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

Good morning, and thank you, everyone, for joining the Virtual UBS Global Healthcare Conference. I'm Colin Bristow, one of the biotech analysts here at UBS. And it's my pleasure to have our top pick C4 here with us today. So speaking on behalf of the company, we have Andrew Hirsch, CEO; Stew Fisher, CFO; and Adam Crystal, CMO. So Andrew and team, thank you for joining us today. Before we start, if anyone has a question, there is a question function in this global meet system. But alternatively, feel free to e-mail me at [email protected], and I'm happy to fill the question for you. So with that, I'll hand it over to Andrew. Please go ahead.

Yes. Thanks, Colin, and thanks to -- for UBS for hosting us. I did want to note that we have some exciting news that we announced this morning. As many of you know, we have been evaluating whether or not to take our EGFR program forward into the clinic and this morning we're really pleased to announce and excited about this program that we are going to move that forward into the clinic. I'm sure we'll talk about that as we get through your list of questions through the chat. But we're happy to answer any questions we can. Obviously, we didn't have any data today. That data is going to be presented at the Keystone meeting, and we'll have an investor event around that to go through a lot of the data. And that's in our press release we issued this morning. So I'll turn it back to you, Colin, and fire away.

Super. And thanks for timing that program advancement with our conference. Appreciate it. So let's just start high level. On a -- from a platform perspective, there's obviously numerous degrader companies out there now. It's a hot space. How is C4's approach and platform differentiated versus your peers?

Yes, it's a great question. One we get a lot. And I think there are clearly differences in approach across the companies. It's really hard for me to compare since I am intimately focused on our platform. And I know what you know and it's publicly available about the others. But what I'll really do is focus on what makes our platform unique, and there are really 4 key elements to that. And first, is really central for us. We focus on optimizing the entire degradation cycle rather than a single step in the process. And so for those of you kind of new to the greater space, one of the key differentiation features of a degrader versus inhibitors is that once they've done their job of enabling the destruction of a single protein molecule, target protein molecule, they can actually go back and do it again, and we call this the catalytic sort of cycle. And that can range anywhere from 300 to 3,000x a minute. And so we believe, though, that optimal degraders do this towards the 3,000x per minute part of the range, and that's where you get maximum efficacy. You get the real benefits of degraders, and you can see differential biology. And so that's -- our lead program is an example of this. I mean if you look at the potency and the catalytic -- the improved catalytic efficiency of our fleet programs, 7455 versus pomalidomide, for example, we actually see -- we induce apoptosis in our in vivo models, and you don't see that. And so that's where really optimizing that catalytic cycle kind of where we make the best degraders. And our entire approach is really designed to accomplish this. And as I said, it's really fundamental to how we think about optimizing degraders. Secondly, kind of following along that, we've developed a framework of assays, methods and tools that we use to really translate the kinetic properties of our degraders that we see in vitro into highly predictive models for in vivo performance. And those tools allow us to really move quickly and with confidence to get to a drug candidate nomination. I'd say the third key feature of our platform is that we really focus deeply on Cereblon. It's a decision the company made very early on. It's one of the E3 ligases that are in the body. And we chose it for 2 reasons. One, it's widely expressed in all tissues and compartments. So it gives us quite wide target latitude. We know that what ever target we want to go after, Cereblon is going to be there to do the work. And secondly, it's the only really clinically validated ligase since it's the ligase of the image drugs, which have been on the market. There's a long track record. And we know -- we know what those do. And the other -- the last reason is we haven't actually seen the need to go beyond that because we've worked on about 45 targets over the history of the company. We've had about a 95% success rate with Cereblon. So we haven't seen the need to do that. And then we've also made a deep investment there. We've got about 15 distinct chemical series of Cereblon binders, and that really enables us to optimize degraders. And then I would say the last key feature is that we have the capability to do what we call MonoDACs or for the generic term molecular glues in BiDACs or heterobifunctional degraders, and both can deliver viable clinical candidates. For example, our lead program is a MonoDAC and then the EGFR program, for example, we announced this morning, as well as the BRD9 program, those are BiDACs. And what that enables us to do is really match the target selection with the discovery optimization process. So there are certain targets that lend themselves to one or the other. And so we have that flexibility given the targets we won't go after. So I think those are the key 4 elements that are really important to how we think about differentiating and what's important for optimizing our degraders. At the end of the day, we think that our peer companies, their approaches are complementary to ours. And given the breadth of the degrader opportunities, I think there's room for everyone, given that there's very little target overlap across the degrader companies. And so I think there's really -- it's a wide open space, as you mentioned earlier, it's really exciting. And we're really, really happy to be a leader in the space.

Great. Thank you. So I mean, it's a good segue actually to the fact that your lead program is a MonoDAC, which is, I think, fairly unique amongst your peers. And can you talk about some of the complexities and technological challenges of developing a MonoDAC versus a BiDAC?

Yes. I mean, sure. I mean I think there's a number of different ones, and I'll let Stew go into some of the technical details. I think for us, we're really flexible. We don't -- we're not challenged either way. Once we get to sort of hit ID, all the tools and processes that I talked about earlier are really the same in terms of optimizing the degrader performance, but I'll let Stew give some of the technical details.

Yes. I think Andrew has really highlighted that these 2 approaches are truly complementary and it's not that we see fundamental differences between the 2 and our ability to optimize them to drugs. That's really been the case, as exemplified by our pipeline, we've been able to advance both BiDACs and a MonoDAC to drug candidate status. So I think it really comes down to target selection and which of these 2 approaches lends itself best to get a good degrader hit that we can then utilize our platform to optimize against. So we don't really see any kind of challenge or benefit to either approach, but they do require a subtly different med-chem approaches, but both of those are certainly within our capabilities.

Okay. Great. And some of your peers are using tissue-specific ligases. Just how do you think about that conception? Is it something that's attractive to you guys at some point?

Yes, I'll field that one. I think the vision of a tissue specific ligase is quite attractive, and it's something we have talked about quite a bit at C4. I think the -- because in that vision, you can see perhaps the most precise medicine available where you'd be able to target a specific indication with a ligase that's expressed in that particular tumor. I think the challenge comes, at least at this point in the field on how to reduce that to practice. On the one hand, one should ask, what kind of targets would lend themselves to that to get maximal benefit of a tissue specific ligase? And on that front, I think if you're -- by and large, I think most of this value in this approach lies in having a more safe profile. And that is that you don't have to worry about off-target or off-tissue mechanistic tox. That's what that targets tissue-specific ligase would bring. I think that opens up targets that would otherwise be panessential or toxic to other tissues. So it opens up more landscape for targets, but I think that does put a lot of pressure then to ensure that, that ligase is only expressed in those tissues. Because then if you use those kind of targets, then you're going to lend -- you're going to lose that benefit of that selectivity. On the other hand, and so -- and I would say one of the challenges I see with the ligase -- the [indiscernible] as it is, is most of these ligase are generally very poorly characterized. We don't really understand their biology. We don't really understand their location and their expression levels or cycling within a different cellular differentiation state. So I think that there's risk in trying to understand that and then deploy it at this stage. I do believe, in the long run, we will do this, but I think there's a lot of fundamental biology and chemistry needed to lift those ligases off the ground and deploy them as druggable entities. We're interested in participating in that effort, but it's not something we see that actually can be reduced to practice in the near-term to deliver medicines to patients. And so that's what we focused on in our first phase of C4, but I think, in our future vision, certainly, it does include downstream efforts in additional ligases to incorporate -- include applications such as tissue-specific ligases.

Okay, great. And then maybe moving to some more specifics of 7455. What is -- do you foresee a potential benefit in an asset that could preferentially target Ikaros versus Aiolos in treating multiple myeloma?

I can fill this, if you like, Andrew.

Yes, Adam, why don't you take care of this one.

It's hard to say with confidence that, that will be the case. And I point to data in really multiple myeloma where maximum cell [indiscernible] of those myeloma cells requires knockout of both IKZF1 and IKZF3. Knocking out either has some effect on cell viability, but maximum effect requires knockout of both. And to some extent, this is true in the NHLs as well. So we believe that our molecule is really optimized for both the multiple myeloma and NHL spaces. Furthermore, I think it's fair to say in our hands the SAR is really inseparable in targeting IKZF1 and IKZF3. It's possible that it could be achieved. But I think certainly a challenge, and we see no benefit in doing so.

So just remind us, the 7455 have equivalent patency and degradation kinetics on both the targets?

That's right.

Okay. Okay. And then I know this is a question you get a lot, but could you just run us through how 745 is differentiated versus Bristol's CELMoD, iberdomide 9480?

And yes, absolutely. So we really optimized CC19 -- I'm sorry, we really optimized CFT7455 for potency on target being IKZF13 as well as catalytic activity. It is a fast molecule. And we really believe this molecule has the opportunity to differentiate on 2 things. One would be simply efficacy. And what I'd point to is the in vivo data we've shared before, demonstrating that we achieved equal efficacy as CC-92480 to a 1, 100th the dose. We believe this positions us well clinically to translate those to efficacy, which may be superior to anything that is currently available, whether approved or experimental. And the second is that really on the back of this exquisite potency in terms of in vivo activity, we believe there's a differentiating opportunity to develop this molecule as a dexamethasone sparing agent. So what I mean by that is that, historically, IMiDs as well as the experimental CELMoDs, have been developed and explored in combination with dexamethasone. And the primary reason for this is because dexamethasone boosts the overall response rate and efficacy of Len and Pom and presumably through 20. However, the feedback that we've gotten that we agree with entirely from KOLs that our molecule is so active as a single agent, meaning no -- I'm sorry, meaning no dex, that we may well be best served by moving it forward in the absence of dexamethasone. And our first-in-human study enables us to find a dose of 7455, both as single agent as well as in parallel in combination with dexamethasone and move either or both those forward into expansion.

Okay. Great. Just 1 question in my email, actually. How important is the rapidity of degradation in terms of realizing a clinical benefit?

I think it's critically important. And I think that the data that speaks to that clearly is actually in the literature as well as data we've generated internally, which demonstrates the effect on cells, which has had at active doses or clinically achievable doses. And I'm speaking vaguely now and I'll clarify, what I mean by effect on cells is stasis versus apoptosis. And to reach the conclusion I'm getting to really is that, with the slower molecules, what's achieved is largely stasis. And with the faster molecules, what's achieved is largely apoptosis. And I believe, to achieve a maximal or optimal antineoplastic effect, you really want to be killing those cells and apoptosis certainly achieve that.

Great. Great. Now this is something we've talked a lot about, but AACR, you presented some data on 7455's activity in Cereblon downregulated or partially depleted H929 cell lines. Could you just -- could you walk us through these data? And what should we take away from it? And why is it important?

Yes, absolutely. So effectively, the model we created there was a model of resistance to Len and Pom, where you take H929, you grow Pom and Len, Cereblon levels drop, they become resistant to Len. Then you switch them to Pom, Cereblon levels drop a little bit further and they're resistant to Pom. And then what you have is a cell line, which has resistance to unit molecules we just discussed in down-regulation of Cereblon. It's important to note that this model is potentially clinically relevant. It's published in the literature that patients who are on IMiDs do demonstrate down-regulation of Cereblon. And to some extent, that correlates with resistance, I don't want to say, to reach a conclusion to what degree it causes resistance, but it's certainly a reasonable mechanism of resistance. Despite that down-regulation of Cereblon and the active resistance to Len and Pom, our molecule retains activity in that cell. And this makes sense, right? One could hypothesize that the cells that learned to survive Len and Pom by downregulating Cereblon to the point that Len and Pom can't drive enough IKZF13 degradation to kill the cells. But our molecule is more potent and faster, effectively, meaning it requires less Cereblon around to do the job, so it retains activity. And in my mind, what this translates to is the very real possibility that our molecule would be active in an analogous setting [indiscernible].

Okay. Great. Do you have -- or have you established a preclinical model, a theoretical threshold of Cereblon depletion that would render 7455 inactive?

That's a super great question. I haven't actually thought about quite in that way before. To be honest, we've looked at the flip side of it, which is, to what degree do we need to degrade target for it to be maximally active. And I think, there, we do want to have greater than 80% target degradation what we've modeled is for a portion of the 24-hour dosing cycle, though I'm confident we'll be able to obtain that for the entirety of a 24-hour dosing cycle. But I think to what degree down-regulation of Cereblon would confer resistance is a really hard question to answer. The literature is actually challenging, meaning one can look at baseline levels of Cereblon before a patient ever gets an IMiD and see whether or not that predicts responsiveness to an IMiD and the short answer is that the literature is mixed. So even that sort of symbol first past question with clinical data is very challenging to answer.

Okay. Now a question we get a lot, I'm sure you do too, but [indiscernible] trials, your -- the Phase I trial initiated in mid-April. In your press release today, you said sites are open and enrolling. We still not had or confirmation of a patient dose. Have there been any delays? Can you just confirm on here that we're on track? Can you talk us through that?

Is super fair question. It's an exciting milestone for us. We are on track. There have been no delays, certainly no significant delays. We are active and ready to enroll, and we are confident that we will meet our goal of achieving the first patient, first dose in the first half this year.

Okay. And now playing those time lines forward, I guess, kind of like a 2-part question. When would you reasonably expect to be able to provide us with the data set from that trial? And then the second part of that question, again, is something we get, I'm sure you do, is, what is the potential for us to see something from you guys before year-end?

Very important question. We have consistently been guiding to 2022 for the release of clinical data. And this is really based on 2 important pieces of data that drive this. One is the timing. If we dose our first patient in the next month or so, we'll have 6 months until the end of the year. In my mind, that is not enough to put together a story of a data set. It's enough time to put together a few data points, which would have a tremendous amount of [ air ] bars in all of those points. And would just be too low to have confidence that it was representative of the true data. And so I think, ideally, what we would like to push forward is a set of representing the RP2D of 1 of the cohorts in our study well, along with the PK and PD and safety data, and hopefully, efficacy data. I don't know how many dose levels it will take to achieve that. And I think that would be the ideal to have a dose. But I think that, at some point next year, it is fair to expect us to have when we would consider a publishable unit, and we would go forward at that time with the aim of presenting it at a medical meeting.

Okay. Great. And then as we think about that readout and the trial design, how many patients' worth of data should we expect to get at that Phase II go-forward dose?

Sure. So there are a few ways to answer that, and they're all straightforward, but they're different. So it depends on whether we're talking about -- there are 3 basic doses that will determine potentially, right? A dose in NHL, a dose in myeloma as single agent and a dose in myeloma with dexamethasone. So each of those, I would imagine, would have in the range of 12 to 15 patients, if we do this well. And in some cases, the data feeds into each other. So it really depends, I would say, something in the range of 12 to 15 at each of those dose levels, may be a little bit lower in some cases. And it depends on whether or not we present the entire data set, right? These are the doses for all of these conditions or we're simply going to publish single agent myeloma single agent NHL.

And just I'll add to -- some of it depends on how many escalation cohorts we go through, right? I mean that's obviously the big driver here. So -- and we don't know until we know.

Absolutely fair.

And again, I think we talked about before, but can you remind us, what are you hoping to see in terms of efficacy on -- at the Phase II dose?

At the Phase II dose. Yes. So I think this is one of the most critical questions for the study, how are we benchmarking efficacy? And I think the trial is intentionally designed both to -- so that the patient population in myeloma mirrors that in the Phase I with CC-92480. This is for 2 reasons. One scientifically, we think is the right population to go into. And second is that it allows us to use that data as a reasonable benchmark. Obviously, Phase I uncontrolled data, not perfect, but the best we can do. That data set is still emerging. But what CC-92480 published in ASCO 2020 was at their RP2D, which was 1 milligram plus dexamethasone, 3 weeks on 1 week off, in about 20 patients, an overall response rate of about 50%. And that is PRS dex. So effectively, we want to compare ourselves to that. Now that number is challenging by something that all such numbers are challenged by, low end, right? I don't know if that 50% is really 35%, 70% or 50%.

Right.

But what we're aiming to be is, at our single-agent RP2D, just just CFT7455, I'd like to see us be as good or better than 92480 plus dex. And with out our molecule 9455 plus dexamethasone, the goal is to be better than the activity achieved with 92480 plus dexamethasone.

And let me just add, though, right, that the data we're talking about sharing in 2022 is dose escalation data. So you're likely to see a mix of patients, right? At different doses versus the data Adam referred to with 92480, which was 20 patients at a single dose. So I think that's the other sort of caveat to just be careful of when folks will inevitably compare the data we released next year.

Sure. And actually, another question on the trial design, but how are you guys managing cytopenias, et cetera? And how does this contrast to the Bristol program?

Yes. I have to say, of course, I don't know the details of how the Bristol program is managing it. There are constraints that any Phase I trial has in terms of what supportive care, meaning growth factor can be given during cycle one. Even during cycle 1, it effectively becomes challenging, if not impossible, to identify the appropriate dose and dose [indiscernible]. But after that DLT evaluation period, we do allow support for cytopenias to be given, really in line with [indiscernible].

Okay. Okay. Great. Maybe we can switch gears now to 8919. So obviously, the talk of today -- and just take us a step back and walk us through this program. How we got to where we are? What you have seen that's now giving you confidence to take this forward?

Sure, Adam, do you want to tackle that?

Yes. I mean, I think it's fair to say we're extraordinarily excited to move this forward. Really, the idea of the molecule frames this out. And what we're putting forward is an orally bioavailable allosteric degrader, which specifically targets EGFR L858R, the driver mutation in something like 40% of EGFR positive non-small cell lung cancer, and also covers the secondary mutations, which occur in EGFR, which often cause resistance by their first, second or third-generation inhibitors such as the T790M gatekeeper mutation, the C797S mutation, which just strides the covalent binding side of Osimertinib. We think there's a very clear path to develop this molecule in the setting of resistance to Osimertinib in patients who have such mutations. And I think it's fair to say that we also think there's a very viable path if the clinical data reads out, as we hope it will, to identifying a role for this molecule in that upfront space. One of the thing -- there are a few things that are different about this molecule than all other kinase inhibitors that are out there in the clinical realm. And I think a couple worth pointing out, obviously, it's a degrader this comes with potential advantages. I think it's fair to say, osimertinib is a wonderful molecule. It's a wonderful kinase inhibitor [indiscernible] selective inhibition of muted EGFR, activity and tolerability. But 2 things I would point out are that, overall, the PFS is 17 months. So while the overall response rate is spending about 80%, 17 months is something that, I think, we, as a community, should aim to do better for in the first-line setting for EGFR immune lung cancer. And I think that a degrader approach might provide the opportunity to sort of break through that, let's call it, 17-month barrier that you see with Osimertinib, that I think would be very challenging to do by improving with the better kinase inhibitor. There are reasons to believe a degrader could do it. The other thing that I would point out is that this is a specific -- an allosteric degrader, which specifically targets the L858R mutation And that L858R mutation is something of a liability for Osimertinib. While the overall PFS in the frontline study, which resulted in the approval of Osimertinib is 17 months. It's different for L858R the activating [indiscernible] mutation. Really, in L858R, it goes down to about 14 months in comparison to about 21 months for Exon 19 deletion, which is to say the bar for this molecule, even in the front-line setting, isn't the overall 17 months plus, but lower 14 months, and we believe that creates a real medical unmet need that our molecule has the potential to address.

And I would add 1 other thing too because it's an allosteric binder. It doesn't compete with the inhibitors. So we'll see what the data looks like, but there is the potential. It doesn't compete, so that could be used in combination with a kinase inhibitor if that was warranted based on the data and based on the treatment approach.

Okay. And timelines for this program? I mean, can you help us think through to when we could potentially see some clinical data?

I mean I think it's easiest to follow the press release on this one, which really points to IND filing mid next year. And I think that following from there, I would apply sort of to your standard time lines for IND filing the first patient in, and then I would expect something in the range of a year before we had the [indiscernible] complete story that we just articulated, referring to publish rather than a few data points.

Okay. Great. Maybe let's talk about the BRD9 program now. So this is a target that's not really been clinically validated. Can you -- what do you foresee as some of the potential risk to this program?

Sure. So I think it's a very fair question in that any target, which has not previously been pursued clinically with the risk that tried and proved clinical targets do not. I think that for this particular molecule, there is a strong scientifically big rationale to believe that effectively therapeutic index will be there. And that's really based on the concept of synthetic reality, meaning that in cancers, which are susceptible to depletion of this target, such as synovial sarcoma or smart V1 [indiscernible] tumors, that susceptibility is specific or particular to the tumor cell by virtue of the genetic lesion that made them susceptible to BRD9, by virtue of the genetic lesions would drive the tumor, they are susceptible to BRD 9 FBSS [indiscernible] location and Sinovant smart V1 [indiscernible] 2 parts and the others. And normal cells don't have that it creating that better. I think the other real piece of data that increases our confidence that BRD9 degradation would be tolerable is data that recently was released at ASH in abstract form by [indiscernible] et al, along with Omar [indiscernible] and others. And really what this showed, which was of note to us, is that additional knock out BRD9 in the hematopoietic compartment really results in these [indiscernible] tolerable [indiscernible]. There's a little bit of a myeloid shift, [indiscernible], there is a little bit of thrombocytopenia, there's a little bit of upregulation of hemoglobin, but these mice were relatively unaffected by that. And so I think there is very real reason to believe that it would be tolerable. I think in terms of efficacy of the data is extraordinarily strong, meaning there's the dependency on BRD9 and exciting for the target, and the reason we like it so much for one of them is that you need a degrader. What I mean by that in a straightforward manner is that, for example, synovial sarcoma cells don't care about BRD9 inhibition. One can take a promo domain inhibitor, put it on synovial sarcoma cells, and they simply don't care, there's no back [indiscernible]. Instead, if you deplete the target basically replicating what's seen when you air pan out or knock out the target, you see a real effect on cell viability. And in vivo, we have demonstrated really nice regressions [indiscernible] with [indiscernible]. So in short, any novel target has risk, but there are very real reasons to believe that going after this good molecule like ours will result in [indiscernible] as well as an appropriate therapy [indiscernible].

Okay. Great. And so obviously, synovial sarcoma a very tough to treat tumor, nothing out there and potentially lending itself to a faster market strategy. Can you outline what that looks like from the development path for your BRD9 degrader? And help us think about what do you think you need to show in a potential Phase II registrational setting?

Absolutely. So I'll be clear, we have yet to finalize our clinical trial design, though I think I know what it's going to look like. I think it will frankly be straightforward. An escalation of single-agent finding the dose in tenant expansion in a limited number of patients, let's say, 20 to 30, to see whether or not we achieve a predetermined signal, which would allow us to open up that expansion further to a number of patients which would allow us to achieve that accelerated approval. I don't want to speculate exactly on what that number is. But I think the fact that it is a relatively rare indication would put downward pressure on the number required and that additional expansion. And I think further speculation is also challenging because it will depend on both the data that we have, the degree of efficacy, which has [indiscernible] as well as interactions with health authorities. So I think, in short, find a dose, expand in a smaller number -- in a small-ish number of patients to get a better sense signal and then determine the minimal number of patients we need to further enroll to get to that accelerated approval.

Great. Thank you. Any other indications that you think targeting BRD9 can lend itself to?

Absolutely. So there is a population of patients who have smart V1 deletion, which biologically creates this dependence to BRD9 as well by a very similar mechanism. This includes things like ATRT and [indiscernible] tumor in [indiscernible] sarcoma. And together, there a meaningful number of patients. We think that the best path to initial registration is in synovial sarcoma, but the data sense that if it works there, it should work in these other tumors as well. And then beyond that, I think it's fair to say that there's a large literature supporting that BRD9 does play a role in other malignancies. And our presence is not exactly clear, the best way to use that in other indications, but there's data in AML, there's data published in ovarian cancer, there is data published in prostate cancer that suggest a role. So we keep very close tabs on this and are hopeful that with time, there will merge very clear opportunities to use BRD9 degrader [indiscernible].

Great. Thank you. Maybe we can move on to the BRAF and RET programs. Can you just -- a similar line of question, can you walk us through what you've seen there that got you excited? And then the sort of development timelines and plans there?

Sure. So I'll speak to BRAF, and then perhaps I'll pass the time to Stew to speak to RET. I think BRAF is extraordinarily exciting for us, really, for 2 reasons -- 3, I would say. The first is the rationale. This molecule effectively degrades Type 1 BRAF mutant, such as the B600E mutant. What this means is that this monomer is degraded, it cannot incorporate into the dimer and can't be active in that dimer setting, which occurs with the approved RAF inhibitors by virtue of paradoxal activation. So it really overcomes a potential liability of all of those approved Type 1 RAF inhibitors, all of the approved RAF inhibitors, which may indeed limit their efficacy and the emergence of resistance. And I think that translates the 2 other things that we're excited about, which is really data. And the data that has me particularly excited to focus into patients is data that we have made public before in the S1 and elsewhere, which demonstrates that our molecules are, in comparison to approved RAF inhibitors, able to achieve a deeper and more durable regression. And what I think fair to say is a model of the upfront setting and better retains activity in the resistance. So in particular, what we did was we knocked in [ NRAS ] 261K, a Clinically bonafide mechanism of resistance to rapid [indiscernible] in melanoma into an a 375 melanoma model, which has B600E. And in that setting, as you would predict, those models are insensitive to RAF inhibitors. However, our molecule retains activity in that setting, suggesting that really could have [indiscernible] activity in the relapse setting that has been hypothesized. So strong rationale, deeper and more durable in the upfront setting that improved RAF inhibitors preclinically and activity in their resistance setting.

Great. Thanks. And then...

Stew, do you want to tackle RET?

Yes. I'll talk on the RET program. So we remain very excited about the RET program potential. But we're also quite aware that this is a competitive space. Presotinib and Selpercatinib are really excellent drugs, and that is playing out in the clinic. They have limitations, for sure, and that is that they're prone, both of them, to solvent front mutation, and that is emerging in clinical data. So the goal for our program is to have a degrader that covers all the potential mutants that are -- that can emerge, both gatekeeper and solvent front mutations in addition to covering the fusions that cause the disease as well. So we're looking for a molecule that is really truly best-in-class. What I would say is that we're making great strides on that. We've actually got molecules, which do cover with within a tenfold potency range all of those mutants and the frontline sort of fusions that drive the disease. But we want to make sure that any molecule we put forward is going to have the profile that can be best-in-class. So we're taking not our time, but we're taking great care in making sure that those compounds that we do advance have all the features of the best-in-class molecules.

Okay. That's great. Maybe just a quick question, switching back to 7455. Bristol seems to have a differentiated strategy in terms of taking different assets into multiple myeloma versus NHL. Do you -- what do you think the reason is for that? And what gives you the confidence that 7455 is kind of the right hammer for both those nails?

Sure. I think I would say that we don't have any particular insight into what's driving BMS' strategy. And so I'm very reluctant to speculate on that. I think that we are driven by the data that we have as well as our understanding of the biology, which we discussed earlier, really that for these indications, you really want to be crushing IKZF1 and IKZF3. So we designed a molecule to do that. We believe we've achieved it. It has the opportunity to be best-in-class. And as such, it's really it's really appropriate, if not optimized, for both of those indications.

Okay. Great. And then maybe just as we go into the end of time, can you talk a little bit about just your cash position and burn rate and the runway that gives you?

Yes, sure. I can [indiscernible]. Yes, I can tackle that. So as we reported in our latest filing, we ended the quarter with $346 million in cash. Obviously, with the announcement this morning, moving the EGFR program into the clinic, that's going to increase our spend a little bit. It's really skewed toward late '22 and '23 as the clinical trial ramps up. Today, we're not really providing an update on runway guidance. We're excited to really move this program forward. But don't have an update for you on runway today.

Okay. That's great. And anything you can tell us about beyond the assets we've talked about, I know you're working on some sort of transcriptional control, you have partnered programs. Anything we should be watching out for that?

Yes. So I'll say, on the partner programs, obviously, we very -- we're not really allowed to talk about many of those. We're excited about the progress we've made. We've made really excellent progress across all 3 of our collaboration partners and the programs within that. And hopefully, as we get closer to moving them, moving those into the clinic, we'll be able to share more. But until then, I really wouldn't expect any any communication -- beyond BRAF, obviously, which is the [indiscernible] the Roche collaboration, there's not really much we're allowed to say. And so I wouldn't expect there to be much there. And then similarly, I think we have some really exciting targets that we're working on across both MonoDAC and BiDAC approaches that are earlier in our pipeline. At this point, we don't really want to share anything about them, including what the targets are or time lines. As we get closer and have kind of line of sight to develop a candidate and think about what that clinical program and timing is, we'll share more. But for now, not much to say. Stew, you want to add anything about how we're thinking about some of these targets, maybe at a high level?

Yes. I mean I think what we're -- at the highest level, we -- it goes back to where we started this conversation, and then we have capabilities both for BiDACs and MonoDACs. I will say this, we're investing heavily in the MonoDAC approach to further expand our capabilities there and really complement the strength that we have in the BiDACs and that we've demonstrated to date. So I would say, going forward, what you'll see is C4 continue to expand on its existing platform, but grow most heavily in the MonoDAC capability because that's where we feel we can truly provide complementary approaches to some of the most challenging targets out there. That's really our vision.

Okay. And then just, I guess, as a continuation from the partnership discussion. As you think about moving towards commercialization, 7455, even maybe sooner H634, would you look to partner these on a commercial basis? Or is this something you want to go out alone?

Yes. So let me talk about 7455 first. I think that we're very comfortable as a small company, taking that program through kind of the Phase I/II that we've outlined and through potential accelerated approval based on what the data looks like. Obviously, this program, we think, has greater potential than just that in penta-refractory myeloma or relapsed refractory non-hodgkin's lymphomas. But we recognize that moving this program forward into second line, third line, et cetera, it's going to require large global studies with comparator arms against the combination regimens that are the main stains of myeloma treatment today. And that's something that we're probably unlikely to be able to execute kind of well and would be best served doing that with a partner. So likely for that, we would. We'd obviously want to keep -- obviously, keep some role in the program. It's an important program for us. We think has really, really large potential. So we're not looking to out-license it. But certainly, we would want to have someone to collaborate with and help really maximize the value of this asset to patients across all lines of therapy and all indications where we think the drug will have utility. BRD9 is a very different story as well as, frankly, I think EGFR. I think those are indications where a company of our size can move that forward on our own and frankly, commercialize ourselves in targeted focused markets. And so that's something when we get there, we'll likely consider doing.

That's great. And I think that brings us to the end of time. So this has been fantastic. Thank you. Thank you, Andrew, Adam, Stew, this is super insightful. If anyone has any follow-ups, feel free to e-mail me at [email protected]. Thank you, everyone, for dialing in, and yes, have a great day. Thanks guys.

Yes. Thanks for having us.

Thank you.

Have a good day, everyone.