Employing Factor IX Variants to Avoid Limitations Imposed by Immune Recognition of AAV Vector in Hemophilia B Gene Therapy

Abstract 3124

Persistent factor IX expression and phenotypic improvement have been achieved in a human clinical trial for hemophilia B using liver-directed adeno-associated virus (AAV) gene therapy vectors. An ongoing clinical trial uses a vector incorporating self-complementing AAV (scAAV) genome form, factor IX codon optimization (FIXopt) and AAV serotype 8 capsid. As was seen in a previous single-strand AAV serotype 2 trial, dose escalation has been associated with apparent immune-mediated transient inflammation of vector-transduced liver, although in contrast to the previous trial persistent FIX expression has been maintained for the first time. Taken together, these important trials define a consistent threshold load of AAV capsid that has stimulated capsid-specific cytotoxic lymphocyte recognition and potential transaminitis. To advance the successes achieved in these trials while providing a clear margin of safety so that this immunogenic threshold need not be approached, we have pursued steps to limit further the AAV capsid load. Single amino acid substitutions at arginine 338 in the FIX catalytic domain generate FIX variants with increased specific activity. We separately substituted either R338A, R338Q, or R338L (FIX Padua) into a codon optimized human factor IX cDNA and evaluated F.IX expression in tissue culture following plasmid DNA transfection of HEK 293t cells. Each R338 substitution improved FIX specific activity, up to 10 times increased over wild type using the R338LFIXopt cDNA. We next generated scAAV8 vectors incorporating a liver-specific transthyretin (TTR) promoter to express optimized codon F.IX cDNA with or without the R338L substitution. FIX^−/− mice receiving portal vein injection of 1 × 10¹⁰ vg/animal (4 ×10¹¹ vg/kg) expressed 86.5% of normal FIX activity at 2 months post-transduction from the WTopt vector and 330% normal from the R338LFIXopt. Incorporation of R338Lopt variant resulted in at least 6 to 10 fold increase in FIX specific activity over a follow-up of > 40 weeks. At ten months following FIX gene delivery, mice underwent a tail transection bleeding challenge. FIX vector mice demonstrated therapeutic protection from this major bleeding challenge and furthermore all survived with no late rebleeding (a hallmark of hemophilic phenotype). Greater than 100% normal human FIX activity was maintained for >40 weeks following treatment with the R338LFIX vector (v. 26.3% at euthanasia in WTopt vector group). The prolonged follow-up permitted extended safety evaluation. Factor IX inhibitor antibodies were not detected in any mice throughout the follow-up; FIX-binding IgG1 and IgG2 were negative also. Thrombin/antithrombin III complexes (TAT) examined at 12 weeks and at >30 weeks of age in R338LFIXopt vector mice did not differ from levels in WTFIXopt vector-treated or age-matched C57Bl/6 hemostatically normal mice. Necropsy at 40–44 weeks after vector (1 year of age) showed only age-related changes with no microvascular or macrovascular thrombosis on H&E staining or specific immunostaining for fibrin/fibrinogen deposition; specific staining for fibrosis within myocardium or other sites was negative. We next synthesized a R338LFIXopt expression cassette containing the LP1 promoter/enhancer/intron sequence being used in the ongoing clinical trial and demonstrated equivalent FIX activity from either promoter construct. We then established that the R338LFIXopt vector gives a predictable dose-response across a range of doses as low as 1x 10¹⁰ vg/kg I.V. and as high as 4 × 10¹² vg/kg I.V. Hemarthrosis is the most common bleeding complication in hemophilia and leads to chronic joint destruction. Bleeding was induced in the joint of FIX^−/− mice that had been transduced 4 weeks earlier with the R338LFIX vector. Joints were collected at 2 weeks after induced bleed and the bleeding-induced joint damage was graded using an established histologic score. I.V. R338LFIXopt vector pretreatment resulted in protection against joint degeneration in a dose-dependent fashion in this most relevant clinical scenario. These preclinical studies demonstrate a safety :efficacy profile to advance hemophilia gene therapy using the scAAV8.R338LFIXopt vector.