ESTRO 37 Abstract book

S1102

ESTRO 37

Material and Methods A total of 61 C57BL/6 adult male mice were used in this study, divided into six groups: the control group (n=12), and five groups with different dose levels of 4 Gy (n=9), 8 Gy (n=8), 12 Gy (n=10), 16 Gy (n=12) and, 20 Gy (n=10). Mice were irradiated using 5-mm circular parallel- opposed fields targeting the upper right lung. After irradiation, all mice were imaged at regular intervals over 39 weeks (i.e., 10 imaging time points). An ROI encompassing the volume of the lung which received maximum dose (cut-off at 80% of maximum lung dose) was delineated on the final CT and used for our analysis. In total, 1384 radiomic features comprising: a) first-order statistics, b) shape and c) filtered texture features were extracted from the ROI. A Wilcoxon rank-sum test was performed to assess the difference in the value of radiomic features across groups. A p-value smaller than 0.05 was considered to be statistically significant. Results We observe statistically significant differences in feature values between the control and irradiated groups. Between 0 and 20 Gy; a total of 459 features were significantly different. For the lowest dose level (4 Gy), significant differences were also found compared to the control group (n=108). Comparing the irradiated groups, we found a statistical difference between 16 and 20 Gy in a total of 188 features, versus 31 features between 8 and 12 Gy(see Table 1). Conclusion This study shows differences in radiomic feature values between groups irradiated with different dose levels. Significant feature differences were observed between the 0 and 4 Gy groups, which were not found using simple CT greyscale difference analysis in previous publications. Changes in radiomic features between baseline CT and follow-up CT scans could potentially be used as non- invasive biomarker to detect changes in the lung after radiation at low dose levels. Longitudinal analysis of earlier time points is currently ongoing.

Conclusion Modest inter-patient variations in organ-mean RBE values were observed for the heart, lungs and thyroid, while only small variations were seen for the brainstem. Greater variations in RBE were seen between estimates from different RBE models, thus RBE models should be used with awareness of this. The estimated median RBE values could be used in assessment of proton CSI treatment plans optimized with RBE 1.1 . EP-2015 Radiomics can detect changes in lung after low dose irradiation: a preclinical study T.A.G. Refaee 1 , A. Ibrahim 1 , H. Woodruff 1,2 , R.T.H. Leijenaar 1 , A. Jochems 1 , R.T.H. Larue 1 , E.E.C. De Jong 1 , F. Verhaegen 2 , L. Dubois 3 , P. Lambin 1 1 Maastricht university, The D-Lab: Decision Support for Precision Medicine- GROW - School for Oncology and Developmental Biology- Maastricht Comprehensive Cancer Centre- Maastricht University Medical Centre, Maastricht, The Netherlands 2 Maastricht university, Maastricht Radiation Oncology Maastro- GROW – School for Oncology and Developmental Biology- Maastricht Comprehensive Cancer Centre- Maastricht University Medical Centre MUMC, Maastricht, The Netherlands 3 Maastricht university, Department of Radiotherapy- GROW – School for Oncology and Developmental Biology- Maastricht Comprehensive Cancer Centre- Maastricht University Medical Centre MUMC- Maastricht, Maastricht, The Netherlands Purpose or Objective In radiotherapy for thoracic cancers, dose constraints are determined by lung toxicity. In this study, we evaluated changes in radiomic features on CT obtained from a radiation-induced lung fibrosis (RILF) study in mice, after partial lung irradiation for different dose levels (Granton et.al, 2014). It would help monitoring radiation-induced changes in low dose areas.

Tabel1: The significant features between groups Groups Number of Features (α=0.05)

0 vs 4 Gy 108 0 vs 8 Gy 186 0 vs 12 Gy 78 0 vs 16 Gy 329 0 vs 20 Gy 459 4 vs 8 Gy 151 4 vs 20 Gy 487 8 vs 12 Gy 31 16 vs 20 Gy 188

Electronic Poster: Physics track: Intra-fraction motion management

EP-2016 A prospective study developing decision algorithm for respiration controlled radiotherapy E. Kim 1 , E.K. Chie 1 , H.C. Kang 1 , S.Y. Park 1 , C.H. Choi 1 , J.M. Park 1 , J.I. Kim 1 , H.G. Wu 1 1 Seoul National University Hospital, Radiation Oncology, Seoul, Korea Republic of Purpose or Objective The primary goal was to develop patient selection decision algorithm based on pre-treatment clinical factors for patients undergoing radiotherapy to