ESTRO 38 Abstract book

S244 ESTRO 38

The updated ICRU90 data has a small effect on k Q , in the order of 0.2%, but this is well within the quoted uncertainty associated with k Q in TRS-398 of 1%. References: 1.Seltzer, S. et al (2016). ICRU Report 90. 2.Muir, B. R., & Rogers, D. W. O. (2010). Medical physics , 37 (11), 5939-5950. F. Ghareeb 1 , J. Lencart 2 , J. Oliveira 1 , J. A. M. Santos 1,2,3 1 Instituto Português de Oncologia do Porto Francisco Gentil- EPE, Medical Physics- Radiobiology and Radiation Protection Group- IPO Porto Research Center CI-IPOP, Porto, Portugal; 2 Instituto Português de Oncologia do Porto Francisco Gentil- EPE, Medical Physics, Porto, Portugal; 3 University of Porto, Instituto de Ciências Biomédicas Abel Salazar ICBAS, Porto, Portugal Purpose or Objective Patients undergoing 3DCRT radiotherapy using linac head equipped with HD120MLC are subjected to out-of-field transmitted localized extra-focal dose (LEFD) in a very specific region in the direction perpendicular to the MLC leaves motion direction [1]. The goal of this work is to study the effect of collimator rotation angle on the location and the magnitude of the extra-focal dose regions in case of VMAT using Monte Carlo (MC) technique. Material and Methods Several VMAT treatment plans were created using Varian Eclipse TPS for the same hypothetical abdominal lesion. The plans were based on a whole-body CT dataset of a pediatric anthropomorphic phantom (CIRS, ATOM model 705-C, Norfolk, VA). Each plan consisted of two 6 MV full arcs and they were optimized to achieve the same PTV and OAR objectives with different collimator rotations ranged from 0º(360º) to 50º(310º). The treatment plans were then simulated with Monte Carlo (MC) technique using PRIMO software. The dose out of the treatment field was evaluated using 10 spherical structures (1 cm 3 ) located along the central axis at several distances from the isocenter (5 spheres) and along the lateral axis (5 spheres) located 7.5 cm to the right of the isocenter (Figure 1a). For each collimator angle all the mean dose values inside each spherical structure were normalized to the corresponding mean dose value inside an identical spherical structure located at the isocenter. Results The localized extra focal dose distribution is located after 11 cm from the isocenter in the direction perpendicular to the leaves motion direction. At 0º of collimator rotation a cylindrical dose distribution with a diameter equal to the Y-jaws aperture is noticed (as shown in Figure 1b). The dose inside the cylinder is up to 3 folds of the background dose. As the collimator angle increases, the dose distribution start to decrease along the central axis and to increase at the peripheral regions (as shown in Figure 1c for a collimator rotation of 30º) until the extra focal dose disappears from all the evaluated structure at a collimator rotation > 40º. The collimator angle dependence of dose in points 1, 9 and 10 are shown in Figure 2. PV-0477 A Monte Carlo study of collimator angle dependence of extra focal dose during VMAT

1 Australian Radiation Protection and Nuclear Safety Laboratory, Primary Standards Dosimetry Laboratory, Melbourne, Australia ; 2 Australian Radiation Protection and Nuclear Safety Laboratory, Australian Clinical Dosimetry Service, Melbourne, Australia Purpose or Objective The publication of the updated key data for ionizing- radiation dosimetry in ICRU90 1 includes changes to the mean excitation energy (I-value) of graphite and water, and the renormalised photoelectric cross sections. Both of these changes will affect the Monte Carlo calculated beam quality correction factor k Q , which is being considered with the update to TRS-398. Here, we have modelled the change that the ICRU90 data has on k Q for several common All modelling was done in the EGSnrc c++ user code egs_chamber. Various ionisation chambers commonly calibrated by the Primary Standards Dosimetry Laboratory (PSDL) at ARPANSA were modelled (Including the NE 2571, PTW 30013, IBA FC65-5, ICA CC-13, and the Exradin A12) according to detector specifications from obtained from the manufacturers, including accurate modelling of the stem. Two source inputs have been used; a commissioned full linac head model (6, 10, 18 MV) and tabulated photon source spectra from both published results (4-25 MV) and results from the ARPANSA modelled linac. The dose to the air cavity in the chamber and the dose to water were scored at 5 cm depth in 60 Co beams and at 10 cm depth in MV photon beams, as per the IAEA TRS-398 protocol, and the ratio used to calculate k Q . The simulations were repeated with the previous (ICRU37) and updated (ICRU90) key datasets. Results All simulations were run until a combined uncertainty from all four measurements was <0.1%. All simulations compare well to TRS-398, agreeing within similar ranges to other published studies. When using the ICRU37 data, results agreed with published values 2 to within 0.4%. When comparing the same inputs with the updated ICRU90 data, the values for k Q were on average 0.2% higher/lower, indicating that the impact of the updated data from ICRU90 has a small effect on k Q , which agrees with other published works. Figure 1 shows the calculated values for the NE-2571 ionisation chamber, including the difference. The difference of 0.2% is small, and within the 1% uncertainty quoted in TRS-398. ionisaiton chambers. Material and Methods

Figure 1: Calculated k Q values for the NE-2571 ionisation chamber using ICRU37 and ICRU90 data, with the differences (positive shows ICRU 90 is higher than ICRU37) between the two datasets. Conclusion Several ionisation chambers have been successfully modelled, allowing for the beam quality correction k Q to be calculated. We have shown good agreement with published protocols and published Monte Carlo studies.