Numerical prediction and experimental measurement of cardiac assistance from skeletal muscle ventricles
DM Pullan, JC Jarvis, AP Shortland and S Salmons.
Depts of Human Anatomy and Cell Biology (JCJ, DMP, SS) and Clinical Engineering (APS), The University of Liverpool, UK and The Cardiothoracic Centre, Liverpool NHS Trust (DMP).
Skeletal muscle ventricles (SMVs) are hydraulic pumps formed from autografts of skeletal muscle arranged to provide extra hydraulic work to the cardiovascular system in heart failure. We have investigated one configuration of such a system in which the SMV acts as an aortic counterpulsator. The SMV is connected by a single conduit to the descending aorta and, by analogy with an intra-aortic balloon pump for example, contracts during diastole to enhance coronary perfusion and relaxes during systole to reduce the work of the left ventricle. We have investigated both by numerical modelling and now by experimental measurement how the degree of SMV assistance depends on the placement of the assist phase within the cardiac cycle. Cylindrical SMVs were made in pigs by wrapping the latissimus dorsi muscle around a PTFE former. The SMVs were electrically stimulated at 1 Hz for 4 weeks to render them resistant to fatigue. In a terminal procedure the SMVs were connected to the descending aorta via a GoreTex conduit. We recorded left ventricular volume via a conductance catheter and SMV volume via a sonomicrometer system. Pressures were measured within the left ventricle, aortic root and SMV. Flow was recorded from the aortic root, and proximal and distal to the site of anastomosis of the SMV conduit with the aorta.
The timing of SMV activation was controlled relative to the prevailing systolic and diastolic durations of the left ventricle. The delay between the QRS complex and the start of SMV activation was varied between 20% and 140% of the systolic duration and the duration of SMV activation was varied between 60% and 140% of the diastolic duration. Various timing combinations were delivered in a pseudo-random order, after a short period with the SMV OFF. The complete matrix was achieved in 6 experiments. In two of the experiments it was also possible to evaluate the isolated pressure-volume characteristic of the SMV.
Results and conclusions: The pressure-volume loops for the LV and SMV show that the timing of SMV action has a profound effect on its influence on the LV. SMV action can reduce the work done by the heart, but can also increase it if the timing is inappropriate. The experimental results contain trends similar to those predicted by our numerical model of the assisted circulation.
The support of the British Heart Foundation is gratefully acknowledged.
