ESTRO 2021 Abstract Book

S885

ESTRO 2021

volume within the brain. Abbreviations: Gy = gray, D1cc = dose to 1 cc of respective volume, MU: monitor unit, PTV= Planning target volume Conclusion Measuring and optimizing doses to cornea, scalp, and hippocampus during brain radiosurgery may control respective doses and help improve QOL in respective aspects, especially if tumors are located away from critical structures such as optic neural structures. TLD verification can improve confidence in delivered doses, especially for superficial structures. PO-1065 Overall survival of patients treated with repeated courses of stereotactic radiation therapy L. Kuntz 1 , P. Meyer 2 , C. Le Fèvre 1 , C. Mazzara 2 , A. Keller 1 , C. Renaud 2 , G. Noel 1 , D. Antoni 3 1 ICANS, Radiation Therapy, Strasbourg, France; 2 ICANS, Medical Physics, Strasbourg, France; 3 ICANS, Radiation Therapy , Strasbourg, France Purpose or Objective Brain metastases (BM) are the most common intracranial malignant neoplasms in adults. Stereotactic radiation therapy (SRT) is one of the most widely used therapeutic modality to treat initial or recurrent BM. Main benefit of SRT is to delay whole brain radiotherapy (WBRT), which causes cognitive toxicities. One of the major SRT advantage is to be repeatable. The aim of this retrospective and monocentric study is to evaluate the efficacy of repeated courses SRT in terms of overall survival (OS) and neurological death-free survival. Materials and Methods We retrospectively reviewed the data of 184 patients treated between 2010 and 2020 with at least two courses of SRT without previous WBRT. Initial patient’s characteristics, patient’s characteristics at each SRT, brain metastases velocity (BMV), delay between SRT, OS and neurological death-free survival were collected and analyzed. Results 184 patients have been treated for 915 BM and received two to six SRT sessions. New SRT session was delivered for local (5.6%) or distant (94.4%) brain recurrence. Median number of BM treated per SRT was 1 (range:1-6; 95%CI 1.9-2.1), for a median of 4 BM treated during all sessions (range:2-19; CI95% 4.5-5.4). Median interval between SRT was 6.2 months (95%CI: 8.2-10.6), there was no significant difference in time interval between sessions. Median follow-up was 18.4 months (min-max 2.1-94.9; 95%CI 23.0-68.6). WHO and Diagnosis-Specific Graded Prognostic Assessment were stable between each session, Karnofsky Performans Status was significantly higher in first SRT (SRT1) than in the following SRT (p=0.02). Recursive Partitioning Analysis (RPA) was statistically improved in SRT1 than in second SRT (SRT2) (p<0.001), but this difference was not found in subsequent sessions. Number of BM between each SRT was not statistically different but there was a tendency for more metastases in SRT1 (p=0.06). Six-, 12- and 24-month OS were 91% (95%CI 88-96), 70% (95%CI 64-77), 38% (95%CI 32-45), respectively. In multivariate analysis, OS was statistically related to the number of SRT (p<0.01), RPA (p=0.01), salvage WBRT (p=0.01), final BMV (p<0.01) (figure1). Ultimately salvage WBRT was delivered in 18.9% of patients after a mean of 2 SRT sessions but death occurred after a median of 7.0 months (95%CI 5.9- 11.7) after WBRT. Patients with kidney or breast cancer, WHO 0-1 at SRT1/SRT2, high RPA at SRT1/SRT2, extracerebral stability at SRT1, no systemic treatment at SRT1/SRT2, were most likely to benefit from more than 3 SRT and therefore to benefit from the avoiding of WBRT (Table1). Also, at the time of SRT3 long-survivor patients were in better general condition than short-survivor patients were at the moment of SRT1. BMV at SRT2 was predictive of OS (p<0.001).