Imaging Coronary Anatomy and Reducing Myocardial Infarction
Udo Hoffmann, M.D., M.P.H.,
and James E. Udelson, M.D.
In 1998, the Journal published one of the early studies evaluating the sensitivity and specificity of coronary computed tomographic angiography (CTA), as compared with invasive coronary angiography, for the detection of obstructive coronary artery disease.1Subsequent studies have established that CTA has excellent sensitivity (95 to 99%) and high specificity (64 to 83%) for the detection of coronary stenoses of 50% or greater.2 An analysis from the Prospective Multicenter Imaging Study for the Evaluation of Chest Pain (PROMISE) showed that CTA predicted subsequent cardiovascular events at least as well as, and perhaps better than, functional testing (C-statistic, 0.72 vs. 0.64; P=0.04).3 The National Institute for Health and Care Excellence of the United Kingdom now suggests that CTA is the most appropriate test in patients with stable chest pain in whom angina pectoris cannot be excluded by means of clinical assessment alone.4
The Scottish Computed Tomography of the Heart (SCOT-HEART) trial investigators now report in the Journal that an initial diagnostic strategy that incorporated CTA in addition to standard care was associated with a 41% lower rate of the primary end point (death from coronary heart disease or nonfatal myocardial infarction) after almost 5 years of follow-up than standard care alone.5 This observed lower rate of the primary end point was driven almost entirely by a lower rate of nonfatal myocardial infarction in the CTA group than in the standard-care group and was achieved without resulting in a higher rate of subsequent invasive coronary angiography or coronary revascularization. This is a new finding that is in contrast to previous trials that have shown higher rates of invasive coronary angiography and coronary revascularization after anatomical imaging with CTA than after functional testing.
The relative risk reductions observed in the SCOT-HEART trial are similar to those observed in recent secondary prevention trials,6 which prompts speculation about the mechanism. In trials of diagnostic testing strategies, it is the downstream management — presumably driven by the testing results — that affects outcomes. It is unlikely that revascularization played a major role in the difference between the groups in the rate of the primary end point, and specifically in the difference in the occurrence of nonfatal myocardial infarction, given the similar rates of revascularization in the two groups. Therefore, the benefit seen in the CTA group might be attributable mostly to changes in medical management that were made on the basis of testing results. The authors speculate that the mechanism by which myocardial infarctions were prevented may, in part, be related to more appropriate use of preventive therapies in the CTA group — a hypothesis that is quite plausible. However, the differences in medical management between the groups were modest (the difference between the groups did not exceed 10 percentage points in either statin or aspirin treatment) and do not seem to be sufficient to explain the much lower rate of myocardial infarction in the CTA group than in the standard-care group.
The findings of the SCOT-HEART trial are in contrast to those of PROMISE, which randomly assigned similar patients to either CTA or functional testing but showed no difference in outcomes over a median 2-year follow-up.7 Although both trials had one group in which management was informed by CTA, an important difference between the two trials was the comparator group. In PROMISE, functional testing was predominantly stress imaging (nuclear or echocardiographic), and only 10% of patients underwent exercise electrocardiographic (ECG) testing. In the SCOT-HEART trial, however, the standard-care strategy was predominantly stress ECG testing, and few patients (approximately 10%) underwent an imaging test. One might conclude, then, that the CTA strategy was associated with fewer myocardial infarctions in the SCOT-HEART trial because of the suboptimal stress ECG comparator and because of the potentially suboptimal management that was based on that strategy. A review of the changes in medications after initial testing (Table S4 in the Supplementary Appendix of the article by the SCOT-HEART Investigators, available at NEJM.org) might support that concept; aspirin use increased after CTA was performed but decreased after exercise ECG was performed, even among patients with a probable or definite diagnosis of coronary heart disease. In contrast, in PROMISE, aspirin use increased after initial testing in both groups.8
The data from the SCOT-HEART trial suggest that management that is informed by results of CTA is associated with a lower rate of myocardial infarction than management that is informed by results of stress ECG testing. There may even be the potential for a lower incidence of myocardial infarction when CTA data are used than when data from any functional test are used. An analysis from PROMISE shows that a substantial proportion of myocardial infarctions occurred in patients with nonobstructive coronary artery disease identified by CTA3 — disease that would not be detected by functional testing. It would be reasonable to consider aggressive secondary prevention in these patients, although this specific approach has not been evaluated in a clinical trial. We also believe that leveraging data from trials such as SCOT-HEART and PROMISE may allow more efficient targeting of noninvasive testing9while continuing to drive improvement in vascular outcomes. The more general message from these trials is that the information provided by a diagnostic test can resonate therapeutically beyond making a correct diagnosis of coronary artery disease and that clinicians should aggressively pursue preventive measures to achieve the best outcomes possible while minimizing daily symptoms.