This article was originally published in a sponsored newsletter.
This month, I have the pleasure of interviewing my friend, Dr. Charles Wykoff, who is director of research at Retina Consultants of Houston.
Dr. Wykoff, recently at the Macula Society meeting, you presented data on a new type of intravitreal injection that could help retinal disease. Can you explain the drug and the setup of the clinical trial?
This is one of the most interesting early-phase data sets I have ever seen. Many of the diseases that we routinely treat — namely wet AMD, diabetic retinopathy, diabetic macular edema (DME), and retinal vein occlusion — are defined by pathologic exudation of fluid, cells, proteins, and lipids. Despite the incredible efficacy of anti-VEGF pharmacotherapies, there remains a clear need for new mechanisms of action to restore and maintain a physiologically intact blood-retinal barrier.
We may be able to address this need through the Wnt pathway and a drug called Restoret. Wnt is a highly conserved, complex pathway that is involved in many stages of development as well as homeostasis in adult tissues. There are many combinations of what creates a multicomponent receptor complex, depending on the tissue type. Retinal endothelial cells have a multicomponent receptor complex that includes FZD4, TSPAN12, and LRP5, all of which are membrane-bound proteins. There is a diffusible ligand called Norrin which, when it binds to this receptor complex, leads to clustering of these 3 components. Following appropriate clustering, intracellular signaling leads to transcriptional activation of a suite of genes driven by beta-catenin. Importantly, many of those genes are involved in development and maintenance of an intact blood-retinal barrier. So, the hypothesis is that we may be able to activate this transmembrane receptor complex to drive increased stability or restoration of the blood-retinal barrier.
The idea here was to develop a Norrin mimetic, or a Wnt agonist. This was accomplished by creating Restoret, which is another name for EYE103 (EyeBio). Restoret is an antibody, but it is not a simple antibody with a single target. It is tetravalent, so it binds to 4 different targets, and it is trispecific, so 3 are distinct molecular targets. On one end of the molecule, 2 different epitopes of LRP5 are bound by a diabody. On the other end of the molecule, FZD4 is bound by 2 FABs (fragment antigen-binding), and these components are joined centrally by a modified FC (fragment crystallizable) domain.
The reason it had to be a complex antibody was that the 3 components of the transmembrane receptor had to be brought together to achieve activation. There are preclinical data demonstrating that this antibody achieves beta-catenin transcriptional activity, which is the goal.
Is Restoret a suspension, solution, or liquid?
It is a fully soluble, human biologic. It is similar to what retina specialists are accustomed to using with ranibizumab (Lucentis; Genentech) or aflibercept (Eyelea; Regeneron) in fluid.
Tell us about the trial design and the results.
This was a 2-part, phase 1b/2a trial called AMARONE. The first part was an open-label multiple ascending dose component, in which 12 patients were treated with 4 doses of Restoret from 0.05 mg to 0.8 mg. These were low doses compared to what we are used to with 6 mg faricimab (Vabysmo; Genentech) or 8 mg aflibercept (Eylea HD). Patients had either treatment-naive DME or treatment-naive neovascular AMD. The DME patients were treated with Restoret monotherapy, while the neovascular AMD patients received combination therapy with separate Restoret and 2 mg aflibercept injections given on the same day. All patients received 3 monthly treatments and were followed through week 12. In part 2 of the study, the phase 2a dose-finding randomized component, Restoret was evaluated in 21 additional eyes with treatment-naive DME, all with monotherapy Restoret treatment.
The primary objective was to assess safety and tolerability, and secondary assessments included all the standard retinal outcomes including visual and anatomic outcomes. Currently, we have 12-week data. There is an extension component in which a portion of patients received additional injections, but those findings have not been presented to date.
Regarding results, the baseline demographics and characteristics were as expected for these disease populations. Visual acuities were approximately 20/63 and 20/125 in the DME and neovascular AMD patients, and mean central subfield thickness (CST) was approximately 500 μm in both groups. So, there was substantial disease activity in both the DME and neovascular AMD populations.
Regarding outcomes, 31 of 33 patients completed the study, and there were no rescue therapies given. Across 97 Restoret injections, there were no drug-related adverse events or any intraocular inflammatory (IOI) events, including no physical exam findings that would suggest IOI. The primary endpoint related to safety was met, and there were no safety concerns observed.
The most interesting part of this data set was the efficacy data. The simple summary is there was a longitudinal increase in BCVA for all the DME patients. They gained 7.2 mean letters at month 1, 9.7 mean letters at month 2, and 11.2 mean letters at month 3 with Restoret monotherapy. There were also corresponding anatomic improvements in central subfield thickness (CST) with Restoret monotherapy as well. There was no VEGF inhibition, and mean CST improved from approximately 500 μm at baseline to approximately 350 μm at week 12. When we evaluated each patient's CST trajectory over time, there was a consistent anatomic response with all patients ending with a lower CST than they had at baseline. There was also an 80% reduction in excess CST, or CST greater than 300 μm, at week 12. There was evidence that all doses of sequential monthly Restoret monotherapy injections were effective in DME patients.
Considering the neovascular AMD data, the key difference here is that the neovascular AMD patients received combination therapy. They received Restoret and aflibercept injections at each visit. Because of this, it is difficult to gauge Restoret efficacy among the wet AMD patients, because we know aflibercept monotherapy is an excellent treatment for wet AMD.
Nonetheless, there were no safety signals among the neovascular AMD patients treated, and they demonstrated robust anatomic and visual improvements. Visual acuity increased 6.8 mean letters, and CST rapidly normalized from above 500 mean μm at baseline to all patients achieving a CST of less than 300 μm by month 1, a notably rapid and dramatic anatomic improvement.
These results are impressive. You seem to be able to get great improvements in vision and corresponding decreases in OCT thickness without having a problem with safety. And given that the mechanism of action is different than VEGF, it can probably be considered in combination with an anti-VEGF in clinical trials. If this medicine gets approved, how do you see it being used in your clinical decision-making process?
I think it is too early to know for sure. We have a fascinating signal of biological efficacy with Restoret where we achieved meaningful drying of the retina and improvement in vision, but we also need to understand how Restoret is differentiated from anti-VEGF. We did not inhibit VEGF in these eyes, but still achieved a robust effect with a completely new mechanism of action.
The Wnt pathway is complicated. There are data, for example, showing that one may be able to achieve reperfusion of damaged vascular beds, and some great published work with activation of the Wnt pathway that led to reperfusion within an oxygen-induced retinopathy preclinical model in animals. There is also the issue of stabilizing the vasculature in a way that is different from VEGF inhibition that may create a durability benefit, and there may be other benefits as well.
That said, in a theoretical world in which Restoret was FDA-approved with no additional data, I would try it in treatment-resistant eyes first. Many patients have done well with anti-VEGF therapy, but some have persistent fluid and vision is not optimal, so I would be excited to try a safe, new drug with a novel mechanism of action to see if, and how, I could improve upon the outcomes achievable with anti-VEGF monotherapy.
Was there a difference in the response of wet AMD vs DME in terms of either increasing vision or magnitude of OCT drying?
In DME, the interesting angle was that Restoret monotherapy was used in the absence of VEGF inhibition, and there was quite a robust, consistent response across patients.
In wet AMD, it was a lot harder to tell because there were only 5 patients, and they all received both separate Restoret and aflibercept injections on the same day. Now that we know that this drug had a meaningful effect in DME, I think it would be reasonable to treat additional wet AMD patients with Restoret monotherapy and follow them closely. When this trial was initiated, we were conservative and used combination therapy for wet AMD patients because the last thing we wanted to do was let someone lose vision with neovascular AMD, a situation that may not be reversible. What we can conclude at this stage is, among the 5 wet AMD patients treated, there was no evidence of any safety concern. We also saw dramatic improvement in anatomy among the neovascular AMD patients, and that deserves further study in larger trials, because it could be a signal that the value of Restoret may be additive beyond anti-VEGF monotherapy. Mean CST at baseline was 523 μm, and all 5 patients improved to less than 300 μm within 1 month. These are certainly improvements that could be achieved with anti-VEGF monotherapy as well, and that could be all it is, but it may also be an indication of a more robust signal than what one would expect with anti-VEGF monotherapy. I am looking forward to future trials that will give us more information.
Given the fact that the bar for anti-VEGF is so high with more than 15 years’ experience, it is hard to supplant that. But you sound excited.
These results are fascinating. We have looked at a lot of other pathways — PDGF (platelet-derived growth factor), kallikrein biology, and integrins, to name just 3 — but we have never seen another pathway that has such a dramatic benefit beyond VEGF inhibition. The Ang/Tie2 pathway is still fascinating to me as well, and I think it still holds unrealized promise. Faricimab inhibits Ang2, allowing Ang1 to activate the Tie2 receptor and I think that is a meaningful addition to the field. Looking forward for this biology, there are other molecules in development that may be able to directly activate Tie2.
But of all the specific molecular pathways studied to date, activation of the Wnt pathway in this uncontrolled trial has had the most dramatic effect that I have ever seen beyond anti-VEGF therapies. We still have yet to determine how exactly it will be differentiated with monotherapy or valuable in combination therapy and I look forward to future controlled trials.
This is incredibly exciting, Dr. Wykoff. Thank you so much for sharing it with us. We are excited to see how the path moves forward, but given these promising data, it seems to be a contender for something that might make it to our treatment arsenal in a few years.
