Good morning. We'll get started with our next company, NGM Biopharmaceuticals, where we have the CEO, David Woodhouse and CFO, Siobhan Mangini. Thanks to the team and the audience there for being part of the B. Riley Oncology Conference. This is Mayank Mamtani, one of the senior biotech analysts on the Health First team. At any point during the fireside chat, at mmamtani@brileyfin.com. Team, great to have you with us today. 2022, to say the least, was an eventful year, or I should say the last 18 months to 24 months have been transformative in terms of focusing the pipeline and corporate priorities towards solid tumor oncology portfolio, which is what we're going to focus on today. Maybe, you know, with that backdrop, David, overview of, both product oncology, but also some of the efforts you've had underway, and then we can dive into the specific Q&A. Yeah, sure. Thanks for inviting us today, Mayank. We appreciate it. Yeah. The maybe just to level set, remind people, NGM is a biologics discovery platform. Our platform really uses an interrogation of biology to undertake target discovery. We have particular expertise, we think, in-house around receptor- ligand deorphanization and just sort of diving further into biology that may not be fully understood. Really, the secret sauce of our platform is around pairing that with our biologics and protein engineering development. We're capable of all forms of antibody development, in terms of the size of the antibody, but also in bispecifics, for instance, and sort of all that new technology that's available in biologics. We use those first as tools to help interrogate the biology, but then ultimately optimize it. What's the right way to modulate that biology? That's been successful over the years for us. We've put a number of programs into the clinic, really our strategy is to get to that proof of concept study in the clinic and then evaluate, you know, whether to take the program forward, whether to take it forward on our own or with a partner. We've got a fairly broad pipeline. As you mentioned, we're focusing in our development efforts on solid tumor oncology, that also leaves a number of assets that we're continuing to think about just in terms of taking it forward with partners. Maybe on, you know, that last point of sort of focus that you have, just from your experience working with, you know, the programs that you have advanced in the past, not just in oncology, but broadly in GA and NASH, what have been the learnings from a platform standpoint, you know, how you've identified some of these, you know, very challenging targets historically part of and thinking of obviously FGF21 and obviously complement. You know, you try to at least apply your protein engineering capabilities to, you know, develop a sort of a antibody or a biologics approach that makes sense. What learned from that as you try to fine-tune some of the current pipeline that you have, David? Yeah. Yeah, it's a great question. It's something we have spent time on, the sort of lessons learned from these various programs. You know, well, first of all, as a level set, I think everybody understands, translational medicine is really tricky, and that's what we're focused on a lot of these cases. You have to take learnings from, you know, the biology side, often working in in vitro or in vivo models, and then, trying to predict what's going to happen in humans, which is a challenge. That's something that's well known. We've tried to and continue to emphasize finding these targets and having some sense of the human biology behind it, whether it's genetics based or whether it's an observation from clinical treatment to begin with, like gastric bypass surgery. We continue to think that's an important way to get some clues about what will work in humans, and putting as much effort as we can in that discovery side into that. The other aspect of it that we're really focusing in on based on these learnings is patient selection once we go into the clinic. We've really had a strategy of speed of enrollment and speed just in conduct of all of our activities. We think there could be benefit in hindsight in terms of really selecting maybe a subset of patients or kind of quote-unquote "higher quality patients" as we've moved into these in some cases to minimize variation, but also just in terms of giving the molecule the best chance to win, particularly in these early stage clinical development. The final lessons learned that I mentioned just a second ago is in several of these studies, there's been the primary endpoint measurement has been a bit of a challenge. Going into these studies when you're really putting a lot of pressure on a particular measurement, whether it's a biopsy, whether it's a retinal scan, really understanding how best to gather the most precise information off of that. Those are, you know, things that actually can apply to oncology trials. For instance, you have CT scans often in measuring these tumors. Certainly patient quality matters a lot in this setting as well. Those are just a few examples of what we're applying going forward. Yeah, definitely, you know, principles with which, you know, studies are executed in cancer, those are absolutely important attributes, patient enrichment, and sort of speed and scale and access to patients. You know, before we get into some of the studies that you're running, the early stage phase I studies, for a number of your different molecules, maybe at a higher level, if you can discuss, sort of your broader portfolio of myeloid checkpoint inhibitors, you know, as a class MDSCs, what makes it interesting in your mind? We obviously know tumor microenvironment sort of is a interesting challenge to work around, specifically with cold tumors. I'm going to assume some of these tumors are sort of initially your target tumor types that you'd look to try to get some activity either as monotherapy and combination. Maybe just at the highest level, if you could hear your strategy on the myeloid checkpoint drug class, and then we can dive into the specific agents. Yeah, sure. Well, let's just start with, you know, the first word in that target, which is myeloid cell. Why are we going after myeloid cells? Well, when we started deploying our platform in oncology, we were certainly inspired by what PD-1 inhibitors have been able to do, and other T-cell checkpoint inhibitors. Yet we felt we'd be late to the game if we just came to try and target some of these other T-cell checkpoint inhibitors. What was known at the time, and continues to be, interesting to us is, that T-cell checkpoint inhibitor obviously only works on that class of cells, the effector cells, or the cells that are engaged to come and attack the tumor. What's pretty well known is that unfortunately, a lot of patients don't respond to T-cell checkpoint inhibitors, or they become refractory over time, and their tumor continues to progress. What got us interested is it's not enough for many patients to simply take the brakes off of the T-cell and remove the inhibition that is created that cell type for attacking the cell. What's also existing in the tumor microenvironment are a number of these myeloid cells, these macrophages and other lineages from the myeloid lineage. What they create is a suppressive cytokine environment around the tumor that's really sort of a signal to the immune system that this is a tissue that should not be attacked. Normally, quite a healthy balance of immune system when deployed properly. In the instance of several of these tumor types, it becomes another block for the immune system. While you can take the T-cell checkpoint off, T-cell gets suppressed as it goes to attack the tumor because of this tumor microenvironment created by the myeloid cells. What we asked the question about seven years ago, what are the key receptors on the myeloid cells that are present in that tumor microenvironment? It's actually a fairly large percentage of the cell volume in a solid tumor is actually myeloid cells. They're present, they're just doing the wrong thing. What we think we've honed in on are a collection of receptors that govern that cytokine profile, that environment that's created around that tumor, to try and make it inflammatory rather than suppressive, and really work in tandem with T-cell checkpoint inhibitors to take brakes off of both axes of the immune system. That's the hypothesis by which we're following. We've got plenty of in vitro data to suggest that, and we're super excited to be in the clinic and see if we can translate this across the three programs we have in this myeloid immuno-oncology platform or portfolio. Right. Maybe just a good segue into talking a little bit about those targets and the uniqueness of, say, ILT2, ILT4, targeting versus maybe doing it as a standalone and then also, you know, the ILT3 work and the LAIR work you guys are doing. Yeah. Let me start with ILT2 and ILT4. LAIR-1 and ILT3 are sort of a separate but related story. Our NGM707 program targets two receptors, ILT2 and ILT4. The reason we took that approach is because the ILT2 and ILT4 receptors are very closely related. In fact, our monkeys have one version, and between monkeys and humans, there's a chromosomal duplication that created two versions that in humans is ILT2 and ILT4. The function of the receptors appears very similar. The expression pattern on myeloid cells appears very similar. There's redundancy within these myeloid cells that they tend to both express ILT2 and ILT4. There's some other interesting aspects of ILT2, but let's just stick with the redundancy hypothesis, which is, seems that you'd be able to either capture more patients' resistance mechanisms, or within a patient, have a deeper overcoming of that resistance mechanism by hitting both receptors. That's been our strategy with NGM707. We're the only company in the clinic that's hitting both receptors at once, as opposed to a couple of companies that are going after ILT4 only. The reason that we're interested in these receptors to begin with, just to tie back to what I was explaining after your first question is, both of these receptors are overexpressed in patients that exhibit this phenotype of resistance to anti-PD-1 inhibitors. There's evidence that these receptors are important to driving that, not just the observations in the clinic through tumor biopsies, but also, when you look at the cytokine profile that when active, these receptors drive that suppressive cytokine environment. When you block them, it takes away and turn that environment and turns it into more of an inflammatory cytokine environment. That's the thesis here clinically is, we think that treating patients in combination with an anti-PD-1 in the clinic, we're using pembrolizumab, will allow us to remove the checkpoint that exists from both the T-cell checkpoint with pembrolizumab and then hopefully myeloid checkpoint with NGM707 and drive greater response. On the LAIR side of story, the ILT3. That NGM438 and NGM831 are antibodies we have directed against LAIR-1 and ILT3 specifically. Both of these programs are oriented around what we term a stromal checkpoint. Both LAIR-1 and ILT3 pair with ligands that are involved in the extracellular matrix. Collagen in the example of LAIR-1 and fibronectin in the example of ILT3. Both the collagen and fibronectin, of course, are part of the ECM that shrouds many solid tumor types, and in fact, many of the more difficult tumor types to treat. Initially, it was thought that that just creates sort of a physical barrier to the immune system. Increasingly, and we have some really interesting histology that shows that, it appears that it's not so much the physical barrier, but as much sort of a signaling barrier that these macrophages may climb towards the tumor, but then when they hit the extracellular matrix, they become dormant or suppressed in some way. So, our research has suggested that it's actually these receptors, LAIR-1 and ILT3, that help drive that suppression of macrophages rather than infiltration and inflammation. We think by blocking that, we can overcome this stromal checkpoint, and it's effectively the same clinical thesis, which is, let's treat patients in combination with an anti-PD-1. Again, in this case, it's pembrolizumab with both programs. That will take care of the T-cell checkpoint piece and then use these programs to test the hypothesis of the stromal checkpoint. There is of course a thesis of combining 831 and 438 in that some patients likely use both resistance mechanisms or their tumors do. It may actually, you could think about a triplet down the road. For now, we're testing them independently to see sort of what tumor types if we can identify what tumor types particularly use this resistance mechanism, and then that will help us define a development path from there. David, this discussion is also sort of topical after, you know, we had a GI cancer sort of panel earlier today and obviously the ASCO GI conference we're kick-starting today. Where, you know, a number of these heavily refractory lapsed colorectal cancers or even like earlier lines of metastatic pancreatic cancer kind of have that expression profile that you're talking about. You know, there was a discussion on that, you know, many redundancies in the pathways, but also that, you know, extracellular matrix being so dense that, you know, you're not getting that penetration. Maybe just talking about the story in the clinic, you know, with starting with maybe seven oh seven where it looks like we did have some initial data at ESMO IO. Do you wanna touch on that? Sure. Just in terms of sequence where we are with the three programs. NGM707, we did put into the clinic first. We've been through monotherapy dose escalation and are in the process of doing the combination dose escalation with pembrolizumab that I referenced a moment ago. What we shared at ESMO IO was the monotherapy data, and Siobhan's gonna put it up on the screen now. Which is, of course, the 1A monotherapy dose escalation is primarily done for safety reasons. Fortunately, we were able to go up to the highest dose we tested without any dose-limiting toxicities. That's 1,800 milligrams. Of course, that's important because when you're inhibiting two receptors at the same time, what we think is a, is a fairly important checkpoint within the immune system, you do run the risk that you push it too far, and you come up with safety issues. We're pleased that we've been able to safely, you know, block both these receptors in these late-stage cancer patients and not observe safety signals. Of course, these early studies are also an opportunity to look for evidence of the type of activity you're looking for from an efficacy point of view as well. Maybe just to set some context, I mentioned some mono-specific ILT4 agents that have entered the clinic ahead of us. We have some context of the type of safety and efficacy that's been observed with only hitting ILT4. Also those agents have proven to be safe, but they haven't shown much monotherapy activity, which isn't too surprising because, you know, the myeloid axis by itself is really governing the cytokine environment, but not so much the tumor destruction aspect of the immune system. We had modest expectations for monotherapy activity going to this small phase Ia study. You know, while small patient numbers, we're encouraged what we've seen so far. We had 24 response evaluable patients involved in the trial. We saw one fairly deep partial response, as you can see on the waterfall chart. We had six patients that showed tumor reduction of some form, with a few others that had stable disease. It's, you know, we think encouraging as monotherapy activity. The difference between monospecific ILT4 and a dual ILT2/ILT4 is that ILT2 is actually expressed on some effector cells. It's expressed on certain B cells, natural killer cells, for instance. We think that it's possible, as we study this drug more, we might be able to tease out some of the monotherapy activity that could be driven from the ILT2 inhibition. What's shown on the right panel of this slide is the sort of translational medicine type of approach I was describing in our platform generally, which is looking for evidence that we can connect the dots to what we've seen preclinically into patients. We've been able to biopsy some patients baseline end of treatment and try and observe whether the phenotypic cell changes are happening on treatment. Specifically, are we getting this transition in the macrophage phenotype to more of an inflammatory phenotype compared to a suppressive one? This marker CD163 is what we're using to track that. You can see that we have, you know, some preliminary evidence here of observing in patients that that's happening. It's something we'll continue to track. If, you know, again, we're encouraged what we're seeing with the monotherapy, just to remind everyone, the strategy here is to do this in combination with the T-cell checkpoint inhibitor. What we're hoping, this sets a nice foundation for activity that really can be, if not additive, synergistic with a T-cell checkpoint inhibitor. understood. Just thinking of this program sort of going forward, David, you know, you how are you sort of doing on the exploring this kind of three-fold therapeutic hypothesis that you have your expand and it go forwards? Have you gone up to a higher dose with colorectal, for example, like you have for pancreatic? Or is it more like a sort of a scramble of patients, and you're trying to tease out, you know, wherever there is a risk? If you could talk about that. Yeah. I think I followed you there. You're breaking up a little bit. Yeah, I mean, it's one of the challenges of these early-stage studies is it is sort of a basket of different tumor types. It's frankly easier to enroll some of the tougher tumor types, as you're referencing, colorectal and pancreatic. We do get, you know, a predominance of those types in these early studies. We have been able to take this up to 1,800 milligrams, which is our highest dose tested. We've got some selection of that. We don't, given the size and scope of this trial, we're really using the 1A and 1B as a signal-seeking exercise to help inform where we may wanna take it into expansion cohorts and sort of these, you know, 20-ish patient, you know, earlier phase I tumor type of e-environment. The three therapeutic hypothesis you referenced is really, you know, if you think about broadly, you can take into three buckets of tumor types. One is you're referencing is sort of the cold tumor types. These, colorectal cancer is probably the, unfortunately, the most well-known or the largest tumor type, which is known to be refractory to T-cell checkpoint inhibitors. There's something extra you need there. That, that's probably the highest bar that we could reach for, but the easiest way to differentiate our approach because there's really no application for T-cell checkpoint inhibitors there. Then you sort of have the other extreme, which is the quote, unquote, "hot tumors," which, you know, for various reasons respond really well to T-cell checkpoint inhibitors. Can you boost that or have some of the patients that become refractory respond to treatment? That's probably a little easier given that you may not need as much of a tip of the immune system to drive that effect. Of course, showing differentiation will take more patience, differentiation from T-cell checkpoint inhibitors alone. Then you have certain tumors in between which may be sort of warm and that, depending on the mutational burden or other aspects of it, that they may be, you know, good candidates for treatment. So those are the three we're exploring. Like I said, what we're hoping is to be able to follow signals we see in these early stage trials to not be doing so much of a guessing game. Then, you know, run very disciplined small cohorts to see if we can see signals and move forward from there. I apologize, you know, my internet here in North Bay is, maybe because of the storm is, not the best today. If you are following me, David, just quickly on the sort of timeline, if you're guided on how you intend to disclose data for some of these combination cohorts, then just broadly for your portfolio and for the ILT space, you know, what are sort of things you are tracking? I know we've seen, you know, some data from larger companies and smaller companies over the course of last. Seems like a, you know, the tip of the iceberg. There's a lot more we have to learn still. What are sort of things to focus on in 2023 from NGM and also externally would be great to hear? Yeah, sure. You know, first of all, we're actually not guiding on timing for our oncology portfolio. We have tried to resist the temptation to do so to arbitrarily set a time point because, you know, you have to not only enroll these patients but also follow them over time. Particularly in this environment, you know, we've heard pretty clearly that it's better to have a more complete data set than dribble it out in medical meetings. That doesn't mean we're not pushing, yeah, as hard as we can and having success in enrolling all these trials. It's just, we wanna make sure we have a really as robust a data set as possible before we disclose. That comes to the second part of your question, which is, well, what happens in the interim? Merck has a mono-specific ILT4 that they've been working on for a while now and appear to have a very broad development strategy across a number of tumor types, co-formulation with pembro. You know, this is all based on ClinicalTrials.gov. We don't have any particular insight. They may disclose data at some point that's an update. It's been over two years since they've given a clinical update on the progress, and so that could be interesting. The other thing we're watching is there's actually a ILT2 mono-specific that's in development in the clinic that Sanofi is developing in collaboration with a company called Beyond Therapeutics. So that'll be kind of interesting to see whatever they can produce with a mono-specific ILT2 because we know there seems to be some basal activity with a mono-specific ILT4. If Sanofi and Beyond can demonstrate some with ILT2, you know, logically, when we're hitting both, we might be able to pair those two activities together. Those are the types of things we're looking for externally. We have no idea whether there will be... I don't think either company has guided on timing. Got it. that it could come out. Got it. Maybe even taking a step back on, you do have, I think, a heme malignancy program also, a bispecific, if you could touch on that. By the way, also maybe give us a little bit of a quick update on other parts of your pipeline. You still have the, you know, the NASH program I think running in the F4 setting. Just in terms of milestones for NGM this year broadly, then I also wanna pull in Siobhan for a minute about talking about the, you know, the balance sheet and the cash runway that you said. Okay, great. Go ahead, David. I'll be quick. Yeah. NGM936 is a bit of an offshoot from our NGM831 program, and that it also is an antibody direct against ILT3, but it's a different strategy. ILT3 plays a role in overexpress in the tumor microenvironment of solid tumors to act as a stromal checkpoint I was referencing. Interestingly, it's become dysregulated in certain hematologic oncology indications such as AML, and we think multiple myeloma as well. In that instance, we're using it as a tag with NGM936 because it's a bispecific. It binds ILT3, but it also binds CD3. The strategy is to be a T-cell engager and bring and activate T-cells in proximity to these dysregulated AML or MM cells. We do not plan to develop an expertise in heme, NGM. This is a program, as we referenced in our press release earlier this year, that we're seeking partners who do have that expertise to help us develop it. In terms of our other programs, we have ALPINE 4, which is our aldafermin cirrhotic NASH trial reading out in the second quarter of this year. We're going to use that trial result in combination with a whole bunch of other data we've established over the years with this program to seek partners. We're continuing to investigate our findings from CATALINA trial on NGM621 and see if there's a path forward with a partner with that program in geographic atrophy. Siobhan, do you wanna hit the nail? Yeah, maybe I can build off of that, in terms of the cash runway. What we've been doing, and we talked about in our press release this year, is actually getting much more tailored and focused on what we're developing in the clinic. That's in solid tumors, the programs that David's been talking about today. Being much clearer operationally that we are not developing ourselves in programs like NASH, Aldafermin, Hem-Onc, so that was NGM936, as well as in geographic atrophy, which is NGM621. We're looking for partners to outlicense those programs. That enables us to have cash runway into the fourth quarter of 2024. In terms of, I think, just what that means for our R&D expenses is you will see a decline in our R&D expenses as we're focusing the four programs that are in the clinic today, all in solid tumor oncology, and, you know, really focusing on partnering these other programs that are effectively non-core to NGM's development capabilities, but coming out of the discovery engine. That's helpful, Siobhan. That does mean, you know, a tailor down in R&D expenses and a sort of a more focused operating structure. Mm-hmm. Is that fair as people think to model? That's fair. 2023 versus 2024. Okay. Got it. I think with that, we've run out of time. Looking forward to the updates, you know, non-oncology and oncology, throughout the course of this year. Again, thank you for being part of this, and I appreciate the audience tuning in also. Well, thank you. Thank you.