KEY POINTS
- This international retrospective study evaluated 271 patients with recurrent IDH-wildtype glioblastoma treated with single-fraction Gamma Knife stereotactic radiosurgery at 15 centers between 2007 and 2025. All patients had undergone previous radiotherapy and received SRS at their first recurrence. Median age was 61 years, median recurrent tumor volume was 3.0 cm³ (IQR 0.8–8.7), and median Karnofsky performance status at recurrence was 80. The primary endpoint was post-recurrence survival, defined from first recurrence to death or last follow-up.
- SRS targeted contrast-enhancing recurrent disease without an additional CTV or PTV margin. The median prescription was 16 Gy (IQR 14–19 Gy), generally prescribed to the 50% isodose line, with median maximum tumor dose 30.9 Gy and median target coverage 100%. Dose selection was individualized according to tumor size, location, and proximity to critical structures. Higher prescription doses were preferentially delivered to smaller tumors, creating an important selection effect: prescription dose and recurrence volume were inversely correlated (Spearman r=−0.54; p<0.001).
- Higher SRS prescription dose was independently associated with longer post-recurrence survival. In multivariable Cox analysis, each additional Gy was associated with lower mortality hazard (HR 0.81 per Gy, 95% CI 0.73–0.90; p<0.001). In the detailed Results section, median survival was 15.8 months after >19 Gy versus 10.1 months after ≤19 Gy (p<0.001). Patients receiving >19 Gy also had longer overall survival measured from initial surgery (29.8 versus 23.3 months; p=0.009). However, post-SRS progression-free survival did not significantly differ between dose groups.
- The survival association persisted after propensity score matching, but the comparison remained observational. Matching produced 118 patients, 59 per dose group, balanced for measured clinical factors, including tumor volume, performance status, MGMT status, multifocality, and salvage treatment. Median post-recurrence survival was 15.8 months with >19 Gy versus 10.3 months with ≤19 Gy (p=0.005), while overall survival was 28.3 versus 22.5 months (p=0.048). Post-SRS progression-free survival remained similar. Residual confounding from tumor biology, clinical selection, and unmeasured treatment factors cannot be excluded.
- Tumor volume and performance status further distinguished prognosis. Recursive partitioning identified a recurrence volume threshold of 7.2 cm³, with dose escalation most strongly associated with favorable outcomes among smaller lesions. Patients with volume ≤7.2 cm³ receiving >19 Gy had median post-recurrence survival of 15.2 months, compared with 11.7 months in the intermediate-risk group and 6.4 months among patients with volume >7.2 cm³ and KPS ≤80. These thresholds were derived within the study population and require external validation.
- Reported symptomatic radiation injury was relatively uncommon, although bevacizumab use was extensive. Symptomatic radiation necrosis occurred in 5.5% and neurological deficits attributed to radiosurgery in 8.2%, including persistent deficits in approximately 6.4%. Following SRS, 67.7% received bevacizumab, with most initiating it within 30 days. Other salvage therapies included temozolomide or lomustine in 52.9%, reoperation in 8.0%, and tumor treating fields in 8.4%. These additional treatments complicate attribution of both survival and toxicity outcomes to radiation dose alone.
- Several reporting and methodological limitations prevent establishing an optimal SRS prescription. Dose was not randomized, high-dose patients had smaller tumors and different treatment characteristics, and the intervals from initial RT to reirradiation were not consistently available. The 19-Gy threshold was selected as the upper quartile of the observed dose distribution, not identified as a biological threshold. Moreover, the abstract reports different median survival estimates from the detailed Results section, and the absence of a corresponding improvement in post-SRS progression-free survival warrants caution when interpreting the survival association.
CLINICAL TAKEAWAY
Higher single-fraction Gamma Knife doses were associated with longer survival in selected patients with small-volume, first-recurrence IDH-wildtype glioblastoma, without a clear increase in reported symptomatic radiation injury. However, this multicenter observational study does not establish that dose escalation itself improves survival, particularly given treatment-selection effects and frequent bevacizumab use. The findings support prospective dose-optimization research rather than routine escalation above existing recommendations.
SOURCE
International Journal of Radiation Oncology, Biology, Physics