Effect of Selective Factor Xa Inhibition on Arterial Thrombus Formation Triggered by Tissue Factor/Factor VIIa or Collagen in an Ex Vivo Model of Shear-Dependent Human Thrombogenesis
Abstract Tick anticoagulant peptide (TAP) is a potent and selective inhibitor of factor Xa. TAP has shown good antithrombotic efficacy in experimental animal models of disseminated intravascular coagulation and venous and arterial thrombogenesis. In the present study we evaluated the effect of recombinant TAP (rTAP) on acute thrombus formation in human nonanticoagulated blood triggered either by tissue factor (TF) or by collagen at arterial shear conditions. The main goal was to establish the role of factor Xa in thrombus formation by use of an optimal inhibitory concentration of rTAP. Blood was drawn directly from an antecubital vein by a pump over the respective thrombogenic surfaces, which were positioned in a parallel-plate perfusion chamber. rTAP was mixed homogeneously into the flowing blood by a heparin-coated device positioned proximal to the perfusion chamber. The passage of blood through this device caused minor activation of coagulation but little activation of platelets. Fibrinopeptide A and β-thromboglobulin levels after 5 minutes of blood perfusion were, on average, 14 ng/mL and 45 IU/mL, respectively. rTAP at a plasma concentration of 0.90 μmol/L completely inhibited TF/factor VIIa–dependent thrombus formation at wall shear rates of 650 and 2600 s−1. These shear conditions are comparable to those in medium-sized arteries and in moderately stenosed small arteries, respectively. In contrast to the TF-coated surface, rTAP was less efficient in reducing collagen-induced thrombus formation. While a significant reduction of 53% was observed at 650 s−1, thrombus formation at 2600 s−1 was not affected by rTAP. Thus, rTAP is an efficient inhibitor of thrombin-driven human thrombus formation on the TF-rich surface but less efficient when thrombus formation is elicited by type III collagen. The lack of antithrombotic effect on collagen type III at 2600 s−1 corroborates earlier findings, showing that collagen-induced thrombus formation in blood from patients with severe factor VIII deficiency is not affected at this blood flow condition and thus is not dependent on the prothrombotic effects of thrombin.
- Received February 28, 1995.
- Accepted August 23, 1995.
The biochemical events that compose the blood coagulation cascade play a major role in pathological vascular thrombus formation. Thrombin is the terminal enzyme in this cascade and has been shown to be the primary physiological mediator of fibrin formation and early platelet activation and recruitment to the growing thrombus.1 2 Thrombin also serves to structurally rigidify the growing thrombus through the activation of the transglutaminase factor XIII, which results in fibrin strand cross-linking.3
Based on the importance of thrombin in thrombosis, it is therefore not surprising that many of the new, more selective anticoagulants have been tailored to specifically inhibit the catalytic activity of thrombin4 5 6 or to inhibit the conversion of prothrombin to thrombin by interfering with the catalytic activity of factor Xa in the prothrombinase complex.7 8 9 Both approaches are promising, since each can effectively attenuate the primary thrombogenic source, which is thrombin activity. This has been demonstrated in vivo, where direct inhibitors of thrombin and factor Xa efficiently interrupt experimental thrombogenesis in various animal models.10 11 12 13 14 15 16 17 18 19 20 21 Furthermore, a number of direct and specific thrombin inhibitors are currently being evaluated in clinical trials as acute antithrombotic agents, with encouraging results thus far.22 23
TAP is a 60-amino-acid protein originally isolated from the soft tick Ornithodoros moubata that has been characterized as a potent and selective inhibitor of factor Xa catalytic activity.24 Antithrombotic efficacy greater than that observed with standard heparin has been demonstrated with rTAP14 in a wide variety of experimental preparations, including a primate model of disseminated intravascular coagulation,14 a rabbit model of stasis-induced venous thrombosis,17 canine models of coronary and femoral artery thrombosis,16 21 and a model of occlusive thrombosis in a prosthetic Dacron graft segment positioned in a femoral arteriovenous shunt in baboons.15 18
In the present study, we used an optimal inhibitory concentration of rTAP to establish the role of factor Xa in thrombus formation in nonanticoagulated human blood in an ex vivo model of shear-dependent thrombosis.25 Thrombus formation was elicited by TF or by collagen fibrils at arterial blood flow conditions,26 27 and it was of particular interest to compare the antithrombotic effect of rTAP on these two thrombogenic surfaces under two different arterial shear conditions (650 and 2600 s−1). The role of thrombin generation in collagen-induced thrombus formation at arterial shear conditions has not been characterized, although a previous study indicated that thrombin is partially required only at intermediate arterial shear rate.28
To evaluate the role of factor Xa in arterial thrombus formation, we had to develop a device allowing infusion and homogeneous mixing of rTAP into the flowing blood stream distal to the donor and proximal to the thrombogenic surface. By connecting this system to the parallel-plate perfusion device, we found that inhibition of factor Xa by rTAP virtually abolished thrombus formation on the TF surface. In contrast, thrombus formation was much less affected by rTAP on the collagen surface.
The mixing device consisted of five separate pieces, as shown in Fig 1⇓. A complex helix made of polypropylene (Extrude, Kerr) was fitted into a cylinder (Fig 1A⇓) made of polymethyl methacrylate with a length of 117 mm, OD of 10 mm, and ID of 7 mm. The dimensions of the helix were 75 mm (length) by 7 mm (ID). To gradually taper off the diameter of the outlet portal in the long cylinder, a short cylinder made of polymethyl methacrylate (length, 10 mm; OD, 7 mm; ID, 4 mm) was placed inside the long cylinder distal to the helix. All materials that came into contact with blood in this disposable device were coated with covalently bound unfractionated heparin (Carmeda AB).29 An inlet piece (Fig 1B⇓) made of polytetrafluoroethylene (Teflon) (Lilaas Finmekaniske AS) allowed infusion of optional solutions into the flowing blood. The IDs of the inlet tube and side tube were 1.8 mm. The inlet tube was coned to 0.8 mm where it entered the mixing device, while an outlet piece (Fig 1C⇓) of Teflon had an ID of 1.8 mm. The mixing device was connected to the ex vivo perfusion system proximal to the parallel-plate perfusion chamber25 by a silicone elastomer (Silastic) tubing (ID, 2.6 mm; Dow Corning Corp) (Fig 2⇓).
Mixing Efficiency of the Device
The mixing device was placed vertically to avoid the persistence of density layers of blood and test agent. The efficiency of mixing was first determined by a dye (0.2% Ponceau red in 3% trichloroacetic acid), with maximal absorbance at 518 nm, mixed with water. In another set of experiments, radiolabeled BSA (125I-BSA; Du Pont–NEN Products) diluted in saline was mixed with blood anticoagulated with trisodium citrate. Ponceau red and 125I-BSA were mixed with water or blood, respectively, into the mixing device by infusion with two separate syringe pumps (Alitea AB). The mixing agent flow rate and blood/water flow rate were 0.2 and 9.8 mL/min, respectively. A special device that divided the cross-sectional area of the blood flow into four equal portions (Fig 1D⇑) was placed about 140 mm distal to the mixing device. The fractions were collected, and the absorbance of Ponceau red (DU-70 Spectrophotometer, Beckman Instruments, Inc) or 125I radioactivity (Cobra II, Auto-Gamma Counting Systems, Packard Instrument Co) in each fraction was measured and calculated as percent of total. Homogeneous mixing should give 25% of the absorbance or 125I activity in each of the four outlets.
Blood Donors and Blood Sampling
Healthy nonsmoking individuals who denied ingestion of aspirin or other nonsteroidal anti-inflammatory drugs for at least 10 days before the experiments gave informed consent to donate blood to perfusion experiments. Venipuncture of an antecubital vein was performed with a No. 19 butterfly infusion set (Abbott Laboratories). The first 4.5 mL of blood was collected for determination of hematological parameters (Auto Counter AC 920, Swelab Instruments). The successive 50 or 40 mL was used for perfusion experiments characterizing the mixing device or for thrombus formation, respectively. Hemoglobin, hematocrit, and platelet count were within the normal ranges for all volunteers. In the perfusion experiments with rTAP, the same individuals donated blood for both rTAP and saline (control) on the same day.
Activation of Coagulation and Platelets and Heparin Release by the Mixing Device
Heparin-coated and noncoated mixing devices were compared to determine the effects on coagulation and platelet activation. Plasma levels of FPA, β-TG, and heparin were measured proximal and distal to the mixing device. After venipuncture of an antecubital vein, 0.9 mL of blood was collected into three Eppendorf tubes, each prefilled with 0.1 mL of appropriate anticoagulant mixtures, for plasma measurement of FPA, β-TG, or heparin (all from Diagnostica Stago). The butterfly infusion set was subsequently connected to the mixing device prefilled with saline (37°C), and the blood was drawn directly from the antecubital vein through the device by an occlusive roller pump (model M312, Gilson). The pump was placed distal to the mixing device, and the blood flow rate was maintained at 10 mL/min. At 3 and 4.5 minutes of perfusion time, blood samples (0.9 mL) were collected immediately distal to the mixing device into a syringe prefilled with 0.1 mL of the anticoagulant mixture of the FPA kit (Diagnostica Stago). The blood sampling was performed in a rubber-coated area of the tubing to prevent leakage after puncture of the Silastic tubing.30 After 5 minutes of perfusion, the roller pump was switched off, and two 0.9-mL blood samples for determination of β-TG and heparin were immediately collected into syringes prefilled with 0.1 mL of the β-TG and heparin anticoagulants (Diatube H, Diagnostica Stago).
Preparation of rTAP
rTAP was prepared from fed-batch fermentation media after expression in the methylotrophic yeast Pichia pastoris as described.31 The purified inhibitor was determined to be >98% pure by a number of analytical criteria, including reverse-phase high-performance liquid chromatography, electrospray mass spectrum analysis, quantitative amino acid analysis, and amino-terminal sequence analysis. The specific content of the lyophilized trifluoroacetic acid salt was determined by quantitative amino acid analysis. The inhibition of human factor Xa amidolytic activity by rTAP was determined and yielded a Ki of 0.187 nmol/L.
All experiments were done from a single lot of rTAP diluted in sterile saline. For perfusion experiments, rTAP was diluted to a concentration of 26 μmol/L (0.21 mg/mL).
Coagulation Assay With rTAP
The anticoagulant activity of rTAP was assessed by a one-stage clotting assay using a Thrombotrack coagulometer (Thrombotrack 4, Nycomed Pharma). Pooled citrated human plasma (100 μL) was preincubated at 37°C for 3 minutes. Different concentrations of rTAP (0 to 33 μmol/L; final concentrations, 0 to 1065 nmol/L) were added to the plasma after 2.5 minutes. Thromborel (Thromborel S; Behringwerke AS), which consists of a lyophilized TF preparation from human placenta, was diluted in 0.9% saline, and 100 μL of this dilution (37°C) was added to the plasma-rTAP mixture. Subsequently, 100 μL of prewarmed (37°C) 25 mmol/L CaCl2 was added, and the clotting time was automatically measured by the coagulometer. Detailed biochemical characterization of this Thromborel preparation was reported previously.26
Preparation of TF Surface
One vial of Thromborel was dispersed in 2 mL deionized water, incubated for 15 minutes at 37°C, and diluted 1:50 in coating buffer (0.1 mol/L sodium carbonate, pH 9.5). Thermanox plastic coverslips (Miles Laboratories) were stored in 70% ethanol and rinsed in deionized water before being incubated at 4°C for ≈17 hours in 2 mL of the Thromborel dilution. The coated coverslips were rinsed six times with sterile PBS and stored in PBS for a maximum of 7 hours before use in perfusion experiments.26 Thrombus formation triggered by this surface is dependent on TF, since inclusion of a monoclonal antibody that is directed against human TF and blocks complexing of TF and factor VIIa interrupts the thrombotic response completely.26
Preparation of Collagen Surface
Type III collagen was purified from a pepsin digest of human placenta by selective salt precipitation.32 Purification, fibril formation, and coating of coverslips were performed as previously reported.27 The collagen-coated coverslips were kept at 21°C for ≈16 hours before use in perfusion experiments. This surface does not activate coagulation.27
Ex Vivo Perfusions With rTAP, Fixation, and Embedding
Ex vivo perfusions33 were performed at 37°C. The mixing device was placed vertically between the donor and a parallel-plate perfusion chamber (Fig 2⇑). Both devices were prefilled with Buffer C (in mmol/L: NaCl 130, KCl 2, NaHCO3 12, CaCl2 2.5, MgCl2 0.9; pH 7.4, 37°C).34 Human nonanticoagulated blood was drawn directly from an antecubital vein and mixed with rTAP or saline (control) in the heparin-coated mixing device as described above. The blood was subsequently exposed to either a TF- or collagen-coated coverslip positioned in the perfusion chamber.25 27 The blood flow rate was maintained at 10 mL/min for 4 minutes by an occlusive roller pump placed distal to the perfusion chamber. Perfusion chambers with different geometrical dimensions of the blood flow channel were used, such that a flow rate of 10 mL/min produced shear rates of 650 and 2600 s−1 at the thrombogenic surface. These wall shear rates are encountered in medium-sized arteries and moderately stenosed small arteries,35 respectively. The reactive surface was exposed to blood for 3.5 minutes because of the 0.5-minute residence time of blood in the mixing device. Either rTAP dissolved in saline or saline (control) was infused with a syringe pump at a flow rate of 0.2 mL/min. The plasma levels of rTAP were calculated from the hematocrit values, and the mean final plasma concentration of rTAP was 0.90 μmol/L (range, 0.83 to 1.09 μmol/L).
The total volume in the tubing from the mixing device to the donor was 0.8 mL, which restricted the possibility of backflow of the injected compound into the antecubital vein of the blood donor during the 4 minutes of perfusion time.
After 4 minutes of blood perfusion, the mixing device and the blood donor were disconnected from the parallel-plate perfusion chamber device, which then immediately was perfused with Buffer C (37°C) for 20 seconds at 10 mL/min, followed by a perfusion with fixation solution consisting of 2.5% glutaraldehyde in 0.1 mol/L cacodylate, pH 7.4 (21°C), for 40 seconds. The flow was not interrupted during the 5-minute period of perfusion with blood, buffer, and fixation solution. The coverslip was removed from the chamber and kept in a freshly prepared fixation solution for 1 hour and was then stored in 0.1 mol/L cacodylate/7% sucrose at 4°C until being embedded in epoxy resin.36
FPA Plasma Levels During Blood Perfusions
Plasma FPA levels were measured proximal to the mixing device before perfusion and distal to the perfusion chamber after 3 minutes of perfusion. Blood samples for quantification of plasma FPA levels were processed further according to the manufacturer of the assay kit (Diagnostica Stago).
Microscopic evaluations of thrombotic deposits were performed on epoxy resin–embedded sections 1 μm thick prepared perpendicular to the direction of the blood flow 1 mm downstream from the upstream edge of the coverslip.37 The sections were stained with toluidine blue and basic fuchsin.38
Percent surface coverage with fibrin (% fibrin deposition), percent platelet-fibrin adhesion (TF surface), and percent platelet-collagen adhesion (collagen surface) were assessed by standard morphometry.38 The evaluations were carried out at ×1000 magnification with a Zeiss standard 25 light microscope (Carl Zeiss).
Computer-assisted morphometry was used to quantify platelet thrombus area (μm2/μm sectional length). Platelet thrombus volume (μm3/μm2) was derived from the sectional thrombus area as previously described.37 The evaluations were performed by a Kontron Vidas image analyzing unit (Zeiss) at magnifications of ×500 or ×2000, depending on the size of the thrombi.
Results are expressed as mean±SEM. Significance for paired data was calculated with a two-tailed Student’s t test. Values of P<.05 were considered significant.
The efficiency of mixing was studied with two different sets of experiments, in which either a dye (0.2% Ponceau red in 3% trichloroacetic acid) or 125I-BSA was used as a mixing agent. The device was found to efficiently mix the dye with water (data not shown) and 125I-BSA with citrated blood. Percentages of 125I radioactivity from the four test outlets were 24.1±0.9%, 23.5±1.1%, 26.9± 4.5%, and 25.5±2.5% (mean±SEM, n=6). The theoretical percentage of mixed label was 25% of the signal in each of the four outlets.
The minimal length of the helix required for homogeneous mixing was 75 mm.
Activation of Coagulation (FPA) and Platelets (β-TG)
The mixing device caused some activation of platelets and coagulation after 5 minutes of perfusion (10 mL/min) of nonanticoagulated blood. The plasma levels of FPA increased, on average, from 2.8 ng/mL at the flow inlet at time 0 to 14 ng/mL at the outlet of the heparin-coated device after 5 minutes of perfusion (P<.05) (Table 1⇓). The increase in FPA was higher in the noncoated device (from 2.8 to 24 ng/mL), although the plasma levels measured at 5 minutes in coated and noncoated devices were not statistically different. The corresponding β-TG values increased, on average, from 25 to 45 IU/mL (P<.05). However, the 5-minute β-TG sample was significantly increased by the noncoated system (85 versus 45 IU/mL; P<.05) (Table 1⇓). Apparently, thrombus formation on collagen at 2600 s−1 was not affected by the slight activation of coagulation and platelets by the mixing device, since the platelet-collagen adhesion of 43.1±3.7% and the platelet thrombus volume of 7.0±2.9 μm3/μm2 in the absence of the mixing device were not significantly different from thrombus formation in the presence of the mixing device (Fig 3A⇓ and 3B⇓).
Heparin release from the heparin-coated device into the perfused blood stream was not detected after 5 minutes of perfusion with nonanticoagulated blood.
Inhibitory Efficacy of rTAP in Plasma
The anticoagulant activity of rTAP was measured in a one-stage clotting assay. rTAP had a dose-dependent inhibitory effect (data not shown), and at 1.07 μmol/L rTAP (final concentration), the clotting time increased from 32 to >590 seconds.
rTAP Perfusion Experiments
Thromborel (TF/phospholipids) or collagen fibrils were exposed to nonanticoagulated human blood mixed with rTAP or saline (control) for 3.5 minutes. The mean final plasma concentration of rTAP was 0.90 μmol/L, and the shear rates at the thrombogenic surfaces were 650 or 2600 s−1.
Effect of rTAP on TF/Factor VIIa–Induced Thrombus Formation
Infusion of rTAP reduced fibrin deposition by 99% at both wall shear rates (650 and 2600 s−1, P<.01) (Fig 4A⇓). Platelets adhered exclusively to the fibrin mesh. rTAP reduced the platelet-fibrin adhesion by 94% at 650 s−1 (P<.05) and by 98% at 2600 s−1 (P<.01) (Fig 4B⇓). Thrombus volume was reduced by 99% at both wall shear rates (P<.001) (Fig 4C⇓).
Effect of rTAP on Collagen-Induced Thrombus Formation
rTAP had no effect on platelet adhesion (Fig 3A⇑) at either 650 or 2600 s−1. However, rTAP reduced the average thrombus volume by 53% at 650 s−1 (P=.03) but had no effect on thrombus volume at 2600 s−1. There was no detectable fibrin deposition on the collagen surface after 3.5 minutes of perfusion at these blood flow conditions.
Effect of rTAP on FPA Plasma Levels
The TF surface resulted in substantial FPA generation. rTAP reduced the FPA levels by 92% at 650 s−1 (P<.05) and by 93% at 2600 s−1 (Table 2⇓). However, the latter reduction was not significant because of the large variation in the control group. In contrast, the collagen surface resulted in little FPA generation, which was not affected by rTAP (Table 2⇓).
Among the various strategies taken in the development of novel antithrombotic therapies, the direct inhibition of factor Xa activity has been shown to be an effective approach. Inhibition of factor Xa has gained particular attention, since it can terminate the amplified burst of thrombin generation that plays an important role in arterial thrombus development. rTAP is a potent and selective inhibitor of factor Xa that has been shown to be superior to unfractionated heparin in animal models of in vivo thrombogenesis15 and shows a better efficacy than hirudin in dogs at doses that may limit the risk of concomitant hemorrhage often encountered in antithrombotic therapy.16 However, the antithrombotic effect of TAP in humans has not been studied either in TF/factor VIIa– or collagen-induced arterial thrombogenesis. The latter issues are of particular importance, since both TF and collagens are exposed to the bloodstream at sites of vascular lesions. Therefore, it was relevant to investigate and compare the antithrombotic effects of rTAP on thrombogenic surfaces consisting of either TF or collagen in nonanticoagulated human blood flowing at different arterial shear conditions. The main goal was to establish the role of factor Xa in thrombus formation. Therefore, a plasma concentration of rTAP known to block thrombus formation in baboons15 was chosen.
rTAP at a plasma concentration of 0.90 μmol/L efficiently inhibited TF/factor VIIa–dependent platelet-rich thrombus formation at shear conditions of both normal (650 s−1) and atherosclerotic (2600 s−1) coronary arteries. However, the collagen-induced thrombus formation was less affected by rTAP, since significant reduction of thrombus formation was observed only at shear rates corresponding to healthy coronary flow but not at a shear rate found at atherosclerotic lesions.
Previous experiments with collagen-induced thrombus formation in blood from patients with severe factor VIII deficiency28 have indicated that the thrombotic response at a shear rate of 650 s−1 is partly thrombin dependent, whereas this is not the case at 2600 s−1. The findings with rTAP are consistent with these results, and it is interesting to note that rTAP reduced the thrombus volume by 53% at 650 s−1, whereas the reduction of thrombus volume in blood from severe hemophiliacs was 69%. In contrast, neither rTAP nor deficiency of factor VIII affected the collagen-induced thrombus formation at 2600 s−1. Thus, thrombus formation triggered by a surface of type III collagen is apparently driven by a platelet-dependent mechanism that does not involve thrombin.
The striking inhibition of TF/factor VIIa–dependent thrombus formation by rTAP is also consistent with previous findings in which preincubation of the same TF-rich surface with a monoclonal antibody directed against human TF abolished thrombus formation as well.26 Therefore, thrombus formation on a TF-coated surface appears to be completely driven by thrombin and as such is dependent on the activity of factor Xa in the prothrombinase complex. These findings demonstrate for the first time that potent and specific factor Xa inhibition is an extremely efficient means for interrupting human, platelet-dependent thrombus formation on a procoagulant surface consisting of TF/phospholipids but is less so on a nonprocoagulant surface consisting of pure type III collagen fibrils.
The TF-dependent thrombus formation was blocked at an average plasma concentration of 0.90 μmol/L rTAP, thus at approximately a sevenfold higher concentration than the factor X concentration present in normal plasma (130 to 140 nmol/L). The antithrombotic efficiency of rTAP at this concentration was simultaneously reflected by the reduction of the plasma levels of FPA sampled downstream of the site of thrombus formation. However, the reduction in thrombus formation on the collagen surface at 650 s−1 was not paralleled by a reduction in FPA. Apparently this was due to the “background noise” triggered by the perfusion chamber device, which yields 15 to 20 ng/mL FPA without the mixing device.39 40 It should also be mentioned that the efficacy of rTAP in this human system remains to be established, since no dose response of the inhibitor was investigated in the present study. This is of interest, because rTAP plasma concentrations ranging from 0.5 to 1.0 μmol/L were previously reported to protect against thrombus formation in baboons.15
Development and characterization of a device that facilitated homogeneous mixing of an antithrombotic test agent into flowing native blood were a prerequisite for the successful completion of this study. Another type of mixing device other than the one we used was recently described,41 but no data on homogeneous mixing were reported. The present device, which makes use of a complex helix with covalently bound unfractionated heparin, mixes the test agent homogeneously in flowing native blood. However, this is at the expense of moderate activation of platelets and coagulation. Fortunately, heparin coating of the surfaces substantially reduces the level of activation. Although the β-TG plasma levels were lowered close to the upper limit of the normal range, the FPA plasma levels still remained significantly higher than the normal range. Nevertheless, these levels are within the range of those frequently encountered in patients suffering from thromboembolic disorders,42 and there was no evidence that this activation affected thrombus formation in the perfusion chamber, at least not on collagen. It is apparent that investigation of experimental antithrombotic agents in native blood in this novel human model of thrombogenesis offers great advantages at the expense, however, of modest upstream activation of the coagulation cascade.
The antithrombotic effect of rTAP on TF/factor VIIa–dependent arterial thrombus formation in humans is striking. The strategy of inhibiting the formation of thrombin and thus thrombus formation at the level of factor Xa seems to be a most efficient approach. However, as demonstrated by Fressinaud et al28 and the present investigation, collagen-induced thrombus formation at high arterial shear conditions (2600 s−1) appears to be thrombin-independent. Since both TF and a variety of collagen types are exposed to the blood flow at vascular lesions, inclusion of a selective inhibitor of platelet adhesion or aggregation in addition to a selective factor Xa inhibitor may be necessary.
Selected Abbreviations and Acronyms
|TAP||=||tick anticoagulant peptide|
We thank Maria J.A.G. Hamers (Nycomed Pharma AS) for excellent technical assistance. Professor Helge Stormorken and Dr Steinar Bergseth (Nycomed Pharma AS) are acknowledged for stimulating discussions.
Bevers EM, Comfurius P, Zwaal RFA. Platelet procoagulant activity: physiological significance and mechanisms of exposure. Blood. 1991;5:146-154.
Henschen A, McDonagh J. Fibrinogen, fibrin and factor XIII. In: Zwaal RFA, Hemker HC, eds. Blood Coagulation. Amsterdam, Netherlands: Elsevier Science Publishers; 1986:171-241.
Hanson SR, Harker LA. Interruption of acute platelet-dependent thrombosis by the synthetic antithrombin D-phenylalanyl-L-prolyl-L-arginyl chloromethyl ketone. Proc Natl Acad Sci U S A. 1988;85:3184-3188.
Jang IK, Gold HK, Leinbach RC, Fallon JT, Collen D. In vivo thrombin inhibition enhances and sustains arterial recanalization with recombinant tissue-type plasminogen activator. Circ Res. 1990;67:1552-1561.
Kelly AB, Marzec UM, Krupski W, Bass A, Cadroy Y, Hanson SR, Harker LA. Hirudin interruption of heparin-resistant arterial thrombus formation in baboons. Blood. 1991;77:1006-1012.
Dunwiddie C, Thornberry NA, Bull HG, Sardana M, Friedman PA, Jacobs JW, Simpson E. Antistasin, a leech-derived inhibitor of factor Xa. J Biol Chem. 1989;264:16694-16699.
Eidt JF, Allison P, Noble S, Ashton J, Golino P, McNatt J, Buja LM, Willerson JT. Thrombin is an important mediator of platelet aggregation in stenosed canine coronary arteries with endothelial injury. J Clin Invest. 1989;84:18-27.
Fitzgerald DJ, Fitzgerald GA. Role of thrombin and thromboxane A2 in reocclusion following coronary thrombolysis with tissue-type plasminogen activator. Proc Natl Acad Sci U S A. 1989;86:7585-7589.
Jang IK, Gold HK, Ziskind AA, Leinbach RC, Fallon JT, Collen D. Prevention of platelet-rich arterial thrombosis by selective thrombin inhibition. Circulation. 1990;81:219-225.
Krupski WC, Bass A, Kelly AB, Marzec UM, Hanson SR, Harker LA. Heparin-resistant thrombus formation by endovascular stents in baboons: interruption by a synthetic antithrombin. Circulation. 1990;81:570-577.
Neeper MP, Waxman L, Smith DE, Schulman CA, Sardana M, Ellis RW, Schaffer LW, Siegl PKS, Vlasuk GP. Characterization of recombinant tick anticoagulant peptide: a highly selective inhibitor of blood coagulation factor Xa. J Biol Chem. 1990;265:17746-17752.
Schaffer LW, Davidson JT, Vlasuk GP, Siegl PKS. Antithrombotic efficacy of recombinant tick anticoagulant peptide: a potent inhibitor of coagulation factor Xa in a primate model of arterial thrombosis. Circulation. 1991;84:1741-1748.
Sitko GR, Ramjit DR, Stabilito II, Lehman D, Lynch JJ, Vlasuk GP. Conjunctive enhancement of enzymatic thrombolysis and prevention of thrombotic reocclusion with the selective factor Xa inhibitor, tick anticoagulant peptide: comparison to hirudin and heparin in a canine model of acute coronary artery thrombosis. Circulation. 1992;85:805-815.
Vlasuk GP, Ramjit D, Fujita T, Dunwiddie CT, Nutt EM, Smith DE, Shebuski RJ. Comparison of the in vivo anticoagulant properties of standard heparin and the highly selective factor Xa inhibitors antistasin and tick anticoagulant peptide (TAP) in a rabbit model of venous thrombosis. Thromb Haemost. 1991;65:257-262.
Kelly AP, Hanson SR, Dunwiddie CT, Harker LA. Safe lasting interruption of arterial thrombosis by transient intravenous administration of tick anticoagulant peptide (TAP). Circulation. 1992;86(suppl I):I-411. Abstract.
Freund M, Mantz F, Nicolini P, Gachet C, Mulvihill J, Meyer L, Beretz A, Cazenave JP. Experimental thrombosis on a collagen coated arterioarterial shunt in rats: a pharmacological model to study antithrombotic agents inhibiting thrombin formation and platelet deposition. Thromb Haemost. 1993;69:515-521.
Mellot MJ, Stranieri MT, Sitko GR, Stabilito II, Lynch JJ, Vlasuk GP. Enhancement of recombinant tissue plasminogen activator-induced reperfusion by recombinant tick anticoagulant peptide, a selective factor Xa inhibitor, in a canine model of femoral arterial thrombosis. Fibrinolysis. 1993;7:195-202.
Gold HK, Torres FW, Garabedian HD, Werner W, Jang IK, Khan A, Hagstrom JN, Yasuda T, Leinbach RC, Newell JB, Bovill EG, Stump DC, Collen D. Evidence for a rebound coagulation phenomenon after cessation of a 4-hour infusion of a specific thrombin inhibitor in patients with unstable angina pectoris. J Am Coll Cardiol. 1993;21:1039-1047.
Waxman L, Smith DE, Arcuri KE, Vlasuk GP. Tick anticoagulant peptide (TAP) is a novel inhibitor of blood coagulation factor Xa. Science. 1990;248:593-596.
Ørvim U, Roald HE, Stephens RW, Roos N, Sakariassen KS. Tissue factor–induced coagulation triggers platelet thrombus formation as efficiently as fibrillar collagen at arterial blood flow conditions. Arterioscler Thromb. 1994;14:1976-1983.
Sakariassen KS, Joss R, Muggli R, Kuhn H, Tschopp TB, Sage H, Baumgartner HR. Collagen type III induced ex vivo thrombogenesis in humans: role of platelets and leukocytes in deposition of fibrin. Arteriosclerosis. 1990;10:276-284.
Fressinaud E, Sakariassen KS, Rothschild C, Baumgartner HR, Meyer D. Shear rate-dependent impairment of thrombus growth on collagen in nonanticoagulated blood from patients with von Willebrand disease and hemophilia A. Blood. 1992;80:988-994.
Roald HE, Ørvim U, Bakken IJ, Barstad RM, Kierulf P, Sakariassen KS. Modulation of thrombotic responses in moderately stenosed arteries by cigarette smoking and aspirin ingestion. Arterioscler Thromb. 1994;14:617-621.
Miller EJ, Rhodes RK. Preparation and characterization of the different types of collagen. Methods Enzymol. 1982;82:33-64.
Turitto VT, Baumgartner HR. Platelet-surface interactions. In: Colman RW, Hirsh J, Marder VJ, Salzman E, eds. Hemostasis and Thrombosis: Basic Principles and Clinical Practice. Philadelphia, Pa: JB Lippincott Co; 1987:555-571.
Baumgartner HR, Muggli R. Adhesion and aggregation: morphological demonstration and quantitation in vivo and in vitro. In: Gordon JL, ed. Platelets in Biology and Pathology. Amsterdam, Netherlands: Elsevier North Holland Biomedical Press; 1976:23-60.
Sakariassen KS, Roald HE, Salatti JA. Ex vivo models for studying thrombosis: special emphasis on shear rate dependent blood-collagen interactions. In: Hwang NHC, Turitto VT, Yen MRT, eds. Advances in Cardiovascular Engineering. New York, NY: Plenum Press; 1992:151-174.
Diquélou A, Lemozy S, Dupouy D, Boneu B, Sakariassen K, Cadroy Y. Effect of blood flow on thrombin generation is dependent on the nature of the thrombogenic surface. Blood. 1994;84:2206-2213.
Kirchhofer D, Tschopp TB, Hadváry P, Baumgartner HR. Endothelial cells stimulated with tumor necrosis factor-α express varying amounts of tissue factor resulting in inhomogeneous fibrin deposition in a native blood flow system: effects of thrombin inhibitors. J Clin Invest. 1994;93:2073-2083.