Validation of QGS and 4D-MSPECT for quantification of left ventricular volumes and ejection fraction from gated 18F-FDG PET: comparison with cardiac MRI.

Abstract

The aim of this study was to validate Quantitative Gated SPECT (QGS) and 4D-MSPECT for assessing left ventricular end-diastolic and systolic volumes (EDV and ESV, respectively) and left ventricular ejection fraction (LVEF) from gated (18)F-FDG PET. Forty-four patients with severe coronary artery disease were examined with gated (18)F-FDG PET (8 gates per cardiac cycle). EDV, ESV, and LVEF were calculated from gated (18)F-FDG PET using QGS and 4D-MSPECT. Within 2 d (median), cardiovascular cine MRI (cMRI) (20 gates per cardiac cycle) was done as a reference. QGS failed to accurately detect myocardial borders in 1 patient; 4D-MSPECT, in 2 patients. For the remaining 42 patients, correlation between the results of gated (18)F-FDG PET and cMRI was high for EDV (R = 0.94 for QGS and 0.94 for 4D-MSPECT), ESV (R = 0.95 for QGS and 0.95 for 4D-MSPECT), and LVEF (R = 0.94 for QGS and 0.90 for 4D-MSPECT). QGS significantly (P < 0.0001) underestimated LVEF, whereas no other parameter differed significantly between gated (18)F-FDG PET and cMRI for either algorithm. Despite small systematic differences that, among other aspects, limit interchangeability, agreement between gated (18)F-FDG PET and cMRI is good across a wide range of clinically relevant volumes and LVEF values assessed by QGS and 4D-MSPECT.