Welcome back to the 44th annual TD Cowen Healthcare Conference. Marc Frahm from the biotech team here. Really happy to have the next session here with Revolution Medicines and their CEO, Mark Goldsmith. Maybe just start it off with, Mark, you wanna just, for people who maybe aren't as familiar in the weeds on Revolution, just give, give a very high-level overview and kind of status update and lay out the major catalysts for the next year or so, and then we'll-
All in one sentence.
One, one sentence. You get 30 seconds. Go.
Thank you, Mark, and thanks to TD Cowen for inviting us to be at the conference. Revolution Medicines is a clinical-stage oncology-focused company. Our particular interest is in RAS-driven cancers, representing about 30% of all human cancers. We have developed a very exciting and very deep pipeline of direct RAS inhibitors. It's a class of inhibitors that's quite differentiated from anything else that's ever been in humans. This is a set of inhibitors that bind to the active or on-state of different RAS mutants. We have three such inhibitors that are in the clinic now, and we have reported initial data on the first two of those. The first of those is a very bold compound called RMC-6236. We refer to it as a RAS multi(ON) inhibitor , on because it's inhibiting the on form, and multi because it inhibits every form of RAS we've ever tested.
So some people refer to it as a pan-RAS inhibitor. We call it a multi-RAS(ON) inhibitor. And although, going into the clinic, the paradigm had been the dogma had been that inhibiting all forms of RAS would be anywhere from causing severe morbidity to lethality. We've disproven that dogma and have now established clearly a very attractive safety profile for RMC-6236 so far, and actually superior, I think, at least in reference to the early G12C RAS(OFF) inhibitors and certainly in reference to chemotherapy. So well-tolerated, and appears to be safe. And we've shown significant anti-tumor activity across RAS-driven cancers that we may wanna get into some more detail about. Our second compound, RMC-6291, is a G12C selective RAS(ON) inhibitor. So it binds to the G12C variant that causes much of lung cancer.
It is operating via the same mechanism that inhibits the on state, that I described earlier. This compound we've also reported initial data for is highly active in G12C cancers and shows interestingly a strong activity even in tumors in patients who have been previously treated with a RAS(OFF) inhibitor like sotorasib or adagrasib and had progressed on those compounds; they have shown good response rates for our compound. Our third and, I'll just say final clinical stage asset from the RAS(ON) inhibitor collection is RMC-9805, a G12D selective inhibitor. RAS G12D is the most common variant of RAS that causes human cancer, so it's of high importance. This is also an on-state inhibitor. It is a covalent inhibitor that binds to the aspartic acid of the RAS G12D, which is a pretty groundbreaking bit of chemistry.
We've not reported much information about 9805 yet. Other than that, it is orally bioavailable, and we are seeing dose-dependent increases in exposure as we dose escalate. So far, it's been safe and well-tolerated.
Okay. Great. Thanks for that. Maybe back to one of the comments you made there of that dogma going into the clinic.
Yeah.
That, you know, inhibiting RAS broadly was anywhere from maybe a modest idea to a the worst idea ever, because of the toxicity that you clearly that's not been panning out. But what level of RAS inhibition do you think you're actually sustaining at those higher doses, like 300 milligrams that you're most interested in, kind of longer term, at least from a monotherapy perspective?
Yeah. We don't really have a measurement of pathway inhibition in these tumors because we haven't biopsied them, and it's a pretty difficult thing to quantitate anyway, even if we did. So the best way I can answer that is by referring back to the preclinical data and to say, what, at what exposure level in mice that's comparable to the exposure level that we achieve in humans at 300 milligrams, what kind of toxicity do we see? And the simple answer is those mice actually do quite well. That's a tolerated exposure level even for months on end. We've treated mice for, you know, 100 days, and we've suppressed their tumors for 100 days. So, we haven't seen much in mice at that dose level that we would have expected to cause problems in the clinic.
And in fact, in the clinic, we haven't seen much that we would expect to limit dosing. Now, there are some side effects. The most prominent side effect in humans is a rash. It's not dissimilar from the EGF receptor antagonist-mediated rashes that many people are familiar with. It tends to be grade one or two. It doesn't tend to require dose reductions or changes in or discontinuations in dose, although occasionally that does happen. And it tends to respond to management practices that are well-established for EGF receptor treatment. So it has not been a significant manifestation from the point of view of clinically treating a patient's tumor, but it is a common side effect in patients that who are treated at the higher doses. We've seen some GI toxicity, some diarrhea, some nausea, and, and some other side effects.
But these have largely been grade 1, grade 2 , and really, relatively de minimis compared to many other oncology drugs, including the RAS G12C(OFF) inhibitors , sotorasib and adagrasib.
Okay. And maybe, again, going into that dose escalation, it may be not so much from the direct inhibition, but what, you know, what are you seeing from a kind of dose response perspective on response rate, on, you know, durability? Like, how would you kind of describe that pattern of those metrics of efficacy?
Yeah. You know, so going back over the history now, it's been about a year and a half, two years since we started studying RMC-6236. We first dosed patients at 10 milligrams per day. We were sufficiently intimidated by the dogma that RMC-6236 must be very toxic that we started at a very low dose just to make sure that we would avoid toxicity. And it turns out even that 10 milligram dose had an anti-tumor effect. We could see effects that were measured as stable disease in a number of patients even at 10 milligrams. And that was true at 20 milligrams and 40 milligrams. And at 80 milligrams, we started entering an exposure level that we expected to induce tumor regressions, actual shrinkage of tumors based on what we had seen in the preclinical species.
And in fact, that's what we began to see at 80 milligrams. We started seeing tumor regressions, partial responses. And as we continued dosing up, and we have now dosed as high as 400 milligrams, we saw some increase in side effects, for sure, but again, not to the level that one would consider intolerable for most patients with life-threatening disease, and not requiring frequent dose interruptions or discontinuations. And I can come back to that point in a moment. And what we reported back in October with RMC-6236 was that we had an aggregate response rate in lung cancer of about 38%. So that is taking all doses from 80 milligrams up and adding those together, we saw about a 38% response rate and in pancreatic cancer, about a 20% response rate.
Now, those numbers, just to sort of, level set for everybody, those numbers are significantly higher than the comparators that is established standard care for second-line treatment of lung or pancreatic cancer. So substantial improvements over those, not necessarily as high as everybody wants to see, but still clearly, differentiated. Since then, we've reported without showing waterfall plots in the detailed, data that, we will eventually show, we did report that as we've dose escalated and as we've observed patients for longer periods of time, we've seen those response rates actually go up since the ESMO report, countering any concern, I think, that they might deteriorate over time, which sometimes happens with early studies. And so in lung cancer, in the aggregate study, we've now seen patients, roughly in the mid-low to mid-40s% response range and in the pancreatic cancer, realm in the mid-20s%.
So we've seen them go up. And then at the highest dose that we've studied most intensively, 300 milligrams, the numbers were even a little higher for both of those groups. So we're seeing good, we call those favorable trends. I don't think the exact number matters too much ultimately, but we're in a range that clearly is differentiated from standard of care if it holds up in a head-to-head kind of analysis. And the most important question on the table today, since these are perfectly solid starting points, the most important question is progression-free survival. Those that will be a critical component of the endpoint for pivotal trials in second-line lung cancer and, frankly, also in first-line lung cancer and pancreatic cancer.
We've not yet reported progression-free survival data for the simple reason that it takes time to allow those data sets to mature for patients to be on the drug long enough so that you can actually reasonably accurately estimate the median PFS. So we've not reported that yet, but we will report that out in the second half of this year, both for lung cancer and for pancreatic cancer. We expect to do so in conjunction with initiating a pivotal trial in second-line pancreatic cancer and a pivotal trial in second-line lung cancer. We may not report those at exactly the same time, but in totality, we should report that information later this year.
So at those disclosures, you're expecting to have, like, mature median PFS data?
Yes.
Okay.
Yeah. I mean, the only reason as I said, the only reason we haven't reported yet is because the data have been too immature. And with PFS estimates, if you're projecting out the right side of the curve, which is where the PFS number is going to be, and you don't have data points that take you out that far, those projections have huge error bars around them.
If you wanna move it to the right here, that might probably help.
So, it's very important not to be commenting on talking about numbers that are really premature, with large error bars. So we're just waiting for those to mature. We're developing those data. And importantly, we're also preparing to share them with the FDA as part of our overall, end-of-phase two package.
And those updates, should we expect those at medical meetings that have their own kind of cadence to them, or is this really more likely to be a standalone kind of Revolution Medicines update when you have that agreement with the FDA on this is the path forward with this dose, with this cohort?
Yeah. You know, in a perfect world, we'd present these first at a medical meeting. But we're not in a perfect world. And medical meetings are only available at certain times of the year, and they require certain lead times to submit abstracts and so on. So, I don't know. I think it's probably more likely that it will not be first at a medical meeting and we'll follow up at a medical meeting. But we'll just have to see.
But the intention is data update with this is the final design of the next step?
Yeah. We think it's important. We've given enough information now, I think, for most investors to have a pretty good idea how these compounds perform. And we don't expect that to change much. We just haven't provided any real insight about durability because we haven't had the mature data to do so. There has been a clue, which is the disease control rate for RMC-6236 in both pancreatic ca second-line pancreatic cancer and second-line lung cancer has been quite high. You know, we've been in the 85%+ range for disease control. And that's probably the best thing to look at as a projector of what's going to happen with the progression-free survival because if you're seeing those curves separating early on, that tends to carry over longer term. So we're quite excited about those disease control rates.
We think they're very important, even though they're not an endpoint per se. But for those who wanna read the tea leaves, I think that's the first thing to start with. And then we'll just simply provide those mature data, when we are able to do so.
And then you know, I think the next step that you have kind of broadly described, maybe we'll start with pancreatic cancer, you know, is a second-line trial kind of versus chemo, not exactly, you know, probably a handful of different chemo regimens, all on a basket. But what is the strategy going forward to the first-line? Do you view this as a replacement for chemotherapies in the first-line, or is this something to add on top of those first-line chemo regimens?
So let me first comment on second line and then, then come to your, your real question. But just since you mentioned it in, in your question, I think gives me the ability to talk about it a little bit further. Our population for both the lung and pancreatic cancer studies so far, the Phase I, Phase II studies, has been a mix of pure second line, real second line, and second line plus, meaning third line or fourth line. And patients have had anywhere from 1 to 7 prior treatments. So really, you're talking second to sixth line or whatever, seventh line. And it is clearly the case that with pancreatic cancer in particular, as one progresses through disease and through lines of therapy, response rates and durability decrease over time.
So we really have a mixed population that's reasonably weighted towards quite advanced stage patients because that's what you get in a clinical trial of that sort. The second-line study that we conduct for both pancreatic and lung, each of those studies will be a much more pure second-line population. So one can infer from that that the benchmark we're reporting for second-line lung and pancreatic cancer are actually the pure baseline benchmarks for those populations. But we're comparing it, our compound, against a much more mixed and more advanced population. So I think that's something to sort of calculate in a little bit. We're already beating those benchmarks as we've described, but I think that will improve, again as we pure you know have a more pure population in the late-stage study. Okay. first-line .
We're very interested, of course, in advancing to first-line. We're actively working on charting the path to first-line. We don't think that it's a multi-year path to get into first-line. There's already interest in first-line with RMC-6236. It's caused it generated enough interest among patients and investigators. And I think the data clearly are pointing in that direction. So we're planning for it. And remind me exactly the question about I what it was.
So, take on chemo, essentially, do you view this as chemo sparing?
Chemo. Okay.
It either replacing some or all of those first-line chemo regimens or versus, I going on top of them?
Yeah. It's an important question. It's one I don't have a very good answer for yet. The most natural and common way to enter first-line is by adding your agent on top of first-line chemotherapy. It's the most protective for patients that they secure at least the benefit of the chemo and then whatever you might add on top of that. On the other hand, chemotherapy is a pretty devastating thing to have to experience. And if you're talking about particularly pancreatic cancer patient whose projected lifespan is not very long to spend a lot of that experiencing chemotherapy, it seems like not a very high-quality-of-life approach. So we'd love to be able to displace and to provide a chemo-free regimen. It really comes down to our estimate of the probability of success.
We wouldn't want to launch a chemo-free regimen that we thought was high risk, that wouldn't make sense in a first-line study, and it wouldn't make sense for patients if there's high risk that they're getting an inferior, that they might be getting an inferior treatment. So we really have to get some feel for what are the reasonable projections there and whether we end up adding to chemo, displacing chemo, or potentially even studying both, in the same trial. It would not be an inefficient use of capital, to compare to the two because they might actually serve different patient populations. Not everybody with first-line diagnosis of pancreatic cancer in particular can take chemotherapy. You know, it's a pretty rough go. And if your performance status is low, you may not be eligible for chemotherapy.
So you'd like to have the option for doctors to decide with their patients what's the right regimen. So in a perfect world, we might actually study both. But we don't know today. That's to be determined. And we're spending some of our time this year developing some data that will help us establish what's the best path. For example, we'd like to combine 6236 with chemotherapy and see if it's tolerated. If it's not tolerated, well, then that rules out that combination. But if it is tolerated, then that could become an arm study. So, we'd also love to see whether a first-line patient who receives simply 6236, how do they do? That's a little bit harder to do because that's already asking somebody to forgo standard of care.
that if that were to happen, if such an experiment were to happen, it almost certainly would require that those patients be ineligible for chemotherapy as the ethical basis for giving them RMC-6236. That's all to be worked out, but we're highly interested in actively working on those possibilities.
I mean, I guess, yeah to that point to that last point of, you know, asking somebody to take particularly monotherapy in the front line is a big ask, both from a, you know, ethical to them but also for regulators to allow it and all that. And to get a type of data set that's mature that tells you what you know, gives you a reasonable prediction of the first-line randomized trial, it's gonna be really difficult. So how predictive is kind of the longer-term outcomes in the second, third, fourth-line patients that you're seeing treating now to what is really gonna happen in the first-line?
Generally speaking, what one sees in second-line improves in first-line patients. Just generally speaking, that's what one sees. It's hard to answer the question of how will we project exactly what that, for example, PFS is in first-line treatment. I don't know what equation we could do, you know, use to, to do that. It may be the case that we'll know it when we see it. You know, I think that all comes down to what do the data show with regard to the second-line median PFS. And hence, you're asking me regularly, "When are we gonna see the median PFS data?" That's because you need to know the answer to that. And investors need to know to understand what are the legs on this compound? Where does it go? Where does it take us?
It's possible that those data will be sufficient for us to already have confidence in first-line. It's possible that they won't. I can't predict the data that we haven't yet finalized, but we'll know that before we have to make that decision.
The intention would be to get a first-line trial up and running, well, before you necessarily even have second-line randomized second-line data?
Yeah. We don't think that it's gonna be necessary to wait for the results of the second-line pivotal trial before starting a first-line study. We could be wrong, but we don't think that that's gonna be required. And given that investigators are already asking us about design of first-line trial and giving us their opinions about it and they're anxious to do some of these preliminary studies that I talked about to evaluate 6236 in combination with chemo in patients, suggests that there's growing enthusiasm for doing that. And when we reach the point where there is enough consensus on it, I think we'll be able to do that. So we wouldn't have any problem running those two trials in parallel, launching the second-line pivotal trial this year and perhaps in 2025, launching a first-line study. That is conceivable.
I can't commit to that today, but that's a perfectly reasonable thing to do. Obviously, in the world of IRA, it becomes important to make your decisions about your development plan much earlier. I, you know, in the old, old days, being a year ago or two years ago, everything was done sequentially. I think we'll see much more compression around that. And we certainly are experiencing that, also.
That's probably a good segue to compressing not just sequentially within a tumor type but also in parallel, multiple tumor types. So maybe transition the conversation to lung cancer for 6236. Just the strategy there, you know, should we also expect mature kind of PFS data at the next update?
I think so. I mean, that's our intention is to really treat these very similarly. We owe our investors, our investigators, mature PFS data, on 6236 in each indication for which we're pursuing a pivotal trial. We're developing those data. Patients have been fully, you know, enrolled, for all the dose cohorts that we need. And we're simply collecting and sort of allowing those patients and their data to mature, and conducting the analyses that need to be done, to do that. They may not arrive at the same time. Just, there are just different logistics for pancreatic versus lung cancer. So we may not be able to present them at exactly the same time. And we may not, therefore, start the pivotal trials at exactly the same time.
As a relatively young company for whom these will be our first two pivotal trials, it's probably not a terrible thing that they might not be started at exactly the same time. We are anxious to start them both. Our current projection is that we ought to be able to start them both, by the end of this year, subject to selecting dose in conjunction with the FDA, subject to FDA agreement about eligibility, trial design, statistical design.
Maybe similar to the prior conversation in pancreatic, just lay out that path to not just, I mean, what you've talked about before and described is, you know, second-line versus docetaxel. But also, you know, what's the path further forward? And, you know, what's the desire to try to remove chemo in this setting?
Yeah. Well, it's pretty desirable for same reasons. I mean, it's the same chemotherapy. And it's just as ugly whether you're a first-line patient or second-line patient. So, you know, in an ideal world, our vision would be ultimately to provide chemo-free regimens with targeted therapy. And because RMC-6236 so far has been well-tolerated and safe, it's actually a very attractive differentiator from chemotherapy. With that said, we've gotta make sure that we're providing the same efficacy benefit or superior efficacy benefit to justify doing that, even though, honestly, I think any patient given the choice between a targeted agent that's well-tolerated and chemotherapy with its profile and even achieving the same clinical outcome, I think most human beings would choose a targeted agent. But that's not our design. That's not our set of assumptions around how to seek approval.
We'll just have to see, can we get to the point where we have confidence that monotherapy could or should deliver that kind of effect?
Sorry, monotherapy for first-line, monotherapy there? Or I would or with in combo just with PD-1s but no chemotherapy?
Sorry. I got lost in pancreatic cancer. So it is true in lung cancer that there is no path forward as far as we know into lung cancer first-line without dealing with PD-1. I, you know, it's very widely used. It's established standard of care. So it is critical to combine with Keytruda or similar PD-1 antibody. And I know you gave me a softball, and I just whiffed on that. But now I'm with you. And so we are actually studying 6236 in combination with PD-1 now. Now, as many of you may know, the RAS off-inhibitors, the G12C off-inhibitors that have been studied in combination with Keytruda have so far shown significant tolerability and safety issues, principally hepatotoxicity as the combined side effect. And that has limited and largely prevented those G12C off-inhibitors from moving into first-line, not completely but largely.
And in fact, I think there was just a study reported by yet another company, this time Merck, about their G12C OFF inhibitor in combination with their Keytruda. And it showed toxicity, hepatotoxicity. So that's been a barrier to first-line that I think has been broadly across the G12C OFF class. And we don't know if that extends to all RAS inhibitors, to G12C only, or only to those specific inhibitors. So we are now studying that with RMC-6236. And I have no data to report today. I can point to, though, that in our monotherapy study with RMC-6236 in lung cancer, roughly a third of the patients had had Keytruda prior to entering the study and had had it within 12 weeks or less of starting RMC-6236.
It turns out - this has now been published by at least one and I think maybe two separate groups - that hepatotoxicity is seen also in patients who have what appears what's, ostensibly sequential treatment by anti-PD-1 and then their compound but really isn't sequential because of the long half-life of Keytruda. And so within 12 weeks, many patients still have a significant level of Keytruda. And so it is actually an early glimpse of what you might see later. And about 30% of those patients, something like that, show hepatotoxicity within that, you know, that 12-week sequential treatment. So that's a marker. And I can't tell you how much confidence one should put in it, but it is a marker.
And we've shown that in our patients, in whom about a third have had prior PD-1 within 12 weeks, we did not see any grade 3 hepatotoxicity. So, you know, everybody will have to decide for themselves how much weight they wanna give that in, trying to sort of see down the road. For us, it's a pretty encouraging sign. But we also feel we need a definitive prospective data set. So we have already begun enrolling. We have already been treating patients with RMC-6236 and Keytruda. And we'll get the answer to that. We expect to disclose that, you know, sometime this year because that will be very important to decisions about first-line lung cancer.
You know, frankly, if it does play out the way that early glimpse, that preview suggested might, that could move 6236 really to the head of the class just on that characteristic alone.
Particularly in lung where G12C is such a big portion of the RAS mutations, how important is your now-open combo with 6291, your G12C inhibitor, in terms of figuring out how exactly to move 6236 forward?
It's a very complicated question that you've just asked. So I'll try to unpack it. G12C represents a little bit less than half of the RAS-mutated lung cancers, not small-cell lung cancers. The other half are mostly G12 non-C, so what we call the G12X population but without the Cs. So that's about equal in size between the two. And then there's one more bit, about 5% of patients who have non-G12 RAS mutations totally. It's not C, and it's not anything else. It's Q61, or it's G13. So the total is about 30% of lung cancer. Our compound today, we've shown activity in that 12%-13% who are G12X but not C. We've not shown any data on C. And we've not shown any data on the non-G12s totally, that final 5%.
So that leaves a very interesting question about what to do in our phase III monotherapy trial with RMC-6236. Do we include the G12C population, or do we not? If we do include them, are they in a core group, or are they in a nested group where they get their own statistical treatment? Likewise, for the non-G12X population, do we include them or not? And if we do, do we include them as a nest? And we're trying to develop some data in the clinic now that might help us address that question. We're testing RMC-6236 in populations outside of that core group that we'd already tested. And preclinically, just to be clear, RMC-6236 is quite active against all forms of RAS that we've ever tested.
But it did show some differential activity in that G12X population, which is why we created the term G12X because it came out of our preclinical data that those tumors were more sensitive. And as a result, we prioritized them in our study. Now, we have to go back and clean that up and find out what happens with 6236 and those other populations. We're trying to study those now to see if we can get at least a qualitative answer that will guide us in the second-line study and, of course, would also then impact the first-line study.
I believe we're supposed to get some data kind of along those lines, in the next quarter or so. Can you frame enough of that presentation in terms of the, you know, how, how much towards the one end of the spectrum being just kind of anecdotal reports in different mutations versus kind of more robust proof of concept in different settings?
It's not gonna achieve the level of proof of concept, although I think what one needs for proof of concept here is quite different from what one needs with a kind of first-study population. But it's really not gonna be a very large population, w e're sampling a lot of different things, so it's impossible to get enough N from that. I think we're looking for a qualitative answer. Does this look like what we saw preclinically, in which case that will send us down one path? Or does it look vastly different from the preclinical data? So I'd say it's more gonna be towards the anecdotal end of the spectrum rather than the super deep quantitative end of the spectrum.
Okay. Unfortunately, that's all the.