ESTRO 36 Abstract Book

S788 ESTRO 36 2017 _______________________________________________________________________________________________

Results The PGs DVH and mean doses were not significantly different according to the two CT modality based calculation. On average, the difference for the mean dose was 0.1 % (SD=0.7 %). The SC D2% doses were slightly significantly higher when calculation is based on the standard CT with a mean value of 42.94 Gy (SD=3.03 Gy) compared to 42.52 Gy (SD=2.76 Gy) when calculated on the 3 density classes. Figure 1 represents dose distribution in sagittal plane calculated on 3 density class CT. Figure 2 shows the 3D gamma index map on the sagittal plane (criteria DTA/DD 1mm/1% local dose, dose threshold 10 %): 91.7 % accepted point; gamma mean value 0.6. Most differences between the two dose distributions seems to appear on bone volumes.

The reproducibility of film measurements was on average of 2%. Measurements made on the Cheese phantom surface without mask showed an overestimation of the TPS of 28.6% with fine grid, which is commonly used in clinic. In presence of the mask there was an improvement of the agreement between EBT3 measurements and TPS estimated doses, achieving 0.7%. A considerable number of measurements was performed on 8 patients. The mean absolute value of the difference between measured and TPS-calculated dose and its standard deviation was 11.6% ± 2.8% for all treatments. The average differences were 9.1% for brain and H&N (in these case measurements were performed with mask), and -9.2% for the sarcoma. Hence, there was an overestimation of the TPS without the thermoplastic mask. Conclusion In vivo surface dose measurements with EBT3 are a useful tool for quality assurance in tomotherapy, since the TPS does not give accurate dose values in the first millimeters of skin. Measurements performed both on phantom and in vivo have shown a bolus effect due to the thermoplastic mask, that compensates for the overestimation of the skin dose calculated by the TPS. EP-1490 A 3-class density method to monitor doses to the parotid glands and spinal cord in oropharynx IMRT N. Perichon 1 , S. Couespel 1 , C. Hervé 1 , O. Henry 1 , C. Lafond 1,2,3 , J. Castelli 2,3,4 , A. Largent 2,3 , O. Acosta 2,3 , E. Chajon 4 , R. De Crevoisier 2,3,4 1 Centre Eugène Marquis, Unité de Physique Médicale, Rennes CEDEX, France 2 INSERM, U1099, RENNES, France Purpose or Objective Within a perspective of dose guided/dose monitoring adaptive radiotherapy, a crucial issue is the possibility to calculate the dose distribution on Cone Beam CT scans (CBCTs). The parotid glands (PGs) and the spinal cord (SC) are among the main organs at risk (OAR) exposed to an overdose during the course of IMRT for oropharynx carcinoma. Dose calculation is particularly complex on non-CT images. One clinically applicable option would be to apply three density classes (soft tissue, air, bone) in the CBCTs, corresponding to the density values of the planning CT. The aim of this study was therefore to estimate the accuracy of the dose distribution calculation within PGs and SC by affectation of three density classes. Material and Methods Fifteen patients receiving IMRT for oropharyngeal cancer had a weekly CT scan along their treatment. OAR and target volumes were manually delineated in each CT. A 3- class tissue (soft tissue, air and bone) segmentation was performed in each CT scan using a manual threshold method: over 110 UH for the bone and under -150 UH for the air contained into the external patient contours. Soft tissue was deduced by Boolean operation from air, bone and external patient contour. Mean density values were affected to the 3 classes in the weekly CTs, corresponding to those read on the planning CT for each patient. A plan was first generated on each planning CT scan using a 3 dose levels simultaneously integrated boost protocol (70Gy/63Gy/56Gy in 35 fractions). The beam parameters defined on the planning CT scan were transferred to each weekly CT. Two dose distributions were then calculated in these CT using an adaptive convolution algorithm: either based on the “standard reference” CTscan, or based on the 3 density class CT scan. The doses to the PGs (DVH and mean dose) and the SC (D2%) calculated according to the two CT modalities were compared (Wilcoxon test). Finally, 3D gamma index were also calculated to compare the 3D dose distributions. We report the results for the first 5 patients. 3 Université Rennes 1, LTSI, RENNES, France 4 Centre Eugène Marquis, Radiation Oncology Department, Rennes CEDEX, France

Conclusion This 3-class density method can be used to monitor the fraction dose in the PGs during oropharynx cancer IMRT. Small significant differences are observed for the highest dose received in the spinal cord, likely due to the bone heterogeneity. EP-1491 Verification of FFF VMAT plans with PDIP and GLAaS algorithms by using the new imager of TrueBeamSTx T. Ercan 1 , A. Levent 2 , T. Cagin 3 , S.M. Igdem 1 1 Gayrettepe Florence Nightingale Hospital, Radiation Oncology, Gayrettepe - Istanbul, Turkey 2 Medideal Medical Projects and Solutions Inc., Medical Physics, Istanbul, Turkey