Clinically Available Prediction Models for IDH and ATRX Gene Status in Adult-type Diffuse Gliomas Based on VASARI Features

Motivation: Glioma patients with different IDH and ATRX mutation status indicate specific classification and different prognosis. Goal(s): To non-invasive predict IDH and ATRX status based on preoperative MRI VASARI features. Approach: Multivariable logistic regression and nomograms were employed to develop the prediction models. Results: Six VASARI features combined with age and relative apparent diffusion coefficient (rADC) contributed to the IDH prediction model, achieving an AUC of 0.96 and 0.92 in the training and validation set. Three VASARI features combined with age and minimum ADC (ADCmin) contributed to the ATRX prediction model, achieving an AUC of 0.76 and 0.71 in the training and validation set. Impact: Identifying IDH mutation and ATRX mutation status based on VASARI features is helpful for glioma risk stratification.