Great. Welcome back to the next session of our 2025 Global Healthcare Conference here in Miami. Welcome, everybody, to my hometown. I'm happy to be hosting Rachel McMinn from Neurogene. Rachel, how are you doing?
I'm good. I'm good. How are you?
I am three minutes late to the interview. Somewhere, my mother is disapproving as we speak. That aside, let's dive right into what I think most people are spending their time asking, certainly asking me and my team about, which is, what are the range of regulatory outcomes for a pivotal Rett program, for a program like yours, not a small molecule? How should people think about the goalposts?
Yeah, so let's just take a huge step back and just briefly introduce Neurogene and then dive in. But Neurogene is a clinical stage gene therapy company. Our lead program is NGN-401 for Rett Syndrome. I know you know this, but just making sure everybody else does. Rett Syndrome is a very devastating neurodevelopmental rare disease with a global prevalence of approximately 1 in 10,000 females. That is a large market opportunity, very attractive, and there are no approved treatments that really address the underlying cause of disease. Clinical manifestations, including severe impairments to hand function, communication, gross motor function, as well as other abnormalities, including seizures and GI dysmotility, breathing abnormalities. This is a very devastating condition.
NGN-401 is unique in that we had to design a regulatory system in order to control precisely the levels of transgene coming off of the vector because of the nuances of the gene itself. Too little causes disease while too much is toxic. All of that is baked into NGN-401. We presented our initial clinical data on four patients in November of 2024 with some very encouraging efficacy data. We are at the point now to bring it back to the question that you actually asked in negotiations and discussions with the FDA of what could this look like. Taking a big step back, NGN-401 was selected as part of the START pilot program under CBER, one of three gene therapies that were selected.
The purpose of START, literally like the spirit and design of START, is to take promising medicines and help accelerate them through the development phase. Developing treatments for rare diseases is very, it's not an easy thing to do when you're pioneering a treatment. As we think about the nice thing about being under START, given the goal of what that program is trying to do, it's obviously very much aligned with what we're trying to do. It allows us a cadence of communication and meetings, interactions to really start to begin to unpack some of these really critical topics of what is the registrational trial elements. When you think about sort of the two biggest elements, it is the control strategy as well as the primary endpoint.
Again, just to kind of high level, we're allowed to be able to do this without all of the ABCD type of meeting series. There's no PDUFA packages. That allows us to begin to tackle these questions. Now let's get to the range of outcomes. On the control strategy, number one question people ask is, do you need a placebo? Are you going to run a placebo? What we've been saying for some time now, and we're continuing to say, is that a placebo really is not warranted in this kind of study. That is because both the ICV procedure as well as the immunosuppression regimen are difficult, if not impossible, to blind.
If the placebo doesn't actually blind, then doing it just for the sake of, well, that's the way gold standard trials are run, it's not going to achieve its intended objective. Now you're taking placebo off the table. Once you've done that, it leaves you with sort of two flavors of options. One is a single arm study. We'll get into what that would look like. The other one is a delayed start, where you're randomizing a population into treatment, non-treatment. Let's kind of tackle each of those.
In order to have a single arm study, you absolutely need, we would need to obviously delve into not only our data, but contextualize that against the natural history and gain alignment with the FDA that basically whatever the changes we see according to the primary endpoint are above and beyond what you would expect to see in the natural history. That is sort of paramount in order to justify. Otherwise, somebody could see changes and then you could argue, OK, those changes could just normally happen. Single arm study requires that alignment on how to analyze the natural history. I do not want to keep talking. Do you want to pause or shall I finish up the randomized control?
I would stop with one quick question.
Yes.
When we talk about natural history, there's a lot of confusion about what that means. Because Rett Syndrome, unlike other very severe pediatric inherited illnesses like SMA, which tends to be fatal very early, has a quite long course. It's tragic and it's complex, but it's long. When you talk about natural history, across what age range of natural history are we talking about? That can be very different for studies of girls that are enrolled quite young as opposed to young women with Rett who will be much farther along in the disease process.
Yeah, absolutely. Age range does matter in this disease. I think we've talked about this, but just for everybody else, in infancy, these babies may not even be diagnosed, but they're developmentally delayed. At some point, they actually start to regress. They're losing some of the simple skills that they acquired. At some point, and that is a differing course. Some patients regress quickly. It takes them some period of time. The population that we are studying is post-regression. They're still young. In the case of this population, it's an age range in the pediatric cohort of 4-10 years old. That was selected specifically because we wanted to rule out patients who are still regressing. If you saw an improvement, you wouldn't know, is that just sort of bouncing around during the regression period?
Because that period is very uneven. It's not a linear course down. Sometimes a kid will stop talking, for example, and then have some words and then stop talking again. If you saw some words, you really wouldn't know, is that because of the drug or just sort of the natural pace of regression? That is why we are studying post-regression at this point in development, because this is a period where it's called pseudo-stationary. Basically, these girls typically don't gain skills. They typically don't improve. When you look at developmental milestones specifically, there are sort of two large buckets. One is a bucket of skills that we just talked about, where the girls had gained them and then they lost them during regression. The proportion of girls gaining back those skills is pretty low. There is a suite of them.
To your point, it's not Zolgensma where you're talking about survival and sitting independently. That was on a relative basis, pretty straightforward. Now you're talking about, are you using words with meaning? Are you able to use a pincer grasp? Are you able to self-feed? There's a long list when you get into gross motor function, sitting independently, walking with or without assistance, being able to stand from a seated position, being able to use a utensil, going up and down the stairs. There is this other bucket of skills that are much more complex that girls with Rett Syndrome typically never learn. 80%+ of them on these more complex skills, it's not that they learned them and lost them. They literally just don't learn them. That's the kind of data that we have that we're able to interrogate, like, what are these girls?
What did they have and lose? What is the likelihood? What proportion of girls actually relearn those skills? Then this other bucket of, I never learned them because they're complex, and that's just part of the natural history of Rett. Does that help explain? We can talk about the older patients separately, but does that help explain at least where we are with pediatrics?
That helps at the table.
Yeah.
I think one of the questions that I get is, OK, a placebo is not blinding, but to what extent does it make sense to do a sham procedure? This, among other things, suggests to me that people have never actually watched an ICV be done. Could you answer whether or not it makes sense for a sham procedure to be the control arm of this study?
We're aware anecdotally that others have tried sham procedure, even with ICV. The feedback that we learned was that it does not actually effectuate the blind, meaning people figure it out. It is pretty straightforward. When you're on gene therapy, you're using steroids to control liver enzymes and patients are on immunosuppression. You cannot take a girl who has Rett Syndrome, do a sham procedure, and then put them on immunosuppression. That is more than minimal harm to a child. You cannot do that. It is not like the DMD case where the standard of care is already immunosuppression. You kind of have these two separate issues. One is, can parents and caregivers sort of figure out that a child did not actually undergo general anesthesia, waking up groggy, not you come out of surgery? That is going to be very difficult to blind.
Then on top of it, you have this immunosuppression issue that is also part and parcel to gene therapy. I think for both those reasons, and I'm happy to talk to these investors that want to see a sham procedure, but when you're introducing more than minimal harm to a child, it's not ethical and it doesn't actually achieve what's been intended.
Great. We nailed down how to think about the control arm, which control arm comparator data implicitly talks about what the bar you're being compared to is.
Exactly.
I'm going to make a little bit of a right turn here into how do you evaluate effect size for a phenotype that is as diverse as Rett against the natural history control that we just talked about?
Some of that is going to obviously depend on what your primary endpoint is. We have not really talked about and disclosed what we are putting forward with the FDA. As you know from our data, there is a series of data. There is the CGI-I or the Clinician Global Impression of Improvement. We know from that data that, again, it is just four girls. It is early days. You have a range of follow-up from 3-15 months with two girls over 12 months and beyond in terms of follow-up, but still a small N. With that caveat, we know that all four of those girls were much improved on CGI-I. When you look at the clinical trials that have been done, again, with the caveat that this is open label, you can see that that has not been achieved with other treatment modalities.
Blinded, not blinded, that has not been achieved. In terms of RSBQ, we saw roughly 25% to not quite 50% reductions in RSBQ when you aggregate the data. Again, that has not been observed previously in other trials. Finally, on the developmental milestones, we just talked about contextualizing that data in these two different buckets of lost but then relearned as well as never learned. Obviously, we've interrogated all of that data. We're looking at all of that. The way it works with regulators is that obviously, we're going to the thought leaders to understand their perspective on what's clinically meaningful. We're also going to the community to understand. We've had that market research and that feedback to understand what's clinically meaningful to them.
We make sure that whatever endpoint that we are going to be selecting, that that is driven by that input. You do not want to end up with an endpoint that does not mean anything to anybody. The point is that whatever data we gather, it is going to be informed by clinical meaningfulness and then justifying that to the FDA. I cannot answer your question specifically on the effect size, but obviously, depending on what endpoint you go to, then you are looking at numbers to basically justify, like, OK, what is sort of the highest level of "spontaneous improvement" that you would see in the natural history? Use that as some sort of cutoff to establish, like, OK, this is what happens most conservatively in the natural history setting. When you power the study, you are obviously looking for a delta over that.
Depending on what you've looked at and what assumptions you're making, that would inform the numbers of patients, for example, that you would need in order to meet statistical significance. Am I answering your question?
You're answering my question as thoroughly as you can.
Yes. As much as I can without telling you exactly what's going on.
We have spent this time mostly sort of setting the table on registrational design. Let's look at the other debate, which I think is one that more has investors that are a little more bearish conservative, which is around the safety. Obviously, there was a tragic safety event. Added dose is no longer being pursued. Lay out a little bit of changes to trial conduct, follow-up, monitoring, et cetera, and sort of paint a picture of how well understood HLH is by clinicians and what the interventions to treat an early elevation in ferritin would be in the study.
Right. Just to set the stage here, we had two patients who had been dosed safely with a 3E15 vg dose. We had had five patients that had been dosed safely with a 1E15 vg dose. While these were enrolling concurrently, we had a third patient that was dosed with that 3E15 vg dose that had this side effect that you're talking about that was ultimately a rare but hyper-inflammatory response. It has become clear that it basically falls under the umbrella of HLH, which, again, very severe, very severe immune reaction. At the time that the case was ongoing, there was very limited information that was available. There is a small number of reports of HLH following treatment with AAV therapy. It had been seen before, but rarely reported.
What we learned, obviously, during that process and subsequently is that if it is detected early and if it is treated early, that HLH is actually reversible. That, obviously, it's still a very terrible thing that happened. If there's a silver lining, we want to make sure to make sure that everybody in gene therapy understands this. The other thing that we learned is that the 3E15 dose, and this is I do not know why this is complicated for people to do the math, but I just want to make it really simple. That dose translates into between 1 and 2E14 vg/ kg. I say that because it's really important for people to understand that the HLH cases that have been reported have all been associated with at least a 1E14 vg/ kg systemic exposure.
That's sort of point number one. Obviously, we have not seen any HLH. We're not aware of any HLH cases in the E13 vg/ kg range. The 1E15 dose that we're moving forward, again, translates into, from an exposure perspective, the E13 vg/ kg range. What did we do? The obvious thing was, let's take out the 3E15 dose because we're seeing very strong efficacy at the 1E15 dose. No reason to continue that dose at this time. That was sort of change number one, most important change. That being said, we had clearance to proceed, and we could have gone straight ahead with continued dosing with making no other modifications to the protocol. As a company, though, knowing what we knew, we didn't feel comfortable with that. To your point about ferritin, we added ferritin as part of our standard monitoring.
The reason why we did so is that ferritin, while it's a nonspecific marker of inflammation, it turns out in HLH, it's pretty dispositive and it's very sensitive. That number, that ferritin number, actually jumps very dramatically in the first few days of following AAV administration. To give you a sense, for the case report that is published in 2022, it's a very nice study. It shows that a baby with a normal ferritin level of 80 ng/ml , 48 hours after he received a treatment course of AAV9, the ferritin number was 3,000 ng/ml . It's not a subtle change. The HLH numbers, HLH as an indication has been around for quite some time. The diagnostic criteria were actually published in 2004. We're building off of more than 20 years' worth of experience of HLH across settings.
We know there that ferritin actually can go into the tens of thousands. What I told you was at 48 hours, it was 3,000. If you go on for days, it's obviously going to continue to go up. As a sponsor, we felt this is a very straightforward, simple blood test to add. We increased the monitoring. While it was pretty close to daily, we just added a few extra days of monitoring just to make sure, again, in the abundance of caution, we certainly wouldn't want to miss if there was something dramatic going on and lose a day. That's really the big change. Certainly, if we were to see a big jump in ferritin along with other markers that are associated with HLH, there are treatments that can mitigate and potentially reverse this hyper-inflammatory response.
These are very standard, again, well-trodden treatments. We did not invent them. High-dose steroids is something that has been proven to work following AAV. The other treatment that has been proven to work is anakinra, which is an IL-1 receptor antagonist. That is just very clearly laid out in our protocol so that, again, if we did see a bump, it is something we will be able to monitor extremely closely and treat to hopefully prevent any untoward complications.
That is helpful. Let's roll out over the course of this year. We're expecting an update around the regulatory path, as we talked about in the first half.
First half, yes.
Expect an update on clinical data in the back half of the year?
Yep.
Set the table a little bit on what we should expect from that clinical data set.
Yeah. There is, as I mentioned, the five patients who were previously dosed, and four of which we had enough information to talk about efficacy. There will be that information to be able to provide an update and follow up on those patients. We are also currently enrolling in the phase I/II an additional sixth set of patients. There are three slots for the pediatric cohort. This is an eight-patient cohort. We had enrolled five. There are three additional patients to add to that. We have also opened up an older age cohort, 11 years and older. In that cohort, we have outlined three patients. We will be able to provide an update, obviously, on the patients who have already been dosed as well as an update on the patients who will have been newly dosed.
Does that change depending on when we get a regulatory update or are these two totally separate processes?
That is a good question. I think it could be either. It really depends on the timing. If they merge to a point where it's possible that some of those patients could end up going into a registrational trial because it just doesn't make sense, then is that theoretically possible? That's theoretically possible. The way that we're thinking about it now is that they are separate. It depends really on the mechanics of the trial itself. Would we amend the current trial to make it registrational? As I said, just the timing of the enrollment of those patients.
That's great. I want to zoom out from what has been thus far, the conversation has been very sort of individual event-focused.
Right.
I know. Zooming out a little bit, one of the conversations that we've had from right when we initiated coverage was debate amongst investors of, well, how complex is it to do an ICV administration versus a competitor who has an intrathecal administration? How difficult and complex? How much of an adoption hurdle could this be? There's a non-trivial number of Rett patients. This is not an ultra-rare disease. It's uncommon and ultra-rare. Walk me through how you think about the infrastructure and the number of physicians that are routinely doing ICVs versus the potential demand from the pool of Rett patients.
Yeah. So look, ultimately, we think that the efficacy, obviously, the efficacy and safety profile of a treatment for a one-time treatment where somebody's going to be living with that treatment decision for the rest of their lives is really important. I know that one of the arguments is like, oh, just go with something that's less invasive. This is a very devastating disease. I don't think it's that simple of a decision. In terms of infrastructure, we've done market research. There are approximately 30,000 ICV procedures performed every year in the U.S. Now, that doesn't mean that all of those people would all of a sudden start treating Rett Syndrome girls. There will have to be certain hospitals that are Rett centers of excellence as well as all of these hospitals have neurosurgeons that are capable of doing ICV.
It is not like a gene therapy-specific unique administration that we have somehow invented. We are really just using an existing procedure that neurosurgeons are incredibly comfortable with already. It is the bread and butter of neurosurgery as an ICV procedure. I think it is going to be all of that. It is an education in the field of having whatever our data set looks like and having families and physicians decide based on that data set what is really the optimal treatment for them.
I think one more question I periodically get around thinking about accumulated diagnosed patient pool. When you think about patients who have tried but whose families have rolled them off of DAYBUE, which is an increasing pool because DAYBUE is not entirely easy to take.
Yeah.
Recognizing, given the direness of the unmet need, I don't think no one is surprised that that approval happened. For patients who have washed off of DAYBUE, would they be eligible for potential registration down the road? Or is that something you want to wait for more of a commercial environment? How do you think?
Oh, no, no. That's part of our protocol. If you're washed off of DAYBUE, you can enter into the trial. There's really no concern. After a certain period of time, patients can actually restart DAYBUE if they wanted to. We don't really view DAYBUE as directly competitive. It's just if patients want to be on it prior to treatment, if they want to be on it some period after, that's fine. The only concern is really specific to taking immunosuppression. We wouldn't want the immunosuppression to be excreted out or metabolized out or have any type of interaction. That's really where we're most concerned, just from a pure safety perspective. Other than that, from a commercial standpoint, DAYBUE is, we're happy that patients have an opportunity to take that drug, come off the drug, take it back again if they'd like.
It would make sense to have a minimal washout period before they enroll or something.
Right. We have a defined washout period as part of our protocol already. Obviously, that can be refined over time. This is not a drug that lingers very long. It is not a hurdle from a commercial perspective.
Great. Zooming forward, when you think about, I'm going to take back a few of the minutes.
Sure.
That I ruined by being late. When you think about zooming forward, you have to a potentially commercial environment with a data set produced in younger girls with Rett. How do you think about accessing the large pool of older girls and young women with Rett? Is that something that you think might require a separate study? Or is that an extension data set? Or should we think of this as most likely a single broad label? How do you think about the strategy? I know this is a little bit dependent upon your conversations with regulators already.
Yeah. I mean, look, we have a pilot cohort that is currently enrolling for these older subjects. From a pure biological standpoint, these girls have neurons. It's not the same as having a neurodegenerative disease where the neurons literally are dying. There is hope that an older patient population should be able to biologically benefit from gene therapy in Rett Syndrome. The magnitude of that benefit has yet to be defined. I think we want to better understand that. Certainly, our view is that we'd like to have as broad of a treatment label as possible. Depending on what those discussions look like with the agency, we would obviously create data in order to create access for those patients as well.
That seems like a totally reasonable place to stop. Thank you.
OK.
Looking forward to continuing the conversation soon.
All right. Thank you, Mani. Thank you, everybody.