MacroGenics, Inc. (MGNX) Earnings Call Transcript & Summary

Okay. Great. Welcome, everyone, to the afternoon session, Large Molecule Drugs for Oncology -- Framing the Tip of the Spear for Novel Antibodies and Protein Therapeutics. I'm Yigal Nochomovitz, one of the biotech analysts here at Citi. It's my pleasure to have with me the leaders of 5 protein antibody companies, apart from Harpoon Therapeutics, Jerry McMahon, President and CEO; from MacroGenics, Scott Koenig, also President and CEO; from Merus, Bill Lundberg, the CEO; from Shattuck Labs, Taylor Schreiber, the CEO; and from Zymeworks, Ali Tehrani, President and CEO. So welcome, gentlemen, to the program, a lot to discuss today. But as a starting point, perhaps you could just give a brief introduction to your company. And in doing so, cover what are the strengths and differentiating qualities of your antibody or protein therapeutic platforms. Maybe we could start with Scott.

Thank you, Yigal, and it's a real pleasure to join this panel with everybody. MacroGenics is developing now for over 20 years antibody-based therapeutics. We have 3 platforms that we've been developing, 8 molecules in clinical development. And as you know, we commercialized one molecule from our platform. So the 3 technologies include our Fc engineering technology to better engage family of Fc receptors to promote better tumor killing through antibody-dependent cellular cytotoxicity. We have 2 molecules in clinical development and also one commercialized using this technology. Second is around a bispecific and multispecific platform. This is our DART technology, which uses a diabody structure. We've utilized this to both promote redirected killing mechanisms as well as combination of our [Technical Difficulty]. And then finally, we have 2 molecules in development for using antibody drug conjugate technology using our linker-toxin technology that we've in-licensed from our partners. Overall, we are platform-agnostic, but the beauty of our technologies is that they use complementary mechanisms. And so these orthogonal mechanisms help to promote better tumor control, including the activation of both innate and specific immunity. And with that, I'll open it up for discussions by the other parties.

Sure. Bill, do you want to give a quick overview from Merus.

Thanks, Yigal. Thanks for the invitation to speak on the panel. Thanks to my colleagues for the opportunity to join the same panel we had -- similar panel we had almost a year ago. And we're here again to talk about progress we've made. What's special about Merus is this common light chain antibody technology. Historically, making a multibinding protein has been difficult, and the older approaches have made these Frankenstein-like structures glomming on one binding domain to another. These have been hard to make, hard to manufacture, hard to purify, behave radically in patients. They have immunogenicity, short half-life, et cetera. And each of our companies has a different approach to solving this problem, in particular, making bispecifics more like human antibodies. But getting the 2 heavy chains and 2 light chains to pair, which can recombine in 9 different ways inside a cell to pair the right way has been challenging. Merus' proprietary approach is to make all the light chains the same, hence, the common light chain notion. And then the problem is simply enforcing the left side heavy-chain binding to the right side heavy chain, which we do with our proprietary charge-based approach. So out of a single cell, we make a monoclonal that is bispecific and its binding. We can now leverage the decades of tools and technologies of monoclonal antibodies, the high-throughput screens, the fully human IgG antibody-like structure, functional modifications like ADCC to select the best and most appropriate bispecific and trispecific in the setting. And our bispecifics work as shown in most recent ASCO presentations for our drug NRG1 fusion cancers. And others have recognized this approach in our partnerships as a really powerful way to make medicines in our partnerships with insight, Loxo Oncology, et cetera. So our company in a nutshell.

Great. Thank you. Jerry?

Yes. So Harpoon started in 2015, founded by Patrick Baeuerle, who was formerly at Micromet, which was a company that was pioneering T cell engagers using molecules that could create a physical connection between T cells and tumor antigens expressed on the surface. And so with that as a goal now to bring that mechanism to solid tumors, we created now 2 platforms. We have another one coming, with really the goal of reprogramming a T cell to direct its energy against tumors and kill them. So the platform, in a nutshell, is a single polypeptide with 3 binding domains, 2 of the 3 derived from [indiscernible], pH-only binders that can show strong potency in T cell-directed killing, and in part, nice drug properties by virtue of binding to human serum albumin. And so these molecules, we now have 4 in the clinic. It's 4 different tumor antigens. They're all in dose escalation and expected to move into Phase II studies. And then our second platform is called ProTriTAC, which uses 3 binding domains, but changes the orientation. And with an engineered albumin domain, we've been able to make the CD3 locus inert until it's activated in the tumor where the albumin binding feature is relieved and allowing the T cell engager to operate then in a tumor microenvironment. So our company focus has been around really trying to optimize the ability to access T cells in solid tumors and direct them to kill various cancers.

Thank you. Taylor?

Great. Thanks, Yigal, and happy to join the other panelists today. Shattuck is similar to the other companies on the panel here, in the sense that, we are also building multifunctional biologics, but different in the sense that we do not use antibody-derived binding domains on the functional ends of what we call the agonist redirected checkpoint or ARC platform. The ARC platform was created to solve a very specific set of problems that applied immuno-oncology and other diseases. And specifically within immuno-oncology, what we wanted to do was to build a biologic that could, on the one side, block immune checkpoint targets like PD-1 or CD47, while on the other side, activating co-stimulatory receptors within what's known as the tumor necrosis factor superfamily of receptors. This is a family of about 30 different receptors that is responsible for fine-tuning adaptive -- both innate and adaptive immune responses and includes targets that many in the audience may have heard of, such as CD40, 4-1BB, OX40 and others. And so we have 2 molecules in a total of 3 different clinical trials right now. Our internal lead molecule is a SIRP-alpha-Fc-CD40 ligand's ARC, which we call 172154. This molecule is designed to simultaneously block the CD47 checkpoint and improve the ability of macrophages to cross-present antigens to T cells via CD40 binding. And the second molecule we have in the clinic is a PD1-Fc-OX40 ligand fusion protein. We call it 279252, that is designed to simultaneously inhibit PD-L1 and L2 on the surface of tumor cells while co-stimulating OX40 on T cells to amplify the magnitude of an adaptive T cell response. So happy to join the panel and entertain your questions.

Thank you. And last but not least, Ali from Zymeworks.

Thank you, Yigal. Thank you for hosting back here on this panel. Just as a brief control to Zymeworks. We were founded on the principle and on the basis of differentiating in protein engineering. We've leveraged that capability in developing a number of platforms, therapeutic platforms, including bispecifics that we have leveraged into 9 active pharma partnerships and the development of 2 clinical assets, zanidatamab and ZW49. For lead clinical candidate, zanidatamab, which is a biparatopic HER2 antibody, we've demonstrated clinical activity in over 300 patients in a broad spectrum of tumor types from breast, the gastric, the pancreatic cancer. And overall, our development strategy for this asset and our commercialization strategy for this asset comes down to 3 components. We have a fast market strategy where we're on track for our first label in the near future in biliary tract cancer. We have breakthrough designation there. We have a display strategy where we are going up against Herceptin and Herceptin plus Perjeta in a number of indications, including gastroesophageal and breast cancer. And we have a diversified strategy where, in combination with other programs, such as CD47 from ALX or Ibrance from Pfizer, we're looking to address the needs of patients where currently the other targeted therapies are unable to do so effectively. Overall, our background, our capabilities, our platforms and protein engineering has enabled us to open the field and to address the shortcomings and the limitations of other approaches from the past to truly put therapeutics on the table for patients and their needs as we go forward.

Okay. Great. Thank you very much. And obviously, you all have different pipelines, which we'll get to a bit later in the conversation. But hopefully, the following question that I'll ask will stir a little bit of debate. I think in some investors' minds, there's some -- there's a notion that some of the complex protein therapeutics, such as the bispecifics, may have to compete with cellular therapies such as the CAR-Ts or the CAR NK. So help us understand why that's possibly not the case? And how you could potentially collaborate or combine with the cellular therapies, as you know they're more and more cellular therapy companies every single day? So maybe, Scott, you want to give that one a shot.

Sure. So it's obviously a complex question that it continues to evolve. Clearly, the greatest responses that one has seen in the CAR-T therapy has been in the case of hematologic malignancies with less of resounding responses observed in solid tumors, although there has been more recent inroads there. The complexity, as everybody on this table knows, with regard to the manufacturing and engineering of cell therapies creates great obstacles. Although, again, there has been some inroads and the issue becomes not only the cost but ultimately the complexity of using these molecules in particular patient populations. A lot of the issues were focused on potential side effect profiles with cytokine release being dramatically accelerated or exaggerated as compared to as we'll get to some of the bispecific technologies. Some of that is obviously trying to be addressed with additional engineering there, but to date, I don't see a resolution of that problem. And the fact of the matter is the persistence, particularly on -- depending on the cell systems being used in the patient can be quite short, depending on which cell type you use for transfer. So clearly, there has been dramatic inroads on the bispecific technologies. I won't address the differences among the different technologies here, and there are some features that could favor one from the others. But the beauty of what, for instance, MacroGenics is doing is that not only can we engineer these molecules to have unique engagement for 2 different molecules, we can do it for 3 different molecules, we can do different stoichiometries. To the one, we can add specificities for multiple antigens so that you ultimately hone in on the right tumor molecule. So -- and a lot of advances, including the work we've done in MacroGenics to mitigate on the redirected killing mechanism, the cytokine release with alternative versions of CD3, other molecules that can be directed to T cells and NK cells. And then finally, the utility of this to combine checkpoint molecules that are quite distinct and actually can enhance the activity as compared to separate antibodies. So I see both the evolution of the bispecific technology and the ease of production and the utility in various different ways, ultimately being an advantage over the cell therapies, which, in fact, have shown very good responses admittedly in hematological malignancies.

Yigal, just to provide a brief comment, as well, after Scott's point of view. I think we are already seeing different use cases, facility with engineering, as Scott has laid out with bispecifics. They're very different types of molecules. One of the most compelling issues is, how do we get solid tumors to respond to immunodirected therapies that are more complex. And one of the best pieces of the data is around RYBREVANT from J&J that's shown recently 64% response rate in exon 14 mutant lung cancer, 100% in frontline lung cancer. And we have a similar molecule in the clinic MCLA-129. And these data really point to the concept that you can engineer target-pair combinations to get the right balance of efficacy and safety that just haven't been tractable with cell therapy approaches. So I think to Scott's point, we have a facility with engineering, our approaches that really can give us unique opportunities above and beyond cell therapies.

And Ali, you wanted to say something.

Yes. Look, if I may take a slightly different angle here, and I'm not going to sort of take the position of one is better than the other or vice versa. I'm going to take the position of, we all have to think of patients first and their needs first and the biology first. There is going to be instances that cellular therapy is going to be clearly more advantageous over bispecifics or anything else. And there's going to be other instances that the biology needs a large molecule or multifunctional molecule. I don't think there's a silver bullet. I don't think there is a situation where one should pick over the other and say mine is better than yours or yours is better than mine. You have to be very, very clear on the problem you're solving and what does it take to solve that problem. And this is -- I'm not going to take the position of Zymeworks is wonderful at everything, and we're going to do everything really wonderfully than others can. I'm going to take the position that we put patients first, and we focus on addressing their needs. And if we can't solve it with a bispecific then when we won't challenge it.

Okay. Makes sense. Anyone else wanted to add anything.

I'll chime in as a non-T cell engager developer here. And just to add that from an outsider perspective. Clearly, one of the goals that both the CAR-T therapies and the engagers are attempting to solve and going about it in slightly different manners, which I think has the ultimate goal of, as Ali was laying out, keeping the patient's interest first and foremost, is to think about the question of response durability and in designing elements of either the engager CAR-T platforms that are able to use the initial specificity of that CAR-T or engager in a manner which then leads to activation of endogenous immunity against a broader profile of tumor antigen targets, both in solid and heme malignancies. And I think that this is an area of opportunity for both of these classes of therapy moving forward that isn't met with the first generation of either.

Why don't we just add on the other additional point is the combinability of these things, both as a way to enhance the durability of the response, say, in the case of an adaptive cell therapy, activating various checkpoint molecules. And also, the application on the bispecifics, not only in terms of mechanisms that we alluded to of redirected killing or checkpoint blockade but you can also use these bispecifics, trispecifics or whatever with toxin conjugates or with cytokine conjugates. So there is, again, it's really up to the imagination now on how the engineered platforms can evolve in various ways.

Okay. Great. So moving on. I have another conceptual question, which I think should also spur some good debate and discussion, and that's the in vivo generation of protein therapeutics. So what I mean by that is that we've obviously seen the advent of new modalities, such as mRNA for the COVID vaccines. But it may be one day possible to use mRNA to deliver the message that actually codes for the protein similarly, gene therapy as a potential way to generate the protein therapeutic interest in vivo. So just wondering, conceptually, as antibody manufacturers, whether you see this as a threat? Or is it really something that's just too far in the future to be concerned about, but it is getting -- going to have more relevance, especially now that we've seen the commercial viability of the mRNA-based approach. Scott?

Yes. So I welcome, obviously, the application of mRNA technology to be able to express variable domains in various forms, either small targeting molecules or free fragments of antibodies and no issue on that. The question that becomes the persistence, obviously with mRNA, there will be some limitations there. But in terms of some of the structures that we just discussed, whether it be bispecifics, trispecifics, these are conjugating, the complexity of these molecules can't be really addressed by just giving a message in. The folding, the purification that we go about, and in some cases, obviously, conjugation to make even more complex molecules can't be solved by that. But under certain circumstances, where it's easier to get an end product like that, I welcome it.

Okay. Any...

I agree with Scott's comments on -- related to the persistence of some of these approaches. That is certainly something that needs to and I think ultimately will be addressed. And then Scott alluded to the fact that for complex antibody-based biologics, delivery via RNA is somewhat complicated if you have to deliver both a heavy and a light chain let alone multiple heavy and/or light chains. ARCs are different from these approaches and that we produce them from a -- as a single transcript from a single plasma. And therefore, the ARC protein is assembled from a single building block. So from that perspective, they may end up being more amenable to RNA-based delivery methods. The other major technical challenge that, over time, I think the RNA delivery approaches will solve but is not yet solved, is delivery methods. Right now, we're really limited to thinking about delivery of RNAs in cancer patients via intratumoral injection. And particularly in light of the persistence question, that may pose challenges from a therapeutic standpoint. But I do think these challenges as we define them today are opportunities for innovation and ultimately the industry will meet and preparing for that, I think, is in all of our interest.

Thank you. Any other thoughts on that topic? Otherwise, we can move on. Okay. Well, we've got 15 minutes left. So I'll just maybe move into some company-specific questions, if we could, in a sort of a lightning round. So maybe Jerry, starting with you and Harpoon. Could you tell us a little bit more about your lead TriTAC HPN424, which targets PSMA for prostate cancer? Could you just give us kind of a synopsis of the activity profile? And how you see it as a differentiated asset? And what do you think the regulatory hurdle is in metastatic CRPC?

Well, late-stage prostate cancer is a very difficult disease, very heterogeneous. We are using this population to hopefully extract clinical benefit for some of these patients with an infusional product. And we continue to determine what is the optimal dose, the regimen for this. We've seen patients out -- we had one patient out 45 weeks with a durable 60% PSA response rate who was refractory to chemo. And then we have another patient who had only a 10% decrease in PSA, yet had a partial response and is on study for something like 9 months. So you have these patients who clearly benefit from a therapy like this, but you're not seeing it routinely enough in the population at the lower doses we've looked at. So we continue to want to optimize the dose with the hope that we'll see more activity in more patients. And when we do, we'll graduate that program. Complete contrast, let me just mention, we have a fourth program that went in the clinic this year that has really not been limited at all with respect to dose escalation, and we've sailed up significant to those levels. So the target-by-target differences between our forward programs is gigantic with respect to CRS, the doses we're exploring, and of course, the clinical indication. So what's really more important for us is, is the choice of target, and how do we navigate any safety liabilities, which as we continue to see evidence of clinical activity. So that's a bigger answer to your question, but I wanted to kind of go from the prostate program to the other programs.

Okay, perfect. And Scott, for you, you obviously have a very big conference coming up next week or the week after at ESMO, where you're going to have data for MGC018, that's your B7-H3 antibody drug conjugate. I believe it's dose expansion data in prostate as well as in lung cancer. So if you could, to the extent that you can discuss it without giving away too many details, what should our expectations be for that readout? What should we look for? And how do you see the responses for those patients deepening over time?

Yes. Well, thanks very much, Yigal. We're very -- looking forward to the presentation at ESMO of MGC018, which is an ADC molecule using a cleavable linker called duocarmycin, a B7-H3-directed target. As you know, we've presented in the last 2 ASCOs, the initial dose escalation data to 4 mg per kg in the last ASCO. In prostate cancer, we had the first part of the expansion cohort of 22 patients with castration-resistant prostate cancer were progressed on chemotherapy and AR inhibitors and showed a 50%, 11 out of 22 of the patients, even in very early stages of treatment already had a PSA of 50 reduction. So what we have coming up at ESMO is 3 more months of data. We finished the full enrollment of the 40 patients of castration-resistant prostate cancer. We have promised at least 30 or more patients coming from that. So the patients we already presented for PSA 50 plus resist the valuable patients based on measurable tumor signs, which amount to at least half of those patients. And you'll see a lot of data with regard to spider plots, age scores on every single tumor type and then a very profound safety database associated with that. This will also be the first presentation of the second cohort of non-small cell lung cancer, where we have again promised at least 10 patients of the 20 that are completely enrolled in that cohort at least 1 to 2 scans, which are done every 9 weeks on that. So first, the preview of the non-small cell lung cancer patient. So these are 2 of 5 expansion cohorts that are ongoing right now, and we look forward to the review by both the patients, investigators as well as the investment community.

Great. And Bill, obviously, your molecule Zenocutuzumab in NRG1 positive tumors has had some very good progress in the clinic. Could you talk a bit about the recent data at ASCO and lung cancer? And how you see this dataset evolving and what your clinical trial strategy is for Zeno?

Yes. And just to frame it out, we have 4 clinical stage programs, each of which are increasingly generating interest for Zeno, which is our furthest along program. We presented data at ASCO showing a 42% response rate in this genetically defined population of previously treated pancreatic cancer, where the comparable therapy is a 7% response rate. And then in the overall population, a 31% response rate in previously treated cancers that are genetically defined. In general, accelerated approval in oncology has relied on ORR as an endpoint in a group of patients of 55 to 100, 110 patients, for the most recent approvals. And we're rapidly approaching that size and scope of type of data collection, where we think there's a real opportunity to consider this as an appropriate set of data for consideration for regulatory authorities. We have said in June, we will be meeting with the FDA and other regulators later in the year. And as we get through that and reach alignment, we will be sharing updates on or before the first half of next year.

Okay. Perfect. Maybe we can move over to Shattuck. Any thoughts -- or if you could just give us a little bit of a deeper understanding of your ARC platform and your bifunctional fusion proteins as well as the value of having the checkpoint binding domains as well as the TNF binding domains in a single molecule.

Sure. Happy to. Thanks, Yigal. So as I noted, the motivation to build this platform was to build a molecule that could simultaneously inhibit an immune checkpoint and activate a co-stimulatory receptor in the tumor necrosis factor superfamily. And the reason that's a unique challenge is because TNF receptors are trimeric receptors that have to assemble into trimers in order to signal. And most attempts across the industry clinically to go after this class of target, which have been occurring for over 30 years, have utilized bivalent IgG-based antibodies. And this is a situation where, in our view, a bivalent drug is simply the wrong tool for the job if that job is to activate a trimeric receptor. And that's what led us to think outside the antibody scaffold and building these molecules. And so the functional domains in -- I'll talk about 172154 as the example. On the one end, we have the human SIRP-alpha extracellular domain, which will bind and inhibit CD47, adjoined via a central Fc domain to the extracellular domain of the human CD40 ligand protein. And because CD40 ligand in all TNF ligands, just like their receptors, self-assemble into trimers, the combined intermolecular interactions between that trimeric structure and the propensity for the central Fc region to dimerize leads to a hexameric structure that's produced out of the box from the chose cells we use to make these proteins. And so the reason -- and so that's the structural reason why we ended up where we did. The reason to link checkpoints in co-stims is because of the fact that for many co-stimulatory receptors, those same receptors that can activate an immune response can be inhibited by checkpoint receptors. And so creating a molecule that can co-locate checkpoint inhibition and immune co-stimulation at the same place and time creates an antitumor effect that preclinically across the platform is far greater than the sum of the parts. And so that's what really led us down this path, and we have an opportunity coming up at SITC in November to present clinical data from the first 2 programs at the same time. And I think we'll provide a read-through to the ARC platform as a whole and answer some of the outstanding questions the industry has had about whether the outcome of efforts to target TNF receptors with bivalent antibodies has more to do with the target or the tools that were used to go after those targets.

Got it. And Ali, you're developing the HER2 bispecific novel antibody Zanidatamab. So I've been curious to find out a little bit more from you as to how you see this asset fitting into a therapeutic landscape, specifically in gastric and breast cancer? And then more specifically, when could we see data that could prove that the thesis that Zani may, in fact, be able to replace Herceptin and Perjeta.

Thank you, Yigal. Well, I'll answer the last part of your question first. You'll get a first look to that end next week at ESMO. As many of you know, next week, we'll be presenting data in the first-line setting in GEA of Zanidatamab plus chemo, which will be the first time we have shown data in a bit of a head-to-head against Herceptin plus chemo in the first-line setting. Naturally, this data has a read-through into breast cancer as well. And as many of your listeners here may know, we also have a frontline study in breast cancer, looking at Zanidatamab plus chemo as well and that data should be presented in the first half of next year. What next week also will show is a bit of a head-to-head against some of the historical data that has been generated in the frontline setting for GEA. The ToGA study ran by Genentech/Roche, where ultimately, Herceptin plus chemo became the standard of care. The JACOB trial where Perjeta was introduced in the setting and ultimately did not result in an approval. And last but not least, the most recent dataset that came out of Merck in the KEYNOTE trial. So next week at ESMO, the dataset is meant to demonstrate the unique mechanism of action, the unique biology that ultimately should position Zanidatamab to become the standard of care in first-line. Followed that is the datasets that will come out from the triplet combination of adding PD-1 into the equation. This is BeiGene's PD-1 tislelizumab. This is a Phase II study that has been underway and has been maturing the dataset on route to the Phase III randomized multi-arm in the frontline setting GA, which is on track to kick off in Q4 of this year. So over the course of the next 6 to 12 months, we will be sharing GEA data. We will be sharing breast cancer data. We will certainly provide updates in our colorectal study that is underway. And we will be pushing the asset more and more into the early lines of setting and ultimately making it the center of care in the area.

Okay. Perfect. We have a bit more time. So I think we can go around the horn again. Jerry, regarding your platform, your ProTriTAC platform, talk a bit about how you -- how that program is designed to minimize some of the safety issues, such as CRS and half-life issues that have been needed to be managed with some of the other bispecifics? What's -- how do you get around that with the ProTriTAC platform?

Yes. Well, the focus is really about target. It's not about the mechanism. I mean, all T cell-directed killing will release cytokines. The question is whether it becomes a problematic clinical issue. You need that mechanism. But your choice of targets makes a big difference. Some targets elaborate CRS more frequently and more profoundly, other targets don't. And some tissues are dispensable like B cells. The only reason we have T cells looking so good in B cells has quickly eliminated the B cell in the body. So in a way, we've eliminated the normal cell in that situation. But obviously, as you move into other targets, you have to be more wary of where the target is expressed because, obviously, you don't want to direct T cells against normal tissues, unless those tissues either are recoverable or dispensable. So that's where ProTriTAC comes in. It allows us to go after cancer targets that would be problematic from a target specific problem. It's not a technology to solve, CRS. It's a technology to allow cytokine mediated killing to occur preferentially in places where tumors grow. And to do that, we rely upon a proteolytic event that clips the blocking agent of our CD3 domain, which is derived from our albumin-binding domain allows that to be relieved and that allows the T cell engager to work preferentially in those tumor microenvironment. So it's really about targeted space. It's not about CRS. Okay?

Yes. Thank you. And Scott, in terms of expanding your pipeline, I'd be curious to get your thoughts on how you're thinking about molecule design for the early-stage programs. It seems that you're focusing maybe increasing on ADCs with some of the newer agents in your pipeline, but I'd just like to get your thoughts whether that's accurate and what the early-stage programs look like in MacroGenics?

Thank you very much, Yigal. We have a very eclectic group of programs in our preclinical development. Later this year, we will initiate the IND and early next year start enrolling patients in the alternative DART molecule targeting CD123 and CD3. This was designed to have, a, an Fc domain to increase half-life and be able to give intermittently, but secondly, making a minor alteration in CD3 to dramatically reduce cytokine release, but maintain the maximum killing effect. This will be initiated in AML, MDS and potentially other indications. And the same specificity is going to be incorporated in other molecules for solid tumor targets. So a refreshing of the whole redirected killing mechanism. We are going to have some additional molecules that we were developing with ADCs. One of the lead molecules of interest is composed of 2 unique antigens or 2 different tumor specificities in which one could then add an ADC component as well. So you get minimum side effects to normal expressing tissues. And then we have other targeting molecules, bispecifics, that are directed in different effective mechanisms, different cell types [Technical Difficulty]. So you should expect over the next couple of years, different molecules from different parts of our platform going forward.

Got it. And Bill, I'm interested in your -- one of your earlier-stage programs. It's the EGFR cMET bispecific MCLA-129. Could you talk a bit about where you plan to develop that asset in which solid tumors? And then as you know, obviously, J&J recently published data from their Phase I with their EGFR cMET bispecific , Amivantamab, at World Lung. How much read-through should one apply to that program relative to what you're doing with MCLA-129?

MCLA-129 is in a dose escalation part of the Phase I trial. And we think it's really exciting for the reasons you just outlined. This target pair combination, EGFR and cMAT, has shown remarkable responses, 100% in frontline lung cancer, significant responses post-EGFR TKI, like, obviously, when in combination with lazertinib as monotherapy striking responses in an exon 20 mutant EGFR and in cMET mutant lung cancer. And those areas where we've seen really striking in better than comparator responses for this target per combination really direct us where to look for opportunities as we reach the recommended Phase II dose and move into dose expansion. We think it's a real opportunity, in part, because Rybrevant or Amivantamab and MCLA-129 are really quite similar molecules. We think that we may have an advantage in the way we have ADCC enhancement and 1 or 2 other characteristics, but that has to be borne out in the clinic. So we think it's a really exciting opportunity. We just have to head down, execute, get the clinical data to prove the hypothesis.

Thanks. And Taylor for you, for your lead asset, 172154, curious how to think about the bar for success there? Would you say it's best to look at the other CD47 focused antibodies? And what is your clinical development strategy going to look like for 154?

Sure. So I mean, first and foremost, this drug has to walk and talk like a CD47 inhibitor, and we've built it to do that. And one of the key outstanding questions folks have appropriately had about that molecule is whether you can safely dose escalate a CD40 agonist to the level you have to achieve for a CD47 inhibitor. If the answer to that question is yes, then it's quite clear that the underlying biology there is compelling. And the ability for CD40 activation to amplify the ability of macrophages, which phagocytose tumor cell to process and present antigens from those cells they've consumed. Because ultimately, it is not the macrophage phagocytic event, which leads to tumor cell debulking, it's the cross-presentation of those antigens to an adaptive immune response. And this molecule is the first in the CD47 class that has that function. So those sort of key dose escalation questions with CD47 inhibition and convincing CD40 engagement are what folks should look for at SITC.

Great. Thank you very much. And Ali, if I could ask you a key piece of your corporate strategy is partnering. You have quite a long and impressive list of pharma partners. Could you talk about that strategy a little bit more? And how that advances your goals at Zymeworks?

Yes. Well, I think you're right. We have all of our partnerships active. And most notably, we announced that at least this year, both Bristol-Myers Squibb and also J&J have advanced programs into the clinic built on our technology, and we expect that trend to continue over the course of the very near future. Additionally, on new platforms that we generated, most recently presented at AACR, our ProTECT platform, which is a combination of masking and co-stimulation capabilities should generate future partnerships and future activities for our own pipeline and other people's pipeline. So we expect more patients being dosed by our partners, and we expect more partnerships to form, and we will continue developing first and best-in-class platforms right next to our multifunctional capabilities.

Okay. Great. And maybe I can just squeeze in one more conceptual question or maybe a philosophical question. All of you are involved or have been involved in Phase I oncology trials to some or to a greater or lesser extent. Just -- what would be your advice in terms of -- what pitfalls should be avoided in the design and conduct of the Phase I oncology trials? How to make those generate the good data and avoid pitfalls? Any thoughts there?

Well, we've taken the philosophy that we want to make these Phase I studies, ones that we can build upon. And so have enough patients with dose escalation so that you can do expansion cohorts very easily and have enough patients in that cohort to come out with a go-no-go decision based on that analysis. The second is, historically -- particularly a smaller company usually rely on the U.S. to do Phase I study. We take an international footprint now to do Phase I studies, which we would have never imagined before. And so particularly getting access to certain tumor types, you actually have to go to Europe and Asia now routinely to get enough patients in to come out with an answer. The other thing is as many around the table have said, partnerships can be very helpful so that the ability for a partner, who you might be working with even on an early-stage molecule to be able to test it in tumor types that they have more greater access to, could be quite valuable.

Thanks, Scott. Any one else wanted to throw in some final thoughts before we...

I think in our case, the protocol evolution and the interactions with the FDA on a Phase I design for T cell engagers has been dramatic over the last couple of years. Our first program that was out of the gate, has been slow and systematically increasing the dose. Our third and fourth programs have done it in half the time with bigger jumps. So there's been an amazing evolution of understanding and protocol development that has really favored the development of T cell engagers, specifically, which have a particular challenge in a Phase I setting. So that's something we've learned over the last few years.

Sure.

I'll offer just a [indiscernible]. Sometimes people run Phase I studies with the attitude of, "We'll see what we'll see and we'll figure out when we get there," to the best of people's ability to have it be hypothesis directed and really clear our priority of what go means and no go means helps immensely once you get to the point where you have data. That kind of rigor and discipline is important.

Thank you. All right. Thank you all so much for the lively discussion. Much appreciated. Best of luck with your clinical catalysts and data presentations for the balance of 2021, and I'm sure we'll see you here again next year. Thank you.

Thank you, Yigal. Thanks everybody.

Thanks, everybody. Thanks, Yigal.

Thank you.