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

All right. I believe we're live now. So I just want to say, thanks, everybody, for taking the time to join the Citi 15th Annual Biopharma Conference. My name is Neena Bitritto-Garg. I'm one of the smid-cap biotech analyst here at Citi. And I'm very pleased to be joined for our next fireside chat by the management team for Kodiak Sciences. So just to start off, Victor, if you want to make some opening remarks just about Kodiak and the exciting things that you guys are doing, that would be great.

Thanks, Neena. Well, we're pleased to be invited to the conference. Thanks for having us. It's an exciting time at Kodiak here. Our ABC platform and our lead molecule, KSI-301, really making a tremendous amount of progress. As you may remember, we're a company that was founded back in 2009. And so we're really in our 10-plus years of design and development of the platform, our lead molecule and now increasingly also our pipeline of molecules on top of the ABC platform. So it's a very exciting time for us, having done first-in-human with KSI-301 back in -- over 2 years ago, now back in mid-2018. So we started off with a 9-patient Phase Ia study in previously treated DME patients and saw some very intriguing data actually in that population. Treatment-resistant patients by and large on a single injection, seeing durable responses in some cases, even beyond 6, 7, 8 months, with some of the physicians calling and saying, "Hey, one of the doses of your medicine reminds me of 3 to 4 doses of EYLEA." That was really based on the hints of excitement from those first 9 patients, previously treated DME by and large. So we stepped up at that time after that into a new idea for a study, what we called our Phase Ib, but which is really a Phase II equivalent study. So it's open-label on KSI, and we took treatment-naive patients into the Ib study. And that's now turned out to be, I think, 121 patients, treatment-naive, 51 in wet AMD, and then I think 30 in each of DME and retinal vein occlusion, RVO. And then we gave 3 loading doses once a month, so over 8 weeks. And then we monitor the patients every month. And we're doing this in 10-plus sites in the United States with top KOLs. So that was -- the 1a was, I think, July first-in-human. And the Ib, I think we did the first patient there in January of 2019. So we're 18-plus months into that, and we've extended it multiple times, and now I think it's going to be a 3-year study in treatment-naive patients evaluated monthly. And for each of those 3 diseases, as we bring the patients back monthly, we have criteria, as you know, to whether we retreat them or not. And we monitor vision and OCT and safety all along the way. And that's really been a wonderful study. It's generating really a tremendously interesting data set in these treatment-naive patients and with KSI-301. And it's allowed us to look at the power, the efficacy of the medicine, of course, the safety and also the durability and how long the patients can go in between injections. And I would say that the data that we're generating in the Ib and the safety, efficacy and durability profile really continues to be beyond our expectations as drug developers. And Jason being a card-carrying ophthalmologist, and we have a deeper ophthalmology team as well under Jason, having hired from a number of Genentech and Bayer, for example. So it's very exciting. And then we stepped from the Phase Ib study. The idea was use that to educate the design of pivotal studies in the 3 big indications, right? Wet AMD, DME and RVO, and why not the earlier diabetic retinopathy as well, right, NPDR. And so that's what we've done. So the DAZZLE wet AMD study, as you know, is our wet AMD pivotal. So as we mentioned back in July at the ASRS meeting and our kind of webcast at that time, DAZZLE is going to be approximately 550 patients. It'll have a very similar patient population to the Phase Ib with 80% plus or minus of patients coming from the United States and 20% plus or minus of patients coming from Europe, by and large. And as we mentioned, DAZZLE was enrolling again very quickly in nearing the end of its enrollment. So we do believe we're likely to be last patient in, hopefully, by the end of this year for DAZZLE. And then with the 12-month end point, that means close to the end of 2021 or early 2022, let's say, Q1 2022, we're looking at top line data for DAZZLE. And we're close to completion of that enrollment, and we're feeling very good about what we can see in the massive study. A lot of conviction and a lot of confidence coming out of that visibility. Now we're also have an aggressive regulatory clin-reg plan that we've mapped out really together in collaboration, I would say, kind of with FDA, that recognizes that there's a lot of knowledge and comfort in a way around the anti-VEGF biologics space. And so therefore, we don't need to run 2 pivotal studies in each of the 3 indications for a total of 6. Rather, our regulatory strategy and the agreements we have are, we're planning to run 2 pivotals in DME, and then one each in wet AMD and RVO and then another one in earlier diabetic retinopathy. That suite of pivotals will be put into one BLA. And we believe we're on track still for that filing in 2022. So to circle back, we did the Phase Ia, moved that into a pretty interesting basket study in Phase Ib, continuing to generate really wonderful data on a monthly basis, the Ib, then into DAZZLE nearing last patient in. Now we're -- we believe also, as we indicated in July and still operating on that trajectory, for the 2 DME studies and the RVO study to have first patient in probably within 30 days is our operational estimate, okay. And we hope that we can enroll those, let's say, aspirationally by mid-summer next year, 2021, maybe push a little bit further out, depending on COVID, for example. But with a 6-month primary endpoint for RVO and a 12-month primary endpoint for DME, that puts us for top line data Q1, Q2 and Q3 of 2022 for the 3 major indications. So there's a lot of execution that we're focused on. We're very pleased with the data coming out of our clinical studies, and we've begun to focus also quite a lot on manufacturing and the validation and PPQ batches. And our objective will be to provide and service the market in the first 12 months with many millions of dosage forms. And hopefully, that would be in prefilled syringe, and we're working on that in parallel with vial format. So we're building the company. We're well capitalized with north of $400 million of cash. And the second tranche of the Baker Brothers royalty deal that we did last December is also poised and essentially driven off of operational milestones in our clinical program. So we also are pleased overall with the competitive other molecules or technologies that are being developed towards retina. And we feel that those profiles, to a certain degree, are a bit underperforming, whereas we hope that we'll continue to outperform even our own expectations. So we look forward to the next set of quarters as we hope to continue to deliver for ourselves and our shareholders and our patients.

Great. Thanks so much. That was a great overview. So I guess just thinking about the platform, kind of the ABC platform, could you just walk us through kind of the technology there? And what the importance is of actually improving durability kind of in retinal diseases? I mean you touched on this a bit, right, and we've seen this play out commercially with EYLEA coming on the market and then Beovu. But if you could just talk a little bit about, yes, the actual technology kind of behind the platform. Yes.

Right. Thank you, Neena. Well, there's 2 ways to think about platforms. One way is to say that as the drug and the platform progresses, you want the platform to kind of disappear in the background and it becomes not relevant, right, because it's about drug development, and it's about the patient and the physician. And they shouldn't have to care about what the platform is, you want to give them a safe, effective medicine that solves an important problem. So one way to think about it is we've done a wonderful job on the design of our platform because we're seeing real safety, efficacy and durability. And physicians are injecting it with the same types of needles and in the same way as the existing medicines, right? There are 25 million-plus intravitreal injections done annually. And so with liquid aqueous optically clear biologics. And that's fundamentally what we've built in a very elegant way, the special kind of sauce is under the hood, and the physician and the patient don't really need to know much about it. So that's really kind of neat. It fades away like a mist and you have this optically clear thing that you're doing in the same way. And physicians can stock their offices, and they can get reimbursed in all the same ways. That's really exciting. On the other hand, of course, the design of the platform, the specifics and the details, that's -- the nuts and bolts of that and how we had a philosophy of working with this phosphorylcholine with the material that's on the surface of all of our cells in our body and how this zwitterionic biophysics of how phosphorylcholine traps water and how it locks it into a confirmation and how it binds many times its weight and water and how that's what biocompatibility in the body is. That water isn't flushing around in the body. We think our -- as a physician, right, our body is more than 2/3 water. But actually, it's not flushing around. The water and the body is structured. It's structured and bound through phosphorylcholine. And as our cells and our proteins and all that move around, they're seeing this water as it's bound to phosphorylcholine, and that's what natural kind of needs. That was a philosophy really back when the company was formed in 2009, and there was this concept of how you operationalize that for drug development. And then if perhaps we were able to build phosphorylcholine-based polymers using what was called the controlled living polymerization manufacturing approach, which was used in other industries to build and design well-characterized materials. What if we could do that with phosphorylcholine, to create concepts around membranes, but not to bring the fatty parts of that by-layer, but just to bring this water-structuring component of that zwitterionic biophysics at phosphorylcholine? And then what if we were able to conjugate that using industry standard conjugation approaches like people use in ADCs, in antibody drug conjugates to build large, highly branched essentially macromolecular water that we bring in the environment of that protein. And what if we vertically integrated into protein engineering and antibody design and what if we work with full antibodies that are really the safest. And what if we made the polymers large and conjugated to create these large conjugates that were safe and stably linked and essentially watery. And what if we brought that, kind of, water-structured environment close to the protein, what would that do? So well, what we found is that we can make those things, and we can make them very clean, and we can make them at scale. And we can make the polymers large. So KSI-301 is 950,000 molecular weight compared to LUCENTIS at 48,000 or EYLEA at like 112,000. So we're manifold larger. But I think the misconception a little bit early on with Kodiak, people thought, well, we're going to make something so large that it will never penetrate into the tissues. It won't be active. And if it is active, it will take several months for many injections to build up the level needed. What people misunderstood is the polymer is not just mass. It's not just about size or the hydrodynamic radius. It's actually a special material that is -- allows this conjugate to noodle its way into the tissues to achieve a very high tissue penetration into the retina. So when we even compare against say EYLEA, which, as I mentioned, is around 112,000 and our conjugate is 950,000, yet we get manifold more of our drug into the retina and the core eye than does EYLEA, even though we're much larger. And that's because this phosphorylcholine creates this very interesting water-structure environment, the macromolecular water. And it also increases the potency of the medicine because this phosphorylcholine favors hydrophilic watery interactions of one amino acid to its binding partner, okay, with the molecule of water in the middle, that's kind of how it happens. That's favored by the phosphorylcholine sheathing on the protein. And hydrophobic interactions of, say, the antibody having a little greasy patch, finding some greasy patch on another area of a cell or an extracellular matrix, that's disfavored in the context of our philosophy of phosphorylcholine. And we're seeing the benefits there as well. And of course, the safety that we're seeing is also really tremendous. So a lot of philosophies around where we came and then really a lot of excellence in the design and a lot of troubleshooting over the years, a lot of problems that we recognized, we were able to fix those and then to get to the next generation. So as a 10-plus year company, really focused on the application of our platform and to build of the ABC platform and asking if we build these things, is that really useful to solve the problem in retina? And the problem, of course, is not safety so much because EYLEA and LUCENTIS are very safe. It's not so much efficacy. More than 2/3 of patients respond well, let's say, in wet AMD to the anti-VEGFs. But really, it's the durability that the anti-VEGF mechanism has a tremendous potential to improve vision immediately for treatment-naive patients and then to maintain that vision over like 10 years. But the problem is that as people space out injections and you get drug holidays, the benefit, say, in wet AMD decreases over time and then patients essentially are burning out. So the durability, the usefulness of that in wet AMD is paramount. Now that really does need to be solved in a way that's familiar and safe. However, there's even future opportunity if you have a platform that drives durability in other diseases like dry AMD or even the retina parts of glaucoma. And macular degeneration is the leading cause of blindness in working-age people in the developed world. And dry AMD is 10x more prevalent really than wet AMD. But there really is no therapy for that, and you really need something that's much more durable because dry AMD is a disease that's, kind of, less acute in a way. Your blindness is emerging slowly over a period of decades, and you really need a medicine that would be, say, 2, 3, 4 times a year, not something that's, like, really once a month or every other month, which is what you currently need for today's existing durability. So -- and even in glaucoma, which is another big disease, there's a lot of components of the retina. And that's really, sort of, more neuro protection. And there, again, you need a platform that can give you a real benefit with 2, 3, 4 injections a year, and to do that safely. And I think what we're validating with KSI-301, we're really showing is the safety, efficacy and durability of our platform, and we have a pipeline that's quite exciting of KSI-501, that's a bi-specific conjugate, and we have our new triplets that actually bring something even more special to the table for things like dry AMD and glaucoma. So we're very pleased, and it's really been a philosophy and a 10-year journey. And it's really accelerating every month now with the Ia study, the Ib study, DAZZLE, the new pivotals, which we call BEACON, GLEAM and GLIMMER that we hope to get started this quarter.

Great. So now I want to talk a little bit about the Phase Ib data that you've presented. So just, kind of, thinking about kind of this trade-off between dosing frequency and efficacy on visual acuity and then on OCT, the CST. I guess how do physicians, kind of, think about that trade-off between the frequency and visual acuity and CST? I mean, how are they generally treating patients? And how well do the kind of assessments that physicians use correlate with the retreatment criteria that you use in the Phase Ib?

Jason?

Yes, sure. So couple of different parts to that question, so let me try to take them in turn. I think if you look across the different ways that people are treating patients right now. The -- many people treat on a, what we call, treatment-extend schedule in the United States. And so they tend to use OCT because that's kind of a measure of disease activity that you can quantify. It's very objective on the images. And vision that we do in clinical trials is not necessarily what they do in clinical practice every day. So -- and then those treatment extend paradigms tend to result in people being treated maybe every 4 weeks, 6 weeks, 8 weeks, sometimes up to 10 weeks with, say, EYLEA, maybe 1/3 of patients are on 4-week dosing, 1/3 of patients 8-week dosing and so on. But overall, I think at the end of the day, what matters to the patient is their vision, right? And they may notice their vision getting a little bit worse, and they'll come in sooner. But generally, the patient is fundamentally concerned about the vision and the goal of the physician is to, more or less, optimize the therapy for the patient like stretch the interval as long as they can while maintaining that vision and a reasonable amount of disease activity. And people debate what reasonable is, is it treated totally dry on the OCT? Do they tolerate some fluid? That's kind of more art than science. And most of the studies show that there's not any meaningful difference in vision if you -- regardless of how you do the dosing. So then to other studies, I think our -- fundamentally, our goal is in the Phase Ib study is to explore what the dosing interval possibilities are, right? And we used retreatment criteria that are very similar to those in all of the other large ongoing studies, where you treat the patient with a bunch of fixed doses and then you follow them and see how long it takes essentially for the disease to come back by a certain amount and then you retreat them again. And that's been the same design of -- similar design for the HAWK and HARRIER studies that Novartis ran for some of the faricimab studies for many of the studies that have gone on in the last, like, 5 years. They're all criteria-based retreatment. And those criteria are never exactly what an individual physician will do in their practice, but it's meant to be sort of the best thinking that you can then apply in a clinical study across a large number of sites. And I think what all of those studies so far have shown is that the results are very similar to fixed interval EYLEA dosing, right? If you look at the results of the Novartis program, for instance, where they did criteria-driven retreatment for brolucizumab versus fixed interval EYLEA because that's the standard comparator, the results are very similar. So then what does that translate into our durability and what we are going to see? I think what we've seen and what we would expect with other anti-VEGFs is that if you treat enough patients, you get like a histogram of how long the patients can be dosed for. Some people require more frequent dosing, some people require less frequent dosing. And we can build that histogram. And what we can clearly see is that the distribution of injection intervals with KSI-301 across wet AMD, across DME and RVO is clearly different than what the expected dosing interval would be with, say, LUCENTIS or EYLEA. Obviously, it's an open-label study. It doesn't have an active comparator, but there's a lot of data on how often people are treated with LUCENTIS or EYLEA in these different diseases. And then we have our data, and that histogram has clearly shifted to the right in terms of durability. So I think no matter which way you want to look at the data, we're clearly showing, I'd say, highly differentiated durability relative to what would be expected with the existing anti-VEGFs. I think I covered all the parts of your question, Neena. Victor, is there anything else to add there?

I think the main thing is the end is that today, there is a distribution of dosing, right, for 2 to 4 weeks, 4 to 6 weeks and 10 to 12 weeks or something for EYLEA and LUCENTIS. And our data is really showing that our distribution has really shifted very far to the right, and there's a lot of white space in the middle. And I think that's what's really important when people see our durability now and when they see it coming out of our pivotals, we think it really will be very differentiated, not only from the labeled regimens for EYLEA and LUCENTIS, but actually from the regimen in terms of how people are using those agents and how we expect people to be able to use our agent. That's what's really important in the end. And we're designing our pivotals to really reflect that white space in the middle, so that we could really be a singular agent, right? We're trying to really build our clinical program, our regulatory program, to really be able to define a first-line agent.

Absolutely. Okay, great. So then in terms of -- I know you touched on in your opening remarks about the, kind of, regulatory strategy and talking with the FDA and regulators about developing this, kind of, comprehensive strategy across multiple indications. But just curious about, I guess, why the need for 2 studies in DME? And then also if you could talk a little bit about -- or I'll ask about DAZZLE later. So yes, we'll start with that. Why 2 studies for DME versus one study for some of the other indications?

Yes. Jason, do you want to kind of take that into a little bit of a collaboration on it?

Yes, sure. So the way that these medicines have been developed in the past, the indications have gone kind of in series. Wet AMD studies, DME studies, RVO studies, there is a sequence. And what FDA has typically wanted are -- if you go back to the law on the guidance, adequate and well-conducted studies, right, as in more than one. And so people would do 2 wet AMD studies, and they'd get that approval. And then DME, they would do 2 studies and then RVO, they would also do 2 studies. So we took all of the precedent in the field as well as our own data and went and had this discussion with the agency. And the efficacy of anti-VEGFs and the similarity of the efficacy is now very well understood as a class, right? There's multiple medicines in the class. And so there's a good understanding of how the medicines behave. And so what we negotiated with the agency is that we could do 2 studies in one indication and then one study each in the subsequent indications so that, that would provide a data set that would allow them to understand the efficacy in -- across the indications, assuming that all of those studies are successful, right? So why 2 studies in DME? We think that that's a really important area where we can be highly differentiated. There's a lot of patients. There's a big unmet need there for a more durable therapy. And -- yes, so I think 2 studies in DME was a very logical approach. And then we can sub -- we have -- the core package will end up being essentially those 4 studies. GLEAM and GLIMMER, which are the 2 studies in DME; plus DAZZLE, which is the ongoing wet AMD study; and then BEACON, which is the vein occlusion study. And then that together would form a comprehensive BLA. And I think actually, putting all the data in front of the agency in -- at one time actually gives them more information, more safety, efficacy and durability information to understand the whole profile of the medicine and make the best regulatory decisions.

Okay. Great. No, that makes a lot of sense. Okay. So then in terms of DAZZLE, I believe there were some modifications made to the design just to make it kind of -- to bolster it as a registrational study. So can you walk through kind of the design of DAZZLE today? And kind of differences versus the Phase Ib that you've made to the inclusion criteria and the retreatment criteria and things like that to kind of tighten up the patient population and the results of the primary endpoint and things like that?

Yes. Jason?

Fundamentally, it's an exploratory study. And DAZZLE is a noninferiority study against a quite active-active comparator. So what we wanted to do was tighten up some of the criteria on the re-groupings because of the criteria-driven dosing to make sure that we're maximizing the likelihood of success of noninferiority. So those criteria really like changes on the margin. For instance, is it like a 50-micron difference versus a 75-micron difference in terms of OCT? Because those criteria, not -- in DAZZLE, the only purpose of the criteria is to determine you want a 12-week, a 16-week or a 20-week dosing interval? Whereas in Phase Ib, those criteria used to determine, did you get treated that day or not? So the purpose is a little bit different. Other than that, broadly, the inclusion and exclusion criteria are similar. The DAZZLE patient population is going to be something like 80% from the United States. So a very similar treatment-naive patient population to what we've seen in the Phase Ib study, which is at a smaller number of sites in the United States. And then we've -- the DAZZLE sample size is now 550 patients, which provides greater than 90% power for the noninferiority comparison against EYLEA. So fundamentally, it's always been designed as a pivotal study. We did increase the sample size to increase the statistical power with, obviously, the same noninferiority margin and inclusion criteria that has always had.

We spent some time on the webcast on July 27, where we went through each of the pivotals. And the indications, and we have a presentation associated with the 27th, that's on our IR page. And it shows the Phase Ib, and it also shows the edits that we've made in the pivotal design for each study in the context of, sort of, what we call the high margin of safety being designed into the pivotal studies. That further increased, kind of, the probability of success, right? Tighter criteria for the disease activity assessments, right? Shorter durability intervals, high statistical power, while maintaining a very similar 80% U.S. population. So for each of wet AMD, DME and RVO, right, this concept of introduce -- using the Phase Ib study to have a high-margin of safety and predictability in each pivotal. And we detailed some of that in the deck that's on the July 27 on the website.

Sure. Absolutely. Okay. So then in terms of thinking about kind of the primary endpoint for DAZZLE, of course, it's best corrected visual acuity. So thinking about the fact that patients can be dosed in different dosing frequency groups and that as over time, we're seeing these wet AMD trials enroll a milder kind of patient population with better visual acuity at baseline, how should we think about the -- all of those dynamics and how that could impact both, the KSI-301 and the EYLEA arms in DAZZLE?

Right. Well, it's a great question. What's really nice about retina is that there are many predicate studies now that have been run, right? ANCHOR and MARINA and View 1 and 2, also HAWK HARRIER are pivotal. And it's a bit of a changing landscape in a predictable way, in particular, say, thinking about baseline vision, right? With patients over time now being diagnosed, for example, a bit earlier and with better vision. And so the opportunity for showing with an anti-VEGF agent, really a strong boost in vision. For example, ANCHOR and MARINA, I think, was around 11 letter mean change in BCVA from baseline through the end of their, say, 52-week or 1-year endpoint. In View 1, 2 with EYLEA, a bit less and HAWK HARRIER in some way in the range of, like, say, even 5-6 letter gain, right, with patients coming in with a better baseline vision. Now we also did some discussion around baseline vision, I think, in the webinar as well. I think our baseline vision in the Phase Ib study is a bit north of that for HAWK HARRIER. And I believe even in DAZZLE, given that we have the 80% overlap in terms of the patient population, right, the baseline vision will be at around the 1b or perhaps even a little bit better. So how does that translate? What's the relevance? I think the most important thing is we want to show the same safety and the same efficacy as LUCENTIS or EYLEA, right? And we want to have a game-changing durability. So on the efficacy then, you take a look at the baseline vision in the Ib, and it looks very strong, perfectly appropriate. And of course, even further supplemented with the very powerful potency that we see in DME and RVO because wet AMD is a more variable, kind of, disease in some way in terms of its response to anti-VEGF therapy, looking very, very appropriate in the Ib. So DAZZLE coming in at fairly high baseline vision, well, why is efficacy important? We want to show the same efficacy, right? So when you think about probability of success of KSI-301 in DAZZLE as a pivotal, the same efficacy, the same safety, i.e., noninferior to EYLEA and game-changing durability. So with baseline visions going up, right, and then the absolute value of the mean change in BCVA over time predictably going down, what's interesting is that the noninferiority margin that we use in these studies in the collaboration with FDA of 4 letters isn't changing. So then as you go from, say, a 12-letter gain to, say, maybe a 6-letter gain, right, the probability, really, of showing noninferiority in some way, right, is much higher in a way. And so it also -- as baseline vision goes up with the ceiling effect and with DAZZLE being with fairly high baseline, mean baseline vision, we have perhaps an increasing probability of really showing good noninferiority with our medicine. So it's kind of interesting. But of course, we should expect that both EYLEA and KSI will show lower mean change in BCVA than View 1, 2, right? And perhaps even a bit lower than for both of those agents in HAWK HARRIER given that baseline visions are climbing. But actually, the consequence of that is it increases the probability of success overall of showing good noninferior efficacy against EYLEA, if that makes sense.

Definitely. No, that makes sense. That makes sense. And as we think about the percentage of patients that are kind of at some of these longer dosing frequency durations in DAZZLE, what percentage of patients would you say in, like, a 16-week interval or a 20-week interval, would you consider to really be kind of game-changing?

Well, it depends a little bit who you talk to and what -- how they think about it and also how the bar -- I think, with KSI-301 generating really powerful durability data, it's really moving our own goalpost. Because in some way, if you spoke to physicians, let's say, 2 years ago, they might say, if you gave me 2 real weeks of durability better than EYLEA, and it's safe and efficacious, I'm going to move all my patients to that. In other words, even a 2-week increment is really, really important. And I think we saw that early on a little bit with Beovu before the people became more focused on the retinal artery occlusion and the vasculitis. But before there was a real focus on that, people -- physicians were starting to take patients that were on a 2-week regimen, and they were then able to move them to 4-week regimen. And that was game-changing for the patient or a patient who's on monthly and they could get the patient to 6 weeks or somebody on 6 weeks getting them to 8 weeks. Those increments are 10 weeks to 12 weeks. Those increments are very powerful in patients' lives, and therefore, the ability for the patient to avoid drug holidays and to really maintain the benefit of anti-VEGF therapy over the long term. So some physicians would say, if you just give me 2 real weeks, that's all I need. Of course, our objective is to do a lot more than that. And as I said, I think the medicine is really over delivering. So -- and as you indicated, there's a bit of nuance in the field around, well, is it the intervals you see in a pivotal study? Or is it the intervals as the physicians wants to make those decisions based on bringing the patient in with OCT, right? In Europe, they may not have the patients see the physician every month, they might not do an OCT every month, right? They'll bring the patient in on its label. Whereas in the United States, the physician may want to do an OCT every month or every other month to determine what happens. I think in our pivotal study, we can fall down tremendously from where we are in our Phase Ib and still be by far the best agent, right? Because our label is already going to be the best label even at the worst durability grouping, right? We're going to have 100% of patients on q.12 weeks or better. Okay. Of course, we're also going to be generating data at 4 months and at 5 months. So what's game-changing? I think where should we set the bar, right? I mean, if we had 50% or more of patients at 4 months or longer, that's already game-changing in the field because even everybody at 12 weeks. But we don't want to fight it out in the trenches at 12 weeks, 3 months. I think our molecule has the potential to really bring 50% of patients to 20 weeks. That's what the Ib data show. We've tightened the criteria a little bit. But certainly, we don't need to hold ourselves to that standard. I think we're falling down quite a lot from the 1b if we don't bring 50% or more to 4 months or longer. So I think we're hopeful that we can do that in wet AMD. And I think that will be a real powerful showcase to really help be very clear, right, that there's a tremendous amount of durability and white space between us and the next closest anti-VEGF biologic.

Absolutely. We only have a couple of minutes here, but I just wanted to touch to on safety. And you did bring this up in your opener as well. But of course, with Beovu kind of seeing some intraocular inflammation signals earlier this year. I guess, could you just talk a little bit more about the safety data that you have so far from the Phase Ib? And what kind of gives you confidence that -- in the safety profile?

Right. Safety is really critical. The molecule needs to have the same safety as LUCENTIS or EYLEA. That's our objective. Of course, we're bringing some additional benefit on the durability, but that really should be the gravy. I think we need the same safety as LUCENTIS and EYLEA. That's what we're seeing. We track the Phase Ib patients on a monthly basis. And also we've done a data cut in the DAZZLE safety database as well. And of course, well masked for vision. And durability, safety in terms of the comparison, EYLEA, is more of an absolute end point. We're very, very pleased with what we're seeing. As we reported, we've looked at data in over 300 patients on KSI-301, representing more than 150 patient years of exposure across treatment-naive populations, which means broad in wet AMD, DME and RVO. And we're seeing safety that's in line or better than LUCENTIS or EYLEA. So that's really, really important. I think the next thing to say is, well, are we afraid or should the community be afraid based on, say, the experience with Abicipar or Beovu? I think what's important to recognize about those molecules is the types of inflammation, which they had, which were quite serious. It's not, I think, accurate to say that they emerged only when they were very late in their development. I think it's really important to recognize that those very severe types of inflammation were present really all the way from the beginning of those programs all the way back in FDA adverse events, databases back to 2015 when really only small numbers of patients, tens of patients, or say, maybe even fewer than 100 patients have been dosed. So the overall percentage of those very severe types of inflammation were present at the beginning and then were fairly consistent through their development and approval. So we're not seeing any of that. We haven't seen any of that yet. We're in many more patients than that. In the eye of every analyst or every investor, every shareholder, every physician, it's a question of, well, how much data do we need to show, right, for people to feel comfortable that we've sort of crossed the rubicon a little bit in terms of having a safety profile along the lines of LUCENTIS or EYLEA? And we're continuing to dose patients who are continuing to enroll more patients where continuing to turn additional pivotal studies on with the agent. We're very pleased with the safety profile. We can't predict the future, but we continue to believe that we have a safety efficacy and durability profile that can meet our objective and be disruptive for patients. And I think if people really focus on the number of patients that we dosed and the number of injections we've given in the variety of the indications, and they really do that true comparison back to Abicipar and brolucizumab, they'll see that our agent is already very differentiated from those agents in terms of safety. And that, in fact, we're really in the game with LUCENTIS and EYLEA very clearly.

Great. So I know we only have 1 minute here, but I did want to ask just kind of a high-level strategic question. So know -- of course, we know the plan now is to initiate these next few studies and potential NDA submission -- or a BLA submission in 2022. But I guess, what happens after that? I mean, what are your thoughts in terms of building a commercial organization, U.S., EU, all those sorts of things?

Right. Well, you -- in this business, it's important to take a long-term view because that way -- we say we make the long-term decisions every day, right? So you have to remain agile and you have to remain open. You want to have a long-term view and think, well, could we be a commercial company in retina? Can we build a global retina franchise company? You have to think that you can so that you make the right long-term decisions. Whether we get there, we just don't know, and we remain agile. I do think, though, as we move through different successive stages of the company, and now we're really in a pivotal study execution phase, I mean we really built the framework for that and are doing well there, we begin to think about, well, how many commercial doses do we need to be providing from a manufacturing standpoint. And we begin to think about what's the commercial complexity in the marketplace. And how do we begin to explain why we believe we can be a first-line agent and what that means in terms of market share and what that means in terms of LUCENTIS and EYLEA biosimilars and our place in that rightful market? And are we the right party to do that? I don't see why we couldn't be the right party and the right franchise to compete there. But as we begin to see ourselves and think of ourselves as a precommercial stage company, and we begin to learn more, it will be exciting to, kind of, see where we go and what we're able to do, but we always remain open and agile.

Great. Well, we are out of time. So I just want to say, thank you so much for taking the time to speak with me. Your insights have been very, very helpful, and I appreciate it again.

Thank you very much, Neena. Thanks for the opportunity to talk about Kodiak. Thanks.

Thanks, Neena.

Of course. Thank you. All right. Bye, guys.