Cancer remains the second leading cause of death worldwide, with almost 10 million deaths in 2022 and nearly 20 million new cases, and an estimated 35 million new cases by 2050, making it a major public health concern. It is predicted that about one in five men or women will develop cancer in their lifetime, while about one in nine men and one in 12 women will die from cancer.1,2 Overall, lung and colorectal cancer are the main causes of death, followed by liver, breast and stomach cancer. The higher incidence is mostly due to lung cancer, breast cancer, colorectal, prostate and stomach cancer.3

The treatment of cancer patients depends on the histology and tumor stage. Surgical resection is the preferred treatment for patients in the early stages of the disease, while chemotherapy and radiotherapy are the standard treatments for those diagnosed in advanced stages. However, even with the development of less toxic and more effective chemotherapy, the lack of precise therapeutic targets and continued tumor resistance to treatments has led to the emergence of precision oncology. This approach is based on new therapies that aim to eradicate oncogenic molecular alterations, select patients, and administer drugs to malignancies with these alterations.4 More recently, immunotherapy has become the most attractive treatment option for various types of cancer with very encouraging results, including melanoma and lung cancer, among others.5 Nevertheless, despite the considerable advances that have been made in treatment modalities, many cancer patients experience recurrence and metastasis.6 In this sense, it is known that cancer cells are heterogeneous and there is solid evidence indicating that chemoresistance, tumor progression and metastasis are linked to a subpopulation of cells with “stem” characteristics called Cancer Stem Cells (CSC).4,7 Understanding the mechanisms that regulate the properties of CSC and developing therapeutic strategies specifically targeting these cells are areas of active research, aimed at improving treatment outcomes and survival rates for cancer patients.7,8 However, few studies have attempted to analyse the scientific production on CSC in terms of derived articles, the content of which is limited only to focus on one or a few specific cancer types.9 Furthermore, the extent of the raw data produced by CSC research is also not known.

In this respect, the availability of open access (OA) publications and the accessibility to the high quality research datasets that underpin these articles is crucial, as it can improve the dynamics of cancer research.10 Among the main problems faced by researchers in the field of oncology, it is worth highlighting the limitations of patient samples obtained by invasive techniques, the lack of reproducibility of published results and the need for OA articles; and even more complex, finding the datasets that support these results.11 Research raw data may be deposited in repositories or included as supplementary material with the papers. The benefit of data sharing has been described in the acceleration of research in medicine12 and recently during the last pandemic, by SARS-CoV-2.13,14

The objective of this work is to evaluate the scientific production on CSC, with particular emphasis on the study of current practices related to OA publications and research data sharing, to examine the availability and type of raw data sets released through scientific publications, signed by at least one Spanish affiliation.

A bibliometric study was conducted in the following stages:

• 1)

  Search strategy. The terms of the search equation were identified from the classification of cancer types established by the U.S. National Cancer Institute (NCI)15 and combined with the terms “Cancer* Stem Cell*” OR “Cell* Cancer* Stem” OR “Stem Cell* Cancer*” (see Supplementary file 1). The bibliographic search was carried out through the Science Citation Index Expanded (SCIE) of the Web of Science (WoS) in January 2021, and the selected period was until December 31, 2020, limiting the results to the Country: Spain. An Access database was generated with a total of 708 retrieved documents which contained the following information: article title, authors’ names, institutional affiliations, journal title, financial support, keywords, and data availability statement.

• 2)

  Comparative study of scientific publications in CSC was developed analysing the journal's openness criteria classified into quartiles (Q) based on the 2019 SCIE edition of the Journal Citation Reports of WoS. Additional data were collected for each journal: publisher, journal's access modality (gold OA, hybrid, traditional), possibility of storing manuscripts in thematic or institutional repositories, reuse policy, policy regarding publication on the official website or by the author, statement of policy on supplementary material, and statement of policy related to deposit of research data.

• 3)

  Analysis of research data attached to the article through dissemination as supplementary material or deposit of research data in repositories. To identify publications with research data available in the form of supplementary material, a secondary search was performed by two independent researchers reviewing manually each paper. The number and types of files located on the articles as supplemental material were registered even if a single article included several different files. On the other hand, the searches of the publications with raw research data deposited in repositories were executed in WoS-SCIE combining the initial search equation related to CSC with the terms referring to the main thematic repositories of biomedical sciences and cancer.13

• 4)

  The title, abstract and keywords of the articles were examined in order to perform a thematic study of the papers on CSC, depending on the type of cancer investigated and the anatomical location according to the U.S. NCI classification.15

• 5)

  Descriptive analyses of variables and crosstables were performed using Microsoft Access database and Excel software. Authors, institutional affiliations, and funding entities were normalised manually. Analysis and visualisation of networks of authors and institutions were performed using Pajek software.

CSC have the ability to self-renew and differentiate into various cell types within the tumor mass, contributing to tumor heterogeneity; are thought to be responsible for initiating tumor formation, which is often resistant to conventional cancer therapies; and have been implicated in the process of metastasis, as they can migrate from the primary tumor site to distant organs and initiate secondary tumor growth.8 Therefore, understanding and targeting CSC are areas of active research in oncology, making the subsequent dissemination of scientific results urgent to accelerate the knowledge of cancer, which is a major public health problem and in which a large amount of economic resources is invested.16 This study shows that the international scientific production in the CSC resulted in 708 documents signed by at least one Spanish institution, with the first article in 1998 and a 30-fold increase in 2009, reaching 100 articles in 2020. A decrease in publications is observed in 2018, which coincides with the discontinuation of the journal Oncotarget from the WoS database. From 2019, scientific production recovers and the journals of the MDPI publishing group begin to gain weight, although it is the Nature publishing group that leads the ranking of publications with associated research data. CSC articles showed the participation of 4,001 authors from 828 institutions, of which only 249 affiliations were Spanish, highlighting 7 authors with more than 20 published articles and 7 institutions that appeared in more than 50 papers, being the top producers the Autonomous University of Madrid, the University of Salamanca, Centro Nacional de Investigaciones Oncológicas (CNIO), ICO, CIBERONC and the Universidad Autónoma de Madrid.

Analyses of 288 journals openness criteria demonstrates that over the years there has been an augment in publications in gold OA journals compared to the hybrid model, surpassing it as of 2018, with hardly any publications in commonly known traditional journals. Furthermore, around 96% of journals publishing CSC articles allowed directly or with some conditions the storage of manuscripts in thematic or institutional repositories, reuse, and publication in websites. This behaviour is particularly consistent with the promotion of OA statements in scientific research between 2002 and 2007 and the subsequent mandatory OA publication of publicly funded research.17,18

It is in this context that the concept of Open Science (OS) is framed, meaning unrestricted access to all aspects of research, so that everyone can follow, use and participate in science. OS includes a growing list of other open areas of interests, not only OA to the publications but also open research data, which is a valuable resource that is often expensive and time-consuming to generate.19 In fact, one of the practices that has gained great momentum in recent years because of its ability to improve quality of research is the sharing of research data.20,21 To promote data sharing in cancer research, initiatives such as OA journals, data repositories, and collaborative research networks have been established. For example, The Cancer Genome Atlas (TGCA) is a project launched in 2005 by the NCI and the National Human Genome Research Institute of the USA to accelerate understanding of the molecular basis of cancer by applying genomic analysis and protein expression profiling maximizing data sharing. In 2013, the Global Alliance for Genomics (GA4GH) was established as a common framework to enable responsible, voluntary and secure sharing of clinical and genomic data.22 In Europe, the Oncology Data Network (ODN) for all cancer types was developed in 2017 with accessibility to all cancer centers and patients.23 Its objective is to establish a comprehensive infrastructure of real-world information on cancer care. The final aim is to inform precision medicine and validate specific therapeutic approaches for each cancer subtype by sharing data derived from routine clinical experience.24

As a further step, the rationale for this study was therefore based on the availability of CSC research and datasets shared as supplementary information in articles or deposited in repositories. Our results show that 303 out of 708 articles contained associated research data (47%), mostly published in Q1 journals (57.3%), demonstrating a relationship between impact journals and their commitment to improving the quality, transparency and reproducibility of the communicated research.

Supplementary material was the preferred method of data sharing, with pdf format being the most common file type used, followed by image and xls files. These findings suggest a similar behavior to previous studies of data sharing in regenerative medicine and stem cell research (45.9%).25 Furthermore, these are promising results, although not sufficient, when compared with the availability of associated data in other fields such as emergency medicine (9.4%),26 general and internal medicine (9.5%),27 dentistry (7.6%) or substance abuse (4.7%).28 In contrast, only 69 articles mentioned data sets deposited in repositories, showing that this is still an uncommon practice, closely related to omics research, referring to data from genomic, transcriptomic, proteomic and metabolomic analyses, and which ranks first in the GEO repository. These results highlight the need to ensure that the research data generated is in line with the FAIR (Findable, Accessible, Interoperable and Reusable) principles developed to identify good practice in data sharing.10

Regarding the main organs affected by cancer in the 708 papers studied, first position is digestive/gastrointestinal cancer (17.2%), followed by neurologic cancer (14.2%), breast cancer (13.5%), hematologic/blood cancer (8.7%) and respiratory/thoracic cancer (3.2%). These results are consistent with the most commonly diagnosed cancers and the leading causes of death,1,3 but also indicate those cancers where the presence of CSC has been identified as playing a significant role in the disease. CSC were first identified in leukemia and are thought to contribute to treatment resistance and disease progression.29 Glioblastoma is a highly aggressive type of brain cancer in which CSCs have the ability to resist radiotherapy and chemotherapy, contributing to tumor recurrence after treatment.9 Similarly, in colorectal cancer, CSC play an important role in tumor initiation and progression, as well as treatment resistance and recurrence, and in breast cancer, CSC are able to self-renew and generate daughter cells that make up the tumor, contributing to its heterogeneity and ability to resist treatment.30

Our findings in cancer research through the CSC study show that OA and data sharing policies are being implemented, but at a much slower rate than might be expected, and these data are consistent with other studies on the availability and capture of study protocols in oncology.31 Journals and publishers have a key role to play in promoting OS policies, and researchers should adopt guidelines accepted by the scientific community when archiving their research results.32 In addition, biomedical research institutions have the responsibility to develop data management and sharing plans supervising OS practices,33 and government agencies should recognise researchers and institutions that adhere to OS mandates.34

The data generated and used during this research are openly available from Zenodo.org public repository at https://doi.org/10.5281/zenodo.11370781.

Cristina Candal Pedreira

The corresponding author, on behalf of the other authors, guarantee the accuracy, transparency and honesty of the data and information contained in the study, that no relevant information has been omitted and that all discrepancies between authors have been adequately resolved and described.

R. Lucas-Domínguez and A. Vidal-Infer: conceptualization, data extraction, data curation, formal analysis, funding acquisition, investigation, methodology, project administration, visualization, writing original draft, review and editing. A. Alonso-Arroyo: data curation, formal analysis, investigation, methodology, visualization, writing original draft, review and editing. B. Tarazona-Álvarez, M. Bolaños-Pizarro and V. Paredes-Gallardo: data extraction, formal analysis, investigation, methodology, review and editing.

This work is part of the R&D&I project of the Spanish Ministry of Science and Innovation, reference PID2019-108579RB-I00, funded by MCIN/AEI/10.13039/501100011033.