Verification of the left ventricular ejection fraction from gated myocardial perfusion studies (GSPECT)

Abstract

Introduction: The perfusion study (which may be obtained using SPECT or GSPECT technology within six hours of administration of the radionuclide Tc-99m-MIBI) reflects the regional blood supply to the left ventricular [LV] myocardium at the time of radionuclide administration (i.e. at rest, at peak exercise, or at peak vasodilatation), while the values of EF, EDV, and ESV measured using GSPECT are parameters of LV contractility at the time of image acquisition (i.e. at rest or in a nearresting state following exercise or vasodilatation). Planar radionuclide ventriculography [RNV] is, however, considered to be the most accurate method for calculating LVEF.

Aims: The main goal of the study was to compare the values of EF obtained by the most frequently used method, GSPECTQGS, and the reference method, RNV - taking into consideration various clinical scenarios (presence or absence of LV dilatation) and various conditions under which GSPECT was recorded (at rest, post-exercise, or post-dipyridamole).

Methods: Two hundred patients (145 males) aged 58+/-11 (18-80) with previously confirmed (n=166, of whom 108 had a history of myocardial infarction) or suspected (n=34) coronary artery disease were included in the study. Ranges of normal values for EF, EDV, and ESV were established based on a group of 26 'normal' subjects. LV dilatation was defined as an EDV >127 ml (at rest, measured by QGS) - this was present in 88 patients. Myocardial perfusion studies were obtained using GSPECT following administration of Tc-99m-MIBI at rest (all patients), as well as one hour after treadmill exercise (138 patients) or dipyridamole administration (48 patients). The resting RNV was conducted within three weeks of the GSPECT exam. The EF values obtained by QGS and RNV were compared for patients with and without LV dilatation. EF, EDV, and ESV values obtained by QGS were compared for resting patients, post-exercise, and post-dipyridamole.

Results: 1. The GSPECT EF values calculated using QGS software, at rest or one hour after treadmill exercise or dipyridamole administration, demonstrated, for the study population as a whole, a significant, strong correlation with the results obtained by the reference method, RNV (correlation coefficient, r> or =0.86). The correlation was stronger in patients with LV dilatation than in those without. Both in patients with and without LV dilatation the correlation of EF with RNV was slightly weaker for postexercise (relative to resting) and post-dipyridamole (relative to post-exercise) measurements. 2. QGS tended to underestimate the absolute values of EF, as compared to RNV. 3. In post-exercise and post-dipyridamole measurements, relative to the resting measurements (in patients with previously diagnosed or suspected coronary artery disease) the mean values of EF were lower while EDV and ESV were higher.

Conclusions: In order to complement data on myocardial perfusion, the GSPECT-QGS technique should be optimally used to calculate LV contractility parameters at rest (as opposed to post-exercise or post-dipyridamole), and include a range of normal values for EF, EDV, and ESV, obtained using QGS. Of note, EF measurements by GSPECT are more accurate for dilated than non-dilated ventricles.