ESTRO 36 Abstract Book

S326 ESTRO 36 _______________________________________________________________________________________________

duration of RT (74.4 vs. 41.2 days, p = .005), but did not delay RT (37.8 vs. 31.4 days, p = .407) or affect MS (25.5 vs. 16.3 months, p = .562). Vascular events did not significantly delay RT (36.7 vs. 31.3 days, p = .188), prolong the duration of RT (33.7 vs. 42.3 days, p = .127), or affect MS (13.1 vs. 16.7 months, p = .915). Conclusion More aggressive treatment strategies such as reoperation should be investigated in patients with interim PD for survival improvement. Since a significant portion of patients demonstrate interim PD, pre-RT MRI is essential for accurate target delineation. Non-PD events before RT may affect the duration of RT, but not survival. PO-0624 Gammaknife Radiosurgery in patients receiving anticancer immunotherapy: efficacy and safety D. Greto 1 , M. Loi 1 , S. Scoccianti 1 , M. Baki 1 , I. Desideri 1 , L. Bordi 2 , P. Bono 2 , I. Meattini 1 , P. Bonomo 1 , F. Terziani 1 , V. Carfora 1 , L. Livi 1 1 Azienda Ospedaliero Universitaria Careggi- Università di Firenze, Radiotherapy, Firenze, Italy 2 Azienda Ospedaliero Universitaria Careggi, Neurosurgery, Firenze, Italy Purpose or Objective Radiosurgery is the treatment of choice for brain metastasis (BMs) in patients with controlled extracranial disease under systemic therapy. In recent years, immune checkpoint inhibitors emerged as a valuable option resulting in unprecedented long-lasting remissions and integrated clinical management in patients affected by metastatic melanoma, lung and kidney cancer. The aim of the study is to assess the safety and the synergistic activity of concurrent immune checkpoint inhibitors and GammaKnife Radio Surgery (GKRS) in a retrospective cohort of patients. Material and Methods We retrospectively reviewed patients undergoing anti CTLA4 and/or anti PDL1 immunotherapy treated with GKRS at our Institution from January 2014 to March 2016 for BMs. Radiosurgery was delivered within 6 months from the last immunotherapy administration. Response to radiosurgery was evaluated according to RECIST criteria by magnetic resonance (MRI) performed at 45 days, three and six months after procedure. Clinical outcome and toxicity were analyzed and correlated to patient characteristics and treatment modalities. Results Twelve patients (5 melanoma, 6 lung, 1 kidney) were eligible for a total number of 61 treated lesions for a median number of 3 lesions per patient (1-16). Median age was 55 years (32-77) ; median GPA was 2 (1-4). GKRS consisted of a single session in all patients for a median dose of 21 Gy (15-24); median treatment volume was 9.45 cm 3 (1.75-220.35). Immunotherapy consisted of Anti CTLA4 (Ipilimumab) in 3 patients, Anti PDL1 (Nivolumab) in 8 patients or both in 1 case, corresponding to 8 (13.1%), 37 (60.7%), and 16 (26.2%) lesions, respectively. No acute neurotoxicity occurred after GKRS. MRI at 45 days showed complete response, partial response and stable disease in 7 (11.5%), 22 (36.1%) and 32 (52.4%) lesions. MRI at 6 months showed progression of treated lesions in 4 (6.6%) cases; five (41.7%) patients experienced distant brain failure. At statistical analysis, local control at 6 months was correlated only to BRAF mutation (p=0.029). At a median follow up of 9.6 months (6.3-30.6) we recorded one death due to brain progression while 5 patients died for extracranial disease; radionecrosis occurred in one case. Conclusion The association of immune checkpoint inhibitors and GKRS is feasible and did not result in severe toxicity. Enhanced local control in GKRS treated BRAF mutated melanoma BMs might result from defective DNA-repair or by

diagnostics, tumor stadium, risk group, localization, initial symptoms, treatment modalities, toxicities of the treatments, and survival outcomes were investigated. Results A total of 22 patients were identified. Median age was 30.2 years (16.5-45.6 years). All of the patients had de novo medulloblastoma and good Karnofsky performance status. The most frequent histologies were classic (n=10) and desmoplastic (n=9). 2 patients were high risk and 19 standard risk patients. After tumor resection, all the patients received craniospinal irradiation with a median dose of 35 Gy and a boost to the posterior fossa with a median dose of 19.8 Gy. Simultaneous chemotherapy with vincristine q1w was given to 20 patients. Sequential chemotherapy with cisplatin/vincristine/CCNU was applied in 15 patients. 3 patients with recurrent disease were detected. The median overall survival has not been reached after a median follow up of 92 months. Estimated 5-year OS was 88% and 10-year OS was 80% respectively. Estimated 5-year and 10-year-locoregional control was 81% respectively. In univariate analysis shorter locoregional treatment time (interval of tumor resection until end of irradiation) was significantly associated with improved OS (p= 0.031) and improved locoregional control (p= 0.049). Furthermore anaplastic histology was associated with worse OS (p<0.01). The most common side effects were hematological toxicities, with only 2 patients having had acute grade 3 toxicities. Conclusion The combined treatment showed a good outcome in patients with adult medulloblastoma. The therapy has been well tolerated. However, further investigations are needed regarding possible prognostic factors. PO-0623 Impact of Interim Events during the Surgery- to-Radiotherapy Interval in Newly Diagnosed Glioblastoma C.W. Wee 1 , E. Kim 1 , I.H. Kim 1 , T.M. Kim 2 , C.K. Park 3 , J.W. Kim 3 , S.H. Choi 4 1 Seoul National University College of Medicine, Department of Radiation Oncology, Seoul, Korea Republic of 2 Seoul National University College of Medicine, Department of Internal Medicine, Seoul, Korea Republic of 3 Seoul National University College of Medicine, Department of Neurosurgery, Seoul, Korea Republic of 4 Seoul National University College of Medicine, Department of Radiology, Seoul, Korea Republic of Purpose or Objective To investigate the impact of interim disease progression (PD) and other events occurring during the surgery-to- radiotherapy interval for treatment and outcome in glioblastoma. Material and Methods A total of 222 patients were planned for radiotherapy (RT) and 173 of them were evaluable for the presence of interim PD by 2 separate MRIs. The size criteria from the updated Response Assessment in Neuro-Oncology criteria was adopted. Results Of the 222 patients, 1 patient couldn’t start the planned RT due to pulmonary complication. All patients with interim events completed the planned RT. Forty-three (24.9%) patients experienced interim PD, and their median survival (MS) was significantly shorter than patients without PD in univariate (13.9 vs. 18.4 months, p = .002) and multivariate analysis [ p = .001, HR 2.018 (95% CI, 1.321–3.082)]. Interim non-PD events were observed in 24 (10.9%) patients but did not affect MS compared to those without non-PD events (20.2 vs. 16.4 months, p = 0.414). Vascular events and infection of the central nervous system occurred in 5 (2.3%) and 9 (4.1%) patients, respectively. Infection significantly prolonged the mean