Xencor, Inc. (XNCR) Earnings Call Transcript & Summary

Good afternoon. Thanks for joining us for another session at the 41st JPMorgan Healthcare Conference. I'm Brian Cheng. I'm one of the senior covering analysts at JPMorgan. Presenting next is Xencor. I'll turn over the stage to the management team. After the presentation, we will follow up with a Q&A session.

Thanks so much, Brian. It's a pleasure to be back at JPMorgan in person. My name is Bassil Dahiyat, I'm the CEO at Xencor. And I'll kick it off by showing you our forward-looking statements slide. So we are a protein engineering company. That's we were spun out of Caltech 25 years ago to do. That's still the core of everything we do. We focus our engineering work on manipulating immunology molecules in the body, mostly antibodies and cytokines. And we've built a very modular toolkit of different domains of these structures that we use to construct drugs, both for ourselves and our partners. We really started out in this work focusing on the bottom half of that structure, antibody Fc domains, building a repertoire or a library of ones that give us access to different functions that we can control, whether extending the half-life of an antibody, controlling its effector function or changing its structural organization. And we have that same philosophy and a whole set of tools that we have in-house for building complete biologic drug candidates. And we've used this approach to have a dual strategy of building a robust internal pipeline that we pursue ourselves, that we support with a partnering effort that consists of everything from licensing intellectual property, some of our tools to partnering and co-development arrangements with large companies to develop our pipeline. So that's sort of the top one company in a nutshell. And to date, we've participated in or created 3 drugs that are now marketed in producing royalties for the company. So the way this works is we see it as layers of value creation that starts -- as the foundation of the base at the bottom is our technology platform. That enables us to have this diversified approach to building the company, where we can create a strong financial foundation with a combination of partnering that helps drive both our technology innovation work as well as supports our internal pipeline. We are -- we finished 2022 at about $610 million compared to starting '22 with about $650 million. That drives really the efforts we have in the clinic, building data sets that can convince us to move our molecules forward, and we have several -- a couple of programs in mid-stage clinical trials. And then always, we're building on top of that new technologies whether it's new bispecific antibody formats like our novel 2 plus 1 engagers, novel engineered cytokines that we've started reporting data on over the last year to create that next wave of candidates as well as technologies that feed down into the bottom layer. So this approach means we have to have a pretty stringent and disciplined way of vetting our candidates because we're constantly creating new approaches and new molecules, and we're always tackling highly risky biology to go after with these new molecules. So we have to be stringent about what we're willing to invest our dollars into to go to where our ultimate aim is, which is late-stage development registration trials of our own. So in our early clinical testing, we decide whether we have sufficient data to pursue further development ultimately into registration-enabling trials. We haven't made pulled that trigger yet internally, whether we partner, like we recently did with our plamotamab CD20 x CD3 about a year ago with Janssen when the risk reward of investing our own dollars in a very competitive and frankly, large indication space, wasn't -- it didn't make sense or sometimes we stop programs if the data warrants it. So that's the overall strategy. And I'll sort of point at the pipeline right now. You can see that it's really dominated by that light blue color, which is at the bottom, you can see means solid tumor targeting. That's where we see is the most opportunity to bring antibody tools, in particular, bispecific antibodies and immune therapies to bear outside of the dominance of checkpoint inhibitors in certain hot tumors, where there's a huge unmet need. We have at the bottom our heme work. And in the middle, an opportunistic molecule that was really built by using these modular tools in a way that just made so much sense. We kind of had to do it. So this kind of breadth in the pipeline means, again, we have to be stringent about which ones we carry forward. Vudalimab is currently in Phase II after promising Phase I data and in prostate cancer that I'll get to in a little bit, but the rest are still in that earlier testing stage, in that Phase I stage. So this is a real snapshot. The only real technology side, I'll show is it's truly an approach of plug-and-play different modules that we've built over the years and that we constantly build new ones of -- to create, whether it's bispecific antibodies with the diversity of structures. There are engineered cytokines with the properties that we need. And the goal is to bring antibody properties, pharmaceutical properties to a broad range of new biologic structures. That means long half-life, easy to make, stable and compatible with standard manufacturing process. You don't have to build your own in-house manufacturing, you're not creating a novel gene therapy or cell therapy. Keep the parts that work great and do the risky biology to create the therapeutic that you want. So this modularity is what drives this partnered pipeline. And much of this is just technology licenses we've done where we've not really devoted any time or effort of dollars but the royalties and milestones and upfronts we received help fund the effort. So at the top, you can see Ultomiris, where our Fc domain extended the half-life of the first-gen molecule to now being every 2 months. That, I think, was approved in late 2019. And at the bottom, you see a wave of bispecific antibodies that as our newest technology that have emerged into the clinic, and we hope to see those advanced as well. And just yesterday, we are -- today, we saw the announcement of obexelimab started Phase III trials at our partner, Zenas. So now I'm going to jump into our pipeline and say that it's organized along 3 structural classes: to the left, our T cell engagers, where we engage either CD3 or CD28 on T cells and recruit that T cell and activate it in the relevant way against tumor-associated antigens directly; there's our dual checkpoint molecules, where we're trying to use the selectivity you can create in a bispecific structure to control high checkpoint inhibitors work and manage some of the toxicities that they have, in particular when you're adding things like CTLA-4 or IQOS to the picture; and then our cytokines. Engineering cytokines to make them more drug-like, which is something that they are not naturally and perhaps can overcome the things that have limited them from being really successful therapeutics. So I'm going to only hit highlights of our pipeline today because of the time and obviously can answer questions about any of our candidates. So the first part I'm going to talk about is actually one of our cytokines. This is a relatively new class, but we just reported data from our second clinical program in November. And the approach is unified and modular, like I've been saying for our whole technology approach. Cytokines are highly potent molecules. That diagram on the right shows you sort of hypothetical activity on the y-axis and time on the x-axis curve if you dose a cytokine into a patient. Native cytokines have enormous potency that's evolved over the millions of years and very short half-life because they're engineered to act locally, right? One cell touches another cell tells us what to do and boom, it goes away and that's it. We want drugs that we can inject and have last for weeks and weeks. That doesn't work when you're injecting something that's hyperpotent so it's toxic. Our approach is bring it back into the drug-like regime of potency. So we dialed down the potency and you can see in that middle panel, 100- to 1,000-fold from native cytokines to create our candidates. So we have a much lower and slower burn like that blue curve on the right. You get activity, it's maintained for a long time because the truth about cytokine is they've also been evolved to rapidly clear themselves. They hit the cell, they hit it hard and then they've evolved to stick to the receptor and the cell gets rid of it. We want to avoid that. So our general approach is reduce the affinity from potency attached to our Fc domains for ease of manufacture and standard processes and stability, and that lets us also array structures in interesting ways. I'll show you in a moment. And so that's the general approach. The program I'll highlight is when we reported data on in November, a wholly owned molecule, XmAb564, it is an engineered IL-2, where we have opposite the usual paradigm, try to enhance its immunosuppressive functions. And we did that by enhancing its binding affinity to 1 of its 3 receptors, the IL-2 receptor alpha or CD25, but then knocking down the potency to its other receptors that drive much of its effective function, the beta gamma by over 400-fold. So you have a molecule that won't stick to cells that really heavily express the beta-gamma like conventional cytotoxic T cells or NK cells, but it sticks great regulatory T cells that have high CD25. So that's the selectivity design we went after. It's on one of our Fc domains. It let us make a structure that has just 1 IL-2, in contrast to some earlier efforts from other companies that have 2 IL-2s that could potentially stick too tightly to the cells you don't want to because there's double the binding power. We wanted to minimize that and our Fc heterodimeric structures for our bispecific let's do that, and we applied our half-life extension technology. So we have that whole package, and we were delighted to report in November, the first clinical data, a single ascending dose study, a safety and pharmacodynamic study in healthy volunteers, where the punch line is, it was a well-tolerated agent, really the adverse events were primarily limited to injection site reactions. It's subcutaneously delivered as you would want for an autoimmune disease drug. And we saw remarkable expansions of the cells we wanted to have grown more because they're immunosuppressive these regulatory T cells, these CD25 bright regulatory T cells, reaching over 100 fold, which we believe is the highest reported from this kind of clinical trial. But the remarkable thing was their exceptional durability. This long half-life we've designed seem to have translated to a really robust and long duration, which is something we're going to try to translate now that we've advanced into our multiple ascending dose study into a longer-dosing interval. The class has so far been limited to every other week dosing for the programs that are in Phase II at some of the large pharmas. We want to extend beyond that, and we believe our molecule has that potential. And I'll note, this is the second potency reduced cytokine that Xencor has created. Our IL15 program for oncology partnered with Genentech, where we split the worldwide P&L 55-45 with them, also had remarkable target cell expansion and exceptional durability with buildup over multiple doses. So here's just a quick snapshot of the data from that IL-2 study. You can see on the left panel, day 21, CD25 bright Tregs for dose levels in the middle green and then the 2 blue 4 and 5, you see significant elevations of the CD25 bright regulatory T cells at 3 weeks. This, we believe, is the best we've seen reported, where competitor molecules in similar settings had all faded away by 3 weeks out. And then at a top dose, this remarkable spike, which tracks the evolution along our potency curve if you track the concentration in the blood to what we would have expected from in vitro data. So this remarkable durability is something we're trying to, like I said, exploit in a multiple ascending dose study, which has started in atopic derm and in psoriasis, and we hope to report data from an early 2024. So this is just the first 2 members of this reduced potency class. The Genentech partner, XmAb306 to the left, the XmAb564 wholly owned at Xencor. This reduced potency paradigm seems to be working nicely, and we're off to the races with the class. We'll report more on this later. And I will say that the IL15 Genentech started the third clinical trial in our collaboration recently in myeloma just this month. So that one is going. We hope to report more data on that soon. So next, I'll touch on a similar kind of approach of tuning a molecule to make it fit purpose. This is our dual checkpoints. And I will say that the key area to focus on here is the right side of the slide where you can see we labeled double positive cells, if we've designed a molecule the target 2 targets, we wanted to only bind the cells that have both of those targets. We want to limit the exposure of the immune system to these highly activating checkpoint inhibitors we have to control and improve tolerability and still hit the cells that are going to be the most active against the tumor. So vudalimab is the lead in that class, it's our PD-1 by CTLA-4 bispecific. We completed Phase 1 about a year ago in a wide range of solid tumors. We had multiple responses in nearly every patient. Those dark balls in the waterfall are patients who've had prior and progressed on prior checkpoint therapy. So it's quite active. It's tolerability profile seem differentiated, less of the colitis you see from sustained CTLA-4 blockade. And again, that was the goal being selective. So our goal was to take this molecule with this interesting activity and chase a signal into the white space. Where can we bring a checkpoint inhibitor or there's still opportunity to not be seventh, eighth, tenth in class of the PD-1 inhibitor, we saw 2 durable responses in prostate cancer, a tumor cold for PD-1 inhibition, but 2 out of 4 in our Phase I dose escalation that were durable PRs in late-line prostate for this molecule. And so we're starting -- we started 2 studies, Phase IIs, that we reported initial data on from the first one a few months ago. It's a Phase II in combination with standard of care chemotherapy in late-line CRPC. We saw multiple PSA50 drops, 3 out of 8, and we saw a durable PR out of those first 8 patients. We did see extra talks from the high-dose chemo we use. We dialed down the chemo, and we're enrolling again and hope to report more data on that study soon. And then the second study that, again, cast our net widely, monotherapy in a high-risk subpopulation of prostate cancer searching for that population that we can advance quickly. That's ongoing in another study, that's also enrolling patients that also have a little checkpoint therapy opportunity, a lot of unmet need in gynecologic tumors, where we also saw good responses in our Phase I. So that's the vudalimab strategy. And now I'll wrap up with our newest molecules, just entering the clinic, both Phase I started dosing patients second half of last year. They both rely on using sort of these tools we had to take the next step in T cell engagement. To the left is sort of the first wave of T cell engagers that have come about in the last decade that sort of stepped over the early scientific experiments from early in our industry's history in the molecules that could be durable action, easy to make. You're starting to see approvals now, where you have 1 binding domain to the T cell, run to the target cell and off you go. You might have a situation now where to expand from where heme malignancies have dominated bispecifics into the realm of solid tumors that we want to get to, to really meet these unmet needs in these poorly treated populations, you need to go after the targets that solid tumors expressed, which are unfortunately almost always expressed on healthy tissue, too, and in contrast to heme tumors. So how do you separate that out? The same way we did with vudalimab, you use this toolkit we've got to build a molecule that's just not going to stick well and you can see to the right to cells that don't express a lot of the antigen like normal cells and those crazy tumor cells that have a lot of the antigen, you design it so that it will stick to those better and then bring that killing power in that selective way. So XmAb819 follows that paradigm. You can see that structure on the left, targets a molecule called ENPP3, a nuclease that's on the surface of kidney cells and is very highly expressed on clear cell renal cell carcinoma, about 85% of the renal cell carcinoma population. Now that's a tumor that's treated well in frontline with checkpoint inhibitors and tyrosine kinase inhibitors, but after that, there's really not a lot other than salvage tyrosine kinase inhibitors. And the CR rate, complete response rate, is only 12% -- everybody does relapse. Huge unmet need for new modalities, and we're stepping into that breach with the molecule with high selectivity for tumor and excellent killing power. We started dosing patients in July. A little later this year as we step up our dose escalation cohorts, we'll guide on data timing for that. And I'll note that our partner, Amgen, is using a molecule we design with exactly the same structural approach to target prostate cancer with a novel antigen called STEAP1. And just last year, they reported really positive data from a Phase I study in prostate cancer for that, where this format was able to, in a way that was well tolerated or reasonably tolerated, knock down the key biomarker for prostate cancer in over 30% of the patient is better than 90%. So they're off to the races with that one. So again, these selective approaches of the next wave of how we're going to get better drugs coming out of our bispecific tools now that we've gotten that first wave done. And I'll wrap with a new mechanism that we're trying to exploit that's enabled by these tools. And I'll quickly say on the left is how nature has T cells get activated. So on the T cell on the left, APC on the right of that, the T cell receptor binds something it's recognizing that it wants to recognized later and kill another cells, and that's called Signal 1, and it turns on the T cell. But that kind of stops and [ peels that ] out unless you have the bottom half Signal 2 by CD28 being bound. Now Signal 2, if you just turn it on by itself, well, that causes an immunocytokine storm, and it's really bad for patients. And that was seen about 15 years ago. These new bispecific tools let us control that. So we only turn on that CD28 in the presence of tumor cells when you have both binding events happen. And that selective control offers the ability to boost up T cell responses in tumors that are normally cold. And there's been a lot of buzz about this from some large companies that have released very early data showing the proof of concept. There's only 3 companies in the world in the clinic now, and we're one of those, we started dosing patients last quarter. And so the molecule we have that's exploiting this is a B7-H3 x CD28, potentially addressing a broad range of tumors in clinical tumors like prostate cancer and small cell lung cancer. The preclinical data was highly selective. And this is a wholly owned asset we hope to advance in our escalation over the next year and get to the place where we can report some data on it soon. I will say that we have a collaboration with Janssen for different CD28 bispecific molecules. It's preclinical, but that we hope to have in the clinic over the next year. So a lot of activity there and a lot of potential for our both Phase IIs and Phase Is to take that next step while we maintain the business with the tools and technologies that we use with our partners. So with that, I will just leave this chart of a lot of activities for next year. Happy to answer questions on them of next year -- this year. This is 2023. And we'll be touching base at lots of conferences throughout the year. Thank you.

Great. Thank you. We'll now switch over to a Q&A session. [Operator Instructions] Bassil, welcome. Thanks for joining us today at the conference.

Thank you, Brian.

Maybe just high-level question to just kick off the conversation. I think one question -- that one issue that I have is that I realized that I do have a lot of questions to go through and we only have 20 minutes to go and you have a lot of answers to go for. But I guess one high-level question is that you have been pretty successfully in performing partnerships. And now that you do have royalty streams coming in and your pipeline portfolio is also evolving -- advancing rapidly. So how should we think about just the strategic mix step here? And let's say, 2, 3 years from now, where we look at the portfolio again, how do you think it would look like?

I hope it's going to look like an asset that we've determined has good proof of concept and a clear registrational path that's wholly owned by Xencor is in late-stage studies or completing them. And meanwhile, other molecules we've got in our pipeline now, we've had clear answers go, no go and are either chugging along behind or partnered if that's the best home for them. And we're good at partnering things when we realize there's a better place for them or we've terminated them because I can assure you there will be multiple new programs advancing because there's all sorts of new science we're continuing to do. And I expect to also have additional new partner programs that we've enabled with -- through licensing or partnering of molecules we built. So the goal, though, is to have our own molecules we're going to take to approval, and hopefully, one day launch ourselves, but be very stringent about which ones we do advance, not just because it's the most advanced program in our pipeline, are we going to take something forward, we want something that when we get to the end, it was worth it. We don't get stuck with something that we've convinced ourselves as good, but it's fourth in class, and is a bigger negative than a positive. So we're trying to use this advantage of this broad pipeline and the capital to do things smart. And we know that that's a tough strategy, so we've got stay on our toes.

Okay. And then maybe let's first touch on your cytokine program because I think usually conversations start with plamotamab and vudalimab. Let's just start with your cytokine program. So maybe hit IL-2 and IL15 at the same time. So you saw the Treg expansion profile. And we also saw the data coming from competitor as well in a similar space in atopic derm specifically. What are you shooting for in terms of the Treg expansion profile that should look like for IL-2 is the atopic derm plus psoriasis. And then for IL15 is multiple myeloma. I assume that fundamentally, you would want to have a very different profile in how you're seeing the -- in expansion...

Right, right. So I'll start with the IL15. And that's in a broad solid tumor basket study with a checkpoint inhibitor, Tecentriq or tazolizumab with Genentech, our partner, that's their checkpoint inhibitor as well as in 2 myeloma studies. And IL15 acts through the IL-2 pathway. It's sort of IL-2's cousin and it binds the beta-gamma receptors. IL15 as nature's can't bind the IL-2 alpha receptor right? I just -- it does stick to it. So for that one, I think we want to have high sustained levels of activated NK and T cell -- cytotoxic T cells to be the sort of feedstock for whether it's a checkpoint inhibitor or a cytotoxic antibody that you're pairing it with. So there's 2 cytotoxic antibodies in myeloma, Genentech and the collaboration, a CD3 bispecific called cevostamab, and they're actually interestingly using a competitor's program tidutamab and IgG antibody, an NK antibody that relies on NK cells for much of what it does. So I think for those populations, you just want a lot of cells that are in the activated phenotype, NK, CD16 positive T cells that have a lot of granzyme be expressed. For the solid tumor as well, you want the T cells for the atezo to work with. And I think for that one, the belief is more is better. There's been no great experiments, historical IL-2s and some -- few IL-15s we've had in oncology have not correlated -- have not really provided enough data on how many cells they have in the blood and how activated they were with response. There's been gaps in data, so we don't really know. And our belief is that we have really unprecedented expansion of these populations in an activated form, we'll see what efficacy that provides. So we're kind of blazing the trail to correlate the science with the clinical response. In the Treg IL-2, it's actually a little bit better. There were early academic experiments done with IL-2 at low doses, which tends to tickle the Tregs a little bit and leave the killer cells learn and even twofold increases in both Tregs, which would correlate to maybe tenfold increases or fivefold increases in the CD25 bright super immunosuppressive Tregs, that was correlated with anecdotal improvements in diseases like RA and psoriasis and lupus. And then this competitor data that emerges in atopic dermatitis, they showed, call it, they used absolute scale of cells, but roughly threefold increase in bulk Tregs and maybe seven to tenfold increase in CD25 brights showed really a promising response in atopic dermatitis and psoriasis, small numbers, but pretty clear. So I think that's the kind of minimum bar we're shooting for in those indications, and we think we can go higher and maybe do better.

So maybe just on -- one more on 306. Your partnership with -- this is a partnered program of Roche. It seems like you're doing 3 different combinations against multiple myeloma...

Two in myeloma, one in solid tumor.

2 -- you're right. So why are those combos makes sense? If you have the bet, is there a favorite child that you want to pick?

So the 2 combos make sense because IL15 and in particular, XmAb306, are really good at boosting NK cell counts in activation state and are good at activating T cell activation state, right? So daratumumab works with NK cells with its Fc domain and cevostamab is a CD3, which works with T cells, so 2 different hypotheses, 2 different cell populations. If I had to pick one that I'm really excited about, I want to say it's the daratumumab combo because we have really, really remarkable NK cell expansion and activation. And that has been anecdotally correlated with limited dartumumab response in some patients. So it seems to match up a solution with a problem as opposed to being really more of an exploratory approach based on first principles like with T cells, where if you have more active T cells, sure T cell therapy should work better. With dara, it's like, "Oh, here's a problem. Let's hit it." So I'm excited by that one.

Okay. Well -- and then maybe just on your new program that you announced this week. So 541 again -- why this specific program? And how do you get to this target? And why does it make sense to roll it out now in the preclinical setting?

Yes. So XmAb541 is our newest CD3 bispecific. And in this solid tumor strategy we have, we're using our CD3s and CD28 to try to go after targets that each of those sort of killing mechanisms complements. With CD3 bispecifics, you want target, just like I showed with our XmAb819 ENPP3 that have really high expression on tumor cells and lower expression on healthy cells. Because we think we have the tools now with these highly selective bispecifics to steer the killing power, the binding and the T cell killing to those high expressing cells. Claudin-6 has that property, and it's in a tumor type with a high unmet need ovarian cancer. So that was how we selected Claudin-6 out of the universe of possible targets. Also, it's a tough target to go after because it's Claudin-6, right? We all know that there's a Claudin-18.2 out there that other people are developing therapeutics [ and against ], including our partner, Astellas, with a CD3 bispecific we made for them. So just a lot of Claudin. They're all called Claudin because they're all kind of a like. So making antibodies that are selective for different Claudin is challenging and putting it in a context of a bispecific is even more so. So that's right up our. So we want to go into things where we have an edge, and that's part of the reason.

And for your ENPP program, just any thoughts on what you want to see in RCC?

Yes. So in RCC, when you've passed through your frontline checkpoint TKI combo and then your sort of salvage TKI, palliative chemo is really sort of 10 to mid-teens response rates and low -- very low durability. So 20% response rate in this Phase I setting with the highly pretreated patients you're getting would be something enough to make us want to expand and really go forward.

Okay. Maybe going back to vudalimab. So can you talk about the -- any direction on your thoughts on combination work in the target indications that you're looking on? And maybe just one more on how -- how differentiated is it from comparing to just the 2 agents approach targeting the same thing?

So it's very different from the 2 agents approach because we really do have conditional blockade of CTLA-4 only when PD-1 is on the same cell and the antibody can engage both. So you really don't hit CTLA-4 broadly. And the hope is that, that translates into the ability to block CTLA-4 in the right cells, but for a long period of time. Combination therapy with CTLA-4 blockade has come down to having low-dose CTLA-4 for only 4 cycles, so 4 doses to limit the damage, but hopefully still prime up the immune system and give that PD-1 inhibitor you're paired with a better chance. So they do seem to work quite differently, and we think we had a different tolerability profile in our Phase I study to help demonstrate that while we still had activity in these hard-to-treat patients. So that's kind of the differentiation piece. And the hope is that, that tolerability profile lets us combine with usually the -- also often harsh combination agents. So in prostate cancer, mCRPC after you failed antigen deprivation, after antigen deprivation has failed you, the patient, right, you get a docetaxel standard of care. Checkpoint inhibitors have been tried. Recently, pembro in a big Phase III on top of docetaxel and failed to show superiority against docetaxel alone. So still wide open there. And after that, you get a salvage with a different taxane, cabazitaxel. So we decided to go into that last line, a very high unmet need. And combined with the sort of standard of care chemo, which in really aggressive prostate cancer is that taxane plus a platinum and then the others are just the taxane or if you're PARP [indiscernible] olaparib. So we set up our study to go really aggressively combined with the really high-dose platinum taxane doubled across the board to try to maximize efficacy, but the chemo seem to be too rough to handle, so we dialed it down and we're enrolling more patients. So I think just because your checkpoint may be more tolerable in the monotherapy, the chemo you're combining with might be nasty, so really be thoughtful. We learned that less and it slowed us down a bit. But I think there's also now opportunity to move earlier up to where the PD-1 inhibition alone didn't seem to do the job with docetaxel. So there's a lot of opportunity. We have to establish that our tolerability and our efficacy meets up with what we would have expected from our Phase I responses and durability.

So do you have a good sense of the optimal regimen for chemo?

Not yet. Not yet, but I think this study with a combination of some patients getting a platinum ataxane because their diseases have severe merits at some patients just getting ataxane is going to give us a really good read on that.

So maybe I just want to touch on plamo as well. So you talked about how you -- the trial for TAVI is now stopped and what is the next potential step for the program because we saw data at ASH last December, it does seem that the data is pretty much the same across some of the bispecific out there. So what is the next strategy between you and your partner changing?

It's really to follow the strategy that Janssen and we agreed to at the beginning of the collaboration 13 months ago in December 2021, but we had really wanted to keep pushing forward something we had just started right before we signed a deal with them, which was this really exciting combination of our CD20, CD3 with a CD19 antibody that have double target hitting in DLBCL. And so that study was actually just outside the collaboration. The collaboration's goal is what is driving the program forward now and what we're all fully intent on, which is advanced plamo, get the safe dose, switched to subcu, which we started last quarter, advanced that, while we build CD28 bispecifics against B cells. I think Janssen -- and I think they've said this publicly, and they've said it with us, I think that there's a very compelling opportunity with CD28 bispecifics to potentially really move the needle with T cell therapies like CD3s both on efficacy and maintain tolerability because of the specificity designs. And so they really wanted to use that combination, which is unique to this collaboration with Xencor to try to build a regimen that's chemo-free and leapfrog the competitors that have gotten against us with their initial monotherapy regimens, but it's all going to shift to their combos, which are dominated by chemo combos. So I think that's the leapfrog strategy that we have. And in the meantime, we've been moving forward with [indiscernible] combo and just the saturation of the lymphoma space with so many clinical trials from late phase bispecifics and CAR-Ts made it extremely challenging to feasibly in a reasonable time frame, complete that TAFA plamo study, and we had to focus.

Okay. And when could we get some update on the next step of the CD28 combo?

So the next step would be getting a molecule into formal development at Janssen. And we hope to have some guidance on that this year and the research collaboration has been going very well. We built upon the work we had done with them with creating a prostate cancer specific CD28 bispecific, where we had candidate selection a little while ago. It's been in formal development. Knock wood, they get into the clinic real soon. They speak for their own time lines, but we're excited that it's approaching. We, of course, have the first Xencor CD28 in the clinic, XmAb808. We started dosing last quarter, and that's the B7-H3 I mentioned.

Okay. And then maybe just lastly on just how you think about resource allocation and cash runway as well. Maybe you can just touch on those.

Yes. So starting the year with 610, we really are fortunate to have these multiple revenue streams that come in because any one revenue stream can be very bumpy, but having enough of them stacked up, kind of gives you some consistency over a 2-, 3-year period. We had a lot of revenue off of our cetrovumab royalty. We created the half-life extended Fc for that wonderful COVID antibody that even helped some of our employees over the course of the pandemic. That's really drying up. The virus moved away, right? So that's -- it's a different sequence space, and that's that. We have a really nice, steady, we think, a consistent ramp of our Ultomiris royalty. And we think Monjuvi is going to be pretty flat. And in the meantime, we've got a variety of milestones coming in our operating models where we predict cash through the end of 2025 with our pretty aggressive clinical development portfolio, we really assume very modest milestone numbers and no new deals. That we, of course, are always actively in discussions to find ways to best leverage our pipeline and our assets. And so what that means is we're at a place where we have a great team and enough resources to drive this pretty broad early phase pipeline, but that's about where we're at, right? We're not going to be able to expand it more. So we've got to be disciplined as we want to advance more new things to pick a winner and scale the company or find more room by canceling programs.

I guess one -- maybe just one last question. It seems we have about 2 minutes left. So one question that I have is that what is the threshold that you're willing to tolerate in terms of -- this is good data, let's keep pushing to the next stage. And -- because last year, you did discontinued 2 programs, and you're focusing on new programs and you're pushing those forward. So how do you think about just -- how do you vision -- envision your tolerance and push forward?

We think that it has to be based on the idea that there could be a really valuable program either in our hands or with a partner. And if it's just something that could make it, but it's not going to be really valuable we'd be foolish to spend our money. We don't have to have guarantees, but we have to have a clear hypothesis for what we could do. For example, vudalimab, it could be one of the first checkpoint in prostate cancer if we can play the hypothesis out properly. It might not, but it could be, right? It has that chance. With plamo, it could certainly be a drug with a wonderful combination with CD28 bispecifics. But on the other hand, we needed a partner with the scale to get us there because just once we start those trials, it's going to be challenging and having a really leading hematology franchise like Janssen working on it, I think just going to make a huge difference. We got to can't hit ourselves. So it has to have a real shot. The programs we canceled, when we compare them to sort of benchmark sort of first generation, we designed our molecules to be better than or in the case that was for our LAG-3 x CTLA-4, for example, we didn't see that we had data suggesting we were better. And that's just not going to be a good use of capital. And we have enough programs. We're not forced to have something going just to say we have something going, right, which doesn't...

And you have a good royalty stream coming in as well.

We do. We do, but boy, we don't want to federate it away.

Okay. Well, thank you so much for your time today.

Thank you for the invitation.

Thanks everyone for joining us. Thank you.