March 25, 2021
by Jesse Harris, Digital Marketing Coordinator, ACD/Labs
Antibody-Drug Conjugates (ADCs) are a class of therapeutic that is attracting a lot of attention. These specialized biopharmaceuticals leverage the specificity of a monoclonal antibody (mAb) to deliver a ‘payload’ drug molecule in a targeted manner. At least 50 biopharmaceutical companies have ADC development programs, with more than 180 ADCs in clinical trials.
Unfortunately, the potential of ADC therapies may be held back by several issues, including ineffective data management. A recent white paper, titled “Addressing the Unique Challenges of Data Management in ADC Development”, explores how the characterization of ADCs requires a variety of analytical techniques due to their complex structure. The white paper also explains how Luminata can simplify ADC development by bringing together many types of analytical data in one program to provide researchers with a holistic perspective for effective decision making.
The authors, Eliot Randle (Ph.D., an Independent Consultant in the Biopharmaceutical Industry) and Joe DiMartino (Luminata Solutions Manager, ACD/Labs), had a lot to say about the future of ADC therapeutics. We had a conversation covering why biopharmaceutical companies are so excited about ADCs, and how collaborative development strategies are particularly advantageous in ADC research.
This interview has been edited for clarity and brevity.
What are some of the features of ADCs that make them special, and how are they different from small-molecule pharmaceuticals?
Eliot: The first thing is that they are a more targeted therapeutic. Rather than just giving somebody a dose of a drug that ultimately gets distributed through the body, ADCs provide far more specificity.
On the other side of things, to achieve that [targeting], they have a rather complex structure compared to a typical pharmaceutical. Therefore, their development and subsequent manufacturing are more challenging than a typical pharmaceutical.
Joe: I agree with Eliot. ADCs are targeted and go where they need to, while small molecules may have more interactions within the body (e.g., cells, tissues, etc.), which may cause harmful effects.
The FDA approved the first ADC about 20 years ago. Now, the market for ADCs is roughly $2-3 billion per year worldwide, but it is estimated to grow to about $13 billion in the next five years. Why is the market growing so rapidly?
Eliot: I think it’s a result of two things. The targeted nature means that ADCs are particularly well suited for use against cancer-based diseases. Rates of cancer are increasing globally, and they continue to increase quite rapidly. That’s not going away. The need for smarter therapeutics to address the rising incidence of cancer is one driver.
I mentioned earlier that manufacturing is a constraint. Being complex to manufacture has historically constrained the development of new therapeutics. The time and resources required are only really available to a limited number of large companies.
However, biopharmaceutical companies now have much more experience working with ADCs. If you look at the pipeline for ADCs and those currently on the market, they are often the result of partnerships. This sharing of expertise, technology, resources, and knowledge has helped drive things forward. Companies have pooled expertise in the different elements of ADC development, and this collaborative approach has helped drive things forward.
ADCs have three main components: the antibody, the drug molecule, and the linker. Sometimes, the linker is discussed as just a tether connecting the two functional portions of the drug. Is it that simple? Or is there more that goes into linker development that is important to understand?
Eliot: The linker does provide the functional role of being the connector, but it’s a bit more than that. You need to stabilize the ADC in the circulatory system, so it’s got to be stable, but this needs to be balanced with being brittle enough to release the drug or the ‘payload’ when the antibody combines with the antigen that it’s targeting.
It’s not enough to just attach, it has got to attach when it’s appropriate to be attached, and it’s got to break when it’s appropriate to break. This idea of being stable outside the cell and then unstable when internalized is a key characteristic. The linker also needs to provide stability during the preparation and storage of the ADC. It’s there to make sure [the ADC] stays intact, not only once you’ve created it but once it has shipped to the patient.
In many cases, the drug portion of an ADC is based on previous anti-cancer agents considered too potent in clinical settings. Why is this? And is this an opportunity for data-savvy pharmaceutical companies to leverage their old research for new applications?
Eliot: A lot of the candidate drugs that didn’t make it to market failed due to non-specific cytotoxic effects throughout the body when administered at the dose required to be effective against the tumor, resulting in severe side effects.
The fact that ADCs enable drugs to be specifically targeted means you can administer them at a lower level yet still achieve a high localized concentration at the tumor. These drug candidates have been given a new lease of life as the payload of an ADC.
To the second part of your question, I think this illustrates the importance of good data management. If a compound fails at an early stage of development, but the results are not as well documented as more successful candidates, or documented in a way that limits their ability to be found, then the potential value of that data for future ADC research is lost. Data-savvy companies realize that data from today’s failures might be useful in the future.
There’s also a broader industry trend of finding better ways to mobilize legacy data. Pharmaceutical companies have such rich databases that finding ways to get more out of them is a big competitive advantage.
Joe: You’re right Jesse, you do not want to lose that legacy data, where it gets stored in an archaic system that no one manages. Having tools to leverage legacy analytical data helps any project going forward, not just by accessibility but also by allowing for machine learning.
Eliot: More generally, there are a growing number of drug repurposing initiatives out there. Effective repositioning requires connecting the dots between different data sources and between recent and legacy data. Legacy high throughput screening data can also have a new lease of life—a compound might have been screened out in a cell-based assay due to its cytotoxicity, but that may be really valuable information when you’re looking for ADC development candidates.
You mentioned earlier that the ADC development work is often shared across organizations. This implies that ADC research and development could be spread across multiple teams—even more than what is typically seen in pharmaceutical research. How does that complicate ADC development?
Eliot: I think you definitely need to bring together people with different expertise. The expertise needed to develop and produce an antibody is different from that required to design a linker molecule or to synthesize the payload.
You definitely need a multi-disciplinary and collaborative approach. You need experts in chemistry, bioprocessing, and immunology as well. They might be brought together as a cross-functional team, or coordinated but separate groups of specialists.
Collaboration and coordination are the keys to success rather than exactly how the development team is structured, particularly when the development is performed across two or more companies. To make that work, the different data generated by each discipline needs to be integrated. Even if they’re part of a single multifunctional team, each discipline generates different types of data. The chemistry data and the biological data are usually held in separate systems. You need some way to connect the dots.
Joe: That’s a huge issue. This is all scattered. Every team has to trust in their abilities to do their portion, but when something does come up, how easy is it for everyone to access that information? Having the capacity to see the bioprocessing information, at the same time as the chemistry, and then notify others if there’s a small molecule or genotoxic impurity that would cause an issue.
That’s a big piece—doing it properly. Having the proper integration where all teams can view a record and view a project together, that’s something that can initiate everything, right? From our side, how are people doing this, how are they viewing it? If there is a problem, what can be done right away? Because by the time you find [a problem], it has to be solved the day before. You’re already behind. There’s already an issue.
How is it that data sharing done? Is there synchronization between these different teams? Do you have any notion of the norms within the industry?
Eliot: In many cases, it’s still done through informal communication or exchanging files by email or via secure shared file stores. This isn’t a particularly great way of doing it. Without a standard format for data exchange, it limits the degree to which data can be integrated. There are plenty of opportunities to do things more effectively and more efficiently and reduce the cost of sharing and managing data, ultimately helping reduce the overall cost of development and manufacturing.
It seems that data management may offer an opportunity for a competitive advantage in the CRO/CXO space.
Joe: Data management is not just crucial for the pharmaceutical sponsor company but also their CMO partners. We have seen globally that unstructured data gets lost or, even worse, not used to the extent that it was created. Programs like Excel were not designed to handle batches of chemical or ADC information. This leads to transcription errors since you have so many source types being added into one location. On top of that, the CMO is not just working with one company but with many. You can start to imagine how many Excel sheets or PDFs need to be created.
To my knowledge, all the ADC therapies on the market are targeting cancer. Thinking about the future of ADC therapeutics, are there non-cancer diseases that researchers are thinking about as targets?
Eliot: Outside of oncology, I see some interesting developments in the area of infectious diseases. If you can use a mAb to selectively target part of a pathogen, whether that’s a virus or a bacterial infection, then that makes it amenable to an ADC-based approach.
I’ve come across people exploring the use of ADCs for metabolic disorders as well. ADC technology is certainly not limited to oncology, but of course, oncology is a huge market and the one that everybody wants to address and solve, so that’s where most of the research has gone.
Is there anything else about the development of ADCs or the future of the ADC therapeutics you would like to mention?
Eliot: Going back to one of my original points–they’ve got huge potential if you can address the problems associated with the high cost of manufacturing. While things have definitely improved over the years, the complexity and cost of manufacturing still constrains the true potential of ADCs. Whether that’s data-related, equipment-related, or process-related, any technology that helps reduce the manufacturing costs will help unleash the growth potential.
Are there any other technologies that are targeting development and manufacturing costs?
Eliot: Not necessarily specifically for ADCs, but more broadly in the biopharmaceutical industry, I am seeing increased use of statistical-based design of experiment approaches to help streamline the development process, for example, to optimize process conditions and formulations. Plus, there’s the adoption of miniaturization and automated technologies, which often go hand-in-hand with the design of experiment approach. Many companies are also applying process analytical technology to enable a more proactive approach to quality.
All of these things come together and can ultimately reduce the timelines and costs associated with the development of therapeutics.
Joe: In my opinion, it’s looking through and getting a better understanding of the needs of bioprocessing. I know there are systems out there. The key is to have a system that allows you to integrate bioprocessing knowledge with the chemical knowledge about the linker and the payload. If you can have that integration, that would be a game-changer.