Biogen Inc. (BIIB) Earnings Call Transcript & Summary

Good morning. My name is Jess, and I will be your conference operator today. At this time, I would like to welcome everyone to the Biogen and AAIC 2026 Webcast. [Operator Instructions] Today's conference is being recorded. Thank you. I would now like to turn the conference over to Tim Power, Head of Investor Relations. Mr. Power, you may begin your conference.

Thanks, Jess, and good afternoon, everybody. Thanks for joining us today. I'd like to start by just pointing out that we'll be making forward-looking statements, which are based on our expectations. These statements are subject to certain risks and uncertainties, and our actual results may differ materially. I encourage you to consult the risk factors discussed in our SEC filings for additional details. Joining me on today's call are Dr. Priya Singhal, Head of Development; and Dr. Diana Gallagher, Head of Clinical Development for immune-mediated and neurodegeneration. You can access the press release and supporting materials regarding today's presentation as well as a replay of the call at biogen.com. And I'll now hand it over to Priya.

Good afternoon. It's great to be speaking with you from AAIC. As you know, we also just had good news for patients with the approval of LEQEMBI Click for treatment initiation for early Alzheimer's disease. We at Biogen have chosen to tackle Alzheimer's disease for several years. And having been at the forefront of innovation in Alzheimer's, it is really good to be here at a key Alzheimer's medical meeting with important progress in the field to share. Today, we are pleased to join you to present data for diranersen, which we believe demonstrates an important advancement in targeting tau. The CELIA study was a pioneering effort. What we have observed is that diranersen offers a differentiated approach that leads to tau lowering, demonstrates robust tau tangles reduction in the brain and importantly, has a substantial impact on multiple clinical endpoints. Several aspects of the SELIAR data are unprecedented. As you know, the purpose of the CELIA study was to determine if we could achieve proof of concept and therefore, identify a dose to advance into Phase III development. We know that many of you have had questions about how we are thinking about this asset because we did not observe a typical dose response with this drug, which was the primary endpoint. What I can tell you today is that we are seeing promising results that importantly provide us with proof of concept and an emerging dose, which is supported by: one, an unprecedented reduction in both tau tangles in the brain and clinical efficacy; two, diranersen being favored for all doses, showing consistency across the data set; three, the 60-milligram dose twice a year, demonstrating the largest and most substantial benefit across nearly all endpoints; and finally, the results that reinforce the prior findings from the smaller Phase Ib study. Furthermore, not only did we see a CDR Sum of Boxes benefit similar to the anti-amyloid therapies at 18 months, but also cognitive efficacy signals not seen before in an Alzheimer's clinical trial for a potential disease-modifying therapy. This is important because while CDR sum of boxes is an important endpoint for clinical research, patients and their doctors care deeply about cognitive endpoints and cognition overall. diranersen now has the potential to be administered just twice a year and has a mechanism that is not associated with ARIA. Now let's walk through what we have learned from Sela with Diana.

Thank you, Priya. Let's start with a reminder of the 2 targets most closely associated with Alzheimer's, amyloid and tau. While the current treatment landscape is focused on amyloid, research has also told us that tau and more specifically tau pathology spreading across the brain is much more temporally linked to cognitive decline. Thus far, as you know, antibody approaches have not been successful in showing clinical benefit. We believe that this may be because they primarily target tau outside the cell and are likely unable to impact the intracellular tau pathology. However, diranersen acts a very different mechanism. Our ASO targets the reduction of the translation of the MAPT gene into tau protein, thereby reducing the production of all isoforms of tau inside the cell, impacting both intracellular and extracellular mechanisms. Our hypothesis remains that reducing tau production results in overall reduction of tau pathology in the brain, potentially translating to a clinical benefit as a result, neither of which have been seen before in any other tau-directed agents tested thus far. So now let's talk about how we tested that hypothesis. On this slide, you'll see our Phase II design. Let's remind you of a few key features. This was an 18-month study recruiting patients with early Alzheimer's and mild cognitive impairment, where patients were randomized to placebo and 3 different dose levels of diranersen. We measured several clinical endpoints, including CDR sum of boxes, ADAS-CoV13, MMSE, ADCS-ADL-MCI as well as modified IADRS and AdCom. We measured total tau in the CSF for all subjects. And then within a substudy of the population, approximately 30% of the subjects, we also measured tau PET to assess changes of tau across different regions of the brain. The primary endpoint was intended to formally assess a dose response to CDR-SB and change from baseline. Before we look at the results, let's review who we recruited into the study. Here, you can see the baseline characteristics of the patients recruited. What you can see is that the arms were generally well balanced across elements such as MCI versus early AD as well as ApoE4 gene status. So now let's talk about what we found in this study. To start, the primary endpoint assessed was whether increasing dose levels improved efficacy compared with a predefined dose response. As you know, we did not meet and observed that dose response pattern in this trial. Importantly, as shown on this slide, diranersen was favored over placebo across nearly every endpoint though, which is very promising. So let's take a close look at what we saw in the data. Starting with CDR sum of boxes. We saw the largest and most substantial effect at the 60-milligram every 6-month dose level. What's most interesting is that for this dose, you're seeing the difference compared to placebo, which is similar to the current anti-amyloid therapy. And when we dig deeper into the data for that 60-milligram dose, we see a very compelling profile emerge. On this slide, we have placed the 60-milligram data from today's AIC presentation into context alongside results from historical amyloid therapy trials. While these are not head-to-head comparisons, we believe they are useful for thinking about our decision to move diranersen forward to Phase III. Recognizing the sample size limitations of the Phase II, what we are seeing is that the benefit on CDRSV is similar to anti-amyloid, but with a greater effect size on cognition than what we have seen with other agents. For example, we observed 42% for ADAS-CoV13 and 50% for MMSE. You'll note that for ADCS, ADL MCI, we were not seeing separation for the 60-milligram dose at this endpoint so far at 18 months. But importantly, we will continue to follow patients and look forward to seeing how this evolves over time, including looking at 24 months. And furthermore, even at the 18-month time point, the story for functional benefit is much broader, as I'll show you on this next slide. And that's because when we examine the individual components of the CDR-SB, the benefit is not driven solely by cognition, but is also evident in the functional domain. As you may recall, CDRSB consist of 6 domains and the top 3 on this slide assess cognition, and you can see consistent treatment effect against each of those measures. But interestingly, the bottom 3 domains assess function, and we also see separation across these as well and a pattern that's reminiscent of what we've seen before on dose response. So we believe the fact that we are seeing impact on both cognition and function is encouraging. On the next slide, we'll move into the biomarker data. And while we see substantial reductions in CSF tau, as you can see in the top panel of this slide, -- what further strengthens our confidence in these clinical findings of CELIA is that we're also seeing reductions in tau tangles in the brain as measured by tau PET. Looking at the bottom panel, you can see that in this subset of approximately 30% of patients where tau PET was measured, we see the burden increasing as would be expected for placebo in early AD. However, in all 3 diranersen arms, tau pathology accumulation was not just slowed but reduced from baseline. This has never been seen before with any other tau-directed agents. And in our view, this is a very important finding. Turning now to safety. What's important to note is that diranersen was well tolerated in the CELIA study with no new safety signals identified as compared to our Phase Ib. Most AEs were mild or moderate and not serious, and the majority of patients completed dosing and even rolled into our ongoing long-term extension study. As we saw in the Phase Ib, confusional state was reported as an adverse event, and there was a higher incidence of confusional state AEs observed at those higher doses. But importantly, AIA, which was not anticipated with this mechanism of action and the results in SILI were consistent with this expectation, it was not observed. I hope the data we overviewed today helps you understand the SILY results and the implications. I'm going to now turn it back over to Priya to close out the discussion.

Thanks, Diana. I would now like to address what is next for diranersen. First, we will continue the long-term extension study of SILIA to evaluate durability as well as safety and tolerability. We expect to have 2-year data for CLIA by the close of 2026. And as more of the CELIA data becomes available, we will look to share and publish these data at medical congresses and in the literature. We have also reviewed the CELIA data with several external experts who are equally excited by the results and agree that advancing diranersen to Phase III is the appropriate next step. We are now working urgently towards a Phase III study design and overall evidence generation package. For both of these, we continue to engage with key leaders in the field and will engage with global regulators as we take a thoughtful approach to determining the ideal design of the Phase III and additional data generation approaches. We look forward to sharing more with you in the future. To conclude, we are excited by the opportunity to bring forward diranersen, which we see as having a compelling and differentiated profile, especially with respect to cognition, a tolerable safety profile without increased risk of ARIA and the potential to be administered just twice a year. And I'll note that the dropout rate in the trial was low and 94% of patients who completed the study have continued into the long-term extension study, which is very encouraging for obvious reasons. Therefore, dosing continues in the LTE, and we have the opportunity to continue to learn more about diranersen, including the 2-year data. I hope that the review today helps you appreciate why we are excited about diranersen's overall potential to be the first tau-targeting medicine for early Alzheimer's disease and the reasons that we are advancing it to Phase III. Tim, please go ahead and open the call for questions.

Thanks. Can we go to the first question, please?

[Operator Instructions] We will move first to Evan Seigerman with BMO Capital Markets.

As we look at the data, is it possible that you're not -- is it possible that we're knocking down tau too much at the high doses to not get an effect? And how do we know the 60-milligram dose is the right dose to take forward? There is a potential benefit of tau in some patients, as you know. And we don't know exactly if you're kind of cutting out too much of that in these patients. So a long way of saying, is 60 milligram the right dose to be taking forward? Should we be looking at lower doses here?

Thanks, Evan, for the question. So yes, I mean, one of the reasons we moved from Phase Ib, where we had in a small sample size tested some of these dosing paradigms to a dose-ranging Phase IIb was to answer the exact question you're pausing there and to determine, do we have a path to Phase III and what is the dose. What we see in that 60-milligram dose, I think, is that we see the reduction in the tau CSF as well as a reduction in the tau tangles, which corroborates what we've seen in Phase Ib. And we see it across every endpoint that we measured of the key secondary endpoints with the exception of the ADS MCI. And so that's really encouraging to us. In terms of assessing whether or not that we have the dose in its totality right, you see this also diranersen is favored across every dose, but our goal here was to identify the dose or the emerging dose that we take forward into Phase III. And we think that's what we've done here.

We'll go next to Salveen Richter with Goldman Sachs.

When we look at the 60-milligram dose arm, the end was small here. There were a lower amount of patients with ApoE4 and there were fewer women in that cohort and lower amyloid PET potentially. So I guess what I'm trying to ask is when you look at all the information you have at hand right now on top of the efficacy metrics, -- just speak to how you think about further elucidating this on top of the dose work in the Phase III trial and the thoughts of doing a combo with amyloid in Phase III? And if I could just also add what's driving the confusion state?

There's a couple of questions there. Maybe we'll take them in order. And I think the first one had to do with understanding what we're seeing in this proof-of-concept study. And it is acknowledged that we were testing 3 different dosing paradigms versus placebo in that 2: 1 to 2:2. And so we think we had nice overall distribution in the baseline demographics, but of course, it's never going to be identical. And we think that when we're looking at that, we see effect sizes that are pretty comparable across -- we've started to look at some subgroup analysis. We will be presenting that in the future. But what I can say directionally is that we're not seeing any significant differences based on some of those demographics that you had mentioned earlier. Of course, as we move into a large Phase III, we'll be able to confirm that in the next study. I think the next question you had was how are we thinking about designing our Phase III. It's a great question, and it's something, of course, we'll be engaging with regulators about in the coming weeks and months. And I think, first, we have to remember, this is the first drug that has shown the ability to hit the target, reduce the tau tangles and show the clinical benefit. So we will be engaging with them to understand what is necessary from a sort of dosing standpoint as well as safety and ultimate efficacy in order to meet expectations there. And so we're really excited about what tau can do for Alzheimer's disease in particular. And we also know that we will be hopefully engaging in a market that continues to have utilization of multiple sort of drugs. So we will be -- we're thinking just as you are about future states where patients based on their personalized biomarkers may sort of use drugs at the same time or sequentially and more to come there. But right now, we're focusing, first and foremost, on the tau monotherapy and then thinking about adjunctive data we can generate in parallel to that.

And finally, I think you had a question, Salveen, about the confusion event.

Sure. So as we mentioned, you did see that there was an increase in confusion in a dose-responsive way that Kathmomory had shown on the podium today. I think what's important to note is that most of those events were mild to moderate. Most of them occurred within the first 7 days of dosing and then resolved within 7 days of dosing and patients were able to treat through that and continued on diranersen. So 94% of patients who were eligible rolled over into the long-term extension, which was a question we didn't know from an intrathecal standpoint, how would patients and families react to this. So we're seeing a lot of retention for this drug and this route of administration, which we think is encouraging as we move towards Phase III.

We will go next to Eric Schmidt with Cantor...

Maybe given this is a bit of a smaller Phase II, how did you guys convince yourself that the placebo group is not behaving apparently? Any comparisons you might have looked at would be helpful. And then I know there was also a question about the rate of completion on the study with the higher dose groups having a lower rate of completion. Could you comment?

Sure. So generally, in looking at placebo, acknowledging this is, as you said, a small study and cross-study comparisons are challenging. We see it's generally consistent with trials or sort of peer trials with mild Alzheimer's or MCI. So for example, we see the CRSC moving at about 2 points. There are some small changes and differences that you can see when you compare every single trial, but overall, we see it moving as expected. And I think you had a second part of your question, which had to do with the higher dose discontinuation. Yes, overall, it is true, and we showed in the disposition on the podium today, what those rates were. I think it's important to note overall, in particular, comparability, particularly between the placebo group and the low-dose group. So that's something we'll continue to study. But again, the fact that we had high completion rates for an Alzheimer's study, particularly one that's intrathecally administered and high rates of rolling over to the long-term extension is something that we were looking to assess and are pleased with seeing.

And I think we're really pleased with the emerging dose and its benefit risk profile, right, because we are seeing very consistent rates of adverse events and rates of discontinuation. So that remains quite promising.

We'll go next to Michael DeFi with with Evercore ISI.

Number one, what could explain the disconnect between tau lowering and clinical response? You think that the PD effect and clinical response will be positively correlated, but we saw the opposite. And also, why do we see cognitive benefit, but no clear functional benefit on ADCS-ADL? And just one more. Why would you say the decrease or benefit on MMSC was so much higher than it was on CDRSB, especially when CDRSB is generally considered to be a more sensitive measurement?

So we'll work to take those in order. And I think one of the first things I want to do in terms of tau lowering is sort of address that. We definitely understand why you're asking this question. It's important to note that it was a sub-study. So of those 131 patients that we were able to have tau and that there's a lot of individual variability in baseline tau levels, particularly in early AD. So we're going to continue to assess all of the different components of that as we -- to look at the variation in baseline tau burden as well as other baseline factors and what of all of those could be sort of rolling up into efficacy differences as well as differential tolerability. So that's a good question and something we're continuing to examine. Your second question, I think, has to do with ADCS-ADL-MCI. It is a functional endpoint, and it is something that we didn't see moving on that 18-month time point. Interestingly, I'd point you back to the Nature Medicine paper where when we looked at another functional endpoint, we looked at week 37 and then week 100. We did see there actually on FAQ that it wasn't moving very much at 37 and then we started to see it at 100. So one of the reasons we have a long-term extension is to continue to see if any of these endpoints, which aren't behaving maybe as we would initially expect change over time. But even separate from whether it does or doesn't move as we look over time, the reason we looked at other composite endpoints was because they have function in them. So as you saw in the podium, the CDR-SB has both cognition and function, and we do see a dose response similar to what we saw with the other endpoints on function. So we believe we're seeing an effect on function. It's not that there's only cognition. We see cognition and function, both in the CDRS-V as well as on the iADRx. So that was why this weight of evidence across multiple endpoints, key secondary endpoints is how we are confident that by targeting tau, we're seeing this clinical benefit. I think your last question was why would there be a disconnect. And that's a great question. We're pleased to see the impact on CDRSV at that level that we are seeing and on cognition and function. But you're right, that effect on MMSE and ADASCo for cognition is really quite remarkable in cross-study comparisons. So it's something we're kind of at this pioneering space with what happens when you hit tau, what clinical benefit can you see. We're definitely seeing a significant impact on cognition alongside function. So we'll continue to examine over time, whether they remain comparable or continue to have this pattern. But I want to emphasize that it's in both cognition and function.

I'll just add that CDR Sum of Boxes, as Diana mentioned, has got domains for cognition as well as function. We're seeing with ADCS that there could be a lag. So we might see this improve at 24 months, and that is something that we're waiting to see the 24-month data. That's going to be an important time point. And I think what's really encouraging about MMSC is that it is a test that physicians administer in the office and seeing such a treatment effect at the 18-month time point is really encouraging. But it could be a function, and I'm speculating here, it could be a function of the fact that tau is really the penultimate sort of protein accumulation before you get symptoms. So maybe there is a differential on cognitive outcomes with an agent that addresses tau. We don't know this. This is speculative, but I would say let's look at the long-term extension data at the 24-month time point and see how these evolve.

We'll go next to Paul Mattias with Stifel.

I was wondering if you looked at inflammatory biomarkers or neurofilament at either of the higher doses and if you see any changes there? And I guess just more broadly, it sounds like you don't want to speculate on the mechanism of the inverse dose response. But at this point, are you -- like would you say you're convinced that there isn't some on-target tox when you lower tau too much?

Yes. We have -- thanks, Paul. We have looked at inflammatory biomarkers. We haven't yet disclosed that data. We look forward to disclosing it with sort of more data. And I would say that it's possible. I think there are several hypotheses as to why a higher dose may not actually translate to clinical efficacy. But I think we would be speculating. And I think it's important for us to kind of come back to the fact that we do have a dose that has emerged with a very encouraging benefit risk profile. As you say, tau does have a role in the normal physiology of the brain, microtubules and neuron. So all of this is possible. But I think we are really encouraged and excited by the fact that we've isolated the dose here.

We will go next to Terence Flynn with Morgan Stanley.

Maybe 2 for me. I was just wondering if you can comment at all on the phosphorylated tau data. It looked like that was something you did not present today, but just wondering if that's consistent with what you saw in Phase I. And then on the Phase III design, how are you thinking about the primary endpoint? I know when you did the sell-side call several weeks ago, you weren't committing to CDR sum of boxes. And now that we've seen some of the other secondary endpoints, looks like the effect size might be more dramatic. So any preliminary thoughts on how you're thinking about primary endpoint for Phase III?

Turning to the -- your first question first on the phosphorylated tau endpoints. We are analyzing that data, and we're looking at it, but haven't disclosed it yet. So we have a lot more data that we're going to be disclosing in the coming months, and we can look forward to sharing that. In terms of your question about what will the primary endpoint for the trial be, we will be engaging with regulators, as I mentioned, both in the U.S. and rest of world. And I think we have not finalized what that will be. We do see an effect on CDRST, which was one of the important things that we wanted to do here. We also see it on IADRS. We also see it on cognition. So again, this weight of evidence we have across multiple endpoints gives us the opportunity to engage in that dialogue. But we have an understanding, particularly in the U.S. of typically what their expectations have been. And so we'll have that conversation with them.

And maybe we can offer that we have looked at p-Tau 181. We know that it's consistent, and we haven't shared details, but we look forward to sharing that.

We'll go next to Michael Yee with UBS.

Two questions. One is, do you believe that the tau reductions are basically pretty much all the same in overlapping confidence intervals on the doses. And so maybe all of those doses pretty much give you the same result. And so when you look at the results there and try to concord that with cognition, that basically all the overlapping confidence intervals are also pretty much the same also on IiARS. So to me, it looks like only CDR sum of the boxes is a standout, but all the other ones basically overlap, particularly IA. So do you agree with that comment? And would you be open to using iADRS, which is arguably a potentially better endpoint for Phase III?

So I think all things are possible, and we're examining the data, as you said, and thinking about engaging with regulators. Again, we'd like to point out, we do see that impact on CDR sum of boxes as well as the iADRS. And so seeing it in both places gives us that optionality and flexibility, which we're encouraged by. And to sort of toggle back to your question it is true, there was a robust impact on the tau biomarkers, both in the lowering in the CSF as well as the impact across the tau PET. And so we continue to examine that data as well. But you're right, we saw a very robust sort of target engagement and reduction of tau tangles across all the doses, which was absolutely encouraging.

And we do think that a lot of those confidence intervals overlapped. So we think that it is -- it's hard to say that there was a dose response on the tau reduction. The second thing I would just say about IDS and CDR sum of boxes, I think that we are -- I just want to be clear that we do believe that there is -- the lowest dose has the largest and most substantial effect. as of now. We will continue to examine this at the 24-month time point that we've called out. We think it's important. And so we'll have to see how that evolves. But as of now, we believe the low dose is distinguishing itself. And so that's important. And I think IDS gives us confidence, right, because we included a slate literally of all cognitive and composite endpoints that other trials in the anti-amyloid have ever used, and we see a consistent signal and consistently better at the lowest dose. So there may be an overlap in some, but we think that the lowest dose is distinguishing itself. And actually, for CDR sum of boxes, even the absolute change at the 18-month time point of about 0.54 is quite impressive. So we're continuing to see whether that will evolve. But yes, we think the lowest dose is declaring itself.

We'll go next to Chris Shaw with JPMorgan.

This is Taylor Hanley on for Chris Shah at JPMorgan. We were just wondering, can you give some thoughts or color on how you're thinking about the commercial potential for Dinersen versus LEQEMBI and COSemLA just when you're comparing both their efficacy and their safety profile?

Sure. It's a great question, and it's one that we've thought about. I mean just stepping back, we know that about 6.5 million patients have Alzheimer's disease. This number keeps growing. There's an annual incidence of about 500,000 patients. I think the anti-amyloids have really done a valiant job and continue to -- this market continues to develop. And specifically, as blood-based biomarkers come in, we think that diagnosis rates will increase. The important thing to remember about an anti-tau agent is that we think it offers a different therapeutic modality for patients who have early Alzheimer's disease, and it's actually quite proximal to their development of symptoms. So while it's early to really paint the whole treatment landscape, we think that this is quite a large market, and there will be different patient segments. And as Diana mentioned, who may be typed on a different biomarker profile and would be suited to different options. For example, while the CDR sum of boxes at the 18-month time frame is consistent with the anti-amyloid, we haven't yet seen the 2-year data, and we know that the other cognitive endpoints are orders of magnitude different. So how will all of this evolve at 24-month time point? What will the Phase III show? And how will we design that Phase III? Those are the questions that we're trying to tackle. But we think there is an important role for a therapeutic modality that tackles tau. And I think that we are also looking and very cognizant of the fact that if we are successful in Phase III, we would launch into a space that does have anti-amyloid treatment. So how do we tackle that? What data do we need to generate to ensure that prescribers are comfortable when they think about all these different patient profiles and treatment paradigm. So I think it's the totality, but we think this is a large opportunity. It's obviously a few years away. We're just at the threshold of starting a Phase III. So a lot of work ahead of us, but we believe it's going to be an important and exciting opportunity.

We'll go next to Mohit Bansal with Wells Fargo.

This is Sadia Rahman on for Mohit. Just wondering if you've looked into any subgroup analyses here yet and if that might give you more confidence that any of the small baseline imbalances, for example, on CDR-SB or CDR global scores here might not be driving any differences in these responses between the different doses?

So it's a great question. And we showed the sort of overall demographics of the patients and the differential in the different dose groups. It's pretty well done for a small sample size, but you called out a few small changes. I think due to time constraints, we weren't able to show everything and that analysis -- subgroup analysis is ongoing. But what I can say is based on what we've been looking at so far, we're seeing efficacy across those patient profiles. So we're obviously ultimately going to consider presenting those things, but there's nothing that we're really seeing differentially at this time.

And I think overall, we're just seeing very consistent overall trends. We've also looked specifically at some very important subgroups like the ApoE4 carriers -- and we know that there's a differential in terms of can anti-amyloids are they being used to treat that population? And actually, we see consistent trends across. Small numbers in a small trial, but quite consistent trend.