Considerations for successful scale-up and viral vector purification

Viral vector manufacturing is critical to the advancement of the field of cell and gene therapy. Viral vectors are used directly as therapies or indirectly for the manufacturing of cell therapy products. During the development of any CGT product, there are typically multiple production scales required, from small scale research production for proof-of-concept studies to larger scale research production for preclinical studies and/or toxicology studies, and finally to GMP production for clinical use. As a product progresses through development, there are parallel needs for not only higher quality material but also progressively larger batches. Therefore, the production process utilized for early-stage work must inevitably be scaled up.

There can be significant differences in the manufacturing process at the benchtop scale versus a clinical scale, as is often the case for products developed in academic laboratories that require larger batch sizes for clinical studies. Similarly, there can also be a need for process scale-up for products already in the clinic, as they progress from early phase to late phase clinical testing and then onward to commercial production. In both cases, process scale-up becomes a critical step along the product development pathway.

To scale up a process, there are several important considerations, starting with the selection of a target batch size. However, process scale-up can encounter challenges. Here, several key considerations will be discussed, and specific examples of potential challenges faced during the scale-up of upstream and downstream production will be presented.

Scale-up considerations

One of the first considerations relates to the production method. Very often, scale-up involves a shift from adherent to suspension culture, though this is not always required. This is because small-scale benchtop production can typically be done more inexpensively in adherent culture, and so this method is commonly used for early proof-of-concept studies. In contrast, suspension culture is regarded as easier to scale and is generally the method of choice for larger-scale preclinical, clinical and commercial production. Likewise, the purification methods used at the smallest scales (often centered on centrifugation) can differ quite substantially from larger scale manufacturing (which tends to be chromatography-based).

Another early consideration is to identify the target production scale which is driven ultimately by the needs of the clinical study (which is itself influenced by vector dose, number of patients, route of administration, potency, etc.) and is also influenced by expected vector productivity. The needs of QC testing for product release and stability must also be considered since these activities can require a significant portion of the final product.

The final scale determines the number of scale-up batches that will be required. Generally, increases in scale should be limited to 5-fold or less. For example, a product to be manufactured at 200L scale for a clinical study with a process developed at 10L would need a scale up batch at 50L, whereas an additional scale up batch would be needed if progressing beyond 200L. It is important to note that scale-up batches may have value beyond confirmation of a production process. The batches may be needed for additional preclinical efficacy testing, dosing and biodistribution studies, formulation development, analytical development, stability studies, or other reasons.

During scale-up, it is always advisable to limit process changes to the extent possible since significant changes may prompt the need for comparability studies. However, in some cases a process needs modification or additional optimization, particularly if it is coupled with a technology transfer, as when production is shifted from one manufacturing site to another. As a few examples, there may be intellectual property barriers to the transfer of cell lines and plasmids, and/or there may be facility or equipment differences between manufacturing sites. Regardless, it is always best to invest in process development and/or optimization as early as possible, so that the process can be relatively fixed before further scale-up. 

Upstream and downstream challenges

Once the required scale has been identified, additional considerations relate to the process itself, for both upstream production and downstream purification. On the upstream side, the key consideration is that cell growth profiles may change as scales increase. Cell doubling times and viable cell densities can be altered in the typical progression from shake flasks to a bioreactor, or from one bioreactor type to another one (such as going from a rocking bioreactor to a stirred tank reactor), due to different shear forces and aeration strategies, among other factors. Different cell growth properties can alter transfection (AAV/LVV production) or infection (adenoviral vector) conditions and thus the amount of vector present at harvest, which impacts the downstream purification. 

“It is always best to invest in process development and/or optimization as early as possible, so that the process can be relatively fixed before further scale-up.”

Taking a closer look at the downstream purification process, key considerations will depend upon the overall manufacturing strategy. As mentioned above, small-scale purification of many vector types often involves density gradient ultracentrifugation. In the case of AAV production, this step may even be included in largescale manufacturing to enrich for full capsids. Although it is effective at this purpose, ultracentrifugation can be cumbersome to perform at clinically relevant scales (such as 200L) for a number of reasons. Ultracentrifugation requires several labor-intensive steps such as the layering of the different density gradients and fraction collection after the run. These steps become progressively more time consuming as scales increase, since larger scales require more tubes. The increase in handling time can be mitigated in part by preparing the fraction collection tubes ahead of time and by using a multichannel pump to perform gradient layering in multiple tubes in parallel. 

Another challenge related to ultracentrifugation is that larger production volumes may require multiple runs to process all of the vector material. Adding an additional centrifuge is one solution, though this is not always feasible due to cost and space limitations. Instead, it is sometimes possible to use larger rotors in existing equipment, allowing more tubes to be centrifuged at once.

Downstream purification at large scale commonly includes one or more chromatography steps, and in these situations considerations such as filter sizes, chromatography column dimensions and volumes, and flow rates are critical. During process scale-up, the goal is to scale linearly to the extent possible, since challenges can arise when this is not the case. Two typical process steps provide examples of challenges that may occur during scale-up. During column chromatography, it is preferable to maintain the same bed height when upsizing a column so that flow rates can be kept constant. However, the amount of vector per unit volume in the crude harvest may differ from one scale to another for reasons mentioned above. To mitigate the risk of overloading the column, it may be advisable to use a slightly larger column size than would be suggested by simple linear scaling.

Challenges can also arise during ultrafiltration/diafiltration by tangential flow filtration, which is often done using hollow fiber membranes. In this case, linearly sized membranes may not be available commercially, forcing the selection of potentially large for the expected volumes at the larger scale. Here, it may be better to select a longer membrane with a slightly ‘undersized’ total area. This strategy allows for flow rates to stay consistent, though it can extend processing times. In the examples above, selection of the appropriate-sized column or TFF module for vector purification should be made carefully and should be guided by data obtained during process development.

Process scale-up is an inevitable part of the development pathway for most CGT products. Although it is not without challenges across both upstream production and downstream purification, risks can be mitigated by rational and data-driven process design.