Developments in Antibody–Drug Conjugates: A Discussion with Thomas Ryll of ImmunoGen

View PDF


As a major class of emerging therapies, antibody-drug conjugates (ADCs) already have gained the attention of biopharmaceutical researchers and manufacturers because they combine both the precision of monoclonal antibodies and the potency of highly potent drug compounds. A few ADCs already have entered the market, but many more candidates are progressing through industry pipelines. Platform processes are not yet universal (and it remains to be seen whether they ever will be), but major ADC developers are establishing their own with significant success. I spoke with Thomas Ryll, head of technical operations at ImmunoGen, to discuss his company’s current work in ADCs, its approach to harnessing innovation to develop products, the progress of ADCs in the industry, and his perspective of their future potential with immuno-oncology.

Current Work
Would you review ImmunoGen’s work on ADCs?

Thomas Ryll, head of technical operations at ImmunoGen

Ryll: ADCs came to the forefront in the 1980s. When antibody technology was developed, researchers determined how to use antibodies to bind to target molecules on cells while delivering a highly toxic payload to enable targeted killing with few side effects. That was the fundamental idea: to find a “magic bullet.”

Since then, companies such as ImmunoGen have been in the ADC field, and obviously, the field hasn’t stood still. There’s a lot of innovation and technology development beyond the early toxins researchers used that had resulted in issues with immunogenicity, for example. Those issues are long gone, and our toxin platforms now are highly potent with few side effects compared with those of classic small molecules. It’s an exciting time in the industry.

ImmunoGen entered the ADC business more than 30 years ago after having been spun out of Dana-Farber Cancer Institute at Harvard. Over the years, we have accumulated deep bench knowledge in this space and have developed a number of innovative approaches. Since I joined the company two years ago, we have transformed from an innovation and platform company into one that uses innovation to drive its own products forward. Our lead ADC product, mirvetuximab soravtansine, is in phase 3 clinical testing for platinum-resistant ovarian cancer, an indication that still has great unmet medical need. Our current focus is primarily on driving mirvetuximab soravtansine forward, and we recently reported pooled data from three phase 1 expansion cohorts at American Society of Clinical Oncology (ASCO), which demonstrate its safety and efficacy profile.

Beyond that, we maintain our focus on innovation in the ADC space. We cannot be and don’t want to be a one-product company. To drive shareholder value in the long-term, we have to work on our pipeline and continue our focus on innovation. That’s what we do with linkers and payloads, and we have some exciting initial data on our new IGN toxin platform, which alkylates DNA.

And finally, we have collaborations with a number of partners using our technology. There are approximately 50 ADCs in clinical testing and of those, approximately 25% use our platform technology

BPI: Would you expand on the company’s platform approach?

Ryll: We have what we call maytansinoid payloads, which are tubulin polymerization inhibitors. Mirvetuximab soravtansine is using the DM4 version of that platform.

The commercial ADC drug Kadcyla from Genentech (Roche) is using our DM1 platform, a predecessor to DM4. DM4 has enhanced bystander activity, which should have more activity in solid tumors. Let me explain how this works: If the target cancer cell sits in a solid tumor environment, the cell will metabolize our drug, and the active component from DM4 will have a higher ability to diffuse from the initial target cell and into the surrounding tumor cells and kill them as well. DM4 now has an enhanced “bystander effect” compared with the DM1 construct. This is a great example of innovation and the progress our scientists have made.

BPI: How is it different from single-agent therapies?

Ryll: Our FORWARD I phase 3 trial, which is currently underway, is targeting monotherapy for platinum-resistant ovarian cancer with a certain phenotype and medium-to-high expression of the folate receptor alpha, which is the target for the antibody. When testing a new class of oncology agents, you have to start working with patients who have had many prior treatments. Typically, treatments for ovarian cancer begins with platinum-based drugs, but patients who are resistant to those drugs need other options. That’s where our focus lies.

Since we see very benign  side effects compared with small-molecule drugs and strong activity, we believe mirvetuximab soravtansine can work as a single-agent therapy. Small-molecule drugs are like blunt weapons. In the case of ovarian cancer, platinum-based drugs, for example, are highly potent initially but have low specificity and thus higher toxicity. But with an ADC, you can go after a specific target on a cancer cell and basically hit target cells without causing a lot of damage to healthy cells.

Beyond monotherapy, mirvetuximab soravtansine could be a perfect combination therapy candidate, given that it is well tolerated. To explore this, we are running a number of phase 1b/2 exploratory trial (FORWARD II) for a combination of mirvetuximab soravtansine and other drugs for platinum-resistant patients.

Addressing Development Challenges
Some issues with ADC development are site-specific conjugation, off-target toxicity, and the need to increase potency and homogeneity. How does ImmunoGen address these?

Ryll: We recognize that what we need going forward is a tool box of solutions to deal with aspects such as specific toxicity, specific engagement in certain tissue-based tumors, aspects of drug release, and aspects of pharmacokinetics. The payloads we have developed and used have different potencies. For example, if you go after a tumor with an extremely high-potent payload, then the amount of drug to be administered is small. If you administer a small amount of drug, the target distribution and pharmacokinetics of the drug determines whether you get an effect that is potent enough. So you may need to play between the potency of the payload and the dose you can administer.

In the future, we will need tools that allow us to titrate potency of payloads to meet pharmacokinetic requirements for certain tumor indications. That’s what we do with a toolbox of different linkers, different payloads with different potencies, different conjugation technologies, and the appropriate antibody and target. By developing this toolbox, we can in the future titrate lower or higher exposure or faster or slower killing to make a drug product fit the need of certain tumor indications.

ImmunoGen employees use isolator containment and powered air purifying respirators (PAPR) while handling potent compounds at the company’s manufacturing facility in Norwood, MA.

BPI: Would you elaborate on the development of the new linkers and payloads?

Ryll: We are excited about our new class of payloads called IGNs. We have our first IGN in a phase 1 clinical dose-escalation study. Later this year, we will file an investigational new drug (IND) application for our second IGN-based ADC, which has an even higher potency.

IGN payloads are DNA-acting. These toxins enter the cell, bind to DNA, and then stop the cell from dividing by preventing DNA replication. Our new IGN class will enable us to develop a wide therapeutic window. Let me explain our working hypothesis: To understand how this works, imagine you have a toxin that binds to DNA and has two active sites. One site binds to one strand of DNA. The other site binds to the second strand of DNA. The results is crosslinking the DNA. This is a very stable structure, and the cell has no way of repairing this. So when that cell goes through the next cell cycle to duplicate DNA, it will go into apoptosis and die. Such a drug is called a crosslinker.

These crosslinking agents have been tested, are extremely potent, and highly toxic. Of course, if you hit normal body cells with such toxins, you can have toxicity. Our researchers, though, have been thinking about how to develop a DNA-acting payload with similar potency to a crosslinker but also less toxicity to normal body cells. The solution our scientists came up with was to design a molecule that has one active side that chemically binds to one DNA strand while the other side attaches itself to DNA but does not chemically bind. We call these alkylating agents.

A fast-growing cancer cell can replicate every day, whereas a normal body cell (such as a progenitor cell in bone marrow) might replicate only every month or so. If you insert an alkylating agent into a fast-growing cancer cell’s DNA, the cell will attempt to double, but the molecule will prevent DNA replication and will kill the cell. That said, when it integrates into a normal body cell, there is much more opportunity for this toxin to degrade before the cell multiplies. Our initial studies have shown that an alkylating molecule like this has similar potency to kill cancer cells but much less toxicity in killing normal progenitor cells. We call this “widening the therapeutic window,” and you can now dose the agent into an area where you efficiently kill cancer cells, but you don’t yet kill the normal body cells. That’s our working hypothesis, and our first molecule is in dose-escalation clinical studies.

BPI: What are some supply chain issues with ADCs, and how do you address those?

Ryll: Because ADCs are made of different components (antibodies, linkers, payloads), supply chain challenges lie when all of these elements come together. To manage your supply chain, you have to select the appropriate partners, ensure that their quality systems are adequate, and that they can deliver on time. You have to manufacture the linker, the drug, and the antibody. So the appropriate manufacturing partners need to be capable of releasing materials on time so that it can ship to the conjugation site. After receiving all elements at the drug-substance manufacturing side, conjugation of the drug to the antibody can happen (producing the drug substance) followed by drug product production at the same or a different time. The process requires management of manufacturing activity timelines, the hold times in between, commitments from each organization, the appropriate quality systems in place, and the appropriate commitment from your partners to deliver on time.

BPI: What do we know about characterizing ADCs and what related technical challenges?

Ryll: That is one of our core competencies. ImmunoGen has focused on innovation of linkers and payloads, conjugation technology, and the analytical characterization underlining all of this. You are only as good as your analytical assay. Analytical capability is key. The challenge is that you have to go beyond the analysis of impurity profiles for small molecules and the analysis of purity profiles of antibodies. Combining an antibody with a small molecule changes the behavior of that antibody, and it becomes more hydrophobic. So a major challenge becomes the formulation strategy and analytical capability to look at the drug substance conjugate in its totality — not only biophysical properties, but also biological properties in terms of specific and unspecific toxicity, for example. In the end, you need to be up to speed on the combination of chemical, biophysical, and biological assays.

BPI: Has a regulatory pathway been established for ADCs, or is more guidance needed?

Ryll: Certainly it’s not as established as it is for small molecules or biologics. At times you might deal with a reviewer who is more familiar on the small-molecule side (for example, if you deal with aspects of starting material designation for your small molecules). Or you might deal with somebody with more biologics experience. Ultimately, you hit this “interface” at the regulatory agencies, and now you need to draw them all in. When it comes to filing, you would write one filing module for an antibody describing your development and manufacturing process. But for an ADC, we have to write more of those sections, including one for the linker, another for the payload, and a section for the antibody (similar to a biologics application). Taking it even further, we would then be required to write a fourth section, which now deals with the production of the conjugated drug substance. The drug-product piece is similar to a standard application. So the process is more complex, with two worlds coming together. There is more work to be done between industry and regulators to inform and speed up some of these processes, but I think the fundamental principles are well established and constant communication with regulatory agencies is important.

Future Work with ADCs
What are some future opportunities you’re looking forward to?

Ryll: I took the opportunity to join ImmunoGen for a couple of reasons. One is the vision for morphing the company from a platform-focused, research-oriented company into a true product-focused company for delivering products to patients. The other is the space I see for ADCs in the future. There is so much going on. There is the new field of immuno-oncology emerging now, and I’m excited about the combined use of ADCs with immuno-oncology drugs. I think ADCs will have a powerful role to play in that future. I’m not an expert in oncology, but I understand enough to know that tumors are very different, and of course, patients are different. If we could treat everyone with the same drugs, we would have an easy play, but we can’t. The future of oncology is combination therapy, and ADCs will play a significant role in that area. Having a precision weapon for tailoring drugs to specific targets with a wide therapeutic window and combining it with other small molecules — maybe immuno-oncology approaches — will be the way to go. I’m excited to be part of building the ADC space, and I’m happy to see that there is a huge number of clinical trials planned and underway. I think this whole area will have significant impact going forward.

Maribel Rios is managing editor at BioProcess International;

Leave a Reply