Peripheral Vascular Pulsatility in Heart Failure Patients with Continuous Flow Centrifuge and Axial Left Ventricular Assist Devices

Abstract

Current continuous flow left ventricular assist devices (LVAD) decrease peripheral vascular pulsatility, which may contribute to side effects such as bleeding, thrombotic events and orthostatic intolerance. PURPOSE: To investigate the impact of manipulating LVAD pump speed, documented as revolutions per minute (RPM), on peripheral (brachial artery) pulsatility index (PI) in 20 heart failure patients implanted with a HeartWare (HVAD, n = 10) or HeartMateII (HMII, n = 10) LVAD. METHODS: Doppler ultrasound blood velocity in the brachial artery was recorded at baseline and 3 minutes after altering RPM, at three different RPM settings above and below baseline (60 RPM increments for HVAD and 200 for HMII). Brachial PI was calculated for each cardiac cycle by dividing the difference between minimum and maximum blood velocity by the time averaged mean blood velocity. LVAD device pulsatility indices that are used clinically were also recorded: maximal blood velocity (HVADVmax) and minimum blood velocity (HVADVmin) (HVAD) and HMIIPI (HMII). Relationships were evaluated using multilevel linear modeling with random intercepts and data are reported as mean±SE. RESULTS: Baseline RPMs were 2509±44 (HVAD) and 9220±75 (HMII). Brachial PI changed significantly across the range of LVAD RPM speeds tested (HVAD: 360; HMII: 1200), from 2.3±0.6 to 4.1±0.9 with the HVAD and from 1.8±0.6 to 3.6±1.0 with the HMII, with no differences in brachial PI between device across relative pump speed stages. Specifically, a 180 RPM decrease of the HVAD resulted in a 0.9±0.1 (37±4%) increase in brachial PI and a 600 RPM decrease in the HMII resulted in a 0.8±0.1 (38±3%) increase. These reductions in pump speed resulted in an ~20.0% fall in LVAD power consumption and a reduction in device reported blood flow of ~9%. Brachial PI correlated with HVAD HWVmax and HWVmin (r = 0.45 and r = -0.31, respectively), and HMII device HMIIPI (r = 0.73), suggesting device derived indices of PI provide a fair to good linear prediction of peripheral vascular pulsatility. CONCLUSION: Reducing HVAD or HMII LVAD pump speed within a clinically acceptable outpatient range yields a measurable and potentially clinically and physiologically meaningful change in peripheral vascular pulsatility, accompanied by substantial power savings.