Good afternoon, welcome to the Jefferies Global Healthcare Conference in New York. My name is Michael Scharfenberger with the Jefferies Investment Banking team. It is my great pleasure to introduce Alexandria Forbes, CEO of MeiraGTx.
Thank you very much. Thanks for inviting me to talk at your conference. I'm just going to go through two programs that Meira has at the moment. We're a company that was built on many different technologies. The main one we intended to do was to control any gene with oral small molecules, which we can now do. It's called our RiboSwitch technology; we're taking that into the clinic this year. I'll just mention that program at the end. We have our own manufacturing, which has helped us have four late-stage programs, two of which I'll talk to you today about, next generation optimization, which is one of the reasons our programs have been successful to date. Quite a deep pipeline, but our four lead programs, two of them are awaiting filing, two of them are finishing pivotal or starting phase III.
Radiation-induced xerostomia, wholly owned in-house, and that is in a pivotal study. 12-month data will be this time next year, and we have Breakthrough Therapy Designation and RMAT, and we're hoping to file and get approval at the end of 2027 with a launch in the beginning of 2028. X-linked RP, we partnered this with Johnson & Johnson. They bought it from us, and then we bought it back from them just recently. This is a program where we're intending to get this approved in Europe and the U.S. and launch it in 2027. AIPL1, another disease for the eye. We developed it philanthropically, and 11 babies who were blind were treated with this drug, and all 11 grew into children that were able to see.
Eli Lilly acquired this from us last year, and they are working steadily to get global approval in multiple markets for these children. We have a collaboration with an AI company in Parkinson's disease. This has three studies. Two are sham-controlled positive phase II studies with UPDRS as the primary endpoint, and that is going into phase III as we speak. We have three deals. The one I've mentioned with Lilly about AIPL1 for blind babies. Sanofi has invested in us a couple of times, and we have ROFRs around our riboswitch and xerostomia, which means they get to look at our data for a few weeks before we're allowed to show it or discuss it with others, and the Hologen deal with the AI company for Parkinson's.
The first program I'm going to talk about today is the aquaporin program, and this is for a completely untreatable condition called late radiation-induced xerostomia. This is actually extremely common. Anyone who knows that patients who've had head and neck cancer and been treated with radiation, all of them will have had xerostomia, and 30% of the people who've been cured of head and neck cancer and treated with radiation maintain xerostomia, moderate to severe, Grade 2, Grade 3, for the rest of their life. That is very bad. They have dry mouth. They can't eat. They can't swallow. They often lose weight. They go on feeding tubes. They have terrible tooth problems and infections in their mouth. They often, because they can't speak or eat, don't go out.
We even spoke to an investor whose family member had xerostomia and eventually refused to go out, and eventually just died. Don't be under the impression that it's a dry mouth, and it's not severe. It's very severe. It's for the rest of your life, and it's completely untreatable because these patients are the patients that don't respond to any other treatment that makes saliva be produced. During radiation, the salivary glands are very sensitive to radiation, and that's why everyone gets xerostomia, which is a dry mouth. In most cases, 60%-70% of cases, the xerostomia resolves over about 12 to maybe 18 months as the gland heals. It's only the patients when they get out to two or three years that we treat who have a completely irreversible, and I suppose it's untreatable condition.
What we do is we take the gene for aquaporin-1, which is a non-polar water channel found in kidney, found in blood cells, and we introduce it to the duct of the parotid gland. Patients open their mouth, and you just squirt the viral vector AAV2- hAQP1 into the remaining glands. That water channel, when in the glands, makes those glands permeable to water, and water flows down the concentration gradient into the ducts and into the mouth. We have done a phase I study, a dose escalation study, where the data was incredibly strong.
We just presented three-year data where it was maintained from 12 months out to three years. When you ask physicians about this condition, having shown them the three-year data, we had huge response from physicians, and this actually is reflected also in the very strong end of enrollment for our pivotal study, which was extremely crowded, has just finished. You'll see some of the comments from physicians. They're dealing with patients who really have no option, and what they really liked about this treatment is that it is disease- modifying. It's a one-time treatment that you can give in surgery. It doesn't hurt. It doesn't need anesthesia. It's obviously you go, and you don't stay there, and it looks like it has these transformative effects that are durable.
From a physician position, when they saw this three-year data, they considered this something that they would be highly likely to treat their patients with. 78% is what they said. We actually used a much lower number when we were modeling this, but this is a very, very good physician buy-in and similar actually buy-in from payers when we did this survey. This is a very large market, clearly an unmet need. 30%, 40% of radiation-induced xerostomia patients. That is just those patients are 165,000 in the U.S. and 20,000 new patients a year. You can appreciate that's a large market for any gene therapy, genetic medicine, but also a continuing market.
We think that we can have reasonable pricing that we've discussed with payers. This should be a multi-billion dollar global market with an ongoing, almost $2 billion a year market just in the U.S. From a point of selling, we discovered that in the U.S., there are 15 centers that treat or have under their care over 60% of these patients. With a very small focused sales force, we could launch this drug ourselves. We think be really successful and have a very good launch. The data that we have recently released is from a phase I study. It had two separate groups. It had a set of cohorts, dose escalating, where one gland was treated, a set of cohorts where both glands were treated. This is three-year data. You can see that blue arrow is the decline in the xerostomia score.
Ideally, you want it to go down at 12 months. It was -21 points in the bilateral, which is the dotted line, and that blue line is -17 points. Why I say those numbers is a eight-point change is clinically meaningful, a 10-point change is transformative, and no one's ever seen a 21-point change before. Really, really strong data at our endpoint, which was the primary endpoint for our pivotal at 12 months. We recently released. Oh, sorry, I forgot to show you three-year. You can see other than a little bit of a blip down, the data was really, really durable out to three years. Right? When you look at individual patients, because we wanted to understand if different patients were going up and down or the population was variable.
What we saw is when we did waterfall plots of every patient, so the bilaterals on the left and the unilateral on the right, we saw these waterfall plots maintained over three years. You can see at the far right, there are a couple of patients that don't respond. It's not perfect, but it's a phase I dose-escalating study with three patients per cohort. We thought this was pretty good. What we did is we looked at every single patient at every single time point, and we compared their XQ score over three years. What you can see here is the darkest red line is the 12-month data, and the lightest red line is the three-year data. There's remarkable consistency. If you responded well at the beginning, you continue to respond well.
If you responded poorly at the beginning, you generally don't respond. The one who responded worst was the worst performing. You can see that guy. I don't have a pointer, that guy on the left, the line goes up. A bit more variability when you don't respond, and most patients respond. Remember a 10-point decline. If you look at unilateral, very similar. You respond well, you continue to respond well. In the unilateral, interestingly, patients, let's see, two and three, you can see they didn't respond at 12 months, they start to respond at 18 months, and they continue to do so. This was very consistent data. We looked at water flow, similar. This is a bit more variable, similar. You can see the 12-month similar out to three years.
When we did the same thing, look at individual patient data, you can again see that there was a lot of consistency. Bad responders on the left, they lost water flow. Good ones on the right gained water flow. We were able to do biopsies, well, NIH did them, and were able to show between one and two years that RNA was present, DNA, and protein. That was very encouraging. This is a design of our current pivotal phase II that will read out its 12-month endpoint this time next year. We've just finished the study now. This has full alignment with the FDA on both CMC and clinical design to support a BLA filing and potential approval if possible.
We also have RMAT, Breakthrough Therapy Designation. We're going to hopefully apply for priority review. I'm not going to go through the summary. That's the program that is wholly owned, very large market, and we're starting to look at not only filing the BLA, but how do we commercialize it in the U.S. ourselves. X-linked RP. We started the company only doing local delivery of small doses of gene therapy, which took us to the eye. I've mentioned the drug for blind babies that we sold to Lilly. This is the opposite of that. It's a very common inherited retinal disease. XLRP, X-linked retinitis pigmentosa, is one of the most common inherited retinal diseases, and the mutations caused by RPGR, the most aggressive and common of XLRP. That's what this is for. Patients start to go blind as children. The peripheral vision disappears.
They can't see in the dark and eventually are legally blind in their third decade. We had partnered this with Johnson & Johnson. They paid 100% of development. We were owed a 20% royalty. They paid another royalty, and we manufactured the product, and they acquired it back from us. They bought the royalties back from us a few years ago. However, they decided not to develop eye programs. They shut their eye programs down at the end of 2024, and this product became available, and we were able to acquire it a few weeks ago for $25 million. In addition to acquiring this product, a number of leaders from J&J have joined us, which is extremely helpful in getting the team together again who worked on the development of this program to move it now towards commercialization. We've done a large phase III.
That phase III had a novel endpoint mandated by the FDA that did not hit its primary, it did hit a trend. Every other of the secondary endpoints and another functional vision endpoint met with very strong significance. We are looking for approval in Europe, U.S., and Japan based on that data supported by the phase I study. You'll see these are quite large studies. A lot of patients have been treated. Very good safety. We are the manufacturer, and we've completed PPQ, and there are no capacity constraints, and it's actually a very low cost of goods when we make it. Just to go through the phase III data, this was a blinded study with two dosing arms and a non-treated arm. The endpoint was the primary.
Well, all the endpoints were at 12 months, and everyone in the non-treated after that period of time went into the long-term follow-up, and most of those, I believe, have now been treated. The primary endpoint was initially a maze, which we spent two years discussing with the FDA, in particular, a gentleman called Wiley Chambers. They would not budge on us developing and using this novel endpoint and a very novel way of scoring this endpoint. We kind of had to do this as a primary. However, not only was it novel, but we were concerned about its variability because we had to build mazes all over the world. It was a 32-site study in Canada, U.S., Europe, U.K., and everyone had to go to different mazes, had different adjudicators, had different people guiding them in the maze.
It was a concern, and it's in the domain of functional vision. It did not hit its primary; there's a trend. It's 2.5 x more likely to respond with this strange responder analysis if you've been treated than not treated. Importantly, in the functional vision domain, which shows the same thing, can you function in dim conditions? What's your mobility like in dim conditions? The LLQ PRO is a very well-validated and well-known PRO that measures exactly the same thing. In the case of that, if you look down here, that PRO was quite a high-ranking secondary. If you look at the exploratory analysis of individual questions, the P value for the question on mobility is very strong. Going back to this, we had good responses in functional vision despite the primary retinal function.
This was the endpoint that the Europeans actually wanted us to use. Retinal sensitivity, central 10 degrees, is how sensitive is your retina when you shine different light on it, and the static perimetry responder is this thing you may have heard about, which is, can you get five points increasing by 7 dB and the same five points over two different time points? That was a very high bar. In both cases, we saw very significant improvements in treated versus untreated. LLVA, that is visual acuity on an ETDRS, on a line chart. This was very strong data. There's a P value. These are the individual patients, LLVA, and you can see that 46% of patients had a two or more line difference in the treated and only two in the untreated.
In fact, there was confounding issues in the baseline of those two untreated, which did not exist in any of the treated. In Europe, this sort of data would get something like Eylea approved. This is very, very strong data. Of late, more recently, the FDA itself has said that LLVA can be a primary. While we missed the primary and only saw a trend, we have unprecedented data across all three domains of vision, very strong. When we look at responders who responded to two or more domains of vision, an endpoint in two or more domains of vision, 40% of treated and 0% of untreated had that response. This isn't cherry-picking because you can mix and match any of the endpoints, and this is them all mixed and matched.
The difference between treated and untreated ranges between about 25%-40%, and that's pretty good as a responder with no cherry-picking. This is a large indication, over 20,000 patients in EU and U.S. It's a small community of physicians, maybe 40-50 in Europe and U.S., who see 80% of the patients. 32 of them were in our study, and we know them all, and over 60 surgeons have been trained. We have a lot of KOL support. We have a lot of European regulatory support, actually, and we have several hundred patients waiting. We have drug that we've made out of PPQ ready for commercial, so that's already been paid for by R&D. We have the capacity to manufacture this drug as much as we need it.
The phase III, supported by phase I and extension, that treated 137 patients, 127 patients, I think. No, 138 patients. Very good safety. Every patient at every time point was checked for any safety issue, and it was really good safety. We have very good leadership from Johnson & Johnson and the very team that developed this with Johnson & Johnson, now working for approval. The EMA has requested a number of times that we file for approval, and that will be happening as we get all the information from J&J, and as soon as we get that information. Also, we'll be filing for meetings and designations with the FDA. Likewise, in Japan, four patients were treated, which is the same as Luxturna got approved on, and the physician who treated them saw some of the best responses in Japan.
One of his patients had a 40-letter improvement in visual acuity. There is a lot of excitement in the regulatory folks and the physicians in Japan. We will also be looking to hire a Japanese group for approval in Japan, which is actually good for pricing because they are keen on making these sorts of drugs available to the Japanese. First, I'll just ask any questions. Oh my God, a question. Yeah.
On the xerostomia product, does administering the product into the duct require cannulating the duct? Part one [audio distortion]
Okay. This can be done by a dentist. This can be done by a surgeon, and you can be trained in about two hours. The patient opens their mouth, and at the back of their mouth, you can see the opening of the parotid. A cannula is inserted into the opening, and the fluid is injected into the volume that, that gland holds. Patients, for example, patients are asked how inconvenient, how painful this is, and when a very stoic gentleman from the north of England who was a miner was asked, "How painful is this from 0- 10?" He went, "Well, somewhere between zero and one." It really is an easy procedure, and the physicians who are doing it say it fits really readily into their clinical practice.
The person who actually administered that to the gentleman I mentioned said he could do up to eight a day in his clinic. It really is easy, and that's one of the reasons you have such buy-in. Patients are very willing to do it, and physicians, it fits into their practice, and it's just a one-time , and then you have a lifelong benefit. Well, you have a benefit in an otherwise lifelong, untreatable disorder. Yes. Another question.
Just for the XLRP, has there been any turnover at the FDA since the submission of the protocols ?
We have not spoken to the FDA yet. What we're doing is we're collecting all the documents from Johnson & Johnson, and while we prepare for EMA, when the IND is transferred to us, we will immediately start filing meeting requests, designations, all those sorts of things. We have not started yet. Yeah.
For xerostomia, it looks like you're collecting data on, like, a PRO. The patients, I guess, how they feel about it. Also, how much saliva they're producing. Which one is the FDA interested in? Which one carries more weight?
The XQ, the PRO.
I imagine that somebody who's had drawing now, let's say three years ago, and got treated, I have a hard time imagining that they can really recall what that was like.
That's a very interesting question, because there are some PROs that say, "Are you better, worse, or the same?" from pre-treatment. This PRO defines xerostomia. It's not how much better are you, but how bad is your dry mouth? How bad is your sleep? How bad is it when you speak? You can appreciate that seeing the consistency we did when it is a PRO, which are more variable, indicates that maybe when the PRO is done at two years and three years, and they come out the same in the same patient, that's pretty compelling that you've had a real change. You're absolutely right about having to remember what it was before. Most PROs wear out that kind of PRO in a couple of years. Xerostomia is only a patient-reported indication.
If you look at the ASCO guidelines, there is no correlation between absolute saliva and xerostomia symptoms. You can only really improve xerostomia by looking at the xerostomia scales, and that's what we did. Our clinical trial design, primary secondaries have been agreed by the FDA, as has the statistical plan in writing in response to a briefing book that we gave them. Riboswitch, completely different topic, but super cool technology. We think one of the coolest technology in genetic medicine in the world. What this allows us to do is control any gene, the production of any therapeutic protein, from peptide, hormone, antibody, Cas9, CAR, anything that you can encode in a gene, we can control the expression of that in a person, well, it's mice at the moment, but in a person using a pill.
We can control it more accurately than you can control by injection. How we do that is a splicing mechanism. Essentially, we put a toxic cassette into our DNA sequence. You put the vector, you can inject it into muscle, lentiviral, naked DNA, whatever you want to do, it can go into CAR T, and that gene sits there, and that toxic cassette is spliced in, and the entire RNA degrades. If you add a small molecule, that alternative splicing happens, the toxic cassette comes out, and RNA is made. This can be used for vectorized antibodies to control antibodies. We've done it for every pharma's biggest antibodies. CAR T, it improves CAR longevity and makes them 4x more potent. We can control editors, we can do hormones, peptides, CNS delivery, many things. These are some of the things we've done.
This is an example of the dose responses that we do very tight dose responses. The thing that we're taking into the clinic this year happens to be one of these hormones and peptides. It is leptin. There's an unmet need for leptin. Injectable leptin is quite toxic and can be immunogenic, and is only used in leptin null kids. On the left, this is in vitro leptin expression in dose response to our molecule. On the right is the serum level. These mice have a gene for leptin that's off all the time in their muscle that we injected one-time injection, and this is in response to an oral dose. You see that dose response. This is leptin minus mice given an oral pill every day or oral molecule every day, and you can see really clear dose response in obesity.
We have done this, given those mice the small molecule every day for 20 months. You can see here the ob mice are the top. The middle mice are mice that have low level leptin, so they're like a hypofunction. They're not a null. The purple line have the controlled leptin. Every time you add the small molecule, their weight is lost. You take the small molecule away, they gain weight. They lose weight when they get the small molecule back. They gain weight, lose weight. Finally, at 274 days, which is a long time after they've been injected and they've been getting the molecule on and off every day, we gave them the molecule for another four or five months, and every day they produce the same amount of leptin and maintain that lost weight.
That's what we're taking into the clinic this year. We have had INTERACT meetings with the FDA. We're currently going into manufacturing of the viral vector, and our small molecule is ready for its GMP and it's ready for the clinic. That's it. Question? Yeah.
Is this leptin in partnership with Sanofi?
No.
No. Okay.
Sanofi has a ROFR on incretins. They did actually ask after the fact, could they include leptin? Now it's only a ROFR. It's only to discuss based on data. We pointed out leptin isn't an incretin.
Just thinking about the mechanism you have. While you have a very tight dose response, you initially have to get the gene therapy, DNA, whatever, RNA, into muscle or wherever into a tissue, but that's where you have the variability?
I don't have the slide. That's super interesting thing. We've done PK studies, right? In fact, as you're alluding to, for example, in the liver, if you dose a number of mice with a gene in the liver, right, you get about a half- log variability between the expression of that gene in the liver. Between mice, yeah between mice. When you express the regulated gene only with a small molecule, every mouse produces exactly the same amount because it's limited by the small molecule. You look, and what our gene is, it's a sensitive measure of small molecules. One small molecule makes one RNA. It very, very accurately covers the different tissue biodistributions.
As long as you get a minimum amount of DNA transfection, you should be able to express that RNA.
We do. We haven't only been developing this, but what we've been doing is improving and optimizing our vectors, promoter, we've got AI-driven promoters and capsid, all sorts of things. We have very, very potent vectors. I mentioned at the beginning, we only do local delivery and somewhat immune-protected doses. That technology was really developed for this. We give really small local doses into muscle, which we manufacture and actually have low cost of goods because it's so potent. Now, we're at a position where with low dose vector and a dose of our even more optimized small molecule, because we can use different small molecules, we have about a pill equivalent to Advil and a single injection into muscle of a relatively low dose AAV. This is not IV large doses.
This is something that costs in the thousands of dollars cost of goods. I can talk to you more if you like.