Abstract Book

S676

ESTRO 37

values and dexamethasone (DEXA) medication were assessed. The analyzed central gustatory structures were: the solitary tract and nucleus, the ventral posteromedial nucleus of the thalamus, the sensory tongue area of the postcentral gyrus, the anterior part of the insula, the frontal operculum, the amygdala and the hypothalamus. These structures were delineated on magnetic resonance tomographies (MRIs) registered to planning-CTs in the Eclipse Treatment Planning System (Varian Medical Systems, Palo Alto, CA, USA). The dose-volume- histograms (DVHs) were analyzed for doses to the total GS and the ipsilateral and contralateral GS (Dmean, Dmax and V5 to V60). Further, Dmean and Dmax for normal brain tissue were recalculated as EQD2 using α/β=3. Reported are average ±standard deviation and median (range) values. DM patients were compared to non-DM patients. Results The median total dose to the PTV was 51.0 Gy (range:18.8–66.0). 8 patients (21.6%) were diagnosed with DM before radiotherapy. 20 patients (54.1%) had at least one BG value >120 mg/dl in the morning. The average BG of all patients during RTx was 155.9±47.1mg/dl – with a significant difference between diabetics(192.8±24.4mg/dl) and non- DM(145.7±39.5mg/dl) (p=0.010). 21 patients (56.8%) had solitary BG values over 200 mg/dl during RTx, out of these 14 (66.7%) were non-DM. The average change of the highest BG compared in all patients from before to during the therapy was 65.0±49.0mg/dl which means an increase of 41.9±48.6%. 30 patients (81.1%) took DEXA, without a significant difference in daily DEXA dose between DM and non-DM(p=0.29). The average Dmean to the total GS in all patients was 36.0±37.0Gy EQD2 with no significant difference between both patient groups (p=0.93), 41.3±11.1Gy to the ipsilateral GS and 31.1±9.1Gy to the contralateral GS. There was a tendency for a higher increase in BG values with more radiation dose to the total GS in the regression-model (b=1.9, R 2 =0.103, p=0.06) (Fig.1). A linear correlation between the increasing maximum BG and the daily (r=-0.16, p=0.41) and cumulative (r=0.17, p=0.41) DEXA medication could not be shown. Conclusion There is an increase in BG during radiotherapy for most patients, with a significant difference in diabetes patients. However, even non-diabetics reach single values over 200 mg/dl. DEXA medication did not significantly influence BG. There was a tendency for a higher increase in BG values with more radiation dose to the GS. EP-1217 SRS for Brain Metastases in Non Small Cell Lung Cancer patients: Evaluation of Prognostic Factors. D.S.L. Fonseca 1 , L. Ercolin 1 , D.F.S. Fantini 1 , D.N.S. Gomes 1 , A.A. Jacinto 1 , M.D. De Mattos 1 , A.B.B. Borges 1 1 Barretos Cancer Hospital, Radiotherapy, Barretos, Brazil Purpose or Objective To evaluate factors that can influence on Overall Survival (OS) in patients with Non Small Cell Lung Cancer (NSCLC) with brain metastases undergoing radiosurgery (SRS) and better understand the main cause of death in this population, which was treated in a Brazilian Cancer Center. Material and Methods Between October 2014 to January 2017 we founded 53 patients with NSCLC who underwent SRS for treatment of brain metastases and they were included in our trial. For the evaluation of OS were analyzed the KPS, the number of brain lesions, the use of Chemotherapy (CHT), the presence of extra-cranial disease, presence of EGFR / ALK mutation, SRS technique, treatment of primary site, control of systemic disease at time of SRS, Ds-GPA/SIR scores and use of Whole Brain Radiotherapy ( WBRT). For

the statistical analysis, we used Kaplan-Meier curves and for the multivariate analysis the Cox Regression model, with statistical significance of <0.05. Results Our results showed a median follow-up of the 7 months (0-31 months). The main histology was adenocarcinoma (81%). Around 17% had EGFR mutation and 5% mutation of the ALK. About of the Performance Status, 88.7% had KPS> 70% at the time of SRS. 39.6% presented with 1 lesion. The most treatment was SRS with single dose (50.9%) and 56.6% did not perform WBRT. The local failure (in the lesion treated) was 18.9% and the regional recurrence (new lesions) was 60,4%. Median overall survival was 9.0 months (0-31 months) with 15% of deaths due to neurological causes. The presence of EGFR mutation, DS-GPA / SIR and KPS> 90% influenced the SG (univariate analysis). In the multivariate analysis, the use of CHT, the EGFR mutation and Ds-GPA influenced OS. Conclusion We showed that the Index Prognostics Tools, as DS-GPA and SIR, can predict Overall Survival. However, the evaluation of the new variables as the use of chemotherapy and specially the EGFR mutation were important in OS. Despite the lower power of this trial (retrospective and few patients), our study suggest that the molecular informations could be incorporated on this tools. Beside that, the use of SRS upfront was safe and feasible in this population. EP-1218 Comparing diffusion weighted MRI with amino acid PET for re-irradiation in recurrent glioblastoma I. Popp 1 , S. Bott 1 , O. Oehlke 1 , M. Mix 2,3 , I. Mader 4 , P.T. Meyer 2,3 , H. Urbach 4 , A.L. Grosu 1,3,5 1 University Medical Center Freiburg, Department of Radiation Oncology, Freiburg im Breisgau, Germany 2 University Medical Center Freiburg, Department of Nuclear Medicine, Freiburg im Breisgau, Germany 3 German Cancer Consortium DKTK, Partner Site Freiburg, Freiburg im Breisgau, Germany 4 University Medical Center Freiburg, Department of Neuroradiology, Freiburg im Breisgau, Germany 5 German Cancer Research Center Heidelberg, Partner Site Freiburg, Freiburg im Breisgau, Germany Purpose or Objective The GTV definition for re-irradiation planning in recurrent glioblastoma is usually based on contrast enhanced MRI and thus on the blood brain barrier disruption, with a specificity of around 50%. This specificity can be increased to 90% by amino acid PET, identifying areas of increased cellular metabolism. Diffusion weighted MRI sequences (DWI) can reveal regions of high cellularity as surrogate for active tumor, as it has clearly been shown for a variety of tumor entities. The possible advantages of integrating DWI-MRI in the GTV definition of recurrent glioblastoma are until now unclear. The objective of the current explorative study was the evaluation of the localization and quality of diffusion restriction foci (GTV ADClow) in comparison to the GTVs defined according to the amino acid PET (GTV-PET). Material and Methods We retrospectively evaluated 41 patients, who received a fractionated stereotactic re-irradiation for recurrent glioblastoma between October 2010 and June 2015. All patients were initially treated by surgery followed by adjuvant radiochemotherapy (59.4 – 60 Gy, with parallel and adjuvant temozolomide). Progress was diagnosed either by repeated resection, biopsy or according to the RANO criteria in serial MRI examinations. The GTV-PET was generated automatically (tumor to background ratio = 1.8) and manually customized. The DWI data were acquired by single-shot spin-echo echo-planar imaging (diffusion gradients in three dimensions, b=0, 1000 s/mm2). The ADC values were calculated automatically