Browsing by Subject "Tumor Modeling"

Now showing 1 - 1 of 1
  • Thumbnail Image
    Item
    Physical Determinants of Glioma Cell Migration and Disease Progression
    (2015-09) Klank, Rebecca
    Glioblastoma (GBM) is a highly aggressive brain cancer (generally, “glioma”) with poor patient prognosis, even with current standard treatments. In order to rationally develop novel treatments that can significantly extend patient survival, we must first understand at a basic scientific level how the disease progresses. GBM is thought to be fatal due to highly invasive cells that migrate beyond the visible bulk tumor and lead to tumor recurrence after therapeutic intervention. Therefore, we sought to investigate what makes GBM cells invasive at the single-cell level (Chapter 1). Using a genetically induced mouse glioma model and confocal imaging of intact tumor-containing brain slices, we found that, consistent with previous biophysical models, glioma cell migration is biphasic with respect to the concentration of the transmembrane cell adhesion molecule CD44. By contrast, cell proliferation is independent of CD44 level. Additionally, mouse model and human patient survival are also biphasic with respect to CD44 level, with poorest prognosis occurring at intermediate CD44 levels. Thus, migration and survival are both biphasic and are anti-correlated to each other, suggesting that CD44-dependent migration directly affects survival outcomes. We next investigated how these single-cell behaviors impact overall tumor growth and progression (Chapter 2). Noticing that previous models of GBM migration use parameter values for migration rate (defined by a diffusion coefficient, also known as a random motility coefficeint) that are much higher than our measurements of single-cell migration behavior in Chapter 1, a Brownian dynamics (BD) approach was used to simulate single-cell growth, proliferation, and migration, and compare model assumptions. These studies showed that employing the physically-based assumption that tumor cells occupy volume, an assumption not captured in current reaction-diffusion (RD) simulations, resulted in increased tumor spreading behavior with the same input parameters. Specifically, non-overlapping cells can enter a jammed regime where interior cells are subdiffusive, and peripheral cells become biased outward and superdiffusive in a quasi-ballistic expansion. Thus, we show that, when we account for volume conservation, the relatively low values of diffusion coefficient, such as what was measured in Chapter 1, can generate fast progressing tumors that are similar to RD simulations which use diffusion coefficients much greater than what is observed experimentally for single migrating cells. Therefore, we suggest that cellular jamming behavior contributes to the fast spreading of GBM tumors, and that subsequent simulations of GBM growth should incorporate this assumption so that models are physically grounded and achieve consistency between single-cell behavior and bulk tumor progression. Overall, these studies demonstrate the potential importance of fundamental physical effects in driving tumor progression generally, and glioblastoma specifically.