Belite Bio, Inc (BLTE) Q4 2022 Earnings Report, Transcript and Summary

Good afternoon and welcome to the Belite Bio 2022 Full Year Financial Results Conference Call. At this time, all attendees are in a listen-only mode. A question-and-answer session will follow the formal presentations. [Operator Instructions] As a reminder, this call is being recorded and a replay will be made available on the Belite website following the event. I'd now like to turn call over to Tom Lin, Chairman and Chief Executive Officer of Belite Bio. Please go ahead, Tom.

Thank you, Tara. Thank you, everyone, for participating in this earnings call. I'd like to take this opportunity to give you an update on our current overview and milestones. So, we're currently 18th month into a 24-month open label Phase 2 study, which we will be presenting that data at ARVO by the end of the month. At the same time, we are also halfway, we recruited 41 subjects out of the 90 subjects into our global Phase 3 Stargardt study and we are expected expecting interim readouts by mid-2024 next year. We recently met with the FDA to request that we increased the sample size from 60 to 90. The reason being that after analyzing very promising top line treatment data from our Phase 2, we ran some simulations and we worked out that if we increase the sample size to 90, we would have more data going to the interim. And that will give us a very good chance of meeting that growth rate we can for efficacy. We've also initiated our global Phase 3 trial in geographic atrophy secondary to dry AMD, which is expected to start enrolling subjects by midyear this year. So with the interim analysis, coming up for the Phase 2 18-month data, we feel that this will probably give a very good reference to what does data study will look out for our Phase 3. Okay. Next slide, please. So, the market opportunity is that, today there's still no treatment for Stargardt disease and currently there's about 30,000 subjects in -- 30 subset of patients in the U.S. with Stargardt and currently there’s no treatment. In terms of AMD, it is still a very large population globally and this is still arising given the aging population worldwide. As you can see on the graph on the right, the graph shows that with more advance age, the more advance the AMD gets. And at the stage of AMD, we know about the wet AMD, the market about $10 billion a year. But geographic atrophy, we expect this to surpass the wet AMD market. Even though there is the entire complements that are getting approved or patents that's approved, we still think that there's still a very good -- there's still an unmet need given there is still a noninvasive form of treatment such as oral therapies which we think would surpass the market for AMD. Next slide please. So I hand now to Nathan to go through the scientific slides.

Thank you, Tom. Hello, hi, Nathan Mata here, CSO for the company. I want to talk about the mechanism of action of tinlarebant. Before I begin that, I want to preface by saying in both these diseases that we're interested in intervening Stargardt disease and geographic atrophy, they're both associated with the presence of binding byproducts which are toxic, and actually cause retinal cell death and tissue. These toxic byproducts are derived from circulating vitamin A. So what our approach is, is to reduce the entry of retinol into the eye as a means of reducing the bisretinoids, these vitamin byproducts that would call as retinal disease and thereabout our orally available once a day drug competes with native vitamin A retinol for binding to retinol binding protein 4 and reduces or limits the amount of retinol entering the eye. When this happens the bisretinoids which are formed from retinol also are reduced because there's a reduced amount of retinoids traversing through the visual cycle. In Stargardt disease, this is the primary culprit of vision loss in these patients is the formation of A2E and related bisretinoids. In geographic atrophy, these molecules accumulate as well, but for a different reason and from a different origin, it's because of a dysfunction of the retinal pigment epithelium. I'd like to now show you clinical presentations of what these diseases look like. On the top of a series of images, retinal photographs from a patient with Stargardt disease and then on bottom series of images, you have a patient with geographic atrophy. We're looking over about a similar -- same period of time roughly 4.5 years in each patient, which you can appreciate if you compare the baseline images to the last images at 55 or 57 months is that in both cases the auto fluorescent area you see around lesions, which are these bright areas of tissue you see in the surrounding, the lesion, they actually spawn the new dead retina. So wherever the autofluorescence is, that is where the lesions grow into. The autofluorescence expands centrifugally and the dead lesion, that black retina follows. So this tells us that the autofluorescence precedes lesion growth in patients with Stargardt disease and geographic atrophy and that an approach to reduce these bisretinoids by limiting retinol should be effective to preserve vision loss in these patients. Next slide, please. I'm showing now our clinical development pathway starting with our first patient population that is our adolescent Stargardt patients. This, of course, after conducting our requisite SAD and MAD studies, we established safety and tolerability of this drug. In the Phase 1b/2 study, we enrolled initially 11 Stargardt subjects and determined in those subjects in the one month Phase 1b that a 5 milligram dose was effective to drive the retinal vitamin protein level down to roughly more than 70% from baseline. We enrolled those subjects then into an open label two-year Phase 2, and we added two additional Stargardt subjects for a total of 13 subjects participating in the ongoing two-year study. I have one year data, safety and efficacy data to share with you today from this Phase 2 study. We also initiated, as Tom mentioned, begun recruiting for our Phase 3 adolescent Stargardt study. Out of a total of 90 subjects, we have recruited roughly 42 subjects to-date. This can be a randomized double-masked global study. Primary endpoint will be looking at the lesion growth rate. Secondarily, we will be concerned with our best corrected visual acuity. We are also now initiating a Phase 3 trial in geographic atrophy. This again will be a two-year study with a similar trial design that of Stargardt, in that the same endpoints are the same, same treatment duration and we will get interim analysis at one year, except this study will have 430 subjects randomized in the same manner as the Stargardt study with 2:1 favoring tinlarebant. And you can see looking forward around 2025 to 2026, that's when the earliest timing we would have for an NDA filing with our promising Phase 3 study in Stargardt. If our Phase 3 study is not that promising, of course, the FDA may require to read another clinical trial. However, they have given us words that they would take our data under review after the first Phase 3 pending really good lesion results and a stabilization of visual acuity. And then certainly after 2026 would be, when we start looking for planned completion of our GA study, and commencement of a second Phase 3 study. Next slide, please. I'd like to now share with you some preclinical -- I'm sorry, some clinical proof-of-concept data that tell us that reducing retinal delivery to the eye would have an effect on slowing lesion growth. In this particular study, this was done in GA patients. This study was done when I was a CSO for another company called Sirion Therapeutics. It was conducted approximately 13 years ago, and the data was published. The reference is shown at the bottom of the slide. In this study, I used a drug that was not designed as a retinal binding protein for antagonists, this drug is called fenretinide. Fenretinide is a synthetic derivative of vitamin A. It was developed as an anti-cancer drug, but has the side effect of reducing retinol delivered to the eye because it binds to RBP4 in the same way that our drug does. However, because it's a retinoid, and because it's a weak binder, it's not expected to have great retinol binding protein for antagonism. However, I repurposed it for that purpose anyway, just because it was the only thing available to establish a proof-of-concept to address the question would reducing circulating retinal entry into the eye have an effect on slowing lesion growth. This was a two-year study placebo-controlled with two treatment arms, 100 milligram and 300 milligram. What we found at the end of the two-year study was that patients had got to a profound reduction of retinol binding protein 4 of at least 70% or more had a statistically significant slowing of lesion growth. The data you're looking at on the right hand side, these histograms show the lesion growth rate in placebo, that's the black bars. And you can see in those subjects, the lesions grew roughly 50% larger relative to baseline. So over two years period they grew 50% more in size. Whereas patients that achieved this reduction of RBP4 of 70% or more had a 25% expansion of legion growth rate, representing a 25% treatment effect on the expansion of lesions over two years. You can see that subjects in the 300 milligram arm who did not achieve at least that 70% reduction of RBP4, did not have any meaningful change in their lesion size growth, and they are comparable to placebo. Another important point outcome of this study was that those patients who achieved that greater than 70% reduction of RBP4 also had an improvement in visual acuity in the sense that they stabilized after 12 months. If you look at this plug on the lower right hand side, you will see those dark green bars stabilized after 12 months at about 6 letters loss. Meanwhile, the placebo group and the other 300 milligrams substitute did not achieve the RBP4 reduction of 70% or more, lost anywhere between 11 to 12 letters over two years, such that at 24 months, we have a 2 line gain relatively speaking in the group that achieved that significant RBP4 reduction versus placebo. The unfortunate thing in this study was that because of the limited bioavailability and lower potency of fenretinide, only 1/3 of subjects in that high dose cohort demonstrated that significant reduction of RBP4. And again, that is largely because of the reduced bioavailability of fenretinide. We asked that this treatment -- that this drug, it is the oral drug, be taken with a high fat meal at dinner to improve absorption. Possibly 2/3 of the subjects in the high dose arm did not comply after about 12 months. And we can see that from the retinol binding profiles, they actually inflicted upwards which showed us they were not taking their drug as prescribed. But that 1/3 of subjects that did in fact comply had a very profound reduction of RBP4 and had this profound reduction of lesion growth and a stabilization of visual acuity. Our drug tinlarebant is also an oral once a day, but it has much greater bioavailability than fenretinide and a much greater potency. You can see here the comparison, tinlarebant binds RBP4 with an affinity of 2 nanomolar. Meanwhile, fenretinide binds with a similar affinity as the native vitamin A retinol at 200 nanomolar. So this means we have 100 fold greater potency, a greater bioavailability and it's not a retinoid so it has a greater safety -- a better safety profile. Next slide please. I'd like to now enter into this discussion with some of the lesion growth rate data we had from that ongoing open label Phase 2 study. But before I get there, I'd like to first describe to you what lesions are actually looking at. There are two types of lesions I showed you in a previous image. There is an autofluorescent lesion which is the first the earliest lesion that is formed in the back of these patients' eyes and again this is because of the bisretinoid accumulation. So you're looking here at a subject with Stargardt disease who has just the autofluorescent lesion, this lesion can be saved, this is rescuable tissue. It is not atrophic dead retina. But you can see that over 12 months, this autofluorescent lesion grows approximately 0.45 millimeters square per year. The point here is that these autofluorescent lesions are not stable, that is they continue to expand and they certainly never regress. Next slide please. Here is a similar image of a patient that has both lesion types, both the autofluorescent lesion as well as the atrophic retinal lesion. So if you look at baseline, that black and demarcated area with the blue outline around it, that is atrophic retinal, which ophthalmologist referred to as definitely decreased autofluorescence, that tissue is not respirable. But to the right of that tissue, that zone of autofluorescence that you see, that speckled area, that is where the bisretinoids are. That is respirable tissue. However, if you look at the 22 month image now, what you see is that lesion, that DDAF lesion has expanded tremendously from 0.82 millimeter square to 2.09 millimeter square. At the same time, the QDAF lesion size, that is the autofluorescence lesion, has shrunk relatively speaking from 1.53 to 0.45. But if you look at the data more carefully, what you can see is it's only because that auto -- sorry, the dead retinal lesion has expanded into the autofluorescent zone. It's not that the autofluorescent lesion has shrunk unnecessarily. In fact, if you look at the perimeter, it's actually expanded outward a little bit, but the dead retinal lesion has occupied its area significantly. The point being here is that the dead retinal lesion grows tremendously in compromising the autofluorescence lesion size and the dead retinal lesion, the DDAF is our primary endpoint. So with that background, I now want to go into our Phase 2 data where we're looking at these autofluorescent lesions, the QDAF as well as the atrophic lesions DDAF. I should mention that in this open label Phase 2, these 13 subjects, adolescents target subjects who are participating had no DDAF lesions at baseline. So we asked two questions. One, what is the time for transition from an autofluorescent lesion to a dead retina lesion? And two, once that dead retina lesion forms, how rapidly does it grow? We want to compare both those values to natural history. All the zeros you see at the baseline at Phase 1 and Phase 2 are because there were no DDAF lesions, no dead retina lesions. But at six months into Phase 2, you see one of the 13 subjects transition to a lesion. This was striking to us because based upon natural history from the extensive prior Stargardt studies, we would have expected 50% of these subjects to transition from QDAF to DDAF, yet we only have one of 13. At the one year time point we still have not another subject transitioning just, that same subject number 11 going from 0.32 at six months to 0.44 millimeters square bilateral lesion growth in both eyes. When we do the cohort mean for growth rate at 12 months, we get a very, very small growth rate to 0.03 millimeter square per year. In order to more faithfully compare our data to natural history, we looked at one of the prospective cohort studies of childhood onset Stargardt published by an author named Georgiou in 2020. And what they looked at here was the combined growth rate of the autofluorescent as well as the dead retinal lesions so the DDAF plus the QDAF. And they saw lesion size a 0.69, that's the growth rate for that combined lesion size. When we do that same measurement in our cohort of subjects, we only see a growth rate of the combined lesion of 0.26 representing a 60% reduction in the lesion growth rate of the combined QDAF and DDAF lesion size, quite profound. Another piece of confirmatory evidence is that if we just look at that QDAF lesion growth from the one year data below, you will see that the acuity of lesion in our subjects had a mean growth of 0.23, whereas the natural history predicts something closer to about 0.5. So here we are seeing about another 50% growth -- reduction in growth. So we believe we are seeing two things, two positive outcomes: One, a slowing of the transitioning from the autofluorescent lesion to the dead retinal lesion. And two, once that dead retinal lesion forms, we are seeing a slowing of the growth of that lesion. And both of these effects are consistent with the MoA that I described to you earlier. Next slide, please. I'm now presenting to you some of the safety data from the one year Phase 2 open label study in adolescent Stargardt subjects. I'll start this by telling you that there have been no systemic safety AEs whatsoever. No severe AEs or SAEs reported and no AEs required discontinuation from treatment. Furthermore, there have been no clinically significant findings in relation to vital signs, physical exams or cardiac health. What we see are two anticipated features of the drug, which we want to see because they are telling us, we are having the intended biological effect on the retina. The first is called chromatopsia, which is an aberration of color vision. This happens when patients transition suddenly from a very dim environment or a very bright environment. This activates cone photoreceptors and they will require chromophore to mediate that light response, because chromophore will only be slowly supplied to cone photoreceptors. There is a delay in the time for them to adequately respond to bright light, and they will show -- they will misfire and electrically present into the visual field artificial hues of color. In this particular study, we are seeing more reports of Xanthopsia which is a yellow hue of color in the visual field. It has been reported as mild and transient, and all patients are dealing with it quite well. No one is really complaining about it. Because once patients understand how to transition from lighting environments to mitigate the severity of these AEs, they can manage them themselves. Same thing for delayed dark adaptation. Delayed dark adaptation is the manifestation of rod photoreceptors. These are photoreceptors in your retina that need a dim light vision. So when patients transition from a very bright light to a very dark environment, that activates rod photoreceptors. And once again, chromophore will only be slowly supplied from our photoreceptors during tinlarebant treatment, the delay in the timing to fill up those rod photoreceptor so they become maximally sensitized to the light is called delayed dark adaptation and it's on the order of 5 to 10 minutes. Most patients who have this disease as well as geographic atrophy have delayed dark adaptation. So they are mostly asymptomatic because they kind of discern the pharmacological addition of delayed dark adaptation on top of their own intrinsic disease caused delayed dark adaptation. Night vision impairment is a more severe exacerbation of the DDA. And this particular subject lasting out to 20 minutes, that is the delay and the ability to accommodate to dim light, and the increasing error score on the FM100 is a more severe manifestation of this Xanthopsia. So, again, a more prolonged manifestation of that theme of covered individual field. But once again, no subjects have less study because of these, and we want to see these AEs because they are telling us we are having the right biological effect on the retina. This is a pharmacologic profile to show you the pharmacodynamic effect of our drug on retinal binding in protein 4. Again, this is data from the Phase 1b, our dose finding, a phase where we have the 11 adolescent Stargardt subjects. You can see here a 5 milligram daily dose drives the retinal binding protein for down to more than 70% over at least a 3 to 4 day period. We are not showing you all the data points here. But by the third day of dosing, they are actually at 70% or more reduction and they stay reduced as long as we give them daily dosing, until we withdraw the drug. And then during that period of drug cessation, you can see a very nice rapid reversibility of the pharmacodynamic effect. So we see a very rapid onset for reducing RBP4 and a very rapid offset. Having a nice reversibility effect is a good thing to have should there be any untoward effects of the drug from -- during long-term treatment. This is now the clinical trial design overview for our Stargardt study. In the middle you have our Phase 2 trial design, which I've already described to you, so I won't belabor that point. But if you look to the far right, this is our DRAGON study, our Phase 3 study for Stargardt. I mentioned before that are open label Phase 2, none of these subjects were required to have DDAF lesions. But in our Phase 3 pivotal study, they will be required to have that because this is the endpoint and we will need some measure of lesion size at baseline to compare subsequent lesion growth rate over the ensuing two years. Of course, it will be a global study double-blind in nature, Two-year randomization favoring tinlarebant as I mentioned, two-year duration with one year interim analysis, and we'll be looking at the same efficacy measures that I told you about, looking at the dead retinal growth, the DDAF lesion growth as the primary endpoint. Secondarily, we'll be looking at autofluorescence that is the QDAF lesion, also be looking at vision and -- as well as a retinal anatomy by spectral-domain optical coherence tomography. And we'll also be measuring light sensitivity of the retina by microperimetry. I mentioned there is an interim analysis of one year. And at the bottom you can see the key exclusion criteria. We will have very specific lesion size cut offs. And that's because prior experience has taught me that lesions of a larger size do not respond to these early intervention therapies because the disease has gone too far. I will now explain -- I'd provide for you our trial designs for the Phase 3 study in geographic atrophy. Before we began designing this study, we wanted to make sure that we had the right dose. Subjects who are elderly GA, subjects are typically larger, they're heavier, higher BMI, of course they are older age. And natural history shows that those two things predict a higher RBP4 level in blood. So I was concerned that we would have to use a higher dose above 5 milligram to achieve the same pharmacodynamic effect that we saw in the Stargardt subjects, we did not have that problem. So here you're seeing the pharmacodynamic profile from a 5 milligram dose in these older substances with higher BMI and higher age range. And you can see we get about a mean reduction by 80%. That was exactly what we saw in the Stargardt subjects. And then when we withdraw the drug over 14-day period, you can see a return of that RBP4 value back toward the baseline value. This is our clinical trial design overview for geographic atrophy. Geographic atrophy has the endpoint that we're at slowing the lesion growth rate. So both in Stargardt disease and in geographic atrophy, we are looking at the exact same endpoint with the exact same imaging modalities that I showed you earlier, those retinal imaging that shows you the autofluorescence as well as the dead retina. We will be targeted patients with small lesion size for the reason I just said earlier. So this is another differentiator for our treatment effect. In addition to oral intervention once a day, we're looking for patients with early stage disease, because these molecules that we are targeting actually are the earliest incipient molecules that start retinal atrophy. So we believe that if we can get to these patients early enough, we can actually halt their disease process and they would never be at risk of losing any vision loss as long as we get them at the right stage. We'll be doing this both for Stargardt disease as well as geographic atrophy. And of course, there's broad potential for this being an oral once a day therapeutic that is really going after the earliest causes of diseases that we believe this could be going into intermediate stages, such as intermediate AMD. We'd have to work out the biomarkers for efficacy there. But there certainly is a pathway going forward. And as I said before, we think in general with these chronic diseases where you have to have treatment for years to decades of your life, an oral once a day therapeutic will be a much more tractable approach than for instance an injectable interventional therapeutic into your eye. Next slide, please. Okay. Hao-Yuan.

Thank you, Nathan. So, in 2022 our R&D expenses increased $1.5 million from $7.4 million in 2021 to $8.9 million in 2022. The increase was primarily due to an increase in wages and salaries due to our R&D team expansion and increased share-based compensation expenses. Our G&A expenses increased by $1.6 million from $2.4 million in 2021 to $4 million in 2022, which was primarily due to an increase in professional service fees and also an increase of our D&O insurance expenses and increase of wages and salaries. In terms of cash, we received $38 million IPO net proceeds in 2022. And as of year-end of 2022, we have cash of $42.1 million, which can last until the end of 2025 and it's enough for us to complete our Phase 2 and Phase 3 Stargardt trial if we do not spend any money on GA. However, considering the promising data that we have seen in the prior time Phase 2 study, our old Phase 2 study and market potential of GA, we do want to start a GA Phase 3 trial with the limited contractual liability in certain countries that we've already received broad interest from the PI and the size, such as in the U.S. and Australia. And with that, we will have cash runway until end of 2024 and be able to complete Phase 2 Stargardt disease trial and obtain the interim result in Phase 3 Stargardt disease trial. And we expect to only expand GA study to more countries when we secure more funding. With that, I'll turn it back to Tom.

Thank you, Hao-Yuan. So the key anticipated milestones for the year. In Q1, we've initiated the PHOENIX Phase 3 study in GA. We've also enrolled 42 subjects into the DRAGON study. And I believe that by I think this month there's probably about 6 to 10 subjects that qualify for the study. And so that will bring you up to 50 by the end of the month. For Q2, we have an ARVO presentation that we will be presenting the 18 month treatment data for the Phase 2 for Stargardt disease. And shortly after that we have a KOL event to discuss this 18-month Phase 2 efficacy and safety data for Stargardt disease. We also expect to initiate enrollment for the PHOENIX Phase 3 study in GA by late Q2. And then in the second half of the year, we'll have the top line 24-month Phase 2 treatment data, which will be expected efficacy and safety data for the two-year study. We will complete the enrollment for the DRAGON study by the end of the year. Hopefully, we'll work -- with the recruitment rate picking up, we should be able to reach that target enrollment by the end of the year. With this, this ends my presentation and I hand it off to Tara to moderate the Q&A.

[Operator Instructions] So our first question comes from Basma Radwan from SVB Securities. Please go ahead, Basma. Basma, you might be on mute? All right. We'll go to the next question. So our next question comes from Jennifer Kim from Cantor Fitzgerald. Please go ahead, Jennifer.

Hi. Thanks for taking my question, and this is very helpful update. And thanks for that. I have a few questions here. First, I know you talked about the reasoning for increasing the size of the Stargardt trial to 90. I just want to clarify, is that to increase the probability for a successful outcome in the study overall? Or is this thinking that it could increase the probability of a positive readout at interim? And then related to that, does the mid-2024 interim assume -- I know you said enrollment completion by the end of the year. But can you remind me of that mid-2024 interim, is that a one year interim readout based on the total patients? Thanks.

Yes. So thanks, Jennifer. So the first question was on -- sorry. You asked three questions over there. The first question -- what was the first question, may I ask?

Just the rationale behind increasing size of trial, yes.

That's right. Yes. So it's actually for both. So number one, that increases probability for interim. And for the final, if we missed out on the intro, but we still have a positive trend that gives us a pretty good opportunity for the final analysis as well. So actually, it's for both.

Okay. And then the timing of the interim in mid-2024. How does that work with -- so is it the enrollment completion is year end, and then the interim...?

Yes. So we would have majority of the patient completing 12 month by that time. I know some patients will be in the six month -- will be six month. But the first quarter or 1/3 of patients would have 18 month data by then. So on average, we will have 12 month data, majority of patients, yes.

Okay. Great. And then my second part of the question is for PHOENIX. Is there a way we can think about, I guess, the pace of enrollment when you start out in those specific targeted countries? And then how that will ramp up once you, I guess, expand into other sites? Or in terms of numbers, like, is there a way to think about that?

Yes. So we started off in Europe, recruiting patients in Europe. So the first batch or the bulk of patients come from Europe.

So she is asking about PHOENIX. So it's a GA study, not a Stargardt.

Oh, sorry. Sorry. So what's the question again? Regarding GA study?

Since the GA study, I think you said you'll target enrollment in specific countries and sites. And then later down the line, you'll expand it to other sites. So is there a way to think about, yes, enrollment before and after?

So we started -- so it’s a global study, right, where we are starting with U.S. And then we will look ourselves to go hopefully to other sites as we complete our financing. So we can have open about 50 sites globally.

And then of those 50 sites, how many sites come from I guess that first -- those first steps into enrollment?

I think it's evenly spread. I don't have the list of sites right here in my hand, but I think its evenly spread between U.S., Europe and Asia.

Our next question comes from Lachlan Hanbury-Brown from William Blair.

This is Lachlan for Tim Lugo. We noticed that you've recently increased the upper age limit for enrollment from 18 to 20 years in DRAGON. I was just wondering if you could talk about the rationale for that, I guess also the rationale for how you chose the original age bracket of 12 to 18 and then why you've expanded that?

Yes, so -- Nathan, you want to take the question?

Yes. Let me just go ahead and take that. So initially, Lachlan, the age range was really sort of capped off by the ODD designation and the Rare Pediatric Disease designation because they categorize adolescents of up to 18 years of age. And so for those designations, we kept that age restriction above for the Phase 2, the open label Phase 2. But after going to the agencies of ODD and the RPD office, we found out that as long as you don't completely transition outside of your adolescent patient range to just primarily get adults and try to get an indication for adults, then you're fine, you still have the ODD status. So the real rationale for going to the -- adding the 19 and 20 years is from investigators that have reached out to us and told us that they have searched at their sites that would meet all qualification criteria for inclusion, except the age range. So we wanted to do two things. One is to satisfy these investigators and get those patients into study. And two, it helped us reach that additional 30 subjects we wanted to get to, to go from 60 to 90 and it didn't hurt any of our orphan drug or RPD designations in the FDA or EMA.

So our next question comes from Yi Chen from H.C. Wainwright.

Just to clarify, could you give us some additional coloring on the enrollment criteria in terms of lesion size QD versus DD in the Phase 3 Stargardt disease trial? And also, could you give us the same for the upcoming GA Phase 3 trials as well in terms of enrollment criteria of the lesion size?

Yes. Good question, Yi. Thank you. I will go ahead and take this. So for the Stargardt disease study, actually for both studies, what we know, again, based upon prior clinical experience is that lesions above a certain size and that roughly is greater than 10 millimeter square, are not that responsive to this type of treatment. We knew that from the fenretinide study. We also knew that from the recently failed tinlarebant study with a mix of stat where they found that their Stargardt disease patients didn't have any treatment effect. But when they did, a post-hoc study looked at patients that came in with smaller lesions at baseline, they had a 40% treatment effect. So we know and there's other published data that show that lesions that are larger, grow slower. I know there's some contention in the field about if that's actually the truth, but I'm convinced by the data I've seen in the literature that larger lesions whether you have Stargardt disease or you have geographic atrophy, they just grow slower. So we're capping off the lesion a small size of 0.05 millimeter square and the upper size would not be any greater than 8 millimeter square. Those will be the size ranges for DDAF lesion. We have no requirement for QDAF lesion size whatsoever.

Okay. So are the patients in the ongoing Phase 2 trial -- is the requirement the same for the patients in ongoing Phase 2 trial and the Phase 3 DRAGON trial?

No. In fact, that's one of the big differences is that those adolescent Stargardt subjects in Phase 2 did not have any DDAF lesion at baseline. And we're sort of -- again this was more proof-of-concept. So we want to see if our drugs have an effect on the conversion from the autofluorescent to the DDAF lesion and also, what is the growth rate of the lesion once it actually happens. In Phase 3, these subjects will be required to have a measurable DDAF lesion of at least 0.05 millimeter square in order to enter the trial, again, because we have to have a baseline measure of atrophy as a sort of a reference point to measure subsequent lesion growth.

I see. And what's the requirement for the dry AMD patients?

It’s going to be the same lesion size range.

Our next question comes from Bruce Jackson from Benchmark.

I have a follow-up question on the PHOENIX trial, the Phase 3 for AMD. You're going to start enrolling during Q2. How long do you think it's going to take to enroll 430 subjects?

So my prior experience tells me 18 months to 24 months.

Our next question comes from Yuan Zhi from B. Riley.

Thank you for taking our question. I have three of them. First, Nathan, can you share additional color regarding the change of enrollment for Stargardt disease? What was the assumption you used for the statistical analysis?

Yes, so for that, you mean for the initial 60% or 60, that one?

Yes, so from 60 to 90 patients, and you mentioned that the interim analysis will have a better chance to reach statistical significance.

Right. So when we initially targeted 60 patients for enrollment, the treatment effect that was postulate, there was a 35% treatment effect. We felt that was a little bit too robust, given, so -- even though we saw very positive data at six months and one year, what we wanted to do was really compare and be more conservative. And so we used the treatment effect that we saw from the Stargardt -- sorry, from the Phase 2 GA study, which is roughly about a 25% treatment effect. So with that treatment effect, a lower treatment effect again being more conservative and optimistic, even though we believed we're going to have something greater than that, we wanted to power our study with much greater power. And so we posited a lower treatment effect, which of course triggered a higher sample size, a larger sample size to meet the power of at least 80% with 90% confidence, and, of course, an alpha of 0.05.

That's very helpful. And the second question. Hao-Yuan, can you run us you're thinking around the strategy to develop GA programs?

Well, I think in terms of the strategy, we will be more looking for our income result on the Stargardt trial, and then we will be open to maybe find a partner for this entire pipeline. I think it's going to be art that you separate the two indications in Stargardt, remains the same drug. So it more likely does -- if we find a partner, they're going to take over both indications. And I think for the potential of the second GA trial, that probably will be run and paid by that pharma.

Got it. And my last question is regarding the FDA's usual requirement announcing safety. Can you remind us, how many patients are required to be enrolled in the Phase 3 trial for GA, of course, adjusted by the duration of treatment?

So this question is really a on a case-by-case basis. The hard sort of log if you will that the FDA takes, is they have a hard requirements for the safety database. So they are required for a safety database which is really the only requirement you have to have in addition to showing statistical significance in your trial. So the safety database that agency wants at least 300 subjects at or above the intended clinical dose for at least one year. And that doesn't mean, it has to be your indicated patients. They can be any patients with any ophthalmic disorder just as long as they have some ophthalmic disease. So you can use those patients to establish a safety database. In terms of what you put into your Phase 3, the agency does not ever opine that maybe you don't have enough or maybe you have too little -- or sorry too much, because it really is a sponsor's decision. You have to understand what the agency is looking for in terms of statistical significance and the safety database. So those are the two requirements that we look to, and everyone should know that we are a very cautious and conservative company. All of our trial designs are vetted before the agency in Type C meetings, before we solidify and move forward towards recruiting.

Great. Thanks for the questions, Yuan. This concludes the question-and-answer session for today's call. I'll now turn it over to Tom for closing remarks.

Thank you, everyone. Thank you for your time and your interest in the Stargardt and AMD program. We will look forward to everyone, keeping up to-date with our data, of our presentation at our KOL meeting. Hao-Yuan, and [Tara] will send out the invite and the information regarding these two events. And thank you again. Thanks.