Ann Thorac Surg 1995;59:74-77
© 1995 The Society of Thoracic Surgeons
Optimizing ``Delay Period'' for Burst Stimulation in Dynamic Cardiomyoplasty
Joe Helou, MD,
Yoshio Misawa, MD,
James A. Stewart, MD,
Michael Colson, MS,
Ray C.-J. Chiu, MD, PhD
Division of Cardiovascular and Thoracic Surgery and Division of Cardiology, McGill University, Montreal, Quebec, Canada; and Medtronic, Inc, Minneapolis, Minnesota
Accepted for publication June 23, 1994.
 |
Abstract
|
|---|
Hemodynamic evidence of systolic assist after dynamic cardiomyoplasty remains inconsistent. One of the relevant factors may be how the burst stimulator is programmed. In 3 patients who underwent cardiomyoplasty for idiopathic dilated cardiomyopathy, we examined the modes used to determine the delay period between the R-wave sensing and the onset of burst stimulation during cardiac systole. These modes include the fixed time mode, the valve-synchronized mode, and the flow-optimized mode. The rationale for choosing these modes and the benefits conferred by each are discussed.
|
Introduction
|
|---|
Clinical trials evaluating dynamic cardiomyoplasty as a therapy for selected patients with heart failure are ongoing [1]. In this procedure, the latissimus dorsi muscle is raised as a pedicled flap and wrapped around the ventricles. After a period of graded stimulation to make the muscle fatigue resistant, burst electrical stimulation is delivered to the muscle in synchrony with cardiac systole. One of the ways in which this procedure may benefit patients suffering from heart failure is to provide systolic assist. Nevertheless, the hemodynamic evidence of systolic assist after dynamic cardiomyoplasty has been inconsistent [2], and the reasons for this inconsistency are not fully understood. The various aspects to the delivery of burst stimulation, such as the burst frequency, burst duration, and synchronizing delay period, may be the relevant determinants of hemodynamic efficacy. The optimization of both the burst frequency [3] and duration [4] has been examined by a number of investigators. In this study, we focused on the methods of selecting the synchronizing delay period between R-wave sensing and the onset of burst stimulation [5] to optimize this important aspect of programming (Fig 1
).
View larger version (16K):
[in this window]
[in a new window]
|
Fig 1. . Definition of synchronizing delay period for burst stimulator used in dynamic cardiomyoplasty. (ECG = electrocardiogram.)
|
|
|
Case Presentations
|
|---|
Patient 1
A 62-year-old man with idiopathic dilated cardiomyopathy and increasingly severe congestive heart failure was referred for heart transplantation but was not accepted into the program because of the shortage of donor hearts and his relatively advanced age. Preoperatively his ejection fraction was 0.09, as determined by echocardiography and radionuclide scanning and he was in New York Heart Association (NYHA) functional class III. Although he had occasional premature ventricular contractions, he was in normal sinus rhythm with no conduction abnormalities. He underwent dynamic cardiomyoplasty in September 1990, and his burst stimulus was programmed to be delivered 25 ms after the R wave. He has done very well postoperatively for the past 3.5 years, with marked improvement in his functional status; at present, he is in NYHA class I. Hemodynamic evidence of systolic assist includes doubling of the stroke volume from 35 to 70 mL and an ejection fraction that is now 0.20 [1].
Patient 2
A 65-year-old dentist with idiopathic dilated cardiomyopathy and progressively worsening congestive heart failure in NYHA class III to IV was advised to seek heart transplantation but was not accepted by several programs because of his age and a history of ulcerative colitis. His preoperative ejection fraction was 0.08 and he had a markedly dilated heart. He was in sinus rhythm with left bundle-branch block and had multiple premature ventricular contractions. He underwent dynamic cardiomyoplasty in November 1991. Because of his left bundle-branch block, the onset of burst stimulation was timed by echocardiography to be synchronized with mitral valve closure (Fig 2). The delay period between the R-wave sensing and the onset of burst stimulation selected for this patient was 100 ms.
View larger version (128K):
[in this window]
[in a new window]
|
Fig 2. . Example of valve-synchronized mode in which the onset of burst stimulation coincides with mitral valve closure, as determined by two-dimensional echocardiography. Arrow shows mitral valve closure. The heart rate to burst stimulation ratio is 2:1.
|
|
One week after operation, he had an episode of sudden ventricular fibrillation, but was successfully resuscitated. Over the following year, he showed functional improvement (NYHA class II); his ejection fraction increased to 0.15 and cardiac index from 1.6 to 2.2 L • min-1 • m-2 [1]. He died suddenly during sleep 15 months after the operation.
Patient 3
A 58-year-old man had idiopathic dilated cardiomyopathy. His congestive heart failure (NYHA class III) was refractory to aggressive medical therapy, and he was considered for heart transplantation. However, because of a recent history of prostatic cancer, he was not deemed a suitable candidate for cardiac transplantation and the subsequent immunosuppression regimen required. He also had a markedly dilated heart with an ejection fraction of 0.09. He was in normal sinus rhythm, but had a left bundle-branch block. Dynamic cardiomyoplasty was performed in June 1993. The onset of burst stimulation was timed to maximize the aortic root blood flow velocity, as determined by Doppler echocardiography. The latissimus dorsi muscle was stimulated at 30 Hz with every other cardiac cycle, and the delay period was varied between 25 and 150 ms in 25-ms increments; the aortic instantaneous peak flow velocity was measured in the left ventricular outflow tract 1 cm below the aortic valve at end expiration (Fig 3). At 6 weeks postoperatively, there was a 30% increase in the aortic peak flow velocity when the burst stimuli were delivered 75 ms after the R wave, as compared to that obtained near the time of mitral valve closure, which was 45 ms after the R wave (Fig 3). This finding was again confirmed 3 months after operation (Figs 3, 4). At present, 8 months after cardiomyoplasty, he is doing well; he is in NYHA class II and his ejection fraction ranges from 0.11 to 0.15.
View larger version (18K):
[in this window]
[in a new window]
|
Fig 3. . Relationship between the delay period and the peak aortic root blood flow velocity, the latter determined by Doppler ultrasonography, in Patient 3 at 6 and 12 weeks after operation. Heart rates remained stable during each programming period.
|
|
View larger version (111K):
[in this window]
[in a new window]
|
Fig 4. . (A) Doppler flow velocity in aortic root obtained at end expiration, with the delay period programmed for 50 ms. Mitral valve closure occurred at 45 ms after the R wave in this patient. (B) Similar flow velocity recording when the delay period was programmed for 75 ms in the same patient. This is the highest peak velocity obtained, and represents the flow-optimized mode of selecting the delay period.
|
|
|
Comment
|
|---|
These cases illustrate three modes of selecting the delay period for the burst stimulation in dynamic cardiomyoplasty: the fixed time mode, the valve-synchronized mode, and the flow-optimized mode.
The fixed time mode was used in patient 1. As the name implies, the timing of the burst stimulus was fixed at 25 ms, corresponding to the usual time interval between the R wave and the onset of mechanical cardiac systole in patients with a normal conduction system. This mode is simple to program because it does not require echocardiography and it has been found to be adequate for many patients with no conduction abnormalities. Difficulties may arise, however, in those patients whose mechanical cardiac systole is delayed because of conduction defects such as left bundle-branch blocks, as illustrated in patient 2. If a fixed time mode of 25 ms is used for the onset of burst stimulation in such patients, the latissimus dorsi muscle wrapped around the heart may contract in late diastole before the onset of mechanical cardiac systole.
This problem has been addressed by several investigators [6] through the use of the valve-synchronized mode. In this mode, two-dimensional echocardiography is used to time the onset of the burst stimulus to occur at mitral valve closure, ensuring that the burst stimulus is delivered at the start of mechanical cardiac systole, and thus preventing interference with diastolic ventricular filling. The firing of the burst stimulus before mitral valve closure and the increasing left ventricular pressure while the mitral valve remains open could aggravate the mitral regurgitation that is so common in patients with dilated cardiomyopathy. The valve-synchronized mode is depicted in Figure 2
. Because the stimulator used (Medtronic SP-1005; Medtronic, Inc, Minneapolis, MN) allows for programming the delay periods in 25-ms increments, the setting nearest to the desired delay is chosen.
Molteni and colleagues [7] have shown that changes in the delay time between the QRS sensing and their single-pulse stimulation in cardiomyoplasty significantly affect the peak flow velocity of blood in the ascending aorta. As shown by our patient 3, the delay periods chosen either by the fixed time mode or by the valve-synchronized mode may not induce the heart to deliver maximum output on the assisted beats. As can be seen in Figures 3 and 4, the maximum flow velocity in the left ventricular outflow tract was obtained with a 75-ms delay period, which resulted in a flow that was 30% greater than that obtained when the onset of the burst stimulus was synchronized with mitral valve closure, that is, 45 ms after the sensing of the R wave. Likewise, the fixed time mode with a 25-ms delay resulted in lower flow velocity. Flow velocity (v) in the ascending aorta with cross-sectional area (A) is related to the stroke volume (SV) of the heart, as SV = A
vdt. When the systolic interval is relatively constant, the changes in maximum flow velocity in the ascending aorta represent a good approximation of the changes in stroke volume. Thus, the rationale for the flow-optimized mode is to maximize stroke volume and cardiac output by the assisted heart. However, this raises the question: why would the delay period chosen to achieve maximum aortic flow velocity not coincide with the delay period selected using the valve-synchronized mode?
Geddes and associates [8] studied the timing of muscle contraction in a canine cardiomyoplasty model, and found the optimal delay period to be an average of 58 ms (range, 40 to 80 ms), which produced maximal augmentation of left ventricular function, including the stroke volume. They theorized that to obtain the maximum precontraction load on the muscle encircling the ventricle, it is desirable to cause it to contract late in the isovolumic period when the ventricles bulge maximally. In dogs, the isovolumic period starts from 10 to 50 ms after the R wave. In humans, who have lower heart rates, the delay will be longer. Furthermore, there are additional delays that are not accounted for with the valve-synchronized mode. These include the conduction velocity of the thoracodorsal nerve, the transmission of the impulse across the neuromuscular junction, the contraction velocities of the myocardium and the skeletal muscle wrap, the vector of the force related to muscle fiber orientation, the fluid dynamics of blood in the ventricle, and so on, all of which exhibit interpatient variability. Thus, one would expect each patient to manifest a unique timing sequence. By timing the onset of the burst stimulus to maximize the desired output, the flow-optimized mode takes into account these additional delays.
The postoperative improvements in the ejection fraction and stroke volume in our patients described here, may, in fact, reflect the effectiveness of cardiomyoplasty in providing systolic assist when an optimal burst stimulation program is chosen, although long-term changes in ejection fraction and stroke volume can readily be influenced by changes in such conditions as preload and afterload. The changes observed in aortic flow velocity over a relatively short time, as illustrated in Figure 3, during which the preload and afterload conditions remained stable, indicate that the magnitude of systolic assist can indeed be affected by the stimulator programming characteristics chosen. The findings yielded by recent clinical and experimental studies of cardiomyoplasty suggest that there are several mechanisms that may be responsible for effecting the clinical improvements seen in more than 80% of the patients who undergo and survive cardiomyoplasty procedures worldwide [10]. These mechanisms include a myocardial sparing effect resulting from reduced myocardial stress [11], the remodeling and delaying of ventricular dilation as a result of the so-called girdling effect [12], and employing the stimulated muscle wrap as the source of collateral blood flow for ischemic myocardium [13] (Fig 5). Nevertheless, improving systolic function without increasing the myocardial stress and oxygen consumption remains a highly desirable goal of cardiomyoplasty [14]. Thus, optimization of the programming aspects of the burst stimulator, including selection of the best delay period between the sensing of the systolic event and the onset of burst delivery, should be an important consideration in the postoperative management and follow-up of patients with cardiomyoplasty. Further trials to confirm the superiority of this method are warranted.
|
Footnotes
|
|---|
Presented at the Cardiovascular Science and Technology Conference, Washington, DC, Dec 9-12, 1993.
Address reprint requests to Dr Chiu, The Montreal General Hospital, 1650 Cedar Ave, Room C9.169, Montreal, Quebec, Canada, H3G 1A4.
|
References
|
|---|
- Chiu RC-J, Odim JNK, Burgess J. Responses to dynamic cardiomyoplasty in idiopathic dilated cardiomyopathy. Am J Cardiol 1993;72:475–9.[Medline]
- Chiu RC-J. Dynamic cardiomyoplasty: an overview. PACE 1991;14:577–584.
- Chiu RC-J, Walsh GL, Dewar ML, DeSimon JH, Khalafalla A, Ianuzzo D. Implantable extra-aortic balloon assist powered by transformed fatigue resistant skeletal muscle. J Thorac Cardiovasc Surg 1987;94:694–701.[Abstract]
- Pekarsky VV, Akhamdov ShD, Dubrovsky IA, et al. Optimal electrical stimulation for latissimus dorsi muscle after dynamic cardiomyoplasty. J Cardiac Surg 1993;8:172–6.[Medline]
- Grandjean PA, Herpers L, Smits K, Bourgeois IM. Performance of the cardiomyostimulator and intramuscular leads in the cardiomyoplasty procedure. In: Chiu RCJ, Bourgeois IM, eds. Transformed muscle for cardiac assist and repair. Mount : Futura, 1990:233.
- Moreira LFP, Stolf NAG, Jatene AD. Benefits of cardiomyoplasty for dilated cardiomyopathy. Semin Thorac Cardiovasc Surg 1991;3:140–4.[Medline]
- Molteni L, Almada HE, Ferreira RF, Ortega D. Assessment of optimal time interval between QRS and single-pulse stimulation in dynamic cardiomyoplasty. In: Chiu RC-J, Bourgeois IM, eds. Transformed muscle for cardiac assist and repair. Mount Kisco, NY: Futura, 1990:195.
- Geddes LA, Janas W, Bourland JD, Cook J, Hinds M. The importance of timing muscle contraction in dynamic cardiomyoplasty. PACE 1993;5:2255–65.
- Mulhorn HT Jr. The application of control theory to physiological systems. Philadelphia: Saunders, 1966:79.
- Grandjean PA, Austin L, Chan S, Terpstra B, Bourgeois IM. Dynamic cardiomyoplasty: clinical follow-up results. J Cardiac Surg 1991;6:80–8.[Medline]
- Nakajima H, Niinami H, Hooper TL, et al. Cardiomyoplasty: probable mechanisms of effectiveness using the pressure-volume relationship. Ann Thorac Surg 1994;57:407–15.[Abstract]
- Capouya ER, Gerber RS, Drinkwater DC, et al. Girdling effect of non-stimulated cardiomyoplasty on left ventricular function. Ann Thorac Surg 1993;56:867–71.[Abstract]
- Mannion JD, Magno MG, Buckman PD, et al. Acute electrical stimulation increases extramyocardial collateral blood flow after a cardiomyoplasty. Ann Thorac Surg 1993;56:1351–8.[Abstract]
- Chiu RC-J. Dynamic cardiomyoplasty: efficacy and mechanisms. Cardiac Chron 1992;6:1–8.
This article has been cited by other articles:
|
|
|
|
 
K. M. Vural, O. Tasdemir, S. D. Kucukaksu, O. K. Tarcan, and K. Bayazit
Optimization of Synchronization Delay in Latissimus Dorsi Dynamic Cardiomyoplasty
Ann. Thorac. Surg.,
May 1, 1998;
65(5):
1231 - 1234.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
H. J. Patel, D. J. Polidori, J. J. Pilla, T. Plappert, D. Kass, M. S. J. Sutton, E. B. Lankford, and M. A. Acker
Stabilization of Chronic Remodeling by Asynchronous Cardiomyoplasty in Dilated Cardiomyopathy : Effects of a Conditioned Muscle Wrap
Circulation,
November 18, 1997;
96(10):
3665 - 3671.
[Abstract]
[Full Text]
|
|
|
|
|
|
|
R. Lorusso, O. Alfieri, F. van der Veen, and J. Schreuder
Optimizing Cardiomyoplasty Results: Importance of Muscle Contraction Timing
Ann. Thorac. Surg.,
December 1, 1995;
60(6):
1862 - 1863.
[Full Text]
|
|
|
Top
Abstract
Introduction
