All right. Good afternoon, everyone. My name is Jeff LaRosa. I'm an equity research VP on the immuno-oncology team here at Leerink. I'm excited to welcome Pyxis Oncology to our annual Global Healthc are Conference. Joining me from Pyxis in the fireside chat is Chief Executive Officer Lara Sullivan. Welcome. Thanks for being here.
Thank you for having me. It's always great to be here.
All right. Let's start off just to set the stage a bit here. Do you think you could provide an overview of just what you're focused on at Pyxis, especially as it pertains to your lead ADC program, PYX-201?
Yeah. We are excited about this year. This is really a year of focused execution for us. I think you may have seen we had discontinued our other clinical asset just prior to the holidays. PYX-201 is sort of all Pyxis all the time. All of our company focus and efforts are really around advancing this asset, which we are very excited about. We had a robust data readout in November. All hands on deck to deliver quality data later this year.
I guess just sort of to kind of dig into the asset a little bit more before we go into the trial, what does it target? What about the construct itself do you think is differentiating and unique relative to the very crowded and competitive field of ADCs?
Yeah. It's obviously been really exciting to see how the ADC field has captured everybody's imaginations over the past few years. Our observation has been that a lot of the entrants have come in with building sort of better mousetraps on existing targets. We felt it was important to actually bring an innovative target into the mix. EDB is a novel target. This is the first time we've seen in clinical development an agent, an ADC targeting EDB. EDB is extracellular domain B. It's a splice variant of fibronectin, which is a key structural component within the tumor microenvironment. The tumor microenvironment provides nutrition, metabolism for the tumor cells, also can be a source of supporting the tumor in its metastasis. Fibronectin, and therefore by extension EDB, provides structural elements to allow the tumor to have its shape.
We believe that in addition to altering the tumor microenvironment and interfering with the tumor's ability to sustain itself and to metastasize, that by also attacking the structural element of the tumor, it's like taking a building down, right? When you detonate the steel pilings of a building, the whole thing comes down. We have the added benefit that this target has that structural component too. Our view on ways to compete and win in the ADC field is obviously to have a strong technical construct with respect to the ADC itself, which we do, but also to bring innovation to the field because we do not think that it's a lot of benefit to bring the 95th HER2 ADC to the mix. At some point, there are only a couple that are going to be out there. HER2 owns that market.
We're better served finding novel targets to address unmet needs. I think despite the advances that are out there, there are still patients that are not served.
Absolutely agree with that. You started the phase I trial. You enrolled quite a bit of patients, a lot of tumor types, dose levels. I think if you could just talk a bit about what you set out to accomplish, what did you learn about the mechanism and the drug from this trial, the key findings and takeaways you think are kind of important. Pilot just from that data set.
Yeah. In addition to the novelty of the target, to your point, the mechanism is also novel. Our target exists in the extracellular matrix. The ADC itself cleaves in the extracellular matrix. That is different than any of the approved ADCs that are out there and the ADCs that are ahead of us in clinical development, where those ADCs are targets that sit on the cell surface and require internalization for cleavage as a primary mode of their mechanism of action. We do not require that internalization. Our thesis has been because every tumor type has an extracellular matrix, we are not restricted to certain tumor types simply because in the cell surface targeting ADCs, they are, right? The target has to be there. Every tumor has an extracellular matrix. Every tumor has fibronectin. Every tumor has EDB splice variants within there, obviously to varying degrees of expression.
What we found is quite high expression across multiple tumor types. Going into the phase one, that left us as a small biotech with an exciting challenge or opportunity, which is bringing this target into humans for the first time, where the target expression showed significant differential expression in multiple tumor types with a novel mechanism of action. The phase one dose escalation had a lot that it was asked to do. It was asked to show tumor regression in multiple tumor types. It was asked to prove the mechanism by clinical responses. Like any other dose escalation, it was asked to identify a dose range. It was asked to do that across multiple tumor types. We ended up, I think, escalating 10 or 11 doses across 9 of the 10 tumor types that enrolled.
If you actually do the math, that's a huge number of combinations of dose and tumor type. We knew going in that at the end of the trial, even with 80 patients enrolled, which was quite high for a phase one, in any one field of tumor type by dose, the end would be rather low because mathematically that's just the way it was going to be. Our objective in terms of the trial, in addition to kind of the usual objectives of what you're trying to learn in a phase one around dose and around tolerability, was to really find the evidence and the signals of where we should continue to develop. We took a broad approach with those multiple tumor types based on the PDX models, based on the IHC analysis that suggested broad tumor activity ahead of time.
That was really for us, I think that study, that dose escalation study, was really sort of the real beginning of understanding the potential power of this agent because at the end of the day, the mouse and all the preclinical stuff is helpful and informative. Nothing substitutes for finally seeing it in patients and finally understanding what does it look like in a breast cancer patient versus a pancreatic versus a head and neck.
Yeah. I think just before the tumors, I think talking about the safety profile because I think it kind of relates to the mechanism in some ways. It's a little different than what you might expect from microtubule inhibiting ADC. What did you see there that you kind of found encouraging or different versus other ADCs that gives you confidence as you hone in on that 5.4 mg/kg or potentially in combinations with Pembro and maybe other agents in the future?
Yeah. We were very pleased with the safety profile that emerged out of the phase one. A couple of things sort of predicted it going into the trial that I think we saw to be true. One was the specificity of the target expression, right? As I mentioned, it's highly expressed across a number of different tumor types. What's also unique about this program is the differential expression, meaning the target is negligibly expressed in normal tissue. Right there, that tells us that if we believe that the ADC construct is well designed and can hold this payload, hold it tight until it gets to where it needs to be released, then you would expect a nice safety profile because you're not drugging normal tissues. Hypothesis number one was the differential target expression was going to help the safety profile.
Hypothesis number two was that the well-designed ADC, which was designed by Pfizer, who originated the program, with a site-specific conjugation chemistry, an optimized MMAE type of payload, a molecule that was designed for stability, that the construct would deliver that payload exactly where it needs to go. Again, we would see a relatively smooth safety profile. In fact, that's exactly what happened. We were very pleased that out of the 80 patients dosed, and we reported as of the data cutoff in November for the data disclosure on 77 of them, that we had no grade 3, 4 instances of neuropathy or ocular tox, which for ADCs are obviously things that can stop a program in their tracks. We had a very modest amount of neutropenia and cutaneous skin lesions that were easily manageable in the case of neutropenia by G-CSF.
In the case of lesions, they were superficial, did not go to the mucosal layer, and very limited in number. We were very encouraged by that. Out of the 77 patients, we only had one patient discontinue for treatment-related reasons. We had no grade 5s related to treatment. I think that's just important to remind everybody because unfortunately, sometimes in the speed that things get reported and the internet and so forth, patients pass away on clinical trials of disease. That was no different in our case. We had patients pass away of progressive disease. These are end-of-life patients. The median prior lines of therapy of our patients enrolled was 4. We were the fifth line. We were dosing patients end of life whose next stop was hospice.
As any patient who comes in, a doctor's best guess is when that's going to happen, right? Sometimes that happens earlier in the clinical trial than you expect. I just want to make clear that there were no grade 5 safety issues associated with the drug. We did have patients pass away of progressive disease. That's very, very normal. When we look at the safety profile compared to the safety profile of other ADCs that are out there that are approved, in our corporate deck, we have a slide that looks at the label for Padcev, Tivdak, et cetera, et cetera.
You can see that when we compare our 5.4 dose, which was the highest dose that we cleared, in these late-line patients, right, compared to the label of earlier line patients of these approved ADCs, we're comparable or better in each of these side effect dimensions. We believe that as we move into earlier lines of therapy and have patients that are less sick, that theme around the safety table will continue to hold. That's our belief. We'll see with the data later this year.
Do you think you could dose escalate further given that profile? Is that something you're sort of tweaking a bit maybe between that next dose after the 5.4? I guess some people have brought up maybe like pneumonitis, but those weren't necessarily treatment-related. I mean, could you just talk about is that something we should be worried about with the mechanism and about maybe if you have the flexibility to dose higher if you need to?
Sure. We cleared the 5.4 dose. I think it was 42 patients maybe that we had dosed at 5.4. We were really fortunate in the way the enrollment accrued as we were escalating that we could see that signal at 5.4 really early on and were able to continue to enroll rapidly there. We did dose up to 8, and then we backed down to a 6.6 dose. Those patients that were recruited at those dose levels, just the luck of the draw ended up being some of our more sick patients in that trial. In the case of 8, I think as we stepped back and looked at the overall data, we felt that was above the MTD. I think that's off the table. The 6.6 data was never fully characterized because of the acuity of the disease that the enrolled patients were experiencing.
Out of an abundance of caution as we were working through the 6.6 dose level, we and the investigators decided to drop them down to 5.4. To your point, based on what we've seen at the 5.4 level, theoretically, I think we could potentially see some additional dosing there. That being said, when you start to look at timelines, building up data sets, involving with the FDA, it may on a practical basis be difficult to really add another dose level. We're leaving that open for consideration as part of this cohort expansion. I think that ultimately, as we get further advanced and we engage in the Project Optimus discussions with FDA, by then we will really have sort of firmly nailed what we think the top dose is that should be engaged around that.
We're very happy with how the 5.4 performed. It's just a question of do you want to squeeze a little bit more out of it or not? We'll see.
That makes sense. I think now it's going into the tumor types and sort of the efficacy preliminary that you saw. Head and neck was the tumor type that really shone and shined through. You are prioritizing on that. There are some skeptics who might diminish that signal as lucky or overinflated given the small sample size, six patients that were at valuable doses, and the context of this 80-patient data set. I mean, I guess why are they wrong, frankly?
Oh, yeah. Yep. I just realized I didn't answer your pneumonitis question. Let me say two sentences on that, and then I'll answer your head and neck. Just quickly on the pneumonitis, I think the pneumonitis obviously is something you really want to be careful about. Lung patients in particular can be susceptible to that. We just paid attention to that as we were thinking about focusing on the head and neck signal and our resources. I think just out of an avoidance of caution, ADCs and lung patients in certain populations can get a little uncomfortable for folks. We had, I think, one pneumonitis grade 1, 2 at 5.4, which essentially is imperceptible to a patient. It's usually an incidental radiographic finding. We had one at 5.4 that was a grade 3, 4.
Overall, I think still in line, but something that we just don't want to in the lung world, I think that's more of a concern than in the head and neck space. Going to your question then about head and neck and the skeptics, I always smile a little bit at this, to be honest with you, because we have so many stakeholders that we are working with with respect to our data and the company. Obviously, the investor community that we value tremendously, the investigator community, and then the strategics, right, who could be potential collaboration partners. I have never heard a skeptical comment about our data from any investigator or any strategic that we've spoken to, not one. Every single one of them has said, quote, "You have a drug." I attribute a lot of the investor skepticism in a couple of ways.
One is bringing a novel target and a novel mechanism at the same time into the marketplace is always hard because investors and humans, right, people in general, prefer pattern recognition. They want to know what does this thing look like so I can predict a response, right? We could not give people that comfort in leading up to this first study. Therefore, the bar of proof is higher, right? The skepticism is going to take longer to resolve. We understand that. I think the second factor is just the general climate, right, is a very risk-off climate. Again, that reason to believe is so much higher. Just as a backdrop, I wanted to share that. Going into the specifics around the data, we had strong tumor regression across six different tumor types. Head and neck was the strongest signal.
Of the six patients that were in the cleared 3.6-5.4 dose range, we had a complete response confirmed by RECIST and two partial responses confirmed by RECIST. 50% ORR in monotherapy. The three patients that were dosed, two at 6.6 and one at 5.4, all who were disallowed, all had tumor regression. These were all complex medical cases that if I were to walk anybody through five minutes on each case, I think you would understand why we believe each of those cases indicates drug activity. I can't count it statistically, but clinically, it's relevant. If one believes that and were to imagine those three patients dosed in a protocol allowable way, either at an earlier line of therapy or not, that would be 6 out of 9 at a monotherapy response. We see that as compelling.
Now we're battling the law of small numbers, right? It goes back to the design of the trial. We had nine tumor types and what, 10 or 11 dose levels? We knew we wouldn't be able to give the investor community an N of 20, right, in a certain tumor type at a certain dose level. We knew we just couldn't. That is why I think we also try to speak qualitatively about what we're seeing as well because I do think there are clues that investors who are really paying attention and doing the clinical work to understand the cases will benefit from. The data is in our deck. We have had HPV positive, HPV negative responders. We've had responders of patients who have not responded to EGFR therapy prior.
That's really important because Merus and Bicara are the two head and neck folks ahead of us battling it out for EGFR. We have patients in our trial who've relapsed or not responded to EGFR and have responded to our agent. These are clues that obviously we have to prove out with a higher N and a more advanced development path. Our conviction is extremely strong that this is a real signal in head and neck and the other tumor types. Head and neck is shouting. The investigator and strategic communities who've looked at our data, every single one have said the same thing.
Yeah, I think there's some confusion when I talk to more skeptical-minded investors. They would say, well, there isn't a biomarker with around ED B FN and its expression. I know you've been doing some digital pathology work around that. I'd love to hear about that. Also, from the sense of when it was always thought of, I think the confusion was this is targeting the stroma, et cetera. Then pancreatic cancer is all the stroma. You had like 17 or so patients in that, but maybe wasn't as robust as we would have thought going into it. Just sort of, I guess, to allay sort of concerns around this lack of biomarker, I mean, is this something that you're actively digging into?
We can expect that be elucidated at some point in the future and kind of why maybe head and neck was the tumor type that really.
Yes. The answer to that is yes, yes, and yes. I can't give you the specifics right now. What I can share with you is that the digital path work is progressing extremely well. It's a digital Path AI-enabled tool. It's allowing us to hypervisualize these structures in the stroma and the extracellular matrix. With 80 patients having dosed in some, obviously, we don't have biopsies for every single one of them. Some of the lower dose levels are less relevant for us from that perspective. We have enough samples that we can see at a much level, a much greater resolution what's going on in that extracellular matrix. I think the classic way people think about biomarkers, and again, maybe it's going back to cell surface targets as the construct, we're in the extracellular matrix.
We have got a lot going on in there, right? There are structural components. There are other cells in the ECM. There is the ECM itself. It is an acidic environment. It is a hypoxic environment. There are all kinds of things that could contribute to a, quote, "biomarker." The Path AI tool is helping us see that and make those correlations. We are planning to share out some of that work over the course of the year. As we get closer to data, I think we will start to see some of the connections there. I think that the headline sort of message I would say related to this that I would advise investors to think about and pay attention to who are following our story this year, the stroma is not a monolith, right? It is not the same in every tumor type.
It even has its own sort of life cycles, if you will, within certain patients. Different tumor types have sort of different themes of stroma that may or may not make those tumor types more or less amenable to this kind of approach. It is also, again, the stroma and the extracellular matrix are a very dynamic environment, right, attracting immune cells too. When we think ahead about our combo coming up with Pembro, thank you, Merck, for the supply, right, huge, huge value to us. I think we're going to have some really great insights of what we're seeing sort of at the bench and then what we're seeing at the bedside as this all comes together over the course of the year. It may not be a classic biomarker in the sense the way people think about it.
I think there's actually things that will triangulate with what we're seeing. That probably will help people get comfort with both the mechanism of what's happening and the clinical data that we've put up and will put up.
Excellent. I think that will be reassuring for a lot of folks. I think let's go forward looking now. You are focusing, at least currently, on head and neck cancer. Can you sort of talk about the various arms and kind of hypotheses you're testing? As a follow-up to that, what are you doing different in your enrollment criteria from what you saw with the head and neck patients maybe in dose escalation that you're now doing in expansion? How do you think that will sort of impact enrollment of patients?
Yep. In terms of we've got two studies kicking off this quarter, the monotherapy expansion and then the combo study with Merck with the Pembro. In terms of the mono, there's two arms there. One is essentially taking the patients who are not responding to current standard of care, so the platinum and PD-1. This is a large population. It's pretty dominant in the U.S. It's an important place to go. The second population is the patients who are not responding to EGFR. We really like this part of the development strategy for a few reasons. One is you do see some geographic differences in prescribing patterns between parts of the world that use EGFR more than, say, maybe the U.S. or other places. The treatment algorithms are a little bit different.
Also, we can start targeting that population right now because those are the current patients who are not responding to Erbitux. These are also the patients, the future patients who do not respond to Merus or Bicara. The way we are expecting the data to kind of roll out in monotherapy is to get the readout on the platinum and PD-1 patient population first because that is here. It is easy to get. The EGFR population requires a more diverse geographic strategy. We want to capture some of the Merus and Bicara patients. We have to mirror a bit on their clinical trial sites. Fortunately, the PIs see who these patients are. We do not have to know whether they were on Merus or Bicara. We just have to know from the eligibility. By having great site relationships with the PIs, we can identify those patients.
That data will come a little bit later. I think we're guiding to first half of 2026 because we want to make sure we capture enough of the Merus/Bicara. We believe that it's great for patients what Merus/ Bicara have put up. There's plenty of room for multiple mechanisms. In fact, finding a spot if we end up right behind them is a great place to be. I actually think as we continue to develop and put up more data, quality data that I think is going to be competitive to, if not superior to their data, the ability to kind of move earlier than them is there. Right now, they have a higher data set, and they have time ahead of us. If we end up starting right behind them as second line, that's fine. Or third line, that's fine.
We're still going to look for first-line opportunities. It is probably going to be a chronology to do so. Just we're not all on the same time pace right now.
Yeah. I guess just really get us something to look forward to in the second half. I believe you're on track for preliminary data in the second half monotherapy outside of the post-EGFR and the combination with Pembro in the front line. What should we expect in terms of patients per cohort, maybe follow-up, and for the monotherapy and early for any of these? What are you really looking for that you think would really get you excited that you've replicated or supported what you did previously and that you actually could have a potentially best-in-class therapeutic?
Absolutely. I think the combo with Pembro is coming neck and neck with the mono. We expect to be able to have some kind of preliminary readout of the combo by the end of the year in the second half as well. Those two are sort of tracking each other. In terms of the N, we haven't disclosed that yet. Our 10-K is coming up. We'll give more visibility before the end of the month out of our 10-K on some of the specifics on the development strategy. I think that Merus and Bicara are setting a new standard. It's great to see. I mean, certainly looking at the waterfalls for the patients, it's really compelling for both of them. There has been a lot of discussion in the marketplace around their, I'm lumping them together. I know there's a difference at the moment.
The fact that HPV status can be a differentiator for either of them or both of them potentially. We have not seen in our data set thus far any difference in HPV status impacting responsiveness. If you think about our mechanism on our target, there is sort of no reason to think HPV status would be a differentiator. Our belief is that we can sort of continue to find a place that is differentiated alongside, behind, ahead, really remaining to be seen. I think that they put up numbers, certainly in the monotherapy, Merus did, right? Merus put up a 35, 33, 37, something like that in the monotherapy. We see that as kind of a minimum. In the combo, it was sort of 64, 65, something like that.
We also recognize, as we've heard from the KOLs and the PIs, that ORR is fine. It's sort of a crude measure I think we all use. At the end of the day, it's durability, right, and overall outcomes for patients. Of course, that takes time in a trial. I think that we've seen with our patient population, and again, the head and neck patients, as we isolated that data from the other tumor types, actually, and it's in our deck, had a longer time on therapy than the rest of our, when the general, right? It was 115 days for the head and neck. I think it was 75 or something like that for the full population. 115 days in a phase one median fifth line is a really long duration on therapy, right?
Those patients have had nice disease control while they've been on therapy. We think that, again, as we get into these studies and as the weeks and months go by that we're accumulating data, that we'll be able to be competitive from a durability perspective too.
We can get an update on those prior phase I patients for head and neck potentially?
We actually decided we're not going to provide any more updates on the phase one data set primarily because so much of it was out there when we did the data disclosure. Obviously, we're following patients who are still on therapy as well. I think in terms of the focus going forward, we want to really orient everybody's attention to the next stage of development where patients will be earlier line, right? We're looking at first and second line in combo. You asked the question around selection criteria. We've tightened up our criteria around patients with prior neuropathies, neutropenias. We've added some prophylactic controls with respect to neutropenia to the protocol. Again, just making sure patients are supported throughout their experience here. Yeah.
Thank you so much. There's a lot we could talk about. Even going past head and neck, but I think we're out of time, unfortunately. Thanks for being here.
Thank you.
It's good to see you this year. I appreciate the conversation.
Yeah. Good to see you. Thank you.