Insights into research on myocardial ischemia/reperfusion injury from 2012 to 2021: a bibliometric analysis

The Web of Science Core Collection (WoSCC) database includes the leading global academic journals, and allows downloading full citation records with new records being updated every day. The WoSCC is the most commonly used database for bibliometric analyses. Previous reports have widely confirmed the validity of bibliometric analyses based solely on the WoSCC database [12, 13, 15]. Therefore, we chose the WoSCC database to retrieve relevant studies using an advanced search strategy. The primary search words and combinations were “myocardial reperfusion injury” and “myocardial ischemia/reperfusion injury”. The following were the selection criteria: (1) publication year: January 2012–December 2021, (2) document type: article or review, and (3) language: English. All searches were completed and downloaded on the same day (April 3, 2022). Details of the retrieval procedure are provided in Additional file 1: Fig. S1.

The search results were screened independently by two investigators to exclude irrelevant, duplicate, and withdrawn articles. Disagreements were resolved by consulting a third investigator.

The WoSCC literature analysis report was used to evaluate publication characteristics, such as output in the form of number of annual publications, journals, authors, citation frequency, impact factor (IF), and Hirsch index (H-index). IF reflects a journal's rank or importance by calculating the average number of times that its selected papers are cited in a particular year. The H-index, indicating that the given author/country has published at least H papers which have each been cited at least H times, was used to evaluate the scientific impact of author or country. GraphPad Prism 7.00 (GraphPad Software, San Diego, USA) was used to analyze publication trends. The growth of publications in the following year was estimated using the polynomial model: \(f\left(x\right)=a{x}^{3}+b{x}^{2}+cx+d\).

CiteSpace, a visualization tool developed by Professor Chaomei Chen, analyzes emerging trends and critical changes in scientific literature [16]. We used the CiteSpace V 5.8.R3 software (Drexel University, Philadelphia, USA) to explore collaboration networks between countries/institutes/authors, perform co-cited reference analysis, and identify keywords with citation bursts [11]. The betweenness centrality for each node in a network was calculated to reveal how likely it is for an arbitrary shortest path to go through the node. The node with a centrality value > 0.1 was found to connect two or more large groups of nodes by being in between them [11]. The burst detection algorithm was used to find out how an abrupt change takes place in an event within a specific duration. In addition, the co-cited references were clustered using the log-likelihood ratio test to identify the major subfields of MI/R injury research. Two indicators, including the Modularity Q (Q value) and Mean Silhouette (S value), were used to evaluate the quality of the cluster network formed. If the Q > 0.3, the cluster structure obtained was significant; when the S > 0.7, the clustering was efficient and convincing [11].

The VOSviewer (Leiden University, Gravenhage, the Netherlands) software creates, visualizes, and explores knowledge maps based on network data. We used VOSviewer 1.6.18 to identify co-cited journals, and display the network structure [11].

After scanning, a total of 8419 English records related to MI/R injury were included, of which 6623 were articles (78.7%) and 1796 were reviews (21.3%). The total citation frequency was 214,942, the H-index count was 156, and the average number of citations per item was 25.53. A flowchart of the literature selection is presented in Fig. 1.

As presented in Fig. 2A, the number of studies on MI/R demonstrated an overall upward trend rising from 629 publications in 2012 to 1024 publications in 2021. According to the polynomial curve fitting of publication growth, the number of publications was estimated to reach 1039 in 2022 (Fig. 2B). The above results demonstrate that MI/R-related studies have received widespread attention from scholars across the world in recent years.

All the studies were distributed among 91 countries and 4847 institutions. As illustrated in Fig. 3A, the largest number of publications originated in China [3805 (45.20%)], followed by the USA [2020 (23.99%)], Germany [517 (6.14%)], England [416 (4.94%)], and Italy [348 (4.13%)]. The top four countries, in terms of centrality, were the USA (0.20), France (0.19), England (0.17), and Italy (0.13), representing close cooperation among themselves. The country with the most average citations was England (51.02). With regard to the H-index, the USA (128), China (91), England (71), Germany (71) and Italy (58) were the top five countries (Fig. 3B).

As depicted in Fig. 3C, Fourth Military Medical University [188 publications (2.23%)], Nanjing Medical University [138 publications (1.64%)], Capital Medical University [131 publications (1.56%)], Wuhan University [131 publications (1.56%)], and Shanghai Jiao Tong University [130 publications (1.54%)] were the largest contributors of MI/R research publications. University College London (UCL) had the highest centrality (0.13) among all the institutions, representing itself as a key hub for promoting cooperation among relevant research institutions. These results indicate that the USA, England, China, Germany, and Italy have exerted a strong scholarly influence in the field of MI/R.

A total of 28,814 authors were involved in MI/R research. As presented in Table 1, the author who published the most publications was Peter Ferdinandy from the Semmelweis University (67 publications), followed by Derek J Hausenloy from University College London (63 publications), Erhe Gao from Temple University (55 publications), Derek M Yellon from University College London (53 publications), and Yang Yang from the Fourth Military Medical University (53 publications). The top 10 contributing authors hailed from institutions in Europe (n = 6), China (n = 3), and North America (n = 1). All the top 10 most published authors had total citation counts of over 1000, and their average number of citations per publication was more than 25. In terms of H-index, the top five authors were Derek J Hausenloy, Derek M Yellon, Gerd Heusch, Rainer Schulz, and Erhe Gao.

The network of author cooperation is presented in Fig. 4. The highest ranked authors, according to centrality, were Jie Wang from Tianjin Medical University (0.42), Zhelong Xu from Tianjin Medical University (0.41), Thomas Krieg (0.29) from University of Cambridge, Sean M Davidson from University College London (0.24), and Hans Erik Botker from University College London (0.22).

The documents were published in 1203 journals, 13 of which had more than 100 publications (Table 2). PloS One accounted for the most output, followed by Oxidative Medicine and Cellular Longevity, Molecular Medicine Reports, International Journal of Molecular Sciences, and Journal of Molecular and Cellular Cardiology. Among the top 10 journals with the most publications, Basic Research in Cardiology had the highest IF (17.165), ranking sixth in terms of number of publications. Most of the productive journals were classified as Q1 or Q2 according to the 2020 Journal Citation Reports. In addition, the number of studies on MI/R injury was normalized to the sum of publications in respective journals. These results showed that the proportion of studies on MI/R injury was highest in Basic Research in Cardiology (22.29% of the total publications per journal), followed by Journal of Molecular and Cellular Cardiology (7.11% of the total publications per journal), Cardiovascular Research (5.81% of the total publications per journal), American Journal of Physiology-Heart and Circulatory Physiology (4.04% of the total publications per journal), and Oxidative Medicine and Cellular Longevity (2.93% of the total publications per journal).

A co-citation relationship between two journals arises when they are both cited in the same publication simultaneously. High co-citation counts imply that the journals produce superior academic material and can, therefore, be considered mainstream. As illustrated in Fig. 5, the top five co-cited journals were Circulation, Circulation Research, Cardiovascular Research, American Journal of Physiology-Heart and Circulatory Physiology, and Journal of Molecular and Cellular Cardiology. The average IF of the top 10 co-cited journals was 12.608 higher than that of the top journals with the most publications (6.659).

As presented in Table 3, the top three cited studies were:

However, it is important to note that the number of citations does not adequately reflect the knowledge base of a particular field, since older articles are typically cited more frequently than newer ones.

Reference co-citation analysis identifies highly co-cited references that are frequently cited together by other articles, and is, therefore, usually applied for exploring research foci in a given academic field. A reference co-citation network with 685 nodes and 1857 connections is presented in Fig. 6A. Of the top 10 co-cited references, one is an original article and nine are reviews (Additional file 1: Table S1). “Myocardial Ischemia–Reperfusion Injury: A Neglected Therapeutic Target” by Derek J Hausenloy published in Journal of Clinical Investigation was the most frequently co-cited publication [18].

A paper with strong citation burst can be viewed as representing a major milestone in the knowledge base. Figure 6A highlights the papers with citation bursts as red dots or circles. All the top 10 co-cited references had a burst strength of more than 10, except the review published by Sean M Davidson titled “Multitarget Strategies to Reduce Myocardial Ischemia/Reperfusion Injury: JACC Review Topic of the Week” [3]. However, the review published in Journal of the American College of Cardiology by Sean M Davidson in 2019 is the latest of the top 10 co-cited references, and therefore, the time lag between publication and citation might have impacted the burst detection.

To further explore characteristics of the temporal evolution of the knowledge base in MI/R research, we performed a cluster analysis using a timeline view tool of CiteSpace (Fig. 6B). The Q value of the network of co-cited references was 0.779, indicating that it obtained a good clustering. The average S value of the network was 0.933, suggesting high homogeneity of the references in a cluster. Among the top 10 clusters, the clusters which were most recently active included #4 lncrna (long non-coding RNA) and #9 ferroptosis. The clusters #0 mitophagy, #1 cardioprotection, #2 inflammation, #5 melatonin, and #7 mitochondria emerged as active areas in the period 2013–2018, while activity in the five clusters appeared to be declining. The clusters #3 remote ischemic preconditioning, #6 postconditioning, and #8 microvascular obstruction were formed by documents with a mean publication year of 2010, thereby representing older clusters in the MI/R knowledge base.

Keywords are the focus of attention in a given scientific paper. A total of 25,662 keywords were extracted by CiteSpace, of which 45 appeared more than 100 times. According to their co-occurrence frequency, the top 20 keywords in MI/R research are listed in Fig. 7A. The high-frequency keywords can be divided into three broad categories: clinical manifestation, molecular mechanism, and cardioprotective strategies for MI/R injury.

To track hotspots and future directions in the field of MI/R research, we further used the burst detection algorithm to analyze the keywords. A keyword with a high burst level indicates that it received greater attention in a specific time period. The top 20 keywords with the strongest appearance bursts are presented in Fig. 7B. Overall, two stages were identified; the first stage lasted from 2012 to 2017, and the second stage spanned from 2018 to 2021. In the first stage, the top keywords were ATP-sensitive potassium (K-ATP) channel, infarct size, matrix metalloproteinase, adenosine, glucose insulin potassium, cytochrome c oxidase, toll-like receptor, tolerance, resveratrol, drug delivery, and flavonoid. The top keywords in the second stage included extracellular vesicles, long noncoding RNA, cell proliferation, microRNA, mitochondrial quality control, mitophagy, biomarker, and mitochondrial biogenesis. These results indicate that the research focus had gradually shifted to acquire an in-depth insight into the pathogenesis underlying MI/R injury.

One of the indications of the direction of development in a field is change in annual output. According to the WoSCC database, the 8419 papers were published in 1203 academic journals by 28,814 authors from 4847 institutions across 91 countries/regions. The overall trend between 2012 and 2021 was a slow increase in publications. Among the 10 highest output countries, China was the only one from the developing world. The USA was ahead of other countries/regions in H-index, while England had the highest average number of citations per publication. Moreover, the top six authors with the highest number of published articles were from European countries. This implies that the European countries exert a positive academic influence on MI/R research. The most active collaborations were observed among the USA, France, England, and Italy. Nine of the top 10 most prolific institutions were from China, demonstrating its fast progress in MI/R research over the past decade. However, the cooperation among Chinese institutions was infrequent, and UCL in England was the main collaborating center. Good partnerships and a high degree of collaboration among countries and institutions are important for high-quality research output. Therefore, Chinese researchers are encouraged to cultivate partnerships with institutions across the globe as their research output continues to expand, which will improve their academic influence further.

Highlighting the contributions of productive and influential researchers could act as a reference for those scholars and institutions looking for potential collaborators on the basis of research directions [11]. On analyzing author distribution, we found that the top three most influential authors according to the H-index were Derek J Hausenloy (University College London, England), Derek M Yellon (University College London, England), and Gerd Heusch (University of Essen Medical School, Germany). These authors published a large number of papers and were also recognized as the Clarivate Highly Cited Researchers in 2021. The above findings suggest that these three authors have made outstanding contributions, and enjoy a strong academic reputation in the field of MI/R research. Furthermore, the map of authors provides information about potential collaborators who were regarded as influencing authors in Europe and Asia.

In terms of a journal’s publications, IF, and ranking, Basic Research in Cardiology is a high-quality journal that exerts considerable influence in the field of MI/R injury. However, mainstream journals in specific fields of study need not have high publication capacity. In fact, we found that some journals with high publications did not receive much attention from the academic community. Quality and influence of a journal can also be determined by its co-citation frequency. The top three journals, in terms of co-citation frequency, were Circulation, Circulation Research, and Cardiovascular Research, with IF values greater than 10. This suggests that these three journals influence research foci, and should thus be considered when tracking progress in MI/R research.

The highly co-cited references can be viewed as the knowledge base in a particular field [11]. References with strong citation bursts could also characterize research foundations of a field [11]. Our reference co-citation analysis brought forward some of the influential MI/R studies with highly co-cited counts and strong citation bursts published between 2012 and 2021. In general, the top 10 co-cited references focused on: (1) the pathophysiology of MI/R, including the spatial and temporal evolution of ischemic and reperfusion damage, the cell death modes involved, and consequent coronary microvascular dysfunction, (2) the cardioprotective role of pharmacological and non-pharmacological strategies, such as local ischemic conditioning, remote ischemic conditioning and cyclosporine, and their underlying signal transduction pathways, and (3) potential reasons for failing to translate the extensive and promising basic scientific findings into successful clinical trials. The results offer readers a map for locating those studies that have made large contributions to the field of MI/R research.

Meanwhile, a cluster analysis was conducted to identify the major subfields in the MI/R research, and a timeline view was used to show the developmental process of these subfields. Over the past decade, there has been considerable progress in the understanding of mechanisms of MI/R injury, and development of strategies to protect the myocardium from ischemia and reperfusion damage [2, 3, 8]. By analyzing the top keywords and co-cited references, we found that the mechanistic research before 2018 had mostly focused on molecules involved in cell death, inflammation, and coronary microvascular obstruction. Apoptosis is a cell death mode characterized by DNA fragmentation, but no inflammatory response. During MI/R, both mitochondria and endoplasmic reticulum are crucial organelles that mediate cardiomyocyte apoptosis. It has been suggested that oxidative stress, mitochondrial permeability transition pore, and the K-ATP channel are important factors that contribute to the mitochondria-dependent apoptotic pathway in cardiomyocytes [20]. MI/R can affect the protein-folding capacity of the endoplasmic reticulum, and activate the endoplasmic reticulum stress response [21]. While endoplasmic reticulum stress initially serves adaptive purposes for restoring proteostasis, severe or persistent stress leads to cell apoptosis. Accumulated evidence implicates that the inhibition of excessive endoplasmic reticulum stress could improve cardiac function following MI/R injury [21, 22]. When MI/R induces necrotic cell death, an inflammatory cascade continues damaging the cardiac tissue [23]. The transcription factor nuclear factor-κB (NF-κB) is a key player in the inflammatory signaling. Previous basic research reported that inhibiting the NF-κB pathway minimizes the release of inflammatory factors and reduces infarct size, thereby protecting against MI/R injury [23]. Pyroptosis is a unique form of programmed cell death, that triggers inflammatory response upon infection or sterile insults [24, 25]. During MI/R, a wide range of danger signals, such as oxygen-free radicals, potassium efflux, and mitochondrial stress are produced by dying cells, which activate the inflammasome protein complex leading to pyroptotic cell death and interleukin-1β-driven inflammation [26, 27]. Experimental data shows that inhibition of inflammasome activation could reduce infarct size and cardiac dysfunction following MI/R [28, 29]. Ischemia/reperfusion injury in the coronary circulation manifests as microvascular obstruction and intramyocardial hemorrhage [30, 31]. The coronary microcirculation plays a crucial role in tissue oxygenation and nutrition exchange, and is an important determinant of the infarct size after MI/R [30, 32]. Experimental studies have shown that melatonin treatment protects cardiac microvasculature against MI/R injury by activating the AMPK pathway [33].

As for cardioprotective strategies, experimental studies in the early years from 2007 to 2012 had revealed the essential signal transduction pathways underlying local ischemic postconditioning and remote ischemic preconditioning [34]. For example, oxidative stress after ischemia/reperfusion produces a large number of reactive oxygen and nitrogen species, leading to cardiac cell death and contractile dysfunction. Regulation of redox balance has been demonstrated as an important target of these conditioning strategies against MI/R injury [35‐37]. In the intermediate years from 2013 to 2018, MI/R research had focused on translating the cardioprotection of conditioning strategies through animal experiments and early proof-of-concept trials in humans with acute myocardial infarction to clinical practice. In addition, a variety of pharmacological approaches to cardioprotection, such as adenosine and glucose insulin potassium, have progressed to clinical testing for treatment of MI/R damage. Unfortunately, most clinical studies fail to provide evidence of the cardioprotective effects of conditioning strategies and pharmacological approaches [2, 38]. There are several reasons why clinical translation fails, and here we present three major ones [39]:

Although all attempts have failed, better adjunct cardioprotection is still a requirement for decreasing the risk of heart failure after acute myocardial infarction [9]. A clinical translation can be successful and improved through better understanding of the pathophysiology of MI/R injury and more thorough and reliable preclinical evaluation [2, 9]. As seen in the current study, the knowledge base of MI/R from 2019 to 2021 turned to research providing new insights into the molecular events triggered by MI/R, such as lncRNA and ferroptosis. However, few studies were performed on animal models with multiple comorbidities receiving “background drug treatment”. Future studies should be based on robust preclinical data to provide effective translations of cardioprotective strategies from experimental to clinical settings [9].

The keywords with strongest citation bursts were analyzed to present those research frontiers that have recently attracted close attention from the academic community [11]. The keywords with strongest citation bursts (until 2021) identified by this study were mitochondrial quality control, non-coding RNAs, cell proliferation, and extracellular vesicles.

Mitochondria are abundantly distributed and plentiful in the heart tissue. Despite the canonical role of mitochondria in regulating cellular respiration and oxidative phosphorylation, recent studies have concluded that mitochondria are crucial in sensing and integrating extracellular and intracellular signals [40]. There is strong evidence that mitochondrial dysfunction is a decisive signal in various modes of cardiomyocyte death during MI/R injury [41]. Pharmacological agents targeting mitochondrial function, such as cyclosporine A, TRO-40303, and MTP-131 have been shown to attenuate cardiac insults and improve contractile function in preclinical animal models. However, the mitoprotective agents as adjunct to reperfusion have not yet translated into clinically beneficial treatments for patients with acute myocardial infarction [42]. The partial reason for this failure could be due to insufficient pathophysiological knowledge of mitochondria. Evidence from recent studies indicates that mitochondria have a quality control system that maintains and restores their structure and function by regulating mitochondrial biogenesis, networks (fission or fusion), and degradation (mitophagy). If the severity of a stimuli overcomes the protective capacity of the quality control system, mitochondria-dependent cell death would be activated to remove the damaged cells. Abnormal mitochondrial quality control is closely related to mitochondrial dysfunction and contributes to MI/R damage [43]. Experimental data suggest that mitochondrial fission and fusion proteins, which play vital roles in the process of mitochondrial quality control, have emerged as novel targets for cardioprotection [43]. Further studies using clinically relevant large animal models of MI/R injury are needed to test whether targeting mitochondrial fission and fusion proteins such as dynamin-related protein 1 (DRP1) and optic atrophy 1 (OPA1) could preserve mitochondrial function and have therapeutic potential to reduce infarct size. These may be the new foci in the field of MI/R research.

In ongoing research on the regulatory mechanisms of MI/R injury, the focus has gradually shifted from coding RNAs to non-coding RNAs (ncRNAs), particularly microRNAs, lncRNAs, and circular RNAs [44]. microRNAs primarily interact with the 3'-untranslated region of the target mRNA, thus preventing protein translation through RNA-interference [45]. lncRNAs can alter gene expression networks in the nucleus and modulate mRNA stability, translation, and protein function in the cytoplasm by directly interacting with RNA, DNA, or protein [46, 47]. The regulatory actions of circular RNAs include host gene regulation, scaffold, and molecular sponge function [48]. ncRNAs can influence the cellular function of all types of cardiac cell, and are critical regulators during MI/R injury [44]. ncRNAs have also been identified as important diagnostic and prognostic biomarkers in cardiac diseases, including MI/R injury. Beyond this, ncRNAs are rapidly developing as targets (e.g., microRNAs, lncRNAs) and tools (e.g., small interfering RNAs) in novel therapeutic strategies. Considering that about 95% of the human genome produces a vast number of ncRNAs with versatile modes of action in the cell, current understanding of the regulation of MI/R injury by ncRNAs is still rather limited [44]. Further studies are required to develop a deeper understanding of the mechanisms and functions of ncRNAs.

Loss of cardiomyocytes is the major contributor to the pathogenesis of MI/R injury [49]. It is well-known that mammalian cardiomyocyte proliferation is active during development and decreases in the early neonatal stage. Recent studies on cardiomyocyte proliferation have overturned traditional beliefs regarding adult cardiomyocytes permanently leaving the cell cycle [49, 50]. By stimulating the endogenous capacity of cardiomyocyte proliferation, the adult heart can achieve myocardium regeneration and meaningful functional recovery after MI/R injury [8, 50, 51]. However, these findings mainly resulted from in vitro experiments or non-mammalian and rodent in vivo studies. Research involving higher animal models or human trials relied on surrogate measures rather than true cardiomyogenesis assessment [51]. Extensive experiments with large mammals and rigorous assessment of cardiomyogenesis are required to advance the notion of cardiac regeneration toward clinical application [51].

Extracellular vesicles are nanosized vesicles with a double-layer lipid membrane. There is increasing evidence that reveals the vital role of extracellular vesicles as cell-to-cell communication vehicles in the development of MI/R injury [52]. Exosomes, a subtype of extracellular vesicles, are released by various types of cells, and are loaded with non-random subsets of bioactive molecules (proteins, lipids, ncRNAs, and mRNAs) that exist in their source cells. Stem cell-derived exosomes have been shown to confer cardioprotective effects, activate regenerative signals, and participate in cardiac repair after MI/R injury [52]. The biological role and diagnostic and therapeutic potential of extracellular vesicles are attracting increasing attention from researchers in the MI/R field.

This study found that the MI/R knowledge base has evolved rapidly over the past 10 years. Early MI/R research was initially focused on pharmacological and non-pharmacological approaches to inducing cardioprotection, and on their signal transduction pathways. However, research has gradually shifted focus to insightful molecular mechanisms triggered by MI/R. The findings of this study would act as a reference for researchers looking for research directions, and lay the foundation for further research on MI/R injury. In addition, our results reveal the most productive authors and institutions, which could help researchers decide on collaborations. Journals publishing high-impact studies should be followed by researchers to track the progress of MI/R research.

However, the present study does have several limitations. First, studies were not retrieved from databases other than the WoSCC database (for instance, Google Scholar or PubMed). Second, only English articles and reviews were included, which could have resulted in a language bias. Third, different H-index values are possible for the same author, because the H-index method is based on the number of publications and citations in a given time period.