Gene Transfer of Canine FVIIa Results in Partial Correction of Canine Hemophilia: A Model for FVIII/FIX Gene-Based Bypassing Agents.

Despite its extensive use particularly in the management of hemophilic inhibitor patients, recombinant Factor VIIa (rhFVIIa) infusion has important limitations stemming from the nature of FVIIa itself, since its short half-life necessitates repeated injections and also carries high treatment costs. To overcome these, we have designed a gene transfer approach using a modified FVII transgene that is cleaved intracellularly and secreted in the active form, FVIIa. Using the human and murine analogue of this engineered transgene we have shown phenotypic correction of hemophilia B mice, following adeno-associated virus (AAV) - mediated, liver-directed gene delivery (Margaritis et al., 2004). In order to demonstrate efficacy in a large animal model of hemophilia, we cloned the canine Factor VII cDNA and generated the canine homologue of our modified transgene (cFVIIa). Recombinant cFVII zymogen and cFVIIa were purified and characterized in vitro in a clotting-based assay using canine reagents only (activated partial thromboplastin time [aPTT]). We found that cFVIIa had activity indistinguishable from rhFVIIa, while cFVII zymogen had negligible activity (5% rhFVIIa). In order to demonstrate in vivo efficacy, we produced 4 lots of an AAV8-based vector directing liver-specific expression of cFVIIa with similar vector titers (2–5 E13 vector genomes [vg]/ml). In hemophilia A (HA) or B (HB) mice, tail-vein delivery of 0.3 – 1.2 E12 vg/mouse (1.2 – 4.8 E13 vg/kg) resulted in long-term normalization of the hemophilic phenotype, demonstrating that cFVIIa can correct the defect in murine hemophilia. We proceeded to infuse 4 hemophilia dogs, with increasing vector doses: HB male (2.06 E13 vg/kg); HA male (6.25 E13 vg/kg); HA female (1.25 E14 vg/kg); HA male (1.25 E14 vg/kg). None of the dogs showed any adverse effects following vector delivery at any dose (the initial HB dog has been followed for almost 2 years [ongoing]). We followed the level of gene expression by clotting assays (prothrombin time [PT]/aPTT) and whole blood clotting time (WBCT). The initial dose of 2.06 E13 vg/kg resulted in a transient reduction in the PT/aPTT/WBCT. A considerable and sustained reduction in PT (18 sec, normal is ∼25 sec), aPTT (19 sec, normal is ∼30 sec, hemophilic is >40sec) and WBCT (25min, normal is ∼15min, hemophilic is >40min) was observed following administration of 6.25 E13 vg/kg in an HA male dog. Two more HA dogs were infused with 1.25 E14 vg/kg (male and female). The female HA dog exhibited only a modest decrease in aPTT (22sec), despite the vector dose increase, and a reduction in WBCT (30min), an observation that could be due to previously described gender-specific effects on gene expression. From preliminary and ongoing observations, the male HA dog infused also exhibited a decrease in WBCT. As an efficacy endpoint, the dogs exhibited a total of 3 bleeding episodes (none likely to be spontaneous, occurred in the lowest dose HB dog) in a cumulative time period of 38.5 months, compared to the expected 16 episodes (Brunetti-Pierri et al., 2005). In summary, our results demonstrate for the first time that gene transfer using a Factor VIII/Factor IX bypassing agent (canine FVIIa) can result in partial correction of the hemophilic phenotype in a large animal model of hemophilia.