MYC-driven Medulloblastoma : New Targeted Therapies and Mechanisms of Recurrence

Sammanfattning: Medulloblastoma is the most common malignant brain tumor of childhood. It arises in the posterior fossa but presents with distinct histological and molecular features. Hence, medulloblastoma is divided into four molecular subgroups, WNT, SHH, Group 3, and Group 4. The overall 5-year survival is ~70% across subgroups but varies with high- and low-risk disease. Standard treatment of medulloblastoma consists of maximal safe tumor resection, radiotherapy, and adjuvant chemotherapy. Despite the rather high success rate of treatment for many patients it also comes with severe long-term debilitating side effects. MYC proteins are master regulators of gene expression often deregulated in cancer. MYC family members MYC and MYCN share similar roles and are found overexpressed or amplified in most medulloblastoma subgroups and correlate with a poor prognosis. Medulloblastoma dissemination and recurrence patterns differ between subgroups but are always associated with a poor prognosis. Recurrent medulloblastoma is not yet curable and will lead to death. In this thesis, we present the first transgenic mouse model of medulloblastoma recurrence and highlight the role of the transcription factor SOX9 in MYC-driven relapse mechanisms. By studying this recurrence model and matched primary-recurrent patient samples we propose a mechanism in which treatment-refractory and quiescent SOX9-positive cells in Group 3 medulloblastoma are necessary for tumor relapse, and how the recurrent tumors can be specifically treated with MGMT inhibitors and doxorubicin.In addition, we address efficient treatment options of primary MYC-driven medulloblastoma where BET bromodomain inhibition (JQ1) in combination with CDK2 inhibition (milciclib) of human Group 3 medulloblastoma will lead to apoptosis and prolonged survival of xenografted mice. This is explained by a dual hit on MYC transcriptional output and MYC protein stability exerted by JQ1 and milciclib respectively. Finally, in a different novel transgenic model of MYC-driven medulloblastoma, we show how temporal Cdk2 knock-out has no effect on MYC protein stability but slows down proliferation and prolongs survival of allografted mice. The need for better treatments and increased understanding of recurrent medulloblastoma is huge. To that end, this thesis focuses on and addresses novel treatments, the role of the cell cycle protein CDK2 as well as relapse mechanisms depending on dormant SOX9-positive cells in highly aggressive MYC-driven medulloblastoma.