Management of IDEAYA Biosciences with us today, just getting settled in. Thanks so much for joining us, Yuji. Hi, guys.
Hi.
So let's start with a brief introduction, I think. Yuji, just in your own words, a couple of minutes here, can you tell us what you've been working on and what you're most excited about heading into 2026?
Yeah, first, great to see you again, and thanks so much to Evercore ISI for hosting our fireside chat again this year. So at IDEAYA Biosciences, we're leading precision medicine oncology. Our lead program, Darovasertib, is in our first-line registrational study in metastatic uveal melanoma. I would say probably the greatest focus point right now is our randomized median progression-free survival result that's upcoming to enable our first potential accelerated approval filing in the U.S. The public guidance for that readout is year-end, which we're obviously very close to approaching year-end, so potentially weeks away to Q1 of next year. In addition, our full enrollment for the median OS component, that enrollment should also be done here very soon. So we have one patient remaining. The target enrollment there was just north of 450 patients, so that's going to be done as well here shortly.
In addition, as you know, we have several additional registrational trials planned for this program. We have a phase 3 randomized neoadjuvant trial that's now up off the ground. Sites have been activated. Patients are in screening, so that's great. We did receive breakthrough therapy designation for that indication earlier in the spring.
Oh, that's great.
Correct. And then lastly, with our new Pfizer, we've also recently announced that we're going to be launching a phase three adjuvant randomized trial the first half of next year. So everything there, we're firing on all cylinders. And then I would say maybe next is around what we're doing on the DLL3 topo ADC. We had a presentation at the World Conference on Lung Cancer fairly recently in small cell lung cancer. In addition, as you know, we have several clinical assets in the MTAP arena, both our phase two MAT2A inhibitor. We have a phase one PRMT5 inhibitor. And we did recently announce we have a third MTAP program that's targeting a major co-alteration called CDKN2A of MTAP. That candidate nomination is coming up here shortly.
A lot to talk about.
Yeah.
So let's start with Darovasertib, since obviously that is the near-term focus. We've talked a bunch about the different potential paths here, the different populations here. But could you remind us more broadly in the entire uveal melanoma, what's the patient journey from diagnosis through later stage treatment, and where do you have the most opportunity to bend the needle for folks?
Yeah, definitely. So first, let's start with that patient journey, which would start from neoadjuvant adjuvant to the metastatic state. Unfortunately, there are no systemic approved therapies in neoadjuvant, no approved systemic therapies in adjuvant. And we believe the majority of the patient population in the metastatic setting also don't have approved therapies available to them.
You're talking about the HLA-A2 negative?
Correct versus positive. Yeah, that's correct. And then in terms of the patient journey, typically the way this happens is a patient may have some visual disturbance in their eye. They go get their eye examined, and they learn that they have ocular melanoma. From that point, the treatment journey, usually the choices that the patient undergoes is either if the tumor is sufficient size, they will get eye resection or what's called enucleation. The majority of patients will then get what's called plaque therapy as an alternative to that, which are these. It's essentially a metal disc that has radiation plaques to try to reduce the ocular tumor size. Some smaller proportion could also get proton beam radiation. After the primary procedure, which is typically one of these surgical procedures, then they'll go into the adjuvant phase postoperative. There, again, there are no available therapies, unfortunately.
And then typically within several years, at least the high metastatic risk population will get metastatic disease within sort of that two- to three-year time frame.
Now, your data has meaningful PFS and OS results already in that metastatic setting, looking good. But one thing that's always interested me is the potential for vision sparing and reduction of both enucleation rates and reduction of the use of plaque radiotherapy in the earlier stage setting, where patients actually might have the opportunity to save some vision. So can you talk to us about what the relevant bars are for success from a patient experience perspective there, and maybe what you would hope to see in the upcoming data?
Yeah, so when you think about our objectives in the pre-metastatic setting for neoadjuvant specifically, our goal there is to save eyes, preserve vision, and then through the adjuvant trial to ultimately hopefully prevent relapse and hopefully save lives there. In terms of what's meaningful in the neoadjuvant, so for the enucleation cohort, the primary endpoint that we got agreement with the FDA is eye preservation. Ideally, we can exceed a 10% eye preservation rate with a 95% confidence interval. Right now, based on a fairly large data set we presented at ESMO a few months ago, the majority of patients were preserving their eyes. So there we are multiples, we believe.
Multiple, yeah.
Yeah, beyond where we feel we need to be. And that was really what we feel was a key driver that got us the breakthrough therapy designation from the FDA based on that data set. For the plaque therapy cohort, it's a visual acuity readout, and it's a 15-letter BCVA vision test. So the way we've powered this study, ideally we see 20% or more patients in the treatment arm versus the control arm pass that 15-letter BCVA test. Both cohorts, we also have a requirement where we have to hit a secondary endpoint of no detriment to event-free survival.
Now, in the radiotherapy cohort, the vision deficit there is being driven by disease, but also primarily by the radiation therapy itself causing damage, is that right?
That's correct. Yeah, so the vision reduction we know does occur as these ocular tumors grow, and so as they get closer or start impeding critical eye structures, we know these patients are losing their vision due to the ocular tumor, and that was, I would say, the significance of the data that we demonstrated during neoadjuvant treatment. We were seeing a mean 14-letter vision gain if you pull those two cohorts together pre that procedure.
Pre-radio.
Yeah, exactly. So that's obviously a good thing because better vision you have going into plaque therapy, you would presume better vision you have coming out. The second, as you mentioned, is it is believed that the radiation is ultimately what's driving vision loss in these patients. So in the control arm, the vast majority of those patients we believe will be legally blind in roughly three years.
That's quite an outcome to bend the needle on for patients. Do you have data on retreatment at all since you're looking at both the metastatic setting and earlier perisurgical settings?
So in terms of if they flow from, let's say, neoadjuvant to adjuvant or neoadjuvant to metastatic.
Yeah, so the eventual relapse cycle.
Yeah, we don't have that data at this point yet.
Okay. In the interest of time, we've got so many programs.
Sure.
Let's move on to DLL3.
Sure.
I was really excited to see some of the initial ADC data you presented earlier this year. It's a great-looking waterfall. It seems like you show good responses across a wide number of doses. You suggested at the time that pushing dose might improve durability, but it doesn't seem to have much of an impact on ORR, at least in the data set that we showed, that we saw. So how do you approach the balance between safety, durability, and what are you looking for as you continue to see follow-up on those patients?
Yeah, no, that's a great question there. So yeah, one of the expansion doses that we're evaluating is 2.4 mg per kg, weekly infusion, once every three weeks, and at that dose, we saw a confirmed response rate in the second line small cell lung cancer setting of 70%. Across the entire dose escalation, we saw a PFS of approximately 6.7 months, so I think based on all measures, we're seeing very robust activity. With ADCs, as you know here, what is critical is around dose intensity, so because think about it similar to chemo, once you start losing any dose intensity, you could essentially start losing efficacy, so there were earlier doses below 2.4. There were low doses that we cleared and then continued on to 2.4.
I think we probably have the ability to go a bit higher if we choose to, but we also feel very good about the 2.4 dose. At that dose, we saw grade three or higher AEs that were less than 20%, which is why we believe we could probably still continue to push the dose.
More maintaining intensity.
Yes, exactly. And really there, the main thing we would look at is the intensity as well as durability.
What were the rates of discontinuations or step-downs or dose interruptions?
So they were fairly infrequent, I would say, in the single-digit % range, maybe low teens % in that area in terms of SAEs as well. So I think right in there is what you would expect for a topo ADC.
Yeah, and given that you've got multiple doses where you seem to have efficacy, it does seem like you have some flexibility to push.
Correct.
And retreat if needed on an individual patient. So as you move into larger cohorts, we've discussed the potential to start seeing post-DLL3 patients or patients who were previously exposed to some other mechanism. What do we know about loss of antigen or mechanisms of resistance to the other DLL3s?
Yeah, no, it's an interesting question. The important aspect to highlight here with DLL3, it's connected to a key transcription factor, survival oncogene called ASCL1. So that directly impacts the expression of DLL3. That's why we don't believe a resistance mechanism is going to be around downregulation of DLL3, primarily because it's going to impact its own survival. So there we do feel confident you can go something like a T-cell engager DLL3 followed by a DLL3 topo ADC. We've been also evaluating the binding epitopes of an agent like Imdelltra versus our DLL3 topo ADC. And we do believe those binding epitopes are distinct, so we could even think about it as a combination.
And even if there are binding epitope-related mechanisms of resistance, they wouldn't hit you.
Correct, yeah.
Great. You've seen data both small cell that we've just been talking about. Obviously, that's been the focus for DLL3, but you also saw some pretty interesting data in NETs. Do you have plans to develop both of these paths in parallel, and when can we start thinking about seeing single indication cohorts?
Yeah, so I would say the primary focus here is small cell lung cancer, neuroendocrine tumors. We believe there are other DLL3 upregulated tumors, such as melanoma as well, that will be of interest.
Excellent. What should we expect to see for this in 2026? I think you gave a very brief overview in your initial remarks, but can you walk us through how big the data sets could be next year and what we should be looking for?
Yeah, I'm going to tag team here with Josh on some of the '26 items. Yep.
Yeah, so we've initiated our U.S. phase one trial for DLL3. Obviously, as you said, the focus is on small cell and NETs. I think we would hope to collaborate with Hungary in providing an update on the phase one data set that we shared at World Lung with more patients and more follow-up. And then additionally, we have about 20 to 25 patients' worth of data in NETs as it is from Hungary. So we'd like to share that data as well. Again, I think these are the questions we're getting is sort of what are we seeing and where are we taking this? And again, that'll be a big focus as well beyond the data is really trying to lay out what our clinical development strategy is for the DLL3 asset.
Right, well, the initial data was very exciting, so I think we're all looking forward to seeing more and more. Let's move on to MAT2A and the rest of the MTAP pathway stuff. Expecting upcoming data there in both UM and non-small cell. Can you remind us what the path is there and what you're going to show next?
Yeah, so for MTAP deletion, there are several strategies we're taking. So as you know, IDE397, MAT2A in phase two, our PRMT5 inhibitor, that's now in phase one. We're also doing a TROP2 combination with Gilead in MTAP deletion, both urothelial and lung cancer. As you mentioned, our guidance currently is we'll have a clinical data update first half of next year. We're targeting a major medical conference for the urothelial cancer update with Gilead. We'll likely focus the second half for lung cancer just to give us more time to enroll and have more data maturity. In addition, we'll be focusing on several other strategies here, including the MAT2A-PRMT5 combo, likely in the lung cancer setting as well.
Yeah, we definitely have to talk about that. For UC, you showed 50% to 7% ORR. It's a very nice response rate. What sort of follow-up are you going to be able to show in the first half since we thought they'd have been durability?
Yeah, so we're starting to see durability. We're seeing multiple patients not just get confirmation scan, but I would say multiple scans beyond that as well. So I think here I would hope we'll have roughly about six months of follow-up at that point. And I would hope in urothelial cancer, ideally we have 20, 30 some odd patients at an expansion dose, something like that.
Reasonable, but maybe not mature PFS data, but we should have some initial looks at durability response.
That's fair, yeah.
Okay, great. And in non-small cell, you mentioned wanting to wait for maybe a second half opportunity to get longer follow-up. When you say that, I'm just doing some mental math. Is it possible you could have nine or 12 months of follow-up by the time you share that data?
For Non-Small Cell Lung Cancer, yeah, if it's towards the end of the year, we'll likely pick a forum in the fall. We know there are several key conferences then. So yes, I think that will be possible. We've already enrolled several patients there, and I know we'll be pushing that enrollment as hard as we can.
Excellent, excellent. And maybe then returning to that combo that you mentioned in phase one, obviously PRMT5 combos, I think make a lot of sense with a couple of different agents in this pathway. Can you talk to us a little bit about where the PRMT5 is currently in the clinic and what should we expect to see from initial combo looks?
Yeah, so we're in phase one. Sites have been activated. Screening has begun, and we'd like to be able to get that into combination studies with IDE397, MAT2A as quickly as we can, so based on our clinical protocol that was okayed by the FDA, we're going to be able to start combination dosing with MAT2A by the time we're in the second cohort. In that second cohort, we do believe we're going to be at an efficacious dose for that combination. So hopefully we'll even have data. We'll have hopefully an opportunity to share some data even as soon as next year.
Matt, that is really directly aiming at a combo opportunity there. That's wonderful.
Correct. Yeah, for lung cancer, that's our primary focus right now.
When we think about the identity of the combo, MAT2A from a biological perspective makes sense. It's not the only combo for PRMT5 that people have floated. Can you talk a little bit about the competitive landscape for those combinations and what makes sense there?
Yeah, we think the other sort of area of focus for companies, including for IDEAYA, is around targeting co-alterations. So probably the two most relevant ones here are RAS, and this is specifically in PDAC. And we think the other one is CDKN2A. So we are looking at several different options around RAS. In addition, around CDKN2A, I think as we mentioned earlier, we now have a candidate that's going to get selected here soon that would enable us to do a wholly owned combination pursuing this MTAP CDKN2A.
Another wholly owned combination.
Correct, exactly, which would allow us to pursue indications like PDAC as well as others pursuing that strategy.
All right, I have three more programs on my list and only three minutes to do them. Let's see how fast we can go. You started to talk a little bit about PARG, IDE161 in combo with the ADCs as well. What's the opportunity there? And can you walk us through what we should be expecting from that biologically?
Yeah, the central focus for us, PARG, we believe is a key DNA damage repair mechanism that, specifically based on its biology, has the opportunity to amplify the durability of topoisomerase payloads specifically.
Regardless of antigen.
Correct, regardless of antigen. So we're quite excited about this. We'll be implementing that with DLL3. We'll be implementing that with our new bispecific that's in phase 1. We're also having several conversations on this to enable these combinations for others as well. So stay tuned on that. And we think we'll be a key driver over the next couple of quarters.
We just saw the IND for B7-H3, PTK7, bispecific ADC. B7-H3 is a target that we followed in a variety of forms for several years. Interesting activity in a variety of different tumor types. Can you tell us a little bit more about your asset as distinct from some of the leaders in the space that we've seen data from already?
Yeah, look, we think this is clearly, A, it's a first-in-class asset, and we believe clearly differentiated from the mono ADCs of either B7-H3 or PTK7. Second, we know that the co-expression of those two antigens are very high in certain tumor types of high priority to IDEAYA, including non-small cell lung cancer, CRC, head and neck, as well as ovarian cancer. And based on our preclinical data, we believe we're going to see greater efficacy in those co-expression populations. So we will implement a strategy around looking at co-expression as a selection tool to really magnify the effect in that target population. So I think we fully anticipate we should be seeing, we think, robust monotherapy activity here as well as an opportunity to combine with PARG, IDE161.
All right, so that's two questions on that then. Biologically, what's the rate of co-expression? What's the size of that patient population compared to the mono expressing populations? Yeah, let's start there.
Yeah, so based on protein expression databases, we believe in lung cancer, it's going to be about a third of those patients. So substantial in colorectal cancer, we believe about 40%-50% of those patients. So it's quite a large population, which is why we actually got quite focused on this bispecific specifically versus other bispecific. So we had looked at a lot of different bispecific opportunities, and this one was quite compelling for us based on both the data we saw as well as this co-expression profile.
So it does sound like a more limited population than the monospecific antibodies, as you might expect. It's not 100% overlap between these two antigens, but hopefully the efficacy will be sufficiently differentiated to make it worthwhile.
As well as a therapeutic window.
Exactly. So can you talk a little bit about the relative binding affinities here and how you've optimized the molecule? What are the physicochemical, biophysical properties of the antibody?
Yeah, we haven't disclosed too much of that. We will be planning to do much more disclosure on that specific question in the first half of next year. So I know abstracts have already been submitted, but what we would just note is these are very high affinity binding for both the B7-H3 and the PTK7 component.
But the hope is to specifically target cells that have dual positivity.
Exactly.
Rather than see off, or rather than see activity on mono-positive cells.
Exactly, yeah. And that's really this question around bispecific as their AND or OR format. So in this case, this bispecific was specifically designed to be an AND format, needs to express both for that binding to and internalization occurrence.
All right, well, we are out of time. I think we got to two ends and three. But in the last segment, last words here, what is the most important thing maybe in the next six months that we should be focused on? We've covered a lot of different topics, but what is the number one message that you'd like to leave with?
Yeah, I think first and foremost for us, it's the randomized PFS results we're coming into here very soon, and hopefully if that's successful, it starts to transition us towards a commercial stage company.
Fabulous. All right, well, we look forward to seeing that and to following up with you as you get closer to commercialization.
Great, thanks so much.
Thanks for your time.
Appreciate the time, yeah.