All right, I think we're ready to get started with our next session. I'm Josh Schimmer from the Cantor Biotech Equity Research Team. Very pleased to introduce from Aardvark Therapeutics, Tien Lee, Chief Executive Officer and Founder. Tien, give us a quick snapshot of Aardvark, and we'll then dive into all things TAS2R agonist.
OK, great. Yeah.
And by the way, Tien's birthday is today. No better way than to celebrate with us, right?
Thank you. Thank you, Josh. I appreciate it. Yeah, so Aardvark is a San Diego-based biotech company. We're focused on small molecule drug development. We have a unique category of drugs. It's a gut-brain signaling drug. It's small molecule, oral, but it's gut-restricted, and it mostly stays in your gut, but it's signaling gut-brain pathway to mediate a bunch of things, including abrogating hunger.
And maybe talk a little bit about that pathway, right? So you consume something bitter. Now it's tasteless. You don't taste it. And yet it gets into the stomach. What's it signaling? What are kind of these upstream effects that ultimately impact appetite?
Sure, sure. First off, it's counterintuitive, but you have taste receptors in your gut, not just your mouth. And as you mentioned, we bypass the oral cavity, so you don't want to taste our drug in your mouth. And it's hitting enteroendocrine cells, which are the cells that help regulate eating behavior. And they elicit the secretion of a number of relevant gut peptide hormones that give rise to a signal to your brain to basically shut down the eating behavior. And we think there are two different axes of drive. One is a pleasure-seeking drive, and one is a pain-avoidance drive. That pain-avoidance drive, we denote as hunger. And that's the pathway we think we're particularly relevant in.
In terms of that signaling pathway, do we think about more of a hormonal effect where the enteroendocrine cells are releasing a substance that travels into the brain? Or is it a direct pathway to the brain via the vagus nerve? Or is it some kind of combination?
Right. We think actually the predominant effect is through the vagus nerve. So the gut peptide hormones that are released are locally hitting the vagal afferents, which then conduct to the nucleus of the solitary tract in the brainstem. And then therefore, they're then indirectly communicating with the hypothalamus that way. And that's the primary driver of the effect. These gut peptide hormones have a really short half-life in the serum, so they degrade rapidly. So we think that the effect is really locally evident.
As we think about that connection from the stomach to the brain via the vagus nerve, and as we think about the deficiency in Prader-Willi, which are the sensing neurons that you're targeting to be able to kind of bypass the pathology of Prader-Willi and then still control appetite again with this pathway?
Sure. Yeah, that's a good question. And we think there's lots of evidence of gut dysfunction, gut-brain signaling dysfunction, Prader-Willi. As you know, there's also a hypothalamic dysfunction that's evident. Their hypothalamus are smaller than normal controls at the age they get their hyperphagia. And one of the drivers of the effect is that the vagus nerve, once stimulated, projects onto the nucleus of the solitary tract in the brainstem. Then second-order neurons project onto the paraventricular nucleus in the hypothalamus, which coincidentally modulate MC4 receptor activity and therefore can bypass or supplement some of the dysfunction if one has some hypothalamic dysfunction.
A lot of the Aardvark focus around the time of the IPO was Prader-Willi. We spent a lot of time talking about the difference between hunger, just what these patients experience, that feeling of starvation, and the mechanism as it applied there. But you've recently now showed some really encouraging preclinical data in general obesity, which I think we think more of not as hunger, but appetite, that I like sweet foods, and so I'm going to overconsume as a result. How do we think about, again, the biology that now seems to not just be addressing hunger for Prader-Willi, but also appetite?
OK, great. Thanks for that question. So in the Venn diagram of effects, there's appetite and hunger can be differentially expressed and important in different parameters. We think Prader-Willi is mostly hunger-driven, but some people have different contributions of hunger and appetite. And to be clear, we think our drug elicits the secretion of a number of gut peptide hormones from enteroendocrine cells. So our drug hits enteroendocrine cells, induces their secretion of a number of gut peptide hormones, including CCK, which we think is very differentiating from other effects. But there are other gut peptide hormones released that we think we're amplifying the endogenous secretion of as well. And our latest data show that the combination with sitagliptin, a DPP-4 inhibitor, which should enhance certain gut peptide hormones, of which CCK is not one of them.
The fact that that combination is synergistic demonstrates that there's more than just CCK. We think we could be amplifying the other signals that could be relevant for appetite as well as hunger, especially when used in combination with sitagliptin.
It actually reminds me of kind of the very early days of Aardvark, even before you found the Prader-Willi indication, as you were kind of explaining the biology to me. You said coffee. Coffee is an appetite suppressant, kind of validating the bitter taste pathway to regulating appetite. Maybe you can elaborate a little bit on that observation and even the origin of the Aardvark story, because I think that's what it was based on.
Oh, great. Yeah, so thank you very much for bringing that up. I mean, it's very fascinating to me. There's epidemiologic data that show that people who drink coffee do lose weight, and it's proportional to their dose, right? And it's not related to whether it's caffeinated or decaffeinated. There's a small correlation, but most of the effect is driven by just drinking coffee, even if it's decaf. And I theorize that perhaps it was the bitterness that really was driving the effect, and that led to one thing to the other. And it made sense. And also, bitter receptors are evolutionarily conserved, and thought that maybe extra oral engagement of those receptors could replicate some of that reduction in eating behavior.
So you had some initial forays into general obesity with ARD-101. Kind of tell us what you did and what you found there, and then maybe set up kind of what you've started to see with ARD-201 and why there are differences.
OK, great. Yeah, so we did do a 20-patient placebo-controlled trial in general obesity. It wasn't a weight loss trial per se. It was just mostly for safety assessment in these patients. Those patients weren't told to diet or exercise, and we saw about a 1% differential with more weight loss with the folks on ARD-101 versus placebo, but what was really confirmatory to us in terms of proving out the MOA was the subjective hunger score reached that sig when we actually looked at the CoEQ scores for self-reported hunger. We also saw a statistical significance p-value of 0.0168 for ghrelin, which we think is a biomarker that might not be the best target, but it's a good biomarker for evidencing self-perception of hunger, and this was with a very mild fasting challenge of just patients told not to eat overnight.
Then the next morning, they were delaying their breakfast time by an hour. And then the ghrelin would rise in the placebo group but stay flat in the ARD-101 group, which was really helpful. And this was at the low dose of ARD-101 only without the benefit of concomitant use of sitagliptin, which we think would really potentiate the overall effect. And benchmarking that with what we've seen preclinically with the combination, we're very hopeful that a higher dose of ARD-101, which is what we're going forward with, in combination with sitagliptin, which represents our ARD-201 formulation, will elicit a very substantial effect.
Why would those two be synergistic TAS2R agonist and DPP-4 inhibitor?
OK, great, so DPP-4 inhibitors, just a quick review, is obviously a very successful class of diabetes medications. It slows down the degradation of endogenous gut peptide hormones, including GLP-1 and GIP, and these hormones normally have a very short in vivo half-life, like two to seven minutes. If you have a DPP-4 inhibitor, it allows those secreted hormones to last longer. We believe that our TAS2R drug induces enteroendocrine cells to secrete more of a panel of gut peptide hormones, including ones that are generally cleaved by DPP-4, so when we use it in combination with a DPP-4 inhibitor, we see really a magnified result, and also, to be clear, DPP-4 on its own is known to be weight neutral, but because we're inducing that extra secretion of endogenous hormone release, we think that that combination really is optimal.
Do you anticipate you're going to get to levels of these endogenous hormones that might start triggering nausea and vomiting like we see with the GLP-1 agonist drugs? Or what gives you confidence that you can kind of thread that therapeutic window without bumping into that side effect too meaningfully?
That's a great question. Yeah, we think that it's intrinsically limited in terms of what the max secretion is, because we're only inducing the natural release of those gut peptide hormones from the enteroendocrine cells. So we're not going to super physiologic levels. So we're just enhancing that natural effect. Whereas with the GLP-1 receptor agonist drugs, it's actually orders of magnitude more than physiologic normal levels. And I think that that is driving some of the nausea observations for that class.
Got it. So maybe coming to the diet-induced obesity mouse model data that you just recently presented, maybe give us a quick overview of what you studied and what you found there.
OK, great. We're very excited about this recent data that we shared in a diet-induced obesity model, which is a model that is the gold standard model used. This is the same model that Lilly, Novo and other pharmas use. Normally, you're reasonably skeptical about preclinical data and what the translational reliability is. This is the case for a lot of diabetes models. It always doesn't correlate one-to-one with the expectations in humans. Except in the diet-induced obesity model, it has been found to be very reliable. Even the relative potencies of GLP-1 drugs are very reflective of what the relative potencies are in humans. We use this gold standard model. We used our ARD-201 formulation. We observed a 19% weight loss in this model in 30 days, which was a phenomenal result, we believe.
To benchmark in the same conditions, tirzepatide has about a 20.5% weight loss in that similar setup in that study, so we have almost a near injectable GLP-1 effect with an oral drug combination that we would not expect to induce nausea due to our experience, and also, the second experiment we did was to have the established weight loss from a GLP-1 and then try different things. I mean, the normal expectation is once you're off of a GLP-1, there's this rapid weight regain. We think that's actually driven a lot by the increased hunger that is a response to the effect of a GLP-1, and in our study, we saw statistically the similar continued maintenance of weight loss equal to an optimal dose of tirzepatide.
The combination of our drug with a micro dose of tirzepatide, 1/10 of the optimal dose, elicited a continued effect that really superseded the peak effect of GLP-1s.
The results were quite remarkable. We actually covered it in a webinar that we have on record. If you want to reach out to us to listen to that, it's a great overview. Around that same time, though, you kind of amended the phase two program in obesity, so I guess first question: was it the data that led you to amend the program? If so, how, and then maybe you can talk about what the changes made were.
OK, thank you. So we were really excited about the data, especially the protection from weight regain, which is actually a very key focus of attention right now, because undoubtedly GLP-1s are effective. But then maintaining that weight, what options are there to continue that weight loss once you want to withdraw from GLP-1? And because the data were so directionally supportive, we wanted to change our original plans. Our original plan was to have an ambitious study where we'd be looking at a variety of different readouts. The initial part of the study would look at placebo-adjusted weight loss and other types of things. And then we'd study the protection from weight regain in the back half of that study. Now we're going to replace that with two different studies, the POWER and the STRENGTH study.
So now upfront, we'll be able to get data sooner with the protection from weight regain. And we'll be actually recruiting patients that have a deeper established weight loss on GLP-1s so that the expectation of rebound is probably likely to be a lot greater than the population we would have otherwise studied. And that's why we're focusing our attention on that.
What do the two phase two trials look like now in terms of dose, patient population, follow-up, and control arms?
OK, great. So the POWER study, which stands for prevention of weight regain, that will be patients likely to have a 15% or greater weight loss on the GLP-1 than they'll be off of it. The dose that we're moving forward with is likely to be 800 milligrams of our ARD-101 formulation with the normal dose of sitagliptin paired with it. And then we will basically see what the protection versus the normal expectation will be for the prevention of weight regain. That's the POWER trial, which is slated to start before the end of this year.
That may not have a placebo control. Is that hard to add a placebo control to basically force everyone off of it?
Yeah, no, this is still in discussion right now, what the best iteration of this trial design will be. And we'll be announcing that soon in terms of what the schema will be for that trial.
OK, and so that's the rebound.
Yes.
POWER, yep.
And then the STRENGTH trial, there's reasons for the acronym as well, right? But sitagliptin is what the S stands for. And then we're looking at weight loss a priori. So it will be the combination of, again, the same dose of 800 milligrams of the ARD-101 formulation plus sitagliptin in a prospective weight loss study. And then we'll compare it to semaglutide, a GLP-1, and also GLP-1 plus ARD-201 together and see what that weight loss is. We haven't yet declared what the GLP-1 dose is. There's thoughts of whether it should be the optimal dose of GLP-1 or maybe a micro dose of GLP-1. And that's what we're contemplating at present.
Could you consider both, or do you feel like you have to choose one or the other?
I think we're all balancing this against budget constraints and so forth, so this is what we're contemplating.
Expected timing for POWER and STRENGTH?
POWER will start before the end of this year. STRENGTH, first half of 2026.
Data readouts expected?
Data readout for POWER will be second half 2026, probably not too far off the timing for the HERO trial readout. Then we haven't declared what the STRENGTH readout expectation is, but we hope to give guidance of that by next quarter.
OK, got it. Are they likely to be comparable in size in terms of patient numbers?
You know, I think our focus is more on POWER than STRENGTH right now. But we haven't nominated the IND, which again is something that we'll be sharing soon.
Why is that? Why prioritizing POWER?
We think the strength of signal and then the read-through in terms of commercial ability. It's been also notable in conversations we had with prospective partners that this is a question that is very high in the mind of a lot of prospective.
So I would imagine if you can lower the dose of GLP-1, that's equally attractive because you're going to fall more into that therapeutic window where nausea and vomiting are as problematic. So how are you thinking about that consideration? Because it would seem to me to be equally compelling, if not more maybe.
Yeah, I want to pursue so many things. They're all important. So I agree with you. Yeah.
Why don't we go to the Prader-Willi program and some recent updates on timelines, and I think even some of the enrollment criteria. So maybe talk a little bit about the evolution of this program, especially as we think back to some of the early patient experiences where it seemed like there were some dramatic super responders. You were kind of working through dose escalation, maybe optimizing in terms of concomitant antipsychotics. And there's been kind of an evolution to the optimal profile here. So kind of frame that from the lens of some of the recent trial updates and what you're continuing to learn about how ARD-101 is best deployed for these patients.
OK, great. Yeah, so we did learn a lot from our phase II programs. We learned that dose timing is contributory to response. That if patients take their drug between meals, it likely is correlated with a better response, and we think that could relate to the drug's concentration in the lumen at the time it engages those enteroendocrine cells. We're also in our phase II; we saw lots of really dramatic responses to treatment, and beyond just the hyperphagia, too, importantly, I just want to emphasize that fact, because we think that gut-brain signaling is deficient in Prader-Willi. And gut-brain signaling is helpful, in theory, not just to abrogate hunger, but also reduction of anxiety, inflammation, and also normalization of that constipation that's very evident for Prader-Willi individuals too, so we actually saw signals of effect in those parameters, even though they weren't formally assessed.
But they are formal secondary endpoints now in our phase III. So that was another learning that was taken forward from their phase II.
Then the recent update in terms of timing and some of the enrollment criteria. Maybe just kind of frame for us what occurred there.
Oh, sure. So we directionally wanted to push the inclusion criteria of age lower than where we are at right now. So just for review, we had ages 17 and up for our phase two. 13 and up is what our current protocol is calling for. But we directionally want to push that lower to possibly seven years old or perhaps even lower and up. Reason being, there's a couple of motivations for this. Number one, we want to make sure the label of the drug, if we're successful, is going to be as expansive as possible in terms of age inclusion. And also, a lot of the historically in this competitive space, the stronger signals of effect generally have come from younger patients. There might be more neuroplasticity. We saw trends of improvement more in younger patients as well, even in our older set.
So, we're formally trying to incrementally get the age inclusion lower. And due to the accommodation of those younger patients, we want to make sure we don't conclude this study without having younger patients as far as the HERO study, which is why the timing of the top line is now pushed to Q3 to accommodate those younger patients.
I think you also loosened some of the restrictions around concomitant atypical antipsychotics. Maybe help explain why they were there in the first place and then why they were modified.
Sure. You know, we thought that this is another, so antipsychotics oftentimes are given with a normal dose and a PRN dose, and it can also introduce other variables that could have made our signal noisy, and we thought that for purity, it would be easier to exclude them, but then we don't actually think that that will mirror the effects. We think the struggle will be helpful in patients with antipsychotics as well. We also don't want to intentionally have a label restriction either, and as a consequence, our current amendment allows for patients on those concomitant psych medications.
OK, got it. One of the key differences between the phase I, II, and the phase III is the duration of follow-up. Maybe you can frame the phase I, II data in that light as we think about what it is you're looking to show in the phase III, which is, I think, 12 weeks now of duration.
Yes. So our study for phase II had a treatment period of only 28 days. Because at the time, we could only dose for 28 days, given the tox package we had now. We can now theoretically dose indefinitely because of the long-term tox data we have. So the readout now will be after 12 weeks of treatment compared to baseline. And that's our primary endpoint measuring HQ-CT. All patients, regardless of whether they're on placebo or active, will be invited to be on an open label extension for at least one year following that.
Got it. And the delta you're powered to show at 12 weeks, what is that? And how does that compare to what you showed in the phase I, II at four weeks?
OK, so right now, we think that it's a safe assumption our drug has at least a placebo-adjusted effect of six points. If we assume six points, we have 90% confidence. Assuming even standard deviation of six, we would only need 23 patients per arm to have a 90% confidence. If we think the true effect size is five over placebo, we'd only need 32 patients per arm, and we right now are positioned for 45 patients per arm.
One thing that does come up often in Prader-Willi clinical trials and interpretation of them is the placebo effect and the behavioral changes that occur if a patient thinks they're on something therapeutic. As you look at that phase I/II, what gives you the confidence that we're not just seeing that kind of placebo dynamic that might be lost in a phase III clinical program?
Of course, it's a very reasonable worry. We all have placebo. It definitely is a real effect. If you look historically in the histories of Prader-Willi trials, in a 28-day period, a placebo contribution is around two points. And if we presume that, then we still think that our drug has a net benefit of six points above that. In the context of Prader-Willi, a lot of the patients behaviorally have very pattern-oriented behavior. So it's understood that things that would convey benefit might take a much longer time to manifest in the HQ-CT, which is why a longer period of treatment is likely to elicit a larger effect size. This also goes to placebo, too. So we also accommodate for that. But there is an expectation of a deepening of score over three months versus one month. But we are trying to reconcile that, too.
We have ways to try to watch out the placebo contribution. For instance, there is a cutoff for inclusion for trials generally, and there is an implicit incentive to err on the side of over-reporting for study inclusion, and once in study, that might lead to a placebo effect if this baseline score changes, so for us, we're doing a second baseline score assessment after the inclusion is already guaranteed, so there's no more implicit pressure to report a higher HQ-CT score, and that second baseline is what we're using as the baseline for the study comparison at the end.
So it was interesting where we've done a number of webinars in the Prader-Willi space, including one with the patient advocacy group. And when we got to ARD-101, their faces lit up, right, as though there was just uniquely tremendous enthusiasm for the product in the community. To what do you attribute that to? What about the data in this context do you think has resonated so strongly? And is it the data? Is it the anecdotes that are going around?
Yeah. So there are more than a few patients that had really dramatic responses to treatment. I think the Prader-Willi community is very tight-knit. And a lot of the patient advocacy groups, there are folks in the leadership that probably are aware of the families that were on the trial. We try not to have direct channels of contact. But I think they secondhand appreciate the dramatic benefit of the drug. I mean, we've had multiple patient families reporting patients leaving food on the plate, which is something you would not expect in Prader-Willi. More than a couple of patients that completely forgot about their meal times, which is something that also is outside the realm of reasonable expectation. Multiple families unlocking the pantry and the refrigerator just because they're encouraged by the profound perceived benefit on the treatment period.
How do you think about limiting the effect that enthusiasm might have on the placebo side of the equation, right? Because the families and kids are going into the trial knowing there's a good chance that they could get an incredibly effective drug. I might imagine the placebo effect might start to creep up.
Yeah, it's always a possibility and always a worry. But we're confident that the underlying mechanism is evident, that it will be difficult to even what we observed. It would be very difficult to reconcile with just the placebo response, especially at the time. We had no expectation ourselves what the rapidity of onset of benefit might be. But it seemed pretty reproducible. I mean, even though our formal assessment period for HQ-CT in our study was 14 days after dosing, anecdotally, many families were reporting benefit within a day or two of first dose, which was a consistent subjective reporting to us, too. So that's another point of encouragement was how quickly it seemed to be active.
In terms of trial enrollment at a time that VYKAT XR is now available and seems to be getting quite strong early traction, has that represented a challenge in enrolling patients? And what is the rationale for them to perhaps enter a trial of ARD-101 and not go on VYKAT?
OK, great. Yeah, so we actually anticipated the approval of VYKAT. And we always had in planning the majority of our clinical sites actually outside the U.S. But even within the U.S., we still have high interest of enrollment. I think there are patient families and PIs that, on balance, like what the side effect profile seems to be like in ARD-101 and are very enthusiastic based on our early data as well, too. So I think we're still getting a lot of inbound interest. And even within the census of patients for the PIs that we have, there seemed to be abundant interest in the trial. So we're actually very confident about full enrollment.
Will you be giving us any interim enrollment updates?
No, we haven't planned. We do have an interim analysis. But I think we haven't declared any point. But we do want to assure that we have confidence that we'll deliver top line data by Q3 of 2026.
What's the interim analysis designed to look at?
Just for making sure we have the proper sizing for conditional powering and also having sufficient numbers of patients that still are HQ-CT score 13 by the second baseline assessment.
OK. And will you announce the results of that interim?
We haven't decided on that yet. But we're contemplating, yes.
And then maybe quickly, the HO indication, what are the timelines for hypothalamic obesity?
No, so actually, because of the interest and impetus to develop ARD-101, we're actually more focused on ARD-101 right now.
Makes sense. So then in the final seconds here, the next steps, I think we've kind of discussed some of the course of this conversation, but what we should be looking for over the next 12 to 18 months.
OK, great. So we will be having a cadence of peer-reviewed publications sharing some of the data sets, both preclinical and clinical, over the next months. I think we hope to have a presence at Obesity Week in November. And then we'll be giving updates and guidance on the POWER and STRENGTH trials by the next quarter as well.
Got it, and so we could potentially be reading out Prader-Willi phase III or the POWER study right around your birthday next month.
That'd be very interesting, yeah.
Great way to end. Thanks, Tien, for coming out, and thanks, everyone, for joining.
Thank you very much. I appreciate it.