Rick Schreurs

134 Chapter 7 photoplethysmography, device recorded electrograms or peak endocardial acceleration (SonR) [23]. Unfortunately, no single study was able to show long-term benefits of acute optimization strategies [24-26]. The second way of optimization is continuous, ambulatory optimization with the use of automated algorithms. Currently several automated algorithms are being used that mainly optimize the settings based on electrical parameters. The Adaptive CRT algorithm provides periodic evaluation of intrinsic conduction intervals to optimize LV-only pacing times with intrinsic RV activation [27, 28]. The SyncAV algorithm periodically measures intrinsic conduction delays and programs the AV-delay in such way to create triple wavefront fusion of BiV activation and native conduction [29, 30]. Finally, QuickOpt performs a rapid automatic AV/VV-delay optimization according to intracardiac electrogram data [31]. In chapter 5 we studied the SonR system, which is by now the only automated algorithm based on mechanical data being used for CRT. The CLEAR and RESPOND-CRT randomized trials have shown that optimization of CRT using the SonR system is non-inferior in terms of responder rate and hospitalization in heart failure patients with wide QRS complex compared to standard echocardiography-based optimization [32-34]. One of the main aims of this thesis was to find improvements for AV-optimization. Using the mean effective AV-delay to optimize pacemaker settings As discussed above and in chapter 4 , the mean effective AV-delay takes into account the mechanical contractions of both atria and both ventricles, which makes it a more reliable parameter to define the pacemaker setting with the highest overall cardiac output. In theory, this parameter can be used for acute optimization during the implantation procedure, although it would need invasive blood pressure monitoring in all heart chambers. Alternatively, doppler echocardiography might be a non-invasive tool to define the RA-RV and LA-LV intervals. The continuous wave doppler signal over the mitral and tricuspid valve gives information on the moment of LA and RA activation based on the A-wave. The timing of LV and RV contraction and subsequently IVD can be derived from the moments of aortic and pulmonary valve opening, which resemblances the time between LV and RV dP/dt max . Conceptually, these measurements could be performed shortly after the pacemaker implantation while BiV pacing at various AV-delays during atrial sense or atrial pace mode. In the porcine model the mean effective AV-delay was ~27% of the RR-interval (HR 100bmp), while in Circadapt it was ~14% (HR 60-100bpm). For patients, however, this value still needs to be determined before it can be used for optimization strategies. The above-described parameters are all echo-derived mechanical parameter that can unfortunately not be used for ambulatory, continuous optimization using automated algorithms. It would be interesting to examine whether electrical parameters can be used as surrogates for mechanical actions. For instance, the RA-RV interval can be derived from the RA and RV lead, while the P-wave width gives information about the IAD. Additionally,