Cardiac resynchronization therapy improves pulmonary function in patients with chronic heart failure and conduction disturbance

VVD). The way VVD is defined differs per implantable device. We investigated if this affects improvement of pump function during BiV pacing. Methods: In 8 canine hearts with chronic LBBB, leading to 17 16% LV hypertrophy and 25 19% LV dilation, BiVS pacing was performed with a large number of combinations of AVDs and VVDs. Two commonly used methods to program VVDs are tested; BiVS1; programming an AVD with respect to the ventricle to be paced first and delaying stimulation of the other one and BiVS2; maintaining a fixed AVD for the LV while tuning the VVD with RV stimulation time. To combine the effects of AVD and VVD and intrinsic RV activation during LBBB we calculated an effective VVD (VVeff). VVeff was defined as the difference between time of LV stimulation and (a) time of intrinsic RV activation (RAVDintr) during LBBB or (b) RV stimulation, whichever occurred earlier. RAVDintr was determined during BiV0 pacing with incremental AVD, and defined as the AVD at which the shape of the QRS duration changed as a results of loss of full capture. Results: RAVDintr correlated linearly to PQ-time (r 0.84) and was 35 4 ms shorter than PQ-time. Maximal improvement of LVdPdtmax that could be achieved with BIVS pacing was 12.2 5.9%. The figure shows that LVdPdtmax depends on both AVD and VVD. Especially at long AVD BiVS1 leads to lower LVdP/dtmax values as large as 5% despite identical applied VVDs. LVdPdtmax had a unique relation with VVeff, independent from AVD and definition of VVD (r 0.91). Conclusion: definition of VVD is important for hemodynamic outcome, but use of VVeff avoids such confusion.