Nanomedicine, Volume IIA: Biocompatibility

© 2003 Robert A. Freitas Jr. All Rights Reserved.

Robert A. Freitas Jr., Nanomedicine, Volume IIA: Biocompatibility, Landes Bioscience, Georgetown, TX, 2003 Pyrolytic or Low Temperature Isotropic Carbon

Originally developed in the early 1960s by Gulf General Atomic as a coating for nuclear fuel rods [938, 1038-1040], pyrolytic carbon is formed in a fluidized bed by the pyrolysis of a gaseous hydrocarbon such as methane depositing carbon onto a preformed substrate such as polycrystalline graphite at 1000-1500 K [903, 955]. Its strength and ability to absorb impact energy is ~4 times greater than that of glassy carbon [903]. Pyrolytic or low temperature isotropic carbon (LTIC) is characterized by a microporous, oxidized, high-energy hydrophobic and domain-mosaic structure [806, 906]. LTIC examined by STM shows atomic lattices with many disordered defects, and patchlike carbon crystallites with sizes of 2-8 nm [792]. As with glassy carbon, the different orientations of the crystallites create surface domains that may differ in surface properties [792], producing a poorly characterized molecular surface.

The pyrolytic carbon surface has strong interactions with adsorbed proteins [806] and even with DNA [807]. LTIC adsorbs and denatures all proteins without preference [806], including albumin, fibrinogen, and some other small proteins. This is probably due to hydrophobic interactions, although the presence of air at the carbon-water interface can prevent surface denaturation of fibrinogen [808]. Protein adsorption has been directly visualized on LTIC [809], and the steady state and time varying kinetics of fibrinogen and albumin protein adsorption onto ULTI [810] (see below) and LTIC [811-813] surfaces have been studied. For instance, the adsorption of human fibrinogen onto LTIC at pH 7.2 and 25 oC was 5.2 mg/m2 (~9100 molecules/micron2) and the net heat of protein sorption was measured as 3-4 x 106 Kcal/mole of adsorbed fibrinogen [811]. Tests for adsorption of bovine fibrinogen and human serum albumin (HSA) from buffered HSA solution found that both proteins are tenaciously bound to Pyrolite (an LTI pyrolytic carbon) [812]. Note the authors: “The [high] thromboresistance of Pyrolite may be partly due to the lowered reactivity of the platelet binding domain, as well as a lessened ability of tightly bound fibrinogen to interact with platelets” [812]. In general, the rate of protein adsorption is high, the surface concentration is large, and the surface strongly retains proteins such that they cannot be displaced by buffer or exchanged by proteins in solution [806]. Thus LTIC accomplishes its blood compatibility by establishing a passivating film of strongly adsorbed bland proteins which do not interact with platelets nor participate in blood coagulation [806]. The adsorption of albumin onto a pyrolytic carbon surface has been computationally simulated using molecular mechanics techniques [4840].

For long-term exposures to blood, pyrolytic carbon is generally considered to be a relatively nonthrombogenic material [808]. This is one reason for its extensive use in artificial heart valves (Section For example, in one experiment [813] mechanical heart valves with pyrolytic carbon surface were implanted in the mitral position of sheep without the administration of post-operative anticoagulants or antiplatelet agents for 2, 4, and 6 weeks, then were removed and examined by scanning electron microscopy. Surfaces appeared clean to the naked eye, but when observed by electron microscopy the surfaces were mottled, mainly by solitary platelets and aggregations, though only a few leukocytes or red blood cells were observed and there were no fibrin clots on the leaflets. The density of platelet deposition was higher in the vicinity of the pivots and near the edges of the leaflets, with the sizes of the platelet aggregations decreasing with longer duration. The outer surfaces of the pivot guards were covered by varying amounts of deposition composed of platelet aggregations and thrombi. There is some evidence that LTIC surfaces may be conditioned (e.g., reducing platelet retention) by adsorption of a passivating protein such as albumin [810, 926]. Platelet adherence on LTIC is remarkably low compared to other implant materials, typically ~0.0043 platelets/micron2 after a 5-minute exposure to fresh human blood flowing at a wall shear rate (Section of 50 sec-1 [1680].

For brief exposures to blood, however, this material is far from ideal. A comparison of the thrombogenicity of heart valve materials found that pyrolytic carbon disks implanted in the intrathoracic venae cavae of anesthetized sheep for only 2 hours elicited significantly more thrombus formation than did titanium or cobalt-chromium disks, and more leukocyte adhesion than on pure titanium disks [814]. Recent investigations [908] of very pure heart-valve pyrolytic carbon suggests this material may have improved properties relative to traditional LTIC, which usually contains substantial amounts of silicon additive (up to 15% Si by weight) that was believed necessary to consistently achieve the hardness required for adequate wear resistance. (The silicon is present in the microstructure as discrete, lacey networks of silicon carbide particles [955].)

LTIC is generally biocompatible with cells. For example, porcine aortic endothelial cells were cultured on pure Dacron and on vapor-deposited pyrolytic carbon-coated Dacron vascular prostheses [815]. Cell adherence was unaffected, but cellular growth occurred only on carbon-coated Dacron. SEM images showed rounded adherent cells on Dacron but extensively spread cells on carbon-coated prostheses [815]. Similarly, an isotropic carbon coating on dental replicas implanted in dog mandibular arches showed good permucous acceptability (hard to obtain in other materials) and good anchorage to the surrounding bone [816]. LTIC dental implants in baboons found a good clinical response [907]. Most specimens showed a complete absence of bone resorption of the alveolar crest and an absence of epithelial cell migration or fibrous tissue formation at the implant-tissue interface. Pyrolytic carbon is also a biologically compatible material for arthroplasty of diseased finger joints [817]. In this study [817], no adverse remodeling or resorption of bone was seen. 94% of the implants had evidence of osseointegration with sclerosis around the end and shaft of the prosthetic stems. A few instances of chronic inflammatory tissue were seen, but there was no evidence of intracellular particles or particulate synovitis.

Once again, however, the material is far from ideal. When LTI pyrolytic carbon transcortical (bone) implants were placed in the femora of mongrel dogs for 6 months, the bone formed a direct appositional interface with the LTI carbon. But the strength of appositional attachment was at least one order of magnitude weaker than bone growth attachment to porous titanium and carbon-coated porous titanium systems which were also tested [818]. (The presence of carbon coating enhanced bone ingrowth [818].) As with glassy carbon, in at least one experiment [801] pyrolytic carbon or Pyrolite implanted in rabbit mandibular tissues for 0.5-3 months elicited fibrous connective tissue capsule formation, multinucleated phagocytic cells, a mild inflammatory infiltrate, and reactive bone. Thin capsules were also observed surrounding the ends of pyrolytic carbon catheters implanted intraperitoneally in dogs for 12 weeks [895].

What about pyrolytic carbon particles? Helbing et al [902] tested 97.3% pure LTIC dust (2.7% graphite) of particle size <~ 1 micron by injecting the particles intravenously, intraperitoneally, and intra- and peri-articularly into 60-day-old Chbb-strain rats. There was a slight inflammatory reaction with an increase in neutrophils in the peritoneal fluid after 24 hours. Some carbon particles were phagocytosed by macrophages. After 12-24 weeks, some foreign body granulomas had formed around large aggregates of carbon particles, but the peritoneal surface was macroscopically shiny and smooth. No foreign-body giant cells were found in the knee joints and there was no evidence of acute inflammatory change. Joint cartilage remained completely unaltered after 6, 12, and 24 weeks. There was no evidence of foreign-body reaction in any of the parenchymal organs. The general conclusion was that tissue tolerance of LTI dust is excellent [902].

Ultra-low-temperature isotropic (ULTI) carbon is a closely related material [923]. By the late 1970s, it became possible to deposit isotropic carbon coatings at nearly room temperature using a hybrid low-pressure vacuum process that does not require the object to be coated to be suspended in a fluidized bed. The steady state and time varying kinetics of protein adsorption of ULTI have been investigated. Flow exposures over ULTI-coated microporous membrane produced a uniform protein coating averaging 1.3 microns in thickness. Adsorption of human fibrinogen onto the ULTI was 53.5 mg/m2 (~94,900 molecules/micron2) and 14.4 mg/m2 (~127,000 molecules/micron2) of albumin, after a 1-hour exposure [810]. Albumin adsorption reaches equilibrium within 15 minutes, while fibrinogen levels are still increasing after 60 minutes, at which time the noncompetitive albumin/fibrinogen adsorption ratio reaches 0.27 [810], comparable to the 0.24 ratio achieved for LTIC [925]. In related experiments [924], the response of ULTI carbon surfaced materials to ex vivo blood flow were evaluated over perfusion periods of 0.5-8 hours. At flow rates with low Reynolds numbers (Section, the carbon attracted fewer and less distorted cellular elements than uncoated microporous membranes and microchannels [924].

Failure strength of ULTI carbon is ~7.5 x 107 N/m2 (impact fracture energy 1.1 x 107 J/m3), compared to 5.5 x 107 N/m2 (6.6 x 106 J/m3) for LTI carbon with Si, 4.5 x 107 N/m2 (3.4 x 106 J/m3) for pure LTIC, and 1.4 x 107 N/m2 (7 x 105 J/m3) for glassy carbon [955]. All these carbons have similar stiffness, with modulus of elasticity of 2.1-2.6 x 1010 N/m2 [955], in a range comparable to bone.


Last updated on 30 April 2004