the literature have shown the utility of coronary CTA in the safe and rapid triage of low to intermediate-risk acute chest pain patients presenting to the ED24-29. One of the clinical advantages of the coronary CTA is that, unlike stress testing that can detect flow limiting stenosis but not CAD; it can detect flow limiting stenosis in addition to the presence or absence of coronary artery disease. The disadvantage of coronary CTA is the exposure to radiation.

Multiple small trials comparing coronary CTA with invasive coronary angiography have demonstrated sensitivities ranging from 86-99% and specificities ranging from 93-97% for the detection of hemodynamically significant CAD. The negative predictive value (NPV) of coronary CTA in these studies was uniformly high, ranging from 95-99%, while rates of non-evaluable segments were low, ranging from 0-12%23, 30-34. Results of the multicenter, international CORE-64 trial were discussed at the 2007 American Heart Association Scientific Sessions evaluating the diagnostic performance of coronary CTA as compared with interventional coronary angiography in 291 patients with suspected or known CAD and Agatston calcium score <600. On a per patient basis, the sensitivity and specificity of coronary CTA for detection of significant CAD were 85% and 90%, while positive and NPV were 91% and 83%35. Moreover, in 2007, Meijboom et al36 reported on the utility of the coronary CTA in symptomatic patients with low, intermediate, or high estimated pretest probability of having significant CAD. Of the 254 patients enrolled in the study, the authors determined that all of the patients in both the low and intermediate-risk groups that were found to not have disease by coronary CTA were later determined to not have disease by invasive coronary angiography. Patients presenting to the ED with acute chest pain could potentially undergo more rapid triage with coronary CTA, speeding diagnosis, lowering hospital and further cost of testing.

Additional reports on cost-analyses in populations when adding coronary CTA to the SOC risk-stratification algorithm (CTA + SOC):

In a study by Goldstein et al published in 2007 in Journal of the American College of Cardiology, the authors reported that ED cost during the index visit was significantly lower in the group that received the coronary CTA versus the group that did not receive the coronary CTA. However, this cost difference was primarily due to a decrease in length of stay in the ED rather than the cost associated with other testing, specifically nuclear imaging. This cost analysis was limited to the index ED visits alone; since the study was powered for outcomes, not cost. No additional cost data was gathered, although outcome follow-up was 6-months.

Preliminary cost-outcome data presented at the Chicago 2008 American College of Cardiology meeting reported that; when including coronary CT calcium scores in the protocol to screen 495 asymptomatic firefighters, researchers were able to effectively triage subjects into follow-up coronary CTA or send them back to work without further workup, thus, decreasing cost. In another report presented, researchers concluded that cost was decreased in an out-patient cardiology clinic by incorporating coronary CTA in the risk-stratification of cardiac patients leading to reduced need for myocardial perfusion imaging and exercise treadmill testing over a 6-month period. Moreover, doctors were able to identify more coronary artery disease and provide more aggressive lipid management.

These data illustrate how cost and outcome studies are at the forefront of the research that is measuring and defining the role of coronary CTA. The coronary CTA is so innovative in the domain of acute chest pain, there is a paucity of published work that looks at the cost; instead, the majority of the literature examines outcomes and is produced primarily by cardiology.

Healthcare cost implications when adding coronary CTA to the SOC (CTA + SOC); and the coronary CTA demonstrates normal coronary arteries:

The cost implications of potentially discharging patients from the ED that are found to be free of CAD without further testing are impressive. In May of 2008, data in abstract form were presented in Academic EM where authors reported good NPV in 568 prospective patients who received coronary CTA in the ED at 30-days. These reports lacked a description of design, methodology, and a stated power calculation of an a priori hypothesis. Thus, the need for more definitive apriori cost and outcome clinical research in the application of coronary CTA in acute chest pain is warranted. The effect of this work will define and change the clinical practice landscape when applying this innovation to the cardiac risk-stratification model, particularly in acutely presenting chest pain. Embracing the incorporation of coronary CTA in acute chest pain is an innovation that will require well performed clinical trials to define the patients who will benefit most; whether it should be applied to asymptomatic patients, low, intermediate, and/or high-risk acute chest pain patients remains the question.