The quantitative study of tissue dynamics has advanced through reaction–diffusion frameworks that capture the interplay between proliferation, motility, and environmental factors. Classical models describe tumour invasion as traveling waves but neglect cellular heterogeneity and phenotypic variability. In cancer biology, therapeutic interventions rarely eradicate all malignant cells: depending on their phase in the cell cycle and local oxygen level, some cells are lethally damaged while others survive in quiescent or sublethally damaged states. These effects demand a multiscale description linking cell-cycle kinetics, oxygen dynamics, and therapy scheduling. We propose a stochastic, multi-scale, age-structured frame work to analyse the dynamics of heterogeneous cell populations under therapy. Each cell is characterised by its time since division and its division rate, which depends on oxygen concentration and therapy dose [1]. In our model, therapy introduces a subpopulation that is not killed but diverted to a damaged phenotype, altering tumour evolution. From the stochastic formulation, a deterministic mean-field limit is derived, to describe the evoluton of these two populations coupled to the oxygen concetration. We then design a hybrid multiscale model that combines the accuracy of the stochastic formulation—applied at the tumour front, where fluctuations dominate—with the computational efficiency of the deterministic mean-field description in the bulk [2]. Numerical simulations demonstrate that the coupled system supports traveling-wave solutions whose velocity and internal structure depend on therapy intensity, frequency, and oxygen regulation. We systematically compare the wave speeds obtained from our model for different doses of therapy and study if the speed primarily depends on the leading-edge proliferation rate, meanwhile the wave morphology and the relative contribution of damaged cells are strongly shaped by therapy scheduling. Longer therapy intervals promote re-entry into therapysensitive phases of the cell cycle, leading to more synchronized, efficient responses and sharper wavefronts, whereas shorter intervals maintain a persistent damaged population that broadens the tumour interface. REFERENCES [1] de la Cruz, et al. (2016). Journal of Theoretical Biology. 407. 10.1016/j.jtbi.2016.07.028 [2] de la Cruz, et al. (2017). J Comput Phys. 2017 Dec 1;350:974-991. doi: 10.1016/j.jcp.2017.09.019.