Background Hemophilia A (HA) and hemophilia B (HB) are inherited bleeding disorders resulting from deficiency of coagulation factors VIII (FVIII) or factor IX (FIX) activity, respectively. HA and HB both occur naturally in dogs. The bleeding phenotype of severe canine HA and HB is analogous to the human patient experience with frequent spontaneous and trauma-induced bleeding. Previous studies of dogs from HA and HB colonies at research institutes have been highly informative about the safety and efficacy of adeno-associated virus (AAV) liver-directed gene therapy for HA and HB. However, these animals are housed in highly controlled environments; in contrast, pet dogs with HA or HB are exposed to real-world challenges that likely affect their bleeding phenotype and the efficacy of novel therapeutics. In this regard, privately owned dogs may better recapitulate the human patient experience. Aims To determine the real-world efficacy of liver-directed AAV gene therapy for HA/HB in privately owned canines with and without inhibitors. Methods Eleven dogs with severe HA and one dog with severe HB were treated with liver-directed AAV gene therapy with canine (c) FVIII or cFIX, as detailed in Table 1. Pre-existing neutralizing antibodies to AAV8 were detected using a luciferase-based transduction inhibition assay. Pre- and post-gene therapy annualized bleeding rates were calculated by cumulative bleeding events divided by the age at gene therapy (pre) or duration of follow-up (post). FVIII activity was determined using a chromogenic assay and FIX activity using a one-stage assay. Results All dogs had AAV8 neutralizing antibody titers ≤ 1:5 at enrollment. The HA dogs (10 males and 1 female) were treated with either dual chain vectors of cFVIII-SQ (n = 2, total dose of 5 x 1013 vg/kg) or single-chain vectors encapsidating high-expression cFVIII variants: cFVIII-ΔF (n = 2, 6x1012 vg/kg) or cFVIII-ΔF/V3 (n = 7, 5.4-9 x 1012 vg/kg). F8 genetic testing revealed intron 22 inversions (n=7), small deletion (n=2), missense (n=1) and complex mutation (n=1) all compatible with severe HA. Of note, one HA dog (PC3) had a pre-existing FVIII inhibitor which was eradicated following gene therapy. In all dogs with available data (n=10), FVIII activity is > 1% with 5 in the moderate hemophilia range and 4 in the mild range with a median FVIII activity of 3.2% (range 1.1-13.8%) after a median duration of 4.1 years (range 2.6-8.9 years). In HA dogs, the ABR post-gene therapy declined by an average of 93%. The sole HB dog carried two missense F9 mutations and was treated with wildtype cFIX at 3 x 1012 vg/kg and has been followed for 4.5 years with FIX activity of 5.7% and has had no bleeding episodes since gene therapy. Three animals have died during follow up. The previously inhibitor positive dog (PC3) died approximately 2 years after gene therapy from an unrelated cause. An HA dog (PC1) treated with dual chain vector with expression in the ~2% normal range died from a bleeding episode at 8.8 years after gene therapy. Notably, many of his bleeds were associated with real-world activities including having his legs being pulled by a toddler. A third HA dog (PC9) was euthanized 3.5 years post-gene therapy after being diagnosed with widely disseminated cancer; the details of the vector integration analysis of his tumor will be reported elsewhere. Conclusion AAV gene therapy was successful at substantially ameliorating the bleeding phenotype in both male and females dogs with > 1 year of follow-up. However, some dogs continued to have bleeding, especially animals with <5% normal cFVIII activity. One animal developed cancer after gene therapy; there were no other safety concerns. As we have described previously, liver-directed AAV gene therapy also induced immune tolerance in an HA dog with a pre-existing inhibitors. We show that AAV liver-directed gene therapy is safe and efficacious in a real-world experience of privately owned dogs with severe HA and HB. Figure 1View largeDownload PPTFigure 1View largeDownload PPT Close modal