We conducted a search of the Web of Science Core Collection (WoSCC, https://www.webofscience.com/wos/woscc/basic-search) using the following terms: “TS = [(“Stem Cell*"OR “Stromal Cell*") AND TS = (“ILD” OR “interstitial lung disease” OR “pulmonary Fibrosis”)]” on 5 August 2022. Only articles and reviews published in English were selected. Other document types, such as early access, book chapters, meeting abstracts, and letters, were excluded. The details of literature selection are provided in Figure 1.

VOSviewer (version 1.6.18) was used to analyze the cooperation between countries and institutions, journals and co-cited journals, authors and co-cited authors, keyword co-occurrence, and network visualization map. In the map, nodes represent various items, such as country, institution, journal, and author, with node size representing the number of included items. The thickness of the lines linking different nodes reflects the degree of cooperation or co-citation of projects (Zhang et al., 2020; Wu et al., 2021).

CiteSpace (version 6.1. R1) was employed to draw the dual-map overlay of scientific journals that published articles on MSCs research in the field of PF. CiteSpace was also used to analyze references with citation bursts. Microsoft Office Excel 2019 was used to manage the data and assist the analysis.

A total of 1,457 studies (984 articles and 473 reviews) that included MSCs and PF as keywords were identified from 2002 to 2021. The number of studies increased steadily from 2002 to 2015, with a more rapid increase beginning in 2016; over 100 studies were published yearly (Figure 2).

The authors were distributed among 54 countries and 1,661 institutions. As shown in Table 1, the top country was the United States of America (United States) (n = 632, 30.5%), followed by China (n = 294, 14.2%), Germany (n = 124, 6.0%), Italy (n = 118, 5.7%), and Japan (n = 105, 5.1%). Literature from the United States and China accounted for almost half of all the retrieved studies. Countries with frequencies ≥6 (n = 31) were selected for visualization analysis, and a cooperation network was constructed according to the number and relationship of publications in each country/region (Figure 3A). Positive cooperation was found between China and the United States, as well as Australia and Japan. The United States also had close collaboration with Japan, Germany, France, and Italy.

The top 10 institutions publishing on MSCs and PF were from four countries. The majority of these institutions (6/10) was in the United States. The University of Michigan (n = 64, 2.8%), University of Pittsburgh (n = 47, 2.0%), University of California Los Angeles (n = 29, 1.3%), Monash University (n = 25, 1.1%), and Harvard University (n = 24, 1.0%) were the top five institutions. Institutions with frequencies ≥13 (n = 37) were selected for visualization analysis, and a cooperation network was constructed according to the number and relationship of each institution (Figure 3B). According to the results, there was a close association between the University of Michigan and the University of California Los Angeles, and similar findings were observed between German CTR Lung Res DZL and Helmholtz Zentrum Munchen, and the University of Michigan and Mayo Clinic. Although a large number of studies was published from Nanjing University, the collaboration of Nanjing University with other institutions was modest.

The literature were published in 508 academic journals. Most studies (n = 39, 7.7%) were published in PLOS ONE, followed by the American Journal of Physiology—Lung Cellular and Molecular Physiology (n = 38, 7.5%), Stem Cell Research and Therapy (n = 37, 7.3%), and the American Journal of Respiratory Cell and Molecular Biology (n = 34, 6.7%). Among the top 15 journals, the American Journal of Respiratory and Critical Care Medicine had the highest impact factor (IF = 30.528), followed by the Journal of Clinical Investigation (IF = 19.456). Journals with ≥8 publications were used to construct the journal network map (Figure 4A).

Among the top 15 co-cited journals, five journals had a total citation of more than 2,000 times (Table 2). The American Journal of Respiratory and Critical Care Medicine (n = 4,307) was the most frequently cited journal, followed by Proceedings of the National Academy of Sciences (n = 2,624), the American Journal of Respiratory Cell and Molecular Biology (n = 2,411), and the Journal of Clinical Investigation (n = 2,399). Among the top 15 journals, the New England Journal of Medicine had the highest impact factor (IF = 176.079), followed by Nature (IF = 69.504). Journals with ≥500 co-citations were analyzed to determine the co-citation network (Figure 4B). In summary, a positive co-citation relationship was observed between the American Journal of Respiratory and Critical Care Medicine and Chest, as well as Proceedings of the National Academy of Sciences and the New England Journal of Medicine.

The dual-map overlay of journals was used to analyze the relationship of citations between journals and co-cited journals (citing journals, left; cited journals, right) (Chen, 2017). The orange and green lines are the main application paths. The orange lines indicate studies published in molecular, biology, genetics, health, nursing, and medicine journals; the primary citations were in molecular, biology, and immunology journals. The green lines indicate studies published in molecular, biology, and genetics journals; the primary cited studies were in medicine, medical, and clinical journals (Figure 5).

A total of 7,842 authors were retrieved during the literature search. As shown in Table 3, the first two authors have published at least 15 studies. A collaboration network was conducted with the authors who published ≥7 papers. According to the number of relevant published studies, Rojas Mauricio, Xiaodong Han, Phan Sem H, and Weiss Daniel J had the largest nodes, and a close collaboration was observed between Rojas Mauricio and Mora Ana L, Noble Paul W and Stripp Barry R, and Henke Craig A and Bitterman Peter B (Figure 6A).

Co-cited authors are defined as authors who are co-cited in a series of publications. Among the 43,654 co-cited authors, 11 of them were co-cited >200 times. Raghu G had the highest number of co-citations (n = 494), followed by Khanna D (n = 373) and Ortiz La (n = 296). Authors with ≥120 co-citations were analyzed using the co-citation network graph (Figure 6B). Many positive cooperative relationships among different co-cited authors were found, such as Khanna D and Steen VD, and Tashkin Dp and Raghu G and King Te.

Co-cited references are defined as those references cited together by other publications. A total of 64,747 co-cited references were found in the publications related to MSCs and PF research. Among the top nine co-cited references related to MSCs and PF research (Table 4), each reference was co-cited ≥100 times. References with ≥60 co-citations were analyzed using the co-citation network diagram (Figure 7A). The data showed that “Rojas M, 2005, Am J Resp Cell Mol” had active co-citation relationships with “Ortiz La, 2003, P Natl Acad Sci United States”.

A citation burst refers to references that are frequently cited by scholars in a particular field over a period of time. Ten references with strong citation bursts were detected using CiteSpace, and the minimum burst duration of literature related to MSCs and PF research was 3 years. In Figure 8, the red bar represents the reference burst. The earliest citation burst occurred in 2004, and the latest occurred in 2015. Among these references, “Bone marriage-derived progenitor cells in Pulmonary Fibrosis” (authored by Hashimoto N et al.) had the strongest citation burst (strength = 16.99), which appeared between 2004 and 2009. The second reference was “Mesenchymal stem cell engraftment in lung is enhanced in response to bleomycin exposure and ameliorates its Fibrotic Effects” (authored by Ortiz L et al.; strength = 13.48; from 2004 to 2008). Overall, the burst strength of the top 10 references ranged from 11.4 to 16.99, while duration was from 3 to 5 years. Table 5 summarizes the main research content of the 10 references with the strongest citation bursts from Figure 8.

The high-frequency keywords in the field of MSCs and PF research are shown in Table 6, amongst which IPF, MSCs, and systemic sclerosis appeared >100 times, representing future research directions. Other keywords, such as inflammation, myofibroblasts, fibroblasts, aging, telomerase or telomere, extracellular matrix, and epithelial-mesenchymal transition, appeared >20 times, which may point to areas of mechanistic studies involving MSCs and PF.

The cluster analysis of keywords was performed using VOSviewer software. Each circle represents a keyword, the size of the circle indicates the extent of positive correlation with the frequency of the keyword, and the line thickness linking the circles indicates the degree of correlation between these keywords. A total of four clusters were obtained (Figure 7B), suggesting four main research directions. The yellow cluster includes systemic sclerosis, rheumatoid arthritis, interstitial lung disease, pathogenesis, and stem cell transplantation. The blue cluster includes aging, stem cells, telomerase or telomere, regeneration, apoptosis, and cancer. The red cluster includes myofibroblasts, fibroblasts, EMT, tumor TGF-β, and cytokines. The green cluster includes PF, MSCs, inflammation, and extracellular vesicles (EVs).

The current status of the research Frontier and the challenges of MSCs and PF are discussed in detail. Since the first isolation by Friedenstein et al. (1970) in 1970, bone marrow-derived mesenchymal stem cells (BM-MSCs) have become the most common source of multipotent cells for transplantation in preclinical and clinical trials. However, harvesting BM-MSCs is a painful, invasive process and carries a risk of viral exposure. In addition, the number, differentiation potential, and maximum longevity of BM-MSCs decreases with age (Bustos et al., 2014; Cárdenes et al., 2018; Nehlin et al., 2019). Compared to BM-MSCs, umbilical cord derived mesenchymal stem cells (UC-MSCs) have a similar profile, are easier to harvest, have no ethical or safety concerns, and are more rapidly refreshed (Ding et al., 2015). At present, many preclinical studies have confirmed that MSCs are safe and effective in the treatment of PF. One meta-analysis (Li et al., 2021) of 24 preclinical studies involving MSCs-targeted PF treatment showed that MSC treatment significantly improved patient survival rate and alleviated PF in animal models. All published clinical trials that have applied MSCs for PF treatment have confirmed the safety of MSC treatment and the benefits for some patients (Table 7). Of the clinical studies included in Table 7, all were collected from clinicalTrials.gov, except the fifth study (No. 5). Recently, some progress has been made in engineering MSCs to regulate and guide their behavior. For example, endocytic superparamagnetic iron oxide nanoparticles have been shown to improve MSC targeting to the lung (Silva et al., 2018).

Due to the physical properties of MSCs, exogenous engraftment of MSCs in vivo after infusion seems to be rare at the site of injury, thus hampering the therapeutic efficacy of MSCs. To avoid cell-related problems, novel applications of MSC-derived EVs has gained interest in tissue fibrosis treatment in recent years. EVs that are 30–5,000 nm in diameter include exosomes (30–100 nm), microvesicles/ectosomes (100–1,000 nm), and apoptotic bodies (100–5,000 nm). EVs can carry a variety of substances including mRNAs, microRNAs, enzymes, and other bioactive molecules (Abraham and Krasnodembskaya, 2020; Kanai et al., 2021). An EV of endocytosis origin is called an exosome, and an EV that forms from outward budding of the plasma membrane is called a microvesicle or microparticles (Gao and Raj, 2020). MSCs-derived EVs can be administered after isolation, characterization, and purification from the conditioned medium (CM) of cultured MSCs. However, there are limited studies on the effect of MSC-EVs on IPF. Mansouri and colleagues proposed that MSC-EVs can shift pro-inflammatory/classical and non-classical monocytes, as well as the alveolar macrophages, in the lung towards monocyte/macrophage profiles, which may reduce cell apoptosis, reverse PF, and improve the Ashcroft score (Mansouri et al., 2019). The effects of MSC-EVs on lung fibroblasts were investigated by Wan et al. (2020), and the results suggested that MSC-EVs can inhibit proliferation, migration, invasion, and differentiation of fibroblasts through overexpression of miR-29-3p in IPF.

The underlying mechanisms of PF, including telomere shortening and EMT, have also been considered research hotspots. The biological process of IPF has been described as an aberrant repair response to repeated alveolar epithelial injury in genetically susceptible aging individuals. Type II alveolar epithelial cells (AEC2) serve as the primary progenitor/stem cells for alveolar regeneration during tissue repair and homeostasis (Selman and Pardo, 2006; Barkauskas et al., 2013). In IPF, AEC2 display telomere shortening, epigenetic changes, and cellular senescence, which lead to impaired renewal capacity and dysfunction. Telomeres are the protective end caps of eukaryotic chromosomes and determine the proliferative lifespan of somatic cells, functioning as the protectors of cell replication (Celtikci et al., 2021). Genetic studies have confirmed that significant telomere shortening was found in 40% of familial IPF cases and 25% of sporadic IPF cases. Moreover, genetic alterations in telomere-related genes (e.g., TERT, TERC, TINF2, RTEL1, PARN, etc.) have been identified. Regulator of telomere length 1 (RTEL1), which is an essential helicase, has been implicated in sporadic IPF (Wang et al., 2020). Telomere shortening leads to the senescence of alveolar epithelial stem cells, and AEC2 apoptosis leads to a significant reduction in cell number, resulting in reduced alveolar regeneration (Zhang et al., 2021). In addition, senescent AEC2 secrete high levels of interleukins, interferons, tumor necrosis factors, colony-stimulating factors, growth factors, TGF-β, and chemotactic cytokines, which promote the differentiation of fibroblasts into myofibroblasts and ongoing tissue remodeling (Zhang et al., 2022). Therefore, telomere dysfunction and AEC2 cell senescence are the main drivers of IPF (Michalski and Schwartz, 2020). Telomerase gene therapy treatment in wild-type and telomerase-deficient mice has been demonstrated to prevent the development of pulmonary profibrotic lesions (Liu et al., 2019; Piñeiro-Hermida et al., 2020). EMT plays an important role in the development of IPF as well. EMT is a process in which fully differentiated epithelial cells gradually transform into mesenchymal cells and lose their original function after exposure to extracellular factors (Nieto et al., 2016), which is an important mechanism leading to the generation of myofibroblasts. During the EMT process, epithelial cells undergo significant changes in transcriptional regulation, cytoskeleton shape, adhesion, and ECM components (Sacks et al., 2018).