Idorsia Ltd (IDIA) Earnings Call Transcript & Summary

Good day and thank you for standing by. Welcome to the daridorexant Phase III program published in The Lancet Neurology Conference Call. [Operator Instructions] I must advise you that this conference call is being recorded today, Thursday, the 20th of January 2022. I would now like to hand over to your speaker for today, Mr. Andrew Weiss. Please go ahead, sir.

Thank you, operator. Good morning, good afternoon, everyone. This is Andrew speaking, Investor Relations at Idorsia. We're here to have a webcast to discuss the newly published results of daridorexant in the Lancet Neurology publication that was made available this morning. With me on the call today are our CEO, Jean-Paul Clozel; our Chief Scientific Officer, Martine Clozel; our Head of Global Clinical Development, Guy Braunstein. And joining us afterwards for the Q&A, we will also have Simon Jose. Next slide. As customary, we will be making forward-looking statements during this webcast. So everybody has, with this slide, been adequately be reminded of the risks and benefits and also possibilities of us making forward-looking statements in this call. Next slide. Jean-Paul, the floor is yours.

Thank you, Andrew. So we are on Slide 3. So in -- 2022 was -- had a very busy start. And I think it's because we are now fully functional and all the teams that it can be in the U.S., in Europe and in Japan are delivering on their promise. I have to say, it's fantastic to see how people are motivated and are working very hard to get these results. Next slide. And today, it was not planned, but a few weeks before the due date, Idorsia received the Japanese PMDA approval for clazosentan called -- brand name is PIVLAZ, for subarachnoid hemorrhage. Unfortunately, we not have time to discuss this very good news. It's very important and congratulations to the Japanese team. But we are going to focus today on the results of daridorexant published in The Lancet Neurology. Our intention today is really to recall how did we really discover daridorexant, and why we are having this fantastic results. So I think that now we can go to the next slide. Next slide. And Martine will tell you the very long way to discover daridorexant before we go to the results. Martine, please.

Thank you, Jean-Paul. Good morning, good afternoon to everyone. It's a great pleasure to be with you today to announce the publication in The Lancet Neurology of the seminal paper on the results of the 2 pivotal Phase III studies of daridorexant just over 10 days after its approval by the FDA. Next slide, please. It was indeed the long road to daridorexant. In 1998, just 1 month about 2 groups published on the discovery of a new peptide neurotransmitter which is a group of some so called hypocretin in the book of Masashi Yanagisawa at that time in Dallas called orexin. It happens that just 1 week before this publication, I had invited Masashi Yanagisawa basically newborn labs at Actelion, just a few months after its startup. Masashi Yanagisawa had been the discoverer of endothelin and we had discovered the first endothelin receptor antagonist. At that time, we immediately started a program because what Masashi was describing was -- he was explaining me that 1 week after he was going to publish and sell on a new peptide and what he was telling me was very particular. And we immediately started the program to try to find out what was the role of orexin and what could be a potential profile if interesting, of blocking the orexin system. It took almost 1 more year for Emmanuel Mignot and Masashi Yanagisawa in 2 different papers to show that orexin was playing a role in wakefulness. And in 2007, we published in Actelion in Nature Medicines that indeed dual orexin receptor antagonist induce sleep in the lab and in man. Our tool at that time was almorexant and we felt that almorexant was not yet a perfect candidate. So we defined a scientific product profile, which would be a better compound to take into account all of our learnings of what was the profile of dual orexin receptor antagonist, but to improve from almorexant. And it took many years of organic chemistry and iterative cycles to arrive to a compound which was fulfilling the scientific product profile, and this optimized candidate was daridorexant. Then entered into man, initiation of Phase III, 20 years after the discovery of orexin. And today, it's a real pleasure to announce a description of this Phase III pivotal program in The Lancet Neurology. Next slide, please. Indeed, orexin is very particular in that -- and that was one of the main reasons for which we started the discovery program even before the publication of the paper describing hypocretin and orexin. Very, very few cells in the brain produce orexin. Only about 70,000 of neurons produce orexin in man and these neurons are located in the lateral and posterolateral hypothalamus. And that was really suggesting because the hypothalamus is a very particular and important organ for the regulation of many functions that there was perhaps an important role of orexin. And indeed, when we discovered our first dual orexin receptor antagonist, there are indeed 2 orexin receptor: OX1 and OX2. We learned that there was a very specific profile of sleep induction. This sleep was not modified as compared to a normal sleep. There was no respiratory depression. The sleep was surmountable in that one could easily wake up from that sleep and that sleep if there was no pressure to sleep. And in particular, upon a wakening, there was no decrease in muscular strength in contrast to GABA agonist. We realized also that chronic administration was not inducing any loss of efficacy and that there was no . Next slide, Slide 9, please. Slide 9. Next slide. Okay. I will continue with the slide. Almorexant was not a perfect drug, as I said, and we continued our drug discovery research. Can you go to the next slide? We continue to try to combine the properties of dual orexin receptor antagonist but with an optimal profile for the sleep drink. Indeed, we wanted to combine this precision mechanism of action, which is very different from a global sedation of the brain as that which can be obtained with GABA receptor modulator, but we wanted to combine this precise mechanism of action with an optimized microcytic profile. And we try to make a next-generation dual orexin receptor antagonist, which would have an optimized pharmacokinetic profile for the right duration throughout the night without residual next morning sleep effect, hoping it could provide an improvement in something which is very important symptom of insomnia, which is impairment of data and functioning. We were really hoping we could improve daytime functioning. So we defined with these properties, what we wanted in the next-generation DORA and this scientific product profile was indeed put in motion. Next slide, Slide 10, please. Our concept was that we should really try to cover about 6 to 8 hours of the night with an optimal drug profile for a sleep drug in insomnia. Next slide. Indeed, many drugs are having a duration which is too long, and that is normal because it's difficult to ask for the duration of the night and not have, as a consequence, residual effect the next morning of somnolence. Next slide. To avoid this residual somnolence, what can happen is to obtain a suboptimal efficacy during the night in order to avoid the next day somnolence effect. So our goal was to try to have an optimal duration of action during the night while operating next morning residual effect. Next slide. And the result of that is daridorexant. Practically, the way we did was to try to have a drug discovery effort where we would try to predict the duration of action in line of our molecules and for predicting the human duration of action, we predicted by PBPK modeling the human pharmacokinetics of our molecules. But first of all, we estimated which orexin-2 receptor occupancy was needed for sleep. OX2 is a receptor, which is the most important for inducing sleep when 1 blocks the orexin system. The combination of the prediction of human pharmacokinetics and orexin-2 receptor occupancy needed for sleep or awakening was allowing us to calculate what was going to be the duration of action at optimally effective doses for our different molecules during the discovery process. And the goal was to continue the discovery program until we would find a compound, which would have an estimated human pharmacokinetics, giving sufficient receptor occupancy for 6 to 8 hours. And you can see here on the pharmacokinetic slide of the Phase I of daridorexant is that after an administration of the compound, there is a maximal concentration, which is reached rapidly, and which declined quite rapidly achieving a 24 hours of extremely low levels. A data where we have lost already about 80% of the plasma concentration, 70% to 80%. And at 24 hours, the concentration residual concentration is very low. Daily administration for this shows that the next full pharmacokinetic curve is similar, super impossible to the first one, there is practically no accumulation which allows to have a day without really significant effective concentrations any longer. Next slide. This important aspect of the scientific product profile was not alone. There was also element in the SPP, the desire for the profile to discover a molecule with optimize efficacy and safety. This is why it took more than 25,000 molecules after almorexant in the project to arrive to 1 molecule. The sequence here on this slide is given by the cascade of test in the program, and you can see how we reduce progressively by iterative cycles, we eliminated many, many of our molecules because besides the optimal duration of action, which we desired, which was modeled by use of the predicted pharmacokinetics and pharmacodynamics, hoping that there would be a good effect on a variable results next morning residual effect, we had a number of additional requirements for designing the optimal molecule. A high inhibitory potency on both orexin-1 and OX2 receptors. Indeed, in addition to antagonist for OX2 receptor from the amount of sleep, we felt important to find the molecule, which will block as much orexin-1 receptor. Indeed, OX1 mediate the link between orexin-producing neurons and systems involved in anxiety, mood, hyperactivity. Therefore, we said that OX1 blocking was important as well to help solve the fear of not being able to sleep characteristic of insomnia. And hopefully help to solve also the data impairment in addition to the improvement in night variables. We, of course, wanted also a high and fast brain penetration to help efficacy and speed of action and also various aspects to optimize safety. Next slide, please, 15. The result of this more than 20 years of drug design is a compound, I think, with outstanding qualities. And indeed, the preclinical characteristics of daridorexant was combining an equipotent antagonism of both orexin receptor. High brain penetration, rapid absorption, and a model duration of action of about 8 hours in insomnia, a very safe compound in toxicology with no teratogenic or carcinogenic potential, chronic efficacy and also a preservation of muscular strength and preservation of memory in the lab upon awakening with no abuse potential. I would like now -- next slide, 16, to hand over to Guy for the presentation of the data published today in The Lancet Neurology of the 2 pivotal Phase III studies of daridorexant. Guy?

Thank you, Martine. Good morning, good afternoon, everybody. Slide 17. We can move on immediately even to Slide 18. And on slide -- next slide, please. On Slide 18, you see immediately the conclusion from The Lancet Neurology article that daridorexant 25 milligrams and 50 milligrams improved sleep outcomes and daridorexant 50 milligram also improved daytime functioning in people with insomnia disorder with a favorable safety profile. What I'm going to do in the next 20, 30 minutes to tell you how we got there. And of course, as you can imagine, there is a wealth of data and results in the publication and in the supplement, and I invite you to read in detail the paper and the supplement because I won't be able to show you everything that is already published. Next slide, please. Before I jump into the study itself, I would like to maybe remind you a little bit about the disease insomnia. And first to mention that insomnia, it's primarily a subjective patient experience. I'm saying that because we place patients in the center of our development program. Aligned with the DSM-5 definition, we know that insomnia affect both night and day. And this is also important to us because, as already mentioned, we want to show the efficacy of the product on the night symptoms as well as the day symptoms. From patient input and patient research that we have conducted, we have identified daytime symptoms and total sleep time as the major concern to them. And this is important also in terms of the choice of the endpoint that I will describe later on. Finally, I would like to mention that in most of the development programs of diagnostics, daytime symptoms have been largely ignored, which means that in the end, we know very little about impact of currently available products on the data and symptoms of patients. This may be due to the desire to ignore it or maybe most likely as well, the fact that there was no tool, no instruments, patient-reported outcome instrument to collect properly the impact of product on the daytime functioning of patients. Next slide, please. Here, it's a schematic of the design of the development program, Phase I, Phase II and Phase III. What I would like to mention here is that at the end of the Phase II, we knew quite a lot about daridorexant. We knew that there was a dose response, efficacy on 9-time parameters, WASO, LSO, LPS, total sleep time. We also knew, and it was very important that the full night was covered with the effect of WASO increasing quarter-by-quarter and culminating the last quarter of the night. What was also very interesting is that from the Phase II study and even from the Phase I, we knew that there was no persistent effect beyond 8 hours. And when patients wake up in the morning after night dosing, there was no sleepiness in the morning. Finally, we also know -- knew at that time that there was a signal on daytime functioning. That was with a good instrument, is analog scale only, but we had the idea that the profile that Martine described before, could actually be reached. However, there were a number of open questions still, the first one being the dose that can provide the best balance between a good night, a good day without morning sleepiness. It's clear that there is a fine balance here between the good effect and the potential hangover effect as shown by Martine on the pharmacokinetic profile of the product. And it's not easy to determine the dose that will provide the best benefit risk from that perspective. And this explains why we had 2 testable doses in the Phase III program, and we decided to test the 10, 25 and 50 milligrams. The other aspect that was still open was to have a proper measure of their time functioning we knew that there was no adequate tool that will allow us to properly measure the data functioning from a patient perspective. And we had a chance to develop this instrument in parallel to the development program of daridorexant including the Phase II study. Finally, we could design the Phase III program through the doses, as I mentioned, 10, 25 and 50 milligrams. We looked at the effect during the night and the morning and the effect during the day. Of course, the safety with long-term treatment was absolutely essential as well as the maintenance effect with chronic treatment. So all of that is actually addressed in the Phase III program. Next slide, please. We are on Slide 21. So the key question here is how can we measure daytime functioning? And as I mentioned, there was no tool available, and we embark into the development and validation of a new instrument called IDSIQ. The basis of this instrument was in the publication in 2007 called the DISS, Data Insomnia Symptoms Scale. This was a good baseline to develop a better instrument, but it was insufficient from a patient input perspective. And during the development of the final instrument, IDSIQ, we had a series of patient interviews that will allow to elicit concept to ensure saturation. We had also connected the brief that allowed to make sure that the patients would understand the questions, be able to respond to them and eventually validate the content of this questionnaire. Following this development, the validation was done through a series of studies, 2 studies: 1 healthy participants and another 1 insomnia patients. These studies were conducted in parallel to the Phase II program of daridorexant. And it's the statistical analysis validated psychometrically, the instruments, in particular, we could show that there was a very different response between healthy subjects and insomnia patients as well as all the characteristics culminating in the final IDSIQ questionnaire, which was eventually accepted by the FDA as appropriately validated instruments. We had several interactions with the FDA and confirmation that was an instrument that would fit the FDA needs. One important element of the validation was to identify the within patient meaningful difference and this is actually published in a peer review journal in 2021. All the details of this development actually published. Maybe one point to mention as well is that this questionnaire was translated into multiple languages, which was useful for inclusion into our program that was multinational. Now that you know how we did it, and this really took several years to maybe 6, 7, 8 years of work to develop this instruments, we can look at the instruments itself on the next slide, Slide 22. So here, you see on the left side of the slide, the exact questions in the right order that were raised to the patient every day. We see the variety of -- the diversity of questions here. And as you see, the phrasing is really coming from the patient themselves. We see the diversity of the question can go from having been able to concentrate, being frustrated, irritable, maybe how much of an effort was it to perform daily activities as well as being tired, mentally, or physically or sleepy. You can see all these questions. These questions are addressing 3 key dimensions of data functioning. The first one is cognition, and the domain score is called alert/cognition in blue made of 6 questions, item 1, 2, 3, 9, 10 and 14. The second one, which is called the mood domain is really looking at the effective dimension of the data and functioning and it's made of 4 questions. And finally, the physical component of data and functioning is assessed by the sleepiness domain score. Sleepiness domain score, which is made of 4 questions: Item 8, how energetic do you feel today; Item 11, how mentally tired did you feel today; Item 12, how physically tired did you feel today; and Item 13, how sleep in the field today? On the right side of the slide, you see a typical screenshot of what the patients were actually seeing when they were answering every day the response to your questions. It's important to mention that this is a daily assessment. The recall period was today, and this avoids the problem of memory bias when the recall period is longer. So now that you are very familiar with the instrument that we have, I can go to the next slide, slide 23, which shows the design of the study, very classical, the screening phase with a double -- a single-blind placebo running phase of 2 to 3 weeks, during which the patient had a polysomnography. And then the patients were randomized to either a high dose, low dose, or placebo for 3 months with an interim look at the first month of V6 with 2 polysomnographynitis, at the end at 3 months again 2 polysomnographynitis. And finally, a single-blind placebo run-out period designed to assess a rebound effect, if any, the withdrawal symptoms. And eventually, the patient could move on to the extension study, so that in total, they could be treated for 1 year or maybe they didn't want that, there was a safety follow-up period. With this 2 identical studies called Study 1 and Study 2. Study 1 had the high dose of 50-milligram and a low dose of 25 milligrams. And Study 2 had the high dose of 25 milligrams and low dose at 10 milligrams. So by doing that, we had 2 arms that were similar in both studies, the 25 milligram and the placebo. And 2 arms that we are different in the 2 studies: the 10-milligram in Study 2 and the 50-milligram in Study 1. Next slide, please. Slide 24. Here, we see the key assessments, as I mentioned before, we assess the polysomnography at baseline at 1 month and 3 months of treatment as well as during the single-blind placebo run-out period. And this was needed for -- to meet regulatory requirements. There is really an appetite from the regulators for objective measures. But we also wanted to have patient-oriented outcome every day because as we know, and as I remind you earlier, Insomnia is primarily a patient subjective experience. And we have 2 categories of patient-oriented outcome instruments or measures. One was the night with the sleep diary questionnaire that I will come back to in a moment; and the IDSIQ questionnaire, which I described before. The next slide show you the sleep diary questionnaire in brief. Again, that was a daily recording. And this questionnaire has actually 2 components, the morning questionnaire made of 10 questions. And from those, the total sleep time is the one that was the most important. As I mentioned, this is one of the main concerns from patients. And there were also 3 visual analog scales that we're assessing in the morning, the quality of the sleep, the deepness of the sleep and how sleepy patients were in the morning. And in the evening, there were additional questions for the day relating to napping during the day as well as 2 visual analog scales again, related to the alertness of the patients during the day and the ability to perform their normal task during the day. Next slide, please. Slide 26 is now showing the key endpoints that we had in the study. As I mentioned, we have objective and subjective endpoints. The 2 primary endpoints were derived from the polysomnography, the wake after sleep onset or WASO and the latency to persistent sleep or LPS. We had 2 secondary endpoints derived from patient's questionnaire, the night, and the day. The first one was subjective to the sleep time by the SDQ that I showed you before and the sleepiness score domain during the day by the IDSIQ questionnaire. As mentioned, we had 2 main endpoints -- 2 main time points, the month 1 and the month 3, where all these endpoints were assessed. And of course, we had 2 dose levels in each study: Study 1, 50 mg, and 25 mg placebo; and Study 2, 25 mg and 10 milligrams placebo. In total, when you look at all these endpoint, time points and doses, you end up with a need to execute 16 comparison to placebo -- excuse me. And what we need to do is to control the type 1 error at a study level, so that the conclusion -- solid conclusion can be made from a statistical perspective. When I presented that years ago now at the beginning of our program, everybody was questioning how can you do that? How can you actually control that whenever across that many hypothesis testing? And this is shown on the next slide. I can spend maybe a minute or 2 describing it because this is really crucially important. So the way to do it, first is to organize this hypothesis testing in a logical way. To do that, we sort of prioritize initially the high dose versus placebo. And this is represented by the green and the blue bubble. And then the low dose represented by the red and the purple bubbles. Within each dose level, we prioritize month 1 and followed by month 3. So this is the green versus the blue and the red versus the purple. And there is a reason to do that. There is no reason to look at low dose if high dose doesn't work. So therefore, we start with the high dose. And there is no reason to look at 3 months if the product doesn't work at 1 month. So this is why we had these 4 consecutive buckets here, high dose placebo at months 1, month 3: and then low dose placebo month 1 and low dose placebo at month 3. However, we still have to deal with the 4 endpoints as each -- for each of these 4 comparisons. And to do that, we prioritize the objective measures was on LPS first, followed by the subjective measures of sTST and IDSIQ. And of course, we want to control the time for all of those. So we start with an alpha level of 0.05 and we initially applied the Bonferroni correction. And in doing that, we divide alpha by 2 to test the hypothesis, the new hypothesis for LPS and the new hypothesis for WASO. Depending on the results of this first comparison, which is tested at 0.025 for LPS and 0.025 for WASO, the propagation of the alpha depends on whether we can reject the new hypothesis on a conclusion or not. If we can reject the hypothesis then the alpha is propagated below following the red arrows. And if it's not -- if we cannot reject the null hypothesis, then the alpha stops there and is, in a way, lost. So if I take the first example of the study, study #1. When we test the first 2 hypothesis: H1, H2, the WASO, LPS, we could reject both of them. And therefore, the 0.025 for H1 and H2 could be preserved and propagated. The propagation was half of it, so 0.0125 was going to H3 from LPS. And similarly, 0.0125 was coming from H1 to H3, which means that in total, the sTST H3 could be tested at 0.025. And half of the alpha was coming from the H2 was going to H6 and half of the alpha from H1 was going to H5 and so on. So here, you see the progression from 1 variable to the other. We were very fortunate in the 301 study that there was no failure of rejecting any hypothesis up to H12. So therefore, up to that level, the alpha will stay intact. And it's only H12 that finally which show you later. And however, H14 and 13 could be tested, H15 as well, and the final one, H16 fell as well. So we were very, very fortunate here that with this structure, we could actually make very robust conclusions on the high dose at month 1 and month 3, the low dose at months 1 and 3 on most of this variable. In the end, 14 of these 16-hypothesis testing, we could reject the null hypothesis and make a conclusion. So this is how we did it. It's fully described in the supplement of The Lancet Neurology article, and I invite you to read the tech that goes with this design and also to look at the Page 19, which shows the alpha for each of these hypothesis and the P values, and you can compare it to and see whether we reject or not the new hypothesis. In addition to these 16 endpoints and measures, we had a number of exploratory endpoints, and they are described on the Slide 28. Next slide, please. And you see that we had also TST by polysomnography as well as a number of variable from PSG, in particular, the percentage of sleep time by each sleep phase. We had, as I mentioned before, the morning VAS for depth and quality of sleep. We had the change in insomnia severity index, a very classical instrument IDSIQ. We had, of course, the IDSIQ domain, the mood domain as well as the commission domain. We had evening VAS that I described and of course, the adverse event withdrawal on the rebound. So with all these endpoints, we have really a rich body of evidence supporting our product. Now that you understand the methodology for the development, we can start to move on to the patient demographics on Slide 29. And this is very classical. You have all the data in the paper. I would just point out that we had approximately 70% female, which is normal for insomnia studies. We had approximately 40% of patients -- elderly patients, which is very important in that indication. And interestingly, on the bottom, you see that we had quite a large number of obese or overweighted patients. And that's, of course, normal for insomnia population. So in total, we had really a population reflecting well the real-life situation. The baseline characteristics are on the next slide, and we see that Slide 30, what we see here is that the WASO approximately was an hour plus 35, 40 minutes. The LPS was approximately 1 hour. And the total sleep time was very short around 5 hours and this really reflects a well-defined insomnia population. Next slide, please. From the set of results that are published in The Lancet, I cannot, of course, in the next minute, show you all the data. But because the results of the 10-milligram didn't provide benefit and the 25-milligram was very consistent between the 2 studies, I decided for the purpose of this presentation to focus on Study 1. The rest you can really see everything is presented in the publication of the supplement. So we now just concentrate on Study 1, the placebo, 25 and 50-milligram dose. Next slide, Slide 32. Let's -- stay with me for a second as I can't describe how the slide is constructed because we'll use the same format for several other endpoints. So on the right of the slide, you see the graph with the observed values, the baseline, the month 1 and the month 3 with 3 lines: the placebo and purple; the 25 mg in blue, light blue; and on green, the 50 milligrams. On the bottom right, what you see are the statistics, the effect of placebo, the effect of daridorexant 25 mg, of daridorexant 50 mg as well as the placebo-corrected differences. On that one, you see the change from baseline. On the graph, you see the values, and, on the table, you see the change from baseline and the placebo-corrected change from baseline. And then you have this red box, which basically provides the conclusion. As we can see immediately here, we have a nice effect on WASO from daridorexant. The p-values in the rest of the presentation are only shown for those that are statistically significant. And we see a highly statistically significant effect of 25 mg and 50 mg as placebo. This is what we concluded at month 1 and month 3. What we see as well is that the improvement is achieved at month 1 and stay the same at month 3 with no further improvement. On the next slide, using -- Slide 33 using an identical format, we see the results for LPS. We see the very nice reduction. The placebo effect seems to be more important here than WASO, but we see an improvement by approximately 30, 35 minutes on the daridorexant 50 milligrams. We see the high level of significance with the P values. We again see that the effect is at month 1 and remained stable at month 3. So these are really a confirmation that daridorexant at 50 mg and 25 milligrams are inducing sleep and maintaining sleep. Now what about the patient feeling. And this is shown on the next slide, Slide 34, where we have the subjective total sleep time. And here, we see a pattern that is maybe slightly different. Of course, now total sleep time has to increase, not decrease on the previous slides. So we see an increase of the total sleep time. You have the numbers here on the table and you have the absolute value in the graph. Well, what we see here is that the effect at month 1 is good, but there is even more effect later on at month 3 and eventually on 50 milligrams at month 3, we see that the increase in total sleep time is reaching on average approximately 1 hour, which is considerable for these patients. Remember that at baseline, they sleep for approximately 5 hours and now on the treatment, they sleep at least 6 hours, which is really an achievement. And as we see the conclusion is daridorexant 25 mg and 50 milligrams are significantly impacting subject to total stim compared to placebo at month 1 and 3. So not only the patients are sleeping better in the sleep lab with the PSG, but also feeling this improvement in total sleep time, they're able to report in by themselves. There are a number of other analysis done during the night. I can only summarize them on Slide 35. And you see here, the proportion of sleep time spent in each sleep stage is preserved. It's affecting both the -- daridorexant is increasing the sleep time for the REM and non-REM sleep identically. There is no preference for 1 at this stage. So the architecture is preserved. We see in the middle that the quality of sleep and the depth of sleep are actually numerically higher for on daridorexant and placebo at month 1 and month 3, which is really good. Again, it's a perception by the patient that the night was actually good, and the ISI is also moving in the right direction. I want to point out that there was no statical testing on these endpoints but numerically, they are all in favor of daridorexant center. Let's move now on the data and functioning. And this is shown on the next slide, which has an identical format, Slide 36, an identical format to the previous slide. We see the graph on the left with the actual measures, and you see the statistics on the right and the conclusion. And here, what is really interesting, again, is that from a patient perception during the day, there is an improvement at month 1, but it continues to improve over time at month 3. And the pattern of patient feeling, as we have seen for sTST and now for IDSIQ is actually slightly different from the PSG. And here, we could make the conclusion that 50 milligrams significantly improved IDSIQ sleepiness domain. However, 25 milligram was not statistically significant, and that's why the p-value is not provided. However, if you look at the paper, you will see the p-values for all these endpoints as well. Looking now on Slide 37. I'm here reporting the results or the other domain. And again, no p-values because although you can see them in the paper, they were not adjusted for multiplicity. So for simplification, I don't provide them, but you can see them in the supplement. And we see an identical pattern of improvement with the placebo effect being relatively modest with the 25 milligrams getting better and the 50 milligram clearly differentiated from the 25 and the placebo. On the mood demand, alert/cognition domain as well as the total sleep time is the total IDSIQ score. So here, I want just to point out that there is consistency at least numerically on this different endpoints derived from the IDSIQ questioner. Finally, just to bring you a few more results on Slide 38. Again, during the day, we had this vast visual analog scale scores for ability to function on data alertness. And this, again, numerically were better at months 1 and 3 on daridorexant on placebo. And we had this morning sleepiness VAS score, which indicates less sleepiness in the morning in the patients treated with daridorexant compared to placebo at month 1 and month 3. Again, these are not statistically tested because they are not in the hierarchy, but the trend and the numerical values are very clear, and you will be able to see all these results again in the supplement. Finally, to finish on the efficacy results, I want just to point out on Slide 39 that the effect on WASO, LPS as well as total sleep time and IDSIQ, were very consistent between the different subgroups adult versus elderly subjects across sex, male and female as well as geographical location. You see here just one example of WASO at month 3, but all the data for all the endpoints at month 1 and month 3 actually presented in the supplement and you will see by yourself that they are very, very consistent between subgroups. Of course, I cannot stop here. I have to talk a little bit about the safety of the product. And I will start by -- and maybe remind first that for transparency, I will show you the results -- the safety results for Study 1 and Study 2, although I focused the efficacy on Study 1. Next slide. Slide 40 shows the adverse event profile, and we see that there are actually relatively few adverse events in general, only 1/3 of the patients have reported one or more adverse events. And this is very balanced between the treatment groups. We see that the adverse events led to treatment discontinuation is very rare. Actually, very few of them. We can more or less count them by hand. And we see that the most frequent adverse events are really nonspecific nasopharyngitis, headache. Maybe it's more interesting to look at the somnolence with 2% on placebo in Study 1 and 1% Study 2. And you see a number that are really very close to that in Study 1 and 2 in daridorexant 2% on 50 milligrams, 4% on 25 milligrams Study 1; and 3% on daridorexant 25 milligrams and 2% on daridorexant 10 milligram in Study 2. So we have really a very, very small difference between daridorexant and placebo. And maybe it's time to also point out that from a safety perspective, the profile of 50 mg doesn't really differentiate from lower doses. We can see also the values to fall or fatigue or dizziness and all of that is extremely consistent with very small number of adverse events reported on daridorexant. The next slide, 41, is a focus on adverse events which might be seen as maybe more concerning, excessive daytime sleepiness, sleep paralysis, hallucinations, suicidal ideation, or self-injury. We had very few of them. We had an adjudication committee independently blindly reviewing these adverse events and you see the data here, they are really very, very important but are rare and somewhat balanced as well between daridorexant and placebo. Finally, on the adverse event description, I would like to point out the effect of gender effects -- sorry, the effect of age, the younger and the older patients, and you see that on Slide 42. On the left-hand side of the slide, you see the younger patients. On the right-hand side, you see other patients. And as you can see, it's very by consistent and even sometimes maybe less adverse events in elderly compared to the younger. Of course, these are small numbers, so it's difficult to make a real conclusion. But if you look at somnolence you have 2.1% on 50 milligram in younger and you have only 0.8%, actually 1 patient in the elderly. If you look at the fall, which is one of the main concerns of agnostic in the early patients, you see actually on the on the right side that there were more -- numerously more patients having falls on placebo than daridorexant. So these are really reassuring findings that the safety and tolerability of daridorexant even at 50 milligrams and in more fragile, elderly patients is very good. We had further safety observation that I would like briefly to mention but all the details again are in the publication. Slide 43, we see no adverse events suggesting the drug misuse, which is an important observation. We see that there was no suggestion of withdrawal symptoms during the run-out phase and also there was also a rebound detected during this placebo run-out phase. Just to finish with a couple of slides on the description of the paper. There are, of course, Slide 44, some strength and limitation of the studies. On the left, you see the strength with the assessment of most component of insomnia in particular, objective, and subjective endpoints, the night and the day which is unique to this development program. We had robust type 1 error control, which is also very important to make robust conclusions. We see that the description of patients is really reflecting the real situation with more female and male and a nice representation of elderly patients, approximately 40%. And something I didn't mention before, but we had a very low rate of missing data and actually also a very low rate of discontinuation of patients. Of course, as normal, we have to balance that with limitations and they are so that I mentioned in the paper, which we can talk briefly about mainly the study focused on moderate-to-severe insomnia. This is reflected by the high size score above 15 as part of the screening values. But as we know, below 15, it's borderline insomnia. So this reflects really the real-life situation. Regarding rate, we may have had maybe underrepresentation of minorities. Maybe more interesting is to see that we had a few patients with CBTi experience. But again, this is reflecting the real-life situation that although CBTi is recommended as first-line, there are actually very few patients that have the time, the energy, or the possibility to find the psychotherapy. There is also a price constraint. So it's difficult to do CBTi and therefore, the fact that we had only a few patients with CBTi reflects the real-life situation. Of course, comorbidities, we didn't have many, but there are reasons for that, that are fully described in the protocol, which is in the supplement. And finally, the issue of having a new instrument like IDSIQ can makes, of course, the discovery is interesting, but more surprising and difficult to interpret for some people. In the end, the next slide is showing again the conclusion from the paper and the fact that there are requirements for adding some of their treatment these days that is to improve both nighttime and daytime symptoms and that we really need to prevent the risk of next morning residual effect. And it seems that -- that's what is written the paper that daridorexant 50 milligrams seems to fulfill these requirements. And we have improvement in sleep variables without excessive sleepiness and good effect on daytime functioning. So this being said, next slide, please. I would like now to maybe give more kind of -- I don't see a conclusion on this development program, and this is shown on the next slide, Slide 47, which will be my last slide. We have a comprehensive sleep efficacy and perceived by the patient fall asleep faster, they still sleep longer. And the total sleep time was increased perceived by them and that was preserving with preservation of the sleep architecture. We also have daytime functioning perceived by the patients with formal statistics on the steepness domain as well as consistent efficacy on the other domain, although not statistically tested. And interesting is to see that this effect is maybe progressing over time and maybe not observed to the maximum effect very shortly. Finally, the safety is good, the low -- very low incidents of adverse events comparable between groups, very few serious adverse events and even mentioned them very, very few. There was no evidence of daridorexant dependence and no rebound. And all of that is probably due to the mode of action that described by Martine before, the precision, mode of action as well as the intrinsic properties of daridorexant. It's a targeted mechanism and an ideal pharmacological profile, which allow us to have all these conclusions and they are not automatically true for other products, but certainly for daridorexant. And with that said, I will stop, I see that it's quite late now and maybe take a few questions, if any.

Okay. Thank you, Guy, for your concluding remarks. We have come to the end of our prepared remarks and the slides and are ready to take the Q&A session full frontal. As this is a scientific call, please I want to remind everybody that please address your questions to this topic. We will have a financial call for the full year results in about 3 weeks' time and would defer those kind of questions to them. Operator, please open the lines.

[Operator Instructions] We have the first question coming from the line of Peter Verdult from Citi.

Peter Verdult from Citi. Given how thorough your presentation was on the science and the clinical data, it was difficult for me not to ask maybe a conventional question or 2. But can I just kick off to Guy, apart from supporting the Sleep Alliance and investigator-led studies. Are there any plans from an Idorsia point of view to conduct some sort of Phase IV program as you release the diary? And then just looking into your early-stage discovery efforts, are there any ongoing research efforts in insomnia or new mechanisms or targets that you would flag at this juncture? I'll pause there and then I have just 2 very quick follow-ups.

Okay. Thank you, Peter. Guy, do you want to take the Phase IV?

Yes, I can maybe start the answer to the question. Of course, with this -- there are 2 things I can mention. First, we are not at the end of the full publication of the data we already have from the trial. So you would see a lot more data coming gradually as we publish. And we also want to do more studies with a product of that quality. There are many areas of research that we are considering and working very diligently on creating new studies so that we understand better the efficacy and the safety of daridorexant within the indication, maybe in some subgroups. So this is work in progress, and I'm sure you will hear about that in the near future.

Okay. Thank you, Guy. Martine, do you want to comment on mode of action or other insomnia product that we may have in early development?

Yes. We have Phase II running at the moment in an indication of binge eating with a selective orexin-1 receptor antagonist. We should have a profile of a compound without a sleep effect and binge eating is a big medical need. We will have the results this year. And then beyond, it's too early to speak. But of course, we are interested in continuing in the fields in which we have been working. But so far, I think it's not yet visible. But it's a lot of research, which has gone into orexin and insomnia understanding.

Thank you, Martine.

And thank you, Andrew. Forgive me. Just if I may just squeeze one last one in just for the team. Look, I know you said you're not going to talk about financials, but I would like just to -- because now you've got the label in hand, you got Jennifer Aniston on board. You know what you're dealing with. You've got The Lancet Neurology paper. I mean the market is penciling in $1 billion at blockbuster status. But when you, as a team, talk about dario-QUVIVIQ, could you at least give us a hint as to what your expectations are and what you're seeing, how big do you think this drug can be?

Thank you, Peter. So Jean-Paul, do you want to take more of a strategic view on this, and then we can give maybe Simon, the opportunity to comment in more details on marketing and tactics here. Jean-Paul, you may be on mute.

I just wanted to say simply, sky is the limit. Frankly, we have an extraordinary drug for millions of people. So I think that this, as I said, this drug will ensure growth for the company for the next 14 years because it's a very long patent life. We will have to work very hard to gather more and more patients. We are going to give in a few weeks, the first -- our first vision for the beginning of the launch. But really, since I always said that zolpidem, Ambien used to sell for $4 billion 15 years ago, we should do better, but it will take time.

Thank you, Jean-Paul. Simon, do you want to have -- try to stitch together how you see the nearer term maybe and how we're putting tactics together with all of these elements that we now have laid out since the approval about 10 days ago.

Yes. I mean I just -- maybe to add a comment to Jean-Paul, I mean just to give you a sort of numerical framework. If a lot of people are talking about price and if you start with suvorexant price as a sort of benchmark. At a suvorexant price, 3% market share in the U.S. alone is $1 billion at list price. So you can look at the label, you can look at what we've got and work out what sort of penetration rates we can get. But for every 3%, you get about $1 billion in the U.S. at this price. So as Jean-Paul says, with the size of the market and the profile of the product that we have it's very significant. And then on top of that, we are actually very excited about Europe. We will be the first daridorexant receptor antagonist in Europe. There's a lot of excitement. There's been no innovation in the European and some new market for 20-odd years. And we're finding a great deal of receptivity towards the profile as we get into our discussions about Europe. So I think that's something else I think you can add on to the opportunity in the U.S.

Thank you, Simon Thank you, Peter.

[Operator Instructions] We have the next question coming from the line of Manos Mastorakis from Deutsche Bank.

Essentially, I just wanted to ask the extent to which you will be using this Lancet publication in your marketing efforts and discussions with physicians and all these engagements.

Thank you, Manos. Simon, do you want to provide a context as to how this fits in our ability to market?

Yes, sure. It's -- so thank you, Manos. It's critical, frankly, because obviously, the depth of data that we're now putting in or putting out there through the publication is much broader than we have in the label, which is often the case. But when you see all the IDSIQ data now in this publication, including exploratory endpoints, the VAS. And I think really critically, the somnolence data, I mean, to be looking at -- there's no difference between doses and somnolence, not that much different to placebo. And as Guy said, striking you look at the elderly population, you've only got 0.8% somnolence. So all of these data are going to be important. They are now in the public domain. And we do expect to be using them proactively and disseminating these data through our different communication channels.

Thank you, Simon.

We have not further question at this time, sir. Please continue.

Okay. Operator, thank you very much. Jean-Paul, do you want to make some concluding remarks.

No, I want to thank Guy and Martine. I just like to repeat that it took us really 25 years to get there, but it was really worthwhile. I think the data are going to frankly -- if at the beginning of the program, we would have said we would have this type of data, we would not have believed it. I think that it shows that to have looked for the ideal profile really made a difference. And I repeat, if you look at the difference within the same drug between 25 and 50, you understand that this program really and the drug have been really allowed us to really discover something which is going more than to improve many patients. I think we are going -- we are dealing with insomnia with a society problem. There is more and more insomnia. We have seen that with COVID, there are 20% more patients than usual who are claiming that they cannot sleep or really a problem. So I think that with this tool, with this drug, we are going to be able to change the life of million and improve the life of millions of patients. So this is very exciting. And I hope that -- and I invite everybody to look at this data to go into the details. It's a lot of data, but it's worthwhile to really make your opinion yourself. And I hope that we have helped everybody to understand better this paper and this development program.

Thank you, Jean-Paul. So we have passed the top of the hour. So it is time to conclude our webcast. Thank you for your ongoing interest in Idorsia. It is an exciting year that we're in right now. Our next scheduled press release is for the full year '21 results scheduled for the 8th of February. Until then, have a good time. Operator, please close down the lines.

Thank you, ladies, and gentlemen. That does conclude our conference today. Thank you for participating. You may now disconnect your lines. Speakers, please stand by.