Abstract 711: The Medulloblastoma Meta-Analysis Portal for integrative exploration of clinical and genomic data from 898 patients

Abstract Medulloblastoma (MB) is one of the most malignant pediatric brain tumors. Molecularly, MB separates into four subgroups—WNT, SHH, Group 3 (G3), and Group 4 (G4)—each with unique clinical features and genetic profiles. WNT tumors have the best survival outcomes and low metastatic potential, whereas G3 tumors exhibit the highest malignancy and poorest outcomes. While WNT and SHH are highly distinct, G3 and G4 tumors are interrelated and when grouped together, these are referred to as G3/G4 tumors. These tumors are genetically complex, with a constellation of molecular features including MYC/MYCN amplifications, OTX2 activation, KBTBD4 mutations, and 8 molecular subgroupings complicating risk stratification. Multi-omics approaches, especially DNA methylation profiling, have advanced our understanding, but further exploration is needed to clarify the genetic heterogeneity of G3/G4 tumors. To address this complexity, we developed the Medulloblastoma Meta-Analysis Portal, incorporating data from 898 patients across three major clinical trials (ACNS0331, ACNS0332, SJMB03) to facilitate cross-cohort data exploration. Our portal offers an array of robust tools for investigating genetic, cytogenetic, and clinical features of medulloblastoma, including custom variable creation, cohort filtering, integration of molecular data with tumor methylome t-SNE maps, and on the fly Cox regression. As a use case the portal enables the assessment of MYC amplification and isochromosome 17 cytogenetic association with progression-free survival of standard-risk G3/G4 tumors. This exercise reveals that MYC amplification is significantly associated with worse outcomes irrespective of isochromosome 17 status suggesting MYC amplification as the prognostic marker and justifying its inclusion as a high-risk feature in risk-stratification algorithms. In the second use case, we overlay mutation status on the methylome t-SNE map to reveal complex relationships between genetic alterations and DNA methylomic state. Given that KBTBD4 is mutated with complex insertions in G3/G4 tumors we explored its relationship among G3/G4 subgroups. By manually reviewing and rescuing mutation calls, the portal uncovered novel mutation patterns associated with distinct methylome subgroups, as well as outlier cases. These findings emphasize the heterogeneity within G3/G4 subgroups and underscore the limitations of current classification methods, which may fail to capture nuanced genetic differences. In summary, the portal is a valuable resource for detailed exploration of medulloblastoma clinical features, methylomic classifications, and driver mutations that can inform both basic research and clinical decision-making. The portal will be freely available at https://proteinpaint.stjude.org/mbmeta. Citation Format: Xin Zhou, Karishma Gangwani, Congyu Lu, Sandeep Dhanda, Airen Peraza, Edgar Sioson, Gavriel Matt, Robin Paul, Jian Wang, Colleen Reilly, Alexander Acic, Kyle Smith, Tong Lin, Qian Li, Paul A. Northcott, Giles W. Robinson. The Medulloblastoma Meta-Analysis Portal for integrative exploration of clinical and genomic data from 898 patients [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2025; Part 1 (Regular Abstracts); 2025 Apr 25-30; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2025;85(8_Suppl_1):Abstract nr 711.