Abstract

Objective: To examine the predictive value of systolic dyssynchrony measured by tissue Doppler velocity versus tissue Doppler strain imaging on long-term outcome after cardiac resynchronisation therapy (CRT).

Design: Cohort study.

Setting: Two university hospitals.

Patients: Two hundred and thirty-nine patients (65 (SD 12) years, 76% males) who underwent CRT.

Interventions: Baseline echocardiography with tissue Doppler imaging (TDI) and clinical follow-up for 37 (20) months.

Main outcome measures: The time to peak systolic velocity during ejection phase (T_s) and the time to peak systolic strain (T_ϵ) were assessed for dyssynchrony, that is the maximal delay in T_s and the maximal delay in T_ϵ among the four left ventricular basal segments. Occurrence of cardiovascular endpoints between patients with and without dyssynchrony was compared by Kaplan–Meier curves, followed by Cox regression analysis for potential predictor(s).

Results: There were 78 (33%) deaths, with cardiovascular causes in 64 (27%) patients, while 136 (57%) patients were hospitalised for cardiovascular events, including decompensated heart failure in 87 (36%) patients. Patients with the maximal delay in T_s of ⩾65 ms showed a lower event rate for cardiovascular mortality (19% vs 38%, logrank χ² = 7.803, p = 0.005) and other prognostic endpoints. In Cox regression analysis, the maximal delay in T_s (hazard ratio (HR) 0.463, 95% CI 0.270 to 0.792, p = 0.005) and ischaemic aetiology (HR 2.716, 95% CI 1.505 to 4.901, p = 0.001) were independent predictors of cardiovascular mortality. In contrast, the maximal delay in T_ϵ of ⩾80 ms failed to predict any cardiovascular event.

Conclusions: Echocardiographic evidence of prepacing systolic dyssynchrony measured by TDI velocity, but not TDI strain, predicted lower long-term cardiovascular events after CRT.