Excellent CEOs from cell therapy companies across the United States. Pascal Touchon from Atara, Nick Leschly at 2seventy bio, and André Choulika from Cellectis. Welcome, everyone. Thank you so much for joining. Maybe just to start, if each of you could spend two or three minutes just giving a high-level overview of your company and what are the strengths and differentiating qualities for each of your cell therapy platforms, and then we can go into more specifics after that. You want to start, Pascal? My pleasure. Thank you for inviting us to this panel. Atara Biotherapeutics is a leading company in the field of Allogeneic cell therapy, developing transformative therapies for patients in need in both oncology and autoimmune disease. In fact, we are leveraging the unique biology of Epstein-Barr Virus (EBV)T-cells to be able to address either EBV associated disease like some cancer and some autoimmune disease like MS, or with the addition of chimeric antigen receptors, then target different solid and liquid tumors. We are arguably the most advanced Allogeneic T-cell therapy in the sense that our first product, Tab-cel, is in final stage of regulatory review in Europe and should be approved maybe by the end of this year. We also have a unique differentiation in the sense that we have already treated more than 500 patients with our Allogeneic T-cell therapy with really clear efficacy in different type of disease as well as favorable safety. A very significant type of clinical experience with a scaled platform that is really gearing towards the first approval very soon. Thanks for having us. Nick Leschly with a newly formed company, 2seventy bio, which we're excited about, a natural extension out of bluebird bio with an exclusive focus on oncology. The whole idea with that was to say, based on the last 10, 12 years in Autologous cell therapy, how can we focus in and deliver more, really to unleash the curative potential of the T-cell, to really get after doing that through various different strategies. It's based on one, our commercial product, which is Abecma in multiple myeloma in partnership with BMS, which is doing wonderful things at this point for myeloma patients. We announced earlier today that we're continuing development in earlier lines, launching a frontline study as well. That's exciting. Can talk more about that, but also more importantly, frankly, over the next five years, how do you make Autologous therapy even better? How do you layer different technologies in to make it that much smarter as it relates to liquid tumors and solid tumors? That's where we have a couple different efforts in AML and NHL, as well as in partnership with Regeneron on the solid tumor side. We'll get into that. The whole idea, we don't think CARs or TCRs are really enough. How do we think about taking T-cell biology and engineering, combining it with extrinsic agents to really get after cancer in a more fundamental way? That's the hope and dream for the next five years. Okay. Cellectis is a cell and gene therapy company. Initially we started, and we are, the core of the company is gene editing. Like 10 years ago, we decided to start focusing gene editing to develop a new type of therapy, which were Allogeneic CAR-Ts. Cellectis started as one of the first company—the first company developing the concept of doing off-the-shelf gene-edited cell therapies for cancer. Like, the company has been focusing a series of different type of indications and product. The first product we've been developing, which is, which was named UCART19 at this time targeting CD19, is now, being licensed to, Allogene Therapeutics and different type of indication, and especially non-Hodgkin's lymphoma that will enter very soon pivotal. After this was BCMA, and then Cellectis is currently developing a series of products in different types of liquid cancers such as UCART22 in acute lymphoblastic leukemia, UCART123 in acute myeloid leukemia, UCARTCS1 in multiple myeloma. Last but not least, we recently filed an IND and got, like, that was granted by the FDA for the development of the first new dual CAR-T targeting CD20 and CD22 together for non-Hodgkin's lymphoma. A CAR-T that we've been manufacturing from A to Z internally in our manufacturing facilities. We're probably going to dose the first patient, I hope, in the coming months in order to open all the sites for this. We're excited about that because, you know, most of the CARs that you see currently are focusing on CD19, and there is so much pressure on this target because it's a validated target by a lot of the companies and trials that are done around, that there is a strong need for people that relapse after CD19. We're excited by the products that Cellectis has been developing that are very creative and efficacious product, we hope. Okay, we'll get into some more company-specific questions down the road. But just to sort of spark the discussion a little bit, just a high-level question, thinking about Autologous versus Allogeneic. We have both strategies. Both are interesting, both are valid. Tell us a little bit about, you know, where do you see the field going? Do you think it's gonna eventually move to Allo, or will there always be a place for Autologous in the treatment landscape for blood disorders and potentially even solid tumors? What are your thoughts? Just it'd be interesting to hear. Should we all start talking at the same time? Let's jump. [crosstalk] You're flanking me here with, I have to go first, and then I get a right punch and a left punch. You may be surprised. I don't think it's nearly as sexy as the question, sort of suggests. Ok Look, I think Allo, there are challenges with Allo and there are upsides with Allo, and over time it will find its right place. It will find patients that it will help, etc. I think Auto's already found its place where it'll help, and it'll continue to get better at the things it's not good at, right? Right now it is more complicated from a manufacturing point of view, etc., but it's also getting a lot better. Now, our focus is we're gonna sort out our respective manufacturing strategies and our scaling capabilities, I think on both sides. It's really ultimately sort of like Gretzky saying, "Don't skate to where the puck is," right? Like, where's it going? Where it's going is who's gonna be the most impactful in cancer and have the best safety profile associated with that? That's been our emphasis. I tend to say, "Listen, I'm actually rooting for Allo, I'm rooting for Auto, I'm rooting for any company in our space 'cause this is hard, right?" It's important. It's important because if any of us got any of these diseases, we don't really care if it's Allo or if it's Auto or if it's base editing or if it's that or if it's that, right? In that sense, I'm a little bit more co-opetition here. We can all learn from each other, and we are all learning from each other, and there are different diseases that are gonna require different things. That sounds awfully like Danish of me, right? In the sense that I'm sort of playing it right down the middle. I will say that as it relates to Auto, we think, "Look, we know it works." We do fundamentally believe it's easier to make something that works better. We don't wanna change that platform for what we're working on right now. To say what we do wanna change is what are the things we're adding to it. Hence, in the case of NHL, what are the problems we're trying to solve? I'm gonna keep some of the biological and manufacturing things consistent 'cause I know I can get it there, I know I can manufacture it. Now the question is how do I beat the issues that are sitting there with NHL, right? Some of the relapse on the CD19. Okay, we're gonna go after CD79A as well as CD20, and then we're gonna play with the cell a little bit more to sort of release the brakes on that to say we think we can then make it even better. That's been the mantra, how do we make it better and continue to make it better? Same with AML, right? It's CD33 is a good target, but it has safety issues. Well, what are you gonna do? How do you release the brakes? How do you pulse it, right? That's where our DARIC platform comes in. We're gonna continue to build and layer with these different tools to say how do we ultimately become an engineering company that sits on top of a T-cell based platform. Then over time, all of us need to know how to engineer. I don't care if it's an Allo cell or it's an Auto cell. Right? You need to know how to attack the cancer in a fundamental way. That's where I think ultimately that learning, those insights, that I wouldn't say is the winner, but that's where you consistently can innovate and outpace cancer. All right, your turn. Yeah. No, I mean, as you know, I was leading the efforts to get the first-ever CAR-T approved and launched in the world. Right. I know that Autologous CAR-T works. Yeah. For sure. I mean, Emily Whitehead just celebrated her 10 years cancer-free recently, and that's a great progress. I think that is amazing. That's why we're here, is to change the lives, transform the lives, to cure these patients. Auto works. I think Allo is really the next step, especially for liquid tumors right now, where we have now found the ways to develop Allogeneic CAR-Ts that are safe and effective. More and more, and especially what we're trying to do, and like many others, to increase persistence of the T-cells, you know, to have this durability of response. That is clearly the next frontier there in terms of the Allogeneic CAR-Ts. Now, solid tumors is another task to address that is very challenging. We are making some activities there. I mean, others are making as well. We have not yet find the right type of design and construct and combination strategy that allows to address solid tumors. We will get there. I'm very confident we'll get there. Now, whether it will be auto first, Allo first, I don't know. I think we need all of these modalities. Clearly Allogeneic CAR-Ts have this advantage of being easier to access patients in general in terms of convenience, time, cost, as well as some benefits of some construct that we are making. For example, in Atara, we're trying to make not only an Allogeneic CAR-T, but an Allogeneic CAR-T with next gen CAR technologies that are going to improve functional persistence and other features that are essential for liquid and solid tumors. That's right. Maybe I'm going to take a different type of angle that might be less sexy, but it's fine. Let's see how many times we can say the word sexy. That'd be fun. It's definitely, I think, part of what. Like, we're an industry. We're not hospitals. We're not physicians that treat patients. We're biotech, and we're supposed to be private companies that are supposed to make money at the end. That's one of the basis of when you take it back, this is the thing. When you look, for example, at the data of CAR-Ts, such as Yescarta, like Gilead is giving $1.5 billion target sales for this year, which is great because it shows that these products, even autologous, can become blockbusters, and that's amazing. However, they have gross margin that are extremely bad, low for the industry, where you have a threshold around like 85%+. Usually people look more into 90%-95%. When you look at the data of ALLO-501A, last one that were presented by Allogene, which is a CAR that we licensed out to them, then you see that the data matches, if not sometimes better, or even with durability that is quite okay with this, with these A utologous CAR-T, with the Allogeneic CAR-T. With the market access, distribution, and performance and gross margin, etc., why on earth the physician that sees a patient that comes in that have non-Hodgkin's lymphoma or whatever indication it is that have two option, either go for an apheresis or just deliver a product that sits in liquid nitrogen somewhere in the hospital and can give an immediate response to it with almost the same performance, if not better sometimes. I think the Allogeneic is going to be an overwhelming wave at the time of commercialization. However, is Autologous going to disappear? That was the basis of your question. The answer is no, because sometimes you will need to go for an Autologous therapy because Allogeneic for XYZ reason didn't work. Mm-hmm. I think you just need lenti, that is your product, and a service that the hospital can provide. When you look at the origin of Autologous CAR-Ts, the origin has been hospitals, not biotechs. While Allogeneic is a real industrial product at the end. Well said. This is not an Allo versus auto question. This is sort of an Allo and Auto CAR-T question versus some of the other cell therapy modalities. You know, we have other approaches, NK cells, TILS are good examples. What are your thoughts about where those intersect? When would they be appropriate? When would they not be appropriate? What are the advantages of doing auto or Allo CAR-T versus an NK cell or a TIL? When might an NK or a TIL be a better approach? I think, I mean, clearly the T-cells are the most effective weapon that the h uman biology has developed over millions of years, and animal biology before, to address some specific needs, in particular for cancer. To me, there is no doubt that the T-cells are better than any other type of cells. Now, the reason that others may be developing other type of cells is to address other type of aspects, such as the Allogeneic aspect. How do you make a safe, effective, and persistent Allogeneic CAR-T or TCR-T? That's another question. We believe that what we're doing at Atara is really aiming at having this safety and persistence in creating Allogeneic CAR-Ts based on EBV T- cells in a very unique way. Now, this aspect of safety, avoiding GvHD and persistence is very important aspect. In that case, why on earth would you want to go to NK cells? What's the point? We need to have enough persistence. NK cells are not persisting by definition. They are being eliminated within a few weeks from the body. You want these cells to persist longer than a few weeks to be able to ensure a durable remission there. Now, on the TCR aspect and TILS, it's probably slightly different because if you need an polyclonal or oligoclonal response based on having several TCR, it's more challenging today, but we are working on that to address that for Allogeneic TCRTs. I mean, some companies are doing autologous TCRTs. TILS are not really well selected just based on the patient type of tumor response. There will be a way, I believe, to have a clear oligoclonal TCR Allogeneic TCRT product. There will be still a place for autologous TCRT product as well that address cancer beyond the surface antigens that we are addressing with CAR-Ts. I still think we're sort of looking at this the wrong way, because I think the currency for what ultimately will play out, again, I come back to not the investor, I just come back to the patient sitting on the other end. What do they care about, right? What does the physician care about in that context is I need something that works as well as it possibly can work, right? I don't care about the margin, with all due respect, right? In a sense, I care, but I care more about having a product that works really well. I think there are things that we're gonna solve. We're gonna solve because it's just a matter of wanting to solve them. Those are things like some of the manufacturing and the scale-up and so forth that we're been hammering away at for years, and we're hammering away at right now. What I think is harder to solve is how to actually navigate to attack the cancer the way we want. You referenced a few of them, right? Solid tumors is a whole another ball of wax. Not having a persistent cell in that context, I think, is gonna be really difficult to build upon that, right? To be able to have the access and be able to deliver the tools you need to deliver. The other thing is, right, TCRs are good at certain things, have a wonderful signal, but you can't really. There aren't enough of them. You can't sort of get what you need to or tackle, whereas CARs have a wonderful recognition system. How do you merge the benefits of both these over time into next generation constructs and then start to build on top of that with? Well, how do the antibodies outside, the extrinsic antibodies, how does PD-1, how do others, how do we connect into the existing universe of treatment paradigms out there with our CAR- T therapy? It's a T-cell-based therapy, but you're still now connecting into the sort of the real world, if you will. That also, I think, is where you're going to ultimately be going. That's where I think the currency in the next five years is going to be, because some of this stuff we're talking about here, yeah, these targets work. Now the question is, how do you connect in and how do you actually layer? How do you improve those technologies? I think that's ultimately what is going to make a bigger difference than an Allo cell, NK cell, Auto cell. It's sort of how do you get to the efficacy side with safety being in tow? Well, TILs, though we have like a collaboration with Iovance, and like there is the first gene-edited TIL that, like was, granted, like an FDA approval for like moving into the clinic. Like the first IND approved, and that's like an exciting moment because I think TILs is definitely a great angle, to tackle solid tumors, and we've seen great data so far. I think like the strongest data on the TILs space, Iovance has been showing this in a very nice way. There is a huge potential for next gen in re-engineering these TILs and make them more and more efficient. It's definitely also an autologous product, that is exciting, I think. On the other side, the NK cells. NK cells so far have been giving some interesting results in certain indications, but with doses that are totally disconnected from what you can give with CAR-Ts. The ability to produce these NKs makes things complex at the end. It means the volume that is injected of these cells is usually two or three orders of magnitude above the volume that you can inject with T-cells, as results that are quite similar and makes things a bit complicated again, with potential type of rejection that you can have, because of the exposure of the CAR-T as an antigen in the patient with the potential to be forced to read those. The TCR-T, I agree with what Nick and Pascal just said. I think it's a type of like recognition between the TCR and the target cell. I think that also the combination between multiple type of technology, between the binders themselves, so like the TCR or the CAR and the gene editing or the different type of operations that they can do by gene editing in the cell can make these things way more potent in the future. I think the first steps are extremely promising. What you will have not in 10 years, but probably in two, three, four years from now, is something that is going to be a whole different space. I totally agree with that point, which is, as we're learning how each of our hammers and nails work, if you will, it's that convergence, whether it's editing or with antibodies. I think that's really hard to do, because in order to scale and run a study that says, "Listen, we're not smart enough to pick one arm, so I'm gonna pick, I'm gonna do four arms with X patients per arm," and then basically systematically kind of get smart with intrinsic sort of engineered cells in combination with or alone, right? To then because that I think ultimately be able to run enough of those experiments. In order to do that, one, you need to be smart enough to design those. You need to have access to both sides of that equation, engineering as well as access to therapeutic sort of other therapeutic modalities. You need to have the vector manufacturing, you need to have the drug product manufacturing, you need to have the money to be able to go after that. Ultimately, that's where we have to go. Now that we're learning how each of the components, these rifle shots, it's just too hard to know, right? Then you're 2-3 years out from your next question. We have to be able to ask and answer questions much faster now that we have what I'd say an underlying base of technologies. You can debate, I think bispecifics are going to come together with CARs and TCRs, right? Allo and Auto. I think antibodies of all sorts are going to come together with this. I think it's almost an obvious statement, right? What's worked in cancer? Combinations. This is now just fancy molecular combinations combined with other agents. It's a really hard study to do. We need the regulators to get on board with their master INDs, which they are doing now. It allows us to be able to manufacture cassettes as well as combination studies. Whoever gets the ability to do that soonest is going to be smarter faster. It could be the combination, could be also the way you design your cell therapy. I mean, we have in, as you know, in the clinic, this particular design where we have a CAR targeting mesothelin, but having a PD-1 dominant negative receptor built into the CAR. I think the idea is to go into that ability to address the tumor microenvironment in solid tumors in a way that allows these cells to be functionally persistent, not to be immediately exhausted, inhibited by the tumor microenvironment. That doesn't mean that it's going to be the product on its own. It might also be combined with other ways to address the tumor microenvironment. Both the combination as well as the addition of specific features to these cells, will allow to address solid tumors in very different way. Yes. Just to build up on what you said, are we on the same page on this? The thing is that it's too sophisticated and there's too many parameters. Protein design, type of cells that you're going to use, the potential combo that you can do potentially with a bispecific or something else, the gene edit or the little tweaks that you're going to do in the genome itself, maybe adding or removing, et cetera, some genes. Many parameters for a single product, you can. It's difficult to get it right from just the start, from the time you file IND. The thing that needs to evolve, I think you're saying the same thing, is probably to have maybe an evolution in the FDA in the way they're seeing these type of products, where you have a channel where you can do multiple type of little trial. You can try it on super small cohort, let's say four, five, six, 10 patient with a certain type of dosing. Then once you get it right and you think that the product is great, then you can pull this out into Phase 2, Phase 3, and registration. Keeping the same way to see the drug development of Phase 1, Phase 2, Phase 3, or the thing fails in Phase 2, and then it's over. It's never over. There is always something that can be changed if you start to see certain type of signals. Sometimes, for example, you add a safety switch, we've done this in the past and say, "Oh, like the safety switch respond to this type of drug that can be given in this way, so you have to remove it." New IND to be filed, etc. These make the things slightly complex and needs to be evolved. There is a need of discussion with the FDA in term of making these type of cell therapies evolving more quickly and with more innovation. You're suggesting some sort of a quote-unquote, "modular IND structure" where you could propose a series of potential strategies or edits or knock-ins, knockouts, and you wouldn't necessarily realize all those in the first trial, but they would be encompassed by. The FDA would give the blessing to that spectrum of possibilities. That's kind of where you're. It's already heading in that direction, and it's already happening in small bits, but it's not quite as what I'd say, engineered into the pathways of what this requires not only is the will. Mm-hmm. You need the intellect to be able to stitch that together, and then you need the money to be able to run that, and then you need the timeframe to do it. None of which the investors like a whole lot, right? 'Cause, "Hey, I just want to know what are you doing and when's it going to give me an answer?" As opposed to saying, "I got four or five shots and I'm going to be really smart afterwards because then I'm going to put the real boat in the water and that's high odds." That's where we're going. Mm-hmm. I think that now the technology is there. No one wanted to do this before because you hadn't proven it worked. Now we know it works, and we have a bunch of different tools that are maturing at the same time. I think the agency's getting there. The CMC group is not quite there. The clinical group's there. Manufacturing group's not quite there, but that's okay because we're not quite there either, right? It's just we say we are, but, you know, it's complicated, right? There's ups and there's downs and there's surprises. As that levels, then you can have more confidence around your constructs. Every time you put something in a construct, you can't take it out, right? As opposed to when you do a combination, you can have one with and without if it's an antibody you're adding or a PD-1. Then if it works, you can maybe build it in right afterwards. This is where we're going. It's just it's gonna be turned into more engineering. It can't, we can't leave the rest of the world behind because it's not a T-cell on its own. Engineered, it's not gonna do everything we want it to do. I think we have seen very recently that the FDA has started to change, as you say, not really a CMC, but for companies, it's a big question right now when you go to IND. Do you go with an IND with the first version, and then you have other versions where you incrementally improve or change some features of the CAR-T? Or do you wait for the final versions to do that? We had a particular case for CD19 Allogeneic CAR-T, where we had a version ready to go to IND, but it was not optimized enough in terms of putting stem cell memory cells into the product as a high percentage of the final product, so you can have more persistence and more functional persistence for longer durability of remission, as it has been proven with Autologous. We decided to delay the IND because of that. As you say, Nick, if there had been a system where we can start the IND and then have that as a modern IND and switch slightly different product into that, we could have gone into the clinic earlier and then built on that experience to build further features, further advances in the product and the use of the product. It is not yet possible, but hopefully it will be. If I may, you're adding a lot of new indications, a lot of new candidates, but ultimately, there's got to be the ability to provide additional supply. A lot of these new candidates and indications are all competing for the same resources. Essentially, you're increasing the size of the toothpaste bottle, but the whole thing, you know, relatively similar in size, so it's still hard to squeeze out. How do you meet the demand? The science is there, but how do you ultimately meet the demand and actually manufacture for these patients? Are you talking about the commercial setting or clinical setting? Getting it to the commercial at the end of the day. A lot more candidates. All these products are competing for the same vectors. They're competing for the same manufacturing. Well, on our side, we solve the problem very simply. It was becoming more and more difficult at the beginning, like, I mean, like B.C., like before COVID. That's good. I'm gonna steal that. To get any type of like, you know, plasmids, messenger RNA, etc. Now, try to buy a messenger RNA today. It's crazy. Like, all the manufacturing is becoming hugely complex. The only way to get out from this, and the only path that we had, is we took an, I wouldn't say it's a miraculous decision, but it was an unbelievably good decision in 2018 to integrate, internalize everything. You can discuss with certainly, like Pascal or Nick or any kind of cell and gene therapy in the space. If you don't make it yourself and would like to subcontract your manufacturing at CDMOs, well, forget the D because they say they D, but you have to transfer all your knowledge to them. Once you explain how to do the things and how to develop it, then they charge you for this, and then you have all the tech transfer. It's a total nightmare. The only way to go is to have everything internal. If you don't have the things internal, then you suffer, and it's extremely painful. I'd like to differ a little bit, but let's address your question. I think the big thing on Allogeneic is a scale-up. We have already a product that's going to be approved soon in Europe, where we make about 400 doses per batch. That's already something important. We are working on a new technology for a second product in multiple sclerosis, where clearly making only 400 doses per batch is not sufficient. We're investing in bioreactors. We know already that we can make thousands, and ultimately we'll make 20,000 doses per batch. That's a huge increase in scale-up, and that's where we need to go as an industry, is to be able to do this type of scale-up in manufacturing from different improvements in our technology. We know we can get there. We are already making thousands of potential doses of the MS product in our kind of pilot phase of manufacturing. That's really the way to go. In that case, we don't have an issue because you put that in inventory. We have made already one year of commercial inventory for Tab-cel for the launch in Europe. One year is already there. We don't need to do permanent manufacturing. We just have it in inventory. The more we increase the manufacturing yield, the manufacturing efficiency, the less we will need access to manufacturing space because you just make inventory, and then you've got your inventory at a relatively low cost of goods. Yeah, I think this is gonna converge, whether it's Allo or Auto, they're gonna get better, they're gonna scale, so the commercial hole will get big because there's a need, right? That's the point is if it's working and it's working really well, and you can have a one and done therapy in any one of these diseases, the commercial world will fit around that, right? The economics, okay, whether it's a 60% margin or an 80% margin, right? They're all still pretty doggone good margins as it relates to that, and they continue to get better and better and better. I still come back to one of the challenges in the near term for all of us is cost of capital is high, right? Yes, you want to own everything at the same time, right? Your money isn't just falling off trees these days, right? That's part of the challenge is how do you find that right balance? When are you ready to invest for the future? More importantly, in the near term, to me, to have this ask and answer engine, you need to have vector access, you need to have drug product access in rapid fashion. There you need to own. Why we're building our manufacturing facility, why we have the vector facility, 'cause you don't want to be held back by anyone in that regard. To me, the clinical translational component is the one that needs to be on roids, 'cause the commercial one, quite honestly, as soon as you start seeing something, you can lean into that and actually scale, and that's getting better and better and better. We see that with Abecma, we see it with Breyanzi, we see it with other players, and that's just the autologous space. Ultimately, I think that converges enough that all you really care about, who has the better product. In terms of like versioning, if you want to develop several versions and don't want too much delay between one version and another, then if you want to restart everything, then you need the plasmid that start. You need the slot for the plasmid. You need to discuss the agreement, etc. Then you get your plasmid, then you have to do the messenger RNA, then you have to another slot, etc. If the thing is delayed, then you lose your slot. If you want to go fast, then you have to do everything internally because you can't. I'm with you. I'm with you. I'm with you. I feel your pain. My CSO over here, she's almost crying. I totally agree with you. It's like, wait, what? That takes six months to get, and then you got. Yeah, I'm totally with you. Again, this is where cost of capital comes in, right? You got to get a little bit lucky with things that are working in your current currency, if you will, in order to support a more aggressive pipeline. I wanted to touch on some relatively specific clinical questions, and then we may have some time to just go company by company and ask some pipeline type questions. Just in terms of the way you think about depth of response, and this is for Auto or Allo. You know, is it more important to have you know, the depth of response driven by the persistence of the cells or how do you think about that, the depth of response? Is the persistence what's driving it the most or are there other factors? I think for durability, persistence is a key aspect there. Okay. Because you need to have your cells being able to be functionally persistent. It's not only to be there, but not to be exhausted too rapidly, to be able to fight any type of environment they encounter in the patients, and to be able to permanently answer to repeated antigen challenge. We've made some experience in looking at repeated antigen challenge and the traditional CAR- T, the first one that were created, the Yescarta and so on, have a tendency to exhaust rapidly. It's fine in CD19, as you say, it's been approved, it's doing well in CD19 because you eliminate the bulk of the tumors. If you look at durable response, all of these autologous CD19 CAR T have a maximum of 40% in the LBCL more than six months here. There is no possibility at this stage to go beyond that. It's not only about antigen escape, it's also about the exhaustion of the cells and the lack of persistence of the cells. Any way to increase persistence through different ways. For example, we are working a lot on the phenotype of the cells. Having more stem cells in a final product is going to increase persistence. Changing the design of co-stimulatory domain that are less prone to T-cell exhaustion like 1XX is going to improve that. Having the ability for example in our Allo CAR- T to maintain the function of TCR, because the TCR is a survival signal for the cells, is going to improve persistence. All of this effort improving persistence should lead to durability of remission. Okay. Not much to add there. I I think this is where auto has a natural, right? Because a natural sort of when I say capability there, 'cause it's your own cells, right? So you're not fighting that out of the gate. What I still come back to, even if you get after that, is what else are you doing to sort of clear the way, if you will, for the T-cell through your extrinsic or intrinsic sort of manipulations of the sort of the therapy or the product profile that you're putting forth. I think that's, we're getting there and that's the cool part. Yeah. Well, when you look at literature in this space, and there is like a lot of papers that has been written on persistence versus response and durability of response. Yeah. It's not that clear-cut. Like papers that makes more correlation between the cell expansion at the time, like the in vivo challenge with the product versus the durability of the response. The better the cell expansion is, the better you get a durable response. What Pascal was saying is totally true, actually. It's the most important thing is the fitness of the cell that you're injecting. If you inject even autologous cells that are exhausted, it's not gonna work well. If you inject cells that are extremely fit, it means they can have a repeat killing that is very prolonged. They can multiply very quickly, so the speed between one time where they l ike the cell division is important and also the recharge between like of the granzymes and the perforins that they're stocking inside, and also the type of cells that you're injecting. If you have all these parameters, then you can inject this to a patient, and you have a very strong expansion and kill at this time, then you'll get a good response. If the cell stays there forever, that's not important. This period is extremely important. The quality of the product that you inject and the fitness of the cell that you inject will determine the time of the durability of the response itself. Th at's why it's very difficult and important to be able to master the manufacturing of the cells in order to be able to deliver product that have a very high fitness. You're saying there's some studies showing that the initial depth of response is the most important driver? Yes. Of durability? Yeah. There are studies like that. There is a paper of, like, Jae Park. In CD19 space only. Yeah, the CD19. There is. There's probably 10 papers on both sides. Yeah. We'll have to leave that for another day then. Yeah. I wanted to touch on two other topics. Lymphodepletion, if you could talk about there's different strategies there, other approaches to engraftment, maybe beyond lymphodepletion, if there are others, and then redosing? How important is redosing in the long-term strategy for Auto or Allo? Yeah, I think, I mean, redosing we have experience because that cell in ATA188, in MS, we are redosing patients, so it's safe and doesn't create any particular risk of GvHD, whatever. I think the ability of redosing is something very important. At the same time, the lymphodepletion is something that we don't need in our EBV associated. That's right. We, you know, we believe on the Allo CAR- T that will go to the clinic, the CD19, next year. We will have the low level of lymphodepletion. What we know is that we don't need to go to high level long lymphodepletion like adding an anti-CD52, because that's interesting, because if persistence was not important, it's interesting to see many companies giving an anti-CD52 there, which is only objective, is to have more persistence and more longer prolonged lymphodepletion, which might create potentially some risk, from that point of view. We like to go at standard low level lymphodepletion at the kind of worst case scenarios to have sufficient engraftment of the Allo CAR- T. You got to be careful with which variable you want to play with, right? Because you end up kind of destroying the therapy a little bit. It's going to just take us a while where we're going to have to push the limits on a few areas, and then we can start dialing things back. I think we're just getting to the point where, hey, before I want to create the space, give the cell the best opportunity to be successful. All right, before you start dialing that back, before you have solved some other problem, right? One thing is that, yes, the health of the cell is great, but what else is standing in the way, right? What in the microenvironment is standing in the way? That's a very different answer in hematologic versus solid tumors. What you might need to do, right? There's some interesting stuff that's happening, not even in our space. Some of the Regeneron data out there, it's interesting with their bispecific and PD-1 that independently don't really work. You put them together and it's gangbusters, right? This is using words like that's the trail to the Holy Grail concept that George is putting out there. I I think it's really interesting though, 'cause it applies to what we're doing here as well, saying, all right, well, that's clearly activating the immune system in some way. Can we control it? Can we support it in a way? The combinations are starting to work, and I think that's what we're all sort of discovering is as good as one individual agent is or might not be, when you start putting them together, you've all of a sudden taken a block that then actually allows the cell to do what it needs to do, the T-cell in particular. That's why I do believe the fundamental base layer here is still T-cell based. I think you said that earlier as well. Okay, I think we have time for probably one company specific question. For Atara, for Tab-cel, just give us the latest on the filing strategy. Is it safe to say at this point that we're not going to need another Phase 3 and that you'll be able to leverage some of the existing data to make the filing happen? Yeah. Following this very constructive dialogue with the FDA, they proposed a possible path to file the BLA in the U.S. without a new clinical study, the commercial product. We are right now gathering clinical data on the use of the commercial product in the pivotal study as well as in compassionate use. We have a number of patients increasing month after month there. We are finalizing with the FDA over the next couple of months. We'll have some meetings with them, and we have a lot of active discussions to try to clarify how many patients in particular they will consider as sufficient. That will then allow us to clarify when we can file the BLA. Meanwhile in Europe, as I said, we are progressing very well and we're expecting very significant steps over the next few months. Our partner, Pierre Fabre, is ready to launch in early 2023. One other follow-up. I said I was going to ask one question, but this is quick. You're planning for a U.S. commercialization partner as well. What are you thinking about there? What type of partner do you want? Since we can ensure all what I would call the back office, we can ensure manufacturing, supply, logistics and so on, we don't need a partner that knows cell therapy. Mm-hmm. We need a partner that has already an existing infrastructure in the field of rare cancers and can deliver to the potential which we believe is largely over $500 million for that product in the U.S. That partner doesn't need to be a cell therapy specialist. We're going to talk to a series of companies that have an interest to have an additional product that is a neutral rare cancer type of product with a significant pricing potential. Okay. David, you have a question for Nick? You announced a new trial starting today. Tell us about the study and what's the design and what you expect to learn. Yeah. Just to back up, this is about Abecma, myeloma, right, or approved in sort of fourth, fifth line. A month ago or so, we announced a positive top line data on our KarMMa-3 study, which is in two to four lines. First randomized controlled study. More details coming towards the end. Very excited about that, both on PFS as well as overall response rate. With that signal, as well as a study that we're running called KarMMa-2, which is exploring different ways to enter into earlier lines, second line and first line, as that data's evolved, we now have announced today that we're going to be initiating along with BMS, KarMMa-9 study, which is a newly diagnosed study in the context that we're strong believers in transplant in that frontline setting. This is really a suboptimal response to transplant. You're going to have transplant. Then if you have a suboptimal response really defined as anything short of a CR, which is about 50%, 60% of the population, that's where we think Abecma can fit in and really come in and say, "You haven't relapsed yet, but we really need to hammer home and finish off the job." That's what we believe between KarMMa-3 and KarMMa-2, we have very high probability because we're big believers in transplant and what it's doing. It can also be in the spirit of what we've been talking about is how do you partner with existing modalities and existing proven treatments? This is exactly what it is. More to come on that as we get towards the end of the year, along with with BMS, our partner, on this front. It's exciting because the big question has been, are you and BMS, are you guys in? Are you not in? Where are you going in earlier lines? Well, here it is, and we're going for it. Looking into the crystal ball, when can we start thinking about label expansion? Well, I think right now the KarMMa-3 study, right? We're obviously gonna engage with the agency to try to get that into the label, right? And then over time, as soon as we get this study going. But this is gonna be a run. Again, speaking of the session, once the battle that plays out tomorrow. No, this is gonna be the next 5-10 years of Abecma and Cilta-cel and probably others that are really crushing myeloma in a big way. We're settling in for the long-term battle on this, and we're excited that BMS is as excited as we are on this, and that's what these studies are now all about. Meanwhile, we're ramping up, continuing to reiterate sort of the top end of our projections for Abecma this year, and so it's going about as well, if not better, as I thought in the context of Abecma. Good news is more myeloma patients are getting better responses. André, you mentioned the CD20 by CD22, which is about to start the study. Yes, this is a very exciting product, because first, I think it's the first dual Allogeneic CAR-T that's entering the clinic. I like this product, and we're excited by this product because it's not a CD19 product. There is a need. Like, you talk to most of the leukemia or lymphoma doctors, they're all looking for products post-19 relapse. Mm-hmm. They have nothing. There is a real demand for this type of products that is exciting. First of all, the dual CAR-T gives you an interesting response because it enhances, augments the number of contact points between the CAR itself and the tumor cells, so it can form a better synapse, so you have a better killing activity driven by the many contact points. Because most of the target that we're targeting can wobbles. They have, like, high expression and low expression sometimes. Most of the relapses are due to lower density of tumor-associated antigen without the ability for the CAR-T to form a synapse to make the kill. That's avoided by having multiple targets that can recognize. It's totally different than injecting a CD20 and a CD22 CAR-T together. You don't have the same killing effect. There is a real clear path to commercialization for a product like this, and we'll see how this is going to be driven. Already all the preclinical data that we have since the beginning are extremely exciting, and I think that this product is off to a great path. Okay. We'll have to leave it there. Thank you all very much. Thank you. Have a good conversation. Thank you, Yigal Nochomovitz. Thank you very much.