Xencor Inc (XNCR) 2017 Q2 法說會逐字稿

Good day, ladies and gentlemen, and welcome to the Q2 2017 Xencor, Inc. Earnings Conference Call. (Operator Instructions) As a reminder, this conference call is being recorded.

I would now like to introduce your host for today's conference, Ms. Hannah Deresiewicz with Stern Investor Relations. You may begin.

Thank you, operator. Good afternoon. This is Hannah Deresiewicz with Stern Investor Relations. Welcome to Xencor's Second Quarter 2017 Financial Results Conference Call. This afternoon, we issued a press release, which outlines the topics we plan to discuss today. The release is available at www.xencor.com. Today on our call, Bassil Dahiyat, Ph. D., President and Chief Executive Officer, will discuss the company's business highlights and provide an update on Xencor's clinical programs and pipeline progress; and John Kuch, Vice President of Finance, will review the financial results from the second quarter of 2017. Then we will open the call up for your questions.

Before we begin, I would like to remind you that during the course of this conference call, Xencor management may make forward-looking statements, including statements regarding the company's future financial and operating results, future market conditions, the plans and objectives of management for future operations, the company's partnering efforts, the company's capital requirements, the company's future product offerings and the company's research and development programs. These forward-looking statements are not historical facts, but rather, are based on Xencor's current expectations and beliefs and are based on information currently available to us. The outcome of the events described in these forward-looking statements is subject to known and unknown risks, uncertainties and other factors that could cause actual results to differ materially from the results anticipated by these forward-looking statements, including but not limited to those factors contained in the Risk Factors section of Xencor's most recently filed annual report on Form 10-K and quarterly report on Form 10-Q.

With that, let me turn it over to Bassil.

Thanks, Hannah, and good afternoon, everyone. Thank you for joining us. Xencor spent the last several years expanding and advancing a broad pipeline of engineered antibody drug candidates based on our XmAb antibody engineering platform, which is a small set of antibody Fc domains whose sequences we've slightly modified to improve natural antibody functions, such as immune regulation and antigens clearance, cytotoxicity, circulating half-life or to create bispecific antibody structures that are stable, long-acting and readily produced.

Today, XmAb-based antibodies are being developed internally by Xencor and by our partners in 11 different clinical stage programs for the treatment of a wide array of diseases. This includes 4 programs that we're advancing internally with multiple additional programs we expect to start clinical trials for in the next 18 months. The breadth of this pipeline is enabled by the XmAb platform's portability, our ability to plug-and-play an Fc domain that we've engineered for a particular property into multiple new antibodies.

Now we had several updates for our internal portfolio in the second quarter. In May, we announced receipt of orphan drug designation for our lead compound, XmAb5871, in patients with IgG4-Related Disease or IgG4-RD. And in June, we announced updated interim data from our ongoing study in this indication, which showed that 14 of 15 treated patients responded to therapy with 6 achieving remission. We expect top line results from the study later this year, and we're engaging with the U.S. FDA to discuss the development plan, future trials and potential registration requirements in this indication.

We also expect top line results from our subcutaneous administration Phase I study of XmAb7195 by year end and initial data from our Phase I study of XmAb14045 and 13676 in 2018, pending alignment with our partner, Novartis.

Overall, we expect to have proof of concept data for these [forward-lead] programs by the end of 2018. This will allow us to select the best program or programs to independently advance into later stages of development as we continue to mature Xencor into a fully integrated biotech company.

And we're planning to file INDs for our next 2 bispecific oncology programs, XmAb's 18087 and 20717, in late 2018 and in 20 -- late 2017, I should say, and in 2018, respectively, with more programs to follow.

With that, I'll now get into the specifics of our clinical and preclinical programs, starting with XmAb5871. Now that's our first-in-class monoclonal antibody currently in Phase II development for IgG4-RD and systemic lupus erythematosus, or SLE. Both IgG4-RD and SLE are diseases with a strong rationale for B-cell inhibition, in which we can execute efficient Phase II trials and potentially address areas of high unmet need.

5871 works by targeting CD19 with its variable domain and with our XmAb immune inhibitor Fc domain used to target Fc gamma RIIb, an Fc receptor that inhibits B-cell function. As the first known drug candidate to target Fc gamma RIIb, 5871 offers patients a differentiated product profile and a unique mechanism of action. It potently and broadly blocks B-cell activation without depleting the B-cells, creating a reversible inhibition that can be stopped to help avoid long-term immunosuppression.

Now as we announced last quarter, subcutaneous dosing of 5871 is well tolerated and supports every other week dosing, which offers patients and doctors a simpler, more flexible treatment option than infusions.

We provided an update for 5871 and IgG4-RD in June. Now this disease, IgG4-RD, is a newly defined fibro-inflammatory autoimmune disorder that typically affects multiple organs causing tumor-like swelling, a variable degree of fibrosis and potentially irreversible organ damage. It's characterized by a lymphoplasmacytic infiltrate in the affected organs rich in IgG4-positive plasma cells. Now there's no approved therapies for the approximately 40,000 people affected in the U.S., and corticosteroids are the current standard of care. We believe we can be at the forefront of providing a treatment option to these patients, and we are discussing future development plans, including clinical trial design and potential registration requirements with the FDA this year.

Now at the European -- the Annual European Congress of Rheumatology, or EULAR, meeting in June, we reported updated interim data from our ongoing open-label, single-arm Phase II study, which showed that 14 of 15 treated patients or 93% achieved a response to therapy of at least a 3-point reduction in the IgG4-RD Responder Index. This includes 12 patients or 80% who received -- achieved this response within 2 weeks of their first dose. Responses continued to deepen over time with 6 patients achieving remission or Responder Index of 0.

Now 5871 continues to be well tolerated with all drug-related adverse events graded as mild or moderate and no AE reported in more than 2 patients. We're very encouraged by the rapid initial responses of therapy observed in our interim data set, and we look forward to reporting top line data from the trial later this year.

We also continue to progress our Phase II study in SLE. As a reminder, this is the randomized, double-blind, placebo-controlled, multiple-dose study designed to evaluate XmAb5871's ability to maintain the improvement in SLE disease activity after a short course of intramuscular steroid therapy and in the absence of immunosuppressive medication. We designed this study to assess 5871's effect on disease activity with a shorter time to endpoint and with fewer patients compared to standard SLE trials. We expect to report data in late 2018 or early 2019.

Turning now to XmAb7195, our first-in-class monoclonal antibody that targets IgE with its variable domain and uses the same XmAb immune inhibitor Fc domain as XmAb5871 to target Fc gamma RIIb. In contrast to approved IgE-targeting therapies, which block IgE receptor binding only, 7195, with its immune inhibitor Fc domain, leverages 3 distinct mechanisms to reduce IgE levels: it sequesters free IgE and blocks its IgE signaling through its receptor; it also suppresses B-cell differentiation into IgE-secreting plasma cells; and finally and critically, it enables -- it rapidly clears IgE from the circulation.

Now while existing therapies validate IgE as a robust target for reducing asthma symptoms and disability, potential -- a potentially significant fraction of patients are not addressed due to suboptimal potency. 7195 has induced a rapid and potent suppression of free IgE below the level of detection from single doses in 75% of high-IgE subjects. Subjects typically contraindicated from receiving existing IgE-reducing -- or IgE-blocking therapies. Now based on these results, we believe 7195 has the potential to be an important treatment for allergic diseases, and we continue to evaluate our subcutaneously administered formulation of 7195 in a Phase Ib trial and are on track to announce top line data later this year.

Now I'll shift to our bispecific oncology pipeline. Our bispecific antibodies are built on a novel Fc domain, which provides a robust scaffold for 2 or more different antigen-binding domains. The results of the single molecule [and] simultaneously binding multiple targets while preserving important properties of native antibodies like long circulating half-life, stability and ease of manufacture.

Our lead bispecific programs are tumor-targeted antibodies that contain a tumor antigen-binding domain on one side and a cytotoxic T-cell-binding domain on the other. They work by activating T-cells at the site of a tumor for highly potent killing of malignant cells. Now importantly, the format of our bispecifics allows us to tune their potency to balance antitumor activity with the reduction of immunotoxicity that can result from T-cell activation.

We're currently evaluating in the clinic 2 bispecific oncology candidates in Phase I studies. The first is XmAb14045 for the treatment of acute myeloid leukemia and other CD123-expressing hematologic malignancies. Now 14045 engages the immune system by binding to CD123, a protein highly expressed in AML and leukemic stem cells and is associated also with poor prognosis. Now 14045 binds to CD3 on the other side of the molecule, and that's a protein outside of toxic T-cells, and it activates the T-cells for highly potent and targeted killing of CD123-expressing tumor cells. The second candidate, XmAb13676, binds the CD3 and CD20, which is a highly expressed antigen on B-cell tumors like CLL and NHL.

Now both programs are partnered with Novartis who licensed ex-U. S. commercial rights from us in June of 2016. And pending alignment with them, we expect to report initial data from both studies in 2018.

Now our expanding bispecific pipeline also includes XmAb18087 and XmAb20717. We plan to file an IND for 18087, which targets SSTR2, or somatostatin receptor 2, and CD3 for the treatment of neuroendocrine tumors later this year. XmAb20717 is our first candidate to simultaneously target 2 T-cell checkpoint targets, PD-1 and CTLA-4, and it's designed for use in multiple oncology indications. As a dual checkpoint antibody, we believe 20717 can improve the selectivity of combination checkpoint inhibitor therapies and eliminate the need for multiple antibodies in combo. We look forward to advancing it into the clinic in 2018. We are -- we're also continuing the preclinical development of several additional programs with additional INDs to follow in 2018.

Next, a quick update on our partnerships. 9 pharmaceutical companies and the NIH are advancing drug candidates either discovered at Xencor or they're relying on our proprietary XmAb Fc domains. This includes 7 programs currently undergoing clinical testing.

In the second quarter, our partner, MorphoSys, initiated the pivotal Phase III portion of its Phase II/III study of MOR208, formerly called XmAb5574, in combination with bendamustine in diffuse large B-cell lymphoma. This triggered a milestone payment to Xencor of $12.5 million. MOR208 uses our cytotoxic Fc domain to target CD19 in development for B-cell malignancies broadly.

Now with this advancement, we now have 3 partner programs in Phase III testing: the others being talacotuzumab, which is in development by CSL Limited and its licensee, Janssen, and it also utilizes our XmAb cytotoxic Fc domain. An undisclosed program is in development by Alexion in Phase III, and that one utilizes our half-life extension Fc domain.

Finally, I'd also like to welcome a new to member at Xencor. In the second quarter, we welcomed Dr. Raphael Clynes as Vice President of Translational Biology. In this newly created role, Dr. Clynes will focus on the biological mechanisms of our antibody drug candidates with a particular focus on immuno-oncology.

And with that, I'll turn the call over to John to review our second quarter and first half financial results.

Thank you, Bassil. This afternoon's press release reported cash, cash equivalents and marketable securities totaling $378.7 million as of June 30, 2017, compared to $403.5 million as of December 31, 2016. The decrease reflects net spending on operations for the first 6 months of 2017.

Total revenue for the second quarter of 2017 was $13.3 million compared to $66 million for the same period in 2016. Revenues for the 6 months ended June 30, 2017, were $17.7 million compared to $73.3 million in the same period of 2016. Revenues earned on the 3- and 6-month periods ended June 30, 2017, were primarily milestones received from our CSL and MorphoSys collaborations compared to revenues for the same period in 2016, which were earned primarily from the company's Novartis and Amgen collaborations.

Research and development expenditures for the second quarter of 2017 were $16.9 million compared to $14.4 million for the same period in 2016. Total R&D expenses for the 6 months ended June 30, 2017, were $32 million compared to $24.4 million in the same period of 2016. The increased R&D spending in the 3 and 6 months ended June 30, 2017, over the same period in 2016 is primarily due to additional spending on Xencor's pipeline of bispecific oncology candidates.

General and administrative expenses in the second quarter of 2017 were $4.1 million compared to $3 million for the same period in 2016. Total G&A expenses for the 6 months ended June 30, 2017, were $8.9 million compared to $7 million for the same period of 2016. The increased spending in G&A in the 3 and 6 months ended June 30, 2017, reflects additional charges for stock-based compensation.

Noncash share-based compensation for the first 6 months of 2017 was $6.6 million compared to $4 million for the first 6 months of 2016.

Net loss for the second quarter of 2017 was $6.9 million or $0.15 on a fully diluted per share basis compared to net income of $47.2 million or $1.13 on a fully diluted per share basis for the same period in 2016. For the 6 months ended June 30, 2017, net loss was $21.5 million or $0.46 on a fully diluted per share basis compared to a net income of $40.8 million or $0.98 on a fully diluted per share basis for the first 6 months of 2016. The loss reported for the 3 and 6 months ended June 30, 2016, compared to income reported for the same periods in 2016 is primarily due to milestone revenue received from CSL and MorphoSys in 2017 compared to revenue recognized from our Novartis and Amgen collaborations in 2016.

The total shares outstanding was 46,854,762 as of June 30, 2017, compared to 40,944,080 as of June 30, 2016. The increase in total shares at June 30, 2017, reflects the sale of shares in the December 2016 financing.

Based on our current operating plans, we expect to have cash to fund research and development programs and operations beyond 2020 and to end 2017 with approximately $340 million in cash, cash equivalents and marketable securities.

With that, we would now like to open the call up for your questions. Operator?

(Operator Instructions) And our first question comes from Michael Schmidt from Leerink Partners.

I had a couple on 5871. And Bassil, if you maybe could talk a little bit more about what the points are that you need to clarify with the FDA ahead of initiating a pivotal program. A question on what pivotal trial could look like. Will this be a randomized controlled study, potentially, or single-arm study? And lastly, is IgG4-Related Disease an indication where one would expect a placebo effect? And maybe if you could just comment on some of those considerations.

Sure. So I guess the points we really want to clarify with the FDA are does the general approach of whatever trial design we propose, is it consistent with their expectations? It's a brand-new disease. There's no experience directly with this disease, and we think dialogue at the end of Phase II is going to be very important, and we'll learn a lot from it. So we fully expect there to be a need for a randomized control trial, not a single-arm trial. I think correlates in autoimmune disease that could be instructive for this study would be those in vasculitides. For example, the trial that was recently described by Dr. Johnston at ACR last fall for giant cell arteritis, which is a form of vasculitis with an antibody drug. IgG4-RD has common features with the vasculitides. It affects multiple organs. There's microvascular damage. There's organ damage caused sometimes by fibrosis. And so I think those are good metrics. The trials, there's 3 that we're aware of that were in the vasculitides for antibodies to get -- well, in 2 cases, approved, and the third case, a successful Phase III announced. They're randomized against standard of care, so that's what we'd expect to have happen here. They were typically on the order of a couple of hundred patients. It all depends on the treatment effect size, and that's one of the things we're looking at carefully, so we can understand how to size our trial. We're certain we would want to do a randomized study against standard of care, and the question is what -- how do the details all look. These are the kind of the points we want to understand best with the FDA. We want to certainly communicate to them the work that's been done to validate and study the IgG4-RD Responder Index, which is based on the Birmingham Vasculitis Activity Scale (sic) [Birmingham Vasculitis Activity Score] or I should say is related to it and sort of was developed using the BVAS as a guidepost. And the BVAS, of course, has been validated by several approved drugs using it. So we want to educate them on the endpoint measurement and then, of course, define how we're going to measure the endpoint. Now in this case, for IgG4-RD, standard of care is fairly simple. It's corticosteroids that are typically used as rescue upon a flare of disease. And so in terms of expecting a placebo effect, I think what we would imagine is that we're going to be looking at really flare rate, and then the rescue with a corticosteroid would typically be done in response to an increasing Responder Index or evidence of symptoms or the like. And so it's hard to define a placebo effect when the event you're measuring is essentially going to be done in the context of somebody who's been tapered off of their steroids because it's given up a [side effect]. You don't keep people on high dose of steroids forever. And so the placebo in a sense here is the watching and waiting for the next dose of steroids rather than some active control arm. Now is there going to be a placebo effect because patients are on study and are being monitored more carefully? I don't know. We don't have any experience there. And again, that's one of the things we'll have to look at in terms of powering it properly.

Yes. Maybe placebo effect was the wrong choice, but -- of words. But I was kind of thinking, is there data that -- what would one expect in terms of IgG4 Responder Index for the control arm, for example, I think that's...

I think what you're going to expect is that the -- there will be an initial rapid response and an effective response, usually, not always, but usually a deep response to corticosteroid therapy. And then very commonly, patients will have a disease flare within some period of time. And so it's expected that what we're going to likely do is, again, based on the similarities to vasculitis and how those trials were run, you'll induce a patient into a response with standard of care, and you'll follow in maintenance, head to head against placebo, so 5871 against placebo. And then if the patient has a relapse of disease, you would treat with corticosteroids as rescue. And so in that sense, the design is -- the efficacy of corticosteroids will help determine how long before a patient will likely relapse, and that'll determine the duration of the trial and the effect size we would expect to see. And there's frankly not that much natural history information. We're learning a lot from our Phase II about how rapidly our drug works. But the natural history on steroids is still based on just a few studies and sets of experiences with this disease. I think that the idea that the endpoint is based on the disease exacerbation is -- which was what was done in these other areas and we're controlling just with placebo in a sense, and then the rescue is with steroids, help somewhat with the idea that you're masking the disease, at least during that interim after you put a patient into response. Did that help? I think the giant cell arteritis, [the after] trial, which was presented in November of this past year and was just published in New England Journal, has that kind of approach as did the others and I think that's instructive.

Our next question comes from David Nierengarten from Wedbush Securities.

First off, I noticed, not sure if this is a change in language, but the bispecific CD3, 123 antibody looking at both AML and other CD123-expressing heme malignancies, just curious, you guys are opening it up to BPDCN, and then other less common heme malignancies. And then the second question was, refresh my memory if the SSTR2 CD3 bispecific has incorporated any of the CD3-tuning technologies and/or research that you guys have been working on.

Sure. Thanks, David. For the 14045, it was initially in the plan, and it's been on our clinicaltrials.gov entry. I believe we've stated this publicly. I'd have to go look at my exact press releases and notes, but I believe we've had this in our public disclosure, that from the very beginning, we designed a trial for CD123-expressing malignancies, and then there's a list of things we include, like, of course, AML, which is by far the most predominant or common of those. BPDCN certainly fits that bill. There's certain types of relapse CML that fit the bill. There's other more rare tumors that do as well. So that's not new. We expect the majority of patients, certainly in the Phase I dose escalation, to be AML because those are the easiest to find, and we just want to accrue this trial rapidly and advance this as quickly as we can. So that's, that one. On the SSTR2 CD3 compound, the XmAb18087, we looked in primate models very carefully at the effect of affinity on both sides, and we selected one that has, I believe we reported this, an affinity in the animal range for the CD3 side and I don't recall the affinity for the SSTR2. But we looked very carefully at the primate models as well as the in vitro with cells to find where we had that window of good activity at hitting the target cells and as well, the reasonable Cytokine Release Syndrome that we managed to get some of our heme-targeting tumors. So we did thoroughly go through it, just the answer you end up with, in that case, will depend on a particular target antigen. So all the -- yes, so we absolutely looked at it, and that's what we got. I believe we presented much of that at AACR this past year.

Yes. Just to back up to the 123, is it under Novartis' control on how to report out the initial data? Because it would be interesting to see if there's any stratification according to CD3, 1 2 -- or CD123 expressions since it does over -- it is more common in AML, but there are certain patients with AML that have higher expression levels, of course, than others. So are there any plans to report on that or discuss those details when the data come out?

Yes. No. We'll try to dive into the different -- whatever different tumor types we might have as well as see if we can glean anything from CD123 expression. I believe, from my understanding, there -- the variations that are seen in AML are typically in a leukemic blast cells, I don't know, I could be wrong, I don't know if it's so much on the blast -- or I'm sorry, on the leukemic stem cells rather than cellular blast cells. But we're doing very thorough work characterizing the phenotypes of these patients' tumors, and so we'll see what we see.

Our next question comes from Arlinda Lee from Canaccord.

I guess, 2 things. One, on 7195, that subcutaneous trial seems to have started around the same time as the 5871. Is there -- can you remind us of the design and if there were any differences to account for maybe why this might take a little bit longer? And then I guess on the bispecifics, can you -- you mentioned in your press release that you were going to talk a little about your tumor microenvironment targeting thing, pre-clinically. What else are you guys thinking about in terms of bispecifics? And might we start to see the Amgen CD3, CD38 partner program in the clinic anytime soon?

Sure. So first, on 7195, we did start it around the same time as the 5871 subcu trial. Of course, the study design here was a little bit more complex. The 7195 subcu trial was a multi-dose, weekly study done in a variety -- a series of doses, with a dose escalation portion, first in healthy and then in atopic patients, so we could examine different IgE levels and build a robust pharmacokinetic and pharmacodynamic data set. So that is IgE reduction response compared to pharmacokinetics subcu, so we can build that across a range of IgE. So it's a somewhat more complex trial and larger trial than simply enrolling healthy volunteers, only like we did for 5871. So the timing is a bit different. That said, we are still on track to report it out -- top line results before the end of this year, so just about a year and maybe a couple of months after we started that. For the tumor microenvironment in bispecifics that we hope to have data later, I guess more information and more description of our molecules later this year, hopefully, at a scientific conference. We have, after our PD-1 CTLA-4 dual blocker, that's XmAb20717, which we expect to have the IND filed in 2018. We expect to have other agents. We've talked about these, our CTLA-4 x LAG -- CTLA-4 x LAG3 bispecific blockers is the next one after that. And then a PD-1 x costimulatory molecule bispecific that inhibits PD-1 and agonizes a costimulatory molecule that's also moving forward, and we have a candidate to clear it with an XmAb number and hope to have the IND for both of those in 2018. Did that answer your question? I wasn't clear on the first part.

Yes. And then maybe lastly on the 13551 partner with Amgen?

Sure. So the CD38 x CD3 targeting program that was part of our Amgen collaboration, I guess we did that deal in September of 2015, so we're going on 2 years. I know that they're still actively advancing. I'm hopeful that they will have it in the clinic soon. I can't really say because that program was fully out-licensed to them, and we're just watching them as they progress it. Yes, we're very excited to see if that one can get in the clinic very soon.

Our next question comes from Edward Tenthoff from Piper Jaffray.

A little higher-level question, if I may, on 7195. First, with some of the newer agents in asthma IL-5 and now with the dupilumab data coming, how do you see sort of the biologic asthma landscape changing? And how could that change sort of the landscape for your compound?

Yes. I think severe asthma has been sort of -- had historically been sort of absent, really, biologically active immune modulators like antibodies. And in stark contrast, like in the world of rheumatology where you have -- after the TNF blockers came out 20 years ago, there was just a flood of other agents. That really created a diverse, multi-layer treatment paradigm in, say, rheumatoid arthritis. And now say, in psoriasis where you have first-line agents, second-line agents, because people typically progress through these agents -- either they're growing responsive or their disease waxes, and they have to have something else to hit it with. And asthma has been completely absent that with the exception of the Xolair tagging along to block IgE. And now that's all changing. There's new agents coming on, like dupilumab, which is an IL-4 receptor blocker, the IL-5 agent, others behind it, I think are going to create this broadening and acceptance of biologic using more expansively in asthma, which I think is great for patients. And I think you have a situation that's again similar to rheumatology. You have sort of an unclear immune ideology of what's driving people's disease -- there's certainly variability. We know that, for example, the IL-4 blockers, they typically try to focus on eosinophilic heavy disease, though eosinophils are probably involved in everybody's allergic asthma. How it expresses in certain patients certainly differs. Nobody has a full understanding of that. And so I think you're going to have a situation that's going to evolve where use of biologics is more accepted, people with more severe disease are going to be able to taper off of with their oral corticosteroids and manage their disease better just generally with more biologic agents, and I think you're going to have multiple lines of therapy. I think it's going to be great for patients. And I think you're going to have in that, we believe, a strong and central role for the well-established utility of IgE blockade. So IgE blockade is not driven by any particular biomarker like some of these therapies, but it's really about the central mediator of the allergic response in disease, IgE binding to your allergen, right, and blocking the [downs and consequences] of that. So we think IgE blocker is going to be central to this paradigm of multiple agents, and we think that with a much more effective blockade or reduction of IgE like we can achieve with 7195, we could play an important role. So I think it's not going to be one biologic. I think it's going to be a broad number.

I think that makes a lot of sense, I agree. Question on 123, obviously, a lot of programs going after that target for AML. How do you sort of see your bispecific maybe differentiating? Obviously, early days, but what's your goal to differentiate that program versus some of the other approaches?

Sure. I think there's a 2 sort of buckets of CD123-targeting therapies. It's certainly an emerging target where there's been a lot of interest in the last few years that have now resulted in a handful of clinical programs. You've got sort of -- take nonbispecific programs sort of as the baseline, programs like the talacotuzumab program that Janssen has going forward in Phase III in AML that actually happens to use our cytotoxic Fc domain to recruit natural killer cells in macrophages. It seemed to, from the reported data, to be relatively well tolerated. It must be showing some responses that are in Phase III [that they've] published very little data. You've got also advancing the diphtheria toxin conjugate to IL-3 itself. So IL-3 is a natural ligand for CD123. That had shown activity, though with significant, I think, toxicities that seem to be mediated by the function of diphtheria toxin itself. So I think what you've established there is that you can have activity and probably efficacy from CD123 targeting, and your toxicity seems to track from your modality, whether it's diphtheria toxin on the one hand or a fairly well tolerated sort of more traditional antibody on the other hand with the talacotuzumab. So it sort of established that target as one, "Okay, look, there's some traction here." It's a heme target that looks like you can hit with something pretty hard, and hitting that target CD123 itself might not be something that a patient can handle. Great. So now you've got the bispecifics where you have the first programs that we're aware of are, for example, a program that is a non-Fc-containing bispecific Macrogenics has developed -- is developing that has a relatively short half-life, say, relative to, say, the Fc-containing CD123 approach that we have, so relatively short half-life. I think it's something that's going to be using a multi-day, if not multi-week, continuous infusion, which creates challenges for delivery as well as having a very, very high potency. Our agent as well as the bispecific -- Fc-containing bispecific agent that Janssen has now reintroduced into Phase I, they have a lower potency, and that might give you a better window to introduce the drug to patients to get it into them, to get it active without having such a profound immune activation that you have difficulty managing the toxicity. That was the whole thesis for our approach. Still very highly potent molecule, maybe 1,000x more potent than a traditional antibody, but tuned down relative to the very high potency, first-generation bispecifics, like the blinatumomab, Amgen program or the CD123 that I just mentioned without an Fc from Macrogenics but also have longer half-life, so the drug can stay around for days, maybe weeks and not just minutes. I think that profile of having something you can dose simply and have manageable [tox] to get to, hopefully, activity that we'll see is a great window to be in.

(Operator Instructions) And our next question comes from Christopher Marai from Nomura Instinet.

And I am showing no further questions from our phone lines. I would now like to turn the conference back over to Bassil Dahiyat for any closing remarks.

Thank you very much, operator, and thank you all for your time. We look forward to providing further updates as we continue to advance our XmAb pipeline later this year. And thanks, again. Bye-bye.

Ladies and gentlemen, thank you for participating in today's conference. This does conclude the program, and you may all disconnect. Everyone, have a wonderful day.