Rick Schreurs

87 SonR for AV-optimization Baseline (BL) measurements were performed during atrial pacing at a rate of ~10 bpm above the intrinsic rhythm. The total AV-block animals were paced at the RV lateral wall at an AV-delay of 150ms. Measurements were performed during variation of AV-delay in LV-only and BiV pacing mode. The AV-intervals were chosen to provide a range of various degrees of fusion between pacing-induced activation and intrinsic (right bundle branch (RBB) or RV lateral wall pacing) activation. In case of BiV pacing the VV-interval was set to zero. BL recordings were repeated after each pacing setting [12]. In three LBBB animals, cardiac contractility was enhanced using gradually increasing dobutamine infusion, with the maximal infusion rate being 40μg/kg/min. Hemodynamic and SonR analysis Data analysis was performed using custom-made MATLAB software. Hemodynamic and accelerometer data were recorded and averaged for two respiratory cycles. Premature beats were excluded from the analysis. PR and QRS duration were measured from the surface ECG. LV maximal (LVP max ) and end-diastolic pressures, and dP/dt max were derived from LV and RV pressure signals. SonR signals were notch filtered before further analysis. SonR1 peak-to-peak amplitudes were calculated for RA and RV [6]. Mechanical interventricular dyssynchrony (MIVD) was determined as the time delay between the normalized upslope of simultaneously recorded LV and RV pressure curves [13]. The time between the moments of LV and RV dP/dt max­ was calculated as an alternative for MIVD. A negative MIVD indicates that the RV is activated before the LV and is typical for LBBB and RV pacing. The optimal AV-delay (AV opt ) was defined for each individual animal and for each pacing setting and defined as the AV-delay with the maximal relative increase in LV dP/dt max . The two shortest AV-delays for each animal were averaged and referred to as AV short . Statistical analysis Continuous data are presented as mean ± standard error of the mean (SEM). Paired and unpaired sampled t-tests were used for comparing dependent and independent continuous variables, respectively. Linear regression analysis was performed, and correlation coefficients are presented as R 2 . A two-sided probability value of <0.05 was considered statistically significant. A Bonferroni correction was applied when multiple groups were compared. Statistics were performed using Microsoft Excel (Microsoft, Redmond, WA). 5