A gene therapy can lose part of its dose inside the tube meant to deliver it. In work presented in May 2026, Ascend Advanced Therapies examined a particularly awkward version of this problem: viral vectors sticking to the inside of long, thin cannulas used to reach the brain. The science had made it all the way to the delivery device. Some of the product still needed help getting through.
That small indignity is a useful introduction to Ascend. The company works in the space between a promising biological idea and something that can be made, measured and supplied repeatedly. Here, a clever discovery keeps acquiring practical conditions. What is in the batch? Can the process scale? Will the dose survive its journey?
- The work: development, testing and manufacturing for advanced therapy companies.
- The bet: turn experienced biotech manufacturing teams into a shared service.
- The detail to watch: two-plasmid AAV production, with quality measured alongside yield.
The factory nobody wanted to rebuild
Consider the manufacturing department inside a therapy developer. Its scientists learn the peculiarities of a product, develop assays and work out how to produce usable material. That knowledge is valuable. Keeping an entire operation occupied, however, depends on the needs and fortunes of a limited drug pipeline.
Ascend emerged from that tension. Founded by Monograph Capital, it acquired capabilities from Freeline Therapeutics and launched publicly in March 2023. Tim Funnell was its founding CEO; manufacturing veteran Mike Stella succeeded him. The company began with people who had worked on therapeutic programs, carrying their practical experience into a service business.
“When we struggled to find our own CDMOs, we became one.”Ascend, on its development roots
The abbreviation means contract development and manufacturing organization. A customer hires the company to help develop a process, test material or manufacture supplies. Ascend offers individual assignments and broader programs. Its customers are therapy developers buying technical capacity and judgment. Patients are the eventual beneficiaries, rather than the people placing the orders.
The April 2024 Beacon Therapeutics deal made the model concrete. Ascend acquired Beacon’s manufacturing team and Alachua, Florida facility. A long-term agreement kept Ascend making Beacon’s products for clinical and commercial use. Beacon could concentrate on development while retaining a manufacturing relationship. Ascend gained an operating facility, experienced colleagues and an ongoing customer.
Later in 2024, a combination with Advanced BioScience Laboratories in Rockville broadened the offer. ABL added experience across other viral modalities and US fill-finish capacity. EW Healthcare Partners also committed capital. The sequence reveals Ascend’s preferred building material: operations with people already inside them.
Two plasmids. Several business decisions.
Inside that network sits a more specific proposition. Adeno-associated virus, or AAV, is used as a delivery vehicle for gene therapies. Manufacturing it involves supplying production cells with genetic instructions. A conventional approach uses three plasmids, circular pieces of DNA that carry different parts of those instructions.
Ascend’s proprietary EpyQ system uses two: one custom plasmid and one supplied off the shelf. The company argues that this reduces complexity and input costs while improving productivity and limiting impurities. Those are product claims to evaluate against a customer’s process. Counting plasmids alone cannot establish whether a particular therapy will be easier to manufacture.
Two inputs. A process still needs product-specific testing.
The commercial terms are revealing. Customers manufacturing at Ascend in Munich or Alachua do not need a separate EpyQ license. External manufacturers pay an annual license fee and obtain the helper/rep plasmid through supply agreements. Published terms specify no milestones or royalties. Ascend can therefore sell manufacturing work and make its technology available beyond its own facilities.
At public launch, cumulative funding stood at $132.5 million, including a Series A led by Abingworth and Petrichor. That figure describes money raised to build the business. It is not a customer’s manufacturing bill. Buyers still need a proposal matched to their product, development stage, analytical needs and production requirements.
Count what is inside
The tempting manufacturing metric is yield: how much came out? In gene therapy, that question immediately needs companions. How much contains the intended genetic material? What unwanted DNA is present? Does the vector perform its intended biological function? An impressive count can conceal a less useful batch.
Ascend advertises more than 60 characterization and release assays. Its long-read sequencing service uses Oxford Nanopore technology to examine vector genomes and impurities, including shortened sequences and DNA carried over from production. Munich’s GMP quality-control certification, announced in June 2025, added a regulated testing capability to its development work.
In November 2025, Ascend reported that two proprietary enhancers increased AAV yields by up to threefold under tested conditions. The company said testing covered its two-plasmid platform and a commercial three-plasmid system, without adverse effects on measured quality parameters. “Up to” matters. The result invites process-specific comparison; it does not promise every customer a threefold improvement.
The same discipline applies to choosing a supplier. A developer can bring an existing process for evaluation, buy a particular analytical service or discuss a more integrated program. The sensible question is which uncertainty the engagement will resolve. A production slot answers a capacity question. An assay may answer a different question entirely, before a team commits to making a much larger batch.
The last few inches
Its 2026 research keeps returning to practical comparisons. A 10-liter AAV9 study evaluated orbital shaking, rocking and stirred-tank bioreactors, measuring quality and potency alongside yield. The team favored stirred tanks for large-scale production. The finding belongs to that experimental setup; extending it requires evidence at the next scale.
Then there is the cannula. The formulation team found that low concentrations of commonly used surfactants could improve vector recovery during delivery. For a developer, the transferable habit is straightforward: test the product with the equipment it will encounter. A manufacturing specification and a delivery device have to get along.
Ascend competes for work a developer might keep internally or award to another viral-vector manufacturer. Its case rests on connected services, proprietary technology and experience inherited from development teams. Outsourcing still requires a suitable process, clear responsibilities and convincing data. The factory becomes shareable. The obligation to understand the product stays with everyone involved.