Biopharmaceuticals, also known as biological drugs or biologics, are manufactured from living organisms or contain living organisms that have been genetically engineered to prevent or treat diseases. Biologics are chemically and structurally complex, and often highly heterogeneous; therefore, controlling and maintaining quality remains a challenge. The potential for new therapeutics to cure and treat previously untreatable diseases is enormous, but there is still a long way to go before they can be manufactured at the required scale, with predictive control of quality, and at a lower cost. We sat down with NIST’s Vijay Srinivasan and Sheng Lin-Gibson to discuss their recent paper on some of the challenges and solutions associated with manufacturing these life-saving drugs.
Lin-Gibson: There are lots of challenges because we're talking about a broad class of therapeutics. We have drugs that look somewhat like small molecules, such as RNA-targeting therapeutics. We have drugs where we have literally gutted a virus to remove the viral genome and replaced it by sequences that, when delivered to patients, can correct the underlying genetic cause of various diseases. We have cells that have been taken from patients, and genetically modified to only recognize and attack cancer cells.
So, the difficulty is that all these drugs are made in different ways. The commonality is that they're all coming from a living system. Biology is incredibly complex. For one, living systems are inherently dynamic and constantly changing. Secondly, they tend to be highly heterogeneous. When we talk about cells coming directly from the patient, we have to worry about patient-to-patient variations. Meanwhile, our understanding of the biosciences continues to evolve at an incredibly rapid pace.
That's really the heart of the manufacturing challenge. How are regulatory agencies going to evaluate the quality and consistency of these products to ensure they're safe and effective? These are manufacturing and measurement challenges that NIST has been actively working with the industry and other stakeholders to address.
Srinivasan: Coming to this as an engineer with a manufacturing and engineering background, what I found was that the living cell is an amazingly complex and wonderful factory. And what we are trying to do in biopharmaceuticals is to coax this natural factory to produce things that we want, which is very different from how we engineers are used to making things.
We're changing the manufacturing paradigm, from large centralized factories to smaller, more distributed manufacturing, because practically any licensed hospital could one day manufacture these drugs in-house, on demand. Unlike other types of manufacturing, these cellular factories don’t need enormous amounts of power. Biological reactions involving human cells occur at around 37 C, which you can get from simply plugging a machine into a wall outlet. Now you’re talking about super small, super distributed manufacturing.
Think about that, compared to the power and infrastructure needed to run a massive chemical plant. This is a fundamentally different way to think about manufacturing. How do you do manufacturing on demand? How do you manufacture at different scales? How do you control quality? Am I making the same drug in Arizona as I am in Maryland?
These are really interesting challenges that are presented by the new paradigm.
Srinivasan: It is very difficult to make these drugs using manual processes, so we have to think about how automation and new manufacturing capabilities, biology, physics and information technology can come together to make these types of products.
Can we envision a future where the therapies are manufactured in a hospital where the patient is being treated?
Hospitals have quality management systems and the control training, so could those evolve to handle distributed biologic manufacturing? What are the additional hurdles that we have to overcome to make that happen?
In manufacturing, we have been talking about mass customization and personalization.
Mass customization and personalization is the same idea as where you don't have a bunch of shirt sizes and then you go pick one close to what your size is, but rather, you can customize things, manufacturing-wise.
That is where the manufacturing industry is moving for some consumer-related products. Now, biologics can be the extreme case of mass customization and personalization. To me, that's very exciting.
This is not just a nice to have, you must have this capability. And now manufacturing is moving in that direction.
Lin-Gibson: The overarching NIST role is addressing the measurement challenges. My colleagues and I in the NIST Material Measurement Laboratory are focused on the measurement challenges for cells themselves, and the tissues, and gene therapies that can be produced from cell-based manufacturing.
Why? What is the difference between measuring a bottle of cells that are ready to be injected into a patient versus vitamins that I'm ready to give to a patient? For a molecule like a vitamin, I need to know the chemical structure, which is pretty well established. I need to know the concentration, and there are established methods for determining this.
I have a vial of cells. Those cells are alive. Each cell is likely to be somewhat different. Some of these cells will be therapeutic; some of these cells may cause unintended adverse effects. How do you determine what and how much is in that vial? Is your measurement comparable to one that was made before? What is your traceability chain? They don't yet have a metrology framework for complex biological systems. We've been working on this with stakeholders nationwide but also globally.
As we think about manufacturing, it’s no longer just about the end product; maybe you want to measure the starting materials, the reagents, or intermediates of a manufacturing process. We now also have to think about the supply chain. What does that look like? What are the reagents and how do you track them? What is the standard for that entire supply chain? How do you document the manufacturing process from start to finish? This is where our work began to converge in a very interesting way with NIST’s Engineering Laboratory.
Srinivasan: That’s another exciting thing about it.
My colleagues and I in the NIST Engineering Laboratory don't do biological measurements. What we do are lots of the documentary standards that deal with how and in what format you capture the information at various stages in these processes.
In general, documentary standards are extremely important, but even more so here because the Food and Drug Administration requires that you gather and keep this information and then you transmit it appropriately along the supply chain.
In my opinion, NIST is uniquely positioned for helping to create documentary standards. We have the Engineering Lab and the Information Technology Lab, both of which contain experts in the engineering and informational aspects of these standards. And then the Material Measurement Lab has expertise in the actual measurement science of these things.
At a recent workshop where many biopharmaceutical industries were represented, one aspect that they talked about was the data standards.
And they felt that this is such a high priority that they want to do it within one year, so it's very exciting that industry sees the importance of standards in order to succeed in this business.
Srinivasan: We have to automate as many of these processes as we can to avoid contamination by viruses or other pathogens, so all aspects of monitoring, control and automation are extremely important. In addition, information about each link in the chain has to be collected, and then exchanged seamlessly among the various players.
What we try to do in our recent paper is explain to this community that there's this vast area that requires collaboration.
Now, industries are really trying to do this and struggling with it. And that's why the roadmaps that were presented in this paper were developed to say that these are the important problems that we need to work on. And they very clearly show why engineers such as the robotics and automation people are getting involved in that.
Our goal is to meet industry in this process and use our unique NIST perspective and expertise to help guide them into creating the data standards that are needed.
Lin-Gibson: Some of the output will be standards, many of which can perhaps be adopted from existing standards intended for more established biotechnology sectors. But we need to have conversations about to what extent can we adopt and learn from existing standards and best practices. So that's where I think a lot of the expertise is required because folks should not be starting from scratch.
On the other hand, when we talk about standards, we are really talking about a lot of different kinds of standards. We also need to address a lot of different measurement challenges before we get too far ahead of ourselves.
I think NIST can really offer a lot to the entire ecosystem.
The new NASEM report on Safeguarding the Bioeconomy suggests Technology Readiness Levels (TRL) (pg 152-154) as a common metric for tracking & developing biotechnologies, used by NSF, NASEM, EU, UK. It highlights advanced manufacturing as a high priority need for many areas. Do you think this is the most useful framework for smart bio-manufacturing, or is there some other intellectual framework, e.g. focused on regulatory approval requirements, that is better or an essential addition? Thanks for your comments.
Emerging biotechnologies involve the convergence of many technologies, so it is important to consider various drivers, factors, requirements and risks. TRL is one useful tool to assess where the technology is on a spectrum starting from idea generation, ending in commercialization. In the context of biomanufacturing, smart manufacturing is an enabler, helping to accelerate the capability to produce, with better control, faster, more cost-effectively, etc. In other words, it should help to overcome the valley of death, shown in figure 5-6.
For many biotechnology products, and all therapeutics covered in our blog, regulatory approval is a must. Considerations for meeting regulatory requirements should start in TLR 3-4 (development stage), if not earlier. Analytical methods should be developed as early as possible to help understand the potential product and inform other decisions. NIST is developing the unpinning measurement sciences and standards to help streamline many aspects of the regulatory approval process.
I suggest starting totally from the beginning. For example, each cell has its own definable limits: possible variations and these can be mapped to create receptor "magnets" to capture exactly the cells that are needed. A completely opposite approach to that you outlined. Why?
Because it is a living mechanic - even sheep do not behave by measuring their length or weight, but we choose from the herd those that meet the current requirement.
The limiting standard should be the requirement as in animal or plant production. Ttechnical limits in live production are useless.
Thank you for your reply.