Patient-specific mathematical radiation oncology: Integrating clinical imaging and modeling to improve outcomes
Session type: Parallel sessions
Glioblastomas are uniformly fatal primary brain tumors that diffusely invade the surrounding normal-appearing tissue at large distances from the frank imageable abnormality. The diffuse invasion combined with the marked phenotypic heterogeneity across patients challenges the applicability of the clinical standard approach of a "one-size-fits-all" radiation treatment planning approach. We will discuss a series of mathematical modeling approaches that:
1) quantify net rates of proliferation and migration of tumor cells to characterize the tumor's growth and invasion along with the linear-quadratic model for response to radiation therapy,
2) allow accurate quantitative prediction of patient-specific radiosensitivity taking as inputs routine clinical pre-treatment MRIs,
3) demonstrate that incorporation of spatial hypoxia information provided by FMISO-PET produce an enhanced accuracy of the spatio-temporal predictions of the tumor evolution through treatment, and
4) illustrate that by tuning to each patients predicted radiosensitivity and tumor growth kinetics, multiobjective evolutionary algorithm for IMRT optimization can generate optimal treatment plans that may dramatically improve the treatment response on a patient-specific basis.
Thus, patient-specific mathematical radiation oncology bridging clinical imaging and modeling has the potential to improve outcomes in an otherwise morbid disease.