The value of multi-parametric magnetic resonance imaging (MRI) radiomics in predicting isocitrate dehydrogenase (IDH) wildtype with telomerase reverse tranase (TERT) promoter mutation in glioma.

Background: Accurate preoperative prediction of gliomas is crucial for formulating personalized treatment decisions and assessing prognosis. With the release of the 2021 World Health Organization (WHO) classification of central nervous system tumors, molecular diagnostics have demonstrated even greater importance. Isocitrate dehydrogenase (IDH) is a significant marker for evaluating glioma prognosis, while the telomerase reverse tranase (TERT) promoter has a dual impact on glioma prognosis. In this study, we aimed to explore the application value of preoperative multi-parametric magnetic resonance imaging (MRI) radiomics in the prediction of IDH wildtype with TERT promoter mutation in glioma. Methods: Preoperative MRI images and genetic data of 415 glioma patients from three centers were retrospectively analyzed, of which 297 patients were categorized into training and test sets, and the remaining 118 patients were classified into independent external test sets. A total of 3,591 radiomics features were extracted from the tumor region of interest (ROI) of T2 fluid-attenuated inversion recovery (T2-FLAIR), contrast-enhanced T1-weighted imaging (CE-T1WI), and apparent diffusion coefficient (ADC) MRI images. Feature selection was performed using the Pearson rank correlation coefficient and the least absolute shrinkage and selection operator (LASSO). Logistic regression was used to construct the imaging histology model and the clinical model, and the combined model was constructed. The models were evaluated by receiver operating characteristic (ROC) curves, area under the curve (AUC), sensitivity and specificity. Results: Radiomics models based on individual sequences and multiple-sequence fusion were constructed. The fused sequence model outperformed the single-sequence models, with AUC values of 0.954, 0.867, and 0.816 in the training set, test set, and external validation set, respectively. Age and grading, as reliable prognostic factors, were used to construct a clinical model. When radiomics features were added, a combined model was established. The combined model demonstrated the highest performance, with AUC values of 0.963, 0.905, and 0.823 in the training set, test set, and external validation set, respectively. Calibration curves and decision curve analysis (DCA) indicated good calibration capability and clinical applicability. Conclusions: Radiomics based on preoperative MRI can effectively predict the molecular subtype of IDH wildtype gliomas with TERT promoter mutation.