Xencor, Inc. Stock

For joining this session. My name is Alec Stranahan. I cover SMID Biotech at Bank of America, and I'm the analyst covering Xencor. And it's my pleasure to introduce Bassil Dahiyat, President and CEO of Xencor; and Dane Leone, Chief Strategy Officer. Guys, thanks for being here.

Thanks so much for having us.

Yes. Looking forward to the discussion. So maybe, Bassil, like just at a high level, what gets you most excited about the company? And which asset is your favorite?

Well, I think the tempo of data that we have starting in the second half of this year with our XmAb819 Phase I, sort of second readout where we hope to have a recommended Phase III dose, a clear pivotal plan and robust efficacy and tolerability data in our target dose range. That is coming on really soon.

So I'm very excited about that. And that's really the beginning of a flow of data that we're going to have our second oncology solid tumor bispecific XmAb541 in gynecologic and GCT and then followed in the new year in 2027 by our first bispecific TL1A containing molecule, XmAb412. We expect to have healthy volunteer PK/PD and safety and immunogenicity data for that one.

And then shortly after that, in the second half of '27, the full Phase IIb readout of our ultra-long-acting anti-TL1A antibody, XmAb942. So the transformation that we undertook a couple of years ago of reorienting the company around high engineering concept but validated biological concept drugs and disciplined clinical development is really going to play out.

So that's what I'm most excited about is that the way the drumbeat that's been a little slow and coming, was really going to hit a rapid tempo this next year of real clinical data that's really going to move the needle. And it's going to allow us to make really critical decisions about which programs to lean into and deeply invest in to reach our true goal, which is to be a commercial stage company, which ones that might benefit from strategic partners or other sources of resources and capital. So all those decisions are coming. So I'm really, really on the edge of my seat as we approach those.

Great. Great. Well, maybe we can start at a high level. You've built several clinical stage programs from your engineering platform. This remains -- it's been differentiated. It remains differentiated, in my view.

And you have several active partnerships as well. I guess, has focus shifted at all towards building your wholly owned assets and pushing those towards commercial stage? Or is the partnering model sort of still central?

The partnering model underpins and shares that common technological base that we have. I think the mindset has shifted in the investment community and to the pharmas that we speak to that, oh, okay, Xencor is really putting its chips down and has very viable clinical assets that are going to give me interpretable near-term clinical data to make go decisions and investment decisions.

That, I think, is something that has been a change over the last couple of years, and it's a welcome and it puts the burden on us. It's a challenge. That base of partnerships and the revenues that it brings in and the validation certainly helps. But I'd say that shift has been very, I think, pronounced in people's minds. And it means it's a show-me story, and I'm very happy about that.

Okay, then show me the clinical data that's going to get you to Phase III in RCC with a differentiated agent. Show me the best-in-class TL1A antibody data, deliver. We're happy to give you a shot.

Yes. And we've got plenty of catalysts look forward to pipeline forward. I guess maybe starting on XmAb819. This is your ENPP3 x CD3 in T-cell. You showed some pretty impressive 25% response rate, 70% disease control in very heavily pretreated patients, manageable CRS, which is -- has dampened activity for others in the class.

I guess what sort of dose expansion data do you need by year-end to sort of justify moving the asset forward?

Yes. So let me take the start with the setup on the data, and then Dane can really dive into what that means for us for our pivotal phase thinking, which has a few options. So for us, we designed this molecule to selectively engage high expressing ENPP3 expressing tumor cells and avoid hitting and killing with this very potent CD3 T-cell mechanism, lower expressing normal tissue.

And we think the data set from last fall at the ESMO meeting really demonstrated that we weren't seeing critical organ toxicities really with CRS, which is on class for CD3s. It was some rash, which is on mechanism for killing ENPP3 positive cells. There are basophils and mast cells, basically allergy driving cells in your blood that have that target.

And so you hit those, you get some rash. But really, the AEs really clear up as you get to the end of the month through that priming regimen and that immune boosting phase, and then you're at a place where it's really a well-tolerated regimen with again, no major organ toxicities that we observed at all and that promising efficacy profile. That was the beginning of getting to our active dose range.

What we spent the last 7 or 8 months doing has been now taking dose levels within that range, enrolling expansion cohorts. We've completed enrollment in the first one. We're enrolling rapidly the second one, and we're really doing a lot of experiments on the side because we have a very, very high interest from investigators and patients, making sure we're optimizing our priming regimen and things like that towards the end of this year, we'll have robust data sets of expansion cohorts at least a couple of doses with their priming regimen set that we could say, we believe this is the right one for recommended Phase III dose.

And depending on the degree of response rate we see and the kind of early durability signals we see, we think that could position us, in particular, for very high unmet need subsets like multiple TKI pre-exposed patients or prior HIF-2 alpha exposed patients, that might afford a more rapid path to approval potentially. So it depends on the data, but the setup is great, and we're going to really deliver robust and much larger data package in that first glimmer last year. Maybe that's the sort of setup for what's next?

Yes, sure. I mean from our view as an organization and as a management team, 819, we're all in on what we see as a very favorable probability of technical success with this program and the ability for it to potentially be the first commercial stage product for Xencor as an organization.

And to do that, we've taking the proof-of-concept data that we presented at the ESMO meeting in October last year to trigger the dose expansions that we need to characterize and do the proper Phase I work to satisfy Project Optimus and work with the health authorities to have an end of Phase I meeting that can then really, to Bassil's point, elucidate the best and most rapid path forward to late-stage development or through late-stage development.

But that's only a part of the story, right? Because let's say, we go with a monotherapy registration-enabling study post IO, post TKI that affords us a flexible kind of second, third line label, much like belzutifan has in the U.S. today that's helping propel that drug to be a $1 billion drug. But we're thinking more broadly than that already. We're starting a pre-TKI study post IO/IO or classically ipi/nivo, which is about 1/3 of the frontline patients in advanced clear cell renal cell carcinoma.

If that proof-of-concept substudy works out for us, that would then catalyze another study that could then allow us to capture the totality of the second line. And an anecdote there is really where cabozantinib is used today, which is a multibillion-dollar drug, largely used in the second line. So we think that the TAM for 819 in clear cell renal cell is going to grow rapidly, and we'll start thinking about frontline opportunities as these second, third-line opportunities start to mature along the clinical development pathway.

But this is a targeted oncology agent, right? And ENPP3 is implicated in a number of different tumor types. And that's why we started screening and are now dosing patients across non-small cell lung cancer, colorectal cancer and papillary renal cell carcinoma that are ENPP3 positive. because we want to pursue that real target oncology development pathway efficiently and rapidly to say, okay, we have primary indications that really make sense where you don't have to screen patients.

We're not preselecting in clear cell renal cell carcinoma because the H-scores are so high and consistent across patients with that histology. In these other tumor types where we would prescreen, we almost look at like a potential basket of very large market opportunities that could ultimately lead to a tumor-agnostic label follow the path of other successful programs like Enhertu. So we're very excited about this.

We think we are novel first-in-class with a really differentiated agent. And we have the internal expertise of T cell engager development through partnerships like Amgen and others and Astellas that give us a lot of insight of how to do this properly and successfully that we're applying to our wholly owned programs. So we're very excited about 819.

Yes. No, that's great. And I guess in terms of how you step through the derisking in the later line CC/RCC, the frontline TKI naive patients or the ones that have progressed on IO and then the broader opportunity set for ENPP3-driven tumors.

Is it sequential? It almost sounds like you're approaching this in parallel in terms of the initial derisking, but then how -- once you get that data in hand, how do you sort of pick and choose?

I think the clear cell opportunity is the one we focused on initially because -- we didn't have to prescreen on the target antigen. We could move rapidly. We could be assured that nearly all the patients, and it's at least 90% of them have the target.

So we could elucidate the biology, come up with a dosing regimen and particularly a priming regimen that we knew was reflective of an ENPP3-rich tumor environment. So we wanted to make sure we could check that developability and safety box in the most rapid and robust way. And that, of course, naturally means clear cell opportunities further ahead, right? We could be starting a pivotal next year if all goes well. That's our plan.

The other ones are just going to beginning to efficacy proof of concept, we hope by next year. We didn't have to redose escalate. We could go with the active dose that was tolerable in this target dose range. So they would then really have kind of independent depending on their own commercial and regulatory pathways, development paths.

And so we would follow the signal where the data takes us. I mean, in late-line colorectal cancer, there are potentially more rapid ways to approval, right? We would see. But clear cell is ahead to be clear.

Parallel is the right word. And that's very contemporary for successful development in clinical oncology is as soon as you have that real signal in the first setting, right, which you can prosecute, in this case, clear cell for us, and you've looked across the board at the emerging data set you have internally to say, okay, we've more heavily pretreated patients seeing clinical efficacy on par with the marketed drugs available in the second and third line, that's worth investment.

That's where we should be putting our dollars to invest and accelerate this program to make sure we cement ourselves as the leaders of this molecular target.

And I guess there's also learnings you could probably apply to the dose from the dose escalation to other tumor types, right? Like do you anticipate having a higher starting dose and, say, CRC?

We're starting at the active doses. We're starting at the doses that we've expanded at.

RD1.

RD1 is being...

Colorectal...

Colorectal Yes, there was no need to do that. Luckily, in oncology, you don't have to rewind and start all over again usually. So we're very happy with that. So we expect to have clarity, is it working? Is it not working much more rapidly.

Okay. And maybe thinking about your Claudin-6 program, too, this is XmAb541. Early responses across -- this is for germ cell and gynecologic tumors, some distinct competitive dynamics across germ cell and ovarian. But I guess, how does the early data sort of shape your thinking around which indication becomes sort of.

It set up the experiment, really. So the experiment that we're doing this year is we have this early data. We were able to accomplish our escalation more rapidly because we've learned a lot about how to escalate solid tumor CD3 is about CRS, about toxicity profiles to where right now, we're doing the real experiment at -- in the right dose range for both germ cell and gynecologic tumors, essentially separately, right?

They're going to have their own different efficacy profiles. And they have their own sort of competitive landscapes in the GCT, rare, but for the people that fail on high-dose chemo, very high unmet need. And so that's one dynamic, but rare, very -- no competition and a very competitive landscape in ovarian with different modalities like ADCs and a rising bar of efficacy. And so there's a sort of a divergence of the 2.

So we're going to have independent data at target dose range in a robust enough number of patients to decide, do we have the efficacy that merits one or the other going forward? And we would be able to, by the end of the year, that should all be brought together in a way that we can clearly articulate, here's the go-forward strategy or are we going to be disciplined with our capital and not further invest in monotherapy.

And maybe to that point, how are you sort of thinking about combos? I think you may have had an update, if I'm remembering correctly, in the 1Q press release around...

Yes, sure. So we're very excited about what's called the Amgen CD28, CD3 T-cell engager combination. We think that's going to be a real potential emerging class of how to more effectively agonize T cells against tumor cells. And there have been a lot of early experiments of trying to combine CD28 with PD-1 or the same antigen targeted CD3. We think empirically, those have not worked, right?

We view good antigen targets that are different, but co-expressed on the tumor cells specifically has the potential future. And from the work that we presented at AACR this year around what's XmAb808, our B7-H3 x CD28, it shows really good co-expression on tumor cells of B7-H3 in Claudin-6. Since we had monotherapy proof-of-concept activity with 541, but we have to do all these expansion cohorts to verify a monotherapy development pathway one way or the other.

We know, unfortunately, for regulatory requirements, we have to start at what's called MABEL for the CD28 dosing combination. So it's going to take time to dose escalate to cohorts where the XmAb808 arm is going to really be clinically active, we think, even though we can start at a clinically active dose of 541.

So we wanted to start this now, not wait because we do think there could be differential efficacy that's worth investigating. And it didn't really make sense to delay because we have to do that dose escalation. So you'll hear us talk about this more, but the AACR poster we thought was really a good starting point and lays out the investment case to do this, even though it's going to take a bit of time to do something differentiated.

Okay. We've seen some companies layer on a CD28 on top of the CD3 going after the same target. I imagine there, you kind of soak up all the binding that you could with one or the other? Maybe if you could speak around the rationale for going out.

Yes. The real primary driver is you want to be able to narrow the cells that you're killing to just the right ones. And if you have strong expression of, say, B7-H3 and Claudin-6 on your tumor cells, the expression of those 2 targets on healthy cells is not necessarily correlated. And in fact, when we checked, they're not.

So the end gate, it must be Claudin-6 and B7-H3 narrows your -- the cells you're killing. So it makes it a more selective antitumor agent, which hopefully avoids off-tumor toxicity. That's the genesis really of the idea. Now on the point that you're going to block up all the target, Typically, with CD3s, not always a potential design, typically with CD3s, the doses are relatively low that you need to get the T-cells going, you're typically not fully occupying them.

So that can be a problem depending on the details, often not. That wasn't even in our thought process. We were like, how do we make this more selective, right? Because CD3s can sometimes fall down. We saw this in Roche's program years ago in colorectal cancer on healthy tissue tox. So it was really a safety.

Yes. It's meant to broaden the therapeutic index of the CD3. Yes. And that's why you need to do the end gate. If you're doing the same antigen target, by definition, you're not broadening the therapeutic index.

You're doubling down on your toxicity.

You're doubling down on everything. And so that's only if for whatever reason, you've gotten to the maximal dose that you've gotten with the CD3 aspect. and can't push that agonism forward, but you have no toxicity and then just need to punch through the agonism with the CD28. That's rarely and almost never the case though.

Yes. Yes. No, that makes a lot of sense. And I guess maybe in the meantime, as you're dose escalating B7-H3, you can test additional combination partners, PARP inhibitors or...

We haven't yet initiated any of that work, but it's very much in mind. I think we want to establish a baseline of monotherapy activity for something that isn't a truly synergistic -- known synergistic mechanism before we engage in that kind of thing. But it is definitely on the mind because I think that that's what the long path in ovarian cancer would entail as you look at the landscape.

I think -- and there's a philosophy we have, too, especially in oncology, we want monotherapy agents unless we own both of those agents ourselves for a combination.

When you move into earlier line after successful monotherapy development and that earlier line requires combination therapy like what is obviously plausible for 819 to move in the front line, that's fine. But developing a fully only combo agent, right, is not that attractive to us.

Yes. I mean that's the benefit of the flexibility of the platform being able to pair the CD3, CD28 with the.

We view CD28 as a life cycle play and evolution of what CD3s can be ultimately and want to continue to be leaders there just as we have with CD3s.

It's early days.

And we have partnerships on the CD28 side. So we're learning a lot as we go to be as smart and effective as we can in the clinic.

Yes. Great. I want to shift gears to your TL1A program. Some pretty exciting recent updates and upcoming updates and we're kind of between updates and we're thinking about the future for TL1A and your next-gen option.

I guess you've shown some pretty encouraging long half-life for XmAb942, that's your TL1A program, durable target suppression on a single dose in healthy volunteers. You've got a Phase IIb enrolling across 3 different doses. What should we sort of be looking forward to by year-end this year to sort of establish either a best-in-class profile or better understand sort of how.

Yes. The real readout for XmAb942 in the Phase IIb study is going to be second half of '27. So when we have our primary endpoint at the 12-week induction for clinical remission. That's the gold standard in ulcerative colitis development. And our goal there is to think about where the first-generation TL1As are in that -- and the close behind in the IL-23s in that 20-ish percent, maybe low 20% placebo-adjusted response rate or remission rate.

So for us, we want to have something that can be -- that can exceed that, 5% to 10% to show that there's a way to break through the ceiling with optimized exposure with a very high potency molecule, 10-picomolar affinity with a very long half-life, 74 days that lets us go to 2, 3-month maintenance. So that had a best-in-class usability profile, if the heightened exposure and potency that we get from our optimized induction regimen as well as the very durable half-life and maintenance, if that can help us break through that ceiling.

So late '27 is going to be really the driver for us understanding how this molecule could be developed, how attractive it's going to be for Phase III development. And at the same time, a variety of peer companies are doing studies in a range of indications for TL1A that we don't have any need to invest our money in exploring whether it works in rheumatoid arthritis or in NASH or whatever, they're doing an experiment for us.

And so we'll be ready with a go-forward Phase III dose well characterized in that time frame, just as the Phase IIIs are reading out for the first gen.

Okay. And the thought process there is that if you see, say, one profile in, say, IBD in terms of the differentiation versus Roche, AbbVie, others, that would be applicable to other indications.

Well, the efficacy differentiation based on that optimized exposure, certainly, it's not been as validated in other indications as it has in IBD. But I think that plus the best-in-class dosing profile, I think, would make it a compelling entrant into what's really going to be one of the few branded biologic markets remaining or biologic drug classes remaining in the early 2030s.

'23s are going to be out, TNFs already out, integrins are out. So a good place to be if we have a best-in-class profile.

Okay. And I guess, what does best-in-class sort of look like to you? Is this the longer half-life, superior target engagement, less frequent dosing?

I think it's efficacy.

Yes. I mean it's a totality statement based on the market research we've done. in understanding how you have not best-in-class, but best in market, right? So the concept here is 942 could come in well early into the branded period for the entire TL1A class.

And the question we had asked when we were doing our market research was what would it take to have a profile or a target product profile that would be able to convert new scripts when we go to market in IBD. And what -- there were a couple of things. One, have you maxed out the exposure response, right?

And that was what we presented at DDW that with our Phase IIb dosing regimen, which is very reasonable in terms of the actual amount of drug being given in both the IV induction and the subcutaneous -- single subcutaneous maintenance dose that we have greater than target -- 90% target inhibition in greater than 90% of the patients, which is almost double what we modeled for the first-gen class at their pivotal dose regimens that they're actually using in their pivotal studies.

And so we think we have a very clear argument there. And so it's to Bassil's point, let's see where this go. Can this give us an extra 5 to 10 points on clinical remission at the induction period? We would hope so, right? And that is obviously the first and foremost most compelling thing on an FDA label or an EMA label, right, is what your efficacy looks like. Secondarily to that is obviously clinical convenience.

The first-generation class are hamstrung by being stuck at this Q4-week maintenance level dosing. We have real-world evidence and from risankizumab, for example, in the IL-23 class wasn't the first to market in IBD, but is fully dominating at Q8-week dosing, right? And so I think when you have people believing you have the best efficacy with the best clinical convenience, you will convert and win those new scripts, right, even into 1 to 2 to 3 years in the branded period of a biologic class.

So we're very confident that the design of XENITH-UC, the Phase IIb study to answer these questions definitely one way or another. And that's what will win that program, additional investment, right, once we see those results. But we shouldn't overlook the XmAb412 program, which is our novel bispecific and the genesis of a potential bispecific platform for autoimmune inflammatory and allergic disease. Do you want to take them through that?

Yes. The challenge in autoimmune disease is these agents are given chronically. They have clear need for simplicity and durability to give these attractive dosing profiles. recapitulate what Dane said, but it's a much harder challenge than an oncology to design bispecifics to fit the bill.

So when we set out to make our TL1A IL-23, we knew we wanted to maximize the inhibition of both targets because both targets clearly benefit from higher exposure. We've seen this in TL1A for TUSCANY-2b. We've seen this in a myriad of studies for IL-23s. More drug is always better. So we wanted to dial up the potency, but we also wanted to have it in a format that could be readily administered in a convenient subcutaneous modest volume shot, right?

So that means you formulate a lot of it into that tube. So the answer to that was rather than make the fastest thing we can make, which we could have had in the clinic probably 9 months ago, we decided to make something that was a very, very high potency molecule in order to achieve that potency in a format that avoids a lot of the liabilities of tetrameric or sort of Frankenstein-looking molecules where just -- there's a whole bunch of binding domains slapped on in order to avoid those stability half-life and immunogenicity liabilities, we made something that looks like a regular antibody and we had to get potency that was subpeicomolar, so femptomolar potency.

You can't do that with a lot of the common tools for making antibody-looking bispecifics, 1 plus 1 formats by common light chains or like Camelid domains or scFvs, you don't have enough sequence diversity to play with. So we built a new structure and a new format for making for making native light stability pairings of distinct light and heavy chains on each side of the antibody, but very simply by throwing them into the same manufacturing, call Zenlock technology.

That lets us optimize each site independently use all the best tools you can to get those very hard to achieve subpicomolar affinities and get us in a small simple molecule that inhibitory potency that can be delivered in a simple subcu injection. That's a tall order for the not quite as optimized format.

And so we're very excited about 412 and also how this positions us for all the other programs in autoimmune and allergic diseases that have the same really stringent requirements for efficacy and for patient experience. And so we've got another program preclinically now that we've made this modular toolkit. We're going to play the Xencor theme again of make a bunch of modular ones. Find the best ones to develop and then maybe partner the others. So it's very exciting.

XmAb412 will be in the clinic in the third quarter, and we'll have first-in-human results in the first half of '27. So that will be another powerful driver of the TL1A pipeline portfolio that we have.

Right, right. And I guess just in the last 30 seconds or so, how do you see the platform? You mentioned sort of the protein engineering piece. How does that feed into sort of the next leg? We didn't even talk about the autoimmune sort of B-cell depleters as well.

Yes. I'll just say we tried to make molecules bespoke for those applications. Taking B-cell depleters into autoimmune disease means you have to have high tolerability and high safety, right? Means CRS cannot be an impediment.

You can't have -- I mean, the reason why CAR-Ts are challenged is preconditioning, high CRS toxicity, you have to have things that are easy to give for rheumatologists. We think we made molecules that thread that potency needle and give you really durable B-cell suppression using a lot of the tricks of the trade we've learned over the years, making bispecifics for ourselves and for others.

Yes. Great. Well, with that, I think we're going to have to leave it there. So Bassil, Dane, thank you so much.

Thank you, Alec.

Thanks for being here.

Thanks for having us.

Hello, and welcome to the Xencor DDW 2026 Webcast. [Operator Instructions] Also, reminder, this conference is being recorded today. If you have any objections, please disconnect at this time.

I will now turn the call over to Charles Liles.

Thank you, and good morning. Yesterday, we issued a press release introducing the topics we plan to discuss on this webcast, including results from the Phase I study of XmAb942 and the preclinical characterization of XmAb412. The press release is available at www.xencor.com.

Providing comments on the call from Xencor are Bassil Dahiyat, President and Chief Executive Officer; John Desjarlais, Executive Vice President and Chief Scientific Officer; Dane Leone, Executive Vice President and Chief Strategy Officer; and Mark Osterman, Senior Vice President and Head of Clinical Development for Xencor's autoimmune programs. We are also joined by Dr. Vipul Jairath, Professor of Medicine in the Division of Gastroenterology at Western University and Chief Medical Officer at Alimentiv, who will join Mark for a discussion of our Phase IIb XENITH-UC study.

After the prepared remarks and presentation, we will open up the call for management to answer your questions. Slides that we are using today should be visible here on the webcast and will be made available for download on the Events & Presentations page of our website around the time our remarks are concluded.

Before we begin, I would like to remind you that during the course of this conference call, Xencor management may make forward-looking statements, including statements regarding the plans and objectives of management, future product offerings and research and development programs as well as future financial and operating results, future market conditions, future operations, the company's partnering efforts and capital requirements. These forward-looking statements are not historical facts, but rather are based on current expectations and beliefs and are based on information currently available to us.

The outcome of the events described in these forward-looking statements are subject to known and unknown risks, uncertainties and other factors that could cause actual results to differ materially from the results anticipated by these forward-looking statements, including, but not limited to, those factors contained in the Risk Factors section or most recently filed annual report on Form 10-K.

With that, I'll pass the call over to Bassil.

Thanks, Charles, and welcome, everyone. Xencor is a clinical-stage biotech company dedicated to making better medicines based on our world-class XmAb protein engineering platform. Our purpose is solving the hardest science and engineering problems to create new drugs and then developing them to advance the standard of care for patients and their caregivers. Our protein design tools and novel antibodies have been validated many, many times over through our commitment to using strategic partnerships when the best development strategy for a program involves another company. That approach has produced real advances in the standard of care for patients and caregivers, and there are many ongoing programs across oncology, autoimmune and infectious disease poised to continue that into the future.

Xencor's wholly owned clinical-stage development pipeline is centered on 6 programs with differentiated molecules, all built with our protein engineering tools to enable better clinical outcomes and utility. Today, we'll discuss our TL1A portfolio, consisting of XmAb942, our potential best-in-class monoclonal antibody that targets TL1A, an emerging biologic target implicated across a growing number of autoimmune and inflammatory diseases; and XmAb412, a novel bispecific antibody targeting TL1A and IL-23p19. XmAb942 is currently enrolled in XENITH-UC, a global Phase IIb study of patients with moderately to severely active ulcerative colitis and XmAb412 is starting Phase I in Q3 of this year.

During the ongoing Digestive Disease Week conference, we're presenting posters updating our progress on these programs. Our poster for XmAb942 provides PK/PD analysis based on our Phase I data that shows how our XENITH-UC study is testing our best-in-class drug exposure and target inhibition [indiscernible] 10 to deliver better clinical results versus those observed from studies of first-gen TL1A-targeted biologics. We plan to provide a progress update on XENITH-UC around year-end and expect the results of the 12-week induction period during 2027.

We're also presenting the preclinical characterization of our XmAb412 bispecific program. Our poster for 412 details the novel design of our bispecific targeting TL1A and IL-23p19 to deliver dual inhibition of these 2 key inflammatory pathways in autoimmune disease. Our first-in-human study is on track to start during the third quarter of this year and is expected to support moving into Phase II development during 2027.

You can see here the large and growing landscape of inflammatory bowel disease therapies, and we're particularly drawn by the opportunity to address the high remaining unmet need by targeting the new and compelling biology of TL1A. We believe our XmAb tools can bring next-generation, highly potent and long-acting molecules to these patients to improve quality of life through lower injection burden and better target inhibition over the course of maintenance therapy.

Our design philosophy for both XmAb942 and 412 was to address the key clinical pain points with our well-validated XmAb antibody modules and tools. For 942, we built a highly stable antibody with class-leading potency and applied our Xtend Fc domain to get a 74-day half-life in humans, a high-concentration formulation and Q12-week dosing.

For 412, we built a novel bispecific 1-plus-1 format that John will tell you about to minimize immune complex formation and simplify formulation. And of course, we used our Xtend Fc for long half-life. We dialed in ultra-high potency, including sub-picomolar binding. So this bispecific can match or beat existing traditional antibodies. We consistently challenge ourselves to push the boundaries of protein engineering to deliver better medicines to patients.

We believe 942 is positioned to be the best-in-class anti-TL1A antibody, which we think creates a significant market opportunity within the emerging TL1A class for the treatment of inflammatory bowel disease, but also longer term positions 942 to be developed for a broad range of autoimmune inflammatory diseases, like the many currently being studied within the class.

XmAb412 is poised to deliver on the promise of dual-targeted therapy for inflammatory bowel disease, which you can see exemplified by the most recent DUET-UC data, where 412 can deliver potentially class-leading combination potency in a single-molecule format that can follow the efficient clinical and commercial pathway of monoclonal antibodies or oral medications and avoid the highly costly and complex development of combinations and co-formulations, like the nearly 3x size of the DUET Phase IIb studies over standard programs.

Now John Desjarlais, our Chief Scientific Officer, will walk us through the preclinical characterization of XmAb942 and introduce our XenLock platform. Go ahead, John.

Thanks, Bassil. To address this need for bispecifics optimized for autoimmune disease, we aim to make a very high-potency and long-acting bispecifics with low immunogenicity potential and stability needed for high-concentration formulation. To do this, we expanded our modular XmAb platform with our new XenLock Fab format that lets us readily combine full Fab domains into highly stable multispecifics in a native-like IgG format. This allows us to generate binding domains from full variable domain diversity and highly optimize their affinity and stability without the limitations of common light chain or VHH approaches.

XenLock lets us assemble these fabs with our XmAb Fc domains and readily produce molecules with ultra-high affinity into whatever format we like. The result is XmAb412, our TL1A x IL-23p19 bispecific with femtomolar affinity or sub-picomolar for IL-23 and single-digit picomolar affinity for TL1A.

We can see how the stringent engineering that we did to make XmAb412 in a very potent molecule. Cell-based assays for both TL1A activity on the left and IL-23 on the right show the same or better potency for suppressing these pathways from this bispecific compared to leading clinical TL1A and marketed IL-23 antibodies, with XmAb412 shown in the dark blue. The XenLock approach lets us get this very high potency from only half the number of binding domains for targets in the monospecific antibodies. This smaller single binding domain construct means less mass to put into the vial when we make high-concentration subcutaneous formulations, a big help for it.

The other big advantage of this 1-plus-1 ultra-high potency format is the much lower propensity to cause immune complex formation and much smaller size of these complexes. In particular, for trimeric targets like TL1A, antibodies with 2 TL1A binding domains can create large immune complexes, as you can see in the schematic. Large immune complex formation is an important driver of antibody drug immunogenicity. But our 1-plus-1 format does not have that potential for a large complex formation as there's only one domain per drug molecule to stick to TL1A.

We experimentally showed this with size exclusion chromatography. You can see just a single peak for XmAb412 in the presence of TL1A compared to the messy mixture of very high molecular weight complexes formed with a 2-plus-2 bispecific antibody using 2 TL1A binding domains, a common format used to make bispecifics quickly, but they potentially suffer from this liability.

Turning to the in vivo data for XmAb412, we show on the left the long half-life observed in nonhuman primates of greater than 20 days, which extrapolates to an expected 60 to 70 days in humans, possibly enabling every 12-week dosing. XmAb412 uses our Xtend Fc domain for long half-life. This PK gives a very durable pharmacodynamic profile, and you can see that the complete suppression of free TL1A below detection limit has outpassed 2 months from a single dose.

In summary, XmAb412 is poised to deliver potentially best-in-class innovation of both TL1A and IL-23. In addition to potency and extended half-life, we have a high-concentration formulation for simple subcu administration ready to go into our Phase I study starting in Q3 of this year. We expect this trial to validate our XenLock approach to making bispecifics tuned for the needs of autoimmune disease and have a second program in preclinical development behind XmAb412.

Mark?

Thanks, John. Our first-in-human healthy participant study of XmAb942 was a randomized, double-blind, placebo-controlled Phase I study with a primary endpoint of safety, a secondary endpoint of pharmacokinetics and exploratory endpoints of immunogenicity and pharmacodynamics.

We studied 4 different dose levels across single-ascending dose and multiple-ascending dose cohorts. The single-ascending dose cohorts tested 3 different dose levels given both subcutaneously and intravenously in 48 healthy volunteers. And the multiple-ascending dose cohorts tested 2 different dose levels, each given 3 times every 4 weeks intravenously in 16 healthy volunteers.

As expected, XmAb942 was found to be safe and well tolerated in healthy participants with a similar rate of treatment-emergent adverse events between the XmAb942 arms and placebo. Specifically, all treatment-emergent adverse events were mild or moderate. The rates of overall treatment-emergent adverse events were similar in XmAb942 and placebo. There are no serious or severe treatment-emergent adverse events and no adverse events that led to discontinuation of either the drug or the study.

Headache was the most common adverse event, occurring in 33% of participants administered XmAb942, compared to a slightly higher 38% of participants administered placebo. There are only 2 definite treatment-related adverse events, both of which were mild: 1 mild injection site reaction and 1 mild administration site bruise, both occurring in the highest dose subcutaneously.

XmAb942 is found to have high and extended exposure after 20 weeks of follow-up in both single-ascending dose and multiple-ascending dose cohorts. The estimated terminal half-life of 74 days supports adequate drug levels at the 12-week dosing interval being used in the ongoing XENITH-UC Phase IIb study.

We see here dose-dependent and durable target engagement with complex soluble TL1A after 20 weeks of follow-up, which again supports effective target inhibition for the 12-week maintenance dosing interval being used in the ongoing XENITH-UC Phase IIb trial. Reduction in free soluble TL1A was dramatic, rapid and durable, and was very consistent with the pharmacodynamic effect on total TL1A, with levels remaining below the lower limit of quantification through 20 weeks.

XmAb942 has demonstrated best-in-class immunogenicity on rates of antidrug antibody positivity and observed neutralizing antibodies, compared to reported rates across healthy participant studies with first-generation anti-TL1A drugs. And importantly, for the target dose induction and maintenance regimens we are using in the ongoing XENITH-UC Phase IIb study, there were no healthy volunteers with neutralizing antibodies and only 25% had positivity for antidrug antibodies at any time point.

In summary, we remain confident that the innovative design of XmAb942 can deliver a best-in-class clinical profile and supports our optimism for success with the ongoing global Phase IIb XENITH-UC study in moderate to severely active ulcerative colitis.

Now it's my distinct pleasure to welcome Vip Jairath, who I've known for years from our academic work and mutual interest in endoscopic and histological assessment of disease activity in inflammatory bowel disease. As I was an inflammatory bowel disease academic on faculty at the University of Pennsylvania for almost 16 years, I had the good fortune to get to know Vip quite well over the years as we thought through a number of clinical and scientific questions together on many occasions.

Vip is a Professor of Medicine and a John and Susan McDonald Endowed Chair in Inflammatory Bowel Disease at Western University in London, Ontario, Canada, and also the Chief Medical Officer at Alimentiv. He has former training in the U.K., in London and Oxford. And of note, Vip has contributed so much to our field so far in his career in terms of clinical trial understanding and endpoints.

Vip, thank you so much for joining us.

Great. Thanks very much, Mark. Pleasure to be with you here and the rest of the team. Thank you for having me.

The clinical remission rates from large, randomized, controlled trials show that more than half of ulcerative colitis patients do not adequately respond to their advanced therapy, irrespective of mechanism of action, although the data appear to be potentially more promising with the TL1A inhibitor class. We've seen this trend now for 2 decades. Vip, what is your feeling regarding the unmet need with respect to efficacy in UC at the present day?

Mark, I'd entirely agree with you, we've seen this disappointing trend in the last 2 decades in ulcerative colitis. We have several approved drugs with different mechanism of actions in ulcerative colitis where we haven't yet broken this efficacy ceiling. So therefore, we still need new mechanisms for patients because there's still a substantial proportion that do not respond to their initial therapy and will sometimes subsequent therapies that they lose response to.

So I agree with you there's a chance that these data and results could be more promising for patients with the TL1A class, particularly given its more pleiotropic nature in terms of mechanism of action.

Thank you, Vip. As has been well-described through a number of studies across multiple drugs and mechanisms of action, including vedolizumab and guselkumab most recently, there is often a strong drug exposure response relationship with inflammatory bowel disease, both in Crohn's disease and also in ulcerative colitis, with higher drug concentrations being associated with higher rates of clinical remission.

The vedolizumab analysis shown in this slide I did with Takeda while I was in academia, and we see the largest delta between highest and lowest concentration strata of the 3 drugs portrayed here, as we used individual patient-level data from the GEMINI 1 randomized controlled trial and actually adjusted for the 5 variables that are known to most prominently affect vedolizumab clearance.

Afimkibart also, not surprisingly, has been reported to exhibit an exposure-response relationship, with higher drug concentration tertiles associated with higher rates of clinical remission in its Phase IIb UC trial. What is your opinion about the importance of drug exposure optimization in IBD and with the TL1A inhibitor class?

Mark, as key observation, I think it's really likely that we'll see a dose exposure-response relationship with all of the TL1A class because that's been demonstrated to robust data across multiple different mechanism of actions in IBD already, including some of these data that you've shown here. And what we see here is that higher drug concentration levels have been associated with higher rates of response.

What we also know is that immunogenicity to IBD drugs is often seen in the setting of low drug concentrations. So optimization of drug exposure is essential to protect against the formation of antidrug antibodies. Because if you develop these, we know it can impact both the rates of response and durability of response to our drugs.

I suspect that the same phenomenon will hold true for the TL1A inhibitors. And therefore, it's really critical that optimization of drug exposure is likely to be important both to maximize the rates of clinical endpoints, which is clinical remission, but also importantly, those sustained long-term durable clinical remission rates that are so important.

Thanks a lot, Vip. We developed a unified quantitative systems pharmacology, QSP, model, integrating clinical and published data on XmAb942, tulisokibart and afimkibart, and extended it to support virtual population simulations and comparative population-level PK/PD predictions across compounds using the expected pivotal dose regimens for tulisokibart and afimkibart, compared to our target Phase IIb dose regimen of XmAb942.

In the induction phase, our QSP model-based projections indicate that XmAb942 provides markedly enhanced TL1A inhibition, which predicted greater than 99% inhibition in 86% of patients, compared to only 30% to 40% of first-generation anti-TL1A antibodies.

Regarding maintenance, it was also shown that higher exposure with afimkibart was associated with higher response rates in the Phase II UC trial. XmAb942 is designed to deliver maximum TL1A inhibition and is projected to achieve greater than 90% inhibition in 90% of patients versus 60% to 70% for first-generation antibodies during maintenance therapy, despite being dosed much less frequency. These higher drug exposures for XmAb942 during both induction and maintenance position our molecule very well to test whether increased exposure improves efficacy.

Vip, what is your impression of our QSP modeling data and the potential efficacy profile of XmAb942?

Yes. Thanks, Mark. I'm frankly impressed with these modeling data for XmAb942. And my reading of this is that Xencor have really optimized the molecule's potential for getting high drug exposures, which are essential, as we just said, both during induction and maintenance. And ultimately, the importance of that is it could translate into high rates of clinical response and remission during the induction period, but then sustained during the maintenance period. And perhaps this will give lasting durable response for patients.

As I mentioned earlier, I do think there'll be an exposure-response relationship for the TL1A class of drugs. And these model differences in TL1A inhibition between XmAb942 and these leading first-generation TL1A competitors is substantial and is compelling. My reading here is that XmAb942 appears to be well positioned and it potentially could be a best-in-class molecule in terms of efficacy. And really I'm excited and the community's excited to see the results of XENITH-UC.

Thanks a lot, Vip. XmAb942 has a terminal half-life, as we mentioned, of 74 days, which not only is approximately 4 times as long as first-generation anti-TL1A antibodies, but also enables, at minimum, every 3-month dosing, which is what we are studying in our Phase IIb UC trial. Compared to the first-generation anti-TL1A antibody competitors, the number of subcutaneous injections administered during maintenance therapy with XmAb942 is significantly less. Vip, how do you feel that this difference in injection burden will be received by patients and their physicians? And what impact do you think that XmAb942 could have on overall patient experience and satisfaction?

Yes, this is a key point. We know from many patient preference studies, and I know from my clinical practice of treating many patients with IBD, that injection burden is important and it matters to patients, also matters to the treating physicians. The key thing here that stands out, there's a markedly reduced injection burden with XmAb942 compared to first-generation anti-TL1A antibody competitors. And this has to be an attractive feature of this next-generation drug. And I think ultimately what I see is this has been enabled by the technology of your protein engineers and they've been able to manufacture this XmAb942 to lengthen its half-life so much.

The infrequent dosing requirement is not only very convenient for patients, but it's also a really nice addition in terms of the potential unpleasantness that patients sometimes report having to inject themselves with the needle. I'm sure patients will almost certainly welcome this improvement compared to first-generation anti-TL1A antibodies. And really taken together with the optimization of drug exposure to help maximize efficacy, I believe that XmAb942 is really uniquely poised to be a best-in-class drug.

Thanks so much, Vip. Appreciate that. So we have designed a proper and rigorous dose-ranging, double-blind, placebo-controlled, randomized Phase IIb trial in ulcerative colitis that can inform a single induction dose selection for Phase III. There are 3 active drug arms, each separated by half logarithm and dose, compared to placebo. Also, there's asymmetric randomization such as the highest dose arm will include the most patients. And in fact, overall, 77% of patients during induction will receive active drug and only 23% will receive placebo. Even the lowest-dose arm is designed to achieve greater than 90% target inhibition over the dosing interval, such that we expect some efficacy even at this lowest dose.

The doses have also been selected to have the highest dose maximize drug exposure to potentially achieve greater induction efficacy than has been observed in the competitor trials. In addition to the convenient and patient-friendly every 12-week maintenance dosing, all patients will receive active drug at week 12 for up to 2 years as there will be an open-label extension for at least an additional year. Vip, what are your thoughts regarding this Phase IIb UC trial design?

Thanks, Mark. As you know, I spend a lot of my time thinking about these things, and I think you've really designed an excellent trial, state-of-the-art trial, it's been very well thought out. A number of reasons I say this. First, it's well-crafted to do a full exploration of induction dose ranging. And that will give you confidence in the results to inform a single-induction dose to be selected for Phase III. And there's a number of other aspects that bring rigor to this Phase IIb.

Secondly, your trial is attractive to patients, I think, for a number of reasons. First thing, obviously, the high probability of getting the active drug. The second thing is the drug doses have been well selected to potentially lead to efficacy, with the highest dose potentially maximizing efficacy due to optimized drug exposure. Third thing, there's only a 12-week placebo period, which is attractive to the sites in terms of recruitment. And the fourth thing is there's very convenient and patient-centric Q12 maintenance dosing.

In addition, there's an open-label extension that enables up to 2 years of free drug for patients. And the final and sixth thing here is I think it's a relatively derisked MoA, given that we've seen 3 positive Phase II trials already of the first-generation anti-TL1A antibodies in UC. So given all of these features, I really expect this to be an impactful trial and XmAb942 will be an impactful drug.

Thanks a lot, Vip. And now as a glimpse into the future, at this DDW, we showcased a poster of our preclinical data of our TL1A x IL-23p19 bispecific molecule, as John has already talked to us about XmAb412. Our protein engineers have built a fantastic molecule in a 1-plus-1 format that enables sub-picomolar affinity. We purposely chose to target p19 rather than p40 for IL-23, given the superior efficacy of p19 blockade over p40 blockade in IBD and also in psoriatic conditions.

XmAb412 was found to robustly suppress both TL1A and IL-23 inflammatory pathways. In cellular assays, XmAb412 demonstrated IC50 values comparable or superior to clinical-stage TL1A inhibitors and approved IL-23 inhibitors. Allometric scaling predicts that XmAb412 will have a half-life between 60 to 70 days in humans. As in nonhuman primates, they achieved a half-life exceeding 20 days, with similar target engagement to monospecific antibodies. XmAb412 also supports a high-concentration, low-viscosity, citrate-free formulation suitable for subcutaneous dosing. Evaluation of XmAb412 in healthy volunteers is expected to begin in the third quarter of this year.

Given the exciting results of J&J's DUET Phase IIb trials being presented here at DDW today, in fact, especially in multi-MoA refractory patients, where the combination of guselkumab and golimumab was found to be meaningfully superior to either monotherapy, we present here the DUET-CD data on this slide. What are your views, Vip, on the potential transformational efficacy that could be achieved with the combination of TL1A with IL-23, the safety profile of such a combination, and the streamlined regulatory and clinical development path offered by a bispecific in this context?

Thanks, Mark. Look, I think the combination of TL1A and IL-23 inhibition is probably the most exciting prospect I see in our field and potentially even more so than anti-TNF and IL-23. And that's given the possibly even more pleiotropic anti-inflammatory and even potentially antifibrotic potential that we see with TL1A inhibition. I expect this combination could be truly transformational and has the potential to yield efficacies that we've not yet seen in IBD and possibly even other immune-mediated inflammatory diseases.

Equally, in this situation, safety, such a combination has to have a favorable safety profile. And I think that's a potential here like JAK inhibitors where we've seen differential efficacy in some indications, but with some safety concerns that make that a long-term use far from ideal.

Given the very good data here showing superiority of combination of TNF IL-23 inhibition versus either monotherapy in predominantly a bio-naive UC population and now the DUET data showing superiority of this combination in the more refractory patients, I do feel that this combination of TL1A and IL-23 could be a very strong first-line and later-line option for patients to deliver transformational efficacy, but in a safe form.

Your new bispecific XmAb412 seems to be expertly constructive with very impressive affinities for both TL1A and IL-23. And I'm glad that you actually chose p19 as your target rather than p40 for the IL-23 blockade given we've seen superior efficacy in IBD and psoriatic conditions that you mentioned by targeting p19.

On top of this, the very long expected half-life and promise of subcutaneous delivery are really the icing on the cake here to make this an ideal drug in many ways. That doesn't even take into account the fact that there are large and distinct advantages from regulatory and clinical development standpoint of testing a bispecific compared to a combination or co-formulation, as comparisons to each monotherapy are not a requirement for a bispecific entity, which not only improves the probability of technical and regulatory success for clinical trials of bispecifics, but it also saves an enormous amount of time and money and expediency in the development programs.

So I'm really looking forward to the commencement of your Phase I program for XmAb412 later this year. And I'm very much looking forward to a Phase II program in 2027. I can say that many of my academic and community IBD colleagues around the world are really eagerly awaiting a Phase II trial of XmAb412 and IBD and really hope to be included as investigators in this program also.

Vip, we really appreciate your time and thoughtful comments. Thank you so much for joining us.

Thank you. A real pleasure.

It's always great to see you, my friend.

Next, Dane Leone, our Chief Strategy Officer, will speak to TL1A strategy and development road map. Dane?

Thank you, Mark and Vip. Great conversation and really appreciate all the insights. Today we've discussed the key points to why our TL1A pipeline is primed to deliver the next generation of advanced therapies for patients living with moderately to severely active inflammatory bowel disease. Our focus on novel best-in-class therapies targeting TL1A aims to deliver both an improvement in clinical outcomes over today's drugs while also providing the convenience of less frequent injectable dosing for a best-in-market biologic class.

We view the future landscape of innovative medicines in IBD to be defined by branded TL1A monotherapy, the emergence of bispecifics and oral inhibitors of traditional biologic targets. Biosimilar versions of today's branded medicines will also continue to play an important role as more affordable monotherapy drugs targeting IL-23 and integrin inhibitors will add to the armamentarium of cost-effective options for patients as the TNF class does today.

Looking ahead at our road map, we are hard at work executing on our time line to enable exciting data-driven decisions during 2027 for the next phase of development for our TL1A pipeline. As you can see, we expect to report the primary analysis set of XENITH-UC 12-week induction period during 2027 and use that data in tandem with the first-in-human XmAb412 study data to optimize our investment across our TL1A pipeline for the next phase of development.

Longer term, we expect the full data of XENITH-UC along with early looks at Phase II patient data for the XmAb412 program to provide us with rational decision-making around planning registration-enabling studies. Of course, data will allow us to evaluate opportunities to optimize the development of our TL1A pipeline through strategic relationships.

Going back now to our overarching business strategy, as Bassil provided in his intro, we view our commitment to delivering innovative medicines as part of a life cycle that we also assess through the lens of business development opportunities. We think there is good precedent that programs delivering on the potential of better medicines in IBD to have optionality when it comes to strategic relationships, and we plan to make these decisions driven by our data as our programs progress.

Our XmAb412 first-in-human study is on track to start during the third quarter of 2026. And for XmAb942, we expect to report an update on enrollment progress around year-end, along with the recommendation of the Data Monitoring Committee on the prespecified blinded interim analysis for XENITH-UC.

2027 will be another exciting year for both programs with the interim results of the first-in-human study of 412 expected during the first half of 2027, which will help validate the pharmacokinetics and pharmacodynamics of the novel bispecific platform that we now call XenLock. The second half of 2027 will provide a clear answer to the pharmacological thesis of XENITH-UC, whether best-in-class drug exposure can support best-in-class clinical results with XmAb942.

In conclusion, the hard work and dedication of the Xencor team to delivering better medicines to patients living with inflammatory bowel disease is at a clinical inflection point, and we are committed to continuing to rapidly advance our TL1A pipeline in the coming years.

And with that, we're ready to take questions. Operator?

[Operator Instructions] Our first question comes from Jonathan Chang with Leerink Partners.

This is Albert Agustinus on for Jonathan Chang. And congratulations on the data. How do you see this 942 data fit into your plan to expand the program into Crohn's disease and indications beyond IBD? What do you need to see from the program for you to make the decision?

Thanks, Jonathan. I'll take that question for the team. It's a great part of the road map that we have and the flexibility to make these decisions that are data-driven. So as been part of the plan since the initiation of the 942 program, we've looked at UC as the first proof of concept on drug exposure, given there's clear benchmarks of what we can deliver with a best-in-class investment market biologic.

So as we think about moving towards the primary analysis set of XENITH-UC in 2027, that can trigger or catalyze additional investment and, obviously, parallel development pathways into Crohn's disease, as has been very traditional with many other IBD programs from our peers. So we feel very confident about the profile here. That said, to be rational about our data-driven decisions, we do want to see that data in the back half of 2027 as a trigger point.

Our next question comes from Tara Bancroft with TD Cowen.

Congrats on the data. This is Ikenna on for Tara. So across the TL1A class, we've seen clinical remission rates pretty high. You guys showed it on the chart there. Given your modeling suggests a deeper and more durable TL1A inhibition for 942, how do you think about the extent to which that could translate to higher remission rates in the Phase IIb? Like where is the ceiling?

So that's a great question. Before I turn it over to Mark and Vip to comment, I don't think anybody knows where the ceiling is. There are some glimmers of data implying there's some directionality from existing studies like TUSCANY-2B, maybe I'll let you guys opine on that. But yes, there's an element of uncertainty inherent, which is why we're so excited about the XENITH-UC experiment. What do you...

Yes, I'll start and turn it over to my friend. Since I'm no longer the KOL. But we see for many of our drugs in IBD, both in UC and in Crohn's disease, that there's efficacy left on the table. We see it both in clinical trials and we see it clinically with patients every day. And then we are confident with the design of this molecule that we will capture additional efficacy given that we can maximize exposure to this drug.

As far as where that ceiling is for this MOA in a completely optimized manner, it's unclear. There are some patients who will never respond to a TL1A drug because their pathobiology is different. And at the current time, we don't have a great way to individually pathotype patients to find out what's driving each of the diseases and which pathways are most important for them. So I think it's unclear what the actual ceiling will be. But I'm confident that there's efficacy left on the table that could potentially be recaptured with this drug.

Yes, I'd agree with that, Mark. I mean I think we've seen with some of the other assets, that we've seen consistent data, which have been 20% to 25% deltas over placebo, which is really better than anything that we have today. And obviously, JAKs are up there with that kind of efficacy, but there's some trade-off with safety. So I think really that's the sort of marginal delta that we've seen in the existing programs. But as you've outlined, there are some potential additional attributes of this that could move that needle a little bit more, and that's incremental and important for our field to get to those benchmarks.

Our next question comes from Etzer Darout with Barclays.

Just one for me. I'm curious about the relative positioning of 412 relative to 942. Really, is just 412 meant as a follow-on or next-gen path, or maybe aiming at a different patient population, refractory disease or TL1A responders versus nonresponders? Just how you're thinking about the 2 molecules coexisting in IBD.

Yes, Etzer. So I think the TL1A class is still revealing itself where it can have an impact. I mean if you recall from our earlier slide, there's more than half a dozen additional autoimmune and inflammatory indications where TL1A is being tested with really very good underlying preclinical and genetic rationale for why it ought to be a good target in those diseases.

So I think when we look at 412 and 942, this TL1A mono and TL1A plus p19 mechanisms, we see sort of, first, an ability to address all of these other indications where p19 might not be a target of use. That's a great additional benefit of 942 as a monotherapy. Of course, there are some other non-IBD indications where IL-23, like in derm, can be very useful.

So first, there's sort of just the ability to be positioned for the long-term future of a broad potential pipeline in a product. Maybe within IBD, Dane, if you can comment on more of the subtleties about how we're using our experimental data coming out of our clinical trials.

Yes. No, absolutely. And I think there's kind of a short-term window of the answer here and then maybe a longer-term window of the answer. So the short-term window of the answer is with 942, obviously, we wanted to provide a best-in-class monotherapy biologic that can be optimized for clinical convenience as well as drug exposures. And I think that's what we're delivering with the 942 program, which gives us a really clear pathway into what we think the next great branded drug class will be for IBD specifically.

But as Bassil mentioned, the opportunity for the TL1A class is rapidly expanding with over an $80 billion potential TAM outside of IBD as it currently stands today in a number of our peer proof-of-concept studies. We'll be watching those studies to see how they read out and feel confident that the clear and thorough dose characterization that we're doing in the XENITH-UC study ongoing today with a readout in 2027 will allow us to actually move into other indications much more rapidly through that knowledge of our PK/PD and dose characterization as it relates to inpatient study with clear disease in a controlled study.

With XmAb412, as Bassil said, that's a very clear, almost rifle shot at certain indications that are implicated in TL1A and IL-23, which are very large indications even outside of IBD. For 412, it's also validating our next-generation bispecific platform that we call XenLock. And so there's a duality there of the importance of the validation of our first-in-human studies for 412 that will have first data in the first half of 2027, allowing us to really invest in that program and platform.

But as we said in the survey and market research we've done with gastroenterologists, 412 represents a real position in the market to help clinicians provide better care for their patients that are not only looking for a best-in-class, next-in-class biologic to use as a first-line agent, but as we said, many clinicians in our surveys and market research would also use a bispecific approach or multi-specific approach when the patients are inadequately responding or nonresponding to a frontline IL-23.

And so there's a lot of different opportunities within our portfolio. But yes, you're absolutely right in your question that 412 and 942 will continue to differentiate from each other as our clinical portfolio advances.

And can I piggyback on another...

Absolutely.

So based on sort of looking at this thing as a clinician, as an epidemiologist, even within the IBD space, there is a home for both of these drugs. Now remember that even though you might have more differentiated efficacy with the combination therapy on a population level, on an individual base level, individual pathotyping, many patients do not require combination therapy in clinical practice or in clinical trials. So there is a home even within IBD space itself for both drugs, that they can both live in the same space and serve different patient populations. I don't know, Vip, if you want to...

Yes. No, I fully agree. I think where bispecifics or the combination assets fit into the treatment paradigm is evolving. I mean, obviously, we're seeing direct results today, which is in a refractory population for a co-formulation. But I think from a bispecific population, there will be patients where you might want to use this first line and patients where you might want to use it downstream for those specific reasons.

But absolutely, I see a space for both of these, and where they land in the treatment paradigm will evolve. But it's a combination of both first line and in refractory patients.

Our next question comes from Timur Ivannikov with Cantor Fitzgerald.

This is Timur on for Steve Seedhouse with Cantor. So can you just talk a little bit more about how you will use 942 monotherapy data and 412 Phase I data to make a follow-on development decision? Just in terms of immunogenicity or inhibition data that you're looking at, since it's not the mechanism is not completely apples-to-apples, you're inhibiting 2 different pathways with the bispecific. So if 412 looks similar in terms of inhibition of the TL1A pathway, how would you position it in terms of 942?

Yes. So it's our data where we're going to be looking for 412 to see what kind of half-life and target coverage it gives us in terms of durability and depth. That will obviously have an important role. And with 942 verifying our thesis of better drug exposure leads to better efficacy is a part of that thinking package.

But remember, there's all of the work going on around the industry validating, or potentially not validating, TL1A in a range of other indications. And so those pieces are all going to come together in that 2027 time frame and help us make decisions how to maximize the value of both assets. That's sort of the big picture. I don't know if there's anything to add to that or.

Yes. I mean the technical point, Timur, is in 2027, we'll have a coinciding of both the first-in-human healthy volunteer data for 412. As Bassil said, will give us PK/PD immunogenicity check on a novel bispecific platform and the drug properties of 412, which we're very excited about. But also we'll have that 12-week primary analysis set and induction data in ulcerative colitis from the XENITH-UC Phase IIb study of 942. And that will help us understand of how to optimize both of those development pathways, potentially in IBD or other areas.

But then you look ahead, we're looking to rapidly move 412, if all goes well in the first-in-human study, into proof-of-concept Phase II studies in patients. And that data we hope will also coincide with the maintenance period readout of the XENITH-UC IIb study, which will occur a little later after the primary analysis set. And those 2 data packages together will be really powerful in informing us on how to optimize both of these programs for longer-term success in registration-enabling studies.

Our next question comes from Sean McCutcheon with Raymond James & Associates.

On 412, can you maybe speak a bit more to the stoichiometric considerations for the bispecific versus combination of monospecifics and your conclusions from the preclinical data that drive confidence in translating this into humans? And maybe to be a bit coy, do you expect the dosing to be driven by the TL1A component?

Yes. That's a good framing of the question. So the beauty of bispecifics is that you don't have to address all the different ways you might dose a combination together. Now someone might flip that around and say, "Oh, I can't optimize the dose of the 2 different arms."

A part of the reason why we really look, zoom out and look at why did Xencor pick this pair of targets as its first real autoimmune bispecific, something that wasn't a T-cell engager, it's because the stoichiometry has been driven by the biology here. We've seen a very, very large amount of data in humans across a decade of the excellent safety record of IL-23p19 inhibition. And in fact, better drug exposure, higher doses, you see this in post-marketing studies, higher doses give you potentially better efficacy as you get more exposure, but also the safety, the excellent safety track record seems to be maintained. Just this past weekend at DDW, I've seen a number of talks about that.

So you've got one drug class that's incredibly well validated to be very safe where more drug seems to maintain that safety. With TL1A, certainly less data is out there. But we've seen across the class now 3 different programs, different drug molecules, different antibodies with different designs, also have just an outstanding track record of safety in all the reported data.

So we have an opportunity here to take 2 targets where we can just turn both knobs to 11, right? And maximize drug inhibition in the molecule and really truly take the best advantage of a bispecific approach and avoid having to do unnecessary dose-ranging and combination sort of dose matrices.

So that was the basis of our design, just push the limit on the affinity, which pushes the potency because these targets are validated to be very safe and more inhibition is believed to be better. I think that's a pretty universal belief. I mean, I don't know, Vip and Mark, if you would agree with that thought. I see them nodding good.

Yes.

So when we look at the preclinical data, aside from the affinity, sub-picomolar on the p19 side, about 1 picomolar on the TL1A side, John and his team did just a remarkable job of design. And the XenLock piece enabling that, as John said. When now you're looking at nonhuman primates for the other half of it, exposure is also about half-life, we see a 20-day half-life that essentially matches the half-life in nonhuman primates we saw for 942, which translated into an over 74-day human half-life. So we're really optimistic that we're going to get that kind of durability and coverage, hopefully, the patient convenience.

And the last piece of the puzzle, when you think about actually making a practical drug, you've got to actually make it. You've got to make kilograms of it, put it in vials, ship it all around the world and then have it be easily injected into patients in a subcu format. The XenLock format put on top of our long track record of XmAb Fc domains -- bispecific Fc domains, we've been able to produce this in standard antibody manufacturing processes at scale and we're delivering drug now to clinical sites in the next couple of months in a very high-concentration formulation with yields that we're quite proud of.

So all the pieces of a biophysical optimized molecule with the right underlying biological thesis that more is better for these 2 targets, I think, really came together beautifully. So it's the perfect first case for an autoimmune bispecific on the biology side.

Our next question comes from Gregory Renza with Truist Securities.

In your preclinical benchmarking studies, you've shown that XmAb might be superior to first-generation TL1A. Just curious if you've done similar experiments with second generations like SPIs, agents, or other assets, you know, out of China. How does it look compared to those? And if I may tag along one more question, how do you view the potential commercial trade-off between your 12-week -- once every 12 weeks, versus maybe a more aggressive or less frequent dosing like Q12-weeks to Q24-weeks, that your second-generation peers are marketing?

So simple one, the first maybe 1/3 of your question was inaudible. I'm sorry, could you possibly repeat your whole question? I just want to make sure we address it properly. We really couldn't hear about the first 1/3.

Sorry about that. Okay. So basically, I'm just curious if you've done any preclinical benchmarking for XmAb942 to the second-generation TL1A. I mean it looks better than first generation. Just curious how it compares to the second-generations TL1A agents. Yes.

Okay. So a benchmark 942 versus other second-gen agents. Well, I mean, we're aware of some data from peer program on initial half-life disclosures that look promising based on using long half-life Fc domains. I don't know that we've seen from any second-generation TL1As any immunogenicity disclosure from human data. I do believe that from published preclinical data, XmAb942 has the highest affinity, and that should translate quite directly to higher potency, again, based on preclinical data of inhibition of the target TL1A on human cells.

I don't know that we've seen any other details of early clinical characterization of other second-gen TL1As in any kind of thorough way. Obviously, we await further data. We know it's a very active class. We feel very confident about our positioning and the very thorough job we did maximizing and making best-in-class potency and half-life for the program as well as a very robust Phase II design.

I think there's a philosophy component here. Xencor sat down to make something that could solve a clear clinical problem and unmet need in inflammatory bowel disease. And we really like the TL1A target from emergent data 3 years ago. And so as world-class protein engineers with a protein engineering platform that's been validated many, many times over, we had our heart and our passion to make these molecules as good as they possibly could be.

And I think it's hard for third parties or third-party protein engineering platforms to have that same commitment and dedication when they're making molecules for other people to develop that they'll never have a part in the clinical development. Our research team walks hand-in-hand with our clinical development team on a daily basis to design the best molecules and solve the clinical problems. And we feel confident that the molecules we make across the entire spectrum of our portfolio are best-in-class designed and delivered medicines at the end of the day.

Now the other component of this is of the second-gen class, we're ahead of everyone else by a long shot in the development pathway to market. That's why we engaged a Phase IIb study right out of the gate from a first-in-human healthy volunteer study and have executed that relentlessly. We started in clinic first-in-human in November of 2024 with XmAb942. Less than a year later, we were in a global Phase IIb study, and we're on track 2 years from that point to deliver a Phase IIb study results that will be registration-enabling in the traditional efficient pathway of biologic drug development. So we feel very good in our market position of delivering the best-in-class and best-in-market biologic targeting TL1A.

Our next question comes from Madeleine Stone with William Blair.

Great. This is Madeleine on for Matt Phipps. One on 9 -- or on 412, actually, could you maybe discuss a bit more on how confident you are that 412 will not lead to an immune complex formation that could drive autoantibody response?

Yes. So what we're very confident of is that the very well sort of understood and accepted possibility of a trimeric ligand like TL1A or other members of the TNF super family, like how a trimeric ligand when you bind it with a bivalent antibody like your standard Y-shaped antibody, that's going to create this ability to form -- easily form these clusters, right? Where you have TL1A trimer 1 binding on the left side of the antibody, TL1A trimer 2 binding on the right side. Meanwhile, TL1A trimer 1 has another antibody in its other side, trimer 2 has yet another antibody in its other side, and you get these big clusters like we had that picture of.

It's simply not, not possible for these multi-molecular clusters to form when you only have one TL1A binding equivalent on your antibody drug. Obviously, you have to make sure your antibody drug is stable and soluble, so it itself is not prone to aggregation, right? So if it is aggregating itself, you can have these artifacts emerge.

So we always strive to make the most stable and soluble antibodies. It's just like the bedrock biophysical underpinning of all of our design work, which I think we've amply demonstrated over the decades. You will never completely eliminate the ability of the immune system to surveil anything you put into the human body. That's just a predicate, right? You put an antibody -- you put a drug into the human body, it will notice. What we believe we've done is greatly significantly reduce the ease of that recognition by the immune system relative to what happens when you have bivalent molecules against a trimeric ligand. So that's the essence of that one design choice that we hope skews the probabilities in the favor of XmAb412.

Our next question comes from Eva Fortea-Verdejo with Wells Fargo Securities.

A quick one from us. Are you considering other potential mechanisms to combine with 942 beyond your TL1A IL-23 bispecific strategy?

So that's a great question about how the future might unfold. There are certainly a lot of other mechanisms that we can imagine combining 942 with as the development of the molecule advances, as we establish our dose efficacy relationship as we hope to do in XENITH-UC and push the program forward. There are certainly other bispecifics we have on -- in our preclinical development and as well as on the design table for a variety of targets using this new XenLock format that lets us get these really high affinities into simple -- simply engineered and produced molecules.

I think our current focus is establishing the efficacy response relationships, hopefully pushing that ceiling of efficacy higher, and then establishing how 412 fits against 942. I think there's just so much uncertainty about how that future is going to unfold. We're just going to wait and see. I don't know if there's anything more concrete we can add on that.

I think from a market research perspective, it goes back to the point we made in the presentation. We believe TL1A is going to be the branded anchor of biologic monotherapy and combination therapy in the future, which is a good thing for our portfolio and retains a lot of utility for the monospecific in development, even though we're also taking this very targeted shot with XmAb412 on TL1A and IL-23p19 biology.

That said, we also don't really envision a future where combining 2 biosimilar targets together in a co-formulated approach is going to be necessarily viable as I think for the benefit of the payers, the system, the patients and their caregivers, as is today already used, you can take 2 biosimilar targets and different mechanisms off the shelf and provide them to your patients in the clinic. And it would be kind of a hard argument to make why someone would pay a branded price that's more expensive than the individual components for their patients.

So we feel very good about TL1A being the right anchor to your question, and fully agree that there will be different targets to engage with it as data continues to evolve in the clinical setting.

Our last question comes from Alec Stranahan with BofA Securities.

This is Matthew on for Alec. On ADAs for 942, the poster pointed to 57% incidence. Can you maybe speak to the kinetics where you saw the decrease in incidence and magnitude with increased drug exposure? Is this driven by small and baseline variability? Or is there something, different hypothesis going on? And then for XENITH-UC, you noted the target dose is at 25%. I guess, is that just based on dose selection or were there other changes to the trial design or patient characteristics?

Well, I'll take the first one on ADA. So this was a healthy volunteer study with an n of 6 per cohort. We did 8 different cohorts, so you can sort of -- IV subcu and we did MAD cohorts. And at the 3 different dose levels tested, both SAD IV subcu and MAD, it was a very consistent pattern of as you gave more drug, the magnitude of your ADA signal and the frequency of positivity went down. So I think the consistency across IV subcu SAD/MAD makes it very clear, that's a real signal. And it's frankly utterly unsurprising given that that is a very, very common, almost the dominant pattern seen in biologic drugs, particularly monoclonal antibody therapeutics in humans. So the more drug you give, the more you sort of swamp out an immune response and tolerize. So that we're very confident of.

And we're very pleased that we saw overall ADA rate, I think it was 57%. It's lower than reported from 2 of the leading first-generation TL1As, and our NAB rate was as well significantly lower than the ones that have reported. So we're happy with that. And furthermore, when we look at our high-dose go-forward dose for XENITH-UC, so the highest induction dose and our selective maintenance dose, it had an even lower ADA rate of 25%.

So we think we're very well positioned with a very favorable profile relative to everything that's been reported by other people in their Phase Is. And that speaks to the quality of the design of the molecule because the underlying biophysical stability of your molecule is always the first thing that your body's immune surveillance has to -- first immune surveillance sees.

So maybe on the second question, I know, Dane, you were saying, you want to touch on that?

Sure. Yes. And obviously, looking forward to seeing you guys at the BofA Healthcare Conference next week, first and foremost. But to your actual question, immunogenicity was one of the key things that we aim to solve for from an engineering perspective when we were looking at the liabilities of the first-generation TL1As. And as Bassil said, our focus on stability, epitope selection and overall kind of design, one would say, of the molecule, I think, has proven out to be best-in-class immunogenicity.

The question from you and many others has been, can you make a molecular target of TL1A in a biologic, have acceptable immunogenicity rates that we can feel confident are not going to be deleterious to efficacy? As is the concern with some of the first-gen TL1A molecules. We think we did it, right? So to the point of an overall rate of 57%, which is lower than the 65% report for tulisokibart and the 98% report for afimkibart, we don't have that liability. And then kind of driving to the point of the dose selection and dose regimen that we're actually using in the XENITH-UC IIB study, we're using obviously a high-dose induction IV along with a single subcutaneous maintenance dose. That dosing regimen has 0% neutralizing antibodies and only a 25% ADA rate.

So we feel very confident of the drug exposure that we're going to get in XENITH-UC not being impacted by immunogenicity, and again, very clearly able for us to elucidate whether best-in-class drug exposure will drive best-in-class clinical efficacy.

We have no further questions at this time. I will now turn the call back over to Bassil Dahiyat for closing remarks.

Thanks very much. And we'd really like to thank everybody who's joined us on this call. We're very excited by what's coming next for our TL1A programs. And we are really looking forward to updating you as we advance XmAb412, our bispecific, into first-in-humans later this year and as we read out data from XmAb942 in 2027. And we're also looking forward to initiating Phase II studies with XmAb412 in 2027. So a very busy year upcoming for us.

And thank you again. Look forward to speaking soon.

This concludes today's webcast. You may now disconnect.

Hello again, everyone. My name is Etzer Darout, one of the senior biotech analysts at Barclays. My pleasure to welcome Xencor to our fireside chat. With us today, we have Bassil Dahiyat, President and Chief Executive Officer; Dane Leone, Executive Vice President of Strategy. And so thank you guys for joining us today. Maybe, Bassil, just to kick us off, just give us a broad overview of Xencor and some of the milestones that you would like to hit in 2026.

So Xencor is a company that was founded around really world-leading protein engineering technology. We've established ourselves over the last couple of decades as people who can create antibodies that can truly have best-in-class properties. There's multiple marketed products we've built. Our technology is used widely around pharma to have created a broad pipeline in oncology, autoimmune disease and other areas.

Right now, we're focused on building our own internal proprietary pipeline of drugs that we've built with these techniques, and we focused in solid tumors, particularly T-cell engagers for solid tumors, a class that's really emerging that we've managed to put ourselves into the forefront of and then best-in-class bispecific and long-acting antibodies in autoimmune disease. Critical for us in the next 12 to 18 months are define fully the dosing regimens and efficacy safety profiles for our 2 lead solid tumor T cell engager CD3 bispecific antibodies, XmAb819 and 541 that we expect to have recommended Phase III doses and pivotal plans by the end of the year, so we can hopefully in 2027, initiate pivotal trials for those. And then as we roll forward into 2027 for our TL1A bispecific and long-acting monospecific have key readouts that are going to really set the stage for their value creation in late phase.

Great. And maybe just kick us off with 819. We're going to have some data, as you mentioned, in the second half of this year. You think about sort of the incremental data update that we could get there, what ultimately strengthens your conviction around pivotal studies in 2027 for that program?

So XmAb819 is our ENPP3 CD3 bispecific antibody. We built it with the format that we designed to help separate out the ability to kill high ENPP3 expressing tumor cells while sparing lower ENPP3 expressing normal tissues. So there's a pretty good spread, but you have to make an antibody that can separate those 2 out. We do it with our 2+1 format. I won't go into the technical details now. We published on that quite a bit. And I think our early clinical data that we published at the triple meeting last year demonstrated really once we deal with the on-class CD3 bispecific on-class effect of CRS in the early weeks as you prime the patient, the tolerability of the agent was excellent after that first month, indicating that it's likely that our design achieved what it wanted to avoiding impacting healthy tissues.

The efficacy there was quite promising, 25% response rate in the most heavily pretreated population seen in renal cell carcinoma. 4 prior lines median, over 1/3 of patients that had prior HIF-2 alpha therapy, which is an emerging class has never been seen before. So that generated a lot of enthusiasm in the investigator community. We have lots of patients who want to come on study, and that's allowing us to deliver by the end of this year, we believe, the kind of data that is going to define really the value proposition we expect would be a significant lever on the community's awareness of what we're doing and the value of it and define our Phase III because early on, when we were just getting the ball rolling last year in 2025, we pulled data from cohorts that had just reached active doses to get an overall sense of the drug and its tolerability and its efficacy. Again, really exciting in that very late line.

This year, we have now the opportunity to characterize in detail with solid numbers of patients, specific dose levels because we've continued to escalate so we can choose which dose level is the right one for Phase III define the priming regimen really carefully and have that specific granular data that shows people we understand the drug, we see its efficacy. As we've been able to enroll more patients, we've really solidified that. And I think we should start to see evidence of durability, at least from some of the cohorts that we had earlier on like late last year, early this year, looking at duration of response to start defining that real winning profile. So I'd say that's really where we think the data this year could be very telling.

Great. And as you think about sort of the potential to move into earlier lines of therapy, as you mentioned, the 25% response rate that you saw in a very late line setting. What gives you more confidence around the ability to move to earlier lines? Is it going to be the durability signal, maybe seeing sort of, again, the response rates in heavily pretreated population, the same? Could we get patients that have less pretreatment where we can maybe see maybe higher responses? What ultimately builds that thesis around...

So it's predicated on the fact that we have a novel mechanism in renal cell carcinoma, right? It has been IO -- sorry, it's been PD-1s. It's been VEGF receptor TKIs and now HIF-2 alpha, which is essentially a tyrosine kinase inhibitor or an enzyme inhibitor as well. We bring a cytotoxic antibody recruiting the immune system into play that's targeted, right? In clear cell renal cell, it is nearly uniformly expressed in almost all patients and other subtypes a little less. So that novelty and the unique profile we have where after priming doses where you get to the CD3 CRS event, it's a very well-tolerated agent gives us a unique opportunity to move forward in line. So with that sort of scientific basis, I'll let Dane talk about how we envision the initial work going into pivotal and how that we can move up and how we're already doing some things there.

Yes. And I think a lot of it comes from what we've seen with the ENPP3 expression levels in clear cell histology and why we're developing a targeted T cell engager in an unselected patient population. So we're not prescreening by any cut point of ENPP3 positivity because the overall positivity rates are so high in this patient population. You have to remember that the monotherapy activity that we've seen in the profile that we presented in October last year, where we achieved a 25% response rate is in really heavily pretreated patients. So a median 4 prior lines, over 2/3 or about 2/3 had 2 or more prior TKIs, about 1/3 had prior HIF-2 alpha inhibitor, all IO, all TKI.

And what that means is the phenotype of kind of the tumor microenvironment invariably changes, right, over those lines of therapy. Our investigators have been very vocal that given the novel mechanism and the first time we've seen kind of an IO-like mechanism work post PD-1 effectively in clear cell patients, we need to explore the true up line monotherapy potential. And so we're working on a protocol right now to engage a substudy in pre-TKI or TKI-naive patients that progressed on IO frontline therapy. Classically, that's defined as a patient that's intermediate to poor risk and is likely progressed on ipi/nivo.

What that does for us is that gives us an opportunity to look in a substudy, a fast substudy, what the monotherapy potential is truer to that original biopsy where 90% of patients have at least adequate ENPP3 expression levels to respond to 819, our therapy. That could be something truly transformational to the entire development landscape. And it's -- for us, getting that signal in parallel while we plan the pivotal development of monotherapy in the later lines really gives us the totality of the clinical development planning of how to maximize the utility of this drug for these patients and clinicians.

Great. And we saw some data from ASCO GU around sort of HIF-2 alpha. Maybe first, your thoughts around ENPP3 as a target and the potential for synergy with other mechanisms that we're seeing sort of come through, if you will, in RCC.

Yes. I think the key for our hypothesis on how we could synergize is that really distinct mechanism of action, you're not going to be piling into the same pathway, right? The other piece is that distinct AE profile you have, in particular with TKIs, also to an extent with HIF-2 alpha, AEs that can be steady over time or build over time even. We have one where it's like -- it happens early and then it really, really seems to taper off. So I think there's a lot of opportunity there. I think the phenomenon we're seeing in RCC is a compression into earlier and earlier line with all 3 mechanisms, creating a significant unmet need third line, certainly, but even going into second line ultimately.

So I think there's a lot of opportunity to advance. Initially, we're going to do monotherapy focus there, and we haven't disclosed what our combination therapy is yet, but it's something on our mind. Do you want to add anything to that?

Yes. By the numbers, what we've seen is for a novel modality, which is most recently Belzutifan with an FDA label that allows use by the clinical community post-IO, post TKI, 1 TKI, PD-1 effectively. Given that compression into earlier line of doublet therapy and potentially even triplet therapy going forward, the clinical community is looking for something that is novel and can be used. And Belzutifan is on track to be a blockbuster $1 billion drug just with the LITESPARK-005 study and the label afforded by that. We would look to replicate that strategy in a large part with some minor changes in detail, but we fully expect 819 with that similar label in the U.S. and EMA to be able to be a blockbuster drug in the late line. And obviously, that opportunity and market that we could capture into an earlier line or second line post ipi/nivo would be a great opportunity as well because that's about 1/3 of patients in the frontline that progress on ipi/nivo.

So we're really excited about the totality here, but your question is right. Ultimately, you do want to explore combination therapy, right? And how you break it down is at the first part, PD-1 in the frontline is here to stay. It's being used in adjuvant setting, right? As we've seen early data sets from other T-cell engager studies with PD-1s, there is potential synergy, and there does not seem to be a penalty for additional CRS. So that's obviously of interest as we explore earlier line, which would obviously be more in the frontline scenarios for our clinical development plan.

HIF-2 alpha is now the kind of the combo of choice for a lot of investigators. Probably below that, the clinical community is maybe a little bit more tepid on TKIs. They're really trying to get away from that access as much as possible, but you never say never. That's kind of how we would rank order the combination opportunities.

And as you think about your initial pivotal study next year, are there key design questions that have yet to be answered? Or do you have a pretty good sense of sort of the strategy around sort of the initial pivotal? And then also maybe how the subcu versus IV could ultimately play a role, whether if not in pivotal, ultimately, when you think about sort of go forward, how that dynamic can evolve?

Yes. I think the design for the pivotal in monotherapy is really crystallizing pretty nicely. We had ad board and steering committee meetings at ASCO GU a couple of weeks ago. I think there was a pretty broad consensus on all the key parameters. I think we have good models from all these precedent studies about what kind of what the bar is. These precedent studies have had typically response rates, 20%, maybe in the high teens and then 5-ish month PFS. So I think that sets a good bar. We'll be disclosing all those kinds of details if all goes well. And as we expect late this year when we -- after we roll out our Phase I data RP3D. So that's all really crystallized. I think the script will be fairly clear for us to follow.

Going to your point on the subcu versus IV, subcu is a little behind. We've established IV dosing first. It could potentially be helpful in commercial. I think one thing to note is even with our IV, we are at Q2-week dosing and maintenance now, and we're going to be exploring Q3 week. So we'll let the efficacy data drive because we don't want to have a more convenient play that leaves efficacy on the table. As we characterize that, we should be able to wrap that up as we get into early '27 and understand exactly how subcu might layer in.

Great. And for the expansion cohorts, are those patients still sort of unselected for ENPP3?

Correct. Unselected. All clear cells.

And maybe if you can talk about sort of maybe enrollment dynamics, how that is going and how you...

Yes. Since we had our data out and we've continued and things have been moving forward, the excitement from the clinical community around having a novel agent with a novel MOA that has activity in these late-line subjects and in particular, that can be given long term and seems tolerable has really generated a lot of patient and investigator interest. And so we've been able to enroll pretty rapidly. We're happy with that. We think we can answer all the key questions with this impulse of patients, all the key questions around priming and final dose this year.

Right. And maybe on 6 CD3 program, you saw some initial early confirmed PRs in ovarian and germ cell tumors. Ultimately, what sort of drives, if you will, the decision around whether or not this is sort of a big opportunity or maybe a more niche opportunity in just smaller tumor types. What could sort of drive.

Yes. This year's data should do that. So we've gotten to target dose ranges late last year. We've coalesced around the right range right now, and we're enrolling across gyne tumors and germ cell tumors to see does the efficacy get us where we need to go. In gyne tumors, particularly ovarian, which is the largest opportunity, it's a competitive space. There's a lot of modalities, a lot of agents. I think we know the bars that we would want to hit in this early data to suggest continued pursuit there. I think in germ cell tumors, there's a really strong unmet need and a different commercial dynamic completely, as you point out, and potentially an opportunity to expand beyond that later line. So the data, response rate data and all those pieces will come together later this year because we have been enrolling. We'll have a knowledge base at the right doses in those indications to make the strategic decisions for each going forward into '27.

And then how much data could we see in the second half of 2026 to kind of help drive?

Yes, I think adequate numbers of patients at our target dose to really establish a recommended Phase III dose. So certainly, at the dose levels, sufficient to give that kind of -- these are expansion cohorts we'll be talking about.

And with this initial data set, I think, with the other sort of -- or the number of tumor types, I guess, in both oncology indications, I mean, certainly, the patient numbers could ultimately balloon if we see additional data in other cancers.

Right. Absolutely.

How are you thinking about potential pivotal and then also commercialization strategy around those, assuming success? And then now you're talking about maybe from thousands of patients to hundreds of thousands of patients.

Yes. So for us, it's about building the basic building blocks with efficacy data right now to define what is that next step that we can then follow on with greater numbers. Just like with our 819 program ENPP3, we've started expansion cohorts in papillary renal cell, in colorectal cancer, MSS and in adeno non-small cell lung cancer, all of those are biomarker selected, but substantial fractions. So move quickly, establish yourself monotherapy in a setting where you could move quickly, get that initial -- our ultimate strategy, get that initial commercial momentum going in a manageable way and then keep bringing things along and bootstrap up.

Great. And then in terms of strategy, you've not shied away from partnerships in the past. Do you see these assets as maybe go it alone assets commercially?

That's our base case. That's our base case, absolutely.

And maybe shift gears to immunology for the remaining time. Obviously, TL1A, long-acting, ulcerative colitis and others competitive areas. You plan to disclose some healthy volunteer data soon. Maybe just what investors should focus in on that data set? Is it half-life, PK, safety, are you thinking about?

Yes. So we disclosed, I think, the real punchline data from our Phase I in healthy volunteers for our long-acting TL1A XmAb942 last year. This is going to be the full Phase I data set with all the follow-up. And really, it's just confirmation. We reported a 71-day plus half-life. We reported really durable suppression of the TL1A levels, which is the key biomarker from genomic biomarker, key levels, single doses out to 16 weeks. No apparent impact of ADA on PK, on PD. Really just -- it's about confirmation of that as we enroll our Phase IIb global study, which is a study that's exploring 3 dose levels. It's well powered 220 completers against placebo to establish what is the Phase III dose so that you can really move forward rapidly and offer to ourselves or whether we partner the ability to get into pivotal trials definitively and quickly without having to do multiple doses and pivotals.

That's at DDW in a few weeks. Also, our bispecific IL-23p19 TL1A program will have its first public data disclosure there. Preclinical. We expect to be in the clinic in the first half -- in the second half in healthy volunteers. And that will be full preclinical workup in vitro in vivo data that can help people look at half-life and potency at really how we design this thing to address the really high challenges for bispecifics in I&I.

Right. And for 942, you think about the XENITH-UC study, when we look at sort of clinical benchmarks today, ultimately, what do you think can position this as a potentially best-in-class in terms of efficacy, durability and then sort of maybe frequency.

Yes, I'll mention the design and maybe Dane can touch on how the clinical strategy plays into that. So we built a molecule with very high potency, de novo TNA binder. We believe higher potency than others have reported. That's important. You want to cover the target, put our extremely well-validated Xtend technology, a couple of marketed products, many, many clinical programs use it to extend the half-life. Again, we got over 70-day half-life. And that hopefully lets us cover the target and get maximal exposure for the longest period of time. Then maybe, Dane, you want to mention how we design the study then?

Sure. Yes. And that was the guiding light for us in the early stages of the TL1A program with XmAb942 under design by the protein engineers was how can we get better exposure in the biologic class for IBD, where we think in many respects, these patients are being kind of underdosed on a target level. We had early evidence of that with vedolizumab, where there was a clear exposure response in the induction period from their studies in ulcerative colitis and Crohn's. We saw that with the IL-23 class that reinduction or extended induction with mirikizumab could capture north of another 50% of patients. And we went with that in mid-program, we finally got the data from the TL1A class, [indiscernible] UC and TUSCANY-2b. And the same thing, you got extended capture of clinical response with longer induction or there was an exposure response that was very clear at over 10% difference in clinical remission on the exposure tertiles in TUSCANY-2b.

So our Phase IIb study design will answer that question at the top dose level that we're dosing in the IIb study, we will have better target engagement based on our models than the pivotal dose regimens being used by our peers with the first-gen TL1A molecules. And I think there's a real there there. We'll see what happens, but it's a very exciting, well-structured, robust study to test that hypothesis while delivering clinical convenience of Q 12-week dosing in the maintenance period with a single subcutaneous injection.

In terms of the market opportunity, we think that kind of the intersection of better efficacy and better clinical convenience allows us to go maybe a couple of years into the branded period of the first [indiscernible], convert that market well into the runway of having all of this entire class remain on brand for many years after, convert the market on new scripts and then ultimately take the dominant market share. We've seen that work time and time again with other biologic classes, including the IL-23 class in diabetes, and we think there would be no difference for the TL1A class. So our design of the IIb is to be as efficient to move through clin reg development and get into commercial launch as fast as we can. And I think we have a good strategy. We'll see how it reads out when we have the primary analysis set in 2027.

Right. And for the 412 program, TL1A, IL-23p19, I guess, maybe conceptually, what's the problem that you're solving? You think about depth of induction for ulcerative colitis, for example, durability, the ability to maybe reach refractory patients, how do you think about sort of the value proposition?

I think it's driven by what's driving all the excitement about combination therapy in I&I in general and in IBD in particular, you get a cap on the percentage of patients that are going to go into true remission, right? People are going to have a deep response, IL-23, it's capped around whatever, 20-ish percent on a placebo-adjusted rate. You got TNF, that's a little lower vedolizumab's a little. How do we get that number higher? How do we reach more patients? The belief is that the different targets are going to reach different patients, hitting both at once at the least should give you some kind of additivity of the people that one and the other doesn't cover, maybe even synergy, we'll see. So it really leverages that growing appreciation that we need to hit more targets in order to reach more patients. That's a very strong driving factor for this bispecific as well.

And for those 2 target antigens, specifically, we have translational data from real clinical studies where in the nonresponding patients to TL1A, TL1A monotherapy you are seeing relative upregulation of IL-23. And that was a supportive basis that these could be orthogonal/synergistic targets to combine into a single bispecific molecule.

Great. And just quickly, you're going from 10, thousands of patients in oncology and potentially millions of patients in immunology. So how are you thinking about that from a strategic standpoint?

Yes. We are keeping our eyes open about when we get to key data inflections and value inflections for the program, what's the best path is bringing more resources on board makes sense. We obviously stay engaged with a lot of parties. We've historically done that. And we're going to make a decision based on the quality of the data and what we need in the next step.

Great. Thank you. So we're up on our time. Bassil, Dane, thank you so much for your participation.

Thank you so much.

Thank you.

Good day, everyone. My name is Megan, and I will be your conference operator today. At this time, I would like to welcome you to Xencor XmAb819 Initial Phase I Dose Escalation Results Webcast. [Operator Instructions].

At this time, I would like to turn the call over to Charles Liles, Senior Director of Corporate Communications and Investor Relations.

Thank you, and good afternoon. Earlier today, we issued a press release announcing positive initial results from our ongoing Phase I dose escalation study of XmAb819 in advanced clear cell renal cell carcinoma. It is available at www.xencor.com.

Providing comments on the call from Xencor are Bassil Dahiyat, President and Chief Executive Officer; John Desjarlais, Executive Vice President and Chief Scientific Officer; and Dane Leone, Executive Vice President and Chief Strategy Officer. We are also joined by Dr. Monty Pal, Professor in the Department of Medical Oncology and Therapeutics Research at City of Hope Comprehensive Cancer Center and Co-Director of its Kidney Cancer Program.

After the prepared remarks and presentation, we will open up the call to address your questions. Slides that we are using today should be visible here on the webcast and will be made available for download on the Events and Presentations page of our website around the time our remarks are concluded.

Before we begin, I would like to remind you that during the course of this conference call, Xencor management may make forward-looking statements, including statements regarding the plans and objectives of management, future product offerings and research and development programs as well as future financial and operating results, future market conditions, future operations, the company's partnering efforts and capital requirements. These forward-looking statements are not historical facts, but rather are based on current expectations and beliefs and are based on information currently available to us.

The outcome of the events described in these forward-looking statements are subject to known and unknown risks, uncertainties and other factors that could cause actual results to differ materially from the results anticipated by these forward-looking statements including, but not limited to, those factors contained in the Risk Factors section of our most recently filed annual report on Form 10-K and quarterly reports on Form 10-Q.

With that, I'll pass the call over to Bassil.

Thanks, Charles, and I'm delighted to welcome everybody. I'd like to start by putting today's XmAb819 program update into the context of our overall strategy and pipeline. Xencor's mission is to bring novel medicines that can advance the care of cancer -- advance the standard of care for patients living with cancer and serious autoimmune disease. Our novel medicines are powered by the XmAb protein engineering platform that's been clinically and commercially validated through numerous Xencor partnerships.

Last September, we realigned our pipeline and focused on our lead first-in-class T-cell engagers, XmAbs 819 and 541. And we also used our bispecific expertise to launch multiple autoimmune disease programs. 2025 has been a very busy year for Xencor. While we're rapidly advancing our wholly owned oncology portfolio, we've also built momentum with our autoimmune programs, with the start of a global Phase II study of XmAb942 along with a Phase I study of plamotamab for the treatment of rheumatoid arthritis.

Today, we're going to focus our discussion on 819 and 541, where we're positioned to have recommended Phase III doses established in 2026. And to build on this momentum, we expect our next first-in-class T-cell engager in the clinic in 2027.

Now John Desjarlais, our Chief Scientific Officer, is going to walk us through our design rationale for XmAb819 and the setup for the Phase I study. Go ahead, John.

Thanks, Bassil. XmAb819 is a first-in-class XmAb 2+1 T-cell engager targeted to ENPP3. Importantly, there are 2 binding domains against the target antigen, a design that we found could promote greater selectivity for tumor cells versus normal cells and an Fc domain to promote antibody-like half-life.

The ENPP3 target, as you can see here from tumor versus normal RNA levels, is selectively overexpressed in kidney cancer, where we focused our initial efforts. You can also see from these RNA patterns that there are multiple expansion opportunities in a number of other histologies, including lung and colorectal cancer.

As shown on the right plot here, XmAb819 demonstrates strong T cell-mediated toxicity against tumor like cell lines with high ENPP3 expression. Now we found that the XmAb 2+1 format, which is now used in 4 clinical stage T-cell engagers gives us a platform where we can tune each molecule for selective engagement in killing of tumor cells versus normal cells.

As you can see from this in vitro activity plot, each drug could be tuned to promote high cytotoxicity against tumor cells with low or minimal activity against low-density cells that look more like normal tissue. This is a well-known concept in the antibody world, whereby bivalent engagement of your target, coupled with affinity tuning, could promote preferential engagement of cells with higher target density.

So this slide details our Phase I dose escalation scheme, very typical for a T-cell engager Phase I, beginning at low doses and ramping up as the standard for any immune agonist. The goal is to establish an effective prime step target dosing strategy that will ultimately support a reduced inpatient monitoring profile while maximizing efficacy.

We initially prioritized intravenous weight-based dosing but with strong recruitment to the study, we've been able to accelerate the subcutaneous [ flat dosing ] cohorts and have recently started catching up to the IV drug exposure levels. Here, we've also bucketed the dosing cohorts into low-dose cohorts and then higher doses that achieve our target dose range as expected from our preclinical work.

The actual dosing regimen, shown here on the bottom consists of a priming step, 1 or 2 step-up doses, followed by weekly target dose for the first cycle. Then as supported by our pharmacokinetic profile, we proceed with biweekly dosing for Cycle 2 onward. Our data cut on September 19 reflects preliminary results of 10 completed IV cohorts and 5 subcutaneous cohorts, which has now triggered the first IV dose expansion cohort. Importantly, we continue to dose escalate both IV and subcu.

So as I mentioned earlier, our preclinical work, coupled with some mechanistic modeling had predicted where we might expect to achieve active concentrations of XmAb819. You can see here for our latest IV cohorts 8 through 10, our [indiscernible] predictions for these cohorts show we've been reaching those active dose ranges, first reaching those levels on a second step-up dose and then achieving consistent exposure in those ranges as we proceed through the first cycle. Our IV half-life is 8 to 9 days, which supports biweekly dosing in later cycles. We've also started characterizing our subcutaneous PK profile, seeing a Tmax of 5.5 days and promising bioavailability in the 55% to 70% range.

Thanks, John. Now we've only enrolled clear cell renal cell cancer patients in our study, and they were very heavily pretreated. This is the first RCC trial that we're aware of with a median of 4 prior systemic therapies with over 1/3 of patients with 5 or more. And also the first trial with a large proportion, 36% of patients with prior HIF-2 alpha inhibitor therapy.

Our goal is to show that our first-in-class T-cell engager can create exciting antitumor activity in a population like this, who received the most current standard of care and we were rechallenged with earlier line therapy, whether it's a checkpoint inhibitor or TKI is ineffective. Also, we've not preselected for ENPP3 expression, but we are testing patient samples retrospectively throughout the study.

We expect nearly all the clear cell renal cell cancer patients to have high ENPP3, but do expect some outliers with lower absent expression based on our earlier work. Now of the 69 patients we've treated, 25% remain on study as of the data cut and the drug has been well tolerated, as you can see, with a low rate of only 4% of patients discontinued for adverse events, specifically 3 individuals, 2 of which were liver enzyme elevations and 1 case of nonfatal myocardial infarct in the presence of CRS and hypotension.

Now AEs overall, were manageable, and we're glad that over 90% of our patients reached their target dose, a key goal in a Phase I study for CD3 T-cell engagers. The low rate of dose reductions and discontinuations supports our view that 819 is well tolerated in patients with very advanced disease. Now we are happy with the primarily low grade and rate of CRS observed with 819. We think this will be very important as we advance further into development and explore reduced inpatient monitoring.

As you can see on this slide, the study was impacted by dose prep dilution errors for the initial priming dose, which we discovered in the early part of 2025. These dose errors correlated to elevated drug levels in serum and a significantly elevated rate and severity of CRS, 28% Grade 3 for these patients across the study. And while the great majority of patients were correctly dosed, the errors had a meaningful impact on CRS rates.

Now when we look at the target dose range for the 18 patients with the intended priming dose, Grade 3 CRS has only occurred in 6% of them or one patient specifically. This is consistent with the overall study Grade 3 CRS rate of 4% for those 51 patients that received the intended priming dose of our study drug and is consistent with the Grade 3 CRS rates for marketed and late-stage CD3 bispecifics.

Since discovering it, we've mitigated the problem by training sites to use only the correct materials and procedures for dose prep and we'll eliminate the root cause entirely in the first half of 2026 when we roll out a low concentration drug product. Thanks to our team's hard work of several months determining root cause, we put this problem behind us.

Now for the full AE picture, we have our overall tornado plot on the left and on the right, the post day 29 AEs plotted, which is after all of the priming and step-up doses. These plots include all patients, whether the correct priming dose prep or not. Now treatment has been well tolerated with related adverse events, mostly consisting of CRS, rash and GI symptoms with low Grade 3 rates. We've had no reports of ICANS of any grade.

Now importantly, related adverse events occur mostly within the first few weeks during the initial priming period and repeat dosing at the target dose level, which, of course, is the highest doses, incurs a few notable side effects as shown on the plot on the right. We're hopeful this profile can be advantageous as we explore combination therapies in the future.

Now the maculopapular rash we've seen was transient and it responds quickly to supportive care. It happened shortly after initial exposure to 819, and we believe is due to the activity of T cells against basophils and mast cells, which express ENPP3.

Now here's our efficacy data. Low dose cohorts on the left, target dose range cohorts on the right. We're excited by the antitumor activity of 819 we're seeing in recent cohorts, which are predicted target exposure range, and we see a clear dose response relative to these lower dose cohorts.

There are marked and sustained reductions in target lesions in heavily pretreated patients, including those with prior progression on Belzutifan, like the rightmost patient in each of the 2 plots. Also, regarding ENPP3 expression, we looked at expression data retrospectively and a very high number were high expressers with just a few outliers. Also very interesting is the patient with a double dagger in the target dose waterfall plot, the fourth from the right.

They had progressive disease at their first scan due to a new lesion emerging even though their target lesion shrank, but they continued treatment and experienced a marked reduction in target lesions of 47%. Per RECIST, they are characterized as progressive disease, but the patient remains on therapy at day 185 with this tumor reduction and full resolution of that non-target lesion.

Now on an evaluable basis, our best overall response rate is 25% with a disease control rate of 70%, which is unprecedented for monotherapy treatment of clear cell renal cell carcinoma patients that have progressed on a median of 4 prior lines of systemic therapy, including having patients that have progressed on prior HIF-2 alpha inhibitor therapy.

Now this exceeds our internal target for triggering our first dose expansion cohort, which was 20% or greater response rate. It's still only early follow-up for our responders. But note, all responders remain on treatment as of the data cut.

Now this is a spider plot for all evaluable patients in the target dose range. You can see that there's consistent target lesion reduction, the majority of the lines going down. Now it's still early for our target dose range cohorts. We are also encouraged that patients can experience deepening responses with time.

This swim plot is another way of looking at time on treatment, and it underscores that patients with an initial response to therapy tend to stay on treatment and have durable reductions in target lesions.

Now we'll be continuing follow-up with these patients, of course, and we're optimistic that continued dose escalation can support further dose response and durability of disease control. And ultimately, of course, we hope that this translates into an improvement in progression-free survival over the current standard of care for these patients with very advanced disease.

Now I'd like to introduce one of our investigators and close collaborators, Dr. Monty Pal from the City of Hope Cancer Center, where he's Co-Director of the Kidney Cancer Program and oversees one of the largest clinical trial portfolios on the West Coast. We're very happy to have Monty's insights and experience in both early and late-phase trials to guide us as we develop 819. Monty?

Thanks so much. It's really been an honor to represent the study here at Target 25. And I enrolled the first patient on this trial several years ago. It's been a wonderful journey since then.

I wanted to highlight a case of a 56-year-old gentleman, very near and dear to my heart. I've been following him for over a decade now. He developed metastatic disease about 5 years ago. And at that point in time, he was started on therapy with cabozantinib and nivolumab. Very quickly after starting, he developed new onset of hepatic metastases. So we transitioned his therapy on to lenvatinib and everolimus. He was on that treatment for a healthy amount of time for about 3 years, but he really sort of suffered from a multitude of toxicities, hand-foot syndrome, diarrhea, fatigue, et cetera.

We enrolled him on the protocol. As you can see in the panel on the left-hand side, he had membrane staining for ENPP3 in less than 25% of tumor, fairly diffuse cytoplasmic staining of ENPP3. With that in mind, he had a really profound response. As you can see in the panels on the right, we've highlighted one of his mediastinal nodes decreasing in size from 28 centimeters and in maximal dimension to 11 millimeters.

So with that in mind, with that sizable reduction, it amounts to about a 63% decrease in this target lesion. I would suggest that this is one representative example. And the RCC investigator community in broad terms has really been looking long and hard for novel targets. To date, we've had VEGF axis inhibition. We've had checkpoint inhibition, but really nothing else beyond that. There's a lot of redundancy in drug development for the most part.

The strategy of ENPP3 targeting is very, very novel. And based on our data, I think it's worthy of much more extensive exploration. The toxicities that we're seeing here, for instance, CRS, were actually common to other bispecifics. And frankly speaking, I think myself, my colleagues are becoming increasingly familiar in how to manage these toxicities.

I personally have been very, very pleased at how many patients have been able to maintain exceptional quality of life on this therapy, particularly after priming. And I think some of the data that you saw on the previous slides alludes to that. Some of these deep and protracted responses that we're seeing in patients who have been exposed to multiple prior therapies really makes me eager to move XmAb819 forward as both as monotherapy and potentially in combinations as well. Thank you.

Thank you, Monty, and thank you to all the investigators in this study and all the patients that have been willing to enter into this clinical study, and we're very excited to be providing this opportunity to do something truly novel and different. And with that, we want to move into how we think XmAb819 can be really truly differentiated and potentially effective treatment option for those patients, all the patients living with advanced clear cell renal cell carcinoma.

We hope to trigger our second IV dose expansion cohort around year-end and are rapidly advancing our subcutaneous dose escalation in our target dose range. As we work toward establishing the target dose regimen to use for our pivotal studies, we're also evaluating broadening the renal cell program in combination testing along with additional ENPP3 expressing tumors. XmAb819 has a significant opportunity to become a blockbuster drug in clear cell renal cell carcinoma.

Global drug sales for RCC are projected to reach $12 billion during 2030. Despite the size of the market, drug revenues are dominated by only 2 mechanisms of action currently, PD-1 inhibitors and VEGF-TKIs. We expect the adoption of Belzutifan, a HIF-2 alpha inhibitor, to continue to increase, but yet there remains a real need for additional treatment options for patients and clinicians.

Patients with advanced clear cell renal cell carcinoma are increasingly treated with the doublet therapy of checkpoint inhibitor plus VEGF-TKI in the first-line setting. Unfortunately, 50% or more of those patients progress and need additional therapy, which as of today is most likely a rechallenge with VEGF TKI or increasingly Belzutifan.

Once patients have progressed on to the third line of treatment, they are effectively out of therapeutic options beyond potentially another rechallenge with a VEGF-TKI, which has limited therapeutic benefit. We view the recent drug labels for approved therapies such as Belzutifan as promising for only requiring exposure to a prior checkpoint inhibitor in prior VEGF-TKI, indicating a similar pivotal design as LITESPARK-005 could provide clinicians with the ability to use XmAb819 in the second-line setting.

We expect the use of Belzutifan will move into earlier lines of therapy as a doublet regimen with lenvatinib upon completion of LITESPARK-011 and potentially become part of a triplet regimen in the frontline with LITESPARK-012. We view XmAb819 monotherapy as a clear contender to be a drug of choice post progression on a checkpoint inhibitor and VEGF-TKI, helping to redefine standard of care for second and third line treatment of clear cell renal cell carcinoma.

Now supporting our excitement around continued development of XmAb819 is the competitive response rate observed to date in our target dose range cohort of 25%. This stands impactful relative to the most recently approved monotherapy drugs for clear cell renal cell carcinoma, which are Belzutifan and tivozanib. XmAb819 is not only the first T-cell engager to demonstrate clinical response for patients that progress on prior therapy with a HIF-2 alpha inhibitor.

But our study also overall enrolled a more heavily pretreated patient population that was enrolled for either the Belzutifan or tivozanib studies. With a median of 4 prior systemic therapies, over 1/3 of patients enrolled in our dose escalation study have been exposed to 5 or more prior regimens with over 60% of those patients having progressed on 2 or more VEGF-TKIs.

Now based on the strength of the data in clear cell renal cell carcinoma, we're excited to be exploring additional development opportunities in other tumor types that exhibit ENPP3 positivity and remains significant unmet need for novel therapies.

As highlighted by Dr. Pal, we have seen significant antitumor activity with XmAb819, even in patients that exhibit lower levels of ENPP3 positivity in their tumor biopsy. We have not decided on a positivity rate yet that we can use for expansion into papillary renal cell colorectal or lung cancer, but our addressable patient opportunity for the XmAb819 program is significant.

With that, I will actually turn it back to Bassil for an exciting early look at XmAb541.

Thanks, Dane. So I'm really happy to present an early look at the progress we've made with XmAb541, which is our first-in-class Claudin-6 targeting CD3 built on the very same 2+1 platform as 819 that John described.

Now we've worked really hard to make the Claudin-6 binder are highly selective for avoiding binding to the closely related Claudins 9, 3 and 4 that are expressed on healthy tissues. And the team has made really rapid progress in less than 18 months, enrolling 8 cohorts, both IV and subcu from our start in April of 2024 and meeting and exceeding our corporate goal to begin characterizing target dose range cohorts by the end of this year. Thanks very much to them.

And our investigators are getting very excited about our first-in-class 2+1 T-cell engager as we've recently achieved active dose levels. At our most recent IV dose level, we have 3 responders out of 9, 2 patients with germ cell tumors and 1 patient with ovarian cancer with tumor reductions for 7 of 9, which is potentially the beginning of a response profile similar to the reports from Claudin-6 ADCs.

Now we still have plenty of dose escalation and dose optimization work ahead of us. But given the progress to date, we're targeting a recommended Phase III dose for XmAb541 during 2026.

As Bassil said, while we still have dose escalation optimization ongoing for XmAb541, we are optimistic in reaching the recommended Phase III dose in 2026. With the emerging antitumor activity observed in this basket cohort of germ cell gynecologic tumors, we are hopeful that XmAb541 could ultimately offer a treatment paradigm free of toxicities associated with chemotherapy and antibody drug conjugates, such as febrile neutropenia and neuropathy.

Each of the current indications represent significant development opportunity for this drug. And as we progress through dose escalation, we'll further define the strategy for our next stage of development.

With that, I'll turn it back over to Bassil for concluding remarks.

Thanks. Now I'll wrap up today by looking at our oncology portfolio as a whole. We have 4 2+1 CD3 bispecifics advancing in solid tumors, each of them showing compelling clinical data. No other platform has that consistency or breadth. XmAb819 and 541 are internal, and our partner, Amgen's Xaluritamig is in Phase III in prostate cancer. And just last week, Astellas' ASP2138 showed very promising data in gastric and gastroesophageal junction cancers at ESMO, and they've said how excited they are for potential next development steps.

Now behind those are Janssen with CD28 T-cell engager in prostate cancer and an undisclosed T-cell engager we hope to have in the clinic in 2027. This validation and more importantly, the opportunity it represents has us very excited for next year and beyond.

We're ready to take questions.

[Operator Instructions]. Our first question will come from Steve Seedhouse.

Great data. I have a few questions, if that's okay. Bassil first, you mentioned you were testing for ENPP3 in the study and actually the case report did emphasize that staining in that particular patient. But across all patients, can you describe whether ENPP3 expression in those tumors correlates with the clinical outcomes that you're seeing?

I mean the vast majority of patients were positive and quite higher for ENPP3. So I don't know if I want to use the word correlation, but we do have a couple of anecdotes. So if you go back to our -- Charles, to our waterfall plot, the left most patients that is the ones with the highest tumor progression and growth in tumor change on the target dose cohort plot, those 2 left most patients were both the 2 that had ENPP3 null by IHC, so using our assay.

So anecdotal because, again, the vast majority were high, but I think that's a really nice validation of the mechanism of the drug, and I think hopefully reads on the ability to set meaningful cutoffs as we go into other tumor types like in colorectal, like in lung that don't have uniformly high or nearly uniformly high expression like clear cell renal cell.

Thank you. The other thing, too, as you have this slide up and just as we're looking at it, despite how advanced this population is that you emphasized in the slides, relative to legacy studies, your deepest PRs you're seeing here are in like the triple refractory post HIF-2 alpha patients.

And I'm wondering if you'll have an opportunity in the dose expansion cohorts or in any of the additional dose escalation cohorts to maybe enroll less refractory patients and see if the efficacy is even better than the already impressive 25% objective response rate.

Thanks, Steve. Great question. I think we take patients as they come. And I think we're very happy with the work of our investigators, Dr. Pal and his colleagues that we have a wait list for this study. And we do not want to turn patients away that are in critical need of care after they progressed on standard of care.

So as it turns out now, we're really not planning for any type of change in the study as we go into dose expansion. And to your point, are quite confident in the clinical profile that we have and the ability of that clinical profile to stand up as we contemplate pivotal studies and starting a pivotal study because these patients really have nothing that can afford them real clinical benefit, which is why there's so much excitement. Dr. Pal, would you like to make a comment?

If you don't mind, I think the data that we're seeing here really gives me the confidence as an investigator to actually enroll my patients into this experience much earlier. So I am keen on bringing patients on in earlier lines of therapy.

Fantastic. Last question from me. I just wanted to ask, the -- so there's 24 patients in the target dose range that were dosed, 4 are excluded from the efficacy analysis, as mentioned in the footnote. But of those 24, 6 received the incorrect dose prep and 18 the correct dose, priming dose prep. How did the 5 RECIST responses distribute between those 2 subgroups?

Yes. So good question. I kind of figured we would get that. There is a higher response rate in those that had the correct dose prep. And I think there's a good actually anecdote here that, that Bassil went over previously with the one patient that did have PDF first scan did receive into correct dose prep. Bassil, maybe you want to take them through that case?

Yes. So yes, majority of responses were for the correct dose prep and the rate was better for correct dose prep. And I think that patient, the fourth from the right on this waterfall with a double dagger, so they had initial progressive disease on their first scan and then they had -- because of a new lesion and then they had both that lesion resolved and the other lesions have a 50% reduction for a long time.

That patient had the incorrect initial dose. They also had, quite consistent with that, a number of AEs like CRS during the priming and step-up that delayed their target -- getting to that top target dose, going up through the priming until day 50. Their first scale was on day 48, and that's when they got that progressive disease.

So we think with resolving this issue, we're hopeful that improves patients' ability to stay on and hopefully has a benefit for them as a result of that.

And that's been the overall goal of the treatment algorithm evolution and getting into the target dose range is where by our preclinical prediction, we thought we would have enough drug exposure to be able to prevent progressive disease in this advanced patient population while getting to the target dose, which can then really elicit a profound clinical response and disease control at the end of the day.

Next question will come from Jonathan Chang with Leerink.

First question, can you provide any color on any clinical differences observed between the high-dose cohorts so far, if any? Or are the numbers too small to really highlight anything at this point?

Yes. I think the numbers are too small to highlight anything. We're -- in order to sort of build the numbers up, we clustered them together in the high dose range. And it's a relatively small span of drug exposures. So I think we feel confident that that's the right way to do it. It's just too small and frankly, too early in follow-up for the last couple of cohorts to say much about individual cohorts.

I think the important point here is we've seen RECIST response in all the target dose range cohorts, which gives us confidence that it matures and marries with our preclinical predicted exposure levels that we need to affect that clinical response. But when we talk about triggering the first dose expansion cohort, which is Cohort 10 IV. We looked internally debated it, and we do think there is a dose response on disease control and durability with that cohort versus, say, Cohort 10, which was the first cohort in the target dose range.

And I don't know, Charles, so you want to go back to the PK slide, but that also matches that concentrated and steady-state AUC coverage that we see with Cohort 10 versus still dropping a little bit below on C trough in Cohort 8. So we're very happy with that. We feel confident that Cohort 10 will be good to have as the first dose expansion cohort in the study. And then again, we hope to trigger the second dose expansion cohort before year-end.

Yes. Understood. And second question, what do you see as the CRS benchmarks for 819? And can you provide any additional color on what happened in the pharmacy error side? And what's been done to ensure this won't happen moving forward?

I'll start with the benchmarks and then Bassil, you want to go into the mitigation on the dose preps. So the benchmarks, we're very happy with. Listen, when you see a 4% overall Grade 3 rate with a T-cell engager, that's eliciting this type of antitumor activity in this advance of a patient population, that's something. And I think that's still only half the story because, again, there's a 0% rate of ICANS. We've seen no ICANs in the study with this mechanism or this drug.

So that compares to, say, IMDELLTRA, which is, I guess, the gold standard T-cell engager in solid tumor oncology right now, that has a single-digit rate of Grade 3 CRS or higher, but also carries a 10% Grade 3 or higher rate of ICANs. And as we know, neurotoxicity is something that is equally scary to the clinical community and hard to deal with patients.

And despite that, tarlatamab is afforded per FDA label to be used in a reduced inpatient monitoring setting, which gives us the confidence in the consistency that we've seen with Grade 3 CRS with correct dose prep even in our target dose range to start exploring that as we continue to advance this program. Bassil?

Yes. So the issue with the dose prep is really one of, I would say, the ports and syringes used to transfer the solution. So you start out with our drug product in a vial at a relatively high concentration. For the priming dose, you have a very low concentration you need to infuse because it's a very small amount of drug for the priming dose.

As you go to higher doses and the step-ups, things are much simpler. But for that when you have to do a couple of dilution steps, so you're moving solutions from this high drug product vial into a bag, an IV bag that you're using to dilute it, and you're taking some from that bag, putting it in another bag to dilute it further for that very low priming dose.

If you're using ports for transfer between those bags and syringes that have unacceptably large dead volumes and not following the right procedures for using the right ports and syringes, you're going to have a little bit of extra of that high dose coming out of the vial sticking around and then it's going to contaminate that last bag, that low dilution bag, low concentration bag with high concentration products.

So that's just something that, by the way, happens with these ports and syringes. And so that didn't become clear until this year as we got to higher priming doses. Earlier events were when our PK assay was a lot -- we were at the bottom end of the range. Now that we've gotten to higher priming doses. The PK issue became apparent we're able to connect the dots with Grade 3 CRS. And then we mitigated by training in the short term to make sure we don't have more of these or try to minimize them that the pharmacy technicians do the right thing.

On the longer-term side, so next half, so it's the first half of 2026, we're introducing a low concentration drug vial formulation for that priming dose to eliminate the possibility altogether, eliminate the root cause of the multistep dilution.

So that's long story short. We put it behind us, and we're really happy when we -- when things are done in the intended dosing regimen like it was for the vast majority of patients, we really get a nice clinical profile.

Understood. Maybe just one last question, if I may. And I'm just trying to map what you just said with the course of the study. Can you provide any color on how enrollment has progressed over time on the study? And do you think it will accelerate now that you're at higher dose levels and as you move into expansion cohorts?

So I will say we've seen it accelerate, I would say, greatly as we've reached these higher dose levels starting around mid of the year. And I know maybe Monty can comment on how he's seeing that as one of our lead investigators.

I have to tell you, seeing these response rates in the salvage population of patients with incredible amounts of previous therapy, I can tell you all of my colleagues are going to be really chomping at the bit to participate in this trial. We already have a cadre of the top centers around the country participating. So I'm very, very excited to see what's going to happen in enrollment trends. And we've already generated a pretty substantial wait list at this point.

Yes. We've not been limited by enrollment at all, and it's really accelerated in the last couple of quarters. So pedal to the metal.

Your next question will come from Sean McCutcheon with Raymond James.

First, maybe can you speak to the dynamics of T cell exhaustion that come into play in this very late line patient population? I think we've seen some preliminary data that less frequent dosing may be beneficial in certain tumor histologies for T-cell engagers.

So how do you think about the T cell fitness in the context of late-line clear cell renal cell carcinoma and the immunophenotype of the tumor microenvironment and how that shifts -- could shift as you move into earlier lines potentially? And then secondarily, could you speak to the IV versus subcu administration on PK profile, tolerability and how you're thinking about a go-forward strategy to that end?

Sure, sure. So on the T cell exhaustion front, I mean, clearly, the activity we're seeing suggests that there's enough patients with active enough T cells to drive the kind of really exciting efficacy in a very late line that we're seeing.

A general truism that we've observed and others have observed is the less heavily pretreated, the generally the healthier and earlier in their disease and in their cytotoxic regimens of therapy that they receive, the more immune competent patients are, right? And so we are hopeful that, that helps as we go into earlier lines.

On the issue of exhaustion, I mean that's something that we definitely have on our mind as we look at patients going out into past, what is it, cycle 2, getting every other week dosing. I think some of the programs I'm aware of you're commenting on, as you extend out the dosing interval, the T cells seem to have a little bit of time to recover.

There's a lot of unknowns here, but I think in general, we feel that earlier and giving T cells in chronic treatment, time to recover is going to benefit. But I think for starters, we're very happy with what we're seeing in this group. So that's a good starting point.

I think it's probably too early for us to break out IV and subcu. The patient numbers, in particular on subcu are pretty low. But this is going to be essentially what we're looking at and comparing over the next year to choose our recommended Phase III dose and regimen. Hopefully, as we progress, we'll have a subcu expansion cohort as one of our expansion cohorts to compare to the IVs and see how that -- how they look when we can enroll a bigger number of patients.

And again, with the demand for the study and the enrollment rates that we're seeing now, we're optimistic we can test multiple of these regimens in short order in 2026 and come up with the best one. Again, a little early to say and compare, but I will say that subcu, we have responses in subcu. And so that makes us definitely want to study it more.

Our next question will come from Matt Phipps with William Blair.

I appreciate all the data for both programs here. I was wondering if maybe you could just comment on what hospitalization requirements you're going to need as you take Cohort 10 into expansion. Is there already an ability to maybe reduce that as you figure out the correct priming dose? And then also, Dr. Pal, maybe you can just give us your view on the rash seen here and how to manage that?

Sure. I'll start and then hand it over to Dr. Pal. In terms of exploring reduced inpatient monitoring, we'll probably follow the same path that our peers have done of prosecuting expansion cohorts to get through maybe a recommended Phase III dose.

And then as you're moving the totality of the program forward, start exploring in parallel a reduced inpatient monitoring cohort. I think we would want to see the data from dose expansion cohorts at that target dose range to then go back and make an argument with the clinical community and potentially regulators that this is a profile that should be explored in reduced inpatient modern setting.

I'll add one more point before you get to that. And note that we do see that attenuation of AEs, CRS, but as well as others as we go from the first priming dose to the next and the next. And then as we get to target dose, you saw really a very small and manageable set of AEs.

So we're optimistic that we can trim that back, but it just takes data, right? It takes data, regulators don't take anything on faith. We're hopeful that we can get there as we build up the numbers like Dane said. Maybe on the rash, Monty?

Indeed, yes, absolutely. With respect to the rash, my experience and that of other investigators has really been sort of that it sort of peaks and then subsides subsequent to the priming period. And this being sort of a mast cell basophil-related phenomenon, I don't think that these symptoms are necessarily durable. They're quite easily managed, and we've built some mitigation strategies with antihistamines to help address that.

And if I can ask one more? Awesome to already see some responses in the 541 program. Any initial insights on the safety profile there? Is there anything on target off tissue to worry about or still just maybe some CRS step-ups to work through?

Yes. I think -- I mean, we're -- it's early days. We're happy. Essentially, the data shows that we hit the corporate goal of starting to characterize the target dose ranges, which we had hoped to do by year-end. And I think the waterfall plot probably exceeds our expectations of where we've been able to get to and basically starting the study in April 2024.

So this has been really proof of concept for the Xencor team that we can take the learnings of all of our programs we've worked through clinically with T-cell engagers and really accelerate this 541 program. So we're very excited at the clinical activity we've gotten to. We have a lot more optimization characterization work to get to now, but it's still premature to really comment on what the profile of the drug is.

Our next question will come from Tara Bancroft with TD Cowen.

So the first question is just a clarity question. So just to be extra clear on your response to, I think it was Steve's third question. Can you confirm what was the response rate among those 18 patients with the priming issue? And were those 5 responders in that group?

It is the PK aberration.

Yes, the PK aberration...

Yes. We've had a higher response rate in the patients with correct dose prep versus those with PK aberrations. As we walk through the one patient that had PDF first scan because they were not able to get to target dose on time due to dose prep errors ultimately did respond when they got to the target dose. So I'm not sure we're going to give a full qualification of what that is and [indiscernible]. But yes, the vast majority of the responders have been in the correct dose prep patient population.

Okay. And then on safety, can you separate out for us a little more the impact of that as well? Like I know you said the 4% Grade 3 CRS with the correct priming. But were there any other safety events that you found were disproportionately impacted as well?

I don't know that we could say specifically. It was just everything seemed a little bit -- patients just seemed a little bit harder to manage, right? Nothing characteristic necessarily that I want to point to or even disclaim necessarily that we had in the correctly dosed patients.

But the CRS was just right in your face, 3 of 6 at the target dose range that were incorrectly dose prepped had grade 3, 1 of 18 had it that were correctly. So nothing else, I guess, was so stark and clear that we could characterize it. And the good thing is, of course, even with the dose prep errors averaged in, still our CRS was in the first few -- early on in the priming and it resolved and overall rates were quite manageable.

Our next question will come from Eva Fortea-Verdejo with Wells Fargo.

Congrats on the progress. Two from us. First, you mentioned combination potential for 819. We're wondering what mechanisms would make the most sense here? How are you thinking about the potential combination strategy moving forward? And the second question, is how should we be thinking about data disclosure in the next 12 months for 819 and 541?

I'll take that second one, and then we can -- maybe Dane can jump on the combo question. For the data, timing, we'll obviously guide as we move forward, and we're ready to have real clarity on when we're going to say what's next. But the next kind of data we're going to want to talk about is our expansion cohorts and how we -- what our determination is of recommended Phase III doses and the characteristics of those recommended Phase III doses for those 2 programs that justify us wanting to move forward, then we'll be able to have a really fulsome discussion of what the next step strategies are for mono and combos and why.

We can talk generally now about that, but I think data is going to rule the day. And again, next year with recommended Phase III doses, we hope to determine, we should be able to have that be the next tranche of information.

Yes, I think we were very purposeful in putting a recommended Phase III dose being established for XmAb819 in 2026. And obviously, when any company gets to that point in development, they would share that data. So I think 2026 is a good placeholder for the next data update, and we'll get more granular as we see the data flow and see where we get and make certain data analysis decisions like looking at IV versus subcu presentation.

In terms of combination therapy, I think the most obvious statement we could make is when you look at the evolving landscape of treatment within clear cell renal cell cancer, you have VEGF-TKIs, PD-1s, still the backbone of first- and second-line therapy, and there would be a great deal of interest of exploring a drug like XmAb819, which really has a limit and low rate of AEs that are seen post day 29 in our study, meaning that there's good combination potential here. And so that's something that will be discussed, and we'll make some decisions on in the coming year. We're very excited about that.

Dr. Pal, do you want to make a comment on that?

Yes. If I may just add to that. I think that in the face of an agent with good activity as monotherapy with nonoverlapping toxicities largely with the classes that you've illustrated VEGF-directed therapies and checkpoint inhibition, I think there is a real pension from investigators to really explore all of these combinations as we move forward.

Our next question will come from Alec Stranahan with Bank of America.

Great to see the update. First, I appreciate you're taking cohorts 8 through 10 as the target dose range based on your preclinical work. Just curious how many of the 38 patients from the low-dose cohorts were in Cohort 7 and which cohort the single PR came from in that low-dose group?

Thanks for the question. I mean I'm not sure I totally understand the gist of it. But yes, I mean, if you think about the low-dose cohorts, again, this is a 3 plus 3 design. We obviously filled in some of those cohorts with additional patients as we probably got to the mid-dose range. So I'm not sure there's a lot to read in there. But yes, as you mentioned, we did actually see a nice response even in the lower dose cohorts.

Yes. I'll say that responder in the low-dose cohorts, I would say, was at a -- not near the top end of the range. And so that happens in these studies with immune therapy. Sometimes there's a patient that's a little off the curve.

But we were, I think, excited that you even see as you get to the right-hand side of the low-dose cohort, there's a number of stable diseases, and that gave us confidence that we really see a dose response now that we're in the target dose range, and we're seeing the deepening of those antitumor responses.

Okay. That makes sense. And I guess in terms of the preclinical work or what you're seeing in terms of PK AUC, is it -- like how steep is the curve through the dose escalation? And I guess, how does that sort of relate to the cutoff at Cohort 8? Just trying to understand the difference between that threshold.

Yes. So look, you have to pick a line for where you -- when you're in the range and when you're not. If you look at that curve, Cohort 8 is just starting to climb up into that range. Cohort 7 is a little below that. There's small multiples that separate cohorts. I would say, though, that if you look at the theoretical basis and the preclinical data usually shows very steep dose responses for T-cell engagers to go from really not engaging to engaging. Then once you engage, it's not all 100% at once.

But that turn on of engagement is because you have to form trimers between a T cell, a tumor cell and the third component being your bispecific. And the more components you have in a system that come together, the more cooperative that is and the steeper the curves. I don't want to make too much quantitatively on it.

I don't want to make it sound like these preclinical models and quantitative systems pharmacology tools that we have are perfectly quantitative. But the concordance was pretty gratifying, I would say, that things really started to move, and you see that, in particular, disease control rate because you have so many bigger numbers, right, compared to OR. That DCR between the low and the high dose, it's really such a stark difference. That gives us confidence that our models were telling us the right sort of semiquantitative range for the immune activation.

I think what we could comment on qualitatively to perhaps help you is there's a concordance on the front end of the dosing regimen for the high dose range or the target dose range. And that's really to Bassil's point, these patients are rapidly progressing.

Our job as a drug developer is to get them as much drug in that first month as possible, balanced against, obviously, toxicity and CRS. That allows you to then step up to the target dose, the effective target dose that really keeps your AUC up there and drives that T cell engagement against the tumor cells. And so that's where you see the consistency of the target dose range is not only at the target dose, but also on the front end of the priming regimen that allows us to deliver or maximize that amount of drug delivered in the first few weeks.

Yes, yes. It's not just week 3 that matters, week 2 can make a difference. And if you -- that's higher than it was in the prior cohort, you're right, there's sort of accumulating effect.

Okay. Okay. That's very clear. And maybe one super quick one. Just on the Claudin-6 study. Curious if a similar priming dose issue was observed here at all? Or if this was maybe just isolated to the 819 study?

Nothing like it whatsoever. There's no such high dilution factors, and we've had no issues of that kind.

Our next question will come from Brian Cheng with JPMorgan.

Brian, are you there?

Can you hear me now?

Yes.

Just one quick one from us. Just wanted to follow up on these PD excursions in those patients who had dose prep errors. Was the excursion higher than expected or lower than you expected than what you initially assumed? I'm just curious how we can reconcile the correlation of these excursions to the lack of response, even though that you are seeing the CRS event.

Oh, because -- I'll talk about that. Anytime you slow down or you create problems where the patient can't get to that second step-up dose, third and then their target dose, you're not getting enough drug in there to start really attacking tumor. When you have this bad dosing on day 1, which causes extra tox, extra CRS, you slow all that stuff down so the patient can't get benefit. I mean, you can't keep a patient giving them the exposure they need they're not going to benefit, right? So that makes perfect sense.

I would say that the signal came to us really strongly when the PK excursions, that sort of 3- to 8-fold difference from what it ought to have been in that first day, that actually matched up beautifully with the effect of dead volume when we got retains from our clinical sites and said, please give us the bags you use that we asked you to save so we can analyze them. And then we've replicated the same sort of fold increases from the dead volume in the ultimate diluted drug concentration.

So everything matched up quantitatively pretty nicely. But the key point is you're saying, oh, you got CRS, you got drug onboard, you should get efficacy. That priming dose is about tolerizing people to get the effective doses. And if that priming dose gets [indiscernible] somehow, you are going to hurt efficacy. So that makes perfect sense.

And Bassil, just to clarify for Brian, the PK excursions are always higher than what is expected.

Oh, yes, of course. I'm sorry, I should have said that. Yes.

There are no more questions at this time. I'd now like to turn the call over to Bassil Dahiyat for closing remarks.

Well, thank you, everyone, today for joining us to review these results and the plans across our oncology portfolio, which you can see we're very, very excited by. But really, the most important thing we can say today is really to thank our patients and their caregivers for their willingness to join this study and try a completely new agent and really walk with us on this journey. So thank you very much to them and of course, to our investigators and the care teams at the wonderful study sites that have been driving this study the whole time. We look forward to further updates on these in our autoimmune programs next year. And please have a wonderful rest of your afternoon. Thank you.