modulation for epilepsy. My name is TJ, and I'll be the facilitator for the presentation today. Before we get started, I would like to take a moment to acquaint you with a few features of this web event technology. At any time, you may adjust your audio using any computer volume settings that you may have. On the right-hand side of your screen, you will see the Q&A window. There's a large window which holds all your sent messages and a smaller text box at the bottom where you will type in your questions. To send a question, click in the text box and type your text. When finished, click the Send button. All questions that you submit are only seen by today's presenters. Your questions will be answered at the end of the webinar. Please note that this webinar does offer continuing education credit after completing the evaluation. You will be taken to the evaluation immediately after the webinar. You will receive an email when the on-demand version is available. Now I'd like to introduce today's moderator, Dr. Chen Wu, AES Online Education Committee member. And now I'd like to turn it over to Dr. Wu to introduce today's presenters. All right, thank you, TJ. Welcome, everybody, to our Ask the Expert webinar focused on new paradigms in intracranial neuromodulation for epilepsy. Again, my name is Chen Wu, and I'll be moderating today's webinar. I'm an epilepsy neurosurgeon at the Vicki and Jack Farber Institute for Neuroscience at Thomas Jefferson University. And I have a particular interest in neuromodulation, especially its applications in epilepsy. In terms of disclosures for myself, I am a consultant for a few of the neuromodulation manufacturers, NeuroPace, Medtronic, Boston Scientific, and Abbott. None of those will be relevant in my role of moderating this session. To review the learning objectives of this webinar, we aim to, one, review the common indications and applications of intracranial neuromodulation for epilepsy, two, evaluate the different thalamic targets, such as the anterior nucleus and centromedian nucleus for intracranial neuromodulation, and three, discuss future applications of intracranial neuromodulation and how these therapies fit into treatment paradigms for epilepsy. We really want this to be an interactive session. So during the webinar, as mentioned, please feel free to enter your questions in the Q&A box. We're going to go through both presentations first. And after the completion of both presentations, we'll move on to the Q&A portion of the webinar, where I will present your questions to the two experts and have a conversation in that manner. We plan about 15 minutes for this last interactive portion, and we'll do our best to respond to as many questions as possible during that time. So with that, I'm excited to introduce our two expert speakers today. We are joined by Patricia Dugan and Dr. Jamie Van Gumpel. Dr. Dugan is an Associate Professor of Neurology at NYU Grossman School of Medicine. She's the Director of the Epilepsy Fellowship Program, Director of the Epilepsy Special Procedures Service, and Co-Director of the NYU EEG Laboratory, and will be speaking today about newer applications of responsive neurostimulation. Dr. Van Gumpel is a Professor of Neurosurgery and Otolaryngology, specializing in, among other things, epilepsy surgery at the Mayo Clinic in Rochester, Minnesota. And he'll be talking about different DBS targets in epilepsy as well. So without further ado, I'll turn the presentation over to Dr. Dugan. Thanks so much, Dr. Wu. So I'm delighted to join everyone this afternoon to talk about new paradigms in intracranial neuromodulation for epilepsy, and I will be covering the RNS system. So these are my relevant disclosures. I receive research funding from the NIH, Neuropace. I also do receive honoraria for speaking engagements from Neuropace. We will be using some industry graphics, and we will have a brief discussion of off-label use of RNS, but I will explain where we are doing. So I also feel obligated to point out I am not a neurosurgeon, but we'll be talking about some neurosurgical procedures. So since we are going to be covering new paradigms or new applications of the RNS system, I will be covering refective surgery combined with the RNS system, RNS to target regional neocortical areas, and utilizing corticothalamic stimulation with the RNS system, particularly treating regional neocortical epilepsy, multifocal epilepsy, generalized epilepsy, and we will also discuss some ongoing studies. So just briefly to review the RNS system for those who may not be particularly familiar with this surgical intervention, this is a therapy that makes use of open-loop stimulation through continuous monitoring of brain activity via ambulatory electrocorticography. So when electrographic signals have been identified as either preictal or ictal, stimulation is delivered in real time to abort seizure activity. So the RNS system also records long-term electrocorticographic, or ECOG, data that gives us the ability to review this data remotely, and it provides insight into long-term response to RNS therapy, medical management, various lifestyle factors, and it also gives us the limited ability to characterize clinical events electrocorticographically. So this is a therapy that is widely available at most comprehensive epilepsy centers in the United States, and it is currently approved for adjunctive therapy in patients with drug-resistant focal epilepsy who are 18 years or older. So this is a current treatment algorithm that we consider when we are thinking about surgical intervention for our patients with drug-resistant focal epilepsy. So remember that if a curative surgical intervention is applicable, this should always be prioritized. But if not, for example, if a patient has a single unresectable seizure focus, you might want to consider an intervention like RNS. The barrier to curative resection or ablation is often the overlap with functional cortex. Perhaps the seizure focus is within a high-risk area for ablation or resection, maybe due to its deep-seated location. Perhaps it is a highly vascular focus, or perhaps the size or extent of the seizure focus simply precludes resection. And that ablation or resection might result in significant cognitive impairment. So for patients who have two seizure foci, they may be eligible for RNS placement. For example, if they have bilateral foci and homologous structures, like the classic case of bilateral temporal lobe epilepsy, or for example, they may have two ipsilateral foci with dual pathologies. So to start, let's discuss RNS in combination with resective surgery. As I mentioned, resective surgery is going to be obviously limited if you're dealing with two foci and homologous brain regions, or if the seizure focus or foci involve eloquent brain regions. So RNS can be employed to augment resective or ablative management. For example, in this paper written by Brandy Ma and Vikram Rao, it very nicely describes the use of RNS to target residual epileptogenic cortex following resection in one of two foci. So in this graphic, this example shows dual pathology within the right hemisphere with a seizure focus in the temporal lobe and a seizure focus in primary motor cortex. So the seizure focus in the temporal lobe was addressed with a right entero-medial temporal lobectomy and the remaining focus in the precentral gyrus, or primary motor cortex, is then treated with RNS. And another application of RNS to supplement surgical decision-making is using the long-term ambulatory ECOG data that is obtained with the device to inform seizure burden. So there was a wonderful paper that was published in 2015 that was led by David King-Stevens that showed that in patients with bilateral temporal lobe epilepsy with bilateral temporal RNS leads, it took about an average of 42 days to produce the first contralateral mesiotemporal lobe seizure. So that highlights the fact that the seizure frequency that we typically observe in inpatient phase two or intrusive monitoring, which is oftentimes one to two weeks in most centers, is usually insufficient in terms of capturing the lateralization of seizure burden from each hemisphere. Larry Hirsch then led a multicenter study of 24 patients with bilateral mesiotemporal RNS leads and tracked their progression to mesiotemporal resection. So in this study, there were 24 patients who met inclusion criteria. Nine of 24 were subsequently found to have exclusively unilateral mesiotemporal lobe seizures. 15 of these 24 had bilateral seizures and 13 of these 15 had greater than 90% coming from one side. And in all cases, surgery was performed on the side where majority of the electrographic seizures arose from. In terms of outcomes, 22 of the 24 had entero-mesiotemporal lobectomies and two had selective amygdalo-hippocampectomies. And the median reduction in disabling seizures at the last follow-up was 100%. 23 of these 24 patients continued to be treated with the RNS system post-resection. 17 were bilateral, five were unilateral, and were contralateral to the side of resection. So of the 15 patients with known bilateral mesiotemporal lobe seizures, post-resection, 13 continued to be monitored bilaterally with RNS, and all continued to have seizures recorded after the mesiotemporal resection, whether it was lateral to the side of the resection, contralateral, or bilateral. But majority of these were clinically seizure-free and the remainder showed greater than 50% reduction in seizure frequency. A group at UC Irvine published their experience combining surgical resection and RNS therapy in patients with multifocal epilepsy. In 10 patients, of these 10, five had resection and ipsilateral RNS placement, while five had resection and contralateral RNS placement. This group observed an average of greater than 80% reduction in seizures at one-year follow-up, and four patients were seizure-free at one year. So the group concluded that this approach of partial or subtotal resection of a multifocal seizure network plus RNS placement could produce worthwhile seizure reduction. So this is an abbreviated table that illustrates the localization of the 10 patients' ictal onset zones and resection sites. It's fairly busy. I apologize for the small font. You'll note that this cohort was composed mostly of patients with variations of temporal, frontal, temporal lobe epilepsy. And the graphic on the right side shows above, there is a lateral view of a left anterior mesial temporal lobectomy with placement of RNS leads in the posterior temporal lobe, posterior to the resection margin. And below that, there's a coronal view of a right anterior mesial temporal lobectomy with RNS placement in the contralateral left mesial temporal lobe. Now let's talk about laser ablation or LIT. Now let's talk about laser ablation or LIT. Laser interstitial thermal therapy or laser ablation is something that has been steadily gaining traction in the treatment of drug-resistant focal epilepsy. So the Stanford paper describes the first published case of laser ablation performed with the RNS neurostimulator or implantable pulse generator, or what I'll refer to as an IPG, and the leads remaining in place during the procedure. So they described the case of a woman with refractory temporal lobe epilepsy who had RNS, the RNS system in the temporal lobe because she initially declined resection. So despite this, she continued to have seizures and subsequently underwent an anterior mesial temporal lobectomy with continued RNS monitoring, posterior to the resection margin, and continued to have frequent focal-aware and focal-unaware seizures. So she subsequently underwent laser ablation of the residual amygdala and piriform cortex with the RNS leads and IPG in place during this laser ablation procedure. So the group concluded that this was safe, but that if this were to be considered, perhaps consider implanting the IPG or neurostimulator contralateral to the side of anticipated ablation in order to minimize MRI artifact during the LIT procedure itself. The group at the University of Wisconsin then described another case of combining RNS and laser ablation in a woman with bilateral temporal lobe epilepsy. In her case, the RNS system showed almost complete lateralization of seizures to the right amygdala and hippocampus. Her IPG was implanted on the right side of the head. So during the procedure, the IPG was removed, but the ferrule or the tray that the IPG sits on was left in place. The right RNS lead in the temporal lobe was removed, but prior to that, the location was used to map the trajectory of the laser ablation for targeting. And despite having the right-sided ferrule in place and the left-sided temporal RNS leads in place, the team did not observe significant artifact or interference during the procedure. So they went on to perform a standard right mesial temporal lobe ablation and then replace the right RNS lead in the hippocampal tail for continued monitoring. So the graphic shows at the top panel representative laser ablation predictions in subsequent ablated regions in the right temporal lobe. And then you see the RNS ECOG tracings below with the top tracing taken the first day post-operatively. And you can see that the tracing from the right posterior hippocampus is very slow and attenuated. And then the ECOG below that eight weeks post-operatively then shows recovery of the tracing in the right hippocampus with a nice continuous theta rhythm and also notably the absence of any interictal epileptiform discharges. So RNS can also be used to treat large neocortical seizure foci or what we refer to as regional neocortical epilepsy. So really extensive seizure foci are often not amenable to surgical resection simply given their size and the likelihood that they overlap or very closely abut functional cortex which would put them at risk for permanent neurological deficits. So a regional neocortical seizure focus would be defined by a seizure focus that spans more than five electrode contacts or greater than four centimeters. So in this case series of 30 patients with regional neocortical epilepsy, RNS leads were placed four centimeters or farther apart flanking the identified borders or margins of the seizure onset zone. And in this series, the median seizure reduction rate was over 75% and 20% of subjects had 95% or greater reduction in their seizure burden. Now corticothalamic stimulation has really captivated the epilepsy world in the last few years. And we can thank the SANTE trial for DBS that really paved the way for this. But the epilepsy community has moved beyond the anterior nucleus of the thalamus and there's been a lot of exploration into stimulation of the centromedian nucleus of the thalamus as well as in the pulmonary. So this comes at a time where we are challenging the conceptual framework of the conventional and rather concrete understanding of epilepsy and anatomic seizure foci to more nebulous concept of seizure networks and treating epilepsy by removing nodes in the seizure network or by modulating the seizure network as a whole. So the premise for targeting thalamic nuclei is that doing so should enable us to treat spatially extensive seizure networks by modulating the extensive thalamic cortical connectivity. So let's review the anatomy and connectivity of these thalamic nuclei. The anterior nucleus is highly connected to the limbic system, the hippocampus. The centromedian nucleus has widespread cortical connections including the frontal cortex and the pulmonary has strong connectivity with the posterior quadrant, particularly visual cortex. So one of the earliest case series that was published that describes RNS in the thalamus targeted the anterior nucleus in three patients with multifocal epilepsy and that was done here at NYU by my colleagues and myself. As we really didn't know what to expect when it came to thalamic seizure detection, each patient in this series had a depth electrode in the anterior nucleus of the thalamus and an ipsilateral cortical strip for seizure detection. The location of the cortical strip was chosen based on the brain region that we felt that would have the most robust involvement in seizure onset. So in this series, it was the parietal cortex, postcentral gyrus and the middle temporal gyrus and the resulting ECOG that we recorded showed spread of seizures from the cortex to the anterior nucleus of the thalamus as well as seizures that were recorded independently within the anterior nucleus without a corresponding cortical signal. So aside from demonstrating that thalamic targeting appeared to be safe and well-tolerated, all three of our patients demonstrated 50% or greater reduction in seizures at their 30-month follow-up. So the top panel of this graphic shows the localization of thalamic depth electrodes and the ECOG below shows a seizure with very robust signal in the thalamic depth and that actually preceded the electrographic seizure activity that was observed in the temporal strip. Dave Burgett's group then at Spectrum then described three patients with posterior quadrant neocortical epilepsy who were treated with RNS in the pulmonary nucleus. Patients with posterior quadrant epilepsy are generally considered suboptimal candidates for resective or ablative surgery, usually because of the likelihood of involvement of visual cortex. So in this case series, one patient had bilateral RNS systems implanted that targeted the pulmonary nuclei and the occipital lobes bilaterally. This is a patient with bilateral occipital epilepsy. And the other two had unilateral right-sided RNS systems with targeting of the pulmonary nucleus and the occipital temporal region in one patient and the parietal lobe in the third patient. All three of these patients were responders with 50% or greater reduction in seizures and two of them actually had 90% or greater reduction in seizures. This group also described a series of seven patients with regional neocortical epilepsy, parietal, frontal, temporal, and opercular onsets with one lead in the centromedian nucleus and one lead in the most actively involved neocortical seizure focus. This study showed that electrographic seizure onsets were detected in both the centromedian nucleus as well as the neocortex in all of the patients. Although there was a tendency for the neocortical onsets to precede thalamic onsets as detected on ECOG. This graphic shows a nice coupling of the cortical and thalamic seizure onsets on the ECOG. And this cohort had a 100% responder rate with at least 50% or greater reduction in disabling seizures. Three out of the seven patients actually achieved greater than 90% seizure reduction. Now, let's talk about corticothalamic stimulation in the central median nucleus, particularly in Lennox-Gastaut patients. So, the Mount Sinai group described two patients with autism and Lennox-Gastaut, or LGS. And again, just to recall, this is a devastating epilepsy syndrome with both generalized and focal seizures, often, well, also characterized by intellectual disability or static encephalopathy. These patients had the RNS system with depth electrodes targeting the bilateral central median nuclei. And you can see examples of that lead placement in the graphic on the right-hand side, as well as the excellent ECOG recordings from the central median nucleus. Both of the patients in this study experienced 75 to 99% clinical seizure reductions over a year of follow-up. The central median nucleus has also been used to target patients with multifocal epilepsy. So, in this series, there are 23 patients, most of whom had central median nucleus targeting. There were a fraction who had depth electrodes in the anterior nucleus. You can see that it's a rather heterogeneous group with a variety of localizations, most of which were bilateral temporal, and most were syndromic in etiology. But regardless, all reported significant improvement in seizure duration, as well as seizure severity. Now, we move on to the newest frontier in RNS stimulation, which is central median targeting for idiopathic generalized epilepsy. Now, if you recall, patients with idiopathic generalized epilepsy are generally not candidates for resective or ablative therapy for, I mean, what I hope would be obvious reasons. There are no approved surgical therapies, you know, approved by the US FDA or Food and Drug Administration for patients with IgE. So, in this very interesting paper, Mark Richardson's group described four patients with drug-resistant IgE who were treated off-label with bilateral RNS leads in the central median nucleus. All four patients experienced 75 to 99% reduction in seizure burden, and all reported significant improvements in quality of life. The RNS ECoG also showed that seizures were very reliably detected in the central median nuclei, and were therefore effective in triggering cortical stimulation. So, in conclusion, the takeaway points really should be that the RNS system can be used to augment surgical resection and or laser ablation. And currently, the applications for RNS for the treatment of drug-resistant focal epilepsy really challenged our understanding of what we really consider focal when we think about focal epilepsy. Again, I think we're moving farther and farther away from our concrete concept of a seizure focus, and simply removing that will provide seizure freedom. I think we've turned more to an understanding and appreciation of seizure networks. The extensive connectivity of thalamic nuclei make this a very promising target for the treatment of regional, multifocal, and possibly also generalized epilepsy, and there are ongoing FDA-approved clinical investigations to evaluate the effectiveness of RNS in the treatment of IgE and LGS. So, thank you very much for your attention. That was a, excuse me, that was a fantastic talk, Dr. Dugan. Thank you for that. I think my talk will complement and be symbiotic with what you said, but I really learned a lot. Thank you for doing that. Thank you to Dr. Wu and the AES for the opportunity to present with a fine panel today. I was asked to cover new paradigms in intracranial neuromodulation for epilepsy, and specific what Dr. Wu had asked me to comment on are more than two leads, so four lead systems targeting two nodes within a network, and again, Dr. Dugan kind of introduced the concept of treating networks rather than focal epilepsy. I think she really tied it up nicely at the end about why we're thinking about doing that now, and how do we make decisions amongst these targets, and why might one choose these targets, and Dr. Dugan's talk also touched on that quite a bit, so we'll cover it twice, so hopefully people will understand it a little bit better. As I said before, my name is Jamie Van Gumpel. I'm a professor of neurosurgery from the Mayo Clinic in Rochester. I do have a fair amount of disclosures to cover for this. I'm a named inventor, and we have intellectual property in a company called Cadence Neuroscience, which will be launching a multinodal seizure treatment with an implantable that will be starting probably in quarter one of 2024, and that's co-owned by the Mayo Clinic, my employer. We will not discuss anything relative to that in this talk. I'm an investigator for Medtronic, EPAS, Slate. Also, we have NIH co-awarded, public-private-funded uroid research with them, and we'll be talking about some off-label DBS in this particular talk, as well as on-label ANT, stock and ownership and consulting contract with NeuroOne, which is a implantable SEG, as well as thin-film electrodes. I'm a primary investigator, as well, in the Nautilus trial, as previously mentioned in the talk, and openly enrolling at the Mayo Clinic as a site, and we're gonna talk about, again, off-label use of DBS mostly, but some RNS, for CMPF, VIM, pitainment stimulation in this talk. I wanna make a shout-out to a person that trained at the Mayo Clinic as a resident, Dr. Irving Cooper, and ultimately finished their career doing remarkable things in New York, tying us again to Dr. Dugan, who, you know, Dr. Irving Cooper originally was the first person to try to treat with implantable stimulation devices, epilepsy, attempting to treat the cerebellum, but also the anterior thalamus, and if anybody here has not read this book, this is an amazing book about surgical courage, and also, you know, what the environment of the scientific community was then, and still is, and how do you navigate that. It's a very good book. It's $5 on Amazon. I highly recommend it if you are interested in, you know, some of the history around these devices and epilepsy. ANT-DBS, as mentioned before, was a Medtronic-led trial that ultimately led to FDA approval on May 1, 2018 in the United States. However, European approval, with CE mark, occurred on September 16, 2010, and there's a fair amount of more experience in Europe with these devices. The therapy was approved as an adjunctive treatment for reducing the frequency of partial onset, both complex and simple seizures, in individuals 18 years of age or older who are refractory and drug-resistant to three or more anti-epileptic medications. The approved target was anterior nucleus thalamus, but I would argue that this has opened up a lot of the ongoing non-FDA-approved off-targeting that we're doing currently. That treatment was for focal-onset seizures, both uni and multifocal, not only housed within the PET-PES circuit, as mentioned previously, but also frontal and temporal, and their subset analysis demonstrated significant advantages over more parietal motor-onset seizures. So it's important, I guess, for thinking about future treatments, that this is mostly showing a lot of efficacy in PET-PES circuit frontal and temporal. And as mentioned previously, we've been entertaining treating more through Centromedian for a variety of reasons, but this nucleus is thought to be the gatekeeper function for the brain itself and widely progresses and is connected throughout all of the brain, as we know from fMRI studies. CMS literature regarding each of these neuromodulatory therapies, such as Tourette syndrome, Parkinson's disease, generalized seizures, as we've mentioned, intractable neuropathic pain and restoring consciousness, in fact. It's a very diverse target. We'll cover that just briefly here. In addition to Dr. Cooper's work, I think we owe a debt of gratitude to a group of a family of neurologists in Mexico City, the Velasco family, that published about this through the early to late 90s and 80s and into the 2000s, really pioneering Centromedian therapy as safe and efficient for treatment of severe epileptic seizures in otherwise intractable patients with conventional therapies. CM stimulation significantly decreased the number of generalized tonic-clonic seizures and partial motor seizures that was thrown throughout their career. And I think this would be a lot more mainstream already, again, if it wasn't performed in Mexico, unfortunately. I think sometimes we have a tendency to demean other countries' research, and I wish we really caught on to this a lot earlier. But this is some of the research we used prior to FDA approval for any of these, inciting this for off-label use, especially I owe a debt of gratitude to the Velasco family. This is something that they were doing back in the 1990s and early, again, 1980s, where they were implanting with stimulation electrodes and mapping with cortical, sorry, scalp electrodes. They were stimulated and trying to figure out what areas of the centromedian nucleus would be responsible for. And through this, they showed that the centromedian nucleus itself had some somatotopic organization in terms of the brain in which what we see was, this little area down here was mostly the parvocellular region responsible for frontal region and frontal projection. The majority of the centromedian nucleus actually was rolandic, or the magnocellular portion was to the motor areas and sensory areas, so post-frontal and early parietal. And type C areas were this anterior portion of centromedian nucleus had a lot of parietal projections. This led, some of their work led to a really outstanding trial performed in Australia with DBS, not RNS in this circumstance, for Lennox-Gastaut syndrome, in which 20 young adults were treated with the main target being centromedian, where 59% of the stimulation group had a greater than 50% reduction at the end-blinded phase, and 50% of all participants had diary-recorded seizures reduced by half, demonstrating a significant efficacy in these patients. And we see a phenomenon similar to what we saw in the ANT, DBS, and RNS trials, where after implantation, those that were in the non-stimulation phase obviously drifted back to baseline, but once stimulation in the non-blinded phase occurred, there was substantial improvement, again, targeting centromedian nucleus in these particular generalized seizures. There's also an option in these patients, so this is a report out of our group, genetic generalized epilepsy, so a single-case report of a patient that was having 48 generalized tonic-clonic seizures per month, implanted with CM-DBS, and had two over the subsequent two years, so a substantial improvement for genetic generalized epilepsy, again, not meeting that IgE category. Pulvinar, most of the evidence for pulvinar, as previously mentioned again by Dr. Dugan, is not necessarily rooted in patients. It's mostly lab work that's been done. Evidence of pulvinar stimulation from those animal studies demonstrates posterior quadrant epilepsies are probably the most effective, so lateral pulvinar probably relates mainly to the visual areas, and that's where obviously the lateral geniculate body interacts with the pulvinar region as well, and protects out to the dorsal-medial-occipital regions and parietal regions. Interestingly, the inside part of pulvinar may actually have mesial-temporal projections, so it might be another alternative target to ANT for treatment, and has temporal neocortex, cingulate, and orbital frontals, so another pepez circuit potential involvement treatment. This was mentioned previously, but I wanted to put this up here because Dr. Dugan went over this very well, better than I could, but the point is is that this is the only patient information out there with pulvinar stimulation. Our center has adopted it, and we've been doing it a lot more frequently, but it's being done a lot, and there's just not a lot of patient information about it. I just wanna mention that, that beyond its corticothalamic connectivity use right now, we don't have a lot of clinical evidence for its use at this point in time. This is my personal series of DBS implantations for epilepsy. 75 total cases of those involving ANT, 60 of those patients that had ANT, as well as some other target. ANT alone in 44 of those patients. Two lead constructs were 54 of them. However, multiple leads, so what Dr. Wu asked me to talk about today, multi-lead constructs were 21 of these patients. 11 of them were experimental IRB approved thalamic plus bilateral hippocampal stimulation with recording devices through Medtronic. Pulvinar and ANT, 1% of those patients. Thalamus plus cortex in five of those patients. ANT plus CM in four of those patients, and CM alone in 10 of those patients. In this particular group, paying attention to more than two target DBS, so four lead insertions, the median age at surgery was 32. 11 patients were female. The most common seizure location was bitemporal in this treatment population. Prior vagal nerve stimulation had been trialed in nine patients or 56% of the patients. Invasive EEG recordings was performed in seven. Median surgical time for implantation of four leads was 5.03, so five hours and three minutes for the total median operative time was 7.56, and that obviously came down with our surgical experience over time. Two patients had transient neurological deficits. One patient had a trace intraventricular hemorrhage. Nobody had permanent deficits with these in particular implants. Three patients required intraoperative repositions of a single lead. Obviously, the more leads we put in, the more likely our third and fourth lead will be mistargeted, so those were to accommodate for that. No infections in this patient group. 12% at last follow-up reported angle class one for six months or greater, 6% at angle class two, 44% at angle class three, and 38% at angle class four. Obviously, this isn't the best way to report this. Approximately 70% of patients had 50% or more of reduction in seizures, but this is what we have right now. In fact, we're talking about the need for a different scale for this prior to logging on for this. Median follow-up was 37 months, so quite a long follow-up for this cohort, and one patient out of all those patients elected to have their system explanted. This is data that was submitted to the AES by Dr. Pierce Peters, one of my research fellows. Give you a quick example of an implant. So this is a patient with frequent falling and injuries, born with developmental delay and ultimately recognized to have seizures at age 15. She presented to us at 29. She had three different types of seizures, generalized tonic-clonic seizures, and these were frequent and difficult to control. Second, absence type, where she had staring and unresponsiveness up to eight to 10 seconds. And the most worrisome were episodes of slow neck flexion and falling. These began in four years prior to us seeing her. They were very frequent. They're difficult to control, and she had had multiple injuries secondary to these. Initial EEG demonstrated 1.5 to 2.5 Hertz spike wave discharges, maximum in the frontal region. Separate EEGs had revealed independent left and right central parietal discharges and often bilateral synchrony, right frontal temporal spike and wave discharges in this patient. And the interpretation was primary generalized form of epilepsy based on generalized indirectal and ictal discharges that were severe and resistant to medications and surgeries. Now, these are the patients that we've implanted over years. And I'll just mention that also in this patient, because of the injuries with falling, we discussed sequenced corpus callosotomy with the family. This was refused, this recommendation. Other facilities, in addition to ours, due to the comorbidities associated with these. She was re-offered, of course, with this in 2014. Lit callosotomy really wasn't an option, although that's become more of an option now. And due to the severity, we offered bilateral ANT and CM stimulation in this patient. I'll just mention prior to FDA approval, we discussed off-label use all the time with the patients and had a philosophy that in severe cases, since we were doing this off-label or essentially eating the cost for the implant, we would oftentimes cover more than one target. In this case, obviously anterior nucleus and CM. And the reason for that was that the components that we would use were spine stimulation hardware, and it was much cheaper than the implants that we would use for typical direct brain implants. And we thought this was more efficacious with the only limitation being that we couldn't get up to 210 Hertz frequency. We were limited to 110 Hertz stimulation frequencies. And the philosophy behind this was my own religion has been to do all the good I could for my fellow man and as little as harm as possible. And I think that is very apropos for this patient. They were looking to avoid comorbidities in this particular case. This was prior to our temporal parietal approach we've been describing for ANT. So these are both bifrontal implants for centromedian and ANT. And this patient went on to have two battery replacements and her batteries were consumed quite quickly over the course of 18 to 24 months. At the end of battery wear down, she would have a flurry of seizures. And then ultimately a rechargeable battery was put in that had seemingly taken care of that. Prior to this talk, she was doing well and had been seizure free for nearly three years. No drop attacks. She was tolerating stimulation well. No changes had been made to her parameters since approximately five years prior to this and had been able to successfully reduce some of her anti-seizure medicines without any breakthrough seizures. This is her programming. I'm not gonna go over that extensively given the time that we have allotted. I'll just say we're also doing some things now with patients with motor seizures. And in fact, including VIM and some of our constructs, of course, off FDA label, no real evidence beyond the fact that the motor related seizures come through this area. So targeting lesions, you know, for ANT-DBS, a recommendation is for focal, uni, multi onset that frontal temporal or papaz circuits, including insular and singular are targeted from and included with ANT-DBS. CM-DBS is used for motor onset seizures, linectis stoen generalized, and genetic as well. And pulvinar, posterior quadrant, and supplemental temporal. It's important when we're putting together these constructs to think about portability. Obviously, RNS and regular two-lead DBS are portable between institutions, and doing these multi-lead constructs does wed the patient to your institution coming back. So if transfer or travel is an issue, obviously this might move you away from some of these treatments. I just wanna again mention in the past that we targeted more than one site off FDA approval as the cost of the device was cheaper, and institution was paying for the procedure as we thought it was the right thing to do. Now, I think we need to think a lot more about this in our patients and the needs of the patients and what we consider risk versus benefit is very important for this. And I'm mentioning this because I see a lot of people reporting multi-lead constructs with very little evidence for them. And I think that we need to think strongly about adding these, although it appears to be safe. I think there's a lot of unanswered questions in these patients. For instance, can personalization of DBS in terms of connectivity between a seizure onset zone and a particular thalamic sub-nuclei improve outcome? We don't know that. Can personalization according to spectral or synchrony characteristics of EEG during seizures better reduce seizures? And in fact, can SEG implants predict this if we're using thalamic SEG? Does it predict these types of arrangements? And will one target be better than another? We don't know that. And will targets within one system lead to synergy? We just don't know that at this point in time. But I think those are some of the real things that we try to grapple with in our institution. So in closing, I wanna say that a team does this. It's not just me. This is my OR team, but obviously we work extensively with our epileptologists in making these decisions. We have a neuromodulation group that's dedicated to that. So four neuroepilepsy doctors that spend time thinking about this with myself and my partner, Dr. Kai Miller. But even in the operating room, these big cases require a big team. And I'm so thankful to have the expertise that we have that exists at the Mayo Clinic and be able to do these things with my friends and family here. I think that's all I have. Thank you for your time. Just as a quick reminder, text chat is located on the right-hand side of your screen. To submit a question, type your question in the small text box at the bottom. And when finished, click the send button. And now I'll turn it back to Dr. Wu. Thank you. And thank you, Dr. Van Koppel, Dr. Dugan. Great talks on both topics here. So we do have a few questions that I'll kind of group together. The first one goes to Dr. Dugan, but I suppose it could be for both. It says, what are your thoughts using VNS as an augmented therapy to some oral treatment you touched on? Yeah, so I guess I'll take that first, right? So there are multiple case series, actually, that have been published looking at the concomitant use of VNS and RNS. Particularly, these are studies done by David Spencer, Leah Ernst, and Danielle Becker's groups, as well as, I think, the group over at Westchester, really looking at people who have both. And there's a suggestion that the two therapies together augment one another. There was an interesting case report of a patient who's had one seizure focus that was more beneficially treated by the VNS device and another seizure focus that appeared to be preferentially treated by the RNS system. Generally, we counsel our patients that having the VNS device implanted prior to having the RNS device is not a deal breaker. It doesn't preclude RNS placement. But I think there is still a lot of study that needs to be done, really investigating the exact mechanism as to how these two neurostimulation therapies seem to augment each other or complement one another. I'll add that Dr. Dugan's excellent commentary that, you know, so we published about this 2020 in the Red Journal in patients that had combined devices. Because as you know, for the EPAS trial, they're telling people they have to have their VNS device removed. And I've made the recommendation for DBS that it should always go on the right side to open up the therapy for VNS. And also, by the way, in the SANTES trial, most of the patients that had ANT-DBS inserted were failures for VNS, right? So we know that the efficacy does not get impacted by the failure of VNS. And I will tell you, I have patients that can still magnet out of seizures when they have them and still, you know, have their VNS on. And they use their ANT to reduce their overall seizure burden. And I would recommend to everybody listening that, you know, these devices work independently. We know fMRI data from them. They have different areas of projection. We don't have any clue still how VNS works. VNS is getting better also. We didn't talk about that with the micro-STIM, but there's so many reasons to believe in that as a therapy too. So I think that in terms of symbiosis though, that paper that we put out that had a fair number of patients with ANT and VNS, we did not see symbiosis with those two treatment types. Great. So the next question comes actually from a few of the attendees. And it's one that I was gonna ask so we're all kind of thinking along the same lines is that, you know, now that we're talking about thalamic stimulation, both on the responsive neurostimulation as well as the deep brain stimulation standpoint, thoughts or of, you had mentioned it, Dr. Van Gompel, of SEG, of the thalamus. One of the attendees asked, is this standard part of your SEG implantation? How do you interpret this information? You know, is this something that we should be moving towards? So I'll go to Dr. Van Gompel first. This is really controversial and we could probably spend the rest of the time talking about this. So we had been implanting thalamic SEG since I think about 2015. And we have not published too much on it because the initial stuff that we did was just extending leads that currently exist under IRB into the thalamus to randomly map it. Okay. Now, with my partner, Dr. Miller, joining the practice, we've very commonly now said, well, we should implant the ANT because it will predict what we're gonna do later on, which I don't know if that's true. And we haven't answered that question. We do do it. We do use it as Dr. Dugan showed. And I love that paper where they did the unilateral ANT implant with the cortical electrodes and RNS that obviously the seizure detection is logarithmically different than in the cortex, but it does have a lot of usefulness in terms of seizure capture. And I think it is useful, but I don't know how to recommend it as a routine part of practice. I'm here curious to see what you guys think because I think there's so much to learn from it yet. I do know it's safe. Dr. Dugan? Yeah, so I agree with Dr. Van Gompel that is very controversial. I know that, so in my institution, so at NYU, we don't routinely sample any of the thalamic nuclei in a standard stereo EEG investigation. I know that there are some centers that do do that routinely. It's not part of our current practice here. In terms of, I think there was another question also, or perhaps I'm jumping ahead, is it necessary to sample these thalamic nuclei when we are considering targeting these for RNS? I don't think that it's necessary. I don't think that, we don't have enough, as Dr. Van Gompel said, we don't have enough published data looking at that. We do have some of these publications that in terms of like proof of principle, they do show that there is, particularly in the pulmonary and in the CMT, there is significant correlation between cortical seizure capture as well as within the centromedian nucleus in particular. We didn't find a great one-to-one correlation in the anterior nucleus of the thalamus at NYU in our own case series. It's a long answer for basically in short, I don't think that it's necessary to sample that a priori when you are considering targeting these nuclei for stimulation. Great, and then, so the other common theme on the question here is probably because of the- Chen, can I ask, Chen, can I ask for your impressions of the last question I asked you? Yeah, you won't let me escape. You can't get out again. No, no, so we are in the process, we're submitting an IRB to do it because we feel like it might be useful, but that if we are going to try to investigate it, we should do it in a, you know, a very formal prospective way with the idea of answering a couple of things of, you know, should we be implanting one or two of the nuclei, so for thalamic, should we maybe perhaps implant both ANT and pulmonary, especially if there's more posterior thalamic involvement or to see is there one better than another? The other question that we had is that would, just like we do for mapping, would any stimulation during the implant change any of the cortical patterns to see if that would help predict anything, almost, you know, thinking of it like we do with other forms of neuromodulation in the trial towards stimulation. So we have a couple of questions that, you know, we want to investigate as well. So, and it's certainly a topic that I talked to a lot of folks about at the last AES meeting, so. So moving on, you know, because of the overlap in talks, there's certainly the question of, well, if you have a generalized epilepsy, what makes us choose RNS versus VBS of central median? Or is it simply a practice preference or is there a particular algorithm? Do you want to go first? I can take that. I'll start, I'll start. Go ahead, go ahead. I'm sure Jamie has some feelings about that. So I think that, you know, that's an excellent question. And that is a very relevant question because that is something that we cover in our surgical conference time and time again. If we're going to be recommending central median nucleus stimulation, it's, well, should it be DBS or should it be RNS? And so again, that's, you know, that's another conversation about, well, do we need open loop or is this closed loop stimulation? You know, there are certain benefits to both. There are drawbacks to both. So a large part of the decision-making is going to be, you know, what are your goals here? Do you need to have the, would this patient in particular benefit from long-term electrocorticography, for example, that would come with RNS selection? Is there, you know, is there a need to quantify seizures? Is there a need to attempt to characterize some clinical events? I mean, again, there is a limited ability to do that with the, you know, ambulatory ECOG that you get with RNS. There are certain patients who may certainly benefit from having open loop stimulation. So I think those are generally the questions that, I mean, those are the things that it gets, the conversation really boils down to those two things. Does the patient definitely need open loop versus closed loop? And then again, is there some benefit to having the long-term ECOG that you get with RNS? Maybe I'll talk about battery life and stuff, but. Yeah, that's why this is awesome to have a neurologist and a neurosurgeon here because we all think about different things. But I think you said it, you know, in the end, the goal, I think we're going to show through Nautilus for IGE that it's going to be effective. That's what the data is probably going to show. We don't know that for sure yet, but all the accumulated literature evidence says it's probably going to be a really, a clear path to being effective. So eventually there's going to be an FDA approved target for it with a device that may wet us to one. However, if the goal is just stimulating CM, it may get down to a variety of other issues, right? So with Neuropace and RNS, we're going to have developed biomarkers over time and it might not just be spiking or seizure detection. And that's going to be a huge asset with the downside of having to replace the battery all the time, because that's going to be one thing that people are going to worry about with that device. It probably is going to stimulate continuously. I don't know how long the batteries are going to last. And as you know, revising that scalp room over time is problematic and I don't know how the FDA is going to view rechargeable devices there. Now for DBS, which has been our preference up until probably five or six years ago, the device you can put in is rechargeable and it's like an iPhone. They can recharge it until they get sick of recharging it, but it's probably going to last 10 to 15 years. It doesn't have that mass up there. There's some advantages, some disadvantages to it, but you don't get biomarker collection, right? And that's, I think that's going to be an issue and a limitation for a long-term. I think right now you also have to think about who the end user is. So if I was putting one in for Dr. Dugan, it sounds like I should really put in an RNS or I should call her. And if I was putting it in for a group that has, you know, a big DBS group and that's what they support mostly, I would probably tell them and say, what's your preference? I am just a technician in a lot of this. So I always think about the end user and what their comfort levels are and how it's going to affect the patient. So that's my thoughts. And I'll just throw in my two cents here is that, you know, for us, we predominantly will do either DBS for ANT and CM. The only reason we're doing, you know, RNS CM is part of the novel novelist study. So kind of the logistic part of it is getting approvals and what we can get through insurance and so we've kind of stuck to, you know, what is more feasible from that standpoint. So kind of just a practical aspect of it. So the next set of questions, it's nice how these groups together and audience thinks in the same way here is that with regards to a little bit more on stimulation parameters, so Dr. Dugan might be a little bit more involved in this aspect. But so there's one question on how stimulation parameters and RNS affect outcomes. And there's also a question similar about stimulation parameters to start with on ANT DBS, because, you know, I think this is an important aspect of that we, you know, treat these therapies as this monolithic entity of, oh, it's just, you know, DBS or RNS, but certainly the way in which you apply it has immense effects that we've seen in certainly in the movement disorder world. So, you know, diving into kind of the different stimulation parameters, I know Dr. VanCampbel had mentioned things about higher frequency capabilities and so forth, so. Right, so that's a challenging question to answer. You know, unfortunately I have to default to, it really has to be very personalized management for, you know, whether it's DBS or RNS. And unfortunately, I mean, this is precisely why we need more studies and why there are some ongoing studies right now to really try to determine what the optimal therapeutic parameters are for these relatively uncharted brain regions. I mean, you know, speaking for RNS, the company has like sort of a menu of options for the more clearly, you know, defined seizure foci, like in the neocortex. I mean, if it's, whether it's bilateral or, you know, unilateral, mesial, temporal, I think there are some tried and true, tried and tested therapeutic parameters for those areas. But when we're talking about thalamic stimulation, it's not clear yet whether the standard high frequency or standard frequency parameters that we use, for example, for the neocortex are going to work. There's some evidence that lower frequency stimulation or stimulation parameters that mimic, you know, what was used in the SANTE trial, for example, for DBS are going to be more effective. And that's why further study is undergoing right now. Just ongoing right now, I should say. Any thoughts, Dr. Van Kompen? Yeah, I mean, so just like the biomarker considerations, you know, there's a lot, you know, obviously when ENT, if you ever listen to Dr. Fisher talk about his biggest worry when we implanted the site was sleep disruption. And there's a lot of now accumulating evidence. I don't want to let the cat out of the bag because we have some nice abstracts at AES about low frequency versus high frequency stimulation and impact in sleep. And these are things we're really going to care about long-term and we're going to, you know, geek out for a number of years about it. But I think the questions are very important because we're still exploring that. You look back to why we're stimulating, what we're stimulating for ENT, which has the most data out of all the sites. These are based on mouse studies mostly, and then some large animal studies. We have, you know, the idea of resting and doing duty cycle for us, for us Santes has to do with, you know, not disrupting sleep and whatnot. These are just where we started with the effective therapies. We're going to get creative over the years trying to figure this out. You know, I see some things in the chats also about the side effects of these. I think it is important that, you know, some of these sites we may, you know, it might be like, and maybe Chen can talk about this a little bit more than I can, but, you know, in the movement disorder world, I know a lot of people avoid STN over GPI for suicidality. I know that mood related issues, we sometimes choose not to use ANTI and I'm hopeful that, you know, medial pulmonary is an alternative target, but we just don't know enough about that. And we don't know if we can program our way out of this. But what we do know is that the stimulation side effects that we see in movement disorders, which are immediate, like pain in the face or things like that, they're not immediate in epilepsy, just like our impact and seizures. So it's really hard to know without good, you know, patient interaction, how well we're helping this. That's why, you know, I think it's awesome to have a group like we have here. We have four neurology, epilepsy neurologists. That's all they do is run a neuromodulation clinic, Brian Lundstrom, Greg Worrell, Nick Gregg, you know, Keith Starnes, that's all they do. And they see it all the time. And I think we're going to need a lot more neurologists like that. Yeah, that's kind of right on exactly my thoughts as well. I think when we talk about stimulation parameters, there's a lot less known in the epilepsy realm than when we talk about it in the disorders, so. Well, we've hit our time limit. I think this was a great session. Thank you both for participating. So on behalf of the AS, I would like to thank everybody for your participation in today's event. A recording of this webinar will be sent to you within seven to 10 days. Please be sure to complete the webinar evaluation in order to claim CME credit. You should be directed to that immediately after the conclusion of this. And with that, this concludes today's presentation. Thank you, everybody. Awesome talk, Dr. Dubin. Likewise. Bye-bye. Have a great day.

intracranial neuromodulation

epilepsy

responsive neurostimulation

deep brain stimulation

surgical resection

laser ablation

anterior nucleus

centromedian nucleus

personalization of stimulation