Ladies and gentlemen, thank you for joining us, and welcome to Belite Bio's phase III DRAGON topline results conference call. After today's prepared remarks, we will host a question-and-answer session. If you would like to ask a question, please raise your hand. If you have dialed into today's call, please press star nine to raise your hand and star six to unmute. I will now hand the conference over to Nathan Mata. Please go ahead. Welcome, everyone, to the webcast presentation of top-line data from our two-year phase III trial of Tinlarebant in Stargardt disease. This pivotal trial is known as the DRAGON Study and is the largest interventional trial ever conducted in adolescent patients with Stargardt disease. I'm Nathan Mata, the Chief Scientific Officer at Belite Bio. Before we begin, let me point out that there will be forward-looking statements on today's call that are based on our current expectations and beliefs. These statements are subject to certain risks and uncertainties, and actual results may differ materially. We encourage you to consult the risk factors discussed in our SEC filings for additional details. Joining me on today's call from Belite Bio are Dr. Tom Lin, Chairman and CEO, and Dr. Hendrik Scholl, Chief Medical Officer. Dr. Scholl was the coordinating principal investigator of the ProgStar Study and, at the time, was professor and endowed chair at the Wilmer Eye Institute at Johns Hopkins University. I'm also pleased to welcome three retinal disease experts who were involved in the DRAGON Study: Dr. Leopold Schmetterer, Head of Ocular Imaging and Scientific Director at Singapore Eye Research Institute and Professor at the Nanyang Technological University. Dr. Leopold Schmetterer is also the DSMB Chair of the DRAGON Study. Next, Dr. Michelle Michaelides, Professor of Ophthalmology at UCL Institute of Ophthalmology and Consultant Ophthalmologist at Moorfields Eye Hospital, and Dr. Helen Sui, Professor and Director of the Department of Ophthalmology at Peking Union Medical College Hospital and Vice President of the Chinese Medical Doctor Association Ophthalmology Committee and Counselor of the Chinese Ophthalmology Genetics Alliance. Incidentally, Dr. Michaelides and Dr. Sui were the two highest enrollers in the DRAGON Study. As the lead developer of the therapeutic approach that you will hear about today, I'm extremely pleased that we were able to report positive results from the top-line DRAGON Study data. Tinlarebant achieves statistically significant and clinically meaningful outcomes, demonstrating its potential to substantially slow disease progression, which is aligned with the therapeutic benefit we observed in earlier studies. This is an exciting time for Belite Bio and personally rewarding to me as we get one step closer to being able to offer a first-ever treatment to people living with this debilitating disease. As Dr. Scholl will soon explain, Stargardt disease is due to genetic mutations which cause an aberrant and excessive accumulation of toxic vitamin A byproducts within the retina, leading to the formation of retinal lesions and progressive loss of vision. Because these toxic byproducts are derived from circulating vitamin A, which enters the visual cycle, our approach is to reduce vitamin A delivery specifically to the eye by targeting the sole carrier protein which facilitates this delivery: retinal binding protein 4. Tinlarebant is an orally administered small molecule once-a-day tablet that targets RBP4 as a means to reduce vitamin A delivery to the eye and slow progression of pathology in Stargardt disease. With that, I'll turn the meeting over to Dr. Scholl to present details of the clinical trial and results, and after his remarks, I will lead a panel discussion with our experts and then open the call up to Q&A. Thank you. Thank you, Nathan. The DRAGON Trial investigated Tinlarebant for the treatment of Stargardt disease. Stargardt disease, and/or ABCA4-associated retinal dystrophy, is a rare and progressive eye disease leading to legal blindness in almost all cases. There is no approved treatment for Stargardt disease. The pathophysiology of Stargardt disease is well understood. The disease is caused by autosomal recessive mutations in the ABCA4 gene, impairing retinoid transport and leading to toxic bisretinoid buildup in the retinal pigment epithelium, or RPE. Bisretinoid accumulation manifests as yellowish flecks underneath the retina, which can be seen clinically. Bisretinoid accumulation causes oxidative stress, toxicity, and eventually cell death of the RPE. Degeneration of the retinal pigment epithelium invariably leads to secondary degeneration of macular photoreceptors. This, in turn, leads to progressive macular atrophy and retinal degeneration. The key symptoms include gradual central vision loss in both eyes, often starting in childhood or early adulthood, blurry or distorted central vision, central blind spots, so-called scotomas, photophobia and delayed dark adaptation, impaired color vision, and difficulty with detailed tasks like reading or recognizing faces. The peripheral vision remains preserved in the majority of cases. The impact on activities of daily living includes challenges with reading, using screens, and recognizing faces, harming education and work. Typically, the inability to drive, reducing mobility and independence, restrictions in sports, leisure, and simple tasks like avoiding trips, social struggles including interactions, relationships, and events, emotional effects like frustration, worry, anger, and depression, and career changes or job loss affecting finances and productivity. Although Stargardt disease, by definition, is a rare disease affecting fewer than 200,000 people in the U.S., it is not so rare and belongs to the most prevalent inherited diseases in medicine. In a recent review article led by our Chief Scientific Officer, Dr. Nathan Mata, based on Moore's recent genetic database and prevalence data, it was reported that in the U.S., there are about 53,000 people affected by Stargardt disease and more than 100,000 in both Europe and China. In standard clinical care, the central retina in Stargardt disease is imaged with fundus photography, optical coherence tomography, or OCT, and fundus autofluorescence imaging. This slide shows a video derived from a sequence of autofluorescence images over a period of four years in a patient affected by Stargardt disease. He was a patient of mine at the Wilmer Eye Institute at Johns Hopkins in the years 2010 to 2016 and also participated in the ProgStar Study. A normal central retina would show a relatively even grayish appearance, with the exception of the optic nerve head, which is the spot in the retina where nerve fibers exit the eye to form the optic nerve and appears as a round black structure on the right side of this image. What we can see in the image is a central area of abnormal autofluorescence. There are some whitish spots with increased signal that is caused by bisretinoid accumulation, and there is a round area of strongly decreased signal where the retinal pigment epithelium degenerated. In such areas, called definitely decreased autofluorescence, or DDAF, the patient can't see anymore. There was a continuous growth of this area over time, and the patient lost visual acuity from 20/40 to 20/80 over those four years. Reading ability for newspaper print is typically lost at a visual acuity of 20/50 or lower. I would like to briefly present the study design of the DRAGON Study. The DRAGON Study was a phase III multicenter randomized double-masked placebo-controlled study to evaluate the safety and efficacy of Tinlarebant in the treatment of Stargardt disease in adolescent subjects. The key inclusion criteria included the clinical diagnosis of Stargardt disease, an age between 12 to 20 years old, at least one mutation identified in the ABCA4 gene, an atrophic lesion, DDAF within three disc areas, and a best-corrected visual acuity of 20/200 or better. Subjects were randomized to either oral Tinlarebant 5 mg per day or placebo with a two-to-one ratio of randomization. The primary efficacy endpoint was the annualized rate of lesion growth in the aggregate area of definitely decreased autofluorescence, or DDAF, from baseline as assessed by fundus autofluorescence imaging at month 25. The key secondary endpoint was the annualized rate of lesion growth in the total area of DAF, decreased autofluorescence. Other secondary measures included visual acuity, photoreceptor morphology assessed by OCT, and pharmacodynamics, namely the change in retinal binding protein 4, or RBP4, levels from baseline to the end of study. Exploratory endpoints included change in retinal sensitivity by microperimetry, amongst others. Baseline DDAF lesion size was fairly balanced between the Tinlarebant and the placebo group in the study eye. As expected in an inherited condition, Stargardt disease affects both eyes, and the clinical appearance in the two eyes is typically similar, however not identical. That the lesion size is exactly the same would rather be uncommon, and the fellow eye may as well show no lesion at all or a very large lesion. Still, the fellow eye mean lesion size was relatively similar to the study eye. The trial design does not allow to achieve perfect balance of the fellow eye between treatment and placebo group, but the DRAGON Trial baseline characteristics still showed a quite similar mean DDAF area in the fellow eye as well. Before we get to the efficacy endpoint results, I would explain the pharmacodynamic effects and the primary endpoint of the DRAGON Trial. This slide shows the results for the biomarker in the systemic circulation, namely percent change from baseline in retinal binding protein 4, or RBP4. The goal of the pharmacodynamic effect was to reach at least 70% reduction in RBP4. The graph shows that daily dosing of 5 mg per day with Tinlarebant led to a sustained 80% reduction of RBP4 with very little variability, and RBP4 levels returned close to the original baseline value by 28 days after treatment when Tinlarebant was discontinued. The ProgStar Study was a large international natural history study of ABCA4-associated Stargardt disease designed to characterize disease progression and establish reliable clinical trial endpoints. Using standardized fundus autofluorescence imaging, ProgStar demonstrated that DDAF, definitely decreased autofluorescence, represents well-demarcated areas of complete RPE loss, and enlarges at a predictable rate, making it a robust primary structural endpoint. The study also defined QDAF, questionably decreased autofluorescence, as regions of partial autofluorescence reduction that eventually evolve into DDAF and thus provides an earlier marker of progression. By combining DDAF and QDAF, namely DAF total decreased autofluorescence, captures the full extent of RPE impairment and offers a sensitive, comprehensive metric across a wider disease spectrum. The ProgStar Study confirmed that these autofluorescence-based measures are highly reproducible, correlate with functional decline, and are suitable for detecting progression over 12 to 24 months. As a result, DDAF, QDAF, and total DAF have become a standard structural outcome measure for interventional trials in Stargardt disease. In ProgStar report number 17, it was shown that the mean DDAF growth rate in the ProgStar cohort over 24 months was 0.74 sq mm per year. DDAF growth as an outcome measure has been accepted by the FDA as an approvable endpoint in treatment trials for Stargardt and was the primary endpoint in the DRAGON Study. Consequently, the primary endpoint in the DRAGON Study was the annualized rate of lesion growth in the aggregate area of DDAF from baseline as assessed by fundus autofluorescence imaging at month 25. Data are shown for the Modified Full Analysis Set, MFAS, which consists of all subjects who were randomly assigned to receive study drug and have received at least one dose of study medication. In addition, the MFAS subjects must have a defined DDAF lesion, meeting the eligibility criteria at baseline, and have at least one post-baseline assessment. Data analysis used a mixed model for repeated measures, MMRM, measuring change from baseline in DDAF in the study eye and including terms for treatment, visit, treatment-visit interaction, baseline focality of lesions, and baseline DDAF lesion size. The Statistical Analysis Plan, SAP, specified an unstructured covariance matrix for the MMRM. The CRO also performed a post-hoc analysis using a first-order autoregressive covariance matrix to account for the longitudinal nature of the data while maintaining model stability in a relatively small sample, such as in the DRAGON Study. This approach was discussed and supported by the DSMB when reviewing the final efficacy and safety data. This slide shows the primary endpoint of the DRAGON Study, namely the change from baseline in DDAF total area in the study eye. Applying an unstructured covariance matrix, the treatment effect size was 35.7% compared to placebo and yielded a p-value of 0.0033. With the first-order autoregressive covariance matrix, the treatment effect size was essentially the same at 35.4%, with p smaller than 0.0001. DDAF lesion growth was slowed down to 0.38 sq mm per year, while a progression of 0.59 sq mm per year was measured in the placebo group. To put this number into perspective, in the ProgStar Study, as mentioned earlier, the mean DDAF lesion growth was 0.74 sq mm per year. As mentioned earlier, Stargardt disease affects both eyes. Because Tinlarebant is an oral therapy, it is expected to benefit both eyes, and therefore the fellow eye was also evaluated. Because of phenotypic disparity between the two eyes, the trial design could not account for all variability in the fellow eye, and thus a treatment effect could not necessarily be expected. Notably, statistically significant treatment effect was observed in the fellow eye. Tinlarebant reduced DDAF lesion growth by 33.6% compared to placebo. Decreased autofluorescence DAF is the composite of all areas with reduced autofluorescence, representing the full extent of retinal pigment epithelium impairment. The ProgStar Study established that DAF growth rate is one of the most sensitive and comprehensive structural progression measures. It correlates with functional outcomes such as microperimetry, sensitivity loss, and future visual acuity decline, and it is highly useful in early and intermediate stages where pure DDAF may be small. DAF was the key secondary endpoint in the DRAGON Study. It was found that Tinlarebant slowed DAF lesion growth by 33.7% compared to placebo. The effect also reached statistical significance. Tinlarebant slowed DAF lesion growth also in the fellow eye by 32.7% compared to placebo, with a p-value of 0.017. Not only were the findings for the primary endpoint supported by the findings of the key secondary endpoint, but the significant findings in the fellow eye must be considered confirmatory evidence. Best-corrected visual acuity in the study eye did not show any significant change over the course of two years, neither for the Tinlarebant nor the placebo group. This is not surprising and is consistent with what we expected. Mean visual acuity at baseline in the Tinlarebant group was 39.9 ETDRS letters at baseline and was 39.7 letters at the end of study visit. Similarly, visual acuity was 39.4 letters for the placebo group at baseline and exactly 40 letters at the end of study visit. An ETDRS score of 39-43 letters corresponds to 20 over 160 Snellen visual acuity. The test-retest variability for ETDRS change scores in Stargardt disease yields a repeatability coefficient of about eight letters, as shown by Parker and colleagues. Observed changes of five to ten letters are frequently indistinguishable from measurement noise. Such minor changes in average visual acuity over two years is in line with the natural history of Stargardt disease and was observed in the ProgStar Study. In the prospective cohort of 434 Stargardt patients in the ProgStar Study, the overall rate of best-corrected visual acuity loss was 0.55 letters per year over the duration of two years. For eyes with baseline best-corrected visual acuity between 20 over 70 and 20 to 100, which is shown in the lower left spaghetti plot and exactly matches the DRAGON cohort, visual acuity declined at a rate of 0.6 letters per year. Now we get to the safety results. Importantly, Tinlarebant maintained an excellent safety and tolerability profile over the two-year course of treatment. The table shows systemic safety and tolerability, namely the number of subjects who experienced at least one non-ocular treatment emergent adverse event. A total of six serious adverse events was reported in the study. All events were non-ocular, with four assessed as unrelated and two assessed as unlikely related to the study treatment. The most reported non-ocular adverse events were nasopharyngitis, headache, and acne. Most events were mild and resolved during the study period. Regarding ocular safety and tolerability, xanthopsia and delayed dark adaptation were the most common drug-related ophthalmic adverse events. The majority of xanthopsia, delayed dark adaptation, and night vision impairment were mild, most resolved while on treatment. There were no serious ocular treatment emergent adverse events. Four TAEs led to study drug discontinuation, and two TAEs led to study discontinuation. In summary, the DRAGON Study met its primary endpoint. A highly statistically significant slowing in DDAF lesion growth was observed in subjects treated with 5 mg per day oral Tinlarebant as compared to placebo. The treatment effect was 36% and must be considered clinically meaningful. The observed treatment effect was supported by the fellow eye data and the key secondary endpoint, a reduction of DAF area growth. The change in best-corrected visual acuity both in the treatment and the placebo group was minimal and is in line with natural history data. The biomarker of Tinlarebant treatment, RBP4 reduction, showed a sustained 80% reduction with very little variability. Tinlarebant at 5 mg per day was safe and well tolerated in adolescent Stargardt patients. In conclusion, this is the first ever oral therapy in a retinal degenerative disease to demonstrate a clinically meaningful slowdown of neurodegeneration. 36% reduction in DDAF lesion growth rate represents a robust and reproducible treatment effect in Stargardt disease. There was an excellent safety and tolerability profile observed across the two years of treatment. The therapy addresses the root pathogenic mechanism, namely bisretinoid accumulation, offering a rational disease-modifying approach where no approved therapies previously existed. There is broad applicability across disease stages, from early ABCA4-mediated changes to more advanced atrophy. On a personal note, I've seen patients with Stargardt disease throughout my career for more than two decades in Germany, the U.K., the U.S., and now in Switzerland, and I can say with confidence and enthusiasm that this represents a true game changer. I would confidently offer this therapy to all of my Stargardt patients. Thank you very much. I hand over now to the KOL panel discussion. Thank you, Hendrik. I'll go ahead and lead the discussion. I'd like to open up by addressing Dr. Michaelides with a question. For those who are not familiar with Stargardt disease, what is this disease and why is slowing of the lesion growth meaningful for patients? Thank you, Nathan. Stargardt disease is an inherited retinal condition that leads to gradual progressive loss of central vision, often starting in childhood or adolescence. Patients may still see reasonably well in the early stages, but underlying damage in the retina, and specifically in the macula, continues to worsen over time. Slowing that damage means delaying the increasing difficulty patients experience with reading, recognizing faces, and seeing at school or work. Even if the vision chart number does not change over one or two years, slowing the disease gives patients many more years of better vision. Thank you, Dr. Michaelides. Appreciate that. I'll now turn to Dr. Sui for another question regarding lesion growth and vision. The study showed a 36% slowing of lesion progression. Can you explain what this means in terms of preservation of vision for patients over time? Yeah. A 36% treatment effect means that, on average, the area of retinal damage grows about 1/3 more slowly in treated patients than it grows on placebo. Loss of vision in Stargardt 1 patients is due to lesion progression into the macula. Slowing lesion progression by 36% will likely lead to a significantly longer period of preserved vision. Thank you. Now to Dr. Scholl. I was wondering, during the presentation, Dr. Scholl mentioned that the CRO performed a post-hoc analysis using a first-order autoregressive covariance structure for the analysis of the primary endpoint. Can you explain why this is a valuable addition to the original statistical analysis plan? Thank you, Nathan, for this question. The mixed model for repeated measures with the first autoregressive structure is considered the most suitable method because it can reliably track changes in patients over time, even with a relatively small sample size, such as in the DRAGON Study. It also reflects the idea that measurements taken closer together tend to be more similar than those taken farther apart, which matches the gradual time-dependent progression typically seen in Stargardt disease. Dr. Schmetterer. As As a follow-up question to the previous one, the p-value for the primary endpoint using the unstructured covariance matrix was very low at 0.0033 and was even lower at 0.0001 using a first-order autoregressive covariance structure. Could you comment on the significance of these values? The very low p-value indicates that it is statistically highly unlikely that the observed 36% difference occurred by chance, given the study design. It reflects not only the magnitude of the effect, but also how consistently it appeared across patients. Such a strong p-value, together with the clean and reliable conduct of the trial, substantially strengthens confidence that the treatment produced a real effect. Thank you very much. Now back to Dr. Michaelides regarding visual acuity again. Visual acuity was relatively unchanged during the trial. I'm wondering, why are these structural results still considered clinically meaningful? In young people with Stargardt disease, visual acuity can remain relatively stable in the short term, even while the retina is undergoing measurable, continuous, irreversible damage. Autofluorescence imaging allows us to see that damage earlier and more precisely than using a standard vision chart, meaning that there is loss of retinal structure before consequent loss of vision. If we wait for vision to drop before acting, we've already lost retinal tissue that we can't get back. Preserving retinal structure preserves vision. It's also important to note that the key secondary endpoint, growth of decreased autofluorescence, or DAF, was also slowed and showed statistical significance. While DDAF reflects more or less complete loss of cells, DAF reflects earlier stages. Slowing down of DAF therefore means that the development of terminally sick cells could be prevented or slowed. I see. Thank you very much for your answer. Going back to Dr. Sui, I was wondering if you could explain to us how does slowing lesion expansion potentially translate into preserving daily vision in the long run? The damaged area we measure is like a no-go zone in the retina where cells are no longer functioning. As that area grows, it encroaches on the central regions of the eye needed for reading, recognizing faces, and doing fine tasks. Slowing the expansion means the healthy areas stay intact longer, so patients may be able to keep doing those everyday tasks independently for more years. Long-term follow-up will be key to quantify exactly how many extra years we gain. The direction we are seeing with Tinlarebant is clearly positive and can have a meaningful impact on people's everyday lives. Thank you very much. I appreciate your response. Now if we turn to safety, Dr. Schmetterer, as our DSMB chair, I wondered if you could answer a couple of questions for me. First is, in terms of the safety outcomes, what stood out to you in this trial, especially given that this is an oral small molecule administered daily for two years? Yeah, thank you, Nathan. This is, of course, a very important question. From the DSMB perspective, what was most reassuring is that there were no pattern of serious safety problems related to the drug over the two-year treatment period. We monitored laboratory tests, systemic side effects, and any new symptoms very closely. There was a zero rate of serious treatment emergent adverse events in both the treatment and placebo arms. Some mild to moderate side effects occurred, as they do with most oral medications, but they were generally manageable. Overall, we considered the safety profile of Tinlarebant very favorable for a chronic therapy. Thank you. The second question, were there any safety signals that you paid particular attention to as the DSMB chair, and how reassuring were the findings? For any new oral drug, particular attention is given to liver and kidney function, as well as potential cardiovascular or neurologic events and any unexpected clusters of symptoms. Unblinded safety data were reviewed at regular intervals, always with the question in mind, is there any emerging pattern that would warrant concern or a change in the study plan? In this trial, no such pattern was identified. This allowed the study to continue to completion, and it itself provides a strong indication that the safety profile of Tinlarebant was acceptable. Thank you very much. Now about the therapy itself in terms of the management of Stargardt disease. Dr. Michaelides, I wondered if you could explain to us, based on these results, how do you think that clinical management of Stargardt disease might change or might be modified if the drug is approved? Sure. Currently, we can readily diagnose Stargardt disease, but we can't offer a treatment that changes its course. We mainly provide low-vision support and coping strategies. Once Tinlarebant is approved, we would, for the first time, be able to intervene with a therapy that slows the underlying disease process. That would shift our focus to earlier diagnosis and earlier treatment, potentially before vision loss occurs. It will transform the care we offer patients and their families from, "We have nothing to slow this," to, "We can actually do something to help preserve your vision for longer." Yes, I see that's very promising. As a follow-up, you mentioned the opportunity to treat early in the disease process before major vision loss occurs. Could there also be a benefit for patients who've already lost significant vision? Absolutely, without doubt. ABCA4-associated retinopathy is always progressive, and it does not stop after the central macula has been lost. In some patients, the visual acuity may deteriorate to less than 2200, but even then, the remaining vision is very useful for patients, and they will still benefit from treatment slowing down deterioration. We want to treat early in the disease natural history in the same way that Tinlarebant targets early stages of the disease process. Yes, there is also a place for Tinlarebant in advanced disease. That's reassuring to know. Thank you. Now back to Dr. Sui. I was wondering, what would you tell a family asking how meaningful this treatment could be for their child? First, let me say that this is an exciting time for us as a practitioner, as we consider the potential to be able to offer our patients a treatment that can change the course of their vision loss. To answer your question specifically, I would explain that this medication does not cure Stargardt disease and does not restore lost vision, but it slows the damage that causes vision to decline. Over the span of many years, that slowing can mean more time reading regular print, more independence at school and work, more time seeing loved ones' face, and a slower transition to severe vision loss. For a progressive disease with no current treatment options, gaining those extra years of better function is highly meaningful, and we now have solid scientific evidence that this drug can provide that benefit while being very well tolerated. Thank you very much, Dr. Sui. Dr. Michaelides, I was wondering if you could tell me, why is it so significant that this is the first interventional trial in Stargardt disease with over 100 participants showing such a strong treatment effect? Absolutely happy to. Stargardt disease is rare, so most studies to date have been small and not optimally designed, which makes it hard to prove that a treatment really works. This is the first large, well-controlled, randomized, multicenter interventional trial with more than 100 patients that not only completed successfully, but showed a clear 36% slowing of disease progression with excellent safety. Tinlarebant sets a new standard for what's possible in rare inherited retinal diseases and provides a template for future therapies in this field. Thank you. As a follow-up, as we explained, DRAGON is the first interventional trial that focused on adolescent patients with Stargardt disease. Is there a difference in the disease severity or progression between adolescents and adults with Stargardt disease? Would you expect this treatment to be as effective in adult Stargardt patients? Very important question. To start, Belite Bio were to be congratulated on focusing on childhood-onset Stargardt disease, also known as childhood-onset ABCA4-associated retinopathy, and also engaging with patients and Stargardt specialists to deliver this landmark trial, something I've been championing for a decade. This is the most severe ABCA4-associated condition. The underlying mechanism of disease for all ABCA4-associated conditions is the same and is thereby targeted by Tinlarebant. Tinlarebant will therefore be effective for all forms of ABCA4-associated retinal disease, including adult-onset Stargardt disease, late-onset Stargardt disease, cone dystrophy, and cone rod dystrophy. Tinlarebant needs to be made available to all patients with a retinal dystrophy caused by ABCA4. That's very exciting to hear. Thank you for your comments. My final question is to our CMO, Dr. Hendrik Scholl. I was wondering, Hendrik, what do you believe are the next steps toward potential approval of Tinlarebant, and what additional evidence do you think the regulators will be looking for? Yeah, thank you, Nathan. The next steps will involve submitting the full data package from this trial, along with supportive data from natural history studies and extension follow-up to regulatory agencies. Regulators will look at how robust and consistent the efficacy signal is, how clinically meaningful the structural endpoints are, and how reassuring the safety profile appears, especially for long-term use in young patients. We believe we have a clear path toward approval of Tinlarebant and for providing access to the treatment for patients suffering from Stargardt disease in the near future. Thank you, Hendrik. That ends our panel discussion. I would like to thank Hendrik and our KOL panelists for your time today. This was a great discussion. As we've shared, Tinlarebant achieves statistically significant and clinically meaningful outcomes, demonstrating its potential to substantially slow disease progression. These robust phase III results represent a transformative moment for Belite Bio and for the Stargardt community. As we move forward with our regulatory interactions globally, we're getting closer to delivering the first approved treatment option for people living with Stargardt disease. It's a very exciting time for us. We would like to extend our sincere gratitude to all of the investigators, patients, and families whose dedication to our efforts made this achievement possible. I'd also like to recognize our incredible team here at Belite Bio. Thank you for your unwavering commitment to our advancing Tinlarebant in degenerative retinal diseases, a mission that can have a profound positive impact on people living with these debilitating diseases. With that, we will open the call for questions. Operator? We will now begin the question-and-answer session. If you would like to ask a question, please raise your hand now. If you have dialed into today's call, please press star nine to raise your hand and star six to unmute. Please stand by while we compile the Q&A roster. Your first question comes from the line of Marc Goodman with Leerink. Your line is open. Please go ahead. Marc, a reminder to kindly unmute yourself. Hi, can you hear me now? Hello? Yes, we hear you. Thank you. Sorry about that. I just want to ask a little more detailed questions about the side effects. More detail, the four TEAEs that led to the drug discontinuation, the two that led to study condition. Can you give us a sense of what were they? I mean, we kind of know what the side effects were going to be, right? The xanthopsia, the late-adaptation. We knew that these were going to be there. Were the percentages basically the same as what we've seen in the past? Were there any surprises? Were these mild only? Were there some that complained that they were serious? I'm just kind of, what more can you give us about the side effect profile there? And then, have you had discussions? When were the last discussions with the regulators? When's a pre-NDA meeting planned? Just give us a sense of that. I know you've started to make those comments about, is one study enough? Do they understand that the BCVA is not going to be held to two years and that that's not really the endpoint to look at? This DDAF is the endpoint. Just give us a sense of what you think the regulators really, truly understand and whether one study is enough. Thanks. Hendrik, do you want to take this? I'm happy to take that. Thank you. Yeah, maybe I would like to start with the statement that Tinlarebant was safe and well tolerated in this phase III registration trial. The overall safety was consistent with our prior studies that obviously you're familiar with. Thank you for your comment. Supports the potential for strong long-term compliance with such an oral once-daily regimen, of Tinlarebant, for this chronic disease. It is important to know that there were no serious treatment-related adverse events observed at all. The most commonly reported ocular adverse events were mild, and as we had discussed before, are in line with the mechanism of action. That includes xanthopsia, a condition in which objects appear yellowish or discolored, if you will. There are other phenomena like protanopsia, where it is reddish and so forth. It is like a discoloration, which has to do with the saturation within cone photoreceptors, but it only lasts for a few seconds or minutes and is typically very mild. The other one being delayed dark adaptation. This is the other photoreceptor type. This is the rod photoreceptor type that is active in low luminance or in the night. The final threshold is not affected. If you wait long enough, the sensitivity of the rod photoreceptors is not affected by Tinlarebant, but it takes longer simply because it takes the substrate needs to get into the photoreceptors, and that takes longer under Tinlarebant treatment. The most commonly reported non-ocular treatment-related adverse events were nasopharyngitis. In all of these cases, they were unrelated: headache and acne. We conclude this is a very favorable tolerability and safety profile. In terms of our interaction with regulators, is one study enough? The answer is the important point is DDAF lesion growth considered an approval for endpoint. The position of the FDA is, yes, absolutely. It is an approval for endpoint because the FDA considers lesion as areas that have no vision or strongly compromised vision. We know that in Stargardt, these areas come with absolute scotoma, so no sensitivity left. If you slow down degeneration of the seeing cells, that has an immediate impact on vision and in long term will have a measurable impact on vision. Therefore, is DDAF accepted by the FDA as the approval for endpoint? Given the robustness of our findings, the confirmatory evidence of the fellow eye makes us confident and hopeful that, yes, one study will be enough for market authorization. Might further add. We have discussions with the major regulatory authorities across the world. Just last month, we've disclosed that we had a discussion with the MHRA and with the China NMPA. Both agencies have actually given us a green light to submit. This is based on the positive internal analysis data, and that is on the DDAF as the primary endpoint. The FDA will really have discussions with the FDA and the data being robust enough, which we believe that this is very robust data given the p-value is smaller than 0.001. We believe we do have a one single study path forward. I hope that answers your questions. Oh, and regarding EMA and Japan, Japan has already given us a green light as well. EMA did not reject or did not oppose the endpoint being destructive endpoint. We believe the regulatory authorities accept this endpoint all around the world. As far as the four events that led to drug discontinuation, were they basically xanthopsia or delayed dark adaptation, or was there something else? Nathan, do you have more detail on that? We do not have the breakout for the reasons, but you can imagine that the majority of those are going to be probably half of those will be, sorry, due to the ocular AEs. Again, we do not have the breakout. Some patients just withdraw consent or sometimes are lost to follow-up. We do not have the final breakout for the subject disposition. Again, because the number is so small, it is not really that meaningful. You can assume that right around half of those, so two of those would be related to the ocular adverse event. Thanks. Your next question comes from the line of Steve Seedhouse with Cantor. Your line is open. Please go ahead. Thank you. Congrats to the entire team on the landmark success here. A few questions I wanted to walk through first. Just back to the broad label expectation spanning all ages, even beyond what was studied in DRAGON and a wide range of lesion sizes. Can you just talk about your conviction there again? I wanted to specifically ask about the fellow eye data here and whether that will sort of help solidify a broad label. Obviously, there were sort of unselected lesions in the fellow eye. I would imagine it would help. Can you get that in the label? Can you bring that up with regulators? Just comment on that, please. Hendrik, you want to take this one again? Yeah, absolutely. There is a guidance of the FDA on confirmatory evidence. I would not blame the FDA that the specialties in medicine that take care of paired organs such as nephrology and ophthalmology are not perfectly represented. What I mean is that the fellow eye data is not listed in the guidance of the FDA. In our opinion, what could be better confirmatory evidence than the treatment effect in the fellow eye that was not the study eye? In our opinion, this is very strong confirmatory evidence that we find a treatment effect even in the fellow eye. As I explained in the presentation, the fellow eye treatment effect was not necessarily expected because you have a patient in front of you. Yes, you have an inherited disease. Yes, it is bilateral, but it does not mean that the retina looks identical. It is not so rare that you have a given DDAF lesion in one eye and the other eye has not developed a lesion yet or has a very large lesion, which means that the variability in the fellow eye was simply larger. Still, we reached statistical significance in the treatment effect for the fellow eye, plus a treatment effect size that was comparable to what we find in the study eye. In our opinion, this is very strong confirmatory evidence. I would like to add again, Stargardt being a medical need, there is nothing that doctors, physicians like Hendrik and Professor Michaelides can offer patients right now. This is a real unmet medical need. Given the DRAGON Study showing such robust treatment effect, which is both on slowing atrophic lesions, but also slowing the progression of toxic bisretinoid-affected cells, which Professor Michaelides talked about, that not only on slowing lesion growth, but also slowing the transition of sick cells to actually end up being dead cells. This is across the board, mid-30% treatment benefit. We believe that Tinlarebant would be broadly applicable across silo populations. Maybe Professor Michaelides, you want to shed some light? All I would add, Tom, is 100% agreement. I mean, from a clinician seeing patients day in, day out, the data presented is just fabulous. I mean, genuinely, it is greater than what I'd anticipated. I had in my mind, as long as we're getting a 15%-20% type slowing, that would be wonderful. To get 36% is truly game-changing. This is a career-defining study. I've got no doubt that this should be offered to all patients with ABCA4 disease as long as they don't have very much end-stage disease. Basically, the vast majority. I think the data is very compelling clinically. I would be shocked if the regulators do not see how reliable, repeatable, and clinically relevant the data is. Also, just on the adherence side to the daily dosing, do you have data available on just the mean number of days missed in a month for patients in the study or just ballpark how good was adherence to the daily regimen? Professor Michaelides, maybe you want to share your personal experience with your patients? Compliance was—thanks, Steve. I mean, compliance, in my experience, was excellent. We had 19 children at the London site, and compliance was super. In fact, when they finished the study, they really wanted to take part in further studies in DRAGON 2. They keep contacting the study coordinator and the fellows asking, "When the data is going to be out? When can we get access to the therapy?" Compliance was super. Really, really good. Great. Steve, if I may add, right, I just showed this slide that was shown in the presentation on our biomarker. When you look at the pharmacodynamic effect of Tinlarebant in the DRAGON trial, please allow me to say this is almost perfect, right? There is almost no variability when you look at the standard error bar, which is super small. It is an almost perfect reduction to 80% with almost no variability. This indirectly shows that patients were sticking to the regimen and would take the medication. Great. Just maybe a multi-part question on the anticipated marketing of Tinlarebant here to close. First, the 50,000 patient estimate, I know you provided a range in your presentation as well. Can you talk about sort of how well established the epidemiology is here and just conviction and diagnosis rate maybe even increasing once there is a treatment available? On the company's planning side, was hoping you could recap your marketing strategy in the U.S. and outside of the U.S., given things are going to move pretty quick here. It sounds like you're going to be filing in the first half of the year. You have breakthrough designation, fast track, etc. Wonder if you could comment on just your strategy, your progress towards building out a commercial organization. Sure. I'll start with the commercial part, and then I'll let Nathan answer the prevalence. Yes, we're coming to the Christmas period. We are looking or we're planning to file first half next year. We need to wait for the study report to come out. We've almost got everything ready, all the documents ready for submission. Of course, we still need to meet with the regulatory agencies and prioritize what will we go first. I think priority will probably be with the FDA. Again, it's still a discussion we need to have with all the regulatory authorities, which has also given us green light to go ahead. This is an internal discussion, and we'll update once we have more progress or more updates available. Nathan, you want to comment on the prevalence of Stargardt patients in the U.S.? Yeah, if we go—sure, sure. If we go back to the slide showing the estimated patient population, of course, this is, as Hendrik mentioned, a review article that we just published. The over/under, if you will, is on the far right-hand side in terms of what we see in the United States. $59,000 being the high estimate, $40,000 being the low estimate. You can look at roughly about 50,000 if you want to call it the mean. Imagine we get roughly 33% of that base. That's actually a very good catch in terms of the market size, particularly when you talk about premium pricing for an orphan disease. Those are the numbers. We don't know exactly what the number of patients that we'll be able to capture in terms of the treatment, but it should be pretty substantial. We estimate at least 1/3 of those subjects. Hendrik, you want to comment on the diagnostic rate and the expectation of increasing diagnosis and with new awareness with available treatment. In your experience, how would this go about? Thank you, Tom. Happy to—the thing is, it's a genetic disease, right? It affects the macula. A patient affected by such a disease would go to the ophthalmologist, would go in the U.S. to a retinal specialist, and would seek care. You see families, obviously, where not just one child in the family is affected. In some children, you make a diagnosis; you can make a diagnosis very early, as Professor Michaelides also explained. You can catch it earlier. The genetic screening is key and is more and more common. The good thing is that although there are many mutations in the ABCA4 gene, if it's biallelic, you have full penetrance, right? Meaning that the disease will manifest. It's not that you carry a mutation on both alleles and you would not get the disease. We call that penetrance. This disease is fully penetrant if you carry mutations in the target gene, ABCA4. Yes, genetic testing is key, but it will be important to confirm the diagnosis in the future when Tinlarebant is available on the market. Yeah, I'd like to refer this to Professor Helen Sui. I think they've got a very established genetic testing established at her clinic or at her organization. Helen, do you want to— Yes. Share some of your experience? Yes. In China, if a child has screening a vision at school, which not passed, the parent always leads their children to see an ophthalmologist. If we presume it's an inherited disorder, we'll do genetic testing. Most of the patients didn't need to pay anything. The hospital will cover the genetic testing. We have a very high molecular diagnosis rate. In our hospital, it's about 100%. Because we are the tertiary hospital, in some secondary hospitals, they also can have about a 50% molecular testing rate. It is not a problem for molecular testing. Thanks so much. Congratulations on a really impressive win here for Stargardt patients. Thank you. Thanks, Steve. We are nearing the end of our call time, and we have time for two more questions. Your next question comes from the line of Yi Chen with H.C. Wainwright. Your line is open. Please go ahead. Hi. Thank you for taking my questions. Congratulations on the data. One question about the placebo BCVA, which also showed no significant change in the trial. Considering that, do you think there will be an issue with payers supporting reimbursement of Tinlarebant with the data showing there is no BCVA change in placebo as well? Also, in real-world practice, do you think compliance for patients will be as high as observed in the clinical trial? Thank you. I guess this is a question for Professor Michaelides and Hendrik. I think they mentioned this on the BCVA part in their talk. Professor Michaelides, you want to go first? Yeah, sure, Tom. Yeah, the visual acuity was anticipated. We would not expect there to be much in the way of change given that period of time and given the variability in measuring vision in Stargardt disease. I would not really have anticipated that. The structural changes are well documented. We know those are very rapid, and they precede functional change. They are anticipated, in my view. Certainly, patients were excellent with compliance before they knew it worked. Now that it works 36% in slowing progression, I've got even less doubt that they'll take the medication religiously. Maybe I'll pass over to Hendrik if he's got anything additional. No, Michelle, you said it perfectly. There's nothing to add, really. Thank you. Thank you. Your next question comes from the line of Greg Suvannavejh with Mizuho. Your line is open. Please go ahead. Greg, a reminder to kindly unmute yourself. Moving on, your next question comes from the line of Bruce Jackson with Benchmark. Your line is open. Please go ahead. Hi, can you hear me okay? Yes. Yeah, perfect. Okay, super. Congratulations on a well-designed, well-executed trial. I want to follow up on the comments about using it in the adult onset population, late onset mitochondrial dystrophy. Is that something that's going to be encompassed in the anticipated labeling, and is there anything else that needs to be done in order to achieve that? Hendrik, you want to take this? Yeah, happy to. It's all about the generalizability of the data. Since the DRAGON trial involved adolescent patients with Stargardt disease, 12-20 years old, we feel that we already covered adults, at least partially, right? In terms of generalizability, given that this is a monogenic disease due to mutations in ABCA4, I don't think it's difficult to argue that this is all the same disease. If you have an onset, let's say, at age 35, right, why is this a different disease than in age of 1-8, age 15? It's exactly the same disease. It's because of this dysfunction of ABCA4 in the photoreceptors of a given retina. We do not believe that this should be difficult or the other way around, right? If the label was restricted, what would you do as a practitioner or a patient when you have a later onset disease due to mutations in the gene? What's the alternative? There is none, really. We feel that we are in a very strong position to get a broad label and to cover with Tinlarebant all these, let's say, currently, in terms of nomenclature, differently named conditions that Professor Michaelides spelled out, namely typical Stargardt disease, late onset Stargardt disease, ABCA4-associated cone dystrophy, and ABCA4-associated cone rod dystrophy. It's all the same disease with slightly different manifestations. Okay. One more follow-up, if I may. Looking at the discontinuation rate, it looks to be quite low. I was wondering if you happen to have any comparisons with other trials or natural history that might give us some additional information on that. That's it for me. Thank you. Hendrik, do you want to follow along with this question? Yeah, I would pass on to Nathan to share our observations from the phase II study, which are perfectly in line. The low discontinuation rate speaks in favor of a very motivated patient population, given that these adolescents are about to lose vision, and this is the potential treatment. Now we know it is the treatment, will remain very motivated to take the drug. Maybe, Nathan, you can share our observations from our phase II clinical trial. Yeah, sure. Thank you. Happy to, Hendrik. I think it's important to note that in the phase II trial, of course, the sample size was much smaller, only 12 patients. No patients were lost to dropout or withdrawal during that study. All 12 completed. If you talk about, yeah, Bruce mentioned other studies, there's really no other studies to compare to. The other studies conducted in Stargardt disease were in largely adult populations using, for instance, a mixed use that was a very aggressive drug. The dropout rate in that trial was well over 25%. I think it actually approached 30%. This is phenomenally low. There is really little to compare to in terms of prior clinical studies. We can only compare to what we've seen in our own prior clinical studies. We're seeing this drug very well tolerated across all patients in this age range, at least. I expect that will be the same in an adult population. Thanks. Thanks, Nathan. There are no further questions at this time. This concludes today's call. Thank you for attending. You may now disconnect.