Background
Systematic review and meta-analysis. 24 studies, 1,867 patients with brain metastases from various solid tumors treated with SRS. Evaluated predictors of radiation necrosis (RN) after SRS for brain metastases, including dose, volume, prior WBRT, and immunotherapy. Redmond KJ et al, IJROBP 2017. PMID 27843035.
Interventions and follow up
Arm A: SRS to brain metastases (Gamma Knife, LINAC, CyberKnife) — dose 12–24 Gy single fractio
Primary endpoint: Radiation necrosis rate; predictors of RN
mFollow up: 12 month
Primary endpoint: Radiation necrosis rate; predictors of RN
mFollow up: 12 month
Results
Comparison: Rates of radiation necrosis by dose, volume (V12 Gy), and clinical predictors
Overall RN rate: 4–9% across studies; symptomatic RN ~3%
V12 Gy >10.5 cm³: Strongest predictor of RN (OR 10.0 vs V12 ≤10.5 cm³)
Prior WBRT: Associated with increased RN risk (OR 2.1)
Immunotherapy concurrent with SRS: Increased RN rate in several studies
Overall RN rate: 4–9% across studies; symptomatic RN ~3%
V12 Gy >10.5 cm³: Strongest predictor of RN (OR 10.0 vs V12 ≤10.5 cm³)
Prior WBRT: Associated with increased RN risk (OR 2.1)
Immunotherapy concurrent with SRS: Increased RN rate in several studies
Adverse events
Main adverse events: Radiation necrosis: MRI-detectable in 4–9%; symptomatic requiring steroids in 3%; requiring surgical resection or bevacizumab in 1–2%. Distinguishing RN from tumor recurrence on MRI is challenging — advanced MRI (perfusion, spectroscopy, PET) may be required.
Conclusions
V12 Gy is the strongest predictor of radiation necrosis after SRS for brain metastases, with V12 >10.5 cm³ associated with 10-fold increased risk. Limiting V12 and avoiding re-irradiation with prior WBRT reduces RN risk. Concurrent immunotherapy (checkpoint inhibitors) may increase RN risk — temporal spacing is advisable.
Key Limitations
Key Limitations: Meta-analysis of heterogeneous studies with different RN definitions, imaging protocols, and treatment parameters. Some RN events may represent pseudoprogression. IO–SRS interaction data are from early retrospective series. Prospective randomized data on IO timing relative to SRS are lacking.
Clinical Context
V12 Gy constraint is standard in SRS planning for brain metastases — most guidelines recommend keeping V12 ≤10 cm³ to minimize RN risk. Bevacizumab is effective for steroid-refractory symptomatic RN. Concurrent checkpoint inhibitor and SRS is generally considered safe but temporal spacing (1 week before/after SRS) is commonly practiced.
References