Ladies and gentlemen, thank you for joining us, and welcome to the Belite Bio Fourth Quarter and Full Year 2024 Earnings Conference Call. — Operator Instructions — I will now hand the conference over to Julie Fallon. Please go ahead.

Hello, and thank you for joining us to discuss Belite Bio’s Fourth Quarter and Full year 2024 financial results. Joining the call today are Dr. Tom Lin, Chairman and CEO of Belite Bio; Dr. Hendrik Scholl, Chief Medical Officer; Dr. Nathan Mata, Chief Scientific Officer; and Hao-Yuan Chuang, Chief Financial Officer. Before we begin, let me point out that we will be making forward-looking statements 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 detail. Now I’ll turn the call over to Dr. Lin.

Thank you for joining today’s call to discuss our fourth quarter and full year 2024 financial results. 2024 was an exciting year for Belite as we continued to make strong progress towards advancing Tinlarebant in patients living with Stargardt disease and Geographic Atrophy. For those who are new to our story, Tinlarebant is a first-in-class oral therapy intended to reduce the accumulation of toxic vitamin A byproducts, which have been implicated in the progression of retinal lesions in patients with Stargardt disease and Ge1 ographic Atrophy. We believe this approach will be effective in slowing or halting lesion growth, which would ultimately preserve vision. It is important to note that our approach focuses on early intervention of emerging retinal pathology that is not mediated by inflammation. We believe that this may be the best approach to potentially slow the pro- gression of Stargardt and GA. To give you some perspective on the importance of this potential therapy, Tinlarebant has been granted Rare Pediatric Disease and Fast Track designations in the U.S. and Pioneer Drug designation in Japan. It has also been granted Orphan Drug designation in the U.S., Europe and Japan. We believe this speaks to the significant unmet need for both indications, as currently, there is no approved treatment for Stargardt disease and no approved oral treatment for GA. And more importantly, we are uniquely positioned, as we are already in global Phase III trials for both indications. So with that, let me provide a high-level overview of the recent progress we have made. We have 2 studies underway with Tinlarebant in patients living with Stargardt disease. These are the Phase III DRAGON trial and the Phase II/III DRAGON 2 trial. As part of the Phase III DRAGON trial, we recently announced that the Data Safety Monitoring Board has completed its interim analysis, which is based on all subjects having completed the 1year assessment period. The DSMB recommended that the trial proceed without sample size increase or modifications, so essentially maintaining the sample size of 104 subjects. In addition, they recommend we submit the data for further regulatory review for drug approval. With the DSMB’s review done, completion of trial is on track for end of this year. The DRAGON 2 trial continues to progress rapidly. We have enrolled 11 of our targeted enrollment of approximately 60 subjects, including about 10 Japanese subjects. Trail from the Japanese subjects is intended to expedite a new drug application in Japan, to which we 2 have already been granted a Pioneer Drug designation. In GA, we also continue to progress in our clinical global Phase III PHOENIX trial, which has already enrolled over 400 subjects to date. We expect to increase the number of subjects to be enrolled in PHOENIX trial from approximately 430 subjects to 500 subjects as we have been making good progress on our subject enrollment. To summarize, with the excellent progress in our Phase III trials and promising interim results from Phase III Stargardt study and a full-year cash runway, we remain well positioned in advancing Tinlarebant as potentially the first oral treatment for people living with degenerative retinal diseases. I’ll now turn over the presentation to Nathan. Nathan, please.

Thank you, Tom. Here, we have an overview of our trial designs in Stargardt disease. As Tom mentioned, there are 2 Phase III trials that we’re currently involved in. The first is called DRAGON. There’s 104 subjects in that trial. The other trial is called DRAGON 2, there are 60 subjects in that trial. You can see the first 3 rows here. This is the areas where these 2 trials are different. Otherwise, the trials are designed identically. So there’s a difference in the number of sample size, as I just said, a difference in the geography. The DRAGON is a global study. DRAGON 2 is focused on geographies in Japan, U.S. and U.K. The DRAGON 1 study, because of the larger sample size, is a 2:1 randomization favoring Tinlarebant, whereas the DRAGON 2 trial has a 1:1 randomization with its 60 subjects. Otherwise, the trials are designed identically. And it’s important to note that the endpoint for drug approval in Stargardt disease and GA is slowing the growth of atrophic lesions. And at the bottom, you can see the key inclusion criteria for subjects involved in DRAGON and DRAGON 2. Here, you see the demographics and baseline characteristics for the adolescent subjects involved in the DRAGON 1 trial. 3 As I mentioned, there are 104 subjects. You can see the mean age is 15.4 years. So these are school-aged children. They have the average height and weight of children that age. On the right-hand side, you see the breakout for male and female, roughly 60% male and 40% female in the study population. And just below that, you see the race distribution, heavily favored towards the Asian population because we did heavily recruit in China. So we have approximately 56% of our Asian population representing the study, about 37% being Caucasian and European and North American and various other categories by approximately 7% to 8%. Here’s an overview of the interim analysis conclusions. As Tom mentioned, the study DRAGON 1 included a sample size re-estimation in which if the DSMB saw a trend towards efficacy at the middle of the study, then they would allow us to include up to additional 30 more patients to maintain that trend towards the end of study so that we could ensure a statistical significant difference by end of study. But in fact, when the DSMB looked at the interim analysis, they felt there was no modification of the study required and that we should continue the study without a sample size increase. I should remind you that the dosage that these children were getting in the DRAGON 1 and DRAGON 2 studies is 5 milligrams daily. This dose has been very well tolerated and deemed safe, a very nice safety profile. Also important to note that at the time of the interim analysis were approximately half of the subjects have already completed 2 years of dosing. The withdrawal rate was 9.6%, which is 10 of 104 subjects. And the withdrawal rate due to ocular adverse events was only 3.8%. That’s 4 of 104 subjects. Visual acuity was stabilized in the majority of subjects, with a mean change of baseline of less than 3 letters, so very well stabilized under both standard and low luminants throughout the 2-year study. But perhaps the most important finding that the DSMB provided for us was what’s pro4 vided at the bottom there. Additional comments, as Tom mentioned, they recommended us to submit the data for further regulatory review for drug approval, indicating perhaps an efficacy signal. And here are the safety data from the DRAGON 1 trial. These are the treatment-emergent adverse events. We fully anticipate to see 2 adverse events in terms of the drug-related ocular event. One is a form of chromatopsia called xanthopsia. This is a yellow hue of color, which appears in the visual field typically upon waking when light essentially drives of this visual AE. This is a transient AE. It’s – lasts seconds to minutes, and no one dropped out a study because of chromatopsia or xanthopsia. Delayed documentation is the other ocular AE that we anticipate based upon the mechanism of Tinlarebant action. This is the opposite of chromatopsy in which going into dark- ness, patients have a longer time to accommodate to dim like settings. This can last 2x to 3x longer than normal, perhaps somewhere between 16 to 20 minutes. Again, it’s reported as mild, it’s transient, and this is not synonymous with night blindness or nyctalopia because these subjects will eventually get back their dark adapted sensitivity. And you can see the distribution on the right-hand side in terms of the number of subjects and percentage in the patient population. The night vision impairment is a more severe exacerbation of delayed dark adaptation in which the delay may be longer than 20 minutes. You can see that occurred in 15 subjects, approximately 14% of the population. And headache was another AE that we found in approximately 7% to 8% of the population. This can happen when subjects strain to use their visual acuity while experiencing these AEs. But importantly, there was no clinically significant findings in relation to vital signs, physical exams, cardiac health or organ function, and the only systemic drug-related AE was acne, which teenage kids can be prone to especially when there’s less vitamin A in the skin. But otherwise, an overall very, very well acceptable safety profile. 5 So here, we see the visual acuity data from the DRAGON 1 study. This is the 2-year data. We’re looking at visual acuity under both standard and low alumina. We see overall stabilized visual acuity. But for a clinical perspective, let’s bring in our CMO, Dr. Hendrik Scholl, for his opinion. Hendrik?

Thank you, Nathan. When considering the clinical relevance of the finding, we have to take into account that when we measure visual acuity, the inter-session variability in normal subjects is 2 letters on an ETDRS chart. And in patients with macular degeneration, it’s up to 5 letters. And that means that the variability that we see on the left for best corrected visual acuity and on the right for low luminance visual acuity is within the standard variability that we find in such patients. Still, when we look at the left and the development of best corrected visual acuity and knowing that 2/3 of the subjects are on Tinlarebant treatment, it’s very, very assuring that was essentially no loss at all of best practical acuity letters on a standard ETDRS chart. So with that, I hand it over back to Nathan.

Thank you, Andrew. Here is our overview of the trial design in Geographic Atrophy. This is our Phase III trial called PHOENIX. As Tom mentioned, we’re going to recruit up to approximately 400 subjects. Right now, to date, we’re right at about 400. We’ve got about 100 more to go. We expect to close that enrollment by end of Q2 of this year. This, of course, is a global study, double-blind, same randomization as we had in DRAGON 1, 2 to 1 favoring Tinlarebant. It’s a 2-year treatment duration. And of course, just like in Stargardt, we’re looking for the slowing of atrophic lesion growth as the primary endpoint for drug approval. But of course, we’re also looking at BCVA, retinal anatomy by SD-OCT and retinal sensitivity by microperimetry. And like the Stargardt DRAGON 1 study, we will 6 have an interim analysis at 1 year. With that, I think I’ll throw it to Hao-Yuan for the financial results. Thank you.

Thank you, Nathan. In 2024, we had R&D expenses of $29.9 million compared to $28.8 million (sic) [ $24.8 million ] in 2023. The increase in R&D expenses was primarily due to an increase in royalty payments for the completion of the Phase II trial and an increase in share-based compensation granted in the Q3 of 2024. On G&A expenses, in 2024, G&A expenses were $10.1 million compared to $6.8 million in 2023. The increase was primarily driven by an increase in share-based compensation granted in Q3 of 2024. On net loss, we had a net loss of $36.1 million in 2024 compared to $31.6 million in 2023. In terms of cash, we had $31.7 million in cash and $113.5 million in investment by end of 2024 as compared with $88.2 million by end of 2023. The investments were in liquidity fund, in time deposit and U.S. treasury bills. One thing to note is that the net cash outflow from operating activities was $29.2 million in 2024, similar to the cash outflow of $29.8 million in 2023. We also raised $15 million in gross proceeds in the registered direct offering in February 2025. We still expect full year’s cash runway without considering the cost for a second GA Phase III study. Thank you. Back to you, operator.

— Operator Instructions — Your first question comes from the line of Marc Goodman with Leerink. Please go ahead. Moving along. Your next question comes from the line of Jennifer Kim with Cantor.

7 Maybe on my first question, starting with Stargardt. So on the DSMB’s recent recommendation, you’ve said that you plan to reach out and I guess, see harmonization across some ex-U.S. regulatory authorities. Could you outline the potential outcomes from those interactions, either positive or negative?

Thanks, Jen. I can take that. So the DSMB has recommended the interim results to be reviewed by regulatory agencies for drug approval, as you pointed out there. We believe this is a very positive outcome, and we’ll be following DSMB’s recommendations to request regulatory review and see whether the agencies confirmed with DSMB’s recommendation. So we believe that – certainly believe that the regulatory agencies will probably align with the DSMB recommendation because it’s not every day that they see – they make this kind of recommendations during interim. But if not the regulate – if not – if they don’t see it that way, then we’ll just move on and carry on with the study.

And then maybe turning to GA. What drove the decision to increase the sample size to 500 patients?

Well, we’re getting – so the GA study has been moving on very smoothly. We want to take this opportunity to enroll more subjects to further boost our chances of success. So based on this current enrollment rate, we should still be able to complete within our expected time line, which is Q3 this year. And the cost will still remain very feasible based on the current recruitment rate.

Okay. If I could squeeze one more question with GA, can you just remind me what is your latest thinking on what the interim analysis will entail and how that sort of feeds into the 8 decision to start a second trial?

You mean for the PHOENIX interim?

For PHOENIX, yes.

Well, we certainly believe that if we get any positive signals from the [ IF ] for PHOENIX, then we will speed up and expedite our second Phase III trial for PHOENIX. For GA, sorry.

Can you hear me okay now? This is Basma for Marc. Our first question is about Stargardt disease. Are you going to be able when you meet with the authorities to – regarding regulatory clarity for the submission? Will you be able to also get confirmation regarding the potential for a broad label, given that the study population so far is only adolescents? And our second question, again, can you give us an update about the current discontinuation rates in the GA trial, the PHOENIX trial?

Sure. So I’ll take the second question, and I’ll leave the first one for Hendrik. So the dropout rate for PHOENIX right now is approximately about 20%. It is very common because majority of the subjects enrolled in the GA trial are elderly population. So for pre- vious studies, we’ve seen a dropout rate of about – more than 30%. In fact, the [ dropout 9 ] rate of vitamin A study, they were just recently presented at JPMorgan, deep dropout rates is more than 30% and certainly more for imixustat and the anticomplement study. And so I’ll let Hendrik to confirm it for the dropout rates for the complement studies as well as answer your questions on the Stargardt label. Hendrik?

Yes. I’m happy to. Thank you, Tom. So what we have seen in natural history studies is that in patients that have an early onset of disease, the disease is generally more severe and shows a faster progression. The ProgStar study has shown that subjects with younger age and early on, it still show a relatively similar progression rate compared to other subjects that were older in the ProgStar study. Not very old, obviously, but it’s still a juvenile macular dystrophy but would be adults. So we believe that the threshold actually to get something approved for a pediatric population would be much, much higher. And we feel that if we can show efficacy in our adolescent population, it should be relatively straightforward to get the drug approved for adults as well.

And the – Hendrik, [ can you also ] mention about the dropout rate for anticomplement as well as for imixustat.

So in the anticomplement studies, dropout rate was in the order of magnitude of 20% to 30%. I believe in the study with imixustat, which has quite extensive side effects, the dropout rate was even higher. I would actually hand over to Nathan Mata, who actually knows a lot about that specific drug and its effect in Geographic Atrophy.

Yes, it was a little over 40%. 10

Your next question comes from the line of Yi Chen with HCW.

Just to clarify, the adolescent Stargardt disease patients being rolled currently each of the drug and trial, what – they represent what percentage of the Stargardt disease patients diagnosed in the real world?

Hendrik, I believe this is a question for you as well.

I’d be happy to take that question. So the typical Stargardt patient would notice first symptoms in the second decade of life. We see patients that have a very early onset as early as 5 years and then there are patients that are later in adulthood, develop the first symptoms. But as Karl Stargardt described the disease in 1909, right, this is a juvenile macular dystrophy, and it typically starts between 10 and 20 years of age with first symptoms. Given that our study population actually includes subjects between 12 and 20 years old, we feel that this is very representative and would show an overlap. If we look at all Stargardt patients that would schedule a visit in clinic, I would believe that, that would rep- resent 2/3 of Stargardt patients that I would see, for example, in my clinic.

So you would expect potential approval in the future for all Stargardt disease patients in that age range, right, not necessarily meeting those – meeting the enrollment criteria in your current trial, correct?

So if I understood the question correctly, the question is... 11

The future prescription label is not restricted to the patient groups you are currently enrolling into a pivotal trial.

Exactly. The answer is no. And there would be no reason why we would not prescribe the drug to, let’s say, an adult patient that developed the first symptoms at age 30 and still shows progression of these DDAF lesions, which is typical also for adult patients.

But do you expect any potential limitation on the payer side for reimbursement if the label is broader than the patients being currently enrolled in the DRAGON trial?

Well, I think we will definitely ask the regulator to try to get the label for the adults, which we think is doable. And we’ll probably just do a PK study to prove that it works the same on the adult patients.

Your next question comes from the line of Bruce Jackson with Benchmark.

First, a housekeeping question about the capital raise you did for $15 million. Have – has anyone exercised the warrants yet attached to that?

Not yet.

Okay. And then in February, you said that your CRO is going to be handling some of the 12 regulatory process for you with the data for DRAGON. So could you just give us a little bit of color on where they are with that process right now and what the next step might be?

Yes. Thanks, Bruce. So we have 2 or 3 different CROs representing us for different jurisdictions. And certainly, they have a procedure – a different procedure or – and certainly different templates for submitting these kinds of regulatory submissions. So right now, it’s in good hands, and they are basically submitting as we speak.

All right. Great. That’s it for me. Thank you.

Your next question comes from the line of Michael Okunewitch with Maxim.

Just first, what kind of difference in lesion growth between placebo and treatment is the DRAGON study powered for?

Nathan, do you want to answer this one?

Yes, it’s powered for a 40% treatment effect with 80% power to detect that effect at the second year.

Okay. Great. Thank you, and congrats on all the progress.

There are no further questions at this time. This concludes today’s call. Thank you for 13 attending. You may now disconnect. Copyright © 2025, S&P Global Market Intelligence. All rights reserved 14