Abstract
BACKGROUND: Uterine leiomyosarcoma (ULMS) is a highly malignant tumor with a poor prognosis. This study aims to explore the potential and significance of novel immune and inflammation-related diagnostic biomarkers in differentiating ULMS from uterine leiomyomas (ULM). METHODS: We analyzed 25 samples of ULMS and 25 samples of ULM from the GEO database (GSE64763). Differentially expressed genes (DEGs) were identified using R software. Different Inflammation- and immunity-related genes (DIIRGs) were derived by intersecting with immune-related and inflammation-related gene sets. Functional enrichment analysis was conducted on DIIRGs utilizing the Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) databases. The Protein-protein interaction (PPI) networks were employed to investigate the interrelationships among various DIIRGs. Two machine learning algorithms were employed for the selection of diagnostic biomarkers. The diagnostic ability was evaluated through receiver operating characteristic (ROC) curves, principal component analysis (PCA), and a nomogram. To further validate our findings, we assessed the diagnostic value of candidate biomarkers in the validation group, including three datasets (GSE9511, GSE68295, and GSE36610), and performed immunohistochemistry (IHC) in clinical tissue samples. Additionally, this study utilized the Cibersort algorithm to determine the composition patterns of 22 immune cell types within ULMS and analyzed correlations between diagnostic markers and immune cells. RESULTS: A total of 1,363 DEGs and 12 DIIRGs were identified between ULMS and ULM. GO analysis revealed that DIIRGs were predominantly enriched in the positive regulation of the release of sequestered calcium ions into the cytosol, cytokine activity, and G protein-coupled receptor binding. KEGG analysis indicated enrichment in several signaling pathways, including cytokine-cytokine receptor interaction, chemokine signaling pathway, neuroactive ligand-receptor interaction, and IL-17 signaling pathway. CALCRL was identified as a potential diagnostic biomarker for ULMS based on machine learning algorithms, demonstrating an area under the curve (AUC) of 0.898. Its low expression correlates with ULMS progression, which was corroborated in the validation cohort (AUC = 0.792) and IHC. Immune infiltration analysis revealed that levels of Macrophages M0 and activated mast cells were elevated in ULMS compared to ULM, whereas levels of activated NK cells, resting mast cells, and Neutrophils were significantly reduced. Furthermore, CALCRL expression exhibited a positive correlation with CD4 memory resting T cells and resting mast cells but a negative correlation with CD8 T cells (P < 0.05). CONCLUSION: Inflammation and immunity play a pivotal role in the pathogenic mechanism of ULMS. Our study findings suggest that CALCRL can serve as an immune-inflammatory biomarker for ULMS, providing a new perspective for exploring the development and diagnosis of ULMS.