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

Hello, everyone. Thanks for joining us. My name is Jonathan Chang. I'm part of the SVB Leerink equity research team. It's my pleasure to host the management team of Xencor, and we have with us today, President and CEO, Bassil Dahiyat. And for the investors joining us, feel free to submit questions or e-mail them to me at [email protected]. And with that, let's get started. Bassil, would you please briefly introduce the company?

Sure, and thanks very much, Jonathan, for the opportunity to participate in the Leerink conference today. Xencor is a protein engineering company. We engineer antibodies and other protein molecules to optimize their functions, often immunologic functions, to make them into drugs. We've built a platform based on this protein engineering approach that's centered on a suite of engineered modules, Fc domains, that are the bottom half of an antibody, from the constant region, that we've manipulated through our protein engineering to give us properties like enhanced half-life, like greater immune function, and to create highly stable structures that are bispecific or multivalent as opposed to natural monospecific antibodies. We built a business that's both around access of our technology to partners as well as our own internal pipeline. On the partnered side, we've licensed the use of these Fc domains as well as molecules we've built around them. There's currently 2 marketed products, Ultomiris with a long half-life at Fc domain from Xencor; and Monjuvi, that was recently approved in diffuse large B-cell lymphoma by our partner MorphoSys. That's a high ADCC Fc domain we put into our CD19 antibody and then partnered with them. There's a lot of other partnerships as well. I'd say, notable amongst those are our Genentech partnership around our cytokine program IL-15 for oncology as well as our recent Janssen collaboration around our new CD28 costimulatory bispecifics. So that business side and the licensing revenues help support the primary goal of the company which is, develop our internal pipeline, which is centered around our bispecific platform and includes the CD20 x CD3 that's entering into Phase II this year in combination with -- actually with the Monjuvi product and lenalidomide, our partner MorphoSys as well as our PD-1 CTLA-4 molecule, XmAb717 where we're initiating a prostate cancer trial, and a suite of 4 other molecules currently in the clinic in oncology that share this bispecific platform. So with that little intro, I can talk about various programs and the technology base, whatever you'd like to speak about, Jonathan.

Great, thanks for the intro. First question, how is Xencor's antibody engineering platform differentiated from others in the competitive landscape?

I think that we've built the most comprehensive suite of Fc domain technology that lets us use in a plug-and-play way various features that we want. For example, our long half-life Fc domain is currently in the Ultomiris product, giving that the every other month dosing from the every other week dosing that the first generation Soliris product had. It's also the same mutations that are in our partner Vir's SARS-CoV-2 antibody, our partner Gilead's HIV antibodies. And we use that functionality in our own bispecific antibodies where longer half-life is attractive, like for example, in our PD-1 CTLA-4 molecule or in our cytokine constructs, IL-15 with Genentech, where you want duration of action. So we've got that plug-and-play portability and that mix and match capability that I think is unique in the industry. You can make a long-acting bispecific, you can make a high cytotoxicity molecule. And then we've coupled that with just the deep knowledge we have in protein engineering, and for example, manipulating cytokines, manipulating antibody binding domains to build constructs, whatever construct is needed for the purpose. And so we have different formats we use. We're not wed to any one thing. And it's all put on this very stable and portable format. And so we have multiple partners with molecules in the clinic using our tools. We don't do anything special for manufacturing or development. We use all the standard off-the-shelf tools that are abundant in the industry, our contract manufacturing is done very simply. And I think that really sets us apart, is that ability to be scaled and speeded up very rapidly across a whole broad front of tools.

Got it. Now you mentioned -- you touched on briefly a number of the partnerships that you've entered into. Can you talk about the collaboration with Genentech for novel IL-15 cytokines?

Sure. We initiated a collaboration with Genentech for, at the time, a preclinical program around the cytokine IL-15. That was 2 years ago now in February of '19. And now the lead program for that, XmAb306, an engineered IL-15, is currently in dose escalation and has been for about 11 months. So the idea for IL-15 is that it is a cytokine that amplifies and activates T cells as well as natural killer cells. In fact, that's how you often seed these cells and culture when you want to keep them happy and alive. So we identified, of course, that cytokines are an area that there's an immense amount of promise in, I think many in the industry recognize that, for amplifying the function of immunotherapies, whether it's a T-cell therapy like a PD-1 inhibitor or an NK-cell activating antibody. And so the challenge with cytokines is they're often very toxic, they're very short acting, and there's various technology approaches people have taken to solve these problems like PEGylation. The approach we took was to look at the biology of the cytokines in general, looking at IL-15 and IL-2, recognizing an IL-15 inherently has a selectivity that's favorable for what we want, in this instance, which is to drive cytotoxic T-cells because it doesn't bind the receptor CD25, that its cousin IL-2 does, but it binds exactly the same downstream receptors or rather, I should say, the same effector receptors as IL-2, the IL-2 receptors beta gamma. So you already have a clean non-CD25 binding cytokine. So that avoids the bias towards regulatory T cells you want to avoid that many have had to engineer around creating not alpha IL-2s. So we start out from that base, and we recognize that toxicity and short half-life of these cytokines is often due to, they hit their target cell, they amplify that target cell, and they create a large sink of receptor and target cell that just sucks away the cytokine. So we went the opposite direction for making a more active molecule, we dialed it down. And this is a theme we're playing across our cytokines, our IL-2 that's about to be in the clinic. We dialed down the potency about as low as you can get it before you shut off all the signaling altogether. We just do that with 1 or 2 mutations of protein engineering. And we use our Fc domain to create a stable, robust structure that can be long acting, and in fact, use our long half-life mutations along top of that. So we have a low potency, but still active molecule on an Fc that we showed preclinically has very long duration of action and good tolerability, we hope best-in-class, for T-cell and NK cell amplification. That led to the partnership with Genentech, where we hope to share in their scale of clinical trial and capability to move this program fast across a broad front if the dose escalation we're doing now is successful. So in that collaboration, we share the profit and loss 55-45; 55% for Genentech, 45% for us worldwide. We got paid $120 million upfront. We have substantial clinical milestones. The goal is for us to be hopefully minimizing our cost through approval, and then we have an option to co-promote in the United States. And then there's additional discovery stage -- or early clinical -- or early preclinical development and discovery stage programs to target these IL-15 constructs to particular cell types. So they expanded dose escalation to their combination with atezolizumab last quarter, and so we're eager to keep moving on that one.

Got it. When could we see initial data from the XmAb306 program?

We're making life hopefully easier for the world and shortening this just to the last 3 digits of our numbers. I've already found it helpful. So for 306, we have to coordinate with Genentech on any kind of disclosures, and we're currently working with them on a publication plan. We did disclose that they did start the dose escalation portion of the Phase I in combination with atezolizumab. I think as new trial activity starts that are inherently public, we'll certainly talk about those. As for data, we're going to have to guide on that later when we have more certainty.

Understood. Switching over to another cytokine program, XmAb564, your IL-2 program. Can you discuss the opportunity for this one in autoimmune diseases?

Ye's, this is using what IL-2 does well, which is binding CD25 and preferring to activate regulatory T-cells and trying to accentuate that property. So using exactly the same philosophy of dial the potency down and use the Fc domain for a robust scaffold that we used for the 306 IL-15, we did exactly the same philosophy of reduced potency here for 564, created a construct with very low potency that is tuned for bias towards binding CD25, which means you're going to preferentially engage regulatory T-cells, not exclusively, but preferentially. The goal there is to amplify regulatory T-cells and to activate them. Because there's a lot of correlative data out there showing that Treg deficiencies or dysfunction can drive autoimmune disease across a broad front. So there's the potential for Treg stimulation to be a broad clinical class across many autoimmune diseases, and maybe move the standard-of-care needle for patients. There's a bunch of activity going on, early clinical activity in a few companies, and we're jumping into that in the next few weeks. We hope that our design for lower potency and long action can give us that sustained but tolerable activation of Tregs, and then we're going to play the clinical hypothesis out, is it showing disease-modifying activity? Lot of excitement around this class. We think we have a molecule that potentially has best-in-class properties, and we'll see. One notable fact of our molecule is this, the Fc domain it's built off of is actually a bispecific or heterodimer Fc, where we have 2 different things on each side. One side has our engineered low potency IL-2 with a CD25 bias or a Treg bias. On the other side, we have nothing. It's empty. We wanted to avoid having 2 IL-2s that might overstimulate the target cell and cause the problem we're trying to solve. We're trying to solve for the problem of toxicity. We don't want to overdo things on activation.

Got it. Switching over, you recently entered into collaboration with Janssen to discover novel CD28 bispecific antibodies for prostate cancer. And your first wholly owned CD28 program targets B7-H3. So first question, what makes B7-H3 a promising target in oncology?

Well, it's great because it's bright on a lot of different tumors. It's pretty selective for being bright on tumors, not uniquely, but it's got a pretty good selectivity profile of tumor versus healthy tissue. It's bright across a whole range of tumors. So if you have an agent that can be, say, a CD28 co-stimulator, that can, hopefully, in a target and hence, tumor-specific way, amplify the action of, say, a checkpoint inhibitor or CD3, it would be wonderful to be able to use that agent across many different tumors and not be limited to making a brand-new CD28 for every target in every tumor type. CD28 is really an ideal target there. And in the places where CD28 is -- sorry, B7-H3 is brightest, it matches up to where there's significant unmet need still in immuno-oncology like in prostate cancer. And we've seen, in fact, for B7-H3s from various modalities people use, whether it's antibody-drug conjugates or whatever, prostate cancer shows promise because that's probably where there's the biggest gap or the brightest and most frequent expression of B7-H3. So we think the target has an early, but relatively well supported hypothesis, and it could be a perfect backbone for bringing co-stimulatory function to amplifying checkpoint therapy as well as CD3 therapy. So that's why we like B7-H3 a lot, and it's a perfect match with this CD28 approach, which is one of being that and switch, that logic gate that really turns on therapies in the tumor environment that's target-dependent and avoids all the peripheral toxicities that might occur, but really could drive greater activity than you would ever see otherwise. That's the promise of CD28 co-stimulation.

You mentioned that there are a bunch of other approaches and mechanisms targeting B7-H3. How should investors be thinking about which mechanism makes the most sense for this target or across or maybe specific to certain indications?

Right, I would say that you got to think about how the particular tumor type is going to -- how responsive that -- rather not responsive, but how sensitive that patient population is to the toxicities that might emerge, right? So we've seen already that the place where B7-H3 seemed to have the most promise is the one where there's the most delta in expression, the brightest expression on the tumor type, suggesting that therapeutic index is important. And for very active modalities like antibody drug conjugates, you have always therapeutic index challenges. And so prostate is maybe where those challenges were most easily met, right? You could give the drug and it really goes to the tumor and your peripheral toxicities are more manageable. I think with the CD28, though you would expect to see some activity from that agent itself, its selectivity inherently and its activity inherently is going to be, in theory, much less toxic. The whole point of this new class of CD28 like we're making, like some competitors are making is to eliminate the super agonism function of CD28 and just get to that signal 2 co-stim function on T-cells. So we hope that this is going to be an agent that's tolerable and really does just turn on other therapies as its primary mode. And so that tolerability, we think, could be really a distinction and broaden the use dramatically.

Got it. What are the time lines associated with this program?

So we have candidates that are in early development. We expect to have an IND filed next year and hopefully in the clinic shortly thereafter, and we're working very hard on that right now. But this class of CD28s, I think, is going to be one where there's going to be more and more activity, and we're excited to be on the very leading edge of that.

Got it. Now switching over to plamotamab, your CD20 x CD3 bispecific. What are your latest thoughts on the competitive landscape for that program? And how do you see plamotamab positioned in that competitively?

Yes. We think that the competitive landscape has emerged from sort of the early very small data sets that looks like we have a range of molecules that have activity, that have good tolerability and that are positioned now for how you're going to develop in B-cell malignancy, where I think everybody agrees the first most prominent unmet need is in relapsed DLBCL. But the long-term goal is to extend into earlier lines of therapy as well as across lymphoma, into follicular lymphoma and other types. People want to displace Rituxan. Right? You now have agents whose single-agent activity is really as good as that very first R-CHOP data that emerged 15-odd years ago that got really the ball rolling for Rituxan. So what do we do now with it? I think that given that you've got a landscape where you've got a lot of people that are just starting that next stage of trial, there's a lot of activity going on, people trying to -- some people trying to get to market fast and follow-up with a bunch of trials that are in combination with whatever, some of our competitors are throwing it into whatever else they've got in their own internal portfolio, the very large companies like Roche, some are trying for single-agent to go fast. We think that lymphoma is going to be about where the combination agents have the most activity. That's what's going to win the day. Because the bar is very high for activity, and people absolutely want to have an agent that is in a regimen, because it's not going to be single agent, that avoids the use of chemotherapy. The long-term toxicities, the acute tolerability issues are just something that people want to get away from like we've seen happen in CLL, in chronic leukemia, with the advent of the BTKs. So our strategy, our development strategy and how we think it's really most critically going to be positioned in the landscape is by combining with the best chemo-free regimen we could think of, that has the most activity and the best tolerability. And that regimen is, ironically enough, Monjuvi, the molecule we made 2 years ago and partnered with MorphoSys, combined with lenalidomide. And so we initiated a clinical collaboration with MorphoSys and Incyte that's going to start in the second half. A Phase II study that is going to look at relapsed/refractory DLBCL as a triplet, potentially registrational, we'll outline the exact details of that trial when we start it, as well as we have plans to follow-up shortly thereafter with a trial that goes into frontline DLBCL, as well as looks at relapsed follicular lymphoma. So we want to have a broad strategy. And we know there's a lot of competition out there. But we think that chemo-free regimen and looking for the most active combination regimen is what's going to, ultimately, win the day on lymphoma.

Got it. Switching over to tidutamab, your SSTR x CD3 bispecific. What is the latest status and development strategy for that program? And what makes SSTR an attractive target in small cell lung cancer.

Right. So latest status is we reported data in October at the North American Neuroendocrine Tumor Society meeting. The Phase I study within neuro -- is going on, I should say, still in neuroendocrine tumors. We're still running the expansion cohorts and treating patients and following them up. SSTR2 is highly expressed in neuroendocrine tumor and in other tumor types that are neuroendocrine lineage like small-cell lung cancer. So sort of prelude to the answering of that question. We showed at NANETS that we had an agent that had a number of patients in long-term stable disease. We've finished that expansion cohort enrollment, and that expansion cohort is ongoing. We expect to update that later this year. NETs are not tumors that are often resist responsive. In fact, the only most recently approved agent, which was a radioconjugate peptide, so a beta emitter, Lutetium radioactive isotope, had a complete response rate -- sorry, a response rate, almost known complete responses of 12%, and the label was based on PFS and OS extension, so a couple of years. So it's a long-term, slow-growing tumor. We saw in that study, good tolerability, mild CRS and -- but yet excellent expansion of activated T cells sustained over months and of large magnitude. And the dose-limiting toxicity was actually probably an on-target SSTR2 toxicity; nausea, vomiting, which makes sense given gastric tissue expression of some SSTR2-positive cells. So from an immunological standpoint, great activation, very mild -- mild CRS, a manageable toxicity of nausea, vomiting at our maximum tolerated dose, we're going to keep looking at NET as it goes forward. That's a long, slow burn watching those patients to see if we're prolonging their lives. And in the meantime, small-cell lung cancer, as well as skin neuroendocrine tumor, Merkel cell carcinoma have unmet needs. They have a high proportion of tumors expressing SSTR2 at good levels. And they're tumors that are responsive to therapy, whether it's chemo or immunotherapies. So we should be able to see a signal there. So starting this first half, we're going to have a study going that is going to enroll people in both small-cell lung cohorts as well as Merkel cell cohorts to see -- and that Merkel cell carcinoma study will start in relapse post-pembro, which is sort of frontline standard of care then also moving to frontline and combo, where response rates are high, but durations are very low. So we see those unmet needs in tumors that express the antigen. It's an active agent, immunologically at least. Can we see real tumor reductions and get clarity on the molecule's activity? And those are indications where we potentially move very rapidly if we have a signal.

Got it. On XmAb717, your PD-1 x CTLA-4 dual checkpoint bispecific, what's the latest status of the Phase I escalation and expansion study?

Right, so I'll reference that off of our data we presented at SITC last fall, where we had pretty mature data from 2 of our 5 expansion cohorts. Those were the melanoma and non-small cell lung patients. We showed multiple responses in each. And we had very immature data from renal cell carcinoma, a basket of sort of unusual tumors. And then the most immature data from our prostate cancer, castrate-resistant prostate cancer cohort. We plan to later this year present more mature data, fully enrolled data and mature data from those other 3 cohorts that I just mentioned. And what we saw in the expansion cohorts already, and we showed this at SITC, was promising data from that small number of patients that were early on in our prostate cancer cohort, multiple PSA reductions, a PR from 1 of the 4 RECIST-evaluable patients. As you know, in CRPC, often, you don't have tumors that are RECIST evaluable, they're either bony mets or lymph node involvement only. So that led us to planning, and we're going to be starting a little bit later this year, a Phase Ib study in prostate cancer for the agent, both monotherapy for 717 as well as in combination with small molecules like chemo or a targeted therapy, where there's a molecular subtype that suggests those chemo agents are or the small molecule is relevant. So we want to move this molecule into areas where there is an unmet need, where immunotherapy can make an impact, and it's not saturated with checkpoint inhibitors already. Prostate cancer is one of those, and if we can show an uptick on the relatively low activity that was seen in the PD-1 inhibitors, I think we've got a good direction to go. We also plan, as these other cohorts mature, to guide on additional next data development in other indications that we've selected based on the data and on the commercial opportunity. So we'll give more info on that a little bit later in the year. So that's where that one is. So we have to wrap up that Phase I expansion data a little bit later this year and go to the next phase.

And how does 717 compare to other PD-1 CTLA-4 dual checkpoint inhibitor programs involvement?

So we engineered it in a way that I think is, it's not unique. I think there's one other program that takes this approach, but it's different from most, which is most dual checkpoint inhibitors and certainly the PD-1 CTLA-4 dual checkpoint bispecifics we're aware of, have binding modes where you bind -- you have 2 antigen binding domains for the PD-1 and 2 for the CTLA-4. So you're basically trying to hit all the PD-1 and CTLA-4 on all the T cells in the body. Whether that T cell has just PD-1, you should stick to it. Whether it was just CTLA-4, you should stick to it. We went with an approach that was purposely exploiting the bispecific platform to try to heighten the selectivity to the T-cells that had both checkpoints expressed. So we made it so that it only had 1 PD-1 binding domain and 1 CTLA-4 binding domain, and we dialed the affinity, in particular, for the CTLA-4 domain into a modest range. So we are greatly preferentially binding cells that only have both those antigens. And those are the cells that are overrepresented in the tumor microenvironment. They're the ones that usually have had some kind of stimulation, some kind of TCR recognizing some epitope on the tumor, and then the checkpoints come up to sort of oppose that. And so those are the cells you'd love to get going the best, and you'd like to avoid other cells that might be driving the toxicity profile of dual checkpoint blockade that you've gotten with, say, PD-1 antibodies plus CTLA-4 antibodies. That's the hypothesis. So so far, what do we know, in about 100 patients at our 10 mg per kg dose that we did the expansion cohorts in, our tolerability profile was generally pretty good. We do see IRAEs. We don't seem to have quite the extent of more the CTLA-4 blockade characteristic AEs that cause problems like colitis and pneumonitis. Of course, as the data set grows, we'll know more. But it seems to have a distinct profile and it's -- well, it's generally well tolerated at that dose, and that's why we did the expansion. So the only other program we're aware of is one large pharma company, AstraZeneca, has a program with a similar design philosophy that's, as far as we're aware, in Phase Ib. So we hope this unique design approach gives us a distinct place in the clinical development.

Got it. And just in our remaining couple of minutes, can you discuss the 2021 catalyst and milestones that investors should be looking forward to?

Sure, so I guess, for the rest of the year, we're expecting to have data readouts, like I said, for the PD-1 CTLA-4 XmAb717 with the full expansion cohorts; an update on the neuroendocrine tumor cohorts to see how long we're going, how well we're doing for tidutamab. We do expect to have plamotamab data where we -- our plan is to wrap up all the Phase I data, our final dose of regimen and present all that monotherapy dose escalation data as we initiate our combo study as well as expansion cohorts and various indications off of that Phase I. We're starting the clinical trials for 717 in prostate, for tidutamab in Merkel cell and small cell, for plamotamab in combo as I mentioned. We're also starting a clinical trial for our IL-2 molecule, 564, the healthy volunteer dose escalation study, single ascending dose is starting soon. And depending on how that progresses, we'll guide on data timing from that and hopefully quickly. And then we'll see how our -- the various partner milestones that I guess, I don't have time to get into, but a pretty busy schedule.

Yes, that could take another while. Well, thank you very much, Bassil. That's all the time we had. Thank you very much for joining us.

Thank you so much, Jonathan. Take care.