Should we use a semi-quantitative assessment of apical rocking to predict outcome in CRT patients?

Abstract Funding Acknowledgements Type of funding sources: None. Introduction Apical rocking (AR) is a systolic septal-to-lateral motion of the apex observed on echocardiography in the apical 4-chamber view in some patients with a ventricular conduction delay. It is a strong predictor for outcome after cardiac resynchronization therapy (CRT) and has been suggested as selection criterion for CRT. Critics argue that simple visual assessment of AR may be too subjective for clinical use. Speckle-tracking based myocardial motion analysis may allow to objectify AR and may reduce a potential observer dependence. Purpose To compare the visual assessment of AR (VAR) to the semi-quantitative assessment of AR (QAR) in patients for the prediction of (i) volume response and (ii) all-cause death after CRT. To assess interobserver agreement for VAR and QAR. Methods We retrospectively analyzed data from 166 patients with a first CRT implantation. All patients had an echo before CRT implantation and approximately 1 year after implantation. VAR and QAR analysis were performed by three experienced observers on echocardiographic apical 4-chamber views before CRT implantation. QAR was obtained by performing speckle-tracking in the apical segments of the lateral and septal wall. A negative lateral displacement with a positive septal displacement during systole was considered as AR (Figure). Areas under the ROC-curves (AUC) for prediction of volume response were determined and compared using the DeLong method. Volume response was defined as a ≥15% decrease in left ventricular end-systolic volume (LVESV) within 1 year. Patients were followed for a mean duration of 5.2 years. Hazard ratios (HR) for all-time all-cause death after CRT were calculated using an univariate Cox proportional-hazards regression model. Interobserver agreement was analysed using the intraclass correlation coefficient (ICC) based on an average measurement, absolute agreement, two-way mixed effects model. Results 77% (n = 122) of patients were volume responders. Seven patients died before follow-up echocardiography. 31% (n = 52) of patients died during follow-up. AUC for volume response were similar for VAR and QAR (0.73 vs 0.72, p = 0.782). HR for all-time all-cause death were similar for VAR (0.39, 0.25–0.61; p<0.001) and QAR (0.37, 0.23–0.59; p<0.001). ICC was good for VAR (0.78; 0.55–0.90) and QAR (0.84; 0.67–0.93). Conclusion Semi-quantification of AR (QAR) using speckle-tracking technology achieves comparable results as the simple visual assessment of AR (VAR) when performed by an experienced reader for prediction of volume response and prediction of all-cause death in patients eligible for CRT. Interobserver agreement for VAR and QAR was similarly good. QAR could be used as confirmation of the visual assessment and may support clinicians less familiar with the visual assessment.