Abstract Book

S107

ESTRO 37

The PTV averaged 70 cc (range: 10-206) for single fraction cases and 122 cc (range: 18- 377) for three fractions cases. The total calculation time required to generate an ECHO plan was on average 147 minutes (range: 21-541), depending mainly on the size of the PTV. 105 plans (71%) were acceptable after a single ECHO run, 29 plans (20%) required 2 nd run (mainly to adjust one parameter that did not meet the clinical criteria) while 13 plans (9%) required 3 rd run (to accommodate specific dosimetric requests from the physician for those particular patients). All plans met or exceeded institutional clinical criteria (Table 1). For the 24Gy single fraction cases, PTV V100% was 93.3% ± 3.4%, PTV V95% was 97.6% ± 2.1%, minimum CTV dose was 17.4 ± 3.6 Gy, and minimum GTV dose was 22.1 ± 3.8 Gy. For the 27Gy three fractions cases, PTV V100% was 94.3% ± 2.5%, PTV V95% was 99.0% ± 1.1%, minimum CTV dose was 22.4 ± 2.5 Gy, and minimum GTV dose was 24.5 ± 3.6 Gy. Dose to spinal cord and other OARs is presented in the table. Better planning efficiency was achieved with ECHO. As a result, we have instituted an accelerated planning schedule for paraspinal patients planned with ECHO. Table 1: ECHO plans dosimetric metrics compared with clinical criteria

Conclusion EL and HA are able to provide clinically acceptable plans for SRS of single and mBM; however, the number of brain metastases treated with one single isocenter appears to influence the performance of the two automated planning solutions differently. Automation reduces planning time substantially compared to manual planning: effective planning time is less than 5 minutes for EL and HA provides a gain in effective planning time due to the automatic selection of beam geometry, but the optimization process still requires comparable planning time as CP. PV-0198 Clinical experience of automated SBRT paraspinal planning with constrained hierarchical optimization L. Hong 1 , M. Zarepisheh 1 , Y. Zhou 1 , Y. Yamada 2 , J. Mechalakos 1 , G. Mageras 1 , M. Hunt 1 , J. Deasy 1 1 Memorial Sloan-Kettering Cancer Center, Medical Physics, New York- NY, USA 2 Memorial Sloan-Kettering Cancer Center, Radiation Oncology, New York- NY, USA Purpose or Objective To present our initial clinical experience with automated treatment planning using expedited constrained hierarchical optimization (ECHO) to improve plan efficiency and quality for SBRT paraspinal plans. Material and Methods 124 patients underwent SBRT radiotherapy with 147 different paraspinal plans from April to Octob er, 2017. 52 patients (60 plans) received 24Gy in a single fraction, and 72 patients (87 plans) received 27Gy in three fractions. The tumor locations were in various parts of the spine with 25, 84, 33 and 5 in C, T, L and S spine respectively. After physician segmentation, a template plan using 9 IMRT fields was set up and sent to ECHO through an Eclipse API plug-in. ECHO would produce a Pareto optimal plan that satisfies hard constrains with best target volume coverage and lowest normal tissue doses. Upon ECHO completion, the planner received an email indicating the plan was ready for review and evaluation against the clinical planning criteria. The plan was accepted if all of the clinical criteria were met. If the plan did not meet all clinical criteria, the planner could adjust a limited number of parameters and initiate another round with ECHO. Results

Conclusion We implemented ECHO in our clinic for automated SBRT paraspinal planning. ECHO enhanced efficiency of planning and shortened the time between simulation and treatment in our clinic. PV-0199 Influence of tumor infiltration on treatment outcome in Gamma Knife Radiosurgery: a modeling study M. Lazzeroni 1,2 , Z. Khazraei Manesh 3 , H. Sandström 2 , P. Barsoum 4 , I. Toma-Dasu 1,2 1 Karolinska Institutet, Medical Radiation Physics- Dept. Oncology-Pathology, Stockholm, Sweden 2 Stockholm University, Medical Radiation Physics- Dept. of Physics, Stockholm, Sweden 3 Royal Institute of Technology, School of Technology and Health, Stockholm, Sweden 4 Karolinska University Hospital, Dept. of Medical Physics, Stockholm, Sweden Purpose or Objective High grade gliomas are brain tumors with widespread infiltration into the surrounding tissues. Gamma Knife TM