| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
J Thorac Cardiovasc Surg 1994;108:420-428
© 1994 Mosby, Inc.
CARDIAC AND PULMONARY REPLACEMENT |
Cleveland, Ohio
Supported by the National Institutes of Health contract N01-HV- 88103.
Received for publication Aug. 4, 1993. Accepted for publication March 16, 1994. Address for reprints: Patrick M. McCarthy, MD, Department of Thoracic and Cardiovascular Surgery, F25, The Cleveland Clinic Foundation, 9500 Euclid Ave., Cleveland, OH 44195.
Abstract
In vitro function of the Cleveland Clinic-Nimbus electrohydraulic total artificial heart met National Heart, Lung, and Blood Institute hemodynamic guidelines for such devices. In a series of in vivo experiments, we implanted the total artificial heart in eight calves (mean weight 87 kg), one for a short-term experiment and seven for long-term experiments. The mean blood flow during support was 7.7 ± 1.6 L/min with left atrial pressure 13 ± 6 mm Hg, right atrial pressure 13 ± 4 mm Hg, and aortic pressure 97 ± 9 mm hg. Maximum pump flow (9.6 L/min) occurred after 4 days of support as a result of the high resting cardiac output of the animals. A 10% to 15% right pump stroke-volume limit effectively balanced atrial pressures, and afterload insensitivity was confirmed by the in vivo studies. Calves tolerated treadmill exercise studies well, with an average duration of 22 minutes and an average top speed of 2.1 mph. The experiments were terminated after 1 day to 120 days of support (mean 32 days). Most experiments were terminated as a result of correctable mechanical problems. In a separate study of six adult human patients undergoing orthotopic cardiac transplantation, five showed an excellent fit for the Cleveland Clinic–Nimbus total artificial heart. Further studies using chest roentgenograms, chest measurements, and transesophageal echocardiography should help predict fit of the total artificial heart in potential candidates. Initial candidates for a "vented-electric" version of the Cleveland Clinic–Nimbus total artificial heart are patients for whom univentricular (left ventricular assist device) support is not appropriate, but who require mechanical support as a bridge to cardiac transplantation. (J THORAC CARDIOVASCSURG1994;108:420-8)
In vitro testing of the Cleveland Clinic–Nimbus (CC-N) total artificial heart (TAH) proved it to be hemodynamically functional and has been reported in the preceding article. 
1 Early clinical collaboration between the TAH research program and an increasingly active heart transplantation and ventricular assist device program has clarified clinical aspects of TAH use. This article reviews our initial in vivo TAH experience with calves, the clinical studies for determining proper fit of the CC-N TAH in human beings, the lessons from the clinical left ventricular assist device (LVAD) program that will help identify expected candidates for TAH insertion, and our preliminary plans for completing the TAH development project through to clinical introduction.
This article summarizes a series of studies performed under a development program supported by the National Institute of Health (Contract No. 1-HV-88103). Continued work on the implantable TAH was funded by the NIH beginning in October 1993 (as phase I of a device readiness program). In this article, therefore, we are describing our earliest efforts in development; many tasks remain under phases I and II of the device readiness program before clinical trials of the implantable TAH will begin (approximately in the year 2001).
IN VITO EXPERIMENTS—METHODS
From April 1991 until December 1992, eight calves (weight 80 to 96 kg; mean 87 kg) received the CC-N TAH. The first experiment was designed as a short-term (24-hour) experiment; the next seven were designed as long-term experiments.
The TAH was implanted through a right thoracotomy. The ventricles were transected and trimmed back to the atrioventricular groove, so that almost all atrial tissue and the atrial septum were preserved. The great vessels were transected above the valve commissures. Strips of Teflon felt were sewn around the anastomosis of the outflow grafts (26 mm low-porosity Dacron fabric) to the great vessels and around the Dacron velour cuffs to the atria (Fig. 1). Each anastomosis was tested for suture-line bleeding with a specially designed device. 2 A balloon was inflated in the atria, and pressurized saline solution (approximately 25 mm Hg pressure) was infused behind the suture line to reveal potential bleeding sites. The atrial cuffs and pulmonary artery graft included pressure monitoring lines that were also used to evacuate air. These lines were exteriorized, and right atrial pressure (RAP), pulmonary artery pressure (PAP), left atrial pressure (LAP), and aortic pressures (AoP) were continuously monitored and recorded.
|
In the first six calves the interventricular space was vented externally, but in the most recent two intrathoracic compliance chambers were included. Air was evacuated easily from ports in the right inflow and left outflow conduits as the pump was filled and cardiopulmonary bypass was weaned. No surgical bleeding occurred in any animal, and six of the seven animals on long-term support were extubated by the first postoperative day. After recovery from anesthesia and extubation, the animals were transferred to a chronic care facility for physiologic monitoring. No aspirin or other anticoagulants were given to the animals because the tissue valves and thromboresistant gelatin coating inside the pump reduce the thromboembolic risk.
A Gould ES 1000, 16-channel electrostatic recorder and monitor (Gould Inc., Valley View, Ohio) continuously recorded LAP, RAP, PAP, and AoP wave forms, right and left pump stroke volume, motor current and frequency, and spool valve position. Simultaneously, these variables were monitored on a 16-channel, personal computer–based data acquisition system, Metrobyte DASH 16 (Motrobyte Corp., Taunton, Mass.). Average values were stored on a hard disk at preset intervals ranging from every 15 seconds in short-term studies to every 20 minutes for continuous chronic care data monitoring.
Arterial and venous blood gases and venous lactate levels were analyzed during level treadmill exercise studies performed biweekly. Progressively increasing workloads were generated by increasing the treadmill speed 0.2 mph every 3 minutes and recording physiologic data at the end of each 3-minute period. 3
IN VITO EXPERIMENTS—RESULTS
General hemodynamic characteristics
In calf 1 the experiment was terminated during the first day (short-term experiment). In calf 2 ventilator failure and irreversible neurologic deficit developed early after the operation so that the study was terminated on day 5. Calf 3 died as the result of a device assembly error on the ninth postoperative day.
In calves 4 through 8, the TAH reached its maximum rate (160 beats/min) and output (almost 10 L/min) in the first few postoperative days (Fig. 2). This result is explainable by the rapid recovery of the calves from the operation and their relatively large resting cardiac output requirement. All five of these calves were exercising on the treadmill by postoperative day 3. This group of animals provided the following hemodynamic performance data.
|
1 In each experiment, the good anatomic fit of the device with unobstructed atrial filling was determined at operation and verified at autopsy. The pump outputs ranged from 4.5 to 9.0 L/min, at LAPs of 4.0 to 13.0 mm Hg and RAPs of 2.0 mm Hg to 13.0 mm Hg. The mean atrial and aortic pressures are listed in Table I. The mean pump flow for all experiments was 7.7 ± 1.6 L/min. As is typical of artificial heart implants in calves, 4 atrial pressures were near normal early in the experiments; however, pressures increased as the calves grew in body weight and the device reached maximum output.
|
|
1,5 The stroke volume limiter is a mechanical stop that limits the end of right blood pump filling. The right pump stroke volume limiter is effective when the pump is running at maximum cardiac output (as during exercise in the anticipated clinical situation) and limits the rise in LAP (which could cause pulmonary congestion in clinical use). 1 Data correlating LAP to right pump filling for calf 4 showed an increase in mean LAP because right pump filling exceeded 85% to 90%. Similar data from calf 5, in which the right pump filling was limited by 15%, indicated that limiting right pump filling by as much as 20% may further minimize LAP at maximum pump output. Data from calf 6 verified that limiting right pump filling by 20% was more effective in balancing the atrial pressure with an almost 1:1 LAP/RAP relationship after the device reached maximum output. For this blood pump, limiting right pump filling by 15% and 20% reduces the right ventricular stroke volume relative to that of the left by 10% and 15%, respectively, and is referred to as a 10% or a 15% stroke volume limiter. In vivo function was insensitive to afterload (Fig. 4) and showed complete ejection of the right and left blood pump against peak right pump outflow pressures of 90 mm Hg, caused by a pulmonary artery graft kink (calf 4), and peak AoP of 140 mm Hg at a pump rate of 160 beats/ min. The constant ejection velocity of the blood pump is apparent in the constant slope of the stroke traces.
|
Table I). None of the animals had clinical sequelae of hemolysis. Experiments 1, 4, 5, 7, and 8 were terminated in response to wear particles in the hydraulic fluid or inadequate hydraulic manifold cleaning procedures. Only two "first-generation" actuators were tested repeatedly in this group of animals. Identifying these correctable problems in the energy converter and actuator assembly has led to revised manufacturing, cleaning, and quality control procedures for the next generation of devices, to be tested in 1994. The remaining experiments were electively ended as a result of a ventilator failure in calf 2, a device assembly error in calf 3, and an abdominal aortic embolus from fungal infection in the aortic graft (thought to have resulted from infection of a percutaneous pressure line) in calf 6.
Exercise studies
Physiologic monitoring during treadmill exercise evaluated pulmonary function, tissue perfusion, and the hemodynamic response to increased venous return. All exercise studies were performed at the maximum output of the device at essentially a fixed cardiac output. For the development of TAHs, these data indicate the obtainable level of exercise tolerance, as well as the hemodynamic stability of future patients sustained on limited TAH output.
As is characteristic of a fixed cardiac output, the maximum oxygen consumption obtainable was limited. The average maximum oxygen consumption was between 780 and 844 ml oxygen per minute in these calves, or between 8.67 and 9.38 ml oxygen per minute per kilogram for their approximate body weight of 90 kg. The average maximum oxygen consumption for an active 40-year-old man performing treadmill exercise is 45.2 ml oxygen per minute per kilogram as compared with approximately 20 ml oxygen per minute per kilogram measured for patients with angina pectoris either alone or in combination with a healed myocardial infarction. 6 A maximum cardiac output of 10 L/min would provide a 70 kg adult of average body surface area (1.75 m2) a maximum oxygen consumption of 15 to 20 ml oxygen perminute per kilogram during exercise. 7 These preliminary data suggest that the maximum activity level of TAH recipients using these first-generation electrically powered devices will be limited. In the calf studies, oxygen extraction at the tissue level is increased to compensate for a fixed cardiac output. Average venous oxygen saturation levels fell below 35%, and the average arteriovenous oxygen content difference reached 83% of maximum after only the third speed increment, to 0.7 mph, at 9 minutes of exercise.
The calves did, however, tolerate exercise well, averaging 22 minutes on the treadmill and reaching speeds of up to 2.1 mph. Average mixed venous lactate levels exceeded the maximum laboratory normal value of 2.2 mmol/L only at speeds exceeding 1.9 mph, after 27 minutes of exercise. No signs of pulmonary congestion were noted during the exercise, and each exercise was terminated as a result of fatigue. This observation was supported by data showing a steady decrease in average arterial carbon dioxide tensions from preexercise levels throughout the duration of the exercise (Fig. 5).
|
4 a maximum fixed cardiac output, and increased venous return with exercise. Both LAP and RAP increased gradually and proportionally as speed increased from resting (preexercise levels) to 1.3 mph at 18 minutes of exercise in which either a 10% or 15% stroke volume limiter was used. The increases of the LAP and RAP during that period were 10.3 ± 2.5 mm Hg and 8.4 ± 1.6 mm Hg, respectively. These preliminary exercise data suggest that at its fixed maximum output, the CC-N TAH is capable of balancing ventricular outputs and stabilizing atrial pressures for up to 18 minutes of increasingly strenuous exercise. An acceptable safe level and duration of exercise such as this is essential so that fatigue, not pulmonary congestion, is the physical sign for terminating an activity.
HUMAN FITTING STUDIES
The CC-N TAH was designed to fit in the chest of most adult patients. Preliminary anatomic studies were performed with chest computed tomographic scans, magnetic resonance imaging scans, and cadaver studies. 8 The cadaver implants gave limited clinical information because the rigidity of the tissues did not approximate the expected clinical situation. Although computed tomographic and magnetic resonance imaging scans provided excellent anatomic information, in clinical practice they may not be obtainable for patients who are morbidly ill and in unstable condition (many are receiving inotropic drugs and are supported with intraaortic balloon pumps) and who are being considered for TAH implantation. We sought easily obtainable anatomic information to accurately predict the safe fit of the pump in the patient.
Therefore, we studied six patients undergoing orthotopic cardiac transplantation to assess anatomic information regarding "fit" of the pump, as well as orientation of the outflow and inflow ports. The information gathered included (1) external chest measurements, (2) plain chest roentgenogram dimensions, (3) intraoperative transesophageal echocardiographic measurements, and (4) intraoperative measurements of intrathoracic dimensions. Transesophageal echocardiography was chosen instead of magnetic resonance imaging and computed tomographic scans because it can be easily performed in critically ill patients. 9 The extensive transesophageal echocardiographic measurements included intracardiac dimensions, as well as measurements of the great vessels. These preliminary data indicate external chest dimensions may be poor predictors of anatomic fit. Estimation of potential chest cavity space on the left side from anteroposterior and lateral chest x-ray films and transesophageal echocardiographic measurements of critical cardiac dimensions may be easily obtained clinical predictors.
All of the initial transplant recipients studied were male, but we plan further study of female transplant recipients and additional study in female cadavers. On the basis of the findings in the initial two patients, the orientation of the left ventricular outflow port was adjusted. In the remaining four patients, the geometric orientation of the inflow and outflow ports was satisfactory. In five patients, chest dimensions were adequate to easily accommodate the TAH heart (Fig. 6). In one patient (6), who had been supported with a HeartMate LVAD (Thermo Cardiosystems, Inc., Woburn, Mass.) for 52 days, the heart had decreased to normal size. Because of the relatively small pericardium (compared to many patients with cardiomyopathy) and surrounding scar tissue, the TAH did not easily fit in the pericardium. Studies in the other five patients with cardiomegaly indicate that the pump is small enough to fit easily in the majority of patients with cardiomyopathy. This human fitting study is ongoing and will include many more patients before the criteria for predicting fit in potential recipients are firmly established.
|
In general, univentricular support with an implantable LVAD has been successful for many patients awaiting cardiac transplantation. 10-12 Long-term LVAD support has certain advantages over TAH replacement. First, the recipient's heart is not removed and therefore "device failure" may not be fatal because the recipient's heart function may be adequate to support circulation, at least temporarily. This advantage may be more apparent than real, however, because LVAD failure for 5 minutes or more may lead to stasis thrombus formation within the LVAD and may prove fatal. Second, although patients may have "biventricular failure," many patients with right ventricular dysfunction may show improvement of right ventricular function while receiving isolated left ventricular support. 9,10,13 Animal experiments also suggest that right ventricular performance improves during left ventricular support. 14 Third, the "vented-electric" portable LVAD has already advanced to the stage that clinical use has been initiated. 15
However, clinical experience with the implantable LVADs used as a bridge to heart transplantation has provided useful lessons regarding certain patient groups. First, patients with prosthetic aortic valves are at high risk during support with an implantable LVAD because the valves do not open and tend to form thrombi. 11,12 Most LVAD investigators consider previous aortic valve replacement a contraindication to use of an implantable LVAD. Second, patients with drug-refractory ventricular tachycardia or fibrillation may be at risk during LVAD support, although some patients have tolerated prolonged episodes of ventricular tachycardia, albeit with decreased LVAD flow. 16,17 Third, patients who have a combination of right ventricular dysfunction and elevated pulmonary vascular resistance may have acute right heart failure. 10,12 This group of patients in particular may be well served by the CC-N TAH because the TAH is afterload insensitive and can pump against PAPs up to 90 mm Hg. The pump should be able to maintain adequate flows, despite high PAPs and pulmonary vascular resistance. Theoretically, even patients with primary pulmonary hypertension or Eisenmenger's syndrome (who frequently die of right heart failure) may benefit from the TAH.
A number of other potential uses for the TAH have been identified. First, patients with acute myocardial infarction and severe cardiogenic shock may not be good candidates for implantable LVADs because acutely infarcted ventricular muscle provides poor fixation for the LVAD inflow cannulas. Also, the potential for ventricular tachyarrhythmias is avoided by using the TAH. Similarly, patients with an irreparable postinfarction ventricular septal defect may be sustained by the TAH.
Some patients with restrictive cardiomyopathy, who have a small left ventricular cavity, may have an anatomic obstruction that inhibits filling through the LVAD inflow cannula, and they may be best served by the TAH. Patients with extensive thrombus in a dilated native heart may also be TAH candidates. The thrombus would serve as an ongoing source of embolus during LVAD support.
Finally, the TAH has the potential to sustain many heart transplant recipients at risk for graft atherosclerosis. For these patients, the TAH has a distinct advantage over the LVAD in that all immunosuppressive drugs can be stopped after TAH insertion. For the same reason, the TAH can sustain transplant recipients who have uncontrolled severe rejection or donor graft failure.
As clinical experience with implantable LVADs accrues, other patient groups who would be better supported by the TAH will undoubtedly be identified. When the implantable LVAD is eventually used for long-term implantation (for patients who are not transplant candidates), additional problems will likely be identified. Eventually a role for both devices will be identified, and with experience it should be possible to predict which device will best serve each patient.
FUTURE GOALS
The mission of the Cleveland Clinic–Nimbus joint venture is to develop a commercially available TAH that will be an alternative to heart transplantation or medical therapy for end-stage heart disease. Several tasks remain to be completed before final design of the system. First, although the in vivo studies verified excellent hemodynamic function of the pump, design, manufacturing, and processing changes related to the hydraulic system must be incorporated and verified in the next generation of the device to improve reliability and duration of support. Second, the gelatin lining of the CC-N TAH is thromboresistant; therefore we do not anticipate prescribing anticoagulants for patients supported by the TAH (similar to the HeartMate LVAD 9). For commercial purposes, however, the gelatin lining is best processed "dry" to facilitate distribution and storage. Before implantation, further work is needed to characterize and to create dry gelatin. Third, animal experiments with the entire implanted system (including stroke volume limiters, compliance chambers, new actuators, internal batteries, and hybrid controllers) will be performed to optimize the hemodynamic function and longevity of the device. Also, further human fitting studies are needed to allow accurate preoperative prediction of candidates for the TAH.
The results in these two articles summarize our development of the implantable TAH from a concept through initial in vivo experiments. Under phases I and II of the NIH device readiness program, many more tasks have to be completed before the implantable TAH is ready for human use, perhaps by the year 2001 when the phase II contracts end. In the meantime, the electric TAH is a "work-in-progress."
With the recent successful introduction of the vented-electric TCI HeartMate LVAD, it has become obvious that patients may attain an acceptable quality of life with a small percutaneous vent and portable batteries. 12 Therefore, the CC-N TAH will be designed as a vented-electric TAH in tandem with the self-contained, permanently implanted TAH. This "intermediate" step (vented-electric) should provide an acceptable quality of life for patients, and, from an engineering standpoint, is easier to achieve at an earlier date (Fig. 7).
|
In conclusion, the in vitro hemodynamic characteristics of the CC-N TAH have been reproduced in vivo. The pump provides an output of approximately 10 L/min within physiologic limits and allows a moderate amount of sustained exertion. Further engineering work remains to finalize pump details. Initial human fitting studies indicate that it should fit the majority of patients currently undergoing heart transplantation. Of the 50,000 to 70,000 patients per year who will be candidates for circulatory support devices by the year 2010, it is as yet unclear which patients will be best supported by implantable LVADs and which will need the TAH. 18 Bridge to heart transplantation experiences with both devices will help determine optimal candidates. The technology to achieve TAH replacement with an electrohydraulic system is well underway. Clinical trials of a vented-electric TAH used as a bridge to heart transplantation could begin before the end of the decade.
ACKNOWLEDGEMENT
We thank C.R. Bard, Inc., Cardiosurgery Division, Billereca, Massachusetts, for donation of fabrics and grafts used in this project.
References
This article has been cited by other articles:
|
|
 |
|
  C. Faber, P. M. McCarthy, N. G. Smedira, J. B. Young, R. C. Starling, and K. J. Hoercher Implantable left ventricular assist device for patients with postinfarction ventricular septal defect J. Thorac. Cardiovasc. Surg., August 1, 2002; 124(2): 400 - 401. [Full Text] [PDF]
|
|
|
|
 |
|
 P. M. McCarthy and W. A. Smith Mechanical Circulatory Support--a Long and Winding Road Science, February 8, 2002; 295(5557): 998 - 999. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. Chinchoy, C. L. Soule, A. J. Houlton, W. J. Gallagher, M. A. Hjelle, T. G. Laske, J. Morissette, and P. A. Iaizzo Isolated four-chamber working swine heart model Ann. Thorac. Surg., November 1, 2000; 70(5): 1607 - 1614. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Massiello, R. Kiraly, K. Butler, S. Himley, J.-F. Chen, and P. M. McCarthy The Cleveland Clinic-Nimbus total artificial heart: Design and in vitro function J. Thorac. Cardiovasc. Surg., September 1, 1994; 108(3): 412 - 419. [Abstract] [Full Text]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |
