ESTRO 36 Abstract Book

S556 ESTRO 36 _______________________________________________________________________________________________

approach did so due to poor PTV coverage (n=5) or unacceptable lung dose (n=2). Of the 18 patients who had an acceptable plan generated (regardless of planning approach), one was achieved with the 8-fraction regime, with the remaining needing the 12-fraction regime. OAR PRV max doses were 2-3.5% over the OAR dose for the conventional and hybrid approaches and 6% for the SABR approach, highlighting the need for IGRT. Conclusion SABR was feasible for approximately half of the locally- advanced NSCLC patients we assessed and for almost all of these cases only a 12-fraction scheme was feasible. If the alternative to SABR is no treatment at all, compromises to tumour coverage or OAR tolerances may be acceptable, increasing feasibility. This data will inform a phase I study testing the safety of SABR for locally advanced NSCLC. PO-1009 Evaluation of setup margins using cone-beam CT for prostate and pelvic nodes irradiation A. Van Nunen 1 , T. Budiharto 1 , B. De Vocht 1 , D. Schuring 1 1 Catharina Ziekenhuis, Radiotherapie, Eindhoven, The Netherlands Purpose or Objective In 2014 radiotherapy for prostate and pelvic nodes was introduced in the Catharina hospital. For this tumour site, CBCT is used for position verification. Due to variation in prostate position in relation to lymph nodes, large setup margins are required to deliver the correct target dose to both volumes. A CTV-PTV margin of 1 cm is used for both prostate and lymph nodes. The aim of this study was to evaluate the required setup margins using different correction and registration strategies. Material and Methods CBCT-scans of 20 patients were included in this study. 220 scans were analysed retrospectively. Patients were treated with an offline SAL correction protocol with an initial action level of 10 mm and a maximum number of 3 measurements. When large day-to-day variations were observed, an online correction protocol was performed. All CBCT-scans were registered automatically using a grey value, seed or bone match algorithm of the XVI software (Elekta, Crawley, UK). For these automatic matches either a clipbox containing bony structures and the entire PTV, a mask consisting of the prostate or a mask consisting of lymph nodes CTV was used (figure 1). Registration of the lymph node area was performed to determine the correlation between bony anatomy and the position of the pelvic lymph nodes. For all these registrations all translations, rotations and table corrections were collected. From these results the random and systematic setup errors were determined. The required setup margins were then calculated using the margin recipe M = 2.5Σ+0.7σ (Σ: systematic error, σ: random error). Poster: RTT track: Image guided radiotherapy and verification protocols

Results There was a large correlation between bony structures and lymph nodes in all directions (correlation coefficient > 0.82). Correlation between bony structures and the prostate position was large in lateral direction and small in longitudinal and vertical direction due to large variation in rectal filling. This resulted in larger margins in this direction. The required setup margins are summarised in Table 1. In this margin calculation, we did not account for rotations, intrafraction variation, delineation, treatment and match uncertainties. When applying online position verification on the prostate, setup margins for lymph nodes must be 0.19 cm (lateral), 0.73 cm (longitudinal) and 0.57 cm (vertical). When applying online setup correction on bony structures, setup margins for the prostate must be 0.25 cm, 0.73 cm and 0.73 cm respectively. Offline setup correction on prostate resulted in the largest margins for both volumes.

Conclusion The required setup margin depends on the applied correction strategy. When applying position verification on bony structures, larger margins are required for CTV prostate. When applying position verification on prostate, larger margins are required for CTV lymph nodes. When applying these margins clinically, additional margins are needed to account for rotations, intrafraction variation, delineation, treatment and match uncertainties. PO-1010 Investigation of reproducibility of bolus position based on kV CBCT imaging E. Dabrowska 1,2 , P. Czuchraniuk 3 , M. Gruda 1 , P. Kukolowicz 1 , A. Zawadzka 1 1 Maria Sklodowska-Curie Memorial Cancer Center and Institute of Oncology, Department of Medical Physics, Warsaw, Poland 2 Faculty of Physics University of Warsaw, Department of Biomedical Physics, Warsaw, Poland 3 Maria Sklodowska-Curie Memorial Cancer Center and