Conference Transcript - ANNX 2026-06-03

Andrew Tsai, Analyst — Jefferies

Ready for our next session. I'm Andrew Tsai, Senior Biotech Analyst at Jefferies. Thanks for joining us. And it's my pleasure to have the Annexon team with me. To my direct left is Doug Love, President, CEO, and to his left, Lloyd Clark, SVP of Ophthalmology. Correct. Welcome, both of you. There may be some people in the audience who are less familiar with the story, so can you talk about the Annexon story, what programs you're working on, and the milestones over the next 6 to 12 months could be helpful.

Douglas Love, CEO

Yeah, happy to do so. So thanks, everyone, for joining us today. We're very delighted to be at this year's Jefferies Conference. So as Andrew said, I'm Doug Love, CEO and President of Anexon Biosciences. What Anexon is doing is pioneering the next generation of complement therapies to treat neuroinflammatory diseases. We are the world's leaders in doing that, and we'll talk more about that. But first, I want to just differentiate us from the other complement approaches that you may be familiar with. We're targeting upstream C1Q, which is the initiator of the classical complement pathway. What that means is we're providing the most complete protection against neuroinflammatory inflammation that is driving a host of differing diseases across the body, brain, and eye. It's quite differentiated from downstream approaches you may be familiar with, like C3 and C5, where they've shown really unique efficacy in certain diseases like PNH, for example, where those are the primary drivers of the inflammatory process. However, they've shown much weaker data, unfortunately, in a host of neuroinflammatory diseases, predominantly neuromuscular diseases, as well as neurodegenerative diseases like Guillain-Barre syndrome, ALS, and geographic atrophy. By targeting C1Q, as I said before, we're blocking the entire inflammatory process and has led to really unique data sets across a range of diseases. We're led by two late-stage registrational programs. One is Guillain-Barre syndrome, the world's most aggressive acute neuromuscular disease where we had a whopping effect in our phase three pivotal program. We'll talk more about that. And then in geographic atrophy, a chronic neurodegenerative disease where we are the only program to show significant vision preservation and, importantly, preservation of the photoreceptor neurons in the eye that are responsible for driving vision. So where are we today? Really an exciting time over the course of 2026 and 2027. We refer to it internally as our winning season, so we're making no bones about that. We've got a host of catalysts on the horizon, starting with Guillain-Barre syndrome or GBS. As I said before, we were already successful in our pivotal phase three study. We have filed for approval in Europe, and we are looking to file for BLA submission in the U.S. Prior to filing for BLA submission, we will release a cohort of data for our ongoing forward study. This is a study of GBS patients in the U.S. and Europe. We anticipate the data set will be somewhere in the neighborhood of five to ten patients. We'll do that before our GA Phase III data readout, and the import of that data set is twofold. One, it's really allowed patients and physicians in the U.S. and Europe to get their hands on TAN-Rupabart and understand the powerful mechanism and impact it's having on GBS. And two, it is contextualizing or demonstrating the consistency and outcomes of the patient population in the U.S. and Europe to the Phase III data set that we ran in Southeast Asia. So really excited about getting that data set out there. Thereafter, we will get data out on ANX 1502, the first and only small molecule oral program in the classical complement space, also really energized by this program. The intent of this program is to target a host of neuromuscular diseases like myasthenia gravis, CIDP. This initial data set will be a relatively small data set, have a dozen or so patients. There we're looking to show normalization of excess complement activity, as well as impact or normalization of key efficacy markers like bilirubin. Finally, the big coup de grace for us at the end of the year in Q4 is geographic atrophy. So this will be our Phase III study that is designed to replicate the findings from our Phase II study, and so we're really excited about that. Along the way, we will continue to do things from a corporate side. This is a wholly owned platform. We are making various moves here to bring in additional non-dilutive dollars and other things, and so you heard about that as well over the balance of this year.

Andrew Tsai, Analyst — Jefferies

Oh, great. So all three programs, to be clear, have catalysts this year. That is correct. Okay, very good. But so maybe starting with GA to start phase three data in Q4, at its core, you're differentiated compared to Appellus Iverick because you can preserve, show the preservation of 15-letter BCVA law. So remind us what you showed in phase two at the equivalent, I guess, phase three dose and time point and so forth, drug versus sham in phase two. and how does that kind of compare to what the other drugs show?

Douglas Love, CEO

Yeah, I'll start, and then I'll invite Lloyd to join us. So I'm really excited about it. There's two components to our data that we want folks to really focus in on. First is the how. How did we demonstrate what is actually functional benefit, i.e. preservation of vision? The how relates to the structure in the eye that is responsible for asional acuity. It sounds simple, but geographic atrophy is a neurodegenerative disease. It has not been discussed as a neurodegenerative disease over the last decade or so. In all neurodegenerative diseases, what you want to protect are your neurons. That is where your function resides. And in the case of geographic atrophy, these are called photoreceptor cells. We had significant preservation in our phase two study across the entire macula, a differentiated impact in the central retina, which I'll have Lloyd talk about, and why that matters for specifically a disease like GA. And then all of that translated to significant preservation of vision on all of the measurements of vision. Best corrected visual acuity, 15-letter loss is the gold standard. Endpoint has been used in all retina programs. And so we're very happy to use that in this program. And we demonstrated that as well as a host of other measures. But maybe I'll turn it over to Dr. Lloyd Clark, SVP of ophthalmology and a currently treating retina specialist.

Lloyd Clark, Other

Yeah, thanks, Doug. Appreciate it very much. i think the key to understanding our our program is to understand that our drug inhibiting c1q has a specific primary effect on photoreceptors c1q is an important localizing molecule on photoreceptor synapses when they're at stress and they're targeted for removal and so it's a understanding the key differentiation point under is important understanding the type of data that doug's talking about that's why we have demonstrated significant protection of photoreceptors across the entire macula but most specifically in the central macula which is crucial for central visual acuity and again this is because of the underlying differentiated mechanism of action so understanding the how is really the first step in understanding the data so recognizing this This is a neuroprotective drug that protects photoreceptors, leads you to the data that Doug described.

Andrew Tsai, Analyst — Jefferies

Got it. And sorry, one more time, the phase two, what did you see then?

Lloyd Clark, Other

Oh, so in terms of the phase two data, we saw overall a 73% risk reduction in 15-letter loss. So our primary endpoint is confirmed 15-letter loss at two consecutive visits in the phase three. Overall, we saw a 73% risk reduction in that endpoint in the monthly group compared to sham. We saw similar directional benefits and function across a number of different measurements, visual acuity, low luminance visual acuity, a number of functional endpoints. So not only did we see directional benefit with Vaughn improvement, but we also saw a dose response. We saw an increased effect in patients treated monthly compared to every other month. And then the other key finding when we talk about function is that we saw a disease-modifying effect in the Phase II clinical trial. So we treated patients either monthly or every other month in the Phase II and saw vision preservation effect with treatment, but then the final six months of the Phase II was off treatment, and there we saw a continuation of vision loss in all three treatment groups consistent with a disease-modifying effect. So all of these are clear measures of visual protection with our C1Q inhibitor.

Andrew Tsai, Analyst — Jefferies

And so Apollos-Iveric, on this same type of gold standard endpoint, what would they show?

Douglas Love, CEO

That's an important point. So neither have demonstrated visual protection in their programs, and so that's just the short answer.

Andrew Tsai, Analyst — Jefferies

Okay, yeah. And so phase three, what are your assumptions based on the phase two, then, drug versus placebo separation?

Lloyd Clark, Other

All right, so in order to construct a phase three clinical trial with an event-based endpoint, the first thing we needed to do was establish an appropriate sham event rate, and we did that using two main data sets. First is our phase two data, both what we've presented publicly as well as other analyses, as well as primarily the lampalizumab natural history cohort. We used those two data sets to establish an appropriate sham event rate we would expect in phase three. In terms of a treatment event rate, largely we used the Phase II results from patients treated monthly with VON approvement. From there, we took conservative adjustments in both to create appropriate treatment effect and sample size to power at a Phase III level with two sub-analyses.

Andrew Tsai, Analyst — Jefferies

And so maybe ask another way, what exactly is clinically meaningful? What kind of placebo-adjusted separation?

Lloyd Clark, Other

Right, so what's clinically meaningful in our program actually is the endpoint itself, which is 15-letter loss. 15-letter loss is really the gold standard clinical trial endpoint in retina. All of the transformative therapies that have been utilized in this space and have made significant meaningful impacts on patients were all approved on the protection against 15-letter loss. And so the clinically meaningful aspect of this endpoint is the actual definition of the endpoint itself. In terms of a delta, really a statistically significant difference between those two groups is a win with a clinically relevant endpoint.

Andrew Tsai, Analyst — Jefferies

And then I think from phase two to phase three, you've also tried to enrich the population to help your chances of phase three success. So what exactly did you do? Like how much can it help by, in your view?

Lloyd Clark, Other

The key enrichment strategy was eliminating enrollment patients with very poor vision. Patients that have extremely poor vision with geographic atrophy don't have 15-letter loss events. And so we excluded a group of patients that was essentially the bottom 20% of the Phase II study that did not have 15-letter loss events in an effort to bolster event rates in a Phase III study. The other main strategy was to adhere to a target of patients that have foveal involvement. It's well known from our studies as well as others that patients with foveal RPE atrophy have much higher rates of vision loss compared to patients that have lesions far from the center of the vision. And so those were two important strategies.

Andrew Tsai, Analyst — Jefferies

And you mentioned this is an event-driven study, so can you help us define what that means exactly?

Lloyd Clark, Other

Right, so the primary outcome measure is a confirmed 15-letter loss at two visits. So if a patient, let's say, for example, at month eight has lost 15 letters of vision compared to baseline, then that needs to be confirmed at month nine in order for it to count as an event. That's the definition of the primary outcome measure and an event in this study.

Andrew Tsai, Analyst — Jefferies

I see. And you're tracking the event rates, again, based on phase two in the other natural history study, it sounds like.

Lloyd Clark, Other

That's correct. we created a model for expected events during the course of the clinical trial and we follow mass event rates um on a regular basis as one of the main metrics in terms of of uh overall execution and we're in line with estimates i see so you you feel quite confident it sounds like the events will trigger sufficient events will trigger in q4 no delays everything that we see today we anticipate no no difference from our estimates which would support a 15 month readout okay yeah we're about two-thirds into the study now andrew and everything is is very much on track all right good and so there are two sub studies to this phase three study you upsize this phase three

Andrew Tsai, Analyst — Jefferies

um uh how are you dividing so for the u.s it sounds like the agreement if i remember correctly The U.S. would like two sub-studies. How are you dividing the sub-studies exactly?

Douglas Love, CEO

Yeah, just for those who may not be familiar so you can understand the regulatory package that we're going to bring forward, we have two agreements. With Europe, it's a single protocol study, 659 patients. We're exceedingly overpowered for that analysis. In the U.S., it's the same single protocol but broken out into two sub-analyses. The populations are split, pre-specified, in alignment with the FDA. we powered the study at the sub-study analysis. So when Dr. Clark refers to a phase three level powering greater than 90% in each sub-study, which is really important for us. And so we've broken up the sub-studies. They're stratified by sites and various characteristics as it relates to patients to ensure that both sub-analyses have similar patient populations. We think this is a much more efficient way to do this. It was actually proposed to us by the FDA itself, and it certainly increases our PTS or reduces the risk that you will have two analyses with differing patient populations. Both analyses will have substantially similar patient populations.

Andrew Tsai, Analyst — Jefferies

Got it. And if one sub-study hits, the other one doesn't, hypothetically speaking. It sounds like you'll still try to file in the U.S., though?

Douglas Love, CEO

Well, first and foremost, we will do the single-study analysis. So that will, from our perspective, establish substantial evidence of effectiveness. If we were in a circumstance where one of the sub-studies did not hit, we would probably use the one sub-study that did hit as supportive evidence of the overall analysis. Of course, we'll have to look at the totality of the data package. We'll look at other endpoints like impact on the EZ analysis, low luminous visual acuity, We'll make that determination then. But we do think there are multiple ways to win on this study.

Andrew Tsai, Analyst — Jefferies

So, and lesion growth, you know, you didn't necessarily see it in phase two by month 12. So when would you expect lesion growth?

Lloyd Clark, Other

So we did not see a protection against lesion growth that was statistically significant in the phase two study. However, it's important to note that in the final six months of the phase two study, we saw a 10.5% reduction in RPE lesion growth in the monthly treated group compared to sham. we will see a further protection of rpe lesion in a 24-month study and that boils down to the mechanism of action again by protecting photoreceptors first there is a secondary protective effect of the rpe because they have this bitrophic bi-directional relationship where there's where they support each other in contrast the currently available therapies have a primary effect on slowing down the clearance of dysfunctional RPE cells that have already lost function above them. And so essentially we're targeting this at an earlier stage than the currently available therapies. But we did demonstrate a reduction in RPE lesion growth and we anticipate seeing a further reduction out to 24 months.

Douglas Love, CEO

Let me just append to that and provide a little context. And the reason we are studying RPE in this program is solely for the investor community. Neither of the regulators in the U.S. or Europe has asked us to study the RPE endpoint. And the reason for that is it was a hypothesis that if you protected RPE sales, which provide trophic support under your neurons, you would ultimately protect your neurons and lead to vision. We now have five years worth of data that show that by protecting RPE sales, you're not leading to vision because you're not able to protect the photoreceptor neurons. The reason for that is you actually lose your photoreceptor neurons before you lose your RPE cells. So RPE assessment is a lagging indicator for geographic atrophy. No one made a mistake there. Technology has advanced. When the field started in the RPE storyline, there was one imaging technique that allowed you to see RPE cells but did not allow you to see what was going on with the neurons in the eye, the photoreceptors. Now with the advent of OCT-EZ, you can plainly see, this has been published by multiple institutions, not an exon, independent of us, that you lose your photoreceptor neurons before you lose your RPE cells. So it's not a regulatory relevant endpoint force in our study, but as Lloyd alluded to, by protecting the photoreceptors, they strengthen, which then strengthens your RPE. and you can look in the second six months of our study where the rate of benefit or protection of lesion growth on the RPE was three to four times greater in the second six months of our study. So we're protecting the photoreceptors. The RPEs begin to get stronger over time. So we anticipate we will see that by month 24. We'll provide that to the investor community, but it will not be part of our regulatory packages because it hasn't been requested by the regulatory bodies.

Andrew Tsai, Analyst — Jefferies

Understood. And the study, let's just say you did hit on month 15. the study still remains blinded or masked out to month 24 for you to generate longer term data correct to be clear okay that's correct very good and then on the safety side of things uh it sounds like you're not expecting any retinitis vasculitis well we didn't see any in the phase two we saw no evidence of occlusive vasculitis in the phase two we also saw no increased rate of choroidal neovascularization in the phase two study which is of critical importance because the currently available therapies have at least a 2x incidence of conversion to wet AMD.

Lloyd Clark, Other

And so those are important signals of differentiation in the phase two. We anticipate that we'll replicate that in the phase three.

Andrew Tsai, Analyst — Jefferies

Oh, replicate. Okay, that's good to hear. And then do you intend to commercialize, if this was approved, do you intend to commercialize GA yourselves, at least in the U.S.?

Douglas Love, CEO

Yeah, absolutely, and we're excited to do so. I mean, if you look at the history of AMD, with the exception of Genentech, All of the companies that have commercialized in the AMD space, whether it's wet and dry, have been smaller biotechs. And you can do that because it's a concentrated physician group. There are about 3,000 retina specialists in the U.S. Two large practice groups control almost 50% of that. And so it's a really efficient commercial perspective. The other aspect of it is we're just highly differentiated. We're differentiated on the efficacy front. We anticipate being differentiated on the safety front. And we haven't talked about route of administration, but maybe you should talk about just kind of the drug itself and from a commercial perspective, why that may be attractive as well.

Lloyd Clark, Other

Yeah, so the drug has a number of advantages. First, it's a small volume injection, much smaller than what's currently available. 100 microliters, which is what the current GA therapies utilize, is really at the upper limit of what's safe to put in the eye. These patients can have significant intraocular pressure elevations, can have problems, frankly, with their vision going black. And so it can be a very disconcerting issue. So having a small volume injection is very, very key for us. Secondly, this is a small antibody fragment in a very non-concentrated solution, very similar to the biologics that we're used to using in the retina space over the last 20 years. And so this type of molecule gives physicians a lot of comfort. We understand what small-sized biologics do in the eye that are not pegylated. And then that's the third piece, is that our molecule does not involve any pegylation for increased durability of action. And as we know, pegylation is certainly a pro-inflammatory component of a number of drugs and has been implicated in severe cases of intraocular inflammation. So a number of product-level differentiation that goes along with the clinical differentiation.

Andrew Tsai, Analyst — Jefferies

Great. And so maybe shifting gears now to a GBS sounds like we can expect five to ten patients worth of Western patients of data in the Ford study I don't know it sounds like before the GA data reads out certainly yes okay and would you same with the filing or not necessarily I'm sorry would you expect to file for GBS in the US before GA data reads out in Q42 I won't give the exact timeline but but we expect to do so in 2026.

Douglas Love, CEO

We're encouraged by this forward data, as I said before. I mean, you know, 90% of our patients show benefit at week one in our phase three program. So that's a pretty high bar to replicate here in the US. It's an open label study. We have a good line of sight into what's going on there. I mean, we look forward to getting that data out. Got it.

Andrew Tsai, Analyst — Jefferies

And at the end of the day, what do you want to see in that data set one more time? I mean, it sounds like you're evaluating, following out patients technically out to, I think, 26 weeks, but do you need to wait that long, or how does this all work?

Douglas Love, CEO

It's a good question. Yeah, it'll be a mix of patients and how long they have been since treatment, but it's important for us to demonstrate a few things. First, PK and PD, and the consistency of that in the patients in the West with the patients who are treated in Southeast Asia. Antibodies perform the same across all geographies, so we expect that's going to be quite good. Secondly, we do want to show the efficacy. we will look at multiple time points for this release. Certainly week one, which is the most prognostic factor for how patients will do over time. We will translate week one into week four and to week eight on the GBS disability scale. We think that's important. And for a handful of those patients, we have data out to week 26. Week 26 is actually quite important. When you look in our phase three study, we had a two and a half times greater likelihood to being fully recovered by week 26. And so we wanna show as much the longer-term data as well. Got it.

Andrew Tsai, Analyst — Jefferies

Yes, and so bigger picture, GBS market opportunity. How many patients suffer from GBS each year? What's the pricing? What could it look like?

Douglas Love, CEO

Yeah, GBS is one of the unique rare diseases where it has indeed a blockbuster opportunity. So about 8,000 patients a year get GBS in the U.S., another 15,000 a year in Europe. Importantly, 90 to 95 percent of these patients get treated currently with IVIG or plasma exchange off-label. Everyone, in effect, gets treated, and that's because it's such a debilitating disease. Day three, you can't work a remote. Day five, you're on a ventilator, so everyone gets treated, and that's a really nice pull from a population perspective. From a commercial perspective and a pricing perspective, one of the key findings in our study as well as what we're seeing in Ford is the huge value benefit we're providing to this population. We've done significant health economics work. We've been presenting it at multiple conferences now over the course of 2026. We'll continue to do that and publications are on the way there, but it costs the U.S. healthcare system greater than $7 billion a year to manage 8,000 patients. So it's an incredibly, incredibly expensive disease. We're getting patients on their feet 30 days sooner off a ventilator 28 days sooner out of the icu 10 days sooner so excuse me these are hard dollar savings which gives us pricing flexibility there we've not released pricing yet the covering analysts have somewhere between 150 100 to 150 000 of course therapy not unreasonable when we talk to many of the sites the practice groups and the payers They kind of analogize to CART-T-type therapies, which is in the neighborhood of $500 to $2 million. We won't go that high, but we do think we have some pricing flexibility with this. And so if you can imagine 90% of patients getting the drug at a reasonable price, it quickly turns into a blockbuster opportunity.

Andrew Tsai, Analyst — Jefferies

Yeah, and interestingly, or maybe ironically, you know, U.S., there has been some kind of, you know, you're trying to figure out with the FDA, but hopefully this data set can work things out once you file. EU, though, you've filed, and ironically, IVIG is approved in the EU. And it seems like things are going well. So this could actually be approved, I guess, early 2027 in the EU.

Douglas Love, CEO

That's our current timing, first part of 2027 for approval in the EU. We had a higher bar from a regulatory perspective in the EU, given that IVIG is approved. It's approved on a compendia, not because they ran studies. But nonetheless, it's approved and being reimbursed there. But Europe's been fantastic. They understand the disease well. There's a ton of expertise in Europe with regard to GBS. We've met with country-specific folks. We've met with the rapporteur. We're on file. We're in active engagement with Europe. And so we're really pleased with the way that's going. And would you intend to market GBS in the U.S. and EU yourselves? EU is a TBD. We definitely are intending to commercialize in the U.S. We like the opportunity tremendously. And the reason being for that is there's about 50, again, similar to GA. There's about 50 large practices in the U.S. that see about half of the GBS patients year over year now for seven years. It's a population-based disease or incidence-based disease, so you know where they are. That's a very efficient footprint. You don't need a lot of sales reps. In fact, hospitals don't want to see sales reps, so it's field managed care, MSL types. Europe is Europe, right? And whether we capitalize and commercialize on Europe is still to be determined. We've had a fair amount of inbound on the non-dilutive side for ex-US deals. We're working through that now and just how that may relate to the MFN topic. And so that will help inform what we do with regard to that, as well as just ensuring that we get value if we do partner this ex-US.

Andrew Tsai, Analyst — Jefferies

And then lastly, on the small molecule program, also data, it sounds like, sorry, is it before GA?

Douglas Love, CEO

We anticipate that before GA, correct.

Andrew Tsai, Analyst — Jefferies

Before GA. And so earlier you said possibly six patients' worth of data, and what is success to you? What is that no-go threshold?

Douglas Love, CEO

Yeah, up to six patients is what we're targeting for this program. So early-stage program, translational stage is the way we think about it. For us, we really want to see, as I said before, normalization of complement. We're in codiglutinin disease where complement is arguably as high as in any disease that you would study, certainly higher than the neuromuscular programs. And so we're looking for complete suppression of C1S in that target versus 75% or 80%, which you see in other neuromuscular diseases. So that's one that's really important. We do want to see patients normalized on bilirubin, which is a key measure of kind of anemia in the disease. And one thing to note, given that there is an approved therapy for codiglutinin disease out there now, we're having to wash patients off their approved therapy, wait until they go into a hemolysis, and then treat them. As a result, what we're seeing is they're not as severe as a treatment-naive patient, so our aim is just really to normalize them back to their baseline versus a comparison to some third-party data set, if you will.

Andrew Tsai, Analyst — Jefferies

I see. And will the data also have hemoglobin, and do you expect to see?

Douglas Love, CEO

We are looking at hemoglobin. Hemoglobin is a bit more variable, not just in our program but in all programs, of codaglutinid disease, and oftentimes can take a bit longer. This study is four weeks, so we will look at hemoglobin as well and provide whatever information we see there. We think that's probably more directional. Bilirubin is certainly much more objective as an endpoint.

Andrew Tsai, Analyst — Jefferies

And the latest iteration of the formulation of this drug, you found out maybe late last year, I forget. Anyway, you found out that, you know, you're not supposed to be taking this with food. But is that going to be the formulation you take forward in phase three, or are you trying to solve this?

Douglas Love, CEO

Really good question. So, you know, what Andrew's referring to is we did see a food, have seen a food effect in this proof of concept study. This is a tablet that has an enteric coded with the aim of getting the drug past the gut and then to release into the system. We're finding that the enteric coding has been underperformed from our perspective. And so we're seeing in some instances where it falls off, patients actually have the drug release in the gut and they have a food effect. What we're doing with this last cohort of patients is, and that'll be less than six. They're fully fasted. It's BID dosing within an hour or two of the meal. There, things seem to be working quite nicely. The question will be, how well does that work when we pull all of the data together? Do we move forward into a late-stage study with the current formulation, or do we stop and solve the interocoded formulation, which seems to be a reasonably straightforward solve on our end? We'll make that determination once we roll the entire data set up.

Andrew Tsai, Analyst — Jefferies

Got it. I think that's all the time we have but it's great to know that you have three important updates later this year so thank you