ESTRO 2020 Abstract Book

S470 ESTRO 2020

Gy. Prognostic factors were identified using the Cox regression method. From these, the first version of the MeniScore for outcome prediction was created. The following parameters were evaluated: resection status, time from surgery to RT, planning volume (PTV), age, Karnofsky-Performance-Score (KPS), and gender. Results Median follow-up was 3.3 years (95%-KI: 3.2-5.6), median OS was 12.0 years, and median PFS 6.8 years. The parameters resection status, PTV, and KPS showed a significant impact on PFS in the univariate analysis. For the continuous variables, cut off values were determined with ROC analysis. The grouped variables PTV, age, and KPS were also significant. Based on the prognostic factors, a scoring scheme was created to determine the MeniScore, see Table 1. A high score means a good prognosis. With p=0.033, the MeniScore was significant regarding PFS. 0 points 0.5 points 1 point Resections status complete incomplete PTV [ml] >100 >50-100 <= 50 Age [years] > 65 <= 65 KPS [%] <= 70 80-90 100 Conclusion The prediction of local progression using the novel MeniScore shows a significant correlation. Further validation is necessary and currently underway. We initiated an international multicenter trial to gather a larger patient cohort and continue analyses. In the future, the MeniScore could be included in treatment recommendations to foster patient individual cancer care. PO-0875 Comparison of risk scores for re-irradiation of patients with recurrent high-grade gliomas (HGG) H. Scholz 1 , S.E. Combs 1 , K.A. Kessel 1 1 Technical University of Munich TUM, Department of Radiation Oncology, Munich, Germany Purpose or Objective This study aims to evaluate the differences between published scoring systems that predict the survival of patients with recurrent gliomas. We investigate the complexity of the score assessments, used parameters, and accuracy regarding prognosis and patient stratification. Material and Methods We identified 244 patients with recurrent high-grade gliomas (HGG) at the department of radiation oncology. All received re-irradiation between 2002 and 2019. Score assessment was done for five recently published risk scores [Kessel et al. Physica Medica 67 (2019) 20-26]. The Combs- , Müller-, Niyazi-, Krauze-, and Kessel-score. Two are extension and modifications of an already existing score. Results The evaluation is ongoing. The first analysis demonstrates that the determined risk group for one patient differed hugely among the five scoring systems. Only in 46 of 244 (18.9%) patients the calculated risk group was identical. The risk groups of the remaining 198 patients varied widely. We found that more than half of the patients calculated with the Combs- or Müller-score resulted in a poor risk group.For the Combs-score, a WHO-grade of IV has immense influence, and patients cannot reach a good prognosis. In the Müller-score, the parameters age and WHO-grade are weighted heavily as patients older than 50 years with a WHO IV glioma will always result in a poor risk group.When using the Niyazi- and Kessel-score, most of our patients resulted in an intermediate risk group. For the Niyazi-score, the Karnofsky-Performance-Score (KPS) is heavily weighted. If the KPS is <70% a good risk group cannot be reached. If the KPS is ≥70%, patients with an age between 53 to 107 years will always result in an

intermediate risk group. The Kessel-score uses six parameters; hence, the influence of the WHO-grade is not as strong, and the chances of reaching an intermediate or even good risk group are higher than with the other scores. Conclusion For HGG, the clinical situation remains demanding. The use of scores for decision making is a valuable tool and will be increasingly used as we enter the era of patient individual cancer care. The current results show that the scoring systems differ extremely regarding used parameters and calculational complexity. More than 80% of patients were assigned to different risk groups depending on the used score scheme. Further analyses will include a correlation with survival. Eventually, we will identify the score that is best in predicting the individual risk of each patient. PO-0876 Stereotactic radiosurgery with single- isocenter dynamic arc therapy of multiple brain metastases R. Bodensohn 1 , A. Kaempfel 1 , D.F. Fleischmann 1 , J. Hofmaier 1 , I. Hadi 1 , S. Garny 1 , M. Reiner 1 , S. Corradini 1 , R. Forbrig 2 , N. Thon 3 , C. Belka 1 , M. Niyazi 1 1 University Hospital- LMU Munich, Department of Radiation Oncology, Munich, Germany ; 2 University Hospital- LMU Munich, Department of Radiology, Munich, Germany ; 3 University Hospital- LMU Munich, Department of Neurosurgery, Munich, Germany Purpose or Objective With progressively advancing systemic treatment options, overall survival for patients with brain metastases is continuously improving. For this reason, it is important to find alternatives to whole-brain radiotherapy for patients with multiple brain metastases, as long-term neurocognitive decline is becoming more apparent. Single- isocenter dynamic conformal arc (SIDCA) therapy displays a possibility to deliver stereotactic radiosurgery (SRS) to patients with multiple metastases in a safe and efficient way. This study analyzes the safety and feasibility of SRS with SIDCA for patients with multiple brain metastases. Material and Methods Patients with multiple brain metastases except small-cell lung cancer (SCLC) were selected to receive stereotactic radiosurgery using SIDCA. All patients, with two or more brain metastases, who received SRS with this technique between November 2017 and June 2019 were included. The Multiple Brain Mets ® SRS application by Brainlab was used for delineation and radiotherapy planning. Follow-up was performed every three months, including a clinical and radiological examination with cerebral magnet resonance imaging (MRI). The data was analyzed by using general descriptive statistics or Kaplan-Meier methodology. Results 65 patients were included in this analysis. The most frequent entity is non-small-cell lung cancer (NSCLC), followed by melanoma and breast cancer. 254 lesions have been treated with this technique so far, with a median of 3 lesions (range 2–12) per patient per therapy session. Median follow-up was 8 months (95% KI 6.7–9.3); follow-up is however still ongoing. Median overall survival was 13 months (95% KI 7.5–18.5), median progression-free survival 9 months (95% KI 5.7–12.3). One patient (1.5%) reacted with tachycardia CTCAE grade 1 during treatment; other than that, there were no acute adverse effects during or directly following treatment. 7 patients (10.8%) showed subacute adverse effects CTCAE grade 1 (all therapy related). 20 patients (30.8%) experienced late adverse effects: 13 patients (20.0%) CTCAE grade 1 (9 of them therapy related (13.8%)), 2 patients (3.1%) CTCAE grade 2 (one of them therapy related (1.5%)), 4 patients (6.2%) CTCAE grade 3 (none of them therapy related). One lesion (0.4%) was a histopathologically verified progressive tumor