Hi, everyone. Thank you for continuing to join us on the Stifel Virtual Oncology event. My name is Brad Canino, Senior Analyst here. Very happy to do a next fireside chat with Pyxis Oncology. We have Lara Sullivan, the CEO. Lara, thank you so much for joining us.
Thanks for having me, Brad. Always great to talk to you.
Of course. Let's start off just with a quick intro to Pyxis, and maybe talk about what you established in November last year in terms of proof of concept from your initial dose escalation data update for MICVO. Pelidotin, I've got to get used to saying that now.
I know, right? It doesn't exactly roll off the tongue, but it will soon. We like to call it MICVO, so that makes it a little easier. Yes, MICVO was formerly known as Pyx 201. And Pyx 201, well, Pyxis is an ADC-focused company. Our lead program is MICVO. It's an extracellular targeting ADC. We completed phase one, part one dose escalation studies last year. We announced our data in November. From a corporate perspective, we decided coming out of that data to focus our clinical pipeline on Pyx 201, MICVO. We attributed our other immuno-oncology-based clinical program at the end of the year. We recently reorganized the footprint of the company around being a single asset company focused on MICVO. We're all in on this asset.
We saw some very exciting efficacy signals across multiple tumor types that came out in the data set in the fall, with a particular emphasis on a strong signal in head and neck. We are focusing our clinical development path forward in head and neck. We have a monotherapy program as well as a combination therapy program in combo with Keytruda, both of which started up this past first quarter, and both of which we expect to give meaningful data readouts in the second half of the year. We will have data from at least 20 patients in monotherapy and at least 20 patients in combo therapy in the second half of the year in head and neck. We also will have an additional head and neck data set in the first half of next year. There are two different arms of the monotherapy data set.
One arm will come second half this year, and the other arm will come second half next year. Ultimately, a total of 60 patients would have at least head and neck data on across the two studies and the two arms within the monotherapy study.
Right. Maybe just to intro MICVO a little bit too, can you speak about the components of the ADC and then maybe marry that to the PK data, the safety data that you saw as well in the phase one in terms of why those were rational design choices for the ADC construct?
Sure. In addition to MICVO being a first-in-concept extracellular targeting ADC targeting EDB, which is a component of the tumor matrix, EDB is a spliced variant of fibronectin, which is a structural component of the tumor. In addition to that element of the biology, we think the ADC construct itself is extremely sound. The ADC was designed by Pfizer. This was the legacy Pfizer technology before they acquired Seagen recently. Pfizer spent a decade or so perfecting different components of the ADC. I know this firsthand. I was a Pfizer executive at the time. I led the strategy group in Pfizer, and part of my remit was technology investment. I can attest to the fact that a significant amount of capital, hundreds of millions of dollars, was spent on ADC development and optimization.
The components of the construct that were selected to become Pyx 201 or MICVO were empirically tested and optimized, each component, the linker, the payload, the conjugation chemistry, repeatedly tested, optimized, perfected, and brought together. In our case, our MICVO has a site-specific protease-cleavable valine-citrulline linker that enables us to have a reduced free payload in the serum. It is a very sort of nicely bound ADC. It takes the payload to the tumor where it needs to with minimal discharge prematurely in serum. You see that reflected in the PK data. The payload itself is a next-generation orostatin payload. It is designed for potent tumor killing, but it also sort of enables and sort of encourages local immuno responses. It is a nice setup for the combo study that we are doing with Keytruda.
In and of itself, as a monotherapy, it's well designed to go exactly in a targeted fashion to the tumor cell, discharge the payload there in the tumor extracellular environment, and with minimal sort of disruption to normal tissues along the way. We saw that proven out in the PK data that we released as part of the data release last fall. We had no antigen sink demonstrated. We showed PK data across the variety of dose levels from 0.3 up through 8. We consistently saw the serum concentration of the full construct being aligned with the serum concentration of the payload. We were not seeing a divergence there.
Yeah. How has the thinking and the basic science behind extracellular payload release for ADCs evolved recently, not just for your own ADC, but for others? I know I saw a Nature paper come through talking about this for Enhertu recently. Talk a little bit about that.
Yeah. Thank you for that question. I think it's actually an incredibly important one. I think that this moment in time, as this information is sort of coming out from Enhertu, is going to be looked at over the next few years as the beginning of a seminal change in thinking around how ADCs work mechanistically. I say that because we have been at the forefront of this with an extracellular targeting ADC for the last three years. We raised our Series B and our IPO on this technology. We've been very clear all the way through that our agent acts in the extracellular environment, that there are proteases in the extracellular environment that cleave the payload and enhance its release, and that the action in the extracellular environment is where it's at.
There had been such a canonical entrenched belief that ADCs have to be targeting cell surface targets and internalized to be cleaved. We had papers from Pfizer scientists back in 2014, or posters, I should say, that showed the protease concentration in the extracellular environment. We also had seen work done by scientists ex US that showed the same. Honestly, I think that there had been just healthy skepticism from a scientific perspective because we weren't yet dosing patients clinically. This was still hypothesis. Now we've got the paper out from Enhertu demonstrating that that extracellular activity is a way bigger important component of the mechanism than anybody had appreciated. Enhertu is the dominant ADC. We all aspire to be like Enhertu. I think Pyxis mechanism is Enhertu on steroids because we don't have any internalization going on. We are strictly extracellular, and we have broad anti-tumor activity.
We showed that in the fall with meaningful tumor regression in six tumor types. We welcome that paper. We're excited about it. The folks who've been working, the scientists who've been working on ADCs for the last decade, who kind of had this glimmer in their eye about this component of the mechanism, I think are finally being validated. I think we are ahead of the game in having the only extracellular targeting ADC that's out there. We're now advancing into phase one/two, cohort expansion. As this mechanistic body of evidence builds from Enhertu, and we contribute it to ourself, which we're doing this year, we've got two posters at AACR. Please check them out in the next two weeks. They're very impactful.
Our commentary around the proof of mechanism, you're going to hear us continue to discuss that over the course of the year, building up to our clinical data, where I think we're going to have that irrefutable evidence with a high end clinically and the mechanistic proof that Pyxis generates, plus the learnings in the field that Enhertu advances that make this concept essentially the new way of thinking about ADCs as opposed to in the past, an exception to the canonical view, which I think is going to be proven to be inaccurate.
Yeah. Great. Now, thinking about your phase one data, first-in-class target, first-in-concept ADC, what degree of safety events did you see at the therapeutic doses and above? What in your mind would you attribute that to, the target, the payload, the linker?
Absolutely. We have a very safe agent. We had no grade five drug-related events. I want to emphasize that because our phase one study, part one, was in extremely sick end-of-life patients, as many oncology phase one part ones are. These are end-of-life patients. They are often, in our case, they had median four prior lines. We were fifth-line therapy. We had patients coming in who passed of progressive disease within weeks or months of enrolling. That has nothing to do with the drug. The safety profile bore out exactly the way we hoped it would and the way we thought it would based on our PK, based on our understanding of the target and the mechanism, essentially as follows: no grade fives that were drug-related at zero. We had one patient discontinue for drug-related reasons out of 77. That's a 3% discontinuation rate.
Pfizer used the exact same ADC construct targeting HER2, a cell surface target, at a 5 per kg dose and had a 50% discontinuation rate. I say this because it cannot be emphasized enough in this age of people paying half attention to an erroneous comment on Twitter that can lead to inaccurate conclusions. The safety profile was extremely favorable. We had no grade 3, grade 4 neuropathies or neutropenias. We had a handful of ocular tox and a handful of skin rash. The ocular was dry eye in nature. It was treatable, reversible. Same with the rash. It did not go into the mucosal level. The very few grade 3, 4 events that we did have were manageable.
When we looked at our safety profile at the 5.4 dose, which is the top end of the therapeutic dose range, it was as good as or better than half a dozen of the approved ADCs that are out there. We have very high confidence as we go into cohort expansion that this is a safe agent, that it's mechanistically at the forefront of where the field is going, and that it has meaningful opportunities for tumor regression, particularly in head and neck, where we had a 50% confirmed ORR, including a complete response. We feel very good about it. I can give you a little bit of a clinical update on the data set since then, if that's helpful as well.
That would be excellent.
Yeah. We had published in November the data, the efficacy evaluable population. There were six head and neck patients. We had dosed nine total. Two of them were at 6.6. These were end-of-life patients. They ended up in hospice, and one passed away of progressive disease. Those were not evaluable. One 5.4 patient that had one dose was not evaluable because that patient did not complete protocol for other non-related reasons. All three of those patients that were not evaluable, by the way, had tumor regression. Of the six that were evaluable, we had two PRs and one CR, all confirmed by RECIST. We have continued to monitor the patients that were on study. One of the PR patients that in November had a -47% tumor reduction is now at -79%, I believe it is.
This patient had gone off the study and actually came back on once we had validated the 4.4 and had significant additional tumor regression. They continued to respond well to this therapy. They've now been dosed a total of 15 plus months and are tolerating it extremely well. I raise this while we all want large ends when we draw conclusions. In early stage phase one, we have to work off of some anecdotal evidence that gives us clues that we're going to validate. This patient is extremely important because this patient is telling us, number one, 4.4 is a great efficacy dose-responsive dose. Number two, durability on therapy and durability of response. Number three, that this patient doesn't have resistance from prior dosage. By the way, all six of our patients that we evaluated had regressed on prior taxane use.
We are able to get tumor regression on patients who progressed on prior taxanes. There are some real important insights here that we are looking forward to proving out with higher ends in our data set later in the year.
Yeah. It's a really encouraging update. I think I like the way you spoke about the anecdotes as helping tell the story here because one of the questions I've gotten from investors is this being a big basket study of multiple tumor types, you have this really encouraging preliminary signal pulled out in head and neck. What is there to believe that that's real? I think that's a real signal. I think you point to one of those. Maybe talk also about the breadth of activity you saw in the different subgroups and across prior treatments and how you think about that as validation as well.
Yeah. The way we think of our job as drug developers, it's like peeling the layers of an onion, right? You've got to peel away to get to the proof points that are validated. You don't get them all at the very first data run, especially for new technology, especially for a new target. It's the first time you're learning things. What we were extremely pleased and what we expected because of the extracellular nature of our target that we would see broad tumor impact, and we did. Now, why did we see variability in that? Why was head and neck so pronounced versus some others?
This is where AI-enabled pathology gets really interesting because we can now visualize at the cell morphology level what is going on in the tumor microenvironment, what's going on in the stroma, and are there differences across tumor types, and how do those differences relate to our therapy? I can tell you, while we'll be commenting on this over the course of the year, we do see things. We've learned things through that AI morphology observation. There's pieces of this related to, for example, protease concentration, potentially protease geography, where it is in the cell, in the tumor cell, in the tumor microenvironment, elements related to pH level, oxygenation level, elements related to different stages of maturity of the stroma. These are all things that we are looking at relative to clinical response.
We also see that a head and neck tumor behaves very differently than a breast cancer tumor, behaves very differently than a sarcoma tumor. You may see certain types of themes in different tumor types of these other factors that we do not typically look at as, quote, biomarkers, but that actually inform the likelihood of susceptibility to a therapy like ours. It absolutely is not an accident that we have strong head and neck data. We have an emerging sort of body of evidence from a mechanistic and a biology perspective that we are going to be putting out over the course of the year, starting with two exciting AACR papers at the end of this, or posters, excuse me, at the end of this month. We hope that investors take the opportunity who are there to come talk to us. Our translational team will be there.
Obviously, they'll be available online, and we're happy to do follow-up from there. That wall of evidence is getting built as to why, Brad, these signals are believable. They're not just a statistical anomaly. There's actual clinical, biological, structural reasons for some of the themes that we're seeing in our data response. That's great because that now can drive us or direct us as drug developers in where the most fertile ground is to support patients.
Right. I know you previewed this a little bit at the top, but what is the plan to validate this signal with the dose expansion studies? Importantly, too, how are you incorporating a vision for how the treatment landscape in head and neck cancer could change in the next few years?
Yeah. We are very pleased, and we have been pleased to watch the activity in the space. It's great to see Bicara Therapeutics coming in with a next-gen EGFR target. I mean, it's a well-known target in head and neck. It's still the standard of care in Europe and ex-U.S. geographies. If you actually step back and think about it, Erbitux as the lead agent in ex-U.S. markets, that was approved in 2000, and here we are in 2025. Patients need next-generation therapies. It is very encouraging to see Bicara Therapeutics taking the lead in that mechanism. There are many segments in the head and neck space that need serving, right? We are seeing observations for those two competitors of potential differences based on HPV status. We have had positive responses in both HPV positive and HPV negative.
There is no reason to believe with our mechanism, HPV status will influence responsiveness. Of course, we're going to try to prove that out with a higher end. Our development strategy is very much taking into account where the standard of care is going to, not just where it is now. In monotherapy, we are developing the therapy in two arms of patients. The first is the patients who are resistant to PD-1 plus platinum, so current standard of care here. The second group is patients who are resistant to EGFR plus PD-1, so primarily ex-US. Those are kind of the two standards of care that are out there.
This design allows us to be agnostic to Bicara Therapeutics, whether they both win or whether one of them wins or whether neither of them wins, because we are going to be capturing the EGFR-resistant patients, whether they are resistant to cetuximab or whether they're resistant to Bicara Therapeutics. We're doing that by leveraging our clinical site network where the Bicara Therapeutics studies are going on right now. Patients who participate in those trials who regress will be available to us to participate in our second monotherapy arm. That's why that data is expected more first half of next year rather than in tandem with the first arm, which is platinum plus PD-1 resistant, because we want to capture some direct Bicara Therapeutics-resistant patients. We know that's where the market's going, assuming they get approval.
It only helps us to see how our agent behaves in those patients now. At the same time, the other arm, if neither of them get approval, we still have that big platinum plus Keytruda arm. Either way, we're agnostic to what goes on. It is a very unique opportunity in the marketplace where you can actually set up a trial where you're agnostic to the outcome of the competitor who's a step ahead of you. That is because our mechanism is differentiated, and we can fill a niche. We can either compete with them or come in right behind them. Either way, there are plenty of patients that need to be served. You'll hear us talk over the course of the year more specifically on those patient pools. That is sort of bucket number one. Bucket number two is the combo therapy with Keytruda.
As you know, Brad, and I think as people who followed us know, we were really pleased to announce the supply agreement with Merck last year. We were one of the handful of companies that they provide Keytruda for free through this supply agreement. I mean, that's a huge, huge multi-million dollar financial benefit to us to not have to pay for that. Marison Bicara don't have that deal, by the way. So Merck saw something in our mechanism, in our data that got them excited to provide us this for free. That's allowing us to really have a robust MICVO plus Keytruda trial. In the dose escalation phase, that's open to multiple tumor types. It's not just head and neck. We'll get some early signal finding in combo through that. The backfills will be focused on head and neck.
I raised that about the dose escalation because that signal finding through dose escalation in other tumor types like breast or sarcomas could form the basis of a lifecycle management plan later. After we prove out in head and neck, we can then, on the backs of investor confidence and additional funding, consider what is the next broad focus of the development plan in other tumor types. For now, we are highly focused on head and neck. I say that because I want to make sure the door is left open for further development of tumor types because going back to the beginning of this conversation, the wall of evidence that Enhertu is putting out mechanistically around cleavage of the ADC and the tumor microenvironment, and that's agnostic to a cell surface target.
We look at Enhertu with its broad footprint, we could be the natural next Enhertu in the ADC space. We're set up as such based on our biology, based on our mechanism. Frankly, that's what I think the strategics that we dialogue with see as the potential in MICVO. I think it's what the market is missing. Our job is to continue to build the wall of evidence, peel the layers of the onion back, put those proof points in place on the biology, the mechanism, and the clinical data. That's our job over the course of the year. I think the data sets that will come clinically by the end of the year will be well contextualized by the mechanistic work that we're doing and, frankly, the learnings that Enhertu is putting out for the field on the mechanism.
Right. Unfortunately, we're running up to time now, but I want to thank you so much for joining us. You've previewed a lot in terms of the proof of mechanism and the clinical data that we should expect from your company over the next 12, 18 months. I'm really looking forward to watching that all progress. Lara, thank you so much for joining us.
Thank you. Thanks for having us. We look forward to staying in touch as this all comes out.