Good afternoon, everyone, and thank you for joining us for this afternoon's session with Generation Bio. I'm Whitney Ijem, one of the biotech analysts here at Canaccord. It is my pleasure to introduce President and CEO Geoff McDonough to tell you a little bit more about Generation Bio before we dive into Q&A. Thank you, Whitney, for the kind introduction and also for the invitation. It's awesome to be here, and the main event here will be our conversation and Q&A, so I'll just use this opportunity to frame what we're up to in a few slides. Generation Bio is about really pushing the limits of what's possible for genetic medicine, and that's from the point of view of delivery and cargo to create durable, redosable genetic medicines for both rare and prevalent diseases. Before I go into the overview, I'll just remind you, we're a publicly traded company, and to apply appropriate caution to forward-looking statements. So as I mentioned, Generation Bio has two platforms. They are mutually reinforcing, but also independently deployable, and we are operating in both those modalities with these platforms. Our delivery platform is a first and best of class as a biologic ligand-targeted cell-targeted lipid nanoparticle delivery system. What's interesting about that is it unlocks redosable in vivo delivery to previously unreachable cell types and tissues. As I'll talk about in the next few slides, in particular, that opens up very interesting selective delivery opportunities in T cells and in HSCs, hematopoietic stem cells. And what's great about this platform is that it's really robust and something that can be deployed to a variety of cell types and tissues over time. On the cargo side, we've started working on DNA in 2017. It's been a long journey, working through a series of challenges for what's required to make DNA the kind of robust, redosable, long-lasting gain-of-function cargo that we all wish it could be, but for many reasons, it's taken a long time to fully instantiate its potential. Turns out what's important is making DNA immune quiet and allowing it to avoid innate immune sensors. So that's really what we have invented and are working with, and it's why I say these two platforms are, mutually, beneficial, but also can be deployed independently. So we are building a portfolio that focuses on novel approaches to three program areas. The first is focused on really owning delivery to T cells, starting with redosable in vivo applications for autoimmune diseases, and applying exactly the same approach with a bespoke ligand for reaching HSCs and building a portfolio based on the ability to selectively deliver to HSCs, starting with sickle cell. So these first two parts of the portfolio are really leading with that ctLNP delivery system, and we're deploying a range of cargoes across these indications. Hemophilia, on the other hand, while it still uses that ctLNP delivery system, is really dependent on immune quiet DNA to get the long-lasting gain-of-function expression that's required to move the needle in heme, a field that is has been evolving rapidly over the last 10 years. The entire platform sits on a base of two very important core elements. One is that the combination of this delivery and platform and cargo platform creates a very low cost of goods and supports our ambition to both deploy products for rare diseases globally, as well as to go after larger mainstream diseases in the future. The second is that the company sits on a platform with very strong cash on the balance sheet, which funds operations through the second half of 2027, about $217 million on the balance sheet at the end of last quarter. So I'll just show a couple of short slides about each of these areas, and hopefully that will set up the conversation, Whitney, in a good way. We are deploying this highly selective and efficient ctLNP delivery system to T cells because we believe it's the underpinning for an ideal approach to redosable in vivo T cell programming, and obviously, creating CAR T is one aspect of what can be done with that capability. I think CAR T becomes really appealing when you can redose it. It allows for a point-of-care treatment. The cargo can be flexible. We can manage the cargo according to the duration of expression that we're looking for, and it should be a modality that fundamentally expands patient access and can displace or expand on what's possible or what's emerging as possible for ex vivo CAR T. Now, the key is to be able to get to T cells efficiently and at a level of transduction expression that's relevant for what will become a new modality for T cell therapeutics, namely in vivo. So each side of this slide is an important part of the equation. On the left, what you can see is very efficient dose-dependent transduction of both circulating and splenic T cells with the use of this very specific ligand against a receptor on T cells. And you can see here at a dose of 0.5 mg per kg, that we're able to transduce and show expression in 10% of circulating and splenic-resident T cells. And you see a nice dose-dependent rise in the ability to transduce T cells all the way up to 2 mg per kg. Now, why are we showing this dose range? LNPs from the perspective of systemic administration, have a TI limit at about 0.5-1 mg/kg. And so, you know, you'll see a lot of transduction data in this in vivo T cell space, but the disclosed doses can be as high as 2 mg or 3 mg/kg. It's just not gonna translate to humans. So it's really important to have the efficiency and the potency of delivery at 0.5 mg/kg or less, and we show transduction all the way down to 0.05 mg/kg on this platform. What unlocks that is the fact that we don't deliver to the liver and the spleen. We deliver less than 1% by volume of every dose to the liver and the spleen. That core characteristic of our LNP, which we refer to as Stealth, it's what's required to achieve this level of potency and efficiency. When we do reach T cells, we are getting very nice expression of CAR on the surface of T cells in vivo, in the range of 3,000-5,000 CAR molecules per T cell. So to put these two datasets into context of ex vivo, it's about 5% of circulating T cells that get reinfused after expansion with ex vivo CAR T. So the ability to reach 5% with a single dose, I think is, is meaningful. And, the correlation between efficacy and CAR T and clinical outcomes is at somewhere in this range of 3,000-5,000 CAR molecules per T cell. So we're in the zone here, even at our lowest doses, and I think that's just where you'd wanna be at this stage. The second thing we're focused on is extending this very specific and selective ctLNP delivery system to HSCs. Why is that? We think it has all of the same benefits of moving current ex vivo therapies to in vivo. That obviously applies in a very direct way to sickle cell. But I think what's really appealing about HSCs is they sit at the apex of most circulating cells and many important stromal cell types. And so the ability to act and to reach primitive HSCs in the marrow and to shift their programming in a way that impacts daughter cells, I think leads to a very interesting set of follow-on indications as well. So I mentioned earlier that access is something that we're very focused on expanding. It's obvious that ex vivo in oncology has certain logistical and industrial limits in terms of what it can expand to, to do for oncology patients. But we're also seeing that opportunity to expand into non-oncology applications, such as autoimmune diseases and infectious diseases. That transition, I think, is really powered by a transition to in vivo, and that's the analogy I wanna make for HSCs. Obviously, we wanna start with sickle disease and beta thal. Those are straightforward, indel-based editing approaches, but it's very interesting to think of how much larger the HSC therapeutic space can become when you can go after a broader range of hematologic disorders. And ultimately, there's the potential to use HSCs as a biofactory. So lastly, for hemophilia, as I mentioned, what's important here is the ability to get very long gain-of-function expression in hepatocytes to address, hemophilia A with the expression of factor VIII. And it's critically important that you do that in a way that's redosable, so that the impact for each patient can be titrated to meet their individual needs. And it's really important to tie those two things together because of the way that the hemophilia A field is evolving. I've been involved in commercialization in heme for 20 years, and the dimensions in which the field is evolving are really important to keep in mind for what's gonna be successful in the future. Extended half-life factors and bispecific antibodies have done a really nice job of elevating the level of coverage, specifically trough coverage, so that patients are now able to expect that they should not be bleeding. So that is now table stakes. You need to have an average annualized bleeding rate of 0. So the dimensions upon which this can really be advanced is duration. So today, the outer limits of duration for EHLs are around a week. Maybe they'll get to 10 days or 14 days for some patients, and for bispecifics, maybe the puck is moving to a month. So the ability to get to really years at a time with a titratable level of coverage, what we're playing for here is expanding the period of time which hemophilia A patients can live in, where they're free of hemophilia. And so this, in our view, would be a definitive move in that direction without any of the drawbacks of virally delivered gene therapy. So that's the goal for Hemophilia A, but obviously underneath it, there's a whole slew of lysosomal disorders that can be treated in a similar way, as well as inborn errors of metabolism. Those are all therapeutic contexts that benefit from the same combination of titration and durable gain of function. So I think that's probably a good umbrella, Whitney, to stop with, and let's start chatting. Let's do it. Excellent. Yeah. So thank you for that. And just to start off on one side of your platform, which is the LNP side. So can you put into context of the broader LNP landscape, your LNPs? Like, what's different? What have you changed? There's a lot of companies out here talking about new LNPs. What have you guys done, and how is that different relative to some others? Yeah. Well, maybe it's worth starting with the reality that putting an antibody or a targeting ligand on an LNP is not a new idea. It dates back to the early 2010s, when Alnylam started working with GalNAc on LNPs. And they published widely that when you put a targeting moiety on an LNP and do nothing to change its fundamental biodistribution, the pull of the clearance of LNPs by the liver and the spleen, this is their job, to absorb particles of about 100 nanometers, is so strong that the targeting ligand makes no difference. So if you look at the half-life of an untargeted traditional LNP, it's about 8-10 minutes in circulation, and if you do nothing more than add a ligand to it, whether it's a sugar or an antibody, it's 8-10 minutes. So the opportunity to get high levels of circulating LNP beyond the liver and the spleen so that those ligands can engage with their target receptors, depends on avoiding that clearance by the liver and the spleen. So we started with a two-step or a two-level ambition. One, let's create a base composition that would avoid that uptake by the liver and the spleen. We call that stealth. And then that ideally should allow us to engineer ligands that are very specific for their target receptor and in circulation long enough to reach them. So in our case, whether it's untargeted or targeted, in non-human primates, we have a half-life in excess of 8 hours, compared to 8-10 minutes, and that's what allows us to have low input doses and high circulating levels of LNP to reach target tissues like T cells and HSCs. Got it. Okay, that's helpful, and to double-click on the stealth relative to the liver, again, a lot of people talk about liver de-targeting or liver sparing. What have you guys shown in that regard? You touched on it in your presentations. Sure. But, like, how much goes to the liver, and again, kind of how differentiated is that relative to- Sure ... what you've seen? You know, I mentioned this, this idea of a therapeutic ceiling for systemic LNPs as something that no one likes to talk about, but in effect, it's about 0.5-1 milligram per kilo, and it's much closer to 0.5 than it is to 1. Mm-hmm. So, the importance of de-targeting the liver and the spleen is not academic. So if you de-target the liver and the spleen, the baseline level of absorption or clearance is 99% by volume. So most companies are saying, "Well, we're de-targeting the liver and the spleen." What they mean is there's relative de-targeting, which is important. Mm-hmm. You know, every amount, every small amount matters, but in general, if you look at the data, it's 95% or 94%. So you're still wedded to high input doses and relatively low and relatively brief amounts of LNP available to reach target tissues. We're delivering less than 1% of every dose, which means that we can dose very, very low, as I mentioned, as low as 0.05 mg/kg and still get that durable, abiding, circulating level of LNP. We think for genetic medicines, potency and TI win, and most genetic medicines systemically are failing on potency, and it's because you can't really dose above 0.5 mg/kg with LNPs. Mm-hmm. Got it, got it. Okay, and does increasing the half-life of these LNPs floating around in the blood, does that introduce new safety concerns or questions? It doesn't. These LNP components are ones that have been administered chronically for a decade or more in a variety of clinical contexts, and we know when we look at our LNP that what drives the decay is the LNP itself coming apart and being processed. So we don't see anything new related to the extended half-life. Got it. Okay. So with the LNPs available to be retargeted, as you mentioned, one of the ongoing efforts is targeting T cells in partnership with Moderna. So can you provide a status update a little bit? What can you tell us around what you're working on? You kind of hinted at a few things, but, but what can you tell us, and I guess what are the key things left to, to figure out or, questions to answer before- Yeah ... you can tell us more? Yeah, great question. So what I showed here in this one slide was a humanized mouse model for both the efficiency at which we reach T cells and the number of CARs that we're expressing in each T cell. We obviously wanna move and recapitulate that data in non-human primates to be absolutely sure that we're seeing the same kinetics and the same efficiency and potency. So that's an important data set that we're evolving at this moment with Moderna. And then obviously, the bigger picture question is: How do you marry cargo with the therapeutic ambition of let's take autoimmunity for, you know, as the kind of prototypical example for CAR T? You want to get very potent expression of a CAR on T cells to get immediate deep depletion of B cells in vivo, and if that can be redosable and titratable, then it just really widens the aperture for what's possible for autoimmune applications of CAR T, and I think that has always been the ambition of our partnership at the front end. Mm-hmm. As I mentioned at the beginning of my presentation, our ambition is to control therapeutic delivery to cell types that have real depth of therapeutic application. So I would argue that programming T cells to kill B cells is just the beginning of what's possible when you can control what T cells are doing. Mm-hmm. But it's the most obvious and the most validated thing to do to begin with. Sure. Moderna will lead the autoimmune deployment of redosable CAR T with mRNA as the cargo. Mm-hmm. That allows us to really learn together, build the pathway into in vivo, and leaves us free to think about how and where does the application of iqDNA and T cells make the most sense. Gotcha. Okay, so if I heard you correctly, you're kind of where you wanna be, and as you talked about transfection rates and kind of relative to in ex vivo CAR T, you're where you wanna be. You're just box checking... Maybe not box checking, but you're- you gotta double-check in non-human primates, but you're good to go there, and now it's just a question of the cargo, the targets maybe, and just work on that side. Yeah. I mean, you know, the field, I don't think it's an exaggeration to say, is investing $hundreds of millions to understand the role of CAR T in autoimmunity... both with auto and allo approaches. In vivo has a lot to offer there. It's probably, in the end, going to be its own modality because the kinetics of T cell reprogramming are so different compared to ex vivo, but it also leaves a lot of those industrial and clinical limitations behind. So I do think there's a big landscape to explore, and I think it makes sense to let the field continue to mature as we perfect the system, and then it's one of the reasons we decided to work with Moderna. They have a really good track record of executing robustly once they've got, you know, conviction in a new field. Mm-hmm. Okay, noted. Noted. So, we'll stay tuned for more there. On the cargo side, you mentioned Moderna is leading the way with mRNA. You guys own DNA, so is that, is that the split between in the collaboration? Is there a scenario where, you know, Moderna's pursuing a target with mRNA, and you guys could pursue the same target with DNA? It's broadly, it's broadly true that they have rights for mRNA, and we have rights for DNA. They have rights to a specific and very limited number of programs, also for DNA, so they have up to three programs in T cells, that they can name. Everything else for DNA, we're free to, to develop on our own. And, you know, the benefit of the partnership is co-investing and learning together in new spaces. So, you know, we'll just be thoughtful about how we do that. We, we don't intend to be in competition with one another. If, if for certain applications, mRNA is the killer app, great. Mm-hmm. In others with DNA is that perfect application, then we'll, we'll pursue that. Okay, fair enough. And just one... Going back to the delivery side, on the targeting side, the "you're where you wanna be" comment, is that true in the sense of the targeting moieties you added to the LNPs as well? Yeah ... relative to optimizing those and/or finding ones that might lead you to different subtypes of T cells? Yeah, such a good question. So this has turned out to be one of the most fun things about this platform, is when you take away the high levels of clearance in the liver and the spleen, you get to very high levels of efficacy and potency with the ligand, and because we're generally working with small format VHH nanobodies, they're highly tunable, and we have developed a very deep capacity to both identify receptors of interest that are ideally uniquely exposed or expressed on target cell types, but then to calibrate how tightly we bind them, how long we bind them, and what happens when they're bound. Are they internalized? Are they activating? Are they not activating? There's a whole area of biology there that's very nuanced, and to your point exactly, we've shown data for three receptors. Every one of those three receptors has turned out to be tunable, so all of those ligand profiles have moved leftwards in terms of potency with further iterations. But to your question, within T cells, there are many, many subtypes, and so as we turn the wheel here, we can look for subtypes of T cells that uniquely express additional receptors, and that goes for the same, whether it's subtypes within T cells or as we move to HSCs or yet other cell types. It's a very robustly applicable knowledge set. Okay. Noted. Moving over to HSCs then, which is the other cell type you guys have talked about targeting, are there different or new questions when targeting those types of cells, or is it pretty translatable once you've figured out T cells, you can change the targeting moiety and go towards HSCs? Yeah. So here's a perfect CEO answer: yes and no. That yes, in the sense that identifying the receptors and running a campaign to pan for, identify, optimize, and then apply the ligands, totally the same and just as robust. It's working just as well on HSCs as it did in T cells, following the same path and just going a bit faster 'cause we've learned better how to do it. Every cell type has its own kind of core biological considerations, which shows up in the PK and the PD. So primitive HSCs, which are the ones you really wanna go for, for sickle cell and for many other applications, they're situated deep in the marrow, and the health of the marrow in every indication is different. So obviously, in sickle, it's not healthy marrow. Mm. That's true in MDS and a variety of other situations, and in other applications for HSCs, the marrow is healthy and normal. This is why I was stressing earlier why potency and selectivity is so important. If you have a cell type that's in a diseased marrow, having an 8-10-minute half-life is just not gonna get it done at doses that are feasible for normal LNPs. So this is one of the considerations for HSCs in sickle is the health of the marrow in sickle patients and how to really be sure that you're getting the time on target to really get the expression and transduction that's required, which is why it's nice to have the flexibility to select several different receptors and tune how you interact with them. Mm-hmm. Okay, and can you talk a little bit more about the approach, from a cargo perspective that you're pursuing here, and maybe relative to the currently approved editing, ex vivo editing approach? Yeah ... in sickle cell? Yeah. I would say right now we're pretty open-minded. We have in our own hands several different systems that can make the changes that are required to be durable to shift the curve for hemoglobin F, and I think we're pretty open-minded about which one will win. I think I mentioned this for T cells. Ex vivo is not in vivo. They are just entirely different. The state of the cells are different, the nature of the genetic cargo is different, the way that cargo is ultimately put back together in vivo is different. So I don't think we wanna make any assumptions about which system will work best in HSCs. We have systems that are working, and the goal is to see them work in humans. We'll probably carry more than one option through higher species until we're really secure that we've got the right combination with our delivery. Okay, and when you say systems, you mean different gene editing systems? Yeah ... or different. Okay, so - Yeah, different editing. CRISPR versus something other- Yeah ... something else? Yeah. Okay, got it. Yeah. Got it. But very much focused on a fetal hemoglobin target?... upregulation approach? Uh, yeah. Okay. Yeah, I mean, I think obviously correcting normal hemoglobin is also perfectly valid, so I don't think we would say it was only hemoglobin F. But I think the ones that are best validated as benchmark systems are- Mm ... HbF at the moment, so. Okay, noted. And then what's the roadmap for this program? Any milestones you're working towards that you're looking to internally, and anything we can expect externally? Yeah, so it'll follow very much the same pathway as T cells. We started in T cells showing efficacy for delivery in humanized mouse, and then we showed with a reporter molecule, then we showed it with the expression of CAR, and then we showed the efficacy of CARs. And we'll do the same thing here for sickle. We'll start with the delivery system in humanized mice, then we'll show with a reporter, then we'll show the delivery with the editor, and then we'll show the impact on the disease, the subclinical marker disease here, hemoglobin F. So I would like to be able to show that progression through the balance of this year and into the first half of next year for delivery, the delivery of edits, and then the efficacy of those edits. Okay. Okay, exciting. We will stay tuned. And then last but not least, on hemophilia A, you, you made a good case. You sort of said, kind of talked about the landscape and how you guys are different, but what's the feedback you have from patients and from physicians in this space? From an investor perspective, there's probably a lot of people that are jaded by some other gene therapy approaches here. Oh, for sure. So what are you- For sure ... hearing from patients and doctors? For sure. Well, let's just say, or let's remind each other that if you went and talked to patients and physicians about their aspirations and their hopes for engaging with gene therapy, they all said the same thing, and they say the same thing for every therapy that's in development for hem A, which is: "How secure can I be that when I get this, I'm going to be free of my hemophilia?" And what they mean is not bleeding, independent of activity level. "And for how long- Mm ... can I get it, and what's the downside?" That safety is, you know, paramount in all diseases, in hemophilia, it's yet more, more nuanced. So when people asked, patients and physicians 10 years ago about viral gene therapy, it was, "You're gonna get it once, it's gonna last forever, and there's no downside. Yeah. So now we're all really surprised that there's no uptake because you get it once, but you can only get it once. It may or may not work for you, and it probably will only last for three or four years, and by the way, you can never get another virus again. So for payers, it's like a big risk and no guarantee. So the failure of viral gene therapy to fully address what patients want is not totally surprising, and what I think we're aiming to do here is work from the platform of titratable, redosable therapies like EHLs and bispecifics, but extend their efficacy. We're keeping redosing and titration and a payer model that payers understand how to work with- Yeah ... but extending their duration for years once they are titrated into range. That feels like a place where those three things that patients want to be true can become true. Sure. But that's our bar. If we don't hit that, we're not gonna, you know, advance a program for Hem A. Okay, fair enough. Fair enough. Excellent. All right. Unfortunately, we're out of time, but thank you so much. My pleasure. That was very helpful. Yeah, it's really fun to be here. Thanks, Whitney.