ESTRO 2025 - Abstract Book

S658

Clinical - CNS

ESTRO 2025

Results: No difference in mOS (p=0 .2) (20.6 months (%95 CI 15.6 – 31.2) vs 18.4 months (%95 CI 14.9 – 19.9)) or in mPFS (p = 0.183) (10.1 months (95% CI 9.4 - 12.5) vs 9.3 months (95% CI 7.5 - 10.3)) was observed respectively between mGB and hGB. However, mOS of mGB without CE and of hGB were respectively 31.2 months (95%CI 19.6 - NR) and 18 months (14.9 - 19.9) (HR 0.45, 95%CI 0.21 - 0.99). The models developed with ML and DP enable us to highlight differences in features between mGB without CE and LGG, with a best-performing ROC AUC of 0.85 for DL. Conclusion: Our study shows similar survival rates for patients with mGB and those with hGB. However, mGB without contrast enhancing seem to achieve a better outcome in comparison with the other glioblastomas when treated with a Stupp protocol, suggesting that stratification according to this absence of CE should be considered in randomized clinical trials. In addition, radiomics have a great potential to add important diagnostic information to many highly relevant clinical questions in brain tumor patients including mGB.

Keywords: Glioblastoma, clinical imaging, deep learning

528

Proffered Paper Tumor Treating Fields and Stereotactic Radiosurgery Based on FET-PET for Recurrent Glioblastoma: A Prospective Trial (TaRrGeT Trial, NCT04671459) Maciej Harat 1,2 , Maciej Blok 2,1 , Izabela Zarębska 1 , Michał Marjański 2 , Marek Harat 3 , Izabela Miechowicz 4 , Magdalena Adamczak Sobczak 1 , Bogdan Małkowski 5 1 Department of Neurooncology and Radiosurgery, Franciszek Lukaszczyk Memorial Oncology Center, Bydgoszcz, Poland. 2 Department of Clinical Medicine, Jan and Jedrzej Sniadecki University of Science and Technology, Bydgoszcz, Poland. 3 Department of Neurosurgery, 10th Military Hospital, Bydgoszcz, Poland. 4 Department of Computer Science and Statistics, Poznan University of Medical Sciences, Poznań, Poland. 5 Department of Nuclear Medicine, Franciszek Lukaszczyk Memorial Oncology Center, Bydgoszcz, Poland Purpose/Objective: Recurrent glioblastoma poses a substantial therapeutic challenge due to its radioresistance and invasive growth patterns, which are often insufficiently visualized on MRI. Tumor Treating Fields (TTFields) have shown efficacy in treating recurrent glioblastoma, while stereotactic radiosurgery (SRS) guided by ¹⁸F-fluoroethyltyrosine (FET)-PET has the potential to minimize geographical misses and overcome radiation resistance. This study hypothesized that combining TTFields with FET-PET-based SRS could provide complementary therapeutic effects, leading to improved outcomes with minimal toxicity. Material/Methods: The TaRrGeT trial is a prospective, single-arm study with historical controls designed to evaluate the effectiveness and safety of TTFields combined with SRS guided by dual-time acquisition of FET-PET in treating recurrent glioblastoma. TTFields therapy was initiated before SRS and continued until death or significant decline in performance status. TTFields treatment was also permitted following subsequent disease progression. The primary endpoint was the one-year survival rate, compared to the EF-11 trial (a prospective study of TTFields alone in recurrent glioblastoma). Propensity score matching was used to reduce bias in the comparison. Secondary endpoints included median overall survival and the incidence of radiation necrosis.

Results: Forty patients with recurrent glioblastoma were enrolled between March 2021 and July 2023. At enrollment, 30% of