5T chemical exchange saturation transfer imaging improves glioma grading and genotyping prediction: a supplement to 3T diffusion and perfusion MRI

Accurate grading and genotyping of gliomas are critical for tailoring personalized therapeutic strategies and predicting patient outcomes. This study investigates the potential of 5T glutamate chemical exchange saturation transfer (GluCEST) imaging in differentiating glioma grades and predicting pivotal molecular biomarker status. We first validated the specificity of GluCEST signals to glutamate via phantom studies, then quantified potential interference from other metabolites. Thirty-four newly diagnosed glioma patients underwent preoperative 5T GluCEST imaging. The correlation between GluCEST values and the Ki-67 labeling index (LI) was analyzed using Spearman’s rank correlation. Furthermore, the capability of GluCEST values to predict glioma grade and genotype was assessed using receiver operating characteristic (ROC) curves and the area under the curve (AUC) metrics. These results were compared to advanced 3T MRI techniques, including relative cerebral blood volume (rCBV), apparent diffusion coefficient (ADC), and fractional anisotropy (FA). GluCEST signals were predominantly driven by glutamate and exhibited a significant positive correlation with phantom glutamate concentration. A strong correlation was also observed between GluCEST values and the Ki-67 LI (r = 0.565, P < 0.001). Notably, GluCEST imaging demonstrated high diagnostic performance in distinguishing low-grade gliomas (LGG) from high-grade gliomas (HGG) (AUC = 0.90, P < 0.001), as well as in predicting IDH mutation status (AUC = 0.88, P < 0.001) and 1p/19q co-deletion status (AUC = 0.87, P < 0.001). By contrast, rCBV and ADC only showed moderate potential in identifying MGMT methylation (AUC = 0.73, P = 0.018) and EGFR amplification (AUC = 0.69, P = 0.049), respectively. In conclusion, 5T GluCEST imaging provides complementary metabolic information to 3T MRI, showing strong potential as a reliable non-invasive tool for differentiating LGG from HGG and for predicting IDH mutation and 1p/19q co-deletion status.