Top-100 Most-cited Articles on Esophageal Cancer: A Bibliometric Analysis

Background A citation classic is a highly cited work in a eld and regarded as an inuential contribution to the eld’s advancements and literature. Analyzing citation classics and top articles promotes the recognition of research trends within a eld. We present the results of the rst analysis to identify the 100 most frequently cited research studies on esophageal cancer or esophagogastric junction cancer using the bibliometric analysis method. Method We searched the Web of Science on September 24, 2020. Articles were listed in descending order by the total number of citations, and the top-100 most-cited original articles on esophageal cancer or esophagogastric junction cancer were extracted and evaluated. Results The top-100 citation classics in esophageal cancer were published from 1981 to 2018. A signicant increase was found in the number of citation classics from the early 1990s to the late 2000s, which was paralleled by an increase in randomized controlled trials focusing on the clinical treatment of tumors. The medians of the total and annual citations in our analysis were 444.50 (interquartile range [IQR] 346.25-684.50) and 30.08 (IQR 19.10-56.60), respectively. The majority of articles were published in the Journal of Clinical Oncology (n=26), originated in the United States (n=38) and focused on clinical therapies (n=59). The median impact factor of the journals was 27.603 (IQR 9.727-32.956). Conclusion Our analysis provides a historical perspective on the scientic progress of esophageal cancer and contributes to the recognition of important advances in this specialty. the REAL-3 study) on the use of panitumumab, an epidermal growth-factor receptor (EGFR) inhibitor, combined with chemotherapy in patients with EGFR positive advanced esophagogastric cancer, was published in 2013. Three articles assessed the value of the vascular EGFR inhibitor in patients with advanced gastric cancer during 2014-2016. The use of immunotherapy for advanced gastric and EGJ cancers has increased in recent years, producing two recent articles. One 1993 study reported the use of metal stents for palliation of esophageal obstruction.


Background
Esophageal cancer is the eighth most common cancer and the sixth cause of death by cancer worldwide according to the Global Cancer Observatory 2018 [1] . The incidence of esophageal cancer is calculated by geographical distribution. Southeast Asia, especially China, is a high incidence area, with an estimated 477,900 new cases and 375,000 deaths in 2015 alone [2] . Due to the anatomic location of the esophagus, its lymphatic drainage ranges from the lower neck to the abdominal trunk, and laterally along the esophageal wall to the surrounding trachea, heart, large vessels and vertebral body.
Therefore, its recurrence and metastasis rates are very high and the overall survival is poor. A large number of research studies have been conducted in recent years with the intent to improve the overall survival rate of esophageal cancer. With the development of radiation technology and new drugs, the multidisciplinary comprehensive treatment of this cancer has gradually matured.
Retrospective analyses of historical trends in research on esophageal cancer and the examination of the literature that has had a signi cant impact on the eld can provide insight into scienti c progress over time and identify areas of greatest interest along with understudied areas. Frequently cited works often represent landmarks in the literature that have gained recognition by researchers in the eld and laid the groundwork for important clinical research ndings [3] . Bibliometric analyses of a discipline's most-cited articles, also called citation classics [4] , provide insight into research areas that have in uenced the discipline. The number of citations and citation density (i.e., annual number of citations since the date of publication) can be used as an indicator of the in uence of a certain publication on a respective eld [5] . Examining citation classics in a eld facilitates the identi cation of important literature and provides insight into research trends over time. Although articles on citation classics in the areas of critical care medicine [6] , pituitary adenoma [7] , neuroimaging [8] , neuro-oncology [9] , ischemic stroke [10] , obstetrics and gynecology [11] and urogynecology [12] have recently been conducted using bibliometric analysis, no studies have examined the research literature on esophageal cancer. Therefore, this study aimed to analyze the 100 most-cited original articles in the eld of esophageal cancer using the bibliometric analysis method and to provide guidance to investigators in this eld.

Search Strategies and Inclusion Criteria
Web of Science Core Collection includes the Science Citation Index Expanded as well as other citation indexes. We searched the Web of Science Core Collection (Clarivate Analytics) on September 24, 2020 using the following terms: [(esophageal cancer) OR (esophagogastric junction cancer) OR (esophageal carcinoma) OR (esophagogastric junction carcinoma) OR (esophagus tumor)].Incites Journal Citation Reports (JCR) was used to identify the journal impact factors.
Search results were sorted by the total number of citations in descending order. Articles were rst screened by title for their relevance to the esophagus, and the remaining articles underwent a review of the abstract or the full text. Only original articles were included. Systematic reviews, secondary analyses (e.g., meta-analyses), case reports or case series, meeting reports and guidelines and consensus statements were excluded from the study because they did not re ect trends in esophageal cancer research. After two independent authors screened the studies, the top-100 most-cited articles were included in the nal analysis.
Approval from an ethics committee to conduct this research was not required because this study did not involve the collection of original data from humans or animals or interventions.

Data Collection
The following parameters were extracted from each article: article title, country of origin, year of publication, journal name, journal impact factor (IF; Data were collected from Incites Journal Citation Reports for the year 2019), total number of citations, number of annual citations, tumor type, study design, topic theme and type of treatment. Country of origin was based on the a liation of the corresponding author. If there were multi-corresponding authors in an article, the rst corresponding author's country was collected. The number of annual citations was calculated by dividing the total number of citations by the number of years, which was from the date of publication to 2018. Study designs included the randomized controlled trial (RCT), prospective study, retrospective study, laboratory investigation and observational study (i.e., cross-sectional study, case-control study or cohort study). In order to assess trends in the literature, articles were classi ed by their primary themes, which included therapies, pathogenesis or clinical presentation and imaging. Therapy-related articles were subdivided into surgery, chemotherapy, radiation therapy, chemotherapy combined with radiotherapy (i.e., chemoradiotherapy), new agents (i.e., immunotherapy) and metal stents.

Statistics
Data are presented using descriptive statistics. Spearman's correlations between continuous variables that were not normally distributed were analyzed using the Statistical Package for Social Sciences software version 23.0 (SPSS Inc., Chicago, IL). A p-value < 0.05 was considered statistically signi cant.

Results
The 100 most-cited articles in the eld of esophageal cancer or esophagogastric junction (EGJ) cancer and the number of citations are presented in Table 1. The total number of citations of the 100 articles ranged from 3,641 to 314, with a median of 444.50 total citations (interquartile range [IQR] 346. 25-684.50) and a median of 30.08 annual citations (IQR 19.10-56.60). The top-3 ranked articles were clinical studies, whose themes were related to perioperative chemotherapy [13] , postoperative chemoradiotherapy for gastric or EGJ adenocarcinoma [14] and preoperative chemoradiotherapy for esophageal squamous cell carcinoma (ESCC) or adenocarcinoma [15] .

Source
The 100 most-cited articles were published in 27 different journals ( Table 2).
The Journal of Clinical Oncology (n=26), Annals of Surgery (n=10) and New England Journal of Medicine (n=9) published the top-3 ranked articles. Journals were classi ed into ve main categories. Fifty-one articles were published in oncology-speci c journals, 17 articles in genera-specialty journals, 15 articles in surgery-speci c journals, 8 articles in gastroenterology and hepatology-speci c journals and 9 articles in various other journals. The journal impact factors ranged from 2.234 to 74.699 (median, 27.603; IQR 9.727-32.956). A moderately strong correlation was found between impact factor and total number of citations reported in a given journal (Spearman's coe cient = 0.552, p < 0.001), and the number of annual citations strongly correlated with the total number of citations (Spearman's coe cient = 0.715, p < 0.001).
The majority of the articles originated in the United States (US) with 38 publications, Japan and Germany, both with 15 publications and the United Kingdom with 9 articles, which was ranked third (Figure 1).

Study Design and Themes
The most frequent study design was the RCT (n=42), followed by the prospective trial (n=17), retrospective study (n=17) and laboratory investigation (n=15). Observational studies accounted for only 9% (9/100) of all studies.
The most common themes were therapies (n=59), followed by pathogenesis or clinical presentation (n=34) and imaging (n=7). Among the therapeutic trials, 22 pertained to chemotherapy, 19 to chemoradiotherapy and 14 to surgery (Figure 2).
Three articles assessed the value of the vascular EGFR inhibitor in patients with advanced gastric cancer during 2014-2016. The use of immunotherapy for advanced gastric and EGJ cancers has increased in recent years, producing two recent articles. One 1993 study reported the use of metal stents for palliation of esophageal obstruction.
Nineteen citations pertained to preoperative treatment. Among these, 12 articles reviewed the use of neoadjuvant concurrent chemoradiotherapy for esophageal cancer, 6 articles were about preoperative chemotherapy and 1 examined the value of preoperative radiotherapy.

Pathogenesis or Clinical Presentation
A total of 34 citations focused on tumor pathogenesis or clinical presentation. Among these, 16 reported clinical or pathological characteristics of esophageal cancer. Among the 18 pathogenesis-related studies, 9 focused on the etiology of exogenous factors, such as obesity, alcohol intake, tobacco smoking, nutrition de ciencies, biotoxins, medicines and gastroesophageal re ux. The remaining 9 articles were related to gene levels, such as p53, microRNA, SOX2, MTS1/CDK4I and PLCE1.

Imaging
Seven articles pertained to imaging (2000-2011). Among these, 5 explored the predictors of response to chemotherapy or chemoradiotherapy among patients with esophageal cancer using positive emission tomography. One study investigated the utility of positive emission tomography for the staging of patients with potentially operable esophageal carcinoma. The remaining study focused on the early detection of super cial squamous cell carcinoma in the head and neck regions and esophagus using narrow band imaging.

Temporal Trends
There has been a sharp increase in the number of cited articles since the early 1990s. The peak period with the largest number of cited articles was 2005-2008 (n=23), with a steep decline afterwards ( Figure 3A). This increase was largely attributable to studies pertaining to therapies ( Figure 3B), especially those focusing on surgery, chemoradiotherapy and chemotherapy ( Figure 3C). Furthermore, immunotherapy, which was a promising treatment was developing rapidly and produced 2 of the most-cited articles within two years. Since the 1990s, RCTs have increased continuously and steadily. Prospective studies have also increased gradually with a decrease in retrospective studies, which nally reached its peak during 2005-2008 ( Figure 3D).

Discussion
Citation analysis can be a useful tool for identifying publications that have a high scienti c impact. A bibliometric analysis of the most-cited articles highlights the most in uential articles, trending topics and proli c institutions that contribute to the evolution of a scienti c subspecialty. Our study depicts the top 100 citation classics that have played a signi cant role in bolstering progress in the eld of esophageal cancer.
In our analysis, the types of pathology found in the patients in most of the studies were adenocarcinoma with or without squamous cell carcinoma, while some of the other studies involving esophageal squamous cell carcinoma were from Japan and a few were from Germany and France. This nding also re ects the characteristics of the distribution of esophageal cancer. The research mostly pertained to clinical treatment, followed by pathogenesis or clinical presentation. With developments in the eld of medicine, the numbers of RCTs and prospective studies of multidisciplinary treatments are booming, thereby providing more evidence-based rationales for clinicians' treatment decisions.
Analyses of the citation classics showed that more than half of the 100 most-cited articles were published in oncology-speci c journals, and 45% of the articles were published in journals with very high impact factors, including Japan Clinical Onclogy, Annals of Surgery and the New England Journal of Medicine. Most of the remaining articles were published in journals with a relatively low impact factor. Thus, the median impact factor was 27.603.
Although the articles published in top journals are cited more frequently, the articles published in top journals, such as the New England Journal of Medicine are relatively rare. Therefore, the correlation between the impact factor and total number of citations of classic papers in this study was not strong. However, a strong correlation was found between the annual and total numbers of citations, indicating the in uence of the research reported in the classic papers in the eld is persistent and has not been diluted by time.
Most of the classic citations were from the US, which is consistent with the results of most other cancer-related studies. This likely re ects a bias of US authors to submit studies for publication in US-based journals and of US reviewers toward articles from authors and studies based in the US [16] . Of note, we did not identify articles in languages other than English; thus, we have not represented the entire body of literature that could possibly be relevant to the eld of esophageal cancer. That said, this same bias toward English-language articles, made clear by past research revealing that non-Englishlanguage articles have low acceptace and citation rates, reduces the likelihood of increasing the number of citations in non-English-language articles [17,18] . The most-cited articles focused mostly on optimal clinical treatment in RCTs along with therapeutic studies. Among these, the classic articles on radical chemoradiotherapy were the RTOG 85-01 [19,20] and RTOG 94-05 studies, [21] which established guidelines for locally advanced esophageal cancer. The studies from Japan revealed that three-eld lymphadenectomy has better local control and overall survival than two-eld lymphadenectomy [22][23][24][25] .
However, three-eld lymphadenectomy could only be performed in a hospital; thus, it has not been adopted worldwide. Furthermore, the persistence and the recurrence rate of esophageal cancer were high (24-58% [19][20][21][26][27][28][29][30] ), and the 5-year overall survival was low (5-25% [31] ) after radical chemoradiotherapy or surgery alone. This nding has promoted multidisciplinary treatment studies, which include surgery combined with chemotherapy or chemoradiotherapy [32][33][34][35][36][37] . Due to the backwardness of radiotherapy technology and the high incidence of surgical complications in earlier years, most of the prospective preoperative chemoradiotherapy studies had insu cient sample sizes or controversial results. In 2012, a Dutch study (known as the CROSS study) con rmed that the use of neoadjuvant chemoradiotherapy for esophageal cancer has better overall survival and disease free survival than surgery alone, which had been recommended as standard treatment for those with locally advanced esophageal cancer [15] .
The long-term follow-up results further con rmed the value of preoperative chemoradiotherapy [38] . Since then, neoadjuvant chemoradiotherapy for esophageal cancer has entered a new era.
Three citation classics in perioperative chemotherapy mainly focused on adenocarcinoma of the esophagus or stomach [13,39,40] , which might not have been suitable for esophageal squamous cell carcinoma. Similarly, the RCT of postoperative chemoradiotherapy also focused on adenocarcinoma of the esophagus or stomach [14] . Postoperative chemotherapy only improved the disease free survival for patients with ESCC [41] . Based on the citation classics, no recommendations have been made for adjuvant therapy after radical esophagectomy for ESCC.
Salvage treatment studies have gradually changed from chemotherapy [42][43][44] to targeted therapy [45][46][47][48] , with the discovery of molecular checkpoints. Of note, immunotherapy has also attracted much attention in recent years, opening a new door to the treatment of esophageal cancer. At present, the role of immunotherapy in neoadjuvant, adjuvant or radical therapy is still unclear, and many studies are ongoing.
Although bibliometric analysis can provide a comprehensive understanding of the important studies on esophageal cancer, citation analysis has drawn criticism owing to the effects of time [16,49,50] . Some hypothesize that older studies have more time to accumulate citations, making them more likely to end up on the list of citation classics. Articles that have become "landmark" studies may be cited less often over time as their conclusions become more commonplace, a phenomenon referred to as "obliteration by incorporation" [16,50] . To mitigate the effect of time in this way, the number of annual citations was introduced. However, the time lag between publication and accumulation of su cient citations for incorporation into a list of citation classics, such as our top-100 classics, must be acknowledged, and one cannot predict which of the more recent articles will rise to be included as a citation classic and within what time frame. For this reason, elds must periodically update their list of classics in keeping with this trend.  Table 2. Journals in which the 100 most-cited articles were published.

Figure 1
Countries of origin of the 100 most-cited articles.

Figure 2
Distributions of the 100 most-cited articles by study design and theme.