ESTRO 38 Abstract book

S682 ESTRO 38

of centroids (MDC) from the reference centroid was 0.27 mm. Mean DICE was 0.90 (range 0.85-0.92).

7–202 months). They mainly had Koos’s stage II (N=37) and stage III (N=22) VS treated with 3 x 7 Gy (N=47) or 5 x 5 Gy (N=17) dose protocols. We reached a median follow-up of 28 months (range: 7–80 months). The treatment was well tolerated with only four grade III acute toxicities (loss of balance for all this four, associated with a headache and tinnitus for one of these) and no grade IV or V. All of them were resolute with symptomatic treatment (corticosteroids). As late toxicity grade III or more we noticed one Bell’s palsy, which was resolutive with symptomatic treatment and one trigeminal neuralgia occurring 16 months after radiation therapy. More than 50% of the population had a stable hearing after treatment. Thirty three patients presented VS necrosis, appearing with a median delay of 7 months (range: 3–16 months). On the 64 patients we noted only 2 recurrences leading to a control rate of 97.1%. These recurrences occurred after 30 and 40 months. These two patients benefited from salvage surgery and were free of any recurrence at the last consultation. They both were treated with upfront radiotherapy with a 3- fractions protocol on Koos’s stage III VS. With few recurrences, the analyze of recurrence factors was narrow but we found as clinical factors influencing the risk of relapse the maximum size of more than 20 mm (p=0.003) and on dosimetrical factors a conformal index <1.1 (p <0.0001). Conclusion Our results suggest that our HFSRT using 21 Gy in 3 fractions or 25 Gy in 5 fractions is a well-tolerated and effective regimen. EP-1238 Inter-observer variability in target delineation for brain metastases in stereotactic radiotherapy S. Di Biase 1 , F. Patani 1 , D. Fasciolo 1 , C. Rosa 1 , C. Di Carlo 1 , A. Allajbej 1 , L. Gasparini 1 , A. Delli Pizzi 2 , M. Trignani 1 , L. Caravatta 1 , D. Genovesi 1 1 "SS. Annunziata" Hospital, Department of Radiotherapy - G. D'annunzio University, Chieti, Italy ; 2 "SS. Annunziata" Hospital, Department of Radiology - G. D'annunzio University, Chieti, Italy Purpose or Objective In stereotactic radiotherapy for brain metastases (BM), target delineation is based on MRI co-registered with simulation-CT and involves a well-defined volume represented by enhancing area on T1 weighted sequences, defined as gross tumor volume (GTV). In these conditions, a limited inter-observer variability (IOV) could be expected. However, studies found significant differences in shape, size, and location in radiosurgery targets showing also an impact in plan conformity. The aim of this study was to quantify the IOV of target delineation in BM. Material and Methods A case of brain metastasis was proposed to 11 Radiation Oncologists (RO) of our Department. RO delineated GTV on T1 MRI co-registered with planning-CT using MIM Maestro software (MIM Software Inc). GTVs were analyzed and compared to a benchmark volume performed by a Neuroradiologist and a RO with expertise in brain radiotherapy jointly. The analysis of the obtained volumes was based on: DICE similarity index, Jaccard index, mean distance to agreement (MDA), Hausdorff distance, GTVs volumes, common volume (AV100), encompassing volume and centroids. Results The lesion had a well-defined small spheric appearance with a homogeneous contrast enhancement and without edema. The reference GTV volume was 1.04 cc. Mean participants volume was 1.14 cc (range 0.88-1.34 cc). Volumetric analysis of GTVs and similarity and discordance analysis of targets are reported in Table 1. Mean distance

Conclusion In contrast to studies reporting IOV in SRT, our work pointed out a little variability. The peculiar appearance of the considered metastasis could have contributed to the agreement obtained. Anyway, our result need to be confirmed taking into account metastases of heterogeneous appearance and in this direction further evaluations are ongoing. Moreover, it should be evaluated the dosimetric implications of small IOV, due to the high dose ad rapid fall-off of stereotactic radiotherapy. EP-1239 Role of perilesional edema in patients with glioblastoma undergoing adjuvant chemo-radiation V. Nardone 1 , M. Gaetano 1 , M.G. Calvanese 1 , A. Di Biase 1 , I. Nuzzo 2 , L. Di Rienzo 1 , I. D'Onofrio 1 , M.G. Calvanese 1 , P. Pastina 3 , C. Guida 1 , C. Vitale 2 , P. Tini 3 , S. Cappabianca 2 , L. Pirtoli 3 1 Ospedale del Mare, Radiation Therapy Unit, Napoli, Italy ; 2 University of Campania “Luigi Vanvitelli”, Department of Precision Medicine, Naples, Italy ; 3 Unit of Radiation Therapy, University Hospital of Siena, Siena, Italy Purpose or Objective The present study was designed to evaluate the role of perilesional edema calculated on pre-surgery MRI as a prognostic factor in patients with glioblastoma undergoing adjuvant chemo-radiation. Material and Methods We performed a retrospective analysis on 140 patients with glioblastoma undergone surgery and adjuvant chemo- radiotherapy (C-RT) between January 2010 and December 2015. Both perilesional edema (PE) and gross tumor volume (GTV) were contoured on MRI. We also calculated PE/GTV ratio, and PE + GTV (TV). Cut-offs were calculated with the software X-Tile. We correlated progression free survival (PFS) and overall survival (OS) with these parameters and the known prognosticators (surgery, KPS, age, MGMT status), with Kaplan Meier method (univariate) and Cox Regression Analysis (multivariate). Results A total of 140 patients were included in the analysis (mean age 61 years, median 64 years, range 35-84 years, 91 males and 49 females). Ninety-nine (70%) developed recurrence, whereas 90 patients (64%) died during the follow up. We weren’t able to calculate a cut-off value for both PE and GTV, but the found a cut-off for the PE/GTV ratio (>4) for OS survival analysis. The significant parameters at univariate analysis were KPS (p<0,001), surgery (p<0,001) and MGMT status (p<0,001)