Welcome to the next session of TD Cowen's Immunology and Inflammation Summit. We're really pleased to have with us Steve Harr, the CEO of Sana Biotechnology, who will be talking about their programs in diabetes, as well as maybe a little bit beyond that, as some of the other things that are going on at Sana as well. Maybe to start off, I will have a holistic question for Steve. We'll turn it over to him in a second to do a kind of a high-level overview. Also, we do want this to be interactive if possible. To the investors on the line, you are welcome to submit questions either through the portal or you can email me directly at mark.fram@tdsecurities.com, and I will work those into the question list as well. Maybe with that, Steve, do you want to just kind of make sure everybody's kind of level set? You did have some updates on your earnings the other day. Just kind of a quick overview of the company, then we'll jump into some of the specifics. Yeah, first of all, thanks for having us. I'm sure you guys know we'll be making forward-looking statements. Please do take a look at our 10Q, where we've spent a lot of time on the risk factors. By the way, our general counsel is pretty clear. He doesn't like us to call these earnings. He calls them results because earnings would be misleading since we are losing money still. A few things. As you noted, we have, I think, two programs that we're super excited about. One is SC451, which is a gene-modified stem cell-derived islet for the treatment of type 1 diabetes. The goal is essentially a functional cure. It's a single treatment where you with normal blood glucose or euglycemia, no more insulin shots, no monitoring, and no immunosuppression. I think all of the proof of concept or the component parts we can get to are there. It is really now moving to our scalable solution, which is what we hope SC451 is, and hopefully beginning a study next year. We did a subtle thing in our press release, which was that we moved from guiding just to attempting to get an IND done next year to also beginning the study next year. I think that gets to the confidence that we have of how well this is going. It is not to say there are not some risks between now and then still, but things seem to be on track or ahead of track with what we are trying to accomplish. It is pretty rare that you find a market with 10 million people-ish. It's growing mid-single digits per year, where there hasn't been a significant innovation in a new therapy or new therapy launched in about 100 years. The patients are grossly unsatisfied. That relates to the burden of the disease and the minute-to-minute, day-to-day challenges of managing blood glucose, as well as the long-term sequelae of things like kidney failure, blindness, amputation, heart attacks from too high a blood sugar, and the short-term risk of death from too low a blood sugar. All of that gives us a chance to create what we hope is a really, really important medicine for many, many patients over a long period of time. Our other thing that we can chat about is an in vivo delivery platform. The first place we're taking that is in vivo CAR T cells. I would argue, based on all the data we've seen, if you happen to be a non-human primate, which I know you're not, that what we've shown is, I think, best-in-class safety and specificity and efficacy. We need to translate that and see how that goes into people. Again, that runs in a kind of similar time frame for beginning human testing without all the human biologic proof of concept that we have in the diabetes program, but in an area where we think that the biologic leap from non-human primates to humans is not zero, but relatively digestible. I'll stop there for a little bit. Okay. Maybe keeping it a little high level to start, just on 451, can you kind of explain what the changes are that you engineer into the cells to kind of make them hypoimmune, and how does that really differ from some of the other approaches to immuno evasion that have been tried over time? There are a few other programs. I think I got it. I I broke up for one second, which is what's different about what we're doing there by also. By the way, the whole company was founded on this idea of trying to overcome allogeneic rejection. With that, what we have learned is we are also overcoming, at least in some settings, autoimmune rejection, right? At the highest level, there are two levels of the immune system. One is the adaptive immune system of B and T cells, and the other is the innate immune system of things like NK cells and macrophages. Grappling with the adaptive immune system is something the field has kind of figured out a while ago, and that is doing things like knocking out MHC class I and class II. Now you no longer are able to show a fingerprint of what's going on inside the cell. The challenge with that is that we've evolved a system to kill these cells if that happens, and that's called natural killer cells. It's been really trying to figure out how do you overcome natural killer cells. What we do, what we figured out was that overexpression of CD47 works, and it works pretty much every time, right? Knocking out MHC class I, knocking out MHC class II, and overexpressing CD47 in concert is the system. It's amazingly elegant, right? It's only three gene modifications. We've now shown that this works in many, many species in many, many settings, including now in two separate cell types across multiple settings in humans. I'm quite optimistic that this is a generalizable effect and that this is something that we're going to be able to turn into a medicine in type 1 diabetes and beyond. Okay. Maybe that's a good transition. You started talking about the IST. I guess what has been learned by that IST in Uppsala? Are the cells still present to this day? How long is this lasting? Yeah. Take a step back. Our goal is to make a gene-modified stem cell-derived islet. That is a consistent product, and we believe it will be scalable. It turns out that's super complicated and really hard. It took us more time than we hoped. In particular, as we were trying to make a gene-modified master cell bank, meaning the starting cell, we kept seeing emergence over time of certain mutations, and we needed to pause and get that right. The team wanted to go and figure out, did all of these exciting data that we had in preclinically translate into humans? That is where this Uppsala investigator-sponsored trial came in. What that is, is it's an islet that has been, islets have been isolated from a recently deceased person, a cadaver. They're gene-modified with the gene that we talked about and transplanted into a person. In this setting, that's not a scalable solution, right? We're not going to gene edit every cell. It's very complicated. You don't necessarily know. You don't have the time to test everything that you would have from a stem cell-derived product, all kinds of issues. It is a wonderful proof of concept of, does the immunology really work? What we learned from this study is that we can transplant cells, these gene-modified islets, into a person with type 1 diabetes and see them survive and function over time with zero evidence of immunologic recognition or rejection of our fully edited cells. That patient, the last time we updated you is six months. I presume the patient will hit a year sometime soon. The original dose was dose that was done the end of last year. Hopefully, we'll have a chance to update you at some point next year on where things are in a year. I presume these cells are going to last for a while. The donor was somebody who was older. He was in his 60s. He had elevated hemoglobin A1c, which means he already had some dysfunction of his beta cells. That's the person who donated these cells. It may exhaust and may not last forever. I think immunologically, we're in a very good place where we don't expect to see immunologic reaction to these cells. Okay. I guess two parts to this question. I think the IST did formally allow more dosing potentially to happen. Any plans to dose an additional patient or two? Kind of related to that, what gives you comfort that that one patient is enough to really, truly de-risk that immunology versus is there some risk that maybe you just got it is one patient and you just got really lucky? Yeah. Let's go through each of those. No plan to dose more patients with a cadaver islet. I mean, we are moving forward full steam ahead with a gene-modified stem cell-derived islet. That we can do at a dose that is clinically relevant for the patient, right? I don't see any benefit to doing this again. I don't know if it'd be any faster anyway than just kind of getting going with a real study. In the real study, we're hoping to provide—we'll see what the starting dose is after we have dialogues with regulators. Our goal would be out of the gate to get this to a level where at the first dose, the patient's getting meaningful clinical benefit, right? That would be your goal, would be a higher dose than you were able to give this one IST patient? Because, you know. Sure. You would look to give. Look at, so Vertex has a stem cell-derived islet program. It has been published in the New England Journal of Medicine. What they did was the first couple of patients got half of their target dose, and everybody else got the target dose. What we gave here was a single-digit% of a target dose. That just has to do with how many cells you can get from an islet, from a cadaver, how many of them are fully gene edited, some other aspects to it. Our goal would be to provide the patient with a much higher dose. I presume that will be done, right? I am not too worried about what the, it is going to be some level like that to start. That gets at will we do this again, right? The answer is we will not do the Uppsala thing. We're going to go forward with a drug. We're making the scalable solution, right? Remind me the other part of the question was. How do you get confident in that? Oh, from one patient? One patient is truly reflective of reality and not maybe some risk that you just got really lucky with one patient. Yeah. I I mean, I think that there's always some risk that what we do is not 100% generalizable, right? And we know that there are Achilles heels to what we do. We'll be very clear. The one thing we don't deal with is if you have pre-existing neutralizing antibodies to a cell surface protein and what we're putting in, that's not dealt with. What we have here is immunologically just very generalizable. I mean, I think whether you look at the number of, first of all, we've tested many, many people in the CAR T program. We actually published immunology from that in a Cell journal this summer, which shows no evidence at all of immune recognition or immune rejection, right? Second, we have done a lot of different non-clinical experiments to see that this was very generalizable across species and in human models. What we want to do is see does it work in humans. I'm quite convinced that this is a generalizable result. I think your point is fair, which is it's going to be 100% generalizable. That's probably some risk to that. I don't think we just got lucky. Okay. In addition to that publication over the summer, you also had a meeting with the FDA, your FDA Interact meeting, kind of to work towards an IND for 451. Just kind of what were the major takeaways from that meeting? And kind of what are the boxes that you still need to check to get to that IND? Yeah. By the way, I would say since that time, we've had dialogues with FDA and regulators in other parts of the world. I think we have real clarity and real alignment around what we need to do. I'm very optimistic in our ability to deliver on these things. Again, there's no guarantee that everything happens perfectly because we're dealing with complicated science, but we seem to be in a good place. There are just two things that need to be done. We need to kind of finish the manufacturing process, lock and transfer that into a GMP facility. We need to complete our GLP or non-clinical pharm tox package, right? That's basically it. Those are not trivial endeavors. Everything we're doing, we've done before. Nothing we're doing is something that's novel. We do have to do things like take all of our manufacturing process and move it from research-grade reagents into GMP reagents. Often, that has to do with just the qualification of the vendors and things like that. That is not risk-free. It is kind of like you're changing your ingredients in a recipe, right? You want it still to taste the same. You may have to play around a little bit. I think this is something we'll be able to navigate. Would you expect to update investors when it's fully into the GMP setting? Or just the next update is, "Hey, we've filed an IND," or even an IND has now been accepted? I don't know. Probably the latter. If something happened along the way that concerned us, we tell you. Otherwise, I think let us just go. There's nothing in there that is so meaningful that you'd just be that interesting to people. Okay. What are kind of those variables? The guidance right now is IND as soon as 2026. What are those variables that lead to more of an H1 versus a second half or maybe unfortunately pushing into 2027? We haven't really given guidance around timing within 2026. I just look at the early as is. That's kind of what we're trying to do, right? Just to be really clear. We've also said in time that we're going to start the trial, right? As I mentioned. We just have to finish those two things. That's it. I mean, what can make it last longer? Tech transfer run doesn't go well. That's probably the thing that you just say, "Oh, well, I got to redo it again," right? It doesn't take forever, but it adds a month or two to what you're doing and depending on how long it takes you to figure it out. I think that could really stretch it if you had some issue pop up in something like a GLP tox study that you wanted to kind of wrestle to the ground. That might be more than a month or two, right? Those are kind of things that could happen. Okay. What type of scale do you expect to be at as this moves initially into the GMP? Kind of where does that need to go over time? What does that process look like? Yeah. I kind of lay out that there are these kind of like four scientific challenges. I laid this out years ago. One is overcoming allogeneic and autoimmune rejection. We've done that. Two is to make a gene-modified GMP master cell bank that retains genomic stability through many, many, many divisions. And that took us a long time. We think we've now done that, right? The third is to be able to make this drug at a purity, potency, and yield to run a phase I study. We've done that, but just barely. I mean, this isn't a it's not like we're making this is like one run and we're done, right? The last one, the fourth one, is to be able to make this at a purity, potency, and yield that is commercially important. We have some work to do to get there, right? We're at a phase I scale. It's fine. We'll get through the phase I. It probably isn't anything that ever rises to your level of concern around scale. It's not going to be what we want to go forward with. I don't think it would be useful to go forward with it because I don't think it's at a scale that really matters. We're making already some really good progress on that next step, but we have more work to do. I kind of look at that as there are two elements to scale. There are two axes. One is the number of doses per manufacturing run, and that's a science problem. Then it's the number of runs you do, and that's more of a capital problem, right? We need to make real progress on the science part of this before we're willing to put more capital into it. I expect we'll make some again, we've made some good headway over the last couple of months. I expect it will be even better a year from now. My goal is, my guess is, again, last thing I'd say is I think timeline to registration study. The rate limiter is most likely to be process lock for our commercial manufacturing process. We may be faster than that, but that's most likely going to be what drives our timeline to then move into registration study and commercial launch. I think the clinical development program here should be relatively straightforward. Assuming it all works. Yes. Of course. This is a minor little thing there. You touched on it a little bit ago, but just kind of, I know you don't have a completely approved IND yet, but the phase I trial design, you might just kind of lay out what your expectations for what that trial looks like. I mean, you started to talk about dose, but patient numbers and follow-up that you need and what's really reasonable to. I don't know, but we know that someone's done this in the field already. Their kind of program is for a different patient population, right? A smaller patient population. It's with the context of immunosuppression. That was a, I don't know, like 13, 14 patient phase I and a 37 patient registration cohort. So 50 patients in total. I think you could argue ours could be smaller because we're not using immunosuppression. I think you could argue ours should be bigger because it's a much broader patient population. Where that ends up, I don't know. If I were betting, I'd just go with the latter because it's a more conservative assumption. Let's figure that out. I don't think it will be an order of magnitude more. It might just be modestly more. I think that's basically within an error bar, that's going to be what I expect we end up doing. Okay. Others have told us we can do it much smaller. I mean, of different geographies around the world. We can't go much faster than that and be ready manufacturing-wise anyway, right? We have some work to do. Okay. I know we're running out of time, but quickly, just on the, you mentioned earlier the In vivo CAR program. Just how do you think about that program? One, positioning it. You could go down the oncology route where CARs are, of course, approved, but you could also go down the autoimmune route where CARs have shown some interesting data. And we're exploring that with the allogeneic approach. Just how do you prioritize that? Healthy volunteers versus just going straight into patients? What does that program look like? I don't see healthy volunteers ever being a part of this program. There's too much risk. I think you have to have equiposity because you're making permanent changes to potentially the genome. Always be in patients. Yeah. We'll see. My overall view of this is do both and do them as quickly as possible. Which one we do first is still something we'll finalize and decide. There is probably an easier risk-benefit assessment to be made for a novel technology and novel therapy in a later-stage cancer patient. That may be where we go first, but let us figure that out. I do think that this is a we've made a lot of progress over the last couple of years. We've markedly changed the probability of success of this program. There's a lot of really great science that happened here in the last, call it, 24 months. We'll kind of outline more going forward that I think can show you that this is something that is pretty exciting. Maybe one last one on the fuse of gens. Just there has been a number of BD deals in the broader In vivo CAR space over the last handful of months. One, what do you make of that and where you're positioned? Also, how do you think through capital allocation here of the resources Sana does have at its disposal to push this program forward versus kind of not distracting from your work in diabetes that's maybe a little bit more near-term into the clinic? We made two super critical bets at the outset of this program. That is that specificity matters. That matters because of manufacturability, immunogenicity, and off-target cell toxicity, right? That may or may not prove to be true, but that's, and I think if it does prove to be true, we've got a better drug. The second is that you want to integrate, meaning that you need a self-replicating, you want to have logarithmic expansion of your T cells because you're trying to take out 100 billion-plus target cells, and you may only get into 100 million T cells at the outset in any of these technologies. Therefore, you need multi-logarithmic expansion. If those two things are true, then I really do think we've got a best-in-class therapy. If they prove not to be true, then we may have made things more complicated, and we fell a little bit behind, right? Ultimately, that's kind of the bet that you have to figure out if you're aligned with us on or not. I do think what's happening in the field is people have made bets either on things that are easier to make and can be mRNA. If it turns out that neither of those things are true, mRNA, LNPs are simple, right? If it turns out that one of them is true, then they're not going to work that well. If it turns out both of them are true, you're going to need what we do. From a capital allocation perspective, we can't screw up diabetes. I think this is just there aren't many opportunities like this where you have a disease of this scale and size with this much proof of biology and this type of potential clinical impact. We will not screw it up. We've been working through ways that we're going to pay for the In vivo CAR T. I've been pretty clear that we come up with the money without distracting from too, without lowering our probability of success in type 1 diabetes. A partner pays for it, or we actually just get an investor to directly invest into it. I think any of those three are still on the table. Okay. I think with the results we have, I had previously said we might just slow down. With all the activity and the results we have, that one's off the table. We will not just kind of let this die. Okay. Okay. Makes sense. Unfortunately, we're out of time, so we're going to have to cut it off there, but it would be plenty more to talk about. Thanks a lot for joining us, Steve, as well as everyone on the line. Thank you.