Good afternoon, everybody. Thank you for joining us. We've got Cokey here from Atara Biotherapeutics. Thank you so much for spending some time with me this afternoon.
Thank you. It's great to be here. Thanks for having us.
All right. Well, let's get started with, let's quickly move through tab-cel and legacy programs, and we'll get to the new stuff. But I'd love to get your sense for what you think Pierre Fabre's potential is in the U.S. for a tab-cel launch, and what sort of royalties, milestones, cadence are you already baking into your assumptions currently, and where do you expect that to develop as we get into the launch next year?
So I think the first thing to start with is we're pretty excited that we're on track for PDUFA in January, which for the company history, it's meaningful.
Yeah.
We're probably a month away. Things seem to be on track with the agency. As you know, the approval on EMA happened first, surprisingly. I would say for the commercial opportunity, we've done our homework, and I think what's interesting is it's been corroborated by our colleagues at Pierre Fabre North America. They've hired out a whole American subsidiary to commercialize this. It's led by Adrian Herrera, who is formerly of Kite. I think the envelope math says the back of the envelope, $500 million in peak sales is still attainable. I think why we thought that and why that's been sort of confirmed by Pierre Fabre is there's probably 70 transplant centers that have used tab-cel. When you do a deeper analysis, there could be up to 200 centers that are ready to use tab-cel, but they lack awareness.
And so filling that gap with a commercial team leads you, if you do all the right math, to that $500 million peak number. And I think it's interesting, the situation in the U.S., because there's been broader adoption in terms of how many centers have used tab-cel. So that's why.
You mean through the clinical program?
Yeah, through the clinical. That's right. I think that's why we're confident that this has legs for the U.S.
Excellent.
I think to your second question, we are looking forward to a milestone at PDUFA, and commensurate with that, we'll transfer all of the tab-cel activities, the CMC activities of Pierre Fabre. Then we are predicting a run rate into 2027. Obviously, it would be transformative for us as a company. In addition to that milestone, we'll get milestones from sales and significant double-digit royalties.
Those aren't currently baked into your runway guidance?
They are. They are. Yeah. Yeah.
All right. So with the remaining time, we can transition to the internal programs in autoimmune where I think the focus obviously has developed here. Let's start with 3219, your CAR-T, your allogeneic CAR-T with CD19 in autoimmune space. Obviously, the market focus for autoimmune has shifted really dramatically away from CAR-T and towards bispecifics over the course of this year. Before we get into the details, I would love to just get your sense from a high level, how is your program different from other cell therapies where people have lost confidence? On the flip side, where do you identify the weaknesses in bispecific programs that we should be on the lookout for as all of this data develops further?
We're pretty differentiated. There's the classic question about where are you going to play and how do you think you're going to win? I think we are proceeding very carefully. If we weren't differentiated, it would be a tough sell. That's number one. We're differentiated because we're a biology-first company, and I think that already bifurcates because most allogeneic approaches to date are gene editing. I think our approach is more rooted in tab-cel. We've treated 600 patients with tab-cel. We see a consistent safety profile, which is to say very minimal adverse events, consistent efficacy over 20 years. We looked at that and said, hey, this is probably a great off-the-shelf program.
In general.
Right, as a platform. So we do minimal modification. We transduced the CAR. And what we engineered was something that looks the most like an autologous alpha-beta T cell. That's what EBV CAR-Ts are, without gene editing from healthy donors. So we're differentiated in a couple of ways. Number one, we've learned how to make these T cells from tab-cel. There's so many lessons about how to tune the T cells for a central memory phenotype, which we think is important. It's been somewhat validated. Bristol, Novartis, T-Charge. There's the 1XX domain, which we think is differentiated. What we see in the 1XX chimeric antigen receptor domain is we think it's a less inflammatory phenotype. And especially for autoimmune patients, that's probably important. The typical CAR-T profile that's autologous, it looks like it can be pretty inflammatory from a cytokine perspective.
So we think all of these elements baked in with the lack of gene editing is really differentiated. And that's why we're excited about the program.
So you're thinking about better tolerability, obviously off-the-shelf dosing being the key differentiators versus cell therapies. But what about bispecifics? Given the promise of the Schett paper was that an immune reset would give durable functional cure for these patients, when folks saw a handful of relapses, even just a handful, it turns that logic from, well, we're going to cure you forever with one admittedly high-priced treatment to you're going to have to come back again and again. And that's why I think people get so focused on bispecifics. Where does Atara fit into that landscape?
So it's provocative to think about bispecifics because there's just ease of administration.
Yeah.
And it's drug-like for pharma. But what I would say is compared to autologous, it would make sense, especially in the community physician, off-the-shelf bispecific makes sense. But I think that when you start to talk about allogeneic and if you have the right allogeneic platform, by and by, it is very much like a bispecific. So then it gets down to the question about efficacy and safety. With the bispecifics, the data are limited, but you see it in oncology. There's the same, because it's a T-cell methodology that you're invoking the immune system, you see CRS. But what you also see in some cases is infection risk. There's also the question of continuous infusion or how long that infusion has to go for the patient to benefit. So I think that's what you present with. And even in large B-cell or myeloma, CAR-Ts are still superior.
On efficacy.
Yeah, on efficacy, certainly. And so the question when you pivot into autoimmunity is, if you're not as efficacious in oncology, would you have that in autoimmunity? I think the data is too early to say that. But what I would say is when you have an allogeneic platform, it solves a lot of the issues here if you think about it, because if you're allogeneic truly, there is none of this lag time between apheresis, putting the patient on bridging, and then infusion. It's really much simpler. And that's been our experience thus far as we enroll patients. So I think the data are so early for bispecifics, it's hard to know. We know that most of the excitement from ACR, though, has been still from CAR-T.
I think largely because of the maturity of the data there. But yeah, certainly.
So I think it's too early to make the call. I think we're still following. What we would say is what good looks like is still CAR-T. I think there's obviously the promise of bispecifics, but it's early days.
Now, when you say an allo program looks more drug-like, looks more like a bispecific, looks drug-like, are we assuming that there's going to be repeat dosing, that there's going to be repeat induction? What are you assuming in terms of, and data is so early here, but is there a cadence of redosing that you think would be acceptable for a high-priced cell therapy? And we'll talk about lymphodepletion in a second, I think, but for a cell therapy that is more challenging to give?
So I think one of the things about allogeneic cell therapy is you can get a COGM that's closer to biologics.
Yeah.
That's number one, that addresses the price. And then I think patients are going to go with what's more efficacious. And so if cell therapy still wins, it looks more like one and done or one every three years.
Sure.
If you can recapitulate that efficacy of autologous CAR-T, but with off-the-shelf convenience. To your point about repeat infusion, that seems challenging from an auto perspective to ask a patient to do that every three years.
To do lymphodepletion every three years.
Now, from an allo perspective, what's interesting is if you bifurcate, again, the gene edited versus what we do, because we're doing partial HLA matching, we have multiple lots up front. So if you can consider, if you had to go to repeat dosing, we're front-loaded to do that because we do all of our manufacturing. So we have four to six lots ready to go. I think it's much simpler to ponder repeat dosing there. My concern for gene edited, and this is maybe in the minutiae, though, if you have to do repeat dosing with one donor for multiple patients, it's a one HLA genotype. What if you have immunogenicity? Then what do you do? Do you go back to the drawing board and do 20 GMP runs?
Find a different donor.
Yeah, and then show comparability? That's an IND amendment at best. Or do you kind of say, we'll figure it out? I think it's just complicated with that approach. It's just different levels of complexity. And it's whether for the gene edited, you do it in the back end, or you try to take the biology and everything you've learned from tab-cel and do it up front.
Now, one piece of feedback that we've heard a lot, both from market and from docs who are looking to enroll in these trials, is a hesitance to put people on harsh lymphodepletion regimens that are necessary for the autologous CAR-Ts. And one thing that's interesting to me about the 3219 program is you sort of have it both ways. In the SLE trial in the IST that's gotten started, there's no lymphodepletion, which I would imagine is a great selling point for patients with less severe disease to not have to get fludarabine. But you do have full lymphodepletion in the LN. So where's the, do you need lymphodepletion? Do you not need lymphodepletion?
I think strategically, let me address the lymphodepletion cohort. That's in lupus nephritis, which is obviously a more serious branch of systemic lupus. I think we're trying to phenocopy the Schett trial.
Sure.
And say, we know what good looks like. And can we recapitulate that with allogeneic cell therapy? But because we've treated over 600 patients with tab-cel, which doesn't use lymphodepletion, we looked at that and said, I think we could try this. Because if you think about tab-cel, it's 20 years old. It's an effector T-cell. We've engineered a better T-cell in the CAR-Ts. So we've engineered the manufacturing, the phenotype. We have a different chimeric antigen receptor. We just think that without lymphodepletion, based on what we see, we have a fair shot. And to your point, when we talk to KOLs, they really are worried about lymphodepletion. We're talking about predominantly female populations. It's a younger disease. It's genotoxic.
Getting flu is really rough.
Yeah, so fludarabine and cyclophosphamide, alkylating agent or DNA damaging agent, people don't want to do that, so we're getting, to your point, a lot of interest in the no lymphodepletion cohort, and that's in moderate to severe lupus.
Have you considered reduced intensity conditioning insofar as it might be necessary to do some T-cell ablation?
Sure.
SLE only, which I know is part of standard of care for some LN patients?
Yeah, absolutely. I think we're trying to focus maniacally because I think it's easy to get diluted. So I think we're just trying to focus on the lymphodepletion and then no lymphodepletion. And what's nice about the field is that there's a lot of people who are pioneers ahead of us that will show us and get the data. And then what we can do is say what worked and what didn't work, and we try it with our platform. So I think that's an experiment that has to be done, but it doesn't necessarily have to be done by Atara.
Sure. Makes sense. So let's talk about trial start and enrollment. Obviously, you're still in early days there. But I know that other companies in this space have had INDs open for a very long time and had a lot of challenge getting patients in trial. Now, you mentioned some enthusiasm, especially for the no lymphodepletion cohort. But can you give me a sense of where you are in enrollment in the early days and given that it is so early, how your expectations are developing as you speak to PIs in the study?
I think it's a great point to make. The ideal lupus patient that Dr. Schett studied is an albatross. It's a unicorn of a patient. It's very difficult to find that patient. There's a lot of people in the space. We're still confident because of probably three things. Number one, we're based on tab-cel. So physicians kind of go, oh, this has been in patients for 600 patients. It's reasonably safe. The second thing is the no lymphodepletion angle, definitely, is something that opens their minds. And I think the third piece is really just that we are differentiated without gene editing. I don't really know exactly what that means all the time, but it's the fact that minimal modification, the autoimmune patient typically is very conservative. Rheumatology doctors are very conservative.
I would say, based on what we see, it's obviously going to be subject to recruitment, but we still think we're on track to report out initial data mid-2025.
What would that, when you say you're reporting initial data, what's a cohort meaningful enough that you would be comfortable reporting it? How many patients do you want to see?
I think, you know, an ideal world is a handful of patients where you see PK/PD, to your point. You want to see safety, which is big. You want to see that immune reset. So you're looking at the B-cell compartment. You're looking at the autoantibody titers and how durable the reduction in titers remain. Because I think that's what we're all waiting for in terms of the T-cell engagers is how long does that immune reset happen?
Yeah.
So I think, you know, we might not have six months of data, but we will at least have a sense that, hey, we see B-cell ablation.
You're getting there or not.
Yeah, absolutely.
Now, we've got five minutes left, so I want to make sure we do touch on the NHL side of the program and 3431. So maybe an NHL first. At this point, NHL obviously is crowded and more mature even in the autoimmune space. Where's the opportunity at this point, given established CD19 players? And when we see data next year, what's your go, no-go bar for further investment on it?
I think non-Hodgkin is pretty different from autoimmune as a hypothesis. The first thing I'd say is what we're seeing with our clinical recruitment is as people sweep into autoimmunity, there's more physicians who are interested in recruiting patients for large B-cell for us, which wasn't expected. Because again, you're competing against approved autologous therapy. So that was surprising. You also, and I have to call this out, T-cell engagers are changing the standard of care in large B-cell as well as myeloma. Even despite that, we're recruiting better than I would have predicted in large B-cell. What does good look like? I think when you look at the data in autologous CAR-T, I think you see 30%-40% complete response rates at six months. That's basically going to be the benchmark for us for the thesis.
But what I think the value of doing the experiment is we think we have an ethical drug. We think we have a superior product. That's number one. But number two, I think we're going to get that PKPD data that is so important for the platform to show people, hey, this is safe, first of all. But second of all, you don't see graft versus host or host versus, or sorry, graft versus host versus graft.
Yeah.
This looks drug-like. But the benefit of having the large B-cell data also is it potentially speeds up your dose escalation in autoimmunity. In other words, if you show safety at two levels, right, it's not serial one by one for the three plus three.
And then you do the staggered.
That's right. Yeah. So there's multiple reasons why large B-cell is important. And I think the last thing is autologous is so dependent on the immune fitness of the patient. So we don't know if that ceiling is because the immune fitness is because it's an older patient that's seen three to five prior lines. What does it look like when you put in healthy donor T cells? It's an open question, and we just want to answer that question in large B-cell.
Makes sense. Fabulous. If you hit those bars that you're hoping to hit in non-Hodgkin's that justify continuing in not just enable autoimmune, but justify continuing in oncology, what sort of capital do you need to then progress that into a larger phase 2?
I've heard different figures. This is very quantifiable, but it's wicked expensive.
Yeah.
I think it's one of those things where we'd have a conversation with a partner that could help us. The one thing we would bring to bear that's an advantage is we know what the path through the regulatory agencies looks like. And to be honest, it's probably simpler and easier to see than gene editing, which I think we learned from tab-cel, which again is going to be the world's first allogeneic T cell therapy. But getting it there was extremely challenging.
We all remember.
So for the whole field who is not quite there, we're looking at our lessons and saying, you know what, we're going to go with what is pivotal ready. What does a commercial product look like? And I think that's an advantage. And then we could leverage a partner to your capital question.
Sure. Okay. Well, then let's, in the last couple of minutes, talk a little bit about 3431 and moving forward into bispecific CAR space, into the more advanced, more highly engineered CAR space. So what are you hoping to see from this program versus other bispecific CARs? Maybe beyond the standard NHL comps, like are there places we should be looking to evaluate the performance of the additional target on top of what you might expect from a CD19?
I think the reason why we have a dual CAR program was we looked at the data from AbelZeta. This is the program J&J in-licensed for $245 million upfront. The overall response rate was above 90%. The complete response rate was 70% or higher. So it looked like having a dual targeting clearly broke through that ceiling of efficacy. But then I think the next question is, do you want to then build out this same network, $2 billion to do 10,000 starts? Or do you go, okay, it's differentiated, and then do you want to go allo to serve the unmet need? I think there's multiple reasons to be excited about CD19/20. So that's what we're watching closely, and you'll see that at ASH.
Yeah.
In terms of autoimmunity, I think we have a wait and see attitude. ImmPACT Bio, for example, is going there. Poseida might go there. There's actually very few allo CD19/20 programs.
Gracell had one online.
That's right. But if the data suggests we could go there, we will.
Yeah. Makes sense. Sure. I suppose if the data suggests that the NHL prospects are better, that logic would be extensible to autoimmune, you'd expect, right?
Yeah. Yeah. Absolutely.
All right. Well, we are out of time, Cokey Nguyen. Thank you so much for joining me.
Thank you.
This was a very productive discussion.
It's great to be here.