EANO consensus prioritizes advanced imaging and symptom-guided treatment for brain radiation necrosis

EANO reached consensus on 53 statements covering radiation necrosis diagnosis, advanced imaging, corticosteroids, bevacizumab, surgery and other treatment options.

KEY POINTS

  • This European Association of Neuro-Oncology (EANO) consensus statement addresses the diagnosis and management of radiation necrosis (RN) after radiotherapy for primary and metastatic brain tumors. A multidisciplinary panel of 20 experts used a three-round modified Delphi process, with consensus defined as at least 80% agreement. Ultimately, 53 of 57 proposed statements were accepted, covering epidemiology, pathophysiology, radiation dose and volume, imaging, histopathology, corticosteroids, bevacizumab, surgery, and laser interstitial thermal therapy (LITT). The document represents structured expert consensus rather than a systematic evidence-graded clinical practice guideline.
  • Radiation necrosis typically develops 6–24 months after brain radiotherapy, although substantially later presentations are possible. Reported occurrence varies widely, approximately 4–30%, depending on disease, treatment, imaging surveillance, and diagnostic criteria. The panel identifies irradiated brain volume, fraction size, cumulative dose, reirradiation, and selected systemic therapies as important risk considerations. For single-fraction SRS, normal-brain V12Gy >10 cm³ is associated with increased RN risk, supporting consideration of fractionated stereotactic radiotherapy for larger lesions. Importantly, the panel found no established evidence that fractionated proton therapy has a different RN incidence from fractionated photon therapy.
  • Conventional contrast-enhanced MRI alone is insufficient to reliably distinguish RN from tumor recurrence. The panel recommends adding advanced MRI techniques, particularly dynamic susceptibility contrast perfusion MRI, and/or MR spectroscopy when new or progressive enhancement raises diagnostic uncertainty. Increased perfusion and elevated choline favor viable tumor, whereas lower perfusion and increased lactate are more suggestive of RN. The strongest diagnostic approach combines advanced MRI with amino acid PET using FET, FDOPA, or MET, although no imaging technique provides perfect discrimination. Reported diagnostic performance for amino acid PET is generally approximately 80–90%, depending on the tracer and clinical population.
  • Histopathology remains the diagnostic reference standard when imaging is inconclusive, but interpretation can be difficult because radiation injury and viable tumor often coexist. In treated gliomas, reactive astrocytes may resemble neoplastic cells, while tumor-specific immunohistochemical markers such as IDH1 R132H or H3 K27M can improve identification of residual tumor in selected molecular subtypes. In brain metastases, lineage-specific immunohistochemistry may help detect small amounts of viable tumor within necrotic tissue. The panel recommends preserving tissue for permanent sections without prior freezing when possible, because freezing artifacts can further compromise interpretation. For uncertain lesions managed without surgery, MRI follow-up at approximately 6–12-week intervals is recommended, with shorter intervals when clinical deterioration occurs.
  • Management should be guided primarily by symptoms rather than radiological abnormalities alone. Asymptomatic RN generally warrants observation and serial imaging, whereas progressive neurological symptoms, including focal deficits, seizures, cognitive deterioration, or increasing intracranial pressure, may require intervention. The consensus identifies symptoms persisting or worsening for more than seven days alongside compatible imaging as a reasonable trigger for treatment. Corticosteroids remain the most accessible initial treatment, with dexamethasone administered at the lowest effective dose and reduced according to clinical improvement rather than MRI findings alone. The document discusses starting doses of approximately 1.5–4 mg daily, increasing to around 8 mg for severe symptoms, but emphasizes that no standardized dose or duration has been validated.
  • Bevacizumab is recommended for selected patients with persistent or corticosteroid-dependent symptomatic RN. The panel reached consensus that escalation should be considered when symptoms remain steroid-dependent after approximately four weeks, particularly when ≥8 mg/day of dexamethasone is required. Reported regimens include 5 mg/kg every two weeks, 7.5 mg/kg every three weeks, and fixed-dose schedules, with no established optimal dose or duration. Bevacizumab can reduce vascular permeability, contrast enhancement, and edema, but clinicians must consider hypertension, thromboembolism, bleeding, and impaired wound healing. Response predictors and the likelihood of RN recurrence after treatment remain insufficiently defined.
  • Surgical resection should be considered whenever safely feasible, particularly when rapid symptom relief, definitive tissue diagnosis, or reduction of prolonged medical treatment is important. Surgery provides both treatment and histopathological assessment, whereas LITT may be considered for selected treatment-resistant lesions unsuitable for conventional resection. However, LITT evidence remains predominantly nonrandomized, and no reliable comparative hierarchy between surgery, bevacizumab, and LITT has been established. The panel also concluded that evidence is insufficient to support routine hyperbaric oxygen therapy for cerebral RN. The overall management framework remains individualized, multidisciplinary, and constrained by the absence of large randomized trials.

CLINICAL TAKEAWAY

The EANO consensus supports a structured approach to brain radiation necrosis that combines advanced MRI, amino acid PET, and selective histopathological confirmation to distinguish treatment injury from tumor progression. Management should focus on symptomatic disease, using the minimum effective corticosteroid dose, bevacizumab for selected steroid-dependent patients, and surgery when feasible. These recommendations provide a practical clinical framework but remain largely consensus-based rather than supported by high-level comparative evidence.

SOURCE

Neuro-Oncology

Browse more research Suggest a correction