Background: Atrioventricular (AV) optimization of cardiac resynchronization therapy (CRT) is typically calculated at rest. However, patients often become symptomatic during exercise.
Objective: In this study, we use acute noninvasive hemodynamics to optimize the AV delay of CRT during exercise and investigate whether this exercise optimum can be predicted from a three-phase resting model.
Methods: In 20 patients with CRT, we adjusted the sensed AV delay while the patient exercised on a treadmill up to a heart rate of 100 bpm to identify the hemodynamically optimal value. Separately, at rest, by pacing with three different configurations and calculating the sensed-paced difference, we calculated an "expected" value for the exercise optimum.
Results: It was possible to perform AV delay optimization while a patient exercised. The resting three-phase model correlated well with the actual exercise optimal AV delay (r = 0.85, mean difference +/- standard deviation [SD] = 3.7 +/- 17 ms). Simply using measurements made at rest during atrial-sensed pacing showed a poorer correlation with exercise (r = 0.64, mean difference +/- SD = 2.2 +/- 24 ms). The three-phase resting model allows improved exercise hemodynamics to be achieved. Programming according to the three-phase resting model yields an exercise blood pressure of only 0.5 mmHg (+/-1.4 mmHg; P = NS) less than the true exercise optimum, whereas programming the resting sensed optimum yields an exercise blood pressure of 1.4 mmHg (+/-2.2 mmHg, P = .02) less than the true optimum.
Conclusions: Using acute noninvasive hemodynamics and a protocol of alternations, it is possible to optimize the AV delay of cardiac resynchronization devices even while a patient exercises. In clinical practice, the exercise optimum AV delay could be determined from three phases of resting measurements, without performing exercise.