Investigating Intra-Tumoral Heterogeneity Driven By Nutrient Dynamics and Radiotherapy Using an In-Silico Avascular Tumor Growth Model
Abstract
Purpose
To develop an in-silico tumor model that incorporates nutrient-driven growth and radiotherapy response to generate spatio-temporal proliferating (P), quiescent (Q), and necrotic (N) cell maps for radiomics-based heterogeneity analysis.
Methods
We implemented a framework of early avascular solid tumor growth under dynamic nutrient conditions based on the Chaplain–Sherratt partial-differential-equation (PDE) model for P/Q/N evolution. Radiotherapy was modeled using linear–quadratic (LQ) survival model applied to proliferating cells, with a delayed kill effect representing post-radiation mitotic death/repair kinetics. A probabilistic dose threshold was used to trigger local extinction of proliferating cells at high dose. Immune-mediated debris removal via necrotic clearance was modeled as a radial-dependent washout of necrotic cells. We calculated statistical energy of P-cell density curves at different radiation doses as an aggregate feature. To generate spatial heterogeneity, a Monte-Carlo scheme was used to sample radial shells and assign cell types based on solutions to the PDE model. Cell coordinates were pixelized into density maps weighted by their entropy to characterize total-cell heterogeneity. As an experiment, we ran the model for different nutrient uptake factors α and radiation doses. Radiomic texture features were extracted from the time-dependent cells maps to quantify biological-driven heterogeneity.
Results
Increasing α lowered nutrient level, narrowed the proliferative rim faster, and expanded Q/N compartments earlier, leading to lower texture heterogeneity and earlier necrotic saturation. Simulated heterogeneity peaked during tumor expansion followed by stabilization. Across modeled radiation doses, the simulation captured P-cell spatial distribution changes in energy over time. The post-irradiation energy decreased, followed by a recovery towards homogeneity of the proliferating cell maps. At later times, this effect decayed as necrotic cells became more dominant.
Conclusion
Our model mechanistically links nutrient dynamics and radiation-induced cell death to intra-tumoral heterogeneity via radiomic features. It provides an in-silico framework for computational radiation-biology experiments.