Akademik Veri Yönetim Sistemi
ESTRO 2026, Stockholm, İsveç, 16 - 19 Mayıs 2026, cilt.218, sa.1, ss.936, (Özet Bildiri)
Purpose/Objective:
Local recurrence after stereotactic radiosurgery (SRS1) for brain metastases presents a therapeutic challenge, and optimal re-irradiation parameters remain unclear. This study evaluated local control (LC) and predictors of radiation necrosis (RN) after re-SRS (SRS2) in previously irradiated brain metastases.
Material/Methods:
From January 2015 to March 2025, 40 patients with 59 recurrent brain metastases treated with SRS2 to the same lesion were retrospectively analyzed. Patients previously treated with whole-brain radiotherapy were excluded. Follow-up MRI scans were analyzed for local failure and radiographic RN according to RANO criteria. Kaplan–Meier method for LC and RN-free survival; Cox regression analysis for predictors and ROC curve analysis for cutoff values were performed using SPSS v27.
Results:
Median follow-up was 11 months (2–83). Median interval between SRS1 and SRS2 was 11 months (3– 43); median SRS2 prescription dose was 24 Gy (14–35) in median of 3 fractions (1–5) and median GTV diameter at SRS2 was 2.1 cm (0.3–7.6). One-year LC was 57.5% and one-year RN rate was 20%. Most RN events (90%) occurred within first year.ROC curve analysis defined optimal cutoff values for doses predictive of RN. SRS2 EQD₂(10) threshold of 33.6 Gy separated higher- and lower-risk groups, with oneyear RN-free survival of 63.8% for > 33.6 Gy and 91.3% for ≤ 33.6 Gy (p = 0.007; AUC = 0.780, 95%CI 0.625– 0.934). For total EQD₂(10) (SRS1 + SRS2), threshold was 75 Gy, yielding RN-free survival of 63.2% vs 91.4% (p = 0.006; AUC = 0.764, 95%CI 0.610–0.918). One-year RNfree survival was 65.8% for SRS2 BED₃ > 83.7 Gy vs 90.9% for ≤ 83.7 Gy (p = 0.013; AUC = 0.785, 95%CI 0.700–0.870) and 65.2% vs 91.0% for total BED₃ > 177 Gy vs ≤ 177 Gy (p = 0.013; AUC = 0.774, 95%CI 0.690– 0.860). Across all comparisons, higher BED and EQD₂ significantly increased RN risk without significantly improving LC.Factors including sex, age, histology, lesion location, re-irradiated metastases number, prior surgery, GTV/PTV diameters/volumes at SRS1–SRS2, conformity/heterogeneity/gradient indices, brain V5– V25 values, concurrent/adjuvant systemic therapy, and whether SRS2 was delivered in single/multiple fractions were not associated with LC or RN (p > 0.05).
Conclusion:
In this study with a limited number of patients, maintaining SRS2 prescription doses below defined threshold values appears to reduce the risk of RN without significantly compromising LC. Optimal dose schemes that balance efficacy and safety for reirradiation should be validated in prospective studies.
Keywords: reirradiation, radionecrosis, brain metastasis