Good afternoon, everyone. Welcome to the next session with Revolution Medicines. I'm Ami Fadia, Biotech Analyst here at Needham & Company. It's my pleasure to be hosting Mark Goldsmith, Chairman, President, and CEO of the company. Mark, I'll give it to you to sort of give us a quick overview of the company, and then let's dive straight into Q&A. Sure. Thanks, Ami. Really appreciate the chance to be with you today. Revolution Medicines is a company focused entirely on discovering and developing new targeted therapeutics for RAS-driven cancers. We primarily build upon a platform, a technology platform, that enables us to drug the active or ON-form of various RAS targets. This has really broken new ground for the industry, and now I think it's become accepted that this may be the preferred way to target RAS cancer drivers. We have brought into the clinic three of our first set of RAS(ON) inhibitors. One is a RAS multi-inhibitor called RMC-6236, which I'm sure you're going to want to talk about. The second is our G12C selective inhibitor called RMC-6291, and the third is our RAS(ON) KRAS G12D selective inhibitor called RMC-9805. We've reported, I think, a pretty good set of data on RMC-6236 and on RMC-6291, and on RMC-9805. We've yet to disclose a formal data set, although we've given some guidance about how things are progressing. Okay. All right. So let's dive into RMC-6236. But before I do, ask you my first question, just a quick reminder to our listeners: if you want me to ask any question, feel free to send it over to me through the dashboard. So, RMC-6236, you've reported some data on in lung cancer and pancreatic cancer patients, and earlier this year you provided us with an update with regards to the response rates that you're seeing. Could you talk about sort of the most recent data that you shared and how that compares with some of the benchmarks in the industry? Sure. Yes. Back in October at the Triple Meeting and the ESMO meeting, we gave our first really substantial reports on RMC-6236. We presented safety and tolerability data that I think indicated a profile that's quite attractive for patients, very well tolerated, and no substantial safety signals even across a wide range of doses. And then we reported, along with that, activity data, principally in the form of waterfall plots, showing disease control rates and objective response rates from the patients we had studied so far in lung cancer, non-small cell lung cancer, and pancreatic cancer. In each case, patients carrying the KRAS a mutation at the KRAS G12 position. We refer to these as G12X tumors, specifically, though, excluding the cysteine mutations, the G12C cancers. In lung cancer, we showed a response rate of roughly 38% with a very high disease control rate, you know, in the 100% range. And in pancreatic cancer, we showed also a high disease control rate in the high 80% range, and an objective response rate of around 20%. Both of these compare quite favorably to benchmarks for second line and more advanced cancer. One element to consider in thinking about those and comparing those is that, in our study, of course, when we say second line, it really means second line, third line, and fourth line, so whoever gets into the study, and in both the lung and pancreatic cancer patients. So it was a heavy bias towards patients who had had multiple lines of therapy, not just pure second line. But we nonetheless compare those numbers to pure second line patients to establish standard of care. I think both for the disease control rates and the objective response rates, these are significantly ahead of those industry benchmarks. Okay, you know, at your latest update, earlier this year, you sort of talked about the response rates improving, relative to what you had shared last year. Yeah. What could be driving that improvement in responses? And, you know, how do you see these response rates translate into durability of efficacy over time? Yeah, sure. Well, one of the challenges in an ongoing study in reporting response rates is that there's a constant dynamic from literally day to day, where patients are being added to the study so they become part of a denominator, which can often happen before they've had time to be evaluated. And so the denominator can change on a regular basis. And then the numerator changes when some patients have had a chance to have one or two scans, and that can happen in a time in which maybe the denominator hasn't changed. So there are all these, all these moving parts. And so getting a bead on a definitive response rate is quite hard to do during an evolving study. You add to that, of course, that we're studying different doses, and then finally that the duration of observation, from the early report is rather short. And so patients are being added. Many of them might have been on for only a single time to get a single scan. And we have observed with RMC-6236 that the longer a patient stays on the drug, on this investigational drug, the greater the likelihood that they will convert from, let's say, stable disease, to a partial response. And we don't exactly know why, that is the case, but it seems to be part of the biology of the multi-inhibitor. And so in that context, reporting a patient collection, early in the treatment course might not give the same results as it would if you looked after a longer period of observation. So the two specific things that we thought would happen would be that as we emphasized higher doses, since the early report did not have many patients at the higher doses, but as we enrolled and studied patients on higher doses, we would see a higher response rate. Also as we followed patients longer term, that we would pick up responses that weren't observed early in the treatment course. Our view is that both of those happened just as predicted. As a result, the trends from the numbers that I commented on earlier have been favorable. We've seen response rates, now, that are evolving upward from those numbers, and we're quite encouraged by that. Ultimately, I, I'd say the exact numbers aren't really what matters. And the reason we even, you know, took the time to make those comments, in the first quarter of this year was just to assure people that the response rates in this study are not deteriorating, which can sometimes happen. But that clearly isn't happening. Response rates have actually become more favorable over time. And at some point they'll settle out and we'll have a particular response rate at the definitive dose that we're moving forward with into the pivotal trials. But it's still a moving thing in a study that's still ongoing. Okay. So you've talked about a kind of update in the second half of the year, which would include a data update along with really sort of revealing kind of what your final pivotal trial design will be in second line. Can you talk about sort of what all needs to be done between now and then for you to be able to finalize that trial design? Yeah, there really kind of three-ish parts. The first part is the most important is completing a mature enough data set with long enough follow-up to be able to really establish a much better number around progression-free survival. We've not reported progression-free survival numbers yet for either pancreatic or lung cancer simply because the data were too immature to do so. It is possible to plot the existing data that we reported back in October and to come up with what appears to be a progression-free survival number. But the confidence intervals around such a number are very wide when the data set isn't mature. And so we have been, you know, very resistant to declaring any sort of numbers until we can have confidence that what we're saying is likely to be representative of what's the truth. Now, I do know a number of people have done those analyses nonetheless and have reported to me and shown me progression-free survival curves that are quite encouraging and certainly enthusiastic to see them. But we've just not been willing to sort of put our imprint on a particular number. So we need real PFS projections from mature data sets. And we have been analyzing our data. We've said that the expectation is engaging with the FDA in the first half of this year. That's what we asserted back in February. I don't have any new update on that. And but checking that very important box of progression-free survival, having confidence around it, and knowing that it suggests a superiority over standard of care is absolutely essential to all of this. The second component is we need to select the dose. And for the dose selection, now in the context of the FDA's Project Optimus paradigm, we have to take a robust data set looking at multiple doses and compare both the tolerability and safety profile and the activity profile to find the Goldilocks number that we can all agree is the right starting point for a pivotal trial. And while we believe in this case we actually have that number, we think 300 mg is a very attractive dose. It's the case that 200 mg is a pretty attractive dose also, and 160 mg is a pretty attractive dose as well. And that's evident from the data that we showed back in October. Those are very active doses. So there is room for some discussion, and it's our obligation to present a robust data package to the FDA to help arrive at a mutually agreed on go-forward dose. And then the third element is to get agreement, an alignment with the FDA around a specific protocol. And, you know, we don't think that that's a particularly, you know, sort of risky topic, but it is an area where there are some bells and whistles and some features that can be dialed in or dialed out. We certainly have to agree on what the endpoints should be, and then we can go forward. So those are the three things that are in progress, if you will, aspects of those. Those are really primarily first half of this year activities. We would expect that coming out of the other end of that is a dose, a trial design, the data to support moving forward relative to established benchmarks. And that information then, we reported back in February that we would share that in the second half of this year, as we go forward to launch a pivotal trial. I, I will point out that the data set for lung is not the same as the data set for pancreatic cancer. The safety data set is the same, but obviously the activity set is different. And the trial designs may not be identical. And so, ultimately, there have to be two different engagements with the FDA around those. And we wouldn't expect them to happen at exactly the same time. We expect them to be phased. So, that is, that's where we stand today. Mm-hmm. So what you mean is you're sort of waiting for the data for both lung and pancreatic to mature to a certain point before you have that conversation with the FDA and then finalize some of the elements of the design, etc.? Yeah. Well, I didn't quite say we're waiting for it. I'm not stating where we are exactly in that process, but that, maturing the data set, maturing the analysis, the analysis is pretty extensive for Project Optimus. You know, it's 100 pages of graphs and charts and so on. It's a lot of work. We need to make sure it's accurate and so on. But completing that analysis, preparing a package with the FDA, submitting it to the FDA, getting feedback and dialogue with the FDA, that's what we're in in the first half of this year. Okay. Now the data that you shared with us last year, in an updated earlier this year was in G12X patient populations. Yeah. Obviously there's sort of that's about maybe half the patient population, and there's sort of the big group of G12C patients and also the non-G12C categories of mutations. Where are you in the process of evaluating or generating any data in those mutations, and how will that be evaluated to, you know, for consideration into inclusion into the pivotal trial? Right. So maybe just to clarify, in lung cancer, the breakdown is, roughly half of the lung cancer patients with a RAS mutation, roughly, half of, maybe I should put it a different way, of the 25% who have a G12 mutation, 25 out of 30% of, of lung cancer tumors that have a G12 mutation, about half of those have a G12X mutation that's not cysteine, and about half of those have a cysteine mutation. So it's roughly 12% and 12%, 12.5%, 12.5%. And then the remaining 5% have a non-G12 mutation. So those are the groups to consider. All of the data that we had reported previously was in the G12X without C mutation. And we have indicated that we are interested in G12Cs, especially for lung cancer. But it's not an essential component. There's a lot of activity in the G12C space, including our own G12C selective inhibitor. G12C is the most competitive space for getting access to patients because there's so many trials going on and there are two approved drugs. So it is hard to access those patients. We have finally now, though, opened a cohort that allows G12C patients to enter. I don't know how many patients we'll get and how fast we'll get them, but we did open a cohort. I should emphasize preclinically, the G12C tumors are highly sensitive to RMC-6236, the G12C lung cancer tumors. It is an extremely good G12C inhibitor, even though we've not reported any human data on G12C. The third bucket for lung cancer is the non-G12 mutations. That includes Q61 and G13. What we had reported earlier in the year is that we would study patients. We would allow patients into the trial who have G13 and Q61 mutations. We opened that cohort. In fact, we have enrolled patients. And just at the AACR, we disclosed a set of cases that we think are really quite interesting. They weren't particularly lung cancer patients, but they were patients with other mutations beyond G12 mutations. And I think that that collection of cases, which included other tumor types as well, showed some very exciting, albeit anecdotal evidence that RMC-6236 is highly active against the KRAS Q61 mutations. So I think that we can feel very good that there's no mechanistic reason why Q61 mutations can't be included in a pivotal trial. And then there was a very interesting case shown of colorectal cancer, a patient who had a BRAF mutation as their primary driver for their colorectal cancer. They were treated with a BRAF inhibitor, and then developed resistance. And the resistance was accompanied by three different RAS mutations emerging in that patient who did not have known RAS mutations previously. And those mutations included a G13 mutation, a G12 mutation, and a Q61 mutation. And that patient then went on RMC-6236 and actually showed a partial response. So I think the evidence now is good, albeit not highly quantitative. We don't have dozens or hundreds of patients to report. And I think we feel confident that we can now incorporate those mutations as eligibility into a pivotal trial because we have less information about those. We have less certainty that what we're seeing is representative, just objectively, will likely statistically do something to make sure that we don't undermine the power for the core group, which is the most important group for us, the G12X group. But we will allow, I expect that we will allow, tumors with mutations beyond G12 to be studied. In pancreatic cancer, the same is true, although G12C is very uncommon. The rank order for mutations in pancreatic cancer, 92% of pancreatic cancer carries a RAS mutation. The rank order is G12V, G12D is the most common, G12V is the second most common, and G12R is the third most common. And all three of those mutations have shown sensitivity objectively in our clinical studies so far. We've reported that previously. In fact, Wei Lin, our Chief Medical Officer, showed a case of pancreatic cancer with a G12D mutation who was treated with RMC-6236 and showed a complete response. Complete responses are very uncommon in pancreatic cancer. So I, I think the aggregate of all of that, including our preclinical data, suggests that we should include those mutations, in our clinical trials. Okay. Very good. Just a clarification on lung, with regards to G12C, sounds like you still need to generate data to decide whether you'll include that or, or not, or. We might, but I think at this point, our confidence level is sufficiently high around G12C from, you know, years of preclinical data that that was never really gonna be especially gating. So I don't think that will be an issue. We may get some G12C patients, but it's just hard to imagine that it's not going to be sensitive to RMC-6236. Yeah, that makes sense. I just received a question from the audience, just with regards, you know, between kind of PDAC and lung, which data are we likely to see first? I presume that you're gonna share a data update combined with sort of the final design that's agreed upon with the FDA together. Is that the right way to think about it? I think for each indication, when we've got sort of the full package, we'll give a full presentation. I think we won't be sort of slicing and dicing it. I think everybody just needs to see it all put together, in part because the data inform the design, so it makes sense for it to all be shown at once. But each indication is separate, as we just discussed. I don't have any update to you on timing other than that we expect both of these in the second half of this year, but we don't expect them necessarily to be at the same time. Mm-hmm. And, at this point, you can't tell us whether, which one will be first? I don't have any comment on that today. Okay. Sounds good. All right. I wanted to also talk about first line. Obviously, the next sort of natural step is to think about how to move these assets forward, or, you know, 6236 forward into first line. What work are you doing this year to determine the best approach for this? Would you, you know, anticipate it to replace chemotherapy or be an add-on to existing standard of care? Yeah. It's a great topic. My simple answer is yes. We expect it to either replace or be an add-on. We don't know yet. What we really need to do is just to evaluate whether RMC-6236 can be combined with chemo. We don't have any reason to believe that it cannot be preclinically. We don't have any evidence that raises a concern. But since we are combining a novel, unapproved drug with an approved drug, we need to just make sure that there's not gonna be an unexpected safety problem. So we will be combining RMC-6236 with chemo, both chemo that's relevant for lung cancer and chemo that's relevant for pancreatic cancer this year. I don't have any update to report on that, but we will be doing that as part of evaluation for what to do, what we might be able to do in first line. It's especially important, in pancreatic cancer, where, you know, both are highly credible. It obviously depends on what we are seeing in second line as to how strong the case is for going into first line. So that's something I'm sure people will be looking for. But on the assumption that we would be able to advance into first line, I can tell you we'd wanna accelerate that. That would not be something we would wait for the end of the pivotal second line trials. We absolutely would move as forward as quickly as we can. And so one thing we need to vet is can RMC-6236 be combined with chemo? Then a separate thing is, does it need to be combined with chemo? Those are kind of two different questions. So I think that's going on this year. Could I just sort of ask you a follow-up on this just in the context of pancreatic cancer? Because obviously that's chemo is the standard of care in first line, but there's also a fair amount of interest in potentially replacing chemo, and chemo doesn't exactly do much for patients anyway to begin with. So I guess what would you need to see in late line patients? And you know, you're already generating data in late line, to sort of feel confident that potentially you know you could explore one or both in first line pancreatic cancer. Yeah, it's a good question. And you're, you're absolutely right. I mean, our, our overarching goal in the long run is to replace chemotherapy. We don't think anybody wants to take chemotherapy or should want to, and particularly in pancreatic cancer. You know, FOLFIRINOX is not a not a fun experience to go through. And, and for such a lethal disease to spend a good part of your remaining life, experiencing chemotherapy, we can all agree is a bad idea. But, it is more common to add agents onto chemotherapy as a stepwise way to go. And we'd certainly wanna make sure that ethically we weren't depriving patients of some of of something that's, that's established. And, and actually in first line pancreatic cancer, I mean, response rates are in the, you know, 30%-ish, 30%-35% range. So that's, it's not that's not a trivial response rate. Durability is limited, but is superior to what it is in second line. Overall survival is somewhat improved by first line chemo, but still not at the level we'd like to really have impact. So I don't think we can discount chemo yet, but we would love to. We would love to be able to prove that it can be discounted. I'd say that we're going to end up making some judgments based on the second line PFS data that we end up seeing that will clearly influence our thinking. I don't mean in the pivotal trial. I mean the data that are currently being matured and analyzed that will help us. Then I just think we need to combine them to establish whether there's a safety issue or not. If there isn't a safety issue, at that point, we just might end up making a judgment call about whether the pivotal trial in first line should be a three-arm or a two-arm trial, just as you raised. Should we, potentially, study both options? From From a doctor's point of view, having both options could be very powerful, because there's some, maybe some patients who say to their doctor, "I want the most aggressive therapy available. I'm willing to go through chemo, and I'd like to have chemo plus that targeted agent." Other patients aren't eligible for chemo, even if they want it, couldn't get it. So those are the two bookends. For those patients, you'd wanna have monotherapy available as an option. I think that that could be a way to rationalize all of this. But let's just see the data mature some. Let's get the safety information, and then we'll make a judgment about what's the best strategy, and we'll consult with investigators to get their advice. And the safety data that you're generating is chemo plus 6236 in perhaps second-line patients? Correct. Yeah. Okay. All right. Yeah. At least initially, I wouldn't rule out that there could be first line patients as well. You know, from an ethics point of view, it's easier to add the experimental agent to chemo for first line pancreatic cancer to make sure you're not depriving them of whatever's established, standard of care. But I think we'd start with second line and then, most likely have first line patients as well. Okay. Now, the conversation for, the path to first line in lung is slightly different because, for a large part, it is IO plus chemotherapy. And the RAS off G12C drugs have found it challenging to combine with IO. Yeah. You know, what do you know about 6236 today that can throw some light into combinability with EMRO? Yes, you know, it's been a big issue for the RAS off inhibitors. And as far as we can tell, for essentially all of them, even the latest, Merck G12C off inhibitor also showed hepatotoxicity signal in combination with their own Keytruda. Although they are moving forward with a first-line study, nonetheless, they're gonna have to navigate that particular issue. And the big question has been, why does that occur? Is that a compound selective issue, a specific issue, or does it have something to do with the biology of KRAS G12C? We don't have any logic for why it has to do with the biology of KRAS G12C. We just don't see how that plays out. We may be wrong, but we're not aware of any compelling biological argument. So it seems much more likely it's some sort of off-target effect from the compound. And in principle, that therefore could be unique to each compound. And many of those RAS off inhibitors have similar chemical structures, so it may not be too surprising that they have similar behavior in this regard. RMC-6236 has completely unrelated chemistry. It's structurally very dissimilar from those RAS off inhibitors. It binds to a completely different site and has a completely different mechanism of action. Preclinically, we did not see hepatotoxicity signal to raise a concern. But most importantly, in our study that we presented in October with about 40 lung cancer patients, roughly half of those patients had received a Keytruda within 90 days of the initiation of RMC-6236. And it turns out that how a patient handles that new agent after receiving PD-1 therapy is a good predictor of what's going to happen if you actually put the two together in the same timeframe. That's based on a paper published by Kathryn Arbour from Memorial Sloan Kettering. But of those patients in our study, we saw no significant hepatotoxicity that raised a concern. So that gives us some encouragement as maybe a preview of what's to come. We now have open a combination study of RMC-6236 plus Keytruda, so a direct concurrent treatment study. And that is underway. We opened it earlier this year, and we hope that by the in the second half of this year, we'll be able to report on it. All we need to see is that it's not eliciting the level of hepatotoxicity that's been seen with the others, in which case it absolutely opens up first line in lung cancer. So if that demonstrates, you know, safety, then would you study further in combination with chemo as well, or would it just be Keytruda plus 6236? Well, that's a layered or subtle question. I'm gonna make it even more complicated by throwing in RMC-6291 into the picture. We, you know, that's our KRAS G12C ON inhibitor, which has performed quite well so far in the clinic. We reported that, and it's an exciting molecule. One of the things that's exciting about it for us is that not only may its own profile be very attractive, but it may do something special when it's combined with RMC-6236. And we have studied this extensively in preclinical models. We've tested whether or not those two agents together delivers anything different from the two as individual agents, single agents, and indeed observed across a number of lung cancer models that showed less sensitivity to either agent alone. When we combined the two agents, it really converted into a real big knockdown, a punch. That has excited us, caused us to wanna bring that into the clinic. Our investigators have been very excited about it, and we opened up that combination, RMC-6236-6291 study, earlier this year as well and have been enrolling patients and dose escalating there. The reason I bring that up is that it is possible if that also is well behaved, that a triplet of two RAS inhibitors plus Keytruda could be imagined to be a way to displace a platinum doublet plus Keytruda, displace the platinum doublet component, the chemo component, and have a new chemo-free first-line lung cancer regimen. That would be really exciting. That could be really game-changing. We're working hard to see if that's possible. So you raise a number of possibilities of how we could move into first line. I've just thrown into the mix another possibility, and I think we have to let these play out a little bit this year before we can come to any conclusions about what's the best approach. Mm-hmm. The one thing that, so obviously some of the cohorts that you talked about, that are underway, to test these combinations, will certainly give us insight into safety of the combinations. But from an efficacy perspective, obviously you're studying them in second line, and you're gonna look at that data to decipher what is the best combination and how do you replace perhaps chemotherapy in first line. How would you determine what is your right benchmark by looking at a combination in second line? Yeah. Well, that's tough. I mean, they're not directly translatable, but it's very common for targeted agents, or for drugs in general, oncology drugs in general, if they performed well in second line to perform even better in first line. I mean, that's the standard framework that people have. You know, it all depends on how much better in second line something performs. It might give you a lot of information about what's gonna happen in first line. If it's marginal, it might be harder to tell. So, I think for this it's important here to break down response rates versus durability. The easiest thing to measure is rates. But the least important thing from a first line perspective is rates, and the most important thing is durability. And that's hard to get quickly. You just, by definition, you can't get durability without waiting and looking over time. So while we may see with 6236 and 6291, for example, we might see a higher response rate. That certainly is possible. And that would be very encouraging because in lung cancer, response rates tend to correspond to durability. So that is possible. But preclinically, the most profound thing that we saw wasn't just rates. It was actually durability. And we did an experiment, where we took seven models of lung cancer that had RAS mutations that showed for one reason or another less sensitivity to either 6236 or 6291 as single agents. And when we combined the two agents and looked over a three-month period, which is not your typical timeframe for running xenografts, that's a long timeframe. It was quite clear that across those seven models, that combination provided enormous durability benefit. Very few animals progressed during that time. That's just not something we'll see overnight, you know, in a quick study in humans. So, I think we're gonna be looking for any signs that we can get, and then we'll have to make some judgments. We'll make a decision about whether we need more data or whether we can decide what to do with the on the basis of the data that we already have. You know, this is a, you know, a real-time kind of thing where we're trying to make the best possible decisions, to have the highest probability of success for having high impact for patients. Those are our measures. Fair enough. I also wanted to touch upon RMC-9805, which is your G12D RAS(ON) inhibitor. And you showed some preclinical data on that, at AACR as well. Could you talk about that, and what did you learn from it? Yeah. Well, RMC-9805 is a pretty remarkable compound. I mean, we feel very fortunate that all three of our first RAS(ON) inhibitors and the first-ever RAS(ON) inhibitors to go into the clinic, period, have differentiated profiles. You know, RMC-6236 is this RAS multi-inhibitor. It's noncovalent. It binds to every variant of RAS we've ever tested. That gives it a particular profile. RMC-6236 is a G12C selective inhibitor. So on some level, it's analogous to the G12C off inhibitors, but it only binds to the on form. So that's a second profile. And then RMC-9805 is a G12D inhibitor, which shares some properties of the first two compounds I just mentioned. It is orally bioavailable. We've demonstrated that preclinically, and we've disclosed publicly that it is behaving well as an orally bioavailable agent in humans, although we've not disclosed data to support that statement yet. It is not only G12D selective, but it is selective as a covalent compound. It binds initially to RAS, and if there's an aspartic acid there at position 12, it then engages covalently with that asp with that, aspartic acid. It's the first drug-like molecule to engage a KRAS G12D target. And frankly, it's the first drug-like molecule to engage any aspartic acid across the entire therapeutic spectrum. So it's really groundbreaking. And our chemistry team supported by biology and pharmacology really has done a remarkable thing by taking advantage of the tri-complex modality of binding that we've described many times over the last few years that allowed them to position the warhead, the covalent warhead, so optimally at the aspartic acid position 12 that it could engage even though it's not a highly reactive warhead. We used a very low reactive warhead, to achieve this, and that gives it high selectivity. We're really excited about it. Obviously, G12D is the single most common RAS mutation in human cancer. So having multiple ways to attack it makes sense from a company strategy point of view. I'd emphasize that RMC-6236 is a pretty darn good KRAS G12D inhibitor, and we now have a lot of data driving us into pivotal trials that is heavily based on KRAS G12D-containing tumors. So, to us, the benchmark is actually now RMC-6236. And, what we don't know is whether RMC-9805 will show activity in tumors that RMC-6236 shows less activity in, or whether RMC-9805 might be combinable with some agents that RMC-6236 can't be combined with, or whether RMC-9805 should simply be combined with RMC-6236 in the same way that I just promoted the idea of RMC-6291 plus RMC-6236. So what we've created, with this pipeline, is a tremendous amount of optionality for us. And given how difficult RAS tumors are to defeat, given how complex they are, we think having this armamentarium of compounds with different profiles gives us the optionality we need to find the best options for patients, and we'll continue doing so. Okay, Mark, this is all we have time for today. So I'll have to close the session. But thank you so much for sharing all the progress of the company, and wish you all the best, as the company makes progress. Thanks. Thank you, Ami.