Overcoming Batch-to-Batch Variation in Viral Vector Manufacturing
Process and product consistency is key to regulatory approval and therapeutic outcomes
Gene therapy was conceptualized alongside a revolution in genetic sequencing and gene delivery technologies. Currently, there are numerous US FDA-approved gene therapy products to treat a variety of conditions.
In addition to improving the vector constructs, optimizing manufacturing and analytics of viral vectors is equally important. Viral vectors used clinically for in vivo therapies are predominantly adeno-associated viruses (AAV) produced by transient transfection of human embryonic kidney (HEK) 293 cells. The transfected cells are then allowed to expand in a bioreactor and are typically lysed to harvest the viral vectors. A series of purification steps follows, and viral titers and the percentage of full capsids are analyzed before formulation and filling. Production of lentiviruses, which are established viral vectors for CAR-T cell therapies and ex vivo gene therapies, has many of the same steps. Owing to the complexity of virus manufacturing, there can be significant batch-to-batch variation. Here, we will seek to understand the factors causing the variation and strategies to overcome them.
Controlling cellular variation factors
Viral vector production exploits cells as biofactories to produce viruses. Therefore, a considerable portion of batch-to-batch variation arises from cellular differences stemming from cell health, culture conditions, and transfection or transduction protocols. While some biological variation will always be inherent, labs and manufacturers can adopt several strategies to minimize cellular variation factors.
The health of the cells is a major factor influencing variations in product quality. Keeping the passage number of the cells between five and 20 will help ensure consistency. At higher passage numbers, mammalian cells display alterations in morphology, grow slower, express fewer proteins, and suffer from poorer transfection efficiency. Older cells may also perform post-translational modifications of viral proteins differently.
The cell culture media can also affect the health of cells, which can result in variations between batches. Manufacturers should ensure that their media and reagents (including serum) are exclusively ordered from known and trusted manufacturers to minimize the risk of low-quality materials compromising the process and product outcomes.
Variation can also be introduced at the expansion stage in bioreactors. When cells are grown suspended in bioreactors, an impeller helps in mixing, aeration, and heat and mass transfer within the vessel. Care must be taken to control the conditions to maximize cell growth and minimize mechanical stresses that can affect cellular health.
The creation of a producer cell line can help overcome slowing cell proliferation and low transfection efficiency. Producer cell lines can be generated by integrating required genes in the host cell genome, although they are challenging to establish.
An important criterion for regulatory approval is to demonstrate single-cell clonality to limit process variability. Therefore, factors such as cell passage number, growth conditions in bioreactors, and choice of transfection method play important roles in minimizing biological fluctuations that ultimately influence product consistency.
Other sources of process variation
After expansion, cells are typically lysed to release viruses inside, followed by treatment with an endonuclease to reduce the length of contaminating DNA outside of cells. Enzyme treatment might need to be repeated to obtain sufficient results. Following treatment, the choice of clarification filters and conditions is essential to maximize recovery. Chromatography-based methods are commonly chosen to separate viruses from contaminants using criteria such as affinity and charge. By using a chromatography matrix that selectively binds the viral vector or binds to negatively charged capsids, many contaminants in the mixture flow through; the viral vectors can be eluted out with a specific release buffer. In the case of charge-based separation, this process selectively releases capsids with different compositions based on binding strengths. However, this “polishing” step is challenging to optimize, as the charge differences between empty and full capsids are small. Chromatography options are highly customizable, and products can meet the needs of both small- and large-scale viral vector manufacturers.
Measurement variation factors
Throughout the production process and at the end, viral vectors need to be assessed for yields/titers, the ratio of full capsids (for AAV), and the level of contaminants including proteins (such as host cell protein) and host cell DNA. Accurate measurements and analytics allow understanding of the variability in process and product.
A historical way to determine viral titer is the plaque assay where serially diluted viral particles are added to host cells. Agarose is then added on top of the cells to prevent viral spread after host cells lyse and release viral particles. Crystal violet is next added to visualize the plaque colonies with microscopy. The viral titer is finally calculated by multiplying the number of plaques with the dilution factor.
While inexpensive to perform, the plaque assay is subject to human error as operators need to count plaque colonies of different sizes manually. One way the industry is trying to overcome this problem is to make use of an automated system to acquire and analyze images. There are also options using primary anti-viral antibodies and fluorescently labeled secondary antibodies, with flow cytometry, to quantify viral titers more objectively, as well as rt-qPCR (real-time quantitative PCR).
Historically, density gradient centrifugation has been used to analyze viral vector composition by isolating full particles from partial and empty capsids. This method is accurate but time-consuming. Other methods such as electron microscopy and mass spectrometry are also used to determine the ratio of full to empty and partially filled viral capsids. They are highly sensitive but performing them is costly, labor intensive, and has low throughput. It is important to evaluate the options including less costly alternatives like UV absorbance as well as upcoming techniques, such as anion exchange chromatography, and to use multiple methods to ensure accurate and robust quantitation.
The levels of nucleic acids outside of viral capsids but remaining in viral vector preparations also play a role in consistent therapeutic outcomes. However, nucleic acids that are trapped inside capsids cannot be removed with nuclease treatment. Thus, these encapsidated contaminants remain in the final product. Genetic engineering of cell lines is one avenue being explored to reduce the level of host cell DNA inside capsids.
Finally, the potency of viral vectors needs to be assessed before use. In addition to the previously determined viral titer, the genome copy number in the host cell by analytics such as dPCR (digital PCR) can quantify the number of viral vector genomes. This value is taken into account when assessing the potential of a viral vector as a gene therapy treatment, as using a low-potency preparation is costly and could affect clinical outcomes.
Gene therapy has progressed tremendously since it was first conceptualized. To make it effectively and efficiently at a large scale to expand accessibility, manufacturing must be further optimized to maximize yield and contaminant removal while reducing batch-to-batch variation. The variation in product yield and quality can stem from both process and product and requires accurate and reproducible ways to measure and analyze the critical qualities of the viral vectors. Better control of how viral vectors are made combined with accurate measurement using multiple methods can reduce the risk of poor-quality viral vector production. Continued innovation in technology and techniques and the careful choice of orthogonal analytics for accurate analyses will help to realize the full therapeutic potential of gene therapy.
