A Novel In Vivo Model to Study the Role of Excess FVIII Function in Venous Thrombosis

Background: Clinical data demonstrate that elevated FVIII activity, even modestly, is a strong and independent risk factor for venous thromboembolic events (VTE) with an, as yet, undefined mechanism. Elucidating the mechanism of FVIII-associated prothrombotic risk has widespread clinical applicability and is particularly relevant for defining the safe therapeutic window of gain-of-function FVIII variants for HA gene therapy. To pursue this line of work, we sought to develop a robust in vivo venous thrombosis model sensitive to excess FVIII function and altered FVIIIa regulation that could be paired with in vitro biochemical studies of FVIIIa function and regulation. Our underlying hypothesis posits that excess FVIIIa generation propagates coagulation, resulting in increased FVIIIa saturation by FIXa that protects against FVIIIa inactivation, thereby leading to both enhanced clot formation as well as reduced FVIIIa regulation.

Methods: We adapted the femoral vein thrombosis model (Gollomp et al., 2018; Welsh et al., 2019) to initiate coagulation with 2% FeCl₃. Clot formation was directly visualized and platelet and fibrin accumulation were quantified by intravital confocal microscopy. Prior to injury, mice were administered fluorescently labeled non-inhibitory antibodies to platelets and fibrin with or without variable amounts of recombinant hFVIII protein in PBS. Then, topical application of 2% FeCl₃ to the femoral vein was applied for 5 minutes. Thereafter, fluorescent images were collected over 30 minutes, and confocal time-lapsed images were analyzed with Slidebook 6 software. In vivo FVIII protein recovery was confirmed by measuring chromogenic FVIII activity from a blood draw collected within 5 minutes of injury; FVIII activity values were within 20% of the stated experimental group.

Results: HA mice demonstrated no platelet or fibrin accumulation. Consistent with model sensitivity to excess FVIII function, mice administered recombinant FVIII to achieve 300% FVIII activity had a 2-fold increase in platelet sum intensity compared to WT mice. To preliminarily support that the model was sensitive to disrupted FVIIIa regulation by activated protein C (APC), the assay was repeated on CRISPR/cas9 generated mice expressing a FVIII variant that cannot be cleaved by APC (FVIII^QQ mice). FVIII^QQ mice demonstrated a 3.3-fold increase in platelet accumulation compared to WT mice and a 1.4-fold increase compared to mice with 300% FVIII:Ag; these data are overall consistent with the established approximate 4-5-fold enhanced potency of FVIII-QQ over FVIII-WT (Wilhelm et al., 2021; Sternberg et al., 2024). Lastly, to confirm that observations may correlate with an established prothrombotic risk in mice and humans that similarly probes the APC pathway, we assayed homozygous FV-Leiden (FVL) mice. Homozygous FVL mice demonstrated a 4.5-fold increase in platelet accumulation compared to WT mice and a 1.8-fold increase compared to mice with 300% FVIII:Ag.

Conclusions: These data support that the 2% FeCl₃ femoral vein injury model is sensitive to excess FVIII function and the impact of APC on FVIIIa regulation, and suggests this model may be used to study FVIII-associated prothrombotic risk in vivo. Long-term, we aim to employ this model to elucidate the mechanisms by which excess FVIII confers prothrombotic risk and to inform a therapeutic window of gain-of-function FVIII variants that bypass mechanisms of FVIIIa-regulation for second-generation hemophilia A gene therapy approaches.