Ley Sander and John S. Duncan Pragmatically, different perspectives shape future expectations, which depend primarily on the roles of stakeholders. People with epilepsy, their families and carers, hope for a timely and accurate diagnosis, effective treatment at the earliest opportunity, and a life without seizures.1 Many people with active epilepsy have an overwhelming wish for seizure freedom to obtain driving privileges. For clinicians, early diagnosis and treatment are also high on the wish list. Scientists aim for a comprehensive understanding of processes that give rise to seizures so that antiepileptogenic treatment can be delivered at an early stage to those at risk. For society, the reduction of the burden imposed by epilepsy is at the forefront. The hope for effective prevention resonates across all stakeholders. Prevention of sudden unexpected death in epilepsy (SUDEP) is an objective for all. Primary and secondary prevention of epilepsy, or rapidly achieving long‐term remission, would make significant inroads into SUDEP incidence. Treating epilepsy, currently, can be frustrating for all involved, as it is difficult to predict treatment outcomes at the onset.1,2 Frustration tends to increase when a therapy is ineffective. The prevailing anti‐epilepsy treatment paradigm is empirical and mainly symptomatic, focusing on suppressing seizures rather than stopping the underlying process.3 Clinicians try anti‐seizure medications (ASM) in sequence, hoping that one will work. The first ASM works in about half of people, regardless of which ASM is used. This ‘trial‐and‐error’ approach does not make the treatment rational and drives frustrations. Non‐ASM options, such as surgical treatment, for instance, may be curative through targeted resections.3 However, only about three per cent of incident cases are candidates, and outcomes are not always predictable.4 Refining surgical targeting and minimising invasiveness could improve outcomes. Epilepsy is often a result of an insult to the brain.3 Despite the potential benefits, ongoing programmes for effective primary epilepsy prevention are lacking.5 Primary prevention comprises the avoidance of causative insults. Secondary prevention involves therapy to someone at risk of developing epilepsy after an initial event. Tertiary prevention delivers treatment after an epileptogenic substrate is established, but before seizures develop.6 Risk factors amenable to primary prevention are mainly perinatal and obstetric complications, cerebrovascular events, traumatic brain injuries (TBI), CNS infections and infestations. These risk factors account for up to a quarter of epilepsy cases in many series. Perinatal events and TBI are the top risk factors for epilepsy in low‐ and medium‐income countries. In high‐income countries, strokes are the leading preventable risk factor for epilepsy, accounting for over half of new‐onset epilepsy in older adults. In endemic areas, brain infestations with Taenia solium leading to neurocysticercosis are the most significant risk factor for focal epilepsy. Feasible primary interventions include improving prenatal care, vaccination programmes for meningitis, and reducing vascular disease risk factors to prevent stroke. Providing impregnated bed nets to prevent cerebral malaria and improving basic sanitation to eliminate neurocysticercosis would significantly decrease the worldwide burden of epilepsy.5 Secondary and tertiary prevention are attractive options, but there is no established knowledge base to support their deployment.5,6 Thus, any current prospects for preventing epilepsy rely on primary prevention. It would be valuable, in the interim, to identify individuals at high risk of epilepsy following TBI or strokes. Predictive tools, such as the SELECT score, predict seizure risk after an ischaemic stroke. Assessing the benefit of prophylactic ASM for those with a high risk of seizures, and trials of anti‐epileptogenic therapies, when available, would be interesting and important. Three distinct factors will determine the future of epilepsy in terms of conceptualisation, enhancing diagnosis, and improving treatments. These factors are the impacts of genomics,7 the use of artificial intelligence (AI),8,9 and a shift away from a phenotypic approach to disease, to a systems‐based perspective.10–12 In addition to these three, a further factor that remains poorly understood but is likely to be of significance are gastrointestinal microbiota.13–15 There is increasing evidence that a gut–brain axis may influence seizure occurrence in some people. Thus, microbiota manipulation could be a promising target in future. Epilepsy, which has traditionally been characterised as a neurological condition defined by unprovoked seizures, is undergoing a redefinition. It is an intricate symptom complex with many somatic, cognitive, and psychiatric comorbidities, which often interact bidirectionally.16–18 For example, people with depression have an increased risk of developing epilepsy, while those with epilepsy are more susceptible to depression.19 Similarly, a bidirectionally shared increased risk exists between generalised epilepsy and type 1 diabetes mellitus20 and between neurodegenerative conditions and epilepsy. Premature mortality is one of the most significant burdens of epilepsy, and much of this increase is attributable to comorbid conditions, which challenge the long‐standing phenotypic approach to medicine.21 The focus now is on the intricate interplay of genetic, developmental, and environmental factors—including social and biological influences—that shape the manifestations of disease. This shift will guide medicine away from a purely disease‐centric approach towards a more integrated systems approach.12 AI, a ubiquitous technological revolution resulting from the rapid development in computing and algorithms, coupled with the explosion of big data, is poised to have a significant impact. It is already having a marked effect in medicine, and epilepsy is beginning to reap its benefits.22–24 AI architectures, inspired by the intricate neuronal connections, are now maturing to enhance the understanding of brain pathologies, including seizure disorders. AI is rapidly improving the accuracy of seizure diagnosis, automating the analysis and interpretation of the clinical history, videos of seizures, electroencephalography, and neuroimaging, and enabling more personalised treatment strategies.25 Machine learning and deep learning models can identify complex brain activity patterns, which will enhance seizure prediction and help localise epileptogenic zones, thus improving diagnostic accuracy, particularly in challenging cases.26 AI‐driven approaches, coupled with whole‐genome data, will be instrumental in developing personalised treatments.27 The goal is to move beyond the traditional empirical trial‐and‐error approach, enhancing individual safety by predicting hazardous side effects of medication and improving the quality of life for people with epilepsy. Autonomous vehicles made possible by AI may also work as a surrogate for driving privileges for some people with active epilepsy. The integration of AI into epilepsy care presents challenges.28 The effectiveness of AI systems centres on access to large and diverse datasets, and the ‘black box’ nature of several currently available algorithms can inhibit clinical trust and lead to accountability issues. There are also some other challenges, such as the potential for ‘algorithm bias’, which may be of concern when the training dataset is not fully representative of the population being assessed.29 Vigorous privacy protection measures also need to be in place, given the sensitivity and ownership of health data. Oversight models to ensure strict adherence to ethical rules and establish safety and transparency in the use of medical AI applications need to be developed and enforced. This entails the constant collaboration between clinicians, data scientists, and regulatory agencies. AI will redefine surgery by identifying potential candidates for surgery earlier than currently and enhancing the precision and accessibility of interventions.22 Increasingly, traditional craniotomies are being replaced by minimally invasive techniques driven by AI. These include stereotactic radiofrequency thermocoagulation, laser interstitial thermal therapy, and focused ultrasound. These techniques can target the epileptogenic zone with less risk, shortening admissions and with quicker recovery times. Neurosurgical procedures assisted by robotic devices will further improve surgical precision and reduce complications. Advances in connectomic mapping will enable the understanding of network connectivity, allowing for surgical interventions personalised to the unique brain networks of an individual. The growing role of neuromodulation, such as responsive neurostimulation and deep brain stimulation, is also significant, particularly with the advent of new AI‐driven, closed‐loop systems. These approaches are increasingly offered to people who are not candidates for resective surgery or whose seizures originate from multiple or eloquent brain regions that cannot be safely resected. AI tools will detect and localise seizure hotspots during surgery, analyse brain signals for precise targeting, and help tailor interventions in real time. The long‐term vision suggests that craniotomies and resective surgery will increasingly become redundant as focal, minimally invasive, and disease‐modifying therapies become safer, more precise, and more broadly available. Ongoing research focused on combining these surgical innovations with genetic profiling, big‐data insights, and advanced neuroimaging, point toward safer, more effective, highly personalised interventions. The future of epilepsy care is intricately linked to advances in genomics, and a personalised approach will only be possible when the full potential of genomics is realised.3,7 Monogenic therapies for epilepsy are still experimental, but clinical trials continue and may hold promise for some people but are unlikely to work universally. Understanding the complete genomic architecture will enable the identification of gene clusters that predispose to seizures and specific gene–environment interactions contributing to seizure development. Analysis of the vast amount of genomic data will crucially depend on AI‐driven bioinformatics. Algorithms will identify subtle genetic patterns associated with different phenotypes and predict responsiveness to specific medications and the likelihood of adverse drug reactions.30 Attention is shifting away from individual organs, molecules, or single genes to encompass the intricate interplay determined and modified by an individual’s genome. Methylation status, stage of brain development, and the internal environment, as mediated by proteomics, metabolomics, and other data, will also play an important role.30 The environment, including social and biological factors such as pollution and global warming, plays a crucial role. Integrating all these factors, intrinsic and extrinsic, congenital, acquired, social and environmental, will demand rigorous assessment of algorithms. The conventional understanding of epilepsy as a singular brain disorder, characterised by a predisposition to seizures and their neurobiological, cognitive, psychological, and social consequences, is undergoing a substantial re‐evaluation.18,31 An evolving view is that epilepsy represents a diverse symptom or manifestation of varied underlying biological processes and moves away from the simplistic assumption that epilepsy is the index condition.18 A deeper understanding of disease‐modifying pathways, such as mTOR, could lead to the development of drugs or processes that reverse epileptogenesis. This necessitates moving away from linear causal models. Frameworks such as the ‘two‐hit hypothesis’, used in oncology, could provide some understanding.32,33 Certain conditions may require multiple ‘hits’ for their development, an example being a somatic mutation in a specific gene associated with epilepsy. Not all people with the same mutation develop seizures, suggesting that other factors are necessary. An epidemiologically sound approach is provided by the ‘sufficient‐component cause model’, which proposes that a disease emerges as a result of a minimal set of contributing factors.34 This could be a combination of genetic predispositions, environmental exposures, and social effects that together create a ‘sufficient cause’ for the development of unprovoked seizures. The model accounts for different risk trajectories and the heterogeneity across lifespan and demographic groups. For example, epilepsy in older adults is frequently linked with cerebrovascular disease and dementia, while in children, developmental and epileptic encephalopathies, often genetically determined, are the prominent cause. In either scenario, seizures may be symptoms of neurodevelopmental or neurodegenerative processes.35–37 The systems‐based framework aligns well with the sufficient‐component cause model and reinforces the notion that no single cause explains most cases of epilepsy.18 Instead, seizures, like cognitive decline or mood disturbance, are expressions of complex biological systems, modulated by internal and external environments. For example, hippocampal malrotation may not lead to epilepsy per se, but in the proper context, for instance, early‐life CNS infection, it may contribute to an epileptogenic cascade. This evolving perspective also requires a re‐examination of comorbidities. The traditional model often places epilepsy as the central ‘index disease’, with comorbidities considered peripheral or secondary.18 This view oversimplifies complex relationships and usually presumes epilepsy to be causally upstream. Mood disorders and cognitive decline often become apparent after epilepsy becomes established and could be a consequence of epilepsy. However, they may also precede epilepsy, and share common underlying risk factors and mechanisms, exemplifying a bidirectional relationship. Different epilepsy syndromes, individual age, and broader social determinants influence the specific nature and interaction of these relationships. The future of disease classification is increasingly shifting away from phenotypically based approaches. There is a growing emphasis on understanding underlying pathophysiological processes or ‘spectra’ that categorises conditions based on their root causes, be they degenerative, infectious, structural, metabolic, or immune‐related.12 Diseases arise from the complex interaction of multiple processes, modulated by individual ‘modifiers’ such as genetic make‐up, environmental exposures, stages of brain development, and social circumstances.12 This requires the adoption of a systems approach, emphasising the need to understand health conditions as a complex interplay within biological, psychological, and social domains. Disease manifestations are often shaped by social and environmental factors, which are potent ‘component causes’. Poverty, pollution, and access to education are not marginal players but are critical health determinants. There is a pressing need to integrate these social determinants into a comprehensive understanding of risk, progression, and inequalities in outcomes. The implications of reframing epilepsy as a symptom are far‐reaching. This requires assessment of underlying causes, taking a holistic approach to clinical care that extends beyond seizure control to address the person’s overall well‐being. An increasing understanding of the pathophysiology of the epilepsies and the role of comorbidities will form the foundation for future disease definitions, moving beyond mere phenotype and symptomatology. In this context, epilepsy will be part of a broader spectrum of transient brain disorders, with migraine, generalised anxiety, and mood disorders, and with each presentation individualised by its specific constellation of causal components. Diagnosis will increasingly rely on systems biology and precision data, moving beyond syndromic labels to mechanistic, individualised insights. Personal profiles will drive treatments—right down to the molecular and social levels—realised through AI‐powered diagnostics, optimally tailored therapies, and, if necessary, minimally invasive interventions. This move from generalised to personalised precision medicine promises not only greater efficacy but also more equitable, person‐centred epilepsy care. Epilepsy, as we know it today, will cease to exist, but unprovoked seizures will continue as a symptom within a spectrum of transient brain symptoms. In the foreseeable future, AI algorithms will analyse symptoms, videos of seizures, clinical signs, EEG, imaging, genomic, metabolomic, and social data to formulate likely diagnoses and treatment plans. An individual’s complete health and social care data will be available to relevant healthcare professionals at all times, to rapidly inform treatment options and provide information pertinent to individuals. Nationwide call centres will likely be available at all times to respond to individuals’ concerns and queries, with algorithm‐driven responses. Whilst these advances will bring efficiency and consistency, it is essential not to lose sight of the fact that people with epilepsy are vulnerable human beings, and there is a vital need for consistent care being provided by an empathic, well‐informed and well‐trained health care professionals.
CHAPTER 26
Future Perspectives for Epilepsy
EXPECTATIONS
PREVENTION
THE FUTURE OF EPILEPSY AS A DIAGNOSIS
THE EXPANDING ROLE OF ARTIFICIAL INTELLIGENCE
IMPACT ON EPILEPSY SURGERY
THE ROLE OF GENOMICS
CHANGING THE CONCEPTS OF EPILEPSY
DELIVERY OF CARE
Related posts:
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
