Licenciate Thesis defense: The virtual tumour - in silico modelling of tumour vasculature, oxygenation and treatment outcome

Poor tumour oxygenation, namely hypoxia, is one of the major challenges that has been recognised in radiotherapy, yet it is not being accounted for in standard treatments. Hypoxia, resulting from a heterogeneous distribution of vessels (chronic hypoxia) or of a loss in vascular perfusion (acute hypoxia), affects all kinds of solid tumours to different extents. Although over-sustained angiogenesis with vascular remodelling is one of the key hallmarks of cancer, the resulting tumour vasculature is often frail and lacking a clear hierarchical structure, hence incapable of maintaining the same nutrients and oxygen supply standards of healthy vascular networks.

It is ascertained that hypoxic cells require an up to three times higher radiation dose than normoxic tissues to achieve the same biological effect. Tumour hypoxia correlates to worse disease prognoses when compared to normoxic tumours. However, many of its biological aspects remain only partially understood. Although from a clinical perspective most of the countermeasures that have been devised to oxygenate or kill hypoxic cells can be evaluated in terms of short- and long-term effects, a clear and pristine understanding of the mechanisms involved in the curative process of hypoxic tumours has not been provided.

From this perspective, in silico modelling of the tumour key radiobiological features could instead represent a new frontier: unprecedented computational power and numerical optimisation routines permit to expand virtually the set of possible microenvironmental situations, with simulations of real treatments and concurrent intercomparison of hypothetical scenarios. The fact that the real vascular anatomy of a deep-seated tumour is not fully accessible – and hence not precisely modellable – could be compensated by a large casuistry of heterogeneous oxygenation patterns provided by the model, with inherent best- and worst- case studies. At the same time, in silico modelling would not replace in vivo functional imaging, but would rather act in synergy with that as an additional layer of study: based on the underlying macroscopic information that for instance positron emission tomography (PET) or magnetic resonance imaging (MRI) could offer, the microscopic radiobiological nature of the tumour could be simulated.

This thesis consists of papers I-II and an introductory overview of the topics, which provide the background needed for their basic understanding. Starting with an account of tumour hypoxia, its radiobiology and the solutions that have been explored so far to counteract its negative effects are also discussed. Paper I is concerned with the presentation of the novel three-dimensional radiobiological model developed, which is the core of a comprehensive project developed during the PhD work; therefore, it will be introduced in the framework of previously designed models that dealt within this subject in the context of radiotherapy. Since this is probably the first model of its kind designed to be implemented into a treatment planning system (TPS), the subsequent natural steps will consist of the integration of the model in a research version of a TPS and study of the efficacy of various treatment scenarios in terms of underlying tumour oxygenation and treatment choices regarding beam quality, fractionation, and total dose.