At noon on Wednesday, 21 June 2017, BioProcess International presented a panel discussion as part of the “Emerging Techniques and Technologies” session of its BPI Theater at the BIO annual convention in San Diego.
Moderated by Susan Dexter (managing director of Latham Biopharm Group), this panel comprised Deborah Harris (vice president of operations at Abzena), Magnus Schroeder (director of downstream process development at CMC Biologics), and Stephen Lam (vice president of operations at Patheon).
There is no debate: Single-use technologies (SUTs) have become essential mainstream tools to lower costs and shorten time to market, extend the viability of smaller-indication products, and provide flexibility in where and how products can be manufactured. The question now is what technologies are best suited to a given bioprocess. This panel addressed factors behind related decisions and asked whether, in hindsight, companies should have made different decisions. In addition, it considered how the advent of SUTs has changed strategies regarding manufacturing in-house or outsourcing.
Dexter said that some of the panelists’ companies had grown first organically and then through acquisition. The latter adds multiple sites globally with different technologies and different capabilities.
As Patheon has grown and acquired multiple sites, what has been its strategy in deciding where their technologies fit and how to run the business? Patheon Biologics was created by bringing together three companies: DSM (Netherlands and Australia sites) and Gallus (Missouri), which had acquired Laureate (New Jersey) a year before. Process development was spread across all those sites. Patheon made a strategic decision to centralize it within two of them. Now that it is comfortable with those capabilities, each site is developing the other’s set of skills. Both sites must have all the necessary skills because customers are bringing in products at different stages. Patheon wants the flexibility to put customers at different sites depending on capacity.
CMC Biologics was made from acquiring companies as well. What was the strategic process linking them all together? CMC began in Denmark as a contract manufacturing organization (CMO) that wanted to bring new technology to the market but needed greater capacity. The company formulated a strategy for enabling itself to manufacture drugs for clients without a huge capital investment. It decided to use 2,000-L single-use bioreactors, adding multiples of them as needed. The company uses a probabilistic model based on Monte Carlo simulations to go through every unit operation looking for bottlenecks and then coming up with solutions to them. The better the quality of data you put into such a model, the better the quality of predictions you get out of it — so the timing is important. CMC also is interested in technology that would allow downstream processes to be scaled out rather than up, especially a continuous process chromatography system.
Typically CMOs don’t innovate technologies or drive innovation. But now, to address uncertain capacity, they are identifying technology needs and gaps and assessing new technologies as part of their business strategy. How do they do that? CMO scientists work with other companies that are developing the technologies, particularly downstream technologies. They also like to work with customers and partner with them to introduce new technologies into their processes. However, that strategy does involve some risk.
Abzena is in a different position: It has acquired PacificGMP, which was not at a commercial scale. So the strategy will be completely different as it starts from a brown-field site. How will this company make decisions about technology and capacity? Abzena’s strategy is to canvas the current instruments and required technology. It is a slow process. Previous clients who have been using older technology are encouraged to let the CMO modernize their processes. At the same time, great benefits come from working with new clients early on. Abzena is trying to address the question of scale-up/-out, working closely with vendors to choose technologies that will suit most clients.
What new technologies are you using in process development to decrease risk and increase speed or provide more data than before? Patheon is using Sartorius Stedim Biotech’s ambr15 microbioreactor and evaluating the ambr250 system. The latter is scalable all the way up to commercial scale. SUT in general has simplified technology transfer, even between a company’s own manufacturing plants. Results are consistent, particularly with systems from the same vendor.
How do you balance the supply chain for single-use and stainless steel systems? Technology transfer upstream is much easier with SUT. Usually a project runs at only one site, but sometimes it is transferred between sites. No stainless steel systems are identical, but single-use systems are, which streamlines transfers and makes them predictable. CMC uses the ambr technology to improve media development, robustness, and optimization. Downstream screens help optimize chromatography steps. Quality by design (QbD) generates relevant data in a meaningful time frame. With high-throughput process development, analytical capabilities must be built in parallel, helping to prevent bottlenecks. CMC uses high-throughput screening downstream to develop new steps rapidly, optimize them, or answer facility-fit questions.
Some of Abzena’s customers want “just-in-time” manufacturing. The company is trying to develop a platform to offer those clients that will speed up process development and get them material for clinical trials as soon as possible. Abzena believes that being prepared with consumables benefits customers by eliminating wait time.
If a customer wanted an unproven fast-track solution, would you be willing to run it to save time? CMOs would consider using an unproven technology for clients after a frank discussion about the risk of time line delays and production. Often such technologies are single-sourced: If the supplying company goes out of business, then the process is gone too. Additionally, regulatory strategies must be thought through. As long as partners are in it together, if things go well, then they have their product quickly, and the CMO can take knowledge gained on to the next client.
How do you store disposables? It is a balancing act between supplying customers readily and spending funds on supplies, which also requires warehouse space. CMC is working actively with its supply chain to manage that and does have sufficient space for storing what is usually needed. Delivery times seem to be increasing, so forecasting must be done carefully.
What about custom-designed products? If you used a custom bag for a client who comes back and needs a phase 2 batch, can you use a standard bag? It is case by case and company by company. Those that have been using SUT for a long time tend to have more custom systems. Early designs didn’t have all the needed feeds, ports, and so on. Later users have been able to take advantage of what the vendors learned and use one standard option. Usually, a CMO in that situation will use the custom bag because it has been proven. You would have to go through another process to change bags. Generally it is less expensive and less time consuming to use standard items, but that’s not always possible. Many processes have special needs that must be met.
How are analytics affected by the use of high-throughput tools? Have any challenges come in matching them up with manufacturing needs? Many analytical tools (e.g., mass spectrometry) have become commoditized and matured sufficiently that companies can be confident in their data. Assays typically are the same. High-throughput screening isn’t all that different, but you might get a difference in resolution. The one area in analytics that does need to develop further is process analytical technology (PAT); it is one of the major missing links for continuous processing.
Does SUT allow you to “release as you go,” so that a product is fully released by the time it is made? Is that part of PAT or in-line testing, so that you don’t have to wait months for data? Some companies are addressing those issues with beta testing. PAT is where the industry is headed, and we need to make it happen. Taking a somewhat controversial view, Schroeder feels that the industry has the analytical tools to enable PAT, but that the problem is more about interfacing those tools with the manufacturing stream. He would like to set the bar higher for downstream processing: “The goal should be a series of downstream steps that are so robust that PAT would not be required.”
What do you see on the five-year horizon? Schroeder feels that the trend upstream of adding multiple bioreactors will continue — and that perfusion reactors will be coming on-line. He also sees the greatest need and opportunity in downstream processing to come up with compact unit operations that can be scaled up through multiplication as well. “And,” he added, “of course, high-speed, high-capacity solutions.”
Harris agreed that downstream processing is where improvements are needed, including more disposable components. Custom disposable columns are costly and tend to get reused.
Lam said that five years isn’t very long for technology. He thinks that there will continue to be improvement in harvest methods, and that is important because as the expression titers go up, risk of a bottleneck increases as well. He also mentioned buffer conditioning “because facilities have a lot of area devoted to storing liquid buffers.”
What challenges come with moving from stainless steel to single-use? Companies should look at bioreactors from a process engineering perspective, compare their functional abilities, and match processes as best as they can. For example, CMC has products running in both a single-use “six-pack” and in conventional stainless steel and can demonstrate full comparability between those options. In general, making the switch has been favorable for that company.
Lam added that it is just like making any other container change. “If you were going between different stainless steel reactors, you would still have to go through the same engineering study, so just treat it like a change of bioreactor.”
Alison Center is editorial assistant (firstname.lastname@example.org) for BioProcess International, PO Box 70, Dexter, OR 97431.