Abstract ID: 185 Monte Carlo simulation of “Out-of-Field” dose distributions in a pediatric phantom in external radiotherapy – Comparison with Gafchromic film measurements

Abstract

Introduction Patients undergoing radiotherapy are exposed to out-of-field scattered and leakage radiation, which may induce secondary cancer in long-term survivors. The induction of secondary cancer in pediatric patients is an increased concern, because they have increasingly higher life expectancy, and tend to receive higher secondary organ doses than adults due to geometrical factors. Radiation-induced thyroid and breast cancers have been observed in pediatrics after receiving doses as low as 100 mGy [1] . While the dosimetric accuracy of treatment planning systems (TPS) is known to decrease in the out-of-field region, Monte Carlo (MC) is nowadays the most reliable method in dose calculation. A PENELOPE based MC code named PRIMO was tested for out-of-field dose estimation in pediatric irradiation. Material and methods a 5-year pediatric anthropomorphic phantom was irradiated by Varian Triology linac, using three 6MV static fields beams for a brain tumor. The treatment plan was calculated in Varian Eclipse TPS, using Analytic Anisotropic Algorithm (AAA). The 2D dose distributions at thyroid and lung levels were measured using Gafcromic EBT3 films. The plan was simulated in PRIMO after adjusting the primary beam parameters based on reference data in water phantom. PRIMO output data was reshaped, manipulated and re-written as DICOM files for comparisons with both measured and TPS calculated dose using 2D Gamma Function (3%, 3 mm) on DoseLab Pro software. Results The maximum measured doses per 1 Gy at isocenter (0 cm) were 1.5 cGy and 8 mGy at thyroid (10.25 cm) and lung levels (15.25 cm) respectively. Comparing simulated and measured dose distribution one obtained 82.6% and 88.8% of gamma points 1 at thyroid and lung levels respectively. The difference between measured and TPS calculated dose distributions could easily be accessed visually. Conclusion For secondary induced cancer risk estimation, the present TPS, even using a reliable algorithm as AAA, is not generally satisfactory for out-of-field dosimetry. Alternative tools, like MC simulations are necessary if correct dose estimation must be obtained for risk models application.