Good morning, everyone. Thanks for coming to TD Cowen's 46th Annual Healthcare Conference. I'm Tara Bancroft, one of the senior biotech anaplysts here. Our next session, we have a fireside chat with Ocular Therapeutix. From Ocular, we have Pravin U. Dugel, the CEO. Thank you so much for joining us. It's been a very eventful weekend and morning for you guys, so I appreciate you guys taking the time to chat with us. Before we get started, I wanna remind anyone who's listening here that you guys can and should raise your hand, chime in if you, if you feel so inclined.
I guess to get started, Pravin, can you give us some high-level thoughts on your recent events, the SOL-1 data, the top line, the full data at The Macula Society this weekend, and everything that's happened in the last two weeks?
Well, it seems like it's been two years or 20 years, Tara. First of all, thank you for having us here. It's really a pleasure to be with you all the time, and the opportunity to speak to everybody is fantastic, and answer any questions you have. What we had was a positive outcome for the first time ever in the history of at least my field in retina for a superiority study. We're extremely proud of that. The study has a SPA. The study is unique. It's one of a kind because that was according to the FDA guidelines. It's a study that has never been seen before or done before. It's a primary endpoint that's never been met before. Many have failed trying to achieve superiority to an anti-VEGF. This is the first one to succeed.
What we did know, and what I warned our team about, is that when the study came out and it was positive, there'd be a lot of questions and there'd be a lot of misunderstanding. That's exactly what happened. I'm not surprised with that whatsoever. What we did was we said, "Look, we have answers to all of these questions, whether straightforward or nefarious, we'll be able to answer that with data." In our little company, we spend night and day basically getting the data and presenting that on Friday at The Macula Society. It's in our website. The entire meeting is there unedited, with all the mistakes and everything else because we wanted to be completely transparent.
We have all the slides out there to answer every one of the questions that was raised, I think. People oftentimes ask me, "Why do you have so many slides?" We have 129. I guess I get whacked for either showing too much or not showing enough. We can't win either way, but, you know, it's as much data as anybody can possibly generate. The other question that's asked of me is to say which slide of this, and this was asked several times yesterday, "Okay, you got 129 slides. Give me 5 that are the most important or give me 2 that you really like." It's really difficult to do that. What I tell people is what I'm most proud of is the consistency of the data.
Every single cut that we do, every single slide that's out there points in the same direction in which the superiority of this drug, and hopefully we'll talk about that.
Yeah. I think that's a good point that you bring up that, you know, there was a lot of data, there was some potential confusion, there were some surprising elements of the data. Maybe you could go through your top one or two or even three points that you think there was the most confusion on. I mean, I can name one that is, did the EYLEA control arm outperform or not? Was that surprising? Maybe you could start there.
Look, it's not a... It's a very fair question. It's not an unfair statement. I'm not gonna push back against that, but I am, I guess. I think it's more nuanced than that. If there's anything to be learned, again, this study had never been done before. We've never done one injection on EYLEA or any anti-VEGF and just watched patients lose vision. We did that because that's what the FDA required. It's in their guidelines. It wasn't our study. It was really their study, and thus the SPA. The biggest learning for me was how very long a time it takes to lose a lot of vision. That's really what it was.
When you think about what has to happen, fluid has to build up, and then the fluid then has to go to the center of vision, affect the fovea, then you have to start losing some vision, and then you have to lose 15 letters of vision. That's a heck of a lot of vision. That's 3 lines of vision. Why 15 letters? Because the FDA is not gonna sway from that. That's doubling of the visual angle. They've always insisted on that, and that's never gonna change, right? When you think of that sequence, what you see is the consistency of the data that has come out. Let's start backwards, right?
Arguably, the smallest delta, although we hit primary endpoint with a p value of 0.0006, you know, which is quite amazing. Arguably the smallest delta is what took the longest, which is to lose 15 letters of vision, right? As you go earlier and earlier, the delta gets bigger and bigger and bigger and actually much more meaningful. As you go to less vision loss or as you go to time to first rescue, that delta gets bigger and bigger. Where the delta is the biggest is also where it's most clinically significant, which is the OCT. The OCT results are absolutely spectacular, right? That's, that's the ultimate, the ultimate gauge of disease control. The consistency of what I'm telling you, that narrative, makes complete sense with the alignment of the data.
Let me put it another way. People are so fixated on the delta. Let's just hypothetically say the delta was 50%. People may be happy saying, "Wow, it's a 50% delta." Suppose there was not much of a separation on the OCT. I mean, how do you explain that? You can't. Right? The fact that everything is aligned and consistent really is the most spectacular thing in this database.
Okay, great. I think maybe going into the different subsets within the data.
Yeah.
-would be really helpful. I mean, you showed us the rescue-free curves, you showed us the ITT curves, both look beautiful. I think what would be really helpful is hearing a little bit more about the patients who were rescued and how they did and how you think that subpopulation actually worked out and what maybe you would show in the future to be more informative about how rescued patients perform relative to EYLEA.
Yeah. Again, I go back to saying the amount of time it took for 15 letter loss, right? I would argue that the... you know, whether it be EYLEA or Lucentis or Vabysmo or even Avastin, not the compounded things you get out there, but the Avastin that was used in the CATT study, for instance, which was pure Avastin, would have probably looked pretty much the same. It's just the, it's just the time that it took. As far as the subcuts go, maybe the best way to handle this is to sort of give you our thinking of what we put in there.
Mm-hmm.
Why we put in what we put in in the press release. You can only put in so much in the press release, and we still are a small company. We're not Merck or Pfizer or J&J. There are only so many biostatisticians we have, and we're with the data for a very short period of time before that press release went out. We sat around and said, "Look, how do we represent the drug and what the drug does the best we can in this press release? What do we show?" Right? If we just said, "Look, let's just show ITT," well, it's kind of messy. It's messy because remember, after the first rescue, how patients were treated was entirely up to the physician, and it's all over the place, right? That was at physician's discretion.
It was very different in the US versus outside the US, even within the US. We looked at that and said, "You know, we're going to get criticized for showing messy data." We said, "Well, the rescue-free patients are really the purest form." We said, "Well, if we do that, we're gonna shoot ourselves in the foot because it's gonna be biased against AXPAXLI." There are many, many more patients who are rescued in the EYLEA arm. What you're left behind is the EYLEA patients that have gone through the rescue gauntlet, so they are the best of the best, you know, versus a vast majority of the AXPAXLI patient. I mean, there's no good thing there, we chose to do that because that was the purest form.
Even so, even though it was a biased patient population, the results, in our opinion, were still spectacularly good in terms of what AXPAXLI did, even given that biased patient population. After that, the questions came up of saying, "Well, give us the ITT," and we did. I think what was expected with the ITT, again, remember it's a patient population, for instance, we track that in the slide with the pie charts, and you can see the small number in the Eylea patient versus the majority in the AXPAXLI patient. In the ITT, what I think was expected from most people that I talked to was, look, you're gonna have a vision loss that's quite significant.
This is a patient population that's designed to lose vision, and they were expecting a vision loss of, you know, eight, 10 letters, and it wasn't. It was actually a small vision loss. It not only validated what we had been saying with the rescue-free patients, but it actually validated the tunability of the drug. Remember, this drug, this hydrogel was designed to have its peak effect and last for 36-40 weeks. After that, the effect is gonna be less. That's exactly what you saw at week 52, is that the effect was starting to trail off, which validates the tunability of this hydrogel. It really won on every aspect that you look at as soon as you start understanding the trial and the, and the drug itself.
Most importantly, what physicians wanna know is, "Can you give me disease control?" Right? For that, you gotta look at the OCT. To me, when you look at the OCT and are able to say that in 9 months, with a single injection, that 56% of patients were stable up to 30 microns. I mean, nobody gives you 30-micron information. I mean, that's minute, and that is stability that's unprecedented. I mean, that should be extremely impressive. On the other hand, the other part that also has been, I think, not talked about much is we as retina clinicians look at retinal hemorrhage, for instance. Retinal hemorrhage, to us, signifies aggressive disease. If you look at the baseline characteristics, there were more patients with retinal hemorrhage in the AXPAXLI arm versus the EYLEA arm. Just happened that way.
At the end, at week 52, it reversed. All the retinal hemorrhages were in the EYLEA arm, meaning that the disease control in this aggressive subpopulation was much better handled by AXPAXLI than by EYLEA. You know, that part is also really important.
Okay. Thanks for all that. Well, while we're on the topic of these curves and efficacy, I think there has been a lot of street focus on what a label could look like. You know, I know you guys just hosted a call, but I'm just coming from a really fantastic retina KOL panel, where, you know, the physicians were actually very clear that a label doesn't really matter to them because they already know the context in which they're gonna use this. I'm curious to get your thoughts on that because there really has been a lot of focus among analysts and investors on what the curves and the label are gonna look like. How is the FDA gonna impute this data? You know, what's the disconnect between us and physicians?
Yeah, Tara, it's a good question. I'll tell you a funny story. You know, when I started talking about the superiority label in this, in this drug months ago, maybe a year ago or something like that, I won't say who, but one of the investors played back a tape of me in one of these panels saying, "I don't care about the label. I never look at the label. Label means nothing to me." I was right then, and your KOLs are right, but it just depends on how you ask the question.
Yeah.
Let me just say this. Our hope and expectation is that we will have the first and only superiority label in our field for obvious reasons. We have a SPA, and we've hit the superiority endpoint. Our expectation is also that given how difficult it is to get to hit the superiority endpoint, no one else is gonna try that, not in my lifetime, right? It's just not gonna happen. We'll be in a different orbit, and that orbit will be completely protected for decades and decades. This is understood by a lot of people already because nobody's just gonna try it. It's just way too hard. What does that do, a superiority label?
If you are selling this drug, if you're marketing this drug, what that potentially does is that it makes you immune from step therapy and protects your pricing. That is extraordinarily important, right? Every study that is being done, every study that is being contemplated in our field is a non-inferiority study. It's a me-too drug. You know that there are biosimilars, you know there are generics, all kinds of things. There's a race to the bottom. That's gonna continue. That's only gonna intensify. That's where step therapy comes in, and that's where payers can go ahead and bargain and everything else, and that's where pricing is not protected. When you've got a superiority label, you have the potential to protect your price. You have the potential to be completely immune from step therapy.
That's a really important thing to anybody marketing and developing this drug. Look at it from the KOL side. The reason I answered that, I still would if I was a KOL, and the reason you got that answer is because the question is being asked in the wrong way, right? If somebody were to ask me, and now I'm talking as a physician, "Does a superiority label matter?" I still would say, "I don't care about the label. I've never looked at a label. I know what to do with the drug.
Mm-hmm.
That would be it, but that's the wrong question. The right question to ask them is: Does it matter to you that you get to choose the drug you start your patients with? Does it matter to you that you don't have to watch your patient fail with the inferior drug that you didn't want in the first place and then have to petition three or four times before you get the drug you wanted? Does it matter to you that the ASP of the drug is as high as possible so that your reimbursement will be as high as possible? Who's gonna say no to any of those questions? That's what a superiority label gets you. Just who you ask and how you ask the question.
What data do you have to protect the performance of AXPAXLI in patients who are poor VEGF responders? At the same time, if that's the population that it gets used, even if it performs even better and it falls far short in duration than you would show, so and so powers being rebranded as being less effective in the long term.
It's a great question. I would urge you to listen to, and I know you were here in the morning, but listen to the webcast that we had this morning, as well as go to our website to look at the entire symposium. Look, here's what I would say, and I think you're asking that question very appropriately at several levels is in terms of how patients would respond and perhaps what the label will look like, et cetera, et cetera. Let me try and answer that in a very comprehensive way. You know, in the 30 years that I've been a PI and I've done this, there's not been a single drug, including the phenomenal anti-VEGF drugs that we have, that has actually done better, okay, done better in a community than in a clinical trial.
It's never happened for logical reasons, because you select the best patients in a clinical trial. This will be the very first drug ever, in my lifetime at least, to do better in the community than in a clinical trial. It's exactly the reverse. That's because the patients that we're using here are the most difficult patients to treat. They are, by definition, designed to lose vision. That's who we selected. They're, by definition, the most anti-VEGF-dependent patients. You're catching a falling knife, right? That's what the FDA required because our requirement was to lose 15 letters of vision. Who would we pick? We would pick the very best patients we can in terms of vision and watch them lose vision, right? There's a ceiling effect that's absolutely clear.
We know that we're also using the most dependent patients in taking the drug away. Clearly, patients in the community are gonna do a heck of a lot better when this drug is out there. There's no doubt about it. Everybody knows that. This drug is gonna perform a lot better in the community than it did in the clinical trial, as good as it did in the clinical trial. Let me give you a very clear example of that. If you were to come to my office, and I had this drug today, I may see 50 patients. Of those 50 patients, maybe two would be new patients. Who would I use this drug in? First of all, there's nothing I'd have to do. There's not a single piece of new equipment I'd buy.
There's no change in my workflow. Nothing changes for me. I look at this, and I go, "Wow, I got a better drug." The experience is exactly the same. Mrs. Jones would come in, who's been on EYLEA every 8 weeks or Vabysmo every 8 weeks now for 2 years, and she's frustrated. I say to Ms. Jones, "Ms. Jones, you know, I think I've got a better drug. Nothing changes here. You'll be treated exactly the same way. You're frustrated. You've been coming here for 2 years, every 8 weeks. This drug I think may be better. Let me give you this drug and come back in 8 weeks." Miss Jones comes back in 8 weeks. Go, "Holy cow, this looks great.
You don't need an injection today, Miss Jones." Maybe we'll increase the durability to maybe at 12 weeks, and so on and so forth.
Question. I guess I'm asking to the extent that you have patients that are poor responders to Behçet.
Yeah.
Is the TKI approach, what data do you have to correlate the response to TKIs in poor Behçet responders?
What do you mean by poor Behçet responders?
You're not getting dryness for the patients who are getting injected. You're not able to extend them at all, or you're seeing them back, you're seeing them back on it, or you're seeing them back there is some fluid accumulation.
That's pretty much everybody, right? I mean, when you do Treat and Extend, if you've done Treat and Extend, what you'll see is that there's a certain amount of fluid that everybody tolerates. It's a question of how much you tolerate. Those are the best patients of all because their durability is king. All you've got to win on is extending. Just like example I'm giving you. There's Mrs. Jones who's coming in every at 8 weeks. If you can extend Mrs. Jones to every 6 months, you know what? She's doing back flows. As I go back to what I said to begin with, this drug is gonna be the first drug that's gonna do better in the community than in the clinical trial, period.
Okay. Maybe we can move on to read through to SOLR. You know, now that we've seen the BCVA curves, including the ITT, where do you think we should see... You know, I know the populations are so different from SOL-1 to SOLR, say we take the three letters that are lost at six months in SOL-1, should we expect the same kind of magnitude of letters lost in SOLR, or should that be different? Just maybe some context now that we have-
Yeah.
actual numbers to work with.
Yeah, it's a great question. Look, one of my goals was to give even more confidence in SOL-1, by the way, diabetic retinopathy too, following these results. Now I'm more confident than ever. Again, it all goes back to the patient population. If you just look at the numbers and say, "Look, almost 80% of patients were rescue-free in the SOL-1 patient population," that in and of itself should be really impressive. Remember, we've got the largest non-inferiority margin that the FDA gives, which is minus 4.5 in the lower bounds, right? It gets even more impressive when you look at the patient population. As I said earlier on, we've got a patient population that's actually the antithesis of SOL-1. This is a patient population in SOL-1 that is absolutely anti-VEGF dependent that is designed to fail. That's the opposite of SOL-1.
In SOL-1, we've gone through an extensive 6-month process with 2 periods to look for fluctuations to weed out the patients, right? Even 30-micron fluctuation, they're weeded out. You've got absolute stability before randomizing these patients. If 80% of patients almost are rescue-free in SOL-1 with this patient population, with the ideal patient population, I expect that number to be far bigger than that.
Okay, great. Maybe a little context for the EYLEA HD arms.
Yeah.
How should those perform? You know, just for some context that, you know, there may be some concerns or confusion amongst the street that EYLEA HD might over-perform expectations. What do you say to that?
Look, first of all, to put in context, the EYLEA HD is in the masking arm, right? It's for numeric analysis. The FDA doesn't care at all. It's not for statistical analysis. The randomization ratio is two, and one. We could have chosen anything. We could have chosen water, we could have chosen 2 milligram EYLEA, we could have chosen anything. We chose high-dose EYLEA because we wanted the commercial advantage of being able to go up against high-dose EYLEA. That was our choice. Will it have any effect that'd be different? Not at all. I mean, high-dose EYLEA or Vabysmo for that matter, has not been shown to be superior to anything at all. They're all non-inferiority studies, right?
If you talk to physicians, they'll tell you that at most, these second-generation anti-VEGFs will last for another two weeks at most. We're in a different orbit. We're talking about a drug that's gonna last for up to 12 months. We have no concerns about high-dose EYLEA whatsoever. In fact, we're more confident than ever that it that we'll have a great deal of commercial advantage in being able to show what our drug does against it.
Yeah. I know you missed it, with your call, but the KOLs on our retina panel said that fewer than 10% of patients on EYLEA HD actually get to 6 months. It's very uncommon.
It's a tiny amount. Again, even when you look at the four-month mark, it's mostly people will say it's about 20% or so.
Mm-hmm. Okay. How about let's move on to safety.
Yeah.
We have another question from the audience first, but we'll repeat it so that we can actually hear you.
In SOL-1
Yeah.
patients in AXPAXLI who get a rescue injection, how is that treated in the primary analysis?
The... It's a great question. Remember, the rescue criteria for SOLR is five letters and 75 microns, right? How is that treated? Well, what the FDA has made clear is that their position on how rescue is looked at has not changed for decades and decades. Let me just repeat that position for you. What they've said is that the first rescue is free, right? As long as it doesn't affect the primary endpoint. What they mean by that is that as long as it's not within three months or so of the primary endpoint. The first rescue is free, right? Every rescue after that is up for review. They've made that very clear. Thank you for asking that question. I think it's a really important question.
It gives us a huge advantage in several ways. First of all, remember when they look at SOL-1, the results of SOL-1, first of all, I don't think they'll have to look at that for regulatory purposes because we're submitting with SOL-1 alone, as you know, right? Let's just say they had to. Even if they did that, they would do that knowing that SOL-1 is already positive. They already know that in SOL-1 this drug is gonna last for, you know, 9-12 months. When they review rescues, that puts us in a great advantage, right? The other thing to also notice is that, look, we're not giving a combination drug, right? We're not giving EYLEA with AXPAXLI. Our single rescue that's free is not already used up. That's important. Thank you for that question.
I'm changing my safety question to more rapid fire since we're quickly running out of time. Maybe just a yes or no question on safety. Did you find anything particularly concerning?
Look, absolutely not. I can address every part of what you want. I can tell you about the floaters. Okay. Let me. I'm sorry.
Mm-hmm.
Yes or no, let me just tell you about the floaters to explain that to you. We never thought that we'd get this much, you know, backlash with floaters. Remember Lucentis phase 3 had a 27% floater rate. When you have floaters, you've got to ask, what are they? Where did you see them? Do they affect vision, right? Let me just tell you what they are. We asked physicians to look for the drug when we expected the drug to elute. Simple as that. How did they look? They looked with a indirect ophthalmoscope way out in the periphery where you see retinal tears. This is something the patient would never, ever be able to see. These are all physician-reported way out in the periphery with a indirect ophthalmoscope. Never patient-reported. These aren't floaters. These are drug particles.
This is exactly what we wanted the doctors to see. In clinical trials, we have something called a MedDRA coding classification. We report things with that classification. There is simply no code for drug particles. It doesn't exist. We encouraged them to use floaters because that way we knew there was something there. There's no code for drug particles. There will be a code for drug particles starting next year because of us, but that's the floaters. As you saw in the slides, they occurred exactly when you expect them to occur, when the drug was gonna elute. Not early, not late, exactly when that was gonna happen. They will also go away, we expect, as the drug goes away. We also made it very clear that it had absolutely no impact on vision.
These floater patients were not the same as those who got cataracts. The cataract rate was well within what you would expect with that age, aging patient population. The cataracts were all nuclear sclerotic cataracts, which is what happens, or mostly nuclear sclerotic cataracts, which is what happens as a result of aging. There's really no safety concern whatsoever.
Okay, great. I want to get to understand a little bit more about your conviction in single trial approval pathway. I know that we just saw that editorial come out last week, they had a whole list of a number of different interesting criteria that a single trial would have to meet. Maybe, just quickly, your conviction in meeting all of those criteria, maybe missing 1 or 2, and which ones would those be?
Look, we're more convinced than ever that we will succeed in submitting with a single trial alone, more than ever. That work has already started way back in December. That work continued throughout our quiet period. We didn't take vacations. We were working on this. When we said this in December, people thought, you know, this was just a joke or a PR stunt. Obviously, it is not. You see the loop closed with the New England Journal article. We also said in The Macula Society that process has already begun. Okay? I wanna make it very clear, that process has already begun. We check every single box in The New England Journal of Medicine article. I saw your excellent report this morning.
What I would say, Tara, is the one thing that I would respectfully disagree with is we have no concerns whatsoever about the control arm. I mean, no concerns. It's exactly what the FDA wanted. They wanted to see a 15 letter separation, that's the way to show them a 15 letter separation, period. They would argue that that is part of the standard of care, which is Treat and Extend. This is a trial that has a SPA. If there's ever gonna be a drug in my field, if there's ever gonna be a drug that's gonna succeed with a single trial, this has to be it. It has a SPA. It's hit a superiority margin that nobody's ever hit with a P value that is extraordinarily robust, it's everything has been signed off already with a SPA.
Okay, great. You know, with the final few moments, I do wanna give you a chance to tell us, now that we have the SOL-1 data, is your level of conviction in NPDR changed, the same, different at all?
I couldn't be happier. The answer is I am more confident than ever. It's not just NPDR, it's NPDR and DME. Okay? I just wanna make sure to make that clear. We expect a broad label for diabetic retinopathy and diabetic macular edema because these labels are given per disease, not the clinical trial. It's no different than in my last company when we studied geographic atrophy. We never looked at a single patient with fovea involving geographic atrophy. Yet IZERVAY's label is very broad for all of geographic atrophy. That's exactly what we expect here. We expect never to have to do another diabetic retinopathy study again. This will be a broad label that will encompass all of diabetic retinal disease. Remember, every patient that has diabetic macular edema also has non-proliferative diabetic retinopathy.
Either way, every patient with diabetic retinopathy and diabetic macular edema will be covered.
Okay. Thank you. With that, we're up with time. Thank you, Pravin, for your time, and thanks everyone for listening.
Thank you.