Complete Correction of Severe Canine Hemophilia B By Skeletal Muscle Directed AAV-Based FIX-Padua Gene Therapy in Inhibitor-Prone Dogs

Hemophilia B (HB) is an X-linked bleeding disorder caused by a deficiency in Factor IX (FIX). Ongoing gene therapy clinical trials for HB using adeno-associated viral (AAV) vectors targeting the liver have demonstrated sustained FIX expression at therapeutic levels and improvement of the bleeding phenotype in severe HB patients. However, one of the major safety concerns is the cellular immune response directed against vector capsid antigens that can limit FIX expression if not promptly treated with transient immunosuppression. Emerging evidence suggests that this complication is vector dose dependent, occurring only at doses of 2 x 10¹² vg/kg or higher.  FIX-Padua, a naturally occurring hyperactive FIX variant we have preciously described (Simioni et al. NEJM 2009), which exhibits ~8-fold increase in specific activity, offers a potential way to decrease the vector dose while maintaining hemostatic efficacy. In preclinical studies, we have recently shown that in severe HB dogs with a high risk for inhibitor formation to canine FIX (cFIX) protein or with preexisting inhibitors to cFIX, liver expression after AAV-cFIX Padua gene therapy resulted in immune tolerance induction and eradication of inhibitory antibodies (Crudele et al., Blood 2015).  In addition, the first description of an ongoing clinical trial with AAV-FIX-Padua reported therapeutic FIX levels without the development of FIX inhibitors. However, the success of liver-directed gene therapy approaches is not applicable to those HB patients with underlying liver disease, mostly due to HCV infection that affects >80% of adults with hemophilia treated with plasma-derived products. Thus, we sought to investigate AAV delivery to skeletal muscle as a potential alternative for these patients.

Clinical trials and preclinical studies in HB dogs have demonstrated that the therapeutic vector dose for AAV muscle gene therapy was a log higher than liver GT when direct multi-site intramuscular injection was used; however, intravascular delivery to the muscle allows the vector dose to be significantly reduced to 2-3 x 10¹² vg/kg. Here, we used a regional peripheral transvenular injection of AAV serotype 6 to express cFIX-Padua in the skeletal muscle of HB dogs.  These dogs are at high risk for inhibitor formation due to the null mutation in their canine F9 gene resulting in no mRNA or protein expression; typically, a single injection of cFIX wild-type protein (0.5 mg) results in the formation of a high titer inhibitor. Two adult dogs received gene therapy with 3 x 10¹² vg/kg of AAV6-cFIX-Padua under the control of a CMV promoter.  One dog (Malani) reached a plateau of 80-100% circulating cFIX activity levels with a follow-up time of 10 months (ongoing). Sustained antigen levels are 7-10%, resulting in ~10-fold higher specific activity than wild type FIX.  The second dog (Una) achieved 45-60% sustained activity with corresponding 4-5% antigen levels (~11-fold increase in specific activity) after 14 months (ongoing observation). Neither dog has developed inhibitors to cFIX. Both dogs are also negative for non-neutralizing IgG1 and IgG2 against cFIX. Whole blood clotting time and thromboelastogram parameters have also normalized in both dogs. Notably, immune tolerance has been maintained despite challenge with 0.5 mg of cFIX-wild type protein. Together, these data support the use of skeletal muscle as a target tissue for the expression of FIX Padua and, in contrast to an early direct AAV-FIX intramuscular injection trial, patients with null mutations can now be enrolled. This is the first demonstration of complete correction of HB in large animal models using AAV gene therapy targeting the skeletal muscle and supports the feasibility and safety for potential clinical studies.