Monte Carlo and film dosimetry with applications in breast cancer radiotherapy

Abstract

The purpose of the present work has been to exploit dosimetric methods and procedures for accurate quantification of dose distributions in general and for breast cancer treatments in particular. Out-of-field dose was in focus, against a background of concern about radiation-induced second cancers. In addition, surface dosimetry was investigated, which is of importance for superficial target volumes like the breast. Different radiotherapy techniques relevant for breast cancer treatment were characterized and compared, with emphasis on the dose to the contralateral breast and the skin region. Furthermore, the dosimetric consequences of removing the flattening filter from a conventional linear accelerator were evaluated. Radiochromic film dosimetry and Monte Carlo simulations were judged to be appropriate dosimetric methods for the problems investigated. The overall uncertainty of the dosimetric film procedure was found to be about 4 % (2-sigma level) at 2 Gy. While the precision of Monte Carlo simulations is statistical in nature, and therefore only limited by the simulation time (or cost), its accuracy depends on the model itself. A thorough benchmarking procedure for the Monte Carlo model was therefore performed. An anthropomorphic female thorax phantom was used for the dosimetric characterization of different breast cancer irradiation techniques. The dosimetry revealed a highly inhomogeneous dose distribution in the contralateral breast during tangential breast irradiation. Mean dose to the CLB was found to be 1.0 % of the target dose for tangential irradiation, i.e. 0.5 Gy for a standard treatment of 50 Gy. However, the CLB dose depends on the use of wedges and on collimator orientation. Superficial (0.15 mm depth) doses in the treated breast depend on the treatment technique and vary between regions, but are mainly within 40 % - 65 % of the target dose. Some skin dose (0-5 mm depth) can be avoided by utilizing non-tangential fields. The doses to superficial parts of the clinical target volume were mainly within 90-95 % of the target dose for all techniques investigated. The energy of the incident electron beam in a flattening-filter-free (FFF) linear accelerator needs to be increased in order to retain the build-up and attenuation characteristics of a conventional photon beam. For the 6 MV photon beam in this work, this was achieved by changing the electron energy from 6.45 MeV to 8.0 MeV. FFF beams represent a general improvement considering reduction of out-of-field doses. By utilizing photon fields from an FFF linear accelerator in tangential breast irradiation, the mean dose to the CLB was reduced by 24 – 27 % compared to that from the conventional accelerator.