I'm one of the senior biotech analyst at Evercore ISI. With us in attendance is Victor Perlroth, CEO of Kodiak.
You got the name wrong.
Kodiak, I'm sorry. Kodiak. Victor, thanks so much for making time for us and for being here. Before we delve into Q&A, maybe just kick things off with some high-level business highlights and look forward to in 2024.
Maybe remind us, because there's a lot of catalysts coming up, because you guys kind of went catalyst-free for a while. So just remind us all the events that are coming up.
Right.
Yeah. Ok, well, thanks, Mike and Umer. Thanks for having me at the conference. Appreciate it. Yeah, it's an exciting and wonderful time at Kodiak, actually. I think a lot of hard work over the past six, 12, 18 months is really coming together. And we're excited about some of our new science, and we're excited about some of our new studies, and we're excited about the operational progress where we are today. I think it's an exciting time for investors to focus on Kodiak. To answer specifically your question, Umer, I think we are coming right up against completing enrollment in our GLOW2, phase III pivotal study for Tarcocimab in diabetic retinopathy over the next couple of months. We should look for completion, last patient in for GLOW2. That will provide the timing with a 48-week endpoint in a few weeks.
My apologies, Victor. Can I just sort of pause? Just because some folks are refreshing on the story. Could you just remind us what's the indication for each of these trials?
Yeah, so the GLOW2 is for diabetic retinopathy.
OK.
Patients that have diabetes in the eye. It's a nice repeat of the GLOW1 study that we can talk about. Soon we'll be last patient in for GLOW2. Within a few months after that, we'll be last patient in for the DAYBREAK study.
OK, so GLOW2 readout would be when?
That'll be about one year from last patient in.
Late next year.
So if we do last patient in towards the end of this year or early next year, 2025, then the readout would be, say, a little bit under 52 weeks from that time point.
GLOW2, late 2025, early 2026, phase III registrational.
Correct.
OK.
For two indications, diabetic retinopathy and retinal vein occlusion. And then we'll be following within a few months, last patient in for the DAYBREAK study, which is for Tarcocimab and also for KSI-501, which is in treatment-naive wet AMD patients. So another catalyst the way I see it.
That's a mid-2026 readout.
So that'll be Q2 2026 readout.
Got it. And I think what's critical, again, I'm just trying to make sure everyone's caught up on what the setup is. This is not the same drug repeated again on these. There's been changes to the formulation. And it's not just a formulation change like inactive ingredients. You guys changed the conjugated versus unconjugated mix. So could you just remind us of that?
Yeah, I think that's important. I mean, obviously, what's really special is it's not a new drug. It's a subtle but important change to the prior drug. So we're able to use the three successful phase III pivotal studies that we ran with Tarcocimab, including the safety database. So that's really important. Having said that, I think the change we've made by having a mixture of the unconjugated form and the conjugated form allows us to bring the key durability elements that we've seen with Tarcocimab with our ABC Platform pretty much in every study that we've run. But it also, the change that we've made by bringing in the unconjugated protein allows us to solve sort of what we call this immediacy gap, where a drug that's built to be large, that's built to drive the durability, took some time for the action.
So by bringing now some of the unconjugated form in a defined ratio, and we can talk about that, allows us to drive that immediacy, which will, I think, allow us to bring both high efficacy and immediacy and high durability as a biologic.
Got it.
So it allows us to really reboot, I would say, entirely, really the company and the technology. And we applied that to Tarcocimab and also into the KSI-501, which is also a conjugated form.
So as a base case, if we just assume that Tarcocimab, based on the prior trial, underperforms Eylea, let's just take that as a base case, that it underperforms in the sense that immediacy of onset and maybe OCT clearance is not as profound. So let's just say that's the base case. Even in that base case, the fact that you also now have some unconjugated form, could you remind us what is that equivalent versus the Eylea dose? Because maybe that's one way for us to start to say, well, there's a variable, and then there's something we know for certain. So maybe let's start with that.
Yeah, absolutely. Well, I think the way we designed, say, the Tarcocimab enhanced formulation, I call it the one plus four, or five milligram total dose of antibody, in which one milligram is now unconjugated and four milligram remains conjugated. So to focus to your question on the unconjugated portion, which drives that immediacy, one milligram of unconjugated is the same as 1.3 milligrams of Eylea in terms of VEGF binding. So the two milligram Eylea dose is the labeled dose, traditional. And it's been tested at a 0.5 mg dose and the two mg dose. And the 0.5 and two were essentially the same or very similar. And we have a 1.3 mg equivalent just of unconjugated. Another way to think about that unconjugated is to say one plus four, so the one mg of unconjugated, is equivalent to 0.7 mg of Lucentis.
Lucentis is marketed dose for wet AMD is 0.5, so we have a super dose of Lucentis plus all of our conjugate, and another way to think about it is Lucentis is marketed dose for diabetes in the eye is 0.3, so we have more than a 2x just of unconjugated, so we have a very powerful drug for immediacy, and we have a very powerful drug for durability that we've shown. That's now combined as our enhanced formulation for Tarcocimab, and then we applied that same framework into our 501 molecule, which is the bispecific conjugate, which is also being tested in the same DAYBREAK study.
What's the FDA feedback? Like, let's say we get this GLOW2 readout late next year. In this trial, you replicate, you're able to hit all these objectives on immediacy of onset and efficacy. I guess, how does FDA pull the data? How do they think about their registrational filing?
Yeah, we engage with them in a meaningful way. So their feedback was to run one additional phase III study with the enhanced formulation. And we decided to take that one step further. So we decided to run two new, two additional phase III studies. And the DAYBREAK study is also a two-year study. So we'll be able to use these two new studies plus the three prior studies to build a package for BLA, which would be for three diseases in the initial filing.
Could your immediacy of onset look better in DME indication versus the wet AMD indication?
I think our view is that it's always going to be helpful to have some unconjugated antibody present for all of the diseases. I think the immediacy gap that we noted is in wet AMD, and our immediacy in DME and RVO and in the diabetic retinopathy was strong, solid, and appropriate. But it can always be helpful because the spectrum of disease severity across patients is very broad in each disease. So you'll always benefit from that immediacy. The other nice thing about having a little bit more immediacy is the way that retina docs treat patients is they often let the patients get treated, and then they let the disease recur, and then the patients come back. So in some way, having immediacy each time will allow you to always better recapture the disease before the durability of the conjugate comes online.
Got it. My last one, and I'm sorry, Mike, I know you had a question too. My last one really was, it looks like in spite of your prior phase III failures, your recruiting went just fine in the new trials. And is that because people knew that you now have some unconjugated with conjugated and unconjugated is going to solve for it? Or is it because of some of the duration data that we saw from previously?
We did the R&D, kind of the capital markets day at the end of September, and we had two top retina physicians there, Dave Brown and Charlie Wyckoff. I think one thing that's helpful to do is for people to go and watch that and to listen in their own words. The retina community and the physicians who know even the Legacy Tarcocimab were excited about the durability and the potency and the efficacy in the majority of the patients. They were looking for more immediacy if it were possible. When we shut down some of the studies previously, patients were upset because the patients who were doing well, which was the majority, they didn't want to get off the medicine. Now we're actually rebuilding and delivering them a better medicine.
I think the physicians, when they learn about where we are, are excited about it. The patients are excited about it. The GLOW2 is really a repeat of the GLOW1 study. We add an extra loading dose, which is only beneficial. And the DAYBREAK study design really brings a lot of new art to the design of these phase III clinical trials in wet AMD. Patients and physicians are excited about that design. So all patients should be able to do very well.
Got it. And do you have any clinical data with this one to four unconjugated to conjugated mix and confirming that it solves for the immediacy of efficacy onset? Do you have some clinical data?
We haven't run that as, like, a separate, say, open-label phase I study. I mean, obviously, we are enrolling. GLOW2 is almost fully enrolled, and DAYBREAK is enrolling well. So what we have visibility into is mass data, and patients are doing well. But we didn't run a specific open-label study.
Did you have that visibility in the prior trial that they weren't doing as well at the immediacy?
I guess I would say if you feel immediacy is going well, what you see is a consistent pattern. In light of all of the many thousands of patients' data that we've accumulated in pivotals now over the years, I guess we may have more insight now than we did before. Previously, if you have variability, you might just attribute that.
You were seeing variability on a blinded basis previously.
Yeah, I would say that. The reality is in wet AMD in particular, disease response is highly variable in general.
And you were seeing variability on BCVA or on OCT or both?
I would say you see this more on OCT, which is really sort of a biomarker.
On OCT. And so you were seeing on a blinded, pooled blinded basis, you were seeing variability. This time around, you're not seeing variability?
I mean, I don't think we can go into too much detail on the mass data. But in general, I'm feeling, I think I feel.
This might be the most important thing on Kodiak I've learned in a long time.
I think with all of the learning and all of the changes, and on top of that, some of the new science of the aqueous taps that we took in the patients and the half-life that we're seeing in humans in retina, all of these things together give me a much higher level of confidence, I guess, than I've ever had that Kodiak can really deliver this high efficacy, high immediacy, high durability profile with our ABC platform. I've never quite been able to really have this level of, I suppose, confidence may be a strong word, but.
Can you take an interim to confirm this immediacy of onset and the OCT depth of response? Or don't need that because the readout's pretty good?
We don't think we need to because the readouts actually were quite close to the enrollment target dates for last patient in. Then it's just basically the arithmetic to get you to the 48 weeks for the. We have the capital to get us to the endpoint, but probably we'll raise some capital in advance of that. I think we're, and not only that, we've just spoken about Tarcocimab and KSI-501, but of course, we have a very interesting program with the bispecific protein alone and the inflammation and that program in our APEX and PEAK and PINNACLE. It's also going very well.
I'm going to go mute, Mike.
One last question on Tarcocimab, Victor. Just curious, have retinal docs expressed an interest in using Tarcocimab in DME? I know your BLA should get approved. It won't encompass that specific indication. But since there is some overlap between DME and DR, maybe talk about that dynamic and what doctors are saying.
Right. Well, physicians are excited about using the drug in DME patients. And the approval that we would hope to get from the GLOW1 study and the GLOW2 in diabetic retinopathy, that will allow physicians to use the drug in DME because every DME patient has diabetic retinopathy.
Actually, the idea of adding another loading dose early in the GLOW2 study, which we didn't do in GLOW1, so now we have three monthly loading doses in GLOW2, that will be very helpful for patients that may have more severe disease, for example, DME patients, for example, although that's not really the target. But the point is it would allow physicians to be reimbursed if they needed and wanted to use three loading doses in more sick patients.
Got it. Got it. Moving on to the DAYBREAK study, focusing on KSI-501, which is bispecific. Just think about the design of that study. Again, it's Tarcocimab as one arm, KSI-501 as a separate arm, and then Aflibercept as a comparator arm. You guys employed unique retreatment criteria, using intraretinal and subretinal fluid as a disease marker rather than vision loss and CST. Again, this is the first trial to ever use this criteria, correct? And was that a kind of a battle with the FDA trying to get them to accept that?
Well, up until now, as a community, we didn't have technology to implement a trial that uses fluid as a biomarker for disease activity in a consistent and reproducible way. So instead, historically, we and other sponsors used sort of a coarse surrogate, the CST or the central subfield thickness. But in wet AMD in particular, the pathology in the eye or the anatomy is highly variable. It's quite horrible. And so CST is actually fairly unreliable, but it was the best thing. So fluid is the best and quintessential biomarker for disease activity in clinical practice. So we like using it directly because it resembles the way that physicians treat patients in their clinics and will treat patients, let's hope, with Tarcocimab if it's approved. And it accomplishes two key goals. One is it will avoid undertreatment of patients, which we don't want.
And at the same time, it will allow us to really safely extend, hopefully, a large number of the patients who can be extended to really demonstrate strong durability. So that's great. And in addition, we have had very productive and positive interactions with FDA on the use of the AI-guided tool, basically.
Got it. Just thinking ahead here, I mean, if there is no discernible difference between Tarcocimab and KSI-501 on the BCVA primary endpoint at one year in the DAYBREAK study, where both proved to be not inferior versus Eylea, is it fair to say that at that point, if that situation pans out, will the subgroup analysis become even more important to see how the IL-6 component may benefit different subpopulations?
Right. So DAYBREAK is a great study. I think it's innovative because we're bringing two medicines against Eylea in wet AMD. So that's efficient and very useful. So the objective for Tarcocimab in DAYBREAK is different than the objective, what we want to explore with KSI-501. So with Tarcocimab, we want to really showcase the immediacy and the durability. But with KSI-501, we decided, well, most people, when they're bringing, because it's a dual mechanism of action, as you mentioned, with the anti-VEGF and the anti-IL-6. So a lot of times, historically, people have tried to explore the benefit of a second mechanism by trying to push durability. But we have the KSI-501 already in our ABC Platform that's going to bring that durability, and we have the immediacy with the unconjugated. So we thought, well, why don't use DAYBREAK to show durability?
Why don't we actually, and the retina community actually suggested it also, is they really tried to show whether this dual mechanism can give you better or stronger efficacy in some manner as the objective. So therefore, we're dosing all the patients every eight weeks or even through the disease activity, the AI-guided tool, maybe even monthly if the patients need that. So we're giving the patients every opportunity and the drug, KSI-501, every opportunity really to show something special. But we only need to do it in the context of a non-inferiority outcome because the goal is to get the drug moving towards it. So I think we don't know whether or not it's possible to show superior efficacy because that's really hard in retina. But we're giving the drug every opportunity to do that.
If we don't see something superior or if we don't see the trend towards something separated in terms of the primary endpoint of mean change, to your point, looking at different large subgroups and trying to really showcase the benefit in these different patient groups will be very useful, and that's pretty similar to kind of what Roche has been trying to do with their drug, Vabysmo, where they didn't show any type of superiority, but they're still on the podium really doing a lot of sub-analyses post hoc to really showcase, oh, well, the drug is really special in this large subgroup of patients.
Got it. And I guess, is DAYBREAK power enough to show stat-sig differences among different subgroups?
I mean, probably not. It depends how powerful the effect would be in that group.
I see. I see. When we think about just how the efficacy of 501 and even 101 compares versus, say, Roche's IL-6, how should we think about that? I mean, because the fact that Roche is only proceeding with their two lowest doses, I mean, how should we think about 501 and 101's efficacy safety platform in that sense?
Right. Well, I like to think about 501 more like the Tarcocimab and Vabysmo and Eylea in the sense that we're really exploring it with Eylea as a comparator and looking at it in the retinal vascular diseases like wet AMD and stuff. The KSI-101, which is the unconjugated protein alone, where we're exploring it in what we call macular edema secondary to inflammation. So against sham in a 16-week primary endpoint. So it's a very different set of patients, very high unmet need patients. These are very sick patients. And physicians tell us they have a lot of disease, and they really want to have a very powerful medicine at a very high formulation strength like a sledgehammer. And so we formulated 101 up to 100 mg per mL, and we're exploring in the APEX study 2.5 milligrams all the way up to 10 milligrams.
And in the Roche Vamikibart program, they did show in their DOVETAIL study that they had a bit of a dose response. So the 0.25 milligram was their lowest dose, and 1 mg was the next. And I think they went up to 2.5 and maybe 4. But there was a bit of a dose response. But I think for reasons of safety and some challenges that they may have seen in the DME indication, for example, or maybe in some. They actually chose their two very lowest doses to progress into their phase III pivotals for that molecule. And so because they did show dose response, that means that they're really on the low end of that efficacy curve because they chose their two lowest doses. So I think they did show strong bioactivity of their molecule.
Our molecule has two mechanisms of action, and we're pushing towards higher dose levels with a higher formulation strength, so I think we're in a good position and quite excited about the progress of the KSI-101 molecule through the APEX phase I study. From a catalyst standpoint, looking to 2025, we'll be looking at releasing some of the open label data, let's say, hopefully in the March, April timeframe of 2025, and beginning the PEAK and PINNACLE pivotals, we're doing the work to get those pivotals going in the Q2 timeframe of 2025.
Got it. One last question for me, given the time. In both cohorts of the phase I APEX study, you're testing up to 10 milligrams. Given that it's so potent, I mean, is 10 milligrams overkill? I mean, what are your thoughts on that?
It may be. The physicians really are encouraging us. Like I said, these are very, very sick patients with very, very severe inflammation with the macular edema leading to blindness. Physicians do want to have access to a very powerful medicine. A key decision point for Kodiak with the KSI-101 program is which two dose levels do we want to be choosing to put into these PEAK and PINNACLE pivotals? Because we have the designs of the pivotals. We just want to figure out, should it be 10 mg and 3.5 mg, or should it be 8 mg and 5 mg or 2.5 and 5? We need to determine that because you're right, this is a lot of really powerful drug, but it's going into very sick eyes.
I see. Listen, Victor, this has been so helpful. Thank you so much for spending time with us.