People with haemophilia A (HA) lack functional factor VIII (FVIII) and typically receive FVIII replacement therapy to prevent or treat bleeds. However, this requires frequent i.v. access, and efficacy is impaired in inhibitor patients. Mim8 is in development as a subcutaneous prophylactic treatment option for people with HA and HA with inhibitors. Like the recently approved emicizumab (Hemlibra®), Mim8 is a FVIII-mimicking human bispecific antibody bridging FIXa and FX. Mim8 is highly specific towards human FIXa (hFIXa) and human FX (hFX), preventing pre-clinical testing in standard rodent haemophilia models. Pharmacologic characterisation can be conducted in vitro and ex vivo utilizing human components. In vivo studies are feasible in primates due to high sequence homology between human and monkey FIX and FX, however, haemophilic mice are used for the most well-established and widely recognized bleeding models. Our aim was to establish a method to evaluate acute effects of Mim8 using in vivo bleeding models in HA mice, and to compare the potency and efficacy of Mim8 to a sequence-identical-analogue (SIA) of emicizumab.
A protocol for dosing HA mice with hFIX and hFX was optimized based on in vitro Thrombin Generation Assay (TGA) in HA mouse plasma spiked with a range of hFIX and hFX concentrations. The thrombin levels required to stop bleeding in the in vivo Tail Vein Transection (TVT) model were known from previous studies. In mouse plasma with a clinically efficacious concentration of 300-350 nM of emicizumab SIA (Mahlangu J et al, N Engl J Med. 2018 Aug 30;379(9)), we found that roughly twice the normal human levels of hFIX and hFX were needed to achieve sufficient thrombin generation for the TVT model. To maintain concentrations at or above this level throughout the bleeding experiments, in vivo doses were set at 1.5mg/kg and 0.9mg/kg, respectively. Based on the in vitro optimization, the haemostatic effect of Mim8 was evaluated by three different methods:
1. Tail Vein Transection (TVT), a venous in vivo bleeding model sensitive to clinical doses of FVIII, and
2. Tail Clip (TC), an arteriovenous in vivo bleeding model with lower sensitivity to FVIII, presumably due to the more severe nature of the bleed, and
3. Ex vivo TGA on plasma from Mim8-dosed mice
Briefly, mice were anaesthetized with isoflurane and dosed with hFIX, hFX and test compound. Thereafter, they were subjected to either the TVT bleeding model, the TC bleeding model, or cardiac puncture for plasma collection and ex vivo TGA. All mice were euthanized without awakening from anaesthesia.
Both compounds were efficacious in vitro in TGA (Figure A). Potency of Mim8 was significantly greater compared to emicizumab SIA; the efficacy of approximately 40 nM Mim8 corresponded to 300 nM emicizumab SIA. At the highest concentrations (>1000 nM), Mim8 efficacy tapered off, but remained superior to 300 nM emicizumab SIA. In TVT in HA mice, bleeding was reduced in a dose-dependent manner with an ED50 of 0.05 mg/kg for Mim8 or 0.7 mg/kg for emicizumab SIA. Statistically significant reduction of blood loss was observed at doses of or above 0.1 mg/kg Mim8 and 10mg/kg emicizumab SIA, corresponding to measured plasma concentrations above 10 nM for Mim8 and 300nM for emicizumab SIA. In the more severe TC model, blood loss was significantly reduced after treatment with 10 mg/kg of Mim8, whereas the tested doses of emicizumab SIA were not efficacious (Figure B). Mice treated with 4.6 and 10 mg/kg of Mim8, corresponding to a plasma concentration of up to approx. 1000nM, bled significantly less than mice treated with emicizumab SIA. In agreement with the in vitro TGA results, the 22 mg/kg dose (plasma concentration >2200 nM) appeared less efficacious; association of FIXa and FX to different Mim8 molecules is the likely cause. The increased potency of Mim8 was confirmed in TGA ex vivo.
In conclusion, we developed a method for evaluating the FVIII-mimetic compounds Mim8 and emicizumab SIA, which require human FIX and FX, in a murine system. This method may be applicable for testing of other FIXa-FX bridging compounds lacking rodent cross-reactivity. In Thrombin Generation Assay and the Tail Vein Transection model, Mim8 showed significantly increased potency compared to emicizumab-SIA, and the observed potency gain corresponded to in vitro findings in the human system. Furthermore, Mim8 could stop a severe bleed in the tail clip model, which was not possible with the tested doses of emicizumab SIA.
Figure
Disclosures
Ley: Novo Nordisk A/S: Employment, Equity Ownership. Holm:Leo Pharma A/S: Employment, Equity Ownership; Novo Nordisk A/S: Employment, Equity Ownership. Elenius:Leo Pharma A/S: Employment, Equity Ownership; Novo Nordisk A/S: Equity Ownership, Other: Previous employment. Holmberg:Novo Nordisk A/S: Employment, Equity Ownership. Bjelke:Novo Nordisk A/S: Employment, Equity Ownership, Patents & Royalties: Patents. Loftager:Novo Nordisk A/S: Employment, Equity Ownership. Hermit:Novo Nordisk A/S: Employment, Equity Ownership, Patents & Royalties: Patents. Hilden:Novo Nordisk A/S: Employment, Equity Ownership, Patents & Royalties. Kjellev:Novo Nordisk A/S: Employment, Equity Ownership.
Thrombin Generation Assay and Its Application in the Clinical Laboratory
