Angiogenesis-associated immune genes as prognostic markers and predictors of immunotherapy response in pancreatic ductal adenocarcinoma

Angiogenesis-associated immune genes as prognostic markers and predictors of immunotherapy response in pancreatic ductal adenocarcinoma | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Angiogenesis-associated immune genes as prognostic markers and predictors of immunotherapy response in pancreatic ductal adenocarcinoma Qing Chang, Qian Wang, Xiumei Jiang, Yongmei Yang, Ailin Qu This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-9374081/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Immunotherapy has emerged as a pivotal approach in cancer treatment, yet its efficacy is influenced by the interactions within the tumor microenvironment (TME). Angiogenesis, the formation of new blood vessels, is a hallmark feature of the TME, particularly in aggressive malignancies such as pancreatic ductal adenocarcinoma (PDAC). This study aims to elucidate angiogenesis patterns in PDAC and investigate their associations with clinical characteristics, TME features and the response to immunotherapy. By analyzing 40 angiogenesis-related genes, 28 angiogenesis-associated immune genes (AIGs) were identified, enabling classification of PDAC patients into two subclusters with distinct clinical and TME profiles. Afterwards, we constructed an AIGScore risk model using least absolute shrinkage and selection operator (LASSO) regression in PDAC, and its reliable predictive ability was confirmed in both training and validation sets. Functional analyses revealed that a high AIGScore was associated with worse prognosis, reduced tumor immune cycle activity, elevated immune and stromal scores and diminished response to anti-PD-1 immunotherapy. Pan-cancer analyses further demonstrated the relevance of the seven AIGs to tumor progression and patient outcomes in other malignancies. In summary, this study introduces a novel AIGScore model with significant prognostic utility for PDAC. The findings provide insights into TME characteristics and offer a potential framework for optimizing immunotherapeutic strategies in patients with PDAC. Health sciences/Biomarkers Biological sciences/Cancer Biological sciences/Computational biology and bioinformatics Biological sciences/Immunology Health sciences/Oncology PDAC angiogenesis AIGScore tumor immune prognosis Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Figure 8 Figure 9 Introduction Pancreatic ductal adenocarcinoma (PDAC) is a highly malignant tumor with an alarming annual increase in incidence rates [ 1 ]. Most patients are diagnosed at advanced stages, characterized by early loco-regional spread and distant metastases, which preclude surgical resection and contribute to the disease’s poor prognosis. The 5-year overall survival rate remains dismal at approximately 10% [ 2 , 3 ]. Despite advancements in systemic chemotherapies such as 5-fluorouracil and gemcitabine combined with albumin-bound (nab) paclitaxel, as well as traditional radiotherapy for metastatic PDAC, their overall efficacy remains unsatisfactory [ 1 ]. Immunotherapy has emerged as a transformative treatment strategy, achieving remarkable outcomes in various cancers, including breast and lung cancers [ 4 ]. However, its application in PDAC is still limited, largely due to the unique immunosuppressive nature of the tumor microenvironment (TME) in this disease. PDAC’s TME is characterized by a scarcity of infiltrating effector T cells and an abundance of immunosuppressive cell [ 5 – 7 ], both of which diminish the effectiveness of immunotherapy. The TME comprises tumor cells, interstitial cells, infiltrating immune cells, nerves and blood vessels, with angiogenesis—a hallmark of the TME—playing a pivotal role. Angiogenesis, the formation of new blood vessels, is essential for tumor growth and progression as it provides oxygen and nutrients to neoplastic cells [ 8 ]. PDAC is a vascularized malignancy and its progression has been correlated with microvessel density within the tumor [ 9 ]. Overexpression of vascular endothelial growth factor (VEGF) has been shown to drive angiogenesis, further promoting PDAC progression [ 10 , 11 ]. Furthermore, tumor angiogenesis disrupts the interaction between immune cells and the vessel wall, facilitating immune evasion and contributing to the immune suppressive environment [ 12 ]. These findings highlight the critical interplay between angiogenesis and tumor immunity, which is crucial for understanding the immune landscape in PDAC. In light of these observations, a deeper understanding of the relationship between angiogenesis and the TME is essential for identifying distinct tumor immunophenotypes and improving the efficacy of immunotherapy in PDAC. In this study, we conducted a comprehensive analysis of angiogenesis-associated immune genes (AIGs) and explored their impact on PDAC development, survival outcomes, and the TME. Furthermore, we developed a novel AIGScore model to quantify angiogenesis patterns for individual PDAC patients, providing a potential framework to enhance immunotherapeutic strategies and optimize patient outcomes. Materials and methods Ethical statement All experimental protocols were approved by the Clinical Research Ethics Committee of Qilu Hospital, Shandong University (Approval No. KYLL-202209-044-1). Informed consent was obtained from all participants. All procedures were conducted in accordance with the relevant guidelines and regulations. Data and sample acquisition Processed transcriptomic data (HTSeq-Counts and HTSeq-FPKM) from the TCGA-PDAC database, comprising 176 tumor and 4 normal samples, were downloaded from the University of California Santa Cruz (UCSC) Xena browser. RNA-seq data for 167 normal pancreatic tissues were obtained from the Genotype-Tissue Expression (GTEx) database, with batch effect correction performed using the ComBat method. The resulting dataset of 176 tumor and 171 normal samples was designated as the training set. Additionally, 160 pancreatic cancer samples with comprehensive clinical data were obtained from two datasets: ICGC-PACA-AU-seq (ICGC-AU, n=81) and GEO (GSE85916, NCI cohort, n=79), forming the validation set. AIGs were identified from the GeneCards database using the term "angiogenesis." The top 40 genes with the highest relevance were selected for further analysis. Ten pairs of PDAC tissues and their adjacent normal tissues were collected from Qilu Hospital, Shandong University. Consensus clustering based on AIGs Consensus clustering of the TCGA-PDAC dataset was performed using the "ConsensusClusterPlus" package with the k-means algorithm. Classification stability was ensured through 1,000 iterations. The optimal cluster number was determined based on the item-consensus plot and relative changes in the area under the cumulative density function (CDF) curves. Characteristics of TME in two subclusters The CIBERSORT algorithm was applied to estimate the relative abundance of 22 immune cell types in TME [13]. The ESTIMATE algorithm was used to evaluate the proportions of infiltrating immune and stromal cells [14]. Differences in the cancer immunity cycle between subclusters were analyzed using the Tracking Tumor Immunophenotype (TIP) website [15]. Pathway enrichment differences were assessed using Gene Set Variation Analysis (GSVA), with the “c2.cp.kegg.v7.0.symbols.gmt” and “h.all.v7.0.symbols.gmt” datasets serving as references [16]. Development and validation of the AIGScore model In the TCGA-PDAC training set, 28 AIGs significantly associated with ImmuneScore ( P <0.05) were identified through univariate Cox regression analysis. These genes were further analyzed using the least absolute shrinkage and selection operator (LASSO) with ten-fold cross-validation to select the optimal lambda value. The AIGScore model was constructed as follows: AIGScore = expression of a gene β1×corresponding coefficient [β1] + expression of a gene β2 ×corresponding coefficient [β2] +…… expression of gene βi× corresponding coefficient [βi]. Patients were stratified into high and low AIGScore subgroups based on the median AIGScore. Kaplan-Meier survival analysis and time-dependent receiver operating characteristic (ROC) curve analysis were performed on the TCGA training set to evaluate the model’s prognostic efficacy. The ICGC and GEO datasets served as validation cohorts to confirm its robustness. Differences in TME and immune cell infiltration between AIGScore subgroups were assessed using ESTIMATE and CIBERSORT algorithms, respectively. Cancer immunotherapy response was predicted using the Tumor Immune Dysfunction and Exclusion (TIDE) algorithm [17] and the Immune Phenotype Score (IPS). Identification of hub genes related to angiogenesis DEGs between angiogenesis-related subgroups were identified using the "limma" package, with thresholds of |log2-fold change|≥2 and P value<0.05. Weighted correlation network analysis (WGCNA) based on the DEGs was performed using "WGCNA" package [18]. A soft-thresholding power of 2 was selected to ensure a scale-free network. The adjacency matrix was transformed into a topological overlap matrix (TOM) to describe gene co-expression relationships, with dissimilarity represented by 1-TOM. Modules were identified using the dynamic hybrid tree-cutting algorithm, with a minimum module size of 30. Among the seven modules generated, the blue module exhibited the strongest correlations with OS.time, AIGScore and ESTIMATE indices. Hub genes were identified within this module using Cytoscape and the cytoHubba plugin. Correlations between hub genes and immune infiltration were calculated using the "corrplot" package. Online databases information The relationship between somatic cell copy number alterations (CNAs) of seven AIGs and tumor-infiltrating immune cells was analyzed using the TIMER database. The average expression distribution of the seven AIGs was obtained from the TISCH database, while trajectory analysis was conducted through the TIGER database. Gene survival analysis was performed using the Kaplan-Meier Plotter website [19]. Detailed information about these online databases is provided in Table S1. Real-time quantitative PCR and immunohistochemistry Total RNA was extracted using TRIzol reagent (Invitrogen). RNA concentration and quality were assessed with a NanoDrop spectrophotometer (Thermo Fisher Scientific, Waltham, MA, USA). cDNA synthesis was carried out using HiScript III RT SuperMix for qPCR (Vazyme, Nanjing, China). RT-qPCR was performed using ChamQ Universal SYBR qPCR Master Mix (Vazyme, Nanjing, China). Primer sequences for each gene are listed in Table S2. Immunohistochemical (IHC) staining images of AIGs were retrieved from the Human Protein Atlas (HPA) database for validation. Statistical analysis All statistical analyses were conducted using R software (version 4.0.2) and GraphPad Prism 8. Correlations between variables were assessed using Pearson's correlation test. Box plots were generated with the Wilcoxon rank-sum test. Survival analyses were performed using the Kaplan-Meier method, with statistical differences evaluated by the log-rank test through the "survminer" and "survival" R packages [20]. A two-tailed P <0.05 was considered statistically significant. Results Landscape of angiogenesis-immune regulators in PDAC First, the top 40 genes most significantly associated with angiogenic activity were selected from the GeneCards database (Table S3). To interrogate the functional role of angiogenesis in tumor immunity in PDAC patients, we calculated ESTIMAT indices of each patient and assessed the correlation between the 40 angiogenesis-related genes and ESTIMATE indices. As a result, 28 genes with a correlation coefficient absolute value >0.3 were identified as the AIGs (Figure 1A). The immune infiltration patterns of these 28 AIGs were further analyzed using the CIBERSORT algorithm, which revealed strong associations between the AIGs and the abundance of 22 types of immune cells (Figure S1). These findings suggest that the identified AIGs exhibit significant immunogenicity within the tumor microenvironment of PDAC. RNA-seq data from the TCGA-PDAC and GTEx databases were analyzed to compare AIG expression between tumor tissues (n = 176) and normal tissues (n = 171). Twenty-five AIGs were found to be differentially expressed, with most showing elevated expression in tumor specimens ( P <0.05, Figure 1B). To evaluate somatic mutation frequencies, gene mutation data for the 28 AIGs were visualized using the cBioPortal database. Genetic alterations were present in 66% of PDAC samples, with FLT4 exhibiting the highest mutation frequency (10%), followed by ANGPT1 and MMP9 (both at 9%) (Figure 1C). A protein-protein interaction (PPI) network was constructed for the 28 AIGs, identifying VEGFA as a central hub within the network (Figure 1D). This tight connectivity suggests both direct (physical) and indirect (functional) interactions among the AIGs, highlighting their potential cooperative roles in angiogenesis and immunity regulation. Cluster analysis based on AIGs in PDAC To systematically classify tumor phenotypic traits in PDAC, consensus clustering analysis was performed on the TCGA-PDAC dataset using the expression profiles of the 28 AIGs. The empirical CDF plots and the consensus matrix identified k = 2 as the optimal number of clusters (Figure S2A-D). As shown in Figure 2A, 176 PDAC patients were divided into two subclusters: Cluster 1 (n = 88) and Cluster 2 (n = 88). Differential expression analysis between the two subclusters revealed 103 upregulated and 57 downregulated genes in Cluster 2 compared to Cluster 1 (Table S4, Figure 2B). Moreover, substantial differences in AIG expression and clinicopathological features were observed between the two subclusters (Figure 2C), indicating distinct tumor microenvironment characteristics. Kaplan-Meier survival analysis demonstrated a significant difference in survival probability between the two subclusters. Patients in Cluster 1 exhibited better overall survival compared to those in Cluster 2 (P = 0.012, Figure 2E), suggesting that the AIG-based clustering effectively stratifies patients by phenotypic and clinical outcomes. Characteristics of TME in two subclusters GSVA indicated that pathways associated with cancer progression and metastasis-such as angiogenesis, TGF-β, hypoxia, KRAS and P53 pathways-were significantly enriched in Cluster 2 (Figure 2D). This enrichment suggests that Cluster 2 exhibits an elevated inflammatory and immune-active status. The immune cell landscape was further analyzed using the CIBERSORT algorithm. Cluster 1 displayed higher proportions of CD8+ T cells, CD4+ memory resting T cells, activated NK cells, resting mast cells, resting dendritic cells and monocytes, which are generally associated with antitumor immunity. Conversely, Cluster 2 exhibited greater infiltration of follicular helper T cells, M0 macrophages, T regulatory (Treg) cells and M2 macrophages, the latter two being immunosuppressive components that can promote tumor progression (Figure 2F). Anti-cancer immune responses progress through a series of steps [21] and Step 2 (cancer antigen presentation) was notably active in Cluster 1 ( P <0.05; Figure 2G). Interestingly, cluster 1 exhibited lower levels of both pro-tumor and anti-tumor immune signatures compared to Cluster 2 (Figure 2H), underscoring a dual role of angiogenesis in cancer development. Cytokine analysis revealed that most cytokines were upregulated in Cluster 2 (Figure 2I), highlighting its more inflammatory microenvironment. No significant differences in tumor mutational burden (TMB) were observed between the two subclusters (Figure S2E). However, cluster 2 had higher immune, stromal and ESTIMATE scores but lower tumor purity than cluster 1 ( P <0.05; Figure S2F), further supporting the notion of a more immunologically and stromally active TME in Cluster 2. Development and validation of the AIGScore model in TCGA-PDAC set In the TCGA-PDAC dataset, univariate Cox regression analysis identified nine AIGs significantly associated with OS ( P <0.05; Figure 3A). These genes were subjected to LASSO Cox regression, resulting in an AIGScore model comprising seven genes (Figure 3B-C, Figure S3A). The AIGScore was calculated as follows: AIGScore= (0.289*ExpFGF2) + (-0.731*ExpADGRB3) + (0.0839*ExpMMP2) + (0.019*ExpTYMP) + (0.483*ExpVASH1) + (-0.068*ExpVEGFD) + (0.184*ExpITGAV). Patients were stratified into AIGScore-high and AIGScore-low groups based on the median AIGScore value (Figure 3D-E). A negative correlation was observed between the AIGScore and survival time (R=-0.481, P <0.001, Figure 3F). Cluster comparison showed that Cluster 2 had significantly higher AIGScores than Cluster 1 (Figure 3G), aligning with its more aggressive pathological features. The alluvial plot (Figure 3H) illustrated the association between AIGScore groups and clinical characteristics, revealing significant differences in T, N, TNM staging and patient outcomes across subgroups (all P <0.05, Figure 3I). These results collectively suggest that higher AIGScores are associated with more malignant features and poorer prognosis. Kaplan-Meier survival analysis confirmed that patients in the AIGScore-high group had significantly lower OS compared to those in the AIGScore-low group (log-rank test, P <0.001, Figure 4A). Time-dependent ROC analysis demonstrated the robust prognostic performance of the AIGScore model, with AUC values of 0.714, 0.725 and 0.740 at 1, 3 and 5 years, respectively (Figure 4B). The AIGScore model was further validated in independent datasets (ICGC and GEO). Consistent with the TCGA-PDAC dataset, the AIGScore-high group exhibited worse OS than the AIGScore-low group in both the ICGC dataset ( P =0.04, Figure 4C) and the GEO dataset ( P =0.009, Figure 4E). Time-dependent ROC analysis also confirmed the predictive efficacy of the model in these validation datasets (Figure 4D, 4F). Assessment of immunity activity and checkpoints in distinct subgroups We investigated the disparities in anticancer immune responses between the AIGScore-high and AIGScore-low groups. The analysis revealed that step 1 of the immune response was significantly elevated in the AIGScore-high group, whereas steps 3-5 were more pronounced in the AIGScore-low group (Figure 5A, B; Table S5). Notably, no significant differences were observed in other steps (Figure S3B-D). Using the ESTIMATE algorithm, we evaluated the correlation between AIGScore and immune cell enrichment. Stromal score, immune score and ESTIMATE score were positively correlated with AIGScore, while tumor purity showed a negative correlation (Figure 5C). These findings highlight the intricate relationship between AIGScore and the tumor microenvironment. To further assess the immunotherapeutic potential, TIDE was employed to predict patient responses within the TCGA-PADC cohort. The results demonstrated that the AIGScore-low group exhibited significantly lower TIDE levels compared to the AIGScore-high group ( P <0.05; Figure 5D), suggesting a superior immunotherapy response in the AIGScore-low group. Moreover, analysis using the IPS algorithm revealed that the AIGScore-low group benefit more from anti-PD1 therapy ( P <0.01; Figure 5E), underscoring its therapeutic implications. Identification of hub genes related to angiogenesis To explore the biological underpinnings of the AIGScore model, we identified 2,556 DEGs between the AIGScore-high and AIGScore-low groups (Figure S3E). These DEGs were utilized to construct gene co-expression networks via weighted correlation network analysis (WGCNA) (Figure S3F-G). By correlating phenotypes defined by OS.time, AIGScore and ESTIMATE indices, we identified seven distinct WGCNA modules (Figure S3H). Among these, the blue module was selected due to its strong correlation with the immune score (r=0.81, P =8 * 10 -40 ) and negative correlation with the risk score (r=-0.41, P =3 *10 -8 ) (Figure S3I-K). Thereafter, we obtained eight hub genes -CD19, PTPRC, CCR7, CD28, IL6, IL13, CD22 and CD40LG-from the blue module. Correlation analysis revealed strong positive associations among these genes (Figure S3L). Additionally, these hub genes exhibited significant positive correlations with the infiltration levels of naive B cells, CD8 T cells and CD4 memory-activated T cells, reinforcing their role in immune regulation (Figure S3M). When comparing AIGScore subgroups, the expression levels of all eight hub genes were significantly higher in the AIGScore-low group (all P <0.05; Figure S3N), implying these genes might play pivotal roles in orchestrating immune activity and angiogenesis within the tumor microenvironment. These findings collectively illustrate the potential of AIGScore as a biomarker for immunological activity and its intricate link to angiogenic processes. Characteristics of the AIGScore model in TME at single-cell landscape Given the relevance of the AIGScore model to the TME and prognosis, we analyzed the heterogeneity of angiogenic enrichment scores at the single-cell level. Fibroblast cells, malignant cells and myeloid cells exhibited higher angiogenic enrichment scores among all cell clusters (Figure 6A-C). We further examined the average expression distribution of the seven AIGs in CRA001160 set. Results showed MMP2 had moderate to high expression in fibroblast and endothelial cells (Figure 6D-E) and TYMP was predominantly expressed in macrophages and monocytes (Figure 6F). FGF2, ADGRB3 and VEGFD were enriched in stromal and immune cells but showed lower expression in malignant cells (Figure S4A-C). VASH1 was highly expressed in dendritic cells, macrophages, monocytes and endothelial cells (Figure 6G). ITGAV displayed broad expression across various cell types (Figure 6H). The expression patterns of these seven AIGs were consistent in the GSE111672 dataset, as detailed in Figure S4D-K. From a functional perspective, anticancer immune pathway analysis indicated heightened activity in the trafficking of immune cells to tumors and immune cell infiltration in the AIGScore-low group (Figure 6I). Meanwhile, GSVA analysis showed pathways such as angiogenesis, TGF-β, hypoxia and MYC targets were significantly upregulated in the AIGScore-high group (Figure 6J). The CNA analysis further revealed that these seven AIGs significantly influenced the infiltration of B cells, CD4+ T cells, CD8+ T cells, macrophages, neutrophils and dendritic cells into the tumor microenvironment (Figure 6K-N, Figure S4L-N). Considering the abundance of TYMP and MMP2 at the single-cell level, we performed trajectory analysis. The results highlighted an imbalance in their expression across different differentiation trajectory states (Figure 6O-Q). Additionally, analysis in the TISIDB database confirmed strong correlations between the seven AIGs and immune-related molecular subtypes (Figure S4O). These findings suggest that the seven AIGs exert pivotal regulatory effects on immune cells and the antitumor immune cycle, emphasizing their integral role in both angiogenesis and immune modulation within the TME. The relationship between AIGs and antitumor immune pathway in pattern map To elucidate the molecular mechanisms linking angiogenesis and the tumor immune cycle, we developed a pattern map summarizing our findings (Figure 7). The map highlights the critical roles of TYMP-enriched macrophages and MMP2-enriched fibroblasts in the angiogenic pathways. These cells not only enhance immune cell trafficking to tumors but also facilitate immune cell infiltration, thereby impacting the antitumor immune cycle. Using the CTD database, we predicted potential therapeutic agents targeting TYMP and MMP2. The results are summarized in Table S6, offering insights into therapeutic strategies aimed at modulating these key angiogenesis-related genes. These results underscore the dual roles of angiogenesis in promoting immune cell activity and shaping the TME, revealing potential targets for therapeutic intervention to enhance antitumor immunity. Pan-cancer analysis of the AIGs expression and their prognostic value Given that angiogenesis and immunity are not unique to PDAC, we hypothesized that the seven AIGs may also play a role in the tumor status and prognosis of other cancers. To test this hypothesis, we conducted a pan-cancer analysis using the TCGA dataset, examining the expression profiles of the seven AIGs across a broad spectrum of cancers. The results revealed that the seven AIGs were differentially expressed between cancerous and normal tissues in most cancer types (Figure 8A-D; Figure S5A-G). Furthermore, we found that the expression levels of these AIGs were significantly correlated with clinical outcomes in each cancer type (Figure 8E-H; Figure S5H-J). These findings suggest that AIGs may serve as important biomarkers for prognosis across various cancers, reinforcing their potential relevance beyond PDAC. Validation of AIGs expression levels in PDAC and paired adjacent tissues To validate the expression of the seven AIGs in PDAC, we performed RT-qPCR and immunohistochemistry on PDAC tissues and paired adjacent normal tissues. The RT-qPCR analysis revealed that the expression levels of FGF2, MMP2, TYMP, VASH1 and ITGAV were significantly higher in PDAC tissues compared to adjacent normal tissues ( P <0.05). The differences in ADGRB3 ( P =0.07) and VEGFD ( P =0.05) expression, although not statistically significant, showed a trend towards higher levels in PDAC tissues (Figure 9A-G). These findings were further corroborated by immunohistochemistry, where we observed consistent expression patterns for FGF2, MMP2, TYMP, VASH1, ITGAV and VEGFD (ADGRB3 data was not available) (Figure S6). Additionally, we conducted correlation analysis between the expression levels of the seven AIGs and various laboratory indicators. We found significant correlations between these genes and several clinical indicators, including lymphocyte and monocyte counts, as well as tumor markers such as CEA and CA125 (Figure 9H). These associations further suggest that the expression of AIGs is linked to key immunological and pathological features of PDAC, providing valuable insights into their potential role as biomarkers for both diagnosis and prognosis in PDAC. Discussion Angiogenesis, a key process in the early stages of tumor progression, is widely recognized as a hallmark of cancer biology [8]. A growing body of evidence underscores the intricate relationship between angiogenesis and immunity, suggesting that targeting angiogenesis could enhance the efficacy of immunotherapies [22, 23]. However, the precise mechanisms linking angiogenesis-related genes and tumor-associated immune infiltrates remain poorly understood. In this study, we utilized single-cell RNA sequencing and multi-omics data to systematically investigate the role of AIGs in tumor immunity, focusing on their prognostic value. We developed an AIGScore model to quantify angiogenesis patterns in PDAC, with the goal of assisting clinicians in devising more effective immunotherapeutic strategies for PDAC patients. Our analysis of gene expression and mutation data for 28 AIGs from the TCGA-PDAC and GTEx databases revealed elevated levels of most AIGs in PDAC tissues, with 66% of PDAC samples exhibiting genetic alterations. Through consensus clustering analysis, we classified PDAC patients into two distinct subclusters: Cluster 1 and Cluster 2. Notably, significant differences in clinical features, immune activities and functional pathways were observed between these subclusters. Angiogenesis has long been implicated in the malignant behavior of various cancers, including PDAC [24-27]. Tumor-associated macrophages, for instance, promote tumor cell migration via VEGF signaling [28], while numerous proangiogenic molecules are linked to immunosuppressive effects within the TME [29-31]. Disruption of immune cell function within the PDAC TME is a key driver of tumor progression [32, 33]. Consistent with previous studies, our GSVA results showed that cancer-related and metastasis-associated pathways were enriched in Cluster 2, which also exhibited higher expression of inflammatory and immunosuppressive signatures. These findings highlight the pivotal role of AIGs in shaping the immunogenic landscape of the PDAC TME. To better understand the prognostic significance of angiogenesis in PDAC, we developed the AIGScore risk model using LASSO regression. The AIGScore was strongly correlated with clinicopathological features of PDAC and stratified analysis demonstrated its independent predictive value for patient survival outcomes. Time-dependent ROC analysis confirmed that the AIGScore model reliably predicted prognosis in both training and validation sets. While previous studies have explored prognostic models in PDAC using other gene panels—such as Luo et al.’s necroptosis-related genes [34]and Chen et al.’s hypoxia-immune signature [35]—our study is the first to construct a prognostic model using seven angiogenesis- and immunity-related genes. We found that PDAC patients with higher AIGScores had elevated immune and stromal scores, along with poorer survival outcomes. High immune and stromal scores have been previously associated with adverse clinicopathological features and poor prognosis in PDAC [36]. These results suggest that AIGScore could serve as a robust marker of PDAC immunogenicity, potentially guiding therapeutic decision-making. Our study further explored the relationship between the AIGScore and response to immunotherapy. Antiangiogenic therapies have been shown to inhibit T cell infiltration and antigen presentation by dendritic cells, potentially leading to immune suppression within the TME [23, 37-39]. Interestingly, our analysis using TIDE and IPS signatures indicated that PDAC patients with low AIGScores had lower TIDE scores and showed better responses to anti-PD1 therapies compared to those with high AIGScores. This finding supports the notion that the AIGScore model can identify PDAC patients who are more likely to benefit from immune checkpoint blockade therapy. Our constructed AIGScore model contained seven AIGs: FGF2, ADGRB3, MMP2, TYMP, VASH1, VEGFD and ITGAV. Several studies have implicated FGF2, MMP2 and TYMP in the proliferation, invasion and migration of PDAC cells [40-42]. In our study, single-cell analysis confirmed that TYMP and MMP2, two critical angiogenic regulators, were specifically expressed in macrophages and fibroblasts within the PDAC TME. Furthermore, CNA analysis revealed that these seven AIGs significantly influenced the infiltration of key immune cells—including B cells, CD4+ T cells, CD8+ T cells, macrophages, neutrophils and dendritic cells—into the tumor. T cells, which mediate tumor cell killing, are essential for maintaining the cancer-immune cycle [43]. Previous studies have demonstrated that macrophages and fibroblasts not only promote tumor angiogenesis but also contribute to vascular stabilization [44, 45]. Based on these findings, we hypothesize that TYMP-high macrophages and MMP2-high fibroblasts play an important role in facilitating T cell infiltration into tumors via angiogenesis, thereby enhancing the antitumor immune response. Despite the strengths of this study, several limitations must be acknowledged. First, while our findings suggest that the AIGScore model may have significant clinical utility in PDAC, further validation in independent patient cohorts and clinical trials is necessary. Second, while we have established an association between angiogenesis-modified immune cell infiltration and survival, the underlying molecular mechanisms driving these interactions remain unclear and warrant further investigation. Additionally, the relatively small sample size of PDAC cases in our study necessitates validation in larger, more diverse cohorts. Therefore, a multi-center, retrospective study is warranted to confirm these findings. Conclusions In conclusion, we have developed a prognostic AIGScore model that effectively stratifies the risk of PDAC patients based on angiogenesis-related immune molecules. Our results suggest that TYMP and MMP2 are critical regulators of T cell trafficking and infiltration in the TME and that assessing individual tumor angiogenesis patterns can provide valuable insights into the TME characteristics. This model holds promise for guiding the development of personalized immunotherapeutic strategies for PDAC, enhancing treatment outcomes through targeted modulation of angiogenesis and immune pathways. Abbreviations AIGs, angiogenesis-associated immune genes; DEG, differentially expressed gene; KM, Kaplan–Meier curves; LASSO, least absolute shrinkage and selection operator; TME, tumor microenvironment; PDAC, pancreatic ductal adenocarcinoma; TIDE, Tumor Immune Dysfunction and Exclusion; WGCNA, weighted correlation network analysis. Declarations Funding No funding was used for the study. Acknowledgements Not applicable for that section. Author contributions AQ conceived and designed the experiments. QC and QW performed all the experiments. XJ analyzed the data. AQ and YY wrote the manuscript. All authors read and approved the final manuscript. Data availability All data and R codes in this study are reasonably available from the corresponding author. The data for this study were obtained from the publicly available databases. Competing interests The authors have no conflicts of interest to declare Ethics declarations All procedures involving human participants were approved by the ethical standards of the Clinical Research Ethics Committee of Qilu Hospital, Shandong University and performed in accordance with the Declaration of Helsinki. References Park, W., Chawla, A. & O'Reilly, E. M. A review of pancreatic cancer-reply. JAMA 326 , 2436–2437 (2021). Siegel, R. L., Giaquinto, A. N. & Jemal, A. Cancer statistics, 2024. CA Cancer J. Clin. 74 , 12–49 (2024). Cancer stat facts: pancreatic cancer. (2021). Available at: https://seer.cancer.gov/statfacts/html/pancreas.html van den Bulk, J., Verdegaal, E. M. & de Miranda, N. F. Cancer immunotherapy: broadening the scope of targetable tumours. Open. Biol. , 8. (2018). Royal, R. E. et al. Phase 2 trial of single agent ipilimumab (anti-ctla-4) for locally advanced or metastatic pancreatic adenocarcinoma. J. Immunother . 33 , 828–833 (2010). Menezes, S., Okail, M. H., Jalil, S., Kocher, H. M. & Cameron, A. Cancer-associated fibroblasts in pancreatic cancer: new subtypes, new markers, new targets. J. Pathol. 257 , 526–544 (2022). Pawlik, T. M. Pancreatic cancer. Surg. Oncol. Clin. N Am. , 30:xiii–xv. (2021). Hanahan, D. Hallmarks of cancer: new dimensions. Cancer Discov . 12 , 31–46 (2022). Dineen, S. P. et al. The adnectin ct-322 is a novel vegf receptor 2 inhibitor that decreases tumor burden in an orthotopic mouse model of pancreatic cancer. BMC Cancer . 8 , 352 (2008). Sun, Y. et al. Toll-like receptor 4 promotes angiogenesis in pancreatic cancer via pi3k/akt signaling. Exp. Cell. Res. 347 , 274–282 (2016). Veikkola, T., Karkkainen, M., Claesson-Welsh, L. & Alitalo, K. Regulation of angiogenesis via vascular endothelial growth factor receptors. Cancer Res. 60 , 203–212 (2000). Zhou, X., Tang, J., Cao, H., Fan, H. & Li, B. Tissue resident regulatory t cells: novel therapeutic targets for human disease. Cell. Mol. Immunol. 12 , 543–552 (2015). Newman, A. M. et al. Robust enumeration of cell subsets from tissue expression profiles. Nat. Methods . 12 , 453–457 (2015). Yoshihara, K. et al. Inferring tumour purity and stromal and immune cell admixture from expression data. Nat. Commun. 4 , 2612 (2013). Xu, L. et al. Tip: a web server for resolving tumor immunophenotype profiling. Cancer Res. 78 , 6575–6580 (2018). Hanzelmann, S., Castelo, R. & Guinney, J. Gsva: gene set variation analysis for microarray and rna-seq data. BMC Bioinform. 14 , 7 (2013). Jiang, P. et al. Signatures of t cell dysfunction and exclusion predict cancer immunotherapy response. Nat. Med. 24 , 1550–1558 (2018). Langfelder, P. & Horvath, S. Wgcna: an r package for weighted correlation network analysis. BMC Bioinform. 9 , 559 (2008). Gyorffy, B. Survival analysis across the entire transcriptome identifies biomarkers with the highest prognostic power in breast cancer. Comput. Struct. Biotechnol. J. 19 , 4101–4109 (2021). Rich, J. T. et al. A practical guide to understanding kaplan-meier curves. Otolaryngol. Head Neck Surg. 143 , 331–336 (2010). Chen, D. S. & Mellman, I. Oncology meets immunology: the cancer-immunity cycle. Immunity 39 , 1–10 (2013). Rahma, O. E. & Hodi, F. S. The intersection between tumor angiogenesis and immune suppression. Clin. Cancer Res. 25 , 5449–5457 (2019). Anker, J. F. et al. Multi-faceted immunomodulatory and tissue-tropic clinical bacterial isolate potentiates prostate cancer immunotherapy. Nat. Commun. 9 , 1591 (2018). Zhang, L. et al. Endothelial dgkg promotes tumor angiogenesis and immune evasion in hepatocellular carcinoma. J. Hepatol. 80 , 82–98 (2024). Jiang, X. et al. The role of microenvironment in tumor angiogenesis. J. Exp. Clin. Cancer Res. 39 , 204 (2020). Annese, T., Tamma, R., De Giorgis, M. & Ribatti, D. Micrornas biogenesis, functions and role in tumor angiogenesis. Front. Oncol. 10 , 581007 (2020). Sajib, S., Zahra, F. T., Lionakis, M. S., German, N. A. & Mikelis, C. M. Mechanisms of angiogenesis in microbe-regulated inflammatory and neoplastic conditions. Angiogenesis 21 , 1–14 (2018). Penny, H. L. et al. Warburg metabolism in tumor-conditioned macrophages promotes metastasis in human pancreatic ductal adenocarcinoma. Oncoimmunology 5 , e1191731 (2016). Fukumura, D., Kloepper, J., Amoozgar, Z., Duda, D. G. & Jain, R. K. Enhancing cancer immunotherapy using antiangiogenics: opportunities and challenges. Nat. Rev. Clin. Oncol. 15 , 325–340 (2018). Tao, L., Huang, G., Shi, S. & Chen, L. Bevacizumab improves the antitumor efficacy of adoptive cytokine-induced killer cells therapy in non-small cell lung cancer models. Med. Oncol. 31 , 777 (2014). Shrimali, R. K. et al. Antiangiogenic agents can increase lymphocyte infiltration into tumor and enhance the effectiveness of adoptive immunotherapy of cancer. Cancer Res. 70 , 6171–6180 (2010). Liu, L. et al. Low intratumoral regulatory t cells and high peritumoral cd8(+) t cells relate to long-term survival in patients with pancreatic ductal adenocarcinoma after pancreatectomy. Cancer Immunol. Immunother . 65 , 73–82 (2016). Duewell, P. et al. Rig-i-like helicases induce immunogenic cell death of pancreatic cancer cells and sensitize tumors toward killing by cd8(+) t cells. Cell. Death Differ. 21 , 1825–1837 (2014). Luo, Z. et al. A panel of necroptosis-related genes predicts the prognosis of pancreatic adenocarcinoma. Transl Oncol. 22 , 101462 (2022). Chen, D. et al. Development and verification of the hypoxia- and immune-associated prognostic signature for pancreatic ductal adenocarcinoma. Front. Immunol. 12 , 728062 (2021). Li, T., Liu, Q., Zhang, R., Liao, Q. & Zhao, Y. Identification of prognosis-related genes and construction of multi-regulatory networks in pancreatic cancer microenvironment by bioinformatics analysis. Cancer Cell. Int. 20 , 341 (2020). Ohm, J. E. & Carbone, D. P. Vegf as a mediator of tumor-associated immunodeficiency. Immunol. Res. 23 , 263–272 (2001). Voron, T. et al. Vegf-a modulates expression of inhibitory checkpoints on cd8 + t cells in tumors. J. Exp. Med. 212 , 139–148 (2015). Albini, A., Bruno, A., Noonan, D. M. & Mortara, L. Contribution to tumor angiogenesis from innate immune cells within the tumor microenvironment: implications for immunotherapy. Front. Immunol. 9 , 527 (2018). Tao, L. et al. Integrated proteomics and phosphoproteomics reveal perturbed regulative pathways in pancreatic ductal adenocarcinoma. Mol. Omics . 17 , 230–240 (2021). Han, Y., Ma, L., Zhao, L., Feng, W. & Zheng, X. Rosmarinic inhibits cell proliferation, invasion and migration via up-regulating mir-506 and suppressing mmp2/16 expression in pancreatic cancer. Biomed. Pharmacother . 115 , 108878 (2019). Coleman, S. J. et al. Nuclear translocation of fgfr1 and fgf2 in pancreatic stellate cells facilitates pancreatic cancer cell invasion. EMBO Mol. Med. 6 , 467–481 (2014). Song, K. H. et al. Hsp90a inhibition promotes anti-tumor immunity by reversing multi-modal resistance and stem-like property of immune-refractory tumors. Nat. Commun. 11 , 562 (2020). Annese, T. et al. Inflammatory infiltrate and angiogenesis in mantle cell lymphoma. Transl Oncol. 13 , 100744 (2020). Mao, X. et al. Crosstalk between cancer-associated fibroblasts and immune cells in the tumor microenvironment: new findings and future perspectives. Mol. Cancer . 20 , 131 (2021). Additional Declarations No competing interests reported. Supplementary Files SupplementaryInformation.pdf TableS1S6.xlsx Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-9374081","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":634244814,"identity":"aa436ad4-5ec0-443e-84f5-d68d18915946","order_by":0,"name":"Qing Chang","email":"","orcid":"","institution":"Qilu Hospital, Shandong University","correspondingAuthor":false,"prefix":"","firstName":"Qing","middleName":"","lastName":"Chang","suffix":""},{"id":634244815,"identity":"67a14d97-876f-4192-b8c3-d301a52e3ab6","order_by":1,"name":"Qian Wang","email":"","orcid":"","institution":"Central Hospital Affiliated to Shandong First Medical University","correspondingAuthor":false,"prefix":"","firstName":"Qian","middleName":"","lastName":"Wang","suffix":""},{"id":634244816,"identity":"2ec3c0b2-28e0-4713-9f44-6122c68e23e9","order_by":2,"name":"Xiumei Jiang","email":"","orcid":"","institution":"Qilu Hospital, Shandong University","correspondingAuthor":false,"prefix":"","firstName":"Xiumei","middleName":"","lastName":"Jiang","suffix":""},{"id":634244819,"identity":"60e99b91-850e-4e81-a452-6893cfc591cd","order_by":3,"name":"Yongmei Yang","email":"","orcid":"","institution":"Qilu Hospital, Shandong University","correspondingAuthor":false,"prefix":"","firstName":"Yongmei","middleName":"","lastName":"Yang","suffix":""},{"id":634244821,"identity":"657c2a02-11ba-4c10-a8fa-d08906e87b1c","order_by":4,"name":"Ailin Qu","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAABO0lEQVRIie3RsUrDQBjA8TsOEoQzXRMO2ldICAhipK9yR6GTQifpIJgYuCyBrikovkIn55SAdQjOgVtSXCtEsjgV7wSnpBScHO4/hCPJj4+7A0Cn+4cNEIzqZm7jgblo60a+MUz5wOpb3k+cJI69rAyGTpqfeZki6Ahxyw0nJ3zquxV1yc+fxwioGCfAKNgyo9QP5s8jCwHUfHAwtCoK21lXwIzFFzNcsAWu8/erUngcAcN55MB3KopI1iXIZpHIbDklYaF/zQVUBJ1ywFYVNRDuEsNmIcFuwVb5BJDzvRhLglpJ7g4RjNf3BFO5/dcpIDAUTBJAJKHuAWKbkTzkXB1yaXjpi5hwBLnz8GZ7y3Ibkx4yLsxt/blXV5mi+utWXD4lcdHsboKRtZms2x7SFwzV+N+FTqfT6f7QNyyUazrjxenMAAAAAElFTkSuQmCC","orcid":"","institution":"Qilu Hospital, Shandong University","correspondingAuthor":true,"prefix":"","firstName":"Ailin","middleName":"","lastName":"Qu","suffix":""}],"badges":[],"createdAt":"2026-04-10 03:39:19","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-9374081/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-9374081/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":108839018,"identity":"e9211fe5-5f71-4861-bc8c-4685b68b8cb9","added_by":"auto","created_at":"2026-05-09 00:40:24","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":8308007,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eIdentification of AIGs related to immune score.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e(A) Association between AIGs and ESTIMATE indices. (B) Expression distributions of AIGs between PDAC and normal tissues. (C) Genetic alteration profiles of AIGs in 184 PDAC patients from the cBioPortal database. (D) The PPI network between AIGs (\u003cem\u003eP\u003c/em\u003e\u0026lt;0.05 *;\u003cem\u003e P\u003c/em\u003e\u0026lt;0.01 **; \u003cem\u003eP\u003c/em\u003e\u0026lt;0.001 ***).\u003c/p\u003e","description":"","filename":"Figure1.png","url":"https://assets-eu.researchsquare.com/files/rs-9374081/v1/cc60e4c6ba9de6d4b8a22cfd.png"},{"id":108838971,"identity":"d53ed19b-7a73-41d6-9c03-92376160bd5c","added_by":"auto","created_at":"2026-05-09 00:40:04","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":7252003,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eThe TME characteristics of two distinct clusters.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e(A) Consensus matrix heatmap defining two clusters (k = 2) and their correlation area. (B) Volcano plots of differential genes expression between the two clusters. (C) Differences in the clinicopathologic characteristics and expression levels of 28 AIGs. (D) GSVA of biological pathways between the two clusters. (E) Kaplan–Meier curves of two clusters. (F) Abundance of 22 infiltrating immune cell types in the two PDAC clusters. (G) The cancer immune cycle between the two clusters. (H) The antitumor and protumor immune signatures between the two clusters. (I) The cytokine expression between the two clusters. (\u003cem\u003eP\u003c/em\u003e\u0026lt;0.05 *; \u003cem\u003eP\u003c/em\u003e\u0026lt;0.01 **; \u003cem\u003eP\u003c/em\u003e\u0026lt;0.001 ***).\u003c/p\u003e","description":"","filename":"Figure2.png","url":"https://assets-eu.researchsquare.com/files/rs-9374081/v1/1359c78b06e51990e9a4e19a.png"},{"id":108838973,"identity":"9de906fd-57a7-4d56-b3c0-92bb69c1899d","added_by":"auto","created_at":"2026-05-09 00:40:05","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":3187412,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eConstruction of the AIGScore model based on AIGs in TCGA-PDAC set.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e(A)Forest plot summary of univariate Cox analysis for AIGs in PDAC. (B-C) The AIGScore model identified using LASSO method. (D-F) Distribution of AIGScores and overall survival status and overall survival time. (G) Differences of the AIGScore values in two distinct clusters. (H) The Sankey diagram between AIGScore subgroup and clinical features in PDAC patients. (I) The AIGScore values in T, N, stage and patients’ outcome subgroups. (\u003cem\u003eP\u003c/em\u003e\u0026lt;0.05 *; \u003cem\u003eP\u003c/em\u003e\u0026lt;0.01 **; \u003cem\u003eP\u003c/em\u003e\u0026lt;0.001 ***)\u003c/p\u003e","description":"","filename":"Figure3.png","url":"https://assets-eu.researchsquare.com/files/rs-9374081/v1/9299887d0565adb3ee2e788d.png"},{"id":108838990,"identity":"f6aa1e37-da29-4649-a919-5a654e62ee26","added_by":"auto","created_at":"2026-05-09 00:40:13","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":5261708,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eThe prognostic ability of AIGScore model in training set and validation set.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e(A-B) Kaplan-Meier curves and time-dependent ROC of the AIGScore in TCGA-PDAC set. (C-D) Kaplan-Meier curves and time-dependent ROC of the AIGScore in ICGC-AU set (n=81). (E-F) Kaplan-Meier curves and time-dependent ROC of the AIGScore in GEO set (n=79).\u003c/p\u003e","description":"","filename":"Figure4.png","url":"https://assets-eu.researchsquare.com/files/rs-9374081/v1/7609f7e3f69dfe025a5f81a0.png"},{"id":108838950,"identity":"2ad7135b-7324-4bb5-ae4b-45d449a7905f","added_by":"auto","created_at":"2026-05-09 00:39:58","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":7551418,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eClinical significance of the AIGScore.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e(A) The cancer-immunity cycle in different AIGScore subgroups. (B) The difference of the cancer-immunity cycle between AIGScore subgroups. (C) The correlation between ESTIMATE indices and AIGScore values. (D) The TIDE scores in AIGScore-high group and AIGScore-low group. (E) The IPS scores in AIGScore-high group and AIGScore-low group. (\u003cem\u003eP\u003c/em\u003e\u0026lt;0.05 *; \u003cem\u003eP\u003c/em\u003e\u0026lt;0.01 **).\u003c/p\u003e","description":"","filename":"Figure5.png","url":"https://assets-eu.researchsquare.com/files/rs-9374081/v1/4ebb4355383be0bf1847974e.png"},{"id":108838972,"identity":"7eeb703c-0245-4010-9332-8c88a04d2ede","added_by":"auto","created_at":"2026-05-09 00:40:05","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":9604292,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eCharacteristics of the AIGs at a single-cell resolution.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e(A-C) The angiogenic enrichment scores enriched among clusters. (D-H) UMAP displayed the distribution of the four AIGs expression at a single-cell level in CRA001160 set. (I-J) The distribution of antitumor immune processes and angiogenic pathways in AIGScore-high and low groups. (K-N) The relationship between CNAs of the four AIGs and tumor-infiltrating immune cells. (O) Differentiation of subclusters of myeloid cells, CD4+ T cells and CD8+ T cells. (P) Expression of TYMP in different subclusters. (Q) Expression of MMP2 in different subclusters (\u003cem\u003eP\u003c/em\u003e\u0026lt;0.05 *; \u003cem\u003eP\u003c/em\u003e\u0026lt;0.01 **; \u003cem\u003eP\u003c/em\u003e\u0026lt;0.001 ***).\u003c/p\u003e","description":"","filename":"figure6.png","url":"https://assets-eu.researchsquare.com/files/rs-9374081/v1/55ccf4c3be820d9addf38131.png"},{"id":108838947,"identity":"0a2f37af-0ca2-420e-bc7f-1bc4fe9d2c7d","added_by":"auto","created_at":"2026-05-09 00:39:58","extension":"png","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":2870810,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eThe pattern map of AIGs and antitumor immune pathways.\u003c/strong\u003e\u003c/p\u003e","description":"","filename":"Figure7.png","url":"https://assets-eu.researchsquare.com/files/rs-9374081/v1/3127a8ddbd8558bfe05ab6fc.png"},{"id":108838974,"identity":"b08d5e15-9766-4f5e-ae8e-35b92399fc4d","added_by":"auto","created_at":"2026-05-09 00:40:05","extension":"png","order_by":8,"title":"Figure 8","display":"","copyAsset":false,"role":"figure","size":7606702,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eThe expression of the four AIGs across cancers and their prognostic value.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e(A-D) The expression distribution of the four AIGs between tumor tissues and normal tissues across cancers. (E-H) Forest plots showing the results of univariate Cox regression analysis of the four AIGs across cancers. (\u003cem\u003eP\u003c/em\u003e\u0026lt;0.05 *; \u003cem\u003eP\u003c/em\u003e\u0026lt;0.01 **; \u003cem\u003eP\u003c/em\u003e\u0026lt;0.001 ***). ACC: adrenocortical cancer, BLCA: bladder cancer, BRCA: breast invasive carcinoma, CESC: cervical cancer; CHOL: cholangiocarcinoma; COAD: colon cancer, DLBC: large B-cell lymphoma, ESCA: Esophageal Cancer.\u003c/p\u003e","description":"","filename":"Figure8.png","url":"https://assets-eu.researchsquare.com/files/rs-9374081/v1/09cfaf0e7bf6ca4233cbb20a.png"},{"id":108838946,"identity":"afd8c8e4-c609-4af7-ad54-fa2186f9b1a1","added_by":"auto","created_at":"2026-05-09 00:39:57","extension":"png","order_by":9,"title":"Figure 9","display":"","copyAsset":false,"role":"figure","size":1888487,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eValidation of the expression of the seven AIGs in PDAC tissues using RT-qPCR.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e(A-G) The expression levels of the seven AIGs in 10 PDAC tissues and paired paraneoplastic tissues. (H) Pearson correlation analysis across the seven AIGs and laboratory indicators in the Qilu cohort.\u003c/p\u003e","description":"","filename":"Figure9.png","url":"https://assets-eu.researchsquare.com/files/rs-9374081/v1/fffe94850855f672e3be8c6c.png"},{"id":109176689,"identity":"25216ad2-d28c-4c7a-92f5-3a7768675959","added_by":"auto","created_at":"2026-05-13 09:32:59","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":48757032,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-9374081/v1/19b1b87c-3146-458d-81e6-e767cdcd502b.pdf"},{"id":108838951,"identity":"287b2d28-692a-4c39-9013-3d259978aa0c","added_by":"auto","created_at":"2026-05-09 00:39:59","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"supplement","size":1759095,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryInformation.pdf","url":"https://assets-eu.researchsquare.com/files/rs-9374081/v1/b6d9a4c68279d8a1c2d5ae03.pdf"},{"id":108838952,"identity":"7476a5e2-93ff-4391-8142-1e74538eb8df","added_by":"auto","created_at":"2026-05-09 00:39:59","extension":"xlsx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":465294,"visible":true,"origin":"","legend":"","description":"","filename":"TableS1S6.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-9374081/v1/201f1e2059bab60037d09ec8.xlsx"}],"financialInterests":"No competing interests reported.","formattedTitle":"Angiogenesis-associated immune genes as prognostic markers and predictors of immunotherapy response in pancreatic ductal adenocarcinoma","fulltext":[{"header":"Introduction","content":"\u003cp\u003ePancreatic ductal adenocarcinoma (PDAC) is a highly malignant tumor with an alarming annual increase in incidence rates [\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e]. Most patients are diagnosed at advanced stages, characterized by early loco-regional spread and distant metastases, which preclude surgical resection and contribute to the disease\u0026rsquo;s poor prognosis. The 5-year overall survival rate remains dismal at approximately 10% [\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e, \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e]. Despite advancements in systemic chemotherapies such as 5-fluorouracil and gemcitabine combined with albumin-bound (nab) paclitaxel, as well as traditional radiotherapy for metastatic PDAC, their overall efficacy remains unsatisfactory [\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eImmunotherapy has emerged as a transformative treatment strategy, achieving remarkable outcomes in various cancers, including breast and lung cancers [\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e]. However, its application in PDAC is still limited, largely due to the unique immunosuppressive nature of the tumor microenvironment (TME) in this disease. PDAC\u0026rsquo;s TME is characterized by a scarcity of infiltrating effector T cells and an abundance of immunosuppressive cell [\u003cspan additionalcitationids=\"CR6\" citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e], both of which diminish the effectiveness of immunotherapy.\u003c/p\u003e \u003cp\u003eThe TME comprises tumor cells, interstitial cells, infiltrating immune cells, nerves and blood vessels, with angiogenesis\u0026mdash;a hallmark of the TME\u0026mdash;playing a pivotal role. Angiogenesis, the formation of new blood vessels, is essential for tumor growth and progression as it provides oxygen and nutrients to neoplastic cells [\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e]. PDAC is a vascularized malignancy and its progression has been correlated with microvessel density within the tumor [\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e]. Overexpression of vascular endothelial growth factor (VEGF) has been shown to drive angiogenesis, further promoting PDAC progression [\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e, \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e]. Furthermore, tumor angiogenesis disrupts the interaction between immune cells and the vessel wall, facilitating immune evasion and contributing to the immune suppressive environment [\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e]. These findings highlight the critical interplay between angiogenesis and tumor immunity, which is crucial for understanding the immune landscape in PDAC.\u003c/p\u003e \u003cp\u003eIn light of these observations, a deeper understanding of the relationship between angiogenesis and the TME is essential for identifying distinct tumor immunophenotypes and improving the efficacy of immunotherapy in PDAC. In this study, we conducted a comprehensive analysis of angiogenesis-associated immune genes (AIGs) and explored their impact on PDAC development, survival outcomes, and the TME. Furthermore, we developed a novel AIGScore model to quantify angiogenesis patterns for individual PDAC patients, providing a potential framework to enhance immunotherapeutic strategies and optimize patient outcomes.\u003c/p\u003e"},{"header":"Materials and methods","content":"\u003cp\u003e\u003cstrong\u003eEthical statement\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAll experimental protocols were approved by the Clinical Research Ethics Committee of Qilu Hospital, Shandong University (Approval No. KYLL-202209-044-1). Informed consent was obtained from all participants. All procedures were conducted in accordance with the relevant guidelines and regulations.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eData and sample acquisition\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eProcessed transcriptomic data (HTSeq-Counts and HTSeq-FPKM) from the TCGA-PDAC database, comprising 176 tumor and 4 normal samples, were downloaded from the University of California Santa Cruz (UCSC) Xena browser. RNA-seq data for 167 normal pancreatic tissues were obtained from the Genotype-Tissue Expression (GTEx) database, with batch effect correction performed using the ComBat method. The resulting dataset of 176 tumor and 171 normal samples was designated as the training set. Additionally, 160 pancreatic cancer samples with comprehensive clinical data were obtained from two datasets: ICGC-PACA-AU-seq (ICGC-AU, n=81) and GEO (GSE85916, NCI cohort, n=79), forming the validation set. AIGs were identified from the GeneCards database using the term \"angiogenesis.\" The top 40 genes with the highest relevance were selected for further analysis. Ten pairs of PDAC tissues and their adjacent normal tissues were collected from Qilu Hospital, Shandong University.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConsensus clustering based on AIGs\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eConsensus clustering of the TCGA-PDAC dataset was performed using the \"ConsensusClusterPlus\" package with the k-means algorithm. Classification stability was ensured through 1,000 iterations. The optimal cluster number was determined based on the item-consensus plot and relative changes in the area under the cumulative density function (CDF) curves.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCharacteristics of TME in two subclusters\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe CIBERSORT algorithm was applied to estimate the relative abundance of 22 immune cell types in TME\u0026nbsp;[13]. The ESTIMATE algorithm was used to evaluate the proportions of infiltrating immune and stromal cells\u0026nbsp;[14]. Differences in the cancer immunity cycle between subclusters were analyzed using the Tracking Tumor Immunophenotype (TIP) website\u0026nbsp;[15]. Pathway enrichment differences were assessed using Gene Set Variation Analysis (GSVA), with the “c2.cp.kegg.v7.0.symbols.gmt” and “h.all.v7.0.symbols.gmt” datasets serving as references\u0026nbsp;[16].\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eDevelopment and validation of the AIGScore model\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eIn the TCGA-PDAC training set, 28 AIGs significantly associated with ImmuneScore (\u003cem\u003eP\u003c/em\u003e\u0026lt;0.05) were identified through univariate Cox regression analysis. These genes were further analyzed using the least absolute shrinkage and selection operator (LASSO) with ten-fold cross-validation to select the optimal lambda value. The AIGScore model was constructed as follows: AIGScore = expression of a gene β1×corresponding coefficient [β1] +\u0026nbsp;expression of a gene β2 ×corresponding coefficient [β2] +…… expression of gene βi× corresponding coefficient [βi]. Patients were stratified into high and low AIGScore subgroups based on the median AIGScore. Kaplan-Meier survival analysis and time-dependent receiver operating characteristic (ROC) curve analysis were performed on the TCGA training set to evaluate the model’s prognostic efficacy. The ICGC and GEO datasets served as validation cohorts to confirm its robustness. Differences in TME and immune cell infiltration between AIGScore subgroups were assessed using ESTIMATE and CIBERSORT algorithms, respectively. Cancer immunotherapy response was predicted using the Tumor Immune Dysfunction and Exclusion (TIDE) algorithm\u0026nbsp;[17]\u0026nbsp; and the Immune Phenotype Score (IPS).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eIdentification of hub genes related to angiogenesis\u003c/strong\u003e\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eDEGs between angiogenesis-related subgroups were identified using the \"limma\" package, with thresholds of |log2-fold change|≥2 and P value\u0026lt;0.05. Weighted correlation network analysis (WGCNA) based on the DEGs was performed using \"WGCNA\" package\u0026nbsp;[18]. A soft-thresholding power of 2 was selected to ensure a scale-free network. The adjacency matrix was transformed into a topological overlap matrix (TOM) to describe gene co-expression relationships, with dissimilarity represented by 1-TOM. Modules were identified using the dynamic hybrid tree-cutting algorithm, with a minimum module size of 30. Among the seven modules generated, the blue module exhibited the strongest correlations with OS.time, AIGScore and ESTIMATE indices. Hub genes were identified within this module using Cytoscape and the cytoHubba plugin. Correlations between hub genes and immune infiltration were calculated using the \"corrplot\" package.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eOnline databases information\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe relationship between somatic cell copy number alterations (CNAs) of seven AIGs and tumor-infiltrating immune cells was analyzed using the TIMER database. The average expression distribution of the seven AIGs was obtained from the TISCH database, while trajectory analysis was conducted through the TIGER database. Gene survival analysis was performed using the Kaplan-Meier Plotter website\u0026nbsp;[19]. Detailed information about these online databases is provided in Table S1.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eReal-time quantitative PCR and immunohistochemistry\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTotal RNA was extracted using TRIzol reagent (Invitrogen). RNA concentration and quality were assessed with a NanoDrop spectrophotometer (Thermo Fisher Scientific, Waltham, MA, USA). cDNA synthesis was carried out using HiScript III RT SuperMix for qPCR (Vazyme, Nanjing, China). RT-qPCR was performed using ChamQ Universal SYBR qPCR Master Mix (Vazyme, Nanjing, China). Primer sequences for each gene are listed in Table S2. Immunohistochemical (IHC) staining images of AIGs were retrieved from the Human Protein Atlas (HPA) database for validation.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eStatistical analysis\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAll statistical analyses were conducted using R software (version 4.0.2) and GraphPad Prism 8. Correlations between variables were assessed using Pearson's correlation test. Box plots were generated with the Wilcoxon rank-sum test. Survival analyses were performed using the Kaplan-Meier method, with statistical differences evaluated by the log-rank test through the \"survminer\" and \"survival\" R packages\u0026nbsp;[20]. A two-tailed \u003cem\u003eP\u003c/em\u003e\u0026lt;0.05 was considered statistically significant.\u003c/p\u003e"},{"header":"Results","content":"\u003cp\u003e\u003cstrong\u003eLandscape of angiogenesis-immune regulators in PDAC\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eFirst, the top 40 genes most significantly associated with angiogenic activity were selected from the GeneCards database (Table S3). To interrogate the functional role of angiogenesis in tumor immunity in PDAC patients, we calculated ESTIMAT indices of each patient and assessed the correlation between the 40 angiogenesis-related genes and ESTIMATE indices. As a result, 28 genes with a correlation coefficient absolute value \u0026gt;0.3 were identified as the AIGs (Figure 1A). The immune infiltration patterns of these 28 AIGs were further analyzed using the CIBERSORT algorithm, which revealed strong associations between the AIGs and the abundance of 22 types of immune cells (Figure S1). These findings suggest that the identified AIGs exhibit significant immunogenicity within the tumor microenvironment of PDAC.\u003c/p\u003e\n\u003cp\u003eRNA-seq data from the TCGA-PDAC and GTEx databases were analyzed to compare AIG expression between tumor tissues (n = 176) and normal tissues (n = 171). Twenty-five AIGs were found to be differentially expressed, with most showing elevated expression in tumor specimens (\u003cem\u003eP\u003c/em\u003e\u0026lt;0.05, Figure 1B). To evaluate somatic mutation frequencies, gene mutation data for the 28 AIGs were visualized using the cBioPortal database. Genetic alterations were present in 66% of PDAC samples, with FLT4 exhibiting the highest mutation frequency (10%), followed by ANGPT1 and MMP9 (both at 9%) (Figure 1C). A protein-protein interaction (PPI) network was constructed for the 28 AIGs, identifying VEGFA as a central hub within the network (Figure 1D). This tight connectivity suggests both direct (physical) and indirect (functional) interactions among the AIGs, highlighting their potential cooperative roles in angiogenesis and immunity regulation.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCluster analysis based on AIGs in PDAC\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo systematically classify tumor phenotypic traits in PDAC, consensus clustering analysis was performed on the TCGA-PDAC dataset using the expression profiles of the 28 AIGs. The empirical CDF plots and the consensus matrix identified k = 2 as the optimal number of clusters (Figure S2A-D). As shown in Figure 2A, 176 PDAC patients were divided into two subclusters: Cluster 1 (n = 88) and Cluster 2 (n = 88). Differential expression analysis between the two subclusters revealed 103 upregulated and 57 downregulated genes in Cluster 2 compared to Cluster 1 (Table S4, Figure 2B). Moreover, substantial differences in AIG expression and clinicopathological features were observed between the two subclusters (Figure 2C), indicating distinct tumor microenvironment characteristics. Kaplan-Meier survival analysis demonstrated a significant difference in survival probability between the two subclusters. Patients in Cluster 1 exhibited better overall survival compared to those in Cluster 2 (P = 0.012, Figure 2E), suggesting that the AIG-based clustering effectively stratifies patients by phenotypic and clinical outcomes.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCharacteristics of TME in two subclusters\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eGSVA indicated that pathways associated with cancer progression and metastasis-such as angiogenesis, TGF-β, hypoxia, KRAS and P53 pathways-were significantly enriched in Cluster 2 (Figure 2D). This enrichment suggests that Cluster 2 exhibits an elevated inflammatory and immune-active status. The immune cell landscape was further analyzed using the CIBERSORT algorithm. Cluster 1 displayed higher proportions of CD8+ T cells, CD4+ memory resting T cells, activated NK cells, resting mast cells, resting dendritic cells and monocytes, which are generally associated with antitumor immunity. Conversely, Cluster 2 exhibited greater infiltration of follicular helper T cells, M0 macrophages, T regulatory (Treg) cells and M2 macrophages, the latter two being immunosuppressive components that can promote tumor progression (Figure 2F). Anti-cancer immune responses progress through a series of steps [21] and Step 2 (cancer antigen presentation) was notably active in Cluster 1 (\u003cem\u003eP\u003c/em\u003e\u0026lt;0.05; Figure 2G). Interestingly, cluster 1 exhibited lower levels of both pro-tumor and anti-tumor immune signatures compared to Cluster 2 (Figure 2H), underscoring a dual role of angiogenesis in cancer development. Cytokine analysis revealed that most cytokines were upregulated in Cluster 2 (Figure 2I), highlighting its more inflammatory microenvironment. No significant differences in tumor mutational burden (TMB) were observed between the two subclusters (Figure S2E). However, cluster 2 had higher immune, stromal and ESTIMATE scores but lower tumor purity than cluster 1 (\u003cem\u003eP\u003c/em\u003e\u0026lt;0.05; Figure S2F), further supporting the notion of a more immunologically and stromally active TME in Cluster 2.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eDevelopment and validation of the AIGScore model in TCGA-PDAC set\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eIn the TCGA-PDAC dataset, univariate Cox regression analysis identified nine AIGs significantly associated with OS (\u003cem\u003eP\u003c/em\u003e\u0026lt;0.05; Figure 3A). These genes were subjected to LASSO Cox regression, resulting in an AIGScore model comprising seven genes (Figure 3B-C, Figure S3A). The AIGScore was calculated as follows: AIGScore= (0.289*ExpFGF2) + (-0.731*ExpADGRB3) + (0.0839*ExpMMP2) + (0.019*ExpTYMP) + (0.483*ExpVASH1) + (-0.068*ExpVEGFD) + (0.184*ExpITGAV). Patients were stratified into AIGScore-high and AIGScore-low groups based on the median AIGScore value (Figure 3D-E). A negative correlation was observed between the AIGScore and survival time (R=-0.481, \u003cem\u003eP\u003c/em\u003e\u0026lt;0.001, Figure 3F). Cluster comparison showed that Cluster 2 had significantly higher AIGScores than Cluster 1 (Figure 3G), aligning with its more aggressive pathological features. The alluvial plot (Figure 3H) illustrated the association between AIGScore groups and clinical characteristics, revealing significant differences in T, N, TNM staging and patient outcomes across subgroups (all \u003cem\u003eP\u003c/em\u003e\u0026lt;0.05, Figure 3I). These results collectively suggest that higher AIGScores are associated with more malignant features and poorer prognosis.\u003c/p\u003e\n\u003cp\u003eKaplan-Meier survival analysis confirmed that patients in the AIGScore-high group had significantly lower OS compared to those in the AIGScore-low group (log-rank test, \u003cem\u003eP\u003c/em\u003e\u0026lt;0.001, Figure 4A). Time-dependent ROC analysis demonstrated the robust prognostic performance of the AIGScore model, with AUC values of 0.714, 0.725 and 0.740 at 1, 3 and 5 years, respectively (Figure 4B). The AIGScore model was further validated in independent datasets (ICGC and GEO). Consistent with the TCGA-PDAC dataset, the AIGScore-high group exhibited worse OS than the AIGScore-low group in both the ICGC dataset (\u003cem\u003eP\u003c/em\u003e=0.04, Figure 4C) and the GEO dataset (\u003cem\u003eP\u003c/em\u003e=0.009, Figure 4E). Time-dependent ROC analysis also confirmed the predictive efficacy of the model in these validation datasets (Figure 4D, 4F).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAssessment of immunity activity and checkpoints in distinct subgroups\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe investigated the disparities in anticancer immune responses between the AIGScore-high and AIGScore-low groups. The analysis revealed that step 1 of the immune response was significantly elevated in the AIGScore-high group, whereas steps 3-5 were more pronounced in the AIGScore-low group (Figure 5A, B; Table S5). Notably, no significant differences were observed in other steps (Figure S3B-D). Using the ESTIMATE algorithm, we evaluated the correlation between AIGScore and immune cell enrichment. Stromal score, immune score and ESTIMATE score were positively correlated with AIGScore, while tumor purity showed a negative correlation (Figure 5C). These findings highlight the intricate relationship between AIGScore and the tumor microenvironment. To further assess the immunotherapeutic potential, TIDE was employed to predict patient responses within the TCGA-PADC cohort. The results demonstrated that the AIGScore-low group exhibited significantly lower TIDE levels compared to the AIGScore-high group (\u003cem\u003eP\u003c/em\u003e\u0026lt;0.05; Figure 5D), suggesting a superior immunotherapy response in the AIGScore-low group. Moreover, analysis using the IPS algorithm revealed that the AIGScore-low group benefit more from anti-PD1 therapy (\u003cem\u003eP\u003c/em\u003e\u0026lt;0.01; Figure 5E), underscoring its therapeutic implications.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eIdentification of hub genes related to angiogenesis\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo explore the biological underpinnings of the AIGScore model, we identified 2,556 DEGs between the AIGScore-high and AIGScore-low groups (Figure S3E). These DEGs were utilized to construct gene co-expression networks via weighted correlation network analysis (WGCNA) (Figure S3F-G). By correlating phenotypes defined by OS.time, AIGScore and ESTIMATE indices, we identified seven distinct WGCNA modules (Figure S3H). Among these, the blue module was selected due to its strong correlation with the immune score (r=0.81, \u003cem\u003eP\u003c/em\u003e=8 * 10\u003csup\u003e-40\u003c/sup\u003e) and negative correlation with the risk score (r=-0.41, \u003cem\u003eP\u003c/em\u003e=3 *10\u003csup\u003e-8\u003c/sup\u003e) (Figure S3I-K). Thereafter, we obtained eight hub genes -CD19, PTPRC, CCR7, CD28, IL6, IL13, CD22 and CD40LG-from the blue module. Correlation analysis revealed strong positive associations among these genes (Figure S3L). Additionally, these hub genes exhibited significant positive correlations with the infiltration levels of naive B cells, CD8 T cells and CD4 memory-activated T cells, reinforcing their role in immune regulation (Figure S3M). When comparing AIGScore subgroups, the expression levels of all eight hub genes were significantly higher in the AIGScore-low group (all \u003cem\u003eP\u003c/em\u003e\u0026lt;0.05; Figure S3N), implying these genes might play pivotal roles in orchestrating immune activity and angiogenesis within the tumor microenvironment. These findings collectively illustrate the potential of AIGScore as a biomarker for immunological activity and its intricate link to angiogenic processes.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCharacteristics of the AIGScore model in TME\u003c/strong\u003e \u003cstrong\u003eat single-cell landscape\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eGiven the relevance of the AIGScore model to the TME and prognosis, we analyzed the heterogeneity of angiogenic enrichment scores at the single-cell level. Fibroblast cells, malignant cells and myeloid cells exhibited higher angiogenic enrichment scores among all cell clusters (Figure 6A-C). We further examined the average expression distribution of the seven AIGs in CRA001160 set. Results showed MMP2 had moderate to high expression in fibroblast and endothelial cells (Figure 6D-E) and TYMP was predominantly expressed in macrophages and monocytes (Figure 6F). FGF2, ADGRB3 and VEGFD were enriched in stromal and immune cells but showed lower expression in malignant cells (Figure S4A-C). VASH1 was highly expressed in dendritic cells, macrophages, monocytes and endothelial cells (Figure 6G). ITGAV displayed broad expression across various cell types (Figure 6H). The expression patterns of these seven AIGs were consistent in the GSE111672 dataset, as detailed in Figure S4D-K. From a functional perspective, anticancer immune pathway analysis indicated heightened activity in the trafficking of immune cells to tumors and immune cell infiltration in the AIGScore-low group (Figure 6I). Meanwhile, GSVA analysis showed pathways such as angiogenesis, TGF-β, hypoxia and MYC targets were significantly upregulated in the AIGScore-high group (Figure 6J).\u003c/p\u003e\n\u003cp\u003eThe CNA analysis further revealed that these seven AIGs significantly influenced the infiltration of B cells, CD4+ T cells, CD8+ T cells, macrophages, neutrophils and dendritic cells into the tumor microenvironment (Figure 6K-N, Figure S4L-N). Considering the abundance of TYMP and MMP2 at the single-cell level, we performed trajectory analysis. The results highlighted an imbalance in their expression across different differentiation trajectory states (Figure 6O-Q). Additionally, analysis in the TISIDB database confirmed strong correlations between the seven AIGs and immune-related molecular subtypes (Figure S4O). These findings suggest that the seven AIGs exert pivotal regulatory effects on immune cells and the antitumor immune cycle, emphasizing their integral role in both angiogenesis and immune modulation within the TME.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eThe relationship between AIGs and antitumor immune pathway in pattern map\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo elucidate the molecular mechanisms linking angiogenesis and the tumor immune cycle, we developed a pattern map summarizing our findings (Figure 7). The map highlights the critical roles of TYMP-enriched macrophages and MMP2-enriched fibroblasts in the angiogenic pathways. These cells not only enhance immune cell trafficking to tumors but also facilitate immune cell infiltration, thereby impacting the antitumor immune cycle. Using the CTD database, we predicted potential therapeutic agents targeting TYMP and MMP2. The results are summarized in Table S6, offering insights into therapeutic strategies aimed at modulating these key angiogenesis-related genes. These results underscore the dual roles of angiogenesis in promoting immune cell activity and shaping the TME, revealing potential targets for therapeutic intervention to enhance antitumor immunity.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ePan-cancer analysis of the AIGs expression and their prognostic value\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eGiven that angiogenesis and immunity are not unique to PDAC, we hypothesized that the seven AIGs may also play a role in the tumor status and prognosis of other cancers. To test this hypothesis, we conducted a pan-cancer analysis using the TCGA dataset, examining the expression profiles of the seven AIGs across a broad spectrum of cancers. The results revealed that the seven AIGs were differentially expressed between cancerous and normal tissues in most cancer types (Figure 8A-D; Figure S5A-G). Furthermore, we found that the expression levels of these AIGs were significantly correlated with clinical outcomes in each cancer type (Figure 8E-H; Figure S5H-J). These findings suggest that AIGs may serve as important biomarkers for prognosis across various cancers, reinforcing their potential relevance beyond PDAC.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eValidation of AIGs expression levels in PDAC and paired adjacent tissues\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo validate the expression of the seven AIGs in PDAC, we performed RT-qPCR and immunohistochemistry on PDAC tissues and paired adjacent normal tissues. The RT-qPCR analysis revealed that the expression levels of FGF2, MMP2, TYMP, VASH1 and ITGAV were significantly higher in PDAC tissues compared to adjacent normal tissues (\u003cem\u003eP\u003c/em\u003e\u0026lt;0.05). The differences in ADGRB3 (\u003cem\u003eP\u003c/em\u003e=0.07) and VEGFD (\u003cem\u003eP\u003c/em\u003e=0.05) expression, although not statistically significant, showed a trend towards higher levels in PDAC tissues (Figure 9A-G). These findings were further corroborated by immunohistochemistry, where we observed consistent expression patterns for FGF2, MMP2, TYMP, VASH1, ITGAV and VEGFD (ADGRB3 data was not available) (Figure S6). Additionally, we conducted correlation analysis between the expression levels of the seven AIGs and various laboratory indicators. We found significant correlations between these genes and several clinical indicators, including lymphocyte and monocyte counts, as well as tumor markers such as CEA and CA125 (Figure 9H). These associations further suggest that the expression of AIGs is linked to key immunological and pathological features of PDAC, providing valuable insights into their potential role as biomarkers for both diagnosis and prognosis in PDAC.\u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eAngiogenesis, a key process in the early stages of tumor progression, is widely recognized as a hallmark of cancer biology\u0026nbsp;[8]. A growing body of evidence underscores the intricate relationship between angiogenesis and immunity, suggesting that targeting angiogenesis could enhance the efficacy of immunotherapies\u0026nbsp;[22, 23]. However, the precise mechanisms linking angiogenesis-related genes and tumor-associated immune infiltrates remain poorly understood. In this study, we utilized single-cell RNA sequencing and multi-omics data to systematically investigate the role of AIGs in tumor immunity, focusing on their prognostic value. We developed an AIGScore model to quantify angiogenesis patterns in PDAC, with the goal of assisting clinicians in devising more effective immunotherapeutic strategies for PDAC patients.\u003c/p\u003e\n\u003cp\u003eOur analysis of gene expression and mutation data for 28 AIGs from the TCGA-PDAC and GTEx databases revealed elevated levels of most AIGs in PDAC tissues, with 66% of PDAC samples exhibiting genetic alterations. Through consensus clustering analysis, we classified PDAC patients into two distinct subclusters: Cluster 1 and Cluster 2. Notably, significant differences in clinical features, immune activities and functional pathways were observed between these subclusters. Angiogenesis has long been implicated in the malignant behavior of various cancers, including PDAC\u0026nbsp;[24-27]. Tumor-associated macrophages, for instance, promote tumor cell migration via VEGF signaling\u0026nbsp;[28], while numerous proangiogenic molecules are linked to immunosuppressive effects within the TME\u0026nbsp;[29-31].\u0026nbsp;Disruption of immune cell function within the PDAC TME is a key driver of tumor progression\u0026nbsp;[32, 33].\u0026nbsp;Consistent with previous studies, our GSVA results showed that cancer-related and metastasis-associated pathways were enriched in Cluster 2, which also exhibited higher expression of inflammatory and immunosuppressive signatures. These findings highlight the pivotal role of AIGs in shaping the immunogenic landscape of the PDAC TME.\u003c/p\u003e\n\u003cp\u003eTo better understand the prognostic significance of angiogenesis in PDAC, we developed the AIGScore risk model using LASSO regression. The AIGScore was strongly correlated with clinicopathological features of PDAC and stratified analysis demonstrated its independent predictive value for patient survival outcomes. Time-dependent ROC analysis confirmed that the AIGScore model reliably predicted prognosis in both training and validation sets. While previous studies have explored prognostic models in PDAC using other gene panels—such as Luo et al.’s necroptosis-related genes\u0026nbsp;[34]and Chen et al.’s hypoxia-immune signature\u0026nbsp;[35]—our study is the first to construct a prognostic model using seven angiogenesis- and immunity-related genes. We found that PDAC patients with higher AIGScores had elevated immune and stromal scores, along with poorer survival outcomes. High immune and stromal scores have been previously associated with adverse clinicopathological features and poor prognosis in PDAC\u0026nbsp;[36]. These results suggest that AIGScore could serve as a robust marker of PDAC immunogenicity, potentially guiding therapeutic decision-making.\u003c/p\u003e\n\u003cp\u003eOur study further explored the relationship between the AIGScore and response to immunotherapy. Antiangiogenic therapies have been shown to inhibit T cell infiltration and antigen presentation by dendritic cells, potentially leading to immune suppression within the TME\u0026nbsp;[23, 37-39]. Interestingly, our analysis using TIDE and IPS signatures indicated that PDAC patients with low AIGScores had lower TIDE scores and showed better responses to anti-PD1 therapies compared to those with high AIGScores. This finding supports the notion that the AIGScore model can identify PDAC patients who are more likely to benefit from immune checkpoint blockade therapy.\u003c/p\u003e\n\u003cp\u003eOur constructed AIGScore model contained seven AIGs: FGF2, ADGRB3, MMP2, TYMP, VASH1, VEGFD and ITGAV. Several studies have implicated FGF2, MMP2 and TYMP in the proliferation, invasion and migration of PDAC cells\u0026nbsp;[40-42]. In our study, single-cell analysis confirmed that TYMP and MMP2, two critical angiogenic regulators, were specifically expressed in macrophages and fibroblasts within the PDAC TME. Furthermore, CNA analysis revealed that these seven AIGs significantly influenced the infiltration of key immune cells—including B cells, CD4+ T cells, CD8+ T cells, macrophages, neutrophils and dendritic cells—into the tumor. T cells, which mediate tumor cell killing, are essential for maintaining the cancer-immune cycle\u0026nbsp;[43]. Previous studies have demonstrated that macrophages and fibroblasts not only promote tumor angiogenesis but also contribute to vascular stabilization\u0026nbsp;[44, 45]. Based on these findings, we hypothesize that TYMP-high macrophages and MMP2-high fibroblasts play an important role in facilitating T cell infiltration into tumors via angiogenesis, thereby enhancing the antitumor immune response.\u003c/p\u003e\n\u003cp\u003eDespite the strengths of this study, several limitations must be acknowledged. First, while our findings suggest that the AIGScore model may have significant clinical utility in PDAC, further validation in independent patient cohorts and clinical trials is necessary. Second, while we have established an association between angiogenesis-modified immune cell infiltration and survival, the underlying molecular mechanisms driving these interactions remain unclear and warrant further investigation. Additionally, the relatively small sample size of PDAC cases in our study necessitates validation in larger, more diverse cohorts. Therefore, a multi-center, retrospective study is warranted to confirm these findings.\u003c/p\u003e"},{"header":"Conclusions","content":"\u003cp\u003eIn conclusion, we have developed a prognostic AIGScore model that effectively stratifies the risk of PDAC patients based on angiogenesis-related immune molecules. Our results suggest that TYMP and MMP2 are critical regulators of T cell trafficking and infiltration in the TME and that assessing individual tumor angiogenesis patterns can provide valuable insights into the TME characteristics. This model holds promise for guiding the development of personalized immunotherapeutic strategies for PDAC, enhancing treatment outcomes through targeted modulation of angiogenesis and immune pathways.\u003c/p\u003e"},{"header":"Abbreviations","content":"\u003cp\u003eAIGs, angiogenesis-associated immune genes; DEG, differentially expressed gene; KM, Kaplan–Meier curves; LASSO, least absolute shrinkage and selection operator; TME, tumor microenvironment; PDAC, pancreatic ductal adenocarcinoma; TIDE, Tumor Immune Dysfunction and Exclusion; WGCNA, weighted correlation network analysis.\u003c/p\u003e\n\u003cp\u003e\u003cbr\u003e\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cem\u003e\u003cstrong\u003eFunding\u003c/strong\u003e\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eNo funding was used for the study.\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cem\u003e\u003cstrong\u003eAcknowledgements\u003c/strong\u003e\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eNot applicable for that section.\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthor contributions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAQ conceived and designed the experiments. QC and QW performed all the experiments. XJ analyzed the data. AQ and YY wrote the manuscript. All authors read and approved the final manuscript.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eData availability\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAll data and R codes in this study are reasonably available from the corresponding author. The data for this study were obtained from the publicly available databases.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCompeting interests\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors have no conflicts of interest to declare\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eEthics declarations\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAll procedures involving human participants were approved by the ethical standards of the Clinical Research Ethics Committee of Qilu Hospital, Shandong University and performed in accordance with the Declaration of Helsinki.\u0026nbsp;\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003ePark, W., Chawla, A. \u0026amp; O'Reilly, E. M. A review of pancreatic cancer-reply. \u003cem\u003eJAMA\u003c/em\u003e \u003cb\u003e326\u003c/b\u003e, 2436\u0026ndash;2437 (2021).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSiegel, R. L., Giaquinto, A. N. \u0026amp; Jemal, A. Cancer statistics, 2024. \u003cem\u003eCA Cancer J. Clin.\u003c/em\u003e \u003cb\u003e74\u003c/b\u003e, 12\u0026ndash;49 (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCancer stat facts: pancreatic cancer. (2021). Available at: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://seer.cancer.gov/statfacts/html/pancreas.html\u003c/span\u003e\u003cspan address=\"https://seer.cancer.gov/statfacts/html/pancreas.html\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003evan den Bulk, J., Verdegaal, E. M. \u0026amp; de Miranda, N. F. Cancer immunotherapy: broadening the scope of targetable tumours. \u003cem\u003eOpen. Biol.\u003c/em\u003e, 8. (2018).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRoyal, R. E. et al. Phase 2 trial of single agent ipilimumab (anti-ctla-4) for locally advanced or metastatic pancreatic adenocarcinoma. \u003cem\u003eJ. Immunother\u003c/em\u003e. \u003cb\u003e33\u003c/b\u003e, 828\u0026ndash;833 (2010).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMenezes, S., Okail, M. H., Jalil, S., Kocher, H. M. \u0026amp; Cameron, A. Cancer-associated fibroblasts in pancreatic cancer: new subtypes, new markers, new targets. \u003cem\u003eJ. Pathol.\u003c/em\u003e \u003cb\u003e257\u003c/b\u003e, 526\u0026ndash;544 (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePawlik, T. M. Pancreatic cancer. \u003cem\u003eSurg. Oncol. Clin. N Am.\u003c/em\u003e, 30:xiii\u0026ndash;xv. (2021).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHanahan, D. Hallmarks of cancer: new dimensions. \u003cem\u003eCancer Discov\u003c/em\u003e. \u003cb\u003e12\u003c/b\u003e, 31\u0026ndash;46 (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDineen, S. P. et al. The adnectin ct-322 is a novel vegf receptor 2 inhibitor that decreases tumor burden in an orthotopic mouse model of pancreatic cancer. \u003cem\u003eBMC Cancer\u003c/em\u003e. \u003cb\u003e8\u003c/b\u003e, 352 (2008).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSun, Y. et al. Toll-like receptor 4 promotes angiogenesis in pancreatic cancer via pi3k/akt signaling. \u003cem\u003eExp. Cell. Res.\u003c/em\u003e \u003cb\u003e347\u003c/b\u003e, 274\u0026ndash;282 (2016).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVeikkola, T., Karkkainen, M., Claesson-Welsh, L. \u0026amp; Alitalo, K. Regulation of angiogenesis via vascular endothelial growth factor receptors. \u003cem\u003eCancer Res.\u003c/em\u003e \u003cb\u003e60\u003c/b\u003e, 203\u0026ndash;212 (2000).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhou, X., Tang, J., Cao, H., Fan, H. \u0026amp; Li, B. Tissue resident regulatory t cells: novel therapeutic targets for human disease. \u003cem\u003eCell. Mol. Immunol.\u003c/em\u003e \u003cb\u003e12\u003c/b\u003e, 543\u0026ndash;552 (2015).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNewman, A. M. et al. Robust enumeration of cell subsets from tissue expression profiles. \u003cem\u003eNat. Methods\u003c/em\u003e. \u003cb\u003e12\u003c/b\u003e, 453\u0026ndash;457 (2015).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYoshihara, K. et al. Inferring tumour purity and stromal and immune cell admixture from expression data. \u003cem\u003eNat. Commun.\u003c/em\u003e \u003cb\u003e4\u003c/b\u003e, 2612 (2013).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eXu, L. et al. Tip: a web server for resolving tumor immunophenotype profiling. \u003cem\u003eCancer Res.\u003c/em\u003e \u003cb\u003e78\u003c/b\u003e, 6575\u0026ndash;6580 (2018).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHanzelmann, S., Castelo, R. \u0026amp; Guinney, J. Gsva: gene set variation analysis for microarray and rna-seq data. \u003cem\u003eBMC Bioinform.\u003c/em\u003e \u003cb\u003e14\u003c/b\u003e, 7 (2013).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJiang, P. et al. Signatures of t cell dysfunction and exclusion predict cancer immunotherapy response. \u003cem\u003eNat. Med.\u003c/em\u003e \u003cb\u003e24\u003c/b\u003e, 1550\u0026ndash;1558 (2018).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLangfelder, P. \u0026amp; Horvath, S. Wgcna: an r package for weighted correlation network analysis. \u003cem\u003eBMC Bioinform.\u003c/em\u003e \u003cb\u003e9\u003c/b\u003e, 559 (2008).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGyorffy, B. Survival analysis across the entire transcriptome identifies biomarkers with the highest prognostic power in breast cancer. \u003cem\u003eComput. Struct. Biotechnol. J.\u003c/em\u003e \u003cb\u003e19\u003c/b\u003e, 4101\u0026ndash;4109 (2021).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRich, J. T. et al. A practical guide to understanding kaplan-meier curves. \u003cem\u003eOtolaryngol. Head Neck Surg.\u003c/em\u003e \u003cb\u003e143\u003c/b\u003e, 331\u0026ndash;336 (2010).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChen, D. S. \u0026amp; Mellman, I. Oncology meets immunology: the cancer-immunity cycle. \u003cem\u003eImmunity\u003c/em\u003e \u003cb\u003e39\u003c/b\u003e, 1\u0026ndash;10 (2013).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRahma, O. E. \u0026amp; Hodi, F. S. The intersection between tumor angiogenesis and immune suppression. \u003cem\u003eClin. Cancer Res.\u003c/em\u003e \u003cb\u003e25\u003c/b\u003e, 5449\u0026ndash;5457 (2019).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAnker, J. F. et al. Multi-faceted immunomodulatory and tissue-tropic clinical bacterial isolate potentiates prostate cancer immunotherapy. \u003cem\u003eNat. Commun.\u003c/em\u003e \u003cb\u003e9\u003c/b\u003e, 1591 (2018).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhang, L. et al. Endothelial dgkg promotes tumor angiogenesis and immune evasion in hepatocellular carcinoma. \u003cem\u003eJ. Hepatol.\u003c/em\u003e \u003cb\u003e80\u003c/b\u003e, 82\u0026ndash;98 (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJiang, X. et al. The role of microenvironment in tumor angiogenesis. \u003cem\u003eJ. Exp. Clin. Cancer Res.\u003c/em\u003e \u003cb\u003e39\u003c/b\u003e, 204 (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAnnese, T., Tamma, R., De Giorgis, M. \u0026amp; Ribatti, D. Micrornas biogenesis, functions and role in tumor angiogenesis. \u003cem\u003eFront. Oncol.\u003c/em\u003e \u003cb\u003e10\u003c/b\u003e, 581007 (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSajib, S., Zahra, F. T., Lionakis, M. S., German, N. A. \u0026amp; Mikelis, C. M. Mechanisms of angiogenesis in microbe-regulated inflammatory and neoplastic conditions. \u003cem\u003eAngiogenesis\u003c/em\u003e \u003cb\u003e21\u003c/b\u003e, 1\u0026ndash;14 (2018).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePenny, H. L. et al. Warburg metabolism in tumor-conditioned macrophages promotes metastasis in human pancreatic ductal adenocarcinoma. \u003cem\u003eOncoimmunology\u003c/em\u003e \u003cb\u003e5\u003c/b\u003e, e1191731 (2016).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFukumura, D., Kloepper, J., Amoozgar, Z., Duda, D. G. \u0026amp; Jain, R. K. Enhancing cancer immunotherapy using antiangiogenics: opportunities and challenges. \u003cem\u003eNat. Rev. Clin. Oncol.\u003c/em\u003e \u003cb\u003e15\u003c/b\u003e, 325\u0026ndash;340 (2018).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTao, L., Huang, G., Shi, S. \u0026amp; Chen, L. Bevacizumab improves the antitumor efficacy of adoptive cytokine-induced killer cells therapy in non-small cell lung cancer models. \u003cem\u003eMed. Oncol.\u003c/em\u003e \u003cb\u003e31\u003c/b\u003e, 777 (2014).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eShrimali, R. K. et al. Antiangiogenic agents can increase lymphocyte infiltration into tumor and enhance the effectiveness of adoptive immunotherapy of cancer. \u003cem\u003eCancer Res.\u003c/em\u003e \u003cb\u003e70\u003c/b\u003e, 6171\u0026ndash;6180 (2010).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLiu, L. et al. Low intratumoral regulatory t cells and high peritumoral cd8(+) t cells relate to long-term survival in patients with pancreatic ductal adenocarcinoma after pancreatectomy. \u003cem\u003eCancer Immunol. Immunother\u003c/em\u003e. \u003cb\u003e65\u003c/b\u003e, 73\u0026ndash;82 (2016).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDuewell, P. et al. Rig-i-like helicases induce immunogenic cell death of pancreatic cancer cells and sensitize tumors toward killing by cd8(+) t cells. \u003cem\u003eCell. Death Differ.\u003c/em\u003e \u003cb\u003e21\u003c/b\u003e, 1825\u0026ndash;1837 (2014).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLuo, Z. et al. A panel of necroptosis-related genes predicts the prognosis of pancreatic adenocarcinoma. \u003cem\u003eTransl Oncol.\u003c/em\u003e \u003cb\u003e22\u003c/b\u003e, 101462 (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChen, D. et al. Development and verification of the hypoxia- and immune-associated prognostic signature for pancreatic ductal adenocarcinoma. \u003cem\u003eFront. Immunol.\u003c/em\u003e \u003cb\u003e12\u003c/b\u003e, 728062 (2021).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLi, T., Liu, Q., Zhang, R., Liao, Q. \u0026amp; Zhao, Y. Identification of prognosis-related genes and construction of multi-regulatory networks in pancreatic cancer microenvironment by bioinformatics analysis. \u003cem\u003eCancer Cell. Int.\u003c/em\u003e \u003cb\u003e20\u003c/b\u003e, 341 (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eOhm, J. E. \u0026amp; Carbone, D. P. Vegf as a mediator of tumor-associated immunodeficiency. \u003cem\u003eImmunol. Res.\u003c/em\u003e \u003cb\u003e23\u003c/b\u003e, 263\u0026ndash;272 (2001).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVoron, T. et al. Vegf-a modulates expression of inhibitory checkpoints on cd8\u0026thinsp;+\u0026thinsp;t cells in tumors. \u003cem\u003eJ. Exp. Med.\u003c/em\u003e \u003cb\u003e212\u003c/b\u003e, 139\u0026ndash;148 (2015).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAlbini, A., Bruno, A., Noonan, D. M. \u0026amp; Mortara, L. Contribution to tumor angiogenesis from innate immune cells within the tumor microenvironment: implications for immunotherapy. \u003cem\u003eFront. Immunol.\u003c/em\u003e \u003cb\u003e9\u003c/b\u003e, 527 (2018).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTao, L. et al. Integrated proteomics and phosphoproteomics reveal perturbed regulative pathways in pancreatic ductal adenocarcinoma. \u003cem\u003eMol. Omics\u003c/em\u003e. \u003cb\u003e17\u003c/b\u003e, 230\u0026ndash;240 (2021).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHan, Y., Ma, L., Zhao, L., Feng, W. \u0026amp; Zheng, X. Rosmarinic inhibits cell proliferation, invasion and migration via up-regulating mir-506 and suppressing mmp2/16 expression in pancreatic cancer. \u003cem\u003eBiomed. Pharmacother\u003c/em\u003e. \u003cb\u003e115\u003c/b\u003e, 108878 (2019).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eColeman, S. J. et al. Nuclear translocation of fgfr1 and fgf2 in pancreatic stellate cells facilitates pancreatic cancer cell invasion. \u003cem\u003eEMBO Mol. Med.\u003c/em\u003e \u003cb\u003e6\u003c/b\u003e, 467\u0026ndash;481 (2014).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSong, K. H. et al. Hsp90a inhibition promotes anti-tumor immunity by reversing multi-modal resistance and stem-like property of immune-refractory tumors. \u003cem\u003eNat. Commun.\u003c/em\u003e \u003cb\u003e11\u003c/b\u003e, 562 (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAnnese, T. et al. Inflammatory infiltrate and angiogenesis in mantle cell lymphoma. \u003cem\u003eTransl Oncol.\u003c/em\u003e \u003cb\u003e13\u003c/b\u003e, 100744 (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMao, X. et al. Crosstalk between cancer-associated fibroblasts and immune cells in the tumor microenvironment: new findings and future perspectives. \u003cem\u003eMol. Cancer\u003c/em\u003e. \u003cb\u003e20\u003c/b\u003e, 131 (2021).\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"PDAC, angiogenesis, AIGScore, tumor immune, prognosis","lastPublishedDoi":"10.21203/rs.3.rs-9374081/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-9374081/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eImmunotherapy has emerged as a pivotal approach in cancer treatment, yet its efficacy is influenced by the interactions within the tumor microenvironment (TME). Angiogenesis, the formation of new blood vessels, is a hallmark feature of the TME, particularly in aggressive malignancies such as pancreatic ductal adenocarcinoma (PDAC). This study aims to elucidate angiogenesis patterns in PDAC and investigate their associations with clinical characteristics, TME features and the response to immunotherapy. By analyzing 40 angiogenesis-related genes, 28 angiogenesis-associated immune genes (AIGs) were identified, enabling classification of PDAC patients into two subclusters with distinct clinical and TME profiles. Afterwards, we constructed an AIGScore risk model using least absolute shrinkage and selection operator (LASSO) regression in PDAC, and its reliable predictive ability was confirmed in both training and validation sets. Functional analyses revealed that a high AIGScore was associated with worse prognosis, reduced tumor immune cycle activity, elevated immune and stromal scores and diminished response to anti-PD-1 immunotherapy. Pan-cancer analyses further demonstrated the relevance of the seven AIGs to tumor progression and patient outcomes in other malignancies. In summary, this study introduces a novel AIGScore model with significant prognostic utility for PDAC. The findings provide insights into TME characteristics and offer a potential framework for optimizing immunotherapeutic strategies in patients with PDAC.\u003c/p\u003e","manuscriptTitle":"Angiogenesis-associated immune genes as prognostic markers and predictors of immunotherapy response in pancreatic ductal adenocarcinoma","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-05-09 00:38:17","doi":"10.21203/rs.3.rs-9374081/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"c5d6cc54-5ef5-4696-8051-b564946a0e52","owner":[],"postedDate":"May 9th, 2026","published":true,"recentEditorialEvents":[{"type":"decision","content":"Withdrawn","date":"2026-05-13T09:08:24+00:00","index":"","fulltext":""},{"type":"decision","content":"Revision requested","date":"2026-05-08T08:13:12+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-05-07T10:29:55+00:00","index":169,"fulltext":""},{"type":"reviewerAgreed","content":"262656547652463310311017280579084830771","date":"2026-05-01T06:19:20+00:00","index":168,"fulltext":""}],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[{"id":67497802,"name":"Health sciences/Biomarkers"},{"id":67497803,"name":"Biological sciences/Cancer"},{"id":67497804,"name":"Biological sciences/Computational biology and bioinformatics"},{"id":67497806,"name":"Biological sciences/Immunology"},{"id":67497808,"name":"Health sciences/Oncology"}],"tags":[],"updatedAt":"2026-05-13T09:30:15+00:00","versionOfRecord":[],"versionCreatedAt":"2026-05-09 00:38:17","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-9374081","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-9374081","identity":"rs-9374081","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}