Gated myocardial perfusion SPECT asynchrony measurements in patients with left bundle branch block.

PURPOSE: This investigation sought to determine which newly available asynchrony parameter derived from gated myocardial perfusion SPECT (GMPS) systolic wall thickening data best distinguishes patients with left bundle branch block (LBBB) from normal subjects.

METHODS AND MATERIALS: Emory Cardiac Toolbox (ECTb) algorithms were used to compute left ventricular (LV) global and regional function and perfusion indices with regional contraction phases for 20 patients with LBBB, and in 9 control (CTL) subjects who had no function or perfusion abnormalities. Histogram plots of phase frequencies versus R-R interval times included phase standard deviation (SD), bandwidth (BW), skewness and kurtosis. Z-score asynchrony measures were derived for phases sampled using the conventional 17-segment model.

RESULTS

In CTLs contraction occurred nearly simultaneously in all segments, while LBBBs exhibited a wide variety of heterogeneous contraction patterns. Global parameters that differed between LBBBs versus CTLs included EF, end-systolic volume and end-diastolic volume, and asynchrony measures that were different included BW, phase SD and z-scores. Z-scores most strongly discriminated LBBBs from CTLs (93% of cases correctly predicted, logistic regression chi(2) = 29.7, P < 0.0001). Z-scores, phase SD and lateral-septal wall timing were highly reproducible (r = 0.99, 0.99 and r = 0.87, respectively), with no significant inter-observer differences.

CONCLUSION

While traditional global function parameters were different in LBBBs and CTLs, asynchrony parameters characterized LBBB most strongly.