uniQure N.V. (QURE) Earnings Call Transcript & Summary

Good morning, everyone, and welcome to uniQure's Virtual Investor Event. I'm Maria Cantor, Chief Corporate Affairs Officer at uniQure. We appreciate you taking time to be with us today as we feature our pipeline program in temporal lobe epilepsy and highlight our recent advances in manufacturing technology. Before we get started, please note that we'll be making a number of forward-looking statements this morning. We suggest to take a moment to review this slide, which contains our forward-looking statements. These statements involve risks and uncertainties, many of which are beyond uniQure's control and actual results could differ materially from these statements. For a detailed description of these risks and uncertainties, we encourage you to review the company's most recent Form 10-Q filed with the Securities and Exchange Commission as well as the company's other SEC filings. Our program will begin this morning with an introduction by Matt Kapusta, uniQure's CEO, on the company's history of leadership in gene therapy, our recent accomplishments and our focus for the next few years. Ricardo Dolmetsch, our President of Research and Development, will explain how the combination of our miQURE and LinQURE technologies will be instrumental in advancing our research pipeline, starting with our program in refractory temporal lobe epilepsy. Our featured speaker, Dr. Ellen Bubrick of Brigham and Women's Hospital in Boston, will provide an educational overview of epilepsy and refractory TLE. We'll then highlight AMT-260 uniQure's gene therapy product candidate for TLE and its anticipated clinical development program. After this, we'll have a 15-minute Q&A session available for questions from our research analysts. We'll then transition to hearing about advances in uniQure's manufacturing platform technologies, and we'll have our second 15-minute Q&A session. For our research analysts, if you would like to submit a question. Please use the "Ask a Question" box that is part of this webcast. We'll answer as many questions as time allows. We also note that biographies of all of this morning's speakers are available as a PDF also on the screen for your convenience. Now here's Matt Kapusta, uniQure's CEO.

Thanks, Maria. And on behalf of the uniQure team, thanks for joining us here at our 2022 virtual R&D investor event. During the quarter of a century ago, uniQure was founded with a singular mission of delivering potentially curative onetime administered genomic medicines with the goal of transforming patients' lives. We delivered on this mission a decade ago with the world first approved AAV gene therapy and last week, we announced another first with the approval of a gene therapy for patients living with hemophilia B. The HEMGENIX approval assures in a new treatment paradigm for patients living with hemophilia B and not only represents a major accomplishment for uniQure but a historic milestone for the entire field of medicine. None of this would even be possible without the tireless dedication of our employees' as well as the commitment from study investigators to patients and their families. I also believe that the successful development of HEMGENIX embodies the spirit of innovation is at the heart of who we are at uniQure and is illustrated on the next slide. Back in 2017, we announced the transition of our hemophilia B program to a next-generation construct with the potentially best-in-class. Despite incorporating a different transgene and the associated regulatory challenges involved with making a change like this midstream, we were able to substantiate manufacturing comparability with regulators which enabled us to rapidly initiate a Phase IIb study in parallel distorting our pivotal HOPE-B registrational trial. And in less than 2 years from study start, we completed enrollment of a 57 patient registrational study across more than 35 sites in the U.S. and Europe. And with the largest trail ever conducted with a hemophilia B gene therapy and we successfully coordinated hundreds of patient follow-up visits in the midst of a global pandemic. This remarkable journey to approval, which took just 5 years from the initial unveiling in 2017 would not have been possible without the innovative thinking demonstrated by our teams on a daily basis and their ability to solve complex cross-functional problems and the strength of our manufacturing and technology core platforms. Next slide, please. While the field of gene therapy has advanced significantly over the years, we're still in the very early innings. That said, the approval of HEMGENIX represents a seminal moment as the field transitions from a largely ultrarare infantile disorders to more prevalent rare diseases. This is a critically important trend in which uniQure wants to play a central role. For example, while hemophilia B is a rare disease, they're actually up to 15,000 people living with hemophilia B in the U.S. and Europe alone making it one of the more prevalent rare disease is currently being addressed by an approved gene therapy. Huntington's disease represents an even larger unmet need with approximately 60,000 to 70,000 people genetically confirmed in the U.S. and Europe and hundreds of thousands more potentially at risk of HD, but remain untested given the lack of therapeutic options. We continue to be very excited about the potential for AMT-130 and look forward to presenting additional data on this important program next year. Temporal lobe epilepsy represents even further step function increase with over 1 million people impacted by this most common form of epilepsy. And despite multiple antiseizure medications available, the reality is that most people with temporal lobe epilepsy do not respond adequately to therapy. This leaves a very significant unmet need and an exciting application for gene therapy that will be a focus of today's presentation. Next slide, please. Opportunities like hemophilia B, Huntington's disease and temporal lobe epilepsy reflect the broad range of gene therapies applicability, and I believe uniQure is in a prime position to take advantage of this opportunity. UniQure has what I believe is the leading validated technology platform, coupled with established and licensed manufacturing and quality management system capabilities. We have a broad proprietary pipeline focused on meaningful unmet needs across neurology and other rare diseases that have the potential to generate significant catalysts over the next 1 to 2 years. And finally, we have the resources, both human and financial, to support continued execution of our plan and for these reasons, we remain very enthusiastic about uniQure's future. Next slide, please. I'd like to now transition to the meat of our presentation. As Maria mentioned, we have a number of distinguished speakers. From the company, we have Ricardo Dolmetsch, uniQure's President of R&D; Rich Porter, former CEO of Corlieve Therapeutics and current Chief Business Officer at UniQure, Andreas Borta, VP of Clinical Development; and Pierre Caloz, our Chief Operating Officer. We're also delighted to have with us Dr. Ellen Bubrick, Associate Chair of Neurology at Brigham Women's. With that, I'd like to hand the presentation over to Ricardo who will discuss our miQURE and LinQURE technology platforms in more detail. Ricardo?

Thank you, Matt. As you heard from Matt, UniQure 20:45 has a long history of innovation in gene therapy. Over the course of our 20-year history, we've had many technological firsts, including the first AAV delivery of a MicroRNA in humans, the first gene therapy for Huntington's disease center to the clinic and the first use of the baculovirus platform to produce GMP quality gene therapies. Our scientific strategy relies on excellence in 3 technological areas. The first pillar of our strategy are our Cargo technologies. The Cargo is the genetically encoded material that is contained in an AAV. It is the sharp end of the therapeutic spear. In the next few slides, I will take you through 2 of our most important technologies, miQURE and LinQURE. These are the technologies that have allowed us to develop and deliver highly potent and specific MicroRNAs that can be used to reduce the expression of any gene in the cell. We have also developed the ability to deliver antibodies using adeno-associated viruses, a technology that we call AbQURE. Finally, we have developed the ability to simultaneously deliver a MicroRNA to reduce the expression of a pathogenic gene and at the same time, to deliver a corrected gene in a single AAV, a technology that we call GoQURE. These technologies give us unprecedented control over the biology of cells and open new vistas for the development of treatments that were not available in the past. Our second technological pillar is the Delivery technologies. The Delivery technologies are like the shaft of the therapeutics spear. HEMGENIX is based on adeno-associated virus 5, a viral serotype that we have pioneered and that have been able to dose safely and effectively in our hemophilia B program. Importantly, we can dose AAV5 irrespectively of the titer of preexisting neutralizing antibodies in humans. This is transformative because it allows us to dose essentially all the population of patients with hemophilia B in our study. We have also been able to use this technology as the basis for our redosing platform, which allows us to provide the second dose of the gene therapy if the first dose doesn't meet the appropriate level. Finally, our Smart AAV and Brain X capsids provide better transduction of the [indiscernible] the brain and have a potential to increase the convenience and efficacy of future therapies. The third pillar of our technological strategy is our Manufacturing platform. We are one of the few GMP manufacturers of gene therapies in the world. We have pioneered large-scale inexpensive baculoviral AAV manufacturing and have developed many of the technologies required to do this. Pierre will bring you up to date on the remarkable improvements we have made to our platform over the last few years. If I could have the next slide. Our event today is focused on AMT-260, which is our gene therapy for refractory temporal lobe epilepsy. A key component of this program is our miQURE technology. RNA interference is a revolutionary way of producing the expression of the gene but optimizing this technology to be able to deliver it safely and effectively has taken us many years. The miQURE scaffold is designed to prevent saturation of the cellular MicroRNA machinery, specifically the risk complex, which can lead to toxicity. It also doesn't generate a passenger strand. Both risk saturation and off-target strands are potentially toxic. In fact, we have shown that our micro scaffold is saved in mice, rats, pigs, nonhuman primates and most recently, in humans. An additional feature of the miQURE platform is that it doesn't generate any immunogenic products, and it can be turned to safely to knock down a gene between 50% and 80%. This provides a margin of safety for genes in which complete loss of function could be deleterious. If I could have the next slide. We have demonstrated the efficacy of our miQURE platform in our AMT-130 program, where we can reduce the expression of mutant Huntington by more than 50% in human patients with Huntington's disease. We have also shown this in a variety of preclinical programs, including our preclinical program for amyotrophic lateral sclerosis, which targets C9orf72. There, we can reduce the expression of C9orf by more than 80% in mice with relatively modest doses of our gene therapy and our excellent safety. If I could have the next slide. The natural evolution of our miQURE platform is our LinQURE platform, which gives us the ability to encode multiple MicroRNAs in a single adeno-associated virus. This is an extremely powerful thing to do because it allows us to generate adeno-associated viruses that are highly effective at reducing the expression of their targets. Potency is critical in gene therapy because it decreases the amount of a virus that needs to be produced and delivered and reduces the potential for toxicity. LinQURE has another exciting application, which is the ability to target multiple genes that contribute to the pathogenesis of the disease. This allows us to explore combinatorial therapy using a single drug, which is a potential game changer in many areas of medicine. If I could on the next slide. AMT-260 is the first example of an AAV gene therapy that uses our LinQURE technology to include multiple MicroRNAs. As you will hear from Rich Porter later in the presentation, AMT-260 targets the GRIK2 gene, which encodes the kainate glutamate receptor. It targets 2 regions of the gene, which make it potent at low concentrations. And as you'll see, highly effective at reducing seizures in a mouse model of temporal lobe epilepsy and quite safe in nonhuman primates. If I could have the next slide. We see AMT-260 as the beginning of a new platform for uniQure. Many diseases of the nervous system are caused by pathological activity of a small set of neurons. The most obvious example of this is focal epilepsy, which is triggered by a small region of inappropriately active neurons. With AMT-260, we are delivering a small amount of AAV to those neurons, which reduces the expression of a neurotransmitter receptor reduces their electrical activity and prevents them from causing a seizure. This approach can be generalized to other types of epilepsy into disorders like tremor, pain syndromes and ultimately to psychiatric disorders. These are all diseases that have at their core be inappropriate activity of a small set of neurons in a specific region of the nerve system. These indications require small amounts of an adeno-associated virus, so the manufacturing effort is comparatively small. The fact that the AAV is not systemically distributed reduces the potential safety risk. In addition, the clinical development plan is potentially faster because the trials address symptoms that are at least in principle, rapidly reversible. We are excited about this new approach, and we'll be announcing additional programs using this platform over the coming years. With this, let me hand it over to Dr. Ellen Bubrick, who will discuss the clinical presentation and the unmet medical need in patients with refractory temporal lobe epilepsy. Ellen?

Thank you very much, Ricardo. I'm very happy to be here to talk about this disorder that's very near and dear to my heart. I've dedicated my career to try and understand and treat and care for patients suffering with this disorder. Just very briefly, of course, I consult and advice for uniQure. I've also done a lot of clinical trials, although I mostly a clinician at heart, I have done clinical research. You can see some of my funding there. So we'll start with just some basic definitions. We'll talk a little bit about the demographics of the disorder, the clinical overview, some of the current treatments and their limitations. As Ricardo already mentioned, the word seizure is used to describe the clinical manifestation of an abnormal excessive expectation of a population of neurons in the cortex. I'll use the word seizure and epilepsy interchangeably, but technically, epilepsy is the tendency, is the disorder, the tendency towards recurrent seizures, unprovoked by systemic or neurologic insults. The International League against epilepsy came out with a really much more comprehensive definition to really try to understand and incorporate all of the -- that these patients experience and endure. The disorder of the brain characterized by an enduring predisposition to generate epileptic seizures and by the neurobiologic, cognitive, psychological and social consequences of this condition, all really important and part of the patient suffering. Next slide. Epilepsy is a very common neurologic disorder. It's up there with Alzheimer's and stroke. It affects 3.5 million people in the U.S. alone. And you can see here, it's 3x is common as Parkinson's and multiple sclerosis combined. I just always point that out when I gave [ thoughts ] because it's -- we don't get as much attention, maybe our funding, but there's a lot of disability associated with this disorder, and it's really very common. The Institutes of Medicine have estimated that 1 in 26 people will develop seizures in their lifetime. So we look at of that 3.5 -- there's 3.5 million people in the U.S. that have epilepsy, 65 million worldwide, hundreds of thousands diagnosed each year and more than 0.5 million in adults over 65 in the U.S. And the majority of these are focal. We'll talk in a minute. There are other types of epilepsy, but the majority are focal seizures. Temporal lobe epilepsy is the most common form. And the majority of patients are drug resistant. So that leaves us with this large number of patients that don't have a lot of great options. As was already mentioned, 0.75 million people in the U.S. alone have refractory TLE in -- and some estimates are actually closer to 1 million. It affects people of all ages, races and ethnic backgrounds and epilepsy can develop at any time in life. Today, we're talking mostly about TLE in adults. The pediatric epilepsy population is really quite separate, and we're just the tip of the iceberg compared to that. But today, we'll focus on adults and it's quite common in adults as well. Next slide. So there's really 2 ways to sort of broadly characterize seizure types and I'll do this just to frame sort of where temporal lobe epilepsy is in this classification. There's focal onset, which you can see there, the little orange area on the brain, where we talk about seizures emanating from one focal area of the brain, abnormal excitation of discharges in one area that can sort of remain local in that area in effect correlating parts of the brain and the body. And sometimes that can involve impaired awareness it can also progress to what used to be called the grand mal seizure of tonic-clonic convulsion. So any focal seizure can have impaired awareness and can progress to a bigger seizure. As opposed to generalized seizures, which you can see the little brain on the left there, are where the brain sort of see us all at once, and that's a separate category. Or the or can be unknown, sometimes you walk in at the end of the seizure you don't know how it started, but eventually, it declares itself. Next slide, please. So as was mentioned, temporal lobe epilepsy is the most common type of epilepsy. I use this slide, the top here from [ Stanley ] study in New England Journal that shows the area of surgical resection when these patients are drug-resistant and need to have surgery because you see how much brand needs to come out. And so you can understand the risk of cognitive deficits and other problems with such a surgery. So you can see laterally there, that's shaded a little bit pink, the anterior temporal lobe and then underneath the surface of the brain, you can see the hippocampus and the amygdala complex. Temporal lobe epilepsy is a clinical diagnosis although often the EEG is abnormal, not always. Often the MRI is abnormal, not always. The most common MRI finding and correlating pathologic substrate with this disorder is mesial temporal or hippocampal sclerosis. And you can see the little arrow there. On the other side, without the arrow, hippocampus means seahorse. So you can see that sort of beautiful seahorse shaped hippocampus. But on the abnormal side, it's knotted up into a tight little white ball, it's gliotic -- it's sclerotic and so you see increased a signal intensity and T2 weighted imaging. And as was mentioned, temporal lobe epilepsy tends to be drug resistant and often involves impaired awareness and correlating disability. Next slide, please. So just briefly, thinking about the clinical syndromes, the clinical situations patients deal with when they have seizures. When the seizures emanate from the temporal lobe, often it can start with an Aura, which can sometimes just be a feeling in terms of called the psychic feeling like a deja vu feeling or jamais vu sensation where sort of suddenly nothing is familiar, even walking down your street is suddenly unfamiliar. It can progress to sort of a dreamy like state with altered awareness. People will say, he was looking right at me, but it was like I was looking through them. He really wasn't paying attention. They weren't there. There was some altered awareness. It can progress to chewing, lip smacking, other automatisms of the mouth and throat and sort of picking or fidgeting, sometimes can be associated with a bad smell olfactory hallucinations and visual as well. Next slide. So that's the beginning of the seizure. You can see that -- it can be just the original Aura, which sometimes if that comes from the amygdala, can be debilitating fear and panic seizures. And the Aura, of course, as we mentioned, can progress. And that's really where the disability starts. It's very common that these people are altered wherever they are. It's sort of unpredicted when these seizures happen, and they're very frequent. There are patients who have been found walking around in the stow confused during a seizure or after a seizure. So even the Aura can be debilitating. We've done epilepsy surgery in patients for Auras alone, they can be so debilitating. But then to be in a state of impaired awareness where you're really vulnerable to injury and other issues and to be walking around, confused and then can progress -- used to be called secondary generalization to what used to be called the Grand mal and now we say bilateral tonic-clonic seizures, where you can see the tonic-clonic, that's sort of the more TV form, where people are like trashing about and fumbling the mouth and on the ground. So any focal seizure can progress to those, and those are extremely dangerous. Next slide, please. The burden that these patients carry is really very, very high, very heavy. Ongoing seizures can be very debilitating often associated with morbidity and mortality. We'll talk about those. The medications that we have available often have significant side effects. I spent most of my day trying to mitigate these kind of side effects that's really very disabling. These patients have ongoing cognitive impairment, mostly memory but also attentional trouble concentrating. There's high rates of comorbid depression and anxiety. There are high rates of seizure-related injuries like burns and falls just last week. I had a patient that didn't come to clinic at the Brigham. If you have epilepsy, and you don't come to clinic, we call you. And the woman was said I'm at the bottom of my stairs and there's -- I'm in a pool of blood. So she had a seizure, I'm sure and fell and so I had to have an ambulance. Go get her and meet her in the emergency room, and that's not the first time I've had to do that. So these patients are vulnerable to being injured quite a bit. There's rates of accidental deaths by drowning and other sorts. There's an elevated suicide rate. There's a sudden unexpected death in epilepsy syndrome, which is as high as 1 in a 100 in a lot of patients that we take care of in surgical epilepsy candidates. There's a significant amount of shame and embarrassment and social stigma associated with this disorder that is part of what impairs their quality of life in a lot of ways and contributes to their suffering. It's important to remember even one seizure per year is associated with a low quality of life. A lot of our patients have many seizures a month or many seizures a week. But even one seizure a year, is associated with the lower quality of life, one seizure a year, keeps your patient from driving, often can keep your patient from gainful employment and prevents fulfilling social integration. Next slide, please. So there are a lot of treatment options in epilepsy, but they all have limitations and their problems, which is why we're always looking for better treatments and partly why I'm very excited to be here today. The mainstay of epilepsy treatment is antiseizure medications, which we'll talk about epilepsy surgery, resection or laser ablation is possible for some as well as implantation of neurostimulatory or neuromodulatory devices such as a Vagus nerve stimulator, responsive neurostimulation and deep brain stimulation. Dietary therapies are still used on occasion, the ketogenic diet is sometimes still used when we're desperate and other alternative therapies as well. Next slide, please. So you can see here, there's a lot of medications that have been approved for epilepsy, the majority from -- since the 1990s. So there's a lot to choose from, but as we'll talk about, there's still a portion of patients that don't respond. Next slide, please. And you can forward one more. I think another there you go. Thank you. So in talking about the medications, which is really important because whenever I give these talk people feel like, well, there's so many medicines out there. Why do you need more treatments for epilepsy? And it's true. There are a lot of medications out there, but they have their limitations and their problems. It's a whole process to sort out which to start with and which to use and whether to use monotherapy, polytherapy, how do it -- but how we will administer it, if they're in the hospital or if we are at home? It's really important to think about the pharmacokinetics and drug-drug interactions. And unfortunately, these drugs have a lot of them. So a lot of time goes into thinking about some of these are inducing agents and induce the metabolism of other seizure drugs or other drugs that patients are on. Some of them are inhibiting agents and inhibit the metabolism of other seizure drugs or other medications that they're on, and doses need to be -- levels need to be drawn and doses need to be adjusted accordingly all the time. And so there's a lot of -- it's a lot of monitoring, a lot of hospital time and a lot of worry for the patients and for us. There's a lot of concomitant disorders to think about when we're treating these patients. As mentioned, there's depression, anxiety. There's insomnia. We always think about weight. So when I'm trying to think about the right drug combination, some of them are helpful for treating depression anxiety. They can help with sleep. Many of them are weight neutral. But some cause weight gain and some caused and some insomnia or sleep disruption and many can cause mood disorders. So it's really important to think about all of those things, and you can already see how complicated this is with someone who's seizing all the time, but trying to mitigate these medication side effects. Some of these are teratogens. So we have to be very careful about using them in people childbearing age. That's a whole topic in and of itself. The side effects, which I'll talk about shortly, but that's a big problem, of course, efficacy and the cost of these medications, which can be prohibitive for some patients. So the side effects are a big problem. And there's always a trade-off between trying to find medications that actually work for these patients and something they can tolerate. And I always say that mitigating side effects is as important as seizure control because it's -- there's a lot of people who just won't take them or can't take them because of the side effects. So these are very general broad categories. Of course, we always worry about rashes and allergy, organ toxicity. We have to do a lot of drug monitoring for these patients. But the main issue is the neurologic, the central nervous system effects. These drugs are not targeted therapies. They work all over the brain. And when someone has a seizure, we give them more drug, and then they get even more side effects. And so there's a lot of cognitive slowing, mood problems, dizziness, drowsiness, ataxia and I mean these people are going to be very highly sedated. It's really disruptive and have to leave their work or can't concentrate. It's really dangerous for some of them. We've talked about drug monitoring. And of course, there's some drug-induced chronic diseases we have to think about as well like osteoporosis. Next slide, please. Even when patients are on medications, there are triggers that can make their medications worse. The medicines are not really a cure or a treatment, they sort of suppress the problem for a while, but even missing a dose of the medication can result in a recurrent seizure even just one dose and people couldn't have seizures after a year or 2 of seize your freedom. Lack of sleep is a big trigger for many. Again, even when they're on medications, the disorders, just like right under the surface there, illness or infection, there can be recurrence with fevers, increased stress level is a real trigger for many people, and we're all under a lot of stress. Excessive alcohol use or drug use can be triggering, certain types of epilepsy can be triggered by flickering lights. Many people don't have triggers that they can identify. They just have the seizures and it's unclear why. Given all the medications that we have and their problems, some of them work, and that's a great day in clinic when they work, and you can get someone feeling good. But that's -- there's a huge population of patients for which that's just not the case. Kwan and Brodie put out this article in New England Journal in 2000, which was really the sort of defining moment for drug-resistant epilepsy -- focal epilepsy, where you can see here that when you start the first seizure drug in someone, you can get about half the patients' seizure-free, 47% can get seizure-free with the drug, and that's great. if someone fails the first drug, you can squeeze out another 13%, we'll get seizure-free with the second drug. But no matter how many different drug trials you try, about that big list that we saw in that graph, there remains over 1/3 that are uncontrolled, that are drug resistant. And so I always think about that in clinic when someone has new seizures and you've already tried one drug, I'm already thinking is there anything else we're going to be able to do for this person if they failed on drug? Because I know there's only a few percentage chance that they're going to get seizure-free with the second drug. Next slide, please. And so after that came out, the International League Against Epilepsy, defined drug-resistant epilepsy is failure of 2 drugs, to tolerated appropriately chosen and used antiepileptic now we call it anti-seizure medication schedules. So only you don't have to fail 2 and your drug resistant. So it's -- we don't spend time anymore trying 10 different meds. If you don't get seizure control with 2, you are unlikely to get seizure control at all with medications alone, which is why we need to think about other kinds of treatments. Next slide, please. So surgery is an option for some patients. It's a huge undertaking, as I'm sure you can understand. The patients go through a lot. You can see a lot of tissue is taken as we discussed, there are significant risks, including hemorrhage, stroke, infection, cognitive decline, is a big risk in the temporal lobe, especially on the dominant side, but either side. And of course, there's a failure rate, there's a risk of recurrent seizures. So it's great when it works, but it's problematic and it has its limitations. Next slide, please. I think there might be one more there. Sorry, yes, thank you. So we got all excited with neurostimulation. A couple of decades ago, thinking this might be a new cure for epilepsy, but it's unfortunately mainly just palliative. There's lots of different ways to stimulate the brain. In epilepsy, we're sort of stuck on the bottom there with Vagus nerves stimulator, Deep brain stimulator and Responsive neurostimulators. That one is goes -- the generator itself goes into full thickness skull you can see there, very invasive permanent implantations. And they're really only palliative when you look at outcomes. And so that's a hard sell. And it's hard to -- if we think we can get some benefit, we do them. But it's really hard to go through this and for only palliative outcome. Next slide, please. So that takes me to my last slide, which is sort of the combination of sort of everything that I wanted to express to you today is to explain the clinical problem. We talked about this is a large population with a heavy disease burden. It's high cost. And there, I mean more than just psychological and physical but actual costs, greater than $10 billion. These patients cost more than $10 billion a year in the U.S. If you think about all the hospitalizations, all the ambulance rides, all the monitoring, all every time the repeat MRI. I mean it's just -- it's astronomical. And that's patients with refractory disease. Drug resistance is a major problem. Greater than 20, so more than 20 new anti-seizure meds have been developed since 1993. We saw in that graph, more than 20 meds. But not one has changed the statistic that 1/3 of these patients just don't respond to them. And so they just don't work for these patients and of all different types of mechanisms now. Epilepsy surgery has its limitations. I love epilepsy surgery. I direct the epilepsy surgery program at the Brigham. When it works, it's great. But only a small percentage of patients are eligible. As we mentioned, there's a risk of cognitive decline and there's worrisome failure rates -- recurrence rates. Implantable devices, as we talked about also have their limitations, they're permanent. A few are eligible for that as well and again, only palliative outcomes. There's really no good evidence that everyone asked us when I give talks about marijuana, so I put a line up there about marijuana and cannabinoids. Despite what is in the media, a lot of hype about it. There's 2 very rare pediatric epilepsy disorders that have some data with marijuana. But the majority of -- we don't recommend it in adults, it just doesn't work, and it can cause a lot more problems. So that's not -- the hype does not reflect the reality. And so that leads us really with this critical unmet need for innovative treatments that we're just -- we're behind a lot of other disorders in terms of how few patients are -- we can treat with the availability of what we have today. So I will finish there. And with that, I'll hand it over to Rich Porter. Thank you.

Thank you, Dr. Bubrick. So I'd like to introduce AMT-260. This is an example of the LinQURE technology as introduced by Ricardo earlier and it's delivered in an AAV9 vector. Illustrated on the slide are 2 engineered MicroRNAs targeting the kainate receptor. A subtype of glutamate ionotropic receptor and the expression of these MicroRNAs are driven by the human neurons when its specific promoter called Synapsin-1. These 2 MicroRNAs target, the Gluk2 subunit of kainate receptors at different regions of the encoding Messenger RNA to prevent the production of the Gluk2 or to knock down the production of the Gluk2 subunit of kainate receptors as illustrated on the sort of cartoon on the bottom of the slide. Next slide, please. So kainate receptors are a subtype of ionotropic glutamate receptors. The mediates along with the NMDA and AMPA receptors, the fast excitatory transmission within the vein. AMPA receptors mediate fast transmission and mediated large depolarization and are rapidly deactivated as indicated on the right side of the panel. Kainate receptors, on the other hand, are more newer modulatory in nature and mediate a comparably smaller synaptic current, but importantly and critically, have a slow off rate. Next one, please. So the consequences of these different properties are illustrated on this slide, where the kainate receptors having slowed deactivation kinetic and AMPA receptors having fast deactivation kinetics. Hence, upon the repeated stimulation or excessive activation of receptors, we have a summation of the synaptic response with kainate receptors, but not with the AMPA sectors and this decreases the threshold for the sustained action potential filing, which acts as a trigger for epileptic discharges. This is why excessive stimulation of kainate receptors has been known in the literature for a decade to be epileptogenic. A specially in hippocampus, which is the most epileptogenic region of the brain. So on the next slide, kainate receptors have been demonstrated to be aberrantly expressed in the hippocampus of temporal lobe epilepsy patients. And their composed -- and kainate receptors are composed of subtypes. We are targeting the Gluk2/5 containing kainate receptors, which have been reported to be involved in epileptogenesis and this is supported by the observation that Gluk2 knockout mice are resistant to pilocarpine a form of induced temporal lobe epilepsy. In the pilocarpine model of temporal lobe epilepsy and in human patients, the triggering brain and so causes neuronal tissue to undergo a major reorganization. Specifically, starting of hippocampal mossy fibers is one of the best documented examples of seizure-induced reactive plasticity in human patients. And this leads to the formation of powerful [ McCullen ] excitatory loops by the recruitment of postsynaptic kainate receptors, and these kainate receptors are hence aberrantly expressed between subtypes of the hippocampal cells, but specifically dentate granule cells. This results in high quick accessibility of the hippocampal circuitry due to the channel kinetics properties that I've just described and generates the epileptiform activity in the hippocampus of human patients and animal models of temporal lobe epilepsy. Given that the pilocarpine model, in our view, best recapitulates the human temporal low epileptic condition, and the Gluk2 transgenic mice are protected from the epileptic seizures demonstrated in the top panel and on the bottom panel shown on the electrophysiological level by local field recordings where in the normal, [indiscernible] mice, you can see the pilo -- consequence of the pilocarpine leads to these sort of excessive epileptic spike activity, which is absent in the Gluk2 knockout mice. Further, while not directly involved in temporal lobe epilepsy per se, it's worth mentioning that there's recently been further genetic evidence for an involvement of Gluk2 hyperactivity in epilepsies where a de novo gain of function mutation has been reported to cause severe intractable epilepsies in children. On the next slide, we illustrate the finding of increased kainate binding in the hippocampus of patients with refractory epilepsy compared to non-epileptic patients. And this graphically on the middle and right-hand side illustrates our hypothesis of the recruitment of the Gluk2/5 containing receptors as a consequence of the new reviving and consequential seizure activity. Next slide, please. So since our project is predicated on this hypothesis of aberrantly increased Gluk2/5 receptor expression and the Gluk2 subunits form the ion channel and they come together and dimerize with Gluk5 receptors, which -- sorry, Gluk5 subunits, which are accessory proteins. The Gluk5 subunit does not form [ mice ] channels, it doesn't get to the synaptic membranes unless it's dimerized with Gluk2. We said about screening over 100 different candidate siRNA guide sequences designed to knock down the GRIK2 gene of the Gluk2 subunit. The most active of these subunits were then further selected and optimized not only to knockdown and to be selective for the Gluk2, but also for optimal drug-like properties and an ability to test in different species for both safety and efficacy. On the next slide, we illustrate this with AMT-260, our optimized construct where we can show a clear dose-dependent expression in the -- when injected into the hippocampus of mice. On the left, we are measuring in black a control vector which is a scrambled sequence. And in orange, we can show dose-dependent increases in the expression of the vector genomes and in the middle and right-hand panels indicated in orange, the dose-dependent increase in the expression of the MicroRNAs has detected by a specific form of PCR, the black control vector doesn't express the MicroRNAs. Next slide, please. So the expression of our specific MicroRNAs have been demonstrated to knockdown the GRIK2 gene that encodes the Gluk2 subunits of kainate receptors and in the top panel, we can see that the high expression of the Gluk2 gene as depicted in red over the cells in the hippocampus, which are blue and in the region of the injection of the vector on the bottom panel, we can show significant knockdown of the GRIK2 gene, which is what we are targeting, and this doesn't lead to any side effects in these mice. Next slide. So in our pilocarpine mice model of temporal lobe epilepsy, we have demonstrated with multiple constructs that we suppressed the epileptic activity of the mice and illustrated in the slide above, we not only suppressed the epileptic activity, we also normalized the behavioral phenotype of device. The pilocarpine model is a very severe model of temporal lobe epilepsy caused as bilateral lesions and a very severe phenotype such that these mice appear to be anxious, aggressive, they no longer maintain normal nesting and grooming behavior, and this can be shown on the left hand image -- top left-hand image of this slide. This same mouse, 2 weeks after having been injected with AMT-260 in the hippocampus, we can see that AMT-260 with stores, the mice to good health, normal grooming and a more normal behavior. So it not only abolishes the seizure activity shown in the top right-hand panel of the slide, but also in objective measures of sort of the health and high productivity of the animals restores these animals to that approaching their normal non-epileptic [indiscernible]. And on the next slide, and we illustrate this with some resected human material. We're very fortunate through our academic collaborations that we have access to biopsy material from patients as described by Dr. Bubrick, who have undergone the section surgery to have their hippocampus when we're due to their uncontrolled temporal lobe epilepsy. The hippocampus are taken from the neurosurgery suite to the labs, and they're prepared into organotypic slices of approximately 350 to 400 microns places are quite thick, which allow us to maintain them in culture for 2 to 4 weeks and maintain their high excitability of the network as recorded electrophysiologically. But with these slides, we then transduced them with AMT-260 or with a control vector and leave them in culture for approximately 12 to 14 days before recording their network activity. And we've performed these studies from patients with a variety of our contracts ranging from 2 years old to 50 years old patients and with different pathologies. And in all cases, where the slices remain viable, we've demonstrated a significant suppression of the epileptiform activity indicated in the traces on the slide, where the top slice from the same patients infected within control vector and the bottom one with AMT-260, you can see that there's a very profound suppression of the spontaneous epileptiform activity indicated by the [indiscernible]. The suppression of this epileptiform activity is associated with a knockdown of the Gluk2 containing receptors, illustrated at both the mRNA and protein levels. So in preparation for administration to patients with refractory temporal lobe epilepsy, we've investigated the injection of our AAV9 vectors into the hippocampus of nonhuman primates. Demonstrated on the left-hand panel is an MRI image depicting the administration of our vectors with gadolinium to allow real-time imaging of the convection-enhanced delivery procedure that we will be using. And you can see that this is largely contained within the hippocampus. And on the right-hand side is an immunocytochemical image of a reporter gene called GFP which indicates that our procedure largely restricts the expression of the gene to the hippocampus as we would wish, illustrating the targeted nature of the gene knockdown. So we're not only targeting a very specific epileptogenic protein, and we're also delivering this in a very focal way to the region that it's needed. On the last slide, I'd like to illustrate the fact that AMT-260 acts in a dose-dependent manner in the nonhuman primates as well. On the graph on the left-hand side, in blue, is a dose-dependent increase in the MicroRNAs that we are expressing. And in orange, this shows that this is associated in the same samples of a dose-dependent decrease in the GRIK2 gene from these animals. The right-hand panel takes this one step further and shows at the cellular level, those expressions which express those cells within the hippocampus, which express the MicroRNAs depicted in red have a knockdown of the GRIK2 gene depicted in -- I'm sorry, the MicroRNA is depicted in green and the cell and the good -- the messenger RNA for the GRIK2 in red. And so you can see that when we get high levels of the green. We don't see the Messenger RNA for the GRIK2 gene. And conversely, we can see high levels of the GRIK2 gene in red, where we don't see the expression of the MicroRNAs. These studies are forming the basis of our IND-enabling activities in preparation for the clinical studies, which will be described by Dr. Borta in the next slides.

Thanks, Richard. I'm really excited to present you the next section providing an outlook on our ongoing IND-enabling activities and how we plan to advance AMT-260 into the clinic. And the first slide you see, as Rich already have illustrated, we have initially shown that the selected road of administration in nonhuman primate is leading to very good focal transduction efficiency after AMT-260 administration. On the right side, you can see these data occurring from nonhuman primates showing a dose range finder, indicating a clear dose dependency with respect to the MicroRNA expression with a significant reduction of GRIK2 mRNA as indicated by the blue line. We are currently running our GLP tox program at a similar dose range level. And after 3 months in life phase, we have no safety findings for AMT-260. The GLP study is still ongoing and is expected to be completed in the second quarter of 2023. Another important IND-enabling activity is our GMP manufacturing which has started in early November and is expected to be completed in Q1 2023. On this slide, you can see images from our previous NHP study investigating the ADL trajectory into the hippocampus. The main area in the brain responsible for triggering seizures in temporal lobe epilepsy. On the left image, you see the projection of the hippocampus in purple on an MRI image and the precise injection into the hippocampus shown in yellow, using gadoteridol, a contrast agent being used routinely for direct intra-parenchymal pain injections. On the right side, you can see the transduction efficacy of the same nonhuman primate brain showing local transduction of the hippocampus, while no transaction occurs outside the focus area. On the next slide, I want to illustrate the route of administration in humans. We aim to reach out from occipital into the hippocampus. This route of administration is well known to the neurosurgeons in the field because it is used for a surgical procedure called laser ablation, which is used to destroy the hippocampus. You can see the white line illustrating our trajectory into the hippocampus. We go through a single bore hole into the brain using a minimal invasive approach with an appropriate cannula. As we move forward, as you can see, we administer step-by-step AMT-260 into the hippocampus to achieve complete coverage of the affected area. So the next slide is talking about our patient population. The patient population that we are targeting is presented here in the middle, you can see the temporal lobe highlighted in blue. As you heard in the previous presentation from Ellen and Richard temporal lobe epilepsy is the most common form of all focal epilepsy. Within the temporal lobe, the hippocampus is located, highlighted here in orange. This part is mainly generating seizures in temporal lobe epilepsy patients. In our first study, as pointed out on the left side, we are targeting subject with refractory temporal lobe epilepsy who have a clear pathology in the hippocampus turned hippocampal sclerosis. In this population, the hippocampus and its neural networks are dysfunctional and cause mainly seizures. This population is the largest subpopulation of all pathologies causing temporal lobe epilepsy. Our goal is to subsequently for other clinical trials to focus on the entire population of temporal lobe epilepsy patients. In our first study, as shown on this slide, the focus is on evaluating the safety and tolerability of AMT-260. On top of this, we also plan to depict the first signals for efficacy. The study is planned to be an open-label, U.S.-centric study with a control group receiving regular anti-seizure medication for a defined period of time. All subjects must have a confirmed diagnosis of refractory mesial temporal lobe epilepsy and will be monitored for a baseline period of 1 month to confirm the seizure severity and frequency. Overall, we plan to have 2 different cohorts with a total of 16 patients treated and 8 patients in the control group. We plan to offer subjects in the control group, the opportunity to cross over into the treatment arm at a later time during the study conduct. Next slide. You can see the current scheduling of the 2 dose cohorts. Each dose cohort will include 8 subjects in the active treatment arm and 4 subjects in the control arm. We plan to have all available data reviewed for each cohort by a data safety monitoring board before continuing dosing with additional subjects in a dose cohort. This will happen on 3 predefined occasions. After a number of subjects have received a dose of AMT-260, we intend to initiate an additional dose cohort. The inclusion and revaluation of subjects in the next cohort follows the same principle as for the first cohort. Next slide. This is an overview of the relevant endpoints we are considering in our phase study. As already said, the focus of this study is clearly on safety and tolerability. In addition, we will address efficacy focusing on seizure reduction. We plan to use multiple imaging modalities to carefully monitor safety, including MRI and also PET. You can see and the disease-related biomarkers on the right side of this slide. We will include a number of EEG measurements and also plan to collect some state-of-the-art biosensor data using an approved demand. On the next slide, the time lines are foreseeing that we are continuing working to complete our IND-enabling activities. This is our main priority for this and the beginning of the next year. We aim to file the IND next year, followed by the start of the first clinical study for AMT-260. The initial data cut will be available in 2024. On the next slide, you see the next steps to advance AMT-260 into the clinical development. The major components are the completion of the GMP 1 manufacturing in and the completion of the GLP tox program in Q2 2023. In order to be ready to start the clinical phase we also finalized the selection of study sites in the U.S. soon. I want to conclude with a summary. As you have seen, we have very convincing preclinical data occurring from the pilot camping model, supported by excellent biodistribution in nonhuman primates. We have been able to confirm target engagement in several species. All data suggest that we have a clinical candidate at hand that has an excellent safety profile. In addition, we are able to perform a targeted delivery procedure into the hippocampus. With that, I will hand over back to Maria for our Q&A session.

Thank you, Andreas. And now we'll have our Q&A session, and it will feature Andreas, Dr. Bubrick, Rich Porter and Ricardo, will all be part of this Q&A session going through some received questions that we have now from a few of our research analysts. If there are others, I would use this opportunity to prompt you to use the ask a box dialogue that is on this webcast to submit your question.

The first question that we received is from Joe Thome with Cowen and Joe asks Dr. Bubrick, you mentioned that not everyone is eligible for surgery or for neurostimulation. Can you please explain who is and who is not eligible? And then again, maybe another question back to Andreas is to restate the expected eligibility criteria for the AMT-260 study?

Yes. So thank you. That's an excellent question. I'm sorry, that wasn't clear. So not everyone with temporal lobe epilepsy is a surgical candidate. The candidate see is based on a few different things. We have to do video EEG monitoring to confirm that the seizures are from the temporal lobe in question. And sometimes there's discrepancy on the EEG and they would require further intraoperative EEG monitoring before we can determine their eligibility. So I guess I should have said they weren't eligible, they're not clearly eligible initially. And so a lot of patients will opt out of going through other kinds of intracranial monitoring and other invasive testing to see if they candidates. Some are ineligible due to comorbid disorders, bleeding diathesis, other kinds of medication requirements, age, sometimes comorbid severe psychiatric disease. It's a -- so it ends up being a small percentage that are referred to epilepsy centers. Not as much that aren't eligible.

Yes. And to add to the second part of the question. So our eligibility criteria for see that we clearly take patients with hippocampus sclerosis as illustrated on the slide in a certain age range, ranging from 18 to 20 -- to 65 years of age, which is a broad age range, but gives us a broad variety across the population to really be able to recruit a sufficient number for our first and initial clinical study.

If I could also just further clarify. When I say eligible for surgery, I'm talking about everyone sort of has to be on board, right? The patient to the whole epilepsy team, neurosurgery. So if you have -- if a patient has mesial temporal sclerosis and we document that the seizure is coming from there, it would be unusual that they're ineligible except for just some of the comorbidities I talked about, but many will decline or are not ready or are just not on board. And so they're not eligible to be -- to really go through with it. They really have to be -- it has to be that they're desperate and they want it. And there's just -- it's a lot of educating them about it over sometimes years to get them on board. So it's -- the total number that ultimately go through epilepsy surgery is relatively low, given the number that have the disease.

Okay. The next question comes in from Judah Frommer from Credit Suisse. And Judah asks if we can provide a little bit more background on the knockout mouse model and how the model has translated to human data in any prior TLE drug development. Rich?

I'm also happy to take that.

And Ricardo? Sure.

Sure. So the Gluk2 knockout is relatively resistant to epilepsy. As you may know, preclinical epilepsy models are among the few epilepsy -- among the few models that are really translational in neuroscience or models of depression and anxiety are not so [ paid ]. But our models of seizures have been really quite good. And the knockout is comparatively resistant that some of the models that are some of the stimulated or used to generate [indiscernible]. Having said that, nobody has actually developed with Gluk2 [indiscernible] tested in a clinic and most of the people come to something that blocks both the kainate receptors as well [indiscernible] receptors and those are relatively potent anti-seizure medicines with a very narrow safety margin and some pretty significant side effects. So what we're trying to do here is to just simply address the small set of cells that have atopic activity of the kainate receptors without affecting the entire brain.

So if I may elaborate, I think that's the critical sort of premise of our project here. A lot of the available therapies, as described by Dr. Bubrick are limited by their side effects often from a dose-limiting perspective. The involvement of kainate receptors is well established in its over activation to cause epilepsy. Even to the point of ingestion of toxicity, domoic acid from contaminated mussels causes seizures, cognitive impairment and unlike in humans. The approach that we're taking to selectively focus on the knockdown of the specifically identified subtype of kainate receptor in the region of interest only through the delivery as described by Andreas earlier, is where we believe we're going to have the differentiation.

Okay. Thank you very much. The next question comes in from Luca Issi with RBC. And Luca asks what level of GRIK2 knockdown is needed to see a clinical effect? Any risk in knocking down too much and causing destruction of normal brain function?

Shall I start, Andreas? We've got -- so what we've seen on the preclinical studies and maybe from a translation perspective in the resected hippocampal slices that I described earlier is that approximately 40% knockdown of the Gluk2 protein as identified by antibodies against the Gluk2 receptor seems to be the threshold to suppress the simultaneous activity in these slices. Further to address the second part of the question, the Gluk2 knockdown mice are relatively normal. They have a very mild phenotype. They behave normally. And in the nonhuman primates, where we have not -- where we've shown that we can knockdown over 90% of the Gluk2 gene, the -- there were no abnormal behaviors or side effects as described by Andreas. So this is why we believe the targeted nature, we have a range of possibilities, and it appears to not induce the sort of side effects that are typical of the general suppression of the excitability of the brain mediated by the existing anti-seizure medications.

And to add to that, from a clinical perspective, we are aiming to reach out only into the hippocampus so that we have a very focal very local administration of the drug. As we have shown in the nonhuman primate, there is no spread into other areas. And with a knockdown approach, another knockout approach, we are pretty certain that this is a very safe procedure and approach with regard to the target that we have at hand.

Yes. And just to make it quite clear, this is, of course, a onetime administration. One of the beauties of gene therapies, and we can just deliver it, and it's for life because neurons don't divide. And so we're fairly certain that we can locally suppress the activity in just those cells after that surgery for a long time.

So a related question now comes in from Rick Miller with Cantor and this is for you, Dr. Bubrick. Given your clinical perspective, what is your sense of what could be the profile of an ideal refractory temporal lobe epilepsy patient for a gene therapy approach like this onetime administration direct into the hippocampus? And what factors do you believe would need to factor into that decision-making, recognizing what you just said a few moments ago about surgery in general. But when considering a gene therapy approach like this onetime administration, what is your sense on what that ideal patient would be considering and how you would look at that as well?

Yes. So when a patient -- when it becomes clear that a patient is drug resistant and we're thinking about other options. This is an option that we'd like to be able to offer. I mean this is something that can be with someone who is a surgical candidate but doesn't want surgery or isn't ready yet. There's a lot of those patients that are couple of years from being ready to take the risk, the significant risk, who don't want an implantable device that is -- has marginal benefit. So this is a lot of patients with drug-resistant epilepsy that have hippocampal sclerosis. So this is sort of the core of the disorder where we would normally be hoping surgery could work. If we can convince them to do it. And if we were able to get a good outcome but then spare them these other lifelong implanted devices and things that are only palliative. So it's really the kind of -- it's not a rare patient. I'm thinking of, it's most people that I see that are deciding what between the not great options that we have that this could be an option for them.

And I can add to that is the gene therapy in our view can be a treatment that comes before resection surgery because the resection is always an option that can come later. So this is something where we give the patient an extra opportunity for efficacy treatment.

Yes, that's exactly right. So it's -- a lot of those patients, maybe they won't need the surgery. They could have this treatment, and maybe we'll treat them and they won't need the surgery. And so then they have undergone a lot less risk in a treatment that has less side effects, potentially less risk, may be able to let them come off of some of their medications and come down with some of those side effects. So there's a lot of opportunity there as well.

Okay. Thank you. The next question is from Joe Schwartz with SVB Securities. And Joe asks, given what is known about the psycho physiology of TLE, in particular, the changes that can occur at the cellular level in the brains of patients with TLE over time, how much of a challenge could issues like hippocampal sclerosis present for AMT-260? Or do you think the disease should be reversible regardless of disease progression based on its mechanism of action?

So one way of thinking about this is to think that what is actually occurring in a sclerotic hippocampus is not that you don't have any neurons left but is that the neurons there are excessively active. You've lost some neurons, and there are more glial cells than you would expect, but there's still plenty of neurons there. And those neurons are triggering the seizures because they have a whole set of recurrent connections. In fact, that is what you see when you actually remove the hippocampus in somebody who has a hippocampal sclerosis. So what we're doing is preventing those neurons from firing atopically and thus causing a seizure that spreads to the rest of the brain. So I think that the therapy should work even if there is hippocampal sclerosis.

Yes, that's exactly right. I mean the -- it's sort of a paradox in a way, clinically. These patients have very poor memory. They have a lot of memory trouble. So their hippocampus is dysfunctional. It's diseased. You can see it knotted up on the MRI yet it's quite electrically active and hyperexcitable causing all of these seizures. So there's enough alive neurons in there that are causing all this trouble. And so -- in terms of the question mentioned reversible, I don't think of it as reversible. I think of it as curative that we use surgery can sometimes be curative. If you take out the main focus that's generating the seizures from the brain, you can render someone seizure-free under the right conditions. The problem is sometimes you have to take so much out or taking it out causes so much trouble with disconnection and bleeding other things that it has trade-offs with other risks. But this -- but it's sort of similar to me, sort of shutting it down, whether by destroying it with laser, resecting it surgically or potentially this technique, you're just shutting down. It's already a dysfunctional organ. You're just shutting down the part that's causing the seizure.

Okay. Thank you. The next question comes from Debjit Chattopadhyay with Guggenheim Partners and Debjit asked for Dr. Bubrick, are the seizure events clustered? Meaning if subjects experienced frequent events heading into the clinical trial, does the frequency continue or attenuate? And how can you capture a clinically meaningful reduction and seizure rate in short studies?

Sometimes seizures cluster, some clinical trials call a cluster, one event, if it's happening over a couple of days. But the main stay for looking at reduction seizure frequency is just seizure logs. So like any clinical trial. So we watch seizure logs. And it's just a matter of the statistics, if they had, however, many a month before, and then you log it for many months after, there's just a statistical, you can see the reduction or in failed trial is not a reduction, but there's enough patients with enough seizures that you can just -- the clustering doesn't tend to be a confound in any way.

And then there's a follow-up as well. Reminding the audience, what is the vector that's used in AMT-260? And how does the transduction efficiency compare versus AAV5?

So the vector is AAV9. And our internal study suggests that AAV9 and AAV5 have transduced neurons equally well. But historically, people have used AAV9 in the CNS, which is one of the reasons why we're using AAV9 in this study. But I should just add that in our Huntington study, we're using AAV5, and we're quite certain that we have very good transduction there as well.

Okay. A question came in from Paul Matteis with Stifel. I think picking up Dr. Bubrick and what you were mentioning earlier, how familiar are doctors treating refractory TLE patients with this procedure that AMT-260 requires? How commonly do they do it? How much of a lift will it be for training to be in this clinical study?

So I mean, I'm sure none have done it with the gene therapy yet because it's yet to be done in these patients. But the procedure itself, as Andreas was mentioning, to do a targeted minimally invasive sort of like a -- it's sort of like a depth electrode that can go straight to the hippocampus is commonly done. We interrogate or investigate the hippocampus very frequently in patients where maybe we can't tell what side the seizure is coming from. Another reason they might be ineligible at first. And so we -- our surgeons at epilepsy centers put in electrodes at the same coordinates from the occipital lobe. Sometimes we'll do orthogonally, sometimes [ neurons ] lobe and the hippocampus is a fairly easy target to get to. So that particular procedure is not new. It's just the injecting of the gene therapy treatment part, but the getting there is routine at epilepsy surgery centers.

I mean let me just add that from the Huntington study, one thing we've learned is that the surgeries are -- I mean, I think the best way of thinking about these surgeries is that there are actually infusions of the gene therapy into the brain. And so we -- surgeons are very familiar with the stereotactic placement of cannulas and electrodes in specific regions of the brain and the infusion itself is comparatively [indiscernible]

A question has come in from Suji Jeong with Jefferies. As a neurons die in aberrant connections are formed in TLE patients, do you think that younger patients might benefit more from AMT-260 versus older patients?

I don't want to interrupt you. If you're about to start, Ricardo.

No, you should go ahead, Ellen.

I was just going to say, it's not -- it's a really good question. It's not so much younger patients, meaning that it would work somewhat better in a pediatric population than an adult. But we do think about duration of disease, duration of epilepsy. That's really important. And sometimes when it's many decades, it's harder to treat. So we do -- that's why I start thinking about if someone fails one drug, and I'm already on the second drug, I'm already thinking is there anything we're going to be able to do for this person either surgically or otherwise because the sooner we can get control, the better. So the longer we let this disorder sort of progress, the -- it's harder to treat a lot of ways. So yes, earlier intervention, we're introducing epilepsy surgery and other options after 2 drugs, which sometimes is you go through in a matter of months after diagnosis.

Okay. Another related question, but I think worth emphasizing, this comes from Ellie Merle with UBS. Again, elaborating on the administration procedure to the hippocampus, what is known about that procedure since we mentioned a lot is known on the procedure? And our confidence in the safety of the procedure as well as our confidence that there wouldn't be spread to other areas beyond the hippocampus?

Andreas, do you want to take that?

I can take that, yes. So this is from the beginning on our thinking that we follow the procedure that is well known in epilepsy surgery using the trajectory that is used for laser ablation positioning. And this is a -- we planned the trajectory on an individual MRI that you do prior to the injection, but you have a trajectory that is scaring critical structures, such as vascular structures and stuff like that. So this is a pretty well-known and pretty safe procedure in that field. And with regard to laser ablation, it's a technique that is destroying by ablation. The entire hippocampus is clearly what we want to avoid. So we have a cannula that is reaching out into the hippocampus to locally administered there. And by planning the trajectory accurately, we can also avoid kind of targeting into the TSF system, for example, which would lead to [indiscernible] which can be clearly avoided by coming from occipital because if you go from the temple side into the hippocampus, there is a higher risk. And this is all the thinking went into our planning for the clinical study that we come from occipital with a safe trajectory reaching out into hippocampus, which is well known in the neurosurgical field.

Yes. I mean when it comes to spreading, we're relatively confident that spread is pretty small based on not only the work we've done in rodents but also the work we've done in nonhuman primates where we can really show that the gene therapy is very localized. And I should say, this is a little different than what we do in Huntington's because, of course, it's a concentration-dependent thing. So at higher concentrations and Huntington's we can get spreading all over the brain, but in this specific case, by using a lower concentration of a virus in a different serotype, we get much more localized effects.

If it helps clarify I see this came up a couple of times. I too was reassured when they were showing the data about the lack of spread and that it stays very focused and targeted. But the actual get in there, like I said, I mean, at our center, it's routine surgical. We record from hippocampus there all the time. I mean a few times a month. Our surgeons are putting these kinds of stereotactic EEGs into the hippocampus. So it may seem very invasive when you're not used to it, but we do them every month, several, sometimes every week. So it is it's considered low risk to us and our team.

Very helpful. Thank you, Dr. Bubrick. And our last 2 questions, the first from Yanan Zhu with Wells Fargo. Dr. Bubrick, what level of reduction in EEG would be considered clinically meaningful upon gene therapy treatment? And then a separate question from Robyn Karnauskas with Truist. How much biological heterogeneity is there among patients with TLE? And is there any color on why some would response to certain meds, if not others?

I'll take the second one first. That's an excellent question. They're both excellent questions. There's a lot of heterogeneity in the disease in terms of how you get there. We don't know if people are born with a tendency to acquire their hippocampus to later in life get sclerotic. Some of these people might have had viruses when they were young and had a high fever, never really got it checked out, but ended up having hippocampus sclerosis. Some people, it can be from an old trauma. There's lots of different ways to get there. But once you're there, it's the exact same disease. They have the same seizure types, the same EEG signature, the same clinical disorder and the treatments are the same. So the heterogeneity is kinds of angles. But the disease itself by that time, there's sort of this final common pathway that they have hippocampal sclerosis and TLE. Why are some people resistant to drugs and others aren't to me is like the Nobel Prize winning question. I don't know but they are -- and like I said, no matter what we do, how many months we try, there's just 1/3 that are. There's something different about their disease. There's something different about it that we have to address it in a different way, the meds aren't going to do it. The first question is -- I'll try to address that. I'm sorry to be vague, but it's not clear. I mean, in a lot of drug trials, we don't use EEG that much because we go by clinical response. We just log seizures. We don't have a lot of great accurate biomarkers. If the EEG doesn't get that much better, I'm not worried if they're seizure free. If their seizures are reduced. It's helpful when you see that. It's reassuring especially in the animal models, I thought what the animal models here were very impressive because that's a common thing that you want to see when you induce an epileptic model in a rat, that type of reduction. That type of reduction and many other types of treatments translate to improved seizure control in humans. So it's helpful to see it in the rodent models. In humans, we do it more as a safety feature, just make sure nothing is getting worse. We check how is it doing? Make sure there's no increased swelling. And yes, there'll be some less discharges, but it's more of a -- it's less helpful than the clinical response.

Very, very helpful. Thank you to all of you. At this point now in the program, we're going to move forward to our presentation from Pierre Caloz, who is our Chief Operating Officer at uniQure and Pierre is going to feature advancements that are being made in manufacturing technology in our platform here at uniQure. Pierre?

Yes. Good morning, good afternoon, everybody. This is a nice meeting you. I hope you still have some energy after this great science to listen to me to operation and to my French accent. I'm happy to discuss with you today about -- maybe you can go to the next slide, to present to you what is our platform and why we believe our platform is quite unique? I will also illustrate the improvement that we do on our platform around speed, commercial viability, quality and regulatory. And almost what is -- the most important is how do we implement that? Because in operation, nothing makes a difference until this is actually implemented. So in operation, our -- this is our vision. Our vision is actually to translate this cutting-edge science that you've learned a little bit more about in the last hour into product for patients. So basically, we are passionate to do this, and we really believe that the how to meet this sense of purpose is actually to capitalize on our production platform. So we do have quite a bit of know-how in terms of how to make AAV, and we want to make sure that we capitalize on this production platform, which is state-of-the-art and that we want to continue to innovate in-house. Here, we have put 2 high-level objectives for our platform. One is to reduce the cost of goods. We believe that actually, I will show you that we are very close to reduce the cost of goods by a factor of 100 or more. And the other part is how can we be quicker? So how can we capitalize on the know-how that we have to make sure that every time we have a late candidate because we have great science that we can move that into an IND as soon as possible. And the benchmark that we put here is about 18 months before the late candidate identification and the IND submission. So everybody talks about platform. I just would like to take a little bit of time to explain to you what we mean by platform. And actually, we have looked at other industries, more mature industries, which are using a platform approach as well. And here, this is an illustration of the car industry, where based on one platform, they are able to actually develop different models. But again, based on the same basic structure. So what is the platform? A platform is actually a set of modules that can be developed independently. But when one module is evolving, it still remain fit for the other modules. So basically, this is a way to capitalize on the know-how to have a modular approach. So we have different teams working on different parts of the module on a car. This is typically this way. And so we are able to evolve your car or your product but capitalizing on the know-how of the platform of the past. That's exactly what we are trying to do with our manufacturing platform for AAV production. So illustrating this for uniQure. We have actually identified several modules for our platform. So this is how our platform looks like. Our platform is made of modules. So we have one module for cell and seeds. We have one module for drug substance, drug production, so upstream parts, let's say, the bioreactor side. One module for the purification, one module for the aseptic filling and then one module for the analytics. So these are the different modules of our platform. And each of those modules are then split into a unit of operation. And we have projects, we have people, we have know-how, we have documents that are organized around this platform. So we are very clear on who is responsible for module 1, 2 or 3, and who is working behind on the unit of operation XYZ, making sure that all those modules can continue to work together that we can capitalize again on the know-how that we have generated, but continue to evolve the platform. By the way, maybe just want to mention here that one of the big secret of the platform here is what is also around the modules, right? There's a lot of know-how or the glue, which is needed between those modules. So the way we collect data, the way we document what we do, the way we process between generation of module 1, 2 and 3 is actually really important to make sure that we have business continuity throughout the different generation of the platform. So what are we trying to achieve? I mean we are always trying to start with the why in mind, right? The famous start with why. So what we are trying to achieve is actually to be faster, cheaper and better. And these are not only concept world or buzz words here, but we are really trying to say, okay, what are the requirements? I mean, speed is one of the big requirements, speed to market. There is a big unmet medical need. And this is really important that we are able to capitalize on the know-how and this concept of platform, we are trying to make sure that we are not starting from scratch that we are able to start with everything that we've learned so far. So typically, what do we mean by speed. We mean that we are able to move from -- that's what I mentioned before from the lead candidate to an IND in 18 months, and we are actually quite happy to compare ourselves to larger CMO that are using, for instance, the HEK platform which are able actually also to -- roughly to meet the same time line. So we are fast. We are as fast as the best out there. In terms of commercial variability, this is the second part. Cost of goods is a very important topic. Cost of goods, we are working very hard. We are using a platform which is based on insect cells. I will explain to you why we believe that this insect sales platform is -- has reduced cost of goods. But we are working very hard on 2 aspects for our cost of goods. One is on the process directly. So to increase the yield or to increase the productivity of our sales, making sure that for each batch, we get more product, but also to get more throughput. So to be able to do more batches, so get more products per batch and get more batches per year. And typically, we are able to improve at the moment by 50x or even 100x the yield or the productivity, and we are also able using lean methodology to improve the throughput, typically 2x, 5x, 10x more per year. In terms of regulatory approval, one of the illustrations that I want to give you in terms of high-quality product is what everybody is talking about. This is the concept of the idea of full versus empty. And we want to be careful here not to go into too details because we still don't have a real accepted definition of full versus empty, but we are very much working on the quality of the product, making sure that we have a robust process and that we are able to reduce the impurities that we have in terms of empty or partial capsids. So I'm not planning to go into every single detail here, but this is illustrating the evolution of our platform. So basically, here, this is not really product specific. This is more illustrative, but we are working with generation of the platform. And as I mentioned, those generations are taking into account new generation of different modules and giving different outputs. And basically, we are trying to evolve this platform to make us cheaper, faster and better. So just to illustrate where we are. So generation one, this is typically a process where we infect cells -- remember that we have insect cells, right, SF+ cells. We, infect cells with 3 different baculovirus. And we are able to get to a certain yield, a certain speed. In some cases, our first-generation process are not made of one reactor only, but a series of smaller reactors, sometimes even with bags type of -- and this is the kind of third-generation process that we have. By the way, something that I want to mention is when we start the program with one generation of the process, Usually, we try to stick with this process generation to the end to make sure that we don't have a different quality of the product throughout the life cycle of the program. So we do have still some process that are close, let's say, to Generation 1. Generation 2 is typically the type of programs that we have already started now where we have a Dual Bac system. Dual Bac systems mean that instead of infecting the cells with 3 baculovirus, we have actually only 2 baculovirus. This is more -- this is optimized. We have actually better yield. And we go in typically 500-liter scale. So this is still single-use technology, fine-dealer scale. And typically here, in the second generation process, we have now optimized quite nicely the purification step to make sure that we have -- we are able to get rid of the empty or the partial capsids. Generation 3 is typically a generation where we increased the speed because we have learned a lot. So we are able to implement this generation quicker. We are able to -- not only to have a better purification process, but also a better upstream process. So the productivity is typically 50x higher at the moment, and we do generate already in the bioreactor less empty capsid, so more full capsid. Here, we illustrate 2 [ baculovirus ] instead of one, meaning that we do actually scale out the bioreactor meaning that instead of harvesting typically one reactor every 2 weeks, but we are able to harvest, let's say, one bioreactor every week, for instance. So this is working on both the process as well as [indiscernible] The future version of the platform that we are working on today. So this is future in terms of product or program, but this is actually what we are doing in process development already today, is to have actually only one baculovirus with the transition of interest, more reactor that we are able to harvest at a higher pace, very high yield per batch and very high throughput as well. So this is the way we evolve the platform to make us cheaper, faster and better. So illustrating some of the improvements that I briefly mentioned before. So we are happy in CMC to be able not to be always anymore on the critical path. Actually, we see that by optimizing the way we capitalize on our platform. Usually, the path to IND is dictated mostly by the in-life studies. So typically, the nonhuman primate studies. So this is always important that we keep our time line as compressed as possible. But when we can get out of the critical path for the overall program, this is a huge achievement for the CMC team. We are comparing ourselves typically with our colleagues that are working on with the Mammalian platform. And we are happy to say that we are actually as quick as the best of them which was initially an issue with the insect cells. We thought we were very good in terms of cost of goods and reliability, but maybe the time to generate the baculovirus that was a little bit longer. We now see that 18 months is actually very good compared to our colleagues from the Mammalian cells platform. We've mentioned here some key innovation projects that we had ongoing in order to decrease the time from lead candidates to IND. Just to mention 1 or 2, we are working using value stream mapping, LEAN projects in order to make sure that everything can be parallelized where possible. And we are also looking at new ways to improve DNA transfer in our SF+ cell lines such as we can save some time, just to mention, too. Moving to cost of goods. The big advantage of insect cells versus mammalian production system is really the fact that we don't need plasmid. So in terms of cost of goods, this is very significant in terms of raw material costs. But this is also very significant because there is just a shortage of capacity to produce plasmid. So that's a cost advantage, but this was also a time advantage. So this is a really huge internally not to have to wait for plasmid and to pay for the typical 3 plasmids that are needed in the HEK production system. If we go to the next slide, we illustrate here that when we look at the world and not only at ourselves, we see that actually most of the companies today that need a high volume of product tended to go -- so high volume of product. What does it mean? It means that if you look at the dose that you need per kilogram or per patient, and you multiply that with the prevalence of the number of people that you need to treat basically, it gives you an idea of how much volume you need to be produced. And so far, we wanted to check against or in the world who was using insect cells or HEK mammalian cells. And we see that insect cells for now has the potential to really produce high volume. And this is what some of our colleagues in other companies have done as well. So the HEK system here, a direct HEK system, typically still is used for smaller volume. This is not completely black and white. But this is giving us a good impression that we are go into the right direction if we need enough value. This slide is actually illustrative mostly. So don't look at the exact number, but this is illustrative of what we do in order to be to decrease actually the cost of goods or to increase the yield per bioreactor run. So typically here a little bit like what we've done with the car industry, we are looking at where can we learn? Where can we accelerate our evolution or our development? The monoclonal antibody world has been able already to intensify the process moving from batch to Fed batch to continuous manufacturing. And we see that following the same type of concept, we are able to actually intensify so have improved media, adding media, continuously adding media, increasing the cell density, so improving the way we grow the sales to make sure that we are able to increase the yield. And typically here, we illustrate that in relatively short time moving from typically Generation 1, 2, 3, we are able to increase the yield per reactor by relatively easily 10x. So far, I mentioned mostly our platform as our commercial platform. What is important is actually our platform is not only at commercial scale, so the 500-liter scale, but we also have scaled down model. So we have scaled down model from 250 ml scale. This is what we mean here by the AMBR 250 high-throughput system, through it 2-liter stirred tank reactor, 50-liter stirred tank reactor and then the 500-liter. And I will show you a bit later that we have actually to scale on both our sites, both in Lexington near Boston as well as in Amsterdam. And by having those different scale and by having those scale qualified or comparable between each other. We can actually play with those scales and typically go very quickly, for instance, for tox study to the commercial scale, so the 500-liter scale. So this is what we have illustrated here with cGMP/tox eventually come back to optimization or robustness study at much smaller scale and go back again to multiple lines of 500-liter if we need more product. So we are able to play with those different scale using the same -- again, the same module, the same unit of operation at different scale, making sure that we don't need to optimize or to do a lot of runs at 500-liter because we have a scale at 250 or to do 2-liter run for early research and optimization right at the beginning, but which is a good scale down model for what is going to happen later on during the life cycle of the program. Of course, the more you move to the right, the lower the cost of goods, so that you want to move as quickly as possible to the right-hand side for clinical trial production or commercial production. This slide is illustrating what I mentioned before about the empty, full and partial capsids. So again, here, we are illustrating the purification step optimization. But actually, what is important to understand is with our second and third generation process we already produce less empty particles and partial particles in the bioreactor. And then even if we do have a low level of empties and partial capsid, we are even further able to discriminate between the different peaks during the purification steps, chromatographic step, such as we can only keep the peak where we have only full particles. I don't want to go too much into details in terms of numbers here because everybody will have different numbers. Again, nobody has the same definition of what empty, full and partial are. But the important point is we minimize the empty particle that we generate at upstream level, and we optimize the full particle that we have by having optimized purification process. So implementation. This is just illustrating briefly how we are moving from -- for each of the module. We have projects to work on modules and unit of operation. And if you click once or twice, please, you will show that moving from, let's say, a product 1 to product 2. We are moving to the right-hand side. Each time that a module is ready for implementation, we can take the product that Ricardo is developing in research and developing the process with the program, which is -- or the generation of the platform, which is ready. So for each of the programs that we use, we are able to move further to the right, making sure that we are actually making our program platform more efficient. Briefly on capacity. I mentioned that we have 2 sites: Lexington, on the left-hand side and Amsterdam on the right-hand side. We now have 3 manufacturing suites for drug substance. The red box here is the one the suite 1 that we use for HEMGENIX and hemophilia B that is today dedicated to hemophilia B. We have a second line where we produce the other clinical trial material that we currently have in our pipeline. And in Lexington in -- sorry, in Amsterdam, we have a third line that we will use for other clinical trial program that we are finalizing to qualify at the moment. In terms of drug product filling, we have capacity or capability and capacity in Lexington, GMP, and we're actually investing into a second line, that's why this is inducted line. And then we have a small-scale drug product line also in Amsterdam that we are starting up. So to wrap up, the platform today, our platform concept is really allowing us to build on the capabilities that we are developing program after program. And this makes us quicker, faster to the market, cheaper in terms of cost of goods and better in terms of quality of the product and reduction of impurities. Thank you very much. And with this, I'm happy to open for questions. Maria?

Thank you, Pierre. And we do have a few questions that have recently come in regarding your presentation, so I'll go to those now.

The first one comes from Robyn Karnauskas with Truist. Are there any specific limitations with insect cells? Why isn't it more broadly utilized?

Yes, that's an excellent question. When we look at our own data, we are wondering why not everybody else is using insect's cells platform. Of course, we know it well. We have developed it for a long time. I guess the benefit of HEK is this is relatively easy to do small experiment quickly. So if you want to generate early AAVs, small volume of -- this is what a lab is for a small lab, this is easy to do. If you want to bring it to make large volume and to go to commercial, something not that many companies have done so far, we believe that our insect cells platform still has a lot of capability. So probably historic -- history and know-how is one aspect. And the other aspect is the capability to do quickly small-scale batches easily, where the HEK cells platform is actually quite easy to use.

Yes. Maybe just a couple of other things, Robyn. Actually insect cell baculoviral technology has been used for a really long time. It's used prevalently in vaccines, even at 10,000-liter capacity. So it's really a well-known expression system. I do think, as Pierre mentioned, that HEK293 is no doubt prevalent because of its ease of use, it's small scale in a lot of academic centers. So -- so that, in my view, is perhaps why we've seen it more prevalently and the work being done in gene therapy.

Another question came in from Suji Jeong with Jefferies. Is there any transduction efficacy difference between AAV produced by insect versus HEK cells?

We've generated quite a bit of data around that. And the data that we have generated do not see -- do not show differences of transduction. Again, I'm not pretending here that we have tested every single program in every single cell line. But all the data that we have, and we do that regularly to make sure that we are not on the wrong way or the wrong path, right? All the data that we have do not show significant differences.

The next question is from Judah Frommer with Credit Suisse. When you look at clinical stage gene therapy companies that are building out their own manufacturing, where do you believe they are versus uniQure in the efficiency of their platform? So Pierre, you addressed that a little earlier in your presentation. Maybe if you want to touch on that again? And then there's just another follow-up on that. Eventually, do you believe that manufacturers will be pushed to lower COGS in order to bring down the cost of therapy?

Yes, that's an excellent question. It is always a bit difficult to know exactly where we are compared to our colleagues, right? I must say here, we are moving from science to technology, but also to go a lot of capabilities. I mean, at some point, manufacturing is not only about producing. This is also about testing. This is about releasing. This is about releasing batch after batch. This is also about being able to release batch being fully GMP being able to close the deviation, which happen in a short amount of time. So I don't think we should reduce this debate of, okay, how many AAV we can produce per batch, although this is important, but also what is the success rate? What is -- the ability to do that in a fully CGMP manner? And we believe that with the recent success that we have, we are actually quite well positioned in the industry. Now briefly about the cost of goods, yes, this is important to be able to lower the cost of goods. We believe that with our insect cells platform, we have an edge here in terms of cost of goods. A big part of the cost is still on the fixed cost. So being able to better use our capacity will lower the cost of goods and being able to use our platform and not having to redevelop everything from scratch, be it in research or in development, we'll be able to reduce not only the cost of goods during production but also the cost of research and the cost of development, which is still actually the biggest part of the cost of a product.

The one thing I'll also add is, if you believe like we do that the future of genetic therapies is moving to more prevalent diseases. There's no doubt that there's going to be more pricing pressure on these therapies irrespective of how powerful and transformative they are -- and so we do think that cost of goods for being able to address some of those more prevalent diseases is going to be critically important. And we do really believe in our platform and our ability to cost effectively address those larger market opportunities.

Excellent. Thank you, Matt and Pierre. We do have just a few minutes left, and we did have some questions that came in regarding our program in TLE and AMT-260. So just for a couple of minutes, I figured that maybe we could bring our whole group back and just answer these last questions. The first 1 comes in from our analysts with [ Kempen ]. Can you again repeat the time lines for the Phase I/II study start and the data, at which point in the study would you be able to release data after cohort 1 already or after both cohorts? And then another question about how long the stagger period may be when you have the DSMB meetings?

Yes. So happy to take that question. So we plan to start the study in 2023. The first cut point, in our view, is after we have 3 patients dosed. So after the second DSMB review of the safety data, and that gives us the opportunity to recruit additional 3 patients into the study. And this is also the time point when we plan to initiate the second cohort. So that will be somewhere in 2024. The second part of the question was on the scheduling. The initial DSMB scheduling foresees that we -- as we focus on safety and tolerability that we look for the first 14 days on the safety, then bring this to the review of the DSMB and then include the next patient into this clinical study. So there is a not 3-month waiting period until we have the next patient dose. It's more in the trend of a few weeks because the DSMB meeting needs to be scheduled. So we have 14 days data, then we schedule the DSMB meeting, and then we have the next patient group recruited into the first cohort.

Okay. Very good. And -- and again, a question about expecting to see a reduction in seizure count at what time, what would, again, the reduction be for it to be meaningful in terms of really looking at end points and on early signs of efficacy?

So for our clinical study, we are aiming to see a seizure reduction of approximately 50%. That is what we are targeting currently as a first sign for efficacy. The hope is that we get more seizure reduction, but this is where we have calculated our sample size towards.

Yes. Let me just add that we -- I think from a -- of course, the ambition of all gene therapies and particularly ours is to render patients as seizure free as possible because I think that's what's truly transformative. And that's where we're going. But of course, we go in stages. And as we ramp up the concentration of the drug and we optimize the administration, we're looking for just signs of efficacy first.

If I can just chime in, too. I mean, the goal is always to try to get seizure freedom. And -- but as was mentioned, this is still early on. But 50% seizure reduction rate, that's in line with FDA approval for meds. It's sort of similar to a med coming through. So it's -- that's considered significant. It's a clinically meaningful reduction if you can get that much, but agreed, there is the possibility for more. We always want to see more. But that's a very common sort starting line.

Excellent. Well, thank you so very much for the additional time to answer those questions that came in. And at this point now, we're at that point in the hour where I'm going to transition back to Matt Kapusta, our CEO, for some closing remarks. Thank you to all of our presenters.

Thanks, Maria, and a sincere thanks to all of you who tuned in for today's presentation. I want to thank all of our presenters today and particularly Dr. Bubrick for all her contributions. We hope after today's presentation, you have a greater appreciation for the significant opportunity in temporal lobe epilepsy as well as our platform and manufacturing capabilities. We at uniQure remain very excited about the future. We look forward to providing you further updates in the new year. Thanks very much.