Allogene Therapeutics, Inc. (ALLO) Earnings Call Transcript & Summary

Good afternoon, everyone. Thank you so much for joining us for this panel. It's the outlook for the allogeneic approach optimizing the clinical profile, and I'm really pleased to have with us David Chang, President, CEO and Co-Founder of Allogene Therapeutics and Steve Harr, President, CEO and Co-Founder of Sana Biotechnology, and I'm joined by my colleague, Matt Dellatorre and I'm especially excited to host this panel because it reminds me of days prior when we would sit up here and talk about the days of Kite and Juno, but maybe just start here.

I have to say when I found out that Steve Harr will be on this panel, I said, "I'm not going to miss this."

Everybody loves a history lesson, don't they?

They do. They do.

So maybe to start here. Both of you were pioneers in the field of autologous CAR T therapy. And you were looking at other modalities as well and looking at other disease areas as kind of the future of where you might go. And you've now both cofounded companies that are focused on taking different approaches, right? So, help us understand how you think about allogeneic therapy or the allogeneic approach as kind of the next frontier of cell therapy, particularly for hematologic malignancies, but now also for autoimmune diseases. And just frame the current state of the field for us, your respective programs and approaches and then your upcoming datasets.

Do you want to start?

Okay. So I think Sana and Allogene were founded at about the same time and both companies were found with a very similar mission, really democratizing the cell therapy, I would say, taking the bespoken manufacturing of autologous to the next step. And certainly, at the time when I finished Kite Pharma and thinking about the next step, the ideas about different indications, hematologic malignancies and solid tumors as well as even autoimmune at that time was discussed. But independent of which indication that you go into, you have to move into a more scalable manufacturing, which is where the allogeneic cell therapy comes. And I would say, over the last 5 years, there has been a lot of advancements that everyone in the field has been able to achieve in the allogeneic CAR T therapy. I think everybody now understands probably the most important facet of making allogeneic cell therapy work is overcoming rejection by the patient's immune system of the allogeneic CAR T cells, which patient's immune system sees as a foreign and tries best to reject it. And there are different approaches that people are taking. Fundamentally, I think you can simplify it as either Cloak or Dagger, making CAR T cells invisible from the patient's immune system or have a way to counter this allo-reactive T cells -- patient's allo-reactive T cells from rejecting the autologous -- allogeneic CAR T cells. And I would say that especially on the [ lagger ] which is the way that Allogene has been advancing the pipeline, there has been a lot of advancement and demonstration that allogeneic CAR T therapy can provide the same level of efficacy as well as durability in the heme malignancies. And I would also say that another thing that everybody and certainly in the morning sessions, there were a lot of discussions about the manufacturing. Manufacturing is not just the facility that were manufacturing the cells but also the manufacturing science. How to make the CAR T cells more potent, understanding the critical attributes of the CAR T cells so that you can continuously improve either through manufacturing or by gene engineering to make the CAR T cells work much better. So at Allogene, our LEAP program is cema-cel. This is a product that we have been studying in relapsed refractory setting for some time, demonstrating that as allogeneic CAR T product using our proprietary lymphodepletion that is augmented with a -- in a biologic anti-CD52 antibody on top of fludarabine and cyclophosphamide that we can achieve the overall response rate, complete remission rate as well as a durable remission on par with the autologous CAR T therapy. But as we are advancing the program, there was also sort of changes that's happening in the large B-cell lymphoma space. And more recently, we decided to really differentiate our development strategy by finding a very unique and differentiated approach of advancing the cema-cel into the front line as a consolidation approach. And this is a study that we are currently conducting as a pivotal study. We announced the initiation of the study back in June of 2024 this year. And this is a program that has -- that's very reaching milestone over the next 3 years, starting initially with the early readout of the study in mid-2025. And in 2026, we have the first efficacy analysis at about the time that we expect to complete the enrollment, that's in the first half of 2026. And that leads to the primary analysis, EFS is the primary end point by the year-end, leading to potential BLA filing in 2027. Our next program that is sort of a next-generation approach towards allogeneic, this is our ALLO-316 which is a CD70 directed CAR T therapy that we have been studying in renal cell carcinoma. So this is a study that we've been doing for some time. And the initial data presentation was made at AACR last year, where we have shown that in the CD70 positive renal cell carcinoma, we can get approximately 30% response rate. This is a data that we plan to update before the year-end in a forum where we will cover more patients who have been treated as well as testing what we believe is -- would be the recommended Phase II dose, both in terms of cell dose and lymphodepletion. And the third program is really trying to create more opportunities with the cell therapy overall. And this is autoimmune disease indications. There were a lot of discussions about autoimmune earlier in the session, so I will not go into why the excitement of cell therapy in the autoimmune space is. So we have a program that was designed from the beginning with the autoimmune indications in mind. So this is ALLO-329, CD19 and CD70 dual CAR that we are using next-generation manufacturing technologies. So manufacturing, we always believe that it has to be simple and scalable. And with ALLO-329, we are using CRISPR-based gene editing using a single cut onto which we use AAV virus to introduce the transgene. And this is a program that's also has 2 important aspects. One, having a CD70 allows this ALLO-329 not only to deplete the B-cells, but also deplete CD70-positive activated T cells. And autoimmune disorder is never a B cell or T cell disorders. It's really the combination of two different lineage of immune cells that contribute to auto immunity. And this program is designed from the beginning to address those 2 aspects in both B cell and T cell aspect of the auto immunity. And also having the CD70 really differentiates what we are doing in the allogeneic CAR T space. CD70 portion brings the Dagger technology. So we talked about Cloak and Dagger. This CD70 allows allogeneic CAR T cells to essentially overcome patients' immune system from rejecting the CAR T. So in ALLO-316 program, we have seen great cell expansion and persistence that's going up to about 2 to 3 months easily using the Dagger approach. So this is embedded into the ALLO-329 as we think about the autoimmunity with a very important goal, which is to minimize or reduce the lymphodepletion as we advance the program into the auto immunity. And lastly, also the much of the earlier conversation covered the need to be able to manufacture and like Lyell like Iovance we invested early on in our manufacturing facility, and we have a wholly owned manufacturing facility about 140,000 square foot facility that at the current -- at the full capacity, can produce upwards of more than 20,000 doses of products for cema-cel. So, just to put that in the context; in 2023, the total number of patients were treated with CAR T commercially, the number is about between 9500 to about 10,000. That's estimated number. And this -- a single facility can produce enough to treat 20,000 facility, and that is really one of the big distinction and advantages of the allogeneic CAR T as we think about the future of the cell therapy.

Let me give it a shot. I'll go back to your original question, the founding of these companies and things like that. I was just thinking a little bit about, as David was talking in answer from Bob Newhart when Larry, Darryl and Darryl, one of the Darryl ran for office, and he got up after a talk and just said Ditto. But I'll try to add a little bit to it, which is one of the beautiful things about autologous CAR T cells is that you can get an early indication of how well they're working in a relatively capital-efficient way. I mean most of these drugs were developed out of academia and things like that. I think one of the real challenges that we all learned as we were doing it was that it wasn't just the complexity of manufacturing, which I think is what people really have focused on, but it's the complexity of release and it's complexity of quality and things like that. And it's a complexity for the patient and the physician, and the apheresis center, things like that around scheduling. And so we really start out going into the allogeneic side with the idea that we could hopefully eliminate much of that, as David said, by really trying to overcome the biology of immune rejection. I think the challenge of it is, though, and this is a particular challenge in a world like today is that it's really capital intensive and it's capital-intensive early, right? The supply chains are complicated, just to make a gene modified products, you're looking at making guide RNAs and mRNA and maybe a viral vector or something like that before you've ever dosed a patient and you have to do that GMP. And you're also putting money into manufacturing facilities. The beautiful part of that, though, is if we happen to get the biology right, they truly are scalable in ways and the cost of goods and our ability to make them feel to a patient and a physician exactly like an antibody is real. And so while it's been painful to put all this capital risk in an era of increasing cost of capital, I think the upshot is that if we happen to get this right, we have a really quick runway and our ability to lock processes and move into scalable registration studies with scalable commercial products. So that's kind of how I thought -- how we've thought about it. We've got a number of different ways we're trying to apply the technology, both within CAR T cells and stem cell-derived therapies. Most of it resides upon the singular kind of insight around what we hope is an ability to overcome allogeneic recognition of these cells, which I'm sure we'll get into as we go forward. But that's a little bit about where we are in what we're up to and why if we happen going to get these right, these allogeneic therapies can be so powerful for patients.

David, let's start with you. So you talked about the ALPHA3 trial here, and you've made the decision to pivot cema-cel to frontline LBCL patients who are MRD positive. Help us understand how the trial is progressing, but also the ability to identify these patients, the ability for this to be a successful commercial opportunity and for the data that you've seen in this subset to translate in the Phase III?

Yes. So just on the background about what we announced earlier this year. We were advancing cema-cel in an allogeneic CAR T product with a mindset of how we advance the autologous CAR T and that it came to a recognition that we are not fully taking advantages of allogeneic approach and also not accounting for some of the barriers as we are trying to move into the earlier line where we would expect the most benefit from a potentially curative treatment. So we made a decision, and this is really coming from 2 supporting evidence. One, from the relapsed/refractory setting, we knew that cema-cel was as good as autologous CAR-T. Yes, the data set is a smaller number of patients. But when we compare our Phase I data to the pivotal data that supported the registration of YESCARTA or BREYANZI, that was very similar. The other thing is really advancement in the other field, which is the in vitro diagnostics. This is using the circulating DNA to identify MRD-positive patients. So marrying those 2 important information together, we came up with a clinical development strategy that is uniquely differentiated and also relatively competition-free in the first-line setting. The first line setting for the large B-cell lymphoma, they are a highly effective treatment, Rituxan-based chemoimmunotherapy that will provide cure in about 60 patients. And even before that, I mean 90% or more of patients will respond and respond very well. But the problem is that about those 30% of the patients who initially have good response when you watch and wait, their disease will progress, and they will have to go through the second and subsequent line of therapy. And the question is how to improve the initial treatment effects so that those 30% of the patients can be managed. And so they do not have to go through the agony of being told that the disease has come back and then go to the different lines of treatment. So what we are doing is letting the patients complete the initial R-CHOP or similar regimen and at the end of the treatment, testing the patients with the presence of MRD. And when they're MRD positive, then they are eligible for the cema-cel ALPHA3 study. And this is a randomized study that we are comparing patients with a single cycle of cema-cel versus watch and wait, which is a standard of care and using the EFS and endpoint and will be sort of carrying out the study. And as I've said earlier, this is a study reaching catalyst over the next 2.5, 3 years. In terms of the study, I mean, much of the work in the first half of the year was trying to get all the ducts lined up to start the study. So announcing that we were moving into the front line was made in January. And then in June, we announced the site activation. And so far, we are busy activating both community-based cancer centers where most frontline patients are cared for as well as academic centers and site activation is tracking very well. And we are also testing patients for the MRD using this proprietary assay that our partner, Foresight Diagnostic, has developed, and that is also going very well as we have projected. So we have screened patients, we are enrolling the patients. And so far, all the earlier signs are looking good.

And you touched on the CD70 targeted CAR T data that's coming by year-end and some data early next year in a different indication. Just put that in context for us versus the clinical data you've seen to date and what you're looking for in these 2 populations?

Yes. So in a solid tumor data, which is our ALLO-316 program, the goal that we are trying to get is getting enough number of patients, and this has to be the right patient population and demonstrating the response that probably at least like -- I like personally like to see more than 1/4, maybe 1/3 of the patients achieving the response. And then the second question is, what is the durability of the response? And those 2 are the things that we are really focusing at this point at the cell dose and lymphodepletion. And currently, for ALLO-316, which has a Dagger technology, we're using the same lymphodepletion that autologous CAR T therapy is using. So we do not have to enhance the lymphodepletion even with an allogeneic CAR T. And we will be updating the Phase I data, as I said before the year-end with additional number of patients that are evaluable as well as more durable follow-up on the patients that have already been treated.

So Steve, we'll see the IST data in type 1 diabetes, hopefully, by year-end, which has been a big focus for investors. Could you maybe just frame for us what you want to see there to move forward with your corresponding iPSC program?

So this is -- I mean we changed text a little bit here. And this is looking at type 1 diabetes. And I think -- most of you recognize type 1 diabetes, as the immune system has attacked and killed all the beta cells in a patient. And because they don't have beta cells, they can't make insulin and they end up -- until 100 years ago, they died from glucose intolerance and now people take insulin. And even with best care today, people have about a 10- to 15-year shorter expected life span and during that time, there is a lot of complexity for the patient and a lot of complications as well. And so what's known is that there have been thousands of these cadaveric derived islets that have been transplanted into patients. In the context of immunosuppression, many patients can go 10 to 20 years with insulin free and then some of them are going on longer, we'll have to see insulin free glucose control, and they really do quite well. The problem is it's not really a scalable source nor is there that many patients for whom lifelong immunosuppression has been on lifelong insulin. We now know from others in the field that you can take stem cells and make beta cells, and it's been reproduced several times now. And in the context of immunosuppression, again, patients can do quite well over time and stay insulin free. But you still have -- that's a more scalable source, but you still have the problem of immunosuppression. So the question we're trying to answer this immunosuppression, sorry, this investigator-sponsored trial is can we get rid of immunosuppression? So we have done this in nonhuman primates and in other animal models and shown that in the context of gene modified primary islet cells from another -- from a similar to the same species you can see long-term insulin free survival in these animals with normal glucose control. So our goal in this study is a little bit different, right? And so what we want to see is that these cells survive and they function. And it's a Phase I first-in-human study. So you have to just take into context what is the dose and what's the goal. So all -- if we can show that the cells survive and function, I would argue a cure for type 1 diabetes is inevitable because now you've kind of completed the circle, everything you need to do to make it happen. We might not be the one to do it but it becomes inevitable. And you need a gene-modified pluripotent stem cells, make them into functional islets at scale and transplant them, right? And so, the goal of this study is to see this. So there are 3 places you can see survival and function. One is you see it on an MRI. And you can see that there's actually a paper last week in cell of a patient who had autologous derived stem cells that were transplant, autologous stem cells that are made to beta cells that were transplanted in the muscle sheath. And the patient has done amazingly well. And you can see those cells. So that would be step 1. Step 2 would be to see C-peptide. And I think that should be where the expectation is as good as we will do because of the first-in-human dosing. See, if you recall, the way that a beta cell makes insulin actually make something called proinsulin. And as proinsulin is secreted, it's cleaved into insulin and C-peptide. So all these patients have no C-peptide. If you see C-peptide in the patient, you now know the patient is able to make their own insulin. They're able to do that in a sustainable fashion, right, without any immunosuppression that we're doing, and there's just -- we'll see how that goes. The third would be able to see patients with better glucose control. That shouldn't be your expectation. I mean it could happen if you get a little bit lucky and that happens, but that would not be the expectation at the doses that we're dosing at. And it's demonstrably lower than, for example, what the doses that was in that cell paper or [indiscernible] So that's a little bit about what to expect. And -- so how do you translate that into a therapy? Our goal is not to gene modify cadaveric islets and transplant those in patients at scale. Our goal is to gene modify pluripotent stem cell, grow it into pancreatic islets and do that at scale. So to do that, we've said you have to overcome 4 scientific challenges really to go for that. One is to make a gene modified pluripotent stem cells master cell bank, where you're really confident that you don't make integrity over time. Remember, you got trillions of divisions that you'll be looking at over time. And in our body, we have basically every single mutation, you can imagine, right? And we'll be growing these cells in a media that's selecting for cells to grow quickly. And that's what you do when you manufacture cells. You want to make sure you're not creating a tumor, right, something that could be a tumor. And so we think we've got that. I think we've got that. It is taken us longer than we thought, it has taken us a few years to really kind of get that right, but we think we have that not guaranteed yet. The second is to be able to manufacture these drugs at a pluripotency and scale to run a Phase I study. Again, you should feel pretty good that we got that. Not guaranteed, once you have to make sure you got the same. Every time you like you change your master if you get the subcon from master cell bank, it's like kind of like a new chocolate for your chocolate cake, you may taste a tiny bit different, and you've got to kind of adjust a few of the other ingredients and maybe we have to play around with the salt through the vanilla or the sugar but will get it right. The third is able to overcome allogeneic and autoimmune recognition and rejection of the cell. Again, you'll learn that from this investigator-sponsored trial. So very shortly, you should feel really good about 3 of those 4 things. The fourth is to be able to make the drug at a pluripotency and yield as truly commercially important. And I think, to be clear, we're pretty far away from that. I think that, that's going to take us some time to do. It is not simple to turn these stem cell-derived products into billions or trillions of cells. We'll get there. The field would get there. I wouldn't be too worried about it. But it's going to take time. And I'm guessing that if we happen to get this right, you're going to be frustrated with us for a long time around the scale of which we're manufacturing this partly that relates to just the overall size of the market, which is really quite big. And partly that relates to this -- where we are in the science of scaling these drugs.

Great. So then we'll also see Phase I data for our lead asset, CD19, SC291 in autoimmune and B-cell malignancies. Maybe could you just kind of frame your expectations that go into those 2 sets? And then also, how should we think about read through, particularly in autoimmune disease?

About what?

About potential read through from those early autoimmune sets?

So, what do you know so -- so we make this drug, right? We put these gene modifications and then we've shown you a few patients worth of data, and it looks like -- they look pretty good. They're early. There are no safety issues that we're seeing. They seem to avoid immune detection by the assays that we have. And for a couple of patients in we had a few responses. That's all I could really ask for at that point. So I think in oncology, the challenge here has been durable complete responses, right, and having a profile that looks like the autologous cells. And then if you happen to get that, you should have an important drug. But as David mentioned, it's a very competitive marketplace, and these therapies are now developing long-term survival data. And you have to kind of find your niche, which Allogene has done quite well. And we have to make sure we find that niche to make sure it's useful to go forward, right? Autoimmune -- and so in many regards, we have to deal with the context of those that came ahead of us because they are a lot ahead of us. And we may or may not be able to figure that out in the near term, right? Because if you're really looking for a 6- to 12-month durable response data, I would argue these to be 12 months if you're not in an aggressive lymphoma, then you've got a kind of -- a lot of the Phase I patients are indolent lymphomas and you got to be patient. And you kind of know that in January, we gave you data with 4 patients that you can't be much better than 4 in 1 year. right? And so there's time you can't accelerate it. So that may take longer to understand the drug profile than people hope. The second will be the autoimmune setting. And I think there it's very, very different. First off, the data from the autologous cells are pretty profound, but they're pretty limited to date, right? I mean you're just beginning to see the emergence of data from company-sponsored studies where it's probably a bit more generalizable patient population than what you have in a single center where the patients are highly selective. You can see like the oldest people in those studies would mostly be younger than the youngest person in this room, right? And so when you get into the real world, particularly in the United States and things, you're just going to have an older patient population with a lot more end-organ damage. Second, you already know we can deplete B cells, right? We've shown you that in the oncology data. We've depleted them pretty well. So really, what I've tried to tell people that all you can really learn in the early parts of the study is, do we have a safety problem? Hopefully not. Do we deplete B cells in the autoimmune setting like we do in oncology? And does that translate into early clinical benefit? If you really want to understand where we stack up competitively, it's going to take time. If for no other reason, we don't know where the competitors are yet, right? I mean there's still -- really a lot of them we don't have much data or they're just beginning clinical studies or they've got a handful of patients. And it will take similarly some time. I think you can tell pretty quickly, do we have a drug or not, right? I'd be pretty surprised if we don't have a drug, right, just given what we've seen in the oncology setting. But that would be what you should hope to learn out of early data. It's going to be very difficult to define exactly where it fits against all these different modalities. I think it will be very simple to define whether it's a drug or not.

David, just following up on that topic. You have an effort here, which you touched on in autoimmune diseases, maybe speak to your approach here where you are bringing in CD70 and you're looking to get rid of lymphodepletion. How do you think about the path?

Yes. I mean our effort into autoimmune. I mean this is really building on the early data set that's coming from the autologous CAR T that's really influencing us. I mean, I have to say everybody in this room has to give a lot of credit to Dr. George Schett for taking autologous CD19 CAR T into autoimmune space. I would say that we consider that back in 2018, 2019, but we didn't have guts to do that because of the safety concerns. And when you think about the autoimmune and the B-cell depletion therapy, especially with a CAR T, I mean, the goal here is really the remission that's durable and going for several years. I mean, I wouldn't say necessarily cure, but taking the treatment that is chronic where a patient has to receive treatment either daily oral or on every 3 to 4 months, if it's an IV or subcu to something that you can just get treated and forget about that you had a disease. I think that's where the promise of CAR T in the autoimmune space is and certainly, in autoimmune also is from the commercial perspective, this is a different beast from any oncology indications that we are talking about. So, with all that in mind, and as I said earlier, we started from the very scratch. What is the target product profile that you would like to have to advanced allogeneic CAR T into the autoimmune space. We started in a very important question. It cannot be just depleting the B cell, you have to address the T cell component. And this is where the CD70 portion comes in. CD70 is an antigen that gets expressed in the activated T cells. And having the CD70 CAR addresses activated T cells, which we believe plays a role in autoimmune disorders. And another thing that we thought very carefully is the question of the lymphodepletion, the rheumatology indications, unlike in oncology, safety stands out as a paramount important aspect more than in any diseases. And here, having a chemotherapy-based lymphodepletion that we always saw as a barrier, and frankly, that was one of the reasons that we do not take CD19 CAR T into autoimmune back in 2018, 2019 in early state of time. But what having the CD70 allows one to sort of ambition is gradually getting rid of the lymphodepletion. I mean the step 1 is already there. I mean we have already talked about for allogeneic CAR T you need more lymphodepletion unless you use a different cloaking approach. But our experience with ALLO-316, which is a clinical asset that's being studied in renal cell cancer is that even as an allogeneic CAR T having the Dagger technology allows you to go through the standard autologous lymphodepletion that's based in Flu/Cy. So -- and not only that, these cells expand very well and persist for quite a long time. And that allows us to really think about can we, in a step-wise address this area of lymphodepletion, either reducing it or eliminating altogether. I mean that's really the target product profile that we thought would meet the demand as well as the opportunities in the autoimmune disorders. So ALLO-329, which is the CD19, CD70 dual CAR, we have really put a lot of effort accelerating the preclinical work. Since the announcement of the decision to advance into the autoimmune, we are essentially finishing the IND work within a 12-month period. And currently, we are planning to submit the IND in the first quarter of 2025 next year with the clinical study start by midyear and hopefully, getting the initial proof of concept by the year-end 2025.

Did you want to comment at all on the lymphodepletion?

Lymphodepletion. I mean I think if you look in the autoimmune setting, the company that can get rid of it will be the company that wins. I mean it's clearly something that is super, super sensitive to both the patients and the physicians. It's -- it does -- I've never seen an exogenous cell therapy product successfully treat patients without lymphodepletion, right? So I think we still have some work to do to understand what the magic of it really is and how do we get rid of it. To the extent that you can it will be super important. Do you have a challenge? I mean, I think people forget the math and like autoimmune, these autoimmune as an example. You have -- we all have around 300 billion B cells, right? And so what you measure in the periphery is somewhere between 1 billion and 5 billion in your bloodstream. So your bloodstream is a trivial portion of the overall B cell repertoire. And most of your B cells reside either in your tissues and tissue germinal centers or on your lymph nodes and spleen. And until you really -- and the game and -- the goal, I think, in autoimmune is to deplete every single malignant B cell, right? And it's a math problem if you don't have really nice cell expansion and things like that. And it's the same thing with oncology, where you probably have to eliminate every B cell to eliminate the tumor, right? Because it's got to be harder to eliminate cancer cells and B cells. And so when you're looking at lymphodepletion as an example, what it seems to drive is great predictor of early data in the autologous setting, a great predictor of early expansion is just your IL-7 and IL-15 levels at the time of infusion, right? An IL-7, IL-15 are homeostatic cytokines that go up when you lymphodeplete somebody. And so we have to figure out how to deal with that problem to get the expansion that's almost certainly necessary because almost no drug that I'm aware of, even like small molecules very really, they have 300 billion active moieties right? And so if you're really trying to get rid of the malignant B-cell or called the pathogenic B cell to be like maybe less judgmental, then you've got to go after some huge percentage of those 300 billion, right? And I think that's the challenge for the cell therapy field and getting rid of lymphodepletion. And it is the challenge for any other therapy that doesn't replicate in vivo.

When it comes to lymphodepletion, I mean if you sort of follow the literature, and Bruce Levine is here, and he can probably talk about lymphodepletion better than anybody else here. But lymphodepletion enhances the pharmacodynamic effect of CAR T. And you have shown it with one of your products when you're at Sana going from cyclophosphamide and then having to add fludarabine but autoimmune disorders, I mean, when you look at the clinical data, one of the most tantalizing findings from the autologous CAR T is you don't really seem to need long-term persistence of the CAR T. I mean it's essentially hit and run. You get rid of -- you deplete the B-cells, and let the normal B-cells recover. That has to be part of the equation because people cannot do without CD19 cells altogether. So there has to be a treatment that depletes but allows the B cells to come back. And when you sort of think about it, the idea of sort of reducing or eliminating. And I will say the idea of reducing the lymphodepletion has already been tested, not intentionally, but if you follow George Schett's data, I mean, he talks about now at least 2 patients for one reason or another, got much reduced lymphodepletion. And the outcome of those patients treated with less lymphodepletion doesn't seem to be any different than the ones who got the standard lymphodepletion. So this is where one can really think about how to manufacture the cells to make it potent and balance out the lymphodepletion and the potency of the drug, so you can get the right equation. And I think that based on all the scientific findings as well as emerging early clinical data is something that's doable.