Medicine is a science of uncertainty and an art of probability.—William Osler (1)

Currently, more than 6 million Americans experience chest pain suggestive of coronary artery disease (CAD) (2). In addition to the physical symptoms, chest pain carries a large emotional burden because of the still unacceptably high rate of cardiovascular death in the community, and the patient's fear that their symptoms may be a harbinger of future heart attack or even death. The evaluation of chest pain is not simply limited to diagnosis and treatment; patients also look to their clinician for reassurance of their future cardiac health.

For decades, exercise electrocardiography (XECG) has been the first-line investigation in the evaluation of suspected CAD (3). This is not simply due to its widespread availability, economy, and diagnostic accuracy. Although the sensitivity and specificity of XECG for obstructive CAD are both modest at around 70% (4), a negative XECG carries an excellent prognosis, with a 5-year survival of 97% (5). Even if we are uncertain about the absence of CAD with a negative XECG, we are reassured by the low probability of future adverse CAD events. Although stress echocardiography and nuclear cardiology have advanced both diagnostic accuracy and the capacity to determine the location and extent of ischemia ((6),7), the landscape has been further changed by coronary computed tomographic angiography (cCTA). With multidetector (64 slice and above) cCTA, the sensitivity for the presence of obstructive CAD approaches or even exceeds 95% ((8),9), and coronary anatomic information previously only available through invasive angiography is obtained. It is therefore not surprising that a burgeoning interest has developed in the application of this new technology into the diagnostic algorithm of chest pain.

In a large, single center retrospective cohort study, Cho et al. (10) compared the prognostic value of 64-slice cCTA versus XECG in a group of 2,977 registry patients with suspected CAD, 95% of whom had angina or angina equivalent. All patients underwent the 2 tests within 90 days, with a median interval of only 8 days. The primary endpoint was the rate of major adverse cardiac events (MACEs), defined as cardiac death, nonfatal myocardial infarction, unstable angina, and revascularization over a median follow-up of 3.3 years. In multiple prognostic models that included clinical risk factors, cCTA provided improved discrimination for future MACEs over XECG alone, whereas XECG did not improve discrimination for future MACEs over cCTA alone. When stratified according to a negative or positive XECG, cCTA had added benefit in predicting MACEs in both strata. However, although XECG provided added benefit in predicting MACEs in patients with moderate or severe CAD on cCTA, it did not provide a benefit in patients with minimal or no CAD on cCTA.

Although one might hypothesize, as the investigators do, that due to its combination of high diagnostic accuracy and improved prognostication, cCTA could become the first-line test of suspected CAD, it is important to take a deeper look at the study findings to determine what benefits would be achieved by this approach. The dominant contributor to future MACEs in the study was coronary revascularization (most of which was percutaneous coronary intervention), which accounted for >80% of all MACEs. Although coronary revascularization was only included in MACEs if it occurred >90 days after the index test, the confounding effect of cCTA driving the study endpoints cannot be ignored, because >60% of coronary revascularizations were for lesions deemed to be obstructive at the index cCTA. Although percutaneous coronary intervention for stable CAD provides excellent control of angina, it has not been shown to reduce the rate of future death or myocardial infarction (11); therefore, any benefit of a cCTA-first approach would be mostly limited to the improved identification of those in whom intervention might provide a symptomatic benefit. The rate of “hard” CAD endpoints, such as death, myocardial infarction, or unstable angina, in patients with a negative XECG was extremely low, with only 6 such events in nearly 2,500 patients during the follow-up period. There are many other aspects that would require clarification before elevating cCTA to the first-line test for suspected CAD, most notably the competing role of functional modalities, such as perfusion imaging and stress echocardiography versus anatomic imaging, the potential harm of radiation, and also cost efficacy.

An alternative interpretation of the study findings is that the current XECG-first algorithm for suspected CAD is appropriate, but those with a positive XECG should undergo cCTA and only proceed to invasive angiography if cCTA demonstrates moderate or worse CAD. Patients with a positive XECG but “negative” cCTA (defined by normal or <40% stenosis) had no significant increase in MACEs compared with those with negative results in both tests, which supports the safety of such an approach. Because only slightly more than one-third of patients who undergo invasive angiography for suspected CAD have been found to have obstructive disease (12), cCTA may play a “gatekeeper” role in preventing unnecessary invasive coronary angiograms. One would speculate that such an approach, were it proven safe and effective, could lead to significant cost savings.

An important shortfall in this study is the lack of information on medical therapies targeted against atherosclerosis, such as antiplatelet agents, angiotensin-converting enzyme inhibitors, and lipid-lowering medications. This is particularly relevant given the absence of data evaluating the treatment of patients with nonobstructive disease identified by cCTA. A patient with a negative XECG would generally be reassured that they have noncardiac pain and often only lifestyle measures would be recommended. However, the management of a patient with noncardiac pain but non-obstructive CAD on cCTA is more confusing. Although such patients may be at higher risk of future CAD events (13), whether this justifies aggressive antiatherosclerotic therapy is unknown. Also, in addition to the uncertainty related to its management, a diagnosis of CAD in even mild form has important legal, occupational, and emotional implications.

The study by Cho et al. (10) provides valuable information relating to the interaction of cCTA findings and XECG with respect to future prognosis, which should aid clinicians in selecting the appropriate investigation for the evaluation of chest pain. Uncertainty remains with respect to exactly where cCTA sits in the investigational algorithm of chest pain, and also as to how we might best manage the growing number of patients with minor CAD identified by cCTA and noncardiac chest pain. However, the additional prognostic information obtained by cCTA may now enable us to provide even greater reassurance to our patients, that the odds of a healthy cardiac future are good.

⁎Editorials published in the Journal of the American College of Cardiology reflect the views of the authors and do not necessarily represent the views of JACC or the American College of Cardiology.