From Jean Bennett, that are all gene augmentation using known vectors, going after different types of inherited retinal degenerations, that potentially show improvement of vision. What you will see in our dataset is we like to always try to show improvement of function, not just preservation of vision. We have our most advanced program is called LCA5. It is by far, but 10x our smallest indication. But it is right in the middle of the pivotal trial. That trial will read out next year with potential BLA filing next year if the data holds. BEST1 is a really interesting story. It is our next pipeline program that we will talk about in some detail, but much larger indication at 23,000 or so patients. We have proof of concept data and a low-dose early cohort with an additional cohort coming in Q1.
We also are about to be dosing patients in two additional mutations, one called RDH12, which is about 10 times the size of a LUXTURNA. It is 2,500 or so patients. MERTK is about the same epidemiology. Both of those are more visual cycle type programs that we are excited to be treating patients in now.
Got it. The BEST1 clinical data that were reported earlier this year, obviously from the low-dose cohort, I think it surprised people not just with the data, but the fact that you had alignment with the FDA when you put out the data. The guessing game was off.
Yes.
Right. Can you sort of talk to that dataset? Because I think initially there was the expectation of the subretinal fluid. But then you had something much more remarkable than just the subretinal fluid.
So Best vitelliform macular dystrophy is really characterized by the presence of vitelliform material. And vitelliform material is lipofuscin or byproducts of the visual cycle and working photoreceptors that is not transported into the circulatory system for disposal in patients who are missing the bestrophin channel. And so fundamentally, the holy grail for this disease is to replace the bestrophin channel and remove the vitelliform material. And that's exactly what we showed in at least two of our BVMD patients in cohort one. Interestingly enough, the coolest part of the dataset for me is that the retinal sensitivity in exactly those same areas dramatically improved. And so the vitelliform material is really a marker of where you have rescuable photoreceptors. And those photoreceptors, when you do rescue them by removing the vitelliform material, the sensitivity improves quite dramatically.
If we had just taken those loci where the vitelliform material was, we would have seen about a 4.5 or so decibel improvement on microperimetry. And as you guys know from the literature, three decibels is the improvement is equivalent of 15 letters on an eye chart. So really dramatic improvement in function of the retina. Now, the subretinal fluid, these patients in cohort one were end-stage patients with kind of a lake of subretinal fluid with the vitelliform material around the periphery of it. And the subretinal fluid did not change, and I think that's a characteristic of the fact that there is very little viable photoreceptors in that subretinal fluid. I think that's kind of like a central graveyard for the retina.
And that was really elucidated by an Italian study that was published in early September, where they showed the natural history of 13 patients over five years. And what they showed was in the area of central fluid, it doesn't really change. Everything just kind of stays the same. But each year, the area where the vitelliform material is, they call it a transitional zone, that area decreases on average by a half a decibel. No patient got better.
They all decreased. And the fluid would just kind of circumferentially get bigger each year. And so I think that the vitelliform material is kind of just where the action is in these patients. What we're going to see in cohort two is much more early-stage patients. And so it's going to be a little bit different because of that than what we saw in cohort one. For that reason, I am pretty excited.
It is also a higher dose cohort.
It is about three times the dose. Yeah. It is a much higher dose. Yeah.
Yeah. I mean, that is the interesting thing about when you think about the pipeline, you can go after loss of function mutations, which maybe is easier not to crack. But then to sort of address dominant negative mutations like you did with,
Yes
BEST.
Yes.
That is where the skepticism was.
Yes.
On paper, it should work, but we did not know how much you needed to go up before you saw the effect.
Yes.
You hit it with the first dose.
Yes
Obviously, expectations now for the second dose are
Higher
higher.
Exactly. Yeah.
I do not know if that is a good thing or a bad thing.
Yeah. I mean, what I would say about that is like, it's a great point. I do think fundamentally the September readout, that was what it was about, was, is it going to work in Best vitelliform macular dystrophy, which is the dominant version? And I think we showed at least in those first few patients that it does, or it's very promising that it might. And now the magnitude of effect that we saw in that first dose was quite high. I think that the idea that a higher dose will cause a marginal improvement over that is a tall bar for the drug product because we're already at a really good place from cohort one. I don't know if we're going to see. I mean, what I would love to see is just consistency.
If we can show 11 more patients, 10 of which are BVMD, if we can just show consistent results in line with low variability compared to what we saw, in line with what we saw in cohort one, we'll be like really, I think we'll be really de-risked going into phase III.
Is there a way to identify patients who are more likely to respond?
Yes. Based on the data we have to date, it's all about the presence of the vitelliform material. So we want patients that have active vitelliform material and vitelliform material in regions that we can treat with the gene therapy, and that is unconfounded with additional pathology, like the lack of subretinal fibrosis, things like that.
So any registration study, you do plan to enrich for these patients?
Yes.
And hence,
Absolutely.
Probability, once you move into phase III,
Yes.
Should be relatively high.
I think that's the key. That to me, that's what I'm most interested in in cohort two, is my ability to use the data from cohort two to increase the prior probability of success in phase III.
Got it. Onzi, you want to have?
Sure. Maybe going to pivotal trial, and the alignment with the FDA about the endpoint. I think you have seven pre-specified endpoints that should show statistical significance in improvement. Maybe can you just talk us through how you choose those points?
Yep. The easiest way to choose the points is just look at the hyperautofluorescent spots on the autofluorescent image. The autofluorescence image is a high-resolution image of the retina, and lipofuscin is bright white. So what you see on our patients, 101, 106 is the most clear example of this we have public so far, is that there's these little white balls on that image. If you were to just pick the loci specifically over those white balls, you would have had four and a half decibel improvement at three months. It's a quite meaningful way to pick them, and it's simple.
And let's say you have more than eight of these spots. How do you choose the seven right ones essentially?
It's a good question because you want to be not only sensitive, but also specific on this, right? Because you're taking an arithmetic average of the spots you choose, so you don't want to have spots that don't move, right? So really, the key in my mind is not only sensitivity, but also specificity. So I want to pick the spots that are most likely to improve. We're actually doing a lot of work in the background right now on characterizing the lesions, like actually characterizing the vitelliform lesions to see if you can put a predictive value on those spots based on the dimensions of those lesions. There is more to come on that.
Like, right now, what we know is that we would meet the bar if we just simply picked the spots, but we do think there is room to improve by what you said, which spots do you pick? Do you pick them all? Are there certain ones that may not move? Are there certain ones that are more likely to move?
Can you talk about the vitelliform material? This gets deposited in the earlier stages of the disease and maybe gets reabsorbed as the disease progresses?
Yeah, it does. So the vitelliform material is actually lipofuscin, which is byproduct of the visual cycle that gets deposited, that basically gets produced by the photoreceptors and transported into the choroidal circulatory system for disposal by the bestrophin channel. So when the bestrophin channel is missing, that vitelliform material backs up. It is like the garbage truck didn't come, right? It is like you get a big mound of this stuff. So early in disease, you actually get this big yellow lesion. They call it an egg yolk. And those patients have very little subretinal fluid, if any, almost none. It is just a vitelliform material. And that vitelliform material puts a metabolic stress on the photoreceptors, which causes the sensitivity of the photoreceptors to light to start to go down. As you get later in the disease process, that egg yolk starts to scramble, right?
The vitelliform material starts to break up, and then you end up with subretinal fluid, and the vitelliform material starts to go away. The reason for that is because the photoreceptors are atrophying over it. So as the photoreceptors atrophy, they are producing less vitelliform material, and so that vitelliform material that is there eventually starts to kind of break up and go away, and you are left with an area just of subretinal fluid. And so in cohort one, all of our patients were sort of that stage 4 or 5, where you have that fluid with minimal vitelliform lesion around the edges. In cohort two, you will actually see majority are stage 3, a couple are stage 4, but you will see at least one stage 2. So the demographics are fundamentally shifting towards earlier stage patients in cohort two.
When you think about the registration study then, do you want to enroll across the spectrum or with an eye on the label, or how should we think ab out that?
I think the most predictable response is going to be in the earlier stage patients. We have a lot to learn on that in cohort two. But based on cohort one, that is what we've seen. I think from a standpoint of increasing the prior probability of success of the phase III, I think the more homogeneous early stage population you can get, the better. From a labeling standpoint, it'll be interesting to see. We haven't had those discussions yet, but I would expect the label would be for patients with Best disease, and that we wouldn't be pigeonholed into an early stage population because we have shown really meaningful functional improvement for patients at end stage, including a 63-year-old who could see on the eye chart for the first time since he was 15 years old. I think that it would be- We'll see.
We'll have those discussions. It would be a travesty if you couldn't offer it to later stage patients, because we've seen a really important improvement in this patient's quality of life.
Got it. On the alignment with the FDA, the greater than three decibel change in microperimetry. That's like a patient reported outcome, right?
Yeah.
How do we understand this patient reported outcome? Because it's in-
Yeah.
In other disease, forget retinal disease, in other diseases, it's usually been a stumbling block.
Yeah. Yep. The way microperimetry is done is the patient is looking into sort of a bowl, and they have a clicker, and when they see a light, they click the button. What the machine does is it actually shines light on different places of the retina, and then it'll shine it multiple times, and it'll vary the intensity to give you a, basically a luminance that they can see out of, and that's what the decibel is. It is a patient reported outcome. It is a subjective outcome, but it's a little bit more objective than a lot of things, because there is a stimulus followed by a reply from the patient. What we've seen in natural history is that normal fluctuation is about a half a decibel. Three decibels is way outside of normal variability.
We've also seen specifically in Best disease, in BVMD, the dominant Best disease, that there's a kind of uniform decrease over time by about a half a decibel a year. That'll be the primary endpoint. There is another PRO component to it where to be a It's a responder analysis between two groups. To be a responder, you have to hit microperimetry that we've talked about. You also have to hit the PGIS, which is Patient Global Impression of Severity, which is a scale, ordinal scale. It's a question, very simple question, rate the severity of your disease, then, say, 0 to 5, something like that. That's how the primary endpoint is set up, is a responder analysis where a responder has an improvement on both.
The-
Triggers
You have to sort of pre-select seven spots, right? So, how do you guarantee fidelity while you select those seven spots?
Pre-specify, it doesn't have to be seven. It has to be at least five. It can be up to as many as you want. You do have to pre-select them before the surgery happens. What we likely will do, and we're working out this methodology with cohort two, but what we likely will do is just have a little committee of key opinion leaders who get together before the patient is operated to pick the loci that we'll include for that patient.
Got it. Onzi, you want to have?
Maybe just how would that approach then translate into real world when you cannot have these parallel committees?
Mm-hmm. So in the real world, microperimetry is really a tool of clinical trial mostly. In the real world, I think that these patients will be treated without a requirement of microperimetry improving. I think that what the physicians in the real world will be more interested in is the sort of activities of daily living improvements, which in our cohort one have been quite dramatic. So, yeah.
Can you talk to the prevalence and diagnosis rates?
Yeah.
How challenging is it to diagnose, and at some point you have to probably start preparing for a commercial launch and
Yeah. Maybe I'll
Um-
Yeah, I'll let Rob talk about the epidemiology and his commercial launch.
Yeah. So, originally we thought the prevalence was 8,400, and that was based off of the documented databases out there for Best disease. We were hearing from various KOLs that they thought the number was quite higher, and George can talk about what that's like in practice. We actually commissioned a third party to go out there and do some work to help validate that, and they went out to 150 eye specialists, really a bottoms-up approach, quantitative approach, came back with a result of about 23,600 patients in the U.S. Do you want to talk about the egg yolk lesion and-
The reason we did that was because BEST1 is unique for IRDs. Most IRDs, when you look at the retina, they look the same, right? It is hard. To diagnose Stargardt, you need to have a Stargardt, you need to have an ABCA4 genetic test. To diagnose Best disease, you do not have to have that, because the egg yolk lesion can be diagnostic, can be pathognomonic. For a majority of physicians out there, when they see an egg yolk lesion, when they see a Best disease patient, they tell the patient they have Best disease. There is nothing you can do about it, and they do not get referred to a tertiary care center or genetically tested. When we were reporting 8,400, we thought that we were doing that from basically the published literature that is out there that typically gets genetic databases and tertiary care centers.
The KOLs were out there telling everybody in the world that we were way off. They were like, "We see, for every Stargardt patient I see, I see a Best patient." It is like, "We are on the order of that." It was like one KOL call where they flat out asked him in a public call, "Why does the company say 8,400?" He is like, "I do not know. They are wrong." Rob and the team, Rob and a guy named Joe Schachle on our team commissioned a more quantitative approach that he described that came out to be- I got a lot of "I told you so's" from the KOLs when we showed them the results. It came out to be what they were saying it was.
On commercial launch, we are starting to think about that now, and it is going to be a very targeted approach. It is going to leverage digital technologies as much as possible, but probably looking at FTEs somewhere in the range of 20-30 in total. It will be very fit to purpose for the launch and for the size of the market opportunity.
Mm-hmm. You think the pivotal study, the registration study gets underway back end of next year? From a-
Data holds, yeah. Mm-hmm.
How quickly can you enroll and sort of, is this a 29, back end 29 kind of approval?
The company hadn't officially guided on that, but yeah, I think that math kind of, like, I think that would be logical. The enrollment, I mean, look, 11 patients from mid-August, and they were all done and treated by October. So pretty fast, like, when it starts going. There was a lot more demand than the 11 patients, too. So the patients are like, there's I think this thing could potentially go quickly. Yeah.
Okay.
Yeah.
Let's talk about the remainder of the pipeline. What's especially with the focus on 2027 now?
Mm-hmm. Well, you want to start with LCA5 and how that's meaningful to us?
Yeah, LCA5, well, we think there's approximately 200 patients, maybe just under 200 patients, but a lot of them are captured at Penn, so we know who they are now. But LCA5 will also come with a voucher, so that's something that we'll be working towards to initiate the auction process. That would probably fall in the 2028 timeline, likely. I think I've seen some of those go off recently at $200, $205, even maybe $225.
$215 was the highest we saw.
$215. We'll be working towards that in 2028. It'll be a great source of capital, fund really the rest of the pipeline for us. We have a number of other exciting programs too. We're making good progress with RDH12 and MERTK.
RDH12 is probably the most imminent one, so we'll be dosing patients any day now on that one. That one is probably the most straightforward read-through to LUXTURNA. It's a RDH12 is visual cycle, LCA, that we should see an improvement. What we'd be looking for is FST within a few months moving. That'll be sort of the guidepost for whether or not we're getting target engagement on that program. MERTK is an exciting one too. Lots of patient for RDH12 and MERTK, lots of patients in the Middle East where they actually get comparable pricing for LUXTURNA as the U.S., but on the order of like 15,000-20,000 in the MENA region. MERTK is actually being sponsored in Abu Dhabi by the Department of Health there. That one is kicking off as we speak as well.
I mean, kicking off as in treating patients imminently. That one's an exciting one too. There's a lot of experience in MERTK, a lot. It's a well-characterized disease. I think both, certainly RDH12, I think both of them have a pretty high probability of success from a technical standpoint. Both of those will read out some data, either next year or early 2028. So potentially have four data readouts next year, the pivotal LCA, the BEST1, RDH12 and MERTK.
Rob, where are you with your ability to fund everything?
Yeah.
I mean, clearly there's this, it's a PRV machine, but yeah.
There's a lot. I think that really speaks to the business model because we estimate from an IND to a BLA, it takes us about three years and costs us about $30 million. So it's a very efficient business model. We put out our update Q3 earnings release this morning. We'll be filing a 10-Q later today. Our cash as of September 30th was about $90 million, and we actually had some warrant exercises. We also have a milestone coming in this quarter from phentolamine, that legacy asset that was licensed, essentially a financial asset for us at this point. So long and short of it is pro forma, we have about $100 million of cash sitting here today. We have a runway into 2029, and the reason for that is we have an additional $70 million of available capacity on our credit facility. So a good long runway.
We can work the math so that the voucher in 2028 could be the next financing for the company. We will just have to see.
Well, congrats on all the progress. I am looking forward to the next data updates next year.
Yeah. Thanks .
Thank you.
Yeah.