VWF-Mediated Platelet Adhesion is Required for Deep Vein Thrombosis in a Flow Restriction Model.

Abstract 473

Deep vein thrombosis (DVT) and its life-threatening complication, pulmonary embolism, are wide-spread in the Western world. Disturbance of blood flow without substantial endothelial denudation is a leading pathogenic factor for non-cancer related DVT. Von Willebrand Factor (VWF), a large multimeric protein, facilitates hemostasis via two separate pathways by stabilizing coagulation Factor VIII (FVIII) and by recruiting platelets to injured vessel wall or thrombi through the interaction with GPIb-alpha. Whereas the role of FVIII in DVT has been suggested by clinical studies (Koster T et al., Lancet, 345(8943):152-5,1995), whether VWF-platelet interaction is implicated in venous thrombosis remains unclear. We utilized murine models of partial and complete flow restriction in the inferior vena cava (IVC) in mice to mimic clinical conditions in which thrombus develops in deep veins. In 8-10 week old C57BL/6 male mice anesthetized by isoflurane-oxygen mixture, IVC and two side branches were ligated by a polypropylene suture immediately below the renal veins to obtain complete blood stasis. For partial flow restriction (stenosis), IVC ligation was performed over a 30G needle and then the needle was removed. Mice were euthanized after 48 h and thrombi from the IVC were taken for analysis. Results were evaluated using the chi-square test. The VWF^-/- mice were completely protected from thrombosis in the stenosis model: none of the 14 VWF^-/- mice developed a thrombus compared to 6/6 wild-type (WT) mice (p<0.001). In the stasis model, a similar albeit less pronounced phenotype was observed (33% of VWF^-/- mice with thrombus, n=9, versus 82% in WT mice, n=11; p<0.03). Stenosis-induced thrombi in WT mice contained abundant amounts of VWF, as was shown by immunostaining. To delineate the involvement of VWF-platelet interactions, we infused WT mice with GPG-290, a recombinant GPIb-alpha N-terminal domain conjugated with human IgG1 Fc fragment. This compound has been shown to inhibit VWF-GPIb-alpha interaction (Hennan JK et al., Thromb Haemost, 95(3):469-75, 2006), but does not interfere with FVIII binding and turnover. Infusion of GPG-290 markedly reduced thrombus development in the stenosis model (3/9 GPG-290-treated WT mice developed DVT versus 9/9 vehicle-treated control mice; p=0.003). Notably, in the absence of blood flow (stasis model), GPG-290 was less effective (83% thrombosis development in vehicle-treated WT, n=6, versus 55.6% thrombosis in GPG-290-treated WT group, n=9; p=0.26). We next addressed the events preceding thrombus formation in the DVT stenosis model using intravital microscopy on living mice. We observed accumulation of fluorescently labeled platelets and leukocytes in the IVC in the area below the suture 6 h after stenosis induction. The amount of both adhering platelets and leukocytes was substantially reduced in VWF^-/- mice compared to WT (approx. 20-fold, p<0.005 and 11-fold, p<0.001, respectively). In conclusion, VWF mediates platelet and leukocyte recruitment to the vessel wall. This initiates thrombus development in the absence of major endothelial injury. Interference with the VWF-GPIb-alpha axis may be a potential target for prophylaxis of deep vein thrombosis.