ESTRO 2020 Abstract Book

S1057 ESTRO 2020

which CNR the CBCT was sufficient. The AD was measured on the CT slice in the isocenter. An exponential regression analysis was performed to assess the relation between the image quality and the absorption, i.e. CNR and the AD. We expect an exponential relationship based on the Lambert-Beer law which describes the absorption of radiation through a medium. Based on this analysis we defined a threshold for the usability of inline CBCT, up to a certain value of the AD. Results The AD of the patients varied from 22.7 cm to 34.5 cm (mean 27.8 cm). The CNR varied from 0.6 to 4.9 (mean 2.4). In 5 patients the CNR of the inline CBCT was ≤ 1.3 and categorized as insufficient. The AD for these patients was ≥ 30.6 cm. The result of the exponential regression analysis is shown in Figure 1.

outside of the PTV. For all other changes (e.g. changes in atelectasis, tumour regression) the action level is set at 1cm. The CBCT scan with clinically applied correction was resampled and sent to the treatment planning system (RayStation v7, Raysearch). The dose was recalculated on CBCT and compared to the planned dose. If a dose difference larger than 2% of the total treatment dose was found for a significant volume or if a hot/cold spot occurs a physician is consulted. The need for further actions was discussed taking into account the dose differences, the remaining amount of fractions and patient specific parameters such as e.g. previous irradiation. Results In the selected period 134 recalculations (11% of all patients, see table 1) were performed due to anatomical changes observed during treatment. In 84% of all recalculations it concerned a pelvis, chest wall / breast or lung case. The following changes were observed: in the pelvis region the body contour was reduced due to weight loss, in the chest wall / breast a change (in-/decrease) of the body contour due to seroma and in the lung changes in the amount of atelectasis. In total 27% of the cases needed a treatment plan adaptation. For pelvis the treatment was adapted in 39%, for chest wall / breast 30% and in lung 20% of cases. These adaptations were replanning based on a repeat CT scan or creating a new plan on the original CT scan with the additional information of the CBCT scan. For chest wall and breast patients we noticed that the dose delivered by a full IMRT or VMAT technique was less robust for anatomical changes compared to a tangential fields technique and therefore we had to decrease our action level for contour changes in that region from 1cm to 0.8cm. Conclusion In 11% of all patients an action level for anatomical changes was exceeded and in 27% of the cases treatment adaptation (i.e. ad-hoc ART) was necessary.

Conclusion The AD of the patient can be used as a predictive factor for inline CBCT quality during SBRT for lumbal and sacral spine metastasis. Based on the exponential AD/CNR relation, we defined a threshold diameter of 32 cm above which the CNR of an inline CBCT scan is too low to correctly perform an automatic registration. PO-1894 AD-HOC adaptive radiotherapy: how often do anatomical changes lead to treatment adaptation? Y. Van Herten 1 , N. Van Wieringen 1 , J. Wiersma 1 , R. De Jong 1 , A. Bel 1 1 Amsterdam UMC - location AMC, Radiotherapy, Amsterdam, The Netherlands Purpose or Objective In our hospital we apply online IGRT protocols based on cone beam CT (CBCT, Elekta) imaging for all indications. Each CBCT is not only used for position verification but also for monitoring anatomical changes which may lead to differences in dose delivery or target coverage. We introduced action levels for anatomical changes for RTTs evaluating the CBCTs. When an action level is exceeded a medical physicist estimates the impact on the dose distribution. The purpose of this study was to determine 1) how often an action level was exceeded and 2) how often this results in a treatment adaption. Material and Methods Data was collected of all anatomical changes exceeding our actions levels between August 2018 and May 2019. Action levels are: body contour changes more than 2cm in general; breast / head and neck more than 1cm (0.5cm for breast boost area) or any change yielding a target position

PO-1895 Introduction of surface imaging as part of repeat CT procedure for proton breast cancer patients M. KuijperS 1 , E. Batin 1 , T. Van Faassen-van Loenen 1 , A. Meijers 1 , A. Crijns 1 , J.A. Langendijk 1 1 University Medical Center Groningen, Department of Radiation Oncology/GPTC, Groningen, The Netherlands Purpose or Objective Control CT imaging (repeat CT) during the course of treatment is becoming a standard for most proton institutions, to evaluate anatomical changes and their impact on the treatment being delivered. For these repeat CTs, patient positioning is often based only on lasers and tattoos and therefore may differ from the treatment position: especially the arm position for breast patients (figure 1). Uncertainties in repositioning may result in unnecessary or suboptimal plan-adaptation. For breast treatments differences in arm positioning may hamper accurate visualization of treatment delivered to axillary/supraclavicular nodes (L1 to L4). Surface imaging for breast patients was introduced in the repeat CT workflow to improve the reproducibility of the treatment