According to previous bibliometric studies, the Web of Science Core Collection (WoSCC) is the most widely used database, with more than ten thousand high-quality journals (Liu et al., 2022; Zhang et al., 2022a; Zhang et al., 2022b). A database from the WoSCC, the Science Citation Index-Expanded, was selected to conduct this study. The literature search for original articles and reviews was conducted independently by two researchers on October 1, 2022, using the search terms: #1: Topic (TS)= (spasmolytic polypeptide-expressing metaplasia OR trefoil factor 2 OR TFF2); #2: TS= (gastric cancer OR gastric carcinoma OR stomach cancer OR stomach neoplasm*); Source of final search term: #1 AND #2. The study period was from January 1, 2002, to October 1, 2022, with English used as the only language. In total, 260 articles and 32 reviews were obtained ( Figure 1 ). Supplementary File S1 displays the included studies.

A range of key information was extracted from the WoSCC database for further analysis in this study. This series contained information such as the year of publication, the number of citations, countries or regions, research organizations, authors, sources, references and keywords. Then, several repetitive keywords, such as nations, organizations, authors and keywords, were combined into one word, spelling errors were corrected, and irrelevant words were removed. Lastly, the bibliometric analysis was completed by importing the cleaned data into Microsoft Excel 365 and bibliometric visualization platforms.

A nation’s productivity is generally measured by the number of publications, and its impact is measured by the number of average citations. In addition, screening procedure based on Price’s Law ( ∑ m + 1 I n ( x ) = N ), was employed in this study to identify representative scholars and key research forces. In this equation, x denoted the number of publications from each author and n (x) denoted the number of authors who have written x number of publications. Additionally, I equal to nmax represented the number of documents produced by the highest-producing author, N represented the total number of authors, and m represented the minimum number of documents produced by the core author. As a result, m = 0.749 × N m a x ≈ 5.3 was used to define authors with ≥6 documents issued as the core author (Knowlson et al., 2022).

Diagrams of visualizing network and knowledge structure were conducted using VOSviewer, a widely used bibliometric visualization software (van Eck and Waltman, 2010), which provides three main visual maps: the network visualization map, time-overlay visualization map and density visualization map. This study used VOSviewer (Version 1.6.18) to analyze co-authorship between countries, organizations, core authors, influential journals and co-occurrence keywords (Kokol et al., 2018). Furthermore, SCImago Graphica (Version 1.0.26) was primarily used to assess the international geographic collaborations and distribution among the top ten productive countries (Hassan-Montero et al., 2022). Additionally, CiteSpace (Version 6.1.R3), another visualization tool invented by Prof. Chaomei Chen (Synnestvedt et al., 2005), was also used in this study to visualize the co-citation analysis of references and authors and identify the keywords and references with the strongest citation bursts. CiteSpace was also used to create a dual map overlay of journals using parameters: duration (2002–2022), years sliced (3 years), the type of node (reference, cited author, and cited journal), selection criteria (g-index = 25), and pruning methods (Pathfinder and pruning sliced networks). An additional Cytoscape plugin, CytoNCA, was used to conduct centrality analyses among the productive countries, organizations and authors of SPEM-GC.

To further review the key pathways and targets related to the regulation of SPEM transformation, GO classification and KEGG pathway (http://www.omicsbean.cn/) methods were employed to identify genes’ functional categories and predict their biological functions. Figures were generated with R using KEGG pathway enrichment analysis and GO enrichment analysis with a corrected p-value < 0.05.

A total of 292 SPEM-GC papers written by 1757 authors at 412 organizations from 41 countries were collected. They were published in 131 journals, cited by 1395 journals and referenced in 8431 articles. The total number of citations was 9345, and the average number of citations per publication was 35.76. The H-index of all the documents was 53, indicating a high academic value and social impact for papers in this field. Based on Figure 2A , we observed that the number of documents issued each year from 2002 to 2012 was relatively stable, except for two small peaks in 2017 and 2020. In contrast, the annual citation curve increased steadily since 2002 and peaked in 2021 when 11,816 citations were recorded.

To determine which countries had the most prominent contribution and cooperation in the field of SPEM-GC, we made a visual analysis of the number of documents issued, citations, and co-occurrence frequencies for 41 countries and regions. The results showed that most SPEM-GC studies were from North America, Europe and East Asia. The number of publications of the top 10 productive countries/regions continued to rise rapidly ( Figure 2B ). As shown in Table 1 , the top ten productive countries/regions had the highest number of publications. The United States published 125 papers in this field, accounting for 40.7% of all papers. Although China ranked second in terms of publications, it had the lowest average citation among the top 10 productive countries.

Co-authorship network maps of the top 10 productive countries were generated using VOSviewer and SCImago ( Figure 2C ). The results showed that the United States was a leading cooperation center in this field, with close ties to Japan, South Korea and China. According to Table 2 , the United States was the most central country (0.47), with Germany and the United Kingdom ranking second (0.35) and third (0.32), respectively. However, a visualization of the number and the year of issuance using VOSviewer indicated that despite European countries such as Finland and France being the first countries to publish the first research in the field, since 2014, Asian countries such as China, South Korea and Singapore have gradually become the main research centers in this field ( Figure 2D ).

Based on Price’s Law, 69.7% of the total articles in this field were published by 18 core authors. More than 50% of the documents issued by core authors defined by Price’s law were evaluated, and the results suggested that the SPEM-GC research field was relatively stable in terms of authorship cooperatives. Table 3 shows the top ten authors, organizations and countries with the most recently published papers in this field. Among them, seven of the top ten authors were from the United States, two were from South Korea, and one was from Japan. In terms of high productivity, the top three authors with the most papers were Americans. They are Goldenring, James R. Vanderbilt University (N = 51, APC = 61.9), Mills, Jason C. Washington University (N = 21, APC = 49.7), and Wang, Tc, Columbia University (N = 19, APC = 90.5). With a centrality of 0.18, Goldenring Jr. was the most cited co-author. Figure 3A shows the authors with a minimum of 6 publications. Based on overlay visualization of co-authorships ( Figure 3B ), the blue cluster is considered a pioneering group for SPEM-GC research. Comparatively, the yellow and Laurel-green cluster authors have published papers in recent years. Upon further visual analysis of the co-cited network, Goldenring James R. authored most of the cited articles along with Mills Jason C., Nam Ki Taek and Petersen Cheritine P. ( Figure 3C ). In terms of burst monitoring of authors ( Figure 3D ), Wang Tc, Lee Hyuk-Joon and Kaminishi M were the top three ranked institutions between 2002 and 2009, followed by Nam and Ki Taek bursting between 2010 and 2012. Presently, the bursting of Choi Eunyoung from 2016 to 2022 and Mills Jason C. indicates their dominance in this field.

We found that articles on SPEM-GC were published in 131 journals. As shown in Figure 4A , 135 documents were published in the top 14 journals (>5 documents in counts), accounting for 44% of the articles included. A total of 31 articles were published in Gastroenterology (IF 2021 = 33.9), followed by the American Journal of Pathology (IF 2021 = 5.8) and Gut (IF 2021 = 31.8). Figure 4B illustrates the number of citations, in which a deeper orange color indicates more co-citations. As shown in Table 4 , most of the top ten JCI district journals with the highest average citation rate in this field were renowned journals over the past 20 years, among which 9 had an H-index over 50. The top journals in gastroenterology were Gastroenterology and Gut, Cancer Research in oncology, and Journal of Pathology in pathology. Among them, the United States and the United Kingdom had four journals each, while Japan and Switzerland had one each.

Using a dual map overlay of relevant journals, we visualized the journal’s citation relationships within related fields ( Figure 4C ). In the map, the labels indicate the journal’s field. Its left side represents cited literature, and its right side represents cited literature. We determined the causal relationship between citations by determining the citation path. Citations made up the applied research and the research basis in this field. The colors signify different citation paths, with orange representing one citation path and green representing another. The orange line indicates that the included articles were mostly associated with Molecular, Biology, and Immunology disciplines. Based on the green path, the articles included in the analysis were found to be mostly distributed in Medicine, Medical and Clinical fields, while their cited papers were mostly distributed in Molecular, Biology, and Genetics.

Reference analysis was conducted, and VOSviewer and CiteSpace were utilized to visualize the references that supported the development of in-depth studies. We first performed a coupled network analysis of 292 articles using the VOSviewer software and generated a visual map based on the top 50 citing references with the strongest association strength ( Figure 5A ). Based on coupling strength, these references were classified into four clusters: Cluster 1 (red dominates), consisting of 16 highly citing references with the theme of “Cellular Reprogramming and Regeneration of SPEM”; Cluster 2 (green dominates), consisting of 13 highly citing references with the theme of “The origin of SPEM”; Cluster 3 (blue dominates) consisting of 11 highly citing references, with the theme of “Oxyntic Atrophy, IM and SEPM”, and; Cluster 4 (yellow dominates) comprising a total of 10 highly citing references, with the theme of “Trefoil Factor Family Peptides”. According to the time-overlay visualization map in Figure 5B , the references in cluster 1 published by Goldenring (2022) (Goldenring and Mills, 2022), Lee (2022) (Lee et al., 2021), Jeong (2021) (Jeong et al., 2021), Bockerstett (2020) (Bockerstett et al., 2020) and Burclaff (2020) (Burclaff et al., 2020), as well as references in cluster 4 published by Hoffmann (2020 and 2022) (Hoffmann, 2012; Hoffmann, 2020), represent the current frontiers of research. A complete list of the top ten citing references for SPEM-GC is shown in Table 5 .

Furthermore, a total of 8431 references were cited at least 23 times based on Price’s law. According to Table 6 , the ten most co-cited references were cited at least 124 times. One of the most co-cited references was an article by Houghton J et al. published in Science in 2004 (n=903). Nine of the top 10 articles were research papers, and one was a Review. As shown in Figure 5C , the reference timeline visualized the evolution of research hotspots over time, and cluster labels were created based on the terms associated with each cluster’s highest frequency. We observed that: cluster #2 (Trefoil factor1), #4 (Gene regulation), #5 (Spasmolytic polypeptide) and #12 (the TTF peptides) started earlier, while cluster #0 (Stem cells), #8 (Differentiation) and #14 (progenitor) could be considered the frontier since they are still ongoing.

Citation bursts are references with a significant increase in citations over time. Figure 5D shows the top 25 citation bursts of the 71 detected. The burst with the strongest strength (strength=10.86) was entitled “A molecular signature of gastric metaplasia arising in response to acute parietal cell loss” (Nozaki et al., 2008), published in Gastroenterology by Koji Nozaki et al. in 2008, with citation bursts between 2010 and 2013. Among the 25 references, 10 (40%) were published in 2017-2022, indicating their importance in this field. Notably, 8 (32%) of these 25 papers were still experiencing citation bursts by the time of writing this manuscript. In light of all these factors, we hypothesized that SPEM-GC would continue to attract attention in the future. Additionally, we also found that 5 papers dealt with the proliferation and lineage conversion of SPEM cells (Engevik et al., 2016; Leushacke et al., 2017; Burclaff et al., 2017; Goldenring, 2018; Radyk et al., 2018; Willet et al., 2018) and 3 dealt with mucosal immune regulation (Nozaki et al., 2008; Petersen et al., 2017; Petersen et al., 2018), indicating that SPEM might significantly influence the research field of gastric carcinogenesis by affecting the above mechanisms.

A total of 1,394 keywords were extracted from the titles and abstracts of the included papers. Using VOSviewer, 65 keywords appeared more than nine times and were used to generate the visual map in Figure 6A . We conducted cluster analysis on high-frequency keywords and divided them into three clusters: Cluster 1 (red dominates), comprising 32 core keywords and was the largest, with the most prevalent keywords being gastric cancer (119 times), expression (79 times), and trefoil peptides (72 times). Cluster 2 (green dominates), comprising 20 core keywords, among which the most common were H. pylori (121 times), SPEM (113 times), and gastric (102 times). Cluster 3 (blue dominates), comprising 13 core keywords, with cancer (79 times), inflammation (34 times) and differentiation (30 times) being the most common. Figure 6B shows the link strength keywords displayed as a density map, in which the colors indicate the total strength of the links. H. pylori (748), gastric cancer (724) and SPEM (706) were the three keywords with the strongest association. Figure 6C shows the top 11 high-frequency keywords over time for each cluster. Six out of 11 clusters are still underway. The largest cluster was spasmolytic polypeptide-expressing metaplasia (#0), followed by oxyntic atrophy (#1), Mongolian gerbils (#4), cellular differentiation (#6), activation (#7) and trefoil factor 1 (#11). As a next step, we analyzed the relationship between keywords, the identified themes and three clusters, which is detailed in Table 7 (Kokol et al., 2022). There were three components to this: expression of gastric cancer genes, H. pylori infection and SPEM, as well as cell differentiation and proliferation.

According to the time-overlay visualization map in Figure 6D , “chief cells”, “progenitor cells”, “progression”, “biomarker” and “atrophy” were the recent keywords in addition to keyword trends. Essentially, these keywords seem to outline the current frontiers of research. Additionally, CiteSpace was used to identify the top 20 keywords with the strongest citation bursts ( Figure 6E ). As indicated by the red part, the keywords indicate a blowout at this stage. The keywords related to “ps2” or “trefoil factor 1” bursts were the strongest (strength = 7.29), followed by intestinal “trefoil factor” (strength = 4.82) and “epidermal growth factor” (strength = 3.46). We also found that the keywords were still emerging in 2022. Thus, the research areas of “chief cell”, “inflammation”, “stem cell”, “atrophy” and “biomarker” might become hot spots in the future.

To further clarify the focus of mechanistic studies in SPEM-GC-related research fields, we extracted and analyzed the keywords of all associated genes in literature, based on which 147 associated gene keywords were identified. Supplementary File S2 lists the relevant statistical results. Among them, the most frequently associated gene was TFF1 (54 times), followed by TFF2 (Petersen et al., 2018) and PSEN2 (Nozaki et al., 2008). Figure 7A lists the keywords having a frequency >5. KEGG pathway enrichment analysis showed that PI3K-Akt, JAK-STAT, HIF-1, MAPK, Hippo, Wnt, VEGF, cell cycle and Ras signaling pathways were the hot research pathways in the field of SPEM-GC. Figure 7B reveals that the immune regulation of Th17 cell differentiation, cytokine-cytokine receptor interaction, and differentiation of Th1 and Th2 cells were also major topics in this field. Further analysis of GO functional annotation and significant enrichment analysis in Figure 7C revealed that Biological Process (BP) was predominantly represented in cell proliferation, single-multicellular organism process, and multicellular organismal process. We also found that Cellular Component (CC) was mainly found in the extracellular region, whereas Molecular Function (MF) was mainly found in identical protein binding and receptor binding sites.

Although SPEM was first reported in 1999 by Schmidts et al. (Schmidt et al., 1999), it was not until 2002 that Halldórsdóttir et al. discovered a link between SPEM and GC (Halldorsdottir et al., 2003). Analysis of the WoSCC database from 2002 to 2022 shows that 292 papers were published in 131 journals by 1757 authors in 412 organizations from 41 nations. Despite relatively stable document volumes, the increasing number of citations indicated that SPEM-GC is becoming increasingly popular. Additionally, SPEM-GC research has steadily grown in the last two decades. According to Table 2 , the United States contributed to the most number of publications on SPEM-GC among the top ten productive institutions, of which eight were from the United States, one from Japan and one from South Korea. The United States maintained its dominant position in SPEM-GC research with a centrality of 0.47. We also found that Germany, the United States, the United Kingdom and China played key roles in the global collaboration of SPEM-GC research. China was second to the United States as the largest issuer in this field, with a total of 71 issues. Despite the low average citation rate of Chinese scholars (van Eck et al., 2010), compared to the United States (Lei et al., 2013) and Japan (Katz et al., 2005), they have gradually become a central region for research over the past few years ( Figure 2D ).

As one of the top 10 and co-cited authors, James R. Goldenring published the greatest number of SPEM-related papers, indicating his prominent contribution to the field ( Table 3 and Figure 3A ). The origin of precancerous lesions in the stomach is the primary focus of Dr. Goldenring (Vanderbilt University). Over the past decade, his team has produced paradigm-shifting data proving that precancerous metaplasia does not arise from altered stomach stem cells but develops from the transdifferentiation of protein-secreting chief cells into metaplastic mucous-producing cells (Goldenring et al; Nam et al., 2010; Mills and Goldenring, 2017). They also studied how immune cell populations (M2-macrophages and type II innate lymphoid cells (ILC2s)) modulated the progression of precancerous lesions from metaplasia to growth and proliferative activity (Meyer et al., 2020). Mills Jason C., a Baylor College of Medicine professor, was identified as one of the most productive SPEM-GC authors and ranked in the top 10 and co-cited authors. His research focuses on multiple regulated mechanisms involved in palingenesis during metaplasia and cancer of the gastric lining. In 2022, James R. Goldenring and Jason C. Mills (Goldenring and Mills, 2022). published an important review summarizing SPEM cell regulation mechanisms and pathophysiology in Gastroenterology. According to their study, gastric glands with mixed incomplete intestinal metaplasia and proliferative SPEM might be at greater risk for cancer and dysplasia. In scenarios with chronic pyloric and intestinal metaplasia, targeting early metaplastic lineages could be an effective approach. According to the journal analysis ( Figure 4 and Table 4 ), Gastroenterology published the most SPEM-GC studies, as well as the most cited ones, including 7 of the top 10 highly cited articles (Judd et al., 2004; Katz et al., 2005; Nozaki et al., 2008; Nam et al., 2010; Lei et al., 2013) ( Table 6 ). Among the top 3 published journals and top 5 co-cited journals, the American Journal of Pathology and Gut played an essential role in research on SPEM-GC. These journals mainly focus on studies from the Molecular, Biology and Genetics fields. Altogether, these results were consistent with dual-map analysis, which revealed that SPEM-GC research is currently strongly focused on molecular biology and immunology.

In some instances, the knowledge base was represented partly by co-cited references cited by scholars involved in related research. There were four references in the top ten co-cited references related to proliferation, differentiation and precancerous changes within the stomach (Houghton et al., 2004; Katz et al., 2005; Nozaki et al., 2008; Nam et al., 2010), two that elaborated on trefoil factor1 or spasmolytic polypeptides (Bossenmeyer-Pourie et al., 2002; Tebbutt et al., 2002), two on gene regulations (Akiyama et al., 2003; Lei et al., 2013), one on immunoregulation (Judd et al., 2004), and one on the molecular mechanisms of gastric epithelial stem cells (Mills and Shivdasani, 2011). Eight references were still in the burst phase and deserve more attention, among which 5 papers focused on SPEM cell proliferation and lineage transformation (Burclaff et al., 2017; Engevik et al., 2016; Goldenring, 2018; Leushacke et al., 2017; Radyk et al., 2018; Willet et al., 2018) and 3 were on mucosal immune regulation (Nozaki et al., 2008; Petersen et al., 2017; Petersen et al., 2018), indicating that SPEM had a significant impact on the research field of gastric carcinogenesis.

For researchers in this age of information explosion and technology, it is essential to keep up with the latest trends in the research field. In bibliometrics, keyword co-occurrences can reflect the focus in specific areas, overlay and timeline views can illustrate the evolution of new hotspots, and the emerging topics in the discipline can be identified through reference clusters and citation bursts (Liu et al., 2022; Li et al., 2022; Zhang et al., 2022a; Yang et al., 2022). As part of this study, we examined reference timeline and burst ( Figures 5C, D ), keyword overlay, co-occurrence, timeline, burst ( Figures 6A–C ), KEGG ( Figure 7B ), and GO annotations ( Figure 7C ) to evaluate the hotspots and frontiers of SPEM-GC research. The 6 hotspots and frontiers of SPEM-GC are discussed below.

Overall, this present study represents the first bibliometric analysis of SPEM-GC-related publications in the past two decades. The presented type of analysis offers a fresh and objective perspective on evolving research topics and trends, which cannot be obtained through traditional reviews. As part of the research, a multidimensional analysis was conducted using various bibliometric software tools to provide more comprehensive results for readers. As a comprehensive guide for future developments in GC research, this study also provides scholars with an overview on SPEM-GC research and an objective guide for the public to understand the significance of SPEM. However, there were some limitations in this study. First, as only WoSCC data were used, it is possible that some relevant studies in PubMed, Scopus and other databases were excluded, resulting in incomplete data collection. However, it should be noted that WoSCC indexes the largest number of articles, ensuring source integrity. Additionally, high-quality articles published in other languages were not considered due to the focus on English-only articles. Lastly, due to methodological limitations of the overall literature quality evaluation system, some newly published high-quality documents might not have been included in the bibliometrics analysis due to their low citations and centrality.

We report the first bibliometric analysis summarizing the knowledge map of research between SPEM and gastric cancer in the last two decades and the potential future research hotspots. We found that the United States had the most high-quality publications and the greatest international cooperation and communications. However, worldwide collaboration among organizations needs to be improved. Additionally, although SPEM could be identified through multiple biomolecular markers as a possible source of GC, there is still a lack of understanding on how SPEM contributes to the early detection and treatment of cancer, indicating a potential research area and discoveries in the future. The findings of this study can be used as a guide for choosing research topics, defining the most promising research frontiers, and selecting appropriate journals for publication. Additionally, we provided information to clinicians and practitioners on new approaches and technologies that might enhance the treatment of H. pylori infection-related gastric mucosal diseases and benefit populations at high risk of developing GC.