Nanobiotix S.A. (NANO) Earnings Call Transcript & Summary

Good day, and thank you for standing by. Welcome to the Nanobiotix Study 102 program update conference call. [Operator Instructions] Please be advised that today's conference is being recorded. At this point, I'll turn the call over to Craig West, Senior Vice President of Investor Relations of Nanobiotix.

Thank you, operator. And moving to Slide 2. Good afternoon and good morning, and welcome to the Nanobiotix Conference Call to discuss the final results of our Phase I 102 study in locally advanced head and neck cancer. Joining me on the call today are Laurent Levy, Co-Founder and Chief Executive Officer; and Bart Van Rhijn, Chief Financial Officer. As a reminder, today's call is being webcast and will be available on our website for replay. I would like to remind you that this call will include forward-looking statements, which may include statements regarding the progress, success and timing of our ongoing and planned clinical trials, collaborations, regulatory filings, date of presentation and future research and development efforts, among other things. These forward-looking statements are based on current information, assumptions and expectations that are subject to change. They are subject to significant risks and uncertainties that could cause the company's actual results to differ materially from our current expectations. Accordingly, you are cautioned not to place undue reliance on forward-looking statements. Please review the full description of risk factors that can be found in the documents we filed with the AMF in France and the SEC in the United States including the RFS and the 6-K that we filed with our half year report on September 26, both of which are available in the Investor Relations section of our website, along with the press release we issued yesterday highlighting the results that we're about to talk to you about now. In addition, any forward-looking statements represent our views only as of today and should not be relied upon as representing our views as of any subsequent date. While we may elect to update these forward-looking statements at some point in the future, Nanobiotix undertakes no obligation to update them to reflect subsequent events or future circumstances. With that said, I'd like to turn the call over to Laurent. Laurent please go ahead.

Thank you, Craig. Before starting and presenting the latest data, we've been releasing on the 102 final expansion phase. I just wanted to remind for the new people around the table, what is our product. So NBTXR3 is a first-in-class radioenhancer. It is made of inorganic crystalline nanoparticle made of a very specific material hafnium oxide. And we have chosen hafnium oxide because this material contain high atomic number electron, the hafnium. And this is this unique property that make those products and those particles able to absorb x-ray. So this product is made to be injected once, only 1 administration directly into the tumor. It is a product that already shown a positive randomized Phase III in locally advanced soft tissue sarcoma. And the manufacturing of this product is purely physical. You absorb the energy of radiation and you deliver a much more damage into the cell. Moving to Slide 4. That's why what you see here on this slide is the pipeline of development we have around NBTXR3. We obviously have a focus in head and neck cancer, and we'll come back to that in a moment, but also have completed or have many other clinical study ongoing. On the top of what is presented on this slide, you may be aware that we have been recently signed an agreement with Janssen that we will expand the use of this product and also will encompass commercialization of this product. And as an additional indication that we'll start under this collaboration, Janssen will be responsible to start a new one in lung cancer Stage III cancer patients. But let's focus on the head and neck for now. Slide 5. When we look at radiation therapy usage in head and neck cancer patient, it is the vast majority of this patient that will receive radiation, around 70 or more percent of patients will get that. And when you have a locally advanced head and neck cancer patient, radiation is often used in combination with chemotherapy. Unfortunately, there are many patients roughly observed that are ineligible to cisplatin and that brings some challenges. Because to get a good local control, usually, you need to have this radiation in combination with cisplatin. It is important to note also that for a good number of locally advanced head and neck cancer patient, there is a large number of comorbidities associated with those patients. In the overall population, it's 20% to 30% of patients that have a high ACCI, which is usually correlated with lower overall survival. When you not talk about elderly patient, there is a big issue here. Most of them being not eligible to cisplatin, you see that it is very hard for them to get some treatment that will be efficient. And usually, for the 30% of head and neck cancer patient being 70 years old or more, they have very poor outcome. In this general population, usually observe a PFS of 9 months and a median overall survival of 12 months. So definitely, there is a large unmet medical need in this patient population. This, on Slide 6 are the literature references we've been using to establish some kind of historical control, but also to establish the baseline of what is the design of the ongoing Phase III in a similar population. It is interesting to note, as an example, for the first paper referenced here by Amini that you get a median overall survival of 12 to 13 months for radiation therapy alone. And for a population that have not that much comorbidity associated to baseline, there is 70% of the patient with modified Charlson Comorbidity Index that is less than 4. On Slide 7, you see the study design of the Phase I. We started with an escalation part that was completed and on 19 patients did not show any DLT. This has been already published. Today, we're talking about the dose expansion that have been done at the recommended Phase II dose that have been recruiting 56 patients. The primary endpoint of this expansion phase was efficacy, overall response rate of the primary tumor, meaning the injected lesion, overall response rate of the injected lesion and non-injected lesion, duration of response, PFS, OS and of course, safety. It is important to note, and it's on Slide 8 that in this particular trial, we've been injecting the primary lesion only, meaning for patient that had invasion of lymph node, there have been no specific injection of nanoparticle here. Nevertheless, both primary lesion and involved lymph node have received 70 gray of radiation, which is the dose given as the standard of care for those patients. Moving to Slide 9, you see the baseline characteristic that really show frail and high comorbidity patient. Of note, we've been running this trial expansion part and have been recruiting 56 patients across 20 sites in 4 different European countries. The importance seems to be noted is 61% of the patients got more than 70 years old and there's been a very high comorbidity in those patients. 67% had an ACCI score of more than 4 or equal to 4. And we've had a small number of p16 positive patients, 26 over 56 patients. In summary, on Slide this, you see that the results we've been finding in terms of safety and feasibility are very consistent with what has been observed in the escalation part of this trial. The number of adverse event represented by NBTXR3, where 1.3% of all adverse event. We've been showing a very good feasibility of injection and 91% of the patients have been able to complete their radiotherapy total dose. It is important to note, and that's presented on Slide 11, while some of the patients haven't been evaluable. And this is for a good number of patients associated with high comorbidity at baseline. To be available, you need to get the dose of NBTXR3 and to have completed the radiotherapy session and having a baseline scan and post treatment scan. For a good number of patients, like 10 out of the 12 nonevaluable, we have been observing this before the end of the treatment or within the first 180 days. So it is important to see that 8 out of 10 patients had severe morbidities, meaning ACCI more or equal than 4. Moving to Slide 12. So here, we see the final results of the response and local regional control of this product in this patient population. So in the variable population, when we look at the response at the injected lesion level, we get a very high overall rate of response, 81.8%, including 63.6% of complete response, which is a very high number if we compare that to literature. Same when we look at the overall response rate that will include injected and non-injected lesion, we get to 79.5% overall response rate. Now if we move to Slide 14, that's a new observation we've been doing in this trial, which is very important. You remind that in the introduction, I did mention that we were irradiated both primary tumor and surrounding in both lymph node with 70 grays, but only the primary tumor has been injected with the product. What we've been observing for the injected lesion is that the median duration response has not been reached. Whereas the overall duration of objective response is 12.4 months, but clearly see that most of the relapse that have been observed in this trial are coming from the surrounding lymph node and not the primary tumor. And that's a very surprising and interesting finding because if you go to Slide 15, then you can see that in head and neck with locally advanced cancer patients, usually, the pattern of failure is different. Most of the relapse will come from local treatment failure, meaning the primary tumor and much less coming from the surrounding lymph node. So the fact we've been able to observe a very different pattern failure may be a good hint about the efficacy of this product in the injected lesion. Moving to Slide 16. We're talking now about the PFS and that's also last evaluation, which is more mature than the one that have been done previously. And we've been reaching in the all treated population, 11.4; and in the evaluable population, 16.9. Both numbers are much higher than what you could expect when you look at the literature in better population, meaning in a general elderly population of head and neck cancer patients receiving radiation alone. Same for OS. When we look at the all treated population, we got to 18.1 months. And when we go for the evaluable population, we have been reaching 23.1. In a nutshell, if you go to Slide 18, you see the Study 102 key messages. First of all, in a very frail population being able to have a new treatment without large safety issue, it's very interesting and very important. And that's a reminder for you, one of the reasons why this patient cannot get better treated because of all the comorbidity and all the issues they may have taking cisplatin or some other chemotherapy in combination with radiation. Safety apart, we've been observing a very high overall response rate, both from a global perspective, locational and also from a local perspective. The median duration of objective response was higher in NBTXR3 after injected lesion. That's, for us, a very important finding, especially when we're going to talk about the design of the 312. The mPFS was 16.9 months and mOS 23.1 months in the available population. And this is longer mPFS and mOS compared with historical data that I mentioned in the introduction that is for the PFS 9 months and for the OS 12 months. And that gives us a lot of confidence about the ongoing Phase III in the similar population. But now let's talk about the Phase III and what's the design and how those findings can help us to understand or anticipate or estimate the outcome of this Phase III. Moving to Slide 19. You see the design of the NANORAY-312, which is a global Phase III registration trial for locally advanced head and neck cancer patient. It is a trial that will recruit 500 patients, randomized 1:1. The control arm will be about radiation alone with the plus/minus cetuximab that will be a choice of physician. For the tested arm, it will be the same plus the addition of our product. The primary endpoint in this trial is PFS and the key secondary is OS. The statistics of this trial are giving the following: so we expect the control arm of the PFS to be 9 months, and the trial will be a success at the end if we get to 13 months as a median PFS. And for the OS, we expect the control arm to be around 12 months and success will be based on a 16 median OS for the tested arm. It is important to note that we've been using different stratification factor in this clinical trial. First of all, the comorbidity index, then the HPV status, the use of cetuximab, which will allow us to balance the population according to those parameter in the different arms. Now if we look at what we have been showing in the Phase I, meaning 16.9 months for the evaluable population in terms of PFS and 23.1 months, in the median OS, we may have a large room to play to be successful in this trial. But there are additional things to make sure and you will look at this on Slide 20, to make sure that comorbidities will not impact or we can mitigate the potential impact of high comorbidity at baseline, we've been having a stratification factor on that particular matter. As a reminder, in the Study 102, the expansion part, we had 67% of patients that had high mCCI. And obviously, this influenced the potential outcome for patients, both in their ability to receive the full treatment, but also overall in a potential survival of those patients. So the stratification factor we've been adding in the NANORAY-312 will mitigate this part. The other one, which is important and will play on the potential efficacy now is, as I mentioned previously, in the expansion part of the study 102, we've been injecting only the primary tumor. In the NANORAY-312, not only we will inject the primary tumor, but we also allow to inject surrounding lymph node. And here, we could expect more efficacy because of this. This latest finding had some impact on the timing for fertility and interim readout. As mentioned, having an average potential PFS of 16.9 may increase the average time to event to observe this in this 312 study. In addition to the potential difference in time to event, we've been giving guidance on the readout based on the hypothesis of the clinical trial that we are mentioning a theoretical recruitment rate with the opening of sites. Now we've been running this trial for quite some time and have a good experience of what's happening in real life. So if we combine the potential expansion of the average time to event and our experience of the current recruitment and the ramp-up that is really increasing at the moment, we should give new guidance for the futility analysis that will occur in H2 '24. And as far as the interim readout is concerned, we should get to the appropriate number of events in H1 '25 and be able to deliver the interim efficacy analysis by mid-25. Moving to Slide 21. As concluding messages, we've been showing in a very frail population, the ability to bring this new treatment into patients. We've been showing very strong efficacy signal in terms of overall response rate, PFS and OS. And compared with historical data, these results potentially strengthen our hypothesis currently being tested in the ongoing global registrational Phase III trial NANORAY-312. And we should remind ourselves that this will be potentially the first indication to be commercialized by J&J and LianBio, assuming the data coming out for the interim and final will be positive. So thank you for your attention, and I will go for question and answer. Operator?

[Operator Instructions] We'll take our first question from Lucy Codrington with Jefferies.

So just on the patient death due to sepsis that was deemed possibly related to the radiotherapy and NBTXR3, I'm just wondering if you could give us any more detail on that. I don't believe that this is something you've observed in all the trials that you've conducted to-date. Just any potential mechanism as to why other than the injection itself introducing, I guess, some form of bacteria, you would expect this to potentially be an issue going forward. And then just on the timings, I guess we knew the median overall survival was around 23 months back in February last year. So why the update to the timelines now? I guess, I just want to know how much of the change comes from these data versus what you're seeing in recruitment? And then on that, whether you could give us a bit of more detail on where you stand with the recruitment at present and whether the -- I guess, the rate of recruitment is going in the right direction. And then just on the delay to the readout by my model, if I assume the EUR 50 million capital raise in line with the J&J commitment at your at least EUR 25 million, I get that my runway comes to just about mid-2025. So I guess, how comfortable are you that you can address the cash runway to reach that read out? And I guess there's no room for any delays beyond this. And I guess is -- are there any key catalysts you can see as a trigger for that raise -- do you think futility is enough? Or is there something else that you are looking to as a potential trigger for that? And I'll jump back in the queue after that. Thank you.

Thank you, Lucy. So maybe we can tackle the first question about the sepsis, we've been observing. So just for you to know, it is not a usual adverse event we observed in our trial. That's one of the isolative case and have any trial with a new product when we can't attribute specifically a side effect coming from something, it's by default attributed to the treatment and the new treatment which is done. So obviously, that's something we are following and there's no real hypothesis behind why NBTXR3 as such could cause that, whereas an injection of something in the body could be responsible for that. But as it is something we do not observe in general, in the hundreds of patients we've been treating. That's something we're obviously following, but we'll not have any specific concern about this. So now about the second question that was about the timing. You're right to mention that we already have reported a median OS of 23.1 months in the recent past. But what is new today is the PFS. And the PFS has matured with the progression of the trial and the latest report for the available population, as I mentioned during the presentation, is 16.9 months, which is more than 5 months versus what has been reported in the past. So that's a big part of why this average time to event will change. Let's remind that OS is not the primary endpoint of this trial, that's PFS. So both futility analysis and interim results will be based on a number of events triggered on the PFS, and that will be 106 for the futility analysis and 283 for the interim analysis on the primary endpoint. So now how can we see what the impact of this increase of potential average time to event versus the recruitment, I think that's a mix of both. And that's why we've been shifting by 6 months the previous guideline we've been given to market. But I'd like to remind that shifting for this reason is a good reason. And because that shows that our product has a potential big impact for those patients and that also increased our confidence in getting positive results in the 312 trial. Now concerning the last question about the money and the cash position. Bart, do you want to take that one?

Yes, happy to, Laurent. Thank you for the question, Lucy. So we've disclosed significant steps to strengthen our balance sheet. As you may recall, at the end of June, we reported a cash balance of almost EUR 22 million and that was prior to an influx of cash resulting from the cash licensing of $30 million from J&J as well as their subscription and the subsequent execution of the first equity tranche of EUR 5 million. That then results in the remaining commitment from J&J for a second equity tranche of EUR 25 million in equity, subject to a co-commitment raised amount of equal size and that positions the company well to act upon any favorable opportunity to extend our cash runway. More notably, it excludes any BD debt, product financing, brands and subsidies and more importantly, any milestones. You will appreciate that we cannot talk about it, but management is confident that we will be able to see a significant development progress and materializing with what we just shared.

Our next question comes from the line of Ben Burnett with Stifel.

I just had a couple of questions. I appreciate all the details on efficacy in both the injected and non-injected lesions. Generally, I think efficacy looks pretty similar when you look at target lesions only or both targeted and nontargeted lesions. Could you talk about like why that is the case? Like why is that data so consistent in your view?

So sorry, I'm not sure if I catch the entire question, maybe rephrase it. If you want to recap what we do observe at the local vision perspective versus the surrounding lymph node and why do we observe what we observe, is that a fair summary?

Yes. Apologies. I think my phone connection is not perfect here. But why is that data so similar when you look at -- and you broaden the efficacy analysis to both targeted and nontargeted lesions. I think it's interesting that you don't actually lose more efficacy when you look at nontargeted lesions. Just curious if you could talk about that. And then I guess the other question I wanted to ask is it sounds like most patients had [Technical Difficulty] at least 90% of dose when they were [Technical Difficulty] trial. What is the reason why they didn't get 100%?

Okay. Thanks [indiscernible] got it now. So let's first start first with the injected versus non-injected lesion. So just as a recap, in general, in head and neck cancer, or locally advanced, you will define the target volume for the radiation therapy that will be based both on the primary tumor and involved lymph node that could surround the tumor. And the idea here is to give the central dose to those 2 lesions to make sure we can get to an appropriate level of control. And as a standard of care in this population, we delivered 70 gray. In our Phase I, we've been injecting the primary lesion only, not the internal, but both received 70 gray. And what we do observe, as I mentioned, is that the median duration of response for the injected lesion has not been reached, whereas the overall duration of response is 12.4 months. So as you mentioned that there's a slight difference between overall response that include noninjected versus just the local response. And that's due to the fact that the -- most of the patients, close to all that have been relapsed, the relapse came from the lymph node itself that received the same dose but no injection of R3, so that's where we get that. And you see that difference to not only in the overall response rate, which is closed, but different, but also in the total number of complete response. So we've been achieving a higher rate of complete response when looking at the injected lesion versus the overall. But what it means is the all results here are mainly driven by what we achieved in the primary lesion with the injection of the product given. And most of all the failures are coming from the surrounding lymph node in this trial. That's why in the Phase III, giving the additional ability to get to the lymph node also could have extra efficacy. But regardless of this, we already observed a very good PFS and a very good OS in this population. Knowing that because of what I said just before, the PFS will be mainly driven, at least the patients that are dropping by the surrounding lymph node. So altogether, that's a good surrogate for us to the potential success of the 312. Now the second question is about the dose that we have to give to patients. You're right, to be evaluable eligible the patient needs to get 90% of the lot. And all patients receive the 90% of the dose minimum. So it's not that 90% of the patients receive the dose that is supposed to get. But all patients treated here have been receiving the appropriate dose of R3 to be eligible for efficacy. And we've taken 90% as a reference because we think 90% of the 22% dose, which is the recommended Phase II dose we have in this trial is an appropriate dose for efficacy.

Okay. That's very helpful. And I guess the reason why they wouldn't get 100%, is that just because there's heterogeneity in manufacturing or heterogeneity in the target lesion size. Why wouldn't patients get 90%? Or why was the variability there?

Well, 90% is just an artificial threshold, right, just to being them eligible. So most of the patients got the appropriate dose that was planned at the beginning.

The next question is from the line of Swayampakula Ramakanth with H.C. Wainwright.

A couple of quick questions. So Laurent, in the reference studies that you talked about, where the median OS was 12 months, what was there -- in one of the studies you said the ACCI was greater than -- greater or equal to 4, but how about the other 2 studies. The reason I'm asking is, in general, do patients generally have a score higher than 4 when they get to clinical trials?

Thanks, RK for the question. I think that's a very important question and a strong differentiation of the patient population we've been treating in the Phase I versus the overall population that you can find in some of the literature. So what you see here, these are the different reference that is presented on Slide 6. So this is coming from a clinical trial that has been done or study. And in this case, we have more like overall population that come into these different series that have been observed. And the overall comorbidity index that you observed in those reference were closer to what we should assume of what a general head and neck elderly population is versus the one we have been treating in our trial. So I did mention in the intro that you should expect 20% to 30% of patients having high comorbidity. And in our case, we've been treating patients that 67% of them had an ACCI more or equal to 4. So that's made a strong difference when we look at the potential overall survival at baseline for this patient. So that's why we differ from what we've been treating in our trial versus what you can see in the different paper in literature. Treatment has been the same, but commodity at baseline is different. And just for information, there's also another important factor that in our trial, there is a good proportion of oral cavity patients that usually have a worse prognosis than older head and neck cancer patients based on location.

Okay. Great. That's helpful. The second question I have is on the 312 study. So you told us that there is a stratification for the comorbidity index. And then the -- in this study or in 312, you're going to be treating not only the primary tumor, but also the lymph node. Are you going to be stratifying the data as well? Because I'm just wondering for the patients that receive treatment on the lymph node, how would that change the PFS and OS?

That's a good question. So yes, first of all, we've been stratifying on different factors and have adopted the statistical hypothesis for that. So all the design of the statistical hypothesis and the stratification factor have been done at the beginning -- before the beginning of the trial. So now about the lymph node injection. We are not stratifying between the patients that receive lymph node injection versus not because they will not get it in any case in the control arm. So the idea here is to measure the overall impact of the treatment where we inject by nature, the primary lesion and potentially the surrounding lymph node. Not all patients will have invasion of lymph node, right? In the Stage 3, 4, you have sometimes patient with a very large tumor and no involvement of lymph node. So that will be part of the treatment. So what we should expect or could expect, we can take as a reference what we have been observing in the Phase I for the impact overall and on the local region. And what we may see in this trial, if we inject the lymph node on the top of the primary, we may have a greater response and potentially a greater PFS, especially because what we've been observing in the Phase I is that most of the relapse are coming from the surrounding lymph node, not from the primary tumor that received the product. So altogether, let's say that, that could be a plus versus what we see today.

Yes, because that's what I was trying to get at, like because would you get a good readout on the abscopal effect as well. So I don't know how much of an impact or a benefit you're going to get from that.

I think in this particular setting, if we think about abscopal effect, which is when you irradiate somewhere in the body a lesion, and because of the priming you trigger, you would have an impact potentially of the immune system somewhere else in the body in a distant lesion. I don't think this local general setting here is a very good setting to observe such effect. Because both lymph node and primary tumor will receive radiation. And the lymph nodes usually this one should be the nest to get the priming or to get this immune response. So the fact that we give for 7 weeks irradiation to the patient, both in primary and lymph node may not allow us to see any immune response in that particular setting, which may not be the case in other settings.

The next question is from the line of Colin Bristow with UBS.

Congrats on the data. Just a couple of kind of follow-up clarification questions from our side. [Technical Difficulty] and I think it's been covered, you're obviously allowing the injection of lymph nodes, but you're also allowing the use of cetuximab. Just in the broader picture of how you view -- just what you -- how you think this will impact the realizable [Technical Difficulty] NBTXR3, it would be great to get your thoughts there. And then secondly, in terms of the 10 deaths, it wasn't specifically clear to me the attribution to NBTXR3. So if you could give a little more color and specifics around that, that would be really helpful.

Thanks, Colin. So let's start with the cetuximab question. So the reason why we have had cetuximab as a potential addition of treatment in the Phase III is because cetuximab for this type of patient is still part of the standard of care, even though the use of this product could be a little challenge given the results that have been shown on elderly population. If you take as an example, the [indiscernible] it will tell you that for elderly patients, cetuximab does not bring additional benefit to radiation. It could be even potentially detrimental. But the fact that cetuximab is still in the guideline, we need to leave the possibility for a physician to use it on the top of radiation. But also to prevent any potential unbalance linked to cetuximab, both from safety and efficacy, we have used that as a stratification factor. So altogether, when we look and think about what could be the addition of cetuximab impact versus what we have been observing in the Phase I where there haven't been any cetuximab treatment, so I would qualify that from neutral to plus. Meaning, if we have an effect in any case, it will be the same in tested arm and control arm. But according to literature, we should not have much effect based on that. What we can exclude is the potential synergy of radiation NANO and cetuximab that haven't been tested in terms of efficacy. So that's why we qualify this cetuximab potential impact of neutral 2+. Also, far from all patients will get cetuximab in this trial, that would be a smaller propulsion. So I hope it does answer your question about cetuximab. So out of the 10 deaths, they won only 1, which was treatment related and potentially to NBTXR3 [indiscernible] have been other type of deaths. And a lot of them have been linked to noncancer-related deaths and link for patients that have high comorbidity especially in the early days of the treatment.

Next question comes from the line of Sebastian [indiscernible] with Kempen.

Congrats on the progress. Just 1 from us. Can you maybe remind us on why you have not injected the [indiscernible] lymph nodes in the Phase I trial? Are there potential risks for toxicity [indiscernible] for injecting the surrounding lymph nodes?

So first of all, the Phase 1 was the first trial in head and neck, and we did not want to multiple the question in this trial. And wanted to have a clear way -- a clear look at what is the impact of the treatment in the primary and what's the potential impact overall with the non-radiated lesion. That's why by designing this trial, we've been putting at endpoint not only the local injected lesion response, but also the overall response to study those things. And that's the study of this. That allowed us to define that it will be a good thing if we could also inject the lymph node, which we did propose for the Phase III. So it was a story of injecting another lymph nodes. In between, we've been running a trial in another setting with our former partner for Asia, which was PharmaEngine in the radio chemo setting. So it's another head and neck cancer population, the one which is this time eligible for chemo plus radiation. And here, we already have been allowed to inject only primary, but surrounding lymph node. And in this trial, we had no specific safety issue in that regard.

Next question is from Lucy Codrington with Jefferies.

Sorry, just a follow-up from me. Just a point of clarification. I was just looking back at my old notes. And at one stage, the historical or, I guess, comparative PFS was 7.3 months as opposed to the 9 months used for the NANORAY trial design. So I was just wondering if you -- if there is a reason for that change, is that just reflecting progress in care or am I reading too much into it? And then apologies if this has been covered already and I missed it, but I guess when we're thinking about the NANORAY-312 trial population, would you say they are going to be more like the evaluable patient population in 102 or more like -- more representative of the intention-to-treat population? And on that, you mentioned that you were stratifying by the ACCI score over 4. But are you expecting a similar disproportionately high representation in 312 or are you expecting it to be closer to the 20% to 30% you would normally expect for head and neck cancer population. Thank you.

Thank you, Lucy. Let's come back to the first question about the reference on PFS. So you're fully right to mention 7.3 months because that's one of a potential reference we have for the trial population, and you're not correct. In the past, our PI for this study have been presenting in one of the head and neck specialized conference from real-world evidence that has been taken from a database in U.S., where they have been looking at all elderly patients getting radiation therapy alone as a treatment with or without cetuximab. And the outcome for this patient were of 7.2 months or 7.3. And there was not much difference with or without cetuximab. But it's not the reference we've been taken for the design of the Phase III. At the beginning of the Phase III before submitting the protocol, we have chosen 9 months as a median PFS, for the control arm and 12 months for the OS based on all the literature we found. We thought that would be better to be conservative here and to take the higher PFS rather than just even though very interesting reference that's coming from real-world evidence that was lower. So in order to maximize the potential efficacy and the potential success stories of the trial, we've been choosing conservative hypothesis. So that's for the PFS and therefore, the PFS reference in our trial always has been 9 months and the OS 12 months for the control arm. Now the other question about the stratification factor based on ACCI. And what shall we expect in the 312 in terms of overall population versus the Phase I. In the Phase I that's something that I did not precise, but which is in the protocol and in previous communication, not only the patient, they are ineligible to cisplatin, but also intolerant to cetuximab. So the baseline we have chosen for the Phase I are really patients that are in-build batches. And that is why we've been observing such a high rate of comorbidity, 67% of the patients versus what we should expect in the overall population. That's why we think in the Phase III overall we should have a population that is closer from the evaluable population, if we can say so. But in any case, by having the stratification factor ACCI more or equal than 4, we will balance this potential effect in both trials and could erase that sort of to have a clean readout with or without a high proportion of high ACCI.

It appears that we have no further questions today. Thank you for participating in today's Nanobiotix's 102 study data call. You may now disconnect from the call. Thank you, and have a great day.

Thank you, everyone. Let's talk very soon to continue to report on progress of NBTXR3 in the different program, including the 312. And also, you may have noticed last week we released some written data generated by MD Anderson on the pancreatic cancer. So stay tuned, we will update you on that very soon. Thank you very much, and have a great day.