Hendrik, you want to take that?
I'm happy to take that. Belite Bio is a biotech company targeting unmet medical need in severe retinal diseases. We have our main compound with the name Tinlarebant that went through a completed phase III clinical trial named DRAGON that we discuss in a moment, I believe, for the indication Stargardt disease, which is a severe inherited retinal disease and is the leading cause of blindness in adolescents and juveniles. We also developed that compound for another indication called geographic atrophy secondary to age-related macular degeneration, which is very similar in its phenotype to Stargardt disease and shares significant parts of pathophysiology with Stargardt disease. This briefly is where we are as a company, and we are happy to take your questions.
Great. I guess with Tinlarebant, can you talk about the mechanism of Tinlarebant and how it potentially addresses disease progression in Stargardt as well as in geographic atrophy?
Certainly. A little bit of background information. There is a very important mechanism that happens all the time in the eye called visual cycle, meaning the regeneration of the photopigment. If there wasn't any regeneration of the photopigment, we could only see once at birth, and it would be dark, meaning the photopigment needs to be regenerated all the time. That process is called the visual cycle, and it's dependent on vitamin A available in the retinal pigment epithelium, which is the support layer of the neuroretina. In Stargardt's disease, which is due to mutations in ABCA4, which is the name of the gene, the corresponding protein is dysfunctional, and therefore, the function of ABCA4 cannot be properly conducted, meaning that what it typically does, namely flipping out used visual pigment, we call that all-trans retinal, which is very toxic itself.
To flip that outside the retinal pigment epithelium in order for it to be regenerated, this function can no longer take place, leading to accumulation of all-trans retinal, which has a propensity to react with itself, and leading to also very toxic so-called bisretinoids. Bis stands for two, retinoid stands for derived from vitamin A. Bisretinoids means these types of toxic molecules that accumulate in the eye because they cannot be further degraded. That is the pathophysiology of what happens in Stargardt disease. Because of this toxicity that happens in the RPE, in photoreceptors, this accumulates over time. It leads to photoreceptor dysfunction and then photoreceptor cell death. The macula is especially affected by the disease, therefore Karl in 1909, 115 years ago or so, described this as a macular disease affecting the very center of the retina.
We have a quite similar pathophysiology in age-related macular degeneration, which is also due to dysfunction of retinal pigment epithelium. We know that there are bisretinoids accumulating in the RPE, and we know, for example, from a very large natural history study that we did at the time with Frank Holz in Germany, the largest natural history study at the time, the FAM-Study, where we could show that areas of increased autofluorescence, which we use clinically to see these areas with increased bisretinoid accumulation, are a main driver of geographic atrophy. The two diseases are, other than the age of the patient, obviously, which is very different because GA is a disease of the very old, while Stargardt disease typically affects children, adolescents, and young adults, is otherwise quite similar.
When you look at the macula and the typical imaging such as OCT, autofluorescence imaging or fundus photography, the two diseases are very similar, sometimes indistinguishable. There is significant overlap between the two diseases, and it makes perfect sense to develop Tinlarebant to make a difference here, namely target retinol binding protein 4, which is important for the retina to get access to vitamin A. Since we just learned that these bisretinoids are made of vitamin A, the idea is we reduce vitamin A availability in the eye. Can we do that specifically because vitamin A is important for other tissues in the human body? The answer is yes, because only the eye expresses the retinal binding protein receptor, the RBP receptor, which makes the eye exclusively dependent on getting access to vitamin A bound to retinol binding protein 4.
Tinlarebant is a small molecule, a retinol binding protein 4 antagonist that allows to very precisely reduce vitamin A bound to RBP4 and thereby reducing the amount of vitamin A available in the visual cycle, meaning in the photoreceptors and the retinal pigment epithelium. This can be done very precisely, almost, let's say, with a precision like gene therapy, because it's dependent on this one receptor. With only 5 milligrams per day, we have shown that we are able to reduce RBP4 bound retinol available for the eye by 80%, pretty much exactly 80%, leaving 20% available. That allow us to essentially arrest the further accumulation of these retinoids at the level of the retina.
You've recently announced that you've completed a rolling NDA filing with the U.S. FDA in Stargardt. Can you talk about, I guess, would you expect to get a priority review based on this filing? What's the FDA's feedback been in terms of your pivotal trial and the endpoints that you've used in the trial? Anything to glean from your pre-NDA discussions with the FDA?
Yeah. Happy to take that question. I would like to start with the fact that the FDA has been involved in this endpoint discussion from the beginning. Beginning meaning when, at the time, Foundation Fighting Blindness, including myself at the time as Professor of Ophthalmology at the Wilmer Eye Institute, which as you know, Johns Hopkins University being just around at Bethesda. When we met with the FDA to discuss the design of the largest natural history study ever conducted in the space called the ProgStar study, where we took the advice of the FDA, how such a clinical study should be designed and put DDAF, definitely decreased autofluorescence, as the primary endpoint.
Obviously, there was precedence because at the time when ProgStar was developed, the first clinical trials in geographic atrophy due to AMD were already conducted, and geographic atrophy measured on autofluorescence images was the accepted primary endpoint at the time for compounds developed for GA treatment. It was a natural development that this became the primary endpoint. In the ProgStar study, we could show that DDAF indeed progresses in a predictive fashion and can be measured precisely by reading centers. We had a central reading center for the ProgStar study, and we could show that the progression rate of these DDAF lesions, very similar to geographic atrophy, progress at a certain pace. It was about 40% of what we see in geographic atrophy, so it's slower than in GA.
Obviously patients are much younger, so they do not have 6 to 8 years of life ahead of them, but 60 to 80 years, so they have many more, 10 times as many years ahead of them on average, when they have received the diagnosis of Stargardt disease. Although it's slower progressing, the impact on the life of the patient is very severe. ProgStar helped to establish this primary endpoint with the help of the FDA. Obviously we have been in constant discussions with the FDA, including the submission of our interim results to the FDA, which resulted in the breakthrough designation. Given that we received breakthrough designation as a company for Tinlarebant for the development of a treatment for Stargardt disease, we believe it is fair to assume that we will receive a priority review for our submission.
That's perfect. How about Europe? Could you file with this data set with the EMA?
With the FDA, with the EMA, we have an approved PIP, pediatric investigation plan. I think it's important to understand that your question likely refers to what happened at the time when lampalizumab was submitted as a possible treatment for GA in Europe. At the time, it's important to understand that the EMA's position that was published clearly referred to the risk-benefit ratio of their treatment. Given the modest treatment effect at the time, and the fact that you only reach that treatment effect by monthly injections into the eye. Now we speak about geographic atrophy, obviously, right? It's a very similar endpoint. The EMA concluded that the risk-benefit ratio was not very favorable. Right?
I think it's important to understand there was a very modest treatment effect when we think about the DERBY, OAKS, and GATHER2 trials for GA, 12%, 21%, and 14% reduction in growth rate when you compare that with 37% for Stargardt disease. The fact that this is an oral treatment with the ease of application and with a very favorable tolerability and safety profile that this may, could lead to a very different perception of this treatment for Stargardt disease as they position themselves earlier when this was submitted as a possible treatment for geographic atrophy. There's no guarantee, obviously. There's more and more data. I would also like to emphasize this, because at the time, the correlation between geographic atrophy and visual acuity loss was already established to a degree.
This is also the case in Stargardt disease, there's more and more data coming out of the ProgStar study. Some will be presented at the Academy. There's also independent data from Stanford University that was presented at the ARVO meeting in May that shows that there is a correlation of DDAF and visual acuity. This is important because regulators clearly understand visual acuity. The FDA definitely understands structural endpoints. The FDA, in my personal opinion, I can say that as a non-U.S. citizen, is a great agency. Right? Has a deep knowledge about structural endpoints in ophthalmology, OCT, and autofluorescence-based endpoints, and clearly understands that. Other agencies also understand the correlation of structure and function.
The more data we have on structure function correlation, which is significant for both Stargardt and GA, the easier it will be to convince agencies that for a condition like Stargardt, where visual acuity decline is significant over the lifetime of the patient, but is very slow in a limited amount of time of a clinical trial, such a year or two years, it's essentially impossible to find any treatment for this disease based on visual acuity alone. In ProgStar, we were able to show that the average visual acuity decline in Stargardt patients is 0.55 letters per year, so half a letter on an ETDRS chart per year.
Given the intersession variability of visual acuity measurements in patients with macular degeneration, which is eight letters, shows that it's essentially impossible, right, in a reasonable amount of time to find or to prove efficacy of any treatment for the disease. Therefore, we need other endpoints that eventually correlate but are much more precise, such as autofluorescence or OCT-based endpoints.
You're currently also running DRAGON II. What's the objective of DRAGON II? I think it's a 73-patient study across U.S., U.K., Japan. Would love to go through the objective of the trial and when can we expect top-line data from the study?
Yeah. I think it's important to understand that we believe that the DRAGON trial already provides compelling evidence that Tinlarebant is safe and efficacious for the disease, for which there are no other treatments available. We also have substantial supporting evidence, not including DRAGON II, but DRAGON II would be another important piece of supporting evidence to support our application based on this one registration trial. DRAGON1 was originally designed to enable approval in Japan. Right? At the time of the implementation of DRAGON1, Japan could not participate in the clinical trial. The Tinlarebant as a compound received the so-called Sakigake, or pioneer drug designation by the Ministry of Health, Labour and Welfare in Japan for the treatment of Stargardt disease.
In order to collect data in Japanese patients, but allowing patients in the U.S. and the U.K. to also participate because many could not join DRAGON1, we initiated the DRAGON II trial with essentially the same design. DRAGON II is designed to enable approval in Japan, and it's completely enrolled. We have a significant number of Japanese patients, but we also have patients from the U.S. and the U.K. participating in the trial.
Basically, similar design.
I think this speaks in favor of our development program. In all of our trials, DRAGON1, DRAGON II, the previous phase II clinical trial, and our clinical trial in geographic atrophy called PHOENIX, we have exactly the same dose of five milligrams and the same frequency per day for the treatment of those diseases. We have substantial PK/PD data to support our market authorization applications. We can follow exactly the same design for those two indications, including very similar designs of DRAGON1 and DRAGON II. There are only two differences. One is that in DRAGON II, instead of a 2-to-1 randomization, we have a 1-to-1 randomization. The other is that we allow a visual acuity to still participate of 2,400. The cutoff was 2,200 in the DRAGON1 trial. The latter just allows more patients to join.
The first, the 1-to-1 randomization actually provides us with a larger power to detect an effect. It's just that a 2-to-1 randomization is easier for enrollment, finding patients because simply the probability that they would have twice as high of a chance to receive the treatment than placebo is clearly a motivation for patients to participate in the trial. Now with our successful phase III or the DRAGON1, the motivation of patients to participate has clearly further increased, and we completed enrollment really fast after the data was presented. The 1-to-1 randomization is superior in terms of statistical power. There was no problem in finding the patients for our DRAGON II trial.
From a safety standpoint, there seemed to be some delayed dark adaptation associated with-
Yeah.
event.
Yeah.
Can you just talk about read-throughs to how FDA will view it? Will they require monitoring of these events? Anything around labeling or REMS?
Yeah.
Maybe you could just talk through your impressions on that and-
Yeah.
Is this something that could impact commercial adoption?
Yeah. This is very important. Obviously, safety is key. What have we found in the trial? We found that the majority of these events, one is being xanthopsia, I'll explain in a moment, and the other is delayed dark adaptation. The majority of the events were mild, and most resolved while on study. There were no serious ocular treatment emergent adverse events. Only four TEAEs led to study drug discontinuation and two of them to study discontinuation. Briefly, what is xanthopsia and what is delayed dark adaptation? Both have to do with visual adaptation, and we have to be clear, visual adaptation problems are part of the underlying disease, of Stargardt disease.
When we as healthy individuals when it comes to retinal health go from a dark room to a very well-lit room, typically early in the morning, there is this process called a light adaptation, which is very fast. It lasts only seconds or maybe a minute or two, where we adapt to bright light. When even we as healthy individuals would take Tinlarebant, because of a slower availability of visual pigment to the photoreceptors, this light adaptation process can be somewhat disturbed, leading to a discoloration of the visual scene. If it's yellowish, it's called xanthopsia. If it's reddish, would be called erythropsia, or if it's bluish, called cyanopsia. There are all these Greek names to this phenomenon. What is it really?
It means that the visual scene is briefly discolored, meaning I look at the screen, I can see your face, I can see the border of my screen, but it could appear reddish or bluish or greenish or yellowish, and that will disappear within a few seconds or minutes. The acuity is not affected, meaning this is a very benign effect. You have to educate your patients that that could happen, right? They need to be prepared, meaning patient education will be key. This is not a major concern. The other is delayed dark adaptation. 95% of our photoreceptors, which may be surprising to some of the audience, is actually rod photoreceptors, the cells that are able to see in the dark and not in the light. I would also like to make that footnote that everything that patients care about is cones, not rods.
Although it is only 5% of the population because we see with cones. When I look at the screen to recognize your face, it is exclusively cones. We need to save cones in the disease. Rods are, I do not want to say unimportant, but clearly secondary, not primary. What is delayed dark adaptation? I mentioned the shortage of vitamin A to photoreceptors that also affect the rod photoreceptors.
Dark adaptation, which is the job of rod photoreceptors, this is a long process. It is not seconds or minutes. This meaning many minutes, typically 20 to 30 minutes. When I go from a very well-lit room to a completely dark room, it would take me more than 20 minutes to fully dark adapt. If I, as a healthy individual, take Tinlarebant, it would take me 25 instead of 20 minutes, as an example. Same for patients. It is not easy to distinguish.
Many patients actually complained about delayed dark adaptation in the placebo group because it is part of the pathophysiology. Clearly more, and it is the mechanism of action in the treatment group. That again means it is a very benign side effect. It only is a delay. The final threshold, meaning how sensitive you can get, right, is not affected by Tinlarebant. The time it takes to get to that threshold is affected, meaning you have to educate your patients, right? Avoid abrupt changes from light to dark to light, right? Take some red LEDs in your bedroom, right? These are very benign side effects that can be managed, but they need patient education. I think we can live with that. Patients will understand that they do it for a reason and their dark adaptation may be somewhat affected.
We do not believe that this is a big issue.
Great. We are about out of time. I had many more questions. Would love to have you on again. We should do a full hour, because I think clearly a lot more to talk about regarding commercial outlook, which I think is very attractive given this would be the first therapy for Stargardt disease as well as the potential in GA. Would love to thank both Hendrik and Haohan for coming here and participating.
Thank you for having us. Thank you very much.
Thanks, everyone.