Good morning, everyone. Welcome. Good morning. Welcome to the NeuroPace Investor Day. This is our first Investor Day, and we're very excited to have you all with us. A lot of familiar faces in the room. We really appreciate you taking this time to learn more about the NeuroPace story. We have an exciting program for you today. We hope you'll come away today knowing more about the company and being just as excited about our future as we are. So let's take a quick run-through of our agenda. In the first part of the day, you'll hear background about the company, our technology, our results in treating patients with focal epilepsy. And then you'll hear from a couple of special guest speakers, as well as an individual who is living with the RNS System every day. After that, we'll have some Q&A. We'll have our panel come up.
Then we'll take a quick break. And after that, in the second part of the day, we'll be more focused on the future. We'll talk about our clinical product and market development efforts. We'll have another special guest speaker. We will give 2025 guidance and then conclude, and lunch will be available outside the room. So let me introduce the folks you'll be hearing from today. We are absolutely honored to have a number of physicians speaking with you today, all of whom are luminaries in the epilepsy field. First, we have Dr. Vikram Rao, epileptologist at UCSF in San Francisco. Vikram has lots of experience treating people with epilepsy. We also have two neurosurgeons who will be with us remotely, Dr. Mark Richardson from Mass General and Harvard and Dr. Sameer Sheth from Baylor College of Medicine.
Our very own Mike McKenna will share his journey with epilepsy and treatment with the RNS System. You'll hear from members of NeuroPace Management. I think many of you know our CEO, Joel Becker, who joined us. Joel joined us about a year and a half ago with significant executive experience at big-name companies that you all know well. I'm Rebecca Kuhn, the company's CFO. You'll hear from Dr. Martha Morrell. Marty is also a luminary in the field. In addition to being NeuroPace Chief Medical Officer, she is also a professor and epileptologist seeing patients at Stanford University. We have Dylan St. John, our Chief of Commercial and Development, who is holding down the fort at home back in Mountain View. Katie Keller, our VP of Marketing. Katie is relatively new at NeuroPace. She brings significant experience developing markets in the cardiac field, as well as neuromodulation.
And we have Brett Wingeier. His last name is a bit difficult to pronounce. I hope I did that okay. Brett was rehired fairly recently, but Brett was actually one of our early employees who was instrumental in developing the RNS System and analyzing our clinical data back in the early days. So with that, turn it over to Joel.
Thank you, Rebecca. I'd also like to offer my welcome and appreciation to everybody who's here, some familiar faces and some new faces. Thanks to everybody for taking the time to be with us. In particular, I'd also like to recognize members of the NeuroPace board that are here with us today, Frank Fischer, our Board Chair. Thank you, Frank, for being here, as well as a couple of very valued advisors from the board, Rocky Kumar, as well as Uri Geiger, folks we rely on a lot, and so we're thankful for their commitment to be here with us today, too. My plan in this section of the agenda is to lay out the business, where we are today and where we're headed, both currently as well as part of our long-range plan.
So that's where we're going to spend the day, is we're going to tell you a little bit about the business, but we're going to really focus on where we're headed strategically and what those plans look like and how we're both in a strong fundamental position as a company, as well as really well positioned for growth here over the next three years, which is the long-range plan period that we'll be talking about today. So with that, we got a lot to cover. I'm going to jump right in. Of course, starting with the forward-looking statement slide, please everybody recognize and pay attention to those elements. Starting and talking to the business and introducing NeuroPace, it's important that we start with what is really the foundation of the business and what animates and catalyzes everything that we do, and that's the mission of the business.
And you can feel when you talk to folks about NeuroPace. You can feel the sense of stewardship that people have toward what we do to help the patients that really need it and what we can do with the RNS System. It's one of the reasons we're so excited about today, is the opportunity to tell you the story about RNS and where we're headed. I was talking with Dr. Morrell earlier, how excited we are for today for the opportunity to tell you that story. And it really starts with how excited we are to be able to help fulfill the mission of the business, which is to transform the lives of patients that are suffering from debilitating seizures. And it matters if we do so and we're in a strong position to get at it. So we got a lot to cover.
I'm going to jump right in. As I mentioned, we start with the business in a strong fundamental position. Those fundamentals have been hard-won over a number of years, but are now in place and really address a significant unmet need. So the unmet need that we recognized in the mission of the business discussion is one that is then evidenced here with a large under-penetrated market. So our core market in excess of $2 billion for the patients that go through Comprehensive Epilepsy Centers every year from an incidence perspective, and north of $50 billion in a total addressable market when we think about the number of patients that are drug-resistant epilepsy patients and the treatment gap that exists today. You'll hear us talk about the treatment gap.
We'll talk about the treatment gap in terms of what that means for patients, what the opportunity is then to help those patients with the business, and then what we're set about doing about it in terms of the key initiatives we have underway, and in addition to the solid fundamental of an unmet need in a large market, we have a very solid fundamental in terms of the differentiated unique nature of our technology. We have the only closed-loop brain-responsive neuromodulation system. It's a few long words, but there's a whole lot that goes behind that, and we're going to spend the better part of the day unpacking what that means in terms of technology differentiation, data development, and then opportunities that that affords us. So speaking of unique technology, why do we think it's differentiated? Well, it shows up in the data.
The clinical evidence shows, and we've got more of it coming. And we'll spend a whole section here today talking about the differentiated outcomes that we continue to see over time. We'll talk about additional data, some things that we haven't talked about before that you'll hear about today, as well as some forward-looking trial work that we're planning on as well. So those are the fundamentals, but the fundamentals are then accentuated with the execution focus that we have in the moment. So we've been generating some momentum really focused on revenue growth, gross margin, and operating expense discipline. We've got the flexibility from a balance sheet perspective to support our operating priorities. And we've got a number of future growth opportunities that we see that we'll spend time on as well. So as we think about that, where have we been?
Most recently, 2024 in review, I saw some of you at J.P. Morgan recently. We talked about some of this, but not all of it. Do pay attention. There is new stuff. Starting with a revenue growth profile in excess of 20% in 2024, we've announced, or pre-announced rather, the range. We haven't given the full readout yet on Q4, but the range is $79.4-$79.9, growth in excess of 20%, 21%-22%. If we adjust for that, if we hold for the sales in 2023 that we got from the implants associated with the NAUTILUS trial, growth was actually over 25% in the overall business. A strong growth momentum that accelerated throughout the year. We grew faster in the second half of the year than we did in the first half of the year.
And so we've got some good momentum from a top-line perspective. And we held on to the discipline with gross margins at the high end of the range, as well as operating expenses at the low end of the range. From an operating perspective, we also executed, with regard to our market development strategies, new prescribers at record highs for the year and every quarter throughout the year. So we continued to retest increasing numbers of active prescribers. Began to demonstrate momentum in the second half of the year with regard to our site of service extension with Project CARE. And then it might seem like a while ago, but it was only in Q1 of 2024 that we completed implants in the NAUTILUS trial.
And we'll talk a fair bit this afternoon about a bunch of the product development work that we've been doing as well, where we've initiated development with a software in our AI-enabled software tools that we're really excited to tell you more about, so some good momentum in 2024, and we're just getting started, and so what we're getting started on is where we're going to spend the rest of the day. And starting with, where are we headed? Well, where we're headed with the business, the vision for the business is to be the recognized leader, to be the standard of care in the space for drug-resistant epilepsy patients, and to fully develop the potential of the system that we have today, both for epilepsy and beyond, so that's where our focus is as an organization.
When we think about that focus, there are really three key pillars that we want you to keep in mind. One, it's a growth story. Both a market expansion, a market growth story, as well as a revenue growth story. We've got indication expansion, site of service expansion, patient and referral network expansion, as well as increasing adoption and utilization in our current Level 4 Comprehensive Epilepsy Centers that are all drivers that are underway during the plan period. They're all in flight today that are going to allow us to expand our markets and increase our revenue profile.
So it's a growth story, but it's a growth story with discipline and financial discipline that you've seen from us in terms of focusing on gross margin improvement, executing from an operating expense allocation perspective, and then both investing in our organization, but also getting leverage from that investment as well, in particular in our commercial team. As we expand indications, that same organization and those same call points allow us to expand and get leverage from our commercial organization. And finally, it's execution story. We are absolutely focused on execution. One of our core values is results orientation. And so we've got a lot of potential in front of us, but we're also very focused on the discipline that goes along with executing on that potential and then operating execution.
When we think about the growth initiatives that really support that growth story, we're going to spend time today talking about three areas of development that are underway to really fund that growth: clinical development, product development, and market development. From a clinical development perspective, we're going to talk about some new data. We're going to talk about our post-approval study data in our adult focal population that has now been submitted to the American Academy of Neurology for publication and presentation. We're going to give you the first tour of that data today, those of you who are in attendance. We'll talk to you about Nautilus and give an update on status with regard to completing follow-up in Nautilus. We're also going to talk to you about a new strategy and approach that we have for expanding indications into pediatrics.
We haven't talked to anybody about that, but we've got an approach for further expanding into the significant segment of the pediatric focal market that we call our NEST Initiative. We also haven't talked about very much and are going to spend time talking about our software development program and our annual cadence that's set up in the plan with AI-enabled software tools, remote programming, and our next-generation platform to enable efficiency, easy use, and increased efficacy for the platform. Those are all new things that we haven't talked to you about at all. And one of the reasons we're excited about being here today is to tell you about our product development platform that we've been working on as well. And then finally, our market development efforts. Sales force expansion, site of service expansion, and direct-to-patient expansion are all things that we'll cover as part of that.
When you roll it all up, this is what the plan looks like over 2025 to 2027, a series of increasing and reinforcing growth initiatives that over the long-range plan, over the period of 2025 to 2027, we plan and have targeted to exceed 20% CAGR over the plan period. And as you can see here in 2025, we've got multiple drivers that we'll spend time talking about, including our core customers expanding awareness and education, as well as new product launches and clinical trial submissions. In 2026, expanding on that with launches into IGE and pediatrics, as well as further launch into the product development portfolio. And then in 2027, really accelerating. So further penetration into those segments, launching remote programming, as well as enhanced user interfaces and continuing to build on the growth that's in place today. Today's business generated over 20% growth in 2024.
We're fixing to continue to do so and to accelerate over 2025, 2026, and 2027. And so we'll spend the balance of the day really putting some color around each of these things for you. In the end, these are our objectives over the next three years: to be the recognized market leader with revenue growth in excess of 20% CAGR over the plan period, execution on indication expansion into the meaningful segments of drug-resistant epilepsy patients. There's a meaningful segment. We want that segment covered by RNS. And we're going to develop the data and the indications to do so, really leading with IGE, idiopathic generalized, and pediatrics. Third, the category leader in efficiency and easy use, leveraging both our AI tools as well as remote capability to do that.
We have a fundamental opportunity to take the capability of the system and make it more scalable and efficient, both for customers as well as, as I talked about, organizational leverage, what we can do with our own organization to scale penetration of the implants and make support of these patients more efficient. Expanding awareness as well as access to RNS through direct-to-consumer campaigns and care. And in the end, doing all that, exiting the plan period with the business in a position to be in cash flow break even. And that's what we're going to do over the next three years and is laid out in this plan and what we're excited to spend the rest of the day telling you about. So with that, I'm going to turn the microphone over to our Chief Medical Officer, Dr. Martha Morrell. Dr.
Morrell is a luminary in the field and is somebody you're going to hear from a couple of different times today. And she's going to start by really giving us a solid foundation on the RNS System and what we see in the focal epilepsy population. Marty?
Thank you, Joel. I'll begin by telling you that Vikram, Mark, and I are very excited to be able to tell you about the NeuroPace story. And just to share with you how I ended up here, I was an academic neurologist running epilepsy programs when Frank Fischer, our founding CEO, came to me with a wild idea. And that was, since patients spend most of their time not having seizures, why don't we treat them just when they do and where they need it?
And that made a lot of sense to me, particularly since I had just come from the room of a patient who we were evaluating for epilepsy surgery. They weren't a candidate, and I realized there was nothing I could do for them. And now I'm in a situation where I can do a lot for my own patients. So I'm going to begin by just talking about epilepsy. So epilepsy is actually not one disorder. It's a group of brain disorders that share in common seizures. And those seizures are unprovoked, they're unpredictable, and they're recurring. I think many people don't realize how common epilepsy is. It's the fourth most common neurological disorder. And what we're talking about that makes so much sense is let's treat a recurrent electrical disturbance with electricity.
We also have an opportunity because there's a substantial number of people with epilepsy that aren't responding to anti-epileptic medications and really don't have other options. I want you to think about the epilepsies as falling into two broad categories. There's the focal epilepsies, and those are seizures that begin in a focus in the brain and then will often spread. There's a point, there's a region, there's a network that we can define as being the generator. Those patients, everybody takes anti-seizure medications, and they're also at this moment in time candidates for neuromodulation. They also, a very few of them, are candidates for a resective or ablative surgery at a focus in the seizure that is safe to remove. That is the minority. The rest of the patients we have really focused on. The other type of epilepsy is generalized epilepsy.
This occurs when there is a discharge that involves the entire brain. The only therapy for these patients is one medication after the other, after the other, because in this group, as with focal epilepsy, 30%-40% do not respond to anti-epileptic medications. We have an international organization called the International League Against Epilepsy that really discusses what the benchmark and what the goal should be. They define drug-resistant epilepsy as failure of two medications that are tolerated, and those medications have not achieved seizure control without life-limiting side effects. This is just, this is a different change in practice. It's two medications, two medications, and we know that there's less than a 5% chance that you're going to have a substantial response. So what should we be doing?
The standard of care, according to the ILAE and our own American Epilepsy Society, is you take two drugs and then you get referred to an epilepsy center for consideration of non-pharmacologic therapies, including neuromodulation and the RNS. What is the reality? This is what we have really had to do. We've had to carry this educational effort, and we've done that. The reality is that patients take one medication after the other, after the other, after the other, even though there's no expectation that they will achieve benefit. Oftentimes, they've experienced 10 to 20 years of intractable epilepsy before they get to an epilepsy center, or maybe they never do. The International League Against Epilepsy also partitions epilepsy into those two main categories that I talked about.
60% of the patients have focal onset seizures, those seizures that begin in a focus or region in the brain. And within that group are those who have seizures beginning from one or two regions or foci. Those are the patients for whom we have an indication for use. Patients may also have foci that are multiple. And this is a patient population that everyone, physicians and NeuroPace, has been thinking about as still amenable to our therapy, but in a more regional sense. The other type of epilepsy is where we're spending a lot of time. And those are that 40% of patients that I mentioned have generalized onset seizures. There's two categories of very different epilepsies within that group. One occurs in otherwise neurologically normal individuals who have epilepsy arising all of the cortex, all of the brain.
Most of these individuals are presumed to have a genetic epilepsy disorder. Those are the patients that we are addressing in our Nautilus trial. The other group are individuals who have epilepsy because there's widespread brain injury or problems in development or just dysfunction. Those are the patients that we are looking at for the Lennox-Gastaut study. We have not forgotten the 15% of patients who are children. We will talk about that. The ultimate goal is to have a very simple indication for use. The indication will be for the treatment of epilepsy. Let me talk about the therapy. The basic concept, as I mentioned, is most of the time the brain is fine.
So let's have a device that is placed at the seizure focus that's looking at the brain electrical activity all of the time that recognizes when there is a variation from normal that could predict seizures or is characteristic of pre-seizure activity. Let's provide an intervention, and then let's see if the patient needs any more treatment. How do we do this? We do this with the neuromodulation device that the whole system is self-contained. So it's placed within the skull, excellent signal-to-noise ratio. Patients can't see it. They can't feel the stimulation. Leads are placed according to the region of interest. And then the device is communicated with by a physician programmer or a remote monitor that the patient takes home. They hold a wand over the device, upload that device, upload information from the device.
And all of that is sent to a secure database called the Patient Data Management System or PDMS. This is a treasure trove. I think about it as the library for every individual patient. All of the device behavior, all of the recordings for each individual is there. So I, with a passport, am able to go into PDMS and open up the library for my patient. What does this mean? Well, I finally know what I'm doing. I can see what's happening. And rather than depend on a seizure diary that is often inaccurate because seizures may occur while patients are asleep, or they may not remember the seizures as part of it, or they just get tired of recording them all, and I can see what's going on and inform my treatment. I can inform how I address programming of the RNS System.
Vikram will talk later about other impacts this has on clinical care. Before the patient, I can turn that screen around and show them their information. Mike will talk about what that means. It's usually a really profound experience, really a moving experience as patients finally see what it is they've been facing. Let me talk about the kind of data that we get. There's a number of data presentations or representations that come from the device, including samples of the intracranial EEG recorded directly from the brain, directly where the problem is. This includes the samples of the intracranial EEG and also a recording of every time a detection has been made, a detection of abnormal activity, and when that has occurred. We can look at the cadence and the schedule. Vikram will talk about that.
What you see in the panel to the right is an electrographic seizure where you see the blue recorded directly from the brain. And what you see in the lower panel is the record of every one of those detections. I can look at that hours over the day, days over the week, weeks, months, and even years. And that really allows me to see objectively how my patient is doing. So that data is used so that I can be a much smarter physician. I can address interventions according to this data that I see, and I am able to monitor really almost in real time, in real time, how my interventions have affected the patient. So what do we have here? Well, there is no other therapy like the RNS System that really lets you know about the educational effort we needed to do from the beginning.
We have, as you will hear, an extraordinarily solid database that demonstrates effectiveness and safety in rigorous clinical trials. We have a device that can provide individualized care. It can be personalized to that patient's seizure signature, and it can be modified. It can be modified to do what the patient needs. It is non-destructive. We are not destroying the brain. We are not taking it away. It also does not preclude other options. We can continue to treat the patient with the other tools that are available. I talked to you about the therapy and the data, but I didn't talk to you about the flexibility. In our clinical trials, we have experience treating patients in every region of the cerebral cortex, including in the hippocampus at the temporal lobe, which is often an origin for seizures.
And now we have increasing experience with treating patients in the thalamus. We are also able to be flexible in what is detected and also how stimulation is delivered. The trials. We began with the first patient implanted in 2004 in a feasibility study for safety and preliminary evidence for efficacy. That was a two-year trial followed by the randomized control trial that led to our approval. That was a two-year trial. And then patients had the opportunity to enter a long-term treatment trial, which the majority of patients, a significant majority of patients chose to do to continue being treated for another five years. You add those up, and that is nine years of data in this patient population. This was the largest epilepsy trial in neuromodulation ever done.
It'sT getting even larger because of an FDA-requested and therefore performed post-approval study that began shortly after approval of first patient implanted in 2014. 324 patients at 32 centers across the United States followed for five years. We will present today the primary effectiveness endpoint, which is at the three-year time point. How well does this work? You know, imagine we have a therapy, a powerful technology, unlike anything that's ever been used before. So we had to look at each other and say, where do we start? We selected a starting point, and we selected correctly because we found in these patients with really intractable epilepsy that there was a substantial response that exceeds that of anti-seizure medications, 44% reduction in seizures. Look at that climb. It almost looks like Joel's figure about NeuroPace up into the right.
We saw an increasing improvement over time with patients having a 75% reduction at the time the trial ended. So what is going on? Well, you know, we got 28% of patients seizure-free for a period of six months or more. And these were patients with close to 20-year history of intractable epilepsy. But what happened during this time is us learning. And it also, what was going on is a unique property of neuromodulation, which is that people get better with time. So let's talk about a more contemporary experience. A group of investigators, including Dr. Rao, who is the principal author of this manuscript, combined 150 patients to say, well, how well is this working in the real world? And the answer is, wow, it's working much better. So we're getting outcomes similar to what we got at five years in our clinical trial.
By three years, we were getting outcomes that exceeded that we saw in the initial clinical trial. We found that the median seizure reduction was 82% within three years, and one in three patients had a 90% or more reduction. What do we see here? Better, faster. We'll talk about why that might be, but the learning curve, I would attribute a lot of it to us understanding what's going on. Much of that was contributed, of course, by the empirical experience in the clinical trials, but much of that is a result of the data. Now, a therapy is not really helpful unless it is meaningful in a patient's life. Of course, seizures are a big part of that. Quality of life, function, well-being are critically important, are as important.
As part of the studies that we did, we did very rigorous assessments of quality of life, of mood, and of cognitive function. We found that there were improvements in overall quality of life. You need to have that if you have a meaningful therapy. But really, we were somewhat surprised to see that there were improvements in areas that are often really problems for people with epilepsy. That includes subjective improvements in cognition and in mental health. Depression is a comorbid condition with epilepsy. Let's look at the quantitative outcomes obtained by neuropsychological or neurocognitive testing. We found that there were no negative mood or cognitive effects. In fact, there were improvements in those areas, particularly in areas that are often challenged in intractable epilepsy. That includes verbal fluency and aspects of memory.
We have a therapy that works and that brings meaningful benefit to patients. And I mentioned the power of the data and really the tools that this therapy brings to the management of persons with epilepsy. And there's no better person to talk, to speak to that than my colleague and friend and frequent collaborator, Dr. Vikram Rao.
Good morning, everyone. Thanks, Marty. It's really a privilege for me to be here to share my experience taking care of people treated with the RNS System now for over 10 years. But I think I'm going to start this talk on the power of RNS data with a thought experiment that I'll ask you to indulge me. So imagine for a moment, imagine that you're an investor or an analyst, and you're charged with picking a stock, but you have no historical information on the past performance of that stock.
And any quantitative information that you might have, what little information you do have about that company is probably inaccurate. And yet you have to do something. I'm not talking about NeuroPace, by the way. I just gave Joel a heart attack. But you have to do something with information that is inadequate. And that feeling, that unsettling feeling is really what it is to be a clinical epileptologist because epilepsy is, in many ways, a black box. There is no blood test that helps us gauge the severity of epilepsy. When we do short-term recordings of brain activity with EEG, they are often normal. Even people who have poorly controlled epilepsy spend the vast majority of their time not having seizures.
And when somebody comes to my clinic and they do the best they can telling me how it's going, how many seizures they've had recently, whether the interventions that we're doing are helping or not helping, do the best they can. And I know that their report is likely to be mostly inaccurate, unfortunately. And yet I have to do something. And I tell our trainees that epilepsy, clinical epilepsy, is the art of making incredibly consequential decisions with woefully inadequate information. And that is where data from the RNS System has been truly transformative. As Marty said, we call it a window to the brain, which is actually not just a marketing phrase. That is actually true. That is true.
This has been kind of a game changer in our field because the RNS System and data from it helps us to learn about a patient's epilepsy, to personalize the therapy that we offer to that individual, to monitor in a quantitative way their disease severity over long periods of time. And all of this, by the way, is with somebody in their home environment, normal activities, on medications over months to years. There is no other way in neurology to do that. And so this slide really shows a few examples. This is certainly not, it's a busy slide, and it's not exhaustive. It just gives you a few anecdotes of what we do as clinicians with this data, this really precious data set that we have. In the upper left, you see what happens when we introduce a new anti-seizure medication.
Sometimes we see a precipitous drop, an immediate drop in the rate of detections of abnormal activity, like the sparks that the device is hearing, suddenly plummet. That portends a favorable long-term response to that medicine. If we don't see a drop like that, that suggests that maybe that medicine isn't the best one for the individual. We talk about tracking seizures, and you might think we could just ask our patients, well, how many seizures are you having and base our therapy based on that? But in fact, it's not that simple. It's a very common scenario in my clinic. When I ask somebody, how many seizures have you had since I last saw you? And they say, I've had three. And I look at their RNS data and say, you've had 30. That happens all the time. But with the RNS System, we can track electrographic seizures.
That just means an electrical recording of a seizure that's objective. I can see it in the data. I can count it. I can trend it. I can relate it to interventions that I'm doing, behavioral interventions that I might make, a new medication, a surgery that I might offer a patient. I can track biomarkers of a disease. I can look at the rate of sparks of abnormal brain activity in our patients, for example, and look at interesting trends like day-night variation, so-called circadian variation. We can look at cyclical fluctuation in relation to things like menstrual cycles. And all of this helps us to optimally manage a given individual. What do we do before the RNS System? I mean, we did something, right?
What we did, the gold standard is still bringing somebody into the hospital to a specialized inpatient unit called the Epilepsy Monitoring Unit or the EMU. And this is a place where we glue on some surface electrodes. We record brain activity over a handful of days, usually not more than a week. Insurance companies give us grief if we keep people longer than a week. And we do all kinds of things. We keep people up at night. We sleep deprive them. We strip them off their medicines. We shine lights at them. We make them have seizures. And after four or five days, we conclude, okay, we understand what's going on with their epilepsy. And that is really kind of a problem because epilepsy, I would argue, is the most dynamic disorder in all of neurology.
And so recording somebody for four or five days is just a massive temporal undersampling problem. Okay, not to milk the finance analogy too much, but that would be like tracking the Dow for like four days and seeing it downtrend and concluding that we're in a recession. I mean, that's ridiculous, right? Because you know that this is an intrinsically noisy variable that has to be followed for longer periods of time. And it's actually the exact same in epilepsy. If I have somebody for four days in my EMU and I record two seizures from the left side, am I very sure that they couldn't possibly have seizures from the other side? Sometimes not. And I think that was really nicely exemplified by this study, kind of a classic study now by one of my colleagues, Dr. King-Stephens. And this is a complicated slide.
Let me walk you through this. It's actually simple what they did. They went to the RNS System clinical trials. They pulled out the 82 patients who had one lead wire in each temporal lobe. Why? Because the clinicians who were taking care of them had reason to suspect that this person, these people, could make seizures independently from both temporal lobes. So they were implanted bilaterally, and then they were followed. And this graph, each bar is one of those individuals. And the length of the bar is showing you the duration of time between when you record one seizure from one side and at least one seizure from the other side, meaning how long did it take to prove that, in fact, yes, they could make seizures from both sides of their brain. And the punchline is that it takes a surprisingly long amount of time.
In this study, an average of almost 42 days, with two-thirds of people taking more than a week to prove that they could make seizures independently from both sides. Why does that matter? That matters because I just finished telling you that our EMU admissions typically don't extend beyond a week, so again, a temporal undersampling problem. This cartoon sort of makes that point in a different way, so the cartoon characters here are shaded either blue, which means all seizures from the left, purple meaning all seizures from the right. And the shades of color are something in between left and right seizures. The top row is what you saw in the EMU. The bottom row is what you see in long-term real-world follow-up with the RNS System.
And the X's indicate the seven out of 12 patients where there's a mismatch, where the conclusion that you made during your short-term recording in the artificial circumstance of the hospital is different than what you found in long-term real-world follow-up. Why does that matter? That matters because if we incorrectly lateralize somebody, we might refer them for an incorrect surgical procedure. And once you take out a part of the brain, you can't put it back. So it's a big deal. It's really a big deal. We don't want to make incorrect surgical decisions that don't help people. There's a lot of data that comes from the RNS System. I made this figure a few years back just to summarize some of them. I wish I had time to really tell you about all of the different forms in which data comes to us.
And clinicians, if you ask them, always want more data, right? But the response to getting that data, usually, in my experience, is sort of a biphasic response. First response is, yay, we have a ton of data to look at. Second response is, oh darn, we have a ton of data to look at. And that is actually the clinical challenge. The challenge is taking mountains of data and distilling from that the core information that actually helps me manage a patient clinically. Because I don't need to see everything. I just need to see the most important things. And so what do I do? Well, typically, I manually sort of look through all the recordings, all the data, extract the things that I think are actually useful. And that is fine. There's a certain fun to that, but it does take time. It is not efficient.
It is not scalable, and luckily, all of this is about to get much more streamlined. This is the landing page for the Patient Data Management System, PDMS, that Dr. Morrell told you about. This is where clinicians go. You fire up a patient's account, and this is where we can go and look at graphical representations of trends of activity. If you're with a patient, you can look at live streaming recordings of brain waves. You can look at a library of stored clips of brain activity data. You can look at current and past programming settings on the device, and the green pane is sort of an aspirational or emerging functionality of PDMS. That's called AI-Powered Seizure Trends, and when clinicians click on that box, it will open up essentially a toolkit of things that are going to make our lives a lot easier.
One of those tools is a sorting tool. So we have this library of many, many hundreds, sometimes thousands of recordings to look at. And sometimes the only ones that I actually care about are the ones that have seizures. That's really what I want to know. That is the essential information. And I can go through with my eyes and pick out the seizure recordings. I can do that. But it turns out that a well-trained AI-based algorithm can do that as well, maybe better than me. And that has now been validated and soon will be a functionality that, with a click of a button, an AI-based tool can pull out all the seizures and show them to me really in a simple way. And we think that that will not only streamline the process of programming these devices, but actually save clinicians a lot of time.
Another huge time-saving tool that will be available is shown here, so one of the most, I'll say tedious, but I'll maybe time-consuming is the better term, is teaching the device how to identify seizures. At the end of the day, the RNS System is a computer, which means we have to tell it what to do, and we want it to detect seizures as early as possible. Why? So that when it detects a seizure, it can intervene and try and stop that seizure. The better we can detect seizures, the more effective the therapy can be, and currently, we configure a variety of computational tools manually to identify seizures with high fidelity. That's fine. That's what we do, but it turns out, again, AI can do that better than me.
Through validated algorithms now, there are ways that AI can select the onset of a seizure and do probably a better job at helping me treat that seizure earlier than I would have otherwise and also save me a lot of time. I take care of about 150-160 people treated with the RNS System at my practice. I would be lying if I said to you that every single one of those people has a fantastic outcome with the RNS System. That is definitely not true. Some people are not doing as well as I would like. That is the reality. But I have to believe that a part of the reason why they're not doing better is that I have not programmed the optimal settings on their device. Why? No lack of effort on my part. It's just that I don't know.
This is largely empirical. Again, it's a largely black box sort of enterprise here. I don't know what the best settings are for that patient. But now that there are actually thousands of people treated with the RNS System, why not leverage that big data and have an algorithm go and find patients with clinical similarities to my non-responding patient and to show me patients who are doing better and leverage knowledge from that to apply to my patient? That seems like a pipe dream that is now actually going to be a reality. Deep learning algorithms can mine large data sets. They can extract information that will suggest settings that will be better for my patient.
A method called Bayesian optimization is just a way to use data from every patient to predict the next best settings for my patient and hopefully achieve better outcomes in a way that I don't have to be any smarter. I can actually lean on the algorithm to guide me, so again, the take-home message from this section really is that the tools that are on the way are not only helping to save clinicians' time and make things easier for us with the presumption that the easier this is for clinicians to use, that will drive volume, but also, more importantly, to improve efficacy for seizure reduction in our patients, so with that, I'm going to turn it over to my colleague, Dr. Richardson, who I think is joining us by Zoom.
Okay, thank you, Vikram. Thanks, everyone, for allowing me to speak to you remotely.
I'm sorry I couldn't be there in person, but I'm very excited to talk about the RNS System, which I think is literally the most exciting thing there is in medicine. Now, obviously, I'm very biased. I'll give you a little bit about my background. I direct the functional neurosurgery program at Massachusetts General Hospital, which means I do epilepsy surgery and other surgeries to implant devices in people, for instance, with movement disorders or psychiatric indications. I've been here for about six years, and previously, I ran the epilepsy surgery program at the University of Pittsburgh. Next slide, please. Okay. Thanks, Marty, for jumping up there and doing that. So there is a huge treatment gap for patients with drug-resistant epilepsy. There are 1.2 million people in the U.S. on medication that is not working, having intractable seizures.
Only 50,000 of those are admitted to an epilepsy monitoring unit, a comprehensive epilepsy center. And less than 1% of this total number are treated surgically, only about 6,000 or 7,000. So here's a little more detail on recent growth in surgical procedures. Epilepsy surgery is growing. We're training a lot of people in functional neurosurgery who want to do epilepsy surgery. In both places, established practices where the epilepsy surgery volume was kind of mid-level, it's exploded both times. These patients are out there waiting. Recently, the total—we'll just stay on that slide for a second—the total number of resections actually declined a little bit. And that's been replaced to some extent by LITT, laser interstitial thermal therapy or laser ablation. You see the RNS implant numbers are increasing.
I think we're really at an inflection point where this is going to increase quite significantly as we have the same biases against resection and destructive procedures in the brain and just increasing knowledge about how to use the data and how to implement RNS therapy in broader populations of patients. Next slide, please. Okay, so why does this treatment gap persist? Quite simply, it's really hard to treat these people. And many have types of seizures and types of seizure networks that you cannot cure with surgery. And so the traditional way of thinking about epilepsy surgery is you have to find one spot in the brain or one region and remove it.
A lot of biases against surgery are because there has also been a philosophy in some centers that you do that at the cost of function in some cases, where the goal is just to make the person seizure-free. There's some biases against traditional surgery. Examples of those cases that are very hard to treat with traditional methods are seizure onsets that start from both sides of the brain, those that start in eloquent cortex. That's cortex that you cannot remove without permanently disabling someone's speech areas or motor control areas. They're part of a larger network. You can remove part of that network, but the seizure still persists. You put someone through the risk of a traditional surgery without giving them a surgical cure too large for complete resection. Those that are coming from multiple areas of the brain are generalized epilepsy.
The reality is that every single one of these patients can be treated with RNS. All of those 1.2 million patients from the slide before, they can all be treated. Some of those are still going to be best suited for potential surgical cure with the traditional resection. If you have a family member or child with epilepsy and they have a spot you can see on the MRI or otherwise diagnosed and it's safe to take it out, it should come out, and they may be seizure-free for the rest of their lives. But the really rewarding thing that has happened in my practice has been the use of this device to bring huge quality-of-life improvements in people who otherwise just continue to suffer and suffer because they have a more difficult seizure network. Next slide, please.
Okay, so what's changing in epilepsy surgery despite the kind of modest changes in the numbers so far? Why do we have such reason to believe that there's room for really dramatic improvement and soon? And these are the reasons, really paradigm shifts in how we think about epilepsy surgery. These have been driven to a large extent by the RNS device because for the first time in the history of treating any brain disorder, you can put a device in and record activity and make decisions based on that activity. And I should say this is a safe surgery and easy surgery. So that's another way it's really differentiated from these other techniques. I think that's important for you to know and something we probably surgeons don't talk about enough. The hard part is figuring out the right place to put the leads.
Vikram has a harder job in some ways interpreting the data and using it than I do to put the device in. That's straightforward. In having increased use of RNS, being able to record and simulate essentially wherever we think is best based on the data we have, we've really changed our thinking, first in terms of defining success. It's really clear that you can absolutely change someone's life without making them seizure-free. You can maximize their quality of life. People can go back to driving sometimes when they have the RNS. They can continue to work. They can reduce the amount of medication they're on. They have some knowledge about their seizure onsets and a likelihood for when they may have a seizure. They have some control. They can exercise some control over their disease. This is huge.
These are really big quality-of-life improvements for the patients. The way we think about the treatment and not having to remove the whole seizure focus, this has brought a lot more patients to epilepsy surgery conference than previously. These patients who weren't thought to have a focus that you could resect safely just wouldn't be presented for a surgical option. It literally wouldn't be discussed. It was, well, there's nothing we can do here. Now, the majority of patients that come through our epilepsy conference are complicated patients where RNS is a potential option, if not the main option going in. And this has been a marked change in the past several years in my practice. And certainly, let's say I started MGH in 2019, and it took a couple of years really for this to kind of turn around.
And now the majority of patients that come through, whenever we do an intracranial monitoring surgery, if we need to do that, we're implanting parts of the brain that are important for figuring out where we might put an RNS. This ties into this localization goals paradigm shift. So not only are we trying to inform our decision for the part of the brain that could be removed, if that isn't an option or if it fails in the future, we need to know where to do neuromodulation because, again, we have the device that can record seizure onsets and stimulate, and that's customizable to each patient's data, including their intracranial data that we can get diagnostically in a diagnostic surgery. Patient management. This is just a massive benefit. Vikram's talked about this, and you'll hear more about this in other talks. Next slide, please.
Okay, so this gives an overview of a couple of things. One is the state of the art in epilepsy surgery, and the other is advantages of the RNS System over other systems that aren't closed-loop systems and can't respond to the EEG tracings in the way the RNS device can, so targeted stimulation is the way that RNS was envisioned. It's the way that most people began using it. You would put this on both sides of the brain, so bilateral mesial temporal lobe epilepsy or temporal lobe epilepsy in general is fairly common. It's one of those types of epilepsies that everyone knows you can't take out both temporal lobes, so you can do neuromodulation and other options where, let's say, it's a language area. Someone doesn't want to take a risk for language dysfunction with a resection or in a motor area. This is a targeted stimulation.
This works very well. What we found over time in combining the RNS with, you might think of two categories of approaches. One is in combination with a traditional surgery like a resection or a laser ablation. The other is using the thalamus. The thalamus is a hub for information flow between cortical areas. It can put the brakes on seizures when you deliver stimulation there. Many types of seizures are recorded very quickly in the thalamus. You can affect a wide cortical area with stimulation in the thalamus. You cannot do that with traditional surgery to affect such a large cortical area. You can only do this with neuromodulation. You don't have to use a thalamus if you have something that's localized to a region of the brain.
It might be better to put the recording and stimulating electrodes both at the cortical level, and then these combination therapies. This is something that we do fairly frequently now. There may be large malformations or multifocal types of networks where you can operate with a traditional surgery in one part of the network, but then follow up, but you know that's not going to be curative, so you follow up planned with RNS therapy to address the rest of the network. This is kind of like doing resection to reduce burden in cancer and then following with adjuvant therapy that's going to keep the patient alive for longer after you do the surgical resection.
The reason I said that slide also illustrates the advantages over, for instance, deep brain stimulation is that you can see through just the schematics on that slide that you can personalize the therapy to the type of epilepsy. The way that people use deep brain stimulation in epilepsy, it's very generic. It's kind of like VNS. They put the leads in part of the brain called the ANT, so Anterior Nucleus of the Thalamus, and you just turn it on. One of those devices now, you can look a little bit at the data, but you can hardly see anything longitudinally like you can with the RNS System in terms of the signal. What the figure on the right shows are the structural connections between three, what we think of as canonical nuclei in the thalamus. There are a number of nuclei in the thalamus.
These are the big ones that historically have been tied to modulation in epilepsy. And you see the ANT connections are in the red. That's connected to about a third of the brain. And these other nuclei that we're using kind of quite often now in RNS, still with ANT, but also centromedian nucleus and the pulvinar, are connected to other areas of the brain and have particular use in different types of epilepsies that can be defined. And the field is doing a much better job of figuring this out. And this is kind of the basis for the Nautilus trial, this idea of the centromedian nucleus being widely connected to areas that are important in generalized epilepsies. This is also the reason. Yeah, next slide, Zoe.
This is also the reason why you see the RNS data supporting the idea that if you choose the right spot, you can put this anywhere in the brain, and it should work for the type of epilepsy that you're treating. And that's supported by the kind of ongoing data coming out from different groups. And it's in contrast to this extended data from DBS studies that show that the efficacy is just not as good when you use a kind of generic approach to put DBS in just one part of the brain or thalamus. Other advantages are illustrated here. The RNS intracranial data, you can pick up a seizure in any of the areas you implant as long as the hypothesis is correct. So in terms, it's really interesting. If you have failure of the RNS device to work, that's very helpful information.
You have objective data that shows you where the seizure is not starting, where it's not propagating, where neuromodulation is not effective. And it's not that difficult to go in and revise one of these systems and put the leads elsewhere based on alternative hypotheses. So you're kind of never at the end of the road with this device. And of course, if it is working, then the data has been used to guide stimulation settings. Importantly, there really have not been acute or chronic side effects with therapeutic stimulation. This could be for a couple of reasons. One is a customizable nature. So the programming neurologist can look at the recordings and make sure they're not stimulating physiologic signals. The open loop or DBS approach, it's the device just on. It's not discriminatory.
It's stimulating whether or not the parts of the brain are doing something that they should be doing or whether or not they're generating a seizure. I think this probably has to do. It leads to this problem of a greater report of things like sleep arousal, memory impairments, depression with ANT stimulation, what we call duty cycle or open loop stimulation. The device is just on all the time. Some of these patients do just fine with that, but others, it's disrupting something in their memory consolidation, their mood regulation, or sleep. These are problems that can be avoided with personalized responsive stimulation. Next slide. Okay. The entire field is aware of this gap in treatment. The ILAE, the International League Against Epilepsy, recently included neuromodulation recommendations in their updated guidelines. This field does move slowly, but things are finally changing.
And this is a big deal to point out to epilepsy groups around the world that, yes, if you have a patient who may not be a candidate for resection, but they're medically intractable, they deserve an epilepsy workup because we have neuromodulation. And we're seeing this really around the world, certainly in the United States, more and more patients that come to epilepsy surgery conference who are neuromodulation candidates primarily. And this is a big change. And Marty mentioned in her opening presentation that's the minority of patients who you can get a surgical cure with the traditional surgery. And I wish that wasn't the case, but it is. And so we're really going to see, I think, think back to that chart in one of my slides earlier that shows the amount of surgeries based by category. We're eventually going to see a big switch there.
The vast majority of epilepsy surgery is going to be neuromodulation. It's not going to be traditional surgeries to take out a part of the brain for a cure because those are the minority of patients. Okay, next slide. So I think this is the last slide. This is the cost of doing nothing. People die, and their quality of life is really horrible. One in 150 people with uncontrolled seizures are going to die from something called SUDEP, that's sudden unexpected death in epilepsy. The amount of lost productivity and quality of life in these patients when they don't get referred and don't get a treatment is quite significant. So there's absolutely no question that the field is going to continue to address and push and push on this.
And it's also quite clear to me that the best treatment we have to do this is responsive neurostimulation for all the reasons that we've talked about here. So I'm obviously very excited to continue. We have a lot of work to do, but it's really exciting, and it's helping a lot of people. It's going to continue to do that. So thanks for the opportunity to talk.
The next individual is this really special presentation. Mike McKenna is going to share with you his story. He is someone who participated in the trials of the RNS System. He is someone who is living with epilepsy. I ask all my patients to talk to Mike to really understand the patient's perspective. So Mike's going to walk us, please.
Thank you, Marty. And thank you, everyone, for being here. Hi, way over there. So just a little background.
As you heard, Mike McKenna. Over eight years ago, I joined NeuroPace as a patient educator after volunteering as an RNS ambassador for about two years. I've seen the importance through the years of patient education for patients who are considering or moving forward with the RNS System. And patients and families have been very appreciative of how NeuroPace provides educational resources for them to understand what this device is. As you heard, I've lived with epilepsy for over 35 years. I mean, 25 years, and I've been using the RNS System for over 16 years. During my RNS journey, I've graduated with a graduate degree in social work from both ASU and Mayo Clinic School of Health Sciences and worked as a social worker in Mayo Clinic in Phoenix before moving to Portland, Oregon, where I live today.
I wrote this down so I can stay on track, keep on the time frame. As you heard, 25 plus years ago, I was living a life that I loved. I was healthy, recently married, living in a newly purchased home, and had just become a partner at a successful construction company building high-end custom homes. Building homes allowed me to be creative and engage my imagination. I had no idea that everything that I loved would soon be shattered by epilepsy, leading me to lose everything. I've lost track of the number of times I've woken up in an ambulance or room after someone called 911 because I had a seizure.
I was experiencing focal aware and secondary generalized seizures that resulted in broken bones, dislocated teeth, I mean, broken teeth, dislocated joints, torn muscles, lacerations, my heart going in AFib or stopping, prosthetic joints and replacements of those joints, and just constant mental chaos. Through the years, I've tried different seizure medications. For about four years, I tried VNS therapy. I completed intracranial testing to figure out if I was going to be able to have right temporal lobe resection surgery, but during that 14-day stay, they decided that that was not an option for me because I had seizures coming from both temporal lobes, so I slowly lost my identity and feeling of self-worth. I felt helpless. I was now my epilepsy diagnosis. I started accepting that my doctors might not find a seizure treatment for me that worked.
I found this treatment insight the motivating factor for me to change my mindset. Instead of me waiting for a seizure treatment to be found, I needed to be in charge and advocate for myself and ask questions, finding ways to move forward. My doctors mentioned that there was this new intracranial medical device, the RNS System. So this is a demo device of what the device looks like. Mine is placed right there. It could be placed on different parts of the skull. For reference, it's about the size of your thumb. So if you look at your thumb, that's a computer, battery, everything's in there. So the doctors told me that this new device could monitor for seizures. It could respond with pulses of stimulation that I won't feel, make EEG recordings of those seizures, and report that data to my doctors. I mean, it sounded unbelievable.
It sounded like something I was really interested in. So I said yes to RNS surgery. I was in the hospital for two nights. Once home, my recovery was very quick. I was sent home with a remote monitor. That's the way I was able to transfer my data and collect it. And I was instructed to do that once a day and transfer my data over the internet once every one to two weeks. It was very easy to do and took about three to four minutes to do. A little over a month after surgery, I had my first follow-up appointment. In the waiting room, I was super nervous and super excited to see what my new RNS device might have captured. And what I saw was this: my first recorded seizure, my seizure fingerprint. I mean, I was in awe seeing this.
These were my seizures that I had always feared the most, the ones that would happen in my sleep, the ones that would result in me dying from SUDEP. I mean, look at the time of this when the seizure happened. It was at 12:57 A.M., within the first hour and a half of me falling asleep. I had an emotional release because I now felt validated that I wasn't just making all this stuff up. I finally felt heard, seen, and acknowledged that, yes, I was experiencing seizures. I felt reassured that the RNS System just might help me. I felt empowered. People living with epilepsy many times refer to seizures as their brain on fire. So when I saw this other piece of data related to this recorded seizure, I was just like, "Wow, like my spectrogram." I'm no expert, so this is my layman's electrographic definition.
This is a visual representation of my brain activity that shows my brain seizure strength because of that seizure and as it progressed. I guess I could say that this shows where my brain caught on fire. Remember I mentioned these seizures in my sleep or what I feared the most? Look at the time of this seizure when it happened, 5:56 A.M. Over the years, my RNS data has shown that my largest seizure happened in the first hour and a half of me falling asleep and the last hour and a half before waking up. On this ECoG, you can also see where the seizure started and the stimulations that were provided to try to stop that seizure. This data helped guide my therapy, and my doctors made some adjustments to my stimulation parameters.
This was over 12 years ago, and this was the last time my stimulation settings were adjusted. This was also the last seizure that I ever experienced. Because I'm no longer experiencing seizures, we're no longer making stimulation adjustments. I now only download my data once every one to two months, more kind of like a maintenance check. When I see my doctors now, it's every 12 to 16 months, and my appointments are about 20 minutes in length. This slide is 16 plus years of all my events, all my data. This shows when my RNS System did something as making an ECoG or providing stimulations based on my seizure detections. My neurostimulator responds with pulses of stimulation within milliseconds of my seizure fingerprint showing up. Stimulation doesn't mean I'm experiencing seizure.
It just means that those spots in my brain that have the potential to grow into seizures are being stopped. With these slides showing my RNS data, you've seen my life-changing outcomes, such as my seizure reductions. You've seen how my therapy was personalized using those unique seizure ECoGs. And you've seen some empowering insights to seizure trends and cycles and how these insights helped guide optimizing my treatment plan. You've heard today how the RNS System is transforming the lives of people suffering with epilepsy. So I thought it would be interesting to use my all events histogram and repurpose it by adding some of my major health and life transformations from the past 16 plus years. So in 2008, I entered the pivotal study, and I started college. And a week and a half after starting college, I had my RNS surgery.
I was back in college after a week and a half. In 2012, that was my last seizure. After making stimulation and medication adjustments, lifestyle changes, I became seizure-free. In 2014, this was my last Mayo Clinic appointment. Being seizure-free, I no longer needed to keep seeing my doctors at Mayo, and so we decided it was time for me just to see a primary care provider. In 2014, after six years, I graduated with master's degrees in social work and was no longer a full-time student. That was really exciting. I was able to read a regular book, and I started planning for a career. In 2014, I mean, 2015, I moved to Oregon, and I bought a small home. No longer needing to live next to Mayo Clinic, I decided to move to a city in Oregon that was more pedestrian-friendly.
With the help of my parents, I bought a small house. In 2016, I started working at NeuroPace. For two years prior, I volunteered as an RNS ambassador, helping try and embrace epilepsy and RNS awareness. In 2016, I started providing patient education on a part-time contracting basis before being hired to do it full-time about eight and a half years ago. In 2017, I stopped my SSDI. After working full-time for a year, I decided to call Social Security and let them know that I no longer needed my monthly disability payments. In 2017, I filed taxes. This was the first time in over 15 years that I earned a salary. On SSDI, I was receiving $1,200 a month, and $850 of that went to rent. So you can imagine living on that. In 2019, my credit score became excellent.
I mean, this opened up the ability for me to get credit and get a loan if I needed it. Before I started working, my credit score was around 520. In 2019, I got my driver's license reinstated. This was a major milestone, I mean, life milestone. I regained a sense of independence by not having to rely on others to go anywhere, and then I just started thinking about my future, investing and planning for retirement one day. You know, living with uncontrolled epilepsy, this was never in my thoughts. So how am I doing today? My epilepsy journey continues today, but what's different since the start of my journey over 25 years ago is today I'm in charge of my journey. I decide who I am, what I'm going to do. Yes, I still have epilepsy.
Yes, I still take one medication at a low dose once a day. And yes, I use the RNS System. But I've been seizure-free for over 13 years. Today, I live a pretty normal life. Is it the same life that I lived 25 years ago? No, but it's so much better. As a NeuroPace patient educator, I get the chance to help patients and families who have questions about using the RNS System as one of their treatment options. These life and education opportunities have given me a reason and a purpose as to what I went through all those years living with uncontrolled seizures. And in closing, this is just a photo of a small group of people. They're all living with the RNS System who have been able to start living their better lives and find and enjoy their life milestones too.
So thank you, everyone, for that. Thank you.
Thank you, Mike. Thank you for sharing your amazing story.
You're welcome.
Now, I'd like to invite our panel, our speakers from the first part of the day to come up, and we'll do a Q&A session. We'll take some questions from the audience. So Vikram, Marty, Joel, we'll have Dr. Richardson.
You've got to stay up here too.
Just sit. Joel. Great.
Thanks.
Why don't you sit down, Marty?
Thank you, Rebecca. Thanks, Mike and Mark and Vikram and Marty. And so we would like to, we are a few minutes behind, so we'll try and move with alacrity here, but take a few questions. Anything from the audience? M ike Kratky.
I'll start. Joel, did you mention accelerating? Thank you. Did you mention accelerating growth between 25 and 27 as part of the LRP?
How should we think about the cadence of growth within that? And then I have one more.
It's a great question. And we do see the cumulative effect of the work that we're doing across the different clinical development, product development, and market development all giving us more shots on goal from a growth perspective. And so I think if you think about the growth that we're delivering today and you think about all that, we do see ongoing opportunities to accelerate growth over the planned period. We're not guiding beyond 25 today, but we do absolutely see where we're building a portfolio of growth opportunities here that we think will contribute to that.
Got it. Understood. And maybe one for Dr. Morrell. Can you talk a little bit about your level of confidence in the Nautilus Study from a scientific rationale standpoint?
What gives you confidence that this is going to be a study that can be effective and successful based on some of the things you mentioned earlier?
You know, we're going to talk about the NAUTILIST this afternoon. And so let me hold on some of that. But I'll tell you that we know epilepsy and we know how the RNS System works. There are sufficient publications that FDA decided they'd grant us a breakthrough designation. And then I'll show you where we are and how we are looking at expansion using really good clinical trial data.
Thanks very much.
You bet. Go ahead.
Thanks so much for taking the time, everyone. Rohan Patel from J.P. Morgan. I guess the first question I had for you. Oh, thanks. The first question I had for you, Joel, was actually about your LRP targets for cash flow break-even.
Just wanted to clarify, is that a 2027 comment over your revenue CAGR LRP? Ad also, if you could just walk through some of the levers there and what's driving the operational leverage over the period.
It is. When we think about cash flow break-even, the point for us is to exit the long-range plan period in a position of being cash flow break-even. So we'll march our way toward it over the plan, but 27 being that period of time where we expect to maintain a cash flow break-even status for the business. And the most important thing, and many of us have talked about this, the most important thing we can do to improve our cash burn rate is to grow the revenue line with gross margins where they are. It's a growth story. And it's a growth story with gross margin discipline and operating expense allocation optimization.
That's what we've been doing. We've made a fair bit of progress on changing cash, and we see that continuing over the plan period. That said, we have not hesitated, and we will not hesitate to invest in the growth opportunities that we see in the business. We think they're real, and they're in flight. And so you got to invest to take advantage of those, and we're going to. But so yeah, it's a growth story. And to Mike's earlier point, the accumulation in my affectionately referred to, we call it the ramp slide, that growth is what really drives the change in the cash profile of the business with then good operating discipline along the way.
Got it. And then I guess the second question I had was for Dr. Rao. Thanks for sharing all that exciting information about some of these AI software tools.
Just kind of, I was hoping if you could elaborate a little bit more on some of the applicability for those tools outside of the traditional kind of epilepsy clinical manifestation. Maybe if you could provide a specific example or maybe something that's where the anonymized data is valuable outside of an area that RNS targets today.
Well, that's a good question. One example that comes to mind is, and it is still in the epilepsy space, but I think it speaks to something broader, which is there are patterns in the data that we have seen by virtue of the chronicity of the recordings. If you do a one-week study, that's one thing. If you look at brain activity that unfolds over months to years, you see something very different.
We can see that there is actually a temporal structure and an organization to the data that we never really knew existed, and that enables all kinds of things, like, for example, anticipating when a person is likely to have their next seizure. This, by the way, is not a functionality of the RNS System in its current form. This is research, but I think it speaks to something broader and an ability to predict or to forecast, as we say, when somebody's going to have their next seizure or be at risk of a sort of significant event. We've never really been able to do that, so that's one example. I don't know if that really gets to your question. Yeah, I guess
I was thinking more like outside of epilepsy, outside of seizures. I see. We will mental health and everything. Like neuropsychiatric seizures.
We'll be talking about that this afternoon. Dr. Samr Sheth, who does a lot of research outside of epilepsy, will be discussing that. And I will also be talking about a use of the data in a collaboration with an early pharma company to demonstrate target engagement, which provides us an opportunity to really think about utilizing this data in a way that is of benefit to NeuroPace and benefit in therapeutic development.
So you have to come back after the break, Rohan.
You do. Yeah, you do.
Mike.
Thank you, Mike Polark, Wolfe Research. Probably for the doctors, jump ball. Question on, let's assume a successful Nautilus study and indication expansion into generalized within neuromodulation. How do you think about segmenting the patient's VNS versus RNS? My understanding is VNS technically doesn't have a generalized label in the U.S., but is used frequently off-label.
I think OUS, there is some specific regulatory language. Is there overlap in those patients? If so, how much you see it? Relative to your practice today, VNS versus RNS, again, assuming a successful Nautilus study, what does the VNS-RNS mix look like in three to five years?
Yeah, I'm going to answer by, again, telling you to come back after lunch. We can kind of show in our post-approval study how practice patterns are changing with RNS coming first. I think Mark Richardson would be a really good person. Mark, how do you decide whether you want to use VNS, the standard fixed anatomic target continuous stimulation DBS, or RNS?
There we go. For our group, this is pretty straightforward now.
I mean, we basically never put in a VNS because we think it has the least chance for efficacy, and you can't report any data. So I describe all the options for the patients, but I tell them that's my least favorite for those reasons. Very rarely someone will choose a VNS. I think if you looked at how quickly the Nautilus trial enrolled, there is a real desire to give these generalized epilepsy patients this therapy and a therapy that's better than VNS because even in centers that do a lot of VNS, the reason they do a lot is because there is no other option. They'd rather do that than nothing. We've, for instance, flipped our pediatric group here who used to do a lot of VNS for things like tuberous sclerosis, other pediatric generalized disorders. Now we go to RNS first.
When you combine the ability to get the data and customize the therapy, you can cover a wide range of these generalized epilepsies without flying blind like you do with VNS. VNS is not in the brain. It's not without side effects. It can be painful. It can affect people's voice and so on. When you explain to patients the difference in potential efficacy, it's really, really rare for someone to say, "Well, I'd rather you put this thing in my neck than do the brain surgery." They understand the brain surgery is really safe and really personalized. This is a very easy decision for us. I think you'd see this trend kind of continue.
Thanks, Mark.
And if I can just comment one second, Joel, I just looked at our 2024 numbers, and we removed more VNSs than we implanted for all the reasons that Mark.
Oh, yeah, I definitely remove more VNS than I implant, for sure. Sorry, Joel, go ahead.
Not at all. We are up against it here. Appreciate all the questions. We are going to take a 10-minute or just maybe a little bit less break, and we'll come back in and start as close to 10:30 A.M. as we can. So thank you.
And we'll invite Dr. Morrell to kick off the second part. Here you go.
O kay. We've been promising all the earlier new data, and I get to deliver it now.
So I'm going to talk about our newer clinical trials, and I'm going to talk about other applications for the RNS system relevant to questions that were asked before the break. We are in clinical discovery mode in two areas. One is to collect data to expand indications for use. I'll share with you the primary efficacy endpoints for the post-approval study. You know about the Nautilus study. NIH has funded a study in a very devastating symptomatic generalized epilepsy, Lennox-Gastaut. I'll share that with you, and exciting news about a true collaboration and partnership with FDA to understand how to use the real-world data that is accumulated in pediatrics. We'll also talk about what AI is doing. We invested in AI before it was popular over five or six years ago, understanding that our database was so massive that there was no way we could figure it out.
We are finding objective neurobiomarkers, and we're using those biomarkers to be really smart about programming, detection, and stimulation, and also tracking the patient's response. So the post-approval study is the largest FDA-required prospective post-approval study. It is a really tight study design. This is for patients newly implanted with the RNS System, the first enrollment being in 2015, and the final implant five years later. We have the final patient to three years, so we've been able to look at the primary effectiveness endpoint submitted to FDA, and it was accepted in December. We will be presenting the results at the American Academy of Neurology in the spring, but you're going to hear them first. So let me talk about the patients who enrolled in this trial.
271 of them stayed in the trial for three years, which is pretty impressive because they were living their lives and moving and starting new jobs, and they stayed in the study. For context, the average frequency of seizures when patients enrolled was 5.7 per month. Similar to the pivotal trial, we had a majority of patients who had more than one focus. We had patients with seizures everywhere, neocortex and also mesial temporal lobe, and then a number of patients who had onsets in both areas. We found that about a third of patients did not have intracranial monitoring before they were treated in the RNS System, similar to our pivotal study, but here's a big difference, and I talked about RNS moving earlier in the treatment paradigm.
So whereas in the pivotal trial, a third had been treated with a prior epilepsy surgery, largely resection, that's down to less than 20% in the post-approval study. And look at the drop in use of vagus nerve stimulator. 32% of patients had been treated with a VNS, and in our more recent study, it's down to under 8%. Here are the results. We have data out to three years, and I'm showing it to you in six-month increments. You'll see the sample size, and in purple, you'll see the median percent change. In parentheses is the response of the upper quartile and the response of the lower quartile. You will recognize that we're seeing very impressive responses really early in the course of treatment, and we're seeing that those improvements accelerate pretty quickly.
In the upper quartile, we have over a 90% reduction with 42% of patients seizure-free for at least six months, and we're up to nearly a quarter of patients who are seizure-free for at least one year. For the lower quartile, I am so optimistic about the opportunity to drive this further given our continuing clinical experience and the value of the AI. But getting to this flexible placement and benefit everywhere, the dark blue is the improvement in all patients, and you'll see in the filled-in blue, the patients with actually the open blue, patients with mesial temporal epilepsy, the filled-in blue, neocortical patients, and stippled patients with onsets in both areas. And so it is clear that the RNS System works for anyone with focal epilepsy. So let's contrast this, and I'm going to talk about Medtronic's registry for epilepsy, and that's using DBS.
These are not comparator, head-to-head comparator studies, but it is a similar patient population, prospective, and they also had retrospective data. All of our data was prospective, but the differences, and this is peer-reviewed literature for the MORE registry, are pretty striking that the median % reduction from year one to year three is not what we see with the RNS System, and this represents current contemporary medical practice. These publications from the MORE registry are very recent. They're from 2024. Did I skip ahead? No, there we go. What can I say in summary about this? We continue to see that efficacy improves over time, and we have a home run in a substantial number of patients. 90% reductions within six months are possible and seizure freedom at three years. Why are things better? You have a sense of all that we have learned.
We have learned what should be detected. When we started out, we waited until patients were well into their seizure in order to intervene, and now we recognize that it doesn't work best if you use it as a seizure stopper. It works better as a neuromodulation device. We understand better where to put the leads and where to provide stimulation, and we know where to start with the stimulation programming from the very beginning, and we know how to fine-tune based on what the patient comes in to tell us. When we say, "How are you doing?" they describe that. So we use that. That's really important, but we also have the biomarkers, and we understand how to move forward with treating according to biomarkers. Now, I told you we haven't forgotten about the 15% of children who have focal epilepsy.
And we, some years ago, started th RESPONSE Study that really shows our commitment to advancing our indication for use to a pediatric population. FDA required a prospective open-label trial, and frankly, in contrast to any study NeuroPace has ever done, we've had challenges in enrollment. And this is very typical for any company that's trying to do a pediatric study for a therapy that is widely available to adults and for which physicians are providing it according to the practice of medicine, not everywhere, but some. So in the meantime, people are publishing all these publications about patients treated, children treated with epilepsy. And we became aware of a database, Pediatric Epilepsy Research Consortium, a 10-year-old database or 10-year-long database of children treated according to the practice of medicine across 27 pediatric epilepsy centers.
So when we put that together and we add to it support from advocacy groups, in fact, a demand that children have official access to this, we were able to partner with NESTcc, the National Evaluation System for health Technology that is funded by FDA in order to promote the use of real-world data to expand indication for use, particularly to underserved population. And we are a perfect project for NESTcc. And as part of this, we are able to collaborate with FDA with several meetings per month, sometimes with over 25 individuals from FDA in attendance to discuss how we can use all of these data sources to support an expansion for indication for use. We are well along in this process. NESTcc uses a group of epidemiologists.
There's a set process by which this is done, which includes looking at the data sets and understanding whether they contain data points that FDA would be able to accept and then design a fit-for-purpose study to access that data set, including a feasibility and a concept protocol, which is where we are right now. We are getting very close to the official NEST mark of approval, and what that means is FDA has agreed in principle that the data is available and that they will accept that as sufficient. We then execute on that. We implement, we do data analysis along with NEST and provide that data to FDA. It has moved along since the data already exists. We believe that this will go very quickly and that we will be able to submit in the second half of this year.
Now, there's some interest in NAUTILIST, idiopathic generalized epilepsy. I mentioned this epilepsy before and told you that this happens in otherwise neurologically normal individuals. Medication is the only option. There is not an option for surgery. The risk for sudden unexpected death in epilepsy is particularly high in this population because these patients have generalized tonic-clonic seizures, which are the seizures that put people at particular risk. We, based on the literature, elected to provide treatment in the centromedian nucleus of the thalamus. We are moving ahead with an indication for use as an adjunctive therapy. FDA allowed us to enroll 12 years of age and up. This is a randomized controlled trial, prospective, multi-center sham stimulation controlled. And Mark Richardson alluded to how rapid the enrollment was. 100 participants enrolled quickly, 87 implanted across 23 centers. The image to your right is pretty profound.
We didn't know whether we could detect seizures in the thalamus. And so what you see right there is the RNS System detecting a seizure in the thalamus, really establishing the feasibility of this approach. Here is the timeline. By providing the publications and experience to FDA, we were granted a breakthrough designation. Many, if not all of you understand that this really speeds up the regulatory path. Our first enrollment was in 2022. Our last enrollment was within 18 months. That's a year faster than I had predicted, really speaking to the interest and need. The final patient will complete the one-year endpoint, which is when we look at our primary effectiveness and safety.
I would love to be able to provide those results in March of 2025, but there's a whole lot of work that happens after the last visit to the time that we are able to submit to FDA. I've listed some of that work. I will just, in the interest of time, tell you it is a great deal. However, we're optimistic that we will be ready to submit to FDA in the second half of 2025, about the time that we submit for pediatrics. We will continue to follow these patients to two years, but the primary endpoints are at that one-year time. We mentioned Lennox-Gastaut. This is the worst type of epilepsy. It happens in childhood, and parents are faced with a child who develops progressive encephalopathy, really deteriorates in front of their eyes, and has multiple daily seizures.
We were able to receive a very sizable grant from National Institutes of Health to fund this entire study. It is being conducted at six clinical sites and, in addition, at a site that is in charge of the neuroimaging part of this study. It is also this study is funding biomarker investigation. All 20 patients are implanted. Safety is very acceptable at this point. This was a five-year grant. The grant completes 2026, and at the end of that, we will have a very good idea about safety, which is the primary endpoint in this feasibility study. We'll have information about efficacy. We'll understand about biomarkers that we can use to drive treatment in this patient population, and we will know ahead of time about where to put the leads using computational methods that we are developing as part of this effort.
So I want to tell you about all the data that we have. We have over 6,500 patients that are being treated, and it's like an oil well. Every day, there is more data. This slide is out of date every time I show it. There's over 19 million stored samples of intracranial EEG. And what we have, we have samples when the patient is at baseline, meaning that when we look at it, we don't see anything happening, samples when there are epileptiform discharges, and we have samples when patients have electrographic seizures. I will tell you that what we discovered is each one of these types of intracranial EEGs are telling us something. And the algorithms we have developed are giving us biomarkers that will correlate with patient-reported seizures. And those biomarkers are what are going to drive our innovation in the application of the existing technology.
I'll just show you one example. Treat to clinical seizures, let's treat to biomarkers. So what we started out is using machine learning tools to look at. In the top panel, you'll see year one to year seven in individuals with mesial temporal lobe epilepsy. The x-axis is the frequency of epileptiform spikes. The y-axis is a frequency spectrum or the frequency domain. In the top row, you see patients who did really well. In the bottom row is patients who didn't do well. They stayed at 25% reduction. What we see is that everybody started out the same way in frequency and epileptiform spikes, but as patients got better and better and better, they had fewer epileptiform spikes. This tells us that we should treat epileptiform spikes.
We then developed deep learning capabilities, and now we're using vision transformers very recently to understand whether we could go deep and identify those patients who are going to be responders and those who are having a less robust response. It was very clear that going down 50 levels in this analysis, which was using a convolutional neural network, that it very clearly distinguished that there are differences in these patients. This is another window to the brain. Can we use this data in general? It has some value. It has value to us because of all of the information it provides us about patients. These are my two patients. Where you see the little round M, that's the patient swiping a magnet to tell me that she has had a seizure. The green bars are when I actually see seizures.
So you see that when this patient tells me she's having a seizure, she is, but she's not able to tell me about all of them. But it's a pretty good indicator. The patient on the bottom is a patient that has been so hard to treat because she has no idea when she's having seizures. And I've been treating her for 20 years blindly. With the RNS System, I now know when she has seizures that she doesn't know about. And in January 2022, I found better detection stimulation programming. And now when she has them, it's when she decides she's not going to take her medication. So what is our data telling us? Well, we look at different types of data. And in the upper left panel, you'll see one Intracranial EEG that's showing that the patient is having prolonged epileptiform activity.
It's not a seizure, but the device is detecting a number of abnormalities. To the right, you see that there is an electrographic seizure. There's a buildup in these discharges. The patient doesn't know that this is a seizure, and then the bottom panel is an electrographic seizure that the patient has symptoms with, so the patient knows. We have looked at all of these types of recordings and have found that those long episodes, upper left panel, are strongly correlated with the patient's clinical experience, and this is an example then how this can be applied to drugs, so to the left, you'll see the spike rate. That's the middle line for a patient who's being treated with the RNS System and was started with a new anti-seizure medication.
We could tell, as is demonstrated in the graph to your right, within one to two weeks whether that drug was going to work or not. That relieves patients of the burden of drug trials that could go on for six to 12 months. It also was very interesting to a company called Rapport. This is a Janssen spinoff. It's a company with a very interesting compound that binds specifically to proteins on the AMPA receptor that are very prevalent in the mesial temporal lobe. We'll guess who has a lot of patients with mesial temporal lobe epilepsy. Now, this company wanted to perform the very first study in people to see if there is target engagement.
Based on that experience to make the decision about whether to invest in clinical development, they came to us and said, "We understand from your publications that long episodes are biomarkers for patient clinical response. And we want your long episodes to be the endpoint of our study." This is a whole new paradigm in very early drug development. So that's very exciting. And we've had a few people ask about potential future applications. And I'm going to ask Samir Sheth, who is an extremely forward-thinking, thoughtful, brilliant neurosurgeon at Baylor, to talk about how he views these opportunities. Samir, there you are. I see you. And I'm going to advance your slides.
Okay. Good morning. Can you hear me okay? We can. Okay. Wonderful. Well, thanks a lot, Marty. And thanks to everybody. Sorry, I couldn't join you all there, but very exciting.
I'm really happy to be here. Next slide, please. So maybe we'll skip to the next one. This is kind of an overview of everything, which I want to come back to. Can we go to the next?
Oh, I'm sorry. This is my turn. Yeah. Do you want to talk about this part, Marty? I jumped ahead, but that's okay. You stay on for this.
As we do, Marty, and Dr. Sheth, thank you. It's just for everybody's benefit here. Oh, actually, Joel didn't talk. No, just real quick. So I spent a little bit of time earlier talking about focus and execution. And everything that we're talking about today has to do with epilepsy, except this piece here. And I don't want anybody to be confused as to whether we're confused as to how important that execution is to us.
At the same time, and the question came earlier today, Rohan, around applications of the RNS platform. And so we wanted to take a minute to give you some perspective for not even so much things that we're initiating, but supporting. But what are some of the external vectors that the field is seeing and is pursuing by way of an understanding of the unique nature of the differentiated technology with RNS and how they see it being applied? So the reason we have this section here today is to give you a little bit of a window for the things that we see when we talk with people in the community and the work that's going on, not distracting us from making sure that we're doing everything we can with epilepsy, but speaking to the potential and the value of the RNS system.
So just that little blinking value-added caveat there. And Marty, please get us started. T
Thank you for protecting us here. Yeah. Let me assure you, we have a relentless and consuming focus, all-consuming focus on epilepsy. But it's impossible to ignore the potential of the platform technology. And if we think about what the RNS System brings, well, first of all, there's a huge safety data set and evidence, demonstration of really impressive efficacy. We are able, as we've shown you, to monitor brain electrical activity. And we can do so anywhere in the brain and therefore, theoretically, for any type of brain disorder. The device lends itself to targeted and personalized therapy. And of course, there is this really exciting and very fast discovery in the area of biomarkers. And then, of course, you don't want to wait until a patient is sick to treat them.
We want to treat them before they have the symptoms of their disease. I mentioned that it's impossible to ignore, and it's impossible to ignore because we are participating in collaboration, sometimes as co-investigators, sometimes in a more distant role for externally funded studies in disorders other than epilepsy, and they fall within two main categories, so I'm going to group them into neuropsychiatric disorders, and you'll see the list of, again, this is externally funded research, mostly NIH in depression, impulse control, loss of control, eating, and PTSD, and the other category falls in the area of memory and cognition, so restoring function. These are also NIH-funded. Many of them have to do with memory in the domains of navigational memory, also cognitive assessment, understanding brain plasticity, understanding memory, and also, there is a trial in autism disorder.
So all of this is happening, bringing people, bringing things to us and us able to support them with our enthusiasm, often our data science and our device. But there are opportunities for providing responsive stimulation. And now I'm going to introduce Dr. Sheth again to share how he views this.
Yeah. Great. Thanks very much. So just in the way of framing, I'm going to talk about sort of two framework-level organizations in terms of the types of disorders outside of epilepsy that we might think of as being opportunities to use RNS. I'll spend a little bit of time on this first one and a bit more time on the second one. So the first one really are those with actually, are these out of order? Can I just check and see what the rest of the next one is? Yeah. Yeah.
I think this is the one you want to talk first now. It's the paroxysmal one first. Yeah. Keep going. Is there another one after? No. No. Okay. Fine, well, then I'll say the first one briefly and then kind of get to, yeah, let's go back one more. Yeah, one more. There we go. Okay. All right. So what isn't on the slide there, but just, there's really nothing to show graphically. But again, in terms of the framework, I mean, there are disorders, let's go forward one, that are paroxysmal, and what I mean by that is that they have specific episodes of when the symptoms manifest, and then the rest of the time, the patient or the person is fairly normal. I mean, think epilepsy. They're not perfectly normal when they're not having a seizure, of course.
But the whole philosophy of RNS, which has really borne out quite well, is that when the brain is not having a problem, when the brain is in a more or less normal, non-pathological state, leave it alone. So that's worked very well for epilepsy, as you've seen. Other disorders outside of epilepsy that have that same paroxysmal nature are, for example, Tourette syndrome or PTSD or addictions. Stay back. I'll ask for the advance. Sorry. So those are paroxysmal disorders that might have a similar kind of biomarker or neural biomarker like we see in epilepsy that we could use to trigger stimulation. I'll spend the rest of the time talking about what's on this slide and going forward, which is a different type of disorder, which is not paroxysmal, but rather it's slowly evolving.
These are disorders that have pathology that doesn't just occur in fits and starts, but rather is there most of the time and maybe evolves in terms of severity. In terms of the de-risking of using a responsive neurostimulation type of system for those, that process has begun, but there's a lot of work to be done still. The philosophy here is different. It's not leave the brain alone and stimulate only when you're going to have a symptom outburst, but rather the symptoms are always present or often present and just slowly varying. And so here, the philosophy is more that the stimulation output adjusts to neural biomarkers that we can measure. And as shown here, just in the parentheses, a lot of times, these biomarkers can evolve slowly. So over days, weeks, months, not moments.
But like what we're used to in epilepsy, the idea that we need to individualize our identification of those neural biomarkers to the individual patient is still very applicable. Examples here are depression, memory disorders, and other psychiatric disorders like OCD and anxiety disorders. These, I think, goes without saying. These are not ones that come and go. It's not like with depression, sometimes you're acutely depressed, and other times you're totally fine, but rather that they vary slowly over time, and hence kind of the motivation for this. Next slide, please. So I'll spend just five minutes focusing on this idea of depression. We've done a lot of work in this field. And again, the idea is to measure to see if we can even identify neural biomarkers of mood states and then use them for adjustment of stimulation.
This is work that's funded on this NIH grant, also from the Brain Initiative, like the one that Marty showed you a minute ago from about LGS. This is one that we got a few years ago for depression. Next, please. So the idea is this. We're very used to doing intracranial monitoring in patients with epilepsy. We've been doing this for decades. It used to be with grids put on the surface of the brain. Now, for 10, 15 years, it's really mostly with stereo EEG, meaning thin, small electrodes that are placed into the individual. And this is done at the stage when we have some idea of where seizures are coming from, but not well enough to come up with a definitive plan. So these electrodes are placed. The patient is kept in the hospital for a week or two, for example.
We allow them to have seizures. And then once they do, now we can tell exactly where they came from. And then we may see, are they a candidate for a resection or neuromodulation like RNS, etc.? So this is a tried-and-true paradigm in epilepsy, but used very seldom outside of epilepsy. So this study is one that we're using. We're borrowing this approach from epilepsy of temporary intracranial neural recording and using it to understand the varying mood and affect and cognition in depression. So we're taking here 10 patients with severe treatment-resistant depression. So these are not patients with epilepsy who might be depressed. Rather, these are people who have typically never had a seizure but have severe depression. And we're implanting them with these temporary monitoring electrodes, spanning these regions that are shown here, kind of across frontal and temporal areas that are relevant for mood management.
We keep them in the hospital in the Epilepsy Monitoring Unit for 10 days. We study their individual neural signatures of mood, affect, cognition, etc. Then after they leave the EMU, we adjust neuromodulation based on that kind of bespoke, tailored neural signature pattern that we have identified. Next, please. On the right here are just three papers that we've published just in the last couple of years. I'm not going to bore you with all the details, but I want to highlight a few salient points. The first one is on that top figure on the top right. The point of this was that we really do have a pretty good developing understanding of what neural signatures and where they come from indicate or predict different mood states. What we did in this study was we took, in this case, three of these patients.
We asked them periodically, five or six or seven times throughout the day, "How are you feeling?" A quick mood questionnaire. It took a couple of minutes to administer. We kind of have, over the course of 10 days, maybe 40 or 50 snippets of time where we have an estimate of their mood severity, their depression severity. At each of those times, we also have the neural recordings in all those brain regions shown on the left. Then what we could do, we could say, "When someone is feeling a little less depressed," they're not, of course, normal or cured, "but when their mood is in a little bit better state, what does the neural activity look like?" In contrast, when they're in a worse state, more depression, what does the neural signature look like at that point?
And then we can do the analyses and say, "What particular neural features in what brain regions predict better versus worse mood state?" And that's what's shown here with the red and the blue. And so the kind of summary from this study is that you can see if you just kind of zoom out or kind of cross your eyes that across these three patients, there's definitely some consistencies. You've got kind of the red on the top in those neural features and the blue on the bottom. So there's some generalization that all three of these patients, and we've done this for now a total of six, do show some consistencies in pattern. But then you can tell if you zoom in, of course, that there are some individual differences.
So now, if we just direct our attention to the bottom right, this is a very recent study that just came out a few months ago, where instead of so we want to do the same thing. What are the neural signatures associated with more versus less depression? But here, instead of labeling them with a survey, "How are you feeling now? How are you feeling now? How are you feeling now?" doing it that way is a bit fatiguing. It's a bit subjective. And so instead, what we did, if you kind of see in the green box, that conversation box, you can see a snippet of what this person said. We basically just use conversational data. So this individual was very talkative, and so we would talk to them about their history of depression and just history of life.
We would use AI to really extract the emotional content from the conversation. So it's kind of a much more objective way of assessing how someone is feeling just based on what they're saying. And by doing that, bottom line, we got a very similar sort of pattern of activity indicating that we can really do a lot of this assessment of mood state, not with surveys that are cumbersome and slow and subjective, but through just the objective extraction of emotion from what someone is saying. Okay. So now, just to kind of summarize the points here, if you could go to the next animation, please. Yeah. So these three bullet points are kind of what I'd like you to take home from this discussion. One is important, just in terms of practicality.
The information that we need to decode, in this case, depression severity or mood, is available using clinical macroelectrodes, meaning the electrodes that we see on typical neuromodulation devices like the RNS device. So that's good to know that we can decode that information from clinical devices. The second one is that these biomarkers do have some general features. It's not that each person is completely different than every other one. That would be pretty hard to imagine a feasible study. So the generalization, again, from the top right, the red on the top, blue on the bottom, is there for everybody that we've studied so far. That speaks to feasibility. But there is some need for individualization because in detail, there are differences. And we're used to this with the RNS device of individualizing that detection ability.
Finally, that bottom bullet point is that neural biomarkers are certainly very interesting. Alongside that, we should, I think, really study these behavioral biomarkers as well. They are an exciting and rich source of information. They provide objective evaluation of how someone is doing, especially with mood disorders, let's say, in a way that clinical measures can't. In terms of temporal frequency, we can't assess someone's depression many times throughout the day. There's just not validated scales for doing that, nor is it practical. Through behavioral markers, for example, what someone is saying, or from smart devices that people are very happy to wear. In fact, they pay for these smart rings and watches and phones and whatnot.
The amount of behavioral information we can get from those is staggering in terms of the size of someone's social network, how mobile they are, how much they interact with others versus they're stuck at home because they're sick. So that behavioral data is very rich. And I think coupled with the neural data will be very exciting. Next, please. So I think really the next step in terms of what we would love to do following this study, which we've done in the next year or two, would be to take this information and employ it on a larger scale with a device like the RNS device that can actually record neural data. That's a big limitation we have in our current trial is that the system we're using doesn't record any neural data.
So if we could actually record data outside of the hospital setting when people are just out living their lives, again, pair that with behavior in terms of sociability, mobility, etc., the hypothesis would be that if we adjust the stimulation output of that device based on that neural and behavioral input and feedback, that we'd be able to, in the case of depression, for example, keep people in a less depressed state more often for longer with less chance of relapse. And then one more click, please. And so really, the additional adjunct that would come out of this is the ability to remotely monitor. This is something that we're really lacking just in the field of neurological and psychiatric care. What the RNS System is allowing us to do for epilepsy is kind of leading the edge with that.
But imagine now applying that to many other disorders that we just don't have the ability to routinely monitor how someone's doing. Okay. Next slide, please. This is my last slide. It's going to be kind of a summary and a little bit of expansion. Just four thoughts. Next, please. The first is that, again, to recapitulate, the ability to continuously monitor neural biomarkers of the target state, I think, is going to change how we address many disorders beyond epilepsy, neurologic, and especially psychiatric. And those include disorders of mood or memory. Next. And then, of course, as we're measuring these biomarkers, we adjust our output based on those in order to match the supply of stimulation to the demand, meaning what the patient is feeling.
And these, again, could be in a paroxysmal way for disorders that have fits and starts of symptom severity or slowly evolving ones like I talked about at some length in terms of depression. Next. This point I mentioned on the previous slide, remote monitoring, I just want to highlight it for a second because this is something that we're really missing. Again, as I said there, with many neuropsychiatric disorders, we just don't have objective biomarkers. All we can do is ask someone, "How do you feel?" And there's standard scales that are good for communicating that information, but they're not good for measuring it in a dense way or in a feasibly deployable way to people just out living their lives. And so the idea in that bottom one is to move beyond symptom-based therapy, right?
So not just kind of symptoms that require cumbersome assessments or clumsy ones, but rather to objective and biomarker-based therapies, again, neural or behavioral. And then final thought is engagement of patients. Again, people are very happy and will themselves pay for the ability to engage with their own health, right? My step count, my readiness potential for my smart ring, whatever that means, right? So we're very obsessed as a culture about knowing how we are doing and being told about that. Now, imagine being able to do that in a clinically meaningful way in terms of you're more likely to have a seizure in the next couple of days, so perhaps be aware of that. Or you're perhaps entering a period of time where you're more susceptible to having a relapse of depression.
So perhaps the stimulation is going to adjust, but also, wouldn't it be good for you to know that to be able to anticipate what's coming? So anyway, those are just my thoughts about kind of responsive neurostimulation beyond epilepsy. I'm happy to discuss if there's time now or later.
Thank you, Samir. I think we have a Q&A at the end of the day. But what we want to talk to you now is about how we're going to take all of this stuff that we've been talking about, the AI that is capable of moving patients to better and better responses in epilepsy and in the future, perhaps engage in some of the activities that Samir spoke about. So I'm going to begin by introducing Dylan St. John and Brett Wingeier.
Good morning. I'm Dylan St. John, Chief of Commercial and Development.
I’m excited to be talking with you today about how we think about product development at NeuroPace. I'm going to start by highlighting and emphasizing some of the actions NeuroPace has taken to reach more patients and more physicians. As you've heard, through our clinical studies, we're working to expand our indication into generalized and pediatric. We've also talked about expanding our site of service through the Project CARE program, which Katie will talk about a little bit more in the next segment. As we reach our customers, we want physicians to be able to enthusiastically recommend the RNS System. We want patients to be able to easily say yes to that recommendation.
And that's why, as an R&D organization, we're focused on the continuous improvement of our product in order to remove potential barriers to adoption so that it is obviously and easily the most best solution available from all perspectives, the most effective, the most efficient, and the most sophisticated in its simplicity. We're going to accomplish these goals by focusing on three key opportunities. First, as always at NeuroPace, I'm going to start with our patients. We're going to leverage the unique and massive data set you've heard about to help improve a therapy effectiveness. And you might be asking yourself, "Why does the NeuroPace team think they can do that?" And the short answer is we already have. You heard that patients got better faster during our post-approval study with a median seizure reduction of 62% at six months and 82% at three years.
The top quartile, better than 90% at six months and 100% total seizure freedom at three years. Those outcomes were possible exactly because we learned from our data. As Dr. Rao mentioned, we're not yet satisfied. We know that there's more improvements that can be gained, and we're going to continue to focus on this into the future. Second, we're going to improve clinic efficiency. We're going to reduce the overall effort it takes to manage a patient's treatment with the RNS System. In a few minutes, Brett's going to walk us through our product pipeline, and he'll highlight some of the improvements we're making that will help drive those efficiencies. Third, we'll improve ease of use.
We want the RNS System to be the easiest to use for patients and physicians, and we know we can make that better by adding additional capabilities: the ability to program the RNS System remotely during a telehealth visit, or the ability to transfer data automatically while a patient sleeps. In order to make the most of these opportunities, we have a few focus areas within product development. First, our data. We have this unique and massive data set, something that Joel has the best description of as an asset to humanity: more than 6,500 implanted patients, more than 26,000 patient implant years, and more than 19 million intracranial EEGs.
We continue to expand our capabilities around machine learning, data learning, and AI in order to take our internal software tools and turn them into customer-facing NeuroPace AI-powered products that will simplify resource-intensive tasks like the data review and therapy and detection configuration that you've heard about. These improvements will enable the physician to be able to scale their practice to treat more patients and enable NeuroPace to more rapidly and more efficiently leverage our resources to scale the business. We're also building our next-generation platforms. The patient data management system, what we call PDMS, is being re-architected from the ground up in order to reclassify portions of the system with FDA. As we generate new learnings and insights from our data set, because we've reclassified portions of the system with FDA, we'll be able to more rapidly iterate and improve NeuroPace AI-powered products.
Our next-generation RNS platform, in addition to the capabilities I mentioned a moment ago, will enable a regular cadence of software and firmware updates, even for those patients already implanted with a neurostimulator, and will deliver enhanced detection and treatment options anchored in our data set. Everything I've discussed, everything Brett's about to discuss, will lead us to the outcomes we care most about: improving therapy effectiveness for our patients, improving clinic efficiency both for the physicians and our field team, and improving ease of use for all those involved. Now, I'm going to hand it over to Brett so he can talk about the details of our product pipeline. Brett.
T hanks, Dylan. I'm Brett Wingeier, Vice President of Research and Development.
So before we dive into specific products, I'd love to just share the cycle that translates all of this into products and better results for the business and for our patients. So we have unique assets like the RNS System and this really world-leading data set. And we have core institutional knowledge at NeuroPace that includes neuroscience, data science, product development. And that lets us leverage these assets into a flow of data-driven products that improve outcomes for our physicians and patients. So that means effectiveness. It means efficiency. It means ease of use. And particularly, let me point out, efficiency also means efficiency for our field team. And that means that we can grow the market, scale the customer base without having to directly scale the organization. And really importantly, the more we grow the market, the more data we collect. So we have a continuously compounding competitive advantage.
Now, ever since we launched the RNS System, we've always had a patient data management system, PDMS. PDMS is where you review settings. You review the data. It's what stores this amazing data set, 19 million intracranial EEG recordings, more than 19 billion rows of event records in the database. It's foundational for pretty much everything we do. Now, systems and technologies evolve, especially stuff on the cloud. So we've been working on a next-generation scalable PDMS that'll support growth into the future. Now, a lot of that work is under the hood, but importantly, for our users and our patients, this is also an architecture that'll support this cadence of data-driven products and usability improvements. So the first such product is this AI-powered seizure classifier that's built right into the front page of PDMS.
Now, let's do a deeper dive on that, which we'll be targeting for launch later this year. Now, every brain is different. That's the power of the RNS System. It detects and treats individual epileptiform activity. That also means for a new patient, somebody has to sit and review a lot of intracranial EEG records and identify what do the seizures look like, where are they coming from. Now, when we turn on the AI-powered seizure classifier, it automatically identifies electrographic seizures, presents them to the user. That makes life easier for physicians and helps our field teams scale to serve more sites and more patients. Now, as soon as you can highlight electrographic seizures in an automated way, you can get a ton more value from those data.
When you can identify and count electrographic seizures, you can show what lead are they on, how many seizures are there. You can present histograms that show the progress of treatment, patients' individual circadian rhythms, multi-day rhythms, and these are insights that translate directly to patient care. Things build still further from there. Automated seizure identification is foundational to a series of tools that all relate to effectiveness and efficiency and usability. The sequence builds eventually toward a system that automatically proposes optimal settings for detection and treatment in any individual patient. That eliminates still more complexity and helps us drive toward greater adoption, grow the market, and better outcomes. Let's go still deeper on how this fits together. Every seizure has an onset, right? The point where the brain, that's the point where the brain transitions into seizure.
Users of the RNS System program detection settings to catch this abnormal activity at the earliest signs of a seizure. Now, to do that, you need to identify the onset. And that's the next step beyond just highlighting seizure records. So the next generation of our data tools will report specifically on when seizures are starting, where exactly are they starting. That makes data review even easier, and it gives more important clinical insight and helps guide where to treat for maximum effectiveness. So the model that we have under development identifies seizure onset time and location with less than one second error compared to markings from trained clinicians. Now, where all this comes together is in a third AI tool that builds on the others.
So remember, we're trying to catch abnormal activity at the earliest signs of a seizure and treat it with the best possible settings for that individual patient. Now, the way that that happens is with years' worth of data and experience. So now's the right time to start building that into a product that proposes optimal settings and makes everything a lot easier to get started. What you see here is one example of how our tool in development identified the very earliest signs of an event. Now, one point of clarification. Remember, all of this modeling and inference and data science happens in the patient data management system on the cloud. The data come from the neurostimulator. They get uploaded to PDMS. The analysis happens there, and the system proposes settings that can get sent down to the neurostimulator.
That drives the very powerful, very flexible detection and treatment tools that are embedded in the neurostimulator. What that means is agility. It means as we continue to learn more, we can develop on that without having to change the underlying hardware. Now, why is all that so important for effectiveness and efficiency? Again, remember, every brain is different. So that's why responsive neurostimulation makes so much sense. It detects and treats individual activity in an individual way, but you can't get there by adding complexity for the user. So automated proposal of settings will make this simple. At this point in the roadmap, the product will propose personalized settings based on what a patient's individual signals look like, their treatment history, everything we know about how to treat epilepsy with responsive stimulation. The physician's still in the loop. The physician can review, modify, and approve the settings.
and then we'll have the opportunity to update parameters with remote programming. Those are two important words. Let's talk about remote programming. Even once we've simplified setup even more, like any active implantable device, the RNS System has settings that sometimes need to be adjusted. Right now, that has to be done in the physician's office. If a patient lives three hours away or 12 hours away, they have to drive to the clinic. And a lot of these folks have to depend on public transport or get a ride from somebody else. Now, one of the strengths of the RNS System is that it comes with a remote monitor that already uploads data right from the patient's home.
Our next step is to allow remote programming so physicians can use that infrastructure to send new treatment settings via the remote monitor and have fully remote visits in conjunction with their own telehealth systems, and that takes away a really key pain point for the physician and for the patient and lets us keep growing the business to more patients and more centers. Now, another way we can make life easier for patients is with automatic data transfer, so right now, in the patient's home, data are uploaded with an inductive wand. That means the patient has some work to do. They have to hold the wand. They have to find time for the upload. Now, we can solve that with longer-range telemetry, which is going to be part of our next-generation hardware platform.
A better patient experience here means we've taken away another barrier to adoption, and we've kept assuring an ever-growing flow of data, and that brings us back to the neurostimulator, which is both the source of the data, and let me remind you that the fact that we have this neurostimulation treatment platform is the reason why all of our data advantages translate directly to outcomes and better treatment for patients. Longer-range telemetry enables automated data upload. That'll be a key feature of our next-generation platform, and we're working on this actively in parallel with our software and data efforts. The new platform will be fully compatible with the RNS System. It'll use the same mechanical envelope. It also updates the microprocessor and the custom IC, and that's exciting because it gives us a platform that we can keep developing on into the future, so let's just summarize.
Here's a wrap-up of our pipeline, our roadmap across two themes: data and the platforms that deliver treatment and help us keep expanding those data. So sitting here in 2025, we're planning to launch our second-generation tablet remote monitor and this update to our patient data management system for scalability, which supports our first AI-driven tool that makes it a lot easier to use the RNS System. Now, in our long-range plan, we have remote programming. We have our next-generation platform project with BLE telemetry, and then a cadence of data products that progressively eliminate still more complexity, make the system easier to use, and increase efficiency in the clinic. And let me close on that. The North Star for our product strategy is we develop things that not just advance the business, but serve the needs of people living with epilepsy.
This is the roadmap that gets us there. Thank you. Let me introduce Katie Keller, who's our Vice President of Marketing.
Thanks, Brett. Good morning, everybody. My name is Katie. I'm our Vice President of Marketing. Pleased to be here with you all today. I have the pleasure of talking about market development, which I think is just so fun. What we're going to be looking at today, you've already been hearing about. We have an under-penetrated market and a market that's growing. We've talked already today about the more than a million patients who have epilepsy that's not responding to drugs. Yet only not even 7,000 of those patients each year make it to a surgical consultation and receive a therapy other than the drugs that have already shown they're not working. A real need for devices like the RNS therapy.
Today, our target addressable market is in adult focal epilepsy. This is a significant opportunity, and we're going to talk a lot today about how we expand adoption in our core market, but we have the opportunity to move into generalized epilepsy as well as addressing the pediatric population. This is going to essentially double our addressable market, really an exciting place to be, and we're on a journey through these market development efforts. This is not something we're starting today. For the last few years, we've been focused on the CEC. You've heard about that today. Those are the comprehensive epilepsy centers where most patients who end up with an RNS device are treated. We have some really great, strong clinical data that differentiates us from our competitors. You saw the post-approval study data earlier today. This is unique and unmatched in the space.
We're also expanding in the community. You guys heard about the care program? We're going to talk more about that today, too. More than 1,800 epileptologists are not in CECs. They're out in the communities, and they follow quite a few patients. So as we expand to build more implanting RNS centers in the community through our Care Center, it really opens the access to more patients. And then, of course, expanding indications continues to be a key strategic initiative for us: the Nautilus trial. We're really excited about the progress there, and that will open the door to generalized epilepsy and also our efforts in the pediatric space make this a really fantastic opportunity for developing the market. Let's talk about each of those in a little bit more detail.
First, our core, those patients who are at CECs, those physicians who are assessing patients at CECs and then implanting them. We will continue driving preference for RNS, driving awareness and adoption for the procedure. We're the only therapy that provides 82% seizure reduction at three years. We're the only therapy that improves cognitive function. Patients can remember better. They can learn better. They can speak better. And that results in an improved quality of life at just a year. Now, it's easy for me to put a bullet point up here on this slide about improvement in quality of life. But how about Mike's story earlier today where we got to see how his life changed really significantly because of this therapy? He was able to get a job, pay his taxes, and now works educating patients about the benefits of RNS. It's super exciting.
I wanted to talk a little bit about our direct-to-consumer strategy. I'm not sure how familiar you are with our pipeline, but we've built a really solid infrastructure for reaching patients directly. This starts with digital advertising. We use platforms like Meta to get patients' attention. They come to our website and take a quiz. And based on how they respond to that quiz, we understand, first of all, if they're a target customer for an RNS device, if they're drug-resistant, over 18. And then we can triage kind of where they are in their journey. We pair them with a nurse navigator, and our nurse navigators do one-on-one education to teach the patients and their family about the device. But then they also help the patient maneuver through their medical community so that they can reach the physicians who can prescribe RNS.
For those patients who are the right fit for our technology, through coordination with our sales team, we help them interact with physicians, and then ultimately, they're implanted with an RNS device. An appreciable portion of our revenue in 2024 is tied to our DTC efforts, and we will continue to expand these. We're expanding our direct patient advertising. We're adding more nurse navigators. We're creating new educational content for this group. We will continue introducing new programs. For example, we're doing new local programs where we partner with physicians so that they can educate their community of epilepsy patients about RNS. It's really exciting to see the way we're growing here. Now, to date, we have really focused these DTC efforts on patients that are in a CEC already. They're the ones that will move fastest through this funnel.
However, as we move forward, we're going to begin advertising to patients who are in the community around our new care centers so that we are really reaching more patients that we can funnel in through the care center. And I'll talk a little bit more about that in a moment. But first, I think it's important to recognize what we're doing with patient education. We're adding nurse navigators, but our three existing nurse navigators spoke with over 1,600 patients in the last year, providing that one-on-one education. And Mike often gets asked by physicians to do patient education. He alone spoke with over 500 patients this year, and more than 90% of them moved on in the process to get an RNS device. That's a very impactful part of our program.
We do monthly virtual chats in the community so patients can interact with physicians and nurse navigators and Mike, ask questions, learn about the device. We also have a host of patient ambassadors that can speak directly to potential patients and their families, as well as help us with outreach in the community. Finally, of course, we have a full sales team that does education with reps and their families, excuse me, patients and their families as they're considering the device and then after they're implanted. So it's a key part of our strategy. Speaking of strategy, let's talk now a little bit more about our outreach in the community or the care program. And this, again, has been a significant program for us. Meaningful revenue in 2024 was generated from the care program.
Now, it's important to realize that it helps us today to expand into the community, but this is part of a broader strategy. As we move to getting the indication for generalized epilepsy, those patients have an easier, simpler workup, and so by activating more centers that can implant RNS today, we're really setting ourselves up to capitalize on that future indication, so within Care, we're starting new centers, places that have a neurosurgeon that isn't affiliated with the CEC, that either have the equipment or are interested in acquiring the equipment so that they can perform implants in their center. This has been very effective. Another part of the Care program is really establishing and confirming the referral pathways from the community into the Care center. That means that we're spending time with neurologists.
We're spending time with epileptologists so that we can educate patients and flow them into the newly implanting care centers. Also, there are some times when the patients are more complex and they need to be worked up and treated at a CEC. So these referral pathways also work in that manner. So as we continue building on the success of the care program, what that looks like this year is we're increasing the number of targets that we have for those new implanting centers. We've created a raft of new marketing materials for our sales team to use with referral physicians in particular. And then we're adding digital marketing so that we can identify and educate those patients that are in the community to fill the funnel for those new implanting centers. And then finally, we're adding some new professional education offerings directly targeted at referring physicians.
Speaking of which, let's talk for just a moment about what we're doing with professional education. We've got a really strong program here already. Last year, we reached almost 600 individual HCPs through our large symposia and programs, our more small regional courses. And then we do a ton of one-on-one, peer-to-peer conversations, a great opportunity for newer physicians to get consultation on their patient selection and their techniques. Additionally, AES is a fantastic opportunity for us to interact and educate physicians. We reached over 550 additional physicians through our programs and symposia at AES, which will continue. As we move into this year, some areas of focus continue to be network stimulation. This is a hot area in the field of neuromodulation. A lot of buzz around this and RNS in particular at AES. We will continue to differentiate our clinical data because what RNS does is outstanding.
So this is a key part of how we educate. We have an intense focus on fellows because fellows, of course, will be the prescribers of the future, but they play an important role today. Are you familiar with the case conference? Oftentimes, before a patient is prescribed RNS or a surgical treatment, a case conference will be a discussion among the cross-functional team of physicians. Oftentimes, a fellow leads these. So there's a really nice opportunity to influence these physicians who are earlier in their career. And then that leads me to really the third pillar of our market development. So we'll continue to focus on the core. We're going to expand in the community. And then we're already preparing for the generalized epilepsy indication expansion. As I mentioned, this represents a significant expansion in the amount of patients who are eligible for RNS.
It's a shorter diagnostic process, which means that moving patients through that funnel is more efficient. And while we don't have the indication today, we're starting now. Right now, we're preparing the market. Based on the level of interest in this room already about the NAUTILIST trial results, I'd say we're doing a fairly decent job here. What we're doing is we're really focusing on education around trial design for NAUTILIST. We're talking about the mechanism of action of RNS. And I will say that the physician community is just as eager to see these results as those of us in this room. And we're really building our launch plan now. We've already started the process of our key messages, putting together our strategy around KOLs and podium platform. We're building the team so that we're ready with strong collateral for physicians, patients, as well as referring physicians.
We've got a plan for a comprehensive digital omnichannel launch. It's super exciting. And as we've been talking about, we have an under-penetrated market that's going to grow. And we need to be poised to take advantage of that. And so we are scaling our marketing and our sales teams. It's really exciting to add some folks to the team that can help with our referral efforts in particular with the ongoing Care initiative in the community. And it's not just a matter of adding more people to the team. We're really working on adding skills to the team so that as we're getting ready to launch, we have the digital marketing capabilities. We have the strong KOL engagement capabilities. And gosh, it's going to be fun. I'm so glad I'm here. I'm so glad you all are here.
And I can't wait to bring up Joel, who's going to share a little bit more about our vision and plans.
Thanks for getting Joel's going to bring Rebecca with him because we're going to talk about.
Oh, did I go the wrong w ay
No, you're good. Hold on. There we go. Vision plans and expectations here for the remainder of the LRP, as well as getting into 2025 in some more detail. Our plan to become a leader, and our plan was to fully develop the RNS platform. Hopefully, you've gotten a flavor for that today. We feel like we have an accelerating plan for the business. What we're doing today is delivering us what we've got today.
And we spent the last three hours you were so gracious with telling you about all the stuff we're going to be doing as well that is additive and cumulative to the strategic position we're in now. And I'm lampooned a lot in the office. I have to do it once. I literally can't wait. I'm pounding my hand on the table with enthusiasm for this and what that's going to mean to patients, what that's going to mean for clinicians, and what that's going to mean for all of us as we really take advantage of everything that we have in front of us. And this is what we plan to do. We're going to execute to this plan. And these are the objectives that are going to drive our behavior every day over the next three years.
And so with that, I want to take a little bit more here then and talk a little bit about the setup for 2024 and the 2025 guide. And I'll turn it to Rebecca to help us walk through that. And I'll stay right where I'm at to offer some commentary as well.
Great. So Joel shared a preview of our 2024 financial results a little bit earlier as he shared our pre-announced revenue grew over 20% in 2024. And if we consider that in 2023, we had revenue from our Nautilus trial. If we adjust for that, our revenue grew over 25%. So given the strength of our performance in 2024, as well as all the exciting initiatives that you've heard about today that are growth catalysts for us in the future, we're sharing our 2025 guidance with revenue of $92-$96 million.
Some of the key considerations underlying that guidance is that we assume revenue will continue to primarily be driven by sales of the RNS systems in our current customer base, our Level 4 centers, and also by increasing momentum that we will realize through our care program as more of those centers do implants and refer patients into our Level 4 centers. We have not included contribution from approved new indications. That's not expected in 2025. We do assume that Dixi will continue to contribute meaningfully to our revenue as it has in the past. And we assume that all of our revenue will be substantially in the United States. Our guidance for gross margin is 73%-75%, realizing improvements as we continue to scale. And our guidance for operating expenses is a total of $92-$95 million. And that includes approximately $11 million in non-cash stock-based compensation.
So with that, Joel, would you like to wrap us up?
Yeah, we'll go to Q&A here and being respectful of everybody's time. The last slide I've got here is a little bit of a personal editorial, but I've been around a while. I feel about right, and I hope it comes through. I feel about the brain today, right now, the way I felt about the left atrium and the left ventricle in 1995. This is the next frontier, and you can see it. And we're in such a good position. I would go back to those fundamentals that we opened with of a differentiated technology that bears up in the evidence in a group of patients that really need the therapy. And we're uniquely well-positioned here to provide those benefits. And the work that we're doing is in flight. You saw a lot of PowerPoint slides today.
It's not a bunch of PowerPoint slides. It's a bunch of stuff that's already underway that we're working on, and we look forward to executing on it and bringing it to you both within epilepsy as well as the value and potential of the platform and beyond. Thanks for everybody's time and attention today. I can fully understand if there are those of you who will need to step out, given that we're running a little bit behind schedule here, but we do have a Q&A session that we're going to stay here for for anybody who'd like to. I'd like to ask Rebecca to please stay and Dr. Morrell and Dr. Wingeier and Katie. Dylan, you're still on the line, right?
Yep. Okay. Great. I'm looking over your shoulder, Joel. All right.
So because I know he's going to mention it to me, I'm going to put our first question to Vic. If I don't, it's going to be an uncomfortable lunch for me.
Thanks so much for taking the question, Vic Chopra from Wells Fargo. So two questions for me. One, maybe just about the 20% CAGR over your LRP. Do you see that being consistent over the planning period? Is it linear? Can you help us with the phasing? And then I had a follow-up, please.
Yeah, I think I mentioned it a little bit earlier. We do see increasing growth drivers over the planned period. We're not guiding beyond 25 for today. But hopefully what you've seen here is you're really building in a crescendo of the things that we think are important to execute on over the period.
We would think that it'd have a positive effect on the business over time too.
Okay. Got it. My follow-up question is, you know, based on our channel checks, epileptologists and docs really seem to like the clinical support aspect that NeuroPace provides. But as the company grows, Joel, how do you think you'll have to evolve to meet a growing number of patients? Thanks.
That's such a great question. It is, in fact, it's vast. So as we grow the number of patients, then that number of patients that needs to be supported, and as Dr. Rao mentioned it, the good news is we got a lot of patients and a lot of data. The bad news is we got a lot of patients and a lot of data. So what are we going to do about that?
That's really the focus that Dylan and Brett walked through around how do we make it as efficient and as easy as possible to scale patients in your practice that gives us leverage both with clinic time as well as field clinical organization time. What we can do then is we can much more quickly get to the data that matters with recommendations that are impactful. We then get leverage through a lot of what we've already built. We absolutely see it. I hope you see in the pipeline here. We feel like we're all over it from a product development perspective to get that leverage through the income statement as we grow the implant numbers. Yeah, Priya.
Yes. I think earlier, Dr. Richardson kind of talked about some of the treatment gaps that still persist today.
And as we talk about the care program and expansion and some of the goals for 2025, I guess any early experiences that you can allude to? And then also on that, what contributions are you considering from the care program over your LRP?
Yep. So we mentioned in Q3, and I would emphasize that for our Q4 as well, that we've started to see a meaningful impact from both implants as well as referrals from the care program. So it's encouraging to see that. And Katie mentioned on her or had on her slide, I don't know if she mentioned it directly, but she had on her slide that our goal here in the coming years is to double that impact. And so we see the momentum today, and it's our intention to build on it.
Again, the care program has the benefit of, or the potential benefit anyway, if we do it correctly, I believe, of both contributing implants and referrals for the current patients, but also positioning us for suggesting a positive clinical study with Nautilus and an idiopathic generalized approval, those patients not having to go through the phase two testing and be referred into a comprehensive epilepsy center. It also positions us then to be able to go direct to RNS with those patients from a phase one testing perspective as well. So the barrier to entry in that population is lower. So having more implanting sites out in the community can give us, again, kind of more shots on goal through that program with the expanding indication. So all of that is intended to work together and be reinforcing.
The work that we're doing today with the care program and the current population is intended then to put us in position to do even more when we work to expand indications.
And then just quick, if I could squeeze one more in. You talked about new product launches and software updates. Just anything to consider from a margin implication perspective going forward with those?
Well, I think there's a couple of things. One, we continue to work with, for example, the hardware introductions that we're making. We continue to work to make sure we have the reliability and the customization into those things like the monitor to put us in a position to benefit cost-wise.
But one of the biggest things I think about from a gross margin perspective and the way we have our product pipeline constructed is with software, the rapidity and the infrastructure requirements to develop a software release versus a hardware release are levels of magnitude different. And so we're putting ourselves in a position here to be able to establish really an annual set of product releases on the current platform that can improve easy use, efficiency, and effectiveness without having to go through what you might think about as a traditional kind of a hardware platform medical device cycle and timeline and margin implications. So I like the balance in the portfolio and the way that can benefit us from a gross margin perspective as well as an opportunity cost perspective of doing it all on hardware. Yeah, Roy.
Yeah, hi, Rohan Patel, J.P. Morgan.
Thanks so much for all that, all around guidance for 2025 and the LRP. I guess just going off of software discussion, I think you mentioned, thank you. I think you mentioned an AI-powered seizure detection launch. I believe it was this year, targeted launch this year. So I guess how do you view that from a monetization standpoint, both this year and longer term? And then maybe if you could, and I have a follow-up, actually.
Yeah, I think from a monetization perspective, there's really a couple of things there.
One, as we are more efficient in managing data and then being more effective in deploying RNS to help patients, we think it both drives initial implants and preference for the RNS system, as well as then again gives us leverage to the majority of the income statement to be able to do so without having to scale the whole organization to do it. Maybe you were alluding to a little bit from a monetization perspective as well. What about monetization of data? And I won't say too much about that other than to say that we absolutely think about, and I think our Rapport relationship is an example of where we can monetize the value of that data in different ways than maybe the more traditional way of simply doing so through a device sale.
Got it.
I know last quarter you talked about how replacement revenues were nearing a trough. Obviously, coming up in the nine years, that should, by expectations, I feel like that should increase from here on out. Is that contemplated in the guide at all? And what are you guys assuming as far as replacement revenue growth for 2025 and over the LRP?
We won't break out in particular replacement revenue growth specifically, but it was absolutely contemplated in the guide. And you're right in characterizing it that we feel like we've kind of hit the midway here of the trough of the Model 300s really having been replaced, but the 320s not having enough life behind them yet to have any kind of a meaningful impact. So we get opportunistic kind of replacements. But the whole replacement cycle is really yet to come on the 320.
But we do absolutely think about it both in terms of the 2025 guidance and then factors into our thinking strategically for the LRP as well.
Got it. Thank you.
Yeah, Frank.
Thanks, Frank Takkinen with Lake Street. I wanted to ask two questions. I'll just ask them upfront. One was kind of moving to frontline therapy. And then secondly, just the concept around network stimulation. With some of the new data that was put out today, is that enough to recommend the RNS System over destructive procedures when it is a focal patient and moving to frontline? And then secondly, is there a future where network stim is the standard of care where you just go straight to network stim and avoid that stage two invasive EEG?
Yeah, go ahead. Should I? It's a great question. Yeah.
I'll do the, I'll talk about the network stim first.
Network stim is a good approach for people with regional or more widespread multifocal onsets. But the data I showed you is not with network stim. The data I showed you is with understanding where seizures are coming from and targeting that area. So I see that we're going to do both. And both are going to be driving, moving our therapy earlier. The data I showed you from the post-approval study already sees that happening. That RNS is certainly at this point moving ahead of VNS, that's for sure. And the numbers of surgeries in the United States, the resective surgeries or ablative surgeries are either staying the same or they're going down. And the RNS is obviously going up Thanks, Marty.
Thank you.
Ross
Hi. Ross Osborne with Canaccord Genuity. So they started with a NAUTILUS trial.
What outcomes or insights outside of primary endpoints do you think will be interesting to the clinical community and serve as a driver of adoption?
Yeah, well, our endpoints, which are public, are the time to the second generalized seizure in the sham stimulation people, assuming that if you can keep people from having a seizure longer, that that shows efficacy. But we'll also be looking at the one-year data compared to the patient's baseline. And we'll also be looking at the numbers of days that people experience with no seizures at all. So those are primary. In addition, we have sleep and quality of life endpoints, behavioral endpoints, and all of those contribute to the overall quality of life.
And Katie, would you discuss what an ideal target center looks like for the care program?
Sure. Yeah.
As we're looking at new implanting centers, where we're looking is places where there is a neurosurgeon. That's key. A center that's interested in drawing in new patients. So again, if they have the equipment needed to do the implantation, that's a great fast starter. Oftentimes, we have to work, find a champion within the organization who will help with getting the equipment on site. And then really just a hunger for addressing more patients within their site. There are some sites that can get up and running really pretty efficiently. We find that there's another group that takes some time to get the infrastructure in place. And those are places where we really can start working on referral relationships in the short term while we're getting the equipment on site.
And then there's some centers that won't ever do new RNS implants, but they have a lot of patients that can benefit from the therapy. And those are areas where we're really working on the referral aspect.
And let me add on the reason that physicians don't send patients to epilepsy centers is very often because they don't want to lose their patient. That's a big motivator. But with the changes in qualification requirements last year, FDA loosened up on that. So there's the opportunity for physicians to refer their patients to a center that can offer the evaluation implant procedure. But now they can get their patients back, and they can manage the programming themselves. So you don't lose the patient that you really care about.
Thank you.
Brad, I'm going to ask that a number of us will be hanging around here, but we're over time.
You've all been very gracious with your time and your interest so far today. I think we're going to try and wrap it up. Again, really appreciate the opportunity for us to talk about the story of what's going on at NeuroPace. We couldn't be more excited. We see so much opportunity in front of us, and we really appreciate your interest and ongoing support. Thanks for being here today.