Re-irradiation for recurrent glioblastoma (GBM): a systematic review and meta-analysis

To determine the efficacy and toxicity of re-irradiation for patients with recurrent GBM. We searched various biomedical databases from 1998 to 2018, for eligible studies where patients were treated with re-irradiation for recurrent GBM. Outcomes of interest were 6 and 12-month overall survival (OS-6, OS-12), 6 and 12-month progression free survival (PFS-6, PFS-12) and serious (Grade 3 +) adverse events (AE). We used the random effects model to pool outcomes across studies and compared pre-defined subgroups using interaction test. Methodological quality of each study was assessed using the Newcastle-Ottawa scoring system. We found 50 eligible non-comparative studies including 2095 patients. Of these, 42% were of good or fair quality. The pooled results were as follows: OS-6 rate 73% (95% confidence interval (CI) 69-77%), OS-12 rate 36% (95% CI 32-40%), PFS-6 rate 43% (95% CI 35-50%), PFS-12 rate 17% (95% CI 13-20%), and Grade 3 + AE rate 7% (95% CI 4-10%). Subgroup analysis showed that prospective studies reported higher toxicity rates, and studies which utilized brachytherapy to have a longer OS-12. Within the external beam radiotherapy group, there was no dose-response [above or below 36 Gy in 2 Gy equivalent doses (EQD2)]. However, a short fractionation regimen (ae<currency> 5 fractions) seemed to provide superior PFS-6. The available evidence, albeit mostly level III, suggests that re-irradiation provides encouraging disease control and survival rates. Toxicity was not uniformly reported, but seemed to be low from the included studies. Randomized controlled trials (RCT) are needed to establish the optimal management strategy for recurrent GBM.