The Web of Science Core Collection (WOSCC) was selected to search for relevant literature. Search terms were developed using synonyms from Medical Subject Headings (MeSH) and related textual terms. Follow the search strategy below: TS= (“peritoneal malignancies” OR “peritoneal cancer” OR “peritoneal malignancies” OR “peritoneal neoplasms” OR “peritoneal malignancy” OR “peritoneal tumor”) AND (“tumor microenvironment” OR “cancer microenvironment”). Further screening was conducted based on the following criteria: (1) publication date between January 1, 2010 and August 12, 2024; (2) language restricted to English to avoid bias from multilingual translations; and (3) document type limited to articles and reviews. The processes of data collection, selection, and extraction were independently conducted by two authors and subsequently cross-validated. Discrepancies during the research process were resolved through discussion or consultation with CH Chen. All relevant literature was retrieved and downloaded in “Plain text” format to minimize potential biases caused by database updates. The content of the records and the cited references were comprehensively documented. All valid data were imported to CiteSpace and deduplicated for subsequent visual analysis. The retrieval process is illustrated in Figure 1 .

Microsoft Office Excel 2019 was used to calculate the total number of annual publications, average citations, annual publications by country, cumulative publications by institutions, journals and authors, as well as the impact factor (IF) and H-index. The research overview, hot spots and trends were obtained using bibliometric methodology (6). VOSviewer, a widely recognized bibliometric analysis software, can generate maps of publications categorized by countries, institutions, journals, and authors. The size, color, and thickness of node connections on the map indicate various relationships and levels of collaboration. CiteSpace (version 6.3.1), another bibliometric analysis software, was used to draw a dual-map overlay and to analyze references using Citation Bursts (7). The R package “bibliometrix” (version 4.3.0) was used to create a global distribution network of publications of TME of PM.

Based on the established search strategy, a total of 862 studies were included, comprising 688 articles and 174 reviews related to the TME in PM. Figure 2 showed the basic characteristics of the 862 publications. Figure 2A illustrates that the number of publications increased steadily from 2010 to 2024, with minor decreases in 2011 and 2013. To date, the 862 publications have received a total of 23,352 citations, resulting in an average citations per publication (CPP) of 32.89. Consistent with the publication trend, the number of annual citations has also gradually increased, peaking at 5,040 citations in 2023. In 2021 and 2022, the number of citations surpassed 4,000, reaching 4,116 and 4,821 citations respectively, as shown in Figure 2B . The H-index of all 862 publications was 82. Specifically, the H index showed a trend of first rising and then falling, with the highest in 2019 which was 32. It is important to note that the figures for publications, annual citations and H-index in 2024 do not represent the complete totals for the entire year.

Publications on the TME in PM originated from 56 countries. Table 1 presents the characteristics of the top ten countries, including four from Europe, three from Asia, two from North America, and one from Oceania. The USA, China and Japan contributed the highest proportion of publications, accounting for 32.60%, 26.22%, and 13.11%, respectively. Together, China, the USA and Japan produced nearly three-quarters of the total publications (73.93%). The H-index, total link strength (TLS), and total citations were largely consistent with the number of publications. Research from the USA, China and Japan was relatively more in-depth and recognized ( Figure 3A ). Strong links were observed among different countries, indicating some overlap in published content. Countries with five or more publications were visualized, and a collaborative network was constructed, revealing close cooperation and exchanges. The USA exhibited the most frequent collaborations with China, Japan, Australia, Canada and so on. China also maintained active collaborations, followed by Japan ( Figures 3B–D ).

1291 institutions have published articles of the TME in PM. Given that the number of publications is consistent across multiple institutions, we have counted the top 13 institutions, which are shown in Table 2 . And the top 13 institutions are located in China, the USA and Japan, with six of them based in China. The University of Texas MD Anderson Cancer Center has the highest publication count with 22 papers, followed by Sichuan University with 20 papers and Fudan University with 19 papers. National Cancer Center occupied the highest contribution of citations and average citation (57.41) while The University of Texas MD Anderson Cancer Center had the highest H-index which was 15. Figure 4 illustrates the strong cooperation among various institutions, including The University of Texas MD Anderson Cancer Center, Indiana University School of Medicine, and University of Pittsburgh.

Authors and co-cited authors were analyzed to identify leading scholars and influential contributors in the field. A total of 5,914 authors contributed to 862 articles on the TME in PM. Due to ties in publication counts, the top 14 authors are listed in Table 3 . Four of the top five authors were from Japan, including Kajiyama Hiroaki (11 articles), Yashiro Masakazu (10 articles), Fushida Sachio (8 articles), and Kinoshita Jun (8 articles). They also achieved the highest H-index (nine, eight or seven). The highest average number of citations(62.86) was recorded by Müller Rolf and Reinartz Silke from Germany. A collaborative network was also established for authors with four or more published articles. Visual analysis revealed close collaborations, particularly among Kajiyama Hiroaki, Yashiro Masakazu, Karen K. L. Chan, and Fushida Sachio ( Figure 5A ). Among 31,744 co-cited authors, 22 were cited more than 50 times. Specifically, Mantovani, A is the most co-cited author (150 times), followed by Kenny H.A. (143 times) and Hanahan D (114 times). Two authors had TLS exceeding 2000, namely Kenny H.A. (TLS=2852) and Nieman K.M. (TLS =2249). Positive collaborations were also evident among different co-cited authors ( Figure 5B ).

The study retrieved articles published in 345 journals, with details of the top 10 journals presented in Table 4 . The top three journals are Cancers (6.5% of 862 articles, 56 papers), International Journal of Molecular Sciences (3.6% of 862 articles, 31 papers), and Frontiers in Immunology (3.25% of 862 articles, 28 papers). Oncogene has the highest average citation (98.47), followed by Cancer Research (83.25 average citations). Cancer Research (IF=12.5) was the journal with the highest IF, and Journal for Immunotherapy of Cancer (IF=10.3) ranked second. Cancer has the highest H-index at 21, followed by Plos One with an H-index of 17. We visually analyzed 39 journals with at least five articles and plotted the network ( Figure 6 ). Figure 6A displays six clusters and 181 links among the journals. Cancer has close citation relationships with journals such as International Journals of Molecular Sciences, International Immunopharmacology, and Plos One. There are also close links between other magazines, including Frontiers in immunology, Frontiers in oncology, etc. Table 5 lists the top ten collectively cited journals, with eight are from the USA and two are from the England. Cancer research (n=2355) and Clinical Cancer Research (n=1275) are two journals with more than 1,000 citations. The co-citation network map in Figure 6B illustrates journals with a minimum of 100 co-citations.

Figure 6C presents a dual-map overlay encompassing all journals. There are one green path and one orange path which represented different citation paths. The orange path indicates that Molecular, Biology, and Immunology are the main fields of articles included. The green path indicates Medicine, Medical and Clinical are the main fields. Interestingly, the fields of Molecular, Biology and Genetics cover most of the cited articles, both in the orange and green paths.

The analysis of co-cited references identified the most influential publications and helped generalize research hotspots. A total of 43854 co-cited references were recorded for TME in PM. Due to ties in citation counts across multiple journals, Table 6 presents the key features of the 12 most highly cited references. The most frequently cited reference was “Ovarian cancer development and metastasis” published in American Journal of Pathology, with 80 citations, followed by Kristin M Nieman’s article “Adipocytes promote ovarian cancer metastasis and provide energy for rapid tumor growth” published in Nature Medicine with 73 citations. Figure 7A illustrates close links among references cited at least 20 times.

Reference with citation bursts indicate publications that were widely cited over a specific period. In this study, the earliest citation burst occurred in 2012, referencing a 2010 publication. Currently, seven references continue to exhibit active citation bursts. The reference “Extracellular vesicles: exosomes, microvesicles, and friends” by Hanahan D demonstrated the strongest citation burst (strength=9.07) during the period from 2012 to 2016. Among the 25 references analyzed, the lowest citation burst strength was 4.9 ( Figure 7B ).

Keywords summarize the principal concepts presented in the article. We create visual maps and perform cluster analysis for keywords appearing more than ten times. There are 300 links and six clusters for 39 keywords, represented by 6 colors. Cluster 1 contained 10 items including cancer-associated fibroblast, chemoresistance, epithelial ovarian cancer, gastric cancer, malignant ascites, pancreatic cancer, proteomics, tumor microenvironment and tumor-associated macrophages. Cluster 2 contained 9 items including cancer, cancer-associated fibroblast, chemotherapy, colorectal cancer, immunotherapy, metastasis, peritoneal carcinomatous, peritoneal metastases and prognosis. Cluster 3 contained 7 items including angiogenesis, cytokines, endometriosis, hypoxia, immunosuppression, inflammation and macrophages. Cluster 4 contained 6 items including ascites, extracellular matrix, mesothelial cells, ovarian cancer, ovarian carcinoma and peritoneal dissemination. Cluster 5 contained 6 items including breast cancer, exosomes, macrophage, microenvironment, omentum and peritoneal cavity. Cluster 6 contained 1 item including extracellular vesicles. Figures 8A, B illustrates the latest perspective of the TME in PM. The yellower points indicate key components, including cancer-associated fibroblasts (CAFs), prognosis, malignant ascites, epithelial, chemoresistance, tumor-associated macrophages, chemotherapy, immunotherapy and other factors.

The keyword timeline map integrates temporal and thematic dimensions, providing a dynamic representation of evolving research trends over time ( Figure 8C ). Most articles were published after 2010. It was found that “ovarian cancer” (cluster #1), and “gastric caner” (cluster #5) remained consistent hotspots from 2010 to 2024. “Endothelial growth factor” (cluster #9) had the shortest hotspot research span. By 2023, “tumor-associated macrophages”(Cluster #2), “hyperthermic intraperitoneal chemotherapy” (Cluster #3), and “peritoneal dissemination” (Cluster#4) continued to be major research hotspots. Notably, “immunotherapy”(Cluster #3), “endometriosis” (Cluster #6) and “peritoneal metastasis” (Cluster #7) emerged as the latest research hotspot in 2024.

Burst keywords are also important indicators reflecting emerging frontiers in the field. The first cited burst keywords appeared in 2010, involving keywords such as in vivo, in vitro and angiogenesis. Over time, immune cells in TME have increasingly become a research focus. Keywords such as cytoreductive surgery, tumor-associated macrophages (TAMs), open label, intraperitoneal chemotherapy and CAFs continued to emerge through 2024, indicating these topics will likely remain future research hotspots deserving high attention ( Figure 8D ).

From 2010 to 2024, publications of the TME in PM steadily increased, with only minor declines in 2011 and 2013, indicating strong interest among researchers. Annual citations reflect research popularity, and our study found that they have increased over time. Since 2020, annual citations have surpassed 3,000. An analysis by country shows that the USA has the highest number of publications, H-index, TLS, and total citations, indicating its leadership in this area of research. China and Japan ranked second and third, respectively. Together, they account for nearly three-quarters (73.93%) of the papers published in China, the USA and Japan, indicating that research in these three countries is more in-depth and recognized in relative terms. Cooperation and exchanges between countries are very close. The USA has a connection strength of 124, demonstrating its close cooperation with China, Japan, Germany, Australia, Canada and other countries. China similarly engages in close collaborations with numerous countries, and Japan is also actively involved in such partnerships. The top 13 institutions are located in China, the USA and Japan. The University of Texas MD Anderson Cancer Center collaborates closely with the Indiana University School of Medicine, the University of Pittsburgh, and the University of Chicago. However, it has limited institutions with institutions in China. Similarly, Fudan University engages more with domestic institutions but lacks cooperation with international ones. The above results show that establishing outstanding research institutions is crucial for enhancing research quality and that strengthening communication among different institutions is equally important.

Among the top five authors, four are from Japan and one is from China: Kajiyama Hiroaki (11 articles), Yashiro Masakazu (10 articles), Fushida Sachio (8 articles), Kinoshita Jun (8 articles) and Karen K. L. Chan (8 articles). Kajiyama Hiroaki, Yashiro Masakazu, Karen K. L. Chan, Fushida Sachio had more connections with other authors. As shown in Table 4 , the current journals with the most research publications are Cancer(56 articles), International Journal of Molecular Sciences (31 articles), and Frontiers in Immunology (28 articles). The journals with the highest IF were Cancer Research (IF=12.5), followed by Journal for Immunotherapy of Cancer (IF=10.3), with all journals having an IF between 2.9 and 12.5. Most of these journals are classified in the JCR Q1 region, indicating their high research value in this field. Among the co-cited journals, most high-impact Q1 journals offer robust evidence for TME research related to PM. All journals having an IF between 2.9 and 72.5. This range of impact factors demonstrates the overall quality of research published in these journals.

Co-cited literature serves as the foundational research in a specific field. The 13 most cited articles form the foundation of the TME field in PM. Except for “Hallmarks of cancer: the next generation, “, a 2011 review in Cell by Professor Douglas Hanahan, the remaining 12 papers focus on ovarian cancer. Research interests encompass the mechanisms by which adipocytes, mesothelial cells, tumor-associated fibroblasts and regulatory T cells promoting peritoneal metastasis of ovarian cancer, as well as the treatment of ovarian ascites and genomic analysis of the disease.

References with a citation burst are those frequently cited by researchers recently, indicating an emerging topic in the field. We observed that the literature with citation bursts first appeared in 2012, published by Ernst Lengyelin 2010, titled “Ovarian cancer development and metastasis” ( 8). This article mentioned that the biological characteristics of ovarian cancer are different from other hematogenous metastases, and peritoneal dissemination is the main mode of metastasis. Ovarian cancer cells can adhere to the peritoneum and settle down, eventually leading to peritoneal metastasis. At present, the bursts of seven studies are still not over in which two studies each reviewed the epidemiology, pathogenesis, and treatment advances of ovarian and peritoneal cancers. Besides, WonJae Lee found that neutrophils have a negative effect on the peritoneal environment, promoting ovarian cancer metastasis (9). Andrew Chow from Weill Cornell Medical College found that peritoneal macrophages had an effect on the adaptive immune function of CD8⁺ T cells (10). Benjamin Izar (11) utilized single-cell RNA sequencing to analyze cells from ascites specimens of patients with high-grade serous ovarian cancer, creating a detailed single-cell landscape. This work laid the foundation for immunotherapy treatments for ovarian cancer.

Hotspots and frontiers are determined by cluster analysis of common keywords and burst keywords. Base on the above, we summarized the frontiers and hotspots of TME research in PM:

In addition, the frontiers and hotspots of the TME have been explored in lung (63), colorectal (64), hepatocellular (65), breast (66, 67), pancreatic (68), hematological malignancies (69) and cervical cancer (70) through bibliometric studies. By keyword analysis, five studies identified immunotherapy as a future research direction, three studies highlighted the critical role of CAFs in regulating the TME, and one study mentioned macrophage polarization. These findings are consistent with the hotspots identified in our study. Other emerging directions, including new technologies, ferroptosis, biomarkers, and nanoparticles, are also worthy of exploration in peritoneal cancer. In terms of methodology, nearly all studies were restricted to English publications, with VOSviewer, CiteSpace, and the R package ‘bibliometrix’ being the most commonly used visualization tools. China and the USA were the countries with the highest number of publications. Related studies were more frequently published in Frontiers in Immunology, Frontiers in oncology, and Cancer. Thus, we conclude that the immune microenvironment is the most critical component of the TME, closely linked to the treatment of most malignancies, and will remain a research hotspot for the foreseeable future.

Although this study followed the principles of bibliometric analysis, several potential limitations exist. First, the analysis was restricted to English-language publications, which may have introduced publication bias. Nevertheless, given the relatively small proportion of non-English studies, the WoSCC still captures the majority of relevant research, ensuring the representativeness of our findings. Second, analyses were limited to articles and reviews published in WoSCC, and articles included in other databases (e.g., PubMed and Scopus) should be followed up jointly to avoid omissions. Third, the literature search concluded in August 2024, which may have resulted in missing newly published studies during the subsequent period. Additionally, some recent high-quality studies may have been overlooked due to their low citation counts. Finally, bibliometric analysis is limited to metadata and does not involve full-text examination, potentially missing significant insights such as authors’ perspectives and future research directions.