Assessment of left ventricular diastolic function by gated single-photon emission tomography: comparison with Doppler echocardiography.

Gated single-photon emission tomography (SPET) is not yet an established procedure for the evaluation of left ventricular (LV) diastolic function. This study examined diastolic function derived from gated SPET in comparison with an established diagnostic tool, Doppler echocardiography. We examined 37 consecutive patients with normal sinus rhythm who underwent gated technetium-99m tetrofosmin SPET. A gated SPET program was used with a temporal resolution of 32 frames per R-R interval. We obtained the Doppler transmitral flow velocity waveform immediately before gated SPET image acquisition. Patients who showed a ratio of peak early transmitral flow velocity to atrial flow velocity (E/A) of >1 or whose R-R intervals differed by >5% between Doppler echocardiography and gated SPET were excluded from this investigation. We compared diastolic indices and presumed corresponding intervals in diastole using the two methods. The peak filling rate (PFR) derived from gated SPET correlated with the Doppler peak velocity of the early transmitral flow (E) wave ( r=0.65) and deceleration of the E wave ( r=0.71). The time to PFR and percent atrial contribution to LV filling from gated SPET correlated excellently with the Doppler LV isovolumic relaxation time ( r=0.93) and the E/A ratio ( r=-0.85), respectively. There was a significant linear correlation in all the intervals from the R wave to the presumed corresponding diastolic points. The point of PFR in gated SPET and the peak of the E wave in Doppler echocardiography generally coincided. The onset of filling in gated SPET tended to be closer to the second heart sound than the start of the E wave in Doppler echocardiography. We conclude that gated SPET permits the assessment of not only myocardial perfusion and LV systolic function but also diastolic function, although there may be some errors in detection of the precise beginning of LV filling.