Pure pyrolytic carbon: preparation and properties of a new material, On-X carbon for mechanical heart valve prostheses.

BACKGROUND AND AIM OF THE STUDY: Historically, the pyrolytic carbon used in mechanical prosthetic heart valves contained small amounts of silicon, this being a necessary additive to achieve consistently the hardness required for wear resistance. New processing technology has allowed the deposition of pyrolytic carbon without silicon, while maintaining adequate hardness to ensure wear resistance. METHODS: A parametric study of coating parameters identified the conditions necessary to produce the optimal pure carbon material. RESULTS: In comparison with silicon-alloyed carbon, the pure carbon was found to be about 20% stronger, have a strain-to-failure about 25% higher and have a greater toughness. CONCLUSIONS: The enhanced strength, deformability and toughness of the new carbon permits designers to utilize component shapes and dimensions that could not be manufactured using the silicon-alloyed carbons. Such design features have hemodynamic benefits resulting in valve performance improvements.

Dynamic characterization of a new accelerated heart valve tester.

This paper presents a new accelerated prosthetic heart valve tester prototype that incorporates a camshaft and poppet valves. A three element Windkessel system is used to mimic the afterload of the human systemic circulation. The device is capable of testing eight valves simultaneously at a rate up to 1,250 cycles/min, while the flow rate, the pressure, and the valve loading can be monitored and adjusted individually. The tester was characterized and calibrated using a set of eight Carpentier-Edwards bioprostheses at a flow rate varying between 3 and 5 L/min. The experiment was carried out with the pressure difference across the closed heart valve maintained between 140 and 190 mmHg. Smooth and complete opening and closing of the valve leaflets was achieved at all cycling rates. This confirms that the velocity profiles approaching the test valves were uniform, an important factor that allows the test valves to open and close synchronously each time.

Pyrolytic carbon indentation crack morphology.

In studying fatigue and fracture behavior of brittle materials, Vickers diamond indentation cracks are often used. Many of the studies of indentation cracks use crack system models such as the radial-median crack or Palmqvist crack. These systems are also used to study small crack growth in brittle materials, and have been studied for pyrolytic carbon. However, the true morphology of these cracks in pyrolytic carbon coatings on graphite substrates have not been described. This study examined Vickers diamond and spherical ball indentation cracks in pyrolytic carbon coatings using several techniques, including serial metallographic cross sections, indentation fracture in bending, acoustic emission, and residual surface indentation scanning. The crack systems developed using these techniques were not typical of either radial median or Palmqvist systems. The morphology is unique to this material, possibly because of the coating thickness limitations. Given the difference in crack system, the application of standard indentation crack equations in studying fracture mechanics, especially for small cracks, must be questioned.

Finite element analysis of indentation tests on pyrolytic carbon.

The stresses which cause failure at contact areas between leaflets and orifices in pyrolytic carbon heart valves are evaluated. These contact stresses have previously been studied using Hertzian crack models that apply to monolithic material. Many heart valves are not monolithic pyrolytic carbon but a pyrolytic carbon deposited on graphite. Contact loads on these layered structures cause initial cracking in the pyrolytic carbon at the interface between pyrolytic carbon and graphite rather than Hertzian surface cracks. Increasing the load on layered structures will cause a secondary cracking (of Hertzian cracks) on the surface. The contact loading was simulated with a 5.1 mm diameter ball pressing against a flat sample of graphite coated with 0.26 mm of pyrolytic carbon on each surface. Finite element analysis of this model calculated the stresses associated with a range of loads causing no cracks through initial interface cracks and secondary surface cracks to complete failure. The calculated stresses are correlated with parallel laboratory experiments. A failure criterion for contact stresses is developed. The initial cracks at the graphite/pyrolytic carbon interface occur when the tensile stress in the pyrolytic carbon reaches 207 to 276 MPa and the compression stress in the graphite reaches 414 to 483 MPa. These initial cracks do not propagate immediately to the surface since they run into a high triaxial compression stress field. Circular surface cracks occur at the edge of the ball/pyrolytic carbon contact area at higher loads. These cracks require a shear stress of about 241 MPa and also require a tensile stress component. The results provide a criterion for designing contact regions in pyrolytic heart valves.

Effect of cavitation on pyrolytic carbon in vitro.

It has long been known that mechanical heart valves, when tested for durability using non-physiologic conditions common in accelerated testers, would cavitate. Until recently, cavitation was never observed in vivo. The discovery that a small number of Edwards-Duromedics heart valve explants indicated signs of cavitation erosion prompted a reassessment of the cavitation erosion potential of pyrolytic carbon (PyC). Analyses of the explanted valves indicated that cavitation may be accentuated by porous regions in the pyrolytic carbon coating from which most mechanical heart valves are constructed. Early studies have shown that for PyC (a) the resistance to cavitation erosion is comparable to that of aluminum, (b) the resistance to cavitation erosion is high initially, but with time the erosion rate accelerates, and (c) the cavitation erosion resistance is somewhat variable. In this study, similar experiments were performed utilizing polished pyrolytic carbon as well as microporous surfaces since microporous surfaces have been implicated as accelerating erosion. Within the accuracy of the measurement, we found no contributing acceleration due to the microporous nature of the pyrolytic carbon surfaces tested when compared to the polished surfaces. Examination of cross sections of samples exposed to cavitation conditions revealed the presence of extensive microcracking even without the presence of substantial surface erosion.

On the durability of pyrolytic carbon in vivo.

The durability of pyrolytic carbon heart valve components was examined from the point of view of number implanted, documentation of wear on explanted components, and from the aspect of fatigue. Failures of pyrolytic carbon components were found to be few in number. A model describing the time course of events and successful usage of pyrolytic carbon components in heart valves was developed. The model is based on the yearly shipments of pyrolytic carbon components or valves, starting with 1969. The model indicates that about 2 million components have been successfully implanted, resulting in accumulative experience of over 10 million patient years. The wear of pyrolytic carbon, based on analysis of explanted heart valves, was found to be minimal. Additional wear data obtained on explanted components confirms the earlier observation that wear in vivo in less than that observed in vitro. The recently discovered fatigue behavior of pyrolytic carbon was found to have no demonstrated practical impact on the durability of pyrolytic carbon components used in existing mechanical heart valves.

A histologic and mechanical evaluation of carbon-coated polyester suture.

Ultra low temperature isotropic (ULTI) carbon-coated polyester suture material was evaluated histologically and mechanically in dogs. These results were compared to those obtained for uncoated polyester and polybutylate coated polyester. The suture materials were used in the repair of the surgically incised medial collateral ligament and subcutaneous tissues to evaluate the potential of the carbon-coated system for ligamentous repairs. Following surgery, the dogs were sacrificed at periods of 1-48 weeks postoperatively for evaluation of tissue biocompatibility and mechanical strength of the materials. The polybutylate-coated polyester suture broke at lower force levels than did comparable sizes of uncoated or carbon-coated polyester. All three types showed a high retention (greater than 98%) of mechanical strength at 48 weeks, often exhibiting an increase in tensile strength due to tissue ingrowth. The histologic response to carbon-coated polyester was equal to or better than the response to either the uncoated polyester or polybutylate-coated polyester. A greater degree of tissue growth into the carbon-coated material was evident at most time periods following an initial acute inflammatory response which was also present in the other materials.

The effect of surface treatments on the interface mechanics of LTI pyrolytic carbon implants.

A mechanical and histological evaluation of LTI pyrolytic carbon implants was undertaken to determine the effect of various surface treatments on the retention characteristics of the implants. Five types of surfaces were evaluated, including as-deposited, fine grit-blasted, coarse grit-blasted, ground, and plasma oxygenated. The four surface treatments were chosen in an attempt to emulate the morphology of the as-deposited implants. The implants were evaluated in vivo by placement transcortically in the femora of adult mongrel dogs for periods of 12 to 24 weeks. Although the as-deposited implants exhibited the greatest interface strength at 12 weeks the results of mechanical testing after 24 weeks implantation indicated no statistically significant difference among the interface strength values or among the interface stiffness values of the implants. The histologic response of the implants was similar; while all implants exhibited areas of direct implant-bone apposition, the as-deposited implants exhibited this behavior to the greatest extent. Thus the ability to duplicate the biological response to the as-deposited LTI carbon surface appears possible by one or more of the treatments evaluated.

Protein adsorption in vitro onto biomaterial surfaces covered with ULTI carbon.

A study is presented which describes the adsorption in vitro of albumin and fibrinogen onto a mini-shunt oxygenation unit. The unit is characterized by flow through 10 etched microchannel conduits with a microporous membrane and conduits are pretreated with a uniform layer of ULTI carbon and evaluated for the steady-state and time varying kinetics of protein adsorption. Corresponding results for oxygenator components not pretreated with ULTI carbon are also included.

Oxygen permeability of carbon-surfaced microporous membranes.

A study was conducted to measure the permeability to oxygen (PmO2) of microporous membranes coated with ultra-low-temperature isotropic (ULTI) carbon. Carbon-surfaced membranes were fabricated by application of a thin complete layer (0.7 to 1.7 microns thick) of ULTI carbon onto one surface of a microporous substrate. Based upon the average values of 27 measurements in a diffusion cell, a mean reduction in PmO2 of only 10% results following the treatment of the microporous substrate with ULTI carbon. The values for PmO2 are not dependent upon the thickness of the carbon film over the range of film thickness of 0.7 to 1.7 microns. When coupled with the known antithrombotic characteristics of ULTI carbon, an oxygenator incorporating carbon-surfaced microporous membranes has the potential for use in clinical extracorporeal membrane oxygenation (ECMO).

The fatigue behavior of vapor-deposited carbon films.

Vapor-deposited carbon films (about 4000 to 5000 A thick) on stainless steel substrate were cyclically loaded to 10(6) cycles. The carbon films did not fail in fatigue at strain levels up to 13.12 x 10(-3). Rather, the failure in the carbon film occurred as a result of plastic deformation in the substrate; i.E., the failure was directly related to the endurance limit of the substrate material, which, when expressed as strain, was measured in this study to be about 8.0 - 10.88 x 10(-3). The endurance limit was also found to be very close to the elastic strain limit of the substrate. The implications of the findings for the use of carbon coated components in prosthetic devices are also discussed.

Gaseous flow through thin carbon films.

Thin carbon films, when used as coatings on prosthetic devices, must be a barrier to gases and physiological fluids. Using CO2 at room temperature, the gas permeability of carbon films ranging in thickness from about 200 to 500A was measured. The average permeability constant of 21 carbon films was determined to be 1.91 (+/- 1.02) x 10(-12) cm3-cm/cm2-sec-mmHg. This value is quite comparable to or smaller than that of nuclear graphites, which are considered to be impermeable to gases.

Developments in carbon prosthetics.

The majority of carbon-coated prosthetic devices in use today are coated with a unique form of carbon, low-temperature isotropic (LTI) carbon. The wide acceptance of this special form of carbon is a direct result of LTI carbon's demonstrated biocompatibility, its mechanical properties, and its inertness. The LTI carbon deposition process, however, places severe constraints on the size and type of substrate that can be coated. The substrates must be small so that they may be supported in a fluidized bed and further must be able to withstand temperatures in excess of 1200 degrees C. Recent technological advancements have removed the requirement that an object to be coated must be suspended in a fluidized bed and have also made possible the deposition of isotropic carbon at near room temperature. These developments expand the application of carbon-surfaced components into areas of prosthetics not previously possible. This paper describes some of the new applications and results.

The adhesion of thin carbon films to metallic substrates.

As part of the development of carbon-coated prosthetic devices, the adhesion of thin carbon films to metallic substrates has been studied. The bond strength of carbon films about 5000 A thick on Ti-6A1-4V and stainless steel was measured in a pull test and found to be greater than 4700 psi. Auger electron spectroscopy showed a reactive film/substrate interface. The ultimate bond strength was found to be dependent on the substrate and the deposition parameters.