Hi, everyone. My name is Liisa Bayko, biotech analyst at Evercore ISI, and today I'm here with Rachel Haurwitz. She's founder and CEO of Caribou. Caribou is working on allogeneic cell therapies. And Rachel and I were just talking about kind of all the disruption in the CAR T space this week with, you know, kind of FDA's commentary on some, you know, kind of talks it's been seeing. But maybe you can, Rachel, just let's just get that out of the way and speak about that right now. Give us your two cents. Actually, and I...
What I was also telling Rachel is, like, I immediately started looking at the stocks and seeing who was affected, and really the allogeneic, you know, kind of cell therapy companies were pretty much unaffected by all this news in terms of the impact on the stock. So I'll turn it over to you. Maybe you can just kind of reflect on it and tell us how you're thinking about-
Yeah.
- the data.
Certainly, Liisa, thank you. Yeah, I think a lot of the chatter has been about, you know, whether lentiviral vectors or other retroviral vectors that are used in making auto CAR-Ts are potentially contributing to some of these SPMs. So to your point, I think that's a really important distinction between auto and certainly what we do in the allogeneic setting. We don't use lentiviral vectors to make our off-the-shelf CAR-Ts. In fact, we take advantage of what we call our chRDNA technology here at Caribou, which is a highly precise genome editing technology. We can do site-specific insertion of our CAR. We can do site-specific insertion, hard to say, of other transgenes as well, while maintaining genomic integrity. So I think we have a really great platform for making these kinds of products.
Okay. Are you kind of worried about any kind of spillover as this gets further investigated into the field, or you pretty much feel this is limited to the lentiviral vector thesis?
I mean, I think these kinds of questions are important to ask regardless of the asset class, right? I think this is always super important. Then you have to put it in context, too, right?
Mm.
We understand, for example, that chemotherapy and radiation is incredibly toxic for patients, yet it's also incredibly effective in a lot of these settings. And so I think it's about putting these understandings in the appropriate context.
Okay, fair enough. So I guess just to start off, just to get everyone on the same page, you talked about your chRDNA technology, but give us a sort of broad strokes overview of Caribou.
Caribou is a CRISPR genome editing company, and at the heart of our organization is what we call the chRDNA technology. It's a far more specific version of CRISPR relative to CRISPR-Cas9, and it allows us to do some of the sophisticated work that I just highlighted a moment ago. Today, we are exclusively focused on oncology. We're developing two cell therapy platforms in parallel: allogeneic CAR-Ts, with a focus on heme malignancies, and allogeneic CAR-NKs, with a focus on solid tumors. We now have three clinical-stage allo CAR-Ts, so a lot of exciting work ongoing.
Great. So let's kind of give us an overview of your portfolio, and then we can go into your lead compound.
Certainly. So those three allo CAR-Ts, creatively named CB-010, CB-011, and CB-012, are focused on three different heme malignancies. CB-010 is for non-Hodgkin lymphoma, CB-011 for multiple myeloma, and CB-012 for AML. In each case, we're really excited to use our genome editing to enhance or armor, if you will, the CAR-T cells to really try to boost their anti-tumor potential. If we take a few steps back, you know, we are fundamental believers that off-the-shelf has to be the strategy to bring CAR-T to broader and broader patient-
Right
... populations. And yet, they're inherently foreign to the patient's immune system and will be fairly rapidly immune-rejected, and so that's why we focus so much of our energy and effort on using our genome editing to enhance these cells. We have different strategies we use to really try to match the biology of the disease to appropriate approaches. For example, in our lead program, we knock out PD-1 to try to prevent premature CAR T cell exhaustion. In multiple myeloma, we actually immune cloak the cells to try to keep them in circulation longer. And then in our AML program, which just recently got the green light from the FDA for a phase I trial, we do both of them together, the PD-1 knockout and immune cloaking.
So let's talk about CB-010, I guess. Tell us about the design of the CAR.
For sure. So we use the same binder that many others do. It's the FMC63 anti-CD19 binder, and we've really built a bespoke CAR around that, having done a tremendous amount of work evaluating a number of the different elements to get to a CAR that we believe is the most appropriate for this setting. We site-specifically insert that into the TRAC gene in the T cell genome, so that really gives us extra bang for our buck. Not only do we get the CAR in a particular place, we also get rid of the T cell receptor through the same edit, which is obviously very important from a safety perspective. And then further, we knock out PD-1 from the CAR T cell genome.
Okay, great. So, tell us more about the data that you've seen so far?
We were really excited about the positive data that we shared in the middle of this year. So we had concluded dose escalation following dosing 16 patients across 3 different dose levels, in every case, a single-dose strategy. And we shared in the middle of this year that across those 16 patients, we saw a 94% overall response rate, a 69% complete response rate, and I think most importantly, 44% of the patients achieved a complete response at 6 months or greater. There's nothing magic about a 6-month CR, but it is a tool in the field that is highly predictive, typically of much longer term outcomes for these patients. So to put 44% in context, with the auto CAR-Ts in the third line setting, about a third of patients achieve a CR at 6 months.
So seeing that 44% of these patients did was very exciting. I'll be cautious, it's an n=16, right? But I think definitely indicates that CB-010 can drive durable responses that rival those of the autologous CAR-Ts.
Super exciting.
We agree.
Okay, so, one of the things I know that you're implementing is a you know a more substantial lymphodepletion strategy. How important do you think that is to the overall efficacy? Is that... There must be something specific that you really need for allogeneic. Is that fair to say? And then-
It, it goes back to the same point where we were talking about a few minutes ago, right? We know that these, these CAR-Ts that are off the shelf, they're foreign, they're going to be immune-rejected. And so a member of our scientific advisory board, who had a lot of experience with this particular lymphodepletion protocol, strongly recommended that we use it, with the goal of creating just a slightly bigger window for initial engraftment and anti-tumor activity. Now, it's the same two drugs that everyone's familiar with. It's cyclophosphamide and fludarabine. It's simply a bit more of each. And we didn't invent it. It's been around more than a decade, developed at the NIH, and used extensively in a number of clinical trials across multiple different assets.
We had a lot of confidence in its safety and utility prior to deploying it in the ongoing ANTLER trial.
Tolerability's been good so far?
It has been, absolutely.
Do you think you can re-dose with this approach?
It's a really interesting question. I will say everything we're doing today is a single-dose strategy, and a departure from that would be data-driven. But operationally, the answer is yes, right? That's, I think, one of the benefits of an allogeneic CAR T, is you have product sitting in the freezer, so you could think about a second dose. Unlike the auto setting, where it's often difficult, if not impossible, to actually do that.
I was just actually specifically thinking about with this lymphodepletion strategy, do you think you could actually give it twice? Would it be safe enough to do that?
Potentially, and I think it also opens the question of, what is the gap in between those doses, and might you be able-
That's fair
... to do them both within a single round of lymphodepletion?
Okay. Hmm, that's an interesting thought as well. Okay. My predecessor here, Josh, had this very strong opinion on, you know, that, that maybe these... It was the lymphodepletion component of the fludarabine side that was actually kind of causing, like, contributing to the efficacy. What do you have to say about that?
I think physicians would love it if a few days of chemotherapy would result in multi-year complete responses for these patients. They don't, right? And so I think it's pretty clear that the long-term outcomes that we can see are not driven by a few days of chemotherapy.
You're very specifically focusing on second-line patients as well, right?
Correct.
Okay. Do you think you could go later in more advanced patients as well?
We have, right?
Okay.
We started in dose escalation, where the vast majority of patients, 12 of 16, were third line or later. Though 4 of the patients in dose escalation were actually primary refractory, so technically second-line patients. Based on really the totality of data, both safety and efficacy, emerging from dose escalation, we got very excited about what we were seeing. We went to the FDA and proposed that we do our dose expansion work focused solely in second-line large B-cell lymphoma patients, and they green-lit that, so that's what we're doing today.
Okay, great. And how does your data so far compare to other allogeneic approaches for CD19?
We think the data compare pretty favorably to what many of our peers have seen. We haven't seen other large, robust data sets where all patients on study are resulting in these kinds of incredibly high overall response rates, or the kinds of six-month CR rates that really rival those of the auto CAR Ts.
So what's your plan to kind of like... Your development plan from here now for this program, Rachel?
Fabulous question. So our focus today, as we just talked about, is second-line LBCL patients, and our objective, our goal, is to ultimately run an initial pivotal study in those second-line LBCL patients. And so the data we're collecting today in dose expansion are really meant to be the foundational data set for driving that. One of the things we've committed to publicly is to provide feedback by the end of this year from discussions with the FDA about exactly that pivotal strategy.
Okay, and so when do you think that could start then?
Fantastic question. Stay tuned for the update.
Okay. Anything we should expect from this program at the upcoming ASH meeting?
We are not sharing-
Okay
... CB-010 data at ASH. We have committed to-
Mm-hmm
... the initial look at dose expansion data in the first half of next year.
Okay, great. Let's move on to your next program, CB-011. I have CB-012 here. Well, anyway, CB-012 is for multiple myeloma, right? Maybe I skipped over one.
CB-011 is multiple myeloma.
Oh, I got it wrong there.
Yes.
Sorry.
No worries.
I was confused. Okay, so tell us about the design of this CAR. I know you talked about the immune cloaking and all of that.
Yes, so CB-011 benefits from a proprietary binder. We actually ran an analysis, a campaign, years ago, evaluating a number of different BCMA binders... and this one really stood head and shoulders over the rest. It was developed by an antibody company. It's a humanized binder, so we actually acquired it, and the IP rights there, too, and have incorporated it in the design of the cell therapy today.
Okay, great. So PD-1 knockout is a really interesting approach. Tell us a little bit more about, like, why you're doing that and exactly what you think it can bring to the table?
Yeah, so maybe ten steps back, as we initially started building our cell therapy pipeline here at Caribou, we thought a lot about the fact that a healthy T cell plus a CAR is just not enough. And so how do you appropriately enhance or armor these therapies to have the right anti-tumor activity? There are obviously many different potential solutions. PD-1 knockout is one of them, immune cloaking is another. And so we also thought a lot about the biology of any given disease and pairing appropriate strategies that match that. So for example, now CB-10 in Non-Hodgkin lymphoma, we use the PD-1 knockout there because the PD-1 axis has been demonstrated to be really important in CAR T cell biology in Non-Hodgkin lymphoma. It's probably not surprising the majority of those tumors are PD-L1 positive, and they also reside in the lymphatics.
The endothelial lining is rich in PD-L1, PD-L2, so that makes a lot of biology sense. The same observations did not exist in the myeloma setting, and so we felt that was a better place to evaluate the immune cloaking strategy, and so that's why it's part of CB-011 today. But I'll fast-forward to CB-012 and just point out, CB-012, which is now moving into the clinic for AML, actually benefits from both of these together, which I think really speaks to the strength of our underlying genome editing platform, right? To make CB-012 requires five edits, three separate knockouts, two separate site-specific gene insertions. I think it's really the Cas12a chRDNA platform that unlocks the possibility of making that kind of therapy.
What's the max in terms of, like, these kinda edits you can get to, do you think? Like, do you think you've maxed out at five?
I doubt we've fully maxed it out, but there's definitely some sort of ceiling, right? As you think about any one of these edits, they're never 100%. And so as you think about then the entire mixture of T cells at the end of a manufacturing run, you've only got a fraction that's a... You know, it's just sort of pure math. It's a fraction-
Yeah
... of a fraction of a fraction.
Right. Right.
We have an internal standard that we'd like to see about half of all of the T cells, at least half of all of the T cells, with all of the consequential edits. We can do that with CB-012 at scale. There might be opportunity to push it further from there, but I think this sort of max in terms of editing really points to the value and the potential importance of an iPSC-derived strategy, which is our approach for our current case, where in the solid tumor setting, you probably need even more armoring, even more edits. And so to achieve all of that at scale, we think an iPSC is a very compelling way to go about that.
Let's talk a little bit about your... Well, let's finish off on these. So for... We talked about CB-011, CB-012. When should we expect to see data from these programs? I know the last one is just going to IND status now, so but when could we expect to see some data?
Great question. So CB-011, the myeloma program, we dosed our first patient at the end of Q1 this year. We've disclosed we've now safely concluded dose level one. We're actively enrolling patients at dose level two. We haven't yet circled a date on the calendar, and really, it'll be data-driven, as we want to make sure we have a sufficient amount of information to start to answer some of the meaningful questions about this program, both safety and efficacy. For CB-012, well, we recently got the green light from the FDA to move this program into the clinic for AML, so the team is extremely busy right now as we're getting sites up and running. And we anticipate initiating enrollment by the middle of next year.
Okay, great. And for your NK program, how are you thinking about dosing there? And, what kind of like... As you shift into NK cells, what kind of other considerations do you have to think about?
Yeah, maybe I'll take a step back and say why NKs and therefore how we think about some of these. CAR-Ts have demonstrated obviously incredible anti-tumor potential in the heme setting, and yet the same really hasn't been true in the solid tumor setting. And so as we spent a lot of time thinking about how to build an immune cell therapy in this context, we really felt that the biology of natural killer cells was a really compelling place to start. NK cells inherently have anti-solid tumor activity, including anti-metastatic activity, so a really nice start point. And then you can layer on a CAR and a number of different genome edits to really enhance that kind of activity.
And now, as you think about sort of the blocking and tackling of how to get all those edits in, it's what took us to the iPSC, as we talked about a moment ago, where at the iPSC stage at Caribou, we can do a multiplicity of genome edits and then pick out a single cell clone that has all the attributes we're looking for, and then using proprietary protocols that we've developed at the company, differentiate that edited iPSC into a CAR-NK. We are still in preclinical research for this particular platform, so a bit earlier than the allo CAR-Ts. But certainly looking at other people's CAR-NK data, I think there's really an interesting difference as you think about the safety profile of some of those programs.
That probably opens the door to the potential for repeat dosing in a way that we simply haven't seen to be necessary in the CAR-T setting.
Why is that?
... I think they're fundamentally different cells, right? How they target, how they proliferate, how they persist or don't, actually varies pretty dramatically across these different cell types. And so I think you get to take advantage of the differences in biology to really, in a bespoke way, build a program that can match the need of the tumor biology.
Do you worry about the difficulty of getting the CAR-NKs into the sort of solid tumor environment? There's obviously a lot of resistance there.
It has to be part of what we go after with genome editing, right? Thinking about how to target, clearly a CAR can be an important part of that. But also how to really enhance the overall biology of the natural killer cells to begin to address the incredibly immune-suppressive tumor microenvironment. Absolutely.
How, how are you thinking about durability and things along those lines with this cell type?
I mean, look, I think it's known that NK cells typically do not persist as long as T cells, and so I think you just have to think about other ways to use the toolbox. And so some of that might be genome editing to enhance the persistence of them, and another simply could be maybe a multi-dose strategy, as most appropriate in that context.
Yeah, what's your angle on kinda how your CAR NKs will differentiate from sort of others in the field?
Right. So I think many have approached CAR-NK development by taking a natural killer cell, whether it's a primary or blood-derived NK, or an iPSC-derived NK, adding a CAR, and giving that a try. Our goal is to use these in the solid tumor setting, where we recognize, you know, many of the challenges you've highlighted, we will have to address or overcome to have appropriate anti-tumor activity. And so, our kind of right-off-the-bat thinking is, this has to be a quite sophisticated product with a multiplicity of edits that can address some of these different challenges simultaneously.
Okay, great. So, Rachel, just in the last 30 seconds we have or so, give us a preview of 2024 for Caribou.
Absolutely. Strong balance sheet, so we have $397 million, a runway for our current plan into the fourth quarter of 2025. So really excited to continue executing on our three clinical stage programs. We've already committed to the first dose expansion update for CB-010 in the first half of next year. I'm obviously very excited to continue dosing patients with CB-011 in the multiple myeloma setting, so stay tuned for initial data there. And for CB-012, very much looking forward to getting our first patient on trial next year.
Great. Thank you so much!
Thank you.