In this thesis a number of aspects of conformal radiotherapy of urogenital tumors have been studied to improve their planning and delivery process. The aims of the study are threefold. First, three-dimensional (3D) data on target delineation, organ motion, and patient setup are analyzed to incorporate into the planning of the treatment of bladder and prostate cancer. The data on bladder delineation uncertainties are obtained from the 3D variation in a multiobserver study, organ motion is quantified by analyzing bladder volumes at different moments during treatment, while the data on setup uncertainties of bladder patients are extracted from portal imaging. The results showed that organ motion is the most important uncertainty during radiotherapy of bladder cancer. As part of a quality assurance procedure of a prostate trial, delineation of the various volumes and treatment planning were compared in the participating institutions. Differences in delineation of the prostate between different observers are small. However, differences in delineation of the rectum greatly influenced the calculated risk of rectal complications. Therefore, applying a stricter definition of delineation of organs-at-risk is needed to retrieve more reliable trial results. Next, a mathematical model for the rectal wall has been developed, incorporating stretching due to variable rectal filling and neighboring structures. The application of the model yields accurate dose distributions in the rectal wall without delineating the inner surface of the rectum. This approach reduces both the workload and variations due to inaccurate delineation of the rectal wall. A second tool has been developed to take into account the additional dose resulting from acquiring electronic portal images used for patient setup verification. In this formalism beam weights and wedge angles of the treatment fields are adjusted in such a way that the dose and dose homogeneity in the PTV are preserved. Finally, the impact of an extensive in vivo dosimetry program on the delivery of conformal treatments of bladder and prostate cancer patients has been assessed. Analysis of measurements performed on more than 500 patients revealed that during the first years a number of relatively large, up to 7%, systematic errors were traced by in vivo dosimetry. In the later period it was possible to trace minor imperfections, even smaller than 1%, using statistical analysis of the results.
Read full abstract