1) Exploring Bias in Medical Data, 2) Signal Transduction Pathways Regulating Neuroblastoma Metastasis, 3) Investigating the Relationship Between the Leptomeningeal Microenvironment and Metastatic Medulloblastoma Cells

Abstracts

Cathy Ong Ly (Supervisor: Chris McIntosh)

Coronary heart disease is the second leading cause of death in Canada affecting approximately 1 in 12 adults ages 20 and over. Two interventions for myocardial revascularization to treat coronary artery disease are percutaneous coronary intervention (PCI) and coronary artery bypass graft (CABG). Comparisons of patient outcomes from these interventions is critical and led to several randomized clinical control trials that have been conducted over the past decade, most notably the SYNTAX trial. The SYNTAX trial, included the development of the SYNTAX score, an angiographic score that is currently used to evaluate the complexity of coronary artery disease. While the SYNTAX score is currently considered when deciding cardiac interventions for patients, studies have shown that the complexity of measuring the SYNTAX score, results in appreciable inter-observer variability.

We hypothesize that machine and deep learning techniques can be applied to clinical features and angiograms for risk stratification of PCI versus CABG to improve on the accuracy of current models.

Bias is an important consideration in medical data. Our efforts to investigate bias in our data and other datasets will consist of building a pipeline to elucidate any underlying biases from clinical features or fluoroscopic scanners. Subsequently, machine and deep learning pipelines will be built to analyze clinical features and X-ray angiograms. The result of this work is to develop a reproducible biomarker to improve the clinical decision-making process for risk-stratification of patients for cardiac intervention.

Jenna Park (Supervisor: Meredith Irwin)

Neuroblastoma (NB), a neural crest-derived tumour, is the most common extracranial pediatric cancer, accounting for 8-10% of all childhood cancers. The most frequent primary tumour site of NB are the adrenal glands, a site of abundant sympathetic neuron innervation, and these tumours are often responsive to various treatment interventions. However, metastasis often arises and occurs in 50-70% of patients. Most metastatic tumours in children over 1.5 years of age are fatal, with a five-year survival rate of less than 50%. Common metastatic sites of NB include the bone and the brain, where neural crest-derived cells are not commonly seen. These primary tumour cells must successfully complete the metastatic cascade to be able to survive and thrive in these niches. Thus, this raises the question of whether or not there are kinase or signaling pathway contributions that enhance the metastatic potential of NB to these niches. Our lab has previously developed a mouse model of NB to study metastasis to the bone and the brain, isolated tumour cells from these metastatic sites, and identified genes regulating their metastatic behavior. However, transcriptional changes do not always inform us about proteomic or post-translational modification changes that regulate intracellular signaling pathways that I hypothesize are required for NB cells to metastasize and hone to the bone or brain. Therefore, I plan to identify the cell surface receptors and the signaling pathways they activate, which I propose respond to growth factors in their environment to enable NB cells to metastasize to and thrive in different sites. I am using mass and phospho-mass spectrometry to identify these proteins and pathways and assessing the activation of receptors and signaling pathways known to be important in other cancers that metastasize to the bone or brain. I will then determine the involvement of candidate receptors and pathways using both genetic and pharmacological methods to block their function. These experiments will potentially discover new therapeutic approaches targeted to metastatic neuroblastoma.

Olga Sirbu (Supervisor: Michael D. Taylor)

While the majority of morbidity and mortality in medulloblastoma patients are due to metastatic disease, most research focuses on the primary tumor due to a shortage of metastatic tissue samples and model systems. Medulloblastoma metastases are found almost exclusively on the leptomeningeal surface of the brain and spinal cord and dissemination occurs haematogenously by spreading through the blood to the leptomeningeal space to form metastases. However, very little is known about the leptomeningeal microenvironment and its relationship with metastasis and there are currently no drugs approved or in clinical trials for the treatment of this disease. Furthermore, all patients who receive treatment for medulloblastoma are presumed to have diffuse leptomeningeal metastases due to the standard of care treatment they receive, which is irradiation of the entire brain and spinal cord to children over 3 years of age. This treatment leaves children with neurological, cognitive, and endocrinological impairments. As such, mortality from medulloblastoma almost always results from metastatic dissemination of the primary tumour to the leptomeninges.

The purpose of my project is to investigate the relationship between the leptomeningeal microenvironment and metastatic medulloblastoma cells and as such, I hypothesize that the leptomeninges are secreting factors that support the survival of metastatic tumour cells in this niche. To do this, I will conduct a secretome analysis using the leptomeninges of two metastatic medulloblastoma mouse models, one of which is a transgenic model and the other is a patient derived xenograft model. By analyzing what proteins are secreted specifically by the leptomeninges, we can see what allows those cancer cells to be able to adapt and modify their surroundings to generate an environment favoring metastatic outgrowth. Furthermore, I will use immunofluorescence to validate the interactions of proteins of interest with their metastatic microenvironment to better understand this niche. I will also characterize the spatial organization of the tumour microenvironment using scanning and transmission electron microscopy to validate my proteomics work and see if there is a difference in metastatic localization along the spinal leptomeninges. Ultimately, I hope to validate my secretome results using the results of single-cell analysis on the metastatic leptomeningeal microenvironment done by a post-doc in my lab and map both transcriptomics and proteomics to pinpoint targetable cells of interest. As such, I believe findings from my project will present an opportunity to improve treatment options for this highly burdened population.