Evaluation of the Infant Jarvik Ventricular Assist Device in a Chronic Juvenile Sheep Model

Abstract

Purpose The development of pediatric mechanical assist devices has been hindered by hemolysis and thrombosis. Following revision of the bearing design and with the support of the NIH initiative Pumps for Kids, Infants and Neonates (PumpKIN), the Infant Jarvik ventricular assist device (VAD) was evaluated in a chronic ovine model. Methods and Materials The Infant Jarvik VAD was implanted in 10 sheep (24kg±3.4) between the left ventricle (LV) and descending aorta. Hemodynamic data was collected via intraoperative echocardiogram and pulmonary artery and pump outflow graft flow probes. Blood was collected weekly for analysis of biocompatibility. Platelet activation was measured with ELISA quantification of p-selectin. Finally, organ pathology and VAD components were examined for evidence of injury or thrombosis. Results Average study duration was 44±19 days. As VAD flow increased, the unloaded heart maintained a cardiac output of 3 L/min, without alteration in LV fractional area change (Fig.1). There was no significant difference between baseline and study values of hematocrit (27v27%, p=0.59), platelet activation (8v9%, p=0.08), platelet count (666vs798x103/mL, p=0.12) or aspartate aminotransferase (102v125U/L, p=0.28). Creatinine (0.51v0.60mg/dL, p=0.03), alanine transaminase (15v31U/L, p=0.001) and plasma free hemoglobin (10v15mg/dL, p=0.02) remained within normal limits. There were 4 subclinical renal infarcts and no brain lesions seen on final pathology. There were two instances of graft thrombosis, one of which was secondary to infection. Conclusions With significant improvement from earlier prototypes, the Infant Jarvik VAD demonstrated adequate circulatory support, biocompatibility and feasibility for anticipated clinical use. [ figure 1 ]