Diagnostic Performance of Dynamic CT Perfusion

Quick Takes

  • In a population with high prevalence of prior PCI or MI, the performance of CT coronary angiography was limited (AUC = 0.65).
  • Addition of quantitative perfusion estimates from dynamic stress perfusion improved diagnostic accuracy at the per-patient level.
  • Combining CT angiography and perfusion led to a sensitivity of 72% and specificity of 75%, respectively, at a per-patient level.

Study Questions:

What is the diagnostic accuracy of dynamic computed tomography (CT) stress perfusion imaging?

Methods:

The authors conducted a prospective multicenter study of dynamic CT stress perfusion imaging across seven centers in Japan using recent-generation CT scanners. Exclusions were: clinical instability, allergy to iodinated contrast, and prior coronary artery bypass grafting. Images were interpreted in blinded fashion and compared to invasive coronary angiography (ICA) with fractional flow reserve (FFR) for lesions between 26% and 90% diameter stenosis. Dynamic CT permitted quantification of myocardial blood flow (MBF) during stress. Receiver-operating characteristic (ROC) analysis and logistic regression were used to assess diagnostic accuracy of imaging findings.

Results:

A total of 174 patients were recruited, of which 157 underwent CT and ICA. Mean age was 66 years, and most were male (71%). Approximately one-third had prior history of myocardial infarction (MI) or percutaneous coronary intervention (PCI). On angiography, approximately half of the patients had ≥1 hemodynamically significant stenoses.

Mean MBF was measured to be 10 ml/min by dynamic CT perfusion and differed between vessels with and without hemodynamic stenoses (p < 0.0001). Delayed enhancement was observed in approximately one-third of patients. Mean radiation exposure for the entire CT examination was 12.8 ± 4.3 mSv. On a per-patient level, the area under the curve (AUC) for the ROC curve was 0.65 for a cutoff of 50% stenosis on CT angiography (CTA). Addition of dynamic CT perfusion increased this to 0.74 (p = 0.01). Sensitivity was 72% for the combined data and specificity was 75%. At the per-vessel level, the AUC of CTA was 0.70 and combined CTA and CT perfusion was 0.73, a nonsignificant improvement (p = 0.199).

Conclusions:

Addition of CT perfusion data to CT coronary angiography led to improved diagnostic accuracy for detection of hemodynamically significant epicardial stenoses.

Perspective:

This prospective, multicenter study of dynamic stress CT perfusion imaging versus ICA plus FFR as a gold standard shows only modest accuracy for CT angiography, which was improved by addition of perfusion data. Overall, this may be due to the relatively high-risk nature of the population, with a relatively high prevalence of established coronary artery disease (CAD) and prior stenting. Addition of perfusion data did improve performance in this challenging population.

Importantly, dynamic stress perfusion imaging allows for quantitative estimation of myocardial blood flow. However, the estimates reported of stress MBF during vasodilation are substantially lower than expected (~1 ml/min) and suggest a methodologic limitation. Invasive and PET measures of stress MBF are typically much higher, even in patients with CAD.

Additionally, the radiation dosimetry in this study is quite a bit higher than CTA-only protocols and may be further underestimated by using older conversion factors. Using modern conversion factors, the dosimetry could be as high as 24 mSv, which is markedly higher than standard practice for competing approaches. Finally, per-vessel analyses did not show consistently improved performance with the addition of perfusion data. This could be due to microvascular and diffuse disease contributing to reduced flow.

Clinical Topics: Cardiac Surgery, Cardiovascular Care Team, Heart Failure and Cardiomyopathies, Invasive Cardiovascular Angiography and Intervention, Noninvasive Imaging, Atherosclerotic Disease (CAD/PAD), Aortic Surgery, Cardiac Surgery and Heart Failure, Interventions and Coronary Artery Disease, Interventions and Imaging, Angiography, Computed Tomography, Nuclear Imaging

Keywords: Computed Tomography Angiography, Coronary Angiography, Coronary Artery Bypass, Coronary Artery Disease, Coronary Stenosis, Diagnostic Imaging, Fractional Flow Reserve, Myocardial, Hemodynamics, Myocardial Infarction, Myocardial Ischemia, Myocardial Perfusion Imaging, Percutaneous Coronary Intervention, Perfusion Imaging, Positron-Emission Tomography, Radiometry, Tomography Scanners, X-Ray Computed, Vasodilation


< Back to Listings