Heparin sorptivity and blood compatibility of carbon surfaces.

Abstract

AbstractThe influences of the conditioning treatments, surface topography, and crystal structure of carbonaceous surfaces on their ability to sorb heparin and their in vivo compatibility with blood were investigated. The results of the sorption studies indicated that the adsorption of heparin on the surfaces of turbostratic and graphitic materials is not crystallographically selective and that the amount adsorbed on relatively smooth surfaces is near the amount expected for monolayer formation. Although the adsorption of heparin on relatively smooth carbon surfaces is not influenced by the presence of benzalkonium chloride, the sorption of heparin in porous carbons can be greatly increased by a pretreatment with benzalkonium chloride. This increase was found to be due to the formation and entrapment of the insoluble heparin‐benzalkonium complex in the accessible porosity. Since the heparin sorptions in Dag‐154 coatings were found to be enhanced by a pretreatment with benzalkonium chloride, it was inferred that these coatings contain accessible porosity and that their initial thromboresistance depends on the formation of the benzalkonium‐heparin complex in pores. In vivo tests showed that polished and outgassed, impermeable isotropic carbons deposited at low temperatures were significantly thromboresistant without the exogenous application of heparin. There was no relationship between the amount of heparin sorbed on these materials and their compatibility with blood. Polishing, for example, which reduced heparin sorption, enhanced the thromboresistance of these carbons, and while chemisorption of oxygen markedly reduced their thromboresistance, it did not influence the amount of heparin that could be sorbed. Although the heparin‐benzalkonium complex sorbed in a porous carbon conferred excellent thromoboresistance in a 2‐hr test, the long‐tern (14‐day) compatibility was not as good as for carbon surfaces that were deposited at low temperatures and then polished and outgassed prior to implanting. In vivo tests of HTI carbon structures and annealed LTI carbons indicate that the blood compatibility of a turbostratic carbon is not significantly dependent on crystallite size, Le. Limited tests of surfaces that had a preponderance of c‐faces oriented parallel to the blood‐carbon interface at the surface suggest that orientations of this sort are better than others.