Search

Abstract Background: Adult glioblastomas (GBMs) are associated with high recurrence and mortality. Personalized treatment based on molecular markers may help improve the prognosis. We aimed to evaluate whether apparent diffusion coefficient (ADC) histogram analysis can better predict MGMT and TERT molecular characteristics and to determine the prognostic relevance of genetic profile in patients with GBM. Materials and Methods: MRI, clinical, and pathological data of 79 patients with GBM were retrospectively collected. The ADC values based on histogram analysis were described using 10th percentile (p10), 90th percentile (p90), mean, median, minimum, maximum, skewness, kurtosis, and entropy. The independent‐sample t test, linear correlation analysis, receiver operating characteristics (ROC) curve analysis, Kaplan–Meier analysis, and Cox proportional hazard regression were performed. Results: MGMT promoter methylation and TERT promoter mutation were detected in 53.2% and 44.3% of GBM patients, respectively. The ADCp10 in MGMT promoter unmethylated group was significantly lower than that in the MGMT promoter methylated group (p = 0.005). There were significant differences in ADCmin, ADCp10, ADCmean, and entropy between TERT promoter mutant and wild‐type groups. Entropy showed the best diagnostic performance in differentiating between positive and negative TERT groups (AUC = 0.722, p = 0.001). Overall survival (OS) showed a positive correlation with ADCmin. The TERT promoter mutation was the only independent prognostic factor for GBM. Conclusions: ADC histogram analysis may be a potential noninvasive biomarker for differentiating MGMT and TERT molecular markers and providing prognostic information for GBM patients.

Abstract Background and purpose As a crucial diagnostic and prognostic biomarker, telomerase reverse transcriptase (TERT) promoter mutation holds immense significance for personalized treatment of patients with glioblastoma (GBM). In this study, we developed a radiomics nomogram to determine the TERT promoter mutation status and assessed its prognostic efficacy in GBM patients. Methods The study retrospectively included 145 GBM patients. A comprehensive set of 3736 radiomics features was extracted from preoperative magnetic resonance imaging, including T2‐weighted imaging, T1‐weighted imaging (T1WI), contrast‐enhanced T1WI, and fluid‐attenuated inversion recovery. The construction of the radiomics model was based on integrating the radiomics signature (rad‐score)with clinical features. Receiver‐operating characteristic curve analysis was employed to evaluate the discriminative ability of the prediction model, and the risk score was used to stratify patient outcomes. Results The least absolute shrinkage and selection operator classifier identified 10 robust features for constructing the prediction model, and the radiomics nomogram exhibited excellent performance in predicting TERT promoter mutation status, with area under the curve values of.906 (95% confidence interval [CI]:.850–.963) and.899 (95% CI:.708–.966) in the training and validation sets, respectively. The clinical utility of the radiomics nomogram is further supported by calibration curve and decision curve analyses. Additionally, the radiomics nomogram effectively stratified GBM patients with significantly different prognoses (HR = 1.767, p = .019). Conclusion The radiomics nomogram holds promise as a modality for evaluating TERT promoter mutations and prognostic outcomes in patients with GBM.

Abstract Background and purpose The presence of TERT promoter mutations has been associated with worse prognosis and resistance to therapy for patients with glioblastoma (GBM). This study aimed to determine whether the combination model of different feature selections and classification algorithms based on multiparameter MRI can be used to predict TERT subtype in GBM patients. Methods A total of 143 patients were included in our retrospective study, and 2553 features were obtained. The datasets were randomly divided into training and test sets in a ratio of 7:3. The synthetic minority oversampling technique was used to achieve data balance. The Pearson correlation coefficients were used for dimension reduction. Three feature selections and five classification algorithms were used to model the selected features. Finally, 10‐fold cross validation was applied to the training dataset. Results A model with eight features generated by recursive feature elimination (RFE) and linear discriminant analysis (LDA) showed the greatest diagnostic performance (area under the curve values for the training, validation, and testing sets: 0.983, 0.964, and 0.926, respectively), followed by relief and random forest (RF), analysis of variance and RF. Furthermore, the relief was the optimal feature selection for separately evaluating those five classification algorithms, and RF was the most preferable algorithm for separately assessing the three feature selectors. ADC entropy was the parameter that made the greatest contribution to the discrimination of TERT mutations. Conclusions Radiomics model generated by RFE and LDA mainly based on ADC entropy showed good performance in predicting TERT promoter mutations in GBM.