ESTRO 36 Abstract Book

S842 ESTRO 36 _______________________________________________________________________________________________

Material and Methods Data of 15 patients, including T1 and T2 weighted MRI and CT scans, were used in this study. The pCT was generated according to the methodology described in Speier et al, by segmenting the T1 w and T2 w MRI volume into 6 tissue classes (grey and white matter, cerebrospinal fluid, bone, skin and air). For each patient, three 18 Gy beams (2 axial and 1 coronal,) were designed on the pCT volume, for a total of 45 analyzed beams. The plan was then copied and transferred onto the CT that represented the ground truth. Range shift (RS) between pCT and CT was computed at R 80 over 10 slices. The acceptance threshold for RS was set to 3.5% of R 80 , according to the clinical guidelines of our Institution. Results The median value of RS was 0.6 mm with lowest and highest absolute values being 0.08 mm and 3.8 mm respectively. 40 out of 45 beams passed the acceptance test. Largest discrepancies occurred in correspondence of the surgical hole of the scalp containing a metal plate. This happened because the segmentation process did not include metal classification, thus mis-assigning the Hounsfield Unit to skin or air. In this circumstance, the planned range on the pCT was deeper than the actual one detected on the CT. Conclusion This study showed the feasibility of using pCT, derived from MRI, for proton therapy treatment. The major benefit of MRI acquisition lies in better soft tissue contrast for tumor and organs at risk delineation. Further improvements of the methodology are required for the correct conversion of metal voxels to electron density. EP-1565 Best of both worlds: 3D-CRT-based VMAT for locoregional irradiation in breast cancer. P.G.M. Van Kollenburg 1 , H.J.M. Meijer 1 , M.C. Kunze- Busch 1 , P. Poortmans 1 1 UMC St Radboud Nijmegen, Department of Radiation Oncology, Nijmegen, The Netherlands Postoperative locoregional radiation therapy (RT) is increasingly applied in breast cancer patients as it has been demonstrated to decrease the risk of any recurrence and breast cancer mortality in patients with node-positive disease after mastectomy or breast conserving therapy. However RT has also been associated with side effects such as fibrosis, cardiac and pulmonary toxicity, impaired shoulder function and the induction of secondary malignancies. It is therefore essential to use treatment techniques that enable the delivery of conformal and homogeneous doses, adequately covering the target volumes and limiting the dose to the organs at risk. The technique should also be robust taking into account changes in the position and the shape of the target volumes during treatment. We hereby present the results of the technique as being used in our department. Material and Methods Materials/Methods: 10 breast cancer patients with and an indication for locoregional RT were selected for dosimetric comparison between 3D-CRT and VMAT. All patients underwent a CT- scan with 3-mm slice thickness. Patients with left-sided breast cancer were scanned and treated with voluntary moderately deep inspiration breathhold. The treatment plans were created in the Pinnacle 3 treatment planning system V.9.10 with the Auto-Planning module, using 6 and/or 10MV beams. For each patient a CTV was delineated based on the ESTRO guidelines. A margin of 7 mm was used to generate a PTV. Purpose or Objective Purpose:

Peak-to-Valley Dose Ratios were used to obtain the ‘valley’ dose displayed in Eclipse. We compared dose profiles generated by Eclipse with Geant4 Monte Carlo simulations and measurements from the Imaging & Medical Beamline at The Australian Synchrotron. We are also in the process of performing a plan comparison study using anonymised patient datasets, comparing kilovoltage MRT plans with clinical megavoltage treatment plans. Results The Eclipse TPS performed well in calculating ‘peak’ doses in a water phantom. Considering the simplicity of the algorithm, the ‘valley’ dose and field profiles were also produced with reasonable accuracy, albeit with some underestimation of the valley dose for larger field sizes. Preliminary studies of megavoltage treatment plan comparisons have been performed. Compared to the clinical megavoltage treatment plans, MRT plans demonstrated adequate target coverage whilst meeting normal tissue dose constraints when target volumes were small and relatively superficial. As expected, planning goals for deep seated tumours and target regions distal to bone could not be met using MRT.

A screen shot of the treatment planning system. The peak dose has been calculated for three treatment fields on a head CT scan. Conclusion There are real advantages to using the familiar environment of Eclipse with a new radiotherapy paradigm such as MRT. Although, there are limitations to our MRT calculation engine in Eclipse and further work is required, the data generated in this work are overall encouraging and indicate that the potential for this calculation engine to be implemented in the future as part of a Phase 1 clinical trial. EP-1564 Dosimetric assessment of pseudo-CT based proton planning G. Pileggi 1 , C. Speier 2 , G. Sharp 3 , C. Catana 4 , D. Izquierdo-Garcia 4 , J. Pursley 3 , J. Seco 5 , M.F. Spadea 1 1 Magna Graecia University, Department of Experimental and Clincal Medicine, Catanzaro, Italy 2 Friedrich-Alexander University Erlangen-Nürnberg, Radiation Oncology, Erlangen, Germany 3 Massachusetts General Hospital, Radiation Oncology, Boston, USA 4 Athinoula A. Martinos Center for Biomedical Imaging, Radiology- MGH, Charlestown, USA 5 Deutsches Krebsforschungszentrum - DKFZ, Radiation Oncology, Heidelberg, Germany Purpose or Objective The aim of this work is to use pseudo-CT (pCT) data, obtained from T1 and T2 weighted MRI, for proton therapy planning.