Past, present, and future of exosomes research in cancer: A bibliometric and visualization analysis

ABSTRACT

Cancer seriously threatens the lives and health of people worldwide, and exosomes seem to play an important role in managing cancer effectively, which has attracted extensive attention from researchers in recent years. This study aimed to scientifically visualize exosomes research in cancer (ERC) through bibliometric analysis, reviewing the past, summarizing the present, and predicting the future, with a view to providing valuable insights for scholars and policy makers. Researches search and data collection from Web of Science Core Collection and clinical trial.gov. Calculations and visualizations were performed using Microsoft Excel, VOSviewer, Bibliometrix R-package, and CiteSpace. As of December 1, 2024, and March 8, 2025, we identified 8,001 ERC-related publications and 107 ERC-related clinical trials, with an increasing trend in annual publications. Our findings supported that China, Nanjing Medical University, and International Journal of Molecular Sciences were the most productive countries, institutions, and journals, respectively. Whiteside, Theresa L. had the most publications, while Théry, C was the most co-cited scholar. In addition, Cancer Research was the most co-cited journal. Spatial and temporal distribution of clinical trials was the same as for publications. High-frequency keywords were “extracellular vesicle,” “microRNA” and “biomarker.” Additional, “surface functionalization,” “plant,” “machine learning,” “nanomaterials,” “promotes metastasis,” “engineered exosomes,” and “macrophage-derived exosomes” were promising research topics. Our study comprehensively and visually summarized the structure, hotspots, and evolutionary trends of ERC. It would inspire subsequent studies from a macroscopic perspective and provide a basis for rational allocation of resources and identification of collaborations among researchers.

Introduction

Cancer continues to be one of the most major diseases threatening the lives and health of people worldwide and is the second leading cause of death after cardiovascular disease.¹ Cancer cells have unique biological characteristics, such as resistance to cell death, unlimited replication ability, tissue infiltration, and avoidance of immune destruction;^2–4 as well as the complex pathogenesis, involving multiple environmental and genetic factors,^5,6 which makes the prevention, diagnosis and treatment of cancer difficult. Cancer carries a huge global burden across all fields and caused 250 million disability-adjusted life years (DALYs) in 2021, an increase of 16.0% compared to 2010.⁷ And the total global economic loss due to cancer management over the next 30 years could be as high as $25.2 trillion.⁸ Therefore, enhancing the efficient diagnosis and treatment of cancer has been an urgent challenge for global scientists.

In recent years, researchers have become increasingly interested in the role of a biological structure known as exosome in cancer.^9,10 Exosome is a type of extracellular vesicle (EV) with a diameter of 40 nm to 160 nm (average 100 nm), originally defined by Johnstone RM in 1987.¹¹ In the beginning, exosomes were regarded as cellular waste and were not given enough attention until the important studies of Raposo G and Zitvogel L, both of whom brought exosomes back to the forefront of people’s minds.^12,13 With further research, it has been found that exosomes contain a diversity of active substances, such as proteins, nucleic acids, and lipids, for cellular communication through intercellular delivery, which plays an important role in widely normal and pathological processes.^14–18 These properties make exosomes also play an important role in cancer development, which provides new insights for efficient cancer management, including as biomarkers, therapeutic targets, and even as drug carriers.¹⁹

Bibliometric analysis provides a visual assessment of the current status and trends in a particular research field by statistically analyzing the published researches, thus providing recommendations for future research.²⁰ Even though there are bibliometric studies that have been published for exosomes research in cancer (ERC), they are only for individual cancer types, such as prostate cancer and ovarian cancer,^21,22 and there does not yet exist an exploration of the evolution of themes and identification of hotspots for the whole field. In this study, we endeavored to summarize the evolution of research in this field, provide perspectives for future research, and offer valuable insights for financing decisions and mutually beneficial cooperation.

Materials and methods

Data source and search strategy

Web of Science Core Collection (WoSCC) is one of the most prestigious databases in the world of scientific research and is well suited for bibliometric analysis because of its comprehensive academic data and citation information.²³ We chose SCI-EXPANDED of the WoSCC as our search database for initial quality control of the documents. Our search specialists and experts in this field had developed a comprehensive search strategy around “cancer” and “exosomes” after in-depth discussions and iterative testing (Table S1), which was not limited by time or language, yet the type of publication was limited to “articles” and “reviews” indexed in WoSCC. We completed this search on December 1, 2024, to avoid bias caused by daily database updates.

Study selection and data extraction

After the retrieved records were imported into Rayyan (https://www.rayyan.ai/), we conducted a two-stage literature screening. Initial screening was performed by two researchers by independently reading the title and abstract, and if necessary, the full text. In case of disagreement, a third researcher arbitrated. It should be emphasized that exosomes belong to one kind of EVs, and there are also apoptotic bodies and microparticles according to the size and traits of different EVs.²⁴ To further control for the fit between the included documents and our study topic, we included only studies related to exosomes and cancer with the exclusion of studies that only dealt with EVs without explicitly mentioning exosomes. Subsequently, we extracted the following data for further analysis: number of publications, countries, institutions, authors, journals, and keywords, as well as the journal Impact Factor (IF) obtained from the 2023 Journal Citation Report (Clarivate Analytics, Philadelphia, USA). It should be noted that we normalized Scotland, Wales, Northern Ireland and England into the UK and Taiwan, Macau and Hong Kong into China.

Aggregate analysis of clinical trials

It is well known that clinical trials are considered to be the mirror of a research field and reflect the development of the field, and in order to explore the trends of ERC in depth, we retrieved all relevant clinical trials of ERC from ClinicalTrials.gov (https://clinicaltrials.gov) on March 8, 2025, based on the above search strategy (Table S1). The retrieved clinical trials were screened by the screening method of two-person cross-checking and a third person intervening to arbitrate when necessary. Data such as national clinical trials (NCT) number, study title, first country of conduction, start data, types of cancer, study status, and related publications were also extracted.

Data analysis and visualization

To minimize the bias of different expressions of the same scientific concept on the results of data analysis, we performed data cleaning before data analysis. In this study, we used Microsoft Excel (Office 2021) to create bar charts, pie charts and line charts and perform trend analysis. VOSviewer (version 1.6.19) is a widely used bibliometric software,²⁵ that we used for academic information summarization, collaboration and co-citation information, and keyword co-occurrence. CiteSpace (version 6.2.R4) is an application that progressively visualizes knowledge domains,²⁶ which we used to produce a dual-map overlay of journals. Bibliometrix R-package (version 4.1.4, R version 4.4.1) was used to perform the visualization of hotspot identification and thematic evolution in ERC. In our study, we explored the collaborations, journals, cited references, research hotspots and trends, and thematic evolutions of ERC, flowchart was shown in Figure 1.

Figure 1.

Flowchart of documents selection and data analysis.

Results

Time signature of publications and citations

A total of 19,066 records were obtained from the database search, and after a two-stage screening process, 8,001 documents that met the study topic were included for further analysis. These included 6,138 original articles (76.72%) and 1,863 reviews (23.28%). Figure 2a showed the temporal distribution of global publications, which combined with the fitted curve showed an overall upward trend in the number of annual publications. It could be divided into four phases: the period before 2004 was the beginning of this field, with no more than 10 publications per year. From 2005 to 2013 was the second phase of steady growth. And 2014 to 2020 saw a surge in the number of publications, achieving a breakthrough from triple to quadruple digits. 2021 to the present was the heyday of this field’s development, with more than 1,000 publications each year. Figure 2b showed the total mean citations per year, which was highest in 1998, 2001, and 2009, all exceeding 20.00, highlighting the scholarly impact of publications in this field in these three years. However, the linear fit curve suggested a negative increase in the total mean citations per year.

Figure 2.

Annual trend chart of publications and citations from 1998 to Nov. 2024. The blue dotted lines indicated the fitting predictive trends, polynomial model with R² = 0.9878 (a) and linear model with R² = 0.1023 (b).

Analysis of publication sources and collaboration

Active countries and institutions

A total of 5,657 institutions from 87 countries contributed to these publications. Table 1 listed the 20 most prolific countries. Among them, China taken the absolute lead with 4,863 publications (60.78%), followed by USA with more than 1000 publications, while the rest of the countries did not have more than 500 publications. Figure 3a on national collaborations in ERC suggested that China, USA and European countries were the three most important components of the collaborative network, and ERC was extensive across the globe.

Table 1.

Brief list of the top 20 most productive countries, institutions, authors and journals.

Items	Rank	Name	Number of Publications (%)	Rank	Name	Number of Publications (%)
Country	1	China	4,863 (60.78%)	11	Spain	134 (1.67%)
	2	USA	1,252 (15.65%)	12	France	129 (1.61%)
	3	Iran	373 (4.66%)	13	Canada	101 (1.26%)
	4	Japan	308 (3.85%)	14	Sweden	81 (1.01%)
	5	Italy	298 (3.72%)	15	Poland	74 (0.92%)
	6	South Korea	273 (3.41%)	16	Saudi Arabia	68 (0.85%)
	7	Germany	250 (3.12%)	17	Portugal	60 (0.75%)
	8	India	217 (2.71%)	18	Russia	60 (0.75%)
	9	Australia	154 (1.92%)	19	Egypt	56 (0.70%)
	10	UK	153 (1.91%)	20	Brazil	53 (0.66%)
Institution	1	Nanjing Medical University	313 (3.91%)	11	Soochow University	137 (1.71%)
	2	Shanghai Jiao Tong University	279 (3.49%)	12	Chinese Academy of Medical Sciences	129 (1.61%)
	3	Fudan University	226 (2.82%)	13	Southern Medical University	125 (1.56%)
	4	Central South University	192 (2.40%)	14	Chinese Academy of Sciences	117 (1.46%)
	5	Zhejiang University	178 (2.22%)	15	Nanjing University	115 (1.44%)
	6	Zhengzhou University	178 (2.22%)	16	Sichuan University	114 (1.42%)
	7	Shandong University	157 (1.96%)	17	Tongji University	112 (1.40%)
	8	Sun Yat-Sen University	150 (1.87%)	18	Capital Medical University	108 (1.35%)
	9	China Medical University	143 (1.79%)	19	Wuhan University	108 (1.35%)
	10	Huazhong University of Science and Technology	138 (1.72%)	20	Southeast University	99 (1.24%)
Author	1	Whiteside, Theresa L.	43 (0.54%)	11	Zhang, Yu	31 (0.39%)
	2	Wang, Wei	37 (0.46%)	12	Wang, Jing	30 (0.37%)
	3	Zhang, Wei	36 (0.45%)	13	Xu, Wenrong	30 (0.37%)
	4	Li, Juan	35 (0.44%)	14	Zitvogel, Laurence	30 (0.37%)
	5	Zhang, Xu	34 (0.42%)	15	Li, Xin	29 (0.36%)
	6	Zhu, Wei	33 (0.41%)	16	Li, Li	28 (0.35%)
	7	Liu, Yang	32 (0.40%)	17	Fais, Stefano	27 (0.34%)
	8	Yang, Yang	32 (0.40%)	18	Liu, Wei	25 (0.31%)
	9	Wang, Lei	31 (0.39%)	19	Qian, Hui	25 (0.31%)
	10	Wang, Yan	31 (0.39%)	20	Zhang, Yi	25 (0.31%)
Journal	1	International Journal of Molecular Sciences	233 (2.91%)	11	Frontiers in Cell and Developmental Biology	86 (1.07%)
	2	Frontiers in Oncology	211 (2.64%)	12	Plos One	84 (1.05%)
	3	Cancers	188 (2.35%)	13	Frontiers in Immunology	80 (1.00%)
	4	Scientific Reports	116 (1.45%)	14	Cancer Cell International	78 (0.97%)
	5	Molecular Cancer	109 (1.36%)	15	Journal of Experimental & Clinical Cancer Research	75 (0.94%)
	6	Cancer Letters	108 (1.35%)	16	Oncogene	64 (0.80%)
	7	Oncotarget*	106 (1.32%)	17	Oncology Letters	61 (0.76%)
	8	Cell Death & Disease	99 (1.24%)	18	International Journal of Nanomedicine	59 (0.74%)
	9	Analytical Chemistry	91 (1.14%)	19	Cells	56 (0.70%)
	10	Biosensors & Bioelectronics	86 (1.07%)	20	Journal of Cancer	55 (0.69%)

*Represented journal that have been removed by SCI in 2018.

Figure 3.

Co-authorship analysis visualization map (a) Collaboration world map among different countries. (b) Collaboration network map of the 61 most productive institutions. (c) Collaboration network map among the 51 most productive authors. (d) Core journals graph based on Bardford’s law.

After analyzing the top 20 highly productive institutions, Nanjing Medical University (313, 3.91%), Shanghai Jiao Tong University (279, 3.49%) and Fudan University (226, 2.82%) ranked the top three with more than 200 publications, respectively. There were 19 institutions having more than 100 publications. In addition, Southeast University has the highest citations per document (46.77) among these 20 institutions (Table S2). And most surprisingly, all 20 organizations are from China. The cluster network in Figure 3b showed the collaboration between the 61 institutions with more than 40 publications, forming 15 clusters. The clusters with red, green, and blue nodes had the strongest collaboration. Among them, Nanjing Medical University had the highest total link strength (302).

Highly productive scholars

With 39,185 authors identified in 8,001 publications, the top 20 most prolific authors in ERC were shown in Table 1, with a total of 624 publications (7.80%). Whiteside, Theresa L. was the most prolific author (43, 0.54%), followed by Wang, Wei (37, 0.46%) and Zhang, Wei (36, 0.45%). In addition, France’s Zitvogel, Laurence had the highest citations per document (308.30) (Table S3). And of those 20 authors, all of them are from China, except for one each from USA, France and Italy. The collaborative network map for the 51 authors with 20 publications and more was shown in Figure 3c, with the red nodes forming the largest cluster. However, it was worth noting that there were as many as four independent nodes.

Overview of core journals

A total of 8,001 documents were published in 1,106 journals, with the top 20 most prolific journals listed in Table 1, contributing 2,045 publications (25.56%). International Journal of Molecular Sciences published the most publications (233, 2.91%), followed by Frontiers in Oncology (211, 2.64%) and Cancers (188, 2.35%), while Molecular Cancer had the highest citations per document (126.49) (Table S4). The average IF of these 20 journals was 7.0, with Molecular Cancer having the highest impact factor (IF = 27.7). These journals cover a wide range of research fields, with a focus on Oncology, Cell Biology and Biochemistry. Switzerland (n = 6), UK (n = 6) and USA (n = 3) were the three countries that operate the most of these journals. In addition, Bradford’s Law states that the source of majority of professional literature is a limited number of specialized core journals,²⁷ and we also conducted a statistical analysis of the core journals in this field, which included 33 journals (Figure 3d).

Co-citation analysis

Influential authors

Co-cited authors are the authors of publications that are cited simultaneously in multiple documents in the analysis, and highly co-cited authors are scholars who have had a profound impact in the research field. There was a total of 109,920 co-cited authors, of which the top 20 highly co-cited authors were listed in Table 2. The most highly co-cited authors are Théry, C from France (citations = 3,152), Kalluri, R from USA (citations = 1,964), and Valadi, H from Italy (citations = 1,374) (Table S5 and Figure 4a).

Table 2.

Brief list of the top 20 most influential authors and journals in publications.

Items	Rank	Name	Citations	Rank	Name	Citations
Author	1	Théry, C	3,152	11	Hoshino, A	964
	2	Kalluri, R	1,964	12	Li, Y	918
	3	Valadi, H	1,374	13	Van Niel, G	899
	4	Zhang, Y	1,369	14	Siegel, RL	878
	5	Raposo, G	1,211	15	Colombo, M	847
	6	Whiteside, TL	1,110	16	Wang, J	836
	7	Taylor, DD	1,088	17	Jemal, A	791
	8	Melo, SA	1,087	18	Li, J	775
	9	Peinado, H	1,060	19	Clayton, A	771
	10	Zhang, L	1,035	20	Liu, Y	766
Journal	1	Cancer Research	10,435	11	International Journal of Molecular Sciences	6,368
	2	Molecular Cancer	9,695	12	Cancer Letters	6,019
	3	Plos One	9,366	13	Journal of Biological Chemistry	5,895
	4	Oncotarget*	8,694	14	Nature Cell Biology	5,875
	5	Nature	7,810	15	Oncogene	5,832
	6	Scientific Reports	6,856	16	Clinical Cancer Research	4,812
	7	Nature Communications	6,803	17	Journal of Immunology	4,586
	8	Cell	6,759	18	Cancers	4,222
	9	Proceedings of The National Academy of Sciences of The United States of America	6,514	19	International Journal of Cancer	4,145
	10	Journal of Extracellular Vesicles	6,475	20	Science	4,099

*Represented journal that have been removed by SCI in 2018.

Figure 4.

Co-citation analysis visualization map (a) Heatmap of highly cited author distribution. (b) Bar chart of top 20 highly cited journals. (c) Dual-map overlay of journals. (d) Reference publication year spectroscopy.

Co-cited journals

A total of 10,912 journals were cited two or more times, and Table 2 listed the 20 most highly co-cited journals, covering a wide range of research fields. Among these, Cancer Research had the highest total number of citations (citations = 10,435), followed by Molecular Cancer (citations = 9,695) and Plos One (citations = 9,366) (Figure 4b). The average IF of these journals was a whopping 15.0, with half of them originating in USA, and UK also operates 35% of these journals (Table S6). Figure 4c showed dual-map overlay of journals, representing the paths of citing and cited journal topics. The colored curves represent the thematic reference paths of the journals on the right citing the journals on the left. The map showed the three main reference paths. The orange path referred to documents published in molecular/biology/immunology mostly were cited in journals in molecular/biology/genetics. The green route referred to documents published in medicine/medical/clinical mostly were cited in journals in molecular/biology/genetics, and health/nursing/medicine.

Co-cited spectroscopy and co-cited documents

Table 3 listed the top ten most co-cited references and co-cited documents, respectively, to reveal important publications in our literature dataset. The most co-cited reference and co-cited document in total were a review²⁸ entitled “Exosome-mediated transfer of mRNAs and microRNAs is a novel mechanism of genetic exchange between cells” and an original article²⁹ entitled “Tumour exosome integrins determine organotropic metastasis,” respectively. Notably, papers published as first authors by Hoshino, Ayuko,²⁹ Peinado, Héctor³⁰ and Melo, Sonia A³¹ in Nature and Nature medicine, respectively, were included in both lists (Table S7 and S8). Reference publication year spectral (RPYS) analysis is a quantitative method for exploring the historical roots of research fields based on an analysis of the frequency of references cited in publications, and quantifying their impact on current research.³² As shown in Figure 4d, the frequency of references showed a general fluctuating trend and has peaks in 2002 and 2018.

Table 3.

Brief list of the top 10 highly cited references and documents.

Rank	Reference				Document
	Paper (first author, year)	Publication Type	DOI	Total Citations	Paper (first author, year)	Publication Type	DOI	Total Citations
1	Valadi, Hadi 2007	Review	10 .1038/ncb1596	1,369	Hoshino, Ayuko 2015	Article	10 .1038/nature15756	838
2	Kalluri, Raghu 2020	Review	10 .1126/science.aau6977	1,117	Peinado, Héctor 2012	Article	10 .1038/nm.2753	789
3	Théry, Clotilde 2002	Review	10 .1038/nri855	905	Melo, Sonia A 2015	Review	10 .1038/nature14581	612
4	Hoshino, Ayuko 2015	Article	10 .1038/nature15756	838	Kalluri, Raghu 2016	Review	10 .1172/JCI81135	545
5	Peinado, Héctor 2012	Article	10 .1038/nm.2753	789	Costa-Silva 2015	Article	10 .1038/ncb3169	537
6	Skog, Johan 2008	Article	10 .1038/ncb1800	749	Chen, Gang 2018	Article	10 .1038/s41586-018-0392-8	463
7	Raposo, Graça 2013	Review	10 .1083/jcb.201211138	706	Melo, Sonia A 2014	Article	10 .1016/j.ccell.2014.09.005	456
8	Théry, Clotilde 2006	Review	10 .1002/0471143030.cb0322s30	690	Tkach, Mercedes 2016	Review	10 .1016/j.cell.2016.01.043	425
9	Jemal, Ahmedin 2011	Article	10 .3322/caac.20107	614	Kamerkar, Sushrut 2017	Article	10 .1038/nature22341	414
10	Melo, Sonia A 2015	Review	10 .1038/nature14581	612	Mashouri, Ladan 2019	Review	10 .1186/s12943-019-0991-5	408

Analysis of co-occurrence keywords

Keywords highly summarize the topic of the documents and can be used as a powerful tool to explore research hotspots and trends in related fields. A total of 15,456 keywords were extracted from the literature dataset, and after excluding nonspecific keywords, we summarized the top 20 keywords with the highest frequency of ERC (Table 4). Among them, “extracellular vesicle” (2,028 times), “microrna” (1,753 times), “biomarker” (1,600 times) and “metastasis” (1,173 times) all exceeded 1,000 times. And “breast cancer,” “lung cancer,” and “colorectal cancer” were the three most frequently occurring cancers. And “metastasis,” “proliferation,” “growth-factor,” “progression,” “invasion” represent research topics more than 600 times. Figure 5a showed the percentage of high-frequency keywords more visually.

Table 4.

Top 20 most frequently occurring keywords.

Rank	Keyword	Occurrences	Rank	Keyword	Occurrences
1	extracellular vesicle	2,028	11	invasion	603
2	microrna	1,753	12	biogenesis	531
3	biomarker	1,600	13	diagnosis	517
4	metastasis	1,173	14	mechanism	501
5	proliferation	805	15	Hepatocellular carcinoma	469
6	breast cancer	796	16	angiogenesis	467
7	growth factor	727	17	tumor microenvironment	458
8	progression	676	18	therapeutic target	453
9	colorectal cancer	655	19	migration	449
10	lung cancer	623	20	mesenchymal stem cell	428

Figure 5.

Keywords and thematic analysis visualization map (a) High-frequency keyword cloud. (b) Co-occurrence network map of top 169 high frequency keywords. (c) Trend topics map of keywords plus. (d) Thematic evolution map of keyword plus.

We set the minimum frequency of occurrence to 60 times, and a total of 169 keywords were detected. Figure 5b illustrated the cluster network consisting of the above keywords, where all closely related keywords were automatically classified into four clusters, indicated by different colors. The red cluster was the largest, mainly describing the involvement of exosomes in cancer development, containing 56 nodes, and consisted of the keywords, “expression,” “metastasis,” “proliferation,” “invasion,” “breast cancer,” and “lung cancer,” et al. The blue cluster contained 45 nodes, which was related to “microrna,” “biomarker,” “diagnosis,” “growth-factor,” “serum,” “liquid biopsy” and so on, mainly covering the diagnostic application of exosomes in cancer. High-frequency keywords in the green cluster included “extracellular vesicle,” “tumor microenvironment,” “dendritic cells,” “macrophage,” “natural-killer-cells” and “bone-marrow,” mainly mentioning exosome-related cells in cancer. The yellow cluster was the smallest, with only 20 keywords, mainly associated with the therapeutic use of exosomes in cancer, including “drug delivery,” “delivery vehicle,” “nanoparticles,” “therapeutic target,” “mesenchymal stem cells,” and so on.

Temporal distribution map of research topics

In order to further analyze the pattern of research topic turnover evolution and explore current research hotspots as well as identify promising research directions, we conducted a topic trend analysis of ERC based on keyword plus. We set the minimum frequency of keyword occurrence to 10 times and the maximum number of representative topics to 4 per year and obtained a total of 57 annual hot topic terms. As shown in Figure 5c, which illustrated the trend of research topics, the top terms that has continued to 2024 included “promotes metastasis,” “engineered exosomes,” “macrophage-derived exosomes,” “machine learning,” “nanomaterials,” “plant,” and “surface functionalization,” indicating current research hotspots with potentially promising for future research. However, the keywords at the bottom represent research topics, such as “class-ii molecules,” “cytotoxic t-lymphocytes,” and “adjuvants,” where research has matured and gradually faded from researchers’ attention.

Dynamic exploration of research trends over time and the evolutionary associations of the corresponding thematic units can better reflect research priorities and trends and is a way of clarifying the veins of the research field, as well as further complementing the identification of research hotspots and future research directions. As shown in Figure 5d, of the five periods we have delineated, the first stage was limited and had few research topics; In the second and third stages, the horizontal expansion of the field was more, entering the period of exploration; In the latter two periods, it was normalized back to the main research hotspots. “Dendritic cells” lasted for three periods from the second period, taking up half of the topics of the second period in the third period (dendritic cells, receptor, in-vivo, maturation, heat-shock-protein, and class-i) and radiating backward to the three topics of the fifth period (extracellular vesicle, microrna, and dendritic cells), and its evolution direction represented the mainstream evolutionary direction. “Breast-cancer,” “microvesicles” and “protein” were also hot topics in the second period, with relatively mainstream evolutionary connections both forward and backward. The hot topics of the third period were undoubtedly “extracellular vesicle,” “biomarker” and “microrna,” which carried over all the topics of the second period, with “biomarker” radiating to four of the five topics of the recent period (extracellular vesicle, metastasis, microrna and quantification), and microrna radiating backwards into three topics (extracellular vesicle, metastasis, and microrna), and mainstream “Dendritic cells” radiating to “extracellular vesicle.” Finally, the latest period has focused on “metastasis,” “microrna,” “quantification” and “dna.”

Summary of the clinical trials

The search returned a total of 196 clinical trial entries, and after two-stage screening process, 106 clinical trials were included for further analysis (Table S9). Figure 6a showed the different study themes in ERC, with the vast majority of studies (85.05%) focusing on biomarkers, followed by mechanism analysis (7.48%) and exosomes therapy (4.67%). China, USA and France occupied the top three countries leading the launch of ERC studies, with a total of 87.85% of the proportion of clinical trials conducted (Figure 6b). Clinical trials for ERC increased gradually from 2010 to 2016, and were at a relative plateau from 2017 to the present, with the highest number of studies conducted in 2020 (Figure 6c). In the conducting clinical trials, the most studied cancer was lung cancer (22, 20.56%), followed by prostate cancer (16, 14.95%) and pancreatic cancer (9.35%) (Figure 6d). Table 5 presented a partial list of clinical trials and findings from publications, with the largest number of prostate cancer (4, 36.36%), which primarily illustrated the superior performance of the urine-derived ExoDx prostate intelliscore (EPI) compared to traditional multivariate risk calculators in a homogenized group of men prior to prostate cancer biopsy. This was followed by thyroid and pancreatic cancers, emphasizing the performance of some exosomal contents as biomarkers, with additional mention of the drug delivery vector capabilities of plant and mammalian exosomes that can be targeted to pancreatic cancer.

Figure 6.

Summary of clinical trials in ERC (a) Pie chart of clinical trial themes. (b) Pie chart of countries leading clinical trials. (c) Annual trend chart of clinical trials conducted. (d) Bar graph of cancer types researched in clinical trials.

Table 5.

Some registered clinical trials with findings in ERC from ClinicalTrials.Gov.

NCT Number	Start Date	Locations (First country)	Types of cancer	Types of research	Source of exosomes	Publication (Title)	Findings
NCT02702856	2014–05	USA	Prostate Cancer	Mechanism analysis	Urine	Kretschmer A, et, al. Pre-diagnosis urine exosomal RNA (ExoDx EPI score) is associated with post-prostatectomy pathology outcome. World J Urol. 2022 Apr;40(4):983–989.	The EPI Score urine biomarker outperformed the multivariate risk calculators in a homogenous risk group of pre-biopsy men. The EPI score was associated with low-risk pathology post-radical prostatectomy.
NCT03031418	2016–09	USA	Prostate Cancer	Biomarker	Urine
NCT03235687	2017–07	USA	Prostate Cancer	Biomarker	Urine
NCT04720599	2020–06	USA	Prostate Cancer	Biomarker	Urine
NCT02862470	2016–08	China	Thyroid Cancer	Biomarker	Urine	Huang TY, et, al. Urinary Exosomal Thyroglobulin in Thyroid Cancer Patients With Post-ablative Therapy: A New Biomarker in Thyroid Cancer. Front Endocrinol (Lausanne). 2020 Jun 16;11:382.	In comparison with serum thyroglobulin, U-Ex Tg can be an important pro-inflammatory predictor and biomarker of thyroid cancer recurrence
NCT03488134	2018–08	China	Thyroid Cancer	Biomarker	Urine
NCT03235687	2017–07	USA	Prostate Cancer	Biomarker	Urine	Tutrone R, et, al. ExoDx prostate test as a predictor of outcomes of high-grade prostate cancer – an interim analysis. Prostate Cancer Prostatic Dis. 2023 Sep;26(3):596–601.	Men who received an EPI low-risk score (<15.6) significantly delayed their first biopsy and remained at very low risk for pathology 2.5 years after the initial study. EPI test risk stratification identified low-risk patients with no detectable standard of care.
NCT04720599	2020–06	USA	Prostate Cancer	Biomarker	Urine	Kretschmer A, et, al. Validation of a CE-IVD, urine exosomal RNA expression assay for risk assessment of prostate cancer prior to biopsy. Sci Rep. 2022 Mar 21;12(1):4777.	EPI-CE provides information beyond standard clinical parameters and provides a better risk assessment prior to MRI, of patients suspected of prostate cancer, than the commonly used multiparametric risk calculators.
NCT03031418	2016–09	USA	Prostate Cancer	Biomarker	Urine	McKiernan J, et, al. A Prospective Adaptive Utility Trial to Validate Performance of a Novel Urine Exosome Gene Expression Assay to Predict High-grade Prostate Cancer in Patients with Prostate-specific Antigen 2-10 ng/ml at Initial Biopsy. Eur Urol. 2018 Dec;74(6):731–738.	EPI is a noninvasive, easy-to-use urine test for gene expression with accurate ≥GG2 risk stratification for GG1 cancers and benign diseases. It improves the identification of patients with higher grade disease and will reduce the total number of unnecessary biopsies.
NCT03488134	2018–08	China	Thyroid Cancer	Biomarker	Urine	Wang CY, et, al. Long-Term Changes of Urinary Exosomal Peptide Levels After Thyroidectomy in Patients with Thyroid Cancer: A Prospective Observational Study. Int J Nanomedicine. 2024 May 23;19: 4667–4677.	In high-risk patients after thyroidectomy, serum protein or urinary exosomal peptide concentrations within a defined range may indicate a lower risk of thyroid cancer recurrence during long-term follow-up.
NCT03102268	2017–05	China	Cholangiocarcinoma	Biomarker	Plasma	Ge X, et, al. The diagnostic value of exosomal miRNAs in human bile of malignant biliary obstructions. Dig Liver Dis. 2021 Jun;53(6):760–765.	The expression of exosomal mir-483-5p and mir-126-3p in the bile samples discriminates between patients with malignant and nonmalignant biliary obstructions.
NCT03608631	2021–01	USA	Pancreatic Cancer	Exosomes therapy	Maximum tolerated dose	Sall IM, et, al. Plant and mammalian-derived extracellular vesicles: a new therapeutic approach for the future. Front Bioeng Biotechnol. 2023 Sep 13;11: 1215650.	Exosomes released by plant and mammalian cells are involved in the physiological and pathological mechanisms of cancer, and they have also been investigated as potential biomarkers for diagnosis and carriers for drug delivery.
NCT02662621	2015–12	France	Malignant Solid Tumor	Biomarker	Blood and urine	Chanteloup G, et, al. Membrane-bound exosomal HSP70 as a biomarker for detection and monitoring of malignant solid tumors: a pilot study. Pilot Feasibility Stud. 2020 Mar 3;6:35.	HSP70-exosomes may be a powerful tool to diagnose cancer and to guide clinicians in therapeutic decision-making, improving patient’s care.
NCT03032913	2017–02	France	Pancreatic Cancer	Biomarker	Blood	Buscail E, et, al. High Clinical Value of Liquid Biopsy to Detect Circulating Tumor Cells and Tumor Exosomes in Pancreatic Ductal Adenocarcinoma Patients Eligible for Up-Front Surgery. Cancers (Basel). 2019 Oct 26;11(11):1656.	Combined CTC and exosome detection to diagnose resectable pancreatic cancers could provide a rapid, reliable, noninvasive decision-making tool in early, potentially curable pancreatic cancer.
NCT03108677	2017–05	China	Osteosarcoma	Biomarker	Blood	Bao Q, et, al. Extracellular Vesicle RNA Sequencing Reveals Dramatic Transcriptomic Alterations Between Metastatic and Primary Osteosarcoma in a Liquid Biopsy Approach. Ann Surg Oncol. 2018 Sep;25(9):2642–2651.	Liquid biopsy-based approaches to track metastatic osteosarcoma transcriptomic alterations are a promising source of prognostic and therapeutic biomarkers.

EPI: ExoDx prostate intelliScore; U-Ex Tg: urinary exosomal thyroglobulin; MRI: Magnetic resonance imaging; CTCL: Circulating tumor cells; GG: Grade group.

Discussion

In the current knowledge explosion, numerous research results are published every day, and in face to the countless amount of literature, bibliometrics is a convenient way that can help researchers to quickly understand the past, present, and future of a new field through scientific statistics and visualization. To the best of our knowledge, this paper is the first study to use bibliometrics to statistically summarize global ERC from a visualization perspective. Based on global publications and clinical trials, we summarized the development of this field and provided valuable insights into hot spots and trends.

Summary of bibliometric results

Analysis of the annual number and the fitted curves of publications for the period 1998 to November 2024 for 8,001 publications revealed a general trend of steady growth in the number of publications in ERC. The first relevant article we detected in this field was a 1998 study by Amigorena, S entitled “Dendritic cell-derived exosomes elicit potent anti-tumor immune responses in vivo,” which described how exosomes produced by tumor peptide-loaded dendritic cells triggered specific cytotoxic T lymphocytes in vivo and eradicated established tumors,³³ kicking off an entire field of research. Before 2013 was the stage of knowledge accumulation in ERC, which experienced a lot of explorations and attempts, as shown in Figures 2 and 5d. The explosive growth in the number of publications in the years following 2013 may be attributed to the fact that the 2013 Nobel Prize in Physiology or Medicine was awarded to scientists working on the mechanisms regulating intercellular vesicle trafficking, explaining the great potential of exosomes research and thus contributing to the vigorous development of the field. Especially in the recent years, the number of annual publications has exceeded 1,000, and it could be expected that this field will remain a research hotspot in the coming period. The total mean citations per year showed a downward trend, which may be due to the small number of studies in the initial stage, while the content played a cornerstone role on the one hand, and to the fact that the studies have become more specialized and in-depth in recent years on the other hand.

A total of 5,657 organizations from 87 countries have participated in the publications, which indicates that ERC is an area of high global interest. Unfortunately, the highly productive countries were still dominated by developed countries, and less developed countries accounted for a small percentage, which may be related to the high cost of research, research heritage and scientists’ international vision in this field. However, it is noteworthy that China, as a developing country, has published more than half of the documents (Table S1) and the top 20 highly productive institutions (Table S2) are all from China, and China has led the most clinical trials (Figure 5b), which may stem from the high importance that the Chinese scientific community attaches to ERC. This is because, as far as we know, as China’s high-level scientific research fund, the National Natural Science Foundation of China (NSFC), has funded a large amount of exosomes research in recent years, which has directly contributed to making China the most productive and influential country in the world in ERC, showing a great potential for innovation in the next few years. China, USA and European countries became the most prominent nodes of the national collaboration network, and in the real world these three parties might be called the preferred countries to collaborate with when conducting research (Figure 3a). The institutional collaboration network centered on Nanjing Medical University, Shanghai Jiaotong University, Shandong University suggested that the highly productive 20 institutional collaborations were biased toward domestic and not frequent enough with foreign institutions, such as the University of Pittsburgh, Tabriz University of Medical Sciences (Figure 3b). If national exchanges can be further promoted to be normalized in the real world, there will probably be more transversal topics generated to further promote research to be more extensive and in-depth.

Professor Whiteside, Theresa L., from University of Pittsburgh, USA, had the highest publications in ERC (n = 43). As a distinguished tumor immunologist, she has been at the forefront of tumor-derived exosomes (TEX) and their role in cancer-induced immune mechanisms, describing the role of TEX in tumor immunity, and their application in cancer surveillance.^34–36 In addition, she has also worked on the role of TEX as medicinal carriers and explored the development of anticancer vaccines.³⁷ She has made great contributions to basic and applied research in this field. French scholar Théry, C was the most highly co-cited author by a wide margin (citations = 3,152), demonstrating her profound influence. As a pioneer in exosomes research, Théry, C established the gold standard for extraction, separation and identification of exosomes, described the physicochemical properties of exosomes, and facilitated the whole field of exosomes research.^38,39 And Théry, C discussed the role of exosomes in physiology and pathology, especially in tumor cell communication and metastasis, further refining immune mechanisms.^40,41 Also, Kalluri, R and Valadi, H are internationally influential scholars in ERC and researchers should focus on the research direction and dynamics of such scholars to seek inspiration. In addition, although it is true that Chinese scholars are leading in the ERC in terms of the number or frequency of publications. However, it cannot be ignored that as for some highly influential documents, most of them are published by scholars from USA and Europe (Table S3). As we all know, high-quality researches are crucial to the development of a discipline, so the academic community should focus on improving the quality of publications while emphasizing the quantity of research outputs to ensure that the research work can have a more far-reaching impact. Moreover, there were more isolated nodes in the author collaboration network map (Figure 3c), which reflected the limitation of author cooperation. As to achieve high-quality research results, Chinese scholars should pay more attention to improving international vision and increasing international cooperation.

Suitable journals for publications in a particular field are also sought by researchers, both to guide them in submitting their manuscripts accurately and as a source of access to cutting-edge developments. After comparative analysis of the number of publications and citations, we believed that researchers could focus on Molecular Cancer, Cancer Letters and International Journal of Molecular Sciences on daily basis. Highly co-cited references usually describe fundamental or breakthrough advances in a field that are important in the field. The references we listed in Table 3 were the classic literature in ERC. For example, Valadi, H et al. in their study published in Nature Cell Biology in 2008 found that messenger RNA (mRNA) and microRNA (miRNA) in exosomes, which can be translated across cells, have important roles in cellular communication.²⁸ Théry, C et al. in their 2002 publication systematically described the composition, biogenesis, and function of exosomes, which is a cornerstone of the field of exosome research.⁴² In addition, a study by Hoshino, A et al. published in Nature in 2015 revealed that tumor cell exosomal integrins mediate organ metastasis, such as exosomal integrins α6β4 and α6β1 are associated with lung metastasis, while exosomal integrin αvβ5 is associated with liver metastasis, which enriches the theory of metastasis of tumor cells.²⁹ Researchers could construct a basic framework for research in ERC by reviewing these documents as described above. It is also important to note that the top two most cited local documents are from Dr. David Lyden’s laboratory at Weill Cornell Medicine (Dr. Hoshino, A et al. 2015 in Nature; Dr. Peinado, H et al. 2012 in Nature Medicine), both of which are landmark studies within the field of ERC, highlighting the outstanding contributions of Dr. David Lyden’s laboratory’s outstanding contributions to ERC and should merit widespread attention from the academic community.

Extensive overview and further exploration of ERC

Keywords cluster analysis and thematic evolution studies gave us a broad overview of current statuses and trends of ERC. Exploring exosome biogenesis is the first step to understand the biological behavior of exosomes. Studies have shown that exosomes originate from the endocytosis pathway, and the typical formation process includes four stages: “early secretory endosome – multivesicular bodies – late endosomes – vesicle release.”^18,43 In the above process, an important role is played by the association of endosomal sorting complexes required for transports (ESCRTs) with a variety of related proteins.^44–46 There are also ESCRTs independent mechanisms that rely on heat shock proteins (HSP), such as HSP70 and HSP90, and cluster of differentiation (CD), such as CD63 and CD81.^47–49 Since exosomes are widely present in various body fluids,¹⁴ efficient isolation of them is paramount for further research. Various exosome isolation methods have been reported, such as ultracentrifugation, ultrafiltration, precipitation, chromatography, immunoaffinity capture and microfluidics, each with its own advantages and disadvantages.^50–52 Quantification of exosomes has been a forefront topic in recent years (Figure 5d), and although several different methods have been developed, including protein concentration, tool-specific measurements, cell equivalents, and so on, there are large limitations, so a more standardized quantification method is still pursued by researchers.^10,53

Exosomes contain a variety of contents such as proteins, lipids, miRNA, long stranded noncoding RNA (lncRNA), circular RNA (cirRNA), which participate in the process of tumor development by transmitting signals in paracrine and autocrine ways.⁵⁴ These biomolecules of TEX undergo significant changes in spatiotemporal polymorphism in their expression profiles compared to those in normal tissue exosomes. And they are interdependent and closely related, structurally co-located, and functionally forming a reticular signaling, finely tuned to tumor invasion, metastasis, and are also altering the internal environment of the patient’s organism, resulting in increasingly malignant biological behavior of the tumor. They play a major role in tumor development. Therefore, these contents can be used as biomarkers for the diagnosis, prognosis or grading of cancer, especially those can be isolated and identified in body fluids, in line with the current concept of “non-invasive diagnosis.” In addition, more than 85% of the currently registered clinical trials have focused on the value of exosomes as biomarkers for clinical applications (Figure 6a), making biomarker research on exosomes a hot topic in recent years. Among them, proteins mainly play the functions of transporting and sorting of exosome contents, characterizing exosomes as well as participating in cytoskeleton construction, such as membrane transport proteins, HSP, CD63, CD81, and integrins.⁵⁵ We observed that miRNAs have been hot research topics (Figure 5), which are widely involved in cancer development as important signaling molecules and can serve as good biomarkers. For example, miR-23a and miR-126 can be used as diagnostic biomarkers for non-small cell lung cancer;^56,57 miR-26 and miR-122 can be used as diagnostics for cholangiocarcinoma;⁵⁸ and miR-1290 and miR-375 can be used in combination to analyze the prognosis of prostate cancer patients.⁵⁹ In addition, cirRNAs have also entered the researchers’ field (Figures 5c). Notably, they function mainly by regulating miRNAs, such as circCCDC66, circ-ABCC1, and circ-IARS^60–62 promotes metastasis by regulating miRNAs expression. Exosome contents are morphologically independent and functionally interconnected, and the study of exosome contents contributes to the understanding of the molecular mechanisms of cancer progression and provides targets for cancer diagnosis and treatment. Additionally, exosomes are praiseworthy for the diagnosis and monitoring of refractory cancers, such as the diagnostic value of p53, p65 in blood exosomes in glioblastoma multiforme (GBM), and the prognostic monitoring ability of the exosome-nanoparticle carrier system.^63,64

Tumor microenvironment (TEM), which consists of extracellular matrix, stromal cells and immunity cells, is a guarantee of the biological behavior of cancer cells and has been popular among researchers in recent years.^65,66 Exosomes are an important component of the TME and play an important role as signaling molecules in cellular interactions. Fibroblasts are a major component of TEM, and their derived exosomes release, e.g. miR-21, miR-143, lncRNA, which are one of the key factors in carcinogenesis, as well as promoting cancer cell proliferation and metastasis by inhibiting oxidative phosphorylation.^67–69 Particularly, the biological behavior of the interaction between TEX and mesenchymal stem cells (MSCs). TEX can regulate cell metabolism and influence epigenetic modifications through the delivery of miRNA, vascular endothelial growth factor (VEGF) and other signaling molecules to promote the transformation of MSC into cancer-associated fibroblasts (CAF) and participate in the progression of cancer cells, for example, exosomes secreted by breast cancer and gastric cancer cells mediate the transfer of transforming growth factor-β (TGF-β) to promote the transformation of MSC into CAF.^70,71 Similarly, MSC-derived exosomes can regulate cancer cell proliferation, migration, invasion and drug resistance by carrying growth factors, proteins and miRNAs. For example, MSC-derived exosomes enhance migration and invasion of gastric cancer cells through epithelial-mesenchymal transition, and MSC-derived exosomes trigger the calcium/calmodulin-dependent protein kinase (CaM-Ks) pathway in gastric cancer cells to induce drug resistance.^72,73 Additionally, the interaction between stem cells and exosomes has been the focus of scholars.^74–76 The twenty-first century has been called the “age of immunity,” and the study of immune mechanisms has been a frontier topic in TEM, in which macrophages have become the “new favorite” of researchers in recent years.⁷⁷ Studies have shown that macrophages are core immune cells that regulate TEM inflammation, and their polarization to the M2 phenotype promotes cancer cell proliferation and metastasis, such as exosomes secreted by M2 macrophages are highly expressed in colorectal cancers miR-21-5p, miR-210-3p, and secreted miR-23a-3p metastasize to hepatocellular carcinoma cells.^78–80 This finding further refines the immunological mechanisms of tumor development and provides inspiration for targeted therapies. Moreover, studies have attempted to elucidate the association of cancer activation pathways, macrophage polarization and exosomes in gliomas.⁸¹ The interwoven direction of macrophage polarization, exosomes, and cancer is a highly promising one for the future (Figure 5c), and researchers should pay more attention to it. Cancer cell metastasis, which leads to a poor prognosis of cancer, is an area of focus for clinicians and researchers, and we can also observe this as a leading topic in Figure 5. More widely studied is the effect of exosomes on epithelial-to-mesenchymal transition (EMT). Exosome contents can act as promoters to influence EMT and thus cancer cell metastasis. For example, studies have reported that integrin alpha 2 subunit (ITAG2), miR-106b-3p, miRNA-6780b-5p, and circ-PDE8A promote metastasis of prostate, colorectal, ovarian, and pancreatic cancer cells, respectively.^82–85 Interestingly, exosome contents can also act as suppressors, for example, studies have shown that miRNA-let7e, miRNA-204 can reduce the metastasis of lung cancer cells, and miRNA-381-3p can inhibit the metastasis of breast cancer cells.^86–88 There are multiple pathways that play a role in this process, yet research remains limited and is a difficult field of ERC.¹⁸ Of course, exosomes also have important roles in tumor growth, tumor vascularity, and immunosuppression.

As for cancer treatment, many drugs have been developed for chemotherapy, or selective radiotherapy, but clinical evidence has shown that both of these modalities may result in damage to normal tissues, or drug-resistance of the cancer cells, resulting in a poor prognosis for patients with cancer. Therefore, delivering drugs accurately to cancer cells to improve the effectiveness of cancer treatment has always been pursued. The use of exosomes as a drug delivery vehicle seems to overcome the above limitations. On the one hand, exosomes can improve the precise delivery of drugs while reducing the destruction of normal tissues based on their natural biological properties such as specific targeting, biocompatibility, and ability to cross cell membranes. On the other hand, there have been evidence that exosomes can effectively reduce chemotherapy or radiotherapy resistance in cancer cells.^89–93 Several experiments have demonstrated the advantages of exosome delivery of chemotherapeutic drugs, for example, exosomes loaded with paclitaxel significantly increased drug uptake and cytotoxicity in cancer cells compared to free drug.⁹⁴ Exosomes can also be targeted for drug delivery through inhibitors of some cancer development pathways.^95,96 Exosomes secreted by dendritic cells have been studied earlier and are relatively well-established drug delivery vehicles; and because of the antigen-presenting properties of their cargoes, they have recently been investigated as antitumor vaccines for personalized immunotherapy of cancer.^97,98 Macrophage and MSC-derived exosomes are more popular drug carriers in recent years.^99–101 It is important to mention that recently researchers have discovered the potential shown by plant-derived exosomes (PDE) in drug delivery, such as ginger-derived exosomes delivering doxorubicin to treat colon cancer and grape-derived exosomes delivering paclitaxel to treat breast cancer, which is considered green and harmless and has better precision, which is certainly an exciting direction for research.^102,103 However, this field is still in its infancy and needs to be further explored.

In addition, engineered exosomes have modified to confer new properties to the exosomes to make drug delivery safer, more accurate, more efficient, and more in line with the requirements for intentional drug carriers,^18,104,105 such as the accurate delivery of the chemotherapeutic drug doxorubicin for triple-negative breast cancer.¹⁰⁶ One of the important fields of research is surface functionalization of exosome, in which specific exosome membrane proteins, such as CD63, CD81, CD9, lactose adhesin are surface modified by genetic engineering method, covalent modification, non-covalent modification, fluorescent labeling, etc., to achieve site-specific drug delivery, in vivo imaging and tracking, and to further enhance the precision of drug delivery and dynamic monitoring of biodistribution.¹⁰⁷ Engineered exosomes exist for many types of anti-tumor therapies, such as photodynamic therapy, radiation therapy, immunotherapy and gene therapy, which are ideal tools for clinical treatment.¹⁰⁸ However, there are many challenges in the translation of engineered exosomes, such as the unknown optimal source and modification strategy of their antitumor activity, and how to design appropriate exosomes in different antitumor therapies.¹⁰⁹ Even though engineered exosomes still have a way to go to achieve mature clinical translation, their great potential in the field of targeted cancer therapy has become an emerging hot topic (Figure 5). With the continuous development of artificial intelligence (AI) in recent years, machine learning has also been widely used in ERC. The establishment of predictive models based on exosomes and multimodal data features with a view to improving the accuracy and sensitivity of early diagnosis, predicting patients’ disease progression, risk of recurrence, and response to treatment, and helping physicians to formulate a more personalized treatment plan is a brand new research direction.^110,111

Recommendations for the future research

Overall, we suggest that subsequent experimental studies could target different types of cancer, select representative cell lines and animal models, and comparatively observe the effects of exosome driver genes on the migration and invasive ability of cancer cells by means of techniques such as gene editing, so as to more precisely analyze the specific molecular mechanisms of exosomes in cancer cell metastasis, and to provide a more solid theoretical foundation for targeted intervention in metastatic process. We propose that multi-center, prospective clinical cohort studies can be conducted in the future, not only focusing on the diagnostic efficacy of a single index of exosome, but also exploring the mode of joint detection of multiple markers, and combining with the big data AI analysis methods to establish more accurate, specific and sensitive prediction models, which could help early and accurate diagnosis of cancer and monitoring of the disease. We recommend further exploring the combination of exosomes and new nanomaterials to build intelligent response exosome drug-carrying systems, which can realize precise drug release according to specific signals of the TEM, and meanwhile carry out systematic ex vivo and in vivo pharmacodynamic and safety evaluation studies to accelerate the transformation of exosome-targeted therapies from basic research to clinical application.

It is also worth noting that, combining the themes in the publications and clinical trials we identified, we recognized that breast, lung, colorectal and prostate cancers are the types of cancers for which more research has been accumulated, and subsequent research can be carried out more broadly and intensively on basis of these. As for the other types of cancers, such as bladder cancer, ovarian cancer, lymphoma, and so on, researchers can learn from the experience of exosome research in the above cancers and combine with their own biological characteristics to form a more specific research program to carry out relevant clinical or preclinical studies. In any case, ERC must be a research field full of potential and vitality, which is worthy of the efforts of scholars.

Limitations of this study

Admittedly, some limitations existed in our work. First, the publications and clinical trials included in the analysis were only from WoSCC and clinical trial.gov, which may have overlooked some relevant publications and clinical trials. Secondly, the screening of researches may be partially biased based on the intellectual background of the researchers. In addition, although data cleaning was performed before the analysis, it could only reduce the bias but not eliminate it completely.

Conclusion

From the perspective of bibliometrics, this study visualized the dynamic trend of ERC between 1998 and November 2024 through cluster analysis, citation analysis, and hotspot analysis, summarized the development history of this field, and inspired subsequent research from a macro perspective. In the whole, researches in this field were on an upward trend. Of the 8,001 publications, China, Nanjing Medical University, Whiteside, Theresa L. from USA and International Journal of Molecular Sciences contributed the most. French scholar Théry, C and Cancer Research were the most highly co-cited author and journal. We found that collaboration in research was limited and academic exchange should be improved. And 107 ERC-related clinical trials had a similar spatial and temporal distribution of publications. In addition, detection of thematic trends and evolution suggested that “microRNA,” “surface functionalization,” “plant,” “machine learning,” “nanomaterials,” “promotes metastasis,” “engineered exosomes,” and “macrophage-derived exosomes” are promising hotspots.

Biography

Qiang Pu is currently a Professor of Thoracic Surgery at West China Hospital of Sichuan University and Section Editor of Video-Assisted Thoracic Surgery. His team focuses on basic and clinical research on lung cancer and lung transplantation. He has authored more than 30 scientific papers and has been responsible for and participated in several research projects.

Funding Statement

The author(s) declared that no financial support was received for the research, authorship, and/or publication of this article.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Author contributions

Conceptualization, Yafei Xie and Qiang Pu

Data curation, Yafei Xie, Xingqi Mi, and Yikai Xing

Formal analysis, Yafei Xie

Methodology, Yafei Xie

Project administration, Yafei Xie, Xingqi Mi

Software, Yafei Xie, Xingqi Mi, and Yikai Xing

Supervision, Zhangyi Dai and Qiang Pu

Validation, Zhangyi Dai

Visualization, Yafei Xie, Xingqi Mi

Writing – original draft, Yafei Xie, Xingqi Mi, and Yikai Xing

Writing – review & editing, Yafei Xie, Zhangyi Dai and Qiang Pu

Data availability statement

The data underlying this article were available in Web of Science Core Collection database.

Ethics approval and consent to participate

As an article in bibliometric analysis, ethical approval and consent to participate were not applicable.

Supplementary material

Supplemental data for this article can be accessed online at https://doi.org/10.1080/21645515.2025.2488551

Abbreviations

AI

artificial intelligence

ITAG2

integrin alpha 2 subunit

CAF

cancer-associated fibroblasts

lncRNA

long stranded noncoding RNA

CaM-Ks

calcium/calmodulin-dependent protein kinase

miRNA

microRNA

CD

cluster of differentiation

MRI

magnetic resonance imaging

cirRNA

circular RNA

mRNA

messenger RNA

CTCL

circulating tumor cells

MSCs

mesenchymal stem cells

DALYs

disability-adjusted life years

NCT

national clinical trials

EMT

epithelial-to-mesenchymal transition

NSFC

national natural science foundation of China

EPI

ExoDx prostate intelliscore

PDE

plant-derived exosomes

ERC

exosomes research in cancer

RPYS

reference publication year spectral

ESCRTs

endosomal sorting complexes required for transports

TEM

Tumor microenvironment

EV

extracellular vesicle

TEX

tumor-derived exosomes

GBM

glioblastoma multiforme

TGF-β

transforming growth factor-β

GG

grade group

U-Ex Tg

urinary exosomal thyroglobulin

HSP

heat shock proteins

VEGF

vascular endothelial growth factor

IF

impact factor

WoSCC

web of science core collection