Bridging cell-scale simulations and radiologic images to explain short-time intratumoral oxygen fluctuations

Radiologic images provide a longitudinal way to monitor tumor responses, but operate on a macroscopic scale and are not able to capture microscopic scale phenomena. We provide a link between the average data recorded in radiological image voxels and the tissue architecture that fills these voxels. Our in silico model includes individual tumor and stromal cells, tumor vasculature, and tumor metabolic landscape. This architecture was based on tissue characteristics acquired from electron paramagnetic resonance (EPR) images. We used this model to optimize vascular influx and cellular uptake schedules to reproduce oxygen fluctuations recorded experimentally. By comparing simulation results within the schedules, we showed that sole alterations in vascular influx were able to reproduce experimental data well. On the other hand, in order to fit experimental data with metabolic changes in tumor cells, the cells would need to increase their oxygen absorption by 50-fold over a period of 3 minutes, which may not be biologically feasible. Additionally, we developed a procedure to identify plausible tissue morphologies for a given temporal series of average data from radiology images. In the future our approach can be used to simulate hypoxia-sensitive anti-cancer treatments on a cell-scale based on clinically-collected images.