The platelet reactivity of vascular graft prostheses: an in vitro model to test the effect of preclotting

Protein Adsorption and Cellular Adhesion and Activation on Biomedical Polymers

The design and development of new biomedical polymers for clinical application in devices, prostheses, and artificial organs requires a basic and fundamental understanding of biological interactions with biomedical polymers. Efforts in our laboratory have been directed towards appreciating the humoral and cellular interactions which govern protein adsorption and cellular adhesion and activation on biomedical polymers. Information and data are presented on protein adsorption from whole human blood, complement activation and receptors, and monocyte/macrophage adhesion and activation with growth factor release. Supported by experimental evidence, concepts regarding protein/polymer, cell/ polymer, cell/protein/polymer, and cell/cell interactions as they are related to in vivo events are presented.

Interaction of a blood coagulation factor on electrically polarized hydroxyapatite surfaces

Although the polarization treatment of hydroxyapatite (HA) remarkably enhances the osteoconductivity, the mechanisms have not yet been completely understood. The interaction of proteins in blood and tissue fluids with biomaterials are reportedly triggers for later cellular responses and played a major role in osteoconductive processes. Considering this, we disclosed the interaction of polarized HA surface with a coagulation factor, fibrin stabilizing factor XIII (FXIII). The HA activated FXIII even in Ca2+ free buffer, based on the SDS-PAGE detections of alpha-polymer and gamma-dimer bands assigned to stabilized fibrin. The Ca2+ ions, possibly released from the HA surfaces, were examined whether they initiate the activation of the FXIII. It was experimentally proved by ICP analysis that the induced large negative charges on the electrically polarized HA significantly increased the released Ca2+ concentration for the short pre-incubation time of 3 min. The more Ca2+ ions released from the negatively charged HA (N-HA) surfaces were more effective in the activation of the FXIII, resulting in the rapider disappearance of the gamma-chain bands in fibrin. The slightly lower Ca2+ concentration in the positively charged HA, compared to the nonpolarized HA activated the FXIII at an almost equal rate. The accelerated activation contributed to the stabilization of fibrin scaffold. Therefore, the polarity difference of the induced charges of the polarized HA surface altered the rate of the FXIII activation. The early stage interaction of the HA surfaces with blood proteins was considered to be an essential process of the accelerated new bone formation near implanted N-HA surface.

Preclinical biocompatibility assessment of the EVAHEART ventricular assist device: Coating comparison and platelet activation

Thromboembolism and bleeding remain significant complications of ventricular assist device (VAD) support. Increasing the amount of biocompatibility data collected during preclinical studies can provide additional criteria to evaluate device refinements, while design changes may be implemented before entering clinical use. Twenty bovines were implanted with the EVAHEART centrifugal VAD for durations from 30 to 196 days. Titanium alloy pumps were coated with either diamond-like carbon or 2-methoxyethyloylphosphoryl choline (MPC). Activated platelets and platelet microaggregates were quantified by flow cytometry, including two new assays to quantify bovine platelets expressing CD62P and CD63. Temporally, all assays were low preoperatively, then significantly increased following VAD implantation, before declining to a lower, but still elevated level over 2-3 weeks. MPC-coated VADs produced significantly fewer activated platelets after implant trauma effects diminished. Three animals receiving no postoperative anticoagulation had similar amounts of circulating activated platelets and platelet microaggregates as animals receiving warfarin anticoagulation. Two new methods to quantify bovine activated platelets using antibodies to CD62P and CD63 were characterized and applied. These measures, along with previously described assays, were able to differentiate between two biocompatible coatings and assess effects of anticoagulation regimen in VAD preclinical testing.

Directed type IV collagen self-assembly on hydroxylated PTFE

A method for the creation of a type IV collagen (CNIV) scaffold on polytetrafluoroethylene (PTFE) for endothelial cell attachment is described. This mimic for the basal lamina can be used in the seeding and retention of endothelial cells for blood contacting devices. The CNIV-PTFE production technique can be defined as three processes: (i) creation of a reactive superacidic/ionic PTFE surface with retained hydrophobic characteristics; (ii) activation of this surface via covalent attachment of N-(3-dimethylaminopropyl)-N'-ethylcarbodiimide (EDC); and (iii) conjugation of the EDC with human CNIV resulting in the covalent binding of protein to the PTFE surface. This article demonstrates exciting new results showing that a reaction of CNIV to this particular surface results in a unique matrix assembly of CNIV scaffolds similar to those found in the basal lamina. This assembly is concentration-dependant, occurring in a narrow window around 0.435 microM. Following the fabrication of the CNIV matrix assembly, porcine aortic endothelial cells (PAEC) were seeded onto this material. Results described in this article demonstrate that the PAEC subsequently aligned with the direction of shear, filled voids created by dead or detached cells, and divided during the 6 h of experimentation. Under static conditions, cells remained viable for 1 week of testing. This was not observed with PAEC attached to glass with adsorbed Vitrogen. In summary, this article describes a novel biotechnological breakthrough that enables the creation of stable endothelial cell monolayers useful for fabricating blood contacting devices.

Biological performances of collagen-based scaffolds for vascular tissue engineering

Collagen is widely used for biomedical applications and it could represent a valid alternative scaffold material for vascular tissue engineering. In this work, reconstituted collagen films were prepared from neutralized acid-soluble solutions for subsequent haemocompatibility and cell viability performance assays. First, haemoglobin-free, thrombelastography and platelet adhesion tests were performed in order to investigate the blood contact performance. Secondly, specimens were seeded with endothelial cells and smooth muscle cells, and cell viability tests were carried out by MTT and SEM. Results show that neutralized acid-soluble type I collagen films do not enhance blood coagulation, do not alter normal viscoelastic properties of blood and slightly activate platelet adhesion and aggregation. Cell culture shows that the samples are adequate substrates to support the adhesion and proliferation of endothelial and smooth muscle cells.

Effect of amniotic fluid on peri-implant capsular formation

Although commonly used biomaterials are physically and chemically stable, nonimmunogenic, and nontoxic, implanted and blood-contact biomaterials trigger a wide variety of unwanted responses, including inflammation, thrombosis, infection, and fibrosis. Peri-implant fibrosis is the response most commonly seen by plastic surgeons. In this study, the authors hypothesized that as hyaluronic acid (HA) reduces scar formation by inhibiting the activity of mononuclear phagocytes and lymphocytes, human amniotic fluid (HAF), which contains high concentrations of HA, HA-stimulating activator (HASA), and other factors, might prevent the formation of fibrous capsules and capsule contracture when applied intraluminally. Two 1 x 1-cm silicone blocks were placed dorsally into separate surgically created pockets underneath the panniculus carnosus muscle, distant from the incisions, in each of the 10 rats in the study. At the time of implant insertion, 2 ml of HAF was instilled into the cranially located pockets in group 1, whereas 2 ml of saline solution was instilled into the caudally located pockets in group 2. After 6 months, intracapsular static and dynamic pressure measurements were made, and then all the peri-implant capsules were excised and fixed in 10% neutral buffered formaldehyde. The thicknesses of the capsules were measured in three different areas of each section, and a mean was calculated. Capsular firmness, according to the static and dynamic pressure readings, was significantly greater in the control group, which had saline solution introduced into the pocket, than in the treatment group, which had HAF used in the same manner. The mean total thickness of the capsules surrounding the implants was 876.7 microm in the control group, as comparied with 466.8 microm in the HAF-treated group. This difference was statistically significant (p < 0.001). Because of its ability to reduce capsular thickness and firmness and also because it can be stored in a freezer if it is prepared in a cell-free manner, HAF would appear to be a useful adjunct in the prevention of capsular contracture formation.

Molecular basis of biomaterial-mediated foreign body reactions

Despite being inert and nontoxic, implanted biomaterials often trigger adverse foreign body reactions such as inflammation, fibrosis, infection, and thrombosis. With regard to the inflammatory responses to biomaterial implants, it was previously found that a crucial precedent event was the spontaneous adsorption and denaturation of fibrinogen on implant surfaces. It was further found that interactions between the phagocyte integrin Mac-1 (CD11b/CD18) and one short sequence within the fibrinogen D domain (gamma 190-202; P1) at least partially explained phagocyte accumulation on implant surfaces. However, the reason that adsorbed fibrinogen is proinflammatory--while soluble fibrinogen clearly is not--remained obscure. In this study, therefore, the question of how fibrinogen is converted to a proinflammatory state when adsorbed to biomaterial surfaces is investigated. In soluble fibrinogen, the 13 amino acid P1 sequence was found to be hidden. However, the adsorption and denaturation of fibrinogen on the surfaces of commonly used biomaterials lead to the exposure of P1 and a second neo-epitope, gamma 377-395 (P2), which also interacts with Mac-1 and is similarly occult in the soluble protein. The extent of biomaterial-mediated P1 and P2 exposure appears directly related to the severity of inflammatory responses to a test panel of biomaterials. Finally, thrombin-mediated conversion of fibrinogen to fibrin also exposes both P1 and P2 epitopes. These observations may help explain both the inflammation caused by many types of implanted biomaterials and that which occurs naturally following thrombotic events. (Blood. 2001;98:1231-1238)

Impregnated polyester arterial prostheses: performance and prospects

Impregnated polyester arterial prostheses have gained wide acceptance by most vascular surgery teams, probably because these prostheses are easy to use, without any preclotting. We offer here a synthesis of the main studies that have appraised the experimental and clinical performance of these prostheses, and we delineate their major prospects.

Blood and tissue compatibility of modified polyester: thrombosis, inflammation, and healing

Poly(ethylene terephthalate) (PET) has been reported in literature to be moderately inflammatory and thrombogenic. To moderate the inflammatory response, PET fabric was surface modified by either Fluoropassiv fluoropolymer (FC), or an RGD-containing peptide (RGD). Samples were subsequently autoclave sterilized and implanted subcutaneously in Sprague Dawley rats for 2 to 4 weeks. Retrieved samples were evaluated histopathologically for indications of material toxicity and healing. Minimal acute or chronic inflammation was associated with the fabrics after 2 and 4 week implant duration. However, fibroblast proliferation into FC modified fabric (PET/FC) was less than that into unmodified (PET) and RGD modified fabric (PET/RGD) after 4 weeks, suggesting that FC modification of PET may inhibit excessive tissue growth. Additional samples of modified and unmodified fabrics were placed in stainless steel mesh cages, which were then implanted subcutaneously for 4 weeks. Cellular exudate was extracted weekly and cell concentrations within the exudate measured. Total leukocyte count (TLC) (reflective of local inflammation) at 1 week for PET/RGD was greater than that for PET/FC and PET. TLCs after 4 week implant decreased for all sample groups. In a separate experiment, PET vascular grafts surface modified by either FC or RGD were contacted 1 h with blood using the baboon arteriovenous (AV) shunt model of thrombosis in both the presence and absence of heparin. Accumulation of 111In labeled platelets (reflective of thrombus accumulation) upon grafts was less in the presence of heparin (effect significant at p = 1.2 x 10(-6), two-way ANOVA). Accumulation (in the presence of heparin) upon PET/RGD was less (p = 0.19), and upon PET/FC significantly less (p = 0.016) than that upon the unmodified PET control, suggesting that FC modification of PET may inhibit thrombus accumulation.

Thrombus formation on polytetrafluoroethylene surfaces: the importance of von Willebrand factor

The importance of von Willebrand factor (vWf) in the formation of platelet-fibrin thrombi on expanded polytetrafluoroethylene (ePTFE) surfaces was studied in an in vitro system, perfusing non-anticoagulated human blood over ePTFE grafts for 3 min at varying shear rates (100, 500 and 1500/s shear). Platelet (111In) and fibrin (125I) deposition was assessed on ePTFE surfaces in the presence and relative absence of vWf, achieved by use of polyclonal anti-vWf antibody (anti-vWf Ab). A total of 29 perfusions were performed. Increasing shear rate was associated with greater platelet deposition in the presence of vWf (p < 0.001). This shear-dependent rise in platelet deposition was not observed when vWf was blocked by anti-vWf Ab (P < 0.1), confirming the role of vWf in platelet deposition at high shear rates. Fibrin deposition increased with increasing shear rate in the presence of vWf (P < 0.01). Inhibiting vWf abolished the shear-dependent increase in fibrin deposition. These data suggest that vWf plays a critical role in platelet and fibrin thrombus formation on ePTFE surfaces. These effects are particularly important under conditions of high shear rate. These mechanisms may lead to the observed pathologic thrombus formation and platelet-dependent neointimal processes occurring at areas of high shear rate within the anastomotic regions of ePTFE grafts.

Metabolic response of granulocytes and platelets to synthetic vascular grafts: preliminary results with an in vitro technique

The metabolism of granulocytes as well as platelets evoked by incubation with different synthetic vascular grafts was monitored during 6-h batch experiments using microcalorimetry. Standard knitted Dacron grafts, ePTFE-grafts, knitted Dacron grafts with collagen impregnation, and knitted Dacron grafts with external collagen-coating were used. The heat production per cell was calculated. A rapid increase of metabolic activity followed by a gradual decrease was demonstrated with both granulocyte suspension and platelet concentrate. Significant differences were obtained between the materials with a maximum response of Dacron grafts with collagen impregnation for both granulocyte and platelet response. The materials had different surface morphologies regarding cell adhesion after incubation as demonstrated with scanning electron microscopy with more pronounced adhesion on the collagen-impregnated grafts. The results suggest that microcalorimetry may be useful for the evaluation of cellular reactions on different biomaterials. However, further studies have to reveal the specificity of the reactions.

Ex vivo and in vitro platelet adhesion on RFGD deposited polymers

Clinical applications of small-diameter synthetic vascular grafts are hindered by their highly thrombogenic surfaces. To develop vascular grafts that resist thrombotic occlusion, a radio frequency glow discharge (RFGD) process was employed to modify the surface of existing graft materials. Ultrathin coatings of RFGD polymers of ethylene (E), tetrafluoroethylene (TFE), and hexamethyldisiloxane (HMDS) were deposited on the lumen of Dacron grafts. Surfaces were characterized by electron spectroscopy for chemical analysis (ESCA). The effect of glow discharge treatments on platelet-graft interactions was evaluated in an ex vivo baboon shunt model. Following placement of an untreated or RFGD-treated graft in the shunt, deposition of 111Indium-labeled platelets was monitored for 60 min by gamma camera imaging. Untreated Dacron rapidly accumulated large numbers of platelets, reaching a plateau in 60 min. HMDS- and TFE-treated Dacron had significantly lower levels of platelet deposition compared to the untreated control. In contrast, the ethylene treatment of Dacron augmented platelet deposition, making it the most platelet-adherent surface studied. In vitro studies were also performed using untreated and RFGD-treated poly (ethylene terephthalate) (PET) coverslips. ESCA verified that the surface composition of the untreated and RFGD-treated coverslips were virtually identical to their untreated and treated Dacron graft counterparts. Samples were incubated in washed baboon platelet suspensions for 2 h at 37 degrees C. Platelet adhesion on the untreated PET was relatively high, and many of the platelets had a completely spread morphology. The HMDS and TFE treatment of PET reduced the number of adherent platelets and prevented platelet spreading on the surface. Platelet adhesion and spreading on the ethylene-treated surface was the highest among the four studied. There is a remarkable linear correlation of the ex vivo and in vitro platelet adhesion data.

Acute thrombogenicity of collagen coating of dacron grafts: an experimental study in sheep

The early stages of platelet accumulation in two types of sealed interposition Dacron grafts implanted in the carotid arteries of sheep have been studied. One type was externally coated with collagen (Haemaguard) while the other experimental conduit had an additional inner lining of the same substance. 32P-labelled platelets were used to assess platelet accumulation and corrections for wall absorption were calculated. The activities in both types of grafts were highest proximally and decreased towards the distal anastomosis. The increases in the doubly-sealed grafts were larger than in those that had been single sealed, presumably reflecting activation of platelets in contact with collagen at the graft-wall and bloodstream interface. In addition, a significantly larger amount of thrombus was formed in the doubly-sealed grafts 4 h after reperfusion. It is reasonable to assume that increased acute thrombogenicity due to direct collagen-blood contact on graft surfaces is unfavourable to long term patency.

Effect of albumin coating on the in vitro blood compatibility of Dacron arterial prostheses

A recirculating in vitro perfusion system was used to assess the effect of albumin precoating on the thrombogenicity of Dacron vascular grafts. A complete analysis of platelet activation was carried out, involving platelet count, release, adhesion and aggregation. Fibrin formation was assessed by measuring fibrinogen levels and fibrinopeptide A production; leucocyte interaction was analysed by measuring total leucocyte count as well as an analysis of cell adhesion to the surface by scanning electron microscopy. The platelet count decreased progressively with perfusion time for Dacron until by 30 min, it had declined to 69% +/- 2% of baseline. The platelet count did not, however, change significantly from baseline when albumin-coated Dacron was tested. Release of platelet factor 4 and beta-thromboglobulin at 180 min for Dacron was 37.8 +/- 29.8 times and 66.9 +/- 18.2 times baseline, respectively, while albumin coating caused significantly less (P less than 0.03) platelet release. Albumin coating diminished coagulation activation and fibrinopeptide A formation. The total leucocyte concentration decreased significantly for Dacron by 180 min, while that for albumin-coated Dacron did not change significantly from baseline levels. Albumin coating produced a film-like covering over the Dacron. For Dacron, there were numerous leucocytes and platelets adherent to the surface, whilst cellular deposition was minimal upon the albumin-coated surface. Thus, albumin coating improved the short-term blood compatibility of Dacron by all of the methods employed in this study.

A clinical survey of aortobifemoral bypass using two inherently different graft types

The performance of knitted Dacron and polytetrafluoroethylene (PTFE) bifurcated grafts are compared in this study of 312 patients at a single institution. Patients of the two graft groups were statistically well-matched in risk factors and degree of distal obstructive disease. Operating time needed to implant either graft was approximately equal. For patients with abdominal aortic aneurysms, mean volume of blood transfused was 2.2 units for Dacron grafts and 0.2 units for PTFE grafts; for patients with aortoiliac occlusive disease, the comparable figures were 1.1 units and 0.1 units, respectively. Four-year cumulative patency for Dacron (90%) and PTFE (97%) grafts were not significantly different (p greater than 0.01). Complications affected 13% of the patients of the Dacron group and 4% of the PTFE group. All six graft infections and all seven graft double-limb thromboses occurred in Dacron grafts. Anastomotic aneurysms, amputations, and late graft revisions occurred with greater frequency in patients with Dacron grafts.