How can CFD support cell and gene therapy processing?

What are stirred tank bioreactors?

Stirred tank bioreactors are widely used for a variety of biochemical processes within laboratories and manufacturing plants. They essentially comprise a cylindrical vessel designed to provide the optimal environment for the growth of microorganisms or cells used in industries like pharmaceuticals, chemicals and food processing. Inside the vessel, there’s a stirrer or agitator, typically linked to an impeller (a rotor used to accelerate the flow of fluid), which is key to the reactor’s core function.

The primary objectives of this process are to augment rates of heat and oxygen transfer, ensure the uniform distribution of components and regulate shear rates to remain beneath the disruption thresholds pertinent to the biologics that are being manufactured. This meticulous control is crucial for preventing potential damage to the biological entities.

The operation of a stirred tank bioreactor involves introducing the cells or microorganisms into the vessel, along with a liquid medium that provides the necessary nutrients for their growth. Oxygen, essential for aerobic biological processes, is usually supplied by gas-sparging (the process of introducing air or oxygen-enriched air into the reactor). The temperature, pH and oxygen levels are closely monitored and controlled to maintain optimal conditions for the biological process. Finally, the end products are then harvested for further processing.

The role of CFD in bioreactor optimisation

Computational Fluid Dynamics has emerged as a significant tool in the field of bioreactor design and optimisation, particularly in stirred tank bioreactors. One of the primary reasons for CFD’s widespread application is the ability to predict the performance of full-scale manufacturing bioreactors. The small scale of some bioreactors makes accurate measurement challenging, meaning scaled-up prototypes are sometimes used to characterise them. While accurate measurements can be achieved on these physical prototypes, these can be costly and time-consuming to produce. CFD offers a viable alternative that has been shown to be comparable to physical prototype models. As a result, CFD can potentially reduce the need for time-consuming and costly bioreactor prototype studies.

In addition to performance prediction, CFD can provide a higher level of precision than experimental parametric studies on the process conditions required in bioreactors, such as pressure, temperature and flow rate. This feature produces valuable insights during the early stages of bioreactor design and process development.

CFD also provides detailed visual data that can complement or even exceed the data obtained from experimental methods. In the context of stirred tank bioreactors, CFD is used to simulate and analyse impeller mixing and gas sparging. It also enables analysts to determine key parameters such as the specific power input, Kolmogorov length (a measure of the smallest eddies in turbulent flow), hydrodynamic stress, mixing time and oxygen transfer rate.

Understanding these parameters plays a critical role in avoiding cell damage, since cells that are larger than the Kolmogorov length scale may be damaged by the shear stress induced by simultaneously flowing in different directions. Cells smaller than the flow will not be impacted this way and will just follow the eddies and avoid damage. By comparing the mean Kolmogorov length scale and average cell size, analysts can identify whether catastrophic cell damage due to mechanical agitation is likely to happen.

In physical testing, these parameters are often masked by natural variation. This makes CFD a valuable tool in the design and optimisation of bioreactors.

Bioreactor design optimisation using CFD

There are several examples of how CFD has been used to optimise the design of bioreactors. For example, Aston University, University of Birmingham and UCL engineered a novel stirred-tank bioreactor vessel for the process development of cell and gene therapy candidates. Unlike similar bioreactors, the design was devoid of baffles (the components used to increase the mixing efficiency inside the bioreactor).

As part of the development, the team needed to determine if the “unbaffled” vessel could maintain the necessary flow patterns required for cell and gene manufacturing. Using CFD simulations, the team determined that the bioreactor could effectively support the culture of T-cells with Dynabeads and human Mesenchymal Stem Cells (hMSCs) on microcarriers. As a result, they determined that the bioreactor would be an efficient tool for both cell and gene therapy manufacturing and potentially autologous manufacturing as well.

Comparing CFD with scale-up prototype and
full-scale bioreactors

As mentioned previously, the small scale of bioreactors means that obtaining accurate measurements is often difficult. Building a full-size bioreactor for testing can be challenging, therefore scale-up prototypes and CFD have become popular alternatives during development.

In a recent study, Bayer AG and Leverkusen compared how these approaches performed when determining power input for low-speed stirred-tank bioreactors. The experimental setups yielded comprehensive data for the validation of the employed CFD setup. The study concluded that the experimental characterisation of a reactor can be considerably more time-intensive compared to a CFD approach. This is as the lead time for specialised sensor equipment and the fabrication of customised scaled-up reactors spans several weeks. In this context, a CFD-assisted workflow can expedite results, thereby accelerating overall project progress, especially as additional characterisation parameters, such as shear stress, can be concurrently obtained. The faster development times for a CFD model over a scaled-up experimental one may be beneficial to note when planning future activities.

Perfusion bioreactors

Perfusion bioreactors are used for cells that require a surface on which to grow. This (usually static) cell substrate is gently perfused with aqueous fluid containing nutrients and other media, such as differentiating agents, to ensure all cells are sufficiently nourished and have the desired properties. These differentiating agents are used to convert stem cells into cells for a particular tissue, such as bone or muscle.

A wide array of perfusion bioreactor setups have been explored to date, ranging from commercial-scale and larger lab-scale processes, to one-off bespoke lab equipment.

How CFD can optimise perfusion bioreactors

CFD is relatively limited in its ability to support the development of perfusion bioreactors, however studies have demonstrated its potential in optimising bioreactor designs and cell culture processes, in order to augment cell proliferation.

For example, factors such as fluid shear and pressure gradients that may cause cell damage can be quantified and minimised using CFD. Additionally, CFD can be used to examine and optimise the perfusion rate of fresh nutrients and other media across different locations on the reactor. This can help accelerate cell and gene development programmes by informing which experiments are likely to be most successful.

From a fluid dynamics perspective, CFD models are typically straightforward to execute. CFD is an ideal tool to understand the flow within complex 3D geometries including the impact of small manufacturing defects. To date, it has been successfully applied to commercial cell processing equipment and bespoke lab setups, including complex 3D scaffolds that mimic biological tissue structures.

Case study – using CFD to troubleshoot cell cube reactors

The following case study illustrates how CFD can be used to understand and resolve issues with poor cell growth within a bioreactor. The Corning® CellCube® Culture System is a commercially available bioreactor that facilitates cell growth on diamond-shaped parallel plates, which are perfused by a diverging-converging flow. During early evaluations of the device, irregular cell growth patterns were identified on these plates, including regions of poor cell growth. A team of researchers conducted investigatory CFD simulations to identify the cause, using flow patterns and other CFD data to explain the majority of the uneven growth patterns.

Using particle-tracking analysis, an in-built CFD tool, to model the trajectories of the cells within the equipment, the team discovered low cell growth at the edges of the front plate could be attributed to initial cell seeding not extending to the plate edges. Meanwhile, regions of poor cell growth at the inlet nozzles and central portion of the back plate were attributed to high shear rates leading to cell detachment.

The CFD analysis led to proposed improvements to eliminate the issues with poor cell growth, including eliminating plate bowing and identifying alternative cell seeding approaches.

Case study – scaffold reactor selection

As mentioned, perfusion reactors can be used to build complex 3D scaffolds that mimic biological tissue structures. In one study, researchers compared the flow patterns through a scaffold cell-culture substrate in two reactor designs, in order to optimise the media perfusion rates to different regions of the scaffold. The porous scaffold provided a three-dimensional representation of the bone microstructure. This environment was designed to engineer bone tissue from mesenchymal stem cells.

Using CFD models, analysts identified notably different flow distribution behaviour between the two reactor designs. One design resulted in much reduced fluid shear and more even nutrient perfusion throughout the scaffold, compared to the other. These modelled findings were confirmed by higher cell growth rates in the improved reactor design.

Overall, CFD proved instrumental in understanding how to optimise shear stress for cell growth and osteogenic differentiation.

Future trends

Cell and gene therapy is a growing area and correspondingly, the application of CFD to support these developments is anticipated to increase. As project teams become more aware of the available tools and their potential to bolster their projects, more and more will be turning to CFD. Importantly, CFD has the potential to help bring down development time scales, a key barrier to accessing these vital medicines.

As described in this article, CFD can play a critical role in helping to solve real world problems, such as the different impacts of flow shear and other variables on poor cell growth, or the optimisation of reactor design to maximise process yield.

It is clear that CFD has much to contribute to the exciting area of cell and gene therapy. It will be exciting to see what the next few years of development hold in store.

Sources

1. Rotondi et al, https://link.springer.com/article/10.1007/s10529-021-03076-3
2. Keller et al, https://onlinelibrary.wiley.com/doi/full/10.1002/cite.202200143
3. Aunins et al. https://aiche.onlinelibrary.wiley.com/doi/abs/10.1021/bp0256521
4. Egger et al. https://pubmed.ncbi.nlm.nih.gov/28952530/