Good morning, everyone. Day one of the Jefferies Healthcare Conference. At least not. Well, there was the private day. I keep forgetting that, but today is for all the public companies. Day one, slot one, I believe. Exactly. That, too. Yeah. So that's how it goes. But my name's Akash Tewari. I cover pharma and biotech companies here at Jefferies. I have the pleasure of hosting Third Harmonic. Natalie, it's so great to have you join. Why don't you give a brief update about Third, the disease areas where it's going for, and maybe any opening remarks that would be helpful? Absolutely. Thank you, Akash and Jefferies. It's wonderful to be here. Third Harmonic is a clinical-stage company, we're proud to say again, focused on KIT inhibition for the treatment of mast cell-mediated diseases. We recently announced that our IND is cleared for our development candidate, THB-335, and we are underway in our phase I SAD/MAD healthy volunteer study. We are intending to pursue chronic spontaneous urticaria as our initial indication, but when you look at the role of mast cells in inflammatory disease, we see that they are a central effector in all of the tissues where they're resonant: the skin, the airways, the gut. And so as we look out over the future of the company and as we think about indication expansion and franchise expansion, we really see the potential for a pipeline and a target approach to a disease area where there's still significant unmet need. That's very helpful. And actually, you know, with KIT inhibition, we obviously have the mAb approaches, companies like Celldex and others, that I think have taught us some of the science and the therapeutic window of kind of these next-gen KIT inhibitors. And then we have a lot of players in the oral space. I have to say, with your recent announcement a couple of weeks ago, it seems like your team is actually in the lead when it comes to the development of oral KITs. How important is that time advantage in your mind? Yeah. Um- I think when you look at urticaria as well. To the bottom line is, I think it is important, but certainly not essential and not what we think about most. Because when you look at the initial indication, CSU, there is one approved product for post-antihistamine patients, and that's Xolair. And it only penetrates about 15% of the market by Novartis's own estimates. And so what we see is that CSU as a disease is really in need of market development. Mm-hmm. There need to be new entrants to the space to drive awareness, diagnosis, treatment. And so the fact that there are multiple KIT inhibitors in development, I think actually inures to our benefit, certainly in the near term. It raises awareness for the target, it raises interest for participating in clinical studies, which I think will actually make it easier, at least in the early days, to advance the program efficiently. That being said, we are pleased to be in the lead, and we certainly intend to press our advantage and move as efficiently as possible through development. We can talk about our plans around that. But again, I think more market entrants is the rising tide floats all boats in this space. That's a really good point, and I agree. I think probably at this stage, everyone looks at things in a very binary manner. But let's start with something even more simple. We've seen really good efficacy with the KIT inhibitors, especially post-Xolair. Mm-hmm. There is a potential, I think, to maybe move up, lines of care as well. But if you were to develop an oral KIT, and those were able to enter the market, from your own kind of patient work, how would that expand the pie in a way that maybe the long-acting mAbs would it? Yeah, great question. Well, we know from our own conversations with KOLs and treating physicians that the black box warning on Xolair is a particular commercial impediment in the dermatology space, in particular. And so we believe that oral entrants into the field is going to, first and foremost, drive the physician treating pool that treats urticaria. And when we look at the side effect profile of KIT inhibitors, the potential for anaphylaxis, which is again in the Xolair label, that's an antibody-mediated effect. It's something that you don't mechanistically run the risk of- Right ... with an oral approach. And so we think that a priori, that's going to make KIT inhibition a more accessible treatment modality for a lot of physicians who are currently passing their urticaria patients on to allergists in order to get treatment. Understood. Actually, maybe to that point, is there maybe a view internally about some of the long-acting mAbs that are hitting KIT right now? Do you feel like there's a potential when it comes to anaphylaxis being on the label to differentiate there, right? Because I think everyone understands Xolair has the black box. Yeah. Is there a view internally that that may also be the case for some of these long-acting mAbs? I think mechanistically, yes, any antibody that's targeting the mast cell runs the risk of paradoxical activation and anaphylaxis. I think it'll ultimately be a question of how the data mature. Right ... on the antibodies and how the label negotiations go at the end. But I think, again, you know, mechanistically, with an intracellular TKI approach to KIT inhibition, it's not an issue. Right. Now, maybe some other topics outside of just anaphylaxis with oral KITs. I mean, I think the other question is, do you need to hit KIT as a target that hard in order to have the same efficacy? You know, you got some data with your first-gen compound, THB-001. Can you talk about what you saw in that study? Yep. What makes you confident about the level of, you know, KIT inhibition you're hitting on an IC level that would get you the same efficacy as the mAbs, but maybe a better safety profile? ... I think ultimately, that question hasn't been answered. I mean, as you pointed out, we got a really tantalizing peek at it in the CIndU study with THB-001, our first-generation compound, where at our starting dose of 200 mg BID, which we considered our suboptimal dose, if you will, we saw four out of five patients respond in a truncated study. Some of those patients, most of those patients, in fact, didn't reach 12 weeks of dosing. And so to your point, it suggests that you might not need to obliterate KIT signaling systemically in order to drive optimal efficacy in urticaria. With an oral small molecule, you have the ability to more finely interrogate that therapeutic window, and potentially titrate into a sweet spot where you're driving optimal efficacy, but getting a little bit more of a margin on safety. That is certainly a really key part of our development strategy. We're keen to move into robust dose-ranging efficiently- Mm. - in our CSU study in order to answer that question. But certainly, preclinically, in the pharmacology models, you see that you might not need to switch to sort of biologic terms rather than sort of mechanistic terms. You might not need to actually deplete mast cells to drive optimal efficacy. Inhibition- Right. - of mast cells might be sufficient. Our data suggests that. So, the next thing to do will be to test that in the clinic. That being said, if that's not the case, and you do need to drive tryptase down to the lower limit of quantitation in order to get optimal efficacy, we know that we can do that. We've shown that with the first generation, and we've certainly shown that, we've tested that and shown it with the second generation, THB-335, as well, preclinically, and that's one of the key questions that we'll be answering in this phase I study. So, you know, we'll talk about 001 and some of the liver tox signals that were shown in the first gen compound. But, when we think about other ways that 335 can kinda differentiate, I know, in terms of food effect, formulation, QD, these are, you know, penetration in the brain and the testes. Can you talk about other ways in which 335 was optimized as a second-gen KIT program? Yeah, absolutely. So, as you talked about, it has been structurally modified to alter the metabolism, which we believe is going to mitigate the DILI risk, the drug-induced liver injury risk that we saw with THB-001. Beyond that, it is very pharmacologically similar. Similarly potent, similar, similarly selective, if anything, slightly more potent and slightly more selective, but certainly within the same ballpark- Right. ... as THB-001. It's peripherally restricted, as you pointed out, whereas 001 was broadly distributed. The brain isn't a target organ for these diseases, and generally speaking, if you don't need a drug in the brain, it's better not to have it there. And we were also optimizing to potentially get a bit more of a margin on the histologic effect in the testes. The PK and potency suggest that, as you pointed out, this is a once-daily dosing product from the get-go. With THB-001, we started with twice-daily dosing. Right. We're formulating our way into once-daily dosing. We will confirm that this is modeling at this point. So again, another one of the important readouts from this phase I study is, is this a once-daily drug now, or do we need to do some additional drug product development in order to get there? If we do, another one of the improvements in 335, very soluble versus 001, so it's gonna be easier to work with from a drug product perspective if we need to do that. Well, solubility, you brought that up. I think that's quite important, especially looking at the first-gen compound, where I think, you know, there was probably a need for some type of formulation work. Can you talk about by having more consistent, less solubility, maybe more consistent PK/PD, how that also helps you dose optimize when you're going into your human doses, right? I mean, 'cause again, if you don't have a food effect, if you don't have the PK variability, I think also getting to that optimal dose for each patient may also become easier. Do you see any advantages with 335 versus 001 in that regard? Well, you know, the improved solubility does a couple of things. So number one, it makes it a more, an easier compound to work with from a CMC development perspective, which is an underappreciated and really important part of product development. The other thing that that allowed us to do was to really push the margins in our preclinical tox studies in a way that we couldn't with 001. In terms of improved solubility having an impact on PK variability, that's again one of the things that we're testing in the ongoing phase I study. And the food effect will be confirmed as well. I think it. There are a lot of factors that go into PK. It's not just potency or solubility. Sure. But I don't actually think that the variability that we saw with THB-001 was out of keeping with what you would expect. With small n's, human beings are complex creatures. You're gonna see some variability. Right. That's why drugs don't get approved on 5 subjects' worth of data. So I think if anything, you know, maybe slightly improved, but that wasn't a design element of what we were trying to solve for because we didn't think that the PK that we saw with 001 was problematic. Understood. Now, Natalie, you mentioned something which I think is quite important because I think a lot of investors, you know, the first-gen compound showed the unexpected DILI, and I think, you know, the first question I'm sure you get in all the meetings is: "Oh, okay, how do we know the next one won't?" And I do think it's notable. You know, it was kind of the end of last year or maybe two years ago, where you said: "Hey, we've done some work. We've identified glutathione. We've understood what happened with 001." But it's not like 335 immediately was in the clinic. Your team did a lot of work pre-clinically. Absolutely. Last year to make sure that from a liver tox perspective, this next gen compound was kind of de-risked. You know, talk to me about the work you did with this compound that was, you know, that makes you more comfortable on the DILI risk. And you hit on something really important with the improved solubility. You're able to maybe explore a therapeutic window with 335 in a way that you didn't with 001. I think that's really important. Tell me about that. Yeah. So the mechanistic work that we did in the first half of last year pointed to this metabolic liability with THB-001. One of the major metabolic pathways that ends in the formation of a diol was generating a reactive epoxide intermediate. And as you pointed out, when you generate a reactive intermediate, the body tries to detoxify it. You see an increase in glutathione adducts, and to the extent that the sort of epoxide generation overwhelms that system, you start to see signs of oxidative stress. Since we had this very clear structural path between the major metabolite and the site of metabolism on the parent, we were able to select for a molecule in the next gen program that functionally block that site of metabolism. And so in testing 335, as you point out, I didn't need investors. They asked to ask the question. I was asking the question of my team every day: How do we know we fixed this? We very thoroughly interrogated the metabolism of 335, above and beyond sort of the IND-enabling ADME package by looking at longer incubation periods, looking in multiple species, looking at the metabolic profile, looking for the presence of glutathione adducts, which we very clearly saw with 001, and we don't see it. As you would expect, when you look at mass spec traces of MetID in human systems, you see a very different metabolic profile between the two molecules. Then the other thing that we did was independent of the mechanistic metabolic work, we set up some phenotypic screens- Mm. in human hepatocellular culture systems. We basically sort of little 3D liver systems, and just looked phenotypically, when you put the drug in the culture, are you seeing... In this case, we were looking at transcriptomics. We were looking at upregulation of genes associated with detoxification and oxidative stress. And what we saw was that over time, we had to take these cultures out on the order of 21 days. You very clearly see an upregulation of sort of the toxicity signals with 001. You don't see it with the next generation. And so we have this nice mechanistically agnostic phenotypic screen as sort of an orthogonal assay against the mechanistic, the metabolic work that we did. And so we're showing up to the starting line of going into the clinic with 335, much more de-risked from- Right ... a liver tox perspective because we have interrogated the hell out of it before we even get into humans. I think that investors, we certainly take heart in the fact that we put all this in front of the FDA, and they said, "We're convinced you're okay to proceed. Understood. That's really helpful. Now, two parts on that. When you think about, you know, you said, "Okay, looking at THB-001, there's stuff that we have to do with our GLP tox studies that might be different." But on your healthy volunteer study, are there any modifications you're making there? You know, is this similar to THB-001, or is it maybe longer dosing, higher dosing that investors should be paying attention to when we're, again, thinking about de-risking this program as- Yeah ... quickly as possible? Fair question. So in terms of the overall design of the phase I study, it's very similar to THB-001, single ascending dose, multiple ascending dose, food effect study in healthy volunteers. 14 days of dosing in the MAD. The liver signals did not show up until week 8 in the CIndU study, and so we wouldn't expect to see transaminitis in- Right ... these patients. That's me knocking on wood. But liver enzyme monitoring is standard in all phase I studies, so we didn't have to introduce any additional sort of monitoring in order to be able to pick this up. Now, that being said, one thing that we can do in a phase I study is look at the metabolic profile of the drug in plasma from these healthy volunteers that are taking the study. So just like, and this is data that we've shared recently, just like in the preclinical work, where we can look at a mass spec trace of the metabolic profile, we can look to make sure in the 335 phase I study, no diol- Right ... which would strongly suggest no epoxide and therefore no oxidative stress. So that's a key—that's probably the most sort of de-risking element of the data readout from a liver tox perspective. But there is a wealth of other really rich data that we get out of this study. We, of course, get to confirm the PK, we look at safety, and then we have the benefit of this incredibly powerful biomarker and serum tryptase. And so I would say we've got a way—we're already showing up significantly de-risked on the DILI. Right. We can confirm that by looking at the metabolic profile of the drug. And in all other aspects, what we learned from the THB-001 experience is that phase I is very predictive of what we're going to see in later stage development. So we actually get a lot of useful de-risking information out of this because we have this really well-characterized, powerful biomarker in serum tryptase. You bring that up. The tryptase is a really great point. I think for a lot of investors, the first question is: Okay, what level of tryptase reduction are they targeting that would translate to, you know, really knowing what the clinical profile of this drug is, you know, 2, 3 years down the road? Talk to us about that, right? What level of tryptase reduction would make you very happy? I think the answer is I would like to see a range of tryptase reduction in my phase I study so that I can study a range in my phase II, because the answer is nobody knows. Right. That's not a hypothesis that you can test with a long half-life antibody that you're dosing at really high doses in order to get the dosing level that you want. Sure. That is sort of a blunt instrument that is going to obliterate KIT signaling. So, the answer is we don't know. Honestly, we don't know. So in designing the phase II study, I think we're looking at, you know, anywhere from 40% to full tryptase reduction. That's kind of the window that we'd like to study. The 40% is, I'll just be honest, arbitrary. I think there are- Yeah ... a couple of hints- Yeah - in dataset data from other studies that suggest that might be sufficient. I'm thinking in particular of the Cahill paper from the New England Journal in 2017, where imatinib was used as a KIT inhibitor for the treatment of severe asthma. They saw in 60 subjects a P value on FEV1 and airway hyperreactivity at six months. It's a pretty astounding finding. I think that's why it ended up in the New England Journal. In looking at mast cells and tryptase in that study—Did I lose my microphone? No, I think we're back on. Yeah. It was about a 40% reduction. Noisy endpoint, small n- Right ... but, there are other data sets like that. I just call that one out anecdotally to suggest that it might be that you, as you said earlier, you can get by with a more modest level of systemic KIT inhibition and still get optimal efficacy. I wanna, I wanna know that if I need to go all the way to the LOQ, I can. Right. But it's really sort of the mid and lower end of that range that we're really interested in, in studying. Really helpful. Natalie, you mentioned you've done a lot of work preclinically de-risking the compound. Has that led to maybe your starting doses being a bit more aggressive in terms of getting into your therapeutic window? How did you think about dose selection in this study? Because, you know, A, it's very competitive. There's obviously a lot of other oral KITs in development, but at the same time- Yeah ... you wanna be as careful as possible. I think in a phase I study, the objective is you want. It's all about showing up to phase II with as much information as possible, and so short-changing your SAD by, you know, skipping a couple of dose levels. I don't think that actually buys you much. Yep ... overall. Your starting dose is set by your tox coverage, and then you just go up by factors of two until you hit an exposure ceiling or some sort of dose-limiting toxicity. Since this is an improved molecule from a solubility perspective, we expect we're actually gonna be able to go up above- Mm ... levels that we were able to achieve with 001, and therefore, have a broader window to study in the MAD. But you don't really do yourself any favors cutting corners in phase I. We have a very aggressive plan for getting to phase II efficiently, and when we do, we wanna run a robust, powered study that will allow us, you know, potentially to move directly into phase II phase III from a single phase II study. So we are keenly focused on moving as quickly as possible. Patients are waiting. This is a competitive field. Remind me, for that healthy volunteer study, when can we expect kind of a full update on that program? We've guided to the first half of 2025. And part of being able to sort of tighten up that window is just getting a look at the exposure range that we can achieve- Right ... which will tell us how many MAD cohorts we need to do in order to really fully characterize the window that we have to operate in. So, the work is ongoing. Study is enrolling. We'll know more as the study goes. Now, you know, I think everyone understands, urticaria and KIT inhibition. I think the question a lot of people have is like, okay, there's a lot of mast cell-mediated diseases. Then the question is, okay, well, what's the next indication? Yeah. That's where no one really has a great idea, right? I think we think about asthma, but you know, I don't think investors are really thinking about KIT inhibition in other targets. You have a unique opportunity here because you mentioned, you know, maybe in asthma, even lower KIT inhibition might have more of a clinical effect. So as you think about a compound like yours, where you're able to kind of explore the range of KIT inhibition, where, from a therapeutic perspective, would KIT inhibitors go next, right? Yeah. Ex urticaria. Yeah. We've talked—we've always talked about severe asthma being a disease target of interest. There's actually a lot of clinical validation for KIT inhibition in severe asthma. There's also been no new oral entrants to that market in over two decades. So a lot of new biologics, but there aren't—beyond the anti-leukotrienes, there are no orals. Mm. So we see a clear commercial white space in a big and growing market. So that's still very much on the table for us. As I mentioned at the top, we know that mast cells are resident in the skin, in the airways, and the gut, and so we think about disease targets in all three tissue compartments. As we think about our indication expansion strategy, there's definitely a second mover advantage here, in that a lot of these questions are being addressed by other companies. Right ... in the KIT space. And so I am all for the indication expansion that is happening in the KIT space more broadly because I learn from that. It's kind of. It's almost free data for me as other KIT companies look at EoE, at PN, at AD, at asthma. Sure. So we think, we think very strategically about the sequencing of our indication expansion, because if I've got data coming from another player in the space, I'm gonna wait for that because it's gonna inure to my benefit, and I'm gonna be able to design a study that is sort of better optimized for signal when I learn from what others are doing. Okay, I'm gonna sneak in one more question, even though we are running slightly out of time, but I think this is really important. You know, I think something you'll hear from investors with the mAb approaches, oh, you know, the oral KITs may not have the same stability on heme tox- Yeah ... than you get with the mAbs. And, you know, you mentioned, you know, these are early, and, you know, end studies, you shouldn't read too much, especially in healthy volunteer trials. So when it comes to the question of selectivity with oral KITs and what is enough for you, what makes you feel confident that, no, that's actually a concern that's not well-founded? Yeah. You will see a stability on these heme markers with oral approaches. Yeah, you know, we've had this discussion a couple of times, and it's a little bit of a head-scratcher for me, honestly. I just don't think the data support it- Right ... even at the small ends that we're currently at. And when you look at the selectivity profile of our compounds, you know, the one anti-target that you might worry about from a heme perspective is FLT3, because that's the compensatory pathway that kicks in in higher order species. I gotta wrap it up. So there is no activity against FLT3, and so I think the data don't support it. Mechanistically, it doesn't make sense. So, it's not one of the things that keeps me up at night. Very important. I wanted to sneak that in. Thank you so much, Natalie. Thank you. I really do appreciate it.