Where the focus on established processes such as monoclonal antibodies is centered on instrumentation and automation, trying to get productivity gains from an already well-established and understood platform, gene therapy is still in its infancy.
I have been fortunate to work alongside many customers and colleagues on a diverse range of biopharmaceutical manufacturing and development processes. Each different process comes with its own unique challenges, perhaps none more so than viral vectors.
Where the current focus on more established processes, such as monoclonal antibodies (mAbs), is heavily centered on instrumentation and automation and generally trying to squeeze out productivity gains from an already well-established and understood platform, gene therapy is still very much in its infancy, with the current focus being on trying to produce robust and scalable processes to move quickly from lab to market. In addition, the desire for platform processes to help reduce development time and ease of scale-up, is very much an ever-present consideration.
Many of the pitfalls associated with traditional biologics also occur in the gene therapy space. However, while there is a comparative lack of data compared to some of the other more well-understood biologics, learnings can certainly be taken from other bioprocesses to guide in development of robust processes.
Depth filtration
In the move toward more flexible, scalable platform processes for the manufacture of gene therapies, ultracentrifugation has been readily displaced by depth filtration as the primary method for the removal of cellular debris, high molecular weight host cell proteins (HCP), and host cell DNA.
Whereas ultracentrifugation can often be performed with minimal optimization, the selection of filter media for cell clarification is crucial to a successful filtration step. Critical considerations to be aware of are the production method of the upstream material: whether this has been generated using an adherent cell line or a suspension cell line, and the subsequent harvest turbidity. Material generated via a suspension platform tends to have a higher turbidity than material generated from an adherent platform, due to the higher cell densities associated with suspension platforms. In general, the higher biomass often associated with a higher turbidity requires filters with a larger surface area and often with a higher nominal retention rating as part of the optimized filtration train. Additionally, the selected filter must be able to scale linearly, often to 200 or 500 L, and this is something that should be considered at the beginning of process development work. Unlike resin-based chromatography, where the sorbent can be packed to meet a specific volume requirement, depth filters are often limited by the formats they are supplied in. Where scale-up is not possible due to a lack of a suitable format, scale-out can be considered. However, this comes with its own challenges – manifolding multiple filters together in parallel is a complex operation that is best avoided. For this reason, it is imperative to consider how the upper limit of your process may be limited by your choice of filter and to identify suitable, scalable filters as soon as possible in the development and optimization process.
Tangential flow filtration
Currently AAV upstream titers are relatively low when compared to the capacity of subsequent chromatography steps. So, to reduce downstream process volume, a tangential flow filtration (TFF) step is commonly used.
Like other steps in the process, TFF is not immune to scale-up problems stemming from development technology choices. To develop a successful TFF step at the bench-scale, it’s important to visualize how the process looks at large-scale. Hold-up volume, minimum working volume, and how process parameters scale with an increase in volume should be considered. Many systems have defined parameters at which they can work (e.g., maximum pump speed, minimum working volume) and need to be considered at the start of the TFF optimization process to prevent any painful and unnecessary bottlenecks later down the line. All of this can be underpinned by Quality by Design (QbD) principles and helps ensure a robust process. It may not be necessary if you are working at 10 L to develop a drug for a rare disease, but with the platform ability of an AAV process, it should be considered that the next therapy may require a volume ten times greater. Building in robustness facilitates scale-up and avoids going back to the drawing board, wasting money, time, and potentially impacting a drug making it to market.
Some, but not all, viral vectors have a tendency to aggregate the more concentrated they become. Combined with the shear inducing recirculation that is traditionally part of TFF, this can quickly become a problem that isn’t always easy to solve. From speaking to customers and colleagues there are two main approaches to dealing with this. The first approach tries to address the problem with formulation studies to determine the best combination of buffer excipients (e.g., salt, sugar, amino acids, etc). The second approach utilizes an additional step whereby endonuclease is recirculated with the viral vector product, either prior to or after concentration. The rationale being that in some cases, aggregation may be brought upon by host cell DNA and not necessarily the vector itself. It can be challenging to determine the root cause for aggregation within a TFF step, but it is something that is wise to address in the developmental stage before scale-up is considered.