Eupraxia Pharmaceuticals Inc. (EPRX) October 22, 2026 Conference Transcript

All right. Good morning, everyone. I'm Albert Lowe, Senior Biotech Analyst, and welcome to our Craig Hallam Biotech Conference. For our live viewers, please feel free to submit your questions on the webcast interface, and I'd be happy to include them in our discussion today. I'd like to introduce James Hellenwell, the CEO of Upraxia Pharmaceuticals. Upraxia's lead asset, EP-104GI is in Phase II trials for eosinophilic esophagitis. They have a very important upcoming catalyst, which is the data readout from the randomized Phase IIb study that's coming up in December. But before we dig into the data, James, maybe you can start off the discussion by just giving us a background on EOE and telling us about your practice approach.

Yeah, absolutely. First, Albert, thank you very much to you and to Craig Hallam for having us today. It's a real privilege to be here. So EOE is a really interesting disease. And in some ways, it's sort of come out of nowhere all of a sudden. And it was thought to be an orphan disease as far 2022, 2023. So people talking about prevalence numbers of the 150,000 in the US, that's changed very dramatically. The last publications looking at the disease have had numbers in the 450 to 550,000. And of course, by the time you're publishing something, you're really looking back two years already. And so estimates are saying that by 2029, 2030, it should be close to a million patients in the United States. Now, a lot of that is actually diagnosis. And it's really interesting because EOE is an inflammatory disease. It's type two inflammation. And frankly, probably is a little misnamed. And that, when we talk about approach, is kind of hidden in there. When we look at the tact most people have taken to date, it has been, let's get rid of eosinophils. If we get rid of eosinophils, the disease will get better. And the answer is that eosinophils do not seem to be actually correlated to clinical outcome. And we've seen lots of examples of that. I think Alicose is a very stark example, where they had 90% plus of patients who actually achieved peak eosinophil count remission, so essentially eradicated the eosinophils, and yet were not achieving a good clinical outcome. And yet we see other places where you see quite a good clinical outcome with very little reduction in peak eosinophil count. And so when we looked at this data, including Dupixin's early phase two data, where they reported a clinical remission rate of 39%. They reported a histologic remission rate of 59%, yet a complete remission, meaning both, of only 11%, really highlighting that disconnect between eosinophils and clinical outcome. Our sort of look at that was you need to move upstream and not just target one thing. And by using a corticosteroid, in our case, fluticasone propionate, which has been used off-label in this disease along with budesonide, which is both on and off-label use, steroids are very upstream in this type 2 inflammatory process. And so they're not just going to hit IL-13, IL-5 or IL-4, they're going to hit all of them. So sort of like T-slip is upstream, steroids are even more upstream than a T-slip. And our approach is to be able to deliver it as opposed to orally, which is how things have been delivered to date, which means it's very difficult to actually coat and maintain contact time with the esophagus. We actually do a submucosal injection delivered by the physician during an annual endoscopy that patients are already going for to be able to ensure that the drug is releasing right in the tissues where the disease is and having an impact there. And so we think it's a unique approach, but with a very well understood API and a well understood mechanism of action.

Right. Great. I think that, you know, and perhaps what you reference this, especially this clinical remission versus histologic outcome kind of shows how this is a challenging setting. And maybe that's a good segue. I know a few weeks ago, you shared a dose escalation study update. Maybe we just highlight some of the new findings from this latest data.

Yeah, absolutely. So we're seeing quite a few really interesting things with our dose escalation study. And so really, that was all about trying to triangulate to the right dose for the second part of the study, which we'll be having the data as you pointed out in December. And what we've really noticed to date is that while we've had a very good impact on eosinophils and we have reduced eosinophils, we do not actually really see a clinical correlation when we look at the individual patient level between when we reduce eosinophils and when patients get better. On the flip side of that, histologically, we're seeing very, very good response. So in EOEHSS, which does measure eosinophils, but it also is measuring other things that are related to inflammation as well as fibrosis. There, what we're seeing is a very strong anti-inflammatory effect. We are also seeing a very strong anti-fibrotic effect. And what we've seen in our latest data is that that effect is maintained over the long term. So in other words, our latest biopsies in that trial are at nine months, and we're seeing that effect, that reduction in inflammation and fibrosis maintained all the way to nine months. From a clinical symptom perspective, what we've been seeing is a maintenance of clinical remission where 64% of our patients are still in clinical remission at 52 weeks. And that's where we would expect that essentially redosing would be because most patients, about 60% of patients today come for an annual endoscopy, just for surveillance. And we imagine that that will rise when a therapy is actually tied to it.

Okay, great. And I guess this remission where you mentioned 64%, I guess, is that maybe the right comparison to this 39% for Dupixent that you mentioned earlier as well?

That is the same metric, obviously different studies, and so not a head-to-head comparison, but both looking at the Straumann dysphagia index. That said, most people watching this who've been looking will probably see that DSQ or the dysphagia symptom questionnaire is really the most common PRO that people look at for how patient symptoms are. We are looking at the DSQ and know that it'll be a pivotal focus for investors coming in December.

Okay. Well, we're talking about, I guess, specific outcome measures. What is the approvable outcome for a potential phase three study for EOE?

Yeah. So it's really interesting. We actually think that that's adapting. So to the earlier conversation where everybody felt if you got rid of eosinophils, you would make patients better, the classic outcomes that a bridge agreed to with the regulators have been piqueosinophil count remission and a PRO. In this case, it's been approved in only two products, both using the dysphagia symptom questionnaire. So I think on the PRO, it's fairly clear the dysphagia symptom questionnaire is very approvable and a great endpoint. On piqueosinophil count, I think that there's a concern that it's really just a surrogate marker. And we know that the FDA doesn't love surrogate markers and instead would prefer evidence of disease change on histology. We've certainly heard from KOLs that they've been arguing to the agency that actually EOHSS is probably a much better marker because it is actually directly measuring the fibrotic state of the esophagus, which is one of the important features of the disease that also drives symptoms. And we've seen some phase four data from dupilumab showing exactly that, really a connection between improvement in EOHSS resulting in an improvement in esophageal distensibility. And those are really the features that lead to a lot of the symptoms along with inflammation. So to come to your question, we believe that likely what's coming is a PRO being DSQ and you will have to have a histologic outcome. We think it's probably going a move to be something like EOHSS, eosinophil counts will then just simply become a key secondary.

Okay, great. It's great that we can get a lot of regular updates from the dose escalation. Do you plan on giving another update this quarter?

We may have a little more this quarter, although it may end up being after our phase 2b update, which is coming in early December.

All right, great. And I guess, you're talking about the endpoints, they really shared a lot of data from the dose escalation study. So as we look ahead to this December readout, which of these data points perhaps provides the most significant read-through into the dose optimization phase 2B study?

So I think that there's two that I think, clearly, I love all the endpoints. I think there's value in all of them. But I think that there's two that investors are probably really focused on for good reason. One is the DSQ, because it allows for really sort of cross comparison across other products, what's been out there. And they're starting to become, you know, sort of a level of where a therapy is really effective and approvable in DSQ. And we haven't shown a PRO relative to placebo yet. So I think that delta from placebo will be, you know, a key endpoint people will be looking for. And I think EOHSS is the other. You know, are we having an impact, as we've seen in our previous study, but is that impact happening in one or both of our doses and in really showing that we're getting the histologic change that we've seen and hope to continue to see? Obviously, other important endpoints, EREFs, which are really the visual scoring of both inflammation and fibrosis. And I think people, you know, there's still interest in eosinophil counts, but certainly not focus on eosinophil counts.

Okay, great. You announced last week that the study had completed enrollment, the FACE-2B study. So maybe you can tell us just briefly about the design of the trial with some of the statistical assumptions and what the statistical analysis reviews have found.

Yeah, absolutely. So it's really, it's 40 patients per arm, and so that's fully recruited. There are still a few patients in the funnel, so there will end up being a little bit more than 120 when all is said and done. Um, when we look at, at the way the study is designed, so it's essentially, um, a control arm and that control arm is getting the 20 injections of our vehicle minus the drug. Um, and so, and patients are blinded and injectors are blinded. Um, assessors are, are also blinded and they are not the injector. There is then a dose A and a dose B. Dose A is six milligrams per injection site. for 120 milligrams, and dose B is eight milligrams per injection site for a total of 160 milligrams. When we were designing this, we really look at the real difference between doses at the high end as really being about durability, more than necessarily, you know, sort of a particular change in those outcome measures in the early place there. So we then have follow-ups with PROs throughout that period. And then at the three-month, six-month, nine-month, and one-year mark, we also have endoscopies. So that'll be measuring EREFs and also getting biopsies. And then at the six-month time point, patients who were randomized to the control group will remain blinded, but they are getting crossed over to one of the two active arms. So patient remains blinded. At that point, the injector is unblinded, but all of the assessors remain blinded as to what's going on. So there will also be a limited amount of crossover patient data that we hope to have in December, which will be quite interesting. And then we follow all the way out to 52 weeks. Very interestingly, based on the data that we have seen from our other trial, the dose escalation, where we're very clearly lasting 52 weeks and seeing the histologic endpoint and the clinical endpoint, we are actually extending this trial beyond 52 weeks. And so we're actually adding another year of follow-up. And the intention with that is that these patients, then we will watch the return of their symptoms and their histologic disease return to be able to time the repeat dose. Once we've started doing repeat dose, we anticipate then we will align that with the one year and then be able to have all these patients essentially then become repeat dose patients over time, so that by the time we get to NDA, we will have had patients who will have had several years of follow-up and several doses.

All right, great. Great to hear the extension design of the study, so you can really zero in on maybe the true duration here. And the duration is the long duration after single administration is a big aspect of this product. You talked about some of these time points, maybe you can tell us a little more about, I guess, what exactly will be in this December interim readout, and maybe at this point, what you see as the bar for success for efficacy and durability.

Yeah, absolutely. So when we look, I mean, the primary focus is at our primary endpoint, which is six months. So at six months, it's essentially the halfway point in the trial. So we'll have about 60 patients who will have achieved that, maybe a few more than that, but who will have achieved that six month time point. And when we look at the powering for the interim, the SDI is about 80% powered to be able to detect a 1.2 point difference with a standard deviation of 0.8. So what is that? That's really sort of a difference that is in line with the sort of response that's seen today with other products. DSQ is a little bit of a noisier baseline. So it's a little bit harder to power. The standard deviation tends to be quite high when you look at other trials. It's ranging between 11, 12 to 15. And so based on that standard deviation and the noise in that, we're 70% powered to detect a 10-point difference improvement from placebo at the interim. And so really, when we look at the PRO powering, a really good result here would be if you have separation from placebo, so you're clearly showing a very strong trend in your PROs that you're different than placebo, but you haven't quite achieved power yet, that would say that you're right in line with other things that we have seen in the literature and really other products. I think if you've got power, as you can see from the way that's powered, if you've got statistical significance, you've clearly got a very strong effect size and probably a much more consistent response with less noise. And it also then would be a very strong indicator that your overall trial is clearly very, very positive. When it comes to the histologic outcome, when we look at EOHSS, again, that's about 80% powered to be able to detect a difference of about 0.18. Again, that'll be a little bit baseline dependent. And so that's assuming looking at a baseline entry point of about 0.5 to 0.55, which is what we'd seen in our previous trial. And so again, really the same thing there. If what you see is a very strong trend to improvement, but without statistical significance, that's still a great outcome. If you have statistical significance, then really you know that you've got a product and you really start leaning into your phase three final design dose selection and moving forward with that process.

All right. Thank you for that very excellent summary. I know you also plan to share some baseline patient data ahead of the December readout. Mimi, can you tell us generally about the patients that were enrolled? And perhaps this would be a good segue into where the treatment paradigm EP-104 could fit in.

Yeah, absolutely. So we want to be sharing prior to the data, the baseline characteristics, um so that when there's a lot of data that's coming in december um so we sort of want to um have people understand who this patient population is um and so uh so we look forward in about a month a little bit less than a month to actually sharing those those baseline characteristics when we have them all together for for who's going to be in this trial part of the reason for that is that severity matters and that's been very clear from from other trials and in ours. You know, what sort of difference can you make? And so how can we then use that to refine expectations for people as to what's coming in, what may be coming in December? And when we think about that, though, there's a couple of aspects. So number one is where do we fit, you know, according to KOLs, what we're hearing from payers in the treatment algorithm. them. So today you start with a PPI, perhaps diet elimination, although that is not a very popular treatment for most people. And then most people then move to an oral steroid. The problem with the oral steroids is not that they don't work. It's that taking them properly and consistently is very, very difficult. And both American College of Gastroenterology and UEG both just happened. And there were quite a few papers there that were really demonstrating this phenomenon of non-compliance with oral steroids and the fact that then symptoms and fibrosis return very, very rapidly. So we look at our product really as being a replacement for that step edit, which is replacing the oral steroids with a physician-administered, really compliance guaranteed, one-year steroid. Ideally, we believe the data is supporting that to date, so we'll have to continue to see that. And so as we look at that, we believe that sitting in there pre-biologics is really where this product sits. Now, there is an interesting piece there. We have allowed into this upcoming trial some patients with mild histology. So these are patients who are not in our primary analysis group from a histologic perspective. So they've got moderate to severe symptoms, but they have very mild piqueosinophil count and very mild inflammation and fibrosis. They really represent the patients today who are on a PPI. Our question of those patients to decide whether in phase three, we might actually look at that population very directly with properly powering that arm is, can you stop the progression of fibrosis that we know occurs in these patients by treating them earlier versus waiting until they have fibrosis before you start treatment? So we think that's not going to be powered, but we look forward to sharing some of that early data of what we're seeing and hope that that's going to be interesting, partly because then the question becomes, should you actually be replacing a proton pump inhibitor earlier? And that has a significant impact on when you would then start treatment and really changes your TAM by moving earlier in the disease.

Right. Okay. Maybe the last few minutes here would be nice to touch on what the commercial setting in EOE looks like. So I believe Dupixin is the only approved drug for long-term treatment for EOE. So do you have any idea of what Dupixin's sales are in EOE? And along these lines, how could EP104 be priced relative to Dupixin?

So on pricing, we hired a third party who went and did a lot of payer analysis based on a TPP. That TPP was essentially efficacy in line with approved products today, but lasting at least nine months. from a physician injection. And really the median answer there was a $50,000 price point. So that was very clear. And also in step edits, it was post-PPI and pre-biologic. So that really aligns with what we were just talking about. When we look at Dupixent, all I can report is what they themselves have reported. So about 10 months ago, they had reported that since launch in 2022, they treated 54,000 patients with Dupixent for EOE. That was just updated. So just 10 months later, that number is now 76,006 since launch. So that's a very significant increase in just 10 months in the number of patients who are on DUPI. And I think it reflects that patients are really needing a good long-term therapy. Dupixent has been wonderful for patients. And those patients that it's working for has provided them an alternative. And so I think we're seeing growth. But to recognize that that is really not large market penetration to date. So clearly a very successful product in helping patients. But that's what's publicly available that we're aware of.

All right, great. And I guess, yeah, sign points to that kind of rapid growth that you discussed earlier as well. So maybe just in this last minute here, can you walk us just through the cadence of events leading up to the December data readout?

Yeah, absolutely. So we're going to be sort of out at some conferences and interacting with investors. There is, as we briefly chatted about, the baseline characteristics will be revealing in mid-November, and then we expect in early December, so within that first two weeks, that we're going to have the Phase 2B interim data release in there. So a very active time over the next seven weeks as we lead into data, which is, you know, about seven, eight weeks away. And then, you know, we think that'll be a very interesting way to cap off 2026.

All right. Excellent. Well, I'll definitely mark early December on my calendar as a critical catalyst for the story here. I'm really looking forward to that. And yeah, thank you, James, for the great discussion this morning.

Thank you, Albert. Thank you to Craig Hallam. And for all those who are out there watching, we really appreciate it.

Absolutely. Thanks, everyone.

Thank you.