Abstract
1704
Objectives Ventricular function is a powerful predictor of survival in patients with heart failure (HF) [1, 2]. However, so far, studies on gated F-18 FDG PET for assessment of the cardiac function are fairly rare [3-5]. The aim of this study is to prospectively compare gated F-18 FDG PET and cardiac MRI for assessment of ventricular volume and ejection fraction (EF) in patients with HF.
Methods Eighty-nine patients with diagnosed HF who underwent both gated F-18 FDG PET/CT and cardiac MRI within 3 days were included. Left ventricular (LV) end-diastolic volume (EDV), end-systolic volume (ESV) and EF were obtained from gated F-18 FDG PET/CT using Quantitative Gated SPECT (QGS) and 4D-MSPECT software, respectively.
Results Correlations of LV EDV, ESV and EF between gated F-18 FDG PET/CT and cardiac MRI were excellent for both QGS (R = 0.92, 0.92 and 0.76, respectively) and 4D-MSPECT (R = 0.93, 0.94 and 0.75, respectively). Bland-Altman analysis revealed significant systemic error that LV EDV (-27.9 ± 37.0 mL; bias significantly different from 0, P = 0.004) and ESV (-18.6 ± 33.8 mL; bias significantly different from 0, P = 0.01) were underestimated by QGS.
Conclusions For patients with HF, despite gated F-18 FDG PET/CT correlated well with cardiac MRI in assessment of ventricular volumes and EF, variability between the two techniques did exist. Hence, in comparison with cardiac MRI, the two imaging techniques should not be used interchangeably and LV volumes and LV EF by gated F-18 FDG PET/CT should be cautiously interpreted.