AB Science S.A. (AB.PA) Earnings Call Transcript & Summary

Ladies and gentlemen, welcome to the AB Science web conference. I will now hand over to Alain Moussy, Jean-Pierre Kinet and Olivier Hermine. Jean Kinet, please go ahead.

Thank you for the introduction. So welcome to the AB Science web conference where we're going to talk about masitinib in pancreatic cancer based on the recent clinical study results. So good morning, good afternoon or good evening for all our participants from the United States and Europe. Along with me, there will be -- and Lauren can go -- Lauren is helping me with the slide show. You can go on Slide 3. So along with me, there will be 2 KOLs. Dr. Julien Taieb, Head of the Gastrology department at hospital Georges Pompidou, one of the largest hospitals in cancer in France, and specializing in pancreatic cancer. And Olivier Hermine. Professor Olivier Hermine with the President for our Scientific Committee of AB Science and is Professor of Hematology at Hospital Necker and member of the Academy of Science. And they are there to present and also respond to your questions at the end of the presentation. Next slide. So I'm going to hand over to Olivier for the scientific rationale. Olivier, we are on Slide 4.

Okay. So hello, everybody. And I would try to explain to you in a few slides what is the rationale on the effectiveness of masitinib in pancreatic cancer. So we know that in tumors, there is some kind of a secretion by immune cells in the tumor and particularly mast cells and macrophages, which are critical components of the tumor microenvironment, which may promote androgenesis, tumor growth, and also contribute to the tumor genesis by suppression of the immune response against the tumor. So in this [indiscernible] microbiome mast cells might be antigen as cells, which may orchestrate the tumor microenvironment and specifically in pancreatic the cancer because they may also promote some code factor to end user proliferation and the adhesion and the metastasis of tumor pancreatic cells. And this [indiscernible] between mast cells and pancreatic cells may contribute to the progression of pancreatic ductal adenocarcinoma. And mast cells may contribute, in this regard, to the aggressiveness of the pancreatic tumors because they produce pro angiogenic factors that has been shown in various publications. And by releasing of pro tumoral factors, also mast cells may promote the progression of pancreatic cells. And it has been shown that the number of mast cells is a prognosis factor of post survival in pancreatic cancer. In addition, mast cells may produce some cytokines, which may skew the polarization of tumor-associated macrophages. And we know that the M1 are of good prognosis by preventing the progression of the tumor and the M2 is of bad prognosis by promoting the angiogenesis and the progression of the tumor. And mast cells is able to skew this polarization towards the M2 bad prognosis macrophages. And we know that masitinib is a highly selective inhibitor of mast cells survival and activation, and by this way, may modulate mast cell-related commonly of the tumor microenvironment thereby inhibiting the tumor growth and also redirect the immune system toward the anti-TS1 M1 phenotype, which prevents the progression of the tumor. So one thing, which is very important to know is that in pancreatic cancer, we know there are always and often associated with pain. And pain is, for us, a flag of increased mast cell activity within the tumor microenvironment, which in turn, as we previously said, may induce the progression of the tumor. So pain might be seen as a prognosis factor reflecting the mast cell activation and the pro tumoral immune response. And it has been shown that mast cell degranulation mediates the cancer-induced pain and that this pain is a clinical predictor of poor prognosis in pancreatic cancer. We know that mast cell infiltration is strongly implicated with development of neuropathic pain in pancreatic cancer, and mast cells contribute to the hypersensitivity and degranulation as a vicinity of the tumors. So by this way, mast cell by producing proteases, may induce cancer pain, which is known, and also may produce some factors, which may remodel the neuronal filtration in pancreatic tumor microenvironment. So it has been shown in regard to this physiopathologic mechanisms that pain is associated with increased tumor recurrence and decreased overall survival. So we can make the hypothesis that pain [indiscernible] correlate to disease progression and significantly of poor survival rate in pancreatic cancer. So to -- in order to block this muscle activation in the hope to reduce the pain and to reduce the tumor progression, and we can see pain as a surrogate marker of mast cell activity, we have used and developed a kinase inhibitor, which name is masitinib, which is able to block mast cell activation. And masitinib is a very peculiar kinase inhibitor, different from the other one in the market, because it is a potent selective inhibitor of c-Kit, which block mast cell proliferation, differentiation and activation. But in addition, block 2 Src kinases, LYN and FYN, which are very critical for mast cell degranulation and mast cell proliferation. So it has been shown in previous work that masitinib is able to block mast cell activation and which may play a role in inflammatory disease and also in oncology. But in addition, we have shown that because masitinib block MCSFR-1 receptor, it's able to block macrophages skewing to the M1 phenotype. So here, we have a dual activity, blocking mast cells activation and blocking the skewing of macrophages. So we hope that we can block the 2 mechanisms that we have described earlier regarding the pain and mast cell activation. So on the Slide 7, we have shown that masitinib is not able to block various cell lines, tumoral cell lines, ex vivo. So the effect of masitinib is not to block directly the cell proliferation in various model of epithelial cancer. But very surprising and interestingly, in accordance to hypothesis, in vivo in mice model, when we inject this tumor in mice, and if we do treat this mice with masitinib, you see a dramatic reduction of tumor volume, which is very important and very unique to masitinib. And we have shown that this decrease of cell proliferation postulate in increase of overall survival. And to make a long story short, we have shown that this tumor, we do see no more mast cells, and we see M1 microenvironment upon treatment by masitinib, which explains and demonstrates what we have hypothesized before. So based on this data, we have moved to clinical development of masitinib in pancreatic cancer.

Thank you, Olivier. So now we move to Slide 8, which is part of the clinical development plan of masitinib in pancreatic cancer. So it's made of 3 clinical studies. The first one is called AB05034 is a Phase IIa with 22 patients. It was an introduction for safety, essentially of our company in pancreatic cancer. Then we engaged into a Phase III, AB07012 at 9-milligram/kg/day of masitinib in combination with gem. We'll come back to it just to refresh the results of the clinical study. And then this new study, AB12005, which is the second Phase III program at a lower dose of 6-milligram/kg/day. Next slide is the data of the first Phase III study, so 7012, which was hypothesis-generating. In this study, we could not demonstrate -- which recruited all patients suffering from pancreatic cancer, we did not see any significant increase in survival for masitinib. Again, the design was masitinib plus gemcitabine versus placebo plus gemcitabine, gemcitabine being the standard of care with 348 patients. So as you can see, no advantage in survival, 7.0 million survival for the placebo for gemcitabine alone and 7.7 with masitinib. However, in a subgroup called the patients with pain, which actually match the scientific rationale described by Olivier, you can see that, first, the placebo group has reduced survival, the median of 5.4 months instead of 7.0, so a reduction of 1.6 months median, and masitinib reversed this poor prognosis and generated a median of 8.0 months. And so this is an additional 2.6 in the subgroup of pain. So the hypothesis was consistent with the scientific rationale. And we thought that masitinib could be used in pancreatic cancer with pain. So we engaged into a next study. Next slide, please, which is the design of study AB12005, where we recruited patients only with pain, consistent with the previous study 7012 with -- at a lower dose because we estimated that the first of 9 milligrams was a little bit too high and created some discontinuation. So the design is patients with pain, 377 patients. The primary endpoint is, of course, survival. And the main inclusion criteria is pain or the use of opioid because it's like also pain in the first line, again, with gemcitabine for everybody and then masitinib versus placebo randomized 221. Next slide is the statistical analysis plan. So we actually have tested the overall population with pain and also a subgroup, which is made of the locally advanced patients, unresectable, which is a subgroup of all patients. There are, on the one hand, the metastatic patients; on the other hand, the ones who are unresectable but are not metastasized yet. And it was important to distinguish those ones because, of course, we consider masitinib as an antimetastatic drug, and so maybe with more efficacy in patients who have not metastasized yet. So it was important to take the 2 together. We had so to speed the alpha, and the threshold was 2.5% and 2.5%. The traditional factors are based on this slide, and we analyzed the population on the MICT, which consists of the ones we have taken at least one treatment and has the necessary pain. It excludes 4 patients who have no impact in the overall analysis if we add them back. So you can see the split of the population, 383 patients and 92 in locally advanced. Next slide is the baseline characteristic. I'm not going to spend time on this one, just to notice that everything is balanced, including the variables that could impact the survival like, [indiscernible] for instance. Next slide is the result of the primary analysis. So we're on Slide 13. And we observed no benefit in the population -- the overall population with pain, which is not so a confirmation of the hypothesis we had in the first study. However, we can see that the study demonstrated a significant increase in survival in the subgroup, which was prespecified, the unresectable locally advanced. So in the unresectable locally advanced, we can see that the p-value is p is equal to 0.0070, so less than 1%. And additional survival median of 1.8 months, and the reduction of the risk is called the hazard ratio of 55 -- 54% of the risk of death with median survival being 11.2 months with masitinib -- sorry, with placebo and 13 months with masitinib. You can see, by the way, in the overall population, that the median survival with placebo is 8.0 months, which is much higher than in the first study where it was 5.4%, but masitinib was consistent between the 2 states. Next slide. Next slide is the survival, but this time, described as survival rate every 6 months. And as you can see, masitinib in the locally advanced resectable populations. As you can see, masitinib increased the survival at each time point, and especially in year 2, where people can survive. And you can see, for instance, at 18 months, around 10% of placebo-treated patients who survive only whereas it's 34% with masitinib, a relative risk ratio of improvement of 3.4, which is quite interesting to note. And in 2 years, it's only 3% of the patients surviving with placebo but 14 -- 15% with masitinib. So there are some patients who are able to survive long term due to masitinib. Next slide. Next slide is the secondary endpoint. So it's important to check consistency. Here, we have a consistency with the PFS, which measures the carrier of death or progression of the tumor, so no impact in the overall population but an impact in the unresectable locally advanced, which is statistically significant with an additional 1.8 months, which is completely consistent with the 1.8 months additional survival that we have seen in the primary analysis. The p-value is significant and the hazard ratio, so the reduction of the risk of progression or death, is also completely consistent because it's 53%. It was 54% in survival. Next slide. Next slide is the response rate. It's important to note we did not expect actually masitinib to change anything on the response rate versus gemcitabine alone because, as Olivier explained, it's an indirect effect. However, we can see in this study that even in the overall population, there is an increase in response rate, 8.2% partial response or complete response versus 5.9%, including a complete response with masitinib in metastatic patients. And in unresectable locally advanced, even better, because placebo is doing 3% of response rate, whereas the masitinib is doing 15 -- in the range of 15%, which is interesting. So masitinib has not only an affect on survival, also in PFS and also on response rate. Next slide. The response rate is important for one thing, by the way, because some people -- some patients will be able to be receptive. And it's important to have this option to be receptive, although they were unresectable at deadline. So the next slide is on pain. It's important to check the hypothesis that masitinib acts on pain by reducing pain. And you can see that the orange line is the placebo, the blue line is the masitinib. So the blue line is always below that in changing pain in the orange line, including in the overall population. So there is an impact in pain as compared to the placebo. And in unresectable locally advanced, you can see it is even a reversal. There is a reduction of pain, where there's still an increase of pain with the placebo. And it is almost significant at all time points. So the hypothesis was correct. Masitinib act by reducing pain, acts on pain, generating tumor mast cells and generating benefits in PFS in response rate in PFS in consistently. Then the next slide is on the safety. So as you remember, we reduced the dose of masitinib for 9-milligram/kg/day in the first study to 6 milligrams, and it hasn't an impact because the safety is, I would say, more than acceptable, actually almost surprisingly good because you can see that there are a little fewer case. But in terms of [ ATL ], the fact that is due to the disease proliferation, there is no difference. But there is a reduction of the nonfatal serious adverse events, 2% or 3%, which represents in relative risk 10%. And the CVA reduced 75% with masitinib versus 83% with the placebos of gem alone. So as the masitinib to gemcitabine, actually, leads to a reduction of severe and serious adverse events, which is quite rare because it's an add-on therapy, so we expect more but never less. But here, in this case, it's less. So the safety profile is favorable. More discontinuation but it's essentially in the metastatic patients. And as you can see, fewer temporary interruptions and even fewer dose reduction, which is good. So the safety profile looks correct. Then we enter into the discussion on Slide 19. So the discussion is on efficacy. We have to check that the survival additional benefit, significant is not due to, let's say, in line 2 after discontinuation of treatment. This balance in chemotherapy taken by the patient. So it's very unlikely why because the benefit is done actually in first line in PFS. The 1.8 months of benefit survival is actually consistent with the 1.8 months of benefit in PFS. Also, we have to check that the placebo has not done surprisingly lower than expected. When you look at the literature in locally advanced pancreatic cancer, you can see the publications all range with a median with gemcitabine ranging from 9 months to 13 months but with an average of 11 months, which is exactly what we have done in our study, what we observed in the study, except that in the published study, patients have 10 or no pain, without any distinctions. And we know pain is a negative prognosis factor. So we could expect it actually and even lower. So a median of 11 months of survival with placebo, in looking at advanced pancreatic cancer resectable is very acceptable. And so we think that this data of survival is robust. Next slide. Next slide is the positioning of masitinib as compared to what exists in pancreatic cancer locally advanced. So first difference is our product is targeted for patients with pain, which is completely new because there are no other products that target patients with pain. Remember pain is a negative prognostic factor. And also, usually, you do not decide really to do surgery in patients with pain. The second differentiating factor is that it is only for unresectable locally advanced pancreatic cancer patients. So we can know that other products that are used like Abraxane has no level at this time in locally advanced pancreatic cancer. It's for patients with metastatic. And FOLFIRINOX is recommend also for metastatic pancreatic cancer. We know, of course, they are used by the practitioners. However, regulatory-wise, they do not have a label. And regulatory-wise, gemcitabine is the product of preference. And this study has compared masitinib with gemcitabine, so with the standard of care regulatory correct. Then positioning versus FOLFIRINOX, what can we say? FOLFIRINOX is not eligible for everybody. It's only [ Eco 0 and 1 ] with few Eco 2 in our studies in locally advanced pancreatic cancer. But the biggest difference is the aged people. FOLFIRINOX is not used with aged people usually starting at 70-year-old. And we have patients in the order of 15%, 20%, which has participated to our clinical studies. And so those patients cannot benefit from FOLFIRINOX and would be ideally suited for masitinib. Versus Abraxane, what we can say is that Abraxane is clearly a product which is used in the United States but is infrequently used in Europe because it's not reimbursed. It's registered but not reimbursed. So difficult for most of the hospitals to use Abraxane, which gives a differentiating position in masitinib if we can get the price, of course. And the safety profile is different. Why? But you have seen it's very acceptable with gemcitabine by reducing and severe. And masitinib is not a chemotherapy like Abraxane or FOLFIRINOX and so do not generate the same side effects. For instance, with chemotherapy, you can expect hemato-tox, peripheral neuropathy, sometimes irreversible, alopecia because it is, et cetera. And here, we propose a targeted tyrosine kinase inhibitor with different safety profile. Next slide. Next slide is the next step. So the next step is to talk to the head authorities, both EMEA and FDA. And we have saw some key and strong points in this study and in the clinical development program. And we have one limitation, which is this limited number of patients with 92 patients in the locally advanced. This could be maybe mitigated when discussing with the authorities with the following arguments. First, this is not an exposed data. These 92 patients locally advanced unresectable come from a prospective study with a prespecified statistical claim, and this is regulatory correct. There is a strong p-value, less than 1% on the primary analysis. This population is significantly relevant because we have a reduction of 54% of the risk of death, completely consistent with the other secondary endpoint. And we have -- we still have to know that locally advanced unresectable pancreatic cancer is still one of the worst cancers. It's not only the metastatic. The locally advanced unresectable is considered as the third worst cancer in terms of prognosis. So there is a high unmet medical need and medicine benefit from [ medical perspective ]. So those arguments might be used to circumvent the limited number of patients. Also, there is another argument. When you look at the publications, there is no publications with more than this number of patients in the literature. And so this study is not small. Actually, it's the biggest number of patients that we have been able to identify in the literature not in cost. And the safety is not a problem because we have, as you know, a huge safety database in oncology that goes beyond these 92 patients, of course. And the benefit is for us is favorable. So this will be the arguments we will plead, so to speak, to the EMEA and FDA to see if we can file and when we're looking to look for recommendations of filing with this data in the 3 studies. The next slide is just the market potential. And so the market potential is what, which has to start with the located advanced pancreatic cancer, which represents 35% roughly of the pancreatic cancer populations and pain, which represents roughly 50%. Although patients -- most of the patients will experience pain at one point in the course of their disease. So according to our calculations, that represents 12,000 patients in the U.S.A. It will be the locally advanced pancreatic cancer resectable with pain. And 16,500 in the EU, so [ front ] of 30,000, and we know the price of the compound that has been registered recently. So that ends the presentation. And the next part of the talk is to receive your questions and answer them. And with me, of course, there will be Olivier Hermine and Julien Taieb.

[Operator Instructions] Okay. I will take the first question in writing. So do you intend to evaluate other cancers without metastatic profile?

Okay, we can respond in 2 parts. Are we exploring today? And will we in the future? So today, we have 2 cancers in the program of masitinib. The first one is pancreatic cancer. So you have seen it. And the second one is prostrate metastatic cancer. So all patients are already metastasized. Still, we can maybe differentiate between early metastatic versus more advanced metastasis. Now to your questions, do these data give us ground -- scientific ground to explore.

Hello.

Yes.

Yes. It's Julien Taieb.

Julien, thank you to join us. So you have been introduced, of course, as our KOL. And we are just starting the Q&A. And the first question was on this basis, do we intend to explore nonmetastatic other cancers. And so I was saying that the clinical program is pancreatic and metastatic prostate, metastatic prostate. We don't have the data. We will lay it in Q1 in 2021. But in the future, with this data, do we want to try other cancers in, let's say, before tumors become metastatic or in adjuvant of any cancer. So I will leave it to you, Olivier and to you, Julien, to see if it gives you ground to use masitinib in other cancers.

I'm Olivier. Can you hear me?

Yes.

Yes.

So what I should say is that what we have seen and based on your mice model disease is that masitinib play a critical role in the tumor microenvironment to prevent progression of the tumor. Obviously, not all tumors will depend on macrophages reduction and mast cells infiltration. So this will make a of treatment, which is driven by the biology of the tumors. We can extract this prospect, not only in pancreatic cancer but in all tumors in which mast cells and macrophages in patients has been shown to a critical role in the progression of the tumors. And if you look at the literature, there's a lot of tumors in which we know that the tumor microenvironment, which includes mast cells and macrophages play a role in the progression of the disease. So here, we have shown mainly a general mechanism of action. And most interestingly, at the scientific view is that we can reduce also, as we have seen, the pain of the patients, which is very important in terms of clinical observation. We've proved what we have hypothesized in our mice model.

What I can add -- it's Julien, is only that we don't have so many -- I am only treating GI tumors. And that there are not so many GI tumor-inducing pain, but pancreatic cancer is especially worrying for that. We can have locally advanced rectal cancer that may be also a target because it's bad located and then has many symptoms, including pain, and some kind of gastric tumors also, but not for the other GI tumors, I think. We don't have these inflammatory reaction, the reaction and this infiltration due to inflammation.

I will take a similar question -- I mean another question on similar topic. Does it imply that our clinical study in prostate cancer is less favorable because it includes patients with metastasis?

Olivier, you want to answer?

Can you repeat the question and...

The question is does it imply that the study in pancreatic cancer, because it's -- in prostate cancer because it includes only metastatic patients, is likely to be negative, so to speak, because we show here that the metastatic patients suffering from pancreatic cancer is negative. So again, we will see when we unblind and we have to wait until Q1 2021. But in pancreatic -- sorry, in prostate cancer, we can differentiate maybe between early metastasis and late metastasis. So it seems that masitinib might be -- might generate benefit for, I would say, patients with high risk to develop metastasis, like, for instance, the unresectable locally advanced pancreatic cancer or in early stage of metastasis and where medicine is not good, is in advanced stage. The problem in pathetic cancer, I will leave it to Julien, he is the expert, is that when you are diagnosed in pancreatic cancer, patients who are diagnosed in metastatic, it's considered as already advanced because we take it too late. Julien, if you want to develop there.

Yes. I think inflammation and pain is also present in metastatic cancer. It's mainly due to bone metastasis that are very frequent in the disease, but it may not be too late. I think we have to look at the results because it may induce quite a substantial better quality of life for the patient, even in prostate cancer and even if I am not an expert of prostrate cancer.

Going back to pancreatic cancer, what is the percentage of pancreatic cancer without metastatic profile?

So this is quite clear. The data are quite robust. So we know that pancreatic cancer, first, is increasing very rapidly. It's going to be probably the first cause of cancer death very soon. In France, it was 5,000 cases in 2005 and 18,000 cases per year in 2018 and is still growing. So first. Second, we know that approximately 16% is in our initially with no metastasis. 25% are borderline resectable or locally advanced. And we have on the top of that the metastatic lesion. So I would say that for the locally advanced setting, including some broad line that it's not too easy to resect. We may reach at least 20%, 25% of the population.

What are the next steps for this indication in locally advanced pancreatic cancer?

And as we said, as far as we sponsor our concern, we are going to discuss with EMA and FDA because we have some regulatory strong arguments. In parallel, we are going with Julien, with others to share this data with the scientific community and discuss because regulatory is one thing. That's not the only thing. And of course, trying to publish and go to scientific conferences.

And when do we expect a response from this health initiative?

I would say, take 6 months usually.

And what would be the main reason agency should not accept to register based on this study?

Well, one of the reasons, as we mentioned, is the limited number of patients. If we had 300 or 400 patients in locally advanced pancreatic cancer, then it will be stronger, of course, to present. Know that we have 92 patients. That's a limitation. There are other drugs in other cancers within registers with fewer patients than that also will get endpoints like the response rate. Here, it's survival, but it's only 92 patients. So we don't know, okay? We have to discuss with the FDA and with the EMA to know.

I will take one last question in writing and then maybe we can take some questions orally. Masitinib is being tested in multiple indications, oncology and non-oncology indications. How are you managing resources?

Well, managing resources is human resources most probably. No, but in terms of organization, we were waiting for reading the Phase III, the late stage to see how we organize AB Science. If the studies are rendered negative, then we could specialize in one, let's say, clear indications. But we have positive, multiple positive Phase IIb/III. And so it seems that we have 4 therapeutic areas in AB Science. The first one is neurology, with neurodegenerative disease, 3 of them, ILS, MS and Alzheimer, still to be known because we don't have the data yet. And then the second one is inflammatory disease with mastocytosis and asthma, positive here already. And we have cancer. Cancer was a question mark. So this study opens the door to oncology. And we are waiting for prostate. So we keep now oncology. And the fourth one, as you know, is infectious disease with COVID, which is completely opportunistic. But we have an antiviral with masitinib. So we intend to organize by targeted area, okay, more and more. And we'll start the necessary resources through human and financial to do the necessary next clinical studies if EMA ask us to do. Julien, I know that you just confirmed because you had some other commitments. Before you leave, I would like to give you a quick review to say whatever you want to say about this data.

I think that locally advanced disease is quite challenging because, of course, for the patient, it's to go to resection and to be improved in overall survival is quite difficult. There is still a hope for cure in some of these patients. So it's a very challenging population. We're not only controlling symptoms and the prolonged survival, but also bring back some patients to surgery for some of them or to closure radiotherapy that may result in a long-lasting control of the disease. So I think it's difficult to move forward in this population. Of course, some chemotherapeutics regime even more intensive are also tested, as you probably mentioned. But remember that these patients are not always in very, very good condition. So having also a treatment that improved gemcitabine that is generally well tolerated is valuable. So I think that these results are quite interesting. You remind everyone that the numbers were a bit low, but this is still with quite significant difference. So it's a very good result. And we don't have -- we haven't seen that since a long time. The last big trial published in [indiscernible] tried to add radiotherapy systematically after chemotherapy because we know that starting with chemotherapy is not a good option from the 2000. And it shows that radiotherapy in all patients was not so necessary or not improve significantly. So I think it makes a long time, we haven't seen anything improve the outcome of these patients.

Thank you, Julien, for your help in developing masitinib, and we hope to do a nice publication. Melanie, do we have any overall questions at this time?

For the moment, we have no question by phone. [Operator Instructions] The first question comes from [ Keane Mcquery ] from Soc Gen.

In the U.S., with the availability of Abraxane, how would you view the slotting of masitinib?

So I can still answer that. I'm still here, if you want. I would say that Abraxane is certainly not validated in locally advanced. We have only -- the best evidence is only a Phase II trial by select studies that was published this year, so a few months ago, showing quite interesting results, but we still don't have any randomized data. That will probably come in, in 2 to 3 years. So for now, gem remain at the standard of care. And in Europe, FOLFIRINOX is also developed. But we will not have the result of different controlled trial before a few years from now.

So for Dr. Julien, if masitinib is approved and available, how would you use it then in your locally advanced patients? Would you -- would that become your next drug of choice?

Yes. Probably. Yes. Yes. Probably. I think we will have 3 situations. First, we stick to the regulatory registration that is more or less obliged in many countries, but sometimes you could go a bit on narrow path. But if you do that, yes, it will be if it's registered the tenders. If you are still wanting to do more by using upfront Abraxane or Folfirinox in these patients, I would say once again that not all the patients are eligible for more aggressive treatment, and masitinib presented I think is quite well tolerated and not so intensive as compared to gem, Abraxane or FOLFIRINOX. And we will still have, I think, a significant subset of patients that will benefit from this and that will be eligible for this kind of intermediate aggressive treatment. Okay. I have to leave you to go back to my ESMO teaching advance course. So Alain, don't hesitate to ask a question by later on, and I will answer.

Julien, thank you very much. I take another question in writing. Is compassionate use possible in this indication?

Yes. So compassionate use is in the framework, depending on the countries. But we like to serve the scientific community. So if we can be helpful, we are. We have compassionate use. We have -- masitinib is already used, is compassionate use in many indications, not something we publicize naturally. But yes, it's possible.