Prevention of unwanted thrombus formation and embolization remain major design challenges in the development of mechanical circulatory support systems. Two approaches have dominated device development: supracellular textures may be applied to induce passivation via the accumulation of a neointima that is strongly attached to the surface,1–3 or very smooth material designed to resist formed element and fibrin deposition may be used.4 Although a featureless surface finish, a bulk material having advantageous surface chemistry, and modification of surface chemistry5,6 have all been found to be useful, deposition-resistant circulatory support devices rely significantly upon maintenance of sufficient wall shear stresses to prevent thrombus formation. Although clinically useful adult devices have been developed, wall shear requirements remain a difficult design constraint, particularly as devices are scaled down in size.7–9
A key event in the development of a thrombus is the adhesion of platelets, which is mediated by binding of integrins on the platelet surface to proteins adsorbed to the material surface, specifically fibrinogen,10 fibronectin,11 vitronectin,12 and von Willebrand factor.13 Fibrinogen has been shown to play the major role in platelet adhesion and aggregation on artificial surfaces,14,15 particularly at low shear stresses.16–18 Binding is accompanied by activation of the platelet, which increases integrin adhesion strength and triggers spreading of the activated platelet on the surface.19
We propose an approach to the use of particular ordered surface topographies to discourage adhesion of platelets by limiting the implant surface area that is available to a platelet for binding and spreading. Specifically, we demonstrate a method whereby a biomedical polyether(urethane urea) film may be fabricated with arrays of pillars having geometries smaller than individual platelets. Pillar diameter, height, and spacing are chosen so as to provide markedly reduced area for interaction between platelet integrins and the protein-coated biomaterial surface. The adhesion of platelets was assessed on the textured materials under a physiologically relevant shear stress range and compared with the adhesion on smooth material. Platelet adhesion at low shear was reduced on textured samples and a lower incidence of fibrin clots was also observed. The fabrication process was found to affect material texture and not surface chemistry. It is anticipated that surface texture may be used as an additional design parameter, in addition to surface chemistry and fluid flow patterns, for minimizing thrombosis in the pediatric circulatory support devices that are being developed.
Materials and Methods
Polyether(Urethane Urea) Submicron Topographies
Polyether(urethane urea) (PUU) sheets were fabricated with surface topographies consisting of ordered arrays of submicron pillars using a soft lithographic two-stage replication molding technique.20 Briefly, a master pattern is created on a silicon wafer using standard microfabrication methods, a poly(dimethylsiloxane) (PDMS) mold is cast, and PUU is cast in to the PDMS to replicate the master pattern. Master patterns were prepared on 150-mm-diameter silicon wafers by coating with hexamethyldisilazane adhesion promoter and a film of UV5 photoresist (Shipley, Rohm and Haas, Philadelphia, PA) of 650–700 nm thickness. The resist film was selectively exposed using a 248-nm-wavelength KrF laser on a 5:1 Reduction Stepper (Nikon NSR Series) using one of two reticule patterns. The first reticule generated resist pillars with 700 nm width and 700 nm spacing, and the second generated 400 nm widths and 400 nm spacing. Pillar height was given by the resist thickness. Unexposed resist was used for smooth control samples. Molds of these master patterns were prepared by casting PDMS elastomer (Sylgard 184, Dow Corning (Robert McKeown, Branchburg, NJ), 10:1 base:curing agent) over the masters to a thickness of 3–4 mm. The PDMS was degassed to 29-inch Hg 10 times in a vacuum desiccator before and after casting to eliminate bubbles. Samples were cured for 4 hours at 65° C under a 29-inch Hg vacuum, to ensure PDMS conformation to the patterns. The PDMS was gently peeled from the master, washed with acetone to remove transferred UV5 and dried at 65°C under a 29-inch Hg vacuum.
The PUU replicas were created in Biospan MS.4 segmented PUU (Polymer Technology Group, Berkeley, CA), a PUU previously used in a left ventricular assist device (LVAD).21 It has a methylenediisocyanate hard segment, a polytetramethlyene oxide soft segment, and an ethylene diamine chain extender, and is chain endcapped with 2,000-molecular-weight PDMS at 0.4% by weight. The polymer is supplied at 22.5% weight/volume in dimethyl acetamide, which is a solvent for the UV5 resist, hence the use of a two-stage replication process. The PDMS molds may also be used repeatedly without loss of replication efficiency,22 whereas the UV5 patterns are damaged by peeling the PDMS and may thus only be used once. PUU was spin cast onto smooth and textured PDMS molds at 500 rpm for 1 minute, degassed, cured for 24 hours at room temperature and then 24 hours at 65°C, both under 29-inch Hg vacuum, which was broken periodically to eliminate bubbles. PUU was gently separated from the PDMS after immersion in dH2O for 1 hour.
Replication efficiency was assessed by cutting 15-mm-diameter samples from the PUU replica with a steel cutter, sputter-coating with a 10-nm gold film, and analyzing via scanning electron microscopy (SEM; Philips XL-20 SEM). Topography dimensions were assessed via atomic force microscopy (AFM; Digital Instruments Multimode, Santa Barbara, CA) using an electron beam deposition tip (MikroMasch, Wilsonville, OR) to minimize tip enlargement effects. Surface chemistry of the PUU replicas was assessed via x-ray photoelectron spectroscopy (XPS; a Kratos Analytical Axis Ultra). Both the front (PDMS-contacting) and rear (air-contacting) sides of PUU replicas were analyzed at 90° takeoff angle and compared to the front and rear of PUU cast against glass. The surface elemental composition was quantified by applying the appropriate relative sensitivity factors for the Kratos XPS to the integrated peak areas, and the approximate sampling depth was 80 Å.
Assessment of Platelet Adhesion
The variation in platelet adhesion with fluid shear stress was assessed using a rotating disk system (RDS; Pine Instruments, PA) as illustrated in Figure 1. The behavior of such systems has been well characterized.23–25 Briefly, under laminar flow conditions, in an “infinite” steady-state system, the wall shear stress varies linearly with radial distance from zero in the center of the PUU sample, and the flux of platelets is uniform across the PUU. Therefore, by assessing the variation in numbers of adhered platelets with radial distance, the variation in platelet adhesion coefficient (AC%) with wall shear stress may be calculated using AC% = 100 × N/(tj), where N is the number of adhered platelets observed in an image, t is the experiment duration, and j is the mass flux of platelets to the imaged PUU surface. PUU samples with 20 mm diameter were cut with steel dies and mounted with industrial strength mounting tape on a stainless steel disk that was scribed with concentric circles at 1-mm radial increments and radii at 60° increments. This was attached to the RDS shaft via a stainless steel threaded piece. The PUU was immersed overnight in ddH2O to remove low-molecular-weight oligomers and equilibrated in 25°C phosphate buffered saline (PBS) for 1 hour.
Bovine platelet-rich plasma (PRP) was prepared from whole blood collected from healthy calves in accordance with institutional policies. Blood was drawn from the jugular vein into a PVC blood transfer bag containing sodium heparin anticoagulant at a final concentration of 3 U/ml. A fraction of blood was sedimented by centrifugation at 600g for 20 minutes and the PRP above the buffy coat was gently removed. An aliquot of the PRP was taken and the platelet concentration was determined by electrical impedance on a Sysmex 9600 (Sysmex, Kobe, Japan). The remaining blood fraction was sedimented by centrifugation at 1,500g for 20 minutes and the platelet-poor plasma above the buffy coat was removed. Platelet-rich and platelet-poor plasma were combined to yield two 50-ml PRP volumes (one PRP sample and one PRP control) with platelet concentrations of 2.5 × 108 platelets/ml in 100-ml PTFE beakers (54 mm internal diameter). Samples were maintained at 25°C throughout the experiment. The PUU was removed from the PBS, lowered into the PRP sample, and rotated for 2 hours at 238 rpm. This rotation speed yielded a Reynold’s number (Re) of ∼2,200 and a boundary layer thickness of ∼780 μm. For Re values < 105, laminar flow is maintained in the boundary layer.26 The boundary layer thickness is much smaller than the 10-mm PUU sample and much larger than the fabricated surface topographies and bovine platelets, justifying the assumption of an infinite disk and ensuring an undisturbed laminar flow field. The thickness is also much larger than the size of the surface topographies and bovine platelets, ensuring that the laminar flow field is not disturbed. The rotator reaches operational speed within 4 milliseconds, justifying the assumption of steady-state conditions for a 2-hour experiment. Under these conditions, the shear rate varied from 0 to 10 dyn/cm2 at 9 mm radial distance, encompassing the shear stress range where platelet adhesion is predominantly mediated by surface-adsorbed fibrinogen16,27 and corresponding to the low shear stresses found in a pediatric ventricular assist device design.7
Following rotation, the PRP was exchanged with 245 ml PBS and 210 ml 1% paraformaldehyde (PFA) by sequential addition and aspiration of 35 ml volumes, and adherent platelets were fixed for 1 hour. The PFA was removed by exchange with 210 ml PBS. The PUU/steel stub assembly was removed and adherent platelets were labeled overnight at 4°C with primary antibodies: mouse anti–bovine αIIbβ3 (VMRD, Pullman, WA) and polyclonal rabbit anti–human CD-62P that demonstrates cross-reactivity with bovine P-selectin (Becton Dickinson, Franklin Lakes, NJ). The primary antibodies were prepared at concentrations of 1.5 μg/ml and 1 μg/ml, respectively, in 6% normal donkey serum (DS) to block nonspecific reactions. Fibrinogen was labeled by the addition of goat anti–bovine fibrinogen (American Diagnostica, Stamford, CA) at a final concentration of 1 μg/ml. After incubation, the PUU was washed with PBS. Both primary antiplatelet antibodies were labeled with R-phycoerythrin (R-PE) conjugated secondary antibodies (donkey antimouse and donkey antirabbit; Becton Dickinson) for 1 hour at room temperature at a 1:100 dilution in 6% DS. The antifibrinogen antibody was labeled with a FITC-conjugated donkey anti-goat secondary (Becton Dickinson) also at a 1:100 dilution in 6% DS. The PUU was washed with PBS and mounted with antifade gel (Biomeda, Foster City, CA) and coverslip for analysis via epifluorescent microscope (Nikon Optiphot 3). Images of 98 × 73-μm PUU areas were acquired at the intersection of each scribed circle with each of the six scribed radii on the steel mounting disk using a 100× oil-immersion objective, and recorded via digital camera (Diagnostic Instruments Spot, Sterling Heights, MI). The number of adherent platelets was determined and the AC% was calculated. Six images were recorded in the center of the PUU in a two-by-three matrix corresponding to a 196 × 219-μm area. Thus six independent AC% measurements were recorded at each radial distance per PUU sample. Platelets were also observed via SEM by taking PUU after PFA fixation and dehydrating with serial dilutions of ethanol from 50% to 100% in 10% increments for 10 minutes each.
Assessment of Nonadherent Platelet Activation
The activation of the bulk PRP by the rotating PUU sample was assessed by comparing the activity of the PRP sample, exposed to the PUU and RDS, and the activity of the PRP control, subject to the same conditions as the sample but without PUU and RDS. Activation was determined at t = 0 (immediately before rotation) and t = 2 hours (immediately after rotation). A 0.5-ml aliquot was taken from both sample and control before and after rotation and fixed by addition of 0.5 ml 1% PFA. An additional 0.45-ml aliquot of PRP control was taken at the start of the experiment and a 50-μl aliquot of 1.9-mg/ml soluble calf skin collagen (Bio/Data, Horsham, PA) was added as positive activation control. After 1 hour, 0.5 ml of 1% PFA was added to fix the platelets. A 25-μl aliquot of fixed PRP was immunofluorescently labeled for flow cytometric analysis (FACS). Platelets were labeled by addition of a 25-μl aliquot of mouse anti–bovine αIIbβ3 (1.5 μg/ml in 6% DS). Activated platelets were labeled by addition of a 25-μl aliquot of rabbit anti–human CD-62P (1 μg/ml in 6% DS) for 1 hour at room temperature to assess P-selectin, which is only expressed on platelets following activation and secretion.17 Secondary labeling was performed by addition of a 50-μl aliquot of 1:100 dilution FITC-conjugated donkey anti-mouse (Becton Dickinson) and 50-μl aliquot of 1:100 dilution R-PE conjugated donkey anti-rabbit (Becton Dickinson), both in 6% DS, incubating for 1 hr at room temperature. Samples were diluted with PBS and a minimum of 104 platelet events was analyzed via Becton Dickinson FACScan, gating on forward/side scatter and FITC fluorescence to ensure that only platelets were interrogated. R-PE fluorescence expression was gated such that mean activity of the sample and control at T = 0 had a nominal value of 2%. Percentage activation for the negative control, PRP sample, and positive control was assessed as the percentage of platelet events with R-PE fluorescence intensity greater than this gate value.
Statistical comparisons were performed using non-parametric analysis of variance (Kruskal-Wallis test) with InStat software (version 3, GraphPad Software, San Diego, CA). Means of experimental data were compared and differences were considered statistically significant at p < 0.05. Significance is denoted in the tables and figures of this report. Significance in platelet adhesion experiments is denoted by an asterisk (*) comparing smooth 700-nm PUU and number sign (#) for smooth 400-nm PUU. In FACS experiments, a dagger (†) denotes significance to collagen-treated positive control. In each case, one symbol denotes p < 0.05, two symbols denote p < 0.01, and three symbols denote p < 0.001.
The two-stage replication molding process was found to readily fabricate the 700-nm and 400-nm resist pattern into large areas of MS.4 PUU. The replication efficiency was determined by SEM assessment of multiple PUU samples and was found to be > 99.7% for 700-nm master patterns (Figure 2a) and > 99.5% for 400-nm masters (Figure 2b). The geometry of the PUU pillars was determined via AFM. Pillars replicated from 700-nm masters had typical height, diameter, and spacing of 650 nm, 700–800 nm, and 600–700 nm, respectively (Figure 2c). Pillars replicated from 400-nm masters had typical height, diameter, and spacing of 600 nm, 400–500 nm, and 300–400 nm, respectively (Figure 2d). It is believed that these pillar widths are overestimated due to the AFM tip enlargement effects, because SEM imaging revealed that the pillar width and spacing are consistent with the 1:1 ratio of the masters. The replication process did not seem to vary across the sample surface, indicating the potential for this process to fabricate ordered surface textures into large areas of biomedical polymers; in this case, with 150 mm diameter. For comparison, a 150-mm silicon wafer is shown alongside a 70-cc LVAD sac and a 15-cc developmental pediatric LVAD sac (Figure 2e). The surface elemental composition of PUU from a replication-molded film and of PUU cast against glass was assessed via XPS (Table 1). The compositions of PUU cast against PDMS, glass, and air were all comparable, suggesting that the replication molding process is only influencing surface topography and not surface chemistry. The silicon content in all samples is higher than that generally expected for PUU, but it is consistent. It is thought that this is caused by accumulation of the 2,000-molecular-weight PDMS used to chain endcap the PUU, rather than any transfer of PDMS from the mold.
The mean variation in platelet AC% with shear stress across a 0–10 dyn/cm2 range is presented in Figure 3 with n = 6 for each PUU surface texture and six arcs assessed per sample. A general trend is found for the AC% to be high at 0 dyn/cm2 and to decrease with increasing shear. Platelet adhesion tended to be constant for shear greater than ∼5 dyn/cm2 and comparable between smooth and textured PUU, with the exception of seven areas on two 700-nm samples where larger numbers of individual and aggregated platelets were observed, reflected in the increased variability and standard error of these data. Experiments performed assessing a 0–67 dyn/cm2 shear stress range found that AC% remained constant and comparable between smooth and textured PUU for shear > 10 dyn/cm2 (data not shown), and thus the only variations occur for shear < 5 dyn/cm2. In this shear stress range, platelet adhesion to 700-nm PUU was reduced compared with smooth PUU, significantly so for shear ≤ 3.3 dyn/cm2. Adhesion to 400-nm PUU was also reduced in this shear range but with less significance. Initial experiments assessing the 0–67 dyn/cm2 shear stress range were performed in the presence of a lateral wobble of ∼100 μm in the threaded shaft. Under these conditions, platelet adhesion was found to be reduced on 700-nm PUU compared with smooth PUU across the entire shear range (data not shown). This wobble would likely generate a small, nonzero shear stress at the center of the PUU sample, and may lead to disturbance of the boundary layer at the sample surface and loss of laminar flow.
Epifluorescent microscopy revealed that four of the six smooth PUU samples demonstrated a high incidence of platelets within cloudy, fluorescent structures in the central regions of the PUU, with these structures often extending through several focal planes. Only two 400-nm and zero 700-nm PUU samples demonstrated similar features. Immunofluorescent labeling for fibrinogen revealed that these cloudy fluorescent structures are most likely fibrin clots (Figure 4a). On 700-nm and 400-nm PUU, and at higher shear on smooth samples, platelets generally were present as individual events with no fibrin clot observed (Figures 4b and 4c). Platelets were usually discoid or occasionally pseudopodal, which is consistent with previous observation that bovine platelets do not exhibit the fully spread morphology found in activated, adherent human platelets because of the lack of open canalicular system.28 Qualitative observation of platelet adhesion via SEM revealed that platelets on 700-nm PUU were generally attached between the pillars, and that the pillars were not filled in by an adsorbed protein layer.
The activation of platelets in the bulk PRP was assessed by FACS in PRP sample, PRP negative control, and collagen-treated positive control. The percentage of cells expressing R-PE = CD-62P, compared with a nominal 2% value at t = 0, are presented in Table 2. All samples exhibited similar percent activation with the exception of the positive control, which was significantly higher than all other samples. The differences between the other samples were not statistically significant. These data indicate no significant activation of platelets in bulk PRP by the PUU during the course of the experiments. Collagen-treated activity is a little lower than expected, probably due to platelet entrapment within thrombi observed after collagen activation.
Two main approaches currently exist for design of nonembolizing blood-contacting surfaces of ventricular assist devices, namely, the application of supracellular textures to promote biomaterial passivation via accumulation of a neointima1–3 or the use of smooth surfaces to minimize accumulation of thrombus components.4 The later approach requires, in addition to proper material selection, surface finish and sometimes treatment, maintenance of sufficient wall shear stress to prevent formed element adhesion. This has been found to be particularly challenging in devices that are scaled to smaller sizes, where wall shear stresses are reduced compared to adult devices.7 The critical shear value (e.g., over 500 s–129) is comparable to the shear range over which platelet adhesion is mediated by integrin binding to adsorbed fibrinogen.16,27
The results presented in this report demonstrate a method whereby large areas of PUU may be readily textured with subplatelet-sized topographies and that these textures lead to a reduction in platelet adhesion in this low shear stress range, reflecting increased washing efficiency. We theorize that this reduction is caused by a reduction in the area with which a platelet may contact the PUU surface, by restricting contact to the tops of the pillars. This would be expected to result in a decreased adhesive force between a platelet and the surface by reducing both the probability of a platelet interacting with an adsorbed fibrinogen molecule and also the total number of integrin-fibrinogen bonds formed between a platelet and the surface, facilitating washing by fluid flow.
Platelet adhesion was reduced on both textured PUU surfaces compared to smooth controls for shear stress below 5 dyn/cm2. The 700-nm pillars led to the largest decrease in platelet adhesion with a statistically significant reduction for shear below 3.3 dyn/cm2. The 400-nm pillars exhibited comparable reductions, but significance was only found at 3.3 dyn/cm2 and ∼0 dyn/cm2. It should be noted that a zero value of shear stress would only occur at the central point of the samples where radial distance is equal to zero. The epifluorescent microscope images obtained in the center of the samples extend to a radial distance of 147 μm and are therefore assessing a shear stress range of 0–0.16 dyn/cm2. This small shear stress seems to be sufficient to wash the PUU surface and reduce platelet adhesion in the presence of surface topographies. Initial experiments performed in the presence of ∼100-μm lateral wobble found a reduction in platelet adhesion to 700-nm PUU across a 0–67 dyn/cm2 shear range, possibly indicative of increased surface texture effectiveness in dynamic, nonsteady state environments such as those found within pulsatile devices. Epifluorescent microscopy of the areas of smooth PUU exposed to low shear stress revealed that large numbers of adherent platelets were found attached within fibrin clots. This was less often the case with the textured PUU samples. It not clear if the platelets attached first and acted as a substrate for the plasma coagulation cascade,30 if the fibrin deposited first and acted to entrap the platelets,31 or whether the two events were simultaneous. The PRP used here was anticoagulated with heparin to limit fibrin production. A localized increase in activated platelets, such as that found in the center of the smooth PUU samples, could potentially provide sufficient substrate to overcome the heparinization, whereas the textured samples had fewer adherent platelets and therefore less substrate for the coagulation cascade. Alternatively, the textured surfaces may limit fibrin polymer and platelet adhesion, explaining the reduced incidence of fibrin clots observed on these materials.
Scanning electron microscopy revealed that the majority of platelets adhered to 700-nm samples were found in the interpillar regions and were not confined to the tops of the pillars. These inter-pillar regions have a diameter of ∼1.3 μm (for 700-nm pillar diameter and spacing), which is comparable to the platelet diameter. Under the hypothesis that the reduction in platelet adhesion is due to the reduction in area with which platelets contact the PUU surface, it is anticipated that optimization of pillar geometry could lead to further decrease in platelet adhesion.
Assessment of the activation of platelets in bulk PRP via FACS demonstrated that, while textured PUU leads to a reduction in platelet adhesion, there was no increase in bulk activation. This demonstrates that the textures and fluid shear were acting to prevent the initial adhesion of platelets and were not simply removing adherent platelets from the sample surface, since adherent platelets would have activated via outside-in signaling events.19 If adherent, activated platelets were removed from the PUU surface, they could aggregate elsewhere with the risk of thromboembolic events. These data support the hypothesis that the surface textures are acting to reduce the probability of a platelet interacting with a fibrinogen molecule, facilitating washing by fluid flow over the surface and preventing platelet adhesion.
A two-stage replication molding technique may be used to pattern large areas of a biomedical PUU with ordered subplatelet-sized surface pillars without affecting material surface chemistry. These pillars acted to reduce platelet adhesion at low shear stress compared with smooth controls without leading to increases in the activation of nonadherent platelets in bulk solution. Adherent platelets were generally found in interpillar spaces, suggesting that pillar geometry may be optimized to further reduce platelet adhesion. The ability to reduce platelet adhesion to surfaces exposed to low shear stress has great applicability to the development of novel blood-contacting devices.
The authors gratefully acknowledge the assistance of the Nanofabrication Facility and Materials Research Institute at the Pennsylvania State University, the Musculoskeletal Research Laboratory in the Department of Orthopaedics and Rehabilitation at Penn State College of Medicine, and Nate Sheaffer of the Cell Science/Flow Cytometry Core Facility of the Section of Research Resources, Penn State College of Medicine. The authors also thank Mallory Balmer (the Penn State University), Hanako Yamanaka, Eric Yeager, Dr. Tigran Khalapyan, and the members of Artificial Organs, Penn State College of Medicine.
1.Rafii S, Oz MC, Seldomridge JA, et al: Characterization of hematopoietic cells arising on the textured surface of left ventricular assist devices. Ann Thorac Surg
60: 1627–1632, 1995.
2.Rose EA, Levin HR, Oz MC, et al: Artificial circulatory support with textured interior surfaces - a counterintuitive approach to minimizing thromboembolism. Circulation
90: 87–91, 1994.
3.Scott-Burden T, Tock CL, Bosely JP, et al: Nonthrombogenic, adhesive cellular lining for left ventricular assist devices. Circulation
98: II339–II345, 1998.
4.Harasaki H, Kiraly R, Nose Y: Blood-blood pump surface interaction, in Szycher M (ed), Biocompatible Polymers, Metals and Composites. Lancaster, PA, Technomic Publishing, 1983, pp. 199–212.
5.Lee JH, Ju YM, Lee WK, Park KD, Kim YH: Platelet adhesion onto segmented polyurethane surfaces modified by PEO- and sulfonated PEO-containing block copolymer additives. J Biomed Mater Res
40: 314–323, 1998.
6.Park JH, Park KD, Bae YH: PDMS-based polyurethanes with MPEG grafts: Synthesis, characterization and platelet adhesion study. Biomaterials
20: 943–953, 1999.
7.Bachmann C, Hugo G, Rosenberg G, et al: Fluid dynamics of a pediatric ventricular assist device. Artif Organs
24: 362–372, 2000.
8.Daily BB, Pettitt TW, Sutera SP, Pierce WS: Pierce-Donachy pediatric VAD: Progress in development. Ann Thorac Surg
61: 437–443, 1996.
9.Hochareon P, Manning KB, Fontaine AA, et al: Fluid dynamic analysis of the 50 cc Penn State artificial heart under physiological operating conditions using particle image velocimetry. J Biomech Eng
126: 585–593, 2004.
10.Nagai H, Handa M, Kawai Y, et al: Evidence that plasma fibrinogen and platelet membrane GPIIb-IIIa are involved in the adhesion of platelets to an artificial surface exposed to plasma. Thromb Res
71: 467–477, 1993.
11.Beumer S, Ijsseldijk MJW, Degroot PG, Sixma JJ: Platelet adhesion to fibronectin in flow-dependence on surface concentration and shear rate: Role of platelet membrane glycoproteins GPIIb/IIIa and Vla-5, and inhibition by heparin. Blood
84: 3724–3733, 1994.
12.Fabrizius-Homan DJ, Cooper SL, Mosher DF: The ex vivo effect of preadsorbed vitronectin on platelet activation. Thromb Haemost
68: 194–202, 1992.
13.Ruggeri ZM, Ware J: Von Willebrand factor. FASEB J
7: 308–16, 1993.
14.Tsai WB, Grunkemeier JM, McFarland CD, Horbett TA: Platelet adhesion to polystyrene-based surfaces preadsorbed with plasmas selectively depleted in fibrinogen, fibronectin, vitronectin, or von Willebrand’s factor. J Biomed Mater Res
60: 348–359, 2002.
15.Sheppeck RA, Bentz M, Dickson C, et al: Examination of the roles of glycoprotein Ib and glycoprotein IIb/IIIa in platelet deposition on an artificial surface using clinical antiplatelet agents and monoclonal antibody blockade. Blood
78: 673–680, 1991.
16.Savage B, Saldivar E, Ruggeri ZM: Initiation of platelet adhesion by arrest onto fibrinogen or translocation on von Willebrand factor. Cell
84: 289–297, 1996.
17.Stenberg PE, McEver RP, Shuman MA, et al: A platelet alpha-granule membrane protein (GMP-140) is expressed on the plasma membrane after activation. J Cell Biol
101: 880–886, 1985.
18.Zaidi TN, McIntire LV, Farrell DH, Thiagarajan P: Adhesion of platelets to surface-bound fibrinogen under flow. Blood
88: 2967–2972, 1996.
19.Shattil SJ, Kashiwagi H, Pampori N: Integrin signaling: the platelet paradigm. Blood
91: 2645–2657, 1998.
20.Milner KR, Balmer M, Donahue HJ, et al: Fabrication of ordered sub-micron topographies on large-area poly(urethane urea) by two stage replication molding, in Taylor DP, Liu J, McIlroy D, Merhari L, Pendry JB, Borenstein JT, Grodzinski P, Lee LP, Wang ZL (eds), Nanoengineered Assemblies and Advanced Micro/Nanosystems: Material Research Society Symposium Proceedings 820
, Warrendale, PA, 2004, pp. R2.8.1–R2.8.6.
21.Liu Q, Runt J, Felder G et al: In vivo and in vitro stability of modified poly(urethane urea) blood sacs. J Biomater Appl
14: 349–366, 2000.
22.Xia YN, McClelland JJ, Gupta R et al: Replica molding using polymeric materials: A practical step toward nanomanufacturing. Adv Mater
9: 147–149, 1997.
23.Wang IW, Anderson JM, Marchant RE: Staphylococcus epidermidis adhesion to hydrophobic biomedical polymer is mediated by platelets. J Infect Dis
167: 329–336, 1993.
24.Benton ER: On the flow due to a rotating disk. J Fluid Mech
24: 781–800, 1966.
25.Daily JW, Nece RE: Chamber dimension effects on induced flow and frictional resistance of enclosed rotating disks. J Basic Eng
26.Levich VG: Physicochemical Hydrodynamics. New Jersey, Prentice-Hall, 1962.
27.Ikeda Y, Handa M, Kawano K, et al: The role of von Willebrand factor and fibrinogen in platelet aggregation under varying shear stress. J Clin Invest
87: 1234–1240, 1991.
28.Grouse LH, Rao GHR, Weiss DJ, et al: Surface-activated bovine platelets do not spread, they unfold. Am J Pathol
136: 399–408, 1990.
29.Hubbell JA, McIntire LV: Visualization and analysis of mural thrombogenesis on collagen, polyurethane and nylon. Biomaterials
7: 354–363, 1986.
30.Weiss HJ, Turitto VT, Baumgartner HR: Role of shear rate and platelets in promoting fibrin formation on rabbit subendothelium. J Clin Invest
78: 1072–1082, 1986.
31.Chiu YL, Chou YL, Jen CYJ: Platelet deposition onto fibrin-coated surfaces under flow conditions. Blood Cells
13: 437–447, 1988.