Simultaneous biplane first-pass radionuclide ventriculography using 99Tcm-tetrofosmin: a comparison with magnetic resonance imaging.

Abstract

First-pass radionuclide ventriculography (FPRNV) using 99Tcm-labelled myocardial perfusion agents allows the assessment of myocardial function and perfusion simultaneously. We have assessed the feasibility of biplane FPRNV using 99Tcm-tetrofosmin, and have validated global and regional functional measurements by comparison with magnetic resonance imaging (MRI). FPRNV was performed at rest in 18 patients referred for assessment of known or suspected coronary artery disease (5 with previous myocardial infarction). A dual-headed camera was used to acquire RAO and LAO projections simultaneously. Left ventricular ejection fraction (LVEF) was calculated using standard methods and regional wall motion was evaluated visually in five myocardial segments using a 4-point scale and also by Fourier analysis. Cine MRI was performed in four oblique planes, LVEF was calculated using a biplane area-length method and regional motion was assessed visually in a similar fashion to FPRNV. Agreement between the techniques for LVEF was good using RAO FPRNV (mean +/- S.D. difference = 0.7 +/- 4.7%), but less good in the LAO projection (difference = 10.5 +/- 7.1%). Wall motion was normal by both FPRNV and MRI in 5 patients without CAD and 3 of the 13 patients with CAD. In the remaining 10 patients, wall motion by MRI was abnormal in 20 segments; FPRNV with visual analysis was abnormal in 8 patients (80%) and 16 (80%) segments. Fourier analysis showed regional abnormalities in 7 patients (70%) and 13 (65%) segments. There was good agreement (kappa = 0.68) between FPRNV (visual analysis) and MRI for wall motion scores, but moderate agreement (kappa = 0.55) between Fourier analysis and MRI. Thus, LVEF measured by FPRNV in the RAO projection agrees well with MRI. Normal regional wall motion is accurately identified, but regional abnormalities are better assessed with MRI.