IDEAYA Biosciences, Inc. (IDYA) Earnings Call Transcript & Summary

Thank you so much to you and Evercore for hosting this fireside chat today with the IDEAYA management team. So IDEAYA Biosciences, we're a leading precision medicine oncology company. We have 9 programs in the clinic. So a very deep and diversified portfolio. Umer, as you know, our most advanced program is Darovasertib, which is now currently in 2 Phase III randomized studies, one in the first-line metastatic uveal melanoma indication and then next in the neoadjuvant indication, we're looking to launch our third Phase III study this half. Obviously, very much in focus is our upcoming guidance around our top line results for the first-line metastatic melanoma setting here. The primary endpoint is median progression-free survival, and I'm sure we'll talk more about that today. But beyond that, we have a very deep pipeline. I would say next is a portfolio of assets in the ADC space. I would say probably most attention is to our DLL3 TOP1 ADC, IDE849. We are guiding towards a clinical data update on that program by the end of the year. And we believe this molecule does have the potential to be a best-in-class asset. As both you, Umer and John know, we did present a very robust single-agent activity as an oral presentation at the World Cancer Lung Conference last September. Beyond that, we also have a deep pipeline in MTAP-deletion. We have 2 clinical assets here, both PRMT5 and MAT2A. We think a lot of interesting opportunities, both as monotherapy as well as combination. Also, as I think both of you have been following, a lot of interesting intersections between RAS and MTAP as well. And then lastly, KAT6/7. So IDE574 has now entered Phase I. We think a really unique opportunity and perhaps one of the largest addressable patient populations, including in breast cancer and colorectal cancer. And Mike Whiter, CSO, is here, and we can definitely dive into the science of that as well. So I think that's hopefully a good summary and encapsulation of the company.

Got it. Okay. Great. Fantastic. So maybe just as we begin, could you just remind us sort of the timing and the expectation for this upcoming readout?

Yes. So the timing is going to be in the next couple of weeks. So we've now guided on or around the very end of March. So we're obviously going towards March now. So the data update is in the near term. In terms of sort of table setting on expectations, so Umer here, our perspective is what we believe we'll see in the control arm is going to be consistent with what's been published in large meta-analysis before. And I know there's been some questions about Ipi/Nivo in the control arm. And I think here, probably the most, at least we feel reliable paper there is the [indiscernible] paper, which was a multicenter study. And there, we believe the PFS is going to come in roughly 2 to 3 months. In the treatment arm, we reported a 7-month PFS in 2 oral presentations, one at ESMO several years ago and then most recently at the Society of Melanoma Research this past fall. And that PFS has been solid, approximately 7 months, including with roughly 2 years of follow-up. So we feel good about that number. And if we -- if that's what we see, we feel good about our prospects to hit [ STA sig ] as our primary endpoint.

Got it. Maybe I ask before we get too far off of what the trial design is, one of the things that I know we've spoken about in the past, and I am continually curious about is given you have an investigator's choice comp arm, Ipi/Nivo probably the most relevant thing that people are asking about, but there are other options. You've got a combination on top of Crizotinib, which is not available in the compound. Can you give us some of your thoughts about what the potential variability is in the [ comp arm ], but what the contribution of parts is in the experimental arm and how we should be balancing the expectations there?

Darrin, do you want to take that?

Well, in terms of the contributions of the 2 components in the Daro/Crizo, that's pretty clear. I mean Darovasertib, we know what it does as a single agent. It's active. We've shown that in Phase I. Crizotinib, however, has limited activity, at least the MET inhibitors in this space. They've been tested before in small numbers in clinical trials, really haven't shown much in the way of activity. However, it's very clear, at least preclinically, when you put the 2 together and you impact both pathways, protein kinase C and c-MET, you can see enhanced antitumor activity, and we've shown that pretty conclusively in the clinic, obviously, with the Daro/Crizo combination relative to what either agent can deliver alone. And also, we've discussed this with the FDA. So contribution of components is pretty clear. We know that you need the MET inhibitor to enhance the activity in the metastatic setting. We also know in the neoadjuvant setting, you don't need the Crizotinib. It doesn't seem to add. We already have profound activity in the neoadjuvant setting with Darovasertib as a single agent. So our thoughts are we don't need the Crizo in that setting, but however, in the metastatic setting, different story. With respect to your other question about the control arm, to be quite honest, there's been at least 2 meta-analysis published, thousands of patients. And I think we feel very, very comfortable with where the control arm sits, whether it's Ipi/Nivo, whether it's ipilimumab, whether it's a single-agent checkpoint inhibitor like nivo or pembro, they all sit right around 2 or 3 months. And so with our long-term follow-up from the 01 trial, 2 years of follow-up where we maintain a rock-solid PFS of 7 months, and we only need to hit maybe 5.5 in order to be -- have a positive study. We think we have a great buffer zone. And again, I think we're just as well positioned as you possibly could be to beat the control arm in this particular situation. And our OS looks profound, too. I mean that PFS translates into another metric, which is OS. We have 21 months of overall survival in the 01 trial. And so those 2 things go hand in hand. If you combine that with the fact that we're tripling the response rate compared to standard of care, we have a really rock-solid duration of response of 9 months, we just think we couldn't be better positioned to take on standard of care in this case.

Well, since you brought up OS, Darrin that was going to be my next question.

John, can we -- maybe just one second, can we just like set the trial? I want to go through all the data very carefully, but maybe let's just set the trial. So can you remind us the design of the trial, just so we're all on the same page?

Sure. So what we're talking about now is metastatic uveal melanoma. So patients with uveal melanoma, who have metastasized. They also have to be HLA-A2 negative, which is probably about 2/3 of the patients with uveal melanoma. And so it's a simple randomized Phase III comparison of either standard of care, which in this case is either single-agent checkpoint inhibitors, a combination of checkpoint inhibitors plus other immunotherapy like Ipi/Nivo or you could, if you happen to be a patient who couldn't get immunotherapy for whatever reason, you could get chemotherapy, but we think that's going to be a very small proportion of the patients treated. And then of course, in the treatment arm on the other side, is the Darovasertib, Crizotinib combination, which we've tested extensively at the 300, 200-milligram dosing. And again, this is a study, where the primary endpoint is progression-free survival. So that will allow us to have a discussion with the FDA. And then, of course, that study will go on to read out for overall survival for full approval.

And what has happened so far in the trial? Has there been any interim looks so far?

No. So this is what's coming up now, our top line to look at our first look at progression-free survival.

How mature do you expect OS to be at the time of PFS maturity?

Very good question. So I think the maturity is one of those things where when we talk about the outcome for this particular study, some of the things that we're likely to talk about is, obviously, the response in the 2 arms, the progression-free survival in the 2 arms, of course, maybe some durability with respect to duration of response. But survival at this stage may be relatively immature. So if we don't say anything about it, it's not because it doesn't look good. It's because there may not be enough data to actually make a call on whether there's any trend, for example. However, there will be opportunities down the road. We'll be obviously filing the NDA. We'll have the opportunity to once again provide the FDA safety information and perhaps efficacy information during the filing period. And then, of course, our first interim analysis for OS comes up at the beginning of next year or the first half of next year. So there will be a couple of opportunities to look at OS. And it's very possible that during that period of time, we'll have enough events -- but at the first look, it's uncertain at this time, if that will be the case to really make a call.

If I -- sorry, John, just before we move on from the trial design, there's a Phase II here and there's a Phase III here. Could you just remind us, have you guys spoken to any dynamics around the Phase II data yet?

No. So at this point, again, this will be the first time that people have seen what we have evaluated with respect to progression-free survival as well as safety comparing the control to the treatment arms. We did have -- this was a sort of a 3-component study. We had a IIa part, where we actually looked at dose optimization to show that the 300, 200-milligram dose going forward was the optimal dose, had the best benefit risk profile for patients. And then the IIb/III is the part that we'll be reading out now, which now accumulates patients to be able to have enough data to show that the progression-free survival, how it compares between the 2 arms, and then we'll follow that up with overall survival. So 3 component study, Part 2a completed, getting IIb/III getting ready to read out and then...

How many patients were in Part IIa?

So Part IIa was designed to look at the 2 doses and without getting into dramatic detail, essentially, it was a 2-stage component. So it depends on how you define it. But long and short of it, it was approximately 80 or 90 subjects to be able to look -- take that first look. Remember, they were randomized 2:1. So 2 to the control arm -- 2 to the treatment arm, 1 to the control arm. So it took about 90 patients to figure out what was the right dose. We had additional patients that came on after that as we were continuing to enroll that were part of the IIa, but they weren't part of the initial analyses. But essentially, that was around that 90 patients or so is all that was required to determine what was the optimal dose and then you move forward from there.

So the reason I was...

And Darrin those 90 patients, those were both A2 positive and negative. Those...

No, no. So this study is only HLA-A2 negative. So if you...

Is it pivotal. But you did have a Phase II that had all HLA subjects.

Yes, that's a different component, and we'll probably get to that later. But that is a patient population that we haven't forgotten about that's critically important to us. We think this treatment should work there, no different than it does in HLA-A2 negatives. However, we're evaluating that patient population in our 01 trial. We'll have nearly 100 patients' worth of data, who are HLA-A2 positive. The objective would be the following. As this study reads out, our view is that it will show an improvement in progression-free survival, ultimately an improvement in overall survival. And in parallel, we'll be able to present data to the FDA as well as the NCCN panel that puts together the guidelines for treatment that shows in 100 patients worth of HLA-A2 positive data from a Phase II trial setting that we have a similar overall response, progression-free survival, overall survival, durability of response that we do from our randomized Phase III trial. And hopefully, that will allow people to look favorably at that data set and utilize it as appropriate.

There's no mechanistic reason to expect your drug to behave any differently in these 2 populations?

Absolutely correct, right? It's all about the mutations that drive the disease to begin with. So these tumors, 95% of the time have mutations in GNAQ, GNA11, G-coupled protein receptors that drive protein kinase C signaling, and that's why Darovasertib is so effective in basically knocking the wheels out of the driver for the disease.

Yes. HLA only, but HLA-A2 only matters because of KIMMTRAK already approved in that...

Exactly.

So maybe just on the trial design again because I just want to make sure -- I want to go through all those things, but just on trial design, there's a Phase II and there's a Phase III. And maybe let me just put up something that -- the reason I'm just trying to go step by step on this. So here's something I was looking at just for my own sort of understanding of the trial. We know there's 440 patients that came in. And you said 80 of them were in Phase II. So that's kind of like consistent with.

That's not what I said exactly. So what I said was...

Maybe I could just step in -- so the number of patients, just to be really exact, the IIa portion is 124 patients.

124 patients.

124 patients, yes. And remember, the dose that's not moved forward gets dropped irregardless, whether we included the IIa in the analysis or not. So it turns out that roughly 75 patients were in the non-move forward dose would have been dropped irregardless. So that leaves you...

Right. So 360 -- sorry, go ahead. Are we get into Phase III?

Yes. So that -- what that leaves you is with 49 patients with the move forward dose as well as the control arm. So that was the question, should those 49 get included as part of the analysis or not. The decision ultimately that was made was that we were going to utilize that IIa for the Project Optimus dose optimization portion at the end. The IIb and III now are the same. It's -- because it's an ITT analysis and because enrollment was ahead of schedule, when we submit -- when we do the analysis for the PFS, the full enrollment for the OS has also now been complete. So the IIb and III are 313 patients. So that's how you get to that 437.

So Phase IIB and Phase III is 313 patients you said?

Total, yes. And that will be for both the PFS as well as OS analysis.

So 313 is the number we're going. Okay. Got it. That makes a lot of sense. And the reason -- that's why I was like -- when John was sort of asking this question, the reason I kept coming back to the trial design is I just wanted to be sure that I understood, PFS -- because the way OS is being analyzed here, and I want to get into the relevance and significance of that in a second as well. But this option 1, 2, 3 for OS analysis depends on what happened on PFS, right? And what I wasn't sure about was it kept saying PFS off of Phase II, like kept saying PFS and Phase II. It's not Phase II. It's PFS across the whole trial, Phase II and Phase III, all 313 patients, correct?

Correct. Yes, yes.

So the sample size of 200 PARG 2.55 hazard ratio. So this is actually 313. So you have more wiggle room technically.

That's correct.

I see. And...

And you also have a subset of these patients that are substantially more mature than the bulk Phase III, the patients that were enrolled later on in the study.

That's correct.

And Yujiro, is it fair to say that the interim is still being triggered by 130 events?

That's correct. It's going to be the first 130 PFS events that will drive that analysis. And as you have there, the OS is 253 OS events.

So interim equals 130 events out of 313 patients.

Exactly.

I mean the trial was designed. The IIa was for optimization, the IIb was for accelerated approval and the III was all for OS. But remember, as Yujiro pointed out, the IIb and the III collapsed together because the enrollment went so quickly. And the additional patients that he mentioned in the IIa are simply because we overenrolled the IIa to make sure we had enough patients to do the analysis.

Does that mean that your 130, 140 PFS events you need for the final read, almost half of those are coming from patients in the go-forward dose in the IIa that are presumably much more advanced. Is that fair?

No, no, no. IIa has only contributed 49 patients. IIb and III is really the trial.

If that 49 patients, presumably most of which -- most of whom have progressed much well advanced of the other patients, that's almost half.

No, no ITT, the way I understood it was a Part IIb and Part III. Am I mistaken there?

Yes, that's right.

You've got it correct. It's the IIb/III that drives both the PFS and the OS analysis.

When did Part IIb start recruiting?

So we noted in December of 2024 that the IDMC selected the move forward dose. So I would assume it's roughly -- I don't know, roughly a month before.

I see, so over the course of '25, basically, you guys recruited it in about 14 months and 2:1 randomization. So 100 patients should all have had -- so I'll tell you what I'm really getting at. Of the 313, you've basically got -- if I can have my thing here, you've basically got 100 versus 200 type of randomization going on. All these 100 patients were presumably recruited over the course of 2025. And by now, almost everybody should have had an event and almost feels like this PFS -- the 130 events could have hit even sooner? Or has it already happened? I don't know if -- maybe it's happened and you guys are not unblinded to it yet.

Yes. So we have said, Umer, as part of this most recent earnings that the 130 PFS events has occurred with BICR. So we're now in the data cleaning process.

Okay. So that would be consistent. The reason Yujiro, you can imagine the reason I was going down that path was if the 130 has not been hit, yes, it sounds like control arm might be outperforming by a little bit, but that's not -- presumably, that would have been...

Or the treatment arm is doing better.

Or treatment. No, Darrin, I was just going by...

With 100 patients in the control arm, the treatment arm would have to be doing really spectacularly to not hit events at all.

Right.

The other reason that I wanted to...

Hang on, John, sorry, hold that thought for a second. Darrin, let's just go with what you just said a little further. Wouldn't that also mean that if the active arm was, let's say, tracking more like a 5-month and compared to at 3 months, these interim would have happened back in like October, November. So we can also rule out the scenario for a more narrow PFS benefit? Or am I going to an extreme there?

Well, listen, I don't want to speculate except to say I know that we've been rock solid in 7 months. So I'd be shocked for a way off on that, but the data is going to speak for itself.

Do you disagree with my logic there?

So repeat that one more time, Umer.

My point was if we know the comparator arm tracking at 3 months or so, and most of these 100 patients would have had a PFS event by late fall last year, let's say, 75 out of them or 80 out of them. In a scenario where the active arm was tracking at a slightly lower-than-expected PFS, let's just call it 5 months for a second, we would have seen perhaps a pretty meaningful number of PFS events in the active arm too, to the point where this interim would have gotten triggered by September, October last year, if the active arm was tracking a little lower than previously. So the fact that this interim did not hit in September, October, doesn't that effectively become a validation for your prior efficacy got replicated?

I mean I think that's a logical conclusion, Umer. I mean, look, I think the key part here is what you mentioned earlier is that it's a 2:1 randomization, right, from the treatment to the control arm. So that control arm has got to be pretty wide range to the right, if that's what's really driving the analysis, right? If it was flipped 2:1 from control to treatment arm. But again, it's 2 to the treatment, 1 to the control arm. So I think what you're saying, just high level makes sense to me. Just factual point because the question around when did the IIb/III portion start. I just was looking at my notes here. So in that release, when we noted the IIb portion started once the IDMC picked the move forward dose, we mentioned that 124 patients in the IIa portion, either way, 75 were going to get dropped because it's not -- that wasn't the move forward dose. We noted in that release, there were 185 patients enrolled. So to your question, even at that time, end of '24, right, we were already roughly 60-some-odd patients into the IIb/III enrollment.

There were 60 patients already enrolled.

In the IIb/III, yes. So in the December '24 release, we noted that there were 185 patients already enrolled. We talked about before was in the IIa. So that by default, that means have to be 61 or more patients were enrolled by the end of 2024.

And Yujiro remind me when did you guys say interim has happened? You guys never said that. You definitely never said that at our conference in December.

That's correct. We did not. So...

It came in '26, whenever it did.

Yes. We clarified that the PFS events were not achieved through the end of last year. End of last year, we still had not achieved.

Yes. I think -- okay. That's very, very helpful. That's very, very helpful. And you will preserve -- and you -- I mean, my understanding is there is an expectation for you to hit OS in this setting. Would you agree with that?

Well, we wouldn't have done the study without it.

Yes. I mean I think all we can see the Umer is that we presented OS data at SMR, right, and we showed roughly over a 21-month OS number. So we feel good about that. It's obviously a single-arm study, relatively small denominator, but we think it's an encouraging signal.

So is it reasonable to assume that this is where you guys want to be in terms of the OS analysis, like the PFS comes in, I don't know, 0.55 or better and then you preserve a lot of alpha for OS?

That's correct.

And if you think about it, I mean, most people would argue that the therapies that are out there for HLA-A2 negatives right now are relatively ineffective, right? So you're talking about a therapy that doesn't do a whole heck of a lot, except for a few rare patients. And you're talking about a therapy that has triple the response rate, double at least the progression-free survival. Overall survival looks amazing. We should hit on overall survival. But you have to do the study. That's what we're doing.

One last thing...

The reason you're asking about OS earlier is about timing on that. So you're going to file with the agency on the basis of PFS data, but you're going to show them the OS data. Presumably, there probably multiple opportunities to the agency to see that. Is there a chance that there's enough follow-up on OS to get to full approval before people expect or on the same time line as the initial submission?

I'm glad you brought that up because there's one thing here you have to know is that there is no crossover permitted until the OS reads out. And so I think that is very important that we need to look at this data and just be aware that patients are going to be looking for options. For HLA-A2-negative disease, there really aren't any good therapies. And so if you're on the control arm and you've got an ineffective immunotherapy, it would be really important to be able to offer what's going on in the treatment arm, if possible. But we can't do that until we cross a threshold where everyone feels comfortable that we're looking at a PFS that's robust. We're looking at an OS that looks like there's a significant trend that it can't -- not unlikely to be reversed. But again, that will be an FDA call, and we'll give them every opportunity to look at the data in order to make that assessment along with the IDMC.

How long will it take -- sorry, John, were you asking the same question? There was an investor question.

Maybe -- no, no, I was also going to ask about the state of the Phase IIa data.

So yes, maybe let's just ask the investor question just before we move on to that. The question was, how long would it take to get the independent review of PFS done? And if you could remind us whether this is investigator-assessed or blinded?

So we're -- this is not a blinded study. However, we have blind independent central review because of that. And so that's part of the analysis. So they are obviously doing the imaging analysis to evaluate whether patients progress. We'll be looking at both investigator-assessed progression as well as independently assessed.

And which is primary?

Blind independent central review.

Okay. Blind independent central review. Okay.

Well I was going to say, well it would take you a few weeks, you said. Is that how long it would take?

Well, it's all in the process of doing that now, right? We're in the process of evaluating -- getting the data into the database, cleaning the data, producing queries. Ultimately, once the database is locked, then basically the data between the blind independent central review and the EDC has to be merged, analyzed statistically, and that's when you'll get the output. And that's what Yujiro was referring to when we're talking about targeting the end of March because that's kind of what's going on at the moment.

Makes sense. Sorry, go ahead, John.

No, I was going to ask about the Phase IIa, which obviously is in hand. And admittedly, there's 75 of those patients that are now on the go-forward dose. But do you have plans to submit that as part of the same package to the agency? Have they seen that data already? When can we expect to see that data as well?

The agency will see everything because what they want to know is what the safety is on every patient that was dosed.

But they haven't seen it yet. They didn't see it as part of the...

But the independent data monitoring committee has seen it, and they're the ones that evaluated it, said it was still -- the appropriate dose to go forward with was the 300, 200. They felt the safety profile was adequate to go forward. And so we should let the study read out, and that's what's happening in the next month.

Will you show that to investors ahead of time as well? I mean the Phase IIa data, I mean, it's -- presumably, it's already mature.

Yes. But remember, it's got much shorter -- it's similar data than you've seen with the 01 trial, right? And so -- and the 01 trial actually has 2 years of follow-up. So it's probably more of the same, except shorter follow-up.

Fair enough.

Okay. John, there were a few questions -- more questions from investors on these. Do you want to go through those first? Or do you want to go through your side first, either way?

Sure. One of the other questions that came that just came up was on -- more on that Phase III expectation side. On the safety side, do we expect to see safety comparable versus Phase I observations? And then what are the specific criteria for dose reduction and what the protocols are in the Phase III part of the trial there?

Well, I'd say we've done a really good job. Investigators learned with 01 how to manage the adverse events associated with the Daro/Crizo combination. So we have a group of people that are familiar with its use. And so I think you'll see a safety profile that looks relatively similar to what you've seen before. I'm pleased to say that because investigators, there's 2 things. People are very motivated to take the therapy, as you might imagine, when you have a disease, where there's no good therapies and the only good one is the one you're on. So they really want to stay on the treatment and investigators are getting better at managing the adverse events associated with it. Because of that, I think you'll see at the end of this trial is a really relatively low dose discontinuation rate due to adverse events. It's going to be more about keeping the patients on therapy. We know that's important. And the KOLs that we've worked with are getting really quite expert in how to manage that. So I think we'll be in good shape there.

What about discontinuation -- not discontinuation, what about dose reductions? I would expect to see in later-stage trials as docs get more familiar with the drug, you see fewer discontinuations. But do you have an expectation for the number of patients who are going to down dose?

We'll have to see about that. But I would say there's a few things that are going on here. And the reason why I can't give you a number. 1, obviously, we're talking about a trial that only a few people can see the raw data to; and 2, talking about experience I've had with 01. But there's 2 things that people do. And again, this is usually due to Grade 3 toxicity, which isn't terribly common, right? But usually, they do 1 or 2 things. They usually do a very short pause. We're being told by some that a 1- or 2-day pause is sometimes all it takes to get rid of some of the toxicity. The other thing they do is they typically lower -- if they're going to lower anything, they don't lower the Crizo, they don't lower the Daro. So they'll lower the Crizo dose slightly. Again, I think with those 2 moves, they're able to keep a motivated patient on therapy for most of the time. And by the way, we have some data accumulating now that also implies that the toxicity happens mostly in the first 2 months of therapy and thereafter kind of weans itself off. So it gets more easily tolerated, it seems, in later time points, and we'll be sharing some of that data as this data evolves. But I think those are what I could say about managing the toxicity. Thus far, the people that are running the trials are getting quite good at it. It's manageable. The dose reductions and discontinuation rates hasn't been high in 01. I don't think it will be any different here. And I think patients just want to stay on the study because the therapy is just so good compared to what the options are.

Makes sense. Umer unless you've got more on 02, should we.

Yes. Maybe just -- and I do want to sort of over the -- in the next 10 minutes or so bring it back to the big picture and some of the other pipeline as well. But could you just remind us for the prior data? I was trying to find it, John, do you have the prior data because I couldn't find it in our folder. I was trying to understand maybe if you could remind us what percentage of patients had liver mets? What percentage of patients had LDH disease? I just find that as we often try to translate from prior data to the new trials, sometimes we can lose track of the baseline. So I just want to make sure I had it for my purposes.

We have to remember with uveal melanoma, virtually everybody has liver metastases, right?

So I was looking at these...

That's where this disease goes. So it's actually to have extra hepatic disease only is pretty uncommon. If you have both, that's even worse.

So Darrin, I was looking at a couple of Nivo/Ipi references like the GEM-1402 trial, it says 80% liver mets and 50% had non-extra liver mets. So I was going to ask.

That's in addition to liver mets, mostly.

Right. Sorry. So you're saying everybody had liver mets in your trial.

Well, the vast majority of patients will have liver mets with this disease because this is where -- that's where the disease metastasizes to. Additionally, some will have disease beyond the liver. And that's why things like liver-directed therapies are so big, right? Because people think, well, I can go after the liver and I can get a lot of it. But eventually, if you go after the liver too much, it will end up somewhere else if you live long enough.

Yes. So Umer, I've got the numbers here. So our frontline study that we reported at SMR.

The prior data.

Yes, the prior data, over 90% of patients were either hepatic only or had hepatic and extrahepatic. So only about 5% had non-liver. The [indiscernible] Phase III study, very similar. So roughly 95% had either hepatic extrahepatic and about 4%, so very similar to what we reported had non-liver disease. In terms of greater than upper limit or normal on LDH status, both our study were roughly 1/3 of patients and the [indiscernible] randomized Phase III are very similar. It was roughly -- so that fine from their New England Journal paper. I'm not looking at that.

Super helpful, it's fairly balanced.

Before we move on, the one other thing that I want to make sure we talk about on 02 is when you -- how long will it take you to prep the filing after the top line?

So just so you know, sort of we're going with sort of average metrics here, right? So we're talking about probably around that 6-month time period in order to put the package together. And then, of course, the FDA has to review it. That's probably another 6 months. So that's sort of general guidepost, but we're obviously going to try to do everything we can to expedite the process for sure.

But that puts us to the first half, that's how we get to that first half '27 expected. Make sense, okay. Okay. On 01 and the A2 positive patients, can you talk about when you'll have a breakout of that and how rapidly NCCN guidelines could proceed after a potential approval next year?

So I can sort of give you a high level of the entire package since we're probably running a little limited on time, but there are multiple things we want to do. We obviously want to submit this, get it to the FDA, get it to the NCCN panel. In parallel, what we'd like to do is get the HLA-A2 positive data out, that 100 patients, get it published so that the timing lines up, where we'll be able to, again, say something about the overall response rate, the progression-free survival, the overall survival, talk to them about the biology, how it's no different when you're going after protein kinase C. And therefore, they'll have a couple of pieces of data. They'll have a randomized Phase III. They'll have a single-arm Phase II both relatively good size. They will show them that the data is very similar. And then the NCCN panel can decide what they want to do that. We suspect that they'll overwhelmingly welcome this treatment into both category of patients very robustly. In addition, we're going to talk to the FDA about the opportunity to potentially get the HLA-A2-positive patients in the label. Could a real-world data set help? Are there other things that we could do in order to make that happen? That's a discussion that has to happen. And then in addition, in parallel to all that is the neoadjuvant trial, the 09 study. We would like to get that published in parallel. Again, we've presented that data recently at ESMO, really exciting tumor shrinkage in the majority of patients, saving more than half of the eyes, improving vision, not only during the neoadjuvant period, but potentially long term based on a visual prognostication score. So it would be great, if they also had that body of data for published in their hands so that they can decide what they want to do with that as our studies continue to evolve.

Now on that neoadjuvant side, because that, as you say, very potentially very exciting. The data you showed at ESMO was excellent, but it's also 5 or 6x the size patient population-wise versus the metastatic setting. So can we talk a little bit about patient journey into the neoadjuvant or adjuvant into the surgical setting, when patients get diagnosed and how physicians are choosing the initiation versus brachytherapy and where Daro would live?

So just from a very high level. So a patient walks in and says, I have trouble seeing optometrist looks in their eye, they say, "Geez, we see something there. I don't like it. I'm going to refer you to an ocular oncologist." They do a workup, a clinical workup and they make the determination that the patient has uveal melanoma. From there, in general, the movement is pretty rapid within weeks, the patient can be set up for either a nucleation, which is the removal of the eye, and those are tumors that are usually large and too large in order to be controlled by radiation or if the tumor is smaller and they can show a plaque to the eye, which contains radiation, they can do that. And both of these things are typically happen within weeks of the diagnosis. And again, the problems that come downstream with that is obviously, if you lose your eye, you lose your vision. And then even if you get plaque brachytherapy, you may have saved your eye, but you may eventually lose your vision due to the radiation doses. And also there are toxicities that happen downstream of radiation, radiation, retinitis, toxicity, et cetera. And so what we hope to see is in the future, someday, a patient will come in and get the diagnosis and instead of being sent off to having something sold to their eye or having a surgical procedure that removes their eye, it will be as simple as, you know what, I'm going to write you a prescription. You're going to take pills for up to 6 months, and then we're going to decide what we need to do. So it will be completely different, paradigm changing for this disease. And now we get prescription and you don't have to go directly into your primary local therapy.

But is that being done in the trial, though, Darrin?

Yes, exactly. That's exactly what's being done in the trial.

So like how it has 1/4 of the patients are in nucleation and the other 3/4 are not. Is that -- is the active arm not removing the eye in the nucleation subset?

So remember, it's almost like 2 studies in 1. So if we focus on the nucleation group. So these are the guys with tumors, they're going to get their eye removed. Basically, they get randomized, either you go straightly to get your eye removed or if you're randomized to the treatment arm, you get 6 months of neoadjuvant therapy. We try to save your eye. If we do what we've done in 09, we're going to save more than half of them, and they can transfer to a less invasive or less surgically extreme kind of treatment.

So you keep the eye, even after that surgery in this group then.

No, no you don't have the surgery. You don't get the surgery, and that's in the control arm, go straight to a nucleation or you don't have the surgery, you get neoadjuvant therapy instead. And if all goes well, you get maybe a plaque place there instead. Obviously, there are some patients that will have to have the surgery because perhaps the tumor didn't shrink.

But the off-ramp from nucleation is radiotherapy. And then in the radiotherapy cohort where there maybe the tumor is small enough where nucleation wouldn't be the primary treatment option.

You're mixing the cohorts again.

That's the second trial.

Just -- so focus on. There's only 2 options. You're either going to have your eye removed or you're going to get neoadjuvant therapy and then hopefully, something else will happen, right? Like they'll get a plaque placed or they'll get proton beam or external beam radiation or something. But...

But that is getting radiation. So you're moving from...

Only in the treatment arm if the eye is saved. Only in the treatment arm if the eye is saved. The enucleation group just gets their eye removed. That's the enucleation cohort.

So Darren, let's just -- sorry, before we go to the non-enucleation cohort, I guess this is a question both for you and Yujiro. Unlike a typical neoadjuvant study, because we're talking eye preservation here, could this form the basis of a very rapid filing in much more than the first-line setting?

So the only limitation is, remember, another one of the important endpoints for the study is event-free survival. So we have to show that we're not creating any detriment to delaying the primary local therapy. So with that has to read out. But the fact that we'll have known about saving the eyes very early, as soon as we hit the EFS events, you're right, we could actually take that part of the study and go talk to the FDA.

And defined by disease progression?

What's that?

Yes, it's relapse. So yes, Umer also, I would just highlight, as you know, typically, a randomized Phase III neoadjuvant study, right, you have to show superiority on EFS. Here, all that FDA has asked us is a no detriment threshold, which is much lower. And because of that, typically for an EFS readout on superiority, you got to wait for most of those events to occur, right? Here, we think with about 1/3 or 40%, we'll be able to establish no detriment. So it could be just generally, as Darrin said, especially if we try to submit that enucleation portion first with the no detriment that's pulled across both cohorts for EFS, I think it would be faster than your typical neoadjuvant study, right?

But [indiscernible] is not sufficient.

Correct. Yes. We want to see no harm to EFS, which at least based on our single-arm data in neoadjuvant, we don't see evidence of that at this point. So...

And the definitions must be pretty well established for what -- who counts as an enucleation eligible patient?

Yes. Correct.

Exactly.

It's all about size. When your eyes -- because the radiation can only penetrate so far. So if the tumor is too big, you have to get enucleated.

Sorry, John, I guess maybe I'm curious how you guys think about it. So 6 months of -- so -- the comparator arm will have 100% enucleation right away. How does the active arm get established versus comparator arm? You just have to go all 6 months to be able to do that? Or could you just do it within the first 3 months?

So once you have a tumor that requires enucleation, you get randomized. One group will go to get their surgery. The other group will get the neoadjuvant therapy.

Right. But how long are they supposed to stay on there to establish that, there's a meaningful differentiation between the 2 arms?

I'm not sure I understand you how long are they staying on the therapy?

My point is let's say we took 120 patients, 60 got immediate enucleation. The other 60 now are on the drug. They didn't get it enucleated for now, but some of them are starting to. I guess, how much into that process on that other sort of -- on that active arm, are you able to say, okay, we're now starting to see that there's a 50% reduction in enucleation because that number itself would keep fading, right? It might start at 100 and it's fading.

We typically know in about 4 to 6 months for that patient...

This is what's happening. They're getting a scan every month, ultrasound every month, and they're getting treated for 6 months. And as soon as the tumor shows the maximum reduction in growth, that's when you take them to your primary...

Which is to say as soon as it starts to grow again, you sort of take them off neoadjuvant and go to the -- either a new vision or radiotherapy?

That happens, yes. That happens in the minority of patients, but most usually, they drop down, they shrink and then they kind of level off at the end, and that's when you take them.

And there's no Ipi/Nivo used in neoadjuvant?

No. I mean remember if you're going to be in a neoadjuvant setting, you have to shrink the tumor. These treatments completely tumor.

Right.

You remember, it's not -- obviously, the drug has got to get to the eye, right? So I mean, I think we're fortunate that clearly, we're having an effect there, but I don't think that's a given. And as Darrin mentioned, IO therapies, here, the objective is to shrink the tumor. And we know that...

IOs get used in metastatic settings, presumably because they may be doing something against the metastases, but they're not going to do a ton in the eye.

Exactly.

Exactly. So -- and Yujiro, wouldn't this recruit like right away?

I mean the Phase I/II single arm was extremely robust once we got the sites up. So that's our expectation. I mean I know the team is also...

There's a tremendous amount of inclusion.

Like with less than 6 months to recruit. So I guess let me just push it to an extreme. You started the trial in early '26. You could be fully recruited by ASCO presumably because people are getting their eyes removed, they will probably say, let me try this out, presumably, it's only 120 patients. And you track them over 6 months, you could hit all the events by December, January. Is that too extreme a scenario?

So for the full trial, we're guiding towards first half of next year, but that includes the plaque cohort. So could the enucleation enroll first and we get that readout first? Yes, that's a possibility, I suppose. But as noted by Darrin earlier, we still need to wait for the pooled analysis for the EFS, right?

Okay. And also -- sorry, my last one on this. There was -- I remember Bristol ran the study back when John and I were covering sort of all the Ipi/Nivo trial readouts. There was a trial called NADINA, I think, in Stage III neoadjuvant. I thought that had become standard of care. Is that not the case? Like you give Ipi/Nivo before resection. I realize it may not have much activity in uveal setting, but in general, isn't everybody is supposed to be getting Ipi/Nivo?

Yes, you're talking about skin melanoma maybe...

Different disease.

Different. That's why this disease is so unique, it's a...

So there were no uveal patients in those trials. Is that right?

Basically, this is a treatment that doesn't work -- the biology is completely different. Tumor mutation burden in this disease, low. RAS mutations, RAS mutation is low, if not at all. So it's just a different disease. It just happens to be -- it's a melanoma, of course, but they're driven -- ours is driven by a different set of mutations. It's wired differently.

I just confirmed.

Why do we give immunotherapy -- because we don't have anything else.

No, they excluded uveal. You're right. I just checked as well. Okay. Great. Great. Wow, so this neoadjuvant, usually, I see neoadjuvant, I was like, okay, that's like 5 years away, but that's not the case here.

We're hoping we can do faster than that.

Okay. Great. So in the last 10 minutes then, I want to do 2 things. A, I want to make sure we take care of any investor questions there are. But also just ahead of that, could you just sort of lay out for us sort of the size of commercial opportunity as well as sort of the highest conviction programs beyond this because I feel like there's other things happening, which we usually catch up on.

Sure. Josh, do you want to take the commercial market piece?

Yes, sure. So on the uveal melanoma side, so in the metastatic setting, we think that's between 4,000 and 5,000 patients incident globally. And that's really U.S., Europe, Australia. That's where those patients are concentrated.

And what's U.S.?

U.S. is roughly 1,500 a year.

U.S. is 1,500. Okay. And you're thinking of pricing as orphan?

So I mean, look, we're a big proponents of value-based pricing. It's going to come down to the sort of benefit that the therapies can offer. I think one benchmark people point to is obviously the [indiscernible] price, which is approaching $90,000 a month. So we'll see what our data looks like, but that's one benchmark out there to consider.

Okay. And then, Yujiro, from your perspective, this takes the company to profitability, this launch in metastatic alone?

So we haven't put out that specific guidance on profitability, Umer, but obviously, some of it will also be driven by our spend. We have other registrational studies across other programs we're looking to launch. So I think some of it will be depending on how we ungate a lot of upcoming studies. But yes, look, we do think depending on the trajectory of that launch, Obviously, we have other readouts, including neoadjuvant that will be trailing. I think that's going to be part of the conversation as we move forward, absolutely.

Makes sense. And if you could remind us the cadence of readouts beyond this program for some of the rest of the pipeline?

Yes, sure. So I would say the next one probably most in focus is our [ DLL3 TOP1 ADC ]. Their enrollment in the U.S. and outside of Asia is starting to pick up. So we are targeting a clinical data update for that program for monotherapy by the end of the year. And at least based on the monotherapy activity that we and our partner are seeing, we do hope there is some monotherapy approval path that's going to be available to us. In addition, we have current guidance around an MTAP clinical data update here with Trodelvy. We'll be pushing as hard as we can. Also, you may know we have a PRMT5 inhibitor that's now in Phase I. And so we'll see how that develops as well. So I would say those are several of the pieces. And I know our bispecific ADC is also in Phase I and that we anticipate that enrollment should also go quick. So we'll be closely monitoring the progress of that. But I would say that's probably another possibility as well.

One thing you didn't just mention that I want to make sure we touch on something we talked about at JPMorgan that really fascinates me every time we talk about it is the PARG, which I know is a potential combination agent for any of the Topo-ADCs. You said you're going to start those combinations with the DLL3 this year. Given how rapidly late-stage patients in small cell progress, fair to say that we could see those initial combination cohorts early in the next year time line, and we'll be able to look at them side-by-side with the DLL3 monotherapy data?

Yes. I mean we're planning -- Darrin, you should comment here, but we're planning to dose that combination here very shortly in about a month or so. So I think we should be in good shape on that one, John. And we've done a lot of work on PARG dose optimization. At least my personal sense is I think we'll get a sense of where we are pretty quickly, right, on that combination. So we could have data fairly rapidly on that. And I know the team is, I would say, very enthusiastic about that.

Even ahead of next year, could we have it...

Yes, it's possible. I think it's going to depend on how much follow-up we want and kind of the expectation around that. But there's obviously a durability piece, but there's also just around just pure response rate and do we...

PD-1, PARG as a mechanism with Topo is that something that we could get ahead of durability?

That's the idea. Yes. That's the idea. I mean that could be a differentiating factor for us really to put us above the rest. And it could apply to other ADCs as well. I mean that could be a unifying message that if you have a PARG inhibitor along with an appropriate ADC good target that carries a Topo payload, this could make a difference. That's -- we hope that will be the first test, but it won't be the last in DLL3.

Makes sense. Okay. So last 2, a couple of investor questions. One, expected timing of filing for accelerated approval after the March readout?

Darrin, do you want to hit that? We haven't given specific...

Yes. We kind of basically spoke to that before, right, about 6 months to put the data together and then 6 months to -- so something that's kind of where we're sitting.

And that 6 months is presumably also to get some OS looks. Is that right, Darrin?

Yes, yes, absolutely. I mean we'll share as much as we can. Again, we'll get some of that at the top line, but it may not be very mature. We'll probably get more of that during the filing period. And then, of course, the first interim will read out the first half of next year.

So you don't need to wait for first interim?

We will be providing follow-up data with the FDA. They're going to want to see follow-up safety data during -- once you initially file, they're going to ask for 120-day follow-up information. And so that could be part of the package potentially.

You give them what you have presumably at that point.

So the idea would be you file by August, September. And then the 120-day update allows you to get that first OS interim in there. Okay. Makes sense.

I wouldn't call it a first interim, I would just call it an evaluation, the data could be available for the FDA to review at that time. The official interim won't happen until the first half of next year.

Per stats plan deal.

That's exactly. Per the statistical analysis plan, right? Exactly.

Okay. The last program I think we didn't touch on is KAT6/7, which obviously is getting started in Phase I imminently. We saw Pfizer make their announcement on the dual inhibitor. Can you give us the 3 -- the 2-minute pitch on where your version of this inhibitor stacks up relative to some of the others we've seen in the clinic?

Sure. Mike do you want to take that?

Yes, absolutely delighted to talk about that program. This is potentially very meaningful for patients. It's part of our central strategy to address tumor heterogeneity. What we have found, what we've shown is you want dual potency on KAT6 and KAT7 in order to be able to intercept these tumor genic transcription factors that are present in breast cancer, lung cancer, colorectal cancer. Our asset with respect to the clinic is the only one that I'm aware of that has that profile. We have not profiled Pfizer's asset, but we do know that the other assets that are out there in the KAT6A/KAT6B space, we think do not have the capability to deliver that profile on KAT7. And that is critical because KAT 7 can substitute for KAT6A, KAT6B and the regulation of chromatin. And that's also critical because we and others have shown that KAT7 brings on a very meaningful biology that will intercept both intrinsic and acquired resistance, and those are 2 of the big conundrums facing precision medicine. So KAT7 inhibition together with KAT6 that can kill tumor-resistant cells that are drug-tolerant persistor cells, therapy-resistant drug-tolerant persistence. So that's an intrinsic heterogeneity mechanism. And then also KAT7 is required for the self-renewal of tumor-initiating progenitor-type tumor cells, these so-called cancer stem cells that are a wellspring of acquired resistance. So a very exciting program, large patient populations available to us based on intercepting lineage-specific transcription factor activity, as I noted, breast cancer, lung cancer, colorectal cancer.

Mike, the one thing that I want to make sure we tease out of what you just said is the dual potency against KAT6 and KAT7. We've seen numbers from Pfizer's program from some of the others that suggest that you get good nanomolar or single-digit nanomolar potency against 6A, 6B and then maybe you could have in the order of 1.5 to 2 orders of magnitude before KAT7 and then 3 orders of magnitude before you're getting to KAT5 and KAT8 where you'll have the toxicity. So when you say you want to hit both at the same time, in your opinion, is getting 10x or 50x more selectivity for 6A, 6B, is that like enough hitting 7? Or do you really need to be right on top as your curves look like?

I think that's a great question, and it's really important for people to be able to actually deeply evaluate the mechanism of action here. So what we have seen is you want dual potency, you want low nanomolar potency on KAT6A/B and 7. I have seen a lot of data, our own data, other assets, data. When people are looking at a 10- to 50-fold window for biochemical activity, what we have seen is that translates to a 200-fold window or more, sometimes as much as a 1,000-fold window with respect to cellular target occupancy. Biochemical activity does not take into account the multivalent complexes that are controlling these enzymes. -- and are making a big contribution to your ability to cover that enzyme in cells and compete with everything else that's going on in that setting. So super important to have a clean line of sight on your PD to understand what your target coverage needs to be. And we have shown empirically that this matters a ton. If you're looking at your preclinical models, PDX is coming from patients that have progressed, for example, on Ibrance fulvestrant, you cannot cover these with the clinical KAT6 inhibitors. You can't even cover them in combination with palbo and fulvestrant, but we can hit them with the KAT6/7 dual. So very, very clean, very, very robust demonstration of the exposure profiles that you need when you have a dual potent KAT6/7 that you can't get if you don't have that low nanomolar potency in that very small window between the 2. And you do have to stay away from KAT5 and KAT8, as you noted.

Yes. Excellent. Well, I think we're unfortunately out of time. I feel like we could get -- Mike, I'd love to have just you on the next webinar, so we can really get nerdy on the biochemistry there. Yes. Fabulous. Obviously, a lot to talk about. We spent most of our time on Daro today, but there's a whole ADC platform, PARG, MTAP and KAD6/7 to talk about as well. So thank you guys so much for being on with us. We really appreciate the time this afternoon. Thank you.

Thank you for having us.

Thanks so much for the time today. Bye for now.

Bye-bye.