What is the Impact of Endothelial-to-Mesenchymal Transition in Solid Tumours: A Qualitative Systematic Review and Quantitative Meta-Analysis

Abstract

Endothelial-to-mesenchymal transition (EndMT) has gained increasing recognition as a crucial mechanism in the progression of solid cancers, influencing tumour heterogeneity, metastasis, and resistance to therapy. However, despite its growing importance, EndMT remains insufficiently studied within the cancer research landscape. In this study, we conduct a systematic review, adhered to the 2020 PRISMA guidelines, of the existing literature on EndMT in solid tumours, examining its functional roles, key biomarkers, underlying mechanisms, experimental models, and potential as a target for therapeutic intervention. Our objective was to identify critical areas where further research is needed. In addition, we performed a meta-analysis to evaluate the variability in the expression of EndMT-related markers and their potential links to patient prognosis. To this aim, literature searches were conducted in major databases including PubMed, Scopus, and Web of Science, covering studies published up to June 2024. The risk of bias of selected articles was evaluated using the OHAT tool, for the in vitro experiments and the SYRCLE tool for studies using animal models. Out of an initial pool of 1,197 articles, 54 studies were selected for data extraction by two independent reviewers. Selected studies were identified according to specific inclusion/exclusion criteria applied through distinct stages like "title and abstract screening", "full text article review" and "article bibliography screening". Our analysis confirms that EndMT is a key contributor to tumour progression and metastasis, but several aspects remain poorly understood, particularly regarding the induction of EndMT in specific cancer types, its role in lymphatic endothelial cells, and its interactions with other stromal elements. We observed substantial heterogeneity in the biomarkers associated with EndMT, as well as variations in the endothelial cell types studied, the functional outcomes, and the molecular mechanisms involved. Our meta-analysis revealed significant variability in the expression of EndMT biomarkers, with notable correlations between changes in the expression of specific genes and patient outcomes, particularly in lung cancer. In conclusion, it is essential for future research to focus on identifying the specific cancer and stromal cell types implicated in EndMT and to standardize endothelial cell models and protocols used for inducing EndMT. Investigating EndMT alongside well-established processes, such as epithelial-to-mesenchymal transition (EMT), and exploring its relationship with cancer-associated fibroblasts (CAFs) may provide valuable insights into its role in tumour biology and its impact on therapy resistance.

Keywords: Biomarkers; Cancer-associated fibroblasts; EndMT-related markers; Endothelial cell; Endothelial-to-mesenchymal transition; Meta-analysis; Metastasis; Therapeutic target; Tumour microenvironment.

Figure 1

Morphological and molecular changes during EndMT. Endothelial-like cells (left, red) are typically characterized by tight intercellular connections, exhibiting a cobblestone morphology and expressing junctional proteins. Biomarkers such as VE-cadherin, CD31, and ZO-1 are indicative of a fully differentiated endothelial phenotype. In the transition phase (center, orange), these endothelial cells begin to lose their junctional integrity, undergoing morphological changes to assume a more spindle-like shape. This phase is characterized by the downregulation of endothelial markers (e.g., VE-cadherin, CD31) and the upregulation of mesenchymal-like markers (e.g., N-cadherin, vimentin), signaling the progression of the transition. In the mesenchymal phase (right, yellow), cells adopt a fibroblast-like phenotype, marked by increased motility and altered cell-cell adhesion. The predominant expression of mesenchymal markers (e.g., N-cadherin, vimentin, fibronectin), coupled with the loss of endothelial markers, indicates a complete switch in phenotype.

Figure 2

Flowchart of study screening and selection process. The screening process consists of four steps: identification, screening, eligibility assessment, and final inclusion.

Figure 3

In vitro and in vivo risk of bias. A: Results from the OHAT risk of bias rating tool assessing the risk of bias in the selected studies that included in vitro experiments. B: Risk of bias of selected articles that included in vivo studies based on animal models using the SYRCLE risk of bias tool. C and D display the independent assignment of risk of bias for each study, in vitro and in vivo, respectively. The graphics illustrate the distribution of bias risks across various study components, with categories ranging from low to high risk.

Figure 4

Direct induction of EndMT and feedback mechanisms. Factors and signals received by endothelial cells from tumour cells, tumour-associated stromal cells (e.g., macrophages), and various microenvironmental conditions (e.g., hypoxia) can promote or enhance signaling pathways involved in the induction of EndMT. Furthermore, endothelial cells undergoing EndMT can provide feedback to these surrounding cells, significantly contributing to tumour progression and metastasis.

Figure 5

Inducers and molecular pathways of EndMT. A diagrammatic representation illustrating the induction of EndMT by various secreted factors and distinct molecular pathways. This figure highlights the complex interactions and signaling mechanisms involved in the transition of endothelial cells to a mesenchymal phenotype.

Figure 6

Gene expression heatmap showing the Log₂(Fold Change) of genes of interest across different cancer types. Expression data (TPM) were retrieved from the UALCAN database. The color scale was adjusted between values of -5 and 5 to enhance the visual representation of expression changes.

Figure 7

Kaplan-Meier survival curves illustrating the impact of gene expression levels on patient survival outcomes. Panel (A) shows overall survival (OS) and relapse-free survival (RFS) for colon cancer, and panel (B) shows the same outcomes for breast cancer. Panel (C) displays OS and first progression (FP) for gastric cancer, while panel (D) shows the same for lung cancer. The plots compare high (red line) vs. low (black line) expression. Each panel includes the hazard ratio (HR) and the associated log-rank p-value, indicating the statistical significance of expression differences on survival.