All right. Welcome back to the 45th Annual TD Cowen Healthcare Conference. I'm Marc Frahm from the biotech team. We're really happy to have with us for the next session here from Revolution Medicines, their CEO and Chairman, Mark Goldsmith. So I think most people in the room are pretty familiar with their program, so we probably will just go straight into some details. But maybe this will be a little bit of a higher-level question that you can ease into, Mark. But can you start off by just kind of laying out the RASolute 302 trials, so the second-line PDAC, and kind of how to think about enrollment here? Enrollment is open, and how that is going, and you've provided that guidance that you think most of it will be done this year.
Yeah. Thanks a lot, Marc. I appreciate the chance to be here. The RASolute 302 trial is a trial in second-line pancreatic cancer. It is comparing treatment with standard-of-care chemotherapy, physician's choice, versus RMC-6236, which is a RAS multi-inhibitor, also called RMC-6236. We announced the first patient in, I think, in October. All the sites that are active are enrolling well, very encouraging. There's a high demand for the trial. We are also expanding the number of sites in the U.S., which is according to a plan we've had all along. And we are now opening EU and Japan sites as well in order to get geographic representation there. Based on relatively early enrollment data, it looks to us as if we should be able to enroll this trial largely within 2025, if not completely.
There may be some patients in Japan, for example, who might come a little bit later because we have to open sites there a little bit later. We have to do a safety run-in trial in Japan by their rules. So there could be a few stragglers. But I think we're doing well.
Is that likely the rate-limiting step? It's kind of some of these geography buckets you need to tick off to completing?
Hard to say. I mean, I think we think it's likely we'll be able to complete enrollment in 2025.
OK. And should investors expect the enrolled population to kind of represent the epidemiology of G12X versus G13X versus Q61X and some of these other buckets of patients as you get to that hierarchical analysis?
We do. We do expect that patients that are currently enrolling in the trial will be similar in profile to patients who we've already studied. It's hard for us to be definitive about that because the eligibility criteria are relatively open, and so patients may or may not exactly match those profiles. But in terms of the distribution of mutations, we don't really see why it would be any different from the sampling that we have so far versus the natural prevalence. And we have tried to match our patient population from the phase II trial with other phase III trials that have been done with chemotherapy, for example, the NALIRIFOX trial. And when we compare the prognostic risk factors between patients who we've reported on and patients who are in that trial, which, of course, resulted in an approval, they're very similar.
In fact, the patients in our trial had slightly worse prognostic factors, not better prognostic factors. So we feel pretty optimistic that the patients in the phase III trial will be similar enough to the patients in our phase II trial that they should have similar results. But I can't see the future. We just have to find out.
But nothing about this kind of stratification factors that may be embedded in there that's going to naturally skew it a bit from the general population?
Not really. No.
OK. And then there are, obviously, two kind of important endpoints here, PFS, OS. How did RevMed kind of approach thinking about alpha split there? And maybe that starts to get into interim analyses. Just how to think about when to cut this trial and how to prioritize OS versus a more rapid answer on PFS?
Yeah. That's not so much a powering question. Really, the FDA has declared publicly that for pancreatic cancer, they think OS should be the primary endpoint. We have co-primary endpoints, but they want to see OS. And I think that's pretty reasonable for a disease that progresses very rapidly and for which, in second line, the PFS benefit is roughly three months and the OS benefit is roughly six months. It's pretty hard to make the argument that they should grant something on the basis of PFS when three months later you're going to have the OS results. That's sort of how they think about it. So we have powered the trial around OS, not around PFS. Also, it's events in OS, basically deaths, that will trigger the first read and any subsequent readouts. So it's OS events that trigger it.
But it doesn't necessarily mean that once a look is taken, that we would see that we've passed the OS point. That number that triggers that is a complicated statistical number that just triggers the look. So it's conceivable that because it's powered for OS, it's overpowered for PFS by definition, that we could cross PFS without having yet crossed OS. That's an intermediate sort of gray zone thing that could happen. Of course, if we cross neither of them, there would be some concern. Why did OS events trigger it and you don't have anything to show for that? That seems unlikely, but that would be close to a futile trial. Or it's possible we cross both, in which case we submit a package to the FDA. It's that gray zone if we've crossed PFS in the middle but haven't crossed OS, what happens there?
And that's partly a question of whether we would even know that or just the Data Monitoring Committee. And it's also then a question of if we were to submit that to the FDA, what would they do about it? But I don't know that that's the most likely scenario, but it is a possibility.
If it is in that gray area, would you anticipate telling investors? Or would you need to go kind of have that conversation with the FDA of, what do we do with this data right now before you would want to talk publicly?
Complicated. But at first, I would just emphasize it's not even clear that we would know that result. So the data monitoring committee, in other words, might say, you should continue the trial. So that might be what they tell us, you should continue the trial. That means it's not futile, but it's not over. And then we might just say, OK, we'll continue the trial. That might be it.
OK, and would you anticipate disclosing when interims have happened?
I don't know the answer to that. That would be sort of a materiality question for us to determine.
OK.
Don't know. What did we see in phase II? Yes. Thanks for the question. So in the phase I, II study, compared to the three and six months that I just mentioned for PFS and OS for chemotherapy, we saw a median PFS of in the mid 8s, eight months+ , 8- 8.5, 8.8, depending on which dose you're talking about. And for OS, for the aggregate of 120 mg- 300 mg cohorts, we saw a 14.5 months median PFS. And we don't have a, I'm sorry, OS. And we don't have an OS for the 300 mg group yet, which presumably is higher than 14.5 months, but we don't know that. It's a really encouraging set of data. And when we talk to investigators, I mean, you can be sure our team has pored over those data.
Is there something here that's, is this somehow a select subgroup among the second-line? We don't believe it is. The feedback we get, we actually get it sometimes from patients themselves, but through the investigators, is that this is what they're seeing, and it's what they were seeing back in October of 2023 when we first announced some data that patients were doing very well. By the way, we don't really record this or quantitate it. I know everybody wants to know PFS and OS because those are provable endpoints, but how do patients do qualitatively, and what the investigators tell us is that patients who start on chemotherapy for second-line typically come in the door with pain, often disabling pain, nausea, weight loss. They're not doing well.
What they've observed is, at least for some number of patients who start on daraxonrasib, the next time they come back into their doctor's office, they walk in instead of coming in in a wheelchair. They look more robust. They say they're not having pain or their pain is much reduced. So qualitatively, and I know I can't quantitate that for you, but this is what we hear from investigators. They can tell that a patient is getting something from daraxonrasib. That's why many investigators, who might have been skeptical in the beginning after they put a few patients on the trial, then call us up and ask if we can give them more slots. That's been our experience.
Given those benchmarks, the historical benchmarks of maybe three months of PFS.
Forget the quality of life. Let's just go back to the.
No, no. Yeah, yeah, back to the analysis. Well, it's the stats that, unfortunately, are going to drive the stock in the meantime until the patient, obviously, the patient experience will drive sales. But just with three months of PFS, six months OS, I mean, medians, you're likely already seeing PFS events in the trial happen, even maybe some OS events, unfortunately, happening in the trial today, given that you started enrollment in October.
I wouldn't know that. I mean, I understand you're saying a projection, but I don't actually factually know that. Yeah.
But when we get to completion of enrollment, whether it's very late this year, early next year, it seems like large portions of the trial are likely to have already progressed and maybe in a decent chunk, unfortunately, to have died. Should we expect interim analyses, that first interim, to be pretty close to completion of enrollment and timeline?
It's a very reasonable theoretical statement.
OK, and then you did mention to the prior question there from the audience just the baseline of what data you've shown in the second-line. Of course, that data, as you said, you don't know the median OS yet, or at least you haven't shown it on the higher dose. Should we expect continued updates from that phase I, II experience as you're doing this enrollment?
I don't know that we would necessarily do that. I mean, if we have something that materially is information, we'll share it. But otherwise, we're in the middle of enrolling a phase III trial. I'm not sure that putting out more data in the middle of that is always the best thing to do.
OK. And then you've also talked about starting to work on the first-line trial, or you have been working on it and getting it up and running by the end of the year. Just maybe to start that conversation, when you look across the experience in pancreatic specifically, but maybe a bit broader than just pancreatic, how much better should we think of regimens doing in the first-line than they do in the second-line?
Yeah. The overall experience with oncology drugs in general, and this includes chemotherapy, but also includes targeted therapies, is that the hazard ratios that are seen in later lines get preserved as you move to earlier lines. That's generally true. Somebody here might know an exception to that statement, but that generally is true. And if that applies here, the great news is that first line chemotherapy, in absolute terms, has done better than second line chemotherapy. We just talked about that, PFS from three months to six months, OS from six months to 11 months. And so if the hazard ratio remains the same, the absolute benefit for patients on RMC-6236 will actually be larger. So we don't have any reason to think that that sort of conventional or typical experience would change with a RAS-driven tumor.
I have heard some questions. Well, maybe later-line pancreatic cancer is more sensitive to a RAS inhibitor. I don't really understand it because the epidemiology, which I think is reflective of the natural history of the disease, is that nearly 100%, let's say 92%+, the pancreatic cancers at diagnosis have a RAS mutation. This is not a mutation that occurs somewhere between first-line and second-line treatment. It's not a resistance mutation. It's a driver mutation. So there's really no logic that we can understand as to why first-line would be less sensitive when historically first-line tumors are more sensitive to treatment, not less sensitive.
And so the data you've shown in later-line patients already is at least directionally a bit better than some of those first-line comps, as well as you'd be hopeful that.
It's numerically. It's not just directionally superior. It's numerically whether you want to, if you want to make that cross-trial comparison.
Yeah. And then you would hope it potentially could get even better, so why is the preferred option from Revolution to do a three-arm trial to also have chemo combo in first line, opposed to you've had the answer now for a year, you've been saying that you think this is a viable option as monotherapy in first line? Why not just start that trial and then maybe come later with chemo if it's better, chemo combo?
I think it would be absolutely credible, and if we were only looking at one dimension of this, we could start a first line trial of chemotherapy versus monotherapy daraxonrasib in first line, just like we've done in second line. Absolutely, I think it's qualified to do that. I think the mitigating factors to that are not so much biology as they are more cultural, that physicians who take care of pancreatic cancer are kind of a hardened core. They take care of patients who are sort of on death's door as they come in to get their first treatment, and they urgently get them into chemotherapy. That's how they think about what they need to do because every week counts.
If you think about median survival being six months or maybe 11 months at best for those patients, but median progression-free survival, six months, you really want to get them treated as quickly as possible, and they're used to doing chemo. To say to folks who really have not had experience with RMC-6236, and I'm not talking about guys at Dana-Farber or Sloan Kettering or MD Anderson who've had lots of experience with RMC-6236, I'm talking about community doctors who would be part of a larger trial, to say, you're going to take these folks who walk in the door, are very sick, have very advanced disease, and just put them on monotherapy RMC-6236 because we say it's better. I think for a lot of doctors, it would be, wait a minute, how about if I give them chemotherapy and I add RMC-6236 on top of that?
That's just how they think about it. And the number of oncologists outside of academic centers who even recognize RAS as the major driver of pancreatic cancer is shockingly low. I'm not going to say the number because it's sort of like hard to believe, but it's shockingly low because in medical school, that's really not taught because there's no action associated with knowing that. And when your head's filled all day with new information in oncology, you sort through what you have to remember. And RAS isn't one of them right now in pancreatic cancer. So it's not as if we have a receptive audience out there automatically outside of the academic centers. And I think I'd also comment there may be additivity from chemotherapy plus a RAS inhibitor. We have seen some additivity in a preclinical model setting.
It's not so profound that we're pounding the table about it. In fact, I'd be OK with running the trial with monotherapy versus chemotherapy, as you suggested. But there could be some additive benefit. There's no system by which we can start a phase III trial in first-line with two arms and then come back in six months and add a third arm. That's really not practical because patients will have been randomized at that point. So that would be starting a whole new trial with a whole new control arm at the cost of $X hundred million while another trial is getting ready to read out, wouldn't really make sense at that point.
If we're going to do it, it makes sense to incorporate them, randomize two to one, which means for patients, the chance that they get access to daraxonrasib is higher than if it's a one-to-one trial. That is an issue. The patients, they read The Wall Street Journal, they read Science, they read, they are looking at everything, and they know about daraxonrasib, and they want to get access to it. A two-to-one randomization is more favorable in that respect. If we can put together a regimen that deals with the high toxicity associated with chemotherapy for pancreatic cancer, it's a very aggressive disease, so you get aggressive chemotherapy.
If there is a regimen that is a clinically credible, widely used regimen that's not going to cause patients to get so sick that they come off of RMC-6236 so that they discontinue or take breaks in RMC-6236, then we'll proceed with that as the third arm. If we can't find such a regimen, we'll just proceed with the two-arm trial. But we're optimistic we should be able to find something.
In first-line, there is gemcitabine. There is FOLFIRINOX that are like the named ones, but there are a number of modified versions out there that you just alluded to. Does it need to be one of those, or are you willing to go down the path of kind of creating your own dose-reduced chemo?
No, no. We're not going to create our own regimen. We really only, if the point is to add to chemotherapy that doctors think works, going and creating a new chemotherapy regimen is just creating a new experimental regimen. So we want to, if we're going to have that third arm, it would be taken from the dose and schedules that are already widely used in pancreatic cancer. Maybe another point of clarification. When we say full dose versus modified dose, anybody who starts on full dose modified FOLFIRINOX is on modified dose for their second or third cycle. Very unlikely that somebody continues on full dose modified FOLFIRINOX for six months. It just doesn't really happen. It's such a demanding, it's maximal tolerated dose. I would even call it barely tolerated maximal dose, and most patients reduce. Most patients have a less than 100% dose intensity.
In fact, the median dose intensity in chemotherapy is about 70% for first line, which means they're not getting a third of the drug that you thought they were getting when you started the regimen. That's very conventional. Many patients get hospitalized and so on. So we want to work within the accepted range, clinically useful chemotherapy.
In terms of getting to that three-arm trial design, is it still the hope, or are you confident that you know there's at least one of these chemo regimens that is, from a safety perspective, viable, and it's figuring out which one is best at this point? Just where are you?
The language we used last week in our earnings call was somewhere between hope and a wish. We said we're optimistic, and I think that's exactly where we are. We're optimistic, but we are still following the cohorts that have been enrolled in this set of regimens, this well-accepted set of regimens, and we will have to make a judgment once we have seen them on for long enough. This isn't a three-week determination. We'd like to see them on drug long enough to be able to tell, are the doctors discontinuing RMC-6236 because of chemotherapy side effects? A very credible concern. If a patient, even on gemcitabine, develops a fever with neutropenia, they will be hospitalized. That's convention. Their chemotherapy will be stopped. IVs will be put in, antibiotics given. I call that a hair-on-fire medical emergency. We have to save your life.
We'll come back to chemotherapy after we've saved your life. Well, during that time, what's going to happen with daraxonrasib? And it's very possible that many doctors would say, I'm just not going to take any anti-cancer medicine right now. I'm going to save your life with antibiotics, and then we'll go back to putting you on treatment. With a RAS-driven disease, where RAS is the motor that drives the tumor, if you take your foot off the brake by taking away the RAS inhibitor, you're going to allow the tumor to regrow. So that's a bad idea from a daraxonrasib or patient perspective.
And so, if that trial is going to start in the second half of the year, I mean, it seems like you need to finalize the design in not that very many months from now, from internally, so that you can have that conversation with the FDA and turn it around into contracts and everything to actually activate sites in the back half of the year. Is that fair?
Yeah. Well, I said we'll initiate the trial. I don't know that we have to have all our, we can't activate sites until after we have FDA agreement. So none of that. We can do a lot of pre-work, but actually launching the trial and enrolling patients comes after we have alignment with the FDA and they say you're cleared. So we do have some time, but yeah, we've got to move fast on this. And that's why we've enrolled those cohorts, and we've been transparent about it. We're studying those cohorts. We'll follow them, and we'll make our best judgment about whether or not to include that cohort. But our intention is just to initiate a trial by sometime this year, the second half of this year.
OK. And you've also committed to an adjuvant trial.
We have.
Outcomes are better there, but still not good at all.
So can I define that for everybody? What adjuvant?
Yes, in a second. But also, is the priority to get this chemo regimen or some sort of chemo combo regimen into the adjuvant setting as well? Or at that point, does the toxicity of chemo start being too much and you want to get monotherapy there by itself?
All right, so an adjuvant, what we mean is resectable disease. That's about 15% of patients who get diagnosed with pancreatic cancer. There's a lesion that's detected in the pancreas that a surgeon says, I can remove that, and there are no other lesions, so it's not metastatic. Often, but not by any rules or algorithm, often there is chemotherapy associated with that surgery perioperatively, either before surgery or after surgery. The range of regimens varies from one doctor to the next. Some give a little of this and a little of that. There really isn't an established adjuvant chemotherapy regimen associated with resectable treatment. Maybe for a few months, some might use FOLFIRINOX, some might use gemcitabine, some might use just gemcitabine or whatever. So that's a little bit tricky, and I think we have some ideas in mind for how to manage that.
Not ready today to describe a trial design, but we have some concepts around that. Once that treatment is done, I think most patients leave their doctor's office thinking they're now cured of their disease because it's been surgically removed. The likelihood of a cure is very low. Most patients have micrometastatic disease somewhere that just wasn't picked up on a scan, and at some point, somewhere between 10- 20 months median, will return for a follow-up scan, and a lesion will be found, and then they will be, by definition, at that point, second-line patients, which may not really be biologically different from an adjuvant patient, but they are defined as different because a lesion appeared after they were resected, so we believe that based on the data we've already discussed for daraxonrasib in the second-line setting, we've already validated that approach.
It's already a justified treatment to use, and therefore it ought to be used in the adjuvant setting, so there is nothing left for us to do from a treatment development perspective to proceed with that adjuvant trial. The only thing we have to deal with is the fact that doctors use different kinds of chemo in association with it, and again, I think as we have more clarity around that, we'll define that explicitly for everybody, and that will be the trial that we'll run.
But is that about defining the control arm, or is that about defining what you put?
Just the overall trial design, and I think we can't go further than that without sort of discussing trial designs right now, but we will share that, and it's our intention to launch that trial, to initiate that trial in 2025 as well.
And maybe this is starting to transition to the rest of the pipeline. In the few minutes we have left, RMC-6236, sorry, the RMC-9805, the G12D, how does that fit into pancreatic cancer?
Yeah. The single most common RAS mutation that drives human cancer is RAS G12D. It's most common in gastrointestinal tumors. It's present in lung cancer too. It's about 4% of lung cancer, but in colorectal and pancreatic cancer, it's really the single most common mutation, so it will at some point be a defined disease. It won't be that you have a RAS tumor. It'll be that you have a G12D tumor. Just like in lung cancer, we don't say you have a RAS cancer, you have a G12C cancer or not, so because of that, we think a mutation selective inhibitor may become part of the regimen for treating patients with a G12D-driven tumor, and right now, the first G12D inhibitor that's really compelling is daraxonrasib. It's active against G12D, but it's not selective for G12D.
We're evaluating zoldonrasib, for which we've shown anti-tumor activity in pancreatic cancers driven by G12D. Whether it will be superior to daraxonrasib or whether it should be added to daraxonrasib is something we're trying to determine now. Our thesis is that it would be additive to daraxonrasib and should go on top of daraxonrasib rather than competing with daraxonrasib. The data have not been produced. We haven't shared any data to validate that assertion.
I guess we first talk to what level of data set do you need to accumulate to kind of start proving that out, at least from a proof of concept perspective?
Yeah. I mean, we've tested this idea with our G12C selective inhibitor called elironrasib or RMC-6291. And in order to, we knew this would be the first clinical test ever of putting two RAS inhibitors together. And we thought if we do it in tumors in an indication where each of the agents is highly active alone, it will be very hard to tell whether there's additivity between them. We'd have to run a big trial to prove that. So instead, we went to a very kind of dire situation, colorectal cancer, late stage, G12C, and that had already previously been treated with and progressed on a G12C inhibitor. So we really went to kind of third line, late stage colorectal cancer. And in that setting, combining daraxonrasib with elironrasib yielded a 25% response rate, which is a very high response rate.
No single agent would have delivered that in that context. That validated, from our point of view, the concept of putting two RAS inhibitors together. Now we're testing G12D inhibitor in the same way with daraxonrasib in various tumor types.
When do you think we'll be able to see that, whether that translated or not? I get the scientific rationale for why it would translate, but when will we learn whether it did happen or didn't?
I think investors will learn after we learn, and we don't know the answer to that now. If we knew that, we would tell you that. We're just now testing that. We've done a dose escalation. Now we have to follow patients and see.
Should we expect that initial learning will again be in CRC because of the lack of single agent?
It's actually in multiple tumor types, but we'll just have to see how that plays out. We have a lot of different cohorts going on, and we'll see what we learn from these as we go.
OK. And then maybe just broadly across daraxonrasib, zoldonrasib, and elironrasib, I think many investors are kind of dismissive of the opportunity in lung cancer because of the sales performance of the G12Cs. What do you think they're missing?
I think in second-line, durability has been relatively short for most of the G12C inhibitors. That's not true for all of them. Divarasib, for example, and they're going for an approval in second-line. Clearly, first-line is a more attractive place to go because it has more patients than those who have progressed to second-line, the possibility of longer treatment duration. I think if you're using the best compounds that could deliver that longer duration, I think there's a compelling unmet need, and there's a compelling commercial case to go with it.
You want to walk through some of that strategy to get to the first line in the particularly choice of kind of backbone, right? Some patients can get PD-1 monotherapy, others it's PD-1 combo with chemo. How much do you need to go on top of that chemo combo, not just PD-1, to kind of really access that opportunity?
Yeah. Well, it depends on the results to some degree, but our guess is that in first-line lung cancer, you either need to use chemo KEYNOTE-189 regimen type thing, or you need to have two RAS inhibitors to replace chemo, a RAS doublet. And we're really excited about that possibility. A chemo-free first-line regimen would mean patients get the doublet of RAS inhibitor plus PD-1 as their first-line treatment, and you move platinum doublet chemotherapy to become a second-line treatment, which leaves docetaxel as a third-line treatment. That could stack up to a significant survival benefit, which is what we're all looking for.
OK. Unfortunately, we're at times where we have to cut it off there. But thanks a lot for joining, Mark, as well as everybody in the room.
Thanks a lot.