Radiation Necrosis Following Stereotactic Radiosurgery: Incidence, Risk Factors, Imaging, and Management

Abstract

Incidence: Radiation necrosis (RN) is a common late complication of intracranial stereotactic radiosurgery (SRS), with reported rates ranging 2–30+% depending on tumor type, dose, and follow-up. For brain metastases, landmark studies report RN in ~8–11% at 1–2 years (RTOG-9005), ~24–26% in early series, and ~5% in modern trials. For benign tumors, meningioma cohorts show RN rates ~3–27% and vestibular schwannoma ~0.5–12%. Fractionated SRS markedly lowers RN incidence for larger lesions. Risk Factors: Key predictors include treatment-related factors (high dose per fraction, large treated volume [especially normal brain V10–V12], re-irradiation) and patient/tumor factors (tumor size, histology, location, concurrent therapies). E.g. normal brain V12 of 5/10/15 cc yields ~10/15/20% RN risk. Concurrent systemic therapy or adjuvant resection also elevate risk (adjusted ORs ~8.2 and 3.7). Prior SRS increases RN risk (∼20% 1-year RN if same lesion was re-treated). Imaging: Distinguishing RN from tumor recurrence is challenging. Conventional MRI often shows ring-enhancement with edema, but this is nonspecific. Advanced MRI (DWI, perfusion, spectroscopy) and metabolic imaging improve accuracy. DWI/ADC tends to be higher in RN. MRI “T1/T2 mismatch” (ill-defined T1 enhancement vs surrounding T2 signal) has ~83% sensitivity/91% specificity. Perfusion MRI is useful: e.g. relative cerebral blood volume (rCBV) cutoff ~1.5–2.1 distinguishes tumor vs RN (sensitivity ~91– 100%, specificity ~72–95%). MR spectroscopy (low Cho/Cr ratio suggests RN). Amino-acid PET tracers (e.g. ^18F-FET, ^18F-FDOPA) outperform FDG; for instance, F-DOPA PET can reach ~90% sensitivity/specificity at optimized uptake thresholds. No single modality is definitive; a multimodal approach (MRI + perfusion + PET) yields best diagnostic accuracy. Management: First-line therapy for symptomatic RN is corticosteroids (rapid edema control). However, prolonged steroids have toxicity. Bevacizumab (anti-VEGF) is highly effective: a meta-analysis reported symptomatic improvement in ~88% of patients, radiographic improvement in ~86%, and steroid reduction in ~45%. Typical regimens are 5–10 mg/kg IV every 2–3 weeks for 4–6 cycles, which often yields dramatic edema resolution, though recurrence after discontinuation can occur. Laser interstitial thermal therapy (LITT) is a minimally invasive option for RN (often concurrently obtaining biopsy); it yields symptomatic control in ~71% (vs ~88% with bevacizumab) and radiographic improvement ~65% (vs 86%), with similar safety profiles. Surgical resection is reserved for large, refractory RN causing mass effect, and yields rapid relief of symptoms and allows tissue diagnosis (histology often shows mix of gliosis, necrosis, or occult tumor). Hyperbaric oxygen has anecdotal support (angiogenesis/edema reduction) but lacks high-level evidence. Follow-up: Close MRI surveillance (NCCN: every 2–3 months for 1–2 years then 4–6 months) is recommended to identify RN and recurrences early.