Good afternoon, and welcome to the Jefferies Healthcare Conference. My name is David Tamburri. We're excited to have you guys. Today, we have Rachel with us from Caribou, and we're super excited. Thank you.
Well, good afternoon, everyone, and a huge thank you to the Jefferies team for the opportunity to participate and to present today. My name is Rachel Haurwitz, and I'm the President and CEO of Caribou, and one of the Co-Founders of the company. And I'm very excited to share a repeat of some of our very recent data release at ASCO just a few days ago. Of course, I will pause and acknowledge that this presentation includes forward-looking information. At the heart of Caribou is our next-generation CRISPR genome editing technology, the chRDNA technology, that allows us to carry out far more specific CRISPR genome editing than first-generation CRISPR-Cas9. Today, we are using this technology to advance a wholly-owned pipeline of off-the-shelf CAR- T cell therapies, three of which we are evaluating across four different clinical trials. We are well-resourced.
We ended Q1 of this year with $346 million cash on hand, allowing us to continue progressing these four clinical stage assets with a runway into the first quarter of 2026. As I mentioned, we have four clinical stage programs evaluating three off-the-shelf CAR- T cell therapies, and today we'll focus in particular on CB-010, our lead program, for which we shared quite a bit of new data just a few days ago at ASCO. CB-010 is an off-the-shelf CAR- T cell therapy targeting CD19, and it's being evaluated in the ongoing ANTLER trial for relapsed or refractory B-cell non-Hodgkin lymphoma. It is also soon to be evaluated in the GALLOP phase I trial in lupus, and we recently received the IND clearance for this program and anticipate initiating the trial by the end of this year.
Also in our pipeline are CB-011 and CB-012. CB-011 is an off-the-shelf CAR-T cell therapy for multiple myeloma targeting BCMA, and it's being evaluated in the ongoing CaMMouflage phase I trial. We anticipate sharing initial data from this study by the end of this year, so stay tuned. CB-012 is our off-the-shelf CAR-T cell therapy for AML, targeting an antigen called CLL-1. We dosed our first patient in this phase I study in Q1 of this year, so please stay tuned as we continue to dose patients and make progress in dose escalation. Across our pipeline, we leverage our CRISPR genome editing technology capabilities to edit, to armor, to enhance the anti-tumor potential of our cell therapies. We have multiple distinct strategies that we bring to bear, including checkpoint disruption, immune cloaking, and the combination of these two approaches.
As far as I know, we are the first to use each of these different strategies in the clinic for allogeneic CAR- T cell therapies. At Caribou, we fundamentally believe patients shouldn't have to wait for treatment, and this has driven our focus on using genome editing to develop off-the-shelf CAR- T cell therapies that can be readily available to patients and available to a much broader patient population than are served by the auto CAR- Ts today. So I'd like to spend the majority of our time today focused on CB-010 and highlighting some of the key updates that we shared just a few days ago at ASCO. CB-010 is an off-the-shelf CAR- T cell therapy targeting CD19, and it benefits from three genome edits. One is, of course, to remove the T-cell receptor to reduce the risk of graft versus host disease.
A second is to site-specifically insert the CD19 CAR, and a third is to get rid of PD-1. As far as I know, CB-010 is the first allogeneic CAR-T in the clinic with a PD-1 knockout, and we believe it really sets it apart in the competitive landscape. The goal of the PD-1 knockout is really to take the brakes off the CAR-T cells themselves, to allow the CAR- Ts to retain a high anti-tumor activity state for a longer period of time. CB-010 is being evaluated in the ongoing ANTLER phase I clinical trial. So I'll use this slide to tell a little bit of a story of what we've done, what we most recently disclosed at ASCO, and some of the key next steps in this particular trial.
About a year ago, we concluded dose escalation after evaluating 16 patients, largely third line and later, each of whom had never had a prior CD19 targeted therapy. We evaluated three different dose levels: 40, 80, and 120 million CAR- T cells, in each case as a single dose regimen. We disclosed these data in the middle of last year, and the headline was that CB-010 can drive the kinds of durable responses achieved by the auto CAR- Ts. We were very encouraged by these data and actually took them to the FDA and asked permission to leapfrog into the second-line setting for our continued clinical development, and they greenlit that. So we've conducted all of our dose expansion work exclusively in second-line large B-cell lymphoma patients.
We've since enrolled more than 30 patients into dose expansion, and just over the past few days at ASCO, we provided an update on 46 patients, including the 16 from dose escalation and the first 30 from dose expansion. Because we moved patient populations, we did additional dose-finding work and actually evaluated all three of the same dose levels in dose expansion. Based on the totality of data, we selected 80 million CAR- T cells as the recommended phase II dose. We also used this data set to answer a question that we've had since before this trial ever started, and that question was: Could partial HLA matching between the patient and the donor who offered the T-cells for the CAR- T product have an impact on outcomes for patients? And the answer is yes.
What we observed is that for patients who received a dose of CB-010, where the patient and the donor shared at least four HLA matches, that's four out of 12, so pretty modest compared to transplant, those patients are having enhanced efficacy. We're very encouraged by this, and I'll share both efficacy and PK data that support these observations today. Based on this, we'd like to prospectively confirm this signal. We've created a new 20-patient cohort, which we'll begin enrolling this month, and we'll ensure that each of those 20 patients intentionally receives a dose of CB-010, where the patient and the donor share at least four HLA matches, and we expect to read those data out in the first half of next year. So I'd like to start by actually looking backwards first.
Part of our update allows us to really dive into just how durable outcomes driven by CB-010 can be. So, as I mentioned, about a year ago, we shared the initial dose escalation data for that full cohort, 16 patients. So what I'm showing here on this slide is all 16 of those patients and now updating by almost a year, so from a data cutoff date in June 2023 to now April 2024. Very encouragingly, we're seeing incredibly durable responses in some of these patients. For example, almost a year ago, there were 4 DLBCL patients still on study in complete response, and now, as of this year's cutoff date in April, all 4 of those patients remain in complete response. So we can get, this is incredibly encouraging evidence of the very durable outcomes that can be achieved with CB-010.
In fact, we already have 2 patients who have successfully concluded the 2-year study in complete response, and we're very excited to see that. Now, the second part of what I'd like to highlight today allows me to zoom all the way back out to all 46 patients who are part of the update at ASCO. As I mentioned, we were able to evaluate 3 different dose levels across these 2 different cohorts, and all 3 dose levels were generally well tolerated. As I mentioned also, we've selected 80 million as the recommended phase II dose. So now diving into this partial HLA matching observation, there were 13 patients who, by chance, received a dose of CB-010, where the patient and the donor shared at least four matched HLA alleles.
For those 13 patients, we've seen a median PFS of 14.4 months, which is incredibly encouraging for this population. I'll dig into these data in detail in just a moment, and I'll also put it in the context of the auto CAR-Ts, because our fundamental objective for CB-010 is to develop an allo CAR-T that can meaningfully rival the autologous CAR-Ts. Based on these very encouraging observations, we're taking forward this partial matching strategy, both in our oncology work and in our lupus trial. Here are the key data that led to our determination and our selection of a minimum of four HLA alleles to match between donor and patient. What you're looking at on the left side of the slide is all 46 patients across escalation and expansion, and we've evaluated progression-free survival and done so across three different cohorts.
So in light gray are patients who received a dose of CB-010, where there were fewer than 2 matches between the donor and the patient. Dark blue is 2-3 matches, and light blue is 4 or more. Again, thinking backwards from our objective of meaningfully rivaling the auto CAR-Ts and the progression-free survival that can be achieved with those products, we identified that a minimum threshold of 4 or more matched alleles achieves that objective based on these data. Now, looking at the right side of the slide, very similar analysis, and this time we've laser-focused on the 40 LBCL patients spanning dose escalation and dose expansion. Again, very similar conclusion that 4+ leads to very encouraging outcomes for these patients. And in fact, the median PFS in that cohort has not yet been reached as of the data cutoff date.
As I mentioned, I wanted to put these data in context of the auto CAR-Ts. So what we've done on this following slide is take the emerging data from ANTLER and try to graph it on top of the data from ZUMA-7. ZUMA-7 is the trial that led to the approval of Yescarta in the second-line setting. Now, I recognize there are a number of caveats associated with making cross-trial comparisons, and I'll highlight a few of them. ZUMA-7 was obviously a concluded study by the time this analysis was done. ANTLER is an ongoing trial. We continue to enroll patients. There were 180 patients in the Yescarta arm in ZUMA-7, and there are 11 LBCL patients in the ANTLER cohort that I'm sharing with you here. Also, the timelines are a little bit different.
So day zero in ANTLER is the day of infusion of CB-010, whereas day zero in ZUMA-7 was the date of randomization, since this was a randomized controlled trial. Also, the first efficacy assessments occurred at different time points, 28 days in ANTLER and 50 in ZUMA-7. All of that said, I think the overlay of these two PFS curves is incredibly encouraging and I hope is emblematic of why we're very excited about the potential that partial HLA matching brings to CB-010 and our objective to meaningfully rival the autologous CAR-T cells with this off-the-shelf strategy. I'd like to dig in for a minute on some of the key patient characteristics for the patients who've been enrolled in this study. And I'll take a step back and share a few stories from some of the physicians who've enrolled patients, in particular, in dose expansion.
Actually, a majority of the physicians who put patients on dose expansion have told us that they specifically chose this trial for patients who could not wait for commercial CAR, which tells you a little something about how aggressive and fast-moving their disease must be if they cannot wait 4, 6, 8 weeks for their auto CAR- T product to be manufactured. And I think this is well understood as you look at some of the quantitative metrics that we can share here. For example, if you look at IPI score or LDH at screening, and even just compare between the dose escalation and dose expansion patients, you can see that a high proportion of the dose expansion patients have quite poor prognostic indicators, including 70% of the dose expansion patients having an IPI score of 2 or greater.
And yet, nonetheless, the data that I just showed you demonstrate that with partial HLA matching, CB-010 can achieve outcomes that meaningfully rival the autologous CAR- Ts. I'd like to highlight safety, as this is critical as well. We're very pleased to see across all 3 dose levels that CB-010 achieves a generally well-tolerated safety profile. In particular, we have seen no cases of Grade 3 or higher CRS across this 46-patient population, which is very encouraging, and no cases of GvHD. We've focused on a handful of key AEs, shown on this slide here, and have shared some of the relevant applicable data from ZUMA-7, evaluating Yescarta as well. If you look at, for example, prolonged cytopenias, it looks like the rates with CB-010 are, are modestly lower than with Yescarta.
Infections are in the same ballpark, and I think quite encouragingly, the rates of Grade 3 ICANS are meaningfully lower with CB-010 than with Yescarta. So turning now to a swimmer plot that focuses on the 13 patients who, by chance, received a dose of CB-010, with at least 4 HLA alleles matched between the donor and the patient. We can see just how durable some of the outcomes have been for the patients who've been on study the longest, and we're very encouraged by the overall response rate for this patient population as well. And it's, it's these data that allowed us to calculate the median PFS of 14.4 months, which again, is very encouraging and comparable to what the auto CAR- Ts can achieve as well.
So zooming back out for a moment, this chart between the left column and the middle column represents all 46 patients, bifurcated either in the left column to those who received a product with 3 or fewer HLA matches, or the middle column, 4 or more. And I think really what's most remarkable is the 6-month PFS metric. It is very clear in these data that a minimum match of 4 alleles is driving meaningfully more durable outcomes for these patients, and we're very encouraged by these data. So I'd like to share some of the PK data that we've collected from these patients, because I think it fundamentally supports the mechanism of action for what we believe is occurring and for why the partial HLA matching is having really important impacts for these patients.
So what you're looking at on the left-hand side of the slide is PK aggregated for all patients on study, and it looks exactly like you would expect. Peak expansion around 7-10 days post-dosing and persistence through about 30 days. But I think the right side of the slide is far more interesting. Here, we've broken up the patients by the same three groups, fewer than 2 matches, 2-3, or 4+. And what you can see is quite stark. If you have an insufficient number of HLA matches, PK is quite negatively impacted, both in terms of peak expansion and persistence. Now, this makes sense. Our expectation is that with insufficient numbers of HLA matches, the patient's immune system is able to quite rapidly reject the therapy and therefore prevent appropriate antitumor activity.
In terms of executing on this partial matching strategy, I'm very pleased to share that it will not in any way impact timing to treatment for patients. We are able to quickly determine a patient's HLA type through a simple blood test or a cheek swab, get those data within about 48 hours, and that empowers our team to select the appropriate dose sitting in the freezer to be shipped to the patient's site, such that this does not in any way impact timelines for treatment for these patients. As we fast-forward to our plans for a phase III pivotal trial in second-line large B-cell lymphoma to kick off in the second half of next year, we have modeled that we need about 13 batches of CB-010 to be able to give approximately 90% of patients a dose with at least four HLA alleles matched.
So we are on our way to creating this inventory such that by the time we initiate the pivotal trial, we would have at least 13 batches on hand. And therefore, our intent is to structure a phase III trial that would be open to any eligible patients, regardless of their HLA type. And so to wrap up on CB-010 and oncology today, I'll highlight again the very encouraging progression-free survival that we've seen, a median of 14.4 months, for the patients who've received a partially HLA-matched dose of CB-010. We're very excited to now open this new 20-patient cohort to prospectively confirm the signal that we've seen, and we'll read out those data in the first half of next year to underpin the initiation of our pivotal trial in the second half of next year.
Two other updates I'd like to share on the oncology front. The first is that for the first time ever, we are actually going to be opening a new cohort to patients who've had prior CD19-targeted therapy and relapsed. We understand this is a growing patient population with very high unmet medical need, and the encouraging safety and efficacy we've seen thus far in CD19-naive patients has motivated our decision to open a 10-patient proof of concept cohort for this population as well, and we plan to share those initial data in the first half of next year, too. Finally, I'm pleased to share that the FDA has given us the green light to now administer CB-010 in the outpatient setting.
So we're very excited to work with actually the majority of sites who we have in our trial today, who already have the capacity to deliver CAR- T outpatient. So with our last few minutes, I'll quickly highlight our three other clinical programs. The second is deploying CB-010 in the context of autoimmune, and we've been heavily motivated by the encouraging safety and efficacy profile in oncology to also deploy CB-010 in autoimmune. I'd like to share one other key piece of translational data that led to our decision to bring CB-010 into the autoimmune setting as well, and that's PD. In this context, specifically, B-cell aplasia. So what you're looking at here is data from our oncology patient cohort, where we're evaluating not only B-cell levels, but also T and NK cell levels.
I think very encouraging, first off. We see quite prolonged and deep B-cell aplasia. In fact, across this cohort, on average, the patient's B cells are below the limit of quantification for an average of 114 days, and on average it takes about 268 days for their B cells to come all the way back up to normal. That's in strong contrast to the T and NK cells, which rebound rapidly within about 3 weeks. We think that's very important in terms of otherwise reconstituting the patient's immune system to give them a, a healthy and functioning ability to fight infections. Now, there's a little bit that we can learn from the small number of patients who've been dosed in academic trials with auto CAR-Ts from the B-cell aplasia that they have experienced. So, drawing on the most recent publication from Dr.
Georg Schett and his colleagues, they demonstrated that their autoimmune patients experienced on average 112 days of B-cell aplasia. Uh, as I just shared, we have seen 114 days, so I'll call this approximately the same, and we think very encouraging for de-risking the use of CB-010 in the autoimmune setting. We're very excited that we have an open IND, and we are on track to initiating the GALLOP phase I study in lupus nephritis and extrarenal lupus by the end of this year. We have two other clinical studies open that I'll briefly highlight and then open the floor for questions. The ongoing CaMMouflage phase I trial is evaluating CB-011, our off-the-shelf CAR-T cell therapy for multiple myeloma, and we are in progress in dose escalation as we speak.
We have concluded dose levels 1 and 2 with no DLTs, and we're actively enrolling at dose level 3, and we plan to share first clinical data from this program by the end of this year. And finally, CB-012 is being evaluated in the AMpLify phase I trial in patients with relapsed or refractory AML, targeting an antigen called CLL-1. We dosed our first patient in the first quarter of this year. We're continuing to enroll patients, and please stay tuned as we share updates on making progress in this study. So finally, to wrap up, a number of key upcoming clinical catalysts. By the end of this year, we will have our first data readout on CB-011, and we plan to initiate the lupus trial for CB-010.
In the first half of next year, two key data readouts for CB-010, one will be the 20-patient cohort of second-line large B-cell lymphoma patients who are CD19 naive, who are all going to be receiving product that has at least four HLA matches, and the second cohort will be our first-ever CD19 relapse patients. So very excited to share those data in the first half of next year, and that work will underpin our ability to initiate the phase III pivotal trial in the second half of next year. And with that, thank you very much, and I'd be happy to take your questions.
I can start with, is the HLA matching specific to your CAR- T or does it apply across the board?
Yeah, great question. For those who didn't hear it, the question was whether HLA matching is specific to CB-010 or might apply to other CAR-Ts. So far as in my expectation is that this is general to allogeneic CAR-Ts. I can't imagine why it would be just specific to CB-010. So I'll say selfishly even, we're starting to look at how we can borrow some of these learnings from CB-010 in both CB-011 and CB-012, though I'll note both of those programs benefit from our immune cloaking strategy, which gets rid of all endogenous Class I presentation.
And then for the pivotal design, did FDA ask to see this data before you can start? Because the data seems clear-cut to start your pivotal.
Great feedback. Thank you. Actually, it was, it was our decision to do this. We felt it was prudent to prospectively confirm this signal in a slightly larger patient population ahead of making the go/no-go decision for the resource-intensive phase III trial that we are excited to initiate in the second half of next year.
Then the other question is, how does it impact the lupus nephritis study? Same...
We are taking these learnings directly into Lupus, so actually we baked the partial HLA matching strategy into the protocol for that program.
Thank you so much.
Hi, thank you. Is a higher level of matching likely to be subject to diminishing returns, or is that just too difficult to do?
It's a fantastic question. I think by chance, the highest match we've seen on studies so far is maybe 6 alleles, 6 out of 12. Certainly our objective is to meaningfully rival the auto CAR- Ts and at a threshold of 4 or more, we see data that do that. Stay tuned. We'll see what kind of matching levels we're able to collect in this next dataset to see if we can comment further than that.
Okay. Good luck.
Thank you. Great. Well, with that, I believe we're-