This Biotech Says It Could Cut Gene Therapy Costs to the Low Thousands by Redesigning the Factory's Recipe

Gene therapy can run into hundreds of thousands of dollars per dose, and most of that is the cost of growing the virus. Most approved gene therapies use a crippled virus called AAV to carry a corrective gene into a patient's cells. The standard process of growing that virus in a 500-liter steel tank called a bioreactor is so inefficient that a single dose takes up a large share of the bill. A Massachusetts biotech named Affinia Therapeutics says the fix is in the factory's recipe, not the virus, and that the change is simple enough to copy.

The lever is the DNA recipes, called plasmids, that tell producer cells how to assemble the virus. Affinia says it ran a high-throughput screen of commercial cell lines and the chemicals used to deliver those recipes into the cells (a step called transient transfection), picked the best combination, and rebuilt the plasmids around that match. Plasmid design, the company argues, is the most impactful knob in the process, more than the bioreactor, the cell line, or the transfection reagent on their own. The result, Affinia claims, is a ten-fold increase in vector genome yield, up to a four-fold increase in the share of fully assembled viral capsids, the protein shells that wrap the gene cargo, and a 740% jump in R&D-scale titers after 2024, with the company now reporting yields above 5×10¹⁵ vg/L at lab scale and above 3×10¹⁵ vg/L at 50-liter bioreactor scale.

The current industry baseline is roughly 3×10¹⁴ vg/L with about 25% recovery at harvest, according to a 2025 industry analysis. A ten-fold jump from that baseline would put Affinia well above it, though the company has not published a head-to-head benchmark against an independently run process, and the figure sits on Affinia's own internal comparisons.

That headroom matters because the industry has spent a decade trying to make AAV cheaper without much success. The viral shell that does the delivering is, in production, also a finicky and inefficient product. Cells that are supposed to assemble the virus often produce empty or partially built capsids that have to be thrown away. The yield problem is not new, and neither are the proposed fixes. A separate route, growing AAV inside herpes-derived helper virus systems, has reported five-to-ten-fold yield gains over the standard polyethyleneimine (PEI) transient transfection process, per a 2022 head-to-head study cited in a 2026 industry landscape. The HSV route starts from a different baseline, which makes the two "10x" numbers hard to compare directly.

Affinia is in licensing discussions with roughly half a dozen drugmakers to adopt the approach, Affinia executives Rob May and Matt Edwards told GEN, and the company presented updated manufacturing and capsid-engineering data on its lead program, AFTX-201 for a hereditary heart condition, at the American Society of Gene and Cell Therapy meeting in late April 2026. Affinia first presented the program, including data on capsids engineered to cross the blood-brain barrier, at ASGCT 2025.

The ceiling on what any of this changes for patients is real. Affinia's target, AAV production in the low single-digit thousands of dollars per patient dose against current costs that can reach hundreds of thousands, applies only to the cost of making the drug. What a hospital or insurer pays is a different layer. Clinical trials, regulatory milestones, hospital markups, and the reimbursement games that determine what insurers and health systems actually pay sit on top of manufacturing. None of those move when the bioreactor shrinks.

The test is whether any of the half-dozen companies Affinia says it is in talks with sign a license and publish their own yield numbers. The plasmid recipe is specific enough to copy; the third-party benchmark has not been run.