Arvinas, Inc. (ARVN) Earnings Call Transcript & Summary

Hi, guys. I'm Ellie Merle. I'm one of the biotech analysts here at UBS. Very happy to have Arvinas here with us at the UBS Healthcare Conference. With us from Arvinas is John Houston, President and CEO; and Ian Taylor, Chief Scientific Officer. Very happy to have you guys here today. And with that, we'll jump right in.

And maybe high level, can you talk about the benefits of a PROTAC versus a small molecule inhibitor and the vision that you see for your platform over the next 1 to 2 years as well as long-term and the applications of PROTAC?

Ellie, great to see you. Yes. I can start off, and I'll hand over to Ian to give more of kind of the rationale for differences between degradation and inhibitor approaches, which we do believe drives a lot of the differential biology that we see preclinically and now we believe seeing in the clinic. Maybe just to start off, just, again, a brief statement of where we are as a company today. We have 2 programs in Phase II in the clinic. ARV-471 is the first one, which we believe has the potential to be a best-in-class estrogen receptor targeting therapy. So far in the clinic, we've seen profound AR degradation, good tumor responses, outstanding safety profile. And our VERITAC Phase II dose expansion trial is ongoing. 110 -- ARV-110 is in patients with late-stage metastatic castration-resistant prostate cancer. And the ARDENT study is the dose expansion stage and also progressing really well. And beyond those 2 lead programs, we have another program, ARV-766, which is one of the AR degrader, which will be in the clinic in the first half of this year. Then around 20 other programs in oncology and neuroscience with the idea of getting to one IND per year from that pipeline. And we should have another 4 or 5 IND filings between now and the end of 2023. So it's an exciting stage for the company. A lot of excitement in the protein degradation space in general, and we are happy to be still very much in the leading edge of this technology and this platform. And we'll be happy to talk about all aspects of our pipeline today. But with that intro, maybe Ian can give you the answer to the specific question you're asking is differences between PROTAC's protein degradation and the inhibitors.

Yes, sure. Thanks, John. Yes. We always look at how PROTAC's protein degradation in general have advantages over small molecule inhibitors. Obviously, one obvious one is strong domain. Small molecule inhibitors inhibit the protein in the enzymatic function, whereas the PROTAC fully degrades the protein. And that's important because a lot of targets have more than just their enzymatic function they serve. As, for example, scaffolding protein. So other proteins bind to them and that adds to their activity. So by just inhibiting the enzymatic function, you're leaving behind that scaffolding function. But by degrading the protein, you're not only moving enzymatic function, but also the scaffolding function, which is going to be a big part of their -- of this activity. We have a number of programs that fit into that category. Of course, also, there's the factor of the iterative or catalytic mechanism of action of the PROTAC. That leads to an event-driven pharmacology. So one PROTAC molecule can degrade multiple. We've measured up to 200 molecules of a protein by a single molecule of PROTAC in some cases. And so you have an event-driven pharmacology, the event being just a binding and then the formation of the trimer complex between the E3 ligase and the target protein that leads to degradation as opposed to occupancy-driven pharmacology, which is an inhibitor. So you need an excess amount of a small molecule inhibitor to have antagonistic inhibitory activity. Actually, PROTACs works [ subs to a key metric ] because of that iterative mechanism of action. So -- and we believe that, that leads to requiring a lower exposure of a PROTAC relative to inhibitor, which can, in the long run, lead to fewer adverse events in the clinic. And of course, you have the whole undruggable class. The undruggables are labeled as such because there's been no small molecule inhibitor that can bind or function. A PROTAC -- the warhead that binds the target protein can bind anywhere. It doesn't have to bind to an enzymatic pocket. It can bind to an allosteric site. It actually doesn't even have to bind very tightly because really what you're trying to do is optimize that trimer complex formation, that protein-protein interaction between the E3 and the target protein. And that's the key parameter, not so much where it's binding or how tightly it's binding. So that's another advantage that you can get into that undruggable space, which has been estimated upwards of 80% of target proteins, are in that undruggable class.

Absolutely. That's a helpful overview. And just in terms of the differences between different types of degraders, such as particularly molecular glues versus heterobifunctional, can you talk about how these are different and what the potential advantages and possible disadvantages are of PROTACs versus molecular glues?

Sure. So PROTAC is a heterobifunctional molecule, as you mentioned. So there's 2 Ms to the molecule held together by a linker, one binds to your target protein of interest and the other to the E3 ligase. So you're dragging an E3 ligase to a target protein and actively forcing into ubiquitination and then, ultimately, degradation. Of course, the linker is a key part of the molecule because it helps orient the target protein in the E3 ligase and really optimize that interaction, optimize the ubiquitination, optimize degradation. A molecular glue is also sometimes called as a mono tech. Really what it does is it binds to a protein, in this case, for degradation would be an E3 ligase, changes the confirmation of that protein basically exposing new surfaces to which a protein, which normally wouldn't bind to it and be degraded, now does bind and is degraded. And so the glue is sort of helping bring those 2 proteins together, mainly indirectly, mainly by causing those new protein-protein interactions. So that's a very -- the thing with the glues is it's hard to prospectively design them to target a protein of interest. Usually, you figure that out is from a phenotypic screen, you kind of figure out the boots dropping backwards, what protein is being effected. That's certainly was the case with the IMiDs, thalidomide, pomalidomide, lenalidomide. And so I think it's still -- I think people recognize it's still very hard to design a glue prospectively for target as opposed to a PROTAC. You know what target you want to degrade. You have a warhead to it. You have an E3 ligase ligand. As I said, you bring the ligase to it, and so you can prospectively degrade. I just find it a much more efficient, if you will, form of degradation than a glue. I think glues are smaller, and that's what people point to and say, "Well, they're going to be more drug like." But I think our data with ARV-110 and ARV-471 has clearly demonstrated that PROTACs can be just as drug like as a traditional small molecule. Yes, they're on the larger side and they're larger than glues, but that doesn't mean that they can't be drug like. Again, we've shown it for 110 and 471. So that advantage/disadvantage argument I don't really buy into because of the clinical data that we've shown. So again I think PROTAC is just again more efficient because you can more prospectively go after degradation as opposed to glue approaches. Maybe in the future, there will be a breakthrough there. But right now, PROTACs clearly have the advantage in that area. And that's a big advantage, in my opinion.

Absolutely. And then just thinking from a kind of platform perspective and all of sort of the years of know-how that you guys have built up and your internal work around assays, the unique PK/PD. Can you describe some of the sort of the processes that you've developed? And I guess, why it's so hard to make a PROTAC? I mean, in principle, it sounds like a simple concept, but I think in application, it's proving to be much harder than maybe the concept suggest. Maybe what are the things and internal know-how that you've developed? And if you could talk a little bit more about that.

Yes. So we've done a lot of investment into the platform, if you will, of PROTACs. I guess I'll take -- to change your question around a little bit. I think actually, what you've seen because so many people are doing so many companies, academic groups -- a lot of people can make PROTACs to degrade. That part actually is not difficult. What's really difficult is turning them into drugs, right? They have the drug like properties that can work in vivo, oral bioavailability crossing the blood-brain barrier. That's when we spent a lot of time really from 2015 onwards in particular focusing on to turn what Craig Crews used to call a molecular chemical biology parlor trick into drugs. And so to your point, we had to revamp some of the assays that are traditionally used in screening cascades because there's no 2 ways around it. PROTAC are class IV -- BCS class IV molecules, mean they have low solubility and low permeability. And so when we're doing permeability assays, we had to revamp them to tailor them to PROTACs. The traditional assays weren't really being informed. Same with plasma or protein binding assays, these traditional ADME/PK type assays that are more designed for traditional small molecules. We had to revamp those. The PD, as you mentioned -- I mean honestly I spend a lot of time on oral bioavailability. But because of the catalytic activity, the inter-mechanism of actions I talked about, you do get extended pharmacodynamics relative to the plasma PK. And so learning those parameters, that extended PD, making the PK/PD correlations to efficacy required a different way of looking at, different calculations, different data sets, all the things. In our deck, I don't know if you have available, on Slide 7, I believe that we have the PROTAC discovery engine laid out. And so we have those 2 bucket -- ligase selection ligand identification, but we also have rapid PROTAC design where really we focus on optimizing the zone of ubiquitination. And for a while, that was more empirical, right? Make a compound, see how it degrades, make another one, sort of the traditional medicinal chemistry SAR, if you will. But over the last several years, in particular, we've gone into more computational model, CAD approaches to be able to model that. Using a new generation of linkers that connect the molecule, again to -- into the molecule, again, to optimize that zone of ubiquitination using artificial intelligence now to do that. These are all advances that we've done to make the process more efficient, more prospective, more predictive, which E3 ligase will pair, which with target protein better, all of those things we've built in to make that process more efficient again around PROTACs. And then, of course, the drug like properties, the oral bioavailability, the blood-brain barrier penetration I mentioned as well. Those are all things that, of course, we view as Arvinas know-how that we built in, again, to make the PROTACs -- these tool molecules into drugs.

Great. That's helpful. And then a question for John, just on kind of strategy and prioritization. There's obviously a lot of areas of potential growth for the field across novel E3 ligase identification, novel ligand identification for the "undruggables" or even just creating PROTACs for the druggable molecules where you see that you can improve on a traditional small molecule inhibitor. How do you prioritize investment across these various sort of spectrums and then thinking about kind of the strategy for the company and further growth?

Yes, great question. Yes, clearly, the company has been highly focused on the oncology part of our pipeline and over the last few years our neuroscience pipeline. And as you could see, you could argue that's a nice balanced risk within our portfolio. We know that neuroscience is tougher. There's a higher risk there. Oncology, we believe, that there's a lower risk, but still a huge unmet need. So we balance the portfolio that way, and we realize those were good choices to make. We've added to the oncology piece by having some immuno-oncology programs. And we're also looking at aspects of the neurological rare disease. So we're adapting even that core strategy. But we are focused on those areas as being the ones that will drive the biggest value and where we think PROTACs can have the biggest impact. Having said that, we also have a significant part of our resource focused on continued platform development. Certainly, the platform we started the company with has moved on considerably since 2013. And that PROTAC discovery engine now has quite a significant ligand discovery capability, ligase evaluation capability, as Ian talked about, artificial intelligence and other aspects of lead optimization that allow us to put in these drug like properties into PROTACs with the necessary requirements, either for oral bioavailability or blood-brain barrier penetrants. And that capability has grown and grown. So it's a balance of significant portfolio activities to keep building the topline in score diseases and also building the platform out. But we do keep an eye out for other potential scenarios in different diseases. I wouldn't necessarily say it was opportunistic, but if there's a target or a very obvious disease area that we think could be positively impacted by a degrader, we will assess that. And also potentially looking at partnerships as it relates to that. We do have some significantly strong partnerships with Genentech, Pfizer and Bayer. We have a joint venture with Bayer also in the agrochem space, just to show you the facile nature of the technologies that can be applied outside of human disease. So we're looking at all the different ideas and opportunities. We want to keep laser-focused on our pipeline, but there are opportunities outside that, that we think we could still grow into over the coming years.

And on the E3 ligase, I'm sure you're probably not going to answer this, but I have to try. How should we think about when we could potentially see an asset with a novel E3 ligase sort of enter IND-enabling work or potential IND filing? I know you guys have spoken a lot about your E3 ligase identification work. Any more color that you can give us beyond that.

Well, I mean, we've now published the first 2 structures for ARV-110 and ARV-471. And you see certainly with those 2 programs, we hijacked to serve on ligase, which I think for some people was surprising. I think the guessing was it was VHL. And when we look at our portfolio, we do still have an array of different ligase that are again deployed. So the answer to the question will be, yes, over the coming years, you'll probably see INDs with different types of ligases. Which ones they are and when they are occurred, we won't say. But we're certainly focused on using the appropriate ligase for the particular target and for the typical -- the type of properties we want to have with the molecule as well. So yes, we're still taking that more diverse approach to ligases than just one single ligase.

Got it. And I see a question from the audience here just in terms of the partnerships with Roche, Pfizer. Can you talk a little bit about sort of the progress that you've made with those collaborations, say, in the past couple of years since those were formed?

Yes. No, we believe they're being very successful. We're not able to talk about them under the contractual agreement between the different companies. They don't want us to talk about the targets or how well they're progressing. The only way we can say they are progressing is pointing to our Qs and showing the milestones that we're getting from those different companies. So programs are progressing, and we are very -- we believe very successful. And finally Genentech added to the collaboration, that was set up in 2015. Then they added again to it in 2017. And Bayer -- we did a discovery deal with them, but also we did that joint venture. So I think, yes, these companies have been very interactive, and it's a very good relationship with them.

Got it. That's helpful color. And then just like big picture, maybe for Ian or -- I guess, I'll turn it to both of you guys for this one. Just in thinking about long term, the mechanisms of resistance against degraders, I know there's some emerging literature around that. How are you thinking about what the key drivers of resistance could be and how you're sort of developing your strategy around that?

Means the experiments on it. So over to you, Ian.

Yes. Sure. So we've certainly generated cell lines that are resistant to 471 and 110 by treating -- in the lab by generating -- by culturing them with increasing concentrations, the traditional way to do it long-term, to try to identify mechanism of resistance. I think there's going to be multiple ways that cells become resistant to PROTACs. And they will become resistant to PROTAC cancer cells. Very good at avoiding and eluding therapeutic approaches. We certainly are aware of the publications that have identified resistant mechanisms through the E3 ligase seen through mutation or translocation because a lot of those publications have been using some of our PROTACs that we had published on earlier, particularly with BRD4. We -- so I suspect it's going to be a -- motor of the E3 could get -- methylated expression gets turned down. Gene gets rearranged, et cetera. In our 110 and 471 resistant cell lines, we've not seen that yet. We've not seen the ability of the -- to grade ER and AR to be lost or other proteins to be lost in the cell lines that we've used to generate those resistant cell lines. Basically, we're still characterizing what their resistance [ basically ] are but they seem to be more ER and AR independent. So -- but it could be that in other cell lines that we -- if we were to create them, we would see an effect on the E3. So that's still be determined. It makes sense that the cell could figure out which E3 ligase is being used. As long as it's not essential for growth, that would be something that could then be shut off and prevent degradation. That's part of our criteria for selecting new E3 ligases. To your question about strategy going forward, which ones, particularly in the cell types that we're interested in may be essential to growth. And therefore, the cell couldn't shut it off and still survive. So those are the kind of things we're looking at. Those are the kind of things we're characterizing in our current resistant cell lines. And like I said, it's likely to be multifactorial in the long run. No surprise there.

And I think just in terms of the ability to measure degradation and sort of the proof point, I know we've seen some initial biopsy data, both from ARV-110 as well as 471. But I guess from a mechanistic and kind of platform approach perspective, how do you measure the difference between degradation versus inhibition in patients and particularly across various tissue types and in terms of being distributed across the body?

Well, I mean, in patients, that's the practical matter, that's really tough to do, right? We've been using for a gout indication, obviously, tumor biopsies pre and post treatment. And with 471, we've gone to the AQUA method, the quantitative immunofluorescence over the traditional chromogenic staning just because that method is clearly very subjective. It requires a pathologist to score the samples and provide an H score. Usually you require multiple pathologists because you will get pathologist to pathologist differences in the scoring. Whereas with the AQUA method, it's more objective. It's a computer -- basically a computer algorithm scoring based on the quantitative immunofluorescence. So it kind of takes that pathologist out of the equation. In patients, from a practical standpoint, especially in a clinical trial, especially in the Phase I dose escalation, where we've generated our samples, you're really just trying to -- the degradation is always going to be a secondary endpoint because you're trying to get to the recommended Phase II dose. You're trying to identify any DLTs, the maximum tolerated dose. You don't want to slow down that aspect of the trial to get -- to measure degradation, which is why we always make the paired biopsies optional, at least during the escalation phase. So that's -- from a pragmatic and practical standpoint, I mean, that's where we're limited. It's going to be impossible to show a difference between degradation and inhibition in patients because you actually have to run that trial. And that's just, again, not practical. So unfortunately, the tools are limited. We're doing -- we're using the ones that we have to the best of our ability in the context of a clinical trial. And that's -- in terms of tissue distribution, again, we can do that much easier preclinically just by using mass spec analysis or even radio label studies down the road. But again, that's really hard to do in the human clinical trial setting. So we're doing the best we can with what we have. And I think the data that we have so far with 471, in particular, because we've gotten more paired biopsies than we did in the prostate cancer trial, just as we predicted, has shown that in all the samples that we have, ER levels are going down, again with that quantitative assay. So we're very pleased with that, but recognize that really going beyond that with any other methodology is really not practical.

Yes. And just a follow-up on that. I mean, there was some variability across patients. I understand the assays and sort of the availability of the data, limited also dose escalation study. But I guess how are you thinking about the feasibility of getting up to that like 90% degradation level in all patients? Is that something that's feasible at higher doses? How are you thinking about the ability, both, I guess, with 471, but also broadly, when you think about the PROTAC platform?

Sure. Well, we've already shown that we can get 90% degradation. Even for 471 trial, even at the lowest dose of 30 mgs, we had paired biopsy that was 90%. Not all of them are there. That's where our averages was around 62%. But we feel like as we get more samples, go to higher doses, that number will go higher. And there will be certainly some tumor -- paired tumor biopsies where we have it, 90%, probably some it will be lower. Whether we need to be at 90% or lower, that's part of the experiment we're doing in the clinical trials. We've certainly seen preclinically that 90%-plus gives us tumor regressions. That's why that was our target, but we certainly knew that having lower levels of degradation also led to robust tumor growth inhibition. So where that range is required to be in humans is something that we're going to try to see as part of this trial. Again, with limited sample numbers that's just going to be inherent to a trial design, I'm not sure how much we'll -- how rigorously we will be able to determine that, but we're certainly going to try. Certainly, there's going to be differences in ER levels, patient-to-patient, tumor-to-tumor. That's been shown historically with all like, for example, the fulvestrant studies that have been done, looking at degradation. And we see the pretreatment range of ER levels -- it's all over the map. And then what you hope to see is that the -- whatever level you're starting with, you're getting down to much lower levels post-treatment. And that's what we've actually shown when you look at our -- when we look at our graph with just a handful of samples we've shown, and then 4 out of 5 of those samples are getting down close to the lower limit of detection. In one case, I think, basically at the low -- really close to the lower limited section of the assay. So that's kind of the trend. And we want to see that for as many samples as possible regardless of what the starting levels are. That, to me, is really the proof of mechanism that will be the strongest. And obviously, that hopefully will come along with higher percentage. But I wouldn't focus so much on the percentage. Where is it getting down? It's getting close to the lower limit of detection. I think that's the goal, right? The goal is to eliminate as much ER in the tumor as possible. And of course, what tumors you're looking at -- fine needle -- core biopsies. There will be tumor heterogeneity, of course, not just with the amount of tumor that's in that particular biopsy from biopsy to biopsy, but the amount of ER in those tumors. So there's always going to be some variability just from the biology of a tumor, right? A human tumor is obviously much different than a cell line xenograft in a mouse, whether it be a PDX or a cell line xenograft. That's just the case. Those are much more homogeneous than human tumors. We're not the first to show that. So that type of variability is there. But we feel like we've taken the variability out of the assay with the AQUA assay with the quantitative immunofluorescence.

Got it. Those are very helpful. Yes, I think that's a good point around looking at not the percentage, but sort of the ending level and what proportion are sort of near that below level of quantification. I guess maybe just pivoting to the neurology side of things. Maybe, John, can you comment from a strategic perspective, maybe why you find neurology a compelling area for PROTACs and sort of your vision there in terms of target selection and how you view sort of the validation of some of these targets like mutant huntingtin, tau, alpha-synuclein?

Yes. Yes, so we're very excited when we saw some data -- now several years back, where we -- the team showed they could degrade tau. This was in a preclinical setting. They were able to show degradation of tau. They then injected that PROTAC into the brains of mice and show you could degrade tau there. And that was the first sign that maybe we could move beyond just an oncology focus into an area like neurodegeneration. Since then, the biggest challenge was can you make PROTACs being penetrant, and that problem was cracked several years back by our team. And now that gave us confidence that we really could build an exciting portfolio of neuroscience opportunities. So why neuroscience? Yes, once you've broken through that kind of technical barrier, then you're right into, are these the right targets? Are these the right diseases? And obviously, in neurodegeneration, massive, massive unmet need as you can tell from the portfolios of many other companies in the space. There's been significant amount of failure or disappointment that particular types of drugs haven't managed to show the effects that they hope for. And the question you have to ask, is that because the targets were not the right target or the targets weren't drugged appropriately? If you take that latter view, we do believe that PROTACs will actually drug these targets very effectively. They will get into the brain. They'll get into the right compartment of the brain. We will show that we're able to bind and show some degradation in these settings. And then it leaves a big question. So when you do that, does targeting tau or alpha-synuclein or huntingtin -- does it actually give you a clinical benefit? So the literature has been saying all along that these are the right targets. I don't think they've been appropriately drugged. And I think our next phase is showing that you can drug these targets appropriately. And then we'll see how well they translate into the clinic. So there's still a risk associated with that. We are mitigating some of that risk by also not jumping right into Alzheimer's or not right into Parkinson's, but looking at tauopathy-based diseases or synucleinopathy-based diseases where we know those particular diseases are driven by that dysfunction in the target. And if you can see a clinical outcome there, that will build confidence that you can move into these more broader diseases. But having said that, Alzheimer is clearly a heterogeneous disease where there's multiple targets or mechanisms in play. So the idea that some kind of monotherapy was going to be a cure was naive. I do believe you need a multiple combination approach in that space. And so we'll be thinking about that as we move forward. So neurodegeneration, neurology, in general, huge opportunity, massive unmet needs. A real requirement out there in many, many populations for drugs. And we believe that PROTACs really could have a position to play over the next several years. So INDs, hopefully, over the next 1 year to 2 years that will get us into that position to test the theory that PROTACs can do that.

Got it. That's exciting. Makes sense. And then just in terms of like the challenges of crossing the blood-brain barrier with the PROTAC being sort of a larger small molecule. I guess, what were some of the key ways that you were able to achieve this without sort of giving away your secrets off, but -- and maybe some of the risks that we should think about as you move from preclinical into the clinic?

Yes. I mean, obviously, people want to know, in the same way that when we worked on making these larger than average small molecules orally bioavailable, how did you do it? Well, we can't. That's a secret. And there's more people understanding how to do it. Similarly with brain penetrants, we've worked on the design of PROTACs to allow us to do that. But yes, we haven't shared any of that. No doubt over the next few years, that will come out, too. But yes, that technology breaks out -- breakthrough we made several years ago, so we'll be able to build on that insight quite considerably. And yes, I'm not going to tell you everything. Obviously, it's a competitive advantage. In terms of challenges, the challenges are going to be like the same challenges that every company that's in neuroscience space find. Like getting into the right patient population, understanding whether this is truly a monogenic-based disease, which most of them aren't. Well, what are the true clinical outcomes you're trying to get to prove that your molecule is having an impact. And so we're taking in that all into consideration as we move molecules to the IND stage and think of the patient populations that we move into for proof-of-concept.

Got it. Makes sense. And then turning, I guess, just to the clinical programs. Maybe just big picture for ER and 471, obviously, a competitive space. A lot of players in development in both the oral search space but also breast cancer requires kind of much larger studies to reach sort of the earlier lines of therapy. How are you thinking about this from a development perspective in terms of segmenting different patient populations as well as longer term potentially a strategic option in terms of collaborations or things like that in terms of funding these probably potentially large trials?

Yes. I'll touch on that and then I can hand over to Ian. So clearly, certainly in the -- with ARV-471 in the metastatic breast cancer space, we recognize it is a highly competitive space. When we moved into this whole area several years back, the view was fulvestrant works as an ER degrader, not a targeted ER degrader like a PROTAC. But the mechanism of degradation had already been proven. The question was, could you come up with a better version of a fulvestrant that was oral, targeted degradation? In other words, getting more degradation. And the principle being, if you get more degradation, you're likely to get maybe -- likely to get a better clinical outcome. And that was our major driver. And of course, then there's all the SERDs that are out there, which is a mixture of degrader, destabilizers, disintegrators. They'll have different names in reality, but they were all ahead of us. And the question was, could we come up with a profile of a PROTAC degrader that would be seen as potentially best-in-class? And we believe with 471, we've got that. Our trials are in really late-stage patients. They are 100% post CDK4/6, which is very different from a lot of the trials you see with SERDs. Over 70% of them are post-fulvestrant. And they've had a median maybe of 4 therapies. So it's a very late-stage population. We went into the trial, really trying to show a safety profile. We had no expectations of seeing efficacy or a CBR rate that was particularly high. So we were incredibly gratified to see not only a safety profile that looked best-in-class, but also efficacy in terms of our CBR rate and responses, which was remarkable for such an early trial in a early stage patient population. So our game plan is to move the -- obviously, the program forward aggressively in Phase II, and we're also looking at combination studies. We've already initiated a combination study with palbociclib, and there'll be other combinations that we can look at. And really to aggressively pursue the program, so we can at least close the gap in some of those SERDs, but also hopefully maintain that best-in-class profile. So when the time it get to the market, it's going to have a significant position to play.

Got it. And then maybe just from a scientific perspective, what drives your conviction that ER degradation would be superior to, say, other forms of ER modulation such as inhibition? And how do you think about this and the key advantages of a PROTAC?

Great question. And Ian, can lax, lurk about degradation and inhibition, which our PROTAC does.

That's right. We've now published the structure of 471. And so people can see that the warhead is lasofoxifene, which is a SERM essentially, so it has antagonist activity. We've certainly shown that when we make a cereblon dead. So it's not able to -- the molecule is not able to engage cereblon and not able to degrade. You still see very potent antagonist activity. And when we compare that to degradation, proliferation, other assays, we see a clear advantage of degrading over simply antagonizing. And we've seen that with -- against other antagonists as well. We've certainly seen it relative to fulvestrant in vivo relating to the points that John had made. All of our in vivo experiments have gone against fulvestrant. So we know that not only is degrading better than inhibiting from those experiments, we also know that more degradation is better than less because we've always been superior to fulvestrant, both in terms of degradation as well as tumor growth inhibition, really tumor regressions. So all of that data, both in vitro and in vivo combined, really told us that the better degrader will win in the long run. And so far, from our studies, comparing against the other SERDs that have reported data at the same stage of development as we are, Phase I dose escalation, 471 does have the best degradation profile as well as the safety and the clinical benefit rate, et cetera. So that's why we really have conviction that, in the long run, even though we are behind and we're working hard to catch up, the 471 -- that profile will continue. And then when we get into earlier lines, first-line metastatic combination with palbociclib or CDK4/6 inhibitor, getting into the adjuvant setting, that will win out. In the long run, 471 as the best degrader, will be the endocrine therapy of choice.

That's helpful. And then maybe just to think about the second half of the year update for 471 to the extent that you can comment, how should we think about what we can expect in terms of patient numbers, data update and maybe what you're looking for? And I understand if you decline to answer, but I figured I'd try.

No, no. I mean, in fact, if you look at Slide 5, it gives you a sense of what we're going to be talking about through this year for both 471 and 110. So if you look at the rest of this year, for 471, we'll actually share the end of Phase I complete data by the end of the year at a scientific conference. We were still in dose escalation as we started Phase II, so we'll give the full story on the Phase I data at some point later this year. Our game plan is also to share interim data on CDK4/6 combo study by the end of the year as well. In terms of initiations from that program, initiate a window of opportunity study and other potential combination studies. So a fair amount of activity around 471 in 2021. And for 110, similarly, share the completed end of Phase I data at the end of the year. And then an interim readout of our ARDENT Phase II study at the end of the year, and we'll also be initiating a combination study for 110. And as I mentioned earlier on, ARV-766, we'll initiate Phase I in the first half of this year. So quite a busy time in terms of data release in the second half of the year and initiations. If you look at 2022, you see similarly, based on where we're -- how we're progressing, we'll have interim VERITAC Phase II data for 471, completion of the CDK4/6 combo data and we'll share any interim data we have for other combinations. At that point, we'll also share the full ARDENT Phase II data and, again, any combination data. And also by then, 766 should be in a position where we can share Phase I data and initiate Phase II for that program. And also by that time, Ian, will be -- will have pushed a number of the rest of the discovery pipeline over into INDs. So yes, quite an exciting period of time coming up over the next half year and year with data releases.

Yes, pretty exciting time. Just a question from the inbox on sort of the safety profile of PROTAC. I guess, given that we're sort of "hijacking" a natural part of the body protein disposal system. How are you thinking about the safety of this long term, particularly given some of the resistance mechanisms that might develop?

Yes. Ian, do you want to take that? I know we've discussed this in little bit of detail.

Sure. No, I mean -- first thing is, yes, certainly, we're hijacking the different disposal system with a specific E3 ligase in the case of 110 and 471 cereblon. But again, because of the potency of the molecules, and they're quite potent, and the iterative action, we're really just using a fraction of the total cerebron that's in the cells. Certainly looked at that preclinically and really a fraction of the capacity of the ubiquitin proteasome system. So the mere effect of hijacking really even theoretically is not a problem. But certainly, we've done the appropriate tox studies in rats and dogs, 28-day GLP tox studies. And really, in both 471 profile was remarkably clean, 110 was also very clean. And there was nothing that was a PROTAC-specific toxicity that showed up. Most of them were target mechanism-related. Particularly with 471, basically, reproductive organ toxicities, basically exactly what you would expect if you're degrading ER. And there was no overlapping tox between 110 and 471. And now everyone sees that the ligase we're using is the same for those and yet there were -- the toxicities were actually completely different, mild in general. So none of the data that we've had -- and certainly, in the clinic, we've seen, again, similar types of observations, really no toxicities are really overlapping that would point to us. That's a protein degradation related toxicity. That's a cereblon hijacking toxicity. That's a PROTAC toxicity just in general. None of that has emerged. So I think that's going to continue. I mean, obviously, we're the first PROTACs in the clinic. So early data, other data can emerge. But I think it's -- there's really nothing that we've seen that points to a PROTAC generic -- general toxicity. And in terms of resistance, again, we'll see how that plays out in the clinic. We're going to try to collect the appropriate samples to try to characterize the mechanisms of resistance. Again, that's not as easy to do in human trials as it is preclinically with cell lines, but we'll look.

Got it. That's very helpful. I think we have 3 minutes left, so I'm going to open it up to the audience if anyone wants to put some Q&A in the question box here on the webcast. But meanwhile, just taking a step back. I mean, how do you, I guess, full strategic, but also kind of maybe execution on target-based. In a couple of years from now, where do you see the company in terms of developing and potentially commercial resources as well as sort of investment in R&D? And what are your priorities in terms of as a company where you want your focus to be, particularly as some of these assays head towards later stage and maybe some of the genetic subtype populations?

Yes. So in terms of the company plan, obviously, we want to grow our business to being a fully integrated biotech company. We have a pipeline that is growing based on our platform, and we have the opportunity to take a number of those assets all the way through to launch into the market. Well, if all things go well, over the next 2 years, we'll be in a position where we have line of sight to thinking about what does a commercial plan look like and building that part of the organization. If you get to like within 2 years of a launch, you need to really start mapping out that whole organization. Bizarre to think about it because just a few years ago, we were preclinical, a private company, and now we're planning what you need to do to think about commercial plans. But that's the exciting part of what a pipeline like this can do and a platform can do. So both 471 and 110 and the rest of the pipeline behind it, we are really using that to scale the organization. We're over 215 employees here in the New Haven area. We just announced that we'll be moving into a new building in New Haven where we will -- it will really allow us to expand significantly. So we're building for the long term, and we have a great substrate to do that.

Okay. Great. Well, with that, I think we'll close it off. But John, Ian, thank you so much. Great speaking with you. As always, I could ask you questions all day, but unfortunately, we have only 45 minutes to ask them. But thank you so much. Really appreciate it, and thanks for attending the conference.

Thanks, Ellie. Good talking to you.

Thank you, Ellie.

Thanks, Bye.

Bye, bye.