Belite Bio, Inc (BLTE) Earnings Call Transcript & Summary

Good afternoon, and welcome to the Belite Bio Q2 2023 Financial Results Conference Call. [Operator Instructions] As a reminder, this call is being recorded, and a replay will be made available on the Belite Bio website following the conclusion of the event. Before we begin, I would like to bring your attention to the forward-looking statement slide. During this call, we may be making forward-looking statements. Please refer to the language on this slide for further reference. On today's call, we have Tom Lin, Chairman and CEO; Nathan Mata, CSO; and Hao-Yuan Chuang, CFO. With that, I'd like to turn the call over to your host, Tom Lin, Chairman and Chief Executive Officer at Belite Bio. Please go ahead, sir.

Thank you, Sarah. Thank you, everyone, for taking the time to join this meeting. I'm Tom Lin, CEO of Belite Bio. I'll start off by giving the overview and the milestones we have achieved so far. So for this that -- so for those that are new to the Belite story, the drug that we are developing Tinlarebant is a novel once-daily oral tablet designed to bind to retinol binding protein, as a means to specifically reduce retinol delivery to the eye. This approach is intended to slow a healthy formation of toxic retinol derived byproducts, which are generated in the visual cycle and are implicated in progression of Stargardt disease and geographic atrophy, secondary to dry AMD. Belite Bio believes that earlier intervention directed at emerging retinal pathology, which is not mediated by [ information ] would be the best approach to potentially slow disease progression in Stargardt disease and GA dry AMD. So there is still significant unmet need for both indications. As currently, there is still no approved treatment for Stargardt disease, and there are currently no approved oral treatment for GA, which oral treatments are expected to capture a much higher -- much wider market for advanced dry AMD. We have so far received Fast Track designation, Rare Pediatric Disease designation and orphan drug designation, which allows us to frequently discuss with FDA of our progress and see how we can expedite the approval of this drug will show positive results from our Phase 3 study. I would also like to mention that we still have a long patent [ live ] with the first composition of matter patent expiring in 2035, and this is without patent extension. And with new patents being filed, which will extend the patent portfolio into [ 2050 ]. Now in terms of important milestones achieved this quarter, our Phase 2 18-month treatment data continues to show slowing of lesion growth. We are also expecting our Phase 2 24-month final data readout in Q4 this year. We've also recently completed enrollment of our global Phase 3 Stargardt trial, and we are now expecting interim readouts around mid-2024. We've also started enrollment for our global Phase 3 trial in GA dry AMD. And with this, I'll pass it over to Nathan to go through the clinical trial results. Nathan?

Yes. Thanks, Tom. So I'd like to first start by providing an overview of the trials we have going on in Stargardt disease. We have 2 studies, as Tom mentioned. We have an ongoing open label Phase 2 study. [ This is a 2-year ] study, which is just about ready to end in October. I'll give you some more information about that as we move forward. But we've got 18-month data to share with you and I'll provide that in a moment. There's also the Phase 3 data, which as Tom mentioned, has recently stopped enrollment. We've met our target. In fact, we've exceeded our target by about 10 subjects. We've got 100 subjects in there. Both of these studies are 2-year studies. They're both looking at the primary endpoint, which is the growth of atrophic lesions that is DDAF, and I'll explain what that is in a moment. So there's a lot of similarities between these designs. The differences are as follows. In the open-label Phase 2, there's only 13 subjects, and these subjects came in with only autofluorescent lesions, and I'll show you some of the biology on how the autofluorescent lesions turn into this atrophic lesion that we call DDAF. So that's one of the differences. The other difference, of course, it's an open-label study. We're looking at the same efficacy measures or the same assessments by imaging modalities, such as fundus fluorescence auto photography to look at the lesion growth, and you can see here at the bottom of the key inclusion criteria were. The Stargardt Phase 3 study is also a 2-year study in design. Of course, it's global. There will be a 2:1 randomization favoring Tinlarebant, and you can see there the various inclusion criteria at the bottom of this slide. Next slide, please. So I want to show you, first, as Tom mentioned, this agent Tinlarebant is a retinol binding protein 4 antagonist. And so the first biomarker you will, if you will, that we will see is a reduction in the retinol binding protein 4 levels in serum. And that's what's shown here from the Phase 2 data out to 18 months. So you see the very first point, which is shown there at 100%, that's before the patients got dosed. And you can see over the period of 18 months, we've achieved about 80% reduction from baseline of retinol binding protein 4. You see here this target threshold was greater than or equal to 70% reduction. This number has been determined in a clinical study in geographic atrophy with a different retinol binding protein 4 antagonists. I'll share that data with you as well. But this has become our market because we believe that you need to achieve at least this level of RBP4 reduction to affect a change in lesion growth. And by the way, the daily oral dose these kids were getting -- these 13 adolescent Stargardt kids is 5 milligrams per day, and no one's left study out to 18 months. I'll go over the safety data as well. Next slide. A little bit about the biology. So early in the disease course, there are only autofluorescent lesions, and that's shown on the left-hand side here, the left image. These lesions are called definitely decreased autofluorescence by ophthalmologists. Basically, what they do, they represent cells laid in with autofluorescent entities. These autofluorescent entities are bisretinoids. These were the agents that we're trying to reduce because these bisretinoids are formed from vitamin A, we've [ reasoned ] that by reducing the amount of retinol going into the eye, we can have effect on reducing the accumulation of these bisretinoids and slow the growth of these autofluorescent lesions. So these autofluorescent lesions are amendable to rescue. But if left alone, which, of course, they have to be because there's no treatment, they will transition into atrophic retinal lesions, which is shown on the right-hand side. You see that black demarcated image that basically is irreversible, photoreceptor cell loss. Those cells are never coming back. That atrophic area is what ophthalmology refer to as definitely decreased autofluorescence and stopping the growth of that lesion type is the primary endpoint. But of course, ophthalmologists look at the combined lesion growth rate because both of these lesions are pathologic. And so, in one study conducted in 2020 by [indiscernible] and co-workers, they found in 53 adolescent Stargardt kids, the growth rate of the combined lesion was roughly about 0.7 millimeter square per year. When we look at that same anatomical feature in our 18-month data and annualize it out to a year, we see a growth of only about 0.28 millimeter square per year. So that represents about a 60% reduction in the combined lesion growth rate based upon comparison to this very well conducted natural history study, which, by the way, at that time was the largest natural history study conducted in adolescent patients. But we are very concerned about comparing the atrophic lesion growth because that is after all the end point. And for that comparison, we had to go to for the largest natural history study of Stargardt's conducted today called ProgStar. This study enrolled hundreds of patients with Stargardt's disease. Many of them were adult patients. But among these patients, there was a small group of 20 subjects that had the exact same baseline characteristics, as our subjects in the open-label Phase 2, that is they were 18 years or younger, and they had no atrophic lesions at baseline only autofluorescence. So we were able to compare the combined lesion growth rate in that ProgStar group to ours, as well as the atrophic lesion growth. The combined lesion growth is shown on the left-hand side.This is called DAF or decreased autofluorescence. So it represents the QDAF area plus the DDAF area. And you can see here out to 18 months, we're getting about a 50% reduction in the combined lesion growth rate. And you remember this slide previously showed you a 60% reduction. So it's a pretty good comparison between these 2 separate and independent natural history studies. When we look at the atrophic lesion growth, as the DDAF, we see at 18 months about 50% -- sorry, 60% reduction in that atrophic lesion growth rate. And noticeably, not many subjects are converting. In fact, there seems to be a slowing of the conversion in our treatment group transitioning from the autofluorescent lesion to the atrophic retinal lesion type, and that is all very consistent with our hypothesis that we would first affect a change on the autofluorescence, and then subsequently a change in the atrophic lesion growth. And we believe that's what these data are showing us. And I should have mentioned, but the investigators [ from both the ] previous study by [indiscernible] and this study, ProgStar, which was Hendrik Scholl, commented that we are seeing a definite bonafide treatment effect in these natural history study comparisons. So that's very promising for us to see. Next slide. This is showing you the visual acuity data. We're showing you both eyes, the study eye and fellow eye. Of course, both eyes are going to get the same treatment because this is an oral systemically applied drug. We're showing you this because in clinical studies, you do have to designate a fellow eye -- sorry, study eye, and then the other eye just becomes a fellow eye. We just want to show you that across 18 months, we're having a stabilization of visual acuity in these subjects, and this is a very promising trend because typically, these subjects lose anywhere from 4 to 6 letters per year. So the fact that we've stabilized over 18 months is a very promising trend. That combined with the slow lesion growth tells us we're affecting exactly what we want to do, stop the lesion growth and eventually have an effect on preserving or improving vision. And you can see there the letters lost is roughly within noise of the variability of the visual acuity assessment. Next slide, please. So now we want to get into the safety data. I should start by saying there have been no systemic toxicities or AEs noted to-date. So no clinically significant findings in relation to vital signs, physical exams, cardiac health or organ functions. What we are seeing are 2 expected features of this therapy, and they're expected because we are reducing the amount of vitamin going into the eye. So we expect effects on rod and cone photoreceptors, which are the 2 photoreceptor cell types in your retina. The first AE we're finding is a form of Chromatopsia called Xanthopsia. This is mediated by cone photoreceptors and it typically happens when patients transition suddenly from a very dark light to a very bright light, or for instance, from waking after sleeping and being exposed to very high room light or sunlight. And so basically, cone photoreceptors are activated, they will demand chromophore. Under our treatment regimen that chromophore doesn't get there quite as quickly, so there will be a delay in the timing for these cone photoreceptors to fill up with chromophore. And during that time, they will misfire and produce these artificial electrical mediated hues of color in the visual field. In this case, Xanthopsia is yellow. But you can see here, the majority of subjects are experiencing Xanthopsia, but no one's leaving study because of it, and in fact, we are seeing some recovery over time. And we're not taking subjects off drug. They are recovering while still getting dosed. The second ocular AE is known as delayed dark adaptation. This is mediated by rod photoreceptors. And again, when rod photoreceptors, when you transition suddenly from a very bright light to a very dim light, rod photoreceptors activate. They require chromophore. There will be a delay in the timing of that chromophore to fill up the rod photoreceptors. And during that time, these rod photoreceptors will not have maximum dim light sensitivity. So there was a delay in the accommodation to dim light. This is not night blindness. I want to make that very clear. This is simply a delay sometimes 8 to 12 minutes, in cases where it's very severe out to 20 minutes in this one subject, it's called night vision impairment. But overall, we're very satisfied with these findings. We basically lost 1 subject to follow-up at 12 months. So out of 13 subjects, we are now at 12 subjects at 18 months, but this is still a very, very promising safety profile. Next slide, please. So now I want to talk about that proof-of-concept study I told you about the 70% marker. How did we get there? Well, this was a study I conducted approximately 12 years, 13 years ago when I was with another company. I always had this idea that reducing retinol delivery [ to eye ] might have an effect on slowing lesion growth. I didn't have a drug to do that with, but I did find an anti-cancer drug called fenretinide, which had a side effect of reducing retinol binding protein 4 in the blood. As I said before, it was developed as an anticancer drug. But in all the cancer studies, what investigators noted was a dose-dependent reduction of RBP4. So I repurposed fenretinide into a 2-year Phase 2 proof-of-concept study enrolling 246 GA patients to see if this drug would have any effect on slowing lesion growth. There were 2 treatment arms and placebo, 100-milligram, [ 300-milligram ], and of course, placebo. I want to show you the lesion growth data just from the high-dose arm and placebo because the middle dose [ 100 milligram ] had absolutely no effect on lesion growth. What you're seeing here on this histogram shown on the left-hand side in the black bars is the lesion growth in the placebo group expressed as a percent increase from baseline. So we're getting about a 50% increase over 24 months in the placebo subjects. In the 300-milligram group, there was something very interesting. There was a group of subjects, who had a very profound reduction of retinol binding protein 4 of at least 70% or more. In those subjects, there was about a 25% slowing of lesion growth over 2 years. In the subjects that did not have this reduction of retinol binding protein 4 of 70% or more, there was absolutely no effect on the lesion growth rate. So we're pretty convinced, especially in GA, that this is the level of reduction that would be required to affect a change in lesion. And of course, this is the same sort of approach that we're applying to Stargardt's disease. An interesting thing about this lesion growth reduction, you'll notice, it started right at about the 12-month time point, and it stabilizes between 18 months and 24 months. But when we look at the visual acuity loss in these subjects, we also noticed in these subjects, they had a preservation of lesion growth that is a reduction of lesion growth, there was also a stabilization of visual acuity loss right at about the same time, 12 months, there was a 6 letter loss, and there was no further loss out to 24 months. Meanwhile, the placebo group and the patients or the subjects that did not get that profound reduction of RBP4 continue to lose vision out to about 11 or 13 letters over the 2 years. So we have a very significant visual acuity game, and a very significant lesion reduction that has never been observed before in a GA study. The problem with this Phase 2 study was that only 1 in 3 subjects actually achieved this profound reduction of RBP4 in the 300-milligram group. And the reasons for that are twofold. One, fenretinide has terrible bioavailability. So we ask subjects to take this drug with a high fat [ meal and dinner ] to increase exposure into the blood. Many patients complied out to about 1 year, but after 1 year, we had a lot of patients following off of that compliance. And we knew that because the RBP levels in these patients would inflect upward, indicating in fact that they're no longer will be having suppression of RBP4. The second problem was the low potency of fenretinide. Fenretinide is a terribly -- terrible drug for [indiscernible] RBP4 antagonist. It presents the same affinity for the target as does a native ligand vitamin A. With Tinlarebant, we have designed a drug that specifically overcomes those deficits of fenretinide. So it has greater bioavailability and a 100-fold greater potency than does fenretinide. So we're convinced with this better purpose-designed RBP4 antagonist, we can achieve at least this benefit and probably even greater because, again, we'll have better compliance, and we'll have greater potency of the drug on target. Next slide, please. So now a little bit of our Phase 3 study in geographic atrophy. This is important to note. So we were concerned that with the higher age and higher BMI of patients have GA versus Stargardt's disease, we would have to do a dose higher than 5 milligram. So we did a PK/PD study with both 5 milligram and 10 milligram. And what we found was that 5-milligram dose produces the same pharmacodynamic profile, as it did in younger subjects. So in these healthy adults, we're seeing about a 80% reduction of RBP4 across the dosing period with this 5 milligram dose. And it's also important to note, and we see this in the adolescent subjects as well, once you withdraw the treatment, the RBP4 level start bounding back upward, showing a nice reversibility of the pharmacodynamic effect, which, of course, is a nice safety feature in the event of any untoward AE or you want to return the patient back to baseline status. Sorry about that. Now a little bit about the clinical design overview for our Phase 3 study we call PHOENIX. This study design is going to be nearly identical to the Phase 3 trial design for Stargardt's, that is it's 2 years in duration. It has the same randomization frequency 2 to 1 favoring Tinlarebant. It has the same endpoint measures. So we're still looking at the same DDAF measure, as a primary measure for efficacy. And of course, we're looking at other measures such as BCVA, and looking at the autofluorescence. There are 2 major differences. One, of course, is the indication, geographic atrophy not Stargardt's; and the second one is that we'll be enrolling up to 430 subjects instead of the 90 that we targeted for the Stargardt's disease study. This, of course, reflects the higher prevalence of GA in the population. But otherwise, these studies are essentially identical. And I think Tom mentioned that we've actually kicked off this study. We've enrolled our first patient, I believe it was last week, and we continue to get more interest and more patients rolling into this Phase 3 study, as we move forward. With that, I believe, I can turn it back to Hao-Yuan, so he can discuss the 2023 Q2 financial results. Thank you.

Thank you, Nathan. So as of June 30, 2023, we have $57.4 million in cash. And for the R&D expenses for the 3 months ended June 30, we had research and development expenses that was about $5.5 million compared to $1.6 million for the same period last year. The increase was mainly due to the expense on the PHOENIX trial and also the increase on the wage and salary due to our R&D team expansion. For the G&A expenses, again, in Q2, we had G&A expenses of $1.4 million compared to $0.9 million for the same period last year. And the increase is due to the increase in professional service fee, and also the wage and salaries. The net loss was $6.8 million this quarter compared to $2.4 million last year for the same quarter. And about the key milestone, as Tom mentioned earlier, so we initiated the study [ STGD1 ], and we just got the first patient in this quarter. And we also fully complete enrollment for the DRAGON study with 100 subjects so far. And we expect to have the 24-month data by Q4 this year. And also, we expect to have the interim result from the Phase 3 DRAGON study in Stargardt disease by mid next year. With that, I'll turn it back to Sarah.

[Operator Instructions] So with that, I'm going to open it up for questions from our first analyst, which is Basma Radwan from Leerink.

This is Basma on for Marc Goodman. We have few questions on the upcoming final readout of the Phase 2 trial in Stargardt disease. The first question is really, what should we expect in terms of the efficacy at month 24, more specifically, I'm talking about the reduction of the DDAF lesion growth rate. Should we expect a similar level of reduction in the lesion growth rate to the level demonstrated at month 18, which is the 50% reduction when you compare it to the next month rose from the ProgStar study? The second question is about the conversion from the QDAF lesion to the DDAF lesion. So at month 18, 5 out of 12 patients on Tinlarebant had DDAF lesions versus 9 out of 20 patients in the ProgStar study. What changes to this proportion should we expect at month 24? And the final question is about any updates you have about dropouts in this study. Do we still expect 12 patients for the time point at month 24? And I do have a follow-up on questions for the Phase 3 DRAGON study, if you don't mind?

Sure. Nathan, you want to take this?

Yes, I'd be happy to. If Hao-Yuan can go exactly -- yes, exactly go here. So the first question related to what we expect to see at 24 months. If you follow these trajectories, particularly for the DDAF, basically, you're going to see the same thing, basically [ have another peak level ]. The lines will continue to sort of track the same way. So we'll get at least a 50% reduction. We already know what the 24-month data looks like in ProgStar. So it's going to go a little bit higher than where it is now. And of course, our DDAF will inflect upwards a little bit as well. But this study is going to end in October. So basically 2.5 months from now. I don't expect that there's going to be any significant change from these trajectories over the next 2 months to 3 months. So I think what you're seeing here is a very good snapshot of what you can expect to see at 24 months. And in terms of the numbers of subjects, I don't expect we'll lose any additional subjects. Again, we lost 1 out of 13 when we first started due to a loss to follow-up at 12 months. No one has left because of any safety or AE concern. So I don't have any real concern about that. Your other question was related to the conversion from autofluorescent lesion to the atrophic retinal lesion. And so, I expect that probably by 18 months, sorry, by 24 months, we should see at least 2 more subjects convert. Again, that's based upon the sort of run rate that we're going. But there will be a significant percentage difference, numerical difference in the percentage of subjects in ProgStar that converted versus the number of subjects in our study that's converted. And our study will show a lower number, which again is consistent with our hypothesis, in our MOA that reducing the autofluorescence will then slow the transitioning of the autofluorescent lesion to the atrophic retinal lesion. So I think I addressed all 3 points, Basma, but please let me know if I missed anything.

That was very helpful. The one question we have about the Phase 3 DRAGON study, it's about the inclusion criteria. So you do specify in the inclusion criteria, lesion size to be within 3 disc areas. Could you provide more color on the rationale behind this inclusion criteria and [ really ]?

Yes. So this goes to our approach for early intervention. So I've done a number of studies in the Stargardt's disease and geographic atrophy, and one thing I consistently seen, and by the way, all of the studies I've done in these diseases have been with oral therapeutics and either visual cycle modulators or RBP4 antagonist to sort of mediate the effect. And so what I've seen consistently in these studies is that lesions that are smaller at baseline tend to respond better to these types of therapeutic approaches. And that's even been shown in natural history studies, where you look at growth rate of lesions that are small versus large, you do see -- tend to see faster lesion growth rates in lesions that are smaller than they tend to sort of slow down, as they get large. And this term on using large and small, of course, is ambiguous. But when I say small lesions, I'm talking about lesions that are less than, for instance, 5 millimeter square and certainly nothing bigger than 10-millimeter square. So anything bigger than 10 millimeter square is what I consider too large. And in fact, that's where inflammation starts kicking in. So it's important to know sort of the chronology of the pathology. Early in the disease course, there's very little inflammation. So when these early lesions, Hao-Yuan, could you go to the lesion comparison, the [ QDF ] (sic) [ QDAF ] and DDAF, yes, sure. So these [ QDF ] (sic) [ QDAF ] lesions that you're seeing here are really the first lesions that are actually going to convert, right, and turn into the atrophic retinal lesion. But once the lesion, if you look at the left-hand -- right-hand side [ right now ], once that atrophic lesion gets too large, there's nothing you can do to slow it down. So the reason that we're specifying less than 3 disc areas is again, based upon all the prior clinical studies that I've done and natural history studies in both GA and Stargardt to show that smaller lesions respond better to treatment. And a sort of a real-world evidence for that is data from the Emixustat study. This is a study conducted by Kubota pharmaceuticals or maybe even called Kubota Vision, formerly Acucela Pharmaceuticals is what it was. And they were advancing a drug called Emixustat, which is an RPE65 inhibitor intended to do the same thing that we're doing, which is reduce the bisretinoids, and in fact, it worked very well in animal models, but it's a very aggressive approach because it hits an enzyme on the visual cycle, that's a rate-limiting enzyme. So anyway, they ran a Phase 3 study with 194 Stargardt subjects, and they didn't reach their endpoint at 2 years. But in a post hoc analysis, what they found was that patients, who came in with smaller lesions at baseline had as much as a 40% slowing of lesion growth. So that's a very, very important note for us, particularly in Stargardt's disease because that's exactly what we're doing, is we're recruiting subject with smaller lesions at baseline. We believe all this clinical evidence and scientific evidence tells us we're doing the right thing for these kids. So again, early intervention is the best way to stop these emerging retinal lesions that could -- will eventually affect vision.

The next question comes from Jennifer Kim at Cantor.

Congrats on the execution this quarter. Maybe to start off with DRAGON. I believe the original announcement for enrollment completion highlighted 90 adolescent patients, and here it says 100 subjects. I was wondering maybe you could provide any color around that difference?

Sure.

Yes. I'd like to do that, if I could. I discussed [Technical Difficulty] I'm sure that was going to come up. But it's important to note that when we stopped the enrollment at sites, there were a number of subjects in the screening queue, we can't just turn those subjects away. So basically, although we stopped accepting new patients, the patients that were in the queue, which amounted to roughly, I think it was about 20 patients in the queue went through screening and of those subjects, approximately 10 qualified for studies. So that's why we went from our target of 90 to approximately 100 subjects to-date.

Okay. Wonderful. And then a follow-up for DRAGON, prior to the interim data next year, are you considering at all disclosing like the baseline characteristics for these patients?

Tom?

Yes. So it's -- right now, we do want to present that data, but it's a discussion with the DSMB and with FDA. So right now, we are still in discussion, but I think this is a discussion when it comes closer to the date.

Okay. Great. And then maybe just as we're thinking about the October 24-month data, are you thinking of a venue for presentation?

Yes. In fact, we are looking forward to presenting this data at AAO this year, the 24-month data readout, which coincides with the AAO conference by end of the year. I think that's in early November.

Okay. And then maybe one, just [indiscernible] more of a broad question. With safety becoming even more of a focus given the concerns with Syfovre, I'm just wondering how does that play into your thinking around the opportunity for an oral once-daily?

Yes. Good question. So we are talking about a quite elderly population in AMD with GA. So I guess, again, oral treatment, a non-invasive treatment is always much more attractive. And those that are concerned with safety and was not given the recent news of Apellis. I think in this elderly population, I think invasive intravitreal injection is always going to be an issue for patients that either want to take the treatment, and I would say, quite some majority of those elderly patients would not want anything needle in the eye. Nathan, do you want to add more color to this?

No. I think there's clearly a treatment burden for the patient with an injectable therapeutic. And that the fact is that there's not going to be a clinically meaningful benefit derived by the patient for at least 2 years, perhaps longer. The same could be said with an oral therapeutic, but there's less of a treatment burden for the patient. So I think when offered the option between [ oral body ] and injectable, obviously, the patient is going to choose the oral. So there will be a greater uptake for an oral therapeutic. And with respect to AEs, such as the [ retinal vasculitis ] that have been observed or vasculitis, as they call it, in the eye. This is even though a rare finding, these are the risk factors associated with injections in the eye, and there will be others. So there's inflammation. There's all types of things that can happen when you puncture an eyeball with a needle and you have to do it repeatedly every other month or every third month, whatever it is, this is a very aggressive invasive treatment therapy. So we believe once an oral therapeutic is approved, I think patients will flock to it and that will certainly detract from the uptake of either Apellis' drug or the Astellas/Iveric drug they just recently got approved. So we're not too concerned about the injectable therapeutics. And I do want to emphasize, as I said before, those therapeutics address late-stage disease because actually, what they're doing is, they're killing an inflammatory response that's driving the disease process. Early in the disease course, there's no inflammation, either in Stargardt's disease or in GA, you just have these incipient biomolecules or factors that are causing sort of retinal pathology, but there's no inflammation here. In fact, inflammation will kicks in later. So those therapeutics, those injectables would not be effective in our patient population. Conversely, our therapeutic would be expected to be beneficial in that later stage sort of as a maintenance therapy for patients, who are sort of getting treatment and they need to sort of keep the geographic atrophy at bay. So we think there's synergy rather than competition, and we're certainly not concerned about any safety concerns that they have because we don't think they're going to affect what we have in terms of an oral therapeutic, which to-date has shown to be very safe and well tolerated in these adolescent Stargardt subjects.

The next question comes from Yi Chen at H. C. Wainwright.

You just talked about the AEs associated with Apellis drug. So just to clarify, you believe that the retinal vasculitis is associated with injection, but not the drug itself in terms of the mechanism of action or complement inhibitors?

It could be a combination of both, Yi. I was just saying that in other studies, for instance, when they first started developing the first anti-VEGF drugs and they had -- I believe that time it was Macugen. They had things like this as well, and it wasn't necessarily attributed to the drug. It was attributed to the procedure. But yes, it is possible that the drug in itself could cause that, but I think that's a rare possibility because it didn't occur, I don't think in their Phase 3 study. So again, once you start getting real world evidence for how this drug is going to be applied, you'll start unearthing some of these potential risks, and these are not manageable risks. These are very serious concerns, where patients are losing vision even though it's a small number of patients. So yes, it could be mediated by the drug, but I believe more about the actual intervention itself. That's just my personal belief.

So Yi, I believe the one of the causes of retinal vasculitis is infection and inflammation and neovascularization. So all that can be associated with intravitreal injections, or invasive treatment that causes that.

So it is reasonable to expect the newly approved Izervay may have those AE as well when it's commercialized, right?

Very possible. I predict yes.

It's possible.

The next question comes from Bruce Jackson at Benchmark.

You mentioned the increase in the study size for the DRAGON trial. Originally, you put it up to 90 patients in order to improve the probability of getting an efficacy signal now that you're at 100, has that increased your confidence that we're going to get an efficacy signal on the results?

Nathan, do you want to take that.

[ I want to take that ]. Yes, yes. So it's not really about -- we always believe that we would be getting an efficacy signal since the 6-month data, right? So in the Stargardt open-label Phase 2 study, we have seen positive data at every 6-month interim analysis now out to 18 months. So it's not so much about changing the treatment effect size. It's about increasing the power that is the confidence and probability that, that effect will be durable through 2 years. So yes, that additional 10 subjects does give us an additional buffer for that power. It doesn't necessarily mean we're going to get greater treatment effect or greater statistical significance at the end of study, but we'll have greater power to say that, that is a true bonafide robust result because, again, there's more patients showing that treatment effect. So it's more about the power than it is rather the treatment effect size.

So Bruce, to add on that. So we run the simulations and all that. So with the added sample size onto the study and now given that we've gone up to 100, I think that gives us a better chance of getting a positive OA statistical significance at interim. So that gives us a better chance of reaching there.

Okay. Great. And then one finance question, I thought I'd throw one in. In terms of your cash balance and your burn rate, how many quarters of cash do you have right now?

Yes. Thank you, Bruce. So we do expect that we have cash [ around ] till end of 2025 with the current rate.

Okay. This concludes the verbal portion of the Q&A session. Hao-Yuan, do we have any questions for the webcast? Or should we hand it back to Tom for concluding remarks.

No. I don't have got any questions here. We can turn it back to Tom. Thank you.

Okay. Well, thanks, everyone, for joining this call, and we look forward to updating everyone on our end of Phase 2 results shortly. Thank you very much.