My name is Jennifer Kim. I'm one of the biotech analysts at Cantor. I'm looking forward to this formal presentation with Belite Bio. We have Nathan Mata, the CSO. Nathan, you can get started when you're ready.
Absolutely. So first, I wanna thank you all for attending this afternoon. I also wanna thank the conference organizers for giving us the opportunity to share some of the goings on at our little company called Belite Bio. So Belite Bio is a biotechnology company based in San Diego, California. We're focused on advancing oral therapeutics to treat degenerative retinal diseases with unmet medical need. The diseases we're working on right now are Stargardt disease, which is a juvenile macular dystrophy that primarily affects children, and on the opposite end of the spectrum, geographic atrophy, which is the late stage of advanced dry AMD. Currently, we're ushering our phase III clinical asset, Tinlarebant, through development in both of these indications.
What's important to know about these indications is, although they are very separate and distinct, they share a common pathophysiology, and that is that both of them show the accumulation of toxic byproducts of vitamin A as a toxin that's implicated in disease progression. Because these toxins are derived from circulating vitamin A, our approach is to reduce the vitamin A going into the eye as a means of reducing these toxins and preserving the integrity of retinal tissue. So we'll go on with... Clicker's not working here. It was working earlier. Okay, so I normally don't show our management slide, but this is an important timing because, we're very, very, very thankful to have a new CMO, actually, our first CMO join the team, the management team. This is Dr. Hendrik Scholl. Dr. Hendrik Scholl is a superstar in Stargardt's disease.
He was the lead PI in the PROGSTAR studies, which were the largest both retrospective and prospective studies of the natural history of Stargardt's disease. So he basically is the world-renowned expert in Stargardt's disease. His expertise in AMD is also not to be mistaken. He's published numbers of papers, hundreds of papers, on the subject, and of course, he has an authority being involved in so many clinical trials in both Stargardt's disease and geographic atrophy. So we're very proud to have Dr. Hendrik Scholl join us as our CMO. I want to first go over to an overview of our clinical development plans, and then we can get into the MOA and disease biology, et cetera.
So currently, what we're doing right now, we finished a phase II, two-year open label study in 13 adolescent Stargardt subjects. And I should mention that our focus really is early intervention, so we're looking at early-stage Stargardt's disease, and we're looking at, if you wanna call it, early-stage geographic atrophy, where there's not an inflammation sort of leading the disease. So we're going after these very early-stage patients. The phase II study, I'll be talking to you about earlier, later on, so I'm showing you some of the safety and efficacy data that came from that study. We've also recently completed enrollment of a two-year phase III study in the same patient population, adolescent Stargardt kids. We enrolled 104 subjects.
You can see there the age classification from twelve to twenty years of age, and we're gonna have an interim analysis coming up at the end of this year. In addition, we are currently recruiting for a Phase III study we call DRAGON II. This study is very, very similar to the DRAGON study. That is the first three-year study. The only difference is the DRAGON II study will focus on geographies in Japan, U.K., and the United States. We're doing this study in collaboration with the Japanese regulatory authorities because we've recently received a pioneer drug designation from that regulatory body. This is basically the same as a clinical breakthrough therapy designation.
So there's a potential with very positive efficacy data in our first phase III, and as well as the phase II, III in DRAGON, that we could get a first approval in Japan before any other country, so that's a very exciting opportunity for us. And then finally, we have our study in geographic atrophy, which is also currently enrolling. We're about halfway through our enrollment goal of 429 subjects. Again, this is a two-year phase III study called PHOENIX. So what does the drug do? So our drug, Tinlarebant, is an oral, once-a-day, small molecule tablet, that essentially, it antagonizes the carrier protein for vitamin A. So it, so it basically antagonizes the protein that takes vitamin A from the liver to the eye, and that's a very important MOA. I'll get into it more later.
The point of this is that we're, again, trying to slow the accumulation of these compounds that are basically vitamin A byproducts, and by doing this, we believe we can have an effect on slowing lesion growth in both of these diseases. As I said before, one of our differentiators, in addition to being an early oral intervention, is early intervention, so we wanna get to lesions before they start becoming less manageable, and that's typically when inflammation kicks in. We wanna get there prior to any inflammatory response. We think this approach will be more effective at treating disease long term because you're going after early retinal pathology prior to there being an unmanageable disease state, and that's primarily, as I said before, driven by inflammation.
Of course, there's a market, huge market opportunity for Stargardt's because there's no approved treatments whatsoever, and for GA, there's no orally approved treatment. So most people who are in the space are familiar with the Apellis drug and the Iveric drug, which are intravitreal injections of complement inhibitors. These work very well in late-stage disease where there's inflammation. They would not be expected to work well in early-stage patients where there's very little inflammation. We have a great regulatory profile, Fast Track designation, Orphan Drug designation in pretty much all the key countries, and a very, very strong patent family. Let's get into the MOA now, the disease biology. So I talked to you about these byproducts of vitamin A.
I'm gonna show you now how they accumulate in both Stargardt disease and geographic atrophy, and it really begins with the formation of this complex here in the liver. So this abbreviation here is the chemical abbreviation for vitamin A. It stands for all-trans-retinol. In the liver, vitamin A, all-trans-retinol, binds to retinol binding protein four, and that creates a binding site for this larger protein called transthyretin. So this is the complex that gets dumped into the circulation from the liver. Two important things about this complex: One, it's a very large-sized complex, so it resists filtration in the kidney. This is how you maintain a high, steady state level of vitamin A in the blood. But secondly, and perhaps more importantly, is the fact that this complex is not needed for uptake of vitamin A in other extrahepatic target tissues other than the eye.
The reason is because the eye expresses a receptor for the retinol binding protein four, protein. Other tissues of the body don't express that receptor, so they don't have a reliance for uptake of vitamin A when it's bound to retinol binding protein four. That's very, very important because this approach that we're doing is a very site-directed approach, in that it will only affect vitamin A homeostasis in the eye, not in other extrahepatic target tissues. Once this complex docks onto the receptor, it goes through a series of enzymatic reactions, where it's eventually converted to rhodopsin. Light activation of the rhodopsin liberates this all-trans retinal species. This is an aldehyde form of vitamin A, which is highly reactive and actually very toxic, and if it lingers within the retina, it can actually destroy retinal membranes.
And so there's a mechanism, a biological mechanism, that is intended to pump this aldehyde out of the retina, and it's mediated by a protein called ABCA4. ABCA4 is essentially an enzymatic flippase, which grabs the retinal from inside the retina, flips it outward, available to another enzyme for further detoxification and re-entry back into the visual cycle. So this is the normal process of vitamin A in a healthy, unaffected eye. But in patients with Stargardt's disease, there are genetic mutations that affect the function of the ABCA4 protein. Consequently, the aldehyde cannot leave the retina as efficiently, and it lingers within the retina, where it condenses upon itself and other aldehyde molecules to form bisretinoids. These are the toxic byproducts of vitamin A that I told you about. They're called bisretinoids.
The primary bisretinoid that's been identified in human tissue is known as A2E, and in cell-based studies and in vitro studies, that molecule has been shown to kill retinal tissue through diverse mechanisms. So in Stargardt's disease, it's very clear from the disease biology that A2E and related bisretinoids are the primary reason, if not the sole reason, for retinal pathology and eventual blindness in Stargardt's disease. Now, these molecules also accumulate in GA, GA as well, but not because of a genetic mutation or a broken ABCA4 pump, but rather because of pathology beneath and above the retinal pigment epithelium, which houses all the enzymes of the visual cycle. This interferes with nutrient transfer across these compartments, and when that happens, these visual cycle enzymes can run amok, and bisretinoids will form locally right within the retinal pigment epithelium.
You can convince yourself that you're actually seeing them in the clinical context because these bisretinoids fluoresce. So under normal fundus autofluorescence photography, you can see not only the retinal lesions, but you can see the autofluorescence. So I wanna show you that right now. So we have two case studies. We have a patient with Stargardt disease on the top, and a patient with GA on the bottom, and you're looking at retinal images taken annually over about four and a half years. Let's focus on the Stargardt disease patient first. You see in this baseline image two central large lesions. This is dead retina, but peripheral to that, those lesions, surrounding on all sides, is this boundary of autofluorescence. That's where the bisretinoids are.
Now, as you move forward through time, from 12 months out to 57 months, what you see is that wherever you have autofluorescence, you will soon have retinal lesions. In fact, the autofluorescence expands in a centrifugal manner, and the lesions sort of follow it, sort of like a fire burning forest, right as it goes along the way. The point is that wherever you have those lesions, retinal autofluorescent lesions, you can expect to see retinal atrophy. You see the same thing in GA, but the presentation's a little bit different. The lesion here is much smaller. It's right dead center in the image there. But if you look very closely, it's actually more obvious at 12 months, you see these punctate areas of autofluorescence.
Now, as you go through time, what you see is those areas of autofluorescence soon become retinal lesions. At the final time point, you can see almost all of those little spots that we saw early on have now all become atrophic. Once again, wherever you see those autofluorescent lesions, you can bet you're gonna have atrophy. If you can stop those autofluorescent lesions from growing, you can have a really pronounced effect on slowing retinal disease. What does our drug do? Our drug, Tinlarebant, as I said, is a small molecule drug. It's a tablet about six millimeters in size, and it basically works in the liver to compete with vitamin A for binding to RBP4, and it doesn't allow that larger protein, transthyretin, to bind to it.
So consequently, what happens is that because of its small size, actually, the complex here actually gets eliminated through the kidney, and this results in a reduction of retinol binding protein four, and because the eye has a unique preference for uptake of vitamin A bound to RBP4, if you reduce RBP4, you're gonna reduce the amount of vitamin A going into the eye, and of course, that will reduce all the retinoids cascading downstream, including A2E, so this is the process whereby we intend to essentially slow the disease by going after these early incipient biomolecules that lead to early-stage retinal pathology, so let's move on. I wanna talk to you about the phase II clinical trial that I mentioned we recently completed, so this study enrolled 13 adolescent subjects, aged 12-18 years of age.
One important thing about this two-year study is that these kids came in with an early form of the disease, so they only have the autofluorescent lesions. They haven't yet converted to the atrophy. So in this study, what we're really interested in looking at is the conversion from one lesion type to the other, and then in those subjects that convert to these atrophic retinal lesions, we wanna measure the growth rate of those lesions because that's the endpoint for approval, is to slow the growth of the atrophic retinal lesions. So that's what we're gonna be looking at in this study. Here is the pharmacokinetic and pharmacodynamic profile of our drug. The blue line here shows you the level of Tinlarebant in blood. The red line shows you the reduction of retinol binding protein 4.
This target threshold that you see of 70% was established in a prior clinical study with a different drug, a different RBP4 antagonist, where we found in GA patients, if you got to a reduction of 70% or more, you actually had a slowing of lesion growth. So that was a very profound study. It was published in twenty thirteen under my name in the journal Retina, so you can look that up if you like. So that became our target threshold, if you will. And you can see with our five milligram dose here, which is what we're giving, kids. In fact, the five milligram dose is what we're using for all of our studies, Stargardt and GA. It produces an 80% reduction of retinol binding protein 4, so well below the target threshold.
It's sustained throughout the dosing treatment period until we withdraw the drug at month 24, and then you see a rapid clearance of Tinlarebant from blood and a rapid increase in retinol binding protein 4 as it goes back toward the baseline value. So this is a very nice reversibility of the pharmacodynamic effect, which is useful should you ever need to return a patient to his or own, his or her own, RBP4 baseline status. So I mentioned we wanna look at the conversion of one lesion type to another, and then the growth of the lesions if we see them. I wanna start by saying... Oh, I should mention that there were 13 subjects enrolled. We lost one subject to follow-up at month 12, so we're just dealing with 12 subjects for the efficacy analysis.
Among those 12 subjects, five of them, that's 42% of the cohort, never converted to an atrophic retinal lesion. Their disease was essentially static. The autofluorescent lesions never grew, and they never converted to atrophy. So that's a very promising sign. And one would ask, "Well, it's a genetic-based disease. Maybe these kids just have very mild or benign mutations." Not the case. All subjects, all 13 subjects, have severe biallelic pathogenic mutations. So the fact that we didn't see conversion in these five subjects is not because of mild or benign mutations. Trust me, they definitely have pathologic disease. Importantly, in those subjects that actually did have atrophic lesions, which is seven of them, they're shown here, the growth rate of those lesions was significantly lower, about 50%, than what is found in natural history.
These natural history data came from the PROGSTAR study, which, as I mentioned before, Hendrik Scholl was the PI of. He gave us access to these data. The database, over 400+ subjects, was largely adult subjects, but there was a population of 50 adolescent subjects that we could use for comparison, and that's what you're seeing here. The growth rate of lesions, atrophic lesions in those adolescent subjects from the Natural History study compared to our study. So we're seeing a profound, profound reduction in the lesion growth rate. Another really important finding that we had just recently was based upon lesions in the macula.
Now, I wanna start by saying, typically, what everyone does in GA or Stargardt disease clinical trials is they use fundus autofluorescence photography that's coupled with a software called RegionFinder, which requires a human reader to go in and score the size of the lesions. This, of course, is prone to subjective reader bias, and it takes time 'cause you have to have two readers, then an arbitrator if they disagree, so it's time-consuming. And of course, there's variance because, again, it's the human eye. So our reading center developed a new AI-based grading algorithm. It's computer-driven, mathematically driven, and it looks pixel by pixel across the density of the retina to look for changes in grayness. So the darker the gray, then, of course, you have a lesion, and it focuses just on the macular area.
So with this new methodology, we asked the reading center to look at our baseline images, and what they found was twelve eyes of eight subjects that actually had macular lesions at baseline. The RegionFinder software, that traditional routine method that everyone uses, never found that. Again, that could be due to the subjective reader bias. But because of that finding with this new grading algorithm, we asked the reading center to go back and reread all the images, and this is what they found. So twelve eyes of eight subjects, sorry, had the macular lesions. This is the growth of lesions over time from that subgroup, and you can see there's fairly linear growth up to about month 16, but then following month 16, there's no further growth of lesions into the macula.
This is significant because what it tells us is that if we kept dosing, we've essentially stopped the disease growth into the macula. And because your macula is your center for visual acuity, you should be preserving visual acuity if you can do that. The data on the other side are essentially the same data, except converted to percent involvement of macular lesion, where 100% would mean the entire six millimeter zone of the macula is occupied by lesion. And you can see here what we have in our kids is no more than 7% involvement. So we're very excited to see this because it's actually consistent with our visual acuity data, which show a stabilization of vision across time. The reason this is important is because these type of kids, these Stargardt kids, they actually have a very progressive disease.
They're typically losing about five to six letters per year. The mean annual loss during our treatment is only 2.5 letters, so the mean loss over 24 months was five letters. This is a really, really good finding that we've essentially stabilized visual acuity in these subjects, and I think it relates to the fact that we've actually slowed or stopped the lesion growth into the macula. Going forward, we wanna talk about the AE profile. This is an important profile because the first thing we've observed at the 24-month safety time point was no drug-related systemic effects whatsoever. Again, this speaks to the specificity of the MOA. As I mentioned, we would only be affecting vitamin metabolism in the eye, not in other extrahepatic target tissues, and the safety data from this profile show it.
What we see in terms of drug-related adverse events are two things that we want to see because they're telling us we're having the intended biological effect on the retina. So the first is a form of chromatopsia called xanthopsia. What is chromatopsia? So chromatopsia is a light-driven event. So, our subjects are typically reporting when they wake up from sleep and see a bright light like that, that activates cone photoreceptors in your eye. Cone photoreceptors confer bright light and color vision, but if you don't have enough vitamin A to feed them, you'll get this startling effect of color, where the cone photoreceptors are sort of electrically misfiring in your retina because they don't have enough vitamin A. They'll do that for seconds to minutes until they fill up with vitamin A, and then everything's fine.
So it's light-triggered, and it can be accommodated for by simply moderating your transition to bright light, and in this case, the kids are all reporting yellow. That's what xanthopsia is. Delayed dark adaptation is the other ocular AE that it's manifested. It's sort of the opposite effect. So this is mediated by a photoreceptor cell type called a rod photoreceptor, which of course confers dim light vision. So when these kids transition from a bright environment like this to a darkened environment, let's say going from a sunny afternoon day into a movie theater, normal person takes, you know, maybe five, six minutes to find your seat. Our treated subjects here will probably take three times longer or maybe four times longer because vitamin A is not filling up those rod photoreceptor, so they'll need time to fill up. Once they fill up, it's fine.
The important thing about delayed dark adaptation, it's part of the disease process in both Stargardt disease and geographic atrophy. So these subjects are actually have learned to accommodate it because they've been dealing it with as part of their disease process throughout life, and again, the chromatopsia, we think, can be very well tolerated. There was no dropouts due to AE in this study, and I can tell you that in our ongoing phase III study, which is more than a year into duration out of the two years, there's only been. There's been less than 4% dropouts due to these ocular AEs. So we're very confident that this drug is very well tolerated in this patient population. A little bit about the other phase III studies, the DRAGON I and the DRAGON II trials.
I just wanna make the point of the similarity of these two studies, and we do that intentionally. In fact, all of the clinical trial designs that we conduct, Stargardt disease or GE, are two-year studies with a one-year interim analysis. They all use the same dose, they all have the same endpoint, same efficacy and safety measures. They only vary in a couple of different aspects. The first, of course, is number of subjects. So in our phase DRAGON, our first phase III study, DRAGON I, we have 104 subjects, and of course, all of them have to start with atrophic lesions.
So in contrast to our phase II study, where we had subjects coming in with only autofluorescent lesions, because the endpoint is slowing the growth of the atrophic lesions, we have to start with some measure of atrophy at baseline, so they all have to have that. So there's a difference in the numbers, that's all. There's a difference in the geography, as I explained before, so we're focused on Japan, U.S., and U.K., whereas this is a more global study. There's a different randomization because, again, we have more subjects. We have the luxury of sort of biasing on the Tinlarebant treatment arm, versus when we have 60 subjects, we have a one-to-one randomization. But other than that, all the other parameters of the DRAGON II study are identical to the DRAGON I study.
Moving forward into the PHOENIX trial, which is our phase III study in geographic atrophy, I wanna start first by saying we were concerned that because of the older age and higher BMI of GA patients, because they would have a higher retinol binding protein 4 level, we may have to use a dose higher than five milligrams to achieve the same pharmacodynamic effect that I showed you earlier, the 80% reduction in retinol binding protein 4. We ran a PK/PD study in older healthy adults to match the higher age range and higher BMI of these GA subjects and found the five-milligram dose performs exactly the same way in these older subjects that as it does in the younger subjects.
We get about an 80% reduction with a 5-milligram dose during the dosing period, and then once we withdraw the dose, we see this very nice reversibility of the pharmacodynamic effect. Again, we can use the same dose in GA, and of course, as I said before, it's the same endpoint. Let's move now to the trial design. Again, very identical to what we see in Stargardt disease. The only real differences are in GA, we're enrolling 430 subjects to reflect the higher prevalence of this disease in the population. Other than that, the study is essentially identical.
So because of the similarity in the pathophysiology between Stargardt disease and geographic atrophy, because of the similarity in the trial designs, the dosing, the endpoint, we think that what we see in our Stargardt studies, which, of course, are proceeding ahead of GA, will be predictive of what we can see in the GA studies, and right now, everything is trending very well, if you look at our phase II data. So with that, I think we come to an end. I finished with about seven and a half minutes to spare, so I can take questions if you like. If it went too fast, I'm happy to go back and re-explain anything, but thank you for your time and attention.
I'm sorry, say that again? No, no, no. So in the open label study, which is not called DRAGON, it's just a phase II study, that was the only study where they had autofluorescent lesions, and we watched the conversion to atrophic lesions. In the phase III studies, for both DRAGON I and DRAGON II, all of them have to have atrophic lesions.
So-
Which statistics are you referring to?
... All the time.
Slowing the growth of the atrophic lesions. Well, actually, we did, because as I said before, we're looking at two parameters in this study. The first is the conversion of the autofluorescence to the atrophic lesion. Then, in those subjects that grow atrophic lesions, we wanna measure the growth rate. That's what you're looking at here. You're looking at the growth rate of atrophic lesions in our treatment group versus PROGSTAR. So this is the endpoint for registration.
You had patients in an earlier study that you have a growth rate on drug?
Yes. So seven of the twelve in this efficacy analysis showed those incident atrophic lesions, and the growth rate of those lesions was 50% lower if you compare it to natural-
In the natural history?
Correct.
The phase III is randomized?
All Phase III
No, but sorry, you have a control group. It's not compared to natural history, or is it?
Oh, absolutely, because it's placebo-controlled. Yeah, so this is an open label study. Other questions? Jennifer.
Thanks for the presentation. If I could ask one question, just because it came out recently. There was a potential oral competitor-
Mm.
in GA and Stargardt. They released some data from their trial in GA.
Yeah.
I'm wondering, what were your takeaways from that?
God, I'm glad you asked that question because that's a very important study. This is critical. What Ms. Kim is referring to is a study conducted by a company called Alkeus. Alkeus also has an oral once a day therapeutic, very different than ours. They're still going after the bisretinoids, these toxic byproducts of vitamin A, but they're doing it differently. They're introducing into subjects an overdose of deuterated vitamin A. So basically, this is a synthetic form of vitamin A. Why deuterated vitamin A? Because that vitamin A doesn't dimerize to form the bisretinoids as readily as native vitamin A. So basically, what they're doing is they're sort of slowing the conversion of the aldehyde into these bisretinoids. Okay, that's the approach, all right? There's a problem with that approach, and there's two problems.
I mean, but they all relate to you're overdosing patients with vitamin A. So the data, Ms. Kim spoke about was that they released, their findings from their geographic atrophy study, where they showed not a statistically significant effect, but a trend towards that. I think the p-value was point oh seven, and the difference between placebo and treatment was point two five millimeters square. So there's something there. It says this MOA does work, but they missed the mark because what they're not addressing is this. So actually, this aldehyde, as I mentioned before, is toxic. In the case of the Alkeus treatment, this aldehyde will have a deuterium on it, which means it won't dimerize as well, but it's still a toxic retinal.
In fact, because they're giving so much of this synthetic vitamin A, there's even more of this toxin, this all-trans-retinal, which is the precursor for the bisretinoids. There's more of it in the treated patients than it would be in placebo because, again, they're overdosing with vitamin A. Because the approach here is you cannot direct this deuterated vitamin A to just the eye. It's gonna go everywhere. So you would basically supplant or replace your native vitamin A with this synthetic vitamin A at a very high dose because it has to swap out the native vitamin A. So it's roughly four to five times greater than the dietary recommendation by, for instance, the World Health Organization, so you're overdosing with vitamin A. So I think the reason they didn't have a more robust effect is because they're fighting the toxicity associated with retinal.
So yeah, they're slowing the formation of A2E because it doesn't dimerize, but they're not addressing this. Our therapy is addressing this because we're reducing all the retinoids cascading through the visual cycle by hitting it right here at this choke point. So we're limiting the amount of vitamin A going in, not increasing it. And so, yeah, we'll have the ocular AEs, but as you've seen before, they're manageable. They won't have that, so they won't have the delayed dark adaptation or the chromatopsia, but again, they will have this issue with respect to dealing with this toxicity of retinal. I hope that answered the question.
Yeah.
Any additional questions? We have a few minutes. No?
Okay. Thanks, Nathan.