Totally implantable artificial hearts and left ventricular assist devices: selecting impermeable polycarbonate...

In the development of a new generation of totally implantable artificial hearts and left ventricular assist devices (VADs) for long-term use, the selection of an acceptable material for the fabrication of the ventricles probably represents one of the greatest challenges. Segmented polyether urethanes used to be the material of choice due to their superior flexural performance, acceptable blood compatibility, and ease of processing. However, because they are known to degrade and to be readily permeable to water, they cannot meet the rigorous requirements needed for a new generation of implantable artificial hearts and VADs. Therefore, the objective of the present study was to identify alternative polymeric materials that would be satisfactory for fabricating the ventricles, and in particular, to determine the water permeability through membranes made from four commercial polycarbonate urethanes (Carbothane PC3570A, Chronoflex AR, Corethane 80A, and Corethane 55D) in comparison to those made from two traditional polyether urethanes (Tecoflex EG80A and Tecothane TT-1074A). In addition to determining the rate of water transmission through the six membranes by exposing them to deionized water, saline, and albumin-Krebs solution under pressure and measuring the displacement of liquid by means of a recently developed capillary method, the inherent surface and chemical properties of the six membranes were characterized by SEM, contact angle measurements, FTIR, DSC, and GPC techniques. The results of the study demonstrated that the rate of water transmission through the four polycarbonate urethane membranes was significantly lower than through the two polyether urethanes. In fact the lowest values were recorded with the two Corethane membranes, and the harder type 55D polymer had a lower value (2.7 x 10(-7) g/s cm2) than the softer 80A version (3.3 x 10(-7) g/s cm2). This level of water vapor permeability, which appears to be controlled primarily by a Fickian diffusion mechanism, is between 2 and 4 times lower than that obtained with traditional polyether urethane membranes of equivalent thickness. The superior performance of the polycarbonate urethanes is likely due to the inherently lower chain mobility of the carbonate structure in the soft segment phase. In addition, the study shows that additional impermeability to water vapor can be achieved by selecting a polyurethane polymer with a high hard segment content, an aromatic rather than aliphatic diisocyanate comonomer, and a more hydrophobic surface. The use of a higher molecular weight polyurethane is not necessarily efficacious if the above requirements are not met. As expected by Raoult's Law, the study found that the use of physiological media instead of deionized water further decreases the rate of water vapor transmission. Because none of today's commercial polyurethanes are totally impervious to water vapor transmission, additional work is needed to develop permeable polymers or to apply additional treatments to existing candidates to achieve an acceptable impermeable ventricle material.