Kodiak Sciences Inc. (KOD) Earnings Call Transcript & Summary

Hello, everyone. My name is Gena Wang. I'm smid-cap biotech analyst at Barclays. Welcome to our second virtual Global Healthcare Conference. First, I wish everyone to stay healthy, and I would like to thank all the participants, investors, companies and especially our event team, our corporate access team who made this virtual healthcare conference possible. With that, I would like to introduce our next presenter, Victor Perlroth from -- Chief Executive Officer from Kodiak. Victor, I hand over to you.

Thank you so much, Gena, and thanks to the Barclays team for having us at the conference. We'll move in to our presentation. Thank you very much. I'm the Chief Executive Officer of Kodiak Sciences. We're an ophthalmology-focused biotech, in particular, focused on new treatments for diseases in retina. Please note, there'll be some forward-looking statements made in the discussion. So where are we with Kodiak as a company and as a biotech? Well, we're running 4 pivotal trials today. One of them having been fully enrolled and the other 3 in enrollment phase. We're looking to bring those studies together into a single BLA filing expected in 2022. Our lead molecule, KSI-301 has now data from more than 2,000 injections in more than 500 patients, representing more than 350 patient-years of exposure in treatment naive and representative populations in 3 of the major diseases of retina, wet AMD, DME, i.e., diabetic macular edema, and RVO, retinal vein occlusion. We're very pleased and are proud of the safety, efficacy and durability data that we're generating with KSI-301. We continue to see a safety profile that's tracking very well with standard of care biologics, LUCENTIS and EYLEA. We're very pleased with the efficacy data as well in the recent Phase Ib study with 1 year data presented in February. We're seeing strong vision and strong retinal anatomy gains and improvements that are in line with current anti-VEGF agents. And on the durability, really beyond our expectation, we're seeing 2 out of every 3 patients going 6 months or longer in all 3 indications, wet AMD, DME and RVO. This data in the Phase Ib study in treatment-naive patients sets us up nicely for our pivotal study program, which we optimize those designs based on the evolving Phase Ib data. Our objective is to show the same safety and efficacy as EYLEA in our pivotal program and to show disruptive durability. The concept is that our durability would be substantively shifted from the next best agent, in particular biological agents, which is -- represents the market today. So with DAZZLE, our wet AMD study enrollment completed, looking towards last patient, last visit towards the end of this year, November-December time frame. And looking at BEACON, our study in retinal vein occlusion; and GLEAM and GLIMMER, our, studies in diabetic macular edema, now enrolling globally. Data from those studies, as I mentioned, is expected also in 2022. So we've made a number of kind of step edits or modifications to the pivotal studies building on the Phase Ib towards optimizing their probability of success. Notably, we're taking only the highest dose, the 5.0 milligrams, forward. We've dramatically increased, of course, the size of the cohorts to increase statistical power. We've tightened the disease criteria for the disease activity assessments that we use for durability interval grouping, which will, on the margin, work to increase vision gains in the pivotals versus EYLEA, while still keeping them close to the Phase Ib, which keeps us close to the types of durability intervals that we're seeing with EYLEA. We're maintaining also a similar 80-plus percent U.S. treatment-naive population, such that the data that we can be generating in the pivotals can track well to the Phase Ib experience. Based on our view on the totality of all the data that we're generating and the response, I think, in the community, we're operating and investing in KSI-301 in our ABC Platform and on Kodiak's journey with conviction. We're on track for a single BLA filing in the major diseases, and we're working hard on scaling manufacturing, such that we won't have a gap at launch to be able to supply growing pieces, growing parts of the market, with an objective of a commercial supply goal of north of 2 million or 2.5 million doses, hopefully, prefilled syringes, even in the first year of launch. And with our ABC Platform, we're developing bispecific and triplet ABC Medicines for other retinal diseases, in particular, multi-mechanism diseases, including dry AMD and glaucoma, which, of course, represent tremendous unmet needs for patients globally. And we're very pleased now even with the faricimab data, the Phase III data having been announced both from a PR standpoint, press release standpoint, and also the scientific data presented most recently at the Angiogenesis retina meeting. We're pleased to have some clarification of the commercial potential for KSI-301, and we're still very pleased that there's really a nice open opportunity for KSI to be the best anti-VEGF biologic with a durability profile and strong safety and efficacy that will make it a unique potential first-line agent. So we have both our KSI-301 molecule here on the top right, which shows you the antibody biopolymer conjugate structure of our ABC Platform. We also have our KSI-501 molecule, which is a bispecific antibody with a similar biopolymer, again as conjugate. And then, of course, we have our new triplet molecules further back in our pipeline that have both hundreds of small molecules, what I call as high stoichiometry of loading of many hundreds of small molecules that can be mechanistic, let's say, in dry AMD biologies, together with bispecific or monospecific antibodies to create bioconjugates for these more complex diseases as we move away from really what's only validated target in retina today, which is VEGF. So focusing on KSI-301, our lead molecule, I think it's important to think a little bit about what profile may be required to meaningfully change the current paradigm for patients, for physicians, for payers in the United States, of course, but also on a global basis. So today, people really are at a less than 3- to 4-month predominant profile with LUCENTIS on a monthly regimen, EYLEA on an every 8-week regimen, but with data showing perhaps half of the EYLEA patients really needing monthly dosing. So as we shift to next-generation agents, we think more about quarterly dosing on a 3-month or maybe a 4-month or maybe a 5- to 6-month predominant regimen. Our data in our Phase Ib study is a 6-month predominant profile in all 3 of the indications, wet AMD, DME and RVO. I think to the extent that our wet AMD pivotal DAZZLE only takes us out to 5 months as the longest grouping, a 5-month predominant profile coming out of the DAZZLE study in wet AMD is, I hope to say, highly likely. Of course, we have to await the data. Our DME data is also very strong in terms of 6 months or longer durability. And in our RVO study, we were also very pleased to see more than 2/3 of the patients going 6 months or longer as well. So -- and we're beginning to think about enrolling to initiating our study in NPDR or nonproliferative diabetic retinopathy, looking at 6-month dosing every 6 months after 3, what we call initiating doses. So what profile could be required, that could be a potential game changer in the field? Well, a 5- to 6-month predominant profile would do that. And I think we're credibly the only agent, especially biological agent, which represents the market today that is in this category. We're very pleased to be executing well on our pivotal program towards evaluating KSI-301's potential. So as I mentioned, in terms of a focus on KSI-301, again, we're taking in the Ib study, 2 out of 3 patients on a greater than 6-month treatment-free interval at year 1 in wet AMD, DME and RVO. And importantly, this study, the Ib, was done in treatment-naive patients, which is the same population that we'll be testing in the pivotal studies and gives a very sort of honest unbiased view of the potential durability that the molecule can deliver. So that's, I think, from a durability standpoint, 2 out of 3 patients at 6 months. Other ways to look at it would be the mean number of injections during year 1, which also includes a loading phase, where we've used 3 loading doses once monthly. And also, we can begin to think in the future about mean number of injections that would be required in year 2 once you're past the loading phase. So in wet AMD, we have a mean of only 2 additional doses after the loading phase. In DME, we had a mean of only 1 additional dose after the loading phase. And in RVO of 1.7, together with safety and efficacy, strongly in line with today's first-line agents. Again, another way to look at the data here on Slide 12. In terms of mean number of doses in the different indications in terms of visual acuity gains at year 1. And also in terms of the retinal anatomy or the fluid change. In our view, both the mean number of doses is really game-changing and our visual acuity very strong in all 3 diseases. And the retinal anatomy gain is also strong and very appropriate. I think it's important when we want to think a little bit about vision at year 1. We can compare data from our Phase Ib study, recognizing that it's not a pivotal study, but we'll be running DAZZLE in wet AMD, and our GLEAM and GLIMMER pivotal studies in DME. But I think what's really important is to look at the baseline vision of these studies. This is a historical view of studies looking at View 1, 2 for EYLEA, HAWK and HARRIER for brolucizumab and the new TENAYA and LUCERNE studies with faricimab. I think if you want to think about the population in these pivotal studies changing over time, that's certainly true. And then you can see the types of visual acuity gains in absolute letters over time. Looking back to View or EYLEA or aflibercept on its every 2-milligram every 8-week regimen, started at a low baseline of around 50 gaining, say, 8.9 letters to reach about 60 letters of vision. But over time, with the baselines increasing, looking at HAWK and HARRIER with a baseline of closer to, say, 60; or in the TENAYA study with a baseline also close to 60. And I think TENAYA is a very good approximation for Kodiak's DAZZLE study. It's really the most proximal study in terms of time and it's a predominant U.S. study. DAZZLE is also, obviously, was just enrolled and is also an 80% predominant U.S. study in terms of treatment-naive patients enrolled. So if you look at our Phase Ib data with the baseline of around 63 to 64 letters, which is similar to where we're going to end up in DAZZLE, with a 5.7 letter gain takes us to a very high ceiling of approaching 70 letters in our wet AMD. So we're approaching really a top line level of vision, which is phenomenal. And then in DME as well, we have a high baseline, but as well, we're achieving a very high ceiling. So we're very pleased with the visual acuity benefits that we're seeing in the Phase Ib and the potential read-through into our pivotals versus EYLEA. As a reminder of our platform, we build novel antibodies. We build our biopolymer out of a special material, called phosphorylcholine, which really comes from our cell surfaces and binds water in a special way. One way to think about the biopolymer is really to think of it as a form of macromolecular water. That when we link it to the antibody, brings a number of very special features and properties. We call that a structured water environment. And it's really how the body deals with water and how biological interactions happen in a cellular environment in the body. And we're essentially bringing that kind of context to the antibody in a very natural way. So that as we inject that optically clear liquid formulation into the eye, it can travel in a very sort of lubricious or frictionless manner deeply into the retina quickly, and when it gets there, it can bind and target the VEGF target in a stereospecific special manner that's very natural in the body. And that science of our platform really helps us to understand the type of durability and efficacy and safety that we're seeing with KSI-301 patients and which extends also into our concepts of our pipeline, KSI-501 and also our new triplets. We call it a generation 2 anti-VEGF agent. It's quite large by virtue of that macromolecular water. We're also dosing at a high dose, so we have a high formulation strength. We have a much longer ocular half-life due to our size, and that translates over time into a many log advantage in terms of concentration in the eye. Again, helping us to understand this type of disruptive durability numbers. We call it a more than the sum of its parts kind of structure. So a class-leading half-life tremendous bioavailability into the retina, even much higher than EYLEA, even though we're sevenfold larger. A very strong potency compared to the predicate agents. And interestingly, what you want is something very large in the eye, so it has a long half-life and a long residence time. But then once you get into the systemic circulation of the body, after you clear from the eye as an integral bioconjugate, because it's not falling apart and not degrading. Once you get into the systemic, you want it to be cleared quickly. And that's actually what our platform delivers. We inhibit through the bioconjugate and through the biopolymer attachment, the recycling of that antibody through what's called the FcRn scavenger receptor pathway, which is what gives antibody systemically a long half-life. That's inhibited in our design, and so our conjugate is cleared very rapidly, mostly through the kidney. Our goal is to work towards developing a first-line agent that can compete effectively with EYLEA, LUCENTIS and faricimab if it's going to be approved. So we're looking at running our -- so this is our full panoply of studies today in wet AMD, our DAZZLE study; in DME, GLEAM and GLIMMER; in RVO, BEACON; and soon are non-proliferative diabetic earlier kind of retinopathy study, GLOW. In the case of wet AMD, DME and RVO, we're looking at going against EYLEA as the active comparator. In DAZZLE, it's a 4 letter, non-inferiority margin for vision. And for DME and RVO, it's a 4.5 letter, non-inferiority margin. And we're looking at every 3-, 4- and 5-month dosing in DAZZLE. In DME, it's a broader spectrum of every 2, 3, 4, 5 or 6 months. And in retinal vein occlusion, it's in the initial 6-month period every 8 weeks. And then in the second 6-month period, it's more of what they call a PRN dosing versus EYLEA. And in GLOW, we'll be doing every 6-month dosing compared to Sham. We're also building capacity in our manufacturing to supply a rapid market uptake of KSI. And we're using EYLEA as an example. If you think about it year 1, 2 and 3 full year, it was able to take 18% and 33% and then 40-plus percent of the market. So we don't know, of course, what kind of market access we can have, but we're beginning to focus on that and thinking of ourselves as sort of in a pre-commercial type stage as we complete manufacturing, execution and focus on execution and quality of execution in our pivotals. We're beginning to think about, well, what it will take if KSI delivers in the pivotals for us as a company to be able to lead -- to supply and to participate as an important agent in the market. Our vision, our 2022 vision is to submit for wet AMD, DME and RVO a single BLA, and that we're on track for that. I think if we wanted to look a little bit quickly at Slide 22, which takes a look at the broader portfolio of clinical studies. The Phase Ib study having announced its 1 year data. DAZZLE being fully enrolled, going towards last patient, last visit for the primary endpoint, November, December of this year. GLEAM and GLIMMER enrolling. BEACON enrolling very strongly. And we hope to get the GLOW study started shortly. I think an important set of activities for Kodiak for this year, as I said, is quality of clinical execution, quality of manufacturing, planning and execution. And also thinking a little bit about fuller label, do we have all the studies that we want, both large pivotals as well as other studies to support the label, particularly initially in the United States. But also thinking about information we would want for the market, and potentially for launch and other types of global regulatory discussions. So that's really the focus for this year. Gena, I think I'll stop there, and we can leave some time for questions.

Okay. Sounds good. Thank you, Victor. And I think that data charts comparison across all different trials regarding the visual acuity, I think that's very informative. So Victor, maybe wanted to ask you, the -- regarding the OCT comparison, any thoughts there? When we look at the OCT comparison to the other studies, of course it's a cross-trial comparison, so we did see a little bit higher the OCT -- the retinal thickness. So any thoughts regarding -- when you're excluding the PED patient, what that data look like? And for your pivotal study, will you try to avoid or minimize the patient with a higher risk or like certain thickness of the OCT due to PED for the pivotal study to minimize any potential risk there?

Right. Well, thanks, Gena. I think that's a really important question. And there have been some questions or some discussion around the strength of the OCT, the strength of the drawing of KSI-301 as an agent. I think it's important to remember, so we do the analysis, as you mentioned, with or without the PED, the PED have quite large retinal thickness and preferred somewhat the averages that are presented. So we did an analysis where we removed PEDs, I think, of larger than 500 microns. Of course, you have PEDs that are smaller. But in order to run a basic analysis, you remove PED patients larger than 500. When we do that, we decrease the retinal thickness to around 320 microns. So I think it's important to recognize that a normal retinal anatomy, say, for a female is around 305 microns in a healthy retina and for a male it's around 315 microns. That's the definition of normal. And so when we're bringing down in our disease population, many of whom are quite old, down to 320 microns is representative of strong and appropriate drying. And it's not useful to compare, as you mentioned, across different studies because everybody uses very different approaches for thinking about how you segment the retina according to the OCT and what is or what isn't presented or what's deemed normal in a comparison. And so that's why, of course, why we're doing an active comparator. So number one is we bring down to very close to the retinal thickness of a normal individual, a healthy individual, 320 versus 305 or 315. The second thing really is to say is we also are not alone, right? We're not just looking at wet AMD, we're also looking at data in RVO and also in DME. And in particular, especially RVO, where you have very high VEGF load, that's really seen as a disease where you can look at the power of drying. And I think what we see broadly across the 3 diseases is that we have a very strong agent in terms of its ability to dry. And wet AMD is a very highly variable disease. So it's perhaps the least good disease to evaluate the power of drying, especially when we have RVO and DME. And I think to your question about, well, could we, in our pivotals, somehow maybe bias or remove patients that may have less good drying or maybe less good responders. I think the answer is that we're not doing that, and that's because we're going into a treatment-naive population. And as I mentioned in the Phase Ib, that really gives us a lot of confidence. There are many companies that are doing studies in sort of populations that go through different introductory pathways or journeys and keep some patients and remove others, and it's difficult to evaluate the power of the agent. In our case, we did treatment naive for the Phase Ib, we're doing treatment naive for the core pivotals, and we think that's a good approach. And as a result, you have some patients in wet AMD who respond better in some patients who respond worse. That's just the nature of the disease. DME and RVO, by and large, are a little bit less variable.

Okay. That makes sense. And then regarding the non-inferiority margin, you mentioned, I just want to make sure I got it right. So for the wet AMD, it's 4 letters; for DME, it's 4.5 letters; and then for RVO, it's 4 letters. Are these right?

No, no. Wet AMD is 4 and then DME and RVO are 4.5.

Okay. So both are 4.5. Okay. That's very helpful. And then any, say, secondary endpoint or any non-inferiority regarding the OCT?

No, I don't think there's any -- there's certainly no primary component. And in terms of secondary components, I think we have a lot of latitude about what we want to do there in terms of the different types of analyses. And there are many different ways and different approaches, like I said, that people evaluate the OCT. And we're generating data in many different formats. And so then we'll have the opportunity to present it, I think, based on the format that represents the best view of the science, I would say, and understate.

Okay. I think that makes sense. And then regarding the treatment interval. So your definition is last visit at 52 weeks or basically the last visit. So any thoughts on the pros and cons in the control study versus other definition you see like some variable different way of defining treatment interval? So any additional color you can give.

Yes. Well, the primary endpoint, of course, is non-inferiority of vision and the durability component or the treatment interval is a really important part of the molecule and the benefit to it, right, to deliver the same visual acuity with patients on many different durability groupings. So when you look at the Phase Ib data that we presented, for example, and we're looking here on the Slide 32, which shows the swim lanes, we're not really aware of people that have presented treatment-naive data with this type of swimmers plot in treatment-naive patients in wet AMD, DME and RVO in this way. And it's really a quite fascinating dataset actually to evaluate response to anti-VEGF agents in the patients. I think when you begin to look at the data in more detail, and you think a little bit about, well, how is it that one represents a durability interval in the study along the way you're at an endpoint. So it's a really interesting conversation, actually, and you can think about, well, what is the time to the first retreatment after the loading phase. But of course, that's less useful. So what we're talking about is looking at, well, what's the durability of the patient at the 1-year endpoint. And so if you remember for DAZZLE, our BCVA or the vision primary is a grouping of 48 and 52 weeks. And so we'll be actually evaluating the durability interval in DAZZLE in the context of that 48- to 52-week kind of interval. And I think that's what I'll say at this point. And I think for DME, it's similar. And then for RVO, as I mentioned, we have a fixed interval of 8 weeks versus 4 weeks for predicate agents in the first 6 months and then the second 6 months is more PRN dosing. So I think -- yes. So I think we'll -- our desire will be to evaluate the durability in DAZZLE for wet AMD. And in GLEAM and GLIMMER, where we have these flexible treatment arms, to evaluate the durability in a similar manner to how we evaluated it for the Phase Ib.

Okay. Thank you. We're running out of time, but this is very comprehensive overview, and we're looking forward to the future data update. And hopefully, we have a new drug class coming to the market in the future. Thank you.

Thanks so much, Gena.

Okay. Thanks a lot.

Thanks a lot. Bye.