The Changing Paradigm of Neuroinflammation in Epilepsy: A Bibliometric Analysis from Microglial Activation to Gut-Brain Axis

Abstract

Background

Neuroinflammation has been increasingly implicated in epileptogenesis and drug-resistant epilepsy, leading to growing research interest beyond the traditional focus on neuronal hyperexcitability. Despite the rapid expansion of research in this interdisciplinary field, a comprehensive structural mapping of its intellectual evolution and emerging frontiers is lacking.

Methods

We performed a systematic bibliometric analysis of 1989 publications (2005–2026) retrieved from Web of Science Core Collection and Scopus. Analytical tools including CiteSpace, VOSviewer, and the Bibliometrix R package were employed to visualize collaboration networks, citation structures, and thematic transitions. The literature search was updated to include publications indexed up to January 4, 2026.

Results

Research output followed an exponential growth pattern (R²=0.97), peaking in 2025. China and the USA dominated the global landscape, with the Journal of Neuroinflammation and Epilepsia identified as core academic hubs. Highly cited works, led by pioneers such as Vezzani A and Aronica E, established a knowledge base centered on microglial activation and cytokine signaling. Keyword clustering identified several interconnected thematic domains. Long-standing topics included oxidative stress and hippocampal vulnerability, whereas recent citation bursts indicated increasing attention to autoimmune encephalitis, blood–brain barrier (BBB) dynamics, and the gut–brain axis. Furthermore, emerging trends highlight the ketogenic diet and network pharmacology as promising immunometabolic strategies for seizure control.

Conclusion

This study provides the first systematic bibliometric landscape of neuroinflammation in epilepsy over the past two decades. Our findings suggest a gradual transition from studies of glial, cytokine, and vascular mechanisms toward translational interest in systemic immune–CNS interactions. These quantitative insights identify the gut-microbiota-inflammation axis and personalized immunotherapy as the next frontiers, offering a strategic roadmap for future disease-modifying interventions in refractory epilepsy.

Introduction

Epilepsy is a chronic neurological disorder affecting approximately 50 million people worldwide and is clinically defined by recurrent, unprovoked seizures.¹ Despite significant advances in antiseizure medications and surgical therapies, a substantial proportion of patients develop drug-resistant epilepsy, highlighting the limitations of treatment strategies that primarily target abnormal neuronal firing.^2,3 In this context, growing experimental and clinical evidence has identified neuroinflammation as a critical driver of seizure generation and epileptogenesis.^4,5 Pro-inflammatory cytokines, such as interleukin-1β, interleukin-6, and tumor necrosis factor-α, can disrupt neuronal and synaptic function, reduce seizure thresholds, and promote recurrent seizures.^6,7 Persistent neuroinflammatory activity has also been closely linked to disease severity and therapeutic resistance, particularly in temporal lobe epilepsy.⁸ Early studies established the foundational role of neuroinflammation in epilepsy, demonstrating that glial activation and pro-inflammatory cytokines such as IL-1β and TNF-α modulate neuronal excitability and seizure susceptibility.^9–11 Subsequent clinical and experimental investigations further confirmed the involvement of neuroinflammatory processes in epileptogenesis and temporal lobe epilepsy, particularly through microglial activation and blood–brain barrier dysfunction.^12–14 These studies collectively laid the groundwork for the subsequent expansion of the field toward systemic immune interactions, metabolic regulation, and gut–brain axis–related mechanisms. Beyond classical inflammatory pathways, recent studies have expanded the conceptual framework of neuroinflammation in epilepsy to include systemic immune–CNS interactions, gut microbiota–related inflammatory mechanisms, and intracellular regulatory pathways such as mammalian target of rapamycin (mTOR) signaling.^15–17 These advances underscore the growing recognition of epilepsy as a disorder arising from dynamic interactions among neural, immune, and metabolic systems rather than isolated neuronal dysfunction.

Despite the rapid growth of related research, the overall knowledge structure of neuroinflammation in epilepsy remains fragmented across disciplines and mechanistic domains. Bibliometric analysis offers a quantitative approach to systematically map research development, identify intellectual foundations, and track thematic evolution based on large-scale citation data. Although a previous bibliometric study identified inflammation as a major research hotspot in epilepsy,¹⁸ the field still lacks a dedicated bibliometric analysis focusing specifically on neuroinflammation in epilepsy, particularly using a broader temporal span and a dual-database approach. Given that neuroinflammation now spans glial biology, cytokine signaling, BBB dysfunction, oxidative stress, autoimmune epilepsy, and gut–brain interactions, an independent bibliometric analysis is needed to clarify how these subfields are connected and how research priorities have changed over time. Therefore, the present study aims to systematically characterize the research landscape of neuroinflammation in epilepsy from 2005 to 2026 using bibliometric and visualization approaches. By identifying major research hotspots, influential contributors, and emerging thematic trends, this work seeks to provide a structured overview of the field and quantitative insights to support future mechanistic studies and translational research.

Methods

Database and Search Strategy

Publications related to neuroinflammation and epilepsy were retrieved from the WoSCC and Scopus, covering the period from January 2005 to January 4, 2026. Both databases were used to ensure comprehensive literature coverage.^19–21 Searches were conducted using terms derived from Medical Subject Headings (MeSH) and their related entry terms, focusing on two core concept domains: neuroinflammation and epilepsy. In WoSCC, the search was performed in the title (TI), abstract (AB), and author keywords (AK) fields, while the equivalent TITLE-ABS-KEY fields were used in Scopus. Boolean operators were applied to combine terms related to neuroinflammation and epilepsy. The complete search strategies for both databases are provided in Table S1. Only English-language articles and reviews were included to ensure consistency in data processing and analysis, which may slightly limit the generalizability of the findings across non-English literature and potentially underrepresent region-specific publications. Non-research document types, including proceedings, corrections, news items, book chapters, retracted publications, and editorials, were excluded. This study did not restrict population characteristics, including age group (children or adults) or experimental species (human or animal studies), in order to ensure comprehensive coverage of the field. All epilepsy-related publications were included regardless of clinical subtype, such as temporal lobe epilepsy, drug-resistant epilepsy, cryptogenic epilepsy, and reflex epilepsy. The search strategy was based on MeSH-derived terms and expanded synonym sets to capture variations in terminology across the literature. As a bibliometric study, the aim was to map the intellectual structure of the field rather than to perform clinical subgroup analyses. Duplicate records were identified and removed using R software and Microsoft Excel. After data cleaning, a total of 1989 publications were retained for bibliometric analysis. The literature selection process is illustrated in Figure S1.

Data Processing

To ensure methodological rigor and reproducibility, additional procedures were implemented. Duplicate records retrieved from WoSCC and Scopus were removed using a combination of automated de-duplication in Bibliometrix R package and manual screening based on DOI, title, and authorship.

In CiteSpace, cluster labels were generated using the log-likelihood ratio (LLR) algorithm. Network structure and clustering consistency were assessed using modularity Q and the weighted mean silhouette coefficient. These methodological settings were applied consistently across all analyses to ensure comparability and reproducibility of results.

CiteSpace

CiteSpace (6.4R1, 64-bit Advanced Edition)²² was used to perform co-citation and co-occurrence analyses. The study period was divided into one-year time slices from 2005 to 2026. For reference co-citation analysis, nodes were selected using the g-index criterion (k = 25), and Pathfinder pruning was applied to reduce redundant network links. Authors, institutions, and keywords were selected as node types. For author and institution analyses, the top 25 most frequent nodes per time slice were retained without pruning. In keyword analysis, the top 25 keywords per slice were selected, and network pruning was conducted using the Pathfinder and merged network algorithms. Records from WoSCC and Scopus were exported in plain text format, including full records and cited references.

VOSviewer

VOSviewer (version 1.6.20)²³ was employed to construct and visualize collaboration and co-occurrence networks among authors, institutions, and keywords. Full counting was applied, and minimum occurrence thresholds were adjusted according to the specific analysis to ensure clear visualization of network structures.

Bibliometrix

The bibliometrix R package (version 4.5.1),²⁴ was used for quantitative bibliometric analyses, including journal and author productivity, historiographic mapping, and research trend evaluation. Bibliometric indicators, including the g-index, h-index, number of citations (NC), and number of publications (NP) were calculated to assess academic influence and research performance.

The Other Tools

Microsoft Excel 2021 (version 16.48) was used for preliminary data processing and for generating basic descriptive charts. Journal impact factors (IFs) and quartile rankings (Q1–Q4) were obtained from the 2021 Journal Citation Reports (JCR).

Results

Annual Publication Trends

From 2005 to 2026, a total of 1989 publications related to neuroinflammation in epilepsy were identified across 620 sources (Table S2). Original research articles predominated, comprising 1425 publications (71.64%), while review articles accounted for 564 publications (28.36%) (Figure 1A). The overall annual growth rate was 11.32%, and the mean document age was 4.99 years (Table S2). Temporal analysis demonstrated a sustained upward trend in publication output (Figure 1B and C). Research activity remained limited from 2005 to 2010, followed by steady growth between 2011 and 2015. A marked increase was observed after 2016, with annual output peaking at 386 publications in 2025. Exponential regression analysis, excluding data from 2026 due to incomplete data collection, revealed a strong exponential growth pattern for both annual and cumulative publications (R² ≈ 0.97; Figure 1D). Collectively, these findings indicate sustained scholarly interest and rapid expansion of research on neuroinflammation in epilepsy.

Figure 1.

Temporal trends and document type distribution of publications on neuroinflammation in epilepsy from 2005 to 2026. (A) Distribution of document types, with articles (71.64%) and reviews (28.36%); (B) Annual publication output (blue bars); (C) Cumulative publication growth over time; (D) Exponential fitting of annual and cumulative publication outputs related to neuroinflammation in epilepsy. The fitting analysis demonstrates an exponential growth pattern with high goodness of fit (R² > 0.97) (data from 2026 were excluded due to incomplete data collection).

Distribution of Countries and Regions

The international collaboration network showed widespread participation across Asia, Europe, North America, and Oceania, with comparatively lower representation from South America and Africa (Figures 2A and B). China and the United States occupied central positions in the network, while European countries—including Germany, Italy, and Spain—formed closely connected regional clusters. Corresponding author analysis indicated that China ranked first in publication output with 362 articles (18.2%), followed by the United States with 267 articles (13.4%) and India with 99 articles (5.0%) (Table S3 and Figure 2C). Publications from China were predominantly single-country publications, whereas the United States and multiple European countries exhibited higher proportions of multi-country publications. The United States showed the highest total citations (13,733) and average citations per article (51.40). National outputs increased steadily after 2010 and accelerated after 2018, with China demonstrating the most rapid growth and the United States maintaining stable high-impact output (Table S3 and Figure 2D).

Figure 2.

Trends in Country/Region and Institutional Contributions and Collaboration Networks Analysis of Neuroinflammation in Epilepsy. (A) Country/Region Collaboration Map, with color intensity indicating national publication output and link width reflecting cross-country collaboration strength; (B) Country/region clustering network (VOSviewer), with node size proportional to publication output; (C) Leading countries by publication volume and collaboration pattern, showing single-country (SCP) and multi-country publications (MCP); (D) Cumulative publication trends of the top five countries (2005–2026); (E) Top ten institutions ranked by total number of publications in neuroinflammation in epilepsy research; (F) Institutional collaboration network, with nodes sized by publication output and colored by collaboration clusters; (G) Cumulative publication trends of the top five affiliations (2005–2026).

Distribution of Institutions

Several institutions emerged as major contributors to research on neuroinflammation in epilepsy (Table S4). The Egyptian Knowledge Bank (EKB) ranked first with 99 publications, followed by the University of California system (68), Monash University (57), the University of London (54), and both Central South University and Tehran University of Medical Sciences (53 each). Additional contributors included Fudan University, the University of Amsterdam, INSERM, and the University of Bonn (Figure 2E). Institutional collaboration analysis revealed a multi-cluster network structure with extensive inter-institutional connections (Figure 2F). High-output institutions occupied central positions with dense collaborative links, highlighting their coordinating roles in inter-institutional research. Temporal overlay analysis indicated limited activity before 2015, followed by rapid growth after 2018 (Figure 2G). While the University of California system and Monash University exhibited earlier growth, EKB and several Asian institutions demonstrated significant recent expansion, reflecting an increasingly diversified global research landscape.

Distribution of Authors and Co-Cited Authors

Several authors demonstrated sustained academic influence in neuroinflammation-related epilepsy research (Table S5). Vezzani A ranked first across multiple bibliometric indicators, with the highest h-index (24), g-index (31), 31 publications, and 3773 total citations. Aronica E and Henshall DC also showed high citation impact and stable productivity. Most highly productive authors began publishing between 2006 and 2015. Author collaboration analysis identified a multi-cluster network structure (Figure 3A). The largest cluster was centered on Vezzani A, while additional clusters led by Aronica E, Henshall DC, and Ravizza T reflected stable collaborative research groups. Co-citation analysis highlighted Vezzani A, Aronica E, and Van Vliet EA as central nodes in the co-citation network (Figure 3B). Overall, the field comprised 9531 authors, with an average of 6.57 authors per publication and an international co-authorship rate of 21.32% (Table S2).

Figure 3.

Network analysis of authors and journals on Neuroinflammation in Epilepsy (2005–2026). (A) Author collaboration network, with node size indicating publication output and colors denoting collaboration clusters; (B) Author co-citation network, where link thickness reflects co-publication strength and colored clusters represent research communities; (C) Core journal identification based on Bradford’s Law; (D) Journal citation network, in which node size corresponds to publication volume, edge width indicates citation strength, and colors denote thematic clusters; (E) Journal co-citation network, showing co-citation intensity by node size and link thickness, with clusters highlighted by color; (F) Publication growth trends of the top five journals, generated using the bibliometrix package; (G) Dual-map overlay of journals showing interdisciplinary citation flows in neuroinflammation research in epilepsy.

Journals and Co-Citation Analysis

According to Bradford’s law, publications on neuroinflammation in epilepsy were concentrated in a limited number of core journals (Figure 3C and Table S6). The Journal of Neuroinflammation ranked first, with the highest h-index (29) and 52 publications, followed by Epilepsia and the International Journal of Molecular Sciences. Other frequently contributing journals included Neurobiology of Disease, Molecular Neurobiology, Frontiers in Cellular Neuroscience, Frontiers in Pharmacology, Neurochemical Research, Epilepsy and Behavior, and Frontiers in Neurology, most of which were classified as JCR Q1–Q2 journals. Journal co-citation analysis revealed a highly interconnected network (Figures 3D and E) with the Journal of Neuroinflammation, Epilepsia, and Molecular Neurobiology occupying central positions. Temporal analysis showed stable publication output in Epilepsia throughout the study period, while the Journal of Neuroinflammation and the International Journal of Molecular Sciences exhibited increased output after 2015 (Figure 3F). Dual-map overlay analysis indicated that publications were primarily located in clinical medicine and neuroscience journals, while their cited knowledge base originated largely from molecular biology and basic biomedical sciences (Figure 3G).

Highly Cited References and Co-Citation Structure

Local citation analysis identified a set of highly influential publications that formed the core knowledge base of neuroinflammation research in epilepsy (Figure 4A–D and Table S7). The most locally cited articles were authored primarily by Vezzani A, including reviews published in Nature Reviews Neurology (2019)¹¹ and Neuropharmacology (2015).¹⁰ Other frequently cited works by Devinsky O,⁹ Webster KM,²⁵ and Aronica E¹² also showed high local citation counts. Co-citation cluster analysis revealed several major thematic clusters (Figure 4B and C). The largest clusters were labeled #0 “oxidative stress”, #3 “brain inflammation”, and #4 “hippocampal damage”, which displayed sustained activity over extended periods. Smaller clusters—including “mast cell”, “microRNA regulation”, and “gut–brain axis”—emerged predominantly after 2018. Citation burst analysis showed that early bursts were associated with foundational studies on brain inflammation and epileptogenesis, whereas more recent bursts (2022–2023) were associated with immune modulation and translational research themes (Figure 4D).

Figure 4.

Reference co-citation and citation burst analysis in neuroinflammation in epilepsy research. (A) Reference co-citation network generated by CiteSpace, where node size reflects reference influence, links denote co-citation relationships, and color gradients indicate publication years; (B) Reference cluster dependency map, with colors representing distinct clusters and links illustrating inter-cluster relationships; (C) Timeline visualization of reference clusters, showing the temporal evolution of co-cited references; (D) Top 25 references exhibiting the strongest citation bursts from 2005 to 2026, with red bars highlighting periods of increased citation activity.

Keyword Analysis

Keyword Frequency and Co-Occurrence Analysis

Keyword co-occurrence networks showed that “neuroinflammation” and “epilepsy” were the most frequent and central terms (Figure 5A). Other high-frequency keywords included “microglia”, “oxidative stress”, “epileptogenesis”, “status epilepticus”, “hippocampus”, and “inflammation”. Temporal overlay analysis revealed that earlier studies focused on epilepsy phenotypes and neuronal injury, whereas more recent keywords were associated with microglial activation, blood–brain barrier dysfunction, gut–brain axis, and neurodegeneration (Figure 5B). Word cloud analyses based on title and author keywords supported these patterns (Figures 5C and D).

Figure 5.

Keyword Analysis and Visualization of neuroinflammation in epilepsy. (A) Keyword co-occurrence network generated by VOSviewer, where node size represents keyword frequency and colors denote thematic clusters (Low-frequency nodes were filtered to improve visualization clarity); (B) Time-overlay keyword co-occurrence network, with node color indicating the average publication year (lighter colors reflect more recent topics); (C) Word cloud of title keywords, with word size proportional to frequency; (D) Word cloud of author keywords, highlighting frequently used terms; (E) Keyword co-occurrence network from CiteSpace, illustrating keyword frequency, relationships, and temporal distribution of research hotspots (Low-frequency nodes were filtered to improve visualization clarity); (F) Timeline visualization of keyword clusters, depicting the temporal evolution of research themes and hotspot shifts; (G) Top 25 keywords with the strongest citation bursts related to neuroinflammation in epilepsy, where red segments mark periods of rapid attention increase; (H) Trend topics analysis showing the development of major research themes over time, with bubble size indicating term frequency.

Keyword Clustering Analysis

Keyword clustering analysis identified multiple thematic clusters (Figures 5E and F). Major clusters included “oxidative stress”, “blood–brain barrier”, “status epilepticus”, “ketogenic diet”, “drug-resistant epilepsy”, and “autoimmune encephalitis”. Timeline visualization showed sustained activity for clusters such as oxidative stress and status epilepticus, while clusters related to ketogenic diet and autoimmune encephalitis exhibited increased activity in more recent years.

Keyword Evolution and Burst Analysis

Keyword citation burst analysis identified terms with rapidly increasing attention during different periods between 2005 and 2026 (Figure 5G). Early burst keywords included “arachidonic acid” and “dentate gyrus”, followed by intermediate bursts related to epilepsy surgery, hippocampal sclerosis, blood–brain barrier, and status epilepticus. In recent years, burst keywords increasingly included “antiseizure medication”, “autoimmune encephalitis”, “vagus nerve stimulation”, and “network pharmacology”. Trend topic analysis showed that emerging keywords were concentrated in the most recent years (Figure 5H).

Discussion

Epilepsy is a highly heterogeneous neurological disorder with complex and incompletely understood pathophysiological mechanisms, which continue to challenge effective disease control.^26,27 Although current therapeutic strategies primarily aim to suppress seizures, approximately one-third of patients remain drug-resistant, underscoring the limitations of symptom-oriented treatment approaches.^28,29 These challenges highlight the need to identify disease-modifying mechanisms that contribute to epileptogenesis and treatment refractoriness. In this context, neuroinflammation has increasingly been investigated as a potential disease-modifying process, with accumulating evidence suggesting its involvement in seizure initiation, network reorganization, and disease progression rather than being a purely secondary response to recurrent seizures.^28,30 Consistent with this conceptual shift, the bibliometric architecture of the field delineates several interconnected domains, including microglial activation and cytokine signaling, blood–brain barrier (BBB) dysfunction, oxidative stress and mitochondrial perturbation, hippocampal vulnerability, and emerging immune-related and peripheral immunometabolic themes such as autoimmune encephalitis and the gut–brain axis. Importantly, these domains emerge from the structural organization of the literature rather than individual studies, providing a quantitative perspective on how mechanistic emphases and translational interests have co-evolved over time.

Microglial Activation and BBB Dysfunction

Keyword co-occurrence and clustering analyses position microglial activation among the most central themes in neuroinflammation-related epilepsy research. Rather than forming an isolated cluster, microglia-related terms exhibit strong linkages with cytokine signaling, BBB dysfunction, and drug-resistant epilepsy, reflecting a recurring glia–vascular immune axis and a gradual shift toward translationally oriented research questions. Across experimental and clinical studies, pro-inflammatory mediators—particularly IL-1β and IL-6—are consistently implicated in altered neuronal excitability and sustained gliosis.^7,31 In temporal lobe epilepsy, persistent microglial activation has been associated with excitotoxicity and network remodeling,³² and microglia-related therapeutic strategies have been frequently reported in preclinical studies and are strongly associated with the microglial activation cluster in co-occurrence analyses.³³

BBB-related terms also emerge as a recurrent and closely connected theme in the bibliometric maps. Mechanistically, BBB disruption may amplify microglial activation, while inflammatory signaling from activated microglia can further destabilize barrier integrity, forming a feed-forward loop.^34,35 In line with this interaction, strategies targeting microglial signaling—such as microRNA modulation or pharmacological inhibition—have been reported to attenuate neuroinflammatory cascades and improve synaptic function in epilepsy models.^33,36 Cytokine programs involved in barrier repair have been proposed in neuroinflammatory injury contexts,³⁷ although epilepsy-specific validation remains limited.

Beyond the core microglia–BBB axis, secondary neuroinflammatory contributors may further modulate disease trajectories. Astrogliosis likely interfaces with microglial activity and synaptic dysfunction,^38,39 while prostaglandin signaling via the COX-2/mPGES-1/PGE2 pathway has been linked to excitatory amplification in seizure models.^40,41 Dysregulated microRNA signaling and mast cell–associated immune signatures have also been reported in epilepsy-related neuroinflammatory settings, although findings remain heterogeneous and context-dependent.^42–44

Oxidative Stress and Hippocampal Vulnerability

Oxidative stress emerges as a long-standing and stable thematic domain in the bibliometric landscape and is consistently linked with neuroinflammation-related keywords. This co-occurrence suggests that redox imbalance is frequently conceptualized as an amplifier within inflammation–seizure interactions rather than an isolated mechanism. Inflammatory activation can increase reactive oxygen species (ROS) production and impair mitochondrial function, while oxidative stress may, in turn, intensify inflammatory signaling, collectively promoting neuronal hyperexcitability.^5,45 Clinical and experimental evidence supports this bidirectional relationship. In temporal lobe epilepsy, ferroptosis-related signatures and microglial oxidative injury correlate with seizure burden, implicating redox-driven microglial dysfunction in disease progression.⁴⁶ Experimental models further demonstrate increased lipid peroxidation alongside reduced antioxidant defenses, indicating a sustained pro-oxidant environment that may reinforce neuroinflammatory cascades.⁴⁷ Oxidative stress has also been linked to BBB injury, providing a mechanistic bridge between metabolic stress and immune amplification.⁴⁸

Hippocampal vulnerability represents another stable bibliometric theme and is tightly coupled with oxidative stress and inflammation. In epileptogenesis models, hippocampal microglial activation and inflammatory gene programs converge on synaptic remodeling and neuronal injury.⁴⁹ Genetic and experimental models of hyperexcitability further support the intrinsic susceptibility of hippocampal circuitry to inflammatory and oxidative insults.⁵⁰ Together, these findings suggest that the hippocampus emerged as a highly connected node in keyword co-occurrence networks, frequently associated with oxidative stress and neuroinflammation-related terms, suggesting its central role in the thematic structure of the field.

Autoimmune Encephalitis–Associated Epilepsy and Immunotherapy

Keyword clustering and citation burst analyses reveal increasing attention to autoimmune encephalitis–associated epilepsy and immune-defined clinical phenotypes. Mechanistic and clinical studies underlying this trend consistently implicate BBB dysfunction, autoantibody-mediated synaptic disturbances, and sustained cytokine signaling in seizure persistence.^51,52 Clinically, autoimmune mechanisms are particularly relevant in new-onset refractory seizures and temporal lobe epilepsy presentations. Antibody-negative autoimmune encephalitis may present as adult-onset epilepsy, highlighting that immune-driven epileptogenesis can occur even in the absence of detectable neuronal antibodies.⁵³ Targeted immunotherapies, including IL-6 receptor blockade, have demonstrated benefit in selected refractory cases.⁵⁴ Early immune intervention is generally associated with improved outcomes compared with antiseizure medications alone, supporting a disease-modifying role for inflammatory pathways.^7,55 Nevertheless, chronic epilepsy may still develop despite treatment, suggesting that delayed inflammatory control can permit lasting network remodeling.⁵⁶ Emerging therapeutic strategies extend beyond acute immunosuppression toward broader immune modulation. Experimental studies suggest that mesenchymal stem cell–derived secretomes can dampen pro-inflammatory signaling in epileptic hippocampal tissue.⁸ Given the heterogeneity of immune-mediated epilepsy, neuroinflammation imaging and immune biomarkers have been proposed to refine patient stratification and guide personalized immunotherapy.^57,58

Gut–Brain Axis and Dietary Modulation

The gut–brain axis has emerged as a growing research frontier in neuroinflammation-related epilepsy. Bibliometric mapping indicates increasing interest in microbiota dysbiosis and intestinal barrier dysfunction as upstream modulators of systemic immune–metabolic signaling that may shape central inflammatory tone and seizure susceptibility.¹⁶ Within this thematic cluster, bibliometric patterns also indicate increasing attention to differences in gut microbiota composition across clinically distinct epilepsy populations, particularly between drug-resistant epilepsy, drug-sensitive epilepsy, and healthy controls. These comparative studies have gradually become a recurrent research focus, reflecting growing interest in disease heterogeneity within neuroinflammation-related epilepsy research.

In addition, emerging evidence suggests a potential association between gut microbiota dysbiosis and variability in antiepileptic drug response.⁵⁹ This may imply that microbial composition could indirectly influence treatment outcomes through immune–metabolic interactions. However, within the scope of bibliometric analysis, these findings should be interpreted as emerging research directions rather than established mechanistic conclusions.

Within this context, dietary interventions represent a practical translational entry point linking peripheral and central inflammatory processes. The ketogenic diet (KD) forms a distinct and increasingly active keyword cluster and is discussed for its dual role in seizure suppression and microbiome remodeling in drug-resistant epilepsy.⁶⁰ Mechanistic studies suggest that KD-induced shifts in microbial composition and metabolite profiles may modulate neuroinflammatory pathways and gut–brain signaling.^60,61 Beyond KD, broader dietary patterns may influence inflammatory load and cognitive vulnerability. Diets enriched in antioxidants and omega-3 fatty acids are generally associated with anti-inflammatory effects, whereas high-fat or nutritionally imbalanced diets may exacerbate neuroinflammatory states.^62,63 Collectively, these findings highlight modifiable peripheral targets—microbiota composition, intestinal barrier integrity, and nutrition—that may complement central immunomodulatory strategies, particularly in drug-resistant epilepsy.^8,33

Limitations

This study provides a systematic bibliometric overview of neuroinflammation research in epilepsy. By integrating multiple databases and applying complementary analytical tools, including CiteSpace, VOSviewer, and Bibliometrix, it enables a multidimensional assessment of collaboration patterns, knowledge structures, and thematic evolution. This integrative approach enhances objectivity and reproducibility while offering a structured perspective on the development of the field. Several limitations should be noted. Bibliometric findings are inherently influenced by database coverage, citation practices, and algorithmic parameters, which may underrepresent recently published or less frequently cited studies. Additionally, restricting the analysis to English-language publications may exclude relevant work in other languages. These limitations reflect the methodological constraints of bibliometric analysis rather than the validity of the observed research trends.

Conclusion

Our findings underscore the increasing centrality of neuroinflammation within epilepsy research. Across the mapped literature, core domains—microglial activation/cytokine signaling, BBB dysfunction, oxidative stress with hippocampal vulnerability, and immune-defined epilepsy phenotypes—form a structured yet heterogeneous knowledge landscape that connects mechanistic work with clinically oriented questions. Meanwhile, emerging attention to peripheral immunometabolic modulation (including the gut–brain axis and dietary strategies) highlights a translational direction toward biomarkers and disease-modifying interventions beyond seizure suppression, particularly reflected in inflammatory and immune-related research signatures identified through bibliometric mapping. These findings further suggest that immunotherapy and microbiota-targeted strategies may warrant prioritized evaluation in future clinical research for refractory epilepsy.

Funding Statement

The authors declare that no funding was received for this study.

Use of Artificial Intelligence

The authors used ChatGPT (OpenAI, GPT-5.3) to assist with language editing and articlerefinement. All content was critically reviewed and verified by the authors to ensure accuracy and integrity.

Abbreviations

WoSCC, Web of Science Core Collection; CNS, central nervous system; mTOR, mammalian target of rapamycin; MeSH, Medical Subject Headings; TI, title; AB, abstract; AK, keywords; CWTS, Centre for Science and Technology Studies; NC, Number of Citations; NP, Number of Publications; IFs, Impact Factors; JCR, Journal Citation Reports; SCP, single-country publications; MCP, multiple-country publications; EKB, The Egyptian Knowledge Bank; TC, Total Citations; BBB, blood-brain barrie; IL-1β, interleukin-1β; IL-6, interleukin-6; TNF-α, tumor necrosis factor-α; ROS, reactive oxygen species; KD, ketogenic diet.

Data Sharing Statement

The datasets analyzed in this study are derived from publicly available databases, including Web of Science Core Collection and Scopus. The data are available from the corresponding author upon reasonable request.

Author Contributions

J.J.L.: Conceptualization, Methodology, Investigation, Data curation, Formal analysis, Visualization, Writing - original draft. G.X.C.: Investigation, Data curation, Formal analysis, Software, Writing - review & editing. Z.Y.B.: Visualization, Validation, Formal analysis, Writing - review & editing. F.H.X.: Conceptualization, Methodology, Supervision, Project administration, Validation, Writing - review & editing. All authors made a significant contribution to the work reported, approved the final version to be published, agreed on the journal to which the article has been submitted, and agree to be accountable for all aspects of the work.

Disclosure

There are no disclosed conflicts of interest for the authors.