Good afternoon. Thank you so much for joining us. It's our pleasure to have with us Sana, and from Sana, President and CEO, Steve Harr. Steve, thank you for being here. To start, can you provide a quick overview of where the company stands today, your strategic focus on the type 1 diabetes and in vivo CAR T franchises, and key catalysts for the second half and beyond? Sure. Well, first of all, Salveen, thank you, and thank you, Goldman Sachs, for having us, and thank you for everybody in the room and online who are joining us. Maybe, just one last thing I'd say, I think you know we'll make forward-looking statements, please refer to our most recent SEC filing, our 10-Q, for risk factors. Take a step back. The company, when we started it, we had a vision, that vision was to try to go after and tackle what we thought were the most important challenges in making our vision a reality, which was turning this idea of engineering cells or modifying cells into medicines. We went after two different platforms and technologies. One, if you take a step back and think about almost every disease you can think of is caused by either a missing, a damaged cell, or a cell that's kind of gone awry. To go after the missing or damaged cells, we wanted to be able to make cells outside the body and transplant them in. To do that, you have to have cells that you can manufacture at scale that will engraft, function, and persist. The largest challenge has been persistence, since the advent of transplant medicine, essentially allogeneic rejection. You put someone else's cells into your body, it will see them as foreign and reject them. We went after this with a thing we call the Hypoimmune Platform, which its goal was to hide cells from detection by the recipient's immune system. We'll come back. We've made tons of progress. I'm going to come to that in one second. The second was, if you just take a step back, the other thing you want to be able to do is fix cells and fix their DNA and their RNA, you can do almost anything you want to a cell in a Petri dish. The real challenge has always been delivery. What we wanted to do was come up with a delivery system that allows us to deliver genetic payloads in a specific and repeatable way. We really focused on cell specificity in doing that's something that we've now created this Fusogen platform, which you brought up. With the Hypoimmune Platform, what we're going after is type 1 diabetes. Again, I'm going to just kind of peel the onion back a little bit. Type 1 diabetes is a giant problem. I have a daughter who's about to graduate from college this weekend. I looked up a few months ago, if you have a 22-year-old female and she's diagnosed with HIV behind door 1, breast cancer behind door 2, or type 1 diabetes behind door 3, it turns out the shortest expected lifespan is actually type 1 diabetes. In that time, they have this challenge of trying to grapple with meals and blood sugars and lows and highs, and they can have blindness, amputation, heart attacks, strokes, all those things. We have to do better for this group of people. It affects 10 million people globally, right? I mean, it's growing 5% a year. It's a really big problem. We know what the issue is. The immune system gets confused and kills the pancreatic beta cell in the patient. Up until 103 years ago, it was a death sentence. They just died. Since that time, we've had insulin, but it's just not good enough. Now I'm going to get into what we do. Again, I'm going to peel the onion back. If you take about 25 years ago, it started with a guy named James Shapiro in Canada. He figured out how to isolate pancreatic islets. We think of it as an islet as a beta cell plus a support structure from a pancreas of someone who died, and transplant them into someone with type 1 diabetes, and it works. It's been published in The New England Journal of Medicine, other places. The challenge is it's not a very scalable or replicable supply source, and patients have to be on lifelong immunosuppression, just like an organ transplant. There aren't that many people for whom immunosuppression is better than lifelong insulin. Thousands of people have gotten it, but its impact's pretty limited. Over the last few years, you've seen several groups take pluripotent stem cells, grow them or differentiate them into islets, and transplant them, and they work. Again, published in The New England Journal of Medicine, and it seems to be more predictable. It works pretty much every time. The patient population is still pretty limited who benefit from it. It's important. I'm not going to say it's not important, but still, there aren't that many people for whom lifelong insulin is better than lifelong, sorry, worse than lifelong immunosuppression. What we've now been able to show is that we can make gene modifications, and we can hide these cells from immune recognition. Again, published in The New England Journal of Medicine. Now you have all the component parts together for a cure, a functional cure. This will happen. I think it's now inevitable. We may not do it. We may stub our toe along the way. I hope it's us, and we're a long ways towards that goal. To your question on that one, we've been working really hard, and we can get into doing what, in moving a gene-modified stem cell-derived pancreatic islet therapy into people that will be a single injection into the muscle and will be a functional cure for people with type 1 diabetes. Normal blood sugar, no insulin, no immunosuppression for life. After, I tell people it's like we've been waiting for Godot, and I think Godot is finally coming. Our goal is to file our IND and begin the phase I study this year. I think it'll be really pretty quick that we begin to understand if this works or not. If you transplant normal cells into someone with type 1 diabetes, they'll be rejected within days. If we see that this person has cells that are living, and it's at a month or so, and they're functioning, they're making insulin themselves, we can get into how you measure that, you'll know that this probably is going to work. The second, that will happen very quickly after we start testing people. The second element will be, are these people ever able to get off insulin? You probably know that within a few months. You can see, first of all, cell survival, then do you get normal glucose with no insulin? Then you're going to want to know how generalizable is this, right? If we're kind of six out of six or seven out of seven at the beginning, you're going to feel really good. If we're kind of like three out of six, four out of eight, something like that, you're going to say, "Give me a few more. The error bars are still pretty big." We'll learn all of this over the next, call it 12 to 18 months. We all know through 2027, we'll be able to figure out even the kind of replicability of this. I think after that, it's pretty straightforward to move into a registration study. It's a lot of work to do. That program, I think, has got a lot of promise and a lot of work, and super excited to see how this turns out. We've also been making this in vivo CAR T platform, you've probably seen there's a lot of excitement in the field about this. CAR Ts generally, over the last 12 years or so, have just had dramatic effects in patients. In particular, you can cure somewhere between a third to 50% of people with lymphoma leukemia and multiple myeloma, we're starting to see its impact in autoimmune disorders. Its utility has been really, really under-penetrated. That relates to both complexity of manufacturing as well as the chemotherapy. The goal of the in vivo CAR T platform is to actually just make the drug inside the patient, give them a single dose of the transgene, they'll make the CAR T in their blood, that will go and find all your pathologic cells, be that tumor cells or B cells in autoimmune. I think if you were to look at all the non-clinical data in non-human primates, we have a best-in-class molecule. Others have moved a bit faster than us now, they're ahead of us, we need to see how it really. Non-human primates don't always predict what happens in humans, we have to see what happens in humans. Our first drug is called SG293. It targets CD19. We'll dose patients this year. Maybe we'll learn if it works or not this year. We don't know what the exact dose will be, if we're on dose, it happens pretty quickly. If we have to go higher, we'll take a little longer. It won't take that much longer. We'll know well in under a year if this is working or not. If it works, you can see a rapid expansion into non-Hodgkin lymphoma to start other tumors. We'll go into autoimmune diseases, and then we have a second drug called SG227 to go after multiple myeloma. A lot will happen for the company, I ask about the next six to 12 months during that time. I'm super excited about both programs because of all of the work that's been done both in the field and inside our company to de-risk them. I love that they're idiosyncratic biologic risk. There's no correlation between what happens to the fusogen in vivo and with our type 1 diabetes. That, I think, means we have a pretty good shot of having at least one and maybe two really important medicines as we look forward. That was probably a longer answer than you wanted. Yeah You wind me up, you can't slow it down. Steve, you've been here since kind of the beginning of this evolution in cell therapy, from the autologous to the allogeneic to the in vivo as where we are today. The optimization efforts that have played out here with the technologies, with delivery have extended, I think, as we've gone on over time. When you look at in vivo, how long do you think it's going to be to kind of optimize for that perfected ability to deliver in patients? I don't know what you mean by optimize. I'm astounded by the fact that in 1995 or so, there was the discovery of the first really powerful B-cell depleting agent, which was rituximab. We're still sitting here 30 years later, 30+ years later, talking about better B-cell depleters. I don't know if I've perfected, but I do think that we're at a place now where they are more than adequate to drive safety and efficacy in a simplified regimen for patients. I think the key elements in moving forward are one, mass deliverability. Ease of use, off-the-shelf like a drug that physicians, payers, and patients are used to. Number two, if we can get rid of this chemotherapy. You always hear about CAR T toxicities. You have to actually read through the labels. It actually took me a while at a CAR T company to figure out how frequently we were causing things like severe infections and other things. You want to get rid of the chemotherapy. The third is it needs to be curative. I think there's something magical that happens when a person who has been kind of near death's doorstep is cured of a disease. Making people live longer is helpful. Actually, having the privilege of uttering the word cure is super important. It needs to happen at least in a third to 50% of people, or it's not competitive with current technologies. I think those are all of the. You can't be adding significant new toxicity. With these in vivo CAR T cells, there's been the emergence of kind of a new toxicity, which is a peri-infusion cardiovascular collapse. We can call it that. I think that's been increasingly managed with a single dose of relatively high-dose steroids. I'm optimistic that our mechanism will allow us to either get around that or have a lot less of that, but you never know till you get into humans. When you get into single treatments with very limited toxicity, like few days of fever kind of thing, leading to cures of very deadly diseases and very prevalent diseases that are very in the autoimmune setting, it's going to be transformative, and they're going to have a very important place across a number of different indications. How is your technology differentiated from other VLPs and the LNP mRNA approaches that are employed by competitors in vivo? Super important question. I'll start. The company made two fundamental assumptions when we started down this path. Again, what you're trying to do is make a CAR T in the body, right? The first assumption that we made was that specificity will matter. By that, we mean only deliver the genetic payload to the T cell. Don't go into liver and lung and a whole bunch of other cells. We think that matters because one, just safety, right? Two is immunogenicity. You get into other cells, you can create an immune response to your therapy. The third is manufacturability. T cells are a small number of the cells in your body, so if only 5% of your drug's getting into T cells, you have to make 20% more or 200x more. The second is that we have a belief that with the CAR T specifically, you want to integrate your signal into your target cell's DNA. The reason is, even the best-case scenario, you might make 100 million CAR T cells. You and I have something like 100 billion B cells, let alone how many cancer cells you might have in the body. It's very difficult to make that math work of eliminating all of the target cells without these CAR T cells expanding. Expansion is basically divide, and your DNA goes with both progeny. Others made the exact opposite bet, and probably LNPs/mRNA is the exact opposite bet, which is one, you don't want to integrate. mRNA is going to be good enough, and it can be safer, right? Because you're not going to break the DNA. Two, it doesn't matter what other cells you get into, just get into enough T cells. If that turns out that they're right, my sense is people would rather take mRNA than DNA, and my other sense is that they're easier to manufacture, they'll probably be very difficult competitors for us. Most of these other virus-like particles, and I can get into why, they've all made the bet that you need to integrate, right? Most of them have made a bet that specificity is less important than we have, right? I have to say their early data, right, you're talking about handfuls of patients to date, but at least from several different sponsors, you have early data that are very compelling, right? We'll have to see if our specificity really does lead to a safety, immunogenicity, less manufacturability advantage or not. I'm optimistic that we'll have an important medicine, and how it fits into these dynamic competitor environments, we'll just have to see over the course of the next year or so. On your in vivo CD19 CAR T candidate, SG293, you presented data at ASGCT demonstrating robust CAR T generation and B-cell depletion in NHPs without lymphodepleting chemotherapy. Can you just frame this data and your confidence in the drug and help us understand when we're going to see next data from this program? There's a lot that I think we've figured out in that, right? We figured out how to deliver genetic payload safely and efficiently to the target cell. We figured out how to do that without going into other cells, right? Including some very easy cells to get into, like the liver. The third thing is we figured out what tricks we needed to do to really get these CAR T cells to expand and function inside, to really deplete the target cell, which while not causing any toxicity issues, right? I don't think that's true for others. We're very optimistic about this working. The next set of data will be in humans, most likely, right? I don't want to guarantee this works, but I'll be surprised if it doesn't work. I think there's still a reasonable chance you run into a safety problem, right? It's just a first in human study, new technology, you have to figure out how do we make sure we don't have that. The bar is higher than it working. The bar is offering a patient the best solution available to it. We'll have to figure those things out as we go forward. I think with those non-human primate data, I think we go in with a lot of optimism and confidence that this will work, right? Now we just have to see how that really translates into people. We'll begin to generate data soon, right? The answer isn't that far away. Ideally, what you'll see is in lymphoma patients that very rapidly, within a month or so or less, they're moving from a very difficult situation into a deep, complete response that hopefully is a durable cure for them. The drug uses CD8 as a source of entry, while many of your competitors use CD3, which has been associated with toxicity. How confident are you that a differentiated CD8 entry mechanism will translate to a meaningfully better tolerability profile in humans? I think it's a good assumption. I don't think it's guaranteed. I was in medical school when this drug, OKT3, came out, and it was a drug that targeted CD3, and you saw these people just go into what we're seeing in some of these other drugs was profound cardiovascular collapse. We were very intentional for several reasons, but that's one of them, of dissociating entry from activation, right? We weren't going to target CD3. We have our own risk with CD8. I think it should be safer, but there's no guarantee, right? Our own risk is that our T cells are made of both CD8 and CD4 cells, right? The patient will have plenty of CD4 cells around to serve as helper cells to kind of kick start the CD8, but we have to see that work. It seems to work in our non-clinical models in animals and non-human primates. If I were on the other side of it, I'd point to that and say that's the thing you have to make sure that you get as good of efficacy with your CD8-only entry. We're optimistic that's true. 99% of the work is done by the CD8 cell, there are people who only get CD8 CAR T cells, and they seem to do pretty well, but we'll have to see how that plays out. Let's say you have positive phase I data here in non-Hodgkin lymphoma. What is that profile that you need to see to decide to move further into oncology and also potentially autoimmune? It's a good question. You hate to be overly precise about phase I data because you're learning as you go, right? I think the first is that it needs to be just the combination of safety and efficacy need to be super compelling. Without getting into-- I'll come back to numbers in a second, but if new safety signals emerge like they did with these first in vivo CAR T cells, we figure out how to manage that. Right? I think that's the first thing we have to really focus on. If you look at the CD19-directed CAR T cells, autologous CAR T cells, yeah, it's kind of like a third of patients or so end up with a durable complete response. We look at large B-cell lymphoma, probably 50% or more with the indolent lymphomas, right? We'll be in the more aggressive lymphomas to start. You'd like to see at least a third or so, right? The air bar is going to be pretty big, right? Ideally, you're well above it because that makes you feel better about the air bar. You'd rather have room to degrade than have to hope that you get better as you move into phase II. Let's just call it a third-ish is the right place to go. The second is that's in oncology, right? You have lymphoma, leukemias you can go after. The other will be in the autoimmune space. Moving the autoimmune space, I'd say two things that you'll be really looking for. One, the safety bar is just higher, right? These are patients who will live for decades. You're trying to offer them a one-time treatment that's curative and allows them to go back to their old way of life or at least gets them into a durable remission for five, 10, 15 years. You can't tolerate the same level of toxicities. Safety will be a bigger question. The second is you could see that we have really deep B-cell depletion and complete B-cell resets, and maybe the efficacy isn't quite there on the cancer side. Your efficacy bar might be a little lower, right? I think we have to kind of see what the profile looks like to see exactly where we go. We have such a great opportunity, I hope, to continue to move forward in lymphoma, to expand into other tumor types, and to move into autoimmune very quickly if the profile allows us to. Pivoting here over to the type 1 diabetes data. We've now seen 14 months of data from the iSP demonstrating long-term cell survival and function without any immunosuppression. How does the iSP differ from your lead program here, UP421- Yeah What key takeaways from the data support that development? Yeah. Making a gene-modified stem cell-derived islet, which is what we're trying to do, turned out to be probably harder than I thought it would be, right? It's taken us some time. During that time, we had an opportunity to learn. What we did with this UP421, the team. By the way, the first time a scientist who drove this program, she's wonderful, came to me, said, "I want to do this. I want to gene modify cadaveric islets, and we'll transplant them into humans and see if they survive and function." I was like, "That's a terrible idea." I'm glad she prevailed, right? I think it really does de-risk our program. What we were trying to do in that study was. It was a 62-year-old person who died suddenly, and he donated his pancreas. The team isolated his pancreatic islets and gene modified them. We were pretty good at the gene modifications, but not perfect, only about. Because you're knocking two genes out, and you're knocking one in. You only end up with, like, 40% of the cells or so being fully modified, right? First difference is you're putting in a pretty dirty product, right? Not dirty, unsafe dirty, but dirty like some of them are fully edited, some of them are not edited at all, some of them are partially edited, right? It's from a 62-year-old who had a hemoglobin A1c of 6.2, so they're not the perfect islets either, right? The second is because all that, it was put in a lower dose, right? Our goal was not to cure the patient. It was to see we're going to put no immunosuppression on board. These cells should be killed within a matter of days. Will they survive and function? When we first learned the guy who got them, he's 42 years old, and he was making insulin for the first time since 1987. I was like, just that statement alone is wild, right? He's had no immunosuppression, and these cells continue to function and survive or survive and function out 14+ months. That's just the last time that we tested them. They'll be tested again shortly. The stem cell-derived islet program, you take one cell, and you modify it. That cell is your. You have 100% confidence of what the genome is, right? You've sequenced everything. We then grow that cell into many, many stem cells, right? Just as an example, it's a billion cell dish per patient. Let's call it a billion and five with release assays and things. That means to treat 1,000 people, that's 1.5 trillion cells, and you're starting at one, right? You know the genome. Now you have to grow them up to a bunch of stem cells, and then you differentiate them into pancreatic islets. From the outset, we will be dosing patients, again with no immunosuppression, again with a single simple procedure into the muscle. This time, we hope it's at a therapeutic dose. This time, 100% of the cells will be edited. That actually should be easier in some regards, right? We know that patients develop an immune response to the partially edited or unedited cells. They develop a robust immune response, or he developed. In this case, hopefully these cells, there's no immune response at all to them, and people will go off and do very well. We'll have to see. Obviously, to shoot for a cure is a high bar, right? If we can do it'll be transformative for type 1 diabetes, and it will be transformative for us. We'll see where we are pretty shortly. Now that the master iPSC cell bank is established with regulatory alignment, can you walk us through the remaining gating factors for the IND? We're pretty far along, and they're becoming fewer and fewer, but they're still not zero. Generally, things you have to do. Align on clinical development plan and really kind of create a clinical protocol. I think you can be confident that's kind of being done. The second is, you have to have a non-clinical testing plan, right? You're looking for evidence of efficacy, biodistribution, GLP toxicology, all those things. That is getting pretty close to being done but not quite done. Everything we're doing that we've done before, it's low probability something happens, but if something bad happens, it could set us back a bit, right? The third is we have to finish our manufacturing tech transfer. What's happened to date is the drug has been manufactured inside of our company by our people. Now we have to transfer CMO, CDMO, where it's done by their people. We're doing that. It will happen, right? It may not happen at exactly the pace we hope it does. I'm optimistic that it will. Those are the only two things left to do, finish the non-clinical testing, finish tech transfer, and actually make the drug in the GMP setting, and then off we go. Can you talk to the phase I study design here, what the inclusion criteria is, what the type of patients and how many you want to enroll, and how you'll select the starting and step-up doses? Yeah. We want the patient population to reflect the need in the real world, it's more or less all comers between the ages of 18 and 65. It's not entirely true. There are a few corner cases. Someone who has a heart attack last week, that would be confusing to put them in a clinical trial. There'll be some things like that won't be in there. But it's 18- 65 year-olds. Hopefully, on the back end of, we'll bring up phase I in a second. That's that question. But on the back end of finishing that initial phase I, we can move into some younger people where there's a lot of desire to have a product and some older people, right, where I think people who are 65 and older still want to get off insulin. It's pretty much, just think of it as all comers, more or less. Yeah. I think the phase I study will be relatively limited in size, call it 12 to 15 patients. Our goal will be to get people off insulin with no immunosuppression. From there, it hopefully is relatively straightforward to move forward. I think we kind of know what the right dose is in the transplant literature. There were thousands of people who have gotten these cadaveric islets, and they get around 1,000 IEQs or insulin equivalents, right? That translates to around a billion cells-ish, right? That's going to end up being more than that within an error bar. The Vertex, which is in a registration study, they're 800 million cells, as an example. Again, that's within the error bar, right, of a billion cells. I'm guessing around where you want to be. Unlike the CAR T, I could see us being logarithmically wrong on the first dose. I'll be surprised if we're off at all, but if we're off by more than 20%, 30%. What is that bar on C-peptide production and length of follow-up that you'd want to see in order to move to a pivotal? I don't think I care as much about C-peptide. Just take a step back. When a beta cell makes insulin, it actually makes something called proinsulin. When it's secreted, it's cleaved into insulin and C-peptide. C-peptide is a one-to-one measure of how much insulin the patient is actually making. The actual number is probably around, most of us walk around with around 200-ish, is the number. What we really want is the patient to have normal blood sugars off insulin. It's not a biomarker-based benefit. I think we're really looking for the clinical benefit. to move forward. For duration, I don't know. I think you're going to want to see this for more than a month, although I don't think anything's going to happen after a month. After six to 12 months, if you're seeing these stable, they should be stable for a prolonged period of time. That's probably the phase I-ish time where you say, "Okay, let's go." The barrier to moving into a registration study is unlikely to be clinical. It's more likely to be manufacturing. Yeah. Right? We have a process that's good enough for phase I. I don't think we would want to commercialize this process yet. There are two elements we want to keep working on. One is scale, and the second is cost of goods. I think we'll go through three periods of manufacturing. One is good enough for phase I. The second is good enough for early commercialization. You need to have that process locked before you can begin a registration study, right? The third will be good enough to treat tens and tens of thousands of people, right? I mean, this is a disease that's so prevalent. If we somehow cure 100,000 people a year and they only need one dose, and it works in everybody, like the most idealistic scenario, all you've done globally is take the growth rate of type 1 diabetes from 5% to 4%. Right? This is a marketplace where we have to be ready to scale and deliver this medicine at a quantity that we haven't seen from cellular therapies to date. How long do you think it'll take you to be good enough for manufacturing for phase I to the pivotal and the early commercial? I wish I knew the answer for sure. We're making a lot of progress. I think it will happen sometime next year. Now, it takes you nine to 12 months to move from locking a process into something that you could move into a registration study. Let's just say that, sometime in 2028, we could be ready to go. I don't know when that will be, though. Yeah. I think that we're too far away from it still, and we have some more work to do. There are always two answers in these manufacturing challenges, which are very simple. I love math. Cells per year is like cells per run times runs per year, right? I think of cells per run as a science problem and scaling up, runs per year as a capital problem and scaling out, right? It can be solvable with the capital and scaling out once you get to a certain threshold. We have to make sure we're at least at that threshold. I have a last question here. Where do you stand now from a balance sheet capacity basis? We ended last quarter with around $100 million. We raised around $95 million last quarter, partly through a collaboration, pretty novel, Mayo Clinic invested in the company, and then we did a small financing with a great healthcare-focused investor. That puts us $195 million pro forma. It gives us money middle of next year-ish. That gives us enough money to turn the card over on both of these drugs, right? I think we'd like to go into that period with a little bit better balance sheet. The Mayo Clinic has an option to invest another $25 million. That expires late this summer. That would make us feel a little better going into things. I think that that's a little bit more of the kind of buffer we need. Fingers crossed that we have some more progress in our relationship with them. That's kind of it. What I like is that it gives us enough money to get through both these data points. We will need to raise more money. I don't love hanging out here, and I know Brian, our CFO, doesn't love. He's only been here for three or four months, and losing sleep over days or weeks of when your data come is not an ideal way to run the company over time, and we want to take the balance sheet, which is still, I think, a weakness of the company, and turn it into something that's neutral. I don't think it'll be a strength, right? We're not going to be AAA rated or anything. Likely, we will wait to do that until we have a much better sense of what our capital needs are, and those capital needs will be defined by the progress that we have from these programs. Our hope is that we have the privilege of accelerating and increasing our investment across both these programs as we move through both these Platforms as we move through next year. I think we'll have access to capital to do that. That's kind of where we are. Great. Well, with that, thank you so much, Steve. Right at zero. I know. Well done. Thanks, everybody. Appreciate your time and attention. Thank you. Thank you.