Abstract 17041: Speckle Tracking Imaging for the Assessment of Cardiac Resynchronization Therapy (START) Study

Abstract

Background: Speckle tracking echocardiography (STE) has a potential to accurately predict responders of cardiac resynchronization therapy (CRT). However, the evidence based on multicenter studies is insufficient. Accordingly, we conducted prospective multi-center study to identify usefulness of STE for the prediction of CRT responders.

Methods: This study design was multi-center prospective cohort study consisted of 17 Japan site. Patients who were newly planned to CRT device implantation with reference to the guidelines were enrolled. All STE data sets were analyzed in the 5 core laboratories.Following dyssynchrony parameters were assessed. Time from QRS to the max peak radial (RS) and circumferential strain (CS) in 12 segments on LV short axis plane, and to the max peak of longitudinal strain in 18 segments on 3 apical LV planes were calculated (Tmax). In the segments with multiple peaks on the time-strain curves, time to the first peak also was assessed (Tfirst). Difference of T between the earliest and latest segment (TD), and standard deviation (SD) of T in each strain component were assessed. CRT responders were defined as LV end-systolic volume (LVESV) reduction ≥ 15% at 6 month after CRT.

Results: 187 patients were primary enrolled, and finally 174 patients were completely reevaluated at 6 month after CRT. Among them, 105 patients were identified as responders. In addition to prevalence of LBBB pattern and significant mitral regurgitation, TD-first and Tfirst-SD of RS and CS were significantly higher in responders than non-responders. In a multivariate logistic regression analysis, Tfirst-SD of CS was selected as an independent predictor of CRT responder (p=0.002, Exp=3.45, 95% CI: 1.75 - 6.81) as well as LBBB (p=0.04, Exp=2.07, 95% CI: 1.00 - 4.28). The area under curve by ROC analysis of Tfirst-SD of CS was highest among clinical and echocardiographic parameters (0.69, 95%CI 0.58 - 0.75). As to inter-observer variability between core laboratories, good reproducibility was observed for Tfirst-SD of CS measurements (n=40 cases, kappa coefficient=0.80).

Conclusions: This prospective multi-center study revealed the feasibility of dyssynchrony assessments by STE, which may improve the ability in prediction of CRT responders.