Transduction efficiency and vector biodistribution following prenatal vs neonatal AAV5 gene therapy for Hemophilia A in a sheep model

Abstract Hemophilia A (HA) patients treated with AAV-based gene therapy (GT) receive vector serotypes that preclinical studies indicated to be hepatotropic. However, the limited clinical biodistribution data available from the single FDA-approved AAV GT product have demonstrated prolonged vector presence in multiple tissues and bodily fluids, including semen. While it is generally accepted that episomal vector genomes are likely responsible for the sustained therapeutic plasma FVIII levels observed in GT clinical trials for HA, preclinical studies in multiple animal models and recent biopsy data from 5 patients who received the FDA-approved product have collectively documented the presence of hundreds to thousands of unique AAV genomic integrations in each recipient. Given the promising clinical outcomes thus far in adults and a CDC report that ~44% of newborns with HA have a bleed by 1 month of age, and nearly 10% receive factor within 24 hrs of birth, there would be clear benefits to administering GT early in life. There are, however, currently no HA GT clinical trials for newborns, infants, or even children, due largely to the lack of preclinical data in these populations. The overall goal of the present studies was to directly compare the on-target (liver) transduction efficiency, vector biodistribution, and incidence of genomic integration following prenatal and neonatal administration of an AAV5 vector encoding a liver codon-optimized fVIII transgene (lcoET3) in the translational sheep model to generate critical preclinical safety and efficacy data to ultimately pave the way for HA GT clinical trials shortly after, or even before, birth. Sheep fetuses at 60-65 days of gestation ( @18-20 wks in human) received 2x1012vg/kg of AAV5-lcoET3 by ultrasound-guided intraperitoneal injection (IUGT recipients; n=4); 3-day old neonatal lambs received the same vector at the same dose via the cephalic vein (NNGT recipients; n=3). Peripheral blood was collected at regular intervals to quantify levels of vector-derived FVIII (ET3) and to assess hematologic and hepatic function. IUGT recipients had normal CBCs and normal liver enzymes at all timepoints, while all NNGT recipients had normal CBCs but increased ALT levels by 7 days post-GT, which returned to normal by 30d. In addition, one NNGT recipient maintained elevated GGT levels up to 30 days post-GT, which subsequently returned to normal (60d) without treatment. ET3-specific ELISA demonstrated that 2 of 3 IUGT recipients had significant plasma levels of ET3 at multiple timepoints during the 1-year analysis – one recipient had 15.7+/-0.5 ng/mL and the other had 5.8+/-0.26 ng/mL at 1 yr of age (455d post-IUGT). In contrast, all NNGT recipients exhibited an initial spike of ET3 in plasma at 7d post-NNGT, but levels then declined to baseline by 30d post-GT. To assess vector distribution, animals were euthanized at 1 yr of age, and all major tissues were analyzed by dPCR with primers to the AAV-ITR to quantify vector copy number (VCN). These analyses revealed widespread vector biodistribution in both groups, with IUGT recipients exhibiting significantly higher VCNs in their liver, lung, spleen, and pancreas, and NNGT recipients having significantly higher VCNs in their heart, kidney, thoracic lymph node, duodenum, and testes. No significant difference in VCN was seen in thymus, mesenteric lymph node, or ovaries between these two groups. While the persistence of AAV genomes within these tissues for 1 yr following administration into fetal and neonatal recipients whose cells/tissues are rapidly cycling/expanding is suggestive of integration into the recipient genome, dPCR cannot distinguish between episomal and integrated AAV genomes. We are currently performing whole genome sequencing using an ultra-long read (>100kb) platform on DNA from the livers of IUGT and NNGT recipients to determine the frequency and location of AAV genomic integration, and to ascertain if the recipient age impacts whether the integrated AAV genomes are intact or fragmented. These data add to the preclinical body of knowledge regarding AAV vector biodistribution in the prenatal and neonatal setting, highlight the potential for enhancing liver transduction efficiency by delivering AAV in utero, and begin to define the potential for genotoxicity of AAV vectors in fetal and neonatal recipients, collectively helping to pave the way for safer and more effective treatments for HA.