Enhanced Efficacy of Platelet-Derived FVIII over Plasma FVIII in the Presence of Anti-Platelet Antagonist

Hemophilia A (HA) is a prime model for gene therapy. The FDA has approved AAV-mediated liver gene therapy to treat patients with HA. However, pediatric and adult patients with severe liver disease or neutralizing antibodies to AAV, which are present in 30-50% of the population, are excluded from AAV liver gene therapy. Our previous studies have demonstrated that platelet-targeted FVIII expression (2bF8) is effective in rescuing the bleeding phenotype of HA (F8^null) mice even in the presence of anti-FVIII inhibitory antibodies. Currently, our gene therapy approach is in Phase I clinical trial. Since antiplatelet drugs are commonly used in the clinic, it is important to investigate how platelet antagonists affect the efficacy of platelet-derived FVIII (plt-F8). Here, we evaluated two categories of platelet antagonists: one is the antibody (Leo.H4) that specifically blocks the αIIbβ3 complex binding to fibrinogen, and the other is the platelet inhibitor [Plavix (Clopidogrel)] that inhibits the P2Y₁₂ subtype of the ADP receptor to investigate how anti-platelet medications affect the clinical efficacy of plt-F8.

Hemostatic properties of plt-F8 were assessed by ex vivo native ROTEM and native whole blood TGA (nWB-TGA) assays and multiple in vivo injury models, including the lateral tail vein transection (TVT) injury model, tail tip transection (TTT) injury model, FeCl₃-induced carotid artery injury model, and 6-hour tail bleeding test. In 2bF8^Tg/F8^null mice with a plt-F8 level of 6.89±0.65 or 12.32±0.73 mU/10⁸ plts, all parameters from ROTEM analysis were comparable to those in WT control mice. However, all parameters, including clotting time, clot formation time, maximal clot firmness, α-angle, and A10 were significantly impaired in the presence of Leo.H4 compared to native samples from 2bF8 mice. In contrast, in samples from WT mice, only α-angle and A10 were significantly tempered in the presence of Leo.H4. Interestingly, peak thrombin and endogenous thrombin potential determined by nWB-TGA were significantly enhanced in the presence of Leo.H4 compared to native samples from 2bF8^Tg/F8^null mice, but there were no significant differences in other parameters. Leo.H4 did not impact the hemostatic parameters determined by nWB-TGA in WT mice. Complete occlusion occurred in the FeCl₃-induced carotid injury in all 2bF8 and WT mice, but none occluded when Leo.H4 was infused. In the TVT injury model, when Leo.H4 was infused, none of the 2bF8^Tg/F8^null mice stopped bleeding in the 10-minute bleeding test, and animals lost a significant amount of blood (525.98±92.06 μL). In contrast, when animals were infused with IgG isotype control, the bleeding time was 0.99±0.05 min, and blood loss was 17.05±9.37 μL. There were no significant differences in the bleeding time and blood loss between the 2bF8 and WT groups infused with IgG in the TVT test.

When mice were treated with Plavix, all 2bF8 and WT animals bled the entire 10 min during the TTT test, while the bleeding time was 0.88±0.39 and 1.43±1.19 min, respectively, in vehicle-treated animals. The primary blood loss in the Plavix-treated animals was 84.1±68.18 and 56.74±49.16 μL in the 2bF8 and WT groups, respectively. The blood loss in these animals was 13- and 16-fold more than in vehicle-treated control animals (6.4±6.19 and 3.5±4.67 μL, respectively, in the 2bF8 and WT groups). In the TVT injury model, the total blood loss was 10- and 14-fold more than in the vehicle controls. In the 6-hour tail clip bleeding test, the bleeding time was significantly longer (3.63±1.06 and 3.8±1.64 hours, respectively) compared to the vehicle-treated groups, in which all animals stopped bleeding in 1 hour, and their remaining hemoglobin levels were significantly lower (57.06±6.14% and 43.99±11.38%) compared to vehicle controls (81.86±12.17% and 95.13±3.21%, respectively) in both the 2bF8 and WT groups. Interestingly, the percentage of the remaining hemoglobin in the 2bF8 group was significantly higher than in the WT group after the Plavix treatment, although the plt-F8 level in the 2bF8 group corresponds to only 24% of that in the WT group.

In conclusion, our results demonstrated that platelet aggregation is critical in maintaining the clinical efficacy of platelet gene therapy in HA and anti-platelet medications can affect the hemostatic efficacy of plt-F8. However, platelet-derived FVIII has better hemostatic efficacy than plasma FVIII in the presence of clopidogrel.

No relevant conflicts of interest to declare.