Abstract Book

S1106

ESTRO 37

motion phases were used in the optimization of PBS plans. Corresponding internal target volumes based on the same criteria were created for DS. The prescribed dose to the CTV was 63 Gy in 15 fractions. A customized and experimentally validated interplay effect routine [2] served to compute 4DDs for PBS based on an empirical beam time model, a 4D CT and deformable image registrations. For DS, the dose delivery was distributed equally over all motion phases to compute 4DDs. Resulting treatment plans were analyzed in terms of DVH metrics, i.e. the percentage of over- (V107%) and underdosage (V95%) and the homogeneity index: Results The percentage of over- and underdosage was zero for DS, whereas for PBS it ranged typically from 5 to 10% for a single fraction and approached 0% when considering all fractions. There was almost no difference in the HI for the static dose and the 4DD in DS. In PBS, the HI was around 10 for a single fraction and went down to below 5 (close to the static case) when averaging over all fractions. OARs proximal to the target, such as ribs, could be spared very well with PBS in cases where significant OAR volumes received the prescribed target dose in DS. Furthermore, the dose to the normal liver tissue was reduced in PBS. Conclusion The developed interplay effect model was proven to be useful for 4DD comparisons in proton therapy. It can help to evaluate the available treatment options for a patient and to access residual uncertainties related to organ motion and proton range. For the investigated HCC patients PBS without further motion mitigation techniques was shown to be inferior to DS in terms of homogeneous target coverage for a single fraction. But the study demonstrated the expected high potential of PBS to spare OARs proximal to the target and liver tissue in comparison to DS. [1] Chang et al., Int J Radiation Oncol Biol Phys 2017, 99:41e50 [2] Pfeiler et al., Z Med Phys. 2017, https://doi.org/10.1016/j.zemedi.2017.07.005 EP-2025 Patient setup accuracy in spinal SBRT J.M. Perez moreno 1 , L. Alonso Iracheta 1 , O. Hernando Requejo 2 , R. Alonso Gutierrez 2 1 Hospital Universitario HM Puerta del Sur, Radiofísica y Protección Radiológica, Móstoles, Spain 2 Hospital Universitario HM Puerta del Sur, Radiation Oncology, Móstoles, Spain Purpose or Objective In the treatment with SBRT of vertebral and paravertebral lesions it is usual not to make any expansion to the defined target volume. In addition, it is very close, if not in contact, with the spinal cord. Since high dose gradients are present in this type of treatment and the regions of interest are small in size, a slight deviation in patient placement during treatment may lead to significant deviations from the treated versus planned schedule. It is intended to evaluate the setup error of patients undergoing SBRT treatments of vertebral and paravertebral lesions. Material and Methods Spinal SBRT treatments have been delivered in a VERSA HD linac using Agility MLC and 6FFF energy. Patient setup have been performed with cone beam (CB) based IGRT XVI R5.0, wich allows CB acquisitions simultaneously to treatment delivery. 80 CB images have been acquired during 23 spinal SBRT treatment fractions of 13 patients (10 single dose tretments). Every treatment consist of 4 full VMAT arcs.

Workflow of treatment delivery is as follows: 1-initial patient setup 2-initial CB to setup error correction 3-infrafraction CB (IFCB) simultaneous to every arc delivery 4-6D correction of setup error before next arc, using HexaPOD robotic couch After every intrafraction image registration, time of imaging and 6D setup error are recorded and evaluated versus hypothetical setup error without any correction. Temporal evolution of setup error during treatment delivery are analyzed too. Results The magnitude of the mean setup error for the analyzed population was 0.91 ± 0.07 mm, less than 1 mm (p = 0.1). If no control or intrafraction correction had been performed, the average placement error would have been 1.43 ± 0.09. The values specified are the mean ± one standard deviation. In 3 cases of 80 image registration performed, a setup error greater than 2 mm have been found. If no intrafraction correction would have been applied, there would have been setup errors greater than 2 mm 13 times of 80 cone beam acquired . Figure 1 shows the temporal evolution of the mean total patient setup error. An increase of its magnitude with time is observed if no corrections were applied during the treatment session, as well as the application of those corrections keeps the maximum setup error around one millimeter as maximum.

Conclusion Our analysis indicate that it is possible to perform spinal SBRT treatments with mean setup errors of less than 1mm, although a larger sample is needed to affirm with greater statistical significance. If no intrafraction corrections were applied according to the proposed protocol, an important part of the treatment would have been delivered with a setup error around 2 mm. This setup error can lead to an increase in absorbed dose of up to 13% in the spinal cord. EP-2026 Breast Radiotherapy : Heart position reproducibility with spirometric DIBH. R. Garcia 1 , P. Mazars 1 , E. Jaegle 1 , V. Bodez 1 , C. Khamphan 1 , M. Alayrach 1 , A. Badey 1 , P. Martinez 1 1 Institut Sainte Catherine, Physics department, Avignon, France Purpose or Objective Deep Inspiration Breath Hold (DIBH) is currently a standard practice to protect the heart during left breast radiotherapy. The increased lung volume pushes the heart in the opposite direction of the breast. This mechanical effect drives the heart away from the irradiate volume. Different systems are used to manage this breathing manoeuver which are based on surrogates. The spirometric method offers a guaranty of inspired air volume reproducibility. However, a question remains about the heart position reproducibility.