Annexon, Inc. (ANNX) Earnings Call Transcript & Summary

Hello, and welcome to Annexon's conference call. At this time, all participants are in listen-only mode. There will be a question-and-answer session after the prepared remarks. Please be advised that this call is being recorded at Annexon's request. I would now like to turn the call over to Lori Murray, Senior Vice President, Investor Relations and Corporate Communications. You may begin.

Thank you, Jonathan, and good afternoon. Earlier today, we issued a press release outlining top line data from our ARCHER Phase 2 clinical trial evaluating ANX007 in patients with geographic atrophy. The press release and the slides to which we'll refer are available under the Events and Presentations page within the Investors section of our website. On the call with me today is our CEO, Douglas Love; as well as our Chief Innovation Officer, Dr. Ted Yednock; and our VP and Head of Ophthalmology, Dr. Donald Fong. In addition, we are delighted to have Dr. Charles Wykoff join us for the Q&A session. Dr. Wykoff is the Director of Research at the Retina Consultants of Texas and an investigator in the ARCHER trial. Today's discussion will include forward-looking statements related to Annexon's current plans and expectations, which are subject to certain risks and uncertainties. Actual results may differ materially due to the various important factors, including those described in our Risk Factors section of our most recent Form 10-Q and other SEC filings. These forward-looking statements represent our views as of this call and should not be relied upon as representing our views as of any subsequent date. We undertake no obligation to publicly update these statements. Now with that, I'd like to turn the call over to Doug.

Thank you, Lori, and thank you all for joining us today. We're excited to report the results of ARCHER, our Phase II clinical trial of ANX007 for the treatment of geographic atrophy, which demonstrated, for the first time, a statistically significant protection against vision loss across multiple endpoints in this patient population. These data are the realization of our founding hypothesis that we can treat complement-mediated diseases differently by stopping the classical complement cascade where it starts on disease tissue. In the case of GA, that C1q localized on photoreceptor synapses, resulting in damage and loss to photoreceptor cells, and ultimately, visual impairment. We'll come back to this concept. Our scientific founder, the Late Dr. Ben Barres, a pioneer in complement mediated neurodegeneration, trailblazed insights that fundamentally shifted the world's understanding of how our immune system can disrupt normal neuronal function in our eyes and brain. Ben hypothesized that by blocking upstream classical activity via C1q, we could provide the most complete protection against harmful inflammation and tissue damage in the body, DI, and the brain. This led to the founding of an Annexon. And since then, we've made significant progress towards translating the scientific hypothesis into therapeutic reality. Ben's original discovery was NDI, where he discovered that C1q led to aberrant removal of neuronal synapses or connections and tissue dysfunction. Over the last 2 years, we presented clinical data in 2 difficult-to-treat diseases, which support that hypothesis. Huntington's disease and ALS, where we showed that by stopping the complement cascade at its start, we can provide meaningful clinical outcomes for patients. Today, we're pleased to report the third supportive data set in a neurodegenerative disease, geographic atrophy, or GA. The top line data from our ARCHER Phase 2 trial show that ANX007 is the first complement agent to demonstrate protection against vision loss following 12 months of treatment in prespecified analyses. Today's top line data supports 4 key differentiators for ANX007. First, the data validate ANX007's unique mechanism of blocking upstream classical complelemt via C1q to preserve photoreceptor function and vision. Second, a statistically significant dose-dependent protection against vision loss was demonstrated in 2 prespecified measures: BCVA as measured by 15-letter loss, and LLVD, and was supported by multiple analyses, including protection against visual loss in both the foveal and non-foveal populations. Third, the primary endpoint of reduction in rate of lesion growth did not reach statistical significance. We believe the results reflect a unique neuroprotective approach of ANX007 and its distinction from the biomarker measurement of lesion growth, and that RPE clearance may be mediated in part by other complement pathways. Importantly, slowing of lesion growth has not been clinically correlated with digital function. And finally, ANX007 demonstrated favorable safety results in the 181 patients treated for 18 months, with no meaningful differences in P&D rates between the sham and treatment groups. In total, we believe these data strongly support the therapeutic potential and market opportunity for ANX007 by protecting against vision loss in GA patients. By way of background on Slide 5, GA is an advanced form of age-related macular degeneration, or AMD. It's a chronic and progressive neurodegenerative disease of the eye that for more than 5 million people worldwide leads to progressive and irreversible vision loss. GA is the leading cause of blindness in people over the age of 65 in developed countries, and the risk of developing GA increases with age. More than 1 million people in the U.S. have been diagnosed with GA and another 4 million people outside of the United States. A person in the early stages of GA is likely to first notice problems with reading or low-light vision. That's because photoreceptors or light-sensitive cells begin to deteriorate in the center of the eye, causing issues with fine vision and color perception. This damage will start as small spots will eventually grow into larger packages. Ultimately, if the disease progresses to advanced stages, permanent blind spots will develop in the center of the visual field. As the photoreceptor cells continue to die, the person will unfortunately lose their vision and the associated independence that goes with that. GA is a disease of vision loss due to damage and loss of functional photoreceptor cells, a type of neuron. Looking at the diagram in the retina and GA on Slide 6, functioning photoreceptor cells shown in green, have a rich synaptic network that links the retina to the brain for vision. Photoreceptors on the edge of the lesion in this diagram, shown in orange, and in the actual microscope image below, are in clear decline. Annexon's similar discovery by Dr. Barres with that C1q drives elimination of synaptic in disease, leading photoreceptor neurons dysfunctional and susceptible to elimination. In preclinical models, we've shown that blocking C1q protects photoreceptor cells synpases, and importantly, photoreceptor cell function. We have also shown that blocking C1q leaves the normal clearance function of the lectin and alternative pathways in place, which is important for removing dysfunctional cells and maintaining tissue homeostasis. Fungicidal fluorescence, or FAF, measures RPE cells, not photoreceptors. And loss of FAF is a measure of RPE clearance as a surrogate biomarker of lesion growth. Importantly, FAF is not a measure of photoreceptor cells or photoreceptor cell function. The goal of ANX007 as a C1q inhibitor is to protect against the loss of photoreceptor synapses to preserve photoreceptor cell function and vision. We demonstrated this in preclinical models and now in our pivotally designed ARCHER trial. Moving to the trial design on Slide 7. Our ARCHER trial is a randomized, multi-center, double-masked, sham-controlled study, comparing safety and efficacy of ANX007 to sham patients with GA. The study enrolled 270 patients stratified by GA lesion size, location, and fellow eye CNV to minimize biases or imbalances in the data. Patients were randomized to receive a 5-milligram intravitreal dose of ANX007 either monthly or every other month, or sham for a total of 12 months, followed by a 6-month off-treatment period. The primary endpoint of the study was the rate of GA things measured by fungicidal fluorescence from baseline at 12 months. Importantly, the real-world clinical goal, however, is preservation of vision. Given the mechanism of ANX007, the ARCHER trial included multiple prespecified endpoints to assess visual acuity. These are best corrected visual acuity referred to as BCVA, low luminance visual acuity, or LLVA, and low luminance visual deficit, or LLVD. BCVA a widely established functional endpoint by the clinical community and regulatory authorities, and represents the gold standard for assessing efficacy in mini retinal diseases. In a BCVA assessment, visual acuity is tested using a specialized eye chart at a premeasured distance and add a certain luminance. For veracity in our study, we defined vision loss based on 2 consecutive visits in which a patient had a loss of 15 or more letters compared to baseline, or if the loss occurred at the end of the study. This approach is designed to limit variability in vision and reading conditions by patients between visits, adding confidence to our data set. Low luminance visual acuity, or LLVA, involves standard BCVA testing in low-light conditions, often by adding a neutral density filter in front of the testing eye. Finally, low luminance visual deficit, or LLVD, is the difference between BCVA and LLVA and assesses retinal health and abnormalities in visual function that would not otherwise be apparent by measuring BCVA alone. In terms of patient demographics on Slide 8, 270 patients were enrolled in our study and were equally split between 3 arms: 89 in ANX007 monthly group, 92 in the every-other-month group, and 89 in the sham group. The trial was predominantly conducted in the United States with approximately 95% of study patients in the U.S. The overall demographics of the patient population were consistent with other GA studies, including age, gender, and ethnicity. The average age in the study was 80 years old and enrollment was generally balanced between males and females. Importantly, patients were split almost equally between foveal and non-foveal. As background, foveal patients are those with vision loss in the central portion of the macula, which is responsible for central vision. Non-foveal patients for those whose lesions are not yet affecting central vision. With that, let's turn to the ARCHER findings, starting with the most important functional endpoint, visual acuity. Starting with BCVA on Slide 10 -- I'm sorry, starting with BCVA, Slide 10 shows the number -- the total number of patients with a 15-letter or greater loss at the end of the 12-month period for each study group. As shown in the bar chart on the left, out of the 181 patients treated with AMX007, only 15 patients, or 8.3%, experienced a 15-letter or more loss. This means that treatment with ANX007, whether monthly or every other month, was able to protect against loss of 15 letters or more for 166 patients. This result was statistically significant with a p-value of 0.002. We view these data as highly differentiated, especially when compared to the 21.3% of patients in the sham group who experienced a loss of 15 letters or greater at month 12. As noted, each patient's best corrected visual acuity was measured at each of their visits. Shown on Slide 11 is an analysis of a risk -- of the risk of a 15-letter or greater loss defined as loss on two consecutive visits were at the end of month 12. ANX007 reduced the risk of having a 15-letter or greater vision loss by 72% for those treated monthly, 48% for those treated every other month, and 59% for the pool of treatment group. This is compared to the sham group, which dropped substantially over the one-year period. These data were statistically significant in AMX-007 monthly group with a p-value of 0.006, nonsignificant in the every-other-monthly group with a p-value of 0.064, and statistically significant in the pool population with a p-value of 0.008. Over the last few years, there's been a significant amount of discussion in the industry on how patients might respond to treatment based on the location of their media. Historically, it has been thought that foveal patients would not be able to benefit from a complement-mediated therapy, whereas non-foveal patients have been expected to see a greater benefit from treatment. We looked at a 15-letter loss for the monthly group compared to sham separated by foveal and non-foveal populations, and the treatment effect was statistically significant at 0.0019. We will look at the same analysis for the every-other-month treatment group, the treatment effect was also statistically significant with a p-value of 0.04. ANX007 pool population also achieved statistical significance with a p-value of 0.0015. In sum, we saw a meaningful treatment effect in both the monthly treatment and every-other-month treatment groups regardless of whether patients were foveal or non-foveal, which we believe strongly supports the therapeutic opportunity in an all-comer population. Moving to the next assessment of visual acuity on Slide 13. We looked at LLVA, which is BCVA assessed in low-light conditions. In this analysis, we observed at 20.3% of patients in the sham group had a 15-letter or greater loss versus 10% for those in the ANX00-monthly group, and 11.8% for those in the every-other-month group. While none of these reached a statistical reduction, this is the first geographic atrophy study that we are aware of that showed a trend of slowing the progression of LLVA. On Slide 14, we show patients BCVA stratified by their retinal health status at baseline. For this assessment, we use the industry standard cutoff of 30 LLVD score. LLVD is a prognostic factor on the health of patients' retinas and evaluation assesses retinal function based on lighting conditions. Patients with a score greater than 30 are considered to have less healthy retinas, while patients with a score of lower than 30 are considered to have healthier retinas. When looking at patients with less healthy retina, 30% of those patients in the sham group experienced a 15-letter or greater loss compared to 16% of those treated with ANX007 either monthly or every other month. Among patients with healthier retinas, 17% of patients in the sham group experienced a 15-letter or greater loss compared with no patients in the ANX007 monthly group and only 6% in the every-other-month group. These data was statistically significant in all three ANX007 groups, with p-values of 0.0016 for the monthly, 0.029 for the every-other-month group, and 0.0009 for the pool population. When comparing the data between ANX007 treatment groups and the sham group, the findings are striking. Treating patients earlier in their disease is always preferred, especially in neurodegenerative disease. When you look at the overall population of people with GA, those most likely to be treated first or those with an LLVD score of greater than 30 as they are showing rapid challenging loss of vision. But there's a large segment of the population with healthier retinas that is just starting to experience the impact of GA on their vision. These data show that those patients may benefit even more from ANX007 with earlier treatment in their disease. In an assessment of LLVD on Slide 15, we looked at patients in two ways. Patients with better vision as measured by BCVA score of greater than or equal to 55, and patients with impaired vision with a BCVA score of less than 55. In the first group of those with better vision, 25.5% of patients in the sham group had a worsening of 15 letters or more, versus 9.4% in the ANX007 monthly group and 15.4% in the every-other-month group. Among those with more impaired vision, 17.9% in the sham group lost 15 or more letters versus 17.7% in the ANX007 monthly group, and just 3.3% in the every-other-month group. ANX007 monthly and pool groups achieved statistical significance, with p-values of 0.0161 and 0.019, respectively, while the every-other-month group showed a nonsignificant trend with a p-value a 0.0597. In total, when assessing all measures of visual acuity, we are extremely encouraged by the preservation of visual function demonstrated by ANX007 in this GA patient population. Now let's turn to GA lesion area. The primary endpoint of the study is based on the GA area. This is a measurement of the loss of RPE cells. RPE cells are needed to nourish photoreceptors, and this loss has been hypothesized to loss of overlying photoreceptors, resulting in vision loss over time. However, meaningful function loss from this mechanism has not been consistently shown, and this separation of loss of RPE and visual function is further confirmed in the ARCHER study. As shown on Slide 17, over the 12-month period, ANX007 had a relatively modest nonsignificant effect of 6% on lesion growth reduction in the monthly treatment group with a p-value of 0.0526, a 1.3% reduction was observed in the every-other-month treatment group with a p-value of 0.896, and a 3.7% reduction was observed in the pool patient population with a p-value of 0.673. These results did not correlate with the preservation of vision function seen on the functional measures. Finally, supporting ANX007 are the safety and tolerability findings. Turning to Slide 19. ANX007 demonstrated favorable safety results when administered either monthly or every other month. CNV incidents, as determined by the investigator, was well balanced between sham and treatment groups. There were 3 cases of endophthalmitis, all of which were determined to be invested -- by the [ investigator ] to be related to the injection procedure rather than the drug. There were also 3 cases of intraocular inflammation observed in fact to be mild. There was one case of iritis treated with topical steroids and resolved within 48 hours. There was one case of vitreous treated only with topical steroids that resolved within a week. And there was one case of anterior uveitis that presented with cataract that was also treated only with topical steroids. In none of the cases was retinal vasculitis observed. Lastly, there was one case of retinal artery occlusion and retinal vasculitis was also not observed in this patient. Overall, we're quite pleased with the safety results observed with ANX007 in this trial. So the totality of the data with ANX007 and ARCHER trial are promising for the field, and most importantly, for GA patients. In our study, 20% of patients in the sham group lost at least 15 letters, or approximately 50% of their vision in just 1 year. To put that in real-world context for patients, that would mean 1 in 5 have significant vision loss likely to impact their activities of daily living. Ultimately, without a treatment that protects against vision loss, millions of people are at risk of losing their sight, along with invaluable human attributes, like the ability to see loved ones' faces or to maintain their own independence. Based on the ARCHER findings, we plan to engage quickly with regulators to discuss next steps and determine the optimal path to advance ANX007 to our patients. In close, we believe the ANX007 represents a new and distinct product opportunity for patients with GA. It represents yet another important step toward achieving our mission here at Annexon of delivering game-changing treatments to patients living with complement-mediated diseases. This is only made possible by the passion and dedication of a lot of people. To our employees, collaborators, and advisers, these data are a great accomplishment, and I thank you sincerely for your warrior spirit and all-for-one commitment. To the physicians, medical teams, and most importantly, the patients, their families and caregivers who participate in our clinical trial, we are eternally grateful and humbled by your support and contributions. The GA field would not be where it is today without you. Before we open the call for your questions, we're delighted to be joined by Dr. Charlie Wykoff who will take -- who will make a few comments from the treating physician perspective. Dr. Wykoff, over to you, sir.

Thank you. Good afternoon, everyone. It's great to be here with everyone and with the Annexon team. Some great summary of the data, fascinating data, a lot to think about here. I think if I summarize this from my perspective as a clinician and the clinical trial is in a very busy space, of course, we have to look at this and think it's frustrating to not meet a primary endpoint. But then very quickly, I think my takeaway is that we're learning something much more important here than just missing a primary endpoint. This trial was structured around the primary endpoint of looking at a biomarker on fungicidal fluorescence that has been validated now as an approvable endpoint for a drug that can slow the progression of geographic atrophy at the anatomic level based on fungicidal fluorescence. And that's an important biomarker for us as a field. And what we're seeing here is a nonsignificant effect on that very specific biomarker. But then when you really step back and think, well, what are we trying to achieve as clinicians for our patients, we're trying to do what we're seeing here, which is a first-time finding in geopathicatrophy at a significant meaningful level like that. And that is preserve visual function. And I think the key point for me is that while this is a complement inhibitor, we are seeing the complexities of the complement pathway. And it's really important what those specific targets are. So a C3 inhibitor, C5 inhibitor, clearly, to me, have very different impacts on preservation of retinal tissue and functional outcomes than now what we're seeing with this 007 investigational product. And so to see preservation of vision in a way that we have not seen any other study is quite meaningful. It's meaningful in the field because it shows that we really can impact visual function, which is what patients absolutely care about through a 12-month endpoint, which is very meaningful. And it also shows us that there's going to be many different endpoints to consider for a host of drugs being studied in geographic atrophy. I think it's overly simplistic to think that all drugs in geographic atrophy are going to have the same effect on every biomarker. And so from another perspective, I see this as complementary to, for example, the other agent that's already FDA-approved, which has been shown to slow the progression of fungicidal fluorescence-based lesion growth. This is quite distinct and very meaningful for patients in a very different way. So maybe I'll pause there and happy to take any other questions.

[Operator Instructions] Our first question comes from the line of Tazeen Ahmad from Bank of America.

Just maybe a couple. If you were to compare your results that you got today versus what the C3s and the C5s have shown already, how would you think about the benefit that you saw in VA versus what the others have not been able to show? I guess, does that give you a particular confidence that focusing on C1q could actually, in the end, end up better? And then can you give us some thoughts this early on why you might have missed the primary endpoint and what you would change for the next study?

Yes. Thanks, Tazeen. Good questions, and we appreciate that and appreciate you joining the call. Really good questions. So as it relates to your first question and the difference in what we've seen in our study versus some of the other downstream approaches, well, it's, as Dr. Wykoff really very nicely stated, we're focused on function. And that is ultimately what patients are concerned with maintaining their vision and their independent lifestyles. And so what we saw was in the study was an impact on function as measured in multiple ways, predominantly looking at BCVA at the 12-month time point. Over a 12-month period of time, we saw that functional benefit also demonstrated regardless of patients' foveal status, whether they were foveal or non-foveal, and we saw it in the healthier eye and the sicker eye. And so that's our objective. We have not seen that in a single one-year study before. And so we're really encouraged by that across the board. I think when you turn to fungicidal fluorescence and the measurement of that biomarker, it's just a very different mechanism. As we said in our prepared remarks, there you're seeking to protect against the loss of RPE cells. That's really mediated through a clearance mechanism with downstream complement approaches, which is very different than targeting the photoreceptor itself, which is where function and vision is housed. And so we just candidly think it's a very different approach. Our hypothesis is that the impact that we had there or the minimal impact that we had there is likely due to one of the other complement pathways playing a role in the clearance of RPE cells, which does not come into play as you attack or approach the photoreceptors.

Yes. Okay. And I'm happy to jump in the queue. I have a thought on that. I think it's really important to move away from the concept of which is better, right? We live in a world where there's, for example, so many different anti-VEGF agents in retina. And so we've kind of gotten new stacks of comparing this drug to that drug, because at the end of the day, they all do something very similar, and we can definitely debate about distinctions between anti-VEGF agents. I think what we're seeing here is a bit of a discussion of better, I think it's more of a question of differentiation. And there's meaningful differentiation here. And I think that there's a plausible mechanism that the Annexon team has been working on for many years, which I think can potentially explain why we're seeing preservation of visual function here, whereas other medications that are really targeted on slowing growth of lesions may not be seen a visual function benefit. Ideally, from a physician perspective, we would have seen much more robust visual function preservation with C3 inhibition, just to focus on the FDA-approved agent out there. And there were some signals with microperimetry, for example, and the foveal occupancy data, but there wasn't a really strong overarching signal and certainly none of the prespecified endpoints seem to show a signal. And that's frustrating as a field because we always -- we built this fungicidal fluorescence story around the concept that eventually we would see a functional benefit, and we certainly talked about that a lot over the years. So with that background, having enrolled a population of patients here, that's actually very similar to the C3 and C5 inhibitor populations. And to see what I think is a strong signal of visual preservation through a 12-month endpoint, I think, is really meaningful because if this biologically is going to continue to have protective effects, I would expect that, that differentiation between active treatment and sham to continue, for example, through an additional year of treatment. So to see this signal within 1 year, I think, clearly differentiates this mechanism of action from the other complement inhibitors that we've been discussing.

And then maybe just one follow-up to what you said. How do you think about the every-month dosing versus every-other-month because then the first year it does look like the every-other-month had a more meaningful impact on preservation of vision?

Yes. We clearly saw a differentiated a dose-dependent impact on function based on every month and every other month. These data are new to us. So we'll guide into that, and hopefully, learn more about that in the coming weeks.

And our next question comes from the line of Joseph Stringer from Needham.

A few from us. Just given that there is no regulatory precedent in GA and endpoints that are established on lesion growth, what's your feeling on what the path forward is from a regulatory perspective and maybe your confidence level that FDA would be flexible on regulatory endpoints in GA? And then...

I think we lost Joey. Operator, are you still there?

Yes. His line is...

I got the first part of his question. I can't hear him.

Yes, I don't hear him now. There he is. It seems like a poor signal.

I'm sorry, can you hear me okay?

We can hear you now.

We heard the first part of your question regarding regulatory approach for vision, but the second part of...

Yes. I'm sorry. For the second part, for BCVA on Slide 11, how does the risk reduction or the response rate for the sham group compared to that of the natural disease progression?

Yes. Great question. So with regard to BCVA as an approvable endpoint, it is the gold standard for assessing vision and has been used as an approvable endpoint in other indications. We anticipate that it is an approvable endpoint here for geographic atrophy. Indeed, it is the most rigorous assessment for visual acuity in the marketplace. And so we're pretty confident on that. That being said, we will go down and talk to the regulators about all of this topic and many others related to this. And then just with regard to the sham rate and natural history, the best example would be lampalizumab where they've kind of published their longitudinal sham data. And we haven't seen it from some of the other programs, happy to look at it. And this is very much comparable with that. But about 20% of our patients had a 50% reduction in under BCVA, so lost 15 letters at month 12. And that's very much close to what we saw in the lampalizumab study.