Ann Thorac Surg 1996;62:1380-1387
© 1996 The Society of Thoracic Surgeons
Original Article: Cardiovascular
Hemodynamic Effects of Different Pacing Ratios in Chronic Dynamic Double Cardiomyoplasty
Ernesto R. Soltero, MD,
Donald H. Glaeser, DSc,
Lloyd H. Michael, PhD,
Craig J. Hartley, PhD,
Nan R. Earle, BS,
Zheng Li, MD,
Gerald M. Lawrie, MD
Department of Surgery and Section of Cardiovascular Sciences, Department of Medicine, Baylor College of Medicine, Houston, Texas
Accepted for publication June 6, 1996.
 |
Abstract
|
|---|
Background. Dynamic cardiomyoplasty is being used clinically worldwide, and evaluated by a clinical trial (phase III) in the United States. Some centers stimulate the skeletal muscle wrap with every heart beat (1:1 [muscle:heart]), whereas others use every other heart beat (1:2). Recent concern over the possible deleterious effects of too-frequent stimulation of the muscles motivated the attempt to evaluate, in a canine model of chronic, double cardiomyoplasty, the effects of two different pacing ratios on several hemodynamic parameters of interest.
Methods. Double cardiomyoplasty was performed using both latissimus dorsi muscles in 11 dogs. Fatigue resistance was achieved using the clinical transformation protocol. At a final experiment, acute cardiac failure was induced by administration of propranolol. Hemodynamic measurements of eight physiologic variables were averaged over complete pacing cycles, including the nonpaced beat at a 1:2 pacing ratio.
Results. The net effects of latissimus dorsi muscle stimulation at each of two pacing ratios were compared using nonparametric statistics. With the exception of left ventricular pressure (p = 0.0262) and its first derivative, dP/dt (p = 0.0099), there was no significant difference between hemodynamic performance at the two pacing ratios.
Conclusions. In this canine model, pacing every other beat produces hemodynamic results that are statistically similar to pacing every beat. Less frequent stimulation of the latissimus dorsi muscle may preserve its function and improve clinical results without compromising hemodynamic benefit.
|
Introduction
|
|---|
Dynamic cardiomyoplasty is currently being evaluated as a therapeutic option for patients suffering from congestive heart failure. The early clinical results have been inconsistent and variable [1–6]. Recent experimental and clinical data suggest possible deleterious changes in the latissimus dorsi muscle (LDM) with chronic electrical stimulation in cardiomyoplasty. Lucas and colleagues [7] observed extensive lipomatosis of the LDM in goats after 12 to 35 weeks of electrical stimulation in a model of left cardiomyoplasty. They proposed several possible causes for the skeletal muscle damage, including overstimulation of the LDM. Similarly, Kratz and colleagues [8] reported their experience in pigs. In the stimulated LDM these investigators observed a significant reduction in myofibril volume concomitant with reduction in cross-sectional area of the LDM, and fibrosis in the distal regions of the muscle. They concluded that the damage observed in the skeletal muscle may explain some of the variability in the clinical results.
The optimal pacing ratio used for skeletal muscle stimulation in cardiomyoplasty has not been defined adequately. In approximately half the patients in the phase I worldwide experience the LDM is stimulated with every heart beat (1:1 pacing ratio); in contrast, the remainder are being stimulated every other heart beat (1:2 pacing ratio) [5].
Several groups have examined the effect of using different pacing ratios to preserve more muscle function in the application of skeletal muscle for cardiac assist [9–14], but a consensus on the optimal pacing ratio has not been reached. The purpose of this study was to evaluate and compare the hemodynamic effects of two different pacing ratios (1:1 versus 1:2) in an experimental model of dynamic double cardiomyoplasty in the setting of pharmacologically induced acute heart failure. Chronically conditioned skeletal muscles were used.
|
Material and Methods
|
|---|
Eleven adult microfilaria-free dogs weighing 15 to 20 kg were used for this investigation. This research was approved by the animal research protocol review committee of Baylor College of Medicine in accordance with the principles specified in the "Guide for the Care and Use of Laboratory Animals" published by the National Institutes of Health (NIH publication 85-23, revised 1985). Anesthesia was induced with intravenous methohexital (10 mg/kg). After endotracheal intubation the animals were ventilated with oxygen using an Ohio ventilator (Ohio Medical Products, Madison, WI) and anesthesia was maintained with isoflurane (1% to 2% to effect). Normal saline solution (40 mL/h) was used for intravenous fluid replacement. Prophylactic antibiotic (cefazolin sodium, 1 g intravenously; SmithKline and French, Philadelphia, PA) was administered before skin incision.
Surgical Technique
The left latissimus dorsi muscle was dissected from the chest wall, spinous, and scapular attachments. Intercostal perforators were ligated, and the humeral tendon was transected. The neurovascular pedicle was identified, and a pacing lead (model 4080; Medtronic, Inc, Minneapolis, MN) was secured around the thoracodorsal nerve. A portion of the third rib was resected, and the muscle was transferred into the chest cavity. The proximal end of the muscle was sutured to the periosteum of the second and fourth ribs to avoid tension on the neurovascular pedicle during stimulation. After the incision was closed, the right latissimus dorsi muscle was harvested, a nerve lead was attached, and the LDM was transferred to the chest in the same fashion. The animals were then placed in a supine position, and the heart was exposed through a median sternotomy. The pericardium was incised and suspended with silk sutures. An epicardial pacing lead (model 6917; Medtronic) was placed in the left ventricular (LV) apex to detect electrical systole. A fluid-filled left atrial catheter was placed for use in the final experiment.
The double cardiomyoplasty was performed using the techniques previously reported by our group [15] except that in these chronic experiments the latissimus dorsi muscles were not stretched back to resting anatomic length, nor were the ischemic distal portions of the LDMs trimmed and discarded. Briefly, the LDMs were placed on top of the heart, and the edges detached from the spinous process were sutured together with interrupted sutures. The muscle flaps were then placed under the heart, and the left LDM was brought over and sutured to the right ventricular epicardium. The right LDM was brought anteriorly over the LV, overlapping the previously sutured left muscle. A firm, but nonrestrictive wrap resulted from this technique. After completion of the double cardiomyoplasty, the nerve stimulating leads and ventricular epicardial sensing lead were connected to the programmable burst stimulator (model SP 1005; Medtronic) and tested. A subcutaneous pocket was made in the animal's abdominal wall for implantation of the cardiomyostimulator. The median sternotomy was closed with heavy silk suture and the subcutaneous tissue and skin were closed with continuous absorbable suture.
The animals were returned to their cages and allowed to recover from anesthesia. Pain was controlled with parenteral intramuscular butorphanol tartrate (1 to 2 mg every 4 hours as needed). Administration of oral antibiotic (cephalexin, 500 mg) was continued daily for 1 week. After a 2-week vascular delay period, conditioning of the latissimus dorsi muscles was initiated with the SP1005 cardiomyostimulator, using the training protocol designed for the current clinical trials [5]. This protocol consists of progressive synchronized skeletal muscle stimulation to slowly increase muscle workload. The cardiomyostimulator was adjusted every 2 weeks to increase the number of pulses in the stimulation burst. The final burst consisted of six pulses with a 33-ms spacing. The animals were inspected daily and their flanks palpated to assure muscle contraction.
At the end of a 12-week conditioning program, the final experiment was conducted. The animals were anesthetized and placed in a supine position. A redo median sternotomy was performed and the ascending aorta was dissected. A cuffed, pulsed Doppler aortic flow probe was placed around the ascending aorta. The right common carotid artery was exposed and a Millar pressure catheter (Millar Instruments, Inc, Houston, TX) was introduced into the LV. The right common femoral artery was exposed, and a second Millar pressure catheter was advanced into the aortic arch. The animals received a titrated dose of up to 5 mg/kg intravenous propranolol to achieve a steady state of acute heart failure. Failure was documented by a decrease in stroke volume, LV pressure, and aortic pressure of approximately 25%. Hemodynamic measurements were then obtained.
Data Acquisition
Data were recorded on a multichannel recorder (Gould, Inc, Cleveland, OH) and simultaneously captured by a Macintosh IIci computer (Apple Computer, Cupertino, CA) with a 16-channel analog-to-digital interface, under control of a data acquisition and analysis package (Acqknowledge; BioPac Systems, Inc, Goleta, CA). Data were recorded under three different conditions in the acutely failed heart. Signals were acquired initially with the cardiomyostimulator off (control protocol). The cardiomyoplasty was then stimulated to contract at a 1:2 ratio, followed, in the succeeding protocol, by a 1:1 ratio (paced protocols).
Hemodynamic Measurements
In each protocol an epoch containing contiguous data was sought. In some animals, segments that were not contiguous had to be analyzed due to occurrence of ectopic beats. If ectopic beats were observed, the pacing cycle (defined as a paced beat and the successor nonpaced beats) containing the ectopy, and the subsequent cycle, were excluded. Because a complete definition of the measurement technique is critical to understanding and comparing results in various reports, a detailed description of this process and the rationale for its selection is provided.
Previous reports have focused on the paced beat alone, either comparing it with a beat in a nonpaced protocol, or in some cases, comparing it with a nonpaced beat that occurred between paced beats. In other reports, the data compared were not clearly defined. It is our belief that evaluation of the hemodynamic effects of cardiomyoplasty, regardless of configuration, should concentrate on measures that are descriptive of the overall effect, rather than on a transient event such as the paced beat. We suggest that there is a critical need, in the literature on cardiomyoplasty, for a standardization of methods of reporting, and suggest ways to meet this need. Thus, we propose that the effects of cardiomyoplasty are best described by a procedure that includes all of the events that occur cyclically. As a result, as the periodic event consists of a paced beat and its successor nonpaced beats, this is the unit, a pacing cycle, for which results are reported. All conclusions are based on averages (cycle average) of data over complete pacing cycles. Parameters of all signals were measured for each beat in the selected epoch.
Statistics
The beginning and end of critical intervals were selected manually within the Acqknowledge package. For example, all pressure measurements were evaluated using the interval bounded by consecutive R-wave peaks in the electrocardiogram. For aortic flow, the upstroke of the flow signal defined the onset of systole, and the negative peak of the first derivative of LV pressure (dP/dt) defined its end. Within-animal statistics (mean, standard deviation, coefficient of variation) were computed for each physiologic variable in each protocol using MUMPS (MGlobal International, Orange, TX). Within the nonpaced protocols, 15 beats were included in these statistics, and in the paced protocols, 10 pacing cycles were used.
A file was produced from these averaged measurements for use by BMDP (BMDP Statistical Software, Inc, Los Angeles, CA). This file contained hemodynamic parameters for the control protocol and each beat class of the paced protocols. Each animal served as its own control. Because the sample size was small and as the data were not normally distributed in several of the variables, nonparametric tests were used to determine statistical significance. For each hemodynamic parameter the Friedman test, a nonparametric two-way repeated measures analysis of variance [16], was used to compare the cycle averages of the paced protocols with each other as well as with the control protocol. A similar set of Friedman tests was used to determine the existence of differences between beats within the 1:2 paced protocol. Wilcoxon tests were then performed, without adjustments for multiple comparisons, on the variables used in the Friedman tests. Because the purpose of multiple comparison adjustment is to make finding significant differences less likely, and, in this instance, we are trying to demonstrate that there is no difference, adjusting for multiple comparisons would create a bias in our favor. Ninety-five percent confidence intervals were constructed for the median differences using the 0.025 and 0.975 quantiles of the Wilcoxon signed rank statistic [16]. In the data, confidence intervals that do not cross zero indicate statistically significant differences, whereas those that do indicate no significant difference.
The power of a nonparametric test is usually given in terms of its asymptotic relative efficiency compared with its parametric analogue for a variety of known distributions. An asymptotic relative efficiency of unity (1.0) indicates power equal to that of the corresponding parametric test. Assuming normality, the asymptotic relative efficiency of the Wilcoxon test compared with the t test is 0.955. When the distribution of the data is not normal, the asymptotic relative efficiency of the nonparametric test can be considerably higher than 1.0 [16]. For each hemodynamic parameter, the post-hoc statistical power to detect a difference as small as 10% between the effects of the two pacing ratios was calculated based on the t test, using the variance of the observed differences and a significance level of 0.05. Because our data were frequently not normally distributed, the power of the tests we used would be expected to be higher than the t test. Therefore, the power values presented here should be considered minimums.
For LV pressure, mean aortic pressure, left atrial pressure, and cardiac output, the power was more than 80%; for stroke volume it was 66%; and for dP/dt it was 33%. This is because, for both of these parameters, the variance of the difference (16 and 28, respectively) is large compared with a physiologically important difference of 10%. The power to detect a 15% difference in effect was 93% for stroke volume and 56% for dP/dt. Assuming 80% power, the minimum detectable difference ranged from 7% for LV pressure to 12% for stroke volume and 21% for dP/dt.
The order in which the paced protocols were carried out was not randomized, that is, 1:2 pacing always preceded 1:1 pacing. Any effect on the results was predicted to be minimal as the duration of these protocols was less than 60 seconds and the muscles had been transformed to a fatigue-resistant state. After equilibration of LV pressure and stroke volume after administration of propranolol, the time interval to completion of the 1:2 protocol was about 7 minutes, and about 10 minutes to completion of the 1:1 protocol.
|
Results
|
|---|
The absolute values of the hemodynamic variables obtained in the two different paced protocols are shown in Table 1
. Stimulation of the double cardiomyoplasty at a 1:2 ratio produced significant (p values less than 0.05 are considered significant) hemodynamic augmentation compared with control (Table 2) during the paced beat in all variables except left atrial pressure, for which p = 0.0547. When both the paced and nonpaced beats are combined (cycle average) and compared with control, significance is "lost" for LV and peak aortic pressure. This is an example of how focusing on the paced beat alone can be misleading.
View this table:
[in this window]
[in a new window]
|
Table 1. . Eleven Dogs With Both 1:2 and 1:1 Pacing Ratios in Propranolol-Induced Heart Failure: Population Description
|
|
Figures 1 and 2 display typical physiologic tracings for the 1:2 and 1:1 pacing ratios, respectively. Figure 3 is a concise summary of the comparison between these two pacing ratios. For each physiologic variable the data displayed are the median of the percentage difference relative to the control, represented by the single dot for each parameter. To this is appended the 95% confidence interval for each variable, so that the hemodynamic similarity is readily demonstrated by the lack of significant differences in six of the eight variables. Although LV pressure (p = 0.0262) and its first derivative (p = 0.0099) are significantly different between the two pacing ratios, mean aortic pressure, which is not significantly different between the two pacing ratios, is the principal driving force for systemic and coronary flow. Figure 4 shows that the 1:2 pacing ratio yields values different from control in stroke volume, peak positive LV dP/dt, mean aortic pressure, heart rate, and cardiac output. Figure 5 displays the corresponding results for the 1:1 pacing ratio, in which stroke volume, peak LV pressure, peak positive LV dP/dt, and cardiac output are significantly different from control. Finally, Figure 6 shows the problem of comparing the paced and nonpaced beats within a protocol. For the 1:2 pacing ratio, all variables except mean aortic pressure are significantly different. Neither heart rate nor cardiac output is displayed here as they are not meaningful within a pacing cycle.
View larger version (82K):
[in this window]
[in a new window]
|
Fig 1. . Typical physiologic tracings at a pacing ratio of 1:2. The stimulated beat can be identified in the electrocardiogram (ECG) signal by the pulse burst. (dP/dt = first derivative of pressure; LAP = left atrial pressure; LV = left ventricular; Press = pressure.)
|
|
View larger version (82K):
[in this window]
[in a new window]
|
Fig 2. . Typical physiologic tracings at a pacing ratio of 1:1. Note that every beat is stimulated, as indicated in the electrocardiogram (ECG). (dP/dt = first derivative of pressure; LAP = left atrial pressure; LV = left ventricular; Press = pressure.)
|
|
View larger version (41K):
[in this window]
[in a new window]
|
Fig 3. . Comparison of the two pacing ratios. Each data point is the median difference between values in the 1:2 and 1:1 pacing ratios. The bars represent 95% confidence intervals, and those that cross the zero line are not significant. (dP/dt = first derivative of pressure; Lt = left; LV = left ventricular.)
|
|
View larger version (39K):
[in this window]
[in a new window]
|
Fig 4. . Comparison of the 1:2 pacing ratio to the control state. The data points represent the median difference between control and the cycle average of values from the 1:2 pacing ratio. The bars are 95% confidence intervals, and those that do not cross the zero line are significantly different. (dP/dt = first derivative of pressure; Lt = left; LV = left ventricular.)
|
|
View larger version (40K):
[in this window]
[in a new window]
|
Fig 5. . Comparison of the values from the 1:1 paced protocol and the control values. Data are presented as in the previous two figures. (dP/dt = first derivative of pressure; Lt = left; LV = left ventricular.)
|
|
View larger version (37K):
[in this window]
[in a new window]
|
Fig 6. . Comparison of the paced beat and the nonpaced beat in the 1:2 protocol. The data presentation is identical to that of the preceding figures. (dP/dt = first derivative of pressure; Lt = left; LV = left ventricular.)
|
|
|
Comment
|
|---|
The optimal pacing ratio in cardiomyoplasty has not been defined. Grandjean and colleagues [5], in a review of the phase I clinical trial, reported that in approximately half of the patients the latissimus dorsi muscle was being stimulated with every heart cycle (1:1 pacing ratio). The rationale for this is the assumption that the skeletal muscle will provide the greatest assistance by squeezing the ventricle with every contraction. However, as recent studies [7, 8] have raised the question of skeletal muscle damage as a by-product of too-frequent contraction, this choice may be inappropriate.
Other investigators have evaluated different pacing ratios in a variety of models using skeletal muscle for circulatory assistance. Lucas and colleagues [7] used single cardiomyoplasty with a conditioned LDM in goats. They observed significant lipomatosis of the muscle and recommended pacing no more frequently than 50 per minute. Pattison and co-workers [9] reported their results with aortomyoplasty in sheep, also reporting skeletal muscle damage with pacing ratios of 1:1 and 1:2. However, because one of the purposes of their study was to demonstrate that augmentation could be achieved soon after the aortomyoplasty, observations were on unconditioned LDMs. Only 1 animal each was paced at 1:1 and 1:2. They recommended against pacing ratios lower than 1:4. Badylak and colleagues [12], using unconditioned skeletal muscle ventricles, reported the degree of hemodynamic augmentation at each of three pacing ratios (1:2, 1:3, and 1:4). Although they did not report on the effects of these rates on the muscle tissue, they did indicate differing degrees of augmentation with the three different pacing regimens. However, they did not report a statistical difference among the different ratios, therefore an optimum value was not suggested.
Schreuder and colleagues [13] have demonstrated decreased LDM blood flow with muscle contraction. These investigators evaluated LDM contraction in patients after cardiomyoplasty. They observed a LDM contraction cycle that lasted 665 ms. They recommended a cycle duration of 1,300 ms (which implies a maximum of 46 muscle contractions/min) to avoid hypoxia and subsequent muscle damage. Chekanov and co-workers [14] evaluated different pacing ratios in 14 cardiomyoplasty patients operated in Russia. In this series of patients, the most significant increases in cardiac index, stroke index, and cardiac output were seen during a stimulation regimen of 1:4 or 1:6.
The purpose of the present study is to compare the two pacing ratios (1:1 versus 1:2) most commonly used clinically. Evaluation took place in a model of double cardiomyoplasty using fatigue-resistant LDMs. It is important in analyzing different pacing ratios to specify which beats are being compared in the different protocols. A standard way of analyzing data of this kind has not been described in the cardiomyoplasty literature, which may contribute to some of the conflicting results. For this reason, in this report we have carefully defined how our data are obtained and results reported. Because this method takes into account all of the events resulting from pacing, we believe it reflects more accurately the aggregate effect.
Acute heart failure was induced by intravenous propranolol. Our data show similar hemodynamic augmentation in both 1:1 and 1:2 pacing ratios. Of note is the finding that during 1:2 pacing, the nonpaced beat is always depressed compared with the paced beat or the control. A possible explanation for this phenomenon is that slow muscle relaxation after the paced beat impairs venous return, thus reducing LV filling. This would be consistent with the observation by Lucas and colleagues [11] of slowing of LDM relaxation after muscle transformation.
Ruggiero and colleagues [17] were the first group to report experience with double cardiomyoplasty in chronic studies. They observed a decrease of 3.7% in cardiac output, a 9.0% reduction in stroke volume, and an increase of 10.6% in LV pressure with double cardiomyoplasty stimulation. In their study, right ventricular pressure, which increased 31.3%, was the only parameter noted to have changed significantly with stimulation. Their earlier acute double cardiomyoplasty study had demonstrated that stimulation increased each of the hemodynamic parameters significantly. The differences observed between their acute and chronic studies were attributed to the changes in fiber type that occur during conditioning of the skeletal muscles, or to the changes in contractile force applied to the heart because of intrathoracic adhesions.
In our laboratory we have been evaluating dynamic double cardiomyoplasty. The rationale for this approach has been the clinical observation of the difficulty of wrapping severely dilated hearts with only one LDM. In contrast to the study by Ruggiero and colleagues, in which the hearts were normal, the studies we report here were conducted with the heart in an acutely failed state. Hemodynamic changes associated with chronic double cardiomyoplasty should be more readily observed in animals that have some impairment in ventricular function. Previously, we have demonstrated good hemodynamic augmentation in both acute and chronic experiments using the double cardiomyoplasty [15; Soltero ER, et al, unpublished results]. Our data now suggest that pacing at a 1:2 ratio will provide similar cardiac assist to that of a 1:1 ratio.
Van Doorn and colleagues [18] evaluated the effects of 1:1 and 1:2 stimulation on blood flow to the in situ, conditioned LDM in sheep. These investigators observed an increase in venous lactate levels after 1:1 stimulation that was not present after 1:2 stimulation. Reactive hyperemia was also noted after 1:1 stimulation, but not after 1:2 stimulation. These researchers concluded that 1:1 stimulation may be detrimental to LDM blood flow. These results may explain, in part, our observation that 1:2 pacing provides similar hemodynamic augmentation as 1:1 pacing. Theoretically, 1:1 stimulation should result in better hemodynamic performance as every heart beat is being assisted by the LDM. On the other hand, if, as van Doorn and colleagues have observed, 1:1 stimulation produces anaerobic metabolism, this is detrimental to skeletal muscle performance and reduces the beneficial effect on cardiac function. Stimulating the LDMs at a 1:2 ratio provided comparable hemodynamic support in our study. Therefore, it is reasonable to assume that the performance of the LDM is being affected by excessive electrical stimulation at a 1:1 ratio that produces muscle ischemia and could produce detrimental effects in clinical cardiomyoplasty.
It is essential that an optimal pacing ratio be determined in the future. A stimulation ratio that balances the need to preserve LDM function and still provides adequate cardiac assistance may improve long-term clinical outcome after cardiomyoplasty. This and other areas need to be thoroughly evaluated before cardiomyoplasty becomes a widely used option for patients suffering from congestive heart failure.
|
Acknowledgments
|
|---|
This work was supported, in part, by a grant from the DeBakey Medical Foundation, One Baylor Plaza, Houston, TX.
The expert technical assistance of Peggy Jackson and Gary Liedtke was instrumental in completing this study. We also thank Medtronic, Inc, for the use of their pacemakers and cardiomyostimulators, a model 9710 programmer, and associated supplies, with special thanks to David Francischelli of Medtronic for his enthusiastic assistance.
|
Footnotes
|
|---|
Address reprint requests to Dr Lawrie, Baylor College of Medicine, 6535 Fannin, MS A-801, Houston, TX 77030.
|
References
|
|---|
- Carpentier AF, Chachques J, Acar C, et al. Dynamic cardiomyoplasty: a seven year clinical experience. J Thorac Cardiovasc Surg 1993;106:42–52.[Abstract]
- Moreira LF, Stolf NA, Bocchi EA, et al. Current expectations in dynamic cardiomyoplasty. Ann Thorac Surg 1993;55:299–303.[Abstract]
- Furnary AP, Magovern JA, Christlieb IY. Clinical cardiomyoplasty: preoperative factors associated with outcome. Ann Thorac Surg 1992;54:1139–43.[Abstract]
- Chiu RCJ. Dynamic cardiomyoplasty: efficacy and mechanisms. Cardiac Chron 1992;6:1–8.
- Grandjean PA, Austin L, Chan S, Terpstra B, Burgeois IM. Dynamic cardiomyoplasty: clinical follow-up results. J Cardiac Surg 1991;6:80–8.[Medline]
- Jegaden O, Delahaye F, Finet G, et al. Late hemodynamic results after cardiomyoplasty in congestive heart failure. Ann Thorac Surg 1994;57:1151–7.[Abstract]
- Lucas CM, Van Der Veen FH, Cheriex EC, et al. Long-term follow-up (12 to 35 weeks) after dynamic cardiomyoplasty. J Am Coll Cardiol 1993;22:758–67.[Abstract]
- Kratz JM, Johnson WS, Mukherjee R, Hu J, Crawford FA, Spinale FG. The relation between latissimus dorsi skeletal muscle structure and contractile function after cardiomyoplasty. J Thorac Cardiovasc Surg 1994;107:868–78.[Abstract/Free Full Text]
- Pattison CW, Cumming DV, Williamson E, et al. Aortic counterpulsation for up to 28 days with autologous latissimus dorsi in sheep. J Thorac Cardiovasc Surg 1991;102:766–73.[Abstract]
- Cumming DV, Pattison CW, Lovegrove CA, et al. Biochemical and structural adaptation of autologous skeletal muscle used for counterpulsation. Int J Cardiol 1991;30:181–90.[Medline]
- Lucas CM, Van Der Veen FH, Cheriex EC, van Ommen V, Penn OC, Wellens HJJ. The importance of muscle relaxation in dynamic cardiomyoplasty. PACE 1992;15:1430–6.[Medline]
- Badylak SF, Stevens L, Janas W, Gray MH, Geddes LA, Voorhees WD. Cardiac assistance with electrically stimulated skeletal muscle. Med Biol Eng Comput 1989;27:159–62.[Medline]
- Schreuder JJ, van der Veen FH, van der Velde ET, et al. Beat-to-beat analysis of left ventricular pressure-volume relation and stroke volume by conductance catheter and aortic modelflow in cardiomyoplasty patients. Circulation 1995;91:2010–7.[Abstract/Free Full Text]
- Chekanov V. After cardiomyoplasty there is no need to stimulate the muscle with every heart contraction [Letter]. Ann Thorac Surg 1994;57:1370–1.
- Soltero ER, Michael LH, Glaeser DH, Hartley CJ, Earle NR, Lawrie GM. New configuration of double cardiomyoplasty based on studies of the length-tension properties of the latissimus dorsi muscle. J Thorac Cardiovasc Surg 1993;106:842–9.[Abstract]
- Conover WJ. Practical non-parametric statistics, 2nd ed. New York: John Wiley, 1980: 278–305.
- Ruggiero R, Thomas GA, Niinami H, et al. Double cardiomyoplasty: acute versus chronic effects. Ann Thorac Surg 1993;56:31–7.[Abstract]
- Van Doorn CAM, Bhabra MS, Hopkinson DN, Barman D, Cranley JJ, Hooper TL. Latissimus dorsi muscle blood flow during synchronized contraction: implications for cardiomyoplasty. Ann Thorac Surg 1996; 61:603–9.[Abstract/Free Full Text]
This article has been cited by other articles:
|
|
|
|
 
H. R. Shah, M. Vaynblat, L. Salciccioli, P. Impellizzeri, J. N. Cunningham Jr, and M. Chiavarelli
Composite cardiac binding in experimental heart failure
Ann. Thorac. Surg.,
February 1, 2000;
69(2):
429 - 434.
[Abstract]
[Full Text]
[PDF]
|
|
|
Top
Abstract
Introduction
