| Abstract | Gliomas are the most prevalent primary malignant brain tumors in adults and are currently incurable. Experimental work done by Assanah, et al, 2006 demonstrated that very similar tumors can be initiated in rats by infecting them with a retrovirus (PDGF-IRES-GFP) that lends the capacity to over-express platelet-derived growth factor (PDGF). The key observation from this study was that the majority of cells within all such tumors were glial progenitor cells uninfected with the retrovirus, comprising 70 – 85% of the tumors. Infected PDGF-IRES-GFP+ glial progenitors made up the remainder. This finding strongly suggested that the observed recruitment to the tumor was a direct result of paracrine PDGF signaling. Given this knowledge, we have created and parameterized a mathematical model of experimental glioma initiation and progression, incorporating the experimentally observed recruitment dynamic. In creating our model, we took a multi-scale continuous modeling approach, treating the various cell populations as densities in tissue and unbound PDGF as a concentration in the extracellular fluid. This model accounts for the dispersal and proliferation of both infected and uninfected glial progenitor populations due to PDGF signaling, the secretion of PDGF into the extracellular space by PDGF-IRES-GFP+ (infected) glial progenitors, and PDGF consumption by both infected and uninfected glial progenitor cell populations. Additionally, we assumed that while the PDGF-IRES-GFP+ cells receive both autocrine and paracrine PDGF signals, the other glial progenitor cells receive only paracrine signals. Recruitment model dynamics, after reaching an MRI-detectable density, exhibit the same velocity of linear radial expansion as displayed by a model tumor without recruitment; thus, our model confirms the reasonability of a recruitment mechanism acting in human glioma. Moreover, the differing effects of paracrine and autocrine signals on the glial progenitors described above could explain the differing rates of progression observed in glioma patients. Our model predicts faster rates of radial expansion at the MRI detectable level for more paracrine-driven tumors than autocrine-driven tumors, which exhibited slower expansion in simulations due to greater diffusivity. |
|---|
