Sana Biotechnology, Inc. ($SANA)

Session with Sana Biotechnology. My name is Alec Stranahan. I cover SMid Biotech at Bank of America, and I'm also the analyst covering Sana. Pleased to be joined today by Steve Harr, President and CEO of Sana. Thanks for being here, Steve.

Thank you for having us.

Yes, I appreciate it. Looking forward to the conversation.

It's actually a dry hot day in Vegas.

Exactly. I guess maybe to start, Steve, I guess, significantly tightened Sana's focus over the past year. How would you sort of describe the company's identity today? Is it primarily a type 1 diabetes company? Is it in vivo CAR T company? Or is it -- is there a common thread maybe underlying?

As you know, we'll make forward-looking statements, so people can take a look at our 10-Q, which was filed last night actually for some really up-to-date risk factors. So Anyway, thank you for having us. And so the I guess what I would say is it's no different than when we started the company. And our goal is to engineer cells, right? And there are 2 different ways that you can engineer cells. One is you do it outside the body. And your goal in that scenario generally is to replace cells that are either missing or damaged, right, and transplant them in and hopefully can replace them. The second is we can engineer them in vivo or inside the person. And there, the goal is to repair a damaged cell or to change the signaling so it can have a different function. And so we continue to do both of those. And as all or many novel biology and things are, it proved to be very challenging, and it's taken us some time to get to where we are. And to your point, I do think we have a clear focus right now. I would say for good or for bad, and as an investor that might be for good, as an operator might be for bad, these are very, very correlated risks, right? And so within type 1 diabetes, this is a really big market. I think it's sometimes easy to forget or underestimate the problem and scale of type 1 diabetes. And to give you a sense of the unmet need, actually was looking this up earlier, I have a 22-year-old daughter. And if your 22-year-old daughter could be diagnosed with breast cancer, HIV or type 1 diabetes, the shortest life expectancy is actually type 1 diabetes, right? And of those, obviously, the day-to-day management of that in terms of like every meal, every activity you do every time you get sick. Am I going to dinner with Brian, who sometimes shows up late, Mickey who's a slow eater, how do I have to modify what I actually -- how much insulin I take. Those things are in their decision-making every day. And the second is it's 10 million people, and people often ask us about competition, and we may go there, I don't know. But I always just say I don't worry about it. That's not being arrogant. It's being -- let's just assume like the greatest thing possible happens, and that is somehow we -- this works, and it works perfectly. It works one treatment and a person never has to take another therapy. It works 100% of the time. And we scale it, and we're curing 100,000 people a year of type 1 diabetes. We'll take the global growth rate from 5% to 4%. It is so much -- if we just launched in the United States, we take the U.S. growth rate from 4% to a negative 1%, right? So it's a large enough market that you have to assume that others will play around. There will be other ways to solve this problem. And I think the more successful all of us are, the better it is. The in vivo CAR T is a totally different risk. It can we -- there's a much more vibrant and dynamic competitive environment we have to go and deal with. And that's something where speed, breadth of clinical development and our overall profile is going to be really important in figuring out where do we fit in as hopefully the best answer for patients. So I'll pause there for my introduction, but that's literally what we're up to.

No, that's great. Maybe we can start with type 1 diabetes. We saw really kind of groundbreaking proof of concept. Now you've got follow-up in that patient up to month 14. I guess what would it take or is it already happening in terms of number of patients, duration of follow-up to really have full conviction in the program before really scaling manufacturing at?

I'd separate those 2. So I want to take a step back. Type 1 diabetes, we know what its problem is, right? It's the immune system comes in and it kills the pancreatic beta cell. The pancreatic beta cell in the patient is the only cell in the body that makes insulin. So up until 100 years ago, that was a death sentence, right? You couldn't get sugar from the bloodstream into the patient's cells. So they start to death ironically, right? Even the sugars are sky high. And there is -- in 1923, there's the advent of exogenous insulin. And people have done okay. They've done pretty well, right, but not great. About 25 years ago, a scientist in Canada, James Shapiro discovered that you could take pancreas from someone who just died, isolate the islet. And I'm going to go back and forth. So just I'll pause here for a second. Pancreatic beta cells are the cells that make insulin. Think of an islet as a beta cell and a support structure. It's the easiest way to think about it. And so to isolate islets and transplant them into patients. And he found that if they got enough islets, they could be insulin-free for a long, long, long time and do very well. But they had to be on lifelong immunosuppression like any organ transplant. So that's not good. There aren't that many people who are lifelong insulin is better or worse than lifelong immunosuppression, right? And getting cells from a cadaver is neither scalable nor replicable, right, it's very, very low quality. So over the last few years, but that's a huge proof of concept. That's step one. Step 2, over the last few years, we've seen several parties take stem cells, grow them and then differentiate them in islets, right, and transplant them. So now you have a much more replicable supply source and it's almost certainly meaningfully more scalable, right? But you still have the problem of immunosuppression. And so what this study did that you mentioned is the first example that I'm aware of where we took gene-modified cells for a patient who was a -- recently deceased donor, right? And we took his pancreas, isolated and gene modified them and transplanted them into a person who had type 1 diabetes since 1987. So this person is now for the last 14 months been making insulin for the first time in over 40 years, and he's doing it with no immunosuppression. So we now know that you can get 3 things, right, long-term transplants to your patients. Second thing, you can make them from stem cells. Third thing, if you put the right gene edits in them, you can no longer -- you can transplant without the need for immunosuppression. So now I would argue a cure is inevitable. Like someone has to take and put the whole thing together, and that's what we're doing with SC451, our drug. So hopefully, we'll know the answer to that within a year. Our goal is to start the study this year and figure things out pretty quickly. But I would argue that the cure is inevitable, just someone has to put all that together. So to your question, what do we need to know for manufacturing, we haven't -- I would think about manufacturing as 3 phases for us. One is good enough for Phase I. We're finishing -- like we're in the midst of making this stuff, right, the tech transfer into a CMO. It's good enough for Phase I, but not much more than that, just to be clear. The second is good enough for a really nice commercial launch. I think of that as thousands and thousands of people per year. We have work to do on that. And it's not an investment number. That's a -- I'll come back to that. And the third is tens and tens of thousands of people per year what we talked about earlier, and that's going to take us some time to get to. That middle one, we're already investing in. But that's -- again, I'm going to divide this into 2 problems, cells per run, number of runs, right? That's how many cells you get a year. Cells per run is a science problem. Number of runs the capital problem, right? So right now, we're on the science problem. And we are going at that full steam. We didn't start doing that, though, until we knew we had the Phase I process done. But right now, we're going full steam with that problem. I'm more optimistic than I was 3 to 6 months ago that we'll get through that. And ultimately, it will be a capital problem. But first, we have to make some science advances to go to. That's how I think about that in terms of answering your question.

No, that's helpful. And I guess in terms of the IND filing that you're working on for 451, is the last gating steps just sort of the GLP tox? Or is it the tech transfer to the CMO that you mentioned? And I guess, how close are those 2 to completion? What are sort of the best estimates for when first patient dosing happens?

I hope they're pretty close, right? Which one ends up being a rate limiter is you never really know until they're done. I would bet you it's probably getting manufacturing all buttoned up. It's a complicated drug to make. I don't want to kind of sugarcoat that. And our Phase I process is a Phase I process. It's not a commercial process. There are a lot of things we still need to automate, scale and clean up. But I think it will be good enough, but we're in the middle of that tech transfer. Things could always go a little better than we planned. That rarely happens. Things can always take a little bit longer than we plan. I think we have some buffer build in our time lines, but that's probably going to take us the longest amount of time. We are also finishing up the nonclinical testing, right, which is a whole host of things related to GLP tox, efficacy, genomic stability, a whole host of things. Again, I think those should go -- those are getting pretty close to being wrapped up. They're not done until they're wrapped up, something could always surprise you, right? And can end up being the long pole in the tent. But for right now, we seem to be on pace with both of them to kind of meet our goal. What we said is we will both hope if the things go as we hope they will, we will both clear our IND this year and get the study going. Exactly when we'll have data will lead to another time to figure out.

Yes. Yes. I think you've noted that proof of concept could come pretty quickly, maybe within like weeks of dosing, whether the cells engraft, whether they evade rejection, produce insulin, et cetera. I guess what are sort of the data points you hope to gain initially to know that the new drug product that you've made is doing what you hoped in replicating.

So -- I mean, just approach this, first of all, like from the most important thing in a first-in-human study with a very -- this is CRISPR gene-edited stem cell-derived islets, right? It's a combination of gene editing, stem cell biology, immunology, it's a novel way of delivering these things. So the first thing is safety we need to get through. But that's -- I know that's not what you're talking about. But once we have safety, if you transplant -- we're going to transplant these cells with no immunosuppression. This is a -- we're transplanting the cell the patient already has a known immune response to, right? It's already killed all their own beta cells. And we're transplanting the allogeneic cell. I mean it's coming from someone else. It should be gone within days, right? And if it's not, and it's there -- just like with that previous proof of concept, we said if it's there at a month, it's going to be there in a year, right? So I think it is there at a month, we're going to feel really good that in this new way of manufacturing this, we've overcome transplant, allogeneic and autoimmune rejection. The next thing is it was a really low dose that we transplanted, right? So I want to -- we don't want to just see a really low dose of -- so take a step back. When your beta cell makes insulin, it actually makes something called pro-insulin. And then when it's secreted from the cell, that's cleaved in a C-peptide and insulin. So the amount of C-peptide in your blood is a direct measurement of how much insulin the patient is making, right? And so we want to see that the C-peptide at 1 month is very robust, right? It's on its way much higher than what it was. And that gives us a sense this is likely going to be that something works. So that's step one. That can happen very quickly. Step 2 is we're actually not trying to overcome allogeneic and autoimmune rejection, right? We're trying to make a drug product where a patient will be able to come off insulin, have normal blood glucose and never be on an immunosuppression, right? So we want to see that. And that likely takes a quarter or 2, right? If you look at -- that usually takes a while to kind of get these cells to really kind of function really well afterwards. So let's say, there's the first step, which is can you transplant them successfully. Second is you get people off insulin. And then the third is how replicable is that, right? And let's just say we're 5 out of 5 or 6 out of 6, you're going to feel really pretty good about it, right? Let's say we're 3 out of 6. You're probably going to say I want to see some more, right? And so exactly how long that takes, we'll have to see. But again, those are the types of things I think we'll figure out next year as we move through the year. So I'd expect pretty early, we'll figure out, hey, do these cells overcome immune rejection and function. And maybe as we move through the middle of the year, this time of the year, next year, we start to understand, have we truly gotten -- and we're going to get people off insulin, right? Do we have the right dose, all that stuff. And then the latter part will be later in the year most likely.

Okay. And when it comes to 451, you've got the things that the company can control, right, on the manufacturing and the lot-to-lot reproducibility. And then you've got the clinical handling at the hospital, right? How does the Mayo partnership that you signed recently sort of establish a good foundation?

That's a super important question. Like one of the things that's concerned us is what happens from the time the drug product leaves our hands to when the patient goes home, right? There's so much that has to happen in that period around these are live cells, right? They have a very short window that they are actually going to live. They have to be taken from their shipping container and prepared for transplant. They then have to actually be transplanted with a procedure that is replicable across many, many surgeons across many, many sites over time, right? They have to standardize how you take care of those patients so that afterwards, we can be confident that they will safely go home and hopefully function quite well. So that's really to start with, I hope we get a lot out of this collaboration. They've been wonderful so far. We're only a few weeks in, but I actually have to say it might be the honeymoon phase, but the first few weeks are better than I hoped. But we're just learning a lot about that portion of it. How do you ensure that we have something that is standardizable, that is safe, that will deliver this product in a predictable way to patients. And that's what it's for. And so far, so good.

Okay.

They're really good at that stuff.

Yes. Yes. Yes. No, that makes a lot of sense. And I guess you alluded to questions that you get around the competitive landscape. I agree like it's such a large TAM. It's kind of a moot point. But Vertex, I think CRISPR has had something in the works for a little while anyway. I guess how do you sort of see fast followers given that you've now kind of proven the mechanism?

Well, I think that we can't claim they're fast followers until our actual drug really works. So let's -- we'll first get there. But again, I think it's -- I would like to be first in what we do. I think there's some real value in that. I also think there's a lot of value from learning from the field. And there's no doubt we've learned from the field. I'll give you an example is one of the things that Vertex saw early in their trials is stem cells sometimes don't engraft perfectly, they die and they release insulin granules. And some people ended up with low blood sugar. Super easy to treat, give them sugar, right? So now we know that, that's one of the things they have to monitor from. And I hope that they can learn from us some things that we do and learn as we progress as well. And our goal here is to make progress for the field. And I'm sure that if we -- if our stuff works, I'm sure we'll have -- we'll do fine, and I'm very confident there will be competitors.

And there's a lot of know-how when it comes to actually making...

Ton of know-how.

And you're actually building that out from scratch.

Well, there's a lot -- like -- yes, it took us years to make -- to actually define this problem. And that is these stem cells are a little bit genomically unstable. When you gene edit and become even more unstable and you end up selecting for, if you're not very, very careful, mutations that you would kind of select for cells that grow quickly. That's called cancer, right? And so that's the know-how it took us to get around. Defining the problem though, I think others now can run more rapidly behind us and say, okay, we have to go after that challenge really early. They know it's there, right? We're pretty transparent about where they can be. We have to be because we're pretty small and this has -- if things had gone perfect, we would have treated the first patient a while ago, right? And we learned, right? It took us some time to make that -- the science really work.

Yes. Yes. Maybe just for the sake of time, I want to talk about the in vivo CAR T platform. This is really the emerging part of the fusogen platform that you guys have built the company off of. Maybe we can talk about 293. I don't know if you want to give a quick sort of intro on that, but whether it's the B-cell depletion you're seeing in monkeys or anything else that you're seeing coming out of that program and sort of what you hope to replicate?

Let me start with there's this in vivo CAR T space. And there are 2 broad sets of technologies and then within one that we're in, they're like cousins, right? And there were 2 very, very essential assumptions we made at the outset of this, right? And that is that you're going to try to make -- so the idea of these in vivo CAR T cells is you're going to deliver some type of genetic material to a T cell, they're going to make it into a CAR T cell that can go and attack some target cell, it's a cancer cell or a B cell or whatever you want. And we made 2 critical assumptions. One is that cell specificity and delivery really matters. You only want to go to T cell. There's an alternative viewpoint, which is you just need to get enough T cells to have efficacy, right? And don't worry about the off-target stuff -- they'll take care of themselves. The second is we thought -- we feel like because you're going to make, I don't know, 100 million CAR T cells and you have tens of billions or hundreds of billions of cells you have to eliminate, you have to get the DNA to integrate into the T cell so that you can get what we see with CAR T cells is a logarithmic expansion and grow. Like every time they see a target cell, they kill it and they divide. You go from 1 to 2 to 4 to 8, you just keep dividing to have many, many cells that will kill. So those are the 2 assumptions. If we're wrong, other technologies will end up being faster and easier, to be very clear, right? And so we believe we're right, but we really believe cell specificity matters and you have to integrate. That's part one. That's part of the competitive landscape. With what we've seen, I think that the big differences that we have versus others in this kind of like virus-like particle space is we're more specific, at least in preclinical settings. And I think we have a different way of entering the cell. Both of those should end up if preclinical science is right, being safety advantages, right? It may prove that that's easy to see in a clinic. It may prove to be really hard in the clinic. It may prove that we have some other problem, right? And so you dose a patient, you don't know. But what we see in the nonhuman primate, which is a really good model, right, of efficacy. So what we're trying to do is we give them a single injection of our drug and it makes a CAR T cell that will target B cells in the monkey. And we see a dose-dependent complete elimination of detectable B cells, right? And that's true. And then whether we look at peripheral B cells, lymph nodes, spleen, and then you see this thing that you have seen with George Schett's data in the autoimmune setting where we have this B cell reset. I mean it's all naive cells that come back. So it's kind of control the lead on that B cell repertoire. And that's something we haven't really seen from others, right, where it's just clearly just naive cells coming back. And we also seem to have a very good toxicity profile. So it's very cells specific. You don't find this in the liver. You just don't find it. I challenge you to look at any other data and see that. You don't find it good out of tissue. You don't find it in the heart. You don't find it in the lung. It is just in T cells, right? The second is there's been this kind of infusion reaction that many have seen in the field that's been pretty challenging. And we don't seem to have that at least in nonhuman primates. So a little bit of a fever. One dose of tylenol was all the animal needed and did very well. And so we think we'll have a differentiated drug on safety, maybe on potency and efficacy because we can have a better safety profile. But these are -- this is a novel category of drugs, and we need to get into humans and see what it looks like, and we'll start doing that this year. So we're optimistic, but we have to see what happens.

Okay. Obviously, this is a pretty hot space. We've seen large pharmas stepping in through M&A. I guess, versus some others that are in the field, would you say it's really the specificity of the edit that's the driver.

It depends. I mean, LNP mRNA, 2 things, right? One is integration, second is specificity. The cells of the delivery, not the edit, the specificity of delivery. So those things that are going to be in the liver, it's just mRNA. When you do the cell divides, mRNA won't go into the progeny, right? And so you only have a limited -- you can redose it, but you have a limited number of cells that can kill. It's unlikely that's as potent. But maybe good enough. I mean, I don't think we know, right? We see that. That hasn't really been tested in humans yet in a really robust way. The second is -- but those -- there have been a number of acquisitions in that space of and off of preclinical data, right? The second is you have these virus-like particles. And there, you integrate your signal into the target cells DNA, right? And there, the difference in what we have versus others, others have had human data when they've done these partnerships and acquisitions. And we need to get that. And I think that will help us understand what we have. And from there, we'll figure out what the right path forward is for us to develop this drug.

Okay. And I guess in terms of gating the go forward, obviously, we'll see how the activity of 293 looks in people. You recently nominated a BCMA directed asset to 227 that could enter clinical study as early as maybe middle of next year. Is that -- is pushing that asset forward sort of dependent on 293? Or are they separate?

Yes. If it turned out that we deliver this -- our delivery vehicle is not safe. we're not going to do 2. If it turns out that our delivery vehicle does not work, we're not going to do 2. If it turns out that it works, we're ready to exploit that opportunity with taking forward this drug 293, which targets CD19 in oncology. We're going to take it forward in autoimmune diseases, and we'll be ready to take forward at BCMA not long after that. And so it's really more ensuring that we're prepared. It's not -- getting prepared is not that expensive, actually executing is expensive. And so -- and both in terms of opportunity cost for patients, but also capital for us. And so if it turns out that stuff isn't safe or effective, we won't take it forward. It's the second drug, right? But if it is, we'll be ready.

Okay. And I guess the initial step is in oncology, autoimmune, you mentioned. Is that something -- I guess it depends on sort of the activity you're seeing, but would you want to invest to push it forward just broadly across both of those verticals yourselves? Or would you seek to partner? And I guess, which indications?

Yes. When you look at -- take the broad category of B-cell depletion, it's hugely competitive, right? You have -- almost every large company has got several mechanisms. They got pills, they've got antibodies. They've got antibody drug conjugates. They've got T cell engagers. They've got CAR T cells. They've got in vivo CAR T cells. They've got all these different ways to go about this. And therefore, being really a little bit broader in our clinical development and getting -- rapidly finding the place where our technology is the best answer for the patient is really important. It's likely going to happen more rapidly and more broadly with a partner than it will alone, right, particularly when we think assuming success that this type 1 diabetes is kind of a generational opportunity. So would we partner that? I think that in most scenarios, if we can find a good partner, it will be partnered. It will end up being that the pie is so much larger with a partner that you can divide some of the economics, right? And when you don't grow the pie, that's hard to partner. In some partnerships, you shrink the pie, right, because you just make decision-making so complicated. But I think that's when the pie grows. And so it seems like the CAR T assets broadly will be things that over time, again, we can't control when we get to the right arrangement. But over time, we will be better off in partnership than trying to do them ourselves. Or type 1 diabetes, I don't know the answer to that. I think of like it's very simple. Does it work? Drugs that will be made. So kind of that question caps out of the back, right? Can you scale it? If it works, can you scale it? It's a totally different technology, so unclear, others can help us. If you can scale it, can you commercialize a curative therapy in a truly broad market, again, something that's going to take some real novel creativity, right? And then the fourth is how do you pay for all that? Very clearly, pharma can help us with the latter. We have to think through the first 3. And if we feel like it's enough of it and they help with #4, we may end up with a partnership. If not, I think it will be really valuable and hold on to for cells for a while.

Yes. Yes. Kind of champagne problems once you see the...

The activity that -- first the assay doesn't work.

All right. Well, with that, I think we'll leave it there since we're out of time. But Steve, thank you so much for the great conversation, and thanks, everyone, for attending.

Yes. Thanks, everybody, for their time and attention, and thank you, Alec.

Thank you.