To the UBS virtual event. I would now like to pass the call over to Eliana Merle. Hey, guys. I'm Ellie Merle, one of the biotech analysts here at UBS. I'm happy to have Adverum here with us for a fireside chat as part of our Virtual Ophthalmology Day. Joining us from Adverum is Linda Rubinstein, the Chief Financial Officer. Linda, thank you so much for joining us. Maybe just to start off, can you discuss the design of your gene therapy for wet AMD and the technology you've utilized that allows for the delivery by intravitreal injection? Yeah. Thank you so much for having Adverum and at your conference and for having me present. Ixo-vec is our most advanced program. It's a phase III-ready gene therapy for the treatment of wet AMD. Ixo-vec uses our proprietary 7m8 capsid, which has been the gold standard for transducing the retina, and we use the capsid to deliver the transgene, which is a codon-optimized version of Eylea, which has been the gold standard for treating ocular wet AMD. Now, 7m8 is engineered for intravitreal delivery. Unlike first-generation ocular gene therapy, which needed to be administered with a complex subretinal surgery that involves removing the vitreous to enable gene therapy to be injected directly into the back of the eye, our vector was created by directed evolution to be able to cross the inner limiting membrane, so it can be administered intravitreally, which is the standard of care for anti-VEGF injections, and still transduce the back of the eye. And so the benefit of that is that it can be a simple in-office procedure. And then the capsid has been the gold standard because it has broad transduction of the retina and the choroid, where the disease is active. So when we turn the eye into a biofactory to make Eylea, it's at the location where it needs to be effective, and I think that's why we've seen such great efficacy. Uh-huh, makes sense, and I guess high level within the wet AMD market, we've seen a lot going on in recent years. Where do you think that gene therapy fits in? I think gene therapy has the potential to be transformational. So as we know, the standard of care until recently has been Eylea, with some market share going to Lucentis. That standard of care is moving to VABYSMO, right, Genentech's new bispecific, and Eylea high dose. Yet in the real world, patients don't get the up to 16 weeks of intervals between dosing that people were hoping for. In fact, one study showed that patients either got a one-and-a-half treatment interval extension or one line of vision. But yet, even with that, two and a half years after launch, VABYSMO has reached over $4 billion in revenue run rate and 30% market share, which speaks to the hunger for better therapeutics. And the TKIs are touting every six-month dosing, but many of their patients require rescue injections, and we haven't yet seen the data from repeat dosing. So even with all of this innovation, only gene therapy offers the potentially transformational benefit of having patients be injection-free, potentially for multiple years, and/or to experience a profound reduction in treatment burden. And there's also significant unmet need. Even with all of these anti-VEGFs, you know, wet AMD has been a leading cause of blindness in the elderly. More than half of patients stop treatment by five years. Others are undertreated, so you have fluid fluctuations and long-term vision deterioration. By maintaining a consistent level of anti-VEGF in the retina, gene therapy may not just reduce the treatment burden, but actually improve long-term vision outcomes. Makes sense. And there's a handful of other gene therapies in development for wet AMD. I guess, what are some potential advantages and differences of your gene therapy approach, relative to the others in development? Sure. The agents that are in development include REGENXBIO's RGX-314 and 4DMT's 4D-150. RGX- 314 has two main differences. One is that they deliver the transgene for Lucentis as opposed to the transgene for Eylea, and the other is it's delivered either subretinally or via suprachoroidal injection. Those are both challenging, right? Subretinal delivery is a surgery. You have to, you know, take out the vitreous, as I mentioned, so it's not scalable. It's not suitable for patients who haven't had cataract surgery, and it hasn't performed as well as the other agents. They get somewhat better efficacy with suprachoroidal injection. That's a challenging injection into a vitreous space that's highly vascularized, and so, they need to use, doses that are much higher, think hundredfold, which has a negative impact on cost of goods. So they don't have the IVT administration advantage that we do. 4D, like Adverum, has a capsid that takes advantage of engineering to allow it to cross the inner limiting membrane and transduce the retina after being delivered by intravitreal injection. So it's an intravitreal gene therapy, and like Ixo-vec, their gene therapy delivers a transgene for a codon-optimized version of aflibercept, but unlike us, they deliver a second transgene, for an anti-VEGF-C RNAi, so having two transgenes may compromise the expression level for each. We haven't seen their, anti-VEGF-C RNAi, levels, and we don't know what benefit they're delivering to patients. And you need RNAi to transduce every cell, unless it's a secreted protein, so I think that's been a challenge. We have the best efficacy of any modality in wet AMD, with treatment burden, injection-free rate, and relative to the other gene therapies, they don't all dry wet eyes and keep them dry. So, I think for gene therapy, anything that's not reaching 50% free of injection long term isn't competitive. And we've seen that, you know, the efficacy with the other agents isn't as strong. So I'll end there. I can keep going, but I'll end there. Yeah, let's talk about some of your data. Yeah. Your first-in-human OPTIC trial has now shown data out to three years of follow-up. Can you give an overview of the design of this trial and some of the major takeaways in terms of what you've seen, from this data? Sure. So our OPTIC trial, which began enrolling almost six years ago, delivered a screening injection to patients, and after we confirmed their VEGF responsiveness, they were delivered Ixo-vec. We followed them for two years, and patients are now in the middle of a three-year extension for a total of five-year follow-up. We tested two doses: 2E11 and 6E11, and we tested two very short prophylactic regimens, and we're the first to use a steroid eye drop for prophylaxis in ocular gene therapy. Patients were allowed to receive supplemental injections. They could receive a dose of Eylea if they had vision deterioration of 10 letters, so two lines on an eye chart, or an increase in fluid of 75 microns from baseline, or a vision-threatening hemorrhage, and then after the first supplemental injection, it was at investigator discretion. So we've now shown three years of clinical efficacy and safety that we presented at AAO last fall, and we've shown aflibercept levels for as long as we've measured out to four and a half years. And the major takeaways from it are that we have had best-in-class efficacy, right? At the 2E11 dose that we're continuing to evaluate in LUNA, over half of patients were injection-free at years one, two, and three. We had a profound treatment burden reduction, with 80% annualized injection reduction at years one, two, and three. And we did this while maintaining vision, drying the eyes, and maintaining that lower CST level. We've seen durable aflibercept levels, where the levels we measured at week 12 have been sustained as long as we've measured them. This undergirds the long-term benefits patients have seen and suggests to us that they may continue to be lifelong, and importantly, we have three years now of safety data, which is important for the field, as we showed at AAO, so no retinitis, choroiditis, hypotony, retinal occlusions. We saw dose-dependent inflammation that didn't impact vision and was addressable with steroids, and despite the suboptimal prophylaxis, we believe the safety profile that we saw at 2E11 represents an approval product, would be enthusiastically taken up by physicians and patients. 40% of patients didn't need any steroids beyond the very short prophylaxis. 14 of 15 were free of inflammation at year one. All were free of inflammation at year two, and the 6E11 patients are doing great as well. We think this is a favorable safety profile, where some patients have inflammation that's manageable with local corticosteroids and ultimately resolves, and that's a worthwhile trade-off for the profound efficacy we've seen with gene therapy. You know, the so what on safety is whether any of these events have an impact on vision, and the answer is no. To take us ahead, we've picked the 6E10 dose for our phase III, which gives us this ten X safety margin versus the highest dose we've tested at 6E11 and three years of follow-up. Great, and, maybe just to discuss the safety in more detail, I guess, what gives you the confidence, either from feedback from the regulators and/or physicians, that this is a safety profile that's manageable? So the key things with safety are, you know, does it impact vision? No. Is it addressable by steroids? Yes. Does it resolve ultimately? Yes. We've seen all of that, in our data in both OPTIC and we're tracking that with LUNA. And so this is a profile that has been acceptable to our investigators who have the experience with gene therapy. And increasingly, we're hearing from the retina specialists as they see, you know, the profound efficacy benefits that patients get, that the trade-off of having, you know, some drops of steroids are an acceptable and worthwhile trade-off. And interestingly, we're seeing that, you know, they're thinking about this for all of their patients, right? So the most heavily treated patients will probably become the first patients that are treated with gene therapy, but I think all of the patients want to have this, or many patients will want to have this transformational benefit. And so we're hearing from the investigators that we talked to, who are all very enthusiastic about joining our phase III, that this is a very compelling profile. Mm-hmm. Great, and you mentioned this briefly, but maybe just can you discuss more about how you're thinking about the durability of the gene therapy's effects, and specifically what you're seeing in terms of the long-term data, any variability and durability across patients? So, the durability we've seen is the best in wet AMD of any modality, so it's better than the standard of care or any non-gene therapy modality in development, and better than the other gene therapies. So at least three years, or could be lifelong. We see that because we have durable protein expression levels, which means the eye will continue to make the anti-VEGF. We see that because we can preserve vision, and we see that because we have, you know, CST levels that go down and stay low. So, just we have the clinical proof in the pudding, you know, out to three years clinically, and we're continuing to follow the patients, and we'll reinforce the story. We haven't seen any predictor for, you know, who will be injection-free versus not. You know, nearly all of the patients receive clinical benefit. If anything, the main predictor, for who will get the best clinical benefit is the patients who've needed fewer injections tend to do better. But all of the patients that we've tested, have done well as long as we've continued to follow them. Mm-hmm. Makes sense, and you've recently presented six-month data at ASRS from the phase II LUNA trial. Can you give us an overview of this trial design and how OPTIC informed changes that you made? Sure. For what it's worth, our OPTIC data is great. We made changes not in response to OPTIC, but rather to a DLT that we saw only at the highest dose tested and in a different patient population that we're not currently pursuing. So for the dose, we decided to test the 2E11 dose from OPTIC and a new lower 6E10 dose. Non-human primate data suggested that we could dose lower and still maintain the same aflibercept levels while reducing inflammation, so this gave us comfort in deciding to test the new lower dose. We also decided to explore enhanced prophylaxis. So the trial design is that patients receive a screening dose of aflibercept two to three weeks prior to Ixo-vec to test for their anti-VEGF responsiveness, and there are no additional loading doses after gene therapy. They receive Ixo-vec, and the primary endpoint is looking at vision at 52 weeks. Patients were randomized to the two doses, 2E11, 6E10, 30 patients at each, and it's a double masked trial, so the physicians and patients are masked as to the dose they received. And again, we'll follow the patients out to 5 years during the extension. And patients were randomized to receive one of four prophylactic regimens, providing longer coverage to address the period of peak immunogenicity. Patients either got drops or drops plus OZURDEX, which is an implantable dexamethasone, or they received one of those local regimens with oral steroids. And the supplemental injection criteria were the same as OPTIC, except that patients have to meet the criteria at all injection points, not just for the first supplemental injection. Great. And maybe can you talk a bit about some of the data that we've seen from the trial and your major takeaways? Sure. We presented six-month data at ASRS in July, and we showed that for baseline characteristics, the patients were very similar to OPTIC, where they needed about 10 annualized injections in the year prior to receiving gene therapy. They had decent vision. The eyes were slightly drier, so I think speaking to physicians' confidence in enrolling patients. And at the 6E10 dose that we're taking forward, we saw 76% of patients who were entirely free of injections, and we saw a 90% reduction in treatment burden. We saw stable vision, and stable CST. And we learned, therefore, that there's the potential for a best-in-class efficacy profile, even at this lower dose, while having the chance to improve the overall product profile. And so we've selected 6E10 as our phase III dose. We did see in a subset of patients, when we looked at the patients who'd received no more than six injections in the year prior, that this broader population of patients were 90% injection-free across both doses, so even better efficacy. And then on safety, again, no Ixo-vec-related SAEs. Inflammation was well managed and addressed with local steroids and resolved. Overall, 91% had zero or trace cells with local prophy at any visit through week 26. Focusing on the 6E10 group receiving steroid eye drops alone, all were off prophy, and none required additional steroids for IOI. When we overlay this to OPTIC, this compares favorably to the 2E11 patients at similar time points, and we now know that we have three years of efficacy data. On the safety data, this overlays well, where we had, you know, in OPTIC, 93% of patients free of inflammation at year one and all at year two. One new endpoint that we reported on was patient preference. We saw that 88% of patients preferred Ixo-vec to their prior intravitreal injections, and 93% would want it in their second eye. This is an important measure because it speaks to not just the effectiveness with efficacy, but also safety and convenience. I think it's an important measure since it reflects feedback from all patients. It includes the steroids they received, including the patients who had got oral steroids, which were not well-tolerated, of course, in elderly population in any inflammation. So I think that speaks to the dramatic demand that we would expect for this product if and when approved. Mm-hmm. Interesting. And in terms of the anatomical endpoints that you've measured, I guess between OPTIC and LUNA, I guess, what have you seen there? We've seen that we can dry out the eyes and, you know, CST is an important measure because when we look at retinal thickness, it's an indirect measure of the amount of fluid in the retina, and it's an important both biomarker and a leading indicator for vision deterioration, so in OPTIC and LUNA, the drier patients at baseline stayed dry. The wetter patients at baseline dried out, and we maintained CST levels, so I think that's great. The other thing to point out is that the field has come to understand in recent years that fluid fluctuations are a key contributor to long-term vision deterioration, even in patients treated with anti-VEGFs. The fluid fluctuations cause the retina to expand and shrink, and this trampoline effect causes a proliferative cascade that leads to scarring and vision loss. What we've shown in a couple of interesting case studies is that we can both dramatically reduce CST levels and fluctuations. We have one patient who received nine injections in the year before OPTIC. This patient had large CST fluctuations, notably after being hospitalized, which is maybe not an uncommon event for a 90-year-old patient, and missing a dose. After one injection of Ixo-vec, her CST levels decreased and stayed low. We had another patient who received nine injections in the year before OPTIC, with highest CST levels around 700, despite being injected every four to five weeks, which is as frequent as practical. You might think that patient was anti-VEGF resistant, but in fact, two weeks after the screening dose, fluid levels were down, so the patient just needed more frequent anti-VEGF. And then with one dose of Ixo-vec, those fluid levels stayed down and remained low. So, we're very pleased with this as maybe a point of differentiator in our efficacy versus other programs in the space. Mm-hmm. Interesting, and so you plan to start a phase III in the first half of next year and disclose more about the design later this year. I guess maybe how are the regulatory conversations going, and I guess what can you say about your phase III design and plans now? So we have been having and will continue to have discussions with regulators. We have PRIME designation in Europe and RMAT and Fast Track, excuse me, designations in the U.S. So this has been an ongoing dialogue, and we're looking forward to disclosing our phase III design in Q4, so stay tuned. That said, there is a lot that we can say about the design now. We've said when asked that we prefer the non-inferiority design, which has been used in nearly all wet AMD trials, right? Over the superiority design that's been pioneered by one of the TKI companies. We know that for gene therapy, our endpoint will have to be BCVA at one year as our primary endpoint. Our comparator will be Eylea 2 mg every eight weeks. The guidelines require either Eylea or Lucentis, and we're delivering Eylea, so it makes sense to compare to it. The debate about sham injections seems to have abated, so shams make sense for masking. Generally, we've seen loading doses being deployed across all of the pivotal trials with varying strategies. Although, I do note that loading doses after the investigational therapy look a lot like supplemental injection, but stay tuned for more. Mm-hmm. And I guess just in terms of like, you know, the cases of inflammation that you saw in some of the clinical trials, I guess, what are you thinking about in terms of the, steroid prophylactic regimen or potential, use of supplemental steroids? So in LUNA, we tested the four different regimens, and we determined that oral doesn't add any incremental benefits, so we won't have any oral steroids as part of phase III. Folks are glad. Nobody wants to be giving oral steroids to this patient population. We have two great regimens that have done really well in LUNA. Based on our 26-week data, you can see that both the difluprednate drops that we delivered or the drops plus OZURDEX had great results, and we've said that we will pick a local regimen, either drops or drops plus OZURDEX for the phase III, and we'll disclose that in Q4 as well. Makes sense, and I guess what were your considerations when thinking about the phase III dose selection, and how what you saw in OPTIC and LUNA helped inform that? At 26 weeks, the LUNA 6E10 data was tracking to be best in class and comparable to our OPTIC 2E11 data, where we've shown durable clinical benefit, including the profound treatment burden reduction and injection-free rates, vision maintenance, and anatomic endpoints. So when we overlay the 6E10 aflibercept levels with those we saw in OPTIC and see that they overlay to the levels where we saw this durable clinical benefit, it made it a good choice, right? We have a potentially even better safety profile than we've seen in OPTIC with the 10 X safety margin and the enhanced prophylaxis. So all of that went into the choice of the phase III dose. Just, you know, thinking forward, like, is there anything you can say about your expectations in terms of enrollment timelines for the phase III program? We're not guiding to enrollment timelines other than saying that we're planning to start in the first half of next year. But I note that there's a great deal of enthusiasm from the investigator community. All of the investigators we've talked to want to be part of the phase III, I'll phrase it that way and you know are very enthusiastic, and we enrolled LUNA pretty quickly. So we're very optimistic about phase III. Mm-hmm. And in terms of the 52-week LUNA update that we're expecting in the fourth quarter, I guess, what should we look to see, and what are your expectations for that data? I think we'll update on the, you know, various, different endpoints that we've been tracking. But our expectations is that consistent with the 26-week LUNA data, we hope to show that we're comparable or better to OPTIC with efficacy, safety, patient preference update as well. Mm-hmm. And as we take a step back, maybe just in terms of what's clinically meaningful, I guess, what do you see as clinically meaningful in terms of the reduction in supplemental or rescue injections and your expectations for what you'll see in phase III? At the broader level, I think anything better than the standard of care, right, is meaningful, right? And, you know, I've mentioned that VABYSMO, which is over $4 billion, hasn't been transformative, and yet, you know, there's such a hunger for any improvement that it's been very widely adopted. Certainly the level of clinical activity we've seen so far in OPTIC and LUNA represent a profound improvement on standard of care, with 80% injection burden reduction and 50% of patients injection-free out to three years. And we think that those will be not just meaningful, but I think potentially profound and speak to, I think, a large potential commercial opportunity as well. We think that ultimately, all patients will want a lifelong treatment for their wet AMD. So I think that, you know, we've heard from KOLs, our KOLs, and this has been echoed by the Street, that, you know, if you have 50% plus injection-free long term, that's both a bar and something that would be hugely beneficial and widely valued. Makes sense. And maybe just can you compare what you've seen in terms of the injection-free rate and treatment burden reduction to what's been reported from, other gene therapies currently in development for wet AMD? Yeah. You have to parse the data, right? There's a lot of, Mm-hmm ... nuance in the details. So as I mentioned, we have, 76% injection-free at six months in OPTIC, sorry, in LUNA, and that's tracking to, what we saw in, OPTIC, where 60% of patients were injection-free at one year, and the vast majority of those, 53%, were injection-free at two and three years, and the treatment burden reduction is tremendous. If we look at REGENXBIO with their subretinal and suprachoroidal, their injection reduction rates are significantly lower, and their injection-free rates are significantly lower. If we look at 4D, they're closer, but also significantly lower than ours. So 4D's reported, two phase II cohorts worth of data. In their dose expansion cohort, which is the most similar to the LUNA population, patients who needed an average of 10 annualized injections, they reported an injection-free rate of 63% versus our 76% at six months. And they reported a 37% injection-free rate at one year versus the 60% that we saw at OPTIC. So we're looking forward to reporting our data in Q4. So there's a dramatic, you know, 15-20 percentage point difference in injection-free rate. They do have lower treatment burden reduction in terms of annualized injections than we do. But I think that, for a relatively short time point, like six months, it, you know, injection-free rates is a more sensitive and indicative measure of relative efficacy. In their second phase II cohort, their population extension cohort, this is a population that received one to six injections in the year prior to gene therapy. As you might expect, when patients needed less anti-VEGF in the year prior to gene therapy, they're gonna do better. There they had a 77% injection-free rate at six months, and they haven't really reported updated data. They have a Kaplan-Meier estimate of the one-year data, but only one patient has reached 52 weeks, so I won't go there. Importantly, that's data in a patient population that had about four and a half actual injections and still a mean of about 10 annualized injections. So when I mentioned our subpopulation of LUNA that had up to six injections in the year prior, they also had about four and a half actual injections and 10 annualized injections, so very comparable. And our efficacy in that population was even higher with, you know, you know, 90% across both doses, 100% at 6E10 injection-free at six months. So that efficacy differential of 15-20 percentage points is even wider at that, in that broader patient population. And I wanna point out that, 4D in the PRISM trial in their population extension cohort added an extra aflibercept dose four weeks after gene therapy. So they're measuring the injection-free rates 20 weeks after the last aflibercept injection. We're measuring the injection-free rates 26 weeks after gene therapy, which is the last anti-VEGF, and only one of our patients received across all 60, and LUNA received a supplemental injection at four weeks. So it's not a necessary injection for ours, for our trial, for our agent. So I think we're very pleased with the profile that we've seen with Ixo-vec, both on its own and certainly compared to competitors. Mm-hmm. Great. That's a really helpful summary. I guess to give that back, can you summarize sort of the most important catalyst that we'll see from Adverum over the next twelve to eighteen months? Yeah, absolutely. We're very excited, as you mentioned, that we're gonna be presenting our 52-week data for LUNA in Q4, and then we're also excited to start our phase III in the first half of next year, and moving aggressively to bring Ixo-vec to patients as quickly as we can. Great. Just a question on manufacturing. Can you elaborate sort of on your manufacturing process and potential for commercial readiness? Yeah, absolutely. So we use a CDMO that has made over a dozen batches of our product, that we've used in our OPTIC and LUNA trials, and that are ready for phase III. We have all of the drug substance in place to run our two pivotal phase III trials. Some of it's filled in vials, the rest will be defrosted and filled as we go. Importantly, these batches have already been made at commercial scale, so we do not need to scale up for the process that we will be using at launch, and you know, the process that we have, which is a suspension process, you know, is a robust, resilient and reliable process. Mm-hmm. Great. In terms of your cash position, can you help us understand your cash runway from here, and particularly as you head into starting a phase III? Yeah, absolutely. We ended June with $174 million in cash, and we've guided that our runway takes us out into late 2025. Mm-hmm. Great. Well, thank you so much for joining us, Linda. It was great learning more about the programs, and thank you so much for everyone on the line for tuning in. Thank you so much for having us. Thanks. Bye.