Epicardial versus endocardial pacing in cardiac resynchronization therapy

Abstract Funding Acknowledgements Type of funding sources: Public grant(s) – National budget only. Main funding source(s): Norwegian Research Council. Background/Introduction The timing of left ventricular pressure rise (Td) relative to QRS-onset is associated with dispersion of regional work distribution and prolongs with dyssynchrony and dyssynergistic contractions. Td is short in normal conduction and shortening of Td can be achieved with effective biventricular pacing (BIVP). The effect of endocardial compared to epicardial pacing on Td is unknown and could be important for the interpretation of Td when different pacing strategies are utilized for CRT. Purpose We wanted to analyze the measurement from QRS-onset to onset of LV pressure rise identified as the peak double derivative of LV pressure rise (Onset of Synergy, OoS) and to the peak pressure rise (Td), and test the effects of cardiac resynchronization therapy with endocardial vs epicardial pacing electrodes in a large animal model of LBBB. Methods 5 mongrel dogs were included in the study and handled according to ethics committee approved protocol/GLP. Pacing electrodes were placed epicardial on the left ventricular (LVepi) and right ventricular (RVepi) free wall, and endocardial electrodes were positioned at the right ventricular septum (RVsept) and in the LV endocardial free wall (LVendo). The atrial electrode was placed on the left atrial appendage. Pressure was measured with a Micro-tip pressure sensor (Millar Inc.) in the LV and sampled at 1000Hz. LBBB was created with a retrograde placed 7F standard ablation catheter and confirmed by surface ECG and four endocardial electrodes. Td was measured from 10 consecutive beats. Mixed models were utilized for the repeated measurements. Results Td was measured to 115±2ms at baseline in LBBB. Td decreased by 5ms from RVepi (114±3ms) to RVendo (109±3ms) regardless of LV electrode position. We then compared the effects of using LV endocardial and epicardial pacing in CRT and found that Td was 6ms higher (p<0.01) in LVepi (115±3ms) vs LVendo (109±3ms) regardless of RV electrode position. The interaction between RV RV and LV revealed that RVepi/LVendo (105±3ms) was significant shorter than RVendo/LVendo (110±3ms, p0.01) and RVepi/LVepi (115±3ms, p<0.01). OoS with RVepi/LVendo was 71±2ms, and a significant relationship between OoS and Td were found (R=0.62, p<0.01) to indicate that the myocardial contraction is affected by pacing already at 71ms after QRS-onset early during the pre-ejection period. Conclusion(s) In this study we demonstrate how endocardial and epicardial pacing applied with CRT may affect contraction patterns of the heart differently. The different effects from pacing can be measured as early as 71ms after QRS-onset. Shortening of Td occurs to a larger extent with pacing from the endocardial LV and epicardial RV, while pacing at two endocardial sites did not result in synergistic effects on Td. This may indicate that there might be an interaction between endocardial pacing sites when applied in combination.