In English, welcome everybody to the web call of AB Science on the latest news, which is the conditional approval of Alsysek in amyotrophic lateral sclerosis. My name is Alain Moussy, I'm the CEO of the company, and with me is Laurent Guy, who will facilitate the execution of this call. So we'll present as usual the presentation. We'll go through it, and at the end, you will be able to ask questions that you're going to ask by typing your questions on the computer, and we'll take it and try to answer. We'll try to keep the time to one hour for this web call. Next slide. The disclaimer. Next slide. So the topic is the conditional approval of masitinib in ALS with EMA. Again, the objective is to give you details about the objections that has been raised by EMA, and the answers that we have proposed, that will also have a second objective, which is to give you more information about the study that substantiate this conditional approval. Next slide. So the first information we received from the CHMP, which is the committee from the EMA, is that it actually validated or confirmed the safety of masitinib in ALS, which is deemed to be acceptable. And so in quote, "Having considered the data from masitinib studies, the safety profile is considered acceptable." Next slide. However, the CHMP was not able to conclude on a favorable benefit risk due to the following methodological issues, and we received four pending objections at this time of the evaluation of the dossier. And we are going to get into the details of each of the objections, of course. So the first objections, I'll try to explain. I do apologize about the technicity of the presentations. I will try to make it understandable, but I'll do my best, and we'll try to precise later if you have questions, if it was not completely clear, of course. So the first issue concerned the GCP. There has been some deviation to the protocol, and the objection that remains at the end is that some identified GCP deviations to the protocol cannot be corrected retrospectively. We are going to tell you more about that. The second objections concern the primary population, which excluded fast progressor, and CHMP disagrees with this exclusion, so we received, in writing, this pending objection, which is the significant effect was not demonstrated in the total population, including the fast, although our protocol didn't say that, but the primary population, excluding the fast. So I'm going to detail that. Then the third objection is more statistical, is the handling of missing data. There are also missing data that we portrayed in our study, and, according to CHMP, there is a preferred methodology called Jump to Reference, and Jump to Reference should be applied to some discontinuations. And they consider that we should have applied this penalty called Jump to Reference to all discontinuation, but not some of the discontinuation, as I will explain later. And the fourth objection concerned subgroups discussions where we used an EMA guideline, and they thought that this target population, subgroup population, was identified post hoc, and so could not be taken into account, despite the strong efficacy as you will see in this subgroup. So that's the issues that remain at this stage of the process. Next slide. Okay, so I'm going to present you the arguments, or the key arguments, that were used trying to respond to those issues. So the first one, again, is pending GCP issues that cannot be corrected retrospectively. Some can be corrected retrospectively, like for instance, the collection of all AEs at the site level, but some cannot. For instance, if there has been some deviations of eligibility criteria, where the patients have been randomized, and so it is not possible to correct them since they have finished the study. So it's true that not everything can be corrected. We corrected whatever could be corrected, and we applied then the guideline of EMA to whatever could not be corrected. What the guideline says is that when it cannot be corrected, the sponsor should measure the impact of what cannot be corrected. And this is exactly what we did, and it breaks it down in two categories. Some findings that potentially could likely influence the benefit risk, or some that may, which is less strong. And as you can see in the green part of the table, this is what the, I would say, findings that were found by the inspections. And of course, we responded to each of them exhaustively, where whatever could not be corrected, we measured the impact using some complex methodology. So that includes, for instance, the violation of inclusion and exclusion criteria. As an example, I don't get into the detail of each of them, but we followed the guideline, and we measured everything, and we proved and demonstrated that there was no impact. Next slide. Then we have to apply a second guideline of EMA that says that, and I will read the right end part of the slide, which is the guideline. So the guideline says, "In superiority studies," which is our case, "findings that do not introduce bias favoring one treatment over the other are relatively unproblematic." So when you prove that there is no impact in whatever deviations that have been found, then you match, and you solve this problem. The other point is, it is important to assess whether the findings affects the primary efficacy, which is the ALSFRS, the functional score, and the safety endpoints. Now, and if those two things apply, then the benefit risk can be evaluated, and there is no more GCP issues. So we followed this guideline. Now, as you have seen, I will start with safety. The safety was unacceptable, so it's not the problem. So let's then move to the primary efficacy variable, which is a functional score. The functional score was not even found by the inspections as a systemic problem. There were very few discrepancies, which had no impact, and we had reassessed anyway this variable, and we found absolutely no impact. So in fact, this primary variable is robust, and the safety is acceptable. So here we have the cornerstone of the benefit-risk, but it's not enough. Then we have to go to all other aspects that could impact or favor or create a bias in favor of the treatment. And as I have showed in the previous slides, we demonstrated that the rest, which is less important, but still, has no impact. So we think we applied the guideline fully, and the guideline says that those GCP issues, although there were some deviations, have no impact, and so the benefit-risk could be evaluated. Next slide. Now, we tackle the second issue. The second issue concerns the exclusion of the fast progressors as the primary population of our study. We transitioned from phase II to phase III, and when we transitioned from phase II to phase III, we excluded the fast progressors from the primary analysis. Here, it's not really the point of the amendment, it's the point of excluding the fast itself, which is a debate with the EMA. Why we did that? We did that because the fast progressors, as it is, actually says progress faster, and there is a high risk that they do not reach the time point, which is week 48, which is a long time point, and they will create missing data and also die. So creating even more, I would say, standard deviations. And in fact, it's exactly what happened. As you can see, on the right-hand side, the normal progressors had a 26%-30% chance to reach the time point, but the fast progressor had more than 50% not to reach the time point. It was a good decision to exclude them, to minimize the missing data, and also because we then have a more homogeneous population, as the guideline recommends, we have a more homogeneous population. So that was justified. Now, the counter. The question we had to answer is the definition set for the fast. We have to define the cut between the fast and not the fast, and it was based on literature. The EMA questions the robustness of the methodology used to define the fast, which is to have a first point, which is the time of the diagnosis, and then a second point, which is the baseline, where we start to randomize the patients. So this is the period where you measure the speed of the progression of the disease. And the question from EMA is to say, "But that is not stable through time," which is true. It's some patients start as a normal progressor and become fast, right? That happens. However, a literature say that from onset of the symptoms, the definition of the speed of progression is what is the most robust and what is the most predictive of survival. So in fact, it's backed up by publications. It's also backed up by the practice in clinical trials, because it has been used by most of the clinical trials, and in particular, the ones that have been registered, like edaravone, registered in the USA, and tofersen, registered both by FDA and by EMA. So we are not the only one to have used the same definition and to have excluded part of the population on the basis of that definition. It's also usable in clinical practice, and we have demonstrated in the process that this is robust enough because we don't ask the patients the precise date of the onset of symptoms, but in fact, the classification would not change if they remember the months of the onset of symptoms, and of course, the patients, they do remember normally the months. And also, we demonstrated that the study is still successful, even in the bracket, which is very large of the cut, because we decided based on the literature, the cut was 1.1. But in fact, the study was successful from 0.8 to 1.4. So in fact, we had a large bracket of flexibility, but still it was a matter of debate, and EMA does not favor or recommend to exclude fast progressors, which by the way, is not a point of discussion, for instance, with Canada or even with FDA. So it's, we would say, rather a specific point for EMA, but at the end of the process, we were unable to convince them that it was the right decisions to make. Next time. Next slide, sorry. So next slide is a technical problem, either, on the statistical treatment of missing data. So as I have said, there is a missing data percentage at week 48, because patients progress in their disease will discontinue and not reach week 48 or even die. So there is a concern, methodological concern. It's how to treat this missing data, because we cannot have, at the end, missing data. So we have to impute a data which does not exist, and how to do that methodologically, and there is no rules, which is set up by the guideline from EMA or from FDA, which is universally recognized. It's not a mathematical model. So there are some kind of recommendation, but not a consensus in the industry. So EMA favors what they call Jump to Reference. Let me explain. If there is a discontinuation in the treatment arm, not the placebo arm, then EMA wants to impute a penalty. The penalty is immediately the patient that will discontinue under a treatment arm will have to be considered as a placebo. So whatever the score at the time of discontinuation, immediately will jump to the reference, which is another term to say placebo. And then we follow the slope of deterioration of the disease of the placebo. So in fact, it's two penalties in once. First, you jump to the reference, the placebo, and then you impute the slope of deterioration of the disease of the placebo. That's the preferred methodology used by EMA, which we applied, of course. So the point is not whether we applied it or not, we did. But when we applied it, we have to applied it on some discontinuation. And the discontinuation that usually are used to apply this penalty, is on what we call the reasons, not at random. So in fact, it's like it's what is related to treatment, like lack of efficacy or toxicity. So when we apply this penalty for those reasons of discontinuation, which is classic, we have a P value of 0.00389, which is significant, and that's what we provided to EMA. And EMA challenged that and said, "Well, the other discontinuations," it's not on this table, but there are minimum, like less than 10, but still, those discontinuations, they wanted that we impute we imputed also penalty, although they are at random. So, okay, so we did it. First, we did it on travel, because travel, we don't know, they cannot travel because of lack of efficacy or because they move to another home. But even if we apply penalty on travel, doesn't change P value 0.00376. But if we apply a penalty on everything, which is what they wanted us to do, then we lose significance, but it's close to significance. The P value is 0.00678, as you can see here. Now, we received two recommendations from EMA in the history of our dossier. The first one is, when you apply discontinuation to penalty, sorry, to all discontinuations, then you can use a calculation to see what's the remaining treatment effect that is necessary for the studies still to be successful. And if this is less than 25%, then it's very good. And in fact, it's 24% in our case. So the recommendation should apply, and applying the very, I would say, the very first one recommendation from EMA that we received before, the study is successful, even in this most conservative pattern. And the second recommendation was to apply another methodology called Copy Reference, Copy Reference and reference. Just for you to understand, it's a progressive return to reference, progressive return to placebo, not a brutal jump as the first methodology. And when we do that, the P value is also significant, 0.00477. So applying the most conservative method, the study is, in fact, positive. Only when we apply discontinuation in every on all discontinuation, we lose it by short, but the recommendation from EMA, the two recommendation we received before should have applied. That's what we told them, but still, they have at this time maintained the objection. Next slide. Then we provided lots of alternative data. Not alternative, but supportive data, to prove that the product has a robust efficacy. The first data we provided was quality of life. Quality of life measured by a score called ALSAQ, and we have used the most conservative method to calculate it, which is on the right-hand side, called multiple imputations, the jump to reference, the copy reference, as you have, as I explained in the previous slide. And using that, it's significant. The P value is 0.05%. So we can say that masitinib has a benefit on quality of life, which is very important for the patients, in particular, in the context of the conditional approval. Then we used another methodology for the primary analysis, that is the functional score called CAFS. That is a ranking that FDA, in fact, prefers, not EMA, and this is close to significance. The study was not designed to demonstrate the significance on the CAP. But it's interesting because it's considered as a very difficult variable to reach, and you see that it's close to significance. Next slide. So the next slide is about time to event measures. So in oncology, we use, for instance, the time to progression or death. It's called PFS, progression-free survival, and we use, of course, survival. So we did the same because this disease is fatal. So the progression-free survival, which measures the earliest event between... So the progression of the disease or the death, which was a pre-specified endpoint, is significant. The P value is 0.00159. You can see the curve on the right-hand side. It's a very good, supportive, variable to have in time to event. And then we've been challenged by potential bias to the progression-free survival, which are the use of what is used against this disease, which is tracheostomy, highly invasive, of course, ventilation, sometimes permanent ventilation, or gastrostomy, which is to have a tube to feed the patients. And that can extend survival, so we need to understand the interaction between that and the PFS. And we first demonstrated that those events are occurs after progression. So what happens is that there is a progression, then those events come, and then deaths come. And so it makes the PFS more robust. Then we have showed to EMA the more I would say integrated endpoint, which is called event-free survival, which integrates everything, like progression, but also tracheostomy, ventilation, gastrostomy, or death. So five events. And this endpoint is significant, 0.00162, which proves that masitinib is able to delay those very negative endpoints. Next slide. So the next slide about the survival. Again, this was not the primary endpoint, but still extremely important, of course, to analyze. So the first thing we did is that, it's the upper part of the slide, is that we, you know, there is in the design of the study, an open-label phase where we unblinded the treatment, and we proposed the placebo to switch to masitinib. So it was important to verify what was the impact of this switch, because we measured the survival long term. And so that analysis there is the switch, the difference between the placebo switch to masitinib and the placebo who stayed under placebo, and it was done 3 years, roughly, in average, after they've been randomized. And you can see here, statistically significant. It means that masitinib was able to give more survival even to placebo, who started the drug 3 years later. Now, that is important to analyze the survival as a whole, because it creates a bias against masitinib, because we switched some placebo to masitinib. So we have to retreat this bias. This bias is retreated, and when it's retreated, that's the table on the bottom part, you can see that survival as a trend. It's close to significant, 0.007%, and it brings six months additional survival. So the, in the population of the primary analysis, it's not significant, but it's a trend. And the guideline of EMA asks a trend, does not ask significance. It ask, it asks a trend in time to event. So here we have significance on PFS, significance on EFS, and trend on survival. Next slide. So then we move to the fourth objection. That objection is because we proposed a to use a guideline from EMA. The guideline is on the subgroup. They said this guideline cannot apply. They cannot take our data on the subgroup that you will see are extremely robust, but they cannot take it for a registration. It's not that they ignore the data completely, but for the purpose of the registration, they feel they cannot, they cannot use it. And why? Because they say that these, these subgroups, I would say, has been identified ex post, which means it was not pre-specified in the protocol, which is true. It was not in the protocol. It has been determined after, in ex post analysis. However, the guideline says that this guideline can apply even from post hoc identification of subgroup. It is written black and white. It says, "It is of interest to identify post hoc subgroup where efficacy and risk benefit is convincing." So that's what we did. And in fact, the guideline says another thing. It says that when the clinical data is statistically persuasive, which is our case in the primary analysis population, but there is a subset of patients, a part of the population, where there is a bias, that's what we have put in bold, then this is of interest to identify the subgroup, the subgroup being the full population minus the subset. And that's what we did. So next slide, I will show you where is the bias in the subset. So this subset of patients is the most severe patients, is the patients' loss of function identified by the score, the functional score called ALSFRS. A loss of function means what? The score will measure 12 functions. A function is what the patients can do, and the loss of function means that it's at zero, so patients cannot do. For instance, cannot speak or cannot walk. They cannot do something. And so when there is a loss of function, it means part of the motor neurons are dead, and the patients will probably die very soon. Of course, those patients are extremely severe. They are most severe of the populations. Now, when we take this subset of patients which was present in our protocol, they were eligible. We see there is a huge imbalance between masitinib and the placebo. The placebo, they were 8%, and the masitinib, they were 20%. And so that creates a bias against, because it's a more severe population. In fact, when we look at the severity of this subset of patients, we can see the table, below, that it's not one, in average for the placebo, there were nine patients with a loss of function, and eight had only one out of 12 possible loss of functions, and one had four. But in the masitinib group, there were 21, and ten had one loss of function, but eleven had two or three or four. So they were even more severe than the placebo. Next slide. So, as per guideline, it is of interest, as the guideline says, to look at the data when you remove this biased subset. So it becomes a subgroup that is called patients prior to any complete loss of function, which is the column in the middle. The left is the primary analysis population, and the middle is the subgroup, okay? And the right-hand side, it's another subgroup. It's a subgroup where we have added the fast, because we have, of course, integrated the fact that CHMP or EMA does not like that way we split the fast. So it's important to include the fast in some analysis to see if it is still beneficial. So in this subgroup, prior to any Copy Reference, you can see the efficacy variables. So the functional score, calculated with Copy Reference, is significant. The CAFS, which was not significant, you remember it was close to significance, now is significant. The quality of life was significant, remains significant. The forced vital capacity, which is the respiratory function, is significant. The PFS moves from four months to nine months benefit, and the OS, which was just a trend, is now significant, with a significant benefit of one year of additional survival, which in the context of the ALS, is, of course, of extreme clinical relevance. And when we add the fast progressor, which is so important for EMA, you can see that the data remains very strong with a non-significant but close to significant benefit in survival, which is of interest. Next slide. So, unfortunately, and despite our arguments, at that time, four objections remains. And because objections remains, we don't have the conditional approval at that time. Okay, so what are the next steps? First step is that we have not received the full feedback from the EMA, and we will have more information in the coming weeks, I would say. Also, I precise that the next step for the CHMP is to have a vote, final vote. We received the feedback, which is a trending vote. The trending vote is negative. So I have to be very clear with the audience, when the trending vote is negative, the final vote, which comes one month later, is usually negative. Not always, but in a large proportion, negative. So I have to be very clear with the audience there. But we are going to have a discussion with EMA, and we receive more information about the reasons why those four objections remain, despite our arguments that we objectively feel were strong. So let's wait. And then what we know, because it's the process, is that there is a re-examination possibility, which is a second cycle. So we are here at the end of the first cycle, but there is a possibility of a second cycle called re-examination. Now, how it works is that this re-examination is done, will be done by a new team, which is always an opportunity, because it's a new, I would say, look at the same data. So new reporter, new co-reporter. Then, in the re-examination, there is the possibility to ask a Scientific Advisory Group, which is an external advice on precise points, which is not really possible in the first cycle. So of course, if we go to re-examinations, we would be very interested to have the opinion of the SAG, because CHMP usually follows the opinion of the SAG. So we can imagine then, in an option of a re-examinations, we could maybe treat those four questions that we have, which relates to why at that time, the arguments that we have developed have not been, I would say, successful. So for the GCP, we applied the guidelines, but still, it has not been, I would say, recognized as successful. We don't know why exactly at that time, okay? But we can go to re-examination, for instance, to debate that. The guideline on the subgroups, it seems that we have applied fairly and objectively the guideline on the subgroups. The objection being post hoc, but the guideline didn't say it's impossible for post hoc. It says it's very, it's of interest in case there is a bias on the subset of patients. The precise wording is "of interest," So why this guideline is not applied in this case at that time, okay? We received two recommendations for handling missing data on the statistics, as I have explained. We applied objectively those two recommendations, you know, the copy reference and the J2R and it has not been, at that time, recognized. We can, of course, ask the opinion of the Scientific Advisory Group, and then we're left with the last objections. We do not refer to guideline and do not refer to previous recommendation, but refer to position of the EMA, which is not, I would say universal, because we have seen it's not the position of other agencies, but still is the position of EMA, which is the exclusion of the fast progressors. They don't like that, that the studies exclude fast progressor, although in some circumstances, they have accepted it. So it's not entirely clear, and we have heard their argument about the stability through time of the definition, how you define a fast progressor, which is a pure technical point, but there are arguments and counter arguments, so it's not obvious, it's a point of debate. So that's what we could discuss, but I try to explain you that it's very technical. It's extremely technical. It refers to guideline, it refers to definition, it refers to statistic, and that is complex, and can be debated as it is sometimes the case when experts treat a complex problem. Next slide. So I wanted, in this presentation, also to give you some comparison with Canada, where we receive also objections. Next slide. Between EMA and Canada. So was it the same? No, it was not the same. It was the same topic, I would say the same section, but not for the precise same reasons. Let's try to explain. For instance, the primary analysis population, as you have seen, EMA is not favorable to excluding the fast progressers. Okay? But what does the Canada say? They had no objections on that. They actually understood the transition from phase II to phase III. They understood that when we transition, amendment could be inevitable and not data-driven. They understood that there was a necessity, as I explained, to select a more homogeneous population to limit heterogeneity. They understood the cut of 1.1, but they had still, at that time of the process, some concerns about whether the amendment was not too late and whether it was sufficiently justified. So in the justifications, as we have done, we have explained it by the necessity to minimize missing data. If we do not exclude the fast progresser, then we will have lots of missing data. Other protocols are not at week 48, they are at week 24. So maybe for other protocol, necessity to minimize to exclude fast at week 24, is less necessary. In fact, it's less, it's less strong. Whether for a study at week 48, it becomes, of course, much stronger. So, that's the main justifications. The second, just... Then, whether the amendment is late, too late or not too late, or acceptable, is a matter of course of debate. So we have arguments. The arguments is that it was done when we transitioned from phase II to phase III so there were obviously some patients already included, but still, we were at 2.5 years from the end, so which is a long period of time. When we did it, where we exclude the fast progressor, there were only 8 fast progressors, and they were spread across 3 arms, which is very, very minimum. We had included 12% of the data already of the patients, and we had acquired 12% of the data. But when we remove those 12%, the study is still successful, so it has absolutely no impact. So we have provided those arguments to Canada, as you know, and they've accepted to go to reexamination. The second objection is the statistical method for imputations. You have seen that EMA, its preferred method is to impute a penalty, and the discussion is even not on whether to impute a penalty or not, we do it. It's on whether it should be imputed on all discontinuation or not all discontinuation. That's rather that, okay? So it's not that... It's not the point of Canada. Health Canada has a different methodology to analyze missing data, and this methodology is very complex because it says that it should be nonlinear. Just for you to understand, some patients discontinue and some die, and because some die, and the distribution is nonlinear, Canada is right, okay? But that statistical specificity, which is extremely complex and specific, I would say, is only for Canada and FDA. They follow, in fact, the FDA guidelines. So EMA doesn't care about that. So they have no objections about the linearity or non-linearity of the distribution. But Canada, yes, because they follow the FDA guidelines. So they ask us to prove that all what we, all of our analysis were also substantiated by positive, data when we use different, methodology, which are nonlinear, okay? So it's complex. But we actually did, and we have, positive data, statistically significant in nonlinear, that we presented, at the time of reconsideration, and, and, and we hope that it can make the point. So then, and by the way, just for you to know, they did not ask us this data before, so unfortunately, otherwise we could have done it before. But, we actually added at the very end, when we received the, the conclusion, so to speak, otherwise we would have provided it before. So that's why reexamination sometimes is useful, because it can respond to questions that could not have been responded. Not because we responded wrongly, just because we didn't have the questions. So then we have a subgroup analysis. So the subgroup analysis is the same issue as for EMA or that time, yes, because they have the same problem, the same post hoc. And, and we say, "Yes, it's post hoc," but still the guideline, we, we do not say should apply, it's not a must, but can apply. It is of interest, in particular, when there is a bias in the subset. So you see, nothing is obvious, everything is technical, and it's an interaction with the agencies. It takes several steps. Sometimes, unfortunately, at the end of the cycle, not everything is responded or completely clear, and this is why sometimes re-examination is of some value. Re-examination has no value if it's to repeat the same argument, because we know already their positions. But we need to have a second look sometimes, or because the arguments have not been entirely considered, and they might be worth to be considered another time, and sometimes with the help of Scientific Advisory Board. So, this is what we have today. It's extremely complex. Usually, sponsors don't do that, what I have done, right now, but we think it is of merit, because what we want is to explain the audience that, the data are good, despite the fact that today, at this stage, we have not been able to convince neither Canada nor EMA. Next slide. So, as a conclusion, I have only one more slide. I would like to tell you why we think the data of the program, are robust, because there is a possible confusion to conclude that because in Canada and EMA, at this stage, do not give a conditional approval, then the data are bad. And this is important to address this question. Next slide. So the question of the conditional approval is to ask an agency whether a single study is acceptable for registration, and there is a guideline on single study approval, and that requires very compelling evidence. So I know that a lot of people in the audience thinks that because ALS is a devastating fatal disease and there is no drug, the agencies should be very flexible, and the hurdle should be low, and it's not the case. They apply the guideline the same way for ALS as for another indications, and their guidelines is very compelling data. So what the definition of very compelling is not known. Is there, I would say, a power to judge what is very compelling and what is not? But that's the question we ask. So to some extent, what I would like, you to understand, is that we ask a question whether the first evidence we have in ALS is very compelling. So the hurdle is very high. That's why it's very difficult to get an approval. Now, we, as a sponsor, when we decide to go for conditional approval, we, of course, consider that hurdle. We think that the data are strong enough to try to pass that hurdle, but it's our opinion. Maybe, the agencies have a different opinion. So maybe we will not convince neither Canada nor EMA that it is very compelling enough, but doesn't mean that the data are bad. That's my point. It's just a question of hurdle with only a first piece of evidence. So we think that the data so far of the masitinib program, preclinical and clinical, are robust and are completely fitted to support a full approval, not a conditional approval, when we will have a confirmatory study, which we hope it will be positive. So the preclinical data, I'm not going to present the details of that, and, we are late anyway, but what I want you to, to understand is that there has been recently two drugs developed, Edaravone and, and AMX. And the Edaravone is still, registered, and then AMX was registered, and their mechanism of action was unknown, but still registered. Here, we have an advantage. We understand how masitinib works and the mechanism of action, which is to target the microenvironment of the motor neuron, and the innate immune system in this microenvironment, and in this network, in this innate immune system, macrophages and microglia, has been published, validated by the scientific community. That is a strong point, because we know that that has an impact on the disease. I don't know whether the impact is, in one study, very compelling or not very compelling, is another story, but the mechanism of action we understand, and we think is very valid and can create a benefit in clinical. Then we have the clinical data. We have only one piece of evidence, which is not a lot, but still is good. What we have seen in this study is that provided that we exclude the fact, which is a matter of debate for EMA, the study is successful. Then there is a debate on penalty, full penalty and so on, but still, we can see a lot of variables which are strong. And there was a bias with the subset of patients who had at least one loss of functions. And when we remove this bias, the study is extremely strong, and that gives the idea of the ultimate benefit of masitinib. So that benefit needs to be replicated.... But provided we replicate, then we have a drug which is very strong. So what I would like you to know and to understand is that, yes, it's disappointing news, and yes, we did all efforts to try to register conditional approval, and we will still continue, apparently. But you should know that masitinib has robust data, which is recognized as such by the scientific community. So we try to ask you to not to differentiate the points of the conditional approval, very high order, from good data or robust data for the future. And with this presentation, I thank you very much. I've taken a lot of time, but it's technical consideration, which is hard to explain, and we will take maybe not all of your questions, but some of your questions. Laurent? Yes, we take questions that are really related to this conditional approval procedure. So first question: If data is strong but agency are not convinced, what can AB Science do? Discuss with the agency. Or maybe, Laurent, if we can go back to the re-examination, the next step, slide. Before. Yes, for Canada. Yes, thank you. So, when we finish a cycle and it's not successful, we have to discuss with the agencies because, obviously, we have a meeting, then we have to leave the meeting, and then we receive an email, okay? So, it's not enough, given the... what is at stake. So, before that, there are other steps, which are unforeseen, which is, of course, a discussion with the agency. So we have to take the necessary time to understand. So the first step is to understand. Here, we do a difficult exercise, which is to give you immediately all information, not all, but the key information that we have. And like me, you would like maybe to know more, but we have to wait the interaction that we will have with the agency. So we will do that, and we will. We work closely with the agencies, with EMA and with Canada, of course. So there is an interaction. To understand better and everything, take the time to understand. Then, we will have a possibility of re-examination. So to your question, is it worth to go to re-examination? But you have seen our arguments. They seem to be valid. If they are not valid, we have to understand why. And sometimes we disagree, and if we disagree, we have the possibility to ask a re-examination. So, let's wait what we are going to do, but please, we share with you the possible arguments for re-examinations. Next question: Have the associations of patients provided any feedback to the decision of the CHMP? Hello. We were supported by the patient associations before we entered in the call. Feedback after, yes, it's a disappointment, like us. I understand that EMA wants a robust sample for this study, and the only way is to have a robust phase III. The current study is not enough and cannot be enough. What is your strategy to tackle this? Well, hello. Let's move forward on GCP. Slide. It depends what you mean by robust, but the fact behind robust, there are everything. No, GCP to the slide, I think it's seven. Slide seven. So, it depends what you mean by robust. It could be robust statistics, it's stats questions, but we can start by robust in terms of GCP, for instance. So, we have to convince the agency that the study is robust enough, first in terms of conduct of the study, and so GCP issues. As you have seen in the guideline, we have to go step by step through all the steps. At that time of the process, they still have concerns. Okay? However, according to us, LSFS is robust. They themselves, EMA, agreed on that, and the safety is acceptable. So why it's not considered robust enough, we have to discuss with the agencies. It's a gray zone. It's a matter of judgment. Just for you to know, the Amylyx, for instance, study that was approved by FDA, had a greater number of deviations to the protocol than our study, but they had three times less patients. So in fact, by patients, we have three times fewer deviations. So it's not that it's unusual to have deviations to the protocol. Still, we need to convince the agencies that it's robust enough. So we are going to discuss with them to see what's the point behind that conclusion from the EMA. Maybe one last question: How long does the re-examination process last? Hello. Let's go to the slide on reexamination. But the reexamination, yeah, thank you. So the reexamination takes. We have to wait for the vote first, then a reexamination takes 4-6 months. So we can expect that. We have to prepare the dossiers, but then it takes 4-6 months. So we can expect an answer by the end of the year, if we go to reexamination. If you had the support of a large pharma partner, could that impact positively the process of acceptance of masitinib from the EMA? That's a frequently asked questions. The data are the data. Now, you have seen in the field of amyotrophic lateral sclerosis, that there are not Big Pharma. Biogen is the biggest, but still not... It's probably the smallest among the big. And they have registered tofersen Qalsody, in a very specific population, which is only 2% of the patients. The patients bear the SOD1 mutations. Big Pharma are on there, first of all, it's a rather an innovation. Innovation companies like us, who try, in one of the most complex diseases of the world, to bring a benefit. So it's first of all, we have to start with what exists. It's, you know, small companies or, or, Biogen, which is the biggest, but not the big one. Now, if a big one would come, would it change the decisions? It's unknown, of course. It's unknown. It will not change the data. Would it change something from the perception of the CHMP? I cannot answer this question. I would like to take, maybe, one last sentence, which is not a question, so I will read it. This is not a question, but a comment which I would like to be addressed also in the report about the presentation of this evening. Each of us was investing in AB Science because we trust the research being done since years. I would like investors to have not a purely short-term financial behaving, but to be an actor of medical progress. Let's be a partner of medical progress tomorrow, and let's not think in terms of financial profitability and behaving. Let's AB Science support for medical progress, and let each of us be part of this challenge and trust. I thank the person who did this comment. Of course, it's comforting when we have not the expected good news. I would like to conclude maybe on the benefit that we showed here, or yes, on the other table is the same, but you can leave it here. So, what can I say? The fight against ALS is a crusade, and it has been 30 years that everybody tries and fail. And 30 years ago, there was riluzole, and 30 years after, there is still riluzole. And riluzole is considered by most of the people as even not very effective or not effective at all. So it is 30 years of failure. Now, we have the chance to have a promising, compound with a very good mechanism of actions, with data which are what they are. They might have limitations. We need an amendment. We need the fast progressor. You can find the limitations you want to the study, but it's promising. Now, I understand the frustration, and I'm the very first one to be, the most frustrated among you, that it, it's hard not to be recognized by the agencies, in a early approval process when we see people, dying like that. But as the, way the industry is structured, we have to accept it. Okay? It's tough, but it's- that's it. So, okay, it's not, a good news, but do you have something better on the market? No. Can something come with a solution to ALS, which brings exceptional survival and an improvement to the people? No. So we have to be realistic. It's very tough. Some indications, some diseases are extremely tough. AB Science has decided, through the mechanism of action of its compound, masitinib, to try to make a progress, a medical progress in those indications. And because it's extremely difficult, because people fail, because it's extremely risky, and because the agencies are a little cautious, because they have seen so many failure, we pay the cost of the innovation today. And you can conclude that, it's unfair, or that we, as managers, are not good. But it's very difficult, but it's the way it is, and we are going to continue. It's not because it's difficult that we're going to drop. We think the compound is good. You have seen the objections, and that's why I share that with you. It's technical, and it's not other than that. And we can maybe convince them, but it takes some time, it takes different cycles. Keep in mind, Amylyx succeeded to convince FDA after three cycles. First one, they failed. They failed. They failed, and exceptionally, FDA gave them a third cycle, because at the request of probably I don't know, the lobbying or whatever, but they succeeded in the third cycle. We are here at the end of the first cycle for EMA, and the beginning of the second cycle for Canada. It's not successful so far, and it's not a reason to conclude the data are bad. It's not a reason to conclude that we should drop the development. We need to stay cold and analyze the data. It's difficult, but we have to continue. And I thank you for your support, and keep you informed about the next development, of course, of our compound. Thank you for this, for your attention and time, for this very difficult and complex presentation. And we'll close there.