Introduction In selecting patients that may benefit from cardiac resynchronisation therapy (CRT), dyssynchrony assessment by echocardiography based only upon the timing of regional contraction is limited by being inherently independent of underlying myocardial contractility. We hypothesised that patient selection may be enhanced using a strain-based parameter based not only the timing of myocardial segmental motion, but also on the amplitude of contraction, a potential measure of contractile reserve. We assessed a combined early and late strain index (ELSI) to predict CRT response.

Methods Speckle tracking radial strain was performed in 67 heart failure patients scheduled for CRT (age 69±9 years, ischaemic 56%, QRS 154±12 ms, NYHA III/IV—63/4, ejection fraction 23±7%). The ELSI was calculated as the sum for each of the 12 non apical segments of the difference in peak radial strain and strain at aortic valve closure. CRT response was defined as a >15% reduction from baseline in LV end systolic volume (LVESV) at 6 months. The predictive value of the ELSI was compared to previously reported dyssynchrony measures including the SD of time to peak myocardial longitudinal velocity of the 12 non apical segments (Ts SD12), the anteroseptal–posterior wall radial strain delay (AS-P delay) and the SD of time to peak radial strain of 12 segments (Rs-SD12).

Results Response to CRT occurred in 38/67 (57%) patients. Significant differences were seen between responders and non responders in the ELSI (91±45 vs 27±14%, p<0.01), AS-P delay (256±158 vs 94±87 ms) and the Rs-SD12 (143±62 vs 75±50 ms). There was no difference in the Ts SD12 between responders and non responders. The ELSI had the best correlation with LVESV reduction (r=0.61, p<0.001) and using an optimal cut-off of 40% (AUC=0.94), the ELSI was able to predict response to CRT with a sensitivity of 93% and specificity of 95%. This was much higher than for the AS-P (cut-off 130 ms, AUC=0.79, sensitivity 71%, specificity 74%) and Rs –SD12 (95 ms, AUC=0.82, sensitivity 73% specificity 75%).

Conclusion A combined early and late strain parameter based on both the timing and amplitude of segmental strain has a stronger predictive value in determining CRT response compared to widely reported dyssynchrony parameters based on segmental timing alone.