Heterogenous radiomics patterns are associated with poor survivals and dysregulated pathways in medulloblastoma

Background. To develop a radiomics signature for predicting overall survival (OS) and progression-free survival (PFS) in patients with medulloblastoma (MB), and to investigate the incremental prognostic value and underlying biological pathways of the radiomics signature.Methods. A radiomics signature was constructed based on five magnetic resonance (MR) sequences (T1, T1c, T2, FLAIR and ADC) from a training cohort (n = 83) using LASSO-Cox model. Association between the signature and survival was evaluated on a testing cohort (n = 83). Incremental prognostic value of the signature beyond clinicomolecular factors was assessed with respect to calibration, discrimination, reclassification and clinical usefulness by a radiomics-clinicomolecular nomogram. Key pathways associated with the signature were identified. Prognostic value of pathway genes was assessed in a public TCGA cohort.Results. The radiomics signature was significantly associated with OS/PFS, independent from clinicomolecular factors. The radiomics-clinicomolecular nomogram predicted OS (C-index 0.762) and PFS (C-index 0.697) better than either the radiomics signature (C-index: OS: 0.649; PFS: 0.593) or the clinicomolecular nomogram (C-index: OS: 0.725; PFS: 0.691) only, with a better calibration and classification accuracy (net reclassification improvement: OS: 0.298, P = 0.022; PFS: 0.252, P = 0.026). Nine pathways were significantly correlated with the radiomics signature. Average expression value of pathway genes achieved significant risk stratification in public cohort (log-rank P = 0.016). Conclusion. This study demonstrated radiomics signature was an independent prognostic marker and offered incremental value over clinicomolecular factors in OS/PFS predictions for MB patients. The signature was associated with dysregulated pathways affecting patient prognosis in MB. Radiomics patterns predict survival of MB