Apellis Pharmaceuticals, Inc. (APLS) Earnings Call Transcript & Summary

[Operator Instructions] It is now my pleasure to turn today's program over to Federico Grossi of Apellis Pharmaceuticals.

Thank you very much. Good morning, everyone, and welcome to Apellis Investor Event on Pegcetacoplan, Advancing the Treatment for Geographic Atrophy or GA. My name is Federico Grossi. I'm Apellis' Chief Medical Officer and will serve as the master of ceremony for today's event. Next slide. Before we begin, I would like to make the point that we'll be making forward-looking statements that are based on the current expectations and beliefs. These statements are subject to certain risks and uncertainties, and our actual results may differ materially. I encourage you to consult the risk factors discussed on our SEC filings for additional details. The next slide. We also want to make you aware of specific disclosures related to our speakers today. So we're joined today -- the next slide, please. We're joined today by a lead of group retinal physicians and several members of our science and R&D leadership team here at Apellis for our GA program. We'd like to begin our presentations with an overview of GA, the natural history of the disease and the significant unmet medical need for new treatments in geographic atrophy. We'll then dive into complement C3 in MD and showcase pegcetacoplan as a potential treatment for this devastating retinal disease. That will include a review of our positive FILLY, our Phase II study, and the upcoming Phase IIIs, DERBY and OAKS studies, that are planned to read out on the third quarter of this year. We'll also touch on artificial intelligence technologies being advanced to support early detection and diagnosis of GA, and we'll end up the session with a QA -- with a Q&A session. Let me now introduce our Co-Founder, CEO and President of Apellis, Cedric Francois, who will provide a brief opening remarks. Cedric, please?

Thank you so much, Federico, and welcome, everyone. Sorry for the small delay. As usual, these days, we have some technical issues. We were so excited to have all of you here today to talk about geographic atrophy, a program that we have been working on for longer than a decade now and which is now coming to fruition with the readout of our 2 Phase III clinical trials in the third quarter of this year. Geographic atrophy is the advanced form of dry macular degeneration. There are no treatments for this disease. Approximately 5 million patients globally are on a path to blindness because of that disease. And we are hopeful and looking forward to potentially bringing the first treatment to patients. On the next slide, a brief summary of what Apellis currently represents. So this is a seminal year for Apellis, where our strategy which is really standing on 3 cornerstones, is going to start coming to fruition. On one hand, many of you are familiar with our program in PNH and the other systemic indications that we treat by giving pegcetacoplan subcutaneously twice per week to patients. And we intend to establish that product as a disruptive therapy across rare complement-driven diseases. Obviously, very key in that program is our PDUFA date on May 14, after which we will launch PNH in the U.S. and progress 4 additional registrational programs. But in the retina, we believe that with geographic atrophy, we have an opportunity not just to create potentially a first drug product for this disease, but it is worth mentioning, and you will hear that throughout today's presentations, that geographic atrophy is what the retinal specialist sees most of the time in their practice, either in its primary form but also when it occurs secondary to anti-VEGF injections, where after a long treatment with anti-VEGF, almost inevitably possibly -- or probably these patients end up developing atrophy as well. Also this year, and you will hear more about this in the quarters to come, we will talk about our new compounds in development. So we have worked very hard in the past couple of years to try to harness what the complement factor C3 can do in a wide range of indications, and we look forward to sharing more on that in the months to come. Three compounds will go into clinical development in the next 24 months. Next slide. And one of the things we thought we would do today is -- so we get many questions from all of you, of course, and we tried to summarize 4, what we believe, are key misconceptions that you will hear a lot about throughout the presentations that will follow. And at the end of the presentation, I will summarize for you why these are misconceptions and why it is important to contextualize the answers to these questions. So the first one is other complement inhibitors have failed. Why should pegcetacoplan work? That's a very valid question, and by the end of these presentations, hopefully, you will understand better why we should feel very good about the data that we have and the readout in the summer. Number two, how can we trust the FILLY data? This was a 246-patient clinical trial. Not only that, however. This was a trial that was designed, at the time, very much aware of all the failures that had happened and knowing that if this trial didn't give us a very clear answer, it would be difficult, maybe impossible for us to find the financing and the support to do our Phase III clinical trials. You will hear much more about that today as well. The third misconception is pegcetacoplan has no impact on visual acuity. It will take years to notice a difference. Why would people take this drug? Again, intuitively, that seems like a good question, but as you will hear in the slides and the minutes to come, that is an important misconception in a disease that severely affects these patients and for which there is currently no treatment. And the fourth misconception is the FILLY trial had safety issues. Pegcetacoplan causes wet AMD and had 2 cases of infectious endophthalmitis. The FILLY trial did have 2 cases of endophthalmitis, was associated with exudations. But the context there is very important, and these are not safety concerns, again, something that we will dive into. But without further ado, I'm going to hand it back to Federico Grossi to give a brief introduction before we go to the presenters, and at the end of the next 2 hours, I will be back with you. Thank you so much for joining, and Federico, back to you.

Thank you. Thank you, Cedric. Now we'll jump on our deep dive on GA. So first, we want to start the day by hearing directly from a patient. Caroline is one of the million patients living out there with GA across the globe that have seen their lives significantly changed due to this devastating disease as GA is. So now we'll see a video of her story. If you can please play the video. [Presentation]

Thank you. Well, Caroline's story is what motivates us here at Apellis to advance our science to fight this disease. So moving on, I would like now to introduce Dr. Caroline Baumal from Tufts Medical Center and New England Eye Center, who will provide us a deep overview of AMD. Dr. Baumal, please.

Thank you. I'm Caroline Baumal, a vitro retina specialist based in Boston, and I'll be giving an overview of age-related macular degeneration, also known as AMD. Next slide. AMD is a degenerative disease of the central region of the retina known as the macula. This macular region is important for color vision as well as many activities responsible for independence such as driving and reading. One of the major risk factors for AMD is age, and this graph shows the increased prevalence of AMD with increased age. Other risk factors include genetic and some lifestyle factors, such as smoking. AMD can be classified as wet or dry and further subclassified based on disease severity. To give you an idea of the impact of AMD, the global prevalence is currently 196 million. And as the population continues to advance and people live longer due to medical advances, the global prevalence of AMD will increase by 100 million, reaching 288 million by 2040. As a retina care provider, it's very evident in my practice that with improved health care and as patients live longer, there's been a notable increase in patients with AMD over the last 2 decades. Next slide. AMD progresses through stages. The initial phase is called early AMD, which is characterized by smaller-sized drusen and minor pigment changes. The next stage is intermediate AMD with larger drusen and retinal pigment epithelial changes. The most advanced stage is called late AMD, which is a term used to describe either dry AMD featuring geographic atrophy or wet AMD with choroidal neovascularization. Next slide. The stage of AMD may correlate with the vision. Early AMD is typically asymptomatic. Intermediate AMD is associated with some vision loss, and late AMD is often characterized by significant vision loss. Patients consider AMD to be one of the most disruptive diseases affecting their quality of life. Next slide. There have been significant breakthrough in the last 15 years which have changed wet AMD from an untreatable to a now treatable disease. This has been a game changer in the fields of retina because approximately 15% of the AMD population has wet AMD, and now many of these patients can be treated with anti-VEGF therapy. Next slide. This graph shows the improvement in vision that happens in wet AMD treated with anti-VEGF agents. Prior to anti-VEGF agents, there was really no effective way to have vision gains in wet AMD, and eventual vision loss would ensue. But the introduction of anti-VEGF agents has reduced blindness in wet AMD, and a recent U.S. [ CLAIMS ] study showed that anti-VEGF agents reduce the onset of severe vision loss or blindness in patients with new wet AMD by 43%. In contrast, there are no effective treatments for dry AMD and geographic atrophy, and this remains one of the most significant unmet needs in retina. Next slide. This is because geographic atrophy is a leading cause of blindness, affecting 5 million people worldwide and 1 million people in the United States. Geographic atrophy accounts for 20% of all legal blindness attributed to AMD, and there is no effective treatment for these people. Next slide. Natural history studies like the PROXIMA A and B study shown on the graph here have confirmed that geographic atrophy growth is progressive, constant and irreversible. Let's run this video that shows a fundus autofluorescence image, and the damage, RPE and photoreceptors is present in the black regions. And you can see that the black area, a damage and geographic atrophy increases oversize and impairs the central vision. Next slide. Geographic atrophy lesions can affect extrafoveal as well as foveal regions. The extrafoveal lesions tend to progress more quickly than the foveal lesions, but the visual acuity is typically lost over time regardless of whether lesions are extrafoveal or foveal, as you will see later on in today's discussions. It's also worth noting that the impact of geographic atrophy on visual acuity can be misleading due to foveal sparing. The fovea can be spared for some time as the geographic atrophy lesions progress around the fovea, preserving the measured visual acuity for some time, while other aspects of visual function are impaired. Next slide. Another important point is that geographic atrophy and wet AMD can coexist in the same eye. A novel, relatively new imaging technology called OCT and geography has demonstrated this, as seen in figure B, where there's an unsuspected focus of choroidal neovascularization at the edge of geographic atrophy. This type of choroidal neovascularization is not leaking, and this is called nonexudative choroidal neovascularization. Figure D shows a double-layer sign on the OCT, and this is another finding that can signify subclinical, nonexudative type 1 choroidal neovascularization. By contrast, wet AMD is characterized by choroidal neovascularization where blood and fluid leak out of the blood vessels. Next slide. Not only can geographic atrophy and wet AMD exist, geographic atrophy can be responsible for the vision loss in wet AMD patients who are undergoing anti-VEGF therapies. Note that in many patients with neovascular AMD and wet AMD, if they're followed for sufficiently long enough, they will eventually have vision loss due to geographic and macular atrophy rather than from choroidal neovascularization itself. Geographic atrophy can also be the end result of other disorders, not just AMD. It can also develop in disorders such as vitelliform macular dystrophy or retinal vein occlusion and lead to vision loss. Next slide. So to wrap it up with my key takeaways. Geographic atrophy, secondary to AMD, is a leading cause of blindness. Geographic atrophy and wet AMD can coexist. And lastly, there is no treatment for the 5 million people worldwide who have geographic atrophy, making this a significant unmet therapeutic need. Thank you.

Good morning. My name is Nancy Holekamp, and it is my pleasure to discuss the natural history and unmet need in GA this morning. I have been a retina specialist for 25 years. The first 15 years were spent as a Professor of Clinical Ophthalmology at the Washington University School of Medicine. The last 10 years have been spent as Director of Retina Services at the Pepose Vision Institute, where I see patients with age-related macular degeneration every day. I see patients with both wet and dry macular degeneration, the advanced form of geographic atrophy, and I participate in numerous clinical trials for both wet and dry macular degeneration. Next slide. So I thought we would begin by looking at a video from the NIH National Eye Institute that demonstrates vision impairment for patients with advanced geographic atrophy. Could we please roll the video? So we're going to take a typical daily activity such as going to the grocery store, and you see the central scotomas that really prevent central vision. It's amazing how much distortion can be caused by geographic atrophy centrally. And really, if you're a person with this visual disability, it's very hard to accomplish the daily tasks of living. So it's quite impactful for patients. Next slide. So with patients with central geographic atrophy, whether it's foveal or non-foveal, there's a loss of ability to see objects clearly. Straight lines appear to be wavy or distorted. This creates visual confusion because people know that a line should be straight and yet their eyes are seeing it as distorted. There's loss of color vision, and these dark areas of gray and white spots in the center of vision really preclude sharp focus and the ability to read and drive, which we will talk about in a bit. Next slide. So a study looked at the qualitative impact of geographic atrophy on patients. And essentially, GA interferes with every aspect of life, both those essential and enjoyable activities, the essential activities being safety, transportation or self-care. And the enjoyable activities would be reading for pleasure or hobbies. And there's this concept called vision-related quality of life. And this progressive, relentless disease has a profound impact, as you can tell by reading these testimonials from patients with geographic atrophy. Next slide. So I've shown you the qualitative impact. We've watched a video that mimics vision of a person with geographic atrophy. We've looked at testimonials. But now it's time to look at some quantitative data. And Apellis has collaborated with retina specialists like myself but also the American Academy of Ophthalmology IRIS Registry to produce the largest natural history study of geographic atrophy. This study includes 69,000 GA patients. And the objective was to evaluate clinical characteristics and disease progression in patients diagnosed with GA secondary to age-related macular degeneration and what we call real-world clinical practice. And they used the American Academy of Ophthalmology IRIS Registry. It's the world's largest specialty clinical data registry with almost 60 million unique patients, reflecting the experience of over 18,000 clinicians in ophthalmology practice. And I'm one of the co-authors on this study, and our methods were to divide the patient population into 2 cohorts. The first cohort has geographic atrophy in both eyes. We call that bilateral GA, and the number of patients there is 44,120 patients. Cohort 2 is where they have geographic atrophy in one eye but wet macular degeneration in the fellow eye, and there's 25,321 patients that were studied. But then as Dr. Baumal mentioned in her presentation, geographic atrophy can exist in 2 locations, extrafoveal, outside the center, or foveal. And we have, therefore, further subcategorized cohort 1 and cohort 2 into these additional groups. First -- next slide, please. So here's -- what I'm going to show you first is that up to 1/3 of study eyes progressed to loss of driving eligibility within 24 months. And I'd like to point out that the study eye for these patients was the better-seeing eye. So if we could have the first build. So at 12 months, you see that between 19.7% and 28% of patients lost the ability to have an unrestricted driver's license. That's vision of 20 -- of less than 20/40. Now if we could have the next build. You see by month 24, this number is now from about 28% to 34.3% of patients with geographic atrophy in their better-seeing eye, losing the ability to qualify for an unrestricted driver's license. And I'd like to point out that whether we're in group Ia, Ib, IIa or IIb, this disease is progressive and relentless and that the pie charts are expanding almost equally across all subgroups. The other thing I'd like to point out is that in the clear part of the pie graph, those are patients who are still driving. But let's have the next slide, and I'll show you a different study, a smaller study. This is a study of 137 patients with advanced GA, of which 55% still have their driver's license. And this is a retrospective burden of illness study. And when asked, 52% of patients with a driver's license said they did not feel confident driving during the day, and almost 90% of patients said they did not feel comfortable driving at night. So although some patients do still qualify for a driver's license, they're not driving because of the profound impact that geographic atrophy has on the daily activities of life. Next slide, please. So we've talked about driving ability, but now let's shift to legal blindness. Legal blindness has vision of 2,200 or worse and really precludes reading or perhaps even holding a job. So if we could have the first build of this slide. So what we see is that at 12 months, between 15.9% and 22.3% of patients across all 4 subgroups progressed to legal blindness, which is vision of 2,200 or worse. And now with the next build, please. We see that these numbers relentlessly progressed, and we have anywhere between 23.5% and 32.2% of patients, again, progressing to legal blindness. And I'd like to remind everyone that the better-seeing eye was the study eye in this natural history study. Next slide, please. We're going to shift gears and look at some other data that we've been able to analyze through this very large natural history study, and that is that wet AMD is not a rare occurrence in patients with GA. We're looking at the progression from GA to new onset wet macular degeneration. And if we look at cohort 1 in those eyes that had geographic atrophy in both eyes, we see that at 12 months, the progression rate was 4.7%, and at 24 months, the progression rate is 8.2%. So that's the natural history, that these eyes are progressing to choroidal neovascularization or wet AMD. But let's look at cohort 2. These are patients where they have GA in one eye and wet macular degeneration in the fellow eye. And we see that the rate of 12 months is now 13.3%, and at 24 months, it's 21.6%. So patients were nearly 3x more likely to develop new onset wet AMD in the study eye if the fellow eye also already had wet AMD, which suggests to me there must be some underlying genetic predisposition to developing wet AMD in patients with GA. Next slide, please. So I'd like to summarize the key takeaways. GA lesion growth is progressive and irreversible. It's relentless for my patients with geographic atrophy. It's also important to know from these natural history studies that wet AMD can occur in GA patients, and its occurrence is considerably higher if wet AMD is present in the fellow eye, and this resonates with my own experience in clinical practice. And finally, many patients lose the ability to drive within 2 years, but we also see that many patients are becoming legally blind in their better-seeing eye within 2 years and that the vision-related quality of life is significantly impacted. And I generally try to see my patients with geographic atrophy about every 6 months, and every 6 months, I tell them that sadly, there's no treatment that I have to offer at the present time that can either slow or prevent this relentless disease. And so there is a significant unmet medical need in patients with geographic atrophy. Thank you.

Thank you, Dr. Holekamp, for underscoring the unmet medical need and the urgent need to develop treatments for this devastating disease. I want to apologize also. I had a bit of a technical glitch and dropped my connection for a few minutes. So now we'll turn on our attention to the role of complement C3 in this disease and hear from Apellis' Chief Innovation Officer, Dr. Lukas Scheibler, who is responsible overseeing our preclinical research and advancing candidates into clinical development to establish proof of concept with his group here at Apellis. So Dr. Scheibler, please.

Thank you, Federico, for the introduction. We have heard from the last 2 presenters about the pathology and the unmet needs in geographic atrophy specifically. I want to dive now into and explain to you why we think the inhibition of component C3 is actually the best approach to sort of stop the progression of this devastating disease. Let me start by describing the evidence -- next slide, please. Let me start by describing the evidence we have that actually complement -- and specifically complement C3 play a key role in the pathogenesis of GA and AMD. First, we know that complement C3 is deposited in the retina of patients with AMD, as shown in green on the right. These are complement depositions on slices of our retina. Second, we know that complement activation products such as, for example, C3A, are markedly elevated in the eye of patients with AMD. And lastly, we know of several indications in the genome encoding complement proteins that are linked to complement overactivation and are linked to the onset of AMD. Based on this evidence, we believe that geographic atrophy results from a failure to block the activity of C3 and clean up C3 deposition in the retina. Next slide. Now pegcetacoplan is a targeted C3 therapy. It consists, as shown on the right, of 2 cyclic peptides that have high affinity to C3 as well as to the aggravation product, C3B. This dual mode of action is very critical, as we will see later when we dive deeper into the complement cascade. The 2 cyclic peptides are linked through a water-soluble packaging. Next slide. Now let's dive deeper into the complement system and the role of C3. The complement system is a complex cascade involving more than 30 proteins. The system's key function are to recruit and control immune cells to the site of activation, remove damaged cells and tissue and kill pathogens. However, when complement is overactivated, like, for example, in GA or AMD, all of these mechanisms can attack and destroy healthy cells and tissues leading to diseases. There are 3 different activation pathways, as shown on the top of the complement system: the classical pathway, the lectin pathway and the alternative pathway. We believe that many of the leases we talk about in the complement field have involved multiple activation pathways, and it's just a single pathway which is activated. Often more, 2 or 3, are really involved. Upon activation of either pathway, C3 is broken into 2 fragments, C3A and C3B. Now C3B is a highly reactive protein that binds to nearby cell surfaces, and this is what we saw in the previous slide with the green deposition I showed you. If not deactivated by control proteins, surface deposited C3B can split, as shown further down, C5 into C5A and C5B, and this will ultimately lead to the membrane attack complex, which is inducing cell death. However, membrane attack complex, or MAC, is not the only complement product with a role in the pathogenesis of AMD. C3A and C5A help to recruit immune cells to the site of activation. And C3B, especially when bound to the surface of a cell at a high-enough concentration, will activate phagocytosis, a cell-removal process controlled by immune cells. By targeting C3, pegcetacoplan blocks all of these actions. First, it binds to C3 and blocks the formation of C3A and C3B. In addition, it also binds to C3B and therefore blocks the activation of C5 and ultimately the formation of MAC. Hence, you can think of pegcetacoplan as essentially a C5 inhibitor as well as a C3 inhibitor. So with this in mind, why do we think that pegcetacoplan will be successful when other complement inhibitors have failed? First, lampalizumab only inhibited the alternative pathway. It had no effect on the classical and the lectin pathway. And therefore, this failure of lampalizumab suggests that these additional pathways play a critical role in AMD as well. On the other hand, C5 inhibitors only prevent the formation of MAC and have no impact on the C3B deposition and the phagocytosis I explained. This may result in a treatment benefit, but as we have learned from our experience in PNH, C3B deposition on cells and the subsequent phagocytosis play a critical role in these diseases. And this is actually responsible for the superiority of a C3 inhibitor over a C5 inhibitor. We, therefore, believe that the unique mode of action of pegcetacoplan is very much differentiated from other approaches and will result in a more comprehensive control of the complement system by blocking the formation of MAC, blocking C3 deposition in the retina and inhibiting the recruitment of immune cells into the retina. C5 inhibitors, Factor D inhibitors cannot achieve this as monotherapies, but pegcetacoplan can. And with this, I would like to hand back to Federico. Thank you.

Thank you, Lukas. It's a great primer on our scientific rationale and the mechanism of pegcetacoplan. That's wonderful. So I want to make everybody aware, if your screen freezes as mine did a while back, try refreshing the screen, and that should solve the issue for you. So now we'll dive into positive and encouraging clinical results from our program, presented by Dr. Jeffrey Heier from Ophthalmic Consultants of Boston. Dr. Heier, please, when you're ready.

Thank you, Federico. Can you hear me well? Great. Thank you. So my name is Jeffrey Heier. I'm the Director of the Retina Service and Retinal Research at Ophthalmic Consultants of Boston. I have had the privilege and pleasure of working with Cedric and Federico and the scientists at Apellis dating back to that time that Federico talks about almost 10 years ago. And so I've seen this program as it has progressed. And in addition, I am the lead investigator on the DERBY study, so it's a privilege to present the results of FILLY to you this morning. Next slide, please. The FILLY study was a Phase II, single-masked, randomized study that enrolled 246 patients across 43 sites. Enrollment was very broad-based. This took essentially a real-world environment. It took patients with CNV in the fellow eye. It took patients who had both foveal and extrafoveal lesions. And it randomized patients 2:2 to 1:1 to pegcetacoplan 15 milligrams monthly, pegcetacoplan 15 milligrams every year per month and then sham monthly or every other month. The sham groups were pooled for the analyses. The -- as we said, the patients were geographic atrophy patients secondary to AMD, and the duration of the study was 18 months. Next slide, please. This is the time line and endpoints of the FILLY study. Patients were imaged at baseline 2, 6, 12 and 18 months. The treatment period was from 0 to 12 months, the observation period then from 12 to 18 months. The primary efficacy endpoint was the change in geographic atrophy lesion size from baseline at month 12. And the primary safety endpoints were the number and severity of local and systemic treatment-emergent adverse events. Next slide, please. As we all know now, pegcetacoplan met its primary endpoint with each arm showing a statistically significant reduction in GA lesion growth versus sham, and this was shown in a dose-dependent fashion. As you see, the pegcetacoplan every other month had a 20% benefit over sham, and the pegcetacoplan monthly had a 29% benefit over sham on reduction of GA lesion growth. Now it's always important when we look at Phase II studies to understand if the sham arm or the control arm behaved as we would expect. Next slide, please. And it was very reassuring to see here that the growth rate in the square root of the sham arm was 0.35, and this was virtually identical to the sham arm in the lampalizumab Phase III trials chroma and spectri, where the rate there was 0.342. So that gives us confidence that the differences from the treatment arm versus sham are, in fact, real. Next slide, please. This is the impact of pegcetacoplan on the absolute GA lesion growth. So what I showed you before was the square root analysis. And just to explain, the square root analysis is often used because it accounts for differences in baseline lesion growth. This is the absolute, untransformed lesion size here. And you can see the results in each of these were essentially identical, 20% and 29% difference in the square root transformation and 20% and 30% difference in the absolute, untransformed lesion size. So again, these both were comparable and what we expect to see. Next slide, please. So this was a post hoc analysis of lesion growth by 6 months periods. And you can see that there was an increasing effect size with the longer duration of treatment. There was a little difference but relatively smaller from 0 to 6 months. But from 6 to 12 months, again, there was a significant decrease in lesion growth that, again, was dose dependent. And again, seeing these differences help to serve as a nice internal control. We then see from 12 to 18 months, when the treatment was stopped, that we start to see this go back to the initial growth rate. Next slide, please. Several multi-variant post hoc and unitary post hoc analyses were done with the intent of looking for factors, looking for risk factors that would show faster growth rate or could potentially impact the study. And so things that were studied include baseline characteristics such as foveal versus extrafoveal lesion site. It looked at things like low luminants deficit. And what the intent was both to see when these were present in the FILLY study, if they behaved as was expected as compared to the literature. And they determined that, in fact, this was the case. And then the second point was to see if these factors had an impact on the overall results of the study. So in other words, did only the lesions that were extrafoveal show the -- an effect and the lesions that, say, were subfoveal did not? And in fact, it was reassuring to see that the benefits were noted across these different risk factors. So pegcetacoplan was consistent among the subgroups, and the risk factors that were identified behaved as we would expect from the literature but not -- were not the sole drivers of the outcomes, the positive outcomes of the FILLY study. Next slide, please. Another important analysis was looking at patients who had bilateral GA. So we know that patients with bilateral GA, the lesion growth tends to be relatively symmetric between the 2 eyes. And when you look at the graph on the left, we see the sham arm, and we see that the fellow eye and study eye had growth that was essentially identical. When we look in the middle at the pegcetacoplan every other month, we see there is a difference between the fellow eye and the study eye of roughly 10%. And when we look at the pegcetacoplan administered monthly, we see that the difference is 23%. So again, the fellow eye served as essentially another internal control, and we see this difference. Next slide, please. Safety is important when you're looking at Phase II studies, well, really looking at any study, but these help give us more information. And in this case, there was -- we did see increased exudations at 12 months. And you see here, it was 16% in the monthly arm, 6% in the every-other-month arm and 1% in the sham. We looked at these very carefully. The exudations were caused by the investigators, not by the reading center. And in several of these cases, the reading center did not detect choroidal neovascularization. How would this happen? Well, you can certainly see cystic changes that occur over areas of atrophy that can be mistaken for CNV, and you can have other subtle changes. In the cases where the reading center did agree that exudation had occurred, it was interesting that 0 cases of these were classical CNV. They tended to be a cult. These patients were treated with anti-VEGF therapy. And because they've been followed carefully, followed monthly, these were caught early, treated early, and patients tended to do very well with no clinically significant impact on vision. In terms of safety in other components of the study, most of what we saw was in line with other studies of intravitreal administered agents. In particular, there were 2 cases of endophthalmitis over roughly 1,500 injections, and this falls in line with what we have seen in other clinical trials. Next slide, please. So as clinicians, as investigators, most of us believe that while attacking geographic atrophy is important and limiting the growth of geographic atrophy is critical, and ability to impact disease earlier on would be extremely important. And looking at endpoints, there's a thing called cRORA, which is complete RPE and outer retinal atrophy, as seen in this image. Next slide, please. But iRORA is an earlier endpoint in the atrophy pathway, and it stands for incomplete RPE and outer retinal atrophy, and is essentially consistent with intermediate dry macular degeneration. So if we could impact disease state earlier, that would have even more of a potential benefit. Next slide, please. SriniVas Sadda and his group presented a late-breaking presentation at your retina last year. And they did a post hoc analysis of FILLY in patients who had received all 12 of their injections and had not developed CNV. And what you see here is at month 6, there was a significant lesser amount of patients who had progressed from iRORA to cRORA in the pegcetacoplan monthly versus the sham arm, almost half of 26% versus 50%. And at month 12, that benefit still remained, where only 50% of these patients had progressed to cRORA in the treated arm versus 82% in the sham arm, so again, showing that pegcetacoplan may be able to not only impact the advanced-stage geographic atrophy but also hopefully impact the earlier stages and hence stop vision loss even earlier on. Next slide, please. So key takeaways from the FILLY study. The FILLY study met/its primary endpoint of reduction in GA lesion group with benefit across geographic atrophy phenotypes. It demonstrated a dose response with sham, and the sham arm did progress at the expected rate, as we've seen in both natural history and other studies. There was an effect in the treated eye versus the contralateral eye in patients who had bilateral geographic atrophy, again, important because these eyes tend to grow at similar rates. It slowed progression from intermediate AMD to geographic atrophy, again, from iRORA to cRORA, suggesting the potential for earlier intervention in the course of GA and dry macular degeneration. And then finally, its safety is in line with other studies of intravitreally administered agents. Thank you very much. I appreciate having had the opportunity to share this data with you.

Thank you, Dr. Heier. Well, these studies set the stage for ongoing Phase II program for which we are really looking forward to the readout in the Q3 of this year. To review DERBY and OAKS, now we'll have Dr. Jeffrey Eisele, a new executive hire at Apellis, who serves as the lead for the Local Chronic Program and oversees all aspects of the GA program as well. So Dr. Eisele, please.

Great. Thank you, Federico. So my name is Jeffrey Eisele, and as Federico said, I'm leading the GA program. I'd just like to provide a short introduction of myself. I'm a statistician by training, and I consider myself a drug developer by profession. I've been working in the pharma industry for over 25 years and in ophthalmology drug development since 2006. And as Federico said, I joined Apellis in September of last year from Novartis, where I led the biostatistics teams in neuroscience and ophthalmology, and I worked on both the Lucentis and DoD programs. I joined Apellis specifically because of this program in geographic atrophy. And I'm here to ensure that we effectively manage the study generated in DERBY and OAKS and to ensure that we have robust analysis strategies that will allow us to best characterize the true effect of pegcetacoplan on GA. So now it's my pleasure to provide an overview of the DERBY and OAKS Phase III studies. There are 3 key points that I would like to emphasize throughout the presentation. First, DERBY and OAKS include a broad GA patient population reflective of real-world GA patients and consistent with the FILLY trial population and study design. Second, we aim to deliver a robust and well-characterized data set in GA in spite of COVID-19. And lastly, as mentioned already, we expect to deliver the top line results in Q3 of this year. So in DERBY and OAKS, we have 2 large, well-controlled Phase III studies to compare pegcetacoplan monthly and every other month intravitreal injections with sham injections. The primary endpoint of change from baseline in total area of GA lesion will be assessed at month 12. The study continues for a total of 24 months, and patients will remain randomized for the duration of the study. The study enrolled over 1,200 patients in 15 months in spite of challenges due to COVID. This speaks to the unmet need and I think, importantly, the patient desire for treatment. As mentioned, the studies include a broad patient population, and that patient population is also more representative than the population studied with Roche's lampalizumab and IVERIC's Zimura. Some key differences from those studies. Fellow eye CNV is allowed in DERBY and OAKS but was excluded from the lampalizumab and Zimura studies. Extrafoveal and foveal lesions are allowed in DERBY and OAKS, while the Zimura study excluded foveal lesions. And more advanced patients are allowed than in the lampalizumab studies. DERBY and OAKS allow baseline BCVA of 24 letters or better, while the lampalizumab studies allowed 49 letters or better. Next slide, please. As mentioned, the primary end point has changed from baseline in total area of GA lesion. The primary readout will be at month 12. The study will continue for a total of 24 months, and safety and efficacy will be assessed at the end of the study as well. As we saw in FILLY, the treatment effect increased from month 6 to 12. And we have now the opportunity in DERBY and OAKS and the potential to observe continued increases in treatment effect out to 24 months. We've included additional visual function and quality-of-life endpoints in DERBY and OAKS that were not included in FILLY. Similar to FILLY, a mixed model with repeated measures will be used for the primary analysis of change from baseline in total area of GA lesion. We have a strategy to manage missed injections specifically due to COVID that is under review with the FDA and EMA. Now why is this actually important? So a patient missing an injection because the physician's office is closed due to pandemic is different than a patient missing an injection because of, say, an adverse event, and they should and must be treated differently. Next slide, please. Here, we summarize some of the key changes made to the DERBY and OAKS studies. First, the primary endpoint does not take square root transformation but uses absolute change in lesion size, and this follows a recommendation from the FDA. There are different approaches to how we could model lesion growth while accounting for differences in baseline lesion size. And as Jeff mentioned in the previous presentation, this is key because larger lesions tend to grow more than smaller lesions. Both approaches are acceptable, and as we saw in FILLY, the square root transformed and the nontransformed results were consistent. DERBY and OAKS are double-masked, which should further minimize the potential for bias. But what does this mean specifically? In FILLY, the patients and study personnel responsible for efficacy assessments were masked to treatment, while the physicians assessing adverse events were not. In DERBY and OAKS, study personnel responsible for both efficacy and safety assessments are masked to treatment, and this should help to minimize any potential for bias in reporting of adverse events. We've also included more frequent assessments, which will provide us with a more complete assessment of lesion growth. Now as a statistician, I want as frequent assessments as possible, as feasible because it will improve my ability to model this longitudinal data. Now imagine if a patient went off-treatment or study at month 11, for whatever reason. I'm in a much better situation to understand the treatment effect at month 12 if my last assessment of lesion growth was at month 10 than if it was at month 6 or baseline. So this is a very important improvement in our ability to actually effectively assess and estimate the true effective treatment. And last, patients experiencing exudations will continue on study and continue to receive treatment, which we expect to reduce the number of study treatment discontinuations. And one additional point here. In the FILLY study, we observed that patients with fellow eye CNV were more likely to have exudations. So for DERBY and OAKS, fellow eye CNV was a stratification factor, which should reduce the potential for chance findings due to baseline -- due to any baseline imbalance. Next slide, please. So there are a number of reasons why we believe DERBY and OAKS will be successful, and I think Jeff outlined in his presentation a number of them. But first and foremost is also that we're basically replicating a successful study that had robust results. Multiple sensitivity analyses were carried out in the primary endpoint, and all confirmed the results of the primary analysis. And extremely important for me, I believe, is that if we assume that all the missing data on the pegcetacoplan monthly arm would be imputed as if it followed the trends of the sham arm, we still retained approximately 80% of the primary analysis treatment effect. We've also learned from FILLY and implemented improvements in the study design that we expect to result in more patients staying on treatment and further reduce the potential for bias in the diagnosis of excitations. And lastly, the studies are already well powered for the primary endpoint in spite of the impact of missed injections due to COVID. And as mentioned before, we have a strategy under review with regulators to mitigate the impact of missed injection, specifically due to COVID. So for all of these reasons, we believe DERBY and OAKS will be successful, and I'm truly excited and anticipate the results that we will have in Q3 of this year. Next slide, please. In summary, the DERBY and OAKS studies are 2 large, well-controlled Phase III studies. They include a broad GA patient population, reflective of real-world GA patients, including patients with fellow eye neovascular AMD and both extrafoveal and foveal GA lesions. The study design and population are consistent with FILLY, which had positive, robust results. And finally, the top line results are expected in Q3 of this year. Thank you. And back to you, Federico.

Thank you, Jeff. Well, as many here on the line, we're looking forward to the results in Q3. So throughout the presentations today, there's clear, there is a high medical need and an urgent need for treatment. But there is also a need for better and early diagnosis to support patient treatment. So now I would like to welcome Dr. Pearse Keane from Moorfields Eye Hospital and UCL Institute of Ophthalmology. Dr. Keane will talk about how artificial intelligence is being developed to support early detection of GA at the general ophthalmology and the optometrist practices. Dr. Keane, please, when you're ready.

Thank you, Federico, for that nice introduction. So first of all, I'd like to thank Apellis for asking me to give this presentation. And as Federico said, I'm going to be speaking a bit today about the early detection of geographic atrophy using artificial intelligence. Now just a little bit more about my background. So I'm a consultant ophthalmologist at Moorfields Eye Hospital in London. And at Moorfields, I specialize in the treatment of retinal diseases, and I have a particular interest and focus on age-related macular degeneration. Now the other hat that I wear is that I'm also an Associate Professor at the Institute of Ophthalmology at the University College London at UCL. And at UCL, I lead a multidisciplinary research group, which aims to enable the development and application of artificial intelligence in health care using ophthalmology as an exemplar for other medical specialties. Now my -- when I talk about artificial intelligence, for the most part, really, I'm referring to a subtype of AI, which we call deep learning, and deep learning is that big advance in the last decade that has lead to profound changes in the tech industry and is starting to profoundly influence health care. And in 2015, Scientific American listed deep learning as one of the world-changing technologies of the year. Now my background in AI-enabled healthcare really began in about 2015, 2016 when I initiated the formal research collaboration agreement between Moorfields Eye Hospital where I'm a consultant and with the artificial intelligence company, DeepMind. Now I'm sure that most of the people on the call will be familiar, but DeepMind were a company that were founded in 2012, and 2 of the 3 cofounders are alumni of University College London, which, as you've just heard, is affiliated with Moorfields. And they were acquired by Google in 2014 reputedly for about GBP 400 million. Now we've been working with them for a number of years, and we published the first results of this work in Nature Medicine in August 2018. And it was really exciting for me that this was published there, but also given the excitement around artificial intelligence that have got global media attention. And just to say where we're -- what we're working on now with Google Health is seeing how we can translate, what we have done in this paper and deployed at scale in the real world. So this is all the background that meant that really, in the past 2 or 3 years, I think it was at EURETINA in 2018 that I was first introduced to Lukas and some of the team at Apellis, that we initiated this collaboration. And there's a number of reasons why I feel passionate about developing artificial intelligence for geographic atrophy. One of the reasons comes from this paper that came out in the British Journal of Ophthalmology earlier on this year, and this is a paper from Germany. And of course, there's many such papers, and you've heard from many of the other speakers about how important a disease age-related macular degeneration is. But I thought it would be nice just to point out one of the lines from the abstract of this paper that particularly caught my attention. And that line was this. "And they estimate that approximately 25% of people aged about 60 years and over in Europe have the early or intermediate forms of AMD." And of course, as you've heard earlier, this is not just a disease that is limited to Europe. This is a global disease. And so if I think about my own parents who are in their 60s, it just really brings home to me that 60 is not old anymore, and we have a global population that is continuing to age. So this is a problem that's huge now and is only going to get bigger in the coming years. Now the second reason why I feel passionate about this specific research relates to the one fact that I would hypothesize that every single patient with AMD knows about the condition. And what I would urge you is if you have a friend or family member who has AMD, you can test what I'm about to say. But what I would say is the one fact that everyone knows is this. There is no treatment for dry AMD. Every patient will tell you this and kind of shake their head in weary resignation as regards that. And so for me, the potential to be involved, even a small way with potentially the first treatment for this condition is particularly exciting. Now coupled with those sort of 2 reasons, I think the other reason why I feel particularly interested in this artificial intelligence work is because there's a kind of -- the stars are aligned to actually maximize the potential for patient benefit with this. And so one of the reasons that I make that statement is because we've seen many profound changes recently in the way that ophthalmology is being implemented. And so in particular, something I thought you might find interesting was an announcement from the optometry chain, Spec Savers, and this was a few years ago now back in 2016. And Spec Savers are the largest optometry chain in the U.K. They're also based in the Nordic regions and Australasia and a number of other countries. But they announced that they were going to roll out OCT scans, and I'll talk a little bit more about what that means, in all of their practices. And given that they have over 700 practices in the United Kingdom alone and do more than 9 million eye examinations per year in the United Kingdom at home, clearly, this is going to have profound effects on the way that we administer eye care in the coming years. So this has started to gain momentum. And so just prior to the pandemic, there were some major advertising campaigns, both in the U.K. and abroad, advertising this increased use of OCT. It's important to emphasize that this is not just Spec Savers, but this is all the major optometry chains as well as independent optometrists, both in the U.K. and in other regions around the world. Now the reason why I say this is because -- that I'm so excited about this is because this really gives us the chance to use artificial intelligence to essentially bring world-class expertise from the hospital into the community and potentially into the home of patients in the longer term. Now the aim with this work that we've been working on over the last year or 2 is essentially to take these OCT scans, and these are -- it's kind of like ultrasound, but it measures light waves instead of sound waves. And in doing so, it gives us very, very high-resolution, 3-dimensional scans of the retina and has become the dominant imaging modality in all of ophthalmology and, as I say, is beginning to be used in the community. And so could we develop specifically a deep learning system that could both detect and quantify geographic atrophy on OCT volume scans and use that tool to facilitate more precise treatment of patients with geographic atrophy and possibly even earlier detection of those -- of the condition in these patients. Now we've been working on this for a while. And the foundation of this work was the data set from the FILLY trial. And we took nearly 1,000 OCT volume scans from this trial from about 200 patients in this trial. And we used that to -- in something called supervised learning to develop this GA segmentation and quantification tool. Now this is work that was done at Moorfield's Reading Center by expert graders. And this is the labeling platform that we use at the reading center for this. And one of the things that I think it's important to highlight, and this is different from some of the other early AI work in this space, is that we've incorporated some of the latest definitions from an international consortium of experts in AMD and retinal imaging. And by doing this, we can get a much more granular quantification of geographic atrophy on OCT scans, so potentially more accurate monitoring of this progression but also picking up those early biomarkers of this disease. Now the other thing that I'm really excited about, about this work is that we've actually done something called external validation. So we haven't just trained and tested on the data from the FILLY study or other data from Apellis. We've actually collected a large external data set from Moorfields. Moorfields has about 30 sites within the London region and southeast of the U.K., and we collected geographic atrophy patients, over 100 patients from -- scans from over 100 patients, and we use that as an independent test set for the system that we've developed. Now what the system essentially does is take these OCT scans, and it automatically segments the individual features of GA such as photoreceptor loss, RPE loss, choroidal hypertransmission and can also combine them as one into this figure of GA. And then ultimately, it can do a reconstruction. So it can provide an area measurement in millimeter squared, which, as you've heard earlier, is one of the main ways that we assess the progression in our clinical trials. Now I'm not going to go into a lot of detail about the results, but this table signifies the main findings. And in particular, if you look on the right, we look -- in AI, we're most interested on the external -- the results and external validation sets. And one of the key numbers here is that we have a dice similarity coefficient, a dice score, which is the main way that we look at a lot of computer -- the efficacy of lot of computer vision algorithms. And we get 0.96, when you compare the model that we've created to consensus of human experts in terms of delineating GA on an OCT scan. And 0.96 is an almost perfect score. Now by comparison then, when we compare individual human experts between themselves, we find that the die score is only 0.8, suggesting, at least in this preliminary work, that we may have some almost superhuman performance with this system. Now where we are with this work is that we have prepared the first manuscript from this work, and it's currently under peer review at a high-impact clinical journal. So we hope that this is going to be accepted for final publication at some point in the coming months. So what are the next steps? Just to begin to finish up. Well, we already started working on the next phase. And so one of the things we want to do is do better structure function modeling using artificial intelligence. And so can we take an OCT scan? And can we predict visual acuity or other indicators of visual function in a much more powerful way using this technology? And perhaps, more excitingly, can we do some sort of future prediction? Can we identify patients that are fast progressors? Can we identify patients that might progress by 20% or more in the next year and incorporate not just OCT data but multimodal data? So other demographic information, genetic information, et cetera, so we can predict those patients that are aggressive or perhaps more benign versions of the disease. So to finish up then, what I'm really excited about is not just publishing a paper but seeing how we can implement this in the real world for patient benefit, how we can go from code to clinic. Now in the first instance, I think that's going to be in the hospital setting. So we can enable retinal specialists to optimize the treatment of patients with geographic atrophy. But I think what's really exciting in the medium-term is how, as I've said, we can bring this into the community so we can get much earlier detection of this disease because one of the things that other speakers have highlighted is that this is a degenerative disease. And once the retina is dead, currently, there's no way that we can regenerate that. So early detection and slowing of progression is going to be key to this. And then lastly, I always provide my e-mail address whenever I give a talk. If anyone would like to get in touch, I'm always happy to answer e-mails. Thank you very much.

Thank you, Dr. Keane. Well, it was amazing enough to see the rocket speed advancement of the imaging technology in the retina for the past 15 years. And we thought that there was going to be a limit to it. But now seeing how artificial intelligence and deep learning can use all that information for early detection, for the taking progression and to help with treatment decisions in this devastating disease is just astonishing. So there is no limits. I want to thank you. I want to thank all the speakers for today's presentations. It was a great session. And now I'll turn it back to Dr. Francois for some concluding remarks, and then we'll turn it back to the operator for a question-and-answer session.

Thank you, Federico. Well, I hope you all enjoyed this as much as I did. It's really a privilege to have such extraordinary key opinion leaders, not just present, but be by our side as we try to develop these therapies. So going back to the key misconceptions that we said we would address today. Starting with the first one, and then after we go through that, we'll have, of course, Q&A sessions in which you will have the time to ask any questions you may have. So starting with the first one. Other complement inhibitors have failed. Why should pegcetacoplan work? And in summary, especially from the presentation that Dr. Scheibler gave you, we believe that geographic atrophy is the consequence of a failure to clean up C3 deposition in the retina where retinal cells and, by extension, neuro cells have to use the same cellular resources to clean up C3 product from the surface as what they need for their homeostatic functions in the case of the retina for the visual cycle. And at some point, especially as we grow older, and these cells have to make a choice, they can start neglecting that cleanup of that C3 product. And when that C3 product then becomes visible to mononuclear cells, they start removing these cells in an inappropriate fashion, leading to geographic atrophy and other neurodegenerative conditions. As a consequence, single-pathway inhibitors, we believe, are insufficient to correct this failure. It is important that you give these cells an opportunity to flip the switch so that they have enough capability to clean up that product without neglecting their homeostatic functions. And if you go with single-pathway inhibitors, there may be a slight correction but we believe not enough to flip that switch. In pegcetacoplan, which is a targeted C3 therapy, but which really has a broad effect on the complement cascade, which inhibits the convertase activity in a broad fashion, both the C3 and the C5 convertases, we believe, is the only molecule that can flip that switch and slow down the disease. And as mentioned earlier by Dr. Eisele, I'm personally especially excited by the second year of dosing that we will have, where I believe it may be possible that we see a further increased effect compared to what we saw from month 6 to month 12 in the FILLY study. The second misconception is that pegcetacoplan has no impact on visual acuity. It will take years to notice the difference. Why would people take this drug? Again, I hope you will take home from the presentations today that visual acuity is a tricky concept because really, what visual acuity refers to is when you're at the ophthalmologist and you are being asked to read a chart on the wall with the letters, right, that is very different from crossing a street in New York, where you need broad peripheral vision and an awareness of what is going on other than just at these letters. And in geographic atrophy, we have this phenomenon that we talked about earlier called foveal sparing, where the central point of your vision is often maintained until the very end when you lose that and where, all of a sudden, you have a precipitous loss in visual acuity. So when you look at the average loss of visual acuity in geographic atrophy patients, which generally accounts for approximately 1 line of vision loss per year, this is driven by the outliers that precipitously lose their vision or their visual acuity, I should say, during that 1 year. So important to bear in mind. What is key here, and I cannot overemphasize this because if there's one thing that I find frustrating is that the misconception that geographic atrophy is a mild disease compared to wet AMD, and maybe you should wait, and maybe you should -- this is a relentless, progressive disease that severely impairs patients' visual function and quality of life. It is a leading cause of blindness, and GA has no approved therapies and remains the most significant unmet need in the retina. And again, just one minor point that Dr. Eisele brought up as well. The rate of enrollment that we had in the Phase III clinical trial, in spite of COVID, in spite of the early issue at the start of the study that we had with our formulation, the rate of enrollment gives us a sense of how strong the need for a therapy in this disease is. Next slide, please. The third misconception or, I should say, point of that. How can we trust the FILLY data? Well, as you've seen in the past hour, the FILLY study met its primary endpoint, demonstrated a dose response, and the sham group progressed at the expected rate. The dropouts were in line with other GA trials. 242 out of 246 patients were included in the primary analysis population, which accounted for missing data with a well-established, mixed-effect model for repeated measures. All sensitivity analyses, all of them, genetic or phenotypic, confirmed the efficacy profile. Post hoc analysis also shows that lesion growth slowed in treated eyes versus contralateral eyes in patients with bilateral GA and that pegcetacoplan slowed the progression from intermediate AMD to GA, which may account for that increased effect over time, right? Imagine you have the area of dead retina, and then you have the zone outside of that, where the cells may not be dead yet, but they are so severely affected that they may not be rescuable anymore. And what we showed at that late-breaking abstract at EURETINA last year is that we have an important and post hoc statistically significant effect on slowing down the progression as you go outside of the lesions. And again, we'll see what happens after 3 years of dosing. And then the fourth misconception is that FILLY had safety issues. Pegcetacoplan causes wet AMD and had 2 cases of infections and endophthalmitis. Well, our response to that is that the FILLY trial had no safety findings that limited DERBY and OAKS. We are studying the exact same patient population, which is a broad population, arguably the broadest of any of the confirmatory programs that are out there. Number two, exudations are an expected occurrence in geographic atrophy. This is not an unusual finding. What was unusual was the imbalance between the 3 groups, which may have been more pronounced because of some of the things that we had in the Phase II clinical trials such as the single-masked. What's key here is that these were, without exception, occult, so-called occult lesion, with not a single classical case of CNV. And there was no medically significant impact on vision as a consequence. The FILLY trial was single-masked, which may have led to greater reporting of these exudations, as I mentioned a minute ago, in the subject randomized to pegcetacoplan compared to the subjects randomized to the sham. DERBY and OAKS, the double-masked, we believe, will continue to correct for that. And then the 2 cases of infectious endophthalmitis, as Dr. Heier already alluded to, are very much in line with what you can expect in clinical trials with intravitreal injections. So as we go and move forward towards the Phase III clinical readout in the third quarter, such a seminal event for our company, for patients, for the field, I think there are 2 important things that I personally find we should all take to heart. Number one, this is a very serious disease. It is the one point in discussions that I have, whether it's with investors or layman people, that can annoy me sometimes is that, that concept of visual acuity especially, et cetera, creates a misconception that this is not a problem for patients until you talk with patients, until you talk with physicians. This is a devastating disease. And someone needs to do something, whether it's us or somebody else. The second point, and one that I'm personally particularly proud of, is we cannot control biology. I can, of course, not promise to you today that complement control at the level of C3 is going to work in the summer. But what I can promise you is that as a company, we have, I think, in an extraordinary way, to make sure that if this trial fails, it will be for biology, and not for mistakes made along the way, in the design of this trial and the way we analyze it and the way in which we control it, the way everything was run. So the answer you will get this summer will be the answer that it may be, and of course, we believe that it will be positive. But that is out of our hands. And with that, I look forward to taking you forward to the Q&A session.

Thank you, Cedric. I'll turn it back to the operator for Q&A.

[Operator Instructions] Your first question is from the line of Jonathan Miller.

Thank you so much for a really great set of KOL presentations and a very clear message about what we should expect from DERBY and OAKS. I guess one thing that I want to focus on now is if visual acuity isn't the right clinical outcome here, then what is? I mean, how do we quantify clinical impact of the swelling of lesion growth that you're hoping to show in DERBY and OAKS? And the corollary to that, of course, is what -- in terms of clinical output -- outcome, what would need to be shown to drive interest from patients in order to get them to come in and get intravitreal injections on a monthly basis?

Thank you. That is a fantastic question. I would like to start, if you don't mind, Jeff Heier, to tell us what the patients would expect from changes in vision.

Sure. I think that is a great question and one that comes up quite frequently. So we're heavily involved in geographic atrophy studies. We're typically one of the top enrollers in most of the studies. And unlike what you would expect, the dropout rate in these studies is remarkably low, despite the fact that patients have a disease that moves very slowly. And is it like wet AMD or DME where you get immediate feedback and immediate gratification? Patients know what the natural history of this disease is. They know that this disease is going to progress, and it's pretty much going to progress with absolute certainty. So patients understand that, and they're highly committed to the therapy. And my experience is not only are patients unlikely to drop out of studies, but the large majority of patients, once they've completed a clinical trial for geographic atrophy, their biggest question is, what can we do now? Is there another study I would qualify for us? Is there something else we can do? I don't want to lose any more vision. And that geographic atrophy is, in essence, vision loss. Even if it's not central vision, it is vision loss, and keeping that lesion small has significant impact.

Okay. Thank you so much, Dr. Heier. And I want to move to Dr. Holekamp, which you've discussed about this in your presentation, if you can give more details to answer the question.

Well, I'd like to go back to the question about measuring visual acuity or having other end points. And the FDA really understood when ophthalmologists went to them with studies on geographic atrophy and said, "Reading an eye chart isn't really a good indicator of the visual performance of these patients." So I'm really glad to see the vision-related quality of life tools being used in DERBY and OAKS. But I'm also happy to see that there's going to be 24 months of follow-up because I just showed data today that it's a relentlessly progressive disease that will eventually affect visual acuity. But the patients understand that they're losing vision in real life even if they can pick out a letter on an eye chart. So I think it's well established we're looking at lesion size. I think it's well established that the lesions are going to progressively, relentlessly grow and that it does affect vision. And I'll just close by saying one other piece of information. I was the lead officer on the Proxima A/Proxima B clinical trial papers. It was a natural history study, and the number one thing we learned from those studies is it's not a slow disease. It's pretty quick, maybe not like wet AMD. But as Cedric said earlier, these patients are losing 1 line of visual acuity a year. So we have to kind of change the way we think about geographic atrophy and its impact on vision.

Thank you so much. Anyone else want to add a brief comment before we move to the next question? Caroline, please, Dr. Baumal?

Thank you. I also think that we're going to be looking at a lot of imaging end points, and patients understand these end points. When we show them a picture, whether the lesions have grown, whether it's using autofluorescence or color imaging, these are the sort of things that we can use AI with. They understand and they can see on the pictures with what vision they have if the treatment is working or not. And also, I think patients who have lost vision in one eye and maybe have extrafoveal lesions in their remaining eye are extremely motivated to save vision in that eye.

Thank you very much. Dr. Keane, I want to give you the opportunity just, one, to see the European view just briefly, and then we'll move on to your next question.

I would just echo Dr. Heier's point at the start. These patients are the most motivated patients of all of the patients that I see with retinal disease. My stereotype, if you like, of the patient with AMD in my clinic is a patient in their 80s who is extremely engaged, extremely knowledgeable, who uses e-readers like their Amazon Kindles, listens to opera with their voice assistants and just wants to really look for this -- look for any possible hope for a disease which has currently no treatment.

Your next question is from the line of Steve Seedhouse.

This is Timur Ivannikov on for Steve Seedhouse. And so our question is in terms of the mechanism, can you elaborate on pegcetacoplan's inhibition of C3 and the degradation products and how this inhibition profile versus the various complement fragments may distinguish pegcetacoplan from other C3 inhibitors? And why could that matter in terms of efficacy in GA?

Thank you. Thank you very much for joining us and great question. Dr. Scheibler, would you like to start with answering Timur?

Sure. I mean, the slide we showed, first of all, is obviously a very simplified mechanism of action, a simplified summary of the complement cascade. As I said, there are more than 30 proteins involved in this entire cascade. And what we specifically left out, for instance, is the amplification loop of the alternative pathway, which we didn't discuss there. But -- so if you have the fact that D inhibitor like lampalizumab, yes, you block the amplification loop and the formation of C3b through the amplification loop. But what you do not inhibit at all is the formation of C3b through the classical and the lectin pathway. In these pathways, and we believe are constantly activated, and even if you block the amplification loop, you will still have the formation of C3b. And over the time, you still have this deposition. And that brings us to what Dr. Francois has said in his summary, that brings us towards this takeover and its status. When you have C3b deposition, it's not a matter of -- it's a matter of when do you have enough to trigger phagocytosis. So there is a certain threshold. Nothing is going to happen until you reach this certain threshold and then the fire starts. And then -- so it may just take a little bit longer. If you block the alternative pathway, until you reach this threshold, it could ultimately -- you're going to reach the threshold, and you're going to have this takeover, which is going to start. In addition, when you then think about C5 on the bottom end, when you think about MAC formation, well, in a bacterial infection, MAC is going to destroy the bacteria right away. Now in our cells of the body, we have mechanisms to stop that from happening. We have preventative mechanism. There are proteins involved, CD59, for instance. And so these proteins deactivate the formation of MAC. And therefore, you don't have immediate cell death. But what you have is sort of you have this constant depolarization of a cell through the hole, which is formed through MAC. Now again, a C5 inhibitor can maybe block them. But when you take all of this together, you have to understand, in order to get rid of C3b, in order to fight against MAC, in order to control all of this, this all takes energy. The cell needs energy. And what we think of what AMD is, AMD is triggered by the hypoxic state. So there's not enough oxygen and as a consequence, not enough energy in this tissue. So if you just block one area of it, then you still have this hypoxic state. You still have everything going on. But what you really do with pegcetacoplan is you attach this energy deficiency, which is ultimately leading to the C3 deposition, which is not removed from the cell, which is leading to MAC formation, which is depolarizing the cells. You have to have all of this from the -- different angles simultaneously. But with this, you give the cells much more room to recover and to be alive and actually focus on what their real focus is, especially in the ERP and the photoreceptor, which drive the visual cycle, which, by its own, is a process which requires a lot of energy, especially at night. I hope this answers your question Freddie?

Yes. Thank you very much, Lukas. Cedric, do you want to add anything?

Just very briefly because we have a lot of questions. But it's -- I think the question also kind of referred to the comparators that are being made between the anti-C3 antibodies that are in development by some of our competitors, much earlier, of course, but interesting. We believe that the key differentiation there is the fact that pegcetacoplan binds to the C3b unit on C3 and by doing that, like Lukas says, controls the convertase activity. Complement is an enzymatic cascade that is out of control. And to bring it under control, we have, by binding to C3, basically an enzymatic approach, a buffer's approach to dampening [ total ] wreckage. If you do not have an effect on C3b, which specifically the antibodies that are in development in GA do not do, we actually believe that, that is a problem. So that's all I wanted to add. Federico, back to you.

Yes. Thank you, Cedric. And we have many questions in the queue, so we'll make them shorter and so -- try to get to everybody. So our next question, please?

Your next question is from the line of Anupam Rama.

Thanks for a productive session here. Can you expand on the strategies you have in place for the missing injections due to COVID? And how we should be thinking about accounting for the missing injections in the stats plan or any potential impact on the efficacy measures?

Thank you, Anupam. Thank you for joining today. Jeff Eisele, can I have you take a stab at that?

Sure. Sure. So we actually have a strategy that we proposed to both the FDA and EMA. And I think at this point, it's best for us to wait until we actually have an alignment with them on the approach that we should take going forward. And once we have that, I think we can make this more widely known.

Your next question is from the line of Derek Archila.

Congrats on the nice presentation today. Just 2 quick questions, one kind of a follow-up to a prior question. So on the biology, I just wanted to know, we haven't seen this really with systemic indications where complement is used. But is there any reason to think there might be a workaround in the complement system for GA? And is this kind of basically the theoretical risk that we may not see great efficacy here? And then the second question would be commercially, just maybe the KOLs' thoughts on patients requiring both anti-VEGF and pegcetacoplan therapy in the future for GA and the wet AMD just because it seems like these patients basically are prone to getting it. So how should we think about the treatment burden of a lot of intravitreal injections?

Yes. Now I take the first part. Yes, of course, there are many -- all these biological ways to go around the pathway for activation. But I think as Dr. Scheibler and Dr. Francois pointed out, by blocking in C3, you are targeting the complement pathway at the heart, in the middle of the cascade. And it's actually the only way to prevent all activation pathways and all effective pathways. So that should be minimized with [indiscernible] possible by inhibiting C3. Now Jeff Heier, if I can start with you again to discuss the treatment of GA and wet AMD at the same time in these patients, if you don't mind addressing that.

Sure. So I think one of the questions we always get is the concern about developing CNV in these patients. And one of the things, just even more basic, is we have a treatment for CNV right now for neovascularization, and that treatment is really quite effective. And I know that if patients have a chance of developing that and I catch it early, I can treat that and, in most cases, control it and minimize vision loss, if not stop vision loss, at least over a period of years. For geographic atrophy, I can't do that, none of us can. And we know that progression is going to occur without treatment, no matter what we do. So patients are very motivated to stop each of these. And I think while certainly, it makes the office visit a little bit more challenging because you have a delay in between treatments if they're on the same day, patients are still motivated to do whatever they can to preserve their vision. And in the studies where they've had both and they have to get treated for both, that's been well tolerated. So I think patients will tolerate it. I think they're motivated to do it. And understanding if the studies continue to show that it benefits them, I think it will be something that would be well adapted.

Thank you, Jeff. And Dr. Baumal, on your presentation, you touched on the [indiscernible] systems of both. So if you want to add anything to that.

Sure. Sure. I think geographic atrophy is actually the end result of multiple diseases and -- which really opens up horizons for geographic atrophy, not just for wet AMD but other types of macular degenerations leading to that. Intravitreal injections, for us as retina specialists, it's a very acceptable type of treatment. And while patients may initially not like that, there is a benefit of giving a medicine directly to the source, directly to the place where it needs to go in the eye, and they have less systemic effects from these medications. And most of us who do a lot of intravitreal injection, we have it down to an art. It's efficient, even more so during COVID times. I mean, my patients come in, they get treated, and they leave very quickly. So I think that it would not be difficult for us to incorporate another agent into our system of treatment.

Thank you very much. We only have 4 minutes left. I will try to move on and answer more questions. Thank you so much, Dr. Baumal. Next question, please.

Your next question is from the line of Justin Kim.

Let me echo, what great presentations there were this morning. Maybe one for our retinal specialist on the call. We've heard a fair bit about bilateral GA and the progression of disease across both eyes. When you think about clinical trial development in GA, can you comment as to the importance of comparing an active treatment against the fellow eye and how much insight can be drawn from these types of results?

Yes. Dr. Holekamp, would you mind answering the question?

Yes. So great question, and I can tell you that studying the fellow eye does provide kind of an internal control for the treated eye. And this disease is bilateral. I can tell you, we've looked at large populations of patients with the bilateral GA, and both eyes often look like 2 peas in a pod, and that the progression rates are very similar across both eyes. I think one of the confounders potentially is that you can never rule out a treatment effect in the contralateral eye when you're treating the first eye. So these analyses do have their limitations. But what I think we're looking at is sensitivity analyses where in a post hoc manner, we look at the data in many different ways. And if all arrows are pointing in the same direction, I think that they're valid. But I can expect that with patients with bilateral GA, we're going to see the same analysis you saw today, which is comparing the treated eye to the fellow eye, again, understanding some caveats.

Thank you very much. Jeff?

Yes. I want to say, I agree with Nancy completely. And looking at the dose-dependent response from the FILLY study gives even more credibility to that concept.

Thank you so much, Jeff. Let's move on to the next question. And we may be able to stay 2, 3 more minutes due to others -- a couple of more questions. So let's go through the next question, please.

Your next question is from the line of Brian Cheng.

I have one quick question for Jeff on the Phase III design. I think today you said that there is stratification based on fellow eye CNV. Is there also a stratification based on lesion size or the proportion of multifocal GA at entry? Any color on that would be great.

Thank you very much. Jeff?

Freddie, it might be easier for you to answer as the study designer, yes.

Yes. So we do have the stratification that you mentioned. We also have 2 groups for location of lesion and also stratifying for lesion size. So there are many stratifications, one at the randomization steps and other for analysis. But always, we guarantee that the study is well balanced. So I think briefly, that answers the question.

Your last question is from the line of Phil Nadeau.

Just one follow-up and then one question for Dr. Heier. First, the follow-up, Cedric, in the past -- this is on missed visits and missed data points. Cedric, in the past, you've suggested that the missed visits in DERBY and OAKS has been consistent with FILLY. Curious to get an update on that. Is that still the case? Or now that COVID's dragged on, have there been somewhat more missed visits? And then second, the question for Dr. Heier is, Dr. Heier, you addressed new onset exudations and the risk there. Curious to get your thoughts on the endophthalmitis. What level of imbalance in the Phase III would concern you?

Thank you so much. So I'll take the first one. The short answer to that question is, yes. Sorry, Federico. So thank you for that question. So the short answer is yes. We were very fortunate. There was a bit of the [ hiccup ] moment in April last year, as you may recall, where patients were worried. There were missed injections at the time. Then a couple of patients kind of checked out and just stayed there. But those that came back into the study got used to these visits again. And the second wave, the third wave have not really significantly impacted the numbers in the DERBY and OAKS study. So we feel very good. Even without any concessions on how the data gets analyzed, we feel very good about the quality of the study so far. And then Jeff, I will hand it over to you for the second piece.

Sure.

So can I make one comment?

Jeff Eisele, you want to add something?

Yes. I just wanted to add quickly, and this would get to an earlier question also, the studies are well powered, regardless of how we analyze or account for the missed injections due to COVID. So we're very confident regardless of how we go about with the imputation of that data. They're well-powered studies.

And on sort of the question on endophthalmitis, obviously, an important question, as any safety issues are. It's hard to give an absolute number right now. I'd want to look at it as compared to the rest of the data, look at it in the different arms, look at it across, see if there was a dose-dependent response. But so far, we've not really seen a signal to suggest that's likely to be a problem. But it will be studied extremely carefully, as well all the safety adverse events.

Thank you so much, Dr. Heier. And I believe we don't have any more questions at this point? No? All right.

Your next -- I'm sorry, sir. Your next question is from the line of Matt Luchini.

Thank you for the nice presentation. For the physicians on the panel, particularly those that don't have really any experience with pegcetacoplan, would love to just get your sense on how, when the drug becomes available, you plan to roll it out in your patient population. Are you anticipating sort of broad use among patients you've already identified? Do you want to get a little bit of firsthand experience in a few patients before trying it out more broadly? I'd just like to understand how you're thinking about that. And then quickly for the company, can you just remind us quickly your -- where you stand from manufacturing, manufacturing capacity, specifically as it relates to GA?

Yes. I would like to start with Dr. Baumal on that question and move to Dr. Keane really quick for the European perspective. So Dr. Baumal, please.

Sure. So since there's no other treatment for this disease, I think that the rollout of this will be probably -- put it -- have quite a few patients in my practice that have geographic atrophy, and I think that would be eager to have this. We always start using a drug in a cautious manner but use it as a way that it was indicated in the study. And I always use clinical studies to guide my management of patients, especially if something is new. And I think that if the study pulls the way that things have been going so far, that the biggest concern for patients will be safety. And if the safety is favorable, then we will adopt this widely. The one thing with geographic atrophy, I think, that will help us with our patients is education. We all saw this with wet AMD and anti-VEGFs. Those patients who were there before anti-VEGFs, they only came in every 6 months because there was no treatment. So we would just monitor them. And then when anti-VEGFs became available, education, optometrists, family members and then people just started coming. And I think the same thing will happen for geographic atrophy. There will be a treatment available. And all of a sudden, everyone is going to find out and people will just start coming.

Thank you. Dr. Keane, briefly, the European perspective.

Well, I would just say in the U.K., we have a National Health Service. And because there's no treatments for geographic atrophy, a lot of these patients actually get diagnosed in the hospital, followed up for a period of time and then followed up in the longer term in the community. I think as Caroline says that when they hear about a treatment, they'll talk to us. But what I would love to do is to be able to take some of the AI tools that we're developing, run them on our systems and identify all the patients who we've previously diagnosed with geographic atrophy who are still alive and contact them and offer them the treatment. Because I can tell you, they're -- I bet you my bottom dollar, they're going to be enthusiastic to explore that further.

Thank you so much. And Cedric, if you don't mind answering the manufacturing question.

Yes. That's easy to answer. So we just disclosed that Bachem is our manufacturer [indiscernible] as well as in the U.S. Drug substance is the same for PNH as it is for the eye, with a big difference that we use above 0.0017 of the dose in the eye as we do systemically. So we have built the capabilities. This will not be a problem, and we've been very happy with -- in our CMC group and the progress there.

Thank you, Cedric. I believe we have 3 more questions?

Yes, sir. Your next question is from the line of Alethia Young.

Maybe just for physicians and fellows, just can you talk a little bit about kind of -- there's obviously 5 million patients globally. But how do you kind of think about early subsets or subgroups of populations that might be of the most need or the lowest-hanging fruit as you head into an initial, hopefully, launch in the coming future?

Thank you. Dr. Holekamp, where would you start? From the patient population, who are the patients with the highest need that you may target first?

I think the low-hanging fruit, if you will, are people who have already lost vision in one eye and still have some foveal sparing or extrafoveal lesions in the remaining eye. And I think my top priority will be to preserve some central vision in the remaining good eye. These patients understand the disease. These patients will be very enthusiastic. They will overcome any fear of having an injection, and they will be my first patients. I think to better understand who the next group of patients will be, we will take -- not only looking at the safety data from the clinical trial, but also the efficacy data, the efficacy of what some groups benefit most and really trying to drive treatment there because at the end of the day, we're all eager to preserve vision in our patients with GA. But I think I know where my starting point is, and that is to preserve vision in my ocular patients.

Your next question is from the line of Joe Stringer.

So first, on just the -- for the KOLs, what would you sort of consider to be a clinically meaningful reduction in lesion growth? And if the Phase III [ approaches ] similar reduction, say, 29%, as in the FILLY trial, what kind of percentage of your patients would you use the drug then? And then just quickly for the Phase III trial again for the rate of exudations and anti-VEGF treatment use. Is there a threshold or a bar that you'd be looking at in terms of, if it's above or below a certain percentage, you would use it in a certain percentage of your patients? Or would it be on a case-by-case or patient-by-patient basis?

Thank you. Thank you very much for the question. Dr. Heier, I would like -- if you answer the first part, and then I'll take the second part.

Sure. Would you mind, the first part, again, was related…

So what is sufficient clinically significant…

Right. Right. So what I will tell you is where we are right now with the numbers that we've seen between 20% and 30%, patients have all been extremely anxious to participate in the studies. And I would expect that they would be equally anxious to receive treatment out of the study. So I think anywhere in that ballpark is huge. Now to some patients, that will mean more than others. If you -- as Nancy said, if you're foveal sparing, anything that could potentially prevent you from involving the center would be critically important. So I think most patients will be willing to do it if those are the numbers we're at.

Thank you very much. So to answer the question whether there's a threshold on the exudations is, the Phase III, there isn't. And the way it's going to be compared is, of course, with the control arms. We expect a better balance between the arms in the Phase III since it's double-masked. And we know what the natural history rate is, which was presented today. So those are going to be the benchmarks for comparison, but there is not a set bar on what we considered to [indiscernible] I think, and this should be put into context with the rest of the data and the natural history data out.

Your last question is from the line of Laura Chico.

This is Kenneth on for Laura. So you noted and we noticed in the inclusion criteria on ClinicalTrials.gov, in OAKS, it includes the criteria on microperimetry, but DERBY does not include this. Wondering if you can comment on this and just on microperimetry in general. Why is it in one study and not the other? And maybe you could help conceptualize the end point.

Yes. I can take that really quick. So microperimetry, we included in the study. It is an exploratory end point at the time that the studies were designed. Maybe it would be used for Europe. These are global studies. Microperimetry is not an easily accessible technology everywhere. So we -- instead of having subgroups of microperimetry in both studies, we decide to have the full sample size for either study to have better power on that technology without having to go over both studies. So those are the main reasons that microperimetry was included in a single study. And Caroline?

I have to comment on that. Yes. So I'm actually a clinical investigator in our center that does microperimetry, and we have a lot of experience with microperimetry. It's an exploratory end point because it is exquisitely hard for patients to do microperimetry, even just from elderly age, notwithstanding the fact that the patients in the study have elderly age and they have geographic atrophy. So that is why it's an exploratory end point. We don't understand all the clinical relevance of microperimetry, and a lot of patients cannot do it well. Like most of our visual function measures, it's not perfect. And the worse your eye disease is, actually the worse you will perform on microperimetry. And also, understanding what the end point means is not so simple with microperimetry. So that's why it's an exploratory end point, and it's not a be-all end point for the FDA.

Thank you very much, Dr. Baumal. I think that was last question. So I want to thank everybody on the phone. All our physicians here today, thank you for being here today, for your help on our programs. Thank you to the Apellis team. And with that, I'll be passing it back to the operator for closing the call. Thank you very much.

Thank you for joining us today. This concludes today's webcast. You may now disconnect.