I'm very happy to be kicking off day two of our healthcare conference with the EyePoint team. I'm joined here by EyePoint President and CEO, Jay Duker, CFO, George Elston, and CMO, Ramiro Ribeiro. Thank you, all three of you, for joining us early on a Tuesday morning. Just a quick note on our research disclaimer. Please feel free to visit morganstanley.com/researchdisclosures for our disclaimer. Let's get into it. It's been an eventful few months for the EyePoint team. Maybe we can start on the recent LUGANO trial readout, the phase III readout, the first of two phase IIIs that you're going to have this year. When we look at the data headline, the study didn't technically meet the primary endpoint in the full intent-to-treat population, but there were a number of encouraging findings underneath that headline.
Would love to hear you guys walk through your interpretation of the data, and importantly, what may have the market missed.
Sure. Thank you very much for the invitation. Thank you for the question. The LUGANO trial was a phase III trial, which studied our vorolanib intravitreal insert, 2.7 milligrams per dose. Vorolanib is a small molecule tyrosine kinase inhibitor that has activity against all the VEGF receptors, PDGF, and JAK1. The control arm in this study was on-label 2 milligram aflibercept. We were studying it in wet age-related macular degeneration. The patient population, 75% of them were treatment naive, and 25% had been previously treated. The primary endpoint of the study was non-inferiority change in visual acuity from day one to the average between week 52 and week 56. As you mentioned, in the full analysis set, we did not meet the primary endpoint. But there were some really encouraging data that came out of the study, including the important secondary endpoints. First of all, safety.
In the prior phase II, phase I trials that we've run with the insert, we really had no safety concerns, and that was really backed up in the phase III. The safety looked quite clean. We also showed really good anatomic control of wet AMD. That's measured in the office by an instrument called Optical Coherence Tomography, or OCT. At the end of the trial, there was only a negligible difference in retinal thickening between the study arm and the control arm. We were able to reduce the treatment burden by over 40%, which, if approved and in the real world, could mean two fewer injections per year on average. Over half of the wet AMD patients went the full year under control with our drug alone. In that group of patients, we were statistically non-inferior in change in visual acuity.
In the overall population, however, one of the things that really stood out as highly unusual is the number of eyes in the control arm that lost 15 or more letters. Historically, that occurs because some eyes just don't respond to anti-VEGFs. This is an elderly population, so some of these patients will lose vision for other reasons, like cataract or glaucoma or dry AMD. Typically, in the 2-milligram EYLEA arm at the end of a one-year trial, you expect between 4% and 5% of the eyes to lose 15 or more letters. In the LUGANO trial, it was only a 0.5% . That again, is statistically highly unusual. We believe that was the primary reason for the miss. It certainly wasn't that our drug wasn't active. Again, looking at the secondary endpoints, our drug was durable and potent and safe.
There appeared to be a significant mismatch in these 15 letter losers. If you go back and look at our phase II trial, the DAVIO 2 trial, 7.7% of the control arm, the same control arm, 2-milligram EYLEA, lost 15 or more letters. So it went from 7.7% to a 0.5% . In the phase II, in our high-dose arm, it was under 4% lost 15 or more letters. When we really broke it down, we ended up with nine patients in the DURAVYU arm that lost 15 or more letters for something other than wet AMD. In those eyes, when we looked at the visual acuities and, of course, the anatomic control and the overall clinical findings, six of them lost vision due to geographic atrophy, the severe form of dry AMD, versus none in the control arm. Again, highly asymmetric and unusual.
Two lost vision due to glaucoma, and one from a successfully repaired retinal detachment, where the patient developed a severe cataract after the successful surgery. So we believe that this asymmetry in the 15 letter losers was the primary reason that we missed the endpoint.
Okay. That's really clear. It sounds like the control arm simply just performed a lot better than you would have expected based on historical trials. In terms of the nine DURAVYU patients that actually experienced a 15 letter or greater vision loss, can you help us understand what you think happened in those nine patients? I know you mentioned geographic atrophy.
The geographic atrophy patients, I think, again, at the end of the trial, if you looked at their anatomic control of their wet AMD, it was quite good. In fact, of those nine patients, three of them never received a supplemental injection. The reason they didn't receive it is the treating physician felt that there was no active wet AMD, and there was no reason to give them another anti-VEGF. We then did some further analysis to look at the possibility of our drug speeding the growth of geographic atrophy. We certainly hadn't seen that in any of the prior studies, and there's no preclinical data to suggest that.
In fact, the analysis that we've done shows quite clear that not only do we not cause new geographic atrophy, but in preexisting geographic atrophy, the rate of progression was no different than the fellow eye, for example, which is a nice internal control, and was actually less than what's been reported as the usual progression of geographic atrophy. What we did find is the eyes in the DURAVYU arm started with geographic atrophy closer to the center of the fovea. That means that even if they progress at the same rate, obviously, if you start closer, you're more likely to have visual loss from that. We think, again, it was just an asymmetric distribution.
The two patients that had a loss from glaucoma, one of them, unfortunately, every time they got an EYLEA shot, which was part of the trial for them to get EYLEA at the beginning, they spiked their pressure very high due to the EYLEA shot. Clearly, if that patient had been randomized into the other arm, they would have received EYLEA and had the same pressure spikes. Again, it appears certainly just a little bit of bad luck with the randomization in those cases.
Okay. That's helpful context. When you think about the 42% reduction in the treatment burden that you highlighted, how do physicians think about that magnitude of reduction in the context of real-world treatment burden in the disease?
I think, again, overall, the physicians we spoke to, and we've spoken to a lot of retinal specialists and shown them the data, they're quite enthusiastic about the secondary endpoints, including that reduction in treatment burden. The expectation or hope going into the trial from retinal specialists was that we would have at least a 20% reduction in treatment burden. When you look at the real-world analysis of medications like high-dose EYLEA and VABYSMO, which have been very successful, these second-generation anti-VEGFs, they appear to reduce treatment burden less than 20%, about one injection less per year on average. Yet they're quite successful. We think that that overall success of the reduction in treatment burden, if approved, would be very enthusiastically accepted by the retinal community.
Okay. Then just going back to the safety profile for a moment. It looked quite clean, as you mentioned, including after repeat dosing. So no meaningful imbalance in intraocular inflammation, no observed retinal vasculitis or severe IOI. How important is the repeat dosing safety experience when you think about the differentiation of your drug?
Well, again, these patients, on average, who are diagnosed with wet AMD, live for over another 10 years. You want to be able to treat them successfully and safely for the rest of their lives. The important thing in wet AMD, when you look at the long-term success, patients, despite our very good short-acting anti-VEGF, still lose vision. The ability to retreat and have long-term suppression of VEGF appears to be associated with much better outcomes in the long term. Obviously, you have to show the safety and efficacy in your trials. We were certainly able to show safety with repeat dosing. Again, we expected that animal data in the previous safety from single injection really gave us confidence that we would be safe.
If a drug can't be repeated, obviously, if it's even a long-acting that might last six months or a year, it doesn't really help us retina specialists treat patients in the long term. You need to have repeat dosing.
Okay. Then just thinking about the takeaways from the study, is there anything you learned from LUGANO about patient selection or baseline patient characteristics that could influence how physicians would use DURAVYU commercially?
We've looked quite a bit at the outcomes to see if there was any indication that, for example, the drug worked better in eyes that were previously treated or eyes that were treatment naive, and we didn't find a difference. We've broken it down by some other parameters, and we'll continue to do that, certainly, because that's something, if approved, is very important. You want to be able to tell the treating physicians which are the eyes that are going to do really well. Especially that 54% that went the full year with our drug alone, as I think I've emphasized over and over, we have these supplement criteria in the trial, but that's not how doctors treat in the real world. They don't supplement per se.
But given that we're a different mechanism of action, I think in the real world, if we're approved, doctors wouldn't hesitate to use both a biologic VEGF blocker and a long-acting TKI together to get, hopefully, a synergistic effect. Again, when we look at those eyes that made it through full year stable on our drug alone, we're hoping to find some biologic marker early on that might indicate that subgroup of patients. But as of yet, when we looked in the phase II, we were not able to discern it, and we'll be doing further studies to try to elucidate that.
Okay, great. Just given the totality of the data and the outcome, what conversations have you had or are you expecting to have with the FDA around this study?
I think, again, putting it in context, the FDA has seen a lot of aflibercept 2 milligram arms and how they do. I think we have a very solid clinical explanation as to why we missed the primary endpoint. There are certainly examples in retina of drugs that hit one phase III and missed a second phase III being approved, and there are plenty of examples outside of ophthalmology of that as well. If the second trial, the LUCIA trial, is positive, we are quite optimistic that the FDA will accept our NDA and review it.
Okay. That is a good segue to LUCIA, which I know is coming up next quarter. On Q4, you have guided the street. This is obviously an extremely important readout for the company. Can you remind us how similar the trial is to LUGANO and whether there are any meaningful differences in sort of the patient population that we should be thinking about?
Ramiro, do you want to answer that?
Yes. In terms of study design, LUGANO and LUCIA are identical. Exactly the same type of study design. In terms of the patient population, the only difference is that LUGANO was 100% U.S. patients, and LUCIA is 80% U.S., 20% international. We published the mask baseline characteristics earlier this year in a medical conference, and the baseline characteristics on a mask fashion was pretty similar. The way that we are analyzing the data, also, of course, pretty similar. We are amending the analysis plan for LUCIA based on the learnings from LUGANO. One new analysis that we are going to be doing is looking at patients that lost 15 letters for reasons not related to wet AMD. Similar to what we did for LUGANO on a post hoc matter, removing those patients, in LUCIA, we are going to pre-specify that and has a pre-specified sensitivity analysis.
The primary endpoint, it is still the same, takes into account all the patients, but for the sensitivity analysis, we will remove those patients.
Ramiro, just going a little bit further on the geographic mix. How do you think that could influence either the control arm performance in LUCIA or maybe just the underlying variability in visual acuity outcomes?
Yeah. I think if you look at recent phase III studies in wet AMD and when they broke down by geographic region, there was no meaningful difference between U.S. and international patients. In the end of the day, wet AMD is a relatively genetic disease, so you see mainly Caucasian patients. The sites that we use outside of the U.S., of course, are sites that have experience in phase III studies. They are located in big cities. We should not expect any difference in terms of geographic region between U.S. patients, international patients.
Okay. The timeline you have outlined is Q4 for LUCIA data and then potential NDA filing in the first half of 2027, right? What would you really need to see in LUCIA to feel comfortable moving forward with the planned NDA?
I think it is actually quite binary. If we have a positive primary endpoint, we intend to move forward with the NDA filing in the first half of next year. From the perspective of acceptance in the retina community and use, I do not think the retinal physicians will care very much about the actual numerical difference, if there is one, between our drug and the control arm, as long as we are approved and safe. We would expect to see similar secondary endpoints in LUCIA that we saw. If we do and we hit the primary endpoint, then I think we intend to file, and I think we will have a good chance of having the drug accepted.
I guess in your conversations with retinal specialists, do you feel like there is still a lot of enthusiasm around the program and kind of belief that you guys have an active drug here?
Oh, I think there is a lot of enthusiasm, yes, in that, again, when they look at the secondary endpoints, in particular, the anatomic control, which is what retinal physicians use on a daily basis to decide how well their drug of choice or their interval is working, I think they are very enthusiastic about it. What a drug like the vorolanib insert can do for them and their patients is, first of all, the most important thing is it will really help long-term visual acuity. Because in the long term, right now, patients are undertreated, and by having an insert that lasts six months or longer, the ability to treat in the long term and have patients come back for visits, I think is going to improve. Retina specialists clearly understand that.
It also gives them flexibility to change their dosing frequency longer if they choose to, and depending on the patient and the situation. I think the enthusiasm is there, and I think that in general, the secondary endpoints, we did better than expected in the retina community, and that is what is really going to make us commercially successful. We have a hurdle we need to get over, which is we need to hit the primary endpoint in the next trial and obviously get approved.
Of course. Yep. Maybe just on the commercial opportunity and the competitive backdrop. The competitive bar has certainly been raised by increasingly durable agents like EYLEA HD and VABYSMO that are on market. What would DURAVYU need to offer just in practice to drive actual switching from those agents?
Well, I think the obvious thing is, again, further length of time between injections. If you really look at the real-world data, whatever agent we use, about 20% of wet AMD patients still have to be treated monthly, and about 1/2 have to be treated every two months or more frequently. There is a huge need out there, even with these newer agents, to try to extend those patients out further. Again, I alluded to this earlier. If you have a patient that you are treating 12 x a year, and let's hypothesize that DURAVYU is safe, effective, tolerable, and approved every six months, and you use it in that patient, and you still need to or choose to use a ligand-blocking biologic in between and say your total 4x a year injections.
Those two, what we called in the trial supplements, those would not be considered supplements in the real world because in the real world, you have gone from 12 injections a year to four injections a year, which is terrific for everybody. If you go back and look at the data from LUGANO, almost 90% of the eyes were controlled with four injections or less per year. Again, going back to the real world, if you ask the question, well, how many high-dose EYLEA patients or VABYSMO patients are treated at a three-month interval or longer? It is not 88%, it is probably a third of that. So that ability to get many patients out to that three month or longer mark, I think is something that is going to be very attractive to everybody, patients, physicians, payers.
Yeah. If I can add to that, I think because the topic switching comes up frequently, and that has been the history of how wet AMD is treated. The most recent drugs say, "We last longer. Use us, stop using the other programs." Our message is quite different. Our message is it is no longer either/or, it is both because we are bringing a new MOA. We are going to be that background foundational medicine, and if the anti-VEGF biologics are needed in between, doctors will use them. It is really a different commercial mind shift in that space.
To your point, George, that means that flexibility to actually give supplemental anti-VEGF injections on top of DURAVYU when necessary. That kind of lowers the bar to entry when it comes to practice.
Yeah, I think it totally does. I think even though we showed over half the patients could go six months on our drug alone, it does not mean that is what the majority of doctors will do. Certainly at the beginning, they are going to want to see how the drug performs. Remember, many of these patients do not have wet AMD in the fellow eye yet, and they are susceptible. I think the idea of going six months or longer between visits is not something most physicians are going to be comfortable with. They are going to bring the patients back to see how they are doing and check the fellow eye.
I think that some patients would be very happy to come back for a check to hear they do not need an injection. As I think I have already mentioned, some doctors, even if the eye is under control, may choose to use the assumption that the two MOAs are going to be synergistic and choose to do that. That is the flexibility that our drug may offer.
Okay. Makes a lot of sense. I am going to just switch gears for a moment to your diabetic macular edema program. Maybe you could, Jay, remind us why you believe DME could be an attractive setting for DURAVYU.
Well, I think first of all, the diabetics, it is a different population from wet AMD patients. They are younger. They are often working. Because they are diabetics, they have multiple doctor visits, and it is really hard for them to have the number of injections that they need. The study suggests in the first year of DME treatment, patients should receive about 11 injections, and in the real world, they receive about three.
Some of that is that because DME is a multifactorial disease, it is not just VEGF-mediated, there is clearly an inflammatory component as well, that it can take multiple injections before the patients realize that they are actually getting better. If you are a diabetic and the physician says, "Look, I have got to give you these monthly injections," you have three, four, five of them, and you do not perceive the benefit, it is very easy to stop coming back, and we think that is a real problem in this population. The thing that really gives us really enthusiasm about our drug in DME was the results of the phase II trial, the VERONA trial. We looked at patients who were previously treated, but all of them had active DME, which means they had decreased vision, and they had fluid on OCT.
If you look at the initial data point after treatment, all patients received treatment on day one, and at week four, the eyes in the DURAVYU arm were already seeing about four letters better than the eyes in the EYLEA arm, and they were also 40 to 50 microns drier on OCT. If we can show that in the phase III, even if at the end of the trial we are non-inferior, we are the same vision, but we can show we can get patients drier and seeing better faster, who would not want to have better vision with fewer injections? I think everybody would, and I think that ability to capture market share in the DME market, which is currently about a $3 billion market, I think it is wide open for a drug like ours.
You mentioned the phase III that are ongoing, phase III COMO and CAPRI trials. They are now both fully enrolled. More than 480 patients in the trial. I know you are expecting top-line data from those trials. I think it is Q4 2027-
Correct.
If I'm not mistaken. Can you walk us through the design of those trials and what you learned from VERONA that may have informed some of that-
Go ahead, Ramiro.
Planning?
Yeah. Similar to our wet AMD, COMO and CAPRI are two identical studies. The main difference, first, on the control arm, we're using on-label aflibercept, which means five monthly injections in the beginning, and then after that is aflibercept every eight weeks. On DURAVYU, based on what we saw in our phase II study with the benefit of DURAVYU starting from day one, we are dosing DURAVYU day one together with aflibercept, and then after that, we have two additional monthly aflibercept, and then DURAVYU is every six months. So day 1, 24, and then 48. The primary endpoint is changed from baseline to average week 52 and 56, so this is similar to our phase III program. As the secondary endpoints, we also have treatment burden, anatomy control, and safety. We also have supplement injections for patients that meet a certain criteria.
We have both naïve patients as well as previously treated. Most of the sites that are part of COMO and CAPRI were also part of our wet AMD study, so those sites, they have experience with phase III studies, and they have experience with DURAVYU, which of course, from an operation perspective, help us a lot.
Ramiro, maybe you can just give some context. I know it's a year away, but what do you think would constitute a compelling clinical profile in DME beyond just non-inferiority? What would we want to see in those phase III trials?
I think the key components are, of course, the non-inferiority, which we have agreement with the FDA to use a non-inferiority margin of 4.5 letters, followed by a reduction in treatment burden, which also is important for DME. Safety, we have a good understanding of safety. We should see the same type of safety profile in DME patients. In addition to that, as Jay mentioned, it's going to be very interesting to assess the effect of the JAK1/IL-6 inhibition that we have with vorolanib in DME patients. This can translate to, first, either a gain in visual acuity earlier than just aflibercept. Then second, which is very important for physicians, is a reduction in the leakage on the fluorescein angiography. We know that long-term patients that have leakage, either on the OCT or on the FA, might have worse outcomes than patients that have dry retina.
Okay, that's helpful context. George, I'm going to turn to you. Just thinking about the balance sheet, you ended the second quarter with about $180 million in cash, and you've got runway into Q4 2027, so you're well-funded here. How do you think about cash utilization following LUGANO? I know you've been investing ahead of the launch, but how do you think about cash spend going forward?
Yeah, no, great question. I think we've got a pretty good track record at EyePoint of managing and controlling our burn. Our cash guidance has been unchanged for over a year. Obviously, with LUGANO missing, a number of things have pushed out, and so we're still very confident in that cash runway is probably a little bit better. But clearly, between now and then, we'll need to add to the balance sheet. We've got a number of catalysts, including LUCIA, including the potential NDA filing and acceptance, and as you mentioned, COMO, CAPRI next year. I think with success in LUCIA, we'll have a number of levers we can pull, including potential synthetic structures on top of equity and other mechanisms like that.
And just on the point of the sort of investment ahead of launch, I know you've been bringing in commercial leadership and kind of building out your manufacturing capacity. Can you talk about sort of the prep work that's going into that?
Yeah. I think what gets lost out there, obviously, there's a lot of focus on the clinical outcomes, and that's clearly important, but when you file an NDA, CMC is just as important, if not more important in the sense that most CRLs happen on the CMC level. We've been in front of manufacturing for several years. We had a state-of-the-art facility built for us by our landlord in Massachusetts, and that has been under scale-up, and we now have batches under stability. We have for registration batches, and we're in the process of scale-up to support that CMC, and the team's just done a remarkable job out there. On the commercial side, it's never too early to get in front of your customers, start talking to payers, and we've built a very small but important team in preparation for launch.
I think the bigger spend on the commercial side will obviously come after NDA acceptance and approval, where we'll start to build that organization out.
Okay.
One more comment, if I may, about the manufacturing. These inserts, there's a lot of technical know-how that goes into making them. The other big effort we needed to do at our facility in Northbridge, Massachusetts, was automate the process. In order to really meet the demands of a successful launch, the inserts really need to be made not by hand, but by machines. Again, George said it, and I'll re-emphasize, our team has done a remarkable job to automate and semi-automate the process from the beginning right straight through to the inspection of the inserts. Formerly, the inserts were inspected individually by hand. A person picked them up with tweezers and weighed them and measured them and looked at them under magnification for impurities.
That's obviously not something that you can scale successfully commercially, and we knew that, and that's one of the reasons we built this facility and brought in the type of automation that's necessary, and we've been able to do that successfully.
Okay, just to give you the last word, Jay. LUCIA is around the corner. A lot to be excited for. What do you think investors will understand with the positive LUCIA data that they may not appreciate today?
I'd like to emphasize the secondary endpoints because that's what's going to make this drug commercially successful. What we're trying to do here is really improve patients' lives by giving them better long-term vision. The pathway to do that is through those secondary endpoints. It's obvious we need to hit a primary to get approval, but we're quite optimistic that with the primary endpoint being hit in LUCIA and the strong secondaries and the clinical explanation for the miss, I think the agency's going to look quite favorably at the application.
Terrific. Well, thank you. We're wishing you success in LUCIA, both for the EyePoint team and for patients. We'll stay tuned, and thanks for joining us today.
Thank you.
Appreciate your time.