1. Introduction

Hyperglycemia, the hallmark of diabetes, results from a diminished ability of the body to produce or respond to insulin, making it a complex metabolic disorder. The World Health Organization reports that 451 million people globally are currently living with diabetes, a number projected to increase to 693 million by 2045.[1] The escalating burden of diabetes affects every country and its population worldwide. Consequently, extensive research has been conducted on management, control, and treatment strategies for diabetes. Currently, daily insulin injections remain the standard treatment for individuals with Type 1 Diabetes (T1D), late-stage Type 2 Diabetes, and certain rare forms of the disease.[2] Although life-saving, daily insulin injections do not entirely mimic the natural blood glucose regulation of β cells. Additionally, this therapy is costly and burdensome for patients with diabetes and does not entirely eliminate the risk of acute and long-term diabetes complications. The current treatment for insulin-dependent diabetes involves pancreatic IT, which involves isolating islets from donors and infusing them percutaneously into the liver via the portal vein.[3–8] Among the available therapies, IT may be the most promising approach for the clinical treatment of diabetes, particularly TD1 mellitus.[9,10] In recent decades, the number of publications on IT has significantly increased, yet no bibliometric study has been conducted. Therefore, this study aims to provide a comprehensive overview of the IT field’s knowledge structure and research trends using bibliometric analysis.

Bibliometric analysis, which employs diverse statistical methods, serves as a quantitative tool for scrutinizing and appraising knowledge carriers within a specific field. It facilitates the formulation of therapeutic guidelines and aids in understanding research trends, knowledge frameworks, collaborative patterns, and prospective development directions for a given subject.[11,12] This approach is widely used in many disciplines, and in this study, it seeks to summarize the features of IT literature, explore research findings and future directions for this topic, and offer standards for future investigations.[13]

2. Materials and methods 2.1. Database

The Web of Science Core Collection (WoSCC) is acknowledged as a premier and extensive database in scientific research, encompassing a vast array of scholarly literature and studies.[14] In this study, we used WoSCC to conduct an exhaustive literature search.

2.2. Search strategies

On October 10, 2023, we retrieved and downloaded all papers from the WoSCC database published between 2000 and 2023 to control for daily variations in database content. Our search employed the following terms: TOPICS = (Islands of Langerhans Transplantation) OR (Transplantation, Islands of Langerhans) OR (Islands of Pancreas Transplantation) OR (Transplantation, Islands of Pancreas) OR (Transplantation, Islet) OR (Pancreatic Islets Transplantation) OR (Islets Transplantation, Pancreatic) OR (Transplantation, Pancreatic Islets) OR (Islet transplantation) OR (Islet transplantations) OR (Transplantations, Islet) OR (Grafting, Islets of Langerhans) OR (Transplantation, Islets of Langerhans).

2.3. Exclusion and inclusion criteria of publications

A total of 10,348 articles on IT were retrieved from the WoSCC database. We excluded 1864 articles classified as conference abstracts, proceedings, case reports, letters, or off-topic papers. Additionally, 96 non-English articles were excluded. The final selection for this study included original studies, reviews, and meta-analyses pertinent to IT-related literature. The literature exclusion and inclusion process is shown in Figure 1. Ethical approval was not necessary for this study because it was bibliometric and did not include human or animal subjects.

Figure 1.:

Flow diagram of the included papers.

2.4. Data extraction and analysis

To remove duplicates, plain text files from WoSCC were imported into CiteSpace. The unique files were then compiled and transferred to Microsoft Excel to analyze the authors, countries, journals, and institutions associated with the publications.

CiteSpace (version 6.1.R1), developed by Professor Chen C, is a software designed for bibliometric analysis and visualization[15]. In this study, we set a time frame from January 2000 to October 2023, and utilized a single time slice in CiteSpace for keyword clustering, burst word analysis, and burst reference analysis.

VOSviewer (version 1.6.18) is a tool designed for bibliometric analysis, adept at distilling essential information from numerous publications.[16] In our study, VOSviewer was instrumental in conducting co-occurrence analysis across 3 modules: density, overlay, and network visualization: for institutions, authors, countries, and keywords. This facilitated the creation of a visual atlas that illustrated the connections between research hotspots and scientific literature across various dimensions. Within the VOSviewer map, each node represents an entity (such as a nation, organization, publication, or author), with the node size and color reflecting the quantity and type of these entities, respectively. In addition, the thickness of the lines between nodes indicates the degree of relatedness or co-citation of the entities.[17,18]

Utilizing the “bibliometrix” R package (version 3.2.1), we generated a global distribution network of topic-related articles.[19]

GraphPad Prism (8.0.2) was used to tally the number of issued documents and to generate line graphs.

Given the nature of this bibliometric analysis, which exclusively examined publicly available literature and did not involve direct research on humans or animals, ethical approval, and patient consent were not required for this study. This approach aligns with the guidelines and exemptions for studies requiring ethical oversight, as our work did not entail any intervention, interaction with human subjects, or the collection of personal data. Therefore, the necessity for approval from an ethics committee or institutional review board was deemed not required.

3. Results 3.1. Annual publications

This study included a total of 8388 articles. A line graph illustrating the annual publication count is presented in Figure 2. The annual rate of publications in IT research exhibited a fluctuating upward trend, marked by significant peaks in 2004, 2011, and 2018, indicative of heightened research activity in those years. The decline in publications in 2023 can be attributed to the timing of the search, but the overall trend continues to show an upward trajectory. Notably, over the past 5 years, IT has maintained a consistent annual publication output, underscoring its status as a current research hotspot that continues to garner attention.

Figure 2.:

Annual publication volume trend from 2000 to 2023.

3.2. Analysis of leading journals

A total of 1155 academic journals have published papers related to IT. Table 1 displays the top 10 journals along with the number of documents they have published. Leading the list with 608 articles is “Transplantation,” followed by “Cell Transplantation” with 504 articles, “Transplantation Proceedings” with 482 articles, and the “American Journal of Transplantation” with 331 articles. Among these top 10 journals, “Diabetes” has received the highest number of references, totaling 23,402 citations, followed by “Transplantation” with 21,444 citations, “American Journal of Transplantation” with 12,849 citations, and “Cell Transplantation” with 11,392 citations. These figures underscore their significant scholarly impact in the field of information technology.

Table 1 - The top 10 productive and cited journals on the research of IT. Rank Journal Publications Count of citations IF (2022) 1 Transplantation 608 21,444 6.20 2 Cell Transplantation 504 11,392 3.30 3 Transplantation Proceedings 482 5271 0.90 4 American Journal of Transplantation 331 12,849 8.80 5 Diabetes 276 23,402 7.70 6 Xenotransplantation 268 5368 3.90 7 Diabetologia 191 9473 8.20 8 Plos One 174 4848 3.70 9 Biomaterials 138 9278 14.00 10 Islets 113 1397 2.20

IT = islet transplantation.

3.3. Analysis of leading countries, regions, and institutions

The top 10 contributing countries and regions are identified through the analysis of cooperative network maps in IT-related research (Table 2 and Fig. 3). Leading the list with 3535 articles is the United States, followed by China (798), Japan (774), Canada (724), and Italy (484). The size of each circle in the network map represents the quantity of published papers for the respective nation, while the thickness of the connecting lines indicates the level of collaboration. This analysis reveals variations in the extent of IT research conducted across different parts of the world. Notably, the top 5 countries in terms of collaboration intensity include Sweden, Japan, Canada, Italy, and the United States. This suggests a growing trend of cross-border and international research collaboration in the field.

Table 2 - The top 10 productive countries/region on the research of IT. Ranking Countries/Region Publications Link strength 1 USA 3535 151,343 2 China 798 14,303 3 Japan 774 21,531 4 Canada 724 38,976 5 Italy 484 22,629 6 Germany 461 18,223 7 Sweden 445 18,804 8 England 438 15,769 9 South Korea 426 9487 10 Switzerland 304 11,179

IT = islet transplantation.

Figure 3.:

The visualization of countries (A) and geographical distribution (B) on IT.

As depicted in Table 3, the University of Alberta (378 articles), University of Miami (353 articles), and University of Minnesota (299 articles) rank as the top 3 universities actively supporting IT research. Following closely is Harvard University (251 articles). When examining the collaboration connection strength and the institutional network collaboration map (Fig. 4), it becomes evident that there is room for enhancing global cross-institutional collaboration in IT research.

Table 3 - The top 10 productive institutions on the research of IT. Ranking Institutions Publications Link Strength 1 University of Alberta 378 380,456 2 University of Miami 353 313,029 3 University of Minnesota 299 241,551 4 Harvard University 251 195,647 5 Uppsala University 205 155,640 6 University of Pittsburgh 182 151,065 7 Kyoto University 176 153,218 8 Seoul National University 158 114,043 9 University of Pennsylvania 145 128,672 10 Karolinska Institute 134 98,889

IT = islet transplantation.

Figure 4.:

The network map of institutions.

3.4. Analysis of leading authors and co-cited authors

In the author cooperation network’s co-occurrence map (Fig. 5A), distinct clusters are represented by different colored circles, and the lines indicate the intensity of collaboration. This analysis reveals the presence of 5 primary author collaboration groups, demonstrating substantial cooperation among authors. Table 4 highlights the top 5 authors in terms of the number of published documents, including Shapiro (126 articles), Ricordi (121 articles), Matsumoto (85 articles), Noguchi (80 articles), and Naziruddin (79 articles). Figure 5B illustrates co-cited authors, with the most frequently co-cited authors being Shapiro (4808 co-citations), Ricordi (2285 co-citations), Ryan (2219 co-citations), Hering (1556 co-citations), and Matsumoto (1275 co-citations). These authors hold significant academic influence in the field of IT.

Table 4 - The top 10 productive and cited authors on the research of IT. Ranking Author Documents Co-cited author Count 1 Shapiro 126 Shapiro 4808 2 Ricordi 121 Ricordi 2285 3 Matsumoto 85 Ryan 2219 4 Noguchi 80 Hering 1556 5 Naziruddin 79 Matsumoto 1275 6 Ricordi 74 Sutherland 893 7 Piemont 73 Robertson 877 8 Berney 71 Noguchi 876 9 Hering 67 Carlsson 843 10 Korsgren 67 Lakey 797

IT = islet transplantation.

Figure 5.:

Cooperation map of authors.

3.5. Analysis of keywords and burst words

The key concepts within an article can be summarized and extracted through keywords, with increased frequency indicating a greater focus on the relevant subject of study. Therefore, by analyzing keywords, it is possible to identify research hotspots and shifts in emphasis.[20] This study analyzed 112 keywords with a frequency of occurrence exceeding 30. Among the most frequently occurring terms were “IT” (1300 instances), “xenotransplantation” (383 occurrences), “apoptosis” (163 occurrences), “stem cells” (158 occurrences), “immunosuppression” (148 occurrences), and “microencapsulation” (105 occurrences). The research hotspots encompassed fundamental medicine, clinical medicine, and epidemiology, as illustrated in Figure 6A and B. To track research progress over time, the keywords were aggregated to create a timeline map (Fig. 7). The study resulted in the formation of 5 clusters: differentiation, hypoxia, science, T cells, and xenotransplantation. “Burst words” represent terms that experience a significant spike in frequency for a short period, indicating research hotspots. The emergence and disappearance timeframes of burst words provide insights into research trends and directions.[21] In this study, Bursting Words focuses on the basic theory and clinical application of IT (Fig. 8).

Figure 6.:

Analysis of keywords. (A) Keyword density map and (B) co-occurring map of keywords.

Figure 7.:

Keyword timeline chart and keyword clusters.

Figure 8.:

Top 25 keywords with the strongest citation bursts.

3.6. Co-cited references and reference with citation bursts

Co-cited Citations: Over the past 30 years, there have been 154,892 co-cited references related to IT research. Each of the top 10 co-cited references listed in Table 5 has garnered a minimum of 141 co-citations, with one receiving more than 412 co-citations. The co-citation network map, composed of references with 150 or more co-citations, is depicted in Figure 8. Furthermore, Figure 9 highlights the active co-citations among references, including instances where “Shapiro, 2000, N Engl J Med” actively co-cites “Ryan, 2005, Diabetes,” “Ryan, 2001, Diabetes,” and “Ricordi, 1988, Diabetes,” among others.

Table 5 - Top 10 co-cited references on research of IT. Ranking Co-cited reference Citations 1 Shapiro AMJ, 2000, N Engl J Med, v343, p230 2748 2 Shapiro AMJ, 2006, N Engl J Med, v355, p1318 1023 3 Ryan EA, 2005, Diabetes, v54, p2060 942 4 Ryan EA, 2001, Diabetes, v50, p710 571 5 Ricordi C, 1988, Diabetes, v37, p413 551 6 Ryan EA, 2002, Diabetes, v51, p2148 435 7 Barton FB, 2012, Diabetes Care, v35, p1436 412 8 Kevin A D’Amour, 2006, Nat Biotechnol, v24, p1392 388 9 Kroon EJ, 2008, Nat Biotechnol, v26, p443 387 10 Lim FL, 1980, Science, v210, p908 384

IT = islet transplantation.

Figure 9.:

The visualization of co-cited references on research of IT. IT = islet transplantation.

Citation burst references are those that have experienced a substantial surge in citations during a specific period, indicating heightened academic interest in a particular topic. In this study, CiteSpace identified 10 references with significant citation bursts, as illustrated in Figure 10. Each bar in Figure 10 corresponds to a year, with red bars highlighting periods characterized by a remarkable surge in citations.[22] Citation bursts were observed spanning from as early as 2001 to as late as 2013. The reference “IT in Seven Patients with TD1 Mellitus Using a Glucocorticoid-Free Immunosuppressive Regimen” by Rajotte et al displayed the most substantial citation burst, with a strength of 389.08, occurring between 2001 and 2008. The reference “International Trial of the Edmonton Protocol for IT,” published in the New England Journal of Medicine by Shapiro et al, experienced notable citation bursts from 2004 to 2014, with a strength of 207.31, making it the second most impactful burst. These 10 references exhibit burst strengths ranging from 62.20 to 389.08 and have a duration of influence lasting between 6 and 7 years. For further details, please refer to Table 6, which outlines the primary research contents of these ten references, ordered in accordance with the literature in Figure 9.

Table 6 - The main research contents of the 10 references with strong citations bursts. Rank Strength Main research content 1 71.41 Human pancreatic duct tissue can be cultured and guided to develop in vitro into islet tissue capable of responding to glucose. This approach holds the potential to provide a new source of pancreatic islet cells for transplantation.[23] 2 389.08 The infusion of an adequate islet mass, along with glucocorticoid-free immunosuppression, can result in achieving insulin independence and maintaining excellent metabolic control.[24] 3 105.28 This approach is effective in managing labile diabetes and offers protection against undiagnosed hypoglycemia, particularly in carefully selected individuals.[25] 4 91.77 It presents a viable alternative for individuals experiencing severe hypoglycemia or glycemic instability.[26] 5 192.08 It can address issues related to hypoglycemia and glucose instability. The majority of participants maintained C-peptide secretion for up to five years, although many eventually required some insulin supplementation.[27] 6 65.39 Following single-donor, marginal-dose IT, this established transplant strategy effectively prevented hypoglycemia and restored insulin independence in all 8 recipients.[28] 7 207.31 In individuals with TD1 mellitus experiencing unstable control, IT using the Edmonton protocol can effectively restore long-term endogenous insulin production and promote glycemic stability.[29] 8 62.20 The insulin content of insulin-expressing cells generated from human embryonic stem cells closely resembles that of adult islets.[30] 9 79.33 Human embryonic stem cells have the capability to differentiate into insulin-secreting cells that can respond to glucose.[31] 10 135.10 Improvements in the major effectiveness and safety outcomes of IT between 1999 and 2006 for transplant patients vs recipients between 2007 and 2010.[32]

IT = islet transplantation.

Figure 10.:

Top 10 references with strong citation bursts. A red bar indicates high citations in that year.

4. Discussion 4.1. General information

With abnormally elevated blood glucose levels serving as an indicator of the endocrine disorder diabetes, it is projected that by 2045, approximately 693 million individuals worldwide will be affected by this condition,[1] bearing a heavy financial burden.[33] Many diabetic patients have discovered that IT is effective in reducing hyperglycemia and enabling them to achieve insulin independence. However, further research is needed to enhance its long-term success rate.[34] As a result, IT has been a hot topic of research in diabetes treatment for more than 2 decades.

From 2000 to 2003, the annual average of published papers was <300, indicating that IT research was in its early stages with an insufficient foundational base. Subsequently, from 2004 to 2009, the field saw an increase in publications, averaging over 350 papers per year. However, from 2010 to 2013, there was a substantial surge in publications, with an annual average of 400 papers. In the past 5 years, the number of publications related to IT literature has remained relatively stable, suggesting that IT-related fields continue to be a prominent area of research.

The United States and China have emerged as leaders in IT research, with the United States publishing 3535 papers, followed by China with 798, and Japan with 774. Among the top 10 research organizations, the majority are based in the United States (n = 5, 50%). Significant cooperation is notably observed among the United States, Canada, Germany, and China, with Germany also actively collaborating with Italy, Japan, and Sweden. However, despite the existing collaborative relationships among certain nations, the extent and intensity of interinstitutional cooperation are not optimal. For instance, collaborations between institutions in Poland and China are limited, and if this continues, it may hinder the progress of the research field in the future. Therefore, we strongly advocate for research institutions worldwide to engage in broader cooperation and dialogue to collectively advance IT research.

“Transplantation” (Impact Factor = 6.20) is the primary publisher of IT research, establishing itself as the leading journal in this field of study. Among all the journals, “N Engl J Med” boasts the highest impact factor (Impact Factor = 158.50), followed by “JAMA” (Impact Factor = 120.70). A glance at co-cited publications reveals that many of them belong to high-impact Q1 journals, underscoring the pivotal role played by these prestigious global journals in fortifying IT research.

From an authorial standpoint, Shapiro, Ricordi, and Matsumoto stand out as prolific contributors. Shapiro has published 126 papers, including 5 that explored the role of stem cells in IT,[35–39] one of these articles talks about the side effects and risks of islet transplants.[40] Ricordi, author of 121 papers, has affirmed in 1 article that the greater omentum can be used as a site for IT.[41] Matsumoto, with 85 published papers, confirms in 1 piece that improving allogeneic IT by suppressing T helper 17 cell and enhancing Treg with histone deacetylase inhibitors.[42]

Shapiro is the most frequently referenced co-cited author, with 4808 citations, followed by Ryan (22