The Web of Science Core Collection (WoSCC) database is widely used in bibliometric analysis (16–18). Aside from its comprehensive bibliometric software capabilities, we also chose it for its reputation as the most influential database when it comes to bibliometric software (19). Web of Science (WoS, Clarivate Analytics, Philadelphia, PA, United States), which contains more than 12,000 international academic journals, is one of the most comprehensive and authoritative database platforms to obtain global academic information. Apart from the general literature search, it also possesses an important function of citation index searching, which is helpful for assessing the academic performance of literature in a specific field (20, 21).

We retrieved the data from WoSCC database on August 2, 2022. The search strategy was (TS = [cardiotoxicity or cardiovascular toxicities or cardiac toxicity or Myocardial toxicity]), and the publication year was limited to 2010- 2022. There was no language limit, and the publication type was limited to only Article or Review. We downloaded all retrieved records together with their cited References in the format of plain text. The files were then transformed to text file with the name of “download *.txt” which could be recognized by Citespace software.

For bibliometric analysis and visualization, we used CiteSpace (version 5.8.R3 [Chaomei Chen, 2006]), VOSviewer (version 1.6.18 [Nees Jan van Eck and Ludo Waltman, 2010]) to analyze the data and create a visual representation of scientific knowledge. The annual publications were analyzed and managed using Microsoft Office Excel 2019. We derived the annual growth trend of publication outputs by analyzing the data and producing the figures using Microsoft Office Excel 2019.

As a commonly used visualization software in bibliometrics, CiteSpace is capable of detecting collaborations, hot spots, internal structures, possible trends, and evolutionary processes in scientific fields [22]. Therefore, CiteSpace was used to generate visual map of distribution of countries/regions and institutions, author, the dual-map of journals, reference timelines, and keyword citation bursts. We choose the following settings: time span is set to 2010–2022 and years per slice is 1; Top N = 50 filters the top 50 authors, organizations, and keywords with the highest frequency in each time slice; the pruning is set to Pathfinder. With the CiteSpace visualization, nodes are sized according to the frequency of co-occurrences, and links show the relationship between co-occurrences. The colors of the nodes and lines change from purple to red as the year 2010 to 2022 progresses. Purple round nodes have a strong centrality (≥0.10), which serves as a role of bridge interconnecting various networks through it (22–24).

As another bibliometric tool, VOSviewer can create and visualize knowledge maps, showing clusters, timeline, or density colors (25, 26). We used it to detect citation of country/region and institution collaboration, as well as co-cited authors, productive journals, and keywords. Data were analyzed using VOSviewer based on the full counting method. Cluster maps represented co-occurrence frequencies based on node size, and nodes of the same color were in the same cluster. In addition, the thickness of links indicated co-occurrence relations that were defined by their strength, which depends on how often two researchers co-authored papers or how frequently two keywords appeared together in papers (26). The co-citation frequency was positively correlated with word size, circular size, and yellow opacity in the density map. The overlay map displays the average publication year in different colors. Additionally, we obtained the impact factor (IF) and Journal citation reports (JCR) divisions of journals on August 2, 2022 from Web of Science.

Among the 8,077 papers we obtained from the WoSCC database, we eventually included 8,074, after removing duplicate papers. Overall, there was an increase in the number of publications each year. Considering Figure 1, the history of research can be divided into three stages: (i) 2010–2013: in the early stages, during which 400–500 publications were published annually. (i) 2014–2018: in the smooth growth stages, during which 500–800 publications were published annually. (iii) 2019–2022: Rapid development phase. The number of papers on cardiotoxicity exceeded 800 in 2020.

A total of 8,074 total publications (TP) were published from 124 countries or regions between 2010 and 2022, with 69% of the total publications coming from the top 10 countries(Table 1). Table 1, Figure 2A shows that most of the articles were from the United States (2,178 articles, 27%), followed by China (1,121 articles, 13.9%), Italy (559, 6.7%), India (338, 4.2%), and England (305, 3.8%). Among all countries, total citations(TC = 81,970) were highest in the United States, indicating its leadership in cardiovascular toxicology. In Figure 2B, the total number of citations for the global article is shown.

In cardiovascular toxicology, the United States led the way, working with Italy and England most closely. On the citation network map (Figure 2C), we visualized the citation relationships between countries/regions. The VOSviewer parameter was set to have a minimum of 50 documents per country, and we received 60 threshold. Clearly, the United States had the strongest total link strength (TLS = 10,594), with close ties to Italy, England, and Canada.

A total of 8,074 publications were published by 6,530 institutions. Table 2 shows the ten most productive institutions. Among them, The University of Texas MD Anderson Cancer Center (TP = 62) was the most prolific institution, followed by University of Toronto (TP = 56) and University of Pennsylvania (TP = 55). By using VOS viewer, we were able to visualize the network of institutions with more than 25 publications. The collaboration between institutions is shown in Figure 3A, containing 52 items grouped into five clusters of different colors. First placed was Vanderbilt University with 166 TLS, Harvard Medical School (TLS = 115) and Mayo Clinic (TLS = 106) ranked second and third, respectively. We then colored the institutions based on when they began to do research on cardiotoxicity, the bluer their colors, the earlier they started and the yellower the color, the later the institution appeared. Harvard Medical School and The University of Texas MD Anderson Cancer Center were pioneers in the field, as illustrated in Figure 3B.

The number of authors was 39,071, and there were 8,074 publications that they had co-authored. Table 3 shows that Zhang, Yun had authored the most publications (n = 53), followed by Wang, Yue (n = 42), Li Ying (n = 38), and Sun, Zhiwei (n = 35) and Zhang, Jing (n = 34). There was a lack of centrality among the top 10 authors. None were more significant than 0.10. Figure 4 shows a certain degree of collaboration between the authors. The circles represent different authors, and the lines between them represent collaboration between them; different colors represent different years, and thicker lines indicate closer cooperation.

Co-cited authors are authors whose papers are cited by one or more other publications at the same time. There were more than 900 citations among the top 10 co-cited authors, as shown in Table 3. Moslehi, Javid (n = 2,230) and Ky, Bonnie (n = 1,375) were the most frequently co-cited authors, followed by Vanholder, Raymond (n = 1,362), Abdollahi, Mohammad (n = 1,251) and Lyon, Alexander R (n = 1,112). The minimum citation threshold of VOS viewer was set to 10 to produce a map with 60 nodes and 6 clusters (Figure 4B). Table 3, Figure 4B shows that co-cited authors with high TLS play an essential bridge role.

In order to find the journals with the most co-citations and published papers, we used the VOS viewer and Citespace software. There were 8,074 papers published in 2,392 academic journals, according to the results. The most published articles are Plos One (n = 93), followed by Cardiovascular Toxicology (n = 81), Frontiers in pharmacology (n = 76), Clinical Toxicology (n = 68), and Journal of Ethnopharmacology (n = 68), as shown in Table 4. There were nine journals in the top 10 that published more than 50 papers, seven of which were in the Q1 category of Journal Citation Reports (JCR). Furthermore, Chemosphere (IF = 8.94) has the highest impact factor (IF) of all these journals.

According to Table 5, the most frequently cited journals were the New England Journal of Medicine (n = 10,674), followed by Journal Of Clinical Oncology (n = 10,063), Journal of Circulation (n = 9,263), Journal of The American College of Cardiology (n = 5,501), and Lancet (n = 5,404). There were six journals with more than 5,000 citations among the top 10 co-cited journals, and New England Journal of Medicine had the most citations overall. The journals with the highest IF were the New England Journal of Medicine (IF = 176.079) among them, followed by Lancet (IF = 202.731), Journal of the American Medical Asociation (IF = 157.335), Journal of Clinical Oncology (IF = 50.717), and Circulation (IF = 39.918). Figure 5A shows the citation patterns of five journals, divided into 4 clusters with 399 links. In Figure 5B, which contains 1,000 items and 7 clusters, the co-citation relationship among different journals is visualized.

The dual-map overlay of journals shows how journals are distributed and how they relate to cited journals (the color path shows the relationship between journals and cited journals). Linked journal citations are displayed on the left, followed by cited journals on the right, with colored paths indicating relationships among the journals. Five main reference paths are shown in Figure 5C. It showed that studies published in “Environment, Toxicology, nutrition”, “Molecular, Biology, Genetics” and “Health, Nursing, Medicine” journals were frequently referenced in articles published in “Molecular, Biology, Immunology” journals and “Medicine, medical, Clinical” journals.

In order to identify the most popular hot spots in a specific field of study, keywords are often used to indicate their frequency of co-occurrence, and the approach is critical for tracking scientific progress. There were 29,533 keywords extracted from 8,074 papers in total. Co-occurrence frequency for the top 10 keywords from WoSCC is shown in Table 6. Using 150 keywords with more than 20 co-occurrences, we created a density map, with which hot topics can be intuitively visualized in the field as a whole. According to Figure 6B, “toxicity”, “oxidative stress”, “cardiovascular disease” and “cardiotoxicity” had the hottest color in the density map, indicating that they were the most prevalent elements.

A cluster analysis was performed on the selected keywords. The index words derived from the keywords were used to mark all clusters. In Figure 6A, 150 high-frequency keywords are grouped into three clusters, which are representative of the three dominant lines of research in the field. With 56 keywords marked in red circles, cluster 1 is the largest cluster, focusing on mechanisms of cardiotoxicity, including “toxicity”, “oxidative stress” and “inflammation”. There are 52 keywords in cluster 2, marked with blue circles, which focus on cardiotoxicity diseases. Among keywords were “heart-failure”, “cancer” and “breast-cancer”. A green circle was drawn around cluster 3, which included 42 keywords relating to risk factors, including “overdose”, “blood pressure”, and “therapy”.

Additionally, we detected keywords with the most citation bursts, which can indicate trends or frontiers in research in recent years. Among them, Figure 10 shows the top 25 keywords. As shown in Figure 6C, the following is a list of keywords that experienced a citation burst after 2017: stem cell (7.87), consumption (9.95), sliver nanoparticle (7.82), diagnosis (8.11), immune checkpoint inhibitor (18.93) and myocarditi (9.91), which were the frontiers of cardiotoxicity.

Figure 7A shows the top 10 most cited references (27–36) over the course of the study, which we analyzed further to understand the development trends. As shown in Table 7, there were co-cited at least 70 times among the top 10 co-cited references. The 2016 ESC position paper on cancer treatments and cardiovascular toxicity, which received 191 citations, was among the most deeply cited. In addition, there are nine research articles and one review article among the top 10 articles. Furthermore, it reflects the fact that professional association guidelines are commonly quoted papers within a discipline because they are closely related to medical practice. Commonly, citation rates are correlated with the time of publication, so current citation counts may not match their current value.

In the references timeline view, you can see how research hotspots evolved over time. Cluster labels were determined by identifying the terms that appeared most frequently in each cluster. As shown in Figure 7B, The cluster 0 (kinase inhibitor) and cluster 12 (sorafenib) appeared first, suggesting that clinical medicine was the primary focus early on. The cluster 7 (lung cancer), cluster 10 (myocarditis) and cluster 11 (immune-checkpoints) occurred in 2010–2019. Cluster 3 (cardio-oncology) appeared in 2014–2022. This indicates that cardio-oncology is a popular topic of current research, suggesting that greater attention is being paid to this area.

This bibliometric analysis of the 8,074 cardiotoxicity-related documents in the WOSCC database for the past 12 years revealed current status, research frontiers and trends in the cardiotoxicity field. Research on cardiotoxicity was in a state of rapid development. Over the past 12 years, the number of publications on cardiotoxicity steadily increased worldwide. There have already been 496 papers published in 2022 so far, which has almost reached half of last year's total, and the outputs were expected to hit a new all-time high. Another obvious advantage of bibliometric analysis provides the ability to visualize global trends from countries/regions, institutions, authors, and journals over time (37).

Table 1 and Figure 2 showed that the United States was the world leader country in cardiotoxicity research, which had accumulated the highest number of critical citations. Noticeably, despite its late start, China is now one of the top productive countries in the world. Besides, Figure 2 showed that active collaborations among different countries, indicating cardiotoxicity had gained worldwide interest, and the United States was the key collaborating center. Italy and England showed the considerable increases in output in recent years, collaborating most closely with the United States.

In our analysis, the top 10 productive institutions were from four different countries, and 7 (70%) were in the United States, with the top 2 institutions respectively from the United States and Canada. In 2000, the University of Texas MD Anderson Center, a famous pioneer, established the first cardio-oncology unit (38), and it is the most prolific institution publishing academic research related to cardiotoxicity. This means that cardiotoxicity is more likely to be noticed by oncologists, mainly due to the treatment of patients with cancer. Nevertheless, as an interdisciplinary field, cardiotoxicity requires more multidisciplinary collaboration and treatment. Moreover, we found no oncology organization among actively collaborating institutions, including the Vanderbilt University, Harvard Medical School, and Mayo Clinic. To prevent incidence of cardiotoxicity, countries and institutions may need to strengthen their cooperation in cardiotoxicity research to further advance the development of cardio-oncology.

By highlighting the contributions of the most productive researchers in a given field, researchers in this field could be inspired by understanding these pioneer researcher's work (39). The top 5 productive authors were from China, while top 5 co-cited authors were all from other countries. Furthermore, we found three researchers, namely Sun, Zhiwei, Lyon, Alexander R., and Duan, Junchao were among both the top 10 most productive authors and the top 10 most co-cited authors, implying that these three authors had an outstanding contribution to cardiotoxicity field.

In this analysis, we found that in 2017, the most co-cited study was published (n = 191) by European Heart Journal co-authored by Jose Luis Zamorano and 17 other outstanding researchers in cardiotoxicity research (27). This practice guideline focuses on cardio-oncology, calling for the awareness and recognition of cardiotoxic side effects during cancer treatment. Since then, countries around the world have gradually begun to recognize and paid attention to the cardiotoxicity related to tumor therapy. Johnson et al. published the second most co-cited paper in the New England Journal of Medicine in 2017 (28). The study reported two cases of melanoma patients who developed fatal myocarditis following combination immunotherapy. Characterizing these severe cardiotoxicity is a major priority, despite the durable antitumor effect produced in cancer patients by combination immune checkpoint inhibition. The mechanism behind its cardiotoxicity has become a research hotspot. Following the European Society of Cardiology Committee, in 2017, the American Society of Cardio-oncology released a similar clinical practice guideline for prevention and monitoring of cardiotoxicity in survivors of adult-onset cancers, which is the third co-cited publication (29). Subsequently, cardio-oncology was well recognized as a new area in the US and most European countries. Daniela Cardinale et al. published the fourth co-cited article in Circulation in 2015 (30). This study prospectively evaluated the incidence, timing, clinical relevance, and response to heart failure therapy of cardiotoxicity following anthracycline-containing therapy and demonstrated that early detection and prompt treatment of cardiotoxicity are critical for substantial recovery of cardiac function. Juan Carlos Plana et al. published the fifth co-cited articles in 2014 (31). This review introduced Definition of Cancer Therapeutics–Related Cardiac Dysfunction (CTRCD), and summarizes the recommended cardio-oncology-echocardiogram protocol.

Overall, the top 5 co-cited references focus on field of cardio-oncology. With the tremendous development of cardio-oncology, more attentions are paid in detection, monitoring, prevention and therapy of cardiovascular diseases(CVDs) occurring in the context of cancer treatment (40). Meanwhile, immune-based therapies have revolutionized cancer treatments, however, the serious and fatal cardiotoxic effects, representing new challenges in cardio-oncology (41).

The top 5 papers were published in 2014–2017 (the top 3 in 2017). This is a critical period for the start and development of cardio-oncology, which suggests that cardiovascular toxicities have become research hotspot and are attracting increasing attention. This year, the International Cardio-Oncology Society published consensus statement seeking to provide a reliance for defining cardiovascular toxicities of cancer therapy including cardiomyopathy/heart failure, myocarditis, vascular toxicity, hypertension, as well as arrhythmias and QTc prolongation (42). It means that tumor cardiotoxicity has become an international public topic, and we believe it will be a high co-cited study. These results of highly co-cited references may be explained by the fact that advances in treatment have led to improved survival for cancer patients.

The cluster of keywords could fully demonstrate frontiers of academic research, and every cluster reflected a different research topic (48). The cardiotoxicity field is divided into three main clusters (Figure 6) by cluster analysis, including the mechanisms, the related disease, and risk factors of cardiovascular toxicities, representing three Emerging aspects of cardiotoxicity research.

Advances in antineoplastic agents research have led to huge improvements in malignant tumors survival rates. Meanwhile, cancer treatments, including cytotoxic chemotherapy, molecular target therapy, and mediastinal site radiotherapy, are thought to be linked with cardiomyocyte ischemia, damage, conduction and rhythm disturbances, ventricular dysfunction, cardiac failure, and several other cardiovascular complications (57). The mechanisms of toxicities caused by antitumor agents on the cardiovascular system are diverse and not fully understood, including direct effects on the heart or vascular system, release of cardiac regulators, oxidative stress and changes in blood coagulation status. Therefore, in this era of treatment, new models for multidisciplinary support team including cardiologists and oncologists are required and attention should be paid to the regular monitoring of patients' heart function during tumor therapy to help mitigate associated cardiovascular complications (32).

However, the available options for cardiac protection are minimal, and current medications to combat chemotherapy-induced side effects generally fail to address the underlying cause and create distress for patients undergoing therapy. In the future we need to discover the different mechanisms of action of chemotherapeutic agents so that the toxicities can be circumvented along with improved clinical experience for patients.