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

All right. Great. Hopefully, everyone is connected. Well, I'm Yigal Nochomovitz, biotech analyst here at Citi in New York City. It is my pleasure to introduce the next iteration in our virtual C-suite fireside chat series with Yujiro Hata, who, of course, needs no introduction, the CEO of IDEAYA. If you guys have questions for Yujiro during the course of the conversation, please e-mail me, [email protected], and I will hopefully see those and relay them over to Yujiro.

So Yujiro, welcome. Thank you. A lot has been happening. You had a great R&D Day at the -- right after Labor Day. A lot of activity at several medical meetings as we well know. So maybe it would be helpful if you could just kind of review maybe some of those key milestone markers with the pipeline. And then we'll move into a discussion across all of the programs, but just that would be a good place to start, I think.

Yes. So first, Yigal, thank you so much for hosting this fireside chat, and thank you to Citi as well. So Yigal, as you mentioned, this past quarter, it probably has been one of our most productive quarters that we've had since the company's founding. As you noted, we had several updates at major medical conferences, including at the World Cancer Lung Conference for DLL3 in Barcelona. We also had an oral presentation at ESMO for darovasertib in neoadjuvant. We also had the Society for Melanoma Research oral presentation at Amsterdam, where we reported our first survival results. And as you know, our lead program is getting closer to our objective around the randomized PFS results to hopefully enable our first accelerated approval filing in the U.S., which right now is guided from year-end. So we're getting about a month to December to Q1. So we are getting closer and closer to that. Enrollment is going great. We're almost finished with full enrollment. So that study is also in the process of getting wound down. So I would say we're probably going to be done with full enrollment for the OS, randomized OS as well portion in about a week or so. Beyond that, we're firing on all cylinders on -- which is continuing to be a growing pipeline, DLL3. We did dose our first patient in the U.S. We're ramping up sites globally to accelerate that. Obviously, a big focus in small cell neuroendocrine tumors as well. And MTAP as well here with our partner, Gilead, on Trodelvy. We picked an expansion dose recently in urothelial cancer. We've dosed several patients already in lung cancer, and that enrollment has been extremely robust. And just as a reminder, as part of our earnings, we did note a confirmed 57% response rate in the MTAP urothelial cancer setting. Beyond that, our PRMT5 asset is now in Phase I. We got IND clearance. We also filed an IND on our B7H3/PTK7 bispecific ADC as well. So a lot happening and we expect hopefully a very catalyst-rich next several quarters.

Okay. That's a good start. Well, maybe if we could just kind of go big picture first and go through what you sort of see as the opportunity across the uveal melanoma landscape, because as you pointed out, we're looking at metastatic, neoadjuvant, adjuvant. And so if you could kind of just outline the scope of those opportunities and how crowded or not crowded each of those are?

Sure. Yes. One of the areas that we're very focused on is to ensure that darovasertib has the opportunity to be the standard of care across the uveal melanoma patient journey. So Yigal, as we just talked about, we're getting close to hopefully our randomized PFS results for frontline metastatic uveal melanoma. As you know, we launched a second registrational study in neoadjuvant uveal melanoma. And now we're guiding with our partner, Servier, to launch a Phase III randomized study in adjuvant in the first half of next year. In terms of metastatic annual incidents, we believe this is about 4,000 to 5,000 patients. Prevalence would be higher than that. We're seeking approval in the A2 negative population where there is nothing approved. So we would be a frontline treatment in the scenario we make it to market. We also believe the [ agent ] is going to be equally active in the A2 positive setting. And we do believe we will capture market share in that segment. For neoadjuvant, that number of annual incidents goes up substantially from 4,000 to 5,000 to at least double that. So we think that's more in the 10,000 to 12,000 range. Here, average treatment duration in neoadjuvant roughly in that 6-month range. And then in adjuvant same thing, 10,000 to 12,000 patients. There, treatment duration, we're anticipating -- or we're discussing with Servier is a 12-month treatment duration. And in the adjuvant setting, we anticipate we'll pursue that with our partner as a daro-crizo combination as well. And also in adjuvant, there's nothing approved, as you know, Yigal. So here, we would anticipate that we'll be randomizing against placebo in the adjuvant setting.

Okay. So I think it would be helpful if you can kind of walk through the neoadjuvant data that you showed at ESMO because there were 2 cohorts. Obviously, the enucleation and the brachytherapy. So what were the key conclusions in each of those? And did they meet or exceed your expectations as far as the endpoints that you're aiming for there?

Yes. So maybe just to table set here, Yigal. We received breakthrough therapy designation from the FDA in the spring based on a good portion of the data that we just shared at ESMO, in particular the enucleation cohort. We shared about 95 patients worth of data. Roughly 60% of that was from the enucleation cohort, the remainder in the plaque therapy. In terms of eye preservation rate, what's in our current Phase III clinical protocol, which we had alignment with the FDA was we needed to exceed -- or we need to exceed a 10% eye preservation rate with a 95% confidence interval to receive a full approval. So that's our primary endpoint for the enucleation cohort. What we demonstrated is the majority of patients, we are preserving their eyes and getting them off of the enucleation track. So we are obviously well above that 10% threshold for full approval. For the plaque therapy cohort, the primary endpoint here is around visual acuity, which is a 15-letter BVCA vision test. Here, I would say several pieces of information that were new. First is that we are seeing very -- what we think is a very meaningful vision improvement during neoadjuvant treatment before the plaque therapy. So Yigal, you may recall, at our R&D Day, we noted a median letter improvement during neoadjuvant treatment of 7 letters at R&D Day. And at ESMO with a larger data set, more follow-up. If you pull both of those 2 cohorts, we saw on average, roughly a 14-letter vision gain. So that's a fairly significant jump from 7 to 14 as we went into more patients with more follow-up. And then lastly, based on the level of consistent and robust ocular tumor shrinkage we're seeing, at least based on visual radiation prediction tools and visual prediction tools, we believe the amount of predicted vision loss is going to be substantially less in the treatment arm versus the control arm. So that's the setup. As you know, there's also a no detriment requirement of event-free survival, which is going to be a pooled analysis across both cohorts. There's no specific hazard ratio we need to hit. All we've said with the FDA is that they need to be overlapping confidence intervals that we're not doing harm to the patient. And we have publicly noted, Yigal, here that so far, we've seen very few events in the single-arm Phase II study. So we feel good about that no detriment to EFS requirement as well.

Okay. And then given the 10% level that you mentioned with regard to eye preservation and then the 10% to 17% range on the improvement in letters. How are your -- the Phase IIIs, the pivotal structured such that you're well positioned to be better than those or at least in that range for the vision and I presume much better on the eye preservation? Just -- if you could just review the cushion you have on the Phase III to get to those -- that type of efficacy.

Yes. So on the eye preservation one, Yigal, I think that one is pretty straightforward, right? Where it's a 10% threshold, 95% confidence interval, which means you want to ideally be at a 15% eye preservation rate or higher. Here, we're already multiples north of that. So I think that one should be relatively straightforward. For the plaque therapy, so the way this is going to work is the treatment versus the control arm. We wanted to be able to demonstrate 20% of patients or greater are passing that 15-letter BVCA vision test in the treatment arm versus the control arm. So that's what we need to show to hit our primary endpoint. So the reason why we feel good about this endpoint is because, as I just noted, Yigal, during the neoadjuvant treatment, the treatment arm patients should have better vision because during neoadjuvant treatment, we're seeing this vision gain that's occurring, which in the control arm, that shouldn't occur. And then second, the patients of the treatment arm should have much smaller ocular tumors because daro is shrinking those tumors and therefore, getting less radiation to the critical eye structures and therefore, less vision loss. So now you have 2 stacking events, right, that should hopefully push it in our favor to be able to hit that primary endpoint.

Okay. And are you -- I mean visual acuity is obviously kind of a gold standard measure. Are you looking at anything else in terms of position of the tumor in the eye or anything related to center versus noncenter lesions looking at retinal imaging, for example? Or is that all secondary?

Yes, those are secondary. So the primary endpoint is around this BVCA 15-letter vision test. We have other secondary endpoints, including response, prevention of macular edema as well as others. So -- but the primary endpoint is this 15-letter BVCA test.

Okay. And then you also reported that apparently a certain percentage of patients had gained over 5 letters consecutively at consecutive visits. Why is that important for people to understand as opposed to not consecutive or over a longer period of time?

Yes. Great question, Yigal. So I know you also track the ocular space. So at least from what we understand, the consecutive visit readout is actually quite important in that -- in the setting of just sort of a visual improvement that it's not just at one time point, but through multiple consecutive visits, they're seeing the sort of consistent maintained visual improvement. So I think that data also improved since our R&D Day. So we're excited to see that as well.

Okay. And just again, on neoadjuvant, the time lines are for data. When is the -- what is the latest statement on that?

Yes. So the way -- yes. So you may have seen that the plaque therapy cohort, we reduced the number of patients we need to enroll from 400 to 330, and that was based on FDA feedback around the statistical plan, and we didn't need to split the alpha between the 2 cohorts. So that's great for us. It makes trial size smaller, which means we can enroll it faster. So right now, what we've said is that we -- our target at least is to try to enroll this study in roughly 5 quarters. The first data we'll likely see will be the eye preservation data from the enucleation cohort. The second data set would most likely be the no detriment to EFS. And because of that no detriment threshold, we believe with about 35% to 40% of the events, we'll know that answer. And then the last one is the visual acuity results from the plaque therapy cohort. And that first analysis will be approximately 18 months from the last patient enrolled. So that's the sequence.

And when did this clock start on these 5 quarters?

So we are right now in that process of -- the study has just launched. So essentially right around now.

Okay. All right. Got it. So then usually, when companies talk about changes to power, usually the press release says that they had to raise things, right? In this case, interestingly, it could go in the other direction. So was that -- clearly, you had the buy-in from FDA and you did it. I just -- was that controversial or not? I mean you could have kept it, I suppose, and had more of a cushion, but it sounds like in this case, it simply was not necessary and not a wise use of R&D capital.

Yes, exactly. No, look, I think the FDA collaboration has been great. We're obviously very grateful for them to provide us that feedback that we didn't need to split the alpha between those 2 endpoints. And yes, sort of simple statistics, not needing to split that alpha enabled us to reduce the patient enrollment number.

Okay. So then -- so that's neoadjuvant. And then adjuvant, anything else you want to -- you briefly mentioned the design there. Can you just go into a little bit more detail perhaps on the timelines? And then...

Yes. So here, Yigal, it's going to likely be in that 400 to 450 patient enrollment range. We want to ensure that this study has sufficient power. Second, with our partner, Servier, we did discuss, do we do it as a company sponsor versus a cooperative group? We decided we're going to be able to move a lot faster to do it as a company-sponsored study with Servier. Second, our partner will be side-by-side with us in terms of to ensure rapid enrollment for this trial. Next is that the patient population will be HLA-A2 agnostic. So we're not going to have any kind of HLA requirement here. In terms of the metastatic risk population, we haven't made the final decision there, but it's most likely going to be the high risk and the high medium risk group. So that's the setup. And in terms of the control arm, as you know, Yigal, there's nothing approved here. So we believe the control arm will be essentially placebo. This will likely be a superiority endpoint around relapse-free survival.

Okay. So then sort of in the marketplace, if people get daro neoadjuvant, I assume they would not be eligible in the adjuvant to get it again? Or like how does that whole logic work?

Yes. So once commercialized -- so one is, as you know, for the neoadjuvant, it's monotherapy. There will be likely some clear out period, which there typically is anyway between neoadjuvant, you have the primary procedure before you start adjuvant care. And for adjuvant, it would be as a combination. So for the actual trial, I would assume here that we will not let patients that were on the neoadjuvant portion flow into the adjuvant, especially on the control arm side, as you can imagine.

Okay. Makes sense. And then -- so of course, the phase -- well, let's talk then about the metastatic MUM, which we're getting very close, as you pointed out. So I think everyone is very familiar, but just to remind everyone, maybe what is the expectation for control and active for PFS? You mentioned your very close to finishing on the OS enrollment. So can you just talk about when that might potentially be available and whether there be any commentary on OS towards the end of the year? Or that would be a future update?

Yes. So for the PFS here, at least we feel based on historical data, the control arm is going to come in somewhere between 2 to 3 months. We feel the treatment arm, as we reported, most recently at ESMO -- sorry, at SMR as well as ESMO earlier, it was a 7-month PFS. I would hope that it's going to land in that 7- to 8-month range, which would put us in a solid position to hit stat sig for accelerated approval. In terms of OS, Yigal, as you know, there, we feel the historical data in treatment-naive has been in that 12- to 13-month range. At SMR, we reported a 21-plus month OS. So I think as long as we're in that range or obviously above, we should be in good shape for the full approval OS as well.

Okay. And at this point, are you -- I mean it's November 6. So are you kind of getting -- are you in the database lock portion of this right now? Or are you still -- that is yet to occur?

That has yet to occur. So the events are continuing. So yes, that's still ongoing in terms of the collection of the event. So the database lock has not occurred.

Okay. And the data from Society of Melanoma Research, which was a very good number, the 2021 and change, how does that -- in terms of thinking about how that would work in translating to the Phase III of those patient populations very similar that, that's a good translatability. Can you make that statement with high confidence that, that over 20 months is a good a good proxy for what could occur in the Phase III?

Yes. Look, it's a frontline patient population. I think, Yigal, if I were to sort of highlight one patient baseline characteristics where maybe we might see some improvement was around the ECOG performance status. So as you may recall, from the SMR data we presented, the ECOG performance 1 percentage of patients was about 40%. ECOG 0 was 60%. And the peer company of ours, randomized Phase III, both in their treatment and control arm, the ECOG performance status percent was about half of that. So ECOG 1 was 20% versus 40%. So I think there, the question is if we had that type of population, might we see some additional benefit in survival? And as you know, Yigal, there is published data on how ECOG performance status does impact things like survival. So that's one for us to watch. But yes, so otherwise, it's a frontline patient population. And otherwise, I would say the patient baseline characteristics are going to be similar.

Okay. And then it sounds like you're not speaking about OS for the myeloma -- for the MUM study, that's not going to be something you're ready to talk about until sometime in 2026. Is that a fair statement?

Yes, when the IDMC looks at the PFS, randomized PFS, Yigal, they will look at OS, but our anticipation is that it's going to be too early.

Okay. Okay. That makes sense. All right. Well, let's move on then to talk about some of the other topics. So IDE397, maybe to start out with just that MTAP deletion market is a very interesting market. What is the commentary there in terms of the opportunity set?

Yes. So for MTAP deletion, this is an indication where there is multiple tumor types that are relevant here. I would say, for us, the tumor types of greater priority for us is lung cancer is one, urothelial cancer, and I would say, PDAC. There are other tumor types as well that we're evaluating. But I would say those are going to be the highest 3 priorities. And obviously, here, Yigal, as you know, we're 1 of only 2 companies that we're aware of that have 2 clinical assets in the MTAP space, obviously, MAT2A and our MTA-cooperative PRMT5 is now in Phase I as well.

Okay. And what -- if you could just kind of give us a quick synopsis of what you've seen so far with 397 in both the monotherapy and you've shown some combo data as well. And then what is the go-forward strategy there for both monotherapy and combo?

Yes. So for IDE397 monotherapy, we've seen a confirmed response rate in that 25% to 30% range. The expansion initiatives have been largely focused on the lung cancer and urothelial cancer. And now we've shifted the focus to the Trodelvy combination, which is data that we shared at our R&D Day recently. And there, as I noted, at the expansion dose that was selected for urothelial cancer, which is 30 mg once a day with IDE397, that's our monotherapy expansion dose. And Trodelvy at 7.5 mg per kg. And so that's an infusion once every 3 weeks. And as you know, Yigal, Trodelvy is approved at 10 mg per kg. So it's below the approved dose. And that dose in urothelial cancer, at least right now, it's a small data set, but we're observing a confirmed response rate that's in the 50s. So that's good to see. We're still monitoring durability. But so far, I would say, we've continued to see not just confirmation scans, but continued confirmation even post the confirmation scan. So I think that's encouraging to see. We're now expanding that work in lung cancer. So we've already dosed several patients there. And then, of course, now our PRMT5 asset is in the clinic. Our plan is to launch that combination work as fast as we can. So based on the starting dose the FDA agreed with us on our PRMT5 molecule, we believe by the second cohort, we should be at an efficacious dose. So -- and we have -- based on the protocol, have the ability to enable that combination once we're in that second dose cohort. So we'll get that moving rapidly. And there, I would say, our #1 priority for that combination will be in lung cancer. We have other, I would just say, strategies here around coalterations without getting into too many specifics, including in our internal pipeline, which is our third MTAP program, which is targeting, we believe, the largest coalteration of MTAP and that would give us 3 potential clinical assets in MTAP, and we would be very unique in that regard. And then all I could say is we're having other conversations about enabling other coalteration combinations in the MTAP arena at this time.

Okay. So for lung cancer, what -- is it going to be the same dose with Trodelvy? Or are you going to -- is that unclear at this point?

So right now, that expansion dose should be very viable in lung cancer. So that's our focus. I think the question is, can we dose Trodelvy a bit higher in lung cancer? Yigal, you may know that urothelial cancer patients, I would say, I'm going to make a general statement here, but are sort of viewed to be typically a bit more fragile than lung cancer patients. So you may have the ability to dose a bit higher. So we may explore that, but we do feel very good about this dose that we have.

Okay. And what type of response rate would be -- would it also be in the 50s that you got for urothelial? Or what are you looking for in the lung cancer to go forward?

So I think for both urothelial and lung, Yigal, that I think as long as we're sort of 40 and above with good durability, I think we're going to be in good shape. Obviously, if we can move 50s and above, even been better, of course. But in terms of kind of a go/no-go and that's what we've publicly communicated, I would say, ideally, 40s and above. Right now in urothelial, we're north of that. Obviously, we want to build more data, a larger data set and obviously more follow-up.

You're keeping the 397 dose the same in long end. It's just Trodelvy that you think you can go higher with in lung. Is that correct?

Potentially, yes. That we may interrogate to see if we might even drive greater efficacy. But obviously, the balance between safety and efficacy is obviously important any time we're doing ADC combination. So we know at that 37.5% at least right now, based on the data we have, which is growing, we feel really good about what we're seeing from a safety perspective. So -- but yes, could we go higher in lung and move that to 10%? I think that's still an open question.

Okay. And then what's the next -- so if these look good, what's the decision point after this? Is it -- do you do -- you go into late-stage studies in one or both of these? Or are you going to sort of determine which is the better place to start a pivotal urothelial or lung with your partner?

Yes. That is the next step. Yigal. Once we have a larger data set, more follow-up, we'll likely give that guidance perhaps in January for next year. But at least in my mind, the next step is a registrational study. And here, yes, that's going to be the decision matrix. Do you go forward in lung only? Do you do both lung and urothelial? Obviously, lung is a larger market. I think the part on urothelial that we're monitoring that we think could actually expand this market quite significantly, Yigal, I'm sure you saw the PADCEV PD-1 data in urothelial that's moving -- potentially moving even before metastatic into the pre-metastatic setting, and that data looks quite encouraging. So if that happens, we believe the frontline metastatic setting in urothelial could really open up. And if we can capture that portion, this market would get much larger.

Okay. And then for the combo with PRMT5, now it's your -- obviously, it's your own molecule. How are you feeling about this combo given some of the past attempts with other partners with PRMT5? How is this looking?

Look, Yigal, for us, to really make a run at frontline lung cancer for MTAP, this is our focus. Based on all the data we have, the preclinical data, we feel that's the best opportunity to really deliver hopefully an exciting result. We know others have launched a PRMT5 combo with chemo. I think that's with a checkpoint inhibitor. We think that's fine. That wouldn't be our preference. We think if we can deliver something exciting with a chemo-free regimen with a checkpoint inhibitor, would really position us in a unique way in frontline lung. So that's going to be our focus. We obviously have to start generating data now, but we feel very good about our MTA-cooperative PRMT5 molecule. We do believe that has a potential best-in-class profile. I'm sure we want to have time to go through that now. This molecule was made fit for purpose to combine with IDE397. So that's going to be our strategy and focus in lung cancer for frontline.

Okay. Right. And then have you talked about the doses? You said you're in the -- the second cohort looks like a good place to be. So what -- can you be more specific there about what the doses are? That's...

Yes, so for IDE397, we're fairly confident we know what that dose needs to be in combination with the clinical stage PRMT5 inhibitor. So we have that data. So we know what that dose should be. And then it's about as we dose escalate with the MTA-cooperative PRMT5 inhibitor. But based on our preclinical evaluation, we believe by the time we've cleared that first cohort in combo with a set dose we know with IDE397, we should be at an efficacious range. So we should be at an efficacious range essentially right out of the gates because we're not going to start the combo work until we get to the second cohort.

Right. Okay. And then in terms -- what would you be watching for in terms of possible like combo talks that needs to be monitored? Is there anything specific that just needs to be kept in mind? Or you wouldn't expect any additional tox on combo -- because of the combo?

Well, I think in this pathway in general, right, for PRMT5, I think probably the main one people have been looking at in terms of target related is around myelosuppression. So I think that's what we'll be looking at. Based on prior work we've done here, at least our perspective is we should have several cohorts hopefully available to us that we could move forward with. That should be efficacious. So -- and that's what we'll prosecute in the clinic and make that determination.

Okay. And someone was just asking me if you could expand a little bit on -- you mentioned these coalterations in MTAP. Can you just expand on that so everyone is clear on what you mean?.

Yes. So we know that MTAP is genetically -- has genetic alterations with other key mutations, including RAS, right? So in PDAC, for example, if you said pan-RAS, RAS more broadly, so KRAS, HRAS, NRAS, you're going to have most of those patients will overlap with MTAP and at least in certain tumors like PDAC. Depends on the tumor type they vary. The other beyond RAS, I would say, people have been looking at, which is CDKN2A. So you may know that CDKN2A MTAP are very often co-deleted because the proximity of those genes are very actually close. So when MTAP is deleted, CDKN2A is often co-deleted. So we've been having a drug discovery program on this for 5 years now. So we've been kind of over this for a long time that this could be a great combination opportunity. But for us, we are very focused, and we wanted to wholly own this. So what we're doing, I can tell you is the chemistry has been very, very complicated, but we're very much nearing a candidate now. And so if we can deliver that, that would be our wholly owned strategy around taking that coalteration approach. So hopefully, with that, we would sort of cover our full basis on what's at least what we think are the most exciting rational combinations. And then in that case, we would wholly own 3 sort of pieces of this lever that would position us very uniquely in the market.

Okay. So that's -- you're not ready to talk about an IND on that yet, I guess.

Candidate this year, yes, I would hope sometime next year, maybe towards the end of the year. And for coalterations, PDAC is one, an obvious one, maybe lung cancer, although I think we will prioritize the MAT2A, PRMT5. But probably for the PDAC 1, the coalteration strategy for us is one we want to consider and ideally with an internal asset as well.

Okay. I want to make sure we get to DLL3. So 849, obviously, a lot of interest in this asset you've got from Hungry. Well, so just essentially tell us what is the thesis on your construct? You've got competitors, as you know, but what makes yours stand out?

Look, we feel that our ADC DLL3 with Topo-1, it's from a very battle-tested platform from Hungary. They've been in several thousand patients across multiple clinical stage programs that have been presented in many major medical conferences. And so we know that this platform, in particular, around the payload linker system, which is a tetrapeptide cleavable linker, which is sort of your classic linker system, which cleaves once it's internalized in the cell from lysosomal degradation. So here, things similar to the Daiichi platform that was developed. And so we feel really good about that platform and its level of validation in patients across many, many studies. So we feel we are positioned to have a potential best-in-class molecule based on what was presented at the World Cancer Lung Conference recently in Barcelona. Just a quick snapshot, at least at the 2.4 mg/kg expansion dose, which is likely going to be one of our expansion doses, what we focused on. And just so you know, that's the starting dose, the FDA has let us begin our U.S. trial with. So our first patient that was dosed was at that dose. We saw a confirmed response rate by RECIST at 70% in the second-line setting. So we feel really good about that number. We've seen a lot more data since the World Cancer Lung Conference. I would say the data continues to remain to be robust. In addition, on the PFS side, across all lines, we reported a PFS of approximately 6.7 months, which, Yigal, as you know, is a very good number in relation to other reporting of PFS, including IMDELLTRA, where they got accelerated approval. It was 4 months and change. So we're obviously north of that. We want to build that data set, build it with a global data set, and that's what we're doing right now. So I think we're in a good spot. And now it's going to be about speed as well as what is our clinical strategy. We're not prepared to disclose at this time all the specifics for competitive reasons, but we would like to launch ideally 2 registrational trials here. One in small cell and then I would say another -- and an another tumor indication. And we do believe, based on the data we're seeing, we should have some monotherapy accelerated approval path available to us.

Interesting. So you're thinking monotherapy accelerated approval because, as you know, some of the competitors are talking about comparator arms and talking about needing to enroll in IMDELLTRA in those arms, which obviously raises the bar for a randomized trial. So is that -- any comments you want to offer on that discussion?

I mean, you can still have an accelerated approval path as a randomized study, right? So I think there, Yigal -- and that's the part we just haven't disclosed yet what our strategy is. But what I can say is we're thinking about this very deeply. I know the clinical team, I know is even going to investigator about next week to kind of dig even deeper on this and get them put on where to focus that monotherapy accelerated approval study. But hopefully, we'll be able to share that publicly here not too far in the distant future. The only other part I should have mentioned to you, Yigal, was at that expansion dose I mentioned because I think maybe there was maybe some lack of clarity on this, we are seeing what we think is a very viable AE profile, where the Grade 3 or higher AE rate at that does, at least right now, was observed at less than 20%. We think we can go even higher than that based on that because as you know, with ADCs, it's very common to have Grade 3 or higher, that's in the 30, 40-plus percent range, including very well in ADCs like in HER2.

So that's still kind of in being determined, 2.4% is not -- you say starting dose, but it could -- there could be a potential for going higher.

We could potentially go higher. I think right now, we're very focused on that 2.4 dose, and we will be doing some dose evaluation, but we feel good about that dose. So -- and that is -- we think it's a very viable dose at this time.

Any thoughts you want to offer on activity in the CNS tissue for your construct?

So the data that was presented at the World Cancer Lung Conference is a response waterfall in patients that have brain metastasis. And what we could say is we're seeing very robust activity there. So we don't see any diminishing of response in that setting. So we feel good about that. And obviously, brain metastasis is something you have to think about in that indication specifically.

Right. Okay. Well, in the final few minutes here, you have a lot of other assets earlier stage or some that are earlier stage, some that are maybe getting close to mid-stage. So could we kind of walk through those in a little bit more rapid fire maybe? I mean which of those would you like to highlight? Pol Theta PARG. Obviously, you have the KAT6 as well. There's a B7H3/PTK7, bispecific ADC. And then there's -- well, we talked a little bit about 892. But so of those other ones, can we just kind of quickly run through those? And where do you see the most opportunities?

Sure. Yes. I can do more rapid fire here, Yigal. So maybe with -- quickly with Pol Theta, so for the listeners here, this is partnered with GSK. Focus is combination with niraparib, specifically in patients with BRCA. That escalation has been going well, has entered the combination phase with niraparib. We do anticipate an expansion dose being selected here relatively soon on that combination, which would trigger an expansion milestone payment. So I think that's all going as planned. For Werner Helicase, as you know, Yigal, we feel we have a truly differentiated best-in-class molecule based on the way it binds to the D1/D2 domain and does not bind with that key cysteine 727 interaction, which both the Novartis and Roche molecule do. And our hypothesis on this is that, that should lead to greater efficacy. That escalation is also going well. And hopefully, we'll be able to give some disclosure update on that with GSK over the next couple of quarters. And there, we do plan to have, I would say, an important focus on gynecological cancers as well as combination with GSK's checkpoint inhibitor, dostarlimab. The next 2 programs here, Yigal, I think that at least I'm excited about. We're quite excited about what we're seeing preclinically with our B7H3/PTK7 bispecific. It's very active preclinically. We do see enhanced efficacy versus B7H3 alone or PTK7 alone, based on the data we have. The co-expression population is large, which is why we pursued this bispecific, in particular, in indications like non-small cell lung cancer and CRC, which are 2 very high priority areas for us. So now that IND has been filed, it's essentially now a clinical stage asset. We will be very focused, I didn't mention it, for DLL3 on the combination with PARG IDE161, which is our other Phase I asset to enhance the durability of Topo ADC. So that preclinical data, which we shared at R&D Day, looks great. And we've demonstrated this with multiple ADCs now. So that's going to be a really key strategy and lever for us for both DLL3 and the bispecific. KAT6/7, we're tracking towards that IND next month. A first-in-class dual inhibitor. One of these have not entered in the clinic, is similar to our bispecific. No bispecific of that format has entered the clinic yet. Clearly, we believe, based on the preclinical data we're seeing, with that dual inhibitor, we should be able to drive greater efficacy. So we're going to be very focused on that asset as well. Breast cancer, lung cancer, obviously, big opportunities. As you can imagine, Yigal, we're not going to be able to do all of this on our own. So we're having a lot of different conversations, a lot of interest across the portfolio. But we're trying to build a world-class precision oncology company, and we're thrilled to really be able to get these next set of assets in the clinic and to move them forward.

I mean there's been a lot of excitement around the Werner Helicase, and I was tracking it for a long time. Is that -- are we close to getting an initial readout there? Because I know you had -- you showed the biomarker data, I believe it was earlier this year, if I'm not mistaken. But what is the time line there? And what would be a good result that would be compelling?

Yes. So yes, we need to be in coordination with GSK, Yigal, on that and kind of the timing of that disclosure. Ideally, it's one we sort of have an expansion dose, and we have a decent amount of data there. So we'll be in coordination. In terms of response rate, I think there will be about, at least in my mind, what do you see in the setting of gynecological cancers. So here, in this case, specifically, indications like ovarian and endometrial cancer. And then I would say the second area is most likely around CRC, which we know MSI-high is very prevalent in both of those settings. And you may see different response rates depending on which of those 2 indications or general indications that you're focused on. So what would we like to see, I mean, Yigal, I think for a monotherapy ideally, right, you're above 30% or higher with good durability. So this will be the first time again, a molecule is going to be tested in the clinic without this cysteine binding interaction. And we know, in the MSI-high setting, in particular, you have to be very thoughtful about point mutations because it's a very rapid point mutation clinical setting.

You're doing high MSI for Werner, and I guess -- say solid tumors, but I guess, as you point out, endometrial would be important. And then you're also doing that with the PARG with KEYTRUDA. So -- but I mean that's a different partner, obviously. So when you think about those 2 programs, they could both proceed together, right? I mean how do you think about that? Is one going to win? Or is it not like that?

Yes. So with PARG IDE161, our focus here, Yigal, is around enabling the Topo ADCs. And the reason why we've been evaluating the KEYTRUDA combo, which were done with that dose evaluation is really answer the question around safety. Could we generate that safety data? And we've done that now. And the reason why we did that was as we enable the ADCs now with the ADCs with PARG, right? As you can imagine, in lung cancer or non-small cell lung cancer, in particular, you're going to likely want to do that with a checkpoint inhibitor. So that's why we have that data to enable as we think forward about the PARG checkpoint ADC combination. So that's where our focus is for that program versus sort of competing with Werner independently in the endometrial setting. So hopefully, that makes sense. But that's our current focus right now.

Okay. And then the KAT6/7 axis is also heating up. There are several other players in the space. You have a dual or I guess, it targets both 6 and 7. What is the rationale for that versus just one of them? What does that get you in having the dual targeting?

Yes. Our CSO, Mike White, is -- we think, is one of the world experts in this arena. But really, the shortened version of it, Yigal, is we believe 7 is a key potential bypass mechanism for KAT6. And so we have a lot of preclinical data to support that statement, including data that we presented publicly. But essentially, we feel by hitting these 2 together, which I would call for now paralogs, we feel we can drive much greater efficacy and durability as well as expand the opportunity into other indications beyond breast cancer, including indications like non-small cell lung cancer. So the chemistry was very complicated. And our team is obviously first rate, and we were able to figure that out. So we'll be in patients here shortly. I should have mentioned also the GLP tox in both species is done. Everything is behind us now. So it's really just about IND assembly, some of the final CMC pieces. But essentially, we have the data we need to file that IND.

I was always impressed with some of your more technical slides where you talked about advanced physics concepts like free energy perturbation, which doesn't come out much obviously, to say the least. But really, the question is, with the obvious acceleration in machine learning and LLMs and everything and AI and AGI, how much more is that helping you in the very early -- in the early innings when you're designing molecules? Are you leveraging that even more because you do have a platform, which I'm sure you're working on all sorts of other things you haven't talked about at all. But how much more is that technology advancement helping you?

Yes, Yigal, look, it's been an amazing voyage because we began investment in this area several years ago, and we began applying it to serval programs like KAT6/7 as well as others. And now it's essentially integrated across all of our drug discovery programs. And I would say it's been a really great surprise how much more efficient it's made us in terms of optimization of certain parameters of lead optimizations. We found that it's not great for certain areas of lead optimization. But extremely effective in others. So I think here, it's really about knowing where to apply it. And I would say what makes us unique versus some of the other players here is that we're applying it to first-in-class targets, right? And we also have a machine learning process and system that's becoming more and more sophisticated because we're actually feeding the results into the next program. And so we probably have one of the largest Helicase databases, polymerase databases of chemistry, and we're feeding that into the machine learning system. So it's actually getting more and more effective and sophisticated in what it's doing. So look for us to double down in that area, Yigal, and that's why we highlighted it at our R&D Day. And I know for our CSO or our Chief Technology Officer, it's definitely a very much of a high priority. So our hope longer term, it should just essentially accelerate our time to IND, right? So by how much? Right now, we're probably seeing about a 30% acceleration, which is big, right? So now the question is, can we even enhance it beyond that? So before it took us 5 years to get Werner. If you could do that in 3, that's big, right? If you could even reduce it more than that, that's a game changer for us.

And then without giving away the house secrets, so to speak, I mean, you mentioned areas where it's more effective and less effective. Just generally speaking, like how do you characterize like that? Where is it -- where are the pitfalls? Where is it -- where do you know that it's just -- it's not going to do a good job? And where is it really, really, really helping you?

Yes. That I can't tell you, Yigal. I'll get in trouble if I told you that, but I do know the answer. So -- but yes, there are definitely [indiscernible] areas. And we actually have great data on it, which I've seen just where -- how predictive AI has been for certain parameters of lead optimization and where in other parameters, it just -- it does not do a very good job. So you do need to know where to apply it effectively. And as companies get more sophisticated and build their own databases, I think companies will get better and better at this, which is great for the whole sector.

Okay. Great. Well, thank you. Very interesting, for sure. And we'll be watching for the MUM data any day now or in the next few weeks, I guess. So thank you. And congratulations on a very, very strong progress over the last 6 months.

Yes. No. Great, Yigal. Thank you so much for the time and really appreciate the chance to talk today.

Okay. Absolutely. Take care.

Bye for now.

Bye-bye.