| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Ann Thorac Surg 1996;62:1698-1706
© 1996 The Society of Thoracic Surgeons
Division of Cardiothoracic Surgery, Department of Surgery, Wayne State University, Detroit, Michigan, and Department of Human Anatomy and Cell Biology, University of Liverpool, Liverpool, United Kingdom
 |
Abstract |
|---|
 TopFootnotes Abstract IntroductionMaterial and MethodsResultsCommentAddendumAcknowledgmentsReferences |
|---|
Methods. The SMVs were constructed from the latissimus dorsi muscle in 10 dogs. The inner surface of each SMV was lined with autologous pericardium harvested at the time of SMV construction. After a 3-week period of vascular delay and 6 weeks of electrical conditioning to convert the muscle to a fatigue-resistant state, SMVs were connected to the descending thoracic aorta and stimulated to contract during cardiac diastole.
Results. Initial hemodynamics revealed that SMV contraction at 33 Hz increased diastolic pressure 24.7% (60.8 ± 7.3 mm Hg versus 80.3 ± 8.8 mm Hg). Skeletal muscle ventricle relaxation decreased presystolic pressure 14.4% (59.9 ± 7.7 mm Hg versus 51.3 ± 7.5 mm Hg) and decreased peak systolic pressure 4.1% (90.2 ± 7.3 mm Hg versus 86.5 ± 5.8 mm Hg). Hemodynamics were assessed at 1 to 2 weeks, then at 1, 2, 3, and 6 months, and at 6-month intervals thereafter. Hemodynamic performance remained stable for the duration of this study. After 2 years of pumping continuously in circulation, SMV contraction resulted in a 34.8% augmentation of diastolic pressure (63.6 ± 6.6 mm Hg versus 85.3 ± 6.4 mm Hg), a 17.2% decrease in presystolic pressure (54.7 ± 3.73 mm Hg versus 45.3 ± 4.1 mm Hg), and a 4.2% decrease in peak systolic pressure (95.3 ± 10.4 mm Hg versus 91.3 ± 12.3 mm Hg). Three dogs survived to 2 years with the SMVs in circulation. No animal showed evidence of thromboembolism during serial echocardiography or at autopsy and no SMVs ruptured.
Conclusions. These data demonstrate that SMVs can provide effective hemodynamic assist over an extended period without specific complications related to the SMVs.
|
Introduction |
|---|
|
TopFootnotesAbstractIntroductionMaterial and MethodsResultsCommentAddendumAcknowledgmentsReferences |
|---|
There is much interest in skeletal muscle as a potential power source for cardiac assistance. Cardiomyoplasty, in which skeletal muscle is wrapped directly around the heart to augment systolic function, has gained the most attention. Currently, cardiomyoplasty is in Food and Drug Administration phase III clinical trials in the United States and more than 500 patients have undergone the procedure worldwide [1]. Although many patients report improvements in symptoms of heart failure, improvement in hemodynamic function has been more difficult to document [2–4]. To harness more of the energy of skeletal muscle contraction for cardiac assistance, we have concentrated on using skeletal muscle to construct auxiliary blood pumps, which we term skeletal muscle ventricles, or SMVs.
The SMVs have been connected to the circulation in various configurations to provide both left and right heart assist [5–7]. However, two problems encountered during in-circulation studies have been SMV cavitary thrombosis, with the risk of subsequent embolism, and SMV rupture, with resultant hemorrhage [8]. In an attempt to decrease the incidence of these complications, we lined the inner surface of the SMV with fresh, nontreated pericardium; the initial hemodynamics on these animals were reported previously [9]. The animals in this study were then allowed to survive long-term with intermittent hemodynamic measurements. In this report, we present the follow-up hemodynamic, morphologic, biochemical, and survival data for this cohort of animals over a 2-year period.
|
Material and Methods |
|---|
|
TopFootnotesAbstractIntroductionMaterial and MethodsResultsCommentAddendumAcknowledgmentsReferences |
|---|
An eighth interspace thoracotomy was used to harvest pericardium between the two phrenic nerves from the base of the heart to the diaphragmatic reflection. The visceral surface of the pericardium was applied to the outer surface of a conically shaped, polypropylene mandrel with a base of 3 cm, a height of 6.5 cm, and a volume of 25 mL. A 5-mm ring of Dacron felt (USCI, Billerica, MA) was placed around the circumference of the mandrel's base and the pericardium was sewn to this ring with 6-0 polypropylene suture. An additional suture line of 6-0 polypropylene was also used to sew the cut edge of the pericardium to itself so that the outer surface of the mandrel was completely covered.
The thoracolumbar fascia from the dorsal edge of the latissimus muscle was sewn to the Dacron felt sewing ring with polypropylene suture starting at the medial aspect of the muscle's origin and wrapping the muscle circumferentially 2 to 2.5 times around the mandrel. Layers of the SMV were sewn to each other with absorbable suture. The SMV was then placed in position on the chest wall near the axilla. A completed SMV is depicted in Figure 1
. The nerve lead was connected to a neurostimulator (Itrel model 7421; Medtronic) that was placed underneath the left rectus abdominis muscle. After stimulator function was confirmed, the stimulator was programmed "off" and the incision was closed in layers.
|
Connection of the Skeletal Muscle Ventricle to the Circulation
After 9 weeks the animals were reanesthetized and the chest was reopened through a fourth interspace thoracotomy. The descending thoracic aorta was exposed and two 12-mm, ring-reinforced Gore-Tex conduits (W.L. Gore & Associates, Flagstaff, AZ) were anastomosed to the aorta in an end-to-side fashion with 6-0 polypropylene suture. The proximal ends of the conduits had been sewn previously to a conically shaped piece of 0.6-mm-thick Gore-Tex cardiovascular patch. This conical "base cap" was then anastomosed to the base of the SMV with polypropylene suture.
Two myocardial sensing electrodes were placed on the surface of the left ventricle (model 6917A-35T: Medtronic). The nerve lead and the myocardial leads were connected to an R-wave synchronous pulse train stimulator (SP1005; Medtronic). After initiation of SMV contraction, the conduits and SMV were filled with saline solution, air was removed, and the clamps released. The aorta was then ligated with a cotton umbilical tape between the two limbs of the graft to direct blood flow entirely through the SMV. The configuration of an SMV in circulation is depicted in Figure 2.
|
The SP1005 cardiomyostimulator (Medtronic) was programmed to deliver a 33-Hz burst frequency stimulus. Each individual pulse had a width of 210 µs and an amplitude of 1 to 2 V. A delay of 175 to 250 ms from the R-wave with a burst duration of 185 to 420 ms was chosen to correspond to each animal's heart rate so that SMV contraction began during early diastole and SMV relaxation commenced before the next systole. The contraction ratio was set at 1:2 (SMV:heart). Measurements were then made at burst frequencies of 33 Hz and 85 Hz.
After hemodynamics were recorded, the lines were removed, the incision was closed, and the animals were allowed to recover. The cardiomyostimulator was programmed for chronic stimulation with a burst frequency of 33 Hz, delay of 175 ms, duration of 225 ms, and a 1:2 contraction ratio. The animals were reanesthetized at 2 weeks and then monthly for 1, 2, and 3 months from the initial operation for repeat hemodynamic recordings. Measurements were again made at 33- and 85-Hz burst frequencies. Femoral and carotid artery pressure was recorded with a fluid-filled pressure transducer (Spectramed, CA). After the third month, recordings were made at 6-month intervals. At the time of each recording, the SMV cavity was examined for thrombus formation with two-dimensional echocardiography (model 17065; Hewlett-Packard, San Diego, CA).
For 4 animals, the cardiomyostimulator was replaced if either the SMV was noted to be noncontractile on examination or a low battery status was detected at the time of a subsequent measurement. Three stimulators were replaced with a second-generation cardiomyostimulator (Prometheus; Medtronic) that has the capability for adjusting the delay and duration for burst stimulation as a function of the R-R interval. For these animals, both the burst duration and the delay was set at 35% to 45% of the R-R interval. The contraction ratio was set to 1:2 with chronic stimulation at 33 Hz. For periodic measurements, stimulation was recorded at 33- and 50-Hz burst frequencies, because 50 Hz was the maximum setting available for this cardiomyostimulator. Because of these changes in stimulator characteristics, hemodynamic recordings during the latter half of this study were made at 33 and 50 Hz in all animals (both the Prometheus and the SP1005 stimulator can be set to 50 Hz) and also at 85Hz in animals with the SP1005 stimulator.
Digitized data were processed using a waveform analysis program (Advanced Codas and Winaq, Dataq Instruments). Data analysis was performed on a portable computer (T3200; Toshiba, Irvine, CA). Data are expressed as the mean ± the standard deviation. Statistical significance was assessed using a statistical software package (INSTAT; Graphipad Software, SanDiego, CA) with a repeated-measures analysis of variance to determine significance between groups. Significance was accepted at a p value of less than 0.05.
Biochemistry
The SMVs from dogs 6 and 8 (see Table 2) were analyzed to determine their myosin heavy chain (MHC) composition. Samples from the base, middle, and apex regions were quick-frozen in liquid nitrogen and stored at -70°C pending analysis. Crude myosin extracts were prepared by high salt extraction and MHCs were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis [11, 12]. The protein bands were stained with Coomassie blue or electroblotted onto a nitrocellulose membrane (Hybond-C, Amersham, UK), where they were probed with a monoclonal antibody specific for fast MHC subtypes [13]. The immunoblot was visualized by a peroxidase-based detection system.
|
|
Results |
|---|
|
TopFootnotesAbstractIntroductionMaterial and MethodsResultsCommentAddendumAcknowledgmentsReferences |
|---|
, and the hemodynamics for all animals over the 2-year time-period are shown in Table 1.
|
|
.
|
In 1 animal, SMV performance during isovolumic contraction was measured by clamping both inflow and outflow conduits during a terminal experiment after 779 days in circulation. Peak pressure developed by the SMV was 180 mm Hg at an afterload of 100 mm Hg. Figure 5 depicts the relationship between preload and afterload during SMV contraction.
|
summarizes the length of time in circulation and cause of death for this cohort of 10 dogs. Three dogs were killed iatrogenically during the first 2 weeks from pharmacologically induced hypotension used to assess SMV function during low cardiac output (results reported previously) [9]. Because of this early iatrogenic mortality, subsequent low-output studies were delayed until after 1 month postoperatively. The remaining 6 animals all survived long term. During its time in circulation, each animal's SMV was studied by echocardiography to check for the presence or absence of thrombus inside the SMV cavity. No SMV developed evidence of thrombus formation at any time during the duration of this experiment. As of January 29, 1996, 2 animals were alive with their SMVs in circulation at 969 and 1,004 days.
Morphology
Each animal underwent an autopsy at the time of its death. This included examination of the SMV, the heart and lungs, and the abdominal viscera, including the kidneys and spinal cord. The SMVs showed evidence of attenuation of muscle tissue and replacement by fatty and connective tissue. The inner lining of the SMV was smooth, without evidence of thrombus formation. Sections through the SMV showed the pericardial layer to be attached to the sewing ring of the SMV. Examination of the viscera and spinal cord revealed no evidence of infarction or embolism.
Biochemistry
The gel in Figure 6a shows the MHC complement of the SMVs from dogs 6 and 8 and representative control muscle samples. The control latissimus dorsi sample expressed three MHCs: the slow isoform (MHC1) and two fast isoforms (MHC2A and an isoform tentatively identified as MHC2D/2X). The SMV samples expressed exclusively MHC1; no fast MHCs were present at detectable levels in any of the samples. This was confirmed by the immunoblot shown in Figure 6b; fast MHC isoforms were detected only in the control latissimus dorsi and soleus samples.
|
|
Comment |
|---|
|
TopFootnotesAbstractIntroductionMaterial and MethodsResultsCommentAddendumAcknowledgmentsReferences |
|---|
Anderson and co-workers [5] looked at different autologous linings for the blood-contacting surface of the SMV. Some SMVs were lined with either pleura or pericardium; the rest had no lining other than the fibrous reaction that formed between the SMV and the mandrel during the conditioning process. The incidence of thrombus formation was 100% in the unlined SMVs and 70% in the SMVs lined with pleura or pericardium. Moreover, 66% of the SMVs that had been in circulation for more than 24 hours ruptured. Pochettino and colleagues [16] were able to decrease the incidence of thrombus formation to 33% by modifying the mandrel, and thereby the SMV cavity, to improve intracavitary blood flow characteristics. Nakajima and colleagues [8] studied a series of 44 SMVs connected to the descending aorta for counterpulsation. They found a 58% incidence of thrombus formation and a 34% incidence of SMV rupture. Skeletal muscle ventricle rupture occurred the first 40 days of placement into the systemic circulation; thrombus formation was generally mild, with extensive thrombus occurring in only 13% of the animals and a 2% incidence of embolic events noted at autopsy.
We have now incorporated slight modifications to SMV construction by using a pericardial lining, by reinforcing the muscle/sewing ring anastomosis with thoracolumbar fascia to increase strength at the muscle/sewing ring interface (the location of SMV rupture), and by using a conically shaped base cap to decrease stress at this point. The animals in which these modifications were made form the basis of the current report.
In our initial assessment of these animals, more extensive hemodynamic measurements were reported [9]. Aortic counterpulsation significantly improved myocardial loading conditions by decreasing presystolic pressure as well as peak arterial pressure on the heart beat after SMV relaxation. Although a decrease in peak left ventricular pressure more truly reflects systolic unloading, the decrease in presystolic pressure that we measured correlated with a decrease in aortic impedance, as well as a significant decrease in the systolic tension-time index of 7% to 8%. Diastolic augmentation improved the diastolic pressure-time index 12% to 18%. The ratio of these, the endocardial viability ratio, increased 25% to 33%, reflecting improvement in the oxygen supply/demand ratio for the heart. Pressure-volume loop analysis revealed that counterpulsation significantly decreased ventricular stroke work by 15% to 19% and decreased the pressure-volume area, which is related to myocardial oxygen consumption, by 16% to 25%. Ventricular contractility, measured by end-systolic elastance and preload recruitable stroke work, was also improved.
For the current study, the above hemodynamic parameters were not reported because left ventricular pressure, needed to calculate the tension-time index, diastolic pressure-time index, and endocardial viability ratio, could not be measured in all animals throughout the study period. This was because repeated carotid cannulation for retrograde placement of a left ventricular micromanometer led to carotid occlusion, eliminating continued access to left ventricular pressure. Therefore, only the arterial pressures (peak systolic, presystolic, and mean diastolic) could be used throughout the study period. This represents a limitation of this study. However, as the arterial waveforms and absolute values for arterial pressure during diastolic augmentation and presystolic unloading remained similar and stable over the 2-year period of follow-up, it is likely that improvements in the load-sensitive parameters tension-time index, diastolic pressure-time index, and endocardial viability ratio, as well as the improvements in the load-insensitive parameters pressure-volume area, end-systolic elastance, and preload recruitable stroke work were sustained during chronic SMV counterpulsation. In addition, measurements available in some of the animals before carotid occlusion occurred revealed similar hemodynamics to our initial recordings.
During the study period, three stimulators were replaced with a newer generation pulse generator (Prometheus, Medtronic), which determined the burst duration and delay as a percentage of the R-R interval rather than a fixed time delay. This is a much more desirable method for chronic stimulation; although fixed time delay can be adjusted and optimized during measurements, stimulation could occur at an inappropriate time during the cardiac cycle when an animal is exercising or excited during chronic stimulation. Previous characterization has shown that a delay and duration of 40% of the R-R interval provides optimal hemodynamics [17]. In the current report, the percentile changes in presystolic pressure were greater during the latter half of the 2-year period. This is may have been attributable to the use of the newer stimulator and the ability to optimize timing. However, in comparing the presystolic pressures obtained from the initial 6 months with subsequent measurements, we found no statistically significant difference.
Few deaths were related directly to performance of the SMV. One dog died postoperatively of a presumed arrhythmia, representing a perioperative mortality of 10%. Three of 4 dogs died during the first 2 weeks during propranolol-induced hypotension. These data, reported earlier, showed that SMVs exhibited improved functional hemodynamics under failure conditions [9]. These deaths are not attributable to SMV function, but represent a learning curve in performing this pharmacologic study. The two deaths caused by infection were related to intermittent hemodynamic monitoring, rather than the function of those SMVs. If these animals are not included in survival statistics, then the survival for SMVs in circulation is 80% at 1 year and 60% at 2 years, with deaths attributable to early perioperative mortality and failure of the pulse generator. Most important, no deaths were attributable to thromboembolism or SMV rupture.
Little is known about the characteristics of the muscular wall of SMVs that have been in circulation over an extended period of time. The findings of Salmons and Sreter [18] were the starting point for our early studies of stimulation of skeletal muscle for use in cardiac assist. We showed that chronic electrical conditioning at 2 Hz up to 1 year was well tolerated and continued to convert the muscle phenotype from that of type II, fast twitch fibers to that of type I, slow twitch fibers [19]. The gel electrophoresis data presented here confirm changes in the MHC isoforms in the muscle of SMVs functioning chronically in the circulation. Fast MHC isoforms, which predominate in the control latissimus dorsi muscle, were undetectable in the muscle from the chronically stimulated SMVs. This is evidence of a complete fast-to-slow fiber-type transformation induced in the muscular wall of the SMV by conditioning and chronic activation.
In summary, this study presents follow-up hemodynamic data for 10 dogs with pericardium-lined SMVs over a 2-year period. Skeletal muscle ventricle function, measured by diastolic augmentation and presystolic unloading, remained stable and reproducible throughout the 2-year period. The potential complications of intracavitary thrombus formation and SMV rupture at the muscle-Dacron felt interface did not occur in any of these animals. Overall survival was determined by additional experimental manipulations and was unrelated to SMV function. Biochemical evidence is presented to confirm that the SMVs maintained MHC expression consistent with their conditioned status during chronic, in-circulation conditions. If these data are reproducible, then the clinical application of SMVs as auxiliary ventricles appears to be feasible.
|
Addendum |
|---|
|
TopFootnotesAbstractIntroductionMaterial and MethodsResultsCommentAddendumAcknowledgmentsReferences |
|---|
|
Acknowledgments |
|---|
|
TopFootnotesAbstractIntroductionMaterial and MethodsResultsCommentAddendumAcknowledgmentsReferences |
|---|
|
Footnotes |
|---|
|
TopFootnotesAbstractIntroductionMaterial and MethodsResultsCommentAddendumAcknowledgmentsReferences |
|---|
Address reprint requests to Dr Stephenson, Division of Cardiothoracic Surgery, Wayne State University, Suite 228 Harper Professional Building, 3990 John R St, Detroit, MI 48201.
|
References |
|---|
|
TopFootnotesAbstractIntroductionMaterial and MethodsResultsCommentAddendumAcknowledgmentsReferences |
|---|
Related Article
This article has been cited by other articles:
|
|
 |
|
  L. I. Astra and L. W. Stephenson Skeletal Muscle as a Myocardial Substitute Experimental Biology and Medicine, July 1, 2000; 224(3): 133 - 140. [Abstract] [Full Text]
|
|
|
|
 |
|
 A. Kalangos, M. Beghetti, A. Baldovinos, D. Vala, T. Bichel, B. Mermillod, N. Murith, I. Oberhansli, B. Friedli, and B. Faidutti AORTIC VALVE REPAIR BY CUSP EXTENSION WITH THE USE OF FRESH AUTOLOGOUS PERICARDIUM IN CHILDREN WITH RHEUMATIC AORTIC INSUFFICIENCY J. Thorac. Cardiovasc. Surg., August 1, 1999; 118(2): 225 - 236. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. A. Thomas, F. A. Baciewicz Jr, R. L. Hammond, K. A. Greer, H. Lu, S. Bastion, P. Jindal, and L. W. Stephenson POWER OUTPUT OF PERICARDIUM-LINED SKELETAL MUSCLE VENTRICLES, LEFT VENTRICULAR APEX TO AORTA CONFIGURATION: UP TO EIGHT MONTHS IN CIRCULATION J. Thorac. Cardiovasc. Surg., December 1, 1998; 116(6): 1029 - 1042. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. W. Stephenson Skeletal Muscle Ventricles in the Pulmonary Circulation: Up to 16 Weeks' Experience Ann. Thorac. Surg., September 1, 1998; 66(3): 977 - 979. [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 |
