Fractionated Radiosurgery Associated With High Rates of Local Control for Large Volume Intact Brain Metastases

Document Type

Conference Proceeding

Publication Date


Publication Title

International Journal of Radiation Oncology Biology Physics



Radiosurgery is a well-established treatment modality for patients with intact brain metastases, and there is a known impact of tumor size and volume on tumor control. Based on desires to improve local control rates and reduce radiation necrosis rates, along with improvements in available technology, fractionated radiosurgery (FSR) is increasingly employed using both linear accelerators (LINAC) and GammaKnife (GK) units. Utilizing our single-institution database, we investigated outcomes for patients with intact brain metastases treated with FSR using LINAC or GK modalities.


Institutional records of patients with intact brain metastases treated with FSR were retrospectively reviewed. Patients treated with either GK or LINAC-based FSR were included. Descriptive analyses were performed of primary histology, lesion location, dose/fractionation, modality, tumor volume, and maximum tumor dimension. T-test and chi-square statistics were performed to compare characteristics and outcomes between patients treated with GK versus LINAC FSRT. Kaplan-Meier was used for survival estimates. Univariate analyses and Cox regression models were generated.


56 patients with 137 brain metastases were reviewed. 67.2% of lesions were treated via GK and 32.8% were treated via LINAC. Median prescription dose was 25 Gy (16-30 Gy) delivered in 2-5 fractions, with a medial lesion volume of 1.27 cc (0.01-77.0 cc). The most common treatment regimens were 24 Gy in 3 fractions and 30 Gy in 5 fractions. The most common primary tumors were lung (38.7%) and breast (28.5%). Median and mean maximum tumor dimension for the entire cohort were 1.73 cm and 2.16 cm (0.27-8.59 cm), respectively. Mean tumor volume was 6.10 cc (0.01-76.97 cc). There was a significant correlation between tumor volume and prescription dose, with larger tumors being more likely to receive a greater number of treatment fractions. The 1-y OS for all patients was 54.1% with a median OS of 15 mos. With a median follow-up of 6 months, the 1-year local control rate was 94.5%. For lesions ≤2 cc, 1-y LC was 97.4% compared to 91% for lesions > 2 cc, though this difference was not statistically significant. On Cox regression, neither maximum tumor dimension nor volume were predictive of local control. Univariate analysis demonstrated increased local recurrence in patients with prior history of neurosurgical intervention. No differences in LC or OS were observed between GK and LINAC.


FSR using 3 to 5 fractions for large volume intact metastases was associated with a high rate of LC compared to rates historically achieved with single fraction radiosurgery for tumors of similar volume. No statistically significant differences in LC or OS were detected according to maximum tumor dimension, volume, fractionation schedule, or treatment modality. FSR should be considered for larger brain metastases to improve LC.




3 Suppl

First Page



American Society for Radiation Oncology (ASTRO) Annual Meeting, October 23-26, 2022, San Antonio, TX.