Hotspots and trends of microglia in ischemic stroke: A bibliometric analysis from 1998 to 2024

Abstract

Background:

Ischemic stroke (IS) represents a predominant cause of disability globally, characterized by significant neuroinflammatory responses facilitated by microglia, the resident immune cells of the central nervous system. A comprehensive understanding the current research landscape in this field is essential for guiding future inquiries. However, a systematic, quantitative panorama analysis of this intersecting field is lacking. This study aims to address this knowledge gap by conducting a bibliometric analysis of publications retrieved from the Web of Science Core Collection database, elucidating historical developments, collaboration networks, and emerging frontiers.

Objective:

This study endeavors to perform a comprehensive bibliometric analysis of microglia research in IS, with the aim of elucidating historical developments, identifying current research frontiers, and proposing future directions through a quantitative assessment of publication trends, collaboration networks, and the evolution of conceptual frameworks.

Methods:

Utilizing controlled vocabulary terms for IS and microglia, we conducted a systematic search of the Web of Science Core Collection database. After applying exclusion criteria, our analysis included 5709 publications (4859 articles; 850 reviews). We employed Bibliometrix R, VOSviewer, and CiteSpace for advanced bibliometric analysis.

Results:

The annual publication count showed exponential growth, with a compound annual growth rate of 14.6% over the past 5 years (2020–2024), indicating robust research momentum. Systematic reviews and meta-analyses accounted for 14.9% of the total, highlighting the growing importance of integrative evidence. China (n = 1852) and the United States (n = 1723) leading in research contributions. Capital Medical University emerged as the most prolific institution (n = 217). Journal of Neuroinflammation was the primary publication venue (n = 142), while Stroke had the highest citation count (8742 citations). Moo-Ho Won was the most productive author (n = 68), and Iadecola C. was the most co-cited researcher (n = 1024). Keyword analysis identified “microglia polarization” (burst strength = 36.5), “NLR family pyrin domain containing 3 inflammasome” (burst strength = 32.1), and “extracellular vesicles” (burst strength = 28.7) as emerging research frontiers.

Conclusion:

Our analysis outlines 3 key research trajectories, molecular mechanisms of microglia activation, particularly NLR family pyrin domain containing 3-mediated pathways; phenotypic modulation of microglia during stroke recovery; and novel roles in angiogenesis and immunomodulation. The emergence of “gut-microbiota-brain axis” and “mitochondrial transfer” as new research themes offers promising directions for investigation. These findings provide a strategic framework for prioritizing research on neuroinflammation and neurorepair after stroke.

1. Introduction

According to the World Stroke Organization 2025, ischemic stroke (IS) is the leading cause of severe disability and the 2nd leading cause of death globally, following coronary heart disease^[1,2] IS accounts for 75% to 80% of all strokes, is characterized by high morbidity, disability, and mortality.^[3] According to the Global Burden of Disease Study 2023, IS is responsible for the loss of approximately 330 million disability-adjusted life years, with patients in China accounting for 40% of this total.^[4] Although thrombolytic and thrombectomy techniques can restore blood flow, their use is limited by a narrow therapeutic time window (<4.5 hours) and their applicability to only 30% to 40% of patients.^[5] Within this critical public health challenge, the neuroinflammatory response is identified as a central mechanism underlying secondary brain injury during both the acute phase of the disease and the recovery process.^[6]

Microglia, the resident immune cells of the central nervous system (CNS), account for 5% to 10% of the total number of CNS cells and serve as the primary responders during the pathological processes associated with IS.^[7] Under physiological conditions, microglia are responsible for maintaining the homeostasis of the neural microenvironment through dynamic synaptic monitoring; under pathological conditions, they are activated by damage-associated molecular patterns, etc, and undergo significant morphological remodeling and functional transformation, transforming from a branched, resting state to an amebic, activated state, and rapidly migrating to the affected area region.^[8] This process involves a complex molecular regulatory network that determines the dual role of microglia in nerve injury and repair-both releasing pro-inflammatory factors to amplify injury and secreting neurotrophic factors to promote repair. Consequently, microglia occupy an important pivotal position in IS. Although numerous research literature has explored the key role of microglia in the pathological mechanisms of stroke, the current field still lacks systematic integration and visual analysis, which hampers the ability to elucidate the dynamic shifts in research paradigms and potential pathways for translation.

Bibliometrics, as a statistical analysis and quantitative tool, employs mathematical and statistical methods to investigate publications, aiming to track the development of specific research fields within defined timeframes.^[9,10] It places particular emphasis on analyzing relevant information within the research domain, such as the influence of publications, contributions from countries/regions/affiliations/journals/authors/references/keywords, and primary future research directions in the field.^[11] Utilizing the Web of Science database and analytical software, this interdisciplinary quantitative approach untangles global research trends and hot topics within a research area.^[12] Visualization analysis software, including CiteSpace (Version 6.4.R1 Advanced Edition; Drexel University, Philadelphia), VOSviewer (Version 1.6.19; Leiden University Centre for Science and Technology Studies (CWTS), Leiden, The Netherlands), and R software (R Foundation for Statistical Computing, Vienna, Austria), are widely utilized in bibliometric analyses.^[13] These 3 software tools were also employed in our study.

Therefore, this study employs bibliometric methods to conduct a panoramic analysis of global literature in the crossover field of stroke and microglia. The aim is to construct a knowledge map, identify core collaboration networks, and emerging trends, and discusses the role of microglia in the pathological processes of ischemic stroke and the possible molecular mechanisms involved. This will provide new ideas for the development of neuroprotective agents and optimization of neuromodulation, and promote the precise translation of basic research.

2. Methods

2.1. Data collection

The Web of Science Core Collection (WoSCC) was used as the main database for data retrieval. As one of the largest and most comprehensive online databases globally, WoSCC contains a vast amount of scientific research with high authority and reference value. In this study, the retrieval formula was set as follows: TS = (“IS” OR “brain infarction” OR “cerebral infarction””OR “ischemic encephalopathy” OR “infarction encephalopathy” OR “brain ischemia” OR “cerebral ischemia”), AND TS = (“Microglias” OR “Microglial Cell” OR “Microglial Cells” OR “Microglial” OR “Microglia”). The search covered the time span from January 1, 1998, to December 31, 2024. The search was restricted to English-language “Articles” and “Review Articles.” Early Access, Book Chapters, Preceding Paper and Retracted Publication were excluded. The retrieval process is presented in Figure 1. To prevent potential biases caused by database updates, all searches and downloads were completed on a single day. All the selected articles were exported in TXT format. Finally, a total of 5709 records, including 4859 articles and 850 reviews, matched the inclusion criteria of the research. Ethical approval was not required for this bibliometric analysis as it utilized publicly available publication data.

Figure 1.

The flowchart of study identification and selection.

2.2. Data analysis

After verifying the exported data, 2 authors (Ma L and Cao JP) consolidated redundant elements. The refined dataset was subsequently imported into the Bibliometrix R package, VOSviewer 1.6.20 and CiteSpace 6.4. R1 for conducting bibliometric analysis. The Bibliometrix R package, VOSviewer and CiteSpace were primarily responsible for conducting the visual analysis of the data, the publication and citation trends of the literature over the years were generated using Microsoft Excel 365. Bibliometrics is an R software package that offers a comprehensive suite of features for conducting quantitative research in scientometrics. VOSviewer is a robust bibliometric analysis software that facilitates the extraction and processing of data. It is mainly used for visualizing collaborative networks among countries, institutions, authors, and journals, as well as co-citation of keyword clusters. In addition, CiteSpace is an extensively used bibliometric analysis software that delivers an easy-to-understand comprehension of research hotspots and evolution processes in specific fields, thereby providing insights into future directions for development. CiteSpace is especially useful for analyzing citation bursts and keyword bursts as a means of identifying research hotspots. Additionally, it offers other visual analysis functions, such as clustering for publication data, drawing keyword timeline graphs, and more, that help researchers gain insight into a discipline’s past and present life. In our study, we utilized VOS viewer to analyze country/region distribution, institution distribution, author collaboration and distribution, as well as keyword distribution and collaboration. Simultaneously, we employed CiteSpace to analyze the dual-map overlay of journals, reference collaboration and distribution, literature bursts, and keyword bursts.

3. Results

3.1. Database search

In this study, a total of 5709 publications were retrieved. These publications involved the contributions of 27,001 authors affiliated with 4001 institutions across 76 countries. They were published in 9596 different journals and referenced a total of 179,285 articles from distinct journals.

3.2. Annual publications and citation trends

To understand the research progress of microglia in the field of IS, we analyzed the number of publications and citation trends of related papers between 1998 and 2024. Figure 2 illustrates the changes in the annual number of publications and their citations in this field, revealing the dynamics of this research direction. In the early stage (1998–2004), the number of relevant annual publications remained at a low level, indicating that the exploration of the mechanism of microglia’s role in IS was relatively limited during this period. From 2005 onwards, the number of publications began to increase gradually, reflecting the gradual recognition of the importance of microglia in the pathophysiological process of stroke. This growth trend accelerated significantly after 2015, reaching a peak (500 publications) by 2024, which fully reflects the dynamic development of the field in recent years. Citation analysis further highlights the academic impact of these studies: each paper was cited an average of 60 times, including 6 ultra-highly cited papers (≥1500 citations per paper). The phenomenon of simultaneous increase in the number of publications and citations not only indicates the increase of scientific output in this field, but also indicates that microglia research in IS continues to be emphasized by the academic community, and its clinical value and mechanistic research significance are becoming more and more prominent. Overall, these data indicate that microglia research in IS has entered a period of rapid development, and the related results not only promote the development of stroke neuroscience, but also provide an important basis for the development of potential therapeutic strategies.

Figure 2.

Annual publications and citation trends on research of microglia in IS. IS = ischemic stroke.

3.3. Distributions of countries/regions

The current landscape of research on microglia in IS reveals contributions from 76 countries or regions worldwide, with the majority of research activity concentrated in the Northern Hemisphere. Notably, collaborations between these countries and regions are also predominantly located within the Northern Hemisphere. However, Australia, situated in the Southern Hemisphere, stands out for its relatively high level of activity in this field, characterized by frequent communication with other nations (Fig. 3A). As detailed in Table 1, China leads globally with 2251 published papers, followed closely by the United States with 1494 publications. These figures underscore the significant influence of these 2 nations in advancing research on microglia and IS. Other key contributors include Germany (500 publications), South Korea (424), and Japan (408). In terms of overall collaboration intensity, the United States ranks 1st with a total link strength of 771, followed by China (462), Germany (391), the United Kingdom (216), and Australia (166).

Figure 3.

Distributions of countries/regions. (A) Countries/regions collaboration map. (B) Distributions of countries/regions and collaboration network.

Table 1.

Top 10 most publication countries/regions related to microglia in IS.

Rank	Country/region	Count	Total link strength
1	Peoples R China	2251	462
2	USA	1494	771
3	Germany	500	391
4	South Korea	424	128
5	Japan	408	146
6	Italy	192	149
7	England	185	216
8	Australia	137	166
9	France	137	150
10	Sweden	114	136

IS = ischemic stroke.

To further analyze international collaboration patterns, we employed VOSviewer for visualization (Fig. 3B). Countries and regions were grouped into 10 clusters based on their level of collaboration. The analysis highlights a high degree of cooperation among nations, with the United States and China emerging as the core nodes of the global collaboration network due to their extensive connections with other countries. Additionally, Germany, the United Kingdom, and Japan demonstrate strong collaborative ties, playing pivotal roles in driving research progress in the field of microglia and stroke. This visualization reinforces the importance of international collaboration in advancing scientific understanding of microglial mechanisms in IS.

3.4. Analysis of affiliations

On the basis of the country distribution analysis, we conducted a systematic assessment of the contribution of research institutions. As illustrated in Table 2, the top 10 institutions are ranked by their publication output, with 5 institutions surpassing the 100-article threshold. Capital Medical University leads the list with 141 publications, trailed by Shanghai Jiao Tong University (121), Hallym University (118), Fudan University (113), and Zhejiang University (105). Notably, Hallym University stands out for its highest total link strength, reflecting robust collaborative ties within the field. Many of these leading institutions are based in China and South Korea, underscoring the pivotal role of these nations in advancing research on microglia and IS. Additionally, institutions such as the University of Pittsburgh and the University of California, San Francisco also feature prominently, underscoring the collaborative nature of this research domain.

Table 2.

Top 10 most publication institutions related to microglia in IS.

Rank	Organization	Count	Total link strength
1	Capital Med Univ	141	259
2	Shanghai Jiao Tong Univ	121	113
3	Hallym Univ	118	302
4	Fudan Univ	113	144
5	Zhejiang Univ	105	109
6	Nanjing Univ	88	153
7	Univ Calif San Francisco	86	118
8	Univ Pittsburgh	86	132
9	Seoul Natl Univ	84	220
10	Kangwon Natl Univ	82	236

IS = ischemic stroke.

We next analyzed the trend of collaboration among research institutions for articles related to microglia and ischemic stroke. The results show that there is a continuous and close collaboration between research institutions. Among them, institutions such as Capital Medical University, Fudan University, and Shanghai Jiaotong University have formed a collaboration network centered on themselves, further consolidating their influence. (Fig. 4A). In addition, collaborative networks among research institutions are important for promoting knowledge sharing and accelerating research progress, and for this reason we visualized and analyzed the collaborative networks using VOSviewer (Fig. 4B). In the visualization graph, institutions such as the University of South Florida, the University of Rochester, and Brigham and Women’s Hospital are predominantly shown in blue, indicating their early entry into the research field. In contrast, Chinese institutions such as Shanghai Jiao Tong University, Fudan University and Capital Medical University are predominantly in light red and red, reflecting China’s late but fruitful contributions to the field in recent years. Meanwhile, South Korean institutions such as Hanlin University, Kangwon National University and Seoul National University are presented in darker colors and larger nodes, suggesting that they have become a significant force in the field and are expected to drive future research progress. These dynamic changes highlight the evolution of the institutional contribution landscape and the growing influence of Korean institutions in shaping the direction of microglia and ischemic stroke research.

Figure 4.

Analysis of affiliations. (A) The clustering analysis of the institutions. (B) Network-view map of institutional collaboration.

3.5. Contributions of journals

A total of 5709 articles on microglia in IS have been published across 9596 journals. As shown in Table 3, the top 10 journals are ranked by their publication output. The Journal of Neuroinflammation tops the list with 227 papers, followed by the Journal of Cerebral Blood Flow and Metabolism (192 papers) and Stroke (158 papers). Among the top 10 journals, 7 are classified in the Q1 quartile, and 3 are in Q2. Notably, the Journal of Neuroinflammation also holds the highest impact factor of 9.3, underscoring its significant influence in the field. In the ranking of the top 10 most cited journals, 5 are from Q1 and 4 from Q2, highlighting their substantial impact and recognition in this research domain. Stroke leads in total citations with 16,392, followed by J Cerebr Blood F Met (12,453) and J Neurosci (11,821). This distribution reflects the high level of research achievement in the field and suggests opportunities for further in-depth exploration. A journal’s impact is closely tied to its citation count, as this metric reflects how frequently its articles are referenced by scholars and researchers. Among the top 10 cited journals, Stroke stands out for its highest citation count, emphasizing its pivotal role in advancing research on microglia in IS.

Table 3.

Top 10 journals and co-cited journals related to microglia in IS.

Rank	Journal	Count	IF (2023)	JCR quantile	Co-cited-journal	Citation	IF (2023)	JCR quantile
1	Journal of Neuroinflammation	227	9.3	Q1	Stroke	16,392	7.9	Q1
2	Journal of Cerebral Blood Flow and Metabolism	192	4.9	Q2	J Cerebr Blood F Met	12,453	4.9	Q2
3	Stroke	158	7.9	Q1	J Neurosci	11,821	4.4	Q1
4	Neuroscience	155	2.9	Q2	Brain Res	7968	2.7	Q3
5	Experimental Neurology	120	4.6	Q1	J Neuroinflamm	7188	9.3	Q1
6	Glia	117	5.4	Q1	P Natl Acad Sci USA	7096	9.4	Q1
7	International Journal of Molecular Sciences	115	4.9	Q1	Glia	6533	5.4	Q1
8	Frontiers in Cellular Neuroscience	80	4.2	Q2	Neuroscience	5479	2.9	Q2
9	Journal of Neuroscience	73	4.4	Q1	J NeuroChem	5667	4.2	Q2
10	Frontiers in Immunology	65	5.7	Q1	J Biol Chem	4900	4.0	Q2

IF= impact factor, IS = ischemic stroke, JCR = Journal Citation Reports.

To systematically assess the scholarly contributions of journals in the field of microglia and IS research, we analyzed the journal clustering characteristics, co-citation relationships, and temporal evolution trends through a multidimensional visualization approach. Journal clustering density analysis (Fig. 5A) showed that the research literature showed obvious clustering characteristics in journal distribution. High-density areas such as red and yellow signify the core knowledge aggregation areas in the field, and journals such as Stroke and J Cerebr Blood F Met are located in these hotspot areas, indicating that they have formed a close academic community through high-frequency cross-citation or convergence of research directions, which provides an objective basis for identifying the core journals in the field. The network analysis of co-cited journals (Fig. 5B) further reveals the knowledge association among journals. Different colors represent heterogeneous clusters, and core journals such as Stroke and J Neurosci occupy the network hub position, and their dense connecting lines indicate that these journals are widely cited in microglia-IS research and play the role of key nodes connecting different research directions. The cross-field citation path analysis (Fig. 5C) highlights the cross-disciplinary characteristics. The left and right color blocks in the figure represent different research fields, such as molecular biology and immunology, and the colored citation curves in the middle of the figure show the obvious direction of knowledge flow-especially, the yellow path shows that research in the field of molecular genetics is more inclined to cite the results of molecular immunology, which confirms that multidisciplinary integration has played an important role in promoting the development of the field. The time-series evolution of journal contributions (Fig. 5D) documents the cumulative trend of publications across journals over the period 1998 to 2024. The steep upward curve of Stroke, J Cerebr Blood F Met reflects its continuous growth in field influence; J Neuroinflamm, J Neurosci has grown significantly faster in recent years, demonstrating its increasing academic status; and journals such as Brain Res, Neuroscience have grown at a flat rate but maintain a steady intellectual output.

Figure 5.

Contributions of journals. (A) Density map of journal clustering analysis of microglia in IS. (B) Co-cited journal clustering analysis of microglia in IS. (C) The dual-map overlay of journal publishing research. (D) Journal/Sources’ Production over time. IS = ischemic stroke.

Overall, these visualizations provide a comprehensive analysis of journal contributions in the study of microglia in IS. By examining journal clustering, citation relationships, and temporal trends, they offer a multidimensional perspective on how different journals contribute to the development of this research field. This analysis highlights the roles of journals in terms of their influence, research focus, and the evolution of research output over time.

3.6. Authors and co-cited authors

Research on microglia in IS has engaged a total of 27,001 authors. As shown in Table 4, the top 10 authors are ranked by publication quantity and co-citation frequency, underscoring key scholars and their impact on this research domain. Moo-Ho Won leads with 71 publications, reflecting significant involvement in this area. In Koo Hwang (53 publications) and Joon Ha Park (49 publications) also demonstrate substantial research output. In terms of total link strength, Moo-Ho Won (1651) and In Koo Hwang (1363) exhibit strong collaborative ties with other researchers. Among the top 10 co-cited authors, Iadecola, C. stands out with 965 citations and a total link strength of 29,465, underscoring his pivotal role in the field. Hu, X.M. (955 citations) and Wang, J. (587 citations) also wield significant influence. Most co-cited authors in the top 10 have been cited hundreds of times, reflecting their notable contributions to research on microglia in IS.

Table 4.

Top 10 authors and co-cited authors related to microglia in IS.

Rank	Author	Count	Total link strength	Co-cited author	Citation	Total link strength
1	Moo-Ho Won	71	1651	Iadecola, C.	965	29,465
2	In Koo Hwang	53	1363	Hu, X.M.	955	27,320
3	Joon Ha Park	49	818	Wang, J.	587	16,133
4	Ji Hyeon Ahn	42	604	Lambertsen, K.L.	513	15,239
5	Jae-Chul Lee	40	612	Dirnagl, U.	499	14,426
6	Jun Hwi Cho	40	706	Ransohoff, R.M.	422	15,723
7	Ki-Yeon Yoo	35	1216	Gelderblom, M.	390	15,079
8	Young-Myeong Kim	34	1050	Lalancette-Hébert, M.	386	12,531
9	Jung Hoon Choi	34	1042	Liesz, A.	379	14,844
10	Dae Won Kim	33	438	Hanisch, U.K.	373	14,184

In order to systematically reveal the characteristics of academic collaborations and research lineage of microglia in ischemic stroke research, this study used VOSviewer to visualize and analyze the author collaboration network and co-citation relationship. Figure 6A shows the results of the visualization of the author collaboration network. Through clustering analysis, the researchers formed 11 distinctive collaboration clusters, each labeled with a different color. In the network diagram, the node size is positively correlated with the authors’ publication volume and academic influence, in which scholars such as “Xu, Yun” and “Chen, Jun” occupy the center of the network, and the nodes are significantly larger than those of other researchers, which indicates that they have an important academic status and extensive research collaboration in this research field. The connectivity between nodes reveals the collaborative relationship between authors, and the dense connectivity within clusters shows a stable teamwork pattern; the simultaneous existence of cross-cluster connections reflects the academic exchanges between different research teams. This network structure clearly demonstrates the research ecology of the field that maintains team-independent research characteristics and promotes cross-team collaboration. Figure 6B presents the results of the analysis of the author co-citation network. There are 3 main clusters in the network: the blue cluster with “Iadecola, C” as the core shows the most extensive connectivity and not only maintains dense co-citation within the cluster, but also establishes the most connections with other clusters; the red cluster represented by “Hu, XM” shows the highest internal connectivity; and the green cluster with “Barone, FC” as the hub is relatively independent, but still has connections with other clusters. The green cluster, with “Barone, FC” as its hub, is relatively independent, but still maintains moderate interactions with the other clusters. Among these co-cited authors, “Iadecola, C” has the highest co-citation frequency, followed by “Hu, XM” and “Dirnagl, U,” who, through their pioneering research results, have become representative figures in the field, and whose work has laid an important foundation for subsequent research. Through these 2 network analyses, we can see that the field has formed a research pattern led by a number of core scholars, with multiple teams collaborating to innovate. This cooperation pattern, which maintains research characteristics and promotes academic exchange, has effectively promoted the in-depth development of microglia in the field of ischemic stroke research. The results of the study not only reveal the current research cooperation dynamics, but also provide an important reference for the possible future direction of academic cooperation.

Figure 6.

Authors and co-cited authors. (A) Coauthor related to microglia in IS. (B) Network of co-cited authors. IS = ischemic stroke.

3.7. Exploring the landscape of stroke and microglia research

The quadrant map in Figure 7A visualizes the interplay between the relevance (centrality) and development (density) of research themes in stroke and microglia studies. This map is divided into 4 quadrants by 2 dashed lines. The upper-right quadrant, labeled Motor Themes, highlights themes that are both highly relevant and well-developed, serving as key drivers of the field’s progress. By contrast, the upper-left quadrant, Niche Themes, showcases topics that have advanced significantly but remain less central to the broader research agenda. Themes like cerebral ischemia, expression, and activation are prominent here, exhibiting a high level of maturity. More refined research can be conducted on this basis to consolidate one’s advantages in the niche field.

Figure 7.

Exploring the landscape of stroke and microglia research. (A) Quadrant map of research themes in stroke and microglia studies. (B) Sankey diagram of journals, themes and authors in stroke-microglia research.SO = the leftmost for journals, ID = the central for research themes, AU = the rightmost representing authors.

In the bottom right quadrant of the research focus, basic research topics such as “focal cerebral ischemia,” “central nervous system,” and “microglia activation” are concentrated. These key issues, as the core foundation for the development of the field, have already demonstrated a high degree of academic relevance and research value. However, there is still significant room for exploration and potential for further development. Given their profound theoretical significance and practical value, it is recommended that researchers conduct systematic and in-depth mechanistic studies and innovative explorations on these basic and urgently needed research topics. This will further uncover their underlying scientific connotations, thus laying a solid foundation for promoting theoretical breakthroughs and technological innovations in this field. Meanwhile, the lower-left quadrant pinpoints themes that are either gaining or losing ground in terms of significance, categorized as emerging or declining themes.

Figure 7B presents a Sankey diagram that elucidates the connections between journals, themes, and authors in stroke-microglia research. This diagram is organized into 3 primary columns: the leftmost (SO) for journals, the central (ID) for research themes, and the rightmost (AU) representing authors. Each column is composed of colored rectangular bars, denoting various entities. The connecting lines between these bars visualize the relationships among journals, themes, and authors, where the thickness of the lines suggests the intensity or frequency of these connections.

The diagram unveils, for instance, that the Journal of Neuroinflammation is tied to the theme of focal cerebral ischemia and features the author Wang J. Likewise, the Journal of Cerebral Blood Flow and Metabolism relates to the theme of expression and involves the author Zhang Y. This graphic representation offers a comprehensive view of the collaborative landscape and research scope within stroke and microglia studies, shedding light on how various journals, themes, and authors intersect and advance the field.

3.8. References and co-cited references

The analysis of literature co-citation (refer to Fig 8A and B) revealed interconnected themes in microglia and IS research. Utilizing CiteSpace, we constructed a keyword-based clustering network (illustrated in Fig. 8C), identifying 9 distinct co-reference clusters. The high modularity (Q = 0.7846) and silhouette scores (weighted mean = 0.9121) attest to the clustering structure’s validity and stability. These clusters uncovered key research themes. Cluster #0, the largest, revolves around cerebral ischemia-reperfusion injury, Toll-like receptor signaling, and inflammatory responses, shedding light on brain injury mechanisms and immune activation. Cluster #1 focuses on cannabinoid receptor-mediated neuroprotection and oxidative stress, hinting at potential therapeutic paths for ischemic damage. Cluster #2 emphasizes microglia, neuroinflammation, and neurodegeneration, mirroring investigations into microglia’s role in chronic neurological disorders. Cluster #3 delves into delayed microglial treatment, the subventricular zone, and neurogenesis, highlighting an interest in poststroke regenerative processes. Cluster #4 spotlights the peripheral benzodiazepine receptor, ethyl pyruvate, and antiinflammatory effects, pointing towards novel pharmacological targets. Temporal trends indicate a growing emphasis on gut microbes, microglia, and mitochondria, signaling a shift in research priorities in recent years.

Figure 8.

References and co-cited references. (A) Literature co-citation network of microglia in IS (by author and year). (B) Literature co-citation network of microglia in IS (by title and year). (C) Cluster view of references on microglia in IS. (D) CiteSpace visualization map of top 25 references with the strongest citation bursts. IS = ischemic stroke.

Figure 8D exhibits the top 25 references with the most notable citation bursts, highlighting their significant contributions to the field. Notably, Hu XM et al “Neuroinflammation in IS: New Developments and Therapeutic Opportunities” (2012) displayed a burst strength of 47.12 from 2012 to 2016, underscoring its foundational importance. Patel AR et al “Microglial Activation in IS” (2013) showed a burst strength of 42.44 from 2013 to 2017, reflecting its impact on comprehending acute neuroinflammatory responses. Ritzel RM et al “Microglial Phenotypes in Acute and Resolving Brain Injury” (2015) demonstrated a lasting influence (burst strength = 33.69, 2015–2019), shaping discussions on microglial polarization. These bursts outline evolving research fronts and underscore seminal works that continue to steer the field.

The combination of co-citation clustering and burst detection provides a rigorous method for mapping research trends. By identifying key themes and influential literature, this analysis assists researchers in aligning with emerging directions and harnessing high-impact studies to advance microglial research in IS.

3.9. The analysis of hotspots and frontiers

3.9.1. Keywords co-occurrence and cluster analysis

In the keyword clustering analysis (Fig. 9A), 5 distinct research themes emerge. The Green Cluster, with its emphasis on “stroke,” “inflammation,” “neuroinflammation,” and “IS,” underscores the involvement of microglia in cerebrovascular inflammation and poststroke neuroinflammatory responses. The Blue Cluster delves into molecular mechanisms, as evidenced by keywords like “expression,” “apoptosis,” “brain,” and “ischemia,” reflecting research on gene regulation and cell death pathways in ischemic brain injury. The Red Cluster highlights studies on ischemia subtypes and their neurological implications, featuring terms such as “cerebral ischemia,” “focal cerebral ischemia,” “central nervous system,” and “brain injury.” Supplementary keywords, including “neurons,” “neuronal death,” “minocycline,” and “nitric oxide synthase,” broaden the scope to encompass neuroprotection, therapeutic targets, and microglia-related signaling. This clustering illuminates current research frontiers, particularly the dual roles of microglia in neuroinflammation and ischemic injury, paving the way for future investigations.

Figure 9.

The analysis of hotspots and frontiers. (A) The clustering of keywords in IS. (B) A map of the conceptual structure of the top 50 terms in terms of word frequency. (C) Keyword timeline viewer. (D) CiteSpace visualization map of top 25 keywords with the strongest citation bursts. IS = ischemic stroke.

The multiple correspondence analysis map (see Fig. 9B) further distills these themes by plotting keywords along dim1 and dim2 axes, which represent feature accumulation percentages. Four thematic groups stand out: the Green Cluster focuses on postocclusion pathophysiology with terms like “Artery occlusion,” “reperfusion injury,” “cell death,” and “brain injury.” The Purple Cluster centers on molecular responses to ischemia, featuring “Focal cerebral ischemia,” “gene expression,” “nitric oxide,” and “oxidative stress.” The Red Cluster highlights preclinical models and neuroprotective strategies, including “Stroke,” “rats,” “apoptosis,” “inhibition,” and “neuroprotection.” Lastly, the Light Blue Cluster bridges neurodegenerative and cerebrovascular research with terms such as “Mouse model,” “in vivo,” “Alzheimer disease,” and “IS.” The proximity of keywords on this map indicates conceptual affinities, aiding researchers in quickly identifying interrelated topics. Ultimately, this multiple correspondence analysis map serves as a valuable tool for navigating the thematic landscape of the field.

3.9.2. Keywords timeline viewer

The keyword co-occurrence network, constructed using VOSviewer in Figure 9C, utilizes a color gradient ranging from blue to red to visually represent the temporal progression of research focal points. The spatial arrangement and hue variations of key nodes, including “microglia,” “stroke,” and “inflammation,” illuminate the dynamic evolution of research priorities in this domain. In terms of Chronological Development of Research Foci, the Early Foundational Research Phase, represented by the Blue Spectrum, features terms like “ischemia” and “apoptosis” in a distinct blue hue, denoting their establishment as foundational concepts circa 2012. The grouping of keywords such as “cell death” and “gene expression” underscores the initial focus on the molecular underpinnings of ischemic injury. Moving into the Intermediate Developmental Phase, represented by the Blue-Green Transition, terms like “cerebral ischemia” and “focal cerebral ischemia” gain prominence around 2015, signaling a shift towards exploring subtype-specific mechanisms. The increased density of the network during this phase highlights thematic diversification. Finally, in the Recent Cutting-Edge Directions, represented by the Red Spectrum, the predominantly red-hued terms “neuroinflammation” and “functional recovery” underscore the current emphasis on neuroprotection and functional restoration. Emergent keywords like “regulatory T cells” and “angiogenesis” hint at new frontiers in immunomodulation and vascular remodeling research. Centrality metrics from the Network Topology Analysis indicate that the “microglia” node exhibits the highest betweenness centrality (0.48), affirming its critical role in bridging diverse research themes. Additionally, node size-frequency analysis reveals notable recent spikes in terms such as “neuroprotection” (2017–2018) and “mechanistic studies” (2019). This analytical approach not only charts the field’s knowledge trajectory but, more significantly, pinpoints emerging nodes, including rapidly growing yet unlabeled terms like “exosomes” and “gut microbiota,” to forge a predictive roadmap for research directions in the next 5 to 10 years.

3.9.3. Burst detection

Bibliometric analysis of the brain injury research domain reveals distinct thematic clusters and temporal evolution patterns, as evidenced by keyword frequency and citation burst analyses. The frequency distribution of keywords in Table 5 demonstrates that “microglia” emerges as the predominant research focus, with 2383 occurrences and a total link strength of 22,257, establishing its central position in current investigations. This is followed closely by “stroke” (1522 occurrences) and “inflammation” (1471 occurrences), which collectively delineate the principal axes of research inquiry. The significant representation of “focal cerebral ischemia” (1130 occurrences) and “cerebral ischemia” (1100 occurrences) underscores the continued importance of ischemia-related pathophysiology in the field.

Table 5.

Top 20 keywords related to microglia in IS.

Rank	Keyword	Occurrences	Total link strength	Rank	Keyword	Occurrences	Total link strength
1	Microglia	2383	22,257	11	Injury	809	7393
2	Stroke	1522	14,106	12	Neuroprotection	778	7523
3	Inflammation	1471	13,943	16	Oxidative Stress	561	5154
4	Focal Cerebral Ischemia	1130	10,633	13	Cells	510	4650
5	Cerebral Ischemia	1100	9884	15	Central Nervous-System	505	4835
6	Neuroinflammation	1046	9673	14	Microglial Activation	502	4696
7	Expression	952	8894	17	Brain Injury	494	4722
8	Activation	951	8801	18	Mechanisms	472	4309
9	IS	843	7047	19	Apoptosis	467	4408
10	Brain	833	7672	20	Apoptosis	467	4408

IS = ischemic stroke.

The temporal analysis of citation bursts (Fig. 9D) reveals a clear trajectory of conceptual development. Early research (1998–2009) was characterized by foundational studies, as exemplified by the strong citation burst for “rat brain” (burst strength = 51.24). In contrast, contemporary research (2016–2024) has shifted toward more sophisticated investigations of “microglial polarization” and “NLR family pyrin domain containing 3 (NLRP3) inflammasome” activation, reflecting an evolution from descriptive to mechanistic approaches. Specifically, burst detection identified “NLRP3 inflammasome” (burst strength = 16.81, 2020–2024) and “EVs” (burst strength = 16.84, 2020–2024) as the most prominent recent frontiers. Particularly noteworthy is the emergence of novel concepts such as “EVs,” indicating potential new frontiers in the field.

Integrated analysis of the data presented in Table 5 and Figure 9D clearly demonstrates 3 fundamental paradigm shifts in brain injury research. A transition from macroscopic characterization to in-depth investigation of microscopic mechanisms. An expansion from single-factor analysis to exploration of multi-system interactions. A progression from basic research to clinical translation. These transformative trends are not only reflected in keyword frequency distributions but are also evidenced by temporal patterns in citation burst strength. Current research priorities particularly focus on 3 key areas: dynamic regulation of microglial activity, molecular mechanisms of neuroinflammation, and discovery of novel therapeutic targets-research directions that are projected to maintain their scientific prominence over the next 5 to 10 years.

4. Discussion

4.1. General research overview

The study employed bibliometric methods to systematically review 5709 research articles on microglia in IS published between 1998 and 2024. Combined with multidimensional visual analysis, it provided a deep insight into the disciplinary development trajectory and academic ecological landscape in this field. The findings revealed that China ranked 1st globally in the number of publications, which strongly attested to its rising academic influence in this domain. Leading research institutions such as Capital Medical University and Hallym University stood out due to their abundant research outputs, while core journals like the “Journal of Neuroinflammation” and “Stroke” served as crucial platforms for disseminating achievements and facilitating academic exchanges. Scholars like Moo-Ho Won and C. Iadecola played pivotal roles in advancing the field, thanks to their highly influential academic contributions.

The study indicated that the field’s development exhibited distinct 3-phase evolution characteristics. The period from 1998 to 2004 was marked as the basic exploration phase, where the annual publication volume remained relatively stable. Research during this phase primarily focused on fundamental mechanisms, and a significant scale effect had not yet formed. Between 2005 and 2014, disciplinary development entered a phase of theoretical breakthroughs, driven by the introduction of the M1/M2 phenotype polarization theory of microglia and the confirmation of the NLRP3 inflammasome’s critical role in neuroinflammation. After 2015, the number of research achievements grew exponentially. The widespread application of single-cell sequencing technology successfully untangled the heterogeneous subpopulations of microglia, while emerging research directions like exosomes significantly expanded the disciplinary boundaries.

In terms of journal analysis, core journals represented by the “Journal of Neuroinflammation” and “Stroke” played an irreplaceable role in academic leadership by continuously publishing high-quality research outcomes. On the other hand, emerging journals provided a broad platform for interdisciplinary crossover studies thanks to their openness and innovativeness. Through visual analysis of dimensions such as publication quantity, citation frequency, journal partitioning, and collaboration networks, the differentiated contributions of various journals in the disciplinary development process became clearly apparent. The international collaboration network exhibited a geographical distribution characterized by “Sino-US dominance and Eurasian follow up.” The technological complementarity among regions drove a shift in cooperation models from the traditional “single-center dominance” to “multi-polar synergy.” This transformation significantly enhanced research efficiency and powerfully promoted the translation of basic research achievements into clinical applications.

4.2. Research hotspots and frontiers

Key themes identified through keyword co-occurrence and citation bursts include neuroinflammation, microglial activation, and ischemic injury mechanisms. High-frequency keywords like “microglia,” “stroke,” and “inflammation” reflect sustained interest in microglia’s dual role in IS: protective phagocytosis vs deleterious inflammation. Recent citation bursts for “microglia polarization” (2019–2024), “NLRP3 inflammasome” (2020–2024), and “EVs” (2020–2024) indicate emerging frontiers, aligning with studies demonstrating microglia’s phenotypic plasticity and their contribution to inflammatory cascades via NLRP3 activation and vesicular communication (as detailed in Table 6).

Table 6.

Summary of key evidence supporting emerging research frontiers (burst keywords).

Burst keyword (strength, period)	Key supporting references (examples)	Study models	Major findings/mechanistic insights	Potential therapeutic implication
Microglia polarization (15.28, 2019–2024))	6. Li Y, et al Neural Regen Res. 2025 22. Li F, et al Metab Brain Dis. 2018	In vitro BV2/HMC3 cells; Rodent MCAO model	Microglia shift from pro-inflammatory M1 (iNOS+, CD86+) to antiinflammatory/reparative M2 (Arg1+, CD206+) phenotype poststroke. M1 exacerbates injury via ROS/TNF-α; M2 promotes repair via BDNF/TGF-β. Phenotype is spatiotemporally dynamic.	Pharmacological (e.g., L-3-n-butylphthalide) or genetic modulation to promote M2 polarization.
NLRP3 inflammasome (16.81, 2020–2024)	6. Li Y, et al Neural Regen Res. 2025 33. Ding J, et al Cell. 2017 34. Yang Y, et al Cell Death Dis. 2019	In vitro priming + activation assays; Rodent MCAO model; NLRP3 knockout mice	A multiprotein complex activated by DAMPs (e.g., K + efflux, mtDNA). Leads to Caspase-1 activation, IL-1β/IL-18 maturation, and pyroptosis (GSDMD cleavage), driving secondary inflammation.	Specific small-molecule inhibitors (e.g., MCC950) show neuroprotection in preclinical models.
Extracellular vesicles (16.84, 2020–2024)	53. Zhang Q, et al J Nanobiotechnol. 2025 51. Song Y, et al Theranostics. 2019 52. Song Y, et al Stroke. 2023	In vitro cell co-culture; Rodent MCAO model with EV injection; Human plasma samples	EVs (especially from M2 microglia) carry miRNAs (e.g., miR-124, miR-25-3p) and proteins that modulate neuroinflammation, promote NSC differentiation, and protect the BBB. Act as intercellular communicators.	Engineered EVs as natural nanocarriers for drug/miRNA delivery (“cell-free therapy”). Diagnostic biomarkers.

Arg1 = arginase 1, BBB = blood-brain barrier, BDNF = brain-derived neurotrophic factor, DAMPs = damage-associated molecular patterns, EVs = extracellular vesicles, GSDMD = Gasdermin D, IL = interleukin, iNOS = inducible nitric oxide synthase, MCAO = middle cerebral artery occlusion, NLRP3 = NLR family pyrin domain containing 3, NSC = neural stem cell, IS = ischemic stroke, ROS = reactive oxygen species, TNF-α = tumor necrosis factor-alpha, TGF-β = transforming growth factor beta.

Notably, microglial polarization emerged as a critical trend. Pro-inflammatory (M1) and antiinflammatory (M2) phenotypes influence ischemic outcomes, with M1 polarization linked to exacerbated neuronal damage and M2 polarization associated with neuroprotection. This aligns with our finding that “neuroprotection” and “oxidative stress” are recurring themes, suggesting therapeutic strategies targeting microglial polarization may hold promise. NLRP 3 inflammasome activation also garnered attention, as it mediates caspase-1 activation and interleukin secretion, driving neuroinflammation in ischemic injury. This mirrors its role in IS’s disease, indicating conserved inflammatory pathways across neurodegenerative conditions. Additionally, “EVs” as a burst keyword highlights their emerging role in intercellular communication, potentially transporting pathological cargo (e.g., pro-inflammatory mediators) or facilitating neurorepair.

4.2.1. Phenotypic polarization and the dynamic balance of neuroinflammation

4.2.1.1. Spatiotemporal polarization characteristics of microglia in IS

Polarization of microglia after IS shows significant time-dependence and spatial heterogeneity.^[14] In the acute phase (0–72 hours), microglia in the core area of injury rapidly polarize toward a pro-inflammatory M1 phenotype, highly express surface markers such as CD86, inducible nitric oxide synthase (iNOS), and CD32, and release pro-inflammatory factors such as tumor necrosis factor-α (TNF-α), interleukin-1β (IL-1β) and other pro-inflammatory factors.^[15–17] In contrast, during the subacute phase (3–14 days), microglia in the semidark band region gradually switched to a reparative M2 phenotype, with high expression of CD206, Arginase1, and Ym1, and secretion of transforming growth factor-β (TGF-β) and insulin-like growth factor 1 and other antiinflammatory factors.^[18,19] Notably, single-cell transcriptomic studies have revealed the complexity of microglia polarization states beyond the traditional M1/M2 dichotomy.^[20] 2025 The team of Chou-Ping Tang^[6] found that there are multiple intermediate state cell subpopulations in poststroke brain tissues such as lipid droplet-rich microglia (LDRM) These cells express myeloid triggering receptor 2 (TREM2) and lipid droplet inclusion protein 2 (PLIN2), which act as a bridge between lipid metabolism disorders and inflammatory responses, and the SORLA gene (sorting protein-associated receptor) affects the phagocytosis of microglia by regulating the TREM2 signaling pathway, which is not entirely dependent on the classical M1/M2 classification framework. In addition, a comprehensive review has systematically argued for targeting microglial polarization as a novel therapeutic strategy against ischemic stroke, summarizing various compounds with polarization-modulating effects.^[21] This further underscores that regulating phenotypic switching is a key focus of translational research.

4.2.1.2. Mechanisms of neurotoxicity

Aberrant activation of M1 microglia can exacerbate the neurological injury process through multidimensional mechanisms.^[22] The 1st is the initiation of the oxidative stress cascade, in which dysfunction of the mitochondrial electron transport chain leads to electron leakage and triggers the explosive generation of reactive oxygen species (ROS) such as superoxide anion (O²⁻-). This oxidative stress state directly attacks the phospholipid bilayer of the neuronal cell membrane, inducing lipid peroxidation and causing DNA strand breaks and base damage, ultimately triggering the neuronal apoptotic program.^[23] Secondly, the positive feedback amplification of the inflammatory network, the TNF-α and IL-1β secreted by M1-type cells activate nuclear factor kappa-light-chain-enhancer of activated B cells (NFκB) through the classical signaling pathway, and this transcription factor enters the nucleus to drive the transcriptional activation of inflammation-related genes, such as COX-2, iNOS, etc, to form a positive feedback loop of pro-inflammatory factor-signal transduction-gene expression. This transcription factor enters the nucleus and drives the transcriptional activation of COX-2, iNOS and other inflammation-related genes, forming a “pro-inflammatory factor-signal transduction-gene expression” positive feedback loop. At the same time, these cytokines destroy the physical barrier between endothelial cells by degrading the tight junction protein complex (e.g., claudin-5 and zonula occludens-1) of the blood-brain barrier (BBB), increasing vascular permeability, which then mediates the abnormal infiltration of peripheral immune cells, such as monocytes and neutrophils, and the formation of “central-peripheral inflammation.” The 3rd aspect is the mechanism of excitotoxic injury, in which M1-type microglia overactivate NMDA receptors on the surface of neurons by abnormally releasing glutamate, which leads to an elevated concentration of this neurotransmitter in the synaptic gap compared to physiologic levels. This process triggers an overload of calcium ion (Ca²⁺) in-flow, activating calcium-dependent proteases, phospholipases, and nucleic acid endonucleases, which ultimately leads to mitochondrial membrane potential collapse, ATP depletion, and acute neuronal necrosis.^[24] Notably, this excitotoxic injury is particularly pronounced in the ischemic semidark band region and positively correlates with the degree of neurologic deficit in clinical patients.

4.2.1.3. Neuroprotective mechanisms

M2 microglia exert neuroprotective and reparative effects through multiple pathways. At the level of inflammation regulation, M2 microglia can secrete antiinflammatory factors such as IL-10 and TGF-β, which inhibit the activation of the NFκB signaling pathway and block the transcription and release of pro-inflammatory mediators such as TNF-α and IL-1β^[25,26]; at the same time, their synthesized antiinflammatory mediators such as resolvins can actively promote the termination of inflammatory response and accelerate the process of tissue repair. In terms of neurotrophic support, the brain-derived neurotrophic factor (BDNF) and glial cell line-derived neurotrophic factor (GDNF) secreted by M2 microglia can accelerate the process of neuronal repair by interacting with the tropomyosin receptor kinase B receptor on the surface of neurons. BDNF and GDNF, which can activate downstream signaling pathways through specific binding to tropomyosin receptor kinase B receptors on the surface of neurons, thereby enhancing synaptic plasticity and promoting neuronal survival, and decreasing neuronal apoptosis under pathological conditions.^[27] In microenvironmental remodeling, M2 microglia efficiently remove cellular debris and denatured proteins after nerve injury through TREM2-mediated phagocytosis, which often forms a physical and biochemical barrier to nerve regeneration, and this phagocytosis significantly improves the microenvironment of nerve regeneration, creating favorable conditions for axon regeneration and synapse reconstruction.^[28,29] The function of phagocytosis can significantly improve the microenvironment of nerve regeneration, creating favorable conditions for axonal regeneration and synapse reconstruction.

M2 microglia constructed a multi-level neuroprotective network through the bidirectional regulation of antiinflammatory factors (inhibition of pro-inflammatory and promotion of antiinflammatory), transcellular signaling of neurotrophic factors, and active clearance of damage-related substances. This mechanism not only reveals the key link of endogenous repair in the CNS, but also provides a potential intervention strategy for the treatment of IS and neurodegenerative diseases with microglia phenotypic regulation as the core.^[30]

4.2.2. Metabolic reprogramming drives neuroinflammation

4.2.2.1. Mitochondrial ysfunction

Under pathological conditions such as ischemia, mitochondrial kinetic homeostasis in microglia suffers disruption. Under normal physiological conditions, mitochondria maintain a balance of morphology and function through fusion and division, whereas after ischemia, the expression of the division protein dynamin 1 is significantly upregulated, while the expression of the fusion proteins mitochondrial fusion proteins 1/2 is downregulated, an imbalance that leads to mitochondrial fragmentation. The altered mitochondrial morphology directly affects the efficiency of its energy metabolism, with inhibition of the oxidative phosphorylation process and insufficient cellular energy production.^[31,32] Mitochondria with impaired function also trigger a series of adverse consequences. Damaged mitochondria release mitochondrial DNA (mtDNA) as well as ROS, which activate the NLRP3 inflammasome, and activation of the NLRP3 inflammasome further contributes to the activation of caspase-1, which in turn mediates the activation of IL-1β. The activation of NLRP3 inflammasome further promotes the activation of caspase-1, which mediates the maturation and secretion of IL-1β, triggering an inflammatory response.^[33,34] Of interest, the study^[35] found that peroxisome proliferator-activated receptor γ coactivator 1α (PGC1α), a key regulator of mitochondrial biogenesis, plays an important role in this pathology, and its downregulation can exacerbate oxidative stress damage and accelerate mitochondrial dysfunction.^[36] In contrast, activation of PGC1α through the use of metformin can effectively restore mitochondrial function and reduce inflammatory responses, providing a new idea for therapeutic intervention.

Mitochondrial dysfunction in microglia after ischemia forms a vicious cycle from imbalance of mitochondrial dynamics, abnormal energy metabolism, to activation of inflammatory signaling pathways, which plays a central role in metabolic reprogramming driving neuroinflammation. Changes in the expression of dynamin-related protein 1 and mitochondrial fusion proteins 1/2, the release of mtDNA and ROS, as well as the regulatory mechanism of PGC1α together constitute this complex These findings not only deepen our understanding of the neurological These findings not only deepen our understanding of the mechanism of neuroinflammation, but also provide an important theoretical basis and potential intervention targets for neuroinflammation therapeutic strategies targeting mitochondrial function.

4.2.2.2. Enhanced glycolysis

In the pathological process of neuroinflammation, adaptive shifts in cellular metabolic patterns are important intrinsic mechanisms driving the escalation of inflammatory responses. Following mitochondrial dysfunction, aberrant activation of glycolytic pathways also plays a key role in microglia-mediated neuroinflammation.^[37] Stimulated by hypoxic conditions, microglia undergo a classic Warburg effect-even after oxygen supply is restored, they tend to use glycolysis rather than oxidative phosphorylation as their primary mode of energy supply.^[38–40] This metabolic shift is mainly regulated by hypoxia-inducible factor1α (HIF1α). When cells are in hypoxia, HIF1α protein stability is significantly increased, and it accelerates glucose uptake and glycolysis by binding to specific DNA sequences and upregulating the expression of glucose transporters (e.g., GLUT1) and glycolysis-related enzymes. In this process, pyruvate kinase M2 (PKM2) plays a unique dual role, forming dimers through conformational changes and translocating to the nucleus to directly phosphorylate signal transducer and activator of transcription 3 (STAT3). Activated STAT3 then binds to the promoter regions of pro-inflammatory genes such as IL1β and TNFα, promoting their transcription and expression, thereby amplifying the inflammatory response. In addition, lactate, the end product of glycolysis, is also involved in the regulation of neuroinflammation.^[41] Lactate activates the Kv1.3 potassium channel in the microglia membrane, prompting the cells to release exosomes containing β-amyloid.^[42] These exosomes not only carry the neurotoxic substance amyloid-beta, but also transmit neurotoxicity to the surrounding cells through intercellular communication, further expanding the scope of inflammatory damage.^[43]

Enhanced microglial cell glycolysis induced by hypoxic environment forms a multidimensional neuroinflammatory regulatory network through HIF1α-mediated metabolic reprogramming, activation of PKM2-STAT3 signaling axis, and lactate-driven exosome release. This series of processes not only altered the energy metabolism pattern of microglia, but also continued to amplify the inflammatory response and exacerbate neurotoxicity through the synergistic effects of metabolites and signaling pathways. In-depth analysis of the mechanism of glycolysis enhancement can help to reveal the new targets of neuroinflammation and provide theoretical basis and intervention direction for the development of neuroprotective strategies targeting metabolic pathways.

4.2.2.3. Lipid metabolism disorders

In recent years, it has been found that LDRM accumulate in large numbers in the ischemic semidark zone region.^[44,45] These cells are unable to effectively metabolize phagocytosed myelin debris, leading to the accumulation of cholesteryl esters (CE) and triacylglycerol and the formation of lipid droplets.^[46] Lipid droplets not only serve as an inflammatory signaling platform to promote NLRP3 inflammatory vesicle assembly, but also hinder M2-type polarization through peroxisome proliferator-activated receptor (PPAR)/retinoid X receptor signaling pathway inhibition.^[47] A study by a team from Xiamen University demonstrated that downregulation of microglia PPARα enhances intracerebral infiltration of peripheral pro-inflammatory monocytes/macrophages (Ly6C⁺) and amplifies the inflammatory cascade response.^[6] In summary, the abnormal accumulation of LDRM in the ischemic semidarktic zone and the disordered lipid metabolism it triggers collectively contribute to the deterioration of the neuroinflammatory response through multiple mechanisms, such as promoting activation of inflammasome vesicles, inhibiting M2-type polarization, and inducing peripheral inflammatory cell infiltration.

4.2.3. Extracellular vesicle-mediated intercellular communication networks

4.2.3.1. Bidirectional regulation of microglial extracellular vesicles

Microglial cell-derived small extracellular vesicles (sEVs)-act as natural nanocarriers carrying proteins, miRNAs, and other bioactive substances that are involved in intercellular communication after stroke.^[48] Firstly, the neurological damage effect of M1 type sEVs, carrying pro-inflammatory factors such as IL-1β and TNF-α, can diffuse to the distal region, activate the transformation of astrocytes to A1 type, and release neurotoxic substances such as complement C3.^[49] In addition, miR155-containing sEVs enhanced neuronal sensitivity to excitotoxicity by inhibiting SOCS1 protein expression.^[50] Second, the neurorestorative effects of M2-type sEVs.^[51] sEVs secreted by M2-type microglia are enriched in miR25-3p and miR93-5p, and these miRNAs are effective in neurotoxicity by targeting transforming growth factor β receptor (TGF-β receptor, TGFBR), phosphatase and tensin homolog and Forkhead Box O3, which promote neural stem cell proliferation and neuronal differentiation.^[52,53]

4.2.3.2. Regulatory role of exocysts from other brain cell sources

Exocysts from astrocytes regulate autophagic flow through the miR181c5p/autophagy related gene 7 axis and inhibit cysteine asparagine 1 (caspase-1)-mediated cellular pyroptosis.^[54] And outer vesicles from endothelial cells, carrying miR132-3p, maintain blood-brain barrier integrity by targeting matrix metalloproteinase9 to protect tight junction proteins.^[55] It has also been found that outer vesicles from neural stem cells (human neural stem cell-derived extracellular vesicles), enriched with miR125a-5p, can be uptaken by microglia and promote their polarization toward the M2-type by inhibiting the toll-like receptor 4/NFκB pathway.^[56]

4.2.4. Microglia and hippocampal neural stem cell interactions and neuroregeneration

4.2.4.1. Effects of the inflammatory microenvironment on neural stem cell fate in the hippocampus

The ability of neural stem cells (NSCs) in the hippocampus to proliferate and differentiate has a direct impact on recovery of cognitive function after stroke. The inflammatory microenvironment, dominated by M1-type microglia, inhibits neuroregeneration through a variety of mechanisms, and it has been found that inflammatory factors can inhibit neurogenesis.^[57,58] For example, IL-6 and TNF-α significantly reduce the ability of NSCs to differentiate into cholinergic neurons by inhibiting the Notch signaling pathway, leading to impaired neurotransmitter synthesis. In addition, oxidative stress injury can also inhibit nerve regeneration. It was found that^[59] excess ROS not only directly damaged mitochondrial DNA of NSCs, but also induced cell cycle arrest through activation of the p38 MAPK pathway. On the contrary, M2-type microglia support NSCs survival and neuronal differentiation by secreting BDNF and GDNF.

4.2.4.2. Microglia-derived extracellular vesicle-mediated regenerative signaling pathways

M2-type microglia-derived sEVs play a central role in promoting neural regeneration. For example, a team from Xiangya Hospital of Central South University demonstrated that^[6] miR25-3p and miR93-5p carried by M2sEVs promote the differentiation of NSCs toward neurons by targeting TGFβ receptor (TGFBR) and deregulating its inhibition of Smad2/3 signaling pathway.^[53] In addition, miR25-3p promotes the proliferation of NSCs by activating the phosphatidylinositol 3-kinase protein kinase B pathway through inhibiting the expression of phosphatase and tensin homolog^[60]; it also inhibits the Forkhead Box O3 transcription factor and reduces apoptosis.

4.2.5. Targeted therapeutic strategies and translational challenges

4.2.5.1. Gene modulation strategies

First, SORLA gene therapy,^[61] a microglia-specific SORLA overexpression model developed by a team from the Huazhong University of Science and Technology (Cx3cr1CreER × SORLAcKI), significantly reduces infarct volume and improves motor function even when the intervention is initiated 24 hours poststroke, leading to a 72% reduction in neuronal apoptosis. The mechanism is to enhance phagocytosis by activating the phosphatidylinositol 3-kinase protein kinase B pathway rather than relying on the traditional M1/M2 phenotype switch. In addition, fenofibrate (an approved lipid-lowering drug),^[62] a PPARα agonist treatment^[6] developed by a team from Xiamen University, inhibits microglial NFκB activation while decreasing intracerebral infiltration of peripheral pro-inflammatory monocytes/macrophages. This strategy has the potential for rapid clinical translation due to a clear drug safety profile.

4.2.5.2. Engineered exocyst vesicle therapy

The M2 microglia-derived sEVs developed at CSU piggybacked on miR25-3p/miR93-5p to break through BBB by intranasal administration, targeting neural stem cells in the hippocampal region to regenerate nerves.^[53] The surface of engineered modified outer vesicles expresses CD47 molecules, which can evade immune clearance^[63]; sEVs carrying Kv1.3 channel inhibitors (e.g., ShK223) can selectively inhibit M1-type microglia activity and reduce pro-inflammatory factor release.^[64] In other words, M2sEVs loaded with miRNAs and target-modified exocysts are currently cutting-edge engineered exocyst therapy.

4.2.5.3. Metabolic intervention drugs

Glycolysis inhibitors, autophagy inducers, and active ingredients of traditional Chinese medicines are the main metabolic intervention drugs currently being developed for microglial cells.^[6] 2 The deoxyglucose and PKM2 tetramerization agonist, TEPP46, can reverse the Warburg effect and inhibit the solidification of pro-inflammatory phenotype. Rapamycin promotes lipid droplet degradation through activation of autophagy, reduces LDRM formation, and improves microglia function. Autophagy inducers significantly promoted microglia transformation to M2 type. Gardenia extract GJ4, curcumin (activates PPARγ), and tretinoin (inhibits NFκB) improve poststroke cognitive function through multi-target regulation of microglia polarization.

4.2.6. Outlook and future directions

In the complex pathological network of IS, microglia exhibit the dual roles of “neurological injury accomplice” and “repair ally.” The dynamic switching of their functional status is essentially the result of the combined effects of phenotypic polarization dynamics, metabolic reprogramming processes, and intercellular communication networks mediated by external vesicles. However, the translation of basic research results to the clinic still faces many challenges, and these challenges profoundly influence future research directions.

From a pathophysiological point of view, the functional state of microglia exhibits significant spatiotemporal heterogeneity, with significant differences in the state at different disease stages (acute, subacute, and recovery) and in different brain regions (infarct core and semidiabloid zone), making it difficult for a single intervention strategy to satisfy the complex clinical needs. In terms of species differences, there are significant differences in gene expression profiles, morphology and function between human and rodent microglia, making it difficult to effectively translate the results of animal model studies into clinical applications. There are also bottlenecks in the drug delivery system, as most drugs are difficult to penetrate BBB efficiently, and although engineered exocysts have potential advantages, their large-scale production and quality control standards have not yet been established.

For the future, the research in this field needs to break through from multidimensional synergistic innovation, and there are 3 strategic directions with great potential: the 1st 1 is the development of spatiotemporal precision intervention strategies. To address the spatial and temporal heterogeneity of microglia, we can realize precise spatial and temporal regulation of the activity of specific microglia subpopulations with the help of optogenetics or chemical genetics; at the same time, based on the natural delivery characteristics of exocysts, we can design engineered carrier systems with the ability of responding to the stage of the disease, delivering antiinflammatory molecules in the acute phase and intelligently releasing neurotrophic factors during the recovery phase, so as to overcome the limitations of a single strategy. The next step is to promote the construction of humanized model systems. In view of the impediment to clinical translation due to species differences, the establishment of a co-culture system of microglia and brain-like organs through induced pluripotent stem cell technology to construct an ischemic microenvironmental model that is more closely related to human pathological characteristics will not only deepen the mechanism study, but also provide a revolutionary technological platform for drug screening. The 3rd is to explore the paradigm of individualized neuroimmunotherapy. Relying on single-cell transcriptome, spatial metabolome and other multi-omics technologies, we can deeply analyze the intrinsic correlation between patients’ gene expression profiles, immunophenotypic characteristics and clinical prognosis, and establish a precise stratified treatment system. For example, by detecting the expression level of SORLA gene, we can screen subgroups of patients with high sensitivity to PPARα agonist treatment, and at the same time breakthroughs in the production and quality control of engineered ex vesicles, so as to push forward the clinical application of targeted microglia treatment strategy.

4.3. Limitations

This study has several limitations. First, the exclusion of nonEnglish and nonWoSCC publications may introduce bias. Specifically, our analysis relied solely on the WoSCC database and did not include publications from Chinese databases such as China National Knowledge Infrastructure or Wanfang. While Chinese institutions lead in publication output, this conclusion is subject to language bias. Research published in Chinese may feature unique characteristics in experimental models (e.g., specific animal strains), intervention protocols (e.g., dose conversion of traditional Chinese medicine compounds), or clinical study designs that are not fully captured in English-dominated international journals. This potential publication bias means our analysis may not fully represent the complete spectrum of research, particularly region-specific investigative trends and methodological nuances. Second, bibliometric analysis relies on keyword frequency and citation metrics, which may not fully capture mechanistic nuances. Third, while China and the U.S. dominate publication output, underrepresented regions (e.g., Australasia) may offer unique insights into IS epidemiology and treatment. Future bibliometric studies should employ cross-database, multi-language systematic searches to provide a more comprehensive assessment of the global research landscape and to explore the convergence or divergence of research paradigms across different academic contexts.

Author contributions

Conceptualization: Li Ma, Jiang-Peng Cao, Xiao-Wei Lin, Yi Guo.

Data curation: Li Ma, Jiang-Peng Cao.

Formal analysis: Li Ma, Jiang-Peng Cao, Jia-Shan Li, Yi Guo.

Funding acquisition: Yi Guo.

Investigation: Li Ma, Xiao-Wei Lin.

Methodology: Xiao-Wei Lin.

Project administration: Li Ma, Jiang-Peng Cao, Ze-Zhi Fan, Yi Guo.

Resources: Yi Guo.

Supervision: Li Ma, Jiang-Peng Cao, Jia-Shan Li, Ze-Zhi Fan, Shuai Chen, Jun-Yi Li, Xiao-Wei Lin.

Validation: Li Ma.

Visualization: Li Ma, Jia-Shan Li.

Writing – original draft: Li Ma.

Writing – review & editing: Li Ma, Jiang-Peng Cao, Jia-Shan Li, Ze-Zhi Fan, Shuai Chen, Jun-Yi Li, Xiao-Wei Lin, Yi Guo.