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

Great. Thanks, everybody, for joining. Thanks for your participation in our Biotech Focus Conference this year. We're really pleased to have the management team of Exelixis, Mike Morrissey, CEO of the company; and Peter Lamb, CSO of the company, to discuss current progress and what's in store for the next 6 to 12 months for the company. Before I begin, I need to let you know that for a full list of disclosures and conflicts, please visit our website. So Mike, it's been a great year for Exelixis, right? CABOMETYX continues to launch well in first-line RCC, catalyst-rich second half, which we'll touch on the 3 Phase III trial readouts. Strong balance sheet, right, $2 billion in the bank. So maybe you could provide us with a quick snapshot of where the business is. And yes, we'll kind of dig deeper into the pipeline and catalysts.

Sounds good. Well, you gave a great intro right there, so I could just pass that on to the next question because that was awesome. Yes. Before I begin, me and Peter will be making forward-looking statements today. So please see our SEC filings for a description of the risks that we face in our business. So yes, so Exelixis, commercial-stage biotech company focused on oncology. Flagship product is -- based is cabozantinib, 2 different forms. Obviously, CABOMETYX is the main driver of revenue growth. Certainly strong year-end 2021 based on the success of the 9ER launch and looking at a pretty robust 2022 as well. We're using the free cash flows and the revenues in our balance sheet, obviously, to reinforce the pipeline. It's really hard in this business to have one blockbuster like cabo. It's even more challenging to do that again and again and again, but that's what our goal is. And we aspire to be a global multiproduct company in the future, and we're working really hard through Peter's team and Peter's efforts on the discovery side. As you know, Vicki Goodman joined to be Head of Development and CMO. And then we've got a real strong commercial team with P.J. Haley and that group in terms of being able to get out and educate docs. So we're fully integrated in the U.S. And with cabo, we have strong partnerships ex U.S. and that certainly, going forward, will be different, where we hope to both develop and commercialize products on a global scale. And certainly, with our balance sheet and our success, we feel pretty good about doing that going forward. So lots to do, busy year in '22 and certainly '23 as well. But a lot of momentum and a lot of excitement at the team here in Alameda and then also on the East Coast in the Philadelphia area. So we're growing and have a lot going on for sure.

Yes, definitely. You can see front and center, you're hiring, if you visit the corporate website. So it's nice to see that.

So let's talk about the 3 Phase III trials that's reading out in the second half of this year, right? So the first one, from a chronology standpoint, it's COSMIC-313. Strong rationale, right? Basically, the hypothesis that you like to talk about is you transition some of the PD patients that you see with 214 to either stable disease or partial response or even complete response. CABOSUN, it's kind of a similar population, intermediate/poor risk. One thing that we've gotten pushback on is really this idea that in the intermediate/poor risk patients, it's not as well VEGF-driven as, say, the favorable risk population. So I'm just curious if you have any kind of thoughts on that dynamic.

Yes. Well, that's a debatable topic. That's based on somewhat dated data. If you look at the totality of data for cabo, I think we could engage in a pretty robust debate there. I guess I'll just add to one thing you said about some of the kind of background for 313. It's really -- you talked about the classical patients, say, from 214 that had PD as their best response. Can we convert those into either a response or stable disease by adding cabo on top? Again, our DCR, our ability to have either disease stabilization or tumor shrinkage, if you look at METEOR, you look at CABOSUN, you look at 9ER, it's in the 90-plus percent range, so pretty strong signal there. But also, I would say, in the maintenance space, looking at the combination of cabo and nivo, again, based on the 9ER data, might be a more effective way to keep patients stable or improving with their disease in the out months once they're past the triplet or the maintenance phase, right? So -- but again, global pivotal trial, that's why we do the experiment to work out all the details and get the -- see the global truth based upon -- or the cosmic truth based upon the global trial. So we'll see that. But Peter, maybe he can speak to some of the topics around more anti-angiogenic phenotype.

Yes. No, I'd be happy to comment. And you're correct, there's certainly some data out there where people have looked at gene expression profiles typically taken from tumors obtained from primary nephrectomy. So that's sort of something to understand. That's what the tumor material is. It's not immediately pretreatment with anything. It's obtained at the time of surgery and then there's a variable amount of time between that when anybody gets treated for -- with anything, be it a VEGFR inhibitor or IO. And there have been attempts to assign signatures either to make an assessment of whether they confer a favorable prognosis or are predictive in any way. That's included in IO-type signatures, as you might imagine, and what people have called these angiogenic profiles. There's a couple of different ones that have been used. And typically, they're just looking literally at genes that are believed to be involved in angiogenesis like VEGF receptors of various kinds, sort of a number of others. I think it's interesting as you start actually digging into the data, there are overall scores assigned. If you look at the individual genes, they do bounce up and down a bit. So it's not a one size fits all, that all genes that are associated with angiogenesis are specifically only associated with particular IMDC subtypes. I mean that's almost certainly not true. Yes, my other word of caution to those gene signatures, sometimes they're gene signatures, right? So at least to my knowledge, they haven't directly been correlated with the degree of vascularization or micro-vascularization of those tumors, in other words, how angiogenic they actually are. So I think saying that driven is maybe a little bit of an overstatement with gene expression data to be clear. And then finally, of course, people have looked at this historically in the context of sunitinib and crizotinib. Also, we don't have data yet whether that will be true of cabozantinib, given its differentiated profile. I think it's quite easy to envisage that things may come out quite differently if we were to do some statements. So I think we're very confident with the patient population that we're enrolling in 313. And as you commented at the beginning, I mean, it's based on solid clinical data about activity in those subtypes.

Great. Great. Excellent. That's a great overview. Thanks for sharing with us. So I just wanted to get a sense of the data release. So basically, this is the primary analysis of the primary end point of PFS in an interim analysis on OS come the July top line release, right, for the 313 study.

That's the guidance, yes.

Okay. Great. Great. Great. So moving on to CONTACT-01, high-risk, high-reward study. I think nobody would doubt the kind of the financial upside for non-small cell lung cancer. So you wanted to revisit the cohort 7 data presented at ASCO. I was just curious, what should investors take away from the 80-patient CONTACT 7 cohort (sic) [ CONTACT-01 cohort 7 ] data?

Yes. So like you said, we had an updated look at cohort 7 at ASCO back when we were in Chicago, meeting with you afterwards. So this is the second time in a month that we've had a chance to talk, so just realized that we're talking on a regular basis, which is great. Yes. So clearly, with all the caveats of it being a single-arm, non-randomized, small, if you will, small number of centers study. The combination looks active in terms of response rate, PFS, overall survival, intervals, those kinds of things, good duration of response as well. But again, I think you have to look at that in the context of what that trial was, which is a single-arm, non-randomized study. And CONTACT-01 is designed to ask the right question the right way relative to a global randomized study against the standard of care. So we're excited to work with Roche on both CONTACT-01 and 02 in prostate cancer and 03 in second-line -- or second-, third-line RCC. And again, we're in the business of getting p-values and raising the bar for standard of care. So whether it be lung cancer or prostate cancer or, in this case, more renal cancer, that's the job that we do and looking forward to having these studies read out with 01 and 03 this year and 02 -- our -- yes, 02 probably next year. So yes, so off we go.

Yes. Yes. Okay. So looking at kind of the field. CONTACT-01, there was also LEAP-008, Lenvima, Keytruda, SAPPHIRE, right, sitravatinib plus Opdivo. They're all sort of looking at the same patient population, but the enrollment criteria is a little bit different, right? Mirati, they basically require patients to stay on therapy for 4 months and then have basically a radiographic relapse. So I'm curious if you have gone through this kind of thought of exercise before. Is that the approach that you can answer questions about the immunomodulatory properties of cabo or basically kind of have -- cast a wide net, just enrolling anybody who had previous exposure to IO or chemotherapy in the front-line setting? I just wanted to get a sense of the thinking behind the enrollment criteria for CONTACT-01.

Yes. Maybe I'll start and Peter can add some color commentary. Again, I wouldn't want to comment on what other companies are doing. I don't know what they've been thinking about and how they've been able to prioritize, what they're hoping to see clinically and the impact on the commercial setting versus their enrollment criteria. So I'll leave it for you to talk to them. But certainly, for us, with -- in the Roche collaboration with CONTACT-01, we wanted to cast the widest net possible. There's many patients that are refractory upfront to either an IO or IO/chemo, either combination or sequence, as well as those that do have a response or have some clinical benefit and then become resistant. So we're trying to cast the widest net as possible to be able to have the greatest impact if the data is supportive of that. So Peter, do you want to...

Yes, sure. I mean the thinking behind that mechanistically and scientifically is simply that there's a rationale in both settings, both in the primary refractory setting and in the acquired resistance, I mean, based on all the data. I mean for the acquired resistance setting, I mean it is unfortunately true, despite a significant amount of effort over the last 10 to 15 years, that the mechanisms driving resistance to checkpoint inhibitors are still relatively ill defined. There are some specific molecular alterations that have been documented, lesions in MHC peptide processing, beta-2 microglobulin, JAK-STAT signaling, for example, but they really appear to make up the minority of patients. So that's not a real general mechanism, although it does occur. There has been a lot of focus on changes in immune cell populations associated with acquired resistance, which as we bring up one of my favorite preclinical experiments with cabozantinib, which was really looking at acquired resistance in the preclinical setting to checkpoint inhibition, essentially what they found in -- I think it was a prostate model they were looking at that it was -- resistance was driven by recruitment of an immunosuppressive neutrophil subset, and that recruitment was driven by MET. And that if you inhibited MET with cabozantinib, you could then block that resistance mechanism from occurring and extend the responses. So the acquired setting is very definitely a rationale for exploring setting further primary refractory. Again, it's a matter of tipping that kind of immune microenvironment into a more immune permissive state or status, which is certainly what we believe cabozantinib does based on the wealth of clinical and preclinical data at this point, thereby enabling the checkpoints to do their thing. So we think looking broadly is the way to go. I'll just add that at this point there are no cross-molecular ways to really subset those things. And if you start looking at acquired resistance, generally, the exact definitions of what constitutes acquired resistance clinically could sometimes be a little complicated.

Got it. Got it. Yes, that's super helpful. So again, that means sometimes investors do like to look at other trials to kind of gauge the probability of success. One trial that really struck me was LEAP-007. So this is basically looking at a front-line Keytruda monotherapy, PD-L1 high population. In that study, Lenvima didn't really add much to Keytruda itself. Is it just in that hot population, right, kind of hot tumor, Keytruda alone is sufficient? I don't know if you have any interpretation.

Yes. Yes. I guess I would be -- I would hesitate to speculate on what happened there. I mean they've missed on a few. They missed on bladder too that had really good Phase II data. So is it a just kind of missed signal with a small population in Phase II? Patient selection? Is there a tolerability issue or tox issue? I mean, it's hard to say. So -- but again, we like our 021 data from cohort 7. And we're in the business to ask the right questions and run the appropriate global pivotal trials to get the answers. And that answer will -- again, as we've guided or Roche has guided, will come out in the second half of the year. So we're excited about that. And like you said, it's a big population. The unmet medical need is very, very significant. And we're hoping to help a lot more patients with cabo if we can. So that's just -- it's all going to be in the data. So we'll just keep going here.

Yes. All right. All right. Fair enough. So moving on to CONTACT-03, right? So there is some debate in the RCC community about the use of the PD-1 and PD-L1 blocker. So I'm just curious if you have any thoughts on that dynamic. Obviously, in the front-line setting, I think most people will get PD-1 blocker, right? Opdivo, Keytruda are very entrenched in that setting. So would it make sense to view a switch to a PD-L1 blocker like atezo be an advantage in the setting? Or maybe your view is that these 2 classes are essentially clinically the same?

Yes. So I mean, it's a question we get a lot and there's always a lot of maybe -- well, there's certainly no data comparing a PD-1 versus a PD-L1. So no one has taken up the charge to actually answer that question in the -- in a randomized sense, even though everybody asked the question, right? So in our hands, right, you look at cohort 1 and cohort 2 from COSMIC-021 where we looked at front-line cabo/atezo in RCC. We saw very, very good response rates, good duration of response, good PFS, all the caveats, non-randomized, blah, blah, blah, but on the same general activity that we've seen with cabo and other PD-1 targeting molecules. So from our point of view, they look roughly equivalent. I think the real question is what is driving resistance for front-line RCC patients that get an IO or an IO-X combination? Is it IO escape? Is it angiogenic escape? Is it something else? And you can ask a lot of questions biologically, but I think the best way to understand things is to go in there pharmacologically and run the pivotal trials. You need to understand what different combinations can do, which is what we've done with CONTACT-03. So again, trial is fully enrolled. We should hope to get data later this half and looking forward to opening the envelope for sure.

Right, right, right. So let me just squeeze one more trial into the discussion, if you don't mind. So study 111, that's basically asking a very similar question to CONTACT-03. So this is basically getting patients in the front-line setting, IO/IO or IO/TKI and then move on to Keytruda, Lenvima, 100-and-so patients, 50-something response rate and 11-plus PFS. So is that something that kind of increases your confidence about CONTACT-03 based on the data that it's -- the efficacy is really robust, and you can basically kind of re-sensitize patients to IO-based therapies?

In the context of a single-arm, non-randomized study, that is just -- it's a snapshot of what an experimental arm could look like based upon the patients that were selected for that study. And that's where you got to be careful. We've seen lots of those single-arm, non-randomized studies not always confirm when they're done in a randomized sense against an active control. So that's -- you just have to run the right experiment. Certainly, you can be either hyper-positive or nervous based upon what you see with that single-arm, non-randomized study. But I think that's the learning, especially in a second-line population, is that you've got to run the randomized study with the appropriate controls and ask the question the right way, right? So we'll see.

Right, right, right. Okay. Great. So moving on to some pipeline assets. So obviously, XL092, a lot of people are really excited about that, but actually -- on getting the STELLAR-303 study up and running. So yes -- so I think one of the more exciting things about that is the ASCO GI presentation from the CRC cohort, especially within the wild-type, the RAS wild-type population. So maybe you can remind us the data that you've seen from 2 trials and really kind of the genesis of the STELLAR-303 study design.

That was good. Peter?

That was yours.

Was it? I thought that was yours. Okay. So yes -- so the 303 design is really inspired by the data that we saw with -- again, as you mentioned, with ASCO GI, where we had either a cabo/atezo combination that we did or a cabo/durva, is that correct, combination that was done via an IST where we saw between 25% and 50% response rate in the RAS wild-type population. So again, standard of care there is essentially regorafenib, so -- which has a pretty low response rate, low PFS, overall metrics. So it's viewed by us to be a pretty, again, enabling and exciting response that we saw, a dataset that we had at ASCO GI with all the caveats that it wasn't randomized against the control arm or it wasn't a large study. So yes, so it's a very simple, straightforward approach, look for KRAS wild-type patients and enroll a broader population, both with the mutant and wild-type. It really focused the primary end point on the wild-type population, and off we go. So...

Okay. Yes. All right. So yes, the next asset is the tissue factor, ADC. I think a lot of people are excited about that. So I think the company talks extensively about the difference in binding profile and the sparing -- the Factor VII sparing properties of the antibody component. So obviously, logically, that would lead to a reduction of bleeding events. And so we're just wondering -- the other, I guess, safety liability with TIVDAK is really the ocular toxicity. So is that something that you see as an improvement for XB002 in the clinical setting? Kind of what's your thought on that?

Yes. So generally speaking, with XB002, I mean that was one case, as you pointed out, where we had an opportunity to go with a different antibody that bound to a different epitope from TIVDAK and there was a clear connection between a clinical AE and that binding site. While we see generally, as we're looking at ADCs, particularly ADCs against known targets or targets which have been the subject of first-generation ADC approaches, our thinking has been to come in with next-generation technology, which is aimed at improving the overall characteristics of those ADCs, hopefully, including also what their tox profile is going to be. And that's really based on, one, going to site-specific conjugation technique. So we have very precise control of the DAR. We have a very homogeneous molecule that kind of helps on the manufacturing side and then like to be able to pay attention specifically to the chemical characteristics of the warhead itself as well as the linker and the linker stability. So going back to XB002, we believe that one of the advantages of that molecule beyond the binding site is the differentiated or a statin that we have there along with the linker technology, which is the Zymeworks ZymeLink technology. So it's not the same as the classic MMAE that's in TIVDAK, it's not the same linker at all. It's highly stable. We're seeing that from a PK point of view, both preclinically and clinically. And historically, certainly, instability of linker payloads is something that's contributed to tox overall. So we're looking for kind of a -- hoping for certainly a global improvement on the tox and the TI overall. And that's something you'll see from us again going forward as we're advancing other ADCs. We had a new ADC going to preclinical development late last year targeting 5T4. That's, again, certainly a target that's been the subject of ADC efforts historically, but we're using the same thinking there with a kind of contemporary linker warhead and site-specific conjugation technology. So that's our general thinking and that's how it applies to XB002.

Yes, yes. Okay. So in terms of the clinical development, there is a combination effort going on in combination with Opdivo in head and neck. And at ASCO, I think, one of the positive surprises is cabo's activity in head and neck. So is there any sort of scientific rationale to kind of do the checkpoint tissue factor ADC and then VEGF, multikinase activity, triple combination? Is that kind of in the store?

I would say, generally speaking, I think combining agents that have activity from the immune side with ADCs makes a lot of sense. If you have the right payload, that's inducing immunogenic cell death, and I think to your point that data in ASCO, I think there's a lot of interest in that. I think there's generally a lot of interest in the field in doing that going forward. And then going to cabozantinib, and yes, same for XL092, right, given the immune-modulatory properties of those compounds, just from that point of view, I would say it would make sense. And then we've got the potential for additional activities, whether they're anti-angiogenic or direct tumoricidal type of activity. So scientifically, it's certainly something that is worthy of thought.

Yes. And again, 313 is a good example of a triplet that makes, -- again, mechanistically and from a rationale point of view, makes a lot of sense. And we'll do more of those going forward. And hopefully, we'll do those triplets with more than one component from the Exelixis portfolio, right? So we've had discussions about 092, 002 plus X across different tumor types. So as we get more data with the doublets, then going to triplets is a natural consequence for sure.

Right, right, right. Okay. Yes. So at ASCO, I believe this is the first time I heard from at least KOLs, independent KOLs, not company representatives, talk about kind of vedotin immunogenic cell death. I don't know, Peter, if you want to comment on that, about kind of the validity of that based on experimental data. Is that effect real or it's kind of an artifact?

No, no, it's certainly preclinical experimental data around a number of different payloads. It's not unique to vedotin by any means, and immunogenic cell death is just characters. Those methods of characterizing cell death as being immunogenic based on display of certain cell surface markers, for example. So it certainly meets that criteria. But again, it's a fairly broad set of payloads that have the potential to induce that. I'd also add, just generally, if you're having a direct tumoricidal effect, one of the things you're doing is then just releasing a lot of tumor antigen into the local microenvironment, which in and of itself tends to be immunostimulatory.

Got it. Okay. Yes, I wanted to talk about the ADAR since it's kind of bubbled up both in the investment community and the scientific community. That's something that you've kind of talked about a lot as a pipeline program. So one thing, if you approach that from a therapeutic standpoint, is really the need to identify cancer cells with interferon signatures to really get the therapeutic index. So how should we think about having a biomarker that measures interferon signatures within the tumor context.

Yes, just to provide some background there. I mean, it's a, I would say, hot target of oncology from a research point of view. It kind of came out of a number of independent screens looking at tumor-specific cellular liabilities or genetic liabilities, where the cells were clearly dependent upon ongoing ADAR activity for survival. And then, long story short, it turned out the reason for that was in the absence of ADAR, you tend to stimulate release of interferon from those cells through activation of internal sensing pathways that sends abnormal nucleic asset structures, in this case RNA structures. And then if those cells have interferon receptors, they then die. They basically commit suicide through that mechanism. And ADAR activity keeps that in check and then allows those tumor cells to continue to grow. And so as you point out, there's been a gene signature on ISG identified that seems to correlate with dependence upon ADAR somewhere between 10% and 30% of all the tumor cells in TCGA, so a pretty broadly applicable mechanism. So that's kind of the state of play right now. I think -- in terms of what specific biomarkers might look like down the road, I think I'm probably getting ahead of myself a little bit, I mean, this is still an early-stage program for us, and as far as I know, for anybody else. But we certainly have a reasonable starting point based on that information to start looking at a number of candidates that could potentially help guide either monitoring of pharmacodynamic activity or ultimately being able to pick out more sensitive patients.

Got it. Okay. Yes. All right. I want to talk about the BioInvent deal that was announced recently. So it has this very unique screening using patient tissues. So we're just curious about what Exelixis could gain from the BioInvent deal that you guys do not have the internal capability to do. And yes, and basically, internally for Exelixis and also you were kind of a broad external collaboration.

Yes, very happy to have gotten this deal done. It's an exciting deal with BioInvent who are a Swedish-based biotech company. And it's sort of in a different category of deal in some ways from all of our previous biotherapeutics-focused deals. I mean, as you know, over the last 2, 3 years, we've done a network of deals that really enable us to go from antibody production or antibody discovery, if you like, and engineering through Invenra collaboration, which we've kind of expanded on a couple of occasions historically and is providing us with antibodies going forward. They provided, for example, the 5T4 antibodies that are in our most recent ADC. Then we have our collaborations around site-specific conjugation and also giving us access to payloads, either through Catalent, Redwood and NBEBI now. We also have a collaboration that enables us to generate conditionally masked antibodies through a collaboration with Adagene. So that's sort of the current engine for our biotherapeutics pipeline that I would say is operating very well right now. With an eye to the future, we have been looking at adding additional collaborations, partnerships, if you like, that may provide us with additional targets and maybe more novel targets. So the BioInvent deal is really the first of those, and we're continuing to look at other opportunities there. So the attractions of that platform were a couple Their platform, one really attractive part is what it starts with is primary human tumor material. So they've established ongoing access to primary human tumor samples that include the tumor cells, but also all the microenvironmental cells, that does not have stromal cells, the immune cells, and like, so it's a real tumor in its real environment, essentially. And their approach is then -- it's basically agnostic. It basically asks the question in various cell types. Using an antibody panning approach, what antibodies can we see bind to those cells that don't bind to normal versions of those cells? So you could see things that pop up in T cells that wouldn't be present in T cells in normal tissues, just as an example. And then once they fish out those antibodies, they immediately have a tool to ask, is there a function associated with the antibody binding to this thing? So it's a target identification and validation platform, the output of which is an antibody that could either be a therapeutic in and of itself based on its functional activity, or could be the starting point for a bispecific or for an ADC. So we're looking at this really as a way to go from tumor material through the potential therapeutic and provide some more novel targets to our overall biotherapeutics approach.

Great. Great. Yes. Thanks for that insight. So moving on to the Oragene deal. I think most people kind of associate the Oragene deal with the CK7 program. But I believe in the press release you also talked about protein degradation, which has been also kind of an interesting topic. Anything that you guys are working on there, especially given Exelixis has been kind of a kinase company historically. You guys are really good at kind of designing things that bind to a pocket. So that's obviously a very well, suitable technical ability when you're designing things like that to gauge the ubiquitin ligase complex.

Yes. Just to speak to the Oragene collaboration a little bit. It's a collaboration we've had for a number of years now. It's a broad collaboration in small molecule discovery. So just for clarity, it's not focused specifically on degraders. They have some experience in making kind of PROTAC-type degraders historically. So it's an area where we have a couple of early-stage programs with them, just sort of exploring that. Looking at the degrader field broadly, I would say it's certainly an area of interest for us for all the obvious reasons. You get access to a wider target base. And to your reference to kinases, obviously, there can be differences and sometimes meaningful differences between inhibiting the kinase in the traditional ATP competitive versus just degrading the whole protein, and there are some interesting programs out there that take that. So I think it's the technology and approach that we want access to. The PROTACs -- I mean, in some ways, the PROTACs have now entered kind of clinical validation stage in terms of what's going on, especially at venous and the like. So there's certainly potential of interest. Its pros and cons, they tend to be fairly large, making orally bioavailable can sometimes be a challenge because of that. On the molecular glue side, historically, that was based more around IMiD chemistry, which is kind of less interesting to us. I think there's been a lot of exploration of that. And what's going to come out of that is what's going to come out of it. But more recently, there have been some, what I call, newer glues -- smaller compounds have been identified or causes that have been identified that have glue-like activity. So I think our view of it is, it's very interesting technology. It's something we want to have access to going forward. And we're kind of looking at the general landscape there right now actively.

Okay. Yes, yes. That's a very interesting budding field. So maybe one of the last technical questions regarding the construct of ADCs. Obviously, you have the flexibility of conjugating not only to a payload. People are looking at immunomodulators, maybe perhaps like a niche field is looking at radioactive isotopes. Just curious about that's a huge white space. Is there any kind of particular focus for the company?

Yes. No, I agree. I mean, I think it's stronger other than everybody. If you look historically at the ADC field, the vast majority of the payloads fall into 1 or 2 mechanistic categories. So really the microtubular disruptors of one kind, of which MMA or statins are the classic, or they are DNA damaging agents of some kind. And obviously, there's kind of rise of ENHERTU and the tecans kind of leading the way there. And I think there's still a lot of clinical validation there, there's a lot of understanding of various iterations of those scaffolds and mechanisms and work. So I think there's definitely more juice to be squeezed out of those particular payloads. That said, I totally agree there's a huge white space. And people seem to have been confined within this box because they knew that works and I suspect there was so much else going on trying to optimize ADCs, they didn't want to start changing that as well. But I think things have changed. So we certainly have an interest in payloads beyond those classical ones, you mentioned immune-stimulatory ones. They're certainly of interest. But I think even going beyond that. I mean, look, we have a pretty deep history in small molecule chemistry generally. It's like, well, can we exploit some of what we know, some of what we've learned, even some of what we did now in the ADC space? And so are there other MOAs that make sense? So kind of on the current research side that's something we're keen to explore. And again, generally, I kind of spoke a little bit to collaborations we have enabling us to generate antibodies and the like. So on the one hand, we're looking to generate panels of antibodies. And on the other hand, obviously, we're looking to get access to a broad range of different payloads, so that we can mix and match as is appropriate for the various tumor types in patient populations that ultimately we want to go into.

Great. Great. Okay. Perfect. So maybe last question just to round everything up, but this is kind of a timing question for the bench trial. So Mike, I was just curious, for these type of bench trials, how long do you think that we might get some sort of final results. Like when do you think Judge Andrews might deliberate in terms of the MS IP lawsuit?

Yes. So we've given quasi guidance at a ruling on the first case sometime late Q3, early Q4. Hard to pinpoint that. That's obviously out of our control. We have the second end process ongoing right now, and that will go trial for the crystal and malate forms of cabozantinib in second quarter of '23 probably. So again, we've got a very extensive patent estate that protects cabo and its solid-state forms that we are going to be continuing to actively protect and litigate as needed, right? So I don't want to get into details here, but I've been really pleased to see you and others on the sell side getting engaged in all the information in the public domain and I think that's great for investors. And we're going to continue to be very diligent on our side behind the scenes and shooting for the best outcome possible.

Okay. I think that's a wrap. Mike and Peter, I really want to appreciate your time and sharing with you kind of the latest update on the company, pipeline progress, clinical advancement, the whole 9 yards. So really appreciate your time today.

You bet.

Thanks, Andy. See you.

All right. Take care.