Everyone, to Novel Mechanisms in Oncology Part 2, because we had a Part 1 on Tuesday. Just a lot of oncology to discuss this time. So we have three really important companies in the space: BeiGene, Mark Lanasa is the Chief Medical Officer in the solid tumors; ORIC Pharmaceuticals, Jacob Chacko, the CEO; and from Syndax Pharma, Michael Metzger, CEO. Welcome, all of you. Thank you so much. It seems it's amazing how quickly a year went by. As a reminder, if you have questions and you'd like to ask anything, just type it into your laptop and I'll see it on the tablet up here. So with that, let's get started. Just to orient everyone, if we could do a quick intro, high level, what the programs are, what's approved or in the works. You have two approvals, so I'm sure you're going to want to talk about that. And then move down the line, and then we can get into more nitty-gritty. Sure. You want me to start? Go ahead, Michael. So first of all, thank you for having us on the panel. Great to be here. Nice to see everyone. So Syndax, for the uninitiated, is focused in the area of targeted oncology in the heme space. And we do have two approvals. Very happy to say, proud of the fact that we had, in the last 90 days or so, two drugs get approved. One is for GVHD, chronic GVHD, third-line GVHD, which is a drug called Niktimvo. So this is a new mechanism, CSF1R antibody used to treat these patients, first of its kind. So that was approved in August this year. And the second agent is now known as Revuforj, Revumenib, first and only menin inhibitor approved for KMT2A acute leukemia in adults and pediatric patients. And we recently had pivotal data in NPM1 as well for adult AML. And that data read out with a positive trial just a few weeks ago. And so we are now in the midst of a launch for Revuforj and soon to be launching Niktimvo as well. So really interesting time for the company as we transition to a commercial stage organization. And we'll look to expand these molecules beyond their first indications into earlier line patients and other important areas as well. Great. My thanks as well, Yigal, to you and to Citi for including BeiGene in this important discussion. My name's Mark Lanasa. I'm the Chief Medical Officer for solid tumors at BeiGene. BeiGene is a mid-sized oncology development organization. We now have about 11,000 global employees. And we're focused specifically on oncology and hematology development. There are some adjacent areas such as INI, where the biology crosses, and we have some effort there as well. We have two globally approved medicines. One is Brukinza, which is our, in our view, best-in-class BTK inhibitor. The other is Tevimbra, which is our PD-1 inhibitor. We have a relatively expansive pipeline that's growing by the day. We had 10 NMEs enter the clinic this year. We have something like 15-20 investigational programs across focus areas of B-cell malignancies, lung cancer, GI malignancies, and breast cancer. Great. Jacob. Yeah, thanks, Yigal. So my name is Jacob Chacko. I'm with ORIC Pharmaceuticals. ORIC stands for Overcoming Resistance in Cancer. In a nutshell, that's the mission of the company. As you think about what areas of cancer we tend to focus on, in general, as you look at the pipeline, we have a big focus on small molecule drug development as opposed to antibodies. And then within small molecule drug development, we do take a precision oncology approach to three specific solid tumors, which is namely prostate, lung, and breast. The two most advanced programs are ORIC-114, which is a non-small cell lung cancer program directed at three different targeted therapy populations within lung. And then ORIC-944 is a prostate cancer program, which is being developed in combination with AR modulators. There's a lot of literature that suggests there's synergy in combining a PRC2 inhibitor, which is what ORIC-944 is, with an AR inhibitor like enzalutamide, daralutamide, or apalutamide, and so that's what we're doing with that program. Both of those programs have finished phase 1 dose escalation. Both are in dose optimization right now, and for both of them, we hope to be in pivotal studies by the end of next year or early 2026, so a lot on our plate from the execution side of things, but a lot of exciting developments in the pipeline. Great. So I asked this question in the first iteration of this panel yesterday, two days ago. But to keep it sort of broad and thematic, there are three sort of important pillars of the stool: biomarkers, targets, and rational combinations. So how do you each think about those domains when putting together your clinical plan? Great to hear. And of course, reference the specifics of your pipeline as you wish. So Michael. Sure. Thank you. Good question. So when we think about targeted therapy, I mean, you start with the premise that you're targeted, right? So that you're going after a specific, in our case, driver mutation or impactful target that really gets at the biology of the disease and is fundamental to that disease. And I think we've proven that with both of our agents. And so when you start with that premise and you prove it as a monotherapy in your trials, the ability to rationally combine from there gets a lot easier, right? When you have to combine an agent in order to advance it in clinical trials, it's always the question of, well, how do the components actually contribute to the efficacy and/or safety of what patient you're trying to treat, basically, with your drug. I think we've fundamentally established that through the mechanism of both of our agents. We've found combinations that we believe are synergistic and also part of the standard of care, specifically around Revuforj. Now we're going to learn a little bit more about that this year at ASH as we roll out more combination data, both in newly diagnosed patients as well as patients in the relapse refractory setting. That work is ongoing. It will help us to expand the use of our medicine beyond the monotherapy indication that we have for KMT2A and hopefully for NPM1 when we submit that application next year. I think it's a very clear strategy of establishing monotherapy, moving to combinations, rational combinations, adding to your safety and efficacy in a positive way. Biomarkers. It's also a benefit to us in our particular population, specifically for KMT2A and NPM1. These are tested. Patients are tested. They're genotyped when they come into treatment, and so we know whether they have a KMT2A rearrangement translocation or if they have an NPM1 mutation, so that's very clear who we're treating and why we're treating them and how the medicine actually impacts. Great. So starting with targets, the way our research and discovery organization approaches this, again, is to go back to what are the tumor types where we're trying to drive impact. So again, that's lung, GI, and breast cancers. And then from there, they'll select targets where being, shall we say, platform agnostic, that there are targets where the introduction of a new platform can bring benefit to patients, hopefully significant incremental benefit to patients. So perhaps one example of that is the EGFR targeted degrader that we are about to bring into the clinic, evidently, that we think that that addresses a specific unmet need for EGFR mutant non-small cell lung cancer for patients who have been previously treated with a tyrosine kinase inhibitor. Another example in the GI space is that we're about to start treatment with an FGFR2b targeting ADC, which was specifically designed to mitigate ocular toxicity associated with the binding of FGFR2b with either of its cognate ligands. From there, biomarker selection can be simple or can be complex. For a target like EGFR, genetically driven patient selection is relatively straightforward, whereas expression-driven patient selection can be fairly complicated. Then the last point of the idea of having these different disease area franchises and advancing multiple targets is that it naturally enables rational and portfolio combinations. For our EGFR targeted degrader, certainly we would like to combine that with our EGFR by c-MET trispecific antibody that recently entered development in the breast cancer space, our novel CDK4. We want to combine with our novel CDK2 and so on. Great. You go. I could talk for an hour about biomarkers, targets, and rational combinations, but I'll give you the 30-second version of it. Go for it. Yeah, which is for us, we won't develop anything that doesn't have a biomarker. I mean, it's just too hard in oncology these days to go forward shooting in the dark without a biomarker. So it wouldn't be of interest to us to develop a candidate that doesn't have a biomarker. I think the real tension is with targets, at least for small oncology companies, which is, as you think about the targets and just the sheer number of oncology companies, I mean, being overly simplistic about it, you can either have highly validated targets where the biology is really easy to understand. EGFR is a great example of that. In the case of our program, ORIC-114, we're looking at three different populations: EGFR exon 20, HER2 exon 20, and EGFR atypical mutations. The biology there is super easy to understand. And the bar, competitively, of what you need to clear to have a best-in-class program is easy to understand. But that attracts a lot of competition. So you end up having to fight with a lot of other companies to enroll those trials and to try to really differentiate yourself with the best-in-class profile. We very intentionally pair that up with a target that I would say is less validated, at least in prostate cancer, with PRC2, namely for ORIC-944, where the biology is harder to understand. And in fact, the true potential of that target is really unlocked with a rational combination with AR inhibitors. For a small biotech company, that's often a tough endeavor to go down that route. Now, in our case, by luck, we have a big pharma that has helped de-risk that rational combination of those two targets in combo with one another. So it makes it a lot easier. But otherwise, I think rational combinations for small biotech companies are quite difficult unless you have been very thoughtful about your target selection and then also the use of biomarkers early on in the trials. Another big industry theme, which is sort of an introduction into the world of drug development recently, is this Project Optimus, which obviously you guys know all too well. So I guess the question is, I remember my sixth-grade teacher giving me busy work just to do, but I didn't learn anything, right? So is this Project Optimus just busy work, or has the FDA actually introduced some additional requirements that have caused you to enhance your understanding of what you're doing with the right dose? Or I'd love to hear your thoughts on that. Michael? Yeah, sure. Well, having gone through the process of getting two drugs approved recently, I can say I think Project Optimus, I think, is a real thing. I don't think it's made up, but I think it's a real thing. Obviously, they're quite serious. The FDA is quite serious about picking the right dose. But I think they've always been quite serious about picking the right dose, and so from our vantage point, what we've always really focused on is doing that and exploring the full range of the dose and really understanding. I mean, for instance, when we started developing Revuforj or Revumenib, we knew there was an impact of CYP inhibitors, right? These are antifungals that patients are on. Our drug is metabolized through the CYP3A4 pathway, not interacting with other drugs, but metabolized by CYP3A4. And so we needed to understand that relative to where we are and where we would be in the dosing range if a patient was on a strong antifungal or a weak antifungal. That took a lot of work, right? And we had to tease that out. And that was properly done early on in development, in the course of the phase 1 trial, in order to establish the proper dose and see where we landed relative to efficacy and safety and having that other factor. So it's a normal thing that people encounter, especially in drug development. But doing it early and actually really painstakingly going through and establishing the dose early and feeling very confident about the dose paid dividends for us, really accelerated our program. And as a result, we have a very, I think, very strong label relative to how do we dose patients, both who are on strong CYPs or who are not on strong CYPs. So again, I think it's just an example of what we did on our own. The agency has their own criteria and their own way of operating, and they sometimes push companies into it. We actually came to the table and established the dose in a way that we thought was appropriate, and it paid dividends. So my view is really quite similar that overall, I would say I think Project Optimus is largely constructive. The core message is to really understand your exposure response relationships early in the development of the program. So for example, with the ADCs that we're bringing through early development, that's critically important because the hematologic and some of the other cytotoxic-like side effects are clearly dose-related. So you'd like to understand what is that exposure safety relationship across a critical range of doses. Now, for some of our biologics that are safe across a broad range of doses, sometimes it can feel a little bit overdone if you're having to enroll a substantial number of patients across two dose levels that are fundamentally likely to be safe. But nevertheless, the focus on exposure response, I think, is the critical issue. The reality is that few things can derail a development program quite like choosing the wrong dose that can really, really set you back. And the other point that I think is important is that because FDA has written guidance documents around this now, that if we're following the guidance documents and are having check-ins with FDA that are otherwise in parallel to the progression of the molecule, that there's minimal time impact overall. As long as we don't have to stop, have a conversation, and then restart, I think it doesn't have a substantial impact. Yeah, I very much agree with what's been said. I think thoughtful drug developers have been adequately focusing on dose selection, as you mentioned, for a long, long time because that's so critical to the success of a program. So I think on the margin, perhaps it introduces a bit of busy work, but it really is more, I think, just trying to more systematically have the broader industry take the same approach that the more thoughtful drug developers have been doing anyway. So we view it as really just a slight extension of what would have already been done anyway without Project Optimus. I'll ask one question on novel approaches to drug development, obviously, and machine learning, AI, putting it all in one bucket. I mean, to what extent are each of you leveraging that new technology, whether it's in terms of target identification or building the molecule, or even something as simple as assisting with the filing, for example, to build the document? I'm just curious, all of the above. It'll be great if we had a machine to do our NDA and our SNDA and all that, but that doesn't seem to be where we're at yet. No, I think, look, we're not in the drug discovery end of the business. I know that there's quite a bit of machine learning and AI work that's gone into target discovery, and that continues to evolve. I think in our case, we've actually taken it more from the commercial end of the story where we're profiling accounts. We're buying data to understand where patients actually are, right? So you can buy lab data. You can identify patients early on in their treatment journey and then use machine learning to understand how they're going to track through their treatment process. And so accessing the patient as quickly as possible and getting them the drug as quickly as possible is absolutely most important within the treatment of AML and ALL. And so we're using AI and machine learning to do that work and actually build our engine on the commercial side. So it's early days, but we're not alone in doing this, but certainly in targeted therapy, taking a page out of the rare disease playbook, you've seen this before, and it's really evolving nicely for identification of patients. On the development side, this algorithm-based programming is extremely powerful, and we're just now beginning to tap into what we can get out of that. One simple example is we're trying to develop algorithms that can teach a computer to do some of the more tedious aspects of clinical data review. We've had edit checks for a long time, but now you can take that to the next level and have cross-checks and those sorts of things that really can, again, take some of the tedious work that the teams, frankly, don't love off of their plate. We're thinking about across the entire life cycle of development, how those types of technologies can be implemented. An interesting facet of your question is how can this be used in the research organization? A recent, not so long ago, conversation I had in this regard is that within our research organization, they have found that AI or algorithm or machine learning can be really helpful for identifying correlations and associations that would be otherwise very difficult to identify. It doesn't necessarily correlate or it doesn't necessarily predict causation. The example that was shared with me is you could ask AI a question, what are genes that are associated with PD-1 resistance? And it will give you answers if you give it a large data set and an algorithm to work through that could identify targets that would not be intuitively logical or might have taken you a long time to find on your own. However, that doesn't mean that those molecules are causative. They might just be changed with whatever the driver is. So there's still an indispensable role for the wet lab research where you can identify interesting targets through correlation, but then you have to go back to the lab and actually prove it to decide, is this something that's worth actually targeting clinically? Quite similar from our experience, Yigal, I think we haven't yet found the AI easy button that helps us discover a drug and take it all the way to NDA. But we're trying to figure out, largely with the help of outside vendors, ways that we can either facilitate the drug development once drugs are in the clinical side, but even more so actually on the discovery side. We have internal discovery capabilities at ORIC, and where we've found at least AI capabilities from these outside vendors to be most helpful is as we look at target selection, as we look at various targets that are thought to be undruggable. It's been quite helpful to start to try to do some work around how AI might enable us to either pick targets or, more importantly, even figure out how to drug things that are considered undruggable. I'd say I always, I'm sort of waiting with bated breath as to what the results are, and I think it'll take some time to see the full results of that come to fruition. But I'm hopeful that the industry continues to utilize AI in a more productive fashion. I personally think it'll be more with the help of outside vendors that are developing these capabilities as opposed to every single company trying to develop the capabilities in-house. I'm not sure that that's really scalable for the industry, but we'll see. Okay, now we can give everyone a chance to "pitch" their own book on each of the pipelines, get into a little bit more detail. Now, we've talked at the beginning about the two approvals. Let's focus on AML since that's the more relevant one for this panel. So tell us a little bit more about the approval and then the recent data in NPM1 and how you see the market evolving. So recent approval in KMT2A acute leukemia. It's a first of its kind. This is about 10% of AML and ALL. This affects both children and adults. So it's a very serious, probably the worst prognosis you could get in AML or ALL, and it affects even very young children. So this is, again, the first of its kind where the agency was supportive of giving a very broad label that crosses over AML and ALL, adults and pediatrics. So we were able to cover the whole landscape, which is extremely exciting. The drug is now available, and I think the label is very clean. So it's easy to administer this drug. It's an oral pill given twice a day and has very significant impact. About two-thirds of the patients actually respond to therapy, and they respond very quickly. So they go get on drug, and usually within a month, they have cleared their tumor. And so that is pretty remarkable. This is an average patient population, about 35 years of age. So they have a lot of living to do, and the fact that they are out of options after their first line of treatment and they come to a drug like this that can get them to a remission, a deep, durable remission within a month and have the chance to actually go on to get a transplant is remarkable. About 5% of patients get transplanted in this setting. And in our trial, we were about 40%, roughly 40% of patients getting transplanted. So that's a very big difference. And now we have the opportunity to put them back on drug in the post-transplant setting. So all of that is, of course, based on a novel agent that came through our pipelines, and we're quite excited about that. In kind of a remarkable fashion, we have a second indication. Usually, you have to wait years for a second indication to bear fruit. NPM1 affects about 30%, 30, 35% of AML patients. It's mostly adults. And the results that we had just a couple of weeks ago in our pivotal trial, 64 adult patients. The safety database is about 84 patients. Primary endpoint is CR/CRh, the same as it is for KMT2A. We had very similar results in terms of point estimate, 23.4%. CR/CRh, about half the patients get a remission. So overall response rate was 47%. Duration of response, 4.7 months. Time on therapy, somewhere around eight months on average. If you count how long it takes to get to a, it's slightly longer to get to a response, and then they stay on drug in this trial so about eight months. Again, in this population, what we showed in this trial was 75% of the patients had received prior Venetoclax, so they were very sick. That's a very bad prognosis once you fail Venetoclax. These patients are older. They're on average 65 years of age versus the 35 years of age for KMT2A. So they have gone through a lot of treatment. They're extremely fragile, and they're able to take an oral pill, and about half of them get to a response and do well, and so many of these patients are actually able to even get a transplant, which you don't see in this population very much, is enabled by this drug. Safety, efficacy, very similar in terms of all measures, including MRD negativity, all very similar to what we saw with KMT2A as monotherapy, and again, this will be the first drug of its kind, we think, approved in the second indication. We'll submit an SNDA next year, and we'll have an approval next year. This is a new class of agents. It's called menin inhibitors. We'll be the first. We are the first to get an approval. We'll be the first on the market selling. Probably depends on who comes next, but maybe a year to two years in advance of competition. KMT2A, there's no approved therapies, and nothing works. Even standard of care doesn't work. NPM1 at this stage of disease, nothing seems to work for these patients as well. Quite an interesting space of business to launch into. And then, as I mentioned earlier, we're working in combination now. We have relapse refractory trials as well as newly diagnosed patients who are combining with venetoclax and another trial in chemotherapy. So we're now getting to see how the drug combines, and it looks really, really good in combination in terms of safety and efficacy that we're adding to the doublet and not giving any worse on the safety side. And that data will start to emerge even more at this year's ASH later this week as we present that data. But very exciting time. Physicians are excited to have a new agent, and they will look to deploy it pretty liberally, we think, across KMT2A and NPM1. How quickly do you think you could launch an earlier line late-stage study based on some of this data? Yeah, so we are actually starting our pivotal trial in unfit patients. So in combination with venetoclax, that's, again, some of the data that we're showing this year at our ASH event. But the pivotal trial will start before the end of this year. So we're getting going there. And we think we'll start a fit patient combination trial with 7+3, our drug in the fit patient population sometime in 2025. Okay. And Mark, we had a pretty extensive conversation a few hours ago on your solid tumor pipeline. Can you give us sort of the maybe highlights of some of the targets that you think are most interesting or most promising across not just small molecules, but the degraders, the ADCs? What would you like to highlight? So again, we're platform agnostic. So we think first about the target, and then we think about what platform to use to go after that target. We have a number of targets that we've brought into the clinic this year that we're very excited about. We recently had our first patient dose with our pan-KRAS inhibitor, which has very interesting characteristics in terms of, as a non-covalent inhibitor, it can cover both the on and off forms, and therefore should cover not only patients who have oncogenic KRAS mutations with high selectivity for KRAS over HRAS and NRAS, but also should cover KRAS amplification, which could be an important mechanism for resistance to the RAS targeting covalent inhibitors. So to go back a step, the way our research organization thinks about this is our aspiration is to have a mix of first-in-class and best-in-class molecules. If we're going after a target where we're looking for a best-in-class profile, an established target, the research organization will also have a TPP. They call it a TPP. It is their preclinical target product profile. And they'll say, what are the characteristics that we're trying to design into a molecule? Because particularly for small molecule chemistry, you're usually trading off one thing for another facet. We have a large preclinical research team of almost 1,200 PhD scientists. At any one time, we have between 50 and 70 preclinical targets under investigation. And it's our aspiration to bring 8 molecules-10 molecules per year into the clinic. So we're only bringing forward those things that are actually hitting that preclinical TPP. So very excited about our pan-KRAS, very excited about our EGFR degrader, highlighting our CDK4 selective inhibitor that we talked about earlier. We're very excited about our internal ADC profile or portfolio. We not long ago had our first patient enrolled in our CEACAM5 or CEA targeting ADC. We are already enrolling quite quickly on that program, and that's advancing well as well. Okay. So Jacob, two very interesting assets that you're developing. Tell us a little bit more about both of them and how they're differentiated, especially for, well, starting with the EGFR. What's special about 114? There are obviously a number of molecules in the clinic. You've got some very good early data, which I think maybe you could highlight again from ESMO 2023. And then on the prostate side, talk about, you mentioned the data from pharma. How does that enhance your conviction on what you're doing and the approach that you're taking in terms of binding the PRC2 complex that may be different? Yeah, so we'll start with ORIC-114, the lung cancer focus program. So I mentioned three populations of interest, EGFR exon 20, HER2 exon 20, and EGFR atypical mutations. Collectively, that's around 6.5%-7% of non-small cell lung cancer. They're constituted by those three populations. There's some homology in those targets, which is why you can have a drug that is equally potent against each of those targets. And those are the three populations that we're looking at with the program. As you look at those populations and in terms of what's been approved, which is not very much, Amivantamab from Janssen is approved in EGFR exon 20. And then there's a couple of other agents approved for the other, for at least for HER2 exon 20. But the main issue has been in two broad categories, which is either safety or lack of CNS activity or both. On the safety side, what I mean is, in particular, if you drill down on EGFR exon 20, I think folks are very familiar with the first-generation compounds, Amivantamab, and also Mobocertinib from Takeda, which was pulled from the market. Both exhibit quite a bit of EGFR wild-type toxicity. So in trying to hit the EGFR exon 20 target, they're also getting too much wild-type activity, which is leading to things like either rash or diarrhea. The next generation of compounds, ourselves included, have been able to dial back the wild-type toxicity and have more selective compounds. So you see a minimization of the EGFR wild-type toxicities. But what you continue to see from nearly every one of those programs is off-target toxicity. So as you look at the kinome trees, they hit multiple off-targets aside from the target of interest. And that leads to things in the safety profile like anemia, QTc prolongation, liver enzyme elevations, things that you really don't want to see in your safety profile of a candidate. With ORIC-114, we've been able to, this is a program that we in-licensed, we've been able to find a program that is quite selective on both of those fronts. And so we seem to be seeing less of the wild-type toxicity and then, importantly, almost no off-target toxicities with 114. And then even more important is the CNS activity. So in all of those populations, you don't have a drug that is approved or in late-stage development that has CNS activity. It is now dogma among lung KOLs that you want a drug to be CNS penetrant because you've got 35% of patients at initial presentation that have CNS metastases. Several patients, some large percentage, have the metastases that you just can't visualize them on a scan. Those are called micrometastases, and nearly every patient will progress in the brain if they're treated with a drug that is not brain penetrant, so that's why it is so critical to have a CNS active compound. You referenced our ESMO 2023 data. At that time, we had dose escalation data in EGFR exon 20 and HER2 exon 20. We hadn't yet started working on atypical mutations at the time. But for those first two populations, we had already seen good evidence of systemic activity and intracranial activity with the safety profile, like I said, that we were shooting for. In the case of PRC2, as I alluded to in my remarks earlier, so with ORIC-944, it's targeting PRC2 in prostate cancer. That's an epigenetic modifier. The biology is a lot more complex there than it is with the lung targets that I just mentioned. The real de-risking event there has been the large pharma I mentioned is Pfizer, where Pfizer is developing a PRC2 inhibitor in combination with their AR inhibitor, enzalutamide, and in at least a phase 1 setting, they've showed in two different metastatic CRPC populations, they're showing a radiographic PFS that is anywhere from 3.5 to 4 times as long as you would expect with enzalutamide alone in those settings. They've also got an open-label randomized study that it sounds like they're going to show the results next year that we get a lot of questions about, and we sort of refer folks to, you should ask Pfizer your questions, not us. And then I think most importantly, within the last three months, they've started two global phase 3 studies in metastatic CRPC populations for that program, again, in combination with enzalutamide. In our case, we've finished single-agent dose escalation with our PRC2 inhibitor and now are exploring dose combinations with the other two major AR inhibitors, one daralutamide from Bayer and the other is apalutamide from Janssen. So we have clinical collaborations with both of those companies. They obviously understand the strategic rationale here of combining a PRC2 inhibitor with an AR inhibitor. We're doing dose-finding work, dose optimization work there with the plan to try to start our own pivotal one or two pivotal studies by the end of next year or early 2026. The real issue that has plagued the PRC2 field has been poor drug properties. So as you look at companies or drugs that have come before, Constellation had an agent in this area, Epizyme with Tazemetostat, Novartis had an EZH2 inhibitor. All of them have been plagued by terrible drug properties on the order, when you think about half-lives on the order of two hours, properties like CYP autoinduction, where you literally are revving up the CYP enzymes that are then metabolizing your drug. And so you get things like dose-dependent decreases in exposure. That's not the kind of things you want to see, obviously. Yeah. And in the case of Pfizer, it seems like they've gotten a compound that's got a bit better of a half-life than what's come before. Doesn't seem to have CYP autoinduction, and they seem, at least in the phase 1 results, to be seeing profound results, synergistic results in combination with an AR inhibitor. In the case of ORIC-944, this is a compound that we acquired from Mirati several years back. Mirati, just as Mark was describing, Mirati laid out a target candidate profile and said, this is the target candidate profile that will optimize a PRC2 inhibitor, and they made a bunch of compounds that target two different subunits of the PRC2 complex, and the one that won, so to speak, was what is now ORIC-944, which was predicted to have a clinical half-life of over 10 hours. In fact, in the clinic, it has a clinical half-life of 20 hours, and so we're doing QD dosing with that drug, whereas Pfizer is doing BID dosing because their half-life appears to be substantially shorter than that. We don't seem to have CYP autoinduction with the compound. So in other words, very good drug properties, good PK, good PD that we've demonstrated thus far. And we're really excited to see what we might be able to see in combination with these other AR inhibitors. When could we expect to see the first DFS data in a randomized setting for you, the two that you mentioned? Yeah. So in the spirit of walk before you run, we started the combo dosing middle of this year. So what I would say is we're not ready yet to commit to when you'll see PFS data, certainly, because the ultimate read here is radiographic PFS that is very likely in the 2026 timeframe. Really, the decision we have in front of us is just given the very strong body of evidence that's already been put up by a competitor: do we wait for two, three years to run out a phase two study and figure out whether we then want to go jump into phase threes, or do we go path two, which is the at-risk path, which is the one that we've chosen, which is that there is such a strong body of evidence that we've already seen here from a competitor. We've already liked what we've seen from a single-agent experience that we're proposing to pick dose within the next year and then start our own one or two pivotal phase 3 studies in metastatic CRPC by the end of next year or early 2026. The real key decision, I think, is actually not whether to start those phase 3s. It's really which population to focus on within metastatic CRPC. There's obviously quite a bit of different prior treatment histories in that population. Pfizer has chosen two particular populations. We'll have to vet out whether those two make sense for us or whether we go a different direction. And then just a question in terms of pipeline development. This is maybe more for Michael and for Jacob. Just in terms of bringing in assets, where do you stand there? You've done an excellent job with these two. So is there more BD coming? And Jacob, same question for you for ORIC. A little bit of a different situation since you have an internal discovery engine. Yeah, in the case of Syndax for sure. I mean, I think that business development is our lifeline, right? This is what we build around. We have a very dedicated effort to bring in additional molecules. We've been focused on targeted oncology and will remain focused there, oncology. And we tend to go earlier stage, right, and try to bring in assets that are sort of at the IND or slightly ahead of the IND stage in order to use our internal capabilities, translational medicine, and clinical development to really differentiate these molecules and develop them well. And we like to see early data, sorry, early economics that are favorable to us so we can deploy our resources appropriately. But that's been the focus, and it'll continue to be the focus. The bar is very high, as you said. We have two approved agents through in-licensing them at an earlier stage. We tend to believe that we can find these types of assets in the market, and we're deployed everywhere looking. We expect to bring additional assets pretty soon, hopefully. Yes, so we've put the pipeline together in a fairly unique way, which is we have both internal discovery as well as we will go out and do targeted BD if we find someone else has got a better or more advanced approach for a target that we like. The two programs I just mentioned to you happen to be in-licensed. We also have a PLK4 program that was developed internally, which is a best-in-class synthetic lethality program for breast cancer. So the internal group continues to work on interesting targets. We tend to try to save the best-in-class, first-in-class kind of, sorry, the first-in-class, novel-in-class targets for them, for the internal discovery group. And then externally, we've been looking for other programs. There's a lot of programs out there that are available because there's a lot of companies that have had funding crunches or just other challenges, and so it's been a target-rich environment. I think the issue is we do extensive diligence, and we're very opportunistic about these things, so if the internal discovery team doesn't bless the program that we're looking at, we'll always try to bring those programs in under MTA, do our own experiments to validate the key hypotheses, and you'd be surprised by the number of frogs you have to kiss to get one to the finish line, then we may not align on economics because we're a frugal group, as we like to say, and so we will continue to look. We have been looking actively, and we hope to bring in other programs in the coming year or two. We would agree that it's a very target-rich environment. And again, within our disease areas of focus, although we do have an internal pipeline, we're certainly always on the lookout for things that, based upon the science, have likelihood of synergistic combinations with our internal portfolio. So we're also on the search for things with that level of complementarity. We're all kissing a lot of frogs. All right. Well, on that note, we'll wrap it up. Thank you all so much. Thanks. All right. Thanks, everyone. Okay. Take care.