Mechanical dyssynchrony or myocardial shortening as MRI predictor of response to biventricular pacing?

Purpose: To investigate whether mechanical dyssynchrony (regional timing differences) or heterogeneity (regional strain differences) in myocardial function should be used to predict the response to cardiac resynchronization therapy (CRT). Materials and Methods: Baseline mechanical function was studied with MRI in 29 patients with chronic heart failure. Using myocardial tagging, two mechanical dyssynchrony parameters were defined: the standard deviation (SD) in onset time (T_onset) and in time to first peak (T _peak,first) of circumferential shortening. Electrical dyssynchrony was described by QRS width. Further, two heterogeneity parameters were defined: the coefficient of variation (CV) in end-systolic strain and the difference between peak septal and lateral strain (DiffSLpeakCS). The relative increase in maximum rate of left ventricle pressure rise (dP/dt_max) quantified the acute response to CRT. Results: The heterogeneity parameters correlated better with acute response (CV: r = 0.58, DiffSLpeakCS: r = 0.63, P < 0.005) than the mechanical dyssynchrony parameters (SD(T_onset): r = 0.36, SD(T_peak,first) r = 0.47, P = 0.01, but similar to electrical dyssynchrony (r = 0.62, P < 0.001). When a heterogeneity parameter was combined with electrical dyssynchrony, the correlation increased (r > 0.70, P_incr < 0.05). Conclusion: Regional heterogeneity in myocardial shortening correlates better with response to CRT than mechanical dyssynchrony, but should be combined with electrical dyssynchrony to improve prediction of response beyond the prediction from electrical dyssynchrony only.