The aim of this study was to determine the impact of prolonged left ventricular assist device (VAD) support on cardiac ventricular geometry and VAD flow kinetics.
Nineteen patients with end-stage heart failure underwent the implantation of HeartMate II rotary flow VADs. Left and right ventricular geometry and VAD flow kinetics were assessed by transthoracic echocardiography early (7 ± 1 days) and late (113 ± 21 days) after VAD implantation.
Left ventricular end-diastolic internal dimension decreased by 21% and 35%, respectively, early and late after VAD implantation ( n = 19; P < .001 vs before VAD implantation). Right ventricular end-diastolic internal dimension did not decrease at either time. Hemodynamic trends were similar. VAD inflow obstruction by myocardium was observed in eight patients, seven of whom demonstrated significantly increased variation of VAD inflow during the cardiac cycle (“pulsatility”) detected by Doppler studies. Medical or surgical intervention returned VAD flow patterns toward baseline in seven of eight patients with VAD obstructions.
Prolonged rotary VAD support unloads the left ventricle, with modest effects on the right ventricle. These changes are often associated with alterations of VAD flow kinetics, requiring therapeutic intervention. These findings indicate the usefulness of echocardiographic surveillance in patients undergoing prolonged VAD support.
Heart failure afflicts 5.7 million adults in the United States, of whom 59,000 will die from end-stage heart failure annually. Heart transplantation is the treatment of choice for medically refractory end-stage heart failure but can be performed in only about 2,500 patients per year. Circulatory support with a left ventricular (LV) assist device (VAD) can prolong survival and improve quality of life for patients unlikely to receive heart transplants and for those awaiting transplantation. The recent introduction of compact and durable rotary VADs has the potential to further prolong patient survival and VAD dwell times. This technological advance presents a new set of opportunities and challenges for physicians caring for patients with end-stage heart failure.
The immediate goals of VAD support are to relieve pulmonary vascular congestion via LV unloading, while maintaining adequate systemic blood flow at physiologic perfusion pressures. We hypothesized that those processes would result in geometric changes within the failing heart. Conversely, we hypothesized that changes in the size and shape of the left ventricle would be associated with clinically relevant alteration of the kinetics of blood delivery to the VAD in a significant number of patients.
This study was approved by the Institutional Review Board of the University of Iowa. Between May 1, 2008, and October 27, 2009, a total of 21 patients required mechanical assist support as a bridge to heart transplantation. Of those, two patients required biventricular mechanical support and are excluded from this report. The remaining 19 patients, who required LV mechanical support only, underwent the implantation of HeartMate II (Thoratec Corporation, Pleasanton, CA) rotary flow VADs. All patients were deemed candidates for heart transplantation by a multidisciplinary heart transplantation selection committee, using accepted criteria. Study data were obtained via retrospective review of clinical, hemodynamic, and imaging data from all 16 of the 19 patients who underwent VAD support for ≥70 days.
Transthoracic echocardiography was performed before VAD implantation, as well as early (7 ± 1 days) and again late (113 ± 21 days) after implantation, for surveillance or on the basis of a change in clinical status. Images were acquired in standard planes with the patient positioned supine or in a shallow left lateral recumbent position using a phased-array probe (nominal transmission frequency, 1–5 MHz) coupled to a Sonos 7500 or iE33 mainframe (Philips Medical Systems, Andover, MA). Ultrasound contrast agents were not used in this study.
LV ejection fraction was measured and reported by a clinical echocardiographer. In most cases, this was accomplished using the biplane method of disks. In cases in which endocardial border identification was suboptimal, ejection fraction was visually estimated. LV end-diastolic internal dimension (EDD) was measured retrospectively by a single author (R.M.W.), reported as the transverse endocardial diameter in the apical four-chamber view, at the level of the papillary muscle tips. Right ventricular (RV) EDD is reported as the transverse diameter at the most basal level of myocardium, just beneath the tricuspid valve, also assessed in the apical four-chamber view. LV length was measured from the center of the mitral annulus to apical endocardium. LV sphericity was calculated as LV EDD/LV length. The presence and severity of mitral regurgitation was evaluated visually on the basis of color Doppler images obtained in multiple planes and assigned a score on the basis of the following convention: 0 = none or trivial, 1 = mild, 2 = moderate, and 3 = severe.
Calculation of VAD Pulsatility Index (PI)
The degree of variation of instantaneous volumetric flow into the VAD during the cardiac cycle can be quantitated by calculating the PI, which is reported by the HeartMate II using the following formula :
VAD PI = ( VAD flow max − VAD flow min ) VAD flow mean × 10 .
Echocardiographic Assessment of VAD PI
LV-to-VAD flow was first assessed qualitatively, using color M-mode visualization obtained from a parasternal view, and again using color Doppler interrogation of flow from an apical view, adjusting the Nyquist limit to optimize visualization of low-velocity VAD inflow. Quantitative assessment was obtained by color Doppler–directed pulsed-wave Doppler velocimetry at the VAD inflow cannula ( Figure 1 ). Because VAD components are rigid (fixed orifice cross-sectional area), VAD volumetric flow should be directly proportional to the mean flow velocity. Thus, Doppler PI) was calculated using the following formula:
Doppler PI = ( Flow velocity max − Flow velocity min ) Flow velocity mean × 10 .