Evaluation of three methods for quantifying valvular regurgitation using gated equilibrium radionuclide ventriculography.

The non-invasive quantification of mitral and aortic regurgitation using the left-to-right stroke count ratio (SCR) calculated with gated equilibrium radionuclide ventriculography (RNV), is affected by the overlap of atria and ventricles and the consequent difficult definition of the ventricular regions of interest (ROI). Various solutions of the problem have been proposed. In this study we evaluated the results obtained with a technique based on visual inspection of the RNV images (variable ROI method--VRI) and those of two approaches which utilize functional images (stroke volume image method--SVI--and Fourier amplitude ratio--FAR), by comparing them with the invasive quantification of valvular regurgitation according to Sandler et al. (stroke volume ratio--SVR). Forty patients (15 controls and 25 valvular patients) were studied. In the control group the range of the SVR was 0.81 +/- 1.11 (mean +/- 1 SD = 1.01 +/- 0.08). The SCR was 0.83-L28 (1.03 +/- 0.15) with VRI, 1.10-1.15 (1.30 +/- 0.14) with SVI and 1.11-1.58 (1.35 +/- 0.17) with FAR. The correlations between SVR and SCR were r = 0.47 (P less than 0.05), r = 0.62 (P less than 0.001) and r = 0.55 (P less than 0.01) respectively with VRI, SVI and FAR. The SCR of valvular patients fell in the range of controls in 11/25 using VRI, 6/25 using SVI and in 4/25 using FAR. This overlap was present in 2/25 with the invasive quantification. Irrespective of the method used, a reliable assessment of the valvular regurgitation was not possible in two patients with severely depressed left ventricular function. We conclude that the use of techniques based on functional images clearly improves the effectiveness of the non-invasive quantification of valvular regurgitation with the SCR even if this cannot be regarded as a substitute for invasive quantification and has a limited reliability in particular groups of patients.