An Immunodeficient Mouse Model With Circulating Human Platelets To Evaluate The Platelet Clearance and Pharmacokinetics Of Platelet-Targeted Coagulation Factor VIIa

Treatment of bleeding episodes in hemophilia patients with inhibitory antibodies to factor VIII or factor IX by recombinant activated factor VII (rFVIIa, NovoSeven) has been sub-optimal partly due to its low affinity to activated platelets and its short half-life in circulation. To develop a rFVIIa variant with enhanced coagulant activity, we have targeted rFVIIa to platelets by monoclonal antibodies that recognize the human platelet receptor αIIbβ3. However, the assessment of pharmacokinetic parameters in mice is limited by the lack of recognition of mouse αIIbβ3 by antibodies to human αIIbβ3. The present study addresses the need to develop appropriate in vivo models to study this new class of bypass therapeutics.

First, we evaluated the survival of human platelets in hemophilia A, NOD/SCID, and NOD/SCID/gamma (NSG) mice. Platelet concentrates were prepared from normal human donors and transfused retro-orbitally into mice. The whole blood from dosed mice was then collected via tail vein laceration at various times and the human platelet counts in blood was determined by flow cytometry after staining with fluorescently labeled antibodies against human CD42b, mouse CD61, human FVII to visualize the human platelets, mouse platelets, and the FVIIa candidate that bound to human platelets, respectively. The half-life of human platelets in NOD/scid/gamma (NSG) mice was approximately 4 hours, which is considerably longer than the 0.8 hour half-life observed in hemophilia A mice.

The effect of platelet-targeted FVIIa variants on the clearance of human platelets was then investigated in NSG mice. The FVIIa candidates were pre-selected for their inactivity toward human platelet activation and aggregation, as determined in a battery of in vitro assays. In agreement with the in vitro results, all of these selected candidates did not affect the clearance of the transfused human platelets when dosed in NSG mice at 5 nmol/kg. In contrast, a control antibody fusion protein that is known to activate platelets and cause thrombocytopenia in vivo led to rapid platelet clearance in NSG mice. Similar results were observed for these proteins in cynomolgous monkeys when dosed at 2 nmol/kg.

The NSG mice with circulating human platelets were also explored to evaluate the clearance of FVIIa candidates that remain platelet-associated in vivo. To improve the pharmacokinetics, we have fused XTEN, a hydrophilic peptide that increases the dynamic radius of payload proteins, to the platelet-targeted FVIIa candidates. When tested in NSG mice model for the clearance of platelet-associated protein, addition of XTEN markedly reduced the clearance rate, resulting in several fold increase in exposure.

Together these data indicate that NSG mice with circulating human platelets can be used to assess the safety and pharmacokinetics of the platelet-targeted FVIIa variants, and the method can be adapted to evaluate other agents designed to utilize platelet-targeting approaches.