Hello, and welcome to the Deutsche Bank Depositary Receipts and Virtual Investor Conference, DBVIC. My name is Afra Aziz from the Deutsche Bank team. I'm pleased to announce that our next presentation will be from Belite Bio, Inc. from Cayman with U.S. headquarters. Before I introduce our speaker, a few points to note. Please submit your questions in the questions box to the right of the slides. Also, all of today's presentations will be recorded and can be accessed via the Deutsche Bank website, adr.db.com. At this point, I'm very pleased to welcome Dr. Nathan Mata, CSO of Belite Bio, which trades on Nasdaq under the symbol BLTE. Over to you, Nathan. Thank you very much. Before I begin, I would like to thank the organizers for the opportunity to present on this forum and share a little bit of information about our company, Belite Bio. Belite Bio is a small biotech company based in San Diego, California. We also have offices in Taipei, Sydney, and Hong Kong. At Belite Bio, we are focused on the development and advancement of oral therapeutics to address unmet needs in medicine. We are currently, right now, adamantly focused on developing our lead asset, Tinlarebant, in two different but somewhat related ophthalmic indications. The first is a genetically inherited disease that primarily affects children. This disease is known as Stargardt's disease. Children who are affected with Stargardt's disease, typically diagnosed from six to eight years of age, and by twenty, they are legally blind. The other disease is sort of the opposite end of the age spectrum. It is the advanced form of dry, age-related macular degeneration, commonly known as geographic atrophy. In both of these diseases, the accumulation of toxic byproducts of vitamin A is implicated in disease progression. Because these byproducts are derived from circulating vitamin A, our approach is to reduce the amount of vitamin A going into the eye as a means of reducing the accumulation of these toxins and slowing the progression of retinal disease. Here we go. Sorry, everyone. So I normally don't show a management slide. I wanted to show this one in particular because we've recently added... We're very excited about adding a CMO to our team. This is Dr. Hendrik Scholl. He was a world-renowned ophthalmologist and retinal specialist, particularly in the area of Stargardt's disease, where he was the lead PI in the largest natural history study of Stargardt's disease conducted to date. That study is known as PROGSTAR, and we'll talk more about that in a moment. But in addition, he's been involved in over 10 clinical studies of both Stargardt's disease and geographic atrophy, and he's published hundreds of papers in these two fields. So we're very, very happy to enrich our already strong management team with the medical and clinical expertise of Dr. Hendrik Scholl. So what are we doing at Belite Bio? Let's give you a little overview of the pipeline. First, in Stargardt disease, we've recently completed earlier this year an open-label phase 2 study, a two-year study in adolescent Stargardt children with our drug, Tinlarebant, and we saw very, very safety, very promising safety and efficacy outcomes. I'll show you some of that data today. We've also recently completed enrollment of a two-year phase III program, also in the same target population, adolescents with Stargardt disease. This study completed enrollment with 104 subjects, aged 12 to 20 years of age. We're expecting the interim analysis from that study later this year, perhaps around December, latest part, probably into January of 2025, and then one year after that, we'll have the top-line two-year data from that phase III study in adolescent Stargardt disease. In addition, we've recently started a another study, a phase 2/3 study that we call DRAGON II. This study is currently recruiting up to sixty subjects. It's identical to the previous study that we call DRAGON, with the exception that this is focusing on geographies in Japan, U.K., and the U.S., and it's really intended to take advantage of a recent designation we got from the Japanese Regulatory Authority called Pioneer Drug Designation, which is essentially the same as a clinical breakthrough therapy designation here in the U.S., and it gives us an opportunity to potentially get registration with just this one phase 2/3 study, pending positive safety and efficacy outcomes. And finally, in geographic atrophy, we're also enrolling a phase III, two-year study that we call PHOENIX. We're enrolling up to four hundred and twenty-nine patients into that study. We're a little more than halfway into that enrollment goal. We expect to close that enrollment probably about Q1 of twenty twenty-five, and then Q1 of twenty twenty-six will be the top line, sorry, the interim data from that one year, and then the following year will be the top-line phase III data for the PHOENIX study in GA. So what is Tinlarebant? So Tinlarebant is a novel oral once a day tablet to target a protein called retinol binding protein 4, that delivers vitamin A from the liver to the eye. This is the only protein that does that, in the body, so we're specifically targeting that protein, as I said before, as a means to slow the accumulation of these toxic vitamin A-derived byproducts, that are called bisretinoids, and I'll show you those in just a moment. The point about these, these compounds, these bisretinoids, is they appear very early in the disease process, so we're really focusing on emerging retinal pathology that's not mediated by inflammation. Once you get to a late-stage disease where there's inflammatory components being elicited, it's much harder to manage these diseases. So again, what, in addition to oral intervention as one differentiator across the therapeutic landscape, we're also going in very early in the disease process to go after these lesions before they become inflamed. In terms of the market opportunity, there's no FDA-approved treatments for Stargardt's, and there's no FDA-approved oral treatments for geographic atrophy. Current therapies, of course, are intravitreal injections of essentially anti-inflammatory compounds that interfere with the cascade of inflammation.... We have Fast Track Designation, Rare Pediatric Disease Designation in the U.S., and of course, Orphan Drug Designation in various territories, including now in Japan. Our patent position is very strong, 14 active patent families. Most of those composition of matter patents, with the last one not expected to expire until 2040, and that's without patent term extension. So let's get into a little bit about the disease biology, just to show you what we're doing here. I mentioned that in both Stargardt's disease and GA, the accumulation of toxic byproducts of vitamin A is implicated in disease progression. I want to show you how those compounds accumulate. It begins here in the visual cycle. This is a schematic of the visual cycle. On the bottom right-hand corner, you see this complex, this ternary complex of three entities. RBP4 is retinol binding protein four, a tROL is the chemical abbreviation for vitamin A. That stands for all-trans-retinol. In the liver, these two bind together, and then this larger protein, transthyretin, binds to it, and this is what gets dumped into the circulation, this very large molecular size complex that's essentially encapsulating vitamin A in its core. And because it's so large, it essentially resists filtration in the kidney. So this is a nice way to maintain a high, steady state level of vitamin A in the blood. But what's really important to note is that only the eye has a preference for uptake of vitamin A bound to RBP4 because of the presence of an RBP4 receptor on its basal surface. Other tissues of the body do not express that receptor, and so they don't rely on delivery of vitamin A bound to retinol binding protein four. They can uptake vitamin A from other sources and from other carriers. So in the eye, when vitamin A, when this complex docks onto the receptor, vitamin A is admitted into the back of the eye, and it goes through a series of enzymatic reactions, where it's eventually converted to rhodopsin in the retina. Light activation of rhodopsin liberates the aldehyde species of vitamin A, which is highly toxic and very reactive. If it stays in the retina, it will actually damage cellular membranes. So the way it gets out is through an active pumping process mediated by a protein called ABCA4. ABCA4 grabs the aldehyde from inside of the retina and flips it outward, availing it to another enzyme for further detoxification and re-entry back into the visual cycle. This is the normal processing 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 can actually condense upon itself, that is, bind with itself, forming these dimers of vitamin A that we call bisretinoids. These are the toxic byproducts that I told you about. These byproducts eventually find their way into the retinal pigment epithelium, where they sit and mature and become very stable cytotoxic entities. So in Stargardt's disease, we know the whole reason for retinal pathology and eventual blindness is because of the accumulation of A2E and related bisretinoids. And of course, as I said, because they're derived from circulating vitamin A, one way to reduce them is to limit the amount of vitamin A going into the visual cycle. In geographic atrophy, these compounds accumulate as well, but not because of a genetic mutation or of a broken ABCA4 pump, but rather because of pathology beneath and above that layer of tissue called the retinal pigment epithelium, where all the visual cycle enzymes reside, and when those enzymes become dysfunctional, you can form those bisretinoids locally, right within the cellular compartment of the retinal pigment epithelium, so in both of these diseases, they accumulate. In Stargardt's, they are the sole cause and probably the primary cause, and in GA, they're probably a component of a larger disease process, but the nice thing about these compounds, at least in terms of clinically, is because they're comprised of vitamin A, they actually fluoresce. And so ophthalmologists can actually visualize them when they're looking at the disease itself, and they use a specialized camera that allows you to look at the retina in the back of the eye. And if you look at the top here, what we have are two case studies. We have a patient on the top with Stargardt disease, and we have a patient on the bottom with geographic atrophy. You're looking at retinal images from these patients taken over four and a half years annually. Let's start with the patient with Stargardt disease in the upper left-hand image. You see these two central areas of blackness. Those are. That's area of dead retina. They're called atrophic lesions, but peripheral to those lesions on all sides is an intense zone of autofluorescence. That's where the bisretinoids are. Now, as you move from the baseline image out towards 57 months, what you see is the autofluorescence area continues to expand in a centrifugal manner, and the lesion, the black area, follows right behind it. Wherever you have autofluorescence, you will soon see atrophic lesions. It goes to show you that if you can get rid of that autofluorescence, i.e., the bisretinoids, you can slow the progression of retinal cell death. Now, let's look at the bottom patient, the GA patient. On the bottom left-hand corner, you see a central lesion. It's much smaller than in the Stargardt patient, but look peripheral to that lesion. You see these little punctate areas of light. They become a little more obvious at 12 months. You can see them surrounding that lesion, sort of like satellites on all sides. And now, if you contrast that 12-month image with the 55-month image, what you can see is those areas that were previously just autofluorescent at 12 months have slowly become atrophic lesions, and they all have little zones of autofluorescence around their perimeter. So again, here's a clear demonstration from a clinical perspective that those molecules are, in fact, present in both tissues, both Stargardt disease and geographic atrophy. So how do we get rid of them? Our drug, Tinlarebant, 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 gets liberated into the circulation is a very small complex of our drug bound to retinol binding protein four, and because it's so small, it gets readily filtered through the kidney. The net effect will be a reduction in the wild type complex, this ternary complex of RBP4, retinol, and TTR, that would normally deliver vitamin A. So when that concentration of that complex goes down, of course, the amount of vitamin A in the eye goes down because, again, the eye requires delivery of vitamin A bound to RBP4, and we're slowly reducing RBP4 as a means of reducing retinal delivery to the eye. Once we reduce retinal delivery into the eye, then all the retinoids cascading downstream, including those bisretinoids, will also be reduced. So that's the mechanism of action, and that's the approach whereby we intend to slow lesion growth and hopefully eventually slow loss of vision in these patients. Moving forward, let's go into the Stargardt phase II clinical trial. I told you that we recently completed. This was an open-label study enrolling 13 adolescent subjects aged 12 to 18 years of age. These subjects came into a study with the very early form of the disease, so they only had the autofluorescent lesions. They haven't yet converted to the atrophic lesions. So the atrophic lesions are referred to as QDAF. That stands for Questionably Decreased Autofluorescence. Just know that they're the autofluorescent lesions. They convert over time to atrophic lesions, those black blotches I showed you earlier in the clinical images, to lesions that are called DDAF, or Definitely Decreased Autofluorescence. Slowing the growth of the DDAF lesions, the atrophic lesions, is the endpoint for registration, pretty much in the FDA and EMA and elsewhere. So we want to slow the growth of the atrophy. So in these patients, because they're coming in early, we want to watch the transition from one lesion type to another to see how many subjects actually convert and over what period of time. And then those subjects that actually convert to these atrophic, retinal lesions, these DDAF lesions, we want to measure the growth rate of those lesions. And you can see there the various criteria. As I mentioned, it's an open-label study. Australia and Taiwan were the sites that were enrolled, two years, looking primarily at safety and tolerability, but also looking at these efficacy measures, as I mentioned, and you can see there the key inclusion criteria. I should mention that at month 12, we lost one of these subjects to follow-up, so the safety and efficacy assessments you'll see were based upon those 12 patients. First, I want to show you the pharmacokinetic and pharmacodynamic profile of the drug. We're using five mg of Tinlarebant daily. Again, this is an oral dose, and that drug reduces retinol binding protein four by a mean 80% reduction. First, the blue line you see in the plot here, or blue-gray, is the level of Tinlarebant in blood during daily dosing. You can see that certainly by month three, we have a steady state concentration of Tinlarebant in blood, and you can see that corresponds to a reduction in retinol binding protein four. That's the red line shown as a% reduction from baseline, and you can see a very steady state reduction, 80%, throughout the dosing period. The reason we indicate here a target threshold of greater than or equal to 70% is because that was the value that was determined to be efficacious in a GA study, that is a study with Geographic Atrophy. It was a proof of concept phase II study done about 13 years ago with a different drug. It was not an RBP4 antagonist per se, but it had that side effect, and that level of reduction was shown to reduce lesion growth in those patients. So we want to get below that, and here we are at 80%. And then finally, once we withdraw the drug over a 1-month period, you can see there's a rebounding of the Retinol Binding Protein 4 right back towards the baseline value as the Tinlarebant leaves the blood, circulation. Let's get into the data now. So remarkably, what we found, I mentioned that we wanted to watch the conversion of one lesion type to another. In five of 12 subjects, that's 42% of the cohort, there was actually no conversion to the atrophic lesions over two years, and the reason that's remarkable is because that's just not seen in adolescent Stargardt Disease. This, this disease is a very rapidly progressive, and in all our kids, I didn't mention it, but we did genotyping. Every single one of them has severe biallelic mutations that would predict pathogenicity, so we know that the reason for not having conversion in those five subjects was not because of a mild or benign mutation. We're likely seeing a drug effect. In those seven subjects of the twelve that did spawn these atrophic lesions, the growth of those lesions was significantly lower than the growth rate observed in natural history. The red line shows you the growth of lesions in those seven subjects in our Tinlarebant study, and the blue line above shows you the growth rate of similar subjects, adolescent Stargardt subjects, taken from the PROGSTAR study. This data was given to us by Dr. Hendrik Scholl. PROGSTAR focused primarily on adult subjects, over four hundred plus subjects. In that larger group, there was a subgroup of about fifty subjects that were adolescent subjects that had similar baseline disease criteria as we had in our study. We could compare apples to apples to see whether or not the growth rates were different. We're seeing here a profound reduction in the lesion growth rate of our subjects compared to natural history. Another interesting finding came from a recently developed imaging grading algorithm from our reading center. The current reading algorithm requires readers essentially to demarcate the boundaries of lesions so that the machine can calculate the area within that area that was demarcated by the reader. Two readers typically have to read an image and come into an agreement with it within a certain variance, or it goes to a third reader. As you can imagine, there's subjective reader bias there because the human eye is different in everybody. Everyone sees images a little differently. Of course, if you have to get an arbitrator every time to tiebreak, then, of course, there's time involved there. Our reading center wanted to remove the subjective reader bias from this traditional method of reading images, and they developed an AI-based essentially a mathematical algorithm, whereby the computer simply looks at the retinal images and scores on a pixel-by-pixel level, the grayness or darkness, if you will, of each pixel. That is, the grayer or darker a pixel is, the more pathologic it is, so it would score based upon that criteria. Again, it's looking pixel by pixel, so much more finely than the human eye can look. Finally, it looks just at the six- mm zone of the macula. Just that area that confers visual acuity. It's focused just primarily on the area that would be compromised in patients that are losing their vision. When we had this methodology against... Looking at our retinal images, our reading center came back and told us there were actually 12 eyes of eight subjects that had macular lesions at baseline that the traditional readings methodology didn't pick up. It just didn't see those lesions. So we were curious to see what the rest of the images would look like using this new grading methodology, and we had the reading center go back and reread all the images, and what they found is what you're looking at here in these two plots. The plot on the left shows you the change in the area of macular involvement over time. The line and symbol figures show you the actual data, and the dotted line shows you a third order polynomial constructed through the data, so you can look at the trend. What you can see here is there's fairly linear growth up to about month 16, and then from month 16 out to month 24, there is no further growth. We've essentially halted growth into the macula, and it sort of makes sense that it would kick in after one year, because early on in the disease, there's some cells that you're just not gonna be able to save. They're probably just too full with the autofluorescent lipofuscin, they're just tipped over, and you just can't get to them. But you can get to those other cells, and it probably takes a year or so to sort of wash those compromised cells out and start sort of replenishing with these healthier cells, or at least protecting them from the accumulation of these toxins. On the right-hand side, you're seeing the same. What we're looking here is looking at the data as a mean% change of macula involvement, where 100% would mean the entire 6-mm zone is filled up with lesion. You can see here that we're not getting any more, they're about 7% involvement, and it doesn't go beyond that, for the last year study. So this is a very promising result, and in fact, it feeds very nicely into the finding that we have very stabilized visual acuity. Typical visual acuity loss in these subjects is somewhere between 5 to 6 letters per year. The mean annual loss that we're seeing is 2.5 letters. So, so we think we're seeing something, you know, promising here. Certainly, it's not improving, but, but it is, it is showing a stabilization of vision. That's how we like to present it. By the way, the reason you're seeing two lines is we're looking at both eyes. One eye is always designated a study eye from which you report your statistical data, and the other eye becomes a fellow eye. We're just showing you both data from both eyes in the interest of being complete. Suffice it to say, we have stabilized visual acuity, and it is consistent with that stabilized lesion involvement into the macular area. Finally, I will talk about the safety from the study. This is, you know, really remarkable, and it's a testament to the specificity of this drug. There were no drug-related systemic AEs whatsoever. All vital signs are clear, physical exams, cardiac health, everything's fine. We have two ocular AEs that we want to see because they're telling us we're having the intended biological effect on the retina. The first is of a color effect, a form of chromatopsia called xanthopsia. This is triggered by onset to bright light. So typically, when kids are waking up, this is what they report in their diaries. When they're waking up from sleep, they get this startling effect when they're seeing light, whether it's sunlight or artificial light, and that really is due to cone photoreceptors in the retina not getting vitamin A quickly enough. Under a normal context, where you're not being treated with our drug, the vitamin A floods in very rapidly. Here, it'll sort of trickle in, so there'll be this sort of startling of color in the visual field for seconds to maybe a few minutes. It's, of course, transient, it's mild, as you can see, it's reported, and of course, it's fully reversible. We haven't lost one subject to chromatopsia. Delayed dark adaptation is the other manifestation. This is the opposite effect, so also light triggered, but in the opposite way. So when subjects transition from a bright light to a dark environment, it takes them a little longer to adapt to that darkness, that dim setting. This is not night blindness, otherwise known as nyctalopia. This is a delay in the ability to accommodate to dim light, and it's typically two to three times longer than normal. In cases where it's out to about twenty minutes, which would be roughly almost four times longer than normal, it's called night vision impairment. We had that in one subject, but again, no one's left study because of either delayed dark adaptation or chromatopsia. Increasing EROS score on the FM-100 is an increasing exacerbation of the chromatopsia, and intermittent headaches we think can occur when subjects struggle to use their visual acuity while they're experiencing these ocular AEs. But overall, really good safety profile. We're happy that we didn't lose one subject at all to any adverse event. Now, going into the DRAGON one and DRAGON two studies, I just want to show here that these studies are actually very, very similar. They're almost identical. There's only a few differences. The first, in the number of subjects, so the phase III DRAGON one is 104 subjects, versus the phase II DRAGON two is 60 subjects. DRAGON one is more global. As I mentioned earlier, DRAGON two is more focused on Japan, U.S., and U.K. The randomizations are a little different. In the DRAGON study, the first one, the large one, we can take advantage of the 2:1 randomization because of the larger sample size. But with 60 subjects in DRAGON II, we're keeping a one-to-one randomization. Otherwise, the studies are identical, and the reason this is important is because it suggests, and given the similarities in the population, that what we see in one study could likely be predictive of what we see in another, because it's the same drug, same endpoint, same study population, et cetera. Moving on now to our PHOENIX study. That's our phase III study in geographic atrophy. It's nice to point out that in this study, we will be able to use the same 5-mg dose that we're using in all of our Stargardt studies. And the reason is because that 5-mg dose produces the same pharmacodynamic reduction of retinol binding protein 4. That is a mean 80% reduction during the dosing period that we saw in the adolescent subjects. The data that you're looking at here is not from GA subjects, but rather it's from older, healthy adults meant to match the higher age range and higher BMI of GA subjects, so this was a PK/PD study. We also did a 10-mg dose just to see what would be required, and it turns out the five-mg dose produces the same 80% reduction and the same reversibility following the end of the dosing period here at 14 days. And again, just like our Stargardt studies, the clinical trial design in Geographic Atrophy is essentially identical. There's only two differences: the larger number of subjects to match the higher prevalence of GA in the population, and of course, the nature of the study itself, looking at GA subjects rather than looking at Stargardts. But again, because of the similarity in the pathophysiology, the similarity in the trial designs, and of course, the identical dose and the identical endpoint, we believe what we see in Stargardts will be predictive of what we can see in GA, and right now, everything is trending very, very positively. So with that, I thank you for your time and attention, and hope you look at us. Look us up at belitebio.com. Thank you.