Predicting the trajectory of radiotherapy response in patients with head and neck cancer using mathematical modeling of MRI-based habitats

Motivation: Intratumor hypoxia in head and neck cancer influences response to radiotherapy. Goal(s): To characterize intratumoral heterogeneity of hypoxia distribution via predictive, MRI-based mathematical modeling. Approach: MRI-based habitats identified in 20 patients informed a mathematical model of tumor response to radiotherapy. Patients were divided into training (75%) and test (25%) sets to optimize model parameters. The optimized parameters and initial habitat conditions from the test-set were then used to predict response during radiotherapy. Results: The biologically-based mathematical model accurately forecasts anticipated treatment response up to week 4 of radiotherapy for both primary and nodal lesions. Impact: MRI-based modeling of intratumoral heterogeneity in hypoxic, perfusion, and cellular status can predict changes in tumor biology in response due to radiotherapy. Patient-specific predictions based on dynamic changes in imaging parameters could be used to identify optimal radiotherapy strategies.