- Email: [email protected]
Hydraulic pouches of canine latissimus dorsi
J
THoRAc CARDIOVASC SURG
91:534-544, 1986
Hydraulic pouches of canine latissimus dorsi Potential for left ventricular assistance We have studied the fatigue rates of hydraulic pouches constructed in the form of a multilayered conical spiral using the latissimus dorsi muscle of 17 beagles. The roles that electrical muscle conditioning and early interruption of collateral blood supply have in the prevention of pouch fatigue were evaluated. The length of time that a pouch could generate flow in a hydraulic test system was measured; afterload was set at 80 mm Hg and preload 24 mm Hg. Pouches (N = 3) fashioned from muscles subject to neither electrical conditioning nor a vascular delay generated an initial flow of 990 ± 346 ml/min, but could sustain flow for only 2.3, 3.8 and 3.6 minutes. Pouches (N = 5) constructed with electrically unconditioned muscles after a vascular delay (median 3 weeks) demonstrated a variable improvement in fatigue rates (initial flow 826 ± 265 m1/min; time to no forward flow, 2.5, 7.5, 7.5, 10, and 200 minutes). Four of six pouches that received the benefit of long-term electrical muscle conditioning and a vascular delay (N = 6) were able to generate flow for a 4 hour period, at which time the experiment was terminated (initial flow 478 ± 204 m1/min; final flow 195 ± 157 ml/min), After the 4 hour fatigue test was completed, one electricaUy conditioned pouch was placed in series with the heart and served as a counterpulsator. The initial volume of blood pumped by the muscle pouch was 262 ml/min or 13.8 % of cardiac output. After the pouch had contracted at a rate of approximately 45 beats/min for 1 hour, the volumeof blood pumped was 178 m1/min, or 11 % of cardiac output In three other animals a pouch was fashioned and then left in situ for a 1 to 3 week period before hydraulic testing. These pouches generated significant initial flows (390 ± 60 m1/min), which demonstrates the feasibility of further study of permanent pouches. These results suggest that permanent electrical muscle conditioning and perhaps a vascular collateral delay might permit an auxiliary skeletal muscle-powered ventricle to assume a portion of left ventricular function.
John D. Mannion, M.D., Robert Hammond, B.S., and Larry W. Stephenson, M.D.,
Philadelphia. Pa.
h e 1 year mortality for the estimated 10,000 patients who present annually in the United States with irreversible congestive heart failure is approximately 50%.1 This high mortality persists despite significant advances made in medical therapy, cardiac transplantation, and the artificial heart program. Since none of these three presently available treatments for severe cardiac pump failure can be expected to benefit a large number of From the Harrison Department or Surgical Research and The Department of Cardiothoracic Surgery. Hospital of the University of Pennsylvania. Philadelphia. Pa. Supported by National Institutes of Health Grant HLBI 27570. Read at the Eleventh Annual Meeting or The Western Thoracic Surgical Association. Incline Village, Nev .. June 16-20. 1985. Address for reprints: John D. Mannion. M.D., 313 Medical Education Building, University or Pennsylvania. Philadelphia. Pa. 19104.
534
patients in the near future, we have been investigating a potential fourth therapeutic alternative: the use of a patient's own skeletal muscle to augment myocardial performance. A functioning skeletal muscle-powered ventricle would not be subject to tissue rejection, would not be hampered by donor shortage, and would obviate the necessity for cumbersome and expensive external power sources. Although theoretically appealing, skeletal musclepowered cardiac assistance has been unworkable thus far because of muscle fatigue.l' Skeletal muscle fiber type, surgical trauma, interruption of collateral blood supply, stimulation rates, pressure versus volume work, and muscle mass-to-ventricular volume ratio are only a few of the variables that have been related to the fatigability of skeletal muscle-powered ventricles. Strategies are needed that would enhance the endurance of
Volume 91 Number 4 April, 1986
auxiliary ventricles by eliminating some of the factors that lead to fatigue. In this experiment three unique approaches were taken. First, to minimize the acute effects of ligation of collateral blood supply, a several week delay was introduced between the interruption of major collateral vessels and the exercise testing of muscles. Second, canine latissimus dorsi muscle was electrically conditioned at a frequency of 2 Hz for 6 weeks, in an attempt to transform fatigue-sensitive, type II fibers into fatigue-resistant, type I fibers. Third, to increase the ratio of wall thickness to muscle pouch volume and thereby lower wall stress, pouches were constructed of latissimus dorsi muscle wrapped in the form of a multilayered conical spiral. This experiment was designed to determine if any combination of electrical conditioning, vascular delay, and a multilayered pouch would minimize muscle fatigue and permit an auxiliary ventricle to assume a portion of left ventricular function. Methods and materials Seventeen adult male beagles, with a mean weight of 13.0 ± 2.1 kg, were studied. Latissimus dorsi pouches were constructed and were subjected to a hydraulic fatigue test, which will be described subsequently. The animals were divided into four groups. In Group I (N = 3) there was no preparation of the latissimus dorsi muscle before the construction of the muscle pouch and the hydraulic test. In Group II (N = 5) the multiple segmental collateral blood vessels to the distal latissimus, arising from the intercostal arteries, were divided, and the muscles were rested for a median of 3 weeks before pouch construction and hydraulic fatigue testing. In Group III (N = 3) a permanent pouch was evaluated. The latissimus was mobilized and a pouch fashioned and left in situ for a 1 to 3 week period. In Group IV (N = 6) the latissimus dorsi muscle was prepared by division of chest wall collaterals and a rest period of 3 weeks, as in Group II. The muscle was then continuously stimulated via the thoracodorsal nerve for 6 weeks before pouch construction. Procedures for vascular delay and electrical conditioning. In Groups II, III, and IV, in which the latissimus muscles were subjected to either vascular collateral delay or electrical conditioning, an initial sterile procedure was performed. The animals were anesthetized, intubated, and draped in a sterile fashion. Preoperative and intraoperative antibiotic prophylaxis with cefazolin sodium (Ancef, 250 mg) was given intravenously. In Group II division of intercostal collaterals necessitated reflection of the muscle from the chest
Hydraulic pouches of canine latissimus dorsi
535
wall, extending from the tip of the scapula to the insertion of the latissimus into the thoracolumbar fascia. The muscle itself was not mobilized, and the division of vessels did not include the collaterals from the trapezius or teres major, but did include the major serratus collateral. The major axial blood supply to the latissimus-the thoracodorsal artery-was not disturbed. The wound was then closed, and the pouch was constructed 3 weeks later, immediately before fatigue testing. In Group III the entire latissimus was mobilized for construction of an in situ pouch. The intercostal collaterals were divided and a plane was developed between the platysma and latissimus. The thoracolumbar fascia was divided, and the muscle was mobilized proximally. The collateral vessels between the trapezius and the latissimus were then interrupted, and the muscle was separated from the teres major. The insertion of the latissimus into the humerus and the thoracodorsal neurovascular pedicle was left intact. With this method it was possible to mobilize the entire latissimus without dividing any muscle fibers. A modified Medtronic pacing electrode was placed around the thoracodorsal nerve to later facilitate hydraulic fatigue testing. The muscles were then wrapped around a 30 ml Teflon cone, the pouch approximated to the chest wall, and the subcutaneous tissue and skin closed over the muscle pouch. Hydraulic fatigue testing was performed 1, 3, and 3 weeks later. Group IV muscles received vascular and electrical conditioning. The intercostal collaterals were divided, as in Group II. A modified Medtronic electrode was then placed around the thoracodorsal nerve and attached to a Medtronic pulse generator, Model 5984 or 8422. The pulse generator was placed in a pocket fashioned beneath the right rectus muscle. The pulse generatorlead system was used to electrically condition Group IV muscles by stimulating the thoracodorsal nerve at a continuous frequency of 2 Hz for a 6 week period. The pulse generator was set to one-half amplitude (3.2 volts) and the pulse width to 220 1J.Sec. At the end of the 6 week period of stimulation, the pouch was constructed and the hydraulic fatigue test performed. The Medtronic electrodes were modified in the following manner. The platinum ring and coil wire lead of a Medtronic Model 6904 bipolar endocardial lead was removed. The ring was fashioned into a half cylinder of 2 to 3 mm radius and placed in a silicone rubber cuff. The coil wire was then reinsulated with silicone rubber tubing. Hydraulic pouch construction. The latissimus dorsi muscle is irregularly shaped but roughly corresponds to a broad-based triangle. The posterior aspect of the
The Journal of Thoracic and Cardiovascular
5 3 6 Mannion, Hammond, Stephenson
Surgery
"1.5 - 2 more \
t urn s
Fig. 1. Example of construction of latissimus multilayered ventricle in form of conical spiral. Representation of hydraulic test system. Preload and afterload was adjustable . A Bjork-Shiley valve was inserted in each limb of the Y.
muscle is thicker than the anterior. A pouch was constructed by wrapping the muscle around a specially prepared condom so that the muscle assumed the form of a conical spiral. An attempt was made to have the thickest portion of the muscle correspond with the widest portion of the skeletal muscle ventricle (Fig. I). The condom received one and one half to two wraps of muscle. The proximal thoracodorsal vessels were positioned on the outside of the muscle wrap. The condom was prepared in the following manner. A 1 em excessivelength of condom was passed over a 5 em multiply perforated 3/8 inch inner diameter (10) silicone tube, which was stretched over a Y.2 inch 10 polycarbonate connector. The connector was passed through an ovoid of low-porosity Teflon felt (USCI, Teflon felt, (07839), approximately 2 by 3 em, which served as the base of the auxiliary ventricle. The latissimus muscle layers were approximated to the Teflon felt with 2-0 Dexon sutures. The connector was made watertight and attached to the hydraulic test circuit. Hydraulic test circuit. The hydraulic system consisted of an adjustable inlet reservoir and outlet column (Fig. 1). The pouch served as the zero reference level. The height of the inlet reservoir was adjusted so that the pouch was exposed to 24 mm Hg filling pressure. The height of the afterload column was adjusted so that the pouch had to contract against a desired pressure. During pouch filling, a 65/35 saline/glycerin mixture would flow from a 2 L reservoir, through a Travenol miniprime heat exchanger, and through an 18 mm Bjork-Shiley prosthetic aortic heart valve proximal to a Y-connector and into the pouch. With contraction of the auxiliary ventricle, the inlet valve would close and a 16 mm
Bjork-Shiley valve on the other side of the Y-connector would open. At the height of the column the flow was diverted back into a catch funnel for measurement and then into the inlet reservoir. Pouch and afterload pressures were recorded. Experimental protocol. After the pouch was constructed, the preload was set to 24 mm Hg and the pouch volume was measured. Afterload was set to 80 mm Hg. The pouch was then made to contract at a rate of 48 beats/min. This was accomplished by stimulating the thoracodorsal nerve at a frequency of 25 Hz for 300 msec, followed by 950 msec of rest. The voltage used provided a supramaximal stimulus of approximately 6 volts. The pulse width of the individual square-wave pulses was 220 lJSec. Pouch pressures were monitored continuously and flows were recorded every 10 minutes. The pouch was stimulated at this rate until it could no longer produce net flow over the column. In those pouches that continued to produce forward flow, the experiment was terminated at the end of 4 hours. In those pouches unable to maintain forward flow, the column was lowered to 70 and then 60 mm Hg, and the time to cessation of forward flow was recorded; At the termination of the procedure, pouch volumes at 24 mm Hg preload pressure were again determined. Body temperature was maintained with a heating blanket, and percent oxygen saturation was kept greater than 90% with supplemental oxygen. Throughout the experiment, the ventilator was adjusted to maintain pH between 7.35 and 7.45. Serial hematocrit values were monitored. An infusion of 5% dextrose in normal saline was given, and infusion volume was varied according to the central venous pressure and cardiac output.
Volume 91 Number 4
Hydraulic pouches of canine latissimus dorsi 5 3 7
April. 1986
Table I. Systolic pressures and flows of pouch when contracting against a glycerin/saline column producing 80 mm Hg of static diastolic pressure Pmax (mm Hg) Initial Group I (N = 3) Unconditioned muscle No vascular delay Acute pouch construction Group II (N = 5) Unconditioned muscle Vascular delay Acute pouch construction Group III (N = 3) Unconditioned muscle Pouch construction Pouch in situ for 3 weeks Group IV (N = 6) Conditioned muscle Vascular delay Acute pouch construction
Flow (mljmin)
I
Final
Duration offlow (min)
Final
Initial
228 ± 28
121 ± 13
990 ± 346
o
2.3, 3.8, 3.6
279 ± 42
134 ± 26
826 ± 265
o
7.5, 10, 2.5, 7.5., 20
160 ± 20
108 ± 23
390 ± 60
o
7, 55,67
214 ± 46
ISS ± 24
478 ± 204
195 ± 157
I
240 (N = 4) 230(N=I)* 120 (N = nt
Legend: Forty-eight pouch contractions per minute were generated by stimulation of the thoracodorsal nerve at 25 Hz frequency for 300 msec followed by 950 msec of rest. Depending on heart rate, 300 msec is the approximate length of diastole in canine species. Voltage used-c-o volts;duration of individual pulses-220 !JSCc. Duration of flow refers to total time that a pouch could produce any flow against 80 mm Hg diastolic afterload. *Experiment stopped when flow reached 30 rnl/rnin. tExperiment terminated when balloon condum ruptured.
After the 4 hour hydraulic test was accomplished in two electrically conditioned pouches, the duration of the pulse train and the frequency of stimulation within a pulse train were varied to establish the relationship between these variables and the generated pressures and flows. In one electrically conditioned animal, at the completion of the 4 hour hydraulic testing, the auxiliary ventricle was placed in series with the heart and served as a counterpulsator. A JA inch ID T system was placed in the distal thoracic aorta and connected to the pouch with a t/2 inch ID silicone tubing. The pouch was stimulated at 25 Hz for 300 msec during diastole at a rate of I :4 or 1: 3 in relation to the animal's heart rate. Synchronization was accomplished by monitoring the electrocardiogram on the amplifier section of an AVCO intra-aortic balloon pump. The R waves were then COunted with an IC4017 decade counterdivider that could provide 1: 1 through 1: 9 synchronization of the stimulator. Flow to and from the pouch was recorded with a Biotronix Model 610 electromagnetic flowmeter, which was calibrated before the experiment. Areas under the flow curves were integrated on a digitizing board connected to an Apple lIe microcomputer, permitting calculation of the volume of blood pumped per COntraction of the muscle pouch. Systemic arterial pressures were recorded through femoral and carotid arterial catheters.
All animals received humane care in compliance with the "Principles of Laboratory Animal Care" formulated by the National Society for Medical Research and the "Guide for the Care and Use of Laboratory Animals" prepared by the National Academy of Sciences. Statistical analysis. Results were expressed as mean ± standard deviation of the mean. Paired and unpaired t tests were used to determine differences between groups, where appropriate. Histologic features. Muscle sections were taken and frozen in liquid nitrogen. Biopsy specimens were prepared with myofibrillar adenosine triphosphatase (ATPase) stains with acid (4.2) and alkaline (lOA) preincubation. Myofibrillar ATPase stains were used to determine the degree of transformation of fast-twitch, fatigue-sensitive fibers into slow-twitch, fatigue-resistant fibers. The slides were reviewed and representative muscle bundles were evaluated on both the acid and alkaline stains. Cells were identified as fast or slow according to Brook's ATPase classification. In conditioned muscles with acid stains, a cell was counted as a slow-twitch type if it stained darkly; with alkaline stains, a cell was counted as a fast-twitch type if it stained darkly. Sections stained with hematoxylin and eosin were reviewed for muscle structure. After the biopsies, the muscle pouches were placed into formalin. All fat was later removed. Muscle weight was determined by adding the biopsy weight to the
The Journal of Thoracic and Cardiovascular
538 Mannion, Hammond, Stephenson
Surgery
I
250-
500-
200-
PEAK
400
0>
::r:
E E W 0:: ::J
(f) (f)
c
150
E
<,
300-
u u
3
100
0
--l l.L
W 0:: 0-
FILLING
50
o
200
100
o
2
3
4
TIME (hours)
Fig. 2. Pouch pressures overa 4 hour period generated by latissimus dorsi pouches constructed of muscle that was subject to a vascular collateral delay and long-term electrical conditioning. Pouches were stimulated at a rate of 48 beats/ min.
weight of the formalin-fixed tissue. The latissimus dorsi mass was then compared with the mass of the left ventricle of the same animals. Results
The performance of the pouches in the hydraulic system against the diastolic afterload of 80 mm Hg is depicted in Table I. For the muscles that were prepared with neither vascular collateral delay nor electrical conditioning, flow ceased in 2 to 4 minutes. A vascular collateral delay alone seemed to enhance the performance of electrically unconditioned muscles, but the results were not striking. Only pouches constructed of electrically conditioned muscles were capable of generating flow after 4 hours of testing. When compared to all electrically unconditioned muscles or to electrically unconditioned muscles with a vascular delay, the prolongation in flow for the electrically conditioned pouches is significant (p < 0.01). Figs. 2 and 3 demonstrate the pressures and flows generated by conditioned muscles over a 4 hour period against 80 mm Hg diastolic pressure. Although there was a gradual loss of pressure over time, the generated pressures were still physiologically significant after 4 hours of hydraulic testing.
o+------,--,-----,---,--~~--,----~
o
2
3
4
T I ME (hours)
Fig. 3. Flows generated by latissimus dorsi pouches over a 4 hour period. The muscles were subject to a vascular collateral delay and long-term electrical stimulation.
That an electrically conditioned pouch failed to produce flow against 80 mm Hg afterload did not necessarily mean that it would fail to produce flow against lower afterloads. Table II depicts the duration of any forward flow against varying afterloads for pouches constructed with unconditioned muscles. Even against a lower afterload, however, only one unconditioned muscle could produce flow for 4 hours. In contrast, four of six conditioned muscles still produced considerable flow against 80 mm Hg afterload after 4 hours. The initial volumes of pouches constructed from electrically conditioned muscle (25 ± 10 rnI) were similar to the initial volumes of pouches fashioned from unconditioned muscles (22 ± 5 rnI). At the completion of the fatigue testing, the final volumes of pouches from the conditioned (35.6 ± 15.7 rnI) and unconditioned (32 ± 14 rnI) muscles were again similar but were significantly higher than their initial volumes (p < 0.01). It is not possible to precisely determine the ejection fraction of the pouches. Initial and final volumes were determined at the pressure at which the preload was orginally set-24 mm Hg. During the hydraulic testing, when the afterload column was full, there was back leakage through the closed Bjork-Shiley valve, raising the filling pressure of the pouch to a mean of 34.6 ± I
Volume 91 Number 4 April, 1986
Hydraulic pouches of canine latissimus dorsi
539
Table II. Total length of time that a pouch constructed from unconditioned muscle could produce any flow against an afterload of 80, 70, or 60 mm Hg Duration of flow (min) against afterload 80 mm Hg Group I Group II Group III
3.2 ± 0.8 45.6 ± 86.4 39.3 ± 28.0
I
70 mm Hg
I
29.5 ± 45.5 27.0 ± 29.8 70.0 ± 88.9
mm Hg. The precise pouch volumes at this pressure are not known. However, the initial and final ejection fraction can be roughly estimated by dividing net forward flow per stroke by the initial and final volumes determined at 24 mm Hg. The mean initial ejection fraction of conditioned muscles (N = 6) calculated in this manner was 0.45 ± 0.27; the mean initial ejection fraction of unconditioned muscles (N = 7) was 0.78 ± 0.34. The difference was significant (p < 0.01). After 4 hours of stimulation, the mean ejection fraction of conditioned muscles (N = 4) was 0.12 ± 0.07. The mean weight of eight pouches of latissimus dorsi muscles was 72 ± 16 gm in comparison with the mean weight of the corresponding left ventricle, which was 62 ± 5 gm. Table III relates the duration of the pulse train to a conditioned muscle with the generated pressures and flows when the number of pulse trains per minute (48 beats/min) and the frequency within a pulse train (25 Hz) were held constant. As the duration of the pulse train increased, the flow increased. When the pulse train duration was held constant at 300 msec (Table IV) and the frequency of stimulation within a pulse train increased, the pressures and flows increased. This relationship also held true for another conditioned pouch tested in this manner. Tracings from a conditioned pouch that served as an aortic counterpulsator for 1 hour are depicted in Fig. 4. The pouch had previously been subjected to a 4 hour hydraulic test in the mock circulation system. The initial volume of blood pumped by the muscle pouch was 262 ml/rnin or 13.8% of cardiac output. After 1 hour the volume was 178 ml/rnin or 11% of cardiac output. All muscles electrically conditioned for a 6 week period were transformed into muscles composed predominately of slow-twitch fibers as determined histologically by characteristic ATPase stains with either acid or alkaline preincubation (Fig. 5). Discussion All muscle, including myocardium, is subject to fatigue depending upon the conditions under which it must work. The overall performance capacity of any
60 mm Hg
Total
54.5 ± 45.1 9.8 ± 6.3 17.5± 10.6
88.0 ± 77.6 95.3 ± 103.0 121 ± 104.0
Table m. Pressures and flows generated by a pouch constructed with latissimus muscle subject to vascular delay and electrical conditioning Duration of pulse train [msec}
Pouch pressure (mmHg)
Flow (mllbeat]
100 200 300 400 500
110/22 140/22 140/35 140/38 m/22
0 0 4.0 8.0 10.0
Legend: Afterload was ~o mrn Hg. Frequency within pulse train was held constant at 25 Hz. Number of ejections per minute was held constant at 4~. Duration of pulse train varied.
Table IV. Pressures and flows generated by a pouch constructed with latissimus muscle subject to a vascular delay and electrical conditioning Frequency of stimuli within pulse train
25 40 50 75
Hz Hz Hz Hz
Pouch pressure (mm Hgj
Flow (mllbeat]
140/35 158/25 158/38 160/38
4.0 7.8 9.1 10.3
Legend: Afterload was ~o mm Hg. Duration of pulse train was held constant at 300 rnscc. Number of ejections per minute was held constant at 48. Frequency of stimuli within pulse train varied.
given muscle is related to a number of variables, anyone of which can significantly affect muscle function. In the attempt to use skeletal muscle to augment cardiac output by creation of an auxiliary ventricle, two major problems are confronted: First, skeletal muscle is inherently more sensitive to fatigue than myocardium. Second, surgical manipulation, including interruption of collateral blood supply and the reshaping of muscle geometry, places further stresses on the limited performance capacity of skeletal muscle. The desire to maximize the function of skeletal muscle ventricles obviously requires methods to minimize factors that contribute to muscle fatigue. This report deals with our first attempt to prepare canine latissimus dorsi muscle with electrical
The Journal of Thoracic and Cardiovascular
540 Mannion, Hammond, Stephenson
Surgery
85-206
85-206
7 MIN OF COUNTERPULSATION
I HR OF COUNTERPULSATION
co
E
<,
160 ,-----..-r-----,---....,.....-------,
~---~--~--~
120
¥
80
E E w
40
+-_-+
~
0
.L.-_--'-
0:::
--"---_ _----'_ _-------' L -_ _---'--
.L.-
_
(f)
~::RJt=
- -
PA ~~
CVP
-
1--1 sec-
Fig. 4. Hemodynamic recordings when the muscle pouch was used for diastolic augmentation in the animal's own circulation. High filling pressures of the pouch are the result of a valveless connection with a systemic circulation. Initial pouch ejection was approximately 6.5 rnl/beat. After 1 hour, ejection was 4.5 ml/beat. Systemic blood pressure recording was obtained from femoral catheter. PA. Pulmonary artery. CVP. Central venous pressure.
and vascular conditioning to test its potential to assume a portion of left ventricular function. Latissimus dorsi muscle was selected for study in this experiment for several reasons. First, the muscle has a natural advantage with regard to electrical conditioning via the nerve because, unlike the rectus and pectoralis muscles, it is innervated by a single nerve, which makes it technically easier to condition. Second, as the data comparing muscle weight with left ventricular weight suggest, the latissimus has enough bulk to theoretically assume at least a portion of left ventricular function. Third, as evidenced in the long-term studies in this experiment, the muscle is noncritical, and its use for pouch construction minimally impairs the normal activities of the animal. Since the concept was introduced by Kantrowitz and Mckinnon' in 1959, many investigators have commented upon the potential of skeletal muscle to assist in cardiac function. Spotnitz, Merker, and Malm" studied the characteristics of rectus hydraulic pouches and noted pressure-volume curves that were reminiscent of Starling's law of the heart. Vachon, Kunov, and Zingg" observed a similar length-tension relationship with canine diaphragmatic pouches. Kusaba's group,' Nakamura and Glenn,' and Termet, Chalineau, and Estourin" demonstrated short-term hemodynamic benefits with cardiac assistance by skeletal muscle. Long-term studies
have also been performed documenting that skeletal muscle can be implanted into the atrium, right ventricle, and most recently left ventricle with resultant skeletal muscle viability.HII The long-term studies did not fully assess the contribution of the skeletal muscle patch to overall cardiac performance; the majority of shortterm studies emphasize that the major problem with skeletal muscle cardiac assistance has been muscle fatigue. One method to lessen the fatigue of auxiliary ventricles is to limit the demands placed upon the muscle. As pointed out by Dewar and associates," it is perhaps unrealistic for skeletal muscle to assume the entire function of the left ventricle. A skeletal muscle assist device that can generate 20% to 25% of cardiac output might be a more achievable goal. The use of skeletal muscle as an auxiliary ventricle, rather than as a replacement for a portion of myocardium, has several advantages: First, the procedure does not involve cardiopulmonary bypass and eliminates untoward effects on the heart. Second, the auxiliary ventricle is not required to contract in synchrony with each heartbeat and can be paced at a less demanding rate, such as 1: 2 or 1: 3. Third, this model permits a more easily quantifiable assessment of the work performed by skeletal muscle assistance, a point that demands careful scrutiny. Perhaps the most promising possibility for overcom-
Volume 91 Number 4 April, 1986
Hydraulic pouches of canine latissimus dorsi
54 1
Fig. 5. Myofibrillar ATPase stains with acid and alkaline preincubation of electrically conditioned and unconditioned muscles. Top row, Unconditioned muscles. Bottom row, Conditioned muscles. Left column, Acid preincubation . Right column, Alkaline preincubation. Note unconditioned muscles contain approximately 50% fast and 50% slow fibers. Virtually 100% of cells from conditioned muscle contain evidence of slow myosin (uniform dark staining of fibers), as determined by acid stains . Alkaline sta ins of conditioned muscle reveals residual fast myosin (darkly stained fibers) in about 20% of the cells conditioned at 2 Hz.
ing the fundamental difficulty with muscle fatigue lies in the ability of skeletal muscle to adopt to different patterns of use. Muscle plasticity has been most clearly demonstrated in rabbit muscle in response to long-term electrical stimulation.'>" After 6 weeks of electrical conditioning,a fast skeletal muscle is transformed into a muscle characterized by a slow contraction time, dense capillary network, large mitochondrial volume, and increased aerobic enzymatic capability. These changes render a muscle relatively resistant to fatigue, We IH- I9a have previouslyextended these observations on electrical muscle conditioning to the canine species and have transformed latissimus dorsi, rectus abdominis, pectoralis, and diaphragm through long-term electrical stimulation. In an isometric fatigue test developed for the latissimus dorsi muscle, we19a have demonstrated that electrical conditioning does impart to the muscle a relative resistance to fatigue, but also results in a decrease in muscle weight, fiber size, and isometric strength. The results of this experiment with the hydraulic pouch model are similar to the results observed with
our linear model. Just as conditioned latissimus generated lower initial isometric tensions but with minimal fatigue rates in the linear model in comparison with their unconditioned counterparts, so auxiliary ventricles constructed of conditioned muscle generated lower initial pressures than comparable unconditioned pouches (Groups II and IV), but were able to sustain them for longer periods. The fact that conditioned pouches could generate significant physiologic pressures with filling volumes of 25 rnl suggests that a diminution in muscle fiber size and a decrease in isometric strength that accompany a muscle's adaptation to a relatively fatigue-resistant state does not necessarily render a muscle less capable of cardiac assistance. The optimal method to electrically precondition a muscle has not yet been determined and may, in fact, be a moot point. The muscle will ultimately adapt to how it is used. A muscle conditioned at 2 Hz for 6 weeks may undergo a further series of changes if it is used permanently for diastolic augmentation. For instance, the residual fast myosin seen in 20% to 25% of fibers
542
The Journal of Thoracic and Cardiovascular Surgery
Mannion, Hammond, Stephenson
after 6 weeks of electrical stimulation with single pulses at 2 Hz might disappear with stimulation with a burst pattern, which is required for diastolic augmentation. We* have demonstrated that diaphragm conditioned for a I year period is different in some respects from diaphragm conditioned for 6 weeks. The present value of electrical conditioning is that muscle conditioned by any pattern of long-term stimulation might give a more realistic appraisal of the ultimate ability of skeletal muscle to assist in cardiac function. Several authors have commented upon the possible adverse effects that interruption of blood supply might have upon muscle function.v' We 19b have previously shown that electrically unconditioned latissimus is especially sensitive to interruption of collateral supply. In the same experiment we have also noted that a delay of exercise testing for a 3 week period after collateral interruption resulted in a considerable recovery of the normal exercise-induced increase in unconditioned muscle blood flow. The vascular delay alone used in muscle preparation in Groups II and III was an effort to see if the muscle pouch function could be improved by delaying exercise several weeks after interruption of its collateral blood supply. However, the delay period between interruption of the chest wall collaterals and creation of the hydraulic pouch in Group II resulted in only minimal improvement in fatigue rates. This theoretically could have resulted from one of several reasons: First, all of the collaterals were not interrupted. Second, newly formed collaterals were again interrupted during pouch construction. Third, other factors such as fiber type play a predominant role in pouch fatigue. The observation that two of three animals in which the pouch was made and left in situ were able to generate flow against 80 mm Hg for approximately I hour may suggest that this method of vascular delay may result in more optimal pouch function. Further studies may clarify possible improvements in pouch function with creation of permanent pouches. The proper method to construct a pouch of skeletal muscle remains to be determined. Nature has provided two examples, the right and left ventricles." The right ventricle is basically a bellows pump, with contraction of the free wall against a relatively immobile septum. This mechanism provides for a highly efficient volume pump, but it is useful only at low pressures. The left ventricle is a muscular cone, with the thickest part of the wall at the 'Acker M, Mannion JD, Brown W, Salmons S, Henriksson J, Bitto T, Gale D, Hammond R, Stephenson L: Canine diaphragm musele after one year of continuous electrical stimulation at 2 Hz and 4 Hz, Its potential as a myocardial substitute, (Submitted for publication.)
widest part of the ventricle and the thinnest at the apex, Ejection of blood is accomplished with circumferential shortening around the short axis. The pump is useful for generating high pressures. The multiple wrap technique was an attempt to increase wall thickness and permit an auxiliary ventricle to assume a portion of left ventricular function. The contour of the wrap-a conical spiralwas an effort to simulate the configuration of the left ventricle, a geometry that was unraveled by Pettigrew. 2 1, 22 We are not certain what advantage this multiple-layer technique provides to the latissimus pouch. Indeed, there may be a disadvantage, in that the inner distal muscle may become relatively ischemic from high intraventricular pressures. Further study might clarify this point. The hydraulic system used in this experiment is not a replica of systemic circulation. It has an infinite compliance, low resistance, and high inertial load. Flows generated in this model would not necessarily be comparable to flows generated over time in an in vivo circulation. However, the system does give some indication of how an auxiliary ventricle might perform against a physiologic pressure load. The observation that one auxiliary ventricle was able to generate significant flow for I hour against systemic resistance holds promise for future investigation in this area. We are not yet certain if the dilatation observed during the hydraulic fatigue testing with the auxiliary ventricle reflects intrinsic muscle failure or unraveling of the muscle pouches. What is certain is that all factors that contribute to fatigue have not yet been determined, let alone eliminated. Further work on electrical muscle conditioning, optimization of muscle blood supply, configuration of auxiliary ventricle, and new modes to harness the energy of skeletal muscle are only a few of the items that need to be investigated in the effort to use skeletal muscle for cardiac circulatory assistance.
REFERENCES Franciosa JA, Wilen M, Zuseke S, Cohn IN: Survival in men with severe chronic left ventricular failure due to either coronary heart disease or idiopathic dilated cardiomyopathy. Am J Cardiol 51:831-836, 1983 2 Kusaba E, Stule W, Sawatoni S, Jaron D, Freed P, Kantrowitz A: A diaprhagmatic graft for augmenting left ventricular function, Trans Am Soc Artif Intern Organs 19:251-257,1973
3 vonRecum A, Stirke JP, Hamada 0, Saba H, Kantrowitz A: Long-term stimulation of a diaphragm muscle pouch, J Surg Res 23:422-427, 1977 4 SpotnitzHM, Merker L, MaImJR: Applied physiology of the canine rectus abdominis. Force-length curves corre-
Volume 91 Number 4 April, 1986
lated with functional characteristic of a rectus powered "ventricle." Potential for cardiac assistance. trans Am Soc Artif Intern Organs 20:747-756, 1974 5 Kantrowtiz A, McKinnon WMP: The experimental use of the diaphragm as an auxiliary myocardium. Surg Forum 9:266-268, 1959 6 Vachon BR, Kunov H, Zingg W: Mechanical properties of diaphragm muscle in dogs. Med Bioi Eng Com put 13:252260, 1975 7 Nakamura K, Glenn W: Graft of the diaphragm as a functioning substitute for the myocardium. An experimental study. J Surg Res 4:435-439, 1964 8 Termet H, Chalencon JL, Estour F: Transplantation sur Ie myocarde d'un muscle strie exciete par pacemaker. Ann Chir Thorac Cardiovasc 5:260-263, 1966 9 Macoviak JA, Stephenson LW, Kelly A, Likoff M, Reichek N, Edmunds LH Jr: Partial replacement of the right ventricle with a synchronously contracting diaphragmatic skeletal muscle autograft. Proceedings of the Third Meeting of the International Society for Artificial Organs 5:Suppl 1:550-555, 1981 10 Macoviak JA, Stephenson LW, Alavi A, Kelly AM, Edmunds LH Jr: Effects of electrical stimulation on diaphragmatic muscle used to enlarge right ventricle. Surgery 90:271-277,1981 II Sola OM, Dillard DH, Ivey TD, Haneda K, Itoh T, Thomas R: Autotransplantation of skeletal muscle into myocardium. Circulation 71:Suppl 2:341-348, 1985 12 Dewar ML, Drinkwater DC, Wittnich C, Chiu RC: Synchronously stimulated skeletal muscle graft for myocardial repair. J THORAC CARDIOVASC SURG 87:325-331, 1984 13 Salmons S, Sreter FA: Significance of impulse activity in the transformation of skeletal muscle type. Nature 263:30-34, 1967 14 Salmons S, Vrbova G: The influence of activity on some contractile characteristics of mammalian fast and slow muscle. J Physiol 201:535-549, 1969 15 Salmons S, Henriksson J: The adaptive response of skeletal muscle to increased use. Muscle Nerve 4:94-105, 1981 16 Cotter MA, Hudlicka 0, Pette D, Staudte HW, Vrobva G: Changes of capillary density and enzyme pattern in fast rabbit muscles during long-term stimulation. J Physiol 230:34-35, 1973 17 Hudlicka 0, Brown M, Cotter M, Smith M, Vrobva G: The effect of long-term stimulation of fast muscles on their blood flow, metabolism and ability to withstand fatigue. Pflugers Arch 369:141-149, 1977 18 Bitto T, Mannion JD, Hammond R, Macoviak lA, Rashkind WJ, Edmunds LH Jr, Stephenson LW: Pectoralis and rectus abdominis for potential correction of congenital heart defects, Pediatric Cardiology, Proceedings of the Second World Congress of Pediatric Cardiology, New York, 1986, Springer-Verlag, pp 609-612 19 Armenti FR, Bitto T, Macoviak JA, Kelly AM, Chase CT, Hoffman BK, Rubenstein NA, St. John Sutton M,
Hydraulic pouches of canine latissimus dorsi
54 3
Edmunds LH Jr, Stephenson LW: Transformation of skeletal muscle for cardiac replacement. Surg Forum 35:258- 260, 1984 19a Mannion JD, Bitto T, Hammond R, Rubenstein NA, Stephenson LW: Histochemical and fatigue characteristics of conditioned canine latissimus dorsi muscle. Circ. Res. 58:298-304, 1986 19b Mannion JD, Velchek M, Alavi A, Stephenson LW: Blood flow in conditioned and unconditioned latissimus dorsi muscle (abstr). Second Vienna Muscle Symposium, 1985, p 28 20 Parmley WW, Talbot L: Heart as a Pump, Handbook of Physiology, Vol I, The Heart, Section 2, The Cardiovascular System, R Berne, ed., Bethesda, 1979, Waverly Press, pp 429-460 21 Streeter DD: Gross morphology and fiber geometry of the heart, Handbook of Physiology, Vol I, The Heart, Section 2, The Cardiovascular System, R Berne, ed., Bethesda, 1979, Waverly Press, pp 61-112 22 Pettigrew J: On the arrangement of the muscular fibers of the ventricular portion of the heart of the mammal. (Croonian Lecture, Abstract). Proc R Soc Land 10:433440, 1860 23 Macoviak JA, Stephenson LW, Spielman S, Greenspan A, Likoff M, St. John Sutton M, Reichek N, Rashkind W, Edmunds LH Jr: Replacement of ventricular myocardium with diaphragmatic skeletal muscle. J THORAC CARDlOVASC SURG 81:519-527,1981
Discussion DR. TOM D. IVEY Seattle, Wash.
The authors have shown that in preconditioned pouches physiologic parameters can be obtained in an in vitro preparation. They also have shown that fatigue is a factor in this preparation. I would like to briefly review a slightly different approach that Dr. David Dillard and Dr. Sola in our laboratory at the University of Washington have used to tackle a similar problem. Our first attempts to use skeletal muscle as a pumping fiber involved the sternohyoid muscle of the dog. These studies showed that when the sternohyoid muscle is divided, viable muscle fibers can be maintained in the myocardial position over a long period. We next turned to the sternocleidomastoid muscle, and these studies showed that the pedicle, when divided after long-term implantation in the left ventricle, would bleed rather briskly in the retrograde fashion, more briskly, in fact, than it would in the antegrade fashion. However, use of this preparation also turned us to a significant problem, that of fibrosis and clotting at the muscle-blood barrier. Our later attempts to tackle this problem, like the authors', involved constructing a permanent pedicle from latissimus dorsi muscle. Our current preparation using the latissimus dorsi pedicle is similar to that used by the authors, in that the pedicle is lifted, mobilized, and then prestimulated in a stretched position according to the following protocol: 2 weeks
544 Mannion, Hammond. Stephenson
of pacing at 30 beats/min, 2 weeks at 60 beats/min, and finally I month at 100 beats/min. These slides depict the hypertrophy that we have seen in the pedicle with a section taken before stretching and stimulation versus another section showing marked hypertrophy and, as the authors have described, conversion to slow-twitch fibers. The fibers in these pedicles are rather healthy and have taken on a rather hypertrophied appearance. We then tackled the problem of fibrosis and thrombosis at the muscle-blood barrier by implanting into the left ventricle without full-thickness excision of the muscle, leaving a thin layer of endocardium. In our preparation, this is roughly equivalent to a 25% left ventricular infarction. We then used synchronous pacing in the VVD mode. Currently our short-term preparations have shown no hemodynamic improvement. Our long-term preparations await study by both sonomicrometry and ventriculography, and we also are working on a right ventricular preparation. I would like to ask Dr. Mannion two questions. In the manuscript he showed that even at 4 hours his preparations did show fatigue, I would like to know what he plans in the future to overcome this problem. Second, how would he modify his preparation to overcome the problem that we have seen, that being encasement of the muscle fibers in fibrous tissue when exposed to the blood barrier?
The Journal of Thoracic and Cardiovascular Surgery
DR. MANNION (Closing) Dr. Ivey, thank you for your perceptive comments. It is true that we have identified an element of fatigue in these pouches, and at this point we do not believe that we have even identified all of the causes that contribute to muscle fatigue. Blood supply contributes to it, fiber type, pH, glycogen content of the muscles, and many other factors that we are not even sure of yet contribute to muscle fatigue. Once we identify more of these factors, we can eliminate more of the causes for fatigue. One of the problems with our model, as Dr. Ivey pointed out, is that it is subject to the same difficulties inherent in the artificial heart program. We have to use a biologic membrane inside the ventricle and then face the attendant problems of embolism and stroke. We do not necessarily know yet that this is the best approach to take. The value of this particular model is that it is easy to quantitate the amount of cardiac type work that a skeletal muscle can do. In past studies at the University of Pennsylvania in which we have placed pedicle flaps onto the heart, we have had difficulty determining exactly what percentage of the cardiac output is caused by the contraction of the skeletal muscle. We think if we can demonstrate with this model that the skeletal muscle can do significant amounts of cardiac type work for prolonged periods, the study will be a success.
THoRAc CARDIOVASC SURG
91:534-544, 1986
Hydraulic pouches of canine latissimus dorsi Potential for left ventricular assistance We have studied the fatigue rates of hydraulic pouches constructed in the form of a multilayered conical spiral using the latissimus dorsi muscle of 17 beagles. The roles that electrical muscle conditioning and early interruption of collateral blood supply have in the prevention of pouch fatigue were evaluated. The length of time that a pouch could generate flow in a hydraulic test system was measured; afterload was set at 80 mm Hg and preload 24 mm Hg. Pouches (N = 3) fashioned from muscles subject to neither electrical conditioning nor a vascular delay generated an initial flow of 990 ± 346 ml/min, but could sustain flow for only 2.3, 3.8 and 3.6 minutes. Pouches (N = 5) constructed with electrically unconditioned muscles after a vascular delay (median 3 weeks) demonstrated a variable improvement in fatigue rates (initial flow 826 ± 265 m1/min; time to no forward flow, 2.5, 7.5, 7.5, 10, and 200 minutes). Four of six pouches that received the benefit of long-term electrical muscle conditioning and a vascular delay (N = 6) were able to generate flow for a 4 hour period, at which time the experiment was terminated (initial flow 478 ± 204 m1/min; final flow 195 ± 157 ml/min), After the 4 hour fatigue test was completed, one electricaUy conditioned pouch was placed in series with the heart and served as a counterpulsator. The initial volume of blood pumped by the muscle pouch was 262 ml/min or 13.8 % of cardiac output. After the pouch had contracted at a rate of approximately 45 beats/min for 1 hour, the volumeof blood pumped was 178 m1/min, or 11 % of cardiac output In three other animals a pouch was fashioned and then left in situ for a 1 to 3 week period before hydraulic testing. These pouches generated significant initial flows (390 ± 60 m1/min), which demonstrates the feasibility of further study of permanent pouches. These results suggest that permanent electrical muscle conditioning and perhaps a vascular collateral delay might permit an auxiliary skeletal muscle-powered ventricle to assume a portion of left ventricular function.
John D. Mannion, M.D., Robert Hammond, B.S., and Larry W. Stephenson, M.D.,
Philadelphia. Pa.
h e 1 year mortality for the estimated 10,000 patients who present annually in the United States with irreversible congestive heart failure is approximately 50%.1 This high mortality persists despite significant advances made in medical therapy, cardiac transplantation, and the artificial heart program. Since none of these three presently available treatments for severe cardiac pump failure can be expected to benefit a large number of From the Harrison Department or Surgical Research and The Department of Cardiothoracic Surgery. Hospital of the University of Pennsylvania. Philadelphia. Pa. Supported by National Institutes of Health Grant HLBI 27570. Read at the Eleventh Annual Meeting or The Western Thoracic Surgical Association. Incline Village, Nev .. June 16-20. 1985. Address for reprints: John D. Mannion. M.D., 313 Medical Education Building, University or Pennsylvania. Philadelphia. Pa. 19104.
534
patients in the near future, we have been investigating a potential fourth therapeutic alternative: the use of a patient's own skeletal muscle to augment myocardial performance. A functioning skeletal muscle-powered ventricle would not be subject to tissue rejection, would not be hampered by donor shortage, and would obviate the necessity for cumbersome and expensive external power sources. Although theoretically appealing, skeletal musclepowered cardiac assistance has been unworkable thus far because of muscle fatigue.l' Skeletal muscle fiber type, surgical trauma, interruption of collateral blood supply, stimulation rates, pressure versus volume work, and muscle mass-to-ventricular volume ratio are only a few of the variables that have been related to the fatigability of skeletal muscle-powered ventricles. Strategies are needed that would enhance the endurance of
Volume 91 Number 4 April, 1986
auxiliary ventricles by eliminating some of the factors that lead to fatigue. In this experiment three unique approaches were taken. First, to minimize the acute effects of ligation of collateral blood supply, a several week delay was introduced between the interruption of major collateral vessels and the exercise testing of muscles. Second, canine latissimus dorsi muscle was electrically conditioned at a frequency of 2 Hz for 6 weeks, in an attempt to transform fatigue-sensitive, type II fibers into fatigue-resistant, type I fibers. Third, to increase the ratio of wall thickness to muscle pouch volume and thereby lower wall stress, pouches were constructed of latissimus dorsi muscle wrapped in the form of a multilayered conical spiral. This experiment was designed to determine if any combination of electrical conditioning, vascular delay, and a multilayered pouch would minimize muscle fatigue and permit an auxiliary ventricle to assume a portion of left ventricular function. Methods and materials Seventeen adult male beagles, with a mean weight of 13.0 ± 2.1 kg, were studied. Latissimus dorsi pouches were constructed and were subjected to a hydraulic fatigue test, which will be described subsequently. The animals were divided into four groups. In Group I (N = 3) there was no preparation of the latissimus dorsi muscle before the construction of the muscle pouch and the hydraulic test. In Group II (N = 5) the multiple segmental collateral blood vessels to the distal latissimus, arising from the intercostal arteries, were divided, and the muscles were rested for a median of 3 weeks before pouch construction and hydraulic fatigue testing. In Group III (N = 3) a permanent pouch was evaluated. The latissimus was mobilized and a pouch fashioned and left in situ for a 1 to 3 week period. In Group IV (N = 6) the latissimus dorsi muscle was prepared by division of chest wall collaterals and a rest period of 3 weeks, as in Group II. The muscle was then continuously stimulated via the thoracodorsal nerve for 6 weeks before pouch construction. Procedures for vascular delay and electrical conditioning. In Groups II, III, and IV, in which the latissimus muscles were subjected to either vascular collateral delay or electrical conditioning, an initial sterile procedure was performed. The animals were anesthetized, intubated, and draped in a sterile fashion. Preoperative and intraoperative antibiotic prophylaxis with cefazolin sodium (Ancef, 250 mg) was given intravenously. In Group II division of intercostal collaterals necessitated reflection of the muscle from the chest
Hydraulic pouches of canine latissimus dorsi
535
wall, extending from the tip of the scapula to the insertion of the latissimus into the thoracolumbar fascia. The muscle itself was not mobilized, and the division of vessels did not include the collaterals from the trapezius or teres major, but did include the major serratus collateral. The major axial blood supply to the latissimus-the thoracodorsal artery-was not disturbed. The wound was then closed, and the pouch was constructed 3 weeks later, immediately before fatigue testing. In Group III the entire latissimus was mobilized for construction of an in situ pouch. The intercostal collaterals were divided and a plane was developed between the platysma and latissimus. The thoracolumbar fascia was divided, and the muscle was mobilized proximally. The collateral vessels between the trapezius and the latissimus were then interrupted, and the muscle was separated from the teres major. The insertion of the latissimus into the humerus and the thoracodorsal neurovascular pedicle was left intact. With this method it was possible to mobilize the entire latissimus without dividing any muscle fibers. A modified Medtronic pacing electrode was placed around the thoracodorsal nerve to later facilitate hydraulic fatigue testing. The muscles were then wrapped around a 30 ml Teflon cone, the pouch approximated to the chest wall, and the subcutaneous tissue and skin closed over the muscle pouch. Hydraulic fatigue testing was performed 1, 3, and 3 weeks later. Group IV muscles received vascular and electrical conditioning. The intercostal collaterals were divided, as in Group II. A modified Medtronic electrode was then placed around the thoracodorsal nerve and attached to a Medtronic pulse generator, Model 5984 or 8422. The pulse generator was placed in a pocket fashioned beneath the right rectus muscle. The pulse generatorlead system was used to electrically condition Group IV muscles by stimulating the thoracodorsal nerve at a continuous frequency of 2 Hz for a 6 week period. The pulse generator was set to one-half amplitude (3.2 volts) and the pulse width to 220 1J.Sec. At the end of the 6 week period of stimulation, the pouch was constructed and the hydraulic fatigue test performed. The Medtronic electrodes were modified in the following manner. The platinum ring and coil wire lead of a Medtronic Model 6904 bipolar endocardial lead was removed. The ring was fashioned into a half cylinder of 2 to 3 mm radius and placed in a silicone rubber cuff. The coil wire was then reinsulated with silicone rubber tubing. Hydraulic pouch construction. The latissimus dorsi muscle is irregularly shaped but roughly corresponds to a broad-based triangle. The posterior aspect of the
The Journal of Thoracic and Cardiovascular
5 3 6 Mannion, Hammond, Stephenson
Surgery
"1.5 - 2 more \
t urn s
Fig. 1. Example of construction of latissimus multilayered ventricle in form of conical spiral. Representation of hydraulic test system. Preload and afterload was adjustable . A Bjork-Shiley valve was inserted in each limb of the Y.
muscle is thicker than the anterior. A pouch was constructed by wrapping the muscle around a specially prepared condom so that the muscle assumed the form of a conical spiral. An attempt was made to have the thickest portion of the muscle correspond with the widest portion of the skeletal muscle ventricle (Fig. I). The condom received one and one half to two wraps of muscle. The proximal thoracodorsal vessels were positioned on the outside of the muscle wrap. The condom was prepared in the following manner. A 1 em excessivelength of condom was passed over a 5 em multiply perforated 3/8 inch inner diameter (10) silicone tube, which was stretched over a Y.2 inch 10 polycarbonate connector. The connector was passed through an ovoid of low-porosity Teflon felt (USCI, Teflon felt, (07839), approximately 2 by 3 em, which served as the base of the auxiliary ventricle. The latissimus muscle layers were approximated to the Teflon felt with 2-0 Dexon sutures. The connector was made watertight and attached to the hydraulic test circuit. Hydraulic test circuit. The hydraulic system consisted of an adjustable inlet reservoir and outlet column (Fig. 1). The pouch served as the zero reference level. The height of the inlet reservoir was adjusted so that the pouch was exposed to 24 mm Hg filling pressure. The height of the afterload column was adjusted so that the pouch had to contract against a desired pressure. During pouch filling, a 65/35 saline/glycerin mixture would flow from a 2 L reservoir, through a Travenol miniprime heat exchanger, and through an 18 mm Bjork-Shiley prosthetic aortic heart valve proximal to a Y-connector and into the pouch. With contraction of the auxiliary ventricle, the inlet valve would close and a 16 mm
Bjork-Shiley valve on the other side of the Y-connector would open. At the height of the column the flow was diverted back into a catch funnel for measurement and then into the inlet reservoir. Pouch and afterload pressures were recorded. Experimental protocol. After the pouch was constructed, the preload was set to 24 mm Hg and the pouch volume was measured. Afterload was set to 80 mm Hg. The pouch was then made to contract at a rate of 48 beats/min. This was accomplished by stimulating the thoracodorsal nerve at a frequency of 25 Hz for 300 msec, followed by 950 msec of rest. The voltage used provided a supramaximal stimulus of approximately 6 volts. The pulse width of the individual square-wave pulses was 220 lJSec. Pouch pressures were monitored continuously and flows were recorded every 10 minutes. The pouch was stimulated at this rate until it could no longer produce net flow over the column. In those pouches that continued to produce forward flow, the experiment was terminated at the end of 4 hours. In those pouches unable to maintain forward flow, the column was lowered to 70 and then 60 mm Hg, and the time to cessation of forward flow was recorded; At the termination of the procedure, pouch volumes at 24 mm Hg preload pressure were again determined. Body temperature was maintained with a heating blanket, and percent oxygen saturation was kept greater than 90% with supplemental oxygen. Throughout the experiment, the ventilator was adjusted to maintain pH between 7.35 and 7.45. Serial hematocrit values were monitored. An infusion of 5% dextrose in normal saline was given, and infusion volume was varied according to the central venous pressure and cardiac output.
Volume 91 Number 4
Hydraulic pouches of canine latissimus dorsi 5 3 7
April. 1986
Table I. Systolic pressures and flows of pouch when contracting against a glycerin/saline column producing 80 mm Hg of static diastolic pressure Pmax (mm Hg) Initial Group I (N = 3) Unconditioned muscle No vascular delay Acute pouch construction Group II (N = 5) Unconditioned muscle Vascular delay Acute pouch construction Group III (N = 3) Unconditioned muscle Pouch construction Pouch in situ for 3 weeks Group IV (N = 6) Conditioned muscle Vascular delay Acute pouch construction
Flow (mljmin)
I
Final
Duration offlow (min)
Final
Initial
228 ± 28
121 ± 13
990 ± 346
o
2.3, 3.8, 3.6
279 ± 42
134 ± 26
826 ± 265
o
7.5, 10, 2.5, 7.5., 20
160 ± 20
108 ± 23
390 ± 60
o
7, 55,67
214 ± 46
ISS ± 24
478 ± 204
195 ± 157
I
240 (N = 4) 230(N=I)* 120 (N = nt
Legend: Forty-eight pouch contractions per minute were generated by stimulation of the thoracodorsal nerve at 25 Hz frequency for 300 msec followed by 950 msec of rest. Depending on heart rate, 300 msec is the approximate length of diastole in canine species. Voltage used-c-o volts;duration of individual pulses-220 !JSCc. Duration of flow refers to total time that a pouch could produce any flow against 80 mm Hg diastolic afterload. *Experiment stopped when flow reached 30 rnl/rnin. tExperiment terminated when balloon condum ruptured.
After the 4 hour hydraulic test was accomplished in two electrically conditioned pouches, the duration of the pulse train and the frequency of stimulation within a pulse train were varied to establish the relationship between these variables and the generated pressures and flows. In one electrically conditioned animal, at the completion of the 4 hour hydraulic testing, the auxiliary ventricle was placed in series with the heart and served as a counterpulsator. A JA inch ID T system was placed in the distal thoracic aorta and connected to the pouch with a t/2 inch ID silicone tubing. The pouch was stimulated at 25 Hz for 300 msec during diastole at a rate of I :4 or 1: 3 in relation to the animal's heart rate. Synchronization was accomplished by monitoring the electrocardiogram on the amplifier section of an AVCO intra-aortic balloon pump. The R waves were then COunted with an IC4017 decade counterdivider that could provide 1: 1 through 1: 9 synchronization of the stimulator. Flow to and from the pouch was recorded with a Biotronix Model 610 electromagnetic flowmeter, which was calibrated before the experiment. Areas under the flow curves were integrated on a digitizing board connected to an Apple lIe microcomputer, permitting calculation of the volume of blood pumped per COntraction of the muscle pouch. Systemic arterial pressures were recorded through femoral and carotid arterial catheters.
All animals received humane care in compliance with the "Principles of Laboratory Animal Care" formulated by the National Society for Medical Research and the "Guide for the Care and Use of Laboratory Animals" prepared by the National Academy of Sciences. Statistical analysis. Results were expressed as mean ± standard deviation of the mean. Paired and unpaired t tests were used to determine differences between groups, where appropriate. Histologic features. Muscle sections were taken and frozen in liquid nitrogen. Biopsy specimens were prepared with myofibrillar adenosine triphosphatase (ATPase) stains with acid (4.2) and alkaline (lOA) preincubation. Myofibrillar ATPase stains were used to determine the degree of transformation of fast-twitch, fatigue-sensitive fibers into slow-twitch, fatigue-resistant fibers. The slides were reviewed and representative muscle bundles were evaluated on both the acid and alkaline stains. Cells were identified as fast or slow according to Brook's ATPase classification. In conditioned muscles with acid stains, a cell was counted as a slow-twitch type if it stained darkly; with alkaline stains, a cell was counted as a fast-twitch type if it stained darkly. Sections stained with hematoxylin and eosin were reviewed for muscle structure. After the biopsies, the muscle pouches were placed into formalin. All fat was later removed. Muscle weight was determined by adding the biopsy weight to the
The Journal of Thoracic and Cardiovascular
538 Mannion, Hammond, Stephenson
Surgery
I
250-
500-
200-
PEAK
400
0>
::r:
E E W 0:: ::J
(f) (f)
c
150
E
<,
300-
u u
3
100
0
--l l.L
W 0:: 0-
FILLING
50
o
200
100
o
2
3
4
TIME (hours)
Fig. 2. Pouch pressures overa 4 hour period generated by latissimus dorsi pouches constructed of muscle that was subject to a vascular collateral delay and long-term electrical conditioning. Pouches were stimulated at a rate of 48 beats/ min.
weight of the formalin-fixed tissue. The latissimus dorsi mass was then compared with the mass of the left ventricle of the same animals. Results
The performance of the pouches in the hydraulic system against the diastolic afterload of 80 mm Hg is depicted in Table I. For the muscles that were prepared with neither vascular collateral delay nor electrical conditioning, flow ceased in 2 to 4 minutes. A vascular collateral delay alone seemed to enhance the performance of electrically unconditioned muscles, but the results were not striking. Only pouches constructed of electrically conditioned muscles were capable of generating flow after 4 hours of testing. When compared to all electrically unconditioned muscles or to electrically unconditioned muscles with a vascular delay, the prolongation in flow for the electrically conditioned pouches is significant (p < 0.01). Figs. 2 and 3 demonstrate the pressures and flows generated by conditioned muscles over a 4 hour period against 80 mm Hg diastolic pressure. Although there was a gradual loss of pressure over time, the generated pressures were still physiologically significant after 4 hours of hydraulic testing.
o+------,--,-----,---,--~~--,----~
o
2
3
4
T I ME (hours)
Fig. 3. Flows generated by latissimus dorsi pouches over a 4 hour period. The muscles were subject to a vascular collateral delay and long-term electrical stimulation.
That an electrically conditioned pouch failed to produce flow against 80 mm Hg afterload did not necessarily mean that it would fail to produce flow against lower afterloads. Table II depicts the duration of any forward flow against varying afterloads for pouches constructed with unconditioned muscles. Even against a lower afterload, however, only one unconditioned muscle could produce flow for 4 hours. In contrast, four of six conditioned muscles still produced considerable flow against 80 mm Hg afterload after 4 hours. The initial volumes of pouches constructed from electrically conditioned muscle (25 ± 10 rnI) were similar to the initial volumes of pouches fashioned from unconditioned muscles (22 ± 5 rnI). At the completion of the fatigue testing, the final volumes of pouches from the conditioned (35.6 ± 15.7 rnI) and unconditioned (32 ± 14 rnI) muscles were again similar but were significantly higher than their initial volumes (p < 0.01). It is not possible to precisely determine the ejection fraction of the pouches. Initial and final volumes were determined at the pressure at which the preload was orginally set-24 mm Hg. During the hydraulic testing, when the afterload column was full, there was back leakage through the closed Bjork-Shiley valve, raising the filling pressure of the pouch to a mean of 34.6 ± I
Volume 91 Number 4 April, 1986
Hydraulic pouches of canine latissimus dorsi
539
Table II. Total length of time that a pouch constructed from unconditioned muscle could produce any flow against an afterload of 80, 70, or 60 mm Hg Duration of flow (min) against afterload 80 mm Hg Group I Group II Group III
3.2 ± 0.8 45.6 ± 86.4 39.3 ± 28.0
I
70 mm Hg
I
29.5 ± 45.5 27.0 ± 29.8 70.0 ± 88.9
mm Hg. The precise pouch volumes at this pressure are not known. However, the initial and final ejection fraction can be roughly estimated by dividing net forward flow per stroke by the initial and final volumes determined at 24 mm Hg. The mean initial ejection fraction of conditioned muscles (N = 6) calculated in this manner was 0.45 ± 0.27; the mean initial ejection fraction of unconditioned muscles (N = 7) was 0.78 ± 0.34. The difference was significant (p < 0.01). After 4 hours of stimulation, the mean ejection fraction of conditioned muscles (N = 4) was 0.12 ± 0.07. The mean weight of eight pouches of latissimus dorsi muscles was 72 ± 16 gm in comparison with the mean weight of the corresponding left ventricle, which was 62 ± 5 gm. Table III relates the duration of the pulse train to a conditioned muscle with the generated pressures and flows when the number of pulse trains per minute (48 beats/min) and the frequency within a pulse train (25 Hz) were held constant. As the duration of the pulse train increased, the flow increased. When the pulse train duration was held constant at 300 msec (Table IV) and the frequency of stimulation within a pulse train increased, the pressures and flows increased. This relationship also held true for another conditioned pouch tested in this manner. Tracings from a conditioned pouch that served as an aortic counterpulsator for 1 hour are depicted in Fig. 4. The pouch had previously been subjected to a 4 hour hydraulic test in the mock circulation system. The initial volume of blood pumped by the muscle pouch was 262 ml/rnin or 13.8% of cardiac output. After 1 hour the volume was 178 ml/rnin or 11% of cardiac output. All muscles electrically conditioned for a 6 week period were transformed into muscles composed predominately of slow-twitch fibers as determined histologically by characteristic ATPase stains with either acid or alkaline preincubation (Fig. 5). Discussion All muscle, including myocardium, is subject to fatigue depending upon the conditions under which it must work. The overall performance capacity of any
60 mm Hg
Total
54.5 ± 45.1 9.8 ± 6.3 17.5± 10.6
88.0 ± 77.6 95.3 ± 103.0 121 ± 104.0
Table m. Pressures and flows generated by a pouch constructed with latissimus muscle subject to vascular delay and electrical conditioning Duration of pulse train [msec}
Pouch pressure (mmHg)
Flow (mllbeat]
100 200 300 400 500
110/22 140/22 140/35 140/38 m/22
0 0 4.0 8.0 10.0
Legend: Afterload was ~o mrn Hg. Frequency within pulse train was held constant at 25 Hz. Number of ejections per minute was held constant at 4~. Duration of pulse train varied.
Table IV. Pressures and flows generated by a pouch constructed with latissimus muscle subject to a vascular delay and electrical conditioning Frequency of stimuli within pulse train
25 40 50 75
Hz Hz Hz Hz
Pouch pressure (mm Hgj
Flow (mllbeat]
140/35 158/25 158/38 160/38
4.0 7.8 9.1 10.3
Legend: Afterload was ~o mm Hg. Duration of pulse train was held constant at 300 rnscc. Number of ejections per minute was held constant at 48. Frequency of stimuli within pulse train varied.
given muscle is related to a number of variables, anyone of which can significantly affect muscle function. In the attempt to use skeletal muscle to augment cardiac output by creation of an auxiliary ventricle, two major problems are confronted: First, skeletal muscle is inherently more sensitive to fatigue than myocardium. Second, surgical manipulation, including interruption of collateral blood supply and the reshaping of muscle geometry, places further stresses on the limited performance capacity of skeletal muscle. The desire to maximize the function of skeletal muscle ventricles obviously requires methods to minimize factors that contribute to muscle fatigue. This report deals with our first attempt to prepare canine latissimus dorsi muscle with electrical
The Journal of Thoracic and Cardiovascular
540 Mannion, Hammond, Stephenson
Surgery
85-206
85-206
7 MIN OF COUNTERPULSATION
I HR OF COUNTERPULSATION
co
E
<,
160 ,-----..-r-----,---....,.....-------,
~---~--~--~
120
¥
80
E E w
40
+-_-+
~
0
.L.-_--'-
0:::
--"---_ _----'_ _-------' L -_ _---'--
.L.-
_
(f)
~::RJt=
- -
PA ~~
CVP
-
1--1 sec-
Fig. 4. Hemodynamic recordings when the muscle pouch was used for diastolic augmentation in the animal's own circulation. High filling pressures of the pouch are the result of a valveless connection with a systemic circulation. Initial pouch ejection was approximately 6.5 rnl/beat. After 1 hour, ejection was 4.5 ml/beat. Systemic blood pressure recording was obtained from femoral catheter. PA. Pulmonary artery. CVP. Central venous pressure.
and vascular conditioning to test its potential to assume a portion of left ventricular function. Latissimus dorsi muscle was selected for study in this experiment for several reasons. First, the muscle has a natural advantage with regard to electrical conditioning via the nerve because, unlike the rectus and pectoralis muscles, it is innervated by a single nerve, which makes it technically easier to condition. Second, as the data comparing muscle weight with left ventricular weight suggest, the latissimus has enough bulk to theoretically assume at least a portion of left ventricular function. Third, as evidenced in the long-term studies in this experiment, the muscle is noncritical, and its use for pouch construction minimally impairs the normal activities of the animal. Since the concept was introduced by Kantrowitz and Mckinnon' in 1959, many investigators have commented upon the potential of skeletal muscle to assist in cardiac function. Spotnitz, Merker, and Malm" studied the characteristics of rectus hydraulic pouches and noted pressure-volume curves that were reminiscent of Starling's law of the heart. Vachon, Kunov, and Zingg" observed a similar length-tension relationship with canine diaphragmatic pouches. Kusaba's group,' Nakamura and Glenn,' and Termet, Chalineau, and Estourin" demonstrated short-term hemodynamic benefits with cardiac assistance by skeletal muscle. Long-term studies
have also been performed documenting that skeletal muscle can be implanted into the atrium, right ventricle, and most recently left ventricle with resultant skeletal muscle viability.HII The long-term studies did not fully assess the contribution of the skeletal muscle patch to overall cardiac performance; the majority of shortterm studies emphasize that the major problem with skeletal muscle cardiac assistance has been muscle fatigue. One method to lessen the fatigue of auxiliary ventricles is to limit the demands placed upon the muscle. As pointed out by Dewar and associates," it is perhaps unrealistic for skeletal muscle to assume the entire function of the left ventricle. A skeletal muscle assist device that can generate 20% to 25% of cardiac output might be a more achievable goal. The use of skeletal muscle as an auxiliary ventricle, rather than as a replacement for a portion of myocardium, has several advantages: First, the procedure does not involve cardiopulmonary bypass and eliminates untoward effects on the heart. Second, the auxiliary ventricle is not required to contract in synchrony with each heartbeat and can be paced at a less demanding rate, such as 1: 2 or 1: 3. Third, this model permits a more easily quantifiable assessment of the work performed by skeletal muscle assistance, a point that demands careful scrutiny. Perhaps the most promising possibility for overcom-
Volume 91 Number 4 April, 1986
Hydraulic pouches of canine latissimus dorsi
54 1
Fig. 5. Myofibrillar ATPase stains with acid and alkaline preincubation of electrically conditioned and unconditioned muscles. Top row, Unconditioned muscles. Bottom row, Conditioned muscles. Left column, Acid preincubation . Right column, Alkaline preincubation. Note unconditioned muscles contain approximately 50% fast and 50% slow fibers. Virtually 100% of cells from conditioned muscle contain evidence of slow myosin (uniform dark staining of fibers), as determined by acid stains . Alkaline sta ins of conditioned muscle reveals residual fast myosin (darkly stained fibers) in about 20% of the cells conditioned at 2 Hz.
ing the fundamental difficulty with muscle fatigue lies in the ability of skeletal muscle to adopt to different patterns of use. Muscle plasticity has been most clearly demonstrated in rabbit muscle in response to long-term electrical stimulation.'>" After 6 weeks of electrical conditioning,a fast skeletal muscle is transformed into a muscle characterized by a slow contraction time, dense capillary network, large mitochondrial volume, and increased aerobic enzymatic capability. These changes render a muscle relatively resistant to fatigue, We IH- I9a have previouslyextended these observations on electrical muscle conditioning to the canine species and have transformed latissimus dorsi, rectus abdominis, pectoralis, and diaphragm through long-term electrical stimulation. In an isometric fatigue test developed for the latissimus dorsi muscle, we19a have demonstrated that electrical conditioning does impart to the muscle a relative resistance to fatigue, but also results in a decrease in muscle weight, fiber size, and isometric strength. The results of this experiment with the hydraulic pouch model are similar to the results observed with
our linear model. Just as conditioned latissimus generated lower initial isometric tensions but with minimal fatigue rates in the linear model in comparison with their unconditioned counterparts, so auxiliary ventricles constructed of conditioned muscle generated lower initial pressures than comparable unconditioned pouches (Groups II and IV), but were able to sustain them for longer periods. The fact that conditioned pouches could generate significant physiologic pressures with filling volumes of 25 rnl suggests that a diminution in muscle fiber size and a decrease in isometric strength that accompany a muscle's adaptation to a relatively fatigue-resistant state does not necessarily render a muscle less capable of cardiac assistance. The optimal method to electrically precondition a muscle has not yet been determined and may, in fact, be a moot point. The muscle will ultimately adapt to how it is used. A muscle conditioned at 2 Hz for 6 weeks may undergo a further series of changes if it is used permanently for diastolic augmentation. For instance, the residual fast myosin seen in 20% to 25% of fibers
542
The Journal of Thoracic and Cardiovascular Surgery
Mannion, Hammond, Stephenson
after 6 weeks of electrical stimulation with single pulses at 2 Hz might disappear with stimulation with a burst pattern, which is required for diastolic augmentation. We* have demonstrated that diaphragm conditioned for a I year period is different in some respects from diaphragm conditioned for 6 weeks. The present value of electrical conditioning is that muscle conditioned by any pattern of long-term stimulation might give a more realistic appraisal of the ultimate ability of skeletal muscle to assist in cardiac function. Several authors have commented upon the possible adverse effects that interruption of blood supply might have upon muscle function.v' We 19b have previously shown that electrically unconditioned latissimus is especially sensitive to interruption of collateral supply. In the same experiment we have also noted that a delay of exercise testing for a 3 week period after collateral interruption resulted in a considerable recovery of the normal exercise-induced increase in unconditioned muscle blood flow. The vascular delay alone used in muscle preparation in Groups II and III was an effort to see if the muscle pouch function could be improved by delaying exercise several weeks after interruption of its collateral blood supply. However, the delay period between interruption of the chest wall collaterals and creation of the hydraulic pouch in Group II resulted in only minimal improvement in fatigue rates. This theoretically could have resulted from one of several reasons: First, all of the collaterals were not interrupted. Second, newly formed collaterals were again interrupted during pouch construction. Third, other factors such as fiber type play a predominant role in pouch fatigue. The observation that two of three animals in which the pouch was made and left in situ were able to generate flow against 80 mm Hg for approximately I hour may suggest that this method of vascular delay may result in more optimal pouch function. Further studies may clarify possible improvements in pouch function with creation of permanent pouches. The proper method to construct a pouch of skeletal muscle remains to be determined. Nature has provided two examples, the right and left ventricles." The right ventricle is basically a bellows pump, with contraction of the free wall against a relatively immobile septum. This mechanism provides for a highly efficient volume pump, but it is useful only at low pressures. The left ventricle is a muscular cone, with the thickest part of the wall at the 'Acker M, Mannion JD, Brown W, Salmons S, Henriksson J, Bitto T, Gale D, Hammond R, Stephenson L: Canine diaphragm musele after one year of continuous electrical stimulation at 2 Hz and 4 Hz, Its potential as a myocardial substitute, (Submitted for publication.)
widest part of the ventricle and the thinnest at the apex, Ejection of blood is accomplished with circumferential shortening around the short axis. The pump is useful for generating high pressures. The multiple wrap technique was an attempt to increase wall thickness and permit an auxiliary ventricle to assume a portion of left ventricular function. The contour of the wrap-a conical spiralwas an effort to simulate the configuration of the left ventricle, a geometry that was unraveled by Pettigrew. 2 1, 22 We are not certain what advantage this multiple-layer technique provides to the latissimus pouch. Indeed, there may be a disadvantage, in that the inner distal muscle may become relatively ischemic from high intraventricular pressures. Further study might clarify this point. The hydraulic system used in this experiment is not a replica of systemic circulation. It has an infinite compliance, low resistance, and high inertial load. Flows generated in this model would not necessarily be comparable to flows generated over time in an in vivo circulation. However, the system does give some indication of how an auxiliary ventricle might perform against a physiologic pressure load. The observation that one auxiliary ventricle was able to generate significant flow for I hour against systemic resistance holds promise for future investigation in this area. We are not yet certain if the dilatation observed during the hydraulic fatigue testing with the auxiliary ventricle reflects intrinsic muscle failure or unraveling of the muscle pouches. What is certain is that all factors that contribute to fatigue have not yet been determined, let alone eliminated. Further work on electrical muscle conditioning, optimization of muscle blood supply, configuration of auxiliary ventricle, and new modes to harness the energy of skeletal muscle are only a few of the items that need to be investigated in the effort to use skeletal muscle for cardiac circulatory assistance.
REFERENCES Franciosa JA, Wilen M, Zuseke S, Cohn IN: Survival in men with severe chronic left ventricular failure due to either coronary heart disease or idiopathic dilated cardiomyopathy. Am J Cardiol 51:831-836, 1983 2 Kusaba E, Stule W, Sawatoni S, Jaron D, Freed P, Kantrowitz A: A diaprhagmatic graft for augmenting left ventricular function, Trans Am Soc Artif Intern Organs 19:251-257,1973
3 vonRecum A, Stirke JP, Hamada 0, Saba H, Kantrowitz A: Long-term stimulation of a diaphragm muscle pouch, J Surg Res 23:422-427, 1977 4 SpotnitzHM, Merker L, MaImJR: Applied physiology of the canine rectus abdominis. Force-length curves corre-
Volume 91 Number 4 April, 1986
lated with functional characteristic of a rectus powered "ventricle." Potential for cardiac assistance. trans Am Soc Artif Intern Organs 20:747-756, 1974 5 Kantrowtiz A, McKinnon WMP: The experimental use of the diaphragm as an auxiliary myocardium. Surg Forum 9:266-268, 1959 6 Vachon BR, Kunov H, Zingg W: Mechanical properties of diaphragm muscle in dogs. Med Bioi Eng Com put 13:252260, 1975 7 Nakamura K, Glenn W: Graft of the diaphragm as a functioning substitute for the myocardium. An experimental study. J Surg Res 4:435-439, 1964 8 Termet H, Chalencon JL, Estour F: Transplantation sur Ie myocarde d'un muscle strie exciete par pacemaker. Ann Chir Thorac Cardiovasc 5:260-263, 1966 9 Macoviak JA, Stephenson LW, Kelly A, Likoff M, Reichek N, Edmunds LH Jr: Partial replacement of the right ventricle with a synchronously contracting diaphragmatic skeletal muscle autograft. Proceedings of the Third Meeting of the International Society for Artificial Organs 5:Suppl 1:550-555, 1981 10 Macoviak JA, Stephenson LW, Alavi A, Kelly AM, Edmunds LH Jr: Effects of electrical stimulation on diaphragmatic muscle used to enlarge right ventricle. Surgery 90:271-277,1981 II Sola OM, Dillard DH, Ivey TD, Haneda K, Itoh T, Thomas R: Autotransplantation of skeletal muscle into myocardium. Circulation 71:Suppl 2:341-348, 1985 12 Dewar ML, Drinkwater DC, Wittnich C, Chiu RC: Synchronously stimulated skeletal muscle graft for myocardial repair. J THORAC CARDIOVASC SURG 87:325-331, 1984 13 Salmons S, Sreter FA: Significance of impulse activity in the transformation of skeletal muscle type. Nature 263:30-34, 1967 14 Salmons S, Vrbova G: The influence of activity on some contractile characteristics of mammalian fast and slow muscle. J Physiol 201:535-549, 1969 15 Salmons S, Henriksson J: The adaptive response of skeletal muscle to increased use. Muscle Nerve 4:94-105, 1981 16 Cotter MA, Hudlicka 0, Pette D, Staudte HW, Vrobva G: Changes of capillary density and enzyme pattern in fast rabbit muscles during long-term stimulation. J Physiol 230:34-35, 1973 17 Hudlicka 0, Brown M, Cotter M, Smith M, Vrobva G: The effect of long-term stimulation of fast muscles on their blood flow, metabolism and ability to withstand fatigue. Pflugers Arch 369:141-149, 1977 18 Bitto T, Mannion JD, Hammond R, Macoviak lA, Rashkind WJ, Edmunds LH Jr, Stephenson LW: Pectoralis and rectus abdominis for potential correction of congenital heart defects, Pediatric Cardiology, Proceedings of the Second World Congress of Pediatric Cardiology, New York, 1986, Springer-Verlag, pp 609-612 19 Armenti FR, Bitto T, Macoviak JA, Kelly AM, Chase CT, Hoffman BK, Rubenstein NA, St. John Sutton M,
Hydraulic pouches of canine latissimus dorsi
54 3
Edmunds LH Jr, Stephenson LW: Transformation of skeletal muscle for cardiac replacement. Surg Forum 35:258- 260, 1984 19a Mannion JD, Bitto T, Hammond R, Rubenstein NA, Stephenson LW: Histochemical and fatigue characteristics of conditioned canine latissimus dorsi muscle. Circ. Res. 58:298-304, 1986 19b Mannion JD, Velchek M, Alavi A, Stephenson LW: Blood flow in conditioned and unconditioned latissimus dorsi muscle (abstr). Second Vienna Muscle Symposium, 1985, p 28 20 Parmley WW, Talbot L: Heart as a Pump, Handbook of Physiology, Vol I, The Heart, Section 2, The Cardiovascular System, R Berne, ed., Bethesda, 1979, Waverly Press, pp 429-460 21 Streeter DD: Gross morphology and fiber geometry of the heart, Handbook of Physiology, Vol I, The Heart, Section 2, The Cardiovascular System, R Berne, ed., Bethesda, 1979, Waverly Press, pp 61-112 22 Pettigrew J: On the arrangement of the muscular fibers of the ventricular portion of the heart of the mammal. (Croonian Lecture, Abstract). Proc R Soc Land 10:433440, 1860 23 Macoviak JA, Stephenson LW, Spielman S, Greenspan A, Likoff M, St. John Sutton M, Reichek N, Rashkind W, Edmunds LH Jr: Replacement of ventricular myocardium with diaphragmatic skeletal muscle. J THORAC CARDlOVASC SURG 81:519-527,1981
Discussion DR. TOM D. IVEY Seattle, Wash.
The authors have shown that in preconditioned pouches physiologic parameters can be obtained in an in vitro preparation. They also have shown that fatigue is a factor in this preparation. I would like to briefly review a slightly different approach that Dr. David Dillard and Dr. Sola in our laboratory at the University of Washington have used to tackle a similar problem. Our first attempts to use skeletal muscle as a pumping fiber involved the sternohyoid muscle of the dog. These studies showed that when the sternohyoid muscle is divided, viable muscle fibers can be maintained in the myocardial position over a long period. We next turned to the sternocleidomastoid muscle, and these studies showed that the pedicle, when divided after long-term implantation in the left ventricle, would bleed rather briskly in the retrograde fashion, more briskly, in fact, than it would in the antegrade fashion. However, use of this preparation also turned us to a significant problem, that of fibrosis and clotting at the muscle-blood barrier. Our later attempts to tackle this problem, like the authors', involved constructing a permanent pedicle from latissimus dorsi muscle. Our current preparation using the latissimus dorsi pedicle is similar to that used by the authors, in that the pedicle is lifted, mobilized, and then prestimulated in a stretched position according to the following protocol: 2 weeks
544 Mannion, Hammond. Stephenson
of pacing at 30 beats/min, 2 weeks at 60 beats/min, and finally I month at 100 beats/min. These slides depict the hypertrophy that we have seen in the pedicle with a section taken before stretching and stimulation versus another section showing marked hypertrophy and, as the authors have described, conversion to slow-twitch fibers. The fibers in these pedicles are rather healthy and have taken on a rather hypertrophied appearance. We then tackled the problem of fibrosis and thrombosis at the muscle-blood barrier by implanting into the left ventricle without full-thickness excision of the muscle, leaving a thin layer of endocardium. In our preparation, this is roughly equivalent to a 25% left ventricular infarction. We then used synchronous pacing in the VVD mode. Currently our short-term preparations have shown no hemodynamic improvement. Our long-term preparations await study by both sonomicrometry and ventriculography, and we also are working on a right ventricular preparation. I would like to ask Dr. Mannion two questions. In the manuscript he showed that even at 4 hours his preparations did show fatigue, I would like to know what he plans in the future to overcome this problem. Second, how would he modify his preparation to overcome the problem that we have seen, that being encasement of the muscle fibers in fibrous tissue when exposed to the blood barrier?
The Journal of Thoracic and Cardiovascular Surgery
DR. MANNION (Closing) Dr. Ivey, thank you for your perceptive comments. It is true that we have identified an element of fatigue in these pouches, and at this point we do not believe that we have even identified all of the causes that contribute to muscle fatigue. Blood supply contributes to it, fiber type, pH, glycogen content of the muscles, and many other factors that we are not even sure of yet contribute to muscle fatigue. Once we identify more of these factors, we can eliminate more of the causes for fatigue. One of the problems with our model, as Dr. Ivey pointed out, is that it is subject to the same difficulties inherent in the artificial heart program. We have to use a biologic membrane inside the ventricle and then face the attendant problems of embolism and stroke. We do not necessarily know yet that this is the best approach to take. The value of this particular model is that it is easy to quantitate the amount of cardiac type work that a skeletal muscle can do. In past studies at the University of Pennsylvania in which we have placed pedicle flaps onto the heart, we have had difficulty determining exactly what percentage of the cardiac output is caused by the contraction of the skeletal muscle. We think if we can demonstrate with this model that the skeletal muscle can do significant amounts of cardiac type work for prolonged periods, the study will be a success.