The aims of this study were as follows: 1) analyze the basic situation and main contributors (authors, journals, institutions, and countries/regions) regarding the application of stem cells for glaucoma; 2) evaluate the cooperation relationship in this domain; 3) map the knowledge network, analyze the Frontier contents, and predict the future directions.

Web of Science (https://login.webofknowledge.com/) was used to perform the search for publications. The search and data acquisition were completed on 9 December 2022. All publications were exported in the format of “Full Record and Cited References” as plain text files. The search formula was as follows [TS= (“stem cell*” OR “progenitor stem cell*”) AND TS= (“glaucoma”)] AND DOP= [(1999-01-01/2022-12-09) AND DT= (Article OR Review)] AND LA= (English). Two investigators (YT and QZ) retrieved and screened the publications. Disagreements were discussed with a third investigator (MM) until a consensus was reached.

1) Web of Science Core Collection (WoSCC)

1) Meeting abstract, editorial materials, corrections, retractions, and book chapters

We mainly used the following software: VOSviewer (version 1.6.11, Leiden University, Leiden, Netherlands), CiteSpace (version 6.1. R1, Chaomei Chen, Drexel University, Philadelphia, PA, United States), Microsoft Excel (Redmond 2010, WA, United States). The data analysis using these applications were exported to a table that summarized the bibliometric parameters, including publication counts, publication years, total citations, average citations, co-citations, titles, countries, authors, journals, keywords, and references. The maps of co-occurrence and cooperation among authors, journals, institutions, and countries/regions, were made by VOSviewer. In the VOSviewer map, the side of the node corresponded with weighting. Links between nodes indicated cooperative relationships or common appearances in the same literature, and link thickness positively correlated with link strength. In the network visualization module, each color represented a different cluster. In the overlay visualization module, node color represented the average publication year. Blue represented the early research phase, yellow represented the recent period. We used CiteSpace to generate a keyword burst map and identify references. The red node highlighted highly cited references that co-occurred in multiple publications, and darker and lighter shades, respectively, indicated early and recent publications. In the burst module, the keywords and references were sorted by the beginning year of the burst.

We retrieved 572 publications from the WoSCC database, with publication dates from 1999 through 2022. Then, we limited the search results to English-language original articles and review articles. Eventually, 518 articles were brought into this analysis. There were 363 (70.08%) original articles and 155 (29.92%) review articles (Figure 1). From 1999 through 2008, the number of publications increased slowly, but there were fewer than 10 publications per year. From 2009 through 2016, the field received more attention, and the number of publications increased significantly. From 2017 through 2021, more than 50 articles in this field were published each year. Although the articles we obtained were only published before 9 December 2022, it can still be seen that the number of articles published in this field has decreased significantly in 2022. The peak publication frequency was in 2021 (n = 71) (Figure 2A). The general number of citations was 11069, with 21.37 citations for each paper. Overall 50 countries/regions, 2380 authors, 216 journals, and 712 institutions were represented. The United States, China, and England were the main contributors (Figure 2B).

Most of the top 10 prolific authors in this field were from the United States (Table 1), and other authors were from China and the United Kingdom. They published 104 articles, which accounted for 20.08% of all included publications. Yiqin Du, from the University of Pittsburgh, published the most articles (n = 16) followed by Keith R. Martin (n = 14) from the University of Cambridge. We also analyzed and ranked the number of co-citations between authors. Of the top 10 co-cited authors, most were from the United States. In terms of co-citations, Harry A. Quigley, Thomas V. Johnson, and Ben Mead were ranked in the top 3. It is worth noting that Johnson ranked highly in terms of citations and co-citations among the top 10 prolific authors and top 10 co-cited authors. Supplementary Figure S1 depicts the cooperative network between authors with more than five publications included in this analysis.

We analyzed the top 10 prolific and co-cited journals separately using VOSviewer (Table 2). A total of 216 journals published relevant articles, and 24 journals published more than five articles included in the analysis. Sorted by the number of papers, the total number of publications in the top 10 productive journals was 149 (28.76%). Investigative Ophthalmology and Visual Science was the most represented journal (n = 39; 1743 citations); Experimental Eye Research (n = 21; 483 citations) and Cornea (n = 14; 294 citations) ranked second and third, respectively. Investigative Ophthalmology and Visual Science (3690 co-citations) and Experimental Eye Research (1416 co-citations) also were the top 2 co-cited journals; Proceedings of the National Academy of Sciences of the United States of America (891 co-citations) ranked third in this regard.

There were 712 organizations, of which 35 organizations published no less than five articles. We listed the top 10 prolific institutions (Figure 3A), eight of which were in the United States, and two were in China. Central South University ranked first (n = 21; 332 total citations; 15.8 average citations); University of Pittsburgh (n = 19; 413 total citations; 21.7 average citations) and the National Institutes of Health (NIH) National Eye Institute (n = 17; 1186 total citations; 69.8 average citations) followed closely behind. We generated a cooperative relationship map (Figure 3B). The University of Pittsburgh cooperated most frequently with other institutions, and it collaborated most closely with Central South University. In the overlay map, we can view the duration (in the year) each institution has published in the field (Figure 3C).

A total of 50 countries/regions were active in this area. There were 24 countries/regions with at least five publications included in the analysis. Figure 4A shows the top 10 countries/regions in terms of the total number of included publications. The top three countries/regions were the United States (n = 251; 6323 total citations; 25.2 average citations), China (n = 116; 1520 total citations; 13.1 average citations), and England (n = 44; 1953 total citations; 44.4 average citations). The total citations among the top three were markedly higher than the others. The cooperation network of these countries/regions is shown in Figure 4B. The scope of collaboration between the United States and other countries/regions was extensive, its key partners were China, England, and Germany.

High-frequency keywords can reflect some knowledge domains’ evolving research frontiers (Liao et al., 2018). We identified 1157 keywords using VOSviewer, and 47 keywords appeared no less than 5 times. After excluding one word that was not associated with the keywords, we subsequently analyzed the remaining 46 keywords. There were 8 clusters (Figure 5A). “Glaucoma”, “retinal ganglion cell”, “stem cell”, “retina”, and “neuroprotection” were the top 5 keywords sorted by frequency of occurrence. In our overlay visualization map (Figure 5B), blue represents the early research phase and yellow represents the recent period. “Transplantation”, “retinal degeneration”, “neurotrophic factor”, “gene therapy”, and “apoptosis” received more attention around 2014. “Exosomes”, “extracellular vesicles”, “growth factors”, “mitochondria”, and “oxidative stress” have become the foci in recent years. CiteSpace allowed us to identify hotspots and research frontiers over time by detecting burst keywords (Chen et al., 2006). The minimum burst duration was set to 1, the red bars in Figure 5C indicate that some words have been cited frequently, while the blue bars indicate words that have been cited infrequently. The diagram exhibits the top 20 keywords with the strongest citation bursts, of which “embryonic stem cells” had the longest duration. “Model” had the highest burst strength. In general, the focus gradually deepened from “stem cell” and “experimental model” of disease to “extracellular matrix”, “retina”, and “trabecular meshwork”.

Table 3 lists the top 10 cited publications. The most cited article was written by Cuenca et al. (2014), published in Progress in Retinal and Eye Research with 259 citations, titled “Cellular responses following retinal injuries and therapeutic approaches for neurodegenerative diseases”. In the article, the authors held that the injury response of retinal neurodegenerative diseases at the cellular and molecular levels is similar. Inflammatory response, oxidative stress and apoptosis may be common features of these diseases, eventually leading to cell death and retinal remodeling. In addition, the authors also summarized the molecular, anatomical and functional changes caused by damage, and the treatment were summarized. This article aimed to help researchers to establish suitable treatment methods for these pathologies (Cuenca et al., 2014). We have obtained key references with high citations in this field using CiteSpace (Figure 6A). The CiteSpace citation burst could identify references focused on by researchers during a specific period (Chen et al., 2006). There were 15 references with the strongest citation bursts when the burst duration was set to 5 years (Figure 6B). “Neuroprotective Effects of Intravitreal Mesenchymal Stem Cell Transplantation in Experimental Glaucoma” by Johnson, had the highest burst strength (n = 12.58) (Johnson et al., 2010). There were four articles with citation bursts ending in 2022 that indicated they had received more attention in recent years (Ohlemacher et al., 2016; Venugopalan et al., 2016; Zhu et al., 2016; Teotia et al., 2017).

Researchers focused on the simple differentiation and replacement of cells before 2009. The types of cells emphasized during this period are also relatively complex, including various stem cells and cells with similar activity (such as Müller cells and TM cells) (Lawrence et al., 2007; Bull et al., 2008; Kelley et al., 2009). Some studies have shown that when human embryonic stem cell-derived neural progenitor cells (NPCs) or retinal cells are transplanted into the subretinal space, they survive for some time, and a few cells could integrate into the retina where they could further differentiate into retinal cells (Banin et al., 2006; Meyer et al., 2006; Lamba et al., 2009). Conditions for inducing embryonic stem cells from mice, monkeys, and humans to differentiate into rod and cone cells have been investigated (Osakada et al., 2008). Another type of cell is bone marrow mesenchymal stem cells, scholars have confirmed that by transplanting this cell, it could be differentiated and partially integrate into the RGC layer to reduce retinal damage (Yu et al., 2006; Inoue et al., 2007; Li et al., 2009). Moreover, other studies have used Müller cells and oligodendrocyte precursor cells (Lawrence et al., 2007; Bull et al., 2008; Bull et al., 2009; Kelley et al., 2009).

From 2009 to 2013, there was growing concern that the effects of stem cell transplantation on the retina may be related to its secretion of neurotrophic factors. Pease et al. (2009) injected an adeno-associated viral vector that contained ciliary-derived neurotrophic factor (CNTF) and brain-derived neurotrophic factor (BNTF) directly into the vitreous body of glaucomatous rats. They found that CNTF had an obvious neuroprotective effect, emphasizing that the action of neurotrophic factors on damaged RGCs are determined by the delivery method, dose, and model of disease (Pease et al., 2009). Several studies have confirmed that bone marrow or dental pulp mesenchymal stem cells, in addition to integration, and differentiation, provide neurotrophic factors to the damaged retina (Johnson et al., 2010; Levkovitch-Verbin et al., 2010; Harper et al., 2011; Johnson and Martin, 2013; Manuguerra-Gagne et al., 2013; Mead et al., 2013; Johnson et al., 2014). Cell therapy can supply more durable, low-dose, multi-factor better protection compared with direct supplementation of nutritional factors (Eiraku et al., 2011; Nakano et al., 2012). Also, the release of neurotrophic factors in cell therapy is based on the host’s retinal condition with less trauma, and it can be engineered to allow donor cells to express specific factors. Concerns have also been raised over the details of how to differentiate stem cells into RGCs (Chen et al., 2010; Meyer et al., 2011; Tucker et al., 2011). During this period, attention began to be paid to the role of induced pluripotent stem cells (iPSCs) and differentiated stem cells in the trabecular meshwork in improving IOP (Kelley et al., 2009; Du et al., 2012; Du et al., 2013).

Between 2014 and 2016, the induction of human pluripotent stem cells became a research hotspot. This was seen as a way to avoid both the ethical issues caused by embryonic stem cells and the host immune rejection. Researchers induced pluripotency in cells and used them to refill damaged areas (Ding et al., 2014; Zhong et al., 2014; Abu-Hassan et al., 2015; Zhu et al., 2016). They also explored the molecular details of iPSCs differentiation and identified differentiated cells (Riazifar et al., 2014; Ohlemacher et al., 2016), and then people gradually focused on targeted induction of IPSC to differentiate into cells with specific functions, such as using CRISPR technology (Sluch et al., 2015; Tanaka et al., 2015; Teotia et al., 2017). Additionally, there are still some researchers continuing to explore the different sources of mesenchymal stem cells (MSCs) and their secretion of nutritional factors on the role of retinal (Mead et al., 2014; Mesentier-Louro et al., 2014; Roubeix et al., 2015).

There was a growing interest in a cell-free treatment approach from 2017 to 2022 (Phan et al., 2018; Rahmani et al., 2020; Han et al., 2021). Exosomes are vesicle structures containing a variety of ribonucleic acids (RNAs) and proteins with diameters of 40–100 nm (Hobor et al., 2018; Cheng et al., 2020b). Mead and Tomarev (2017) have made an outstanding contribution to stem cell therapy for glaucoma. They transported exosomes extracted from bone marrow mesenchymal stem cells (BMSCs) into the eyes, which were found to accelerate the regeneration of RGCs and their axons, and this therapeutic effect was weakened after knocking out Argonaute2, which is a kind of microRNA (miR) for the first time (Mead and Tomarev, 2017). They demonstrated for the first time that BMSC-derived exosomes may exert a protective action on RGCs through miRNAs (Mead and Tomarev, 2017). Exosomes effectively avoid a variety of complications in cell therapy, such as immune rejection, tumorigenesis, and difficulties in integrating with host tissue. Exosomes have become a potential resource for cell and gene therapy due to their small size, easy storage, and non-proliferation (Lu et al., 2019; Feng et al., 2021). Mesenchymal cells can produce more exosomes than other cells (Yeo et al., 2013; Phan et al., 2018).

Because of the remarkable advantages of cell-free therapy, may be the reason for the significant decrease in the research on stem cell therapy for glaucoma in 2022, as the shifts in a more optimized direction. Increasing attention to the use of extracellular vesicles as therapeutic agents or drug carriers (Anand et al., 2022; Sanie-Jahromi et al., 2022). The role of exosome content gradually emerges. Yu et al. (2022) found that tumor necrosis factor-α mediates the neuroprotective effect of miR-21-5p-rich exosomes in glaucoma through the maternally expressed gene 3/programmed cell death four axis. Studies have shown that miR-143-3p, miR-125b-5p, and miR-1260b in aqueous humor may be therapeutic intervention targets, and microRNA can be an early biomarker or therapeutic strategy by regulating its level (Martinez and Peplow, 2022). RGCs are a group of neurons located at the distal end of the eye, their axons contain large amounts of mitochondria (Anand et al., 2022). There are studies found that mitochondrial dysfunction or abnormal mitophagy is involved in glaucomatous damage and affects neurogenesis and neuronal viability (Rosignol et al., 2020; Ozgen et al., 2022). At the same time, some researchers were paying increasing attention to the genetic aspects of the field. Wei et al. (2022) discovered key genes that regulate self-renewal and differentiation of NPCs by using 5-(p-hydroxyphenyl)-1,2-dithione-3-thione to treat mammalian NPCs to promote their differentiation into neurons and oligodendrocytes and found that mRNA and protein expression of key genes regulating NPCs self-renewal and differentiation decreased, such as β-catenin. Optineurin (E50K) mutant astrocytes caused neurodegeneration in healthy RGCs (Gomes et al., 2022). Xue et al. (2022) investigated the common genetic causes of age-related macular degeneration, diabetic retinopathy, glaucoma, retinal detachment, and myopia, they found that genes associated with common genetic sites were involved in neuronal differentiation and eye development systems. Single-cell RNA sequencing data showed that their gene expression from pluripotent progenitor cells into retinal cells increased during retinal development (Xue et al., 2022).

Notably, iPSCs still received attention in 2022 (Coulon et al., 2022). The trabecular meshwork cells can mediate the regulation of intraocular pressure by regulating aqueous humor flow (Zhu et al., 2017). Wang et al. (2022) designed a nanoparticle to label trabecular meshwork cells, and the use of nanoparticles not only facilitated long-term tracking, but also improved the transmission accuracy of transplanted cells in vivo experiments (Wang et al., 2022). It is more important that the use of magnet temporarily enhanced the effectiveness of cell-based therapy in glaucoma-related pathology, and promoted the clinical transformation of stem cell-based glaucoma therapy (Zhu et al., 2016; Wang et al., 2022).

A total of 47 keywords with no less than five occurrences were identified and divided into 8 clusters. The top 3 clusters are red (13 items), green (7 items), and blue (7 items) in Figure 5A. It is worth noting that some studies have focused on the neuroprotective role of mesenchymal stem cells in damaged RGCs through extracellular vesicles (exosomes, microscopic vesicles) (Mead et al., 2018; Mathew et al., 2019; Mead et al., 2020; Mathew et al., 2021), which may also involve apoptosis and glial cell-mediated inflammatory responses (Alghamdi et al., 2020; Yu et al., 2021). According to the analysis in the overlay visualization module, research on exosomes, extracellular vesicles, and oxidative stress in this field have been a hot topic in the last few years (Mathew et al., 2021; Reboussin et al., 2022; Yu et al., 2022). The following three themes may be worthy of further study in the future: 1) the neuroprotective effect of mesenchymal stem cell-derived exosomes in RGCs, 2) growth factors, MSCs, and iPSCs in regenerative medicine for glaucoma, and 3) oxidative stress involving in the injury mechanism of glaucoma.

First, because the visualization software has limitations on the format of the data, we only collected text data from the WoSCC. However, as the most widely used database in the world, the WoSCC met our analytical demands (Shi et al., 2022). Second, we adopted English publications and limited their types, and we only analyzed original articles and reviews. Third, the indicators related to citations will change over time. We should integrate various indexes to analyze Journal Impact Factor, Source Normalized Impact per Paper, and SCImago Journal Rank (Deng et al., 2020). Fourth, updating the database may affect our results, but this effect should be slight.