Expression data from 30 medulloblastomas

Pediatric medulloblastoma is considered a highly heterogeneous disease, and a new strategy of risk stratification to optimize therapeutic outcomes is required. We aimed to investigate a new risk-stratification approach based on expression profiles of medulloblastoma cohorts. We analyzed gene expression profiles of 30 primary medulloblastomas and detected strong evidence that poor survival outcome was significantly associated with mRNA expression profiles of 17p loss. However, it was not supported in independent cohorts from previously published data (n=100). We speculated that this controversy might come from complex conditions of two important prognostic determinants, loss of tumor suppressors (chromosome 17p) and high expression of oncogenes, c-myc (MYC) or N-myc (MYCN). Simultaneous consideration of these two factors led to a new subgrouping of patients, exhibiting obviously different survival expectancies between the subgroups. Patients with up-regulated WNT signalings were always pre-defined as an independent subgroup, which ultimately removed confounding effect arising from contradictory outcome, favorable prognosis of WNT medulloblastomas despite their high MYC/MYCN expression level. We also found that age is a significant prognostic marker after adjusting for 17p and MYC/MYCN status. Diminished survival in age <3 years was more substantial in groups with high expression of MYC/MYCN or 17p loss, indicating survival outcome might be coordinately affected by these three factors. We suggest a more tailored and easily applicable subgrouping system based on expression profiles of chromosome 17p and MYC/MYCN, while separating WNT medulloblastoma as an independent subgroup, which could provide the basis for a novel risk-stratification strategy in pediatric medulloblastoma. We analyzed gene expression profiles of 30 primary medulloblastomas with the aim of providing a new risk-stratification approach based on expression profiles of medulloblastoma cohorts.