Comparative treatment planning in external radiotherapy of malignant tumours : Potential gains using protons

University dissertation from Uppsala : Acta Universitatis Upsaliensis

Abstract: It is well known that protons have physical characteristics superior to conventional radiation qualities in external beam radiotherapy. The primary aim of this work is to determine if the physical advantages also may lead to clinical advantages. A second aim is to evaluate the treatment planning algorithms incorporated into a 3D treatment planning system, used in the prediction, Helax-TMSTM. The potential benefits of using protons instead of conventional radiation qualities is evaluated by comparing treatment plans. Dose-response models of tumour control probability (TCP) and normal tissue complication probability (NTCP) were used to predict the treatment outcome for the different plans. Uncertainties in the results were studied in sensitivity analyses.Comparative studies were performed in four different tumour types, locally advanced rectal cancer, a paraspinal tumour, oesophageal cancer and left-sided node-positive breast cancer. Advantages with protons were seen in all cases, but the advantages were of different size.Proton therapy in patients with oesophageal carcinoma increases the TCP from 6 to 49% if a risk of 1% in the spinal cord and a total NTCP for the two lungs equivalent to a one-sided pulmectomy is allowed. This gain is relatively insensitive to variations, within reasonable limits, in the dose-response parameters.Proton therapy reduces the risk for cardiac mortality and radiation pneumonitis to almost zero when treating node-positive left-sided breast cancer after breast-conserving surgery.The evaluation of the treatment planning algorithms shows that pencil kernel algorithms are required in order to accurately calculate dose distributions in heterogeneous patient volumes. The scattering phenomenon in the vicinity of interfaces between different materials is modelled. The magnitude of the effect is underestimated (less than 1%) due to the inherent approximations.

  This dissertation MIGHT be available in PDF-format. Check this page to see if it is available for download.