How accurate is a CT-based dose calculation on a pencil beam TPS for a patient with a metallic prosthesis?

The accuracy of a CT-based dose calculation on a treatment planning system (TPS) for a radiotherapy patient with a metallic prosthesis has not previously been reported. In this study, the accuracy of the CT-based inhomogeneity correction on a pencil beam TPS (Helax TMS) was determined in a phantom containing a metallic prosthesis. A steel prosthesis phantom and a titanium prosthesis phantom were investigated. The phantoms were CT-scanned and dose plans produced on the TPS, using the CT images to provide density information for the inhomogeneity corrections. Verification measurements were performed on a linear accelerator for 6 and 15 MV x-rays. Measured dose profiles at three different depths were compared to the calculations of the TPS. For the titanium prosthesis and for 6 MV x-rays, the TPS overestimated the beam attenuation by approximately 20% at 15 and 20 cm depths in the phantom. This is due to a limitation in the density allocation of this TPS: any Hounsfield number (HN) above a certain threshold is allocated the density of steel. For the steel prosthesis, the TPS performed the correct mapping of HN to mass density. The dose calculation was within 6% for 6 MV x-rays at 15 and 20 cm depths. However. the accuracy of dose calculation varied with beam energy and depth, with large errors in the region close to the prosthesis. The TPS overestimated the dose by 11% for 6 MV and 15% for 15 MV x-rays at 11 cm depth. 2.5 cm beyond the steel prosthesis. These results highlight the limitations in the density allocation of this TPS and demonstrate shortcomings in the pencil beam dose calculation.