Global research trends and hotspots in prognostic prediction models for pancreatic cancer: a bibliometric analysis

Abstract

Background: Pancreatic cancer is a highly aggressive malignancy of the digestive system, characterized by insidious onset and rapid progression. Most cases are diagnosed at advanced stages, complicating surgical resection and presenting significant challenges for clinical treatment. Recent advancements have emphasized individualized treatment strategies tailored to patients' specific conditions. Consequently, accurate preoperative assessment is crucial, highlighting the urgent need to develop more reliable predictive models to guide personalized treatment plans.

Methods: A systematic literature search was conducted using Web of Science Core Collection (WoSCC) database, covering publications from January 1, 1995, to October 25, 2024. A comprehensive bibliometric analysis was performed employing analytical tools such as VOSviewer, CiteSpace and Microsoft Excel.

Results: This study includes 919 publications authored by 6716 researchers from 3727 institutions in 222 countries and regions. The articles were published in 301 journals, with 1,640 distinct keywords and 25,910 references. China led in publication volume, while the United States garnered the most citations. The top three research institutions in this field were Fudan University, Shanghai Jiao Tong University, and Sun Yat-sen University. Yu Xianjun from Fudan University emerged as the most prolific author with the highest citation count. Frontiers in Oncology had the highest publication volume, while the Annals of Surgery received the most citations. Medical imaging, biochemistry, immunology, bioinformatics, genetics, and interdisciplinary integrative research are the main research disciplines in the field of prognosis prediction for pancreatic cancer. The results of keyword co-occurrence and literature co-citation analysis revealed emerging hotspots and trends in this field, including CA19-9, CT, inflammation, machine learning, tumor microenvironment, radiomics, genes, nomograms, randomized controlled trials, long-term survival, and metastasis.

Conclusion: This bibliometric analysis provides an overview of research conducted over the past three decades, offering insights into the current state of knowledge and outlining directions for future studies on prognosis prediction models for pancreatic cancer. Biochemical indicators have consistently emerged as key research focal points. The tumor microenvironment represents a currently popular research direction, while bioinformatics, medical imaging, and artificial intelligence are gaining traction as future trends in this field. In the future, prognostic models for pancreatic cancer require further refinement to ensure reliable guidance for therapeutic decision-making.

Keywords: bibliometrics; pancreatic cancer; prediction model; prognosis; trends.

Figure 1

Flowchart of the literature screening process.

Figure 2

The global number of publications. (A) A line chart of the number of annual publications. The green dot represents the specific data of each year, and the red line represents the trend line of the data. The diagonal line in the figure represents the growth trend. (B) A clustered bar graph of the cumulative number of publications.

Figure 3

(A) Line chart of annual publication volume of the top five countries. This line chart shows the publishing volume trend of five countries (China, the United States, Japan, Germany and South Korea) from 2015 to 2024. (B) Annual publication heat map of the top five countries in terms of publication volume. Use different color shades to indicate the change in the number of publications per year. (C) Country/region collaboration map for prognostic prediction model for pancreatic cancer studies. Through the lines of different colors, it illustrates the academic exchanges between countries.

Figure 4

Institutional network map for prognostic prediction model for pancreatic cancer.

Figure 5

Network diagram of author collaborations for prognostic prediction model for pancreatic cancer.

Figure 6

The citation and co-citation analysis of journal. (A) Network map of citation analysis of journals with more than five publications.(B)Network map of co-citation analysis of journals with more than one hundred citations.

Figure 7

Double mapping overlay of journal research. The journals citing others are positioned on the left, while those being cited appear on the right.

Figure 8

Co-occurrence analysis of keywords. (A) Co-occurrence clustering map of keywords. (B) The keyword frequency-time map. The size of the circles is correlated with the frequency of keyword occurrences; blue represents earlier keyword occurrences, yellow represents later keyword occurrences (C) Density visualization of keywords.

Figure 9

(A) The top 20 keywords with the strongest citation bursts. The bar graph on the right shows the position of the reference burst duration (red part) of each keyword in the whole time span (2003-2024). (B) Keyword clustering time diagram. Different colored paths and node groups represent different clustering of research topics.