Left Ventricular Decompression During Speed Optimization Ramps in Patients Supported by Continuous-Flow Left Ventricular Assist Devices: Device-Specific Performance Characteristics and Impact on Diagnostic Algorithms

Background: Echocardiographic ramp tests have been widely used to help guide speed adjustments and for identification of potential device malfunctions in patients with axial continuous-flow left ventricular assist devices (LVADs) (Heartmate II LVAD [HMII]). Recently, the use of centrifugal-flow LVADs (Heartware LVAD [HVAD]) has been on the rise. The purpose of this study was to evaluate the utility of ramp tests for assessing ventricular decompression in HVAD patients.

Methods and results: In this prospective study, ramp tests were performed before index hospitalization discharge or at the time of device malfunction. Vital signs, device parameters (including flow), and echocardiographic parameters (including left ventricular end-diastolic dimension [LVEDD], frequency of aortic valve [AV] opening, and valvular insufficiency) were recorded in increments of 100 rpm, from 2,300 rpm to 3,200 rpm. Twenty-six ramp tests were performed, 19 for speed optimization and 7 for device malfunction assessment. The average speed after the speed optimization ramp tests was 2,534.74 ± 156.32 RPM, and the AV closed at a mean speed of 2,751.77 ± 227.16 rpm, with 1 patient's valve remaining open at the maximum speed. The reduction in LVEDD for each speed increase was significantly different when the AV was open or closed, at -0.09 cm/increment and -0.15 cm/increment, respectively (P = .013), which is significantly different than previously established HMII LVEDD slopes. There were also significant changes in overall device flow (P = .001), upper flow (P = .031), and lower flow (P = .003) after AV closure. The power slope did not change significantly after the AV closed (P = .656). Five of the 19 tests were stopped before completion owing to suction events, but all tests reached ≥3,000 rpm.

Conclusions: The parameter slopes for the HMII cannot be directly applied to ramp studies in HVAD patients. Overall, the LVEDD slope is drastically smaller in magnitude than the previously reported HMII findings, and speed adjustments were not based on the degree of left ventricular unloading. Therefore, the slope of the LVEDD-rpm relationship is not likely to be helpful in evaluating HVAD function.