Accuracy of Computed Tomographic Angiography and Single-Photon Emission Computed Tomography-Acquired Myocardial Perfusion Imaging for the Diagnosis of Coronary Artery Disease

Abstract

Establishing the diagnosis of coronary artery disease (CAD) in symptomatic patients allows appropriately allocating preventative measures. Single-photon emission computed tomography (CT)-acquired myocardial perfusion imaging (SPECT-MPI) is frequently used for the evaluation of CAD, but coronary CT angiography (CTA) has emerged as a valid alternative. Methods and Results-We compared the accuracy of SPECT-MPI and CTA for the diagnosis of CAD in 391 symptomatic patients who were prospectively enrolled in a multicenter study after clinical referral for cardiac catheterization. The area under the receiver operating characteristic curve was used to evaluate the diagnostic accuracy of CTA and SPECT-MPI for identifying patients with CAD defined as the presence of ≥1 coronary artery with ≥50% lumen stenosis by quantitative coronary angiography. Sensitivity to identify patients with CAD was greater for CTA than SPECT-MPI (0.92 versus 0.62, respectively; P<0.001), resulting in greater overall accuracy (area under the receiver operating characteristic curve, 0.91 [95% confidence interval, 0.88-0.94] versus 0.69 [0.64-0.74]; P<0.001). Results were similar in patients without previous history of CAD (area under the receiver operating characteristic curve, 0.92 [0.89-0.96] versus 0.67 [0.61-0.73]; P<0.001) and also for the secondary end points of ≥70% stenosis and multivessel disease, as well as subgroups, except for patients with a calcium score of ≥400 and those with high-risk anatomy in whom the overall accuracy was similar because CTA's superior sensitivity was offset by lower specificity in these settings. Radiation doses were 3.9 mSv for CTA and 9.8 for SPECT-MPI (P<0.001). Conclusions-CTA is more accurate than SPECT-MPI for the diagnosis of CAD as defined by conventional angiography and may be underused for this purpose in symptomatic patients.