Roche Holding AG (ROG) Earnings Call Transcript & Summary

Ladies and gentlemen, welcome to the Roche Virtual Event on Key Data Presented at the 2021 MDA Virtual Clinical and Scientific Conference. My name is Henrick, and I'm the technical operator for today's call. Kindly note that the webinar is being recorded. [Operator Instructions] At this time, it's my pleasure to introduce you to Karl Mahler, Head of Investor Relations and Group Planning. Karl, the stage is yours.

Yes. Thanks a lot, Henrick. This is really kind of a routine now in these Zoom meetings with you. Welcome to our second Science Day this year. If you could kindly go to the agenda. This is -- yes, here exactly. I'm today with Paulo Fontoura. He is the Head of Development for CNS and Rare Diseases, but in the meantime, he's also heading up the whole [ NMO ] department in terms of clinical lead, development lead. And Simona, she is the counterpart basically, what the marketing end is concerned. So she is leading our therapeutic area, CNS, from a business point of view. We have 1 hour, 0.5 hours presentation and 0.5 hours Q&A. [Operator Instructions] And before we -- and even before we start, I wanted to thank already Sabine Borngräber for helping taking the lead on this event today. Maybe we can go to the next chart. Yes. Here, remarks from my side. I mean, the year actually started really well. I mean, end of last year was good with the PDS, Port Delivery System, which read out positively. This year, we had the reading out of the faricimab data in AMD and DMD, and so that was actually really good news for us in terms of the start. Then we have -- we do expect, in the not-too-distant future, the antibody to read out, outpatient study, the 2076 (sic) [ 2067 ]. There is more to come on Polivy this year. We have more to come on the adjuvant non-small cell lung cancer somewhere in the year 2021. So lots of material in popular news flow to come in terms of major readouts. I also wanted to highlight that we still have a Diagnostic Day next week on Tuesday where we all can learn a bit about our products here in the business for Roche. And that is really a good event as well, like today. If you go to the next slide, please. I mean, we in Roche have never given up on the commitment to neuroscience. And I would say, if you look at the past and you take a y and x axis, and we look at the low cost and high cost on the one axis and you look at the low probability and high probability on the other axis, if you look where is the highest risk and the highest investment, actually, you end up at the CNS area. And we have -- this was the reason for some of the companies, why they just had said goodbye to CNS in the past. And I have to say Roche never said goodbye. We always had a reasonable level of investment in CNS for the very reason that we have about 600 diseases of the nervous system. We have 700 million cases, neurological disorders. Actually, it's the highest rate of disability burden to society, the second-highest death rate and we have a huge burden to society. And in the meantime, it paid off because other companies tried to get in, and this is usually very difficult to close the gap again. But we have continued our investment. And if you take Paulo as an example. I mean, Paulo was heading up the CNS from a research point of view earlier on. Now he is doing the development part. Simona actually was launching OCREVUS. So there is a good success story in CNS. And you -- we have about 14 pipeline assets, 115 clinical trials. We have 4,000 scientists still dedicated to this one. And maybe we are approaching a new wave. That was maybe oncology before. It's CNS now. And we understand the diseases better than we did. And while we do not highlight the Alzheimer part, of course, you can ask any kind of questions on our own CNS franchise, but only to that. And with this one, Paulo, over to you.

Thank you so much, Karl. And good afternoon. Hello, everybody. So I'm very happy to be here with you today to present some of the data that we showed this week at MDA around the Part 2 data for our SUNFISH trial. So can I go to the next slide, please? These are the investigators and our disclosures. So let me start first by talking a little bit about SUNFISH design. So this is a trial in patients with Type 2 and Type 3 SMA aged between 2 and 25 years old. These are nonambulant patients, Type 2 nonambulant and Type 3 who are able to sit independently. But we've allowed patients who have had either scoliosis or surgery for scoliosis in this trial. So the primary readout of this trial was at month 12, and the primary endpoint was change from baseline in the MFM32 total score. And as we've presented before, this was a positive study that showed a clear benefit in terms of clinical outcomes. Next slide. And what we're talking about today is the 24-month data. The 24-month data includes patients who have been on risdiplam for the full 2 years duration, and at the same time, patients who were on placebo and switched over to risdiplam for the second year. And of course, you have the same clinical outcomes, but these are exploratory now. We have already had a major database lock. So we're going to keep looking at the same type of clinical endpoints: MFM32 responder rates, upper Limb, Hammersmith, et cetera, but also obviously for safety. Next slide. So just to go back a little bit in terms of the baseline characteristics for these patients. This was a very broad range of patients with Type 2 and Type 3 SMA, as young as 2 years old and as old as 25. You can see here, there's a good distribution across this age range. It is 4 age brackets. There is a good balance between gender. The majority of patients were SMA Type 2, and a large majority of them have 3 copies of the SMN2 gene. Next slide. In terms, again, of baseline characteristics from the clinical standpoint. This is important to bear in mind because the placebo patients are -- have slight differences which may or may not have their role in the clinical outcome. So when you look at the percentage of patients with scoliosis at baseline, there's a higher percentage of those in the placebo arm. When you see people who have had surgery, there's a slightly higher prevalence there as well. The total scores on MFM, RULM and the Hammersmith are roughly comparable. As we pointed out before, we actually started out with a population which was significantly disabled at baseline, with a significant percentage of patients with really low score, particularly on the Hammersmith. Next slide. So starting with the primary endpoint, and this is change from baseline on the MFM32 total score. And as you know, this is a very broad motor scale. It measures a variety of motor functions. What you're seeing here on the left-hand side is the overall population on the risdiplam arm versus placebo up until 1 year. And obviously, from month 12 to month 24, it's only the risdiplam patients who have been on therapy for 2 years. And you see roughly the same pattern, which is a continuous stabilization or improvement in function. Now of course, we no longer have a placebo-controlled arm. So on the right-hand side, we're giving you a little bit of a picture of what natural history looks like. At between months 12 and 24 in this population, which is roughly similar to the one that we already have, you continue to see a marked decline in motor function, in this case, between 2 and 3 points on average. And therefore, this puts into perspective the results that we have on patients in the second year of risdiplam treatment. Next slide. Another way of looking at these data is looking at responders. So on the left-hand bars, you see people who have improved at least 1 -- to at least 3 points, which means they have a significant, really significant, motor outcome. And on the right-hand side, people who are either stable or have improved. Again, these data are another way of looking at this. You see roughly it's the same percentage of patients, more or less. There are some differences here, but this is within the statistical variance. And of course, as you can see as well, we don't have the full data set yet for the second-year patients because some of these data are still being collected now. But roughly, we see the same picture, a significant percentage of patients who either have significant improvement or stabilization and an improvement. Next slide, please. Other endpoints that we've presented on are the revised upper limb module and the Hammersmith Motor Functional Score. You see on the left-hand side the revised upper limb module, so this is an upper limb functional scale. Very important, these patients, particularly in the more advanced, more disabled patient, because upper limb function becomes tremendously relevant for their functional outcomes for every day. And you see a continuous improvement on the second year of therapy. On the right-hand side, on the Hammersmith, this was actually one of the points that, when we presented the first year data, was, let's say, not totally -- we're not totally happy with that because, clearly, there wasn't separation from placebo, even though numerically, we were better. But now as you see in the second year, there's a continuous improvement on the Hammersmith as well, which is very much in line with the other motor functional scores. Next slide, please. Now switching gears a little bit to talk about the SMAIS. This is really a scale developed by patients, with patients and together with us that measures the level of help that patients with SMA need for daily activities. It measures 29 items of daily living. It's self-reported by the SMA patients older than 12 years old. And it's reported by caregivers for all patients above 2 years old. And you see here at the bottom, a little bit of an idea how do we score this from this score of 0, in which an individual cannot perform that activity without help, until a score of 2, which only needs a small amount or no help. Next slide. So looking at these data. On the left-hand side, you see the data for caregivers. And again, you see not just the impact that risdiplam has on motor function, but also impact on activities of daily living. The SMAIS clearly increases over the first year and then over the second year. So these patients are becoming more independent and capable of doing their activities of daily living. On the right-hand side, you see a similar picture. You see a stabilization, at least. I'll point out that these are mainly older, more disabled patients, right? So again, if you bear in mind that the overall picture we got when we looked at the functional scores and at the motor scores with that, for younger patients, you usually see a remarkable increase in motor functions; whereas in older, more disabled patients, more about stabilization. These data are very much in line with that picture. Next slide. Now just looking at the patients on placebo, the 60 patients on placebo who switched over to risdiplam. As I pointed out, there are some imbalances in baseline severity scores. And also, of course, now all these patients are 1 year older. Despite that, what you see here is either a stabilization or a slight improvement on these motor functional scores, the MFM32, the revised upper limb module and the Hammersmith. Again, very much in line with the other data. Next slide. And when you look at the SMAIS, again, the same picture. Improvement in autonomy measured by the scale, both by caregivers and by patients. Next slide, please. The adverse event profile. This is extremely important because for chronic therapy, obviously, something we care deeply about, what we saw actually was a decrease in the number of AEs, Grade 3 to 5 AEs and treatment-related AEs, over the second year of treatment. You see here, again, very, very balanced. There were no treatment-related AEs that led to withdrawal or discontinuation. Our ophthalmological monitoring program, which is very comprehensive and very robust, has continued to really show no evidence in humans of the retinal findings that we're observing in the preclinical tox studies. And all other parameters have been stable over time. Next slide, please. So overall and in conclusion. These are really exciting data for us. On one hand, we've been able to show that motor function was either stable or continued to improve during the second year of treatment. Very important outcome for these patients. Secondly, that we do see on an activity of daily living, more of a patient-reported outcome, the SMAIS, that there is continuous improvement or stabilization in these patients. All these changes that we saw in the 2-year duration treatment, we also started to see happening in patients who switched over from placebo to risdiplam. And of course the safety profile. The decrease in adverse events and no new adverse events is really great news for these patients. So these are important data that confirm the long-term efficacy and the safety of risdiplam in a really broad population of patients with Type 2 and Type 3 SMA. Next slide. I just want to finalize by thanking all the patients, all the families, all the investigators who continue to support this program. Thank you.

Hi, everyone. Good afternoon. My name is Simona Skerjanec, and I am the Therapeutic Area Head in Neuroscience and Rare Diseases in Global Product Strategy. The partner to Paulo, as Karl introduced us earlier. So next slide, please. As you know, and as Karl mentioned, we are in neuroscience rare disease. And we have a broad pipeline that's leading the way to first-in-class medicines. In neuroimmunology, we're leading with OCREVUS, as you know very well. We also have a medicine in Neuromyelitis Optica that's just been launched in the U.S. and a few other countries. In neurodegeneration, we've been in Alzheimer's disease, are in Alzheimer's disease formally. We're also in a rare disease, in neurodegenerative rare disease, in Huntington's disease. And we have -- we continue to have programs in Parkinson's disease. In neuromuscular, Paulo just covered a number of -- and some new data from Evrysdi, so SMA. We're also partnering with Sarepta in Duchenne muscular dystrophy with gene therapy. And in neurodevelopmental and psychiatry area, we are developing in a rare neurodevelopmental disease called Angelman syndrome, we're developing a first-in-class antisense oligonucleotide. And we're also going into schizophrenia with a product called ralmitaront. I always have trouble pronouncing that one. Next slide, please. We are not only broadening our portfolio, so just a different view, we have a strong momentum with our successful launches. So we understand more and more the biology in this area, but we also know how to bring these medicines to patients around the world. OCREVUS was launched in 2017. It was considered one of the best launches in the industry. And this year, we're launching Evrysdi and Enspryng, and we hope to be able to bring along the portfolio of products from Phase II and Phase III as well as earlier products in Phase I. So good commitment to this area, I would say. On the next slide. Beyond biology, it's important, especially in neuroscience, to think about other modalities that can either help us help the patients and practitioners measure the efficacy and capture episodic and rare events in a continuous and longitudinal way. But it's also important for us to do this in development. Karl mentioned in his introduction slide that there's a high investment, high risk, but there's also a great opportunity because there's so much unmet need out there in neuroscience and neurological rare diseases. So it's important for us to think about how technology can help us minimize or reduce some of these risks, and some of these digital -- some of that digital technology can certainly help us make better decisions. Here's an example of where we are with the wearable disease-agnostic technology that we're testing in DMD and can help in measuring patients' ability to ambulate at home. And we probably talked about FLOODLIGHT before. We are now actually launching it, and it's been launched in a couple of countries already. But our plan is to launch it in about 10 countries. And hopefully, this will become a very important tool for patients and physicians to uncover earlier progression of their disease and hopefully allow them to treat it better earlier. To the next slide, please. So just a little bit of an update about other medicines in our franchise that are on the market. As you know, OCREVUS has been shifting standard of care. We are now -- it's been available. It's been more than 200,000 patients globally. And I believe that the -- what we have is a robust, consistent and sustained impact on slowing disability. What's shown here on the left side are 2 graphs that provide data on what's important to patients, which is their progression measured by EDSS on the top. And what's quite important on the bottom is a proportion of patients that reached EDSS score of greater than 6, which is a critically important milestone of independence because it represents the requirement of a walking aid. And as you can see with the data, there's almost 50% risk reduction in needing a walking aid when patients are treated with OCREVUS. And this is across 6.5 years of follow-up. And on the right-hand side is how I believe this data translates into growth in the market. And this data has been -- I believe that you've seen this data or you've heard this data. This is end of the year '20. We have been able to show continued growth, with overall 25% total market share in the U.S. and the new market share remaining at 40%. We were able to do this despite COVID effects that we all know impacted patients with chronic diseases that required to go to the hospital. So -- and maybe just to close this OCREVUS topic. We continue to believe that growth will continue in 2021. Next slide, please. The unmet need in MS remains, and the unmet need remains in continuing to impact the course of the disease and slow down progression more and more. That's why we are investing into another Phase III study with OCREVUS, this time studying a higher dose in both patient populations, RMS and PPMS, because our analysis of Phase III -- the exposure response analysis in our Phase III studies showed that higher dose could potentially lower the risk of disability and would not compromise safety. We're quite along with enrollment. We're almost between -- more than 1/4 enrolled in these studies. So we're hoping that we can even further reduce disability progression in both of these patient populations. And then we also have started Phase III with highly differentiated and potentially best-in-class BTK inhibitor in fenebrutinib. We are enrolling in both of these studies. And to just highlight our commitment to showing reduction in disability progression rather than just relapses, our primary endpoint here is linked to that, because we believe that that's where the unmet need and the value in the system still exists. Going to next slide, please. Just a quick note on NMO. So as I said, it has launched in 16 countries to date. It is a rare disease that is highly debilitating. In fact, in -- about half of the patients are blind within 5 years and require a wheelchair. Enspryng is highly effective. It has good tolerability and safety profile. It has -- it's convenient. It's a subcutaneous dosing. And we believe it's competitively priced. We have good -- we made good progress in the U.S. We have a number of patients enrolled in our Access Solutions. We have good payer coverage so far now. This year, we have to pull it through so that we bring these patients -- to bring this drug to the patients. Moving on to the next one. So here's the topic of the day, SMA, which, to me, what's really interesting is that despite a number of treatment options, majority of the patients remain untreated today. And what's also quite important probably is the prevalence will likely increase as more treatments become available. On the bottom, there's just a diagram of the different types of patients and prevalence and incidents. And maybe one thing that I would just focus on that is perhaps interesting is that -- and we don't hear so much about, is that majority of patients are actually adults. So if you go to the next slide, it's -- part of the reason I bring this up is that the significant unmet needs remain and a lot of barriers exist to different patient populations, but they are most pronounced in adult patients. And Evrysdi has -- with Evrysdi, we've studied a very broad population, Paulo was describing some of that data, with a highly clinically meaningful efficacy in different patient population and heterogeneous patient population, where we've seen sustained benefit. And what's also important, especially during these time, is that the home administration, oral administration, allows -- reduces the burden to patients, caregivers and the health care system. What's interesting to me here is also that we have a reason to believe, because the biology show that SMN protein is increased not only in the central nervous system, which obviously is a key, but also peripherally. And if you go to the next slide, this is the sort of thing that the neurologists tend to be very interested in. And here's a survey that we did in the U.S., major European markets, Brazil, Canada and China, that shows that these neurologists really care about whether or not this protein is -- the impact of deficiency of this protein in CNS and peripheral. But what's even more interesting to me is the bottom part, that more than 60% of them think that it is essential to treat SMA by targeting both the CNS and peripheral tissue, which we've shown with Evrysdi. We go to the next one. So I firmly believe that Evrysdi has the potential to be preferred SMA treatment option for patients, doctors and payers. We've talked a bit about patients already, broad patient population types, range of ages and the convenience of at-home administration. The doctors believe there's a compelling benefit/risk profile. And broad population again and the well-tolerated drug. But what's also important, I mean, I haven't mentioned that, is that the payers believe that we have a clear value proposition. We're responsible in our pricing. And we're taking the cost out of the system by bringing this drug to the patients in a home setting. So there's no need for hospital administration, anesthesia or hospitalization. And with that, I believe I have one more slide. Yes. So we are on track and moving fast around the globe with our approval -- regulatory approvals and launches. So we've been approved in the U.S., as we said, since last August and gained share very quickly, more than 10% in the first -- I think it was actually 5 months, not 6 months. And we have a broad use across patient populations and the fastest uptake. So this kind of reminds me a bit of OCREVUS, to be honest, when I look at this. And we have -- we're very excited about CHMP positive opinion that we've just received. Look forward to launch in Europe soon and approval in China. More data coming out this year as well on the bottom here. So we hope we can continue to bring -- continue to deliver good news to the physicians and patients as we gain access around the world. And I think this is the last slide. And with that, I'm going to pass it back to Karl.

Yes. Thanks a lot, Simona. We have really plenty of questions. I was surprised. We have about 140 people on the call, which is a lot for such a specialized call. So thanks for your interest in Roche. There is a physician from [indiscernible]. And usually, we don't have physicians on the call. But I still wanted to start with a question from [ Mr. Sarnin ]. He was asking, "If you want to switch from Spinraza to Evrysdi, what is the waiting period if you want to switch?" Maybe a very specialized question, maybe a detailed one. But I don't know who could take that one. Maybe Paulo, you?

Yes. No, I'm happy to take that. And again, there's a lot of questions around that. We are very -- we're very committed to generating data, okay? So up until we have those data, it's hard to make specific recommendations on how to do things, right? We have an ongoing study called JEWELFISH, in which we are allowing people with -- who are being treated either with Spinraza or with Zolgensma to switch over to start being treated with risdiplam. And we've presented some of these data, and we'll present more data later this year, data hopefully around efficacy already. The data we already have around safety does not generate -- and so basically, the safety profile has not changed based on those data. We do see the same increases in SMN protein for patients who haven't been on Spinraza or Zolgensma, roughly. So you get the same biological benefit. Of course, we can only make specific recommendations on how to do things once we have the final trial data for JEWELFISH, and we hope to have that by the end of this year.

Yes. Thank you. And I would say, let's go to the telephone line. Wimal, you were the first to ask a question, so you're also the first to get an opportunity to ask it here live. Wimal? We can't hear you. [Operator Instructions] Okay. Maybe that -- so then the next one would be Simon Baker.

Thanks so much, Karl. Hopefully, you can hear me.

Yes. We can hear you. Yes.

Questions if I may. Firstly, on the primary endpoints, I wonder if you could give us any more detail on the change in MFM32 score in terms of if you saw any clustering in domain 1, 2 or 3 of the score, or whether the benefit we're seeing more broadly across the whole of the scoring range. And then secondly, really just getting your thoughts on the broader use of splicing modulators. One of your neighbors is obviously looking at splicing modulators in Huntington's, PTC have patented risdiplam in Huntington. So I just wondered what your thoughts and aspirations were for risdiplam in that, and indeed, other segments.

Shall I take this, Karl?

Yes, please.

Yes, sure. So your first question is really interesting. We don't really have the data to show you right now. It is something we're looking at. The data is still coming in, as I mentioned. So don't really have -- even all the full year -- the full 2-year data for all the risdiplam patients, these are analyses that we will be performing. And as soon as possible, we will be making them public at upcoming conferences. So I would ask you to bear with us a little bit. Regarding your second question, I do believe there's something interesting about being able to modify splicing with small molecules. And it's no surprise that our colleagues at PTC and our colleagues across the river are interested in that. Specifically, I don't really want to speculate around indications or molecules, et cetera. One thing that we have learned is importance of specificity in terms of safety, right? And that's why Evrysdi, really, we think, has this remarkable profile. Because in a very broad analysis we did of genomic profiling, we really did not see significant modulation of splicing in many other targets. There's only 2 or -- 2 out of 3, which are the ones that actually led us to put in place such a comprehensive monitoring program for safety. So regarding Evrysdi specifically, we have no plan to studying it in Huntington's, if that's the question as well.

Next, I want to hear a question from the chat here from Stephen Scala from Cowen. He was wondering -- I think this is also going to you, Paulo, "For gantenerumab in Alzheimer, is Roche conducting any kind of patient selection based on tau imaging or any other kind of patient selection, let's say, meant to identify mild Alzheimer?"

Well, the short answer is yes. It's kind of a broad question. When we put together the maybe GRADUATE Phase III programs, we do have -- we're actually pioneers there. We do have very specific molecular diagnostic criteria for entry. And those have to do either with PET imaging on amyloid or CSF testing for amyloid and p-tau and tau. So those are characteristics at baseline. We did not have available broad-scale tau imaging for those programs. We have put tau imaging in other programs, such as our tau program. Therefore -- and we are developing our own tau PET ligand, so in the future, that might be possible. But if your question is specifically around the Phase III GRADUATE studies, we do not have tau imaging there.

I think the question was maybe also because of the leading data. And if we can have any kind of a read-through.

Yes. No, it is an interesting question. At the end, I think we will have some baseline data on CSF molecular markers and therefore we may be able to retrospectively do those types of analysis. We did not prospectively use that as a randomization criteria. Just comparing the inclusion/exclusion criteria for patients in both the donanemab trials, because that's the question, and the gantenerumab trials, there's really not much difference. So given that our trials are very large, there are over 1,000 patients each, I'm sure that randomization would have worked and make sure that we have sufficient samples to do those analyses.

Next one, I'll take one from the line, Matthew Weston from Crédit Suisse.

Karl, hopefully, you can hear me. Three quick ones, if I can. What impact do you think the data that's just been presented and the upcoming JEWELFISH data will have on the launch trajectory of risdiplam? Do you think we'll see an inflection? Do you think one data set is more important than another? I'd be very interested as to how we see the commercial progress. Secondly, Simona, you mentioned Angelman. I just wondered if your Phase I study is fully recruited and when we expect to see the first results. And then secondly, Simona, maybe this is a little bit mean. But your pitch for Enspryng was identical to Novartis' pitch for Kesimpta in terms of subcutaneous home administration, the ease, the benefits to patients. I just wondered what's different about NMO versus MS that you think subcu is better in NMO, but you think a regular infusion is better in MS?

Do you want me to do the last one first, maybe?

Okay -- Simona, please. Yes, yes. Absolutely, go ahead. Yes. Yes.

Okay. So maybe, one, obviously, anything that's convenient for patients is -- it has a great -- it has a good impact, right? Now what we need to remember with OCREVUS is it's not a monthly IV, OCREVUS is a twice-a-year infusion. And even when I was in that program, the patients would say, "If I can only have MS twice a year, that's better than anything else." And I think that's what's coming through. So I would -- it's not just the mode of administration, but it's also the frequency that matters here. Then I'll just go backwards. So let's -- so quickly about Angelman's. We are in Phase I recruiting. I honestly cannot remember exactly what the time lines are. Maybe Paulo does. But -- so I'll pass that one on when I'm done. But we're recruiting. And in the earliest days, as you know, very well, we want to be very thoughtful about safety. So the recruitment in the early stages is perhaps a little bit slower than one would do when you're in a bigger Phase III program. And what was your first question? Now I've forgotten.

This is the market inflection of what the importance of the data is. And if you're happy with the current launch. Maybe you can talk about the current launch, how it's going. And maybe which other kind of data you would look forward to in order to even have even a better one, better launch.

Okay, with the data. Yes. So I would -- so what I would say is the entirety of the data, which really is studied in the heterogeneous population, right, is I think what's driving this market. And as I said earlier, Evrysdi really had -- we've studied with Evrysdi a broad population. And we're going to -- that data that's going to continue to come out, I think, is going to solidify that even more. Of course, when one sees data like we saw today, that gives you a greater confidence in the benefit/risk. And so I believe that the neurologists, that tend to be a little bit more conservative maybe than some other specialties, usually react positively to data like that. So I hope, and it's our ambition, to be the market leader in SMA by the end of the year.

Okay, thank you. Maybe, Paulo, you could go for the timelines, Angelman's status a bit more.

Sure. Yes, I'm happy to. As Simona mentioned, we're just in the beginning of Phase I actually, okay? And this is a dose-escalation study in patients with Angelman's between 5 and 12 years old. And we're going below that age as young as 2 years old. So as you may realize, these are difficult trials to recruit for and to maintain on. The primary objective here is really safety. Of course, we have a number of exploratory efficacy markers. And therefore, as soon as we have data, we will be very happy to share them with the community. But at the same time, we're -- I'm just cautioning everybody, we're in the beginning of this journey. If I may just supplement what Simona said about Enspryng, because I kind of take that personally a little bit. Because I agree with you, all things being equal, subcu versus IV may be a convenience play. All things are not equal. OCREVUS has much better data, right? Because we have PPMS data, we have progression data, which is really, really outstanding. And including for Enspryng, all things are not equal. We're talking -- it's not just convenience we're talking about. We're talking about a very different safety profile between drugs that do have black box warnings, that do have a lot of severe adverse events associated with. And in comparison with that, Enspryng is really a very well-tolerated drug, a very clean drug. And secondly, it's the only drug that has data both as monotherapy and on top of standard of care with the immunosuppressants, steroids, et cetera. And about 50% of all patients actually are currently on something. So it is really a better choice just to allow them to stay on their current therapy and adding Enspryng on top. I would just say that.

Thank you, Paulo. It's even better if the medical partner believes in that than the business partner. So it's a better answer, thank you.

But it's not about -- [ it's how they use supplies ], so I have to say Simona launched OCREVUS and Paulo developed OCREVUS. So I would say you have the 2 ends of that, of that drug.

Many thanks, indeed, to all of you. Extremely comprehensive.

Thank you. Thank you, Matthew. I think there are maybe 2 questions here from the chat from Michael Leuchten and from Marietta Miemietz. They both have the same questions on the high dose for OCREVUS. "What is the reason for the high dose? And why do you undergo that journey?" Maybe, I don't know, Paulo, why don't you answer this question from...

Yes. Sure. I'm looking at the chat, actually. "Why aren't the benefits picked up on the original dose-finding work?" This is actually a really interesting question. Because the reason we're doing these studies now is precisely because we picked up on that higher benefit, okay? So when we looked at the dose we have currently approved for 600 milligrams, and you just do an exposure efficacy analysis on those data, it is very clear, first of all, that there are -- there's a spread of exposures, as you would expect with any drug. There are people who get more exposure, people who get less exposure. So when we looked then at how that correlated with CDP progression basically, right? What we observed is that the lower quartile was really not getting as much as the higher quartile. There was really a separation there. And therefore, our thought experiment was, if we can bring everybody up to that higher quartile, so even the people with the lowest exposure would be in the higher quartile of the current dose, what did that do to progression? And it's model data, of course. But the data we've shown is that presumably, you might even be able to double yet that impact on progression. So it is, actually, because of the data we observed in the clinical program, we decided to go for this higher dose with the hope that this would maximize impact, okay?

Thank you. Maybe we can take the next question from Emmanuel Papadakis from Deutsche Bank via the telephone lines.

Thanks, Karl. Yes, Emmanuel Papadakis from DB. Just a couple on Evrysdi, please. First, a question on the data, perhaps for Paulo. Last year, you were kind enough to provide the cut of MFM32 improvement by age subgroup, but not HFMSE. This year, you haven't actually given us either. So could you just talk a little bit about the trends by age subgroups? Simona highlighted the importance of the older Type 2, 3 population. So was the trend consistent? And then perhaps a question on the commercial launch. I'm a little surprised you just provided the same numbers you gave us for the Q4 results, over 1,000 patients, 10% share. We're obviously a couple of months later. Is there anything you can comment on the trends in the last couple of months? You did comment at the Q4 results that it was around 1/3 naive, 2/3 previously treated. I think you said it was broadly equal to Spinraza, Zolgensma. So any color as well in terms of the types of patients and prior therapies would also be helpful.

Yes. The lack of update, you can put on me, I have to say. Because unusually, forgive me.

We'll forgive you, Karl.

We usually give updates always at the quarterly calls, when we have really -- because of compliance, I mean, we really want to give the same information to a broader audience. But maybe Simona, you can give an intention or some ideas on where it's going at the moment. Yes. I think you mentioned that you want to become the market leader in that market. So I guess it's improved -- it's improving quarter-over-quarter. So this is what I assume. Yes.

Yes. So maybe I'll just answer this. So I would just say that we see -- we are ambitious, and we think that we can grow here. And the population trends I'd say are probably similar abroad, and they include the patients who were on prior therapy, including Spinraza and Zolgensma. I hope this gives you a little bit of a -- of what comes with the quarterly update.

It's only a little bit.

Yes. And on your first question, I think it's a really good question. And it's the same answer as for the gentleman who asked about the clustering in domain 1, 2 or 3. We just haven't done that analysis yet. And as soon as we have, we'll be presenting those data as well.

Thank you, Emmanuel, for your questions. There was one here. I take also this one here from the chat, from Michael Leuchten again. He was asking about the type 1 patients. I think you touched on it. "How has been the experience so far in terms of add-on therapy? Do you been checking the share sampling? And why -- what is the development milestones that triggers Evrysdi use in patients that previously received chemotherapy?"

Yes. I think a lot of these data will come out when we have the full results from JEWELFISH. We do have the original ones, right, this is the baseline characteristics. And it's not -- there's no specific development milestone as far as I'm concerned. I mean, when we ask reasons for switch, there's -- I'd say about 1/3 of patients who quote as a reason for wanting to switch, concerns about efficacy, whatever those are. And you have to realize that for parents, this is a really emotional situation because you want to maximize the chances of benefits for your baby or your young child. Therefore, there may actually not be any specific development milestone. They just may be a question of you wanting to give your child the best possible choice. So there's some concerns around that. There's some concerns around safety. There's multiple reasons, I'd say. In general, people are wanting to experiment with the therapy, I think. Like Evrysdi that has, I would say, really a very convincing package of efficacy and safety data, basically.

Thank you. Next one is via the telephone line, Luisa Hector.

So I wonder whether we could just get a little bit of any insights you have, having seen -- in Alzheimer's, having seen Lilly's TRAILBLAZER study thinking about gantenerumab coming. Is there anything you'd like to point out in terms of your patient population? Your endpoints? Anything that gave you perhaps more confidence in your study after seeing that data? And secondly, I mean, amazing job that your studies, all these studies running and intact through COVID. So I'm wondering whether you have a fair number of perhaps rare disease studies here within this division. And perhaps they are more of a decentralized nature, more -- or perhaps were already more suitable to virtual clinical trial processes. So were there -- anything you could tell us in terms of how -- id you have to invest even more heavily to keep these trials running? Is the fact that you have your sensors and devices for continuous monitoring -- you showed the example in DMD, for example. Were you already well positioned here? And could you feed that across other areas within Roche? So just anything on the key learnings through COVID and how well positioned you were and how that may have advanced some of your other studies.

Yes. Sure. So starting maybe with the first one around the read-through from donanemab. The way I look at this, I mean, for me, the top line message is that it is always good to see that targeting the same target as we are, so targeting amyloid in this case, can result in clinical benefit. And we've seen it across a range of trials now with some different molecules, aducanumab, donanemab, [ lakenumab ], gantenerumab. I think we're all seeing the same picture, which is targeting specific forms of amyloids, specifically the fibrillary forms of amyloid and plaque, can have clinical benefit. And that's so -- it's encouraging to see that again demonstrated with the new molecule, right? It is one very specific target because it's the pyroglutamate form of amyloid, which is very plaque-specific. And as you probably know, gantenerumab has a broader binding profile. It binds very well to plaque as well, but also to fibrillary Filipino oligomeric forms. And therefore, we actually hope that will maximize clinical benefit. In terms of endpoints, et cetera, I know there's been a lot of speculation about the iADRS that was a primary endpoint for this TRAILBLAZER trial. That's really a refactoring of other endpoints that we have in our trials, like with CDR Sum of Boxes, the MMSE, the ADCS-ADL. So it's not really a new endpoint, it's a new way of looking at the same outcomes. They did have the same endpoints there, right? So they did have the CDR, the MMSE. Again, the -- my gestalt of the whole thing is that there's a clinical benefit, it's not very large. It's around 20% on the CDR, which is interesting, I would say, but again, very, very small study. So based on that, if anything, I take more confidence that gantenerumab is a really well-designed program. Our GRADUATE trials with everybody on the maximum high dose, very simple titration regimen, very large study, over 1,000 patients, 116 weeks of treatment, so really maximizing the chances of success, I feel really, really confident that we'll see conclusive data. Let's put it that way. I don't want to bias anybody. Of course, I think it's -- I'm hopeful, but it is important that we design good experiments that give us clear answers, okay? And I think that's what we'll get here. Regarding your second question, I think this is something that in which we did invest a lot. We went to -- all the way out, I would say, almost on a site-by-site basis, not just in rare diseases, but including like in the gantenerumab program, in our Huntington's program. I mean, we really established very close partnerships with the sites. We realized very early on that a lot of this was actually about allowing, for example, safety monitoring to be done in a slightly different way, to be done at home or virtually. Some of these devices, as you mentioned, also helped. So we really put like our patient-centricity hat on and really tried to make trials as easy as possible for patients and for those sites. And I think it has paid out generally. So -- and this is something not just in neuroscience and rare diseases, we're doing it across the whole of our pipeline.

I hope that addresses your questions, Luisa. Next one would be Keyur.

Two questions, please. The first one to you, Simona. I think you kind of compared the Evrysdi launch to kind of the early signs from the OCREVUS launch. Just wondering, kind of longer term, do you see the commercial potential for Evrysdi to be close to what OCREVUS is doing right now? Or how do you see the ultimate commercial potential for that asset? And then as you're launching through Europe, how are you thinking about the price points kind of from a European perspective? Anything you might be able to add there would be great. And then separately, Paulo, for you. Paulo, can you just remind us what are the primary endpoints on the gantenerumab study are powered to show? And when you might actually get to a potential interim analysis there? How should we be thinking about the next steps on that -- or the next thing we will hear on that?

Yes. So maybe I'll start. So it's -- your question around comparing Evrysdi to OCREVUS, it's a little bit more qualitative, not quantitative, okay? But what I see here is that we have a great medicine that is very much appreciated by patients and patient community, which was very similar to what we -- what I've observed, what we've observed with OCREVUS. And so a lot of the drivers that are meaningful to patients, I think, are similar to OCREVUS. Also, the sustainable effect that you saw today, it's also something that matters in this patient population, clearly, and the benefit/risk profile. We are here with neurologists, probably even more important in this patient -- SMA patient population. I think we see similar -- I see similarities there. So I think for that reason, I believe that the trajectory is kind of similar in the marketplace. It translates into that and will translate into leadership in a very short period of time, right? So -- like OCREVUS did. So we -- that's part of reason. In terms of commercial potential, we are talking about very different patient populations here. So I think that it would be impossible to expect similar numbers, if you will, because of -- simply because of the patient population and the size of the patient population. There was a question about the price. I think we're -- as you know, we're very competitively priced in the U.S. I think the same will be true in Europe. And I think there's a lot of value in what we bring, as I mentioned. And the value proposition that has been well accepted in the U.S., I think will be well accepted in Europe as well. I can tell you, before I came to this job, back to neuroscience, I was actually a General Manager in Portugal. And I can tell you that there was a huge interest from payers to bring this in. So I think we're going to have -- the value proposition that we have, I think will convince payers in Europe as well.

Thank you. Paulo?

Yes. So the primary endpoint on our GRADUATE programs is change from baseline on the CDR Sum of Boxes at 116 weeks or at 27 months. And the reason is that that's -- there's multiple reasons. One of that is that, first of all, it is a primary endpoint which is regulatorily accepted for filing and approval. It's the same primary endpoint that Biogen has on their program as well. And this is -- the reason why regulators have accepted the CDR, is that it measures both cognition and function across both prodromal and mild patients. So it is -- that's the reason why. We do have other endpoints. We're measuring ADAS-Cog, ADCS-ADL and MMSE. So all those are there, but that's the primary endpoint. In terms of powering, we really don't disclose details of powering too much. I would say, generally, we power studies for what we think is a clinically meaningful difference. So that's roughly the majority of experts will tell you, that in terms of consensus, it's between 20%, 30%, something like that. So we feel very confident we are powered to see that. And of course, in terms of interim analysis, the same thing. We don't disclose details of interim analysis. If we do them, when we do them, yes, that's -- there's obvious statistical and regulatory reasons to keep these things pretty tight.

Yes. Thank you. Thanks for your question, Keyur. We are coming to an end. Maybe 2 more questions. One from the chat, and then the last one would be for Richard Parkes. But one on the chat. I'm not entirely sure if we can address it, but this is from Peter Welford from Jefferies. He asks, "What proportion of patients starting Enspryng are naive versus or Rituxan or C5 therapies?" I'm not really sure if you can address that one, but it's -- I still wanted to read it for you.

I actually don't have that information, I'm sorry. So this is something that's a big interest. Maybe what we could do is we could try to follow up on that. For sure. Okay.

Yes, we'll do that. Yes, yes. Peter, thanks for the question. We'll get back to you. Then the last one is for Richard Parkes.

Richard Parkes from Exane BNP. Just coming back to gantenerumab and the donanemab data that we've just seen, I think many people kind of walked away from that presentation being more confident that if you're engaging beta amyloid, reducing beta amyloid, there is a clinical benefit. But it looks like that benefit might be quite modest. So can you just talk about your thoughts on kind of effect size that you need to show to be clinically relevant and what you've discussed with the regulatory agencies? And then on top of that, if that's the case, that there's clinical benefit, but it's modest. It looks like beta amyloid reduction is maybe not enough to really sort of drive a real change in this disease. So can you talk about what other efforts you're doing in terms of combinations? Maybe following beta amyloid treatment with an anti-tau antibody. Can you just talk about what progress you're making there?

Yes. No, I'm very happy to. First of all, I agree with you that the picture I got as well was one of encouragement, given that, again, it's another independent trial that looks at targeting amyloid and gives us some clinical benefit. Of course, there are differences between molecules, there are differences in terms of dose, in terms of target, in terms of duration of exposure, et cetera. So yes, that modest clinical benefit one has seen there does not necessarily mean that's what we're going to see with gantenerumab, or for that matter, what we've seen with aducanumab. So there are differences, right? And there's no real consensus, there's no real regulatory bar here about what's clinically meaningful. I mean, statistically significant is one thing. And of course, that's why we power studies to detect a certain change. Again, if you talk to people in the field, generally, between 20% and 30% is considered to be the minimum bar for what is meaningful, yes? But I do agree with you, which is that, that is not a cure by far. And it's certainly a first step. So that's the way I kind of look at this. It's a little bit what we were doing with cancer therapy a few years ago, right? You have to have a foundation on which you build. So you need a first drug that shifts the course, and that's what we hope drugs like gantenerumab are going to be, they're going to be a first step in the right direction. And then of course you need to think about either combination or sequencing in different populations. We are very interested in that. We do have other internal programs. So we -- as you probably know, we have 2 tau antibodies in Phase II that we're studying now in prodromal, mild and moderate patients. Both now -- or I should say now in monotherapy, but we do have interest in looking at potential combination therapies. For combination therapies, of course, there's one thing that we're still all hoping for, which is we need that foundational basis. We need the first therapy to be approved to be able to combine something with. Otherwise, it's just too complicated from a design standpoint. So we're all hoping that, yes, for example, gantenerumab will be successful, and then we can start really actively planning for trials with combinations of taus and gantenerumab, et cetera. And then there are, of course, other mechanisms we're interested in which are, let's say, further in the preclinical stage. And we're not disclosing them now, but really around immune response, around senescence, around apoptosis. I mean, there's a variety of mechanisms that I think are going to be necessary if one would hope to achieve a higher percentage of shift in these curves and even hopefully a complete suppression of this disease.

Thanks a lot to all of you. Thanks for your interest in Roche. Maybe had 140 people roughly on the call, which is really good. I wanted to thank Gerard Tobin, Sabine from our team, also Melanie, for helping organizing the whole thing, preparing the slides. For the speakers, Simona, you, Paulo. And wishing all of you a nice day and stay healthy. All the best to you. Bye-bye. If there are any further questions, please let us know and drop us an email. Bye-bye.

Thank you. Bye.

Thank you.

Bye-bye.