Dokument

Summary:
Non-small cell lung cancer (NSCLC) is the number one cause of cancer-related deaths worldwide. Furthermore, it is predicted that the incidence and mortality will further increase due to smoking, increasing pollution of the environment, and an aging population. For patients unable to undergo surgery corresponding radiotherapy concepts are necessary. However, due to the vicinity of critical organs like the heart, esophagus, trachea, larger blood vessels, and the spinal cord a dose escalation is not always easily achievable using photon-based radiotherapy. Proton therapy (PT) has the potential to deposit a conformal dose in the target volume while better sparing surrounding normal tissue and hence could be beneficial for lung cancer patients. However, there are various challenges connected to proton therapy in general and proton therapy of lung cancer patients in particular. One of these challenges arises from the structure of the lung tissue itself: due to the microscopic density heterogeneity the proton dose distribution is degraded resulting in a broader Bragg peak and a wider distal dose fall-off. This modulation effect can significantly influence the dose distribution in patients resulting in a lower dose deposited in the target volume and higher doses deposited in distal normal tissue and organs at risk (OAR). Since the microscopic structure of the lung tissue is not fully resolved in clinical treatment-planning CT-images, a consideration of the Bragg peak degradation is not possible with current state-of-the-art treatment-planning systems (TPS). In this dissertation, a mathematical model is used to describe the effects due to the Bragg peak degradation. The strength of the degradation is quantified using the material characteristic "modulation power". Microscopic heterogeneous voxelized geometries are used to generate degraded dose distributions with the help of Monte Carlo (MC) simulations. Subsequently, these geometries representing human lung tissue are replaced by clinical voxels with an edge length of 2 mm. Hence, the transition from the microscopic lung tissue as it is present in the patient to coarser clinical CT-structures that cannot resolve the fine lung structure is performed. By modulating the density of each clinical voxel the Bragg peak degradation can be reproduced. Hence, a solution is found to reproduce the lung modulation effects on clinical CT-images. Using this technique, a CT-based phantom study was designed to estimate the effects of the Bragg peak degradation for realistic patient anatomies. Different tumor volumes located at different depths in the lung were investigated. It was shown that, if the lung modulation effects are not accounted for during the treatment-planning process, the dose deposited in the target volume is overestimated and the dose deposited in distal normal tissue is underestimated. This effect increases with an increasing depth of the tumor in lung and a decreasing extent of the tumor in beam direction. At last, the effects were investigated for clinical treatment plans for lung cancer patients. The overestimation of the mean dose in the CTV was 5% at maximum and in the order of 2% on average. The effect on OARs distal to the target volume was negligible for all cases investigated. The investigation of treatment plans confirms that the lung modulation effects are clinically tolerable to a certain degree in the current clinical context considering the various more critical dose uncertainties due to motion and range uncertainties in proton therapy. Nevertheless, PTV concepts were presented that could compensate for the lung modulation effects.

Das Dokument ist im Internet frei zugänglich - Hinweise zu den Nutzungsrechten