Skeletal Muscle Ventricles, Left Ventricular Apex-to-Aorta Configuration
1 to 11 Weeks in Circulation
Background Skeletal muscle ventricles (SMVs) have been used in animals in a variety of configurations to provide circulatory assistance. Long-term survival and function have been demonstrated. Our laboratory recently obtained promising short-term hemodynamic data in a left ventricular apex-to-aorta model.
Methods and Results SMVs were constructed from the left latissimus dorsi muscle in five adult mongrel dogs. After a 3-week period of vascular delay and 5 to 7 weeks of electrical conditioning, valved conduits were used to connect the left ventricular apex to the SMV and the SMV to the descending aorta. The SMV was then stimulated to contract during cardiac diastole. Initial measurements showed a significant increase in the mean femoral diastolic pressure (62±6 versus 51±5 mm Hg, P<.05). There was also a decrease in the left ventricular tension-time index (11.5±2.5 versus 14.6±2.1 mm Hg·s, P<.05), indicating a decrease in the work requirement of the left ventricle. During SMV stimulation, the majority of flow (65%) was through the SMV circuit and was associated with reversal of flow in the proximal descending thoracic aorta. The longest-surviving animal survived 76 days, at which time pressure augmentation was still seen (mean femoral diastolic pressure, 63±0.9 versus 50±1.2 mm Hg, P<.05).
Conclusions Survival beyond the acute setting is possible with this model. Diastolic pressure augmentation can be effectively maintained over time.
Our laboratory has constructed pumping chambers from latissimus dorsi muscle in dogs. These chambers are referred to as SMVs. They have been coupled to the canine circulation in a variety of configurations to assist the left or right heart.1 2 3
Recently, acute studies in a model in which the SMV is connected from the LV apex to the descending aorta by two valved conduits have yielded promising hemodynamic data.4 5 The purpose of this experiment was to determine whether minor design modifications would result in survival beyond the acute setting.
Construction of SMV
Five adult mongrel dogs weighing between 18 and 25 kg were operated on in accordance with the Guide for the Care and Use of Laboratory Animals published by the National Institutes of Health (NIH publication 85-23, revised 1985). After induction of anesthesia with thiamylal sodium (15 mg/kg), the animals were intubated and placed on a ventilator, and anesthesia was maintained with 1% to 2% isoflurane.
The left latissimus dorsi muscle was mobilized through an incision that extended from the midaxillary line to the tip of the 12th rib. The origin of the muscle from the posterior spinous processes and 11th and 12th ribs was divided, and the muscle was elevated from the chest wall, remaining attached at its tendinous insertion to the humerus. Care was taken to avoid injury to the neurovascular bundle. A bipolar nerve lead (Medtronic, Inc) was then placed around the thoracodorsal nerve for future stimulation. This lead was connected to a neurostimulator, which was then placed under the left rectus abdominis muscle through a separate incision. The muscle was then wrapped 1.5 to 2 times around a cylindrical plastic mandrel (diameter, 3.4 cm; volume, 2.5 mL/kg). A Teflon felt (USCI) sewing ring was used to secure the muscle edge to the base of the mandrel with 5-0 polypropylene suture. The muscle layers were secured with several interrupted absorbable sutures, and the muscle pouch was then secured to the chest wall with absorbable suture. The wound was closed in layers.
All animals underwent a 3-week period of rest to allow the latissimus to recover from ischemia induced by division of the collateral blood supply. After this, the neurostimulator was activated to deliver a 2-Hz continuous pulse of 1-V amplitude and 210-μs duration. Stimulation was continued for 5 to 7 weeks to induce transformation of the muscle from a fatigue-sensitive to a fatigue-resistant state.
Connection to Circulation
After the conditioning period of the SMV was complete, a second operation was performed. The animals were again induced, intubated, and maintained on inhalational anesthesia. A left thoracotomy was performed through the fourth intercostal space, and the pericardium was incised to gain access to the LV apex. The plastic mandrel was removed from the SMV, and two sensing leads were placed on the right ventricle.
The SMV was connected to the circulation with two woven Dacron tubes that each contained a 12-mm-diameter porcine valve (Hancock; Medtronic, Inc). These two grafts were sewn to a Gore-tex (W.L. Gore and Associates) base cap that was then sutured with continuous running 4-0 polypropylene suture to the base of the SMV. The afferent graft was anastomosed to the LV apex without cardiopulmonary bypass. A plastic connector (22 mm long×10 mm in diameter; Medtronic, Inc) was used to couple the apex to the afferent limb of the graft. A double purse-string suture of 4-0 polyester suture was placed around the apex, and then single 4-0 polyester sutures with reinforced Teflon pledgets were placed circumferentially around the purse-string suture. A 10-mm ventriculotome was then used to excise a cylindrical portion of the apex, and the plastic coupler was quickly inserted. The purse-string suture was tightened, and the pledgetted sutures were placed through the rim of the connector and tied in place. A running 6-0 polypropylene suture was used to connect the plastic piece to the afferent Dacron graft, and then the efferent limb of the graft was anastomosed to the descending aorta in an end-to-side fashion with a running 6-0 polypropylene suture (Fig 1⇓). The limbs of the grafts were aspirated of air and filled with 0.9% sodium chloride solution. After connection to the circulation, a rate-responsive synchronized pulse-train stimulator (Prometheus; Medtronic, Inc) was connected to the nerve lead and sensing leads. The thoracic aorta just distal to the left subclavian artery was then narrowed by ≈50% with umbilical tape. This was done to increase preferential flow from the LV apex through the SMV circuit instead of through the aortic valve. After initial hemodynamic measurements, the wound was closed in layers and the animal was allowed to recover. The animal was maintained on oral aspirin thereafter for its antiplatelet effect.
SMV Stimulation Protocol
The stimulator was programmed to deliver a 33-Hz or 50-Hz burst of electrical stimulation with a delay of 30% to 35% of the RR interval to provide maximum diastolic counterpulsation, with a burst duration of 30% of the RR interval, a 2- to 3-V amplitude, and at a 1:2 or 1:1 contraction ratio with the heart. Long-term, the dogs were maintained at a 33-Hz burst frequency in a 1:2 contraction ratio with the heart.
Immediately after connection to the circulation, initial hemodynamic and SMV functional measurements were made. Further measurements were made at periodic intervals under general anesthesia. The stimulator was adjusted to deliver different burst frequencies at different contraction ratios for the purpose of measurements, and then the stimulator was returned to the 33-Hz, 1:2 stimulation mode. Femoral and carotid pressure measurements were made with a fluid-filled manometer line connected to a transducer. LV pressure was measured with a microtransducer-tipped catheter (Millar Instruments, Inc). Cardiac output was measured with a 16-mm ultrasonic flow probe (Transonic Systems, Inc) placed around the pulmonary artery, and the SMV flow, proximal descending thoracic aortic flow, and distal descending thoracic aortic flow were measured with similar 10-mm flow probes. Data were collected with a Gould ES 1000B recording and display system and simultaneously with a Codas on-line data collection program (Dataq Instruments, Inc). Flow probes were left in place for long-term measurements.
Cardiac output, SMV flow, and proximal and distal thoracic aortic flow were calculated by averaging the flow over time with a Windaq data analysis program (Dataq Instruments). Mean pressure was calculated as the average pressure over time. TTI was calculated by integration of the area under the LV pressure curve. The passive flow through the SMV circuit was calculated as the distal thoracic flow minus the proximal flow. The SMV flow during active contraction was calculated as the distal thoracic flow plus the negative (reversed) proximal thoracic flow. All data were taken over a period of at least 20 cardiac cycles.
Data are expressed as the mean±SD. Statistical analysis was performed by a paired t test.
Animals survived after the SMVs were connected from the LV apex to the aorta from 10 to 76 days. The causes of death are listed in Table 1⇓. The dog that survived 76 days was euthanized after terminal hemodynamic measurements because of sepsis that most likely resulted from an indwelling flow probe. Histological examination of the proximal, middle, and distal portions of this animal's muscle pouch revealed viable skeletal muscle without fibrosis in all areas. There was some fibrosis distally that probably represented the fascial layer that was included in the construction. Minimal atrophy was present in the proximal area of the chamber near the sewing ring, affecting perhaps 10% of the muscle. All animals had functional SMVs at the time of death. Autopsy demonstrated a thin layer of adherent thrombus in the SMV chamber and on the porcine valves in all animals but no evidence of distal embolization or valvular dysfunction in any dog.
The initial measurements revealed that with the SMV off, 53% of the cardiac output flowed through the SMV circuit. This is shown in the flow tracings as greater flow in the distal descending aorta than in the proximal descending aorta (Fig 2a⇓). At 33-Hz stimulation of the SMV and 1:2 contraction ratio, this amount of blood flow increased to 65%. During the systolic phase of the cardiac cycle, blood is ejected both through the aortic valve and through the SMV circuit. There is an increase in distal flow during this period from blood flowing passively through the SMV. During SMV contraction, there is reversal of flow in the thoracic aorta just proximal to the efferent limb of the SMV (point A, Fig 1⇑) as the luminal pressure inside the pouch is elevated above the pressure in the aorta. During this period, most of the distal aortic flow and the reversed flow in the proximal descending thoracic aorta are being pumped by the SMV contraction (Fig 2b⇓). At 33 Hz and a 1:1 contraction ratio, the flow through the SMV circuit increased to 95% of the cardiac output. The majority of flow in the proximal descending thoracic aorta is reversed in this case, as would be expected, because only 5% of the cardiac output crossed the aortic valve (Fig 2c⇓).
These flow characteristics are substantially different from those in our previous study. In our earlier experiments, we did not narrow the aorta distal to the subclavian artery, and blood flow through the SMV occurred only with SMV contraction. This resulted in a substantial period of time during which there was minimal flow in the SMV circuit (Fig 3a and 3b⇓⇓).4 In contrast, our present model with the aortic narrowing shows flow with every cardiac contraction and additional flow augmentation with contraction of the SMV.
The LV TTI decreased during SMV contraction in the three animals measured at the initial procedure and ranged from a 9.5% decrease (13.3±0.5 versus 14.7±0.5 mm Hg·s) to a 31.5% decrease (8.5±0.2 versus 12.4±0.3 mm Hg·s). The mean reduction was 22.3% (11.5±2.5 versus 14.6±2.1 mm Hg·s, P<.05) (Fig 4⇓). This reduction was not apparent, however, in the longest-surviving dog at 35 days (25.85±0.62 versus 25.55±0.81 mm Hg·s, P=NS) or at 76 days (20.29±0.73 versus 20.46±0.58 mm Hg·s, P=NS). In the two animals not studied at the initial procedure, a Millar catheter was not available because of a technical malfunction.
Initially, the mean femoral diastolic pressure was increased from 11% (60±1 versus 54±0.9 mm Hg, P<.05) to 28% (68±0.7 versus 53±1.2 mm Hg, P<.05). The mean difference was 21% (62±6 versus 51±5 mm Hg, P<.05). This increase in femoral diastolic pressure was maintained over time, with the dog surviving 76 days demonstrating an increase from 50±1.2 to 63±0.9 mm Hg (P<.05) (Fig 5⇓). Three of the four dogs with hemodynamic measurements made after the initial implant showed consistent diastolic augmentation. One animal died before a second study could be performed (Table 2⇓).
The decrease in SMV EDP ranged from 96% (1.6±0.8 versus 42±3 mm Hg, P<.05) to 23% (40±4 versus 52±1 mm Hg, P<.05). The mean decrease was 76% (12±16 versus 45±7 mm Hg, P<.05) (Fig 6⇓). This decrease in the SMV EDP was maintained over time, with the longest-surviving dog demonstrating a decrease of 53% at day 76 (18±5 versus 38±3 mm Hg, P<.05). There was a trend toward decreasing performance of the SMV in this animal, as demonstrated by a decline in pressure reduction over time (Fig 7⇓).
The characteristic waveforms at 76 days for the longest-surviving dog are shown in Fig 8⇓. Substantial diastolic augmentation is still seen in the femoral arterial pressure tracing and the intraluminal SMV pressure tracing. There is some augmentation of the carotid diastolic pressure, although to a lesser degree than the femoral artery pressure. This is because the area of aortic narrowing is distal to the carotid artery but proximal to the efferent limb of the SMV.
The use of a conduit from the LV apex to the aorta is not new. In 1910, Carrel6 developed a technique for connecting the LV apex to the descending aorta and suggested this as a temporary bypass, allowing time for surgical treatment of the diseased proximal aorta. In the 1950s, several investigators reported on the use of valved conduits from the LV apex to the aorta in experimental animals and suggested this technique as a method of treating aortic valvular stenosis.7 8 Cooley et al9 subsequently used such conduits clinically for the treatment of “tunnel” aortic stenosis, aortic stenosis with severe calcification of the ascending aorta, and other related conditions, and Brown et al10 reported on the combined repair of aortic stenosis and aortic coarctation with a combined valved conduit/patch aortoplasty technique.
Acute studies using skeletal muscle power to augment blood flow with the LV apex-to-aorta configuration have previously been reported. Brister and colleagues11 constructed SMVs from unconditioned rectus abdominis muscle wrapped around a dilated conduit that was connected from the LV apex to the descending aorta. There was a valve present in the afferent limb of the conduit. They were able to show augmentation of either systolic or diastolic pressure, depending on the synchronization of the SMV contraction with the cardiac cycle. Stevens et al12 performed studies with an unconditioned rectus muscle pouch in dogs with heart failure induced by β-blockade. These SMVs were wrapped around a pouch made of compressible polymer and had a tilting-disk mechanical valve in place in both the inflow and outflow conduit. There was a significant increase in cardiac output and mean diastolic pressure with SMV contraction; however, there was a substantial decline in function over the 1-hour duration of the experiment. This was thought to be due to muscle fatigue.
This report represents experience with chronic LV apex-to-aorta SMVs. Our previous attempts were associated with a high incidence of thrombosis within the SMV system, thus limiting the potential for long-term survival. Our failure to achieve survival beyond the acute setting earlier was thought to be due to valve dysfunction that led to sluggish blood flow and clot formation. Our present study suggests that it was most likely the sluggish blood flow in the SMV circuit when the SMV was not contracting that led to thrombus formation. The addition of aortic narrowing just distal to the left subclavian artery but proximal to the efferent limb of the SMV causes blood to be preferentially shunted through the SMV circuit, even when the SMV is not contracting (Fig 3a and 3b⇑⇑).4 This led to minimal thrombus formation in our present series and enabled all animals to survive at least 10 days with a functioning pumping chamber.
The causes of death in the present study are similar to those in other long-term SMV studies in which the blood was not routed from the LV apex. Disruption of the vascular anastomoses occurred in two SMVs in the present series and probably represents the early phase of the “learning curve” associated with this model. Sweeney et al13 reported on the 10-year clinical experience of the Texas Heart Institute from 1975 to 1984, in which conduit disruption from the LV apex occurred in 4 of their 34 patients who survived the initial operation. We have constructed SMVs in our laboratory with other configurations that have remained intact while pumping blood in the circulation for >3 years (unpublished observations).
Indwelling flow probes were left in place in our animals to assess the flow characteristics of the SMVs long-term, and two of the deaths were directly attributable to these (one from a flow probe erosion through the aorta and one from infection). The open tract through the animal's skin was the probable source of the infection.
This present model should provide an improved blood supply to the muscular walls of the SMV in comparison with the diastolic counterpulsator model. In previous studies with the SMV as a counterpulsator, the SMV lumen is exposed to a pressure at or above aortic diastolic pressure throughout the cardiac cycle. This most likely leads to capillary compression in the innermost layer of muscle and could cause long-term ischemia and muscle deterioration.14 Our laboratory has previously reported on the long-term histological changes seen in SMVs over time.15 In contrast, this present model has a substantial period of time during which the SMV luminal pressure is at or below LV diastolic pressure, thereby allowing for increased blood flow to the muscular walls of the pump. Our longest-surviving animal demonstrated viable skeletal muscle in all areas of the pumping chamber at postmortem histological examination.
Previously, investigators have used an electrical circuit model to analyze various SMV configurations and concluded that the LV apex-to-aorta model is the most effective in terms of pressure augmentation and LV unloading.16 In this present study, we have shown a substantial decrease in the systolic TTI (average reduction of 22.3%), which is similar to the results of previous short-term studies performed with the LV apex-to-aorta model in our laboratory. Previously, the model of SMVs as aortic diastolic counterpulsators has demonstrated a decrease in load-dependent and load-independent indexes of LV function, although the TTI in these studies was decreased by only 7% to 8%.17 It therefore appears that this present model is more effective in terms of LV unloading.
Augmentation of femoral pressure and decrease in SMV EDP was maintained over time with this model. However, our longest-surviving dog, which initially demonstrated a decrease in SMV EDP of 85%, had decreased to only 50% by the terminal experiment. This dog, however, had developed a wound infection and was septic at the time of the terminal experiment. Our experience has demonstrated that sepsis can lead to a deterioration of SMV function.
In summary, the addition of aortic narrowing in this model improves the flow characteristics through the SMV circuit and allows for survival beyond the acute setting without the problem of thrombosis. Our previous studies with long-term SMV counterpulsation have demonstrated improved survival and functional statistics as refinement of the model progresses. With more experience, we anticipate that good long-term function and survival will be achieved.
Selected Abbreviations and Acronyms
|SMV||=||skeletal muscle ventricle|
This work was supported by NIH grant NHLB-34778-11. Dr Greer was supported by NIH National Research Service Award grant F32-H209292-01.
Reprint requests to Larry W. Stephenson, MD, Chief, Division of Cardiothoracic Surgery, Harper Hospital, Suite 228, 3990 John R St, Detroit, MI 48201-2097. E-mail firstname.lastname@example.org.
- Received May 8, 1996.
- Revision received August 14, 1996.
- Accepted August 28, 1996.
- Copyright © 1997 by American Heart Association
Bailey CP, Glover RP, O'Neill TJE, Ramirez HPR. Surgical relief of aortic stenosis. J Thorac Surg. 1950;20:516-541.
Sarnoff SJ, Donovan TJ, Case RB. The surgical relief of aortic stenosis by means of apical-aortic valvular anastomosis. Circulation. 1955;11:564-575.
Brister S, Fradet G, Dewar M, Wittnich C, Lough J, Chiu R. Transforming skeletal muscle for myocardial assist: a feasibility study. Can J Surg. 1985;4:341-344.
Mannion JD, Velchik MA, Acker M, Hammond R, Staum M, Alavi A, Duckett S, Stephenson LW. Transmural blood flow of multi-layered latissimus dorsi skeletal muscle ventricles during circulatory assistance. Trans Am Soc Artif Intern Organs. 1986;94:733-746.