PURPOSE:Glioblastomamultiforme(GBM)is amalignant braintumor withpoor prognosis and highmorbidity due to its invasiveness. Hypoxia-drivenmotility(HM) and concentration-driven motility (CM) are two mechanisms ofGBM invasion in the brain. The use of anti-angiogenic drugs has uncoverednew progression patterns of GBM associated with significant differences inoverall survival times. Here, we test the hypotheses that the types and rates ofGBM motility predict its progression pattern and the patients’ survival times.METHODS:We applied a mathematical model of GBMgrowth and invasionin humans to simulate a clinical trial and study the effects of the rate and mechanismof motility on the patterns of progression and on survival times.RESULTS: The motility phenotype appears to determine the progressionpattern as well as the survival time of a patient treated by anti-angiogenesis.Highly-dispersive tumors are associated with the longest survival times (p ,0.001) and with progression by Expanding FLAIR. Moderately-Dispersivetumors are associated with short survival times and with progression byExpanding FLAIR + Necrosis. Tumors withHMare associated with the shortestsurvival times and with progression by Expanding Necrosis. The survivaltimes of the latter are similar to non-responders. This investigation also uncoveredtheHM-CM principle: the aggressive HM-dependent phenotype surfacesonly when therate of CM is low in both untreated and bevacizumab-treatedGBM.CONCLUSIONS: Finding that the motility phenotype is a fundamentalproperty that controls progression and survival times, has biological, clinicaland therapeutic implications.< Réduire