39

Barriers and enablers to antiseizure medication adherence in children with epilepsy: a qualitative study

Eric Amankona Abrefa Kyeremaa¹, Andy Stewart², Caroline Smith³, Charlotte Lawthom⁴, Majed Alorabi¹, Meera Naran⁵, Sion Scott³, David Wright¹

¹University Of Leicester, ²Desitin Pharma Ltd, ³University of East Anglia , ⁴Aneurin Bevan University Health Board, ⁵Patient and Public Involvement Lead

Background: Antiseizure medications (ASMs) help to control seizures in about 70% of people with epilepsy [1]; however, up to 79% of children with epilepsy do not take their ASMs as prescribed [2]. Although a range of interventions have been developed to improve ASM adherence, evidence for their effectiveness in paediatric epilepsy is limited [3].

Aim: The aim of this study is to explore the barriers and enablers to ASM adherence in children with epilepsy and their carers to inform the development of a tailored intervention.

Method: A qualitative study was conducted between November 2025 and February 2026 using semi-structured interviews among children with epilepsy aged 5 to 12 years and their carers. Participants were recruited using maximum variation sampling through epilepsy charities. Interviews were conducted on Microsoft Teams lasting 35 to 84 minutes. Data were analysed using reflective thematic analysis in NVivo 14, with attention to both child and carer perspectives.

Results: Ten interviews were conducted; three with carers only and seven with children and their carers. Four main themes emerged from the analysis which were; medication-related beliefs and experiences, organisation of medicine management, physical and formulation related challenges and the role of social and professional support. While children were primarily concerned about medication formulation, particularly taste, parents were more concerned about medication side effects and reported that establishing routines supported adherence. Children with recognised learning disabilities additionally expressed concerns about the colour of their medication.

Discussion/Conclusion: This study highlights the importance of prescribing child-appropriate formulations for child with epilepsy. Healthcare professionals (HCPs) should proactively identify and address formulation-related challenges to support adherence. HCPs should prioritise clear communication about treatment expectations, address concerns about efficacy and adverse effects particularly when seizure is uncontrolled. Prescribers should consider specifying particular brands or formulations to minimise unintended switching which may affect medication adherence.

References: 1. World Health Organization. Epilepsy. 2024 [cited 2026 29/03/2026]; Available from: https://www.who.int/news-room/fact-sheets/detail/epilepsy.

2. Malek, N., C.A. Heath, and J. Greene, A review of medication adherence in people with epilepsy. Acta Neurol Scand, 2017. 135(5): p. 507-515.

3. Kangwal, C., et al., Interventions to promote medication adherence among children with epilepsy: An integrative review. Journal of Pediatric Nursing, 2024. 78: p. e51-e58.

40

Effectiveness and Safety of Cannabidiol in Adult Patients with Lennox-Gastaut Syndrome: A Post hoc Analysis of Phase 3 and Open-Label Extension Data

Matt Callaghan¹, Kishan Vyas², Teresa Greco³, Miranda Harrison², Simona Lattanzi⁴

¹Jazz Pharmaceuticals, ²Jazz Pharmaceuticals, UK Ltd., ³Jazz Pharmaceuticals, Inc., Gentium Srl, ⁴Department of Experimental and Clinical Medicine, Neurological Clinic, Marche Polytechnic University

Purpose: The randomised controlled trials (RCTs) of highly purified cannabidiol (CBD; Epidyolex® [EU]/Epidiolex® [US], 100 mg/mL oral solution) in patients with Lennox-Gastaut syndrome (LGS) reflected mixed-age populations, including paediatric patients. Analysing adult data from the pooled phase 3 RCTs (NCT02224560, NCT02224690) and their open-label extension (OLE; NCT02224573) allows the evaluation of CBD effectiveness and safety specifically in adults with LGS.

Method: This post hoc analysis included adults (≥18 years at randomisation) in the RCTs (RCT Set; 10 or 20 mg/kg/day CBD [pooled], or placebo), and those who continued in the OLE (OLE/Safety Set). Baseline characteristics, demographics, efficacy, and safety data were analysed at RCT/OLE baselines and ends (RCT Week 14, OLE Week 156) using descriptive statistics. This exploratory analysis was not intended to compare CBD versus placebo.

Results: The RCT and OLE sets, respectively, included 125 and 116 adults (49% and 48% male; median 8 prior and 3 current antiseizure medications). Median (Q1, Q3) age was 24.8 (20.5, 31.5) and 25.2 (20.7, 32.3) years; baseline 28-day drop-seizure frequency was 59.7 (25.4, 101.0) and 58.8 (24.0, 97.5). One- and 3-year OLE retention rates were, respectively, 80% (95% CI: 72, 86) and 70% (60, 77). In the pooled CBD group (N=74), median (Q1, Q3) drop- and total-seizure frequency reductions from baseline to RCT end were, respectively, -33.9% (-59.6, 8.6) and -31.6% (-54.5, -1.4); overall OLE-end reductions were -68.8% (-84.8, -40.7) and -64.6% (-80.6, -37.0). Response (≥50% reduction) for drop and total seizures was seen in 29% and 31% of patients in the CBD group at RCT end. Most frequent adverse events in the Safety Set were diarrhoea (51%), convulsion (37%), pyrexia (28%), and somnolence (28%).

Conclusions: This descriptive post hoc analysis supports the effectiveness and tolerability of CBD for adults with LGS. The safety profile was consistent with previous studies. 

Funding: Jazz Pharmaceuticals, Inc.

First presented at EEC 2026 (Vyas et al. [2026]. Abstract 415).

41

Use of Fenfluramine and Cannabidiol in daily practice: An update of a retrospective analysis of German prescription claims

Felix von Podewils¹, Milka Pringsheim^2,3, Claudio Schiener⁴, Elian Amani⁴, Manuela Molzan⁵, Viraj Pindoria⁶, Luis Möckel⁵, Iryna Leunikava⁵

¹Universitätsmedizin Greifswald, Klinik und Poliklinik für Neurologie, Epilepsiezentrum, ²Schön Klinik Vogtareuth, Neuropädiatrie/Epilepsizentrum, ³Deutsches Herzzentrum München, Klinik für Kinderkardiologie und angeborene Herzfehler, ⁴Insight Health, Waldems-Esch, ⁵UCB, ⁶UCB

Introduction: Fenfluramine (FFA) and cannabidiol (CBD [in conjunction with clobazam]) are indicated for the treatment of seizures associated with Lennox-Gastaut syndrome (LGS) and Dravet syndrome (DS) as add-on therapies for patients ≥2 years old in Germany. A retrospective prescription claims analysis reported first evidence of the use of FFA and CBD in patients with LGS or DS in Germany. The present data are an update of this analysis with a greater sample size and a longer follow-up period.

Methods: Prescription claims from December 2020 until October 2024 were analysed covering 64 million statutorily health-insured patients. Date of first FFA or CBD prescription was the index date. DS group: patients with stiripentol and without felbamate or rufinamide. Group LGS: felbamate or rufinamide and without stiripentol. Excluded patients: patients who received CBD and everolismus or vigabatrin. Patient characteristics, change in prior and concomitant antiseizure medications (co-ASMs) were analysed.

Results: A total of 494 patients who received FFA (mean age±SD [standard deviation] 14.8±13.9 years; 35.2% female; 39.1% male) and 1,509 patients who received CBD (18.5±16.1 years; 36.3% female; 38.9% male) were identified. Of these, 70 FFA (17.1±13.6 years) and 238 CBD (17.7±13.4 years) receiving patients were assigned to LGS and 108 FFA (12.7±14.8 years) and 87 CBD (14.9±13.3 years) patients were assigned to DS. Remaining 316 FFA and 1,184 CBD receiving patients could not be assigned to LGS or DS based on available data. For FFA patients with 1, 2, and 3 years of pre-index data, mean number±SD of ASMs received pre-index were 3.2±1.6 (median 3.0; n=443), 3.5±1.6 (3.0; n=304), 3.6±1.6 (4.0; n=214), and for CBD patients, 2.7±1.6 (3.0; n=1,231), 2.8±1.7 (3.0; n=793) and 2.9±1.8 (3.0; n=463), respectively. In the total FFA population, reductions in mean number of co-ASM prescriptions of 11%, 17% and 38% were observed for the periods 15-180, 181-365 and 366-546 days post-index relative to prior ASM prescriptions in the reference pre-index period of −180 to −15 days (Table 1). In those treated with CBD, the number of co-ASMs decreased by 8%, 9% and 24% for the respective periods versus the reference period. Reduction of co-ASMs was strongest in LGS patients compared to DS patients treated with FFA.

Conclusions: the reduction of co-ASM prescriptions was numerically higher in FFA patients, especially in the LGS group.

UCB-funded

42

Association of Fenfluramine Treatment and Everyday Executive Functioning in Adult Patients With Lennox-Gastaut Syndrome

Delphine Breuillard, Kelly Knupp, Adam Strzelczyk, Danielle M. Andrade, Patrick Healy, Jayne Abraham, Clare Makepeace, Amélie Lothe, Rima Nabbout

¹Necker-Enfants Malades Hospital, Assistance Publique-Hôpitaux de Paris (AP-HP), Université de Paris Cité, ²University of Colorado, Anschutz Medical Campus, ³Goethe University Frankfurt, Epilepsy Center Frankfurt Rhine-Main, University Medicine Frankfurt, ⁴Institute of Medical Science, University of Toronto, ⁵UCB, ⁶UCB, ⁷UCB, ⁸Reference Centre for Rare Epilepsies, Necker Enfants Malades Hospital, APHP, Member of the European Reference Network (ERN) EpiCARE, ⁹Institut Imagine, U 1163, ¹⁰Université Paris Cité

Rational: A previous post hoc analysis of the fenfluramine phase 3 randomized controlled trial (RCT; NCT03355209) was performed to evaluate everyday executive function (EEF) by Behavior Rating Inventory of Executive Functioning®-Adult Version (BRIEF®-A) and demonstrated improvement in patients (pts) aged 18-35y with LGS. In this post hoc analysis, we describe fenfluramine-associated changes in EEF by BRIEF®-A scores in patients from this RCT and its open-label extension (OLE; final database lock).

Methods: Patients with LGS (2–35 years) were randomized (14-week RCT, NCT03355209) to fenfluramine 0.7 mg/kg/d (maximum 26 mg/d), 0.2 mg/kg/d, or placebo; then, could enter the OLE (fenfluramine 0.2 mg/kg/d flexibly titrated up to 0.7 mg/kg/d).

Patients (18–35 years) with caregiver-completed baseline and RCT end-of-study (EOS; Day 99) BRIEF®-A were included in RCT analyses; OLE analyses included patients with BRIEF®-A at RCT baseline and OLE Month 12. Clinically meaningful improvement/worsening (Reliable Change Index≥90%/≥80%) in median BRIEF®-A T-scores (Behavioral Regulation Index [BRI], Metacognition Index [MI], Global Executive Composite [GEC]), changes in T≥65 patient frequency, and Spearman’s correlations between change in seizures associated with a fall and T-scores were analyzed.

Results: RCT and OLE data from 67 (fenfluramine [0.7 mg/kg/d, n=18; 0.2 mg/kg/d, n=24]; placebo, n=25) and 41 adults with LGS, respectively, were included.

On GEC, percentage of fenfluramine -treated patients with T≥65 at RCT baseline ranged from 36%–56%. In fenfluramine groups, percentage of patients with T≥65 numerically decreased from baseline to RCT EOS on MI (fenfluramine 0.2: 54%–33%; fenfluramine 0.7: 67%–56%) and GEC (fenfluramine 0.2: 50%–33%; fenfluramine 0.7: 56%–50%) but numerically increased in the placebo group (MI: 44%–52%; GEC: 36%–52%). In the OLE, percentage of patients with T ≥65 treated with fenfluramine <0.3 mg/kg/d numerically decreased on BRI and GEC (26%–11%; 47%–32%); a decrease in percentage of patients was also observed with fenfluramine ≥0.5 mg/kg/d on GEC (25%–0%). Median changes in T-scores for all BRIEF®-A Indexes/Composite indicated improvement or no change in RCT fenfluramine and OLE groups and worsened in the placebo group. Correlations between change in frequency of seizures associated with a fall and BRIEF®-A Indexes/Composite were negligible to weak (RCT: −0.232 to 0.024; OLE: −0.239 to −0.206).

Conclusion: Adults with LGS treated with fenfluramine had clinically meaningful EEF improvements during a 14-week RCT and its 12-month OLE. Reductions in seizures associated with a fall and EEF improvements were partially independent of each other.

UCB-funded

43

Bexicaserin for Seizure Treatment in Developmental and Epileptic Encephalopathies: Interim Analysis of an Expanded Access Program for Participants on Treatment for up to 2 Years

Caitlin Sylvia¹, Shikha Polega¹, Nadine Knowles¹, Randall Kaye², Henrik Loft¹, Tolga Uz¹

¹H. Lundbeck A/S, ²Longboard Pharmaceuticals (now a part of H. Lundbeck A/S)

Objectives: Developmental and epileptic encephalopathies (DEEs) are the most severe epilepsies, characterized by treatment-resistant seizures, frequent epileptiform activity, and developmental slowing or regression [1]. The clinical trial program for bexicaserin – an investigational, highly selective superagonist of the 5-hydroxytryptamine type 2C (5-HT2C) receptor – includes the first regulatory-endorsed study design to enroll patients with any type of DEE in a single trial. Bexicaserin was well tolerated in the Phase 1b/2a PACIFIC randomized controlled trial (RCT) and 12-month open-label extension (OLE), with a favorable safety profile and similar countable motor seizure frequency reductions in participants with Dravet syndrome (DS), Lennox-Gastaut syndrome (LGS), and other DEEs (DEE Other). The expanded access program (EAP) provides further access to bexicaserin in this underserved population. Here, we report an interim analysis of the EAP in the group of participants who rolled over from the OLE and have received up to 2 years of bexicaserin treatment (OLE + EAP).

Methods: This analysis included participants who completed the PACIFIC RCT and OLE and enrolled in the EAP. Data reflect staggered entry into the EAP and/or staggered seizure diary data entry.

Results: Of 38 participants who completed the PACIFIC OLE, 36 enrolled immediately in the EAP and continued their maintenance dosing regimen. At approximately 18 months’ treatment (OLE + EAP), 1 participant was lost to follow-up and 1 participant was reported as deceased (unrelated to study drug). The remaining 34 EAP participants on bexicaserin received approximately 24 months of treatment. Compared with the safety profile during the OLE, no new safety signals were observed. Relative to RCT baseline, the median percentage change in countable motor seizure frequency was −60.2% (n=30) at ~18 months and −53.7% (n=17) at ~24 months. Seizure frequency reductions were seen across participant subgroups at ~18 months (DS, −73.0% [n=3]; LGS, −51.2% [n=12]; DEE Other, −61.5% [n=15]) and ~24 months of treatment (DS, −100% [n=1]; LGS, −37.1% [n=6]; DEE Other, −73% [n=10]).

Conclusions: Bexicaserin continues to exhibit a favorable safety and tolerability profile in participants with a variety of DEEs for up to 2 years of treatment. Comparable reductions in countable motor seizure frequencies between the PACIFIC RCT, OLE, and EAP reinforce the sustainability and consistency of bexicaserin efficacy across DEEs, further validating its progression into Phase 3 trials.

Funding: This study was sponsored by Longboard Pharmaceuticals, Inc., now a part of H. Lundbeck A/S.

Reference: 1. Scheffer I, et al. Epilepsia. 2025;66:1014–1023.

44

Bexicaserin Seizure Treatment in Developmental and Epileptic Encephalopathies: Response Over Time and Responder Rates in the Phase 1b/2a PACIFIC Trial Open-Label Extension

David G. Vossler¹, Dennis Dlugos², Kate Riney^3,4, Nadine Knowles⁵, Tolga Uz⁶, Randall Kaye⁷, Sidsel Jensen⁶, Shikha Polega⁶

¹University of Washington School of Medicine, ²Children’s Hospital of Philadelphia, ³Queensland Children’s Hospital, ⁴University of Queensland, ⁵H. Lundbeck A/S, ⁶H. Lundbeck A/S, ⁷Longboard Pharmaceuticals (now part of H. Lundbeck A/S)

Objectives: In the Phase 1b/2a PACIFIC randomized controlled trial (RCT) and its 12-month open-label extension (OLE), bexicaserin – an investigational, highly selective 5-hydroxytryptamine type 2C (5-HT2C) receptor superagonist – was well tolerated, had a favorable safety profile, and demonstrated reductions in countable motor seizure frequency that included clinically meaningful responder rates in participants with a variety of developmental and epileptic encephalopathies (DEEs; Dravet syndrome, Lennox-Gastaut syndrome, and other DEEs). This post-hoc analysis evaluated the effect of bexicaserin on countable motor seizure frequency reductions over time in the RCT and OLE, as well as the sustainability of responses for up to 1 year, including ≥50%, ≥75%, and ≥90% responder rates.

Methods:Participants were aged 12–65 years and had a diagnosis of DEE; they could enter the OLE after completing the RCT. Baseline seizure frequency was determined from the 28-day screening period before initiation of bexicaserin or placebo. The median percentage change from baseline in countable motor seizure frequency was assessed during the RCT (bexicaserin vs placebo; Weeks 1–2/1–6/1–8/total RCT) and OLE (bexicaserin only; Month(s) 1/1–6/1–9/total OLE). The sustainability of response was evaluated as proportions of participants with ≥50%, ≥75%, and ≥90% reductions in seizure frequency during the OLE.

Results:Median percentage change from baseline in countable motor seizure frequency with bexicaserin (n=35) diverged from placebo (n=9) by Weeks 1–2 (−49.6% vs 1.5%) and was sustained for Weeks 1–6 (−58.0% vs −24.3%), Weeks 1–8 (−60.5% vs −22.0%), and over the entire RCT (−59.8% vs −17.4%). In the OLE (n=40), seizure frequency reductions were maintained through 1 year (−59.3%). Responder rates (≥50%, ≥75%, ≥90%) with bexicaserin during the RCT (60%, 31.4%, 11.4%) were maintained in the OLE (55%, 40%, 17.5%); placebo administration in the RCT only yielded a ≥50% responder rate in 33% (n=3/9) of participants, with no ≥75% or ≥90% responders.

Conclusions: Bexicaserin showed a rapid onset in countable motor seizure frequency reductions across DEEs that was sustained over time, with consistent responder rates throughout the RCT and OLE. The data indicate no loss of long-term therapeutic benefit with bexicaserin, reinforcing the rationale for the ongoing global Phase 3 DEEp program.

Funding: This study was sponsored by Longboard Pharmaceuticals, Inc., now a part of H. Lundbeck A/S.

45

Unintended consequences? Potential harms associated with the prescribing of midazolam as a rescue medication for people with epilepsy

Phil Tittensor^1,2, Zygimantas Puras⁴, Charlotte Hodgetts³, Sally-Ann Remnant⁵, Rohit Shankar³

¹Royal Wolverhampton NHS Trust, ²University of Wolverhampton, ³University of Plymouth, ⁴Southern Trust SHSCT, ⁵University Hospitals Sussex NHS Foundation Trust

Purpose: Midazolam is a short-acting benzodiazepine used in the emergency management of seizures mainly within the community. Despite its clinical effectiveness, there is limited literature specifically examining misuse, diversion, and harm related to midazolam prescribed for epilepsy. Community prescribing of midazolam presents unique challenges, as administration is frequently undertaken by caregivers or family members rather than healthcare professionals.

Method: We undertook a literature review of 70 papers for existing safeguards for midazolam misuse and developed an initial survey to gather commentary using a modified DELPHI method. From this review, a pilot questionnaire was developed and distributed to 80 members of the Epilepsy Nurses Association (ESNA) at their conference in June 2026. The results informed a final questionnaire, which was sent to epilepsy specialist nurses, pharmacists, doctors, service leads and safeguarding leads, from organisations including the International League Against Epilepsy - ILAE British Branch, ESNA, Cornwall Intellectual Disabilities Equitable Research network (CIDER), and NHS Trusts. Ethical approval was obtained from the University of Plymouth.

Results: Clinician perspectives for the potential of misuse, diversion, and harm associated with midazolam prescriptions intended for epilepsy-related use will be analysed. We will identify high-risk prescribing patterns, opportunities for inappropriate self-administration, and improvements in the detection of patients who obtain midazolam frequently or administer midazolam outside of recommended clinical indications. Final data will be presented at the Scientific Meeting.

Conclusion: Preliminary evidence suggests substantial knowledge gaps and variability in practice surrounding community midazolam use. This study seeks to provide clinician-derived data to inform future guidance, education, and safeguarding measures, with the aim of improving patient safety and reducing preventable harm associated with emergency seizure medication in the community.

46

Drivers of prescribing decision-making for nasal midazolam spray for the treatment of acute seizures in Sweden

Annabel Bowden¹, Tommy Stödberg^2,3

¹Spire Outcomes Limited, ²Neuropediatric Unit, Department of Women’s and Children’s Health, Karolinska Institutet, ³Department of Pediatric Neurology, Astrid Lindgren Children’s Hospital, Karolinska University Hospital

Purpose: Midazolam is effective in the management of acute seizures; currently available formulations can be difficult to administer, compromising seizure control. This study sought to identify drivers for prescribing a novel, unlicensed, single-dose nasal midazolam spray for the management of acute seizures.

Methods: Healthcare professionals from adult and paediatric centres in Sweden documented their reasons for prescribing a single-dose nasal midazolam spray for the prehospital management of acute seizures and sedation, in cases where rectal or buccal rescue medication had been experienced as inefficient or otherwise problematic. The data, recorded in 2024, consisted of unstructured free text. Analysis focused on cases that specifically referenced epilepsy. Thematic analysis with inductive and deductive coding was used to identify themes and categories in the data. An initial coding framework was developed from a review of literature and patient-reported outcome measures and developed iteratively. MAXQDA software was used to facilitate the thematic analysis.

Results: From 228 records, data were available for 161 patients with reference to an epilepsy diagnosis and/or various phenotypes, etiologies and comorbidities recorded. Nasal midazolam spray 2.5 mg was administered to 61 patients and 5 mg to 100, for the treatment of seizures. The initial coding framework comprised nine parent codes which were refined through the thematic analysis to ten. The most common reason given for prescribing nasal midazolam was mode of administration/method of treatment (n=122, 76%), with prescribers referring to difficulties with buccal administration (e.g. because of issues with drool and mucus), rectal administration or intravenous administration. The second most common reason for prescribing the nasal formulation was the ability to treat seizures effectively (n=105, 65%), with references to prolonged seizures that are difficult to manage, a perception that nasal treatment is most effective, and concerns with Buccolam® (midazolam) oromucosal solution and Stesolid® (rectal diazepam) in the treatment of prolonged seizures. Further reasons for prescribing single-dose nasal midazolam included ease of use (n=30, 19%), side-effects (n=11, 7%), social acceptability/privacy (n=5, 3%) (rectal administration being considered unacceptable) and frequency of use in patients experiencing daily seizures (n=5, 3%).

Conclusions: Despite an inherent selection bias, given all patients had prior difficulties or dissatisfaction with other administration routes, this analysis indicates a preference among prescribers for single-dose nasal midazolam spray over alternatives for the management of acute seizures in some patients, scenarios and semiologies, due to perceived better effectiveness and ease of use. These findings will inform future qualitative research with parents/caregivers.

47

Does a Short Shelf-Life of Epilepsy Rescue Medications Lead to an Increased Wastage Burden? The Impact of Buccal Midazolam Shelf-Life on UK Patients and Pharmacists

Matthias Koepp¹, Tim Rider², Phil Tittensor³, Ramandeep Sandhu⁴

¹Department of Clinical & Experimental Epilepsy, UCL Queen Square Institute of Neurology, ²The Leeds Road Practice, ³The Royal Wolverhampton NHS Trust, ⁴SDSmyhealthcare GP Federation, West Heath Medical Centre

Background: Unused or partially used medicines are estimated to generate £300 million of waste annually in England, around half of which is considered avoidable. In epilepsy, key buccal midazolam optimisation challenges include medicine wastage and repeat prescribing requirements. Disposal of controlled drugs also creates administrative burden and expense for pharmacies. This market research evaluated the impact of buccal midazolam shelf-life on patients, carers and pharmacists, focusing on wastage, operational burden, patient experience and quality-of-life.

Methods: Patients (n=21) and carers (n=25) completed a 12-question online survey, distributed through Epilepsy Action, assessing experiences with rescue medication and expiry-related concerns. A separate 28-question online survey was completed by UK pharmacists involved in epilepsy rescue medication supply, procurement and demand planning (community pharmacist n=80; primary care network (PCN) pharmacists n=20). Pharmacists were recruited through a specialist market research company using verified healthcare professional panels.

Results: Over the preceding 6 months, 65% (30/46) of patients/carers reported discarding at least one expired unit of rescue medication, with a mean of 2.5 discarded units per patient over 6 months. One in five participants (20%, 9/46) required an emergency prescription due to expired medication, while 13% (6/46) reported being unable to administer medication when needed. Expiry dates also negatively affected wellbeing, with 24% (11/46) reporting stress or anxiety related to medication expiry. Most patients/carers agreed that longer shelf-life would be beneficial (78%, 36/46) and would reduce waste (76%, 35/46). Among pharmacists, 21% of Buccolam units (n=96) and 16% of Epistatus units (n=71) arrived at the pharmacy with ≤6 months remaining shelf-life. Similarly, 25% of Buccolam units (n=76) and 15% of Epistatus units (n=60) dispensed to patients had ≤6 months shelf-life remaining. Disposal of expired-stock contributed additional burden, with 57% (55/96) and 42% (30/71) of pharmacists reporting disposal of Buccolam and Epistatus stock, respectively, within the past 6 months. Pharmacists agreed that extending shelf-life could reduce medicines wastage (94%, 94/100), lost revenue (92%, 92/100) and disposal cost (88%, 88/100).

Conclusion: Short shelf-life of epilepsy rescue medications contributes to avoidable medicine wastage, increased pharmacy and healthcare burden and patient/carer anxiety regarding access to emergency seizure treatment. Extending shelf-life may improve medicines optimisation, reduce operational inefficiencies and support patient confidence and continuity of care. Further evaluation of the socio-economic impact of extended rescue medication shelf-life is warranted to quantify the health-economic and service-level benefits.

Alturix funded this survey-based research. Alturix and all authors participated in the development of this abstract.

48

Real-World Prescribing Patterns of Buccal Midazolam: Implications for Pack Size and Resource Utilisation

Alison Paterson

Neuraxpharm

Background: Prolonged acute convulsive seizures (PACS) require timely rescue medication to reduce seizure duration and the risk of progression to status epilepticus. In the United Kingdom, two branded buccal midazolam oromucosal solutions are available: Epistatus®, supplied as single syringes, and Buccolam®, supplied in four-syringe packs. Previous Delphi-based modelling of PACS frequency and prescribing suggested that 4 pack Buccolam® was being wastefully prescribed to low risk PACS patients who may be more economically managed using Epistatus® single-unit dispensing and therefore reduce waste. There is limited real-world evidence describing how pack size aligns with prescribing patterns in patients who require multiple rescue doses annually.

Methods: This study describes real-world prescribing of branded buccal midazolam in Scotland, the relationship between dispensed syringe volumes, patient-level utilisation, cost and estimated resource use. Retail pharmacy prescription data were obtained from IQVIA, covering approximately 51% of community pharmacies in Scotland over a 12-month period (May 2024 to April 2025). Metrics included syringes dispensed per patient per year, syringes per prescription, and the distribution of patients receiving three or fewer versus four or more syringes annually. Cost estimates used British National Formulary list prices, applying conservative assumptions due to the absence of dosage-level data

Results: A total of 5,300 prescriptions were analysed, representing 19,926 syringes dispensed across both products. Of these, 28.7% were Buccolam® syringes and 71.3% were Epistatus® syringes. The mean number of Epistatus® syringes dispensed per patient was 7.32 per year, with an average of 3.4 syringes per prescription. Overall, 65.7% of patients receiving Epistatus® were dispensed four or more syringes annually, accounting for 91.3% of all Epistatus® syringes. Patients requiring four or more syringes annually would incur much lower medication costs if prescribed Buccolam® four-syringe packs (£296,391) compared with Epistatus® single syringes (£566,869).

Conclusions: In this real-world Scottish cohort, most Epistatus® single syringes were dispensed to patients requiring four or more rescue doses annually. These findings contrast with prior model-based analyses using six-month time horizons, highlighting the value of real-world prescribing data when evaluating pack size strategies. Resource implications appear sensitive to patient-level demand and pack size. Multi-unit pack formats appear well aligned with the needs of patients requiring higher annual syringe volumes and may support more efficient resource utilisation in this population. Further evaluation across broader UK populations may support evidence-informed prescribing and health-economic decision-making for the community management of PACS.

49

Drug-Resistant Temporal Lobe Epilepsy Responding To Weekly Pulsed Clobazam Therapy

Amogh Ravi, Sean Slaght

University Hospital Southampton

Benzodiazepines are used in epilepsy as adjunctive anti-seizure medications (ASM), rescue therapy for seizure clusters and status epilepticus, and prophylactically in catamenial epilepsy. The long-term efficacy with regular use is often limited by tolerance. We describe a patient with temporal lobe epilepsy refractory to multiple ASMs who achieved sustained seizure freedom with an unconventional weekly pulsed clobazam regime.

A 19-year-old right-handed woman of normal developmental background and no significant medical history developed her first bilateral tonic-clonic seizure from sleep. Subsequently, she experienced focal impaired awareness seizures characterised by oral automatisms, chewing, lip-smacking, deja-vu, gradual rightward head version which often progressed to bilateral tonic-clonic seizures. Seizures initially occurred at approximately 27-day intervals. Following epilepsy onset, temporary amenorrhoea obscured a suspected catamenial pattern.

Despite treatment with lamotrigine, levetiracetam, lacosamide, pregabalin, perampanel and regular clobazam, seizures remained drug resistant. Clustering was common and frequently resulted in hospital admissions with status epilepticus, responding to intravenous phenytoin.

Investigations for epilepsy surgery demonstrated a normal magnetic resonance imaging (MRI) of her brain. Repeated prolonged video electro-encephalograms failed to capture habitual seizures despite drug reduction but demonstrated inter-ictal left temporal abnormalities. Neuropsychological assessment revealed impaired verbal memory, particularly when compared to her non-verbal memory and impaired naming consistent with dominant temporal lobe dysfunction. A functional MRI confirmed left hemispheric language dominance. FDG-PET demonstrated subtle left temporal hypometabolism within normal statistical limits. The investigations were supportive of a left temporal lobe onset epilepsy but inconclusive of it. A surgical option was not pursued further due risk of post-operative language deficits and uncertain seizure freedom.

The patient trialled a progesterone implant for 6 months with no meaningful effect on seizure control. Reintroduction of clobazam or an increased rescue dose of 30mg when on a regular 10mg dose was effective in abating a cluster of seizures but the response was found to be diminishing after successive months. 6 years after the onset of seizures, patient independently trialled clobazam 20mg once a week alongside maintenance lacosamide. She achieved seizure freedom for over a year on this regime and has even applied for a driving licence.

To our knowledge, this is the first reported case of sustained seizure freedom using a weekly pulsed oral clobazam regime. Intermittent scheduled benzodiazepine administration may reduce tolerance while maintaining anti-seizure efficacy and could represent a novel strategy in selected patients with drug-refractory epilepsy who exhibit transient responsiveness to benzodiazepines. Further investigation of this approach is warranted.

50

Clinical outcomes following discontinuation of sodium valproate in women of child- bearing age: A retrospective review from NHS Greater Glasgow and Clyde (NHSGGC)

Mya Win¹, Sean MacBride-Stewart², Craig Heath^1,3

¹Department of Neurology, Queen Elizabeth University Hospital, Glasgow., ²Medicines Management Resources, NHS Greater Glasgow and Clyde, Glasgow., ³School of Health and Wellbeing, University of Glasgow

Introduction: In response to new regulatory measure by the Medicines and Healthcare products Regulatory Agency (MHRA) regarding the use of sodium valproate in women of childbearing potential, women with epilepsy dispensed valproate were recalled and treatment plans reviewed. The aim was to reduce exposure to this teratogenic medicine. However, concerns were raised regarding the impact valproate withdrawal may have on epilepsy control, particularly in women with generalised epilepsy.

Aim: This study aims to describe the clinical outcomes of valproate withdrawal in women of childbearing potential within NHSGGC.

Methods: Patients were identified through Pharmacy Information System (PIS), an administrative dataset containing all community pharmacy dispensed medication within NHSGGC. Adult women were included if they were between the age of 16-54 and had discontinued valproate since 2017. Discontinuation was defined as an absence of prescription for a period of at least 365 days.

We conducted a retrospective case note to obtain key clinical variables including indication for use. Adverse clinical outcomes were considered at the last point of contact and defined as deterioration in seizure control, either by emergence of breakthrough seizures or increase in seizure frequency from baseline, epilepsy related unscheduled care (unplanned emergency appointment with epilepsy service, epilepsy related Emergency Department (ED) attendance or hospital admission), and epilepsy related deaths.

Results: Since March 2017, 94 women discontinued valproate as a result of the MHRA process. Mean age at withdrawal was 35.63 years (SD ±8.69). Mean follow up period was two years. A generalised epilepsy was identified in 42/94 (44.7%). 61/94 (64.9%) had been dispensed valproate for more than ten years. 39/94 (41.5%) had complete seizure freedom within preceding 12 months. At the time of withdrawal, 58/94 (61.7%) women were on at least another adjunctive anti-seizure medication. Deterioration in seizure control was observed in 10/94 (10.6%); 3 patients experienced breakthrough seizures following a mean seizure-free period of 56 months, 2 patients with myoclonic jerks developed generalised tonic-clonic seizures at 10.5 months, and 5 patients experienced >50% increase in seizure frequency. Epilepsy related healthcare encounters were noted in 8/10.

Two deaths were recorded during 2-year follow up, one within a week of valproate withdrawal as a result of SUDEP. The other patient’s cause of death was unknown.

Conclusion: Although in the majority of cases, withdrawal of valproate was undertaken without adverse effects within NHSGGC, it is important that women with epilepsy are counselled regarding the risk of adverse outcomes including SUDEP.

51

Towards an emulated sequential target trial design of valproate discontinuation: baseline characteristics, hospitalisations and deaths

Sneha Mary George^1,2, Molly Wells³, Georgie Schofield⁴, Hayley Jones⁴, Alan Griffiths, Naheed Tahir, Laura J. Bonnett⁵, Gregory Y. H Lip^2,6, Iain Buchan⁴, Anthony G. Marson^1,7, Glen P. Martin³, Matthew Sperrin³, Gashirai Mbizvo^1,2,7

¹Institute of Systems, Molecular and Integrative Biology, University of Liverpool, ²Liverpool Centre for Cardiovascular Science, University of Liverpool, ³Faculty of Biology, Division of Informatics, Imaging and Data Science, University of Manchester, ⁴Department of Public Health, Policy and Systems, University of Liverpool, ⁵University of Liverpool Department of Biostatistics, Liverpool, ⁶Department of Clinical Medicine, Danish Centre for Health Services Research, Aalborg University, ⁷The Walton Centre NHS Foundation Trust, Liverpool

Background: Valproate is the most effective treatment for generalised epilepsy. Current guidance restricts its use in women of childbearing potential and more recently men aged <55 years also. It is unclear whether those who discontinue valproate will come to harm. This study aims to establish that and develop a tool to predict which the safest drug switches are. Our approach is an emulation of a target trial designed to reproduce the conditions of a randomised controlled trial, thereby minimising bias and allowing more confident causal inferences to be made from the observational data. The study is ongoing. Presented here are steps taken thus far to engineer the data, select covariates, model outcomes, design sequential trials, and involve the public.

Method: Anonymised UK Clinical Practice Research Datalink (CPRD) data were processed using SQL and DuckDB to create a research database linking primary care with mortality, A&E, inpatient, outpatient, pregnancy, and deprivation data. An analysis-ready dataset was used to design a series of emulated target trials each month from 01/01/2013-11/09/2023 (129 trials). Each trial included patients aged 16-54 years with epilepsy and ≥2 valproate prescriptions in the previous 12 months. A directed acyclic graph (DAG), domain knowledge, and public consultation informed baseline covariate selection. These included age, sex, ethnicity, deprivation, seizure frequency, number of antiseizure medications (ASMs), hospitalisations, outpatient neurology attendance, alcohol excess, neurodevelopmental disorders, psychiatric comorbidities, valproate treatment duration, and Charlson Comorbidity Index (CCI). Their distributions were assessed in relation to subsequent hospitalisation and mortality outcomes. A public involvement workshop refined covariates and outcomes.

Result: An 11.5 GB health dataset was engineered, containing 72m consultations, 198m observations, 138m prescriptions. Of 19,357 eligible patients in the first sequential trial, 63.8% were men, 36.2% women. All-cause mortality was higher in patients with alcohol excess, neurodevelopmental disorders, and psychiatric comorbidity compared with those without these conditions (21.0% vs 7.0%, 10.1% vs 7.3%, and 14.4% vs 6.9%, respectively). Mortality also increased with higher ASM use (1: 6.6%; 2: 9.0%; ≥3: 11.9%) and CCI scores (0: 4.6%; 1-3: 11.9%; >3: 29.0%;). All cause hospitalisation was also higher with alcohol excess (91.8% vs 82.0%), psychiatric comorbidity (93.0% vs 80.0%), and increasing ASM use (1: 79.6%, 2: 86.0%, ≥3: 90.3%).

Conclusion: CPRD data can support sequential emulated target trials of valproate discontinuation. Hospitalisation and mortality varied by comorbidity, psychiatric history, alcohol excess and ASM burden. These findings support careful covariate adjustment when estimating discontinuation and treatment switching safety.

52

The Patient Voice: Valproate, Topiramate and MHRA Regulation

Jane Hanna¹, Professor Ley Sander², Faye Waddams³, Sarah Jones⁴, Rachel Arkell⁵, Elly Nowell¹, Ben Donovan¹, Heather Angus-Leppen⁶

¹SUDEP Action, ²UCL Queen Square Institute of Neurology, ³Crown Prosecution Service, ⁴Royal National Orthopaedic Hospital, ⁵Kent Law School, University of Kent, ⁶University of East London, Royal Free London NHS Foundation Trust, UCL Queen Square Institute of Neurology

Valproate and Topiramate use is increasingly restricted by government regulations, such as those of the MHRA. The views of people with epilepsy and their families affected by these are largely unstudied.

This qualitative project examines the views of 22 people with epilepsy and their families about valproate prescription and restrictions. Themes explored include informed consent, shared decision making, negative consequences of avoiding valproate or topiramate (including SUDEP), missed opportunities and disruption of belief systems.

There were 5/22 deaths in the group and in 13/22 increased seizures and/or serious injuries. All 22 reported significant disruptions in life choices, employment and/or education. Recommendations to align the regulations to human rights and true informed decision making are made.

53

Cenobamate in the treatment of rare monogenic epilepsies

Ben Waters¹, Alexander Hagan¹, Vivienne Evans², Hester Garratt², Ela Akay², Shan Ellawela², Dr Rhys Thomas^1,2

¹School of Medicine, Newcastle University, ²Department of Neurology, Newcastle upon Tyne Hospitals NHS Foundation Trust, Royal Victoria Infirmary

Introduction: Cenobamate is a novel antiseizure medication with a dual mechanism of action, involving the inhibition of sodium currents and positive allosteric modulation of GABA-A receptors. In the United Kingdom, cenobamate is currently approved as adjunctive treatment for focal-onset seizures in adults with drug-resistant epilepsy. There is growing recognition of its additional therapeutic role in treating generalised seizures, with cenobamate now approved for both focal and generalised seizures in the United States. The current project aims to evaluate the utility of cenobamate on seizure frequency in rare monogenic epilepsies.

Methods: We conducted a retrospective review of patients with treatment-resistant, monogenic epilepsies being treated with cenobamate between October 2021 and February 2026. Patients were categorised into three groups based on their genetic epilepsy; group 1 (possible indication of cenobamate - TSC and LGS), group 2 (replication of cenobamate efficacy- SCN1A, SCN2A, SCN8A), group 3 (novel use of cenobamate - CHD2, DCX, POLG mitochondrial). Seizure frequency was reported by patients and/or caregivers at follow-up appointment as >25% reduction, >50% reduction, >75% reduction, no change, or worsening of seizures.

Results: Eighteen patients (11 males, 7 females) were identified, with an average age of epilepsy onset of 10 years (12 months - 14 years 6 months). The average age at which cenobamate was initiated was 28 years 5 months (21 years 6 months - 32 years). At the time of initiation, patients were on a median of 3 other anti-seizure medications (range 2-5). If discontinued (n = 7), the average duration of cenobamate treatment was 29 months. The greatest response to cenobamate was seen in group 1 (n = 9) with 56% of participants demonstrating a reduction of >50% in seizure frequency at follow-up appointment, of which 2 patients achieved seizure freedom. Group 2 (n = 6) demonstrated the poorest response to cenobamate with one participant becoming seizure-free and three participants reporting worsening seizure frequency. Group 3 (n = 3) demonstrated varying degrees of seizure reduction.

Discussion: Our findings indicate the potential utility of cenobamate beyond focal-onset seizures, with observed clinical benefit across a spectrum of epilepsy classifications - generalised, focal, mixed, and DEEs. While careful interpretation is required due to the small cohort size, heterogeneous epilepsy phenotypes, incomplete reporting and subjective interpretation of seizure frequency, our findings offer preliminary evidence supporting the use of cenobamate in broader epilepsy populations.

54

Long-Term Retention of Cenobamate in a Large Single-Centre Cohort of Adults With Drug-Resistant Epilepsy

Luisa Delazer^1,2,3, Isabella Farren^1,2,3, Kazuki Fukuma^1,2, Kajaal Shah⁴, Evelyn Frank⁴, Archie Jaques¹, Rezane Kajashi¹, Dilahxsi Vijayakumar¹, Fei Zhu^1,2, Adolfo Mazzeo^1,2, Thaera Arafat^1,2, Nicholas Fearns^1,2, Sanjeev Rajakulendran^1,2, Simona Balestrini^1,2, Matthias Koepp^1,2, Josemir Sander^1,2,3

¹UCL Queen Square Institute of Neurology, ²National Hospital for Neurology & Neurosurgery, ³Chalfont Centre for Epilepsy, ⁴Pharmacy, National Hospital for Neurology & Neurosurgery

Rationale: Cenobamate (CNB) demonstrated substantial efficacy in drug-resistant epilepsy (DRE), yet large observational studies evaluating long-term retention are scarce. Retention rate is a composite of drug efficacy and tolerability and provides a pragmatic measure of treatment effectiveness in clinical practice. We assessed long-term CNB retention in a large cohort of people attending tertiary epilepsy care centre.

Methods: We prospectively enrolled adults with epilepsy who had failed at least two antiseizure medications (ASMs) at the time of the prescription of CNB at the National Hospital of Neurology and Neurosurgery London at Queen Square and Chalfont sites. We retrospectively reviewed the records of those who were first prescribed CNB between 18 May 2021 and 30 August 2024. Follow-up data were collected up to May 2026. The primary outcome was CNB retention, defined as ongoing CNB therapy at the last available follow-up. Retention was estimated using Kaplan–Meier survival analysis.

Results: A total of 1,037 people were prescribed CNB during the study period. Of these, 1,000 initiated treatment and were included in the analysis. At last follow-up, 774 (77.4%) remained on CNB. Mean treatment duration was 24.9 months (SD 12.9), and median treatment duration was 26.5 months (IQR 16.5–34.7). Kaplan–Meier retention estimates were 89.1% (95% CI 87.2–91.1) at 6 months, 84.8% (82.6–87.1) at 12 months, 81.5% (79.1–84.0) at 18 months, 79.1% (76.6–81.7) at 24 months, 75.5% (72.6–78.5) at 36 months, and 69.3% (62.8–76.5) at 48 months. Among those who discontinued treatment, median time to discontinuation was 6.8 months (IQR 2.0–14.7).

Conclusions: In this large cohort of adults with DRE, CNB demonstrated high long-term treatment retention, with approximately 80% remaining on treatment after 2 years. Median retention was not reached, supporting sustained treatment effectiveness and tolerability in routine clinical practice. Most discontinuations occurred within the first year, suggesting that people who tolerate and benefit from CNB early may be more likely to remain on treatment long term.

55

Post-Licensing Population Data on Cenobamate Exposure and Mortality in Epilepsy

Isabella Farren^1,2, Luisa Delazer^1,2,3, Kazuki Fukuma^1,2,4, Kai Michael Schubert^1,2,5, Kajal Shah⁶, Katarzyna Saunders², Sanjeev Rajakulendran^1,2,7, Simona Balestrini^1,2,7,8, Josemir Sander^1,2,7,9, Matthias Koepp^1,2,7

¹UCL Institute of Neurology, ²National Hospital of Neurology and Neurosurgery, ³Department of Neurology, Medical University of Innsbruck, ⁴Department of Neurology, National Cerebral and Cardiovascular Center, ⁵Department of Neurology, Clinical Neuroscience Center, University Hospital and University of Zurich, ⁶Pharmacy, National Hospital for Neurology & Neurosurger, ⁷Chalfont Centre for Epilepsy, ⁸Neuroscience and Human Genetics Department, Meyer Children’s Hospital IRCCS, ⁹Department of Neurology, West China Hospital, Sichuan University

Purpose: Cenobamate became available for routine clinical use in the UK in 2022. However, post-licensing data on its potential impact on mortality are limited.

Method: As part of routine retention audits of newer antiseizure medications, we identified all individuals who initiated cenobamate within a single tertiary epilepsy service. This provided a denominator-based assessment across the full spectrum of epilepsy severity. Cenobamate exposure was modelled as a time-dependent variable, with only on-treatment time contributing.

Results: A total of 12,737 clinic attendees contributed approximately 36,272 patient-years of follow up. Between March 1, 2022, and March 31, 2026, CNB was prescribed to 1,454 individuals, yielding approximately 2,299 patient years of cenobamate exposure. Eleven deaths occurred during cenobamate treatment (4.78 per 1,000 patient years), compared with 299 deaths during 34,098 patient-years without cenobamate exposure (8.77 per 1,000 patient years), corresponding to an incidence ratio of 0.55 (95% CI, 0.30-0.96; p=0.048).

Conclusion: Although immortal time bias cannot be fully excluded and cause-of-death adjudication was incomplete, these findings are relevant to the discussion of sudden unexpected death in epilepsy (SUDEP). Because SUDEP is intrinsically under-reported and remains a diagnosis of exclusion, all-cause mortality represents a conservative but robust endpoint. Given the dominant contribution of uncontrolled convulsive seizures to SUDEP risk, these post-licensing data raise the critical question of whether potent seizure suppression with cenobamate may translate into a measurable reduction in epilepsy related mortality.

56

Evaluating the impact of adjunctive cenobamate in the management of patients with drug resistant focal epilepsy- a retrospective real world observational study

Craig Heath, Eleanor Arthur, Yousif Mohamed, Sarah Nichol, Sean MacBride Stewart, Giorgio Di Dato, Stephen Bolan, Alex Marshall, Susan Yule

NHS Greater Glasgow and Clyde

Introduction: Cenobamate (CNB) is the latest anti-seizure medication (ASM) to be licensed in the UK as an adjunctive treatment for adults with focal epilepsy. Clinical trials and expanded access programmes have shown positive results in terms of seizure reduction and retention rates, however further evidence is required to assess the performance of cenobamate in real world practice.

Since approval by the Scottish Medicines Consortium, we have been collecting audit data on all patients receiving cenobamate. Thanks to the unique clinical and administrative data available within NHS GGC, this study has the potential to provide unique insight into this ASM in a complete cohort of patients with treatment resistant epilepsy.

Methods: Patients living with epilepsy (PWE) dispensed CNB within NHS GGC between March 2022 and December 2025 were identified from the Prescribing Information System. Patients with generalised epilepsy and those younger than 18 years old at initiation were excluded.

Key clinical variables were obtained by manual review of patient records. The primary outcome of interest was change in seizure frequency. Secondary outcomes included epilepsy-related admissions, changes in ASMs and healthcare resource utilisation costs.

Results: 290 patients received at least 1 CNB prescription. 

Median age of patients at CNB initiation was 40.3 years (IQR 29.8–53.8)

232/290 (80%) had a favourable response - as of most recent point of contact 34/290 (11.7%) patients were seizure free, 137/290 (47.2%) had a greater than 50% reduction in seizure frequency, 61/290 (21.1%) had a less than 50% reduction in seizure frequency. 58/290 (20%) discontinued cenobamate - 40 stopped due to side effects, 11 due to lack of efficacy and the remainder did not cite a specific reason.

Median length of treatment was 542 days (IQR 329 - 846 days), with a median dose of 200mg/d (range 25-400mg/d). median number of previous ASMs was 5 (range 1-22).

4 patients died during the study period (one as a result of SUDEP).

Early analysis suggests a reduction in health care resource utilisation as a result of a decrease in epilepsy-related emergency department (ED) attendances.

Conclusion: These results offer insights for clinicians regarding the real-world utility of CNB in a cohort of patients with drug resistant focal epilepsy. A positive response of 80% in a population of PWE who had previously failed to respond to, on average 5 previous ASMs, and a reduction in health care resource use, highlights that cenobamate should be considered early within this population.

57

What can existing evidence tell us about treatment decisions after the first antiseizure medication failure in focal epilepsy?

Isaac Egesa^1,2, Gashirai Mbizvo^2,3,4, Richard Emsley⁵, Laura J. Bonnett¹, Anthony G. Marson^2,4, Catrin Tudur Smith¹

¹Department of Health Data Science, Institute of Population Health, University of Liverpool, ²Liverpool Interdisciplinary Neuroscience Centre, University of Liverpool, ³Liverpool Centre for Cardiovascular Science, University of Liverpool, ⁴The Walton Centre NHS Foundation Trust, Liverpool, UK, ⁵Institute of Psychiatry, Psychology and Neuroscience, King’s College London

Background: The first antiseizure medication (ASM) tried in epilepsy often fails because it does not control seizures or causes unacceptable side effects. Clinicians and patients are then faced with a difficult decision about whether to switch or add another ASM, and which ASM to use. We currently do not know which decision is best or what patient characteristics should guide it, despite a growing number of available ASMs.

Aim: To assess whether evidence can inform subsequent treatment decisions after the first ASM failure in focal epilepsy and identify the uncertainties requiring further comparative evaluation.

Methods: We synthesised four sources of evidence: a systematic review of randomised trials of subsequent ASM effectiveness after first ASM failure; a scoping review of prognostic factors; treatment pathways observed in Standard and Newer Antiepileptic Drug (SANAD) trials; and a survey of UK epilepsy clinicians. We examined what each source contributed to three decisions: why treatment failed, whether to switch or add treatment, and which ASM to select next.

Results: Six trials involving 961 patients evaluated subsequent ASMs after the first failed. The trials were small, heterogeneous and did not support one ASM over the other or switch over add-on strategy to be superior for seizure remission. Prognostic factors focused mainly on factors assessed at diagnosis rather at the point of first ASM failure. Understanding the reason for first ASM failure could help in inform subsequent treatment decisions. In SANAD and SANAD II, 38.6% and 31.7% of participants failed their initial ASM and started at least one further ASM. The most frequently selected second ASMs were carbamazepine, lamotrigine and valproate in SANAD, and lamotrigine, levetiracetam and carbamazepine in SANAD II. After failure due to inadequate seizure control, switching versus adding ASM was 60:40 in SANAD and 47:53 in SANAD II. Forty-six UK clinicians also differed in whether they would switch (48.0%) or add (52.0%) treatment, and their preferred next ASM depended on the drug that had failed.

Conclusion: The available evidence shows discrepancies in practice on treatment decisions after first ASM failure. Sequential decisions are poorly supported because trials tend to evaluate a single initial choice in isolation. Future evaluation should instead capture the patient’s whole treatment journey, randomise successive decisions and tailor them by reason for failure. Existing data can be used to design and test the feasibility of such an approach.

58

Psychological Predictors of Antiseizure Medication Side Effects: Findings from a Systematic Review and Ongoing Longitudinal Study

Liam Jarvis¹, Adam Noble¹, Rajiv Mohanraj^2,3, Paul Christiansen¹, Saif Ahmed¹, Chris Huntley¹

¹University Of Liverpool, ²University of Manchester, ³Salford Royal Hospital

Background: Antiseizure medications (ASMs) are the cornerstone of epilepsy treatment, yet their benefits are often complicated by adverse effects; ~60% of people with epilepsy (PWE) reporting at least one. While pharmacological factors such as dosage explain some variation in side-effect burden, psychological factors may also contribute. Metacognitive beliefs—beliefs about one’s own thinking—may represent one mechanism. This project comprises two complementary studies: a systematic review of psychological predictors of ASM side effects and an ongoing 9-month longitudinal study testing a theory-driven psychological model.

Methods: Study 1: A PROSPERO-registered review examined psychological factors associated with ASM side effects in adults with epilepsy. SCOPUS and Ovid MEDLINE were searched from 1971 to 2025. Screening, extraction, and risk of bias evaluations were conducted with two reviewers.

Study 2: An ongoing three-wave longitudinal study is testing whether metacognitive beliefs, derived from the Self-Regulatory Executive Function (S-REF) model, predict ASM side effect experience, medication adherence, and healthcare utilisation. Adults with self-reported epilepsy prescribed ≥1 ASM, and significant others, were recruited through a regional NHS epilepsy service and national epilepsy networks. Time 1 measures included epilepsy characteristics, metacognitive beliefs, side-effect burden, adherence, and healthcare utilisation.

Results: Study 1: Forty-three studies met the inclusion criteria, comprising over 31,000 participants and evaluating side effects associated with 24 ASMs. Most were cross-sectional. Side effects were assessed using patient-reported measures (predominantly the Liverpool Adverse Events Profile) or healthcare record review. Although a broad range of psychological factors was examined, these were often poorly defined and rarely theory-driven. Mental health variables (e.g., depression and anxiety) showed the most consistent associations with side-effect burden, with small-to-moderate effects often remaining after adjustment for epilepsy- and treatment-related factors.

Study 2: 723 participants (407 PWE and 316 significant others) have completed the Time 1 assessment. Cross-sectional analyses are underway to test whether metacognitive beliefs explain side-effect burden beyond established demographic, epilepsy, and treatment-related predictors. Preliminary findings will be presented.

Conclusions: This programme of research is the first to systematically synthesise evidence on psychological predictors of ASM side effects while also evaluating a theory-driven model of individual vulnerability. The review identifies mental health as a robust correlate of side-effect burden while highlighting the lack of theory-driven research into underlying psychological mechanisms. Findings from the longitudinal study will provide the first test of whether metacognitive beliefs contribute to ASM side effects in PWE and their significant others, informing psychologically informed side-effect management.

59

Blown Pupils and a Normal Scan: Severe Carbamazepine Toxicity Mimicking Acute Structural Brain Injury: A case report

Donald Adjorlolo, Joanna Kitley

University Hospital Southampton NHS Foundation Trust

Background: Severe carbamazepine toxicity is an uncommon but important cause of acute symptomatic seizures and profound neurological dysfunction. Patients may present with coma, absent brainstem reflexes which mimicks acute structural brain injury, however, with normal neuroimaging. Early recognition is essential, as prompt toxin elimination therapy can result in complete neurological recovery.

Case: A previously fit and well 17-year-old male collapsed at home with a generalised tonic-clonic seizure. The seizure did not terminate following 10 mg intravenous (IV) diazepam but ceased after administration of 2 mg IV midazolam. Due to reduced consciousness (Glasgow Coma Scale score of 3), he underwent pre-hospital rapid sequence induction and intubation by the Helicopter Emergency Medical Service.

On arrival, he was noted to have bilaterally dilated pupils. Initial CT brain imaging was normal. Following withdrawal of sedation, neurological examination remained abnormal with persistent coma and absent corneal, gag and cough reflexes, raising concern for severe brainstem dysfunction. Repeat CT brain imaging, CT angiography, MRI brain, MR venography and cerebrospinal fluid analysis were all unremarkable. Electroencephalography showed diffuse encephalopathy without evidence of ongoing epileptiform activity. Nerve conduction studies showed a diffuse large fibre sensorimotor neuropathy affecting both motor and sensory fibres, suggesting a toxic-metabolic process. This was done to evaluate for a peripheral neuromuscular disorder.

Further collateral history identified possible ingestion of carbamazepine prescribed for his sister, who has epilepsy. Serum carbamazepine concentration was 38 mg/L. Treatment with repeated activated charcoal and renal replacement therapy was initiated, resulting in rapid recovery of brainstem reflexes, purposeful motor responses and progressive neurological improvement.

Conclusion: This case highlights severe carbamazepine toxicity as a reversible mimic of acute structural brain injury. Acute symptomatic seizures may be followed by profound but reversible brainstem dysfunction despite normal neuroimaging. Clinicians should consider antiseizure medication toxicity in patients presenting with unexplained coma after seizures, particularly when the neurological examination appears disproportionately severe in relation to neuroimaging findings. Early toxicological evaluation, collateral medication history and prompt initiation of enhanced elimination therapies can be lifesaving and may prevent inappropriate neuroprognostication.

60

VNS Therapy is Associated with Reduced Severe Disabling Seizure Burden in Children Early in Treatment (3 months) of DRE with Continued Improvement Over 36 Months

J Wheless¹, M Zafar², G Motamedi³, F Babtain⁴, MP Fry⁵, G Giannicola⁶, R El Tahry⁷, G Groepper⁸, TJ O-Brien⁹, P Lyons¹⁰, K Liow¹¹, J O’Flaherty⁶, K Nichol⁶, M Dibue⁶, C Gordon⁶, F Beraldi⁶, A Sen¹², M Boffini⁶, COREVNS Study Group

¹University of Tennessee Health Science Center & Le Bonheur Children’s Hospital, ²Duke University Hospital, ³Medstar Georgetown University Hospital, Department of Neurology, ⁴King Faisal Specialist Hospital and Research Centre, Neurosciences Department, ⁵National Hospital for Neurology and Neurosurgery, Department of Neurosurgery, ⁶LivaNova PLC (or a subsidiary), ⁷Universite Catholique de Louvain, Centre for Refractory Epilepsy, Department of Neurology, Cliniques Universitaires Saint-Luc, and Institute of Neuroscience, ⁸Johannes Kepler University, ⁸Department of Pediatrics and Adolescent Medicine, ⁹epartment of Medicine, Royal Melbourne Hospital, University of Melbourne, Parkville, Department of Neuroscience, Central Clinical School, Monash University, Melbourne, Department of Neurology, Alfred Hospital, ¹⁰Virginia Comprehensive Epilepsy Program, ¹¹Comprehensive Epilepsy Center, Hawaii Pacific Neuroscience, Hawaii Pacific Neuroscience, University of Hawai’i at Mānoa, John A. Burns School of Medicine, ¹²Oxford Epilepsy Research Group, John Radcliffe Hospital

Purpose: Evaluate the effectiveness and safety of VNS Therapy in children <18 years of age.

Method: The CORE-VNS (NCT03529045) study collected data on seizure and non-seizure outcomes following treatment with VNS Therapy. Participants <18 (children) who received their first VNS Therapy were selected for this analysis. Seizure frequency reduction of defined most severe seizures and patient-described outcome measures (quality of life and quality of sleep) were collected at baseline and at 3, 6, 12, 24, and 36 months.

Results: VNS therapy devices were implanted for the first time into 241 (54% male, 47% female, median age=9 years, median duration of epilepsy = 5.5 years, median number of anti-seizure medications at baseline = 6) paediatric study participants (0-18 years of age). The responder rate (≥50% reduction in seizure frequency over the previous 3 months compared to baseline) for all paediatric participants was 69.1% (all seizures), 76.2% (focal seizures), and 63.3% (generalized seizures). An ≥80% reduction in seizure frequency was also noted in 51.7%, 60.1%, and 46.9%, for all seizures, focal, and generalised, respectively. Freedom from all seizures was reported by 19.7% of paediatric participants at the 36-month study visit and most participant reported either improvement or no change in quality of life. The most commonly reported treatment emergent adverse events included respiratory, thoracic and mediastinal disorders including dysphonia and cough.

Conclusion: This analysis of the CORE-VNS Study represents a large cohort of prospectively gathered paediatric data on the effectiveness of responsive adjunctive VNS therapy for DRE. Seizure reduction was noted, regardless of type, over the course of 36 months and seizure freedom was reported as early as the first 3-month visit. The long-term experience of VNS therapy in paediatric patients demonstrates significant therapeutic promise.

61

CORE-VNS Study Full Cohort at 36 Months Demonstrates Significant Reduction in Severe Disabling Seizures

A Sen¹, P Kwan², R Verner³, J Wheless⁴, R El Tahry⁵, G Motamedi⁶, F Fahoum⁷, TJ O’Brien², K Keough⁸, J Boggs⁹, A Suller Marti¹⁰, M Zafar¹¹, Dr Rhys Thomas¹², K Liow¹³, M Tzadok¹⁴, P Lyons¹⁵, M Keezer¹⁶, S Baeesa¹⁷, J Valeriano¹⁸, KA Myers¹⁹, F Babtain¹⁷, R Lee²⁰, Y Al Said¹⁷, J Ferreira²¹, K Kotulska²², J O’Flaherty³, C Gordon³, F Beraldi²³, K Eggleston³, K Nichol³, COREVNS Study Group

¹LivaNova PLC (or a subsidiary), London, United Kingdom, ²Oxford Epilepsy Research Group, John Radcliffe Hospital , ³The Alfred Hospital. Monash University and Royal Melbourne Hospital, ⁴University of Tennessee Health Science Center & Le Bonheur Children’s Hospital, ⁵Universite Catholique de Louvain, Centre for Refractory Epilepsy, Department of Neurology, Cliniques Universitaires Saint-Luc, and Institute of Neuroscience, ⁶Medstar Georgetown University Hospital, Department of Neurology, ⁷Medical & Health Sciences, Tel Aviv University; Neurological Institute, Tel Aviv Sourasky Medical Center, ⁸Child Neurology Consultants of Austin, ⁹Comprehensive Epilepsy Center, Wake Forest University, ¹⁰Department of Clinical Neurological Sciences, Department of Pediatrics, Schulich School of Medicine and Dentistry, Western University, ¹¹Duke University Hospital, ¹²Translational and Clinical Research Institute, Newcastle University, Royal Victoria Infirmary Hospital, ¹³Comprehensive Epilepsy Center, Hawaii Pacific Neuroscience, Hawaii Pacific Neuroscience, University of Hawai’i at Mānoa, John A. Burns School of Medicine, ¹⁴Pediatric Neurology Unit, Edmond and Lily Safra Children’s Hospital, Sheba Medical Center, Tel- Hashomer, Sackler Faculty of Medicine, Tel Aviv University, ¹⁵Virginia Comprehensive Epilepsy Program, ¹⁶Universite de Montreal, Department of Neurosciences and School of Public Health, ¹⁷King Faisal Specialist Hospital and Research Centre, Neurosciences Department, ¹⁸Allegheny General Hospital, ¹⁹Research Institute of the McGill University Medical Center, ²⁰Ascension Via Christi Health, ²¹St. Joseph’s Children’s Hospital, School of Medicine, Department of Pediatrics, and Pediatric Neurology, ²²The Children’s Memorial Health Institute, Department of Neurology and Epileptology, ²³Valos, LivaNova Partner CRO

Purpose: In this analysis, we describe how VNS Therapy reduces the frequency of severe seizures that result in loss of consciousness in people receiving their first neuromodulation device.

Method: Participants enrolled in the CORE-VNS (NCT03529045) study who received their first VNS Therapy implant were selected. Seizure frequency changes using the 2017 ILAE classification scheme for each seizure type as well as seizure and non-seizure outcomes and safety were collected. Participants were asked to identify their most disabling seizure type at each visit and report on the severity of that seizure type (which may have changed throughout the study). In this analysis, severe seizures that result in loss-of-consciousness are analysed: Focal Impaired Awareness Motor and Non-motor (FIA-M, FIANM), Focal-to-Bilateral Tonic-Clonic (FBTC), and Generalised Tonic-Clonic (GTC).

Results: Of those meeting criteria, 531 proceeded to first implantation, 526 had at least one post-implantation follow-up visit, and 426 completed the study with at least 36 months of follow-up. At 36 months the median change in FIA-M seizures was -80.1% (95%CI: -90, -66.7), FIA-NM seizures was -94.4% (95%CI:-100, -77.3), FBTC seizures was -95.0% (95%CI: -100, -72.7), and GTC seizures was -84.0% (95%CI: -100, -66.7). The median monthly frequency of GTC seizures decreased from 4.0 (95%CI: 2.3, 6.7) at baseline to 1.00 (95%CI: 0.0, 1.7) at 36 months. Post-ictal severity for the patient’s self-reported “most disabling seizure” also decreased. Finally, participants tended to report less use of rescue medication, the emergency department, and seizure-related hospitalisation in the follow-up. For all participants, n=92/794 (11.6%) reported an adverse event that was causally, possibly, or probably related. Commonly reported adverse events were consistent with the safety profile of VNS Therapy and included dysphonia, dyspnea, and cough.

Conclusion: VNS Therapy reduces the frequency of severe disabling seizures with loss of consciousness, resulting in a reduced risk of serious health sequelae.

62

Structural morphometry distinguishes vagus nerve stimulation outcomes in individuals with epilepsy

Harry Clifford¹, Sonja Fenske¹, Peter Taylor^1,2,3

¹CNNP Lab, ²Faculty of Medical Sciences, ³UCL Queen Square institute of Neurology

Introduction: Vagus nerve stimulation (VNS) is a commonly used treatment for drug-resistant epilepsy. However, response to VNS is variable, with understanding of this response variability being limited. The vagus afferent network (VagAN), the neural circuits thought to be involved in response to vagal afferents, has been implicated in response prediction in modalities such as diffusion weighted imaging, functional imaging, and neurophysiology. Structural abnormality in the VagAN from T1w MRI has not yet been studied.

Methods: We assessed if VagAN abnormalities, derived from pre-implantation T1w MRI, were associated with VNS response in 92 individuals with epilepsy two-years post-implantation (42 responders, 50 non-responders). We used the first principal component (PC1) from a robust principal component analysis for a variety of region sets trained using 100 healthy controls. We investigated two primary hypotheses. (i) If VagAN regions’ PC1 differentiates response groups, (ii) if other region sets did not differentiate response groups.

Results: Non-responders had significantly higher VagAN PC1 abnormality scores than both responders (r = 0.30, p = 0.008) and controls (r = 0.37, p < 0.005). In contrast, non-VagAN PC1 showed no discriminatory value (r = −0.03, p > 0.05). VagAN outperformed all other region sets.

Significance: Structural abnormality of VagAN was associated with non-response to VNS. This suggests that volume abnormalities in key VNS regions correlate negatively with reduction of seizures. Abnormalities in other regions, such as the non-VagAN region set, were not associated with VNS response. This metric, when combined with other modalities, may be a useful measure to improve the selection of individuals with drug-resistant epilepsy for VNS therapy.

63

Vagus nerve stimulation in adults in Newcastle: a retrospective notes audit

Alexander Wright-Todd^1,2, Rhys Thomas^2,1, Hester Garratt²

¹Medical School, Newcastle University, ²Newcastle upon Tyne NHS Hospitals NHS Foundation Trust

Introduction: Vagus nerve stimulation therapy is a recognized treatment for drug refractory epilepsy and has been established in the UK for decades. We set out to review our clinical service.

Methods: We used the electronic patient record system to identify patient demographics, to classify their epilepsy and capture change in seizures following VNS implantation. We categorised this a worse, improved (but not seizure free), seizure free and unclear/unascertainable. We looked at patterns of seizures at two and at five years where data were available.

Results: We identified 139 unique patients. As of June 2026, 121 were alive (87%) and 73 were male (52.5%). 57 had an intellectual disability (41%). 94 had focal epilepsy (68%), 27 had DEE (19%) and 16 had GGE (12%), 2 had types of epilepsy that were unable to be determined (1%). The median age of VNS implantation was 30 years old; the youngest patient was 3 years and the oldest 68 years.

10% had seizure aggravation, and 45% were improved; including the 10% who were seizure free. In 45% we could either not identify a seizure change or there was no discernable difference.

VNS peak efficacy was seen at two years – with no significant changes seen between two and five years; noting that 73 people could not be used for the five year analysis: unable to determine original seizure frequency, unable to determine frequency at 5 years, patients deceased, no notes either from time of VNS implant or from after 5 years. Of the 10 with a DEE at five years we saw a 70% improvement (20% unchanged and 10% worse)

Conclusion: Clinical records are commonly used for retrospective analysis but data are frequently captured in a structure that prevents us to analyse the impact of our clinical decisions. For VNS therapy this is regrettable as not only is the treatment invasive, but we need the ability to better select who will and who won’t respond to neuromodulation. The NIHR funded VNS-ALERT randomised clinical trial will seek to report on VNS therapy efficacy and cost efficacy across the UK.

64

Optimising EASEE® Device Placement through Integrated multimodal Structural and Functional Imaging, and Electroclinical Modelling

Sherry Liu, Hak Kei (Antonia) Wong, Thomas Jensen, Anna Miserocchi, Mr Andrew Mcevoy, Roman Rodionov, John S Duncan, Meneka K Sidhu

University College London, Institute of Neurology, Department of Epilepsy, Queen Square Institute of Neurology

Introduction: Epilepsy is recognised as a network disorder, in which abnormalities in distributed networks may contribute to seizure generation and propagation. This concept is especially relevant for patients unsuitable for resective surgery, who may benefit from network-level modulation. The EASEE® system is an emerging form of individualized epicranial neuromodulation, with approximately 60% of patients achieving >50% reduction in seizures at 2 years. Optimal device placement remains uncertain in current practice, particularly in individuals with prior resections, in whom the impact of large surgical cavities on stimulation efficacy is poorly understood. We developed a novel approach for optimising EASEE® device placement and report preliminary 6-month outcomes in first four patients, including individuals with previous cerebral resection.

Methods: We developed a workflow integrating 3D-neuroimaging data to guide individualized EASEE® implantation planning using our in-house surgical planning software, EpiNav. Structural lesions, previous resection cavities, and relevant functional abnormalities, including PET and SPECT data were integrated into 3D models with scalp and intracranial EEG data to define the most likely region of seizure generation. A virtual EASEE® electrode simulation model was then created using SimNibs to simulate volume of tissue receiving various levels of stimulation (V/m), with a current of 2 and 4mA, to optimise stimulation of the putative epileptogenic zone, accounting for resection cavities, and differential electrical conductivity.

Results: To date, ten patients underwent EASEE® implantation using this approach, four of which had available 6-month follow up data. 3/4 patients had >50% reduction in seizures, including two patients who had prior resection.

Conclusion: Our preliminary experience suggests that individualized, image-guided EASEE® placement is feasible, including in patients with previous resections. Integrating structural, functional, and electroclinical data may help to optimise electrode positioning, particularly when conventional anatomical targeting is difficult because of prior surgery or complex epileptogenic networks. Early outcomes are encouraging, but larger cohorts and longer follow-up are underway to determine whether this approach improves patient selection, device placement, and seizure outcomes.

65

SEEG for Delineating the Epileptogenic Zone in Extensive Structural Lesions

Mohammed Elkider, Samden Lhatoo

University of Texas Health Science Center at Houston (UTHealth Houston)

Background: SEEG may be particularly beneficial in patients with extensive structural lesions, such as large focal cortical dysplasia, post-traumatic encephalomalacia, polymicrogyria, or perinatal stroke, because the visible lesion on MRI does not necessarily correspond to the full extent of the epileptogenic zone (EZ).

Methods: We report a 19-year-old right-handed male with drug-resistant focal epilepsy following two intracerebral hemorrhages due to ruptured arteriovenous malformation. He underwent left craniotomy at age six and developed medically refractory epilepsy after recurrent hemorrhage in 2017. MRI demonstrated extensive left parietal, posterior temporal, and occipital encephalomalacia. SEEG implantation sampled temporal, parietal, and occipital regions to define the EZ and perform functional cortical mapping.

Results: Interictal SEEG demonstrated continuous slowing over the left temporal pole, periodic discharges in the left mesial occipital and medial temporo-occipital (lingual) regions, and persistent epileptiform activity involving the left lateral posterior temporal cortex and angular gyrus. Frequent epileptiform discharges were also recorded from the amygdala with propagation to the mesial hippocampus and temporal pole, predominantly during sleep.

Twenty-one electrographic seizures were recorded. Ten originated from the left angular gyrus with early involvement of the left lateral posterior temporal cortex, eight from the medial temporo-occipital (lingual) region, two from the calcarine cortex, and one from the occipital cuneus. Two habitual electroclinical seizures were captured, both characterized by impaired awareness and rightward eye deviation, with one preceded by apnea. In both seizures, ictal onset localized to the left angular gyrus and lateral posterior temporal cortex, followed by spread to the amygdala and mesial hippocampus after 5–6 seconds, excluding the mesial temporal structures as the primary seizure onset zone.

Electrical stimulation of the lingual and calcarine cortices elicited contralateral visual phenomena. Language mapping demonstrated impaired visual naming with posterior basal temporal stimulation, while language, calculation, and working memory were preserved during stimulation of the angular gyrus and lateral posterior temporal cortex. SEEG localized a non-contiguous posterior quadrant EZ involving the left angular gyrus, posterior temporal cortex, lingual gyrus, and cuneus.

Conclusions: This case demonstrates the value of SEEG in patients with extensive post-hemorrhagic encephalomalacia, where MRI alone may not accurately define the epileptogenic network. SEEG excluded mesial temporal seizure onset, identified a multifocal posterior quadrant EZ, and provided essential functional mapping that predicted complete right homonymous hemianopia if the entire EZ were resected. These findings directly informed surgical decision-making by balancing the likelihood of seizure freedom against anticipated visual morbidity.

66

Role of MEG in SEEG Implantation Planning

Mohammed Elkider, Pedro Balaguera, Abdulrahman Alwaki

University of Texas Health Science Center at Houston (UTHealth Houston)

Background: Magnetoencephalography (MEG) is an important non-invasive functional imaging modality in the presurgical evaluation of drug-resistant epilepsy. In addition to localizing interictal epileptiform activity, MEG can inform stereoelectroencephalography (SEEG) implantation by identifying epileptogenic regions that may not be evident on MRI or scalp EEG. Incorporating MEG into SEEG planning may optimize electrode placement, reduce unnecessary sampling, and improve localization of the seizure onset zone, particularly in MRI-negative or multilobar epilepsy.

Methods: We report a 44-year-old right-handed man with medically refractory focal epilepsy. His habitual seizures began with arousal from sleep followed by sitting upright, right facial numbness and a pulling sensation associated with an urge to touch his face. He subsequently developed body-directed automatisms characterized by alternating hand movements covering his mouth, accompanied by sialorrhea without impaired awareness. Seizures lasted 20–30 seconds with rapid recovery. Scalp EEG demonstrated repetitive spikes over the left temporo-centro-parietal region (T7/P7>C3/P3), consistent with a tangential dipole across the insular-opercular region. MRI was initially interpreted as non-lesional. MEG demonstrated abundant clustered dipoles with uniform orientation localized to the left lateral frontal cortex immediately anterior to the precentral sulcus, guiding SEEG implantation.

Results: Phase II SEEG evaluation captured five habitual electroclinical seizures. Ictal onset consistently localized to the left frontal operculum (MSIN contacts 5–10), with early propagation to the left rolandic operculum and subcentral gyrus (PSIN contacts 3–11). Interictal recordings demonstrated continuous focal slowing and near-continuous epileptiform discharges involving the same regions, supporting a highly localized epileptogenic network. Functional cortical stimulation showed preserved visual naming, auditory naming, and reading during stimulation of the superior temporal gyrus, pars opercularis, and adjacent frontal operculum. Stimulation of MSIN contacts 7–8 and 8–9 reliably reproduced the patient’s habitual aura and electroclinical seizure, further confirming localization of the epileptogenic zone.

Conclusions: SEEG localized the epileptogenic zone to the left frontal operculum, concordant with the tightly clustered MEG dipoles. The combined electrophysiological findings were highly suggestive of focal cortical dysplasia. Prompted by the concordant MEG and SEEG findings, repeat review of the MRI identified subtle imaging features suspicious for focal cortical dysplasia that had not been recognized initially. This case illustrates the complementary role of MEG in SEEG implantation planning for MRI-negative epilepsy, demonstrating how MEG-guided electrode placement can accurately localize the epileptogenic zone, improve diagnostic confidence, and uncover subtle structural abnormalities that directly influence surgical planning.

67

Murray Alexander Falconer (1910–1977): his contribution to the modern International League Against Epilepsy (ILAE) classification of epilepsies and British epilepsy surgery

Amisha Vastani, Harutomo Hasegawa, Richard Selway

Department of Neurosurgery, King’s College London

A pioneer of epilepsy surgery whose work led to describing mesial temporal sclerosis, focal cortical dysplasia and the en bloc temporal lobectomy. We present M. Falconer’s scientific and operative contribution to the modern ILAE framework, emphasising pathological substrates, surgical outcomes, and prevention strategies defined during his 26-year tenure directing the Guy’s–Maudsley Neurosurgical Unit.

Design: Structured historical literature review alongside his personal archive at King’s containing his and colleagues’ work. We outline key publications and operative innovations, appraising each against contemporary ILAE position papers on hippocampal sclerosis, focal cortical dysplasia, seizure classification, and surgical outcome measurement.

Subject: M. Falconer FRCS FRACS (Dunedin 1910 – London 1977): Nuffield Dominions Clinical Fellow, Oxford 1938 (under Sir Hugh Cairns); RAMC Major, St Hugh’s Military Hospital for Head Injuries 1940–1945; Director, Guy’s–Maudsley Neurosurgical Unit 1949–1975. Produced approximately 50 peer-reviewed epilepsy papers in collaboration with Hill, Pond, Meyer, Serafetinides, Corsellis, Bruton, Taylor, Davidson, and Engel.

Method: A 1953–1975 chronology of scientific publications constructed from his personal archive, cross-referenced with contemporary ILAE position papers: HS-ILAE 2013; FCD ILAE 2011 and 2022; ILAE Classification of Epilepsies 2017; and the Updated Seizure Classification 2025.

Results: Key publications and ILAE relevance: the 1953 JNNP report of paediatric temporal-lobe epilepsy cured by lobectomy; the 1954 Lancet paraphilic automatism abolished by resection; the 1955 Lancet series (n=31; first standardised en bloc anterior temporal lobectomy, the operative method whose intact specimens made ILAE neuropathological classification achievable); 1957 Journal of Mental Science psychiatric outcome study, republished Epilepsy & Behavior 2004; 1958 Brain paper on Meyer’s loop visual-field defects; 1963 JNNP 100-patient ictal-speech series; 1964 Archives of Neurology paper formally naming mesial temporal sclerosis as the dominant substrate, the HS-ILAE 2013 Type 1 entity; 1971 JNNP focal cortical dysplasia (Taylor, Falconer, Bruton, Corsellis), eponymous Taylor-type FCD and foundation of ILAE FCD Type II 2011/2022; 1973 NEJM behavioural reversibility (n>250); 1974 Lancet initial-precipitating-injury MTS-prevention paper, read to the ILAE British Branch, National Hospital Queen Square, 26 April 1974, and cited in NICE guidance; 1975 Brain (Engel, Driver, Falconer) and the Engel outcome scale; and 1975 Lancet paediatric series (n=40, 1–24-year follow-up).

Conclusion: Across four decades, Falconer established the operative and pathological framework upon which the modern ILAE classifications of hippocampal sclerosis, focal cortical dysplasia, and surgical outcome are built. His previously undescribed clinical archive at King’s offers a new primary source for understanding his contributions to the ILAE and modern epilepsy management.

68

Focal Cortical Dysplasia in Epilepsy Surgery Candidates: A 10-year Experience at a Regional Centre 

John Holden, Leach Veronica, Livingstone Shona, Athanasios Grivas, Russell Hewett

Institute Of Neurological Sciences, Glasgow

Background: Focal Cortical Dysplasia (FCD) is a major structural cause of focal drug-resistant epilepsy (DRE), accounting for up to 14% of surgical resections in the European Epilepsy Brain Bank (EEBB) registry [1]. This regional, 10-year retrospective study aims to evaluate the prevalence, patient phenotype, and surgical outcomes of histologically proven FCD. 

Methods: Electronic patient records from all cases discussed at the regional epilepsy surgery multidisciplinary team (MDT) meeting (January 2015–December 2025) were retrospectively reviewed. Patients were classified into two cohorts: pathologically proven FCD (FCD-P) and radiologically suspected FCD without surgical resection (FCD-R). Extracted data included demographics, age at seizure onset/surgery, electro-clinical findings, neuroimaging, surgical interventions, histopathology (ILAE consensus guidelines), and postoperative outcomes (Engel/ILAE scales). 

Results: Of 216 unique cases discussed, 55 (25%) underwent an epilepsy surgical procedure; 6 patients had historical neurosurgical procedures. Resections accounted for 93% (n=57) of total interventions. Among the 57 cases with histopathological data, the most frequent diagnoses were low grade epilepsy-associated tumours (n=17), vascular malformations (n=10), isolated FCD (n=10), hippocampal sclerosis (HS; n=7), and non-diagnostic pathology (n=7). Structural dual pathology (combined HS and FCD; ILAE FCD Type III) was identified in 3 cases. Within the FCD-P cohort, the commonest sub-classification was ILAE FCD Type II (n=8; 4 Type IIa, 3 Type Type IIb, 1 unspecified). An additional 12 cases (FCD-R) demonstrated radiological features consistent with possible FCD but did not progress to surgery. 

The combined FCD cohort (FCD-P + FCD-R, n=25) had a median age of seizure onset of 9 years and a median age at MDT discussion of 35 years. Seizure frequency was predominantly daily (n=12) or weekly (n=11). Patients were prescribed a median of 3 anti-seizure medications (ASMs), with historical exposure to a median of 4 others; levetiracetam and lamotrigine were most common. Postoperatively, 61.5% of the FCD-P cohort (8/13) achieved an Engel Classification score of I. 

Of note, 58 of 216 unique cases discussed (27%) with refractory focal epilepsy presented with normal MRI findings. 

Conclusions: FCD represents a substantial proportion of the surgical cohort at our centre. Patients face a high seizure and treatment burden, yet operative management yields highly satisfactory outcomes in the majority of cases. These findings emphasize the benefit of epilepsy surgery in management of FCD. 

 References: 

1. Blümcke, I., et al. (2017). “Histopathological Findings in Brain Tissue from Epilepsy Surgery.” New England Journal of Medicine, 377(17), 1648–1656.

69

Early UK experience with robot-assisted Visualase MR-guided LITT for drug-resistant epilepsy: workflow, technical metrics and outcomes in the first 22 cases

Amr Moursi, Thangaraj Munusamy, Nandini Mullatti, Harutomo Hasegawa, Richard Selway

King’s College Hospital NHS Foundation Trust

Purpose: Magnetic resonance-guided laser interstitial thermal therapy (MRg-LITT) is a minimally invasive option for drug-resistant focal epilepsy, particularly for deep or surgically challenging targets. UK experience remains limited. We describe a single-centre early experience with robot-assisted Medtronic Visualase MRg-LITT, focusing on indications, technical metrics and outcomes.

Methods: We retrospectively reviewed the first 22 consecutive patients who underwent robot-assisted Visualase MRg-LITT for epilepsy at King’s College Hospital between July 2024 and June 2026. Data included demographics, indication, preoperative workup, applicator and ablation parameters, delivered energy, follow-up duration, Engel outcome and complications. 

Results: Twenty-two patients were treated: 20 adults and 2 paediatric patients; 15 were male. Median age was 32 years (range 12-56) and median epilepsy duration was 21 years. Indications were hypothalamic hamartoma (HH) in 11, mesial temporal or temporal targets in 5, cavernoma in 2, periventricular heterotopia in 2 and other focal cortical lesions in 2. A single trajectory was used in all cases. A 3 mm diffusing tip was used in 12 patients and a 10 mm tip in 10. Median ablation runs were 2.5 (range 1-5), with a median of 1.5 pull-backs and median ablation time of 628 seconds (range 120-1554). Median mean power was 4.95 W (range 2.71-8.50), approximately 33% of the 15 W maximum. Engel outcome was documented in 21 patients at median follow-up of 7 months (range 0-19): Engel I in 14 (67%), Engel II in 1 and Engel III in 6. In patients with at least six months’ follow-up (N=11), Engel I was achieved in 4 (36%), Engel II in 1 and Engel III in 6. Hypothalamic weight gain occurred in six HH patients. One small localised haemorrhage resolved without neurological deficit. No permanent neurological deficits were observed.

Conclusion: This early UK experience demonstrates the feasibility and safety of a robot-assisted Visualase MRg-LITT epilepsy service across heterogeneous indications with no permanent neurological morbidity. Applicator selection, ablation time, power, and delivered energy were reproducibly measurable and may support standardised reporting. These are early results from a small, recent series with a short follow-up. Longer follow-up is essential before firm conclusions can be drawn, and careful counselling is needed.

70

Reduced Preoperative Piriform-Insula Connectivity Is Associated with 5-Year Seizure Outcome After Anterior Temporal Lobe Resection

Hak Kei (Antonia) Wong, Sherry Liu, Jane de Tisi, Boyuan Song, Anna Miserocchi, Andrew Mcevoy, John S Duncan, Louis Lemieux, Umair Javaid Chaudhary, Meneka K Sidhu

University College London, Institute of Neurology, Department of Epilepsy, Queen Square Institute of Neurology

Introduction: Piriform cortex has large scale network interactions: orbitofrontal-thalamic, limbic and anterior insula integrating salience network. It plays a critical role in seizure initiation, propagation and termination. This study investigates preoperative piriform cortex resting-state fMRI (rsfMRI) functional connectivity (FC) profiles associated with longer-term seizure outcomes after anterior temporal lobe resection (ATLR) for unilateral temporal lobe epilepsy (TLE).

Method: 13 healthy controls and 27 TLE patients (16 left) who underwent ATLR had preoperative rsfMRI. Seed-based FC analyses from piriform were performed using CONN22 toolbox-v2407. The effect of 5-year postoperative seizure outcome (ILAE 1-6) was investigated as a covariate (controlling for focal to bilateral tonic-clonic seizures, age at onset of epilepsy, pathology, preoperative seizure frequency). All results were reported with false discovery rate correction (FDR p<0.05) unless otherwise stated.

Results: 50% had ILAE 1-2 outcomes. Compared with controls, left TLE showed reduced connectivity between left piriform and bilateral insulae. In RTLE, there was reduced right piriform-salience network connectivity (uncorrected p=0.008) compared to controls. 

In combined left and right TLE analyses, worse seizure outcomes were associated with reduced left piriform-salience network, bilateral insulae and planum polare connectivity.

In left TLE, worse seizure outcome was associated with decreased left piriform-contralateral insula connectivity. A similar trend was observed for left piriform-salience network connectivity (p-FDR=0.056). 

In right TLE, seizure outcome was not significantly associated with any connectivity patterns, possibly reflecting a type II error due to smaller sample size.

Conclusion: We showed bilateral disruption of piriform cortex connectivity primarily to insulae in TLE. This hypoconnectivity was associated with worse seizure outcome after ATLR implying that more diffuse piriform network disruption is a potential biomarker for seizure outcome after ATLR. This may provide a neurobiological understanding of why adequate removal of the piriform cortex is associated with better longer term seizure free outcomes after ATLR.