Effects of Exercise During Long-term Support With a Left Ventricular Assist Device
Results of the Experience With Left Ventricular Assist Device With Exercise (EVADE) Pilot Trial
Background Long-term implantation of a left ventricular assist device (LVAD) may be a future alternative treatment for end-stage heart failure. The objective of the present study was to determine the hemodynamic effects of supine bicycle exercise and functional capacity during upright treadmill exercise in 10 patients after LVAD implantation placed for refractory heart failure as a bridge to cardiac transplantation.
Methods and Results With supine bicycle exercise, 46±25 days after device placement, heart and LVAD rates increased in parallel from 87±12 to 117±14 bpm and 82±18 to 107±21 bpm, respectively. Peak O2 consumption was 8.2±1.7 mL O2·kg−1·min−1. Fick systemic blood flow rose from 5.0±1.2 to 7.8±2.5 L/min. Right atrial and pulmonary capillary wedge pressures increased from 6±4 and 5±3 mm Hg to 12±5 and 13±8 mm Hg, respectively. End-diastolic left ventricular dimension increased from 3.9±1.3 to 4.8±1.6 cm; however, right ventricular dimension decreased from 3.2±1.0 to 2.3±0.9 cm. With upright treadmill exercise, peak O2 consumption was 14.1±2.9 mL O2·kg−1·min−1.
Conclusions This study indicates that exercise during long-term LVAD support is safe and is not limited by right heart decompensation. It also justifies a larger study to examine how exercise after LVAD implantation compares with that after cardiac transplantation.
Recently, the Food and Drug Administration approved the use of a pneumatic LVAD as a bridge to cardiac transplantation.1 This decision was prompted by improvements in device efficacy associated with a reduction in thromboembolic complications.2 3 4 Investigational use of a related but more portable vented electric LVAD in an outpatient setting for >500 days in humans suggests that long-term use of this or a related system may serve as an alternative to medical therapy or cardiac transplantation for selected patients in the future.3 5 6 7
We have previously reported methods of evaluating patients after pneumatic LVAD implantation during supine bicycle and treadmill exercise.8 9 The objective of the present study was to determine the hemodynamic effects of supine bicycle exercise and functional capacity during upright treadmill testing in a group of 10 patients with end-stage heart failure awaiting cardiac transplantation during long-term implantation of an LVAD. This was done to assess the safety and physiology of exercise testing before a planned multicenter trial comparing exercise capacity after LVAD placement with that after cardiac transplant, the EVADE trial.
The use of the LVAD in patients with refractory heart failure awaiting cardiac transplantation was approved by the Sharp Memorial Hospital Investigational Review Committee. Patients gave written informed consent for each protocol before testing.
A Thermo Cardiosystems Inc LVAD was urgently placed in 8 men and 2 women, 47±7 years old, who were in imminent danger of death within 24 hours. Patients had refractory and progressive heart failure secondary to idiopathic (n=6) or ischemic (n=4) cardiomyopathy despite maximal doses of inotropic medication and possibly intra-aortic balloon pump placement. Patients experienced symptoms of heart failure for 27±23 months before LVAD implantation.
All patients had previously been approved for cardiac transplantation. An additional 3 patients underwent LVAD placement during the time period of the study but were not able to participate because of either persistence of multisystem organ failure secondary to preoperative circulatory collapse (n=2) or aortic dissection bleeding at the outflow cannula insertion site (n=1). All study patients were successfully bridged to cardiac transplantation.
Device implantation was performed by standard techniques previously described in detail,10 including an LV apical inflow cannula, a pneumatically (IP, n=8) or electrically (VE, n=2) powered pump in the left upper abdominal quadrant, and an ascending aortic outflow cannula. The anterior pericardium was removed, and the free ends were left wide open. Unidirectional flow through the device was regulated by the use of one-way porcine xenograft valves. The LVAD functioned in a fill-to-empty mode, ejecting a nearly constant stroke volume after complete LVAD chamber filling. Pneumatic LVAD ejection duration was manually set to achieve complete emptying of the device to limit blood stasis. Anticoagulation therapy consisted of daily doses of 325 mg aspirin and 150 mg dipyridamole. No thromboembolic complications were observed. At explant, only typical postthoracotomy pericardial adhesions were found.
At the time of the study, 4 patients were on vasodilators and 3 on diuretics. Patients underwent right heart catheterization with continuous oxygen consumption determination at rest and during supine bicycle exercise after postoperative recovery and rehabilitation ≈1 to 2 months after LVAD placement. Doppler echocardiography was recorded at rest and at peak exercise. Nine patients also underwent graded upright treadmill testing with oxygen consumption measurements. One subject underwent successful cardiac transplantation before treadmill testing. Exercise was terminated when patients were limited by symptoms of fatigue or shortness of breath. All patients received stable daily doses of medication at the time of protocol testing.
Hemodynamics During Supine Bicycle Exercise
An 8F thermodilution Swan-Ganz catheter with pulmonary artery tip oximeter (Edwards) was placed through the right internal jugular vein. The tip oximeter was calibrated in vivo to a pulmonary artery blood sample oxygen content value. A left radial arterial line was placed percutaneously for continuous arterial pressure measurement. Continuous oxygen consumption and respiratory exchange ratio (R, V̇co2/V̇o2) were measured with a ventilatory analyzer (Sensormedics System 4400). Arterial oxygen saturation was monitored with a finger probe (Oxisensor 24, Nellcor). Supervised supine bicycle exercise was begun at 50-W power output with a protocol increasing the workload 25 W every 3 minutes until fatigue. Measurements from the control unit of the assist device were recorded throughout exercise, including stroke volume, rate, and calculated LVAD output. Systemic blood flow was determined by the Fick method, which was calculated as oxygen consumption divided by the difference between systemic arterial and pulmonary arterial oxygen contents. To determine whether the increase in oxygen consumption with exercise was associated with an appropriate increase in cardiac output, exercise factor and exercise index were measured. An increase of >600 mL/min in cardiac output per 100 mL/min increase in oxygen consumption and an exercise index of >0.8 would indicate normal responses.11
Doppler echocardiography was performed during supine bicycle exercise with an Acuson 128XP/5 system and a 6-MHz transducer. Optimal acoustic windows to view both the RV and LV were obtained from a modified apical four-chamber view in 6 patients and a parasternal long-axis view in 2. Adequate echocardiographic views could not be obtained in 2 patients. Sequential 5-beat averages (5 of 8 patients) or representative single beats (3 of 8 patients) were used for echocardiographic geometric measurements.
Data were analyzed for RV end-diastolic dimension, RV area, RVSF, and LV end-diastolic dimension. RVSF was calculated by the formula RVSF=(4×π×area)/(perimeter)2, as previously described,12 where RVSF=1 for a perfect circle.
Treadmill Exercise Measurements
Initially, graded upright treadmill exercise was performed according to a modified Bruce protocol with 3-minute stages (n=3), but to allow a greater range of functional capacity testing, a modified Naughton protocol with 2-minute stages (n=6) was subsequently used. Oxygen consumption, respiratory exchange ratio, and LVAD performance were measured.
All statistical analyses were conducted with SPSS software (SPSS Inc, SPSS for Windows, Version 6.1.2). Paired comparisons for nonparametric data were made with the Wilcoxon signed-rank test. Unpaired comparisons involving parametric data were determined with Student’s t test. Correlation coefficients were derived from linear regression. All data are presented as mean±SD.
Hemodynamic and Oxygen Consumption Measurements During Supine Bicycle Exercise
Patients performed supine exercise tests 46±25 days after LVAD implantation (Table⇓). Maximum power output was 108±29 W. Average maximal exercise duration was 10.4±2.6 minutes. Heart rate (87±12 to 117±14 bpm, P<.05) and LVAD rate (82±18 to 107±21 strokes/min, P<.01) increased in parallel during exercise (Fig 1⇓). Mean arterial pressure increased from 94±13 to 103±13 mm Hg (P<.01) as Fick systemic blood flow rose from 5.0±1.2 to 7.8±2.5 L/min (P<.01). RV stroke volume derived from Fick systemic blood flow and native heart rate was 58±12 mL at baseline and 67±19 mL with exercise (P=.09). Pulmonary artery pressure increased with exercise from 17±3 to 31±7 mm Hg (P<.001). Right atrial pressure increased from 6±4 to 12±5 mm Hg (P<.001). Pulmonary capillary wedge pressure increased from 5±3 to 13±8 mm Hg (P<.001). Systemic vascular resistance fell from 1478±369 to 899±357 dyne·s·cm−5 (P<.01).
V̇o2 increased from 3.2±0.6 to 8.2±1.7 mL O2·kg−1·min−1 (P<.001). Despite an average hemoglobin of 10.5±1.2, no patient had an elevated plasma-free hemoglobin level to indicate intravascular hemolysis. Pulmonary artery oxygen saturation decreased progressively throughout exercise as arterial oxygen saturation remained unchanged. An exercise factor of 6.6±4.0 mL/min CO per mL/min V̇o2 and an exercise index of 0.89±0.17 resulted.
Changes in LVAD Function During Supine Bicycle Exercise
LVAD output rose from 5.4±0.9 to 7.0±1.4 L/min (P<.001). LVAD stroke volume remained essentially constant, with only small variations during exercise (65±6 versus 66±7 mL).
Initially, Fick systemic blood flow approximated LVAD output at rest (5.0±1.2 and 5.4±0.9 L/min, respectively, Fig 2⇓). From rest to peak exercise, however, Fick systemic blood flow increased significantly more than LVAD output (2.8±1.9 versus 1.6±1.1 L/min, P<.05). By Doppler echocardiography, minimal or no aortic valve opening or flow was observed at rest but both were frequently noted during exercise.
LV and RV Echocardiography Changes During Supine Bicycle Exercise
LV end-diastolic dimension increased significantly from 3.9±1.3 cm at rest to 4.8±1.6 cm with exercise (P<.05; Fig 3⇓, top). RV end-diastolic dimension, however, decreased from 3.2±1.0 cm at rest to 2.3±0.9 cm with exercise (P<.01; Fig 3⇓, bottom). RV area also decreased significantly, from 18.8±6.3 to 15.3±5.4 cm2 (P<.05). RV shape factor showed a trend to decrease with exercise, from 0.69±0.10 at baseline to 0.64±0.03 (P=.15). There was no significant correlation between decreases in RV dimension and increases in LV dimension. In addition, there were no differences in changes in pulmonary wedge pressure between patients with small (n=4) versus large (n=4) increases in LV dimension with exercise.
Hemodynamic and Oxygen Consumption Measurements During Upright Treadmill Exercise
Patients performed upright treadmill exercise testing 50±23 days after LVAD implantation (Table⇑). Maximum workload achieved ranged from 2.0 mph at 3.5% grade to 3.0 mph at 7.5% grade. Heart rate increased from 95±7 bpm at rest to 147±17 bpm at peak exercise (P<.001). LVAD rate increased from a baseline value of 72±13 bpm to a peak exercise value of 111±21 bpm (P<.001). There was no significant change in systolic or diastolic blood pressure during treadmill exercise. LVAD output also increased with exercise from 4.4±0.8 to 7.6±1.9 L/min (P<.001), with minimal change in LVAD stroke volume. Oxygen consumption increased from 4.8±1.7 to 14.1±2.9 mL O2·kg−1·min−1 with treadmill exercise (P<.001) (relative workload, 4.0±0.8 METs; range, 2.8 to 5.3 METs).
Comparative Hemodynamic and Oxygen Consumption Measurements Between Supine Bicycle and Upright Treadmill Exercise
Although baseline heart rates were not significantly different between bicycle and treadmill exercise, the peak exercise heart rate was significantly greater for treadmill than for supine bicycle (147±17 and 117±14 bpm, respectively, P<.01). Systolic blood pressure was greater at rest and at maximal exercise during supine bicycle exercise (P<.01). Assist device output at baseline was lower during baseline treadmill evaluation than during supine bicycle evaluation (4.4±0.8 versus 5.4±0.9 L/min, P<.01) but was similar at maximal exercise. Oxygen consumption was greater during treadmill testing both at rest (P<.01) and at maximal exertion (P<.01). There was a trend toward higher respiratory exchange ratio, R, values during baseline treadmill evaluation than during baseline bicycle evaluation (P=.06). This difference became significant at peak exercise, measuring 1.33±0.23 on the treadmill versus 1.10±0.12 on the bicycle (P<.01).
At rest, the LV and LVAD act as series pumps with minimal or no blood flow through the native aortic valve (Fig 4⇓). With exercise, increases in venous return and LV filling pressure and dimension are associated with decreases in RV end-diastolic dimension. If the LV can eject with increasing left heart filling pressure, then it functions in parallel with the LVAD, contributing to systemic circulation directly through ejection across the aortic valve and indirectly through LVAD filling.
With initiation of active LVAD pumping support, a septal shift toward the LV and an increase in RV chamber size have been reported.11 13 14 This study demonstrated that with exercise, RV size decreased despite an increase in RV filling pressure, implying a decrease in diastolic ventricular chamber compliance with exercise. The factors that determine diastolic ventricular interdependence are complex and include septal shift, pericardial or mediastinal constraint, and initial LV and RV volumes mediated by ventricular circumferential fibers.15 The finding that the right atrial and pulmonary capillary wedge pressures were similar at rest and increased by similar amounts would suggest the presence of pericardial or mediastinal constraint.16 17 A trend toward decreased circularity of the RV shape factor is consistent with a return of septal position from left to right. Despite the decrease in RV end-diastolic dimension, RV stroke volume still showed a trend toward being higher with exercise compared with baseline.
Thus, after the immediate postoperative phase, when acute RV failure is a major cause of morbidity and mortality, increases in systemic flow with exercise appear to be limited by the LV-LVAD complex and not by intrinsic RV function. Systemic cardiac output is only one of the mechanisms limiting exercise performance in patients with heart failure, in addition to other mechanisms of muscle fatigue or respiratory insufficiency.15 A normal increase in cardiac output with exercise, however, is an important factor that can contribute to patient rehabilitation. The finding that the average mixed venous saturation decreased to 38±10% with supine exercise suggests that systemic delivery of oxygen was a limiting factor in some but not all of the patients.18 Circulating neurohormone levels were not directly measured in this study. Future LVAD designs might attempt to achieve even higher maximum outputs at physiological left heart filling pressures to permit greater exercise circulatory reserve and capacity.
Maximum V̇o2 is a well-characterized indicator of functional status and marker for prognosis in patients with advanced heart failure.19 20 It may be more reproducible than exercise duration, distance, or other assessments of functional reserve.21 As in normal subjects and patients with coronary artery disease, we found that higher oxygen consumptions were achieved during upright than with supine exercise.22 23 Higher treadmill respiratory exchange ratio values imply a consistent occurrence of an anaerobic state during upright exercise but not necessarily with supine bicycle exercise. Contrary to normal subjects22 and orthotopic cardiac transplantation patients,24 there was no increase in systolic or pulse pressure with upright exercise, in part reflecting the constant stroke volume of the LVAD output. In these patients, 50±23 days after LVAD implantation, peak oxygen consumption with upright treadmill exercise of 14.1±2.9 mL O2·kg−1·min−1 was below that of normal subjects20 but approached those found in early studies of patients at 1 year after transplantation24 and in more recent studies of patients at 3 months after transplantation (17.5±3.6 mL O2·kg−1·min−1).25
A limitation of this study is that all patients were approved for and awaiting cardiac transplantation, and as a result, one cannot extrapolate the results of this study to the larger population of patients with advanced heart failure. Whether these exercise findings would have improved over a longer postoperative time is unknown, but it is likely.26 Placement of the LVAD in patients before a manifest premoribund state might show an improved functional outcome beyond that observed here. No comparison was possible between patients’ exercise capacity on maximal medical therapy before LVAD implantation due to their refractory heart failure state and associated inability to undergo exercise of any kind.
The present study, performed during the recuperation phase of bridge to cardiac transplantation, indicates that exercise during long-term implantation of an LVAD is safe and is not limited by right heart decompensation. Exercise capacity, however, is just one determinant of quality of life after any treatment of advanced heart failure. Other recent studies have found marked improvements in renal and hepatic function,4 27 enhanced RV function,3 28 29 reversal of severe ventricular dilatation,30 and improvement of general physiological status4 28 after LVAD placement.
Given a shortage of donor heart organs, selected patients with contraindications to cardiac transplantation with end-stage heart failure might be suitable candidates for permanent implantation of an LVAD. Although presently available devices are not totally implantable, permanent implantation would avoid limitations of cardiac transplantation, including the need for long-term immunosuppressive therapy and repeated invasive biopsy. This preliminary study provides a rationale for a comparison of functional capacity after LVAD implantation with that after subsequent cardiac transplantation. This is the primary focus of the ongoing EVADE trial. A direct comparison of LVAD implantation with medical therapy for patients with advanced heart failure will have to await the results of other ongoing randomized trials comparing these two divergent therapeutic strategies.
Selected Abbreviations and Acronyms
|EVADE||=||Experience with Left Ventricular Assist Device and Exercise trial|
|LV||=||left ventricle, left ventricular|
|LVAD||=||left ventricular assist device|
|RV||=||right ventricle, right ventricular|
|RVSF||=||RV shape factor|
This study was supported in part by the Rose Azus Cardiac Research Education Fund, Sharp Hospital Foundation, San Diego, Calif, the San Diego Foundation for Cardiovascular Research and Education, and the Barona Casino Research and Education Fund. We gratefully acknowledge Mary Lou McNeill for her expertise in the analysis of echocardiograms. We also appreciate the assistance of Ian Gocka for statistical analyses.
Reprint requests to Brian Jaski, MD, San Diego Cardiac Center, 8010 Frost St, Suite 200, San Diego, CA 92123.
Presented in part at the 68th Scientific Sessions of the American Heart Association, Anaheim, Calif, November 13-16, 1995, and published in abstract form (Circulation. 1995;92[suppl I]:I-49).
- Received September 4, 1996.
- Revision received November 4, 1996.
- Accepted November 23, 1996.
- Copyright © 1997 by American Heart Association
Medical Devices, Diagnostics and Instrumentation Reports. Chevy Chase, MD: F-D-C Reports, Inc; October 10, 1994.
Oz MC, Goldstein DJ, Pepino P, Weinberg M, Thompson SM, Catanese KA, Vargo RL, McCarthy PM, Rose EA, Levin HR. Screening scale predicts patients successfully receiving long-term implantable left ventricular assist devices. Circulation. 1995;92(suppl II):II-169-II-173.
McCarthy PM. HeartMate implantable left ventricular assist device: bridge to transplantation and future applications. Ann Thorac Surg. 1995;59:S46-S51.
Frazier OH. First use of an untethered, vented electric left ventricular assist device for long-term support. Circulation. 1994;89:2908-2914.
Grossman W. Cardiac Catheterization and Angiography. 3rd ed. Philadelphia, Pa: Lea & Febiger; 1986:253.
Wilson JR, Mancini DM. Factors contributing to the exercise limitation of heart failure. J Am Coll Cardiol. 1993;22(4 suppl A):93A-98A.
Janicki JS. Influence of the pericardium and ventricular interdependence on left ventricular diastolic and systolic function in patients with heart failure. Circulation. 1990;81(suppl III):III-15-III-20.
American College of Sports Medicine. Guidelines for Exercise Testing and Prescription. 4th ed. Philadelphia, Pa: Lea & Febiger; 1991:87-88.
Weber KT, Kinasewitz GT, Janicki JS, Fishman AP. Oxygen utili-zation and ventilation during exercise in patients with chronic cardiac failure. Circulation. 1982;65:1213-1223.
Niederberger M, Bruce RA, Kusumi F, Whitkanack S. Disparities in ventilatory and circulatory responses to bicycle and treadmill exercise. Br Heart J. 1974;36:377-382.
Ehrman J, Keteyian S, Fedel F, Rhoads K, Levine TB, Shepard R. Cardiovascular responses of heart transplant recipients to graded exercise testing. J Appl Physiol. 1992;73:260-264.
Foray AT, Williams DL, Oz MC, Mancini DM. Submaximal exercise capacity in patients with left ventricular assist devices improves over time. ISHLT Convention, 1996. J Heart Lung Transplant. 1996;15:586. Abstract.
Levin HR, Oz MC, Chen JM, Packer M, Rose EA, Burkhoff D. Reversal of chronic ventricular dilation in patients with end-stage cardiomyopathy by prolonged mechanical unloading. Circulation. 1995;91:2717-2720.