ESTRO 2020 Abstract book

S825 ESTRO 2020

radiation of left-sided breast cancer and consequently reduces cumulative mortality risks. [1] Taylor, C. et al. (2017). Journal of Clinical Oncology , 35 (15), 1641. [2] Darby, S. C. et al. (2013). New England Journal of Medicine , 368 (11), 987-998. PO-1453 Aperture shape controller tool impact assessment on the quality of arc therapy treatment plans I. Tyburska 1 , A. Zawadzka 1 , A. Zaleska 2 , D. Szałkowski 1 , P. Kukołowicz 1 1 Maria Sklodowska-Curie Memorial Cancer Center and Institute of Oncology, Medical Physics Department, Warszawa, Poland ; 2 Mazowicki Szpital Onkologiczny, Medical Physics Department, Warszawa, Poland Purpose or Objective In the Eclipse Treatment Planning System (TPS) for Volumetric Modulated Arc Therapy (VMAT) plans, from version 15.6, a new tool Aperture Shape Controler (ASC) is available. Its use reduces the complexity of the leaf movement of the multileaf collimator (MLC) and increases the minimum dimension of the segments. The work aim was to check how the respective levels of ASCs affect the quality of the treatment plan and the results of dosimetric verification. Material and Methods For a group of 10 patients after breast mastectomy (SCAR) and 10 patients with prostate cancer (PROSTATE) VMAT plans with six different ASC levels were prepared (off, very low, low, moderate, high and very high). The calculation algorithm Anisotropic Analytical Algorithm (v15.6) and 6MV beam were used. For all plans, D98% statistics in PTV and the number of monitor units were compared (MU). Additionally, for SCARP plans mean lung dose (MLD) and mean hear dose were analysed. For PROSTATE plans the boost technique with a total dose of 66.15 Gy/ 70.20 Gy delivered in 27 fractions was used. The V60 statistics for rectum and bladder were compared. A dosimetric verification using an Electronic Portal Imaging Device (EPID) was performed for each plan. The measurement and calculation results were analysed using the gamma method (2%/2mm). All above differences were evaluated towards respective plans with ASC option off. Results In SCARP plans, D98% statistics decreased as ASC level was increased, but the differences were clinically insignificant (0,3% lower on average). Dose differences in the MLD and heart were minimal and for 84% of the plans did not exceed 0,2 Gy. In the PROSTATE plans, D98% statistics depended very little on the ASC level used. Similar conclusions were obtained for V60 statistic in rectum and bladder. MU number decreased with the increase of ASC in the PROSTATE plans. No systematic MU changes were noted for SCARP plans. For gamma evaluation, the improvement of the results was observed with an increase in ASC level for both locations. In 72% and 69% of measured fields, respectively for PROSTATE and SCAR, the best gamma results were obtained for a very high level. Conclusion ASC tool can help to achieve better dosimetric verification results and consequently to reduce the inaccuracy in dose delivery without any loss in the quality of the treatment plan. PO-1454 Evaluation of two arc trajectory optimization algorithms for intracranial tumours VMAT planning T. Ventura 1,2,3 , H. Rocha 2,4,5 , B. Da Costa Ferreira 2,6,7 , J. Dias 2,4,5 , M.D.C. Lopes 1,2,3 1 Portuguese Oncology Institute of Coimbra, Medical Physics Department, Coimbra, Portugal ; 2 INESC Coimbra, Inesc Coimbra, Coimbra, Portugal ; 3 University of Aveiro, Physics Department, Aveiro, Portugal ; 4 University of Coimbra, Faculty of Economy, Coimbra, Portugal ;

5 University of Coimbra, Centre for Business and Economics Research, Coimbra, Portugal ; 6 Institute Polytechnic of Porto, Health School Polytechnic of Porto, Porto, Portugal ; 7 University of Aveiro, I3N Physics Department, Aveiro, Portugal Purpose or Objective To assess and compare plan quality and deliverability of two arc trajectory optimization methods for VMAT planning of intracranial tumours. Material and Methods For this study, four cases were selected among a set of meningioma cases previously studied. The arc trajectory optimization followed a two-step approach, where dosimetric and geometric features were combined. In the first step, non-coplanar beam angle optimization was used to search for feasible anchor points of the arc trajectory. In the second step, the anchor points were linearly connected. Two beam angle optimization algorithms ( B and i ) were used to optimize 5 non-coplanar beams. Algorithm B belongs to the continuous search space class and uses a pattern search method to find the best possible beam angle ensemble. Algorithm i is based in a discrete iterative combinatorial approach. For the arc trajectory definition, it was considered that the gantry and the couch could move simultaneously and with different speeds. All fluence map optimizations were performed using an automated multicriterial optimization calculation engine (Erasmus-iCycle). Plan quality was evaluated through SPIDERplan score and the Paddick gradient index. The arc trajectories were qualitatively evaluated. Results Both algorithms produced plans with high dose distribution quality. Moderate quality plan enhancements were presented by algorithm i, due to the higher level of sparing of the optical structures. Nevertheless, the gradient index was lower for algorithm B , which means it presents steeper dose falloffs outside the PTV and a lower dose received by the normal tissues. The two algorithms showed different arc trajectories (Figure 1). Algorithm B presented arc trajectories with higher gantry rotation amplitudes, covering almost all angles, and smooth couch displacements. In algorithm i, for patients #1 and #3, similar shape trajectories to algorithm B were generated. For patient #2 and #4, sharp couch rotations associated to reduced gantry amplitudes created winding arc trajectories that can compromise the treatment delivery and be very uncomfortable for the patient.