Abstract
INTRODUCTION: Glioblastoma (GBM) is the most aggressive primary brain tumor in adults, with a high recurrence rate. Differentiating tumor progression from treatment-induced changes remains a diagnostic challenge. RESEARCH QUESTION: Can FET-PET reliably distinguish true GBM recurrence from post-therapeutic effects across multiple recurrence stages, and how does MGMT promoter methylation influence diagnostic accuracy? MATERIAL AND METHODS: We performed a retrospective, single-center study correlating PET findings with histopathological results from surgery or biopsy at first, second, and third recurrence. Sensitivity and specificity were calculated using three interpretive scenarios for uncertain PET results. Subgroup analysis based on MGMT methylation was also performed. RESULTS: 960 patients with GBM were treated at our department between 2006 and 2021, of whom 347 (36.1 %) had one tumor recurrence during follow-up with 156 (45.0 %) FET-PET scans available. 95 patients (9.9 %) had a second recurrence with a FET-PET conducted in 37 of these (39.0 %), whereas 23 patients (2.4 %) had a third recurrence with a FET-PET available in 8 patients (34.8 %). Sensitivity was 95 % at first recurrence, 96 % at second, and 83 % at third. Specificity was low overall (13 %, 0 %, 0 %). No statistically robust differences in diagnostic performance were observed between MGMT subgroups (p > 0.4 for all). DISCUSSION AND CONCLUSION: PET imaging demonstrated high sensitivity for detecting GBM recurrence but showed variable specificity. Notably, sensitivity declined with an increasing number of recurrences, suggesting that cumulative treatment effects and greater tumor heterogeneity over time may affect diagnostic performance. Further prospective studies are needed to refine diagnostic thresholds and improve clinical decision-making.