ESTRO 2021 Abstract Book

S864

ESTRO 2021

FFF-VMAT Hypofractionated Radiotherapy with SIB associated with concomitant/adjuvant TMZ in not eligible for maximal safe surgical resection patients affected by GBM is an effective and safe treatment compared with the current literature. To confirm these results, prospective study could be warranted.

PO-1035 Re-Irradiation with high-dose stereotactic radiotherapy for recurrent High-grade Gliomas A. Mousli 1 , V. Bocquet 2 , B. Frederick 3 , C. Louis 3 , S. Biver 3 , S. Philippi 3 , P. Nickers 3 , M. Untereiner 4 , G. Vogin 3 1 centre François Baclesse, Radiaiton Oncology Department, Esch-Sur-Alzette, Luxembourg; 2 competence Center For Methodology And Statistics, Luxembourg Institute Of Health, Department Of Population Health, Luxembourg , Luxembourg; 3 centre François Baclesse, Radiaiton Oncology Department, Esch-Sur-Alzette, Luxembourg; 4 centre François Baclesse, Radiaiton Oncology Department,, Esch-Sur-Alzette, Luxembourg Purpose or Objective Treatment of recurrent high-grade gliomas (HGG) is not well standardized, as options include surgical (re)intervention, re-irradiation, chemotherapy - eventually combined. The aim of our study was to evaluate the efficacy and the toxicity of high-dose fractionated stereotactic reirradiation (fSRT) for recurrent HGG Materials and Methods From 2014 to 2019, 18 patients were reirradiated to a median dose of 36 Gy (range 25-36) in 6 fractions (5-6) on the 80% isodose after a primary post-operative chemoradiation. The overall survival (OS) was defined as the time between the primary surgery and death, progression free survival as the time between fSRT and disease progression according to RANO criteria. The Kaplan-Meier method was conducted. Results At the time of recurrence, the mean age was 55 years (21-75). Seven (39%) of the 18 patients underwent prior surgery. Median time from previous RT to fSRT was 13 months (5-84). The median OS after (initial) diagnosis was 37 months with an actuarial 1-,2- 3-years OS rate of 100%, 80% and 54%, respectively. After fSRT, the median OS was 8.6 months with an actuarial 6 and 12- months rate of 81 % IC 95% = (52- 93%) and 38 % IC95% = (14-61%) respectively. The PFS was 44 % at 6 months and 17 % at 1 year. On univariate analysis of prognostic factors, tumor volume (<10 mL) was found to significantly and positively influence OS (p= 0.02) but did not affect PFS (p = 0.6). Radionecrosis was identified in 16 % of the cases after the follow-up period. Conclusion FSRT remains a safe and effective treatment option for recurrent HGG. Additionally, treatment volume predicts survival and may be used in decision making in the salvage setting. Due to the usual diffuse extension of recurrent HGG, prospective trials are needed to improve targeting of the most proliferative sub volumes and allow dose escalation. PO-1036 Malignant Y. Takase 1 , M. Kawamura 2 , R. Nakahara 2 , J. Itoh 2 , Y. Oie 2 , M. Okumura 2 , T. Kamomae 1 , Y. Itoh 2 , T. Ono 2 , S. Naganawa 2 1 Nagoya university, Department of Radiology, nagoya, Japan; 2 Nagoya university, Department of Radiology, Nagoya, Japan Purpose or Objective The clinical target volume (CTV) of malignant glioma is determined by enlarging the gross tumor volume (GTV) by more than 15 mm and considering the anatomical structures. There are complex anatomical structures in the cranium, including the brain parenchyma, which is vulnerable to tumor invasion, and the cerebral sickle, which is less vulnerable to invasion. These complex structures may be the cause of the differences in CTV between planners of radiotherapy. Therefore, in this study, we evaluated whether there is a difference in CTV between planners due to anatomical structures. Materials and Methods Methods: Head MRI without gross brain tumor was used for this study. Six virtual GTVs with a diameter of 25 mm were set up in the vicinity of various anatomical structures on the brain parenchyma. Seven physicians with at least three years of radiotherapy experience participated in the study. The seven planners were asked to set a 20 mm area from each GTV as the CTV, taking into account the anatomical boundaries. To ensure that the interpretation of the brain parenchyma did not affect the results, all planners were given a common segmentation of the brain parenchyma. For each region, the degree of agreement between the CTVs of the seven planners was evaluated using the kappa value and the DICE coefficient. For each GTV, the CTV using geodetic distance was made and compared with the CTV of each planner. Results The kappa values for all six regions were greater than 0.97, which meant all were in good agreement. When we focused on the area where the CTV was expanded to the contralateral side through the corpus callosum, CTV only in the contralateral part was moderately consistent with the kappa value of 0.59. In the area around the region with low Kappa value, the CTV set by the planner tended to be smaller than the CTV segmentated with the geodetic distance. In particular, for slices in which the corpus callosum was not depicted in the horizontal section, the tendency for the planner's CTV to be smaller was remarkable. Conclusion Regardless of the anatomical boundaries where the GTVs were in close proximity, the overall agreement of the CTVs between the planners was good. However, when setting the CTV through the corpus callosum to the contralateral brain parenchyma, the agreement on the other side was moderate and tended to set a smaller CTV compared to the CTV using geodesic distance.

PO-1037 Hypofractionated Radiation Therapy For Non-Elderly Patients With Glioblastoma