Construction and Validation of a New Model for the Prediction of Rupture in Patients with Intracranial Aneurysms.

BACKGROUND

Despite progress in the detection of biological molecules that contribute to intracranial aneurysm (IA) development, many pathophysiological mechanisms remain unclear, particularly with regard to predicting IA rupture. In this study, we aimed to identify hub genes and construct a new model to predict IA rupture.

METHODS

Four datasets (62 ruptured IAs, 16 unruptured IAs, and 31 normal controls) were downloaded from the Gene Expression Omnibus. Differentially expressed genes (DEGs) were identified between the IAs and normal controls. All overlapping genes were analyzed using weighted gene co-expression network analysis. Functional enrichment analyses were performed using key modules. We then intersected the key module genes with DEGs. Protein-protein interaction networks were assessed to identify key hub genes. Least absolute shrinkage and selection operator logistic regression analysis was performed to construct a prediction model. A receiver operating characteristic curve was constructed to evaluate the reliability of the scoring system.

RESULTS

After intersection and normalization, 433 DEGs were identified and 15,388 genes were selected for weighted gene co-expression network analysis. The black module with 1145 genes exhibited the highest correlation with IA rupture. Many potential mechanisms are involved, such as the inflammatory response, innate immune response, extracellular exosome, and extracellular space. Thirty hub genes were selected from the protein-protein interaction, and 4 independent risk genes, TNFAIP6, NCF2, OSM, and IRAK3, were identified in the least absolute shrinkage and selection operator logistic regression model.

CONCLUSIONS

Our prediction model not only serves as a useful tool for assessing the risk of IA rupture, but the key genes identified herein could also serve as biomarkers and therapeutic targets.