Apple Domain-Specific Anti-Factor XI Antibodies Inhibit Venous-Type Thrombosis with Improved Hemostatic Safety Profiles Compared to Enoxaparin in Primates

Abstract 1173

Severe inherited factor XI (FXI) deficiency decreases the risk of ischemic thromboembolic stroke and deep vein thrombosis, but can also cause bleeding in some patients. We therefore sought to further explore the roles of FXI in hemostasis and thrombosis in a baboon model, utilizing the anticoagulant monoclonal antibodies 14E11 and 1A6 that target specific functional domains of FXI. The 14E11 antibody binds to the apple 2 domain of FXI and inhibits FXI activation by factor XIIa (FXIIa), while not significantly inhibiting its activation by thrombin or the ability of FXIa to activate factor IX. 1A6 binds to the apple 3 domain and inhibits the feedback activation of FXI by thrombin as well as the activation of FIX by FXIa. Pre-treatment of baboons with 14E11, 1A6 (both 100 μg/kg, iv bolus), or the low-molecular-weight heparin enoxaparin (0.7 mg/kg iv bolus + 0.3 mg/kg/hr continuous infusion), prolonged the aPTT of baboons up to 3-fold. The antithrombotic effect of the anticoagulants were evaluated by measuring the accumulation of platelets and fibrin within high shear collagen-coated vascular graft segments (4mm i.d.), which models arterial-type thrombosis, as well as in a distal silicone expansion chamber (9mm i.d.) positioned downstream of the graft, which models the slower flow profile of venous thrombosis. The thrombogenic devices were deployed into flow restricted (100ml/min) chronic arteriovenous shunts for 1 hour. The treatments reduced platelet accumulation within the thrombus expansion chamber by 89%, 94%, and 93% in 14E11-, 1A6-, and enoxaparin-treated animals, respectively (P<0.001, n=5–6 for each treatment). Fibrin accumulation was also comparably lower by 80–90% in all treatment groups. While both 14E11 and 1A6 were remarkably efficacious against venous-type thrombogenesis, 1A6 was more antithrombotic than 14E11 in the higher shear proximal grafts that accumulate platelet-rich arterial-type thrombi (P=0.001, 14E11 vs. 1A6). The antithrombotic efficacy of 14E11 indicates that FXI activation by FXIIa may be a sufficient driver for acute venous thrombosis, whereas the efficacy of 1A6 to limit both arterial-and venous-type thrombogenesis indicates that the feedback activation of FXI by thrombin may be an important driver for arterial thrombosis. These data suggest that the mechanism of FXI activation and/or activity may vary depending on the different shear and flow conditions that occur during venous and arterial thrombosis. The effect of anticoagulation on hemostasis was assessed by measuring the template bleeding time (BT) in conjunction with high dose aspirin treatment (32 mg/kg). The average BT was not significantly increased by the FXI antibodies or enoxaparin treatment alone. Aspirin increased the BT by 51% (from 3.5±0.4min to 5.3±0.9min), while co-administration of aspirin with enoxaparin further increased the BT by 66% to 8.8±0.6min. Co-administration of aspirin with 14E11 or 1A6 resulted in BTs that were statistically indistinguishable from the BT prolongation with aspirin alone (4.9±0.7 min and 5.2±0.5 min respectively). Thus, in the presence of aspirin, anticoagulation by inhibiting plasma FXI procoagulant activity (1A6) or the FXI-mediated procoagulant activity of FXIIa (14E11) was hemostatically safer than anticoagulation with enoxaparin. In summary, our data suggest that targeting specific molecular pathways at the FXI/FXII axis with inhibitors specific for FXI apple domains may help delineate the roles of FXI in thrombosis and hemostasis, and that selective inhibition of FXIa activity or FXI activation may be effective and safer than enoxaparin for pharmacological thromboprophylaxis.