Secondary Logo

Journal Logo

Adult Circulatory Support

Exercise Performance During the First Two Years After Left Ventricular Assist Device Implantation

Marko, Christiane*; Xhelili, Endri; Lackner, Thomas; Zimpfer, Daniel§; Schima, Heinrich†§¶; Moscato, Francesco†¶

Author Information
doi: 10.1097/MAT.0000000000000569
  • Free


An increasing number of patients receive left ventricular assist devices (LVADs) not only as a bridge to transplantation but also for their lifetimes, as a destination therapy.1 Although their quality of life and their heart failure symptoms improve immediately after LVAD implantation,2 most patients still have a markedly impaired exercise capacity.3–6 Dunlay et al.7 showed in their study only slight improvement of the peak oxygen consumption (VO2) (+0.9 ml/kg/min) when comparing pre-LVAD implantation with 1 year later. This improvement was not statistically significant, with most of the patients (72%) having a peak VO2 <14 ml/min/kg independent on indication, etiology, or age at implant. A better Interagency Registry for Mechanically Assisted Circulatory Support (INTERMACS) level at implantation also did not correlate with improved peak VO2.8 Leibner et al.9 found in their population that after a slight increase of peak VO2 within the first 3 to 6 months post-LVAD implantation, no further changes could be observed up to one year, and the percentage of the predicted value remained severely limited.

More recently, the number of patients on long-term LVAD support has been substantially growing, leading to cardiac rehabilitation (CR) programs aimed at improving the limited functional capacity and the health status and at the optimization of life quality in this particular patient population. Studies are available concerning the effectiveness and safety of exercise training during a CR program just after LVAD implantation10–15; however, the available evidence is still not strong enough to draw conclusive guidelines for LVAD patients.16 To the authors knowledge, data are lacking in addition to these reports about the long-term exercise performance related to a second CR stay and whether this effectively improves patients’ physical performance. In this study, we report a single-center experience with a series of patients who underwent two CRs within the first 2 years after LVAD implantation.


A retrospective observational cohort study was approved by the Institutional Review Board of Lower Austria. Data from LVAD patients were collected and retrospectively analyzed among all patients who underwent a stationary rehabilitation program twice between September 2010 and March 2014. In our center, patients are usually informed at the end of the first CR about the possibility of a second CR in about a year. After this year has passed, if patients are still on device and if there are no contraindications for a second CR stay (i.e., occurrence of severe adverse event), patients are contacted by phone and referred to the rehabilitation clinic. Among the patients who underwent both rehabilitation stays, those with a time between surgery and first rehabilitation beginning greater than 120 days were excluded from the analysis. Data related to the training sessions and the exercise tests during both CR stays were used to evaluate the changes in physical performances within the first 2 years after pump implant and the margins for improvement available from a repeated, later CR.

Rehabilitation Program, Medical Training Therapy, and Diagnostic Tests

For both rehabilitation periods, the same rehabilitation program and diagnostic tests were performed as reported in a previous study.15

The rehabilitation program consisted of aerobic and strength trainings, as well as walking and gymnastics. For the determination of the intensity of the training sessions, we used the perceived level of exertion according to the Borg scale.17 An exertion of somewhat harder intensity (grade 13 in the Borg scale) was attempted. Aerobic training consisted of bicycle ergometry with an interval training schema. Strength training consisted of two series of 12 repetitions of five muscle groups of the lower extremities. Walking and gymnastic training was organized into groups with the aim of shifting the patient into a better performing group. There were five different walking groups each walking along paths covering different distances, elevation, and duration (W1 was the group with the lowest distance and duration, W5 with the highest). The gymnastic groups were also divided into five grades performing different exercises (coordination, strength, and balance training) with different intensities and in different body positions.

A cardiopulmonary exercise test (CPET) was routinely performed at the end of each rehabilitation period, and its protocol was adapted to the estimated workload of the patient with an aimed duration of 8–12 minutes. A 6 minute walk test (6MWT)18 was introduced in the clinical routine only in February 2013, and data were therefore available only for the second CR.

Blood serum concentration of N-terminal pro-brain natriuretic peptide (NT-proBNP), evaluated using an immunologic test, the glomerular filtration rate (GFR), calculated using the modification of diet in renal disease formula, the alanine aminotransferase (ALT), and the hemoglobin concentration were also measured.

The pharmacologic heart failure therapy and its changes between the two rehabilitations periods were also analyzed. The most used angiotensin converting enzyme –inhibitors (such as ramipril and lisinopril), β-blocker (bisoprolol), and diuretics (furosemide, spironolactone, and xipamid) were considered.

Acquired Data and Statistical Analysis

The number of training sessions, the training intensity and duration for the bicycle ergometry training, and the lifted weight for strength training were recorded. Walking and gymnastic groups in which the patient trained were recorded as well. Adverse events and complications associated with training or leading to a withdrawal from the training were also carefully documented.

The bicycle training was performed in an interval mode using standardized modules with increasing intensity of the peak workload and the recovery phases. The high-intensity phase lasted 20–30 seconds and the recovery phase, 60 seconds. To get a more precise quantification of the interval training, mean workload (p_mean) was calculated by weighting the workloads for both the high-intensity and the recovery phases by the respective durations using the below equation: p_mean = (p_peak × t_peak + p_rec × t_rec)/(t_peak + t_rec). Here, p_peak represents the peak workload at the phase of high training intensity, t_peak, the duration of the high-intensity phase, p_rec, the workload of the recovery phase, and t_rec, the duration of the recovery phase.19 Values were calculated over the whole active training time, excluding the 3 minutes warm-up at the beginning and the 3 minutes cool-down phase with no or very low load.

The following parameters from the CPET were analyzed: duration of the test, peak workload, percentage of expected workload, relative peak VO2, absolute peak VO2 normalized by body weight, percentage of peak VO2 according to Wassermann formula, metabolic equivalents, workload at the first ventilatory threshold 1 (VT1), VO2 at VT1, percentage of predicted peak VO2 at VT1, respiratory exchange ratio (RER) at peak VO2, and ventilatory equivalent ratio for carbon dioxide (VE/VCO2) at VT1.

Four time points were considered in the analysis: the beginning and the end of the 1st rehabilitation period (in the following text denoted by B1 and E1) as well as the beginning and the end of the 2nd rehabilitation period (in the following text denoted by B2 and E2). Where not otherwise specified, variables are considered averaged during the whole rehabilitation stay.

To statistically assess the observed changes, Student’s t-test and the Wilcoxon signed-rank test for paired differences were used. The latter test was used when the variables were not normally distributed. Normality was tested using the Shapiro–Wilk test. A level of significance p < 0.05 was considered statistically significant. Statistical analysis was performed using SPSS 22.0 (IBM, Armonk, NY).



Of the 91 patients who attended a CR just after LVAD implantation, between September 2010 and July 2013, 32 were eligible for a second rehabilitation. Patients were excluded if they were not on device (transplanted n = 20, dead n = 17, and weaned n = 2), or if they had at least one adverse event (n = 20), which was either severe or did not result in a full recovery. From the 32 eligible patients, 15 refused a second CR for personal reasons, and 1 patient completed the 2nd CR in another center. Among the 16 patients who attended the second rehabilitation, one was excluded from the study because of a prolonged time (>120 days) between surgery and commencement of the CR. For a schematic tabulation of the study patients, please refer to Figure 1. The remaining 15 patients came directly from the hospital to the rehabilitation center for the first time 39 ± 18 days after LVAD implantation and stayed for 35 ± 6 days and for the second time 542 ± 197 days after implantation, staying 27 ± 7 days. The demographics, clinical characteristics, and comorbidities of the study subjects are summarized in Table 1. Among these parameters, a remarkable increase of the body weight at the second CR compared with the first CR should be noted (from 82.5 ± 10.9 kg to 92.6 ± 10.2 kg; p < 0.001). One patient received an implantable cardioverter defibrillator (ICD) and one patient a cardiac resynchronization therapy (CRT) before attending the second rehabilitation stay. Nicotine abuse decreased significantly in the second stay; no patient was a smoker at this time, compared with four in the first CR (p = 0.032). Six of the study patients had an adverse event between the two CRs; however, these were not judged as a contraindication for a second CR.

Table 1.
Table 1.:
Baseline Demographics, Laboratory Tests, and Most Used Medication at the End of Both CRs (E1–E2)
Figure 1.
Figure 1.:
Study flow diagram.

Medical Training Therapy

During the 2nd CR (B2 vs. E2), an improvement of the duration of the training and of the average workload for the bicycle ergometry training (Table 2) was observed. Also, the average lifted weight of five muscle groups increased (Table 3).

Table 2.
Table 2.:
Bicycle Ergometry Training Results at the Four Investigated Time-Points: Beginning and End of the First and the Second Cardiac Rehabilitation Stay (B1, E1, B2, E2, Respectively)
Table 3.
Table 3.:
Strength Training Results at the Four Investigated Time-Points: Beginning and End of the First and the Second Cardiac Rehabilitation Stay (B1, E1, B2, E2, Respectively)

All patients performed bicycle training during the first CR. One of the patients could not perform bicycle training at the second CR because of orthopedic problems. Compared the end of the first CR with the end of the second CR (E1 vs. E2), the number and duration of the training sessions and the achieved mean workload in the bicycle ergometer training improved (Table 2).

All patients performed strength training during both CR. The average weight used for the five muscle groups improved as shown in Table 3.

Patients also showed improvement in their gymnastic group (2 ± 1 to 3 ± 1; p = 0.014) and walking group (2 ± 1 to 3 ± 1; p = 0.001). There were no training-related complications recorded during both CR periods.

Diagnostic Tests

The results from the analysis of the CPET are shown in Table 4. Two patients were not able to perform the CPET because of mask phobia and orthopedic problems CR, the remaining 13 performed all the CPET at the end of the first period. At the end of the second CR, the patient with mask phobia could perform only a cardiac stress test; therefore, data about ventilation are available only for 13 of the 14 patients. Peak workload and the percentage of the expected workload increased significantly. However, relative peak VO2 did not increase, and even slightly decreased although not significantly, and absolute peak VO2 increased, but not significantly. A considerable gain of body weight (82.5 ± 10:9 vs. 92.6 ± 10.2 kg; p < 0.001) was observed in the same period (Table 1). The percentage of the predicted peak VO2 at the VT1 slightly increased, although not significantly. The VE/VCO2 at the VT1 slightly decreased. All patients performed the CPET until exhaustion, reaching a RER of 1.09 ± 0.08, very close to the value of 1.1 that is accepted as an excellent exercise effort.20 The number of patients who reached a RER >1.1 was five during the first CR and six during the second CR.

Table 4.
Table 4.:
Cardiopulmonary Exercise Test

Because of a later introduction of the 6MWT, only 10 patients performed the 6MWT during the second CR. These could significantly improve their walked distance during the training at the 2nd CR stay from 421 ± 126 m to 480 ± 133 m (p = 0.040).

A clear reduction of the NT-proBNP from 3325 ± 3234 pg/ml to 1390 ± 1212 pg/ml (p = 0.003) was observed between the two CR periods. Renal function decreased as indicated from the increase of serum creatinine from 1.3 ± 0.6 mg/dl to 1.66 ± 0.6 (p = 0.002). GFR decreased from 70.8 ± 40.8 to 53.0 ± 23.8 ml/min (p = 0.003). A slight increase in hemoglobin from 11.3 ± 1.5 vs. 12.3 ± 2.1 g/dl (p = 0.108) was observed between the two periods, but still staying beneath the normal range. The ALT stayed at the normal range during this period (Table 1).

Pump speed setting did not change between the two CRs leading to the same mean pump flow rate during the first and second CR: 5.7 ± 0.5 vs. 5.3 ± 1.0 L/min (p = 0.278), respectively.

The only significance in pharmacologic therapy was the dosage of bisoprolol, which increased from an average of 2.7 ± 1.6 to 4.6 ± 4.1 mg/day (p = 0.027). In five patients, the administration of spironolactone was discontinued (Table 1).


Fifteen patients finished two CR stays within the first 2 years after LVAD implantation with observable improvement of muscular strength and coordination, as well as of submaximal aerobic activity. The workload during the bicycle interval training was seen to be greater at the second CR (Table 2). A significant increase of strength in four of the five muscle groups targeted by the training program was also observed (Table 3). Further, patients were able to nearly double the walking distance during training and to perform more challenging gymnastic exercises.

Despite this general improvement, the relative cardiopulmonary performance remained nearly the same (Table 4). A considerable weight gain was observed in the same period (Table 1), with all but one patient gaining about 10 kg since their first CR. This might reflect the reversal of several deleterious factors in advanced heart failure patients, that is for example, systemic inflammatory and neurohumoral activation. The increase in weight might have influenced the slightly decreased peak VO2. Considering the absolute peak VO2 (not normalized by body weight), an increase, not statistically significant, could be observed. The peak workload, on the other hand, significantly increased probably because of an improvement in neuromuscular coordination and more practice during bicycling. Despite a higher workload, the percent of its expected value remains rather low (42 ± 12%). The low peak VO2 values are comparable to data from other studies with patients on support more than 1 year. Leibner et al.9 describes in his population a peak VO2 of 11.2 ± 1.7 ml/kg/min. In 70% of the patients from the study by Dunlay et al.,7 the peak VO2 was lower than 14 ml/kg/min. The severely reduced left ventricular function, despite mechanical circulatory support, and depressed right ventricular function are among the potential factors contributing to the limited exercise capacity. Unfortunately, because of sparse echocardiographic data, the relative importance of these factors could not be addressed in the current study. In addition, respiratory abnormalities, peripheral factors including deconditioned skeletal muscle, and a pump set at a constant speed also appear as limiting factors.21 The reverse remodeling of myocardial cellular hypertrophy, eventually leading to atrophy without functional improvement of cardiac function, may also contribute to reduced exercise capacity.22 However, this aspect could not be assessed because histologic evaluation of cellular remodeling was unavailable.

The scientific community has started questioning the validity of peak VO2 as a parameter that describes the condition of LVAD patients, favoring parameters from submaximal tests (e.g., 6MWT) to reflect the condition of these patients.23 For heart failure patients, the 6MWT distance and the peak VO2 are linearly correlated.21,23 The 6MWT distance apparently also reaches a plateau. In a previous large multicenter destination therapy trial, the mean distance covered in the 6MWT after 1 year remained nearly the same as that measured 3 months post-LVAD implantation (319 ± 191 vs. 318 ± 164 m).24 In another retrospective study, the mean distance was 393 ± 290 m 1 year after implantation.25 In this study, the distance covered at the end of the 2nd CR rose to 480 ± 133 m, remarkably better than the cited values approaching the 10th percentile of a 50 to 59 year old healthy man.26 Because the 6MWT was introduced later, we could not determine the improvement relative to the first CR. However, the longer distance, together with other training variables, may indicate a benefit of a second CR, especially in the submaximal load range. A reason for the dissociation between maximal and submaximal performance may be that LVADs are able to provide sufficient support at rest and during submaximal work, but insufficient support during maximal performance, because of the constant pump speed. It has indeed already been shown that exercise capacity with increasing pump speed can be improved compared with that at constant pump speed,21 especially for patients with more preserved right ventricular function.27

The value of the percentage of the predicted maximal VO2 at the VT1, a parameter derived from submaximal workload and therefore independent of patient motivation was 30 ± 7%, below the threshold of 40%,28 expressing again deconditioning of the skeletal muscle after two CRs. This deconditioning also shows that although at the first CR all patients were instructed on how to continue the training at home, they were probably not following these recommendations. Notwithstanding potential remains to improve aerobic muscle metabolism with regular supervised exercise during an ambulatory CR program.

The clear reduction of the NT-proBNP to1390 ± 1212 pg/ml hints at an improvement of the heart failure condition and hemodynamics, but it is still in a high range. At the second CR, the only observed changes in medication were the discontinuation of diuretics in five patients, probably because of an observed deterioration of renal function and an increase in β-blocker dosage, as recommended for LVAD patients.29 Because of this increase, the resting heart rate recorded before the CPET was lower at the second CR in comparison with the first CR (70 ± 9 vs. 83 ± 15 bpm; p = 0.012). However, the peak heart rate during the CPET was unchanged (111 ± 29 vs. 107 ± 24 bpm; p = 0.545). Thus, it seems unlikely that chronotropic incompetence would lead to a lack of improvement in the peak VO2.

Certain limitations on these results are because of the small sample size and the retrospective character of this study. From the 91 patients that performed the first CR, only 32 were eligible for the second CR, with 16 performing it in our center and one in another center (Figure 1). Nevertheless, the patients in this study yield data similar to one of our previous studies which included 41 consecutive patients,10 suggesting the representativeness of this sample. Still there is a risk of a biased selected population, because of the low numbers. Nonetheless, the observations presented here should be valuable for future comparison. As already mentioned, systematic evaluation of any specific echocardiographic data about the right ventricle and the aortic and mitral valve was not possible because of sparse data. Nevertheless, it would be very interesting to correlate these data with the parameters of the CPET especially the VO2 max. Unfortunately, the VE/VCO2 at VT1 and not the minute ventilation carbon dioxide production slope (VE/VCO2 slope) was documented; the latter would have been a much better prognostic parameter in the CPET.


This study shows that a clear, long-term improvement in submaximal physical activity can be achieved, but in the absence of changes in peak oxygen uptake. A second CR about 1.5 years after LVAD implantation improves physical performance, indicating that training of the peripheral musculature is apparently improved and that a longer continuation of medically guided exercise training in an ambulatory setting appears warranted.


1. Kirklin JK, Naftel DC, Pagani FD, et al. Seventh INTERMACS annual report: 15,000 patients and counting. J Heart Lung Transplant 2015.34: 14951504.
2. Jakovljevic DG, McDiarmid A, Hallsworth K, et al. Effect of left ventricular assist device implantation and heart transplantation on habitual physical activity and quality of life. Am J Cardiol 2014.114: 8893.
3. Compostella L, Russo N, Setzu T, Compostella C, Bellotto F. Exercise performance of chronic heart failure patients in the early period of support by an axial-flow left ventricular assist device as destination therapy. Artif Organs 2014.38: 366373.
4. Corrà U, Pistono M, Mezzani A, et al. Cardiovascular prevention and rehabilitation for patients with ventricular assist device from exercise therapy to long-term therapy. Part I: Exercise therapy. Monaldi Arch Chest Dis 2011.76: 2732.
5. Jung MH, Gustafsson F. Exercise in heart failure patients supported with a left ventricular assist device. J Heart Lung Transplant 2015.34: 489496.
6. Loyaga-Rendon RY, Plaisance EP, Arena R, Shah K. Exercise physiology, testing, and training in patients supported by a left ventricular assist device. J Heart Lung Transplant 2015.34: 10051016.
7. Dunlay SM, Allison TG, Pereira NL. Changes in cardiopulmonary exercise testing parameters following continuous flow left ventricular assist device implantation and heart transplantation. J Card Fail 2014.20: 548554.
8. Benton CR, Sayer G, Nair AP, et al. Left ventricular assist devices improve functional class without normalizing peak oxygen consumption. ASAIO J 2015.61: 237243.
9. Leibner ES, Cysyk J, Eleuteri K, El-Banayosy A, Boehmer JP, Pae WE. Changes in the functional status measures of heart failure patients with mechanical assist devices. ASAIO J 2013.59: 117122.
10. Alsara O, Reeves RK, Pyfferoen MD, et al. Inpatient rehabilitation outcomes for patients receiving left ventricular assist device. Am J Phys Med Rehabil 2014.93: 860868.
11. Ben Gal T, Piepoli MF, Corrà U, et al.; Committee on Exercise Physiology & Training of Heart Failure Association and endorsed by Cardiac Rehabilitation Section of European Association for Cardiovascular Rehabilitation and Prevention of ESC: Exercise programs for LVAD supported patients: A snapshot from the ESC affiliated countries. Int J Cardiol 2015.201: 215219.
12. English ML, Speed J. Effectiveness of acute inpatient rehabilitation after left ventricular assist device placement. Am J Phys Med Rehabil 2013.92: 621626.
13. Hayes K, Leet AS, Bradley SJ, Holland AE. Effects of exercise training on exercise capacity and quality of life in patients with a left ventricular assist device: a preliminary randomized controlled trial. J Heart Lung Transplant 2012.31: 729734.
14. Kerrigan DJ, Williams CT, Ehrman JK, et al. Cardiac rehabilitation improves functional capacity and patient-reported health status in patients with continuous-flow left ventricular assist devices: The Rehab-VAD randomized controlled trial. JACC Heart Fail 2014.2: 653659.
15. Marko C, Danzinger G, Käferbäck M, et al. Safety and efficacy of cardiac rehabilitation for patients with continuous flow left ventricular assist devices. Eur J Prev Cardiol 2015.22: 13781384.
16. Scheiderer R, Belden C, Schwab D, Haney C, Paz J. Exercise guidelines for inpatients following ventricular assist device placement: A systematic review of the literature. Cardiopulm Phys Ther J 2013.24: 3542.
17. Borg GA. Psychophysical bases of perceived exertion. Med Sci Sports Exerc 1982.14: 377381.
18. ATS Committee on Proficiency Standards for Clinical Pulmonary Function Laboratories. ATS statement: Guidelines for the six-minute walk test. Am J RespirCrit Care Med. 2002;166:111117
19. Tschakert G, Hofmann P. High-intensity intermittent exercise: Methodological and physiological aspects. Int J Sports Physiol Perform 2013.8: 600610.
20. Guazzi M, Adams V, Conraads V, et al.; European Association for Cardiovascular Prevention & Rehabilitation; American Heart Association: EACPR/AHA Scientific Statement. Clinical recommendations for cardiopulmonary exercise testing data assessment in specific patient populations. Circulation 2012.126: 22612274.
21. Jung MH, Hansen PB, Sander K, et al. Effect of increasing pump speed during exercise on peak oxygen uptake in heart failure patients supported with a continuous-flow left ventricular assist device. A double-blind randomized study. Eur J Heart Fail 2014.16: 403408.
22. Ambardekar AV, Dorosz JL, Cleveland JC Jr, Lindenfeld J, Buttrick PM. Longitudinal left ventricular structural and functional imaging during full support with continuous-flow ventricular assist devices: A retrospective, preliminary analysis. J Heart Lung Transplant 2012.31: 13111313.
23. Nahumi N, Morrison KA, Garan AR, Uriel N, Jorde UP. Peak exercise capacity is a poor indicator of functional capacity for patients supported by a continuous-flow left ventricular assist device. J Heart Lung Transplant 2014.33: 213215.
24. Slaughter MS, Rogers JG, Milano CA, et al.; HeartMate II Investigators: Advanced heart failure treated with continuous-flow left ventricular assist device. N Engl J Med 2009.361: 22412251.
25. Allen JG, Weiss ES, Schaffer JM, et al. Quality of life and functional status in patients surviving 12 months after left ventricular assist device implantation. J Heart Lung Transplant 2010.29: 278285.
26. Casanova C, Celli BR, Barria P, et al.; Six Minute Walk Distance Project (ALAT): The 6-min walk distance in healthy subjects: Reference standards from seven countries. Eur Respir J 2011.37: 150156.
27. Mezzani A, Pistono M, Corra U, et al. Systemic perfusion at peak incremental exercise in left ventricular assist device recipients: Partitioning pump and native left ventricle relative contribution. IJC Heart Vessels 2014;4:4045
28. Mezzani A, Agostoni P, Cohen-Solal A, et al. Standards for the use of cardiopulmonary exercise testing for the functional evaluation of cardiac patients: a report from the Exercise Physiology Section of the European Association for Cardiovascular Prevention and Rehabilitation. Eur J Cardiovasc Prev Rehabil 2009.16: 249267.
29. Pistono M, Corrà U, Gnemmi M, et al. Cardiovascular prevention and rehabilitation for patients with ventricular assist device from exercise therapy to long-term therapy. Part II: long-term therapy. Monaldi Arch Chest Dis 2011.76: 136145.

continuous-flow left ventricular assist devices; cardiac rehabilitation; exercise performance

Copyright © 2017 by the ASAIO