Autolus Therapeutics plc (AUTL) Earnings Call Transcript & Summary

Good day, and welcome to the Autolus Therapeutics EHA Investor Call. [Operator Instructions] As a reminder, this call is being recorded. I would now like to turn the call over to CEO, Christian Itin. You may begin.

Thank you, Michelle, and good morning or good afternoon, everyone, and thank you for joining us to take part in today's call to discuss the data presented at the European Hematology Association Congress that has recently been held in Vienna. Joining us today, we're extremely pleased to welcome Dr. Steven Horwitz from the Department of Medicine, Lymphoma Service at Memorial Sloan Kettering Cancer Center and a Professor of Medicine at Weill Cornell Medical College, Cornell University; and Dr. Kate Cwynarski, Chair of the U.K. T cell Lymphoma Group and Consultant Hematologist, University College London Hospital; Dr. Martin Pule, our Chief Scientific Officer; and Dr. Edgar Braendle, our Chief Development Officer, will also participate in this call. Before we begin, I would like to remind you that during today's call, our discussion will contain forward-looking statements. Please make sure you're familiar with our disclaimer on Slide #2. On Slide #3, you will see the agenda for today, which is as follows: Dr. Horwitz, who is also Chair of the NCCN T Cell and Cutaneous Lymphoma Guidelines Committee, will provide an overview on T cell lymphoma, and will describe the unmet need and cover the current treatment options. Martin will then introduce our AUTO4/5 program with Dr. Cwynarski summarizing the AUTO4 data she presented in an oral presentation at EHA. Martin will then round up with the next steps for AUTO4. We'll then switch gears with Martin introducing obe-cel and we'll cover the AUTO1/22 data in pediatric ALL patients that we presented in a second oral presentation EHA and also cover the updated obe-cel data in NHL and primary CNS lymphoma patients. We'll finish today's call with a Q&A. So now I'd like to turn over the call over to Dr. Horwitz. Dr. Horwitz?

Thank you. And I hope everyone can hear me okay. It's a pleasure to join you today. So what I was going to do is just go over a little bit of the landscape in T cell lymphomas, how we treat those patients standardly, what the area of unmet need is and then some of the newer therapies setting up. I think what you'll hear is very exciting data with the cell therapy program. The next slide, I think, is my disclosures. Following along, I'm seeing Slide 4. If you can go to Slide 5. And then you can jump to Slide 6, which is just a bullet point of what I'm talking about. And I apologize, on the webinar, I'm still seeing slide -- there we go. I think there's a delay as we are across the Atlantic. So if you advance to Slide 6. So what I'll just do is a little bit of background. I know this is a sophisticated audience, but maybe not a completely lymphoma-focused audience and then talk about upfront approaches, dose intensification, patient selection, some of the new therapies and then relapse again. So the next slide, please. So I think we think of -- Slide 7 is what I'll be on, T cell lymphoma as an uncommon non-Hodgkin's lymphoma. So we have incidence data in the U.S. that if we look at all lymphoid neoplasms, there's about 80,000 or so new diagnoses a year. And we think somewhere between 7% and 10% of those are T cell or NK cell neoplasms, which we tend to lump together. Next slide. I think there is a bit of a delay. So I'll just keep going and you guys try to keep up. So I'm on Slide 8. So we think of T cell lymphoma as overall heterogeneous and somewhat of a poor prognosis. So out of those 6,000 or 7,000 cases a year, there's 29 or 30 different subtypes. So there's a lot of individualization in terms of clinical presentation, some present primarily in the blood, the leukemic; some present primarily in the skin, those are the cutaneous; and then there's -- the most common subtypes are sort of the nodal and extranodal. I have them color-coded there based on what's most common and what's less common. And I think one of the things that maybe we'll set up to later talk is that when we're looking at new therapies, we're often targeting things -- specific features of the tumor. And that could be surface features, that could be genetic features, but you get the heterogeneity. And then when you're thinking about a surface target as something as fundamental as a T cell receptor, you do have the opportunity to really cross subtypes in terms of your efficacy. Next slide, Slide 9. So I'll just go over the prognosis then with standard therapies. And for the systemic T cell lymphoma, standard therapies are really still combination chemotherapies often with a dose intensification. So then if you can go to Slide 10 on my deck, which shows survival curves from a large national registry from Sweden looking at over 700 patients with newly diagnosed T cell lymphoma. And these really represent the most common systemic types of T cell lymphoma, which are the most common T cell lymphomas. And you'll see the blue curve at the top is one less common subtype. So that's about the fourth or fifth most common subtype ALK-positive anaplastic large cell lymphoma. So that, it develops primarily in children or young adults. And with combination chemotherapy, the results are not spectacular, but the majority of people are cured and you see that there with long-term remission on the progression-free survival curve. When you look at the other subtypes, and again, this is basically with combination chemotherapy regimens, CHOP or [indiscernible] regimens, which are 4 and 5 drug chemotherapy regimens, you see the majority of patients either don't respond and those are that very steep drop of the curve in the first 6 months or relapse after initial therapy. And then that's where those curves sloped down kind of between year 1 and 4. And really a minority of patients have long-term remission from initial therapy. And depending on the subtype, that may be as high as 25% or 30% for ALK-negative ALCL and probably below 10 -- 20% for some of the more aggressive T cell lymphomas. So next slide. This next slide just shows the way I think of conceptually thinking about that heterogeneous group of diseases in terms of how we approach them. So I have them grouped within -- in the middle -- I'm on Slide 11 now, in the middle, are those where combination chemotherapy is reasonable, if you see those in the light blue. So those are the more common nodal types. And the reason we think combination chemotherapy is reasonable is that there is a cure rate, albeit that cure rate may be between 20% and 30% for most of those subtypes. On the left, there's a group of diseases where CHOP-based chemotherapy is not recommended. And some of those are aggressive T cell lymphomas where CHOP therapy or chemotherapy like that is really ineffective. And some of those are more indolent or chronic diseases where combination chemotherapy may work but the remissions are generally short-lived, and those are patients that may have a better long-term prognosis where they need continuous therapy. And in that case, therapy [indiscernible] some accumulations of side effects unless they're curative are preferred. And then on the right, there's a handful of aggressive diseases with very little data or too little data to make a recommendation. So next slide, or Slide 12. So historically, when we've looked at new therapies or others have looked at new therapies for T cell lymphoma, it's really been more of the same. So if chemotherapy worked for some, does more chemotherapy work for more? And what you can see in Slide 12 is sort of summary data from a series of studies done primarily in Germany. This is really before rituximab changed the fundamental treatment for B cell lymphoma where all aggressive lymphomas were treated with combination chemotherapy and not chemo-immunotherapy. And if you look at this accumulation of data, this is adding etoposide to CHOP, this is giving some people at every 2 weeks and every -- versus an every 3-week regimen, so dose intensification with more chemotherapy, increased dose density with a shorter treatment intervals. You can see that in some patients, we may do a little better in terms of progression-free survival. The caveat here is that this data is limited to younger patients under age 60 and most of these patients had lower risk scores. That was based on some of the clinical trials data done at that time. If you look in terms of the toxicity cost on the right, when we do dose intensify, there is a consequence in terms of hematologic toxicities. And in younger patients, aged 18 to 60, we can usually get away with it with more [ heme tox ], but patients clinically do okay. But once you get over age 60, you see clinical consequences, which can either be life-threatening or blunt any attempt to increase dose or intensive high dose of chemotherapy and then that increase in toxicity kind of washes out any net increase in efficacy. If you go to the next slide, which is Slide 13, it's sort of the maximalist chemotherapy approach that we and others practice for appropriate patients. And this is just some Phase II data from a Nordic study, where patients got CHOP plus etoposide as an induction. So that's their initial chemotherapy. Again, this is Slide 13. And then that was followed by high-dose therapy and autologous stem cell transplant, so truly maximal chemotherapy. So these patients are a bit selective because they enrolled on a clinical trial that involve very intensive therapy. So most of these patients are under 60 and have good performance status. And what you can see, I think if you just look at the numbers in the boxes on the left and right with overall progression-free survival, and I'll focus on the progression-free survival because that's really the readout from the benefits of therapy, is that if we really give as much chemotherapy as a human can tolerate, we can push those progression-free survival rates up to about 40% for many subtypes, maybe into the 40% to 50% for angioimmunoblastic and higher for ALK-negative anaplastic large cell lymphoma. So that's sort of the most we can do with maximizing chemotherapy. Next slide. So other attempts have been made to add to it. So if we sort of follow a paradigm where we have another active drug that's not just chemotherapy, can we do better? So Slide 14 just introduces the topic of adding alternate therapies, non-chemotherapeutic approaches to a CHOP backbone. And if you just jump ahead then to Slide 15, this is really the one success we have, and there's been a bunch without success. But the one main success we have has been the ECHELON-2 study, where we added brentuximab vedotin, which, as you know, is an antibody drug conjugate targeting CD30 and treated specifically patients with CD30-expressing T cell lymphomas. These are patients who were treated in the upfront setting, and they received either brentuximab plus a CHOP-based regimen with placebo vincristine, concerns about overlapping neurotoxicity if we added vincristine, or CHOP plus placebo brentuximab. And I think one of the keys from this study when we're looking at the data is the majority of patients on the study had anaplastic large cell lymphoma, which is a universally CD30-expressing lymphoma; and other subtypes for CD30 that's more variable, patients had to have at least 10% CD30 expression on their tumor cells. So if you go to the next slide, Slide 16, that just shows the 5-year update from -- for progression-free and overall survival. And what you can see is that there was an incremental benefit adding brentuximab vedotin to CHOP, both in terms of progression-free survival -- this again will be Slide 16. And those curves stay separate even with long-term follow-up. You see the median progression-free survival was almost 3x as high with brentuximab vedotin plus CHP in the green curve and the overall survival -- the median overall survival was not met with either arm, but a significant improvement from BV-CHP. And you can see the delta there. While the relative improvement was vast, the overall improvement was between 15% and 20%, absolute difference. Next slide. Slide 17 looks at the subtypes outside of anaplastic large cell lymphoma. So the majority of patients, again, in the study, about 70% had anaplastic large cell lymphoma, that one subtype, and that's where we see the greatest benefit. On Slide 17, we went and looked at the other subtypes of T cell lymphoma. This was not a preplanned analysis. These patients were not stratified, but what you can see is that the benefits of brentuximab plus CHP in the green curves are a little less distinct for peripheral T cell lymphoma, not otherwise specified; and in angioimmunoblastic T cell lymphoma, sort of unclear that the brentuximab added. [ As it caused ] the AITL curves, those curves, the blue curves with CHOP are really higher than anything we have in the literature. So that may reflect some patient selection or CD30 expression on AITL or just the randomness of small numbers. Next slide, Slide 18. So other attempts have been done to sort of recapitulate this. Can we add other active agents to chemotherapy and better for all? And the next slide is primarily a French study, adding romidepsin to CHOP and randomizing [ data against CHOP ] for untreated patients with peripheral T cell lymphoma. Again, romidepsin is a histone deacetylase inhibitor that, at least in the U.S., is approved for cutaneous T cell lymphoma and until recently was approved for peripheral T cell lymphoma. And you see here, this was a negative study with really no benefit in the blue green curve, adding romidepsin to CHOP in terms of progression-free or overall survival. Next slide. Part of the explanation for that, if you look at Slide 19, maybe that romidepsin did add significant toxicity. So on Slide 19, there's a table on the left that shows that patients getting romidepsin had higher rates of dose reductions and dose interruptions due to toxicity. And then if you see the forest plot on the right -- and this will be Slide 19, I'm still seeing Slide 18. The forest plot on the right, you see that the other issue may be that this was an all-comers study or no selection based on who enrolled in this study. We now know with some of this knowledge that since the study was designed, that patients with angioimmunoblastic T cell lymphoma, if any, have the greatest chance of response to romidepsin. And when you look at those hazard ratios, there's a sense that maybe if you could find the study to that group and expand it a little bit, you might find significant benefit. Nonetheless, this was a negative study. So if you go to Slide 20, you'll just see a list of attempts at adding new therapies to chemotherapy backbone, you'll see that most of them have been negative studies, either due to lack of efficacy or in some cases, like alemtuzumab to increase toxicity. 20. And then if you -- that's sort of a number of negative studies. And if you want to move to Slide 21. I think one of the issues that comes across when we're trying to use new targeted therapies is that there's a tremendous heterogeneity as we're understanding more about the biology in terms of the genetics, the molecular pathogenesis and you can just -- back to Slide 21. And you can just see over on the left there that there's a number of pathways that are deregulated, and these may have different targets when we're thinking about targeted therapies, either at signaling pathways or epigenetics. Next slide. And you'll see on Slide 22 that maybe one of the reasons that some of the genetically targeted therapies have been less successful in T cell lymphoma is that there's extreme heterogeneity. And this is a series of 133 cases, again, in Slide 22 , looking at mutation. And what you'll see is that there's a vast array of mutations. And outside of things like TET2 and TP53, most of the mutations are in less than 5% of cases. So there's a lot of heterogeneity that isn't well addressed by genetically targeted therapies. So next slide, Slide 23. So I think when we think about adding to combination chemotherapy, at least with a targeted agent, we know that for anaplastic large cell lymphoma -- or with a very powerful drug like brentuximab vedotin, if we target the universally CD30-expressing T cell lymphomas, we can have benefit. I think when you look at other subtypes of T cell lymphoma, sort of in that middle bullet there PTCL, NOS, AITL and this goes for a number of other subtypes, we really can't draw definitive conclusion by adding a single drug. And some of that is due to the heterogeneity of the diseases we're treating. So I think as we're trying to go forward, adding to new therapies for T cell lymphoma, I think we need to be able to enrich for those most likely to benefit. And if it's a targeted therapy, that target should be broadly expressed and also not add significant toxicity. So I'll just give a few slides on relapsed T cell lymphoma with what we do standardly now and then turn it over to Martin. So Slide 24, just transitions to relapsed T cell lymphoma. Slide 25 shows data with allogeneic stem cell transplant in the relapse setting. So we know that in patients with relapsed T cell lymphoma, they can sometimes be cured with allogeneic stem cell transplant. Again, this is on Slide 25 now, if you can catch up. And that, that cure rate is about 40%. And most of the cures are in patients who have a complete remission. So when we're thinking about treatments in the relapsed setting, we're either thinking about it in a younger, healthier patients getting a complete response to bridge them to allo or in a patient where allo transplant is not feasible due to all the reasons allo transplant is not the right treatment for everyone, we're looking at drugs that can maintain remission. And right now, for our current agents that usually means on continuous therapy. So if you go to Slide 26, this shows some of the standard drugs that are FDA-approved in the U.S. or that we have available here for routine use in T cell lymphomas. And what you see in this table, again, Slide 26, is that most of these drugs have response rates between 25% and 30%. Sorry, if you -- there you go. So if you look at the -- histone deacetylase inhibitors, you have belinostat at the top and romidepsin at the bottom, their pivotal studies had response rates between 25% and 30% with only about 10% to 15% of patients having complete responses. And we think a complete response would be important to bridge to transplant. When you look at the duration of remission of those drugs, you do see that patients can be on those drugs long term with continuous therapy to try to maintain remission. Very similar data with pralatrexate, which is an anti-folate, overall response rate just under 30% and a complete response rate of 10%. But in that subset of responders, you can sometimes maintain remission. Of course, the outlier in the middle of brentuximab vedotin for anaplastic large cell lymphoma with a very high response rate in that subtype. If you go to Slide 27, you see the progression-free survival curves to sort of graphically display what we achieve and what we don't achieve with drugs in the relapsed setting. So on Slide 27, you'll just see the PFS curves from the pivotal studies of these drugs showing that the majority of patients either don't respond initially or relapse after brief responses. And there's a subset of patients, more in the 10% to 15% range that really have long-term benefit. So there's a tremendous unmet need in T cell lymphoma [ therapy ] that is both effective, and if it had the potential of creating durable remissions, would be incredibly valuable and one could think of many ways that could be applied to improve treatment for our patients. If you just go to the next slide, 28, I just list some other emerging therapies in T cell lymphoma. A number of drugs are being studied, inhibitors of the JAK-STAT pathway, inhibitors of the PI3 kinase pathway, EZH 1/2 inhibitors. And these drugs all fit sort of what we've seen before, which is they are probably targeting vulnerable subsets, and they don't cause durable remissions off of therapy. So at least for the top 3 bullets, you would need to maintain patients on therapy to get durable remissions. And then very early days for immunotherapy, which I think we'll talk about more in terms of some cell therapy studies being explored, some checkpoint inhibitor studies and some anti-CD47 strategies, and most of these studies are in their Phase I experience. So just to conclude on Slide 29. Peripheral T cell lymphoma, if we think of first-line treatment and then where we go from that, peripheral T cell lymphoma remains heterogeneous and poor prognosis for too many of our patients. Again Slide 29 is my conclusion. We do have cures for some patients with combination chemotherapy, and that's probably about 20%, 25% overall, maybe as high as 45% for those that can endure most intensive therapy. And some improvement for those particularly with anaplastic large cell lymphoma, where we see adding brentuximab really provides significant benefit. But overall, in the universe of T cell lymphoma, there remains a high unmet need. Most of the attempts to date to improve therapy have been adding active agents to upfront therapy. And I showed you that we have had some success there, but also a lot of failures in being able to do that, both due to our inability to enrich populations and our inability to minimize toxicity. And then I listed for you some of the newer approaches that are underway. So I will conclude there, and thank you.

Thank you very much. It's a wonderful presentation. And if we now move to Slide 31. So as Dr. Horwitz has shown you, T cell lymphomas represent an area of significant unmet clinical need with a majority of patients being either refractory or relapsing after initial treatment. Standard of care is variable and often based on intensive treatment. But despite this, the median overall survival of 5 years, it is about 1/3 of patients. For patients with relapsed or refractory disease, their prognosis is even worse with a median overall survival of less than 6 months. In terms of immunotherapy, brentuximab can be useful to patients who are CD30 positive, but this is mostly the patients with an ALCL subtype of T cell lymphoma. And I think it's fair to say that generally, T cell lymphoma has not benefited from immunotherapy the same way that B cell lymphomas have. And if we go to the next slide, the reason why, of course, is that when targeting B cell malignancies, we target pan B cell antigens, and pan B cell depletion is reasonably well tolerated. Doing the same thing in targeting a pan T cell antigen would result in unacceptable immunosuppression. If we put that aside for one moment and we were to pick the perfect antigen to target in these diseases, it would be the T cell receptor CD3 complex because it's expressed at high level on most subtypes and in nearly all cases. And the way that the biology of this receptor actually gives us a therapeutic strategy that can stop us from having pan T cell depletion by targeting it. So most people know that the alpha and beta chains with the T cell receptor rearranged during T cell development. But quite a few people have forgotten that actually during this rearrangements, the T cell has to pick what beats a constant region gene it takes. So either a T cell -- it either becomes a TRBC1, T cell expressing the TRBC1 and beta constant region or a TRBC2 T cell. So in our peripheral blood, there's a mixture of these two cell types. T cell lymphoma is obviously clonal, so it's going to be entirely one or the other. So this then gives us a therapeutic strategy when, for instance, if you have a patient with a TRBC1-positive lymphoma, it should be all TRBC1 positive. And if you target it, you should be able to target all of the lymphoma cells, but only about half of the normal T cell compartment. So that's the therapeutic strategy. And if we go on to Slide 33, I think this shows you how useful a target the CD3 T cell complex is. So it's expressed on 95% of T cell lymphoma subtypes. And the exception really are these extremely rare lymphomas that come from NK cells or gamma delta T cells, but all the other alpha-beta-derived T cell lymphoma subtypes will express the CD3 TCR complex. There are other potential T cell lymphoma targets. These either [ are problems ] one way or the other, either they're expressed in normal T cells as well or T cells when they become activated or else they're only expressed on a proportion of the entire set of T cell malignancies. If we go on to Slide 34, I said that -- this fact about TRBC1 or TRBC2 gene selection is often forgotten. I think if you look on the left of the alignment, you'll see why this is often forgotten because the two chains are really almost identical. So there's only two amino acid inversion, that the amino terminus is a constant region that is really targetable. And remarkably, we've identified an antibody that's exquisitely selective for TRBC1 and we've flipped the selectivity of this antibody from TRBC1 to TRBC2. So now we have two antibodies, one that binds to TRBC1 and one that binds to TRBC2. And I think the nicest illustration is that scatter plot at the bottom middle of Slide 34, where you can see normal presence peripheral blood and you can see two nice populations of either TRBC1- or TRBC2-positive T cells. If we go on to Slide 35, we've converted these two antibodies into chimeric antigen receptors. And on the top right, you can see some very simple [ killing ] data with [indiscernible] that are natively TRBC1-positive or [indiscernible] that have been genomically engineered to be TRBC2-positive or [indiscernible] that are TCR-negative. And you can see a nice selectivity of our TRBC1 and TRBC2 CARs with a CD19 CAR as a negative control. And we can reproduce the same findings in an in vivo model in grafting TRBC1 or TRBC2 T cell lines and treating with either TRBC1 or TRBC2 CAR T cells. So if we move on then to Slide 36. So the AUTO strategy for T cell lymphoma is a companion diagnostic initially to identify what patients with relapsed/refractory T-cell lymphoma is either TRBC1 and TRBC2. And then TRBC1-positive patients are treated on the currently open AUTO4 study and in future, we'll open a TRBC2 study targeting TRBC2. And Julie, if you go to Slide 37, just a very quick word about companion diagnostics. You can actually figure out if a patient is TRBC1 or TRBC2. It's relatively straightforward NGS pipe assays that looks at VDJ recombination. So that's something that's fairly standard you can do on genomic DNA. We've also developed antibodies that work in fixed tissue, so you can do it by FFPE, if you want to. And also, we can use the same antibodies by flow, if you do lymphoma diagnosis by flow, which is amenable for some subtypes. And I think having introduced the concept, I will hand over to my colleague, Kate, who's going to just run you through what she's presented at EHA. So Kate, over to you.

Thanks very much. I presented this data in Vienna at the European Hematology Meeting on Saturday. So I presented the initial data from our Phase I study of AUTO4 in patients with relapsed refractory TRBC1-positive peripheral T cell lymphoma. So this has a -- sorry, next slide, please, a two-part consent. So we have seen patients to screen for the tumor. We can use archived tissue to assess whether it's TRBC1-positiveTRBC1-positive or not using next-generation sequencing as outlined by Martin. And those that have TRBC1 tumors then can be screened in Part B. We do allow bridging therapy. And you can see here that from leukapheresis to [ QP release ], there's a median of 20 days and the efficacy and safety follow-up with first disease assessment at day 28. Next slide, please. To date, we've treated 10 patients, a median age of 55 and median prior lines of therapy of 3. Advanced stage in the majority, so 80% have Stage 3 to 4 disease and half of them peripheral T cell lymphoma NOS; 4 patients with angioimmunoblastic T cell lymphoma; and just one without negative ALCL. 1/3 of the patients had a prior autograft; 70% ECOG 1 and bridging therapy was used in 70%. Next slide, please. So to date, as I've mentioned, we have treated on four different -- in four different cohorts with increasing dose escalation from 25 million cells up to 450 million cells. You can see here, there have been no limiting toxicities. And only mild cytopenias likely to be because of the FC condition of fludarabine and cyclophosphamide as this was seen at low dose cohorts. In terms of any grade CRS, you can see we just have one grade 3 CRS and no grade items at all. So generally, really well tolerated. And in terms of data, this is certainly encouraging, although this is very early Phase I data. So if you look at the two highest doses we've used about 225 and 450, so all four of those patients were in complete metabolic remission at month 1, and 3 out of 4 at 3 months, 2 ongoing remissions. And this is in the early phase study, really encouraging. Next slide. Sorry, I think you have -- so as I mentioned -- next slide. So if you go back to the next slide, I'll wait for that to be seen. I'll just mention that here, you can see the highest doses mentioned at the top, and the green line represents complete remission. And you can see at that 1-month and 3-month intervals, with encouraging responses with two of those patients still in remission, really encouraging for Phase I. So this is a PET scan from one of our patients I've treated. They've had four lines of therapy. They've never achieved complete remission. They were in complete metabolic remission at day 28, and that's 3 months still in complete metabolic remission and remains in clinical remission at month 5 or waiting the 6th month scan. Next slide, please. You can see that the PET scan, they go complete remission at day 28 in 3 months. So next slide, please. So we've detected CAR T cells in the lymph node biopsy for one of our patients who achieved complete remission. But to date, we haven't identified CAR-T expansion either by PCR or flow in the peripheral blood. Next slide, please. I'll hand over to Martin. And unfortunately, I have a clinic that's just starting. And after a jubilee weekend of being away at EHA, I can't miss this one. But thank you very much.

No, thank you so much, Kate. Thank you. Really appreciate it. If you can go on to Slide 46. So to summarize, AUTO4 is generally well tolerated. The early efficacy is encouraging, particularly looking at the higher dose levels. We've detected CAR T cells in lymph nodes, but we haven't seen expansion in peripheral blood at this point. The study is ongoing. We are also optimizing our manufacturing process. And the plan is to enroll further patients and try and define the recommended Phase II dose in preparation for a pivotal study. So with that then, I'm going to change track. And if we go to Slide 48, Julie, we're going to -- I'm going to talk about obe-cel now. And many of you know about obe-cel, to remind you, this is a CD19 CAR with a fast off-rate binder. The idea here is to reduce the amount of inflammation per target cell encounter, which in turn, will reduce the amount of immunotoxicity in the patients, reduce exhaustion in the CAR T cell and improve engraftment and overall persistence. And those of you not familiar, I refer you to a paper by Sara Ghorashian at Nature Medicine in 2019. So if we go on to Slide 49. So again, many of you will know that obe-cel is currently being tested in a Phase II study of adult relapse/refractory B-ALL. But what we presented at EHA was some ongoing work with obe-cel in NHL, primary CNS lymphoma and also a 21, 19 variant of obe-cel in pediatric ALL. So these are the areas I will quickly touch on next. So if we go on to Slide 50. So the ALLCAR19 study is currently a basket study where we're testing obe-cel in a variety of B cell malignancies, and we presented an update as a poster. And if you go through this data very quickly, our follicular lymphoma cohort, and we've treated seven so far. Sadly, we lost one patient to COVID, but all other patients remain alive and well in complete remission. We've treated six patients -- I'm sorry, seven patients with diffuse large B cell lymphoma and six remain in complete metabolic remission. So if you look at the median follow-up of -- for the follicular and diffuse large B cells, they're 9.2 months and up to 19.1 months. So that data continues to accumulate. And we've treated three mantle cell lymphoma patients, one relapse in 6 months, actually with a skin relapse that was salvaged with radiotherapy and an allograft, but the other patients remain in a metabolic remission. We've also started treating some CLL patients, and from the three patients that have reached the initial evaluation, two have gone into a molecular CR in their bone marrow and -- but have some residual lymphadenopathy on CT scan. And just to remind you that CLL patients don't tend to get PET CT scan. So this is just a regular CT scan. And what we have in these patients is some lymph nodes that are larger than 1 centimeter. And the significance of these lymph nodes at this point is still undefined, those two patients remain well. So that's really our update with ALLCAR19. So if we move on to the next on Slide 51. So we're also exploring obe-cel in primary CNS lymphoma, and primary CNS lymphoma is a type of diffuse large B cell lymphoma, but because of its anatomical location, it has a particularly poor prognosis, initial treatment is often intensive and performance status of these patients tend to be poor. So we've tried obe-cel in these patients. And actually, we haven't seen any CRS -- any significant CRS or any Grade 3 or 4 CRS. We have seen some ICANS. We haven't seen any Grade 3 or 4 ICANS in ALLCAR19, for example. But in these patients, we did see two. I think one of them may well have been due to progressive disease at the time and the other patient with a pontine lesion, which is quite a difficult anatomical location. If we go on to Slide 52. We put the slide here to show you that despite these patients having no disease outside of the CNS and having had lots of rituximab treatment, we still see really nice expansion of obe-cel in peripheral blood. And then if we go on to Slide 53, just to give you a sense of results. I mean I think the first thing to look at is the MRI scans on the right. And you can see quite a large parenchymal lesion in the top left that is completely resolved. And I think this is -- I put this here to show you that obe-cel can result in remissions in patients with quite large parenchymal lesions, which is interesting. The other thing is to remind me to tell you that actually you can't do PET CT easily for these patients because the brain takes up glucose. And I think when you see residual abnormalities on an MRI, it is difficult to know exactly what that means. And with that in mind, if we look at our outcome, our overall response rate was 4 out of 6. And we have seen some patients converting from, for example, stable disease to a completely normal MRI. But again, this is still very, very early data. So kind of looking where we are with sort of this sort of franchise for obe-cel, we can see a favorable and consistent safety profile across all indications of patients evaluable for efficacy across mantle, diffuse large B cell lymphoma, follicular and CLL, the overall response rate was 90%. And the B-NHL cohorts, the complete response rate was 94%. The CLL cohort is interesting, and we have to see what that sort of low-volume residual lymphadenopathy on CT means. And I think as well as having a good CR rate and a good -- and excellent safety, I think as more and more time goes on, you can see that these responses are durable. I mean the next steps with this is really trying to treat 10 patients of all those different indications on the ALLCAR19 study and to continue finishing our primary CNS lymphoma study, which we aim to recruit 12 patients in total. So that's obe-cel. And now if we go on to Slide 55, let's take another step forward to AUTO1/22. So we have tested obe-cel in children with relapsed/refractory ALL, and we see the same as we see in adults. However, in pediatric leukemia, since children are better able to tolerate toxicity, toxicity isn't really a major issue. But what is the major issue is antigen negative relapse or CD19 loss. The obvious solution to this is to target a second early pan B cell antigen and the one -- other antigen you can go for is CD22. CD22 is actually a much more challenging antigen than CD19 being bulky and expressed in lower density. We have a lot of experience from previous programs in targeting CD22. We've developed a very, very sensitive CD22 CAR T sensitive to very low antigen density. AUTO1/22 is a cotransduction. So there are cells -- so the obe-cel component of it is left as unchanged as possible. And we've tested this in a Phase I study in children with relapsed/refractory B-ALL. So if we go to Slide 56. So the main thing to note here is that Kymriah standard of care in the U.K. where these patients are treated. So these patients were Kymriah-ineligible at the point of recruitment to this study. And the other thing to point out is actually four patients had received Kymriah and three had CD19-negative disease coming into the study. So if we go on to Slide 57, manufacturing led to, I think, Sara Ghorashian presented this, led to a nice product profile with balanced CD19, CD22 in double CAR positive T cells. And we saw a nice expansion and engraftment in patients. Whenever you take a new CAR to clinic, there's always questions about what its safety profile would be. And actually, we didn't, I think, have a penalty -- a safety penalty here for co-targeting CD22 in these patients. There was one patient who did have a severe neurotoxicity, but we went and extensively investigated this patient, including a brain biopsy. And our conclusion is that this was most likely due to fludarabine, although obviously a CAR T-related neurotoxicity can't exclude it. But apart from that one patient, the AUTO1/22 is really very well tolerated. And if we go on to Slide 58, you can get a sense of where these patients are. So nine patients were in molecular remission at day 28. And we have had two patients who have had a frank relapse. Neither of these two patients had any antigen loss, and we've had one patient that became MRD positive at 12 months. The other thing to point out is that three of the patients who were CD19-negative going into the study went into a molecular remission. So if we go on to Slide 59. So where are we with this study? We show that we can make the products, the products expand and persist and they have a favorable safety profile. We can see that the 22 CAR is working for those patients who've gone into remission with CD19-negative disease. And given this patient cohort, we're pleased with the response rate and the ongoing remission rate. And to date, albeit with limited follow-up, we have not observed any antigen-negative relapse. And I think with that, I can hand back over to Christian to close up. Thanks for your attention.

Thanks, Martin. Obviously, when we look at where we are with the company, obviously, there's an enormous focus on delivering the pivotal study for obe-cel, the FELIX study. And this is clearly where the primary focus, obviously, of our activities are. But obviously, what was exciting about EHA for us this year is to provide not only an update on the continued progression of obe-cel in additional exploration and additional indications, but also we start to see very interesting profiles for AUTO1/22 in pediatric ALL and in particular for AUTO4 in T cell lymphoma, which is obviously a new set of targets that we're going after there as well as, obviously, a completely new clinical experience that we have with the approach that we're taking. We're also in the process of transitioning again in our collaboration that we're running with UCL. And you've seen, obviously, some of the output of their collaboration in today's call. The AUTO8 program in multiple myeloma, but also the AUTO6NG program later this year in neuroblastoma. With that, I would, first of all, would like to thank all the presenters for excellent presentations. Thank you very much for taking the time today. And at this point, actually, I'd like to open up the presentation for Q&A.

[Operator Instructions] Our first question comes from Matt Phipps with William Blair.

Nice updates at EHA over the weekend. I'm going to ask some AUTO4 questions. What do you think is the reason why you're not seeing or detecting the CAR T cells in the periphery and only find them in the lymph nodes? And is that something that you think is just from these small patient numbers? Or is that just how this therapy is going to work, I guess?

Yes, it's Martin. So I mean, I can try and answer that. So I think this is a new target, new set of diseases. So we don't really know what to expect. I guess from the decel experience, you would expect some peripheral expansion and persistence, but I guess this is something new for us, so we don't really know. I mean, I think the one thing that will be interesting to see is when we treat patients with CAR T cells that have been made with an improved and shortened process because those cells may have a higher capacity to expand and persist. So that will be an interesting next step with this.

Okay. Have you been able to look at just the presence of TRBC1/2 cells in the periphery to see if you're getting kind of aplasia of that specific compartment?

We have looked and at this stage, in the peripheral blood at least, we haven't seen a selective aplasia.

Okay. Interesting. So I guess, what -- you mentioned some improved process. I guess, is it just shortening the process, again, trying to get more of a naive cell type? And then also, what is the time line, do you think, getting the ICE or other diagnosis as it does seem like the NGS screening is taking a while and not really making enrollment as easy as it should be?

Yes. I mean I think in terms of the NGS I think -- in terms of the companion diagnostic, I think we've shown that it's possible to do by those three methodologies. And I think that's something that -- which is obviously a really big for a step because it's quite tricky. But I don't -- can't give you any particular time line when we can formalize those companion diagnostics.

Okay. Then one last question. Time line for getting AUTO5 into the clinic?

So with regards to AUTO5 -- yes, thank you. So first of all, you had a question related to the process. So one of the things that we're doing, obviously, is we're taking a lot of the experience that we can on obe-cel and we're applying that also to the AUTO4 process. And the goal, as you pointed out, is to shorten the manufacturing process and also with that, create a product that actually has a less differentiated phenotype. So that's clear one of the things we're going into and which we'll explore during the later part of this year into early next year. With regards to the time line around the AUTO5 program, obviously, the focus at this point has been to get the AUTO4 proof-of-concept done, which I think we're starting to get a good feel for. Obviously, we'd like to see more follow-up on the patients as well. And we're obviously in the process to run it through the IND-enabling studies in AUTO5. The goal is that we should be in a position during the course of next year to move also with AUTO5 into the clinical development as well.

Is Dr. Horwitz still on the line? Can I...

Yes, I'm here.

Just one question for you, if I may. Would you want to move these patients to an allogeneic transplant that are in CR? Do you worry about missing your chance on, I guess, a transplant?

You're saying if they receive investigational therapy like this?

Yes, yes.

Yes. I mean our main problem is not getting CRs. So there would be few patients who I think that would be a high risk. In some of the CD5 CAR T cell studies that are, again, in early phases because of concerns about long-term T cell aplasia, they do have allo backups in some of those patients, and there's just a few treated have gone to allo after CR or just been observed. So my sense is with a treatment like this, at the beginning, if you're seeing CRs, some of these patients -- and Martin can comment on how the study is being run, were not part of the study. Some of these patients could be observed and you're looking for durable responses. There might be other individual patients who you capitalize on the CR and take them to allo at that point if that was part of the plan. But allo has a 10% or so treatment-related mortality. So there would be a lot of ways to do better.

Our next question comes from Mara Goldstein with Mizuho.

I just have a question on the AUTO4 results. And the patients that had that long-term PR with an ongoing response, can you provide us with some -- just some background and details on that patient?

Martin, can you take that?

Yes, I can take that. This is a patient that really never went into remission. And I think actually, interestingly, the patient then relapsed quite a long time later with skin relapse, which is quite interesting. And yes, and I think it will be very interesting for us to present what happened to this patient next. I think the patient is actually -- still remains quite well. I don't know what else I can really say about them, they had refractory PTCL. They went into a metabolic remission state and relatively recently relapsed with an isolated skin relapse.

Okay. And then on the AUTO1/22 program. For the patients who had prior Kymriah, are you able to isolate out their results relative to the rest of the cohort?

Yes. So there are four of them. One of them didn't respond at all. So that was a patient who didn't -- Kymriah didn't [ last ] and AUTO1/22 didn't. There were two patients who were CD19-negative and went into remission. And those patients stayed in remission. There was one patient who's very interesting actually, who was 19-negative-and didn't respond. And actually bone marrow at day 28 showed double negative disease, which was interesting. So those are those four patients.

Our next question comes from Asthika Goonewardene with Truist.

So just like to dig back into AUTO4 a little bit more. Martin, you mentioned about your change of product. You're going to have maybe a shorter manufacturing to promote a more nice phenotype. Can you talk to us a little bit about what else you're doing to change your product? Is it just the short expansion? Are there any cytokines being used? And I guess the related question I have here is, if you change the manufacturing, how does this not require a new IND? And then I got a quick follow-up after that.

Yes. It's Martin...

Please go ahead, Martin.

Yes. So it's Martin here. So there's two things you can do to increase the stemness and the naivety of your product. You can shorten it, obviously. And then the other thing you can do is you can very simply increase the number of T cells going into the production because that means you can make the dose for a shorter period of time. So those are the two maneuvers. They sound really, really simple, but actually, the two things that really make a big difference. In terms of do you need a new IND? No, and I think it's very, very common in CAR T cell studies that you file an amendment, major amendments with a modified manufacturing process and you can keep on going. I'm almost embarrassed to say that nearly all of my Phase I studies I've ever done, we've changed manufacturing at some point during the study. So in my hands, at least it's not uncommon at all, and I think it's relatively common. Obviously, if you -- a pivotal study, you don't have that flexibility, but for a Phase I study, you do.

Got it. Okay. So then, can you just give us a little bit more clarity on the time lines on when you plan on getting this new AUTO4 into the clinic? And how many patients would you need to treat in Phase I before you can make go or no-go decision to pivotal with this new AUTO4 product?

I think Edgar or Chris, you might want to take that.

I can take that. So the plan is obviously to implement these changes, which are, frankly, being done now. And we expect to treat in the range of about six patients to get a feel for obviously, the impact of the modifications that we've done, and that will get us into the early part of next year plus follow-up in these patients. So it gets us into that first half of next year in terms of actually having data for those patients and an understanding of the performance of the changes.

Great. And then a real quick one on AUTO1. We've obviously seen some data from Bristol and also coming from the Fred Hutch about how combination with the CD19 CAR T plus IMBRUVICA has a synergistic effect there in terms of CAR T cell fitness and long-term durability of response. Just trying to think what's -- in light of that, what's your next steps for taking AUTO1 into something like CLL and would maybe want to try and look at doing a combo study as well.

I think at this point...

Yes. It's Martin -- sorry, Christian, go on. Go on.

Yes. I think at this point, what we're planning to do is really kind of as Martin pointed out, complete the current cohort and get a proper understanding of the activity of the product on its own. I think the analysis with ibrutinib is interesting. And obviously, it's a common background that we have these patients being treated with. And so it's certainly something we'll be looking at. But I think at this point, what we really want to understand is the basic activity of obe-cel on its own. And I think once we know that information, I think at that point, we'll make decisions around whether to take the program forward as it is or whether an additional use of ibrutinib might actually be beneficial.

Got it, guys. And then just a last quick question, and I'll hop back in the queue. What can we expect at ASH in terms of updates from any of these three programs that you presented here at EHA?

Right. So in terms of our outlook for ASH, obviously, we're going to have about 6 months additional data in terms of follow-up, as I think that's going to be very meaningful particularly on the old CAR cohort, particular cohort where we've already seen data getting us somewhere between 12 and 18 months of follow-up on that patient group. I think we want to see that the cohorts filled out. And obviously, also additional follow-up in particular on the DLBCL and [ mantle cell ] patients. And I think that will start to be a very informative set of information. AUTO1/22, obviously, there is a few more patients that can be involved. So I think there's a few more patients that were going to be active at study. But I think what's going to be very informative is the performance of the product in the patients that are currently treated, which would give us all the patients beyond kind of the 6 months follow-up. And with that, I think, a pretty clear view on whether or not we have the impact that we desired on the -- with regards to antigen loss and frankly eliminating an antigen loss in that -- in potentially relapsing patients. With that, we should have enough follow-up to actually be clear that indeed, we've kind of gotten through that. And then finally, on AUTO4, as we indicated, obviously, more follow-up on the current patient set, and then obviously, the additional patients that are planning to dose on -- with the modified manufacturing process. So those are kind of the key updates for the end of the year on those programs.

Our next question comes from Kelly Shi with Jefferies.

My first question is, is the expression level of TRBC1 and 2 variable from patient to patient? Do you see a treatment outcome, I mean the impact of the treatment outcome due to the different expression level? And also the second question is what is the manufacturing success rate from the first -- the 10 patient data for AUTO4 program?

Martin?

So TRBC1 and 2 expression on most T cell lymphomas is really quite high and quite consistent. I think if you look at Maciocia, et al., our original paper in Nature Medicine, I think that's got a nice summary of the expression density. So we'll expect exception is HTLV-driven T cell lymphoma ATLL, where expression tend a little bit lower. So generally, expression tends to be higher than the normal T cells. In terms of this study, we didn't see any correlation between expression level going into the study and outcome. So that's -- that bit of your question. I think in terms of manufacturing, I think we had a high manufacturing success rate, which I think was 100% actually for this study.

Our next question comes from Chang Liu with Needham & Company.

This is Chang for Gil. So we just have a question on AUTO4. It seems the lymphocytes recovery is fairly quick in the patients after AUTO4 treatment. So did you have a chance to say if the T cells recovered are TRBC1-positive or TRBC2 positive?

Okay, Martin, you want to take that?

So we looked and we have not seen any skewing of TRBC1 and 2 in recovery. So we did not see any selective depletion of one compartment or the other in the peripheral blood.

[Operator Instructions] Our next question comes from Eric Joseph with JPMorgan.

Just a few on AUTO4 from us. The first is whether there's any clustering of responses by PTCL subtype, whether there's any concentration of responses among those that were either not otherwise specified or the [ angioblastic suicide ]? And then secondly, I just wanted to follow up on the question from the point related to detection of CAR T in the peripheral blood. I guess, to what extent is seeing -- is peripheral blood detection going to inform or help define the [ recommended ] Phase II dose? And then I have a follow-up here.

Yes. It's Martin. I can try and answer those. I think the answer to your first question is no. I don't think we can say there's a particular histological subtype. I mean, and again, because the numbers are so few. Even if there was, I don't think it would be meaningful at this stage. So I think it's too early to tell. In terms of peripheral blood and making peripheral blood marking and making decisions, because it's a new target and a new disease setting, we don't really know what to expect. I think we'll be primarily guided by what we see in patients in terms of toxicity responses.

Okay. In the abstract, there's -- you comment a little bit on the screen positivity rate. And I guess the implied screen [ failure ] rate is about 30%. I'm just trying to get a better sense of what resulted in, I guess, patients screening out or whether that might be reflective of the addressable population, I guess, in the real clinical setting here? And then as it relates to sort of the adaptive manufacturing process going forward, is there a potential -- or do you see the potential, I guess, with these optimizations to allow patient redosing? And if redosing is something that you might look to explore perhaps in the patient -- in situations where a patient goes -- undergoes partial responses to their initial response, I guess, to the protocol as currently designed for redosing?

Yes. So I can try to -- so your first question, screening. I think obviously, this study is designed in two parts. So there's a screening part and a treatment part. And I think one of the problems we encountered with the screening is that we were right in the middle of COVID. So things were getting slow and then patients would progress. I think the TRBC1 to 2 ratio is 40% to 60%. So that's kind of what you expected with one concentration versus another. So I guess we weren't far off. And then I think in terms of manufacturing and redosing, there has not been very much luck with CAR T redosing, generally speaking, particularly in autologous setting. I mean, it's certainly something we could think about trying but it's not something that's really on my mind at this point in time, generally speaking, because it hasn't been a strategy that's worked particularly well. But never say never.

[Operator Instructions] There are no further questions. I'd like to turn the call back over to Christian Itin for any closing remarks.

Well, thank you very much, everybody, for joining, and particularly thanks to Dr. Horwitz and Dr. Cwynarski for the presentations, which were fantastic. And obviously, we're looking forward to keeping you updated on progress of the programs and wish you a good week. Thanks, everybody.

This concludes the program. You may now disconnect. Everyone, have a great day.