Roche Holding AG (ROG) Earnings Call Transcript & Summary

Kindly note that the webinar is being recorded. [Operator Instructions] One last remark. If you would like to follow the presented slides on your end as well, please feel free to go to roche.com/investors to download the presentation. At this time, it's my pleasure to introduce you to Karl Mahler, Head of Investor Relations and Group Planning. Karl, the stage is yours.

Yes. Thanks a lot, Henrik, for the kind introduction. Welcome to our ASH science call today. This is the last one for this year. It's close to Christmas, and this is also a testimony of the good science, which we had in the past year, I think, on various occasions. So very excited to present an update on the probably largest and broadest hematology pipeline in the industry. And maybe from my side, to highlight definitely 2 things, there's a lot of things you'll hear today, but maybe 2 things. One is the Phase II for our bispecifics in mosunetuzumab and glofitamab, which got an oral presentation here. And one other highlight definitely was the late-breaking abstracts for POLARIX. I mean it was the first improvement you will hear about it since 20 years in first-line DLBCL. And going with the presentation, there was, in parallel, a presentation or a publication in the New England Journal of Medicine. And in summary, strong progression-free survival improvement and good safety. Looking at the feedback so far on the data, which we have presented, specifically also here on the POLARIX, very good feedback. So there was an [ Evoloid ] pharma survey, which was just put out after ASH with key opinion leaders, asking about the future standard of care and really practice changing data. And one, actually, we have to say also, the CAR-Ts actually had a good show up here at this year's ASH. But everybody also mentioned POLARIX has the potential in the standard of care and in line with the recent -- and this is also in line with the recent physician survey we did on the data and the strengths of the data. So let's go to the overview and the agenda, please. Yes. So this is the program for today. We have with us Peter Ahnesorg. He is the Franchise Head of Hematology in the Global Product Strategy group. And as I said before, I mean, this is the broadest and deepest hematology pipeline in the industry. We have 6 different approaches. We have in 8 therapeutic areas. So there is a lot of data now already available and to come. Then we have Charlie Fuchs. He will talk about the 5 assets, focus on 5 assets. He's the Global Head of Oncology and Hematology, Product Development. Just to mention here, Charlie just published a book on cancer. So I also want to get some benefit for this one. So I just wanted to mention that, Charlie, that is a standard book in cancer treatment. You can buy it on the book market. And then we have Franck Morschhauser with us, a leading investigator, a key opinion leader in -- for hematology, working on Polivy and new standards of care in hematology but also on the bispecifics, not just on the Polivy. And with this one, over to you, Peter.

Thanks a lot, Karl. Welcome, everybody, to this investor call at the end of ASH 2021. My name is Peter Ahnesorg. I'm the Franchise Head for Hematology at the Roche and Global Product Strategy, and I'm extremely pleased to be with you today. This has really been an historic ASH for our organization. We've contributed almost 100 abstracts to this meeting, showcasing advances in research across a multitude of malignant and benign hematologic indications. We had more than 30 oral presentations, and most importantly, we presented 3 data sets that we believe have the possibility to change standard of care in the future: the POLARIX study in frontline DLBCL; the HAVEN 6 study with Hemlibra in mild-to-moderate hemophilia A patients; and pivotal data for our bispecific antibody, mosunetuzumab in follicular lymphoma. So a lot to talk about. But before we hear from Charlie about some of the detail, I just wanted to take a couple of minutes here to talk more comprehensively about our hematology portfolio and why we're extremely optimistic about the outlook. And if we go to the next slide, as you've heard at Pharma Day, if you joined that and in another communication from Roche, we really have a bold vision in providing more patient benefit, but also being mindful about the impact to health care systems and to society that our therapies have. And that's also the guiding principle of our strategy in hematology, and there's 3 key pillars to how we're thinking about moving forward our portfolio. The first one is to really focus on our areas of strength, and that's why you see us bring forward important new advances in non-Hodgkin's lymphoma, an area where Roche has been present for the past 2 decades. And we're extremely excited about continuing to move forward in that space. And similar things apply for AML and hemophilia. Secondly, we're not going to stop there, but we're really targeting to go into areas that are new to us. And so most importantly, really, to call out multiple myeloma, where we believe we have a very strong pipeline, and you'll hear some about that. And then thirdly, we believe we continuously need to innovate to stay relevant and to move our pipeline forward. And you see that, if you, for example, look at our bispecifics program, where we're really taking a diverse number of approaches in how we're engineering these antibodies to really feed -- meet different needs of the market out there. But also in how we're approaching combination therapies, the partnerships we're striking and how we're exploring novel endpoints. So on the next slide, and I'm not going to go very deep into this, you can just see the breadth our portfolio across both malignant as well as benign hematology and across a multitude of different modalities. But as you can appreciate on the next slide, we really understand that just the breadth of the portfolio in and of itself is not going to make us successful. We're really building on our ability to understand the space, and we've demonstrated that. And if you just look at the number of breakthrough therapy designations in the U.S., that's really a reflection of the impact we've been able to make in hematology and the knowledge we have in that space. But more importantly, really, we're quite focused on going to lines of therapy where we can meaningfully cure patients because we believe that is really where future value lies for health care systems and not only sort of addressing late-line therapy, but ultimately curing patients in earlier stages of the disease. And one great example of that, if we go to the next slide, is how we're thinking about first-line DLBCL, where for the past 20 years, obviously, MabThera/Rituxan has been instrumental in defining therapy for patients. And we've had quite good success with roughly 60% of the patients being cured in that setting. But that conversely means that 4 out of 10 patients still remain uncured and we're convinced that those patients deserve new treatment options, and they deserve these treatment options in a timely fashion. And the reality is that in these frontline settings, therapy goes on for -- well, the treatment effect of therapy goes on for a long time. So focusing on overall survival is really difficult. And therefore, it's important for us to bring forward endpoints like progression-free survival, that are appropriate because they demonstrate meaningful benefit for patients. And in frontline DLBCL, that's clearly the case where preventing patients from progressing, preventing relapse, preventing subsequent therapy that is often costly and certainly psychologically, really detrimental for patients, is something that is very meaningful. And on top of that, as you can appreciate on the right side of the slide, it's also been demonstrated and accepted in the community quite broadly that PFS in this setting is predictive of a treatment effect on overall survival ultimately. So if you go to the next slide, I'm not going to go into the details of the POLARIX study. We have Professor Morschhauser here who is a distinguished expert and President of the LYSA Group, so he will speak about POLARIX to you. But just sort of pointing out that we're really excited about these results. We're excited about them because there is many patients in the first-line setting of diffused large B-cell lymphoma that deserve better, new therapies. And we believe that with the first study that's positive here in almost 20 years and there's been 11 negative studies in the meantime, we're really making a meaningful advance here. On top of that, there are some distinct benefits of Polivy that we're quite excited about. One is that physicians all across the world already have experience with this drug. And second, it's an off-the-shelf and fixed-duration therapy. So it has a lot of advantages also with regards to payers, understanding what does, of course, the therapy cost and being able to appreciate that in the context of preventing progression for patients. And lastly, we know a lot about Polivy, and we, most importantly understand that it has a very well-tolerated safety profile that is well documented in many patients suffering from non-Hodgkin's lymphoma. And so we're also quite confident that this drug can be easily moved to the frontline setting, and we'll see some of the data, obviously, from Professor Morschhauser. But we're not stopping with Polivy. So if we go to the next slide, you will also hear some of the updates from ASH from Charlie Fuchs later on about our bispecifics program. But what I wanted to point out here is really how much as a company we're focusing on developing differentiated and well-designed bispecific antibodies. And we've demonstrated that with Hemlibra in hemophilia. And we've demonstrated it again earlier this year with [ Bovismo ] in ophthalmology, 2 really carefully designed bispecific antibodies. And we're excited to bring this forward in NHL, and we have 2 differentiated molecules, mosunetuzumab and glofitamab. And we believe that because they are different, we'll be able to address very different needs that exist for patients with non-Hodgkin's lymphoma that have different types of diseases, are in different care settings, and we believe these 2 molecules can really make a meaningful benefit for those patients. And in addition, have significant advantages over, for example, CAR-T therapy as they are much more readily available, can be given in a community setting, don't require hospitalization, for example. So a lot that we're excited about here. On the next slide, I just briefly wanted to point out that sort of some of the corner posts of our development strategy for these bispecific antibodies is also to move to the curative setting. So we definitely see opportunities in the frontline setting in non-Hodgkin's lymphoma for these drugs. But also, we see a huge potential to combine. And the other sort of really great thing with bispecific antibodies is their ability to combine with other modalities. And so we're quite excited to explore that. And as you can see from the slide, we have quite a broad program already in place that is exploring that. On the next slide, what I wanted to talk about is another bispecific antibody, cevostamab, which is actually in multiple myeloma. So I said in the beginning that this is an area where we are really looking to venture into. And what's very interesting and unique about cevostamab is that in what is a very crowded field in multiple myeloma, with a lot of BCMA-targeted agents, this bispecific antibody has a very unique target, which is FcRH5, and so we believe that sets it apart nicely. And Charlie will present some of the data that we showed at ASH a little later on, but we are excited about moving this molecule forward. We have a -- we're planning for a broad program, and we think this is going to be a first really important step into multiple myeloma for Roche. On the next slide and to conclude my presentation on the malignant side of hematology, I just wanted to point out how strong the year has been for Venclexta. We're really seeing this as being a standard in CLL and AML, and we're seeing very strong performance in those spaces. And we're also really excited about the future for the molecule as we, in collaboration with our partner, AbbVie here, are thinking to hopefully have 7 new indications for Venclexta in the next 5 years. So a lot more to come here. And also, if we look into next year, specifically, we're excited to be reading out data in multiple myeloma, and we're looking forward to making progress in MDS. So 2 important steps forward for this molecule. If we move to the benign side of hematology, we're obviously very much focused on hemophilia A. And the progress we've been making with Hemlibra this year has really been profound despite the pandemic. We're seeing very strong growth in uptake. And truly in many countries across the world, Hemlibra is now an accepted standard of care. As you know, we have a very broad study program with more than 1,000 patients. And we have about 13,000 active patients on commercial Hemlibra at the moment. So a lot really to be very proud of here. And at ASH, we demonstrated new data from the HAVEN 6 study in the mild to moderate population, which we believe is further going to substantiate the benefit that Hemlibra is bringing to hemophilia A patients. And on the next slide, I wanted to just briefly also give a shout out to our partners at Spark Therapeutics, who is really our hub for gene therapy. And we're very excited about the progress that they are making and we are making in collaboration with them on SPK-8011 in hemophilia A. We do think that gene therapy still has a road to go in hemophilia A. And we definitely believe that the bispecific class and Hemlibra specifically will continue to remain a really big driver in that space. But we also think that there is opportunity for gene therapy and we're very excited about the progress and getting to a place where we can really demonstrate predictability and durability for these gene therapies, which are really going to be the 2 important criteria to watch out for to determine how successful these therapies can be for patients with hemophilia A. And then finally, before I pass it on to Charlie, I wanted to illustrate that beyond hemophilia A, we're also looking at other benign hematologic diseases. And most importantly, we're excited about crovalimab, our anti-C5 monoclonal antibody. This year has really been a terrific year for crovalimab. We've made tremendous progress in the development program for the molecule. And next year, we're really looking forward to reading out the first pivotal data in PNH, which is going to be an exciting milestone. And this year, we were also able to expand the program in atypical hemolytic uremic syndrome, where we've been able to initiate studies. And also, really importantly, in sickle cell disease, where we believe that this molecule can make a big impact. And so we're excited that the clinical program is also moving forward in that indication. So with that brief overview over our heme pipeline, I hope that I'll leave you with some of the same confidence that we have in the future and what we're going to bring to patients with blood disorders. And to focus a bit more specifically now on what we've presented at ASH and how to think about that, it's my distinct pleasure to introduce you to Charlie Fuchs, who is our Global Head of Development in Oncology and Hematology. So over to you, Charlie.

Peter, thank you. And Karl, thank you for the plug for the book. It's very kind of you. As Peter very nicely discussed, this year's ASH meeting saw an unprecedented number of high-profile presentations coming out of our portfolio. And while I won't be able to cover all of those data sets, I would like to share with you some key studies highlighting the work in Hemlibra, our CD20/CD3 bispecifics, cevostamab and Venclexta. I also want to point out that beyond today's speakers, we also have with us today Marion Ott, Global Head of our Myeloid Malignancies and Multiple Myeloma Franchise; Ginna Laport, leader of our Lymphoma and CLL Franchise; and John Pasi, Global Head of our Rare Blood Disorders Franchise. And Marion, Ginna and John will be available to answer your questions during the Q&A session. As Peter outlined, Hemlibra has transformed the treatment landscape for people with severe hemophilia A. Nonetheless, in patients with mild to moderate hemophilia A, factor VIII deficiency may still result in spontaneous bleeding, particularly into joints, resulting in joint damage, pain, impaired mobility and reduced quality of life. As such, beyond severe hemophilia A, mild to moderate disease represents an important unmet need. And at ASH this past week, we presented results from the HAVEN 6 trial evaluating the safety and efficacy of Hemlibra in patients with mild to moderate disease for whom factor VIII prophylaxis is warranted as deemed by the treating physician. In this interim analysis of 71 patients, 72 patients -- 72% had moderate severity. And among all participants, the mean number of bleeds in the past 24 months -- or 24 weeks rather, was 3.4. With respect to the secondary endpoints of the trial, participants on Hemlibra reported an improvement in treatment burden as assessed by the CATCH score and an improvement in mean total hemophilia joint health score. Additionally, on Hemlibra, 96% of patients prefer the Hemlibra over their previous treatment. Turning on the next slide to the primary results. Across the majority of bleeding event categories, the median annual bleeding rate was 0, and all model estimates showed an annualized bleeding rate of 2.3 or less per year with several well below 1. As shown in the bar graph on the right, more than 90% of patients reported having 0 treated joint bleeds. Finally, no new safety signals were identified. There were no thrombotic or microangiopathic events or deaths reported, and there were no adverse events that led to treatment discontinuation, reduction or interruption. We believe that the HAVEN 6 data demonstrate that Hemlibra offers a favorable safety profile and an effective treatment option for patients with mild to moderate hemophilia A, while meaningfully reducing the treatment burden. Turning next to our CD20/CD3 bispecifics. We presented results from our pivotal Phase II trial of mosunetuzumab monotherapy in patients with relapsed/refractory follicular lymphoma. As seen in the baseline characteristics, patients on this trial were heavily pretreated and had refractory disease. Patients have received a median of 3 prior lines of therapy, 53% were double refractory and 53% had experienced progression or relapse within 24 months of diagnosis as characterized by POD24. As the waterfall plot documents, mosun conferred an impressive depth of response, with an 80% overall response rate and a 60% complete response rate by independent review, a complete response rate that is significantly greater than the 14% historic CR rate in this setting with prior therapies. Additionally, on the next slide, the median duration of response was 22.8 months and the median progression-free survival was 17.9 months, but realize these are relatively immature data because follow-up is still ongoing. With respect to tolerability, cytokine release syndrome, CRS of any grade occurred in 44% of patients, which was predominantly grade 1 and 2. Adverse events leading to mosun discontinuation was very uncommon at 2%. Finally, and no less importantly, mosun was routinely administered as an outpatient. So in sum, mosunetuzumab is the first T-cell engaging bispecific antibody to demonstrate clinically meaningful efficacy in patients with relapsed/refractory follicular lymphoma and thereby offers a first-in-class, off-the-shelf outpatient option for these patients that we believe can also be administered in community settings. Turning to the next slide, beyond monotherapy. We are equally optimistic and enthusiastic about the potential for mosun in combinations and in earlier lines of therapy. Specifically, we reported the results of our Phase Ib/II study for the combination of mosun with polatuzumab in patients with relapsed/refractory B-cell lymphoma. Patients on this trial were heavily pretreated, having received a median of 3 prior lines of therapy. Moreover, 40% of patients have received prior CAR-T therapy and 76% were refractory to their last therapy. As shown on the right-hand side of the slide, among all patients with diffuse large cell lymphoma, who received the recommended Phase II dose, the overall response rate was 65%, with a complete response rate of 48%, which was virtually identical in patients who had received prior CAR-T therapy. Follow-up data in this trial are still relatively immature. However, of the 29 patients who achieved complete response, 28 remain in CR and the 1 patient with progressive disease actually received retreatment and reachieved a new complete response. CRS of any grade was reported in 17.5% of patients and was confined to Grade 1 and 2, with only 1 Grade 2 event. Ultimately, we believe that mosun-Pola demonstrates compelling efficacy and very manageable tolerability in this heavily pretreated population. And we are launching a Phase III study of this regimen in the second-line therapy of DLBCL patients. Of note, we also presented data at ASH for the combination of mosun and lenalidomide in relapsed/refractory follicular lymphoma, showing a 90% overall response and a 66% CR rate in that population with very manageable tolerability. In light of those results, we have also started a Phase III trial of mosun/lenalidomide in relapsed/refractory follicular lymphoma. At the same time, turning to the next slide, we reported on a number of studies with glofitamab, including this trial in patients with relapsed/refractory mantle cell lymphoma. As you know, mantle cell is an aggressive subtype of non-Hodgkin's lymphoma. And patients with progressive disease following a BTK inhibitor have a very poor prognosis. In this trial, the majority of patients had previously received the BTK inhibitor and were refractory to prior therapy. Nonetheless, among all patients, the overall response rate was 81%, with a complete response rate of 67%. Although follow-up remains immature, the swimmers plot on the right shows that the vast majority of patients continue to have ongoing responses at the time of the data cutoff. Finally, most CRS events were grade 1 or 2, and there were no adverse events leading to treatment discontinuation. We believe that this fixed duration off-the-shelf therapy of glofit in mantle cell lymphoma supports further investigation as an important new treatment option for patients with this malignancy. Turning next to multiple myeloma, we presented updated results from our Phase I dose escalation trial of cevostamab in relapsed/refractory multiple myeloma, assessing both a single step-up and a double step-up dose schedule in cycle 1 to mitigate CRS. Patients on this trial were heavily pretreated, having received a median of 6 prior lines of therapy and 85% were triple-class refractory. Turning to the results on the next slide. Among patients who received the target dose of 132 milligrams or higher, the objective response rate in this heavily pretreated, highly refractory patient population was 57%, with 33% of patients experiencing VGPR or better. In addition, MRD negativity as assessed by next-generation sequencing was detected in 7 of 10 evaluable patients with VGPR or better. With follow-up ongoing, these responses appear quite durable with a median duration of response now in the range of 11.5 months. 81% of patients experienced CRS, which were predominantly grade 1 or 2 and AEs leading to treatment discontinuation were uncommon. Based on these results, cevostamab appears to be a unique new efficacious treatment option for multi myeloma. This is the only FcRH5-targeted agent, and therefore, beyond monotherapy, offers the potential to combine with other agents currently in practice or in development, including BCMA-targeted therapies. As such, we will be expanding our clinical development program in 2022. Turning next, as Peter described, Venclexta, our Bcl-2 inhibitor has an important impact on the care of patients across the gamut of hematologic malignancies. In higher-risk MDS, myeloblast overexpress anti-apoptotic proteins, including Bcl-2 and are this -- are thus potentially vulnerable to directed therapies like Venclexta and azacitidine that can inhibit these pathways. At ASH this week, we provided updated results of our Phase Ib trial for the combination of Venclexta and azacitidine in treatment-naive, higher-risk MDS. 84% of patients responded to treatment representing a combination of complete response or marrow complete response. The median time to response was just under 1 month and follow up -- with follow-up still maturing, median duration of response is over 1 year. Additionally, clinical and molecular responses were observed across the spectrum of mutational events, including poor prognosis mutations. Building on our investment in Venclexta, we launched a Phase III trial of Ven/aza in higher-risk MDS in late 2020, which offers the promise of meaningfully improving the treatment landscape for MDS patients. In closing, at this year's ASH, we presented a wide range of exciting new results from our portfolio. And my colleagues and I, as I mentioned, will be happy to take your questions later. But first, let me now turn to Professor Franck Morschhauser, a thought leader in the field of lymphoma, an innovator and frankly, a leader of our POLARIX trial to present the results of yet another practice-changing study, the study of POLARIX. Franck, thank you.

Thank you so much, Charlie, for this very kind introduction. On behalf of all the POLARIX investigators, it's my great pleasure to present you the data of this trial. Next slide. So first of all, to set the stage, it should be very important to remind everybody that much of what is in the standard of care in first-line DLBCL for over 20 years, there have been many attempts to improve over R-CHOP with introduction of new drugs, dose intensity, maintenance regimen. But at least 13 big randomized Phase III trial failed to demonstrate any significant improvement. With this standard of care, let's also keep in mind that up to 40% of the patient either will have refractory disease or would relapse following complete remission after R-CHOP. And this clearly underlines the unmet need in this population. Next slide. CD79b is a component of the BCR receptor and polatuzumab vedotin is an antibody drug conjugate composed of an anti-CD79b monoclonal antibody coupled to an MMAE potent microtubule inhibitor that is internalized upon antibody binding to make this effect. Single-agent data with polatuzumab vedotin were promising with overall response rate of 52%. And then pola was incorporated with CHOP and R-CH, the pola combination, that looked extremely promising in the Phase Ib setting, with a 77% on peak response rate and a promising 18 months PFS that led to think of a randomized Phase III trial to compare with the standard of care. Next slide. And this was the purpose of the POLARIX study, which was a randomized, double-blind study, placebo-controlled international Phase III randomized trial. Patients were eligible if they had previously untreated diffuse large B-cell lymphoma, who are aged 18 to 80 years, had an IPI score of 2 to 5, that means incorporating intermediate and high-risk patients according to IPI and a reasonable PS of 0 to 2. Patient were randomly assigned in a 1:1 ratio to either Pola-R-CHP or R-CHOP given for up to 6 cycles plus 2 additional cycles of Rituxan. Patients were stratified according to their IPI score 2 versus 3 to 5, bulky disease less than 7.5 or more than that, and the geographic region, Western Europe, U.S., Canada and Australia versus Asia versus rest of the world. Next slide. The primary endpoint was the investigator-assessed progression-free survival, and the secondary endpoint examined hierarchically were event-free survival, complete response rate at the end of treatment according to the central review and overall survival. We also had an additional sensitivity analysis on disease-free survival as assessed by the investigator. Other key secondary endpoint are the safety endpoint with the incidence, nature and severity of adverse events. 875 patients were enrolled in this trial and would have this first analysis and primary analysis. There was a need to have at least 228 PFS events, with a median follow-up of at least 2 years for all the patients included. And at the time of this primary analysis, the median follow-up was 28.2 months. Next slide. These are the baseline characteristics of the patient. You can see here that we are well balanced between the 2 groups. The median age was 65, close to 2/3 of the patients had a high-risk IPI score of 3 to 5. Most patients had advanced disease according to Ann Arbor staging and based on characteristics but also well balanced in terms of the biologic subgroups, CEO classifications, double/triple hit or double expressions. Next slide. With the median follow-up of 28 months, the risk of progression, relapse or death was significantly lower in the Pola-R-CHP group than in the R-CHOP group, with a stratified hazard ratio of 0.73. That means that the milestone analysis showed that the percentage of patients surviving without progression at 2 years was 6.5% higher with R-CH and Pola than with R-CHOP, meaning a 27% reduction in the relative risk of disease progression, relapse or death versus R-CHOP. Next slide. Similarly, the event-free survival was significantly better for R-CHOP, with a hazard ratio of 0.75. Remember that event-free survival also included beyond progression, death and relapse, and then introduction of another treatment that was not recommending the protocol, so subsequent treatment or a positive biopsy at the end of Q2. So a very meaningful result. Next slide. The CR rate was 78% with a R-CHP Pola versus 74% with R-CHOP. The difference was not statistically significant. Still, the disease-free survival as assessed by investigator was significantly better with a hazard ratio of 0.7, probably meaning that the quality of the CR achieved with R-CHP Pola was better than with R-CHOP, leading to a better disease control. Next slide. With 28 months median follow-up, that means a relatively short median follow-up in terms of overall survival, there was no difference at this point with a hazard ratio of 0.94. It is important in this context of so far, lack of difference in overall survival to see that the percentage of patients receiving a subsequent treatment for disease progression or relapse was always higher in the R-CHOP group, 30% versus 22% for any treatment, 13% versus 9.3% for predetermined or own preplanned radiotherapy, 23.5% versus 17% for any kind of systemic therapy, 7.1% versus 4% for autologous stem cell transplantation and 3.6% versus 2% for CAR-T. Next slide. The safety profile was pretty similar between the 2 groups in terms of the overall incidence of adverse events or the percentage of Grade 3, 4 or Grade 5 for serious adverse events. What's interesting to note is that adverse events leading to either discontinuation of any drug, especially polatuzumab vedotin or vincristine or dose reduction of any study drug was numerically lower in the R-CHP Pola arm. Next slide. On this tornado plot, we see that there is a kind of mirror showing that the safety profile was really similar overall. What slightly stands out with a higher incidence of diarrhea with R-CHP Pola and a slightly higher incidence of febrile neutropenia, 13% versus 8%. But please keep in mind that this is well within the range of what has been observed in other trial. Remember that, for example, in the GOYA trial, which is the most recent one with all R-CHP Pola, the upper incidence of febrile neutropenia was 15%. So nothing to us deemed as clinician -- clinically relevant. Next slide. So in conclusion, I think we can say that Pola-R-CHP significantly prolongs PFS compared to R-CHOP. Hazard ratio was statistically significant, and the hypothesis preplan was reached. In patients with intermediate and high-risk, previously untreated diffuse large B-cell lymphoma, very clearly, the safety profiles of R-CHP Pola and R-CHOP were comparable. We are now doing virtual analysis to look at various subgroups, but this is more hypothesis testing. And clearly, the product results support the use of Pola-R-CHP in the initial management of patient with issues on B-cell lymphoma. I'm happy to take a few questions. This was my last slide, and I turn over again to Karl.

Here we go. So thanks a lot for all the presenters, maybe a bit of a housekeeping. We have about 200 people on the call, which is a lot, I have to say. So thanks for your interest in Roche and the presentation today. Before we go into the Q&A, I already wanted to thank the speakers for their contribution today. I wanted to also to thank [ Lauren ] from our U.S. team, who did the heavy lifting on all the slides and the storyline development. So thanks a lot, [ Lauren ]. I wanted to thank [ Sonia ] already for all the organizational things to set up the call. We have about 45 minutes for Q&A now. And today, we have to be in time, I have to say, because I know that Charlie and the whole hematology team, they have a Christmas party today, a virtual Christmas party. And he asked me really to be in time here to close because they are -- they will be the key speakers. So all of them, for this party, we don't want to intervene here.

So what we can do here, I will open the questions now. We have about 7 questions now in the line, and we have about 6 here on the chat, and I will try to mix it a bit in the sense of clustering the questions together and mention also who asked the question. So Simon Baker would be the first one. [Operator Instructions] So Simon, I'll open the line.

Thank you for the presentation. I will stick to two. Starting with POLARIX, Professor Morschhauser, you did say that the subgroup analysis was exploratory, but I'll ask anyway, if there are any analysis that you can share at this stage in terms of performance by subgroup. And also one thing that I noticed was the -- going back to the performance of R-CHOP in GOYA and comparing it with POLARIX, it's remarkably similar. And over time, we often see standard -- the performance of standard of care improve, but it stayed very, very constant there. I just wonder if there are any reasons for that. Obviously, it makes the trial comparisons easier, but it's slightly unusual. And then moving on to Hemlibra, Roche is probably unique in being able to give an unbiased answer on gene therapy versus existing therapies because you've got a foot in both camps. So I wonder if you could give your thoughts on what gene therapy needs to demonstrate to displace or compete with existing products like Hemlibra. And also just a quick one on the dose frequency, is there any clinical or commercial justification for pushing beyond 4-weekly dosing?

Yes, maybe I can just add here because we had some questions here in the chat for you, Franck. What -- how important are these data for you as a clinician? Let's say, is it practice changing for you, the POLARIX data that was just really down here from -- so maybe if you can just put that into your answer for the first question.

Yes. Yes. So very clearly, we have been dying for a positive trial in the last 20 years. As I said, 13 big trials that were well designed with new drugs, dose intensity, whatever, all failed. So for the first time, after 20 years, after R-CHOP data, we have a positive trial with a significant improvement in terms of progression-free survival. And to me, this is extremely meaningful. And that without any costs in terms of additional toxicity, similar safety. So I think this is the best answer that can be said to tell you how important is this data margin. Now to answer the two questions of Simon. The first one, the subgroup analysis. I really need to remind you that the study was neither designed nor [ forward ] to look for this subgroup analysis and significance at this point. So when you look at those estimates, it's really to generate some hypothesis and see what the future can tell, especially the biology subgroups, for example. And so far, no conclusions can be drawn from this data. And the take-home message is that the POLARIX data applied to the strict eligibility criteria, IPI 2 to 5, and no subgroup analysis should be done to select a group of patients that should only get R-CHP Pola. Regarding your second question, we were surprised to see that the standard arm was doing so well, of course. As you said, the more we go with clinical trials, the more we get the impression that the standard of care is always doing better. We have had this impression also with ABC, you may remember. The ABC patient was supposed to do so poorly. And when you look in the current -- in the last trial, the ABC patients treated with R-CHOP were at 15% higher. So in this context of such a high bar with the standard of care, it's even more impressive to see a benefit in terms of progression-free survival.

Yes. Thank you, Franck. Maybe for the gene therapy, Charlie, could you maybe just give it a start, yes?

Simon, thank you for that question. And as I mentioned earlier and Peter did, we really think Hemlibra sets a high bar in terms of transforming the therapeutic landscape. But we're obviously excited about the data from Spark and other gene therapies. And let me introduce John Pasi, who leads our rare blood disorder franchise, just to offer some additional insights. John?

Yes. Thank you, Charlie. I think the first thing for me to say is just to reiterate what Charlie said that Hemlibra has set an incredibly high bar in terms of how we look at hemophilia and severe hemophilia now. And that needs to be set in context with the huge excitement that there is within the community about the possibility of gene therapy. And I think it comes to the question that was asked about what do we need to show. So in terms of what we've got, we are in a very fortunate position in Roche having Spark as a member of the team, complementing our existing abilities to expand into gene therapy. There are still unmet needs for patients. The bar has been set very high with Hemlibra and that is absolutely true. But coming to your question, what do we actually need, I think if you look at anybody's presentations about gene therapy, whether it's from industry or whether it's from experts in the field, there are 4 words that keep coming through: reliability, variability, durability and safety. And at the moment, we've got a lot of data from Phase I studies. We need to see Phase III data across the board to decide how best we can take these forward. But it's a really exciting time in terms of opportunities for what we have. The other question, I think, you mentioned was about 4-weekly or greater dosing for Hemlibra, is there a need for it. I think patients will always look to see that hemophilia has got a treatment burden. And there is a psychological burden of always having to treat. So there's always a greater need to expand and increase, if that is possible in any way. So I think there is always a view that this is something we're going to be looking to do. Hemophilia is very exciting at the moment. Lots of opportunities, but there are still some questions out there across the board.

Yes. Thanks a lot, Simon, for the questions. Next one would be Emmanuel Papadakis from Deutsche Bank.

And Karl, since I know this will be one of your last events, perhaps I can take the opportunity to thank you for all your collaboration and support over the years and wish you the best for the future. So perhaps I'll take a follow-up on POLARIX please, just on the overall survival data. And then perhaps a question on the FcRH5 asset. So you highlighted the correlation, PFS and OS in DLBCL at the onset of the presentation. How confident are you that this PFS benefit will actually translate into an OS benefit over time? Have you got any comments on the early OS inversion we saw in those 2 curves? And one for Franck, in particular, how important do you think OS is in driving clinical utilization broadly across the globe? And I'm thinking particularly outside the U.S. here? And then the question on the FcRH5 would be, I mean, interesting Phase I update. It seems to have moved, Marion, I remembered us discussing this a year ago when we were optimistic that perhaps you could move fairly quickly into a pivotal setting, but we're still at Phase I. The data relative to the GPRC5D update we've seen at this ASH again seems perhaps slightly inferior. And also, that asset has moved somewhat faster in terms of development plans. So I would love to hear your perspective on how those 2 targets stack up relative to each other and what exactly your plans to expand the clinical development program in '22 will look like. Will that include a BCMA combination, for example?

We had -- basically, just to close the line here, similar question from Marietta Miemietz, just on the first one, on importance of PFS, OS separation of curves, how important it is for a physician to use those data? Will it be the future standard of care? Just to close the loop here.

So again, it was extremely important for -- to have a positive trend with an improvement in PFS after 20 years of failure. This is the first step. Without this first step, nothing is possible. The second is that the median follow-up is still short. But regarding the possible translation of this PFS benefiting an OS, I think we should be very reasonable in terms of our expectations with OS now. This has been the dogma before R-CHOP. After R-CHOP, we have many trials that showed a PFS improvement without any OS benefit. So let's have reasonable expectations here. That does not mean that with the longer follow-up, we won't have something. But if we do not have something, this does not diminish at all to meet the value of the state. I just would like to point out that in this ASH meeting, we had 3 randomized trial in the relapsed/refractory setting for diffuse large B-cell lymphoma, CAR-T versus autologous stem cell transplantation, 2 were positive. And for these 2, 1 with a reasonable median follow-up, hazard ratio was 0.36. And in these 2 trials with such a hazard ratio and a PFS or EFS benefit, there was no significant OS benefit at this time. So this shows you that things have changed, the dogma might not be as true as before. And that with these new therapies, you have to have a new thinking on this translation. So to me, this PFS benefit is this very strong method, and this would not be diminished by a lack of OS benefit in the future, especially looking at the incidence of subsequent treatment in both arm and the lower percentage of patients getting a subsequent treatment in R-CHP Pola. This is a more stronger argument to me.

Thank you, Franck. Multiple myeloma, Marion, maybe if you can -- it's one for you.

Yes. Thank you for that question. First, I want to mention that for FcRH5 is we have here the potential that it's really like the first-in-class. In terms of the target, what we know so far is that we don't have any soluble forms so we don't have any sheddings. And what is important to us too, it was cheaper as C5 being mentioned, we don't have any really clear on-target toxicities for FcRH5. For the dose, we are currently -- I would like to highlight that we are still in dose escalation. It is important to mention that when we further increase the dose, we do see higher responses, but we can also maintain the safety profile and here specifically, the CRS. So the dose optimization is currently absolutely key. So that's why I really cannot comment at the moment on the exact sequencing. We are interested to go very much into earlier lines, especially with standard of care combinations. We're interested to combine it BCMA. I also want to highlight that we really have encouraging activity in heavily pretreated patients, specifically post BCMA, post CAR-T and also post ADC. So this is also potentially positioning additionally here as well. We have internally also an interesting early pipeline as well where we're also internally combined with novel-novel combinations as well.

Thank you, Emmanuel, for the flowers. And thanks for all your continued support. I was really privileged working with all of you. I have to say. So the next one is from an anonymous attendee, but it's going to you, Ginna. It's in light of the POLARIX data. Will Roche move forward, glofitamab or mosunetuzumab, into first-line DLBCL? It's a combination. How do you see the future standard of care developing?

Yes. Thank you, Karl. I see the future standard of care developing with POLARIX and beyond POLARIX. We are exploring both glofitamab and mosunetuzumab in the first-line setting with other combinations with polatuzumab, our antibody drug conjugate, but also with standard chemotherapy.

Okay. So we keep it broad then?

Yes. Those results will be presented at future meetings.

Okay. Good. So more to talk about in the future. Next one, I'll take from the line here. Richard Vosser. I open your line, Richard from JPMorgan.

A couple of questions, please. Just to go back, I realize you mentioned we shouldn't look at subgroup analyses. But I think the presenter at ASH was asked a question about high-risk patients and the idea that they may do better than the POLARIX trial. And I think he's -- the suggestion was that they would do better. So just -- could you just expand on that a little bit and whether that is in the data? And then second question on mosun, data looks really strong in follicular lymphoma. Just thinking about the subcutaneous version seems to be going relatively slowly, prioritizing IV at the moment. Just thoughts on that, given the competition is coming with subcutaneous versions as well and staying ahead here.

Yes. When we look at the forest plot for the different estimates in the predefined subgroup, it looks like the benefit was higher for IPI 3-5 versus IPI 2. And this was the sense of the answer of [ RVTE ] when he was asked the question. But that does not mean to me that there is no benefit for the IPI 2 patients since the study should be taken as a whole. Now high-risk group might be seen in many other different ways. And this is why we are now performing other studies to look at the ctDNA levels to look at the metabolic tumor volume. Maybe this is more meaningful to try to segregate patients that will drive the highest benefit from the drugs. But so far, it's really in investigation.

Yes, I think we had some questions also here on the chat, if you should do now focus future trials only on certain subgroups and neglect the overall picture. But I guess you gave the answer, no, we shouldn't. And we should really take an exploratory analysis, an exploratory analysis in the totality of the data, the totality of the data and take it as a future standard of care. So subcut, maybe one for you, Peter?

Yes. Thanks a lot, Karl. And great question, right? So I think across the board, we're seeing subcut used for bispecifics to manage CRS. We're actually really happy with the profile that mosunetuzumab is showing at the moment. And we believe that hospitalization is a key factor, and we don't require mandatory hospitalization, which we believe will really enable practitioners across the board to use mosunetuzumab at launch. And so we are looking into subcut, and that is certainly something that we're curious to see the data. But we're also actually quite confident that what we have in terms of the profile is going to be very interesting. And in a space where infused medicines are used very regularly, so we don't really think that subcut, yes or no, is going to be a big determinator on the success of the drug.

And maybe you can -- one from the chat here is asking, can you provide more details around the ability of mosunetuzumab to be utilized in your community setting? I mean it goes a bit longer, the same question, maybe you can also address this one?

Yes. No, thanks, Karl. I mean, look, everybody is very excited about CAR-Ts, right? And it's amazing technology and certainly something we're also interested in. But what we're seeing even after years now is that it remains very limited in terms of the ability of physicians to give it to patients, and there is many patients with non-Hodgkin's lymphoma that don't have access, right? And I think what we're very excited about, in particular, with mosunetuzumab is that a drug that's infused can be given in any sort of practice setting, doesn't require mandatory hospitalization, really has a profile that is going to be quite broadly applicable and hence, be able to address many patients with non-Hodgkin's lymphoma and certainly follicular lymphoma, which is going to be our launch indication.

Yes. Thank you. Thanks for the question, Richard. And I hope we could address it. Luisa Hector would be the next one.

Thank you to everyone for the presentation. So still on POLARIX, I wonder when we might see overall survival data that could be more meaningful. And I was intrigued by Dr. Morschhauser's statement on lack of OS benefit would not diminish the result. Just -- that's great. I'm just wondering what the regulators might feel about that. Can we be confident that the regulators would have the same view on that OS endpoint? And maybe I can have a quick question on the bispecifics. Maybe just a reminder about how -- a little bit more about how you're differentiating those, because I can see now it does look like in second-line DLBCL, you've got both mosun and glofitamab going into Phase III there. So how do you see those 2 being positioned? I know that the combinations are different, Polivy and then the GemOx, but how do you see those being used by different patient groups, perhaps?

Maybe I can add one here, one for you, Peter, most likely, it's from [ Sakata Roy ]. He was asking about the uptake of those data, the POLARIX data in the market. So would you expect, let's say, more a faster uptake, a slow uptake based on the subgroups? And how would you say -- I mean, I think the question is going to both to Franck and to Peter, to you.

So maybe I shall start answering Luisa's questions. I think the next analysis for looking at the PFS and OS data will be in 1 year from now. But we have discussed within the Steering Committee the possibility of having a much longer follow-up and to generate those data to eventually see an overall survival benefit. What the regulator are going to think of that, they live on the old dogma and rules. And they probably have to get the full picture with the reports of the current physicians and their impression of the data to best interpret this PFS benefit and the lack of OS benefit so far. Just the way I did it and other of my colleagues are thinking to really reconsider whether a lack of OS benefit would really affect their appreciation of this data, but I cannot answer for them.

Thanks a lot -- please, Charlie, sorry.

Yes. I would just -- Franck, thank you. I would just add, Luisa, that as far as the Food and Drug Administration is concerned, progression-free survival is an established approvable endpoint in the first-line therapy for large cell lymphoma. And moreover, studies of many thousands of patients in first-line large cell lymphoma showed that PFS is an effective, valid proxy of OS. As I think Franck indicates, the field has changed over the past 2 decades for the better, including the availability of salvage therapies. So even with recurrence, the vast majority of patients fortunately are alive at 2 years. And so I think as Franck points out, probably considerably longer follow-up is going to be needed for OS in this study, recognizing that regulators do accept PFS. Peter, I'm sorry, I didn't mean to interrupt you.

No. No. Thanks. I think great points, Charlie. And Luisa, maybe building on sort of more the uptake questions, right? I mean, what we saw in POLARIX is a very meaningful improvement in progression-free survival. And that is the primary endpoint of the study. So this is essentially exactly what we had planned for, right? And at this point, we never expected to see overall survival, right? As Charlie pointed out, at this point, the vast majority of patients who have been part of the study are still alive, right? There are no OS events to report. So it's not surprising that we're not detecting a difference, regardless of whether Polivy would or would not have an impact on that. And so it's been part of our launch plan from the very beginning to have this type of very meaningful PFS data. And when we've explored that with physicians all across the world, to Franck's earlier point, after 20 years of not being able to move the needle, we do expect that this will be used and taken. And frankly, I think patients don't want to be part of the 40%, right? I mean if you're a frontline DLBCL patient, then you know that there's a hope for cure and you have a 60% to 40% probability, right, you want to do everything to push your luck and also avoid that psychological trauma of having to go through multiple relapses, thinking about subsequent therapies. And from a health care system perspective, we're actually also really quite confident that preventing future therapy is going to be very meaningful in terms of the cost impact.

Yes. For the bispecific one, you're on mute.

Yes. Okay. Yes, I'll take that one. Thank you. So for the bispecifics and about our plans for second-line DLBCL, we do have a Phase III trial of Polivy and GemOx versus Rituxan and GemOx. We also have a Phase III trial of glofit-GemOx versus Rituxan and GemOx. To start with the glofit plus GemOx, that is a trial that's ongoing to combine, obviously, glofitamab with chemotherapy because early results show encouraging efficacy and safety. But Polivy and GemOx, we're also exploring that combination because we understand that across globally, access -- there are access issues. Not all countries will have access to both bispecifics and/or Polivy. I should also mention that Polivy-GemOx Phase III trial is a post-marketing requirement to give full approval to Polivy plus bendamustine-Rituxan in the second-line and third-line setting. One last comment, too. We're also combining mosunetuzumab with polatuzumab as well in the relapsed/refractory DLBCL setting. That trial has not started yet, however.

Yes, we get a lot of questions on potential combinations. I can imagine it's not only because of Roche but also because of competitors on how to combine maybe the PI3 kinase inhibitors, the immuno-oncology assets and so on. Would you like to give a kind of a more general view on how you see the development? I mean, you have already answered a lot of the questions, but I guess your answer is that you really want to keep it as broad and open as possible. But the question is, how do you want to position this POLARIX data in the future? Will the things be now compared to POLARIX in the first-line DLBCL? Or will they be used in combination? I mean, maybe this is one of the questions which we have here from the chat. How do you want to position our POLARIX data? Is it -- will be standard of care and they have to combine it? Or is it in combination always with POLARIX now in first line?

Well, we hope that POLARIX will become the standard of care for the population study as people know that it was for the IPI 2 to 5 population. We also are planning to launch other trials in the first-line setting with POLARIX, with a POLARIX backbone and perhaps add some of our bispecifics to this. Looking at specific populations, I think Franck mentioned that we plan to look at a segment, what we call ctDNA positive, the patients with high molecular loads, as possibly a prognostic tool in addition or in place of IPI. So we're looking at -- in the first-line setting, we're looking at other subsets, too. We're looking at the elderly population. We have -- currently, we have a study of mosunetuzumab plus polatuzumab in the elderly frail population. That's the population who perhaps cannot tolerate chemotherapy. So we're looking at different populations of that large first-line DLBCL entire population.

Yes. Thank you. Luisa, I hope we could address your questions. I open the next one for James. I open your line, James Quigley.

Just a couple of questions. So one on POLARIX and then one on myeloma there. I don't want to labor the subtype -- subgroup analysis point too much. But in the forest plot, another population that did very, very well was the ABC subtype population. So with the cell of origin testing on diagnosis, is that common? Or is that only in sort of specific circumstances? Just want to get an idea of if you will know -- for patients, if there are ABC subtypes upfront or not. And then on the -- again, on the subgroup analyses, there wasn't an OS curve for sort of the populations of that data. So in the older populations in the ABC subgroups and the IPI 2 to 5, if you do a Kaplan-Meier OS curve, is it starting to separate already? Or is it very much similar to the whole population? And then in myeloma for cevostamab, is the potential combination with BCMA is a key consideration cytokine release syndrome as we've seen with both bispecifics? And then a quick one on myeloma. You need to factor in the t(11:14). Is that something you're going to pursue given the data we've seen so far, just because I think Peter mentioned that cevostamab will be the first to move into multiple myeloma?

POLARIX, one for you again, Franck.

Yes. So as you may know, there has been a lot of debate on cell of origin. And again, after probably 10 to 15 years trying to use that to classify and to design trials, this has been now slowly abandoned. Still, everybody or older physicians or the clinical scientists always look at the subgroup analysis still incorporating cell of origin, just because it's easy to compare with the other trials. Polatuzumab is CEO agnostic, molecularly agnostic. It should not be impacted by cell of origin. And still, we get the impression that in this subgroup, there is a clear benefit. We have not looked at OS according to that. But another point I would like to underline is that you probably heard that at this conference, there was a kind of humbling on the molecular classification in diffuse large B-cell lymphoma. Everybody realized that it's still a very long way to go to get a valid subclassification that could lead to stratification in the first-line setting at the time of randomization. Again, so far, it's just an element of different comparison, and it cannot be an argument to make a choice to drive the therapy.

Thank you. Multiple myeloma, Marion, maybe we are going to you?

Yes. A short answer to the BCMA combination, this is something we're really very interested in, and we are here close to, let's say, we have collaborations with external partners to start this combination soon. And for the venetoclax 11:14 population, this is -- thank you for the question. So we -- this is -- we really have exciting data for venetoclax, and this is also absolutely what we want to combine with and where preclinical data are looking really very promising and we were also planning to do so.

Okay. Yes. Thanks a lot. Did we address all your questions, James? And the next one would be Peter Welford from Jefferies.

A few left. So just on POLARIX. I'm sorry, this is coming back to the subgroups again. But this is slightly different, I guess. But I'm curious, if you look at the over 60s and also you look at those without bulky disease, which I guess are 2 fairly sort of unequivocal classifications, if you like, for people. It does look as though those 2 groups basically drove all the benefits. But I guess, rather than that restricting the drug, I guess, I'm more curious, is there some medical explanation do you think as to why those with no bulky disease or equally those that were over 60 could drive most benefit with Polivy? And I guess I'm thinking in particular, the toxicity of the various different profiles, have you looked at, for example, did you get more withdrawals or something due to the side effects and maybe the role of vincristine or something? Or is there something maybe that can explain that? Or is it just do you think -- because they're quite big groups. Is this just something you think that's just sort of an oddity in data? The next one then is just on the bispecifics for CD20. Firstly, just on mosunetuzumab. I think it was a comment that was made because the team have administered as outpatient treatment. Curious if you can just explain, is -- will the first dose or the dose of escalation, will that always be, do you think, given as an in-hospital administration? Is this -- is it really after the first dose that we should then think about this as outpatient for treatment? Or can this be used outpatient from the very first dose of therapy is what I'm asking? And if you could just then explain as well, you made the comment about subcut versus IV. I guess part of the argument, you're right, because in terms of the profile, it's interesting, obviously, from a financial benefit. But also, I guess, from a timing perspective, can you just talk a little bit about some of what was made about this for Herceptin and Rituxan, the length of time it would take to infuse subcuts versus IV to give us some sort of idea of the potential duration and time we're talking in the chair with these respective therapies?

POLARIX, again, Franck.

So regarding the subgroups, one point that is important is that the percentage of patients with ABC increases with age. So there is a mix in terms -- I'm not surprised to see that there is a benefit over 60 and also a benefit over ABC. And this, we need to dig a little bit more. As I said, ABC by itself is not enough. We have to do more molecular classifications and use, for example, the new classifications that have been published by [ Capri ] to test that. And even that, we're still -- it needs to be further refined to try to find a scientific explanation for this possible benefit in those subgroups. Whether there was a difference in terms of toxicity with vincristine in elderly patient, then it comes subgroup of subgroups of patients. It might be interesting to have a look at that, and we have not done that so far. But I have no data to provide on that side, but this is certainly something we should look at.

And Peter, to your question about mosun, so the bottom line is that all doses of mosun are administered as an outpatient. There is no hospitalization required nor was there any hospitalization required in our study. So it's 100% outpatient. With regard to the subcu, before I turn to Peter, I'll just tell you, from an oncologist standpoint, no question. If you can reduce chair time, that's something that cancer centers and patients and nurses welcome. So there is that advantage. But Peter, did you want to add?

Yes. No, Charlie, I completely agree, right? And Peter, as you nicely point out, we have a long legacy of exploring subcut formulations, and we are certainly doing that for the bispecifics. I guess what I tried to say earlier is because -- the question was asked, obviously, in a competitive context. And from a time perspective, right, I'd much rather get an IV and don't have to be hospitalized than getting subcut and still having to have a mandatory hospitalization, which takes much more time and much more resources. So I think that's a little bit of comparison we're looking at, at the moment, and that's why we're quite confident that we have a good profile here with mosunetuzumab despite the fact that we won't have subcut in the very beginning. But on the long run, right, we may even be able to build on top of that if the data for subcut point pans out nicely.

There was, I think, one question also on the regulatory status filing question. So as always, we will keep you updated as soon as we have heard back from an official, let's say, a letter from the health authorities. So we are in active good positive discussions with all of them. I just can really say that. And usually, we do updates then at the full year, so you should be in a position -- when we update, we're in a position to have a bit more of confirmed, let's say, letters from health authorities at that point in time. So this is all progressing well. Keyur, you have the privilege to have the last question here on the line. So I open your line, Keyur from Goldman Sachs.

And I will not bore anybody with further subgroup questions on POLARIX. So let me take kind of 2 separate ones. The first one for you, Dr. Pasi, and good to see you on the corporate side after so many years in academia. But you've very eloquently in the past spoken about the need for therapy beyond kind of Hemlibra. So as you think about the Roche portfolio beyond Hemlibra, can you kind of give us a sense for what is exciting you? And as you think about kind of what that hemophilia treatment landscape looks like over the next 3 to 5 years, kind of how do you see that developing? And what role the early-stage Roche compounds might play on that? And then secondly, on crovalimab. Maybe I'm misunderstanding the slides, on Slide #18, but can you just help us understand kind of what the primary endpoint on the Phase III is and kind of given if it is indeed kind of noninferiority to Soliris, kind of given how established Soliris is, given kind of Ultomiris, is this a good enough clinical outcome at this point of time? Or do you actually need to show superiority for commercial success?

Okay. Thank you. Thank you very much. So yes, beyond Hemlibra, well, clearly, we're looking at all sorts of things because there is still significant unmet need in the hemophilia world. Hemlibra has changed the game absolutely dramatically. And as you point out, that doesn't stop us wanting to look forward to see what else we can do. So there is a lot of fundamental thinking going on, researching into molecular mechanisms to see if there are other modalities that may be applicable in the future. Obviously, we talked about gene therapy already, and this is something that we're very keen to explore with our partners in Spark. I think it -- we've talked about how Hemlibra has changed things, and undoubtedly, it has. But it has also changed patients' expectations to a large degree because they really just strive for more because they've seen such a huge improvement in what they can have. So it is sort of beholden to all of us to continue to look, to continue to see what more we can do, because we have taken a big step. We've taken a huge step with Hemlibra, but that doesn't necessarily mean that's the last step that we need to look for. And so we are looking at different things at a preclinical level, research, understanding all those sort of components to drive things forward to improve medicines for patients as they come through. In terms of crovalimab, yes, we are looking at non-inferiority at the moment. What we can also say though is that for patients who are able to receive crovalimab, they will get a big reduction in treatment burden because this is going to be a subcutaneous therapy. So I think we've got to also weigh that into the bargain as well.

Very good. So I would say that was it. Thanks again to the speakers. I think we have addressed also all the questions here in the chat. Thank you, Franck, Peter, Charles, Ginna, Marion, John. Thanks for all the help over the past years. I wish you all the best. And stay safe and healthy, and wish you all the best and good luck then. Bye-bye. Have a good festivals -- Christmas festivals for you, Charles. All the best to you.

Thank you. Thank you for your leadership, Karl.

Thank you, too. Bye-bye.

Sure. My pleasure. Bye-bye. All the best. Bye-bye.