1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Conclusions
  7. References

©2012 Wiley Periodicals, Inc.

The aim of this study was to determine the effects of the HeartWare left ventricular assist device (HVAD) (HeartWare International, Inc, Framingham, MA) on hemodynamics, ventricular function, and exercise capacity. Between July 2009 and March 2011, 46 patients with advanced heart failure had implantation of the HVAD. Of these patients, 30 had subsequent assessments for transplantation and these patients formed the study cohort. Thirty patients had assessments at a mean of 201±86 days and 13 went on to have a second assessment at 351±86 days. There were marked improvements in hemodynamics and exercise capacity at both assessments, with significant decreases in right and left heart filling pressures; increases in cardiac index, New York Heart Association class, and peak exercise oxygen consumption; and reduction in pulmonary vascular resistance (all P<.05). Left ventricular end-diastolic and end-systolic dimensions were unchanged at the first assessment, although there was a significant reduction in end-diastolic diameter at the second assessment (P<.01). There were no significant changes in a visual grade of right ventricular function. After up to 1 year of support, the HVAD system results in significant benefits in overall heart failure status, with improved hemodynamics and exercise capacity. This occurs in the absence of marked changes in left ventricular size or right ventricular function.

The surgical management of advanced heart failure (HF) is in a state of flux, with, on one hand, increased demand for transplants unmet by either stable or decreasing numbers of donors,1,2 and, on the other hand, increased success of left ventricular (LV) assist devices (LVADs). The Heartmate II axial flow device (Thoratec Corp, Pleasanton, CA) has been shown to provide effective support in patients awaiting heart transplant,3 and is more effective in terms of both mortality and morbidity than the pulsatile Heartmate XVE pulsatile device (Thoratec Corp) for patients not suitable for transplant treated as destination therapy.4

The HeartWare ventricular assist device system (HVAD) (HeartWare International, Inc, Framingham, MA) is a centrifugal flow pump consisting of an impellar suspended by magnets and a displacement volume of 50 mL.5 The device consists of an inflow cannula placed through the apex of the left ventricle and the pump, which is placed within the pericardial space. Initial experience was reported by Wieselthaler and colleagues6 in a study of 23 patients who had the device implanted under the indication of “bridge to transplant.” With a mean duration of support of 167±143 days, actuarial survival was 91% at 6 months and 86% at 1 year. Subsequently, a study of 50 heart transplant candidates reported actuarial survival rates at 6, 12, and 24 months of 90%, 84%, and 79%.7

Whereas these studies have concentrated on survival, the clinical effects of this device in advanced HF on long-term hemodynamics, left and right ventricular (RV) function, and exercise capacity have not as yet been reported. We report our experience in a group of 30 patients who had the HeartWare device implanted as a bridge to transplant who subsequently underwent assessment with right heart catheterization, echocardiography, and cardiopulmonary stress testing.

Materials and Methods

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Conclusions
  7. References


Between July 2009 and March 2011, 46 patients with advanced HF had implantation of the HVAD under the indication of bridge to transplant or bridge to decision. Data were collected retrospectively until November 1, 2011. Survival curves for the total 46 patient cohort are presented in Figure 1. Of these patients, 30 had subsequent assessments for transplant and these patients form the study cohort. Thirteen patients had a second subsequent assessment and these data were also analyzed. The reasons that 16 patients did not undergo assessment are as follows: 8 patients died, 1 patient had a heart transplant, 1 had device explantation for myocardial recovery, and 6 patients were ongoing with device support although had not as of yet undergone an assessment. The characteristics of the 30 patient cohort are shown in Table I and baseline medications are shown in Table II. Of note, where applicable, we have taken oral medications as the last outpatient stable regimen in patients who were decompensated and required inotropes before device implantation. Most patients were taking furosemide as a loop diuretic and those taking bumetamide were, for analysis purposes, considered to be taking a dose of furosemide 40 mg for every 1 mg of bumetamide. Following implantation of the device, patients were treated with warfarin for a target international normalized ratio of 2.7 and an antiplatelet regimen guided by thromboelastography.8


Figure 1.  Kaplan-Meier survival curves of patient survival after HeartWare left ventricular assist device (HVAD) implant, with events of transplant or explant for recovery censored.

Download figure to PowerPoint

Table I.   Patient Characteristics
Age, y47±12
Women, %27
Height, cm175±9
Weight, kg82±13
Body mass index27±3
Diagnosis, %
 Ischemic heart disease47
 Idiopathic dilated cardiomyopathy50
 Adult congenital heart disease3
Duration of heart failure <6 mo17
Intermacs, %
Table II.   Medications
 BaselineFirst AssessmentSecond Assessment
  1. Abbreviations: ACE, angiotensin-converting enzyme; IABP, intra-aortic balloon pump.

ACE inhibitor, %71100100
β-Blocker, %6183100
Aldosterone antagonist, %593546
Furosemide, mg190±97155±85222±85
Sildenafil, %01715
Inotropes, %6300
IABP, %2000


The transplant assessment included routine biochemistries, right heart catheterization, echocardiography, and cardiopulmonary stress test. This is typically performed about 6 months after the device implantation after recovery from surgery with a view to being listed for transplantation. Assessments were then repeated at approximately 6-month intervals. All assessments were done on full HVAD flow. The right heart catheterization was performed with a Swan-Ganz catheter with central venous access and cardiac outputs measured with the thermodilution technique. Transpulmonary gradient was defined as the mean pulmonary artery pressure – pulmonary artery wedge pressure. Pulmonary vascular resistance index was calculated as: (transpulmonary gradient/cardiac index) × 80. RV stroke work index was defined as: (mean pulmonary artery pressure – mean right atrial pressure) × (stroke volume/body surface area). A cardiopulmonary stress test for determination of peak oxygen consumption (VO2), percent of predicted peak VO2, VO2 at anaerobic threshold, and the ventilatory equivalent for carbon dioxide (VE/VCO2) was performed on a bicycle ergometer (Sensormedics, Yorba Linda, CA). Anaerobic threshold was achieved in all tests and determined by the V slope method.9 VE/VCO2 was defined as the ratio of minute ventilation (VE) to carbon dioxide production (VCO2) at anaerobic threshold.10 Not all patients were able to undergo cardiopulmonary stress testing. For the baseline assessment, 18 patients had a stress test, for the first postoperative assessment there were 20 patients, and for the second assessment 8. Data analyzed from the echocardiograms were LV end-diastolic and end-systolic dimensions, and as a measure of RV function a scale of 1 to 5 based on visual assessment of RV function ranging from 1 as normal and 5 as severely impaired.

Statistical Analysis

Data are expressed as mean±standard deviation. Paired t tests were used to compare individually the first and second assessments with the baseline data, with the exception of the cardiopulmonary stress test data, for which unpaired t tests are used because there were fewer paired comparisons.


  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Conclusions
  7. References

Of the 30 patients at the time of data analysis (November 1, 2011), 3 patients died, 4 had undergone transplantation, and the other 23 remained on support. No patients required any form of RV mechanical support, although 63% of patients required inotropes, 20% an intra-aortic balloon pump, and 10% preoperative ultrafiltration to treat RV volume overload preoperatively (Table II). Postoperatively standard HF medications were used, and patients required in general a high dose of diuretics to treat underlying RV dysfunction to ensure that they were edema-free. Sildenafil was used only in patients with elevated pulmonary vascular resistance or in those who were intolerant to angiotensin-converting enzyme inhibitors due to renal insufficiency. The mean pump speed was 2525±174 revolutions per minute as set intraoperatively to allow for minimal aortic valve opening, and postoperatively this was not changed routinely.

The first postoperative assessment occurred at an average of 201±86 days and the second at 351±86 days. At the time of the first postoperative assessment, there were significant improvements in submaximal functional class as determined by New York Heart Association class and maximal exercise capacity on cardiopulmonary stress testing indicating improvement in HF status (Table III). This was manifested with a significant increase in peak VO2, percent of predicted peak VO2, and the ventilatory equivalent for carbon dioxide. Data at the second assessment showed similar and sustained benefits in terms of submaximal and maximal exercise capacity and, in addition, showed that VO2 at anaerobic threshold was also significantly reduced. There were also significant increases in serum sodium and reductions in bilirubin indicating reduced neurohumoral activation11 and reduced hepatic congestion, respectively. Bilirubin is known to correlate closely with right atrial pressure in HF patients.12 Hemoglobin was decreased as a result of some perioperative blood loss. Of note, the significantly lower bilirubin level indicated that this lower hemoglobin was not related to hemolysis.

Table III.   Functional and Exercise-Related Data and Biochemistries
 BaselineFirst Postoperative AssessmentSecond Postoperative Assessment
  1. Abbreviations: NYHA, New York Heart Association; VCO2, carbon dioxide excretion; VE, minute ventilation; VO2, oxygen consumption. aP<.01 vs baseline. bP<.05 vs baseline.

Time to follow-up, d201±86351±86
NYHA class3.6±0.52.3±0.7a1.7±0.8a
Peak VO2, mL/kg/min9.9±2.114.3±5.1a14.6±4.6a
Percentage of predicted peak VO232.3±7.441.4±12.7b42.0±15.0
VO2 at anerobic threshold, mL/kg/min8.2±1.19.9±2.610.3±2.2b
Hemoglobin, g/dL12.8±1.811.6±2.0b11.9±1.6
Sodium, mmol/L135±6137±4b138±2b
Creatinine, μmol/L113±37102±36104±23
Bilirubin, μmol/L32±2614±13a11±6a

There were no significant changes in heart rate or blood pressure (Table IV). There were marked changes in hemodynamics at the first assessment that were sustained at the second assessment. This included reductions in all right and left heart pressures, increased stroke volume index and cardiac index, and increased pulmonary arterial oxygen saturation. There was a significant decrease in pulmonary vascular resistance index, and as there was no significant change in transpulmonary gradient, this was due solely to the increase in cardiac index. RV stroke work index was significantly reduced at the first assessment. Changes in hemodynamics were much more marked than changes in echocardiographic parameters (Figure 2). RV function grade was not significantly changed throughout the study period. At the time of the second assessment, there was a significant reduction in LV end-diastolic diameter, although this was not seen at the first assessment. There were no significant changes in LV end-systolic diameter.

Table IV.   Hemodynamics and Echocardiography Data
 BaselineFirst Postoperative AssessmentSecond Postoperative Assessment
  1. Abbreviations: BP, blood pressure; LV, left ventricular; O2, oxygen; RV, right ventricular. aP<.01 vs baseline. bP<.05 vs baseline.

Time to follow-up, d201±86351±86
Heart rate, beats per min89±2189±1083±12
Systolic BP, mm Hg98±1298±1188±15
Diastolic BP, mm Hg61±1368±1361±13
Right atrial pressure, mm Hg13±68±6a8±5
Pulmonary arterial systolic pressure, mm Hg53±1537±15a34±10a
Pulmonary arterial diastolic pressure, mm Hg25±616±8a14±6a
Pulmonary arterial mean pressure, mm Hg37±825±10a22±7a
Pulmonary arterial wedge pressure, mm Hg26±615±7a13±6a
Cardiac index, L/min/m21.7±0.42.3±0.4a2.2±0.2a
Transpulmonary gradient, mm Hg11±59±59±4
Pulmonary vascular resistance index, dyn sec cm5/m2639±471315±154a321±122
Pulmonary arterial O2 saturation, %56±863±8b61±6
RV function grade3.3±1.33.0±1.13.5±1.1
RV stroke volume index, mL/m220.8±6.426.2±4.5a26.6±4.5
RV stroke work index, mm Hg mL/m2438±160405±201b354±125
LV end-diastolic dimension, mm7.2±1.06.9±1.26.3±1.6a
LV end-systolic dimension, mm6.4±1.26.3±1.36.0±2.0

Figure 2.  (A) Pulmonary arterial wedge pressure (PAW) and left ventricular (LV) end-diastolic diameter (LVEDD) over time. At the time of the first postoperative assessment, there were significant decreases in PAW, although no changes in LVEDD. At the second assessment, there was sustained reduction in PAW and also a significant reduction in LVEDD compared with preoperative values. (B) Right atrial (RA) pressure and right ventricular (RV) function grade over time. At the time of the first assessment, there was a significant reduction in RA pressure and at the second assessment this was not significantly different compared with preoperative values, although of a similar mean value of the first assessment. RV function grade did not change compared with preoperative values. *P<.05, **P<.01 both compared with preoperation.

Download figure to PowerPoint


  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Conclusions
  7. References

This is the first study to document the hemodynamic, echocardiographic, and functional effects of the HVAD in patients with advanced HF being supported under the indication of bridge to transplant. After up to 1 year of support, we have shown marked changes in hemodynamics, improved functional status, and improvement of biochemical markers of neurohumoral activation and hepatic congestion. Echocardiographic measures of LV volumes change to a much lesser extent, and there is no change in RV function. These data are previously unreported, because past studies have concentrated on outcomes,6,7 and also reported effects of the HVAD on biochemistries. Strueber and colleagues7 reported hemodynamics after 48 hours of support, compared with up to approximately 1 year in the present study.

Pressure vs Volume Changes

Whereas this is the first report of HVAD hemodynamics, other investigators have looked at the hemodynamic effects of the Heartmate II axial flow device, which is another continuous-flow device. Haft and colleagues13 compared the Heartmate XVE pulsatile device with the Heartmate II. They showed that both devices provided marked hemodynamic improvement, similar to what we have demonstrated with the HVAD. The Heartmate XVE produced greater effects on LV end-diastolic volumes than the Heartmate II (Heartmate XVE: 7.4±1.2 mm baseline – 5.0± 1.2 mm at 3 months vs Heartmate II: 7.3±1.1 – 6.0± 1.0 mm, P<.05 vs Heartmate XVE at 3 months). In comparison with our data (Table IV), the LV unloading is even less with the HVAD at 6 months compared with the Heartmate II. Similarly, Kato and colleagues14 showed that continuous-flow devices produce fewer effects than pulsatile devices on LV dimensions (particularly end-systolic dimension in this report), although, again, the effects of the HVAD are less than the other continuous-flow devices in this study. It is interesting to note that despite the absence of marked changes in left volumes with the HVAD, peak exercise capacity is similar with the Heartmate II device as shown by Haft and colleagues, who reported a peak exercise VO2 of 15.6±4.7 mL/kg/min (treadmill test). We report a peak exercise VO2 of 14.3±5.1 mL/kg/min on a bicycle, which is expected to produce a somewhat lower value of peak exercise VO2 than that achieved with a treadmill.15 Similarly the improvement in biochemistries that we report is similar to the other studies with more marked ventricular unloading. Thus, ventricular volume unloading does not seem to be essential for improvement in HF status, although this may have an impact on any potential for myocardial recovery.16 We also do not know what impact the speed of the device has on ventricular unloading, and it would be important for future prospective studies to look at the hemodynamic effects of different speeds. Hayward and colleagues17 have studied changing pump speeds with the Ventr- Assist centrifugal LVAD (Ventracor, Sydney, Australia), showing a strong correlation of device speed with flow rates, although with noted marked interpatient variation, suggesting that device speeds need to be individually tailored to patients.

Effects on RV Function

The effects of continuous-flow ventricular assist devices on RV function are as of yet unclear, although our echocardiographic and hemodynamic data suggest that overall RV function does not improve. RV function does not visually change. There is also a decrease in RV stroke work index, and this likely reflects the marked changes in both preload and afterload that the right ventricle experiences after the HVAD coupled with a relatively small increase in stroke volume. It is known that the Heartmate II devices are associated with less right HF than the Heartmate XVE device, the reasons for which are also unclear, although they may relate to reduced complications that might lead to right HF such as bleeding, renal failure, and infection.4 Maeder and colleagues18 have shown that the VentrAssist LVAD did not change RV function when using a visually assessed scale of RV function similar to our method. What is notable in the current study is that no patients required any form of temporary RV mechanical support, with an average RV function of moderate (grade 3) to moderate to severe (grade 4). Indeed no patients had biventricular assist devices implanted during the study period at all at this center. This is most likely a result of intensive preoperative and postoperative management of RV volume overload,19 although whether there are any specific effects of the HVAD on RV function is unclear. Notably, our patients were taking high doses of diuretics after device implantation, reflecting some degree of persistent RV dysfunction (Table II). Consistent with our results, Palardy and colleagues20 studied the effects of inpatient optimization with diuretic therapy and inotropes preimplantation of an LVAD on RV function, and then postoperative RV function after 3 months of support in 20 patients (9 pulsatile and 11 axial flow). After a median support time of 123 days, RV size and global RV dysfunction failed to improve.


The current study, while providing unique data, is a single-center investigation and so will need to be replicated by other center experiences, although the hemodynamic and functional effects are clear. All investigations were a part of standard clinical practice. Fewer than half the patients underwent the second assessment, and this is merely a reflection of the longer time with the device that these patients had, although this may have introduced some bias into the results of the second assessment. Assessment of RV function is difficult, without any agreed gold standard, so this is limited in the current study.


  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Conclusions
  7. References

The HVAD results in marked and sustained improvements in hemodynamics, biochemistries, and exercise capacity in advanced HF, although effects on LV volume unloading occur to a much lesser extent.

Conflicts of interest:  None.


  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Conclusions
  7. References
  • 1
    MacGowan GA, Parry G, Schueler S, Hasan A. The decline in heart transplantation in the UK, Editorial. BMJ. 2011;342:d2483.
  • 2
    Deng MC, Ardehali A, Shemin R, et al. Relative roles of heart transplantation and long-term mechanical circulatory support in contemporary management of advanced heart failure – a critical appraisal 10 years after rematch. Eur J Cardiothorac Surg. 2011;40:781782.
  • 3
    Miller LW, Pagani FD, Russell SD, et al. Use of a continuous-flow device in patients awaiting heart transplantation. N Engl J Med. 2007;357(9):885896.
  • 4
    Slaughter MS, Rogers JG, Milano CA, et al. Advanced heart failure treated with a continuous-flow left ventricular assist device. N Engl J Med. 2009;361(23):22412251.
  • 5
    Larose JA, Tamez D, Ashenuga M, Reyes C. Design concepts and principle of operation of the HeartWare ventricular assist system. ASAIO J. 2010;56(4):285289.
  • 6
    Wieselthaler GM, O’Driscoll G, Jansz P, Khaghani A, Strueber M; HVAD Clinical Investigators. Initial clinical experience with a novel left ventricular assist device with a magnetically levitated rotor in a multi-institutional trial. J Heart Lung Transplant 2010;29(11):12181225.
  • 7
    Strueber M, O’Driscoll G, Jansz P, Khaghani A, Levy WC, Wieselthaler GM; HeartWare Investigators. Multicenter evaluation of an intrapericardial left ventricular assist system. J Am Coll Cardiol 2011;57(12):13751382.
  • 8
    Gordon B, Biss T, Butt T, et al. Platelet mapping thromboelastography for individualized antiplatelet treatment after implantation of the HeartWare ventricular assist device. J Heart Lung Transplant. 2011;30:S208S209. Abstract.
  • 9
    Wasserman K, Beaver WL, Whipps BJ. Gas exchange theory and the lactic acidosis (anaerobic) threshold. Circulation. 1990;81(suppl II):II-14II-30.
  • 10
    MacGowan GA, Cecchetti A, Murali S. Ventilatory drive during exercise in congestive heart failure. J Cardiac Failure. 1997;3:257262.
  • 11
    Lee WH, Packer M. Prognostic importance of serum sodium concentration and its modification by converting enzyme inhibition in patients with severe chronic heart failure. Circulation. 1986;73:257267.
  • 12
    MacGowan GA, Neely D, Peaston R, et al. Evaluation of NT-proBNP to predict outcomes in advanced heart failure. Int J Clin Pract. 2010;64:892899.
  • 13
    Haft J, Armstrong W, Dyke DB, et al. Hemodynamic and exercise performance with pulsatile and continuous-flow left ventricular assist devices. Circulation. 2007;116:I-8I-15.
  • 14
    Kato TS, Chokshi A, Singh P, et al. Effects of continuous-flow vs pulsatile-flow left ventricular assist devices on myocardial unloading and remodeling. Circ Heart Fail. 2011;4:546553.
  • 15
    Page E, Cohen-Solal A, Jondeau G, et al. Comparison of treadmill and bicycle exercise in patients with chronic heart failure. Chest. 1994;106(4):10021006.
  • 16
    Krabatsch T, Schweiger M, Dandel M, et al. Is bridge to recovery more likely with pulsatile left ventricular assist devices than with nonpulsatile-flow systems? Ann Thorac Surg. 2011;91(5):13351340.
  • 17
    Hayward CS, Salamonsen R, Keogh AM, et al. Effect of alteration in pump speed on pump output and left ventricular filling with continuous-flow left ventricular assist device. ASAIO J. 2011;57(6):495500.
  • 18
    Maeder MT, Leet A, Ross A, et al. Changes in right ventricular function during continuous-low left ventricular assist device support. J Heart Lung Transplant. 2009;28:360366.
  • 19
    Butt TA, Oezalp F, O’Leary D, et al. Right ventricular optimisation in patients with biventricular failure receiving a left ventricular assist device – a safe strategy to avoid mechanical RV support. J Heart Lung Transplant. 2011;30:S44S45, Abstract.
  • 20
    Palardy M, Nohria A, Rivero J, et al. Right ventricular dysfunction during intensive pharmacologic unloading persists after mechanical unloading. J Card Fail. 2010;16:218224.