ESTRO 2020 Abstract book

S965 ESTRO 2020

accumulated on the pCT. The accumulated doses and the “worst case” uncertainty scenarios (combinations of 5 mm in all the cardinal directions and 3% range error, 14 in total) were used to evaluate the robustness of the plans. Results One of the most noticeable variations was the shrinkage of the parotid gland, which on average was around 26%. The largest reduction occurred for the parotids of patient 5 with 31% and 39% for the ipsilateral and contralateral parotid, which was reflected in the dosimetric results. The plans resulted in adequate CTV coverage for all the patients in the accumulated dose and in the uncertainty scenarios. The dose to the organs at risk evaluated (parotid glands, spinal cord, brainstem) did not exceed the clinical constraints, neither in the accumulated dose nor in the uncertainty scenarios. The difference between accumulated dose to OAR and nominal plan was on average within +/- 5% (see figure 1) for all the patients except for patient 5, who showed an increase in the mean dose to the ipsilateral parotid gland of 31% (from 18.5 Gy to 24.5 Gy, see figure 2).

Conclusion For prostate patients treated on MRIdian, daily propagation of auto-deformed contours resulted in consistently larger volumes for rectum, bladder and femoral heads, with maximum distances between contours up to 23 mm. Consequently, this leads to significantly different DVH parameters for rectum and bladder, with the risk of underestimating the dose to these structures. Naturally, the clinical impact of these dosimetric differences must be weighed against any extra contouring and treatment time in a daily adaptive workflow. PO-1664 Standard robust optimization of IMPT for nasopharyngeal cancer covers anatomical changes R. Argota Perez 1 , L.P. Kaplan 1 , M.B. Sharma 1 , K. Jensen 2 , U.V. Elstrøm 1 , A. Vestergaard 2 , J. Breede Baltzer Petersen 1,2 , S.S. Korremann 1,2,3 1 Aarhus University Hospital, Department of Oncology, Aarhus N, Denmark ; 2 Aarhus University Hospital, Danish Center for Particle Therapy, Aarhus N, Denmark ; 3 Aarhus University, Department of Clinical Medicine, Aarhus N, Denmark Purpose or Objective Despite its advantages regarding sparing of organs at risk, proton therapy is very sensitive to the many uncertainties during treatment. Robustness towards setup and range uncertainties has been added to treatment planning systems (TPS) but inter-fractional anatomical variations are still a pending issue. The aim of this study is to test how well the geometrical and range robustness included in the TPS deals with anatomical variations in head and neck patients. Material and Methods Five nasopharyngeal cancer patients, treated previously in our center with photon radiotherapy, were used retrospectively for this study. All patients showed various anatomical variations (weight loss, filling/emptying of nasal cavities, tumor and OAR shrinkage) during their treatment. New proton treatment plans were made for this study using Eclipse v13.7. Plans consisted of three beams with multiple-field optimization (robust optimization: ±5mm shift in the cardinal directions and ±3% range uncertainty). Using Velocity v4.0, virtual CT scans (vCT) were obtained by deforming the planning CT (pCT) to the anatomy of the daily CBCTs. The plans were recalculated on the vCT and the resulting dose

Conclusion Target volume coverage was not compromised for the cohort of patients studied. Even though the dose to the