Considerations for Scaling Up Purification Processes

Scaling for Success

Scale-up of chromatography operations is usually achieved by increasing the column diameter while maintaining the resin bed height and linear flow rate. This ensures that the residence time is the same at all scales of operation. Other factors that should not change as chromatography operations are scaled include the buffer setpoints and tolerances in terms of pH and conductivity ranges, as well as quality inputs and outputs. Some parameters will change during scale-up, necessitating optimization of the workflow in terms of volumes, processing time, and equipment.

Among the most fundamental factors to consider when scaling chromatography steps is whether the resin used at lab scale is available for clinical- and commercial-scale operations; for example, size exclusion chromatography is not scalable. Resin cost can also become problematic; at a much larger scale, some resins may become prohibitively expensive.

Even if a resin is available in large quantities and is reasonably priced, process developers should remain mindful of their technical and physical limitations in terms of their scalability. If the chromatography process does not scale in a linear fashion, an investment in optimization will be essential. It is also important to note that smaller-diameter chromatography columns have wall effects that are not present in larger-diameter columns; as such, a high pressure drop at a lower flow rate may be observed with the larger scale process.

This is also an opportune time to determine whether all the existing process steps are necessary. It may be possible to combine or reorder the columns to gain efficiencies such as a reduction in the amount of buffer exchange.

Efficiency and capacity may also be improved by using another type of resin with different properties; an example would be switching to a resin with a better salt tolerance, which can lead to less or even no dilution, resulting in a substantial time savings at larger production scales.

A trial of the optimized downstream workflow is typically conducted in the form of one or more pilot runs in order to validate the process, confirm process reproducibility, and verify stability of the target protein. The process development group will have determined the set of analytical tests needed to validate that the mAb is the same as what was produced at lab scale, and the tests to confirm successful removal of impurities.

Once the pilot runs have demonstrated that the process has been successfully scaled, engineering runs are performed at full scale as a final step prior to full-scale production. Engineering runs are used to test production equipment, finalize batch production records, train operators, and finalize a process control strategy (Atkinson 2011).

Another important consideration related to scale-up is the destination of the process and any associated limitations at that location. It is critical to know the capabilities of the receiving site and the team to which the process is being transferred. While it may seem obvious, the site must have the necessary column, equipment, materials, space, and expertise to operate the process.

Gaining Efficiencies with Prepacked Columns

Ensuring the highest quality chromatographic separation depends on many factors, including the technique used for packing the column. Well-packed columns, in which media beds are homogeneous, stable, and continuous from top to bottom, provide for the best separations. Successful column packing ensures proper mobile-phase distribution and resin contact, which helps to ensure desired yield, separation, and product quality. Poor-quality packing can lead to solute band broadening and compromise quality parameters. Lack of consistency and performance of chromatography columns at pilot- and production-scale can prove extremely costly in terms of disruptions and delays in the manufacturing workflow and batches that may have to be scrapped.

For lab-scale separations, columns are relatively easy to pack, and disposable prepacked columns are available. At larger scales, however, packing and qualification can be challenging, as industrial, large-scale columns have many different configurations. Additional challenges are presented by the fact that production-scale columns are typically used over a number of cycles. As such, process developers must monitor whether the pressure of the packed column is stable throughout each run, and from one run to another. After a few cycles, impurities may build up, affecting column performance. Changes in column performance over time with repeated use must be evaluated to determine the number of cycles the column can be successfully used in and how stable the resin is.

Given these challenges and the skill needed to pack large-scale columns, automated packing may be an option. Aldington and Bonnerjea (2007) offer a precaution, however. While equipment is available to automate packing of large-diameter columns (>30 cm) by pumping in a resin slurry, each type of resin behaves differently in terms of the conditions required to optimize packing. The authors note that extensive trials are needed to develop robust and optimal packing and unpacking procedures for pack-in-place columns.

The use of prepacked pilot- and production-scale columns is an attractive alternative to shorten workflows, decrease variability, and eliminate the need for validation protocols and physical characterization of the column using asymmetry testing, for example. A prepacked column simply needs to be taken from the warehouse and connected to the process, eliminating many time-consuming and labor-intensive steps. This approach is particularly valuable for smaller and/or early-stage biotech companies with limited resources. Prepacked columns also support process intensification efforts and facilitate tech transfers around the world by reducing the time needed to get a column in place and by reducing inconsistencies that can result from different packing techniques.

Schweiger et al. (2019) demonstrated the packing quality, consistency, reproducibility, protein binding capacity, and separation efficiency of prepacked columns ranging from 1 ml column volume (0.5 cm diameter × 5 cm length) to a 57 L industrial-scale column (60 cm diameter × 20 cm length).

Their conclusions were summarized as follows:

• Uniformity of column packing across column sizes was confirmed using acetone pulses, which are susceptible to changes in the packing structure
• Acetone pulse injections confirmed that all the columns were packed to the same quality attributes measured by height equivalent to theoretical plate (HETP) and asymmetry
• Equilibrium and dynamic binding capacities for a model protein showed only slight variations with scale, which were explained by small changes in the salt concentration of the loading buffer
• The purity of the elution pool of three proteins was equivalent across all column scales despite differences in retention time and peak width, which were due to variance in the sharpness of the conductivity change attributed to mixing and extra-column effects being more prominent with the small-scale columns

Conclusion

The production processes used to manufacture mAbs must be efficiently and effectively scaled to meet patient demand and must ensure the integrity and purity of the target protein. While scale-up of the chromatography steps in the downstream workflow can be relatively straightforward, consideration of several key factors can help ensure a successful result. Among the things to keep in mind are whether the resin is available in the necessary quantity and whether the resin can in fact be scaled. Technical and physical limitations of the resins must able be taken into account.

To facilitate scale-up of chromatography steps, prepacked columns can be used. Their consistency, reproducibility, protein binding capacity, and separation efficiency from lab to production scale has been demonstrated and this approach offers significant time and labor savings.

References

Aldington S and Bonnerjea J. (2007). Scale-up of monoclonal antibody purification processes. J Chromatogr B Analyt Technol Biomed Life Sci 848, 64–78. https://doi.org/10.1016/j.jchromb.2006.11.032, accessed February 26, 2023.

Atkinson EM (2011). Why not do an engineering lot? Contract Pharma. https://www.contractpharma.com/issues/2011-10/view_bio-news-amp-views/why-not-do-an-engineering-lot, accessed February 26, 2023.

Schweiger S et al. (2019). Packing quality, protein binding capacity and separation efficiency of pre-packed columns ranging from 1 mL laboratory to 57 L industrial scale. J Chromatogr A 1591, 79–86. https://doi.org/10.1016/j.chroma.2019.01.014, accessed February 26, 2023.