ESTRO 37 Abstract book

ESTRO 37

S527

registered to the planCT using BA following re-calculation of the plan on the fx-CTs. Plan selection used a BA-match to the planCT for selection of the smallest PTV in the library that encompassed the delineated CTVs in each fx- CTs. The dose delivered using plan selection and the conventional approach was compared for all targets based on dose to 98% of the target (D98) and for organs of risk using generalised equivalent uniform dose (gEUD) for rectum (volume parameter k=12), bladder (k=8), small and large bowel (both k=4). Results Due to correlation in the anterior-posterior (AP) and superior-inferior (SI) direction, the plan library consisted of 3 p/sv-PTVs: 2 p/sv-CTVs shifted ± 1.5 standard deviation of the systematic error in AP (± 5.0mm) and SI (±4.6mm) direction and 1 p/sv-CTV in its planCT position (Fig.1). These CTVs used 5mm isotropic PTV margins. The conventional plan used p/sv-PTV margins of 5, 10 and 10mm in left-right, AP and SI directions, respectively. The plans in the library were used in the majority of the fx-CTs, but the conventional plan had to be used at least once for 4 of the 5 patients. D98-values were similar for all targets in plan selection and conventional delivery. Significant dose reductions from plan selection were obtained for rectum (3.1 Gy) and bladder (1.9 Gy), while the average doses to small and large bowel were comparable for the two delivery approaches (Fig.2).

PO-0960 Plan selection in proton therapy for simultaneous treatment of multiple targets in the male pelvis B. Abal 1 , G.M. Engesth 2 , L.B. Hysing 1,2 , K. Ytre-Hauge 1 , S. Thörnqvist 1,2 1 University of Bergen, Faculty of Mathematics and Natural Sciences, Bergen, Norway 2 Haukeland University Hospital, Department of oncology and medical physics, Bergen, Norway Purpose or Objective Proton therapy (PT) to the prostate and the pelvic lymph nodes is challenged by the different motion patterns of the targets, which can affect the proton range. The aim of this study was to evaluate if a plan library with shifted positions of the prostate/seminal vesicle planning target volumes (p/sv-PTV) in relation to the lymph node target can improve PT for locally advanced prostate cancer. Material and Methods Target motion was analysed using a database of rigid shifts of prostate gold-markers relative to bony anatomy (BA), obtained from 18 patients each with 7-10 inter- fraction CTs (fx-CTs) in addition to their planCT. A plan library with different p/sv-PTVs was defined from the distribution and direction of the shifts in the database. The shifts were also input for calculation of p/sv-PTV with a standard margin recipe used in conventional treatment delivery. The lymph node PTV was identical in all plans for both the plan library and the conventional delivery. Intensity modulated PT with two lateral opposing fields delivering 67.5 GyRBE to the p/sv-PTV and 50 GyRBE to the lymph node PTV were optimised in Eclipse (Varian medical systems). Plan selection and conventional treatment delivery were applied retrospectively for 5 patients with 8-9 fx-CTs, excluded from the database used to define the plan library. For the conventional delivery, the fx-CTs were rigidly

Conclusion Plan selection from the population-based plan library was feasible for all patients. As compared to the conventional plan, plan selection reduced dose to the rectum and bladder without compromising target coverage. PO-0961 Anatomical predictors of differences between planned and delivered dose to Head and neck OARs D. Noble 1 , K. Harrison 2 , P. Yeap 2 , S. Seah 2 , Y. Zheng 3 , A. Bates 4 , M. Romanchikova 5 , L. Shelley 6 , S. Thomas 5 , M. Parker 2 , N. Burnet 1 1 University of Cambridge, Department of Oncology, Cambridge, United Kingdom 2 University of Cambridge, Cavendish Laboratory, Cambridge, United Kingdom 3 University of Cambridge, School of Clinical Medicine, Cambridge, United Kingdom 4 Cambridge University Hospital's NHS Foundation Trust,