... Yes, sir. Okay, good morning. Welcome everyone to day two of the Cantor Global Healthcare Conference. It's a privilege to welcome next participating company, Xencor, and I'm joined by CEO Basil Dahiyat and EVP and CSO, Dane Leone. Thanks everyone in the room and on the webcast for joining us. I'm Steve Seiden, biotech analyst with Cantor. Maybe to start, would love to invite perhaps you, Basil, just to give an introduction to current state of affairs at Xencor. Busy time at the company, obviously, really interesting strategic evolution as well, advancing into some immunology indications. How has that been going? How's that message been resonating? And would love to get an update to start, and we'll dive into the Q&A.
Yeah, I'll sort of take a step back. We've always been a company that strives to or a clinical stage company that strives to build drugs that we can advance in severe diseases using, you know, the platform we've created, our XmAb protein engineering platform over the last couple of decades, to create differentiated molecules that can address, you know, real unmet needs for patients. We've used that across different therapeutic areas with partners for many years, and we have, you know, more recently added specific indications against targets we thought our toolkit could really make a difference for in autoimmune disease, to build off of the long-standing core we've had in oncology.
You know, our specialty is creating molecules where the structure is gonna make something new happen, whether it's our bispecific antibody T cell engagers for cancer or tuned for autoimmune disease use to deplete pathogenic B cells, or whether it's optimized monoclonal antibodies with, you know, extremely long half-life and high potency for autoimmune diseases, or novel structures that use bispecific tools to create multiple inhibition of targets in autoimmune disease. So you could see the thread is always build molecules we can advance to the clinic and get to key inflection points using this protein engineering. So it's all kind of a unified theme. It's just where we lean in at any given time.
I think the message has been resonating quite well because people really wanna hear about something where the drug can position itself in its own competitive landscape with unique differentiators.
Okay, and just on, I mean, the bispecific T cell engager side, obviously, I think this is one of the more near-term upcoming catalysts for the company, which is 819, the ENPP3 T cell engager in clear cell renal cell carcinoma. I think you've guided for an update before the end of the year, perhaps at a medical meeting. Maybe you can just recap that program, timeline, and expectations for that initial data.
Yeah, sure. Maybe I'll introduce the design of the molecule, where we've gotten so far clinically, and then, you know, I know Dane always has some good comments on how that positions in the landscape. So the program is XmAb819. It targets a molecule called ENPP3, which is very, very a surface protein that's expressed very, very brightly on renal cell carcinoma. It's expressed in a smaller proportion of patients, but also very strongly in other forms of not just clear cell renal cell carcinoma, but papillary, as well as a variety of other solid tumors. We built a bispecific molecule that engages that side, the tumor target side, with high avidity. So that is the actual binding affinity, if you measure it, it's modest, right?
But we built the structure using our XmAb tools to create two binding domains to that molecule, the ENPP3 side, so that the low binding affinity of any one of those domains won't really stick when there's not a lot of target around, like when you're on a healthy cell and there's not a lot of handholds for that domain to grab, not a lot of ENPP3s to grab. But when there's a high density on a tumor cell, you've got both domains that are able to engage multiple copies of that target, and that high expression of the tumor cell drives binding. And then when you've got that happening, that then triggers the other side of the molecule, the T cell engager side, the CD3, to bind on a T cell in a way that agonizes it and turns on its cytotoxic function.
It dumps perforins and granzymes, blows apart the target cell, and it just creates a general immune uplift to keep going. And so that's the design of the molecule. So that came out of an effort to look for targets that had that property of high on tumor, low on healthy tissue, because we believed we'd built a tool, our XmAb two plus one design, with those two domains against the tumor and one against the T cell. That two plus one design lets us exploit that difference in target density. So when you have new tools like these XmAb two plus one bispecifics, you can go after biologies that have been left behind. And we were the first ENPP3 T cell engager in the clinic.
I believe there's only one other program we're aware of actively in the clinic, and are both competitor and partner, Janssen, or J&J Biotech now, I guess they call themselves. We guided data for the program from our dose escalation cohorts this fourth quarter. In fact, we'll say today that we're gonna be presenting a poster at the triple meeting, the AACR-NCI EORTC triple meeting in Boston in October, so we'll have that poster. We'll be able to present data on the dose escalation to date, and in particular, the initial part of the dose escalation that we've gotten to in the target range of doses for the molecule, so I know the positioning of it, and people always ask us, "So wh-...
Tell us about why RCC and what matters there. You know, maybe you can jump in on that, Dane.
Sure, happy to. Yeah, no, very excited to get some preliminary data from the dose escalation phase and you know kind of demonstrate why we believe in the program and the advancement of eight-one-nine as monotherapy and kind of the development pathway going forward. So advanced clear cell renal cell carcinoma is complicated from a clinical perspective. You ask a treating oncologist what they use for first-line and second-line therapy, and invariably you get different answers.
What kind of remains true, though, is in the frontline setting, you're in the frontline metastatic setting, you're generally going to get a PD-1, either pembrolizumab or nivolumab, and that's usually either plus or minus a VEGF-TKI of the clinician's choice and comfort, and maybe less so in contemporary times, but still, we do see it, ipilimumab as a combination therapy. What's different about this is you're not seeing frontline chemo. So chemo is not effective in clear cell renal cell carcinoma, so you're immediately going to kind of advance targeted therapy.
Thereafter, the first line becomes a complicated question because there's really a lack of modalities or differentiated modalities available to the clinicians and their patients, and usually, you're rechallenging with a VEGF-TKI, and, you know, increasingly cabozantinib, and you do not rechallenge with a PD-1 as over multiple studies, it's just been proven that there's no efficacy, and it's not beneficial to the patient, and, you know, unfortunately, invariably, those patients progress past the second line, and then you're thinking about the third line, what the salvage line is. The good news for the clinical community patients is belzutifan does have a differentiated mechanism of action, has been approved, is experiencing rapid uptake as a different type of targeted therapy, targeting HIF-2 alpha.
And that is, you know, definitely an advancement, and probably sets, you know, what we would say is a contemporary benchmark for monotherapy development in the space. So in the phase 1 dose escalation of belzutifan, and clear cell renal cell carcinoma, that was, you know, a largely heavily pretreated patient population, where they had seen a prior PD-1 and a prior VEGF-TKI. They had about a 25% response rate, which is great for the indication, but obviously still a little low. And that was fairly stable and translated into the pivotal study, where it was about a 22% overall response rate.
Good mDOR, but on PFS, obviously, did not really move the needle a lot, so about a 5.6-month median PFS in their pivotal LITESPARK-005 study, which was the same point estimate for everolimus, which is, you know, the de facto standard of care is still salvage line therapy that I think the clinical community is trying to move beyond, for a number of reasons. But, yeah, it's been a very tricky setting. Response rates are generally low for these patients that are heavily pretreated, and the PFS range for these salvage line patients hasn't really moved out of that five- to six-month range.
So, you know, with all that said, I'd caveat, obviously, we're doing a preliminary look at dose escalation while dose escalation remains ongoing, and we look forward to moving towards expansion phase and thereafter. So, we'll be focused on initial anti-tumor activity, as Basil said, to those, you know, first couple cohorts that are in target dose range and be looking at, you know, obviously, things like mDOR for those patients that had clinical response. So I think it'll be good to get out there.
We'll present, obviously, a full characterization of the data to date, which will include the safety and how we're thinking about the algorithm, which I guess we haven't touched on yet, but is a unique feature of the T-cell engagers, where you're managing through CRS and kind of those first several weeks to get that patient to target dose while trying to do everything you can to maximize the drug exposure, because obviously, these patients are rapidly progressing, and you need as much drug exposure on board to attenuate that disease progression.
Really looking forward to having that first data in October, and the outlook for eight-one-nine is, I think, very exciting in clear cell renal cell carcinoma, and we've seen, you know, the drugs that make it become billion-dollar drugs, and I think that substantiates the reason that we're gonna put so much time and effort into the development.
Great preview. Just follow up on that programming, ENPP3, do we know as you talked about the different lines of therapy in clear cell renal cell, do we know about the expression of that antigen, and if that differs with refractory status or exposure to any of the PD-1s or VEGFs, et cetera?
We're treating patients in a later line. This is a phase 1, multiple, you know, they've all had VEGF-TKI, they've all had PD-1, sometimes multiple times. And we see uniformly high expression of ENPP3 in the tumor samples we've taken from our patients. We feel pretty good that the initial characterization of ENPP3 and where it's expressed in disease carries through all the modern lines of therapy.
Great. Maybe we can put a pin in that and discuss some of the other programs. Obviously, a lot going on at Xencor. I wanna make sure we cover as much ground as possible. I mean, the introduction of the TL1A assets into the portfolio, obviously, that's a very intriguing target. There's some very promising data, and there's been a lot of strategic activity in that space also. Maybe you can just talk about that opportunity set you see for TL1A and where you anticipate over the next three to five years, ultimately honing your focus for that asset, 'cause it can be IBD, but-
... it can be a lot more also.
We saw in TL1A a target that was sort of the perfect kind of one for us to go after to try to make a best-in-class molecule with our protein engineering tools. Just seeing exciting phase two data from a couple of programs showing in really hard-to-treat IBD potential best-in-disease efficacy, but they're just exiting phase two, and they're molecules with clear liabilities, whether it's you know modest potency, need for multiple dosing frequencies, you know unclear dose, a dose-response relationship or potential for immunogenicity, and yet we still had years to go before any drug was launched, so now is the time to jump in. You know, it's hard to make a biobetter against something that's already been on the market five years, so this was great timing for us.
So we took an incredibly well-validated toolkit of ours, our ability to make high-potency antibodies and our ability to make long half-life antibodies with our Fc engineering to get extended half-life, you know, multiple marketed products that, that have used that technology from us in many, many pre-clinical programs at, at partners or academic collaborators, and so we built XmAb942. And then, you know, we do see a potential broad indication space there. We're focused on the highly validated ulcerative colitis. We're starting this quarter our phase 2b dose, you know, multiple dose randomized study in ulcerative colitis, which we've talked about. It's called Zenith UC, and we'll be hopefully following up in other indications that we'll announce as we go forward.
Crohn's disease is an obvious one, but we're looking at a variety of them 'cause this could be your sort of classic product in a pipeline. You know, and how that plays out and how we think about the next agent, right, target TL1A and use our tools. You know, that's when you go to the bispecific approach in autoimmune disease, which is nascent, but is it gonna be a very important class. You know, Dane, why don't you touch on some of the drivers for what brought us to wanting to make a bispecific with TL1A binding as an anchor, and some of the interesting studies that are coming up soon from big pharma?
Yeah. No, happy to. Yeah, it's the TL1A pipeline that we are developing has, you know, a breadth of opportunity, to your point, Steve, on the TL1A monospecific and obviously, the big question for the class is: Is there really an antifibrotic component-
Mm-hmm
... to targeting TL1A? And I think there's good supportive science to that theory, and now we're just starting to I think test that out in the clinic in early stage in some of our peer studies. We're obviously very interested in that angle as well and are evaluating different opportunity sets to explore that while we push ahead in ulcerative colitis. And I think in ulcerative colitis for a quick moment provides a unique opportunity, as Basil mentioned, to get fast to market after maybe a first-in-class that has some liabilities and test our thesis, where better drug exposure will lead to better clinical outcomes for patients. We've had good validation of that with recent clinical publications, especially from the Pfizer TUSCANY IIB study supporting that thesis.
I think our phase IIB is designed well to
Mm-hmm
... explore that possibility. I think later this year we'll provide more timelines around the phase IIB study as we're going through the global rollout right now and are very excited to, you know, figure out how to bring that data forward and share it as fast as possible as that study is going on. But with regards to the bispecific, which will be first in human in twenty twenty-six, this is kind of the killer app in certain indications where IL-23 has been validated as a bona fide, you know, best-in-disease, as Basil said, target. And that goes, you know, not just in IBD, but in some other disease areas as well. And the biology of TL1A seems to match that as potentially synergistic or potentially orthogonal to some degree in those different disease areas.
And what we liked about putting both those targets in a bispecific is they both have very individually safe profiles in the clinic, meaning you have a lot of play with the dose intensity and the exposure that you can give to patients, which is a lot different than some of the other biologic targets, including the TNF class, for example. So we felt, our engineers felt that we could fix the stoichiometry, dial in the monovalent potency to be equipotent to the parental monospecifics that we were looking at for each of the targets. And that could be something that could be developed like a monospecific drug, very effectively through regulatory development and through CMC into commercial. That would just be truly differentiated.
You know, Basil brought up some of the kind of peer studies that will likely read out this year. Obviously, everyone's waited with bated breath on the J&J DUET study, which is a combination of a TNF and an IL-23 in IBD. You know, I think there's probably a read-through more to the companies that are doing kind of one-one, you know, monospecific combinations or co-formulation approaches, but I think nonetheless, seeing if there is true synergy or kind of a one plus one equals two or a one plus one equals three type outcome for patients with a TNF and IL-23 is still a good experiment. We just view TL1A as a superior target and, obviously, you know, maybe differentiate from the bispecific that's in the clinic targeting IL-23.
We're going after P19 versus P40, and P19 as a subunit of IL-23 in head-to-head studies has outperformed P40, so we're really happy with our construct and, you know, all focus on getting this first-in-human study up and started next year.
And we've certainly seen, you mentioned not only orthogonal, but maybe there's actually synergy to these mechanisms, and we've seen some of that literature, too. So that's gonna be, I think, at the level of IL-17 signaling, which it's gonna be fascinating to see how that plays out. This is bispecific too for a moment. Does that... Everyone's okay here?
Drop my mic, sorry.
Does the bispecific have a role outside of IBD, in psoriasis or in psoriatic arthritis or RA, or does anything stand out as maybe interesting to pursue that?
For the bispecific?
For the bispecific, yeah.
I think we've seen good validation for IL-23 in various skin diseases, right?
Yeah.
And so that's a potential direction to take. I think, dermatology, you know, has some opportunities. I think right now we've got a bunch of data sets coming from Big Pharma that we're gonna look at across a different range of TL1A indications. We've got the IL-23. I think for starters, there's a lot of wood to chop in IBD.
Yeah. Maybe back to the 942, then. This is the long-acting TL1A. I mean, it's if you think about sort of some of the analogs or what we can draw, you know, inspiration from, or learning from precedents in immunology, I mean, we're still this month coming up on, like, you know, the third or fourth iteration of IL-17 data, and like, it just seems very clearly the first movers in any of these spaces, TL1A included, are not gonna be optimized antibodies, and you touched on that.
Well, it-
And so we could stipulate to the fact that they could be improved on, I think.
Absolutely. Absolutely. Not only that, these are diseases where there's cycling therapies. I think the earliest story I can think of is going from Remicade to Humira, right? Lower immunogenicity, easier to use, subQ format, you know, easy dosing, and yet we still had two or three TNF alpha antibodies coming after that that were quite successful, that found particular indications to focus on that had been left behind or offered a different, say, dosing schedule that would be more convenient. From Xencor's own technology portfolio, Ultomiris, the anti-C5 antibody, to go every two-month dosing versus every two-week dosing for Soliris. It's parent anti-C5 antibody that's at AZ Alexion. There's absolutely room to improve these things to give patients better options. We're seeing in HAE now a trend towards, you know, longer-acting agents.
Instead of every two weeks, you've got whether it's RNA or whether it's long-acting antibodies. There's loads of analogs, I think, for making better ones where, in particular, where there's large indications with waxing and waning diseases and multiple opportunities and different indications across the, you know, the immune space. We think the opportunity is still enormous for these molecules.
Just on some of the specific data you've presented, I mean, but how should we think about bioavailability in the gut versus in circulation, where clearly there's pretty profound half-life for nine four two? I mean, is there a way to model this accurately, or is just-
Yeah, well-
... we'll see.
We believe, again, the caveat's always that the data that you're using for the models is always not as much as you would like. We've been very aggressive at adopting and using quantitative systems pharmacology models over the last decade to actually understand better how to even design our molecules. The complexity of bispecific antibodies when you have different targets, and I'm talking about like on the T cell engager or CD3 side, different targets with different recycling times, with different densities and different tissues, really can be a challenge for understanding: how do I make a molecule that's gonna act and be present in the right compartment long enough? You know, we've had success now, our partner Amgen, with xaluritamig in phase 3. That was...
There was a lot of thought on how we built that molecule. That's the STEAP1, CD3. Same approach we took for the TL1A antibody design. We know we're gonna get long-acting molecules in the blood with our Xtend Fc domain for a long half-life. How does that transfer over into the gut where you're gonna have partial transfer? There is some data on that that you can model from animal models. Very high antigen load in the gut, very leaky. There's you're losing your drug. So we took the approach of you really have to get very high exposures in the blood. And we did model it.
We presented the data that, you know, our design metric was we wanna have 99% inhibition of TL1A in the target compartment, in the gut, based on our QSP model for the dosing regimen we came up with. And so with such a long-acting molecule, 71+ day half-life was, I believe, the latest number that we're you know, as the data continues to mature, we hope to present, you know, more refined data. But that long half-life gives us that-
Mm-hmm
... 99% inhibition for an every three-month dosing regimen in maintenance phase, and so we use those tools to really guide us, and I think we've got something that should be really, really attractive in XmA942.
I'll give you the counterpoint to the modeling, right? I mean, obviously, our peers that have developed an IBD are rational in terms of how they dose and design their clinical algorithms. And yet, over the past, you know, 12-18 months, we've seen time and time again that for whatever reason, you have, you know, this patient set that come out of induction period, have not responded, you reinduce them or do you give them extended induction, and roughly you're getting a 50% response rate.
That tells you that almost invariably the modeling is wrong in terms of the drug exposure that these patients need to have, and that goes all the way back to, I think, our first comments of our design thesis, that we knew there was exposure response that did not align with dose response, and that's kind of the why of that is because modeling the soluble TL1A engagement and knockdown in the gut compartment is complicated. It is a complicated model, and so that's our hope. I mean, in a nutshell, that's our hope, is that we can deliver outcomes in an induction period that look more like induction plus reinduction for these first gens.
We obviously even saw that with the TL1A class, I think, in the Artemis UC study, where they did reinduce patients that did not respond to that induction period, and I think got somewhere around a 50% response rate. Saw it with the LUCENT study as well with mirikizumab.
As you think about those precedent datasets in IBD as well for the TL1As, I mean, is there anything, obviously, you have the molecular advantages, we think. Is there anything from a purely trial design standpoint, I mean, there were some patient selection, you know, sort of approaches that you can learn from those and that you can optimize, or is it gonna be pretty straightforward?
I think that there is a, there's a sort of reality of how you're gonna enroll these studies, where you're gonna have to go across many, many different countries, many, many different sites. You always want quality sites, and you wanna have a patient mix that's going to give you a, a sort of representative mix. So prior advanced biologics, you know, you wanna have the right range. You don't wanna skew one way or the other. It could throw off your powering assumptions. This is just about quality trial approach, right? I think the place that we really did a lot of work was on the dose regimen, to have we believe we have a very well-tolerated agent, a high-intensity induction regimen, right? Followed by that high exposure, long half-life maintenance phase to give us the right mix.
Maybe in the last few minutes, I wanted to get your perspective also on just B-cell depletion, B-cell-depleting antibody programs and autoimmune disease. I mean, in general, you know, there's been some fascinating and inspiring data from CAR T in this space, but you know, from a therapeutic index standpoint, who knows, for some of these diseases? But from an efficacy standpoint, I mean, do you think you can get deep enough depletion to really achieve sort of like the immune system reset and a durable cure with a depleting antibody, as you can seemingly get with a cell therapy?
I think we don't have a good understanding of that even for the cell therapies. Are we truly getting immune resets? There's anecdata there, right? There's not large datasets. I think the promise is there, which is why we're chasing after it. I think beyond just the therapeutic index issues around cell therapies, just the simple logistical challenges of getting a rheumatic disease patient or a neurology patient to undergo the difficult preconditioning regimens, the hospitalization, the multiple sort of weeks of running around to get the therapy, versus a relatively straightforward, you know, drug. What do they always call it? They call it a therapy in a bottle, right?
Which is like every therapy we've ever heard of for the last hundred and fifty years, is a therapy in a bottle. The world does not revolve around cell therapies. But I think an analog we have is from oncology, where you have complete response rates for CD20, CD3 antibodies that rival the CAR Ts once you actually look at the CAR T therapies on a level playing field with true intent to treat, counting the denominator properly, looking at relevant lines of therapy. So I think we have precedent for these kinds of truly, truly deep and durable effects from oncology.
I think the anecdata of patients who are in long-term remission is something that we do believe deserves real follow-up with agents that can deplete B cells much more deeply in the tissues than the tried-and-true Rituxan, though I will say there are anecdata, rare, but anecdata of Rituxan, one induction regimen, and a patient's done forever, right? So clearly, there's something real about this B-cell reset hypothesis. We just need good drugs. You know, once you get beyond the cell therapy hype cycle, good drugs to actually address the need to have deep B-cell depletion in a tolerable and practical regimen.
Just to close in the last minute, wanted to widen the aperture a little bit, and we've talked about, obviously, the breadth of development that you have going on, and historically, you've been prolific developing novel antibodies. I mean, do you have... As you sit here today and you forecast, you know, the outlook for Xencor over the next few, you know, call it three years, do you have enough on your hands currently with the bispecific program, the TL1As, you know, the B-cell, depleting, portfolio? Or are you gonna continue to, you know, develop novel assets, look for strategic partners for those? What's the strategic focus going forward?
We're always gonna try to exploit our engineering tools to make the next generation of great drugs, right? We didn't know we were gonna be making bispecific T-cell engagers a decade ago. We started working on that because we thought that there was an area, that there was a real unmet need, that engineering could help. So absolutely, we're gonna keep pushing forward into novel structures and novel modalities, but the core focus of the company is getting our clinical data and getting the results that we need. The research piece is about, you know, being creative and following your nose for what might be next, but we got a lot of stuff to do, a lot of wood to chop the next three years in the clinic.
We're looking forward to following along, and it's gonna be an exciting time. Looking forward to also the data at the back end of this year at the triple meeting. So thanks so much, Dane and Basil, for the time. Thanks to everyone in the room and on the webcast for listening, and hope everyone has a great day at the conference today. Thank you.
Thank you, Steve.