Emerging trends and hotspots in cervical intraepithelial neoplasia research from 2013 to 2023: A bibliometric analysis.

Liya Wang: Writing – review & editing, Writing – original draft, Supervision, Project administration, Methodology, Investigation, Data curation, Conceptualization. Bingying Sun: Writing – original draft, Validation, Resources, Methodology, Data curation. Ji Xu: Writing – original draft, Visualization, Resources, Methodology, Data curation. Dan Cao: Writing – review & editing, Resources, Funding acquisition. Yi Chen: Writing – review & editing, Visualization, Resources, Methodology. Ying Xu: Writing – review & editing, Supervision, Project administration, Investigation, Conceptualization. Dan Wu: Writing – review & editing, Supervision, Project administration, Investigation, Funding acquisition, Conceptualization.

Informed consent was not necessary for this study as all data utilized were obtained from the WoSCC database and did not involve any animal or human subjects.

This work was supported by the 10.13039/501100004921 Shanghai Jiao Tong University Medicine Engineering Fund (No. YG2022QN111) and the Application of minimally noninvasive screening program in standardized diagnosis and treatment of gynecological tumors during pregnancy (No. SHDC12023119).

A detailed literature search was carried out using the Science-Citation-Index-Expanded (SCI-E) of the-Web-of-Science-Core-Collection (WoSCC) database (2013–2023) on March 1, 2024. To mitigate the impact of frequent database updates, we conducted our search and data retrieval within a single day. Relevant publications were obtained from WoSCC by searching TS= ((“cervical intraepithelial neoplasia” or “cervical intraepithelial neoplasm” or “cervical intraepithelial lesion” or “cervical squamous intraepithelial lesion” or “cervical precancerous lesion” or “cervical intraepithelial neoplastic” or “cervical adenocarcinoma in situ” or “cervical glandular intraepithelial neoplasia” or “cervical carcinoma in situ” or “cervical cancer in situ” or “endocervical glandular dysplasia” or “cervical epithelial dysplasia” or “cervical dysplasia”)) AND Language = English AND Document type = Article. To confirm the accuracy of bibliometric analysis, two researchers (Liya Wang and Bingying Sun) examined all articles collected through search strategies based on their publication years, abstracts, and titles. The following exclusion criteria were applied: (i) unrelated to CIN; (ii) non-article document types, such as reviews, meta-analysis, proceeding papers, editorial materials, book chapters, news items, corrections, meeting abstracts, and retracted publications; and (iii) duplicate publications. In total, 4677 articles were identified. The detailed screening process is depicted in Fig. 1 . Fig. 1 Flowchart showing the inclusion and exclusion of publications in the CIN research. Fig. 1 Flowchart showing the inclusion and exclusion of publications in the CIN research. We downloaded all data, comprising full records and cited references of the retrieved articles, from the WoSCC database in TXT format. Subsequently, the data were imported into the bibliometric online analysis platform ( https://bibliometric.com ), CiteSpace V6.1.R6 (Drexel University, PA, USA), and VOSviewer 1.6.19 (Leiden University, The Netherlands). We performed an extensive bibliometric analysis covering various publication characteristics, such as authors, regions/countries, institutions, co-cited references, journals, and keywords exhibiting the most robust citation bursts. We extracted some bibliometric indicators, including the publication numbers and citation times from the text data downloaded from the WoSCC database. CiteSpace, established by Chen, stands as the most commonly utilized bibliometric analysis software at present [ 25 ]. CiteSpace serves as a valuable visual analysis tool utilized to uncover emerging trends and advancements within a research field [ 26 ]. We used CiteSpace (V6.1.R6) for analyzing the distribution of journals, authors, institutions, co-cited reference analysis and keyword burst identification to pinpoint current focal points and estimate future research trajectories in CIN. Based on the data exported from CiteSpace, we used Microsoft Office Excel 2019 to create an annual curve of the publication count in the CIN field over the past 11 years and analyze publication trends using an exponential growth function. The VOSviewer software can be used to construct visual bibliometric maps [ 27 ] and build networks between countries/regions based on collaborative data. We utilized the bibliometric-online-analysis-platform ( https://bibliometric.com ) to examine the annual publication count and publication trends of the top 10 regions/countries, as well as to assess the publication count from the top 10 most productive journals. The R package bibliometrix was utilized to generate a keyword cloud consisting of the top 100 most frequently occurring keywords. In addition, in specific cases, H-index is used to evaluate publications of regions/countries, journals, as well as individual academic achievements [ 28 ].

In this research, 7997 articles published between 2013 and 2023 satisfied the criteria for inclusion. In total, 3320 articles were excluded due to lack of relevance ( Fig. 1 ). Finally, 4677 articles were extracted from WoSCC that were published between 2013 and 2023. The total number of annual publications is shown at the top of each bar ( Fig. 2 A). The annual publication count serves as an indicator of the pace of research advancement in this field and reflects the level of attention dedicated to CIN research. From 2013 to 2023, the publication count regarding this topic remained at more than 360. The trendline of the total publication count conforms to a quadratic function curve, exhibiting R 2 of 0.935. Fig. 2 Analysis of the Volume and Trends of CIN Publications from 2013 to 2023: (A) Examination of Publication Quantity and Trends regarding CIN from 2013 to 2023, results extracted from Web of Science. (B) Assessment of Publication Quantity and Growth Trends among the Top 10 Regions/Countries in the CIN Field from 2013 to 2023. Fig. 2 Analysis of the Volume and Trends of CIN Publications from 2013 to 2023: (A) Examination of Publication Quantity and Trends regarding CIN from 2013 to 2023, results extracted from Web of Science. (B) Assessment of Publication Quantity and Growth Trends among the Top 10 Regions/Countries in the CIN Field from 2013 to 2023. The analysis of article publication across various regions/countries was conducted using the bibliometric-online-analysis platform to identify the leading contributors during the last 11 years. In the histogram presented in Fig. 2 B, the top 10 regions/countries accounted for the majority of publications during this period. Notably, USA and China have consistently held prominent positions in CIN research. USA maintained its dominance in the field until 2016, after which China surpassed it in publication output from 2017 onward. We investigated the cooperation between scholars, institutes, and countries and identified the main collaborative partners using the VOSviewer software. In total, 4677 papers were published in 135 regions/countries from 2013 to 2023. Fig. 3 A shows the top 33 cooperating countries in the field of CIN research. We found that USA was the most actively engaged country in international collaboration, and China had the highest publication count. In addition, the USA and China demonstrated the most intimate collaboration. Fig. 3 Collaborative network map of regions/countries and institutions. (A) The collaborative efforts of the top 33 countries. (B) The Leading 10 Most Prolific Institutions. The circles depict various regions/countries and institutions, with circle size correlating to the number of publications. Link thickness indicates the level of collaboration intensity. Color-coded clusters represent regions/countries and institutions with robust collaborative ties. Fig. 3 Collaborative network map of regions/countries and institutions. (A) The collaborative efforts of the top 33 countries. (B) The Leading 10 Most Prolific Institutions. The circles depict various regions/countries and institutions, with circle size correlating to the number of publications. Link thickness indicates the level of collaboration intensity. Color-coded clusters represent regions/countries and institutions with robust collaborative ties. We utilized CiteSpace software to import TXT files and elucidate the patterns of inter-institutional cooperation. After deduplication by CiteSpace, 4677 publications were included in subsequent analyses. Fig. 3 B shows the ten institutions with the highest productivity. The National Cancer Institute (NCI) had the highest publication count (141) and most frequently collaborated with other institutions. Among them, 2 were from USA, including the NCI (141) and Albert Einstein College of Medicine (91). These are prestigious higher education or medical centers in USA with strong scientific research capabilities. Another four institutions were from China, namely the Chinese Academy of Medical Sciences (72), Peking University (62), Zhejiang University (61) and Peking Union Medical College (57). The remaining four institutions with the highest publication numbers were from Sweden, the United Kingdom, the Netherlands and France, which also contributed significantly to the CIN field. During the last 11 years, 21,154 researchers have involved in CIN research. Among these authors, 382 researchers published over 10 articles, making major contributions to publication output. Fig. 4 displays the top 11 most prolific authors. Schiffman Mark, from the NIH National Cancer Institute, contributed the highest number of articles (92), followed by Castle Philip E from Duke University (91) and Wentzensen Nicolas from the NCI (63). Interestingly, two authors together occupied the tenth place, namely Heideman Danielle and Snijders Peter from Vrije University Amsterdam (Netherlands) (40). The collaborative network diagram of the top 10 prolific authors demonstrated strong cooperative relationships. Fig. 4 CiteSpace network illustrating authorship connections in CIN research. The figure displays the top 11 authors with the highest publication count. Circles denote authors, while links among circles depict collaborations among authors. Fig. 4 CiteSpace network illustrating authorship connections in CIN research. The figure displays the top 11 authors with the highest publication count. Circles denote authors, while links among circles depict collaborations among authors. We utilized the bibliometric-online-analysis-platform to assess journal impact. Our analysis covered 4677 original articles published across 680 academic journals. The top 10 journals with the highest citation counts are listed in Table 1 . Articles published in the International Journal of Cancer were cited most frequently in the past 11 years (4357 times), followed by PLOS One (4,127), Gynecologic Oncology (2,209), Journal of Lower Genital Tract Disease (1,820), BMC Cancer (1,109), Journal of Medical Virology (860), Oncology Letters (698), Archives of Gynecology and Obstetrics (617), International Journal of Gynecology & Obstetrics ( 516 ), and Diagnostic Cytopathology (464). Five journals were from USA, and the others were from Switzerland, the United Kingdom, Greece, Germany, and Netherlands. The International Journal of Cancer , based in Switzerland, achieved the highest average citation count per paper at 25.04. There were only 3 Q1 journals based on JCR partitions, among which the Journal of Medical Virology had the highest impact factor (12.7). This finding indicates that CIN research still needs improvements. Table 1 Ranking of the top 10 most cited journals in the CIN research from 2013 to 2023, arranged by total citations. Table 1 Rank Journal Title Frequency Total citations Average citation per paper H-index Impact factor (2022) JCR Country 1 INTERNATIONAL JOURNAL OF CANCER 174 4357 25.04 38 6.4 Q1 Switzerland 2 PLOS ONE 167 4127 24.71 32 3.7 Q2 USA 3 GYNECOLOGIC ONCOLOGY 96 2209 23.01 28 4.7 Q1 USA 4 JOURNAL OF LOWER GENITAL TRACT DISEASE 185 1820 9.84 20 3.7 Q2 USA 5 BMC CANCER 80 1109 13.86 22 3.8 Q2 UK 6 JOURNAL OF MEDICAL VIROLOGY 79 860 10.89 18 12.7 Q1 USA 7 ONCOLOGY LETTERS 70 698 9.97 15 2.9 Q3 GREECE 8 ARCHIVES OF GYNECOLOGY AND OBSTETRICS 88 617 7.01 12 2.9 Q3 Germany 9 INTERNATIONAL JOURNAL OF GYNECOLOGY & OBSTETRICS 73 516 7.07 13 2.6 Q3 Netherlands 10 DIAGNOSTIC CYTOPATHOLOGY 70 464 6.63 11 3.8 Q2 USA Ranking of the top 10 most cited journals in the CIN research from 2013 to 2023, arranged by total citations. We also identified the ten most cited papers from 4677 articles. Table 2 highlights the most cited article published in NEJM in 2015, with 927 citations [ 29 ]. The study proposed that the 9-valent-HPV-vaccine (9vHPV) could prevent infections and high-grade vaginal, vulvar, or cervical diseases associated with HPV-31/33/45/52/58 in vulnerable populations and triggers antibody response to HPV-6/11/16/18. The article with the second-highest number of citations was published in PLOS One in 2013, accumulating 525 citations [ 30 ]. This paper indicated that in addition to screening for and treating CIN, maintaining a healthy BMI through a focus on physical activities can prevent cervical cancer. The third most cited article demonstrated that VGX-3100 can cause histopathological regression of CIN2/3 related to HPV-16/18, serving as a non-surgical therapeutic option for CIN2/3 [ 31 ]. Among the ten most cited articles, 3 were published in Lancet , 3 were published in PLOS One , and the others were published in NEJM , Nature Genetics , Scientific Reports , and JNCI . Table 2 The highest-cited 10 articles among the 4677 retrieved publications on CIN research from 2013 to 2023, arranged by citation frequency. Table 2 Rank Title First Author Source Year Cited Frequency DOI 1 A 9-Valent HPV Vaccine against Infection and Intraepithelial Neoplasia in Women Joura, EA NEW ENGLAND JOURNAL OF MEDICINE 2015 927 10.1056/NEJMoa1405044 2 Mild Obesity, Physical Activity, Calorie Intake, and the Risks of Cervical Intraepithelial Neoplasia and Cervical Cancer Lee, JK PLOS ONE 2013 525 10.1371/journal.pone.0066555 3 Safety, efficacy, and immunogenicity of VGX-3100, a therapeutic synthetic DNA vaccine targeting human papillomavirus 16 and 18 E6 and E7 proteins for cervical intraepithelial neoplasia 2/3: a randomised, double-blind, placebo-controlled phase 2b trial Trimble, CL LANCET 2015 436 10.1016/S0140-6736(15)00239-1 4 Genome-wide profiling of HPV integration in cervical cancer identifies clustered genomic hot spots and a potential microhomology-mediated integration mechanism Hu, Z NATURE GENETICS 2015 329 10.1038/ng.3178 5 Gankyrin Is Frequently Overexpressed in Cervical High Grade Disease and Is Associated with Cervical Carcinogenesis and Metastasis Liu, Y PLOS ONE 2014 279 10.1371/journal.pone.0095043 6 Cervical intraepithelial neoplasia disease progression is associated with increased vaginal microbiome diversity Mitra, A SCIENTIFIC REPORTS 2015 273 10.1038/srep16865 7 Final efficacy, immunogenicity, and safety analyses of a nine-valent human papillomavirus vaccine in women aged 16–26 years: a randomised, double-blind trial Huh, WK LANCET 2017 267 10.1016/S0140-6736(17)31821-4 8 The effects of the national HPV vaccination programme in England, UK, on cervical cancer and grade 3 cervical intraepithelial neoplasia incidence: a register-based observational study Falcaro, M LANCET 2021 255 10.1016/S0140-6736(21)02178-4 9 Cervical Microbiome and Cytokine Profile at Various Stages of Cervical Cancer: A Pilot Study Audirac-Chalifour, A PLOS ONE 2016 244 10.1371/journal.pone.0153274 10 An Observational Study of Deep Learning and Automated Evaluation of Cervical Images for Cancer Screening Hu, LM JNCI-JOURNAL OF THE NATIONAL CANCER INSTITUTE 2019 207 10.1093/jnci/djy225 The highest-cited 10 articles among the 4677 retrieved publications on CIN research from 2013 to 2023, arranged by citation frequency. Furthermore, we computed the total citation number for all papers published by each country and compiled a list of the top ten countries ( Fig. 5 ). During this 11-year period, USA ranked first with 21,451 citations. China and the UK ranked 2nd and 3rd with 14,553 and 8384 citations, respectively. The tenth place was occupied by Japan, with 1739 citations. Fig. 5 Bar graph depicting the total publication count from 4677 papers published from countries between 2013 and 2023. This figure displays the top ten countries with the highest total citation number. Bar represent countries, with its length directly proportional to the total citation number. Fig. 5 Bar graph depicting the total publication count from 4677 papers published from countries between 2013 and 2023. This figure displays the top ten countries with the highest total citation number. Bar represent countries, with its length directly proportional to the total citation number. We determined the average citations per article per country by dividing each country's total citations by its total publications. Afterwards, we identified the top 3 countries with the highest average citations per article, including the United Kingdom (UK) (28.13), the Netherlands (23.97), and France (23.05) ( Table 3 ). Articles from these countries, except for those from Japan (9.06), had an average of over 10 citations. Table 3 The leading 10 nations with the most significant aggregate citation count in CIN research from 2013 to 2023, ordered by the average citation number. Table 3 Rank Country Number of Publications Total number of citations H-index Average number of citations 1 UK 298 8384 45 28.13 2 Netherlands 218 5226 40 23.97 3 France 189 4357 35 23.05 4 USA 977 21451 66 21.96 5 Germany 201 4319 34 21.49 6 South Korea 180 2791 24 15.51 7 Brazil 245 3745 25 15.29 8 Italy 222 3287 28 14.81 9 China 1219 14553 43 11.94 10 Japan 192 1739 21 9.06 The leading 10 nations with the most significant aggregate citation count in CIN research from 2013 to 2023, ordered by the average citation number. CiteSpace was used to analyze 4677 papers retrieved from WoSCC and their 60666 references (excluding self-citations) to assess the association between 2 articles by visualizing their co-citation patterns. Fig. 6 A presents the map of co-citation authors for CIN research in CiteSpace. Specifically, the figure shows the authors with the highest co-citation frequency, listed in the top 10. Fig. 6 Network Analysis of Citations in CIN Research between 2013 and 2023: (A) Visualization of co-citation patterns using CiteSpace, encompassing 60,666 references related to CIN research. Each node denotes an author, and links between nodes signify that the authors' publications were cited as references in the same article. Node size correlates with citation frequency, with denser red nodes indicating top-cited papers recently, and denser purple nodes indicating top-cited references in previous years. The figure highlights the top 10 most cited authors in CIN research. (B) Identification of the top 20 references with the most notable citation bursts in CIN research publication from 2013 to 2023. References highlighted in red experienced a rapid surge in citation frequency during this period, while blue denotes comparatively less cited periods. Fig. 6 Network Analysis of Citations in CIN Research between 2013 and 2023: (A) Visualization of co-citation patterns using CiteSpace, encompassing 60,666 references related to CIN research. Each node denotes an author, and links between nodes signify that the authors' publications were cited as references in the same article. Node size correlates with citation frequency, with denser red nodes indicating top-cited papers recently, and denser purple nodes indicating top-cited references in previous years. The figure highlights the top 10 most cited authors in CIN research. (B) Identification of the top 20 references with the most notable citation bursts in CIN research publication from 2013 to 2023. References highlighted in red experienced a rapid surge in citation frequency during this period, while blue denotes comparatively less cited periods. Table 4 shows the top ten references, ranked by the co-citation frequency. The most cited reference was published by JAMA in 2002 [ 32 ]. This article updated the Bethesda System terminology concerning the reporting of cervical cytology findings. The 2001 Bethesda System terminology represents significant progress in our biological comprehension of cervical neoplasia and advancements in cervical screening technology. The second most cited article, published in the Journal of Pathology in 1999, emphasized the importance of HPV testing alongside, or as a potential substitute for, cervical cytology in routine cervical screening [ 33 ]. The highly cited papers listed in Table 4 are widely acknowledged as seminal contributions to CIN research. Table 4 The top 10 references with the highest citation count among the 4677 papers in CIN research between 2013 and 2023, organized by citation frequency. Table 4 Rank Title First Author Source Year Cited Frequency DOI 1 The 2001 Bethesda System - Terminology for reporting results of cervical cytology Solomon, D JAMA-JOURNAL OF THE AMERICAN MEDICAL ASSOCIATION 2002 557 10.1001/jama.287.16.2114 2 Human papillomavirus is a necessary cause of invasive cervical cancer worldwide Walboomers, JMM JOURNAL OF PATHOLOGY 1999 528 10.1002/(SICI)1096–9896(199909)189:1  3.0.CO; 2-F 3 Human papillomavirus genotype attribution in invasive cervical cancer: a retrospective cross-sectional worldwide study de Sanjose, S LANCET ONCOLOGY 2010 393 10.1016/S1470-2045(10)70230-8 4 Efficacy of HPV-based screening for prevention of invasive cervical cancer: follow-up of four European randomised controlled trials Ronco, G LANCET 2014 359 10.1016/s0140-6736(13)62218-7 5 Epidemiologic classification of human papillomavirus types associated with cervical cancer Muñoz, N NEW ENGLAND JOURNAL OF MEDICINE 2003 351 10.1056/NEJMoa021641 6 2012 Updated Consensus Guidelines for the Management of Abnormal Cervical Cancer Screening Tests and Cancer Precursors Massad, LS JOURNAL OF LOWER GENITAL TRACT DISEASE 2013 284 10.1097/LGT.0b013e318287d329 7 Human papillomavirus and cervical cancer Schiffman, M LANCET 2007 281 10.1016/S0140-6736(07)61416-0 8 American Cancer Society, American Society for Colposcopy and Cervical Pathology, and American Society for Clinical Pathology Screening Guidelines for the Prevention and Early Detection of Cervical Cancer Saslow, D AMERICAN JOURNAL OF CLINICAL PATHOLOGY 2012 237 10.1309/AJCPTGD94EVRSJCG 9 2019 ASCCP Risk-Based Management Consensus Guidelines for Abnormal Cervical Cancer Screening Tests and Cancer Precursors Perkins, RB JOURNAL OF LOWER GENITAL TRACT DISEASE 2020 232 10.1097/LGT.0000000000000525 10 Evidence Regarding Human Papillomavirus Testing in Secondary Prevention of Cervical Cancer Arbyn, M VACCINE 2012 231 10.1016/j.vaccine.2012.06.095 The top 10 references with the highest citation count among the 4677 papers in CIN research between 2013 and 2023, organized by citation frequency. Furthermore, we conducted an analysis of strong citation bursts among references in this field to identify research hotspots. Fig. 6 B lists the top 20 references exhibiting the most pronounced citation bursts between 2013 and 2023. 'Begin' indicates the year when the references were initially cited, while 'End' signifies the year of its last citation. Six of the articles continued to receive significant citations. A worldwide analysis published by Marc Arbyn et al. was the most recent reference burst occurred in 2021 and extended until the conclusion of 2023 [ 34 ]. The researchers described the incidence pattern and mortality rates of cervical cancer alongside HPV prevalence data from 185 countries in 2018, enabling a thorough initial evaluation of the global burden of cervical cancer. Chen WD et al. published a multicenter study in CA-Cancer J Clin in 2017 with the longest citation duration [ 35 ]. The authors estimated the incidence of new cases and mortality rates from cancer from 72 regions of China in 2015 and found that while developed countries witnessed declining incidence trends, China experienced a significant rise in cervical cancer incidence due to deficiencies in Pap screening tests. The highest strength was 83.07 in a consensus on risk-based management guidelines, published in the Journal of Lower Genital Tract Disease by Perkins RB et al. in 2020, which burst in 2021 [ 36 ]. In this study, the American-Society-of-Colposcopy-and-Cervical-Pathology (ASCCP) recommended incorporating HPV testing into risk stratification for managing abnormal cervical cancer screening results and pre-cancerous lesions. The CiteSpace graph depicting co-citations was clustered based on keywords derived from the references of 4677 papers ( Fig. 7 A). The investigation into co-citation clusters indicated the most significant terms for CIN research based on hierarchical cluster labels were # 0 prevalence, # 1 human papillomavirus, # 2 DNA methylation, # 3 p16, # 4 methylation, # 5 conization, # 6 VALGENT, # 7 deep learning, # 8 vaginal microbiome, and # 9 immunohistochemistry. A negative correlation was observed between the number of cluster labels and the number of articles included in each cluster. Alternatively, the cluster labeled as #0 encompassed the highest publication count among the 60666 references. A summary of clusters is provided in Table 5 . Fig. 7 Analysis of Clustered Network of CIN Publications from 2013 to 2023: (A) Visual representation of the clustering network of co-citations using CiteSpace. The figure highlights the top 10 largest citation clusters. (B) Timeline view of co-cited references pertaining to CIN. Each horizontal line denotes a cluster, with smaller label numbers indicating larger clusters. Node size corresponds to co-citation frequency, and links represent co-citation relationship. The colors of lines and nodes denote the different years cited. Cluster labels are displayed on the right side. Fig. 7 Table 5 Overview of 10 clusters of cited works in CIN research. Table 5 Cluster ID Term Size Silhouette a 0 prevalence 130 0.824 1 human papillomavirus 123 0.774 2 dna methylation 93 0.754 3 p16 86 0.829 4 methylation 77 0.895 5 conization 71 0.82 6 valgent 69 0.821 7 deep learning 61 0.908 8 vaginal microbiome 44 0.862 9 immunohistochemistry 41 0.893 a The silhouette value greater than 0.5 suggests that the clustering outcomes are reliable. Analysis of Clustered Network of CIN Publications from 2013 to 2023: (A) Visual representation of the clustering network of co-citations using CiteSpace. The figure highlights the top 10 largest citation clusters. (B) Timeline view of co-cited references pertaining to CIN. Each horizontal line denotes a cluster, with smaller label numbers indicating larger clusters. Node size corresponds to co-citation frequency, and links represent co-citation relationship. The colors of lines and nodes denote the different years cited. Cluster labels are displayed on the right side. Overview of 10 clusters of cited works in CIN research. The silhouette value greater than 0.5 suggests that the clustering outcomes are reliable. A timeline perspective is displayed in Fig. 7 B to clearly describe the shift of hotspots over the past 11 years. We found that the focus of research on CIN shifted from prevalence, human papillomavirus, methylation, and immunohistochemistry to DNA methylation, p16, conization, deep learning, and vaginal microbiome. A word cloud was produced for the top 100 most frequently occurring keywords in CIN researches ( Fig. 8 A). The font size denotes the frequency of using keywords. After filtering out keywords deemed irrelevant for our research trend analysis, cancer, risk, human-papillomavirus, intraepithelial neoplasia, prevalence, cytology, lesions, infection, management, and neoplasia were the top ten highest frequency keywords. Fig. 8 Examination of Keyword Patterns and Publication Bursts in CIN Studies from 2013 to 2023: (A) Visualization of the top 100 most frequently keywords in CIN research via word cloud analysis. (B) Identification of the top 15 keywords with the most notable citation bursts, organized by starting year. Keywords highlighted in red experienced a sudden surge in their usage frequency during that period, while blue indicates comparatively less popular times. Fig. 8 Examination of Keyword Patterns and Publication Bursts in CIN Studies from 2013 to 2023: (A) Visualization of the top 100 most frequently keywords in CIN research via word cloud analysis. (B) Identification of the top 15 keywords with the most notable citation bursts, organized by starting year. Keywords highlighted in red experienced a sudden surge in their usage frequency during that period, while blue indicates comparatively less popular times. Keyword burst detection facilitates the swift identification of emerging research hotspots within a specific field. Fig. 8 B displays the top 15 references with the most robust keyword bursts in CIN research between 2013 and 2023. “PCR” was the most robust burst keyword (strength 10.66) from 2013 to 2023, followed by “photodynamic therapy” (strength 9.58), “deep learning” (strength 8.55), and “metaanalysis” (strength 7.82). “machine learning”, “treatment failure”, “vaginal microbiota”, “early detection of cancer”, “deep learning”, “photodynamic therapy”, “cancer screening test”, and “excision” were found as the emerging research focal points.

In summary, this represents the inaugural bibliometric and visual analysis of previous studies on CIN. Using bibliometric atlas, we analyzed the studies on CIN from 2013 to 2023. In the past 11 years, the publications of CIN have been in a relatively steady trend. USA and China have made the most significant contribution to the CIN research field. USA dominated international collaboration, while China has published more articles since 2017. Recent studies have focused on photodynamic therapy and deep learning. Photodynamic therapy may be an effective technology for treating CIN. Deep learning suggests that computer-related technology may be an important method for diagnosing CIN in the future. Researchers and CIN related clinicians can utilize our findings to enhance their comprehension of the research hotspots and development trends in CIN field over the past 11 years. Additionally, this study can help CIN related clinicians to understand the advanced technologies for screening and treating CIN, and then apply them to clinical practice.

This research represents the inaugural bibliometric and visual exploration of CIN. Currently, the research hotspots in the field of CIN mainly focus on deep learning, machine learning, and photodynamic therapy, representing new technologies in diagnosing and treating CIN. Through our bibliometric analysis, scholars interested in CIN can gain a comprehensive overview of the field and swiftly identify the most recent research trends. Global publication on CIN during the past 11 years exhibited a relatively stable trend. The USA demonstrated the highest level of international collaboration, whereas China had the highest publication count. Our results showed that USA maintains a leading role, totalling 21,451 citations in 977 articles. USA also has two of the top 10 most productive research organizations, with the NCI being the most productive institution worldwide. In addition, four of the top eleven authors in this field were American, including Schiffman Mark, Castle Philip E, Wentzensen Nicolas, and Stoler Mark H. The publication count in China has surpassed USA since 2017, and four of the top ten institutions ranked by publication output were from China. These findings highlight the significant contributions and potential scientific innovations from Chinese efforts in CIN research. Articles from the United Kingdom (UK) had the highest average number of citations, followed by the Netherlands and France, highlighting the excellence of publications from European countries. Although China ranked 1st regarding the total publication count, the average number of citations per article was only 11.94. Moreover, there were only two Chinese authors among the top 10 authors, including Xun Zhang who occupied the fifth position and You-Lin Qiao who occupied the ninth position. These findings indicate that China still has a long way to go in the field of CIN research. As the most productive institution worldwide, NCI conducted a large-scale study and illuminated that the risk of precancerous lesion progression differed substantially by 13 different high-risk HPV subtypes, with HPV16 inducing uniquely elevated risk. For all infections caused by HPV subtypes, annual rates of progression dramatically decreased over the first three years. By year five, the annual rates of progression were less than 0.1 % for most HPV subtypes (except HPV18) and remained low in the following years [ 37 ]. Over the past 11 years, Schiffman Mark from the NCI contributed the highest number of articles focusing on cervical smear screening and HPV infection-related screening. His article published in the International Journal of Cancer was a cohort study on partial HPV typing and cytological classification for cervical cancer screening [ 38 ]. His latest study measured the effect of HPV16 E7 nucleotide variants on E7 protein expression and transformation in cancer-free subjects [ 39 ]. The second most productive researcher, Castle Philip E, recommended that screening intervals of every 5 years or more might be feasible and safe after one or more negative HPV and cytology co-testing against cervical cancer [ 40 ]. He also compared differences in the risk of HSIL in women with or without human immunodeficiency virus (HIV) in combination with co-testing [ 40 ]. Co-cited cluster analysis can provide valuable insights into the cutting-edge developments in scientific works. We found that “prevalence”, “human papillomavirus”, “methylation”, and “immunohistochemistry” were the early main focuses in CIN research. As early as 1894, research has demonstrated that HPV-16 is strongly associated with CIN, which has considerable clinical significance in the differential diagnosis, prognosis, and management of early cervical cancer [ 41 ]. In the early days, scholars paid more attention to the risk of CIN associated with HPV infection [ [42] , [43] , [44] ]. The main research themes have been “DNA methylation”, “deep learning”, “p16”, “vaginal microbiome”, and “conization” in recent years. Methylation has always been a research hotspot in the past 11 years. DNA methylation is a common occurrence in the progression from HPV infection to cervical precancerous lesions [ 45 ]. Previous studies have shown that cytosine methylation in CpG dinucleotides, also known as CpG sites, is an important cellular mechanism regulating gene expression. Increased methylation of multiple CpG sites in various regions of the HPV-16 genome correlated with the presence of CIN3 [ 46 ]. FAM19A4/miR124-2 DNA methylation tests combined with HPV genotyping can be utilized to support a wait-and-see follow-up approach for female patients with CIN2/3 and avoid overtreatment [ 47 ]. With the development of artificial intelligence (AI), disease models based on histological images and machine deep learning are increasingly used to diagnose various tumors. Researchers began to develop a computerized reading system to evaluate cervical cytological specimens for diagnosis [ 48 , 49 ]. Since 2015, more studies founded that vaginal microbiome diversity was closely related with the progression of CIN [ [50] , [51] , [52] ]. Keyword bursts represent keywords extensively cited in publications, aiding in the recognition of trends and focal points within a specific research domain. Our results showed that the strongest keyword has been polymerase chain reaction (PCR) in the CIN research field over the past 11 years. It is a key technology for early diagnosis of HPV infection and has been widely valued by scholars. As early as 1989, PCR was used to identify HPV in CIN [ 53 ]. A study compared the clinical performance of 4 low-cost HR-HPV assays on the basis of real-time PCR. The study applied the Cobas test for detecting CIN2+ lesions and accelerating the control of cervical cancer in low-resource settings [ 54 ]. The second keyword, “photodynamic therapy”, began in 2020 with a strength of 9.58, persisting until the conclusion of 2023. Despite large loop, cold knife conization, cryotherapy, and laser surgery are typical treatments of high-grade CIN, several severe side effects cannot be ignored. Thus, more researchers pay attention to a non-invasive treatment intervention that reduces the advancement of the disease. Photodynamic therapy (PDT) is a non-invasive technique based on the interaction of light, photosensitizer, and oxygen. It selectively destroys abnormal tissues through photodynamic reactions. Most studies reported the use of PDT for treating CIN1/2 [ [55] , [56] , [57] ]. Compared with other invasive treatments, photodynamic therapy can preserve cervical function [ 16 ] and has advantages in eliminating HPV infection and mitigating the recurrence rate [ 15 ]. Prior studies have shown the effect of PDT on local immune responses, eradicating pathogenic microorganisms of CIN1/2 [ 58 , 59 ]. In addition, PDT presented the best results two years after treatment of CIN1/2, preventing the progression of CIN [ 56 ]. The keywords “deep learning”, “machine learning”, “vaginal microbiota”, “cancer screening test”, and “treatment failure” are also current research hotspots. Recent studies have focused on HR-HPV genotype distribution and screening in CIN [ [60] , [61] , [62] ] and predictors of treatment failure (residual or recurrent CIN) [ 63 , 64 ]. With the rapid advancement of computer simulation technology, deep learning has been increasingly applied to the diagnosis of CIN. The application of this technology in cervical cytology can replace some repetitive human labor [ 65 ]. Deep learning technology can help colposcopists identify cervical lesions and reduce the rate of missed diagnosis [ 66 ]. Our study possesses certain limitations. Firstly, owing to constraints within the CiteSpace software, only data extracted from the SCI-E of the WoSCC database were subject to visual analysis, and other databases, such as PubMed and Embase, were not searched, which might have increased the risk of bias. Second, owing to the limited search strategies, literature types, and language limitations, this research only concentrated on papers published in English. English remains the predominant language for academic journals. Consequently, there is a possibility of overlooking articles that do not adhere to these criteria. Third, due to the manual exclusion of articles unrelated to CIN, there may be some bias in accuracy. Therefore, in order to ensure the accuracy of exclusion and reduce bias, we have adopted a strategy of two researchers excluding the retrieved articles separately. If there are any questions about certain articles, a third researcher will be asked to make the final judgment. Finally, bibliometric analysis may introduce a possible time-effect bias, as older articles tend to accumulate more citations. Additionally, factors like self-citation can influence citation rates. To mitigate such errors, we conducted a comprehensive review of the included papers.

Cervical-intraepithelial-neoplasia (CIN) comprises a spectrum of precancerous lesions associated with the development of cervical invasive carcinoma in women. CIN can be divided into 3 grades, namely CIN1, CIN2, and CIN3. Low-grade-intraepithelial-lesions (LSIL) are equivalent to CIN1, and high-grade-intraepithelial-lesions (HSIL) include CIN3 and most of CIN2. Most LSIL resolve naturally, but HSIL may progress into cervical cancer. It ranks as the 4th most frequently diagnosed cancer among women globally, with more than 604,000 new cases worldwide and more than 341,000 deaths [ 1 ]. The persistent infection of human-papillomavirus (HPV 16/18/31/33/35/39/45/51/52/56/58/59/68) contributes to the risks of developing cervical cancer and precancerous lesions [ 2 ]. Other risk factors include immunosuppression (altered vaginal microbiota, immunosuppressive medications, and HIV infection), experiencing a young age at first pregnancy, having multiple sexual partners, and smoking [ 3 ]. HPV genotyping testing combined with cervical exfoliation cytology can be carried out to evaluate the risks of CIN and cervical cancer [ 2 ]. Moreover, methylation detection [ 4 ], HPV integration detection, HPV DNA load detection, immunocytochemical staining, and artificial intelligence (AI) technology [ 5 ] have certain application prospects in screening, but still large-sized prospective studies are needed. Currently, HPV vaccination can achieve primary prevention of CIN [ 6 ]. Colposcopy or direct vision cervical histological biopsy is considered the benchmark for diagnosis. Approximately 60 % of LSIL resolve naturally and can be followed up. Cervical conization is a traditional treatment strategy for HSIL, including cold-knife-conization (CKC) and loop-electrosurgical-excision-procedure (LEEP) [ 7 ]. Cervical conization may reduce the likelihood of HSIL advancing to invasive cancer. Additionally, ablation therapy, including cryotherapy, laser therapy, and electrocoagulation therapy, can also be applied to some HSIL. However, there is still a long-term risk of persistence, recurrence, or even progression to invasive cancer after resection. Studies have shown that the risk of persistence and recurrence after the treatment of HSIL is also significantly higher than that after the treatment of LSIL [ 8 ]. Short-term complications, such as infection [ 9 ], bleeding [ 10 ], and injury, as well as long-term complications, such as cervical stenosis [ 11 ], can occur after treatment, affect pregnancy outcomes [ 12 ], and lead to endometriosis [ 13 ]. Recently, photodynamic therapy has been gradually popularized as a highly targeted, non-invasive therapy [ 14 ]. Although photodynamic therapy offers benefits in maintaining the integrity of normal tissue structures, preserving fertilities, resolving HPV infections, and decreasing recurrence rates compared with conventional therapy [ 15 , 16 ], follow-up management after treating HSIL is pivotal. The high prevalence of CIN worldwide has put tremendous pressure on women and imposed an economic burden, increasing the need for comprehensive research in this field. Yet, there is presently a deficiency in objective analysis based on big data analytics for the research overview and hotspots of CIN over the past 11 years. Bibliometric analysis refers to a thorough and timely review of publications within a specified timeframe, examining factors such as publication count, geographic distribution, authorship, keywords, references, and other relevant parameters [ [17] , [18] , [19] , [20] , [21] ]. Bibliometric analysis can provide researchers with valuable insights into prevailing research focal points and anticipated future trends [ 22 , 23 ]. The application of bibliometric analysis in the medical field is particularly common, enabling researchers to analyze the research hotspots and development trends of a certain field on macroscopic and microscopic levels [ 24 ]. Despite our knowledge, no bibliometric analysis focusing on CIN has been published. Thus, this study aims to bridge this gap by examining publication trends in CIN from 2013 to 2023, with the goal of contributing to the advancement of research in this field.

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Data will be provided upon request.