Rick Schreurs

77 Effective mechanical atrioventricular delay were replicated by Kyriacou et al in CRT patients in atrial sense mode in rest and during exercise and in atrial pace mode at different pacing rates [12]. Furthermore, computer simulations from our group have shown that the mean-eAVD corrected for RR shows less variability than absolute mean effective AV-delay when simulations were performed at different pacing rates [18]. The mean-eAVD as calculated in pigs was ~50ms longer in A-S mode compared to A-P for any ventricular pacing side, while this difference nearly completely disappeared when correcting it for heart rate. This provides further evidence for the influence of the heart rate on the optimal AV-delay. Clinical perspective and future directions The mean effective AV-delay is a global, bilateral measure of AV-coupling and is a better parameter for determining the optimal AV-delay in a wider variety of pacing settings. To implement the mean effective AV-delay concept clinically, detailed information on activation of RA, LA, RV and LV is necessary. Electrical RA-RV intervals can be derived from the pacemaker (sensed or paced RA-RV interval). P-wave width may be, and sometimes is, used to define IAD. Interventricular dyssynchrony can be derived from the QRS complex in case of clear dyssynchrony or from interlead conduction times [23-26]. A combination of these electrical parameters might subsequently be used to calculate mean effective AV-delay. For deriving mean effective AV-delay from mechanical information, echocardiography can be used to define right and left sided AV-delay by examining tricuspid and mitral filling patterns. While a single measurement of mean effective AV-delay in rest is possible using the abovementioned electrical and mechanical information, ambulatory repeated measurements can so far only be performed using the electrical information. Implanted accelerometers may be used to assess interventricular dyssynchrony by analyzing splitting of the second heart sound ( chapter 5, [27]), but further studies are needed to derive also atrial information from these sensors. Automatic optimization algorithms based on electrical signals have already been implemented in commercially available pacemakers. However, randomized clinical trials did not always show significant improvements compared to echocardiographic optimization or default settings [20, 28, 29], possibly because they focus on left-sided optimization of AV-delay. The CLEAR and RESPOND trials comparing the mechanical SonR accelerometer- based optimization to default settings in CRT patients indeed showed comparable clinical outcome in both groups [21, 30], suggesting that such mechanical sensors may be at least as useful as echocardiographic measures. It would be even more interesting to explore the use of an algorithm to automatically define the optimal mean effective AV-delay based on a combination of pacing lead electrograms and accelerometer signals. 4