Prevention of vascular graft occlusion and thrombus-associated thrombin generation by inhibition of factor XI

Abstract

The protease thrombin is required for normal hemostasis and pathologic thrombogenesis. Since the mechanism of coagulation factor XI (FXI)-dependent thrombus growth remains unclear, we investigated the contribution of FXI to thrombus formation in a primate thrombosis model. Pretreatment of baboons with a novel anti-human FXI monoclonal antibody (aXIMab; 2 mg/kg) inhibited plasma FXI by at least 99% for 10 days, and suppressed thrombin-antithrombin (TAT) complex and beta-thromboglobulin (betaTG) formation measured immediately downstream from thrombi forming within collagen-coated vascular grafts. FXI inhibition with aXIMab limited platelet and fibrin deposition in 4-mm diameter grafts without an apparent increase in D-dimer release from thrombi, and prevented the occlusion of 2-mm diameter grafts without affecting template bleeding times. In comparison, pretreatment with aspirin (32 mg/kg) prolonged bleeding times but failed to prevent graft occlusion, supporting the concept that FXI blockade may offer therapeutic advantages over other antithrombotic agents in terms of bleeding complications. In whole blood, aXIMab prevented fibrin formation in a collagen-coated flow chamber, independent of factor XII and factor VII. These data suggest that endogenous FXI contributes to arterial thrombus propagation through a striking amplification of thrombin generation at the thrombus luminal surface.

Figure 1

Thrombogenic vascular graft and blood collection device. Thrombogenesis started and thrombi developed in a collagen-coated (4-mm internal diameter [id]) expanded-polytetrafluoroethylene vascular graft that was deployed for 60 minutes into a chronic high flow arteriovenous shunt in healthy baboons. Blood samples were drawn from the coagulation marker concentration boundary layer by a syringe pump 1 cm downstream from acutely developing thrombi. PPACK anticoagulant was infused 3 mm proximal to the sample port to prevent the sample port from occluding during the 1-hour study. Blood flow through the graft was maintained at a fixed rate of 100 mL/min for the entirety of each study by proximal clamping.

Figure 2

The anti-FXI monoclonal antibody binds human and baboon FXI, and inhibits fibrin formation in FXII-inhibited or deficient human blood under flow. (A) Binding of the anti-FXI monoclonal antibody (aXIMab) to the FXI dimer (160 kDa) in platelet free NHP and NBP was demonstrated by Western blotting, developed using a secondary anti–mouse IgG antibody. aXIMab binding was minimal in FXI-depleted plasma or plasma from a FXI-deficient patient. (B) aXIMab prevented visible fibrin formation in FXII-inhibited or deficient blood under flow. Human whole blood, anticoagulated with CTI (40 μg/mL) to inhibit FXIIa, or reconstituted FXII deficient human blood was perfused through collagen-coated capillary tubes (thrombus chambers) at a shear rate of 265 s⁻¹ for 10 minutes. Before each experiment, blood was incubated with unfractionated heparin (15 U/mL), anti-TF antibody (20 μg/mL) plus FVIIai (1 μg/mL), aXIMab (20 μg/mL), CTI (40 μg/mL) for reconstituted blood where indicated, or PBS vehicle (−), as marked above each panel. Images were obtained via Kohler-illuminated Nomarski differential interference contrast microscopy with a Zeiss Axiovert 200M microscope using a Zeiss 63× oil-immersion 1.40 NA plan-apochromat lens (Zeiss, Göttingen, Germany). Images were captured using a Zeiss AxioCam with Slidebook 4.0 (Intelligent Imaging Innovations, Denver, CO) after 3 minutes of perfusion with modified Tyrodes buffer to wash the thrombus of unbound cells. All experiments were performed at 37°C. Each image is representative of 2 or 3 experiments.

Figure 3

Sustained inhibition of circulating FXI procoagulant activity after administration of aXIMab. An intravenous injection of aXIMab (2 mg/kg) was given over 5 minutes to a single baboon. Plasma samples were collected into citrate anticoagulant and tested over 4 weeks for (A) FXI procoagulant activity, FXI antigen (FXI:Ag), and (B) inhibitor levels, with each time point being the mean of duplicate measurements. The FXI:Ag ELISA could be used to detect both free and complexed FXI. Because the Bethesda assay detected only free FXI inhibitor (aXIMab), FXI:Ag and FXI activity at low inhibitor levels did not correlate until all complexes were cleared from circulation. The 1-hour time points for all levels were omitted for clarity. (C) A Western blot of 1-μL NHP and NBP samples size-fractionated by nonreducing 7.5% SDS-PAGE. Detection was with a polyclonal antibody against human FXI. The 5 lanes on the right represent samples before (0) or 1, 6, 8, and 27 days after infusion of aXIMab. XI indicates 100 ng purified human FXI; NHP, normal human plasma; NBP, normal baboon plasma; and XI-DP, FXI-deficient human plasma.

Figure 4

FXI inhibition reduces platelet and fibrin deposition on collagen-coated vascular grafts. Effects of FXI inhibition on (A) platelet and (B) fibrin deposition on collagen-coated (4-mm id) vascular grafts. The grafts were placed in vehicle-treated (n = 8) or aXIMab-treated (n = 6) animals (≥ 99% FXI inhibition). Blood was allowed to flow through the devices at 100 mL/min, producing an initial average wall shear rate of 265 s⁻¹. Significance levels (**P < .01, ***P = .001) were calculated by the 2-tailed Student t test. Values are means plus or minus SEM.

Figure 5

FXI inhibition reduces thrombin generation and platelet activation. Local and systemic TAT (A,B) and βTG (C,D) levels were monitored during thrombus formation. Blood samples were also tested for the fibrinolysis product D-dimer (E). Local values are those taken from the near wall, low flow concentration boundary layer, 1 cm distal to the growing thrombus, over the course of 10 minutes before its designated time; systemic samples were taken from the arteriovenous shunt proximal to the thrombogenic device. Zero time points in all groups are from samples taken systemically immediately before each study. FXI inhibition (n = 6) was associated with a reduction in local thrombin formation and platelet activation, which translated into lower systemic TAT and βTG levels in aXIMab-treated animals than in collagen controls (n = 7), within 60 minutes after treatment. No significant change in D-dimer release during acute thrombus formation (≤ 60 minutes) was detected in aXIMab-treated and control animals. The silicone rubber tubing without the collagen-coated graft segment (n = 5) did not induce significant increases in coagulation or platelet activation during these studies. Values are means plus or minus SEM.

Figure 6

FXI inhibition limits vascular graft-associated platelet deposition under high arterial shear. Effects of FXI inhibition or ASA administration on platelet deposition on collagen-coated (2-mm id) vascular grafts are shown. Collagen-coated vascular graft segments were placed in permanent arteriovenous shunts in untreated (n = 9), ASA-treated (n = 6), and aXIMab-treated (n = 5) animals. Blood was allowed to flow through the grafts at a rate of 100 mL/min, producing an average initial wall shear rate of 2120 s⁻¹. The flow was maintained by the pulsatile arterial pressure until the graft occluded (defined as ≤ 20 mL/min flow rate). Thrombi that formed in the grafts in the aXIMab-treated animals were unstable, embolized more frequently than in ASA-treated animals, and did not occlude the grafts for at least 60 minutes. Significance levels (*P = .05; **P < .01) pertain to comparisons with untreated controls, using the log-rank test, with the nonoccluded devices being censored. Values are means plus or minus SEM.