| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Ann Thorac Surg 1996;61:603-609
© 1996 The Society of Thoracic Surgeons
Department of Cardiothoracic Surgery, Wythenshawe Hospital, and Biological Services Unit, University of Manchester, Manchester, United Kingdom
Accepted for publication September 20, 1995.
 |
Abstract |
|---|
 TopFootnotes Abstract IntroductionMaterial and MethodsResultsCommentAcknowledgmentsReferences |
|---|
Methods. In situ, preconditioned latissimus dorsi muscles in 5 sheep were stimulated in either 1:1 (muscle:heart) or 1:2 synchrony with the systolic phase of the cardiac cycle, using a burst duration of either 21% or 35% of the cycle. Thoracodorsal artery blood flow and thoracodorsal venous lactate concentrations were measured before and immediately after a 3-minute period of stimulation.
Results. The exercise-induced augmentation of thoracodorsal artery blood flow was significantly (p < 0.05) less with a 1:2 regimen than a 1:1 regimen, for both a 21% (88%; 95% confidence interval [CI], 55.6% to 127.3% versus 138.9%; CI, 97.6% to 188.8%) and 35% burst duration (123.2%; CI, 84.7% to 169.9% versus 167.0; CI, 120.8% to 222.6%). After cessation of stimulation, reactive hyperaemia was observed in 3 of 5 animals with 1:1 21% burst stimulation, and in 5 of 5 animals with a 35% burst duration, but was not seen after 1:2 regimens. A significant (p < 0.01) increase in thoracodorsal venous lactate levels was present after 1:1 35% burst stimulation (34.9%; CI, 9.9% to 65.6%), but lactate levels tended to fall when a 1:2 ratio was used (15.9%; CI, -3.2% to 31.5%; p < 0.1).
Conclusions. One-to-one stimulation regimens may be detrimental to latissimus dorsi blood flow, and an adaptive, rather than fixed, burst duration may be preferable. These findings have important implications for the cardiomyoplasty procedure.
|
Introduction |
|---|
|
TopFootnotesAbstractIntroductionMaterial and MethodsResultsCommentAcknowledgmentsReferences |
|---|
One factor that has been identified as a possible cause of muscle damage after CMP is muscle ischemia [7, 8]. Division of collateral vessels at the time of muscle mobilization for CMP may render the distal part of the muscle acutely ischemic, although it has been demonstrated that over a 3-week period this deficit recovers to some extent [11]. However, perhaps a more important consideration is the abnormal working conditions imposed on the muscle during cardiac assistance. Skeletal muscle blood supply takes place predominantly during cardiac systole, and thus during synchronized systolic LD contraction, muscle blood supply may be compromised at a time of high metabolic demand. To investigate the effect of synchronized LD contraction on muscle blood flow, the following series of experiments were performed on sheep.
|
Material and Methods |
|---|
|
TopFootnotesAbstractIntroductionMaterial and MethodsResultsCommentAcknowledgmentsReferences |
|---|
Anesthetic Techniques
General anesthesia was induced and maintained via an endotracheal tube with a mixture of equal volumes of oxygen and nitrous oxide in combination with halothane (1% to 2%). Anesthesia was supplemented with morphine (0.1 mg/kg) intravenously as an initial bolus, followed by 0.025 mg/kg increments as required. The lungs were mechanically ventilated using a standard volume divider (Blease, Chesham, UK). The initial minute volume (approximately 10 mL/kg) was adjusted according to arterial blood gas estimations to maintain an arterial pH between 7.35 and 7.55, carbon dioxide tension between 40 and 55 mm Hg, and oxygen tension greater than 100 mm Hg. A rumen tube was inserted. Maintenance fluid was with warmed crystalloid (0.18% NaCl/4% glucose) at a rate of 10 mL•kg-1•h-1, and blood loss was replaced with a mixture of warmed crystalloid (0.9% NaCl) and colloid (Gelofusine; Braun Medical Ltd, Aylesbury, UK), maintaining a hematocrit of 25% to 30%. A heating blanket regulated pharyngeal temperature between 38° and 40°C, as is normal for sheep. Electrocardiogram and during the final experiment also carotid arterial pressure, were continuously monitored. The animal was kept in a steady state for at least 30 minutes before data acquisition commenced.
For the initial operation a single 1.5-g dose of cefuroxime sodium (Glaxo Laboratories Ltd, Uxbridge, UK) was given intravenously for antimicrobial prophylaxis and diclofenac sodium (Geigy Pharmaceuticals, Horsham, UK) 1.5 mg/kg intramuscularly was used for postoperative analgesia.
Operative Procedures
During a preliminary procedure, and under aseptic conditions, a left flank incision was made extending from the axilla diagonally to the 11th rib. Two intramuscular wire electrodes (SP6500; Medtronic, Minneapolis, USA) were placed 5 cm apart across the proximal part of the LD muscle and connected to a bipolar stimulator (Itrel SP 4721, Medtronic or Optima MPT, Teletronics, Englewood, CO), which was positioned below the left rectus muscle. The wound was closed in layers. Muscle stimulation was initiated within the next 2 days using 2-Hz continuous stimulation at 2.5 to 3 V (2 to 4 times threshold) with an electrical pulse width of 210 ms (Itrel) or 250 ms (Optima MPT). All animals made an uneventful recovery.
After an 8- to 9-week period of muscle conditioning, a second, terminal procedure was performed. With the animal positioned on the right side, the thoracodorsal artery (TDA) and vein (TDV) were dissected within the neurovascular pedicle of the left LD muscle. A 2-mm ultrasonic flow probe (Transonic Systems Inc, Ithaca, NY) was placed around the TDA, and a 22 G-gauge cannula was inserted into the TDV to allow blood sampling for serum lactate measurements. To ensure that changes in blood flow could be attributed solely to the pattern of synchronized stimulation, and not to devascularization caused by muscle mobilization, the LD muscle was left undisturbed with its collateral circulation intact. The carotid artery was cannulated to allow sampling of blood gases and monitoring of blood pressure. The animal was systemically heparinized with 5,000 IU of heparin sodium (CP Parmaceuticals Ltd, Wrexham, UK) to help maintain patency of the vascular lines. A left thoracotomy was performed at the fourth interspace and an epicardial sensing wire (SP 6500; Medtronic) was placed on the left ventricle. The muscle stimulating wires were retrieved from the rectus sheath and connected, together with the cardiac lead, to the appropriate channels of a synchronizing pulse train generator (SP4710; Medtronic). A 20-mm ultrasonic flow probe (Transonic) was placed around the main pulmonary artery for continuous monitoring of cardiac output.
Protocol of Investigation
The muscle stimulation programme aimed to simulate Chachques and associates' protocol for chronic cardiac assistance [12]. To mirror the effect of a 30-Hz burst of 185 ms duration (pulse width, 198 µs; pulse interval, 31 ms) at a heart rate of 70 beats/min, which occupies 21.6% of the R-R interval, the cardiomyostimulator was programmed to deliver an adaptive burst occupying 21% of the R-R interval. To evaluate the effect of tachycardia, whereby the absolute burst occupies a larger proportion of the R-R interval, the stimulator was subsequently programmed to deliver a 35% adaptive burst. This would be the equivalent of a 185-ms burst at a heart rate of 114 beats/min. The muscle was stimulated at 5 V to ensure full recruitment of fibers. The left front leg of the animal was free to move during muscle stimulation, allowing the LD to perform predominantly nonisometric work as in cardiomyoplasty.
Measurements
Thoracodorsal arterial blood flow was recorded continuously during the experiment. One-milliliter aliquots of blood for lactate measurement were taken from the TDV immediately before the onset of each period of stimulation and within 30 seconds after termination of stimulation. The blood samples were immediately centrifuged at 2,500 g for 3 minutes, and the supernatant was removed and stored at -20°C awaiting further analysis. Serum lactate measurements were done using the Cobas Bio centrifugal analyzer with a fluorimetric attachment (Roche Products Ltd, Welwyn Garden City, UK) [13].
All animals underwent the following stimulation protocols in the same order:
In protocol 1, to investigate the effect of timing of the burst stimulus in relation to the cardiac cycle, the muscle was initially stimulated to contract during cardiac systole with the onset of stimulation 55 ms after the R wave. After a rest period the muscle was then stimulated to contract during diastole (onset of stimulation 245 ms after the R wave). In protocol 2, the effect of the ratio of systolic muscle assist was studied by first stimulating the muscle in a 1:1 muscle:cardiac mode, followed by a period at a 1:2 ratio. In protocol 3, to investigate the effect of a more prolonged burst duration during the cardiac cycle, the muscle was stimulated with burst durations of 21% and 35% of the R-R interval. The effect of an extended burst was studied for both a 1:1 and 1:2 assist ratio.
Each period of stimulation lasted for 3 minutes followed by a 15-minute recovery period before the onset of the next period of stimulation. These parameters were chosen after preliminary experiments in our laboratory had shown that following onset of stimulation of conditioned LD muscle, the TDA blood flow reached a maximum within 2 minutes and remained at this level during the rest of the stimulation period. In the same experiments it had also been shown that blood flow as well as TDV lactate levels returned to baseline within 10 minutes of cessation of stimulation.
Data Collection and Analysis
All hemodynamic parameters were continuously recorded on an IBM-compatible personal computor using CODAS acquisition software (Dataq Instruments Inc, Akron, OH) and analyzed using WINDAQ/EX (Dataq) postacquisition software.
Statistical Analysis
The TDA flow and lactate data were found to be log-normally distributed and converted to natural logarithms for analysis. After analysis the results were transformed back into the original units and presented as geometric means with their 95% confidence intervals (CIs) (the latter were used to illustrate dispersion, as standard deviations and standard errors cannot be transformed back directly from natural logarithms). The changes in each experiment were evaluated using repeated-measures analysis of variance, and results were expressed as absolute values along with proportional changes from baseline as appropriate. The computations were done using the GLIM 3.77 (Royal Statistical Society, London, UK) statistical computer package. Statistical significance was set at the 5% level.
|
Results |
|---|
|
TopFootnotesAbstractIntroductionMaterial and MethodsResultsCommentAcknowledgmentsReferences |
|---|
). All muscle stimulation regimens significantly increased blood flow compared with baseline (p < 0.001). Maximal blood flow was achieved within 1.5 minutes of stimulation, after which no further change was observed during the 3-minute protocol. Although muscle force was not measured, there appeared to be no obvious fatigue during the stimulation periods, except when a 1:1 protocol with a 35% burst duration was used. Maximum TDA blood flow, however, did not change after muscle fatigue had been observed in this group.
|
).
|
). Cycle-by-cycle analysis at maximum blood flow during 1:2 assist with 21% burst showed a highly significant (p < 0.001) increase of 14.7% (CI, 8.4% to 21.4%) in TDA blood flow in cycles without LD stimulation compared with cycles with muscle stimulation (Table 2). This difference was also highly significant (p < 0.001) and even more pronounced when a 35% burst was used (25.6%; CI, 19.6% to 31.9%).
|
|
). After the initial 1:1 systolic contraction with a 21% burst duration a significant (p < 0.001) increase in mean lactate concentration of 52.4% (CI, 24.3% to 87.2%) was observed, although after a further two such stimulation periods no significant change in TDV lactate level from the preceding baseline was seen. In addition, after diastolic 1:1 synchronization no significant change in mean TDV lactate level was measured. However, extended burst stimulation of 35% R-R duration resulted in a significant (p < 0.01) increase in mean serum lactate concentration of 34.9% (CI, 9.9% to 65.6%). Stimulation of the muscle at a 1:2 ratio resulted in a decrease in mean lactate concentration compared with baseline, although this decrease approached significance (p < 0.1) only when a 35% burst was used (15.9%; CI, -3.2% to 31.5%) (Fig 3).
|
|
|
Comment |
|---|
|
TopFootnotesAbstractIntroductionMaterial and MethodsResultsCommentAcknowledgmentsReferences |
|---|
The present study has demonstrated that the LD muscle may be compromised metabolically during certain patterns of synchronous systolic stimulation. The muscle was optimally prepared to deal with the increased metabolic demand; the anatomic blood supply was left intact, and in addition the muscle had been preconditioned with 2-Hz continuous electrical stimulation for 8 to 9 weeks. This protocol of electrical conditioning is known to effect complete fiber-type transformation in both dogs and sheep [17, 18], with resultant fatigue resistance and improved oxidative capacity of the muscle. The muscle stimulation patterns that were chosen reflected those used in clinical cardiomyoplasty. A 1:1 muscle:cardiac stimulation ratio, even at the most ``favorable'' setting of 21% burst duration, resulted in a degree of reactive hyperemia in 3 of 5 animals, and this equated with a small increase in serum lactate concentrations from the muscle. However when the burst duration was extended to 35%, all 5 animals exhibited a hyperemic response, along with a 34.9% increase in TDV serum lactate concentration. In contrast, no reactive hyperemia was observed at 1:2 assist ratios, when TDV lactate measurements tended to fall below baseline levels. Thus at 1:1 assist ratios there is evidence of anaerobic muscle metabolism, which appears to be more pronounced when a longer burst duration is used. However, when 1:2 stimulation protocols are used the metabolic supply:demand ratio does not appear to be compromised even when an extended burst duration is applied.
The possible reason for the observed anaerobic metabolism needs further discussion. It can be argued that when stimulated in a 1:1 mode, the muscle performed twice as much work as when a 1:2 mode was used. Because the LD muscles had been transformed, there should have been no problem to generate increased adenosine triphosphate through aerobic metabolic pathways. However, if the necessary oxygen supply was not delivered because of inadequate LD blood flow, a change to glycolytic adenosine triphosphate generation may have resulted. Cycle-by-cycle analysis of TDA blood flow during 1:2 regimens showed that cycles without LD stimulation carried on average 14.7% (21% burst) or 25.6% (35% burst) more blood than those with LD stimulation, indicating that during muscle contraction blood flow was impaired. That no reactive hyperemia was observed suggests that the period between bursts was long enough to achieve full muscle relaxation and to correct a metabolic deficit before the onset of the next stimulus. However when a 1:1 regimen was used, there was not enough time between individual bursts, and the metabolic deficit could only be corrected during a period of reactive hyperemia after stimulation had ceased. It is therefore not surprising that after an extended burst duration, when there was even less time between individual contractions, a more extensive reactive hyperemia was seen. Work by Lucas and associates [19] on isometric force measurements of preconditioned in situ goat LD muscle supports this theory. Using a stimulation program similar to ours, they showed that when a 185-ms burst was applied at a rate of 50 bursts/min, there was complete relaxation of the muscle between bursts. However when 100 bursts/min were delivered, the muscle failed to achieve full relaxation before the onset of the next contraction. This effect was even more pronounced when an extended burst duration of 240 ms was applied. Skeletal muscle force was well maintained during the increase in the number of bursts. Although in this study muscle stimulation was not synchronized with heart rate, the number of bursts per minute was equivalent to a 1:2 and 1:1 assist ratio at a heart rate of 100 beats/min, as was used in our experiment.
One limitation of our study was that the TDA blood flow measured with the ultrasonic flow probe only constituted part of the LD muscle blood flow. Ideally total muscle blood flow should have been measured. However, this would have involved the use of microsphere techniques, which have inherent inaccuracies compared with ultrasonic flow probes. Furthermore, microsphere methods do not permit the measurement of changes in blood flow within short time periods, and therefore cycle-by-cycle analysis of muscle blood flow during 1:2 stimulation regimens would not have been possible. It may be expected, however, that because of the uniform fiber distribution of transformed muscles, changes in TDA blood flow are a reflection of those in total muscle blood flow. The results in this study have therefore been expressed both as absolute measurements and as relative change from baseline.
The effects of burst duration on blood flow of the mobilized LD muscle have been reported by Gealow and colleagues [20]. In a study on skeletal muscle ventricles it was shown that with high heart rates and a fixed burst duration there was not only reduced muscle blood flow and inadequate muscle relaxation between stimuli, but also impaired functional performance of the skeletal muscle ventricle. It may be expected that stimulation of the muscle during cardiac systole, when muscle blood flow predominantly takes place, affects blood flow to a larger extent than when the same stimulus is applied during diastole. In our experiment, however, mean exercise-induced TDA blood flow and TDV lactate levels were not significantly different between these stimulation regimens. This suggests that during muscle exercise, more important factors than cardiac cycle affect the magnitude of muscle blood flow.
Chronic ischemia can lead to permanent muscle injury, and this may be the reason why clinical results after CMP have been variable. In general most patients have experienced an improvement in their symptoms of heart failure [1–5], although around 15% have not improved or have deteriorated. Of those who have improved, measurements of conventional indices of cardiac performance, such as cardiac output, pulmonary artery wedge pressure, ejection fraction, and exercise performance, have yielded inconsistent results. More recently, animal studies have raised the profile of LD muscle integrity after CMP. In dogs, sheep and goats, variable degrees of degenerative muscle change have been observed [7–9], specifically with loss of viable muscle and replacement by fat and connective tissue. Of more concern are recent reports from a prominent group with one of the largest experiences of clinical CMP, who have observed similar degenerative changes using nuclear magnetic resonance scanning [10]. The same group has recently also reported similar findings in a larger group of patients assessed by computed tomographic scan [21]. This group has been a particular advocate of 1:1 assist settings after CMP. Unfortunately there are few additional data on human LD muscle integrity, even from postmortem studies, although, anecdotally, degenerative change has not been a prominent feature in LD specimens examined by Carpentier's group (personal communication). Interestingly this group has followed a protocol of 1:2 assist synchronization in the majority of their patients [4].
To maintain a balance in the demand:supply ratio during synchronized muscle stimulation it is necessary to match the duration of burst stimulation with that of muscle relaxation. In theory there are several ways in which this can be achieved. First, the muscle:cardiac assist ratio can be increased at higher heart rates, so that the absolute time for muscle relaxation remains more or less constant. This technique has been applied to a limited extend with the SP1005 (Medtronic) cardiomyostimulator, whereby at a heart rate of 110 beats/min the assist mode changes from a 1:1 to a 1:2 ratio. Although this method preserves muscle relaxation time at higher heart rates, no adjustments are being made in the duration of the stimulation period. Therefore, at high heart rates the fixed burst stimulus ``spills over'' into diastole, which in CMP may result in impairment of ventricular filling as well as coronary perfusion. In a recent study by Tsukube and associates [22] it was found that systolic as well as diastolic coronary blood flow was enhanced after CMP in dogs. However in this study nonconditioned LD muscles were used, and it is to be expected that with transformed, and therefore much slower, muscles, coronary perfusion is more likely to be impaired, especially when extended periods of stimulation are used.
An alternative technique to preserve muscle blood supply is to program the burst duration as a relative time of the R-R interval. This option is now available in the newer generation of cardiomyostimulators. There is experimental evidence [20], however, that at higher heart rates the pulse train duration of the adaptive 30-Hz burst is not long enough to allow optimal force generation of the muscle. Higher burst frequencies may have to be used in adaptive bursts of short duration to ensure an adequate stimulus for complete contraction of the muscle and preservation of force. Although extensive manipulation of the burst stimulus is now possible with the latest generation of cardiomyostimulators, more fundamental studies into the effects of synchronized burst stimulation are needed to rationalize the design of stimulation regimens and improve clinical results after CMP.
|
Acknowledgments |
|---|
|
TopFootnotesAbstractIntroductionMaterial and MethodsResultsCommentAcknowledgmentsReferences |
|---|
We thank Trudy E. Shaw, Brian P. Landamore, and Ann E. Tomlin for their technical assistance and Timothy Rainey for performing the lactate measurements. We are grateful to Mr E. Brian Farragher, honorary lecturer in medical statistics, University of Manchester, for his help with the statistical analysis.
|
Footnotes |
|---|
|
TopFootnotesAbstractIntroductionMaterial and MethodsResultsCommentAcknowledgmentsReferences |
|---|
|
References |
|---|
|
TopFootnotesAbstractIntroductionMaterial and MethodsResultsCommentAcknowledgmentsReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  S. Salmons Cardiac assistance from skeletal muscle: a reappraisal Eur. J. Cardiothorac. Surg., February 1, 2009; 35(2): 204 - 213. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Benicio, L. F. P. Moreira, F. Bacal, N. A. G. Stolf, and S. A. Oliveira Reevaluation of long-term outcomes of dynamic cardiomyoplasty Ann. Thorac. Surg., September 1, 2003; 76(3): 821 - 827. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. T.M. Tang, P. Geraghty, M. J. Dascombe, J. C. Jarvis, S. Salmons, and T. L. Hooper Nitroglycerine reduces neutrophil activation and acute damage in latissimus dorsi muscle grafts Ann. Thorac. Surg., December 1, 1998; 66(6): 2015 - 2021. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. Lorusso, O. Alfieri, U. Carraro, J. J. Schreuder, and H. J.J. Wellens Preserved skeletal muscle structure with modified electrical stimulation protocol in a cardiomyoplasty patient: a clinico-pathological report Eur. J. Cardiothorac. Surg., February 1, 1998; 13(2): 213 - 215. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. A. M. van Doorn, H. Degens, M. S. Bhabra, C. B. W. Till, T. E. Shaw, J. C. Jarvis, S. Salmons, and T. L. Hooper Intramural Blood Flow of Skeletal Muscle Ventricles Functioning as Aortic Counterpulsators Ann. Thorac. Surg., July 1, 1997; 64(1): 86 - 93. [Abstract] [Full Text]
|
|
|
|
|
|
E. R. Soltero, D. H. Glaeser, L. H. Michael, C. J. Hartley, N. R. Earle, Z. Li, and G. M. Lawrie Hemodynamic Effects of Different Pacing Ratios in Chronic Dynamic Double Cardiomyoplasty Ann. Thorac. Surg., November 1, 1996; 62(5): 1380 - 1387. [Abstract] [Full Text]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |
