ESTRO 38 Abstract book

S938 ESTRO 38

the GLAaS principles to a new delivery system, Varian Halcyon, a 6FFF linac mounted on an O-ring gantry, equipped with a dual-layer stacked MLC with an effective shaping capability of 5mm at isocenter. Material and Methods Dosimetric data from 3 institutes were collected at dmax for primary and transmitted radiation to be correlated with Halcyon digital megavoltage imager (DMI) signal: pixel-by-pixel response changes were modelled on time basis, differentiating on segment sizes and beam quality, i.e. primary or below MLC radiation. The satisfactory GLAaS algorithm configuration allowed pre-treatment QA verifications in all the centers. For 3 different sites (prostate, head-and-heck and breast), each center pooled 5 patients from its own database to be acquired with DMI and other QA devices locally available (ArcCHECK, MapCHECK, Octavius, ion chamber point dose). For all centers, DMI images were analysed versus TPS dose water matrices with GLAaS algorithm and predicted fluences with Varian Portal Dosimetry (VPD). Among the previous 45 plans, a sub-set of 6 was shared among all to investigate the single machine delivery behavior respect to calculation based on the same pre- configured and not customizable beam modelling (Varian AAA dose algorithm). Results Adapting the GLAaS principles to Halcyon technology, it was possible to conveniently model the DMI dose response. Although in the CIAO area the global result is mainly affected by primary radiation modelling (ima1) due to the low MLC transmission (0.5% per layer), a dedicate model for transmission guarantees a more robust and accurate evaluation of the measured dose.

question the usefulness of changing the volume of interest (PTV or GTV) for dose prescription as suggested by Lacornerie et al. (2004). The purpose of our study is to define a method that ensures robustness when prescribing dose using a MonteCarlo [MC] algorithm. Material and Methods Treatment plans of 30 patients were recalculated from RT algorithm to MC algorithm and three different indicators to prescribe dose were studied: D 95%PTV (ESTRO ACROP), D 50%GTV (Lacornerie et al.) and D 98%GTV (D near-minGTV ). On these indicators, overall doses prescription deviations were analyzed in function of tumor volume. Dose ratios to the indicator between MC and RT were calculated as it reflects the dose degradation. In a second part, using the same data, a prescription of 3x15Gy to D 95%PTV , as recommended by ESTRO ACROP, with MC algorithm is calculated and the doses to the three indicators are compared to the ones determined in the first part. Results When prescribing D 95%PTV =54Gy with a type A algorithm, doses to D 95%PTV , D 50%GTV and D 98%GTV with a MC algorithms are respectively, for GTV≤10cm³ (21 patients) 41.3± 3.2Gy, 53.8± 3.5 Gy and 49.1± 3.0 Gy and when GTV>10 cm³ (9 patients) 43.7± 3.5Gy, 56.2± 1.5Gy and 50.2± 1.7 Gy. Mean dose ratio between MC and RT algorithms are respectively 77.4± 6.1%, 86.8± 4.2% and 84.6± 4.1%. Prescribing doses of 3x15Gy to D 95% PTV with a MC algorithm results, for GTV≤10cm³, in doses to D 95%PTV , D 50%GTV and D 98%GTV of respectively 45Gy, 58.7± 2.6Gy and 53.6± 2.0Gy and for GTV>10 cm³ to 45Gy, 58.2± 4.0 Gy and 51.9 ± 3.1 Gy. Thus, following ESTRO ACROP prescription results in a dose increase to each indicator, for GTV≤10cm³ of respectively 8.2, 8.3 and 8.4 % and for GTV>10cm³ of 2.9, 3.4 and 3.3 %. Conclusion Despite ESTRO ACROP recommendation, it has been shown in this study that D 95% PTV is not a robust indicator to prescribe dose as, for a unique dose prescription, large doses fluctuations received by the tumors are observed depending on their volume: doses of 3x18Gy (GTV≤10 cm³) and of 3x17Gy (10cm³

Overall, pre-treament QA was fully satisfactory, with differences in terms of specific results depending on the devices; in case of DMI images, the gamma agreement index (GAI) was greater in comparison versus predicted /VPD than versus calculated /GLAaS doses (e.g GAI for 3%- 3mm: global VPD>99% vs GLAaS>97%, local VPD>98% vs GLAaS>92%).

Conclusion The extension of GLAaS to Halcyon offers the opportunity to easily set-up flexible and reliable verification, allowing a straight but independent comparison between EPID measurements and TPS water dose maps. EP-1741 Evaluation of the application of NIPAM gel dosimeter to quantify dynamic dose effects J.C. Sun 1,2 , Y.W. Tsang 1 , B.T. Hsieh 2 , K.Y. Cheng 2 1 Department of Radiation Oncology, Chia-Yi Christian hospital, Chiayi City, Taiwan ; 2 Department of Medical