Abstract Book

S1125

ESTRO 37

EP-2054 CBCT verified variation of PTV position in DIBH treatment of left breast T. Koivumäki 1 1 Central Finland Central Hospital, Medical Physics, Jyväskylä, Finland Purpose or Objective The aim of this study was to evaluate the variation in the position of the planning target volume (PTV) during the course of left breast irradiation in DIBH. Material and Methods The data of 15 left-sided breast cancer patients was retrospectively reviewed for this study. A moderate DIBH (approximately 80% of maximum inspiration) was visually instructed with the aid of an optical respiratory monitoring system (C-RAD AB, Uppsala, Sweden). The skin mark –based patient setup was verified for each treatment fraction with pre-treatment CBCT. The patients were treated with Infinity linear accelerators equipped with XVI imaging system (Elekta AB, Stockholm, Sweden). The CBCT images were registered with the planning CT in the Mosaiq system (v2.62) to record the final patient setup position in relation to skin mark – based setup. The registration was performed in 6D recording cranio-caudal (CC), anterior-posterior (AP) and left-right (LR) transitions as well as rotations in coronal, sagittal and transversal plane. The patients were treated according to a hypofractionated scheme (15x2.67Gy) resulting in 225 analyzed fractions. Results The mean shifts from the skin mark –based PTV setup were 1±5 mm anteriorly, 0±5 mm in left-right direction and 3±4 mm in caudally (Fig. 1). Mean absolute shifts from the skin mark based PTV setup were 3±3 mm, 3±3 mm and 4±3 mm in AP, LR and CC directions, respectively. When considering the typical 5mm tolerance margin used to create PTV from CTV, the shifts were smaller than 5mm in 82%, 80% and 69% of the fractions when considering AP, LR and CC directions independently. The difference between the planned position and the skin mark –based setup in AP, LR and CC directions was equal to or larger than 10 mm in 7%, 5% and 4% of fractions. The mean rotations were 0.6±1.5°, 0.3±1.7° and -0.2±1.9° in coronal, sagittal and transversal planes, respectively (Fig. 2). The corresponding absolute rotations were 1.1±1.2°, 1.4±1.1° and 1.3±1.3°. The rotational error in patient setup may add up a significant component to the total setup error. To illustrate, a rotation of 1° corresponds to a 1.7 mm shift at a distance of 10 cm.

Figure 2. Histogram of the rotations in three orthogonal planes. For readability, the data is presented in 1° bins. In addition, 2 fractions with coronal rotation of more than 5.5° and 2 fractions with transversal rotation of less than -5.5° were omitted from the figure. Conclusion Daily CBCT was found beneficial in verifying patient setup in DIBH treatments of left breast. The CBCT scans provide anatomical information on soft tissues, whereas planar images provide only 2D data mainly from the bony structures. Importantly, even more than one fourth of fractions would have been missed more than 5 mm in some orthogonal direction in this study population if only skin mark –based setup were used. A dose-optimized CBCT protocol is recommended for daily CBCT to minimize patient exposure. EP-2055 Comparison of two correction protocols for IGRT in prostate cancer patients treated with IMRT M. Altabas Gonzalez 1 , X. Maldonado 1 , X. Fa 2 , J. Planas 3 , A. Celma 3 , A. Pons 1 , M. Ruiz 1 , J. Mateos 1 , B. Pérez 1 , F. Alvarenga 1 , R. Vergés 1 , D. García 1 , J. Giralt 1 1 Hospital Universitario Vall d'Hebron, Radiation Oncology, Barcelona, Spain 2 Hospital Universitario Vall d'Hebron, Medical Physics, Barcelona, Spain 3 Hospital Universitario Vall d'Hebron, Urology, Barcelona, Spain Purpose or Objective To compare our conventional off-line non action level (NAL-3) error correction protocol by means of portal imaging to the on-line IGRT protocol with fiducial markers in terms of compliance to our CTV-PTV margins design. Material and Methods Between 01/2016 and 04/2017, twenty patients with localized prostate cancer were treated with IMRT-IGRT to a dose of 77 Gy (2.2 Gy/d) with daily correction of the target position based on kilovoltage imaging of implanted prostatic fiducial markers. This group of patients was prospectively compared with a cohort of 115 patients treated with IMRT to the same prescription dose with a tight off-line NAL-3 protocol using bony references with a weekly correction algorithm if the matching was >3mm. All patients were treated with the same CTV-PTV margins (6 mm). The width of the margins was calculated with the McKenzie/Van Herk/Minjheer formula. We compared the compliance of the CTV-PTV margins regarding the image guided protocol used. The statistical analysis of data was performed using independent t-test and Mann- Whitney U test. Results In the off-line group, 1150 set of images were analyzed. 47% of the patients required a change in the set up, and in 11% more than once. In 4 patients the initial set up was over the threshold limits and was repositioned. In the On-Line IGRT with fiducial markers group, 627 images

Figure 1. Histogram of the shifts in three orthogonal directions.