Integration of MRI radiomics and germline genetics to predict the IDH mutation status of gliomas

Gliomas are highly fatal and heterogeneous brain tumors. Molecular subtyping is critical for accurate diagnosis and prediction of patient outcomes, with isocitrate dehydrogenase ( IDH ) mutations being the most informative tumor feature. Molecular subtyping currently relies on resected tumor samples, highlighting the need for non-invasive, preoperative biomarkers. We investigated the integration of glioma polygenic risk scores (PRS) and radiomic features for prediction of IDH mutation status. The elastic net classifier was trained on a panel of 256 radiomic features from preoperative MRI scans, a germline PRS for IDH mutation and demographic information from 159 glioma cases in The Cancer Genome Atlas. Combining radiomics features with the PRS increased the area under the receiver operating characteristic curve (AUC) for distinguishing IDH-wildtype vs. IDH-mutant glioma from 0.824 to 0.890 (P _ΔAUC =0.0016). Incorporating age at diagnosis and sex further improved the classifier (AUC=0.920). Our multimodal classifier also predicted survival. Patients predicted to have IDH-mutant vs. IDH-wildtype tumors had significantly lower mortality risk (hazard ratio (HR)=0.27, 95% CI: 0.14-0.51, P=6.3×10 ⁻⁵ ), comparable to prognostic trajectories observed for biopsy-confirmed IDH mutation status. In conclusion, our study shows that augmenting imaging-based classifiers with genetic risk profiles may help delineate molecular subtypes and improve the timely, non-invasive clinical assessment of glioma patients.