Allogene Therapeutics, Inc. (ALLO) Earnings Call Transcript & Summary

Good morning. I'm Salveen Richter, biotechnology analyst at Goldman Sachs. We're pleased to have Allogene with us here this morning. And with us, we have David Chang, President and CEO.

David, you recently presented initial data from the Phase I ALLO-501 study, your allogeneic CAR T at ASCO this year. Could you start by briefly recapping these findings and help put into context for us, how we should think about these results compared to the early autologous CAR T data?

Yes, Salveen, it's a pleasure to be invited to this conference. And too bad that we have to hold this as a virtual conference. I mean, I always love to go down to the Goldman Sachs conference at a wonderful time in June in Southern California, but such is the life nowadays. So we have pretty exciting and busy second quarter, sort of cleaning up the ALLO-501 data [ we to ] presented at ASCO. And then after that, there were many other activities that kept us quite busy. So the first data set of our allogeneic CAR T that we presented was a CD19 CAR program, ALLO-501, which was a Phase I data that -- where we explored the safety, different cell doses ranging from 40 to 320 -- 60, excuse me, as well as try a different lymphodepletion regimen that contained either low or high dose of ALLO-647, which is our CD52 monoclonal antibody. So in the study, we enrolled both follicular and DLBCL patients on a first-come basis in a 3 plus 3 design, and we completed a dose escalation using the low dose lymphodepletion at ALLO-647 of 39-milligram dose. And we also tested a number of patients at higher 90-milligram ALLO-647, which is still ongoing. So initial data set included heavily pretreated patients, median prior therapy number was 4 and the patient population was representing about 2/3 DLBCL and 1/3 follicular lymphoma. So in the data set that included altogether 22 patients, of whom 19 of them were available for the efficacy evaluation, defined as at least having had the first tumor response assessment, which occurs at month 1, 28 days after the cells are given. So a couple of interesting findings from the study. I mean, obviously, as a Phase I, safety is the utmost importance. And given this is an allogeneic, one of the focus of the safety findings were on the graft versus host. In 22 patients, we didn't have a single case of graft versus host, which gives -- at least reassures us that editing out the T cell receptor is a great means to minimize the risk of GvHD. Obviously, we will have to continue to treat more patients, but the initial sign is that GvHD will not be much of an issue. The second one is related to -- some of the cytokine related syndromes as well as neurotoxicity. We are quite pleased to see that overall incidence of cytokine release syndrome that's Grade 3 or higher, there's only 1 case. And then in terms of neurotoxicity, altogether, we didn't see any -- we didn't have any reported cases on neurotoxicity. So in terms of safety, our assessment is that this is numerically on par or slightly better than what has been reported for the autologous CD19 CAR T. Now when it comes to the efficacy, which is what everybody was looking into. In the overall study, of the 19 evaluable patients, our numbers was a response rate of -- I don't want to misput any numbers, 63% including 7 patients who had a complete response, which gives response -- complete response rate of 37%. Now that's overall, and we look into how the 39-milligram versus 90-milligram. 90-milligram lymphodepletion with 647 gave a deeper tumor response, more patients having the complete remission. And also, another thing that we did was looking into the patients who are CAR T naive. So we had some patients who are previously treated and were refractory to autologous CAR T that we enrolled in the study. So looking into how we're going to move forward and trying to make a comparison, we did a data cut based on CAR T naive patients. So in the CAR T naive patient population, the response rate was higher. And in fact, in the patient population that we believe is more relevant as we think about the Phase II, higher lymphodepletion who are CAR T naive, the numbers are small, 6 patients. There we had 5 patients responding, including 4 patients who had a complete remission. So that's the data set. At the time of the data cut, 75% of the patients who responded were continuing to be in response. I would say this is an area of great interest that requires additional time follow-up, which we are doing right now.

Great. And you talked about GvHD, but as you think about some of the hurdles that you faced going to an allogeneic approach from autologous. Where do you stand today? What answers do you have here?

So 2 answers, for allogeneic. I mean, we always view that 2 most important things that we have to control was really the immunological process that results in graft-versus-host disease. So this is AlloCAR T cells reacting to the normal tissues of the patients -- in the patient's body. At this point, I think this is becoming lesser and lesser of a concern. We are not seeing much of the GvHD -- GvH after editing up the T cell receptor locus. So I think this is really where the science plays itself out because many years of transplant biology research have shown that T cells mediate GvH and the T cell receptor is the protein that is responsible, and we edit out. So scientifically as well as clinical data that we are seeing are very consistent with all the existing data, which we are quite happy with. And the second one is the reversal. Patients body -- patients' immune system rejecting the AlloCAR T cells before they have antitumor effect. We believe that this is probably where the key difference in the allogeneic therapy will play out. How to control that process? There are many different sort of approaches, and companies are taking different approaches. But our approach is based on creating a selective as well as more controllable lymphodepletion using antibody ALLO-647. So ALLO-647 is anti-CD52. CD52 is present in the lymphocytes and antibody mediates depletion of the lymphocytes. In order to do that, we have to protect our own AlloCAR T cells. Otherwise, anybody will also deplete our own AlloCAR T cells. So we accomplished this by editing our CD52 in the setting of using ALLO-647. We think this is a quite differentiated and quite unique approach, and our data supports that using ALLO-647 together with editing out CD52 allows the cell expansion and the initial tumor response as we have just talked about.

Great. And how should we think about the response rate as a function of histology here? Did you see any differences between the DLBCL and the follicular lymphoma patients? And what is the mechanism by which there could be a differentiated response here?

Yes. So the second question, I think, easier question. I don't think there is any -- the mechanism that will lead to differences in the response rate. Other than what has been reported in the clinical study at ASCO, the ZUMA-5 study, this is the YESCARTA study in follicular lymphoma was presented. And people are responding to that release of data because the response rate compared to what has been reported in the large cell lymphoma, or DLBCL numerically was higher. But there are many nuancing to this. In our case, we had limited number of follicular lymphoma patients, 8 patients out of 22. And when we look at the response rate as well as complete remission rate as well and the people who are still in response, we did not see any major differences. Obviously, these are small numbers, which people can be misled because when the numbers are in the single-digit, 1 patient can make 10% to 15% difference in the response rate. So for that reason, we aggregated data without breaking the data set into follicular lymphoma. But as I said, CAR T works both on the DLBCL patients as well as follicular lymphoma patients equally well. And I think there's a lot more for us to learn about exactly what the differences between those 2 tumor types are.

And how should we think about the lack of response that was seen in the refractory patients to autologous CAR T. Is this just a matter of optimization of the preconditioning regimen or cell dose? And then you have spoken previously about how these reasons also hold true to patients -- to why a patient may be refractory to autologous CAR T. So can you just review these and whether those reasons hold here in allogeneic?

So the ALPHA study, we amended protocol earlier this year to allow prior CD directed therapy. Usually, when you're going after 1 target, you don't want patients who had treatment going after the same target. That was the initial approach. But as we develop more confidence in what the therapy is doing, we allow the prior CD19 treated patients. So we have 4 patients who are autologous CAR T refractory. So we went after sort of some aggressive way, in the patients who did not respond at all or patient had very short duration of response, less than 3 month to the autologous CAR T therapy. So what we really wanted to test there is these patients who do not respond so-called refractory is the problem with the cells that you manufacture in the autologous setting. It's an individualized manufacturing. So every cell product is different. Is that the underlying reason for these patients not responding or is it something intrinsic to the tumor itself? So being able to do the study in an allogeneic setting, where we don't have concerns about the consistency of the cell potency because we're using the same cells in different patients. We felt that this was a perfect setting to test that. And we hope what we are seeing, this is early on. 3 out of 3 patients were evaluated for response, and none of them responded to the -- our ALLO-501. So I think this is something that we have learned and the field has learned. And certainly, what is to be followed is some translational research, trying to understand the differences in the tumor that may prevent the CAR T therapy to work. Interestingly, as we are doing this, there was a paper that was published that implicated that if you have a defect in intrinsic death pathway, so this is how T cells or other cells are -- triggered to undergo apoptosis. So if you have a defect in the extrinsic pathway, those tumors are, in a preclinical models, resistant to the CAR T therapy. So here, as we are doing the clinical study, we are also finding on some corroborating preclinical data. And this is the lead that we're going to go after as we investigate the tumors genetically to see what accounted for the lack of responses that we have seen, both in the autologous CAR T setting as well as in the subsequent ALLO-501 treatment.

And then one of the interesting aspects revealed here was the potential to redose patients and achieve a deeper response. Can you just speak to this and whether all patients who lose a response assuming CD19 positive would be eligible to be retreated?

So in a way, we have kept that option available for patients in the Phase I study. So any time when the patient progresses, we offer the retreatment. So in that setting, we had an opportunity to treat one patient where we have an efficacy data available. So this is a patient initially treated with a low dose lymphodepletion, had a partial response that progressed at month 2. At month 5, we made a decision to retreat the patient -- well, the investigator made a decision to retreat the patient. So we re-lymphodepleted using higher dose ALLO-647. So it was a lymphodepletion with a 90 milligrams of 647 and treated the patient with the same lot of the cells at the same cell dose as the first time. Fortunately, with the second treatment, patient's initial response was, again, partial response at month 1 that evolved into a complete remission by month 2. So we are very interested in this early data. Certainly, it tells that retreatment can work. So this is what I would consider as a reactive retreatment when the patient need rises. Another concept that is being entertained, and this is both investigator's interest as well as us wanting to know how to continue to improve the treatment effect is whether you should just upfront redose the patient giving more cells in 2 separate doses. That's an early concept, and we are sort of reviewing the available data. But the reason that this becomes somewhat attractive is 2 things. One, as an allogeneic, redosing becomes relatively an easy issue. And second, we have the benefit of ALLO-647, which allows us to re-lymphodeplete without using the chemotherapy, simply just giving ALLO-647. So many things just factors in a very favorable way for us to redose. What we're going to do is, we'll just follow-up the patients who are already treated, how durable their responses are. And based on that, we'll make some decision whether we test the redosing formally or we just entertain as a reactive treatment as we advance the program.

Great. And in terms of longer-term follow-up of these patients that have been treated, when should we expect to see this data? And how long of a follow-up do you think is necessary to meaningly -- or to kind of inform what -- meaningfully, how to think about durability here?

Yes. So especially the patients who are treated with the high dose ALLO-647, this is the cohort that we are very interested in, and they were all treated early this year and continue to be treated. In terms of the length of follow-up, I think, generally, about 3 month follow-up gives you a pretty good sense. If someone is still in response at month 3, their probability to continue to have response in subsequent follow-up, month 6 or 9 months after, is very high. So in additional follow-up, we believe that after patient is treated, about 6 to 7 months gives a pretty good sense. From the recent ASCO presentation, we want to see additional about 4 to 6 more month of follow-up before we present the data. So whether that will be ASH or sometime early in 2021, we haven't decided what the right timing is.

And then maybe just a final question with regard to this program specifically, but you have a bunch of levers that you can play with to optimize the program. Is there anything else that you're looking to do right now?

The initial data that we saw, I mean, it was positive in many ways. I mean, the safety was -- we are extremely pleased with the safety. Initial responses were very robust and very comparable to what has been reported with autologous CAR T plus the fact that we were treating essentially everybody who was enrolled. And we also saw that as we go up dose of 647, we got much better lymphodepletion between 39 and 90-milligram in terms of the lymphodepletion. And the duration of lymphodepletion, we were able to control it much better. So we have, we believe, all the right components. Now the question is, what other things can we sort of do to further optimize? And before that, our initial thought doing it to optimize anymore, if the follow-up of the patient shows the durable response, I think basically, we have all the right components in terms of what to do with the lymphodepletion and the cell dose. We have all the necessary information to move to the Phase II. If there is some progression and if you want to continue to push the boundary, it comes down to the sequence of how we optimize and whether we consider the redosing upfront. I think those are 2 different levers that we are entertaining. And obviously, there will be a data-driven decision that we will make later this year.

Great. And then your second-generation ALLO-501 is devoid of the rituximab off switch to target a broader subset of patients here regardless of the prior exposure to rituximab. Can you start by reviewing the manufacturing change that was necessary between this first and second-generation constructs? And then remind us about the trial design, how it will be accelerated versus ALPHA and then how you bring that into the pivotal process here?

Okay. So in terms of the 501A, precisely, as you have said, it took out the Rituxan switch, which was a part that was built into the initial construct of 501. Keep in mind the 501 construct was initially generated many years ago when there was a very -- a hotly debated question about whether you need a safety switch for any chimeric antigen receptor that you're introducing at the clinics. 4, 5 years later now, everybody feels very comfortable. You don't need such suicide switch. And in our case, the suicide switch was not only we felt it's not necessary, but it was little bit of obstacle because we had to wait for the patient from the last Rituxan infusion before we can put him onto the study. So early on, we made a decision that Rituxan recognition domain has to go up. So what that resulted is coming up with a construct that has the same amino acid sequence in terms of the chimeric antigen receptor, but no Rituxan switch. But the vector-wise, that's a different vector altogether, which is a critical element in cell manufacturing. So we had to consider this as a different product because that was considered -- that was a pretty significant change. Other than that, other than having to use a different viral vector, manufacturing process, I would say there are some incremental changes to make it more efficient, but fundamentally, nothing really changed, which is the reason why also, when we look at the preclinical data, I mean, in terms of the potency and everything else, it is almost overlapping to what 501 has shown in preclinical experiments. So moving forward, we are assuming that this will behave the same way as 501. But like anything else, we are taking a very cautious approach of making sure that the safety and efficacy has not -- did not shift as we take -- took out the switch. So that's the reason that we are doing ALPHA2 Phase I study, which will proceed as a simple dose escalation. The cell dose range from the 501 study was from 40 million cells to 360. We're going to do the dose escalation in a very abbreviated way. Treating fewer patients then 3 and then complete the dose escalation, at which point, we will have enough confidence to simply phase in 501A into the pivotal study. So that's the plan. So where we stand right now, wait for the follow-up on patients in the 501 study and complete the dose escalation in the 501A. And actually, the time line is very coordinated that, that sort of ends to where we can make a decision on the Phase II sometime in 2021.

And while early, maybe with respect to the commercial opportunity here, how do you view the potential for an anti-CD19 AlloCAR T given the data that you've seen thus far?

Well, we see a lot of opportunities. I think, one, I think the use of the CAR T, as the people understand how to administer the CAR T therapy and also better control the adverse events. As Kite is showing the label extension into different tumor types is occurring [Audio Gap] large cell lymphoma. They had a very exciting data in mantle cell lymphoma that was followed up with most recent data updates on the follicular lymphoma. So the kind of label extension is one. The second one is moving to earlier lines of therapy and moving out of the relapsed and refractory and moving to the second line in the case of DLBCL or earlier lines for the follicular and other lymphoma. I think -- so that's -- so basically, what I consider as an expanded use across different histology and the different lines of therapy. That's one. Where we are most interested in is whether the therapy can be made more amenable for outpatient treatment. So if we can do that, and certainly, the safety data that we have seen in ALPHA1 suggests that there's a high potential we can do this as an outpatient. Now when you can do that in addition to the label extension into different indications -- lines and different histology, being able to give it as an outpatient, I think that will greatly increase the usage. And then, of course, for allogeneic, we are doing this as an off-the-shelf on-demand. So as an outpatient, on-demand off-the-shelf, I believe that allogeneic cell therapy is really moving the CAR T therapy to the more standard way treatment delivery, while keeping the treatment still as a onetime therapy. So with that mix of different combinations, I think there is a great potential for the allogenic CAR T therapy to be more widely used.

Got it. And so moving to your second program, ALLO-715, you're looking at a Phase I trial here in relapsed/refractory multiple myeloma. Can you walk us through the trial design and what learnings from 501 are being applied here?

Yes. So this is ALLO-715 BCMA CAR T program, which we initiated study in the -- about -- large being about 7 months ago last fall. So the study is enrolling patients with relapsed and refractory multiple myeloma who have had treated with proteasome inhibitor, IMiDs as well as CD38-directed therapy. So these are really refractory patients. So the trial is proceeding as a cell dose exploration study, 3 plus 3 design with the primary objective of determining the safety as well as a secondary objective of looking at the early signs of efficacy. So we are completing the dose escalation. There are 3 planned doses. We can also add additional dose. But right now, we are completing the initial planned dose escalation from 40 to -- which was from 40 million cells to 320 million cells. So BCMA is also an area of great interest. I mean, the initial data set, we are planning to present the initial data in the second half of this year, most likely in the fourth quarter. Other things that we are doing with the BCMA in multiple myeloma indication is exploring other ways to increase the treatment effect. So 2 other things that we are doing. One, initiating a study investigating ALLO-715 with a gamma secretase inhibitor, nirogacestat, which is in collaboration with SpringWorks. We expect to start that study in the second half of this year. And the primary sort of hypothesis that we are testing is BCMA in multiple myeloma normally gets clipped off by gamma secretase that's present in the multiple myeloma cells. If we were to prevent that clipping, we can retain high levels of BCMA expression in the multiple myeloma. And with a higher target density, we expect there will be more activation of CAR T cells, and the deeper initial killing. And that's the underlying hypothesis. And both preclinical as well as early clinical data coming from autologous BCMA CAR T are well supporting that hypothesis. So we are very excited about moving the program in that direction. The third one that we are doing is really transformative concept of introducing TurboCAR into our BCMA. So this is something that we presented at the recent ASGCT meeting. And it goes back to what activates the T cells in the normal situation. It requires signal 1 coming from the T cell receptor and the signal 2 coming from the co-stimulatory molecule. Both signal 1 and signal 2 are built into the chimeric antigen receptor construct, and that's taken care of. What so far nobody has tested is the third crucial signaling activation signal that comes from the cytokines. So to address that issue, we built in signaling domain of cytokine receptors into our engineered CAR T cells. So that's the concept of TurboCAR, and this is our ALLO-605 program that we intend to introduce into the clinics in 2021.

And then, David, I apologize if you already talked about this a little bit, but could you just comment on the key difference in the preconditioning regimens between both programs? And secondly, how we think about whether there's variations in the respective tumors as we look at multiple myeloma versus NHL as to how receptive they are to CAR T therapy, particularly in the sense of durability?

So the first question in terms of lymphodepletion, differences between 501 and 715, 2 AlloCAR T programs. We start initial study with the same lymphodepletion. And in the 501, we went up to the high from 39 ALLO-647 dose to 90-milligram. And we are planning to test something similar in ALLO-715. Another thing that we are doing as we're increasing the ALLO-647 dose is trying to test whether lymphodepletion can be achieved using biologic agent without using chemotherapy. So this is something that is built into the 715 protocol and this will be tested in a stepwise fashion. Obviously, we want to achieve the optimal lymphodepletion that we want to get. And once we get to that, we'll start teasing off one by one, the 2 chemotherapy components, fludarabine and cyclophosphamide. The second question about the durability of response. I think here I definitely see, in multiple myeloma, what the autologous BCMA CAR T program is showing is a really deep response and progression-free survival that's nearing about 9 to 10 months -- probably, let's say, about a year, which is really a spectacular result given comparing with the existing therapy. The thing that's somewhat different between DLBCL and multiple myeloma is, in multiple myeloma, there appears to be late relapse in a patient who had MRD-negative response who seemed to relapse even after a year. What's going on in that setting? I think that's still under investigation. And also linking the previous discussion that we had about retreatment, the fact that allogeneic CAR T therapy makes the retreatment a lot easier, I think this really, for me, I think for us, it's creating opportunities for allogeneic CAR T, where in a patient, if they progress after 6, 9 months of response, retreatment becomes a much easier discussion in such setting with allogeneic cell therapy.

And then you have an IND for an anti-CD70 CAR T, which is on track to be filed by year-end, looking at both solid and liquid tumors. Could you provide a high level of overview as to why CD70 is an attractive target? And then your decision-making process as you look to your portfolio as to why this was the one you wanted to focus on initially?

So CD70, we were interested in this target for some time. One, when you look at the expression profile, it -- we believe the lack of CD70 expression in most normal tissues provides enough therapeutic index to apply technologies -- CAR T technologies. The second one is, CD70 in certain tumors, the expression is almost close to 100%, and the expression is at high level. And one great example is the renal cell cancer. We also know that CD70 is expressed in many different liquid tumors beginning with AML, but the list goes on to T cell lymphoma, leukemia, B-cell leukemia, lymphoma. So there's a lot of other hematologic malignancies where CD70 can play a role. One of the biggest challenge with the CD70 is that CD70 is also expressed in the T cells, which makes -- making chimeric antigen receptor against the CD70 somewhat challenging and that's where we spend a lot of time finding a construct that doesn't do that. Essentially, that doesn't cause so-called the fratricide, CAR T cells killing each other. So we have that construct, and we're very excited to advance it into the clinics. And here, because it is expressed in solid tumor, our interest is really testing the hypothesis and beginning to explore how to effectively look for the opportunity to improve the advancement of CAR Ts into the solid tumor. So that's one, and that will be the first IND study. And the second one, obviously, based on what I just said, we are very interested in how CD70 could play a role in AML treatment, which is another -- I would say, still an indication that has not been fully addressed with a CAR T approach and also has a high unmet medical need. So that's more or less how our thinking process about how to sequence 2 different IND studies and also using CD70 as a way to start investigating the use of CAR T in solid tumor. And as it happened, renal cell cancer is a very interesting cancer, not just because of the CD70 expression levels, but also this is immunologically active tumors. We certainly know that a lot of checkpoint inhibitors have shown efficacy. So I think the tumor microenvironment of renal cell cancer may be more amenable to the CAR T therapy. So in many ways, I mean, those are the sort of thinking processes. But in terms of portfolio management, obviously, we are making big bets in both CD19 and BCMA. We will continue to do so, and we will continue to look for the next-generation technology that will improve from what we are doing. So that's in the one category. And the second one is really growing the use of the CAR T therapy in settings other than hematologic malignancies. So CD70 is the first one, and we also have plans for DLL3, which is our next product candidate. And we are doing so-called IND-enabling work to allow -- that will allow us to introduce DLL3 targeting CAR into the clinics.

Great. And David, a last question here. So you have been amassing novel technologies and also doing collaborations. Could you just touch on the collaboration with Notch to develop iPSC-derived therapies for blood cancers here and the impetus behind this move and when we might see something come out of that collaboration?

So it is an early collaboration. I mean, what we are looking for in the iPSC area are 2 things: one, ability to engineer the T cells at a clonal level. In the iPSCs, you can do one gene editing, one gene engineering, isolate the clone that has that exact change that you want and then do more additional gene engineering. So essentially, you are doing the gene engineering in a clonal way, which gives you 100% gene editing and gene engineering. That's a big plus compared to what we're doing. So that is one. And much of the work we are doing in-house to make sure that we choose the right gene engineering and gene editing. The second one is how to differentiate iPSCs into functioning potent lymphocyte. And here, the T cells historically is what has given the high level of response that we have seen in the cell therapy. And we are interested in technologies that will allow the differentiation of iPSC into functioning T cells, which is where the Notch Therapeutics is coming in with their proprietary technology that is fee-based that allows the differentiation of iPSC into the T cells. So that's the underlying rationale for this partnership. And this is early on. And why we're doing this? I mean, certainly, iPSC, they can potentially provide another transformative changes in the cell therapy, being able to engineer CAR T cells from a renewable cell source in such a complete way in terms of gene editing and gene engineering. This will take some time. I think this is such transformative changes. We are approaching it very carefully. We know that others are already in this space. But we are approaching this as sort of next-generation technology going beyond what we are doing with the healthy donors. So in terms of when the product can go into the IND, I think it's a little bit too early to sort of set any kind of time line. We just want to do this right.

Right. Well, with that, thank you so much David.

Salveen, it's great to see you and have a great day, and thanks for hosting us at this Goldman Healthcare Conference.

Thank you.

Okay.