ESTRO 38 Abstract book

S672 ESTRO 38

P. Windisch 1 , M. Röhrich 2 , S. Regnery 1 , E. Tonndorf- Martini 1 , T. Held 1 , K. Lang 1 , D. Bernhardt 1 , S. Rieken 1 , U. Haberkorn 2 , J. Debus 1 , S. Adeberg 1 1 Heidelberg University Hospital, Department of Radiation Oncology, Heidelberg, Germany ; 2 Heidelberg University Hospital, Department of Nuclear Medicine, Heidelberg, Germany Purpose or Objective To assess target volume alterations when contouring is based on PET-CT with 68Ga-radiolabeled ligands of Fibroblast Activating Protein, FAPI-02 and FAPI-04, versus conventional pretreatment imaging with contrast Fourteen Glioblastoma (GBM) patients treated between 10/2017 and 08/2018 received an additional FAPI-PET prior to treatment. As one patient’s tumor, the only one with an IDH-mutation, showed very low FAPI-enhancement compared to healthy appearing normal tissue, analysis was carried out with the remaining 13 patients. Three different FAPI-GTVs were created using a 5-, 7- and 10- fold threshold of increased uptake compared to normal tissue (FAPIx5, FAPIx7, FAPIx10). Following GTV delineation, according to the EORTC guidelines, based on thresholds the volumes were corrected for false-positive enhancement. MRI-GTVs were created based on T1-weighted Gd- enhancement. Based on these GTVs, MRI-CTVs were created by adding a 20 mm margin. Applying another 3-5 mm margin to the MRI-CTV resulted in the MRI-PTV. Results GTVs based on MRI resulted in a median volume of 33.8 ml (range: 0.5 - 88.2 ml) while GTVs based on FAPIx5, x7 and x10 resulted in volumes of 49.6 ml, 33.7 ml and 23.8 ml respectively (ranges: 3.5 - 115.6 ml, 2.1 - 99.7 ml and 1.4 - 76.7 ml). FAPI-GTVs were significantly different from MRI-GTVs for FAPIx5 (p = 0.022) and FAPIx10 (p = 0.068) but not FAPIx7 (p = 0.95). As MRI- and FAPI-GTVs were not congruent, adding FAPI- GTVs to MRI-GTVs to create a combined MRI/FAPI-GTV resulted in an increase of the median volume of 17.7 ml, 8.9 ml and 3.9 ml (ranges: 2.9 - 49.3 ml, 1.6 - 32.4 ml and 0.8 - 18.7 ml) for FAPIx5, x7 and x10 respectively which corresponds to increases of the MRI-GTV by 55.6%, 27.3% and 17.2% (ranges: 20.0 - 586.3%, 12.4 - 311.8% and 3.9 - 164.7%). Changes to the MRI-GTVs caused by adding FAPI- GTVs were highly significant for all FAPI-PET thresholds (p < 0.001). Median MRI-PTV was 271.5 ml (112.4 - 490.9 ml).To determine whether FAPI-GTVs were included in the MRI- PTV, we calculated the overlap between both volumes. While in almost all patients PET-GTVs were completely included in the much larger MRI-PTVs as expected, one patient’s FAPI-GTV was not entirely covered by the MRI- PTV (30.9 ml, 21.3 ml and 13.5 ml of uncovered FAPI-GTV respectively). No complications occurred during either MRI or FAPI-PET. Conclusion FAPI-PET can provide additional insight into GBM spread compared to conventional pretreatment imaging and might therefore prove to be especially useful for treatment planning in radiation oncology but also for surgical interventions. Further research is warranted to assess appropriate thresholds for different patients and possible prognostic significance of tracer uptake. EP-1216 Single- versus 2-session Gamma Knife surgery for symptomatic midsize brain metastases S. Yomo 1 1 Aizawa Hospital, Division of Radiation Oncology- Aizawa Comprehensive Cancer Center, Matsumoto, Japan enhanced CT and MRI. Material and Methods

low risk of recurrence (0-150 points), and high risk (151- 350 points). Conclusion The Ayala nomogram shows validity in our population and allows users to integrate the information from different variables to provide precise risk stratification. The Rodrigues nomogram don´t show validity in our population. We have validated these nomograms in our Spanish population and we are considering new lines of study. EP-1214 Patterns of care: Treatment of glioblastoma in elderly patients C. Straube 1 , S. Antoni 2 , P. Schaffer 3 , J. Gempt 4 , C. Zimmer 5 , B. Meyer 6 , S.E. Combs 1 , F. Schmidt-Graf 2 1 Klinikum rechts der Isar- TU München, Department of Radiation Oncology, München, Germany ; 2 Klinikum rechts der Isar- TU München, Department of Neurology, München, Germany ; 3 MVZ InnMed Oberaudorf, Praxis für Strahlentherapie, Oberaudorf, Germany ; 4 Klinikum rechts der Isar- TU München, Department of Neuosurgery, München, Germany ; 5 Klinikum rechts der Isar- TU München, Department of Neuroradiology, München, Germany ; 6 Klinikum rechts der Isar- TU München, Department of Neurosurgery, München, Germany Purpose or Objective The optimal treatment for elderly glioblastoma patients is certainly unknown. We surveyed the pattern of care and the outcome of elderly glioblastoma (GBM) patients treated with varying adjuvant treatments, including radiotherapy (RT), temozolomide regimens, best Patients older than 65 years who underwent surgery for GBM between 2010 and 2016 at the Klinikum rechts der Isar, Munich, and who were treated with adjuvant therapies or underwent BSC were eligible. We analysed the median overall survival (mOS) by the Kaplan-Meier method and extracted prognostic factors with Cox regression modeling. Results 200 patients (106 female, 94 male) with a median age of 74 years (65-91 years) and a median Karnofsky performance status before adjuvant therapy of 80 (20- 100) met the criteria. MGMT promotor was evaluated in 109 patients and was methylated in 40 cases, 123 were IDH wild type and 77 unknown. 45 patients underwent biopsy (8 got resection afterward), 92 partial and 63 gross total resections. mOS of the entire cohort was 5.5 months. 76 patients underwent RT only (mOS 4.96 months), 10 received chemotherapy (CT) (mOS 9.90 months) and 54 RT and concomitant CT (mOS 13.73 months). Active adjuvant treatments with radiation therapy (HR 1.52 if no RT), MGMT promoter methylation (HR 0.49 if present), as well as the presence of paresis (HR 0.28 if no present), were significant prognostic factors in a multivariate model. 52 patients underwent treatment for progressive disease (mOS 13.77 months), including 23 patients undergoing second surgery, 14 patients (mOS 13.77 months) salvage RT or RCT and 15 patients (mOS 11,45 months) salvage CT. Patients undergoing BSC were generally in a worse physical condition (KPS 60 vs. KPS 80, p<0.001) and older (77.5y vs. 73.8y, p<0.001). Conclusion Our data underline the efficacy of active adjuvant treatment for GBM, optimally with RCT, even in an elderly cohort. As expected, the performance status after surgery is the most important predictor for BSC. EP-1215 Fibroblast Activating Protein specific PET for advanced target volume delineation in Glioblastoma supportive care (BSC). Material and Methods