A bibliometric and visual analysis of cancer-associated fibroblasts

Resting mesenchymal cells, commonly known as fibroblasts, are found in normal tissues, residing within the extracellular matrix (ECM) composed of interstitial fibers. These fibroblasts undergo activation in response to environmental factors, particularly during processes such as wound healing, tissue inflammation, and organ fibrosis (1). In contrast, fibroblasts in tumour tissue transform into cancer-associated fibroblasts (CAFs) by receiving signals from nearby tumourigenic cells. CAFs are the predominant stromal cells within the tumour microenvironment (TME). They possess the ability to secrete a diverse range of cytokines and metabolites, thereby impeding immune cell functionality and fostering the advancement of tumours, invasion, angiogenesis, therapeutic resistance, and immune evasion (2–6). As a result, CAFs have emerged as a significant focus in the field of cancer treatment.

In response to tumour-derived stimuli, CAFs are primarily generated by fibroblasts and/or stellate cells residing in the pancreas and hepatic, but CAFs may also arise from bone marrow mesenchymal stem cells that are recruited to the tumour or undergo transformation from adipocytes, pericytes, and endothelial cells (7). Fibroblasts can serve as biomarkers indicating the presence of an immunosuppressive TME, while CAFs contribute to tumour progression by inducing immunosuppression (8). On the one hand, CAFs can establish and modify the ECM framework, thereby facilitating the invasion of tumour cells within the TME. Additionally, CAFs can secrete growth factors, cytokines, and chemokines, which interact with cancer cells or other stromal cells, ultimately impeding the activity of immune cells (9). For example, CAFs secrete transforming growth factor-β (TGF-β), stromal-derived factor 1 (SDF-1), platelet-derived growth factor (PDGF), vascular endothelial growth factor (VEGF) and interleukins (IL), which generate tumour vasculature, promote immunosuppression and enhance drug resistance (10). Furthermore, CAFs secret IL-6 and C-X-C motif chemokine ligand 2 (CXCL12) to strengthen the population of regulatory T cells within the TME and activate factor-related apoptosis ligand and programmed death receptor ligand-2 to induce cytotoxic T cells death in an antigen-dependent manner, thereby inhibiting anti-tumour immunity and immunotherapy (11). On the other hand, CAFs secret large amounts of collagen, fibronectin, etc., to promote the solidification of the extra-tumoural stroma, and the ECM is more abundant, concentrated, and hard in the tumour than in the surrounding healthy tissues, which leads to an increase in the pressure of the interstitial fluid and makes it difficult for the tumour to obtain nutrients, oxygen, immune cells, and therapeutic drugs (12). Moreover, CAFs have the ability to induce Epithelial to mesenchymal transition (EMT), along with aerobic glycolysis, to produce more metabolites such as pyruvate, lactate, and fatty acids, which promote tumour metabolism and nourishment of cancer cells and affect metabolic reprogramming of tumours (13).

CAFs exhibit considerable heterogeneity and plasticity, giving rise to various subtypes that may exert opposing cellular functions (3). α-smooth muscle actin (α-SMA), S100 calcium binding protein A4/fibroblast specific protein 1, fibroblast activation protein (FAP), and Desmin are widely recognized as prevalent activation markers of CAFs (14). Markers for various subtypes of CAFs exhibit variability. Upon receiving diverse signals from other tumour components, such as cytokines, chemokines, or growth factors, stromal cells can generate heterogeneous populations of CAFs with distinct phenotypes and functions. These functions extend beyond the conventional tumour-promoting role (15) and may encompass diverse tumour-suppressive functions. Recently, there has been a notable increase in the volume of publications pertaining to research on CAFs. However, the intricate and uncertain mechanism of action exhibited by heterogeneous subpopulations of CAFs poses a significant challenge in the field of precision tumour therapy. Consequently, considering their intricate involvement in tumour therapy, it is crucial to conduct a quantitative assessment of the current state, research focus areas, and future directions regarding CAFs. Bibliometric analysis, an interdisciplinary field that combines mathematics, statistics, and bibliography, is employed to quantify and visualize research trends. This method has emerged as a crucial approach in assessing the quality and impact of scholarly works (16, 17). The objective of this study is to investigate the research profile of CAFs with the aim of generating novel insights and potential avenues for future research within the domain of CAFs.

We retrieved a total of 7788 published papers from WOSCC, set up exclusion criteria based on publication time, article type, and publication language, and finally screened 5925 papers for bibliometric analysis. The specific flow chart was shown in Figure 1.

CAFs are the most numerous stromal cells in TME, which can not only directly alter the immunosuppressive effect of TME through their own metabolism but also influence the aggregation and function of immune cells by secreting a large number of factors, thus reducing the body’s immune surveillance of the tumour cells, and making the tumour more susceptible to immune escape (21, 22). In light of the therapeutic potential of CAFs within the TME, we conducted a bibliometric analysis using publication data from the Web of Science (WOS). We aimed to elucidate the current development trends and prominent research directions about CAFs. Based on the consistent upward trajectory observed in global publications, it is evident that CAFs have garnered significant interest among researchers, as evidenced by the substantial number of publications exceeding 1,000 in 2022. Furthermore, as of the present search, approximately 900 publications have already been released in 2023, indicating a significant ongoing research potential within the field of CAFs.

In terms of countries/regions and institutions, it could be seen that the number of publications and the number of citations were not precisely the same. Although the USA had the second-highest number of publications, it had the highest average number of citations and displayed extensive collaborative centrality with other countries/regions and institutions. Additionally, it is noteworthy that 7 out of the top 10 institutions in terms of publications were situated in the USA, thus substantiating its position as a frontrunner in CAFs research technology and its substantial contributions to the field. China recorded the highest volume of publications, with Shanghai Jiao Tong University emerging as the foremost institution in terms of publication count. However, it is noteworthy that the average citation count was relatively low, suggesting a limited degree of collaboration with other countries/regions and institutions. These findings suggest that while China has shown a significant commitment to research in the field of CAFs, with substantial investments in research personnel and funding, the quality of academic papers produced in China displays considerable variation. This is also a side effect of the fact that the size of a country’s population introduces a bias in research. The USA has a smaller population than China, so the number of studies will be smaller than China. However, one of the drawbacks of overpopulation is that it produces many low-quality articles, which can have a detrimental effect on academics. It is imperative to prioritize the quality of research papers and foster increased collaboration and cooperation with esteemed institutions in foreign countries, aiming to augment academic influence. Meanwhile, the top 10 countries/regions or organisations in terms of the number of publications were mostly in the developed European and American countries, and there was a high degree of cooperation among them, which confirmed their advanced technology and sensitivity to the research hotspots in the medical field industry, especially in the field of CAFs. It is easy to see that the wealth of different countries is also a potential reason for the existence of a research bias, as this would limit investment in health research. Developed countries have more wealth, so that research can be funded more, is more intensive and yields more results. It is also worth noting that an author may be involved in several working affiliations simultaneously, but only one may be the primary, real working institution. This will inevitably result in a less precise count of authors and institutions. This may be caused by the fact that the author is attached to more than one institution, which may be an honour, whereas there is only one main institution. Or the author changes the body of work, but the paper retains the initial body of work. One way to really solve this statistical problem is to rely on the fact that authors need to mark their main working institution when papers are reviewed. Secondly, the literature database needs to be constantly updated with algorithms that can promptly and accurately label and update the institutions where authors are located at the time of publication and limit the number of institutions to which the literature is exported. Thirdly, it is checked and verified through manual auditing by many people, but this is a very heavy workload that could perhaps be aided by artificial intelligence.

Sotgia, Federica, Martinet, Wim, and Lisanti, Michael P emerged as trailblazers and frontrunners in the domain of CAFs, exhibiting exceptional prowess in authorship and co-authorship analyses. These esteemed individuals demonstrated remarkable leadership by virtue of their substantial contributions in terms of publication count, citation count, and H-index. Their research pursuits demonstrated significant overlap, and their scholarly works primarily showcased collaborative efforts. They focused on how certain proteins and genes drive the phenotype of CAFs, as well as exploring the mechanisms by which CAFs promote tumour growth and progression through tumour metabolism, such as aerobic glycolysis and mitochondrial oxidative metabolism, and confirming the important role of oxidative stress on CAFs’ metabolism (23–30). Studies were also involved in the metabolic effects of inflammatory signals such as NF-κB with oncogenes on TME and worked on the development of anticancer drugs (31–33). Meanwhile, from the author cooperation graph, it could be found that there was active cooperation among authors from different countries/regions and institutions, but the cooperation was still dominated by the same countries/regions and institutions.

Cancers emerged as the journal with the highest number of publications, whereas Cancer Research garnered the highest number of citations. Most of the journals occupying top positions in publications and citations were situated within the JCR Q1 region, signifying their significant academic influence. Notably, Frontiers in Oncology, Frontiers in Immunology, Cancers, Oncogene, International Journal of Molecular Sciences, and Cancer Letters serve as the primary platforms for disseminating information on CAFs at present. Simultaneously, Cell, Nature, and their affiliated sub-journals emerged as preeminent publications within the domain of CAFs, garnering substantial citation counts. Consequently, researchers ought to diligently enhance the paper’s caliber in subsequent stages, with the ultimate objective of securing publication in Cell, Nature, or their associated sub-journals.

The disciplinary analysis of journals revealed that research in the field of CAFs was concentrated in the disciplines of medicine, medical, and clinical, as well as in the disciplines of molecular, biology, and immunology. Additionally, the disciplines of health, nursing, and medicine, along with the disciplines of molecular, biology, and genetics, provided a strong knowledge base for research in these fields. Publications in the molecular, biology, and immunology disciplines were mainly influenced by publications in the molecular, biology, and genetics disciplines. Combined with the bi-plot superimposition analysis, the research articles in the field of CAFs were mainly published in journals in the disciplines of molecular, biology, immunology, and medicine. Molecular, biology, immunology, medicine, and genetics were the main research disciplines in this field. Meanwhile, we can see that the research of CAFs is multidisciplinary and multidisciplinary, and the communication and mutual influence among various disciplines also help to promote the progress and development of the field.

Analysis of the references showed that 7 of the top 10 most cited articles were clinical trial articles, two reviews and one theoretical framework. These publications primarily focused on breast, pancreatic, and skin cancers, thereby reflecting the extensive research background on CAFs and carrying significant implications for the advancement of CAF-related research. Co-citation and timeline analyses can better reflect the direction of research in a particular field and the development of academic crossover, and it can be seen that the research centre of gravity of CAFs has begun to shift towards extracellular vesicle,extracellular matrix,pan-cancer analysis,single-cell atla,preclinical evaluation and so on. In terms of the literature explosion, we found that most of the early studies on CAFs were focused on breast cancer; for example, CAFs promoted breast cancer growth and angiogenesis through the secretion of SDF-1/CXCL12, and induced metastasis through the secretion of the chemokine ligand 5 (19, 34). In addition, CAFs mediated the pro-inflammatory effects of tumours through NF-κB signalling (35). In the medium term, the direction of research shifted from the cancer cells to the microenvironment in which they were embedded, such as glycolysis and interactions between CAFs of different phenotypes in the tumour stroma (36, 37). The role of CAFs in TME also received increasing attention (38, 39). At the later stage, the focus of research on CAFs was reflected in the heterogeneity of CAFs and their different phenotypes, and exploring targeting the phenotypic heterogeneity of CAFs for anti-tumour effects, as well as more in-depth research on CAFs in TME (1, 40–43). Furthermore, Pancreatic Ductal Adenocarcinoma emerged as a prominent focus in CAFs research (44, 45). However, our investigation revealed a dearth of literature on CAFs in the cancer treatment context, with limited representation of various cancer types. Consequently, it is imperative for future scholars to augment research endeavours in the realm of clinical treatment and delve into the correlation between CAFs and specific cancer types, such as the association between CAFs and lung cancer alongside pulmonary fibrosis.

From the keywords co-occurrence and clustering analyses, the research direction on CAFs shifted from the molecular biological properties of CAFs and the mechanism of CAFs in the pathological process of tumour diseases to the cancer therapy targeting CAFs. CAFs were clearly confirmed to facilitate tumour growth, metastasis and invasion, etc., and were closely associated with tumour progression, especially in breast, colorectal and pancreatic cancers.

From the keywords burst, the research direction of CAFs was roughly the same; here, we focused on analysing the molecular targets and mechanisms of action of CAFs with tumours. Angiogenesis is necessary for tumour cell proliferation and survival (46). The keyword hotspots of early CAFs were all related to angiogenesis. CAFs could secrete a large number of growth factors, cytokines and chemokines and mediate related signalling pathways such as TGF-β/TGF-β receptor, VEGF/VEGF receptor and SDF-1 (CXC12)/CXC chemokine receptor type 4, or synthesize related proteins such as podoplanin, FAP and Winglesstype MMTV integration site family member 2 (47), or secret MMP-9/13 to promote tumour angiogenesis (48–51).

At mid-term, oxidative stress had the highest citation burst intensity among keywords. Active expression of TGF-β resulted in the deletion of Caveolin-1, which led to metabolic reprogramming of CAFs and promoted their autophagy and aerobic glycolysis, resulting in oxidative stress (52). The excessive presence of oxidative stress within cells can induce modifications in DNA bases, leading to the formation of strand breaks, activation of proto-oncogenes, and inactivation of oncogenes. The progression and development of tumours are intricately linked to these molecular events. Oxidative stress in the body primarily generated reactive oxygen species (ROS), with hydrogen peroxide (H₂O₂) being the principal ROS produced by tumour cells. H₂O₂ was a major signalling molecule that could promote cancer cell proliferation, survival and invasion through redox signalling (53). Interestingly, another study showed that impaired TGF-β signalling led to higher H₂O₂ production by CAFs (54). This proves that our studies on CAFs are far from sufficient, and more experiments are still needed to clarify their relationship. Targeting the redox pathway seems to be a potential therapeutic strategy for targeting CAFs against tumours. Also, myofibroblast differentiation had a higher bursting intensity, which played a pivotal role in the malignant progression of lobular tumours. Tumour cells could activate mesenchymal fibroblasts, leading to their differentiation into CAFs with myofibroblast characteristics, thereby enhancing the aggressiveness of the cancer (55).

At a later stage, the investigation of cancer therapy targeting CAFs emerged as a prominent area of research. Signal transducer and activator of transcription-3 (STAT3) was identified as a critical mediator of the oncogenic effects of CAFs and was a key transcription factor regulating the function of CAFs while crosstalking with tumour cells and immune cells within the TME (56, 57). For instance, the pro-inflammatory cytokine IL-6 stimulated the expression of Twist1 in normal fibroblasts, leading to their transdifferentiation into CAFs through STAT3 phosphorylation, thereby facilitating tumour invasion (58). So, targeting STAT3 is a promising therapeutic target. Meanwhile, CAFs could provide cancer cells with protective ecological niches for anti-cancer drugs, while the symbiotic relationship between tumours and CAFs provided key resources, such as growth factors and nutrients, for optimal tumour growth and proliferation, so CAFs promoted tumour therapeutic resistance (59, 60). A study showed that drug-resistant tumour tissues from gastric cancer patients were enriched in CAFs, which promoted the development of chemoresistance by secreting IL-11 and activating the IL-11/IL-11 receptor/glycoprotein 130/Janus kinase/STAT3 anti-apoptotic signalling pathway in gastric cancer cells, thus targeting IL-11 to CAFs could alleviate drug resistance (61). In addition to resistance to cancer treatment drugs, CAFs are also involved in resistance to radiotherapy. Although IR was effective in killing cancer cells (62), CAFs could attenuate IR-induced cancer cell death by modulating the DNA damage response. Simultaneously, the activation of CAFs by IR might induce cancer cell proliferation, migration, and radiotherapy tolerance (63). In contrast, CAFs subjected to IR lost their pro-tumour growth potential in vivo, which might be related to affecting angiogenesis and tumour engraftment (64). IR inhibited the proliferation, migration and invasive ability of CAFs in lung cancer, which might be related to the power of IR to promote MMP-3, inhibit MMP-1, and enhance integrins α2, β1 and α5 to stabilise lesion exposure (65). It can be seen that the effect and mechanism of action of IR on CAFs are more complex and may also be related to the radiotherapy dose and changes in the secretion rate and composition of extracellular vesicles (66), which need to be further researched and investigated. Another hotspot in the keyword explosion was GNPs, which had been applied to the field of CAFs due to their unique biological functions. GNPs converted activated CAFs to a quiescent phenotype, which might be related to the endogenous synthesis of lipids induced by GNPs (67). Meanwhile, GNPs could inhibit the activation of CAFs and exert anti-CAFs effects, which might be due to their ability to alter a variety of fibroblast activating or inactivating proteins secreted by cancer cells and TME cells such as TGF-β1 and PDGF-1 (68, 69). Meanwhile, the high uptake of GNPs by CAFs in radiotherapy confirmed that GNPs could be an effective sensitiser in future cancer radiotherapy, demonstrating the very high therapeutic potential of cancer nanotechnology (70, 71).

The utilization of bibliometric analysis serves as a crucial approach for investigating the trajectory of development and identifying prominent areas of research within the domain associated with CAFs. However, it is imperative to acknowledge that this study methodology is not without its constraints. Firstly, the reliance on the WOSCC database for retrieving literature information may lead to the exclusion of pertinent data from alternative databases. Secondly, the omission of publications from the year 2023 may result in a failure to capture the most recent research hotspots and trends. Thirdly, despite employing double-checking measures, the necessity for manual correction of overlapping or inaccurate elements remains unavoidable. Fourthly, the limitations of the research algorithms of the bibliometrics themselves have allowed the contributions of many emerging researchers to be overlooked, while information bias between different measurement software has led to inconsistent results. Finally, the removal of articles other than Article and Review from the inclusion criteria may introduce some bias, although the overall bias is manageable due to the relatively small number of exclusions.

In summary, the field of research on CAFs is experiencing rapid advancements, and there is a growing anticipation that cancer therapy targeting CAFs will emerge as a prominent area of investigation. Furthermore, the next phase requires extensive international collaboration to provide new insights and directions for the ongoing exploration and advancement of CAFs.

The original contributions presented in the study are included in the article/. Further inquiries can be directed to the corresponding authors. 1

W-CY: Investigation, Writing – original draft, Writing – review & editing. J-XZ: Investigation, Writing – original draft, Writing – review & editing. H-BC: Investigation, Writing – original draft. YY: Investigation, Writing – review & editing. Y-PZ: Investigation, Writing – review & editing. H-LZ: Investigation, Writing – review & editing. M-HL: Conceptualization, Methodology, Supervision, Writing – review & editing. W-LQ: Conceptualization, Methodology, Supervision, Writing – review & editing. H-GZ: Conceptualization, Methodology, Supervision, Writing – review & editing.