Front. Neurol.Frontiers in NeurologyFront. Neurol.1664-2295Frontiers Media S.A.10.3389/fneur.2025.1644835NeurologyReviewPrecision medicine for epilepsy: challenges and perspectives for an optimized clinical care pathwayMorita-ShermanMarcia1†TrinkaEugen234†KwanPatrick567†IkedaAkio8†ChoMin1†HampelHarald1*†1Eisai Inc., Nutley, NJ, United States2Department of Neurology, Neurointensive Care, and Neurorehabilitation, Christian Doppler University Hospital, Paracelsus Medical University, Member of EpiCARE, Salzburg, Austria3Neuroscience Institute, Center for Cognitive Neuroscience, Christian Doppler University Hospital, Salzburg, Austria4Karl Landsteiner Institute for Clinical Neurosciences, Salzburg, Austria5Department of Neuroscience, School of Translational Medicine, Monash University, Melbourne, VIC, Australia6Department of Neurology, Alfred Health, Melbourne, VIC, Australia7Department of Neurology, Melbourne Brain Centre at Royal Melbourne Hospital, University of Melbourne, Melbourne, VIC, Australia8Department of Epilepsy, Movement Disorders and Physiology Kyoto University Graduate School of Medicine Shogoin, Kyoto, Japan
Edited by: Georgia Ramantani, University Children’s Hospital Zurich, Switzerland
Reviewed by: Kapil Gururangan, Northwestern University, United States
Md. Al Hasibuzzaman, Ningbo First Hospital, China
Antonio Giulio Gennari, University Children’s Hospital Zurich, Switzerland
*Correspondence: Harald Hampel, Harald_Hampel@eisai.com
This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
Epilepsies are a common, but heterogeneous group of brain disorders, characterized by an enduring predisposition to recurrent epileptic seizures. Recognizing epilepsies as a disease spectrum offers compelling opportunities to implement precision medicine in routine care. In this narrative review, we assess the status and development of precision epilepsy, compare its implementation with the advanced model of precision oncology, and discuss strategies to advance the implementation of precision medicine in epilepsy care. We aim to raise awareness about the current state-of-the-art approaches in precision epilepsy, emphasizing their potential to optimize epilepsy care. Rapid technological innovations provide the foundation to improve epilepsy research and management including the establishment of multi-dimensional biomarkers to aid disease subtyping and treatment decision. We also introduce emerging digital health technologies that will transform seizure monitoring and prediction. Advances in data science and artificial intelligence will deepen our knowledge of epilepsies, and may deconstruct and systematize historical, clinical, and descriptive concepts. Following a thorough examination of the current epilepsy landscape – including obstacles against precision medicine implementation and clinical adoption - we envision that the path toward precision epilepsy care lies in studies uncovering the mechanisms underlying systems-biology and neurophysiology-based epileptogenesis using technological innovations, such as genetic testing, fluid indicators, neuroimaging, neurophysiology, and wearable devices. We review the literature based on four core pillars - biomarkers, digital technologies, systems medicine, and data science - to pinpoint the unmet need for epilepsies and thus revolutionize disease management strategies.
precision medicine in epilepsysystems medicinebiomarkersdigital technologiesdata sciencesection-at-acceptanceEpilepsyIntroduction
Epilepsies are a heterogeneous group of brain disorders (1), characterized by a continuing predisposition to epileptic seizures, with cognitive, psychological, and social consequences. Epilepsies impose a high burden on patients, caregivers, and society, disproportionately affecting low- and middle-income countries (2, 3).
Pharmacological treatment remains the primary method for achieving long-term seizure control despite limitations such as side effects and drug interactions (4). Existing challenges in epilepsy care include drug resistance, the lack of preventive or disease-modifying treatment (5, 6), and the absence of established biomarkers to predict seizure onset and treatment outcomes. Despite the ever-growing number of anti-seizure medications (ASMs) that target diverse molecular mechanisms, such as selectively modifying preictal-ictal transition currents, enhancing inhibition, blocking ionotropic glutamate receptors, and through interactions with elements of the synaptic release machinery; or by combinations of these mechanisms (7) approximately 30% of patients on ASM treatment continue to experience seizures, as observed over the past three decades (8, 9). Drug resistance remains a clinical challenge, largely due to an incomplete understanding of the underlying mechanisms of drug resistance. New treatment advances, such as the potential use of focused ultrasound to allow for selective blood–brain barrier permeability and localized drug delivery, hold promise for preventing seizures (10). Patients who fail to respond to treatment with ASM require a diagnostic re-assessment to confirm and refine the epilepsy diagnosis and should be evaluated for other treatment options, including surgical interventions, neuromodulation, and dietary therapies (4).
In the absence of disease-modifying treatments, a biology-based, holistic framework is needed to replace the current symptom-oriented care model (11, 12). The World Health Organization’s Intersectoral Global Action Plan (WHO-IGAP) on epilepsy aims to improve care in epilepsies by proposing strategies to reduce epilepsy stigma and burden, and promote research innovation (2, 3). The WHO-IGAP identifies precision medicine (PM) as a promising innovation with the potential to accelerate the improvement of epilepsy care (5, 13).
PM is defined as “prevention and treatment strategies that take the individual variability into account” and was proposed as a holistic framework to address the complexity of a multifactorial disease spectrum (14). Some treatment paradigms or decision-support tools have been inaccurately described as PM. This misinterpretation considers “precision” to mean localization, personalization, or prognostication. While these are components of a PM model for epilepsy, they do not encompass the full scope. PM is tailored to an individual’s specific multi-dimensional characteristics and consequently has the potential to revolutionize epilepsy care by improving early detection, diagnosis, classification, prevention, and treatment (13). While some use personalized medicine and PM interchangeably, they are not synonymous. Personalized medicine is a targeted approach that guides treatment decision-making based on evidence, individual conditions (genetic profile, environment, and lifestyle) and clinical skills of healthcare providers. The existing model of ASM selection is an example of personalized medicine. ASM selection considers clinical and socioeconomic factors and leverages evidence-based data, physician experiences, and patient preferences. Rather, PM is an extension of the traditional personalized care model by offering more precise and individualized care through the utilization of sophisticated diagnostic tools enabled by recent technological advances (13, 15).
Epilepsies provide compelling opportunities for PM given their measurable outcomes, known risk factors, and established animal models (16). To approximate and appreciate the special complexity of the human brain, affected by pathophysiological mechanisms leading to diverse clinical phenotypes, this review will explore PM in epilepsy using a previously established general PM framework for neurology and psychiatry (17) focusing on four main pillars: biomarkers, digital technologies, systems medicine, and data science, and will discuss major limitations to implementation compared to the ones faced by oncology. We highlight, based on examples, how the pillars of PM in epilepsy have already been established and how we can leverage the experience gained in oncology to further expedite the implementation of PM in epilepsy care. Ultimately, a better understanding of the current state of PM in epilepsy will illuminate the path forward. Finally, in this narrative review, we aim to raise awareness of the state-of-the-art in epilepsy care and how precision epilepsy can enhance the management of these complex diseases. While numerous examples are described in this review, providing a comprehensive overview of all tools currently available in epilepsy care is beyond the scope of this review.
PM in epilepsy: modeled from oncology
PM-oriented approach was pioneered and substantially advanced in oncology. Learnings from this development serves as a conceptual framework for other complex disease areas. Data extracted from tissue samples shared through collaborative studies led to the identification of specific gene alterations associated with many cancer types (18). The Cancer Genome Atlas (TCGA) collaborative research initiative (started in 2006) analyzed numerous cancer tissues and generated over 2.5 petabytes of genomic, epigenomic, transcriptomic, and proteomic data that is publicly available to researchers worldwide (19). TCGA played an essential catalytic and leadership role in oncology research.
The availability of large datasets from various omics levels enabled the use of computational tools, allowing an in-depth understanding of cancer at the system level. Challenges related to harmonization, volume, and complexity of data were, and still are, being addressed with modern data science techniques (19). This comprehensive genetic characterization of cancer informed the development of biomarkers and therapeutic targets, leading to the transition from a broad cytotoxic treatment to targeted therapies based on the molecular characteristics of individual tumors (18).
The successful integration of biomarkers, digital technologies, data science, and system biology principles shaped the landscape of PM in oncology (17). Following the success of PM in oncology, epilepsy research and clinical care took the first steps toward PM. Many conceptual and technical challenges faced by the epilepsy field coincide with those encountered by oncology.
The application of PM to epilepsy is limited by the complex pathophysiology and etiology of epilepsy and epileptogenesis and the interactions between pathological and physiological excitation and inhibition. Additionally, the dynamic nature of epileptiform activities over time, which includes both short-term fluctuations occurring within seconds to minutes around each seizure and the long-term changes that occur over several years can hinder the application of PM. Lastly, our limited understanding of the interactions among mechanisms at various scales—ranging from cell intrinsic changes to multiple cell types and their interaction among one another, and local neural-glial and long distance neural networks that involve multiple brain regions—adds complexity to the implementation of PM in epilepsy (20). Unlike oncology, where identified biomarkers have a direct relationship to the pathogenic mechanisms of cancer, in epilepsy, biomarkers may represent epiphenomena or be confused with the etiology of the diverse underlying mechanisms and different diseases causing epilepsies.
It is crucial to recognize these differences as they relate to our foundational understanding of epilepsy mechanisms that the pillars of precision epilepsy stand upon and not allow these complexities to hinder our progress in advancing the field.
The pillars of pm in epilepsyBiomarkers
Biomarkers are indicators of physiological and pathogenic processes or responses to an exposure or intervention (12, 21). Biomarker categories include susceptibility, diagnostic, monitoring, prognostic, predictive, pharmacodynamic treatment/response, and safety (Table 1) (21). In oncology, biomarkers such as genetic profiling and liquid biopsies revolutionized treatment decisions (22, 23). Continued advances in the development, validation, and application of diverse biomarkers are pivotal in PM (12, 24), as they bridge the gap between disease pathophysiology and clinical care.
Examples of different categories of biomarkers available in epilepsy care.
Category
Examples
Susceptibility/risk
Epileptiform abnormality as a susceptibility/risk biomarker for seizure recurrence after a single unprovoked seizure
Diagnostic
Genetic testing (e.g., SCN1A mutations as a diagnostic biomarker for Dravet syndrome, Neuroimaging tools, Multimodal imaging co-registration tools, Automated detection of brain abnormalities in MRI, Quantitative evaluation of Neuroimaging, invasive and non-invasive EEG data, Detection of autoantibodies in autoimmune epilepsies).
Monitoring
Blood levels of lamotrigine as a monitoring biomarker to adjust treatment dose during pregnancy, continuous EEG in status epilepticus, seizure detection devices.
Prognostic
MRI as a monitoring biomarker for patients with brain tumors
Predictive
Neurophysiological testing as a predictive biomarker to identify patients at risk for memory and language deficits following epilepsy surgery, multimodal imaging co-registration tools (the comparison between functional MRI results and structural abnormalities facilitates the prediction of potential postoperative deficits)
Pharmacodynamic response
EEG patterns as a pharmacodynamic biomarker to assess response to status epilepticus treatment
Safety
Human leukocyte antigen (HLA-B*15:02) allele as a genetic biomarker to identify patients at risk for hypersensitivity to carbamazepine
EEG, electroencephalography; HLA, human leukocyte antigen; MRI, magnetic resonance imaging.
An integrated multi-modality biomarker approach is crucial for the integrative holistic conception of epilepsy (24). Biomarkers, based on genetic, neuroimaging, electrophysiology, and body fluids assessments, were already incorporated into epilepsy clinical practice. We provide examples of biomarkers currently being used toward a more “personalized and sub-population tailored” clinical approach and discuss their growing role in the clinical implementation of PM. Integrating data from various modalities will enhance our understanding of the complexities of epilepsy in a more holistic way, allowing us to move from personalized clinical care to a PM approach involving individual level differences. We do not intend, however, to provide a comprehensive review of all available biomarkers currently being used and studied in epilepsies.
Genetic biomarkers
About 70 to 80% of epilepsies are estimated to have genetic variants underlying the disease (25). Clinical tests using comprehensive gene panels, exomes, or genomes (26) led to the identification of hundreds of genes associated with monogenic and polygenic epilepsies (27). These mutations affect the function of ion channels, neurotransmitter receptors, and other molecular components leading to seizures (Box 1).
Key take away points.
Epilepsies provide compelling opportunities for PM given their heterogeneity, measurable outcomes, known risk factors, and established animal models.
Advances in biomarkers, digital technologies, systems medicine, and data science are the key pillars of implementing PM in epilepsy.
By comparing the oncology and epilepsy fields, we can gain insights into implementing PM in epilepsy care successfully.
The path toward PM in epilepsy care lies in studies aimed at understanding systems-biology epileptogenesis using technological innovations.
PM has the potential to optimize epilepsy care in the predictive, preventive, personalized, and participatory dimensions as envisioned by the P4 paradigm.
The role of genetics in epilepsies is more complex than in other diseases since a similar clinical phenotype may originate from different genetic mechanisms (e.g., in West syndrome) (27, 28) and mutations in one gene can be associated with a wide range of phenotypes (e.g., in epilepsies mediated by mutations in the SCN2A gene) (26). Despite this high genetic heterogeneity and pleiotropy, significant strides were made to uncover novel genes associated with epilepsies (18).
Elucidating genetic profiles in epilepsy clinical practice advanced diagnosis, the development of tailored treatments, prediction of ASM response, drug metabolism, and risk of adverse events. The continued development and characterization of genetic models of epilepsies (in vivo and in vitro) can inform us about relevant epilepsy pathways and additional points of intervention.
Dravet syndrome (DS) is an example of PM being applied in epilepsy clinical care (Box 2). DS is an infantile-onset developmental and epileptic encephalopathy associated with pathogenic variants in the SCN1A gene, which encodes a subunit of a sodium channel in the brain (29, 30). A deeper understanding of the disease at a molecular level and the availability of confirmatory genetic biomarkers revolutionized the management of DS. A precise diagnosis offers the opportunity to customize treatment by avoiding ASMs with sodium channel blocking property that may worsen seizures and negatively impact cognitive development. It also allows genetic and prognosis counseling and enables approaches to prevent status epilepticus by providing guidance on antipyretics, vaccinations, and rescue medication (30).
Precision medicine in Dravet syndrome.
Our understanding of Dravet syndrome (DS) has gone from zero to a detailed, yet individualized, comprehension of its mechanism in less than 50 years (31). Clinically, DS is characterized by the intractable, prolonged seizures alongside fevers and cognitive impairment (32, 33). DS onset peaks in the first year of life. The association between DS and de novo pathogenic variants in the SCN1a gene was first reported in 2001 and has been repeatedly confirmed since then (32). Studies on how SCN1a mutations affect the structure and activity of the sodium channels enhanced our understanding of DS (33). Various SCN1a alterations correspond with different levels of loss-of-function, and less frequently, gain-of-function mutations in the sodium channel, explaining the wide variability in DS severity and therapeutic response to antiseizure medications (29, 34). Mutations on the SCN1A gene are often de novo and constitutional; however, they can also be inherited from a parent with a less severe phenotype or present as somatic mosaicism (33), further highlighting the complexity of genetics in epilepsies. This mechanistic understanding of DS allowed the establishment of animal models with specific mutations and thus offered opportunities for targeted treatments. Even though there is still a lot to be researched in DS, the acquired knowledge on the pathophysiology of DS, its complexity, and individual variability has already led to innovations in prevention and treatment strategies. The future of PM in epilepsy will likely follow a similar path (i.e., a new finding of the pathophysiological mechanism will lead to follow-up questions, which will eventually contribute to the decoding of epilepsy etiology).
Identifying SCN1A mutations also has implications for developing new treatments, with some drugs being tested and approved specifically for DS (30). It also allowed the development of disease-modifying therapies, such as genetic therapies and antisense oligonucleotides that restore the sodium channel, some of which are already being tested in clinical trials (30). DS exemplifies how genetic biomarkers can lead to a systems medicine level understanding of epilepsy pathology and transform clinical care. A PM approach to DS encompassing changes from genetic mutations to protein expression, channel functions, and environmental influences, has significantly altered how we comprehend, manage, and develop new treatments for patients with DS.
Genetic biomarkers in epilepsy can also be used to tailor treatment selection in clinical practice. For instance, the human leukocyte antigen (HLA-B*15:02) allele was found to be highly associated with carbamazepine-induced severe skin reactions in broad Asian populations. Carrying out HLA-B*15:02 allele genotyping helps guide treatment selection by identifying individuals with an increased risk of severe skin reactions to carbamazepine (35, 36).
Despite the high genetic heterogeneity and pleiotropy observed in epilepsies, similarly to oncology, genetic will play a major role toward advancing PM in epilepsy.
Neuroimaging biomarkers
Epilepsy surgery is one of the most compelling examples of how a multi-modal approach can improve epilepsy care. To achieve successful surgical outcomes and preserve eloquent brain areas, it is imperative to accurately identify the epileptogenic zone. Structural, functional and metabolic neuroimaging biomarkers including magnetic resonance imaging (MRI), fluorodeoxyglucose positron emission tomography (FDG-PET), and single-photon emission computerized tomography (SPECT), are often part of the epilepsy pre-surgical evaluations (12). The concordance of multimodality data, along with safety considerations, is critical in defining the extent of the surgical resection in epilepsy (37).
Technological advances have facilitated the interpretation and integration of multimodal data. For example, post-processing tools are used to geometrically align images so that corresponding pixels/voxels represent the same structure (38). Continued improvements in co-registration tools have allowed their use in epilepsy surgery evaluations enabling physicians to overlay images from different modalities to facilitate the evaluation of concordance between finding from different modalities. Co-registration tools also offer a unique opportunity to evaluate cross-sectional or longitudinal data with a high degree of precision (39).
The incorporation of new MRI sequences, such as arterial spin labeling, into presurgical planning can enhance patient prognostication beyond structural imaging (40). In addition, the use of translational molecular imaging with its variety of molecular probes [e.g., translocator protein 18 kDa (TSPO) PET ligands] has the potential to be leveraged for diagnostic imaging, patient stratification, therapy monitoring, and drug development as described in other pathologies (41). Improvements in imaging quality, such as higher field MRI, combined with machine learning (ML) tools applied to MRI have shown compelling results in detecting subtle malformations of cortical development (42, 43), while automatic quantitative assessments have enabled more accurate evaluations of structural abnormalities (39). Some of these tools are already available in clinical practice.
In epilepsies, certain features and patterns extracted from brain images can serve as biomarkers, providing valuable information about structure, function, and connectivity. Following the same path as oncology, neuroimaging tools can be utilized not only to define etiology but also hold significant potential to serve as biomarkers for PM care.
Electrophysiological biomarkers
Analyses of ictal and interictal electroencephalogram (EEG) patterns are critical for defining surgical care, diagnosis, treatment, and prognosis in epilepsies (37). EEG can access brain activity either in an invasive or non-invasive manner. Advances in invasive EEG monitoring exemplify how PM can be applied to epilepsy clinical practice. The development of a tailored electrode placement map, based on the convergence of multimodal assessment data, is crucial to successfully delineate the epileptogenic zone in surgical cases (37). By co-registering CT and MRI, physicians can visualize the precise location of the invasive electrode, thus facilitating the interpretation of invasive EEG data and allowing comparison of EEG results with the location of other structural and metabolic findings. Additional tools, including magnetoencephalography (MEG), electric, and magnetic source imaging will further improve epilepsy evaluation (44). In epilepsy clinical practice and research, improvements in the automatic analysis of EEG patterns and seizure detection are also providing valuable information about brain activity with the potential to advance precision epilepsy further. For instance, mathematical modeling has identified additional EEG patterns such as infra-slow (e.g., ictal Direct Current (DC) shifts, red slow) and high-frequency oscillations, which has refined the localization of the epileptogenic zone (45–47). In the field of seizure forecasting, data acquired from recording devices, such as deep brain stimulation (DBS) or responsive neurostimulation (RNS) devices, and the development of long-term EEG devices will enable novel biomarkers for a seizure forecasting (48). The detection of critical slowing down preceding seizures suggests that slow waves could be used as indicators in seizure forecast algorithms (49). In addition, studies on seizure distribution following the circadian, multiday, or seasonal rhythms further demonstrated the potential use of EEG data to predict seizure recurrence, especially when combined with other physiological measures and environmental influences (48). Continued advancement in the automated detection of EEG patterns hold significant potential for improving the field.
Over the past decade, lags in the analysis of EEG recordings come from the large data volume, confining the application in clinical practice. New tools, such as ML and mathematical models, will enable a broader use of EEG as a biomarker (50–52).
Bio fluids biomarkers
Body fluid matrices are attractive because they can illustrate biological changes in the brain while circumventing the blood–brain barrier and are a relatively inexpensive and minimally invasive source of additional biomarkers. Variations in inflammatory cytokines, redox states, hormone levels, and signaling pathways are correlated with neurologic dysfunction (12, 53).
In epilepsy research, there is a need to explore the use of fluid biomarkers (CSF and serum) in clinical care further. For example, high levels of neurofilament light chain (NfL) in status epilepticus represent a promising biomarker of seizure-related neuronal damage (54). Increased levels of certain proteins, such as S100B and glial fibrillary acidic protein (GFAP), in the blood or cerebrospinal fluid, may indicate neuronal damage and could be associated with epilepsy severity (55).
Advances in the field of autoimmune epilepsies exemplify the role of fluid biomarkers in PM and how they can transform clinical practice. The detection of autoantibodies in some types of epilepsies (such as the anti-NMDA receptor antibodies) allows classification of autoimmune epilepsies as clinically independent entities that can be managed with tailored treatments targeting the autoimmune response (56).
Digital technologies in precision epilepsy
Digital technology is defined as systems that use computing platforms, connectivity, software, and/or sensors for healthcare and related uses (57). There is rapid growth of clinical data in electronic health records and other health-related information databases via digital technologies, such as wearable devices, smartphones, and edge computing (17).
In oncology, digital solutions that have furthered PM include digital patient records, digital pathology, digital platforms for clinical trials, and the ability to share data digitally among healthcare providers and institutions. The same tools are available in the epilepsy field and are essential to foster collaboration between centers. In epilepsies, there are multiple applications of digital tools for PM, ranging from facilitating sharing of research data to developing devices to aid patient care.
Seizure detection devices exemplify how digital technologies can play a role in PM for epilepsy care and research. Generalized tonic–clonic seizures and focal to bilateral tonic–clonic seizures can now be accurately detected by wearable devices, lowering risk for injuries and sudden death (58, 59). Wearable devices for seizure detection were created based on the insights from the dynamics and rhythms of ictal and interictal activities acquired from EEG data (58) combined with known motor and autonomic features of seizures. Data collected by wearable devices, such as surface electrocardiogram, electromyography, and wrist accelerometer, allowed further advances in automated seizure detection (59). Commercially available video monitoring and infrared-based devices also have the potential to be used for seizure detection but need further development (60). Seizure detection devices can also increase accurate measurements of seizure outcomes (e.g., seizure frequency and duration); this information has typically been collected via patient-reported data, which can be inaccurate and biased (13, 61). Since epilepsy treatment is based mainly on seizure outcomes, this lack of precision directly affects clinical management.
Despite encouraging advances in the field, the sensitivity of seizure-detection devices is paramount to optimizing epilepsy management. False alarms, particularly for seizures without convulsions, can have a detrimental impact on patients with epilepsy and their caregivers. Well-designed studies are warranted to evaluate the effectiveness, usability, adoption, and clinical impact of these devices (59, 62). Digital devices also allow the collection of a diverse repertoire of disease-related phenotypes in a convenient, unobtrusive, and longitudinal manner (17) providing an opportunity to further advance the understanding of epilepsies, seizure patterns, and triggers.
Digital biomarkers in epilepsy have the potential to address several challenges in the field that significantly impact treatment decisions, including seizure prediction, detection, and quantification. The ongoing development and adoption of digital tools will continue to shape the landscape of PM, fostering collaborations and research breakthroughs.
Systems medicine
Systems medicine is a translational extension of systems biology. Systems biology refers to an integrative research approach designed to address the complexity of biological systems (63). The main principle of systems biology is that a simultaneous multitude of molecular interactions from various levels occurring at any one time, are combined in a holistic manner to produce a phenotype. A systems biology approach uses computational and mathematical tools to elucidate causative dynamics, intermediate endophenotypes, and/or clinical features of a condition (17, 63, 64).
In oncology, the systems biology approach adopting integrative omics enabled a deeper understanding of disease pathophysiology by taking into consideration genetic profiling, molecular interactions, signaling pathways, and regulatory networks (18). The addition of “omics” to a term implies a comprehensive, or global, assessment of a set of molecules or components within a particular biological system (65). Advances in omics technologies allowed researchers to integrate information from different levels (genetic, transcriptomic, proteomic, and metabolic) to build comprehensive models of biological systems (66). Systems biology provides the analytical and computational framework to integrate, interpret, and model omics data, enhancing our understanding of normal versus pathologic conditions (63). Insights gained from systems biology have furthered systems medicine in oncology by informing the development of new diagnostic and therapeutic strategies.
Like its application in oncology, systems medicine has the potential to decipher the complexity underlying epilepsies. Even though we have an incomplete understanding of many epilepsy syndromes at the systems level, there are some examples of how systems medicine has advanced epilepsy care.
Epilepsy associated with tuberous sclerosis complex (TSC) represents a unique model of disease that exemplifies how system medicine can improve care. TSC is a rare multisystem genetic disorder characterized by the development of tumors and caused by mutations in TSC1 or TSC2 genes, leading to dysregulation of the mammalian target of rapamycin (mTOR) pathway. Gene mutations in TSC have downstream effects on neurobiological systems in a dynamic and complex way that results in considerable phenotypic variability (67) with different degrees of developmental disabilities and seizure burden (67, 68). In-vitro and in-vivo studies have undercover the roles of TSC1 and TSC2 genes, which include tumor suppression, neuronal network development, morphology and function, oxidative stress, inflammation, and regulation of specific microRNAs in the neurological environment (67).
A system medicine approach continues to contribute to advances in the development of biomarkers and therapies for TSC and has led to improved targeted treatments (e.g., mTOR inhibitors and gene therapy), tailored ASM selection, and advances in epilepsy surgery (67). A recent trial showed that preventive treatment of TSC with ASM can modify the natural history of seizures (69), further emphasizing how PM has the potential to change epilepsy care.
By integrating various types of biological data and employing computational models, systems medicine will allow us to further understand and incorporate the heterogeneity and multi-factorial nature of epilepsy toward clinical care.
Data science in precision epilepsy
Data science is an interdisciplinary field that encompasses a range of techniques, including statistical analysis, ML, and data visualization, to uncover patterns, trends, and relationships in complex datasets from various domains (70).
Advances in biomarkers and digital health technology and the creation of epilepsy consortia, databases, and repositories generated a massive amount of data (12, 16, 71, 72). Despite being a challenge for traditional statistical approaches, the availability of comprehensive datasets provided an opportunity to incorporate complexity and use a system medicine approach to further our understanding of epilepsies.
Innovative analytical tools based on artificial intelligence (AI) are rapidly taking on a leading role in all fields of science and society and have the potential to revolutionize epilepsy research. For example, ML can identify patterns hidden in large-scale unprocessed data using supervised or unsupervised models (17, 71, 72). ML tools to predict drug treatment or surgical outcomes, or automatically detect MRI abnormalities are already being developed for epilepsy care (73). Deep learning, a subset of ML, employes multi-layered neuronal networks to analyze complex patterns, enabling the interpretation of data features and relationships (17). Unsupervised deep learning can reveal the pathological heterogeneity underlying epilepsies by clustering high-dimensional data and has recently been employed to classify EEG patterns for epilepsies (74, 75). Other computing innovations such as molecular dynamics modeling, whole-cell simulation, tailored animal models, and “digital twin,” provide a glimpse into the potential clinical applications of AI-based PM in epilepsy research (76–78).
As data availability progressively increases, the limited power of classical computing will significantly hinder the analysis of large datasets in epilepsy research (79). Quantum computing provides a potential solution with the ability to handle complex analytical demands in shortened times (80). Enhanced ML models allow integration of sparse and noisy data to overcome limitations imposed by traditional approaches. Interpretability of findings, validity of data outputs, and ethical concerns are some barriers to implementation of AI in epilepsy and in the healthcare industry. Together, data science and AI will broaden the range of research tools to better pinpoint the mechanisms underlying epileptogenesis, and help develop disease-modifying, individualized treatments (13).
Discussion
As illustrated by the examples above, the pillars of PM in epilepsy have already been established. The convergence and integration of the pillars of PM will forge a robust foundation, enabling the field to thrive and advance. By examining the incorporation of PM in epilepsy, we can identify variances and parallels between oncology and epilepsy and gain insights on how to overcome barriers and successfully implement PM in epilepsy care.
Following the knowledge acquired from precision oncology, there is a need to further integrate data from different disease measurement modalities and shift from the single-center study paradigm to a large interdisciplinary multi-center collaborative effort to fully understand epilepsies. To break down information silos and foster deeper collaboration, updates on health information policies and laws, and the development of clear rules for collaboration are needed. Concerns related to data sharing can be overcome by improvements in de-identification technology, data homogenization, and the secure transfer of medical data or federated learning (12, 19, 81). Further advancements and data integration across the pillars of PM in epilepsy are crucial for its full implementation.
Incorporating PM in epilepsy presents a set of unique challenges. In contrast to oncology, where abnormal tissues and cells are accessible for analysis, in epilepsies, such histological data is accessible in only a small proportion of epilepsy cases. Furthermore, while oncology treatment outcomes can be objectively measured by tumor growth and mortality, epilepsy outcomes rely mostly on seizure frequency, a subjective measurement. Outcomes based on seizure-count are not always precise due to the episodic nature of seizures, the reliance on the caregiver’s description, and the wide variability of seizure patterns (61). Advances in digital technologies and biomarkers can help overcome this barrier.
Differently from oncology, epilepsies are a group of biologically complex, heterogeneous central nervous system disorders that can occur due to an extensive range of potentially overlapping etiologies (e.g., genetic, brain insults, metabolic, immune, and idiopathic). Seizures can occur as the main feature of an epilepsy syndrome (e.g., juvenile myoclonic epilepsy), or they can be a component of a more complex disease (e.g., developmental and epileptic encephalopathies) (82, 83). Additionally, epilepsies frequently co-occur with other conditions, such as mood disorders, sleep disturbances, and cognitive impairments (84). In other words, epilepsy is not a single disease, but a spectrum of disorders characterized by recurrent seizures, further emphasizing the extent of its complexity. The gradual evolution of PM will likely follow a self-fueled trajectory characterized by continuous integration between the four pillars and a dynamic and interconnected network in which advancements in one area contribute to improvement in others. As the field matures, these feedback loops and interdependencies are expected to drive continuous progress, making PM an increasingly powerful and integral part of healthcare. This integrated framework has the potential to gradually optimize clinical care in the predictive, preventive, personalized, and participatory dimensions across a variety of epilepsy syndromes as envisioned by the P4 paradigm (85) (Figure 1).
The shifting landscape toward precision epilepsy care. Epilepsies are a group of biologically complex and heterogeneous disorders. With the availability of novel biomarkers and digital tools, combined with advances in computational and mathematical tools (data science), there is an opportunity to integrate and analyze multimodal data at all levels. The systems medicine approach offers great potential to overcome the complexity of phenotypic heterogeneity and the multifactorial nature of epilepsy. A holistic understanding of epilepsies will ultimately lead to optimized clinical care, moving us toward precision medicine and allowing us to fully implement the P4 paradigm: predictive, preventive, participatory, and personalized care. CT, computed tomography; DTI, diffusion tensor imaging; EEG, electroencephalography. fMRI, functional magnetic resonance imaging; MRI, magnetic resonance imaging; PET, positron emission tomography; SPECT, single-photon emission computerized tomography.
Diagram depicting the multifactorial nature of epilepsy, highlighting key factors that influence its pathophysiology, including brain structure, brain function, genetics, environment, demographics, and individual characteristics. Each factor can be assessed through different biomarkers and digital technologies such as MRI and genetic tests. The analysis and integration of these multimodal data, facilitated by advances in data science, will allow us to conduct a comprehensive analysis of patients within a systems medicine framework. This holistic approach will optimize clinical care, advance precision medicine, and enable implementation of the four-pillar framework: predictive, participatory, preventive, and personalized clinical care.
Several factors will foster further development of PM to improve the quality of care in epilepsies, with education and participation of healthcare providers being critical to this endeavor. For the full implementation of PM, health information policies and laws must be updated to ensure the safe use of AI across healthcare sectors and provide appropriate insurance coverage for broad access to biomarker tests and digital health services. Ultimately, the goal of PM is to help patients with epilepsy achieve improved clinical outcomes and reduced socioeconomic burden.
According to the Gartner Hype Cycle, emerging technologies follow five phases: technology trigger, peak of inflated expectation, trough of disillusionment, slope of enlightenment, and plateau of productivity (13). We are hopefully moving toward the slope of enlightenment to bring PM in epilepsy to fruition soon.
Conclusion
This narrative review provides an overview of recent achievements and existing challenges for the implementation of PM in epilepsy. In both epilepsy and oncology fields, the availability of data collected from validated biomarkers and digital technologies can inform the system medicine approach and, with innovations in the data science field, advance the development of PM-based clinical care. The complexity and diversity of epilepsy pathophysiology, challenges in outcome measurements, and the diversity of treatment options (leading to a lack of abnormal tissue samples for analysis) contribute to the delayed implementation of PM in comparison to oncology.
To advance PM in epilepsy, we need to shift from the traditional clinical phenotype-oriented mindset to a biomarker-guided multimodal and clinical-biological model encompassing the multifactorial nature of epilepsy across stages. New exploratory, hypothesis-independent (multi-) omics technologies combined with computational (multi-modal) neuroimaging and neurophysiology advances can identify and validate additional biomarkers and harness the advantages offered by systems medicine to help implement PM-based prevention, early detection, diagnosis, classification, and treatment, resembling the progress observed in oncology.
The author(s) declare that financial support was received for the research and/or publication of this article. This research was funded by Eisai Inc., USA, which provided financial support for the publication fees and medical writing assistance associated with this manuscript.
We want to express our sincere gratitude to Dr. David B Goldstein for revising the article and providing invaluable contributions. We thank Helen Hu and Christina Castelino for providing support and insightful comments throughout this project. Medical writing support, under the supervision and direction of the authors, was provided by Can Huang, PhD, on behalf of CMC AFFINITY, a division of IPG Health Medical Communications, funded by Eisai Inc., in accordance with Good Publication Practice (GPP 2022) guidelines.
Conflict of interest
HH is an employee of Eisai and serves as reviewing editor for the Journal Alzheimer’s & Dementia. HH is inventor of 11 patents and has received no royalties for: In Vitro Multiparameter Determination Method for The Diagnosis and Early Diagnosis of Neurodegenerative Disorders patent no. 8916388; In Vitro Procedure for Diagnosis and Early Diagnosis of Neurodegenerative Diseases patent no. 8298784; Neurodegenerative Markers for Psychiatric Conditions publication no. 20120196300; In Vitro Multiparameter Determination Method for The Diagnosis and Early Diagnosis of Neurodegenerative Disorders publication no. 20100062463; In Vitro Method for The Diagnosis and Early Diagnosis of Neurodegenerative Disorders publication no. 20100035286; In Vitro Procedure for Diagnosis and Early Diagnosis of Neurodegenerative Diseases publication no. 20090263822; In Vitro Method for The Diagnosis of Neurodegenerative Diseases patent no. 7547553; CSF Diagnostic in Vitro Method for Diagnosis of Dementias and Neuroinflammatory Diseases publication no. 20080206797; In Vitro Method for The Diagnosis of Neurodegenerative Diseases publication no. 20080199966; Neurodegenerative Markers for Psychiatric Conditions publication no. 20080131921; Method for diagnosis of dementias and neuroinflammatory diseases based on an increased level of procalcitonin in cerebrospinal fluid: US patent no. 10921330. MM-S is a full-time employee of Eisai. ET has received consultancy fees from Arvelle Therapeutics, Argenx, Clexio, Celegene, UCB Pharma, Eisai, Epilog, Bial, Medtronic, Everpharma, Biogen, Takeda, Liva-Nova, Newbridge, Sunovion, GW Pharmaceuticals, and Marinus; speaker fees from Arvelle Therapeutics, Bial, Biogen, Böhringer Ingelheim, Eisai, Everpharma, GSK, GW Pharmaceuticals, Jazz Pharmaceuticals, Hikma, Liva-Nova, Newbridge, Novartis, Sanofi, Sandoz, STOKE Therapeutics, and UCB Pharma; research funding (directly, or to his institution) from GSK, Biogen, Eisai, Novartis, Red Bull, Bayer, and UCB Pharma all outside the submitted work. Eugen Trinka has received research grants from Austrian Science Fund (FWF), Österreichische Nationalbank, and the European Union. Eugen Trinka is Co-founder of PrevEp Inc. and the CEO of Neuroconsult Ges.m.b. H. ET is co-director of the European Consortium on Epilepsy Trials (ECET). AI belongs to the Department of Epilepsy, Movement Disorders, and Physiology. Between 2018 and 2023, this department is the Industry-Academia Collaboration Courses supported by Eisai, Nihon Kohden, Otsuka Pharmaceutical, and UCB Japan, and since 2023, supported by Sumitomo Pharma, Nihon Kohden Ltd., and RICOH Company. PK and/or his institution has received research support or consultancy fees from Angelini, Eisai, Jazz Pharmaceuticals, LivaNova, SK Life Science and UCB Pharma. MC is a full-time employee of Eisai.
The author(s) declared that they were an editorial board member of Frontiers, at the time of submission. This had no impact on the peer review process and the final decision.
Generative AI statement
The authors declare that Generative AI was used in the creation of this manuscript. Grammarly was used in the preparation of this manuscript for copy editing purposes.
Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.
Publisher’s note
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.
ReferencesFisherRSAcevedoCArzimanoglouABogaczACrossJHElgerCE. Ilae official report: a practical clinical definition of epilepsy. Epilepsia. (2014) 55:475–82. doi: 10.1111/epi.12550, PMID: 24730690Group GBDNDC. Global, regional, and national burden of neurological disorders during 1990-2015: a systematic analysis for the global burden of disease study 2015. Lancet Neurol. (2017) 16:877–97. doi: 10.1016/S1474-4422(17)30299-5, PMID: 28931491GrisoldWFreedmanMGouiderRGuekhtALewisSMedinaM. The Intersectoral global action plan (Igap): a unique opportunity for neurology across the globe. J Neurol Sci. (2023) 449:120645. doi: 10.1016/j.jns.2023.120645, PMID: 37062176FattorussoAMatricardiSMencaroniEDell’IsolaGBDi CaraGStrianoP. The Pharmacoresistant epilepsy: an overview on existent and new emerging therapies. Front Neurol. (2021) 12:674483. doi: 10.3389/fneur.2021.674483, PMID: 34239494TrinkaEBrigoF. Antiepileptogenesis in humans: disappointing clinical evidence and ways to move forward. Curr Opin Neurol. (2014) 27:227–35. doi: 10.1097/WCO.0000000000000067, PMID: 24556736KoeppMTrinkaELoscherWKleinP. Prevention of Epileptogenesis - are we there yet?Curr Opin Neurol. (2024). doi: 10.1097/WCO.0000000000001256, PMID: 38345421LoscherWKleinP. The pharmacology and clinical efficacy of Antiseizure medications: from bromide salts to Cenobamate and beyond. CNS Drugs. (2021) 35:935–63. doi: 10.1007/s40263-021-00827-8, PMID: 34145528DevinskyOVezzaniAO’BrienTJJetteNSchefferIEde CurtisM. Epilepsy. Nat Rev Dis Primers. (2018) 4:18024. doi: 10.1038/nrdp.2018.24KearneyHByrneSCavalleriGLDelantyN. Tackling epilepsy with high-definition precision medicine: a review. JAMA Neurol. (2019) 76:1109–16. doi: 10.1001/jamaneurol.2019.2384, PMID: 31380988HogeveenLBoonPMertensAVerhagenLVonckK. Therapeutic use of transcranial ultrasound for epilepsy: a review. Heliyon. (2025) 11:e43001. doi: 10.1016/j.heliyon.2025.e43001LercheH. Drug-resistant epilepsy - time to target mechanisms. Nat Rev Neurol. (2020) 16:595–6. Epub 2020/10/08. doi: 10.1038/s41582-020-00419-y, PMID: 33024326SimonatoMAgostonDVBrooks-KayalADullaCFuremanBHenshallDC. Identification of clinically relevant biomarkers of Epileptogenesis - a strategic roadmap. Nat Rev Neurol. (2021) 17:231–42. doi: 10.1038/s41582-021-00461-4, PMID: 33594276JosephsonCBWiebeS. Precision medicine: academic dreaming or clinical reality?Epilepsia. (2021) 62:S78–89. doi: 10.1111/epi.16739CollinsFSVarmusH. A new initiative on precision medicine. N Engl J Med. (2015) 372:793–5. doi: 10.1056/NEJMp1500523, PMID: 25635347McAlisterFALaupacisAArmstrongPW. Finding the right balance between precision medicine and personalized care. CMAJ. (2017) 189:E1065–8. doi: 10.1503/cmaj.170107, PMID: 28827437Epi PMC. A roadmap for precision medicine in the epilepsies. Lancet Neurol. (2015) 14:1219–28. doi: 10.1016/S1474-4422(15)00199-4HampelHGaoPCummingsJToschiNThompsonPMHuY. The foundation and architecture of precision medicine in neurology and psychiatry. Trends Neurosci. (2023) 46:176–98. doi: 10.1016/j.tins.2022.12.004, PMID: 36642626DuggerSAPlattAGoldsteinDB. Drug development in the era of precision medicine. Nat Rev Drug Discov. (2018) 17:183–96. doi: 10.1038/nrd.2017.226, PMID: 29217837GaoGFParkerJSReynoldsSMSilvaTCWangLBZhouW. Before and after: comparison of legacy and harmonized Tcga genomic data commons’ data. Cell Syst. (2019) 9:24–34.e10. doi: 10.1016/j.cels.2019.06.006, PMID: 31344359StoberTMBatulinDTrieschJNarayananRJedlickaP. Degeneracy in epilepsy: multiple routes to Hyperexcitable brain circuits and their repair. Commun Biol. (2023) 6:479. doi: 10.1038/s42003-023-04823-0, PMID: 37137938FDA-NIH Biomarker Working Group. Best (Biomarkers, Endpoints, and Other Tools) Resource. Silver Spring, MD: Food and Drug Administration (US) (2016).ConnalSCameronJMSalaABrennanPMPalmerDSPalmerJD. Liquid biopsies: the future of cancer early detection. J Transl Med. (2023) 21:118. doi: 10.1186/s12967-023-03960-8, PMID: 36774504MarcinakCTMurtazaMWilkeLG. Genomic profiling and liquid biopsies for breast Cancer. Surg Clin North Am. (2023) 103:49–61. doi: 10.1016/j.suc.2022.08.003, PMID: 36410353EngelJJrPitkanenALoebJADudekFEBertramEH3rdColeAJ. Epilepsy biomarkers. Epilepsia. (2013) 54:61–9. doi: 10.1111/epi.12299, PMID: 23909854BlazekovicAGotovac JercicKMeglajSDuranovicVPrpicILozicB. Genetics of pediatric epilepsy: next-generation sequencing in clinical practice. Genes. (2022) 13:466. doi: 10.3390/genes13081466, PMID: 36011376BayatABayatMRubboliGMollerRS. Epilepsy syndromes in the first year of life and usefulness of genetic testing for precision therapy. Genes. (2021) 12:1051. doi: 10.3390/genes12071051, PMID: 34356067GuerriniRContiVMantegazzaMBalestriniSGalanopoulouASBenfenatiF. Developmental and epileptic encephalopathies: from genetic heterogeneity to phenotypic continuum. Physiol Rev. (2023) 103:433–513. doi: 10.1152/physrev.00063.2021, PMID: 35951482RastinCSchenkelLCSadikovicB. Complexity in genetic epilepsies: a comprehensive review. Int J Mol Sci. (2023) 24:4606. doi: 10.3390/ijms241914606, PMID: 37834053LoscherWPotschkaHSisodiyaSMVezzaniA. Drug resistance in epilepsy: clinical impact, potential mechanisms, and new innovative treatment options. Pharmacol Rev. (2020) 72:606–38. doi: 10.1124/pr.120.019539, PMID: 32540959WirrellECHoodVKnuppKGMeskisMANabboutRSchefferIE. International consensus on diagnosis and Management of Dravet Syndrome. Epilepsia. (2022) 63:1761–77. doi: 10.1111/epi.17274, PMID: 35490361DravetC. Dravet syndrome history. Dev Med Child Neurol. (2011) 53:1–6. doi: 10.1111/j.1469-8749.2011.03964.xPeruccaPBahloMBerkovicSF. The genetics of epilepsy. Annu Rev Genomics Hum Genet. (2020) 21:205–30. doi: 10.1146/annurev-genom-120219-074937MeiDCeticaVMariniCGuerriniR. Dravet syndrome as part of the clinical and genetic spectrum of sodium channel epilepsies and encephalopathies. Epilepsia. (2019) 60:S2–7. doi: 10.1111/epi.16054, PMID: 31904125SchefferIENabboutR. Scn1a-related phenotypes: epilepsy and beyond. Epilepsia. (2019) 60 Suppl 3:S17–24. doi: 10.1111/epi.16386, PMID: 31904117TangamornsuksanWChaiyakunaprukNSomkruaRLohitnavyMTassaneeyakulW. Relationship between the Hla-B*1502 allele and carbamazepine-induced Stevens-Johnson syndrome and toxic epidermal necrolysis: a systematic review and Meta-analysis. JAMA Dermatol. (2013) 149:1025–32. doi: 10.1001/jamadermatol.2013.4114, PMID: 23884208ManCBKwanPBaumLYuELauKMChengAS. Association between HLA-B*1502 allele and antiepileptic drug-induced cutaneous reactions in Han Chinese. Epilepsia. (2007) 48:1015–8. doi: 10.1111/j.1528-1167.2007.01022.x, PMID: 17509004PapadelisCPerryMS. Localizing the epileptogenic zone with novel biomarkers. Semin Pediatr Neurol. (2021) 39:100919. doi: 10.1016/j.spen.2021.100919, PMID: 34620466RuberTDavidBElgerCE. Mri in epilepsy: clinical standard and evolution. Curr Opin Neurol. (2018) 31:223–31. doi: 10.1097/WCO.0000000000000539, PMID: 29389747BernasconiABernasconiN. The role of MRI in the treatment of drug-resistant focal epilepsy. Eur Neurol. (2022) 85:333–41. doi: 10.1159/000525262, PMID: 35705017GennariAGGaitoLCserpanDKottkeRKrayenbuhlNRueggerA. Brain perfusion imaging by arterial spin labeling predicts postsurgical seizure freedom in pediatric focal Lesional epilepsy: a pilot study. Epilepsia. (2025) 66:2353–64. doi: 10.1111/epi.18375, PMID: 40099955HaiderAElghazawyNHDawoudAGebhardCWichmannTSipplW. Translational molecular imaging and drug development in Parkinson’s disease. Mol Neurodegener. (2023) 18:11. doi: 10.1186/s13024-023-00600-z, PMID: 36759912CendesF. Neuroimaging in investigation of patients with epilepsy. Continuum. (2013) 19:623–42. doi: 10.1212/01.CON.0000431379.29065.d3, PMID: 23739101GillRSLeeHMCaldairouBHongSJBarbaCDeleoF. Multicenter validation of a deep learning detection algorithm for focal cortical dysplasia. Neurology. (2021) 97:e1571–82. doi: 10.1212/WNL.0000000000012698, PMID: 34521691TrinkaEKoeppMKalssGKobulashviliT. Evidence based noninvasive presurgical evaluation for patients with drug resistant epilepsies. Curr Opin Neurol. (2024) 37:141–51. doi: 10.1097/WCO.0000000000001253, PMID: 38334495IkedaATakeyamaHBernardCNakataniMShimotakeADaifuM. Active direct current (dc) shifts and “red slow”: two new concepts for seizure mechanisms and identification of the epileptogenic zone. Neurosci Res. (2020) 156:95–101. doi: 10.1016/j.neures.2020.01.014, PMID: 32045575Andrade MachadoROttersonSE. Infralow Activity on Intracranial Eeg: A Systematic Review: Characteristics, Recording Methods and Predictive Value of the Zone to Remove. Clin EEG Neurosci. (2025) 2:15500594251336845. doi: 10.1177/15500594251336845, PMID: 34620466NakataniMInouchiMDaifu-KobayashiMMuraiTTogawaJKajikawaS. Ictal direct current shifts contribute to defining the Core ictal focus in epilepsy surgery. Brain Commun. (2022) 4:222. doi: 10.1093/braincomms/fcac222, PMID: 36381989StirlingRECookMJGraydenDBKarolyPJ. Seizure forecasting and cyclic control of seizures. Epilepsia. (2021) 62:S2–S14. doi: 10.1111/epi.16541, PMID: 32712968MaturanaMIMeiselCDellKKarolyPJD’SouzaWGraydenDB. Critical slowing down as a biomarker for seizure susceptibility. Nat Commun. (2020) 11:2172. doi: 10.1038/s41467-020-15908-3, PMID: 32358560GersterMTaherHSkochAHlinkaJGuyeMBartolomeiF. Patient-specific network connectivity combined with a next generation neural mass model to test clinical hypothesis of seizure propagation. Front Syst Neurosci. (2021) 15:675272. doi: 10.3389/fnsys.2021.675272, PMID: 34539355IkedaA. Epilepsy research in 2021: multidisciplinary achievements. Lancet Neurol. (2022) 21:8–10. doi: 10.1016/S1474-4422(21)00421-X, PMID: 34942141ValerianiDO’FlynnLCWorthleyASichaniAHSimonyanK. Multimodal collaborative brain-computer interfaces aid human-machine team decision-making in a pandemic scenario. J Neural Eng. (2022) 19:96a5. doi: 10.1088/1741-2552/ac96a5RavizzaTOnatFYBrooks-KayalARDepaulisAGalanopoulouASMazaratiA. Wonoep appraisal: biomarkers of epilepsy-associated comorbidities. Epilepsia. (2017) 58:331–42. doi: 10.1111/epi.13652, PMID: 28035782GiovanniniGBedinRFerraroDVaudanoAEMandrioliJMelettiS. Serum Neurofilament light as biomarker of seizure-related neuronal injury in status epilepticus. Epilepsia. (2022) 63:e23–9. doi: 10.1111/epi.17132, PMID: 34806176GiovanniniGMelettiS. Fluid biomarkers of neuro-glial injury in human status epilepticus: a systematic review. Int J Mol Sci. (2023) 24:519. doi: 10.3390/ijms241512519, PMID: 37569895LeviteMGoldbergH. Autoimmune epilepsy - novel multidisciplinary analysis, discoveries and insights. Front Immunol. (2021) 12:762743. doi: 10.3389/fimmu.2021.762743, PMID: 35095841U.S. FDA. (2023). Digital health technologies for remote data acquisition in clinical investigations: guidance for industry, investigators, and other stakeholders. Available online at: https://www.fda.gov/regulatory-information/search-fda-guidance-documents/digital-health-technologies-remote-data-acquisition-clinical-investigations. [Accessed April 4, 2025].BrinkmannBHKarolyPJNurseESDumanisSBNasseriMVianaPF. Seizure diaries and forecasting with wearables: epilepsy monitoring outside the clinic. Front Neurol. (2021) 12:690404. doi: 10.3389/fneur.2021.690404, PMID: 34326807BeniczkySWiebeSJeppesenJTatumWOBrazdilMWangY. Automated seizure detection using wearable devices: a clinical practice guideline of the international league against epilepsy and the International Federation of Clinical Neurophysiology. Clin Neurophysiol. (2021) 132:1173–84. doi: 10.1016/j.clinph.2020.12.009, PMID: 33678577JoryCShankarRCokerDMcLeanBHannaJNewmanC. Safe and sound? A systematic literature review of seizure detection methods for personal use. Seizure. (2016) 36:4–15. doi: 10.1016/j.seizure.2016.01.013, PMID: 26859097CookMJO’BrienTJBerkovicSFMurphyMMorokoffAFabinyiG. Prediction of seizure likelihood with a long-term, implanted seizure advisory system in patients with drug-resistant epilepsy: a first-in-man study. Lancet Neurol. (2013) 12:563–71. doi: 10.1016/S1474-4422(13)70075-9, PMID: 23642342NaganurVSivathambooSChenZKusmakarSAntonic-BakerAO’BrienTJ. Automated seizure detection with noninvasive wearable devices: a systematic review and Meta-analysis. Epilepsia. (2022) 63:1930–41. doi: 10.1111/epi.17297, PMID: 35545836AuffrayCChenZHoodL. Systems medicine: the future of medical genomics and healthcare. Genome Med. (2009) 1:2. doi: 10.1186/gm2, PMID: 19348689MahoneyJMMillsJDMuhlebnerANoebelsJPotschkaHSimonatoM. 2017 Wonoep appraisal: studying epilepsy as a network disease using systems biology approaches. Epilepsia. (2019) 60:1045–53. doi: 10.1111/epi.15216, PMID: 31087652HasinYSeldinMLusisA. Multi-omics approaches to disease. Genome Biol. (2017) 18:83. doi: 10.1186/s13059-017-1215-1, PMID: 28476144BabuMSnyderM. Multi-omics profiling for health. Mol Cell Proteomics. (2023) 22:100561. doi: 10.1016/j.mcpro.2023.100561, PMID: 37119971CuratoloPSpecchioNAronicaE. Advances in the genetics and neuropathology of tuberous sclerosis complex: edging closer to targeted therapy. Lancet Neurol. (2022) 21:843–56. doi: 10.1016/S1474-4422(22)00213-7, PMID: 35963265IhnenSKZCapalJKHornPSGriffithMSahinMBebinEM. Epilepsy is heterogeneous in early-life tuberous sclerosis complex. Pediatr Neurol. (2021) 123:1–9. doi: 10.1016/j.pediatrneurol.2021.06.012, PMID: 34343869KotulskaKKwiatkowskiDJCuratoloPWeschkeBRineyKJansenF. Prevention of epilepsy in infants with tuberous sclerosis complex in the Epistop trial. Ann Neurol. (2021) 89:304–14. doi: 10.1002/ana.25956, PMID: 33180985SubrahmanyaSVGShettyDKPatilVHameedBMZPaulRSmritiK. The role of data science in healthcare advancements: applications, benefits, and future prospects. Ir J Med Sci. (2022) 191:1473–83. doi: 10.1007/s11845-021-02730-z, PMID: 34398394ThompsonPMJahanshadNChingCRKSalminenLEThomopoulosSIBrightJ. Enigma and global neuroscience: a decade of large-scale studies of the brain in health and disease across more than 40 countries. Transl Psychiatry. (2020) 10:100. doi: 10.1038/s41398-020-0705-1, PMID: 32198361SmolyanskyEDHakeemHGeZChenZKwanP. Machine learning models for decision support in epilepsy management: a critical review. Epilepsy Behav. (2021) 123:108273. doi: 10.1016/j.yebeh.2021.108273, PMID: 34507093HakeemHFengWChenZChoongJBrodieMJFongSL. Development and validation of a deep learning model for predicting treatment response in patients with newly diagnosed epilepsy. JAMA Neurol. (2022) 79:986–96. doi: 10.1001/jamaneurol.2022.2514, PMID: 36036923NhuDJanmohamedMAntonic-BakerAPeruccaPO’BrienTJGilliganAK. Deep learning for automated epileptiform discharge detection from scalp EEG: a systematic review. J Neural Eng. (2022) 19:9644. doi: 10.1088/1741-2552/ac9644JanmohamedMNhuDKuhlmannLGilliganATanCWPeruccaP. Moving the field forward: detection of epileptiform abnormalities on scalp electroencephalography using deep learning-clinical application perspectives. Brain Commun. (2022) 4:fcac218. doi: 10.1093/braincomms/fcac218, PMID: 36092304FeigMSugitaY. Whole-cell models and simulations in molecular detail. Annu Rev Cell Dev Biol. (2019) 35:191–211. doi: 10.1146/annurev-cellbio-100617-062542, PMID: 31299173KuhlmanBBradleyP. Advances in protein structure prediction and design. Nat Rev Mol Cell Biol. (2019) 20:681–97. doi: 10.1038/s41580-019-0163-x, PMID: 31417196DepannemaeckerDCarluMBouteJDestexheA. A model for the propagation of seizure activity in Normal brain tissue. eNeuro. (2022) 9:ENEURO.0234–21.2022. doi: 10.1523/ENEURO.0234-21.2022, PMID: 36323513LeeJGJunSChoYWLeeHKimGBSeoJB. Deep learning in medical imaging: general overview. Korean J Radiol. (2017) 18:570–84. doi: 10.3348/kjr.2017.18.4.570, PMID: 28670152BacherDAminiAFriedmanDDoyleWPaciaSKuznieckyR. Validation of an Eeg seizure detection paradigm optimized for clinical use in a chronically implanted subcutaneous device. J Neurosci Methods. (2021) 358:109220. doi: 10.1016/j.jneumeth.2021.109220, PMID: 33971201RiekeNHancoxJLiWMilletariFRothHRAlbarqouniS. The future of digital health with federated learning. NPJ Digit Med. (2020) 3:119. doi: 10.1038/s41746-020-00323-1, PMID: 33015372FisherRSCrossJHFrenchJAHigurashiNHirschEJansenFE. Operational classification of seizure types by the international league against epilepsy: position paper of the Ilae Commission for Classification and Terminology. Epilepsia. (2017) 58:522–30. doi: 10.1111/epi.13670, PMID: 28276060SchefferIEBerkovicSCapovillaGConnollyMBFrenchJGuilhotoL. Ilae classification of the epilepsies: position paper of the Ilae commission for classification and terminology. Epilepsia. (2017) 58:512–21. doi: 10.1111/epi.13709, PMID: 28276062HermannBPStruckAFBuschRMReyesAKaestnerEMcDonaldCR. Neurobehavioural comorbidities of epilepsy: towards a network-based precision taxonomy. Nat Rev Neurol. (2021) 17:731–46. doi: 10.1038/s41582-021-00555-z, PMID: 34552218HoodLBallingRAuffrayC. Revolutionizing medicine in the 21st century through systems approaches. Biotechnol J. (2012) 7:992–1001. doi: 10.1002/biot.201100306, PMID: 22815171GlossaryAI
Artificial intelligence
ASM
Anti-seizure medication
DBS
Deep brain stimulation
DC
Direct Current
DS
Dravet syndrome
EEG
Electroencephalogram
FDG-PET
Fluorodeoxyglucose positron emission tomography
GFAP
Glial fibrillary acidic protein
HLA
Human leukocyte antigen
MEG
Magnetoencephalography
ML
Machine learning
MRI
Magnetic resonance imaging
mTOR
mechanistic target of rapamycin
NfL
Neurofilament light chain
PM
Precision medicine
RNS
Responsive neurostimulation
SPECT
Single-photon emission computerized tomography
TCGA
The Cancer Genome Atlas
TSC
Tuberous sclerosis complex
TSPO
Translocator protein 18 kDa
WHO-IGAP
World Health Organization’s Intersectoral Global Action Plan