ESTRO 38 Abstract book

S693 ESTRO 38

A total of 688 tumors were included, 530 treated with SRS and 158 treated with FSRT. The median radiographic and clinical follow up was 5 months and 7 months, respectively. All subsequent comparisons are given as SRS versus FSRT. The median tumor diameter was 0.82 cm and 1.65 cm (p<0.001). The KM estimate of 12-month LC among tumors ≤4cm was 97% and 95% (p=0.196). KM estimates of 12-month LC for tumors ≤2cm was 98% and 100% (p=0.352), >2 to ≤3cm was 91% and 100% (p=0.439), and >3 to ≤4cm was 80% and 74% (p=0.863). Additionally, KM estimate of 12-month freedom from significant CNS toxicity was 99% and 97% (p=0.308) overall and 89% and 100% (p=0.128) for tumors >3 to ≤4 cm. Only tumor volume (HR 1.141; p<0.001) was predictive of local failure in a multivariable analysis that included treatment type (HR 0.927; p=0.909).

22.8 months (range 1 to 71.4 months). The maximum number of BMs treated in a single GK radiosurgery per patient ranged from 4 to 18 (median 7), and the median number of total BMs treated over 1-5 courses of GK was 9 (range 4 to 37). The median overall survival for all patients was 33.4 months from the time of BM diagnosis. Overall survival was better in patients with fewer total number of lesions developed /treated (70.8 months for patients with 4-10 lesions, 34.5 months for those with 11-19 lesions and 27.7 months for those with >20 lesions), although this difference was not statistically significant (p=0.37). The overall survival was not affected by the number of GK courses or the maximum number of BMs treated in a single GK session. Patients who received CNS-active TKIs had longer median survival (53.4 months) than those who received non-CNS active TKIs (33.4 months) (p=0.8), and in this population receiving immunotherapy did not influence overall survival. Time to CNS progression after 1 st course of GK radiosurgery was 10.1 months, and 6.2 and 6.3months after the 2 nd and 3 rd course of GK, respectively (p=0.002). Six patients received WBRT after the initial GK (crude rate of 18.8%), and the median time to WBRT was 13.6 months (range 10 to 23.9). Radiosurgery was well tolerated. Symptomatic radiation necrosis (RN) was observed in 4 of 405 treated BMs (0.98%) or 4 of 32 patients (12.5%). Conclusion In patients with multiple BMs from EGFR-mutated and ALK- rearranged NSCLC, the optimal initial management remains controversial. The results from our current retrospective study support the use of initial radiosurgery without WBRT as an effective and safe strategy for metastatic NSCLC patients with driven mutation and more than 4 BMs. EP-1260 Outcomes of Local Control and CNS Toxicity with Single and Hypofractionated SRS for Brain Metastases M. Patel 1 , S.R. Marcrom 1 , R.A. Popple 1 , A.M. McDonald 1 , K.O. Riley 2 , B.L. Guthrie 2 , J.M. Markert 2 , C.D. Willey 1 , M. Bredel 1 , J.B. Fiveash 1 1 University of Alabama-Birmingham, Department of Radiation Oncology, Birmingham, USA ; 2 University of Alabama-Birmingham, Department of Neurosurgery, Birmingham, USA Purpose or Objective Radiosurgery is a standard management strategy of brain metastases in select patients. The role of single-fraction stereotactic radiosurgery (SRS) and hypofractionated stereotactic radiation therapy (FSRT) is currently evolving, with significant variability in utilization across institutions. We sought to retrospectively evaluate the impact of tumor size and treatment type (SRS vs. FSRT) on local control (LC) and CNS toxicity. Material and Methods All brain metastases from 2015-2017 treated with single- isocenter volumetric modulated arc therapy (VMAT) LINAC radiosurgery with a 6 DOF couch were captured. Tumors were included for analysis if they received SRS, FSRT, had no prior surgery or radiation, and had available follow-up imaging. The dose was prescribed to the edge of the GTV with no PTV margin. Local failure was defined as a 25% increase in tumor diameter on follow-up MRI or pathologic confirmation of tumor recurrence. Locally controlled tumors were censored at the time of last MRI follow-up, death, or whole brain radiation for distant brain failure. Significant CNS toxicity was per the RTOG definition as irreversible Grade 3 or higher toxicity. LC and CNS toxicity were evaluated by the Kaplan-Meier (KM) method with the log-rank test utilized to compare subgroups. A multivariable analysis (Cox Proportional Hazards model) of potential factors affecting LC was performed. Results

Conclusion SRS and FSRT appear to have a high rate of local tumor control with a low rate of significant CNS toxicity. LC and CNS toxicity outcomes appear to be driven by tumor volume without being significantly affected by treatment type. FSRT appears to offer a safe and effective alternative to SRS in both large and small tumors. In patients with multiple metastases in which a large target requires fractionation, utilizing the same dose and fractionation across all treated targets, regardless of target size, offers an opportunity for improved efficiency in treatment planning and treatment delivery without compromising outcomes. EP-1261 MRI-guided SABR of spinal metastases: comparison of Co-60 and linac treatments R. Llorente 1 , C. Takita 1 , R. Yechieli 1 , J.C. Ford 1 , K. Brown 1 , M. Samuels 1 , E.A. Mellon 1 1 University of Miami/Sylvester Comprehensive Cancer Center /Jackson Memorial Hospital, Radiation Oncology, Miami, USA Purpose or Objective MRI-guided radiotherapy systems visualize the spinal cord and cauda equina for spine SABR (stereotactic ablative radiotherapy) setup and tracking of motion during treatment. However, concerns have been raised regarding the dosimetric limitations of the first generation of Co-60 MRI-guided systems. In prior work, we demonstrated excellent visualization of spinal structures and tolerance for Co-60 MRI-guided spine treatments. In this work we demonstrate dosimetric improvements for linac based MRI guided SABR based on recalculation of the original Co-60 plans. Material and Methods Ten spinal metastases were treated with Co-60 MRI-guided radiotherapy at our institution from May 2016 to January 2017. Eight received a single 16 Gy dose and two received 30 and 24Gy in three fractions. Coverage was at least 80% This abstract is part of the media p ogramme and will be released o the day of its presentation