Abstract
Integration of naturally occurring adeno-associated viruses (AAV; wild-type AAV [wtAAV]) and those used in gene therapy (recombinant AAV [rAAV]) into host genomic DNA has been documented for over two decades. Results from mouse and dog studies have raised concerns of insertional mutagenesis and clonal expansion following AAV exposure, particularly in the context of gene therapy. This study aimed to characterize the genomic location, abundance, and expansion of wtAAV and rAAV integrations in macaque and human genomes. Using an unbiased, next-generation sequencing-based approach, we identified the genome-wide integration loci in tissue samples (primarily liver) in 168 nonhuman primates (NHPs) and 85 humans naïve to rAAV exposure and 86 NHPs treated with rAAV in preclinical studies. Our results suggest that rAAV and wtAAV integrations exhibit similar, broad distribution patterns across species, with a higher frequency in genomic regions highly vulnerable to DNA damage or close to highly transcribed genes. rAAV exhibited a higher abundance of unique integration loci, whereas wtAAV integration loci were associated with greater clonal expansion. This expansive and detailed characterization of AAV integration in NHPs and humans provides key translational insights, with important implications for the safety of rAAV as a gene therapy vector.