Comprehensive effects of LVAD speed changes on alveolar gas exchange, sleep ventilatory pattern, and exercise performance

Background. Increasing left ventricular assist device (LVAD) pump speed according to patients’ activity is a fascinating hypothesis. This paper is an analysis of the short-term effects of LVAD speed increase on cardiopulmonary exercise performance, muscle oxygenation (NIRS), lung diffusion capacity (DLCO and DLNO), and sleep quality.
Methods. We analyzed CPET, DLCO and DLNO, and sleep in 33 Jarvik 2000 LVAD patients. After a maximal cardiopulmonary exercise test (CPET, n=28), patients underwent two maximal CPETs with LVAD speed randomly set at 3 or increased from 3 to 5 during effort (n=15). Then, at LVAD speed randomly set at 2 or 4, we performed: a) constant workload CPETs assessing O2 kinetics, cardiac output (CO), and muscle oxygenation (n=15); b) resting DLCO and DLNO (n=18); c) nocturnal cardiorespiratory monitoring (n=29).
Results. The progressive pump speed increase raised peak VO2 (12.5±2.5 ml/min/kg vs. 11.7±2.8 at speed 3; p=0.001). During constant workload, from speed 2 to 4, CO increased (3.18±0.76 L/min vs. 3.69±0.75, p=0.015, at rest; 5.91±1.31 L/min vs. 6.69±0.99, p=0.014, during exercise), and system efficiency (τ= 65.8±15.1 sec vs. 49.9±14.8, p=0.002) and muscle oxygenation improved. At speed 4, DLCO decreased, and obstructive apneas increased despite a significant apnea/hypopnea index and a reduction of central apneas.
Conclusion. Short-term LVAD speed increase improves exercise performance, CO, O2 kinetics, and muscle oxygenation. However, it deteriorates lung diffusion, and it increases obstructive apneas, likely due to an increase of intrathoracic fluids. Self-adjusting LVAD speed is a fascinating but possibly unsafe option, probably requiring a monitoring of intrathoracic fluids.