Radiation Necrosis after Proton Therapy In Meningioma and Low-Grade Glioma: Incidence, Clinical Variables Association and Evolution Patterns
Abstract
Purpose
To investigate the incidence of radiation necrosis (RN) after proton therapy (PT) in low-grade brain tumors and characterize RN temporal evolution.
Methods
We retrospectively reviewed 41 meningioma and 61 low-grade glioma (LGG) patients treated with PT at our institution (2020-2023), with no prior cranial radiotherapy and ≥ 2.5 years follow-up. RN was determined based on follow-up MRIs and biopsy when available. Differences in continuous clinical variables between meningioma RN and control groups were evaluated using the Mann-Whitney U test, including lateral ventricles (LV) volume at simulation, age, and CTV volume. We assessed RN temporal evolution using longitudinal T1-weighted post-contrast images, and identified two distinct patterns: focal enhancement that subsequently regressed after treatment (Group 1); wall-enhancing lesion with an enlarging dark necrotic core (Group 2).
Results
In meningioma cohort, 9 (22%) patients developed RN (mean: 15 months, range 4-34) after PT. In LGG cohort, 20 (33%) patients developed new enhancement suspicious for RN (mean: 15 months, range: 2-48), including 6 biopsy-confirmed cases. LV volume was larger in meningioma RN patients than in controls (median: 30 cm3 [IQR 16-52] vs 17 cm3 [IQR 12-28]), corresponding to a mean difference of 13 cm3 with evidence of association (p < 0.05), while age and CTV volume were similar between groups. Meningioma RN cases demonstrated both Group 1 (3/9) and Group 2 (6/9) patterns, whereas LGG cases uniformly exhibited Group 2. LGG underwent more frequent surveillance MRI than meningioma (median interval: 3 months vs 6-12 months), which more often captured early RN stages and temporal evolution.
Conclusion
LV volume at simulation may be a potential early imaging biomarker for RN susceptibility after PT for meningioma. RN demonstrated specific evolution patterns. Larger cohorts with standardized imaging intervals and multimodal MRI are needed to validate these findings and relate imaging patterns to symptoms, treatment, and outcomes.