Hi, everyone. I'm here with Yujiro Hata, the CEO of IDEAYA Biosciences. You guys have quite a lot going on in 2026. Maybe for those less familiar with your story, can you just provide a, you know, a brief overview of the company and maybe some milestones that you want to flag for this year?
Yeah, no, definitely. Laura, great to see you. Thank you to Stifel for hosting this event with us today. IDEAYA Biosciences, we're a leading precision medicine oncology company. darovasertib, which is our lead program, we just had positive top-line results for the primary endpoint and median progression-free survival. We also recently received RTOR to enable an expedite NDA submission process. The pre-submission module 1 is about to be submitted. We also have a ASCO late breaker oral presentation on Monday, June first. We're excited about that. This is in the first line HLA-A2 negative metastatic uveal melanoma setting.
As part of that update, we'll also be giving full waterfall plot, safety tables, things of that sort, which we did not as part of the top-line results to date. In addition, we have a very deep pipeline, and I would say the next asset to focus on is IDE849, our phase I/II DLL3 Topo ADC. We have two clinical data updates we're guiding towards by the end of the year. That will also include from our partner, Hengrui Pharma, a full response rate, PFS, and 12-month OS landmark OS data, including for small cell lung cancer as well as neuroendocrine carcinoma. It would be a large data set, over 100 patients. We'll have our data as well.
They're planning to start a phase III registrational study this year. We're also planning to start a registrational study. Beyond that, as you know, Laura, we're highly invested in MTAP, an area you know very well. We believe we have a first in class, best in class, MTA-cooperative PRMT5 inhibitor. Dose escalation is going well there. We should be starting combination studies shortly. We also have a potential first and best in class MAT2A inhibitor, IDE397, which is in phase II. We'll also anticipate in focus for us strategically is in the Ras KRAS space. We do anticipate entering into that area with a significant focus in the area of pancreatic cancer.
Lastly, KAT6/7, IDE574 in phase I, significant focus on breast cancer, prostate, CRC, as well as other indications. I know we covered a lot of ground there, as you can see, a very deep pipeline, and also, a very late-stage asset that's hopefully now, approaching commercial phase.
Absolutely. You guys recently had a very highly anticipated clinical milestone, which was, you know, the top line of your phase I/II, OptimUM-02 study. Can you just, you know, remind, or, you know, investors who are new to the story, you know, contextualize what you've presented to date, versus what, you know, really is available for these patients as standard of care right now, or lack thereof standard of care right now?
Laura, here, unfortunately for these patients, first, there's nothing approved by the U.S. FDA in HLA-A2 negative metastatic uveal melanoma. You know, really just an extraordinary unmet need, and that's really the hole we're trying to fill. I would say right now, what patients get is ipi/nivo, and then PD-1 alone, sometimes the chemo DTIC. In the randomized study, we showed a 6.9-month PFS in the treatment arm, 3.1 months in the control arm. That led to a hazard ratio of 0.42 and a P value of less than 0.0001. The large majority of those patients got ipi/nivo. Exactly as we had expected, the PFS landed right at 3 months.
We do anticipate, at least based on published data, the OS in that arm we anticipate will come out likely in that 12-13-month range. As you know, Laura, in the past, at the Society for Melanoma Research in an oral presentation last fall, we presented an OS over 21 months. Obviously a single arm data, but, based on historicals, we feel really good where we are and, obviously, clearly mission accomplished, as it relates to the hazard ratio and P value. The last piece I will mention is around response rate. Single-digit % response rate for the control arm, which again, just shows you how problematic this indication is. We showed a response rate that's approaching 40%.
Remarkably, our response rate went up in the randomized phase III under central review versus investigator scoring, which, you know, which doesn't happen a lot, but in this case it did. We were thrilled to see that.
Absolutely. Can you provide a little bit more granularity on what additional data we should expect at ASCO?
We'll have, I would say, more color on the response rate waterfalls. We did show numerical values, but as you know, hear a lot of people like to look at the water actual waterfalls and how deep were those responses. I think big picture, I think at least our perspective, it's a very clear data set. Very clean, very clear. In addition, we did not talk about investigator scoring response rate and PFS, we'll fill in that box as well. We'll show by central review, by investigator, a full response rate waterfalls, treatment and control arm by central review, as well as investigators. I think it'll be really good in terms of it'll just fill in any gaps there may have been.
I know we're gonna be also looking at some subset analysis on efficacy as well, so I think that will be helpful. Then on safety. You know, we did give a general update on safety, sort of several sentences in a paragraph. We've said in the past, it's been consistent. We, we've reported it in the past. It's consistent with what crizotinib has reported in the past. Obviously, that drug has been approved for over a decade. It's always nice to see those safety tables.
Absolutely. Regarding your registrational path here, you know, you guys are filing off of this PFS endpoint for accelerated approval to be supported for a full approval with OS at a later date. Fair to assume that we shouldn't see any look at OS data at ASCO, that's something that is gonna come in the future?
Laura, we did mention, and thank you for bringing that up. We did mention as part of the top-line results that we have seen a preliminary trend in OS, and that is based on a specific hazard ratio that was observed at the time of the data cut at the end of January. The follow-up has been fairly short, right? It's just over 10 months. Even with that, we're seeing a trend in OS, which is great. The interim OS analysis has now currently projected for middle of next year, you know, we don't anticipate refreshing that as part of ASCO.
Okay. How should we think about registrational path, you know, beyond the HLA-A2 negative population? You know, what's the strategy there?
Yeah. Here I think it's gonna be fairly straightforward. You know, with our tour now, you know, we're gonna get more interaction with the FDA, the ability to submit as per these 3 pre-submission modules from now to, you know, when the full submission will occur. Our focus for that will be around frontline HLA-A2 negative. Laura, as you know, we are generating a large dataset in HLA-A2 positive in a single arm fashion. We noted recently in our public guidance that we're gonna give a fairly sizable, about 100 patients, clinical data update at a major medical conference in the second half. That will include response rate PFS as well as survival data in the frontline and all-comer population. That data will get submitted as part of our NDA.
You know, just so it's clear, this is not base case, but upside is, you know, can we have that dialogue with the NDA and have some form of it on the label. You know, that we think the data is very robust and it's an area of high met need. Base case here for A2 positive will be around, as you know, compendia NCCN guideline strategy. There, I think, a nice part of the strategy that we'll pursue is we're gonna try to pursue it as HLA agnostic for the NCCN guidelines.
Got it.
would enable us to, you know, hopefully provide this therapy to patients that, you know, they don't need to go and get HLA testing. Which we think, puts us at a, you know, a good situation, you know, ultimately if this drug gets approved in, in the commercial setting.
Absolutely. Can you contextualize what the TAM here is in the first line and what the breakdown is of HLA positive versus negative?
Yeah. In terms of total patient pool, now I'm talking about the global population, which is largely driven by the U.S. and Europe. Annual incidents, we believe is in that 4,000-5,000 range. In the pre-metastatic setting it's about double that, 10,000-12,000 patients. We do believe, in the split between HLA-A2 negative and positive, at least based on our clinical screening results. It's roughly two-thirds HLA-A2 negative and one-third HLA-A2 positive. As you know, the competing agent that's commercial has generated, you know, roughly an annual revenue clip of about $430 million.
If you kind of do that simple math, if the majority is negative, and we believe we'll capture part of, positive, you know, you start getting to fairly, you know, interesting and attractive numbers fairly quickly.
That's sort of how we're thinking about it. Obviously, pricing and duration are gonna be key variables. We are seeing similar to tebentafusp patients getting treated beyond progression as well. I think that's also positive. You know, last part I will mention, Laura, because we have an oral tablet also, you know, we do think that, you know, gives us a nice product profile, especially for trying to penetrate that community setting.
Yeah. You know, you just noted that, you know, expanding into the, you know, adjuvant or neoadjuvant could potentially double the TAM here. You guys are running registrational enabling studies there. Can you provide a little detail on where you guys are with those studies?
There are areas that, you know, we think provide really just great opportunity and also, again, highlighting just extraordinarily high unmet need. There's nothing approved in the neoadjuvant setting, nothing approved in the adjuvant setting for uveal melanoma. Also, we did receive breakthrough therapy designation in the neoadjuvant setting as well. I think hopefully establishes FDA's perspective on the data that we've published in the past. The two randomized Phase III studies. Neoadjuvant, I'll kind of give the high level summary. It's broken out into two cohorts. One is enucleation, the second is plaque brachytherapy. For the enucleation cohort, the primary endpoint is eye preservation, and there we want to exceed a lower bound of 10%.
We're well above that right now, so we're seeing a majority of patients eye preservation. I think that one, we feel extremely good about, and that was the data we feel gave us the Breakthrough Therapy Designation.
Yeah.
For the plaque therapy cohort, it's a 15-letter BCVA vision test, and we need to show 20% improvement in the treatment versus control arm, post the treatment and then ultimately plaque procedure. For both cohorts, there's a pooled analysis around no detriment for event-free survival.
That's the neoadjuvant study. That enrollment is ongoing. We're continuing to activate sites. The second study we're just about to launch with Servier. We did have a positive Type B meeting with the FDA and endorsement around the adjuvant study design. Servier will be doing this study alongside with us. That study is another randomized Phase III study. This is going to be randomized against observation, there's nothing in that control arm. This is a classic adjuvant trial, which is superiority for relapse-free survival, about 450 patients. We think that trial is our largest market, right?
Yeah.
10,000-12,000 patients, no HLA restriction. The treatment duration will be 1 year with the daro-crizo combo. As you know, in the adjuvant setting, sometimes patients will take it, right, as long as they can.
Because of, you know, the consequences of metastatic disease. I know we and Servier are really very excited about that study and to get that off the ground.
Absolutely. You know, you guys are set to transition to being a commercial company. As you guys prepare for, you know, this and a launch, you know, what activities do you guys have ongoing? Is there salesforce prep? Are you anything else related to that?
The organization is full force right now, so we're hiring. We just hired a head of VP, head of sales, really across the board. You know, we'll now be going through that process of building the sales organization and preparation. You know, we're continuing to build out as it relates to our readiness on the supply chain and commercial supply chain as well. It's really firing on all cylinders. I think the good news also for us here is because we also have a global partner in Servier, that allows us to really focus the execution to ensure we have a really strong U.S. launch. We feel really good about it. Laura, as you know, it's not a lot of times where companies get to launch in the first-line setting where essentially you're not competing with any other approved agent.
Yeah.
Yeah.
That's for sure. Maybe pivoting more into your, you know, the rest of your pipeline, perhaps starting with DLL3. IDEAYA's gonna have data this year, as well as your Chinese partner is gonna present longer follow-up and durability data as well in the back half of this year. Can you put in perspective for us, you know, how many patients from each data set we should expect to see, and, you know, perhaps maybe the differences in length of follow-up of you guys versus where the Chinese partner is, and how that should reflect in the extent of data we see?
Our partner, Hengrui Pharma, the Chinese data set, it's gonna be sizable. It's gonna be data all at expansion doses. It's gonna be over 100 patients. That will include, majority will be in small cell lung cancer, but they will also show for the first time, data neuroendocrine carcinoma. Here, I think what's, I think the really good news is the data that was presented last year, I think the response rates were impressive. The PFS was very encouraging, but the follow-up was fairly limited. I think when you looked at the censoring, and it's like, okay, where is this all gonna, you know, fall out? I think the good news is now there's more than 1 year of follow-up now that that's been had, and the denominator's increased.
I think people will have really good visibility about what we have. In particular on this question around durability, 'cause as you know, Laura, obviously it's great to have a high confirmed response rate, but you also wanna make sure you have a PFS and hopefully this early preliminary landmark OS data that positions as well. For us in the U.S. and outside of the, in other regions of Asia and Europe, I think we're still seeing how much data we'll have. I suspect it'll be over 30, maybe even over 40 patients, so it will be a sizable data set. I think it's gonna be largely based on how much follow-up we have. What we can tell you is several things so far in our experience in the U.S.
First is the data that we've seen thus far, it looks essentially the same, than what Hengrui Pharma's been observing. That's great. Our data's very, very consistent with what our partner's been observing. If anything, it actually feels like we may be seeing a bit better safety in terms of being able to dose above 2.4 mgs per kg. You may know that Hengrui Pharma's, that's gonna be one of their key expansion doses in their registrational study. We've been dosing above that now. You may know our closest peer here has been capped at a 1.6 mgs per kg dose. Even at 2.4, right, that means we're delivering about 50% more payload.
If we could go even beyond that, the question will be, can we deliver a higher confirmed response rate? Will that translate to a higher PFS and ultimately, OS? The last part I will mention, our view based on the data we have and seen is, we think not only do we have the opportunity to be best in class in the DLL3 Topo ADC area, we feel we have the opportunity to be best in class across all DLL3 modalities. Hopefully this data that we'll share in the second half, will be able to, you know, support our enthusiasm around the profile that we're seeing.
Thank you. I very much so look forward to those catalysts this year. For the sake of time, I could ask a bunch more questions here.
Yeah
you know, as you noted early on, I do know the PRMT5 space very well, and, you know, I am generally very excited about this space. Curious to get your take. You know, you guys are focusing in on an internal combination here. You also referenced perhaps moving into the RAS space as well. You know, we've seen this, you know, very quick shift into a targeted therapy or precision oncology combinations here. Can you provide a little bit more insight how specifically for your asset, you know, you plan to position here and, you know, whether or not you think the shift as a space, you know, has been warranted?
Yeah. Look, you know, Laura, I don't think there is a company that's been doing clinical trial enrollment in MTAP deletion than longer than we have. We have a lot of.
Yeah
experience. We've been, you know, with 2 clinical programs across 2 independent targets. I think here, you know, Laura, our view is the details really matter. Even on the MTA-cooperative side, you know, when you look at sort of the profile assets, they come in different categories, right? Some are brain penetrant, some are not. Some are MTA-cooperative, some are both MTA-cooperative and SAM-cooperative. Which we don't believe is the ideal profile from a therapeutic index perspective, we think ultimately that can have an opportunity to play out in the clinic, especially when you're gonna enable a combination between PRMT5 and MAT2A. That lack of SAM cooperativity is gonna be essential. Also, as you noted around combinations, right? A lot of the focus here could be the, in the area of PDAC.
As we know there with PDAC, there is no benefit to have brain penetrance, right? Our view based on the epigenetic biology of both PRMT5 and MAT2A, we don't believe brain penetrance is the right profile. We think it's more of a liability than a benefit, especially in indications, like PDAC. At least our view is we've seen that data already play out with certain brain penetrant PRMT5 inhibitors based on those AE tables. I know I said a lot there. In our mind, there's really 2 strategies for combination. A, we have a significant focus on PDAC. Our MTA-cooperative PRMT5 escalation's going very well. Based on the early human PK data we've seen, we feel we're gonna have a favorable pill burden size.
We are gonna be in a position to start combination studies here very shortly in a month or so, and we'll start expansion as well, here in the second half. Our two strategies around combinations, one is to really fully suppress this pathway of PRMT5. We think to do that because of certain bypass mechanisms that exist and that we've published on multiple times, the optimal approach is to hit both the PRMT5 and MAT2A target. The indication we think that's gonna be the most important is a non-small cell lung cancer. However, as I mentioned in the beginning, you wanna have a non-SAM cooperative molecule to test that hypothesis. At least that's what we believe in particular, because the key AE to look for there is gonna be around myelosuppression.
Any SAM cooperativity you have, you can exacerbate that, especially in this combination, 'cause what we're trying to do is put this pathway in hyperdrive when we enable that.
Yeah.
For PDAC, here our focus has some similarities, but key differences. We know in PDAC, one of the key alterations you have to really go through and manage is KRAS, right?
There, you know, clear opportunity, combine it with KRAS-related assets, with PRMT5, we think there could be an opportunity for a doublet and a triplet here, which would be our differentiation. As I mentioned, we think we have a best-in-class asset to combine it with KRAS because of lack of brain penetrance for PDAC and this piece that we don't have SAM cooperativity.
Yeah.
That's really, we think, our path forward. Look for us to push on all fronts here. As you know, we have a 3rd MTAP program around CDKN2A, which is completely novel. No one has delivered this, we're getting closer and closer to the clinic now with that because that's now guided for the first half. Just so you know, CDKN2A loss, which would be defined as mutation, deletion or methylation, is present in 70% of PDAC.
Yeah.
Not 40%, 70%. You could imagine doing a doublet with that, right, with a pan-RAS or a pan-KRAS, and now you're covering 70% of PDAC. You know, I was gonna maybe show, shoot up a slide right now to show you, but we have some really exciting data. In pancreatic KRAS models as a monotherapy with our lead molecule, we're showing tumor regressions pre-clinically with this asset. That's now, you know, going into, you know, later stage toxicology studies. I think that's also gonna be an asset we think will differentiate us. We're thinking about this area, hopefully you can see.
Yeah
for the long term.
Are you guys, in any ongoing kind of conversations regarding, you know, maybe drug supply partner, for a RAS inhibitor?
I think that's probably a good way to describe it, but I can't say much more than that, Laura.
So-
You know, I think there's a lot of ways we're thinking about the area of RAS, KRAS. I think our view is there's probably, you know, multiple opportunities there.
wanna be thoughtful about how we enter that area.
Absolutely. I'll stay tuned on that one.
Yeah.
I do have a question for the audience. You know, given how long you guys have been pursuing this target, do you have any thoughts on the MTAP deletion or you know or mutation screening? You know, do you think the current the current binary MTAP positive negative framework is sufficient for patient selection? You know, any learnings that you think that you as a company can add to those phase?
Great question here, Laura. I think the short answer here is yes, we do think MTAP deletion on its own is sufficient as a selection biomarker. I think where the nuances come in is around the elevated MTA, right, in that tumor setting versus the wild-type setting across indications, and are they similar? I think that answer is not as clear. That's why we mentioned to you the MAT2A PRMT5 combo. We think that's most optimally gonna be positioned in lung cancer. When you start going into different indications, that answer changes, right? That's why we think for indications like pancreatic cancer, we could also evaluate that triplet, but that's where we also think there could be value to, you know, specifically target. We know that there's another very key genetic alteration.
that's present.
Yeah.
Like KRAS. Hopefully that's helpful. Yeah, what we can tell you is all of those indications are definitely not created equal. I think we've been fairly, you know, transparent about that perspective. Beyond that, you know, let's say a biomarker with an MTAP, we don't think that's needed. We think it's really about sort of indication focus.
Got it. We're running a little low on time. I do wanna ask 1 question about your KAT6/7.
Sure
You guys have, you know, stressed that this is an equipotent dual inhibitor of both KAT6 and KAT7. You know, biologically, why do you guys think it's important to hit both versus, you know, some of your peers focused on just KAT6?
Yeah, I think, you know, frankly, this question around KAT6/7, Laura, is one that I think people have been aware of for some time, and the importance of also hitting KAT7. I think frankly, the reason why people, we hadn't seen them in the clinic until us and Pfizer, you may know their next gen molecule is a KAT6/7. They just started in the clinic in January, right now it's a 2-horse race there. They just entered the clinic, right? Years and years after KAT6. The reason for that is really simple. The chemistry was extraordinarily difficult. It was very much of a bear project for us, and we were insistent that the profile we needed to achieve, was this dual potency between those two, what we describe as paralogs.
In terms of the biological importance here, Laura, I think what I would highlight, it's sort of similar to what we're doing with MAT2A between PRMT5, or as you know, whether it's in the CDK2 and CDK4 space other paralogs that we're aware of. I would think about them as paralogs, and they have mutual dependency. There's what we've shown and published most recently at AACR, bypass mechanisms that we've identified that no matter how hard you hit KAT6 alone, you're just not gonna get the full activity you can. If you can precisely hit KAT6, KAT7, spare 5 and 8.
We think that can really take this whole area to the next level. Really exciting. I think one of the key questions for us, which, you know, we'll answer here, dose escalation is going well, so we can state that, is can we show strong monotherapy activity without fulvestrant, you know, the SERDs class? I think that will be a key question. If we can, clear differentiation. Because we have the 6/7 profile, we think we can also take this beyond just breast cancer into indications like colorectal cancer. You'll see with Pfizer, they have a significant focus on colorectal cancer, which we believe is because of the underlying biology. The last part I'll mention, Laura, is, you know, clearly this whole ESR1 mutant story in breast cancer and how to deal with that.
We saw the recent AstraZeneca with the FDA around this emergence with ctDNA versus progression. They're also recognizing how important to address this ESR1 mutant is for the SERDs class. Clear opportunity there. We think KAT6/7 would be the superior approach. Second, you may have seen, which was our deraxanrasib pan-RAS combination data with KAT6/7 in KRAS CRC. We're actually seeing quite strong activity in KRAS CRC in combination with deraxanrasib. As you know, in that area, you know, with EGFR combinations, it's the overlapping skin toxicity. Right here, that's not a limiting AE for us. I think that's a clear opportunity as well.
Makes sense. Clearly, the 25 minutes is nowhere near enough time to really get into the weeds as much as we want.
Yeah.
I look forward to future conversations with you, and I, you know, thank you for your time today, for sharing your story, and, you know, very much so looking forward to Catalyst coming later this year.
Definitely. Okay, thank you so much, Laura, and thank you to Stifel again.
Of course. Thank you.
Bye for now.