Abstract BACKGROUND The management of recurrent glioblastoma (rGBM) remains a clinical challenge, with only limited therapeutic options available to date. Reirradiation may offer a progression-free survival (PFS) benefit in selected cases, but data are scarce. MATERIALS AND METHODS Consecutive patients with GBM (WHO grade 4, IDH wildtype) who underwent at least one additional course of cranial radiotherapy for suspected or histopathologically confirmed rGBM at a tertiary neuro-oncological center were retrospectively analyzed. The primary endpoint was PFS, secondary endpoints included reirradiation-related adverse event rates, with a particular focus on radiation necrosis (RN). RESULTS Fifty-nine patients were included with a median follow-up (range) of 8.7 (0.5-48.0) months after reirradiation. The median time to first recurrence was 15 (4-89) months, with the majority of recurrences occurring in-field (59.7%). The EQD2α/β=10 ranged from 31.3-80.2 Gy with a median prescription dose of 42 Gy. Reirradiation was combined with systemic therapy in 81.4% of patients. No grade 3-5 acute reirradiation-related adverse events were observed. RN was diagnosed in 16.9% of patients (80% grade 2 and 20% grade 3), with a notably low rate in those receiving anti-VEGF therapy. RN risk was independent of reirradiation volume or dose (p = 0.15 and 0.43, respectively). The overall response rate following reirradiation was 83.6% and the median PFS 5.9 (0.5-48.0) months. Concomitant chemotherapy or anti-VEGF therapy was significantly associated with improved outcomes (p = 0.049), whereas smaller reirradiation volumes demonstrated a non-significant trend towards longer PFS (p = 0.23). CONCLUSION In this retrospective analysis, reirradiation for rGBM was feasible and safe, conferring a potential PFS benefit in selected patients. Bevacizumab emerged as a particularly promising combination partner, contributing to both RN prevention and enhanced efficacy. Prospective trials are warranted to guide evidence-based decision-making in this complex patient population.
Dejonckheere et al. (Wed,) studied this question.