Hello, welcome to the 30th Deutsche Bank Depositary Receipts Virtual Investor Conference, DBVIC. My name is Zafar Aziz from the DR Investor Relations Advisory Team at Deutsche Bank. I'm pleased to announce that our next presentation will be from Belite Bio. Before handing over to our presenter, some points to note. Please submit your questions at any time throughout the presentation. All of our presentations will be recorded and can be accessed via the Deutsche Bank website, adr.db.com. At this point, I'm very pleased to welcome our speaker from Belite Bio.
You are live.
Thank you. Hello. I would like to welcome everybody to the presentation of Belite Bio. Just have to share my screen anymore. Here we are. All right. I hope my screen can be seen. The title of my talk is Early Intervention with an Oral Treatment for Macular Degeneration. This is our forward-looking statement and legal disclaimer. My name is Hendrik Scholl. I'm the Chief Medical Officer of... I must say that I can't see my slides here on the screen, or I can't see... Are my slides being shown?
Yes. Right now, I see the management team.
Okay.
Now the pipeline.
Okay, good. What we see right now? We see the Belite Bio executive team?
Yes, now disclaimer.
Okay.
My name is Hendrik Scholl. I'm the Chief Medical Officer of Belite Bio. I'm a retinal specialist by training. Started my training in Germany, then the U.K., and then led the research at the Department of Ophthalmology at the University of Bonn, Germany, before recruited to Johns Hopkins as a Professor of Ophthalmology, where I led retinal dystrophies. At the time, at the Wilmer Eye Institute, we implemented the largest natural history study for Stargardt disease named the ProgStar study. I will get to the ProgStar study later in my presentation. In 2016, I was recruited back to Europe, led the Department of Ophthalmology at the University of Basel, and 2 years ago, joined Belite Bio as Chief Medical Officer. This is our forward-looking statement.
I will get to our Belite Bio pipeline overview now. Belite Bio has Tinlarebant in development. I must admit that the slide is very small that I can't see it really. I pretty know what's on the slide. The developing Tinlarebant for Stargardt disease. We completed a 2-year treatment trial in a phase II in 13 subjects. We completed a phase III clinical trial. Let me try to make my screen bigger here so that I can see the slide. We completed a 2-year treatment trial. Our global trial, the so-called DRAGON trial on 104 subjects.
We are currently running a phase II/III 2-year treatment trial called DRAGON II, which is currently ongoing in 73 subjects. We started our U.S. FDA rolling submission this month. We also have a treatment trial in geographic atrophy, which is called PHOENIX, which is a phase III 2-year treatment trial, which is currently ongoing and is completely enrolled with 530 subjects. Tinlarebant is a novel once daily oral tablet designed to bind to serum Retinol-Binding Protein 4 or RBP4 as a means to specifically reduce retinal delivery to the eye. This approach is intended to slow or halt the formation of toxic retinal delivery byproducts that are generated in the visual cycle and are implicated in progression of Stargardt disease and GA.
Belite Bio believes that intervention directed at the root cause of retinal pathology, namely bisretinoid-mediated toxicity but not complement activation, will be the best approach to potentially slow disease progression of Stargardt disease and geographic atrophy. We have an unmet market opportunity. There's no FDA-approved treatment for Stargardt, and there are no FDA-approved oral, orally administered treatments for geographic atrophy. We have breakthrough therapy, fast track in rare pediatric disease designation in the U.S. and orphan drug designation in the U.S., in Europe and Japan, and pioneer drug designation called Sakigake designation in Japan for Stargardt disease. We have 14 active patent families. We have composition of matter patent until at least 2040 with our patent term extension. Let's get to the Tinlarebant overview. First, the market opportunity.
We recently completed a survey in collaboration with large academic centers looking into the prevalence of Stargardt disease. That resulted in an estimated number of patients in the U.S. of more than 50,000 and more than 100,000 in China and in Europe. When we get to geographic atrophy, which is obviously age-dependent because it's the late atrophic form of age-related macular degeneration, an estimated pretty much exactly 1 million people in the United States and approximately 5 million people globally have geographic atrophy in at least one eye. Here we get to the normal processing of vitamin A in the visual cycle, which is important to understand the mechanism of action of Tinlarebant.
Vitamin A or retinol is transported to the eye in a protein complex with RBP4 and transthyretin, TTR. Unlike other tissues, the eye has a unique requirement for uptake of retinol bound to RBP4 due to the abundant expression of an RBP receptor. The RBP-RBP4 receptor, called STRA6, binds specifically to retinol RBP and TTR. Internalized retinol is used to generate rhodopsin in the visual cycle. In Stargardt patients, dysfunction of the ABCA4 protein results in accumulation of all-trans retinal, which is very toxic itself, leading to the formation of retinal dimer lipid complexes, including A2E. The deposition of these complexes into the RPE results in accumulation of cytotoxic bisretinoids and retinal cell death through diverse mechanisms. Tinlarebant, shown on the lower half of the slide, is a small molecule, an RBP4 antagonist.
Has a hundredfold greater affinity for RBP4 compared to retinol and competes with retinol for binding to RBP4 and does not allow the binding of TTR. The Tinlarebant RBP4 complex is liberated into the circulation, eliminated through the kidney due to its small size, and this results in reduced retinol delivery to the eye, and this in turn reduces the amount of toxic retinal and A2E. There is similar pathology of Stargardt disease and GA. Stargardt disease and GA show similar phenotype, where loss of retinal pigment epithelium and photoreceptors in the center of the retina, this center is called the macula, leads to loss of vision. Fundus autofluorescence images. There are two examples shown on the right of the slide.
On the left a patient with Stargardt disease aged 17 years old, and on the right, a patient with geographic atrophy aged 71 years. Fundus autofluorescence imaging allows to visualize bisretinoid content in the retinal pigment epithelium and shows loss of signal in degenerated areas simply because RPE has degenerated and there is no signal coming from these cells. An increase in signal peripheral to these lesions, which is typical for both Stargardt and geographic atrophy. Slowing or halting the spread of retinal lesions is the intended effect of Tinlarebant treatment. Now we get to the DRAGON clinical trial. The DRAGON and the DRAGON II clinical trial have essentially the same design and are phase III clinical trials for the treatment of Stargardt disease.
In the left column we see the design of the DRAGON trial, on the right of the DRAGON II trial. 104 subjects were enrolled and completed the DRAGON trial. We have currently 73 subjects enrolled in the DRAGON II trial. Both are global clinical trials. Randomization is 2 to 1, was 2 to 1 in the DRAGON trial, and is currently 1 to 1 in the DRAGON II trial. It's a double blind placebo-controlled clinical trial with a treatment duration of 2 years. The primary outcome measure is DDAF. I will explain that in a moment. Definitely decreased autofluorescence, lesion growth rate, and safety and tolerability. Other measures include questionably decreased autofluorescence or QDAF, best corrected visual acuity, spectral domain OCT outcome measures, and microperimetry as an exploratory endpoint.
We performed an interim analysis in the DRAGON trial and anticipate to perform one in the DRAGON II trial. Key inclusion criteria was age between 12 and 20 years old, so an adolescent patient population. Diagnosed with Stargardt disease with at least 1 mutation identified in the target gene with the name ABCA4, and an atrophic lesion size within 3 disc areas, and the best corrected visual acuity of 20/200 or better in the DRAGON and 20/400 in the DRAGON II trial. This slide shows the results for the biomarker in the systemic circulation, namely the % change from baseline in Retinol-Binding Protein 4, RBP4. The goal of the pharmacodynamic effect was to reach at least 70% reduction in RBP4.
The graph shows that daily dosing of five milligrams per day with Tinlarebant led to a sustained 80% reduction of RBP4 with very little variability. RBP4 levels returned to close to the original baseline value by 28 days of the treatment when Tinlarebant was discontinued. I mentioned the ProgStar study, which 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. It also defined QDAF, questionably decreased autofluorescence, as regions of partial autofluorescence reduction that eventually evolve into DDAF, and thus provides an earlier mark of progression.
By combining DDF and QDF, namely DAF, decreased autofluorescence or total decreased autofluorescence, captures the full extent of RPE impairment and offers a sensitive, comprehensive metric across a wider disease spectrum. DAF was the key secondary endpoint in the DRAGON trial. The ProgStar study confirmed that these autofluorescence-based measures are highly reproducible, correlate with functional decline, and are suitable for detecting progression over periods of 12 to 24 months. As a result, DDF, QDF, and total DDAF have become the standard structural outcome measures for interventional trials in Stargardt disease. In ProgStar report number 17, it was shown that the mean DDF growth rate in the ProgStar cohort over 24 months was 0.74 square millimeters per year. DDF 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 trial.
Consequently, the primary endpoint in DRAGON was the annualized rate of lesion growth in the aggregate area of DAF from baseline as assessed by fundus autofluorescence imaging at month 24. Data will be 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 1 dose of study medication. In addition, the MFAS subjects must have defined DAF lesions meeting the eligibility criteria at baseline and have at least 1 post-baseline assessment. Data analysis used a mixed model for repeated measures, MMRM, measuring change from baseline in DAF in the study eye and including terms for treatment visit, treatment visit interaction, baseline focality, and baseline DAF lesion size. The statistical analysis plan 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 trial. This approach was discussed and supported by our DSMB when reviewing the final efficacy and safety data. This slide shows the primary endpoint of the DRAGON trial, 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 maintained at 35.4%, but the standard error was significantly smaller and resulted in a P value which was smaller than 0.0001.
DDF lesion growth was slowed down to 0.38 square millimeters per year, while a progression of 0.59 was measured in the placebo group. To put this number into perspective, as mentioned earlier in the ProgStar study, the mean DDF lesion growth was 0.74 square millimeters per year. Also mentioned earlier, it's an obvious thing, Stargardt disease is an inherited disease affecting both eyes, and because Tinlarebant is an oral therapy, we are not used any longer to oral therapies in ophthalmology for the last 2 decades. This oral therapy is expected to benefit both eyes and therefore the fellow eye was also evaluated. Because of phenotypic disparity between the 2 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, a statistically significant treatment effect was observed in the fellow eye. Treatment with Tinlarebant reduced DDAF lesion growth by 33.6% compared with placebo. Decreased autofluorescence, DAF, as mentioned earlier, is the composite of all areas which reduce autofluorescence, representing the full extent of RPE impairment. The ProgStar study established that DAF growth rate is one of the most sensitive and comprehensive structural progression measures, and it correlates with functional outcomes such as microperimetry sensitivity loss and future visual acuity decline. It is highly useful in early intermediate stages where pure DDAF may be small. DAF was the key secondary endpoint in the DRAGON trial. It was found that Tinlarebant slowed DAF lesion growth by 33.7% compared to placebo, and this 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 reaching statistical significance at 0.017. Not only were the findings for the primary endpoint supported by the findings of the key secondary endpoint, but a 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 the 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 and was 39.7 letters at the end of study.
Similarly, visual acuity was 39.4 letters for the placebo group at baseline, exactly 40 letters at the end of study visit. An ETDRS letter score of 39 to 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 8 letters, as shown by Parker and colleagues, observed changes of 5-10 letters are frequently indistinguishable from measurement noise. Such a minor change in average visual acuity over the 2 years is in line with the natural history of Stargardt disease and was indeed observed in the ProgStar study. This is shown on this slide.
In the prospective cohort of 434 Stargardt disease patients in the ProgStar study, the overall rate of best-corrected visual acuity loss was only 0.55 letters per year over the 2 years. For eyes with baseline best-corrected visual acuity between 20/70 and 20/200, which is shown in the lower left spaghetti plot, and exactly matches the DRAGON cohort, visual acuity decline declined at a rate of 0.6 letters per year. Let's get to the safety results. Importantly, Tinlarebant maintained an excellent safety and tolerability profile over the 2-year course of the treatment. The table shows systemic safety and tolerability, namely the number of subjects who experience at least one non-ocular Treatment-Emergent Adverse Event. A total of 6 SAEs was reported in the study. By the way, 4 in the placebo group and only 2 in the treatment group.
All events were non-ocular, with 4 assessed as unrelated and 2 assessed as unlikely related to the study treatment. The most reported non-ocular EAs were nasopharyngitis, headache, and acne. Most events were mild and resolved during the study period. Regarding ocular safety and tolerability, xanthopsia, meaning a yellowish appearance of the visual scene where everything can be seen but it's tinted, like, in yellow for seconds or minutes, and delayed dark adaptation and night vision impairment were mild. Most resolved while on treatment. There were no serious ocular treatment emergent events. 4 TEAEs led to study drug discontinuation and 2 TEAEs led to study discontinuation.
Briefly on our PHOENIX trial in geographic atrophy, we first performed a phase I clinical trial to look at PK/PD in an elderly population, which allowed to conclude that the PK/PD profile in the elderly is exactly the same as in an adolescent patient cohort that was studied in the DRAGON trial. This slide shows the clinical trial design of our PHOENIX trial. 530 subjects had been enrolled into the trial. It's a global trial. Our efficacy endpoint is the reduction in atrophic lesion growth as measured by fundus autofluorescence imaging and is the FDA accepted primary endpoint. With this early intervention, we target patients with small lesions due to the potential slow disease progress at an early stage.
We have an oral treatment. It's exactly the same dose, 5 milligrams per day, which is well suited for long-term treatment for chronic diseases. We have a broad potential with a primary focus on geographic atrophy, but the potential to treat earlier stages, including intermediate AMD. In summary, the DRAGON trial met its primary endpoint. A highly statistically significant slowing in DDAF lesion growth was observed in subjects treated with 5 milligrams 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 milligrams 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 2 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's 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, starting in Germany, then the U.K., then more than six years in the U.S., and now in Switzerland. 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. With this, I would like to conclude my presentation and happy to turn over to the Q&A. My apologies for this little glitch in the beginning because I anticipated to actually see my slides. I only saw these very small ones. The good thing is I have very high visual acuity and was able to even read very small print. This is exactly what happens with Stargardt patients.
They lose exactly this ability and are then extremely disabled and end at a visual acuity of about 20/800 or worse. Thank you very much. I get to. One question is, "With visual acuity stable in DRAGON, could you run trials in earlier stage patients before atrophy develops?" This is an excellent question. Actually, this refers to what we found in ProgStar report number 5, looking at retrospective patients over a five-year period and measured or detected the incidence of DDAF lesions. Indeed, incident DDAF is a good endpoint actually. As I mentioned, as shown in ProgStar report number 5, it needed a five-year period to come up with reasonable and statistically significant numbers.
The answer is a principle, yes, but it's very difficult because the duration of the trial essentially would be prohibitive in order to find a statistically significant treatment difference between placebo and in treatment. I would like to add and use the question, which is really good. Namely, what is the anticipated treatment effect for patients that have not yet developed a lesion? Although this was not studied in the trial, but this was discussed after I presented the top-line results on December first in a Q&A by Professor Michel Michaelides, our top enroller at the Moorfields Eye Hospital in London, in the United Kingdom.
He mentioned that he would anticipate an even more pronounced or even more beneficial treatment effect in patients that have not yet developed significant visual acuity loss and maybe not even a lesion. Indeed, the patients in the DRAGON trial had a mean visual acuity, I mentioned that early, of 20/160, meaning that they are on average already underneath the threshold of reading ability. What I wanna say is that in the future, I anticipate that patients before they reach that threshold, losing reading ability, would benefit the most from treatment with Tinlarebant. Another question is, given the rarity of Stargardt disease, how are you approaching regulatory pathways like FDA? This has been discussed with the FDA in-depth.
We had discussed the interim results with the FDA. We were given a breakthrough designation by the FDA. We had our pre-NDA meeting with the FDA, we started our submission this month, a rolling submission to the FDA. Everything is according to plan. We know that the FDA accepts this endpoint as a primary endpoint, DDAF. This was already discussed when the ProgStar study was initiated in 2011. We visited the FDA with Foundation Fighting Blindness and discussed the design of the ProgStar study at the time.
Obviously, there's precedence because the lesion type that is being used as a primary endpoint is more or less the same than that used in geographic atrophy that led to the approval of the two injectables currently in the United States. Another question is, the company's position to be the first approved treatment ever for Stargardt. How does that change your commercial strategy? Not entirely clear why it changes our commercial strategy. It is true that we may run into a complex situation when we find a significant treatment effect for geographic atrophy as well.
This has to do with a significant pricing difference, given that Stargardt is a rare disease, not so rare, by the way, but is an often indication compared to geographic atrophy, which, as I mentioned in my presentation, affects pretty much exactly 1 million patients in the United States. We have a commercial strategy in place in order to not lose out on the commercial opportunity for Stargardt disease, which is very significant. Another question that I see in the list here is, Belite Bio is achieving 80%-90% of before reduction. Is that better than previous attempts with this mechanism? The answer is a definite yes.
There was a precursor, so to speak, trying to reduce RBP4, but that did not allow a sustained reduction of RBP4 and not to that degree. I think this made the difference because any precursor so far has failed. You need to reach at least 70% reduction of RBP4 to see this treatment effect. At least this is indicated by the preclinical and clinical data that we have available to date. It's now 3:30, or 3:30 P.M. my time in Switzerland. I believe I have come to the end of this Q&A. I'd like to thank the audience for these excellent questions.
I will be available for any follow-up questions or any other questions that are present. Please be in touch. Again, I would like to thank the organizers for the invitation to speak. Thank you.