ESTRO 38 Abstract book

S1087 ESTRO 38

Results Mean and SD of positional shifts, systematic and random errors of overall HN and all sub-regions were mostly within 1mm and 1 o (Table I), indicating the excellent positional reproducibility achieved on the MR-sim. Inter-fractional positional shifts in sub-regions were highly correlated with (highest in SI, lowest in roll) and insignificantly different from those in overall HN except for roll in hypopharynx, mainly attributed to the high motion flexibility in the neck even under immobilization ( p =0.039). Sub-regions did exhibit notably different, although insignificant, shift patterns from the overall HN, in particular in roll and yaw, as shown by the box-plot in Fig. 1. Overall HN mainly underestimated the inter-fractional shift magnitude, range and variability in sub-regions. In terms of margin setting, according to van Herk target margin recipe of 2.5Σ+0.7σ, larger target margins might be needed in NP for positional compensation than in overall HN and other sub-regions.

order to derive PTV margins, we adapted the margin recipe described by Nijkamp et al (Radiother Oncol 2012). Results Systematic- and random errors were ranging from 2 mm SD in the upper-lateral region of the mesorectum up to 5 mm SD in the upper-anterior region (figure 1). Local group mean variation was relatively small, ranging between -1 mm in the anterior region and 2mm in the lower-posterior region. Derived PTV margins were smallest in the upper- lateral region (6 mm) and largest in the upper-anterior region (16 mm).

Conclusion Even in rectal cancer patients irradiated in prone position and using a belly-board, mesorectal shape variation is heterogeneous and largest in the upper-anterior region. Derived PTV margins are less than described in literature, ranging from 7 mm in lateral direction up to 16 mm at the upper-anterior region of the mesorectum. EP-1990 Assessment of Inter-Sessional Positional Reproducibility in the HN Sub-Regions Using 1.5T MR- Sim Y. Zhou 1 , J. Yuan 1 , K.F. Cheng 2 , W.W.K. Fung 2 , O.L. Wong 1 , G. Chiu 2 , K.Y. Cheung 1 , S.K. Yu 1 1 Hong Kong Sanatorium & Hospital, Medical Physics and Research Department, Happy Valley, Hong Kong SAR China ; 2 Hong Kong Sanatorium & Hospital, Department of Radiotherapy, Happy Valley, Hong Kong SAR China Purpose or Objective We hypothesize that head-and-neck (HN) sub-regions might show different inter-fractional positional characteristics due to the complicated HN anatomies and different sub-regional motion flexibilities. As such, we attempt to utilize the superior MRI soft tissue contrast to assess these inter-fractional positional characteristics in the HN sub-regions at 1.5T. Material and Methods 14 healthy volunteers were recruited. Each subject was immobilized with a 5-point thermoplastic mask, and received 4 MRI scans using a T1-SPACE sequence (TR/TE = 420/7.2ms, isotropic voxel size = 1.05mm, scan time = 301s) on a 1.5T MR-sim to simulate RT fractions. Sub- regions of brain, nasopharynx (NP), oropharynx, and hypopharynx were manually drawn on the reference images acquired in the first MRI scan. Inter-fractional positional variations in translation and rotation were assessed by rigidly registering the overall HN images and each sub-region to the reference images. Correlations of sub-region positional shifts with that of overall HN were assessed using Pearson correlation. The mean and SD of positional shifts, systematic error (Σ) and random error (σ) of each sub-regional anatomy were calculated, and compared using paired Wilcoxon signed rank test.

Conclusion Sub-millimeter inter-fractional positional reproducibility could be achieved in HN sub-regionals on an MR-sim. HN sub-regions showed notably different inter-fractional positional characteristics and this should be taken into account for positional and motion compensation. EP-1991 PTV margin evaluation for pediatric craniospinal irradiation with 3D and 2D position verification E. Slooten 1 , N. Van Wieringen 1 , R. De Jong 1 , B. Balgobind 1 , S. Huijskens 1 , C. Windmeijer 1 , I. Van Dijk 1 , A. Bel 1 1 Academic Medical Center, Radiotherapie, Amsterdam, The Netherlands