Precision BioSciences, Inc. (DTIL) Earnings Call Transcript & Summary

Welcome to the Precision BioSciences CAR T Program Update Conference Call. [Operator Instructions] Thank you. Alex Kelly, Chief Financial Officer. You may begin your conference.

Good morning, everybody. This is Alex, and thank you for joining our conference call today to discuss our CAR T results for both our Azer-cel lead program for potential first-in-class program for CAR T relapse patients and also an update on PBCAR19B stealth cell which is designed to be our potential best-in-class allogeneic CAR T program. So this morning, I'd like to remind you before we begin that we may make some forward-looking statements during the call. And those forward-looking statements have been identified in our document. You can see them here on the screen. You can also refer to our most recent 10-Q for all the risk factors that may cause our actual results to differ materially from any statements we make today. With that, let's move on with the call. We have two speakers today. First, we'll have Michael Amoroso, our Chief Executive Officer and then also Alan List, our Chief Medical Officer, will join for the clinical update. So with that said, let me hand the call over to Michael Amoroso.

Good morning. Thank you, Alex, and welcome to our investor community. Thank you for joining us this morning. Today, Dr. Alan List, our CMO, and myself, will walk you through our 2023 CAR T updates. We'll give you an update on the totality of our lead candidate, Azer-cel, the experience to date starting in non-Hodgkin lymphoma and adult acute lymphoblastic leukemia, but ultimately focusing on the CAR T relapse setting in diffuse large B cell lymphoma. I'll tell you a bit more this morning about this setting, the urgent need in this growing population. Alan will then walk you through all of the clinical data updates from our Phase Ib expansion with Azer-cel. And then he'll move on to our second-generation product, PBCAR19B stealth cell and tell you about our Phase I readout, where we've established a safe and potentially effective dose. Ultimately, we will give an overall update on the progress from our cell therapy platform, one half of our business. The other half, of course, being our in vivo gene editing with ARCUS. Now for today's conference. First, Azer-cel, focused on the autologous CAR T DLBCL relapsed setting. Over the past few years, Precision has generated a robust data package across 84 patients with non-Hodgkin's lymphoma or adult acute lymphoblastic leukemia, showing meaningful activity and acceptable safety profile. Dr. List will tell you and give you updates on that safety profile today from our new regimen. The data was most compelling in our signal search into diffuse large B cell population and equals 18 patients who are CAR T relapsed. In this setting, Azer-cel achieved an 83% overall response rate, a 61% complete response and durability. A duration for the responders at 6 months, 55% of patients in response were ongoing. In our latest cohort that Dr. List will share with you today, Azer-cel's safety profile has been ameliorated where you will see no Grade 3 or greater allogeneic CAR T-related side effects. Next step with Azer-cel. We will meet with the FDA in June. We will discuss the potential path forward for a Phase II study at all recommended dose of 500 million cells of Azer-cel plus flu-cytoxan 750. The focus of our meeting objectives will be on study design, trial size and endpoints. Next, we'll move on to our second-generation CD19 positive CAR T programs for CAR T naive patients in the earlier line aggressive lymphoma setting, DLBCL, PBCAR19B stealth cell. In our Phase I trial, safety and preliminary efficacy was on par with today's approved autologous CAR T therapies. Stealth cell achieved a 71% overall response with no allogeneic CAR T-related adverse events, Grade 3 or greater. More specifically, in the diffuse large B cell population, an 80% overall response was achieved with a 60% MRD-negative complete response, full molecular remission. The 19B stealth cell proof of concept, the immune cloaking to overcome T and NK cell host immune rejection has been achieved. Finally, building upon Precision's strong cell therapy foundation. Precision has now optimized our cell therapy manufacturing platform using ARCUS for CAR T insertion at the TRAC locus and it has now been validated in 2 clinical candidate programs. Precision's platform, and organizational cell therapy capabilities, are primed and ready for advancement in hematologic malignancies, solid tumors, with potential for autoimmune disease. Next, what you will see is the total body of evidence for Azer-cel. Across 84 patients, across hematologic malignancies, non-Hodgkin lymphoma and adult ALL, across doses and different LD regimens, Azer-cel has demonstrated strong activity. Our goal over the past year was to truly hone and tailor that activity for the right patient at the right dose with the right optimized Azer-cel product. That patient is the auto CAR T relapse patient population, as you can see on Slide 6. In 18 patients treated with prior CAR T therapy, Azer-cel has achieved an 83% response rate, 61% complete response and durability of 55% of these patients still an ongoing response at 6 months or longer. In fact, the median duration of the responders in the 6-month group was 431 days. Azer-cel has the potential to become a standard of care in this population with dire need. As we advance to Slide 7, I think it's important to discuss why would a CAR T product work in a diffuse large B cell patient who has failed prior autologous CAR T. And there is a biologic rationale that really may support our proceed. This patient in the auto CAR T relapsed setting has impaired immune integrity when making their autologous CAR T. This patient doesn't fail CAR T therapy. Unfortunately, their CAR T therapy fails them. Their poor immune integrity leads to product attributes of an autologous product in this patient that are not optimal for getting the disease at bay. Azer-cel from a healthy donor may be more effective than making an autologous product in this setting. It's also important to understand why this patient does not progress. Many questions have come in from my investor friends on, do these patients still have CD19 positive disease? And the short answer is an emphatic yes. In fact, from all prospective data across all of these patient set, 85% of all patients who progress with diffuse large B cell lymphoma in the later line setting, have CD19-positive disease as a primary driver. Next, it's important to set the stage for, unfortunately, how dire and poor the prognosis is when you progress from a prior auto CAR T therapy. There is no approved standard of care in the major markets of the world today. The data you see before you is from the U.S. consortium and some other cuts of real-world data. When you aggregate the data in patients who have failed prior autologous CAR T therapy, you see response rates in the range of 20% to 30%. Unfortunately, those responses don't last long, as you can see that progression-free interval is no greater than a median of about 1.8 months. Overall survival in this population is 4 to 6 months in the treated population. However, that's only 3 out of every 4 patients who make it to next therapy. If you're not in that group, 1 out of 4 go straight to palliative care, their survival is less than a month. This aggregates for about 3 to 4 months of survival in this population, dire need. And of course, anything we put in the scheduled population, a major reason why they're excluded from those trials is very difficult, and safety has to be paramount and first in this population. I'd like to share with you today what we believe is a go-forward target profile, a bare minimum threshold that we have to overcome, and we'll have this conversation in the upcoming weeks with the FDA. This is in no way to highlight our data today. but we want to show you what has been our North Star as we develop and Dr. List moves these programs forward. From a response rate standpoint, we're targeting at least half the population in a response. That duration of response has to last at least 3 months, 50% longer than we see with 1.8 months today. Median survival, we want to see at least half the population greater than 6 months. We are truly looking for advancement in this population. And of utmost important, especially for our presentation today, we must see no treatment-related Grade 5 events in this population or at least that's the North Star goal. When we last left you at the midyear point last year, this was an area that we needed overtly to clear, and Dr. List will share our success in this area today. Finally, the CAR T relapse population is a large and growing group. The advent of autologous CAR T, Yescarta, Breyanzi, Kymriah is one of the greatest breakthroughs of hematologic malignancies. That being said, only 3.5 out of every 10 patients will be cured. That means 60% to 65% of the population will move onward with their disease, mostly with CD19 positive disease. And right now, they have dire poor outcomes. This market will only grow in the years to come in the world major markets. As you know, Yescarta has moved to a second-line standard of care. Even in that setting, there is still similar recurrence rates as you see on this slide. With no further ado, Dr. List, can you tell us a bit about our Azer-cel updated data?

Thank you, Michael, and good morning, everyone. Let's move on to Slide 12. You may recall that in our ASCO update last year, we shared data showing that Azer-cel was highly active in CAR T relapsed subjects conditioned with cytoxan intensified with the depletion regimens, demonstrating 100% response rate and preliminary providing evidence of durability among 11 evaluable subjects. The challenge that remained in this heavily pretreated population was safety, which is needed to be improved for an acceptable registration path. Slide 13. Since then, we have improved the therapeutic index to position Azer-cel as a potential Phase II registration study. Specifically, we continue to optimize our manufacturing process to further improve product attributes and maximize cell potency. We also had favorable feedback from the FDA on our CMC process changes, in which they shared alignment on our manufacturing changes and the analytics. And most importantly, to improve safety, we reduced lymphodepletion dose intensity while continuing to maintain efficacy. Next slide. Two specific changes are responsible for the improvement in therapeutic index. First was a reduction in total fludarabine exposure by removing the fourth dose of fludarabine, and secondly, manufacturing optimization, [indiscernible] product with improved fitness as measured by the non-apoptotic fraction in addition to a controlled reduction in the CD4/CD8 ratio that lessens the risk of inflammatory toxicity. Slide 15. We assessed 2 lower-intensity lymphodepletion regimens, beginning first with standard lymphodepletion, which is 3 days of fludarabine and cytoxan 500 milligrams per meter squared for 3 days, and a cytoxan-intensified variant, what we call FluCy750. What you see on this slide are the CAR T kinetics assessed by flow cytometry for standard lymphodepletion, in the light blue, and FluCy750 in the dark blue. Standard lymphodepletion yield with only modest CAR T expansion and area under the curve. Because of that, we pivoted to FluCy750 based on our data showing that homeostatic cytokines, which drive CAR T expansion, are maximized at a cytoxan dose of 750 per meter squared without increasing inflammatory cytokines that contribute to toxicity. FluCy750 gave rise to a 58-fold increase in CAR T peak expansion, reaching a median of 44,000 per mL and a 51-fold increase in the area under the curve, compared to standard lymphodepletion without a change in the adverse event profile, which I'll share with you shortly. Let me first share the safety data from the lower dose LD regimens. On Slide 17, on the left-hand side of this table, you can see the toxicities observed in the cytoxan-intensive regimens reported last year. And on the right, are the adverse events with stand-alone completion in the FluCy750 regimen. Using the optimized product, there are no adverse events specifically related to Azer-cel or the LD regimen that specifically included Grade 3 or greater CRS, ICANS, infection or GvHD. Importantly, there were no treatment-related deaths with either standard lymphodepletion or the FluCy750 regimen. Next slide. As you can see on this slide, compared to the autologous CAR T FDA-approved products, the CAR T-related adverse event profile for Azer-cel with FluCy750 in this heavily treated CAR T relapsed patient population compares quite favorably with that reported for the approved products in the third-line setting. Going to Slide 19. I'll now share with you an update on the durability of the ASCO 2022 subjects as well as the correlation with molecular response to Azer-cel treatment. The response to the first 6 patients shown here were initially shared at ASH 2021, reflecting results of patients treated with the enhanced lymphodepletion regimen and 3x10^6 cells per kilogram CAR T cell dose. As you can see in this swimmer plot, the updated swimmer plots, 3 of these 6 subjects remained in remission beyond 6 months with 2 remaining in remission beyond 1.5 years. The second cohort, shared 1 year ago, which received a higher flat cell dose of 500 million cells with the modified lymphodepletion regimen, which is 4 days of flu therapy and cytoxan750, 3 subjects also reached day 180 or longer in response. While for 2 of these subjects, D and K investigators elected to consolidate with allogenic stem cell transplant. And these subjects remain in an ongoing response well beyond 1 year. Overall, there was an overall response rate among the 11 evaluable subjects with a corresponding -- 100% overall response rate with the corresponding CR rate of 82%. 6 or 55% of those evaluable patients remained in response to day 180 or later, and 4 or 36% remained in response beyond the year. Safety remained an issue in this cohort. However, excluding the treatment-related deaths, 57% of subjects remain disease-free more than 1 year. Compared to the active cell consortium experience, the response rate, duration of response and progression-free survival with Azer-cel in this cohort was far superior in outcome. Slide 21. Knowing that our most recent cohort has limited duration of follow-up, we included data on molecular response. And -- or as you can see in the ASCO '22 cohort as well as the most recent data set. As you can see in this figure, which is the ZUMA-1 axi-cel trial data, clearance of circulating tumor DNA by day 28, measured using the adaptive PCR-based clonoSEQ assay was a powerful predictor of durable response to axi-cel treatment. Overall, a molecular response characterized by achievement of early MRD negativity was more powerful than day 28 PET scan metabolic response. Okay. On -- this is Slide 22. As you can see in the updated swimmer plot from last year, durable clinical response is closely aligned with early clearance of circulating tumor DNA. Overall, among the 8 clonal type of valuable subjects, 7 achieved a molecular response or MRD negativity by day 28, suggesting deep and durable responses to Azer-cel. Let me now share with you the results of treatment in the most recent cohort of 7 subjects receiving reduced lymphodepletion intensity with a dose of 500 million cells. The overall and molecular responsive treatment in the 5 subjects who received Azer-cel and FluCy750 as well as the response duration are summarized on this slide. As noted earlier, deep CAR T expansion with standard lymphodepletion was modest in the first 2 subjects, among whom experienced 1 CR of 60 days duration. Using the FluCy750 lymphodepletion regimen, we achieved an overall response rate of 60%, including 2 PRs and 1 complete response, with 2 of 3 evaluable subjects achieving an MRD negative response. Subject O achieved an MRD-negative complete response, however, progressed at day 90 with a confirmed loss of the CD19 antigen. The results we just received on subject Q appears to be at the very least a very good PR. Although follow-up is short, results from additional candidates continue to inform Azer-cel efficacy. Nevertheless, we feel that these preliminary data demonstrate safety and remote deep molecular responses, which support an effective and safe product for this patient population with high unmet need. Slide 25. Comparing these data to the target product profile we feel is necessary for an FDA approvable product in the auto CAR relapsed setting. These preliminary data exceed the overall response rate of 50% and, in particular, the safety profile with no Grade 3 or greater CAR T-related adverse events. The molecular response data are promising and we will rate validation by the duration of response. Okay. To summarize on Slide 26. What I shared with you today, Azer-cel was highly active in CD19-positive CAR T relapse subjects, demonstrating an overall response rate of 83% among 18 evaluable subjects with 61% CR rate and a duration of response of 6 months or longer than in 55% of those subjects that had sufficient follow-up. In the modified lymphodepletion cohort, deep responses permitted transition to allogeneic stem cell transplanted to subjects. In the latest cohort, the safety profile is pristine with no Grade-3 or greater allogeneic CAR T-related adverse events in this heavily pretreated patient population. At the recommended Phase II dose of 500 million cells with FluCy750, we achieved comparable overall response rate with preliminary molecular response in 55% of evaluable subject. Also, in December of last year, we received favorable cell manufacturing feedback from the FDA that confirmed alignment on chemistry, manufacturing process changes and analytical controls, that support ongoing life-stage development of Azer-cel. We have secured a clinical meeting with the FDA to discuss a potential Phase II study that includes an acceptable registration path, study size and endpoints that will take place by the end of June. Next trial, we utilize our next-generation product with the highest potency to date. Move on to Slide 27. And now we'll follow with this stealth cell data, or CD19B, which was engineered to be cloaked to evade allogeneic immune rejection in CD19 positive CAR T naive subjects. As you can see here on Slide 28, stealth cell includes 2 additions to the CAR T construct that include an shRNA to knock down related to micro microglobulin RNA to reduce MHC Class I antigen expression by 90% and prevent rejection by host T cells, as well as an HLA-E transgene to prevent NK cell recognition. Remember that stealth was designed to compete with autologous CAR T products and displace it from its current FDA indication in the second line, CAR T naive subjects in particular. Slide 29. To assess the biologic principle, we compared T cell and NK cell recovery in diffuse large B cell subjects who received stealth at the dose level 2 or 540 million cell dose to Azer-cel treated subjects who received the same lymphodepletion regimen of FluCy750 in 500 million CAR T dose. Keep in mind that an HLA mismatched allogeneic cellular product will listen in hyperacute rejection by T and NK cells following recovery from lymphodepletion. As you can see in this figure, T cell recovery on the left, and NK cell recovery on the right, was delayed 1 week or longer in subjects receiving stealth compared to those receiving a Azer-cel at the same lymphodepletion regimen, indicating that the cloaking components of the construct mitigated immune rejection. The CAR T-related adverse events in the stealth cell treated subjects is summarized here. 3 subjects received standard lymphodepletion and 4 subjects received the FluCy750 lymphodepletion regimen to compare to the Azer-cel experience. As you can see, none of the subjects bring us grade 3 or greater adverse events, which included CRS, ICANS, GvHD, infection or deaths, confirming safety of the 2 lymphodepletion regimens and the CAR T dose levels evaluated. The overall response to treatment at the 540 million cell dose and early -- had early clearance -- as well as the early clearance of circulating tumor DNA as summarized in this table, at the first dose level of 270 million cells using product that was produced before the manufacturing optimization, 1 of 3 subjects achieved a PR. What you see here are the results from dose level 2, using the optimized product, and the stealth cell achieved an overall response rate of 71% across both the standard lymphodepletion and FluCy750 lymphodepletion regimens. Most importantly, among the 5 diffuse large B cell subjects, highlighted in the blue box, overall response rate was 80%, with 60% of the subjects achieving an MRD negative complete response. Slide 32. To summarize our stealth cell experience to date, and we have shown the stealth construct has demonstrated biologic proof of concept, evidenced by the delay in host T and NK cell recovery compared to Azer-cel in diffuse large B-cell subjects. At the 540 million cell dose with the FluCy750 lymphodepletion regimen, stealth cell showed an acceptable safety profile with no Grade 3 or greater allogeneic CAR T-related adverse events and a high rate of MRD negative response with preliminary evidence of durability. In subjects with diffuse large B cell lymphoma, in particular, overall response rate is compelling and comparable or superior to autologous CAR T in the third-line setting with an overall response rate of 80% and a 60% MRD negative complete response rate. We feel that the 540 million cell dose with FluCy750 is the go-forward recommended Phase II dose in lymphodepletion regimens. Let me quickly summarize the assets and capabilities of the next generation of Precision's allogeneic CAR T program. The clinical data that we have shared with you today reflects the product of Precision Bio's deep capabilities that enable us to move rapidly from discovery to the early development of donor-derived allogeneic cell therapies. This includes expertise in preclinical research to rapidly design and test new constructs and an experienced CMC and manufacturing team that continually iterates improvements to process development as well as experience quality and clinical operations team. This positions us to tackle opportunities in solid tumors as well as in autoimmune disorders. For solid tumors, in particular, we are able to address the key challenges such as tumor heterogeneity and antigen escape, T-cell trafficking and tumor infiltration as well as the immunosuppressive tumor microenvironment. This is complemented by the potential to advance allogeneic CAR T cell therapy into autoimmune disorders, such as lupus or MS and others. And now I'll hand it over to our CEO, Michael Amoroso.

Thank you, Alan. Sum up our presentation today, Precision BioScience is leading CAR T platform. Precision has validated its proprietary cell therapy platform using markets across now 2 clinical candidates, we believe. Azer-cel is active with an acceptable safety profile in the CAR T relapse setting. The next steps for Azer-cel are a clinical meeting with the FDA in the month of June. Here, we will pursue the conversation around size of trial, study design and endpoints for a potential Phase II. Our second-generation CD19 positive stealth cell, PBCAR19B, for CAR T naive patients earlier line DLBCL has now shown that stealth cell's proof of concept on the construct has been demonstrated. Acceptable safety from Phase I and preliminary efficacy have been demonstrated. There has been a very compelling signal in the diffuse log B cell that should be further validated in the next stage of development. We see this as an earlier line trial, a large and an ultimately randomized if stealth cell continues to have the activity it shows. And this is an area we'll seek partnership. Next, Precision's platform-wide manufacturing optimizations implemented in 2022 with a high CAR T insertion rate using ARCUS are the foundation for the improved potency of our product and control for safety purposes. Precision's state-of-the-art cell therapy capabilities support potential and ongoing collaborations in hematologic malignancies, solid tumors and in autoimmune disorders. Thank you for your time today. We will now open it up to our investor community for questions and answers.

[Operator Instructions] And your first question comes from the line of Andrea Tan from Goldman Sachs.

Alan, maybe I could ask you about -- more about the correlation between MRD negativity and what that implies for durability in the post auto CAR T population. You did have the 1 patient that achieved negativity, but then progressed later. So just curious more on your thoughts there. And then in a related manner, as you look to PBCAR19B in the CAR T naive patients, how should we think about the meaningfulness of MRD negativity versus the 6 months complete response rates that we traditionally look for in this population.

Yes. Thanks for the question, Andrea. Well, I think the MRD negativity for us -- based upon the autologous CAR T experience, is a very powerful predictor of durability of response. That 1 patient in the Azer-cel FluCy750 cohort that did progress despite achieving negativity early on, and remember, lost the CD19 antigen. And for that reason, I think we have a good rationale or explanation of why that occurred. We added this in at this time, and I don't think any other allogeneic CAR T program has included looking at clearance of circulating tumor DNA. In our stealth cell experience so far, of those patients, we have 3 CRs, 50% CR rate, each one of those is MRD negative. So we fully expect those to be durable. One of those is already out beyond 150 days. Right now, I think is 6 months durability going to be the key? Absolutely. For an approvable product like this is going that would potentially go up against autologous CAR T, it's going to have to be 6 months to 1 year. We're just using this as a potential surrogate of the depth of response and potential for durability.

And your next question comes from the line of Soumit Roy from Jones Research.

Congratulations on the data. Alan, if you could just elaborate a little bit on the difference between the lymphodepletion last ASCO versus now? It looks like it's 1 fewer day of fludarabine. And is that what you ascribe the main success is? There were 3 Grade 5 events in the modified lymphodepletion last year. Is it because of that extra day of flu? Or is it just a patient characteristic that have changed between prior 5 patients and now additional new 7 patients?

Thanks, Soumit. It's a very good question, and thank you for asking that. It's a combination of 2 things. Remember in that patient population, we had very high peak expansion to levels that were above 100,000 per mL. That combined with the fourth day of fludarabine in patients that have been previously received fludarabine for their autologous CAR T, put them at risk for the fludarabine related neurotoxicity. They had that maximal huge peak expansion simply because they had a large CD4 cell dose that was able to push the CD8. What we've done to control -- we've done 2 things that help control this. One is removing the fourth day of fludarabine, as you pointed out. So we are back down to 3 days of fludarabine. But optimizing our manufacturing process, where we have now a controlled reduction in the CD4/CD8 ratio. So we control for that CD4 cell dose, so we won't have excessive CAR T expansion [indiscernible]. Remember, CAR T did 2 things. They drive the CD8 expansion, but they also the source of inflammatory cytokines in this related inflammatory toxicities. So those 2 things, reducing or eliminating the fourth day of fludarabine and having better control of the manufacturing of the CD4/CD8 ratio contributed to the lack of Grade 3 or greater toxicities that we see now.

Your next question comes from the line of Kostas Biliouris from BMO Capital Markets.

Congrats on the data. Maybe a couple of follow-ups from us. How high is the antigen loss risk here given that these patients have already been exposed to CD19, and we saw 1 case? Do you think this will be common moving forward? Or this is not really a concern? And then I have a couple of follow-ups.

Thank you, Kostas. So as far as the frequency advantage in loss, there's 2 ways to look at this. One is in the patients that we are seeing and prospectively looking at and other than our data, I'm not aware of any other prospective data looking at this, it's about 85% of patients still are CD19 positive when we are treating them. Now it doesn't mean there can't be antigen escape after therapy. And that's what we saw in that 1 Azer-cel patient. Even though they were CD19 positive in the beginning, had an MRD-negative complete response, they eventually did relapse at day 90 with CD19 negativity. That's the first case that we've experienced in all of our data, but with 1 caveat. Obviously, when patients do progress, they're all studied and there's no obligation for a repeat biopsy.

Great and any early comments around the PFS that you see so far? Any quantitative color around the PFS so far?

Yes. Remember that we had 2 patients at day 90 or longer. And let me just pull that up so I can just refresh my memory specifically. So that's looking at duration of response in the Azer-cel patients. And -- here we go. The other ones are just too early. We have -- the other PR is at just at day 28 this week, so we can't say anything. But I think what we can say and those people that have sufficient follow-up, obviously, we're beyond the -- what we think is the target product profile. And that is from the consortium data, that progression free response is only 1.8 months, and we already have responses that are going out to day 90.

Kostas, this is Michael. The only thing I would add there is when you look at the 18 patients, the CAR T relapsed, the 55% duration of response in 6 months, yes, some of those patients come from the modified regimen. But what I would tell you is the commonality cytoxan750, and our optimized manufactured cells. But I think we feel really good. The point was for us, we thought we had the efficacy well beyond the efficacy last year at the midpoint. We needed to improve the safety. I think that was really what we did. I think the 750 patients, the 5 right now, we're tracking durability, those are really confirmatory for us. We needed to get there in safety, but we feel the activity is at least indicative of what we've seen in the totality of the N equals 18. So Remember, 1.8 months is the median progression-free interval for the, dare I call it, standard of care today, nothing approved in this setting right now, Kostas. So we feel we're well beyond that at this point in time.

Perfect. And maybe one quick follow-up on the patients who achieved negative -- MRD negativity. Any baseline characteristics that you observed there and you can use moving forward to maybe kind of do case and selection more targeted?

So Kostas, what we've done is look at, certainly in tumor DNA at -- starting at day 7 through day 28. What we didn't do was look at screening, certainly in tumor DNA. There is published data for autologous CAR T showing that levels at baseline as well as day 0 reduction of 300-fold is predictive for durable response. So there are several parameters to look at. What we used was just clearance by day 28 as our initial assessment.

Your next question comes from the line of Maury Raycroft from Jefferies LLC.

This is [indiscernible] on for Maury. So I wanted to ask you, can you provide more context on these 5 new DLBCL treated patients versus the 11 prior treated patients? Like, how did the baseline profile compare? And how much of the responses may have been driven by higher lymphodepletion regimen prior?

[indiscernible], you broke up there a little bit. You're a little low. Could you repeat the question, please?

So can you provide more context on these 5 new DLBCL patients versus the 11 prior treated patients? Like, how do their baseline profile compare? And how much of the prior responses may have been driven by higher lymphodepletion regimen?

Well, I can tell you that we've looked at duration of response to the prior CAR T, and we see no relationship between duration of response in this setting or in the prior cohorts in response to Azer-cel. So it doesn't appear to be related at all in that setting, which I think is consistent with the idea that you relapse from autologous CAR T simply because you have poor product attributes to begin with, which can be addressed by an allogeneic CAR T. The other point is these patients, like many of our other ones, so the overall line of therapy is the fourth to fifth line for the majority of these patients. So this is much more advanced patients that you see in many of the other trials. These were also -- had pretty heavy volume of disease. That last patient was a very good PR, possibly a CR. That patient has some of the biggest volume of disease I've ever seen.

[indiscernible], it's Michael. The only thing I would add there, really important, the lymphodepletion helps you get past that initial first 14-day host immune rejection, right? In order to stay in response, obviously, you need to see your CAR is working and you need to see your CAR is peaking, your CAR is working, and that's what drives your durability. So you see why Alan -- the slide in front of you right now, Slide 15 shows bringing the peak around the 44,000 range, the main peak. It's in line with the ZUMA-1 durable responders. The major thing that drove this data is the optimization of some of the product, too. We went up in self-fitness potency, and we tightened the control around the 4 to 8 ratio. So yes, taking 1 less day of LD improve safety, you're dialing up the potency of your product through self fitness and also tightening the 4 multiplier for expansion. So I'd say the #1 thing driving it is the CAR, not the LD. The LD helps you for your day 28 response to get past that first immune rejection.

Got it. And then for the stealth program, what did the CAR T peak and kinetics look compared to the Azer-cel?

Yes. Thanks for asking that. We didn't include that on this slide, but it's very comparable to what we see with Azer-cel. The AUC, it looks -- in the preliminary, it look comparing to diffuse large B cell patients, which is what obviously Azer-cel is receiving, the AUC looks higher, which is what you would expect to see with proof of principle with the cloaking change that we made to the constraint.