Abstract
Despite the misconception that adeno-associated virus (AAV) gene therapy vectors are nonintegrating, they can integrate into the host genome at a low but nonnegligible frequency, posing a theoretical risk of tumorigenesis. While AAV integration can trigger hepatocellular carcinoma in mice, no such association has been established in humans. None of the 10 cancer cases reported in AAV vector recipients so far has shown evidence that AAV integration drives tumorigenesis. However, the strength of the evidence from molecular analyses differed significantly across these cases. The scope and conclusiveness of causality assessments depended on sample quality and cross-validation using complementary analytical methods. For example, poor sample quality precluded a conclusive analysis in a case of a spinal cord tumor. Conversely, comprehensive analyses provided strong evidence that AAV integration was not the causative factor in a case of hepatocellular carcinoma. These findings underscore the need for standardization, global long-term follow-up, and careful communication of outcomes.