Rick Schreurs

30 Chapter 2 Conv CRT Multisite Multipoint Endo optimal Endo worst * Conv CRT Dual Triple Quadruple Septuple Endocardial * * 15 10 0 5 Change in LV dP/dt max (%) 20 Medium Best Worst Endocardial Multisite or multipoint Conv CRT Patients Canine LBBB model Figure 5. Percentage change in LV dP/dt max versus baseline during conventional (Conv) CRT (white bars), multisite or multipoint (grey bars) and endocardial pacing (black bars) measured in a canine LBBB model (panel A) and in patients (panel B). The values in the bars are presented as mean+SEM. The three lines in panel A are the pooled data for the two ‘worst’, the three ‘intermediate’ and two ‘best’ single LV pacing sites (red, black and green, respectively). * indicates p≤0.05 vs. conventional CRT. Reproduced with permission from the Heart Rhythm Society (panel A) [38] and the European society of Cardiology (panel B) [40]. Optimization of timing of pacing All CRT devices have the option to adjust the AV- and VV-delay to the individual patient. Figure 6A shows schematically how these different settings can influence the electrical activation of the LV. The VV-intervals (y-axis) can be chosen such that either the RV or the LV is activated first or that both are activated simultaneously. This determines the relative contribution to the complete electrical activation of the LV by activation fronts initiated by the LV pacing site and the RV pacing site and/or intrinsic activation from the RBB. AV- delays (x-axis) shorter, equal to or longer than the intrinsic AV-delay influence the relative contribution of the intrinsic and paced activation wavefronts to the complete activation of the LV as well ( Figure 6A ). These settings affect pump function primarily by optimizing the degree of resynchronization [12, 46]. Figure 6B shows the relative changes in LV dP/dt max , stroke work and electrical resynchronization in response to 100 different combinations of LV (x-axis) and RV (y-axis) AV-delays. Note the similar leftward turn (white arrows) of the optimal values for all three parameters, occurring at an RV AV-delay just shy of their respective intrinsic PQ duration. Several parameters have been used to judge which AV- and VV-delays result in the best CRT-response. These include diastolic filling time, aortic velocity time integral, arterial blood pressure, LV dP/dt max , stroke work and vectorcardiography [47, 48]. However, a major problem seems to be that the increase in pump function is in the same order of magnitude as the biological variability. According to basic rules for accurate measurements, reliable results are only observed if repeated measurements are performed [49]. The fact that