香京julia种子在线播放

Subject words were searched using the Medical Subject Headings (Mesh) database of Pubmed (https://pubmed.ncbi.nlm.nih.gov/) to make the retrieval more comprehensive and accurate. All the related literature about the relationship between ctDNA and breast cancer was gathered from Web of Science (WoS) Core Collection database with no limitation for citation index, including Science Citation Index Expanded (SCI-EXPANDED), Social Sciences Citation Index (SSCI), Arts & Humanities Citation Index (ESCI), Current Chemical Reactions (CCR-EXPANDED), Index Chemicus (IC). The query parameter was designed as followed: #1, TS=(“breast neoplasm*”) OR TS=(“breast tumor*”) OR TS=(“breast cancer*”) OR TS=(“mammary cancer*”) OR TS=(“mammary neoplasm*”) OR TS=(“breast carcinoma*”) OR TS=(“mammary carcinoma*”); #2, TS=(“Circulating Tumor DNA”) OR TS=(“Cell-Free Tumor DNA”) OR TS=(“ctDNA”); #3, “#1”, and “#2”. A total of 1097 pieces of literature were got in the preliminary search. And then, we did some screening: the publication year was refined from January 1st, 2012 to December 31st, 2021, the document type was set to articles or reviews, and the language was restricted to English. And we also excluded the articles published in 2022. All the searches were performed on August 28, 2022 and finally, we got 739 publications, including 472 articles and 267 reviews, accounting for 63.870% and 36.130%, respectively. The flow diagram of data collection and filtering was shown in Figure 1 .

With the function of “analyze results” in WoS, we did the preliminary analysis of the retrieved literature, including author, publication year, document type, research direction, country, affiliated institution, publisher. The function of “citation report” was used to acquire the information about the number and citation frequency of each paper.

Then, the information of the literature retrieved by WoS was exported in plain txt format and imported into the tools of biblioshiny (the shiny app for bibliometrix) and VOSviewer for further analysis of annual production, authors, countries, journals, institutions, keywords, and citations.

Bibliometrix package is a scientometric analysis software based on R language. In this study, we used the biblioshiny website with R language (version R 4.1.0) to analyze the bibliographic information of retrieved literature. An overview of basic information contained the time span, annual growth rate, average citations per doc, document contents, document types etc. Descriptive analysis and visual presentation about the annual production, average citation per year, contribution of countries, organizations, and journals, authors publications and impact were acquired. The author impact was evaluated by Hirsch h-index, g-index, m-index. The h-index indicated the impact of a researcher’s scientific outputs, defined as the largest number of articles published by a researcher which have been cited at least h times (18, 19).

VOSviewer (version 1.6.18), a computer program for constructing and visualizing bibliometric maps in an intuitive and intelligible view (20). In this study, VOSviewer was applied to perform the co-authorships analysis of countries, organizations, and authors, bibliographic coupling analysis of journals, co-citation analysis of cited reference, co-occurrence analysis of keywords. We also used this software to gain the total link strength among different countries, organizations, and authors.

IBM SPSS (version 26.0) was used for statistics analysis. We set the data type as numerical variables and used Spearman correlation analysis to analyze the correlations between selected variables. All the tests were two-sided, and p-values < 0.05 were considered statistically significant.

A total number of 739 documents contributing to this research field from 2012 to 2021 was obtained from WoS. Figure 2A showed the change in annual publication volume from 2012 to 2021 (r² = 0.973, [CI, 0.806 to 1.000]; p < 0.001). From 8 articles published in 2012 to 130 articles published in 2021, the overall number of articles published showed an upward trend, with an average annual growth rate of 36.31%. The highest number of publications was produced in 2020 (131, 17.72%). From 2012 to 2014, there were few articles in this field, with a total of 31 articles published in the three years, reflecting that this field arouse little concern from the researchers at that time. From 2015 to 2018, the annual output increased rapidly, with a median annual output growth rate of 34.65% (2015 had the highest output growth rate of 65.11%), indicating a sharp rise in academic attention and study in this field during this period. There was a small decline in publishing in 2019 but it remained high. From 2019 to 2021, the number of publications continued to increase slowly, indicating that researchers or scholars have kept a watchful eye on this field.

Among all the retrieved articles, a total of 37,987 citations were received and the average citation frequency of each article was 51.4 times. Figure 2B showed the average citations per year from 2012 to 2021. The number of citations increased sharply in 2014, reaching the highest peak (n=47.68) in the past decade. In 2018. the citations per article reached another small peak (n=14.1) and then it continued to decline until it reached the lowest average citation frequency in 2021 (n=6.47).

Country analysis helps to reveal the geographical distribution of relevant papers in the field. There are 52 countries involved in the 739 retrieved documents contributing to this research area. The top 10 cited countries according to the rank of total citations were performed in Table 1 . We can learn that United States (US) has the largest number of published articles (266 papers), accounting for 35.99%; and then China (148 papers, 20.02%) followed by Italy (86 papers, 11.63%) and the United Kingdom (UK) (73 papers, 9.87%). The number of citations of a country’s published literature is also an important factor to measure its influence on the field. US is still the top cited country with 20,225 times, followed by UK (8,377 citations), Italy (7,913 citations). Australia (7,338 citations), Germany (6,856 citations). However, in terms of total citations per production, Sweden (436 citations), Australia (236.7 citations), and Brazil (192.1 citations) took the top 3 countries, respectively. Obviously, the influence of US on this field is beyond doubt. Moreover, it is worth noting that although the publications of Italy, UK and Australia is less than 100, their publications are cited more frequently than those of China.

And then VOSviewer was used to conduct co-authorships analysis of countries with the threshold set as 3 documents. The size of nodes represents the quantity of literature published in each country, and the larger the node is, the more the country produced. The connection between nodes represents the cooperation between countries, and the thicker the line, the closer the cooperation between countries. It is apparent that the United States was the central country in this field, and has a close cooperation with China, Italy, UK, France, and Germany ( Figure 3 ). The total intensity of links of USA (212) is significantly higher than other countries. In addition, although China had a higher academic output, the total link intensity (69) was less than Italy (129) and UK (124).

Setting a threshold of 5 documents of an organization in the co-authorships analysis of organizations, and we got 91 of 1375. Figure 4 provides a network visualization map of the major organizations; Among them, the top 5 for publications were University of Texas MD Anderson Cancer Center (n=28), Harvard Medical School (n=26), Institute of Cancer Research (n=23), Memorial Sloan-Kettering Cancer Center (n=22) and Royal Marsden Hospital (n=21). However, the top 3 with most citations were Memorial Sloan-Kettering Cancer Center (6219 citations), University of Turin (4259 citations), and University of Cambridge (4085 citations). The top three in total link strength were Institute of Cancer Research (43), Royal Marsden Hospital (43), and Memorial Sloan-Kettering Cancer Center (39) ( Table 2 ). Notably, of the top 10 most productive organizations, five are from the United States, three are from the United Kingdom, and the other two are from Italy and Belgium.

The analysis of journals in related research fields can provide references for researchers to get better submission selection. There are 300 journals active in the field; By setting 5 as the minimum number of documents of a source, 35 of them meet the thresholds and the visualization of sources analysis was got ( Figure 5A ). Clinical Cancer Research got the largest number of publications (n=32), followed by Cancers (n=29), the third to fifth were Oncotarget (n=24), Breast Cancer Research and Treatment (n=17), and Frontiers in Oncology (n=17). The total number of articles published in the top 10 journals was 179, accounting for 24.22%. Science Translational Medicine had the most citations (5205 citations) with only 9 documents. And the next were Clinical Cancer Research (2489 citations), Journal of Clinical Oncology (2138 citations), Cancer Discovery (1761 citations) and Annals of Oncology (1297 citations). The h-index is an indicator reflecting the quality of scientific publications (21). In this study, Clinical Cancer Research ranked first with the h-index of 24, and followed by Oncotarget and Annals of Oncology, with the same h-index of 16, and Molecular Oncology (h-index = 11). Of the top 10 journals based on publications, Clinical Cancer Research was the first to start publishing articles in this field in 2012 and 70% of them started publishing in 2014-2015. ( Supplementary Table S1 ).

From the overlay visualization map of sources analysis in VOSviewer ( Figure 5B ), as the blue color represent the earlier average publication year while the yellow color represents the more recent average publication year, we found that Science Translational Medicine, Cancer Research, and Oncotarget have been active in this field at least 5 years ago. Moreover, although Cancers, and Frontiers in Oncology have become active in the past 2 years, the number of published articles has already ranked among the top 5.

There are 46 research categories and 70 publishers involved in the retrieved literature. The top 5 subject categories and publishers based on the number of publications are listed in Supplementary Table S2 . The most represented one is “Oncology” (n=429, 58.051%), accounting for more than half of all documents. Followed by “Research Experimental Medicine” (n=65, 8.796%) and “Cell Biology” (n=60, 8.119%). Regarding publishers, “Springer Nature” ranked the first (n=155, 20.974%), followed by “Elsevier” (n=113, 15.291%), “Amer Assoc Cancer Research” (n=57, 7.713%), and “Mdpi” (n=51, 6.901%).

Through statistical analysis, 5017 authors contributed to all the 739 documents, 15 among them were single-authored documents. The average number of authors per document was 6.78, and the average number of co-authors per document was 9.56. Cristofanilli M from US was the most prolific author (19 articles), followed by Turner NC, from UK (n=17) and Ma F, from US (n=16) ( Table 3 ). Ranked by h-index, Turner NC was the top 1 (h-index = 15), and then Garcia-murillas I (h-index = 13), Hhrebien S (h-index = 12), Cristofanilli M (h-index = 11). M-index is an indicator for “scientific quality” which was corrected for the time of scientific activity (21). Turner NC and Garcia-murillas I also occupied the top two places, with the m-index of 1.875 and 1.675, respectively. Cristofanilli M ranked No.3 (m-index = 1.571) and followed by Hrebien S (m-index = 1.500). In terms of the most cited authors in this fields, the results were a little different. Carlos C ranked first, with 7 published papers receiving 3723 citations, indicating that his literature has made a significant contribution to the development of the field. And it was followed by Nitzan R (3520 citations), and Sarah-Jane D (2405 citations), with publication count less than 10 articles, suggesting a great impact of each of their literature on this field.

Co-authorship analysis between authors was conducted using VOSviewer, and 79 co-authors wrote more than 5 articles. Preserving the largest set of 67 connected items and we got the Visualization Map ( Figure 6 ). Cristofanilli M had the most total link strength (100), and was followed by Ma F (99), Yi ZB (96), Xu BH (90), Guan XW (81). The co-citation analysis of cited authors is shown in Supplementary Figure S1 , where a total of 15 authors were cited more than 150 times. The top five cited authors are Dawson SJ (321), Diehl F (302), Bettegowda, C (275), Garcia-Murillas I (271), and Mmurtaza M (243).

We used bibliometrix to analyze the specific information of the top 10 documents in this field ( Supplementary Table S3 ). All these articles were all published between 2012 and 2018. 70% of them published before 2015. Among them, the most cited paper was “Detection of circulating tumor DNA in early- and late-stage human malignancies (Bettegowda, C 2014)” published in Science Translation Medicine Right, with the total citations 2655, analyzing the early detection of DNA in human malignancies (22). Dawson et al. published an article in The New England Journal of Medicine in 2013 titled “Analysis of Contaminated Tumor DNA to Monitor Breast Cancer”, with the second highest citations (1,488). In this article, it revealed that contaminated tumor DNA is a highly sensitive and informative biomarker of breast cancer (23). The third is “Liquid Biopsies: Genotyping Circulating Tumor DNA” published by Diaz, LA and Bardelli, A in the Journal of Clinical Oncology in 2014, receiving 1374 direct citations. This paper describes the sensitivity and accuracy of ctDNA based liquid biopsy to type the mutated genes of tumor cells, which is of great significance for clinical research (24). From the perspective of content, these three papers are related to the theoretical basis of ctDNA as a monitoring index for the detection of early breast cancer, which are representative and influential basic papers in this field and have far-reaching influence and significance for subsequent research.

From the co-citations analysis of cited references in VOSviewer, we can learn that 215 of the 24894 co-cited references had been cited at a minimum number of 20 times. In the network visualization, all the cited references are divided into four clusters and larger node represents more reference cited ( Figure 7A ). The top cluster with 96 items is shown in red, indicating the most interesting research area. The redder color means more citations in the density view ( Figure 7B ). Supplementary Table S4 listed the top ten most co-cited references. The top five cited references are Dawson SJ (2013; 310 citations), Bettegowda C (2014; 266 citations), Garcia-murillas I (2015; 228 citations), Diehl F (2008; 189 citations), Murtaza M (2013; 173 citations).

291 keywords were obtained to perform keyword analysis by setting the occurrence frequency to more than 5. With the function of clustering on VOSviewer, all keywords were divided into four clusters shown in different colors ( Supplementary Figure S2A ). The size of an item’s circle and label is proportional to its importance. The items in the same cluster with the same color mean higher correlation with each other (20). The red cluster and the green cluster are the two main clusters in this field. The red cluster contains 117 keywords, mainly including circulating tumor DNA, breast cancer, mutations, therapy, heterogeneity and so on. The green cluster contains 108 keywords, including cell-free DNA, plasma, acquired resistance, and so on. The blue cluster is the third largest cluster with 65 items, including liquid biopsy, survival, metastatic breast cancer (MBC), circulating tumor cell and so on. In the density visualization map, the color of an item depends on its frequency of occurrence, which is easy for us to pay attention to the most important areas in a map (20). The redder color means the heavier weight. As shown in Supplementary Figure S2B , the most frequent keywords excluding the search subjects were “liquid biopsy” (n=196), “plasma” (n=112), “mutations” (n=101), “metastatic breast cancer” (n=100), “acquired-resistance” (n=98) and so on. What’s more, the overlay visualization map showed the average publication year in which keywords appears in these publications ( Supplementary Figure S2C ). Yellow means the later it occurred in the articles. We can learn that the focus of research areas ranged from the expression of ctDNA genes in various cancers to the value of peripheral blood ctDNA as a biomarker and then to the diagnostic and prognostic value of ctDNA in BC.

The word or phrases that frequently appear in the title of an article’s cited reference were extracted and defined as keywords plus, which is used to identify research topics more descriptively and objectively (25). From the co-occurrence analysis of keywords plus analysis in VOSviewer ( Supplementary Figure S3A ), except from the search subjects, the top 5 with most cited keywords were “cell-free DNA” (n=140), “plasma” (n=105), “metastatic breast cancer” (n=100), “acquired-resistance” (n=98), and “mutations” (n=98). The overlay visualization map of keywords plus analysis showed the most recent keywords, including “liquid biopsy”, “cell-free DNA”, “ESR1 mutations”, “survival”, “resistance”, “double-blind”, “palbociclib”, and “fulvestrant”. ( Supplementary Figure S3B ).