Radiosurgery of limited brain metastases from primary solid tumor: results of the randomized phase III trial (NCT02355613) comparing treatments executed with a specialized or a C-arm linac-based platform

BACKGROUND: Comparative prospective data regarding different radiosurgery (SRS) modalities for treating brain metastases (BMs) from solid tumors are not available. To investigate with a single institute phase III randomized trial whether SRS executed with linac (Arm-B) is superior to a dedicated multi-source gamma-ray stereotactic platform (Arm-A). METHODS: Adults patients with 1–4 BMs from solid tumors up to 30 mm in maximum diameter were randomly assigned to arms A and B. The primary endpoint was cumulative incidence of symptomatic (grade 2–3) radionecrosis (CIRN). Secondary endpoints were local progression cumulative incidence (CILP), distant brain failure, disease-free survival (DFS), and overall survival (OS). RESULTS: A total of 251 patients were randomly assigned to Arm-A (121) or Arm-B (130). The 1-year RN cumulative incidence was 6.7% in whole cohort, 3.8% (95% CI 1.9–7.4%) in Arm-B, and 9.3% (95% CI 6.2–13.8%) in the Arm-A (p = 0.43). CIRN was influenced by target volume irradiated only for the Arm-A (p << 0.001; HR 1.36 [95% CI 1.25–1.48]). Symptomatic RN occurred in 56 cases at a median time of 10.3 months (range 1.15–54.8 months), 27 in the Arm-B at a median time of 15.9 months (range 4.9–54.8 months), and 29 in the Arm-A at a median time of 6.9 months (1.2–32.3 months), without statistically significant differences between the two arms. No statistically significant differences were recorded between the two arms in CILP, BDF, DFS or OS. The mean beam-on time to deliver SRS was 49.0 ± 36.2 min in Arm-A, and 3.1 ± 1.6 min in Arm-B. CONCLUSIONS: Given the technical differences between the treatment platforms investigated in this single-institution study, linac-based SRS (Arm-B) did not lead to significantly lower grade 2–3 RN rates versus the multi-source gamma-ray system (Arm-A) in a population of patients with limited brain metastases of small volume. No significant difference in local control was observed between both arms. For Arm-B, the treatment delivery time was significantly lower than for Arm-A. Trial registration: ClinicalTrials.gov Identifier NCT02355613. SUPPLEMENTARY INFORMATION: The online version contains supplementary material available at 10.1186/s13014-023-02216-5.