Great, thanks everybody. Luca is a senior biotech analyst here at RBC Capital Markets, and today is our great privilege to have Caribou for Fireside Chat. Representing the company, we have Rachel Haurwitz, co-founder, President, and Chief Executive Officer of the company. Rachel, thanks so much for joining us today. How are you doing?
Doing great, Luca. Thanks for having us. We really appreciate it.
Super, super. I have a long list of questions here, but maybe before we ask you about individual programs, can you just maybe give us a little bit of a big picture update of what progress has the organization made over the last few months, and most importantly, what's ahead here for Caribou.
Absolutely, it's been a big year for us already and quite a bit more yet to come. So already this year we've dosed the first patient in our AML CAR-T trial, which we're super excited to have that up and running. And also recently we disclosed that we're now expanding the clinical development of our lead asset CB-010 into autoimmune in addition to the ongoing oncology work that we're continuing to do. So excited to have the green light from the FDA for that IND, and we anticipate initiating that lupus trial by the end of this year. And stay tuned; in just a few weeks we'll be seeing a bunch of folks at ASCO where we'll be sharing the initial dose expansion data from CB-010 and second-line large B-cell lymphoma patients.
Excited in particular to be able to dig in on some of the data we've talked about for a few weeks now, really this intriguing observation that patients who by chance received partially HLA-matched product are having better outcomes, and it'll be an opportunity for us to dig in and really talk about key next steps including enrolling a new 20-patient cohort to investigate this in further detail that we anticipate will underpin the phase three trial we intend to initiate next year.
That is super helpful. Maybe before we go into the HLA matching component, which is obviously very important, maybe bigger picture, when we think about ASCO, walk us through a little bit of a data preview. What do you think would be a kind of base case scenario, worst case scenario, and upside scenario for Caribou as we see that data.
So we've enrolled 30 patients in dose expansion, these are all second-line large B-cell lymphoma patients. We've actually explored three different dose levels, the same three dose levels we looked at in dose escalation initially, though you'll see a bias towards the one that we've selected as the recommended phase 2 dose. So this will be a chance for us to look at safety data, efficacy data, translational data including PK and PD, and to disclose the RP2D. And on top of all of that we'll be able to give an update on the patients from dose escalation who remained on study as of our last update about a year ago.
How many patients were still in dose escalation? Was it seven?
Seven yes.
Okay, so you're going to show us data only for dose 7 or also? Okay.
Well, good question. So we'll be able to update for those who remained on study where they are now, how they are doing. And the overall HLA matching observation, it's true across the board, and so we'll be able to hit on what we've observed across escalation and expansion.
Got it, got it. You may have already answered this question, but when you think about HLA matching, were you testing a hypothesis here? Like, were you proactively testing for HLA matching, knowing that there was a possibility that that was going to be maybe better data than non-HLA matched or partially HLA matched? Or was this more like a chance finding as you were slicing and dicing your patients across the different subgroup? I think that's an important dichotomy, so any call there would be helpful.
Yeah, I love the question. It's a biological question we had from day zero, and in fact we set up the study so that we could intentionally collect HLA typing information on every patient and every donor. It's obviously taken us some time to collect a large enough data set. Now between escalation and expansion we're talking about 46 patients, and that's given us enough data to really have a robust signal where we do think there's clearly a there there that this partial HLA matching is driving better outcomes for patients.
Super helpful. Are all the patients tested in the trial so far received T cells from the same batch or for multiple batch and if it is multiple batches can you just give us a sense of how large of a sample size do we have do we have again in the context of only 37 patients but how generalizable is this finding so any color on how many batches you use to dose these 37 patients?
Yeah, multiple batches, so we've used a little more than 6 different batches of material in the clinic for those 46+ patients to date.
Gotcha, gotcha, very helpful. And then I'm just kind of reflecting on HLA more broadly. I believe except for like identical twins, right, HLA matching comes with shades of gray, right. So what can you tell us about patients that are closer to a perfect match, maybe seeing better data than patients that are a partial match? I think that's going to be important to see, you know, kind of the totality of the evidence. Any, any thoughts on that one.
Yeah, I can tell you what we'll be able to share at ASCO, and then I can hit some key themes today. At ASCO, you'll be able to see us take our data and slice it by different levels of partial matching, and we'll be able to do that both, for example, for efficacy as well as PK. What I can share with you today ahead of having some of those discrete data in front of us is we're not talking about bone marrow transplant, right? We're not looking for these people's long lost twins. It's a fairly modest partial match that drives these better outcomes for patients.
Gotcha, gotcha. Helpful. And maybe for the audience, can you maybe remind us how are you testing for HLA? And I'm assuming this is a simple saliva test. One, is that correct? And two, is that commonly reimbursed by payers? Like, how should we think about that part?
Yeah, it's really interesting, so it can be done either through a blood draw or through a cheek swab, so a couple different ways to get the data, 1-2 day turnaround, so this is super fast. And it's not typically done for lymphoma patients today unless they are being considered for an allotransplant, in which case it would be considered standard of care in that context.
Okay, okay, that's helpful. Are you also testing for HLA antibodies? I'm just the only reason why I'm asking is because, you know, if you have received a transfusion at some point in your life, you may have antibodies against HLA subtypes is different than your own HLA subtype, so how and I think that's regularly tested for when we think about transplant, so are you are you testing for HLA antibodies or how should we think about it.
Absolutely, it's a great question. Yes, so another key piece of the puzzle is going forward we'll ensure that patients do not receive a product to which they have donor-specific antibodies. So what are donor-specific antibodies? They're a subset of anti-HLA antibodies, right? Any of us for a variety of reasons may have anti-HLA antibodies. They're only relevant if they target one or more of the antigens that are expressed on the surface of the product. So again, this is data that we can use to make sure that we're selecting the right dose for patients that both meets our partial HLA matching threshold and make sure that they're not donor-specific antibodies either.
Gotcha, also like you know obviously women after having kids they also have other antibodies against other HLA subtypes so I think that's quite important. Can you maybe talk about the inventory? I think you mentioned that you're confident you can run the next pivotal trial in allogeneic; however, can you give us a sense of how many batches off the shelf you kind of need at this point to actually execute on such trial.
At ASCO I can be even more specific than I can today, but suffice to say it's a fairly modest number of batches from distinct donors that we will need to be able to supply a phase 3 as an all-comer where we would have confidence that we could partially match the vast majority of patients on study. To help bookend that a little bit, you can think about the fact that we intend to have that in hand by next year so it gives you a sense for how quickly we can build that inventory. But also as we're looking backwards at the data we'll be sharing at ASCO, by chance about a third of these patients received sufficiently partially matched product so that also gives you a sense that it can't be too high a matching threshold.
Got it, got it. Helpful. Maybe one last one on HLA subtypes, you know, obviously HLA subtypes depends on demographics and obviously in the U.S. we have the benefit of having folks that comes from historically all over the world but maybe in other geographies this can be different. How should we think about that especially in the context of clinical trials? Unfortunately there are some minorities that are not represented in some of those clinical trials so how should we think about that in the context of your clinical trial as well as how you could potentially use this in clinical practice.
Yeah, really important questions. Look, the fact that the U.S. is a melting pot is a huge win in this specific context, as it is for many other contexts as well. So we get to draw from a very diverse donor pool as we think about building our inventory of CB-010 going forward. So we've certainly factored in some of this thinking as we think not only about deploying CB-010 in clinical trial sites here in the U.S. but also ex-U.S. as well, and already today we're working with sites in Israel and in Australia.
Got it, got it, very helpful. I think you already alluded to this earlier, but I want to make sure I capture it. Obviously, responses is what matters, but what can you tell us about the PK, PKPD, I should say, and how consistent is the data between the partial HLA matching not only with the clinical evidence but also with the PK and the evidence of T cell expansion? There any thoughts?
Yeah, these are great questions. Really, the PK and the efficacy hang nicely together and tell a consistent story, so excited to dive into those details at ASCO as we think about the role of partial HLA matching. ASCO will also be the first opportunity for us to share PD data. We've never shared pharmacodynamic data before, and some of these data are actually really underpinning our enthusiasm for deploying CB-010 in lupus as well. So what we've described thematically and we'll share the data in detail at ASCO is that the patients on ANTLER, our oncology trial, have experienced B-cell aplasia both in terms of depth and duration on par with what Dr. Georg Schett and his colleagues have published in terms of the lupus patients they've treated with auto-CARs
So we think that's a really compelling piece of evidence to suggest the role that CB-010 could play in treating lupus patients as well.
Gotcha, so you're basically using T cells for PK and B cells for PD. That's kind of a simple way to think about it.
Exactly.
Okay, okay, that's actually very helpful. What are what are realistic expectations for that data set? Again, I appreciate you already have a pretty compelling data so far. Now you have this new wrinkle with the HLA matching. Is it best case scenario that your data set matches what we've seen for autologous CAR-T or can there be a scenario where this data is even better than what we've seen from or is that is in unrealistic expectations because if you get the T cells from the same individuals and you're infusing them back in the same individuals you're always going to have better data versus allogeneic like any thoughts on on that front.
I'll set expectations for ASCO and then maybe we can talk about the future. Our fundamental objective for CB-010 has always been and continues to be to develop an allogeneic CAR-T cell therapy that can meaningfully rival the autologous CAR-T cells. And so I would say as you look at the ASCO data in a few weeks, really focus on the subset of patients who received partial HLA matched product and we believe that gives a strong signal that we're on the right path in terms of meeting our objectives.
Now in terms of measuring this, this is obviously still an ongoing study, whereas last year we were able to wait with dose escalation until we had at least six months to follow up on all 16 patients. This is still an ongoing study. Not all of these patients have hit the six-month mark yet. So you'll see us use PFS as a metric to measure the durability outcomes for these patients and to use that to benchmark against, for example, ZUMA-7 and what YESCARTA has done for patients in the second-line setting. Now I will caveat all of this by saying we're also working in a really interesting patient population. Actually, the majority of patients who've been put on this portion of our trial, their physicians told us they picked this trial because they thought these patients could not wait for autocars.
Now, I'm not telling you we're then taking the ZUMA-7 benchmark and factoring it down by some multiple or some fraction. I'm still telling you the bar is the same in terms of the objective that we are aiming for with this asset.
Got it, very helpful. Again, we're going to see that at ASCO, but how are you thinking about dosing? When I step back and look at the broader field of allo-CARs, I can't say that I've seen a clear dose response across multiple data sets, so in the context of this not being a traditional oral small molecule or monoclonal antibody, how do you even pick the dose? Like, walk me through your thinking here.
Yeah, there's more of a thresholding effect than maybe a linear dose responsiveness as you look across a number of cell therapy assets. We have made our RP2D determination. We'll disclose what that is at ASCO, and I'll say the totality of data went into that decision. We looked at safety. We looked at efficacy. We looked at PK. We looked at PD. We used every bit of information that we had to inform the appropriate determination.
Got it, got it, very helpful. I think you spoke about this in the past, and obviously we need to see the data, but how are you thinking about potential for a trial design the second line settings? Obviously TRANSFORM and ZUMA-7 you already mentioned to either the two comps here. I was always under the impression that you were leaning more toward a design similar to TRANSFORM and ZUMA-7. Any thoughts on how you're thinking big picture of a potential for a phase 3 and design there.
Yeah, that's still consistent. So we had a discussion with the FDA towards the end of last year where they endorsed use of a control arm that looks a lot like both of those trials, actually. So that would be chemoimmunotherapy followed by stem cell transplant. The two trials are actually quite similar to each other. Probably a key difference to highlight is that TRANSFORM allowed a crossover design whereas ZUMA-7 did not, and I think that drove the meaningful differences in patient numbers. TRANSFORM was concluded with a much smaller end than ZUMA-7. So suffice to say, we're trying to take as many learnings from those experiences as possible as we plan for the phase 3.
What's your view on CAR-T in general between late line and early line? I'm under the impression looking at some of these more novel approaches that in early line settings when maybe there's less tumor burden to begin with maybe the data is better, which is somewhat counterintuitive if you will, but that's what some of these trials are telling us. Is that consistent with your thinking or your understanding and that's maybe one of the reasons why you decide to move earlier in the treatment paradigm.
Yeah, we're really excited that in the lymphoma setting a few of the autocars have been successful in moving earlier and earlier. Clearly, as you compare the second-line and the third-line outcomes, there are, it appears, modest improvements in the second-line setting in terms of outcomes over the third-line, and perhaps that's not terribly surprising in terms of catching the disease earlier in its progression. And so we're very excited that we got the green light from the FDA to do all of our dose expansion work in second-line patients. As far as I know, CB-010 is the first and still only allocars being developed in the second-line setting, and we see tremendous potential for those patients.
Got it. I know we have a limited amount of time here, so maybe I'll pivot to I&I for a minute, reflecting on that New England Journal of Medicine publication actually showing fantastic data out of 15 patients, right? That's all we have, and the other day we did the math on how many companies have actually pivoted to I&I in some sort form or shape, and we actually counted 22 companies. So the question is, is this getting too crowded too fast? That's question one, and two is this the structural problem that we have in biotech more broadly where you have like one scientific breakthrough and then you have like so many companies going after the same well, like any thoughts on everything I just said.
Yeah, it's a lovely set of questions, and I'll answer it through the lens of why did we choose to jump in. And obviously it started with the academic data, which are, to your point, small n but profound outcomes for these patients. It appears that some sort of, you know, B-cell aplasia driven immune reset can have just an incredible outcome for these patients.
So why why us? Why CB-010? We're able to leverage the tremendous work that we've done in oncology. You know, we've dosed, as I said, more than 46 patients at this point, so we've built quite a good understanding in terms of safety, efficacy, and some of these key translational data points that give us a lot of confidence to say the encouraging data we've seen in oncology have a lot of crossover potential into lupus and potentially other autoimmune indications as well. So we get to leverage a lot of those learnings and capabilities. I think it de-risks some of the work that we're doing because we already know CB-010 can drive the kinds of durable responses in non-Hodgkin lymphoma that the autocars can drive, and as I shared a moment ago, we've seen B-cell aplasia on par with what Dr. Schett and his colleagues have shared.
No easy for me to say I drank the Kool-Aid a long time ago but I'm a deep believer that in the autoimmune setting an allogeneic strategy is really important over an autologous approach for this broader patient population for a number of reasons. As we look at that space there are actually very few allogeneic cell therapies that are moving into the clinic for these lupus patients and we're excited that CB-010 is one of the first ones. We also get to leverage all of our ongoing oncology work which includes of course 30 open clinical trial sites some of which could be great partners for us for our work on lupus as well which we think could give us a leg up in terms of actually executing on the lupus trial.
Got it, super helpful. Maybe when I think about I&I and you know allo CARs it feels to me that in oncology what matters is T cell persistence versus maybe in I&I all that matters is T cell expansion so one is that consistent with your understanding and two if that is correct is there an opportunity to maybe take off some of the lymph depletion that we're using and maybe use lower doses you know cyclophosphamide like again how are you thinking about T cell expansion versus T cell persistence and how are you thinking about implications for potentially using a different type of lymph depletion regimen versus what we have been using so far in oncology.
I'll cut to the chase and say that's what we've done on lymph depletion, so we're actually using a less deep lymph depletion approach in lupus compared to what we've used in oncology to date. Same two drugs, cyclophosphamide and fludarabine, we're simply using more of each in oncology relative to autoimmune. I would say the way I think about the role of persistence, it's maybe a little bit different from how you articulated it in oncology. I think about functional persistence, so for example with CB10 we've knocked out PD-1. That shouldn't change the PK and data that we shared, you know, some time ago at EHA. We demonstrated that within about 30 days CB10 is gone in peripheral blood, so it's not about persistence in time but it's the functionality while it persists. And our expectation is that the knockout of PD-1 really takes the brakes off the CAR-T cells and enables them to pack a better anti-tumor punch during the period of time while they are present. And I think a similar thought process can be brought to autoimmune as well. Fundamentally, it's really the same mechanism of action, right? You're trying to get rid of the B cells.
Super helpful, just maybe a quick follow-up here. When you mention lower doses of lymph depletion, are you referring to lower doses of lymph depletion versus CD19 or versus BCMA? Because obviously for CD19 you're using higher doses versus BCMA, so how should think about both that.
Versus the CD19 oncology approach that's right.
On par with the BCMA.
I think it's a bit more than what we're using in the myeloma setting right now. If you look across the cell therapies that are being brought into the clinic for lupus, there's a pretty tight range of the doses of Cy and flu that everyone is using, and we're in the same ballpark.
Okay, that's helpful. That's helpful. Let's maybe talk about CB-011, maybe the program that doesn't get the amount of airtime that it deserves. So maybe just walk us through why you're excited about that approach and maybe what's next for that program.
Yeah, so CB-012 is our AML allogeneic CAR-T going after CLL-1. We dosed our first patient in Q1 of this year, so very excited to have that program in the clinic in patients. I think AML is the perfect example of a disease where you have to develop an allogeneic cell therapy. I think your average AML patient has such poor quality T cells that we're not going to see autologous CAR-Ts broadly serve that patient population. I also think target selection is really critical in this patient population, and we got really excited about CLL-1 in particular because it's highly expressed on AML tumor cells. It's even expressed on leukemic stem cells, but importantly it is not expressed on healthy hematopoietic stem cells, so it sets up the long-term vision that CB-012 could potentially have disease modifying activity on its own and not necessarily have to be a bridge to transplant for these patients. Some folks at MSK had done quite a lot of work on this target. They'd even developed fully human binders for it, so we've exclusively in-licensed those for CB-012 and are using one of them in the asset in the clinic today.
Got it, super helpful. Maybe in the last minute, let's touch upon BCMA. Maybe just walk us through differentiations versus what's out there. Obviously, Legend sets a pretty high bar for everybody in this space, so as we see that data toward the end of the year, what's a go no-go decision to continue to invest in that program versus maybe prioritizing other programs at this point.
Perfect question, so we ask KOLs this question all the time. What do they need to see from an allocars for it to be important and relevant for their treatment approaches? And what we hear from them uniformly is that they would love to see an allocars in the same ballpark as the bispecifics. That's the only other broadly available off-the-shelf approach, so that means, you know, let's call it 60%-70% overall response rate. And importantly, they have expectations that an allocars would have lower infection challenges than what their patients experience with the bispecifics today.
Got it, got it. Super helpful. Maybe last one: how are you thinking about business development at this point? Not that we have to revisit the whole AbbVie situation, but like, walk us through how you're thinking about BD going forward.
Yeah, great question. Look, we fully own all of our clinical stage assets today, which I think is a tremendous advantage for an organization like us. It gives us a lot of optionality as we continue to think about how best to deploy our resources, you know. For example, we ended Q1 with $346 million in the bank, so strong balance sheet and ability to continue advancing those programs while balancing it with thinking about if and when might be the right time to bring in a partner to either bring in expertise or to accelerate something that we couldn't do on our own, to say nothing of the opportunity that our entire genome editing and cell therapy platforms bring in terms of the potential for new preclinical or discovery stage programs as well.
Got it. I thought it was the last one, but last one is now any update on the Chief Medical Officer search at this point. Anything you can share with us at this point?
Yeah, excited by the progress that we've made, you know. I'm a picky person with a high bar and really pleased with the caliber of candidates we're able to interact with.
Fantastic, a lot more questions but no more time. Rachel, thanks so much for joining us. Thanks, everyone, for joining the conversations and all the best with the rest of the conference.
Thanks, Luca. I appreciate it.
Anytime, thank you, Rachel.