For cancer. The company currently has three clinical-stage candidates, including its lead ADC asset, PYX-201, and two immuno-oncology candidates, PYX-106 and PYX-107. So to discuss the company's portfolio, let's get started with Lara. Lara, thank you for coming down, and appreciate you accepting our invitation to talk to our audience today. So to start off, Lara, stromal tissue around cancers have actually been a major issue for many therapies, specifically ADCs, them not being able to cross and get to the tumor. But however, at Pyxis, your team has found a solution for this, and so just so that we understand your technology a little better, please highlight what your team has achieved so far, especially on the technology side of things.
Sure. Thank you. It's a pleasure to be here, and we appreciate the time and interest of everyone in the room to hear our discussion today. So Pyxis, while we have the three clinical candidates, I'd say the real focus of the company is on PYX-201, which is our ADC program targeting the stroma. The approach we're taking is to target a splice variant of fibronectin in the stroma, which serves as a structural component of the stroma. So that helps the tumor maintain its infrastructure, its metabolic properties, and all the things that the tumor does to keep itself alive.
In the past, what I think we've seen with other ADCs, a belief that the only way to really leverage the modality was to target the tumor cell itself and to have the ADC internalized into the tumor cell, where it would then cleave and kill the tumor from the inside out. What we've seen through our preclinical work that was done by Pfizer, who originated the target and has been enhanced and built upon by our team, is that the conditions for payload cleavage actually exist in the stroma. The right proteases are there, the right acidic environment is there, and in fact, by taking down the structure of the tumor, you can take the tumor down as well.
So when our clinical data comes out later this fall, we believe that not only will the data be of interest for the program, but it will also enhance an understanding of how the ADCs can be used and to reopen interest in stroma as a rich source of targets going forward that, frankly, have been overlooked over the last thirty years.
So, this splice variant, EDB-FN, and how specific is it to the tumor stroma itself?
So we love this target. This target is highly specific. It's expressed quite intensively within the tumor stroma across a variety of solid tumor types, and when you look at the companion normal tissue, the expression level is zero. So the gradient, tumor type after tumor type, from tumor to normal tissue, is extremely steep. We think that bodes extremely well for this type of modality because it, you know, it reduces the potential for side effects happening in normal tissue. You know, at the end of the day, the patients, when they're taking ADCs, are sort of battling target-mediated tox as well as platform-mediated tox.
By having such a specific target expression pattern, you know, we're minimizing the need to deal with target-specific tox and just trying to dial in, you know, where the optimal amount of payload dosing for the patient can be.
Talking about the three candidates, and let's start with PYX-201. What's the mechanism of action for 201?
So, as I mentioned, 201 targets the stroma, and so it cleaves within the stroma. We think the you know primary components of the mechanism include passive diffusion of the payload into the tumor cell to kill it through that mechanism. We believe there's a heightened bystander effect. T hat's going on with the payload cleaved in the stroma, and then the local sort of marshaling of the immune system with some immunomodulatory effects. We think the contribution of all three of those components is what's driving the work of this ADC. Pfizer had put out some AACR posters back in 2014 showing the nature of the proteases in the stroma, and we've been continuing to build upon that work. So, with our data disclosure this fall, we'll include some additional insights into the mechanism-
Okay
As well as the clinical data itself.
You are currently in phase I dose escalation study. So what's the design of the study? And, you know, how high of a dose have you tested so far?
So we have both a new target, new biology, and a new mechanism that we're working on at once, which, you know, is no small feat, right, in terms of getting understanding across all these dimensions. So because of that, we chose to design a basket trial for the phase I dose escalation with ten tumor types. We had pretty compelling PDX models, and the IHC analysis showing the target expression was quite prominent across multiple tumor types. And so we didn't want to solely rely just on the preclinical data that wasn't able to really discriminate enough, "Go here, don't go here," because it looked good across multiple models. So we said we should test all of those in the clinic the best that we can, recognizing we're a small biotech, we're limited in resources.
So we decided to do a basket trial of ten tumor types, and we've been very careful, particularly over the course of this year, as we've gotten into that therapeutic dose range, to make sure that we are allocating the patient enrollment across the tumor types, thoughtfully, so that as the data comes out, we have a decisionable, evaluable N for each of these tumor types. So the ten tumor types include head and neck, lung, pancreatic, ovarian, sarcoma, triple-negative breast, HR-positive breast, thyroid, HCC, and RCC.
And the eight milligram per kilogram dose is the highest?
Yes.
Okay.
In terms of the dose escalation, we started out dosing at zero point three way back in March of 2023. Again, you know, we've been very cautious and deliberate as we've done our dose escalation because this is the first time this target has been explored clinically with this modality. We used a very patient dose escalation design. We announced in March of this year that we had cleared the five point four mg per kg dose level, which was very significant because this particular linker and payload combination, it's a modified auristatin. Again, the IP came from Pfizer.
Had been seen in the clinic, that linker payload combination, in the legacy Pfizer HER2 ADC, which has since been stopped for commercial reasons. And in that program, that linker payload was not able to dose above 4 mg per kg. So they got to 5, and they had a 50% discontinuation rate. In March of this year, we cleared 5.4, and we announced we were beginning to dose at 8. And so the leap from 5.4 to 8 was driven by our PK analysis, suggesting that, you know, that was sort of a meaningful differential to be able to assess, for patients.
Since then, we have just announced continued dose enrollment or patient enrollment as opposed to dose levels, because I think as most people in this room know, you know, dose escalation is also an art as well as a science. Especially when you get to higher dose levels, for us, we wanna get to that MTD, and so you wanna be very careful in managing, you know, the significant AEs that come as you get close to MTD. So you do a lot of sort of small steps up. Maybe you do some interpolation, you know, between dose levels. You allow investigators a little more freedom to dose escalate or to deescalate, all in service of learning more about the patients.
So 8 was the highest dose that we publicly disclosed, and 5.4 was the highest dose that we had publicly disclosed that we had cleared. And then we'll provide full insight into the full dosing regimens as part of the data disclosure later this fall.
Yeah, is there any possibility to go higher than eight milligrams per kilogram, or should we start being concerned about higher doses of auristatin?
Yeah. So I think, you know, even with such a specific target, that's where the payload, right, the payload can kind of put the brakes on it for the patient population. So we're not gonna comment, today, on the ceiling of the dose level or where we've gotten thus far. But I will say that as part of the data disclosure, we'll provide clarity around what that dose range or dose envelope is, that we feel very comfortable, sort of is lower, and we're approaching sort of on the upper side. And then we'll rely on the next stage of development, including some of the dose optimization work that needs to be done to fine-tune, where exactly the dose level will land.
And then, you know, since you said that you're testing out 10 different key tumors, and you're gonna put out some data.
Yes
Later this year. So what would be the criteria for you to kind of figure out which are the real indications that you would take forward, you know, beyond what you're testing now?
Yes, so we announced a couple weeks ago with our 10-Q that we had dosed 72 patients already. We expect to dose somewhere around 80 by the time the data disclosure comes out. You know, even if you just did straight math, which, of course, none of this is that linear, that's 8 patients per tumor type. That's not enough of a concentration to help with decision making.
So to help alleviate the pressure of that, about six months ago, we decided to enhance our recruitment in certain tumor types to build a higher N at these higher dose levels, so that as we put the data out, we might have, you know, 10 patients at a therapeutic dose level in one tumor type and 8 in another tumor type, so that would allow us to make those decisions. So how we're making those decisions is a combination, truly is art and science, meaning, you know, there's the straight math of X number of patients, and how many responses do you see, and what's that response rate look like? But of course, that's in the context of an all-comers, heavily pretreated patient population, low N.
So you need to be able to contextualize that math with an understanding of the qualitative nature of that signal, right? So the math could mask something very interesting going on in certain tumor types, where perhaps you have five patients out of eight that are stable disease, but if they're all - 25%, there's probably a great signal there.
Yeah.
Maybe doing the fine-tuning around the patient selection is gonna help that in the next phase of development. We start with an understanding of where the current emerging standard of care treatment expectations are. We assess where we are relative to that, and then we layer on a qualitative assessment around the strength or the quality of the signal and how it's been tested in the context of this study, how it could potentially be tested in the context of the next phase of development, where we'll have more specific patient inclusion criteria based on our learnings from this, as well as more experienced clinical trialists now that, you know, we've been at it for a year and a half and a better understanding of the agent ourselves.
So all of that together, when we disclose our data, will provide clarity around, here's the tumor types that we're taking forward and why, what the stage of development will be, and the catalyst associated with it, which tumor types we're setting aside because we didn't think it was, you know, strong enough to be competitive, either clinically, or commercially. And then there'll probably be some in the middle that the end just wasn't high enough.
Yeah.
To get a strong enough decision one way or the other, and we'll share how we plan to further evaluate those.
Okay, and then, you know, as you know, ADCs end up being part of a combination therapy at some point.
Yes.
For 201, you know, what sort of drugs makes scientific sense?
Yeah
In terms of being part of a combination therapy?
Based on the mechanism we have, and also some of the preclinical work that's been published already, we see combination with PD-1, PD-L1 as being a really interesting place to explore. Obviously, the ADC plus checkpoint inhibitors, and we all saw the Keytruda plus Padsev data last year. That was super exciting. That's kind of a no-brainer from a combination potential to take a look at. I also think that, you know, as we get more granular by tumor type around what our next steps are, there could be combination partners that vary based on tumor type, right? If we were to go forward in ovarian, you know, we might wanna look at combination opportunities in one area that looks different than if we're going forward with combination opportunities in lung.
The other thing I would put forward is that this concept of ADC plus ADC combo is not sort of introducing fear the way it might have three years ago. We heard from a lot of big pharmas at AACR earlier this year when they talked about their pipeline strategies, and we've heard in private conversations a real interest in figuring out how to layer ADCs with one another. We, I think, are uniquely positioned compared to all the others that are out there to be a potential combo partner of choice because we're targeting the stroma.
So I think the way we can layer and manage the tolerability profiles is a real advantage compared to trying to put together two internalizing, you know, cell surface targeting ADCs, who could potentially have synergistic or augmenting toxicity profiles, as opposed to ours, where it could be more complementary. So I think that the monotherapy data earns us the right to participate in the combination regimens, but I think the combination regimens could be a real upside for us once we really get the lay of the land on where we're going from a monotherapy perspective because of the mechanism.
Okay. So the PYX-106 is the second molecule that is in the clinic for you. So what's the target there, and what's its mechanism of action?
Yeah, so, PYX-106 is targeting Siglec-15, so it's a novel checkpoint inhibitor, and we think that it has, you know, potential in the PD-1 or PD-L1 refractory or non-responding populations. That asset is in a similar basket trial design to what we have for 201 , but we indicated pretty early on we were focusing recruitment around lung, and that was based on the clinical experiences of a prior molecule against this target. That data is expected by year-end, so we've been guiding that our data for 201 will come out before Thanksgiving, sometime between now and Thanksgiving, not Wednesday before Thanksgiving, and 106 will come out before the end of the year, so sometime in December, effectively.
Okay. So is there a plan to look at other tumors outside of non-small cell lung cancer with PYX-106? And when would you make that decision if you decide to do so?
Yeah. So we had, I think, a total of nine tumor types, so eight ones besides lung cancer in the basket recruitment. I think, you know, we also tended to see from the investigators some concentration in certain tumor types. And so I would say that one is probably less diverse in terms of how the data will read out by tumor type than what we have for 201 . But similar calculus in terms of decision making, you know, taking a look at what the clinical findings are, taking a look at where the current emerging standard of care is, and then assessing, do we think this can be competitive, and potentially win or have a place in the current treatment regimen? And if yes, take it forward, and if not, then, you know, move on to the next thing.
In terms of Siglec-15 or targeting Siglec-15, is that a better target for being a monotherapy, or do you think there, too, you will have to end up, you know, having to combine with something else?
Yeah. So I think, you know, our philosophy is, no matter what modality we have in-house, the therapeutic has to earn its stripes as a monotherapy to earn that right. You know, we've seen too many biotechs who have just mediocre data or middling data, and they hope that combination regimen will rescue them, and it just. It rarely works.
I think it has to stand on its own as a monotherapy agent. I do think that, you know, combination regimen, obviously, you know, is sort of a bread and butter, but we would not take it forward with hopes that combo regimen would be where it would declare itself. It has to declare itself on its own.
So beyond these two, which are currently in the clinic, you know, what else is there in the pipeline, that can come to the clinic, and what could be the cadence of that happening?
We have an additional clinical asset, 107, that you mentioned earlier, that came to us from our acquisition of Apexigen last summer, which is a CD40 agonist.
Best-in-class. It's been through phase II in liposarcoma and melanoma. It came over on hold, and we have kept it there pending the data readouts of these two programs before making any kind of big portfolio commitment for later-stage development for that asset. We also have a huge amount of IP between our founding immuno-oncology IP out of Tom Gajewski's lab at the University of Chicago. Between the IP that came over from Pfizer with our PYX-201 license, we have all the linker payload conjugation chemistry IP that they built up over a decade and spent hundreds of millions of dollars optimizing. And then, we have Apexigen's rabbit-based antibody IP. So there's a lot to choose from, and we haven't commented publicly on what our plans are for the next wave of pipeline build.
I would say that, you know, as we work through kind of the clinical readouts this year, the overall shape of the pipeline will start to sharpen, and we'll be communicating around that kind of in a staged series of events.
So, as on closing out.
What's your current cash position, and what sort of a runway would you get from that?
Yeah, so our last announced cash position's $157 million. It takes us through the second half of 2026, so it covers the next stage of development coming out of these data readouts. We have a catalyst-rich calendar that we expect for the next, you know, 12+ months. Again, we'll share that with the announcement of the data, and you know, depending on the market conditions, the right timing for the next raise could be associated with this data or potentially the next set of catalysts. We'll have to see what happens.
Okay. Thank you. Thanks for this.
Thank you.
Good luck.
Thank you.