Utilizing bioinformatics for integrated analysis of multiple genes in the diagnosis and pathogenesis of metastatic pheochromocytoma and paraganglioma

The aim of the study was to investigate effective diagnostic molecular markers and the specific mechanisms of metastatic pheochromocytomas and paragangliomas (PPGLs). Data were collected from GEO datasets GSE67066 and GSE60458. The R software and various packages were utilized for the analysis of differentially expressed genes, Gene Ontology analysis, Kyoto Encyclopedia of Genes and Genomes analysis, receiver operating characteristic curve assessment, logistic model construction, and correlation analysis. The NetworkAnalyst tool was used to analyze gene-miRNA interactions and signaling networks. In addition, the TIMER database was used to estimate the immune scores. A total of 203 and 499 differentially expressed genes were identified in GSE67066 and GSE60458, respectively. These genes are implicated in cytokine and cytokine receptor interactions, extracellular matrix–receptor interactions, and platelet activation signaling pathways. Notably, MAMLD1, UST, MATN2, LPL, TWIST1, SFRP4, FRMD6, RBM24, PRIMA1, LYPD1, KCND2, CAMK2N1, SPOCK3, and ALPK3 were identified as the key genes. Among them, MATN2 and TWIST1 were found to be coexpressed with epithelial-mesenchymal transition–linked markers, whereas KCND2 and LPL exhibited associations with immune checkpoint expression and immune cell infiltration. Eight miRNAs were identified as potential regulators of key gene expression, and it was noted that TWIST1 might be regulated by SUZ12. Notably, the area under the curve of the 4-gene model for distinguishing between malignant and benign groups was calculated to be 0.918. The combined gene and mRNA expression model enhances the diagnostic accuracy of assessing PPGL metastatic potential. These findings suggest that multiple genes may play a role in the metastasis of PPGLs through the epithelial-mesenchymal transition and may influence the immune microenvironment.