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

Welcome to the Jefferies Virtual Health Care Conference. My name is Erin Primdahl, and it is my great pleasure to introduce Bassil Dahiyat, President and CEO of Xencor. If time allows, we will take a few questions from the audience at the end of the session. [Operator Instructions] I will now pass it over to Bassil.

Thanks, Erin. It's a pleasure to be attending the Jefferies Virtual Conference this year. I'm looking forward to telling everybody a little bit about Xencor. This is our forward-looking statement slide. So Xencor is a company that does antibody engineering. We specifically focus on trying to engineer the structure and immune functions of antibodies to make them better drugs. We do this by focusing on engineering the bottom half of antibodies that are colored here in blue, which is called the Fc domain. We've engineered a whole host of variant Fc domains where we introduced specific amino acid mutations and these Fc domains are called our XmAb Fc domains. And these allow us to manipulate the immune functions of antibodies, which are actually controlled by Fc domains interacting with immune cells through a set of receptors. It helps us enhance the half-life of antibodies. That's another function Fc domains confer on antibodies is long half-life, and we can further accentuate that. And they're also the structural core of antibodies, what makes them so stable in this dimeric Y shape structure. And we can manipulate that to create a multitude of structures with different sides and different halves or bispecific and multispecific antibodies, all by focusing on the Fc domain. So we've used this approach and built a large intellectual property portfolio, patents around these novel compositions of matter of these Fc domains and we've used that to build out a large pipeline of molecules. We focused recently on our bispecific technology and produced antibody and cytokine bispecific structures, 9 of which are currently in Phase I studies at both Xencor and at our partners. And so we have an internal pipeline we've been focusing a lot of work on that is focused on oncology using bispecifics. We also have a number of partners, and this partnership model is enabled by the plug-and-play nature of this Fc technology. We don't have to redesign an Fc domain to create a new drug. We simply populate it with the right antibody engineering of different variable domains or the right cytokine that we've protein engineered, and there's a new drug. And so those are some of our partners shown at the bottom. I'll note there's 2 marketed drugs that were created with Xencor technology, Ultomiris, which is Alexion's second-generation C5 inhibitor, which is a twice -- every other month dosing, enabled by our XmAb Xtend technology and Monjuvi, which is just approved for second-line diffuse large B-cell lymphoma CD19 antibody that Xencor initiated the development of and engineered with our high cytotoxicity Fc. So that's a snapshot of the company. We have an internal pipeline and partnered all built off of this Fc technology platform. So what does the Fc domain approach allow you to do? It just really opens up a new axis of differentiation for antibodies whereas standard antibody discovery, you engineer the tips of the antibody to bind to a new antigen. So you make an antibody to bind CD20, and that can create a drug like Rituxan, and it does wonderful things. You can build bispecific antibodies that take the tips of different antibodies and melt them together. But what an Fc domain does is it allows you to create entirely new features that add-on to or further differentiate whatever that antibody happens to be sticking to you. So for example, Soliris is a C5 antibody and with our Xtend Fc domain, we can make it instead of an every other week dose drug and every other month dose drug, a huge advantage for chronic autoimmune disease patients. Similarly for bispecific antibodies, we can use our bispecific Fc domain and these 2 purple shades on each side of this Fc are a little bit different, representing the slight differences in their amino acid sequence, that we can use to now bring together 2 different sides, 2 different nonidentical sides of an antibody structure and create bispecifics that retain all of the beneficial properties of a monoclonal antibody, their long half-life and stability, et cetera. So these are the 4 Fc domains that we've created over the years that are the primary drivers of our pipeline and our partnering. There's a lot of different engineering that's got -- that's taken place at Xencor around Fc domains. These are the 4 that we've used most. Each one was based on a lot of structural design work and hundreds of different variants that were tested, not just for the desired biological property, but for good biophysical properties and manufacturability. In all cases, we wanted to be able to plug into standard antibody discovery approaches. So we have our long half-life extend Fc domain and the approach there was to enhance interaction that's already there naturally to a receptor called FcRn that gives you long half-life and accentuate it. But do that selectively. So 2 amino acid changes allowed us to do that. That retains 99.5% of the structure of the antibody, that is of the 225 amino acids in the Fc only to have changed. So we're very close to native because you don't want to break what's already great, right? Antibodies already good drugs. And that same Xtend Fc domain is in Ultomiris, is in a number of partnered programs as well. The same approach for our high cytotoxicity Fc domain in the drug Monjuvi, our immune inhibitor Fc domain, which down regulates immune function instead of enhancing it like our cytotoxic Fc domain. And then our bispecific Fc, all the same structural approach, the same parsimonious design to give us very native-like structures. So this is a snapshot of our internal pipeline. And I'm going to highlight a few of the programs on it today. The development pipeline that we're running internally, and these are programs that can be partnered, like our Genentech collaboration. But these are where we're putting significant dollars, time and expertise into their further development. And that's in contrast to our technology licenses. I'll comment on what the Fc domain is as the basis for most of these agents, there's a lot of purple here. That's the color that represents our bispecific Fc domain. This is -- this bispecific technology toolkit that's emerged over the last few years at Xencor and a few other companies, has really opened the door on a lot of new biology we can use to attack different diseases, in particular, in oncology. And so we've got a focus now using these bispecific Fc domains in oncology. You can see here, there are 7 programs in the clinic at Xencor. Two of them are partnered. One with Novartis, outside the United States for our AML program, vibecotamab; and one with Genentech worth of 55-45 profit and loss worldwide for our IL-15 molecule. It's a site engineered cytokine for enhancing T cell function and enhancing T cell targeted therapies in cancer broadly, that's in Phase I with Genentech. So a lot of work going on here. I'll dive into a few of the programs, but the strategy is to use this highly differentiated approach of making bispecific antibodies to create new drugs. So this is our technology partner pipeline. Most -- not all, but most of these molecules Xencor never touched. We'd created the Fc domains, a partner identified a need or we contacted a potential partner and said, hey, this can make a new drug for you or can make a current drug better, like Alexion's Ultomiris, they take a license to the Fc domain technology, and they use it. It's very portable, right? We've already made -- done the engineering work. And so this is a very efficient resource and cost-wise approach to partnering. We don't spend time and effort in our precious FTEs building drugs for other people usually, right? We enable them. And you can see a number of these programs across different therapeutic areas. Inflammatory disease for Ultomiris, you've got allergy, our partnership with Aimmune, which is now Nestlé, acquired them recently. Our infectious disease work with Gilead, VIR and the NIH, in HIV and other viral diseases, additional oncology partnerships. So there's a range of opportunities here that we can tap that are outside the scope of what Xencor can do with the focus that we need as a clinical stage biotech company that's trying to move a product to registrational studies and hopefully OneData market. So now I'll dive into our bispecific technology and programs a little bit. The basis of these programs is this Fc engineering work we've done where we've created Fc demands and we generally use just one. It suits many purposes, where the goal was to have natural antibody properties. The first wave of bispecifics that came around about 15 years ago -- 10, 15 years ago, didn't look much like antibodies because the way you could stitch together 2 different antibody binding domains was to literally just link them with a peptide linker. So you lose this Fc structure, but that's what gives the antibodies their long half-life. You go from weeks to a short number of days or even hours without it. It gives them their high stability, their ease of purification and their simplicity and manufacture there. Their portable manufacturing that's so easily done nowadays with all the capacity that we have. So what we did was we said, well, look, how do we replicate this but have an antibody with 2 different sides that we can decorate with whatever we want. We needed to build an Fc domain that let us, first of all, express both halfs of it from different genes, different constructs in the same manufacturing subline but have it spontaneously assemble. And then have it be highly stable, but also to be able to readily separate it from the commission of things that occur whenever you do production of any kind of molecule in a cellular system. And so we built all that and that ended up giving us now a readymade scaffold that we could apply to almost any different combination of binding domains or cytokines that gives us antibody like half-life in vivo was easily produced and purified using standard antibody manufacturing processes. We don't have to reinvent the wheel here. We just plug into what our CROs can already do, and it's very stable. And so this is the paradigm we have that's led to us expanding and having such a broad partnered internal pipeline so rapidly, you can literally plug-and-play on the standard Fc format binding domains from different donor antibodies. Our CD3 bispecific molecules, we have a fixed CD3 on one side. You just pop on different ones. Our dual checkpoints and co-stims built off the same scaffold and our engineered cytokine. So it truly is a plug-and-play toolkit. We think of them as legos. So there's 3 specific applications we focused on as we try to build out our clinical pipeline off of this platform within oncology and specifically immuno-oncology. We have our T-cell engagers that bind directly to T cells on one side using binding domain to the antigen CD3 and bring that cytotoxic killing to some tumor cell. We have our dual checkpoint or checkpoint co-stim molecules that strive to have a selective activation of target cells that have checkpoints for pressing -- sorry, T cells that have checkpoints for pressing their function and particularly focusing on the cells that have multiple checkpoints, which tend to be the ones that are most prevalent in the tumor microenvironment. And we have several programs there. And then our cytokines. The first one got in the clinic in March of this year, our IL-15 for stimulation of NK and T cells in our collaboration with Genentech, XmAb24306. And again, you can see we built structures based on this bispecific Fc domain that allow us to have both targeted and nontargeted, highly stable, long-acting molecules. So the CD3, and I'll touch on our plamotamab and Tidutamab programs today, given the time we have, they all share the same structural scaffold, the same CD3 binding domain and different antigen affinities. And you can see they're all in, in vitro model, our in vivo models, long-acting, quite potent and tolerable. You can give these as a bolus injection and have them tolerated and look for anti -- and we see potent antitumor activity preclinically. All 3 have had -- have shown a promising activity in Phase I studies, and we're currently starting a next round of studies for plamotamab and Tidutamab, and hopefully vibecotamab to follow shortly thereafter. So I'll touch on our CD20 x CD3 plamotamab. It's our B-cell malignancy targeting agent. It targets CD20, a very well-validated target for B-cell malignancy on one side, a universal -- or I should say, B-cell restricted and on all B-cell lineages or rather all B cells in lymphoma and CLL target; and then CD3 on the other side. It redirects the killing of T cell against this target cell when you have both ends of it bound, and that's an important point. We got good pharmacokinetics. And we've turned off all of the natural immune functions, the NK cell and macrophage recruitment of the Fc domain to keep this a very clean mechanism and to reduce potential cross-linking of CD3 and potential side effects. So this is some data we presented last year at the American Society Hematology meeting. During dose escalation, we started seeing activity at 80 micrograms/kilogram due to the very potent molecules. And very promising responses that seemed potentially dose-dependent. I believe this was about a 27% complete response rate and an overall response rate of just under 40% in heavily pretreated, relapsed and refractory diffuse large B-cell lymphoma patients. And with this promising activity as a baseline, we've continued dose escalating. We're currently wrapping that up right now, and we expect to have our dose and regimen fixed going into later-stage studies in the new year. Time on treatment, it's still fairly early at that data cut, but we have a number of patients out past 6 months and number ongoing. And we also looked at how the toxicity of this agent might play relative to earlier generation bispecifics. And we saw the promise of having a potency that is still very, very potent relative to traditional antibodies like Rituxan, but much less potent than that first wave of nonnative structured bispecific antibodies. And so the adverse event that was most commonly observed was, of course, cytokine release syndrome, which is universal for these CD3 bispecifics. And we saw CRS for majority of patients at least one event, but the vast majority of those events were Grade I or Grade II. And there was really not much in the way of CNS involvement from our CRS, which is a potential problem that's been happening with CAR T therapies and high potency -- really high potency bispecifics. So it was a promising toxicity profile, very well tolerated by the patients in this regimen. And you could see here an important observation, the CRS by dose day. So the first dose of the patient got the second, et cetera, you can see that both the severity grade and the frequency plummets upon repeat dosing. And that's because we believe the immune system is getting conditioned to seeing the drug. And so we've used that to design dosing regimens that start off a little lower on day 1 and then escalate to mitigate further the CRS. And the data I just showed didn't have all that baked in, we were learning as we went. And the regimen -- the go-forward regimens, we -- we're optimistic should be tolerable and have quite manageable CRS. So in summary, coming out of that data, we had a well-tolerated agent with very encouraging signs of clinical activity and very difficult to treat patients. In addition to the DLBCL, we had responses -- complete responses in Waldenström macroglobulinemia and richer transformation as well as in follicular lymphoma, the vast majority of our patients were in diffuse large B-cell lymphoma. Our plans going forward are informed by a very recently announced global collaboration with MorphoSys and Insight, where we're combining plamotamab with the MorphoSys and Insight agent Monjuvi, which we, of course, at Xencor had created and licensed to MorphoSys about a decade ago. So it's exciting to again work with the molecule that we built here and engineered for high cytotoxicity. And we believe that in the context of lymphoma therapy, because there are multiple mechanisms of action that can treat these diseases, chemo, targeted agents, various antibodies, combination therapies are where patients are going to get the best results and where the field is going to go. With this study that we're looking at doing a Phase I/II study combining the anti-CD19 tafacitamab with our anti-CD20, the NK and macrophage-mediated killing from tafacitamab with our T cell-mediated killing, it's a terrific confluence of orthogonal mechanisms that we hope can complement each other. We do expect to start these studies next year. So now I'll move to Tidutamab. This is our first solid tumor targeting bispecific -- CD3 bispecific. It targets somatostatin receptor 2, or SSTR2 which is a target highly expressed in neuroendocrine tumors as well as in other tumors like Merkel cell carcinoma and small cell lung cancer. These are all tumors that arise out of various neuroendocrine cell lineages. An ongoing Phase I study right now with -- in patients with net or neuroendocrine tumors, and these are the classic mid-gut net typically affecting the pancreas, various GI tissues as well as in GIST, a tumor type that also expresses. This was our dose escalation study where we started off with a dose on day 1 and then a step-up to a higher dose on day 7 to match this sort of looked to replicate this reduction in cytokine release syndrome that seems to be universal for CD3 that happened on repeat dose. And about a month ago, we reported initial data from this dose escalation and some expansion cohort patients in net for 27 patients at the North American Neuroendocrine Tumor Society or NENETS meeting. They were heavily pretreated patients with various lesions. And interestingly, 56% had received SSTR2 targeted radionuclide therapy or radioactive isotope therapy. So again, a very difficult patient population. This is the -- the drug was generally well tolerated. And interestingly, we did reach a dose-limiting toxicity at a starting dose of 1 microgram/kg going to 2 mcg/kg on week 2, that was actually nausea and vomiting. So GI toxicity, which makes sense because SSTR2 is expressing gut tissue and it's quite likely this is an on-target phenomenon. Cytokine release syndrome was observed and at our maximum tolerated dose and the recommended expansion dose was very mild to moderate, only Grade I and II. And in general, the CRS was mild to moderate. We did see some nausea vomiting, but it was quite manageable. So with this profile, we're very encouraged that we have a solid tumor targeting agent that engages CD3 on T cells and can be tolerated. And how is it doing engaging the immune system? It's doing a great job engaging the immune system, it turns out, we get both initial and sustained activation of CD8-positive T cells or your effector T cells, the killer T cells. So you see a dose-dependent and these colors show you the dose step-ups in blue as our recommended expansion dose. You can see here after a month or 2 months, you still see an expansion of these proliferating T cells at our go-forward dose as well as the PD-1 positive cells, which is a sign of activation at our go-forward dose. So they come up early and they stay up. So that's great. We did not see any objective responses, not surprising in a neuroendocrine tumor population where they tend to not show objective tumor shrinkage by RECIST criteria. We are -- we do have a number of patients on follow-up with stable disease and we'll continue that follow-up and recruit more and follow them for -- follow them throughout the course of -- the rest of the trial. Because of the activity agent, it's good tolerability, the signs of biomarker function, we are moving into 2 additional indications, Merkel cell carcinoma and small cell lung cancer. We do expect to start that in the first half of next year, phase Ib studies with high SSTR2 expressing tumors in agents that are no -- in diseases that are known to be responsive to immune therapy. So we should be able to have a more rapid look at the activity of the agent while we still follow the net patients to see how they do, again, where for them, it's duration on treatment that's the real marker of clinical benefit. So with that program, I will switch gears quickly to our tumor microenvironment assets, which we just reported the first data for recently. In this case, we're looking at rather than recruiting at T cell directly to kill a target cell, we're just targeting the T cell now and trying to get at 2 checkpoint markers. And in particular, we've engineered the molecule to have modest binding to each of the individual antigens. But when you have both antigens present on the cell, both sides can hook on and you get highly avid binding. That, we believe, based on a very large literature should bias us towards the T cells that are in the tumor microenvironment, which tend to have multiple checkpoints up regulated these tumor infiltrating lymphocytes or TILS. So those are the cells you want to get to. So we design this agent to have that selectivity to hopefully open up a tolerability profile as well as engage just the right T cells. So we have 3 agents in the clinic now. I'm going to focus on XmAb20717 now. It's sort of the prototype of the class and it engages the 2 best validated checkpoint targets where T cell repressing targets that we want to block and let the T cell wake up, PD-1 and CTLA-4. In particular, CTLA-4 has had challenges that are exacerbated with toxicity when it's used in conjunction with PD-1, characteristic immune-related -- immune activation related adverse events like severe colitis and pneumonitis. And we were trying to create an agent that had the potential to maybe be -- open up a tolerability window to enhance our ability to dose PD-1 and CTLA-4 inhibition long-term. And we just reported the Phase I data for this program, XmAb20717. We reported very initial dose escalation data at ASCO in June and a broader data set from our expansion cohorts at SITC just a couple of weeks ago. So the Phase I study was designed with dosing every other week, in a range of tumor patients. We had dose escalation starting at 0.15 mg/kg going to a preplanned top dose of 10 mg/kg, which was tolerated, and we extended that to 15 mg/kg with potential to go to 20 mg/kg if we pass 15 mg/kg. And meanwhile, we expanded a 10 mg/kg in preplanned expansion cohorts in melanoma, non-small cell and renal cell carcinoma. These are indications where there's approved PD-1 antibodies. And therefore, we expect the great majority of patients in those expansion cohorts to have prior immune checkpoint therapy and have failed on that therapy and progress to us. And 2 cohorts that are patients that have -- are not in indications with prior checkpoint approvals. Those are castration-resistant prostate cancer and a basket of different tumor types. We have plans for additional cohorts in select populations. So we fully enrolled all the cohorts, except for the renal cell carcinoma cohort. So this is the adverse event profile, it was generally well tolerated. The most common AEs were immune-related, and you can see here in color code, Grade III in red. The most common Grade 3 immune-related adverse events were rash or other skin itching or laboratory elevations of liver enzymes, transaminase increases where we did have Grade IV. And then a much less common range of other adverse events, many immune-related, not all. So we did have 2 Grade Vs that were in a patient with pancreatic metastases that had pancreatitis in presence of our drug. And we had a patient with respiratory failure and myocarditis where they had a long history of cardiac events, including atrial fibrillation, a dual lead pacemaker and they had a lung tumor that was adjacent to the cardiac tissue. We've kept enrolling subsequent to these events and are moving forward with the program rapidly. So here's an initial look at efficacy. I'll note that these dots here mean patients had prior checkpoint exposure. So all but one of our responders had prior checkpoint therapy. I believe our overall response rate in our efficacy evaluable population was 19% across these range of tumors, 3 in melanoma, 2 in lung and in ovarian and in renal cell. Most interesting, we did have an objective response, resist response still ongoing for a patient with prostate cancer. And we're still following up with a number of these patients. That patient is right here, unconfirmed PR to date, and note that out of a very early data cut, we have multiple patients with PSA reductions, including one with a 85% PSA reduction. So a very promising start. One in 4 patients evaluable had a PR and PSA reductions to this very early look at our prostate cohort. And we are starting a Phase Ib in multiple molecular subtypes in the first half of next year, we expect in prostate cancer with this agent to pursue this very early glimmer of a signal. Last, I'll just mention that our IL-15 program is in the clinic with Genentech, our partner, that's XmAb24306. We did a 55-45 worldwide profit and loss split about 1.5 years ago, and we're aggressively pursuing that molecule in monotherapy as well as soon we hope if the trial goes well in combination with atezolizumab, Genentech's PD-L1 inhibitor in a wide range of tumor types to see if we can enhance the activity of PD-L1 blockade. So it's a rapidly expanding pipeline, many different modalities, all in Phase I. We hope to pick the winners from this based on clinical data and aggressively pursue those and be very stringent about only advancing the programs that merit it. So I'll end with just a quick look at some of the very busy schedule we had and achievements of this year. And I will say we have a strong balance sheet, just under $600 million in cash at close of the last quarter and runway into 2024 we -- is what we're guiding. So with that, I'll wrap up. I'm not sure we have time for any questions, but thank you very much for your attention.

Thank you, Bassil. With that, we'll conclude the presentation there. Have a nice evening.