C4 Therapeutics, Inc. (CCCC) Earnings Call Transcript & Summary

Hello, everyone. I'm Nader Sohraby from the JPMorgan Healthcare Investment Banking team. On behalf of JPMorgan, it's my pleasure to introduce Andrew Hirsch, the President and CEO of C4 Therapeutics. As a reminder, we'll try to reserve a few minutes at the end for Q&A. So if you have any questions, please submit them online using the ask-a-question feature at the bottom of the screen. Thank you, and I'll now turn it over to Andrew.

Thank you, Nader, for the introduction. Good afternoon, and thank you for joining our presentation today. Before I begin, I'd like to thank JPMorgan for inviting us to participate in the conference. I'm really pleased to provide an update on C4 Therapeutics today. I'm sure that most of you know the drill by now, but you should all have access to the slides. I’ll indicate which slide number I'm on as I move through the presentation. After the presentation, Stew Fisher, our CSO; Adam Crystal, our CMO; and Bill McKee, our CFO, will join me for Q&A. I will be making forward-looking statements today as part of my presentation, and Slide 2 contains our legal disclaimer on this manner. I'll start the presentation on Slide 3. C4 Therapeutics is focused on making medicines that use a novel approach called targeted protein degradation, which we believe has the potential to transform the treatment of disease. For most of our industry's history, the focus has been on small molecule medicines that inhibit the activity of disease-causing proteins. This approach has been successful and has led to meaningful improvements in patient outcomes. However, it's constrained by 2 factors: first, small molecule development has been limited to 15% or less of the proteome, where we can find ways to bind to active sites on proteins to effectively inhibit activity; and second, for those targets where inhibitors have been approved, selective pressure ultimately results in resistance that render inhibitor approaches ineffective over time, as demonstrated in the image of a melanoma patient treated with an inhibitor in the middle panel of this slide. By harnessing the body's natural process for destroying unwanted proteins, targeted protein degradation has the potential to overcome both of these limitations and transform patient care. What's even more exciting is that unlike other novel treatment modalities, this approach has already been validated in the clinic with the IMiD class of medicines, which are degraders. They make up a meaningful percentage of the $63 billion in oncology revenue from small molecules highlighted on this page. Importantly, the approved IMiDs weren't intentionally designed as degraders. And in fact, by today's standards, they're relatively weak degraders. Nevertheless, they're the backbone of multiple myeloma therapy and have led to meaningful improvements in patient outcomes. What's so exciting about this field is that we now have the science for rationally design highly potent degraders across a wide range of targets. As Slide 4 shows, C4 Therapeutics is a leader in targeted protein degradation, with a platform to efficiently and rationally develop highly potent degrader medicines. We've developed a robust pipeline with 4 programs expected to be in the clinic by the end of 2022. These programs demonstrate the potential of degraders and have an opportunity to address unmet need in large groups of patients. We're also pursuing 14 additional programs in earlier stages of preclinical development across our internal pipeline and with our collaborators, which are going after equally attractive targets. These platform partnerships enable us to extend the potential of the platform. In addition, with the proceeds from our recent IPO, we have the capital to fund our operating plan through the end of 2023. Now while some of you may be familiar with protein degradation, I'll start off with a brief introduction for those of you who aren't. On Slide 5, we're showing a cartoon of the body's natural process to destroy unwanted proteins. When a protein is identified as needing to be destroyed, for example, in the case of a misfolded protein depicted here, the cell will recruit an enzyme called an E3 ligase to attach to the protein, identified as step 1 on this slide. In step 2, the E3 ligase will tag the protein of interest with a tagging protein called ubiquitin, depicted here in bright green. That lets the cell know that the protein is ready for disruption. In the next step, the ligase separates from the tagged protein, which then makes its way to proteasome in step 5, where it's broken down into amino acids, although ligase continues its surveillance for additional proteins. What targeted protein degradation does is co-op this process through a small molecule medicine to destroy disease-causing proteins that the ligase would otherwise ignore. Slide 6 shows how this works. When you add a targeted protein degrader, pictured in this graphic as the green dumbbell-shaped item to a tumor cell, it initially binds to the disease-causing protein, which is depicted in step 1 on this slide. This binary complex of target and degrader then encounters an E3 ligase and forms what is known as a ternary complex, comprised of the target, the degrader and the E3 ligase shown in step 2. Once this complex is formed, the E3 ligase can recognize the target protein, and it then transfers the tagging protein shown in step 3 in bright green onto the target protein. Once tagged, the entire complex disassociates and releases the tagged protein, so it can be brought to the proteasome for degradation in step 5. However, and this is really important for how degraders function, this dissociation releases the degrader unchanged as shown in step 4. This allows the degrader to go back to step 1 and bind to another target protein and repeat the process. This process is what is known as the catalytic cycle and has key implications for how degraders are designed and optimized. It's been shown that catalytic cycle can be very efficient with anywhere from 300 to 3,000 cycles per minute for an optimized degrader. At C4, this is central to how we design degraders. Our goal is to create degraders which are maximally efficient in completing the entire catalytic cycle, not an individual step in the process. This is because our experience has shown that only maximally efficient degraders offer significant advantages over traditional inhibitors, which I'll outline on Slide 7. First and foremost, degraders are highly potent. Traditional inhibitors require one-to-one binding and high exposure to maintain disease control. In contrast, because of the catalytic cycle I just described, degraders amplify their potency because each degrader molecule can destroy more than one disease-causing protein molecule. Furthermore, because degraders are destroying the protein rather than inhibiting its activity, the only path to restore the activity of the disease-causing protein is its resynthesis. That means that the pharmacological effects can be sustained beyond the degraders' lifetime in the body. These combined effects mean that degraders can provide deeper and more durable effects in vivo than inhibitors. And you'll see examples of this in data from our most advanced programs. Second, fast degradation can lead to enhanced cellular responses, as the rate of protein depletion can also amplify the pharmacological effect of degrader. Third, the catalytic cycle can also improve the selectivity profile over degrader molecule because each step in the catalytic cycle can act as a filter. For example, the degrader may bind to other targets, but not all the targets will form ternary complexes with the ligase, and not all of the ternary complexes will tag the target for destruction. This filtering process is amplified through our drug discovery process, as we optimize our degraders for the specific target of interest. We've demonstrated this selectivity in our 2 most advanced programs, where both candidates have single selectivity for their targets. Lastly, this approach broadens the target landscape, as I referred to earlier. Traditional inhibitors require optimization of binding affinities to active sites that block the function of the target. In contrast, since the degraders function isn't to block the activity, but to connect the target protein with an E3 ligase, as long as a degrader can bind to a target, it can be degraded. This opens the landscape to targets previously deemed undruggable, and our BRD9 program is a good example of this. To summarize, there are a number of advantages to targeted protein degradation, however, it's important to note that they're only derived from maximally efficient degraders that optimize the catalytic cycle. At C4T, our goal has been to make the most catalytically active molecules. We've invested in our platform that allows us to do that for each program. Slide 9 outlines the 4 key elements of our TORPEDO platform that enabled this. First, our platform is flexible since we can utilize MonoDACs, or molecular glues, or BiDACs, or heterobifunctional degraders, depending on the features the protein we're targeting for destruction. This gives our chemists additional tools to bring an E3 ligase and disease-causing protein together to be tagged for destruction. Second, we focus our optimization efforts on making the most catalytically efficient degraders that optimize the overall degradation process instead of individual steps such as only optimizing ternary complex formation. This is because our data show that degraders with these properties lead to the most potent degrader medicines. Third, we've made a significant investment in computational methods and tools and enhanced our ability to rationally design degraders with improved or altered ternary complexes. These in silico models allow us to design degraders with enhanced potency and selectivity across our pipeline. Our cellular degradation assays provide volumes of high-quality data, which we analyze using a proprietary and unique framework based on fundamental enzymology principles. These models allow us to robustly predict both the depth and duration of the target degradation at any dose. The last important feature of our platform is our decision to use Cereblon as our ligase of choice for all of our key drug discovery efforts. This was driven by 2 key factors: first, Cereblon is the only E3 ligase that is clinically validated, as is the ligase used by the IMiD drugs, we have a rich and successful clinical history with these agents that demonstrate it can be safe and effective in the clinical setting; second, Cereblon is universally distributed across all tissues and cellular compartments in the body. This wide distribution provides great latitude in target selection and disease indication. The next 4 slides expand on these points in more detail, but at the interest of time, I'll move on to Slide 14. Our TORPEDO platform has led to a robust pipeline, with 4 programs expected to be in the clinic by 2022 and 5 programs in earlier stages of discovery, and all of our core focus is in oncology. There are also 9 additional programs with our collaboration partners that are outlined on Slide 15. These platform collaborations extend the potential of the platform beyond what we can do on our own. The Roche partnership was signed with the company's very beginning and is focused on 6 oncology programs. The Calico agreement was entered into in 2017 and focuses on Calico's disease areas of focus, namely diseases of aging. And lastly, the Biogen collaboration takes our platform into neurology, which is an exciting new therapeutic area for targeted protein degradation. Our lead program, CFT7455 can be fairly thought of as a next-generation IMiD, as it targets the same targets IKZF1/3 as lenalidomide and pomalidomide. An overview is found on Slide 17. It's well-known that depletion of these targets kills myeloma cells. In addition, there's an opportunity for the same mechanism in lymphomas. In both indications, we believe that the marketed medicines aren't potent enough to achieve maximal efficacy in these patient populations. Based on preclinical data that I'll share on the next few slides, we believe there's a transformative opportunity in multiple myeloma and non-Hodgkin's lymphoma indications for 7455. Slide 18 contains some of our foundational in vivo data, which has us so excited to move this molecule into clinical development. On the left panel, we're showing results from our study in the H929 xenograft model, which is the workhorse of multiple myeloma translational research. There are 3 key takeaways from the study: first, 7455 is markedly more active than pomalidomide at the clinically relevant dose; second, treatment with 7455 results in deep regression at doses as low as 30 micrograms per kilogram as a single agent; and lastly, in comparison to CC-92480, BMS' most advanced clinical stage IMiD, 7455 achieves comparable efficacy at 1/100 of the dose. On the right panel, we're showing the same experiment at the 100-microgram per kilogram dose, but plotted on a per mouse basis over an extended period of time. While we see the same tumor regression during the dosing period we see in the panel on the left, after dosing is stopped, those CRs were durable in all but 1 mouse out to 50-plus days. Slide 19 contains data that speaks to the relevant activity of 7455 in the setting of pomalidomide failures. This experiment uses a multiple myeloma model that's relatively insensitive to pomalidomide. On the left, you can see that 7455 demonstrated dose-dependent efficacy from 3 to 100 micrograms per kilogram, with tumor regression evident at doses of 10 micrograms per kilogram or higher. On the right, you can see that 7455 is active in pomalidomide failures. In this experiment, the tumors have grown for 17 days while being dosed with pomalidomide at a human equivalent dose with no signs of efficacy. At day 18, they were switched to 7455 at 30 micrograms per kilogram, and that resulted in a rapid and deep tumor regression, demonstrating that 7455 penetrates large tissues and is efficacious in rapidly growing IMiD-resistant tumors. Our IND for 7455 was filed in December and is currently under review by FDA. We expect to initiate our Phase I study in the first half of this year, and Slide 20 outlines the Phase I/II design, which allows us to accomplish 3 objectives. First, it will allow the identification of a discrete dose in both multiple myeloma and non-Hodgkin's lymphoma, which is critical because historically, these 2 patient populations have tolerated different doses of IMiDs. Second, by including cohort B1 and B2, it allows us to explore 7455 with and without dexamethasone in parallel, which is typically given with IMiDs to boost monotherapy efficacy. Given the potency of our degrader, we believe that we have the potential to drive meaningful efficacy as a monotherapy. Lastly, it allows us to collect early efficacy data, albeit in a small number of patients and the indications of interest. We're preparing now to begin our first clinical study with 7455 and look forward to updating you on its progress. Moving on to our BRD9 degrader, CFT8634, which can be found on Slide 22. BRD9 is a target you may be less familiar with. It's a member of the BAF complex, which is responsible for chromatin remodeling. Mutations in discrete subunits of the BAF complex results in discrete malignancies. One such mutation, the SS18-SSX translocation, results in BRD9 dependency and is a hallmark of synovial sarcoma. Additionally, SMARCB1 deletions are seen in other tumors and results in the same BRD9 dependency. These are examples of synthetic lethality. Only in the presence of the mutations as the cell have a unique dependency on BRD9, which provides an attractive target and minimizes the risk of on-target toxicity since healthy cells are not dependent on BRD9. There are 3 reasons why we're excited about this target. First is the biology that I just described. The second is that synovial sarcoma is an area of high unmet need. First-line therapy provides a median PFS of about 7 months and median survivals approximately 18 months, and there are no approved second line therapies. As a result, we believe that accelerated approval can be achieved with a second line uncontrolled study. And lastly, this is a target that can only be drugged with a degrader approach, and this is outlined in the cartoon on Slide 23. Inhibitors of BRD9 have been shown to be pharmacologically inactive in synovial sarcoma models. This is because existing BRD9 inhibitors inhibit only the bromodomain of BRD9, while the oncogenicity of BRD9 depends on sub-domains not addressed by traditional inhibitors. So inhibiting the bromodomain has no effect on synovial cell viability as shown in the left image. Our degrader eliminates the entire protein as shown on the right, thus replicating the potent antitumor effect we see in BRD9 knockout models. This is demonstrated on Slide 24. We've been able to design a potent and selective degrader of BRD9, CFT8634, which has demonstrated robust dose-dependent responses in the Yamato xenograft model and patient-derived xenograft models. We expect to submit an IND for this program in the second half of this year. Our BRAF degrader, which is partnered with Roche, is summarized on Slide 26. It was designed to overcome the liabilities of approved BRAF inhibitors. Inhibition of BRAF causes paradoxical RAF activation, which can result in diminished efficacy. In addition, resistance to BAF inhibition occurs over time, often driven by mechanisms which drive BRAF dimerization and capitalize on this paradoxical activation, such as BRAF amplification or the BRAF splice variant. By degrading the target instead of inhibiting it, it prevents the paradoxical RAF activation that allows for a deeper elimination of mutant BRAF signaling to create deeper and more durable responses. This is shown graphically in the cartoon on Slide 27. Slide 28 demonstrates this benefit with some of our compounds. On the left panel, we show improved efficacy compared to encorafenib with 2 of our degraders. And on the right panel, we show more durable efficacy with the third degrader when compared to dabrafenib. Importantly, this data demonstrates the point with 3 different degrader compounds, illustrating the breadth of our platform. We are in late lead optimization activities with this program and expect to begin IND-enabling studies this year. Similarly, the goal of our RET program is to develop a degrader that addresses the resistance that develops to approved RET inhibitors. An overview of the opportunity appears on Slide 30. While impressive efficacy has been observed with approved first-line RET inhibitors, there are no approved therapies for second-line or later treatment. As illustrated on Slide 31 in the upper path. RET inhibitors are effective in reducing signaling and halting cell growth. But like most inhibitors, secondary RET mutations occur, as depicted in step 3, which results in loss of the inhibitors activity, while a degrader's function is independent of the mutations and remains potent even in the setting of resistance mutations. Additionally, there is the potential to use a RET degrader in the first-line setting to potentially avoid the emergence of RET mutations. The next slide, Slide 32, shows our in vitro response data compared to the approved RET inhibitors and the next-generation RET inhibitor in early clinical development. What you can see here is that our degrader is the only molecule that shows potent activity across all 4, including both the gatekeeper and the solvent front mutation. Our RET program is currently in late lead optimization, and we expect to start IND-enabling studies this year. Putting all these programs together on Slide 33, we have a compelling pipeline of 4 programs directed to cancer indications with meaningful patient populations, where there is a significant unmet need not addressed by current therapies and are tailor-made for a targeted degrader approach. As we outlined in our press release last week, we have a number of important milestones across all 4 of these programs that allow us to make progress toward our goal to have these programs in the clinic by the end of next year. These are shown on Slide 34. As I mentioned earlier, for 7455, we're awaiting FDA feedback on our IND submission and are ready to start our Phase I/II study in the first half of this year. For 8634, our BRD9 program, we're targeting an IND submission in the second half of this year. And for BRAF and RET, we expect to initiate IND-enabling studies this year to position them to enter the clinic in 2022. So bringing this back together on Slide 35, we've really established a leading platform to develop highly-potent degrader medicines. We have a rapidly advancing pipeline with 4 programs that address meaningful areas of unmet need expected to be in the clinic by the end of 2022 and a robust pipeline of degrader medicines behind that. Our 3 collaborations enable us to expand the platform beyond which we can do on our own as well as provide an important source of nondilutive capital, and we have the capital resources to execute against the plans I just shared with you. Let me close on Slide 36 by thanking the entire team at C4T for all their incredible work to get us to this point, and look forward to updating you on all of our progress over the course of the year. Thank you all for your time. And now we have time for some questions. As mentioned earlier, please utilize the ask-a-question button.

Andrew, I'm going to read questions to you. The first one is, how do you prioritize technologies and target selection? And then specifically, how do you think about what targets are better suited for a MonoDAC or a BiDAC approach?

Yes. So maybe I'll start out at a high level on how we think about target selection, and then I'll turn it over to Stew to address the sort of the technical part of the question. So as we think about target selection, the technology is so broad, as I outlined, in terms of being able to go after multiple targets, we really have to be careful about how we select our target. The first filter is oncology, that's where we're going to focus as a company. And then within oncology, as we think about targets, really, the first thing we have to have is a strong degrader rationale. So is there a reason why you need a degrader because of inhibitor or some other way and do the same thing, it may not make sense for us to develop a degrader. So we have to have a strong rationale for why a degrader is the only way to go after that target. And then in cancer, we want to have an indication where there's genetics and there's a clear genetic link to that cancer, and those genetics are drivers, drivers of that cancer. We also want to evaluate the clinical unmet need [ of that ] indication as well as the commercial opportunity. And then I think lastly, we want to make sure there is at least a way to get chemistry to target the proteins we want to destroy, at least some starting chemistry. And as I mentioned earlier, the important thing is we don't need to have highly-potent ligands that -- because we're not trying to inhibit the activity, we're really just trying to attach to the protein of interest and get in close proximity to the ligase. So Stew, I'll flip it over to you for the other part of the question.

Thanks, Andrew. So I think the second part of that question was really a compare and contrast to the MonoDAC, or molecular glue approach, and the BiDAC approach, or the bifunctional one. And I think, first and foremost, you can see in our pipeline, the 2 lead assets represent examples of each. And so from my perspective, it's not really a technical hurdle in terms of optimizing these molecules to make them drug-like. We certainly have been able to do that in both classes, including what can some consider to be more challenging in the BiDAC space. So I don't really see the technical challenges being one preferred -- one approach preferred over the other, but there are meaningful differences in how you would apply that target selection framework that Andrew just lined out in terms of which approach to use because there are some technical aspects there that are meaningful. First and foremost, BiDAC approaches are easier to conceive in that if you have a chemical matter or you can find chemical matter that binds to a target protein, that actually allows you to then rationally build out a degrader. As I mentioned, there are med chem challenges there to overcome. These are larger molecules. But in our experience, we've been able to deliver those drug-like molecules in a relatively fast fashion. In contrast, the MonoDACs don't have that specific interaction built into the molecules. It's more of a protein surface protein interaction that's formed upon binding of the molecule. So there, we take, and believe that this is the norm, to take a more empirical screening approach to find targets that have activity in there. Then from there, we can use the TORPEDO platform to optimize the catalytic activity and so forth to get them to drug-like properties.

Okay. The next question is, can you please discuss your BD strategy around out-licensing assets versus developing them internally?

Sure. So look, our goal is to build a fully integrated company, and we want to really focus our platform and our team's efforts on making programs that we develop and commercialize. I think from a BD strategy perspective, as I mentioned earlier, we have these 3 platform collaborations. For now, that's really all we want to have. We don't want to expand because that would kind of take away from the resources working on our internal pipeline. From a product-by-product strategy, I think an example, our lead program is a good example. I think we're very comfortable taking the program through the initial clinical development plan that I outlined and potentially to accelerate approval. Given the dynamics in the myeloma space as well as the lymphoma space, being able to bring that program to earlier lines of therapy I think will be a challenge for a company of our size. So I certainly see a case to be made where we partnered that program at that stage to help a global partner with a little bit more development muscle and resources as that's more of a combination approach, so thinking about programs like that. But at the end of the day, our goal is not to become the CRO of the industry. So we'll look at collaborations where they make sense, but our goal, again, is just to make programs for our own pipeline.

Okay. And then a question somewhat related on the BD front. The question is, what can you say about potential business development combination strategies for myeloma specifically? And for an example, can you do something similar to SpringWorks' BCMA combinations? And then the second part of that question is just given the IMiD sales of about $15 billion, how are we thinking about the commercial opportunity for CFT7455?

Yes. So let me -- I'll start with the second one first because that's the easiest. It's too early to really predict what that opportunity is. Certainly, it's quite meaningful, but we're not even in the clinic yet. So we'll hold off from providing any the opportunity guidance until we see some clinical data. It's certainly in terms of combination strategies, I could certainly see doing kind of clinical development combinations where there's an important combination partner that we might want to do and put into development, and Adam may be able to talk about specific ideas there. But that's certainly something that we've thought about. Ideally, we could do 1 collaboration could bring both the sort of global clinical development reach as well as the right combination partners. But we'll see what the data looks like from the study we're about to do and then evaluate our path forward from there.

Yes. That about sums up what I would say. I think that it's clear the ultimate path for this molecule in myeloma may well lie in combination, not just with dexamethasone as we're doing in the first-in-human, but proteasome inhibitors and CD38 antibodies as well, but also the opportunity for combinations with some of the emerging therapeutic modalities, whether they be BCMA-targeted therapies or bispecifics. We certainly explore those areas now in terms of preclinical data, but I think, ultimately, we need to see clinical data in order to best determine the path forward with some of these attractive combinations.

So a couple of related questions. One, are there any significant risk to using targeted protein degradation in our technology? And then somewhat related, what are the biggest hurdles in terms of targeted protein degradation?

If I let, Stew, do you want to tackle that one? You're on mute.

So I would say that the hurdles or the risks going forward are somewhat project-dependent. And I say that because in large measure, the IMiDs have already cleared the way to understand what are the challenges or lack thereof of using Cereblon as a ligase for this modality. And the -- it still remains that the resistance seen for IMiDs is relatively low, and those relapses that are seen are largely based on reversal or lower Cereblon expression. So making compounds more potent does blunt that mechanism. In terms of hurdles in this space, I think this is an area where prior experience and say, small molecule drug discovery, is helpful, but can throw up blind spots and that these -- these molecules work through this catalytic mechanism, as we've highlighted. And in that way, the direct linkages to sort of conventional approaches of optimizing affinity and binding can be and often are found to be in our hands somewhat counterproductive. To optimize catalytic efficiency, we often have to reduce the binding affinity to make sure these molecules remain -- reduce the affinity, say, the target or the ligase binding affinity to make sure that they maintain or build higher catalytic efficiencies. And then the more complex BiDAC molecules, again, I think it is really following where the med chem takes you. These larger molecules have special properties. Highly-efficient molecules can overcome some of the traditional barriers to small molecule drug discovery, such as overcoming permeability issues or high clearance issues. A good catalytic degrader can work quickly and doesn't need to get very high concentrations inside the cell to exert its mode of action. So those are the things that, as long as you're open to these sort of perspectives, I think these hurdles can be overcome, but they are a challenge in terms of understanding these factors and then leveraging them.

Great. And our last question relates to CFT7455 and when we may expect to see data. And Adam, the question is specifically 2021 versus 2022.

Sure. So we are -- we submitted our IND at the end of last year, and we anticipate beginning dosing patients this half. I think it's fair to say that escalation will take time, and we need to get the dose right. I think it's unlikely that we will finish escalation this year, and we intend to release data at a medical meeting when we have a complete story to tell. And by that, I hopefully we meet a dose which we will move into expansion. I anticipate that will take about a year, perhaps more depending on the rate of enrollment and how many cohorts we need to enroll to determine that dose. And I think that, that is too early to say what that time frame looks like. But specifically, 2021 or '22, I think the answer is not likely in 2021, likely in 2022.

Great. Well, I just want to close by thanking everyone for your attention today and look forward to updating you in the future. Have a great afternoon.