On Wednesday, 21 June 2017, BioProcess International presented a panel discussion from 1:20 to 2:30 pm as part of the “Emerging Techniques and Technologies” session of its BPI Theater at the BIO annual convention in San Diego.

Left to right: Joshua Speidel, William Taylor, Diane Retallack, and Rachel Felberbaum (PHOTO BY LEAH J. ROSIN)

Moderated by Joshua Speidel (managing director of the commercial practice team at Latham Biopharm Group), this panel comprised Diane Retallack (senior director of upstream processing and intellectual property at Pfenex), Rachel Felberbaum (senior director of business development at Protein Sciences Corporation), and William Taylor (vice president of business development at Protalix BioTherapeutics).

Platform approaches to vaccine drug-substance manufacturing play a key role in accelerating the product development cycle time. The benefits are clear: Patients have earlier access to treatment, and product sponsor investments are derisked from regulatory and manufacturing perspectives. Vaccines present unique challenges for manufacturing platform development because of their diversity. They may be composed of inactivated whole viruses or bacteria, recombinant proteins, subunit particles, nanoparticles, or live-attenuated infectious agents, and they may or may not include adjuvants. Thus, a plethora of technologies are involved in their manufacture. This panel discussed practical implementation of a platform approach to vaccine manufacturing, considering when platforms would be appropriate, potential risks, key milestones in platform acceptance, and the long-term value of such an approach.

Discussion
Speidel pointed out that the panelists’ companies represent different stages in platform development.

How do you define a platform? All three panelists agreed on defining vaccine platforms based on expression systems. Felberbaum described making a recombinant baculovirus to express a selected antigen specifically. Protein Sciences uses one main baculovirus and universal cell line — US Food and Drug Administration (FDA) approved — for all products. To produce influenza vaccines, the company has an approved downstream process using the same purification steps with the same filters and resins to get pure protein. Taylor added that Protalix had to develop and scale up its process by building proprietary bioreactors to enable robust and scalable production using plant cells.

How did you decide that a platform would be only the expression system but not include downstream processing? Taylor said that the downstream aspect is protein specific. “You address that as one of the unique characteristics of the protein target you are trying to express.” Retallack said that her company’s main focus with a platform was to prevent compromises to protein quality (e.g., refolding and losing much protein during processing). However, Pfenex is always looking for ways to speed downstream development and lower the cost of production.

How has the platform influenced your selection of the antigens that you include in your vaccines or products? To gain an advantage in fast production, Protein Sciences does restrict the antigens to those already known and defined. Besides working on influenza, the company is making a Zika vaccine and already has made products for Ebola, severe acute respiratory syndrome (SARS), and rabies. In each case, it is generally the surface glycoprotein that is purified.

Pfenex makes its own vaccines and provides a platform for other companies with challenging proteins. Although focused on already identified proteins, the platform does lend itself to discovery as well and can produce different proteins in active form for screening. Protalix primarily focuses on therapeutic proteins but also is looking at virus-like particles that plant cells could produce.

How early in the process of selecting targets and products for a platform do manufacturing scale, protein stability, and so on make a difference? Taylor said that if a candidate cannot be successfully scaled up, then Protalix wants to know that early. Some classes of proteins have proved difficult to express on traditional platforms but might work well on the plant-cell platform. If that can express difficult proteins, then it could have a competitive advantage.

Felberbaum explained that one of the attractive features of the baculovirus–insect-cell platform is its scalability. Bioreactor size is the limiting factor. Protein Sciences produces Flublock vaccine at 21,000-L scale. The company’s scientists started with an academic insect cell line, optimized it for manufacturing, and developed from that a unique cell line (ExpressXXplus). Those cells have thicker cell walls than the originals did and are robust for scaling up. Protein Sciences focuses on tried-and-true processes so that clients can make a transition seamlessly to current good manufacturing practice (CGMP) compliance.

At Pfenex, Pseudomonas fluorescens was selected because it can make proteins at very high titers in soluble form (and it already had been used). As different processes are developed for the target proteins, developers are constantly thinking about the next step. The final strain selection is made with the downstream process in mind.

Did you have to develop critical reagents or analytical tools when developing your platform? The answers included host-specific cell assays, high-throughput techniques, release methods, and glycan analysis specific to plants.

What is the greatest regulatory hurdle associated with developing new products and new platforms? For Protein Sciences, Felberbaum shared that it came down to getting the FDA comfortable with a novel platform. The company developed Flublock vaccine, performed all related clinical studies, and submitted a biologics license application (BLA) in 2008. The FDA reviewed that, all the clinical data looked good, and there were no questions about it. Then the company moved on to focus on chemistry, manufacturing, and controls (CMC). The safety bar was high because influenza vaccine is given to healthy adults for flu prevention. The FDA had to be cautious in its approval of a new product made on a novel platform.

It took five years of working with the FDA for Protein Sciences to get approval. The safety data were fine, but the concern was whether something could be hidden. One assurance required by the agency was for the company to show that no known or unknown viruses were present in the insect cells. But Felberbaum asked, “How do you prove that you have no unknown viruses?” Regulators wanted more sensitive assays, then finally gave approval after five years. This was a major hurdle, but now the company has a universal cell line that is qualified and that will streamline its process for future products. The cell line stays the same; only the baculovirus changes for each new project.

Retallack said that it is important to consider how the FDA will react to a new expression system. As Pfenex began to develop a new platform for human therapeutics, staff members went to the FDA for a lunch-and-learn session to talk about the platform: its similarities to and differences from Escherichia coli, and its safety profile. Because the cell line started as a service for other pharma companies having difficulty expressing certain proteins, the company could point to proteins that were already being produced for clinical trials before making its own products. Pfenex has not had too many questions to date about the safety profile of its own platform or products.

Taylor said that it is never fun being the first in front of the FDA with anything new. “You should try to anticipate the agency’s questions, try to prepare your case ahead of time, but don’t discount the odd question coming in.” When Protalix brought forth its product, it demonstrated that viral clearance/inactivation was unnecessary because there is no cross transmission of animal viruses with plants and that plant barriers do not allow the transmission of human and animal viruses.

What are the advantages and disadvantages of describing a platform rather than a product? It all depends on who you are and with whom you’re talking. A service company marketing itself as being capable of producing proteins that others can’t make definitely wants to describe its creation as a platform. Companies having difficulties with certain types of proteins will be attracted to that. But investors are more interested in product companies because they see more value there. A platform is great, but what you’re doing with it is important. Making a product will move a company forward.

The seasonal influenza vaccine is a licensed product. However, governments working with Protein Sciences on pandemic flu preparedness are interested in the baculovirus–insect-cell platform as well. They want to be able to make a pandemic vaccine fast domestically. And Protein Sciences can convince people to invest in a manufacturing facility for making a seasonal flu vaccine that can be converted quickly in an emergency to make a pandemic vaccine.

It seems that Protalix has made a decision to be a product company. Was that a transition, and can you describe that? Taylor replied that the platform came first. Scientists and engineers had to solve scalability, translate that science from their laboratory for use in industry, and then make a decision about which product to put into the pipeline first. In the early days, you need to convince people of the rationale, the science, and the technology behind a platform. Once Protalix had gotten a product through clinical testing and approved by the FDA, it attracted Pfizer as a licensing partner, which then in for due diligence before taking it on. From a business development perspective, once that first product is approved and on the market, that answers a lot of questions about a platform. It’s been validated multiple ways in a commercial sense: It’s scalable and approved by the agency to make safe and efficacious products. Multiple world regulatory agencies have inspected the facility for GMP compliance. Finally, it has gone through a large pharmaceutical company’s due diligence.

Is it possible to partner with Big Pharma without having a product of your own? How can you convince them that your platform will be appropriate for them? You can partner with Big Pharma if your platform fills a need for them. The goal for a novel platform is to demonstrate that it can meet an unmet need. Diversity of expression strategy is a good thing.

Are those relationships codevelopment agreements, licenses, or something else? For Protein Sciences, large companies generally are more interested in licensing and taking the platform in-house. With small companies, the company has more product development or codevelopment arrangements. Other companies already have a proof-of-concept product using a baculovirus– insect-cell system and are interested in moving their product into the clinic using the Protein Sciences platform. Pfenex started out more as a fee-for-service and licensing company, but it has evolved into partnerships.

What advice might you offer to biotechnology companies and academia about how to develop a product/platform? Felberbaum said that if you want to demonstrate that a platform is widely applicable and can express thousands of proteins, then you need to demonstrate quickly how good it is. For potential pandemic strains, Protein Sciences has demonstrated in an important proofof-concept project that a product could be completed in 21 days. “At some point,” said Felberbaum, “you need to focus and not be divided trying to produce multiple products at the same time.” Protein Sciences made a conscious decision to focus on flu and demonstrate the advantages of its platform.

Retallack pointed to regulatory interactions, saying that it’s important to consider those when designing a new platform. “Get feedback from the agency to understand its concerns. Also get feedback from partners or others in Big Pharma to understand the issues out there and the types of problems you might be able to solve.”

Taylor ended by suggesting that companies should demonstrate their platform’s robustness (in range and class and type of proteins) as well as its applicability across a wide range, or to showcase its differentiating features (e.g., speed or certain types of molecules) or what unmet need the platform targets. Then get validating clients and customers. “Once you get deals down with a few companies, suddenly others will be interested. If you want to evolve from a platform company into a product company, be very careful choosing your lead candidate.”

Alison Center is editorial assistant (acenter@bioprocessintl.com) for BioProcess International, PO Box 70, Dexter, OR 97431.