Roche Holding AG (ROG) Earnings Call Transcript & Summary

Ladies and gentlemen, welcome to the Roche live audio webcast on the new AAN 2020 data. I am Sandra, the Chorus Call operator. [Operator Instructions] And the conference is being recorded. [Operator Instructions] The conference must not be recorded for publication or broadcast. At this time, it's my pleasure to hand over to Karl Mahler, Head of Investor Relations. Please go ahead, sir.

Yes. Thanks a lot, Sandra. Sorry for the slight delay. Maybe that's a part of our new world here. But what I can confirm to you that you can by now download the slides from the AAN presentation right now because the presentations just finalized at AAN, the official one, and you can download by now all the package which you want. So feel free to do it. Good afternoon, good morning, ladies and gentlemen. A warm welcome from my side. I'm very happy that we have Professor Laurent Servais with us. He is in U.K. Oxford at the Neuromuscular Center, and he will share with us the FIREFISH Part 2 data for risdiplam that were originally accepted for presentation at the AAN 2020. This will be followed by Paulo Fontoura, our Global Head, Neuroscience and Rare Diseases Clinical Development at Roche. Paulo will update us on the clinical program in patients with Type 1, 2 and 3 SMA. And after this, we will share -- he will share data on OCREVUS originally selected for presentation at AAN and the latest information on the shorter infusion opportunity. Then we have the Q&A. And before I forget it, I also wanted to thank Sabine Borngräber from our team. She was really instrumental in setting up this call, helping to prepare the presentation and so on. So in view of the fact that we are a bit late, only a few remarks from my side. Slide 5, you have seen the overview of expected key Phase III pivotal readouts, 2 of them have already read out positively, the Venclexta VIALE-A and the first-line AML and the FIREFISH Part 2 for risdiplam, more to come, with etrolizumab port delivery system. I can confirm that all these trials, the pivotal trials are actually really ready for readout this year as we have previously announced. And then on Slide 6, I'm leaving over to neuroscience. This is basically the topic of today's call. We are very proud to be launching 2 new molecular entities in neuroscience, risdiplam and satralizumab. And with the data which we are going to present today, which we are sharing with you, we are, let's say, looking forward also to round up the profile for patients, for physicians for further use. And with this one, I wanted to hand over to Professor Laurent, please.

Thank you very much. So I'm going to present you the data on behalf of a very large group of investigators. As you may imagine, next slide, Slide 8, we have a couple of disclosures, all of us. So Slide 9, so we talk about Type 1 SMA, which is a severe progressive disease. Patients used to die before the age of 2 years in absence of treatment, and it's highly -- it's very unusual to see an improvement in these patients. When a patient has been diagnosed, he is only deteriorating. FIREFISH concern -- is a confirmatory study of efficacy and safety for risdiplam. Risdiplam is an orally administered compound that is centrally and peripherally distributed and that interferes with the pre-mRNA splicing of SMN2 to increase the level of functional SMN protein. So that's the 12-month results of FIREFISH Part 2 that's included 41 patients. And so that's the 12-month results. So we will present today the primary endpoint as well as additional motor milestone, survival and swallowing data. So this is a global study, including Russia, China, Saudi Arabia, Brazil, United States and countries in Europe. And the inclusion criteria includes only symptomatic patients starting the symptoms before the age of -- between 1 month and 3 months. So patients were included between 1 month and 7 months. And they have all 2 copies of SMN2 genes. So the primary endpoint is the proportion of infants sitting without support for 5 seconds at month 12. So it's defined by the BSID-III. And there were key additional endpoints like time to death or permanent ventilation, achievement of motor milestone at month 12 as measured by HINE-2 and the proportion of infants who achieve either an increase of 4 points, is at a total score of 40 points on the CHOP-INTEND. We will also detail the bulbar function or the ability to swallow and to feed and an interesting outcome, which is the number of nights that you spend in the hospital during the trial. Of course, we'll discuss about this. So first thing first, the primary endpoint. So across the 41 patients who were included in the trial, Slide #11. No, sorry, first, the baseline characteristic of these patients. These patients are pretty old for Type 1, 5.3 months. And they were typical Type 1b patients. So with the onset of symptoms between the age of 1 month and 3 months and a disease duration that was quite long, 3.4 months. And this reflects into the CHOP-INTEND score at baseline of only 22, with some patients scoring as low as 8 at baseline. So let's go to Slide 12 for the primary endpoint. 29% of patients achieved after 12 months the ability to sit and support for at least 5 seconds. So it's 12 out of 41 patients. And in comparison with the natural history study, it's, of course, many more because you don't expect actually any patients to keep the sitting position in the natural history study. So the p-value is far below, 0.0001. Regarding the survival, next slide. This is actually the event-free survival. So among 41 patients, 38 survived and 35 were event-free, which means that they did not reach either more than 16 hours ventilation, either invasive ventilation. I must underline that if during the trial a patient had achieved this respiratory endpoint, even if afterwards is doing better, it will still be considered as having achieved this ventilatory endpoint. So at the end of the day, 93% of patients were alive and 85% then achieved one of the respiratory endpoint by 12 months. This is, of course, very different of what we know from patients with 2 copies of SMN2. Next slide, Slide #14, the HINE score. As you know, HINE is divided in -- HINE-2 is divided in several assessments. You have the head control, sitting, voluntary grasp, ability to kick in supine, rolling, crawling, standing, walking. You it's -- can score it most of time from 0 to 4 or from 0 to 3 for some items like the grasp or the standing and 0 to 2 for the sitting for the head control. So most of time patients in SMA do not progress in any on these items. So next slide, Slide #15. For head control, 76% of patients improved. It was wobbles for 32% of them and maintaining upright all the time the head for 44% of them. For sitting, 61% of patients improved on this item. 17% were only sitting with support at hips, 20% props, 15% stable sit and 10% pivots. For standing, 22% improved, 17% actually were supporting weight and 5% were standing with support. And for walking, 2% were actually bouncing. So altogether, 78% of infants responded to treatment using the HINE-2 scale as predefined response criteria. This is Slide 24. Let's move to Slide 25. And this is the CHOP-INTEND on the 38 survival. The 2 values that we had to predefined was either an improve by more than 4 points that define a responder, either the fact that the patient reached plus 40 points on the CHOP-INTEND. As you can see, all but 1 patient who has survived, improved by more than 4 points on the CHOP-INTEND. And the median change was 20 points. And this increase in the CHOP-INTEND was very heterogeneous with some patients improving up to 45 points and some patients improving by 6 points. So altogether, it's all but 1 patient who survived, who improved by more than 4 points on the CHOP-INTEND. Next slide shows that this progress goes quite fast with a median increase at 2 months that is 7 points, which shows that actually the mode of action is quite rapid. And then there is a continuous improvement over the different visits. It is actually 56% of patients who achieve a score of more than 40 on the CHOP-INTEND after 12 months. In natural history study, this most of time do not happen. Next slide, which is Slide 27, 2 important feature for bulbar function which are swallowing and feeding. In terms of swallowing, if we look at the 38 patients who survived after 12 months, 95% of them, actually 36, maintained the ability to swallow. If we go to feeding, among the 38 patients, 89% of them were able to feed orally and 74% of them were able to be fed exclusively orally without any support. And this, of course, contrast broadly with what we know from natural history study in which no patients after the age of 12 months can be fed independently. Let's move to Slide 32. An important finding is the number of nights that the patient spend in the hospital. And this is the number of night regardless of the reason why you spent a night in the hospital. So it can be for elective installation of an invasive ventilation to address respiratory, whatever. And we know that the number of nights that you spend in the hospital is a very important point for quality of life of the whole family and that untreated patients may spend weeks in the hospital and then go back home for 1 or 2 days and then come back to the hospital. Here in this study, 49% of patients did not spend a single night in the hospital over 1 year, which translate into a major improvement of quality of life for the families. And then you have some patients who spend 1, 3 nights, up to 7 nights, which is probably not for pneumonia, but more for elective hospitalization. In natural history study, it's between 4 and 7 hospitalization every year, taking into account that this hospitalization may be quite long. Here, the median hospitalization, the rate was 1.3 per year, so below what we know from natural history study. Regarding safety, there were many adverse event, of course, but none of them were actually led to withdrawal of treatment discontinuation. Next slide, which is Slide 34, about the serious adverse event. The most common, as we could expect, were upper respiratory tract infection. And actually, the frequency of this serious adverse event declined through the course of the disease, with a decline by more than 50% after 6 months of treatment. And there was no risdiplam-associated ophthalmologic findings in the study as in the other studies of risdiplam. The most common side effect that was associated with treatment was actually the rash, the skin rash that was self-resolving. So altogether if I want to summarize this data, it's Slide 35. 29% of babies able to sit independently without any support for at least 5 seconds. 93% of survivors and 85% of event-free survivors. 95% of patients alive will maintain the ability to swallow. 49% of patients who did not spend a single night in the hospital, no drug-related safety findings leading to withdrawal and no ophthalmological finding and motor milestones that were achieved in a very significant percentage of patients. So just want to conclude on Slide 36 on a big, big, big thank you for everybody who participated in the data collection in this very large study. So happy to take your questions.

Well, maybe if we could maybe first do Paulo Fontoura's part shortly and then go back to the Q&A so that we have all in one package if that's fair to you.

Okay. Yes. Sure.

Thank you. Thank you.

All right. I guess I'll take it over from this. Yes. So we're on Slide 37. Good afternoon, everyone. Thank you for joining. So on Slide 38, we just wanted to show you our current most advanced pipeline in neuroscience. And neuroscience is one of the fastest areas of growth for Roche, generally. And we see here an opportunity to really develop life-changing medications for a number of really, really severe conditions. Obviously, you're very familiar with OCREVUS and the impact that it has had in the MS field. We're planning on launching this year, hopefully, 2 new therapies, satralizumab for NMO spectrum disorder and risdiplam, which you've just heard about for spinal muscular atrophy as the first oral therapy. And more broadly, as you can see here, we're really aiming for medicines, which should be either first or best-in-class and would really represent a fundamental breakthrough in terms of how these patients are being treated. Next slide, please. Just to briefly highlight as well on satralizumab that we actually just now published in Lancet Neurology the results of SAkuraStar, our second pivotal trial of satralizumab and NMOSD. This is the monotherapy study. We had previously published the SAkuraSky study in The New England Journal of Medicine, which looked at patients who were treated on baseline corticosteroids and immunosuppressant. And both of these data sets together just really emphasize the very broad and robust safety and efficacy profile that we see with satralizumab in these patients. And we're still on track to launch in the U.S. or EU in the second half of this year. Next slide, please. So coming back to the risdiplam program in spinal muscular atrophy, really the point we wanted to make is that when we designed this program, we wanted to go very, very broadly and look at patients across the whole range of phenotypes, both Type 1, Type 2 and Type 3, but also patients who were both naïve and had been treated previously with other therapies. And what we're seeing really is a very compelling efficacy and safety profile which is very consistent across this age range. So we have positive efficacy in Type 1 infants, the data from the FIREFISH study which Dr. Servais just mentioned to you. We have positive efficacy data in a very large placebo-controlled trial in a broad spectrum of Type 2 and 3 patients with the SUNFISH study. And we had very consistent safety as well which is equally important with the population. And so we believe that risdiplam could potentially be the treatment of choice for a majority of patients living with spinal muscular atrophy. In the next slide, we're showing you a little bit why we say that. So our program was really purposely designed to look at a broad range of age and disability type. So not only Type 1, 2 and 3 SMA, naïve and pretreated, but also including your real-world spectrum of SMA patients, patients with scoliosis, contractures, really low baseline motor scale scores. Now, of course, there's been a lot of attention and a lot of studies in the earlier, more younger population of Type 1 babies or patients under 5 years old where most of the trial had been conducted. But in our risdiplam program, we've really spread out and encompassed the whole phenotype and range in terms of babies and the mobility scores and actually extended to the patients beyond 5 years old, which actually constitute the most prevalent population, about 85% globally. So this is a very large prevalent population usually to right now really does not have good treatment options. And so we feel that our data really supports the broad use of risdiplam across all of these patients. Next slide, please. Just to give you a brief update on the JEWELFISH study. So the JEWELFISH is a very large open-label study of non-naïve patients that have already completed their enrollment. As you can see here, we have 174 patients. And we have collected data on the prior SMA therapy. As you can see here, a broad range of other therapies, including Spinraza and Zolgensma as well. The data that we're going to show at the American Academy this year was preliminary safety data from 45 patients who had received risdiplam for about 30 months. And really, the safety profile is very consistent to the ones that we've observed in FIREFISH and SUNFISH as well. We'll be -- we are collecting efficacy, which obviously is exploratory in this case, and we hope to report that in 2021. And also, we are collecting data on the reasons for switching therapy from their prior disease modification therapies with risdiplam, which we hope to present later this year. Next slide, please. Also to finalize on our regulatory update. As you know, recently, we've announced that the FDA has moved backwards the PDUFA date to 24th of August, which is really due to their need to review a very large data set of SUNFISH Part 2 data, which we think is actually great because it hopefully will mean that with very broad label that we'll encompass all types of SMA patients. We're on track to submit mid-2020 to the EMA the fully integrated data set. And we do have, as you know, unique EU PRIME designation. And finally, just a couple of weeks ago, we've announced that we filed in China. And therefore, we hope to have a fast review there as well and bring risdiplam to patients in Asia soon. Next slide. So switching gears a little bit and I wanted to give you an update on OCREVUS and the data that is going to be shown in the next slide. First, just to set the ground a little bit. OCREVUS is really the first and only treatment approved for both relapsing and primary progressive MS forms. And right now, it is the market leader in the U.S. And we think that is really a representation of the benefit that physicians and patients recognize in the therapy which, first of all, has really outstanding efficacy data, has very good safety profile with over 150,000 patients treated already. No one confounded PML cases so far. It's a very convenient therapy. Actually, it's twice yearly. And you can see here on the left-hand side of this graph here, we actually show the adherence. And a very substantial, over 90% adherence to therapy compared to other modalities such as other IV drugs, oral therapies and injectables. And finally, it is a very accessible therapy with a pricing strategy that enables the very broad payer coverage. And as you can see here on the right-hand side as well, we do continue to grow the share of OCREVUS. And that is based, I think, on this very convincing benefit/risk profile as well as this convenient profile. Next slide. So what we do think is that there's -- while OCREVUS is obviously an amazing drug, and it's great news for patients, one aspect of its benefit that we'd like to really highlight is the impact on progression. And progression from a clinical standpoint for MS patients, it's really what matters. Long term visibility progression is the biggest determinant of long term health outcomes. So 2 on 2 of the data sets we're going to show this year at the American Academy, one of which is around visibility progression in patients with relapsing forms and the consequences and the risk of reaching and EDSS score of 6, which is walking aid required for ambulation. We do know that after EDSS 6, there's a much higher risk of irreversible disability. So while in the previous EDSS stages, patient scores can fluctuate and you can become better occasionally. After EDSS 6, the chances are much, much higher that you're going to progress and become wheelchair-bound and have a much worse health outcome. And also that it is from EDSS 6 that the impact on quality of life and employment in this case really declines. So we thought this was a really important clinical milestone to look at. And the second thing is really the impact on tissue damage. And in this case, we're looking at thalamic atrophy. Now thalamic is a deep gray matter structure under the basal ganglia which integrates a lot of the information about motor and sensory control, attention and cognition. And therefore, it is clinically very relevant and is also not as susceptible to variability as looking at, for example, cortical atrophy. Next slide, please. So I'm showing you here the OCREVUS 6-month (sic) [ 6-year ] follow-up data from the Phase III OPERA study in relapsing MS. As you know, after the first double-blind controlled period, all patients were put on OCREVUS. And as you can see here, we do see a very substantial reduction in the risk of reaching EDSS 6, which now at year 6 is about 50% risk reduction overall. Of course, as I mentioned, this is a clinically very important milestone. So the fact that we're seeing a half, a 50% reduction in risk is really remarkable and we think compares very, very favorably with any other drug out there and just confirms the very robust profile that OCREVUS has in terms of being the drug with the best checks on progression overall. Next slide, please. And here are the data for the thalamic atrophy. So you see both on the left-hand side data from OCREVUS versus interferon beta-1a in the OPERA studies. And on the right-hand side, the same data for OCREVUS versus placebo in the ORATORIO PPMS study. Again, a very robust suppression of the progression of thalamic atrophy. And we feel this represents, again, the impact that OCREVUS has in terms of protecting brain tissue loss. Next slide, please. Now while these data are very, very important, what we realize as well is that OCREVUS is a very convenient therapy given twice a year, but the most we can do to make it even easier for patients to access, particularly in these difficult times right now when more challenging access to health care facilities, probably less infusion [indiscernible] capacity, we thought that this is really important to look at faster infusion of OCREVUS to essentially halve the administration time. And the data here, the data from the ENSEMBLE PLUS study compares essentially the conventional infusion rate with total administration time at 3.5 hours to a shorter infusion which is only 2 hours. And obviously, the big concern here would be about tolerability and safety. And we can show here on the right-hand side, it is really remarkable that the overall infusion-related reaction rate is essentially the same between these 2 arms. So we are not seeing an increase in safety events with faster infusion, which is obviously great news for these patients and we believe supports the approval of this faster infusion rate. So actually, we have filed both in the U.S. and EU in April of this year, and we hope to have a fast review of this filing. I think that health authorities recognize the importance of a faster infusion regimen, not only overall in terms of patient convenience, but also during these more difficult times now with the pandemic in terms of being able to have more patient access to therapy. Next slide. So in summary, the story for OCREVUS continues to be a very, very positive one with over 150,000 patients treated globally, very consistent and very favorable benefit/risk profile. Again, we showed new data now that shows a clinically meaningful, about 50% reduction in disability progression for very clinically meaningful milestone of EDSS 6. Also a very substantial reduction in thalamic volume loss, which again represents the effect on tissue protection of OCREVUS. And finally, the really great news for patients with a shorter infusion regimen that allows them to have easier access and more, even more convenience, which has just been filed with FDA and EMA. And I believe that's my last slide. So I guess I'll stop now and turn it over for questions. Thank you.

Yes. Thank you, both of you. And operator, Sandra, we are now ready for questions, please. But just as a kind of information for all of you, we have about 150 participants on the conference call via the telephone line and about 250 over the webcast. So it's quite a large conference call. So please.

[Operator Instructions] The first question comes from Hector, Luisa from Berenberg.

It's Luisa Hector at Berenberg. My questions are for Professor Servais. You noted that the age of the babies was quite high in the study at around 5 months. Would you expect earlier treatment to yield even better results and -- or perhaps less side effects? And when you look at that cohort of infants that were treated, would you now at this point, so after a year of therapy, consider them eligible for gene therapy? And the same, if you're seeing an infant who's had gene therapy, would you be comfortable in treating them with risdiplam?

So thank you very much for this multiple aspect question. So first point, even if I don't have the data in hand to answer you, but this is my pure speculation. Yes, I believe that treating early will definitely show you better results. And this is true not only for risdiplam, but for any disease-modifying treatment in spinal muscular atrophy. There are lots of evidences about that. We have recently review it and published it. The first author is Tamara Dangouloff, and there are plenty of preclinical models to show that. The second part of the question regards side effects. The side effects that can be attributed to treatment are, of course, very mild. It's like for the rash, but there is no severe side effect or safety issue related to treatment. So I do not anticipate that this could decrease because it's already very low. If you refer to serious adverse events, like for instance, pneumonia, certainly, if we treat patients earlier, we will have less pneumonia and less respiratory tract infections in these patients. So then come the third part of the question. Should these patients now be eligible for gene therapy? Well, within the European label, the answer is yes. They are -- the European label is very broad. And within the U.S. label, the answer is also yes because the U.S. label include patients up to the age of 2 years, and these patients are not 2 years at the time and after the final endpoint. But the question is, is it meaningful to switch a patient from risdiplam to nusinersen or to gene therapy or to switch a patient or to give a patient who is already on -- who has already received a gene therapy another medication? There is not the singlest evidence around to say that it is meaningful. So I would say that if someone is happy to pay for that and that the parents are happy to go for that, then comes the question of safety. There are not so many cases reported. As far as I know, whatever is the sense of the direction of the switch, there were not unexpected findings reported, but the number of cases that are known are extremely -- is extremely low. So I will not make it a rule. But so far, in the very limited numbers of cases that are now who have been switching either if they've been reported informally or not, there is not kind of unexpected events. But there is not the singlest evidence that it's -- that you may by shifting SMN acting disease-modifying treatment that you could get an additional benefit. So I think that this will probably be more a question of the route of administration and how convenient it is to deliver the medication to the patient.

Thank you. Luisa, does that answer your questions?

Yes. That's super, Karl.

There's a question over the web from Steve Scala from Cowen. He asked various questions. He was asking why the decrease in the proportion of patients sitting at the primary endpoint versus that what we saw in Part 1. So why is there a difference between Part 1 and Part 2? And I think you already answered that question. Is the physician concerned on the respiratory infections with pneumonia or any kind of other adverse events? And the last one is, does Roche anticipate any ocular monitoring on the label? So basically 3 questions.

Well, so starting with the last one. I can just tell you that across the different trials with risdiplam, there was not a single ophthalmological finding. So I think the show should now be over because these ophthalmological investigations are super burdensome. I mean we're talking about examining a kid 8 months old. And for those of you who have already seen an 8 months old baby, they don't use to open their eyes when you ask them to do or to do exactly what you ask them to do. So it's super burdensome. And for the ophthalmologist, they would never agree to go to a much broader population of patients. So this has to be decreased or suppress for the label. Otherwise, it's going to be very difficult to organize, at least in the sites in which I've been working, and I've been working in 3 sites. So that's the first point. And the second part of the question is the proportion of sitters. It depends if you take cohort A plus B in the Part 1. If you take cohort B, yes, it's slightly below. I think it was 41% in cohort B and 29% here. First thing first, not sure it's significant because we talk about persons. But when you have 17 patients because actually in cohort B of Part 1 it was 17 patients. A single patient will account for 6%, right? So 1 patient more, 1 patient less means 6%. So I'm not sure that the difference is significant. If we have to find some explanation, we cannot find it on the baseline because the baseline characteristic, patients who are even older in the Part 1 study. I would say that the Part 1 study, most of patients actually were included in 2 sites in Milano and in Paris with quite good standard of care. And the FIREFISH Part 2 was a much more international study, including centers in Brazil, in China. And so I think it's quite a common feature, especially in neuromuscular disease when you move to Phase II studies in a few selected centers with patients who are very closely followed up to a much broader pool of centers. So probably it's slightly below. I'm not sure it's significant. We must take into account the very low number of patients before speaking in terms of percent. And I think that the main explanation resides in the different sites, in the different of the geography of patients.

Next question, please. Maybe we can take one over the phone.

The next question comes from Michael Leuchten from UBS.

Two, please. One, just going back to Professor Servais. If I look at Slide 25, the percentage change in CHOP-INTEND scores. Do I take the comment, if we were to overlay this with age of patient or duration of disease, the younger patients would sit to the left of that chart and the older ones sit to the right of the chart? And then a question for Paulo. Thank you for the new patient share chart. What's the reason why it's sort of stuck at 40%, which is very good? But the plateau here, given the broad label that OCREVUS has, what's stopping it from going higher from here? It's been sitting at that level for a little while.

So I can take the first part of the question. I would anticipate that indeed, the younger patients will be the best performer. This being said, the analysis to identify the best responders is ongoing, and it's ongoing not only on FIREFISH Part 2, but also to pull FIREFISH Part 2 and Part 1 to get a broader population which will allow to test more robustly the different outcomes. So this is a work in progress. And I anticipate that, yes, but I don't have the data to formally answer you today.

Yes. On the 40% number, Paulo, maybe you give it a shot. But maybe the 40% increase of new patients here, it's really not bad, I have to say. The market might forgive us maybe 10% a year ago. And now you're not happy with 40% now. But maybe Paulo, you can answer.

Yes. Maybe a couple of comments. First of all, I think, Karl, you're kind of hitting on something which is, so this is a number that won't grow to 100%. I mean, we have 15 approved medicines on the market, so there's lots of patient choices. And obviously, at some point, yes, you can only grow so much in being first to brand, right? The second point, which may be true, although pure speculation is, we are aware that during the current pandemic, there are some potential either hesitations about starting new therapies or things like that. And therefore, it's not unexpected that you'd see the numbers not rising as fast. But again, we have no data on that so far. Suffice it to say that we think that the data we're putting out really will continue to support patients and physicians in their decisions to either start patients on OCREVUS or switch to OCREVUS. I don't know if Paul Byrne is on the call and maybe you want to offer something.

Yes. Thanks, Paulo. Thanks, Michael, for the question. So I think maybe to reiterate what Karl said, obviously, 40% is a pretty big number in terms of new-to-brand share. So we're looking at new patient starts and switch patients here. And OCREVUS for some time now has continued to be the #1 prescribed for new patient starts. And actually, Michael, you will see that this actually has moved quite considerably, probably from the sort of hovering below the 40% that it's now sort of at around 44% from the most recent data that we have, which is a 3-month rolling average data that you see there, which is now at 44%. And actually, we're quite encouraged in terms of the sort of movement into earlier line use as well that we have been seeing towards the back end of last year, which has given us a strong momentum and reflected in the results that you have seen in quarter 1.

The next question comes from Elizabeth Walton from Crédit Suisse.

It's Matthew Weston from Crédit Suisse. Two questions, please. The first, Professor Servais. The upper respiratory tract infections and pneumonia, you mentioned them as being disease-related. But in studies in SMA, we do see an imbalance, which tends to see a greater frequency in the treatment group. Can you just explain why we see that? Is it to do with solid food, which leads to more problems in the treated patients? Anyway, I'd love to understand why that's drug-related, but the imbalance goes against the drug. And then a question for Paulo. It's the first time we've had an opportunity to speak to you since the Huntington's program was fully recruited, the GENERATION study in Phase III. If you could just remind us of your expectations on time lines for that important program, that would be a great help.

One, for this specific study, there was no control group. So it's difficult to make your opinion. This being said, if you have a high rate of upper respiratory tract infection, I would consider this as an achievement because it means that the adverse event is recorded as an upper respiratory tract infection and does not go down as a low respiratory tract infection. Upper respiratory tract infection is something extremely common in kids less than 1 year. It's supposed to keep in upper respiratory tract, not going low. The problem that we face in SMA Type 1 patients is that it goes low. So a high number of respiratory tract is for me the sign that actually doesn't move to pneumonia immediately. Regarding the rate of pneumonia, something interesting here because there's no control group is that actually it decreased with time. And they were half. The rate of pneumonia was 50% lower in the second 6-month period. So when you compare months 6 to 12 and months 0 to 6, you divide the rate of pneumonia by 2, which means that actually when the patients get stronger, the number of pneumonia decrease. So I think that it's how I would understand the issue. The other point is that in an untreated patient, I would not anticipate to have a lot of pneumonia because if they have a pneumonia, most of the time, they will not survive. So if the patients survived to pneumonia and can get a second one, at least it means that you have survived the first one. It's a little bit cynical, but unfortunately, that's the tragic history of SMA Type 1. Thank you.

Yes. And regarding your question on the Huntington's program, I think you're referring just for everyone on the call. We recently announced the completion of enrollment of GENERATION HD1, which is our first pivotal Phase III study in Huntington's with early manifest disease. Now of course, the treatment duration is 1 year so the primary endpoint should read about a year from now. We are keeping a very close eye on that study. At the same time, as you know, we have other programs which are ongoing, including the continuation of the Phase Ib trial, which is open label and where we know patients are on drug. In terms of the regulatory time lines, again, our primary focus is on executing these programs really, really well and try to achieve a label and a convincing data set that would enable access worldwide. Of course, we do keep monitoring the data from the open-label trial. And if at any point, we do see that there is sufficient evidence there to approach health authorities that option is still there. And as I said previously, we are not going to rush this. I think it's important that as soon as we're convinced that we have enough data and that we approach regulators and they've been open to that. But of course, our primary focus is still thinking that the GENERATION HD1 is going to be the primary source of data for our filing.

Next question comes from Richard Vosser from JPMorgan.

A couple of questions for Professor Servais. So firstly on the comparison of the CHOP-INTEND score to other therapies and by age. Compared to Spinraza, how do you see this data? How will it influence your treatment decision based on the data that's come out from FIREFISH? And then a question also on SUNFISH, which I think you were involved in as well. Is there a cutoff in terms of age based on the SUNFISH Part 2 data where you wouldn't decide to treat based on the data in SUNFISH or would you think to treat patients of all ages?

Thank you for the questions. So for question number one, if you have a look at first, the population of ENDEAR are pretty much with the population of FIREFISH Part 2 in terms of baseline so we can compare them. I will not say the same for START, for instance, or STR1VE U.S. who are younger. But for ENDEAR, we can compare. Of course, we must be careful because on one side, you have a double-blind, randomized, placebo-controlled study. And on the other one, you have an open-label study with a concomitant approved medication. So it makes things quite different. But nevertheless, you have another limitation for the comparison is that the ENDEAR study was interrupted after 9 months. And if you look at the data of the evolution of the CHOP-INTEND in the patients, you see that we still observe an increase between the age of 10 and 12 months. So the follow-up if you want to compare with ENDEAR is a little bit longer here. Nevertheless, if you look at the data from the chart in The New England, you don't have -- certainly not 90% of responders on the CHOP-INTEND, right? And the median increase was not 20 points, it was lower. So it's certainly noninferior, I would say. That's certainly the least we can say. The same apply for sitters. You are not at 29% of sitters. I agree that the follow-up was shorter. And in terms of survivors, you had certainly not 85% of event-free survivors, you had like something 66. And then the follow-up here is longer. So it plays more in favor of risdiplam. So with all the limitations that I've raised for comparing the 2 drugs, the least we can say is that there is noninferiority with the existing treatment. Then I'll leave to Roche to try to match exactly the population and to try to figure out if there is reason to believe that the drug is even working better than nusinersen. Something that is not in ENDEAR, but that is more from our clinical experience and it's definitely not the secret because it comes again and again during communication in congresses that Type 1 patients treated with nusinersen, most of them experience bulbar problems and sooner or later difficulties in swallowing or feeding. I think that it's a great achievement with a population of patients with more than 5 months old at the beginning of the study to get such a high number of patients with possibility to swallow at the end of the study after 12 months. So as long as you have certainly no inferiority and maybe a better efficacy, as long as you don't have any safety issue, an oral drug versus an intrathecal drug is something quite attractive, not only in the situation of a lockdown when it's difficult to make your patients come into the hospital, but in general, I mean, I've never met someone happy to have an intrathecal injection or a mother that was happy to bring a child for an intrathecal injection. So if we can go for an oral therapy, it's going to be much better for patients and for the hospital because intrathecal injections, when you have a lot of patients, it's quite burdensome. So to summarize, I'll say there is certainly no inferiority on several outcomes like CHOP-INTEND responders or survivors. Swallowing ability, probably even better. But of course, we have to be very careful in this comparison because it's not exactly the same duration and with all the limitation that I have raised. And on the safety, both drugs do compare. But on the burden of administration for patients and for physicians, there is no matter of comparison, of course, between an oral and an intrathecal drug. Then comes your question about SUNFISH. It doesn't come to me as a surprise that younger patients respond better than older patients. I mean it could have been largely anticipated. It's not a surprise. For me, it's a surprise that you can see improvement in all the patients. It was, for me, a surprise to see that all the patients with nusinersen, I could see an improvement. I've published these patients and others have also. And we see also patients on risdiplam improving. In my experience, I have observed patients at a quite advanced stage of disease with nusinersen still improving. So if I started treatment with nusinersen in patients who were Type 1 40 years old, Type 2 50, 55 years old, so if I did using an intrathecal procedure, as you might imagine, I will have absolutely no problem in starting an oral medication as long as, of course, it's reimbursed by the social security. So the basic and short way to answer your question is no age limit for me.

Thank you. Maybe Paulo, you would like to further comment on?

No. I mean there's really nothing else to add. I think Dr. Servais covered it very, very well.

Yes. Yes. Okay. We have 2 more questions in line. And of course, we want to make up for the slight delay. Maybe we can take 2 more questions, the 2 last questions which are in the queue. Operator, Sandra, please.

And the next question is from Andrew Baum from Citi.

Okay. Perfect. Yes. Thank you. There is one left. Okay. That's good.

So my question relates to Medicare coverage of patients with MS, in particular, in relation to OCREVUS. I think Medicare is about 30% of the total MS patient population, but lower than that for OCREVUS. My question is, given it would seem to be financially advantageous for Medicare patients to be reimbursed under Part B given the lesser copay with Medicare compared to the copay, why isn't it higher? Is that because just inconvenience of patients being unable to travel to infusion clinic or is there some other reason? I'm obviously thinking about the extent to which you could limit the competition from ofatumumab given the lower financial burden for the Medicare patients.

Maybe, yes, maybe, Karl, I could?

Yes, Paul, you can take it. Yes. Yes. Paul is the Lifecycle Leader for OCREVUS. So please, yes.

Yes. Thank you for the question. And maybe I just would start in a more general way, I think, which is, whether it's Part B or Part D, that we're actually very pleased in terms of the coverage we have at the moment. We have 4 of the top national health plans now covering OCREVUS in RMS with no step edits. We have 98% of insured patients who actually have coverage or access now to OCREVUS. And actually, 82% of these have no step edits on the RMS side and 96% on the PPMS side. So I would say, whether it's commercial or whether it's Medicare, we have sort of very strong and broad access. And so we anticipate that we're in a very strong position, irrespective of who may come or what molecules may launch in the marketplace. We also, I think, what's sort of interesting and important here is that a lot of infusions now actually are taking place outside of the hospital setting as well, either in private clinics or also in ambulatory infusion settings. So in terms of just access, either from a reimbursement perspective or in terms of site of care access, we're very pleased in terms of what we see in terms of progress in the U.S. right now. So I think I would approach it more on that basis in terms of where we're at and how we're set up. We also interestingly have set up and we've been doing this now for quite some time over the last 12 months is what we call patient navigators as well who actually are dedicated to supporting patients through the reimbursement process and also trying to find site of care as well as alternative infusion sites outside the hospital setting. And all in all, I think we are in a very strong position in terms of access for patients, whether you're looking at it from reimbursement point of view or from a site of care point of view.

Yes. Thank you. There was a last question coming in to Professor Servais from Jefferies from Colin White. He was just wondering about how you see the, let's say, the treatment schedule evolving for patients, let's say, Type 1, Type 2, Type 3 because we do have good news for patients now, various choices. And I guess also the physicians will have to make a choice or they see the landscape evolving. What is your take on how the whole thing could move in, let's say, a year or 2 years or whatever it takes. Maybe this is the last question from the web which we can take.

Well, it's quite a complex question because, as you might imagine, this will depend broadly from the different population of patients that we have. I think, first, we must make the difference between what I would call the prevalent population, so the population who is here today and the population who is newly diagnosed. And in this newly diagnosed patient, we must make the difference between patients who are diagnosed because of newborn screening, which is an increasing pool of patients, and patients who are diagnosed because of symptoms. And then, of course, this population will belong to certain countries and certain regions. And this is going to be a moving target with the first population that decrease and an increasing population of patients who have been diagnosed by newborn screening. And this population of patients diagnosed by newborn screening at one time could become symptomatic and then it's going to constitute a fourth group, a fourth population of patients treated at birth who finally develop symptoms. So this is a very moving target. And two, broadly as far as your question. I anticipate quite some changes because, obviously, if for all the patients you have the access to an oral medication, it's quite a good opportunity to make things simpler, more simple by just giving an oral medication. I'm pretty sure that many patients will be and many parents will be very attracted by a one shot therapy at birth, especially for patients issued from newborn screening because these are parents who are still with the dream of a normal kit of medication. So the concept of a medication that could possibly put all of this behind you as a parent will be extremely attractive. Then for some patients issued of newborn screening, you can imagine that the balance between safety and efficacy and unmet need will come into consideration. And I think, especially about patients with 4 copies of SMN2, in patients with 4 copies of SMN2, that the balance of the benefit and the risk is completely different because you cannot accept any kind of risk in these patients because you are even not sure when they will develop the symptoms and even sometimes if they will develop the symptoms. I'm pretty sure that the treatment landscape and the profile of treated patients will broadly change. And the best way to figure out is to try to isolate this different subpopulation and to try to imagine what will happen. Of course, for all, the easiest case to answer is certainly patients issued by newborn screening, especially with 2 or with 3 copies, all the patients who are too heavy to receive gene therapy and who have scoliosis or spinal fusion that makes intrathecal injection sometimes very challenging. And between these 2 extremes, as you might imagine, there are different scenarios and possibilities according to the different country.

Thanks a lot. I mean it's good news that we have so many new alternatives evolving. I mean there was nothing a few years ago. Now we have Spinraza, risdiplam, gene therapy. There are lots of choices. This is good news for patients, good news for you as a physician because you have good choices as well for parents. And with this one, I wanted to thank you. I wanted to thank all participants. Thanks for your interest in Roche and looking forward to hopefully seeing you in person at a later point in time somewhere in the world and not only over the phone. And thanks to our speakers again. Thanks to Sabine for organizing the call and all the best to you. Thank you.

Bye-bye. Thank you.

Thank you. Thanks a lot. Much appreciate it. Thank you.

Ladies and gentlemen, the conference is now over. Thank you for choosing Chorus Call. Thank you for participating in the conference. You may now disconnect your lines. Goodbye.