Medulloblastoma (MB) is the most common malignant pediatric brain tumour. Of its four distinct molecular subgroups (WNT, SHH, Group 3, and Group 4), Group 3 MB patients hold the worst clinical prognosis due to their high incidence of tumour recurrence and metastasis to the spinal leptomeninges. Relapsed Group 3 MB patients, particularly those with MYC amplification (known as Group 3𝛾), carry a mortality rate of nearly 100%, given the paucity of effective therapeutic options currently available. The most common cause of mortality amongst these patients is leptomeningeal metastasis, yet this metastatic disease is poorly characterized. Our understanding of MB tumour recurrence and leptomeningeal metastasis is further encumbered by the rare clinical opportunities at which specimens may be collected from relapsed patients. We were able to circumvent this obstacle by establishing a patient-derived xenograft mouse-adapted therapy model, which mimics the treatments administered in the clinic and in turn, recapitulates both local and metastatic human MB recurrence. This model system enabled the collection of valuable, patient-derived Group 3𝛾 MB tumour cells from the spinal leptomeninges at relapse. It provides an opportunity for chemical screening, with the aim of identifying small molecule inhibitors capable of eradicating these cells. Existing studies on MB leptomeningeal dissemination at diagnosis suggest that effective treatments may target signalling proteins, such as phosphatidylinositol 3-kinases and histone deacetylases (HDACs). Therefore, I hypothesized that selective kinase and HDAC inhibitors would pose as effective therapies against Group 3𝛾 MB metastatic recurrences. In this thesis, I conducted a high-throughput screen of 640 kinase inhibitors and robust testing of novel HDAC6-selective inhibitors against these treatment-refractory, metastatic cells. Here, I showed that metastatic recurrences of Group 3𝛾 MB are therapeutically vulnerable to selective inhibitors of checkpoint kinase 1 (CHK1), platelet-derived growth factor receptor beta (PDGFRβ), and HDAC6. Functional studies demonstrated that these inhibitors selectively target the aggressive, migratory Group 3𝛾 MB cells, while sparing healthy neural stem cells. They also showed effective blood-brain-barrier penetration in silico and in vitro, while significantly reducing MB tumour cell properties that are often associated with treatment failure. Taken together, my thesis highlights specific inhibitors of CHK1, PDGFRβ, and HDAC6 that effectively target MB tumour cells that fuel treatment-refractory leptomeningeal metastases. With additional preclinical validation, these compounds may serve as potent therapeutic options to extend survival and improve the quality of life for patients that are ostensibly limited to palliation. / Thesis / Master of Science (MSc) / Medulloblastoma is an aggressive brain cancer in children. While current standard treatment improves patient survival, 30-40% of all medulloblastoma patients relapse at local (brain) or metastatic (spine) sites. Medulloblastoma metastatic recurrences remain poorly understood, yet they result in an alarmingly high mortality rate amongst patients due to inadequate treatment options currently available. Specific molecular targets are common in both medulloblastoma and metastatic cancer research. These targets are particularly important in governing cell signalling pathways that regulate tumour growth and migration. Therefore, treatment against these targets may be effective at preventing medulloblastoma metastatic recurrences. As the collection of local and metastatic tumour samples at patient relapse are rare, the Singh lab developed an animal model that mimics human medulloblastoma recurrence. In this thesis, recurrent medulloblastoma metastases were isolated from our established animal model and tested against compounds that inhibit the specific molecular targets previously implicated in medulloblastoma and other metastatic tumours. We identified potent compounds that effectively target these metastatic cells. Next, we determined which compounds spared healthy cells and were able to penetrate the brain, given our future objective of targeting these MB cells from their source to ultimately prevent metastasis. The identified compounds substantially reduced the ability of these cells to divide. With further investigation, these compounds may pose as effective therapeutic agents to treat human medulloblastoma patients with metastatic recurrences.

abstract: Achieving effective drug concentrations within the central nervous system (CNS) remains one of the greatest challenges for the treatment of brain tumors. The presence of the blood-brain barrier and blood-spinal cord barrier severely restricts the blood-to-CNS entry of nearly all systemically administered therapeutics, often leading to the development of peripheral toxicities before a treatment benefit is observed. To circumvent systemic barriers, intrathecal (IT) injection of therapeutics directly into the cerebrospinal fluid (CSF) surrounding the brain and spinal cord has been used as an alternative administration route; however, its widespread translation to the clinic has been hindered by poor drug pharmacokinetics (PK), including rapid clearance, inadequate distribution, as well as toxicity. One strategy to overcome the limitations of free drug PK and improve drug efficacy is to encapsulate drug within nanoparticles (NP), which solubilize hydrophobic molecules for sustained release in physiological environments. In this thesis, we will develop NP delivery strategies for brain tumor therapy in two model systems: glioblastoma (GBM), the most common and deadly malignant primary brain tumor, and medulloblastoma, the most common pediatric brain tumor. In the first research chapter, we developed 120 nm poly(lactic acid-co-glycolic acid) NPs encapsulating the chemotherapy, camptothecin, for intravenous delivery to GBM. NP encapsulation of camptothecin was shown to reduce the drug’s toxicity and enable effective delivery to orthotopic GBM. To build off the success of intravenous NP, the second research chapter explored the utility of 100 nm PEGylated NPs for use with IT administration. Using in vivo imaging and ex vivo tissue slices, we found the NPs were rapidly transported by the convective forces of the CSF along the entire neuraxis and were retained for over 3 weeks. Based on their wide spread delivery and prolonged circulation, we examine the ability of the NPs to localize with tumor lesions in a leptomeningeal metastasis (LM) model of medulloblastoma. NPs administered to LM bearing mice were shown to penetrate into LM mets seeded within the meninges around the brain. These data show the potential to translate our success with intravenous NPs for GBM to improve IT chemotherapy delivery to LM. / Dissertation/Thesis / Doctoral Dissertation Biomedical Engineering 2018