Scott Krzywonos:

The debates at AAO Retina Subspecialty Day have a reputation for examining unresolved issues. Where did the debate fall on whether or not wet AMD tolerates fluid?

Greg Nothstein:

I'm Greg Nothstein, and he's Scott Krzywonos. And this is New Retina Radio from Retina Today and Bryn Mawr Communications.

Dr. Avni Finn joined the show to recap her side of the wet AMD fluid debate and shared what the opposite side had to argue as well. How well does she buy her own argument?

Scott Krzywonos:

And Dr. Darius Moshfeghi joins the show to share findings from the SOL-1 study, which examined OTX-TKI for wet AMD. What did the researchers find, and how did the study authors leverage AI to simulate real-world treatment activity? Join us on this episode to find out.

The debate session at AAO Retina Subspecialty Day included a back-and-forth between doctors, Avni Finn and Anat Loewenstein about whether wet AMD tolerates subretinal fluid. Both are friends of the pod, and today, Dr. Finn is joining us to break it all down.

Greg Nothstein:

Dr. Finn is an associate professor of ophthalmology at the Vanderbilt Eye Institute in Nashville. Dr. Finn, welcome to New Retina Radio.

Avni Finn, MD:

Thank you so much for having me. It's great to be here with you both.

Greg Nothstein:

Before we get started, tell us about how you ended up in the position of arguing that wet AMD does not tolerate subretinal fluid.

Avni Finn, MD:

I was asked to assume this position by our esteemed Sub Day program directors, Dr. Shah and Do, and, of course, I politely accepted this position and honor.

Greg Nothstein:

So now let's turn to the presentation itself. Where did you open your argument?

Avni Finn, MD:

So I initially started my argument by setting the stage for why this is even a conversation. There was a point in our practices where we did not tolerate any subretinal fluid, and I think that there have been some more recent studies that have caused us to reconsider the paradigm of treating patients until dry. One of those studies is the FLUID trial, which looked at tolerating patients with subretinal fluid and how they did two years after that subretinal fluid was tolerated.

Scott Krzywonos:

And you also talked about the CATT study. What were the findings from CATT, this long-ago study that's still very relevant?

Avni Finn, MD:

Yeah, so the CATT post-hoc analyses were a little bit older, but I think they were some of the initial studies that called into question whether subretinal fluid is really that bad. And what these post-hoc analyses of the CATT trial showed was that when we look at two and five-year outcomes, that patients that had some subretinal fluid at baseline did better than patients that had other types of fluid, like intraretinal fluid. So just more support for whether or not we should really think of subretinal fluid as all that bad or not.

Scott Krzywonos:

So the CATT and the FLUID trial data suggests that maybe some subretinal fluid is tolerable, but in making your argument, you were a bit more skeptical. Tell us where you turn to next in your discussion.

Avni Finn, MD:

Sure. So I think it's important to think about this from the perspective of the data and really look at what CATT and the FLUID trial told us. And although the FLUID trial did show us that some subretinal fluid could be tolerated, it did not test unlimited tolerance of subretinal fluid or really tell us that subretinal fluid is safe in the long term. So some key points that I made in the debate was that in the FLUID trial, there was only fluid tolerated when it was less than 200 micrometers

It did not establish whether subretinal fluid was safe in the long term. And it merely showed us that in some select patients with wet AMD, small amounts of subretinal fluid can sometimes be tolerated without sacrificing vision at year two. We don't know what happens after that in these patients.

Greg Nothstein:

SSo during this session, your main argument was that reduced subretinal fluid is linked to improved vision. What's the evidence there?

Avni Finn, MD:

So there's some good evidence for this in large analyses. So one study showed that if you look at about a thousand eyes that are treatment-naive, and this was a postdoc analysis of the HARBOR trial, showed that decreasing subretinal fluid actually led to improvement in vision. So in that particular analysis, a 100 nanoliter decrease in subretinal fluid led to about a six-letter increase in vision. So again, building some evidence for us that not all subretinal fluid is benign and that treating subretinal fluid can, in fact, improve visual function in patients.

Scott Krzywonos:

Talk to us a little bit about the HAWK and HARRIER study, which talked about fluctuations in fluid and the potential effects that has on visual function.

Avni Finn, MD:

Yeah, Scott, I think that's a really important point. We have more and more evidence that fluid fluctuation is really important to vision. And in particular, this postdoc analysis of HAWK and HARRIER showed that when we had residual fluid after loading doses of treatment, whether that fluid was intraretinal, subretinal, or a large PED volume, that was independently associated with progressive vision loss during the maintenance phase. So that residual subretinal fluid did make a difference to long-term visual outcomes.

Greg Nothstein:

All right, Dr. Finn, we have to know, though, you presented these arguments, but is this how you would treat your patients?

Avni Finn, MD:

Well, hopefully I presented a really good argument, but I have to say this is not always how I treat my patients. I do think that treatment has to be individualized and that there are some patients in whom subretinal fluid can be tolerated. In my particular practice, there are patients that you treat very frequently, and they will always have some residual subretinal fluid and just a small amount of persistent fluid and very constant, steady vision.

Those are the patients who I think it's okay to treat at a continued interval and tolerate that small amount of subretinal fluid as long as you're not seeing large increases in that subretinal fluid or decline in vision. So I do think it really has to be personalized to the individual patient, what you're seeing on OCT imaging, as well as vision.

Greg Nothstein:

Very interesting to see both sides of the coin here in part of this presentation, but Dr. Finn, we want to thank you so much for joining us here on New Retina Radio.

Avni Finn, MD:

Thank you for having me.

Scott Krzywonos:

Retina specialists have been trying to extend treatment intervals for wet AMD patients for two decades. They've had some success, but a major change could be on the horizon. Dr. Darius Moshfeghi shared data from the phase three SOL-1 study, which assessed OTX-TKI for wet AMD at the AAO Retina Subspecialty Day this year in New Orleans. He's here to tell us more.

Dr. Moshfeghi is the Chief of the Retina Division and Professor of Ophthalmology at the Stanford University School of Medicine in Palo Alto, California. Dr. Moshfeghi, welcome to New Retina Radio.

Darius Moshfeghi, MD:

Thank you, Scott. It's always a pleasure to be back here with you.

Scott Krzywonos:

So before we get to the SOL-1 results themselves, we should note something distinct about the study, which is that it's a superiority study. It is not a non-inferiority study. What does that mean, and why does it matter?

Darius Moshfeghi, MD:

So when you look at benchmark clinical trials, you're basically having two components to that. One is a superiority outcome, and the second outcome that is of relevance is that you're beating visual acuity at 15 letters in macular degeneration. And that's important because if you go all the way back to the original two trials that have guided the anti-VEGF error, which are MARINA and ANCHOR, MARINA was the last one to go up against the placebo arm. And then, of course, ANCHOR went up against photodynamic therapy.

And every subsequent anti-VEGF trial up until the present day has been a non-inferiority outcome. And we often forget or confuse that non-inferiority may be the same as equivalent or equal. In fact, it is not. They have a non-inferiority visual acuity outcome, which is typically on the order of minus three to minus five letters at whatever your assessment time period is. So if you have minus three or minus five letters at 52 weeks, every 52 weeks, you could potentially continue to lose vision even though you're "non-inferior".

And what we've seen as we extend our treatment-free intervals, we go from four weeks to six weeks to eight weeks out to 14, 16, 20 weeks, we have a concomitant decrease in visual acuity over time, and that's baked in. Every non-inferiority trial will have that unless it's a true equivalence visual acuity outcome. And so the opposite is occurring in superiority. Even if the visual acuity is decreasing over time, you know it's superior to your comparator arm. Okay.

If you have to meet that metric of 15 letters, whether it's maintenance of vision, less loss, or actual gain of 15 letters, and the FDA allows you to choose any one of those three metrics, superiority means that you're better than the other outcome. Treatment-free intervals have increased with ranibizumab from four weeks all the way out to 24 weeks with ranibizumab in the Port Delivery System. But over that same time, we've seen as you go out two, three, four, five years, you have anticipated visual loss and our patients are giving up all their visual acuity gains.

I like to say that by extending the intervals when we're using these non-inferiority trials, all we're doing is really creating a soft landing in the legal blindness as you start moving out to three, and four, and five years.

Scott Krzywonos:

Yeah, from that perspective, it's kind of a lousy minimum viable goal.

Darius Moshfeghi, MD:

Absolutely. And like I said previously, I've never had a patient that says, "Give me the non-inferior medication, Dr. Moshfeghi." They always want the superiority drug, right?

Scott Krzywonos:

So let's move on to SOL-1 itself. Can you give us a quick primer on just the design of the study before we get into the data?

Darius Moshfeghi, MD:

Sure. I mean, famously, this trial had stringent criteria to get in, and that was a safety measure. Patients had to be proven responders, both with respect to visual acuity, they had to be able to get to 20/20 or gain 10 letters of visual acuity, and they had to be able to dry out the retina dehydration from greater than 350 microns, and this was after two loading doses of aflibercept. So you started at time zero, which was screening, and at baseline, patients had 350 microns and 70 letters, and they both had every four weeks for two doses, they received aflibercept

And then at screening, at that point, it was a randomization. And if you had met those two criteria, you had dehydrated the retina and you had gotten either 20/20 or gained 10 letters of vision, then you were eligible to be randomized. At that point, you could only receive either ocular therapeutics or aflibercept, and then you were followed for as long as you could go with rescue criteria, either being loss of 15 letters of vision or a vision-threatening hemorrhage.

Scott Krzywonos:

And that loss of 15 letters of vision, once they hit 15 letters, that was considered that they've lost vision. So a patient could have lost, say, seven letters over the span of 30 or 36 weeks, and they would have been still-

Darius Moshfeghi, MD:

Yeah, accumulated visual acuity loss.

Scott Krzywonos:

Yeah. Okay. So let's get to the data itself then. What did SOL-1 find when it came to data?

Darius Moshfeghi, MD:

Top-line data was superior at the primary endpoint, which was 36 weeks. And so they had 74% versus 56%, and this was highly statistically significant. And when you look at week 52, again, there was a 50% treatment benefit. It was 66% versus 44%.

Scott Krzywonos:

Are there any theories as to why patients could go so long without losing vision as defined by the study?

Darius Moshfeghi, MD:

Yeah, so why is it durable? So these other medications haven't been, and there's nothing special about TKIs themselves that would increase their durability, but remember, it's embedded within a hydrogel platform. And so you're getting continuous release of the TKI, fresh TKI, which is causing a sustained VEGF inhibition. And remember, TKIs work intracellularly, so it takes a little while before you get to that steady state, somewhere around four weeks or so, which is why we loaded them up at baseline with those aflibercepts, and then we checked to see if they were responders. But at that point, once you get to steady state, you should have this sustained VEGF inhibition effect, which could benefit both function, visual acuity, as well as anatomic findings such as dehydrating the retina.

Scott Krzywonos:

Got it. So giving them that loading dose, it gives them a running start by the time they experience the TKI exposure.

Darius Moshfeghi, MD:

That's absolutely correct. And because randomization did not occur until after those two loading doses, randomization is when you start your tracking of visual acuity. So the company understood upfront that all they could really do was lose visual acuity because they had gained the vision, then they randomized, which is different than every other clinical trial, which started at a baseline, and then they experienced all the visual acuities.

But still, at the end of 52 weeks, they had a net gain of ... I mean, a net loss of three and a half letters, but they were still up six and a half letters from the time of screening, because it just depends on where you measure. From the patient's perspective, they all gained vision.

Scott Krzywonos:

You mentioned the hydrogel. I think it was visible in a significant number of patients over the course of the study, right?

Darius Moshfeghi, MD:

That's a great question, Scott. So, at six months, 90% of the patients had visible hydrogel, and at nine months, about 50% of the patients had visible hydrogel, which is why we believe that they were continuing to have this sustained effect through 36 weeks and actually continuing out to 52 weeks.

Scott Krzywonos:

We started this conversation talking about how longer duration treatments could extend intervals, and I know that the SOL-1 study looked at the interval to re-treatments. What did you guys find there?

Darius Moshfeghi, MD:

That was probably the most remarkable thing. Treatment benefit basically mirrored the primary outcomes as far as durability goes. It was about 75% versus 56%. However, the time to first rescue showed a six-month benefit at week 52 over aflibercept, which is shocking because currently the longest approved second-generation anti-VEGF that's out there is only approved at 16 weeks. And so you're talking a 24-week benefit over that.

Scott Krzywonos:

Yeah, I mean, that could be significant for patients if it were to get into their eyes in the real world.

Darius Moshfeghi, MD:

And I like to think of it as a safety issue, Scott. We have patients, they go out, they have family, they travel, they get sick, and maybe they're on a 14 or 16-week interval, and then they miss their appointment. They're now at risk for a hemorrhage or atrophy or a retinal pigment epithelial tear. They miss for two or three months. Here, we have dosing that we know that 66% of patients are going to make it to a year. Okay. 75% of those patients are going to make it to 36 weeks. So we have a lot of confidence that we can miss a month or two here for a patient that we don't have to really worry about, are they going to have this really catastrophic outcome?

Scott Krzywonos:

What you hear about all the time, we use the term non-compliance, which, in a way, makes it sound like the patient decided to not turn their homework in on time, but sometimes these patients get sick, they're hospitalized, they can't come in for their appointment. Somehow life gets in the way. There are various reasons why they don't show up. Sometimes the doctor's not available, sometimes the doctor's at AAO and isn't able to schedule them at that exact interval.

Darius Moshfeghi, MD:

Yeah, life happens.

Scott Krzywonos:

The study used AI in two very interesting ways here. So you and your colleagues used it to measure dryness, and then also to estimate how real-world treatment patterns would look if OTX-TKI were to make it into doctors' hands. Tell us about the first one. Tell us about using AI for dryness measurements.

Darius Moshfeghi, MD:

Famously, the SOL-1 trial did not use re-treatment based on the presence of fluid, which is what all of our other clinical trials have done, and this was causing a lot of consternation in the retina community. And so one of the arguments against the trial was this is not how we treat patients in clinical practice, so how relevant is this data?

And we thought about that, and I think it's a fairly reasoned approach and critique of the clinical trial, but we had met with the FDA, and this is the way that they had said that we should design our trial. And so you have these competing tensions, what retina specialists do in their own practice and what the regulatory agency is going to allow you to do in order to achieve a certain label indication. So what we did is we looked at what percentage of our patients are going to do well if they're dry. But here's the problem. If you put three retina specialists in a room, they're going to come up with dryness in different ways.

And so when we looked at the literature, there's an evolving consensus that's coming out of the home OCT monitoring where, if you look at the amount of intraretinal fluid and the amount of subretinal fluid and you add them together and they exceed 10 nanoliters, that's an inflection point that you would either bring the patient in for evaluation or you would actually trigger a treatment, and that has now been adopted by the DRCR. So what we did is we defined 10 nanoliters or less as quantitatively dry, and then we did an analysis of all the patients in both arms, and it turned out that 70% of the OTX-TKI arm was dry eight weeks after the randomization, and 75% of the aflibercept arm was dry at the same time point, which is not surprising because remember I said there's a delay in the intracellular action of the continuous release OTX-TKI, whereas aflibercept is this magnificent anti-VEGF that works almost instantaneously, so it gets that immediate effect.

But what we see is that over time, that 70% in the OTX-TKI arm stays dry out through 36 weeks, and the numbers are pretty remarkable, right? If we look at the percentage of patients in the OTX-TKI arm, 75% and then 68% maintain that dryness, whereas you saw a pretty steep dropoff in the aflibercept arm, only 60% of the original 75% and less than half at week 36, which argues for why the durability that's conferred by the hydrogel platform and the continuous release.

Scott Krzywonos:

Yeah, it's certainly interesting. I think it's a very clever way to assess dryness because, like you said, you could put a handful of retina specialists in a room and everyone has their own definition of dryness, but if there's an emerging consensus that a negligible amount of intraretinal and subretinal fluid collectively, and that threshold is 10 nanoliters, if that's what a real-world specialist would consider dry approximately, I think, yeah, those numbers are really impressive to get up to 36 weeks and have more than two-thirds of patients on the OTX-TKI drug dry, quantitatively dry, seems pretty important to me.

Darius Moshfeghi, MD:

One other aspect of it is that in the treat-and-extend regimen model, if you will, people believe that if you are treating to always maintain the patient dry, let's say at a 16-week interval, then you will achieve maximal visual acuity. However, we know from the real-world data that as we go and extend these treatment intervals further and further out, we see we get continuing vision loss. It's pretty linear over time, which the treat-and-extend model is based on the concept that there's no benefit to treating earlier when it's dry.

However, when you look at the fluid curves in the always dry, the quantitatively dry group for OTX-TKI and compare it to the quantitatively dry aflibercept group, which is bolus anti-VEGF, you see that there's a stabilization of the visual acuity, whereas for the OTX-TKI arm, whereas on the aflibercept arm, it dips down and then gradually recovers, which argues that continuous release is giving you a vision benefit despite the retina being dry. To me, that is like the single greatest argument in favor of continuous release.

Scott Krzywonos:

Give us a quick safety update. Anything to note?

Darius Moshfeghi, MD:

No. I mean, the most common findings were floaters, and they occurred around the time that we would expect the hydrogel to begin to break up and the drug to continue to elute. There were no visual acuity outcomes associated with that. They had nine cases of inflammation, seven of them in the OTX-TKI arm. None of them were occlusive vasculitis. There were no cases of endophthalmitis. They were either mild or moderate, and some resolved pretty quickly without visual sequela.

Scott Krzywonos:

What should we be keeping an eye on? What's next for OTX-TKI?

Darius Moshfeghi, MD:

Well, we're continuing to do a refined analysis of the effect of fluid on this, and then we're awaiting the results of SOL-R, which will come out sometime in 2028, and that will be exciting.

Scott Krzywonos:

All right, so we've got a little bit more than a year before we hear about SOL-R.

Darius Moshfeghi, MD:

In the meantime, I know they're preparing their submission for the FDA, so that'll be exciting, too.

Scott Krzywonos:

Interesting. So they're going to use a single study submission to FDA rather than the old-school two-study approach?

Darius Moshfeghi, MD:

Yeah, and it's based on they had the SPA, the agreement in advance with the FDA, they beat a superiority

Scott Krzywonos:

The Special Protocol Assessment. Yeah.

Darius Moshfeghi, MD:

Correct.

Scott Krzywonos:

Or agreement. Yeah.

Darius Moshfeghi, MD:

They adhered to the terms of the SPA. They beat a superiority outcome against a clear benchmark standard, and that standard was superiority with respect to visual acuity at 15 letters. And so this is a very compelling narrative.

Scott Krzywonos:

All right, then all eyes on FDA.

Darius Moshfeghi, MD:

Can I bring up one other point that may have gotten lost in this? When you look at that quantitative dryness, people are saying, "Well, in the real world, we're going to treat for fluid," and I get that. But if you're a fluid-based re-treater, you see the separation of the 70% and the 30% immediately at eight weeks. They break off. And so when you look at those curves, the people who are never dry immediately go shooting high up, and that occurs both in the OTX-TKI arm and in the aflibercept arm, which goes to show that 30% and 25% of the patient population tend to be fluid gainers and always have fluid around.

Scott Krzywonos:

And make themselves known early.

Darius Moshfeghi, MD:

Exactly. And you can clearly differentiate them. So when people come up to me and say, "Well, does this only apply to this highly selected population that you used in this study?", my rebuttal is I would treat the entire field of my patients if the drug were approved because I could clearly identify which ones would respond by eight weeks.

Scott Krzywonos:

Yeah, that is certainly interesting because that's something that you and your colleagues have talked about struggling with in the past. How do you know what patients are going to respond to which treatment? You see it a lot in diabetic eye disease, right?

Darius Moshfeghi, MD:

Sure.

Scott Krzywonos:

Is it a VEGF issue? Is it a cytokine issue? Is it some other inflammation issue? And now it looks like we're going to start having some of that discussion in wet AMD also.

Darius Moshfeghi, MD:

But when you apply this strict criteria of that quantitative dryness, it's very clear that that separation occurs at eight weeks.

Scott Krzywonos:

Dr. Darius Moshfeghi, thank you for coming on New Retina Radio.

Darius Moshfeghi, MD:

Scott, always a pleasure. Thank you for having me.

Greg Nothstein:

That concludes our first episode of coverage of AAO from New Orleans just a few weeks ago.

Scott Krzywonos:

We have one more episode coming up, so be sure to subscribe to the show and look out for it in your feed.