Radiation Necrosis following SRS or FSRT with Higher BED for Brain Metastases >20 mm

Abstract

In the RTOG 90-05 study, the maximum tolerated dose of single-fraction radiosurgery (SRS) for lesions 21-30 mm was 18 Gy (BED 45 Gy12). However, patients of the RTOG study were previously irradiated with median 60 Gy (brain tumors) or 30 Gy (brain metastasis). Thus, the max. tolerated BED for lesions not irradiated before may be >45 Gy12. This study investigated safety and efficacy of SRS or fractionated stereotactic radiotherapy (FSRT) with BED >49 Gy12 for radiotherapy-naive brain metastases >20 mm. Data of 169 patients who received SRS or FSRT alone with BED >49 Gy12 for 1-4 brain metastases (2011-2022) were retrospectively analyzed. Thirty-two patients had lesions >20 mm (21-30 mm, n = 27; 21-40 mm, n = 4; >40 mm, n = 1). SRS was performed with 20 Gy (n = 9; 53.3 Gy12) or 19 Gy (n = 2; 49.1 Gy12), prescribed to the 60-80% isodose line (PTV = GTV). FSRT was Linac-based delivering 3 x 10 Gy (n = 3; 55 Gy12), 3 x 11 Gy (n = 4; 63.3 Gy12), 7 x 6 Gy (n = 1; 63 Gy12), 9 x 5 Gy (n = 1; 63.8 Gy12), 10 x 4 Gy (n = 11; 53.3 Gy12), or 12 x 4 Gy (n = 1; 64 Gy12), prescribed to the 80% isodose line (PTV = GTV+1-2 mm). SRS and FSRT were compared for grade ≥2 radiation necrosis (RN), local control (LC) and survival (OS). Fractionation [SRS (n = 89) vs FSRT (n = 48)] and lesion size (≤20 vs >20 mm) were studied in terms of RN rates. Kaplan-Meier method and log-rank test were used for univariate analyses and the Cox proportional hazards model for multivariate analyses. In patients with lesions >20 mm, SRS and FSRT were compared (Fisher's exact/Chi square test) for age, gender, KPS, year of SRS/FSRT, primary tumor, number/site of lesions, systemic treatment, and extracranial metastases. In patients with lesions >20 mm, significantly more patients receiving SRS had a KPS >80 than patients treated with FSRT (91% vs 29%, p = 0.002); otherwise, distributions of patient characteristics were balanced. In patients with lesions >20 mm, SRS was associated with significantly higher RN rates at 1 year (50% vs 9%) and 2 years (50% vs 9%) on univariate analysis (p = 0.012), which remained on multivariate analysis (HR: 0.10; 95% CI: 0.01-0.94, p = 0.044). No significant differences were found for LC (p = 0.21) at 1 year (75% vs 70%) and 2 years (75% vs 53%), and OS (p = 0.09) at 1 year (36% vs 59%) and 2 years (27% vs 46%). In patients with lesions ≤20 mm, 1- and 2-year RN rates following SRS were both 4%, compared to 0% and 15% respectively after FSRT (p = 0.60). In the SRS group, RN rates were significantly higher with lesions >20 mm (vs ≤20 mm) on univariate (p<0.001) and multivariate (HR: 3.82, 95% CI: 1.70-8.58, p = 0.001) analyses. In the FSRT group, max. diameter had no significant impact on RN rates (p = 0.93). Given the limitations of this study, FSRT with BED >49 Gy12 appears associated with a low risk of RN when used for brain metastases >20 mm. SRS with 1x19-20 Gy may not be optimal for metastases >20 mm but appears safe for lesions ≤20 mm. Prospective trials are warranted.