Good afternoon and welcome to day two of the 2024 Bank of America Healthcare Conference. Thanks for joining this session with Xencor. My name is Alec Stranahan. I'm Vice President and Senior Biotech Analyst covering Xencor here at B of A, and I'm pleased to be joined by Bassil Dahiyat, President and Chief Executive Officer of Xencor. Bassil, thanks for being here.
Oh, thank you so much, Alec. It's a pleasure. A lot of energy in the conference today, so having a great day.
Yeah, good to hear it. Good to hear it. So I've got some questions here, so we'll just do it fireside chat format. If anyone in the audience has questions, feel free to raise your hand, and a mic will be brought over. But Bassil, maybe just to start at a high level, could you maybe walk us through your approach or the philosophy that Xencor has to drug development, whether it's discovering new assets or pushing ongoing studies through points of value inflection?
So we are always pushing hard on building out our protein engineering capabilities. We've been engineering antibodies now for about 20 years and have always sought you know, to be at the very leading edge. We created the first high-effector function Fc engineered antibodies. We helped pioneer long half-life extension through antibody engineering. And we, you know, came in early into the new wave of bispecific antibodies, particularly T-cell engagers, but then broader than that, that emerged about a decade ago, and we're still pushing forward in that way, pushing at new biologies we can access with protein engineering. So that's always central to what we do. And the purpose is to feed our internal pipeline. We want to be a company that develops and markets its own drugs one day.
Along the way, as we create new science and as we learn more about the molecules we've built, sometimes it makes sense to take a molecule and find the right partner for it so that it can move on, we can get value from it, but have it, have it have a much higher chance of success than if we tried to hold onto everything on our own. You know, and that creates room in our own pipeline for the things that we think are best suited for our, relatively small biotech company to internally develop. So it's really about using the technology to create molecules that we can push forward. Over the last couple of years, we've focused on the internal early clinical development, hopefully maturing soon, of our T-cell engagers.
That's, I think, really the focal point of the company, is exploiting what we've created in the labs over the prior years and getting that key clinical data to validate our own molecules and advance them.
Great. And I guess from a, you kind of touched on this, but from a business development perspective, where's sort of your sweet spot in terms of running larger studies on your own versus partnering out? And do you think there's a scenario where you would ever bring a drug through approval and commercialize it yourself, or is the R&D side sort of where you think you can create the most value?
We absolutely believe there's scenarios where we could take a molecule all the way through and commercialize it ourselves. I think you have to, it has to be the right fit where you can support the late-phase trials, cap, with the right capital and operationally to get a good answer that is an answer that truly does create a differentiated product profile that will be successful in the market. We've seen, you know, some small companies do everything they can, heroic efforts to get a drug approved, but it's very challenged because the competitive landscape shifted underneath their feet, and it ended up not working out and not letting them grow their company further. So we're very cognizant of having to make sure we fit in that environment three or four years later.
You know, an example of a molecule where we think there's enormous potential in it but it didn't make sense for us to carry it forward alone is our plamotamab program. We had developed and, you know, nearly completed phase 1 for our CD20, CD3, T-cell engager for lymphoma and other B-cell malignancies, and looked at the competitive landscape where Roche and AbbVie were a little bit ahead of us, and that early label with monotherapy data in late line to get that jump into the market was getting harder for us to imagine getting to and competing. So we found a tremendous partner, best heme oncology company in the world, I would argue, Janssen, to develop it with us. That's a case where the business environment really decides for us what the best way to advance is.
The beauty of it is we have that strength of that research engine that creates a rich pipeline all the time. When we did that deal, we had our ENPP3/CD3, a renal cell carcinoma T-cell engager, really the only in class there at the time, and still it's very little competition in that arena where if we can show a win, that's a really clear line you could imagine to designing amenable and feasible late-stage studies and hitting a market that would be receptive. So we want to think about those pieces, not feel like we're tied to any one asset, because we're not. We have that we have that benefit. So our T-cell engager programs, we think, all have that fit of having a clarity of outcome when we do clinical studies.
Do we hit certain metrics on response rate, on duration of response, that we can then say, "Okay, we now have high confidence in a late-phase study that we could run on our own." Not all assets have that, and so we try to be thoughtful about it.
Right. Then it sounds like there is a competitive landscape piece. Beyond just the data and the profile of the asset, there's, you know, it takes years to even get to that point where you can say, "Yes, this is a drug that could benefit patients." And, you know, things may change by that point, right?
Absolutely. You see that happen all the time. Again, we see it with such cases that we've studied very carefully from other companies with really great technology who've done a heroic job in clinical development where the world's just different when you get to the finish line, right?
Yeah.
and so, you know, sometimes, we proactively look at a field where we've created great science and decide it makes sense to just do this with a partnering approach. I would say infectious disease is exactly in that vein. That is a very highly specialized area for therapeutic development. Markets are very complex, very challenging to commercialize. There's some companies that do it well. We create technologies, in particular ones that let an antibody last a really long time, our Xtend technology, which is commercialized now in ULTOMIRIS, in Xevudy, which is the sotrovimab COVID antibody, is in a range of infectious disease antibodies by our partners at Vir or the NIH in doing all sorts of wonderful work in third-world countries. We're not going to, we're going to proactively say we are going to partner in those areas.
Yeah.
It makes no sense for us to think we can build that expertise and market drugs on our own there. In fact, we were very excited to see our Xtend technology show durable protection in a recent clinical study published in the New England Journal by our wonderful colleagues and collaborators at the NIH. Durable protection for six months from a single dose of an Xtend engineered antibody, our half-life extension technology, protecting children in Mali from malaria. And I think that's the kind of half-life extension that can play in a lot of markets and a lot of applications like ULTOMIRIS, like sotrovimab, and we, we hope others in the future. But to see it helping people in that kind of setting, where modern advanced technologies can have an impact is really, really, really, really fulfilling.
Great. Great. And I want to talk about, you know, what's going on in the in-house pipeline, but maybe one last high-level question. Obviously, you've had a very successful track record getting partners signed up. What's the draw for the partners in terms of what do you guys bring to the table that's unique or differentiate from your platform?
I would say, we have a breadth of tools that we've created that we've built always from one central foundation, which is the biophysics of the molecule have to be great. We are engineering proteins, particularly antibodies and antibody derivatives. Antibodies are wonderfully evolved to have beautiful biophysical properties. They're super stable. You can purify them, store them frozen, store them in liquid for months. Their half-lives are already great, one, because they have this beautiful Fc domain-mediated recycling that they do through the FcRn receptor where our Xtend technology can exploit that to greatly extend that already robust half-life. But there are also long half-lives because they're super stable. They've evolved to be that way. You don't want to mess that up.
Too often, when people engineer molecules, antibodies, to make them do some clever trick, right, whether it's I want it to only activate in here or I want it to have a certain potency or whatnot, or I want it in a different format, you break what's good. You know, we are protein engineers. We're biophysicists. We always make sure that that's really robust. And the way that's translated into sort of a manifestation that you can look at is partners know they can just take one of our tools or take a molecule we built and just plug it into what they're doing. They don't have to worry about manufacturing. It behaves like an antibody. They just take it wherever they normally do it.
We have our molecules, bispecifics with complex engineering we've done, manufactured by our partners at Amgen, at Novartis, at Janssen, at, I'm forgetting some, Astellas, at, you name it, right? We do it at various CROs. They know it's going to do what we said it does. All they have to worry about is was their therapeutic hypothesis right? Did I pick the right target? Did I pick the right, you know, potencies and affinities and all those things you don't know about the biology? But they know that it's not going to create problems. That, I think, has led to this breadth of partnerships, right? That's led to our own rich pipeline that we're not stuck hanging onto any one molecule. So I think that was that is a key differentiator.
So even though a lot of bispecifics look the same these days, and in truth, a lot of companies can do it, technology edges don't last that long in our world, right? You know, I mean, like, how many new iPhones have you had in the last decade, right? But certain principles are going to translate over the long haul, and that's one of them.
Okay. Yeah. No, that's great. Maybe we can start, you know, looking at the pipeline. First with 819. This is your ENPP3 T-cell engager. Could you maybe speak to some of the data we've seen, from your T-cell engager collaborations and how this feeds into your confidence for, for 819?
Absolutely. So, XmAb819 is a bispecific T-cell engager. It binds a target that's very bright in renal cell carcinoma, particularly clear cell, the predominant type, and it binds CD3 on the other side. So it brings those cytotoxic T-cells together with the tumor cell. By binding CD3, it activates it, and it kills. It's a well-proven mechanistic class, T-cell engagers. The trick with solid tumors, where there's an enormous opportunity for T-cell engagers, but there hadn't been much progress until recently, is solid tumor targets are overexpressed in the solid tumor. But these are solid organ cells. That means almost all targets you have are also going to have some expression on important solid organs.
Unlike blood cells, where we know we can blast them with radiotherapies, we can blast them with really potent T-cell engagers, you don't want to blast every healthy cell in a patient's lung or kidney or thyroid or whatever. It's dangerous. So we have to separate out those activities. That's where these new engineering approaches we've taken, we think, are starting to bear fruit. XmAb819 has two binding domains to ENPP3 and one to CD3. Those binding domains to ENPP3, converse to what you might think, don't stick very well individually. But when there's two of them, they can grab on, but only when there's a lot of target to grab. That's on the tumor cell. Smaller copy number of target on the healthy cell. That's why we picked ENPP3. That gives us that selectivity to only bind and hopefully only kill those high-expressing tumor cells.
So that's the hypothesis. We love how we designed the molecule. All the in vitro properties are great. In vivo work is great. We're in the clinic. We've been escalating. It's been going well, but we're still moving up now, right? Now go to our partner that I think you alluded to, Amgen, with a molecule we created a little earlier called now xaluritamig. It was AMG 509. That binds. It's exactly the same format design as 819. Two binding domains to this, its own target, STEAP1, a prostate cancer target, one to CD3. Tricky target, hard to bind to. Amgen came to us with that target because they couldn't make molecules with other formats that actually even stuck to and killed prostate tumor cells in vitro.
With our 2+1 format, with some engineering work we did, it did a nice job of that. It seemed to also separate from cells that don't express much STEAP1 on it. So it had that selectivity. So that's what the 2+1 hopes to offer. Now in humans, they disclosed data at ESMO. We were extremely happy with what we saw. In late-line prostate cancer patients, they had a 40% RECIST response rate in the RECIST-evaluable population. That's about 60% of the study. Overall, they had a 60% PSA 50 rate. The durability is early, but it's looking really promising. They've expanded the program. They're going great guns on that one. That's not an easy target, right? And I think this 2+1 selectivity approach had its first real test, and it passed with flying colors. So we're very excited by that validation.
We're very hopeful that our ENPP3 program can have similar kinds of success in renal cell carcinoma, an area that is ripe for a cytotoxic agent to jump in, where right now checkpoint inhibitors and TKIs are really the only game in town and leave a lot of room for improvement. We have a second program with the same format, subtle differences in design, but the same format. Our Claudin 6 x CD3 called XmAb541, we just started dosing patients a month or two ago. That's expressed heavily in ovarian cancer but also a variety of other histologies. And, you know, we think that this new format can really open the door for solid tumor CD3s in a much bigger way, and we think that's a huge opportunity, so. So, thank you to our partners at Amgen. Now we have other partners.
Astellas is advancing a Claudin 18.2 x CD3, same format. We recently saw they disclose that they are in expansion cohorts. They started about two years ago. I don't think they've shown any data, but we, we think that progress is a good sign. So I think momentum's building.
Yeah. And I will say, you know, all three of those targets, ENPP3, STEAP1, Claudin 6, they're not, like, the canonical targets, but they are, like, the emerging next-gen targets, and they're very interesting. Maybe on circling back on ENPP3, I think this is one target that has had some development in the past, and it's a little bit checkered, so to speak. You know, how does you did a great job painting a picture of how the T-cell engager, how you've developed it, could maybe overcome some of these limitations, but is that really the stickiness piece, or is it the dual binding that's?
Yeah. I would say with ENPP3, I'm sure you're alluding to the earlier development that was done with an antibody-drug conjugate about 7 or 8 years ago using sort of earlier generation auristatin technology and MMAF, antibody-drug conjugate to ENPP3, and they used it in renal cell carcinoma in phase 1, and they saw activity. They saw RECIST responses across multiple patients, very heavily pretreated renal cell carcinoma patients. They also saw significant toxicity that was the characteristic toxicity of the warhead. In particular, these early generation MMAFs with non-cleavable linkers with or hydrophobic linkers with hydrophilic, more hydrophilic warheads, they saw ocular tox where ENPP3 is very unlikely to have much to do with that, but boy, you see a lot of it with early generation auristatins.
And so, you know, we think that was really a choice of warhead and mechanism rather than the target, but they saw responses in a very difficult-to-treat tumor, which we thought was very encouraging. Immunotherapy, CD3 T-cell engagers, we think it's just fundamentally different. We have been escalating. We're still escalating. That's going well, and so we're optimistic.
Okay. And, you know, you mentioned the data for xaluritamig, AMG 509, ESMO last year. There was a lot of excitement around that early data, and I think it was a positive surprise for a lot of people. Could you maybe talk through your excitement for this asset, and, you know, when we could expect the next wave of updates from this program? Is that kind of under Amgen's control?
Yes. Amgen controls that. We have a, we, we built the molecule, have it under license, then we are not co-developing it. We do receive, obviously, disclosures from time to time, but they make all the decisions. What we do know is they've guided that they're greatly expanding the program. They greatly expanded their phase 1. They've moved into earlier line combination with taxane into, you know, immediately post-androgen deprivation, prostate cancer. They've moved into castrate-sensitive with, with combinations with, ADTs. They're looking at reduced monitoring of patients for CRS to sort of move more to the outpatient setting, and they guided late this year, early next year, they should have next wave of plans. That's what we know. We know they maintain strong enthusiasm and investment in it, but we really can't say more than that.
You know, I think the strongest data in that line of prostate cancer patients that we've seen.
Okay. How about for 819 and 541? These are, you know, your wholly owned assets. Any updates on timing to data?
So on timing to data disclosure, we usually, as a company, make those disclosures of when to expect data. When we get a little closer to them, when we have certainty, we have disclosed that we expect to be at our target doses and characterizing those this year. We're for 819. And like I said, we're escalating well, and we would be able to make decisions on next steps of development, which, if all goes well, would be expansion cohorts in monotherapy, perhaps in combination as well. We strive to move earlier line, of course, than this sort of third line that we would likely be in initially. And so that's, we expect to have a lot of information by the end of this year. When we share it would emerge, you know, when we're really close to ready to sharing it.
We are very mindful of the competitive landscape. We do know there's a competitor that just started an ENPP3 CD3 combination molecule in the clinic, a much, much, much, much bigger competitor than we are, one that we know very well, J&J. We have a lot of respect for them. And so we want to be very mindful of not teaching about dosing regimens and priming and step-up dosing strategies too early. So we're going to be really careful about that. Now for XmAb541, our Claudin 6, we just dosed our first patient a few months ago, and we are escalating now, but it's a bit early.
Okay.
It's a bit early.
Okay. That makes sense. And I want to next move to vudalimab. And you know, you, you mentioned a couple of times, you know, watching the competitive landscape. You know, how does launch of Pluvicto feed into how you're thinking about vudalimab's potential?
Yeah. I think it's going to certainly have a big impact, especially as now they're moving into pre-taxane with the PSMAfore data. I think is very likely in our read to place it into that setting. How broadly it gets used, access issues, I can't speak to that. I do know that there's a lot of belief in the MOA, you know. We think that's actually really promising for a checkpoint therapy in prostate cancer if one can be made. We know that all the PD1 inhibitors that have been tried have not really gotten anywhere, not a lot of activity.
You know, we've seen recent data from academic groups showing that just short, essentially, priming use of Pluvicto followed by maintenance PD-1 therapy can have really excellent, at least medium-term outcomes for patients with PSA drops and good PFS. So I think that really bodes well for potentially thinking about combination uses with the radiotherapy. But that's a little bit down the line. You know, we are encouraged by the monotherapy activity we disclosed earlier this year for vudalimab, very late-line patients, all with measurable disease, which means they're very advanced and very sick, you know, PSA90s and resistant responses in a significant percentage. So we're optimistic that a checkpoint will have a very important could have a very important role in that setting, whether it's with Pluvicto or not.
And I think it's going to be interesting to see how a checkpoint and a CD3 like xaluritamig could potentially even play together.
Right. And, you know, given the non-overlapping AE profiles like that, that's a positive setup for the combos.
Yeah. And we are expecting to see the first for any CD3 bispecific solid tumor, the first real chunk of combination data with a checkpoint inhibitor in the next, you know, little while, a year or two maybe, from Amgen's DLL3 program, tarlatamab, where they've moved into frontline in small cell in combination with PDL1, you know, a Tecentriq or durva. I'm very eager to see how that plays out. To your point, non-overlapping toxicities, there's potential there, but we're going to have to wait and see.
Right. This is a question we get. It's more of a fundamental question about the rationale underpinning vudalimab, which is, how does this stack up versus ipi-nivo combo, right?
Yeah. We specifically designed the molecule to be limited in the T cells it binds to ones that express both targets. So ipi, nivo is going to have ipi stick to any cell that has enough CTLA-4 on it, nivo to any cell that has enough PD-1 on it. And some of those cells are going to have both. Some are going to have one or the other. We purposely designed vudalimab to really be very weak at binding CTLA-4 alone. You really need to have the PD-1 engagement happen on a cell that has PD-1 on it that cell would also you'd be able to block CTLA-4. We have reams of preclinical data showing that it's really very selective for those. So we're selecting for the double checkpoint-positive cell, which tends to be the ones overrepresented in the tumor microenvironment. So that's, that's a distinct difference.
We hope that that gets that selectivity maybe changes our AE profile. I think that the data we have now from 200 patients suggests we are getting a bit of a different profile, but that we don't give up much on efficacy, right? I think the challenge with any kind of use of CTLA-4 blockers is you can only give them for a little bit, right? We know that the way they've made ipi-nivo combos tolerable is by really limiting the ipi, lower dose, maybe 4 doses, maybe 6, but short-term, and then you're done, almost like a priming or an induction phase. Our goal is to have longer-term with that. I think some of the early data we've shown with, you know, really durable RECIST PRs and PSA 90s in late-line prostate cancer patients suggest we are getting that CTLA-4 boost.
Now, can we manage it, you know, and how does it play out as we get to bigger patient numbers?
Right. Right. And I guess in terms of typical AEs you would expect, I think autoimmune hepatitis was one that popped up, which feels more like a fringe event than anything. It, you know, anything that you would highlight, I think it was one patient out of.
Yeah. We have over 200 patients treated with relevant doses, say 10 mg/kg up or the equivalent flat doses of vudalimab roughly across a range of indications, you know, and we saw one very unfortunate event, grade 5 autoimmune hepatitis where a patient with very involved disease and multiple adverse events that had happened that were immune-related, grade 3s, but they were in response so the doctor didn't want to take them off drug, right? Really late-line prostate cancer patient. They also had some kidney AEs. They had other enzyme being tweaked. They didn't want to take them off drug, and so they eventually expired, and attribution was to autoimmune hepatitis.
I would say of the 200 patients we treated, our rates of any kind of hepatic, IRAE toxicity, using very broad term sets to pull it out of our database, compared to the ipi/nivo published labeled data, suggests that our hepatic events are roughly equal, if anything maybe a touch lower. So we don't think that there's a broad pattern there, though we are very, you know, very sad about that event, a patient who was doing well on the drug and, you know, they didn't want to take them off.
Right. Right.
So we think we've continued to enroll patients. We've progressed. We've had no real slowdowns as a result. Sometimes, you know, with these really sick patients, there's a lot of difficult choices our treating physicians make, and I don't envy them. But we're advancing. We expect in the first half of next year to have enough data from our monotherapy cohort as well as from our taxane combination cohort to make clear decisions on what we want to do in prostate cancer there. You could go either way depending on the data that you see. And we are also in our frontline lung study in combo with chemo, really chasing after some exciting very early data that AstraZeneca, our competitor, saw with their PD-1 CTLA-4.
We should have enough early data there early next year to decide whether we're going to go into the big randomized portion of that study, so, because we're looking at safety first. So yeah. So that's how vudalimab's stacking up.
Great. Looking forward to those updates early next year. Maybe the next asset, 808, so this is your first internal CD28 bispecific to enter the clinic. Maybe at a high level, Bassil, could you maybe talk to the CD28 arm of your platform, pun not intended.
Yeah. I know, right?
and how this is maybe differentiated versus other CD8, CD28 developers.
Sure. So the idea with CD28 is how do you harness this super powerful part of T cell stimulation, which is that so-called signal two. The T cell receptor where CD3 resides is signal one. There's signal one from T cell receptor engagement happening that when you use a PD1 therapy, you're taking the brakes off of, right? Signal two through CD28 is how you keep that activity going. You sustain it. You amplify it. Signal two. So CTLA-4 blocks signal two, so ipi actually releases the brake on signal two. Our CD28 engager, unlike the ancient CD28s tried two decades ago that had way too much activity, so-called super agonism, turns a T cell on absent of any kind of target engagement on the tumor or T cell receptor engagement. That's bad.
There have been some early data sets from competitor CD28 bispecifics that attempt this tumor selective activation where only when the other arm of the bispecific is sticking to the tumor do you turn it on, that have shown some promising activity but also toxicity from very early data sets. Ours are about 20-fold less potent. We purposely pushed on the safety side there, and we hope that was the right decision now. We, we think maybe in retrospect from seeing some competitor data with the more toxin you'd want, that it might have been the right decision. We picked B7-H3 because it's expressed on a lot of tumors. The initial phase one work is in combo with pembro, though we do have ambitions to combine with CD3s as well because any signal one, CD3 or PD-1 is a, is a potential combo agent.
We're looking at a lot of tumor types, a lot of prostate cancer because it's an early read from PSAs but also triple negative breast and RCC and, and melanoma and some others. So we're looking to see whether we can boost up that PD1 activity, early on. It's, it's a really promising but very early new T cell engager.
Yeah. I think within the class, a lot of people are looking forward to Regeneron's EGFR x CD28 data at ASCO.
Yes. We are as well.
Yeah. Anything that you're looking out for in that update?
Gosh. Yeah. I want to see the dose levels they got to when they started seeing boosting of their PD-1, cemiplimab's activity. And I want to see are they separating toxicity from efficacy even on a patient-by-patient basis because for their earlier PSA, PSMA combination program, it seemed like patient-by-patient, efficacy and tox.
Right.
I want to see how that plays out as well as the severity of tox.
Right. Right. Okay. Well, I think we're about up on time. So unfortunately, we'll have to leave it there. But Bassil, really appreciate the, the engaging discussion and for your participation in the conference.
Oh, thank you, Alec. We're having a great time.
Thanks a lot.