ESTRO 38 Abstract book

S531 ESTRO 38

assess the inter-DIBH uncertainty (Fig1a). To assess inter- DIBH uncertainty during treatment, pre-treatment CBCT scans were acquired for daily setup (CBCT1) and post- treatment (CBCT2) at the first three fractions and once a week afterwards. For LC with tumour (T) and lymph node (N) involvement, T and N were analysed separately. SL targets were mostly mediastinal (M) and grouped with the mediastinal N targets. Systematic ( Ʃ ) and random (σ) errors were calculated using the van Herk formalism. Intra-DIBH uncertainty was measured by intra-treatment 5Hz fluoroscopic kV imaging of gold markers implanted at N-site for LC (Fig1b) and motion trajectories were derived (Fig1c). The interquartile motion range (IQR) was determined for each treatment field.

Conclusion Spine tracking with CK is highly efficient also for non- immobilized patients: the residual error seems to be time independent and is close to 0. If tracking is not employed (I.e.: with other delivery modalities), shifts show a dependence on time. For delivery times exceeding 5' in the considered non-immobilized patient scenario, tracking seems to be highly recommendable. PO-0974 Intra-fractional stability of Deep Inspiration Breath Hold during RT for lung and lymphoma cancer D. Sloth Møller 1 , M.L. Schmidt 1 , T. Ravkilde 1 , P.R. Poulsen 1 , J. Hansen 1 , E.S. Worm 1 , H.H. Schmidt 2 , M.M. Knap 2 , A. Safwat 2 , H.K. Rose 2 , L. Hoffmann 1 1 Aarhus University Hospital, Department of Medical Physics, Aarhus C, Denmark ; 2 Aarhus University Hospital, Department of Oncology, Aarhus C, Denmark Purpose or Objective Deep Inspiration Breath Hold (DIBH) during thoracic RT is attractive as it may reduce dose to the lungs and heart compared to free breathing RT. However, intra-fractional geometric instability during several breath holds may decrease the target coverage. We investigate intra- fractional uncertainties during planning, before/after each fraction delivery, and during actual field delivery. Material and Methods Twenty-two patients (15 lung cancer (LC) and 7 sarcoma/lymphoma (SL)) were treated with DIBH-RT. The RPM system with an external marker (EM) placed caudally on the thoracic cage was used as a surrogate for the DIBH level. For all scans and treatments, the DIBH level measured by EM was < 2mm. Four DIBH planning-CT (pCT) scans were acquired: one for RT planning and three to

Results The inter-DIBH uncertainty analysis (Fig 2) for LC T sites showed low stability in longitudinal position and poor correspondence between the uncertainties derived from DIBH pCTs and inter-DIBH CBCT1/2. In some patients, only minor overlap between motion ranges was seen and the mean value of the range was shifted from zero. For M sites the stability was higher and the correspondence better due to target location closer to EM, but the DIBH pCTs were still poor predictors of the actual inter-DIBH uncertainty during treatment. Errors in mm for the LC T sites were Ʃ LR , Ʃ AP, Ʃ CC =0.7,1.5,2.3, σ LR , σ AP, σ CC =0.8,1.5,1.4 (by CT) and Ʃ LR , Ʃ AP, Ʃ CC =0.8,1.1,1.6, σ LR , σ AP, σ CC =1.2,1.8,2.9 (by CBCT). For M sites Ʃ LR , Ʃ AP, Ʃ CC =1.1,1.1,2.0, σ LR , σ AP, σ CC =0.9,0.9,1.7 (CT) and Ʃ LR , Ʃ AP, Ʃ CC =0.7,1.2,1.1, σ LR , σ AP, σ CC =0.9,1.3,1.6 (CBCT). These errors do not reflect the difference between CT and CBCT observed for the individual patients (Fig 2). Fig 1c shows examples of the intra-DIBH uncertainty. The mean IQR over all DIBH’s for each patient ranged from 0.4mm to 1.9mm, in good agreement with the 2mm allowed EM window. The max IQR, however, ranged from 2.8mm to 16.4mm and 1%-23% of all DIBH’s of each patient had IQR>2mm.