Long-term outcomes by clinical and molecular risk stratification in patients with medulloblastoma receiving...

Risk stratification and radiotherapy optimization for medulloblastoma (MB) increasingly rely on molecular classification, yet large-scale, cross-regional, real-world evidence remains limited.

We assembled a global MB database (Chinese cohort: N = 729; international: N = 494). Overall survival (OS)/progression-free survival (PFS) and risk stratification were analyzed by integrating clinicopathological data with molecular subgroup-specific genetic events using Kaplan-Meier and Cox regression models.

While conventional clinical risk stratification effectively separated outcomes, metastatic stage alone, although a crucial factor, failed to adequately stratify survival across molecular subgroups. The average-risk group showed superior 5-year cumulative OS (84.4%) and PFS (77.5%), compared to the high-risk group (77.2% and 68.2%). Key genetic events, including TP53 mutation/17p loss (Sonic Hedgehog, SHH), MYC amplification (group_3), and CDK6 activation (group_4), were strongly associated with survival. Integrating genetic events with clinical features significantly facilitated prognostic discrimination. Importantly, radiotherapy dose-response effects were highly context dependent within defined molecular subgroups. No significant differences in OS or PFS were observed between craniospinal irradiation (CSI) doses of 30.6 and 36.0 Gy in high-risk patients with SHH, group_3, and group_4 receiving chemotherapy. In particular, 36.0 Gy CSI was warranted for aggressive MYC-amplified group_3 MB, even without dissemination.

Integrating subgroup-specific genetic events with clinical features facilitates MB risk stratification, informs context-dependent radiotherapy consideration, and provides actionable candidates for prospective validation.

This work was funded by the Natural Science Foundation of Beijing (JQ24040 and L2510021) and the National Natural Science Foundation of China (82273343 and 82573715).