Cellectis SA (CLLS) 2017 Q3 法說會逐字稿

Greetings, and welcome to Cellectis Third Quarter 2017 Financial Results.

(Operator Instructions) I would now like to turn the conference over to your host, Simon Harnest, Vice President of Corporate Strategy and Finance. Thank you. You may begin.

Thank you very much. And welcome, everyone, to Cellectis' Third Quarter 2017 Financial Results Conference Call. Joining me on the call today with prepared remarks are André Choulika, our Chairman and Chief Executive Officer; Dr. Mathieu Simon, our Chief Medical Officer; and Eric Dutang, our Chief Financial Officer. Yesterday evening, Cellectis issued a press release reporting our financial results for the third quarter ended September 30, 2017. This press release is available on our website at www.cellectis.com.

As a reminder, we will make forward-looking statements regarding financial outlook, in addition to regulatory and product development plans. These statements are subject to risks and uncertainties that may cause actual results to differ from those forecasted. A description of these risks can be found in our most recent Form 20-F on file with the SEC.

I would now like to turn the call over to André.

Well, thank you, Simon. And good morning, everyone. So the third quarter of 2017 has been one of the most remarkable ones in our company's history. In June, we initiated Phase I clinical trial for UCART123, our first wholly-owned allogenic CAR-T cell program in AML and BPDCN patients. In July, we successfully spun out and conducted an IPO for our plant subsidiary -- science subsidiary Calyxt, creating a tremendous potential of new value creation for shareholders. And most recently, our partner Servier has released the first set of interim clinical Phase I data for UCART19, the first-ever gene-edited allogeneic CAR-T program in pediatric and adult ALL patients.

Dr. Simon will go into more details about our clinical program UCART123 and will discuss the ASH abstract by Servier on UCART19 before handing the call over to our CFO, Eric Dutang, who will give an update on our cash position and other financials.

With that, I will pass the call over to Dr. Mathieu Simon. Mathieu?

Thank you, André. I would like to start with discussing UCART123, our first wholly-owned CAR-T development program to reach patients in clinical trials. This is already the second TALEN gene-edited CAR-T program after UCART19 that is now applied in the clinical setting. UCART123 is a gene-edited healthy-donor derived CAR-T cells that targets CD123, an antigen expressed at the surface of leukemic stem cells in AML as well as on leukemic and other tumoral cells such as in BPDCN, for your information, blastic plasmacytoid dendritic cell neoplasm.

AML is a devastating clonal hematopoietic stem cell neoplasm that is characterized by uncontrolled proliferation and accumulation of leukemic blasts in bone marrow, peripheral blood and occasionally, in other tissues. These cells disrupt normal hematopoiesis and rapidly cause bone marrow failure and death. In the U.S., there is an estimated 20,000 AML cases per year, with other 10,000 estimated deaths per year.

In July, we initiated 2 separate Phase I clinical trials, one in AML patients at Weill Cornell Medical Center in New York, led by Professor Gail Roboz; and one in BPDCN patients at MD Anderson Cancer Center, led by Dr. Naveen Pemmaraju.

Following the death of a patient in the BPDCN trial in September, the FDA placed both UCART123 trials on clinical hold. We conducted a thorough investigation into this patient death and communicated all findings to the FDA. In response to our analysis, the FDA has recently lifted the clinical hold on both studies. Cellectis agreed with the FDA to the following main revision to be implemented in the Phase I UCART123 protocol: First, a decrease of the cohort dose level to 6.25 x 10 to the 4th UCART123 cells per kilo; a decrease of the cyclophosphamide dose of the lympho-depleting regimen to 750-milligram per square meter per day over 3 days, with a maximum daily dose of 1.33 grams of cyclophosphamide.

Second, inclusion of specific criteria at day 0, the day of the UCART123 fusion, such as no new uncontrolled infection after receipt of lymphodepletion, afebrile status, of all that replacement dose of corticosteroids, no organ dysfunction since eligibility screening. Third, the provision to ensure that the next 3 patients to be treated in each protocol will be under the age of 65. Fourth, a provision to ensure that the enrollment will be staggered across the UCART123 Phase I protocols, with at least 28 days between the enrollments of 2 patients across the 2 studies.

What we have learned over the past few weeks about UCART123 has strongly encouraged us to move forward with this (inaudible) therapy and clinical trial. Once patient enrollment will resume, we will have a better idea about time lines, and we are planning to share an interim clinical update on UCART123 in 2018.

The UCART allogeneic CAR-T platform is continuing to advance, showing great promise in the clinical setting in patients with clear unmet medical need. On November 1, Servier announced the presentation of 2 ASH abstracts for UCART19, describing 2 initial dose cohorts in a Phase I clinical trial in adult and pediatric ALL patients. The adult Phase I trial, or CALM, is currently ongoing at King's College London, led by Dr. Reuben Benjamin. The pediatric Phase I trial, or PALL, is currently ongoing at University College London in the U.K., led by Professor Waseem Qasim. The primary objective of the adult study is to assess the correct dose of UCART19 by investigating up to 4 dose levels in separate sequential cohorts, starting with the lowest dose of 1 x 10 to the 5th cells per kilogram of body weight.

Only patients with relapsed/refractory CD19-positive ALL, who have exhausted all available treatment options, were eligible. This does not mean that UCART19 is only eligible to this patient population, but it is the higher hurdle to tackle before extending to an easier-to-treat patient population. This is the first set of development of this allogeneic CAR-T therapy showing huge potential. Patients are closely monitored for safety and anti-leukemic activity until the end of the study 3 months after UCART19 administration.

In the pediatric study, only patients with high-risk relapsed/refractory CD19-positive B-ALL and no therapeutic options were enrolled. In addition, all enrolled patients were unsuitable or had failed generation of an autologous CAR-T cell product. Some of the patients had undergone and failed a previous allo stem cell transplant. A fixed dose of about 1 to 2 x 10 to the 6th cells per kilogram was injected.

Here again, this doesn't mean that this therapy is eligible only to patients that are not suitable for an autologous CAR-T cell therapy, but UCART19 has the higher hurdle to take.

Following a conservative trial protocol, if the patient achieves molecular remission, an allo stem cell transplant is scheduled within 6 to 12 weeks after CAR-T infusion and the trial period continues for a further 12 months before entering into a long-term follow-up study. Again, it doesn't mean that this therapy is a bridge for transplant only. In this early Phase I study, we are only paving the way to a standalone application.

Impressively, the ASH abstract show that 5 out of 5 children and 4 out of 6 adults and 100% and 60% of patients, respectively, were cancer-free 28 days after UCART19 injections.

In the long-term follow-up, 60% of the adult population and 60% of the pediatric patients, including the 2 compassionate use cases from 2015, are still in complete remission. Again, these encouraging results are already seen at the very first dose level.

Here, I would like to point out one very important point: 1 patient who relapsed with CD19-positive disease 2 months past initial UCART19 infusion, was reinjected with fludarabine, cyclophosphamide lymphodepletion, no alemtuzumab, and the same dose of UCART19, achieving MRD-negative status again. This is the first successful proof-of-concept for re-dosing of an allogeneic CAR-T injection. It also defeats the argument of the need for long-term persistence of CAR-Ts to achieve long-term complete remission that always come at a cost, such as lymphopenia, which is a long-term, heavy side effect in autologous CAR-T therapy.

From a safety perspective, UCART19 showed a relatively mild level of cytokine release syndrome, with 1 grade 1, 3 grade 2, 1 grade 3 in pediatrics; and 1 grade 1, 4 grade 2 in adult patients, with 1 outlier at a grade 4 CRS. Important to note that this patient had an extremely high tumor burden of 100% tumor in bone marrow at the time of enrollment and was completely refractory to any other treatment. Signs of GvHD were also mild, with only 1 adult and 2 pediatric patients showing reversible grade 1 GvHD. No significant neurotoxicity was seen, with only 1 adult and 2 pediatric patients showing grade 1 neurotox, which recovered without treatment. The lymphodepletion in these UCART19 Phase I studies is much stronger than in autologous CAR trials combining cyclophosphamide and fludarabine with the addition of alemtuzumab. We believe that the strong lymphodepletion with alemtuzumab may be responsible for the slow recovery of blood cell count, which is something our partners, Servier and Pfizer, are examining.

Looking at our experience with UCART123, we have seen that allogeneic CAR-T cells can in graft and expand at a similar starting dose but without alemtuzumab preconditioning.

All in all, we are extremely encouraged by the safety and efficacy of UCART19 in these very early dose-finding clinical trials. The key question behind the safety in this early UCART19 protocol were the following: Can allogeneic cells in graft, expand and hopefully bring hard-to-treat patients into complete remission? The answer is yes. We are excited for Servier and Pfizer to expand this study into multiple clinical sites in the U.S. and Europe, as noted in the abstract. With that, I would like to turn the call back to -- over to André.

Well, thank you so much, Mathieu. This is a very nice section. And I would like also to reiterate the robustness of our manufacturing process. For UCART123 in 2016, our cost of goods per vial was already less than $4,000 for a dose at 6.25 x 10 to the 5th cells per kg for a patient of approximately 70 kilos, and we now have already manufactured enough vials to cover the ongoing Phase I dose-finding trial as well as the planned expansion study.

With early proof-of-concept for UCART19 and UCART123, I believe we have clearly shown that the age of the allogeneic CAR-Ts is here, and we're planning to stay in the lead. We're also pushing forward the manufacturing of UCART22 in B-cell ALL and non-Hodgkin lymphoma as well as UCARTCS1 in multiple myeloma with an IND filing planned next year. Our goal is to invest in expanding our management capabilities, which will enable us to produce multiple CAR-T cell programs simultaneously, bringing CAR-T therapies faster into different types of cancer indications.

Switching gears, I wanted to say a few words about Calyxt, our plant science subsidiary. This company has shown an impressive performance, both from an execution and of its business plan as well as a return on investment. The successful IPO this summer [led] the ground for us to build a public shareholder base independent to Cellectis. However, we are proud to remind our shareholders that Cellectis owns close to 80% of the shares in Calyxt, and we see that this is a beautiful growth story which will provide a valuable asset for Cellectis in the decades to come.

If you would like to hear the details of the Calyxt story, I encourage you to listen to the webcast replay for their Q3 earnings call, which just took place this morning. We're finishing the third quarter of 2017 with a strong cash position of over $300 million, which provides us with a cash runway into 2020.

With that, I'll turn it over to Eric for a discussion on our financial results. Eric Dutang?

Thank you, André. As previously remarked, we have changed the reporting currency of the group consolidated financial statement from euro to U.S. dollar, to improve the comparison with PF, which primarily presents their financial statements in U.S. dollar.

The financial highlights for the third quarter financial results was a successful initial public offering of plant science subsidiary Calyxt, Inc. Calyxt received net proceeds of $58 million after the issuance of 8 million shares at $8 per share. As part of this IPO, Cellectis S.A. purchased 2.5 million shares of common stock of Calyxt for a value of $20 million, which is included in the net proceeds of $58 million.

Cellectis remains the majority shareholder of Calyxt, with 79.8% of (inaudible). As of September 30, 2017, Cellectis had $304 million in cash, cash equivalents and current financial assets compared to $291 million as of December 31, 2016. The $13 million increase over the 9-month period was primarily driven by the following elements: $38 million proceeds received as part of the Calyxt IPO; $7 million proceeds from a sale-leaseback transaction at Calyxt; $13 million ForEx impact; which were partially offset by $43 million net cash flows used by operating activities.

The team continues to execute the strategy with financial discipline. We expect that the cash, cash equivalents and current financial assets position of $304 million as of September 30, 2017, will be sufficient to fund our current operation into 2020.

Total revenues were $27 million for the 9-month period ended September 30, 2017, compared to $43 million in 2016. Collaboration revenues decreased by $16 million, which was notably explained by $8 million one-time milestone revenue recorded in the second quarter of 2016 in connection with UCART19 and also explained by $5 million decrease in revenue recognition of upfront payments.

Total operating expenses were $92 million for the 9-month period ended September 30, 2017, compared to $90 million in 2016. Excluding non-cash stock-based compensation expenses of $39 million in 2017 and $44 million in 2016, adjusted total operating expenses increased by $7 million from $46 million in 2016 to $53 million in 2017.

For the 9-month period, R&D expenses remained stable at $58 million. Excluding non-cash stock-based compensation expenses, adjusted R&D expenses increased by $7 million from $33 million in 2016 to $40 million in 2017. This increase in adjusted R&D expenses was notably related to UCART123, UCARTCS1, UCART22 and other product (inaudible) development, including payments to third parties and costs related to clinical trials, purchases of biological materials and expenses associated with the use of laboratories and other facilities.

For the 9-month period, we recorded $32 million and $31 million of SG&A expenses in 2017 and 2016, respectively. Excluding non-cash stock-based compensation expenses, adjusted SG&A expenses remained stable at $12 million between both periods. Net loss attributable to shareholders of Cellectis was $72 million for the 9-month period in 2017 or $2.03 per share compared to a net loss of $54 million in 2016 or $1.53 per share. Excluding non-cash stock-based compensation expenses, adjusted loss attributable to shareholders of Cellectis for the 9-month period ended September 30, 2017, was $33 million or $0.94 per share compared to $9 million in 2016 or $0.27 per share.

I will now turn the presentation back over to André for closing remarks.

Thank you, Eric. Cellectis has today a strong balance sheet, and we remain well positioned to continue our program into the year ahead. I look forward to updating you on our progress over the coming months. And we strongly believe that the efficiency and accessibility of our allogeneic CAR-T cell platform will ultimately enable patients around the world to truly benefit from this groundbreaking treatment approach. I want to thank you much and very much for your attention. And I would like to open the call to questions. Joining me for the Q&A will be Eric Dutang, Dr. Mathieu Simon and Simon Harnest.

Operator, please go ahead.

(Operator Instructions) Our first question is from Christopher Marai with Nomura Instinet.

I was wondering if we could just touch upon UCART123 for a second. Specifically, you broke out capillary leak syndrome there in your data set. Speaking with some KOLs at SITC, we heard that even in CD19 treated patients with some of the autologous approaches, they see capillary leak syndrome, along with grade 4 CRS. Those seem to come together. So I was wondering if you had any sense, in your follow-up with FDA and in your analysis of the patient death, whether or not this was a target-mediated capillary leak syndrome or if this might have been more due to the CRS? And I have a follow-up.

Mathieu, do you want to get this question?

Yes, I think if the FDA has decided to lift the hold, that's probably because at this point in time, we're excluding the target-mediated syndrome. And we are more thinking about some pre-medical condition of the patient. That's why now we are much stricter on the selection criteria. We are going to treat younger patients, at least for the next 3 patients, 65 years old. We are going to be very careful about the entry criteria. The patient has to be totally without infections, for example. All those kind of precondition that can trigger these kinds of side effect. But at this point, I will exclude the targeted-mediated syndrome for CLL. I mean, at least that's the assumption that the PIs, ourselves and the FDA [is thinking about]. But we are going, of course, to monitor the next few patients.

Okay, great. And then I guess, just thinking about the data you provided for us on the cost of goods and manufacturing, I understand that the GvHD in the trials, for both UCART19 and 123 was quite mild. But can you remind us of the percentage of cells that have the TCR knockout in both those products? Are they the same? Was it higher in 123 versus 19? Just I can't recall the number, 99 point something percent. But what was that?

Well, yes, actually, I can take this one. The criteria for like the release of the batch is higher than this -- the filtration of our cells currently lead to 99.5%, which is a below-detection level, so we do not detect any TCR in the batches we've released, and we believe that our CAR-Ts have like -- actually, there is -- as you see, currently there is no GvHD. And the criteria for the release is largely below the 99.5%. It's 99.5% and below detection level.

Okay, great. Is there any ability to improve that? And then just maybe one last quick one on your CS1-targeted UCART. Maybe curious if you could elaborate a little bit more on why CS1 rather than the BCMA target that seems relatively validated in the autologous side of things?

Just to answer very quickly on the question of can you improve this, of course we can improve this. And we do have developed technologies to increase this to a higher level, even if we think that we are in a very safe place here. Since the first CAR we injected 2.5 years ago, the worst thing that has been seen is a grade 2 CRS on one of the first patients that could not be attributed, potentially, to our CAR-T, because the analysis of the cells showed maybe that potentially the skin rash that was there might have been attributed potentially to the bone marrow transplant that was performed prior to the CAR-T injection that we had before. But it was essentially suspected mild grade 1 GvHD that disappeared very quickly, and it's suspected. And in adults, there is no GvHD that has been analyzed up to now. So we think that to date, at the current level, Cellectis has a very high safety level for the TCR-positive cells. And for CS1, yes, like the [indiscernible], it's a reverse process for production. It means that we have to knock out first the CS1 gene in order to suppress from CD8-positive cells the CS1 gene and to keep a perfect balance between CD4s and CD8s in the vials we're producing, which is a guarantee that's a better product at the end and to suppress cross-T-cell reaction. Therefore, the gene-editing is a very important step in the preparation of CS1 CAR-T. And CS1, we believe, is a great target for multiple myeloma. So we're moving ahead with this CAR-T, and we hope to file an IND by next year.

Okay. And then just following up on the TCR knockout. Was there any improvement between the UCART19 knockout levels and the UCART123 product that you can quantify for us?

No, actually, it's essentially the same technology that has been used during the process and like the first batches that have been made were made at UCL. It's only a big difference between the first injection that we had, like 2.5 years ago, and the current batches. The current batches are done at a contract manufacturing organization we are working with called CELLforCURE, while the first batches were -- had been manufactured in an academic environment at University College London. That's the only difference between the 2, but we are still using the same technology, even though the technology has evolved a bit.

Okay. So both have the 99.5% below the limited detection TCR knockout?

Yes.

Our next question is from Wangzhi Li with Ladenburg Thalmann.

So I have a few. I would start with asking about UCART19. I think you also touch a little bit in the opening remarks. I don't know if you can provide any additional color regarding the UCART19 trial in the U.S. I know it's in Pfizer's hands, but any color -- additional color on the time line and the trial design? For example, it's going to be expanded to patients also eligible for (inaudible) CAR-T? Are you going to make the transplant an option versus a requirement? And also, thoughts on removing the -- potentially removing the alemtuzumab for lymphodepletion? Any color on that would be highly appreciated.

Mathieu, do you want to take this one?

Yes. Of course, the decision is more in the hands of Servier and Pfizer indirectly. I think the decision has been taken to expand those studies in the U.S. And very soon, you will hear from a very large and well-renowned institution, starting the clinical programs. Basically, the protocol, especially on the adult side, is going to mimic what has been done [as similar] in the U.K. at King's College. What we have to take in consideration is probably what are we going to see in the future. We are going to see less difficult-to-treat population. Because in the first U.K. study, as you know, we have very hard-to-treat population. If you take, for example, the first low-dose cohort, 5 out of 6 patients had already undergone transplants. And those patients had relapsed between 4 and 6 months. That means that they were in pretty bad shape. I mean, so the goal is to expand the study in the U.S. to probably get to a more, if I can say, normal population, less hard to treat, but it was the first attempt. It was the way to go at that time when we decided to go. Regarding the question of alemtuzumab, I think there are some discussion to look at protocols without alemtuzumab, with alemtuzumab. I don't want to speak on the behalf of my partners, but maybe in expansion phases, they will address that discussion through different arms. That's my feeling. Is alemtuzumab mandatory for (inaudible) of expansion? That's a critical question. We have shown in our -- the first studies have shown that you can see CAR-T circulating between day 7 and day 42 and even longer in the first compassionate programs. We have shown that in our UCART123, without alemtuzumab, we have also a very good persistence. I think that would be the call of our final 2 partners. It's clear that probably there is -- we could improve the way we address lymphodepletion in the future. I mean, it's -- because the way we have -- we lymphodepleted those patients was pretty high. We're giving 1 milligram of alemtuzumab per kilo; 1,200 milligram total dose per square meter of cyclophosphamide; and fludarabine, 120-microgram per square meter, total dose. And I know that in the future, they will do some changes in the protocol.

Okay, great. Very helpful. And then switching gears to UCART123. I don't know if you can provide any color on the autopsy finding and maybe any additional color on what you think is the most specific cause of the death of the patient, the BPDCN patient?

Well, it's always very difficult to comment, because we have kept the discussion with FDA, I think that's the only channel with -- that was for us extremely important. I mean, I think it's very much too early to say. I mean, when we started those studies, I mean, we were very much afraid of lack of persistence, lack of engraftment, lack of CRS. And to our surprise, we have CRS, we have persistence, but we have also some side effects. So I think now, what we are doing is starting to -- we are going to start with a much slower dose, it's 1 lug below, with a better population. And the goal is really to start and to come back to you with new findings, let's say, we said in 2018, but I'd -- the study is going to resume in the coming days or weeks, once the IRB of both institutions are going to clear the FDA recommendation. But I think it's too early to draw a conclusion, except the one that we have CRS, we have activity and we have expansion. That's the only thing I can do at this point.

Okay. Maybe last question on the finance part. So the USD 304 million cash, how much of that includes the cash of Calyxt? I think they are integrated together, right?

The cash position at Calyxt is -- was $62 million at the end of September 30, 2017. So you have $238 million for the therapeutic activities.

Okay. And maybe a last question. So any guidance internally, milestone payments from Servier for the UCART19 program? As you know, you now have treated about a dozen patients. Any -- does that trigger any milestone payments?

We don't expect to receive a milestone payment in 2017.

Well, when we give a forecast like through 2020, it's usually expecting with 0 milestones. That's our position.

Our next question is from Peter Lawson with SunTrust Robinson Humphrey.

André, just on the 123 trial, how long would you have to exclude the older patients? And do you think that's going to have an impact on the eventual label?

Well, you have to understand that we're still in a dose-finding moment. And when we started with like 6.25 x 10 to the 5th, we had approximately no idea if this dose was -- actually, every CAR is a different CAR. Every target is a different target. Every -- like each time it's a new story. So we tried to make an educated guess with 6.25 x 10 to the 5th. Currently, we're going to a lower dose. The first idea was we had no alemtuzumab preconditioning, because the only cyclo/fluda injection, so we don't know this. Now we're in a dose finding [mode]. This dose finding, we think it's reasonable to move forward in a more selected way with patients in order to try to find the right dose per kg. We're at 6.25 x 10 to the 4th now. We're going to re-escalate, potentially. We'll see. Once we get there and we start to go into the expansion phase, you see that in the expansion phase, the clinical trial for AML, for example, or BPDCN -- for AML, it's 2/3 of the patients are first-line patient with a complex or poor cytogenetics. And for BPDCN, it's half of the patients will be first-line. So the population will totally change only for the expansion phase. Here, we do from 9 to 12 patients in Phase I, and then it will be a total different world. Cellectis has no intention to limit itself to what we are currently seeking to find in the dose finding for UCART123. It's the same thing for 19. If you do dose finding, it doesn't mean that it's only if like the patients are not responding or not eligible for autologous CAR-T. This doesn't mean that. Who can do more -- do the more can do the less. So we will see afterwards. And moving forward, we'll start to expand in other indications and so on. But for now, we have to find exactly what is the right dose per kg. And this is what we're doing. So don't take what we are doing currently as probably like the filing for a BLA. One will go for a BLA. We'll probably change this, and we'll try to find the right population, and we think that allogeneic CAR-Ts for distribution, market access, cost and [seriously] like easy-to-access for patients. The fact that a patient doesn't have to wait for producers of raw materials, will expand this in a very dramatic way. But now we're still in the dose-finding phase. So no, the age is only to find the right dose.

And I don't see any delays in the recruitment of patients, because we have committed to treat the next 3 patients at 65 years old. But the average age for diagnosis in AML in the U.S. is 67 years old. So I don't think it's a problem at all.

André, you're kind of in an interesting position where you can re-administer the therapy. What kind of persistence do you think you see with the first administration? Has the therapy first (inaudible) administration?

Well, we believe our cells have a very significant persistence. When you look at persistence we had like for the 19 trial, it goes up to 80 days. But Mathieu said something that, obviously, people have a hard time to understand. It's there is the ability to re-dose. One of the patients relapsed CD19-positive. Two months later, after being MRD -- like patient was diagnosed MRD-negative, relapsed 2 months later and CD19-positive, was few months later re-dosed and was put back in MRD negativity. So the argument of the persistence has a very mild importance. And also, we are still at start doses here. For example, for 19, is like 5 x 10 to the 5th per kg. And you know that not 100% of these cells are CD52, so alemtuzumab-resistant. So probably a bit above 50%, between 50% and 60% of these cells are CD52-positive. So it's between 50,000 and 60,000 cells per kg that are injected. And we're moving to the next cohort now, which is 1 lug above. So you have to know that we're still, step-wise, moving forward into trying to find exactly what is the right dose. But going forward, like the argument of persistence over several years, this always come at the price, which is a disease linked to the persistence, where you knock out a full compartment of cells. Here you have a real drug that you can give several times in a row. And if the patient relapses, you can re-dose the drug. If it's CD19-positive, then you can re-dose the CD19 with UCART19. For example, if it's CD19-negative, you can re-dose with another CAR. For example, if we can have UCART22 to go into clinic, it can be re-dosed with UCART22 and so on. So you can see that the span of the potential of the therapy is just showing a small ray of light today. But the potential is huge in the future.

And just finally, do you think you need to knock out MHC class, I guess, I or II?

Well, for now you see that, for example, UCART123, there was no alemtuzumab preconditioning in the patient. It was only cyclo/fluda, and the cells expanded well. For 19, like the cells stayed for over 80 days with an MHC there. If there is a need in the future to see if we have to go and knock out MHC, we will do it. But if you pile up knockout that are not compulsory for the performance of the product, it's better to avoid this, because you can do other knockouts that can be good for the performance such as, for example, we can knock out PD-1 or add other features. Or for example, like a CAR-T like CS1, UCARTCS1, you can knock out CS1. For 38, you can knock out 38. So the number of knockouts don't have to be piled up just for the beauty of it, it has to have a necessity. If the cells, for example, 123 was only a plain TCR cell knockout, no CD52, can't engraft and expand in the patient, then well, I can ask you the question, why would you add MHC knockout. For persistence? Well, here comes like the second argument that I said before. Well, if the CAR gets rejected, then we can re-dose the patient afterwards. We have the longer experience in the field of allogeneic CAR-T [for now], and we have been doing a series of patients, and we're starting to have a very good insight of how the cells expand, behave, the durability of these cells in the patient, where they go, what they do. So I think that now, we can have like a deep information about that.

Our next question is from Nick Abbott with Wells Fargo.

It's Nick. Jim's traveling this morning. But first of all, congratulations on getting the clinical hold lift. I'm sure that's a great relief. And so maybe just starting there on 123, did you discuss a split dosing regimen as another potential way to try and address the safety challenge? And part of my reason for asking that is, we just saw this root cause analysis on JCAR015 by Juno, and it was fairly clear that you could identify which patients were most likely to get into trouble early on. So I was wondering whether you discussed a split dosing. I know some of the protocols have day 1, 2, 3, but obviously, it doesn't have to be that.

No, we did not discuss split dosing, because according to our PIs and ourselves, split dosing is very useful when you address infusion toxicities, and we have not seen infusion toxicities. We have seen more some CRS. So in that case, we don't believe that it is the avenue to pursue. That's why, I mean, very clearly, we got to the consensus with the FDA that diminishing the dose was more the way to go, rather than to split dose between day 1 and day 2. Because at the end, the total dose would have been the same given. So we went more through a better and more careful selection of the patients. Keep in mind that our first patient that unfortunately died was a difficult patient to treat. So I think it was not a classical patient that usually we include in those very early dose-escalating Phase I. So it's more careful selection of patients. And where you've seen the dose -- because again, it came as a surprise. A lot of people had expected that an allogeneic CAR-T were not that persistent, that an allogenic CAR-T were not as expanded. And what we have seen in the first 2 CD123 patients is completely the contrary. So we believe that starting with a lower dose, where the toxicity, the side effect would be more manageable, it would be easier to really define what's the optimum dose and the optimum -- and the safety profile of this drug more quickly and in a more organized way.

And as you go forward, do you intend that dose level 2 would be a lug higher, or you wouldn't make such a leap?

No, the plan is to go back to...

I guess it depends on what happens with dose level 1, really, doesn't it?

Well, the plan is to go back -- if we see that the more careful selection of the patients give us the appropriate result, then we'll come back to the original first dose. I mean, that's the plan we have with the FDA. I mean, that's -- if we don't show activities. If we show activities and if we have a good side effect, fine, that may be the dose. But in that case, we will do some confirmatory patients to confirm the dose. But our plan is really to go back to the 65-year-old population, to eliminate some comorbidities like the infection status at the time of the injection, which is a pretty important one, because that's a triggering event of the CRS. Also, probably, we're not going to give G-CSF, filgrastim, during the first 3 weeks, because that's also something that can enhance the cytokine expression. So we have decided to postpone the use of filgrastim after day 28, so (inaudible). And then we'll come back, if we have less -- lower activity, we'll come back to the original dosing.

Okay. And as part of your discussion to the FDA, did you review the protocol for using rituximab as a kill switch?

Not at all, not at all. That has never been challenged, never been in question. I mean, in fact FDA went, the discussion we have had with FDA was very much the same and the conclusion we got from our PI and different safety boards, so the 2 institutions, MD Anderson and Cornell, Weill.

Okay. Excellent. And presumably, as you think about the clinical trials for CS1 and 22, the experience you've had with 123 will influence your starting dose there or -- I mean, I know André said every CAR will be targeted different. But do you feel like going forward, you're going to be starting in the 10 to the 4 range?

I think each target is different. I think 22 mimics much more the CD19 story. So at this point, we have not filed the IND. But the plan is probably not to do very much different than what we do today with CD19. CS1 is a completely different story, because it's a different target, different kind of population also. We have not yet assessed the dosing, but again, we believe that each CAR is different. Each CAR is different, it's also very much patient-dependent, comorbidities, tumor burden, so it's a difficult science. But the goal is really for each program to individualize the dosing scheme.

Okay. And then maybe just one more for me, and that is -- the question was asked earlier about BCMA versus CS1. Have you looked -- I mean, I know your partner is developing a BCMA CAR, but have you looked preclinically at the relative merit of a CS1 CAR versus a BCMA CAR? I mean, obviously, again, we heard at SITC about soluble BCMA, and that's been discussed quite widely as potentially reducing the efficacy of a BCMA CAR. But what are your thoughts about BCMA versus CS1?

Well, as you know, we are collaborating with the MD Anderson with one of the best multi myeloma doctors, Professor Orlowski. We recognize that BCMA is a very good target. But according to Professor Orlowski, CD38 or CS1 could be also 2 very good targets. CS1 has the advantage that, first of all, we are the only company that can gene-edit the CAR-T in order to remove CS1 on the surface of our own T-cells in order to avoid cell destruction. So that's already a point where we have an advantage over competition. And also, we like very much the CS1 expression. It's a very pure expression. There's very few off-target toxicities, very few off-target expressions. We also love CD38, because we looked at these products. It's -- the antibody, daratumumab has been extremely successful in clinic. It's a more complex development because CD38 is expressed in other tissues, other than the myeloma cells. That's why we have decided to give full priority, I mean, to CS1 and we're going to start as soon as possible with Professor Orlowski. We love this target. And you know that Pfizer, our partner, is making progress on BCMA [the rest of the way]. But BCMA is a very good target. We had also that in mind before we signed the deal with Pfizer. They are 2 very good targets.

And we have a follow-up question from Wangzhi Li.

So I just have a quick follow-up. Is there any thoughts on UCART19 in DLBCL? Because it's a larger indication than ALL and also I think at the autologous CD19 CAR-T data kind of indicated that the persistence requirement is maybe less in DLBCL versus ALL, because the NCRR study showed that there autologous CD19 CAR-T is also about 1 month persistence. But the (inaudible) remission without a transplant. So I don't know what's your plan or thoughts on expanding UCART19 into DLBCL. Any color there?

Well, I think it would be a very logical strategy to go after this disease. But that's going to be the decision of our 2 partners, Servier and Pfizer. And keeping in mind that we are still at the very early stage. We are today looking at the products in ALL, with very difficult-to-treat population, with very, very low dose. We are -- as André was saying, we are 1 to 2 x 10 to the 6th cells per kilo. And keeping in mind that the results that we are presenting today have been optimum [least] dose. If you look at the (inaudible) the patient information prospectus, they are recommending to use this product at 1.2 to 2.5 x 10 to the 8th kilos. So we are at the very preliminary stage in the clinical development and very happy to report those, what we call very impressive -- very outstanding results for such a low dose. Regarding the strategy of our partners, I think it would make sense, once you have developed a product in ALL, to look at all of the lymphoma mutations and probably also to think about CLL. But that's a decision that would be taken by our 2 partners. And I'm sure that looking at the size of the market, looking at the similar translational signs that may apply to other diseases like (inaudible). Now we know that we have a good [owning], that we persistence, that we can go into the [lymph nodes] because probably, we have the sufficient persistence. It makes a lot of sense to go after this indication. But it's -- today, UCART19 is their product.

Ladies and gentlemen, we have reached the end of our question-and-answer session. I would like to turn the call back over to Simon Harnest for closing remarks.

Yes, thank you very much again to everyone for the great questions, and thank you to the team for this great Q&A session. Feel free to shoot me an e-mail directly or call me directly at any time if you have any follow-up questions, and we're looking forward to seeing you soon at the upcoming conferences. Thank you so much.

Thank you. This concludes today's conference. You may disconnect your lines at this time. And thank you for your participation.