SANOFI AVENTIS (WWW.SANOFI.COM)

On Tuesday, 21 April 2015, Eric S. Langer (managing partner at BioPlan Associates) chaired a midday roundtable titled, “Deciding on Single-Use vs. Stainless Steel Bioprocessing Strategy: What Do CMOs Know That Biopharmas Don’t?” Langer brought together these industry experts to discuss choosing between stainless steel and single-use technologies (SUTs) for different applications:

• William Hartzel (director of strategic execution at Catalent Pharma Solutions)
• Steven Perry (vice president of technical operations at Cook Pharmica)
• Joanna Pezzini (bioprocess engineer at MedImmune)
• Daniel Vellom (senior director of global technology innovation at Sanofi Pasteur)
• Sue Behrens (senior director of process technology at IPS-Integrated Project Services, Inc.).

William Hartzel (Catalent Pharma Solutions)
Hartzel emphasized defining goals and objectives first off, acknowledging that those do vary between drug companies and service providers. Each company must examine its philosophy and business models, then align its manufacturing strategies with them. A main focus for contract manufacturing organizations (CMOs) is providing speed and flexibility in multiproduct facilities for their customers. Hartzel pointed to cell- expression technology as an example. He described Catalent’s transition from working with transgenic cattle to cell culture in stainless steel reactors, then from those reactors into SUT. The latter was the more straightforward change. In 2008, the company began an expansion before an economic recession necessitated putting that on hold. When the project was reengaged in 2011, Catalent already had purchased stainless steel equipment.

Upon weighing the economics of energy, labor, consumables, containment, and flexibility, the company found that despite its investment in stainless steel, the capital that option would yet require (in infrastructure, support, and timing) actually made choosing SUT no more expensive. It would reduce risk and time in construction, and the company was able to sell its stainless steel reactors. “We were able to turn a building into a GMP revenue- producing facility in one year,” Hartzel said, “for a fraction of the cost of a stainless steel or fixed system.” The results provide multiproduct efficiency through closed systems, allowing the company to handle up to nine different products in parallel. Transitioning between customers is important to CMOs, but so is transfer of processes from customers. Hartzel pointed out that using SUTs facilitates in-transfer of client processes that have been developed in such systems, which is increasingly the case.

He emphasized working closely with SUT suppliers and treating them as partners. “Ultimately,” he said, “you are outsourcing the validation. You’re not buying bags, you’re not buying tubes; you’re buying validation packages.” So the supplier relationship is critical, especially in regard to change control and inventory management. For transitioning to SUT, employees may need specialized training.

Company goals change by development stage. At later points, when a company plans to run the same process consistently, building fixed assets then makes more sense. Flexibility and elimination of cross- contamination potential (as in early stages and multiproduct facilities) are major drivers for choosing SUT.

Steven Perry (Cook Pharmica)
Perry’s company uses SUT in buffer and media preparation. Buffers and media are prepared ahead and used as needed. Although the seed train mostly involves shake flasks, it also includes rocking and stirred-tank single-use bioreactors. In downstream processing, all mixing (e.g., low-pH hold, viral inactivation) is in single-use systems, as is in-process storage up to 1,000-L scale. Depth filtration, virus filtration, and sterile filtration are also all single use applications.

Stainless steel and reusable equipment is found in large-scale (2,500 L) production areas and 600-L scale-up plants. Cook Pharmica’s facility was built in 2004, when single-use bioreactors were still fairly new and limited to smaller scales. Centrifugation, harvest tanks, chromatography systems, and ultrafiltration/diafiltration (UF/DF) are traditional stainless steel, reusable components. Perry discussed the rationale behind his company’s hybrid approach, pointing specifically to schedule decoupling and risk mitigation.

Cook Pharmica accommodates a number of different processes as a CMO. Some use many different buffers and media; others are fairly straightforward, using just a few. Decoupling buffer/media preparation from the actual processing allows the company to level out its labor needs by scheduling those operations well in advance of when materials will be needed. This also aids in risk mitigation. Prepared buffers and media are sterile filtered and stored under appropriate conditions until needed. Decoupling those preparations from bioreactor operations, for example, is an alternative to preparing media in a separate tank and then filtering it directly into a bioreactor. It allows the company to test for bioburden, for example, before using the materials. If something goes wrong, the problem can be identified before they are committed to a process.

Perry then brought up some SUT challenges, particularly in inventory and supply chain management. A single-use strategy usually involves customized assemblies or bags, which makes working with vendors important. If a CMO requires many different configurations, lead-time issues can arise and inventory costs can be high. Managing supply chains and inventories is critical. CMOs must provide flexibility while controlling cost.

Vendor quality makes a difference. When users are “outsourcing” part of their system assembly and validation processes, they need to work with suppliers as partners. SUT companies need good quality systems in place. For example, a bag vendor typically tests its bags for leaks, but it also should test the (often custom) bag connections as well. Design limitations are another consideration.

Set-up can be a challenge. Operators need to be well versed in SUT and get enough practice using such systems. “If you don’t set up a bag correctly in a single-use mixer,” said Perry, “you can get bag folding, kinked tubing, and so forth.” As new types of connections and components are implemented, operators must practice and be tested before working in real process environments. Some new aseptic connectors, for example, require specific techniques.

Perry also highlighted leachables and extractables testing. Material contact can be an issue particularly in upstream processes. “We’ve seen some impact on cell growth in the past and have had to work around that,” he said. It is important to work with a scale- down model of each disposable system before implementing it at full scale.

Cook’s approach is to use well- established SUT with low risk — only where it truly provides an operational advantage. Continually evaluating single-use applications, the company is now considering larger single-use bioreactors. But operational details cannot be overlooked. Practice trials, engineering tests, and operator training are all important when implementing a single-use system. A CMO wants flexibility but needs to be somewhat risk averse. Reliability is important to its clients. Perry said that his company has chosen to stay with traditional reusable systems, so far, for centrifugation and chromatography. It will consider single-use options as they arise but currently is not “ready to make the leap” for those.

Joanna Pezzini (MedImmune)
Pezzini’s group at MedImmune is involved in technology transfer. Her team has evaluated a single-use UF/ DF skid for 500-L and 2,000-L GMP facilities and also at a 500-L pilot facility. The GMP operations run in primarily stainless steel facilities where the company is considering the switch to SUT; the non-GMP facility is more single use overall. Pezzini reviewed some drivers behind the decision to purchase single-use UF/ DF skids including the costs, benefits, and technology challenges.

The single-use skid provides performance improvements and flexibility, as well as automation. Capital cost is about 50–66% of that for a stainless steel system. And the team found that the single-use skid is nicely portable. Consumables costs must be considered, however. “We expect to break even with about 30 batches running on this single-use system,” Pezzini said. So the number of batches a company runs each year figures into the SUT–stainless steel decision.

Her team compared the costs of using reusable and single-use membranes, finding that the latter are slightly more expensive, but that the difference is made up by cleaning cost savings. It takes fewer batches at large scale to equalize costs. Eliminating cleaning steps provides significant validation cost savings. Pezzini’s group saved time in commissioning the GMP suite by performing all commissioning, testing, and automation work at the pilot facility. They found operation of the single-use UF/DF to be easy, with smooth automation.

Many challenges specific to the UF/DF skid can also be found with other SUTs. First, extractables and leachables had to be assessed. MedImmune had to rely primarily on vendor calibrations and certificates for instruments in this UF/DF system, which was an unfamiliar approach.

Kit reliability hasn’t been perfect, Pezzini says, “but it’s been pretty good.” SUT pressure ratings were lower than for a stainless steel system. Turbine flow meters aren’t accurate enough to calculate diafiltration volumes. Now using a floor scale to collect that volume and measure it, the team is actively seeking a disposable flowmeter that could fill that role. Pezzini’s group also found that bag mixing was limited to ≤20 L because bags rely on retentate flow alone for mixing. So the team is also evaluating different mixing systems (e.g., impeller-based).

Because the single-use UF/DF skid provides greater flexibility than a traditional stainless steel system, and because it can be moved easily, MedImmune was able to buy one skid instead of two for the 2,000-L scale. However, Pezzini cautioned, associated cost savings depend on use rates. Although some technology challenges remain, she feels that her team will resolve them.

Pezzini said that the SUT-or- reusable question may not be as different for CMOs and drug companies as some people think. Budget and development phases are more important criteria. “If an established company has a cleaning system in place, then it may not realize great benefit in going disposable.” But companies starting up new facilities or requiring more flexibility would find disposable equipment to be much more useful.

Daniel Vellom (Sanofi Pasteur)
Vellom discussed his company’s choice of SUT over stainless steel for some vaccine products. The vaccine industry takes a relatively conservative approach to new technologies because most of its products are administered to healthy people. With a diversity of product types, decisions regarding process technology must be made case by case.

The initial SUT advantage relates to capital expenditure. Rather than buying tons of stainless steel and installing clean-steam and clean-in-place (CIP) systems, a company builds cleanrooms with some basic utilities (e.g., gases and air flow). “You don’t have to validate and clean and continuously revalidate it,” Vellom pointed out. That represents an ongoing advantage. SUT product- contact surfaces come from a vendor, which shifts the responsibility of ensuring their readiness from the drug company to those suppliers. However, it creates new issues on the end-user side (e.g., leachables, extractables, and particulates).

“We’ve started to look at the concept of modular or mobile units,” Vellom said, pointing to G-CON Manufacturing, Inc. as a specific source.

Challenges: He said that SUT operations require special training. Bags, tubing, and manifolds are relatively fragile compared with stainless steel equipment. Operators accustomed to triclamps now have to carefully unfold and manipulate plastic bags and tubing. Vendors are learning how to better package and present their products for preventing complications. But in converting facilities, Vellom says Sanofi has run into some problems with storage space because disposable-component packaging has to be robust. So operators met with the vendor to discuss their respective needs.

“We were working with a bioreactor manufacturer, and it wasn’t clear to us or them how robust different pieces had to be,” Vellom explained, “and we had some intermittent failures.” Some processes were stretching the tested limits of certain components, and operators did not recognize where the weak points were. So he urged quality and materials management groups to meet with process development and manufacturing from the start. Each specialty has a perspective the others may not see. When asked about appropriate training programs, Vellom said that Sanofi brings suppliers in to do training. They know their products and how to handle them.

Another question related to manual manipulations. A stainless steel bioreactor is a fixed piece of equipment, often automated. SUT can be more complicated and require more manual labor for set-up and operation. A 2,000-L bioreactor uses a large bag with elaborate tubing. Putting things together correctly and breaking them down after use can be technically demanding. Operators need to be nimble, and standard operating procedures (SOPs) need to be specific.

Another participant asked about failure rates, suggesting that companies relying on stainless steel typically achieve 95% success rates with contamination as their principal risk. Vellom countered that Sanofi gets about the same with SUT — with good training.

Some participants asked about supply chain issues. Users depend on suppliers to deliver what they need when they need it, making supplier quality management important. Vellom said that gamma-irradiated components typically have a two-year shelf life, which aids in inventory management.

Sue BehrenS (IPS)
Behrens described a project her team has worked on with the US Department of Defense (DoD) involving advanced manufacturing for medical countermeasures. The goal was to build a facility that could make anything, at any time, in any amount. That “really is the kind of project that’s well-suited for single-use,” Behrens said. She went on to explain that people make a lot of assumptions when comparing SUT with stainless steel.

Speed: Some people boldly assert that SUT implementation is faster. When a company doesn’t have time for new process development after clinical material is made with SUT, it can’t then move to stainless steel. Behrens mentioned the examples of Ebola, anthrax, and influenza vaccine makers that need product ready to go in three to six months. They are looking to SUT to make that happen. But, she said, if a company already has stainless steel with CIP and SIP systems in place, then getting a process going in that facility can be pretty straightforward. Managers need to consider their investments, their companies’ business situations, and where their manufacturing processes stand.

Cost: Proponents say that SUT is less expensive, and Behrens asserted that initial capital investment can be lower and that facility size needs are reduced. SUT allows companies to delay some costs. But large-scale manufacturing can complicate matters, making reusable stainless steel equipment look like a better option. Managers have to make related decisions before they know the full story, so Behrens advised getting the perspective of marketing and regulatory groups to help understand where a given product is going.

Ballrooms: SUT facilitates open architecture, allowing for the ballroom concept. All operations can occur in one space, ostensibly with operators cross-trained to work throughout a process (from upstream to downstream and back). But that can be operationally challenging. Bioprocessing is complex. Each unit operation requires intensive training, and mistakes can be costly. So SUT doesn’t yet have full acceptance in large companies. Even as regulators accept such technology and the ballroom concept, the industry is still concerned about associated business risk. When millions of dollars in product cost are on the line, a problem in a ballroom facility can put an entire business at risk.

Flexibility: SUT is considered to be more flexible than stainless steel. As an example, Behrens pointed to the DoD project’s facility in Florida. It has manufacturing modules for upstream and downstream processing in a ballroom setting. Stainless steel infrastructure doesn’t offer the same kind of mobility.

Many vaccines do not survive shipping well, but some people who need them are in remote areas. Manufacturing facilities based on SUTs could be set up relatively quickly around the world (e.g., during a pandemic when borders must close to prevent further spread). However, Behrens said, some constraints come along with that approach. A process depends on what’s available from the SUT innovators. For example, if a process requires a certain ion-exchange resin that’s not available in a single-use configuration, then even getting a supplier to prepack columns requires validation (which takes time).

Equivalency: Many proponents assert that SUT is operationally equivalent to stainless steel. However, although SUT and reusable technologies can produce the same end product at the same quality, they arrive at that destination by different routes. The transition takes time and consideration. Behrens reiterated her warning about cross- training. Lean strategies may suggest cross-training in which an upstream operator could perform a check of a column downstream. But Behrens said, “You lose that ability to build expertise when you do that.”

Panel Discussion
That segued into a wider panel discussion, beginning with the question, “Are you generally pleased with the process performance of bio- bags?” Perry (Cook Pharmica) reported good performance with single-use bioreactors at smaller scales. Failure modes are different for SUT and stainless steel. With the latter, contamination usually can be attributed to a cleaning or component problem. But single-use components come presterilized. Connections are often an issue.

Vellom (Sanofi Pasteur) brought up robustness. “Our partners at Genzyme are working on extended, semicontinuous perfusion processes. They ran a 50-L single-use bioreactor for 160 days straight, stopping only because of the holidays. It just kept on going.” That demonstrated both robustness in both the components and how they are combined into a system.

The idea of continuous processing generated much interest, and one audience member asked about downstream hardware connectivity. “We spend a lot of time sorting out the connectivity issue,” Vellom said. As 21st-century terrorism has made vaccines for bioweapons a government priority, vaccine companies must move fast. Sanofi spent a lot of time assessing available technologies (e.g., tube welders) for their robustness and ease of use. Now aseptic connectors and disconnectors are making things easier. “The more you move to those kinds of built-in solutions,” he said, “the better off you are.”

View the full presentation video