Optimization of Programming in Dynamic Cardiomyoplasty : Enhancing Synchronization of Muscle Wrap and Ventricular Contractions

Introduction: To obtain maximal benefit from dynamic cardiomyoplasty, stimulation parameters of the cardiomyostimulator need to be optimized. The two components for optimization include the muscle channel and the synchronization channel. This study looks at the issue of timing muscle contraction to ventricular contraction. Specifically, we compared 2 methods of obtaining optimal synchronization delay, 1) using the current clinical method of echocardiographic assessment of mitral valve closure with onset of burst stimulation, and 2) simultaneous open measurements of muscle and ventricular pressure generation. Methods: A left latissimus dorsi (LD) cardiomyoplasty was performed in 4 dogs. Using a previously established protocol for muscle transformation, the LD muscle was continuously stimulated over 4 weeks with an epineural cuff and the Itrel II myostimulator. After the training period, a Cardiomyostimulator (model #4710) was used for assist, and optimal synchronization timing was assessed. First, m-mode echocardiography was used to optimize timing of burst stimulation spikes with closure of mitral valve. Secondly, a median sternotomy was performed, and Millar pressure measuring catheters were inserted into the left ventricle (LV) and the LD muscle. Optimization of synchronization delay was again assessed with simultaneous measurement of ventricular pressure and the LD intramuscular pressure. Optimal delay for the open method was determined by best overlap of peak pressures of muscle and ventricle. Comparison was made between optimal synchronization obtained by echocardiography and the open method. Results: All dogs survived the surgery without complication, and underwent the full period of muscle transformation, followed by optimization studies. One dog (#4) appeared to be in heart failure during the optimization study. In all dogs, optimal stimulation delay as measured by the catheter technique was shorter than that determined through m-mode echocardiography. Of note, 1) peak pressure (PMax) generation of muscle occurred 70-100 msec after LV PMax in 3 of 4 dogs (in one dog with failure, muscle PMax occurred 10 msec after LV PMax), and 2) in 3 of 4 dogs, duration of muscle contraction exceeded ventricular contraction, making diastolic impairment difficult to avoid. Conclusions: From our results, we conclude that optimal synchronization delay occurs at a shorter delay than as determined by the echocardiographic method. In chronically stimulated muscle, Pmax occurs later than LV Pmax, and duration of muscle contraction is longer than LV contraction. A dilemma exists for optimizing timing of muscle contraction where maximal systolic assist is achieved with minimal diastolic interference. Because peak LV wall stress occurs early in systole, and most LV filling occurs early in diastole, we suggest that muscle contraction should occur as early as possible even at the cost of some late diastolic interference.