MDB-75. Molecular Pathology, Treatment and Prognosis of Infant Sonic Hedgehog Medulloblastoma: A Global Multi-Cohort Study

Abstract BACKGROUND Clinical studies in infant medulloblastoma (iMB; 5 years) have, to date, focussed on modestly-sized or national trials cohorts and have not directly compared therapeutic approaches or placed these in the context of dedicated biomarker studies, to inform future treatment strategies. METHODS We assembled a multi-national cohort of molecularly and clinically-annotated iMBs (5 years at diagnosis; n=646), the largest to date. We investigated molecular pathology, treatments received, and relationships to outcome within this population. RESULTS The SHH group (iMBSHH; n=267, 40%) predominated, encompassing SHH-1 (37.7%, median age 2.0 years), SHH-2 (47.0%, 1.4 years) and SHH-3 (14.4%, 3.0 years) WHO subgroups. MBEN histology was significantly enriched in SHH-2, and SUFU mutation and MYCN amplification in SHH-1 and SHH-3, respectively. Notably, TP53 mutations were identified in all subgroups; in SHH-1 and SHH-2 (each n=3) tumours lacked features typically associated with TP53-mutated SHH-3 (LCA histology and MYCN amplification) and did not have a worse survival. Upfront radiation-sparing treatments were used in 132/267 children and comprised regimens founded on intraventricular methotrexate (IVT-MTX; 54.5%), high-dose (HDCTx; 22.0%) or standard-dose (23.5%) chemotherapy. Across all radio-naive iMBSHH, non-DN/MBEN histology (HR 2.76, CI 1.26-6.01, p=0.011) and SHH-1 (vs SHH-2, HR 2.51, CI 1.25-5.01, p=0.009) conferred worse PFS in univariable analysis; subgroup was the only independently prognostic risk-factor (multivariable analysis; SHH-1 HR 2.65, CI 1.27-5.51, p=0.009). 5-year OS for HDCTx recipients was very favourable in SHH-1 (100%) and SHH-2 (93.3%), and subgroup-dependent for IVT-MTX (SHH-1; 76.4%, SHH-2; 100%, p=0.009). In CSI-treated iMBSHH (n=49), non-DN/MBEN histology (HR 10.23, CI 2.22-47.04, p=0.003) and MYCN amplification (HR 7.35, CI 2.01-26.87, p=0.003) conferred worse PFS. CONCLUSION iMBSHH outcomes in this cohort are dependent upon WHO subgroup, histology and therapy received. These findings provide an evidence-based foundation for selection of cohorts and therapies for prospective assessment in forthcoming iMBSHH clinical trials (e.g. SIOP-CONNECT-YC-MB-LR).