Ejection fractions determined by cardiac computed tomographic angiography and single photon emission computed tomographic myocardial perfusion imaging are not interchangeable: Evidence of significant and sex-associated disparities

Purpose:: Left ventricular ejection fraction (LVEF) determines patient management and is a standard part of cardiac imaging evaluation. Intermodality comparisons are useful in determining whether modalities are interchangeable. Multidetector Computed tomography (CT) and single photon emission CT (SPECT) myocardial perfusion imaging (MPI) have been compared in only a few cohorts. We compared these modalities in a sex-balanced group. Materials and Methods:: Sixty outpatients (30 women 30 men) referred for MPI underwent CT on the same day. We calculated LVEF, end-diastolic volume (EDV), and end-systolic volume (ESV) from CT and MPI datasets using a commercially available, semi-automated routine and quantitative gated SPECT (QGS) respectively. Correlations, t-tests and Bland-Altman plots were performed for ESV, EDV, and LVEF. Bivariate and multiple regression analyses for LVEF were performed for both modalities. The subgroups for men and women were analyzed. Results:: Computed tomography showed moderate to high correlations with SPECT for LVEF (0.62), EDV (0.70), and ESV (0.63). End-diastolic volume and LVEF were significantly higher on CT as compared to SPECT (P < 0.001 each). Multiple regression analysis showed a significant relationship between sex and LVEF (P < 0.0001) on SPECT but not on CT. In men, there were significantly higher EDV (P = 0.014) and LVEF (P < 0.001) on CT compared with SPECT, but there was no difference in ESV. For women, there were significantly higher EDV and ESV on CT (P < 0.001 each), but no difference in LVEF. Conclusions:: Left ventricular volumes and LVEF differed significantly on CT compared with SPECT and varied according to sex. Therefore, left ventricular volumes and LVEF values on CT and SPECT are not interchangeable.