| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Ann Thorac Surg 1999;68:780-784
© 1999 The Society of Thoracic Surgeons
a Whalen Biomedical Laboratories, Somerville, Massachusetts, USA
b The University of Utah, Salt Lake City, Utah, USA
Address reprint requests to Dr Whalen, Whalen Biomedical Laboratories, 11 Miller St, Somerville, MA 02143com.com
e-mail: rlwhalen{at}ix.net
Presented at the Fourth International Conference on Circulatory Support Devices for Severe Cardiac Failure, Houston, TX, Oct 3–5, 1997.
Abstract
Background. We are developing and testing a new ventricular assist device (VAD) to be powered by conditioned skeletal muscle.
Methods. To evaluate the VAD hardware and to develop a muscle training regimen, 8 calves have been used in studies in which the right latissimus dorsi muscle was employed. The experiments were carried out to an approximately 4-month duration.
Results. There was significant conversion of type II (fast twitch) to type I (slow twitch) muscle fibers. This did not correlate well, however, with device performance. The device stroke volumes ranged from approximately 17 to 90 cc. This variability of outcome occurred despite the fact that identical hardware, surgical procedures, and training regimens were employed.
Conclusions. The results from the first eight studies lead us to speculate that perfusion may be important even when the muscle is working at pressures much lower than systemic blood pressure levels. In an attempt to augment tissue perfusion, we plan to investigate thermally induced angiogenesis as a possible mechanism for increasing blood flow to the tissue.
T he idea of employing conditioned skeletal muscle (CSM) as a power source for mechanical circulatory support (MCS) is uniquely appealing. At this time, CSM is the only method for providing MCS being studied that does not require a source of external power for the implanted device. This is a significant advantage. In addition to the obvious quality of life limitations associated with devices that require an external power input, there are also basic technical issues that must be resolved relating to either percutaneous or transcutaneous energy delivery to an implanted blood pump, as well as volume compensation for the case of pulsatile output devices that are totally intracorporeal.
We are developing a new ventricular assist device (VAD) that is to be powered by CSM. The conceptual foundation for this device rests on: the ongoing clinical experience with dynamic cardiomyoplasty [1], the pioneering work with skeletal muscle ventricles (SMVs) by Stephenson and associates [2]; and the development of dynamic conditioning for training skeletal muscle as described by Guldner and associates [3].
A block diagram of the CSM-powered VAD is shown in Figure 1. An implanted myostimulator is used to electrically condition the latissimus dorsi muscle (LDM). This muscle has the advantage of having a single major neural innervation and blood supply (the thoracodorsal nerve and artery respectively), as well as the fact that there is no major clinical detriment associated with its use.
|
|
The potential of skeletal muscle to pump blood in the systemic circulation is more clearly demonstrated by experiments performed with SMVs. In that technique, a blood pumping chamber is fabricated from the skeletal muscle itself. Generally, the SMV is lined with an autologous tissue such as pericardium to provide blood compatibility. In the laboratory of Stephenson, a 4-year-duration canine study with a functioning SMV was recently electively terminated (Stephenson L, personal communication). To our knowledge, this is by far the longest duration in vivo study with a functioning blood pump that has yet been achieved. While this study used the SMV as an in-series device to provide diastolic augmentation, SMVs are now being employed in a left ventricular apex-to-aorta model in that laboratory with some success [5]. These results, as well as the long-term follow-up from dynamic cardiomyoplasty, strongly suggest that transformed skeletal muscle has the potential to provide long-term cardiac support [6].
A primary limitation of both dynamic cardiomyoplasty and SMVs, however, is that both modalities provide no immediate benefit to the recipient. That is, there is no significant circulatory support provided by the skeletal muscle until it has been conditioned to perform continuous work. Its utility in critically ill patients is thus limited, and these modalities are not readily useful in those patients who most need circulatory support. The CSM-powered VAD configuration we are developing overcomes this significant limitation of the use of skeletal muscle as a power source for circulatory support by providing circulatory support during the period of muscle conditioning. This is accomplished by driving the blood pump pneumatically with an external power source.
During muscle conditioning, the LDM is dynamically trained using an extracorporeal load device (ELD). The ELD enables the pressure load seen by the muscle-wrapped drive bladder to be incrementally increased as muscle performance improves. The ELD also permits the pressure/volume work being done by the muscle to be determined on a beat-by-beat basis. This ensures that the VAD will be converted to muscle-powered operation only when it has been demonstrated that the CSM is capable of generating sufficient power to pump blood in the systemic circulation.
Recently, Guldner and associates [3] have described an improved method for training skeletal muscle to perform continuous work, which they have termed "dynamic conditioning." In this regimen, an implantable training appliance whose mechanical properties are controlled externally is used to train skeletal muscle in two distinct phases: rate conditioning at low load pressures, and load conditioning at increasing afterload pressures. This methodology makes intuitive sense in that the transformation from anaerobic to aerobic metabolic pathways that occurs during skeletal muscle conditioning almost surely occurs over an extended period of time, and it may not be beneficial for the muscle to work continuously against high pressures before it is metabolically adapted to do so.
We have taken this concept a step further. It is a logical extension of dynamic training to utilize an externally powered blood pump to support and stabilize the circulation of a recipient while the skeletal muscle is being independently trained. This is a fundamental rationale for the design of the CSM-powered VAD we are developing. By supporting the circulation with a prosthetic blood pump during this training process, it also will be possible to utilize the device in recipients with profound cardiac failure, overcoming a current limitation of both cardiomyoplasty and SMVs.
Material and methods
The most crucial component in the new VAD is the drive bladder. It converts muscle contraction into the hydraulic output needed to drive the blood pump. The drive bladder has unusual mechanical requirements. It does not collapse on each beat, but rather expands and contracts to change its volume. In order to prevent the formation of a thickened fibrous tissue capsule at the junction between the LDM and the bladder, the exterior surface of the drive bladder employs a microporous textured surface that promotes connective tissue ingrowth. Attachment of the LDM to the drive bladder eliminates relative motion at the interface between the elastomer and the muscle.
The flexing part of the drive bladder wall is made thin to minimize work performed by the LDM against the elasticity of the bladder wall, a system inefficiency. In the region of the drive bladder in contact with the LDM, the wall thickness is on the order of 0.35 mm, or about two to three times the wall thickness of a typical intraaortic balloon pump. The compliance of the drive bladder over the nominal 60-mL stroke volume of the VAD is in the range of 12–15 mm Hg.
Because it is so thin, the drive bladder is not able to support its own weight. To maintain its shape, the drive bladder is supported with a rigid internal support structure that also serves to eliminate longitudinal compliance, another potential system inefficiency. A cross-sectional drawing of the drive bladder/support assembly used in the in vivo studies is shown in Figure 3.
|
|
The drive bladder was inflated to a pressure of approximately 20 mm Hg to place some tension on the muscle. The muscle-wrapped drive bladder was then positioned within the chest, the myostimulator placed in a subcutaneous pocket, and the incision closed. No stimulation was used during a vascular delay period of 2–3 weeks duration. This period also provided for attachment of the LDM to the textured surface of the drive bladder.
Stimulation was applied with a pulse amplitude of 5 V, burst frequency of 30 Hz, and a pulse width of 0.25 ms with six pulses per burst. This burst profile was not changed during the studies. Stimulation was initiated at a rate of 2 bursts/min, and on weekly intervals the burst rate was increased to 4, 6, 8, 16, 32, and 64 bursts/min. The experiments were continued for 3–4 months.
Results
All animals tolerated the surgical procedure well. One animal was lost to the study when it dislodged the implanted device at approximately 10 weeks postoperatively, resulting in a pneumothorax. There was also one mechanical failure, the fracture of a plastic right angle pneumatic fitting on the driver bladder. (It has since been replaced with a metal fitting.) There were no failures of the elastomeric drive bladder itself, and approximately 18 months of accumulated in vivo operating time has now been achieved with this component. As the mechanical endurance of the muscle-wrapped drive bladder cannot be effectively evaluated in vitro, this is an important result.
The pneumatic pressure excursion in the external accumulator tank was recorded periodically to provide a measure of the device stroke volume. Figure 5 illustrates a typical pressure recording. The stimulation rate here was 32 bursts/min, and the pressure change generated was approximately 20 mm Hg, corresponding to a stroke volume of 70 cc. Device stroke volumes ranged from 17 cc to approximately 90 cc in this series of studies. This rather wide range of outcome was not anticipated, since identical hardware, surgical procedures, and training regimens had been employed.
|
|
Comment
The variability of outcome as measured by the device stroke volume in this series of studies was perplexing. As mentioned earlier, the implantable hardware and the surgical procedure employed were identical. Furthermore, the tension on the muscle wrap was controlled for the duration of each study by maintaining the internal pressure of the drive bladder at a fixed value. Thus, the potential for the tightness of the muscle wrap (or similar procedural differences) to contribute to the observed variability is minimal.
These results lead us to speculate that even during rate conditioning, when the LDM is working at low load pressures, perfusion may be a significant factor determining LDM performance. This premise is somewhat counterintuitive, however. It had been our intention to examine methods for augmenting blood flow in the muscle during load conditioning, when the pressure work being done by the muscle is increased. That perfusion may be a limiting factor in determining muscle performance even at low pressures suggests that any attempts to improve blood flow to the tissue should be initiated early.
Our first approach to augmenting LDM perfusion will be through the use of heat addition to the tissue. Thermally induced angiogenesis has the potential to provide a simple, nonpharmacological method for improving muscle perfusion. Based on an observation of Norman and associates [7], it will be tested by incorporating an electrical resistance heater in the hydraulic drive bladder. This heater will elevate the temperature of the drive fluid. As the lowest thermal resistance pathway by far is through the thin-walled drive bladder in contact with the muscle, this heat will be preferentially dissipated in the LDM.
Load training in which the work performed by the LDM is incrementally increased will be accomplished using a hydraulically operated drive bladder. An extracorporeal load device (ELD) consisting of a spring loaded piston is to be employed. The ELD permits the preload and afterload on the drive bladder to be independently adjusted. By measuring the piston stroke length and internal pressure, a pressure/volume loop to determine the work being performed by the LDM on each beat may be calculated. This will insure that the muscle power is adequate for pumping blood in the systemic circulation so as to avoid low flow states when the muscle is used to drive the blood pump.
Acknowledgments
This work was supported by contract NO1-HV-58158 from the National Heart, Lung, and Blood Institute. We would like to acknowledge the advice, encouragement, and support of Dr Larry W. Stephenson, who serves as a Consultant in this program, and Drs Robert L. Hammond and Kevin A. Greer from his laboratory, who assisted us in our initial in vivo studies.
Footnotes
Drs Robert L. Whalen, John C. Norman, and Messrs Christopher L. Richards, Gilbert W. Lim, and Craig W. Sherman, are employees of Whalen Biomedical Incorporated. Dr Whalen is a stockholder in Whalen Biomedical Incorporated.
References
This article has been cited by other articles:
|
|
 |
|
  K. J. Gustafson, J. D. Sweeney, J. Gibney, and L. A. Fiebig-Mathine Skeletal muscle ventricle pressure-volume properties conform to dynamic and static conditioning Ann. Thorac. Surg., September 1, 2003; 76(3): 828 - 835. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. R. Trumble and J. A. Magovern Method for measuring long-term function of muscle-powered implants via radiotelemetry J Appl Physiol, May 1, 2001; 90(5): 1977 - 1985. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 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 |
