Automation is key in cell and gene therapy manufacturing
Although the list of cell and gene therapies is growing, prices remain high. Now developers are working on a road map to streamline production and drop costs
The great opportunities of cell and gene therapy (CGT) are self-evident. Life sciences businesses can find a new purpose in a generation of products that could treat or cure millions.
But the CGT industry is early on in its learning curve. Therapies currently priced in the hundreds of thousands or millions of dollars will struggle to find wide payer adoption, however innovative the emerging pricing and payment models may become.
To achieve significant cost savings the sector must move from the manual ad-hoc hospital labs, where the early therapies have been delivered, to scaled-up and far more automated processes that are faster and more reliable.
To succeed in this endeavour, the industry must overcome a number of challenges, as we explore below.
The constraints on scaling CGT manufacturing include:
• Current processes are slow, inconsistent and expensive
Many therapies are developed in piecemeal fashion in open systems that can be slow, replete with inefficiencies and open to errors, says Patrick Stragier, Vice President of Operations, Catalent Cell Therapy.
“Most of the time, processes have been developed in a university lab or hospital environment where the developers are not accustomed to thinking about manufacturing and production scale. The focus is mostly on the cure, not the process. They use what they have available to them in the lab, which is small-scale, open systems. Open systems are not optimal; they can bring contaminants and are often labour intensive. In a closed system you gain a lot of time and potentially reduce the risk of contamination.”
• The lack of a GMP mindset results in many process bottlenecks
CGT, by its nature requires more of a specialist approach as compared to traditional biologics, that often comes down to the creation of single batch therapy for a single patient. While this may seem to require a bespoke approach, in fact, a good manufacturing practice (GMP) approach can streamline and harmonise these processes quite effectively.
"As soon as possible in development, homegrown or academic labs should collaborate with CDMOs to develop processes that can be optimally designed with future scalability and material sourcing in mind,” says Stragier. “It must be reproducible.”
This extends to automating, as far as possible, a robust Quality Control and documentation process, says Jason C. Foster, Chief Executive Officer, of CGT manufacturing platform creator Ori Biotech. “There is a QC release process for every single patient that entails a whole bunch of testing that has to be done and can take up to a week.
“In today’s analogue world, that could run to 200 pages of a paper batch record and someone has to put their name to and say it is safe. This is fine for producing a few products per month but doesn’t scale to the level needed to make these products widely available for patients.”
Human talent is another bottleneck. There is a growing demand for expertise in some key areas, notably manufacturing scale up and laboratory expertise, that is leading to a shortage of the right talent in the geographies where CGT clusters have grown up.
Other quirks of the current CGT production approach, such as the heavy consumption of bespoke single-use plastics, the requisitioning of which can itself often create long lead times, also need to be addressed.
• CGT R&D is not like big pharma R&D
Veteran producers of mature molecules often think they understand the complexities of R&D but CGT can differ in important ways, such as potency tests, says Stragier. “They are familiar with potency tests for proteins and small molecules but it’s not the same in CGT. The reactions are quite different.”
The application of animal models in pre-clinical development is also different, adds Stragier. “In CGT, animal models may not be applicable to humans. Sometimes that concept is overlooked by companies that are unfamiliar with the emerging field of cell and gene therapies.”
The way forward
Needless to say, the industry is working on solutions to the constraints outlined above and manufacturers are making rapid progress.
• A modular and platform-based manufacturing approach
Approaches are changing, moving away from batch-by-batch lab work, says Emily Moran, Senior Director of Viral Vector Manufacturing at The Discovery Labs Center for Breakthrough Medicines. “What I see evolving is a more standard platform approach to buckets of manufacturing. That will become more widely accepted.”
“CGT Programmes are different from large-scale pharma manufacturing processes and are therefore better suited for specialists with an experienced track record”, says Moran. “There is a high level of complexity and variability associated CGT manufacturing. CGT assets are usually ‘small different’, not ‘big same’. CDMOs are built to adapt with dedicated testing labs and segregation control in every clean room. It is really about agility in addition to scale.”
Harnessing data will also be central in speeding innovation and unblocking bottlenecks with the platform approach, says Foster. “Ori can aggregate and analyse data across the user platform in real time and spot areas of variation, helping developers learn more about their process development more quickly. This has not been available with the first-generation processes.”
• A new generation of modern, flexible ‘end-to-end’ facilities
Key to this new approach will be dedicated facilities and extensive process automation that removes bottlenecks. Existing processes tend to involve a lot of “moving fluids around, adding fluids in and taking them out. There is yet to be a holistic approach that removes those manual steps and the variation that goes with them,” says Foster.
Robust, reliable and repeatable manufacturing will be achieved by air proofing and automating processes in the new generation of facilities, which are already taking shape. The 600,000 sq ft Center for Breakthrough Medicines, for example, is claimed to be the largest in the world with scope to accommodate developers from start-up to scale up and beyond.
One of the important features of the next generation of automated facilities is modularity, says Foster. Modular facilities in which swappable components can be wheeled in and out depending on production requirements could add flexibility, shrink their physical footprint, require fewer skilled employees and take less time and money to build.
Another important feature of these next-generation facilities will be on-site testing. Not needing to send away for testing saves time and improves batch success rates by enabling control automation with in-process testing and realtime feedback loops.
“Can you or your CDMO do that testing in house? You want to know what is happening inside your bioreactor. Are your cells growing? Is the virus replicating? Depending on a third-party tester can mean weeks to a result which can negatively impact both your batch and lot release time,” says Moran.
Being able to analyse what is happening in a process in real time and to interveneimmediately can improve success rates significantly. “The more we can do to understand what is happening in real time, the better,” says Foster. “If you can cut some of the experimental process from two weeks into two hours, you can shorten time to market to get these treatments to patients faster.”
The nature of the therapy defines the manufacturing process and the more ‘bespoke’ autologous therapies (where the therapy is made from and then injected back into the same donor) will require a more decentralised approach, closer to the patient, perhaps in dedicated hospital clean rooms.
Allogeneic therapies (where cells from a donor can be used to treat another person or multiple people) by contrast, are more amenable to larger scale, more centralised production, possibly with a range of such therapies manufactured in a single plant on a continent-wide basis. While less common right now, the potential graft-versus-host disease risks of the approach are being tackled and in the long-term Stragier expects it to become the most common approach.
• A growing appreciation of the need for a manufacturing-first approach
Considering scalability from the very beginning is a vital way to de-risk and speed the commercialisation process for CGT.
Thinking aboutscalability earlier in the process even at the pre-clinical stage is important, says Foster. “We have seen many of these therapies get to phase one or two without a plan of how to reach manufacturingcommercial scale. Let's all think about scalability from the very beginning of the academic process if an indication has more than a few thousand patients. Then we can design a process we know can reach patients at scale."
The features of modern facilities should also include unit operations that allow for variations in supply and demand, multi-product capabilities, electronic batch recording and QR code generation, says Stragier.
This manufacturing-first mindset should extend to distribution as well, which has very different requirements when compared to many pharmaceutical supply chains. Advanced cold chain shipping and warehousing solutions offer greater scope for a more centralised manufacturing operation. It will also be an important part of meeting regulatory concerns, adds Moran. “Regulatory agencies want to know that your product is stable and safe.”
• The advent of a collaborative ecosystem
Achieving faster and cheaper CGT manufacturing will also require much closer cross-disciplinary working to realise efficiencies, says Stragier. “We need biologists, mechanical and bio-engineers, GMP industry-minded experts and logistical and IT experts to optimize these systems to work in harmony correctly. It is very important that we develop good collaboration between these multi-disciplinary teams.”
Through close collaboration, different disciplines are likely to learn from each other, and could identify areas of crossover from other industries, all contributing to the speed of manufacturing innovations and advances, adds Stragier.
An effective cross-disciplinary approach will also include the equipment manufacturers, says Moran. “Equipment vendors are key stakeholders here. They are catching up with integrated, automated solutions for small scale processes that are reliable but also more affordable and easier to use. The Discovery Labs has great relationships with vendors and when you need a complex mix of vendors for different processes, our under-one-roof concept really helps.”
Orders of magnitude cost declines
The advances outlined above have the potential to realise dramatic cost savings, says Foster: “The cost problem is multi factorial. Making manufacturing more consistent, reproducible and automated will lower the cost substantially.”
He predicts first-generation advances in a new platform-based approach could lower costs by 40-50% with a longer-term roadmap for reducing them by 80-90%.
Stragier points to the massive cost reductions in monoclonal antibodies in the 1990s from tens of thousands of dollars to around a dollar per gramme as a good precedent to aim to follow. “I expect it will be 10 or 100 times less expensive if we can solve the problem of scale and manufacturability.”
If the advances, and the resulting cost declines come as rapidly as hoped for, the idea of such personalised therapies at scale, may be more than simply an oxymoron.