ESTRO 37 Abstract book

S1054

ESTRO 37

and small bowel (p>0.05). In addition, decrease of NTID, volume receiving dose >5Gy and >10Gy were found in 6X- 6X as compared 4X-4X and 4X-6X (p>0.05). However, 6X- 6X showed lesser MU of 8.92% compared to 4X-4X and 15.11% to 4X-6X (p<0.05). Whereas, delivery time was lesser in 6X-6X as compared to 4X-4X and 6X-6X (p<0.05). Conclusion For CSI VMAT plans, 6MV found to be more suitable photon energy as compare to 4MV and 4MV&6MV in terms of better plan quality, higher dose rate, shorter treatment time and lesser MU. EP-1937 Dosimetric impact of stabilizing spinal implants in proton therapy: titanium vs. carbon systems B.S. Müller 1 , Y.M. Ryang 2 , F. Ringel 3 , M. Oechsner 1 , H. Specht 1 , B. Meyer 2 , S.E. Combs 1,4 , J.J. Wilkens 1 1 Klinikum rechts der Isar- Technical University of Munich, Department of Radiation Oncology, Munich, Germany 2 Klinikum rechts der Isar- Technical University of Munich, Department of Neurosurgery, Munich, Germany 3 University of Mainz, Department of Neurosurgery, Mainz, Germany 4 Institute of Innovative Radiotherapy, Helmholtz Zentrum München, Neuherberg, Germany Purpose or Objective Over the last years carbon pedicle screw systems were introduced for spinal stabilization, which promise to reduce dosimetric uncertainties in radiotherapy (RT) compared to metallic implants. We investigated the impact of uncertainties in assigned relative stopping powers (SP) of spinal stabilizing implants on dosimetric quality for intensity modulated proton therapy (IMPT). In particular we compare titanium to carbon systems. Material and Methods The retrospective planning study is based on computed tomography (CT) images of ten patients, who had undergone surgery for spinal stabilization prior to RT treatment, and were previously treated with photon RT at our institute. Five patients had titanium (TI), five carbon (CA) systems. Carbon pedicle systems (Icotec, Altstätten, Switzerland) comprised radiolucent non- metallic screws, titanium coated regions in the pedicle area and titanium rods. To simulate uncertainties in the assigned Hounsfield units (HU) of titanium, corresponding regions were delineated in all CTs and assigned with varying SPs. For each patient, one IMPT plan of three fields was optimized onto the planning CT with a SP of 3.2 (prescribed dose to planning target volume (PTV): 30 Gy à 3 Gy). These plans were recalculated onto the identical CT with reduced SP (SP=3.0) and increased SP (SP=3.4). Dose distributions were analyzed by PTV coverage, homogeneity and several dose volume histogram (DVH) values. Additionally, CT artifacts were contoured and resulting volumes evaluated. RT planning was performed in matRad (www.matrad.org). Results Volumes of CT artifacts and metallic components were smaller in CTs of CA compared to TI systems, with mean volumes of V CA (artifacts)=5.9±1.2 cm 3 vs. V TI (artifacts)=18.7±7.4 cm 3 and V CA (metal)=10.1±0.2 cm 3 vs. V TI (metal)=24.4±5.4 cm 3 . The initial plan qualities of both implant types were comparable, with a slightly superior PTV coverage and homogeneity in CA plans. The impact of varying SPs on DVH criteria was larger for TI. For a SP increase of 0.2, i.e. SP=3.4, the average minimum dose received by 1 cm 3 decreased by ΔD min (PTV)=5.9±2.9% for TI and ΔD min (PTV)=1.5±0.5% for CA (see figure, with box plots showing the median minimum dose, its interquartile range (box) and most extreme values). Recalculated doses of both implant types resulted in more heterogeneous target coverage, with a more pronounced deviation from the original dose

for TI. The mean standard deviations inside the PTV increased e.g. by Δσ TI (PTV)=7.8±4.2% and Δσ CA (PTV)=2.4±0.9% for SP=3.0.

Conclusion Dosimetric uncertainties given by inaccurate HU assignment and its conversion to SP are reduced by utilizing carbon based implants instead of pure titanium, which decreases the risk of tumor underdosage. However, due to the remaining metallic component, some uncertainties in dose distributions still remain. Generally, decreasing the fraction of metal does not only improve dosimetric accuracy, but is also considered to be beneficial for RT planning through more accurate contouring. EP-1938 Comparative study of Auto plan and manual plan for nasopharyngeal carcinoma IMRT radiotherapy X. Xin 1 1 Sichuan Cancer Hospital, Radiotherapy center, Chengdu, China Purpose or Objective Auto planning might reduces the manual time required for optimization and could also potentially improve the overall plan quality. This study is aim to demonstrate the statistical comparison of automatic (AU) and manually (MA) generated nasopharyngeal carcinoma intensity modulated radiation therapy (IMRT) plans. Material and Methods The study included in total, 97 nasopharyngeal carcinoma patients, admitted to our hospital during 2015–2016. The patients underwent IMRT treatments. The clinical and delivered plans were performed with Eclipse (Version 11.0) using manual optimization. The same plans were optimized successively in Pinnacle TM3 (version 9.20) treatment planning system using auto plan software package module (provided by MD Anderson Cancer Center) and then exported to MATLAB software. D95, D98 were calculated for the targets and Dmax and Dmean for