Abstract Book

S686

ESTRO 37

Purpose or Objective To validate the Graded Prognostic Assessment (GPA) in a cohort of patients with brain metastasis treated with Stereotactic radiosurgery (SRS). Material and Methods We retrospectively reviewed our centre experience treating patients with SRS for brain metastasis between January 2011 and December 2016. We analyzed tumour type, RT dose, whether previous whole brain radiotherapy (WBRT) had been administrated or not, immobilization devices, prognostic factors recognized by the GPA, overall survival (OS), and local progression free survival (LPFS). We used the GPA app to calculate the estimated survival expectancy (ES) for each case. We used Kaplan-Meier to compare the ES with the OS, among all patients and adjusted by tumor type. We used the log- rank test to measure whether significant survival differences were present. Results Seventy eight patients underwent SRS for brain metastasis. In 7 patients it was no possible to calculate the GPA index because the tumour type was not one of the indexed ones. The most common primary tumours were lung adenocarcinomas (42.3%), followed by gastrointestinal cancers (18.3%) and breast cancers (15.5%). The medium dose was 19.1 Gy (14 Gy-25Gy). An optically-guided frameless system was employed in 59.7% patients and a conventional stereotactic head frame was employed in 40.3%. WBRT was administrated in 55% of patients. Neither the immobilization device nor the WBRT impacted in the OS. With a median follow up of 12.1 months, the median OS was 12.1 months and the median GPA ES was 13 months. The OS at 6, 12 and 24 months were 76.1%, 53.4% and 28.1%, respectively. The GPA ES at 6, 12 and 24 months were 77.4%, 64.5% and 29.2%, respectively. Not significant differences were found neither in the global analyses, nor in the subgroups analyses. The LPFS was 10 months. Conclusion Our results confirm that GPA is a useful tool that accurately predicts the life expectancy of patients with brain metastasis treated with SRS in our environment. SRS achieves good rates of local control in patients with brain metastasis. The administration of WBRT and the employement of frameless devices did not impact on the OS. EP-1239 Ph II randomized trial comparing cognitive outcomes of proton vs. photon radiation for glioblastoma C. Chung 1 , P. Brown 2 , D. Liu 1 , D. Grosshans 1 , S. Dibaj 1 , N. Guha-Thakurta 1 , J. Li 1 , S. McGovern 1 , M. McAleer 1 , A. Ghia 1 , A. Paulino 1 , E. Sulman 1 , M. Penas-Prado 1 , J. De Groot 1 , A. Heimberger 1 , J. Wang 1 , T. Armstrong 3 , M. Gilbert 3 , A. Mahajan 2 , J. Wefel 1 1 UT MD Anderson Cancer Center, Radiation Oncology, Houston- TX, USA 2 Mayo Clinic, Radiation Oncology, Rochester, USA 3 National Cancer Institute, Neuro-Oncology Branch, Bethesda, USA Purpose or Objective The primary objective of this study was to determine if time to cognitive failure is longer in patients with newly diagnosed glioblastoma treated with proton therapy compared to photon intensity modulated radiotherapy (IMRT). Secondary objectives included evaluation of the interaction between treatment type and brain dosimetry and tumor-related characteristics on cognitive function. Material and Methods Eligible patients were randomized to proton therapy or IMRT. Randomization was stratified for RPA class (III-IV vs. V), Mini Mental Status Examination score (21-26 vs. 27-30) and age (< 65 vs. 65 or older). The primary endpoint was time to cognitive failure on any of the 6

Clinical Trial Battery (CTB) outcomes (HVLT-R, TMT, COWA) with failure defined as a decline that met or exceeded the reliable change index (RCI). Individual profiles of each cognitive test score and the CTB Composite score were plotted as a function of time for each patient. Loess method was applied to fit a smoothed line to the data excluding outliers. Results A total of 90 patients were enrolled (of which 58 had evaluable cognitive measures (25 protons; 33 photons). These patients had a median follow-up of 14.5 months (Range: 0.1 - 32.1 months). There was no significant difference in time to cognitive failure on any cognitive test between treatment arms (p= 0.47). Similarly, no difference was noted in time to cognitive failure when measured by CTB Composite score (p=0.25). In order to adjust for unbalanced sample sizes in each arm, we simulated a Kaplan-Meier plot that duplicated the 5 top cognitive performing patients in the proton arm, which denoted no significant difference between arms even despite simulation of 5 additional high performing proton patients (p=0.62). Cognitive failure rate at 4 months and 8 months following baseline evaluation was estimated 0.04 and 0.36 for patients treated with protons and 0.03 and 0.22 for patients treated with photons. Loess plots suggesting change over time are being investigated with repeated measures analysis. There was more acute grade 2 or greater toxicities in patients who received IMRT (n=21) vs. protons (n=9). Conclusion Preliminary results of this study suggest proton therapy is not associated with a delay in time to cognitive failure but did reduce overall toxicity. Additional evaluation of the impact of tumor location and volume, radiation dosimetry, and tumor molecular subtypes on cognition is being completed. Larger randomized trials are needed to determine the impact of proton vs. photon radiotherapy on glioblastoma tumor control and survival. EP-1240 Oral mucositis, pain, IV opioid use and in- patient days after TBI delivered in 3 vs 6 fractions L.S. Fog 1 , H. Sengeløv 2 , M. Schmidt 2 , P.M. Petersen 1 , L. Specht 1 1 Rigshospitalet- University of Copenhagen, Dept. of Oncology, Copenhagen, Denmark 2 Rigshospitalet- University of Copenhagen, Dept. of Haematology, Copenhagen, Denmark Purpose or Objective This study investigated the grade of oral mucositis, pain, the use of IV opioids and enteral and parenteral nutrition and days from allogeneic bone marrow transplant until discharge from hospital for 65 consecutive total body irradiation (TBI) patients. Material and Methods This retrospective study included 65 consecutive patients receiving 12 Gy of TBI in our clinic. The first 32 patients (the TBI3 patients) received 12 Gy in 3 fractions over 3 days with lung shielding on day 2. The last 33 patients 12 Gy in 6 fractions of 2 Gy with lung shielding on fractions 1,3 and 5 and with at least 6 hours between fractions (the TBI6 patients). The TBI3/TBI6 patients had AML (13/18), ALL (13/10). MDS (3/3), CML (1/3) and Non-Hodgkin’s lymphoma (1/0). Condition consisted of cyclophosphamide (16/25), VP16 (15/8) or cyclophosphamide and fludarabine (0/1) in addition to TBI. Data was recorded for 21 days following bone marrow transplant (BMT). Electronic Poster: Clinical track: Haematology