Researchers at St. Jude Children’s Research Hospital examined data from three clinical trials involving nearly 900 children with the most common pediatric brain tumor, medulloblastoma. The analysis incorporated genomic, molecular, and survival information to develop new risk categories that allow treatment to be safely lowered, thereby reducing long‑term side effects.
The study, published in Neuro-Oncology and Cancer Research, shows that about 40 % of patients may receive reduced doses of craniospinal radiation, and nearly all can receive less chemotherapy than in the original trials while maintaining equal or improved survival rates.
Lead author Dr. Giles Robinson explained, “By re‑examining trial outcomes with molecular data—such as DNA sequences, methylation signatures, and chromosome changes—we have identified patients who can safely receive less intensive therapy.”
Using cross‑study comparisons, the team identified new prognostic factors. For instance, patients whose tumors fall into the dominant G3 or G4 molecular classes can be subdivided by chromosomal gains or losses, methylation subgroup, and MYC amplifications, enabling a four‑tier risk stratification from low to high intensity.
“Medulloblastoma is highly heterogeneous,” said Robinson. “Our findings provide a clearer framework for tailoring treatment intensity so that only the most aggressive tumors receive the most aggressive therapy.”
To support clinical implementation, the researchers are launching a trial that will test these risk categories, facilitated in part by a web‑based portal.
Recognizing the complexity of the molecular data, the team created the Medulloblastoma Meta‑Analysis (MB‑meta) Portal. Combining the datasets from the two studies, the portal offers a user‑friendly interface where physicians, scientists, and patients can select demographic, clinical, and molecular parameters and view projected survival curves for matching individuals.
“The portal makes it straightforward for anyone to generate survival predictions,” said co‑author Dr. Xin Zhou, of St. Jude’s Department of Computational Biology. “It translates complex data into actionable insights.”
As a demonstration, the portal was used to investigate novel mutations in the KBTBD4 gene. Two distinct mutation subgroups were identified, suggesting new ways to classify these patients and revealing previously unrecognized disease biology that may inform future therapies.
Robinson added, “By sharing this tool and data, we invite the research community to build upon our work, ultimately improving quality of life for future medulloblastoma survivors.”