Predicting p53 Status in IDH-Mutant Gliomas Using MRI-Based Radiomic Model

Area of Science:

• Neuro-oncology
• Medical imaging
• Radiomics

Background:

• Accurate p53 status prediction is crucial for molecular stratification of isocitrate dehydrogenase-mutant (IDH-mt) gliomas.
• Noninvasive methods for assessing p53 status in IDH-mt gliomas are challenging but clinically significant.

Purpose of the Study:

• To evaluate the diagnostic efficacy of radiomics using pre-surgery contrast-enhanced T1-weighted imaging (CE-T1WI) for predicting p53 status in IDH-mt gliomas.
• To develop and validate radiomic models and a clinical-radiomics nomogram for p53 status prediction.

Main Methods:

• Radiomic features were extracted from three volumes of interest (entire tumor, peritumoral edema, and combined) in 78 IDH-mt glioma patients.
• Machine learning algorithms were used for feature selection and model development (Rad_VOI^T, Rad_VOI^PE, Rad_VOI^{T + PE}).
• A clinical-radiomics nomogram was constructed integrating age and radiomic scores, with performance assessed by ROC analysis.

Main Results:

• The peritumoral edema (VOI^PE) radiomic model showed the highest predictive performance (AUC 0.811 in training, 0.810 in validation).
• A clinical-radiomics nomogram incorporating age and radiomic scores achieved excellent discrimination (AUC 0.969 in training, 0.929 in validation).

Conclusions:

• CE-T1WI-based radiomic models can noninvasively predict p53 mutation status in IDH-mt gliomas.
• Textural features of peritumoral edema may better reflect tumor heterogeneity associated with p53 status.