Moderator Tom Ransohoff, with Jorg Thommes, Chris Love, Rajesh Beri, and Geoffrey Hodge
For biopharmaceuticals to mature as a process industry, companies need to embrace the ability to adopt new technologies and bring new operational approaches to their biomanufacturing facilities. In this roundtable about adopting and implementing new technologies in the biologics industry, moderator Tom Ransohoff stressed the importance of understanding drivers for adopting new technologies and of developing enabling processes that factor in associated risks and challenges.
He offered the example of single-use (SU) bioreactors as a successful adoption of new technologies. “SU bioreactors have changed the way we sterilize the vessels, the way we connect to them, the way we sample them. So companies needed to develop new skills and develop new approaches for using the technology.”
Now, 20 years after introduction of the first SU bioreactor, that technology is a de facto standard for clinical and commercial manufacturing facilities at the ≤2,000-L scale. But SU bioreactors haven’t displaced conventional stainless steel equipment — which still dominates the landscape volumetrically for biologics production. And SU bioreactors are not envisioned to displace stainless steel completely any time soon, although technology continues to evolve and mature for the industry.
Forces driving adoption of SU bioreactors include speed and reduced capital — the need to bring manufacturing facilities on quickly and economically. Other needs include improving product quality and ensuring compliance with regulations that drive the need for new technologies. So cost often is not a major driver.
Ransohoff noted that highly regulated industries are conservative toward taking on risks inherent with adopting new technology — risks that can include significant penalties for delays and setbacks. The need to move programs forward quickly might seem to provide no good time to adopt new technology. But he argued that the biopharmaceutical industry puts relatively little investment into new technologies compared with other industries. As he invited the panelists to offer examples of successful technology adoption, he also asked them to consider whether their industry is spending at the right level for new technology development or is doing things about the way it should.
Successful Technology Adoption
Beri’s company has ongoing projects toward improving cell lines and expression systems while fortifying and simplifying its media feeds. It continues to invest in platform processes that incorporate new technologies, particularly sensors, and to realize the full potential of process analytical technologies (PAT) for real-time process monitoring and control. Lonza also has a strong interest in manufacturing execution systems (MES), working toward digitizing biomanufacturing eventually.
One good time to think about implementing new technologies is in the early stages of process development, perhaps when constructing a new facility. Adoption is more challenging in existing facilities and processes. Beri’s example of successful adoption involved converting conventional at-line UV spectrophotometers from traditional fixed-path, fixed wavelength units to variable-path wavelength spectrophotometers. As a contract development and manufacturing organization (CDMO), his company had multiple customer projects running in two manufacturing suites, so real-time adoption seemed impossible. Nevertheless, Lonza followed good risk-assessment strategies, made use of available protein standards, and showed good compatibility between the new and existing spectrophotometer results. Customer support helped the company to overcome many related challenges by recognizing the shared value of that adoption.
Incremental Advances: Hodge’s perspective was that advancements within the industry have not come from a few big inventions, but from the cumulative effect of many small innovations implemented over time. He asked how innovations get implemented in an industry that’s as conservative and risk-averse as biopharmaceuticals. From a biopharmaceutical innovator’s perspective, an incremental change to save perhaps 5% on cost of goods or shorten somebody’s timeline by 5% seems like solid justification. But from a consumer’s point of view, is such a small gain worth the very significant risk of an untried technology?
Hodge’s example of a successful adoption came from his work at Millennium Pharmaceuticals in a process development (PD) laboratory that was interested in SU technology. That group spliced together an entire antibody process at bench scale in SU components using SU sensors. It borrowed heavily from the medical industry’s use of many SU components as an early adopter of wave-motion bioreactor technology.
At that time Millennium was making a transition from drug-product manufacturer to service provider. It was building a manufacturing facility with an ambitious timeline for getting a product into the clinic, but none of the CMOs at the time could take on technology transfer and offer the needed capacity to meet that deadline.
So he described creation of a modular biomanufacturing process using SU components in a small cleanroom. And when Millennium moved out of biopharmaceuticals, its displaced biologics group gave birth to Xcellerex — which turned that modular technology into a product. The new company developed the first SU stirred-tank bioreactors that ultimately scaled to 2,000 L. Later, Xcellerex had one of its first big commercial successes when it installed multiple 2,000-L bioreactors instead of stainless steel at Shire’s Atlas facility, saving that company 18 months in bringing its facility on-line. None of these efforts were aimed at incremental improvements but, rather, toward meeting major unmet needs.
Hodge concluded that “as a product developer, you need to look for those places where you don’t just incrementally improve on a technology. You need to find a place where you can make a huge step change for one company, and then you can establish a baseline and some confidence in the product. Eventually it will be adopted by the industry.”
Three Drivers of Transformation
Love suggested that the panelists consider three key ingredients driving transformative solutions: need, disruptive technologies, and an ecosystem. In looking at the industry’s successful use of platform processes, he noted that in the 1980s, biomanufacturing of monoclonal antibodies (MAbs) with Chinese hamster ovary (CHO) cells came about through a combination of those three ingredients. The need was for medicines based on biologics. The disruptive technology was recombinant DNA. And the ecosystem comprised academics, industry, and government stakeholders who could think collectively think about how to enable it all as a platform. Incremental innovations happened since then, but those three ingredients at that time led to the production of many biopharmaceuticals.
Today, the needs are more diverse, but disruptive technologies enable learning biology directly from patients by reading out their genomes, from sequencing small bits of DNA, or from single cells from tissue samples. That gives developers great precision in thinking about an indication, defining it, and identifying a target. Pipelines are filling with ideas about where and how companies can innovate new solutions for drug products.
Challenging new modalities such as cell and gene therapies still need to be made amid constrained pipelines and limited resources.
Disruptive Biology: Love suggested that the disruptive technology today may not be another reactor or resin, but the biology itself, which in biomanufacturing is accessible now in ways that it has never been before through tools used in research and discovery. Transformative genetic engineering technologies now provide the ability to write information back into that biology, writing it into needed solutions that simply have not been possible before.
He asked, “What if you could fine-tune the medium based on what is actually needed when you need it? We’ve done some of that in the lab at MIT thinking about new media formulations for Pichia pastoris, a yeast organism, that are now fully defined based on transcriptomic information about what’s required in a fermentor during the cultivation. This one simple example shows how the lines are blurred between strain engineering and process engineering informed by the biology itself.”
His third example illustrated the ecosystem element. Today’s industry is robust in part because of leaders who were well trained in places in many different institutions. Those leaders worked closely with others in industry and the government to accelerate many ideas that are now appreciated today as standard processes. A robust ecosystem in which information is shared broadly is necessary to promote innovations — CHO cells are a very good example of what can develop from that. He concluded by saying that the industry’s goal should be to accelerate this learning. “Not only does it create the workforce and leadership necessary to be able to sustain new manufacturing strategies, but it also helps create a broader share of risk in establishing the right science and the right understanding of these technologies to bring them forward to regulators with confidence.”
When an Ecosystem Is Absent: Thommes agreed that given a need, disruptive technology, and ecosystem there always is a good time to implement new technology. But all three components are necessary. As an example, he said that IDEC Pharmaceuticals explored continuous chromatography back in 2001. The need was clear, and the innovative technology was developed. But the process was not implemented because the necessary ecosystem was absent. It would have been custom built, custom designed, and used by only that one group.
Thommes added that nothing is wrong with a custom-built system. But his group was trying to address its capacity and productivity problem with an engineering solution, whereas the resin manufacturer basically used chemistry to improve productivity.
By contrast, today there is an ecosystem to support continuous processing, with knowledge about how to make biologics continuously and cell culture for delivering a continuous stream for receipt in chromatography. There is a vendor infrastructure for support. What the IDEC group failed to do 15 years ago can happen today because all three drivers are present.
Still Stainless Steel?
An audience member asked why companies are still building large stainless steel facilities when the industry has so many single-use, flexible options. The panelists agreed that the industry still has demands that require both approaches. Some legacy processes are treating large patient populations with expression titers <5 g/L. For those existing businesses and successes in biosimilar approvals, the industry needs large expansions. But on the clinical side, SU systems can produce higher titers cost-effectively for lower patient populations. So for now, both stainless steel and SU implementations are increasing.
Thommes added that the industry continues to see increases in both stainless steel and SU infrastructure because of a multiplicity of needs and problems. The SU environment allows quick turnovers for small batches and efficient facility use. But for manufacturing huge-demand products, when combining the productivity and the cost of goods of a well-run, large stainless-steel infrastructure, the SU infrastructure might never compete cost-wise for continually operating facilities. Therefore, companies with such products invest in large infrastructure, and companies with very mixed portfolios invest in single use. Large companies that are committed to a multiplicity of modalities will not find one solution to address all needs.
Hodge pointed out that smaller markets requiring frequent turnover between batches and products favor the SU model. But that “there’s some cut-off point where, if you’re making the same drug at a very large scale, stainless steel is just going to be more cost effective.” Although single-use has been advancing the scope of its applications, he wondered how continuous biomanufacturing will change the equation. One big advantage of single use is the ability to change over equipment quickly and reduce cleaning and steaming costs. Continuous manufacturing would appear to reduce some of that benefit.
Love pointed to the value of looking outside the biopharmaceutical industry for examples of why things happen. Using parts manufacturing as an example, he cited 3D printers as a disruptive additive manufacturing technology that creates the opportunity to make unique, specialized one-off solutions with specific capabilities. It allows a company to prototype and move ideas forward faster. The industry still needs products that it can manufacture at low cost at very large volumes. Drawing from other industries for guidance about how to approach manufacturing innovations can be a valuable guide toward developing fast, agile solutions. He offered an example of work at the Massachusetts Institute of Technology toward solutions for intensifying processes. Such approaches eventually could reduce the need for large facilities.
Reducing Supply Costs for Developing Industries
An audience member asked for an update on efforts through the Bill and Melinda Gates Foundation to bring the costs of manufacturing supplies down for developing countries.
Thommes answered that the foundation has been active in thinking about alternative hosts and about decentralized, perhaps localized microfacilities. Love added that a broader program with the foundation is addressing ultra-low–cost vaccine production, with a goal (metrics) of 15 cents per dose in a vial at 40 million doses per year. With a glass vial for a single dose currently costing six cents, economies of scale cannot be used as an argument to reach that target point. His company is working with the University of College London and the University of Kansas on a platform for recombinant solutions to this, and another group is working toward inactivated viruses or attenuated virus solutions for a polio vaccine in particular. Through cross modeling, they know that the key drivers will be small, modular, highly intense processes. Downstream processes would look very different if products from alternative hosts could have 80–90% purity directly out of a reactor.
Ransohoff returned to the importance of an ecosystem for supporting manufacturing innovations, focusing on the value of collaboration both within industry and between it, academia, and government. He mentioned the work of the BioPhorum Operations Group (BPOG) toward developing a technology roadmap and the Biotechnology Innovation Group (BIO), where industry veterans are advising companies about requirements for new technologies. Other initiatives including the Defense Advanced Research Projects Agency (DARPA) and the National Institute for Innovation in Manufacturing (NIIMBLE). He asked, “What role do these types of efforts play in this process, and how can or should we expand our efforts to improve such collaborations?”
Beri spoke about BPOG, which began as an on-line collaboration community. Participants realized that the Internet (a key enabling technology) enabled manufacturing groups to collaborate effectively on-line, networking across multiple time zones. In addition to biomanufacturing work streams, the group began looking at technology solutions to address future needs, examining cost, quality, and technology drivers. The resulting industry roadmap, three-years in the making, focuses on traditional recombinant protein manufacturing. He noted that since publication of that roadmap, member companies have launched nearly a dozen critical projects.
Hodge added that collaborations accelerate innovation by bringing people together who have similar interests, needs, and technologies, increasing their opportunities for finding and addressing unmet needs. He said that “industry representatives working together with government and academia can help push technology to the point that companies and investors are willing to put some time and money into it.”
Thommes highlighted NIIMBLE, part of the Manufacturing USA network, as a public–private partnership that addresses large-scale problems better than people can individually, within their company niches. He noted that sharing initiatives helps lessen the risk in introducing technology innovations.
An important quality of a sustainable ecosystem is long-term commitment from all stakeholders, whether that entails time, intellectual input, or money. NIIMBLE provides one avenue for exploring advanced technologies for manufacturing. Other groups, such as the US Food and Drug Administration’s Emerging Technology team and its equivalent in the European Medicines Agency, provide an avenue to start conversations early among researchers, academics, and regulators. Love added that “from an ecosystem standpoint, finding the right way to seed and establish long-term sustainability as an industry is going to be really critical as we look at advancing new technologies and the people necessary to bring them forward into the next generation.”
Ransohoff posed a final topic for the panel: “In what areas do you see the greatest need for new technologies, and what excites you the most? What area in particular do you see as a big driver moving new technologies forward?”
Love pointed to the opportunity to rethink how medicines are delivered to patients. Currently an entire supply-chain cycle takes 18–24 months. He asked, “what if, instead, a pharmacy could make the drugs a patient needs on demand, release them in real time, and have them ready within days? Technologies exist to do this now.” He broadened that to addressing the ~8,000 rare diseases for which medicines and specialty products are in high demand, asking, “What if you could make those materials when and where you needed them? What would it mean for global access, to be able to bring medicines to patients anywhere when they needed them? It is possible to do this today. It’s possible to move to an agile supply chain. And there are substantial benefits to thinking about a more uniform supply chain of raw materials and a short time of holding finished product.”
Thommes added that his interest is in development of an alternative host expression system to lessen the time from gene to IND to three to four months (from 12). That could bring more potential candidates into the clinic and “democratize access to clinical candidates, at which point maybe an academic lab could run clinical studies.”
Hodge pointed to the cell therapy world, where some of the above processes already are being realized. Autologous cell therapies involve taking blood from a patient, making a product for that patient, and then giving it back. Current challenges have to do with reducing that cycle time from a month or several weeks to a number of days as well as making hundreds or thousands of batches in parallel instead of scaling up. Challenges remaining for regenerative medicines include different approaches to analytics that involve running hundreds or thousands of product quality tests in parallel. “Probably the biggest need for innovation in the cell therapy space is analytics: both new techniques and ways to automate them to turn around lots of batches in parallel very quickly.”
Beri expressed his excitement in the increase of truly curative therapies for oncology. The ability to better understand each person’s disease and lifestyle and maybe make even antibodies more personalized would require huge investments, not only in biology, but also multiplexed manufacturing. “My goal,” he concluded, “is a cure for every disease.”
Watch the full interviews online at www.bioprocessintl.com/BIO-Theater-2018.