Rick Schreurs

75 Effective mechanical atrioventricular delay determination of IAD. Our study used atrial pressure rise times, while previous studies used the moment of atrial contraction onset, derived from tissue Doppler imaging or electrical mapping in the coronary sinus. During A-S mode left- and right-eAVD were longer than during A-P mode for every AV- delay. This can be explained by the fact that atrial activation and possibly contraction had already started before atrial activity was sensed by the pacemaker in A-S mode. The interval between the atrial pace spike and atrial dP/dt max was ~100ms in A-P mode, while the time between the moment of atrial sensing and atrial dP/dt max in A-S mode was only ~30ms. When keeping the paced AV-delay stable the effective AV-delay will consequently be longer because of the earlier activation of the atria. These differences between the A-S and A-P mode were clearly visible in the LV diastolic filling patterns. Since left-eAVD was longer during A-S, the time between the mitral E and A-wave was longer for every corresponding AV-delay than during A-P. This resulted in the absence of A-wave truncation at the shortest AV-delay, preventing a decrease in forward flow due to suboptimal ventricular filling. Furthermore, E and A-wave fusion happened at a shorter AV-delay in A-S compared to the A-P mode, supporting the idea that a shorter AV-delay should be programmed when using VDD pacemaker therapy. The increase in IAD when switching from A-S to A-P was larger during LV pacing (16ms) than during BiV (7-9ms) and RV (7-8ms). A possible explanation might be that during LV pacing at short interval the early LV contraction and pressure development also increases – through upward motion of the mitral valve leaflets – atrial pressure development and thereby shift the moment of atrial dP/dt max . Interestingly, also IVD seems to differ between A-S and A-P mode, especially during LV (from 19 to 13ms) and RV (-9 to -16ms). This could be a result of changes in ventricular filling due to prolonged atrial contraction in A-P mode, causing ventricular dP/dt­ max to occur at a slightly different moment. The increase in optimal AV-delay between A-S and A-P was comparable to the increase in IAD during BiV and RV pacing. However, for LV pacing the increase in optimal AV-delay (29ms) was larger than the increase in IAD (16ms). The latter might be also explained by the greater IVD during LV pacing compared to RV and BiV pacing. Other clinical studies found correlations between the IAD and the optimal AV-delay [11, 12, 14], though the ratio was not 1:1 in all cases, suggesting that there are other factors besides IAD that influence the optimal AV-delay. The influence of interventricular dyssynchrony (pacing site) on optimal AV-delay To our knowledge this is the first (experimental) study indicating that the optimal AV- delay is influenced by ventricular pacing site. This evidence comes predominantly from the measurements during LV pacing. 4