OR WAIT 15 SECS
Randi Hernandez was science editor at BioPharm International from September 2014 to May 2017.
Scale-up of complex, innovative products requires commercialization models that are sustainable.
There are many difficulties that can arise when attempting to predict process scale performance. As small-scale systems are phased to production volumes, biologics can be particularly problematic, as there are so many variables at play with living organisms. There also are different problems that arise whether a company is working with mammalian or microbial cells.
Manufacturing scale-up is often a key element of a larger technology transfer project, and can include the transition from a development pilot plant to a commercial manufacturing plant or from a developer to a contract manufacturing organization or contract and development manufacturing organization (CMO or CDMO).
Author Cynthia Challener wrote that scale-up range has decreased. The severity of each scale-up is no longer as great as it once was in previous years, because in some cases, therapies are becoming more personalized and batches are not as large (1). Even though scale-up procedures may not be as large of productions as they once were, there is still a lot riding on the their successful implementation: Matzmorr asserted that every month lost during a technology transfer carries a price tag of $80 million (2). According to Richard Snyder of CMO Brammer Bio, two of the most common problems that plague scale-up operations are maintaining yield during production and handling an impurity not seen at small scale that may require modification of purification parameters. Other crucial issues include titer differences in cell culture and aggregation challenges during downstream processing, adds Kumar Dhanasekharan, director of process development at Cook Pharmica.
Failure to scale can occur when there are issues with mass transfer or mixing, if larger equipment for scaling is not available and if development is using equipment that cannot be scaled (i.e., in unit operations such as size-exclusion chromatography). The key to understanding a process is to make sure that it has not been created in a way that would make it impossible to scale in the long run, says Peter Levison, PhD, senior marketing director for downstream processing at Pall. “This can be a particular problem with chromatography, where the process performs very well at the small scale, but is impossible to bring to a larger scale.” This phenomenon can be observed with what is known in the industry as the “wall effect”, which Levison says occurs in a “process with columns of less that 10 cm diameter where the wall adds a lot of support to the bed, enabling the achievement of high flow rates-flow rates that are not possible to replicate at the larger scale.” Because of potential problems such as the wall effect, Levison states it is crucial to ensure that scale-down process parameters are aligned to the performance of the large-scale equipment.
Specific scale-up parameters may include adjusting for volume-to-surface ratio of process tanks for mixing steps with or without heat transfer; volume-to-membrane-surface ratio for ultra- and microfiltration steps; cooling rates for freezing processes; and maintaining constant bed heights and process cycles for chromatographic steps (2). Although scale-up challenges can vary, “applying appropriate scale-up principles and understanding the implications of scale-dependent parameters such as gassing strategy, pH control, agitation in cell culture, linear velocity, and residence time in chromatography scale-up provides a good starting point,” states Dhanasekharan. It is possible to avoid most of the challenges of scale-up by achieving “molecule structure/stability understanding through early developability studies” and by being actively involved in cell-line development and clone selection, Dhanasekharan asserts. Notes George Yeh, president of Taiwan Liposome Company (TLC), a biopharmaceutical company specializing in lipid-based formulations and scale-up of parenteral drugs, planning for large-scale production from initial stages of product development “has a clear impact on the final cost of goods.”
Product-specific considerations of scale-up
Scale-up from an academic setting can be particularly challenging, notes Snyder, because cell lines and/or genetic constructs are not characterized or are in limited quantities. Sometimes, culture conditions and purification parameters are not amenable to scalable formats, he says, and must be redesigned for commercial applications. Yeh comments, “The major difference between industrial and academic settings is that reproducibility of results from academic research is much lower because of a lack of standardized procedures.”
When a pharmaceutical company is working with a CMO, the CMO may find the processes that have been provided to them are not well designed and must be reworked to achieve the desired outcome. Choosing an appropriate, possibly local, CMO based on the nature of the product and the target market may be the best strategy, says Yeh, as a product may require additional scale-up activities. Dominik Wegmann, PhD, head of manufacturing science and technology, microbial biopharmaceuticals, Lonza, says that a typical example of a common scale-up issue for a CMO is when a centrifugation step with a beaker or tubular bowl centrifuge has to be transferred to a disk stack centrifugation step.
According to Cynthia Challener, microbial cultures perform better in stainless steel, while mammalian systems that do not require high heat removal and large amounts of oxygen are better suited for single-use bioreactors (1). According to Mohsan Khan, technical director, process development sciences, Lonza, some of the problems that are specific to mammalian systems include a faster decline phase at large scale, which could lead to “premature reduction in product accumulation,” or high levels of dissolved CO2 in culture, which could lead to high osmolality, cell growth cessation, and could potentially impair “the cell’s ability to correctly perform post-translation modification of the expressed product.” For microbial models, unexpected low titers and unexpected filter clogs are common, and finding ideal stirring conditions at large scale can also be a challenge. Sometimes, the filtration step in microbial models is not scalable due to facility restrictions, and sometimes the “linear flow needs to be lowered due to the volumetric flow restrictions in the upscale facility,” Wegmann adds.
The introduction of off-the-shelf equipment has negatively affected processes and has led to poor performance in some scale-up attempts and process transfers across a manufacturing network, says Khan, especially in the case of disposable bioreactors. Some of the disposable bioreactor off-the-shelf designs can be associated with batch-to-batch variability, and “adsorption of some hydrophobic cell-culture nutrients-such as cholesterol on the plastic film-can introduce additional operational challenges for some processes, cell types, and products that are sensitive to such components.”
Scale-up of complex products
Scale-up can be more of a daunting task for large-molecule products and engineered proteins. “More complex antibodies such as bispecifics, engineered Fc regions such as extensions, CH1-CL domain swap, cytokine fusion proteins, etc., can have challenges such as lower titer, product variants such as missing one light chain, and permutations of different fragments,” says Dhanasekharan. “This can make development more complex and will require longer timelines.”
There are many current constraints with scale-up of cell- and gene-therapy products, specifically. Alexey Bersenev, MD, PhD, director, advanced cell therapy lab at Yale University, wrote in a blog entry that large scale-up of cell culture is not always desirable (3). Cell-therapy processes initially designed in T-flasks (such as those for human mesenchymal stem/stromal [hMSC] cells or pluripotent stem cells) typically have to be redesigned to meet commercial demand (4), and going from flasks to microcarriers-and then harvesting cells from microcarriers via a recovery step-introduces a lot of new variables. Stem cells are sensitive to cues from the microculture environment. Maintaining potency during scale-up is necessary, and there is a concern that large-scale commercialization of some cell therapies could lead to cell senescence, with a limited ability to expand (5). Says Kim Bure, director, regenerative medicine at Sartorius Stedim Biotech, for large-scale expansion, “there should not be a knock-on effect of scale-up, yet oftentimes, this [scale-up optimization] work has not been done and the yield or product efficacy suffers.”
Eytan Abraham, PhD, head of cell-therapy research and technology at Lonza, states that for cell therapies, the biggest manufacturing challenge is scaling the processes without losing product quality. “In some cases, this scaling requires significant changes to manufacturing platforms-for instance, moving from 2D to bioreactors and the concomitant use of appropriate downstream solutions.” Abraham asserts that the changes are challenging “not only from a hardware and know-how perspective, but also from a product comparability perspective.” Comments Bure, the square footage required to scale up these therapies is “enormous,” and the move to single-cell, aggregate, or microcarrier suspension culture must be optimized. Therapeutic efficacy must be preserved at the same time, says Bure: “For the best results, many factors in parallel and conjunction must be assessed for a thorough design of experiment (DoE), resulting in longer timelines and lower success rates for translation.”
For autologous cell therapies, scale-out is the main issue, says Abraham. “The solution is closing and automating the process to reduce labor and space while reducing cost of goods and increasing process control.” Conversely, scale-up and timely processing are concerns for allogeneic programs.
Although advances in 2D and 3D cell culture and microcarrier bioreactor technology may help to improve the commercial manufacture of certain allogeneic cell therapies, there is still work to be done to improve the performance of scaffolds (5, 6). In fact, recent studies have looked at standard manufacturing principles from the textile industry to help investigators more efficiently scale up the manufacturing process involved in cell and tissue engineering scaffolds. The authors of the aforementioned study say that using a woven scaffold instead of a scaffold made through the creation of an electric field (electrospinning) is a more cost-effective method of scaling up (7). In a separate study, T. Ma et al. wrote that in addition to improvements in microcarriers, “incorporation of novel 3D scaffolds in [a] scalable bioreactor system may be an alternative that provides optimal cellular microenvironments while maintaining scalability for large-scale expansion” (6).
Because there are specific considerations for certain cell types, it’s necessary to weigh scale-up experiments appropriately. Investigators should not eliminate or exclude potential high producers (in cell culture) on the basis of data in shake flasks without making sure other process variables have been adequately controlled (8). Gonçalves et al. saw in an experiment that lack of control of pH and dissolved oxygen in shake flasks reduced pDNA production and increased the levels of acetate in the samples. Thus, the authors wrote, “productivity data obtained from shake flask experiments often fails to predict the outcome of pDNA production in bench-scale bioreactors …” (8). As Snyder notes, when scaling up transient transfection of mammalian cells using pDNA from shake flask scale to reactor scale, “the transfection method dictates the time available for mixing the DNA and addition to the culture at different scales.”
In particular, innovative products, such as antibody-drug conjugates (ADCs) and cell therapies, can make scaling up more complex, because both scale-up and scale-out must be considered. Because ADCs are heterogenous mixtures, drug loading, monomeric purity, and linkage sites should be controlled by the process as much as possible, says Laurent Ducry, PhD, commercial development of ADCs, Lonza. Ducry adds, “Scaling-up an ADC process can be challenging if manufacturability was not sufficiently considered during process development.”
Preparing for successful scale-up
A successful scale-up is one in which there is a strong process development phase to optimize cell-culture conditions in scaled-down reactors, followed by the use of a reactor platform that allows proportional scale-up, says Snyder, making sure control parameters transfer seamlessly from small to large bioreactors. “Differences in scale-up stem from the fundamental engineering principles of surface area-to-volume ratio change upon scale-up, leading to different rates of mass transfer, mixing, etc.,” adds Dhanasekharan. “Applying the engineering principles appropriately enables successful scale-up irrespective of [if a developer is using] single-use bioreactors or stainless-steel reactors” for final scale-up steps.
Dhanasekharan says that simulation and bioprocess modeling can help reduce time spent on scale-up. In modeling, the engineering design space for bioreactor operation is mapped out, and this blueprint serves as a basis for scale-up. Although Dhanasekharan mentions that it is also a good idea to execute engineered batches at full scale when possible, he also says that the cost of running an engineering batch is significant. As a result of this cost, he says, more and more companies are proceeding with GMP batch at risk to save money.
Indeed, proper scale-up planning is crucial to test the efficiency of a process at large scale, especially when it comes to bioreactors in different locations or facilities. “Characterization of bioreactor kLa and mixing behavior and establishment of scale-independent models are important tools and are used in performing comparisons between different vessels during process transfer and scale up/down,” says Khan. “These allow a rapid assessment of a vessel’s ability to support the oxygen and mixing demands for a given manufacturing process, correctly specify appropriate agitation rates and gassing strategies to meet cell-culture needs, and prevent metabolic CO2 accumulation or excessive CO2 stripping from the culture.” However, Khan warns, “the utility of such models is limited by the quality of the data used in establishing the models and potential differences in the methodology used for determining the data.”
Post-scale up, Khan notes that the biggest contribution to process variability usually comes from “the procedural drifts and inconsistency in raw material quality.” He concludes, “The inability to consistently control process pH can also create process variability and can arise from introduction of different pH sensors and procedures used for sensor calibration and in-process corrections.”
1. C. Challener, BioPharm Int. 28 (6), pp. 20–23 (2015).
2. W. Matzmorr, BioPharm Int. 29 (4), pp. 46–48 (2016).
3. Alexey Bersenev, "Is problem of cell therapy scale overblown?,” Stemcellassays.com, http://stemcellassays.com/2015/12/is-problem-of-cell-therapy-scale-up-overblown/, accessed April 6, 2016.
4. Q.A. Rafiq and C.J. Hewitt, Pharm. Bioproc. 3 (2), pp. 97–99 (April 2015).
5. C. Challener, BioPharm Int. 29 (5), pp. 13–19 (2016).
6. T. Ma, A-C Tsai, and Y. Liu, Biochem. Eng. J. 108, pp. 44–50 (2016).
7. University of Missouri College of Engineering, "Dean, Colleagues Working to Make Tissue Engineering Cheaper, More Scalable," Press Release, March 28, 2016.
8. G.A.L. Gonçalves et al., Vaccine 32, pp. 2847–2850 (2014).
Article DetailsBioPharm International
Vol. 29, No. 6
Citation: When referring to this article, please cite it as R. Hernandez, "Scale-Up of Complex Biologics," BioPharm International 29 (6) 2016.