Good afternoon, everyone. Thank you so much for joining us at the UBS Healthcare Conference. Very happy to have Revolution Medicines here with us today for a fireside chat. Joining us from RevMed is Mark Goldsmith, President and CEO. Mark, thank you so much for making the time out of your very busy schedule for this fireside chat. Maybe just to kick it off, you guys have a pretty unique platform technology approach to treating RAS-driven cancers. Can you just give us an overview of your platform, how it works, as well as the pipeline overall?
Yeah, thanks very much, Ellie. I appreciate the chance to be here at the UBS meeting and the chance to talk with you. Revolution Medicines is focused essentially exclusively on RAS-addicted or RAS-driven cancers. RAS mutations are found in roughly 30% of all human solid tumors and represent a significant unmet need. We have taken an approach to inhibiting RAS as a cancer driver that is differentiated from others in the field. We wanted to find a way to bind to the active or on state of RAS, which we call RAS(ON). That's the form of RAS that causes excessive signaling and causes tumors to grow. And we ultimately acquired a technology as a foundation for doing that. That technology involves designing a small molecule compound that will enter a cell, bind to a highly ubiquitous protein called cyclophilin A.
And that combination of the small molecule and the endogenous cyclophilin A has new binding properties. It can now bind to the on or activated state of RAS. And that's something that had never been achieved before with a small molecule. So this represented to us a very differentiated opportunity, and hence we acquired that technology by acquiring a company called Warp Drive Bio. We've now, since then, that was in 2018, since then have designed more than 20,000 compounds like this. They're very sophisticated molecules. Each of them has a different property. They may be targeted to a particular mutation or to a broad class of mutations. And based on this technology, we now have three clinical stage assets, each of which has shown proof of concept, clinical proof of concept in human cancer.
The most advanced of which is now in a phase III pivotal trial in pancreatic cancer. I'm sure we'll have a chance to talk about that. This technology has gone from its very earliest stages, when nobody was sure if it could work, to now being a very productive, competitive, industrial strength drug discovery platform.
Great. And yeah, maybe starting with pancreatic, how do you think about the opportunity here and the unmet need? And then we can go into some of the specifics of the clinical data.
There's an enormous unmet need in pancreatic cancer. Many people know somebody or know somebody who knows somebody who's had pancreatic cancer, and likely they're not here with us any longer. There's 60,000 new cases of pancreatic cancer each year. There have not really been highly effective treatments for pancreatic cancer. The current standard of care is chemotherapy, which provides a few months of progression-free survival typically in later lines of treatment, and a few more than that in earlier lines of treatments, and treatment is not well tolerated. It's sort of a dismal experience, unfortunately, so there's an enormous unmet need for first line and second line and later lines of pancreatic cancer treatment. Many companies have essentially broken their picks trying to drug pancreatic cancer, but if one can crack that nut, the opportunity is quite large, and patients will be very thankful.
The key discovery that really drives all of this for us is that essentially all of pancreatic cancer is driven by RAS or RAS mutations. That makes it a unique disease in that it is essentially entirely a RAS-driven disease, and so if we can effectively target the RAS signaling that drives any of pancreatic cancer, we can probably target all of the RAS-mediated pancreatic cancer and essentially all of pancreatic cancer in that regard. Fortunately, our first compound, RMC-6236, which is designed on the basis of the technology I described earlier, it's an oral once-a-day drug, investigational drug. It binds to all forms of RAS, KRAS, NRAS, and HRAS, and it binds to virtually all mutant forms of RAS, regardless of where the mutation is, and with that very bold profile, we brought that into the clinic about two years ago. We did a dose escalation.
We began seeing clinical activity very early in the course of that dose escalation, and we've now proven through extensive study of several hundred patients who have been treated with RMC-6236 that it's a very active anti-cancer agent in RAS-driven cancers of multiple types, and in particular in pancreatic cancer, and we've now just shown a few months ago and then confirmed with a more recent update in October data in second-line pancreatic cancer that carried with it a median PFS of 8.5 months. The standard of care chemotherapy is known to provide about three months of median PFS, and a median overall survival of 14.5 months, and the median overall survival from standard chemotherapy is around six, six to seven months, so if those numbers could hold up in a randomized controlled trial, then that would represent a remarkable advance for patients.
The drug is also well tolerated at those doses, and most patients can stay on drug for extended periods of time, so we're very excited about it. I think the whole community of oncologists is excited about it, and our investors have seemed to have thrown their weight behind it as well. We think that this phase III randomized controlled trial comparing RMC-6236 to current standard of care chemotherapy will help us determine definitively whether or not that's the case, and we hope to be able to provide it to patients with a formal full approval at some point in the future.
Yeah, absolutely. Very, very exciting data. Can you give us more details on the design of this phase III program?
Yeah, at its core, it's a fairly simple trial, but there are some elements to it that are a little bit sophisticated. At its core, patients who enroll, any patient with pancreatic cancer who meets standard eligibility criteria can enroll in that trial. They'll have to meet certain criteria, but it's not a high bar. Once they enroll, they'll be randomized either to standard of care chemotherapy or to RMC-6236, 300 milligrams a day. In the standard of care arm, it's up to the individual's investigator or doctor to make the decision about which form of standard of care chemotherapy they will provide. And that typically would be either a gemcitabine-based or a 5-FU-based regimen, since those are the two that are commonly in practice for pancreatic cancer. We call that physician's choice. And that will allow us essentially to compare against everything that's used for pancreatic cancer today.
Now, where this becomes a little bit more sophisticated than that simple design is that patients will be stratified on the basis of RAS genotyping of their tumor. So it's not required at entry into the study, but it will be analyzed when they're in the study, and if the patients have a mutation at position 12 in RAS, we call these G12X mutations, then patients will find themselves analyzed as part of a core statistical group called the G12X group, and in that grouping, we will assess PFS and OS as the dual primary endpoints. Beyond that, there will be patients who don't qualify for that G12X subset, and they then get analyzed as part of a full or all RAS mutant or all pancreatic cancer group after the core group has met its statistical significance.
And that allows us to stratify patients into those for whom we have the highest confidence that they could benefit from 6236. That's that G12X subgroup. And then more broadly, all pancreatic cancer patients, which could give us the chance to have an all-comers label if the data support that.
Great. And just from the data we've seen so far, how are you thinking about the efficacy and the different mutation types?
Yeah. So there are multiple mutant forms of RAS that can be found in pancreatic cancer. The most common form is KRAS G12D. The second most common form is KRAS G12V. The third most common form is G12R. But then there's a handful of others that show up with much less frequency. By the way, for those who are familiar with KRAS G12C, it represents about 1% or 2% of pancreatic cancer. So it's a very uncommon mutation in pancreatic cancer. In preclinical work, essentially every mutant, oncogenic mutant that we've tested is highly sensitive to RMC-6236. It binds to all these different mutant forms as well as the wild type form of RAS. What we can't answer today is whether or not patients with those different mutations will have some degree of variance in how their tumors inside a human patient will behave relative to the drug.
We just don't know that. We have seen activity across multiple mutant forms, but quantitating that in a rigorous way that allows us to answer Ellie's question: is D or V more sensitive than R or is R more sensitive than S, for example. We probably just don't have enough data. We certainly don't have enough data today to be able to answer that. We suspect that they're all sensitive on some level, and ultimately, what will happen in the clinical trial is in that all-comers group, they'll simply all be represented. They'll probably be represented in a way that's similar to their natural epidemiologic or demographic representation in human cancers, and we'll see whether in aggregate it passes that statistical test.
Great. And in terms of the enrollment, since you've, I think, announced you dosed your first patient back in October, how should we think about the enrollment timelines for this study as well as the timelines overall for when we could see data?
Yeah, well, we're not making a projection externally about the timeline. We did provide some sort of speculative view of that some number of months ago when we first laid out this potential trial design. But now that we're in it, we don't really have anything to say. But the mechanics of it are that once the trial is fully enrolled, then we have a data safety monitoring committee who will confidentially look at the data when the patient population in aggregate reaches a certain number of overall survival events, deaths, essentially. And that number is a number we will have provided to the data safety monitoring committee as representing a number that triggers the analysis and allows them to look to see, is the trial still delivering something and should it continue?
In that analysis, because of the way the trial is designed, the likelihood that we meet the PFS differentiation that we need to meet is higher at that first look than it is the likelihood of the OS benefit meeting that need because we've powered it around OS, which means we've overpowered it around PFS. So we don't know whether at the first look we will have crossed one bar or both bars or neither bar, but that would be found out at that time. If we've passed the PFS bar but haven't yet achieved an OS differentiation, the study will just continue until we have another opportunity to look at the data.
Great. That makes sense. Turning to first line in pancreatic, how should we think about your strategy there and the timeframe for when we could learn more?
It's very important for us to advance this compound into first line as soon as we practically can. As I mentioned earlier, when I gave the statistics from the second line data set that we reported in October, you might have noticed that those numbers for median PFS and median OS were greater than the numbers that are well established in the literature based on standard chemotherapy. We can't make that direct comparison in a cross-study kind of way, but it does raise the possibility that RMC-6236 offers a material benefit over standard care. Because of that, even in first line, because of that, we'd like to get that study going to give patients a chance potentially to benefit and to give us the opportunity to test that hypothesis.
The key here, though, is that we need to make sure that we understand what are the right treatment arms to have in that trial. Clearly, we'll have a chemotherapy benchmark arm, a reference arm. Clearly, we'll have RMC-6236 monotherapy as an experimental arm. The major question is, will we also have an arm in which we combine RMC-6236 with some form of chemotherapy? And that really is a question of whether we can do it. If we can do it, we are likely to do it. And what I mean by can do it is patients who are taking chemotherapy typically struggle through it. It's not a particularly easy regimen to take. Patients don't always feel well while they're taking chemotherapy, and they continue taking chemotherapy as long as they can tolerate it and until disease progresses.
So there isn't really a window to give another drug where they're not already taking a fairly aggressive medicine in the background. What we don't know is whether adding RMC-6236 then is tolerable for patients who are already marginally tolerating chemotherapy. We just simply don't know. It's something we have to answer empirically. And we are currently evaluating patients who are receiving both chemotherapy and RMC-6236 to see whether we can develop a combination regimen that justifies and that merits bringing it in as a third line as a third arm in our randomized phase I, phase III first line trial. So that's what's going on now that we'll be gating to that final trial design.
Once we make that decision, it will be a full go forward to do all the things we need to do to design a trial to bring it to the FDA to have those discussions to obtain alignment and to proceed with the trial.
Where are you with that combination work?
It's underway. I can't really give a minute, a blow by blow as to exactly where we are in it, but we are giving chemotherapy and RMC-6236 together. This is a dose escalation trial design, and we have two different things we can escalate. One is the chemotherapy, and the other is 6236. And of course, there are multiple forms of chemotherapy. So there are several things we could be evaluating, and we don't want to spend several years doing that and postpone launching a first line trial, but we also don't want to launch the first line trial in an uninformed way. So we're moving as swiftly as we practically and ethically can in that dose escalation.
Is there a scenario where you would move multiple different combinations forward in terms of just the dose levels?
Unlikely. Anything is possible, but unlikely. I think if we find a regimen that we feel is fair to patients, that it gives them the potential to have the benefit of chemotherapy and doesn't increase the probability that they'll discontinue the treatment regimen, then we'll go forward with that regimen. I don't think we feel obligated to have multiple chemotherapy regimens.
Absolutely. Makes sense. Maybe turning to lung, can you give us an overview of the data that we have seen so far and what we can expect from this update before the end of the year?
Yes. So a year ago, a little bit more than a year ago, we presented our first monotherapy clinical data on RMC-6236 in non-small cell lung cancer. Those early results from a relatively small patient set were quite encouraging. We saw reasonable response rates. We saw generally good tolerability. We did not have any durability evidence at that point in time. We had response rates, but no durability evidence. One, we did provide some swimmer plots that people could make some sort of judgments from, but we weren't able to provide any real estimate of durability. So that's the key to the disclosure that's coming up soon, sometime in the second half of this fourth quarter. And that is not just response rates, which in pancreatic cancer, which are probably not generally sufficient to drive this decision in lung cancer.
It's really about median PFS, and that's what we're looking for. And that's what I think investors should be expecting.
Maybe just any more clarity on number of patients, anything more that we can expect to see in this updated cut?
Not really. We'll have enough data that drives our decision-making. So we generally feel if it's enough for us, it should be enough for our investors. And it also has to drive alignment with the FDA. So they serve as another party in that decision-making. We expect to share the data with everybody this quarter. We expect to reach alignment with the FDA in the first quarter of 2025 as to the dose, as to the final trial design, and to get clearance to proceed with the phase III trial.
In terms of the way you think about what is compelling from a durability perspective, not asking the specifics of the data, but just in general your philosophy around how you look at the data and the relevant benchmarks, what are the considerations here?
The trial design is going to be focused on standard of care for previously treated lung cancer, and that's docetaxel. We know what the results are with standard of care, docetaxel in second line and later lung cancer. Typically, patients have a median PFS of in the four to four and a half month range. That needle has moved a little bit in the G12C subset of lung cancer patients. We can talk a little bit about the demographics in a moment, but the G12C subset represents something like 40% of the RAS-driven cancer. 60% is not included in the G12C subset. In that G12C subset, we know from adagrasib and sotorasib, a median PFS from their more mature data sets around five and a half, six months in that ballpark.
That has represented sufficient evidence to be encouraging, but I think for many people has not driven a lot of excitement that this is really moving things far enough for patients that there's real room to move and improvement to make upon that. It's really a judgment for each person to make what number one has to achieve to be positive. You and your colleagues have speculated about what those numbers might be quite extensively in the research analysis business. We won't provide a number, but we are looking for, we hope to achieve material improvement for patients. That's what we're all about, not just marginal improvement.
Absolutely. And what's the latest in terms of the phase III planning? I guess I think historically you've said maybe starting before the end of the year. Now we're going to learn more about it in 1Q. Just the status there?
Well, that is the status that we are analyzing data. We'll be prepared to share those data in this fourth quarter, which is what we had indicated all along we would provide. The logistics of it are we just don't see a chance to actually start that trial by the end of the quarter now. We were quite optimistic a year ago when we first set this timeline. We did achieve it. In fact, we overachieved with pancreatic cancer. We're a little bit behind with lung cancer. I don't think there's a particularly big story there. It just is what it is. We have a lot going on, a lot of things that have to be balanced against each other, and there are a lot of logistics that go into it. But we're quite excited about it. I did want to come back to the demographics question, though.
I pointed out that G12C represents 40% of RAS-driven lung cancer, which means 60% of RAS-driven lung cancer is not G12C, and that is served today by no targeted therapy, not even an accelerated approval RAS inhibitor, so the opportunity for those patients is docetaxel. That's a really core population for us with RMC-6236, and that will represent the core of what we're interested in in the trial design. Once again, we'll use a nested design of some sort where we'll put those patients at the core. We'll put G12C and other patients in an outer ring and analyze them in a sequential way to make sure that we have the greatest chance of having a successful trial based on the core before we evaluate the broader population.
Absolutely. Makes sense. And you have a number of different combinations that you're exploring in lung. Can you walk us through high level the many different combo studies you have underway and when we could get enough data from those that could inform potentially moving forward with a combo strategy?
Yeah, it's a good way to put it. I appreciate the way you set that up. I mean, we're doing these exploratory studies now, all of these exploratory studies to inform trial design for earlier stage advancement of these compounds into late stage trials, earlier lines of therapy. In lung cancer, it's quite clear that first line is dominated by a combination of chemotherapy and pembrolizumab. Nobody likes chemotherapy, but pembrolizumab is widely adopted. Essentially, everybody with lung cancer now will receive pembrolizumab with or without chemotherapy, and if they don't get it immediately, they'll get it sometime after pembrolizumab, so that's what we have to both compete with and combine with. We don't think that there's any obvious regimen that one could go forward with in first line that does not involve a PD-1 axis inhibitor, pembrolizumab or something analogous to it.
The challenge in the field has been that the G12C inhibitors, particularly the first few that were entered into the race, adagrasib, sotorasib, but even others that came later have all had some degree of difficulty combining with pembrolizumab because they carried into that combination some level of hepatotoxicity as monotherapy, which then got amplified when combined with pembrolizumab where the immune response is somewhat unleashed by suppressing the PD-1, PD-L1 axis. And that toxicity has become clinically significant and become a real barrier to moving compounds into earlier lines unless one significantly dose reduces in order to avoid that combination tox. So that's the hurdle. We have to overcome that hurdle. And we're encouraged by a couple of observations.
First, that for all of RevMed's compounds, the chemical backbone that's the foundation of the technology I described earlier is entirely independent of the chemistry that's been used in most of the G12C inhibitors. It really didn't derive from them. It didn't have any connection to them. It's just de novo chemistry. So there's no obvious prediction that those compounds should carry with them hepatotoxicity. Second, in preclinical work, we saw no significant hepatotoxicity for what that's worth. Most importantly, we've seen very little hepatotoxicity in the many hundreds of patients who have been treated with RMC-6236, whether in lung cancer or elsewhere.
And finally, in patients we reported last year who have lung cancer, had lung cancer, were treated with RMC-6236, sometime shortly after they had been treated with pembrolizumab, those patients also did not show any evidence of significant hepatotoxicity, which is very encouraging because it's now been established that those patients typically would have the same level of hepatotoxicity as patients who actually are concurrently receiving both pembrolizumab plus the experimental compound. And so we do have a preview based on roughly half of the lung cancer patients we reported last year. That is encouraging. It gives us optimism. Now we're running that study live. Concurrently, we're administering RMC-6236 to patients with pembrolizumab, and we will report on that in this fourth quarter. And that's really a safety question. If we clear that safety hurdle, then we have a real path to first line.
And that first line might be simply combining RMC-6236 plus pembrolizumab. We might combine it with chemotherapy plus pembrolizumab, or we might combine it with another RAS inhibitor. And so that leads me into another realm here. We have a deep bench, a deep pipeline of RAS(ON) inhibitors. We have three in the clinic, as I've mentioned. We have multiple others that we've described publicly, but have not yet advanced into the clinic. We have 20,000 other compounds behind those, and we have very extensive research effort trying to create next generation compounds that are superior to the ones that we have in our hands today. So there's a lot to come. Some of the compounds have overlapping target populations. For example, our G12C inhibitor and RMC-6236 could both be used potentially in patients who carry tumors that harbor a G12C mutation.
Our team, our discovery team at some point wondered whether combining our RAS multi-inhibitor with a RAS mutant selective inhibitor, so RMC-6236 plus RMC-6291, whether that would behave any differently in our extensive animal model collection compared to either of the agents alone, and in particular, they picked out models in which either agent alone delivered some level of anti-tumor benefit, but didn't knock it out of the park, so they picked relatively insensitive tumor models and then asked whether when you combine those agents, that would change the outcome, and indeed, across many different models now, we've been studying this for years, we've seen a substantial increase in the power of putting two RAS(ON) inhibitors together. It's not necessarily the case that any two RAS(ON) inhibitors would do it. I'm specifically describing combining a RAS multi-selective inhibitor with a RAS mutant selective G12C inhibitor.
But the results made it very obvious that there was some benefit to combining them. And what we think is happening is that it's causing deeper and more continuous suppression of RAS signaling, which is what these compounds are intended to do. Based on that, we have now advanced into the clinic that combination as well. So it is a first in human evaluation of combining two RAS inhibitors and specifically two RAS(ON) inhibitors. We're uniquely positioned to do that, and we're now running that analysis. Partly, we want to make sure that there's no toxicity concern that emerges from putting the two together. We don't really anticipate that, but that, of course, is a hurdle we have to clear.
And most importantly, we want to see, is there any context in which putting those two together gives us a very obvious signal that this is superior to the monotherapy. And we will provide an initial report on that. It's a first sort of glimpse under the hood of that particular combination engine this coming quarter. This quarter, the coming second half of the quarter. If that is really a viable or compelling combination regimen, and we've already then, let's say, have established that pembrolizumab can be combined with these agents, it would set us up for a first line study in which we might replace the chemotherapy platinum doublet with a RAS(ON) inhibitor doublet. And that would be a chemo-free front line treatment regimen. So it would be two RAS(ON) inhibitors plus pembrolizumab. And it's an ambitious idea.
It seemed particularly ambitious a couple of years ago when we first started talking about it, but now it actually seems very real to us. And we'll talk with you when we're able to share the data in the coming weeks, over the next six weeks, as to why we think this could be a compelling idea. So that does create some real options for us in front line, both unfortunately and fortunately. We'd rather have options, but then we'll have to choose what we're really going to do.
Have you dosed any patients with the pembro and then both of the KRAS inhibitors?
We haven't done that. Our feeling is the most important thing to nail down is that each of the agents alone with pembrolizumab is safe and well tolerated. Once we've done that, then if we've also proven that those two compounds can be combined, I think we would feel comfortable that we could advance into a triplet combination.
Absolutely. Makes sense. And a lot of combination work going on, and particularly in the G12C population. That's interesting in the context of lung. Any implications as we think about the potential second line study design?
I think it's unlikely that we would make the decision about exactly how to do those combinations in first line in time to do it in second line. I think we're anxious to start the second line trial. You raise the specter that, are we running behind? We don't feel we're running behind, but we want to get moving on that second line trial, and it does take some time to design an entirely new trial and to obtain FDA alignment around that and to prepare globally to run such a trial. And it also takes a lot longer to read out than a second line trial does, so I think they're kind of independent events, both worthwhile and both interesting. There is some potential read-through to the rest of the pipeline, though.
This combination 6291 plus 6236, to us, is a very good prototype of what a combination might look like. If we're unsuccessful, that's disappointing. But if we're successful, it raises the possibility that another mutant selective agent could be added to RMC-6236. 6236 could become a backbone or a platform on which other mutant selective inhibitors are provided based on the patient's diagnosis and initial tumor genotyping. And we do have a second RAS mutant selective inhibitor in clinical study and has achieved proof of concept. That's our KRAS G12D selective inhibitor, RMC-9805. We just reported the first ever data on that compound at the Triple Meeting in October.
We reported a good sizable number of patients from a dose escalation, and we showed that at our preferred dose, which is 1200 milligrams, it's very well tolerated, actually has an exceptional profile based on that initial data set, and delivered an anti-tumor punch in pancreatic cancer with a response rate of about 30%, at least at this stage, and so we feel quite encouraged by RMC-9805, so encouraged that we've already begun combining it with RMC-6236. It's behind RMC-6291, so it will take some time to develop those data, but we are dose escalating in that context as well, so we're trying to learn generalizable lessons from the experiments we're doing.
We are at the vanguard in this field with these really remarkable compounds, and we are taking the responsibility to investigate them thoroughly to establish the principles by which we will go forward to continue trying to defeat this terrible set of diseases.
Absolutely. How do we assess if the combination meaningfully adds to efficacy?
Yeah. We don't expect at this stage to have enough data to be able to compare response rates across one's 30% and one's 33%. We won't be able to do that. I mean, that would take hundreds and hundreds of patients and years of observation. But we will look for settings in which we don't expect either agent delivers the kind of punch that we need, and to then ask whether in those settings there's any standout activity from that combination. That's how we did things in the preclinical work, and that could be a model for the coming clinical data.
Great, well, Mark, thank you so much.
Thank you. Thanks for having me.
Thank you.