ESTRO 36 Abstract Book

S186 ESTRO 36 2017 _______________________________________________________________________________________________

A. Baker 1 , T. Hague 2 , Y. Tsang 1 , P.J. Hoskin 1 1 Mount Vernon Hospital, Department of Radiotherapy, Northwood Middlesex, United Kingdom 2 Mount Vernon Hospital, Department of Radiotherapy Physics, Northwood Middlesex, United Kingdom Purpose or Objective Plan of the day (PoD) ART for cervical cancer patients can potentially reduce toxicity and the risk of geometrical miss but may be resource intensive. In order to implement accurate PoD for these patients this study aimed to assess the accuracy of adaptive online plan selection and linac resource impact. Material and Methods An initial patient cohort had planning CTs acquired with an empty and full bladder and an intermediate MRI. CTVs were outlined on each of the datasets to include uterus and proximal vagina, from which an ITV and PTV were defined with further nodal volumes as required. VMAT plans were created depending on the amount of uterine movement, with a further plan using the previous standard technique as a backup. Online daily CBCT was performed for all patients with additional kV planar images used for nodal positioning in one patient and for pelvic tilt in another. Plan selection following online registration using a combination of bony anatomy and soft tissue was performed by 2 members of the project team (observers) who had attended an anatomical training session and had a range of experience with female pelvic CBCT analysis. A 3mm margin between the visible target anatomy and the PTV contour was allowed for intrafraction motion. This was assessed through the addition of weekly post-treatment CBCTs. In- room time (patient enter to exit) was recorded at each session and patients were booked into the departmental 20 minutes time slot for ART. A consensus standard PoD was agreed offline by an experienced clinician and RTT. Offline analysis was performed to measure concordance with the consensus standard PoD and the online decision. Results A total of 100 online PoD evaluations plus 600 offline evaluations, by 6 observers, were used for the analysis. The median concordance between the consensus standard PoD and the online plan selection was 98%. Where poor concordance was observed between online plan selection and the consensus standard PoD, a safe larger volume option was chosen online. Post-treatment CBCT’s showed target anatomy was covered in all but 1 case. In-room timing ranged from 10 – 30mins with a median time of 19mins. The median score of the 4 observers offline compared to the consensus standard was 86%. The range between individuals was 76%- 96% and between patients was 78 – 96%.

plans needs to be built to achieve the full potential of ART. OC-0354 Dosimetric impact of anatomical changes in photon and particle therapy for pancreatic cancer K. Crama 1 , A.C. Houweling 1 , J. Visser 1 , K. Fukata 2 , C.R.N. Rasch 1 , T. Ohno 2 , A. Bel 1 , A. Van der Horst 1 1 Academic Medical Center, Department of Radiation Oncology, Amsterdam, The Netherlands 2 Gunma University, Gunma University Heavy Ion Medical Center, Maebashi, Japan Purpose or Objective Radiotherapy of pancreatic tumors is toxic due to the high dose to surrounding organs-at-risk (OARs). Irradiation with charged particles is characterized by a sharp dose fall-off around the target area. Compared to photon therapy, OARs can be further spared while delivering a high dose to the tumor. Treatment planning studies in pancreatic cancer patients have shown this benefit of charged particle therapy over photon therapy. However, intra- and interfractional changes may greatly affect the robustness of particle therapy. Past studies only investigated differences in planned dose; studies comparing the robustness of different modalities have not been published yet. We compared the dosimetric impact of interfractional anatomical changes (i.e. body contour differences, gastrointestinal gas volume changes and setup errors) in photon, proton and carbon ion therapy for pancreatic cancer patients. Intrafractional changes were not taken into account in this study. Material and Methods Photon, proton and carbon ion treatment plans (36 Gy, 12 fractions) were created for 9 patients. For the particle therapy plans, the relative radiobiological effectiveness was taken into account. To simulate daily online setup correction, the CBCTs were rigidly registered (only translations) to the planning CT using fiducial markers. Fraction dose calculation was then made possible by deformable registering the planning CT to each of the 12 CBCTs. Gastrointestinal gas was delineated on each CBCT and copied to the deformed CT, a relative density override was applied for dose calculation (0.01). Fraction doses were accumulated rigidly. To compare planned and accumulated dose, for each radiotherapy modality, dose volume histogram (DVH) parameters of the planned and accumulated dose were determined for the internal gross tumor volume (iGTV), internal clinical target volume (iCTV) and OARS (duodenum, stomach, kidneys, liver and spinal cord).

Conclusion High online concordance of 98% with the consensus standard PoD demonstrates that the initial training equipped the team with appropriate knowledge to perform accurate plan selection. A combination of 2 observers online achieve closer results to the consensus standard rather than individually. The joint decision making can be performed within the standard departmental ART time slot of 20 minutes. The CBCT data, consensus standard PoD and anatomy training can be used as part of the assessment programme for future RTT observers. Greater confidence in choosing smaller volume