Solving Vector Yield Challenges Starts Long Before Manufacturing

By Erin Harris, Editor-In-Chief, Cell & Gene
Follow Me On Twitter @ErinHarris_1

One of the recurring themes throughout my conversations with cell and gene therapy leaders is that manufacturing challenges rarely begin in manufacturing. They often originate much earlier, during the biological decisions that shape vector design, tissue targeting, and platform strategy.

That reality became clear during our recent Cell & Gene Live, Solving Vector Yield Challenges in In Vivo Cell and Gene Therapy, where I was joined by Karen Kozarsky, PhD, President and CSO of Vector Biopartners, and Robert Bell, PhD, CSO at Ascidian Therapeutics. While vector yield is often discussed as a manufacturing problem, both experts emphasized that solving it requires developers to think about biology, process development, and commercialization simultaneously rather than sequentially.

Biology and Manufacturability Must Advance Together

For Dr. Kozarsky, the challenge is balancing therapeutic performance with the practical realities of producing vectors on a commercial scale. “We clearly need to find vectors, delivery vehicles that do what we want them to do to have sufficient efficacy,” she said. “At the same time, we also must balance it with the reality. Can we actually make these? Can we make them in sufficient quantities and at a price at which we can make this hopefully somewhat affordable.”

That balancing act begins with delivery biology. Developers must determine whether a vector can physically reach the target tissue, cross biological barriers, recognize the appropriate receptors, enter the cell, and ultimately deliver its genetic payload to the nucleus. Every step influences whether a therapy succeeds clinically and whether it can eventually become commercially viable.

Dr. Kozarsky noted that many organizations still approach efficacy first and manufacturability second. While understandable, she believes that strategy often creates unnecessary delays. “It really is a question of trying to deal with both issues of efficacy and manufacturability more or less simultaneously,” she explained.

Dr. Bell echoed that perspective from the standpoint of a therapeutic developer. At Ascidian Therapeutics, every program begins with the disease itself, followed by careful evaluation of delivery options, existing clinical experience, and manufacturing feasibility.

“We really start with the problem at hand,” Bell said. “We think about the disease, the target tissue, the unmet need, how much efficacy is needed, and then we look to see what tools exist today with clinical precedent and what the manufacturability looks like.”

Rather than abandoning promising programs when challenges arise, Dr. Bell said his team evaluates where additional investment can reduce risk while keeping development on track.

Breaking Down the Silos

Both experts repeatedly returned to a common obstacle facing the industry. Discovery scientists and manufacturing teams have historically worked independently, often creating challenges later in development.

Dr. Bell believes the industry is beginning to change that mindset. “It’s really important to try to break down any artificial barriers between discovery biology and the manufacturing colleagues that are really critical,” he said.

When biology and manufacturing teams work together early, organizations can identify issues involving vector production, purification, analytical testing, and scalability before they become expensive late-stage problems.

Dr. Kozarsky agreed that understanding the biology inside the production cell remains one of the industry’s greatest opportunities for improving vector yield. Because AAV production depends on living cells, manufacturing efficiency is influenced by complex cellular processes that researchers are still learning to optimize.

“My hope is that additional work on what’s happening inside the cell and how we can optimize it may actually lead to some improvements downstream,” she said.

Dr. Bell added another biological consideration that often receives less attention.

“AAV production is somewhat inherently cytotoxic to the cells that it’s in,” he explained. Better understanding how production cells respond metabolically during vector assembly could ultimately improve manufacturing yields by helping cells tolerate the stress associated with viral production.

Better Screening Creates Better Candidates

Capsid selection represents another area where early decisions can dramatically influence downstream success. Dr. Kozarsky emphasized that developers must carefully consider how they evaluate new vectors because laboratory models do not always predict clinical performance. “People always say you get what you screen for,” she said.

While in vitro testing remains valuable, cells grown in culture behave differently than cells in living tissues. Even more importantly, many traditional screening approaches measure DNA delivery rather than successful transgene expression, potentially overlooking critical biological processes required for effective therapy.

Instead, Dr. Kozarsky encouraged developers to build screening systems that more closely resemble human biology while evaluating complete transduction rather than simply measuring vector uptake.

Dr. Bell suggested adding another important filter much earlier in development. Before advancing promising capsids into animal studies, developers should determine whether those vectors can be manufactured using commercially relevant production methods.

“If you can de-risk and weed out capsids that are already predicted to have manufacturing challenges before you then take them into the in vivo system, I think we’ll see some more successes,” he said.

That philosophy perfectly captured one of the strongest messages from our discussion. In vivo gene therapy development no longer allows biology and manufacturing to operate independently. Every early scientific decision has downstream implications for scalability, regulatory readiness, and ultimately patient access.

Smart Tradeoffs Drive Smarter Development

One of the most interesting parts of our discussion centered on the constant tradeoffs developers face. The most biologically elegant solution is not always the one that makes the most sense for clinical development.

Dr. Kozarsky shared an example from her experience developing therapies for spinal cord disorders. Rather than pursuing an entirely new capsid, her team leveraged the existing clinical experience behind AAV9 while improving the delivery method to achieve broader distribution throughout the spinal cord.

“Could there have been another capsid out there that would have hit the entire spinal cord with straightforward intrathecal delivery? It’s possible,” she said. “But it was going to take quite some time to identify those.”

Instead, the team focused on improving delivery while benefiting from the extensive safety and manufacturing knowledge already available for an established capsid.

Dr. Bell noted that similar tradeoffs extend beyond the capsid itself. Payload design, promoters, enhancers, and other regulatory elements can improve biological specificity but also complicate manufacturing.

“You really have to be partnering early on with your manufacturing colleagues or partners that you’re going to work with to understand where there may be a required focus for optimization,” he said. That collaboration becomes especially important as therapies move toward commercialization.

Early Means Earlier than Most Companies Think

Throughout our discussion, I repeatedly returned to one question that surfaces in nearly every Cell & Gene Live conversation. What does “early” really mean?

For Dr. Bell, it begins as soon as organizations can evaluate whether a program can be manufactured reproducibly, characterized with robust analytical assays, and eventually produced at a commercially realistic cost.

“If the answer to any of those is we’re not sure, that doesn’t necessarily kill the idea,” he said. “But it would definitely change how aggressively you would want to resource it and what experiments you can do early on to de-risk or start to answer some of those questions.”

Dr. Kozarsky added that developers should engage manufacturing partners well before final process development begins. Even small-scale manufacturing runs using the intended therapeutic payload can reveal challenges that might otherwise emerge much later.

She also encouraged companies to involve CDMOs while lead candidates are still being selected so manufacturing considerations become part of the development strategy instead of an afterthought.

Collaboration is Becoming a Competitive Advantage

Later in the discussion, Vishruth Gowda from Fujifilm Biotechnologies joined us to provide the CDMO perspective. His observations reinforced many of the themes that Drs. Bell and Kozarsky had already highlighted.

“What really stood out to me was how we are talking about robust manufacturability even when we are talking about early development,” Gowda said. He noted that only a few years ago many companies prioritized reaching first in human studies as quickly as possible. Today, organizations increasingly recognize that every development decision should support the eventual biologics license application and commercial launch.

Gowda also emphasized the importance of integrating manufacturing expertise directly into development teams. “Breaking those boundaries down and getting manufacturing involved early has become one of the most effective ways to create scalable, robust production processes,” he said.

That advice reflects a broader shift taking place across the industry. Manufacturing is no longer simply an operational function that begins after discovery. It has become a strategic capability that influences platform design, vector selection, process development, regulatory strategy, and long-term commercial success.

Looking Beyond Individual Programs

As our conversation concluded, I asked both panelists what they would change about how the industry approaches vector development.

Dr. Bell encouraged developers to think beyond building the perfect vector for a single therapy. “I’d really like to see more collaboration and sharing of tools and reagents that can help develop more medicines instead of it being done in a more bespoke manner,” he said.

Dr. Kozarsky agreed and pointed to another challenge that is often underestimated. “I think the one thing people don’t always understand is how much time it really takes to identify a new capsid, characterize it, test it, and develop the manufacturing process around it,” she said.

That reality reinforces why every decision surrounding vector design should carefully balance innovation with practicality.

One of my biggest takeaways from this discussion is that improving vector yield is no longer simply about making more material. It is about making better decisions earlier. The organizations that successfully connect biology, manufacturing, and commercialization from the very beginning will be best positioned to bring the next generation of in vivo cell and gene therapies to patients.

If you would like to hear the complete discussion, including additional insights on capsid engineering, manufacturability, platform strategy, and audience questions, you can watch the full-length Cell & Gene Live here.