Abstract Book

S1009

ESTRO 37

case scenarios fulfilled the dose requirement for the GTV leading to an insufficient robustness. Conversely, more than 90% of the D 95% values to the CTV were robust against uncertainties, with S3 being most robust (97%) .

evaluation. DIs included: PTV1 D95%, cord D2%, brainstem D2%, left_ and right_parotid D2% and D mean , esophagus D50%, larynx D mean and mandible D2%. Absolute percent differences between DIs from subsequent CTs were tallied. Furthermore, statistical tests on the absolute dose differences of the DIs between CT1 and CT2 were performed for each case and optimization scheme. In addition, the clinical plans, used for actual treatment, were also analyzed for DI differences between CT1 and CT2. Results Regardless of the optimization modality the changes were small with few exceptions. The average absolute percent change between CT1 and CT2 for each DI is listed in the table, where the values in bold represent the higher value between DVH- and Energy-based optimized plans. Change for 6 out of the 10 DIs was larger for DVH plans compared to Energy plans, but there were no statistically significant differences for any DI when comparing the absolute doses between CT1 and CT2. This result indicates that the DI differences are drawn from a distribution with mean zero and some symmetric spread around that mean. The high average absolute value percent change for the esophagus D50 was due to an outlier where the absolute doses were very low. For this case CT2 was difficult to register to CT1 due to neck tilt and noticeable area shrinkage, leading to higher dose to the volumes of interest. The changes observed in the clinical plans were comparable to the tested plans, with slightly smaller changes, with the exception of mean dose to the left parotid, which was largely due again to one outlier.

Conclusion Disregarding the inter- and intra-fractional organ motion, dose escalation is possible using robust MFO-IMPT plans with three beams, of which at least one non-conformal. EP-1868 Study on the dependence of Energy-based inverse optimization on the changing anatomy for HNSCC. I. Mihaylov 1 , M. Couto 1 , N. Elsayyad 1 , C. Takita 1 , M. Samuels 1 1 University of Miami, Radiation Oncology, Miami, USA Purpose or Objective To compare the stability of Energy-based inverse optimization on changing anatomy in head-and-neck squamous cell carcinoma (HNSCC). Material and Methods Twelve HN patients were CT scanned before RT (CT1) and mid-treatment (CT2) during the treatment course. All targets (PTVs) and organs at risk (OARs) were deformably propagated from CT1 to CT2. For each case Energy- and DVH-based plans were generated on the planning CT1, using nine equispaced step-and-shoot IMRT beams. DVH- based plans were used as a benchmark since current clinical practice is based predominantly on this type of optimization. During optimization, doses to the organs at risk were iteratively lowered until the standard deviation across the PTV was ~4.0%. All plans were normalized to 95% of the prescription dose to the primary PTV: 70 Gy in 33 fractions. The plans were then transferred to the subsequent CTs without any modification/re- optimization, simulating the actual treatment scenario. Dose indices (DIs) were obtained for plan stability

Conclusion Even though Energy-based depends explicitly (through the optimization cost function) on the density distribution within volumes, the changes that occur during radiation treatment are small and comparable to that of DVH-based plans for HNSCCC. With either optimization large differences may occur due to noticeable weight loss and/or positioning errors. EP-1869 Dosimetric plan comparison of VMAT vs. IMRT in 257 patients with head-and-neck and prostate cancer C. Ostheimer 1 , S. Ensminger 1 , M. Janich 1 , F. Sieker 1 , V. Izaguirre 1 , R. Gerlach 1 , D. Vordermark 1 1 Universitaetsklinikum HalleSaale, Radiation Oncology, Halle, Germany Purpose or Objective In modern radiation therapy, complex tumor sites, particularly with concave shapes and a close proximity to organs at risk (OAR) are treated with advanced radiation techniques such as intensity-modulated approaches. Characteristically, the latter feature a dose delivery with variable intensity across multiple radiation beams or arcs. The planning target volume (PTV) receives a