Roche Holding AG (ROG) Earnings Call Transcript & Summary

Ladies and gentlemen, welcome to Roche virtual neurology investor event. My name is Henrik, and I'm the technical operator for today's call. Kindly note that the webinar is being recorded. [Operator Instructions] One last remark: If you would like to follow the presented slides on your end as well, please feel free to go to roche.com/investors to download the presentation. At this time, it's our pleasure to introduce you to Bruno Eschli, Head of Investor Relations. Bruno, the stage is yours.

Thanks a lot, Henrik. And then can I have, please, the agenda slide, first? So welcome to our first IR call in 2024, focusing on our neurology franchise and especially on the latest data which were presented at MDA a week ago and just over the weekend at AD/PD. So let me quickly take you through the agenda. We have 3 speakers with us today: first, Paulo Fontoura, our global head of neuroscience, immunology, ophthalmology, cardiometabolic, infectious and rare diseases clinical development. Paulo will provide us a quick update on our latest -- or late-stage neurology pipeline. Our second speaker with us is Dr. Alex Murphy, a senior clinical director within product development in neuroscience. Alex will lead us through the Phase III EMBARK results in DMD as they were presented a week ago at the MDA conference. And finally, we have Azad Bonni with us, our global head of neuroscience and rare diseases from pRED, so early development. He will provide us an update on the early-stage neurology pipeline. And Azad will focus on data which just got presented over the weekend at AD/PD and which we are quite excited about. These will be the Phase I, Phase II dose-finding data for trontinemab, our novel brainshuttle [ molecules ] antibody; and then secondly, the Phase II 4-year [ open-label ] extension data, PASADENA, from prasinezumab in Parkinson's disease. Overall, we have 90 minutes scheduled for the call, roughly 50 minutes for the presentations. And then we have, I think, 40 minutes for the Q&A. For the Q&A session, we will also be accompanied by Samir Megateli, our Global Franchise Head, Neuromuscular Diseases, and life cycle leader for [indiscernible]. Can I ask, please, the next slide? Just here as a quick reminder to our neurology franchise. I think Roche has recently become the #1 in neurology in terms of sales. Neurology sales account already for 1/5 of our total pharma sales, and this is expected to continue to grow in the double digits just based on the -- our market portfolio. As outlined here, you can see that we have the 3 products where we expect they will have additional opportunities ahead. Especially to highlight here: For Ocrevus, there is the Ocrevus 6-month subcutaneous approval to come mid-year in the -- in Europe and a bit later in the United States, which will give us an additional incremental sales opportunity. And also here to flag, Ocrevus high dose data which are expected in 2025 and which gives us additional then development opportunities, if positive. For Evrysdi, I just want to call out that we have reached now the target of being the most used SMA therapy globally. With around 15,000 patients, that's, I think, a significant milestone. And also, with the uptake rates we see here in all the countries, we believe that there is significant more growth to come for this compound. And then I think the third one here, to Enspryng, you see that we have 4,000 patients now in NMOSD. There is high treatment satisfaction with the monthly subcutaneous administration. And also here we have a development program running, I think Paulo will touch on it, with additional opportunities. Can we go to the next slide, please? On this slide, I just wanted to -- this is a slide actually Thomas has shown for the first time at the full year '23 results presentation. It shows a bit the late-stage pipeline updates which are scheduled for '24, '25. And you see here in the red box some of the molecules we will focus on today, especially about -- on the latest data presented for Elevidys, prasinezumab and trontinemab. Fenebrutinib, the data, we had touched upon in a previous IR call which was back in October. And looking forward, I also wanted to highlight here that we have upcoming IR events. We will touch on other parts of this news flow. We have the Diagnostics Day scheduled for May 22. We will focus on mass spec, CGM, NGS and novel point-of-care solution. And we also plan for an ASCO oncology update. The date needs still to be determined, but this development -- this IR call will then focus on the development programs for solid tumors, especially on the molecules listed here, tiragolumab, inavolisib, divarasib, giredestrant; and also on hematology development programs for Columvi, Lunsumio, Polivy also our allogeneic CAR-Ts. Can you, please, move to the next slide? This slide, I just wanted to show to give a bit the bigger picture how do the molecules we touch upon today fit into the [ bigger scheme ] of our emerging late-stage pipeline. On this chart, you can see the building blocks for mid- and long-term growth at Roche. On the left side, you see the launched portfolio, which is to deliver good growth momentum in the coming years, at least until '26, '27; and as mentioned on my previous slide, Ocrevus and Evrysdi here with further growth and development potential. In the middle section, you see all the -- which is defined here as the mid-term opportunities. You see all the NMEs which could be launched until the end of 2026. For neurology, this includes Enspryng in various other autoimmune diseases. You see gMG, MOGAD or AIE; and Elevidys in DMD and fenebrutinib in RMS and PPMS with readouts to come in '25. On the right then, you see the potential for long-term opportunities. These are all the NMEs to be filed after 2026. And of course, the molecules trontinemab in AD and prasinezumab in Parkinson's disease would fall into this box; as well as our anti-latent myostatin antibody in SMA, FSHA (sic) [ FSHD ] and potentially in combination with GLP-1 and GIPs. And with that, I will finish and hand over to Paulo for an update on our late-stage neurology pipeline. Paulo, please?

Thanks, Bruno. Good afternoon, everyone, good morning, wherever you are. So I'm going to talk you a little bit through our pipeline and then hand over to my colleagues for some more details. Can I have some -- to the next slide, please? So this slide depicts our current pipeline. And as you can see, it's a very broad, very deep pipeline with 4 medicines already launched in the market, several programs in all stages of development and, also you can see here, color coded by the general areas of interest. So we have a lot of projects around neuroimmunology and neurodegenerative diseases and neuromuscular, which are our 3 core focuses; and then some also other emerging pipeline. Also included here are the diversity of modalities that we usually go for now, so both small and large molecule but also gene therapies and antisense. So really again a very broad, industry-leading pipeline in neuroscience. Next slide. As Bruno mentioned, we're going to have a range of events and data flow happening over the next 1 to 2 years. You see here across our main 3 franchises, MS, neuromuscular and neurodegenerative, lots of data coming in from the Ocrevus subcu program and the high-dose program, from the -- our fenebrutinib BTKi program as well in 2025. In neuromuscular, obviously we'll get an update from the Elevidys team, but also then we're looking forward to regulatory interactions and submissions this year. And we have 2 active program which are recruiting right now. And for neurodegeneration and -- my colleague Azad Bonni is going to tell you about trontinemab and prasinezumab, where we're making really fast progress towards accelerating these programs to Phase III development. Next slide, please. And very briefly as well, I wanted to touch on the integration we have between Pharmaceuticals and Diagnostics at Roche. And of course, now neurology has become, for our Diagnostics colleagues, one of the main areas of interest as well. You see here there's 2 main focuses. One is Alzheimer's disease, where we have already launched the golden standard diagnostic CSF tests on the Elecsys platform for pTau181 and A beta 42; and that's approved by FDA. And at the same time, we're working fast on screening and triage tests, including an amyloid plasma panel that will be a rule-out test. So it would exclude patients at risk for Alzheimer's. And we are also working at the same time in a rule-in test using pTau217, which we'll move forward [ to development as we start ]. MS is the second big area for Diagnostics. And the main focus here is on developing the Elecsys NfL blood assay that would be used for disease monitoring not just in MS but potentially as well in Alzheimer's disease and other neurodegenerative conditions. Next slide. So I'll briefly touch on a few programs before handing over to my colleagues, the first of which is the Ocrevus program. As you know, we finalized the trials, the OCARINA trials, or subcu development. And these have demonstrated both the safety, tolerability but also the efficacy of the subcu every-6-months regimen. This is a significant advancement for these patients because it reduces administration time to just 10 minutes. And also it makes Ocrevus available for a wider set of treatment settings than currently is available. So especially for community hospitals and clinics, they don't have as much IV capacity or for at-home administration. The subcu formulation of Ocrevus really opens up a lot of potential for those patients. At the same time, we're fast finalizing our Ocrevus high-dose program, here on the right-hand side of this slide, that is based on our observation from our Phase III programs that higher levels of exposure to Ocrevus lead to higher levels of B cell depletion and commensurately to higher levels of progression, suppression. So we're really excited about this as an opportunity to really push forward the already really good effects we see in terms of control of disease progression in both relapsing and PPMS patients. Next slide, please. Our second program in MS that I wanted to highlight is fenebrutinib. This is a highly selective, noncovalently bound brain-penetrant BTK inhibitor that's in Phase III trials right now. Of course, we feel that fenebrutinib has a best-in-class and best-in-disease profile. It's [ the only ] noncovalent and reversible BTK inhibitor with the highest potency and the highest selectivity, which allows us to dose pretty high and maintain IC90 suppression of target over 24 hours with a very good safety profile. Last year, we presented the data from the Phase II FENopta trials in relapsing MS, this is here on the right hand of your slide, that showed convincingly the effects that fenebrutinib has in terms of suppression of disease activity as measured by total T1 new Gd+ lesions, which is a hallmark of inflammatory MS activity. Now we also observed in this trial CSF concentration levels that are above IC90 over the 24-hour period, which is a key aspect of this molecule because it leads us to believe we can actually suppress inflammation and microglial activation in the brain as well. Next slide, switching gears now to Enspryng. Enspryng is being developed apart from NMOSD in a range of other neurological conditions in which IL-6 suppression, we think, may play a role. And these are generalized myasthenia gravis, autoimmune encephalitis and MOGAD. As you can see here on the left-hand side, IL-6 suppression plays actually a fundamental role in several aspects of the immune response, whether that's the activation of B cells and plasmablasts to produce pathogenic antibodies or the activation and differentiation of proinflammatory T cells or actually controlling the blood-brain barrier integrity. And because of these data, we actually moved forward in trials in all these indications. I'm showing you here just a brief snippet of our preclinical data for an anti-IL-6 receptor antibody in generalized myasthenia gravis animal models that shows convincingly the control of disease that's possible with this mechanism. So our Phase III luminescence (sic) [ LUMINESCE ] trial is ended at -- recruitment in August last year. And we're expecting results, clinical results, in the first half of this year; and we're looking forward to those as well. Next slide. Finally, the last molecule I really wanted to highlight is our anti-latent myostatin monoclonal antibody GYM329. This is an antibody that's been engineered by our colleagues at Chugai to have both sweeping and recycling technology, which means that it achieves very deep levels of suppression of myostatin signaling with only once-a-month subcu dosing because the antibody gets recycled and goes on to suppress further. Now this is a very exciting mechanism for neuromuscular disorders, in this case. We're studying it both in spinal muscular atrophy and FSHD. These are both conditions with large unmet medical needs. And you see here in the middle of this slide the data from animal models, in which the use of this antibody, in combination with risdiplam, Evrysdi, shows convincingly increase in muscle map as well as increase in muscle strength in these animal models. Because of these data, we're now in the middle of recruiting Phase II trials both in SMA as well as in FSHD, as you can see here on the right-hand side of this panel. And we're looking forward to seeing data from these programs later on in this year. And I think, with that, I'm going to turn it over to Alex Murphy, who's going to tell you about the EMBARK data. Thank you.

Thank you very much. So first of all, I'm going to take through a little bit about the background. And then we'll move on to the data that was presented at the recent MDA congress by Professor Mendell. So if we move on to the next slide, please. So this slide is all about trying to get across really, first of all, what these patients go through and their families go through throughout their lives but also as a highlight that there is a big unmet medical need in Duchenne muscular dystrophy. It is a relentlessly progressive disease which is present and the damage is being done from birth. This diagram in the middle, if you see, that -- patients from 0 to 4 years. They really start to be a little bit slower than the average child in terms of their ability to start walking. And they don't meet their motor milestones at the same time as other children. And then as they get to 5 to 7 years of age, they really reach their peak motor function here. They are getting slightly better, but they are always lagging far behind their friends and their peers. As they go between 8 to 11 years and in their early teens, this is where they really start to slow down. They start to use wheelchairs more. And as their peers and friends are becoming more independent, these children are relying more and more on wheelchairs and caregivers to help them. As they go even older, towards the right of the slide, you see that their upper limb function is affected and they start to lose the ability to use their arms and eventually their hands. Their respiratory muscles and their heart muscles are also affected. And this leads then to a premature death at the age of, on average, 28 years of age. The treatment options are currently really limited. They're only really trying to slow progression and improve quality of life. There's no treatment which affects the underlying cause of Duchenne muscular dystrophy which is available to the majority of patients and applicable to all. Corticosteroids are the current standard of care. And supportive care is the other aspect that we see most frequently. Can we move on to the next slide, please? Thank you. So how do we measure Duchenne? And how do we measure ambulation? There's a few -- there's 2 slides here which I'm going to go through just to show you how the current thinking is in the field. There is the North Star ambulatory scale. This is a unidimensional functional scale. It looks at 17 different items. You can see them listed there on the left. Patients score either 0, 1 or 2, depending on how well they achieve the task they're asked to do. If they score a 0, they are unable to complete the task. If they score a 2, they do the task completely with no problem at all. And if they score a 1, then they have to use some sort of compensatory mechanism to help them to complete the task, but they do manage to complete it. Now while this is a scale which has been used widely in clinic and is highly clinically relevant, it's not a perfect scale. And we know that it was designed originally to show decline for patients over a number of years. And in certain populations, it may not be as sensitive as other measures. I'm going to show you an example of that now. If we look to the right of the diagram, we're just going to see a short graphic. And this is just showing you 2 patients. Now both of these patients, they score a 1 on the North Star score because both are able to do the task. However, you can see there's quite a wide difference in terms of their ability to get up from the floor here. The time that's taken is very, very different, but that's not being relayed by the North Star score. The overall score for both of these patients on this task is a 1. So we're going to move on to the next slide, please. And this is just going to say a little bit more about the timed function tests. So these are potentially more responsive to disease change, particularly in some populations. It -- if we'll start to the left of this diagram, we can see here this is just calling out, timed function tests, they are lost in a fairly sequential manner. The time to rise is particularly one which is -- which happens earlier. [ It starts ] to decline at an earlier age than the other timed function tests, but you can see here that they are quite sensitive in terms of disease progression. To the right of -- sorry. I should say, to the middle, actually, we have the time to rise test here. And this is actually something which is taken from a large study, 293 patients, so very large, for Duchenne -- of patients in the U.K. And they were looking at the age between 6 and 7.5 years of age and the time to rise at that age and whether that could help to predict loss of ambulation. And so if -- what they found was really, if you have a time to rise greater than 5 seconds, you are more likely to be in this blue line you can see in the middle there, which the patients are losing ambulation at a much earlier age, whereas if we compare that to the patients who are below 5, so those are the other 2 bars, you can see that they lose them at a later age. And so we use this as a milestone really to say that, once you get over 5 seconds, if you're at that kind of age between 6 and 7.5 years of age, then you're more likely to lose your ambulation at an earlier age. The other thing on this slide you can see, to the right, is a highly objective measure of ambulation. This measures the stride velocity 95th centile or what we call the SV95C. You use a wearable device. You can see how it's placed on the child there. This measures the stride velocity over a period of several weeks. The child basically puts this on in the morning, puts their sock over the top of it and then just forgets all about it, but it gives nice, highly objective evidence, where as I said the child really forgets they're wearing it. [ So it gives ] nice real-world evidence of how ambulant and how quickly they're able to move about. This has been qualified by EMA. The SV95C has been qualified for use as a primary end point in clinical trials of Duchenne muscular dystrophy and, as I say, highly objective evidence. So we move on to next slide, please. So this brings me on to some of the slides that were presented at the MDA meeting recently by Professor Mendell. Can we move on to the next slide? So just at the very top of the slide here, you can see that, as of February 2024, we know that delandistrogene moxeparvovec, the gene therapy, has been approved in U.S.A., UAE, Qatar and Kuwait for patients between 4 to 5 years of age with Duchenne muscular dystrophy. And we're going to talk a little bit about the EMBARK trial, also known as the 301 study. This is a 2-part randomized double-blind placebo-controlled trial of patients who are between [ 4, 4-plus ] and 8 years of age. We're going to show the 1-year safety and functional outcome measures. And to the right of the diagram, you can just see a small graphic which just depicts the gene therapy. I think the most important part for safety is, of course, the vector, which is AAVrh74. In the middle, you can see the transgene. This is the most important part of -- it's kind of the package that we're trying to deliver into the cells. And this produces a shortened version of the dystrophin protein, with all of the important parts inside, to give muscle membrane stability and to try and do the same thing that dystrophin does, dystrophin protein. If we move on to the next slide, please. Thank you. So I'll just take you very briefly through the study design. So we had 125 patients, who were randomized in a 1:1 ratio to either receive gene therapy or placebo. A subgroup of patients had a muscle biopsy at week 12, but all patients, if they received gene therapy in part 1, they receive placebo in part 2. And all those patients who received placebo, they receive gene therapy in part 2. Patients then at the end of part 2 get taken into the [ 305 ] study, which is a long-term follow-up study for up to 5 years post dosing. Patients were stratified along the lines of both age, as you can see at the top of the diagram; and the North Star score. The key inclusion criteria that I would just draw your attention to include the North Star score of greater than 16 and less than 29 and a time to rise of less than 5 seconds at screening. These were put in place to homogenize the population. We also had, as you can see to the right -- we've already talked about the end points here, but just to call these out: As a primary, we saw the North Star score to week 52. We had -- 2 key secondary functional end points were identified, which were the time to rise and 10-meter walk/run. And we had some secondary functional end points. SV95C was one of these; 100-meter walk/run and time to ascend 4 steps. If we move on to the next slide, please. I'm not going to take too long on this. This is patient demographics. Really I think the key message here is we're -- thanks to stratification, we are really quite well balanced in terms of both age and functional ability at baseline. You can see the North Star score and the time to rise both fairly similar, and the ranges were fairly similar as well. If we move on to the next slide, please. The overall safety was very positive. We saw that the safety profile of our gene therapy was consistent with our experience from early-phase studies. We saw no new signals. The adverse events were medically manageable with appropriate monitoring and treatment where appropriate. And there were no clinically relevant complement activations. We've not seen that throughout our program. We've had no deaths and no study discontinuations. If we move on to the next slide, please. So that brings me on to the efficacy results. So you can see here this is the primary end point, the North Star score to week 52. And the kind of -- I suppose I -- we struggle to describe -- the color purple, I suppose, the top line, is the gene therapy group, whereas the gray line is the placebo group. You can see there is a clear separation at all time points here. However, this is not statistically significant at week 52. We're going to move on then to show you a little bit about the rest of the functional assessments. And you can see here this is using standardized statistics. Everything which is basically to the right of that line in the middle is favoring the gene therapy group. So you can see here, if you just look from a distance, a very consistent effect which is being shown, that we're seeing, the secondary -- the key secondaries and the other functional assessments, they're all in favor of the gene therapy group. Now as you can see, we just mention the primary did not reach the statistical significance. However, the key secondaries, the time to rise and 10-meter walk/run, these were both nominally statistically significant and highly clinically relevant results. We also saw for the SV95C, which is this objective measure. You can see again this was -- this is actually nominally statically significant in favor of the gene therapy. The 100-meter walk/run was again in favor of the gene therapy but not statistically significant. And the ascend 4 stairs, again, was nominally statistically significant and again in favor of the gene therapy. So you can see. Just by looking at this slide, you can see that here there is consistent evidence of a treatment effect which is going on within this population from the study. If you move on to the next slide, please, just to focus a little bit on the 2 key secondary outcome measures, the time to rise and the 10-meter walk/run. Here it's worth saying that this is a timed function test. So it's how long it takes you to do a task, so if it takes you less time to do the task, that's an improvement. So we're looking here for a minus number, which will show an improvement. And you can see at the top of this slide that we saw a between-group difference of minus 0.64 seconds between the treated group and the placebo group, again in favor of the treatment group. If you move on to the next slide, please. And of course, I mentioned earlier on about this very clinically important milestone of going over 5 seconds in terms of the time to rise. And so we looked at this in a post hoc analysis. And again here we see the gene therapy group showing more evidence that it is having a disease-modifying effect, with the patients in the treatment group here being -- only 3% of them went over this time to rise of greater than 5 seconds, whereas it's 16% in the placebo group, showing a reduction in odds of 91%, which was statistically significant. If you move on to the next slide, please. Here again this is the 10-meter walk/run. Again you can see that there is a clear difference between both of these different groups, with the treatment group having a between-group difference overall of minus 0.42 seconds, again in favor of the treatment group. And again this is seen as being clinically relevant. If we move to the next slide. So overall, our conclusions. The safety findings, we didn't see any new safety signals. And we demonstrated a manageable benefit-risk profile. We -- as I mentioned, though we didn't reach statistical significance in terms of the primary end point, however, between-group differences favoring gene therapy on the secondary functional end points indicate potential for long-term disease modification in Duchenne muscular dystrophy. And we also discussed about the post hoc analysis here of the time to rise, where we showed fewer patients in the treated group getting over this important milestone of greater than 5 seconds, which is a prognostic marker for earlier loss of ambulation. So we conclude here the totality of evidence indicates that the gene therapy produces potential benefit and disease trajectory modification, versus placebo, with a manageable safety profile. If you move on to the last slide, I think. So this just really highlights our overall clinical development program. This is the largest trial program to support the broader Duchenne patient populations, I would just draw out here, as opposed to studies which are the actively recruiting one, the [ 302 ] study with the younger patients from 0 to less than 4 years of age. And then we have the [ 303 ] study, which is our patients who are in the late ambulant phase and the non-ambulant patients there as well, that is currently recruiting. And I think, with that, I will hand over to my colleague, Dr. Bonni.

Thank you, Alex. So it's great to be here, and thank you for joining. So I'm going to give you an overview of our early-stage pipeline in neurology and give you a couple of updates on a couple of exciting projects. So if you go to the next slide, you will see a slide that you saw earlier from my colleague Paulo which shows the clinical development portfolio. And I just want to make a couple of points here. One is that this is really an industry-leading portfolio, including in the early stage Phase I, Phase II trials. And as was mentioned earlier, we're in broad areas of neurology. And in the early stage, we're in a number of diseases, including but not limited to Alzheimer's disease, Parkinson's and MS. And of course, there's also a large research portfolio not shown here that precedes this. And the other point that I want to focus on is the technology that we're using; and in particular the different types of technologies that we're using, including the brainshuttle technology, which I will highlight actually in the first update on trontinemab. So if you go to the next slide. Basically here what I wanted to tell you was that I'm going to tell you about this update which was just given a couple of days ago at AD/PD. And in this project, as you may know, we're using trontinemab, which is a novel molecular entity that deploys the brainshuttle technology that's been developed here in pRED, at Roche pRED, for the last over 15 years. And actually what I'm going to do first is -- I wanted to start out with a video to help make the points about how this technology will help us generally and including in this particular project, so next, we will go to that video, which will be about a couple of minutes.

As the brain's gatekeeper, the blood-brain barrier represents a significant challenge to medicines that need to be delivered to the brain, limiting access to potential treatments such as for Alzheimer's disease. To overcome this obstacle, Roche has developed the brainshuttle, an innovative technology that uses transferrin receptors on endothelial cells to smoothly cross the blood-brain barrier while carrying its therapeutic cargo. Once inside, the brainshuttle antibody disperses throughout the brain to reach its site of action; in this case, amyloid plaques, the hallmark of Alzheimer's disease. Bound to the plaques, the antibody engages with receptors on microglia and triggers the clearing of amyloid from the brain. Unlike conventional antibodies whose uptake and distribution is limited to areas where the barrier between brain and vasculature is not as tight, the brainshuttle enables a high concentration of the Alzheimer's plaque-clearing antibodies to all affected brain regions, accessing even deeper regions through the dense capillary network. This more homogenous brain uptake can potentially lead to many more benefits such as reduced local vascular inflammation with antibody treatment for Alzheimer's disease. The brainshuttle represents a breakthrough that offers many new possibilities for the efficient transportation of large molecules across the blood-brain barrier, ushering in more effective treatments for a variety of brain diseases. At Roche, we embrace this exciting potential, a ray of hope for anyone battling a brain disorder.

So as you can see, we're excited about this technology broadly. I'm going to focus here on trontinemab with this technology. And this basically -- I'll be able, through the slide, to go through some of the key points that were made in the video. So essentially trontinemab is this novel antibody -- molecular entity that targets aggregated Abeta and amyloid plaques. And essentially what it does -- through the transferrin receptor, through that fragment of the agent, binds to the transferrin receptor and hijacks this transferrin receptor system to essentially allow for enabled transcytosis across the blood-brain barrier. As transferrin receptor is enriched in the capillaries of the brain, that allows this to happen specifically in the brain. And this is done in a way that actually does not interfere with the function of transferrin. And so once it's in the brain parenchyma, of course, it leads to its action, which as was shown in the video. And so from a pharmacological approach, this technology allows us to get higher exposure, superior target engagement and a broader distribution in the brain. And the next slide makes that point more clearly in mouse studies in this particular experiment where an IgG or a brainshuttle version of the IgG have been injected intravenously. And 5 days later, we're looking at the fluorescence of whole-brain mounts. You can see on the left side, with the standard IgG, you find this in the areas that are -- in the choroid plexus as well as in the ventricles and the meninges. There's a lot less, as is typical, in the parenchyma itself. However, with the brainshuttle IgG to the same target, you see -- on the right side, you see it's much broader and homogeneous; and this allows also accessing the deeper regions. So now I'm going to give you -- go to the -- going to the next slide, I'm going to now tell you about -- specifically about trontinemab. So as you know, this is a Phase IIa, Phase Ib study, a dose escalation study. And essentially here what we have are 4 cohorts. This is a staggered, parallel-group adaptive design. And we're going from cohort 1, at 0.2 milligrams per kilo, all the way to cohort 4, at 3.6 milligram per kilo. And the study participants are randomized 4:1 to active versus placebo within each cohort. Now we are also taking the cohorts that are -- where there would be encouraging results, to expand them over this year. And so this will give us a broader understanding of the safety and profile; as well as a more robust assessment of the safety, PK, PD and immunogenicity. And as you may know, at CTAD last year, the first 3 -- the results of the first 3 cohorts were presented, which I will show. And then at AD/PD, we showed the results of the fourth cohort. I should add that this is done in participants who are amyloid positive. They either have MCI, mild cognitive impairment, due to Alzheimer's disease or early Alzheimer's disease. And this is basically the injections are every -- once every 4 weeks, for a total of 7 doses. So if you go to the next slide, first, I'm going to tell you about the -- sort of the baseline characteristics of cohort 4 relative to cohorts 1 through 3. I do want to say, of course, that for cohort 4 here we're looking at 12-week data, whereas for the first 3 cohorts, we had 28-week data that were shown, but by and large, the characteristics are fairly similar down the line, as you can see. Now if we go to the next slide. These were the data that were shown at CTAD last year, so what you find here -- so basically what we're looking at here on the Y axis is the mean amyloid reduction from baseline in centiloids. And so what you find is that there is a dose-dependent reduction in the first 3 cohorts. So with 0.2 milligrams per kilo, you see there is a 28 weeks difference of 20 centiloids. And then you see 31 at -- with 0.6 milligram per kilo. And at 1.8 milligram per kilo, you see a rather robust reduction at 28 weeks that's at 84 centiloid difference. And you actually also see, as you can see on this slide, at week 12, there's a 62 centiloid mean difference, so the question is what happens with the higher dose as you go from 0.6 to 1.8. And that's what was shown at AD/PD just a couple of days ago and shown on the next slide. So as you can see, there is actually further pronounced reduction of the amyloid load at 12 weeks. This is where we have the data of -- currently for this cohort. So you're seeing that there's actually a reduction of 91 centiloid mean in the participants where this has been studied and analyzed, so even deeper than what you see at 28 weeks for the 1.8 milligram per kilo. And if you go to the next slide, now we're looking at the mean amyloid burden actually both to get a sense of the baseline and also to get a sense of the mean amyloid burden at 12 and 28 weeks. And so for the 3.6 milligram per kilo, what you're finding is that we actually have, at 12 weeks, more -- about half of the participants have -- are becoming amyloid negative. And actually, in 4 of these 5 patients, they're deeply amyloid negative in that it's less than the 11 centiloid. As you may know, the 24 centiloid is the threshold for amyloid positivity. And this is happening even though we found that -- in this cohort actually that the baseline was high at 120. So the take-home message from these data or from these results is that trontinemab is moving a lot of amyloid in a short period of time, which is really quite remarkable, in my view. Let's go to the next slide, which I want to get -- give a sense of the safety profile. So here we're looking at 12 weeks for cohort 4, as I mentioned, whereas for the first 3 cohorts, we're looking at the 28 weeks. And basically what we -- what I can say is there -- for cohort 4 there have been 0 deaths and no serious adverse events that are related to the drug -- to the study drug. There were 2 serious adverse events, as shown in the right column, but these were deemed by the principal investigators to be unrelated to the study drug. And so let's go to the next slide. And I want to tell you about infusion-related reactions and ARIA, so on the top part of this table you -- is the illustration of the numbers of the -- and percentages of the infusion-related reactions. Let's start with cohort 3. As you can see, in cohort 3, in the third column, we do see common infusion-related reactions. This would -- this occurred at the first dose and this occurred prior to premedication. With a protocol amendment, we implemented premedication with paracetamol or NSAID. And as you can see, in cohort 4, we actually see fewer now, significantly fewer, infusion-related reactions. And in ongoing observations, we see even less. And we expect that this could be suppressed even more with use of corticosteroids. So essentially suppression of the IRRs. We see we had cases of mild reversible anemia in cohort 3. And so far in cohort 4, we just have one case of mild anemia. Now let's go to the bottom part of the bottom table, looking at ARIA. And of course, with the caveat that this is still at 12 weeks, it's remarkably we see no cases of ARIA thus far in cohort 4. As you know, in cohort -- you may know, in cohort 3, we had 2 participants with ARIA. I should say that all of these results are shown blinded to the individual, so we're showing you cohort-level data. And that is important to maintain the blind status of the individual participants in the study. Now if you go to the next slide. This shows, basically gives you a summary of the findings that I have conveyed to you; and to reiterate that trontinemab is this novel, new -- it's a new molecular entity that uses, deploys the brainshuttle technology developed in pRED. And this is through transferrin receptor mediated transcytosis at the capillary level. And what we have found, thus far, is that we have rapid and robust amyloid plaque reduction at relatively low doses. And so far, we see this, that about half or more than half of the patients that we have looked at are amyloid negative at 12 weeks with the 3.6 milligram per kilo cohort. Of course, we need to -- these are still a few patients, but we need to expand and continue this. And that's really the last point, which is those results, together with the sustained low ARIA incidents and overall favorable safety and tolerability profile, support further investigation of this molecule in this ongoing study. And on my last slide for trontinemab, I want to put it a bit in context as shown on the next slide. And of course, this is obviously -- were -- this is not a head-to-head comparison but just in terms of context relative to what's been published. As you see there, we have graphs here with a number of colors. So the orange and blue, those are from gantenerumab. And the gray and yellow are lecanemab and donanemab. So for those two, as you can see, and for gant as well, there is amyloid reduction that takes a number of months to get to amyloid negativity. And what we're showing now to you with trontinemab, at 3.6 milligram per kilo every 4 weeks, at 12 weeks, we're seeing about half the patients are becoming amyloid negative at this early stage, so really quite -- again a quite remarkable result. And now I'd like to go to the update on prasinezumab. This was also presented a few days ago, on Wednesday, at AD/PD. So just to remind you: In PASADENA, we're assessing the effect of prasinezumab, shown on the next slide, in Parkinson's disease. Of course, in Parkinson's disease, symptomatic treatments are available. However, no treatments are available to slow the progression of the disease. Prasinezumab is a humanized IgG1 monoclonal antibody that selectively binds aggregated alpha-synuclein. The idea is that aggregated alpha-synuclein, in the course of the disease, is released from degenerating neurons; and that, that leads to non-cell-autonomous effects basically through -- on other cells leading to degeneration of other cells that are exposed to aggregated extracellular alpha-synuclein. So the idea with prasinezumab is that this selectively binds aggregated alpha-synuclein in the extracellular milieu. And that should lead to prevention of cell-to-cell transfer of pathogenic alpha-synuclein aggregates and degeneration. And as you may know, in the PASADENA trial, we had evidence of signals on a prespecified secondary end point of MDS-UPDRS Part III that was corroborated with digital readouts as well. And so we wanted to see -- we have been following these patients in an open-label extension to see what happens over time after the Parts I and II of the trial. Now the next slide will give you a sense of how we're doing this. So in -- under PASADENA, you'll see that we have 2 arms, the delayed-start and the early-start cohorts. So in the early start, these participants received prasinezumab from the get-go, from the start, whereas in the delayed start, they received it at year 1. And the study ended at Part II, at year 2. And then the open-label extension begins from the end of year 2. And now we have data that I can show you up to the end of year 4. Now what we're doing here to kind of get a sense Because we're looking at open-label extension. To get a sense of how these results -- how to interpret them, we are comparing them to essentially a cohort, an observational study from the PPMI cohort supported by The Michael J. Fox Foundation. And we are doing this using an approach that's shown on the next slide, basically where you balance the characteristics after weighting with propensity scores. And so basically when you look at this on this table. The second column has PASADENA with [ NF 271 ]. And look at the sort of the right-hand part, in bold, the PPMI weighted. Basically when you do this, you ensure that the characteristics are similar. And so the SMD, which is basically a mean difference that you're looking at, that should be less than 0.2. We see, across the board down the line of these characteristics, we are -- we have similar characteristics. So this gives us the opportunity to make a comparison to essentially real-world data, although this is an observational study, so let's go to the next slide, which shows the data. So what you will see now is that, with the open-label extension, the Y axis is basically adjusted mean change from the baseline for the MDS-UPDRS Part III scale in the OFF period, which is basically a -- reflects motor signs. And here what you see, in the gray, is that, every year, there is an increase of about 3 points. And this is consistent with what is seen in practice and in the literature. And for the PASADENA prasinezumab-treated individuals, what you find for both the early- and delayed-start cohorts, by end of year 1, you don't see much of a difference when compared to the observational study data. However, starting at year 2 and all the way to year 4, you do see separation, by year 4, about 51% to 65% difference relative to delayed start and early start. So of course, these are open label -- this is an open-label extension. Of -- and needs to be interpreted with that in mind. On the other hand, it is interesting to see that these patients who are on prasinezumab are actually quite stable in terms of the Part III scale. Now let's go to the next slide, which is the last data slide. And here we're looking at Part II. And this is motor experiences of daily living, so essentially their motor symptoms. And basically, in the literature, it's known that this moves less fast than Part III. And so what we're finding is that actually, even with Part II with the -- in the open-label extension, once you get to 3 years and in particular 4 years, that -- there appears to be a different separation from the observational study points, as shown here, with a 40% to 48% difference. And so the next slide just summarizes our findings that were presented for the PASADENA open-label extension. Essentially what we're seeing in this open-label extension is slowing of progression of MDS-UPDRS Part III as well as Part II. And I will let you read the rest of the slide. Of course, these findings are exploratory and need to be confirmed in an independent trial such as the Phase IIb PADOVA study and its open-label extension. And that is shown on the next slide, and this is really the last slide. So this is an ongoing study. And data are expected in the second half of this year, where we're looking at patients that are a little bit later in course of the disease but still early-stage disease, on stable L-dopa -- or on stable symptomatic treatment, randomized to placebo or prasinezumab and treated for a 70-week period. And the primary end point here is time to confirmed motor progression event at 5 or greater points on the MDS-UPDRS Part III in the OFF state. And so this -- if this trial proves positive, obviously this will be an important result. And of course, we're awaiting the results of this trial. So with that, go to the next slide. This is our "statement of purpose" slide. And I thank you for your attention.

Thanks, Azad. And with that, I think we open the Q&A session.

And the first question comes from Andrew Baum from Citi.

So just could you talk to your development plan for trontinemab in Alzheimer's? Are you going straight into Phase III? You're doing a sort of seamless Phase II, Phase III. And could you also talk to, to what extent is it possible to use the transferrin shuttle, both bispecifics and ADCs, for brain mets? So I'm curious on that. And then separately, on the [ side allos ], could you talk to when you anticipate to initiate a trial in autoimmune disease with the CD19, CD20?

Do you want me to start with tronti?

Yes, please.

Okay, I'm going to start. And Paulo, please add. So in terms of tronti, we're really, really focused on the current trial; and that is the Phase IIa trial. And what I can say is that we are completing this, what you saw in the slide, the cohort; and also, of course, looking to expand the doses that look promising, as you saw, to get a sense of -- we need to get to a good number of patients, around a couple of hundred patients, to see that -- to confirm that these data, which look very promising -- that these data hold up and are confirmed. And of course, we need to have enough data on safety as well. Now just to say, from the literature, what we can say, and from the field, what we have learned, is that the faster and the deeper that you bring down amyloid, the more likely there's correlation in all of these studies in terms of an effect on the clinical point. In terms of plans beyond that, really we have to wait, in my view, but Paulo, please add, to see what the results of the Phase IIa will be.

Yes. Nothing much to add, Andrew. I mean we're still obviously very excited about the data. And we'll try to move as fast as possible, but there's still data that's coming in that's going to be really relevant for us to inform in terms of trials and designs. And your second question, I'm not quite sure I understood. That was about bispecifics with CD19...

Bispecifics...

Well, yes. I mean if you've got -- well, if -- for a patient with brain mets, obviously, having greater concentration of a relevant...

Yes, yes.

Right. So it would seem to be an obvious and given [ they're ] larger molecules, unless there's some structural reason why you can't use...

Yes, maybe Azad is better placed than I am to actually talk about that. I -- my experience with the brainshuttle is that each molecule kind of behaves in its own way, so there's a lot of optimization that needs to happen. I don't think in principle there's any major limitations, but Azad, I mean, you are the real expert here.

Yes. So for the brainshuttle, Andrew, the -- in terms of we actually have another molecule in early development in Phase I. And that is the brainshuttle CD20 for MS. And of course, in the research phase, we are -- this is an area that we've been working on for some time. And now that we see the results with trontinemab, we're looking at a whole slew of possibilities in terms of antibodies, to start with, of course, with the focus being on neurological disorders and -- but also, beyond that, modality. So that's a really good question and something that is of great interest to us.

Andrew, did we answer all your questions?

Yes. That's fine.

Okay, then we move on. And the next question will come from Tim Anderson from Wolfe Research.

Just going back to trontinemab. Where is Roche headed now in terms of the debate about continuous dosing versus finite dosing? You're obviously getting very rapid plaque reduction. In the past, you've aligned with Eisai and Biogen in terms of feeling you need to dose these antibodies continuously, but when I'm looking at the tox profile and just the rapidity of plaque reduction, it kind of begs that question again. And on anemia. On Slide 43, of a footnote; and you say that you think anemia might be related to frequent phlebotomy. And just kind of I guess I'm a little bit skeptical of that because there's a mechanistic reason, which is high expression of transferrin receptor in red blood cells. So how do you know that's not really what's driving anemia? And again that goes back to this idea of maybe you only dose these antibodies until you get plaque negativity.

I can start with the second question and then -- [ and hand it ]. So I -- so first of all, of course, we are very looking at this carefully. You have to remember we're looking at these in cohorts, right, not at the individual level. The anemia, it's -- turns out to be mild and reversible. And of course, we're dealing with patients who are -- participants who are sort of at an later stage of life and they may have low iron levels. And so we think that, just from that perspective, this is something that can be addressed. The reason that, that is shown as a potential explanation, that it could be from frequent blood draws, is that we're seeing this across, not just -- we're seeing it also in placebo, so it's probably more reflective of the population that we're dealing with rather than the study drug. I take your point. It's well taken because there is -- there can be an explanation. Do remember that we kind of selected doses that would not lead to effects elsewhere, so that is something to also bear in mind.

Yes. Maybe I can comment on the first question. I think it's actually still too early to say whether intermittent dosing, in whichever form one thinks about it, will lead to the same clinical benefits or not. So our base case plan still is that we're going to keep dosing, also because what you measure by amyloid path is a fraction of the available amyloid-binding potential of trontinemab, right? Trontinemab binds both to plaque, oligomers and fibrils. What I think it does open up the possibility is for less-frequent dosing to achieve the same clinical benefits. And that's obviously something we'd might be interested in mostly from a patient convenience and health care burden standpoint, but this concept of dosing to negativity and then stopping and seeing what happens, I think that opens up a lot of clinical questions which are not answered yet, so that's still not our plan.

Can I just -- is it possible that you would run a trial in Phase III that would actually have 2 different arms, 1 that would be finite dosing and then 1 that would be continuous dosing? Because that's the only way you're going to know if there's clinical benefits from a Phase III like that.

Yes. I mean everything is possible. Not everything is probable, right? Right now what our base case is -- as I mentioned, is we're going to keep dosing chronically to find out what is the best efficacy that can be achieved with trontinemab. Whether -- in the future, based on data, which they may become available and which includes obviously long-term safety, tolerability, adverse event profiles, et cetera, if that becomes an option that's interesting, of course, we have to do that experiment, but that is not our base case right now.

Tim, I just wanted to add back to your first question. When I say anemia is mild, it's asymptomatic as well. Just so that -- it's quite mild, just to make the point...

Okay, next question would go to Richard Vosser of JPMorgan.

Two questions, please. Just on the trontinemab data, I noticed again that not everybody is reaching the dose time points. So 5 patients didn't reach the 28-week data for 1.8 milligrams. 4 didn't reach the 12 weeks for 3.6, so if we want to understand the amyloid reductions, do we need to know what reductions are for those patients? I mean I can understand patients might discontinue, but when we looked at the open-label extension data for gantenerumab back in the day, they showed a good reduction in plaques, but ultimately that didn't translate in GRADUATE, in the GRADUATE trial. So I'm -- I suppose what I'm asking is how reproducible is this. And then the second question is just looking at the functional end points. So going on to the DMD program. The 100-meter walk distance and the 10 meter, [ one was static ], the 10 meter. And the other one, the 100 meter, wasn't. I would have thought the 100 meter would give a -- maybe a better representation of the benefits because it's a longer distance. So just what do the regulators think about that, do you think? And I suppose what's also taking the time for filing? What does the regulators want from you that you need to produce for -- all the data seems to be in house for a while, so why filing in the middle of the year as well?

I can start with the first. So in term -- to your first question, for the cohort 4 at 3.6 milligrams, I -- we have just shown 12-week data. We don't yet have the 28-week data, as you could see from the graph. And as we -- as the graphs showed, in 5 out of 8 patients, you see that they are becoming amyloid negative. Again, for the actual data in terms of it's shown for -- as a group, bear in mind that -- as I mentioned, that it turns out that in this group the baseline is 120, which is actually higher than what typically has been the case in our other -- definitely in our other cohorts but also in what you see in the trials where people -- so you, we're starting with a very high -- relatively high level of amyloid in the group and we're bringing it down quite a bit as a mean, as you saw, with minus 91.

Yes, I understand that, but 12 patients start and only 8 patients reach 12 weeks, so I'm just -- the -- what's happening to the other patients? Do they actually see a reduction? Do they discontinue? What happens to them?

Okay. So basically the -- we're looking at the ones that were analyzed. They're still in process, so there will be more data that will become available.

And the same for the 28-week data. I mean we've had those data for a while, since October, for the 28-week data [ for the cohort before ].

Yes, yes. That's right. There will be more data as more -- as we analyze more data, well, it'll -- in due course, it will become available. Did that answer your question?

Okay. And -- somewhat, yes.

Maybe I can add. I mean, what you have to understand, this is a Phase I trial, so it's not like everybody gets dosed at the same time, right? So we're -- as we recruit, they progress. And we get the end points as they come in, right? So the data is just coming in staggered and not in boluses, if you want. What Azad has shown you is the data as it is now, so he's literally giving you like the top line results as they stand at this point.

Sure.

You had a second question, I think...

[ Yes. Alex, do you want to talk about that ]?

[ It was towards ] the DMD team. Yes, I'm happy to comment on this. So thank you very much for the [indiscernible]. So the -- it was around the 100-meter walk/run, wasn't it? And I think, for the second part, it was regulators aspect. I think I will pass it over to Samir at the end. But I think the key thing I'd point to, first of all, is, I mean, the totality of data. We showed that slide earlier on which just showed the consistency of the functional end points. I -- obviously, the 100-meter walk/run, as you said, it wasn't nominally statistically significant. We know the 100-meter walk/run. It is a longer test for these boys. These are very young boys 4 to 7 years of age. I'm sure, if you -- if ever you have kids, then you'll know that they can be distractable. So the longer the timed function test is, the more likely it's going to be a little bit more variable. The 100-meter walk/run also looks at -- has other aspects to it as well because it's a little bit more endurance. It's not just something that any kid could do just straight off the bat, so I think it's important to say, things like the 10-meter walk/run, the time to rise, they are directly correlated. Well, the time to rise particularly is correlated to this loss of ambulation, but the other thing I would really point to here is the SV95C, which really does -- in some ways gives even more data on the endurance. Also, just the general mobility, the real-world data aspect of it, it really should not be overlooked. And it gives objective evidence. And while no one has used -- to my reckoning, no one has used the 100-meter walk/run as a primary outcome measure, we know that [ several of these others ], including the time to rise, the 4-stair climb and the SV95C, they're all validated as primary outcome measures. And so they're all highly sensitive. No one has used the 100-meter walk/run, perhaps for a good reason, but in spite of all of that, it was in favor of the gene therapy group, so I think it's really worth pointing out we see this consistent evidence. And I'll hand over to Samir, about the regulator aspect.

Yes. Thanks, Alex. Just I'll also reinforce what Alex just said. Also, a digital end point is an objective measure. And this has been also validated by EMA as a primary outcome measure as of last year. So we are engaging with the different regulators according to plan. And we are planning to submit by mid of this year, so we took our time to look really carefully at the data to build a robust dossier. And we are having [ fruitful discussion ] with the different regulators.

Then we have the next question coming from Stephen Scala from Cowen.

Two questions. First, Lilly mentioned last week that it also has a brainshuttle program. Curious if Roche has visibility on how the Roche and Lilly programs are similar or different. And secondly is the view that faster plaque reduction is beneficial based solely on donanemab clinical results. And therefore, are you of the view that donanemab is superior to lecanemab based on the clinical data we have to date?

Shall I start? Maybe I'll start with the second question. I mean obviously we'll -- I'm not going to comment on donanemab versus, well, lecanemab. I think it's fair to say that our own modeling shows that there's a pretty good correlation between depth of amyloid removal and clinical benefit. So that alone is a, I think, pretty solid fact by now. The second is that the speed at which you reach that amyloid negativity status does seem to matter. And therefore, that's why we're excited about trontinemab. And regarding your first question. Again Azad is the expert here. My -- in my experience, there's lots of different companies with different platforms looking at this. Really the only that has clinical data, so far, that's pretty robust is ours, so in theory, these things may look equal or different, et cetera, but I think the proof is in the pudding literally. It took us a long time to get to having clinical data. These are not technologies that are easy to manipulate, and therefore, I would hesitate to make a lot of comparisons between drugs that have different levels of evidence associated with them, or platforms, in this case, but Azad, I don't know if you want to add anything.

Just a bit, not much to add. I -- for the first point actually, it's based on a lot of trials and agents. It's not just on a couple of agents, this correlation of speed and effect, so it's not based on one. And in terms of the other companies, yes, really little to add. We know about the -- a number of companies that are interested in brainshuttle. Of course, this is, has been the holy grail in neuroscience therapeutics. What we can say is, for Alzheimer's disease, trontinemab is the most advanced. And we have the results that I've shown you thus far that have been shared with you.

We go on to Colin White from UBS.

Can you hear me?

Yes, we can hear you...

Colin White from UBS. I have a question on prasinezumab, for me, please. When we see the results of the PADOVA study: Obviously a significant proportion of the patients have been treated with L-dopa. And it's a little bit more of an advanced population, so I'm just wondering what you think that might mean in terms of what we see; also thinking about it in terms of how you've shown today, that in the first year, you might not see much of the separation, but then you might see more of an effect in later years. So just how -- what should we be looking for when we get the PADOVA data? What do you expect to see? And then -- and the second question I had was about the design of a potential Phase III clinical study. If you were to design a Phase III clinical study, like, what sort of a duration of an effect would you be looking to demonstrate in a Phase III study?

I can start, if you -- so basically -- thank you for the question. Yes, the -- in PADOVA, the population is slightly more advanced than PASADENA. And however, what we're -- and they are on stable symptomatic and 3/4 on L-dopa and 1/4 on MAO-B. What we can say is that, of course, the time to event is the primary end point here. And that allows us -- now first of all, stepping back: In the general population, patients are diagnosed with Parkinson's disease. Within a number of months, they often go on to L-dopa, so we think it is important to test the idea that -- prasinezumab targeting aggregated alpha-synuclein, on top of L-dopa. Now in terms of the primary end point, however, that will help us, it's time to event. And that time-to-event end point helps to mitigate changes in L-dopa, as well as we're doing this in the OFF state. So for those -- so we think that, that will help in terms of telling us whether there is an effect even on top of L-dopa. And in terms of the Phase III, I will hand it to Paulo.

Yes. Thanks. And I mean clearly a lot depends on the outcomes of PADOVA; and the type of effect size we see; and yes, the profile of the medicine. Our preliminary conversations with regulators are very much towards coming up with a way to measure clinical progression that is relevant and meaningful but also achievable in a short-duration trial. So if the profile that, as I've -- showed you for the PASADENA data that -- the separation between progression curves, for lack of a better word, takes some time to be achieved, then obviously we'd be interested in more of a time-to-event design, right? These are all things that we are discussing with regulators right now. I think there's openness to consider these types of different approaches, also because this is really the first new mode of action and potentially a first disease-modifying therapy for Parkinson's in over 40 years. So I think for now we are, let's say, considering several options in terms of trials and design. Our assumption is that we will have to do at least one Phase III trial even with a positive PADOVA, but we have designed PADOVA in a way that could be supportive of a filing later on as well, so in that sense, all of this is being thought through as we speak and -- waiting for the data.

Okay. I think, if there are no further questions, then we are basically at the end of our event for today. I would like -- there's more questions coming in now, quite -- a couple of questions. Sorry. We have to go on. Next question then goes, next questions go to Emmanuel Papadakis from Deutsche Bank.

Bruno, maybe I'll jump in if there's no more intelligent questions [indiscernible]. A quick one on LUMINESCE: You mentioned briefly the preclinical data and [ basis for bleeding IL-6 ] may have a role in gMG. Could you just talk a little bit about clinical data we've seen from either some of your own trials? Or other molecules in development lead you to believe you may see positive clinical results in the Phase III -- and indeed what you will consider a relevant benchmark. We've had several competing modalities readout in the myasthenia gravis space over the last couple of years, so do you expect to be able to exceed or surpass some of those clinical benchmarks?

Sure. I mean the data I have shown is just preclinical data. There's not a lot more data. To be fair: There are some open-label trials or investigator-initiated studies that have looked at the use of tocilizumab in this disease. And those data also seem to show a significant clinical benefit for patients with myasthenia. Now of course, we don't have more data than this and which is why we designed luminescence (sic) [ LUMINESCE ], to actually address that question conclusively. And again we'll know the data soon enough. Now the expectation for this medicine is that in terms of differentiation. What we would like to see is a medicine that has a significantly better efficacy profile than standard of care; and equal or even potentially superior to, let's say, first-line switchers for FcRns and drugs of that nature. Now of course, we believe there's a big unmet need in this population still. Even though there are numerous approved drugs, there's still about 30% of patients that really don't have a significant clinical benefit and about 60% of them don't really achieve remission of symptoms, so the idea here is really to be able to bring not just a substantially better drug in terms of acute efficacy, if you want, but also something that in chronic dosing would really result in higher levels of remission, for lack of a better word. And at the same time, these are drugs, the IL-6 receptor blockers, for which we have a lot of experience, including in terms of safety. And therefore, what we hope to bring in terms of positioning, if you want, would be a medicine that would be used in generalized myasthenia as, let's say, on top of the still first line, steroids, immunosuppressants, et cetera, with a significant benefit in terms of efficacy but also very safe and very convenient for patients. And we think that's -- that will be a really valuable profile.

Emmanuel, does this answer your questions?

Yes. That's perfect.

Next questions would go to Harry Gillis from Berenberg.

Yes. So just going back to trontinemab, I was just wondering what your latest thinking is as it relates to the rate and depth of amyloid clearance and how that relates to rates of ARIA; and then perhaps, I mean, if there is a positive association, why you think trontinemab, including the highest dose, [ sort of ] seems to have low levels of ARIA.

All right. So I'll start. And that's a really great question. And as you mentioned, you see that we have very -- in this cohort, we see the deep reduction of amyloid. However, so far, we see no cases of ARIA. We think -- this is speculation. We think that actually the route of entry of the antibody may have something to do with ARIA in that there is some evidence that ARIA may result from antibody engaging amyloid in the perivascular space around vessels. And because trontinemab undergoing transcytosis in the capillaries, that you don't have -- and it's in very low doses, that, that is not happening. Again this is speculation, but basically what this would suggest, if confirmed, if shown further, is that you actually can uncouple the ability to reduce parenchymal amyloid in the brain from ARIA.

Thanks, Azad. Harry, does -- did it answer your questions?

Yes. No, that was great.

Then there is one more question from Stephen Scala from Cowen.

A bit of a bigger picture question, but Roche has 18 programs in neuroscience, between Phases I and III. And 10 are in 3 areas: Alzheimer's, MS and Parkinson's. Alzheimer's has proven to be a difficult area to develop commercially successful drugs. MS is an area where there's already a lot of alternatives. Is this viewed as an ideal mix for the neuroscience business; and -- or will it be different, say, in 5 years? And if a broadening is viewed as desirable, how will that be achieved?

Maybe I can kick it off. That's kind of an interesting question because you're asking me to predict the future. And I'd say right now -- I don't know. First, I don't know if there's an ideal profile. I'd certainly say that, everything we have in the pipeline right now, we believe strongly that they could be best-in-disease medicines, transformational medicines that would achieve a lot of benefits for patients and therefore generate value for the company as well and for investors as well. Alzheimer's has traditionally been a difficult area. And I think we possibly have turned the corner in the past couple of years of doing these first approvals. And that also sets us up, I think, now with trontinemab to really develop something better than what exists right now, so in some ways, the door is opening and it's an opportunity for us. And in MS, even though you mentioned correctly that there's a lot of alternatives, there's still a lot of remaining medical need, especially for drugs that tackle disease progression, so CDP progression. And that's a bit what we're trying to do with fenebrutinib and if, let's say, the focus of future developments. And other areas of neuroscience. I mean obviously there are still a tremendous medical need, 2 of the biggest ones being, for example, stroke or adult [ psychiatry ], but again, we only go into these areas when we have convincing evidence based on science that we can actually achieve something meaningful in those areas. That has been the challenge for a lot of companies in this area. Actually recently we've had our own challenges in the [ psychiatry ] field because it is hard to develop. In 5 years, if the science matures enough and we have a breakthrough, of course, we're ready to step back in. I mean we are a company driven by science. Azad, I don't know if you want to add anything from your side.

Yes. I think that -- a little bit to add. I think that's really a great question and a big picture, as you mentioned, a big picture question, so let me give you a big picture response, which is I think that, of course, first of all, there's a lot of unmet need. So clearly there's that. And in terms of success, I think Roche's history really illustrates that neuroscience is really in a great spot now. If you go -- before 2017, all the way to the '60s, there were really a handful of launches, mostly dominated by the benzodiazepines. And that's there are good reasons for that. However, in the last couple of decades, there have been amazing advances in human genetics and technologies in our understanding of disease biology. And as you can see from Roche, since 2017, there have been 4 launches in this area, so I think we're actually on the cusp of major findings and discoveries in neuroscience therapeutics. And I think that, of course, I'm excited about trontinemab, but I think that bodes well. And I think, once you have more success, then agree with Paulo that we will be able to then extend this into even other areas within neuroscience, but I don't know what that will look like exactly in 5 years.

Okay, there are currently no more questions. Maybe I'll pause here for a second [ if there are one or more ] questions. All right, if not, then I think we will close the call...

There's a couple of questions, Bruno, on the Q&A, on the Zoom Q&A...

Yes, but I think they are -- I could quickly read them aloud. I think they are pretty straightforward. So will the slides be available later on the investor web page? For sure, as always. And will the Phase III tronti study contain a subcu arm, Paulo? It's a bit early to...

A bit early to talk about that, yes.

Yes. And then there is a question here on the mid-year submissions for Elevidys. Have the EMA [indiscernible] the early meetings? And is there a way to consider the FDA action to help EMA decisions? Samir, this goes to you.

Yes. Happy [ to pick it up, Bruno ]. So basically we don't disclose the nature of our interaction with regulators, but we can just say that these are really collaborative, constructive. And we are moving forward with our submission with EMA. Regarding FDA, as you know, regulators are independent from each other. Of course, [ they talk ]. They look at what they are doing, but they are independent.

And then this also goes to you, Samir, one question here on the competitive landscape in DMD gene therapy. So what is Roche's view on the competitive landscape and especially mentioning here Pfizer's potential Phase III readout this year?

Yes, we are looking forward, basically. You'll see the data from the [ CRO ] Phase III trial, which is expected by the end of the year. And of course, we cannot comment until we see the data.

Yes, okay. I think, with that, we have covered the questions here which were handed in the writing, in the chat. And if there are no further questions, then I would close the call. I would like to thank the speakers again for all their commitment and their work; and then also, from the IR team, Loren Kalm and [ Anita Tang ], who were working on the slide decks together with the speakers; and also [ Sameer Schwab ], for, from organizing the event. I hope this was a helpful event for you to get to know the latest from the immunology pipeline. If there are any open questions, then the IR team is happy to take your questions. Please reach out. And with that, I would like to wish you a good day. Bye.