ABC | Volume 113, Nº6, December 2019

Original Article Ket et al. Myocardial ischemia by SPECT and CCTA Arq Bras Cardiol. 2019; 113(6):1092-1101 Figure 2 – Comparison between myocardial perfusion images with stress perfusion defects on computed tomography (CT) and on single-photon emission computed tomography (SPECT). Concordant example of a same patient with significant obstructive anterior descending (LAD) coronary artery disease. Table 3 – Perfusion defects on scintigraphy (SPECT) and myocardial perfusion CT in relation to obstructive CAD (n = 35) Perfusion defects Positive SPECT* Negative SPECT* Positive CT** Negative CT** Obstructive CAD 10 5 14 1 Non-obstructive CAD 10 10 5 15 Two-sided Fisher’s exact test for SPECT (*p = 0.49) and for CT (**p = 0.0001). CAD: coronary artery disease; SPECT: Single-photon emission computed tomography; CT: computed tomography. Table 4 – False-positives on myocardial scintigraphy Cause of false-positive Positive SPECT Negative SPECT Deep myocardial bridge 2 2 Anatomical variation (short anterior descending artery) 1 1 Low levels (tracer leakage) 1 1 Patient with a 40% LAD stenosis 1 1 Patient with coronary-cavitary microfistulas 1 1 Others (microcirculation disease?) 4 4 SPECT: Single-photon emission computed tomography; CT: computed tomography. (CI) range of 0.67 – 0.94 (p < 0.001). On the other hand, SPECT myocardial perfusion had an AUC of 0.58, with a CI range of 0.40 – 0.74 (p < 0.001) (Figure 4). Assessment of correlation between observers of computed tomography for perfusion imaging Excellent intra- and inter observer correlation was reported in the assessment of stress perfusion, with an ICC of 0.90 (0.87‑0.92) and 0.94 (0.93-0.96), respectively. The intraobserver correlation of perfusion at rest was also excellent, with an ICC of 0.96 (0.95-0.97). For interobserver correlation of perfusion at rest the result was good, with an ICC of 0.71 (0.63- 0.78). Discussion In this study, it was possible to assess the diagnostic performance of myocardial perfusion by CCTA for the detection of significant obstructive CAD in relation to SPECT. The perfusion findings of scintigraphy with 99m Tc-sestamibi were compared with the findings of myocardial perfusion by 64-detector row computed tomography. As a strength of this study, we highlight the simultaneous use of the same pharmacological stress agent for CT perfusion image acquisition, and the administration of the radiotracer, which enables performance of CT and subsequent scintigraphy image acquisition, because it lacks significant redistribution. Another important data was the possibility for anatomical localization and correlation with the presence of myocardial perfusion defects by SPECT. In this study, it was also possible to understand why the defect was not detected by CCTA and to describe SPECT false positives. If we assess myocardial perfusion alone, an intermediate correlation between CT and scintigraphy images will be found, especially because the sensitivity of CT perfusion sensitivity for perfusion defects detection on SPECT was 70%, with a 1096