Biogen Inc. (BIIB) Earnings Call Transcript & Summary

Great. Thank you very much, everybody. It's my pleasure to be hosting Priya Singhal, Biogen's Interim Head of Research and Development, this morning. We'll have a lot to talk about as it relates to Alzheimer's, neuropsychiatry, antisense oligonucleotides and anything else Priya thinks that I'm overlooking here that definitely needs more attention.

But maybe we can start with conversation around Alzheimer's, Priya. And I assume there's not a ton you can say, but I'd be remiss at not at least asking for an update on where things stand ahead of the final NCD and if you think there's any sort of plausible path to materially modifying things as we get the final decision?

Thank you, Paul. It's a pleasure to be here with you and Stifel. I will just start off by saying that before we begin, I'd like to point out that I will be making forward-looking statements. They're going to be based on our current beliefs and expectations. These expectations and statements are subject to certain risks and uncertainties. I encourage you to consult the risk factors discussed in our SEC filings for additional detail. Today, I'm also going to be discussing ADUHELM, approved last year by the FDA for treatment of Alzheimer's disease under the accelerated approval pathway, and please see the full prescribing information and patient medication guide including warnings and precautions at aduhelm.com. Thank you for that question, Paul. So I can't really speculate on what CMS will ultimately decide to do with the final NCD that is expected next month. But maybe I can walk through a few of the critical points here as I see them. What I can say is that the concerns highlighted by the draft NCD are also key areas for our focus at Biogen. These include the general additional clinical benefit on anti-amyloid therapies, generating this, providing additional information on the associated risks, including ARIA, and generating additional data on safety and benefit risk for anti-amyloid therapies in patients from underrepresented communities. They're all very critical areas for us. And on behalf of Biogen, I can say that we remain very focused on generating this additional data to characterize the safety and efficacy profile of ADUHELM and then communicating it in the appropriate forum, as you may have evidenced by our recent publication of the Phase III aducanumab data. So we are fortunate to have a really large database, more than 3,000 patients. And recently, at the AD/PD conference, we presented new data from the Phase III long-term extension study which shows that aducanumab treatment continues to reduce the 2 key Alzheimer's disease pathologies, which is the amyloid-beta plaques and also the plasma tau p-181, in patients for up to 2.5 years -- treated for up to 2.5 years. So this shows that continuing dosing of aducanumab, you see continued reduction in both of the core pathologies for Alzheimer's disease. And I think that we also presented this data showing that clinical decline was reduced in participants who had the plasma p-tau reduction at 78 weeks. So taken together, we feel that all the data that we've shown further informs the scientific evidence for amyloid as a surrogate biomarker and the importance of continuing treatment. Importantly, I think what I really want to highlight here is that we have, beyond our Phase III studies, 2 clinical trials. So we have the EMBARK re-dosing study, and we have the ENVISION study, which is our confirmatory trial, where we expect to initiate screening in May, so shortly from now. And beyond that, we have a multipronged approach to generate real-world data. All of this -- I mean, we have 3 opportunities here that we are taking upon ourselves to generate the real-world data. ICARE AD registry, which is up to 6,000 participants and at least 16% of U.S. participants from black and Hispanic populations. Second, we are also initiating a clinical data research network, or CDRM, more than 50,000 patients. This will aim to collect a wide range of data on both individuals treated with anti-amyloid antibodies but also untreated patients. So it will be the opportunity to provide additional data on treatment response versus standard of care. And it's modeled on other approaches where we have a lot of experience, like MS PATHS and expertise in the MS area. Finally, we'll also be working on Medicare claims analysis. The main aspect here really is that all these approaches are viable only if they are paired with full coverage for the FDA-approved agents in the class, so that patients in the real-world setting actually have access to these agents. So we are -- at this point, I'll just say that we are respectfully urging the CMS to align Medicare coverage for the class of anti-amyloid directed therapies with the FDA-labeled indications for these products, consistent with the population studied in the registration trials but also guided by expert recommendations for clinical practice. I hope that addresses the question.

Yes. We'll see what happens. In the meantime, I guess, do you want to kind of just set up the conversation on lecanemab by saying -- by sort of outlining what are the major differences between this compound and what are the major similarities between this and aducanumab?

Sure. So first of all, we are very excited that lecanemab is going to have a readout in this fall of 2022. The Clarity study will read out. We are very excited that we are potentially able to provide 2 of the 4 anti-amyloid therapies, either approved or in late-stage development. And we really are looking forward to providing these options to patients and people suffering from Alzheimer's disease. So as you know, we are developing lecanemab with Eisai, our collaboration partner here. I'll first just speak a little bit about the similarities and then talk about some of the differences between the 2 therapies, lecanemab and aducanumab. So they have several similarities. Number one, they are both quite specific for binding aggregated forms of amyloid beta. They have full effective function, important for inducing amyloid clearance by microglia. We do believe we have data demonstrating robust removal of amyloid plaque in the brain in humans with both therapies, and we have evidence of potential slowing of clinical decline. And of course, as you know, aducanumab has received accelerated approval in the U.S. The differences here are that lecanemab is dosed every 2 weeks versus every 4 weeks for aducanumab. Lecanemab is also being evaluated in preclinical Alzheimer's disease in the AHEAD 3-45 study, which was initiated in 2020. And in the Phase IIb study, lecanemab did not utilize a titration period. So patients could start immediately on the target dose and showed a relatively low rate of ARIA of approximately 10%. Lecanemab, as you may know, has been granted Breakthrough Therapy designation in June of last year and then fast-tracked in December 2021. And we are looking out and waiting for the readout of the Phase III study, which is a very large study, about 1,800 patients on this dose, randomized controlled trial. So we are really looking forward to the results, which we expect in the fall of this year, so in a few months from now.

I'm on mute. I do this once a panel. Thank you. I think as investors and analysts try to forecast the probability of success, obviously the Phase IIb hit its primary endpoint, but the 2 major caveats, right, that are always pointed out are the APOE4 imbalance and then the fact that CDR-SB missed significance, but is the primary in the Phase III. So I guess, what allows your excitement to transcend those 2 confounding variables as it relates to the prior data?

Thank you, great questions and very important points. So I think the first one I'll crack first, which is I think the premise of your question is essentially whether carriers progress at a faster rate compared to noncarriers, and therefore, having fewer noncarriers in the high-dose arm relative to placebo, could that have driven the observed slowing of clinical decline in the high-dose group, which is obviously the selected dose. So while that's an important hypothesis, I think what's really critical here is that Eisai has actually presented placebo data from the core 201 Phase IIb study suggesting that the APOE4 carriers and noncarriers progress -- from the placebo groups, this is -- at a similar rate. And then through separate analysis they've also concluded that disease progression at 18 months was not driven by ApoE4 status. So that is lecanemab data that makes me feel quite confident that APOE4 here isn't driving that superiority that we saw. But secondly, I think looking at the therapy, the class, we have our own data from aducanumab and we presented this in our briefing book for the Advisory Committee November 6, 2020, that shows that actually individuals in the Phase III aducanumab trials, that class, were classified as rapid progressors. We looked at whether there were any differentiating factors, and from the study sample that we had, based on demographics and baseline data characteristics, including APOE4 status, we did not see any difference either. Now granted, this was a small subset of 31 patients who were classified as rapid progressors, but if you triangulate that information, it is actually quite reassuring. So we don't believe that it's being driven by the APOE4 status. So that's one. The second part that you mentioned is also very critical. So CDR-Sum of Boxes remains a very, very important measure of both cognition and function, and it has been established as an appropriate measure in early symptomatic Alzheimer's disease. Importantly, it's been endorsed, as you know, by the FDA as an integrated scale that not only adequately and meaningfully assesses both daily function but also cognition in early AD and is acceptable as a single primary efficacy outcome measure. So you're right that it is intrinsically meaningful and there could be other scales that could be meaningful as well, but they would be composites such as iADRS. We've seen Lilly use that ADCOMS. These might be more sensitive, but we have not seen them being established as acceptable single primary endpoints for efficacy. So there again, I think while you see, and we expect to see, that readout in the Clarity AD study from lecanemab, we have also chosen this as our primary endpoint for our ENVISION study, which is our confirmatory trial for aducanumab, which we expect to screen patients in May. So again, I think the Envision primary endpoint is going to be CDR-Sum of Boxes as we have already shared publicly at 18 months after treatment initiation with ADUHELM or placebo. So we feel that it's very important for us to stay with CDR-Sum of Boxes.

Yes. Okay. Great. Do you have a high degree of confidence that, if the study succeeds, lecanemab may be an approval product?

Thank you. Yes, I think at this point, we have a very high degree of confidence, and there are a few reasons for this. One is that it's a very large and robust study, with just that 1 dose randomized controlled placebo is the comparator. It's about 1,800 participants, single drug arm study. We think that the positive readout will help support prior findings that indicated that targeting the pathological hallmark of Alzheimer's disease, in this case amyloid plaques, can improve clinical outcomes. Second, Eisai has already initiated rolling submission for lecanemab to the FDA under the accelerated approval pathway. They expect to complete it, as they've stated publicly, by Q2 this year. So it's already on track. And following their engagement with FDA, they have also stated -- Eisai has also stated that FDA has agreed that the results of the Clarity AD, when completed, can serve as a confirmatory study to verify clinical benefit of lecanemab. So I think that's quite confidence-inspiring at this point. And of course, the data will speak for itself. But we remain quite confident that, yes, we have tried to address all aspects here and that the readout will be treated as a confirmatory study.

Yes. Yes. Okay. Anything to add on this or ADUHELM before we move on to the rest of the pipeline?

Well, I'll just say that we are very excited. ADUHELM is the only approved therapy right now by accelerated pathway. This is very exciting. FDA reviewed this data with great rigor, high totality. And importantly, the data continue to reinforce our belief that amyloid reduction is a critical surrogate and that there's potential to impact clinical decline, as we saw in EMERGE and hope to confirm again in Envision. So we remain very excited about aducanumab. And then again, I'll just reiterate, we are very excited about the upcoming readout of lecanemab. We will continue to engage with all our stakeholders. And I hope that CMS will make the right decision and will provide this to Medicare beneficiaries, which we believe would be the right thing to do.

Yes. Yes. Okay. All right. Very good. Can we talk about zuranolone? This is a drug I've followed for a really long time, and I was excited for Biogen when you did the deal with Sage. That data as well, right, again, right or wrong, in the investment community has been controversial, and the dosing approach and all that -- we're going to get into it -- but I guess when Biogen did the deal with Sage, one of the hopes was that they were increasing the dose, maybe the MOUNTAIN study on effect size could be an outlier. And since we've got data from WATERFALL and CORAL that are both positive, but the effect size for this drug is still smaller, maybe, than some other antidepressants, with all the caveats across trial comparisons. So may be just sort of bottom line, sitting at Biogen today ahead of a filing, are you as excited about this drug as you were 1 year, 1.5 years ago?

Great question. Yes, I'll say that we are -- I am excited, Biogen is excited about zuranolone. And there are many reasons. One is that it's been a very comprehensive development platform, the landscape program for MDD, major depressive disorder, and the NEST for the postpartum depression. So one is it's across several trials attempting to answer different questions that, in a composite way, the totality of the data are extremely compelling. The other is that depression is, again, a very high patient burden, high burden of disease globally in the U.S. And it's also a very common comorbidity in the patients and caregivers across several of our current therapeutic areas. So we believe that there's a lot of synergy and a lot of expertise -- scientific expertise, but also just interest, that we have in developing within the neuropsychiatry sort of framework and therapeutic area. Now given our engagement with KMEs and other market researchers, psychiatrists and PCPs, we think that there's -- there could be a lot of excitement. There is already, but I think with the totality of the data and a potentially approved therapy, there could be a lot of excitement, because of the rapid onset antidepressant, new mechanism of action, well tolerated, and can be used broadly in patients with MDD and potentially PPD. Through the LANDSCAPE program, there are 4 positive randomized controlled trials in MDD, MDD-201, WATERFALL, Shionogi Phase 2 and CORAL. And important insights fall also from the SHORELINE study, which is the largest prospective naturalistic study in MDD, I think not to be underestimated in my view. And while all of these are answering different questions, what we have seen consistently is a profile from zuranolone: rapid reduction of depressive symptoms as early as day 3; consistent safety and tolerability profile without the observed weight gain, sexual dysfunction, suicidal ideation and low discontinuation rate due to adverse events; improvements in quality of life; overall health across domains of feeling, functioning, well-being reported by patients and continued after completion of treatment; and a short course. I mean, imagine a future where patients with this kind of burden of disease can only need to take 1 to 2 therapies that are 2 weeks each in a 365 day period. That would be pretty much a paradigm shift for patients and for physicians who are treating these patients. So the flexible treatment approach and the optionality that ACPs would have and patients would have. So finally, I'll just say that when we look across the studies, we see that the patients with the highest unmet need, elevated anxiety -- so looking at a subgroup, baseline HAMD score of 20 or greater, traditionally harder to treat patients -- they respond well to zuranolone. So this is really, really becoming more and more compelling to me as I see the breadth of what are the options that we could offer patients. And then finally, I think switching gears to PPD. This becomes really, really important, right? So there's no real approved therapies apart from IV brexanolone. And obviously, while it is effective, it comes with a convenience factor that mothers will need to make choices on. So if -- we are very excited about the SKYLARK study reading out midyear, and it has the potential for being first-in-class oral treatment option as a 2-week short course treatment for mothers suffering from PPD. So I think taken all together, I am very excited. I think that we have to really consider this totality and it could be a paradigm shift, honestly, for patients and for prescribers alike.

Yes. Yes, yes. Okay. Okay. Very good. One question on SHORELINE. I think we've -- I've maybe asked you this off-line one time, but one of the most -- I agree with everything you said about this drug and especially PPD, it can be tremendous. For SHORELINE, I think the one thing we've been trying to kind of figure out from our team is the regulatory utility of the study. And if this was another run of the mill, chronically administered antidepressant, you'd probably be able to show acute efficacy open-label extension data, get approval and then maintenance of effect is shown later. Given the novelty of how this drug is dosed, it seems, to me at least, like the question of maintenance of benefit comes into play earlier if I were a regulator. Do you agree and do you feel like SHORELINE is just a safety study, or does it as an open-label trial definitively show maintenance of benefit? Honestly, I'm curious if you think this question is on point or if it's not even relevant, too.

Yes. It's a good question. I mean I think we can't speculate at this point, I can't speculate how the FDA might use or interpret the data from SHORELINE, but I think it's powerful data. It's important data that prescribers need and it's the largest naturalistic study. So I think that the primary objective was safety. So it will definitely complement the safety database, which is very critical in antidepressant -- with antidepressant therapies, and we've spoken about it multiple times even just today. But I think overall, it provides run alone real-world data insights. So I think that it will be a totality, and we can't predict what the label is going to say and such, but I think everything will be informative and will add to what we will be -- how we will be able to use it in the real world.

Maybe to ask this in a different way, right, as from a clinician's perspective, given that most clinicians perceive depression as chronic, do you think they'll have the data they need to have confidence to kind of maybe take someone who's had depression for a long time off therapy just after 2 weeks and sort of go with it and hope that it stays and hope that they don't lose them to follow up or to a relapse or something like that? Do you feel like there's enough information and guidance for prescribers on that?

I think there is certainly important information that we will glean from the totality of the LANDSCAPE program. And I think that there's a lot of work to be done in assessing how we might want to -- how prescribers might want to use this, whether it would be co-initiation with antidepressants and then phasing out or it would be as needed and creating optionality. But I think, again, I'll come back to the point that the rapid onset, in an area where you have to take drug for 6 to 8 weeks just to see any benefit, that's a big deal. So I think that that will become very important. Of course, I think we will continue to generate more real-world data as, hopefully, the drug gets approved, we will get more data. So we'll continue to see how this will evolve in the real world. And so I can't predict exactly. But I think what we do have here is actually pre-approval data in different settings, the composite and totality of which would be highly informative to a prescriber. I think that's what we're going to see, yes.

Okay. Okay. So to round this out, we're going to get the PPD data midyear and then completing the filing second half, correct?

That's right.

Okay. Okay. Excellent. Looking forward to that. We can talk about some of the ASO programs. So I don't know if there's any opening comments you could make pre on the C9orf data that were announced this morning. And I guess, in your view, what are, if any, of the implications for just ASO technology and neurology given you have a number of other assets there?

Sure. Great question. So first I just want to say -- obviously, this news just came out today, and I just want to say how incredibly grateful we all are at Biogen for the selfless commitment of individuals with ALS who participated in the study and the community's dedication to advancing research for this devastating disease. So I am -- this was obviously not the result we all expected or wanted. But this is important data that is going to help us get eventually to a therapy that does make an impact and is meaningful for patients with ALS. So having said that, I want to take a moment to just say that C9orf72-associated ALS, it's a very complex genetic form of ALS. And there are multiple mechanisms by which the scientific community, so the wider scientific community, believes that C9orf72 gene causes disease. And it's important to consider this today in light of the results. So we designed the BIIB078 study to test the prevailing hypothesis that mechanisms of disease for C9orf72-associated ALS were caused by the toxicity -- it's a toxic gene of function, and were caused by toxicity associated with the repeat-containing RNA and corresponding dipeptide protein that is then translated from that repeat-containing RNA. What the results of the Phase I study show us is that while BIIB078 was generally well tolerated, these results don't support that hypothesis that we were testing. So we weren't able to prove the hypothesis. And they suggest -- actually, there could be many, many other hypotheses, so it's more hypothesis generating, but they do suggest that the disease mechanisms in C9orf72 genetic ALS are far more complex. Because in those cohorts up to 16 milligrams, we saw no differences between the BIIB078 groups and the placebo groups. And participants in the BIIB078 90-milligram dose cohort trended towards a greater decline than in the placebo group on all secondary efficacy endpoints. And because BIIB078 did not show any clinical benefit, we have decided to discontinue the development. So this is really important, because as you step back and look at ASOs -- and I'll specifically comment on our other ASO, which is also attempting to address a genetic form of ALS SOD1 where there's also a toxic gain of function. So there's the SOD1 protein. And there are very, very important differences here because they are completely distinct subtypes. And not only are they different based on genetics, but they are also different based on pathophysiological diseases. So one of the important pieces that we believe is going to be an important area is that TDP-43 pathology is observed in C9orf72 ALS but not present in SOD1 ALS. So we will be looking at this very carefully, and it is possible that BIIB078, which was designed to exclusively target the mutant hexanucleotide repeat expansion-containing transcripts, reducing levels of that is not sufficient, because the propagation is by other mechanisms. So we will need to look at this. I'm very -- we are humble. We have post mortem data, we have other data. We'll be looking at this very carefully. Obviously, what we've released today is just high-level top line data, and we want to release it as soon as possible. So we are still analyzing the data. We're still looking at the totality. But I think the totality of data will be important. And then just turning back to tofersen, we've seen that actually it has an impact on the SOD1 levels, and we've also seen an impact on measures of clinical efficacy and the secondary endpoints, so that is a very different story. We continue to engage with regulators, it's a very different picture from the C9orf72 BIIB078 trial, and I'll stop there.

What do you see as next steps for SOD1?

Yes. So the next steps, they are very important next steps. So number one, we remain engaged with regulators, including the FDA. We continue to have our open-label extension, which is a very, very important aspect of the continuation of VALOR, which was the pivotal trial. And we have also initiated enrollment in our ATLAS study, which is presymptomatic -- SOD1 ALS -- and which has a very important design. So I think that this is very important. We continue to recruit for ATLAS. And we also, in recognition of the unmet need and our observations to date where we missed the primary endpoint, which is ALSFRS, but we saw a lot of movement on clinical end points, secondary end points, we continue to have an open -- an expanded access program in many countries around the world. So these are very, very important sources of data, but they remain open because we believe in the potential of tofersen and the impact that tofersen could have on SOD1 ALS. So that's where we are. Now I don't know if you are aware, but we also have BIIB105, and this is sort of alluding to your question on ASOs. We also have BIIB105. It's an antisense oligonucleotide as well, and it tackles ataxin 2, which is currently in Phase I. So we really believe that genetically validated targets are important, not across just ALS, but across our entire neuroscience franchise and pipeline. And ataxin 2 is also genetically validated target with intermediate repeat polyQ expansion in ataxin 2, which results in 7x increased risk of ALS and is associated with a more aggressive course. But the importance here, I think, that may not be obvious, is that ataxin 2 also could reduce -- targeting that could also reduce TDP-43 pathology, which we know is -- could be central and is observed in many patients with ALS. So this is already in Phase I. And so we really are -- we do believe in the platform, we believe that it's a very effective way. We've seen this with SPINRAZA. We have data from NURTURE where a few years ago, before 2016, really infants didn't see their lives extend beyond the age of 2, and now we have patients who were dosed presymptomatically NURTURE who are seeing their fifth birthdays.

Yes, it's unbelievable.

As this is unbelievable. I think there are lot of -- and I think it's also opened up and paved the way, right, for more innovation in the space. So we hope to be pioneering similarly in the [indiscernible].

Yes. Can I ask you a question about -- so I had a conversation with a really -- I won't name, but somebody who works in ASO development in neuro and had a really smart thought. And the premise of this conversation was SPINRAZA has this incredible efficacy, but then the next set of ASO CNS programs have the tofersen data that was a little bit more equivocal, and then Huntington's and now C9orf didn't work. Do you think that there's -- it's difficult in some of these diseases, where more brand areas are involved, to deliver an intrathecal ASO. Could that be part of the reason why? And is there anything you can do to address that? Or what's your view?

Yes. So I think overall, before we get into modality, which is ASO is a type of modality to address a druggable target, I think we need to think about the fact that the disease biology for neuroscience, I mean, it's really quite complex. They have different targets and many of them have distinct biology. So I don't think that we can have a universal set of learnings about and apply them sort of in a black-and-white manner. Just for example, as I mentioned, with the difference between C9orf72 toxic gene of function versus SOD1 [ and ] propagation potentially through TDP-43 or not, that could be one very important area. The other is that we may be mixing up a few things, like, for example, in ALS we believe that the ALSFRS really isn't a very meaningful endpoint to measure simply because of its variability, unlike CDR-Sum of Boxes that we just spoke of. So I want to make sure that we really look at the data carefully. So I think the way we do drug development at Biogen, we're looking at the disease biology, versus looking at the modality and what can it address. We're looking at the disease biology and then we're working backwards to the target where the science is breaking and then applying the modality. So I think there are absolutely learnings from this area, and we are not rushing. We are being very thoughtful. We are pausing. We're looking at this data, but we don't believe that you can apply it universally.

Yes. What do you think is Biogen's most underappreciated pipeline asset?

Thank you. I think a lot of our pipeline, unfortunately, is really not appreciated. And so I do appreciate the opportunity to speak about it.

What would be your favorite, though? If you had to pick. Can you pick one?

Sorry?

Are you able to pick one as your favorite underappreciated asset?

Yes, I actually can pick one. So stroke is an important area for our focus, and we are very excited with our data from BIIB131. You may know it as TMS-007. We licensed it in from Japan. And I think what's really important here is we regard ourselves as pioneers, we believe we have the expertise, and we also believe that the data that we see are important. And we make our decisions all based on data and evidence and also on the high unmet need, right? So if you step back, stroke, again very high global burden of disease. And today, really, the innovation is tPA, which is, as you know, a thrombolytic. But the limitation here is that you have to administer it within 4.5 hours of onset -- acute onset of stroke. And this is potentially to address the safety aspects of tPA. So in our Phase IIa study of BIIB131, we saw acute stroke patients getting randomized BIIB131 for placebo up to 12 hours after their last known normal. And the primary objective was safety, but out of the 52 patients who received BIIB131, none experienced symptomatic intracranial hemorrhage compared to 1 out of the 38 patients who received placebo. This was despite the extended treatment window where patients on an average received BIIB131 at 9.5 hours after onset, so more than double of the standard of care. But the other important piece here is that in this Phase IIa study we were able to show a statistically significant improvement versus placebo on the MRS, or the modified Rankin Scale, which is a registrational endpoint for functional independence and improvement in blood flow. We saw rate of recanalization, or improvement of vessel blood flow, in patients who had a visible vessel occlusion approximately 58% at 24 hours in those who received 131 compared to 27% in patients who received placebo. So we think it does have the potential. Obviously, it still needs to be developed. It's a mid-stage asset, mid- to late stage, but we believe it has the potential to be a best-in-class thrombolytic for the treatment of acute ischemic stroke by extending the time window, having safety and efficacy, and we're looking at the next steps. Now in addition to that, I also want to comment on BIIB104, where we expect to read out midyear of this year -- I know you asked me for one, but I'll take the liberty of two.

Yes. That's okay. Two's cool.

And that's another important program within our neuropsychiatry pipeline. Obviously, we have a front runner in zuranolone, but we have BIIB104 that will read out. This is an alpha receptor positive allosteric modulator. And we're looking at this in a Phase II study, again, very high unmet need, cognitive impairment associated with schizophrenia. It's got Fast Track designation. And you know how serious schizophrenia is, serious mental disease, set of symptoms. And we will be looking at the [ MCCB ] working memory domain. And this will work if it -- the hypothesis here is that it will impact the functioning of the NMDA receptor, which then also impacts synaptic plasticity. So in a Phase Ib study, we showed that it has an improvement in working memory, and I'm hoping that the TALLY study will read out this year and will further reinforce our belief in 104.

What -- how wide is your therapeutic index for 104? I think AMPAs in the past have had some [ tox ] related to seizures at too high doses?

Yes. Well, we believe that we have adequate therapeutic index. We are testing 2 doses in our Phase II trial, so I think we will -- so far, I think I'm quite cautiously optimistic.

Okay. Okay. Maybe to finish it off, where do you see the greatest opportunity for pipeline build-out right now, maybe therapeutic area, stage of development?

Yes, great question. I think for us, the most important piece here is that we have 4 drivers of growth, 4 pillars. Neuroscience, specialized immunology, digital health and biosimilars. We are looking across the 4 pillars to diversify risk. But that's not all, right? We're looking at genetically validated targets. I spoke about some of these, those help derisk early. But we are also looking at building where we have significant presence in late stage. And the ones, obviously, that are frontrunners in this area, neuroscience, of course, with Alzheimer's, Parkinson's behind it, MS, where we believe we are leaders, SMA, they continue to be very critical. But we're also focusing in neuropsychiatry, where I've spoken of zuranolone and then 104, we'll see how that feeds out, and in addition, specialized immunology, where we've got 2 assets for SLE in Phase III. So a lot of potential to build out our neuropsychiatry and specialized immunology franchises and a lot of overlaps with our expected expertise -- existing expertise, rather, in these platforms and a lot of overlap with MS, immunology and other areas. But it will always be driven for us by data and by the meaningfulness of impact to diseases with high unmet need. That has to come first for us.

Yes. Yes. Okay. All right, great. Any closing remarks you'd like to make before we finish up?

Well, I think that I'm very excited about the opportunities that lay ahead. We have some challenges, of course, but that's what drug development is all about. And I think we have a very resilient organization. I'm very excited by the leadership that Michel and the Executive Committee provide and by my R&D organization as well as the R&D leadership team. So I know that we can face the challenges and we have the pipeline and the substrate to bring a lot of benefit and meaningful change and benefit to patients' lives. So I'm very grateful and appreciative for all the discussion and thank you for the opportunity.

Yes. No, thank you so much for joining. I really appreciate it. Thanks, everybody, for listening in. I really appreciate that as well, and I hope to see you on the next panel. But, Priya, have a great rest of your day. Thank you.

Thank you. Bye-bye.