Scaling-up, converting a laboratory-based process so it can be run in large vessels, used to be core to every biomanufacturing strategy. But in recent years, advances in single-use technologies have made scaling-out—performing multiple laboratory-scale processes in parallel—a viable option for biomanufacturing firms looking to increase production volume.
There are pros and cons to each approach. Scale-out requires lower upfront investment due to minimized process changes and the ability to leverage knowledge from clinical production. However, reproducibility can be difficult to achieve. Scale-up involves more process changes. As a result, it can increase costs initially. But the approach can achieve lower cost of goods and improve supply chain security in the long term.
Deciding which of these strategies will be most effective for a project requires that developers take a huge range of variables into account.
Scale-out or scale-up?
When biomanufacturers decide between scale-out and scale-up, they should consider market dynamics, asserts Jian Dong, senior vice president of global manufacturing at contractor WuXi Biologics. In his view, the relevant factors include “the risks associated with process scale-up and its impact on product quality, the scale of manufacturing to meet and adapt to changes in product market demand, and speed to market.”
In other words, the scale-out/scale-up decision hinges on commercial strategy, which varies among manufacturers. According to Michelle Stafford, global marketing manager at GE Healthcare Life Sciences, a manufacturer’s commercial strategy reflects factors such as production goals, geographical constraints, buy/build preferences, and partnership decisions. Other relevant factors, she points out, include the size of the manufacturer’s portfolio and the types of products the portfolio contains. Manufacturers will also have products in different stages of clinical development.
Product quality is another important consideration, asserts Joe Makowiecki, enterprise solutions architect, GE Healthcare Life Sciences. “Scale changes are needed to meet product quantity requirements as [the manufacturer] moves through the phases from development to clinical production and on to commercial production,” he says. “There are options on how to get there—scale-up or scale-out.” Makowiecki adds that most firms using GE’s FlexFactory technology have adopted scale-out strategies, with processes typically involving two to six 2000-L bioreactors and one or two downstream trains.
“Deciding on scaling-out versus scaling-up is not a binary decision,” states Mark Santos, associate director of commercial development at Lonza. “There are various scenarios and risk profiles that will result in pros and cons to both approaches.” In sum, Santos continues, all the contingencies suggest that decisions about scaling should be shaped not only by commercial concerns, but also by product characteristics.
Scale-up may be preferable in the production of various biotherapeutics. An example cited by Santos is the manufacture of antibody therapies. Antibody therapies are low-titer products that need to be administered in large doses over a long period of time. Consequently, as Santos advises, the manufacture of these products “would most likely need to be scaled-up,” that is, the antibody therapies would need “to be produced in large-scale bioreactors.”
In addition, scale-up is more suitable for some cell lines than others. “NS0 cells,” Santos points out, “need cholesterol-containing feed, which results in adhesion issues in single-use technology, slowing growth rate. In such scenarios, it may be beneficial to leverage a stainless-steel production platform to minimize impact on growth.”
In contrast, mesenchymal stem cells (MSCs) are less well suited to scale-up. MSC-based therapies can undergo only a limited amount of cell fold expansion before their therapeutic effect begins to diminish (Front. Immunol. 2016; 7: 504). For products such as these, running multiple small bioreactors would be preferable to a large single reactor.
In cell therapy, product characteristics directly influence which scale-up approach is used, says Thomas Heathman, PhD, business leader, North America, Hitachi Chemical Advanced Therapeutics Solutions. “The decision to scale-up or scale-out,” he elaborates, “is mainly based on whether the cell therapy product is allogeneic or autologous in nature.
“If the product is autologous, that is, if the cells are taken from a patient and given back to the same patient, then the manufacturing process must be scaled-out to meet an increasing demand. If the product is allogeneic, however, then there is an opportunity to scale-up the manufacturing process.”
Most immunotherapies—including CAR T-cell therapies—in biopharma industry pipelines are autologous, indicates a report (“Medicines in Development for Cell and Gene Therapy”) issued last year by the industry group PhRMA. And at present, most of these immunotherapies are made using scaled-out processes.
Whether this changes as the cell therapy sector matures and processes become more standardized remains to be seen, but research in this area continues. “There are several companies exploring the potential to develop immunotherapies,” notes Heathman. “The manufacturing processes can be scaled up in an approach that is far more akin to traditional biopharmaceuticals and that has the potential to provide much lower-cost cell therapy products.”
Cost is another obvious consideration. The challenge when choosing to scale-up or scale-out is understanding how to compare costs, Heathman says: “In general, the overall cost to perform process runs in a scale-up manufacturing approach is higher than scale-out as the volumes of critical reagents are higher and manufacturing equipment is larger and more expensive.
“However, when considered on a per-product basis, scale-up processes are generally less expensive as the cost is divided by the number of products that can be generated per batch, whereas in a scale-out approach, the cost is multiplied by the number of products, as the number of products per batch is always one.”
Costs not related to the process also need to be factored in according to Dong, who cites reduced cleaning and energy costs as major advantages of the scale-out approach: “The running cost of disposable bioreactor bags is offset by the lower cost in energy consumption as scale-out does not require clean-in-place or sterilization-in-place operations and [does not incur] the higher facility and equipment depreciation costs associated with scale-up facilities.”
Although this observation is true for single-use-only facilities, it does not apply to hybrid plants, which require clean-in-place and sterilization-in-place operations, points out Santos. “Cleaning requirements for scaled-out processes in which all or most technology is disposable are much less stringent,” he continues. “The risk of cross-contamination is mitigated through disposal of all product contact surfaces between batches and products.
“This is not necessarily the case where disposable technology is applied to established systems where permanent equipment is also installed and requires full cleaning. This is a common occurrence in industry, so the gains [derived from less stringent] cleaning requirements may be limited.”
Consistency is another important consideration when deciding how to scale a process. In scaled-out processes, the use of multiple culture vessels introduces a potential source of variability that does not occur in scaled-up processes, which typically rely on a single vessel.
According to Santos, “Confirming consistency can require more steps when scaling-up compared to scaling-out. This is typically due to the equipment changes required during scale-up, as the disposable equipment typically leveraged at smaller scales reaches its limit.”
“Consistency variation between batches,” he continues, “can be mitigated [by developing] a robust technology transfer strategy and [using] on-line/in-line process monitoring with real-time feedback.”
The extent to which such variability should put a developer off using a scale-out approach depends on the product, according to Heathman. He cites cell therapies as an example: “For the manufacture of cell therapy products, the input material typically consists of apheresis or bone marrow from a patient or a donor. This means that the variation between manufactured batches for either scale-up or scale-out is high and that product consistency is much lower than it is with traditional biopharmaceutical products.”
For patient-specific products, consistency is even less of an issue. “Considering that for autologous products, the input material is typically being received from a patient in a critical condition, the quality and consistency of the starting material is much lower than for allogeneic products,” Heathman says. “Therefore, this impacts the entire manufacturing process, all the way to the finished product.”
This view is shared by Philip G. Vanek, PhD, general manager, cell and gene therapy strategy, GE Healthcare Life Sciences. “For cell and gene therapy,” he explains, “the decision is not driven by product consistency but therapy type. The difficulty in parallelizing unit operations to develop a ‘scale-out’ but pooled product for allogeneic applications is one of biological consistency across each unit of the operation.
“Since biological comparability is difficult to measure in even the simplest models, the challenge of pooled products is difficult to realize in cell and gene therapy from a regulatory approval perspective. Most important, there hasn’t been an economic incentive to move to parallelized production outside of the autologous workflows supported today.”
Scale-Out Turns Over a New Leaf
For the last 40 years, bioproduction has largely relied on culturing populations of single cells, particularly mammalian and microbial cells. Instead of growing single cells in cultures, in tanks or bags, it is possible to use plant-based expression systems. One such system is called called Hypertrans®. It was developed by Leaf Expression Systems, which asserts that the system represents a shift from single cells grown in cultures in tanks or bags to a sustainable green future where whole plants can be used as single-use bioreactors.
Many studies have already shown that eukaryotic bioprocessing results in reliable and predictable product quality. Following up on this point, Leaf says that in many instances, plant-based systems achieved therapeutic expression even though all other attempted systems failed to adequately express the desired protein. The company adds that the process is highly responsive, taking as little as 12 weeks from amino acid sequence to shipped drug substance, making it suitable for rapid early development of diagnostics, vaccines, and therapeutics.
According to Leaf, the impact on scale-out strategy is clear—the same process can be used throughout scale-out, with minimal management issues as processing moves from microgram quantities (from one leaf), to milligram quantities (from one plant), to gram and kilogram quantities (from multiple plants). Process economics can also be defined at an early stage, resulting in lower costs of goods than those available with traditional stirred-tank bioreactor technology.
As well as offering strategic advantages, plant-based expression avoids some of the widely documented challenges with existing biomanufacturing processes—including the need to deal with batch infections, leachables/extractables, and one- to three-day batch changeovers—even if the processes involve single-use systems and incorporate disposable lined bioreactors that permit the safe disposal of plastics.