Differentiation between infarction and cardiomyopathy in left bundle branch block by parametric radionuclide imaging

In LBBB the diagnosis of the underlying disease, i.e. infarction (MI) or cardiomyopathy (CMP), is still a challenge. In this study, the usefulness of radionuclide ventriculography (RNV) for solving this problem noninvasively was tested. The hypothesis was that quantitation of the space-time sequence of left ventricular (LV) contraction by Fourier analysis may differentiate among MI, CMP, and LBBB-specific wall motion abnormalities (WMA). Forty-seven patients (pts) with LBBB were divided into 3 groups: 1) 20 pts with MI, 2) 7 pts with CMP, 3) 20 pts without MI or CMP. Regional parametric amplitude and phase data of eight LV sectors were calculated and expressed in standard deviation units (SDU) of normal. Out of the many regional data only the values of the mean amplitude and the standard deviation of the mean phase shift showed high discriminative power for separating the three groups, with an average squared canonic correlation of 0.5 and a Wilks lambda of 0.22, respectively. In other words, 17 of 20 pts in group 1, 6 of 7 pts in group 2, and 19 of 20 pts in group 3 were truly diagnosed by RNV alone. Such a separation also seems reasonable physiologically, since the mean amplitude reflects global left ventricular function and standard deviation of phase is a measure of time homogeneity of contraction. Furthermore, there was a significant septal phase delay of -4.92 +/- 5.3 SDU in 15 pts with septal MI compared to group 3 pts, who had a subtle premature septal motion of +0.75 +/- 2.64 SDU, whereas the septal amplitude was not different and slightly decreased in both.(ABSTRACT TRUNCATED AT 250 WORDS)