An Integrated Multi-Omics Analysis Identifying Immune Subtypes of Pancreatic Cancer.

Limited studies have explored novel pancreatic cancer (PC) subtypes or prognostic biomarkers based on the altered activity of relevant signaling pathway gene sets. Here, we employed non-negative matrix factorization (NMF) to identify three immune subtypes of PC based on C7 immunologic signature gene set activity in PC and normal samples. Cluster 1, the immune-inflamed subtype, showed a higher response rate to immune checkpoint blockade (ICB) and had the lowest tumor immune dysfunction and exclusion (TIDE) scores. Cluster 2, the immune-excluded subtype, exhibited strong associations with stromal activation, characterized by elevated expression levels of transforming growth factor (TGF)-ß, cell adhesion, extracellular matrix remodeling, and epithelial-to-mesenchymal transition (EMT) related genes. Cluster 3, the immune-desert subtype, displayed limited immune activity. For prognostic prediction, we developed an immune-related prognostic risk model (IRPM) based on four immune-related prognostic genes in pancreatic cancer, RHOF, CEP250, TSC1, and KIF20B. The IRPM demonstrated excellent prognostic efficacy and successful validation in an external cohort. Notably, the key gene in the prognostic model, RHOF, exerted significant influence on the proliferation, migration, and invasion of pancreatic cancer cells through in vitro experiments. Furthermore, we conducted a comprehensive analysis of somatic mutational landscapes and immune landscapes in PC patients with different IRPM risk scores. Our findings accurately stratified patients based on their immune microenvironment and predicted immunotherapy responses, offering valuable insights for clinicians in developing more targeted clinical strategies.

Identification of m6A-associated LncRNAs as predict factors for the immune infiltration and prognosis of thyroid cancer.

OBJECTIVE: This study aims to evaluate the prognostic value of m6A-associated long noncoding RNAs (lncRNAs) and their interaction with tumour microenvironment in thyroid cancer (THCA). METHODS: The clinical and gene expression data of tumours from 502 patients with THCA and 58 adjacent normal tissues were retrieved from The Cancer Genome Atlas (TCGA)-THCA dataset. The Pearson test was utilized to identify potential m6A-associated lncRNAs (p < 0.001 and Pearson correlation coefficient > 0.4). Quantitative real-time polymerase chain reaction was performed to verify the expression levels of lncRNAs in tissues. MTT, EdU, colony formation and wound-healing assays were performed to determine the functions of m6A-associated lncRNAs in THCA cell proliferation and metastasis. RESULTS: M6A-associated lncRNAs were identified in three cluster groups. A significant survival difference was found among them, with cluster 1 patients showing worse survival. Moreover, lower immune and estimate scores were correlated to poorer prognosis, and CD8+ T cell and memory CD4+ T cell levels were increased in cluster 1. Cluster 2, with better overall survival, had high expression of PD-L1 and CTLA-4. Eleven of the m6A-associated lncRNAs were screened to establish the risk model, including AC007365.1, AC008555.1, AC040160.1, AC064807.1, AC126773.4, AL023583.1, AL512306.2, EIF2AK3-DT, LINC00667, LYPLAL1-DT and MIR181A2HG. Based on the median risk score, THCA patients were stratified into low-risk and high-risk groups. Overall survival analysis showed a dramatic difference between the two groups. qRCR was performed to verify the expression levels of lncRNA (LYPLAL1-DT, EIF2AK3-DT and MIR181A2HG) in THCA and adjacent normal tissues. Furthermore, functional experiments showed that knockdown of MIR181A2HG obviously inhibited the proliferation and migration of papillary thyroid cancer (PTC) cells in vitro, whereas LYPLAL1-DT overexpression promoted PTC cell proliferation and migration. CONCLUSIONS: Eleven of the m6A-associated lncRNAs were identified as a risk model to predict clinical outcomes and provide a novel and efficient immunotherapeutic strategy for THCA patients.Key messagesm6A-associated lncRNAs can be used to predict the clinical outcomes of thyroid cancer patients.An m6A-associated lncRNAs risk model, which can accurately evaluate the immune status and risk stratification in individual thyroid cancer patients, was established.Knockdown/overexpression of representative lncRNAs in the risk model significantly affected the proliferation and migration of papillary thyroid cancer cells.