Exelixis, Inc. (EXEL) Earnings Call Transcript & Summary

Excellent. So welcome to the day 2 of our Biotech Focus Conference. We're very privileged to have the Chief Scientific Officer, Peter Lamb, to kind of talk about this evolution of Exelixis. So before, as you look into the history, the company was doing genetic works and then evolve into a small molecule expert and now expanding into the biologics world with either modified antibody or ADC. So I believe it's the first time that you'll give us a very in-depth overview of this effort. So I appreciate that, Peter, and I'll turn the floor to you.

Great. Thanks, Andy. Really appreciate the opportunity to present today. And as you said, I will be focusing really on our kind of strategy and progress in terms of building a next-generation oncology pipeline, kind of following on behind cabozantinib. The usual safe harbor. So I'll just start with sort of a state-of-the-company overview, if you like. We are a successful commercial company at this point based on the CABOMETYX franchise, CABOMETYX being the tablet form of cabozantinib only drug. It is currently the #1 prescribed tyrosine kinase inhibitor, both in RCC and second-line HCC. It's doing over $1 billion a year right now in global net revenue, approved in lots of countries. And most recently, of course, we had the, I think, very compelling data that came out of the CheckMate 9ER study combining cabozantinib with nivolumab, again in renal cell carcinoma. And that launch is ongoing as of this year as well. So beyond that, we have a very robust clinical program that's ongoing, again, largely focused on cabozantinib, multiple registrational trials are underway right now. A lot of that is focused on, again, combinations with immune checkpoint inhibitor. So behind that, we now have some compounds in early clinical development. The first of those was XL092, essentially our next-generation cabozantinib, which is moving through Phase I trials, both as a single agent and in combination with immune checkpoint inhibitors. And then what I'll talk about a little bit today are our 2 newest compounds or agents to hit the clinic XL102 and XB002. So from a pipeline point of view, we're developing a broad early pipeline that encompasses both small molecules and biologics. Obviously, we built on our prior history. We had a very robust discovery effort that went on during the 2000s. We have kind of legacy infrastructure there that's been very helpful in terms of reinitiating that effort. And then those efforts are complemented by a series of collaborations, both on the small molecule and biologic side, which again, I will speak to today. So we have a whole series of collaborations and partnerships that kind of span the breadth of the business on the commercial side with CABOMETYX or commercial partners, [indiscernible]. In terms of clinical collaborations, we have multiple ones of those, which again, I'm largely focused on getting access to immune checkpoint inhibitors that are being combined with cabozantinib, but also now with XL092. And then from a pipeline expansion point of view, there's a whole network of collaboration that expand those small molecules and biologics. And again, I'll go through some of these individual collaborations later. From a strategy point of view, with the commercial success of CABOMETYX and cabozantinib, this gave us a lot of options and flexibility. So we've been looking really to invest that money in multiple areas. One, obviously, we wanted to invest further in CABOMETYX and expand the indications in which it was approved. And I think half a dozen ongoing pivotal trials right now that are [indiscernible] doing exactly that. And then really the topic for today was really building -- how to build a pipeline beyond or behind cabozantinib and again, we're looking at targeting novel promising mechanisms of action or next-gen approaches building off previous clinical data. And we have 3 compounds in the clinic. They are the first results of that particular strategy. And then broadly speaking, we wanted to expand both our kind of throughput and also capabilities from a small molecule point of view. And move on into Biologics. I mean Exelixis is certainly is a small molecule company by history. But as we were thinking about doing is, we saw absolutely no reason why we couldn't effectively commercialize or develop and commercialize biologics of various types. And obviously, if you look at the evolution of oncology treatments over the last couple of decades, having access to biologics is something that's very desirable from a science and modality point of view. So I think the other thing that's kind of exciting for us and especially for me right now is we're in a position where we don't really have to box ourselves in scientifically in terms of what we do. Historically, again, we may have had more of a kinase focus in oncology. We don't really need to do that anymore. We don't need to be like an epigenetics company or a metabolism company. We can really go from a mechanism point of view, where we feel the science is most compelling from our point of view. And then with the addition of capabilities to advance biologics, we can then marry that with what we see as the most appropriate therapeutic approach. So from a partnering strategy point of view, you can look at this in kind of 3 different buckets as illustrated here. From an MOA point of view, again, we're looking very broadly at all of the science in oncology right now. And obviously, the attractive areas here we're seeing sort of first-in-class, and I think some of the things I'm going to talk about today fall into that bucket. And then best-in-class, again, agents or compounds that really build off prior clinical experience in such a way that one may advance something into the clinic that may succeed, we had a similar approach that failed in the past. So you're sort of building off those clinical learnings. And one of the agents are talking about today also falls into that category. So with those approaches that tends to give you a differentiated approach. And obviously, for preference, one would like to have a mechanism that is fairly broadly applicable in oncology. So again, there are attractive opportunities, which might be a bit more restrictive, but which are driven by a clear bio market strategy. So state wise for us, obviously, we've looked very hard at a lot of clinical assets. We're looking at near clinical or kind of late preclinical kind of assets as well. But we also spent some time looking at the innovative technology platforms, particularly on the biologics side. And then modality, small molecules, obviously, as per our history and now we've added biologics capability. That give us the capability to advance monoclonals, bispecifics, multi specifics, ADCs, which I'll talk about a fair amount today. On the parting front, we're also looking at various fusion proteins, cytokine fusions, for example, are a very attractive area right now. Now there are some things that aren't on this list, and we were not -- we have, for the moment, at least, not decided to look closely at things like cell therapies, although it's a very interesting area scientifically, I think that's probably not an area that we would want to get into at least in the short term. So sort of by discussing the small molecule pipeline and strategy. Obviously, again, building off our history here. Very exciting. We've recently completed some brand-new laboratory space that will greatly increase our -- both our footprint and our capabilities. The group we have in place right now is already fully enabled from a small molecule point of view and has all of the functions that one would expect to go from -- to enable us to find leads, to optimize those leads, to characterize from a biology and pharmacology point of view. But with the new labs in place, obviously, our capacity and our capability should increase greatly. I'm also excited to say we do have a new Head of Discovery in place, who's going to be leading this particular effort and expansion that was laid out by our own Board. So the first compound to advance from these efforts was XL092, dictated now in the clinic. I'm really not going to speak to that more today. We have a couple of additional lead optimization programs against, I think, exciting targets that are ongoing and we're now running multiple high-throughput input as well. So those assets are augmented by a couple of collaborations that we have in the small molecule area, 1 with Aurigene and 1 with a company called StemSynergy. And the thinking here is to really complement what we're doing here at Alameda and add additional bandwidth to enable us to prosecute more targets, but also to add additional expertise. These companies have either interest in chemical matter and/or experience with certain discovery techniques that we haven't prosecuted previously. So just a little about Aurigene. Aurigene is a Bangalore-based discovery company that has actually been able to advance compounds into the clinic. We have a broad agreement with them, whereby they will prosecute numbers of targets and then we have the ability to option those compounds as they [indiscernible] pretty much up to IND. They're also fully enabled from the discovery point of view, they can do everything from lead finding through to all the kind of necessary activities that one needs to file an IND. And again, they bring complementary approaches to what we're doing ourselves. Specifically, for example, they have a lot of experience in optimizing covalent inhibitors and one of the compounds I'm going to talk about today is a covalent inhibitor that already has advanced. They also have some initial, I would say, experience with PROTACs, an induced protein degradation and have some pretty interesting approaches there. This really gives us more tools basically with which to prosecute various targets. There's a little about StemSynergy. This is a small company that's really very much focused on a novel approach to drugging the Wnt/ß-catenin pathway. So I would actually characterize this as probably the major pathway under right pathway in oncology right now. It is broadly activated in a wide variety of different cancers, which is notably colorectal carcinoma where its activated through various genetic lesions, most prominently loss of APC. This then results in active beta-catenin and drive the whole program of kind of pro oncogenic processes. So they had identified some interesting early compounds that bound to a kinase called Casein Kinase 1 alpha. And when you think about kinases normally, of course, we think about wanting to inhibit those and all the kinase drugs out there in oncology are inhibitors. But interestingly, their compound was actually an activator of this kinase. And this makes sense since CK1 alpha actually phosphorolates beta-catenin and that phosphorolation site is important for mediating the degradation of beta-catenin. So CK1 alpha activity tends to be down-regulated in tumors. The approach here as well that's upregulated again, and that then should induce beta-catenin. And that's exactly what these compounds do in preclinical models and actually they see a pretty compelling efficacy in patients deriving xenograft from colorectal carcinoma. So we're currently further optimizing these compounds and characterizing more broadly their activity and toxicity. But I think if we can get these as a goal line, it will be a very novel and potentially impactful approach to the drug in a [indiscernible]. Now I'll move over to the kind of biologics side of the pipeline, which again, is newer, and this has been a couple of years being built during which time we've put in place a variety of different partnerships or in-licensing arrangements or a variety of different companies. And as I show you, this is not done randomly, but it's done really with an eye to giving ourselves the ability to advance biologics of multiple types and with a specific focus, I would say, on antibody for a conjugate. So why ADCs? Well, first I just talk a little bit similar to everybody what exactly is an antibody drug conjugate. And what is that you need access to in order to discover and advance them . The concept, obviously, has been around for quite a while. You take what I call the targeting moiety here, which typically is just an antibody, an ITT, although you can certainly use other formats of targeting moiety. And then you have a cytotoxic payload traditionally is a very potent one. And then that halo is attached to a linker. It's a linker that may cleave in the intracellar environment or it may stay attached, both levers out there. Then you need a method of joining the linker payload to the antibody. Historically, it is a stochastic nonspecific kind of format. As I'll talk about, there are some pretty elegant technologies now that really enable you to control this process. So as I said, this -- the idea behind ADCs has been around for well over 20 years, and there's a fair amount of activity in this space. It did lead to some early successes, but there was an enormous amount of failure in the field. We think getting into the clinic and not advancing either for activity or often toxicity purposes. But I think it's fair to say that there's a real renaissance in the field right now. I think over the last -- since the beginning of 2019, I believe there are 6 approvals. Many of them accelerated approvals for ADCs. And I think this has really driven off an understanding or a much better understanding of what's required to make an ADC both active, tolerable and also importantly, manufacturable. So that's what we were looking to achieve in terms of putting together the technologies that we would use to advance ADC. This is how it fits together in our current platform. From the targeting moiety antibody point of view, we've done basically 2 acquisitions, if you like, we got some -- an interesting panel of antibodies from WuXi. Recently, we conclude an arrangement with GamaMabs that get access to an antibody platform. Talk a little bit more about what I mean by that in a moment. Primarily, we're getting antibodies through an arrangement with Invenra, who I'll talk about in a moment. We have very highly enabled from a de novo antibody discovery point of view. And then coupled with that, I think our most recent arrangement with Adagene, very interesting antibody masking technology that may well kind of enable improved therapeutic indices for ADCs and other kinds of biologics. And from the actual ADC platform point of view, the antibodies from these -- from the left then feed into a couple of platform collaborations that we have, either with Redwood/Catalent and NBE Therapeutics. Both these companies have developed I think very elegant ways of in a controlled fashion, being able to link payloads to antibody, and I'll speak a little bit more about those technologies in a moment. And also, these companies have access to an interesting range of different payload types. Again, they all fall into the category of being potent cytotoxics. We've got [indiscernible], we've got various microtubule disrupting agents here and also NBE very potent [indiscernible] agents. So at the top I've got in blue here EXEL Chemistry. And as we were looking at surveying the ADC landscape, both current and historical, I think it's fair to say that the vast majority of the payload, from a mechanism of action point of view, are actually fall into a fairly limited bucket. Like I said, microtubule disruptors or various kinds have certainly been clinically validated with the auristatins probably being the lead class there. And then the other payloads DNA damaging agents of various types. And there is very good reasons to use these payloads in this particular application. But I think it occurred to us and others that there are opportunities here maybe to expand the mechanism space from a payload point of view. Quiet have seen some of that out there right now. There's a lot of interest in having immunostimulatory payloads attached to antibodies. We certainly view that as an interesting approach. But we think there are opportunities beyond that as well. And obviously, we have a deep history in small molecule chemistry. So we're going to be exploring a number of additional mechanisms going forward with our own internal chemistry efforts. So we think this platform really kind of maximizes the kind of flexibility and optionality that we have in terms of making ADCs and enables us to kind of come at it from a variety of different ways. When we put together, we have choices essentially at all points in the ADC discovery process. I think this is helpful. I think one of the lessons from the past is ADCs are not -- conceptually, it looks like plug and play, but every new one is a bit of an adventure and needs to be specifically optimized each time. So you need -- again, you need choices at every single point, so that you maximize your chances of actually getting something that's advanceable. We do have the opportunity now to kind of key off what we have done with our collaboration with Iconic Therapeutics with the tissue factor ADC that I'll talk about. This is a best-in-class approach with really keying off some clinical data that have been generated by a first-in-class tissue factor ADC. And I think there are now a lot of opportunities actually to go back and revisit targets that had historically been prosecuted as ADC, but have failed for a variety of reasons in the clinic based on better understanding and the current generation of the ADC technologies. And then beyond that, we have a lot of flexibility in terms of how we can innovate in addition to having a kind of standard monoclonal. We have the ability to make biparatopic or bispecific ADCs through the Invenra collaboration. And we also have the ability here to start to be a bit innovative from a conjugation and payload point of view as I just described. So a little bit about our Invenra collaboration. This is a biologic company Madison, Wisconsin. Has a significant history in terms of being able to make antibodies. They have multiple approaches. They have a pretty productive in vitro phase display approach, which is now supplemented by the ability to generate antibodies in camelids, specifically our packers, which is, I think, a very interesting approach and results in small antibodies. Then they have their own proprietary bispecifics platform, which is called the B-Body bispecific platform that the number of formats they're shown in the cartoon on the right, it could be bispecifics. You can then double up, if you want to get an improved kind of [indiscernible] effect for one of the arms of your bispecific or in principle, you can make a trispecific molecule here as well. So they do all the antibody lead generation, discovery and some of the early characterization of the antibody. This arrangement, we would take over basically the point where there's something that's ready to go into preclinical development. We have multiple programs with them ongoing, both antibody discovery and bispecific. Happily, the initial sets of antibodies have been delivered already to our kind of newer ADC partners who are thrilled with the panel of antibodies that they got. So this part of the process seems to be working well. Just a quick word about the GamaMabs arrangement or acquisition. GamaMabs again, a biologics company that was based in France. They have had historically a program for an antibody, which is targeting a protein called anti-mullerian hormone receptor 2. It's normally responsible for male sexual differentiation during development. But from an ADC point of view, the attractive thing here is that the normal tissue distribution in the adults is highly restricted, which is basically what you want. And then combined with that, we see significant levels of overexpression or re-exploration of AMHR2 in pretty significant tumor types, including ones that have great interest to us, renal and hepatocellular carcinoma as shown in the table. GamaMabs had actually done some -- a bit of preclinical work in terms of taking their lead antibody and making a kind of research-grade ADC from it, and some of that actually work quite nicely. So we know there's protein and specifically this antibody can work as an ADC. So I'm really calling out a platform acquisition here because there's quite a lot that comes with it. There was a clinical-grade antibody, GM102, which they had actually to advance just as a naked antibody into the clinic. Around that, there's a whole halo of additional antibodies with different affinities, different species, specificities and the like, which are always great to have. We do also have now -- we'll also be getting the master cell bank for GM102. Obviously, this puts us ahead of the game a little bit in terms of time lines for going from an antibody to a kind of ADC that's ready for the clinic. They developed a lot of associated reagents as well along with data that I think will really be helpful to us as we advance these program as many can see. Quick word about Redwood and Redwood/Catalent and NBE in terms of their conjugation technologies. Redwood/Catalent technology is called SMARTTag developed in Carolyn Bertozzi's lab at Stanford and essentially relies on a cell line expresses an enzyme or formal glycine generating enzyme. What this does is that it modifies cysteines that are present in the antibody in the context of a short specificity sequence into aldehyde, and those aldehydes can then be chemically linked very specifically to your linker payload. So this gives you, one, it's a very stable linker and two, gives you a pretty precise control of drug antibody ratio and homogeneous product, which gives you a real kind of head start from a manufacturing point of view. They also have access to a whole range of payloads, and are actually developing more proprietary payloads themselves, various classes. And we at Exelixis develop custom made or EXEL generated payloads, and could feed very nicely into that platform. So NBE also has an enzymatic based method of conjugation, it's very different, it's called SMAC, sortase-mediated antibody conjugation. Here, they just put a tag on the C terminus of the light chain, the heavy chain or both. And then on their linker -- payload linker, they have a coli glycine stretch and then an enzyme called sortase, an in vitro couple with those 2 things, a very high efficiency, and it's very specific for that linker. So again, you can control the DAR very precisely, 2 or 4. A couple of efficiencies are very high, so you got a very homogeneous product. So again, a very, I think, elegant approach to perform in this kind of conjugation. And then they have their in-house proprietary anthracycline-based payload. It is a DNA intercalator as our traditional anthracyclines, but it's not cardiotoxic. And also it's not a P-gp substrate. So I'll say it's a big advantage. And finally, from a platform point of view, just a word about Adagene, so-called safe body platform. This is a masking technology where they screen for peptides that will sit in the binding side of the antibody. And essentially block it, really reducing its ability to bind to the target. And then this peptide is linked to the antibody with a short linker, which contains various protease cleavage sites. And the thinking here is that the tumor microenvironment is known to be very rich in proteases, where it's typically in normal tissues, so you have pretty low protease levels. So in the tumor, the marking peptide will be clicked off, the antibody will then be able to bind to its target, will get internalized, and we'll have a psychotoxic effect as it's linked to an appropriate payload. However, normal tissues, this unmasking will be a lot less effective. So you really have the chance of reducing your -- of your on-target, but off-tumor effects and thereby improve the therapeutic index. And they've done some work certainly with their in-house CTLA-4 antibody showing that, that is in fact the case. So I'll just finish up here with some comments about the 2 newest compounds that we've advanced into the clinic. One is a small molecule, XL102, that we in-licensed from Aurigene last year. I'll talk a little bit more about that. And the other is really in a sense something of a landmark for us. It's really the first biologic that we've advanced into the clinic, XB002, next-generation tissue factor ADC. Again, I'll give you a little bit of data on that. And that was in licensed from Iconic also last year, and it's now in Phase I. So just a word about 102. This is a CDK7 inhibitor. It's highly selective, it's covalent, it's orally bioavailable. CDK7 has been a target of some interest for a few years now. It looks originally not as well as understood is what I'll call a traditional CDK, CDK1 and 2, for example. But it clearly has a very important role in modulating the cell cycle. It phosphorylates and activate CDK1, CDK2, also CDK4 and 6. So it really helps control self cycle transitions from G1 to S, but also G2 to M, [indiscernible] at multiple points in the cell cycle. There's also potentially a role for CDK7 in regulating transcription various areas, may well phosphorylate C terminus on on RNA polymerase II, thereby promoting various transcriptional activity. So one of the things that's attractive about CDK7 is potentially broadly applicable both as a single agent and in combination with certain agents, specifically SERMS in breast cancer and one actually don't have to slide things like enzalutamide in prostate cancer as well. Preclinically, this compound is highly active as a single agent in multiple different tumor types. What I'm showing here is triple negative breast cancer. And so we think this has the potential to be best in class due to its kind of combination of high selectivity, good potency and good oral bioavailability. Quick word about XB002. As I said, this is a tissue factor targeting antibody drug conjugate. The tissue factors interest in protease, it's is normally -- it is normally expressed on normal tissue. But it's expressed -- is sort of hidden because it's not access -- normally accessible to the blood or the plasma compartment. When there's some tissue injury or wounding that, that kind of wall is breached, the blood plasma then comes in contact with tissue factor, which invites to factor 7, and this then -- is what initiates coagulation under normal circumstances. So tissue factor is actually overexpressed in many solid tumors. Again, a whole range of [indiscernible] over expresses and there's some thinking. It's quite interesting that some of the coagulopathies that have seen in cancer patients, maybe because of the inappropriate exposure of tissue factor on the tumor to the blood. So this is a clinically validated target. There is an antibody that's in the clinic [indiscernible] genetics called tisotumab vedotin. It's shown really quiet interesting activity in cervical cancer. Interestingly, some of the side effects of the, quoting, quite a few patients were bleeding adverse events. And this is due to the fact that the antibody they use interfere with the normal function of tissue factor. It actually blocks the binding factor 7. This also uses an extended first generation link to payload. It's [indiscernible]. So we think XB002 has some significant advantages over the first-generation approach. One, we have a next-generation linker payloads, it was actually the BLA technology from Zymeworks to improve stability, improve biophysical properties, for example. And this is manifested by improved therapeutic index if you just compare it to previous generation of linkers in preclinical studies. And crucially, this antibody does not interfere with the binding tissue factor. So in multiple different assays. We see absolutely no inhibition of coagulation. So we anticipate that as an AE, that will go away in the clinic. A quick overview of the early development plans for both of these compounds. 102 is currently in the dose escalation phase as a single agent. We do plan to combine it, both with [indiscernible] and with abiraterone, multitargeting hormone receptor positive breast cancer and metastatic castrate-resistant prostate cancer, respectively. Beyond that, we have some thoughts about additional cohort expansions as a single agent as well. And again, potentially a broadly active kind of mechanism of action here. Likewise, for 002, Phase I is now now proceeding. And you can see the list there of cancers that have significant tissue factor over expression that we'll be taking a look at. So that's sort of the lightning talk of the early pipeline and some of the thinking behind us. We are advancing a robust pipeline of preclinical and now early clinical programs, with compounds now in Phase I. We're looking at potentially doing 1 more IND this year that we'll talk about later on. And we now have both small molecules and biologics in the clinic and also from a preclinical point of view multiple programs. So there's a lot going on between what we've assembled on the preclinical side. You count it out, it comes to around 25 different preclinical programs. And if you kind of look at the person power that's working on this, both here and at our partners, it's around about 200 people. So I think we've got a lot of momentum. Obviously, not all of these programs are going to be successful. That's just the nature of the business. But I think we are expecting to have multiple additional development candidates and INDs over the next couple of years. So we've been fairly busy on the business development side. I just want to leave it with the fact that we're certainly not done. There's still a lot of very intensive business development efforts and scouting that's kind of ongoing, again, looking at assets for small molecules and biologics that are either in the clinic or close to the clinic as well as kind of continued effort looking at additional platform. So I'll leave it there. I'd be happy to take any questions.

Yes. Great. Thanks, Peter. That's a great presentation. It's a side of Exelixis. I think we haven't really seen before. So I appreciate definitely kind of the unveiling of the R&D prowess that you guys have. So just a couple of questions in. And I also appreciate kind of your kind of walking through the rationale behind building out different kind of tool kits to really tackle the ADC field. So that was really helpful. So in terms of ADC, just I kind of this idea of using old antibodies or at least validated antibodies as kind of a foundation to build ADC assets. I'm just wondering if you have any thoughts about this trend, just like what's old is new, right? So we saw Eisai and Bristol-Myers using that pole receptor alpha antibody. You see that theme with being the HER2-positive field, right? So the herceptin core is now featured in 2 ADCs. So just curious if that's something that you want to build upon, obviously, with the GamaMab collaboration, that's also another approach that you guys are doing. So yes, so I'm curious if you can comment on that?

Yes, I think it is a bit of a theme. And I think it speaks to something I touched on earlier in terms of a greatly improved understanding of what it takes to make an ADC that has a reasonable therapeutic index and some additional advances from a technology point of view, from a payload point of view. So I think it's a potential fertile field in terms of going back to antibodies or ADCs or maybe if we tried 10 years ago or 5 years ago and then fail for one reason or another. I think this in many cases, you could probably put your finger on, well, they have the wrong payload. They had a linker that wasn't stable enough, and that really led to some problematic AEs or toxicities. So I think there's a number of antibodies that are almost certainly well worth revisiting with the current generation of technologies because I don't think there's anything -- also know if there's anything intrinsically wrong with the targets that they were aimed at. So I think it's certainly an interesting approach. As you pointed out, essentially, that's what we're doing with the GamaMabs platform going forward and others obviously have the same kind of thought.

Well, great. Thanks for sharing your perspective on that. So I'm curious about the kind of the small molecule aspect of things. One thing that really intrigued me from the presentation is the CK1 alpha activator. We're very used to learning about how you develop and push forward an inhibitor. So maybe share with us kind of the differences -- difference in terms of development route that you would take in terms of advancing an activator versus an inhibitor. And I guess the second part of the question is, how leverageable is that knowledge and know-how in terms of read-through to other potential oncogenic activators in the body?

Yes. No, it's a really interesting question. I mean if you look at oncology drugs broadly for the most part, the vast majority of them are inhibiting something, some oncogenic process or multiple long-age processes. And then certainly in oncology, to think less about activating things. Generally speaking, of course, if you look at other therapeutic areas, they are activated at various kinds that will be GPCRs or nuclear hormone receptor. And I think one interesting other example would be things like the [indiscernible] approach to CF, where they have some molecules that essentially restore function to a damaged protein. So in a sense, it's an activator that kind of shape small molecule chaperone type of approaches. I think it's -- in terms of advancing compound doesn't -- I guess our experience has been -- I haven't presented any specific issues from a regular optimization point of view. You do need to get some comfort with what degree of activation do we actually need. And how does that impact potentially your therapeutic index going forward. So -- but if you have those kinds of considerations with inhibitors as well, how much do I need to inhibit? Is that 80% or 95% to actually get a therapeutic effect. So those are some of the things we thought about. Things like the expression level of the specific protein that you're activating can also be crucial out of that vary between tumor types, like may well impact how active compound with this kind of approach can be in various different settings. I think whether it's more broadly applicable. I would say, look, there are certainly opportunities. There are some that are being active on now, I would say, the classic one in terms of an activator per se is really attempts to restore the function of p53 with small molecules. A number of companies are targeting various mutated forms of p53 hoping to get enough function back in the p53 to really called tumor cells there. So I think that's a potentially -- that's an example of an interesting approach in oncology that's sort of related.

Yes. Yes, very fascinating. So going back to ADCs, I think like you alluded to maybe 5 and 10 years ago, there's been a lot of mistakes. There's been a lot of just learning. And comparing kind of the environment before, there's basically kind of a handful of companies advancing these types of niche at that time, assets. Now there are CROs are going to kind of help you out. So as you get closer into establishing yourself as a leader in ADCs, how do you think about differentiation? It becomes more and more important as there's more candidates in the pipeline in a differentiation. And how do you kind of think about strategic positioning of products?

Yes. I mean, differentiation, I would say, generally, whether it's ADCs or any of the other approach we're taking is clearly a very important factor. And look, I think it's becoming even more important in the sense just because of the advances that there have been in oncology therapy as of yet. There was a time when agents could be approved on relatively what we would consider modest gains in whatever PFS or OS. That's probably still true in our few tumors where there are a lot of tumors where there are now multiple therapeutic options, and you really want to go in with an approach where you're really going to make a significant difference in that you're able to do that with reasonable tolerability, then you can differentiate. I think one of the attractive things about ADCs is that when you get it right and they do work, the activity is evident very early, and it's usually quite striking. And I think we've seen that with the wave of recent approvals. In my view is at least if you have an ADC and you're not really seeing that much as exciting even in Phase Ib, we probably got it wrong somewhere. As an aside to that interestingly, of course, that does then can often provide because of the differentiation and just the magnitude of the activity can provide approaches to accelerated approvals. And again, I think we saw that with some of the ADCs that were just approved. So it's a mixture of things, we've got to get the ADC right. Obviously, the choice of target can be differentiating. As we talked about a little bit with 002, even the exact nature of the antibody in the epitope can provide differentiation as well. And then hopefully, down the road, if we're successfully being a bit more adventurous around the payloads, that opens up a huge landscape potentially additional differentiation.

Yes. Great. Yes. Thanks for that perspective. It's very fascinating is how fast this field has transformed, like you said in the past 2 or 3 years, a lot of excitement in the field. So I want to come back to Invenra's kind of bispecific and trispecific constructs. I think both in the checkpoint inhibitor field and also in the cell therapy field, we're kind of expanding the focus from T cells to other types of modalities. So maybe you could comment on just how modular and generalizable these bispecifics or trispecific SAR in terms of just kind of thinking beyond T cells into other dendritic cells or NK cells and things like that?

Yes. I think the principle has sort of been established with T cell engagers, you have one part of our bispecific, which anchors you to the tumor or the stroma and then another part that has a functional impact on an immune cell. I think that concept is, in principle, completely transferable. I also agree that I think it's a very attractive way to go. I think ways of trying to modulate components of the innate immune response in the tumor microenvironment is likely to become a very important area for additional therapeutic development. So in principle, yes, you can do exactly what I say we can come in with the things that might help activate NK cells or kind of modulate the myeloid phenotype in so those are, in fact, approaches that we are taking with Invenra in a general sense. So we'll see. I mean it does provide some challenges, and there are certainly challenges in advancing T cell engagers as well. Some of them are on the preclinical side in terms of having the appropriate models and kind of trying to get things to work in mice, you need -- sometimes you need pretty complex humanized mice to really get this -- see some kind of effect. So I think that's always the challenge in the field it's likely to be with as we move into other immune cell type engages as well. And then, yes, I think it's a matter of understanding that we're going to need to continue to increase our understanding of what the immune cell environment is in specific tumor types at specific stages following various treatment regimes. To understand best, I think, were to test out these new kinds of immune cell engagers. So there's certainly some work to be done there, but I think it's potentially a very exciting area and certainly one that can be approached with the Invenra platform.

Great. So maybe one just quick question to wrap up. I think you alluded to this, Peter, basically about the fact that in the past -- if you look at the past 6 ADC approvals, they were all based on accelerated approval pathway. So response rate, like you kind of mentioned in your earlier Q&A, which is basically you see activity, single agent activity right away in a very rapid manner. So just curious about your view in terms of the regulatory path. Is that kind of a standard now for all ADC players out there, basically, that's kind of where you should focus on? It's a kind of a rapid to-market strategy.

Yes. I think generally, yes, I'll just speak for files. I mean, I think generally, with ADCs or any of the other assets that we're advancing. I mean we certainly look at whether there's a potential to have an accelerated approval based on the usual kind of expanded Phase Ib that have become very common in the oncology space. And again, if you have an agent that has sufficient activity or specifically targeted to the correct patient population. That's normally what gives you the opportunity to go ahead and how that level of activity that you can do it. So it's something we think about a lot. It's something that we think about in the context of all of the programs we're advancing, whether they're ADCs or small molecules. As I said, I do think it's striking that I think the characteristics of ADCs where when they're active, they're really active, it sort of lends itself to that particular strategy. And we've seen it a lot. So it's certainly something we're thinking about as well.

Great. I think that's all the questions I have. Peter and Susan, thank you very much for making this event available really appreciate your presentation and had unveiling this part of Exelixis, both for us and our clients. Yes. So with that, I'll close this session, and thanks again.

Thanks, Andy.

Thank you, Andy, and thanks to William Blair for hosting it.