Kodiak Sciences, Inc. Stock

Thank you for standing by, and welcome to Kodiak Sciences DAYBREAK webcast. [Operator Instructions] Please be advised that today's conference is being recorded. Now it's my pleasure to turn the call to Victor Perlroth, Chief Executive Officer. Please proceed.

Thank you for joining the Kodiak Sciences DAYBREAK Topline Data Results Conference Call. I'm Victor Perlroth, Kodiak's Chairman and CEO. The webcast portion of this call contains a slide presentation that we will refer to during the call. Those following along on the phone who wish to access the slide portion of this presentation may do so on the Investor Relations section of the Kodiak website. An archive of the webcast will be available on our website.

Let's turn to Slide #2. I would also like to remind you that remarks made on this call today include forward-looking statements about Kodiak. A more complete description of these and other material risks can be found in Kodiak's filings with the Securities and Exchange Commission, including its Form 10-Q for the quarterly period ended September -- June 30, 2026, which has been filed with the SEC.

Kodiak does not undertake any obligation to update publicly any forward-looking statements, whether as a result of new information, future events or otherwise. Turning to Slide #3. It's with pleasure that I welcome all of you to our DAYBREAK primary endpoint results webcast.

Turning very briefly to the Slide #4. where we can take a moment to reflect on the length and also the importance of the journey that we have been traveling together with our Kodiak stakeholders of employees, patients, physicians, financiers, family and friends. In particular, we like to refer to this 4-quadrant plot here on Slide #4.

As Dr. Brown said it nicely in his quote in our press release, 20 years after Lucentis was approved for wet AMD, the Holy Grail remains the same, robust anatomic disease control and maximal visual acuity gains, i.e., immediacy/efficacy, sustained with less frequent dosing, i.e., durability. Before DAYBREAK and in particular, Zenkuda in DAYBREAK, durability always came with great sacrifice. And despite best efforts over these 20 years, our results of a community have been incremental. But today and together, we describe a tremendous step forward, a step that builds on Kodiak's 17-year history following what I like to call the gentle arrows of science and the harmony of company building and retinal medicines development.

On this plot, the upper right quadrant has always been our target zone. I think we've hit the target directly in the middle in the upper right quadrant. On that note, we'd like the data to speak for itself. Turning to Slide #5.

Joining me today from Kodiak is Pablo Velazquez-Martin, Kodiak's Chief Medical Officer. And joining us today also are Dr. Charles Wykoff and Dr. David Brown. Both are distinguished physician scientists with an important history of retinal drug development. After our prepared remarks, we will open the call for Q&A.

Turning to Slide #6. I'm now pleased to turn the call over to Dr. Velazquez-Martin. Pablo?

Thank you, Victor. I would like to take a minute to bring us all the way back 2 years ago, and we can move to Slide #7, almost to this date to the Investor R&D Day in Kodiak -- sorry, for Kodiak in September 2024 in New York City to remind us how we got here to this moment. At that time, we emphasized our focus to be a learning organization and strengthen our commitment to the journey from data to insights, to learnings to actions. At that time, we made a variety of critical course correct decisions that led us here. In Slide #7, we continue this intentional journey got us here today.

It is my privilege in Slide #8 to introduce this presentation, the primary endpoint results of the DAYBREAK Phase III study in treatment-naive with age-related macular degeneration exploring both of our antibody biopolymer conjugate investigational therapies, Zenkuda and tabirafusp alfa tedromer, which will also be called during this presentation as tabirafusp-ted.

One important reminder in Slide #8 -- sorry, Slide #9, DAYBREAK's innovative design enable an independent evaluation of [indiscernible] tabirafusp-ted at full clinical and statistical significance. This is important because each molecule had a different objective in this trial. Aligned with this principle, Dr. Wykoff will present the DAYBREAK results for each molecule compared to aflibercept separately in independent sections.

We move to Slide #10, please. And with that, I am happy to pass the presentation to Dr. Wykoff, who will present the efficacy and safety results for Zenkuda first. Charlie?

Thank you. Yes. Good morning. Great to be here with you. It's an exciting time in retina. A lot of ongoing clinical trial programs, always enjoyable to digest new data, particularly when it's as positive as I think you will see that this is. You can go forward to the next slide.

So DAYBREAK, fascinating Phase III clinical trial for many reasons. First is that it compared 2 different therapeutics to the same control arm. So a very efficient trial design to digest a tremendous amount of data. We'll first unpack the tarcocimab data and then dive into the tabirafusp data. So with tarcocimab, the comparator here was aflibercept fixed Q8-week dosing, 2 milligrams every 8 weeks after 3 monthly loading doses, 225 patients, really the gold standard EYLEA control arm, always important to think about the control arm first.

And then secondly, the tarcocimab arm was 4 monthly loading doses upfront, 225 patients and then a highly differentiated trial design here, and we'll go into detail about that on the next slide in just a moment. But these patients were then evaluated every 4 weeks and could receive dosing as frequently as every 4 weeks, but could also be observed at those visits depending on their disease activity. Now we've talked a lot about disease activity criteria in this space for the last many years, and this trial was extremely strict on when patients were retreated, and we'll dive into that. The primary endpoint was change in visual acuity there across 3 of the time points, weeks 40, 44 and 48, again, comparing tarcocimab with aflibercept, ultimately, though, a 2-year trial. And all of these were treatment-naive wet AMD patients. You can go forward.

In many ways, this may be the most important slide. Can you go to Slide #12, please? Thank you. Because the retreatment criteria in any clinical trial trying to look at durability are critically important. And most clinical trials use a retreatment threshold of increased CST of a certain threshold, 50, 75, 100 microns, for example, associated with visual acuity loss, usually 5, 10, even 15 letters of visual acuity loss. This trial was highly unique and none of those requirements were necessary. Specifically, we did not wait for patients to experience visual acuity loss, and there was no CST threshold that would drive retreatment. Instead, we use an AI-based fluidics analysis to look for the smallest amount of intraretinal and/or subretinal fluid to initiate a therapy.

We were trying to use practical criteria. In the real world, when we see fluid, we reinject and we don't go longer in between injections. And that's what we tried to recapitulate in this clinical trial using a point-of-care deployed AI algorithm.

So on this slide, you can see examples of what this looks like. On the left is an eye that was dry that did not have significant intraretinal or subretinal fluid. In fact, you see no fluid on that central B-scan. And then on the right, you see those small intraretinal cystic spaces highlighted in red, which is an indication of recurrent disease activity that would not have met any of the other retreatment criteria that I'm aware of in any advanced clinical trial used in retina previously, but it did cross the threshold for volumetric increase, and therefore, that eye was retreated at that time point before there was vision loss and before there was a meaningful increase in CST. You can go forward 1 slide.

So baseline characteristics for tarcocimab versus aflibercept comparison were balanced. Visual acuity was about 20/60, and there was balance in the BCVA category, the low luminance visual acuity deficit and CST was about 350 microns in both arms. About 40% of all patients were phakic. Go forward 1 slide.

So this slide demonstrates mean visual acuity change over time and absolute visual acuity actually plotted here on the y-axis between both of the arms. And importantly, you can see the primary endpoint was met with high statistical significance, demonstrating not only noninferiority but equivalence of tarcocimab versus aflibercept. You can see at the end there, the aflibercept arm had gained 7.6 mean letters compared to 7.2 with tarcocimab. And you see numerical fluctuations across the year. Of interest, you see actually slightly higher numerical visual acuity gains to the loading phase there with tarcocimab compared to aflibercept. But again, the key message here is really overlapping visual acuity curves from baseline through the end of 1 year at the primary endpoint. You can go forward 1 slide.

Looking at the similar trajectory here with absolute CST, again, you see overlapping trajectories. This may be the most important data point for tarcocimab for me. If you look at the loading phases there through 8, 12 weeks, you see overlapping CST curves, no indication of a separation between these curves, meaning that we have achieved that immediacy result where we have achieved complete drying relative to the aflibercept gold standard monthly dosed control arm there in that loading phase.

And then after that, you see that characteristic sawtooth pattern in gray when you switch to the aflibercept every 8-week 2-milligram dosing pattern. And again, you see minor fluctuation with tarcocimab. But again, these values are overlapping, demonstrating equivalence of anatomic outcomes from a CST perspective through the week 48 endpoint. You can go forward.

So what does this look like from a durability perspective? Because remember, the goal here was to achieve a completely dry retina and then maintain that retina at a dry state, right? We were not allowing significant recurrence of fluid before these eyes got retreated. That is very differentiated from most of the other programs, and I think is a much clear reflection of how we manage these patients in our routine clinical practice.

So despite that strict treat to dry and maintain dry approach, we still had a very large number of patients at an every 24-week or 6-month interval. So 54% majority of patients achieved that, again, within that 48-week time window. Another approximately 30% of patients achieved Q12, Q16, or Q20. And then reflecting the real-world nature of these treatment-naive patients, about 15% to 17% were in every 4- or 8-week interval.

I really like looking at that because it shows me that, yes, this is a representative patient population, what we expect in the real world. Some patients do need very frequent dosing, and I'm glad that we have included that population of patients in this clinical trial and yet the large majority are at every 24-week dosing without recurrence of fluid. You can go forward.

From a safety perspective, we saw no cases of intraocular inflammation here in the tarcocimab or aflibercept arms, no occlusive retinal vasculitis, no endophthalmitis. There were 3 cataracts, 1 in the tarcocimab arm for 0.5% and 2 in the aflibercept arm at 0.9%. Go forward 1 slide.

So this slide summarizes that tarcocimab arm comparison with aflibercept, [indiscernible], as it's also known. And again, the primary endpoint here was met showing equivalents through the loading phases and through the extended dosing phases through week 48 from both an anatomic and visual outcome perspective, while a majority of patients were at every 6-month dosing intervals. Importantly, from a clinician's perspective, a multi-indication BLA is planned by Kodiak across 5 positive Phase III clinical trials involving DR, RVO and of course, wet AMD with this DAYBREAK trial. You can go forward. You can go to 2 slides, actually go to Slide #20, please.

So now we're going to switch over and look at the comparison with tabirafusp. So this is a comparison using the same fixed EYLEA control arm there, aflibercept 2 milligrams every 8 weeks. Again, that same 225 patients. Now we're comparing that with a separate statistical analysis plan with tabirafusp. This drug was given again, 5-milligram doses. Dave will go into details about what these drugs look like in the dosing.

And this was 4 monthly doses through week 12, followed by fixed every 8-week dosing. That's that solid green boxes there at 20, 28, 36 and 44 weeks. And importantly, in between, patients could also receive additional monthly doses if there was presence of disease activity using that same AI algorithm I discussed before, and Dave will go into more details.

Again, the objective here was to give patients what they need on a very individualized patient-specific basis on an ongoing basis. So if patients had disease activity, we did not want to extend them. We wanted to be very aggressive and treat them as they deserve to be treated. But if there's no fluid, we did want them to be able to go every 8-week dosing. And again, the goal here with this trial was to really look at patients within the treatment-naive population and see if there could be a population that could do better with IL-6 inhibition, you can go forward.

Baseline characteristics were similarly balanced here. We saw baseline visual acuity about 20/60. BCVA categorical and low luminance deficit ranges were again balanced. CST was about 355 microns in both populations and again, 40% phakic. You can go forward.

The primary endpoint in this trial comparison was also met with equivalence demonstrated between tabirafusp as well as aflibercept. You can see at the end of the 40-week period a gain of 7.6 versus 6.3 mean letters. And again, noninferiority was demonstrated with high statistical significance, the p-value at the bottom there, 0.0036 with minor visual acuity fluctuation across the duration of that 1-year endpoint. You can go forward.

So Slide #23 here demonstrates CST outcomes. And again, this was actually an alpha-protected p-value in this comparison. So this anatomic endpoint also demonstrated noninferiority and actual equivalence between tabirafusp as well as aflibercept every 8-week dosing. And you can see minor CST fluctuations again, that classic sawtooth pattern in the gray population of aflibercept Q8-week dosing as well as some minor fluctuation with the tabirafusp arm and then nearly identical CST improvements in both populations at week 48. You can go forward.

From a safety perspective, similar to the tarcocimab comparison here, no cases of occlusive retinal vasculitis, no cases of endophthalmitis. There was 1 intraocular inflammatory event in the tabirafusp arm at 0.4%. Again, there were 2 cataracts in this comparison, none in the tabirafusp arm and 2 in the aflibercept population. You can go forward.

So Slide #26, my last slide here is just summarizing this tabirafusp analysis. You can see that tabirafusp met the primary endpoint in wet AMD, demonstrating equivalence with both visual acuity and CST outcomes in a comparison with fixed gold standard 2-milligram aflibercept arms. And looking forward, this drug is being studied in an ongoing DME program also, trying to understand who the patients are that may be able to benefit from additional IL-6 inhibition.

With that, I'll pass it over to my friend and colleague, Dr. David Brown. Thank you.

All right. Next slide. The biggest problem we have, we have great drugs. We have great drugs that solve blindness. The biggest problem is all our current drugs require frequent injections, especially for a population that has fast drug clearance. Even though a treatment-naive population, it's only 5% or 10% that need monthly dosing. In our clinics, it's 30% to 40% because those patients have to come every month, and we know their kids, we know their families. We'd love to have a drug that maintains visual acuity, and we see these great patients a little less often. Next slide.

So what Zenkuda is interesting. They have a conjugated ABC protein, which is very high molecular weight. And we know from all kinds of pharmacology studies that the larger the molecular weight, the slower the drug clearance. And so you actually have about a 20-day drug clearance of this big ABC conjugated protein. In the previous studies with Zenkuda, it was mainly -- it was only the conjugated protein. And while you saw longer dosing, you didn't see an immediate improvement in anatomy and visual acuity like we see with our standard of care great drugs like aflibercept, EYLEA.

What this is, is a 2-in-1 formulation, it's 20% unconjugated protein and then 80% conjugated protein. So you're basically getting -- you're getting a rapid response from your unconjugated protein while your conjugated protein provides long-lasting and longer-acting durability. Next slide.

Again, like I talked about, you have strong immediacy. What that really means is just disease control. You want the OCT curve or you want the fluid to go away. And correspondingly, you want improved visual acuity that comes with that. The durability or the ability to last without having injections all the time, I think, comes from this conjugated platform exactly as it's been designed by Kodiak. It's a phosphorylcholine-based polymer that's conjugated to the antibody. Again, the half-life is approximately 20 days. Our current drugs' half-life is 7 to 9 days or so, but that's in the average patient. There's good data that with faricimab, 10% to 20% of the population has a half-life about half that. And that's probably the same with all of our drugs. Some people have a longer half-life, some people have a shorter half-life. Next slide.

So both of the conjugated and unconjugated have a high binding affinity. So the binding affinity is with maximum inhibition is really good numbers, very comparable to aflibercept. The inhibition of a cell culture media, same thing, very similar to aflibercept. The difference is one of them is available immediately, the unconjugated -- and then one is more sustained release or you can look at it as sustained release, which gives you that long-acting end of this curve. Next slide.

I love this slide. We push for Kodiak to really look at their mean ocular half-life in humans. This isn't done in many drugs. It's in faricimab in the FDA filing. And so that's where this data comes from. The mean half-life in faricimab is 7.5 days. And again, that comes from their FDA filing. Zenkuda is 19.8 days. And I think it really is the main reason why you're seeing the more than 50% getting up to 6 months, like Charlie showed. Next slide.

This was calculated in the aqueous humor fluid of the patients in the Phase Ib study. I applaud Kodiak for doing this. I beat on them and said, boy, this is super important. And it really shows us what you're looking for in a drug if you're trying to get extended durability. It doesn't mean that you're going to get 6-month dosing in everybody. If you look at this study, 10 or so percent needed monthly dosing. That being said, that's the majority of patients had a very long duration between treatment intervals. Next slide.

Just showing you more of that Phase Ib study. And again, this is the fully conjugated protein. Here, you have the half-life varies from 6.1 days to 39.8 days, very similar to what we see in the first -- filing data, and I'm sure it's the same with aflibercept and ranibizumab. Some patients just clear drug fast. That's a function of pseudophakia, how much the vitreous is syneretic, how long the eye is, et cetera. But it just shows you we're very different. I love the design of this trial because it's not one size fits all. It's trying to figure out how do we get each patient the maximum effect within the realms of their biologic clearance. Next slide.

State-of-the-art today is faricimab and EYLEA HD. Even though in the clinical trials, which were a little artificial -- they were artificial because they required some vision loss and substantial fluid before you could get retreated, you had 77% and 79% and 70% of the faricimab trials go up to 16-week interval. A recent study by Bill Freeman showed that in real world, it's about 25% with EYLEA HD and about 21% with faricimab study. Next slide.

Discrepancy is 2 things. One, real world is different. We know from the CATT trials that a certain percentage of patients, 3 shots and you're done. And so those all go away in a treatment-naive trial. However, in a real-world trial, the patients that stay in our clinic that need ongoing injections probably have a faster clearance if you test them all and very few of these patients can switch and go to these lengths that you see on the TV commercials of 3 and 4 months. Next slide.

I do applaud Kodiak for this retreatment criteria. It is PRN, and so it's not going to be quite as good in real life as you see in this trial, but it was a very tight PRN. The AI reading read just very small amounts of fluid, which is much better than what you saw as retreatment criteria in our Phase III pivotal trials of faricimab and aflibercept 8 milligram. Next slide.

So here, you got treatment of retinal fluid monthly, individualized treatment. Again, it treated very small amounts of fluid. It is PRN, which I'm not happy about in the clinic, but I think it's the best you can do in a clinical trial. And I think it's very good to not allow much fluid. That's why you have maintenance of the visual acuity curves. A lot of the other trials out there for TKIs and gene therapy have a very lenient -- you have to have a lot of fluid before they let you retreat, and that's why you see those visual acuity curves dropping off, which is, in my mind, not acceptable. Next slide.

So 54% of patients got 6-month dosing interval. That's super impressive. If this is anywhere near this in clinical practice, it certainly will set a new standard of care for durability. Next slide.

And in conclusion, we really are excited to be able to see what this [indiscernible] can do in the clinic, right? They're impressive. This is where the field needs to go. I love the clinical trial design because with it, they can go to the FDA and say they can get monthly dosing and a certain number of patients are going to need that. The strong fluid and vision gains in the learning phase and then the maintenance of both the OCT curve and the vision curve are super exciting. The safety is consistent with our existing anti-VEGFs. And I applaud the development team and Victor and Kodiak for helping us move the game forward.

With that, I'm going to pass it off to Victor for Kodiak's next frontier.

Well, thank you, Pablo, and thank you, Charlie and Dave for walking us through the results, the exciting results both for Zenkuda and for tabirafusp-ted. And thank you, Dave, for putting a bit of the results into the historical context and into the context of your clinics. Let's turn to the Slide #41 for a very quick conclusion, and then we'll move into a few Q&A. I just wanted everybody to take a moment to think after the 17 years of development, it takes a long time to try to move innovation in drug development. I think we've achieved a definitive milestone here in the field of retina. We're proud to have been part of that.

And we're poised as a company for upcoming regulatory in terms of BLA filing that we are planning to do this year for Zenkuda in wet AMD in diabetic retinopathy following our GLOW1 and GLOW2 studies, and in RVO following our successful BEACON study. And we're excited to do the work to file the BLA and hopefully move from a pre-commercial stage company into a commercial stage company.

We're excited from the tabirafusp-ted standpoint. We're very pleased that our KSI-501 molecule met the primary endpoint in DAYBREAK, and we'll continue to evaluate different subgroups of patients within the wet AMD population to identify which populations there really can benefit from that dual mechanism and the anti-inflammatory action of the 501 molecule. And of course, we already are enrolling and randomizing patients in the ALTO Phase III DME study. We're continuing to accelerate site activations and randomizations. We're powering in that study for superiority and we're very excited to see where we can take the 501 molecule into these inflammatory vascular diseases. And we think that can be a good partner for Zenkuda commercially.

And of course, with our KSI-101 molecule, where we have ongoing clinical trials in the PEAK and PINNACLE pivotal program, we remain extremely excited about what we're seeing. With KSI-101, obviously, we had very strong data in the APEX study and have moved APEX into the PEAK and PINNACLE program. We remain on track to release top-line data for KSI-101 in PEAK before Christmas. So we're looking forward to that milestone as well. And then following from that, we'll have the top-line data from the second pivotal analysis for the PEAK and PINNACLE program that will be expected in the second quarter of 2027. And hopefully, we could follow that with a quick BLA for KSI-101 as well.

So as you look and think about Kodiak and you think about year-end 2027, and we think about hopefully approval for Zenkuda and potentially as we look probably not the end of '27, but let's say, early '28 approval for KSI-101 and also, of course, continued progress and acceleration of the tabirafusp-ted program in the background in DME. We think about Kodiak as a 3-product company, that's very exciting, led by the science and I think the validation that we see on the ABC platform, also validating some of our pipeline and development work looking at future retinal diseases such as glaucoma, geographic atrophy and additional bispecific work in ocular inflammation.

And we have had also very exciting work on our VETi retinal analyzers and engagement tools, and that's something that we can talk about a different day. I think I'll stop here, and we'll begin to take a few questions. Thank you.

[Operator Instructions] One moment for our first question, please. It comes from Anupam Rama with JPMorgan.

Congrats on the data. Really great to see. I was wondering if I could ask 2 quick questions of the KOLs on the line. So the first one is with the profile of tarcocimab or Zenkuda in wet AMD, where do you see the product fitting into what is a fairly competitive treatment landscape? And which types of patients do you think would be best for therapy given the emerging profile that we're seeing here?

And then the second question on -- for the KOLs, what percent of patients are super difficult to treat in wet AMD and maybe have some sort of inflammatory component from which they may benefit from an IL-6 mechanism when thinking about KSI-501 or...

Yes. Charles Wykoff here. I don't know if you can hear me, but happy to take that question first. I appreciate the perspective that you're taking, which I completely agree with, that we need to start thinking about wet AMD patients as more of a heterogeneous disease population than we have historically. Historically, we have a one-size-fits-all. And I think we're getting to a place in retina where we can be more nuanced.

So to take the second part of the question first to tabirafusp, I do think that there are some patients with an inflammatory component. I think we start to dive into the data to kind of understand who those patients might be in this particular clinical trial. On average, we're seeing that we've met equivalence here and non-inferiority with aflibercept, but are there subpopulations that may be doing even better with IL-6 inhibition, I think those analyses still need to come. I do think there will be those patients, and we'll see. On the other side, I think that tarcocimab could be useful for a broad range of patients with wet AMD, right? The value of this clinical trial program is that we're showing that we can dry equivalent to fixed EYLEA monthly dosing during the loading phase and then a very large number of patients can go longer in between injections than what we're used to with our current biologics.

So from that perspective, I think this could be broadly useful for a large proportion of patients with wet AMD, both the patients with high need dosing frequencies and also those patients that are already on extended intervals because we can hopefully go longer based on the half-life data that Dave beautifully unpacked, right? This biopolymer conjugate looks like it's meaningfully extending the intraocular half-life for this anti-VEGF component in these patients, which I think would be quite clinically meaningful for a large number of patients.

Yes, I agree with Charlie. There's lots of patients out there that -- if you -- to answer the first question, it's in a busy tertiary care clinic like Charlie and I have, it's probably 30% of the waiting room needs monthly treatment with faricimab or EYLEA HD. So that population is really out there. Will this molecule move that population? I don't know. We'll have to see.

In terms of the IL-6 question, I think it's much more likely that with diabetes, we're going to see an IL-6 responder group. AMD, there's less evidence of IL-6 or inflammation. But in the DME population, certainly, that's why steroids have persisted all these years is there's a certain part of the population that really has inflammatory components.

Congrats again on the data, guys.

Our next question comes from Maury Raycroft with Jefferies.

This is Amin on for Maury. Congrats on the data. A couple of questions. First, for Dr. Brown and Dr. Wykoff, how do the treatment criteria compare with your real-world practice? And would you expect durability to be -- to change outside the trial? And then I have a follow-up.

Yes. So in the real world, most of us don't allow any fluid. So in the real world, we figure out a dosing interval and we stay at that interval. The difference in this trial, and you had to do it in the clinical trial is here, you're actually waiting for fluid to occur even though it's a very tiny amount of fluid.

So in the clinic, you're going to use this a little less than what you see in this trial. How much less, you really won't be able to tell until you get it out there. But unlike previous trials that wanted 50 microns of fluid or 100 microns of fluid and commensurate vision loss, this is much more real world. This is very small amounts of fluid, any appreciable amount essentially with the AI read. And so this is as close as you're going to get to the way we treat in the clinic and not -- given the trial constraints, you can't do treat and extend in a clinical trial and get a drug approved.

Yes, I agree with those comments. I think the biggest differentiator of this clinical trial program, DAYBREAK, is that retreatment algorithm that Dave described and we showed the images for, right? Most other clinical trial programs that are showing extended dosing intervals between treatments or gene therapy or TKIs are using a combination of increased CST and vision loss to drive retreatment or to determine your treatment interval. No one in clinical practice does that. We don't wait for visual acuity loss and meaningful change in CST. CST is a number that most retina specialists are not routinely using to drive retreatment in wet AMD. We use the presence or absence of fluid to determine what that interval is going to be.

And this is really the first pivotal trial to use an AI algorithm, and I applaud them for doing that. This is sort of a first of its kind, and I really think this is where future trials need to go to try to find these recurrences and disease activity before there's vision loss and before there's significant architectural distortion of that central foveal zone.

Sounds good. And for tabirafusp, what do you think explains the numerically lower letter gain with this drug versus aflibercept? Is there any reason to believe it's mechanism related? And does it change your view on the ALTO trial?

Yes. This is Pablo Velazquez-Martin. Thank you for the question. I think like this is a positive study and both BCVA and drying met statistical significance. I think that's important to remind ourselves. I think that the message -- the comment from Dr. Wykoff and Dr. Brown about finding which patients benefited the most from the IL-6 inhibition in this treatment-naive patient population, it's our next task and to understand the data better, of course.

As Dr. Brown was mentioning, wet AMD and particularly treatment-naive overall broad population might not have -- is not probably the best place for an IL-6 inhibition or inflammatory inhibition to shine. And that's what basically validates our development path, which we're moving to DME. And that will be my take on the results. Charlie, do you want to add anything else?

Yes. It's a good question. And I don't know the answer to that. I can tell you statistically, not only has non-inferiority been demonstrated, but my understanding is that equivalent here has also been demonstrated. So I do think from a statistical perspective, these are overlapping, at least the way the trial statistical analysis plan was defined. I think that that's critically important.

Beyond that, these subpopulations will be really interesting, right? I think we're learning so much about new biologies in retina, and it's fantastic for patients, right? We have Ang2 biology that we've learned about now for a while. We're going to get a lot more data on direct Tie2 activation, Wnt activation and in this case, IL-6 inhibition. So there's a lot more that we're learning about these different disease states. And I believe strongly that we are going to migrate more to an oncology sort of approach here where we're being more specific about patient subpopulations that could benefit from additional modulation of different biologic pathways beyond VEGF inhibition.

It comes from the line of Umer Raffat with Evercore ISI.

First of all, congrats to the whole Kodiak team for persevering through this over the last 5 years. I have 3 questions, if I may. First, perhaps for Charlie. Charlie, do you think this is standard of care based on all the data we're looking at, #1.

Second, maybe for David. David, can you speak to how the AI could be implemented in practice or that a little bit of fluid method that retina practices do use will effectively mimic that? I'm just trying to think about how to sort of implement and find that first 1/3 of the patients that do need the additional supplementation.

And the last one, Victor, for you and maybe for your stats team, there's something I was looking at and which may be a labeling thing. On the OCT graph that Charlie leaned into, it shows 133 micron drop for aflibercept in the table, the least squared means when tarcocimab analysis is shown, but it shows 138.8 when that aflibercept arm is shown again for the 501. And similarly, I felt like in the tarcocimab OCT disclosure, maybe the gray versus blue flip the way the OCT is shown, it looks higher even though the number is deeper. So I just wanted to clarify that.

I'll take the first of those, thanks. So it's the standard of care. It's interesting. Whenever you do a prospective clinical trial, right, -- there's a lot of bias there. Every single trial is different than routine clinical practice patients for a lot of reasons. And I think the biggest disconnect that we've had as retina specialists as practicing doctors over the last few years has been this disconnect between what we do in clinical practice, which is treat to dry and then go as long as we can in between those doses based on anatomic recurrence of disease and/or patient symptomatic recurrence of symptoms.

And the disconnect has been in clinical trials because clinical trials have sort of instituted these thresholds, which are more than that, where they've required visual acuity loss and anatomic worsening beyond what we would allow in clinical practice before retreatment or beside or before determining what that longest acceptable interval is. And that's been a disconnect because we look at these clinical trials that require, for example, 15-letter loss, and that's just not standard of care in any of these diseases and especially wet AMD. Even 10-letter loss is a lot of visual acuity you require a patient to lose before you retreat them. The 5 or greater letter loss is, I think, a little more acceptable for some of those clinical trials. But again, that's not what we do in routine clinical practice.

So I think this is the closest pivotal trial that we've ever seen in the modern era for wet AMD management, getting at how doctors actually think about managing these patients in the real world. I think this will be more closely aligned with drug treatment intervals that are achievable in the real world for a broad treatment-naive population. Now if this is approved and we use previously treated patients in the real world, that's going to be a different outcome than, again, a broad population like this that's treatment naive those patients a priori are going to need more frequent dosing.

Umer, this is Dave Brown. Good to talk to you again. In terms of AI, it's fascinating. AI is rapidly getting into our imaging. I think AI will be useful in the clinics to help doctors find small amounts of fluid that if you're looking at a 94 scan -- volume scan, it takes a while to scroll through all that and kind of like a radiology read where it shows you sort of where to look, I think that's going to help.

In the clinics, though, the real-world story is that with step therapies, most insurance companies make you either step through Avastin and/or probably an aflibercept biosimilar at some point, you're going to have a much tougher population that moves on to these advanced drugs. And so you've already weeded out the easy ones with that step therapy. I think you're not going to see these kind of results because unless an insurance company figures out that, boy, it's cheaper to give 2 or 3 doses of this drug than to make the doctors go through Avastin and aflibercept biosims. I think we're going to -- the real world is we're going to be in the United States stuck with that paradigm.

Thank you, Umer -- here for the third question. They are not -- like the numbers are not slipped. So in the OCT graph, just looking at tarcocimab as an example, on Slide #15, it's observed CST. So if you look at the baseline of the aflibercept arm, it is slightly higher, and that's why it doesn't fully represent the change at the end. So the numbers are correct. It's minus 138 and minus 132 for aflibercept and tarcocimab, respectively, at week 48.

Regarding your second comment, the difference between those numbers and the numbers in the table below is the numbers in the table below use a MMRM model. So that's the square means numbers, whereas the number on the top -- on the right on the graph is just the mean change observed -- so that's the difference. The one on the bottom of the table is a statistical number, which is the one that is used for a statistical test.

Okay. But because the EYLEA look different between the 2 tables when the analysis on OCT was shown?

Yes. And as we [ prepped ] at the beginning of the study, so the study -- sorry, the 2 studies are independent. The analysis and the models and everything are independent. I do think that they're identical. We can double check for that, but I'm pretty sure that they're the same.

One moment for our next question that comes from Michael Yee with UBS.

Congrats again, Victor and the team. Maybe a question on 501. I think there's not really any debate on the AMD stuff. I think you guys wrote in the slides the new benchmark. So just thinking about the read-through then for DME in the ALTO study, maybe Dr. Charlie and David could comment, do they seek or would they be pleased to see better BCVA, higher BCVA, superior BCVA? Or would greater dosing be attractive? Maybe just comment about where you could see significant opportunity here in DME given AMD is not such an inflammatory-driven disease. And the second question is beyond DME, how does the data today change any of your thinking about your confidence on MESI, which, of course, is reading out in a couple of months?

Yes, I'm happy to jump in there first. Thanks for the question. I agree with your premise and other comments here by many that DME historically has looked at as more of an inflammatory disease state. Certainly agree with that. I do think there is broad heterogeneity in both wet AMD and DME that's been underappreciated historically, mostly because we didn't have the tools to address these different subpopulations in unique ways.

So I think we need to get more sophisticated as a specialty at understanding who these subpopulations of patients are. Credit to patients to groups that have been looking at things like intraretinal hyperreflective foci and the amount of DRIL and retinal nonperfusion, right? There's a lot of imaging biomarkers that can be leveraged in the DME and DR population to try to identify who those subpopulations of patients are. So I think we have a lot more to learn there.

Specifically to your question, sure, I'd love to see improved mean visual acuity outcomes across the entire population of patients and better anatomic drying. I'm not sure if that will be achieved. I certainly hope it will be. But what I expect will be achieved will be very similar outcomes to aflibercept 2 milligrams with the potential for some subpopulations of patients to be doing meaningfully better from a clinical perspective.

Thank you, Mike. This is Pablo again. So just to remind the audience, so tabirafusp alfa tedromer and KSI-101 share the same protein. That's where the relevance of the question comes from. How do we think about MESI. I think that it's 2 different compounds and 2 different patient populations and 2 different everything. That's why we're doing them separately, the full development plan. We're extremely excited about what the results will be for MESI and for 101, as you mentioned, it's coming up very soon. And about the patient population and the response and the evaluation, Charlie, do you want to comment? You're an investigator in the PEAK and PINNACLE clinical trials.

Yes, PEAK and PINNACLE -- fascinating study population where clearly, that's an inflammatory-driven group of diagnosis. I really like the approach that Kodiak has taken, and I hope others take it also, where they're sort of broadening that potential funnel for patients with inflammatory associated macular edema, right? We're certainly aware of the core uveitis associated diagnosis that are associated with inflammation. But there are many other categories of disease are being studied in there as well. And it will be fascinating to look at those different subpopulations of patients based on their entering diagnosis to get a feeling for how that IL-6 inhibition can benefit them clinically.

Well, thanks, everybody, for joining us for the DAYBREAK primary top line data, WebEx and results discussion, I think it is a very good discussion. We were privileged to have Dr. Wykoff and Brown participate. Thanks also to Pablo. I think we've achieved something very important within the wet AMD field with Zenkuda.

I think in summary, on these different retreatment criteria, what we tried to accomplish in DAYBREAK was something as close as possible to what the retina doctors do in their clinic. So they might print out the OCT CST -- the B-scan of the OCT image. If there is any fluid present, the physician would treat that patient. The application of our tool was only to provide a consistency across the study and to accomplish the same thing.

We set the thresholds for treating based on intraretinal or subretinal fluid if fluid was present at the limit of detection reproducibly of the machine, then it would trigger retreatment for the patient in DAYBREAK. And having said that, we were able to bring a majority of the patients coming out of the loading phase and at the endpoint, through the endpoint, a majority of them on 6-month dosing.

And as Dave says, and Charlie says the population in the real world may be a little bit different. But the point is, as we showed, Dave showed the curves of the aqueous taps and the ocular half-lives that every patient has a different ocular half-life. And now with the combination of the free protein and the conjugated protein, the 2 and 1, our concept is to provide the best possible immediacy and the best possible durability for each patient. And to have the retina physicians to do their job and to have the best drug to be able to provide the best journey for that patient. And we had a clinical trial that best approximated what they do every day in their clinic. And that's very exciting to see that together with the safety that we've also witnessed in this study so far.

With the 501, very excited to accelerate it in DME. We're glad to have a positive study in wet AMD, very exciting, I think, because of the strength of the Zenkuda data in DAYBREAK, then you're asking questions about the tabirafus-ted or the 501. I think we're very excited about these results in DAYBREAK and very excited about the potential for our superiority study design for 501 in DME.

And KSI-101 and MESI is a very different animal, very excited to be in macular edema secondary to inflammation, which represents a very broad population with a very high unmet need and excited in the development of Kodiak in my own career to be able to bring KSI-101 as a career medicine into a commercial setting for patients. Thanks for joining us. We look forward to the next steps of the journey together. With that, we'll sign off.

And ladies and gentlemen, thank you for participating in today's conference, and you may now disconnect.

All right. Welcome, everyone, to the 44th Annual J.P. Morgan Healthcare Conference. My name is Anupam Rama. I'm one of the senior biotech analysts here at J.P. Morgan. I'm joined by my squad, Priyanka Grover, Joyce Zhou and Rati Pinhe. Our next presenting company is Kodiak. And presenting on behalf of the company, we have CEO, Victor Perlroth. Victor?

Thank you, Anupam. Well, each year at the J.P. Morgan, the metronome of the different years, it's great to be here. Thank you. But I also remember the R&D Day that Kodiak had in July of last year. I think at that time, I said I couldn't be more excited and enthusiastic about what we had in front of us. And now looking 6 months later, we've made tremendous progress on the plans that we articulated at that time. I think what I also like to say is many good things that could have happened did happen and many bad things that could have happened didn't. I think as I look and as we look at Kodiak to 2026 and into 2027, it's a tremendously exciting time for us. I continue to be very pleased with all of the progress and everything that I see at the company.

We're going to, as usual, run through quite a lot of content. There is a lot under the hood at Kodiak, and we'll try to do that in a time-efficient manner. First, remember our forward-looking statements. Please read carefully all of the SEC documentations, including the recent materials in the last quarter associated with our capital raising. So at Kodiak, as you know, we're now definitively pre-commercial stage retina-focused biotech. We have our KSI-101 asset, which provides in and of itself a very simple reason to believe in the potential for Kodiak and for value creation with very strong data out of our APEX studies, a mechanism of action around IL-6 and IL-6 plus VEGF that's highly validated now within the scientific community.

Phase III PEAK and PINNACLE studies that are enrolling quickly and ahead of our expectations with top line data expected in the fourth quarter of this year and into 2027 and a tremendously interesting commercial opportunity for this bispecific protein asset. So a very interesting reason to believe in Kodiak. Don't forget, though, tarcocimab and KSI-501 are conjugates, so we look at them at science-based heavy weights. Tarcocimab is targeting a BLA in mid-2026, let's say, after the readout in the DAYBREAK study in 3 different retinal diseases. Retinal vein occlusion, a very exciting and fairly large, let's say, $3 billion opportunity, tarcocimab being a medicine that in the BEACON study showed very strong data, both in terms of its power and in terms of its unusual durability head-to-head against EYLEA.

And then, of course, in diabetic retinopathy, we'll have the GLOW1 and GLOW2 readouts, GLOW1 previously and GLOW2 in March, let's say, of 2026. And the wet AMD readout in the DAYBREAK study in the third quarter. And 501, a bispecific conjugate with our science of ABC, we hope the VEGF trap and the IL-6 will deliver something special for that molecule in the retinal vascular diseases. So 2 very interesting assets. They're not super -- I would say that the potential value and opportunity of these assets is underappreciated. But it's a show-me business, and we're going to generate data in GLOW2, and we're going to generate and share the data in DAYBREAK and we'll all turn those data cards together and see the value of those programs in the context of Kodiak's portfolio of 3 programs.

And in our deeper in our pipeline, we have additional bispecifics around KSI-102 and 103 that we're excited about. In inflammation, we have very interesting bioconjugate-based programs in massive disease areas to come such as glaucoma of the retina and also geographic atrophy of the retina. We don't talk a lot, but we have had tremendous progress in our VETi headsets. And we also, of course, have Kodiak's commercial manufacturing facility partnered with Lonza, the URSUS facility, and we completed our validation batches and released those for the antibody, for the polymer and for the conjugates.

So tremendous progress with inside of Kodiak, a very interesting portfolio of assets with 4 Phase III programs in progress, 3 Phase III programs on track for readouts this year. Wow. As a reminder, 3 Phase III programs in retina, they really follow scientifically from Kodiak's approach to drug development, we're at a decisive moment today. As I mentioned 6 months ago, tremendous progress since then. I still believe we represent a very compelling investment opportunity for all of you. This is a nice slide, Slide 4, where we list our expected catalysts. So on the KSI-101 program, we'll have 24-week data from the Phase I APEX study in February of this year. We'll have top line data expected for the Phase III PEAK study in MESI in the fourth quarter of this year, and we'll have top line data from the Phase III PINNACLE study expected, let's say, in Q2 of 2027.

So an exciting program and an exciting set of catalysts. For tarcocimab, we'll have top line data from the repeat GLOW2 study this quarter. That's our expectation. We'll have top line data from the Phase III DAYBREAK study in wet AMD in the third quarter, and we expect to closely thereafter file the BLA for tarcocimab in wet AMD, diabetic retinopathy and retinal vein occlusion. And the KSI-501, as you know, we'll have the same top line data from the Phase III DAYBREAK study in the third quarter. We look to the VEGF trap and the anti-IL-6 antibody in the context of our conjugate to explore the potential for improved efficacy beyond anti-VEGF monotherapy to show strong immediacy and to show an industry-leading durability.

So with 101, we're looking at a greenfield market opportunity in very sick patients, very attractive opportunity. With tarcocimab and 501, we call that the quadrant of core unmet need, where as a biologic, we want to be able to deliver strong immediacy and an industry-leading durability as a biologic in the same therapy at the same time. Very unique. It's an area like the Wayne Gretzky. We're still trying to skate to where the puck wants to be. And I think with our bioconjugate portfolio and the studies we're running, we're going to turn those data cards, and we'll find out whether we're there. And we very well may be that would very exciting.

So as I mentioned, the assets are on track for top line data readouts in '26 and into 2027. When you look at Kodiak, as we call ourselves pre-commercial, what does that really mean? Well, being in retina, there's not that many companies in retina and with retinal biologics and with the 3 different programs, there's a very interesting potential for Kodiak to begin to see a sustainable revenue drip beginning in 2027 that can grow aggressively across the 3 different assets. Let's stay tuned. And on Slide 7, just a reminder that within our bioconjugate technology, we also have next-generation versions of that where we bring not just antibodies as conjugates, but also small molecules and RNAi and a variety of different types of diverse APIs into the polymer conjugate, in particular for retina, but some early activities systemically.

As drug developers in biotech and pharma, the concept of having a proprietary platform where we can really do multifunctional and modular drug development across a diverse set of APIs in parallel is very interesting, very unique and I think part of a longer-term value build for Kodiak. So as I mentioned, we have 3 Phase III programs. We have a variety of earlier pipeline assets, all designed to address the leading causes of vision loss. So obviously, our KSI-101 in MESI and our tarcocimab and KSI-501 in the retinal vascular diseases. As I mentioned, I'm quite excited about the potential for our glaucoma Duet, including an NLRP3 inhibitor that's quite potent, designed in-house and also IOP lowering elements together and also some earlier GA Duet, which, of course, also represents large opportunities for Kodiak as a retina-focused franchise company.

What about tarcocimab and 501 very briefly. So there remains, despite commercial complexity, valuable open space in the $15 billion retinal vascular diseases market despite the availability of approved biologics and despite clinical trials of new exploratory technologies. So the concept of an upper right quadrant medicine, a biologic that can provide both high immediacy and high durability at the same time and in the same product. So there are technologies, what I call sort of lower right quadrant molecules that are being explored in clinical trials, including implants, gene therapies that are engineering for high durability, but importantly, lack any concept or potential for immediacy.

So those really are maintenance agents that will end up being used in subsets of patients, but they'll always be relying on the mainstay biologics such as ours to establish and to reestablish disease control. So there's a very interesting open space in the market for a biologic that can provide high immediacy and high durability at the same time in the same therapy. And that's what we're trying to develop with our ABC platform. And if you look here on the right, the concept of having a bolus and a basal, right, delivery, a strong early pulse and longest possible durability in the same medicine. That's what Kodiak has been trying to do. So here, Slide 18, what we see is that immediacy with intravitreal biologics is achieved with lower molecular weight biologics.

But the durability increases as the molecular weight increases. And of course, our conjugated antibody biopolymer conjugates are quite large. And we proved definitively that our conjugates have the longest half-life as a biologic. So we have tarcocimab, a mean ocular half-life in humans of 20 days, which is threefold longer than, say, intravitreal biologics such as faricimab. So very interesting and very compelling data in humans. And we generated that in the sophisticated studies on Slide 20 and Slide 21, and it's very interesting science. And we demonstrated that the 20-day half-life, which, as I mentioned, is 3x longer than the smaller biologics like EYLEA and faricimab was consistent across the 3 major retinal vascular diseases.

So as we look backwards, right, did this 3x longer durability come at a cost? And well, it seemed yes. The immediacy seemed to be the cost, right? In the wet AMD monthly DAYLIGHT study that we ran with the pre-changed tarcocimab formulation, we saw a deficit in the loading phase right in the immediacy of the effect, but after the loading phase, the drawing potential of the potency was comparable. So to fix that, looking backwards, we decided to add free protein and conjugated protein together into our commercial formulations for both tarcocimab and KSI-501, right? So for tarcocimab, we have still a 5 mg dose, but 4 mg of conjugate, but we added 1 mg of the free protein. And we applied that platform upgrade into the KSI-501 program as well.

In this case, 1.5 mg of the 101 equivalent, right, of the free protein and then 3.5 mg of the conjugate. And that was designed to confer a number of manufacturing and safety benefits and importantly, improved immediacy. So the unconjugated portion of this commercial formulation for tarcocimab itself contains a high molar equivalent, just of the unconjugated portion to approved intravitreal biologics. So the 1 mg is equivalent to a 0.7 equivalent molar dose, okay, which is the same as a 0.7 mg dose of LUCENTIS, right, with its marketed dose being 0.5 or a 1.3 mg dose of EYLEA with its marketed dose being 2 mg or a 2 mg equivalent dose of faricimab with its marketed dose being 6 mg. So that's just the free protein portion.

So it was added to bring a powerful and immediate disease control capability into the commercial formulation while also improving manufacturability, dose administration and patient safety. So we can apply that into some modeling where you can look at patients in retina that have either a short half-life and need monthly dosing, let's say, in the clinic or an average patient that may need every 8- to 12-week dosing or on the right, a patient that may do well on every 3- to 4-month dosing. And in that case, our new formulations, our new generation of commercial formulations have free protein that can deliver something very similar to like an EYLEA or faricimab, but we also have our conjugate that can deliver that much longer durability.

And we can do that in each type of patient, we can deliver something better. And that's what we hope to be able to show in the DAYBREAK study. So again, our conjugates are designed now to bring the best of both worlds of an immediacy that's not weak but is strong and the best durability as a biologic at the same time and in the same therapy. So our ongoing and planned studies, we hope will support a BLA for tarcocimab right after the DAYBREAK readout and a potential KSI-501 BLA, let's say, in 2027, while enabling complementary and potentially differentiated commercial profiles. So let's remind ourselves, tarcocimab did show very successful data in the retinal vein occlusion area in the Phase III BEACON study.

RVO is a $3 billion market today. And the current agents have failed to address the key unmet need for better efficacy and better durability. So we delivered 6-month durability in 75% of the patients in RVO and BEACON. And we think tarcocimab can be an excellent drug for patients in RVO. And we did a head-to-head individualized dosing of tarcocimab in RVO and BEACON against aflibercept in the second 6 months in 75% of the patients, you can look at the vision RVO OCT, did extremely well against EYLEA head-to-head and 75% of them had 6 months or longer durability. So it shows you the power of the old tarcocimab and what the new tarcocimab can be even better.

And certainly, we demonstrated very high efficacy and industry-leading durability in diabetic retinopathy in the Phase III GLOW1 study with 100% of the patients were treated every 6 months. And actually, quite a large number of patients are seemingly treated with anti-VEGFs in the U.S. for diabetic retinopathy. In fact, recently estimating as many as 850,000 patients. I like on the Slide 34, just a visualization, right? Why aren't more patients using anti-VEGFs when they have DR? I mean if I had my eyeball on the left with all the hemorrhages and the clotting that's going to go on in the retina, I'd rather take twice a year tarcocimab and have my retina look like the one on the right.

So we need to get through GLOW2. We need to get through our BLA, but we need to make tarcocimab available for these patients. So what are the potential implications and the upsides for tarcocimab in the GLOW2 study? So the first thing with top line data expected at the end of Q1 2026, which is really right around the corner. So it will be the first Phase III readout using the commercial formulation for tarcocimab, providing important safety looking. It will be the second pivotal superiority study with 100% of patients on every 6-month, i.e., twice a year dosing versus sham, okay? So it will be a repeat of that. So we expect strong pivotal data and strong endpoint data for the primary and secondary endpoints.

And it's a similar study design as the successful GLOW1 study. So if it's successful, we'll have a BLA-ready profile in DR and RVO. Now the ongoing Phase III DAYBREAK study in wet AMD explores in the definitive manner, the immediacy through the loading phase, the real-world durability for tarcocimab, okay? And also for KSI-501, the potential for showing better efficacy. So in DAYBREAK for wet AMD, what are the potential implications and upsides looking at the DAYBREAK top line data readout for tarcocimab, which we expect in Q3? Well, we'd expect to show strong and immediate disease control in the loading phase, closing that gap and at the same time, showing non-inferior visual acuity gains against EYLEA Q8 and also demonstrating long interval dosing with a very flexible 1-month through 6-month label.

So we kind of have an illustrative view. And what about for 501? What are the value implications of DAYBREAK? So it will be year 1 top line data in Q3. We'll be evaluating Q4, Q8, KSI-501 in wet AMD versus EYLEA Q8. So there's the potential to show better visual acuity gains than anti-VEGF monotherapy. We don't know. Let's remember that we're seeing very powerful data with our dual IL-6 VEGF KSI-101 in patients, right, very sick patients. There should be some translation into some of these wet AMD patients, wet TBD. And we hope to be able to show, of course, based on the commercial formulation, strong and immediate disease control.

So revisiting the core unmet need and based on the science of high immediacy and high durability, we believe that our conjugates, tarcocimab and 501 are poised to be credible contenders and/or to fill uniquely as Gen 2 agents, what we call this golden triangle. So we'll be seeing that this year together. And of course, we've been making and continue to make significant investments in commercial manufacturing, which positions us well to be able to launch multiple ABC products into these large and growing markets. I'm glad to report that in the last year or so, we've completed the BLA facing commercial scale validation batches. We've manufactured those and released those for the antibody, the biopolymer and the bioconjugates.

So tremendous manufacturing progress in support of Kodiak's BLA. So moving quickly into one of our key assets, perhaps the most interesting asset from the standpoint of investors, KSI-101 and MESI. So we won't go through all of the background that we've been through over the last 1 or 2 quarters, but macular edema is a common clinical presentation of a wide spectrum of diseases caused by inflammation and/or VEGF overexpression, but it leads to a common phenotype of tremendous fluid under the retina, what you can see here. Now what is MESI? Well, you can see these different visualizations on Slide 45, whether it's anterior or intermediate, posterior or pan or whether it's idiopathic or whether it's caused by a known stimulus such as postoperative or juvenile idiopathic arthritis, for example, they all have a similar look, okay?

It's a leading cause of vision loss, okay? And what causes it? Fundamentally, it's an autoimmune attack of the body against the retina or in particular, against the barrier, the blood retinal barrier. And when that breaks down, you get IL-6 and VEGF that create a local amplification that leads to this fluid, okay? So the unmet need is real. People are using corticosteroids today, mostly as intravitreal injections that can lead to permanent glaucoma and permanent cataracts that are difficult to replace. So Slide 52 shows you some of the complexities of the surgical repairs or Slide 53 showing you some of these really horrific patient journeys for patients that are on a multi-procedure path towards blindness.

So there's an unmet need for something that's safer and has high efficacy as a biologic. So IL-6, through a lot of the data that Roche generated in patients is a validated mechanism through their Phase I data and dovetail and generally repeated into their Phase III. So -- and while intravitreal IL-6 monotherapy is helpful, 50% of the patients still have persistent fluid, okay, leaving room for a more potent and/or broader spectrum of therapy. So luckily, no on-target adverse events really associated with IL-6 biologic inhibition by Roche were noted in their pivotal studies, derisking IL-6 inhibition in inflammatory macular edema. So how can this be -- the unmet need be addressed by Kodiak? Well, our KSI-101 molecule design, a very nice bispecific program.

So it has a VEGF trap in line with an anti-IL-6 antibody. We formulate that to 100 mg per mL for a 50 and 100 microliter dosing testing 5 mg and 10 mg in patients with modified Fc. So it's immunologically inert. It's very potent. It improves the tight junctions that are, as I mentioned, the core problem is the breakdown of the barrier. So dual inhibition by 101 confers superior normalization of the tight junction barrier, and there's a synergistic effect that we've seen. And in fact, it can restore the barrier resistance from strong preexisting insult better than can be done by anti-IL-6 or anti-VEGF monotherapies alone.

So we think here, it's poised to fill a very interesting unmet need in the greenfield market opportunity where patients are being given horrific therapies, intravitreal steroids that are creating tremendous side effects and our drug so far, knock on wood, doesn't seem to have any of those issues. So we've shown 20-week data in the Phase I APEX study. The data looked very strong. Very interestingly, we dose once a month, 0, 4, 8, 12. And even though we stopped dosing, we see through week 16 and 20, a continuous improvement in the vision in the mean change and in the observed vision over time. So we're planning to share the 24-week data at the Angiogenesis meeting, I believe, in early February.

Fundamentally, the safety continues to look very good. So what we would anticipate to see at Angiogenesis, whether the vision begins to go down or whether the vision continues to improve, they're both great outcomes for the therapy. More than half of the patients are achieving the greater than 15 or equal letter gain, which is critical to the key secondary endpoint in our PEAK and PINNACLE study. And we're seeing very strong, very rapid improvement in the fluid in the patients. And importantly, more than 90% of the patients are achieving absence of any retinal fluid. So it's showing itself to be very strong, very strong agent and very good safety.

How it fits in? Well, Roche generated nice data in Phase I. They generated -- we generated what looks to be stronger data, okay? When we do like a comparison against Roche's vamikibart program, a single dose of our agent seems to deliver the majority of the response. And as we give several additional monthly doses of our agent, we get too deep into this dryness corridor where multiple doses of the Roche anti-IL-6 doesn't quite take the patient where they really need to go. I think that picture speaks a thousand words, basically.

So what about the Phase III programs that are enrolling now, PEAK and PINNACLE, looking at 2 dose levels, 5 mg and 10 milligrams against sham, fixed monthly dosing through week 20 and 24 and then individualized dosing through weeks 48. They're actively enrolling now on a global basis. Importantly, as we looked at the Roche data, we decided that it would be useful for us also because the programs are enrolling more rapidly, let's say, than we had thought. We're going to make the studies larger. And also, we've clarified through Type C meeting with FDA, both on the primary endpoint, but more importantly, on the key secondary endpoint for the programs, which I detail here.

The populations are a little bit different. They're both very attractive for drug development in MESI. So as I mentioned, we've increased the size of the PEAK and PINNACLE program on Slide 92. Patient enrollment is faster than expected. We don't expect major changes to the expected timelines based on increasing the size of the studies. And as I mentioned, we've aligned this -- the program with FDA Type C. So our original idea around PEAK separate from PINNACLE, we've decided to make PEAK larger in and of itself and also to extend that to a larger number of patients. So initially, our thought was maybe 150 patients, but now we're looking at for the PEAK core group to be 300.

After we finished enrolling that because PEAK is generally enrolling 3 to 2 to PINNACLE, we'll continue to enroll patients into what we call sort of PEAK-2, okay? And then we'll combine the PEAK-2 population with the PINNACLE population to create a second study, which we call pivotal analysis 2, and that will be ready in the second quarter of 2027. So not spending a lot of time to talk about the economic opportunity for 101, but I think it is important that ocular inflammation is the fourth leading cause of blindness among working age adults in the developed world.

MESI, which is a serious complication of ocular inflammation is the primary contributor to vision loss and blindness. So there's a tremendous unmet need. There is no biologic available today. We have a pretty attractive initial KSI-101 addressable MESI population in the United States of north of 150,000 patients. A very interesting and differentiated potent, high-strength dual MOA biologic, generating very strong data, exciting data and very safe data so far in patients, and we're putting those into a larger pivotal program, and we're expecting top line data there in the fourth quarter of this year and then into the first half of 2027.

So in summary, looking back 6 months and looking to where we are today and then thinking about where we're going to be in 6 months and 12 months or perhaps at next year's meeting, I hope that we'll be able to say that we're achieving this amazing curve of value because we're able to turn the Phase III data cards on these 3 very interesting molecules and generate data that's important for patients and therefore, for Kodiak and our investors. Thank you very much.

Thank you. I'll ask the first couple of questions, but there will be an opportunity for folks in the audience to ask a question as well. So feel free to raise your hand. Victor, I wonder on KSI-101 on PEAK and PINNACLE, if you could expand a little bit more on the rationale for the enrollment dynamic that you talked about? And was that any at all driven by what you learned from Roche and their update?

Yes. Thanks. Yes, we've spent a lot of time. We do spend a lot of time on the KSI-101 program. And on the PEAK and PINNACLE studies, we're engaged in a zone of rapid sight activation. We're enrolling a lot of patients and we're treating a lot of patients. And under this master protocol concept that we have, where essentially PEAK and PINNACLE are under 1 master protocol. So -- and as I mentioned, we're seeing about 3 PEAK patients randomized to every 2 PINNACLE patients. Obviously, when you look at the Phase I APEX data, the PEAK equivalent patients had around, let's say, 80% 3-line gain rate.

And the APEX patients that were the PINNACLE equivalent patients were lower in terms of their 3-line gain rate. That's really the critical key secondary endpoint from an FDA reg standpoint. So as we look at KSI-101, we need to make sure that we're going to deliver that as an approvable molecule for patients and for Kodiak. So the first thing that we did when we looked at the Roche study is I think that their Phase IIIs replicated their Phase I dovetail where they had 25% to 30% 3-line gain rate with that asset, anti-IL-6 only at a low dose.

So -- but then maybe their studies weren't powered properly for a placebo rate of, say, 15%, so as we looked at PEAK, we think PEAK is highly powered, but we didn't want to have any risks. So rather than being an equal 50 per group, we decided to go to an equals 100 per group to be fully powered. I mean they would have achieved their endpoint had they been at 100 patients per group, even though I think we have a much more potent and a more powerful molecule. So we took PEAK to 300. Then at the same time, once we do that, we know that we're going to continue to have PEAK patients that are coming into the master protocol for screening.

So we don't want to throw those away. So we decided just what will define our preliminary analysis of the first, say, nominally 300 patients, but we'll continue to enroll patients into PEAK. And then meanwhile, PINNACLE is enrolling. So we'll wrap those together and have a hierarchical alpha that looks at PEAK, then into the total population. So we think that's not too dissimilar from what we did in our BEACON study, where we had a combination of the BRVO, CRVO populations.

But we think this is almost -- you can't say risk-free in the business, but this is looking at the Roche data, looking at what happened to them from Phase I into Phase III, where their SANDCAT study didn't meet the endpoint, but we, of course, have a much stronger molecule. But we didn't want to rest on that alone. We decided to flow all of these ideas into sort of an enhanced pivotal program in the context of the master protocol that gives us PEAK and then PEAK plus PINNACLE together with a very high confidence that we can deliver the molecule to the investment community and to the patients.

Just on PEAK and PINNACLE though, like can you talk to us about the patient populations being enrolled in both in this dynamic of 3-line gainer versus 3-line preservation, right? And why that's important?

Yes. Well, what we decided to do is to really power the studies for 3-line gain and not necessarily have to rely on preservation of preventing 3-line loss because we have data from APEX in 3-line gain and of course, Roche has data in 3-line gain, but we didn't have specific data, let's say, on preservation, right, preventing 3-line loss. So we do believe that's important and powerful, but we decided we didn't need to take any risks of assuming anything in the PINNACLE, and we decided to power that analysis for 3-line gain. So really, it's like a risk minimization. We have data towards one. We didn't have data towards the other. We decided to dig deeper in the one where we have the data.

Questions from the audience? Feel free to raise your hand.

Just remind us on tarcocimab, BLA is being filed in 3Q '26. So it's going to be the spectrum of indications, RVO, wet AMD and DR.

Yes. That's correct. I think it's interesting now because we're just right in front of the data from GLOW2 that we should evaluate in that moment, what is tarcocimab coming out of GLOW2 because we won't have the DAYBREAK data yet in wet AMD. I think tarcocimab can generate from the GLOW2 a profile that in and of itself should be very valuable, too positive, let's hope, studies in DR, right? Nice safety of this post-change formulation and bringing through the BEACON study, the RVO indication. So we could file.

That's a fileable profile. It doesn't matter what happens in wet with DAYBREAK. Obviously, we're very excited and very optimistic about DAYBREAK. But to be honest, like a twice-a year biologic for DR that really opens that up and also has 3 loading doses, so it could be used in DME patients and the DR twice a year profile and the ability to sell into RVO, where we have a very powerful action, and we had 75% of the patients giving the same outcome as EYLEA in the second 6 months.

So I think it's a very valuable profile in those 2 diseases. Of course, we'll finish some of the drug product manufacturing, and we'll wait for DAYBREAK because we're optimistic and then we'll be essentially ready to rapidly throw that clinical data from DAYBREAK until like an integrated efficacy and safety and file the BLA very shortly after, let's hope for a mid-September top line data readout for DAYBREAK.

And then just remind us of what you think a win scenario here is in GLOW2. You reviewed GLOW1 was positive. You've made some changes. Is it broadly replication of what we already know? Or would you be looking for something a little bit different given the changes?

Well, the most important is to have a very strong statistically significant superiority on the primary endpoint. Also to show good strength on the key secondary endpoint of preventing the sight threatening complications. But I think we hopefully will achieve that. I think showing nice safety of the post-change formulation will be important. We have that third loading dose. The populations are a little bit different between the GLOW1 and the GLOW2 studies.

But I think we shouldn't get lost in what the different percentages are. I think it shouldn't be a close win the superiority. Just like in GLOW1, it should be a definitive statistical win for tarcocimab in DR plus very good safety that signals that we've made some nice safety changes into the formulation. So safety and a definitive statistical superiority on the primary and key secondary endpoint, I think, will be very valuable and should be good enough.

How -- questions from the audience? How should we think about this dynamic of tarcocimab and the data you're going to have there and the 501 data if both of these products eventually come to market, right? Like how does Kodiak -- do you market both? Do you -- what do you do?

Well, we see sort of 101 and 501 a little bit as being like a franchise. And then the tarcocimab versus the 501, that's just different optionality and lucky, we're in control of how we do our own life cycle management. So what's important for 501, I mean, first of all, it will be interesting with FDA, is it true that maybe there's a possibility to file for approval based on single studies that are adequate, well controlled and definitive.

If that were true, there's no company better in my view than Kodiak to profit from that because we'll have tarcocimab, of course, with many studies but we'll have the DAYBREAK study with 501, which is a singleton study in wet AMD. We'll, of course, have PEAK at the end of this year plus singleton study that feeds into PINNACLE. But we are thinking about, well, for 501, what additional studies and indications should we study.

And we're nominally thinking about initiating pivotal or pivotals in the second quarter of this year to help feed how we think about the evolution of the KSI-501 asset within the retinal vascular diseases. I mean the simple answer is that we control both molecules. We haven't partnered them, and we'll be able to use them commercially to maximize, let's say, the patient outcomes and the revenue.

Maybe...

It's a good problem to have if it looks good.

Maybe final question for me. You recently did a financing. You've got these catalysts coming up. I'm assuming all the milestones are covered, but what isn't covered and what -- as you head into important readouts?

Right. Well, we were very pleased with the financing that we did and in particular, the participation by very high-quality investor groups. I think what we wanted to do as a key objective was to rebuild the Kodiak cap table to try to get us closer to where we had been before, okay, such that we have the quality that can really build Kodiak from, say, a $20 a share company to, I don't know, $50 or $100, I have no idea.

So I think we achieved that, and we also recapitalized the company and raised a bit more capital than we had anticipated, and that's a very good thing. So we definitively have capital through GLOW2, DAYBREAK and PEAK readouts. I think based on operational discipline; we could get through the PINNACLE readout. But we have to believe that we will be achieving some of our science-based outcomes in the pivotals and that those will provide additional opportunities for raising incremental capital like along the way.

Thank you, Victor.

Cool. Okay. Well, thanks a lot, Anupam. Thanks, everybody. We'll see you next year.