Belite Bio, Inc (BLTE) Q1 2023 Earnings Call Transcript

Belite Bio, Inc (NASDAQ:BLTE) Q1 2023 Earnings Conference Call May 11, 2023 8:00 AM ET

Company Participants

Tom Lin - Chairman and Chief Executive Officer

Nathan Mata - Chief Scientific Officer

Hao-Yuan Chuang - Chief Financial Officer

Conference Call Participants

Jennifer Kim - Cantor Fitzgerald & Co.

Timothy Lugo - William Blair & Company LLC

Yi Chen - H.C. Wainwright & Co, LLC

Bruce Jackson - The Benchmark Company

Yuan Zhi - B. Riley Securities

Operator

Good morning, and welcome to the Belite Bio Quarter One 2023 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. As a reminder, this call is being recorded and a replay will be made available on the Belite Bio website following the conclusion of the event.

Before we begin, I'd like to reference you to the forward-looking statements and legal disclaimer slide on the screen. Joining us on today's call is Belite Bio’s Chairman and Chief Executive Officer, Dr. Tom Lin; Chief Scientific Officer, Dr. Nathan Mata; and Chief Financial Office, Hao-Yuan Chuang.

With that, I will now turn the call over to Tom.

Tom Lin

Thank you, Tara. Thank you, everyone, for joining our financial results for first quarter 2023. Joining me today is our CSO, Nathan Mata, who will be updating our recent clinical trial progress; and our CFO, Hao-Yuan, who will be updating our Q1 financial results.

For those that are new to Belite story, the company is developing a novel oral once-a-day treatment that potentially slows disease progression in Stargardt disease and Geographic Atrophy in dry AMD. There is still a significant unmet need for both indications as currently there is no approved treatment for Stargardt disease. And there are currently no approved oral treatments or no approved non-invasive treatments for Geographic Atrophy, which is predominantly in an elderly patient population.

We are currently in global Phase III for both indications. And so far, we have been granted fast track designation, rare pediatric disease designation, and orphan drug designations.

As mentioned, we are currently in Phase III development, and we have composition of matter patents until at least 2035. And with patent term extension and new patents to be filed, which will take us well into the 2040. So we still have a very long patent life on this drug.

So for the Stargardt indication, the Phase III is already two-thirds enrolled with estimated interim readouts by May 2024. We've also just recently presented a very promising 18-month treatment results from our Phase II, which our CSO will be presenting the results later on. We also expect to have the 24-month final data readouts for the Phase II study in Q4 this year. For the Geographic Atrophy in dry AMD indication, the Phase III is expected to start enrolling subjects in mid-2023.

And now I would like to pass it on to our CSO to give an update on our clinical trials. Nathan?

Nathan Mata

Yes. Thanks, Tom. Hello, everyone. I'm Nathan Mata. So as Tom mentioned, this is oral once-a-day treatment. The drug is called Tinlarebant. This drug is a retinol binding protein for antagonists. So our approach is to use this drug to limit the delivery of retinol to the eye as a means of reducing the toxins bisretinoid by byproducts that have been implicated in disease progression of both Stargardt disease and advanced dry AMD.

So in both of these studies, we are trying to slow the growth rate of retinal lesions. That is the FDA primary accepted endpoint for both Stargardt disease and Geographic Atrophy. What you are looking at in front of you are the trial design overviews for our ongoing open label Phase II, where I'll share some 18-month data with you.

In fact, we've had interim analyses that’s periodically six months throughout the study and we published – or sorry presented those data at International Ophthalmology Conferences, AAO and ARVO. And we just recently presented the 18-month data at ARVO.

You can see here the overview of the study and the Phase II outline 13 subjects. These subjects came in with no atrophic retinal lesions, but they had very prominent autofluorescent lesion, which are lesions that precede the atrophic lesion growth in Stargardt disease. These subjects came predominantly from Australia and Taiwan. It is an open label study, as I mentioned, a two-year in duration.

We are looking primarily at safety and tolerability. We've already established the optimal dose at 5 milligrams daily. In fact, that dose is effective to achieve the same pharmacodynamic response in both adolescent Stargardt patients and elderly healthy patients. I'll show you some of that data as we move forward.

We are looking at the growth rate by an instrumentation called fundus autofluorescence fundus autofluorescence photography. This allows us to visualize both the atrophic lesion, which of course, we are trying to slow as well as the precedent atrophic lesions which are, as I said, precede the growth of atrophic lesions and, in fact, are where the bisretinoids are sort of residing. You can see the key inclusion criteria at the bottom.

The Phase III study that we call DRAGON is actually enrolling. We are up to about, I believe 58 patients out of 90 that we are targeting. These subjects will all have DDAF at baseline because again this is a pivotal study. It will have a placebo control and we need some reference point for the endpoint, which is of course slowing the growth of the atrophic lesions.

Of course, this will be a global study of Stargardt being an orphan disease, hard to find these subjects, especially since we are going after adolescent subjects, not adult subjects. So thankfully, the enrollment is going quite well at this point. You can see the randomization here, two to one favoring tinlarebant, two-year treatment duration with a one-year interim analysis. And of course, we are looking at the same efficacy measures and safety measures as we looked at in Phase II.

The inclusion criteria are a little different. You can see there at the bottom. We've increased the upper age range to 20. We've increased – we've also put in a size stipulation for the lesion size. So there is an upper end of the lesion, but that's not because to learn about what work with larger lesions, covered in a two-year study that gives us the best opportunity to observe a treatment effect versus placebo.

Next slide, please. So as I mentioned before, this drug targets retinol binding protein 4, which is the sole carrier for delivery of retinol from the liver to the eye. Important to note that this protein is not needed for retinol delivery to other tissues of the body because they don't require RBP4, they lack a receptor that the eye possesses in sufficient quantity. So the eye has this pendants on delivery of retinol, bound to RBP4. This is why this is a fairly direct site-directed approach.

You can see here that over periods of dosing, in the ongoing Phase II up to 18 months, we are getting a sustained reduction of retinol binding protein 4 of at least 70% or more. The reason we target this particular dose is because of data from a precedent clinical study in Geographic Atrophy where a different RBP4 antagonist was used in those patients – those subjects it was found that subjects who got to at least the 70% reduction or more had a slowing of lesion growth. And I'll share that data with you as we move forward.

And as I mentioned before, this dose, daily dose is 5 milligrams, and you can see here approximate 80% reduction, mean reduction of RBP4 throughout the treatment trial. There's no tachyphylaxis, that means rebounding of the pharmacodynamic effect.

Next slide. I want to now share with you some of the lesion data from our 18-month open-label study, the Phase 2. But before we do that, I need to sort of give you an orientation for what we are looking at. We are looking at two different lesion types. The first one you can see here on the left is known as a questionably decreased autofluorescence lesion. You can see the boundaries they are outlined in blue.

This is what ophthalmologists refer to it as QDAF lesion. It's visualized here as you can see as I mentioned by fundus autofluorescence imaging. The important thing about these lesions is that they maybe amenable to rescue because the tissue is not atrophic yet. It is simply filled up with all of these autofluorescence bis-retinoids compound, so it's on the verge of converting.

But if we get to it early enough, it's possible that we maybe able to slow its conversion into dead retinal tissue and we'll talk about that more in a moment. And now as we transition to the right, you can see how that autofluorescence lesion converts into an atrophic retinal lesion. This is what ophthalmology referred to as definitely decreased autofluorescence. I refer to it as dead retina because that in fact is the indication of irreversible loss of photoreceptor cells. There's no coming back from that.

So when investigators look at the growth rate, they look at it in a couple of different ways. One way is to look at the aggregate growth of both the autofluorescence and atrophic retinal lesions. That's referred to as decreased autofluorescence or DAF. Another way to look at the lesion growth is independently the autofluorescence lesion growth versus the dead retina growth and of course, the dead retina growth is what the FDA is most interested in, in terms of a primary endpoint for efficacy.

We look now to the literature to see if this data, the growth that we are seeing in our subjects is comparable to or less than or greater than something that's published in natural history. And we look at a recent study published by Georgia in 2020, where they quantified in 53 adolescent subjects, the growth rate of the aggregate that is the DAF lesion, the autofluorescence plus the atrophic, they see a growth rate of about 0.7 millimeters square per year.

When we do that same analysis in our cohort at 18 months and annualize the data, we see a growth rate of about 0.28 millimeter square per year. This represents approximately a 60% reduction in the growth rate of the aggregate retinal lesions, both the autofluorescence and the dead retinal lesions, again compared to natural history.

Next slide, please. We now look to comparisons versus ProgStar. So if you are not familiar with ProgStar, this was an international effort by global retinal specialists and ophthalmologists to better characterize the natural history of disease progression in Stargardt patients. They looked at both adolescent and adult patients, but it was skewed more heavily towards adults.

What we have in our cohort, our Phase II cohort are subjects that are 18 years or less and they have no DDF baseline. Well, it turns out subjects with that exact baseline characteristic, we are present in the ProgStar cohort. And when we look at the growth rate of lesions, both the aggregate lesion, the DAF shown on the left and the atrophic retinal lesion by itself on the right, we see the growth rate fairly linear. That's the gray or blue line, you see, that's the ProgStar data roughly anywhere from about 0.8 to 1 a millimeter square change of those retinal lesions per year.

And then we compare that to our data shown in the red line here, the DAF lesion on the left, the DDAF lesion on the right. And you can see if you look at the trajectory, there's roughly about a 50% reduction in the growth rate of the combined retinal lesion and separately the atrophic retinal lesion compared to ProgStar. So in the previous study, we had a 60% reduction in growth compared to natural history. Here we are seeing at about a 50% reduction, looking at both the primary endpoint measure as well as the aggregate lesion growth rate measure.

Another important point to measure is that not all of these subjects converted to an atrophic retinal lesion. In fact, the majority did not. Seven of 12 subjects did not develop any atrophic retinal lesions at the 18-month time point. That's roughly 60% of the cohort.

Next slide, please. Now we are looking at the Visual Acuity Data. We are very happy to see this actually. This shows a stabilization of vision. Most ophthalmologists who see adolescent Stargardt patients tell us that we should be seeing a more significant loss of vision just based upon the natural progression of the disease. Here we are seeing a stabilization, we are not losing any more than about three letters over 18 months, which is well within the test retest variability of BCVA. So really there's nothing to be said with respect to BCVA other than we are stabilizing it, which is always a promising treatment trend.

Next slide. Now we are talking about the safety data. So again, this is 18-month safety data. We look systemically for AEs, of course, in these subjects, because this is an oral drug and you would expect to be of systemic side effects, but in fact, we see no systemic side effects whatsoever. No severe or moderate drug related AEs reported, no AEs requiring discontinuation from treatment and no clinically significant finding signs in relation to vital signs, physical exams or cardiac health, very important point.

And as I said before, this is a fairly sight directed approach. So what we are seeing are primarily two ocular drug-related adverse events, which we want to see. These are anticipated features of the drug and they tell us we are having the intended biological effect on the retina. The first is a form of chromatopsia, which is an aberration of color vision.

This happens when patients transition suddenly from a very dark environment to a very bright environment. This activates cone photoreceptors in your retina. Cone photoreceptors mediate bright light and color vision. So when you stimulate them with bright light or color, they want chromophore. And because under our treatment regimen that chromophore will only be slowly supplied to the photoreceptors, there's a delay in their timing to fill up and they will misfire during this period and produce these transient use of color in the visual field. In this case, yellow has been most frequently reported by our patients, but you'll note most of the subjects are reporting this as mild. No one's leaving study because of it, and of course, it's transient. So it goes away after several minutes.

The other manifestation that we are noting is delayed dark adaptation. Again, accepted and anticipated feature of the treatment. This happens, and this is mediated by rod photoreceptors, which mediate dim light vision. So when patients transition suddenly from a very bright light to a very dim light, this activates rod photoreceptors. Of course, they'll need chromophore just as cones would in the bright light. They don't get it, and they'll misfire during this period. They will not have their full dim light accommodation. So patients will wait several minutes before they can actually accommodate to that dim light. We have mitigating measures for both of these AEs. They simply are to slow patients transition from extremes of lighting environment.

So in the instance of Chromatopsia, you slow your transition from the dark to the light. And in the instance of delayed dark adaptation, you slow your transition from light to dark. This is particularly important for elderly patients in our GA study. But again, all these AEs are mild and well-tolerated with respect to delayed dark adaptation. This is a manifestation of the disease process. So most patients can even discern the pharmacological added delayed dark adaptation on top of their own intrinsic disease cause delayed dark adaptation.

Next slide, Hao-Yuan. So I mentioned earlier that we had clinical proof-of-concept from a different retinol binding protein 4 antagonist. I call this a surrogate molecule. I did this study when I was with another company approximately 12 years ago, trying to answer the question, would reduction of retinal delivery to the eye reduce lesion growth rate in patients with either Stargardt disease or Geographic Atrophy? At that time, we didn't know much about Stargardt's disease. There was no clinical – sorry, regulatory path toward approval.

But for Geographic Atrophy, there was. And in fact, that study was easy to do because there were lots of patients available. The drug used is called Fenretinide. This drug was chosen because although developed as an anti-cancer drug, it has a side effect of reducing retinol binding protein 4 in the circulation. So as shown here, this is a synthetic retinoid. It was not designed as a retinol binding protein 4 antagonists. It was designed as an anti-cancer drug. So I repurposed this drug in a Phase II proof-of-concept study with 246 patients that had Geographic Atrophy including a placebo control. So there was placebo and two dosing arms, a 100 milligram arm and a 300 milligram arm.

I'd now like to show you the lesion growth data. Over the duration of that study, we are not showing the 100 milligram cohort because there was no treatment effect whatsoever. You can see in the black bars here, the placebo growth from baseline grew 50%. So basically a 50% increase in this area of the size of the lesions in those patients, they got the placebo treatment. In the 300 milligram group, there was something very interesting. There was a group of subjects that did not achieve a reduction of retinol binding protein 4 of 70% or more. Those subjects showed the same growth rate as placebo.

But in the 300-milligram group subjects who achieved at least a 70% reduction of retinol binding protein 4 or more had this profound reduction of lesion growth, approximately a 25% reduction over the two-year study. I do get often asked from time to time, how did we find out that this treatment effect occurred? Actually, at end of study, there was no treatment effect whatsoever, but in the interim analysis, we noticed a treatment effect. And I wondered, where did it go. And so I started looking at the pharmacodynamic data and we saw that there was an inflection of retinol binding protein 4 upward at the 12 month point. A lot of these patients either stopped taking it or it stopped having bioavailability effect in these patients.

So the other point I want to point out is that at the 12 month time point, when we look at the Visual Acuity loss, Hao-Yuan. We see a stabilization in those patients who had a treatment effect against lesion growth. So you can see here in those patients who had at least a 70% reduction of retinol binding protein 4, they had a slowing, right about the 12-month time point, they didn't lose any more than six letters. In the meantime, placebo and those patients in the 300-milligram group that did not achieve that large RBP4 reduction, they lost as much as two-lines relative to the other group, the responding group. So this is roughly a two-line gain, it's approximated exactly based upon. I'm looking at the difference between placebo and the treatment arm. We have a 25% slowing of lesion growth and a stabilization of vision representing a two-line gain at the end of this study.

As I mentioned before, it's unfortunate this study only one of every three subjects actually met this requirement for the retinol binding protein 4 reduction. And we believe that's due to largely the lowered bioavailability of the drug. We asked this drug to be taken with a high fat meal. A lot of elderly patients don't want to comply with that. So we didn't get a lot of good bioavailability exposure. The other problem was of fenretinide is not very potent. It has the same potency for the target as does the native ligand retinol.

Our drug Tinlarebant overcomes all of the negative aspects of fenretinide. First, it has much greater potency, a 100-fold greater potency than fenretinide. Two, it has much greater bioavailability, so it's much more water soluble. Patients won't have to take this drug with a high fat meal. And thirdly, maybe even more importantly, it's not a retinoid. So it has a much cleaner safety profile, and be much more tractable for long-term chronic dosing in diseases like Stargardt disease and Geographic Atrophy.

Speaking of Geographic Atrophy, let's talk about our Phase III study design in GA. I mentioned earlier that the five milligram dose was effective to achieve the same pharmacodynamic response in both adolescent subjects as well as elderly healthy adults. You are seeing here a pharmacodynamic profile obtained from adults to match the higher BMI and higher age range of geographic atrophy patients. These patients are taking – these volunteers took five milligrams daily. And you see a very nice profound reduction of retinol binding protein 4 following the initial doses. Its stays reduced during daily dosing until we withdraw the drug. We didn't show you that in the adolescent patients, but it's true for them as well. When we withdraw the drug, the retinol binding protein 4 rebounds back toward the baseline, which is a nice effect to have the reversibility of the pharmacodynamic effect, should there be any untoward effects with regard to safety or long-term dosing of Tinlarebant.

Next slide. So this is the clinical trial design overview for the Phase III GA study that we call PHOENIX. This endpoint is going to be exactly the same in Stargardt disease. The duration is the same, same primary endpoint, same imaging modalities. The only real difference between these two studies, of course, is the indication, that being GA versus Stargardt disease, and of course the size.

We are looking at approximately 430 subjects to be targeted. We are actually starting that enrollment – actually starting this month and next month. This will be a global double-blind study. Same randomization as we saw in the Phase III Stargardt study, 2:1 Tinlarebant – favoring Tinlarebant rather. As I mentioned, two-year treatment duration, looking at exactly all the same primary efficacy measures that we looked at in Stargardt. And of course, there will be a one-year interim analysis.

So with that now, I think I'll throw it over back to Hao-Yuan, so he could talk about the Q1 2023 financial results. Hao-Yuan?

Hao-Yuan Chuang

Thank you, Nathan, and thank you, everyone for joining this earnings call. So for Q1, on the income statement, our R&D expenses were $5.7 million compared to $0.9 million for the same period in 2022. The increase was primarily due to an increase in expenses related to the DRAGON and PHOENIX trials, and an increase in wages due to our R&D team expansion.

Our G&A expenses were $1.2 million compared to $0.2 million for the same period in 2022. The increase was primary due to increase in professional service fee, insurance premium for D&O liability insurance and wages. In total, our net loss was $6.9 million compared to a net loss of $1.1 million for the same period in 2022. As of the end of March, we had $37.8 million cash, and we expect this will take us to the end of 2024.

Now just to recap, our key milestone include that we have initiated the PHOENIX study in Q1 and we have announced the 18-month data from the Stargardt disease Phase II on April 25th on ARVO. Currently, we have enrolled 58 subjects for the DRAGON study, and we expect to enroll the first patient for PHOENIX study around mid this year.

In the second half of 2023, we expect to have complete the Phase II Stargardt disease study with the 24-month data. And we expect to complete the enrollment for the drug and study as well. In the first half of 2024, we expect to have the interim result for the DRAGON study.

This concludes our earnings call presentation. And now I would like to turn to Tara for Q&A. Thank you.

Question-and-Answer Session

Operator

Great. Thanks, Hao-Yuan. [Operator Instructions] So our first question comes from Jennifer Kim from Cantor Fitzgerald. Please go ahead, Jennifer.

Jennifer Kim

Hey, good morning. Congrats on another quarter and thanks for taking my questions. Maybe to start off with the DRAGON trial. The additional enrollment in this patients, I'm wondering so far, can you breakdown where those patients are coming from in terms of the clinical trial sites, and especially those I think 16 more patients this quarter. Where are those patients coming from? And then my second question, more just a general question. For BCVA loss over 18 months, I'm just wondering what kind of loss would one anticipate in patients with DDAF at baseline. Thanks.

Tom Lin

Sure. I'll take the first question. I'll let Nathan answer the second question. So far the majority of patients are coming from – I would say majority of patients coming from Europe. And then with sites that just started in China about a few months ago and given a large population in China, that recruitment has caught up pretty fast with Asian patients from Asia especially from China. So I would say the bulk of the studies so far, the enrollment so far from Europe especially from UK, Moorfields [indiscernible] and then from Europe, aside from Asia and Australia. Nathan, the visual acuity.

Nathan Mata

Sure. Yes. In terms of loss of visual acuity in Stargardt subjects who have atrophic retinal lesions at baseline, it really depends on where that lesion is. So most ophthalmologists will tell you if it's an foveal involved lesion, you can expect about a line of loss that is five letters per year. Our subjects came in with foveal involved autofluorescence, which converted to atrophic lesion. So they're having lesions essentially in the phobia, and yet we're having fairly stabilized vision. So we believe. And if you look at some of the quantitative autofluorescence data, we believe that we are actually clearing some of the autofluorescence away from the fovea. So we think that's why we're getting a stabilization. But to answer your question, typically you see somewhere between five to six letters per year, so roughly about a line in these types of patients when they are foveal involved.

Jennifer Kim

Okay. That's helpful. Thanks guys.

Operator

Thanks for the questions, Jennifer. So our next question comes from Tim Lugo from William Blair. Please go ahead, Tim.

Timothy Lugo

Thanks for taking the question. And is the 24-month data expected to be at AAO? Is that the likely spot for the update?

Tom Lin

Yes.

Timothy Lugo

And I guess, what are your expectations given the 18-month data for that update?

Tom Lin

So I don't quite get that. I think you're breaking off a bit.

Timothy Lugo

I'm sorry. Just what are your expectations for the 24-month data given the 18-month data? Are you just looking for stability? I guess improvements.

Tom Lin

Nathan, do you want to take it or I can do it.

Nathan Mata

Yes. I mean, I can take that. So if you look at the trajectory, I don't know if we can go back to Slide 9 to show the lesion growth data relative to ProgStar. When you ask, what we can expect? We're seeing a very little linear trajectory. It's just a different slope than natural history. So if you look at, for instance, I guess the DAF would be a good example. The difference between natural history and our growth is, as I said before, about 50%. So we expect to see this continued through 24 months because imagine both of these lines just basically continued outward for another six months. Same thing with respect to DDAF.

You see here an inflection downward at 12 months. Again, these are limited numbers of patients. But generally speaking, these are fairly linear growth rates. We expect to see that linearity continue through 24 months. I expect that we will see the same treatment trend that is roughly a 50% difference in the growth rate at 24 months versus what we're seeing here in ProgStar. And again, we do have the 24 month data for ProgStar, so we know exactly what that looks like. We believe we'll be well under that.

Timothy Lugo

Okay, I understand. And since you have the 24-month data, and I'm sure you've shown this to us over the past few days as well as during these KOL calls. But can you remind us what the 24-month data looks like for ProgStar?

Nathan Mata

Yes. We haven't shown the 24-month ProgStar data yet. Because what I can tell you is that, again, if you just draw these lines out linearly, that's exactly what the 24-month data in ProgStar shows. So you can expect another roughly a 0.2 millimeter square of growth on the DAF. So again, just stretching it outward. And the same thing could be said with respect to DDAF. And again, we'll show that, when we're comparing the ProgStar 24-month data to our 24-month data. Right now, we have nothing to compare to. So it doesn't make much sense to show the ProgStar 24-month data. But I can tell you it is quite linear. And during our KOL event with Hendrik Scholl, he did comment in fact that he sees very linear growth throughout time points up to two years.

Timothy Lugo

Fantastic. And how long do you expect enrollment to last for PHOENIX?

Nathan Mata

Yes. That's a good question. Right now – so we're just starting to enroll this month. I'm expecting it'll take at least a year with our strong network to start getting up to the 430 subjects, but potentially up to a year and a half, if things don't go our way.

Timothy Lugo

Understood. Thank you for all the questions.

Nathan Mata

Sure.

Operator

Thanks, Tim. Our next question comes from Yi Chen from H.C. Wainwright. Please go ahead, Yi.

Yi Chen

Thank you for taking my questions. My first question is, does the FDA require a certain percentage of the patients in the DRAGON trial to be recruited from the U.S.?

Nathan Mata

They don't have any guidance, Yi. Regarding the percentage or number of patients, they want some representation. In the U.S. typically what we do is after we've completed enrollment, we will communicate with the FDA what our subject composition looks like. We have regular meetings with the FDA, thanks to our fast track designation. We have that access. So they will be feeding back to us information with respect to demographics. But there is nothing in the guidance and nothing certainly in a regulatory statute that says you have to have X number of patients for an international Phase III study to be considered for an NDA in the U.S., you have to have some U.S. representation. The percentage on that is a little vague.

Hao-Yuan Chuang

Yes. If I may, I would like to add that, we were not to worry about that. Given, we're not like some of the studies, international study that they have the majority of the patient or even more, or even 80%, 90% of the patient coming from one Asian country. In our study, we do have – majority still come from Europe. So we don't expect we'll run into any problem instead of having too much focus on non-U.S. countries or race.

Nathan Mata

That's a good point, Hao-Yuan. Thanks for mentioning. Because after all, what the agency looks at is the races, right? So Caucasian, the European population represents the Caucasian race, which is the predominant race in America. As long as there's some distribution of African-American and potentially Hispanic, they like to see that, but you can't always get that. Particularly in Stargardt’s disease where you're not just the Stargardt’s disease, which as a whole is limited, but now we're going after adolescent Stargardt disease, which is even more limiting. So the agency is willing to give us some leniency with respect to demographics because of that rarity of finding these patients.

Yi Chen

Got it. At this point, is it safe to assume that the DRAGON trial could complete enrollment around mid-2023?

Nathan Mata

Probably by July.

Yi Chen

Yes. Okay. Thank you. And my next question is, so you have the PHOENIX pivotal trial, starting enrollment and together with DRAGON trial enrolling. How should we project the operating expenses going forward?

Hao-Yuan Chuang

Well, I would say, for the DRAGON study, I think the cost is probably around $15 million to $20 million total. For the PHOENIX study, we're expecting something about $40 million to $45 million total. But bear in mind, we don't need all those money on day one. It's going to be spread out in three to four years time. So we wouldn't have too much cash pressure on the first year.

Yi Chen

Got it. Thank you.

Hao-Yuan Chuang

Thank you.

Operator

Thanks, Yi. Our next question comes from Bruce Jackson from Benchmark. Please go ahead, Bruce.

Bruce Jackson

Hi. Thank you for taking my questions. I wanted to ask a few follow-up questions about Page 9, which we have up here on the screen right now. Can you tell us about the error bars on those lines? How are the – what do they represent? And how are they calculated?

Nathan Mata

That is simple standard error of the mean. So basically looking at the deviation about the mean.

Bruce Jackson

And then as we take those bars out to the 24-month time period, what would we see on the ProgStar line?

Nathan Mata

It's going to be very, very similar. I know that it looks like they're getting larger, the standard errors. Especially if you look at DAF, they don't get much larger. They basically stay about that same as a 24-month. So you're seeing roughly – there will still be – if we see a treatment trend as we're seeing here, there will still be a statistical difference, significant difference between LBS-CTO2 patients and ProgStar, but the center deviations do not get any more large than they are here at 18 months for ProgStar.

Bruce Jackson

Okay, great. And then just one follow-up modeling question. Can we use the operating profile from the current quarter and kind of extrapolate that out for the next couple of quarters? Or will the start of the PHOENIX trial have any impact on the back half of the year?

Hao-Yuan Chuang

Well, for this quarter we have higher expenses because when you start a new trial, you will have like a down payment, usually you have some milestone payment that is subject to the achievement of say, select the first site, getting the contrast site, getting the first patient screen, getting the first patient dose, et cetera. So usually you will have slightly higher expenses on the starting of a trial. So I do not expect you can just times it four to get an annual expenses. I do expect that the next three quarters will be lower than this quarter in terms of the income statement.

Bruce Jackson

All right. Perfect. Thank you very much.

Hao-Yuan Chuang

Thank you.

Operator

Thanks for the questions, Bruce. So our next question comes from Yuan Zhi from B. Riley. Please go ahead, Yuan.

Yuan Zhi

Good morning. Thank you for taking our questions and thank you for the two recent KOL seminar. I have a layered question related to the assumption for your Phase III Stargardt DRAGON trial. It would be helpful if you can summarize them for us. So first, learning from ProgStar and your ongoing Phase II program, do you expect lesion growth or DDAF in the Phase III placebo arm to be slower or faster in the DRAGON trial? And then based on your 18 months data, do you expect lesion growth in the Phase III treatment arm to be slower or faster in the DRAGON trial?

Nathan Mata

So I think with respect to placebo in the DRAGON study, we hope to see what we're seeing here. We hope that these natural history data are going to be exactly what we get in a placebo group. And I'll remind you that the slide prior to this where we look at DAF in a separate study in 53 adolescent children, we saw 60% difference. It's a reduction in our growth rate compared to that natural history study. Here we're getting a 50% reduction. So across two different studies with pretty large numbers of patients for adolescent Stargardt disease, we're getting pretty much the same treatment trend. So I believe that what we'll see once we do our Phase III, which is actually ongoing, that the placebo growth should look something like what we saw here in ProgStar as well as what we saw in Georgiou et al. 2020 because within those two studies it's fairly the same.

And we expect that our group would be at least 40% to 50% better than that. Again, that's just based upon this limited data. But if we are thinking optimistically, the placebo groups that are shown here for natural history are looking consistent and we hope that our treatment trend will be the same as we're seeing here. We haven't changed the dose, we haven't changed the patients. The only thing we really did is cap the upper end of the lesion size and require that all patients have DDAF at baseline. So we asked this question just recently to an investigator at KOL, would he expect to see a difference in our lesion growth in this study where we have DDAF versus in the Phase II where we only had QDAF?

And as he mentioned, these are evolution of the same disease. So the autofluorescent lesions turn into the atrophic lesion. So if we can stop the growth rate of the combined lesion is what we're doing here and we're stopping the growth rate of just the atrophic lesion, which again you're seeing here versus natural history. Then it goes to say that we will have the same effect on the DDAF lesion type as we're having on the conversion. So we're looking both at the conversion, but also what we're looking at right here is the growth rate of the incident DDAF lesion. So this is basically looking like whatever patients had DDAF at baseline, let's call 12-month the baseline, what was that growth rate over six months and then annualized that. So based on these data, we are seeing a very positive treatment effect that we believe will be somewhere around 40% to 50% better than the growth rate in the placebo arm.

Yuan Zhi

Got it. Thanks.

Tom Lin

Yes. Just to add, so bear in mind [indiscernible] somewhere between like 20% I believe. Is that Nathan around about 20%?

Nathan Mata

That's about right.

Tom Lin

Yes. So what we're seeing here is 50% to 60%, and even with half, that's about 25%. So I think if we are getting these treatment effects, I think we're very comfortable.

Nathan Mata

And there will very unlikely be any risk for CNV conversion as you see with the Astellas drug. Again, this is an oral once-a-day, it doesn't target anything, but retinal delivery to the eye. Whereas these drugs that Astellas and Iveric are developing our anti-inflammatory agents, so they're very ubiquitous. So they have their actions pretty much everywhere. You will get off-target effects from those drugs. So we think our drug overall will be a much more safe approach, especially for long-term dosing, which is needed in Stargardt disease and Geographic Atrophy.

Yuan Zhi

All right. Thanks for the helpful color.

Nathan Mata

Sure.

Operator

Thanks, Yuan. This concludes our Q&A session. I'll now turn it back over to Tom for closing remarks.

Tom Lin

Thanks. Thank you, everyone, and we look forward to updating everyone on our next call [indiscernible] Hao-Yuan just mentioned, that's in Q4 we should have Phase II completion, the final analysis data as well as completing enrollment for our Phase III. So looking forward to that and thank you, everyone. Thanks again. Bye-bye.