Preclinical characterization of a recombinant adeno-associated virus type 1-pseudotyped vector demonstrates dose-dependent injection site inflammation and dissemination of vector genomes to distant sites

To translate the potential advantages of recombinant adeno-associated virus type 1 ( rAAV1) vectors into a clinical application for muscle-directed gene therapy for alpha(1)-antitrypsin ( AAT) deficiency, we performed safety studies in 170 C57BL/6 mice and 26 New Zealand White rabbits. A mouse toxicology study included 8 cohorts of 10 mice each ( 5 per sex). Mice were killed either 21 or 90 days after intramuscular injection of doses ranging up to 1 x 10(13) vector genomes ( VG), equivalent to 4 x 10(14) VG/kg. A mouse biodistribution study was performed in 5 cohorts of 10 mice, receiving intramuscular injections at the same doses; as well as in a lower dose cohort ( 3 x 10(8) VG; equivalent to 1.2 x 10(10) VG/kg); and in 4 other cohorts ( excluding the vehicle control) injected with identical doses intravenously. Finally, biodistribution was examined in rabbits, with serial collection of blood and semen, as well as terminal tissue collection. Two significant findings were present, both of which were dose dependent. First, inflammatory cell infiltrates were detected at the site of injection 21, 60, or 90 days after intramuscular injection of 1 x 10(13) VG. This was not associated with loss of transgene expression. Second, vector DNA sequences were detected in most animals, levels being highest with the highest doses and earliest time points. Vector DNA was also present in liver, spleen, kidneys, and a number of other organs, including the gonads of animals receiving the highest dose. Likewise, vector DNA was present in the semen of male rabbits at higher doses. The copy number of vector DNA in the blood and semen declined over time throughout the study. These two dose-dependent findings have served to guide to the design of a phase 1 human trial of rAAV1-AAT.