The effects of hypoxia on the theoretical modelling of tumour control probability.

Theoretical modelling of tumour response is increasingly used for the prediction of treatment result and has even been proposed as ranking criteria in some algorithms for treatment planning. Tumour response to radiation is greatly influenced by the details of tumour microenvironment, especially hypoxia, that unfortunately are not always taken into consideration for these simulations. This paper intends to investigate the effects of various assumptions regarding hypoxia distribution in tumours on the predictions of treatment outcome. A previously developed model for simulating theoretically the oxygenation in biologically relevant tissues, including results from oxygen diffusion, consumption and perfusion limitations in tumours, was used to investigate the effects of the different aspects of hypoxia on the predictions of treatment outcome. Thus, both the continuous distribution of values and the temporal variation of hypoxia patterns were taken into consideration and were compared with a 'black-and-white' simplification with a fully hypoxic compartment and a fully oxic one. It was found that the full distribution of oxygenation in the tissue is needed for accurate results. The 'black-and-white' simplification, while showing the same general trends for the predictions of radiation response, could lead to serious over-estimations of the tumour control probability. It was also found that the presence of some hypoxia for every treatment fraction leads to a decrease in the predicted local control, regardless of the change of the hypoxic pattern throughout the duration of the whole treatment. The results thus suggest that the assumptions regarding tumour hypoxia influence very much the predictions of treatment outcome and therefore they have to be very carefully incorporated into the theoretical modelling.