A computational model for phenotype-structured tumour dynamics with reduced microvascular oxygen and event-based radiotherapy

79/2026

A computational model for phenotype-structured tumour dynamics with reduced microvascular oxygen and event-based radiotherapy

Wednesday 21st October 2026

Vitullo, P.; Macaluso, C.; Gazzoni, S.; Cassanmagnago, A.; Dimola, N.; Possenti, L.; Zunino, P.

Tumour response to radiotherapy is strongly influenced by spatially heterogeneous oxygenation and by the ability of tumour-cell populations to adapt their phenotypic composition to the local microenvironment. We present a computational framework that couples phenotype-structured tumour dynamics with microvascular oxygen transport and fractionated radiotherapy. The tumour population evolves in physical and phenotype space, while tissue oxygenation is obtained from a reduced-order approximation of a mixed-dimensional vascular–tissue transport model. Radiotherapy is represented through event-aligned survival updates, allowing treatment events to remain distinct from the smooth inter-fraction tumour evolution. The numerical formulation combines finite elements in physical space, a conservative discretization of phenotype space and implicit time integration within a sequential tumour–oxygen coupling strategy. Its numerical behaviour is examined through physical-space resolution sensitivity, temporal and phenotype refinement, and comparison of the oxygen surrogate with full-order solutions. The refinement studies show approximately first-order behaviour for the assessed temporal tumour-burden trajectories and approximately second-order behaviour for phenotype-class-resolved trajectory quantities. The framework is then used to investigate fractionated radiotherapy across vascular environments differing in network pruning and oxygen supply. The simulations show that the microvascular oxygen environment substantially alters the accumulated treatment response and the onset of numerical extinction across the sampled dose courses. Across the explored vascular map, inlet oxygen provides the clearest organization of hypoxic exposure and phenotype composition, while vascular pruning introduces an additional, non-uniform modulation. Lower oxygen supply is associated with greater enrichment of stem-like phenotypes and a smaller radiotherapy-associated reduction within the model. These results provide a numerically assessed framework for studying how microvascular oxygen heterogeneity and phenotype adaptation jointly shape modelled tumour response to fractionated radiotherapy.