Variations in energy spectra and water-to-material stopping-power ratios in three-dimensional conformal and intensity-modulated photon fields.

Because of complex dose distributions and dose gradients that are created in three-dimensional conformal radiotherapy (3D-CRT) and intensity-modulated radiation therapy (IMRT), photon- and electron-energy spectra might change significantly with spatial locations and doses. This study examined variations in photon- and electron-energy spectra in 3D-CRT and IMRT photon fields. The effects of spectral variations on water-to-material stopping-power ratios used in Monte Carlo treatment planning systems and the responses of energy-dependent dosimeters, such as thermoluminescent dosimeters (TLDs) and radiographic films were further studied. The EGSnrc Monte Carlo code was used to simulate megavoltage 3D-CRT and IMRT photon fields. The photon- and electron-energy spectra were calculated in 3D water phantoms and anthropomorphic phantoms based on the fluence scored in voxel grids. We then obtained the water-to-material stopping-power ratios in the local voxels using the Spencer-Attix cavity theory. Changes in the responses of films and TLDs were estimated based on the calculated local energy spectra and published data on the dosimeter energy dependency. Results showed that the photon-energy spectra strongly depended on spatial positions and doses in both the 3D-CRT and IMRT fields. The relative fraction of low-energy photons (< 100 keV) increased inversely with the photon dose in low-dose regions of the fields. A similar but smaller effect was observed for electrons in the phantoms. The maximum variation of the water-to-material stopping-power ratio over the range of calculated dose for both 3D-CRT and IMRT was negligible (< 1.0%) for ICRU tissue, cortical bone, and soft bone and less than 3.6% for dry air and lung. Because of spectral softening at low doses, radiographic films in the phantoms could over-respond to dose by more than 30%, whereas the over-response of TLDs was less than 10%. Thus, spatial variations of the photon- and electron-energy spectra should be considered as important factors in 3D-CRT and IMRT dosimetry.