ESTRO 36 Abstract Book

S980 ESTRO 36 2017 _______________________________________________________________________________________________

K. Koiwai 1 , A. Shinoda 1 , T. Ozawa 1 , H. Matsushita 1 , A. Fukazawa 1 , K. Sakai 1 , M. Kadoya 1 1 Shinshu University School of Med., Radiology, Matsumoto, Japan Purpose or Objective In radiotherapy for head and neck cancers, postural reproducibility is sometimes deteriorated due to the non- rigidity of the head and neck. Image-guided corrections cannot completely compensate for the deterioration. We evaluated whether aligning the chest with a treatment couch just before immobilisation improved the postural reproducibility in radiotherapy for head and neck cancers. Material and Methods Fifty patients with head and neck cancers who underwent radiotherapy from February 2015 to January 2016 were selected for this study. Twenty five patients were treated with the “aligning-the-chest” method, and others not. When we acquired planning CT images of the patients with the aligning-the-chest method, we drew marks on the skin of their anterior chest in line with the fixed marks on the both sides of the couch of the CT scanner which were geometrically the same as on the treatment couch. On the set-up for the treatment, the marks on their chest were aligned with the fixed marks on the both sides of the treatment couch just before immobilisation. For all patients, 2 oblique X-ray images (image 1 and image 2) were obtained on the first session of the treatment. We generated digital reconstructed radiographies (DRRs) representing an ideal posture from the planning CT images with the maximal similarity to the corresponding X-ray images with 6 degree-of-freedom registration. On the both X-ray images and DRRs, we measured the mandibuler- spinal angle that was defined as the angle formed by the line projected parallel to the base of the mandible and the line projected parallel to the alignment of the lower cervical spine. As an index for postural reproducibility, the difference of the mandibuler-spinal angle between the X- ray image and the DRR (DMSA) was calculated in each patient. The raw and absolute DMSA were analysed. Results In the mean value of the raw DMSA, there was no difference between the patients with the aligning-the- chest method and without the method (image 1 and 2, - 0.5° vs. 1.7° and 1.2° vs. 3.0°, p = 0.1713 and p = 0.1072, respectively). The mean value of the absolute DMSA in the patients with the aligning-the-chest method was significantly smaller than that in the patients without the method (image 1 and 2, 1.3° vs. 4.7° and 2.0° vs. 5.7°, p = 0.0004 and p < 0.0001, respectively). Conclusion Aligning the chest with a treatment couch before immobilisation improved postural reproducibility in radiotherapy for head and neck cancers. EP-1812 A study on the patient positioning accuracy for stereotactic radiotherapy of brain lesions O. Pashkovskaya 1 , E. Polovnikov 1 1 NSRICP, Oncology and radiotherapy, Novosibirsk, Russian Federation Purpose or Objective The purpose of this study is an investigation of the patient positioning accuracy (inter-fraction and intra-fraction shifts) for stereotactic radiation treatment of the brain lesions using invasive Lexsell Coordinate Frame G®, non- invasive vacuum-activated head frame system HeadFix® and individual thermoplastic masks, and comparison position errors of these systems. Material and Methods A total of 532 patients, which were treated for benign and malignant brain tumors in our center, represented study population. The invasive Lexsell Coordinate Frame G® (LFG), vacuum-activated head frame system HeadFix® (HF) and individual thermoplastic masks (TM) were used to

immobilize patients during pre-radiation preparation and radiation treatment. Dose distributions were typically planned with a volumetric modulated arc therapy technique and noncomplanar arcs if needed for optimal dose distributions. Positioning accuracy was estimated using linac-mounted cone beam computer tomography (CBCT) scan system. The target intra-fraction offsets were determined by comparison target positions before and after the irradiation procedures with planning positions. To determine inter-fraction shifts data from daily CBCT scans were used. Statistical samples of target offsets in six coordinates were collected and statistical analysis was carried out. Results For 386 patients TMs were used, for 31 LFG and for others 115 HF. Analysis of inter-fraction shifts, based on scaterplots, showed no significant correlations and selected systematic errors in patient’s positioning. Quantitative estimations of mean values and standard deviations also demonstrated acceptable accuracy. It was concluded that it’s necessary to do daily CBCT control, if shifts at least in one direction are greater or close to 5 mm during first four procedures. This is due to patient’s specific characteristic and TM doesn’t provide sufficient immobilization. Mask remaking does not always solve the problem. As the result of intra-fraction offset estimations, the position errors for LFG and HF were less than 0.3 mm in coordinates and less than 0.3 degrees in rotation. Conclusion Study of inter-fraction shifts showed that TMs provide acceptable setup reproducibility, as consistent with PTV. With respect to intra-fraction shifts, due to the slight differences of the offset error results, the choice between LFG and HF should be determined by the individual patient characteristics and does not affect on the accuracy of the delivered dose. EP-1813 The investigation of the immobilization devices and localization methods for brain cancer in P- SRS. R.D. Lin 1 , J. Gao 1 , P. Chen 1 1 Chang Gung medical hospital, Proton and radiation therapy center, Taoyuan, Taiwan Purpose or Objective There are two aims of this study: one is to assess the accuracy of the Base of Skull (BoS) immobilization cast and BRW headring for proton stereotactic radiosurgery (P- SRS). The other one is to investigate the accuracy of imaging-guided localization method by using bony structure and fiducial markers as different landmarks. Material and Methods To assess the accuracy of the immobilization devices, we immobilized the CIRS Radiosurgery Head Phantom with hidden target by BoS cast and headring, respectively. Then we implemented the imaging guidance by using bony structure. Compare to the localization method, the CIRS Radiosurgery Head Phantom with hidden target was immobilized by headring followed by implementation of the imaging guided process using bony structure and fiducial markers, respectively. The overall couch shifts were performed by the radiation therapists, the deviation with respect to the proton isocenter was measured by using the hidden target with paired t-test and 3D vector . Results For BoS cast and headring, the deviations (mm) were 0.54±0.68 and 0.48±0.28 (p ＝ 0.412) in right-left (RL), 0.22±0.28 and -0.11±0.34 (p ＝ 0.186) in superior-inferior (SI), -0.30±0.11 and -0.15±0.22 (p ＝ 0.167) in anterior- posterior (AP) directions, and 0.85±0.42 and 0.60±0.32 (p ＝ 0.017) in 3D vector, respectively. For the imaging- guided localization method, the deviations (mm) of bony