Innovation in AAV: Breaking through bottlenecks in yield, safety, and cost
At this year’s European Society of Gene & Cell Therapy (ESGCT) meeting, one theme resonated across sessions: adeno-associated viruses (AAVs) are entering a new era of innovation. The field is moving beyond solving manufacturing bottlenecks to designing viral vectors that are more efficient, safer, and ultimately cheaper to deliver at scale.
AAV remains the gold standard for gene delivery due to its safety profile, tissue tropism, and ability to achieve long-term expression. Yet it is limited by a cargo capacity of approximately 4.7 kilobases. To compensate, developers often deliver truncated or split versions of genes, or administer large vector doses to achieve sufficient expression. This creates two compounding problems: increased manufacturing burden and heightened risk of immune responses.
Solving these challenges requires progress on two fronts: boosting yield through better manufacturing and engineering capsids and payloads that require lower doses.
1. Making AAV production more efficient
Even as gene therapy scales toward commercialization, AAV manufacturing remains one of the costliest and most technically complex steps. Improvements in process design are helping to change that.
Recent work by groups such as Oxford Biomedica, Lonza, and PackGene highlights advances in suspension-based production and optimized plasmid systems that are increasing yield while reducing the proportion of empty capsids and impurities. For example, Oxford Biomedica’s dual-plasmid transfection system streamlines AAV assembly and has demonstrated titers with over 90% full capsids, while Lonza’s stable producer cell line platform integrates the required helper and replication sequences directly into the cell genome, enabling large-scale production without repeated transfections.
Many biotech and pharma companies are also developing innovative in-house manufacturing capabilities. Such platforms are gradually transforming AAV production to allow higher vector yield per batch. This will ultimately lead to a lower cost per dose, which is a necessary step toward making gene therapy commercially viable.
2. Engineering vectors to work smarter
In parallel, researchers are working on techniques to reduce the amount of vector required in the first place. This is where capsid engineering, dual-delivery strategies, and mini-gene constructs are changing the game.
By engineering capsids to better target specific tissues, scientists can achieve therapeutic effects with lower doses, mitigating both toxicity and cost. For example, MyoAAVs have been designed to preferentially transduce muscle tissue in models of Duchenne muscular dystrophy (DMD), allowing for lower systemic exposure. Sarepta Therapeutics entered a deal with The Broad Institute for the generation of next-generation MyoAAV capsids for their neuromuscular programs.
Similarly, researchers highlighted several engineered capsids with enhanced tropism for the central nervous system, which was previously challenging due to the low penetrance of the blood-brain-barrier. In a speaker session at ESGCT, Shape Therapeutics announced that their next-generation AAV5 capsid, SHP-DB1, efficiently targeted the substantia nigra in non-human primates.
Parallel innovations in payload design are also expanding what AAV can carry. Mini-gene and micro-dystrophin constructs have enabled functional expression for DMD. Several biotech companies, such as Genethon and REGENXBIO, are running trials for AAV-based delivery of microdystrophin in patients with DMD. Early phase results presented at ESGCT by representatives from both sponsors reveal therapeutic efficacy and high tolerability/safety profiles.
Another approach to overcoming AAV’s cargo limit is the use of split-intein systems for large genes. The technique was validated for use with large dystrophin constructs by Jeffrey Chamberlain’s lab. Two AAV vectors each deliver half of the dystrophin protein, which then self-assembles inside the cell through protein trans-splicing. The strategy has since been adopted commercially to expand AAV’s therapeutic reach. Dual-vector systems are also being used to reconstitute large transgenes at the DNA level through recombination within the cell nucleus.
AAVantgarde Bio is actively applying both methods in its lead programs, using a dual hybrid platform for Usher syndrome type 1B (USH1B) and split inteins for Stargardt disease. Early data from AAVantgarde’s LUCE trial for USH1B were also presented at ESGCT and showed positive safety profiles and functional outcomes.
3. The case for collaboration and data sharing
One recurring message at ESGCT was that no single company can solve AAV’s efficiency problem in isolation. The field’s progress increasingly depends on shared learning, whether through open data on capsid performance, improved analytics for vector quality, or standardized assays for genome integrity. A more transparent ecosystem could accelerate improvements across both manufacturing and design, shortening the path from discovery to clinic.
The evolution of AAV is a story of convergence. Manufacturing advances are making it possible to produce high-quality vectors more efficiently, while engineering breakthroughs broaden AAV’s therapeutic scope and versatility while directly impacting cost and scalability. Lower doses mean smaller manufacturing runs, less vector per patient, and reduced risk of immune-related complications. Together, these efforts are redefining what scalable, accessible gene therapy looks like.
To explore additional barriers and breakthroughs shaping the AAV field today, access our whitepaper AAV gene therapy trends: Navigating the boom in viral vectors.