Hello, everyone. Welcome to our fireside chat with AN2 Therapeutics. It's my pleasure to be joined by Eric Easom, CEO, here to give us an update. Thanks so much. Why don't we start with a brief level set, just to make sure everyone has a good basis to understand where you are in your development journey. Can you provide us with a brief high-level overview of where we are with EBO-301?
Sure. I'd love to. Thank you. Thank you for the opportunity to be here. We're a boron chemistry company focused on infectious diseases and oncology. Our lead program is a compound called epetraborole, and it is being studied for nontuberculous mycobacteria lung disease caused by MAC, or Mycobacterium avium complex. Epetraborole is a novel mechanism oral compound, and we've been studying that in a seamless phase II, phase III pivotal study. The design is basically our drug on top of a background regimen compared to placebo on top of a background regimen. The phase II part of that enrolled 80 patients, randomized 1:1, and read out in August of last year.
There we saw clinical efficacy in the top line primary endpoint and also showed that we could validate the PRO, which was the primary endpoint called MACrO2 , was the PRO that we developed in-house. We did that. It was well tolerated, as expected, but the culture conversion rate was lower than expected. We were basically similar to placebo at that time. We knew to continue on with the phase III, which was running in parallel, we had 97 patients in the phase III at that point, in addition to the 80 in the phase II, would be a pretty big endeavor to show a difference enough to have the power to show that using a discrete kind of top line clinical endpoint.
We decided to stop the phase III, slow down, and really look at the data, see what we could learn, and then go forward with that. We had a pre-specified secondary endpoint called quality of life, bronchiectasis, respiratory demand. As a PRO, we were nominally stat-sig on that in the 80-patient phase II trial. Around that same time, Insmed announced that that was their primary endpoint. Insmed has the only FDA-approved drug for this indication. We had that. We took our MACrO2, that was our primary endpoint, and turned that into a continuous scale, not a discrete scale, a 100-point scale. We also were in post-hoc analysis for stat-sig on that. We had two clinical endpoints that were nominally statistically significant. This became pretty interesting. That is essentially where we are.
We just announced that we've picked QOL-B, respiratory demand as our primary endpoint. We've updated the statistical analysis plan. We submitted that to the FDA. We're going to read out the phase III data, the 97 patients in the second quarter of this year. It's really quite interesting, I think, from our perspective.
Right. You had really sick patients in the phase II.
Thank you for bringing that up.
Probably really sick in phase III, very advanced cavitary disease, resistant mutations, et cetera.
All the above, yeah. I mean, there were three really significant features. One, the average duration of disease was nine years. We had more than 50% cavitary and fibrocavitary, which is the most severe form of disease. We had 2/3 of the patients were resistant to amikacin, which is Insmed's drug. 20% were on amikacin itself and refractory to amikacin. These were substantially worse, more severe than what had been previously studied. Our whole hypothesis was, let's run a study like Insmed's and show superiority and get this drug out. In reality, they've been on the market for six years, and the patients that are available for the study were much, much more severe.
Why do you think you saw such a different picture according to the culture conversion, which did not work versus the PROs, which did?
Yeah. I mean, this was a vexing question for us, too, when talking to the KOLs. As soon as they saw the enrollment criteria, again, we had nine years on average. Insmed had three years on average of disease. We had 50% cavitary. They had 12%. We had almost 2/3 of the patients were on multi-drug resistant regimen. We are putting a single active drug on top of a failing regimen and expecting that. As soon as they looked at that, they said, you do not even have to tell us. We know you are not going to convert these patients. I think it was a lot of that. Now, just to remind you, Insmed's culture conversion rate was good enough to get approved. It was 29% versus 9%, but that is pretty low.
They didn't show any clinical benefit in their phase III study in the less severe patients. We do think that epetraborole is an antibiotic. Sputum is a very noisy, messy signal. You don't know the quality of the sample. Did it spit or was it sputum or kind of all these variables? It's difficult. Unless you culture convert, it's hard to tell. I think it was the types of patients, but we're pretty excited to think that the clinical, I should mention the sole primary endpoint for FDA guidance is a clinical endpoint as measured by PRO. We have two of those. It's pretty clear that secondary endpoints aren't required for approval. Things like culture conversion wouldn't be expected.
OK. Very helpful. How is the phase III study powered on the QOL-B primary endpoint?
The phase II was powered to validate the PROs, so MACrO2 and QOL-B, both of our PROs.
Sorry.
Yeah. No, that's fine. They were powered. We had 80 patients in that 1:1. In the phase III, we have 97, and it's 2:1 on epetraborole. We think there's a reasonable chance that we've got the power here to show a difference. Ideally, we would have a few hundred patients, I think, but we've got what we've got. We've stopped the study, and I think we'll see how that goes.
OK. Does it matter which domains of the QOL-B generate more of the power?
Yeah. I mean, I don't think so. I think if I understand your question, the respiratory domain, which is a segment of QOL-B, is the one that's most relevant to NTM, and that's the one Insmed and the FDA aligned on. That has nine questions in there that are very specific to the NTM disease, like coughing with sputum, coughing without, chest tightness, things like that. There was one question that was removed in their agreement, which was the color of sputum. Insmed's only using eight questions. We think that's the one they've removed. We don't formally know that, but we believe that's the one. For us, either with or without that doesn't matter. We're comfortable. We picked QOL-B just to anticipate maybe what you might ask over our own internally developed MACrO2, just because it seems to be the one that Insmed used.
It's commonly used in other NTM studies in the literature with academic groups and so forth. We're indifferent. Both of them showed nominally stat-sig, so we could use either one, but we think it's probably better, just simpler with the FDA to pick the one that's most common and is already precedented by their agreement on the primary endpoint for the ENCORE study that Insmed's running.
OK. Great. When will you unveil the data?
Q2 is what we've said. We submitted a revised SAP in our press release a few weeks ago. We're giving the FDA some time to look at that. They normally wouldn't comment on this unless they have any major issues, which we don't believe because they've already agreed to it with Insmed. We'll give them time kind of to respond, and then we'll unblind it. It could be in Q2, early Q2 time frame. Right around the corner.
Yep. OK.
Once we have that, we decided to go ahead and do that rather than spend all the time writing briefing books and getting in front of them with all these data. We have cleaned the phase III database. We locked it. It is ready to go. We essentially just need to, at some point here, push the button, if you will, in simple terms. We wanted the power of two studies. If both of them are showing the same signal, then it is pretty confident that that is not occurring by chance. These patients are, it is life-threatening. They are kind of at the end of their journey, and physicians are just trying to maintain their health. Really, the KOLs want a drug with this profile, oral, well tolerated, and shows some clinical improvement because they do not believe they can culture convert patients at this late stage.
Right. Previously, you paused things when you were cleaning the phase II database and noticed some things that looked curious. Have you detected any similar patterns in the phase III?
Yeah. We did pause the enrollment of the phase III in February of last year, driven primarily because we saw overall aggregate blinded low culture conversion rate. We're all ID people. We want to kill the bacteria. That's how we're all predispositioned to do. I think we paused that. I think the trends we think in the phase III are similar in that respect. With the clinical piece, because it's the primary endpoint, you typically would not look at that. We have not assessed aggregate levels. It'd be too hard to tell anyway because the signal is much higher than the culture conversion rates, which were pretty low on aggregate, like 10% in the total.
OK.
Yeah.
What degree of magnitude are we talking from the phase II on QOL-B benefit?
Yeah. At six months, we were a 7.2 improvement versus a 0.2 for placebo. There is a nice separation in the curves between the two at three months and at six months. It is with a p value of 0.03. It was a pretty clear separation from the beginning.
OK. Did you notice any patterns in the patients who tended to do better on the PROs versus those that did not based on any baseline criteria in particular?
Yeah. I think not surprisingly, maybe, as those who are worse off tend to do better because they have more capacity to improve. I think given those trends, we like the way the phase III sits because the baseline characteristics are equal or worse than the phase II data. I think that plays well for patients. If you start lower on the scale, you just got more room to feel better.
What about, can we talk about the microbiology a little bit and the potency of EBO-301 relative to ARIKAYCE, for example? How does it perform on MICs and things like that?
Yeah. In all of our preclinical panels, which were done on a few hundred isolates from the U.S. and Japan, the MIC90 was, I think, 4 micromolar, somewhere in that range. The break points on ARIKAYCE are 128 micromolar. You can just get more in the lung, but it's just got a higher break point. It's always a matter of how much can you get safely into the person to kill them. The thing that was really interesting is two-thirds of the patients in our trial were resistant to amikacin, meaning they had an MIC that was higher than those break points, which is shocking. A lot of patients out there, well, it's not shocking in that they've been on treatment for nine years and probably had everything under the sun thrown at them. Bacteria that has had time to evolve and change.
EBO-301 is a fairly potent drug, similar to what you see with the macrolides, which are the workhorse with MICs in the kind of MIC90 ranges in the 8-16. These skewed a little bit higher, but I think an important point we also haven't mentioned, we didn't see a lot of people speculated that because this drug is monotherapy in UTI trials previously that showed great efficacy but had a few cases of breakthrough resistance while on treatment. We didn't see any of these big jumps in MICs that we saw previously or that we can see in vitro; we can create the conditions as monotherapy to see. When you use combination therapy, even if there are background drugs that are resistant, there's enough activity that it prevented that it appears from happening. That's a really important finding, I think, from this study.
OK. Where would you see EBO-301 fitting in in the treatment landscape? These are very refractory patients. That's where ARIKAYCE is currently approved. First part of the question is, within refractory, where would it fit in? What are the plans for, do you have plans or aspirations to develop it in front line?
Totally. Yeah. I mean, that's going to be our initial indication is the data we have. We're happy to play in the refractory space. The feedback we get around a profile like this is physicians don't particularly like, I don't know what the word is, forcing older patients with challenges to take a drug that's difficult to administer through an inhalation and all the toxicity. They like an oral profile. I think that's well tolerated. I think that still positions quite nicely relative to ARIKAYCE, which is currently selling what, $400 million in refractory. It's probably underachieved because it's a challenging drug to use. We would ultimately see this profile best fit in front line, in the front line setting. That's where we would go.
The other huge unmet need that is not talked about a whole lot, but we just published MIC data in abscesses, and we've published in vivo data as well in preclinical models. It's a very favorable profile. The studies are more challenging, but we also think that's a billion-dollar-plus market just in the U.S. alone for abscesses. These are high unmet needs, and patients don't have a lot of options. That is where we would take the drug, same as we had planned before.
OK. Help us think through the scenarios after you unblind the phase III data and presumably engage with the FDA. What do you think the likelihood that you could actually file on this data is versus needing to do another study?
Yeah. That's a tough one. I mean, I think in best case, we've got two studies. If they show good results, these patients, this is a serious life-threatening condition, and they don't have any options essentially at this point. They're refractory even to the only drug approved. I think we've got a pretty good case to say patients should have access to this. We're happy to run post-approval confirmatory studies like Insmed has done in the front line setting. I think what are the odds of that? I don't know. It's a small, it's two studies showing that. It's unlikely by chance that they would show that, but the safety database is small, but they're high unmet needs of patients. We think that's one.
The other is post-confirmation study, or the other scenario would be we have to run another study, but we would have two studies. The risk profile changes quite favorably in terms of being able to show that again in a third study would be a lot less risky. I think any of those are good, but for patients' sake, I think the argument would be they have zero options essentially at this point.
Right. What is the safety and tolerability profile like?
It was good. I think it was well tolerated. Almost all of the adverse events were considered mild. The most prominent is GI, which is common in antibiotics. It's generally transient as it adjusts the microbiome or so forth. We see that, but they're mild, didn't lead to, I don't think, any discontinuations that I can think of. The other one is there is a hemoglobin reduction that also is very slow in reduction and mild generally. It just goes down over the first month, and then kind of there's a compensatory effect, and it kind of stabilizes. When you take the drug off, it returns to normal. I think those are the two well-known described safety signals, none of which are, I think, all that concerning.
I think that's why we say it was generally well tolerated and feel comfortable with long-term use of it.
OK. Let's turn to the rest of your pipeline. How about you talk us through AN2-502998 and Chagas disease?
Yeah. No. I mean, we are a boron chemistry company. This is a compound we discovered when we were at Anacor. Chagas disease infects an estimated 7 million or more people, mostly in Latin America. The vector, the kissing bug, infected kissing bug is present in Florida. Might be under your chair.
No.
All over the southern United States, there's over 300,000 patients in the United States with Chagas. What happens is the parasites go into your muscle. It's a silent but deadly disease, and 30% of those develop pretty severe cardiac complications that will lead ultimately to aortic aneurysms and heart failure and things like this. It's a silent killer. We have the only drug we know that cures chronic Chagas disease. Most people don't know they have acute.
You get a mild flu-like symptom, and they think it's a cold or something else. They don't diagnose it. They have chronic Chagas. We think there's a significant number of those patients in the U.S., and there's simple diagnostics, the serology test to test and treat, and that would be the method. We've, in preclinical, naturally infected monkeys who get this in Texas at a well-known cancer institute. They eat the bugs because they're outdoors. We've been 100% curative. We've been following this group of monkeys for over five years now. We're doing a second study now, but it's the only drug that does that. Merck had a compound that failed in the clinic a decade or so ago. Eisai, the Japanese pharmaceutical, also had another. Posaconazole and ravuconazole were the two compounds out of the same class.
Benznidazole is approved for acute Chagas in the U.S. It's highly toxic, and it cures in the monkey model at around 30%. It's not, every less than 1% of patients use it for chronic Chagas. There's really no treatment for disease that is, it's the leading cause of infectious disease death in South America. I mean, this is a massive problem that most of us haven't heard, but there's a lot of it in the U.S., Canada, Europe, places where Latin American people go to live. There also is transmission in the U.S. We're confident. We have a drug for African sleeping sickness that we were part of the discovery of. It's an oxaborole, single oral dose cure for stage II African sleeping sickness that's before the EMA right now, same mechanism of action, completely novel mechanism.
I could go on and on about boron chemistry, but this compound will go into phase I this year, phase II next year. Within our cash runway, we'll have phase II proof of concept. That's Chagas. We're also working on, we announced we've got two cancer projects. One will go into development this year. That's a completely novel, very exciting target we haven't revealed yet for competitive reasons, but we believe we have a best of class profile. We know that because we have very potent compounds with better PK and very selective for the target that's not on any drug yet. That's a very big value driver. We have a completely novel first-in-class that's kind of bubbled up through our chemistry that's also really exciting. We've got that and some other projects. Boron chemistry, we're good at developing and discovering drugs.
We put 20 drugs in development from Anacor in lots of strange global health diseases, but the platform is applicable to inflammation, oncology, infectious diseases, and has a lot of power. We think the market misunderstands our pricing, but we have NTM as an inflection near term, and we've got Chagas, and we've got our pipeline. I haven't even mentioned melioidosis, which is another indication we plan to start a phase II proof of concept study later this year on. I think we're pretty excited about all this in terms of creating optionality to address some pretty serious problems and the cash flow to do that.
Can you help us understand the resources the company has and the runway it provides?
Yeah. We had a restructuring at the top line data to really make sure we had time to assess the data and invest in our pipeline. We have a cash runway to the end of 2027 to advance all those topics on the NTM front to get to the top line data. If the data is good, I think we'll have a different picture altogether with a phase III asset that could be potentially registerable.
Yep.
We're feeling good about things. It's been an up-and-down story, but I think as we look back, I think we've made the right decisions to really try to give ourselves optionality and a path forward.
Excellent. Good luck, and we are waiting with bated breath to see the results in the second quarter.
Yeah. No, thank you. We are too. Thank you very much.