Flexible Manufacturing

Flexibility has quickly become one of the most noticeable buzzwords of the bioprocessing industry. Understanding what constitutes a “flexible” process ranges from the simple application of one specific type of technology (e.g., single-use systems, automation, standard controls) to a somewhat extreme concept of a “throw-away” process. But whatever the definition, the factors leading to the need for more flexible approaches to biomanufacture are clear: Rapid, sometimes unexpected, changes in a company’s business situation and/or product portfolio (whether for patients or clinical trials) require that companies adjust to meet those challenges in very short time.

Achieving Flexibility

Although blockbusters are still dominating current markets, biopharmaceutical companies have been quietly shifting from the blockbuster model to diversified portfolios with big and small products on the horizon. To develop those products, manufacturers have been forced into a position of having to introduce flexibility into their manufacturing capability that didn’t exist before. Ran Zheng (executive director and plant manager at Amgen) notes three major factors in the industry’s general shift toward flexible manufacturing: a volatile market with potential changes in healthcare and Medicare policies in the United States and increasing competition from emerging markets; more diverse products (particularly personalized medicines) in the pipeline and commercialization, demanding facilities to be multiproduct capable with shorter change-overs; and advances in science and process technologies providing higher potencies and titers.

Business: “The flexibility we talk about today is a new way of operating the business,” says Parish Galliher (founder and chief technology officer of Xcellerex). Galliher points to three major business factors contributing to the current need for flexibility the result of several economic, political, regulatory, and market pressures on the industry. First, the economic situation is requiring companies to do more with less money, less cash, and less economic risk. “Manufacturers have learned they need to invest in manufacturing assets much more carefully. Historically, when a company made a decision to build a plant or add capacity, it was committing large amounts of nonproductive capital for many years. Meanwhile, the plant was being built for a drug whose clinical and commercial success was still unknown. That is an enormously risky proposition that few companies are still willing to consider.”

To further mitigate such capital investment risk, Xcellerex added another dimension of flexibility to its FlexFactory biomanufacturing platform. Galliher adds, “Our CMO team can operate a FlexFactory line for a customer during the early, riskier period of a drug’s clinical and commercial life, allowing the client to avoid the capital investment altogether. When appropriate milestones have been hit, we can then ‘TransPlant’ the entire operation from our facility to the client’s, enabling them to take control of a prevalidated line with pretrained operators. This essentially takes the capital investment risk out of the equation for a customer by bridging over it.”

Galliher says the second factor is that companies can’t take the risk of the totality of the capital cost. There’s less cash available from venture capital groups and the investment community, and more companies are practicing cash preservation. The shift to smaller, targeted end-markets for drugs also contributes to this scaled-back appetite for investment. “They want to install less and yet have more capability. So spending less, doing it in less time, and not risking so much capital is forcing a more agile and less complex investment. It’s really forcing the need for flexibility such that even though they’re spending, they’ve got the kind of agility they need for their drug pipeline.”

The third factor is the increasing competition in the industry. Biosimilars are making their way to the market, and innovator companies are under increasing economic pressure to compete on price per dose. As a result, they have to be able to repeatedly modify or set up, produce drugs, more quickly and with less investment, less effort, and less labor than ever before.

“The only way they can do that is by designing in flexibility,” Galliher says. Outsourcing is another way to achieve flexibility without committing the capital and resources to produce internally. But there are limitations to that, he cautions. “It can become complicated and expensive. So you might spend less on capital to outsource, but most people would agree that it’s not cheaper to outsource on an operating cost basis because you’re paying a third party its profit margin. So on one hand you have the flexibility of not having to build your own; but on the other hand, once you’re locked into a CMO, you lose some flexibility.”

Plant Operations: A flexible manufacturing space allows a facility to operate at different scales and/or produce multiple products without too much cost and effort in change-out. “We’ve had customers tell us that one day they’ll get a request for running a preclinical study and need 30 grams of protein, and next day they get another request for a different molecule for 1,200 grams of protein for stability studies and clinical studies,” says Mani Krishnan (director of Millipore’s Mobius single-use product line). “They will have huge swings in requirements for what molecule they need to make and how much of that molecule they need. Even the processes may be slightly different.” Krishnan says some customers may be running a traditional three-column process, but then require another column for additional purification that a particular molecule requires. “That is the flexibility customers are looking for, particularly for molecules in early phase clinical trials. That flexibility is much easier for customers in a single-use setting than in a traditional stainless steel setting.”

Processing: For Barry Holtz (cofounder of G-Con LLC) flexibility means that an operator can use multiple upstream and multiple downstream processes and convert different systems to adapt to those processes. For example, his company uses plants upstream in its Project GreenVax collaboration with the Texas A&M University system. “If I want to then be able to make a vaccine antigen or a virus-like particle,” he says, “the question becomes, ‘ How fast can I switch to a second process?’”

A flexible process can respond to challenges such as a new project that unexpectedly needs to be put through clinical trials or a project that comes from an acquisition, partnership, or merger — all without requiring many complicated adjustments in manufacturing hardware and setup, major reinvestment, major waiting times, or changes that need regulatory approval before the challenge can be addressed. “The world is full of those challenges and changes in what you might have planned,” says Günter Jagschies (senior director of strategic customer relations, biotechnologies R&D at GE Healthcare Life Sciences). “If a project failure completely changes the scheduling of everything you had planned, that should not create a problem. You never really know which of your molecules will be successful in the clinic.”However, he explains, “It doesn’t mean is you are like a flag in the wind. You’re not flexible in the sense that you’re careless. It’s still with stringent GMPs and very responsible.”

The QbD Link

The FDA’s quality by design (QbD) initiative emphasizes the achievement of product quality through process understanding, monitoring, and control. Manufacturers identify critical process parameters (CPPs) and the direct effects they have on quality and other product attributes, which helps define a particular “design space.” The bioprocessing industry has recently begun to adopt this approach, although the bridges between bioprocess conditions, product quality, and therapeutic effects are not yet well established. Unlike the general consensus just a few years ago, however, most manufacturers now believe that QbD for bioprocessing is a real possibility. “When you take biotechnology apart objectively, it’s not like there are 10,000 processes or a great revolutionary new bioseparation technology every day,” says Holtz. “They are all unique, but they all use the same fundamental building blocks such as chromatographies, membrane processes, and bioreactor processes.”

“One reason bioprocessing was not flexible in the past is that the level of knowledge was too low and uncertainties were too many,” says Jagschies. “You cannot really achieve flexibility without knowing your processes well.” Just a few years ago, bioprocess flexibility increased safety risks. Now, the QbD concept ensures process understanding. “Once you have that,” he goes on, “then you can consider flexibility, and you can make changes without going back to the regulators — within certain limits, at least. It is tightly linked: Flexibility is linked to process understanding, and process understanding is generated in an approach such as QbD.”

James Blackwell (senior consultant at BioProcess Consultants) agrees: “Flexible manufacturing is consistent with FDA’s current QbD initiatives. There is a real link between QbD and what can be done in terms of flexible manufacturing. In a flexible manufacturing environment, once you have defined a design space, then you would have the flexibility to make process changes within that design space with no prior approval from regulators. QbD focuses on critical quality attributes and the desired product quality profile. Each major step or unit operation in a process has an output that can be defined. Once you’ve characterized that for your process, then you can have some flexibility in how to achieve the output for those unit operations, including with different configurations or at different scales.”

Leaders in adopting QbD right now are the larger companies that have been involved in various aspects of modeling and thinking processes through. Because of that, there may be a perception on the part of smaller companies that adopting a flexible, QbD-based approach will greatly increase expenses and may become more of a regulatory burden than a benefit. As such companies adopt these precepts, however, they can improve their performance and efficiencies as well as the quality of their products in a way that makes good business sense. “There is enough of a history of these ideas in other industries, and there has been enough discussion of it — almost a decade at this point,” Blackwell says. “I don’t think it’s a question of whether it will work; it’s a question of what’s it going to cost and what will it mean for each individual company.”

A Flexibility Tool Kit

The technical tools that will enable flexible processing include single-use systems, centralized process control, automation, and in some cases, modularity.

Single-Use Technologies: Although a well-designed traditional (stainless steel) facility can provide flexibility, the goal is often better facilitated with the proper implementation of disposables. “Single-use makes the flexibility envelope larger,” says Millipore’s Krishnan. “It expands the operational window by providing more freedom.” Certain unit operations (e.g., pilot-scale filtrations) are well suited for flexible, single-use technologies. Krishnan’s company recently designed its Mobius FlexReady system comprising single-use bags, filter assemblies, and tubing manifolds that can be configured to accommodate various scales. “Essentially, we’ve taken elements from the hardware design and applied that in single-use technology. It is not tubing slapped together with a bag and a filter. A lot of design effort has gone into it. We spent a lot of time making sure that we are not compromising processing efficiency.” Certain basic building blocks are available, and operators can build on them and assemble various configurations as needed. If 200 L of one buffer needs to be filtered, an operator clicks a 200-L bag in place, whereas for 1,200 L of another buffer, the operator clicks six 200-L bags with a different filter on a manifold. “That level of flexibility is very difficult to achieve in a stainless steel system,” says Krishnan.

Millipore’s objective is to encompass a whole bioprocess for a recombinant protein — from a bioreactor to the final bulk product — in a fully disposable flow path. Last year, the company launched four Mobius FlexReady systems (and it plans to launch more this year) that accommodate some operations from bioreactor to final bulk using completely single-use technologies. “Once you have a fully disposable flow path,” Krishnan explains, “you can consider a manufacturing facility where there are no CIP [clean-in-place] or SIP [steam-in-place] drops. That certainly makes you pause and take a look at how facilities need to be designed in the future.”

But deciding just how much single-use technology to implement requires some consideration. “Certainly single-use components and equipment provide one way of creating flexibility,” says Jagschies, “but there’s more to it than that. In theory, today you could probably build an entire single-use manufacturing facility; but in reality, flexibility would then need to be balanced against economical considerations, such as the cost of disposable units. If you run many batches of the same product, it might be better to use stainless steel units.”

But Krishnan believes increases in efficiency with single-use technologies should be considered as cost reductions. “If a system is designed efficiently such that someone can run a process, get a very high yield from the system, and spend less time installing devices and flow paths into a single-use assembly — that increases efficiency and a cost reduction. The idea is not to be cheaper; the idea is to be more efficient from a processing standpoint.”

And disposable equipment is hampered by scale limitations. Although 2,000-L bioreactor units exist, Jagschies notes that they are not yet really proven. Moreover, he says, “Sophisticated sensor technology has not been developed for disposables yet. And there is still a little work to do in proving that a disposable bioreactor works exactly in the same way in all cases as a stainless steel bioreactor would. Sometimes it’s just the experience people don’t have yet. It will also depend also on what kind of production cells you have: Mammalian cells obviously work (they are the most well tested), but other types of production cells that produce higher cell masses (such as yeast or E. coli) are not as well tested yet in disposable equipment. There are still gaps in experience with this.”

A MANUFACTURER’S PERSPECTIVE

During a discussion with BioProcess International, Ran Zheng (executive director and plant manager at Amgen) outlined some key aspects of operational flexibility from the perspective of a major biopharmaceutical manufacturer.

“Reconfigurability and modularity are key features of a flexible facility. Reconfigurability allows quick change-over while maintaining CGMP compliance regarding flows and segregations. A modular facility can be easily expanded, contracted, or relocated to anticipate changes in production demand. Together these features help reduce cost through maximization of asset and operational efficiencies. A modular facility can replace conventional design and construction and is a trend to watch.”

Reliability and Standardization of Single-Use Systems: “Single-use technologies provide great operational flexibility and efficiency. Combined with modular facility design, their portability enables reconfigurability of production space. Reliability of single-use systems and associated consumable parts ensure robust operations and consistencies of process and product quality. Standardization of disposable parts must be well balanced. Whereas customization brings some convenience to customers, the lack of standardization and interchangeability is likely to reduce flexibility and increase cost. The industry needs to work closely with suppliers to develop a coherent strategy to improve reliability and standardization of single-use systems and thus reduce costs.”

“Automation and information technologies help achieve high levels of operational flexibility and productivity. Currently the application of these in biomanufacturing lags behind other industries. The integration of process control, monitoring, and data management systems will turn manufacturing facilities into information centers, create a feedback loop to the design space for processes and products, and help with QbD and continuous improvement. Wireless technology can offer tremendous freedom on production floors, from control and monitoring to electronic documentation and raw material reconciliation. There are some great opportunities to be captured in this area to further enhance operational flexibility.”

Flexible, High-Performing Workforce Model: “To achieve operational flexibility, the workforce model must be compatible with a company’s production requirements, facility design, and operating philosophy. The workforce model for large-volume, stable-demand, single-product manufacturing is unlikely to fit multiproduct facilities with unpredictable demands. And job requirements for manufacturing staff in flexible plants will have a much stronger emphasis on technical skills, reliability, and adaptability to a changing environment.”

Sustained Operational Excellence: “As a facility’s flexibility increases, so does the complexity of operations. Process simplification, waste reduction, and error proofing become even more critical to handle different products, frequent change-overs, and varying lengths of manufacturing campaigns.”

There is a lot of discussion at most major biotech companies about the extent to which the industry will adopt single-use technologies for commercial manufacturing in the future. And some industry experts point out that not every flexible operation involves disposables. “If you look at some of the largest MAb manufacturers,” says Krishnan, “they can produce perhaps as many as five proteins within the same facility. Contract manufacturers also manufacture multiple proteins within the same facility, and they’re not all single-use based.”

“I think there will end up being no single answer,” says Blackwell. He points out that some companies will adopt certain philosophies internally and stick with them. Others will be more resistant to thinking about disposables and what the benefits might be for them, so not everyone has adopted or will adopt the opinion that the industry should move toward having as much single-use technology as possible. “Part of what is not understood is the real true economics of moving to all single-use technologies,” he goes on. “Nevertheless, it’s clear that for a smaller-demand product, single use would make more sense. Once you get to extremely large volumes and demands, then it makes less sense economically and may not be operationally feasible. So the transition of where it does and doesn’t make sense isn’t clearly understood in the industry. “Blackwell believes the industry should conduct benchmarking studies and share data to develop a broader consensus of how these technologies can be used going forward. “As with automation, compatibility has been a huge problem,” he says. “Bringing in and integrating single-use technology must involve making the different technologies compatible with each other and having standards to make connections.”

Process Control: The ability to control all unit operations within a flexible design is critical. Certainly a process that must quickly respond to changes in scalability, configuration, and capacity must either comprise equipment with open architecture controls or be able to work with a single control platform for any configuration. Sartorius Stedim has developed a version of the latter. At the heart its FlexAct system is a central operating module (COM) platform for controlling various preconfigured unit operations. In its basic “cart” configuration, a touch-screen operator interface runs digital control software used across the platform (for bioreactors, crossflow systems, and so forth). That’s prewired in to communicate with a variety of sensors, scales, and a Watson-Marlow 720 pump to control flow based on different control loops that can be designed in. Currently, the system has been configured for buffer preparation (≤1,000 L), with the prewired operator interface used to communicate with pumps, scales, pH meters, and temperature probes. Future configurations and process analytics are currently under development for viral inactivation and cell harvest operations.

Although each operation has its own specific needs, flexibility is facilitated by a single FlexAct COM unit. “The FlexAct COM system works with any of these operations,” says Paul Priebe (director of fluid management technologies at Sartorius Stedim), “depending on the sensors and single-use consumables that are used—whether for buffer or media preparation, ultrafiltration, diafiltration, or mixing—what style of pump is used, or additional bags.” For example if you need to run a postuse integrity test of filters used in your operation, the system provides room and a well-designed space for a Sartocheck 4 integrity tester from the same company.

In developing the FlexAct platform, Sartorius Stedim wanted to enable the same level of process automation control and monitoring with single-use surfaces that a company could implement in multiuse product surfaces. “In the early days of single use, a very common application would be to filter into a sterile bag to hold media until the time to use it,” says Priebe. “However, one of the most basic parameters in a normal flow filtration process is to watch the pressure differential across the membrane filter—and in many cases, because of the logistics, that parameter was not monitored during the filtration process involving single-use components.”

Modularity and Automation: Like flexibility, modularity is quickly becoming an industry buzzword. Günter Jagschies says, “Modular to me means that you use the same equipment in different processes. The equipment is designed so it can be used in many different scenarios, the connectors are all standardized, and certain flow rate ranges are available that fit to most process steps. And then you have an automation system that can handle very many different processes without having to be completely reprogrammed or redesigned. That is modular.” Examples include a purification system module and a filtration module. “That creates flexibility because you don’t have to go back and buy new equipment, and you know from the beginning that you have everything you need. You don’t need to go back to the planning process, either. And that is more widely spread than one would think within one company. People are starting to standardize things more and modularize things more. But I’m not so sure you can say that for entire industry yet.”

G-Con has developed modular, portable cleanroom pods that combine single-use systems with automation. These pods can be hooked up to steam all other central services. “When we sell a manufacturing solution to a customer (e.g., GreenVax),” says Holtz, “we sell a process that is already prevalidated through operational qualification and already integrated into a master process suite using the usual integration software process control system.” Expansion is possible by configuring the independent pods in parallel, rather than incrementally. Each pod has its own down-in/down-outs, its own redundant air conditioning, and a complete filtration environment.

“You literally bring them in and hook them up to a hallway, to a clean corridor, and you’re ready,” Holtz explains. “Moving them is also easy because they have air bearings so that with compressed air, two people can move a 20-ton, 42-foot by 18-foot cleanroom into the building with one hand each.” The pods can be moved into a facility’s grayspace that has nominal air conditioning. They cool the air further and get it to normal cleanroom operating temperatures and control humidity and filtration. They are set up for vapor-phase hydrogen peroxide (VPHP) sterilization. Once a process is run, the pods can be disassembled, refurbished, and recommissioned.

Xcellerex also has combined modularity in its Flex Factory systems, as well. Unlike the G-Con pods, these systems collapse into a minimal microenvironment around manufacturing equipment. Each module can be tailored to the equipment it’s meant for, widening or lengthening to accommodate different machinery. These modular systems are automated to track performance during manufacture. “We’re trying to maximize the grey space of a facility,” says Galliher, “because grey space is cheaper to build, and you’re not as restricted from a regulatory standpoint to work in a grey space. This keeps a clean room space to a minimum area. That also frees up operators to spend most of their time in the grey space and not have to go in and out of clean rooms to do their work.”

From an automation standpoint, the Xcellerex systems provide real-time information about how manufacturing is going, how well it’s being done, and whether deviations occur. “The idea is to be able to set up a run, do a batch or batches, shut it down, and switch to the next production,” Galliher continues. “Automation helps us practice flexibility in the true sense, which is to make it now, know what you’re making, know how well you made it, and then move on and be able to do that freely.” A different drug can be run in each machine because each is separated from the other(s) by its own module, and these modules can then be operated simultaneously. Because Flex Factory modules have their own heating, ventilation, and air-condition (HVAC) systems, the building becomes basically a clean shell, without SIP or clean-in-place (CIP) utilities. “The HVAC system in the building is about one tenth the size that it normally is in manufacturing areas,” Galliher says. “This means the building is about 40% smaller than a typical facility that has multiple cleanrooms for manufacturing different steps. All those cleanrooms have to be serviced by corridors and gowning rooms. And when you add those up, they add up to 35– 40% of a plant’s footprint.”

Process and Equipment Design: Flexibility is certainly possible without single-use technology. Sometimes it is a matter of proper plant design. Blackwell reports, “I’ve seen transfer panels within facilities, for example, that offer lots of flexibility in terms of operations, and your segregation and flows can also give you a lot of flexibility.” Traditional stainless steel-based designs, of course, can be used to run multiple products with stringent cleaning validation protocols. Production at various scales, however, becomes more problematic with traditional hard-fixed stainless steel reactors. “It’s certainly possible,” Blackwell says, “but I think it’s more difficult to get flexibility in these operations.”

Flexibility can be created, however, by designing a process in a smarter way. Jagschies explains, “You can, for instance, reduce the number of steps, simplify the analytical effort, and avoid the intermediate steps that don’t contribute to purification or processing of the product but are just more adjustments that happen in hold tanks. If you remove those from your process design, you can greatly increase flexibility.”Shortening the time for any individual step also creates additional flexibility. “If you could shorten a longer step,” he says, pointing to bioreactors as a good example, “then you would actually reduce the timeline. That would give you the flexibility to perform other tasks in the time you saved.”

Currently, monoclonal antibodies require 50- to 500-kg or even ton-scale capacities. Of course, not all future molecules will be MAbs. Typical recombinant proteins, for example, fall in the range of a few to perhaps 10 kilograms. This fact, combined with increasing product titers and bioreactor productivities, is leading many to believe that huge facilities with tens of thousands of liters of bioreactor capacity will no longer be needed for most if not all future products. “If we look at the whole biopharmaceutical range of drugs,” says Jagschies, “then it is probably not necessary to have such huge bioreactor capacity in the future. So facilities will likely be much smaller and much more flexible also as a result.”

Equipment engineers also have been paying attention to the industry’s needs for smaller-footprint modular units. For example, at a major industry conference last month, ATMI Life Sciences announced the commercial availability of its PadReactor bioreactor based on PadMixer technology. The space-efficient square tank contains bags composed of multilayered, animal-derived component free, virus-tested, gamma-irradiated film — the same one used in ATMI’s modular mixing and storage systems. The streamlined bioreactor was designed with flexible factory setup, scale-up, validation, and portability in mind, says JeffCraig (ATMI’s director of marketing).

“One main driver of this technology (in addition to the advantages of single-use systems) is engineered flexibility —when you need to design or redesign a facility in response to market conditions or process variation. These technologies are faster to acquire, install, and validate than stainless steel but also offer compelling technical, operational, and economic advantages.” This modular bioreactor is scalable from 5 L to 1,000 L, with footprints ranging from 20 cm × 21 cm to 1.6 m × 1.5 m, and is compatible with and connectable to ATMI single-use mixing storage systems. In a square tank, corners serve as natural baffles, so less mixing power is required. “Anyone with engineering perspectives knows that a square mixes better than a circle with less power,” says Craig, “and we take advantage of that. Lower shear means happier cells. The mixing element is a shaft suspended in the middle of the square tank that rotates like a swizzle stick with a paddle and sparger on its end. The system is designed with either an integrated Pierre-Guerrin or open-architecture configuration for any other controller.”

According to Craig, the latter allows a system to be integrated with other systems in place or preferred by users. “We find that small biopharmaceutical companies that are concerned with time-to-market in making clinical materials have aggressively adopted these technologies. We also have found that early adopters—the preclinical labs of large multinational companies (including those that produce vaccines) — are also using this technology. Their goal is to pass scalable technology from development to GMP large-scale production. Several years ago this may have not been the case, but that’s definitely where we are today.”

The Economics of Bioprocess Flexibility

How flexible can you afford to be? As with the adaptation of disposables and QbD, implementation of a flexible strategy is driven by the economics, size, and capabilities of each company.

Quantifying the monetary cost and savings for a single-use process or facility can be a difficult exercise because many factors involved in the cost of stainless steel processes/factories were not captured because originally there was no alternative. One example is the cost of making water for injection (WFI). “The savings in WFI is quite significant with single use,” says Jerold Martin, senior vice-president of global scientific affairs at Pall Life Sciences, and Chairman of the Board and Technology Committee for BPSA. “Although it’s relatively easy to capture the savings in volumes of water required, many companies find it difficult to determine the actual cost to make WFI. It is quite clear, though, that there are savings to be had, and decisions for single use are being made even when cost savings analyses are unresolved.”

Flexibility will improve a company’s economics, but for a given manufacturing operation, sometimes reducing initial capital cost is more desirable. If flexibility is mainly implemented with the integration of disposables, for example, there is a definite trade off. Although the initial capital cost is lower because a single-use facility is less expensive than one based on reusable equipement (with fewer cleaning and utility requirements, less associated validation, and shorter timelines to initiate a functional facility), the economic trade off with single-use technology is going to be higher direct and variable costs per batch than with a stainless-steel–based process due to higher consumables cost. Blackwell says, “Nonetheless, flexibility and the capital savings of single-use technologies are compelling to people. But where that balance is in terms of larger-scale, multiproduct operations—and how you do that exactly in an optimal way—isn’t really clearly understood right now. That’s why there is a lot of discussion about how we do it, the best way to do it, and the economic and technical risks involved.”

Priebe agrees: “Facility requirements for clean steam, the need for WFI [water for injection] and CIP equipment is reduced or eliminated when a facility integrates a high percentage of single-use product-contact surfaces. However, this is not to say the industry will go to 100% single use because there are still processes that are cost prohibitive and will stay with multiuse systems.” Regarding the belief that facility HVAC requirements will be significantly affected by the use of microenvironments, he responds, “That is a new idea, and we’ll see how that goes. I think most conservative types will stay with the same type of traditional facility design from an environment standpoint.”

Pilot and Small-Scale Companies: Disposables become impractical beyond a certain scale. Manufacturing with single-use technologies is not suitable for processes involving 20,000-L bioreactors, for example, or larger. Nor are disposables likely to be practical for dedicated facilities that commercially produce the same molecule 50 times a year. But for the early clinical-trial molecule stages, batch sizes are smaller (producing 1 or 2 kg of protein or less), and some bioprocessors at that stage may be producing multiple molecules in the same facility.

Small companies producing molecules that show promising preclinical or clinical trial results are sought by big biopharmaceutical companies for acquisition or in licensing. “At a pilot or clinical production scale, the horizons on drug development are very long because they don’t even know which molecules are going to be part of what they will be asked to produce,” says Priebe. “Flexibility allows them to make an investment without knowing what molecule they’re investing in. That is a big driver for flexibility in clinical manufacturing scale operations. They simply don’t know what molecules specifically or what class or what processes are going to be used to make that molecule.”

Smaller companies may be turning to other, more economical solutions than installing flexible or disposable systems or even setting up new facilities. “Many small companies don’t have their own facilities,” says Jagschies, “but they can get the capacity they need from either contract manufacturing organizations (CMOs) or maybe make a deal with one of the larger ones. I think companies will increasingly look at those options of sharing existing installed capacity.”

Blackwell agrees. “Small companies are not going to be thinking ahead to flexible manufacturing. Many are too focused on getting their early product approvals through clinical milestones, so they tend not to be thinking that far ahead. Certainly some of them have been adapting and developing in using single-use technologies for their needs for flexibilities in clinical manufacturing, but they’re not drawing the links yet to QbD.”

Midsized Companies: Flexible manufacturing by means of disposables and even modular units may be suitable for some midcap companies. Holtz provides one scenario of a typical midcap biotech company that, for example, has a phase 2b trial going in a cancer vaccine and finds that its data safety monitoring board sees merit in the data so far and grants authorization to keep going. “The first thought you have,” he explains, “is how to figure out how to manufacture the phase 3 and launch levels. You want to launch from whatever facility you make phase materials in. So if you’re in research trying to work, you’ve got a year to get a permit, let alone build a greenfield site that is flexible and redundant. By the time you build that site and build a hardwall facility and validate it, it’s probably about two-and-a-half to three years. During all that time, your opportunity costs are increasing because you can’t serve your patients until you’re ready to go. So you have your entire organization twiddling its thumbs for that period.” It is for such scenarios that a company can go into a “flexspace” by leasing a very high-quality warehouse and treating it as gray space to build into. “So while you’re building your flexible manufacturing pods, you’re validating your process in a nonlinear manner. Your cleanrooms are up and running and able to be validated before they are even run into the GMP envelope. You can do all that in parallel, and it drops a three-year project to a one-and-a-half or one-year project, depending on how complex your process is. It’s less capital expense and less opportunity lost,” says Holtz.

Big Biotechs and CMOs: According to Blackwell, those who will be the real pioneers in terms of what flexible commercial manufacturing means will be the larger players: established biotech companies and leaders in the field, as well as larger CMOs. “I think they benefit from flexible operations as long as the process output is the same,” he explains. “I could see that for a company trying to establish manufacturing for a new product in a new facility, the most economic decision might be to start with some initial (smaller-scale) capacity with stainless steel bioreactors and achieve lower cost of goods from direct or variable costs. However, there is always uncertainty around market projections and penetrations of a new product into the market. Because of that uncertainty, you wouldn’t want to build a facility of fixed assets that is too large for your needs. So it would be desirable to design a flexible facility with single-use bioreactors. Under this scenario, your increased consumable costs would be balanced by a reduction in direct costs per batch, while your overall risk was reduced.”

The challenge is somewhat different for big biotech companies that have highly variable facility use rates. “I think the challenge is to find really good ways to use these facilities in the most economical way,” says Jagschies. “A company may make other changes to its facility to increase flexibility than would a company that does not have a facility. I’m not sure any of the large companies with overcapacity right now will build additional capacity anytime soon.”

CMOs, however, are always seeking capacity for many different projects. Such companies have been thinking about flexibility for some time already. “Flexibility has always been part of their lifeblood,” says Jagschies. “They always have to respond to their customers’ requests quickly. They just haven’t called it that because flexibility has been a buzzword now for only about a year or two.”

Although most large companies have enough capacity and will not soon build new capacity, they might rework existing capacity or facilities. Many smaller companies are trying to find partners for manufacturing. “The exception to that, of course,” Jagschies reminds us, “would be companies in developing markets (such as India or China), where they will still build facilities because they want to have production where they are located, and they also think they can do it cheaper in those countries. So for them, the preferred option is to build new capacity, new facilities, even if they could go to someone or to a CMO, but they want to be local.”

About the Author

Author Details
Maribel Rios is managing editor of BioProcess International, 1574 Coburg Rd. #242, Eugene, OR 97401; 1- 646-957-8884;[email protected].