Induction of ventricular collapse by an axial flow blood pump.

An important consideration for clinical application of rotary blood pump based ventricular assist is the avoidance of ventricular collapse due to excessive operating speed. Because healthy animals do not typically demonstrate this phenomenon, it is difficult to evaluate control algorithms for avoiding suction in vivo. An acute hemodynamic study was thus conducted to determine the conditions under which suction could be induced. A 70 kg calf was implanted with an axial flow assist device (Nimbus/UoP IVAS; Nimbus Inc., Rancho Cordova, CA) cannulated from the left ventricular apex to ascending aorta. On initiation of pump operation, several vasoactive interventions were performed to alter preload, afterload, and contractility of the left ventricle. Initially, dobutamine increased contractility and heart rate ([HR] = 139; baseline = 70), but ventricular collapse was not achievable, even at the maximal pump speed of 15,000 rpm. Norepinephrine decreased HR (HR = 60), increased contractility, and increased systemic vascular resistance ([SVR] = 24; baseline = 15), resulting in ventricular collapse at a pump speed of 14,000 rpm. Isoproterenol (beta agonist) increased HR (HR = 103) and decreased SVR (SVR = 12), but ventricular collapse was not achieved. Inferior vena cava occlusion reduced preload, and ventricular collapse was achieved at speeds as low as 11,000 rpm. Esmolol (beta1 antagonist) decreased HR (HR = 55) and contractility, and ventricular collapse was achieved at 11,500 rpm. Episodes of ventricular collapse were characterized initially by the pump output exceeding the venous return and the aortic valve remaining closed throughout the cardiac cycle. If continued, the mitral valve would remain open throughout the cardiac cycle. Using these unique states of the mitral and aortic valves, the onset of ventricular collapse could reliably be identified. It is hoped that the ability to detect the onset of ventricular collapse, rather than the event itself, will assist in the development and the evaluation of control algorithms for rotary ventricular assist devices.