Thank you. Thank you. All right, for our next session. Well, first, let me just say I'm Marc Goodman. I'm one of the biopharma analysts at Leerink . Thanks everybody for joining us. Our next session is with Belite Bio. We're lucky enough to have the CMO, Hendrik Scholl. Obviously, if you've watched the story of this stock, it's been quite an interesting and fun ride over the past year, and we're very hopeful that this drug is gonna get over the goal line nicely with this really good data. Thank you very much for joining us. I thought we'd start out just to kinda level set everybody and talk about, you know, just Stargardt, what is Stargardt? What's the problem? What needs to be kinda addressed?
We'll talk about your product and the mechanism of action of your product. I think that's a nice segue. Thank you.
Very good. Very good. Thank you.
Yeah.
Stargardt, obviously described by a German ophthalmologist with the last name Stargardt in 1907, is orphan disease. It's not so rare. We get to that in a moment. It affects the center of the retina, and it leads to blindness in almost all cases. It's recessively inherited and is due to mutations in a gene with the name ABCA4. When you think about the retina, it's the leading tissue in the human body being affected by mutations in genes. We have more than 500 genes that, if mutated, give rise to a retinal degenerative disease. Stargardt sticks out because it's by far the most prevalent disease gene, accounts for about 30% of all cases of inherited retinal diseases.
Currently, we believe there are pretty much exactly, 53,000 patients affected in the United States. It's not so rare.
$50,000?
Yes.
Yeah.
The disease is in phenotypic expression identical all over the world in different geographies and races. The prevalence is slightly different in populations of African, European, and East Asian descent, but when we take all populations together, it's about 1 in 7,000. Amongst all the inherited diseases in medicine that are recessively inherited, it's in the top 10. It's typically number three or so in the ranking of the most prevalent inherited diseases in medicine. It's not so rare. Clearly comes with a significant market opportunity given the fact that Stargardt disease is an completely unmet medical need. There's currently absolutely nothing that is approved. Patients are being advised to protect their eyes against bright sunlight, which is theoretically helpful.
They should not supplement vitamin A, which, for many, in the beginning at least, appears counterintuitive because the retina needs vitamin A. The human body needs vitamin A. In other diseases, sometimes my colleagues would recommend to supplement vitamin A. In Stargardt, this is not recommended.
Talk about your product and how it works.
Happy to. When you think about ABCA4, its role in the retina, you first have to know that in the retina there's the visual pigment, which is a miracle because that's our connection to the outside world. Without this molecule, we would not be able to see. This molecule is our link to the outside world in a chemical process. Light hits rhodopsin, that's the name of that compound made of vitamin A, and this goes through a number of conformational changes, and eventually this leads to a mechanism that then activates the cell, the next cell, the next cell, and eventually the brain, and this is the visual percept.
If there wasn't the recycling of that molecule, we would only see once at birth and then never again, and therefore there's another process in the visual cycle, in the, in the photoreceptors called the visual cycle. That is the regeneration of that, of that photopigment. It needs to be regenerated all the time. ABCA4 comes downstream in the moment when the visual pigment goes through conformational processes and we find a molecule with the name all-trans retinal, which is very toxic. This needs to be detoxified. It needs to be flipped outside the cell. ABCA4 has this flippase activity. If it's mutated, we have dysfunction of the protein.
This flippase activity goes down to nil, and all-trans-retinal accumulates in the photoreceptors, and it reacts with other molecules of all-trans-retinal and forms so-called bisretinal. Bis meaning two. Retinal comes from retinol or vitamin A, meaning molecules that are made out of two molecules of vitamin A. These molecules accumulate in the pigment epithelium, which is the support layer of the retina, and are toxic. The idea was can we do something that reduces the formation of these toxic bisretinals? The answer is obviously yes. Otherwise, I wouldn't sit here. It's the idea that retinal delivery to the eye should be reduced. Can we reduce vitamin A delivery to the human body? We can, but that would not be healthy.
We would only like to reduce retinal delivery to the eye and to no other organ in the human body. That is possible because the eye expresses a receptor to suck in vitamin A, namely the RBP receptor, retinol binding protein receptor, and only the eye expresses this receptor. That's the reason why you can modulate the intake of vitamin A into photoreceptors and no other organ in the human body by reducing available product that would be sucked in through the RBP receptor into the photoreceptors. Our idea, Belite Bio's idea is we take an antagonist of RBP4, retinol binding protein four, that leads to reduction of available product for the RBP receptor. The product is called Tinlarebant. Tinlarebant reacts with retinol.
This is then eliminated through the kidney. We have much less substrate available that is being sucked into the photoreceptors. We found in early phase clinical studies that 5 mg per day only is sufficient to reduce retinol and RBP in the systemic circulation by 80%, leaving 20% remaining. That allows to significantly reduce toxic accumulation at the level of photoreceptors in Stargardt disease.
That's the 5 mg that you were testing.
Exactly
in the early study.
It's a pill, one pill a day. It's maybe, for people that are active across all disciplines in medicine, nothing very surprising that we have an oral product. In retina, it is surprising because when you look at the products over the last 20 years, almost everything was eye-specific, being an injection, a device, a surgical approach or what have you. Now we come up with a systemic drug that is taken orally and obviously has a treatment effect in both eyes simultaneously.
Before we get to the data from the phase III study, just talk about the early work that you did and what gave you confidence that you had the right dose and just, you know, walk through that a little.
In preclinical, in pre-preclinical studies, you can simulate Stargardt disease in mouse models, and it was shown that if you reduce RBP4 by 80% or you leave 20% RBP for retinol available for photoreceptors and RPE, that you have a very strong treatment effect, and you essentially rescue a mouse model. You're in a so-called double knockout mouse model, RDH8 plus ABCA4. You find that you can actually stop photoreceptor degeneration completely when you treat with the equivalent dose of 5 mg per day, Tinlarebant. With that knowledge, we went into a phase I-B/II clinical trial. This was an open label interventional trial in adolescent Stargardt patients in our trial in Taiwan.
In the trial, we found that vision was stabilized over the course of two years in the study. Very few patients transitioned to late phase disease, and there's a feature we can actually measure. We'll get to that when we get to the DRAGON trial that we used as the main outcome measure in the registration trial, namely a lesion that we call DDAF, definitely decreased autofluorescence, which is approvable endpoint by the FDA. These lesions could be measured in their growth rate over time in the phase II clinical trial. Since it was open label, the comparator was a natural history study, the ProgStar study, and we found that the progression rate in these treated patients was only about 50% compared to those in natural history studies.
How old is that natural history data?
The, the ProgStar study was actually implemented at the time when I was professor at, the Wilmer Eye Institute at Johns Hopkins. Foundation Fighting Blindness that funded this large worldwide natural history study suffered from the fact that at the time, 40 mouse models could be successfully treated with various inherited eye diseases, but only one gene therapy made it to the market in 2016. Until to date, there's only this one available. One of the problems was there was no outcome measure. That also applied for Stargardt disease, and that led to the idea, let's characterize the natural history of Stargardt disease, and let's establish outcome measures in the disease.
We implemented this worldwide natural history study in 2014 and saw about 300 patients, 291, in 10 centers around the world every six months with imaging and visual function tests. We were able to characterize the natural history, and we were able to show that specific lesions, people that spend time with geographic atrophy or drugs approved for geographic atrophy know that you can measure these lesions in geographic atrophy. We call them geographic atrophy. In Stargardt, we are a bit more precise. We call them definitely decreased autofluorescence, DDAF. These DDAF lesions were shown to be that we can precisely measure them, and we can precisely characterize the progression, namely the growth rate of these lesions.
subsequently, this was approved by the FDA as an outcome measure.
Okay, let's move to the phase III study.
Happy to. The phase III trial, the so-called DRAGON trial, was a multicenter, interventional, double-masked placebo-controlled clinical trial, in which 104 patients were enrolled worldwide in the U.S., in the U.K., in Europe, in China, in Taiwan, in Australia. These patients were randomized 2- 1 to either receive 5 mg per day of Tinlarebant, and one-third of patients received placebo. Patients needed to have a clinical diagnosis of Stargardt disease, and at least 1 mutation in the target gene, namely ABCA4. Obviously, they needed to have a lesion that can be tracked over time, these DDAF lesions, definitely decreased autofluorescence, and they needed to have minimum visual acuity of 20 to 100.
Patients were investigated every 3 months with visual function tests and imaging over a period of two years. We did an interim analysis after all patients had at least received 12 months of treatment. We had our final readout in November and presented on December 1st our top line data, where all the patients completed the trial. We could show that our primary endpoint was met. We could with 5 mg per day over two years reduce the progression rate of DDAF growth rates by 36%.
This was highly statistically significant, yielding a P value of 0.0033, applying a structured covariance matrix to the data, which was recommended by our DSMB, led to a P value that was smaller than 0.0001. This was highly statistically significant. We also found that our secondary endpoint, called DAF, definitely decrease autofluorescence, DAF. Not definitely decrease, it includes also questionably decreased autofluorescence, meaning it's not as dark, it's just decreased DAF, meaning that the underlying image tissue, the retinal epithelium, is sick but not completely degenerated. Meaning that if we see an efficacy signal for that outcome measure, it means that not only can we prevent complete neurodegeneration, we can also prevent the formation of sickness of those cells that eventually turn into complete neurodegeneration.
For that secondary outcome measure, we had essentially the same treatment effect, namely, 34%, and also reached statistical significance, which is interesting and nice. This is again, has become relatively rare in retina, but we have a fellow eye treatment effect. We have a systemic treatment, so obviously we looked at the fellow eye, and in the fellow eye, we had essentially the same effect size. 33% effect size in the fellow eye. This also reached statistical significance, which when you think about it's like a miracle because everything was optimized for the study eye, right? Inclusion criteria, exclusion criteria, lesion size, what have you. For the fellow eye, all comers were enrolled, right? Every fellow eye of a given study eye of every patient.
Still we hit the statistical significance level for the primary endpoint and also for the secondary endpoint.
What % of patients have Stargardt in both eyes?
All.
They do.
Uh, e-
It's just a matter of when it starts to show.
Exactly right. When I see... I still see patients one day a week in Zürich, Switzerland. When I look in the right eye, I would typically know how the left eye looks like, but it's not identical. The two eyes look similar, but not identical. Again, right, finding a treatment effect of the same magnitude and reaching statistical significance for the fellow eye, for me at least, was not anticipated.
Hmm. Interesting. Okay, so what do we do now? What's the plan?
Let me before I comment on that, I would like to point out that we discuss a very safe drug.
Oh yeah, we didn't talk about safety.
Yeah, our drug, we encountered six serious adverse events. Four of the six in the placebo group, two in the treatment group. None of them was considered related to the study drug. There were only essentially only ophthalmic treatment emergent adverse events, and these were completely anticipated and in line with the mechanism of action. When you think about it, substrate is being reduced flowing into the photoreceptors, meaning less visual pigment is available. That should, and does, immediately affect the time it takes for the eye to adapt to darkness, to adapt to light. What does it mean? When you enter into a completely dark room, it would take you 20 minutes to reach the final threshold of the maximum sensitivity, right? To be fully sensitive in the darkness.
If you take Tinlarebant, it's 5-10 minutes longer. You reach the final threshold. It just takes longer. You leave the room, go out in the sun. It takes a few seconds, maybe one or two minutes, to be fully light adapted. This is also affected by Tinlarebant. When you do that, and you take Tinlarebant, the visual scene can appear discolored. If it's yellowish, we call it xanthopsia. If it's reddish, we call it protanopia, and so forth. This discoloration of the visual scene is something that would be anticipated, and this was indeed observed in the clinical trial. A third or so of patients experienced this discoloration of the visual scene lasting for seconds or minutes called xanthopsia or dyschromatopsia.
Another third also noticed this lengthier or longer time to dark adapt or dark adaptation problem. These were the most prevalent adverse events. Almost all of them rated mild, and many of these instances resolved with time. Although still at the end, we had been able to measure the little difference between Tinlarebant and placebo-treated patients in terms of dark adaptation time.
Basically, a physician sits down with a patient and says, "Look, if you don't do anything, you could lose vision. You could go blind, or you could take this product, which hopefully will be approved. It's the only one, and you'll have a 30 something% benefit, and you have to wait 5 to 10 minutes longer when you go from a dark room to a light room or a light room to get back to normal versus somebody who doesn't have the. That's the pitch.
Exactly, Marc. To make it clear, yes, of course, we see an effect on the rod visual system, which are the photoreceptors that are able to see in the dark, it's interesting to know that 20x as many photoreceptors in the human retina are made for vision in the dark, only 5% for normal light conditions. Our normal life is spent with cones and not with rods. There's a condition called stationary congenital night blindness, CSNB, meaning that the rod photoreceptors are dysfunctional. There's no rod vision essentially, right? Many of these patients are undiagnosed. They are perfectly fine, right? They don't see as well in the night, they can read, recognize faces. They can do everything, right?
While if you have only 5% of your photoreceptors, but all the cones dysfunctional, this is also a disease that exists called achromatopsia, your visual acuity will be 2,400, so 5%, right? You're extremely sensitive to light, and you're extremely handicapped. You cannot read, you cannot recognize faces, and so forth. Why do I mention that? I mention that because I feel as a treating physician, to save the cones on the expense that the rod need more time dark adapt is not a difficult sell, right? What counts for patients in the long run is can they retain their cone mediated vision? Can they retain their reading ability? Can they retain recognizing faces for longer? This is what counts for patients.
Okay. When do we file the product? How is that going?
Top line data presented December 1st last year. We are in the process of finalizing the clinical study report, the CSR. We promised to submit to the FDA in the 1st half of the year. We are definitely on track. We are likely ahead of schedule. We anticipate final feedback approval from the FDA in the 1st quarter of next year, and then being on the market in the United States.
Excellent. The plan for other regions of the world?
Essentially, we are in communication with essentially all regulators around the world. I presented the interim data in Tokyo last April. I presented the interim data to the NMPA in Beijing last September. We will definitely file a submission to the EMA, to the MHRA in the United Kingdom. We aim for submission for Swissmedic in Switzerland, to in Australia. We started a subsidiary in Japan. We have a team in Japan now, and we have a second trial ongoing, the so-called DRAGON II trial, to fulfill the regulatory requirements in Japan to get the product approved in Japan.
We have the Sakigake designation in Japan, meaning that the Japanese government aims to approve Tinlarebant as the first country in the world, which is ambitious. We obviously collaborate with them to make that a realistic goal. Therefore, we have the DRAGON II trial running, which was promised to enroll 10 patients. We have currently 15, and we took the advantage of that trial to also allow patients in the U.K. and the U.S. being enrolled into the trial. We have currently 72 patients enrolled into this second registration trial that we only do for the Japanese authority because Work with the assumption that we will only need one trial for all regulatory agencies.
72 in total, but 15 are Japanese patients.
15 of the 72 are Japanese.
Right. We don't need the endpoint for those 15 patients? Just to make sure we're clear.
The Japanese authority, the PMDA, understands that it is impossible to show a statistically significant treatment effect in only 15 patients.
Mm-hmm.
They wanted us to do a PK/PD study as part of this exercise in Japan. We did that. We completed it and presented it last May. No surprise, obviously, PK/PD was exactly the same in Japanese patients. The same is anticipated for the behavior in placebo-controlled exposure over two years in these Japanese patients. The approval process in Japan will be based on the DRAGON I clinical trial, plus the data that we have collected until we submit to the PMDA.
All the filings external from the United States will all occur, what, over the next 12 months kind of thing? Is that?
Correct
kind of how you're thinking about it?
That's correct, yes.
Just to make sure. The U.S. first and just, then what's next? What's the second most important?
Yeah, second most it will likely be Japan.
Uh-huh.
Right? or EMA or MHRA.
Do those regions also have the same prevalence as the U.S.?
Good question. Right. The prevalence of ABCA4 mutations worldwide is the highest in populations of African descent.
Uh.
Second highest of European descent and third highest East Asian descent. In Japan and in China, the prevalence is roughly 1 in 10,000-11,000, while in populations of European descent is closer to 1 in 6,500 or 7,000.
Okay. There's another study that's being run with this drug for GA.
Geographic atrophy.
... GAs. Yeah, talk about that study.
Exactly. Happy to. The so-called PHOENIX trial. The PHOENIX trial was completely enrolled ahead of schedule, mid of last year. 530 patients that are being randomized to receive exactly the same dose, namely 5 mg per day, against placebo, to reduce the growth rate of geographic atrophy secondary to age-related macular degeneration. Phenotypically, a very similar disease. It comes with essentially the same features. It affects the macula. It has these lesions that can be measured and tracked over time, meaning that we have essentially the same outcome measure. It's just called differently. It's called geographic atrophy, measured on fundus autofluorescence imaging.
Since it's not the same disease, we cannot promise we will have the same treatment efficacy that we observed in Stargardt disease, but we believe that the current bar of the approved therapies in the United States, targeting complement component C3 and C5, having shown a treatment effect of 13%, 21%, and 14% in the OAKS, DERBY, and GATHER2 trials is so low that we should easily beat that. We have an oral compound. We feel in the moment we hit that target that our compound will be the standard of care compared to the injectables that we currently have that need to be injected monthly in order to have this treatment effect that I just explained.
The interim look is. What's the plan? What are you thinking?
The plan would be to do an interim analysis as well as we did for the DRAGON trial. Now we get to commercialization and commercial implications. It will put us in a complex situation in the moment we see strong efficacy in GA because we would not like to have our Stargardt market affected, right? We have several strategies that we developed inside Belite Bio to deal with that. As an example, we have 425 RBP4 antagonists in our IP portfolio that we can use and develop as an alternative therapy to not put our Stargardt market in danger and still develop a therapy for geographic atrophy.
Hmm. Interesting. That would be a plan to put in place next year? I mean, this is.
Yes. Yes.
this is all next year.
Yes. That would be second half of next year.
Okay. Got it. Got it. What else? What did we miss? Anything? I guess commercialization plans. Maybe you can talk about that for the last minute.
We are on track. We hired the most important positions already, our head of commercial in United States. It's clearly important, when you think about reaching the patients in the United States, what is important. Genetic testing is important. We talk about a monogenic disease. Given the anticipated price tag of Tinlarebant, clearly genetic confirmation will be required, meaning that we will help to lower the bar, that patients will receive genetic testing and genetic confirmation of mutations in ABCA4. This is one of our biggest efforts that we are currently implementing for the United States to have as many patients as possible genetically confirmed in order to then make them candidates for Tinlarebant.
Interesting. Part of the hiring has taken place already?
Yes.
Yeah. ready to go in the U.S.
Yes.
That's the plan. Good. Good. Good. Good. Thank you. That was very good.
Thank you, Marc.
Yeah. Anybody? Any questions anybody on just in the last minute?