Preoperative Fractionated Stereotactic Radiosurgery for Brain Metastases: A Prospective Pilot Study at a Single Institution.

Purpose: Brain metastases are a common occurrence in patients with advanced malignancy. Treatment ranges from surgical resection, stereotactic radiosurgery (SRS), whole-brain radiation therapy (WBRT), or some combination thereof. Traditionally, SRS has followed surgical resection to help reduce the risk of local recurrence. In recent years, interest has arisen in treating these patients in the preoperative setting to help reduce the risk of leptomeningeal disease (LMD) and radiation necrosis while preserving rates of local control. This prospective pilot study, originally planned as a phase 2 protocol, investigated the use of 3-fraction preoperative SRS on the Gamma Knife (GK) Icon with the goal of examining those outcomes. Methods and Materials: Patients who had symptomatic and a limited number of brain metastases amenable to surgical resection were prospectively enrolled in this study. Patients were treated with fractionated SRS on the GK Icon to 24 to 27 Gy in 3 fractions followed by surgical resection within 2 weeks. Postoperative MRIs were obtained to assess the extent of resection. Patients were then followed up with standard-of-care MRIs every 3 months for the first 2 years. Rates of local control, distant brain failure, LMD, and delivery of WBRT were recorded on follow-up. Results: Twenty patients were prospectively enrolled in the trial, and an additional 11 were treated off study using 1 to 3 fractions between April 2021 and June 2024. The median age was 64 (31-81), and 55% were females. Primary malignancies were mainly non-small cell lung cancer, melanoma, and esophageal cancer. The median prescription dose was 27 Gy (15-27 Gy) in 3 (1-3) fractions. Surgery was performed at a median of 1 day post-SRS (0-11). All available postoperative MRIs showed gross total resection. Median follow-up was 9 months. One patient experienced a local failure, yielding a 1-year local control rate of 95%. The rate of distant brain failure at 6 and 12 months was 38% at each time point. Three patients (9.6%) developed LMD in follow-up. Six patients ultimately received WBRT, yielding a 1-year WBRT-free survival of 78%. Overall survival at 1 year was 52%. There were no predictors of overall survival, local failure, or distant brain failure on Cox regression. Two patients experienced short-term toxicity, including grade 2 intracranial swelling and one grade 5 hemorrhagic stroke related to a hypertensive crisis. There was no recorded radionecrosis in follow-up. Conclusions: Fractionated preoperative SRS appears to be safe and effective with high rates of local control and low rates of LMD and radionecrosis. Results of ongoing phase 3 studies directly comparing preoperative to postoperative SRS will help further define the appropriate sequencing of therapies.