All right, welcome back to the 45th Annual TD Cowen Healthcare Conference. I'm Marc Frahm from the Biotech team. We're really excited to have for the next session Sana and the CEO, Steve Harr, to discuss their efforts at allogeneic cell therapy broadly. We'll get into a number of specific questions that I've prepared, also anybody from the audience, if you have questions, we're happy to try to take those as well. Maybe to start off with, Steve, do you want to just kind of high-level status update for anybody who's maybe a little bit less familiar with what's been going on at Sana? Then we'll jump into the questions. Yeah, first of all, thank you, Marc, for having us. Thank you, Cowen. I'm sure you guys know we're making some forward-looking statements, and you can peruse our 10-Q and 10-K and look at the risk factors that you spent a lot of time with us. We're about six years old now. We started the company on two ideas. We wanted to build kind of, hopefully, one of the leading cell and gene therapy companies of the era. We tried to solve—we've been trying to solve two big scientific problems. One is to overcome allogeneic rejection. We want to make cell therapies in a way where you can make them at scale, and they'll graft, function, and persist. Really, the first thing is to overcome allogeneic rejection. The second area of platform we've been building is an in vivo delivery capability. I think you guys all recognize you can more or less do whatever you want to a genome in a petri dish, and the hard part is getting it to the right cell. Those are the areas we started the company. We've actually made a lot of progress. Right now, we have four different categories of drugs we'll go through. One is in type 1 diabetes. I'm guessing we're going to spend a lot of time there. I think the data we showed in early January get to the point where it's very clear now you can transplant a pancreatic islet cell in type 1 diabetes, and a patient can make their own insulin, and they can do this without any immunosuppression. I'd say at this point, what that really does is it makes a cure more or less inevitable for this disease. I'm optimistic that we'll be asked to do it. We'll tell you why. There are some things we still have to accomplish. The second area that we're developing medicines is with allogeneic CAR T- cells. We're doing this with, again, hypoimmune or these gene-edited cells that hide them from the immune system. We're looking to develop them in two different areas. One is in B-cell-mediated autoimmune diseases. These are diseases like Lupus, multiple Sclerosis, V asculitis, others. The idea here is to give a CAR T and kind of do a control alt delete of the immune system or the B-cell repertoire and get a patient off drug and allow them to live a normal life again for years to come. We have that as well in cancer, which we'll get into. The third area we've been working on is in vivo delivery. Here we do cell-specific delivery. We do it to a T-cell. We've now shown in non-human primates we can get B-cell depletion with B-cell reset from a single therapy and with no lymphodepleting chemotherapy. That is on its road towards an IND. Those are the three big areas or four big areas we've been going forward. We're happy to go into any of them. I think most of the focus will likely be on type 1 diabetes. It's a very transformative therapy and something that we're very excited about. We'll take questions anywhere. We'll go from there. Ok, thanks for that overview, Steve. Maybe we will start with the diabetes side since that was also the most recent kind of data update. With the cadaver-based product where you have the clinical data now, walk through the amount of C- peptide that was being produced and kind of how to think about that maybe on a per-cell basis and what that means for getting to that functional cure? Yeah, so I'm going to take a step back. Type 1 diabetes is a really simple disease. The immune system, for some reason, goes awry, and it kills the beta cell, pancreatic beta cell inside the patient. This patient can no longer regulate their own glucose. Up until the 1920s, it was a death sentence. We discovered, someone discovered insulin, and patients have done reasonably well, but it's a very difficult disease. About 25 years ago, a team that was first started out of Canada recognized that they had started to transplant islet cells. The way they would do it is they would get them from cadavers. A person dies, they isolate the islet. They then will give these islets to a person with type 1 diabetes. It works. Patients are now off insulin for 15-plus years. There are two challenges with this, though. Number one, it's not a great supply. It's not scalable, and it's not very replicable. Number two is there aren't that many people for whom lifelong immunosuppression is better than lifelong insulin. It hasn't been important, but it hasn't really transformed the field. Over the last few years, we've seen several parties now take stem cells and make them into islets and transplant them into people. What you've seen is all of those patients are off insulin if they can tolerate immunosuppression. What you see is a more scalable, a more replicable supply source. You still have a big challenge, and that is that there just aren't that many people that this is better for. The key to making a cure is to get people off of immunosuppression. We're working on that. Our main drug is a gene-modified stem cell-derived pancreatic islet program that we will transplant into a patient's arm one time. The goal is to have normal blood glucose with no insulin and no immunosuppression for life. That's the goal. I think that's inevitable. What we did to prove that we could overcome the immune system, as we've done this in monkeys, we've done this in mice, was we went into humans. We took cadaveric islets, so the source that people frequently use, and we gene-modified them with our hypoimmune edits. We transplanted them in a first-in-human low-dose study into people. The goal of this study was to see that these cells survived and functioned with no immunosuppression. It worked. The patient is making his own insulin for the first time in over 30 years. It's spectacular. A couple of things that come to Marc's question. It's a first-in-human safety study. The dose was lower than what you would want to see to get patients off of insulin. That's not surprising. That's what we did. That isn't the challenge of the field. We already know what the dose is. The dose is around a billion cells. We'll do it. It won't be a problem. The dose we put in is around 5% of a typical dose. What you see is that we make about 5% of the amount of insulin that you'd want. The way that insulin is measured, when we each make insulin, we make something called proinsulin first. When it's secreted from your beta cell, it's cut into C-peptide and insulin in a one-to-one ratio. If you measure C- peptide in the blood, you're measuring how much insulin the patient is making on their own. There is no other source of C- peptide. If you give an insulin shot, no C- peptide is in that shot. It's a one-to-one. This is a patient who had no C- peptide measurable for years and years and years and now has stable C- peptide, except when he eats, and then it goes up, as you might hope. These are functioning beta cells. Dosing, the higher dose is not the challenge of the field. The challenge of the field would be doing this. We can get into real challenges. That's kind of where we are. I guess it's fair to say, since you described the amount of C- peptide being produced on a per-cell basis seems pretty consistent with cadaver, that it's fair to say the genetic manipulations or time outside the body of that process is not kind of impacting functionality of the cells. We're not in any way impacting the function of these cells. That's true in vitro. And now we've shown that in vivo. Yes, the gene modifications we make don't change their ability to do what beta cells do, which is glucose-sensitive insulin secretion. That's all we want. Glucose-sensitive, glucose goes- up, insulin goes- up, glucose goes- down, insulin goes- down. That's all you want. Yep. Some investors have noted, if you think of vaccines or other times when you're trying to actually generate, not avoid, but generate an immune response, there certainly can be dose responses to that. Maybe talk through kind of the amount of effectively antigen you're putting in here and that you've shown you're actually avoiding the immune response. Because there's some investor fear that as you dose escalate and get to that real functional cure type of dose, that you may find that you didn't avoid the immune system enough. I guarantee you no immunologist has ever said that. Maybe an investor has said that. That is not like a useful, there are a lot of things to be concerned about. Those increasing antigen here to a level will not be the thing to worry about. Just think of a vaccine. I mean, it's a little tiny dose of some antigen or protein that you put into the muscle of a patient. We're putting in millions and millions and millions of cells that are functioning, creating insulin, and they're surviving. You don't see that. You don't sit around and you don't give people half of a liver because you want to decrease the immunogenicity of the liver transplant. You don't give them half of a lung lobe because you want to do something. What drives immunogenicity is the immune system, is the immunogenicity of the antigen. You already know. What we showed in the data—I'll take a step back. These cadaveric islets that we're doing here, these primary islets, we're gene-modifying them. It's kind of a—this is not our product. The product is a mix of three types of cells. You have fully gene-edited cells. We've shown you they're surviving. They're functioning. You have partially gene-edited cells. Then you have cells that weren't edited at all. They're all in the mix. The patients generate a robust immune response to the unedited or partially edited cells. Those are all dead. They're all dead. This is actually incredible. You have a pre-existing immune response to the cell type, the beta cell. We put it in, and the patient generates an immune response to the unedited cells. These cells still live. You're not going to see, as we go into particularly a stem cell-derived product, 100% of the cells will be gene-modified. They're perfect. You're going to start with one perfect cell and grow trillions of cells. That's the goal to treat many, many patients. They will all have all of the correct genetics in them. At the time of that initial data release, you'd only tested one dose within the IST. There was maybe some internal debate of whether you should maybe have another trial to test higher doses and get closer to that real functional cure type of dose within the cadaver set. Just latest thoughts on whether that's a value versus just parking it and waiting until you're ready to go with the stem cell-derived product? Yeah, so the impetus of this question, I think, is what we showed is that a patient at low dose makes insulin, does very well. This isn't a person who's been cured of their type 1 diabetes. We need a higher dose. Wouldn't it be really nice if what we did for a patient is to actually get off of insulin with normal blood glucoses and normal immune suppression? There are two ways to do that. It's to keep doing this cadaveric thing at higher doses, or it's just to go to the stem cell. I want the entire company and all of our capital focused on the stem cell. We have said we might do a higher dose because I do think scientifically it's interesting. We're doing something that's a little different than the field. We're putting it in the arm. It's a single injection in the arm. We can learn more about putting it in the arm. We might do it. I think if we do it, we'll be with someone else paying for the study. I think our shareholders' capital should be focused on the drug product that we're making. We've de-risked it now. We know the immunology works. Now we need to turn this into the stem cell-derived product. That's where we're focused. I'm not sure we might do it. I'm not sure it would even beat the timelines to a stem cell-derived product. It's just not that far away. Ok. I think that's a good transition. Just on getting that stem cell-derived product ready, one of the big steps, obviously, is finalizing the GMP manufacturing process. Just kind of latest on the status of getting to a GMP protocol for making cells. The GMP protocol is not hard. Let me outline for you. There are four scientific steps that we have to overcome to make this a really important drug product. One, overcome allogeneic and autoimmune rejection of a cell product. I think we've done that. Two, we have to make drug at a purity, potency, and yield to run a phase I study. We've done that. Others have. We have. I don't think that's the challenge. Number three, we have to take a single cell. That's a master cell bank. Create a GMP gene-modified pluripotent stem cell master cell bank where we're really comfortable with the genomics over time. Remember, we're going to put these cells into patients who would otherwise be living for decades. We have to take our safety kind of obligations very seriously. One of the things, every time a cell divides, you make a mistake or two. Those mutations could be anywhere. Most of it doesn't matter for us. It's non-coding DNA. What we don't want to do is create and select for mutations in gene repair enzymes. Because then you might be putting the patient at a higher risk over time that those cells we transplant become a cancer. That's been hard for us. It's taken us years. I think we've done it. We've now done it a couple of times in the research setting. We could either re-GMP those, which is just a matter of documentation and sterility testing. I say just. It's a lot of work, and it's not a guarantee. We're doing them in GMP settings. We're doing them in parallel. It will happen. The fourth scientific challenge is making this drug at a purity, potency, and yield to be really commercially important. We have a long way to go there, just to be very clear. I don't want you to be afraid. We're years away. It is a scientific challenge they're going to have to overcome. Just to put it into perspective, I think the dose is around a billion cells. It's been true for everything that's been done in the field so far. If you want to run a phase I study, it's 10 patients-15 patients. It's 10 billion-15 billion cells. If you want to treat 1% of the population per year in the world, it's around 100 trillion cells. It's gigantic. If we're treating 100,000 people, that would be an awesome drug. I think it would be one of, it would be a really important drug for the company and commercially. It is a lot. Again, to do that, if you just limited the United States and you treat 100,000 people, you could do that for over a decade and still not treat half of the people in the United States. I would challenge you to find a patient with type 1 diabetes, if you give them a single treatment of cells into their forearm muscle with no more insulin, no more blood monitoring, and no immunosuppression. Who would not want that drug? We do have a lot of work to do to make this really work. That partly relates to the size of the market and the opportunity in front of us. That partly relates to the status of the science. I mean, we have work to do to get there. Those are our four scientific challenges. Number one, done. Allogeneic autoimmune rejection. Number two, done. Make enough drug, purity, potency, and yield for phase I. Number three, great master cell bank. Not maybe. I think we got it. But I want to make sure. We'll get this very shortly. Four, a lot of work to do. Maybe taking this in order. On number three, the making the master cell bank, you said it's taking you some time. Is that about developing the assays to reassure yourself, or because you were actually seeing cell lines kind of diverge in ways that you didn't want? I would tell you, if I gave you a cell line and asked you to genetically give it to you and let you grow it for a year, you would have all kinds of genetic mutations all over it. It is very challenging. It has to do with three elements. One of them is the quality of the cell you start with. I think we understand that, what we want to do now. Two, the conditions around which you do it. Again, I think you've gotten a lot better. Three is luck. It will not happen every time you do it, even if you take care of one and two. You are just looking for a single cell. The reason I'm reticent to give you exact timelines is I'd rather be accurate than precise. Being accurate exactly around when luck will shine upon you is difficult. I think we've probably done that now. We just need to confirm that. It's a year of testing to ensure you have it. We're going through many, many, many, many different—you want to divide these things and take one cell and make it into trillions. You want to differentiate them into islets. You want to test the genome all the way through. You're going to then test these in vivo. It's a lot. I think we got it. In your dual track, you mentioned there's kind of two ways to get there. One is kind of going back and redocumenting some things. Just to put that in perspective, probably every embryonic stem cell line on the planet that you think about was once a research grade line that was re-GMP. That is what people are doing. This is doable. It is documentation and testing. You have to make sure that if it has been, for example, exposed to a murine protein or something like that, you have tested for every virus that could leap from a mouse to a human. You are documenting everything that it has been exposed to back into the supply chain. It is not just about the company. It is a lot about documenting what happened inside of other places. It is very doable, though. Very, very doable. It is not done overnight. Yeah. At what point is it just that it is taking too long to kind of get back and find all the source material? Is it still sterile? Yeah. Is it still sterile versus just doing it kind of doing it a second at a time? We're doing them both. Just run them in parallel. What is it sounds like it's at least a year plus of process. What's kind of the range of likely timelines there to get to a certified GMP? We haven't said it. I know it's the thing that people really want to know. When we say it, we'll tell everybody. I think we're in a place where we can give you some real granularity around that. Ok. The other one was the scale-up. Yeah. Yeah. Do you need to be there before you? No. You need to be there before you start a registration study. Not before you start a human study. That's not a problem. We can make enough for phase I. When you lock your—the way, again, people, one of the great things about a lot of drugs is the product is the product. You can characterize it. When you're in these cells, we're back into the early days of the biologics. You used to hear the process is the product. That's not entirely true. Characterizing a cell product and then changing your manufacturing process is complicated. You want to launch your registration study with a process that you're comfortable launching the drug with. If you change it, you have some real comparability challenges ahead of you. Over time, my guess is we will not launch at our final scale. We will scale more and more and more over time. We will launch at a scale, hopefully, that is commercially important. There are two elements that you can scale up. You can scale up. We will do both. I mean, presumably, to launch a phase I, I know you don't need to be there yet. Presumably, you want some reasonable line of sight that it's there so that you don't have to do too many changes once you've won the success scenario, showed POC with the cells in phase I. Sorry. I couldn't have to do what you said? Too many changes between phase I and phase III. You probably want, is it fair to say, you want line of sight to a phase III registrational process? I'd say that's typical. I would say that's what you almost always want. I think that no matter what you do here, there's going to be some changes that are important. We've not constrained the phase I process by what we think we need for a registration study. I mean, that would be what we would do in every other thing we've done. I think this is, there's never been a gene-modified stem cell-derived product that's gone into phase, it's gone through something like this. I just think we're going to have to be right instead of the fact you can do a lot to test these cells that you couldn't do before. You've got gene expression. You can do a lot of things. You can do epigenetic markers. You can do all kinds of things. You're going to have to, there's going to be some element of changes. It's inevitable. Ok. Maybe now we've got about a little under 10 minutes. Maybe we need to lock up. I'll keep track of it over here. We'll move to Gleam and the SLE or the autoimmune opportunity. It's broader than just SLE. I'm going to just take a step back. I want to make sure that people leave with this. I kind of use this analogy. I hope it's helpful. It's as if I think of it as like we're building a, we're trying to build a big electric, we're trying to build Tesla or something. The equivalent here is that people have been making electric cars for years, but they had to be plugged into the wall. They proved the technology, but they weren't that useful. What kind of this investigator-sponsored trial did is we proved we make a lithium battery, and we can make it. It's really good. Now the next step is to show you we put it all together, and it really works. Right now, we make the product at the scale of a roadster. The next thing after we make it the roadster is we're going to have to kind of figure out that the supply chain, the distribution channels, the economics, and the scale to really do this in an important way. We will do it. Someone will do this. It's not a question if someone will do this, in my mind. There are only three things that get in the way of this being a really important drug: time and capital, safety. We have to get safety right. You can't have the first three lithium batteries blow up on a fire. It wouldn't have gone forward. Scale. Those three things are the elements of success here. Not on the autoimmune. It'd be so many autoimmune diseases. Yeah. You want to level set kind of what do you think your cells have already established in the bit of NHL experience that you have released? What does that target look like in autoimmunity in terms of how deep and how long B-cell suppression you need to have? CAR T-cells, autologous CAR T-cells. I did them. I was involved with them. They work. They're difficult to scale. We can get into why. Allogeneic CAR T-cells. The challenge has been that when you put someone else's cells into you, two things can go wrong. One, they will try to kill you, graft versus host disease. Two, your immune system will try to kill them, host versus graft disease. The first part of that is actually pretty simple. Many people have shown you can overcome it. The key to that is just knocking out the T-cell receptor. That works. We're over 99% pure in our product. Host versus graft has been the real challenge. The reason that that is, is like any drug, area under the curve matters. If the patient's immune system kills the CAR T-cell, there just isn't enough drug product long enough to work. What has been established in the autologous CAR T-cell space in the, let me just take a step back. What's been established by our drug in the allogeneic CAR T-cell space in the oncology setting? We have a CD19 targeted drug. We started in oncology. Number one, the immune system does not see it, just like we talked about before. Number two, we get, and I want to talk about what's important here. We have a very nice safety profile. We get a dose-dependent B-cell depletion. The goal in the autoimmune setting, what we've seen from the data, and particularly from Germany, is quite clear. That is in the right patient population, a single dose of an autologous CAR T-cell can kind of do control alt-delete for the immune system. You are going to knock out the B-cell repertoire. What you are going to see come back are naive B-cells that no longer have these autoimmune antibodies being made. What we need to see is that we can see what we saw in oncology, that we get safe deep B-cell depletion. We want to see that translate into clinical benefit. That is basically what we are looking for. My take of what you really want, you just want the cells to stick around long enough to hit control alt-delete. You need to hit control alt-delete. One of the things people will look at is peripheral B-cells. How many B-cells are circulating in the blood? We each have about a billion B-cells in our bloodstream. We have about 200 billion in our body. It is a very poor marker of the problem. The reason CAR T-cells are such a great treatment in this area is, number one, they have a big, vast bio-distribution. They go into all your tissues. They can find B-cells no matter where they are and get rid of them. Two, they grow. It is very difficult to get a drug product on board that will kill 200 billion cells or something like that. You get this logarithmic increase of cells with the CAR T-cell. That is why this really works so well. What do we want to see? I want to see the most important biomarker, I think, in this field is that you see B-cells go away. When they come back, they are all naive B-cells. If they come back and they are memory cells or things like that, and you see this sometimes, the disease will recur. I mean, unless you got super lucky, it's going to recur. You really want to see this that you do not just want to turn off your computer. You want control alt delete. You want to reset it. It is not just sleeping. That is the key. What we are looking for here is to see in a high percentage of patients that we get a complete B-cell reset. We want to see that translate into a clinical benefit as measured in a lupus setting by things like a lowering of the SLEDAI score. You want to see the patient off of drugs. Now, I think you could be fooled and see a patient off of drugs or have a SLEDAI score go down just from the lymphodepleting chemotherapy that people get before a CAR T. You will not see the B-cell reset, though. That will give you a sense of how well we're doing, the difference between I would call an Ok, good, or great drug. I'm pretty confident we'll have an Ok drug based on the cancer data and things. We want to know, is it Ok, good, or great, though? That requires we get that complete B-cell reset in a high percentage of people. Again, patients often off all drugs and that they're doing really well in the clinical biomarkers with the SLEDAI score and things like that. I mean, obviously, the dream is that you really are curing, that you end up curing a lot of patients. If it turns out to be they do relapse eventually, but it's some period of time down the road, what do you think that minimum kind of profile is to justify the upfront chemo, the cell therapy? Granted, yours is less involved than an autologous product. Your question is the right one. I mean, the limiter here is the lymphodepleting chemotherapy. If it were not for that, if you had to treat once a year, it would be no big deal. With that, or every six months, I mean, that would be no big deal even. With that, I think you have to have many years. I do not know what that year is. If it is three, if it is five. It is not one. It is not enough. I mean, I do not think patients will go for that. Is that for the typical patient? Or is it enough if you have a tail? I think it all depends on the size of the tail and the patient population. We'll have to see what the safety profile looks like over time. It is a competitive market. The other thing is, don't forget, I mean, there's just like you've got T-cell engagers. You've got natural killer cells. You've got autologous CAR T-cells. You've got allogeneic CAR T-cells. We're going to tell you about in vivo CAR T-cells where you don't have any lymphodepletion and things like that. As all of those things evolve, the answer to your question is dependent upon how the field evolves. There is no question that the field's evolving in a way that that bar is going up, not down. We have to have enough that we clear the bar and we can skate to where the puck will be in more than one year from now. I mean, if you do have kind of the typical patient receiving it, getting a few years, several years of durable remission, what level of kind of the CRS, ICANS that may happen acutely? What, what? CRS, ICANS, the. Oh, again, in all. What is acceptable? You can't have much CRS or ICANS in this patient population. I'll start with the fact if you look in the oncology setting today, in the autologous, just generally, most patients will die without seeing a CAR T-cell. Many patients would benefit from a CAR T-cell. Why are they not getting a CAR T-cell? In general, if you go into the large academic centers, you're not limited by access, by manufacturing. You're limited by number of beds. When you're giving these drugs in hospital centers, you're losing money, and they're complicated to give. If you can get out of the CRS and ICANS problems and give this as an outpatient, it becomes a much more attractive drug for people over time. You're not going to want much of that. ICANS in particular. I mean, if you have a Grade 1 CRS, it's a fever. Let's not forget that. You can deal with a patient can deal with a fever. Grade 3 ICANS, where you're really not thinking straight for a while, or you have a seizure or things like those are bad. Those are things we have to avoid. In terms of your trial, it's been enrolling. It allows both lupus patients. We're a decent amount of kind of proof of concept is out there for autologous. It also allows the vasculitis patients in too. Can you just speak to where the enrollment has been? Are you seeing roughly equal mix? We just opened the vasculitis cohort. It is not that long ago. It has basically been lupus patients to date. Ok. What's kind of triggering to showing data? What's the data set need to justify? I'm pretty sure I can show you based on the oncology data, we can safely deplete B-cells. I want you to know if it's an Ok or good or great drug. I'm not too worried about this thing works or not. I'm worried about does it work well enough to justify continuing to invest meaningfully in this drug? Does it just work well enough for a partner to invest in meaningfully in this drug? Does it not work well enough? We should just not keep going here because we've got a lot of other things we could be doing. That, to me, is the key. I mean, that's how I think about it. Ok. You brought up the idea of a partner helping you. If you see early signs that this really is working in lupus and/or vasculitis, does this naturally need a partner to kind of expand out across the opportunity set? I mean, I think generally, we do not 100% will and rise to all of our assets. That will not continue. We will not be able to develop four assets through commercialization on our own. We will either have to focus the company or have partnerships. There is no question about that. I think that generally, it is hard. The type 1 diabetes space, as an example, is a space where the challenges are things that our company is uniquely set up to grapple with. If you move into B-cell mediated autoimmune disorders, the day we market, the day we partner that drug, it becomes more valuable. Because it is a land grab. There are so many different places that you could take a drug like this. We are a company that is both capital and capacity constrained. We just do not have the same global footprint. We don't have the same disease expertise as a number of other companies would have. Yes, it makes sense for that drug to be partnered. It's one that would be more partnered. It would be more valuable the day we partner it. Yep. Ok. Unfortunately, we're out of time. We're going to have to cut it off there. Thanks a lot, Steve, as well as everybody in the room. Thank you, everybody. Thank you, Marc.