Welcome, everyone, to this session of the Morgan Stanley Global Healthcare Conference. We are very excited to have Belite with us this morning with Hao-Yuan and Hendrik representing the company. Thanks for being here, guys. I am Judah Frommer, one of the SMID-cap biotech analysts here at Morgan Stanley. For important research disclosures, please visit www.morganstanley.com/researchdisclosures. With that, it is an exciting time for Belite with a potential first-in-disease approval coming up. Before we dive in, maybe can you give the audience a quick intro to the company for those who may be less familiar with the story and how the company was founded?
Sure. Belite started as a spin-off of the parent company, Lin BioScience, which is also listed in Taiwan. We currently have our leading pipeline called tinlarebant in both Stargardt disease and geographic atrophy. In Stargardt, we have completed our phase III study and submitted for the FDA, which they already accept the NDA with the priority review, with the PDUFA date for next February 12th. We are also running in parallel a second phase III Stargardt trial called DRAGON II. That is mainly for the reservation for Japan. For GA, we have a fully enrolled global phase III study called PHOENIX is ongoing.
Okay, great. Like you mentioned, your drug tinlarebant could be the first approved therapy for Stargardt disease. Can you tell us a little bit about Stargardt disease? How many patients does it affect? What does the natural history look like here?
Yeah, happy to. Stargardt disease was described by Karl Stargardt about a century ago. It is amongst all the genes that, if mutated, would lead to retinal disease by far the most prevalent. The disease has, in populations of European descent, a prevalence of about one in 6,500. The prevalence is somewhat lower in East Asia and actually higher in populations of African descent. The estimate given the penetrance and mutation rate in the gene that 47,000- 59,000 patients in the U.S. are affected by the disease. We currently, that is the conclusion from our research data, that 20,000 patients are currently clinically diagnosed. Amongst those 20,000 clinically diagnosed patients, more than half carry a genetic confirmation, meaning that at least one mutation was detected in the ABCA4 gene.
Okay, great. This could be the first approved therapy for Stargardt disease, but maybe just spend a minute on standard of care. How are patients currently managed, what's frequency of visit and treatment?
Yeah. Stargardt disease is a very important cause of legal blindness. Amongst all the inherited diseases, it is by far the leading cause of blindness in adolescents and young adults. The disease typically starts in childhood and early adulthood and invariably gets worse and worse, leads to legal blindness in almost all cases. That is rare, but it can happen, can lead to no light perception. There's no light detection, that no light detection remains, meaning the disease kicks in early on in life and is severe, meaning that there's almost no patient that would not have any diagnosis. Because if you have such a severe disease and you cannot be corrected with glasses and your vision is down to 20/200 or a similar level, it would be an unlikely scenario that there was never any diagnosis made. The question is how precise is the diagnosis?
Because clearly, in order to initiate treatment, we need a precise diagnosis of Stargardt disease. It's not so difficult, I have to say. I speak as a clinician, and I still see patients one day a week in Switzerland. The disease can readily be diagnosed, especially nowadays with imaging that we have in the clinic, which is fast and reliable and very precise. To make the diagnosis is not so difficult. But we anticipate that when tinlarebant is on the market, that the vast majority of patients would need to be or will be followed by a retinal specialist, meaning they need to be referred to a retinal specialist in the future. Many patients are currently being seen by retinal specialists. Amongst all of those patients, again, the majority will be seen by so-called IRD, inherited retinal disease specialists.
We have a very good understanding in the U.S. where are the centers of excellence and where are those IRD specialists located.
Okay, great. Switching to tinlarebant itself, can you tell us a bit about the drug's mechanism? How does it benefit Stargardt patients, and what did you see in your phase III DRAGON study that kind of is proof point that the drug's working?
Yeah. Karl Stargardt in 1909 described a macular degeneration. By the way, it's very similar to age-related macular degeneration. It's just the patients are 60- 80 years younger, where the macular shows atrophy, and around the atrophy, it is surrounded by flecks. He called it flecks at the time, yellowish flecks. We know exactly what these flecks are made of, namely toxic bisretinoids that accumulate in the retina in Stargardt disease. When we use so-called autofluorescence imaging, which is a standard tool in retina clinics nowadays, where you shine in blue light and we detect the outcoming light from the retina, we see that these flecks light up, and those flecks indicate the cause of the disease.
The cause of the disease is this constant accumulation of toxic bisretinoids, which cannot be detoxified, because the protein that would be important to do so is dysfunctional because of the mutated underlying gene with the name ABCA4. Those ABCA4 mutations, therefore lead to dysfunction of ABCA4 and accumulation of toxic bisretinoids. The consequence of the disease of these toxic bisretinoids is then macular degeneration, and this again can be very precisely imaged with fundus autofluorescence imaging. Tinlarebant allows to address A2E and bisretinoid accumulation in photoreceptors and the underlying tissue called the retinal pigment epithelium by targeting the so-called RBP receptor that is unique in the eye. The eye has a unique requirement to access vitamin A due to this receptor, no other organ.
Tinlarebant allows systemically to lower retinol bound to retinol-binding protein 4, therefore, less compound is available for the receptor called the RBP receptor, or specifically STRA6. Because of lowering influx of vitamin A into photoreceptors and no other organ in the human body, the underlying material, if you will, the substrate of forming the visual pigment in the photoreceptors, this process is called the visual cycle, is reduced, and that allows to reduce A2E accumulation and therefore to preserve vision.
Great. I guess just turning to what you saw in your phase III program in the first DRAGON study. What was, I guess, best evidence of efficacy of the drug here?
The best outcome measures in macular degeneration have been developed over the last two decades, in conjunction with the FDA. The FDA is very aware that the underlying reason to see is photoreceptors that are being preserved. If you lose photoreceptors, you lose vision, and therefore preservation of photoreceptors is clinically valuable and beneficial. Therefore, the FDA, one and a half decades ago accepted preservation of photoreceptors as imaged by autofluorescence imaging as an approval endpoint, and that led to the approval of SYFOVRE and IZERVAY for geographic atrophy. In Stargardt disease the situation is essentially identical, because we have a more technical, more precise term, if I may say so. Namely, we call these areas DDAF, definitely decreased autofluorescence, that were shown to progress at a predictable rate in the worldwide natural history study called ProgStar, and can be precisely measured from autofluorescence images.
This is an approval endpoint recognized by the FDA as something that reflects health or decline of photoreceptors and therefore eventually will correlate with visual function. This exact endpoint was used in the DRAGON trial. Over the course of two years against placebo, it was shown that the progression rate or the deterioration of those lesions was slowed down by more than a third, namely by 36% over the duration of the trial. Leading to a major difference between treatment and placebo in the resulting area of DDAF or essentially lesion size.
Okay. A question we get is on the separation from placebo on visual function, and you mentioned the ProgStar study. Was it a surprise that you didn't see that separation on visual function, or is that something that might be expected given the natural history here?
It's a very important question. At the end, what counts for patients is can vision be preserved? The challenge that we have and that started already in geographic atrophy is that the progression is still relatively slow and measurement of visual acuity is quite variable. That introduces really a problem when you think about that the intersession variability of measuring acuity is eight letters on a visual acuity chart. We call that ETDRS, Early Treatment of Diabetic Retinopathy Study chart, that has 100 letters. If you have a variability of eight letters between measurements, that is a significant variability that is very difficult to be managed in a clinical trial of a reasonable duration in a reasonable amount of subjects.
That is recognized by the FDA, that if you have something that can be measured much more precisely, namely lesion size on autofluorescence images, that that allows to detect something that is clinically beneficial much earlier than the effect in visual acuity. When we get to Stargardt specifically, I mentioned the worldwide natural history study, ProgStar. ProgStar has shown that the progression rate of visual acuity loss over time was 0.55 letters per year, half a letter per year. When you apply the clinical significance level that the FDA defined, I think a couple of decades ago, namely 15 letters on a visual acuity chart, and you hypothesize that you have a treatment that would actually allow to arrest progression completely, you have to run the trial for 30 years. The FDA understands that this is unreasonable, and therefore, accepted DDAF progression as an approvable endpoint.
Long story short, no, that was not surprising given half a letter loss on average per year. This means that in a two year trial, you'll not be able to see anything that would even apply in a two year trial. We can, however, make conclusions from ProgStar data that show when you look at cross-sectional correlations between the size of DDAF and visual acuity, that there is a significant correlation.
Okay, great. It sounds like from your conversations, FDA has certainly embraced the DDAF as a proxy for visual acuity. Can you give us an idea of how other regulators are thinking about the correlation here and acceptance from their perspectives?
Yeah. This is new territory.
We believe that getting approval from the FDA will be a major sign to the communities, including regulators worldwide. I think your question may stem back from the experience that we have seen in the approval process of injectables worldwide for geographic atrophy. We have to take into account that Stargardt disease is not geographic atrophy, and there are two major differences. Maybe the most important is that we speak about a pretty young patient population, and we speak about a very severe disease that kicks in early in life. Also when we think about what EMA published or announced when they looked into other compounds that would promise to slow down GA progression, they took into account safety and tolerability plus efficacy.
When we look at the efficacy signal that we have seen in the DRAGON trial, 36%, and put this into perspective, plus the really, obviously, you can say I am biased, but when I look at the tolerability profile, in my opinion, this was excellent. It is just a pill a day without the need of monthly injections in the doctor's office. When we take efficacy and tolerability into account, plus a pediatric and severe disease, we anticipate that the conclusion of other regulators will be positive.
Okay, great. Maybe just a bit more on the tolerability profile, specifically the side effect profile. How do side effects relate to the mechanism of tinlarebant, and how well characterized are those?
Yes. This is very important because when we look at tolerability and safety of this drug, it must be excellent because the drug is intended to be given for a very long time, if not lifelong. It must have an excellent tolerability and safety profile. When we think about tinlarebant, it specifically acts on retinol-binding protein 4, and that again will and should, and this is what we see in the trial, only have an effect on the target tissue that expresses the RBP receptor, and that is the eye and nowhere else in the human body. That's exactly what we then observed in the DRAGON trial. When we look at systemic safety, we only observed over two years, six SAEs, four of them in the placebo group, two of them in the treatment group, none of them related to the treatment.
When we get to ocular tolerability and safety, we see on target effects. We see two main effects. One is on the rod photoreceptors, the cells that are able to see in the dark, and then the other is on the cone photoreceptors. These are the cells that are able to see under normal light conditions and allow us to have high acuity. I start with the first. The effect that we see is, we call it delayed dark adaptation, meaning if you take tinlarebant, it would not take you 20 or 30 minutes to get to your final threshold of sensitivity when you'll be in the dark and being exposed to very dim light, but it may take you + 5 minutes. So instead of 30 minutes, 35 minutes, or 20 minutes or then 25 minutes.
This is in a predictable and anticipated effect that actually shows the drug works, and it's easily manageable. Rod vision is not completely unimportant to patients, but what counts for patients is can we preserve their vision under normal light conditions? Can we preserve acuity? Can we preserve the ability to recognize faces? All of these functions are mediated by the cone photoreceptors, not the rod photoreceptors. When we get to the cone photoreceptors, there's also an on-target effect. Now we talk about another adaptation process going from dark to light. This is a very fast process. We may have observed that this morning for everybody who woke up, put on the lights. This is a process that takes only seconds or maybe a few minutes, but it's a very fast process in which we adapt to a bright environment coming out of the dark.
During this process, the visual scene, if you are on tinlarebant, can be discolored, meaning some false colors. It's not that you do not see colors, it's just that the visual scene appears yellowish tinted, then we call it xanthopsia, or bluish tinted, we call it cyanopsia. There are all these fancy Greek names for the specific colors that you may perceive. This is a predictable, very benign side effect that only lasts for seconds or minutes and then fades away. Doesn't really interfere with other functions of vision such as discrimination, reading, recognizing faces. But it will need patient education, and tell patients, if you take the drug, that may happen. It's very benign. It will fade away, and it's easily manageable.
Okay, great. And maybe just tying together two reasonably common questions we get, and we have talked about this many times before, but just the confidence in needing a single Phase III study for registration in the U.S. specifically. What kind of contributes to the confidence there? And on maybe a related point, this is obviously an oral drug. You could potentially see benefit in both eyes. So what can you tell us about the fellow eye you have seen in DRAGON?
Sure. I can address the first one. So apparently, we have very constructive, supportive kind of correspondence with the FDA. They have always been very consistent saying that it is going to be a review issue. But given this is a rare disease, they do see there is a possibility for single study approval. Yeah. It is too early for us to comment about that for the FDA. But we have to prize the FDA that they have very strong knowledge and experience about Stargardt disease and DDAF, all these endpoint, and they have approved a 2-H drug based on that. So, we have to say FDA is actually very, very knowledgeable about this, so we are not too worried about that.
Okay.
Yeah. I would like to add another comment and answer the second question and can connect the two.
Yeah.
Which is straightforward. Namely, we did not just submit the trial data to the FDA, but there is guidance of the FDA for single trial approval. That is so-called confirmatory evidence. We put together a large comprehensive package of confirmatory evidence along this guidance of the FDA. As an example, natural history data from the ProgStar study. We have access to the complete ProgStar data set of all these hundreds of patients that were studied worldwide. We can boost the data that in our placebo group in the DRAGON trial significantly, I have to say. This is one piece amongst other pieces of confirmatory evidence that we have submitted to the FDA. That includes, and this brings me to your second question, fellow eye data. The guidance does not include specifics on paired organ systems such as kidney or eye. I don't blame the FDA.
But our fellow eye data is, in my opinion, very strong confirmatory evidence because obviously you cannot optimize your trial for the fellow eye. You have to pick a study eye and then conduct the trial on the study eye. But of course, the patients had two eyes, and we studied the second or the fellow eye as well. The fellow eye, in not all instances, would have a lesion that could be measured or the lesion was very large or too large. Still, the data we collected in the fellow eye reached, not only an efficacy level of almost the same, 33% slowing down of lesion growth, but that measurement also reached statistical significance.
Yeah.
That is something special. I'm practicing for more than two decades now. But over the last decade, pretty much anything that was approved in this space was eye specific. Was it surgery? Was it a device? Was it injections into the eye or gene therapy even? Everything was eye specific. Now, we come up with an oral therapy somewhat unusual in the space that you give a pill a day, and then you see the treatment effect in both eyes. But that's exactly the case with tinlarebant.
Okay, great. Just switching to the potential label for tinlarebant, right? Your phase III investigated adolescents for the most part, and you're going for approval in adults and adolescents. I guess, can you talk about the confidence in adults beyond 20 years old being included in the label and what's the supporting evidence for that?
This is very important, and we all know that conducting a clinical trial is a somewhat artificial endeavor, and you have to do something that is then generalizable to the total population out there. That applies to the endpoint. You need to use an approvable endpoint in your clinical study to be generalizable to the total population out there. You have to enroll a very specific population that still allows generalization to all the patients out there. We included a pediatric or young adult plus the pediatric and adolescent plus young adult population in the study. We did that in a disease that is caused by dysfunction of ABCA4. That applies to, obviously, every patient that carries a diagnosis of Stargardt disease. We do not see any limitation of generalizability to the total population out there.
That also is in line with our correspondence with all regulators we are in discussion with. We have seen how the FDA has dealt with this situation in other conditions, such as geographic atrophy. It is just the other side of the spectrum, right? When we look at the label of the complement inhibitors that were approved for geographic atrophy, although the trial has a relatively narrow range of age, rightly so, there was no age limit in the label that the FDA imposed on the approval of those two drugs.
Okay, great. With priority review, you could launch in the U.S. potentially next year. Maybe I will just kind of loop two questions together here. I guess in terms of setting up commercial infrastructure for the launch, what do you need to be doing? Then also maybe if you could loop in getting patients on drugs specifically, what does a patient need to do to obtain tinlarebant in the commercial setting? How does genetic confirmation factor into that?
Sure. Well, as we disclosed on our commercial day a few weeks ago, there are a few important aspects. First, we do see that there is about 20,000 Stargardt disease patients already clinically diagnosed. Many of them will be with the IRD expert and the retina specialist. That probably will be our team's focus, to making sure that people are well aware, potentially, there is a treatment coming, what does that endpoint mean. So you get the physicians on board about the treatment effect, and I think that is the first thing. Second, the disease awareness will be important outside of that retina specialist IRD community into the general ophthalmologist, low vision optometrists, that they may see Stargardt disease patient from time to time as well.
And we would love to have them to be able to tell, hey, this could be a Stargardt disease patient. Maybe you prescribe a genetic testing, which there are free testing out there. Or maybe just refer to a retina specialist who have very good understanding about Stargardt disease patient. So these two will be the focus and together with some marketing effort, in-person, on-person, newsletter, working with the patient advocacy group to make sure the patient is also well aware this is coming, so they could reach out to their physicians as well.
Okay, great. And we talked a little bit earlier about the side effect profile here. This is a question we're getting as I think we approach potential launch. How important will patient and caregiver education around safety and side effect profile be? Are there any learnings from the clinical program from tinlarebant you can leverage there?
Yeah. So we discussed earlier about safety and tolerability and the two main effects that we see on target, delayed dark adaptation and dyschromatopsia effect. I call it discoloration of the visual scene. When we look at the DRAGON trial, about a third of patients would mention one or the other over the course of two years, but typically in the beginning and no longer in the end of the trial. So, since these are adaptation processes, obviously, there appears to be an adaptation to these adaptation processes in the patients. So, we believe these side effects are easily manageable if the patient would develop them. We also have to be clear when it comes to clinical practice and the introduction of tinlarebant as the first treatment for Stargardt, what that actually means. So we have not been used in the IRD space to have a treatment.
So, amongst all the 300 disease gene, there's just one, namely RPE65. The trade name is LUXTURNA, but the gene therapy was introduced a decade ago. And for all the other genes, there's absolutely nothing available. And I remember the time when the first anti-VEGF treatments were introduced in the market, and for retinal specialists, it was very unlikely that we would inject patients every month. In over one year, I would inject two patients in the operating theater with very severe disease into the eye. It was absolutely not conceivable that we would inject, I don't know, 100 patients per day in a clinic with these injectables. But then the treatment was available, and there was no alternative, and it worked great, and of course, we would inject them.
If we come across skepticism right now, oh, this is a systemic treatment, and potentially there are some side effects. But what exactly is the alternative? The alternative is to do nothing. I do not believe that neither healthcare providers nor patients would like to go that route. As opposed to gene therapy or inject something in the eye, you do not get rid of it. With tinlarebant, you can simply discontinue the therapy if there was ever a problem, and then see how the patient is doing. I do not believe that those two benign side effects that are on target would really prevent anybody from neither prescribing nor receiving the therapy.
What is your latest thinking on potential pricing for tinlarebant in the commercial setting? Can you talk about how your payer interactions may support your thoughts there?
We have gone through a round of payer research or survey, and we can say that the payers are very supportive. They are well aware of this disease cause, in all cases, legally blind, and many of the patients are still very young, may have average life expectancy of 60 years.
None of the payers that we talk to seem to be not being very friendly and supportive about this treatment. The price range, apparently still too early t o confirm a price. We do see this to be a rare disease pricing, and it seems that the payers are very okay with that as well.
Okay, that's great. Just maybe one more before we get into our little mini survey that we are asking everybody. You finished enrollment in the DRAGON II study. Maybe just remind us the purpose of that study and when we can see data from that.
The DRAGON II study was initiated because Japan was not fast enough to participate in DRAGON I, and then provided Belite Bio with the Sakigake designation, meaning breakthrough designation. Obviously, the PMDA, the regulatory agency in Japan, would like to see some data in Japanese patients. That led to the birth of the DRAGON II program, which is a phase I-B, PK/PD/phase III clinical trial. We took an opportunistic approach to allow patients in the U.S. and in the U.K. that were still waiting to participate in this DRAGON II program. To be clear, DRAGON II is being conducted for the PMDA for approval in Japan. We have collected and made publicly available the PK/PD data. We presented that already one and a half years ago at the ARVO vision conference, and confirmed that the PK/PD profile in Japanese patients is, not surprisingly, no different to populations of other descent.
We are currently conducting this placebo-controlled phase III program. We are fully enrolled since January, have 15 patients from Japan in the trial. Overall, 73 patients in a one-to-one randomization. Otherwise, an essentially identical clinical trial design to DRAGON I. We have actually never discussed about the DRAGON II program with the FDA, so the FDA-
Sure.
does not expect us to submit anything from that program for our NDA in the U.S. In Japan, we submitted also our NDA that is essentially based on the DRAGON data.
Yeah.
PMDA is fully aware that since they want to approve tinlarebant as the first country in the world, which is ambitious, but we like that ambition, that this process is essentially based on the DRAGON I data, and we submitted the package, and we announced that, I believe, beginning of last week.
Okay, great. With that, I think we're up against time. We could keep going, but thank you very much for being here, guys. We really appreciate it.
Thank you for having us.
All right.
Thank you.
Great.