Deciphering Molecular Biomarkers Governing Medulloblastoma Prognosis and Leptomeningeal Metastasis(2/2)

Project: A - Government Institutionb - Ministry of Science and Technology

Project Details


Since medulloblastoma (MB) discovered in 1920s, it became the most common malignant cerebellar tumor in children. MB has 4 histological subtypes: classic, desmoplastic/nodular (DNMB), extensive nodularity (MBEN), and large-cell or anaplastic (LC/A). In 2011, a large cohort genomic study classified MB to 4 molecular subgroups: SHH, WNT, Group 3, and Group 4. These molecular variants correlate with the demographic and clinical differences. In Taiwan, molecular subgrouping of MB is just established in only a few institutes using IHC staining only. However, for MB, the most significant clinical risk stratification is CSF (M1) and leptomeningeal metastasis (M2-3) at diagnosis. Patients with M1-3 tumor belong to high-risk group and need higher dose of craniospinal irradiation (CSI) for treatment, but still have poorer survival. Integrating with our clinical experience and relevant reported studies, the most significant factor influencing prognosis and therapeutic strategy is still leptomeningeal metastasis at diagnosis (Fig 1A, B) [17,34]. At present, the diagnosis of M2-3 metastasis relies on MRI of brain and spine studies. Therefore, to establishment of stream model of MB molecular subgrouping and deciphering metastasis genes of MB is the keystone of molecular identification of metastatic tumor, establishment of therapeutic target, and the capability of providing personalized targeted therapy to improve treatment outcome and increase cure rate. Preliminary preparations of this project included : 1. Based on clinical imaging studies, we have identified patients with or without leptomeningeal metastasis in a cohort series 152 MBs in children [17]. 2. Based on 22 reported subgroup-specific genes, in combination of microarray and RNA-Seq datasets from collected tumor tissues, we have subgrouped 23 MBs into 4 molecular subgroups (Fig. 2A). 3. Within these 4 subgroups, by further analysis of our 23 MBs microarray and RNA-Seq datasets, we found out 18 novel subgroup-specific signature genes for MB molecular subgrouping (Fig. 2B). 4. By applying 3 reported IHC specific markers for MB molecular subgrouping: beta-catanin, YAP1, and Filamin A, we establish our laboratory capability to subclassify MB into 3 subgroups (WNT, SHH, Non-WNT/SHH) by immunostaining (Fig. 3). 5. By applying qPCR and used 4 subgroup-specific signature genes: NPR3 and IMPG2, UNC5D, and GRM8, our lab can also subclassify non-WNT/SHH Group into Group 3 and Group 4 (Fig. 4). 6. Among the 23 MBs with transcriptome analysis, comparing with the cases with or without metastatic tumor at diagnosis, we identified 15 genes related to cell migration and highly correlated with tumor metastasis (Fig. 5). 7. We have also successfully established non-metastatic MB mouse model using Doy-Luc-cells and leptomeningeal-specific metastatic MB mouse model using Daoy-NMYC-Luc cell. These two models can be used for in vivo and ex vivo (metastatic MB tissue brain sliced tissue culture) studies to observe the migration/leptomemingeal-specific metastasis and to study the microenvironment and molecular mechanism facilitating metastasis. (Fig. 6, 7). 8. We have successfully established an patient derived xenograft (PDX) AT/RT mouse model (P2^P3 stage at present moment) (Fig. 8A, B, C). This experience can be applied to establish the PDX leptomeningeal-specific metastatic MB mouse model to support the related research project of metastatic MB. All of the above preliminary preparations are the bases of our clinical and laboratory supports to this project. Specific aims Aim 1: To establish stream model of MB molecular subgrouping and to decipher novel subgroup-specific signature genes for precise MB molecular subgrouping. Aim 2: To decipher leptomeningeal-specific metastasis genes for molecular diagnosis of leptomeningeal-specific metastatic MB and for prognosis evaluation, and personalized treatment. Aim 3: To establish organotypic brain slice culture technique and laboratory environment for the ex vi^vo study of leptomeningeal-specific metastatic MB using metastatic MB mouse model. The purpose is to observe the migration/metastasis of tumor cell from implanted site of brain to leptomeninges and to study the microenvironment and molecular mechanism involved in tumor metastasis for the establishment of targeted therapy. Aim 4: To establish PDX leptomeningeal-specific metastatic MB mouse model, a model more natural and close to the real clinical metastatic tumor condition. The animal model can be used as a platform for of the therapeutic effect study of targeted drug, a step before effective, personalized, targeted therapy.
Effective start/end date8/1/187/31/19


  • Medulloblastoma
  • Leptomeningeal-specific metastasis
  • Metastasis genes
  • Molecular subgroup
  • Metastasis mouse model
  • Brain slice culture