02172nas a2200265   4500000000100000000000100001008004100002260001500043653002100058653001900079653003100098653001400129653002000143653002400163653001300187100002300200700002000223700001600243700002300259245008600282300001100368490000800379520150500387022001401892       2026                            d  c2026-03-0510aAAV gene therapy10aImmunogenicity10amicrophysiological systems10aorganoids10aorgans-on-chips10aPreclinical testing10atoxicity1 aPatrizia Tornabene1 aEstelle Berreur1 aAdrian Roth1 aMagdalena Kasendra00aMicrophysiological systems to accelerate and de-risk AAV gene therapy development  a1158410 v2333 aWhile adeno-associated viral (AAV) vectors are the leading platform in gene therapy - with eight regulatory-approved products spanning multiple disease areas - their clinical performance remains only partially predictable. A major challenge lies in the discrepancy between preclinical model predictions and clinical outcomes, particularly with respect to adverse events such as liver toxicity, immune responses, and tissue tropism. Microphysiological systems (MPS), including organoids and organs-on-chips, offer a promising and powerful approach to address some of these challenges. As emerging "New Approach Methodologies (NAMs)" they provide a human-relevant context to evaluate AAV vectors for tropism, safety, and efficacy, thereby helping to de-risk and accelerate their development. This trajectory is further reinforced by recent initiatives such as the FDA Modernization Act 2.0 and the NIH's Office of Research Innovation, Validation, and Application (ORIVA), both of which promote the adoption and qualification of human-based models in biomedical research. This review summarizes the latest advances in the application of MPS technologies to AAV evaluation, outlines their current limitations and proposes a framework for their integration into AAV development pipeline. Ultimately, MPS could serve as a valuable "human-relevant filter" to prioritize vectors with the highest likelihood of clinical success, ensuring that only the safest and most effective candidates advance to patients.  a1872-8294