Abstract
Background/Aim: Prostate cancer, a leading global malignancy, exhibits variable progression influenced by angiogenesis, the formation of new blood vessels critical for tumor growth and metastasis. We investigated the impact of genetic variants of angiogenesis-related genes on the survival outcomes of patients with prostate cancer receiving androgen deprivation therapy (ADT).
Materials and Methods: We conducted a genetic association study of 87 single-nucleotide polymorphisms across seven angiogenic genes in 630 patients with prostate cancer undergoing ADT. Survival analysis was used to assess progression-free survival (PFS) and overall survival (OS). Functional analyses, including gene ontology and pathway enrichment, were performed to elucidate the underlying biological mechanisms.
Results: ANGPT2 rs2959822 was significantly associated with PFS [hazard ratio (HR)=1.22, p=0.015] and OS (HR=1.22, p=0.021). The minor allele A increased the risk of disease progression and mortality. Functional analyses revealed that rs2959822 influenced ANGPT2 expression. Elevated ANGPT2 expression was correlated with higher Gleason score, advanced tumor stage, and shorter PFS. Gene set enrichment analysis linked ANGPT2 to epithelial-mesenchymal transition (EMT), demonstrating positive correlations with several key EMT genes, along with increased immune cell infiltration, indicating its multifaceted oncogenic roles.
Conclusion: ANGPT2 rs2959822 influences the survival outcomes of patients with prostate cancer undergoing ADT. In addition to angiogenesis, ANGPT2 plays a critical role in prostate cancer progression by promoting EMT and modulating the tumor immune microenvironment.
Introduction
Prostate cancer represents a formidable global health challenge, characterized by its high incidence and complex management landscape. It is the most prevalent cancer among men across 112 countries, accounting for 15% of all cancer diagnoses (1), with an estimated 313,780 new cases and 35,770 deaths projected in the United States alone for 2025 (2). Globally, annual cases are expected to rise from 1.4 million in 2020 to 2.9 million by 2040, driven by aging populations and extended life expectancy (1). Disease progression varies widely, with localized cases boasting a 5-year survival rate of 99%, whereas metastatic disease plummets to 37% (3), underscoring the critical need for early detection and precise staging. Treatment options include active surveillance for low-risk cases, surgery and radiation for localized disease, and androgen deprivation therapy (ADT) or chemotherapy for advanced stages. However, challenges such as treatment resistance and side effects persist. The prediction of prostate cancer progression relies on clinical markers, such as prostate-specific antigen (PSA) levels, Gleason scores, tumor stages, and patient factors, including age and comorbidities, with genetic influences, such as mutations in BRCA1/2, ATM, and CHEK2 (4), contributing up to 60% of the risk and driving more aggressive phenotypes (5). These complexities highlight the urgent need for innovative research to refine diagnostics, optimize therapies, and address disparities in prostate cancer care worldwide (6).
Angiogenesis, the process of new blood vessel formation from pre-existing vasculature, is a cornerstone of cancer progression, enabling tumors to surpass a critical size of 1-2 mm3 by securing oxygen and nutrients essential for growth, invasion, and metastasis (7). This tightly regulated mechanism hinges on a delicate balance between pro-angiogenic factors, such as vascular endothelial growth factor (VEGF), fibroblast growth factor (FGF)-2, and transforming growth factor-β, and endogenous inhibitors, such as angiostatin and endostatin. In prostate cancer, hypoxia within the tumor microenvironment triggers an “angiogenic switch”, upregulating VEGF via hypoxia-inducible factor-1α, alongside other factors such as cyclooxygenase-2, which collectively stimulate endothelial cell migration, proliferation, and tube formation (8). These new vessels, stabilized by integrins and angiopoietins (ANGPTs), increase microvessel density, a prognostic marker associated with aggressive disease and poor outcomes (8). While inhibitors, such as angiostatin, induce endothelial apoptosis, their capacity to counteract this angiogenic drive is often overwhelmed in advanced tumors. Overexpression of VEGF, with its diverse isoforms, has emerged as a key driver strongly linked to prostate tumor expansion, metastatic potential (9), and immunosuppression through the upregulation of programmed death ligand 1, which hinders T-cell activity by fostering an immunosuppressive tumor microenvironment (10). Clinical efforts to target angiogenesis, exemplified by bevacizumab (a VEGF-A inhibitor), have yielded mixed results: a phase II trial in hormone-sensitive prostate cancer showed improved relapse-free survival, yet a phase III study in metastatic, castration-resistant cases found no survival benefit and heightened toxicity (11). Similarly, sunitinib and lenalidomide have shown limited efficacy in advanced disease, with the 27% discontinuation rate of sunitinib due to toxicity and lenalidomide being linked to worse survival, reflecting challenges such as pathway redundancy and resistance (11). These limited efficacies in late-stage disease contrast with modest efficacy in earlier stages, suggesting a stage-specific therapeutic promise. Furthermore, single-nucleotide polymorphisms (SNPs) in angiogenesis-related genes show inconsistent associations with prostate cancer risk and prognosis; some studies associate them with aggressive disease (12), whereas others report no clear link (13). Genome-wide association studies have identified a few SNPs in angiogenic genes, with only one near FGF receptor 2 associated with aggressive prostate cancer and an intronic variant in FGF10 linked to susceptibility, emphasizing the need for larger, conclusive studies (14, 15). Collectively, these findings suggest that angiogenesis is a critical yet challenging target in prostate cancer, underscoring the urgent need for further research to develop effective therapies.
Given the pivotal role of angiogenesis in cancer progression, we hypothesized that variants of angiogenesis-related genes might influence survival outcomes in patients with prostate cancer undergoing ADT. We conducted a genetic association study of 87 SNPs across seven key angiogenesis-related genes in a cohort of 630 patients. Using survival analysis, we identified genetic variants that were significantly associated with prostate cancer progression and all-cause mortality. In addition, functional analyses, including gene ontology and pathway enrichment, were performed to elucidate biological mechanisms, particularly the role of ANGPT2-responsive genes, in prostate cancer prognosis following ADT.
Materials and Methods
Study population and outcome evaluation. This investigation involved 630 individuals with confirmed prostate cancer who received ADT at three institutions in Taiwan: National Taiwan University Hospital, Kaohsiung Medical University Hospital, and Kaohsiung Veterans General Hospital (16, 17). Ethical approval was granted by the Institutional Review Board of the Kaohsiung Medical University Hospital (KMU-HIRB-2013132). The study complied with the Declaration of Helsinki and Good Clinical Practice standards. All participants provided written informed consent prior to enrollment. Clinical and pathological data were retrieved from the institutional medical records. The primary outcomes assessed were progression-free survival (PFS), measured as the interval from ADT initiation to disease progression (including biochemical recurrence, local or regional failure, nodal involvement, distant metastases, and prostate cancer-specific death), and overall survival (OS), defined as the time from the start of ADT to death from any cause. Disease progression occurred in 518 patients during a median follow-up of 165.8 months, and 413 deaths were recorded (18). Key clinical variables, such as age, PSA level at ADT initiation, clinical stage, Gleason score, PSA nadir, and time to PSA nadir, were significantly associated with PFS and OS (p<0.05).
SNP identification and genotyping. Haplotype-tagging SNPs (htSNPs) were chosen from seven genes involved in tumor angiogenesis – angiopoietin 1 (ANGPT1), angiopoietin 2 (ANGPT2), fibroblast growth factor 2 (FGF2), hypoxia-inducible factor 1 subunit alpha (HIF1A), integrin subunit alpha V (ITGAV), integrin subunit beta 3 (ITGB3), and vascular endothelial growth factor A (VEGFA) – using Haploview v4.2 and the tagger algorithm (19). The selection was based on data from the 1000 Genomes Project for Han Chinese populations (Beijing and Southern Han Chinese), targeting SNPs with a minor allele frequency (MAF) >0.05 and a linkage disequilibrium threshold of r2>0.8. Genomic DNA was isolated from 5 ml of whole blood collected in EDTA tubes using the QIAamp DNA Blood Maxi Kit (Qiagen, Valencia, CA, USA) following the manufacturer’s protocol. Samples were processed within 24 h of collection and stored at −80°C until analysis. Genotyping was performed at the National Center for Genome Medicine in Taiwan using the Affymetrix Axiom Genotyping Array (Thermo Fisher Scientific, Waltham, MA, USA) (20). For each sample, 200 ng of genomic DNA was amplified, fragmented and hybridized to an Affymetrix Axiom Genome-Wide TWB 2.0 Array Plate (containing 686,463 SNPs) following the manufacturer’s automated protocol. To monitor reproducibility, two blinded duplicate samples were included. Plates were washed, stained and scanned on an Affymetrix GeneTitan Multi-Channel instrument, and raw data were processed using the Axiom Analysis Suite. Genotype calling was performed using default parameters with standard quality control (QC) filters: dish-QC ≥0.82 and per-sample call rate ≥97%, and any sample or plate failing these criteria was reprocessed. SNPs were excluded from further analysis if they had a MAF <0.03, genotyping call rates <95%, or deviations from Hardy-Weinberg equilibrium (p<0.001). Finally, 87 htSNPs were retained for subsequent analyses. The average genotype call rate for these SNPs was 99.7%, and the concordance rate among the blind duplicate QC samples was 100%.
Bioinformatic analyses. To evaluate the functional implications of rs2959822, we utilized HaploReg v4.2 and the FIVEx database (21, 22). HaploReg provided insights into the potential regulatory effects of SNPs by analyzing their position relative to enhancer or promoter regions, transcription factor binding sites, and evolutionary conservation, drawing on ENCODE and Roadmap Epigenomics data. FIVEx facilitated expression quantitative trait locus (eQTL) and splicing quantitative trait locus (sQTL) analyses using linear regression to link rs2959822 genotypes with ANGPT2 expression and transcript isoforms across various tissues, including the prostate. To investigate ANGPT2 expression patterns and their clinical relevance in prostate cancer, we analyzed the GSE21032 dataset and The Cancer Genome Atlas Prostate Adenocarcinoma (TCGA PRAD) cohort. Molecular pathways associated with ANGPT2 were explored using LinkedOmics, which performed gene ontology and hallmark pathway enrichment via gene set enrichment analysis (GSEA) (23). Genes were ranked by Pearson’s correlation with ANGPT2 expression, and GSEA enrichment scores were calculated using a weighted Kolmogorov-Smirnov statistic (24). Statistical significance was assessed using 1,000 permutations, and false discovery rates (FDR) were corrected using the Benjamini-Hochberg method. The prognostic value of ANGPT2 and the epithelial-mesenchymal transition (EMT) markers, fibronectin 1 (FN1), collagen type I alpha 1 chain (COL1A1), and periostin (POSTN), was evaluated in the TCGA PRAD dataset. Additionally, ANGPT2 copy number alterations were examined using Genomic Identification of Significant Targets In Cancer (GISTIC) 2.0 on TCGA PRAD data, categorizing alterations as deep deletion (−2), arm-level deletion (−1), diploid (0), arm-level gain (1), or high amplification (≥2). Immune cell infiltration relative to ANGPT2 expression was assessed using the Tumor Immune Estimation Resource (TIMER), which estimates the abundance of B cells, CD4+ T cells, CD8+ T cells, macrophages, neutrophils, and dendritic cells based on TCGA PRAD gene expression profiles (25).
Statistical analyses. Data analysis was conducted using SPSS v19.0.0 (IBM, Armonk, NY, USA), with statistical significance defined as a two-sided p-value <0.05. Survival differences were evaluated using Kaplan-Meier curves and log-rank tests. Univariate and multivariate Cox regression models were used to estimate hazard ratios (HRs) and 95% confidence intervals (CIs) to assess the relationship between clinical factors or genotypes and patient outcomes. Correlations between ANGPT2 expression and tumor features were determined using Pearson’s and Spearman’s correlation coefficients.
Results
To explore the link between tumor angiogenesis and prostate cancer progression, we evaluated 87 SNPs across seven genes associated with this pathway, focusing on their effects on survival outcomes following ADT. Of these, three SNPs, ITGAV rs140419946, ANGPT2 rs2959822, and ANGPT1 rs146659513, were significantly associated with PFS, while two SNPs, ITGAV rs61763608 and ANGPT2 rs2959822, were significantly associated with OS (p<0.05; Figure 1). The SNP rs2959822 in ANGPT2 was significantly associated with PFS and OS. Specifically, the minor allele A of rs2959822, compared with the major allele G, increased the risk of disease progression by 22% (HR=1.22, 95% CI=1.04-1.42, p=0.015, Table I) and all-cause mortality by the same margin (HR=1.22, 95% CI=1.03-1.45, p=0.021). After adjusting for clinical variables in the multivariate analysis, rs2959822 remained an independent predictor of PFS and OS (p≤0.012, Table I).
Manhattan plots illustrating associations between 87 single-nucleotide polymorphisms (SNPs) in seven angiogenesis-related genes and survival outcomes in patients with prostate cancer on androgen-deprivation therapy. (A) Progression-free survival and (B) overall survival are shown as -log10(p) values on the Y-axis and SNP chromosomal positions on the X-axis. The blue line denotes the p=0.05 significance threshold. Significant SNPs are highlighted with red circles and labeled.
Association of ANGPT2 rs2959822 with progression-free and overall survival in patients with prostate cancer receiving androgen deprivation therapy.
To assess the functional relevance of rs2959822, we used HaploReg, which indicated that this intronic SNP lies within regions marked by enhancer histone signatures, DNase I hypersensitivity, and transcription regulatory elements across multiple tissues, suggesting a regulatory role (Figure 2A). Further evidence from the FIVEx database showed that rs2959822 influenced ANGPT2 expression in various tissues, including the prostate tissue in sQTL analysis (p=0.011; Figure 2B and C).
Functional analysis of ANGPT2 rs2959822. (A) HaploReg annotations showing the regulatory features of rs2959822. (B) Expression quantitative trait loci and (C) splicing quantitative trait loci analyses linking rs2959822 to ANGPT2 expression across tissues. The blue line marks p=0.05. Inverted triangles indicate a negative effect on ANGPT2 expression; circles denote non-significant effects. Sample sizes are noted in parentheses.
We examined ANGPT2 expression in prostate cancer using the GSE21032 and TCGA PRAD datasets. ANGPT2 levels were markedly elevated in primary and metastatic prostate cancers compared with those in normal tissues, with higher expression linked to increased Gleason scores in GSE21032 (p<0.001, Figure 3A). In the TCGA PRAD cohort, elevated ANGPT2 expression correlated with higher Gleason scores and more advanced tumor stages (p≤0.015, Figure 3B). Patients with higher ANGPT2 expression also experienced shorter PFS (p=0.022; Figure 3B), suggesting its role in driving tumor advancement.
Clinical relevance of ANGPT2 in prostate cancer. (A) GSE21032 and (B) TCGA PRAD datasets depict ANGPT2 expression trends with cancer progression (left), Gleason score (middle left), tumor stage (middle right), and progression-free survival (right). Spearman’s rho reflects correlation strength. Sample sizes are noted in parentheses.
To uncover the biological contributions of ANGPT2 in prostate cancer, we analyzed the genes correlated with its expression in TCGA PRAD data. We identified 3,327 genes with positive correlations and 2,164 with negative correlations (Pearson’s correlation FDR<0.01). GSEA of this ranked gene list revealed that positively correlated genes were enriched in cellular components such as protein complexes involved in cell adhesion, collagen-containing extracellular matrix, and collagen trimers (Figure 4A). These genes also participated in processes such as endothelium development, sprouting angiogenesis, and heart valve development (Figure 4B) and exhibited molecular functions, including extracellular matrix structural constituents, transmembrane receptor protein kinase activity, and growth factor binding (Figure 4C). Hallmark pathway analysis highlighted enrichment in the EMT and cell cycle progression, particularly in the G2/M checkpoint and mitotic spindle assembly (Figure 4D).
Enrichment analysis of the genes associated with ANGPT2 expression. The top 10 gene ontology terms for (A) cellular components, (B) biological processes, and (C) molecular functions are shown along with (D) the most enriched hallmark pathways. Bubble size reflects the gene ratio; color indicates FDR significance.
GSEA identified EMT as the most enriched pathway (normalized enrichment score=2.198, FDR <2.2×10−16), suggesting the involvement of ANGPT2 in this process. To confirm this, we assessed the correlation between ANGPT2 and key EMT hub genes in the protein-protein interaction network. ANGPT2 expression was positively correlated with the top three hub genes, FN1, COL1A1, and POSTN (p<0.001; Figure 5, left). COL1A1 and POSTN were significantly upregulated as prostate cancer progressed (p≤0.047; Figure 5, middle), and higher COL1A1 expression was associated with poorer survival (p=0.017; Figure 5, right), supporting the idea that ANGPT2 may promote oncogenesis via EMT gene activation.
Relationships between ANGPT2 and epithelial-mesenchymal transition hub genes in prostate cancer. Left: Positive correlation between ANGPT2 and (A) FN1, (B) COL1A1, and (C) POSTN. Middle: Expression of FN1, COL1A1, and POSTN during cancer progression. Right: Prognostic impact of FN1, COL1A1, and POSTN expression. Pearson’s r and Spearman’s rho reflects correlation strength. Sample sizes are noted in parentheses.
Considering the potential influence of ANGPT2 on the tumor immune microenvironment (TIME), we investigated its relationship with immune cell infiltration. ANGPT2 deletion was associated with decreased macrophages and neutrophils (Figure 6A), whereas higher ANGPT2 expression was positively correlated with the infiltration of B cells, CD8+ T cells, CD4+ T cells, macrophages, neutrophils, and dendritic cells in prostate cancer (Figure 6B). These observations suggest that ANGPT2 shapes immune cell dynamics in the tumor microenvironment.
ANGPT2 expression and immune cell infiltration in prostate cancer. (A) Reduced macrophage and neutrophil infiltration with ANGPT2 deletion compared with normal samples via the Wilcoxon rank-sum test (*p<0.05, **p<0.01). (B) Positive correlation between ANGPT2 expression and infiltration of B cells, CD8+ T cells, CD4+ T cells, macrophages, neutrophils, and dendritic cells.
Discussion
This study provides compelling evidence linking genetic variants of angiogenesis-related genes to survival outcomes in patients with prostate cancer undergoing ADT. Other studies have similarly highlighted the impact of genetic variants on ADT response, with those affecting polypyrimidine tract binding protein 1 and DNA methyltransferase 3 alpha influence clinical outcomes in patients receiving ADT (26, 27). As illustrated in Figure 7, ANGPT2 rs2959822 emerged as an independent predictor of both PFS and OS in our study, underscoring its potential clinical significance. Our findings suggest that this genetic variant influences ANGPT2 expression, which in turn correlates with aggressive tumor characteristics, including higher Gleason scores and advanced tumor stages. Moreover, the observed associations underscore the critical role of angiogenesis in prostate cancer progression and the potential of utilizing genetic markers to personalize ADT strategies. These results provide valuable insights into the complex interplay between genetic predisposition and the clinical outcomes of prostate cancer.
Study design and key findings.
SNP rs2959822, which resides within an intronic region of ANGPT2, exhibits chromatin modifications characteristic of enhancers, coupled with DNase I hypersensitive sites and alterations in transcription factor-binding motifs. These features strongly suggest a regulatory role of ANGPT2 expression. Our sQTL analysis confirmed that rs2959822 is associated with reduced expression of the ANGPT2 transcript ENST00000338312 in prostate tissues. Physiologically, ANGPT2 destabilizes blood vessels by antagonizing the TEK receptor tyrosine kinase Tie2, facilitating vascular remodeling; these processes are crucial for wound healing. However, in prostate cancer, ANGPT2 promotes tumor growth by enhancing the vascular supply (28). This oncogenic role is underscored by its marked elevation in aggressive metastatic prostate cancer, where it is correlated with increased vascular density, higher histological grade, and diminished survival, thereby establishing it as a significant prognostic marker. Furthermore, observations in gastric cancer, where ANGPT2 overexpression is linked to EMT markers such as vimentin (29), suggest a potential contribution to prostate cancer cell invasiveness. Soluble ANGPT2 has also been identified as a potential prognostic biomarker in colorectal cancer with peritoneal carcinomatosis, where elevated levels are negatively associated with overall survival (30). Notably, ANGPT2 significantly modulates the TIME, fostering immunosuppression. In melanoma, it upregulates PD-L1 expression in macrophages, impairing antitumor immunity (31), a mechanism likely relevant to the immune-evasive phenotype of prostate cancer. Mechanistically, ANGPT2 drives tumor progression by synergistically enhancing angiogenesis with VEGF, promoting EMT through stromal remodeling, and fostering an immunosuppressive niche via IL-10 and regulatory T-cell expansion (32). These effects, which have been observed in various cancers, appear to be applicable to prostate cancer. While clinical trials of the ANGPT2 inhibitor MEDI3617 have demonstrated promise in melanoma by enhancing immune checkpoint therapy through improved tumor perfusion (33), and genetic variants in ANGPT2 have been implicated in the treatment response in breast, hepatocellular, and colorectal cancers (34-36), a definitive link between specific ANGPT2 SNPs and prostate cancer prognosis remains unexplored.
GSEA of ANGPT2-associated expression networks revealed a significant positive correlation between ANGPT2 expression and key EMT genes, particularly COL1A1. This gene is significantly upregulated as prostate cancer progresses, and elevated COL1A1 expression is associated with poorer survival outcomes. COL1A1 drives tumor EMT by stiffening the extracellular matrix, thereby promoting cancer cell invasion. This mechanism is exemplified in colorectal cancer, where COL1A1 upregulates the Wnt/planar cell polarity pathway, enhancing mesenchymal traits through Rac1-GTP and RhoA-GTP signaling (37). In prostate cancer, COL1A1 overexpression likely contributes to EMT and accelerates cancer progression (38). In addition, COL1A1 is overexpressed in lung cancer, where it serves as a prognostic biomarker linked to poor outcomes and is correlated with immune cell infiltration, significantly affecting the TIME (39). These findings underscore the dual role of ANGPT2 in modulating the EMT and TIME, suggesting its potential as a therapeutic target. However, prostate-specific studies are essential to validate these effects and elucidate the precise mechanisms by which ANGPT2 and COL1A1 interact to drive prostate cancer progression.
This study has several strengths, including a well-characterized cohort of 630 ADT-treated patients with prostate cancer from multiple institutions, which enhances generalizability within this population. A comprehensive analysis of 87 htSNPs across seven angiogenesis-related genes, combined with robust survival and functional analyses, offers a detailed examination of genetic influences on prognosis. Long-term follow-up (median 165.8 months) bolsters the reliability of the outcomes, and integrating functional genomics data elucidates the potential mechanisms behind the identified genetic associations. However, limitations include its retrospective design and focus on a Taiwanese cohort, which may reduce its applicability to diverse ethnic groups owing to genetic differences. Restricting the analysis to preselected angiogenic genes might overlook other relevant pathways. This observational approach limits causal conclusions, and while the functional role of rs2959822 has been inferred bioinformatically, it lacks direct experimental confirmation. Finally, unmeasured confounders, such as lifestyle factors, may have affected the results, necessitating cautious interpretation and further validation.
Conclusion
The study highlights the multifaceted role of angiogenesis in prostate cancer progression, particularly through the lens of genetic variants and ANGPT2-related mechanisms in patients undergoing ADT. The identification of ANGPT2 rs2959822 as an independent predictor of PFS and OS highlights its potential as a prognostic biomarker, supported by its regulatory influence on ANGPT2 expression and correlation with aggressive disease features. Functional analyses further reveal the contribution of ANGPT2 to tumor advancement via EMT and TIME modulation, suggesting a broader oncogenic role beyond mere vascular support. These findings advocate for a deeper exploration of angiogenesis-targeting strategies, potentially refining therapeutic approaches to improve outcomes in prostate cancer, especially in ADT-treated cohorts, where resistance remains a critical challenge.
Acknowledgements
The Authors thank Chao-Shih Chen for data analysis, and the National Centre for Genome Medicine, Taiwan, for technical support. The results published here are based in part on data generated by the 1000 Genomes and TCGA projects.
Footnotes
Conflicts of Interest
The Authors declare that they have no potential conflicts of interest in regard to this study.
Authors’ Contributions
Yei-Tsung Chen: Conceptualization, Funding acquisition, Investigation, Visualization, Writing–original draft, Writing–review and editing. Pin-Yi Chen: Conceptualization, Data curation, Formal analysis, Writing–original draft, Writing–review and editing. Chi-Fen Chang: Investigation, Visualization, Writing–review and editing. Chao-Yuan Huang: Conceptualization, Methodology, Writing–review and editing. Chia-Cheng Yu: Conceptualization, Methodology, Writing–review and editing. Victor C. Lin: Conceptualization, Methodology, Writing–review and editing. Hao-Han Chang: Investigation, Visualization, Writing–review and editing. Te-Ling Lu: Formal analysis, Visualization, Writing–review and editing. Shu-Pin Huang: Conceptualization, Funding acquisition, Investigation, Visualization, Writing–original draft, Writing–review and editing. Bo-Ying Bao: Conceptualization, Data curation, Formal analysis, Funding acquisition, Writing–original draft, Writing–review and editing.
Funding
This work was supported by the National Science and Technology Council of Taiwan (grant nos: 110-2320-B-A49A-515, 110-2314-B-002-113, 111-2314-B-002-240-MY3, 111-2320-B-039-021-MY3, 111-2218-E-037-001, 112-2218-E-037-001, 113-2218-E-037-001, 112-2314-B-037-127, 113-2314-B-037-016, 114-2314-B-037-005, and 114-2314-B-037-033), the National Health Research Institute (grant no: NHRI-EX113-11313SI), the Kaohsiung Medical University (grant nos: KMUH111-1R58, KMUH112-2R59, and KMUH113-3R52), and the China Medical University (grant nos: CMU111-MF-09, CMU112-MF-10, and CMU113-MF-11). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Artificial Intelligence (AI) Disclosure
During the preparation of this manuscript, a large language model (Google Gemini) was used solely for language editing and stylistic improvements in select paragraphs. No sections involving the generation, analysis, or interpretation of research data were produced by generative AI. All scientific content was created and verified by the authors. Furthermore, no figures or visual data were generated or modified using generative AI or machine learning–based image enhancement tools.
- Received May 9, 2025.
- Revision received July 29, 2025.
- Accepted August 6, 2025.
- Copyright © 2025 The Author(s). Published by the International Institute of Anticancer Research.
This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY-NC-ND) 4.0 international license (https://creativecommons.org/licenses/by-nc-nd/4.0).
