Key considerations for scalable retroviral vector production—with a focus on lentivirus

What are you currently working on?

AC: I’m currently focused on advancing both process and product development within Donaldson’s Life Sciences businesses, particularly around our advanced manufacturing technologies. A key part of my work involves supporting emerging therapeutic modalities like viral vectors and nucleic acids. I also oversee the development and use of advanced cost of goods modeling capacities for bioprocessing. My background is in chemical and biochemical engineering, and I’ve held a mix of technical and commercial roles across academia and industry, which gives me a broad perspective on how to bring innovative technologies to market.

What unique benefits—and drawbacks—do retroviruses offer for cell and gene therapy applications?

AC: Two members of the retrovirus family are of particular interest for advanced therapies: lentiviruses (LVs) and to a lesser extent, gamma retroviruses (gRVs). Retroviruses are mainly known and used for their ability to transduce dividing and non-dividing cells, which is particularly useful in genetically-modified cell therapies, such as chimeric-antigen receptor T cell therapies (CAR-T), sometimes referred to as ex vivo gene therapies (GTs).

As of today, ten market-approved ex vivo gene therapies use retroviruses—eight of which use lentiviruses, and two use gamma-retroviruses. They are also currently investigated in hundreds of clinical trials. It’s important to note that although the market-approved retrovirus therapies are all ex vivo therapies, there’s a growing interest in using retroviruses, mainly LV in this case, for in vivo therapeutics. One recent example is the platform developed by EsoBiotec, a Belgium-based company developing a therapy using lentiviruses for in vivo use, that was recently acquired by AstraZeneca.

When it comes to drawbacks, LVs are originally derived from HIV, so there is a level of biosafety quality control that needs to be ensured. The LV used in therapies is replication-deficient, but this needs to be tested with every batch and justifies extensive quality control. On the other hand, gRVs pose an inherent higher risk of insertional mutagenesis, which has been linked to an increased chance of leukemia in clinical trials in the early 2000s. For this reason, LVs are now generally favored.

Beyond safety, a key challenge with these viruses is simply that they are hard to manufacture, especially in a cost-effective manner. This means gene therapy drug products are typically associated with high price tags, which impact patient accessibility and in some cases even the commercial viability of these life-saving medicines.

Could you frame for us the current challenges in the production of LV and gRV vectors for applications in cell therapy, as you see them?

AC: It’s important to remember that in almost all cases LVs and gRVs are produced in mammalian cells via either transient transfection, or sometimes stable production in more modern processes. One of the challenges of producing these vectors in a cost-effective manner is that depending on the target indication, the quantity of viral vector required can vary up to a million times from relatively low-dose ex vivo gene therapies to high dose in vivo therapies, with an additional impact of disease prevalence and patient population size.

This breadth makes it challenging to produce retroviruses cost-effectively with standard manufacturing technologies. It means that there is not simply a one-size-fits-all system for manufacturing, and flexibility is needed. For instance, an ex vivo GT or an in vivo GT for ultra-rare disease may be cost-effectively produced with a bench-top production system, while on the other end of the spectrum, a highly automated, industrialized-size manufacturing platform will be needed for large indication in vivo gene therapies.

In a typical process, the cells need to be grown to a target density upon which production will start. If using transfection, the most common method today, this will be initiated with the addition of plasmid DNA and a transfection agent. This is a delicate step as the transfection mix is both time and shear-sensitive, and can lead to cell toxicity. Stable producer cell lines don’t require this step because they already possess the inherent genetic information to produce the virus, but it takes time to develop a robust cell line. As of today the titers obtained from producer cells are still lower than in transfection, which means that this option hasn’t yet been widely adopted—although I anticipate it will be in the future as technology improves.

A key consideration for both LV and gRV is the use of serum as a cell culture supplement. By definition, serum is chemically undefined. It’s an animal-sourced reagent, and requires complex procurement and quality control to use. When the vectors are used as an ingredient to cell transduction, such as for CAR-T production, the purity requirements aren’t as stringent as for direct injection. However, as therapies evolve and move into more in vivo uses, there will be increased emphasis and scrutiny on process-related impurities. Cost and procurement considerations aside, this is likely to keep pushing developers away from using serum, and the use of suspension allows this in some instances.

Finally, another important challenge is identifying the appropriate manufacturing platform. Developers will usually start with developing processes at the lab scale using adherent cells because this process is simpler to adopt at the small scale. Upon successful transition to the next stage, manufacturing will need to be considered with commercial needs in mind. Typically, the decision developers will need to make is whether to stay with an adherent system or to move to suspension if the target production scale justifies it.

Can you expand on the relative pros and cons of adherent versus suspension cell culture approaches?

AC: For suspension-adapted cells, stirred-tank bioreactors (STRs) have long been the gold standard. They have been used for decades and have a proven ability to meet the high-demand requirements for many biologics—starting with antibodies and now moving into the field of viral vector production. They scale well and they offer flexible capacity for many applications. They also provide the ability to produce in serum-free conditions, but there are cases for which their basic design can be a hurdle.

For example, for high cell density perfusion cultures, which is a growing trend in viral vector production, the set up can be complex to run and also requires a cell retention device, which is a somewhat difficult process to both develop and operate. Another hurdle is the simple fact that at large scale, you will be producing a very large volume of harvest material of relatively dilute product, which is also highly heterogeneous in nature and will contain cells, often a lot of cell debris, host cell impurities such as DNA and proteins, and the product. These components need to be removed, which is both time-consuming and costly, and can lead to reduced downstream processing efficiency and yield. Finally, STRs run by agitating cells in liquid medium and as such can generate high hydrodynamic shear, especially at the impeller tip. Cells and product are constantly exposed to this because they are freely suspended and in contact with these impellers. This might not always be an issue depending on application, but LVs and gRVs are notoriously shear-sensitive, and some cell types are too, so this can result in product loss. Bubble damage linked to aggressive oxygen sparging can also be a recurring challenge.

As for adherent cell lines, there are two broad categories of manufacturing technologies. We have traditional flatware, such as multi-layer culture dishes or single-layer culture dishes at the lab scale, or we have what we call fixed-bed reactors. Traditional flatware is perfectly suitable for R&D and is widely used across labs and sometimes also for very small-scale production. However, it doesn’t offer the same level of process control as a bioreactor would. Additionally, flatware approaches can scale out, but not up, which leads to unsustainable costs for commercial production.

Fixed-bed bioreactors on the other hand, provide a controlled environment in the same way that a STR does, with pH, temperature, and DO control, as well as GMP-compatible data recording and access control to monitor the batch. They also can be used all the way from lab to commercial scale.

How can structured fixed-bed bioreactors help to address these issues?

AC: First generation fixed-bed bioreactors based on randomly packed matrix have helped address some limitations of adherent cell-based production, but face scalability and reproducibility issues due to variable compaction levels and lack of intermediate scale options. Next-gen fixed-bed reactors, such as the scale-X™ platform from Univercells Technologies (a Donaldson Life Sciences business), have been developed as a result of these limitations.

The scale-X bioreactor can be more accurately described as a structured fixed-bed. Indeed, the surface provided for cell growth is homogeneous throughout the reactor, and the packing density of the material that is designed for cell anchorage is the same throughout the whole vessel. This provides predictable and consistent cell growth, and therefore production of the viral vector.

A structured fixed-bed reactor is characterized by its surface area for cell growth, the same way flatware would be. This can also be equated to the volume of an STR. The smallest scale-X bioreactor, the scale-X nexo, is 0.5 m² which is roughly equivalent to a 1 L STR. For mid-scale production the scale-X carbo comes both in 10 and 30 m² and can produce the equivalent of up to a 200 L STR. At the top end of the scale we have the scale-X nitro, which is either 200 or 600 m² and can yield throughput equivalent to 2,000 L or more.

There are a number of advantages to choosing a structured fixed-bed platform. Firstly, the cells are retained within the bed itself. This means that processes which could benefit from perfusion or intermittent harvest—such as for retroviruses which have a biological capacity to ‘bud’ from cells—don’t require an external cell retention device, unlike with an STR. You can perform intermittent harvest, which opens process design possibilities such as the ability to collect fractions of the harvest at fixed time points along the process. You also have opportunities to collect them in a way that protects the virus, such as to cool them right after harvest, for instance, to maintain product integrity.

Secondly, as the cells are protected within the fixed-bed, they are also protected from shear linked to impeller and sparging, which helps support high viabilities, and, typically, increased specific viral productivities compared to alternative technologies.

Finally, thanks to the structured fixed-bed design, reproducible cell entrapment, growth and productivities can be achieved linearly throughout scale-up, from 0.5 m² to 600 m². This is a major advantage for process development and scale-up, especially in a time- and cost-constrained setting.

…and what about the impact on downstream processing?

AC: The ability to easily produce in a perfusion setup offers advance for further processing, as the product can be directly stored in more appropriate conditions—for example at lower temperature. This helps improve titers and product quality, facilitating downstream operations.

Interestingly, it has also been shown that a fixed-bed approach reduces contaminants such as debris, proteins and host cell DNA in the harvest. This is the result of a simple physical effect wherein the fixed bed acts a bit like a filter. Cells will anchor and attach as they grow, but even as they lyse a large proportion of the contaminants remain trapped within the fixed bed. The harvest collected will therefore be cleaner when compared to an STR, where all of the debris and impurities are resuspended along with the product. This results in a significant improvement in the downstream processability of the broth.

The filter area needed to clarify the broth prior to the capture step will be smaller, and also the efficiency of the following chromatographic steps, if used, will be enhanced thanks to the lower contaminant burden.

What advice would you give to developers looking to evaluate which production process will be most suitable for their own application?

AC: Evaluation should always start with estimating your commercial requirements in terms of viral vector quantities. This will define the scale at which production will need to take place once the process is developed and safety and efficacy has been demonstrated at the clinical stage. Next, there are a number of strategic considerations to carefully evaluate, including in-house production vs outsourcing to a CMDO, the type of expression system (transient transfection versus stable producer), the size and breadth of the drug pipeline; and linked to that the eventual choice to go for a platform approach.

Manufacturing considerations, including scale-up strategy, should be assessed as early as possible, as this will have a significant impact on development speed and on the final manufacturing cost of goods. In turn this will impact the final return on investment and margin of producing the drug product. A poorly designed process can significantly impact production costs and could even delay market entry.

The flexibility of your manufacturing solution should also be considered. For example, the scale-X carbo enables the production of manufacturing batches releasable under GMP guidelines using a compact benchtop system. This is cost-effective as it contributes to reducing large capital and operating expenses or commitments during the clinical phase, and can delay them to a stage where the risk of failure is lower. Avoiding spending too much upfront, and instead only doing so when your chances of success have increased, is an efficient use of your capital.

Particularly for retrovirus products, the ability to increase product yields by implementing semi or continuous product harvest strategies by design, without a cell retention device, is highly cost-effective. Even above the cost of plasmid DNA or transfection reagents, the most effective way to reduce the cost of goods per dose is to increase the total amount of product that can be produced from the reactor. The higher your productivity, the more significant the impact on reducing cost.

To conclude, it all comes down to scalability, speed, and cost. Choosing a manufacturing platform that supports low-footprint, low-cost production at high yields while supporting rapid scale-up can be instrumental in achieving commercial success and getting these life-saving therapies to those in need. At Donaldson Life Sciences, our Univercells Technologies business is committed to making this a reality.

Biography

Alex Chatel is a Senior Product Manager at Donaldson Life Sciences, based in Lyon, France, where he leads product initiatives in viral vectors, nucleic acids, and vaccines. He also supports Univercells Technologies, a Donaldson Life Sciences business located in Brussels, Belgium, where Alex developed and launched the scale-X technology prior to his current role He has also held positions as an Enterprise Fellow in technology transfer at University College London, London, UK and as a Research Engineer at GlaxoSmithKline in Stevenage, UK. Alex holds an EngD in Biochemical Engineering from University College London and an MEng in Chemical Engineering from The University of Manchester, Manchester, UK.

Affiliation

Alex Chatel, Senior Product Manager, Donaldson Life Sciences, Lyon, France

Article & Copyright Information

Copyright: Published by Cell & Gene Therapy Insights under Creative Commons License Deed CC BY NC ND 4.0 which allows anyone to copy, distribute, and transmit the article provided it is properly attributed in the manner specified below. No commercial use without permission.

Attribution: Copyright © 2025 Univercells Technologies. Published by Cell & Gene Therapy Insights under Creative Commons License Deed CC BY NC ND 4.0.

Article source: Invited.

Revised manuscript received: Jun 4, 2025.

Publication date: June 16, 2025.