Beyond the gene: Targeting neuroinflammation in Huntington's disease

Dr. Rajeev Kumar: It's a pleasure to be here. Thank you

Dr. Sarah Camargos: So while HD is caused by a well-defined genetic mutation, there is a growing recognition that neuroinflammation may contribute to disease progression. [00:01:00] ANX005 is a monoclonal antibody targeting C1q, a component of complementary pathway. Could you briefly explain why the complement system is relevant in Huntington's disease?

Dr. Rajeev Kumar: Well, thank you for the question. The complement system is something most of us learned about in medical school, and we think about as part of the immune response to infection. It tags pathogens and damaged cells for clearance. But in the brain, it's produced locally, and it has quite a different role.

During normal development the complement proteins, especially C1q at the top of the cascade, tag excess synapses so that microglia can prune them, and that's healthy and is part of how the brain refines normal brain circuits. However, in HD, the pruning system seems to get pathologically reactivated, and there's really great preclinical work, especially the paper by Wilton and colleagues from Nature Medicine from 2023, which demonstrates that in Huntington models, C1q tags cortico-striatal [00:02:00] synapses.

Then microglia eliminate them, and then you lose the connections early on, and this tracks with the early symptoms in HD. And this is reinforced by human postmortem evidence in that we see that in Huntington brains, we see upregulated classical complement components at the synapse. And indeed, if we look at patients, CSF complement activation markers like C4a are elevated, and elevation correlates with disease progression and disease stage.

So there's really mechanistic evidence that complement isn't just a bystander of neuroinflammation, but actually might be a driver of aberrant synaptic pruning and drive disease progression for many patients.

Dr. Sarah Camargos: For clinicians who are not immunologists, what does C1q inhibition mean in practical terms? Why C1q was selected as a therapeutic target?

Dr. Rajeev Kumar: Sure. In practical terms, complement cascade is a chain reaction, [00:03:00] and C1q is at the very top. It's the recognition molecule. Once it binds to a target, everything occurs downstream in motion. And ANX005 is a monoclonal antibody that binds C1q, to prevent it from initiating the cascade. Now, we want to start at the top rather than something downstream for two reasons.

First, if you block C1q, you prevent synaptic tagging at its source. Rather than trying to mop up everything after the tagging has already occurred and you're already getting damage, you can prevent that tagging. And second, it's pretty selective. The complement system has multiple arms. It has a classical complement pathway, but it also has the alternative and lectin pathways, but these converge further on at C3.

So, you block at C3, you shut down all complement-mediated activity, which could put you at more infection risk. By just blocking the classical complement arm, we're leaving those other areas intact, and we're going after the pathologic process which we think is pathologic pruning, and we leave the alternative pathway and background surveillance activity intact.

Hopefully, that's pretty [00:04:00] clear. We want to get synaptic tagging and prevent the broader neuroinflammatory loop.

Dr. Sarah Camargos: So you reduce the off-target effects, right? Doing this.

Dr. Rajeev Kumar: Yeah. And indeed, in our study and in subsequent large-- many studies now that have been done with this ANX-003, which now has a funny name called Tanruprubart. Yep, that's the name because it's now before the FDA and the European Medical Association for approval for Guillain-Barré syndrome.

Because of this, we haven't seen serious infections related to this suppression of complement activity by targeting C1q. So that's a nice safety feature.

Dr. Sarah Camargos: So your study reports the CAG age product or CAP score. Could you explain why this measure represents when studying participants and preclinical participants with Huntington's disease?

Dr. Rajeev Kumar: Sure. Well, CAP or CAG age product, it's a way of capturing the cumulative lifetime exposure to the toxic effects of mutant huntingtin as just a single [00:05:00] number. So the idea is that the burden of disease depends on the degree of CAG repeat expansion. The longer repeat is more toxic and causes earlier disease, and how long the person has lived being exposed to that.

So the CAP multiplies these things together. The duration, the severity of the expansion times the age, and if it's above a threshold, then one typically becomes symptomatic. So it's like pack years in smoking. Pretty easy. And indeed, this is a good way of enriching for our patient population who are early manifest or are late pre-symptomatic disease, and really within the window where we think that we can measure progression in a fairly short study and enrich for patients who are likely to benefit from the therapy.

Dr. Sarah Camargos: Please guide us through the outcomes and assessments of your study.

Dr. Rajeev Kumar: Sure. This is a phase 1b multi-center, open-label, single-arm study. So everybody got the active drug. There's no placebo. We enrolled twenty-eight patients. [00:06:00] Again, ninety percent were manifest, ten percent were late pre-manifest. Everybody was genetically confirmed, and they had preserved function. So the independence score was at least eighty out of one hundred.

Initially, patients received a prolonged twenty-one-hour loading infusion done slowly in order to reduce potential for adverse effects, which we can certainly go over today. Patient received a second dose, which was given faster at days five or six, and then starting at two weeks into the intervention arm, they received an infusion every two weeks until week twenty-two.

Then patients had ongoing follow-up to look for adverse effects and efficacy, as well as the pharmacokinetic and pharmacodynamic markers at weeks twenty-four, twenty-eight, and thirty-six. So they had a twelve-week, twenty-four-week treatment window, and then between weeks twenty-four and thirty-six, a twelve-week off-treatment window.

And that was actually quite interesting. And the primary objectives, of course, were safety and tolerability, the pharmacokinetics of the drug in serum and CSF. And then we [00:07:00] want to look for pharmacodynamic markers of target engagement, C1q levels, downstream complement markers such as C4a, C3, C3a, plus NfL as a marker of neurofilament injury.

So we also had exploratory clinical activity, which was quite interesting. So we assessed the Composite Unified Huntington's Rating Scale, or cUHDRS, and its components. Includes the Total Functional Capacity score, the Symbol Digit Modality score, Stroop word reading, and also, of course, the Total Motor score.

So the CUHDRS is a composite score that integrates all of these functional motor and cognitive measures in a single measure, and it correlates well with functional decline and also brain atrophy. So we didn't do imaging measures in this study, but it's something we're certainly considering in a follow-up study.

Dr. Sarah Camargos: And what were the most important findings regarding safety and target engagement?

Dr. Rajeev Kumar: Sure. So let's maybe talk about target engagement [00:08:00] first. So we achieve full saturation of C1q in serum and CSF really from the very beginning, earliest time points measures, which is at six weeks. And this target engagement was sustained throughout the treatment period and then gradually waned in the off treatment period. So CSF drug levels were low, just .2 - .25 percent.

And that's really what we'd expect with an IV-administered monoclonal antibody. But there was enough that crossed the blood-brain barrier so that we saw pharmacodynamic engagement. And it's pretty interesting is that when you get full saturation of C1q, it triggers a measure of target engagement, which is actually that saturating C1q triggers a self-limited fluid phase cleavage of C4a.

This is not on a surface of tissue, so it's not damaging. And so it's a marker. And this is actually probably responsible for some of the infusion-related reactions, which I'll talk about in a second. And then we also saw significant reductions in [00:09:00] downstream markers of complement activation, such as CSF, C3, and C3a.

And these improvements in these complement markers persisted into the off-treatment period. Now, on the safety, there are two main things. One is that when you give the drug, every single patient has an infusion-related reaction, and that's quite interesting and quite dramatic. And that's the reason why the first infusion is so long, twenty-one hours to give the drug...

Yeah. So overnight. It's not an easy thing to do because if you go slowly, you get less severe reaction, and you pre-medicate with H1 and H2 blockers, acetaminophen, and also corticosteroids. And this reduces the severity of the reaction, which is predominantly a maculopapular rash.

And it's itchy, pruritic, and some people will get flushing. These are pretty manageable. They're not severe. Most of the patients get grade one or grade two reactions, and it goes away. And what's interesting, this seems to be a [00:10:00] pseudoallergy related to that complement activation that occurs. Remember that fluid phase reaction I mentioned, a C4 to C4a, another anaphylactoid is being released, not IgE mediated.

And in fact, once you get full C1q saturation, which occurs with that first dose, you really don't see it with any subsequent doses. That's why the other doses can be given much faster because you already have full C1q activation. You don't get these infusion-related reactions. So that's a really interesting thing, and it really fits the mechanism of action.

The other important safety issue is treatment-emergent adverse events, some of which can be serious. And so we saw really an autoimmune phenomena in patients who had an elevated baseline ANA. And so one patient got a lupus-like syndrome, one person got ruminatus, and one person had autoimmune hemolytic anemia.

These all resolved by stopping therapy and/or treating — For example, the patient who had the lupus-like syndrome was managed with steroids and [00:11:00] hydroxychloroquine. Now, this is not surprising because there's evidence suggesting patients who have a deficiency of early complement components such as C1q can get lupus.

It predisposes to this. For future studies, we need to have more aggressive ANA monitoring and potentially exclude those patients who are ANA positive and ongoing monitoring for potential autoimmune phenomena. Because this was a short study; patients were only on drug for twenty-four weeks. It's possible there may be more autoimmune phenomena with more chronic suppression, and that's something that needs to be assessed in longer phase three studies.

Dr. Sarah Camargos: Yeah, but I remember, you measured the ANA antibody in all of them at the baseline were less than 160.

Dr. Rajeev Kumar: Yeah. The three people who are positive, even relatively low titers can induce or can predispose to this. So it's only in those three patients. So that seems to be a clear marker. And so this needs to be taken into account, I [00:12:00] think, as use of this drug potentially becomes more widespread, for example, in Guillain-Barré syndrome or potentially in other neurodegenerative diseases where neuroinflammation may be playing a driving role.

Dr. Sarah Camargos: Were there any biomarkers that appeared particularly promising for future trials? 

Dr. Rajeev Kumar: Yeah, it was really interesting in that we found that the baseline level of complement activation or complement activity measured by the ratio of C4a to C4 ratio seemed to predict treatment response. So C4a is a marker of how active the classical complement pathway a- is, and it's elevated in patients with manifest Huntington.

As I mentioned before, it seems to be associated with disease progression and disease stage. So if we stratify the patients who have higher C4a to C4 ratio and compare it to the lower ratio, we can see a difference. The patients who had the higher ratio, seemingly more neural inflammation or complement activation, [00:13:00] seemed to show the clinical benefit.

And we could use that to stratify a subsequent trial and also possibly enrich our population for subsequent trials. And the target engagement markers of C4a rise and suppression of C3a, C and C3 were valuable. They confirmed that the drug does what we want it to. It's on target, engages, and it has a durable effect in suppressing complement activation.

One thing that we were disappointed was the neurofilament light. It really remained pretty stable. It really, I would say, was not informative here. It's a short study, only twenty-four weeks. We know that in HD, there's a lot of inter-patient variability in the NfL levels.

The dynamic range for NfL is pretty low in HD, different than in ALS, where it can be quite marked. And, there have been some therapies, for example, in ALS, which has shown NfL levels can be suppressed. We didn't see that here. It's something we need to think about in a [00:14:00] larger, longer study, and it makes me think that we need to rely more on volumetric MRI imaging measures as a secondary marker of outcomes when we do a Phase III double-blind, placebo-controlled trial.

Dr. Sarah Camargos: Okay. Although this is a primarily a safety study, were there any clinical signals that you found encouraging, especially on this subset of patients with the ratio C4/ C4a? 

Dr. Rajeev Kumar: Yeah, for sure. This higher baseline complement activity, like you said, in those people that had stabilization or improvement throughout the twenty-four week period of treatment, and this was sustained into the off-treatment window, while the lower activity group declined pretty much in parallel with the natural history cohort taken from, for example, from the TRACK-HD study.

And, if you looked at those patients with a higher complement activity, about 75% of those people improved, whereas only 36% of the lower complement activity level improved. And this was quite uniform across functional measures, cognitive [00:15:00] func- measures, and also motor functions.

So that's good. And there was improvement by week six, which suggests that this is not suppression of neurodegeneration, but there was potentially rescue of dysfunctional but not-yet-dead neurons and synapses by removing C1q from those synapses. Those sick synapses, we'll call them, now became more functional, and there was symptomatic improvement.

We need to see what happens with long-term therapy in terms of potentially suppressing secondary neurodegeneration. But I think we should be pretty clear and honest. This was a small study, open label, single arm, no placebo, and the responder analysis that I've indicated between the high and low complement activity was a post hoc analysis.

Although, the investigators, patients, sponsor was all blinded because that analysis was not done till after the study was completed. So that's good and gives me confidence about it. And the effects across those different clinical measures was all quite uniform, and that gives me more confidence.

But, [00:16:00] it's a small study. The statistical considerations are that, this was nominally statistically significant. We did a lot of statistical comparisons, and there was no correction for multiple comparisons. So we have to be honest and recognize the limitations of our study, but very much a great hypothesis-generating study.

Gives us confidence to go forward with a really a big study. This helped to prove the biologic concept, though.

Dr. Sarah Camargos: And do you think the complement inhibition could apply to also later stages of disease, not only the early Huntington's?

Dr. Rajeev Kumar: Yeah, I think it's certainly possible. But, the mechanistic way this works suggests that we should go earlier. In fact, even to the pre-symptomatic stage because we know that there's early complement deposition. We have a lot of synapses and lots of neurons to potentially preserve. And the early symptomatic improvements suggest that we can rescue viable and sick, but not-yet-dead [00:17:00] neurons. So we wanna predict them earlier before synapses and neurons loss. Now, that doesn't mean that we can't necessarily have a useful interaction in later stage disease, but all of this suggests that we should probably go earlier rather than later. I think we want to use now the ISS to enroll patients across a number of different stages, as this therapy is developed. But I think the first large placebo-controlled trial needs to be done in patients who have early symptomatic disease or late pre-manifest disease. You know, ISS early stage three, stage two, late stage one. 

Dr. Sarah Camargos: Is the complement system primarily a marker of ongoing neurodegeneration, or do you believe it actively drives disease progression? 

Dr. Rajeev Kumar: I think there's quite a bit of evidence that it's actually a driver as opposed to being a bystander, we'll call it. So if we look at the preclinical evidence, in Huntington models, if we block C1q, it preserves the synapses and rescues cognitive deficits. And we can see that complement activation and [00:18:00] synaptic targeting occur earlier, and synapse loss seems to occur earlier, not just as a late footprint.

Plus, we saw that patients who had more active complement activation were the ones who responded. So that suggests that it's actually part of causation rather than just a biomarker. this, our study, it provides more credible case that complement drives progression, but we need to confirm this in a larger, longer study for sure. 

Dr. Sarah Camargos: For sure. What would be the next step to know whether this approach truly modifies disease progression? You talk about the MRI neurofilament, longer duration...

Dr. Rajeev Kumar: I think we need to do all of those things. We need to do a randomized double-blind placebo-controlled size. It has to be adequate duration. It has to do a number of things. We need to look at cUHGRS levels and measures. It declines by about one point per year. A clinically meaningful effect is maybe twenty, thirty percent slowing.[00:19:00] 

So I would say we need an eighteen, twenty-four month study. We need to pre-specify and enrich for patients who have high complement activity and stratify by that. Maybe power by this high biomarker subgroup. And, we have to include not only these complement measures, NfL, but a volumetric MRI, I think is needed.

So this is the study. It's gonna be a lot of work to do it. It's gonna require a lot of resources. I'm hopeful that we will do it. and I think Eneaxon, the sponsor, is very interested in doing it. Of course, the resources required are a lot. Eneaxon is a small company. I am hopeful that their application for, in Guillain-Barré syndrome will be approved. The company will start to have an approved product, which is this drug. We'll have funds then to invest back to get an indication in Huntington. Let us hope for that. 

Dr. Sarah Camargos: Could complement inhibition eventually be combined with huntingtin-lowering therapies? 

Dr. Rajeev Kumar: Well, I think mechanistically it makes tremendous sense. It's really rational. Huntingtin [00:20:00] lowering goes after the upstream genetic driver, right? The mutant huntingtin protein itself, somatic expansion, the root cause. Complement inhibition acts downstream, protecting synapses from the damaging consequences of genetic driver. In microglial activation, we have to think about it. microglial activation is driven by mutant huntingtin. That potentially releases C1q, which tags the synapse, which starts this terrible self-reinforcing cascade. If we can knock down mutant huntingtin, plus we can downstream affect microglial activation and reduce C1q-mediated aberrant synaptic pruning.

We can attack two different nodes simultaneously and really have a greater effect. But, combining two therapies is very complicated. I think we need to show efficacy in huntingtin lowering therapies. We're on that rate with the recent uniQure data. Let's hope for an approval there. And, I think that we need to separately show efficacy of the anti-C1q therapy, and then we put them together. 

Dr. Sarah Camargos: [00:21:00] Great. Do you foresee this therapeutic strategy being applied to other neurodegenerative diseases, as you mentioned, as it was for Guillain-Barré? 

Dr. Rajeev Kumar: Yeah, I think that there is a lot of optimism that might be the case because there is evidence that the, classic complement system-driven excessive synaptic pruning may be present. For example, in Alzheimer's disease, there's evidence to show that C1q tags synapses even before plaques occur.

If you block that pathway, you can prevent synapse loss and that maybe genetic factors like APOE, TREM2 reinforce that. Parkinson has some post-mortem and biomarker associations. In ALS, there's complement deposition on motor neurons, but there's other evidence in ALS that might not be such a big marker, especially early on, maybe later. Complement activation is important.

So I think it, it's credible that there are specific patients or subgroups of patients within each of those neurodegenerative diseases that complement activation is an important driver. [00:22:00] These baseline measure of complement activity may be a way to find those patients and then potentially do trials in those individual subgroups within those different neurodegenerative diseases.

So I'm optimistic. 

Dr. Sarah Camargos: Me too. Of course, this is, still early stage research, so the quick questions were safety, target engagement, biomarkers, and whether larger control trials can show meaningful clinical benefit, right? 

Dr. Rajeev Kumar: For sure. 

Dr. Sarah Camargos: So, Dr. Rajeev, thank you. We appreciate your time and willingness to participate in the podcast, and we look forward to seeing how this line of research evolves and contributes to future treatments for Huntington's disease. 

Dr. Rajeev Kumar: Thank you for this opportunity. 

Dr. Sarah Camargos: Thank you. [00:23:00]