Low Incidence of Myocardial Recovery After Left Ventricular Assist Device Implantation in Patients With Chronic Heart Failure
Background—Mechanical, histological, and biochemical improvement has been described in patients after left ventricular assist device (LVAD) support. Explantation of the LVADs without heart transplantation has been described in selected patients who received this therapy as a bridge to transplantation.
Methods and Results—A retrospective review of patients receiving a mechanical bridge to transplantation at Columbia Presbyterian Hospital after July 21, 1991, was performed to determine the incidence of patients in whom the device was successfully explanted. From August 1, 1996, to February 1, 1998, we prospectively attempted to identify potential explant candidates by the use of exercise testing. During this time, we recruited 39 consecutive patients after insertion of the Thermo Cardiosystems vented electric device to participate in the following study. Approximately 3 months after device implantation, a maximal exercise test with hemodynamic monitoring and respiratory gas analysis was performed with the LVAD in the automated mode. The electric device was interfaced with a pneumatic console such that the rate could be decreased to 20 cycles/min. Hemodynamic measurements were recorded as the device rate was decreased. A repeat exercise test was then performed if the patient remained hemodynamically stable. A retrospective chart review of 111 LVAD recipients at our institution identified only 5 successful explant patients. Eighteen of the 39 patients were studied. Fifteen patients exercised with maximal device support. At peak exercise, V̇o2 averaged 14.5±3.6 mL · kg−1 · min−1; LVAD flow, 8.0±1.3 L/min; Fick cardiac output, 11.4±3.3 L/min; and pulmonary capillary wedge pressure, 13±4 mm Hg. Seven patients remained normotensive and could exercise at a fixed rate of 20 cycles/min. In these patients, peak V̇o2 declined from 17.3±3.9 to 13.0±6.1 mL · kg−1 · min−1. In one of these patients, the device was explanted.
Conclusions—Significant myocardial recovery after LVAD therapy in patients with end-stage congestive heart failure occurs in a small percentage of patients. Most of these patients have dilated cardiomyopathy. Exercise testing may be a useful modality to identify those patients in whom the device can be explanted.
Improvement of indices of cardiac function has been observed in patients on mechanical left ventricular assist devices (LVADs) awaiting cardiac transplantation.1 2 3 4 LVADs provide profound LV volume and pressure unloading while simultaneously restoring total systemic blood flow in end-stage congestive heart failure (CHF) patients. Mechanical, histological, and metabolic improvement after LVAD support has been demonstrated in some patients. Levin et al5 demonstrated reverse remodeling with a decreased LV mass and leftward shifts of the end-diastolic pressure-volume relationship in LVAD-supported patients. A reduction of cellular edema,6 more efficient myocardial mitochondria function,7 improved myocyte contraction,8 decreased apoptosis,9 and reversal of neurohormonal stimulation10 have also been described. Recently, Muller et al4 reported the explantation of LVADs, ie, the removal of mechanical support in lieu of transplantation in 5 of 17 patients with dilated cardiomyopathy. A retrospective review of our experience with 111 Thermo Cardiosystems (TCI) devices implanted as a bridge to cardiac transplantation was performed to identify explant patients. Only 5 patients were identified.
From August 1, 1996, to February 1, 1998, we attempted to prospectively identify patients with sufficient myocardial recovery after LVAD insertion by the use of exercise testing with respiratory gas analysis, echocardiographic, and hemodynamic measurements. Measurements were recorded in the fully supported state and as patients were withdrawn from device support. We reasoned that hemodynamic parameters measured at rest and during stress with the ventricle reloaded could potentially quantify myocardial recovery and guide therapy, ie, whether to proceed to transplantation or to explant the device.
After 1991, 111 patients underwent placement of an LVAD as a bridge to transplantation at Columbia Presbyterian Medical Center. All patients had New York Heart Association class IV CHF, a cardiac index <2 L · min−1 · m−2, pulmonary capillary wedge pressure >25 mm Hg, systolic blood pressure <80 mm Hg, and LV ejection fraction (LVEF) <20%. All patients were supported with multiple inotropic agents and/or intra-aortic balloon counterpulsation. The average age of these patients was 50±13 years; 89 were men and 22 were women. The cause of CHF was coronary artery disease in 60 patients and dilated cardiomyopathy in 51 patients. Duration of implant averaged 100±97 days.
A retrospective chart review was performed to identify all explant patients. Clinical outcome and assessment of LV function were retrieved.
From August 1, 1996, to February 1, 1998, 39 patients had a TCI vented electric device implanted. Eighteen patients were studied. Patients were excluded for the following reasons: perioperative mortality (n=9), awaiting study (n=2), refusal (n=4), extensive cerebrovascular accident (n=1), active infection (n=1), and transplanted before study scheduled (n=4).
The average age of the 13 men and 5 women studied was 47±13 years, 15 of the 18 patients were ambulatory, and preimplant LVEF averaged 19±5%. Twelve patients had dilated cardiomyopathy and 6 patients coronary artery disease. New York Heart Association CHF class was class I in 2 patients, class II in 9, class III in 5, and class IV in 2. Studies were performed an average of 86 days after implantation (range, 34 to 216 days). Medical therapy included digoxin in 4 patients, diuretics in 11, and ACE inhibitors in 5. One patient was on β-blockers.
Under local anesthesia, a Swan-Ganz catheter (Abbott) was inserted. In nonambulatory patients, resting hemodynamic, echocardiographic, and LVAD sensor measurements were recorded with the patient in the auto mode of the device (ie, full hemodynamic support) and then with the device at the lowest fixed rate, ie, 50 cycles/min. The patients were then heparinized. The electrical device was interfaced with the pneumatic stroke volume limiter so that the device rate could be reduced to 20 cycles/min. Downtitration of the LVAD rate by 10 cycles/min was performed with hemodynamic and echocardiographic measurements every 10 minutes until the patient developed symptoms or a rate of 20 cycles/min was achieved. Because the stroke volume of the device is ≈80 mL, a fixed rate of 20 cycles/min generates ≈1.6 L/min flow.
In ambulatory patients, after insertion of a Swan-Ganz catheter, hemodynamic, echocardiographic, metabolic, LVAD sensor, and exercise measurements were made during exercise in the automated mode of the device, during downtitration of the device, and during exercise at a fixed rate of 20 cycles/min. Patients performed exercise in the fasting state and on their usual medical regimen. The subject was seated on a Monark ergometer and connected to a Medical Graphics 2001 Metabolic Cart with a disposable Pneumotach. The transducer was positioned at the level of the fourth intercostal space in the midaxillary line. Resting pulmonary artery, pulmonary wedge, and right atrial pressures and respiratory gases were measured, and blood samples were obtained. Arterial blood pressure was assessed by cuff sphygmomanometer. Baseline echocardiogram was recorded from the parasternal position. After a 3-minute rest period, the patient began bicycle exercise at 0 W. Work was increased by 25-W increments every 3 minutes until exhaustion. Respiratory gas and heart rate measurements were made continuously. Blood sampling, arterial blood pressure, and LVAD sensor measurements were performed during the last minute of each exercise stage. Echocardiograms were recorded in the first minute of recovery.
The arteriovenous oxygen difference was calculated as (arterial−venous O2 saturation)×(1.34 mL O2/g hemoglobin)×(hemoglo-bin concentration). Cardiac output was calculated by use of the Fick equation.
Two-dimensional Doppler echocardiography was performed with a Hewlett-Packard Sonos 500 imaging system and recorded with simultaneous ECG. The data were saved on 1/2-in VHS tape for subsequent analysis. Parasternal long- and short-axis views with color Doppler were recorded. End-diastolic and end-systolic diameters of the left ventricle were measured from the parasternal short-axis views. Average values of 3 successive beats were used for statistical analysis. Fractional shortening was calculated as the difference between end-diastolic and end-systolic diameters divided by the end-diastolic diameter. LV mass was calculated from the formula of Devereux.11 Aortic valve opening in the parasternal long-axis view was quantified as a percentage at rest and at peak exercise of the number of beats with an opened aortic valve over the 10 consecutive cardiac beats examined. Color Doppler recordings for determination of device inflow valve insufficiency were also recorded.
Statistical analysis was performed with paired and nonpaired t tests as appropriate.
Clinical Outcome of Explant Patients
Only 5 patients at our institution have had the devices successfully explanted (Table 1⇓, Figure 1⇓). Patient 1 was a 18-year-old man with a cardiomyopathy who was supported with the device for 186 days. Explantation rather than transplantation was performed. His LVEF decreased rapidly from 72% to 20%, and he died of CHF 3 months after explantation. Patient 2 was a 47-year-old man with coronary artery disease who underwent coronary artery bypass graft surgery, which was complicated by heart failure and ventricular tachycardia. He underwent LVAD placement and was supported for 101 days before explantation due to device infection. He remained in class II CHF until he died suddenly. There was minimal change in his LVEF over these 2 years. Patient 3 is a 30-year-old man with a cardiomyopathy complicated by atrial tachyarrhythmias. Explantation rather than transplantation was performed. The duration of LVAD support was 66 days. At 22 months after explantation, he was readmitted with severe CHF requiring inotropic support. His LVEF had slowly deteriorated from 45% to 28% over 18 months. He had a cardiac arrest and required reimplantation of the LVAD 684 days after explantation. The patient sustained anoxic brain damage after arrest. Sixty days after the second device implantation, the LVAD failed. Unsupported LVEF had again risen to 50%.
Patient 4 is a 47-year-old man with cardiomyopathy who was supported for 58 days before explantation for device infection. He remains in class I CHF 15 months after explantation. His LVEF has remained stable. Patient 5’s device was explanted 380 days after device insertion because of infection. Despite treatment with diuretics, digoxin, ACE inhibition, and carvedilol, she experienced a rapid decline in LV function and required reimplantation of the device at 170 days.
Exercise Studies in the Automated Mode
Rest and peak exercise parameters are shown in Table 2⇓. Peak V̇o2 averaged 14.6 mL · kg−1 · min−1 in the automated mode of the device. At rest, Fick estimates of resting cardiac output and LVAD sensor measurements were not significantly different. Only 3 patients had intermittent aortic valve opening. At peak exercise, Fick cardiac outputs were significantly greater than LVAD sensor measurements, with aortic valve opening occurring in at least 7 patients.
Rest and peak exercise echocardiograms demonstrated that LV cavity size was normal in all but 2 patients. Both of these patients had Doppler signals suggesting inflow valve insufficiency, with LV end-diastolic diameter at rest of 5.9 and 7 cm, respectively. LV mass measurements averaged 186±97 g, reduced from the preimplant estimates of LV mass, which was 332±81 g (P<0.001). Both of these estimates were significantly below the actual explanted heart weights of 400±55 g.
Hemodynamic and Echocardiographic Measurements With Downtitration of the Device
Hemodynamic measurements recorded during LVAD downtitration are shown in Table 3⇓. Most patients could not tolerate downtitration of the device. With downtitration, heart rate increased, mean arterial blood pressure fell from 91 to 70 mm Hg, pulmonary artery capillary wedge pressure rose, and cardiac output fell significantly. LV diastolic dimension increased. Aortic valve opening was now present in almost all patients. Figure 2⇓ shows parasternal short-axis views at rest of end systole and end diastole with maximal and minimal LVAD support for 1 patient.
Dizziness and/or diaphoresis occurred in 11 of 18 subjects. Only 7 patients remained hemodynamically stable at a fixed rate of 20 cycles/min.
Exercise at a Fixed Rate of 20 Cycles/min
The clinical characteristics of the 7 patients able to exercise at a fixed rate of 20 cycles/min are shown in Table 4⇓. Six of the 7 patients had cardiomyopathy. The mean duration of implant was 4 months. The exercise response in the automated and fixed modes for these 7 patients is shown in Table 5⇓. Because of technical problems, we were unable to measure Fick cardiac outputs in patient 1. In patient 2, echocardiographic images were suboptimal.
In these 7 patients, peak V̇o2 decreased from 17.3 to 13 mL · kg−1 · min−1. For patient 2, exercise in the automated mode showed comparable peak LVAD sensor and Fick cardiac output. During exercise in the fixed rate, Fick cardiac output was severely reduced. This suggested that the device could not be explanted in this patient. In contrast, patients 3 and 4 had higher Fick cardiac outputs than were recorded by the LVAD sensor during exercise in the automated mode. In the fixed mode, peak Fick cardiac outputs were maintained, peak V̇o2 approximated 20 mL · kg−1 · min−1, and aortic valve opening was present. However, patient 3 showed LV dilatation with decreased fractional shortening, whereas patient 4 had a normal-size left ventricle with a normal fractional shortening. This suggested that patient 4 had complete recovery.
At the time of transplantation, with the device off, patients 1, 2, and 6 had poor LV function and proceeded to transplantation. In patients 3 and 7, LV function could not be evaluated because of surgical problems. Patient 4 underwent successful explantation. Patient 5 had LV function in the range of 35%. Because right nuclear ejection fraction measured by first-pass radionuclide examination was 14% and peak V̇o2 was only 9.9 mL · kg−1 · min−1 with the device at a fixed rate, he proceeded to transplantation.
Several studies suggest that the profound ventricular pressure and volume unloading provided by LVAD support leads to reverse remodeling on genetic, biochemical, histological, and functional levels.1 2 3 4 5 6 7 8 9 10 12 Normalization of the LV pressure-volume relationship with regression of LV hypertrophy has been reported.5 A decrease in fibrosis4 and reduction in myocardial matrix metalloproteinase expression may play a role in reverse remodeling.12 Decrease in stretch, profound volume/pressure unloading, increased myocardial perfusion, normalization of neurohormones, and reduction in cytokine release are all potential mechanisms for the reversal of the remodeling process.
In our present study, retrospective analysis identified few patients in whom the devices could be successfully explanted. Similarly, prospective analysis using hemodynamic, echocardiographic, and exercise measurements at rest and during ventricular loading revealed few patients with adequate myocardial recovery to permit explantation. Despite the small number of patients either retrospectively or prospectively identified as explant candidates, the results of this study are strikingly important. All of these patients had presumably irreversible end-stage heart failure, yet these patients demonstrated that their LV function could be dramatically modified and almost returned to normal for short periods.
A recent report from Germany by Muller et al4 identified 5 of 17 patients with dilated cardiomyopathy supported by an LVAD for >160 days in whom the devices were successfully explanted and who maintained normal cardiac function. Our report is similar to that of Muller et al, because the majority of our explant candidates also had dilated cardiomyopathy, and in our prospective analysis, most patients able to support themselves with the ventricle reloaded had dilated cardiomyopathies. In contrast to the German study, in which 29% of the dilated cardiomyopathy patients were explant candidates, the frequency of this phenomenon in our patient population was much lower, including only 5% of all our patients and 8% of the patients with dilated cardiomyopathies. All our patients were on multiple inotropic agents before device insertion; thus, the ability of patients to exercise at a low fixed rate demonstrates intermediate degrees of myocardial recovery. The frequency of partial recovery increases to 50% (6 of 12) in patients with dilated cardiomyopathy and 17% (1 of 6) in patients with coronary artery disease.
In the German report, all LVAD patients received treatment with digoxin, diuretics, ACE inhibitors, β-blockers, coenzyme Q10, and other nutritional supplements. Our study differs from that of Muller et al4 in that our LVAD patients did not receive the benefit of maximal medical therapy. In our study, <20% of patients received ACE inhibitors or β-blockers. Initiation of aggressive heart failure therapy might have improved our explantation rates.
Our study also differed from that of Muller et al4 in that LVAD support duration was shorter and a rapid rather than prolonged weaning was used to assess recovery. In our study, duration of device support was determined by the waiting time for transplantation, which averaged ≈100 days. Our rapid weaning may not have allowed the neurohormonal system and/or an atrophied muscle sufficient time to readjust to the reduction in cardiac output or to the increased workload. This is unlikely because, although hypertrophy regresses in device-supported patients, significant hypertrophy persists, as evidenced by the increased weights of the explanted hearts in our sample.
Another important difference between the present report and that of Muller et al4 is that normalization of LV function in our explant patients has not been sustained despite aggressive medical therapy. Three of our 5 patients have experienced redilation of the LV, with a decrease in LVEF and recurrence of symptoms after device removal.
A possible explanation for device removal is spontaneous recovery, which has been described in up to 37% of recent-onset cardiomyopathy.13 In a study of 49 patients with new-onset CHF referred for transplant evaluation, 27% showed improvement, with LVEF increasing from 22% to 49%.14 In the NIH trial using immunosuppressive therapy in patients with biopsy-proven myocarditis, LVEF improved significantly.15 Shorter duration of disease and higher LVEF at presentation are predictive of recovery. It is unlikely that our observations are due to spontaneous recovery. None of our explant patients had new-onset CHF or acute myocarditis. Moreover, 3 of 5 patients have developed recurrence of severe heart failure after device explantation. In 1 patient, reinsertion after LVAD again resulted in normalization of EF, thus demonstrating the reproducibility of the phenomenon.
The small percentage of explant candidates may be due to device malfunction, ie, regurgitation or obstruction at the inflow valve, negating the extreme ventricular volume or pressure unloading. Inflow valve regurgitation interferes with reverse remodeling.16 Again, this explanation is unlikely, because the abnormal Doppler flow pattern at the inflow valve characteristic of this malfunction was observed in only 2 patients.
Exercise Performance of LVAD Patients
Measurements recorded during exercise in the automated mode of the device are comparable to previous reports of exercise performance in LVAD patients. Average reported peak V̇o2 during upright exercise of patients with the Novacor or TCI devices ranges from 14 to 17 mL · kg−1 · min−1.17 18 19 20 Recently, we compared the exercise hemodynamic response of LVAD patients with that of patients with severe heart failure awaiting cardiac transplantation.21 The reduction in pulmonary pressures and peak exercise performance is similar to that in this prior report.21
Clinical Parameters to Assess Recovery
Our findings suggest that the ability to exercise at a low fixed rate to a V̇o2 >20 mL · kg−1 · min−1 and/or a peak cardiac output >10 L/min provides sufficient cardiovascular reserve to tolerate explantation. Echocardiographic parameters such as consistent aortic valve opening, normal shortening fraction, and absence of marked ventricular dilation may also help to distinguish explant candidates. Additional studies are needed to create an accurate profile to identify patients with sufficient recovery.
This study is limited by the small sample size with a low enrollment rate due to complications related to the device and/or unpredictable timing of cardiac transplantation. Patients were frequently reluctant to proceed with a test that would decrease device support.
A major limitation of this study is that it is descriptive. No attempt was made to identify the mechanism for recovery. Investigators at our institution are currently studying the molecular effects of LVAD support.
Inability to completely turn off the device limited our ability to assess recovery of LVAD function. Although we elected to use exercise as a stress to evaluate LV recovery with the device set at a low fixed rate, other interventions, such as volume loading or pharmacological challenges, might also have provided clinically relevant information.
The observations described in this report are important in that they illustrate that patients at the end stages of heart failure still possess myocardium that can be restored to normal or near-normal function. Restoration of function and structure in what had been believed to be an irreversible state demonstrates the plasticity of the myocardium. The low incidence of myocardial recovery described may be improved with medical and/or surgical adjuvant therapies. Patients with intermediate degrees of recovery could be identified and targeted for more aggressive CHF therapy and restudied. In addition, alternative surgical approaches, such as explantation plus mitral valve repair and/or ventricular reduction surgery, could be considered.
In the future, heart failure could conceivably be reversed rather than just managed. Implantation of an LVAD is not a practical solution for large numbers of patients because of the high cost and invasiveness of this procedure. Nonetheless, the clinical application of these devices may be able to provide insight into the mechanisms that signal reverse remodeling and thus open the way to cure.
This study was supported by a grant from the Division of Research Resources, General Clinical Research Centers Program, NIH 5-M01-RR-00645.
Reprint requests to Donna M. Mancini, MD, Division of Circulatory Physiology, Department of Medicine, Columbia Presbyterian Medical Center, 622 W 168th St, New York, NY 10032.
- Received May 27, 1998.
- Revision received July 29, 1998.
- Accepted August 13, 1998.
- Copyright © 1998 by American Heart Association
Muller J, Wallukat G, Weng Y, Dandel M, Spiegelsberger S, Semrau S, Brandes K, Theodoridis V, Loebe M, Meyer R, Hetzer R. Weaning from mechanical support in patients with dilated cardiomyopathy. Circulation. 1997;96:542–549.
Levin H, Oz M, Chen J, Packer M, Rose E, Burkhoff D. Reversal of chronic ventricular dilation in patients with end-stage cardiomyopathy by prolonged mechanical unloading. Circulation. 1995;91:2717–2720.
Dipla K, Mattiello JA, Jeevanandam V, Paolone AM, Margulies KB, Houser SR. Improved myocyte function following mechanical support of the left ventricle in humans with severe heart failure. Circulation. 1997;96(suppl I):I-298. Abstract.
Belland S, Grunstein R, Jeevanandam V, Eisen H. The effect of sustained mechanical support with left ventricular assist devices on myocardial apoptosis in patients with severe dilated cardiomyopathy. J Heart Lung Transplant. 1998;17:83A. Abstract.
Devereux R, Reichek N. Echocardiographic determination of left ventricular mass in man: anatomic validation of the method. Circulation. 1977;55:613–618.
Belland S, Jeevanandam V, Eisen H. Reduced myocardial matrix metalloproteinase expression as a result of sustained mechanical support with left ventricular assist devices in patients with severe dilated cardiomyopathy. J Heart Lung Transplant. 1998;17:84A. Abstract.
Moazami N, Argenziano M, Kohmoto T, Levin H, Rose E, Oz M, Burkhoff D. Inflow valve regurgitation interferes with reverse remodeling during LVAD support. Circulation. 1996;94(suppl I):I-293. Abstract.
Murali S, Uretsky B, Estrada-Quintero T, Tokarczyk T, Cannon Y, Betchart A, Kormos K, Griffith B. Metabolic and ventilatory responses to exercise in heart failure patients on a left ventricular assist system support. Circulation. 1991;84(suppl II):II-354. Abstract.
Foray A, Williams D, Reemtsma K, Oz M, Mancini D. Assessment of submaximal exercise capacity in patients with left ventricular assist devices. Circulation. 1995;1996:94(suppl I):I-222–I-226.
Jaski BE, Kim J, Maly RS, Branch KR, Adamson R, Favrot LK, Smith SC Jr, Dembitsky WP. Effects of exercise during long-term support with a left ventricular assist device. Circulation. 1997;95:2401–2406.
Mancini D, Goldsmith R, Levin H, Beniaminovitz A, Rose E, Catanese K, Flannery M, Oz M. Comparison of exercise hemo- dynamic measurements in patients with severe heart failure and left ventricular assist devices. Circulation. 1998;98:1178–1183.Some patients who have received left ventricular assist device (LVAD) support as a bridge to transplant have developed myocardial recovery, which allowed explantation of the device. To determine the incidence of explantation in mechanically bridged patients, we performed a retrospective review of 111 LVAD patients. Prospective identification of explant candidates was attempted by exercise testing with hemodynamic, echocardiographic, and ventilatory measurements with maximal and minimal LVAD support. The device was explanted in only 5 patients. Seven of the 39 patients could exercise with the device at a low fixed rate. The device was explanted in 1 of these patients. Significant myocardial recovery after LVAD therapy occurs in only a small percentage of patients.