Welcome to our next Fireside Chat. I'm Robert Burns, a managing director and senior biotech analyst at H.C. Wainwright, and I'm joined by Rachel Haurwitz, the CEO of Caribou. Rachel, thank you for joining us today.
Rob, thanks so much for having me, really appreciate it.
Why don't we just dive in? From a high-level perspective, for those who may be unfamiliar with Caribou, could you provide an overview of the company and its pipeline?
Absolutely. Caribou is a CRISPR genome editing company solely focused on leveraging our technology platform to develop allogeneic off-the-shelf CAR T-cell therapies for patients with hematologic malignancies. I hope, Rob, we get a chance to discuss both of them in detail today. The lead program is vispa-cel. It's an off-the-shelf CAR T targeting CD19 for non-Hodgkin lymphoma. This program is done with phase I. It's pivotal ready, excitingly, it has demonstrated in phase I data that look just like the published data for the auto CAR Ts. We think this unlocks a tremendous opportunity to bring the benefit of cell therapy to a much broader swath of patients than can benefit from the N of one bespoke complicated process of autologous CAR T-cell therapies. Second in the pipeline is CB-011.
This is our allo CAR T targeting BCMA for multiple myeloma. This program is in the middle of phase I. We've shared some really encouraging phase I dose escalation data so far, demonstrating responses that actually far exceed our benchmark for success. We were hoping for at least bispecific-like response rates, as that's really the only other asset class that's readily available off the shelf for these patients. We've actually thus far seen response rates that are approaching those of the auto CAR Ts. In data we recently shared at EHA, we've been able to demonstrate just how durable these responses can be in late-line, heavily pretreated patients. We're actively enrolling patients in dose expansion, and we've committed to sharing initial dose expansion data by the end of this year.
Before we get to your pipeline, can you please discuss the advantages of the chRDNA platform relative to other gene editing tools and how you've used that platform in designing vispa-cel and CB-011?
Thank you for the question. I think the chRDNA technology is our key to success. The advantage of the chRDNA platform relative to other CRISPR technologies is its specificity. Through engineering new guides that are part DNA and part RNA instead of just all RNA, we've been able to improve editing specificity by several orders of magnitude. Therefore, what this has unlocked for us for our clinical pipeline is the ability to address areas of biology that have been well understood but have not been acted on for a long period of time. If you look at our approach to allo CAR Ts, it's really quite different from a lot of the companies who are a little earlier than us to the clinic, most of whom are no longer in this space because their programs did not achieve auto CAR T-like outcomes for patients.
If you think back several years, Rob, I know you'll remember this, most of those early players were largely doing the same thing, taking a healthy donor T-cell, getting it to express a CAR through one means or another, removing the T-cell receptor, and called that a product. Our team looked at that and just sort of scratched our heads because we know that these allo cells are foreign to the patient's immune system, and therefore they're going to be rejected a lot more rapidly than your average auto CAR. We felt we need to use our genome editing to address that. In our myeloma program, we do that quite literally. We actually deploy a multi-edit strategy for immune cloaking to try to slow down immune-mediated rejection to buy additional time for additional anti-tumor activity. I think that's working.
That program has PK, has persistence of the CAR T cells that's about twice as long as our uncloaked program. I think it's been key to the success that we've seen in multiple myeloma so far. With vispa-cel and lymphoma, we actually went an orthogonal direction. We're not trying to impact how long the CAR T cells persist. They're around about a month. We're instead trying to enhance their potency, and we do that by knocking out PD-1. We obviously did not invent the importance of PD-1 in CAR T or even T cell biology, but we're the first to do that with an allogeneic CAR T. Of course, the idea is to take the brakes off the CAR T cells to really drive a better anticancer activity during that window of time where the cells are present.
All the biology I just walked you through, Rob, we didn't invent it, but we are the first to have an editing technology that actually allows us to address all of these with high efficiency and high specificity, and it's what really sets our programs apart.
I completely understand. It clearly sets an advantage for you guys relative to other competitors. Coming out of ASCO and EHA, I wanted to get your thoughts on the second-line LBCL opportunity and how it might evolve over time given the data that we saw for Rapcabtagene autoleucel for the CD20 bispecifics and the frontMIND trial. Also, when we factor in ZUMA-23 and the ALPHA3 interim data that we saw earlier this year.
Great and robust question. I'll try to tackle it a couple of different ways. Maybe I'll start with what is happening in the second line setting today. Today there are about 12,000 patients treated annually in the U.S. with second-line large B-cell lymphoma. I think no one would argue that the gold standard top choice is an autologous CAR T-cell therapy. Yet only 25% of those 12,000 patients actually get an auto CAR T. It's not because there's some close trade-off with another asset class that's almost as good and maybe has some other advantages. There's a huge gap in outcomes relative to anything else. It's really driven by two key challenges that prevent so many of those patients from getting an auto CAR T. The first challenge is urgency of need. A meaningful fraction of these patients have quite significant and rapidly progressing disease.
They simply cannot wait the couple of months, three months, it takes to go through the referral patterns, apheresis, bridging therapy while their product is manufactured, and finally get their dose. They need something now. The second is geography. The majority of these patients are cared for in the community setting, yet it's really only the top-tier academic sites that are able to deliver auto CAR T-cell therapy. For a number of very understandable reasons, a meaningful fraction of patients cannot afford to pick up and move and take a caretaker with them to a large city to be close to one of these academic sites long enough to get a dose. You asked a great question about the many evolving elements of the landscape, and yes, there's some very interesting work being done, both in the second-line and in the front-line setting.
None of them, other than what we're trying to do with vispa-cel, really address either of those two key issues with respect to getting a CAR T-cell therapy with potentially curative intent to a larger fraction of patients. Vispa-cel, of course, has the tremendous benefit of being off-the-shelf. It is readily available, so that long timeline, we can just collapse down to zero using vispa-cel, and I actually mean zero. Even in our phase I study, we had some patients conclude their eligibility paperwork and begin lymphodepletion on the same day. Even some of these faster auto CAR T-manufacturing timelines don't really chip away at those much longer timelines. If you think about the bulk of the time being the blocking and tackling of referral patterns, waiting to schedule an apheresis slot, there's still a lot of QC that happens, even if the manufacturing itself is faster.
It cuts down the timeline, but not enough to bridge the gap for most of these patients. As we think about some of the front-line studies that you've highlighted, whether it's, say, an auto CAR T for very specific high-risk patients or the allo CAR T strategy that Allogene has, these are serving a very different segment of the patient population than the one we expect to serve. We don't see a lot of impact on vispa-cel's opportunity from those areas of development.
Since you presented updated data from the ANTLER trial, can you walk us through what was incrementally new and how it might differentiate vispa-cel relative to the FDA-approved CD19-targeted CAR Ts?
Yeah, great question. We were so excited that the vispa-cel abstract was picked for an oral presentation at EHA just a couple of weeks ago. What we were able to share, very excitingly, is more of the same. We were able to continue to demonstrate that vispa-cel drives auto CAR T-like outcomes for patients. This dataset is now sufficiently mature, with a median follow-up of about 16 months to measure median PFS of 17.1 months. It's actually numerically longer than the published data for the auto CAR Ts, 14.8 and 14.9 months. Ours is a smaller dataset. We'll put an error bar on it. I'm happy to say those are the same, and that's incredibly exciting. That is the very high bar we set for ourselves many years ago when we initially developed this program.
We wanted to develop an allo CAR T for lymphoma that could drive outcomes just like the auto CAR Ts, and these data continue to demonstrate that vispa-cel fits that challenge. Importantly, Rob-
Sorry
sorry. Part of what we could share as well is continued update on the safety side.
Yeah.
This is a really key piece of the puzzle. As you know, there are important differences in the safety profiles for the two commercially available auto CAR Ts, and those differences result in real-world outcomes that matter. One of those two products, axi-cel, is typically delivered inpatient as a result of the AE profile and the care model necessary to keep an eye on those patients, compared to liso-cel, which has very low rates of Grade 3 ICANS or Grade 3 CRS. Patients typically receive liso-cel outpatient. The safety data that we have for vispa-cel look extremely similar to liso-cel. Very low rates of Grade 3 CRS or neurotoxicity, and it gives us a lot of confidence for our ability to deliver vispa-cel outpatient and in the community setting.
That's definitely a differentiating factor relative to the FDA-approved CAR Ts. One of the things that we saw at EHA as well was real-world analysis comparing the CD19 CAR Ts versus the CD20-targeted bispecifics. The CAR Ts obviously won out generally, but we also saw the phase III data from the SUNMO trial, which was evaluating mosunetuzumab plus Polivy. When I looked at that dataset, we saw relatively comparable mPFS and complete response rates with only a 0.7% Grade 3 CRS. How do you think about that dataset in the second-line setting, especially now that Roche has filed its sBLA for that combo?
It's a great observation and a great question, I'll maybe address it in two separate ways. One is thematically. One of the key themes we constantly hear from KOLs is how they think about the trade-off between the asset class of CAR Ts versus the asset class of antibody-based drugs, specifically in lymphoma. What we uniformly hear from our friends in the field is that they will pick a CAR T every day of the week in front of an antibody-based therapy. When we say, "Great for us, but why?" Their answer is CAR Ts have continued to demonstrate curative potential for this patient population, whereas they are not convinced that the antibody-based drugs have true curative intent.
That's a really important trade-off, especially if you're thinking then about CD19 specifically as the target and how you make trade-offs there, but also generally, regardless of the target. We're hearing a lot about that. Specifically, for the Mosun-Pola data, one of the interesting things that's happening in the field is that front-line utilization of polatuzumab is continuing to increase. The fraction of second-line patients who are naive to polatuzumab is actually going down. I think there will be some interesting questions specifically about this regimen and who might be able to benefit, given the increasing utilization of polatuzumab in the frontline setting.
I completely agree with you there. Do you think that there might be a shift with the market penetration of Polivy now that the frontMIND trial has read out?
Yeah. It's a great question. Certainly, what we're hearing, both from our friends at community sites as well as academic sites, is that polatuzumab is here to stay in the frontline setting, and I think we should expect to continue to see greater utilization of it.
Okay. One of the things that I also noticed was that there was a meta-analysis that compared the efficacy of CD19 CAR Ts in a single target versus the dual targeted one, CD19, CD20, or CD19, CD22. Generally, the dual targeted ones won out from an efficacy perspective, although it wasn't stat sig. When we look at the competitive landscape, we're seeing several of these agents being advanced in the clinic. How are you thinking about that competitive threat with the dual targeted agents?
Yeah. It's a really important question, and I'll step back for a moment and highlight how we think about the development of vispa-cel, because it's really differentiated from all of the auto CAR Ts. Our strategy for the phase III pivotal that we're planning, and we've recently achieved alignment on with the FDA, is to focus explicitly on a patient population who is not getting auto CAR T, either because they're medically ineligible or they're facing access challenges, and these patients are also not eligible for auto transplant. We're really focused on, if you look at the commercial data, the 75% of patients who are not getting auto CAR T.
The dual targeted assets that you've highlighted, the way I think about them is they're all competing for the same 25% of patients who can actually get to an academic site and wait long enough to get an auto CAR T. I think it's super scientifically interesting to think about the role of dual targeting versus single targeting. I agree with you. It's interesting. Let's continue to watch the field. It doesn't solve the fundamental reasons why patients are not getting auto CAR T. It's not a lack of efficacy. It's a lack of access. It's a lack of being able to wait, and vispa-cel addresses those in a way that the dual auto CARs cannot.
Shifting back to the SUNMO trial. Obviously, that trial was performed in ASCT-ineligible patients. I believe the hazard ratio for PFS was 0.38, which is pretty significant. When we think about the ANTLER trial, how are you thinking about the powering and what you need to demonstrate from a survival perspective to be competitive with that agent?
Yeah, that's a great question. I'll first highlight how the study itself is designed. The study is designed, as I mentioned a moment ago, to focus on explicitly the patient population that's not getting auto CAR T and not getting auto transplant. It therefore allows us to design and run a randomized controlled trial where the control arm is receiving the grab bag of immuno-chemotherapy regimens that patients get in real life today in second line when they're not getting an auto CAR. We're leveraging four different options for the investigators' choices to contain polatuzumab for the ones who don't get pola in the front line. To the conversation we had just a moment ago, two of them don't contain polatuzumab. The total list is pola-BR, pola-R-GemOx, R-GemOx, and tafalen.
As you look at that cohort of approaches, median PFS is something like 4.5 months for these regimens. Compare that against the 17.1 months we've seen for vispa-cel, I think that should get anyone excited. Perhaps even more importantly than just where does the curve cross the midline, it's what is the actual shape of the curve. If you look at the vispa-cel efficacy data, you see an auto CAR T-like plateau. You see more than 40% of patients maintaining these incredibly long, durable responses, which is so exciting to see. Whereas with these other regimens, it is just a drift back down to baseline. The only question is exactly how fast it gets there. With a 250-patient population, we believe we have powered this such that we have quite a bit of power to detect success in almost any scenario.
Why don't we shift gears now to CB-011. Can you update the investors as to what you presented there and how it stacks up against the BCMA or the GPRC5D, what have you?
Great question. Obviously, there's a lot going on for patients in multiple myeloma, and it's also such a humbling disease to think about how many different lines of therapy these patients chew through and how complicated these combinations can get. We're really excited about the role that CB-011, as an off-the-shelf CAR T, could play in being a one-and-done therapy that could provide patients a treatment-free interval. That is really valuable in the landscape of drugs for these patients. Specifically with respect to the data that we've shared, and most recently at EHA, again, we were so thrilled that this abstract was also accepted for an oral presentation. We've been able to tell a pretty robust story from our phase I dose escalation data set. We've enrolled 48 patients in dose escalation.
Through that, we were able to evaluate two different lymphodepletion regimens, multiple cell therapy doses, and at the end of last year, we had disclosed which dose we had chosen as our recommended dose for expansion, a single dose of 450 million CAR T cells. At the end of last year, as part of that initial data update, we were able to share, we already had a dozen BCMA-naive patients enrolled in that RDE cohort, and we were able to share efficacy data on that cohort. That's where the theme came from that I mentioned a few minutes ago of being able to see almost auto CAR T-like response rates in these patients, where we were hoping to at least meet the benchmark of bispecific-like responses.
Just recently at EHA, we were able to provide an update with several additional months of follow-up on those same 12 BCMA-naive patients. The story only gets better.
Yeah.
As a result of further deepening of a response, we're now up to an 83% CR, SCR rate, 92% overall response rate, and I think most importantly, really encouraging evidence of durability. Half of those patients remained in response at 15 months and counting. If you think about this late-line, heavily pre-treated patient population, that is a remarkable outcome. What we've also started doing since then is enrolling patients in dose expansion. I should say since the disclosure at the end of last year, we've been busy working on that this year. Our dose expansion work continues to focus on BCMA-naive patients, and it is also including BCMA-exposed patients. We know that the utilization of BCMA-targeted assets is only increasing and getting earlier and earlier in the myeloma treatment landscape.
As we think forward in the not-too-distant future, we believe the vast majority of late-line myeloma patients will have had at least one, if not multiple, BCMA-targeted therapies. We're really eager to see what role CB-011 can play in providing benefit for those patients. The one other thing we were able to tuck into that EHA data disclosure was a single-patient case study from a patient who'd had prior BCMA-targeted therapy and then came on to CaMMouflage, our phase I study. This patient is the perfect example of how humbling multiple myeloma is. He had had 8 prior lines of therapy. He actually achieved a CR following frontline therapy and then never again achieved complete remission following any one of the additional lines of therapy, including CARVYKTI. He had a dose of cilta-cel, to which he achieved a VGPR.
He ended up on CaMMouflage for his ninth line of therapy, and remarkably, at day 28, was already in complete remission. As of the data cutoff for the EHA presentation, he was out a little past three months, still in CR. Early evidence of the benefit CB-011 can bring those patients, too.
I must say that's quite impressive, given his ninth line of prior lines of therapy, and he already got cilta-cel as well. When we think about the extensive competitive landscape in multiple myeloma, how are you thinking about the go/no-go signal for CB-011 in dose expansion for both the BCMA-naive and BCMA-exposed patients?
I'll start with BCMA-exposed. I think that's probably the best near-term opportunity for CB-011 to really compete in the broader competitive landscape. As I mentioned a few minutes ago, we expect that this late-line patient population will be heavily BCMA pre-treated.
Yeah.
As you know, as the field continues to demonstrate, it is really hard to find efficacious agents for patients who've had one, let alone multiple, prior BCMA-targeted therapies. We think this is where CB-011 has a really unique role to play because whether it's an auto CAR, a bispecific, or an in vivo CAR, all of those approaches necessitate the patient's own T-cells to do the heavy lifting. It's why this chronic utilization of these different assets results in dramatic worsening of the T-cell compartment and worse and worse efficacy in later and later lines of therapy. CB-011 is completely different. It is built on healthy donor T-cells, and therefore, we are able to just completely bypass the patient's immune system. We don't need it to do the heavy lifting. We can provide the healthy donor T-cells.
Perhaps even for that one patient case study that we've shared already, perhaps that's part of the reason why he's able to experience a complete remission now in such a late line of therapy.
That completely makes sense from a scientific perspective. When we think about the data that's coming at the end of this year for dose expansion, how should we be thinking about that data set?
Great question. It will be both dose expansion in BCMA-naive and dose expansion in BCMA-exposed. I think both of those stories stand on their own as we think about the potential to serve both kinds of patients. We're excited to leverage those data to make some of our initial go, no-go decisions around the best place for future development of CB-011. I'll also highlight, Rob, this program recently received RMAT designation from the FDA, that will facilitate our ability to more proactively engage with the FDA and get some of their feedback as we think about development next steps. We expect to have a first engagement with them ahead of this data disclosure.
Okay. Sounds good. Obviously, there's been a lot of enthusiasm with regard to the in vivo CAR T space, and we saw data from Legend Biotech's in vivo CD19/CD20-targeted agent, LB2501, and then Eli Lilly's KLN-1010. Where do you think that that modality will fit into the treatment paradigm of hematologic malignancies, and where does allogeneic therapies slot in then?
Yeah, great question. Obviously, it is extremely early days in in vivo land. That data set that you just referenced, I think had 12 patients, six of whom actually.
Yeah. Very few.
were in the cohort where they saw some efficacy with very short follow-up. Certainly, I think intriguing. We'll be keeping an eye on it. I know many others will. Especially with vispa-cel, we have a dramatic time advantage, right? We have a pivotal-ready program and can see line of sight to potential commercial utilization of vispa-cel on the other side of an approval. Zooming out for a moment, not just thinking about the blocking and tackling of time, we see this incredible value of healthy donor T cells to do the anticancer activity, whether it's in myeloma for the reasons we talked about a few minutes ago, or in lymphoma. We're really excited about an allo strategy really being beneficial to these broader patient populations. I'll also highlight, Rob, obviously one of the key challenges for auto is just how expensive it is, right?
Yeah.
Cost of goods sold and the time associated with it is very costly to the healthcare system and has created some significant challenges as we think about the businesses that are being built with these programs. vispa-cel and CB-011 are, again, completely different. I joke with my team sometimes I wish we could call our therapy something other than cell therapy because they're just so different. With vispa-cel, for example, we anticipate cost of goods sold at launch to be 96% lower than the auto CAR Ts.
Oh, wow.
It's a cheap biologic at that point. If you think about some of the numbers that are being quoted for the COGS of in vivo CAR Ts, they're actually the same, if not a little more-
Yeah
expensive than what we anticipate for vispa-cel. We really feel like vispa-cel has the best of both worlds.
How are you thinking about potential pricing of vispa-cel then, given the lower COGS that you guys have?
Important question. I think I'll have to say probably a bit premature for us to comment specifically on pricing. However, what I can say is that COGS profile gives us tremendous flexibility as we think about a pricing strategy for these different patient populations.
Okay. Last question from me. Could you remind us what your end 1Q26 cash was, and what sort of catalyst can we expect other than the CB-011 update by the end of this year?
Yeah, great question. We last reported almost $120 million on the balance sheet, and that funds our current runway into the second half of 2027. That gives us the ability to do the dose expansion work that we've been discussing for CB-011, and as we've highlighted, get to that initial dose expansion data disclosure by the end of this year. It also allows us to invest in the critical startup activities to get ready for the ANTLER-3 vispa-cel phase III study.
Awesome. Well, I really appreciate your time, Rachel. Thank you for joining us today.
Thanks, Rob.
Awesome.