Autolus Therapeutics plc (AUTL) Earnings Call Transcript & Summary

Hello, ladies and gentlemen, and welcome to the Autolus Therapeutics ASCO Analyst Conference Call. As a reminder, this conference call is being recorded. I would now like to turn the conference over to your host, Julia Wilson. Please go ahead.

Thank you, Jonathan. Good afternoon or good evening, everyone, and thank you for joining us to take part in today's call to discuss the data from the pivotal FELIX study of obe-cel presented today at the ASCO conference. I'm Julia Wilson, Communications consultants of Autolus. With me today are Dr. Christian Itin, our Chief Executive Officer; and Dr. Claire Roddie, Associate Professor at UCL, an honorary consultant Hematologist at UCLH and lead investigator on the FELIX study. Before we begin, I would like to remind you that during today's call, we will make statements related to our business that are forward-looking under federal securities laws and the safe harbor provisions of the Private Securities Litigation Reform Act of 1995. These may include, but are not limited to, statements regarding the status of clinical trials and development under regulatory time lines for our product candidates and our expectations regarding our cash runway. These statements are subject to a variety of risks and uncertainties that could cause actual results to differ materially from expectations and reflect our views only until today. We assume no obligation to update any such forward-looking statements. For a discussion of the material risks and uncertainties that could affect our actual results, please refer to the risks identified in today's press release and our SEC filings both available on the Investors section of our website. On Slide 3, you will see the agenda for today's call, which is as follows. Christian will make some opening comments. Claire will then go over the slide she presented earlier today in the ASCO oral presentation before Christian will wrap up with how Autolus is building the obe-cel opportunity on next steps. Finally, we will, of course, welcome your questions. Over to you, Christian.

Well, thanks a lot, Julia, and thanks a lot for joining us at this sunny day in Chicago. It's been quite a remarkable day, I think, for us at the company presenting the first pivotal study, and I would say it's our great pleasure to have Claire Roddie here to actually go through the data and, frankly, give you an opportunity who may not have had the actually the possibility to be here at the conference to actually get a firsthand presentation of the data. In the distance here, we have soldier field, and we're just actually experiencing sound waves coming over from a concert that's sort of getting ready, and we're sort of seeing, obviously, whether that sound will sort of overpower us or not, but we'll sort of do our best at our end. It's a Taylor Swift concert. So you may actually sort of recognize the tunes. So with that, what I'd like to do is actually hand over to Claire, who will run us through the presentation and obviously give us a firsthand account of the data for the FELIX study.

Thanks very much, Christian, and thanks for joining us. So I'm going to talk today about the FELIX study, the top line results. I'm Claire Roddie, I'm based at UCL in London, so I think we're familiar with the background here. We know that CD19 CAR T cell therapy really truly has revolutionized the field of relapsed and refractory acute lymphoblastic leukemia. And I think we recognize also that obe-cel is really a very unique autologous CD19 CAR by virtue of its fast off with CD19 binding domain. Of course, the binder were selected to really as a reaction to other CD19 CAR T cells in this space, where the issues around safety and persistence were a problem. Obe-cel doesn't have those problems. Obe-cel has got an excellent safety profile and excellent persistence profile. The clinical activity of obe-cel has already been tested in some Phase I studies as I'm sure you're aware, in pediatric and adult acute lymphoblastic leukemia. And of course, at UCL, we've been testing it in a range of other B-cell malignancies. But from the point of view of today, we're really going to focus on the adult patients population with B-ALL that we've treated on the FELIX study. So again, most of you will probably be familiar with this schema and the trial design. And essentially, what happens is the patients are screened and [indiscernible]. They then undergo bridging therapy at the investigating clinicians' discretion. And then when they have a product available, they're admitted to the hospital for preconditioning chemotherapy with fludarabine and cyclophosphamide, followed by a split there's CAR-T cell infusion over day 1 and day 10, and the criteria in order to receive the day-10 dose and the absence of severe immunotoxicity in that intervening phase. There's another additional safety measure that's been implemented in this study, and that is to titrate the day-1 dose of CAR T cells to the pre lymphodepletion bone marrow biopsy result. So for instance, patients with bone marrow blasts with less than 20% will receive a higher initial dose of CAR T cells on day 1 at 100 million cells. And patients with more than 20% blasts will receive a lower day 1 dose, and that is 10 million cells. These patients will go on to receive the dose 2 in the [indiscernible] of immunotoxicity to a total dose of 410 million CAR-Ts. And it's important to note that irrespective of the differences in the pre-lymphodepletion bone marrow burden, actually 94% of all infused patients received both of these obe-cel infusions to a total dose of 410 million CAR T cells. So let's focus -- sorry, do you need to move on to the next slide, Julia. I just moved it forward myself. Well, we'll go on -- we'll carry on now we'll look a bit at the eligibility endpoints and disposition of the patients. So important to note the eligibility criteria. This was an adult ALL study to patient needed to be for the age of 18. They needed to have 5% or more blast in the bone marrow to be eligible for the Cohort IIA. And in terms of the endpoint, the primary endpoint is really mainly CR or CRi by central assessment with secondary endpoints including duration of response, MRD negativity and safety parameters. Now the 112 patients were enrolled on the study. 84% of those enrolled patients were infused with obe-cel. There were some dropouts, and that was mainly due to death and progressive disease. But ultimately, 94 patients were infused with product, and the median duration of follow-up we have for those patients is 9.5 months. Now in terms of the baseline characteristics, there's a lot of information in this slide. We note that the median age of the patients was 50, but we actually did treat one patient who was 81 years old. Other notable factors here, you can see that these are high-risk patients, a lot of them are heavily pretreated with 3 or more lines of prior therapy. And 31% and 53% are refractory to last lines therapy, 38% of have the prior allogeneic stem cell transplant, and 67% have been exposed to either blinatumumab or inotuzumab. So again, this also denote patients who've been through more lines of therapy. Now in terms of the disease burden of these patients, I think this is important to emphasize at the point of screening for the trial, there were 50% bone marrow blast, the median bone marrow blast burden for the patient cohort. And then when you look at the point of literal depletion, the bone marrow burden is still high with a median of 41% blast. So we know that this is a high-risk population for CAR T cell therapy. We also have to point out that there was a sort of over representation, if you will, of extramedullary disease in this cohort with 19% of patients having extramedullary sites. And again, we know that this potential sort of a per prognosis in the ALL field, extramedullary result disease is per [indiscernible]. Now I think this is what makes this slide particularly important. We know that these patients have lots of disease. We know this got circulating disease in a proportion of cases, a lot of extramedullary-fied older patients. So the fact that the manufacturing data is so crisp plane is really encouraging to us. I think on the left-hand side of the slide, what you can see here is that 96% of products reached the target still. That's 410 million cells, and this is irrespective of whether you've got high disease burden. You can see that even those with more than 75% blasts, which dose. In the middle of the slide, you can see the median transduction efficiency of 72%. And I think you have to look closely at this because those results really are clustered very tightly around that 72% mark, telling us again about how consistent this process, this manufacturing process that we're using is. And the last point of this slide is about this vein to release time. I think what you can see here is that we have a median of 21 days for that vein to release. And that's really important for this patient population, in particular, because they're often in urgent need of clinical treatment, and they often have roughly progressive disease. So this is a great place to be in terms of delivery of drug to site. Now in terms of the disease response, well, you can see it very clearly, mark here, 76% of infused patients achieved CR or CRi. And I think, again, that's in the context of an extremely high disease burden for a large proportion of these patients. It's also an older patient population, a lot of comorbid patients. So this is very encouraging data for us to see. And what's more of these responses are deep responses, with 97% of responders achieving MRD-negative disease response as calculated by blue cytometry. And then in terms of whether or not these responses are durable, well, this graph speaks for itself, 61% of responders are in ongoing remission without subsequent anticancer therapies at a median follow-up of 9.5 months. So again, this is really important data. We can see that the duration is really impressive here. There were a few patients who proceeded to stem cell [indiscernible] whilst in remission, but these were sensitive for the purposes of this pictorial representation. Now let's look a little bit about the subgroup analysis here, CR/CRi assessment. There's lots of different criteria that we've looked at here fascinating and to get some early insights into what factors may predispose to respond or not. We denied with extramedullary disease prior to prior to lymphodepletion and high bone marrow blast percentage. And by that, I mean more than 75% blast preconditioning, those both were associated with a higher risk in terms of this sub-analysis. But let's focus on the toxicity on this next slide. I think we've really alluded to the fact that a lot of these patients are older, a lot of them are comorbid, and they have limited tolerance for high-grade immunotoxicity. So in terms of CRS and ICANS, I think what we can see in the far right-hand column is of all of these patients, we saw minimal Grade 3 or more CRS despite the heavy disease burden in this group. You can see it was only 3% of patients experienced Grade 3 CRS event. In terms of ICANS again, limited Grade 3. ICAN events affecting only 7% of patients. And if you track into the table, you can see that the majority of those events we're quite predictable in that they occurred in patients with high disease burden pre-lymphodeplition. So this is sort of a population that could potentially be watched more closely for this side effect. And tocilizumab and steroids were used, of course, for CRS in 56% and 17% of patients, respectively, and there were very few patients who require vasopressor support for CRS up 3%. And the ICANS we've alluded to already that it seems to be associated with higher bone marrow blasts preconditions [indiscernible]. In terms of treatment, the adverse events, there were lots of things listed in this table. But I guess some common things draw your attention to are neutropenia, effective 36%; thrombocytopenia, effective 25%; febrile neutropenia 25%; anemia, 19%. There was 1 death on study that was related to HLH and neutropenic sepsis that was attributed to obe-cel. The expansion and persistence data that I've seen on this slide is particularly impressive. So when we run the ALLCAR19 study, which was the Phase I test of obe-cel in adult B-ALL. You can see that we did some pharmacokinetics in this table on the right of the slide. And when we compare our experience of this early pharmacokinetics with that, that's achieved in the FELIX study, you can see how comparable those measures are. We get great peak expansion, and the Tmax is equivalent across the 2 studies. And the area under the curve over the first 28 days very much mirrors what we saw in the Phase I study, suggesting that we're effectively -- we're delivering the same drug to this patient population, which is very heartening for us to see. If you look on the left-hand side, you can see a nice plot of the CAR and engraftment and persistence by QPCR, and what most patients have achieved a peak Cmax of around about 100,000 copies per microgram of genomic DNA, which is an excellent initial expansion. And what's really nice to chart here is essentially where the graph plateaus out. And you can see that patients even at 6 months on 10 months and the further point to follow up here have bottoming out at about 1,000 copies per microgram of genomic DNA. So what this is telling us in the these patients that are followed up to this time point that they have got ongoing CAR T cell persistence we knew from our early experience on the ALLCAR19 study that this is important or has been shown to be important for our long-term responders. So if we have to conclude, essentially, despite this patient population that we've acknowledged have got high disease burden, have got lots of per risk features, are older and comorbid, Irrespective of all those factors, obe-cel can deliver and complete responses in 76% of patients and not only complete responses, but MRD negative deep responses. So this is again -- this is something that we're very proud to be able to share. And what's more, these responses can be durable with 9.5 months of follow-up, 61% of responders remain in remission, which again is it's a very important piece of data. I think one thing that's really important for us clinicians, particularly is the tolerability of this therapy. So in contrast to some other agents that are available, obe-cel resulted in very low rates of Grade 3 CRS. As you can see here, it's 3% of patients and very low rates of Grade 3 ICANS at 7% of patients. So I think this, again, the tolerability in these permit patients is a huge factor in the decision-making about which drugs to give. And let's just pay a bit of attention to the manufacturing process. Irrespective of disease burden, [indiscernible] and adverse prognostic factors, actually, the manufacturing was pretty straightforward. And the majority of patients for whom leukophoresis was received achieved a released product through this nice consistent manufacturing process. And of course, we've alluded to the CAR T cell engraftment and persistence, which we believe from a kind of a pharmacologic perspective is very important for durable responses. It's really heartening to see that parallel pharmacokinetic data with our ALLCAR19 study. So with that, we'd just like to thank everybody who's been involved in the study, the patients and their families and so on, the study investigators and coordinators, the health care staff at the study sites who've worked very hard on this study and also a therapeutic for sponsoring the study also. So with that, I'll come back for questions.

Thank you much, Claire for a very exciting presentation, very clear presentation. What I'd like to do now in the next few slides is really look to put the data in perspective with regards to the other therapies that are available to these patients. And moving to Slide 21, what you can see here is sort of a comparison between the studies that actually led to the approval for Blincyto as well as Besponsa, and you can see the summary of the key data that we're seeing there. If we see how -- if we're looking at FELIX, what you can see is that clearly, obe-cel stacks up very well both on the efficacy with a very significant level of activity, including when we're looking at intent to treat, which is what normally isn't really calculated or shown our intent to treat actually is at 64%, which is still an outcome for a personalized therapy. When we look at the median duration of response, we're stacking up very well. Obviously, what you have seen from the curve that Claire was presenting is that we still have a lot of patients that are in the early part of the curve. So with 9.5 months median follow-up, I would say, were much shorter than what we currently see as a median duration of response, which was at 14 months. So we're -- obviously, we'll see the maturation of this curve as we go through the course of this year and obviously expect a key update at the ASH meeting at the end of the year. But this is actually looking very encouraging. It gives us, I think, a lot of confidence in terms of the ability to translate this very nice level of initial activity from a response perspective, the excellent persistence that we're seeing that this also will translate into an excellent outcome from a durability perspective. Now as Claire pointed out, the safety is remarkable, and we shouldn't take safety for granted. We needed to understand that these patients are truly in a very difficult position. They have gone through a very significant set of therapeutic treatments, and they tend to be in very poor condition. And so when you actually have patients, particularly [ adult ] patients and you drive high-grade meritoxicity or high-grade CRS, there's a very high risk to these patients. And that is something that we shouldn't take lightly. Now what you can see here is that as we're comparing to Blincyto that, in fact, the safety profile that we've been observing in the FELIX study actually is better than what was observed to Blincyto. And I think that is important to keep in mind when we think about the opportunity for this product to really become a significant component of the backbone of therapy for these patients over time. So this is obviously with regards to Blincyto as well as Besponsa. When we look to the approved CAR T therapy Tecartus, what was very interesting in today's presentation or today's session is to the real-world data presentation as well from Tecartus and was interesting for 2 reasons. First of all, what we do know from the Tecartus experience is that the actual real-world data between the original ZUMA-2 study as well as then the real-world data was extremely close together. They were tracking each other remarkably. And two, what was quite interesting to see is obviously that this isn't the case between ZUMA-3, the approval study and the new study, the ROCCO study in terms of the real-world experience. What was quite obviously is different that I think we need to keep in mind also as we sort of we're thinking about the FELIX study is that we've done this in the same period of time, most during the course of 2022 at the height of the pandemic and under very, very stable conditions. And indeed, we do see an impact clearly in the numbers. The patient populations are different when we look at the real-world experience, about 31% of the patients actually were in complete remission at apheresis. So these are patients that had either no disease burden whatsoever, sero molecular complete remission or they might have had minimal residual disease, but they're clearly in complete remission on the morphological disease. So that obviously is important to keep in mind, and also the denominators obviously have been shifting somewhat. Clearly, the product is active. When we do the calculation, we believe the product is showing equal or comparable activity to what was seen in ZUMA-3 in a comparable patient pool. What was quite interesting and not surprising, this obviously was a very difficult environment, and patients were very aggressively managed for cytokine release syndrome. And that's obviously the key component there is obviously the use and early use of steroids. In fact, actually, that was helpful. It reduced the overall grade 3/4 to 5%, 6%. But the flip side was that it actually did lead to very significant levels of high-grade neurotoxicity which, in fact, when we look at where they occurred, which appeared to be in patients with higher disease burden would suggest that every other patient, more than 50% of the patients have experienced high-grade neurotoxicity, which obviously is very challenging to manage. And in fact, 6 patients died with ICANS and infection. So what it suggests is they're clearly controlling this type of a profile in a real-world setting clearly is challenging, and I think it's something that I'm sure Claire can comment on in the Q&A section as well. As you can see, again, subsequent stem cell transplant quite similar to what we've seen before. So overall, I think it highlights the fact that safety -- managing safety is difficult in these patients. And I think it puts in perspective, I think, the remarkable safety outcome and activity outcome that we've seen across the FELIX study. Now when we look at sort of the commercial opportunity, we also do realize that, in fact, the safety profile is the key enabler for Blincyto to actually grow the way the product has been growing, and it has been growing remarkably. We currently expect that Blincyto will reach approximately $800 million in sales this year, and this is really to do with the fact that the program is really manageable. And patients can actually be managed not only in the academic transplant centers, but also nonacademic transplant centers and nonacademic hospitals as well. And I think that is important because it actually allows patients to be treated closer to home. It takes away risk, and particularly, the risk that we've seen during COVID disease-related to travel exposure to infection, et cetera. And I think it is an important component. Given that, in fact, obe-cel actually has a safety profile that is better than Blincyto safety profile based on the studies that we have now to compare to would suggest that we're very well positioned to actually have provide broad access across not just the academic transplant centers, but beyond for a product with the profile that obe-cel has. So with that, I'd like to just spend a little bit of time talking about manufacturing. And I think you heard from Claire before and you saw the data that we obviously had a very remarkable robustness on the manufacturing side. And I think this has to -- I think I'd like to sort of allude to that and move on that a bit further. So first of all, when we think about product supply, and we think about the critical success factors in a personalized cell therapy One of the real challenges that we have with these types of therapies is that we have to have a very reliable, timely delivery of every batch with a consistent quality. Because every batch we do not deliver either on time or do not deliver to quality, we actually have a patient at risk that will not have an opportunity to benefit. This is very different from any other therapeutic mortality, where your manufacturing is dissociated from treatment. Here, it is integrated and integral connected. So that is the challenge. Now what that requires you is to actually develop a manufacturing process that can perform consistently with a very wide range of patient material. And as you heard from Claire before, one of the challenges we have with these patients, particularly as they have high [ tumor ] burden, is that there is an infusion of leukemic cells from the bone marrow into the blood. And when you collect then the T cells from the blood, you also will collect obviously, their leukemic cells. So the majority of the cells you do collect are actually leukemic cells. That's a very, very difficult starting point for any manufacturing process because not only do you have to extract and isolate your T cells in the first place, but you also actually have an environment that really is challenging for T cells to be active in. And that actually is a very, very difficult starting point. The second aspect is clearly turnaround time. And I think it's important to note, obviously, we did manufacture for 24 centers in the U.S. We manufactured in the U.K. And we were able to actually have a vein-to-vein, median vein-to-vein time of 35 days. This compares to the real-world study, the ROCCO situation with commercial manufacturing established in the U.S. and only U.S. centers involved the 31 days. And I think it points out to the fact that we're highly competitive in terms of our manufacturing turnaround time. And as we'll talk about a little later, obviously, we're actually on track to massively reduce that time. Now when you actually think about the importance of product consistency, but also economies that we have to achieve to actually get products manufactured at reasonable cost, you do actually have to sort of get to a manufacturing process that at least is in part automated. And this is one of the core principles that we're using, is to actually have a level of automation. Now the other component, which is really critical with these types of manufacturing processes are people. It starts with the leadership team that we have in place at Autolus, which is really outstanding on the product delivery side, and gives us really a level of attention, drive and continuous improvement, which are absolutely critical. You do need a highly trained and motivated workforce. And in fact, we had to operate and set up an actual training center to train our operators because the people with the skills we need for this type of manufacturer do not exist in the workforce. So that's an investment of about a 1-year investment in every single operator, and that obviously becomes a key value and a key element in terms of the success. And then you have to establish a culture that really focuses on continuous improvement, and that sort of requires you to really put in place an operational excellence program. Now the final bit I wanted to talk about, which is scale of operation, and this is critical in 2 dimensions. When we look at the CAR-T experience, and I think the challenge to a lot of the centers have that deliver CAR-T therapy is that it is often difficult to get access to CAR-T therapy because the manufacturing capacity isn't there that you need to actually be able to reliably deliver. There's also quality issues, which actually make the matter worse. But capacity is important. So what we really need to do is we have to actually have to put in place a capacity in terms of its size that matches the target population that this product is looking to address. That's one parameter. The second parameter is you have to put that capacity at the right level to serve the population but also make sure it's at the right level to not actually overbuild the capacity. Because if you do that, your overall economics on the cost of goods actually goes down the drain. So you actually have to find the right balance between capacity and the actual size. Now when we look to think about the study that we've just conducted here, obviously, what we have done is that we really have to execute in an extremely difficult environment. And that, in essence, allowed us to pressure test every aspect of the product supply part. It started obviously with the semi-automated manufacturing process, which was critical that we optimize to manage this wide range of apheresis materials. We were doing a lot of work in fine-tuning process controls and release analytics. Training center already mentioned, which was absolutely important to establish and maintain the workforce. But we also operated with a commercial model with 2 shifts, 7 days a week, and that gives you actually the type of operation that is also the commercial operation. The excellence program was very important, and it actually has allowed us to continuously improve our overall operations and, with that, our operating cost. And the final part, which was really challenging was that the logistics part, we have 24 centers in the U.S., where we have to deal with all the COVID restrictions, we had transatlantic flights that were down to 5% of normal, and we're able to deliver product for each and one of these patients. And I think that is absolutely important. But what it did, really, this environment, is it really allowed us to look at every individual component of the product supply chain and pressure test it and ensure that we actually have a very robust overall setup, and it does show very nicely in the data that Claire just walked you through in terms of overall performance. And I'm sure you do realize that those type of data are not often shown. This is every individual batch shown on this depiction, and that's normally not done. And it's not done because, typically, this is difficult to actually show consistency and tightness of data. The commercial manufacturing facility, The Nucleus that we set up is designed to actually serve 2,000 patients in its initial build-out, which is 2/3 of all relapsed/refractory ALL patients in the U.S. and Europe combined. So the sizing allows us to actually fully deliver and serve this market and be able to actually capture the opportunity in full. The challenge that we were dealing with was that we actually were -- had to start this setup for this facility as we're actually gearing up to start the Phase II portion of the FELIX study. This was, therefore, had to be done in a quite unusual way and in a much shorter time frame than any other facility that actually has been pulled together in aerospace. We could do it because the design was such that we had a modular build, which allowed us to build most of the components of the building off-site, 70% built off-site, and it actually allowed us to actually reduce the build time by 60%. So to put this in actual dates, we had groundbreaking in November 8, 2021. In November 22 -- 22nd of November in 2022, we actually did take our first clean room into operation and had our first Prodigy machine that you see here on the right-hand side, the suite in actually full during the capacity challenge. The first machine got online December 14, 2022. And by May this year, we actually have a capacity. So this is where we are. So this is on track. It's robust, it's reliable and it's at the right scale. So with that, just a few words in terms of outlook and next steps. Obviously, we just went through the great presentation of the data for the FELIX study. We'll have obviously, this type of presentation at the EHA in just a few days in Frankfurt. And obviously, what we're focusing on internally is really getting ready to file for a BLA towards the end of the year and also planning to file with the European agency and the MHRA in the first quarter and through the first half of next year. The manufacturing, which just talked about, I think it's important. We're as exactly where we want to be. We're exactly on time and execution perspective, we want to be. We expect to receive a GMP license from the MHRA during the second half of the year and, with that, are going to be in a very good position to support the BLA filing and other regulatory filings there on forward. From a commercial perspective, obviously, with clinical data in hand from a pivotal study, that's a good time to actually go out and talk to physicians about the profile of the product. Medical affairs is obviously a key set of activities we're engaging in. But we're also actually articulating the value proposition to the payers, to the hospitals, the physicians, but also, obviously, to the patients, and that's obviously a key part of what we're engaging on the commercial activities. And finally, we're starting the work to actually onboard centers so that when the product gets through to approval, the centers are in a position to actually deliver the therapy to patients. So those are kind of the key activities that we have lined up, and we believe that puts us in a very interesting spot. From an execution perspective, we have a fantastic data [ that's fetch ] from the FELIX study. We're actually in a position where we can supply the product, and we can do this reliably, and we're now in a position to actually gear up to go through the regulatory process and prepare for commercialization. And with that, I think we're at the end of the formal presentation, and we're happy to take questions.

[Operator Instructions] And our first question comes from the line of Kelly Shi from Jefferies.

Congrats on great results on FELIX. I have a question regarding the safety profile. The neurotox is particularly very outstanding. So on your cellular kinetics chart, obe-cel was observed with rapid cell expression -- expansion and grid persistence relative to other 9 CAR-Ts. So regarding the cost of neurotox, besides tumor burden, I think in the past, it has been proposed that the frequency of neurotox occurrence was somewhat associated with the speed of the cell expansion. And clearly, this is not the case for obe-cel. So what plays the magic for achieving such a clean safety profile, especially on neurotox?

Kelly. Well, first of all, a really good question. And obviously, neurotox is a type of toxicity that's sort of been dealing with probably for 21 years by now, having seen the first neurotoxic 2002 with Blincyto and working with my old team to figure out how to manage it, how to better understand it. What is interesting about obe-cel is that obe-cel obviously leads to very, very significant expansion. So the growth of the CAR T cells in [ vivo ] is very, very significant. But that growth is driven by the actual kill that is occurring and the actual removal of target sales, and it goes alongside that. What we have observed, and this is work that goes back almost -- it actually goes back a good 20 years. What we see is that CAR T cells actually do repeat cell killing and go into serial killing mode, that they actually do stop producing cytokines. And actually, their profile starts to shift. In other words, you can have a lot of high level of activity and very minimal titer candles. And other thing you can see. And typically, when you looked at that, if you were taking a plant for that, you would see significant cytokine release on first kill. If you then actually take those same cells on the second kill, your cytokine release goes down to about 10%. If you take the cells again and run another kill, you actually have virtually no measurable cytokine release. So you actually have adaptation. And one of the hallmarks that we're seeing with obe-cel in its design is that it is really designed to give you very efficient serial killing because it actually does not get stuck on the target cell once the kill is delivered. The cell can actually recycle go into productive next kill. And what that allows you to do is really get through that adaptation. And with that, actually get to a mode where the cells are active, but they actually stop producing a lot of these inflammatory mediators cytokines, obviously being an obvious one category, but other categories as well. And I think that is really at the whole market, it goes very much back to the actual design of obe-cel, which enables the product to actually behave in a much more physiological way, much more like a normal T cell engager and with that avoids this very high level of induction of neurotoxicity.

Very insightful. And I also have a follow-up here. I'm curious, how is the enrollment in the morphological cohorts ongoing? When should we expect the data? And how are you planning to use the data in expanding the patients which can be treated with obe-cel? Could this data be included in the label?

So first of all, I think what you were asking was related to the MRD cohort, not the morphological cohort.

Yes.

So obviously, the key activity, obviously, and the focus on this presentation and also the focus from a data perspective, you submit for a registration is all on the morphological cohort. Any data that we have on patients with low disease burden is supportive, but actually will not in of itself support the actual approval. We'll provide an update on those activities as we go through the second half of the year. And one of the things that certainly we're starting to look at also much more carefully in the current dataset is to what extent tumor burden impacts the activity, the toxicity, as well as long-term outcome. And there's quite a lot of work and lot of data, a very rich dataset that we have even with the current cohort that I think will give us a lot of information in that regard.

And our next question comes from the line of Eric Joseph from JPMorgan.

Very encouraging to get an update. Can you talk a little bit about the duration of response among patients that entered with baseline extramedullary disease? And how sort of the activity in that population compares with outcomes seen with CAR-T currently in a similar patient setting to extended study? And then I have a follow-up.

Well, thanks for joining, Eric. So in terms of patients with extramedullary disease, as you've seen, we have about 19% of the patients who have extramedullary disease. We're starting to actually collect that data. I don't think we have sufficient patients that are far off out to really, I think, analyze the questions that you're asking. I think the very interesting question is definitely some areas that we're going to look at with quite a bit of intent, and we'll certainly communicate around that. But we do need more follow-up on these patients. I think to have a good understanding of how the data is tracking and also how it compares to patients who do not have extramedullary disease.

Okay. Got it. And maybe just one on the manufacturing side. I wonder whether there's a minimum transaction threshold that would need to be surpassed as part of the release criteria and particularly sort of in the commercial setting for product to qualify as a suitable product, I guess, actually as reimbursable product. How does that relate, specifically those that tender has been defined, how do we -- how should we be thinking about that in the context of the 72% transduction efficiency that you're showing here?

Right. So first of all, yes, you do have a minimal level of transduction efficiency that you need to have -- as you have a minimal level of dose as well as a few other parameters in terms of actual cell kill activity and so on that are all constitute release parameters and the release criteria. And you're correct, those need to be met. Now what I think was really important in the presentation is that you can see that the transduction efficiency actually is very tightly packed. And they're very tightly packed squarely above 50%. And with just hardly any product actually that was below that level. The minimum level of transduction efficiency is substantially lower than 50% that you need to hit. And in fact, just around 15%. That gives you still at that level an active product. So one of the things that we're also doing, and this is also one of the more detailed analysis that will be shared with the regulators is obviously also what is the outcome of patients that have across the range of transduction efficiencies. And obviously, as you can see from the data, as it is summarized here, it is highly consistent. So we're in really good shape with that regard, and we believe that we're going to have a very high percentage of the products that will be in specification, or to put it differently, very, very few samples of products that will be high as that. The current study was 94%.

And our next question comes from the line of Gil Blum from Needham.

I'd like to add my congratulations. Very interesting presentation today. So maybe to start -- just to give us an idea of how much more follow-up do you expect will be required to give a better understanding of comparable long-term remission as it compares to toxicity tecartus. I know we're going to have another data read out at ASH, but isn't that time point position.

Gil, thanks for joining. It's a really good question. I mean, normally, what you'd like to be at to sort of actually have I think a good understanding that the numbers that you look at on duration of response for TETRA are really solid and are not moving anymore. You try to go to about 1.5x the time of that median in terms of follow-ups. In other words, if you want to have 14 months of follow-up, and when I say this is a median follow-up and it's solid, the data is solid, you would have to have at least a median overall follow-up in the study of 21 months. So that's when it gets really robust. I think we start to get a very good level of understanding as we're sort of getting towards the end of the year. And we see that the trend, and it's not unusual that you didn't actually see those numbers actually continue to move further back and actually give you sort of an increased level of outcome over time. And we've seen that with other programs as well. So given where we are at 9.5 months follow-up, we're already at, if you want to call the median at 14 months is quite remarkable. And obviously, we have a significant proportion of the patients. As indicated, 61% of the patients that actually had no event whatsoever, no other therapy, and you can actually see those patients continue over time. I think it is a very good opportunity to sort of see that data mature and sort of will give us certainly by the end of the year and then by us for next year, I think, a very solid view.

Very helpful. Just maybe to kind of reiterate this, although you discussed that, just to put in context the importance of the reduced rates of ICAN, especially when compared to what you saw with the real-world Tecartus. I thought it was interesting that one of the panelists who presented that data basically said that, that will be a reason to pick one treatment over another, if you can elaborate there.

Yes. And I think I probably would like to hand it over to Claire as you're the treating physician, you're dealing with those diseases.

I mean yes, I think I'm going to speak on behalf of my -- a lot of my ALL colleagues. These are a really tricky to treat patient population. They were quite comorbid by the time they get to us, and I think toxicity is sort of like a really top priority for us. I think that ICAN data from the real-world experience would put a lot of clinicians off. I mean, I think we already understand the toxicity profile of Tecartus. We understand the kind of the low durability of responses. And as the clinician body, we are really looking forward to obe-cel being more widely available. I speak certainly on behalf of all my colleagues in the U.K. We discussed this quite frequently, and I think across the U.S. also. We've got 2 big advantages here. We've got the toxicity profile across this high-risk group, and we've got the persistence profile as well. I think those 2 things maybe make the decision incredibly easy for us. So I don't know if that answers your question, but I think that's the clinician's perspective on the data we heard today.

That's definitely helpful. Maybe kind of the last one on the angle for the ROCCO study. So just from what you've seen, were there any clear manufacturing challenges in the real-world setting of Tecartus? This is just particularly interesting because it's also under in COVID, so it might give you a better reference point for obe-cel.

It's a good question, but I don't think we have information. What we do know from the ZUMA-3 study is that the 77% of the patients that were at apheresis got dosed versus the 84% that we had in the Felix study and obviously was pre-COVID. There was no information actually given how many patients were intended to be treated on the real-world study. So there is actually no information that was provided. What was clear is that the vein-to-vein time was highly comparable with our vein-to-vein time, although, obviously, this is a commercial setup. And at this point, I think certainly the best supply chain on the CAR T side.

And our next question comes from the line of Yanan Zhu from Wells Fargo.

And also, I wanted to add my congratulations. The neurotoxicity profile when presented in the same session with Tecartus real world really stood out as being a better profile. So I think I have a couple of questions for Dr. Roddie. In terms of -- we are trying to compare efficacy from the FELIX study with that of Tecartus, what should we keep in mind in terms of baseline severity of the patients or prior treatment? Anything that you would highlight for us in terms of the baseline characteristics of the patients from those 2 studies?

I'm sorry, carry on.

If I may just mention, the second question is with regard to the single case of HLH or neutropenia sepsis, that was deemed related to obe-cel treatment. I was wondering, is that really the obe-cel effect? Or is that a lymphodepletion effect? And how -- what have we seen in the FELIX -- sorry, in the ZUMA-3 study has the similar case or cases happened with the Tecartus studies?

Okay. Well, I guess in terms of answering your first question, I mean, you're asking maybe sort of what were the differences in the demographics between the patients on the ZUMA-3 and the FELIX study. And I think we can sort of quite comfortably say that there was more extramedullary disease on the FELIX study. So again, a really high-risk group of patients we accepted onto the study. We certainly had older patients on our study compared to ZUMA-3. So I guess beyond that, the cohorts were partly similar. I think in terms of the therapies that we're permitted sort of pre-FELIX versus pre ZUMA-3, excluded those who have been exposed to blinatumomab, thereby taking out that population that might be at risk of CD19 negative relapsed. Of course, on FELIX, we accepted those studies. So maybe we had a more permissive eligibility or entry criteria. So those would be the sort of the difference of [indiscernible] between the 2 studies. But I think we were more permissive, and we probably had maybe potentially some higher risk patients in our cohort. And then in terms of HLH, I mean, HLH is obviously a well-recognized side effect as were associated with CAR T cell therapy. It's more commonly seen with CD22 CAR, but it's well recognized in the CD19 space also. And it happens as a sort of secularly following a prior CRS. I mean it's sort of like at a CAR T expansion and the absence of antigen. It was not -- we're not unfamiliar with it. We see it in patients, and it seems to be more concentrated in patients with high bone marrow disease as well. So in some regards, I'm not surprised that we saw [indiscernible] HLH, I think probably the kind of what precipitated the grade 5 event here with the neutropenia expenses. and in the context of profile neutropenia in the age. I think that's quite commonly observed. So unfortunately, I think this probably was related to the sort of effect of there may be the cumulative effect of the lymphodepletion on CAR T, but it is certainly a CAR T associated phenomenon. If that answers your question.

Got it. Super helpful. If I may add one more question. The real-world data for Tecartus does show that Tecartus seems to be effective with the CNS disease patients. Obviously, CNS disease patients are excluded -- have been excluded from any of the trial of Tecartus or obe-cell. I was wondering, so first of all, do you think obe-cel might also have that kind of property in CNS positive patients? And also, obviously, we know the neurotoxicity of obe-cel is obviously much better than Tecartus. But would that have implications of efficacy in CNS? I guess, are there 2 things are somehow connected?

So again, in the context of CNS involvement by leukemia, I mean there's a precedent through the pediatric data to show the CAR T cells traffic very nicely to the CNS, and they can eradicate leptomeningeal disease. So I do think that, that is something that's new. I think that's something that's appreciated. And obviously, studies have been designed to be -- to control to exclude those patients just because of concerns regarding a heightened risk of neurotoxicity. I think the bottom line is on the ALLCAR19 study, which is a study I ran at UCL of 20 patients. We've routinely checked the CSF on those patients at day 14 and day 28, looking for CAR T cell infiltration into the CSF. Consistently across the whole cohort, we saw really nice trafficking to the CSF and sort of like high CAR T cell expansion within the CSF, and this is across the board. This is patient who didn't have any neurotoxicity. So I don't think there's any question that CAR T cell can traffic, and obe-cel can prospect to the -- and I guess the next question will be, it's trafficking and become a radicate disease and potentially that's something that could be the objective further study or an extension cohort on the FELIX study potentially.

I mean it's an area obviously that we also had sort of seen in the context of the pediatric work that we've been doing, also actually on the context of AUTO1/22, where we have children with CNS involvement. And it's interesting, it doesn't necessarily actually drive neurotoxicity in these cases, which is typically, you think at first class, those things are associated. They may not be associated, and we clearly saw activity in addition to the trafficking that Claire just mentioned.

And our next question comes from the line of Matthew Phipps from William Blair.

Congrats on the updated data. I was wondering if I could ask a question for Dr. Roddie. When you're looking at data sets just generally or talking to patients, I guess what do you emphasize on the efficacy side, things like median DOR, sort of landmark DOR, CSF, maybe a landmark for the therapists. Just curious what you look for there as kind of highest priority.

I mean, yes, so I think there's lots of different hurdles to jump over here in terms of your efficacy. Your efficacy can be whether or not you can generate a product. So that's efficacy #1, efficacy #2, whether you can get that product into your patients in a timely way. The next binary assessment is whether or not we can achieve a response, so it's the day 20 assessment. These are all important factors that we consider when we're looking at the possibility of a product to take sort of more broadly in patients. We've got our kind of assessment to day 28. I give you that the binary yes or no. And then we could get into the nuts and bolts of whether or not we've got [indiscernible] survival and overall survival. So I think those are 2 maybe important cases information as well. And we look forward to being able to analyze [indiscernible] more detail and present towards the end of the year because those are very sort of meaningful -- clinically meaningful piece of the data that are important to patients and their families as well. But we're not in a position yet to be able to, unfortunately, present the update for this cohort. I don't know, Christian, that you've got an answer.

Well, I think part of what I think if I understood the question correctly, Matt, was that we wanted to understand, I mean, what parameters really matter and give you a good understanding for the performance of the product. And one of the interesting things when we think about the ALLCAR study was, obviously, we had sort of a clear early indicator for longer-term outcome, which was persisting CAR T cells. But then actually, we did see obviously the decline of inventory survival over time, but then a stabilization from 12 months onwards. And that was actually, in our case, at 46%, so technically below the median. So the median was what it was. But the actual story was not the median. The story was that we have a proportion of patients in long-term remission, which ended up actually being 35% of patients in long-term remissions. And that actually was ultimately really the key element in those patients that did not receive any additional therapy. So I think that is sort of, I think, a key parameter. So it's both median if you just assume it's a straight decline and a straight line decline and the median is meaningful. But what certainly becomes more meaningful, if you see stabilization of these curves and you actually have a true long-term benefit in some of these patients.

If I can ask one more quick one. Are you seeing similar rates of cause of relapse such as maybe CD19-negative disease? Just thinking about if we're hoping that the persistence improvement drives that long scale, just obviously making sure there's no kind of differences and reasons why these patients might be relapsing.

I mean we have got some sort of early binary data on the calls of relapse on the acetic between the half and 2/3 of patients of CD19-negative relapse of the relapsed event. And I guess that stands to read. I mean what we're talking about here is we're essentially putting a kind of selection pressure, if you will, on the leukemia blast here by virtue of the persisting highly proliferative CAR. So again, this is something that we saw in the Phase I dataset in that the pediatrician is also saw the Phase I B-ALL study in their hands is CP negative relapse is seen on this study, more so than CD19 positive relapsed this would be the case with Tecartus.

The other parameter we were looking at is how the product was performing the studies were performing at a 6-month time point, where we obviously with 9.5 months of follow-up. The 6 months data, we believe, is firm. And what we can actually see is that the data tracks extremely well between the various studies. We don't see any differences between them. So if you look at those parameters where we have solidity in the curve, actually, we do see that it matters very well, including, as you see in the data on persistence, which is actually matching very well. So we have everything we see at this point appears to tell us that we're really tracking to what we have seen prior.

And our next question comes from the line of Asthika Goonewardene from Truist.

I would also like to add my congratulations on a good-looking dataset and also executing on the study here. I got a 2-part question for Dr. Roddie. And then, Christian, I got one for you afterwards. Dr. Roddie, when do you typically expect to make the decision to give a patient on remission, a sensor transplant? And then the other part is that you reported that at this data card on FELIX, about 61% of patients remain in remission without further treatment. So related to the earlier part of my question, what proportion of these patients in remission have gone past this window when the decision to put them on transplant is made?

Yes. I mean this is something that we debate amongst ALL physicians the world over, because it's not an easy decision to make. And I think that allogeneic stem cell transfer, as you know, has got huge limitations in terms of treatment related to mortality, donor availability, et cetera. And so I think our -- in our own practice, usually, what we would do for patients is transplant naive and haven't had a prior transplant. Then in that setting, we would cancel the patient after they received CAR T cell therapy. And if they were ongoing in ongoing remission after their CAR T cell therapy, but really importantly, if they have ongoing CAR-T-cell persistence, then I think the argument is that in the context of sort of a very sensitive MRD assay on the ability to track your CAR T cell engraftment, our usual practice is to watch and wait. And that's made more attractive if you have, say, another available line of therapy that you could potentially use should they require it, should they relapse and you wanted to bridge them to a transplant. So that would be our usual practice. On the other hand, if you've got a patient who has lost their CAR T cell engraftment, then that perhaps and the recovery of their B cells, and that may be a slightly different in clinical situation. And so then you could look to the -- we take guidance from the pediatric data, and the pediatric data would suggest that the patients who tend to relapse with the loss of B cells is often those who have high baseline disease burden. So when the patients who came in with high baseline disease burden and who lost their CAR-T cell quite early, their B cells, that would probably precipitate a decision to do a term funds and remission. So I think that, that would be a sort of like a kind of a sort of a summary of a very practical approach to allogeneic transplantation in patients who are in remission post CAR-T in the adult setting. On the other hand, patients who had a prior transplant. I think the data for second transplant is still poor. There was an EBMT study that we published in 2019 with Arnon Nagler the main author, and the outcomes were absolutely abysmal by overall survival of around 7%. So again, that's a population where you just continue to watch and wait.

Okay. And if you were to make a guestimate here about what proportion of the patients who are in remission -- remaining in remission right now may be considered for sensor transplant? Or could you maybe quantify how much the do you think you could keep in this just watch and wait setting?

I'm not sure that I have an ability to sort of answer that question, what proportion could be -- would be -- could be transplanted. That's something and probably need to go away and look up and get back to you because there's lots of different factors at play here, not least, patient age, donor availability, baseline disease burden like I say, and the ongoing CAR T cells persistence on the ongoing B-cell lately. So there's multiple factors here that we need to consider.

I think it's fair to say, Asthika, though, that when you look across the various studies that were conducted in the space, that most of the transplantations actually happen early and not late. And when you look at where we are with the study, obviously, there's still a few patients that are in the earlier segments that are within 2 months to maybe 5 months away from the original treatment, there may be a chance that maybe some few patients might still actually experience a transplant. But everybody that's sort of further out the probability of those patients to going on to treat is relatively low. Overall, we see the same dynamic in terms of transplant that we've seen in the other studies, and we don't expect the numbers to actually differ in a significant way.

Got it. That's very helpful, guys. And then, Christian, something for you, please. I've got to change in my notes here, and it seems like 72% transduction efficiency is quite high for CAR-T manufacturing. What's enabling you to do this and to get you there? And then related, when you get high transduction efficiency, are you also getting a higher vector copy number? And related to that, does the higher vector copy number also give you better ability?

Really interesting set of questions. So first of all, the transduction efficiency is something actually that you typically look to dial in. And you want to dial in at a level that is above 50%. Clearly, we felt strongly that the range of rate is actually really good. You also want to make sure you're not actually overdo it because if you do it, you may have then actually have multiple insertion copies per cell, and you may actually get to more distribution. So in order to actually get basically single insertion copy per cell but actually almost all the cells are a very, very high percentage of the cells expressing the CAR T, the chimeric acid in receptor sort of the sweet spot where you want to go, so that you don't have a huge variability in expression across the cells but actually keep it relatively tight, but at the same time, go to the upper end of the range where most of the CAR-T versus the T cells that you actually have in your product are indeed expressed in CARS. So that's where you -- that's the sweet spot you want to go to. And that's actually what we're dialing in from a process development perspective to actually go with a level of transduction that gets us exactly to that point. And I think that's a key parameter also for the consistency of the outcomes that we're seeing.

Just to speak to the VCN data from the Phase I study, essentially, it was less than for all of these despite the high transduction efficiency, and I think the transduction efficiency is partly as we select in the manufacturer process, we select for T-cells. But I think we also have a nice stable CAR. Some of the contracts that you see out there are quite unstable on the cell surface, but we have a really stable expression on the cell surface and we have a nice modality for staining for the CAR expression as well. So all of this, I think, feeds into that really nice high transduction efficiency you see.

And our next question comes from the line of Mara Goldstein from Mazo.

So there's been a lot of questions asked, but I guess this is for Dr. Roddie. And I know we talked about obviously the differences between Tecartus and obe-cel and some of the demographic differences. But I'm curious, though, when you look at the 2 products together, how in a real world situation do you make a determination given the differences in the adverse event profile to use Tecartus over obe-cel or obe-cel over Tecartus. I'm wondering if you can comment on that.

I'm just wondering -- I mean you've obviously heard the data of what would you choose, if it for the patient or a patient relative going into the clinic room, and you saw those presentations, what would you ask for?

Right. Well, I guess that's my question.

Yes, well, the point is that you've heard the data as well. I mean I think that my position on this really clear. One of the most difficult things in the hospital setting is having a treatment that is so incredibly toxic that you have to give to a little very elderly patients often, very frail patients often. You know it's going to land them in the intensive care. The net is going to land them on a whole lot of immunosuppression, and it's going to then engender elusive infections that may well and end up with the Grade 5 event. That is just not a scenario that I feel comfortable with. But feel patients -- when we cancel patients about Tecartus and in the U.K. up to date, it's been in the contacted model, cell lymphoma. We're very frank, and we talk about, I mean, at a 10% treatment-related mortality. If you see this [ Tecartus ] in the [indiscernible] cell lymphoma setting, and we have limited experience in the B-ALL7 recently improved in the U.K. That's not something that's effective. It's not really feasible deliver that to elderly patients. And what's more is because of the toxicity profile in the highest business of neurotoxicity and the protracted course of then steroids and all the infections, we then end up with patients you're sort of essentially occupying a hospital bed for what can be sort of weekly months at the time. They then become really disabled, and there's a protracted period of busy therapy that's required to get them back again. All of these features, all cumulatively impact quality of life and the impact the feasibility of this as a therapy. And I just think that we really sort of need to completely swivel here and change our perspective and make toxicity really a priority monitoring toxicity, having a therapy that potentially if you come in with a sort of a limited disease burden. If you could have your therapy potentially as a lot patients, you don't even need to be in a hospital. These are things that are really appealing to patients. They're appealing to patients. Fondly, they're appealing to physicians. They're appealing in sort of [ economists ], et cetera. Having a product these been delivered safety with minimal toxicity, potentially also as my patient. I just don't even feel that this there's a competition whenever you look at it from that perspective. And then I think coupled on top of this, and I know I sound like a broken record, but having a product that's associated with persistence that in a Phase I study, if we correlate the persistence with long-term durable response rate, which is clearly not apparent with the alternative agent you mentioned, I mean I just don't see where there's a competition here. I just really don't. And that's why I ask you, just as you're coming from a less sort of endless position on this. You're seeing the data fresh. That's why I was interested in your perspective.

And our next question comes from the line of Rajan Sharma from Goldman Sachs.

I just had a couple left, actually. So firstly, just could you provide any more visibility on the 11 patient deaths between enrollment and infusion and what the drivers may have been there? And then secondly, just on the veins delivery. So I think you talked about the median of 21 days. What was the range there? And what could have been driving kind of a faster or slower delivery?

I think I take the second question. I think you go for the first.

When you say, sorry, do you mean the 11 patients, so when you're talking about the patients who enrolled and those [indiscernible] didn't receive their product?

Yes. Exactly. I think there was 112 enrolled and 18 discontinued and then another...

The majority of those patients have progressive disease. And basically, they have progressive disease. And unfortunately, it wasn't possible to continue. And this is something that we see again, in the CAR-T space quite a lot because, as you're aware, you need a good performance status to be able to receive a product. And whenever you've got a performance status of [indiscernible] and potentially your bad bank, it's just not a safe option for you. So in those sorts of settings whenever the performance status is close by and talking in the context of progressive disease refractory to bridging, That is the major reason why patients don't proceed. I just -- I think we have a slide, we can maybe go back to in a little bit more detail to you. There were a few other clauses that have been an equal well for other reasons why patients didn't get there.

Okay. And within those 11, do you know what proportion actually received the bridging therapy?

So I can't -- again, we can probably dig that data out for you. So I don't have the granularity of that, but our publications will be able to do that.

Yes. I mean most of the patients obviously do receive across all the studies that were conducted in this space to receive bridging therapy of some sort. This is a very, very aggressive former disease, very rapidly progressing. And the patients that obviously did not actually get a chance to get dosed are basically patients, you have to move on palliative care. So they're at a point where therapy didn't make any sense anymore. And I think that is sort of just -- I think to put it in perspective, this is a true difficult patient population that are in pretty horrible condition. And so that's why you lose a component there. You obviously link the question to the turnaround time. So first of all, range of delivery, we did show the exact for each patient was shown. And if you looked at the chart, yet every single patient delivery time actually on that chart. What you see it is actually amassed around 21 days, and it goes between about 18 days and about 25 days. Part of the difference there is logistics depending on whether there are any issues on actual air travel. And as I indicated, we did this study during the height of the pandemic where air travel actually was significantly disrupted, not just domestically within the U.S. but also transatlantically. So that has some impact. I think what's important to understand, and this is what I was highlighting before, is that our vein-to-vein time, so the time from taking the cells to infusing the cells, was virtually identical with the real-world study conducted by Kite, which also gives you a very, I think, good appreciation that, first of all, what we've been doing, what we've been able to do is highly competitive. What is driving in terms of the turnaround time, the 2 components that drive that, one is the time that you need to manufacture. And that obviously doesn't change much once you actually lock in your process, and that is in the range of 6 to 7 days of actual manufacturing. The remainder of the time actually is related to product release, and that is analytics. A lot of that actually is linked to the determination of sterility, which, at this point, still requires you to actually around place-based assays. That is likely going to change over time. What our current estimate is based on the changes that we have implemented on the analytical side, which is in actually conversation with the agencies will get us to a point that at the time of launch, we're going to be in the range of about 16 days from vein to certification. So about 6 days less to what we had in the study, 5 to 6 days less of what we had in the study, and we expect that to be further reduced with improvement on the sterility testing that should get us actually between 10 and 14 days. So that's where we're going to be. So this is going to be highly competitive and very well manageable.

And our final question for today comes from the line of Sebastian Banesto from Vanchem.

Congrats on the presentation results today. My first one is on the correlation between the high tumor burden and the Grade 3 CRS and ICANS. Can you maybe comment on the possibility to later in the real world for obe-cel treatment [ 75 ] patients for treating in the -- in or outpatient setting based on like tumor burden at baseline?

So I think a really good question. I think what we have quoted on the slide, and I'll let Claire answer some of the more specific points, but what you -- what we've indicated in the presentation is that we have 7 patients that experienced high-grade ICANS. Six of the 7 patients had more than 75% tumor lymphodepletion. In other words, this is [indiscernible] that even with bridging could not actually be brought down at all, and so this is a very extreme portion or a very extreme end of the spectrum of tumor burden. And obviously gives you -- before you dose gives you a very clear indication that if the patient has this very high level of tumor burden at the time where you make the going to lymphodepletion, that you actually have a patient that you want to watch more carefully. The other proportion of the patient, obviously, had very infrequent, just about 2% or 3% of probability of developing a high-grade ICAN. And that obviously gives you an ability to have a -- place this patient in a segment of the hospital, which would be in the U.S. often be considered hospital outpatient type of setting. As we go forward, obviously, we'll learn more about, I think, the -- how this -- the adverse event profile evolves. But I think what we're going to be seeing or we're looking at, to your point, is also look at the start to revert that inclusion. But the much more predictive one that we're seeing is not an inclusion. It is a tumor burden at lymphodepletion because the trajectory that a patient is on how fast the disease progresses can vary quite a bit between patients. There can have patients that go from mineral residual disease to full-blown disease within 3 weeks. You can have other patients that come in with a medium level of tumor burden, and they're still at the same median level of tumor burden at the time of lymphodepletion. So the real determinant here that we find is really the tumor burden at lymphodepletion, which is also the time where the physician actually needs to make a decision where you actually talk to the patient, we can prepare the patient. So it actually gives you a very good handle to actually prepare, also think about the resources you may need, et cetera, and plan from hospital utilization, resource utilization perspective, actually do the planning accordingly. Maybe Claire, you can maybe give a bit more color around that.

Yes. I mean I would concur with that. I think that's where this product really has a huge potential and that is sort of in terms of in versus outpatient stratification. And I think the data is pretty much as clear as it could be that high disease burden is the risk factor for the sort of grade 3 ICANS in this cohort. And they just need to -- even more scrutiy and they need maybe a bit of closer follow-up, but what that means is that the other patients who potentially can have a less invasive post-infusion [indiscernible] experience, which would be ideal for them. I mean as we're talking about patient experience and quality of life measures here as well as economics and so on.

Great. That's really helpful. And then I was also wondering regarding the step dosing for clarification. Can you just mention a number of patients who did not receive the second dose? And what were the reasons to be high in not receiving that second dose? And were those specific patients also taken along in the efficacy and safety analysis?

Yes. So those patients were included in the efficacy and safety analysis, and there were 6 patients who only received one dose. Three of those were due to immunotoxicity. There was a couple of ICANS cases and one CRS. There was one patient in the cohort who [indiscernible] who had [ frank ] progressive disease, unfortunately, after the first dose, and it was felt to be inappropriate to do some with a second dose. And there was 1 death, 1 early death, and 1 manufacturer bid some issues. So I think the bottom line is that there were only 6 patients. Remember that we're talking about almost 100 patients treated. So it was by far by the minority of patients who received one dose, and the majority, the reason for that was because of the new toxicity.

And obviously, the very reason for designing this particular approach, to make sure that the physician stays in control of the therapy and what's going to happen to the patient and do not basically just get into an overshooting activity. But obviously, with the excellent safety profile we've seen in the product, actually, this is much less required to be used. But obviously, it's very helpful if you have patients that have extreme conditions, and you could still give you the safety to actually manage these patients and give you the tools to do that.

Great, my final question is maybe a little bit naive, but I saw -- I was kind of intrigued by the -- how older age correlated higher overall response rate in the [indiscernible]. Can you maybe elaborate on whether this is typical in adult ALL? Or what the possible explanations are behind this effect?

You want to go?

Well, I mean, I think the bottom line is that I think we've got to recognize that -- [indiscernible] patients sort of 18 to 24 year old patients often have quite aggressive disease phenotypes, and they are traditionally quite a difficult patient group to treat successfully and to get into remission. So often teen and young adults, we do encounter problems in the relapsed/refractory setting. And often with the older patients, again, they're carefully selective population in some ways. We take -- we have treated an 81 year old patient, but that patient had a performance status is one in order to qualify for the study. So I think that's my take on it is that the biology of the younger patients disease is actually quite difficult for us to manage with any therapy, and that's my interpretation of what you see is the inverse of what you might expect.

It's also consistent with what we have seen in the development for Blincyto as well, and I think the background that we experienced at that point was that the young patients or the younger adults are patients that actually have a start of their disease as kids. So they go through hell, lots of lines of therapies, and they're really at the very end of the line by the time they actually make it to their 20s, and they end up actually in a study like this. So they tend to be very, very heavily pretreated. Now conversely, the old patients or older patients actually tend to be treated much more with -- much less intensity, which is the story of ALL, is that the old patients or adult patients couldn't take the toxicity kids could take. So actually have less intensity of treatment. And with that, the immune system of these patients are actually still in better shape. So the cells are less selected out by the other therapies, but also the T cells in these patients still actually, actually are less damaged. So paradoxically, that leads into the situation, they actually get a better product and responsive -- a more responsive disease in all the patients. We've seen very similar outcome with Blincyto in that development. So it's paradoxical ou first class, but actually, it's linked to the intensity of treatment.

This does conclude the question-and-answer session of today's program. I'd like to hand the program back to Dr. Christian Itin for any further remarks.

Well, first of all, thanks all for joining. It's been obviously a fantastic day for us. These are kind of days we're kind of working towards for years to sort of we actually get the type of data that we have now with obe-cel. We think we have a real differentiated product. We have one that addresses the need in this particular indication, and we're looking forward to keeping you updated on how the program evolves and how we're driving towards filing for registration. But exciting day, a great place to be, and thanks for all the support from all of you as we sort of are on this journey, and you've been following us as we went through that. Thank you very much.

Thank you, ladies and gentlemen, for your participation in today's conference. This does conclude the program. You may now disconnect. Good day.