The Web of Science Core Collection (WOSCC) is a high-quality, up-to-date, error-free database of more than 12,000 of the most influential and valuable scientific journals. All relevant references have been searched and updated on September 18, 2023, in the Science Citation Index Expanded/SCI-E database in the WOSCC, which is most suitable for bibliometric analysis (17, 18). Radiotherapy (#1) and cardiac toxicity-related (#2) terms are mainly used as descriptors. A Boolean algorithm consisting of “#1 AND #2” is used to ensure that all items retrieved fall within the scope of RIHD. The specific search query is illustrated in Figure 1. The study covers research from 1991–2023, with 1612 papers obtained from the WOSCC SCI-E database, including all clinical and preclinical papers. Previous studies have shown that the inclusion of English publications can ensure the accuracy of the analysis (11). Therefore non-English papers are excluded. With the literature types being screened, letters (n = 4), bulletins (n = 2), and book chapters (n = 1) are excluded, which may be non-evidence-based and fallacious (12). Finally, a total of 1,554 publications are included in the final bibliometric and visual analysis. In conclusion, the relevant data in the form of plain text format from WOSCC are exported, including title, author, year of publication, country, institution, keywords, citations, abstracts, and references.

The study group adopted Microsoft Excel 2020 to draw the annual publication graph and a chart of the top 10 countries in terms of the number of publications, and so on. The group also used bibliometrics analysis software (VOSviewer 1.6.11, CiteSpace 6.2. R4, and Bibliometrix) to analyze and visualize the data of the above 1554 documents.

CiteSpace is a free Java app developed by Professor Chaomei Chen. It detects and visualizes trends and patterns in scientific papers through a progressive knowledge domain visualization methodology that visualizes maps of highly cited and critical documents within knowledge domains (19). CiteSpace produces a number of important metrics. The centrality based on structural void theory measures the number of times a node is located on the shortest path between other nodes. Nodes with higher centrality are considered as critical hubs and play important roles in the information conversion process (20). The clustering method of CiteSpace is spectral clustering, which uses each vertex in the graph as a cluster and merges the clusters by calculating the similarity between different vertices (21). In the clustering analysis, the Q-score measures the extent to which the network can be divided into clusters, and the S-score can verify the internal consistency of data clustering. The closer these two scores are to +1, the more effective the clustering is (22). The burstness demonstrates a specific duration when a sudden change in the frequency of an element occurs, thus identifying emerging terms, which can represent trends and frontiers to some extent (23). In the graphs drawn by CiteSpace, each node represents one element (author, country, institution, or keyword), and the size of the nodes represents the number of publications (author, country, institution) or frequency of occurrence (keyword). In addition, the connecting lines between the nodes represent co-occurrence or co-citation relationships, the thickness of the lines represents the thickness of the correlations, the red nodes represent nodes that are in hot states in the study, and nodes in the purple outer circle have high centrality (centrality > 0.1) (9). With Citespace, the study group drew the cooperation map of the country, the co-cited reference and Keywords time zone diagram, and the references and keywords burst maps to realize the visualization analysis of research status, hotspots, and frontiers.

VOSviewer is a software tool for creating and browsing maps based on web data, primarily for analyzing scholarly records (24). The clustering algorithm on which VOSviewer is based is VOS clustering, which is similar to the Modularity network clustering algorithm (25). In the graphs drawn by VOSviewer, different nodes represent different elements (co-cited references, journals, authors), and the size of the nodes is proportional to the number of publications or the frequency of keywords. Lines between the nodes indicate collaborations or co-citation relationships. Different colored nodes and lines represent different clusters or years (26). With VOSviewer, the study group has implemented a visual analysis of author collaborations and a keywords co-occurrence map.

The study group then used the bibliometric R-package (https://www.bibliometrix.org/home/) to import the data from WOSCC exported to Rstudio software. Then, the impact factor (IF), H-index, and co-citation times of authors and journals in 2023 were obtained. Bibliometrix and Biblioshiny are open-source software packages for use in the R environment. Bibliometrix performs the entire scientific data analysis process, while Biblioshiny enables users to create visualizations based on an interactive web interface (27). Since no patients have participated in this study and the data are all obtained from WOSCC, informed consent and ethical approval are not required. The specific research process is as follows in Figure 1.

Co-cited references are references cited by more than one publication. Co-cited reference maps are generated with corresponding clusters, which can mention landmark references and research clusters and reflect the trends and frontiers of research in a particular field (28). The retrieved original documents are categorized into 15 clusters using cluster analysis, as shown in Figure 3A. Notably, these clusters are well structured (Q = 0.8424) and highly credible (S = 0.9118). It is observed that the publications in each cluster are closely related and coordinated within a particular domain. The cluster labeled “contemporary view” (cluster #0) takes the largest portion, followed by “non-small cell lung cancer” (cluster #1), “cancer therapy cardiotoxicity” (cluster #2), “thoracic radiation” (cluster #3), and “hodgkin lymphoma” (cluster #4). The latest research on RIHD has focused on thoracic cancer, including breast cancer, NSCLC, and Hodgkin’s lymphoma. The purple clusters on the left represent the foundation of the discipline from a certain point of view, including “preclinical studies toxicity (cluster #11), “pediatric Hodgkins disease” (cluster #9), and “cardiotoxicity” (cluster #12). Other important clusters, such as “heart dose” (cluster #8), “breast cancer patient “(cluster #14), “cardiovascular care,” and “childhood cancer survivors” (cluster #13), may represent shifts or breakthroughs in this field.

After sorting out the top 10 cited references on RIHD, it is found that, as shown in Table 2, all these documents have been published in high-quality journals. Among them, six publications are related to breast cancer, two are on NSCLC, one is on Hodgkin lymphoma, and one is a review of this field, which fully demonstrates the close relationship and hot position of thoracic radiotherapy, especially breast cancer in this field. Risk of Ischemic Heart Disease in Women After Radiotherapy for Breast Cancer is the most frequently cited reference, in keeping with the fact that the largest node in the co-cited reference map represents Darby SC (2013) (29). The results of this study can often be seen in RIHD-related papers. However, highly cited publications were published earlier than the late, high-quality ones, possibly causing the latter to appear later in low citation. The map of the top 25 references with the strongest citation bursts can make up for this deficiency. As shown in Figure 3B, you can see the most recently cited papers. The time cutoff is 2023. They are Piroth MD, 2019 (strength: 11.21, time span: 2021-2023) (30), Taunk NK, 2015 (strength: 11.58, time span: 2020-2023) (31), Atkins KM, 2019 (strength: 14.78, time span: 2020-2023) (32), etc.

This study demonstrates the publishing level in the field of RIHD, the collaboration among countries, authors, and journals, as well as the present research focus and trends through bibliometrics and visualization analysis. From 1991 to 2023, a scientific search yielded 1554 relevant works authored by 8764 individuals from 70 countries and published in 433 academic journals.

Prior to the year of the first analysis paper in this study, only 2 relevant documents were retrieved from WOSCC. The paper published in 1979 indicated that radiotherapy might enhance the cardiotoxicity of adriamycin in patients with stage IV breast cancer, and the other published in 1981 made the initial attempt to describe cardiotoxicity after chemotherapy and radiotherapy (35, 36). The annual publication volume serves as a significant metric for assessing the robustness and expansion of RIHD during a specific time frame. Prior to 2013, the annual output had increased gradually but consistently, with a general upward trend. The study conducted by Darby SC in 2013 subsequently accelerated the rate of subsequent annual publications. It is the most cited publication and is regarded as a significant advance in RIHD research. A novel linear correlation between radiation dose and cardiotoxicity was established in this nearly 50-year-long multicenter cohort study conducted in Sweden and Denmark: the mean heart dose (MHD) was found to be linearly correlated with the incidence of major coronary events in breast cancer patients. The increase in incidence was 7.4% per gray (95% confidence interval: 2.9% to 14.5%; P<0.001). Cardiac events manifested as long-term effects and women with cardiac risk factors had a higher risk of developing RIHD after radiation. These findings served as the foundation for subsequent investigations (29). Furthermore, it is worth noting that around 71% of the extracted publications were published subsequent to 2013, indicating that RIHD has gathered a growing level of scholarly attention in recent times. The analysis of country and author shows that the United States has contributed the largest number of publications and the largest number of high-quality authors and institutions, with Marks LB being the author with the highest H-index. Notwithstanding its dominant position, the United States does not possess a “monopoly” in this field. China, the only developing nation to rank among the top ten nations in terms of publications, is presently ranked second in terms of the greatest quantity of RIHD research outputs. While Italy may not have produced the most publications, it has demonstrated the most intimate collaboration with other nations that have the highest centrality. Countries and authors have collectively established an intimate network of collaboration and endeavored to advance the progress of this domain, thereby unequivocally exposing RIHD as a worldwide predicament. The most productive and influential journal is INTERNATIONAL JOURNAL OF RADIATION ONCOLOGY BIOLOGY PHYSICS. By diligently monitoring these authors, and journals, researchers can remain informed about the most recent developments in the respective field.

Over the past 40 years, research has increased our understanding of the mechanism of RIHD. Acute and persistent inflammation, endothelial dysfunction, and reactive oxygen species (ROS) all contribute to chronic fibrosis of the myocardium, which is a critical component of RIHD (37). Oxidative stress is the most recent area of this field. Radiation breaks down the respiratory chain of mitochondrial metabolism in old endothelial cells. This makes mitochondria more permeable and creates a lot of ROS (1). When ROS exceeds the capacity of intracellular antioxidants, large molecules such as DNA, proteins, and lipids are damaged. Moreover, ROS can initiate crucial mechanisms that result in apoptosis and necrosis, including mitochondrial permeability transition and calcium release. Excess ROS mediates a range of molecular signaling pathways, promoting the release of related cell factors leading to inflammation and fibrosis in RIHD, such as tumor necrosis factor, transforming growth factor-β, IL-4, nuclear factor-kappa β, etc (38, 39).

Research on RIHD has primarily been carried out in animal models, contributing to the understanding of the biological pathways and mechanisms involved in normal tissue toxicity in radiotherapy, identifying potential therapeutic targets at the cellular and molecular levels, and translating to clinical studies to better determine radiation dose (40). Currently, there is a trend to evolve from whole thorax to whole heart to partial heart irradiation in animal models used for RIHD (41). Whole thorax irradiation is a method that has been widely used in the past and does not require precise image guidance during radiation (42). However, its radiation field includes both the lungs and the heart, making it difficult to determine that the damage is solely from cardiac irradiation (43). Whole heart irradiation is a relatively new technique that requires imaging to accurately deliver radiation to the heart using a beam size that is focused primarily on the heart, thus limiting the lung dose (44). MHD is an important measure for evaluating the level of radiation exposure to the heart in cardiac procedure (29). However, more and more studies suggest that there is a closer relationship between RIHD and cardiac substructures receiving high dose of radiation, among which the dose of coronary artery is considered to be an independent factor in the assessment of RIHD and has a strong association with risk of ischemic heart disease (45). Therefore, there is a need to enhance preclinical models of the substructures of the heart, including key components such as the coronary artery, in order to accurately predict risk factors and successfully develop interventions for RIHD. Radiotherapy is frequently used alongside surgery, chemotherapy, hormones, and immunotherapy in cancer treatment, all of which may lead to cardiotoxicity (42, 46, 47). As immune checkpoint inhibitors are increasingly used in combination with radiotherapy in the treatment of NSCIL, the cardiotoxicity associated with them, which is relatively rare but has a high mortality rate, is of concern. Chemotherapy is an independent risk factor for cardiotoxicity, with medicines like doxorubicin, 5-fluorouracil, and anthracycline being recognized for causing cardiotoxicity (47). An analysis of the SEER database revealed a notable rise in the occurrence of cardiac events in patients who received a substantial cumulative dosage of anthracyclines along with adjuvant radiotherapy (RR: 1.26; 95% CI: 1.12-1.42) (48). A study using a New Zealand white rabbit model showed that complete cardiac irradiation plus adriamycin had a synergistic impact on cardiotoxicity (49). Animal models can be utilized to investigate the interactions between various medicines and their impact on short and long-term cardiovascular outcomes, as well as to enhance the timing and order of combined treatments. Preclinical animal models of mice, rats, rabbits, dogs, pigs, and nonhuman primates are often used in RIHD research (50). The rodent is a valuable study model that provides insights into many proteins and processes related to RIHD. The rabbit shares similarities with humans in cardiovascular physiology, particularly in ion channel and calcium transporter function (51). In pigs, the coronary circulation is comparable to that of young individuals, but in dogs, it resembles that of older individuals with ischemic heart disease (52, 53). Yet, rodents are restricted by their substantial anatomical disparities from humans, whereas rabbits, dogs, and primates are restricted by financial and moral considerations. Additionally, factors such as animal age, size, and sex must be meticulously taken into account when choosing models, as they can significantly impact experimental outcomes (41, 50).

Analysis of keywords shows that RIHD-associated cancer types are mainly referred to thoracic cancer, including breast cancer, NSCLC and Hodgkin’s lymphoma, which most of the latest trials are based on (4). Many publications on RIHD are based on breast cancer and have resulted in numerous established findings in the field. As the majority of breast cancer patients are female, “women” is a common keyword. The linear relationship between cardiotoxicity and radiation dose was first put forward in breast cancer and a similar linear relationship was also observed in Hodgkin’s lymphoma (29, 54). The RIHD of breast cancer has a significant laterality, which means a significantly increased risk of cardiovascular death after radiation in patients with left-sided breast cancer happens (55). It was previously thought that heart exposure to radiation was not important in NSCLC due to the short survival time and the well-known risks of pneumonia and esophagitis. But recent investigations have increasingly cast doubt on this perspective. The 74GY group in the randomized, phase III study of RTOG0617 showed significantly poorer overall survival compared to the 60GY group, with 20.3 months against 28.7 months, respectively (p = 0.004). Researchers analyzed several factors and found that the cardiac dose of the 74GY group was significantly higher than that of the 60GY group, which likely accounts for this phenomenon (56). Available studies suggest that NSCLC patients may experience RIHD up to two years after radiation exposure (57). Hodgkin’s lymphoma is common in young people and typically has a favorable prognosis, leading to survivors being consistently at risk of RIHD. Hodgkin’s lymphoma exemplifies change in radiotherapy plan. In order to minimize cardiac radiation, Hodgkin’s lymphoma radiation ranges from the entire field to the affected nodules, with the dose decreased from 44 GY to 20 GY (58, 59). Radiotherapy in conjunction with chemotherapy is now the predominant therapeutic approach for Hodgkin’s lymphoma, leading to a significant decrease in the occurrence of cardiac events (60).

Research on the risk factors associated with RIHD has consistently been the primary emphasis, which can be categorized as radiation-related or patient-related. Risk factors associated with radiation include total radiation dosage, average cardiac dose, primary cancer location, heart volume exposed to radiation, and whether shielding is used during therapy (3). MHD is a crucial sign for evaluating cardiac exposure. Recent studies indicate that targeting radiation to specific cardiac substructures is better predictive of cardiotoxicity (45). Further research is required to determine the most accurate metrics for predicting cardiotoxicity. RIHD can be influenced by inherent risk factors in patients, such as traditional cardiovascular risk factors (age >65 years, diabetes mellitus, smoking, and obesity), high blood pressure, sedentary lifestyle, history of cardiovascular disease, and young age at the time of radiotherapy (61). The incidence of RIHD is higher than that of cancer survivors without these risk factors, as demonstrated by a high-quality cohort Dutch study (62). RIHD following radiation at a young age has distinct characteristics among these risk factors. Thanks to modern medical advances, the long-term survival rate for childhood cancer has improved dramatically. However, the risk for RIHD is higher in children from the onset of radiation because there is more time for classical cardiovascular risk factors to act simultaneously on children (63). Moreover, children’s underdeveloped cardiovascular tissue may be more vulnerable to radiation exposure (62). Mulrooney and his colleagues conducted a study on 14,358 children who underwent radiation between 1970 and 1986. They compared these youngsters with their siblings to investigate the risk of developing heart disease later in life. The cumulative incidence of RIHD increased by 5–6 times compared to their siblings over time (64). Cardiovascular events are the primary cause of non-cancer mortality in childhood cancer survivors, indicating a need for additional study in this population.

Cardiovascular events after radiotherapy mainly include coronary artery disease, valvular disease, cardiomyopathy, pericardial disease, and conduction system abnormalities. These cases differ in duration and characteristics but typically result in heart failure and mortality (5). Coronary artery disease is the most prevalent kind of heart damage. The condition may manifest as early symptoms of acute coronary syndrome even sudden cardiac death, but typically occurs around 15 years following exposure to radiation (65). The lesions occur mainly at the opening or proximal part of the coronary arteries, especially the left anterior descending branch of the heart, which is highly susceptible to radiation (66). Imaging tools such as echocardiography, cardiovascular computed tomography, cardiovascular magnetic resonance (CMR), and nuclear cardiology aid in the early detection of cardiac events. Each of these tools has its specific advantages (3). Echocardiography is the predominant and initial imaging technique due to its low risk and easy accessibility. This method is commonly utilized to evaluate the heart’s systolic and diastolic functioning and is more effective in identifying valvular and pericardial disorders (63). Cardiovascular computed tomography is particularly advantageous for detecting coronary stenosis and determining the amount of calcium in the coronary arteries, as well as the presence of soft plaque burdens (67). CMR is quite accurate in detecting and diagnosing ischemic heart disease (68). Using the serum marker is a promising method for identifying and tracking RIHD in clinical practice. Currently, the relationship between conventional cardiac markers such as troponin T, BNP, and NT-pro BNP and radiation is uncertain (69). Markers of vascular endothelial cell inflammation such as placental growth factor (PlGF), tumor necrosis factor α, IL-2, IL-6, and IL-10 have been observed to rise in mice models following radiation in preclinical studies. Notably, there is a strong correlation between PlGF levels and MHD, with a correlation coefficient of 0.89 (70). Additional preclinical and clinical trials are required to confirm their effectiveness (NCT04305613, NCT03978377).

To enhance the long-term cardiac outcomes of cancer survivors, it is crucial to use technologies rationally to identify at-risk individuals before and after therapy, ensuring reliable and early detection of anomalies and prompting intervention to prevent poor outcomes (71). Preventing traditional cardiovascular risk factors is beneficial, including educating patients about maintaining a healthy lifestyle, quitting smoking, and lowering obesity, hypertension, and diabetes. An investigation by SEER-Medicare on breast cancer patients found that the absolute risk of cardiac death in the left breast decreased from 13% (1973–1979) to 9.5% (1980–1984) to 5.8% (1985–1989), possibly due to improvements in radiation dose reduction (72). Some ways for reducing radiation dose have been explored. Several straightforward physical techniques, including deep inspiratory breath-holding (DIBH), breathing gating, and posture change, can effectively decrease the cardiac radiation dosage (73–75). DIBH requires patients to take deep breaths during treatment to shift the heart away from the front chest wall and the expanded lung tissue in the middle to reduce the cardiac dose. MHD and the dose to the anterior descending branch of the left coronary artery both decrease notably. However, patients should be instructed carefully to enhance their compliance (73). Progress in radiation technology has been crucial in decreasing radiation exposure to the heart. Advances in radiotherapy techniques, from 2D to 3D conformal radiation therapy (3DCRT) to rotational methods like IMRT and VMAT to PT, have significantly decreased radiation exposure to nearby organs at risk and expanded the treatments’ capability to encompass intricate anatomical structures (3, 76–78). Prior to the advent of CT in the 1980s, radiation therapy used 2D radiographic film planning, making it challenging to prevent and measure heart radiation. Following the extensive utilization of CT, 3DCRT allowed for quantification of the cardiac dose, but cardiac avoidance often resulted in underdose to the target area. IMRT and VMAT, developed since the 1990s, were able to better modulate the dose from each beam through multileaf collimation thus allowing for a more conformal delivery of radiation (79). IMRT has become a standard of care radiotherapy technique for NSCLC and is also widely used to treat patients with left-sided breast cancer (80). A study on left-sided breast cancer patients who had radiotherapy found that the VMAT group had a considerably lower coronary heart dosage compared to the 3DCRT group (81). PT is the latest treatment that uses proton particles instead of traditional X-rays to treat cancer targets, keeping the radiation dose to the heart at a low level and enhance the lethality of cancer cells (78). PT uses the Bragg peak phenomenon to deliver the highest dose at a precise tissue depth, ensuring the tumor receives the intended dose while sparing normal tissue from exposure, hence reducing toxicity to healthy tissue (82). In a phase IIB randomized controlled trial, patients with locally advanced esophageal cancer were compared PT and IMRT. The study revealed that patients treated with PT had a significantly lower MHD compared to those in the IMRT group (19.8 vs. 11.3 GY; P < 0.001). During the follow-up period, the PT group experienced fewer cardiac incidents compared to the IMRT group (83). PT has gained more attention recently for reducing cardiac radiation dose and has been studied in several malignancies, including breast cancer and Hodgkin’s lymphoma. As RIHD is closely linked to the amount of radiation the heart receives, it may be deduced that the decrease in cardiac radiation dose from PT helps reduce RIHD. Nevertheless, assessing the therapeutic impact of this may necessitate an extended follow-up period, as indicated by the following clinical trial registrations: NCT02603341, NCT04361240, NCT04305613, and NCT01993810. Aspirin, statins, and colchicine, which are anti-inflammatory medications, have been commonly used to avoid negative cardiovascular events, and some preclinical research indicates that these treatments may also help prevent RIHD (5, 84). However, there is little clinical trial data on these radiation mitigants, and there are no existing guidelines on the use of anti-inflammatory medications for the prevention of RIHD. Anti-inflammatory drugs show tremendous potential in the field of RIHD (85).

Radiotherapy techniques have improved considerably over the past 40 years, and the radiation dose is well controlled, so the credibility of old research data has been doubted. The probability of RIHD increases with increasing radiation dose and guidelines state that limiting MHD is a benefit for cancer patients (86). However, as radiation technology evolves in a more conformal direction, such as IMRT and PT, some studies have shown the radiation dose to substructures is a better predictor of the risk of developing RIHD than MHD, so is it more beneficial for researchers to limit the radiation dose to substructures? (45). Animal models aid in the study of heart substructure, and various institutions have developed cardiac contour atlases to aid in substructure delineation. Several cardiac contour atlases have been established and developed, such as Duane, Feng, and Kong’s atlases (87–89). Preclinical trials and animal models also help to elucidate the pathways and mechanisms of RIHD and explore potential therapeutic targets. Anti-inflammatory drugs such as statins, aspirin, and colchicine may be beneficial in preventing and mitigating RIHD, and some animal models have explored the mechanisms involved, but more preclinical trials and high-quality clinical trials are required to back it up (85). Multimodal therapies such as surgery plus chemotherapy and radiotherapy are recognized as cancer treatments, and it needs to reconsider the potential synergistic cardiac-related adverse events where animal models can play a huge role (41). However, the patient situation in the real world is complex and varied, and so far there is no preclinical model that can perfectly reproduce the complex physiopathology of RIHD. The development of a reliable and effective preclinical model that better simulates the real-world situation is particularly important (90). The serum marker can reflect RIHD to a certain extent, but the mechanism and validity of these markers cannot be determined. Some prospective studies are underway (69, 70). More high-quality prospective studies examining advanced radiotherapy techniques such as PT and long-term monitoring of follow-up data are needed to understand whether these advanced techniques can ensure the efficacy of radiotherapy while reducing RIHD and mortality, as strict cardiac dose restriction may compromise the effectiveness of radiotherapy in the treatment of cancer patients (91). In order to optimize the long-term survival rate of cancer patients with RIHD, the department of cardiac-oncology has been established, and oncology and cardiology organizations around the world have fully recognized the importance of collaboration. Through a global multidisciplinary collaboration, multiple expert consensus guidelines have been published for RIHD limitation strategies based on available data and cancer types. Relevant people should be familiar with these guidelines for screening and mitigation of RIHD in adults and children, which advocate cardiovascular risk assessment and reduction before and after radiotherapy, as well as cardiovascular imaging at appropriate follow-up intervals for early identification of subclinical cardiovascular disease and for the management of RIHD throughout its course (92–94).