Revolution Medicines, Inc. (RVMD) Earnings Call Transcript & Summary

Good day, and thank you for standing by. And welcome to Revolution Medicines, Inc. Conference Call. [Operator Instructions] Please be advised that today's conference is being recorded. I would now like to hand the conference over to your speaker today, Ryan Asay, Senior Vice President, Corporate Affairs, please go ahead.

Thank you, and welcome everyone to our webcast where we'll summarize data that were presented at the EORTC-NCI-AACR or triple meeting this week and provide additional context. Joining me on today's call are Dr. Mark Goldsmith, Revolution Medicines' Chairman and Chief Executive Officer; and Dr. Wei Lin, our Chief Medical Officer. Dr. Steve Kelsey, our President of R&D will join us for the Q&A portion of today's call. As we begin our presentation, I would ask that you review our legal disclaimer on Slide 2. With that, I'll turn the call over to Dr. Mark Goldsmith, our Chairman and Chief Executive Officer. Mark?

Thanks, Ryan, and thank you to everyone who has joined us today. As outlined on Slide 3, at Revolution Medicines, our mission is to revolutionize treatment for patients with RAS-addicted cancers through the discovery, development and delivery of innovative targeted medicines. Today, we'll summarize the data presented at the triple meeting this week that highlights the progress we are making toward that goal. As a reminder, we have a deep portfolio of targeted RAS(ON) inhibitors including 3 programs that have demonstrated compelling clinical proof of concept. Our remarks today will focus on 2 of these clinical stage programs, RMC-6236, our RAS(ON) multi-selective inhibitor and RMC-9805, our RAS(ON) G12D selective inhibitor in previously treated pancreatic ductal adenocarcinoma or PDAC. In Slide 4 is a reminder that RAS acts as an intracellular switch, transitioning between an inactive GDP-bound, a state we refer to as RAS(OFF) and an active GTP-bound state we call RAS(ON). Tightly regulated RAS(ON) proteins are needed to control normal cell growth. Mutations in RAS proteins can drive uncontrolled RAS(ON) signaling pathways that favor cell growth survival and oncogenesis. Pancreatic cancer is a devastating disease. It is the third leading cause of cancer mortality in the U.S. with approximately 60,000 people expected to be diagnosed and 50,000 people expected to die this year alone. PDAC is known to be a RAS driven cancer. Over 90% of PDAC tumors harbor an oncogenic RAS mutation, and RAS proteins are also likely to play a role in many of the remaining 10%. G12D is the single most common RAS mutation. There's clearly a need to improve outcomes in patients with metastatic PDAC currently treated in the second line setting with standard of care chemotherapy-based regimens. I'll briefly highlight on Slide 5 the unique approach we have taken to targeting RAS. Our innovative compounds bind to and inhibit the active or ON form of RAS. We call these compounds tri-complex inhibitors because they are designed to bind selectively and with high affinity to a relatively flat surface on the RAS target by forming a tri-complex structure that contains RAS, a highly abundant cellular chaperone protein and the tri-complex inhibitor bound in a pocket between them. Preclinically, these compounds directly and rapidly inhibit the intended oncogenic RAS(ON) cancer drivers in tumor cells, inducing deep and durable suppression of RAS cancer signaling by a mechanism of action that is unique and differentiated from first generation RAS(OFF) inhibitors. We have now demonstrated clinical proof of concept with our first 3 programs. And today, we will review new clinical data from 2 of these investigational drugs that provide further validation of this approach. RMC-6236, our oral RAS(ON) multi-selective inhibitor has been clinically validated across diverse oncogenic RAS variants including G12D. RMC-9805, our oral RAS(ON) G12D selective covalent inhibitor has been clinically validated against oncogenic RAS G12D. As context for these new results, please review the table on Slide 6 that summarizes the results of previously published Phase III clinical studies of various chemotherapy regimens in patients with metastatic PDAC. The results across treatment regimens consistently show a need to improve clinical outcomes for previously treated patients. These reported findings from clinical trial settings show that regardless of the chemotherapy regimen evaluated, the median progression-free survival ranges between 2 and 3.5 months in the second line setting with median overall survival of 6 to 7 months. The numbers in the third line plus setting are even lower. It's also important to note that these are aggressive chemotherapy regimens with approximately 2/3 of patients typically requiring a dose modification. Patients with metastatic pancreatic cancer deserve better outcomes, meaningfully longer survival with a better quality of life. And these are the aims of our development programs. I'll now turn the call over to Dr. Wei Lin, our Chief Medical Officer, to summarize RevMed data presented at the medical conference this week regarding patients with metastatic pancreatic cancer treated with RMC-6236 or RMC-9805. Wei?

Thank you, Mark. We're pleased today to update the data we first shared with you at our July 15 Investor Event. We believe that these more mature data strengthen the efficacy proof of concept we presented in July and if reproduced in the ongoing Phase III study would change the standard of care for patients with previously treated metastatic pancreatic cancer. In addition, the safety profile of RMC-6236 has remained stable with longer follow-up, suggesting that it may be given for a longer-term as a monotherapy with the aim of providing prolonged clinical benefit. On Slide 8, as of the July 23 data cutoff date, which represents 2.5 months longer follow-up than the analysis we shared on July 15, 127 patients with pancreatic cancer have been treated with RMC-6236 at doses between 160 and 300 milligrams daily. Pooling the results from the 160 to 300 milligram dose levels provides a more robust assessment. Among them, 57 patients received RMC-6236 as second line therapy. The distribution of RAS mutations in tumors of these patients is approximately representative of what we expect to see in our ongoing Phase III trial. With the majority of patients having tumors harboring a KRAS G12X mutation and a minority harboring KRAS, NRAS or HRAS mutations at other immunoassay positions. The proportion of patients with poor prognostic factors such as ECOG performance status 1 or liver metastases is similar to or higher than the proportion in other trials in previously treated metastatic pancreatic cancer, such as the NAPOLI 1 study that supported approval of the Navari chemotherapy regimen. So in addition to RAS mutation status, we also believe that other demographic and risk factors of patients reported in this analysis are representative of those we expect to see in our ongoing Phase III trial in second line pancreatic cancer. Moving to Slide 9. Here, we see a more mature progression-free survival analysis of pancreatic cancer patients treated with RMC-6236 at doses between 160 to 300 milligrams daily. This second line pancreatic cancer patient population is more relevant to our ongoing Phase III registration trial. The Kaplan-Meier curves are represented by the solid green and yellow lines on the left and the median PFS estimates are listed in the top table on the right. In patients who received RMC-6236 as a second line therapy, the median follow-up is 6 months for patients with KRAS G12X pancreatic cancer and 6.2 months for the broader RAS mutant population, and the minimum follow-up is approximately 3 months. The median PFS in the KRAS G12X mutant population is estimated to be 8.5 months compared to 8.1 months in our July presentation. The median PFS in the broader RAS mutant population is estimated to be 7.6 months, which is the same as in the July presentation. The PFS rate of the 3-month mark was proportionate patients who are progression-free at 3 months is approximately 90%. These stable or even improved results after longer follow-up increase our confidence in the estimates and potential translatability to our pivotal trial. On Slide 10, we show overall survival, which of course is the ultimate measure of efficacy. We've just seen that the median PFS for RMC-6236 in second line metastatic pancreatic cancer has been substantiated in a more mature dataset. Now we're pleased to report that with a longer follow-up, we have our first estimate of median OS in second line metastatic pancreatic cancer, and it is 14.5 months. The estimate is the same in KRAS G12X and the broader RAS mutant populations. The lower bound of the 95% confidence interval is 8.8 months in both populations. The OS rate at 6 months, which is the proportion of patients who remain alive 6 months after starting treatment with RMC-6236 is approximately 90% in both the KRAS G12 and the broader RAS mutant populations. We're encouraged by the consistency between PFS and OS outcomes and within the 2 mutant defined populations, as well as their consistency over time. The best change in target lesions and the objective response rates are presented in Slide 11. While the overall shape of the waterfall plot has remained the same compared to our July presentation, some patients have experienced deepening of response, which is consistent with the observation that only half of all responders achieved their initial RECIST response within the first 2 months of treatment with RMC-6236. Similar to the July presentation, the pooled ORR analysis of 160 to 300 milligrams daily presented in the table on the right includes all KRAS G12X patients received their first dose of RMC-6236 at least 14 weeks before the data cutoff date, which is a period that allows for 2 potential on treatment CT tumors assessments. Unlike the July presentation, here we have analyzed separately the ORR for the second line population and the third line population. Due to the longer follow-up, the overall efficacy evaluable population inclusive of all second line and third line plus patients has increased from 97 to 105. The ORR in second line patients was 29% and in third line plus patients was 22%. The DCR for both populations was approximately 90%. These numbers are consistent with the ORRs we reported in July, consistent with the general trend to greater efficacy when a therapy is used in an earlier line treatment and supportive of the early observation as responses increased with longer time on drug. As shown in Slide 12, these findings were obtained at well tolerated doses of RMC-6236. The treatment related adverse events or TRAEs in pancreatic cancer patients treated at doses between 160 and 300 milligrams daily of RMC-6236 are presented in this table. The overall safety profile has remained consistent with what we presented in July. While there were some modest numerical changes in the adverse event rates, we believe these do not meaningfully change the overall safety profile. Both any grade and Grade 3 or higher TRAEs are listed here for the 127 patients treated at these dose levels. Again, we present here a pooled safety analysis because the adverse event rates were similar across these dose levels and the larger database provides a more robust assessment of the safety profile. The TRAEs that occurred in 10% or more patients are listed in the top half of the table. They mainly fall into 3 categories, dermatologic, which include rash and paronychia, GI, which include nausea, vomiting and diarrhea, and mucosal, which include stomatitis and mucosal inflammation. Most adverse events were low grade with 29% of patients experiencing Grade 3 or higher adverse events. These adverse events have generally been manageable and reversible. Rash, a bundle term that included acneiform and maculopapular as well as other skin reactions occurred in 91% of patients. These were mainly low grade with around 8% of patients experiencing Grade 3 rash. These rash events were predominantly acneiform, early in onset and responsive to the same management used for EGFR inhibitors. Therefore, we have recommended rash prophylaxis with oral antibiotics in the ongoing randomized Phase III second line pancreatic cancer trial. Based on the experience of rash prophylaxis for EGFR inhibitors, we believe we may see a reduction in both the frequency and severity of rash and improvement in the tolerability of RMC-6236 within the introduction of prophylaxis. Stomatitis and mucosal inflammation occurred in 31% and 13% of patients respectively. These adverse events are similar to those associated with mTOR inhibitors where a simple steroid mouthwash has been effective as prophylaxis. We have recommended this prophylaxis as well in our Phase III trial aiming to further improve overall tolerability. The lower half of the table lists several notable adverse events that are uncommon in patients treated with RMC-6236, but are worth highlighting because of the implications on the combinability of RMC-6236 with other anticancer therapies used in first line PDAC, colorectal cancer and non-small cell lung cancer. The low rates of cytopenias and liver enzyme elevations suggest improved chance for combinability with chemotherapy and immunotherapy respectively. On Slide 13, we see how the TRAEs affected patients' ability to remain on RMC-6236 treatment. As of the data cutoff, the median time these patients have been on treatment with RMC-6236 have been extended to 5.3 months with this analysis. Even without adjusting for treatment duration, the dose modification rate remains favorable. Around 1/3 of patients required any dose modification of RMC-6236. The majority of these were dose interruptions with only around 1/5 requiring dose reductions. Remarkably, none of these 127 patients discontinued RMC-6236 due to TRAEs. These numbers translated to a mean dose intensity of 92%. We project that the high dose intensity and modest dose modification rates may contribute to clinical benefit, and we hope the tolerability profile confers an improved quality of life for patients on RMC-6236. Slide 14 shows the design for our registrational trial to enable the first approval of RMC-6236 in second line pancreatic cancer, which is called RASolute 302. In this randomized Phase III trial, all patients with metastatic pancreatic cancer previously treated it with one line of systemic therapy will be eligible, including patients with tumors harboring RAS G12X mutations, other RAS mutations and without an identified RAS mutation. The exceptions are patients with tumors harboring an actionable mutation for which an approved target therapy is available, for example, a BRAF mutation. We plan to randomize approximately 460 patients to either monotherapy RMC-6236 or an investigator's choice of standard therapy. The dual primary endpoints will be PFS and OS in patients with pancreatic cancer harboring a RAS G12X mutation. Should efficacy be demonstrated in this core population, PFS and OS will be evaluated in all enrolled patients. Additional secondary endpoints include ORR, DOR and quality of life measures. We're pleased that we have begun dosing patients on this trial. Moving to Slide 15. To summarize, with 2.5 months of additional follow-up since our July presentation, the PFS estimates in second line metastatic pancreatic cancer have remained consistent or improved, with a median PFS of 8.5 months in patients with PDAC harboring a G12 mutation. Our first estimate of median OS is 14.5 months, consistent with the observation that 90% of patients who started treatment with RMC-6236 were still alive after 6 months. The RASolute 302 Phase III trial in second line pancreatic cancer is ongoing. Enthusiasm from pancreatic cancer investigators and patients worldwide has been high. The primary analysis for PFS and OS will be vet driven. Now I want to provide some additional context shown on Slide 16. Earlier in the presentation, we showed a slide indicating the median overall survival in second line pancreatic cancer is between 6.1 months and 6.9 months for patients receiving standard care chemotherapy. This slide lists the 3 approved therapies for patients being treated in the first line metastatic pancreatic cancer. The median survival from the registrational studies for these 3 first line patient populations range from 8.5 months to 11.1 months. The unmet need for patients with metastatic pancreatic cancer across all lines of therapy remains very high. The encouraging data just described for RMC-6236 in the second line pancreatic cancer treatment setting also support the evaluation of RMC-6236 with monotherapy in first line and early disease settings. And we are currently designing a first line registrational trial. Transitioning to RMC-9805, I'll briefly discuss preliminary safety and antitumor activity from the ongoing Phase I study of our oral RAS(ON) G12D selective tri-complex inhibitor in patients with KRAS G12D pancreatic cancer. Moving to Slide 18. RMC-9805-001 is a first in human study of RMC-9805, evaluating the safety, tolerability and preliminary antitumor activity of RMC-9805 in patients with advanced solid tumors. The study evaluated RMC-9805 as monotherapy, which consisted of dose escalation in Part 1, followed by dose optimization and expansion in Part 2. Slide 19 shows that RMC-9805 exhibited a dose dependent increase in exposure from 150 milligrams to 1,200 milligrams once daily dose. Shown on the left is the main steady-state concentrations from Cycle 1 Day 15 following once daily administration. Shown on the right is the individual steady-state blood AUC data for both daily and BID schedules. Comparable exposure was observed between daily and BID schedules at equivalent daily dose levels. The dash line represents the mouse 100 milligram per kilogram human equivalent AUC and is the dose that induced robust tumor regression in the majority of PDAC preclinical models. Oral dosing in humans achieved comparable dose exposure levels and objective responses were achieved in the clinic at doses greater than or equal to 600 milligram daily dose as predicted by preclinical models. The PK data support a 1,200 milligram once daily dose as a candidate recommended Phase II dose in patients with pancreatic cancer. Slide 20 shows the waterfall plot of best percent of change in tumor size for all pancreatic cancer patients who received their first dose of RMC-9805 at least 14 weeks before the data cutoff date, which allows for 2 potential scans. The green bars represent patients treated with RMC-9805 at 1,200 milligrams daily and the yellow bar represent patients treated with less than 1,200 milligrams daily. Among pancreatic cancer patients treated with 1,200 milligrams daily dose the objective response rate was 30%, which includes patients with a confirmed response or a response that is pending confirmation. The disease control rate was 80%. The arrows under the waterfall denote the patients who remain on study treatment. As the numbers of post baseline scans listed below the waterfall suggest, the duration of follow-up so far is short. Therefore, other efficacy endpoints such as PFS and OS will need to wait for longer follow-up. RMC-9805 was well tolerated across all dose levels as reported at the E&A meeting. For this summary, as shown on Slide 21, I'll focus my remarks on the candidate RP2D of 1,200 milligrams, which included 60 patients at 1,200 milligrams daily and 39 patients at 600 milligrams BID. Treatment-related AEs occurring in 10% or more of the 99 patients are shown here. GI toxicities and nausea, diarrhea and vomiting were the most common TRAEs, all of which were Grade 1 or 2 in severity. They were manageable and reversible. In fact, over 80% of nausea were Grade 1 and around half of these events occurred and resolved in the first 4 days of starting treatment. 10% of patients at the 1,200 milligrams dose also experienced Grade 1 rash, which is likely due to some activity against wild-type RAS protein at the highest dose evaluated. No treatment related Grade 4 or 5 adverse events or serious adverse events have been reported. ALT and AST elevations were uncommon and mostly Grade 1 and only 1 patient experienced a Grade 3 effect, which predicts a low risk for hepatotoxicity when combined with immunotherapy. This exceptional safety profile translated to a high degree of tolerability with only 4% of patients requiring a dose reduction due to TRAEs and no patients had a dose discontinuation due to TRAEs. In fact, as was highlighted by the same by the case report by Dr. David Hong shared at his oral presentation at E&A, more than one patient being treated with RMC-9805 has asked their investigator whether they were on placebo. These questions captured the patient experience with RMC-9805 better than any safety table we can show. In summary, on Slide 22, RMC-9805 is a potent mutant selective covalent inhibitor of RAV on G12D. RMC-9805 has shown encouraging antitumor activity in patients with KRAS G12D pancreatic cancer. However, assessment of durability will require longer follow-up. RMC-9805 is well tolerated at all dose levels evaluated, including the candidate RP2D of 1,200 milligrams daily. The preliminary safety and efficacy data support the ongoing development of RMC-9805 as a single agent and in combination with other therapies, including our RAS(ON) multi-selective inhibitor RMC-6236 in pancreatic cancer as well as in other solid tumors such as non-small cell lung cancer and colorectal cancer. With that, I'll pass the mic back to Mark.

Thank you, Wei. I'll conclude our prepared remarks on Slide 24 with a reminder about why this work is so important and why these results have the potential to be so impactful. Metastatic pancreatic cancer is a devastating disease and there is an urgent need to provide new and better treatment options to patients with both improved efficacy and greater tolerability. PDAC is a RAS driven cancer with the vast majority of tumors having a KRAS G12X mutation. Of these, G12D is the single most common RAS mutation representing approximately 40% of all pancreatic cancer cases. RMC-6236 and RMC-9805 are the first 2 oral investigational targeted therapies to demonstrate encouraging clinical activity in KRAS G12D cancers, and they have done so at doses with acceptable tolerability profiles. RMC-6236 has shown acceptable tolerability and a promising efficacy profile, which has driven us to initiate the ongoing Phase III trial in second line PDAC. RMC-9805 has shown acceptable tolerability and encouraging initial antitumor activity with mature ORR data and durability assessments such as PFS and OS pending longer follow-up. While RMC-6236 and RMC-9805 are likely to serve partially overlapping patient populations, these RAS targeted therapies have distinct mechanisms of action and complementary profiles that could enable multiple treatment options consistent with our aim to deliver the best possible clinical outcomes to the broadest number of patients. To support our commitment to serve the widest range of patients with best possible treatment regimens, in addition to exploring the potential of these investigational drugs as monotherapies, we are investigating multiple rational combination strategies, including as RAS inhibitor doublets and with non-RAS directed therapies. These approaches serve our objective to develop meaningful new treatment options for patients with RAS addicted cancers across a range of tumor types, ages of disease and lines of therapy. The data updates we've outlined today highlight the significant opportunity we believe we have to deliver on this objective and the initiation of our first Phase III clinical study in second line PDAC is an exciting step on our path to doing so. With that, I'll turn the call over to the operator for the Q&A session.

[Operator Instructions] And our first question comes from Michael Schmidt from Guggenheim.

Congrats on the data presentations this week. I just had a clarification on the 9805 response rate data, Mark. So when I look at the waterfall chart, I see actually 14 responses at 1,200, 8 ongoing unconfirmed PRs and 6 confirmed PRs which would result in a 35% response rate. Can you just clarify what the -- how you end up at that 30% response rate? And then I had a follow-up.

Yes. Michael, thanks for your question. I'm going to ask Wei Lin, our Chief Medical Officer to comment on that.

Yes. So that what you're referring to, I believe, is a difference between how you count the waterfall versus the table. The waterfall include all the patients who are evaluable versus the table included only patients who had 14 weeks of a minimum follow-up of 14 weeks, meaning that they were dosed 14 weeks before the sort of inclusion on the tables. We had at least dosed for 14 weeks before, so which includes the potential for 2 scans. So that's the difference. So the -- in some ways, that's why the calculation of the ORR is different between what you can count from the waterfall and the table.

Okay. That makes sense. And then obviously, the data looks -- for 9805 looks very encouraging. And I was just wondering, how you expect the data to improve with further follow-up? When we look at the 6236 data, obviously, the data matured nicely over time. And I was wondering if you were to expect similar, longer duration responses with the 9805 program?

Yes. Thanks, Michael. Of course, it's hard to predict the future. So I think we want to avoid making predictions about it. But our experience with RMC-6236 is as you well know and just described was that we were initially reporting a response rate of around 20% and we did not have any durability data. And over time that response rate stabilized at a significantly higher number. So that may happen here. It's difficult in an ongoing dose escalation study to sort of hone in on what the real response rate is until that study is no longer enrolling patients and every patient has been followed for a number of cycles simply because the denominator and the numerator might move at different rates and so the numbers could go up or down at various points. So it is certainly possible that we'll see some change over time, but we can't make any predictions about that right now.

And our next question comes from Marc Frahm from TD Cowen.

Congrats on both data sets here at the triple. Maybe first on Slide 19 where you showed the PK curves for 9805. Can you just kind of explain why the efficacy analysis is cut at 1,200 milligrams since the exposures look pretty similar once you're up above 900 milligrams. And as you mentioned in the prepared remarks, you're seeing responses at 600 milligrams and above? Just, why is 1,200 milligrams the right cut point?

Maybe Dr. Kelsey wants to comment on that.

1,200 is our candidate recommended Phase II dose, which is why we focused our attention on that. I think to include any other dose level in the analysis at this stage may be confusing in the long run. The exposures, by the way, are not completely identical between the 1,200 milligram dose levels and the 900 milligram dose levels. I think if you -- the plots that you see on the right were mainly put there to illustrate the comparison with the mouse data 100 mgs per kg. But if you look at the plot on the left, you'll see that the brownie gold line is clearly higher than the red line. And that -- admittedly that data is only for patients that were dosed once a day versus twice a day. But we are seeing slightly higher exposures and slightly higher Cmax with the 1,200 milligram dose level. So this is not strictly speaking like the RMC-6236 analysis where we pulled 160 milligrams to 300 milligrams. I think with the safety profile we have, we're fairly confident that we're going to make a very strong case from moving forward to 1,200 milligrams daily as our dose. So that's the focus on that dose for all the efficacy data.

I'd like to add just one point to Steve's comment, which is keep in mind that RMC-6236 is a non-covalent compound. And so the area under the curve analysis that we've historically shown for that is probably the most relevant parameter for projecting target coverage. RMC-9805 is covalent compound and the AUC is probably not the most informative parameter to look at. So I think as Steve said, with the good tolerability up to and including 1,200 milligrams, the fairly remarkable safety profile, it makes sense for us to drive that number up to 1,200. There's really no disadvantage to doing so that we can see and so we'll proceed with that.

Okay. That's really helpful. Then just also can you characterize the level of inhibition you think you're driving here with 9805 versus, say, with a G12D patient might be experiencing at the go forward dose with 6236. Like, is this an order of magnitude more inhibition, something more subtle? Just how would you kind of characterize the relative inhibition between the 2 assets?

Something more subtle would be the answer to that question. Maybe Steve can [indiscernible ]

Well, I think it depends on how you measure it. We obviously are not measuring it in patients. The only surrogate of clinical activity that we measure in patients is ctDNA data, which we will collate and disclose in due course. But if we look in the mouse models, the combination of the covalent binding of 9805 to G12D with the noncovalent inhibition of G12D that's due to the tri-complex is basically means that we have to measure downstream markers of pathway activation. And when we measure downstream marks of pathway activation in preclinical models treated with 9805 100 milligrams per kilogram, we see 100% inhibition of pathway activation. So I -- it's very hard to compare against 6236, which also induces 100% inhibition of downstream pathway activation when you dose the equivalent of 300 milligrams daily. So by definite -- I mean, clearly, if one of them is inducing 99% pathway inhibition, the other one is inducing 99.9% pathway inhibition, then yes, we have an order of magnitude increase, but we can't detect that unfortunately.

And our next question comes from Eric Joseph from JPMorgan.

And, yes, let me add my congrats on this great data set at the conference. Just on the -- a couple of questions on the 6236 update. Just looking at the maturing safety profile, you still have very low rates of Grade 3 elevated transaminase levels. But can you elaborate a little bit on sort of the onset of Grade 3 and higher events that the few that you're seeing here and whether there were any sort of other clinical factors contributing to that signal? And then just on dose intensity, it's come down a little bit since the July presentation. Can you say where mean dose levels are shaking out and whether any patients are managed at doses less than 160 milligrams?

Thanks, Eric. Thanks for your questions. I think Wei can take these on one at a time. So the first question had to do with, are there -- is there any -- are any of these events occurring later in time? And is there anything else you could say with regard to the liver enzyme increases?

Yes. They do occur over a spectrum of period of time. I think majority of the AST, ALT do occur early in the first 3 cycles. And then it becomes less and less common with longer follow-up if they don't emerge early. I think the numerical -- even though there are slight numerical changes with these, I think just given the patient population that's to be expected. There's in general, there's some variability. When we look at these liver enzyme elevations, in fact, just looking at shift tables, there's as many patients who have increases, they have declining. So I think given this is predominantly -- this is a pre-pancreatic cancer patient with typically majority of them have liver metastases. You do expect some variability in liver enzyme just over time as you follow-up. We don't think the numerical changes between this data cut and the July data cut is materially different. And so from a clinical perspective, I don't think we're seeing any new signal regarding delivering some. And so we're still, I think, fairly optimistic about this level as a early indicator of potential compatibility with immunotherapy. Obviously, we're generating that data with IO and then when that data is available, we'll be sharing that.

And then the second question, I think, has to do with the dose intensity slide and whether there's a difference between 92% and.

Again, I don't think there's -- this is materially different. I think achieving a over 90% dose intensity is just remarkable itself. I think with longer follow-up, obviously, you're going to have some variability in that regard. And then there's more opportunity for dosing drop in reduction. But regardless, I think getting a dose over 90% over this longer duration, this number of patients, I think, just makes us feel fairly confident we have the right dose.

And our next question comes from Jonathan Chang from Leerink Partners.

Congrats on the data. First question, can you discuss how you're thinking about a frontline pancreatic cancer strategy across your pipeline?

Yes. That's not a short question or at least it's not a short answer, but we're certainly very committed to advancing the frontline. And I think we've already publicly disclosed that we are working to develop a strategy around RMC-6236. Now it appears we may have that opportunity with RMC-9805. And in fact, one of the most common questions we were asked here at this meeting was how quickly can we move 9805 into first line. Those may not be mutually exclusive ideas because certainly the possibility of combining 9805 with 6236 is something that we're currently exploring. It may be possible to combine with chemotherapy. And so we do need to generate some additional data that will help us define at a minimum which of those combinations are safe and whether or not we study them in a single frontline study or in more than one frontline study. So there are a lot of options being created here and we certainly can't test all of them in a Phase III context. So we'll try to prioritize. But it is certainly possible that we'll pursue more than one first line strategy in parallel because patients need it.

Understood. And the follow-up is with the data presented at the triple meeting, what is your latest thinking on the combinability of 6236 with mutant selective inhibitors in your pipeline and with anti-PD-1?

Yes. Thanks. I don't think we can give any specific update on that because we haven't shared any public information yet. Of course, the 6236, 6291 study is underway and had a head start compared to the 9805, 6236 study. And we have publicly committed to providing an initial peak under the hood at that combination by the end of the year. The 9805 study, I don't think is in -- is going to be ready to disclose that information. But the 6236 plus 6291 experiment is sort of the point of the spear for establishing that that strategy of immune selective with a multi-selective is valid and would certainly be encouraging around the possibility of other combinations like 9805 and other mutant selective inhibitors as well. But that's pending and we'll provide that update later this quarter.

And our next question comes from Ellie Merle from UBS.