Abstract Book

S1128

ESTRO 37

Purpose or Objective Wider shifts affect the anterior limit of the uterus and their compensation leads to a significant internal margin from CTV to PTV despite the use of IGRT. The goal of this work was to quantify uterus organ motion (OM) uncertainties and to define individual and population internal margin. Material and Methods 11 patients with cervix tumor treated with External Beam Radiotherapy and Brachytherapy were studied. Using Eclipse contouring software, uterine, rectal and bladder delineations (ROIs) were performed on simulation CT (CT sim) and on all available CBCTs; ROIs volumes in cc were recorded. The analysis of the first 5 CBCTs (UTERUS 1-5) compared with CT sim (UTERUS 0) provided information on the interfraction OM of the uterus and allowed patients’ ITV definition. The distances UTERUS 0 - ITV were measured in each patient on 3 levels (A, roof of the acetabulum and B, pubic symphysis, in axial plane; C, promontory L5-S1, in coronal plane), A and B provided the anterior shifts of uterus and cervix, respectively; C the cranio-caudal shifts of the fundus. The shifts x, y, z of the center of mass (COM), of the uterus on the CT sim and CBCTs, in all patients, throughout the duration of the treatment, were recorded and used to calculate the population internal margin (van Herk’s formula). Results Variation in volume (cc) of the uterus, linked to the regression of the tumor, were recorded (fig 1). The correlation between volume reduction and elapsed time, throughout the duration of the treatment, was studied using linear regression; a significant change was reported when uterus volume at the CT sim was > 200cc (P value < 0.05). In the period between CT sim and CBCT 1 ( 22 +/- 9 days) no substantial changes in the size of the uterus was recorded (P value = 0.49). With the application of multiple linear regression,no correlation was found between the shifts of COM along x, y, z and bladder and rectum filling. The median value of the uterus displacements was 0.73 cm (range 0.12- 3.07) at point A, 0.54 cm (range 0.4- 2.48) at point B, 0.5 cm (range 0.1-1.51) at point C; 90th percentile values, at the same levels A, B and C were 2.6 cm, 2.38 cm and 1.3 cm, respectively. The recorded shifts of COM(x, y, z) were used to calculate the population internal margin (van Herk’s formula) that resulted 0.82 cm ,x (LL); 1.73 cm, y (AP) and 1.93 cm, z (CC). Conclusion Our data, regarding the measurements of the distances between the limits of the UTERUS 0 and ITV, led to a margin of 2.6 cm (point A), 2.38 cm (point B) ; margins based on recorded shifts of COM resulted inadequate in AP (1.73cm). These results highlight the complexity of uterus OM suggesting that an adaptive approach would be needed and margins should be customized for each patient.

Moreover, daily Image Guidance has to be mandatory to also correct daily set up errors and to check the displacement of target and OARs linked to tumor regression. EP-2059 Real time optical surface IGRT: A mono- institutional prospective study of 110 patients D. Reitz 1 , G. Carl 1 , S. Schönecker 1 , P. Freislederer 1 , M. Pazos 1 , M. Reiner 1 , M. Niyazi 1 , U. Ganswindt 1 , C. Belka 1 , S. Corradini 1 1 Department of Radiation Oncology- University Hospital- LMU Munich, Radiation Oncology, Munich, Germany Purpose or Objective Although the increasing clinical adoption of image-guided radiotherapy (RT) has allowed minimizing patients’ set-up uncertainties, reproducible patient positioning remains a crucial issue in modern RT. Recently, optical surface scanners, based on the absence of any additional radiation exposure, have been introduced into clinical practice in addition to well-established position verification systems. The aim of this prospective study was to evaluate theoretical setup errors of the optical surface scanner Catalyst HD ™ (C-RAD AB, Uppsala, Sweden) on different anatomic regions. Material and Methods Following patient positioning using conventional skin marks and positioning lasers, three-dimensional deviations detected by the Catalyst TM system were recorded and a cone-beam CT (CBCT) was performed. Deviations of the surface-based method were compared to CBCT corrections as gold standard. A theoretical Catalyst TM setup error was calculated between the translational deviations of the surface scanner and the laser positioning. One-sided t-tests were used for statistical analysis. Results Between 10/2016 and 06/2017 a total of 1902 fractions in 110 patients (66 male and 44 female) were evaluated. The distribution of fractions was as follows: Head: 689, Thorax: 460, Abdomen: 630 and Extremities: 123. The data analysis revealed total deviations of 0.09 ± 2.03 mm for the lateral axis, 0.07 ± 3.21 mm for the longitudinal and 0.44 ± 3.08 mm vertical axis for the Catalyst TM system, compared to -0.6 ± 3.54 mm lateral, 0.53 ± 3.47 mm longitudinal and 0.19 ± 3.49 mm vertical for the laser positioning compared to CBCT. The calculated theoretical setup error was -0.15 ± 4.21 mm lateral, 0.46 ± 4.47 mm longitudinal and -0.25 ± 4.54 mm vertical. Stratifying data by anatomic regions, the lowest positional deviations were found in the intra-cranial treatments. Larger deviations occurred in the thoracic and abdominal region (see Fig. 1). A statistical comparison using Two one-sided t-tests (Test references: Head ± 0.5 mm, Thorax ± 1.0 mm, Abdominal ± 1.5 mm and Extremities ± 2.0 mm, Alpha = 0.05) showed a general concordance of the two methods (p ≤ 0.036), excluding the vertical direction of the abdominal region (p = 0.198), where Catalyst TM achieved a higher precision (Mean value of deviation -0.59 mm vs. 0.68 mm for the laser).