Welcome to the Cantor Fitzgerald Global Healthcare Conference. I'm Pete Stavropoulos, a biotech analyst with Cantor. With us, we have Annexon, a company I cover, and pleased to introduce Doug Love, the CEO, and Lloyd Clark, VP of Ophthalmology Strategy. Welcome. Let's start off with a brief introduction of yourselves and snapshot of the company.
Yeah. Thanks, Pete. Thanks, Cantor. Happy to be here. I think this is our eighth year in a row, and always delighted to join this conference. I'm Doug Love, as Pete said, CEO of Annexon, founding CEO. I've been in the chair the past 11 years, and we're excited about the progress of the company since then. For those who are less familiar with Annexon, we are a complement-focused company, really stopping neuroinflammation right where it starts, on disease tissue in the body, brain, and eye. Over the course of 11 years, we've advanced multiple drug candidates. We're happy to be here today with two late-stage programs, one in Guillain-Barré syndrome, where we were successful in a large phase III pivotal study and are now on file in Europe and will file shortly in the U.S.
Following that, we have a large geographic atrophy program, the only program in the world to demonstrate significant vision preservation, both on all of the measures of visual function as well as significant protection of photoreceptor neurons responsible for visual acuity. So we're in a large phase III program there, which we'll read out in Q4. We're excited about what we've done over the last 11 years and more excited about what's ahead for us. Our mission is really to impact millions of patients worldwide living with devastating neuroinflammatory diseases for which there are no disease-modifying therapies. We're right on the cusp of doing that. So joining me today is Dr. Lloyd Clark, our SVP of Ophthalmology Strategy, and I don't know if, Lloyd, you'd like to just quickly introduce yourself.
Thanks, Doug. Pete, thanks for having us. 25-year career as a practicing retina specialist, extensive experience on the drug development side. Joined Annexon about a year and a half ago as we build out an ophthalmology business unit. Really excited to be here and excited about our opportunities with vonaprument.
All right. Look forward to hearing what you have to say. Let's start off with the geographic atrophy. There are two approved drugs, both complement inhibitors. In a couple of sentences, how do you think about the unmet need in GA, and what can vonaprument bring to the table that these other complement inhibitors cannot?
Yeah. Unfortunately, the unmet need for geographic atrophy is immense. This is a disease with 8 million patients worldwide, leading cause of blindness, and the two first-generation drugs that are currently on the market really provide symptomatic benefit to the disease. That is, they are protecting against a structure in the eye that has not demonstrated protection of visual acuity. The opportunity is to do just that, to protect vision in patients with geographic atrophy with our therapy.
All right. In May 2023, I remember the day very clearly, you disclosed outcomes from the phase II ARCHER study. Outcomes were a bit surprising, but when you zoom out, the outcomes were a lot better than expected. Walk us through the initial data that suggested functional benefit.
Yeah. Two components to the study. One was an aspect of structural assessment on RPE cells, which have historically been thought of as the structure in the eye that was necessary to protect to ultimately demonstrate visual preservation. There, we were not statistically significant on that particular endpoint. We did show a nice trend, which increased over the course of the 12-month study. However, what we did do in the study is demonstrate significant vision preservation, importantly, on best-corrected visual acuity 15 letter loss. This is the gold standard endpoint for visual acuity, and we demonstrated a dose-dependent response there on that endpoint, as well as its sister endpoint, low luminance visual acuity, or LLVA, and then LLVD, et cetera.
In other words, we were statistically significant on all of the functional assessments of visual acuity in this study, all of which was dose responsive, which we think is really, really important. As I said before, associated with that, the how we ultimately demonstrated that, and that was by having significant preservation of the photoreceptor neurons in your eye that are responsible for visual acuity. It is important to note that your photoreceptor neurons are what are lost first in geographic atrophy. Following that, you lose your RPE cells underneath that. By protecting your photoreceptor neurons, not only are you strengthening the overall neuronal unit, which includes your RPE cells, it does translate to visual outcomes. From a safety perspective, the drug performed really nicely from that perspective as well.
All right. The phase II, you also showed a 73% reduction in the risk of greater than 15 letter loss at 12 months. How compelling is this magnitude of effect?
Maybe I will bring that over to Lloyd as a treating retina specialist.
Well, first of all, I think it is important to understand that the endpoint is the most compelling piece. Losing 15 letters of vision typically renders a patient unable to drive, unable to read small print. This gold standard endpoint, confirmed 15 letter loss, first of all, is incredibly important to think about, and any statistically significant difference in that endpoint is highly clinically relevant for patients. Now, specifically, the risk reduction is incredibly important. That is a very, very easy number to talk to patients about. To be able to cut the likelihood of losing that clinically relevant endpoint by almost 75% is incredibly compelling data.
Or maybe said differently, three out of four patients had their vision protected in our study, and that's really a really high bar.
All right. An interesting observation, and I think you brought it up, Doug, a minute ago, is that you did not demonstrate a statistically significant reduction in RPE loss over a full 12 months. However, treatment effect improved and strengthened in the second six months of the study. So what does FAF measure in terms of cell types, and is there a high correlation with visual function and changes in the FAF?
Really technical question.
Yes.
Let me start there, and I'm going to get Lloyd in on that. FAF is a measurement of fundus autofluorescence of RPE health underneath the neurons. And so the hypothesis was that if you protected these RPE cells, they're a surrogate for ultimately functional vision protection. What we now know, not based just on our data, but the entire field, is that by protecting RPE cells, that does not translate to functional benefit. But maybe, Lloyd, you can talk about it and the mechanism.
Well, in terms of mechanistically, what we know about blocking C1q is that you are specifically protecting photoreceptors first. That is key to the understanding of the differentiated mechanism of vonaprument and its ability to block C1q, which has a specific recognition function on photoreceptors under stress. There is also a bidirectional relationship between photoreceptors and RPE cells in the eye. If you protect photoreceptors, then downstream, you will protect the underlying RPE cells. That is consistent with what we saw in the clinical studies. We saw a marginal protective effect in the first six months, but we saw a 10.5% reduction in RPE lesion growth in the second six months, consistent with that mechanism of protecting photoreceptors first, leading to RPE preservation. The mechanism adds up with the endpoint that we saw on structure.
Now, for us, RPE lesion growth is not a central part of our program because, again, our structural endpoint are photoreceptors as measured by ellipsoid zone. But we do understand the interest in this endpoint given the historical importance with the currently available drugs.
All right. Would you expect that difference to continue to widen?
We do. We do expect that to widen based on our observations in the phase II study, as well as our understanding of the mechanism of why that is the case.
There's another anatomical measure, ellipsoid zone, which you just mentioned. Can you describe what this is and how important of a marker is it when assesses retinal health in GA, and what's your take on vonaprument's effect on the structure?
Yeah. Ellipsoid zone is a measure within the OCT imaging that's performed on patients at every visit. It's an overall measure of photoreceptor health, and so it's a direct measure of the integrity of the photoreceptors. And photoreceptors is where vision lives in the retina. So it's a more meaningful endpoint in a neurodegenerative disease. It's a more meaningful anatomic endpoint for a drug that demonstrates a functional benefit. So pay very close attention to ellipsoid zone data as our data is presented next year to understand the biologic effect of our drug because we think about ellipsoid zone data in terms of its effects on the ARCHER II program as a measure of biologic activity of the drug.
I'll just also add on, just from an industry perspective, it's where the field has shifted to.
Yeah.
Over the last 10 years, prior to that, RPE was the storyline. If you look at oncoming studies that are trying to have a neuroprotective approach, the agencies in both jurisdictions in the U.S. and in Europe now recognize EZ as the measurement for protection from a neuroprotective perspective. Whereas RPE has become just a luxury or an excess endpoint to add into your programs.
Yeah. We did also a KOL call, and we asked him about this structural measure, and he said it is becoming much more relevant. They are shifting away from the FAF. How do these data compare to, I think, SYFOVRE also measured EZ zone, especially in the central subfield? How do your data compare to theirs?
Yeah. Maybe I will start, and then Lloyd, please dive in on this. There are two elements that you want to measure with EZ. You want to look at your impact on photoreceptor health across the entire eye or entire retina, the pan-macular, if you will. Then more specifically in the central fovea or central retina. Why that is important is geographic atrophy is a disease where you lose central vision. You lose the ability to make out faces, as Lloyd said, to drive cars, et cetera. So you want to look at photoreceptor health in both areas. When you look at the entire pan-macula, we were very similar to SYFOVRE, or they are very similar to our data with regard to that.
However, when you look in the central fovea, where the disease emanates from, we had 50%-60% protection by year one with our approach, which are whopping effects. No other therapy has shown that, and it really does evidence why we are able to show consistent dose-dependent vision preservation on every one of the endpoints we looked at from a functional perspective.
Nice.
Yeah. That is it.
That's great.
Nothing to add.
All right. Does preservation of EZ provide evidence that C1q inhibition is directly protecting the photoreceptor synapses rather than simply showing expansion of atrophic lesions?
Well, it's important to understand the mechanism. C1q specifically binds to photoreceptor synapses, and that's what differentiates it mechanistically from downstream complement inhibitors that play more of a role in cleaning up dead and dysfunctional cells. Preservation of ellipsoid zone demonstrates preservation of photoreceptors, and that is specifically due to blockage of C1q, which has this very specific mechanism which clearly differentiates itself from other complement agents.
Okay. So vonaprument demonstrated vision preservation across multiple baseline patient characteristics. Does this suggest that the biology of C1q inhibition is applicable across a broad population and GA population, or do you anticipate particular patient subgroups to derive better g reater benefit.
I would point you to two specific subgroups that we've identified in the phase II data. The first are patients with foveal involvement of their RPE. Those are patients that are at high risk for rapid vision loss. Those patients had a substantially increased effect of vonaprument compared to patients that had non-foveal lesions. Those are patients on the precipice of losing vision. Those are patients that you would consider to be more severe from an anatomic perspective. In contrast, there's also patients that have what we would call less severe photoreceptor damage globally. Those are patients that have a low light visual deficit of less than 30. That's a generally accepted cutoff. These are patients that have less broad photoreceptor damage. In that group, we saw no 15 letter vision loss events in patients treated in the monthly group compared to 17% in the sham group.
On both ends of the spectrum, we see a significant clinical benefit of vonaprument, both patients with foveal disease, which demonstrate vision at the risk of being lost very rapidly, as well as patients with less severe disease. We think there's broad application, and we're in particular very encouraged by the data in patients with less severe disease.
When you do speak to your consultants, your KOLs, how compelling do they find the totality of the phase II data?
Well, I think that what's most compelling in the data, I think for my colleagues that are in the retina space, is the fact that there's, across multiple different measures, there's directional dose-dependent vision protection in these patients. Not just one finding, not just in a certain subgroup, but across a myriad of different measurements. It's consistent. I think that that gives my colleagues a lot of confidence. In addition, I think the other piece is the off-treatment effect.
Yeah.
When you see how well patients did during the first 12 months of treatment, then in that final six months of the phase II study where everyone went off of treatment, there's an immediate return and increased number of vision loss events. Those two factors, I think, play into the confidence of this molecule as we move through the phase III program.
We have evidence of that confidence, right? When we ran the phase III program, we were able to recruit that program six weeks ahead of schedule and over-enroll it by 30 patients. It speaks to the demand for this type of an approach and the understanding by the retina community of the opportunity here to protect vision.
Okay.
That's the phase II enrollment or phase III?
That's the phase III enrollment.
Okay.
Correct.
All right. You are currently in a phase III, the ARCHER II study. It's underway. Just go over the study design, the primary endpoint, and then we'll get into some details.
Yeah. So pretty simple study design. We have two groups. We have patients treated monthly with vonaprument versus a sham group. We have 2:1 randomization in favor of the patients under active treatment. A total of 659 patients were enrolled in the U.S. and Europe. The primary endpoint is the percentage of patients that have confirmed 15 letter loss at two consecutive visits. If you have a 15 letter loss event at month six, it has to be followed up at month seven with a confirmatory visit. We now have dual primary endpoints of both month 15 and month 24. So we'll be evaluating efficacy at both time points given our recent revision to the clinical trial.
What were some of the key learnings from the phase II that you've-
Yeah, a couple of important learnings. First was, we knew that we wanted to replicate the cohorts that we saw in the phase II. For example, we had a target foveal enrollment rate consistent with what we saw in the phase II, and we have matched that. You will see that at an upcoming meeting. Otherwise, we matched the cohorts from phase II into phase III, except for one key learning, which was that we eliminated patients with very poor vision in the phase III study below 45 letters because those patients did not have 15 letter loss events. That was a key learning that shaped our enrollment criteria for the phase III.
In two consecutive visits, they need to have vision loss. What is the general variability from month to month, let us say?
Oh, you can have quite a bit of variability. What that second visit does is it eliminates the noise associated with variability in visual acuity measurements in retina clinical trials in general, but in particular in patients with dry AMD. It is really important. You have roughly 50% of patients will have a 15 letter loss visit that is not replicated at the next month.
How many?
Up to 50%.
Up to 50%.
But with the second visit, when it is confirmed with the second visit-
They do not come back.
virtually they never come back. It is really confirming. It squeezes, as Lloyd said, all the variability out of the assessment.
Okay. Can you just walk us through the powering assumptions for the study and what exactly you powered to show?
Yeah. We have parallel statistical plans. We have a single study of 659 patients. In the U.S., we have a statistical plan for filing that involves two sub-analyses, splitting that in half. When you look at the two sub-analyses, we are powered at a standard phase III greater than 90% level at the sub-studies. As you can imagine, we are significantly overpowered in the single analysis that we will use for filing in Europe.
All right. Assumption on sham rate?
Yeah. We used the data from our phase II study to inform our estimates for sham event rates, combined with a myriad of other publicly available data points. For our treatment effect, we largely used the treatment effect from the monthly group in the phase II study. From both of those estimates, we took conservative adjustments down and down on the sham rate, up on the treatment effect, to establish what we thought was a conservative treatment delta, and then built the study around that delta.
Can you provide us the details of
We've not provided those publicly.
All right. Have you provided the dropout rate or retention rate?
Yeah, that's one thing we're very encouraged about. We've not shared in a public forum the dropout rate, but after 12 months, we are well below what we anticipated based on our experience in the ARCHER II study, as well as what we anticipated in the registration studies for other medications, significantly lower. We're very encouraged by the excitement in the community and the retention.
What we have said with regard to the dropout rate, it's less than 10%, so exceedingly high or seemingly low, as Lloyd said, and our compliance rate greater than 90%. Both of those are really high landmark outcomes for us in the study. The study's really well run. We're really encouraged by that.
All right. How did you drive the high compliance rate?
Well, the patients
you need to get into the-
physicians believe in this opportunity, the patients believe in this opportunity, and the drug really performs well. One of the things that's nice about vonaprument is its dosing approach. It's a really very limited viscosity. It's non-pegylated. It's a very simple administration, and it's very tolerable. Then obviously, there's the hope that we're going to replicate what we saw in the phase II. Patients are showing up across the globe in the U.S. and Europe and getting every single dose, every month. As I said, it's greater than 90%, and that's fantastic. Better than what we anticipated.
Touched on it before, but I guess in August you disclosed that you're basically going to take two bites at the apple, with a 15 and 24-month assessment. Just walk us through the details of that change and especially from a powering perspective and the rationale for actually going to 24 months.
Yeah. Before we get into the details, maybe just set the stage for what we're doing and why we're doing it. As you asked, we've added a second endpoint into the study with month 24. This is a separate endpoint, independent of month 15, so we applied separate alpha to this endpoint to win at either month 15 or month 24 or both. We anticipate winning at both. Important to note that month 15 continues to be our base case. A lot of reasons for why we're really excited about where we are with month 15. First and foremost, we've recruited a very similar patient population as the phase II. We'll be releasing baseline characteristics in a couple of weeks at a Retina Society meeting where you all will see that. Want you to pay attention to a couple of things there.
First, note how consistent the patient profiles are with the profiles we saw, what we recruited in the phase II study. Secondly, really important to look at where vonaprument had really outsized treatment effects. If you look, for example, and Lloyd alluded to this before, if you look at the patients who had LLVD less than 30 in our phase II study, zero out of 56 patients lost vision, whereas 17% did in the sham arm. That is a group that we were really particularly focused on in the phase III study, and then the foveal population that Lloyd alluded to. Patients on the cusp of losing their vision, which represents a sizable aspect of the phase II study, also in the phase III study. We like the consistency in the being able to replicate what we did in the phase II study.
Then, of course, we are tracking masked events, that is, loss of 15 letter loss month-over-month over the course of the study, and it very much is in line with what we projected for this study. Month 15 is our base case, and we have maintained that. Because the way the study was executed by the team, they did a fantastic job with that. We picked up power over the course of the study, and we deployed that extra power towards month 24, giving us an opportunity to win at month 24 while not impacting or prejudicing the month 15 endpoint. We really like that. From a commercial perspective, having month 24 in your data helps immensely from a payer perspective and a reimbursement perspective. The physicians in the community absolutely love month 24, and it allows us to build, in effect, a moat around this asset.
Anyone who is going to follow this program, developing their own C1q program, are going to have to run 24-month studies to beat us, and that is going to be a hard thing to do. We think we are building an asset where we have IP out to the 2040s that is really formidable, an opportunity to drive immense value for the company over time. Maybe, Lloyd, just talk about—
Yeah
—the powering piece.
Yeah, so briefly, we picked up powering through the execution of the study, through over-enrollment, through increased patient retention, and by extending the primary time point from month 12 to month 15. That gave us a larger cohort of patients with more time to demonstrate more events. That gave us XX power. How were we going to deploy it? We made the decision to deploy it at a second endpoint. From there, we could utilize time, another nine months, for an increased treatment delta. So we're using a combination of more patients in the early part of the study and more time at the later part of the study to generate alpha at month 24 without disadvantaging month 15. So it's a very creative strategy and one that I think it has the potential to reap significant benefits.
All right. So just to make sure I understand correctly, no matter what, even if you hit at 15 months, you're still going to continue the study in one way or another blinded, out to 24?
Well, we're required to do that.
Okay
In terms of safety. Yeah, absolutely.
That is a requirement.
But-
For all GA studies. That is the standard.
Okay.
Yeah. All right. What will the DMC see, in 4Q 2026, and what will you disclose? Will you just disclose that you will continue the analyses of the two subgroups
Right
or are you going to give an overall outcome or just a P value?
In Q4, the DMC will make a determination over the single large analysis. Assuming we were to win on that, then that would be the creation of the sub studies. Then shortly after that, the DSMC would then meet and evaluate the sub-analyses. At a point where we win on all three analyses, that is when we would unmask the study as if we had not added the second endpoint and be prepared to file. We will continue to function as a masked study in terms of patients and in terms of sites out to month 24 due to safety. But if we have all three positive analyses, we will unmask the study, present the data, and prepare for filing. Anything short of that, the communication will be that we will continue to month 24.
It is important to note the way we are handling the data at month 15 is to ensure data integrity out to month 24. To do that, the company has to be masked to the data, and the communication has to be really limited. So we will not be reporting P values at month 15. It will be, we have won, or the study is continuing on. The DMC could also report out that the study is futile. We think that is unlikely given that we have-
There is a futility analysis.
There is a futility analysis.
Sure.
They've been able to assess for futility over the course of the study. We haven't received that, and we don't anticipate it, but they absolutely have the opportunity to do that. We will then, as Lloyd said, report out win or go forward, win or go forward, and that's the way we'll do that.
All right. Let's say, assume they say go forward, meaning go out to month 24, how should we interpret that? Does the first look also, you just mentioned it has a futility analysis.
Yeah, I think the interpretation is we're around the hoop, and with more time, as Lloyd alluded to, you give yourself two things. You give the opportunity to have more event losses on the sham arm, which if you look at our phase II data, after month six, you begin to see the curve separate. We'd expect that it will continue to separate after month 12, after month 15, all the way to month 24. In fact, if you look at historic GA studies that go out to 24 months, it's exactly what you see. You also give vonaprument a longer period of time to have a treatment effect. You're winning on both sides of the barbell, if you will. We think that's actually a very positive thing.
All right. Just quickly, market opportunity, where exactly do you see this fitting in?
Yeah, immense. We think it's a replacement therapy for the current therapies out there. All patients are focused on their vision preservation, right? If you're coming in to get treated at a vulnerable time in your life where you're seeking to maintain your independence, any therapy that is going to maintain your vision, if not enhance it, you're absolutely going to dive in on it. We're going to make sure we get this drug available to patients worldwide.
It's also important to recognize that only 20% of patients with geographic atrophy are currently being treated today by only 20% of the retina specialists. There is a significant opportunity for market growth, for physician engagement, for patient activation with a drug that has a functional benefit. Yeah. Like I said before, we hosted a KOL call, and we've spoken to other KOLs, and same thing. If this actually replicates the phase II, right to the front
There we go
of the pact. So.
You said it more succinctly.
Thank you. All right, so last question. I know you have the GBS program. We obviously don't have time.
Oh, we can get it in, Pete.
We'll just say quickly, very pleased by that. I think you all saw the open label data that we shared with GBS.
Yes
About a month or so ago. Couple things to note. 100% response rate at week one. This is not a mild response. These are patients getting out of bed within one week and returning to their lives out of the hospital. It has never been seen before. Highly consistent with what we saw with our phase III data. We had a 90% response rate. That is really important from a filing perspective, supports what we have done with our filing already in the EU, and it really ensures that we will file with the FDA by the end of this year. We are very excited about this GBS program.
Looking forward to that. Last question. We are sitting here 12 months from now, what would you like to say have been the key value-creating accomplishments?
Well, everything, right? We anticipate winning on geographic atrophy and vonaprument at month 15. We anticipate winning at month 24 as well, and then GBS making sure that it is approved by that period of time and launched into the marketplace. Our mission is to serve millions of patients suffering from devastating neurodegenerative diseases, and we expect to be well in our way of achieving that mission at that point.
Well, Doug, Lloyd, thank you very much for joining us
Thank you.
and participation in our healthcare conference as well as Jen. So look forward to the progress, look forward to the Q4 readout, and pretty exciting times.
Thank you so much, Pete. Really appreciate it. Thank you all.