Great. Thanks, everyone, for continuing to join us on the Stifel Virtual Oncology Days. My name is Brad Canino, Senior Analyst here, and I'm happy to do the next fireside with Pyxis. I've got Lara Sullivan, the CEO, and her team is also on and may be joining for some parts of the Q&A. But Lara, thanks so much for agreeing to sit down on the fireside with us. Thank you for having us. We really appreciate the opportunity to share our story and give you the latest updates on what's going on with the company and the pipeline. I'd love to start with an introduction of the company. As I look at this space, to me, you definitely have the most unique focus when it comes to ADCs, a unique approach. Can we start there with a background on the company and then weave in some of the company's priorities for this year as well? Sure. So our company actually was born out of immuno-oncology IP from Thomas Gajewski's lab at the University of Chicago. I joined shortly after the founding in 2019, and it was clear to me that target stage I/O discovery sort of pipeline was going to be a tough road as the sole focus for the company in terms of meeting milestones and financing interest from the market, the conditions at the time. So we quickly began to look at modality expansion because we had a talented scientific team. And we started looking at ADCs at the end of 2019, early 2020, largely driven by the fact that a lot of ADCs have a local immunomodulatory effect. And so a lot of our scientific expertise would map very nicely onto the ADC modality. So we began to evaluate a variety of ADCs that were available for licensing at the end of 2019, early 2020, from a variety of folks in the Asian markets as well as the American markets. And the opportunity to bring in the entirety of Pfizer's preclinical ADC toolkit and pipeline came our way. I'm an ex-Pfizer alum. I founded and spun out SpringWorks Therapeutics. So I have experience with Pfizer spinning out technology. And we connected and brought the technology in, which was the entirety of the toolkit of linkers, payloads, and conjugation chemistries that all have been optimized over a decade-plus of years of time, as well as millions and millions and millions of dollars, which I had familiarity with from my time in Pfizer. So that really was a sort of critical component of the company history because that's where we really made the decision to become an ADC-dominant company. And it turned out to be very great timing because we completed the license the end of 2020. And the ADC market just really began to take off with a variety of acquisitions. VelosBio seems like ages ago now, but that was in the early heydays of the modality and in HER2 with sort of its juggernaut of approvals going along. So we have the technology toolkit in-house. Our lead program, PYX-201, is a novel target. It's a first-in-class, first-in-concept ADC targeting the stroma. So it's a novel approach with novel tumor biology. And we've continued to maintain a modest opportunistic presence in I/O in the background. So our secondary program targets Siglec-15. We have some earlier stage assets on the shelf available for licensing in the I/O modality as well. Really, the best way to think of us is as an ADC-focused company with a bit of a nod opportunistically to our I/O heritage. Yeah, great. Now, I'm going to ask for another history lesson. Specifically, why and how EDB was elucidated as a good target. And then since it's not a classic cancer cell surface target where we've been pulling most of the ADCs, why try to throw a concentrated blast of chemo at it? Yeah, so I think the history of EDB is really interesting. And it's one of those stories that just demonstrates the importance of following science and intellectual curiosity, regardless of where the dominant thinking is at the time. So an investigator named Dario Neri identified the target in Switzerland at his lab in the university. And EDB is a splice variant of fibronectin. It's fetally derived. The IHC analysis that we've done and some of the earlier work that Dario did, as well as Pfizer, who ended up licensing the target from Dario, demonstrates exquisite target specificity. So we see this target expressed quite prominently in a variety of solid tumor tissues and the companion normal tissue negligible expression. So the target expression gradient is pretty remarkable. So Dario had identified it, and I think he's always been a passionate investigator of the stroma. Ahead of his time, sort of focused on stroma, agnostic to modality. He also took some other sort of biological agents to address some of the needs in the stroma. Pfizer got intrigued by this research and licensed the target in the mid-2010s. That was when Pfizer was doing the bulk of its ADC work before the acquisition of Seagen. Pfizer began to progress the work. As you noted, it's a target in the stroma. It's not technically internalizing the way cell surface targets for ADCs classically are. The industry over the last 30 years has held this paradigm that for an ADC to work, it must be internalized because it must be cleaved inside the tumor cell. What Pfizer researchers demonstrated, which was actually in a paper published, I'm sorry, a poster published at AACR in 2014, that the nature of the proteases that cleave these ADCs intracellularly are also present in the tumor microenvironment. So the conditions for cleavage exist in the stroma. I think this has been underappreciated by the general medical community and scientific research community who've been very focused on the canonical idea that an ADC must be internalized. So Dario and Pfizer, I think, were ahead of their time. When Pfizer made the corporate decision to externalize its ADC technology in the 2018-2019 timeframe, this program was the major program that went across with all of that technology. We've continued to do preclinical and translational work on this target. We continue to see the stroma as a really sort of next stage of biology in the ADC field. And we're looking forward to sharing the clinical data in the fall this year that I think will validate a lot of these hypotheses that have been out there for over a decade around this type of mechanism of action, but which I think because ADCs have been such a hard modality historically to wrangle, you have to get your linker right, your payload right, your target right, your chemistry right. So you have four ways to technically fail before you get into the clinic. And when we saw Mylotarg approved in 2000 and the next ADC not approved until a decade later, that was driven by all those technical failures. People were not in the midst of all those technical failures, people were not willing to take target and mechanistic risk while they were still figuring out how to make an ADC work. Now, 30 years in, we've really figured out how ADCs work. I think it opens up a whole new avenue of targets and a whole new avenue mechanistically for ADCs. I think our clinical data in the fall is going to really kick off some serious interest and inquiry around that concept. I think generally for investors, ADCs have been seen as highly investable because they tend to have the ability to produce strong single-agent activity. What I want to ask is, do the preclinical data for PYX-201 suggest that targeting the stroma instead of the cancer cell directly could also produce compelling single-agent activity? Yes. So we shared a variety of our PDX models at AACR last week. And those posters are available on our website, which demonstrated pretty exciting and demonstrable single-agent activity in the preclinical setting. So again, sort of the target specificity, the mechanistic understanding, and the PDX models combined got us extremely enthusiastic about bringing this agent into the clinic where we think our dose escalation study that's ongoing right now will demonstrate efficacy signals when we give the data readout in the fall. So the preclinical models are indicative of such. And we're right in the midst of dose escalation, getting hopefully either in or at those therapeutic dose ranges. You never know really until you're at the end of the trial and have all the data and can see exactly where the high and low dose are. That data will come in the fall, and I think will demonstrate this hypothesis being validated. Yeah. And then the other side of that question is, does the tumor expression pattern that you alluded to allow for the potential of safe combinations with other therapies? Yeah. So I think one of the most exciting things we heard at the AACR Strategics Conference, which was two days before the official start of AACR, where a dozen big pharmas came in with their clinical heads and talked about their pipeline strategies in oncology, and nearly all of them, I think there was only one that didn't mention ADCs. So most of them mentioned ADCs, and most of them mentioned ADC plus I/O combos. And I think that the idea of the direct tumor killing and really ramping up the immune system to fight the tumor, especially locally, is a very compelling strategy. We saw that with the bladder cancer data last year, of course. And so there's a lot of excitement, enthusiasm about that. Then when I think about our agent, because we're coming at this from the stroma, and everyone else, regardless of modality, is coming at it from that sort of cell surface target perspective, we could have a really unique position as almost a universal combo agent. You could imagine combining us with I/O. You can imagine as we start to think, we in the industry start to think about ADC, ADC combos. Merck already has a name for them called DADs, double ADCs. We have a stromal targeting ADC. Everyone else is targeting the cell surface. You could imagine figuring out the right load of burden of payload for the patient with those two mechanistic approaches could be a really powerful approach. We definitely are eager to watch what else is going on in the combo space. I think as our dose escalation monotherapy trial reads out and helps us understand where the therapeutic dose range is, what the preferred monotherapy dose is, that will be a helpful guide into what a potential combination strategy could be here. And obviously, the preclinical models are important where we can test how our 201 agent partners are combined with a variety of other agents and then figure out what could make the most sense for the clinic. Yeah. What monotherapy side effect profile should be anticipated? What comes from the chemo payload and ADC construct, and what comes from targeting EDB? Yeah, so this is really when you have a new target and new biology, it's almost like every week is a great new learning experience, right? Because you're contributing to your knowledge of understanding what the target does as well as how the mechanism works. And so with such a clean target, with such a specific target, and we found pretty minimal in our preclinical models target-specific toxicity concerns, our main sort of observation area for our agent is around those ADC or payload-specific type of toxicity events. So we really keep a close eye for neutropenias, neuropathies, ocular tox. And from a target-specific perspective, the clinic is teaching us, are there any potential patterns that we might observe as we go through dose escalation that suggests a target-specific mediated response? So that's a real-time learning. There's not much precedent out there from a target-specific safety profile that we have at our disposal other than our preclinical models. And again, we all know the limitations, benefits, and limitations of preclinical models. So I would say overall for us, keeping an eye on the platform, the ADC-specific toxicity is our main focus and worry area. And if we observe any sort of target-specific patterns that come out of the dosing regimen as we continue to advance it, then obviously we'll figure out how to manage those and get a better understanding of what those could look like. Okay. And now can you discuss the scope of the ongoing phase one clinical trial and any important elements that you think investors should know? Sure. So we started dosing first quarter of last year. We're a year and a month or so into the trial. We announced in March with our 10-K that we had already dosed 37 patients. The trial is sized for 45. We can typically go up about 10% or 20% over that number without too much fuss. So I would say probably the full data readout will be somewhere between 45-55 patients. And we'll be prepared to share the full safety, efficacy, and PK data this fall, which is probably sometime September-October timeframe. The trial is designed as a basket trial. So 10 tumor types were identified to participate in the trial. And you don't often see those kinds of designs in small biotechs. The reason we chose to do that is because we showed in those PDX models responsiveness, sensitivity across a variety of tumor types. When you look at our target specificity or target expression and the companion specificity across multiple tumor types, again, you see a real strong message around target expression. So we didn't want to leave tumor types behind in the dose escalation trial. So we decided to go with the basket of 10 tumor types. We have dose escalated now through the 5.4 mg per k cohort. We got the approval from the Dose Escalation Committee, which is the group of investigators that evaluate safety, in March, to go up to 8 mg per k. So we're now getting to dose levels that are very exciting and somewhat unprecedented in these types of payloads, these microtubule inhibitors. I think that that's where the mechanistic understanding starts to come into play. Is there something about the drugging the stroma versus cell internalization? Cell surface targets with internalization that's allowing us to have a higher dose range here. We're very excited to see how our eight mg per k dose behaves during this escalation. Excuse me. At this point in the trial, when you're four or five months out from data, you're starting to get to the top of your escalation or close to it. The data that we're going to amass over the next few months is going to be very telling for us. Is eight the max dose? Is it an intermediate dose? We can escalate further. You start to do some of those titrations at this stage. The last point I guess I would raise about the trial design is, as I mentioned, there were 10 tumor types that were identified to participate in the Phase 1. We also announced in March that we were enriching from this point forward at these higher dose levels for patients with pancreatic, head and neck, ovarian, and lung. The decision to do that is driven by the fact that, again, we look at our trusty PDX models, we look at our understanding of the target, we look at the marketplace, current and emerging standard of care, and we also incorporate real-time clinical feedback within the trial. We wanted to make sure that we had a high enough N in particular tumor types so that as we're in that therapeutic dose range, it guides us to making good decisions about cohort expansion from there. So it doesn't sort of help us figure out what to do if we have 10 tumor types at each dose level, 1 patient at each dose level. So eventually, I would love to have the chance to, once we have that defined dose level, I would love to have the chance to go back and potentially pick up those other 6 tumor types and test them again at that higher dose level. But to make the data meaningful and decisionable and actionable for us and the investor community, we felt with the consultation of the investigators working on the trial with us that enriching for these 4 tumor types would be the best way to really understand the potential power of this agent. Yeah, right. And now you alluded to some of this in your previous comments, but the March update also came with a decision to push out the public data release to investors. Anything else you would say about what led to that decision? Yeah, thanks for asking that because I think it's always helpful to have the opportunity to explain those kinds of things. So our trial operationally has been operating extremely well. We have waiting lists at every site to participate in this trial. If a patient's screen fails, doesn't meet the inclusion criteria, there's another patient to replace it like that. So when a timeline adjustment happens in a program like ours, the reason for that is related to the ability to continue to dose escalate to higher doses than what we might have originally expected. So our adjustment of the timeline reflects the fact that the program has been able to dose escalate to eight. And we might see we can dose escalate beyond that. When we did our initial trial planning, including our timelines, we looked at the Pfizer HER2 clinical asset, which went through Phase 1 and has the same linker and same payload as our agent. We looked at their dose escalation curve, we looked at their max dose, and we used that to model our planning. Two things that are important and now are playing into this. HER2 is a different target than EDB. It has its own toxicity and safety profile. It's a cell surface target. Mechanistically, it's different than ours. Our trial planning with our initial timeline assumed a max dose of 5. The way Pfizer capped out at 5, they actually had a 50% discontinuation rate at 5. Their real dose was somewhere between 3-4. Here we are where we announced in March that our 5.4 mg per k dose with 10 patients was very well tolerated and only had 6% grade three AEs through that 5.4 cohort. So right away, you see that the timeline now is driven by the agent itself telling us that it can go higher into escalation, which just takes some extra time. And so we made the decision to communicate that to the investment community in March as soon as we were made aware of the ability to do that coming out of the Dose Escalation Committee meeting. We had that meeting mid-March, and we made the announcement to the investment community a day or two later. And if it had not been tolerated at the 5.4, that would have been our max dose. And we would have kept the same timelines for earlier in the year. So this is a net good news event for this agent because the ability to have such a clean target with tolerability being the way it has been through 5.4 bodes very well for the patients to be able to have continued drug exposure and ultimately have potentially efficacious responses out of it. So it's a little bit nuanced, and I appreciate the chance to explain that to the investment community. Yeah. I think the question that arises from that is in regard to the good safeties, well, does this pose a risk that the payload is actually not undergoing that cleavage extracellularly? Because without any efficacy disclosure for context, it's hard to know that you're getting that cleavage and payload diffusion occurring within the tumor. How would you address that question and concern at this stage? Yeah, so one of the things we're evaluating very closely are the PK profiles of our patients. And again, that data comes in in batches with each cohort. We have not disclosed that yet. We will disclose that with the data in the fall. And I think what the investment community will see through the PK data and the clinical data and the understanding of the IHC analysis and the mechanism that there's a lot of great reasons to believe that the payload and the ADC is behaving just as one would expect. And I appreciate that's probably a missing piece of the puzzle for the investment community here. As a general principle, we prefer to when you've got an in-process dose escalation going on, we prefer to give a fuller picture than interim looks because I think we've all seen in many of our peer companies when they've disclosed data in the middle of an escalation, either it's people are holding them to full data explanations or announcements, and they're not reaching it because they're just not there, and that's problematic. Or they say something and then it changes before the full data readout, neither of which is good for the investment community. So I think that when all said and done and all the data is disclosed, the clear linkage between the mechanism, the behavior of the payload, and the clinical data will be seen. And between now and then, we're excited about continuing to treat patients at these dose levels that are beyond our wildest expectations. Yeah. Okay. Now, the phase 1 is a little complicated. There are a lot of tumor types. There are a lot of doses. So when you do update the data in the fall, how do you hope to define activity with all that going on in the study? Is it important to have a clear ORR from this initial readout of this specific experiment? How are you thinking about that? Yeah, so I think this is one of the challenges of just also small biotechs doing phase one studies compared to pharmas, right? Because we sort of live and die on these phase ones, and we all have to make decisions on very small ends and information that's generated while we're learning. And it's harder when you're learning for a new agent with a new mechanism than, say, the 92nd HER2 ADC that somebody just tweaked a linker or something, right? Those guys know what those trials are going to look like. They can have a very specific ORR or disease control rate to shoot for, and they have a very specific population. So we know we're a little more complicated for folks. And I personally believe that's where opportunities for asymmetrical upside come, right? Because if you do the work and you really understand and believe in the hypothesis, the rewards can come. So for us, what we've done is we have defined for each of the 10 tumor types what the response rate, aspirational response rate would look like in light of merging and current standard of care. Now, because we've dosed 37 patients already, we may dose up to, say, 55. We have 10-15 patients plus or minus left across four tumor types. I think all of us with our statisticians had it on would say, well, can you really calculate a response rate from that? And I'm sure we in the investment community will do that math. But it's not going to be a statistically significant response rate. If it's an N of four patients in pancreatic versus an N of eight in ovarian. But it's going to be informative for sure because when you see signs of efficacy like PRs or significant chunks of stable disease that contribute to your understanding that, hey, this drug has biological activity. It's telling us where it wants us to take it. I think the mandate for us is going to be to follow that signal into cohort expansion. So we're not going to be so rigid that if we had, say, four patients in pancreatic and the response desired, well, if you have one even out of four, you've already beat the current standard of care. So that's a bad example. Let's say out of lung, right? If we had four lung and we had one, I don't think we would be irresponsible to say it's only a 25% response rate. Therefore, it's not going to be competitive. I think what we do is really understand that patient vignette, what's going on with the other 3 lung patients, and is that worthy of a cohort expansion? And so I suspect some of these tumor types are going to be more definitive in informing our decision-making if their N is higher. But even the ones with a lower N, I think, are going to be actionable for us to make some determination around what to do about that. And so the next stage of the study that we would announce as part of the data release will be around the dose optimization and selection. And depending on the escalation goes, sometimes you can accomplish that within the dose escalation if you're titrating up and down. FDA might say, okay, that did satisfy Project Optimus, but now with Optimus, they usually like to see a dedicated cohort on that. So we're assuming the plan for that, although again, depending on what happens in the next few months, that may or may not ultimately be what we need to do. And then we would announce the cohort expansions. And so that'll obviously be driven by what we learn over the next few months. And also, if the data warrants it, that we might consider going back and doing another mini basket with the other six tumor types at that therapeutic dose range. So there could be a whole host of next studies that get teed up coming out of this dose escalation that could set up a catalyst-rich calendar for the next year or two after our data release. Very helpful. Now, last question for me because you'll be at the forefront of this novel biology with an older version of ADCs and the toolkit there. Are you working on a next-generation version of ADC technology to go after EDB? Yeah, so we're one of the few well, I guess there's a handful of us out there that have sort of complete toolkits in biotech, right? Which a lot of these toolkits are held captive in the hands of big pharma. So for making improvements on existing targets or drugging new targets with our existing toolkit, there's only a few places people in the industry can go. And we are fortunate that we do have that full toolkit from Pfizer where they optimized each component of the linker, the payload, and the conjugation chemistry. And then we also, when we acquired Apexigen in August in a stock-for-stock transaction, it was a modest transaction, $10 million value for Apexigen. We acquired their rabbit-based antibody platform. So now we have, in addition to this great toolkit that was originated at Pfizer, we have this rabbit-based antibody toolkit that can help us build the full suite of ADCs. So we are looking at the best ways to leverage our technology. Whether we decide to do the next-gen EDB out of our data or think about ways to apply that technology to other targets is still to be determined. And I think the data is going to drive our decision-making on that. I think one thing that's really nice about our target and is, again, unique right now in the ADC field is because it's such a novel mechanism and people are waiting to be convinced that you can drug the stroma, that hopefully when we convince them, we're years ahead of anybody else with any other stromal target. No one's going to be able to come right behind us on this. So we'll be shaping the field in how to think about the stroma from this target's perspective. And then I think, Brad, to your question, and big pharmas are great at this because they have great resources, right? They're always building the backup and the backup to the backup and continuing to tweak and perfect. For us, sometimes that can be us in biotech, that can be a little more costly because it could also put the milestones farther away. So I think we feel really great about this agent. And part of the reason for that with this ADC construct is I know personally, since I was at Pfizer when they did these ADC investments, I ran the strategy group that supported the investment into this technology. So I have firsthand experience with it. I know that Pfizer did that empirical testing. So with this specific construct for this target, they picked the best choices of what they had available. Whereas small biotechs sometimes make the best hypothetical choices and hope in the clinic it shows, and then they have to go back. Here, we have the benefit of that empirical testing that Pfizer did preclinically to put all those pieces in place of this construct. Yeah. All right. Very good. Really enjoyed the conversation.