Repeat stereotactic radiosurgery for recurrent brain metastases: a retrospective comparison of local progression and distant brain metastases after prior radiosurgery

Aim This study evaluated the efficacy and safety of repeat stereotactic radiosurgery (SRS) for recurrent brain metastases (BMs), focusing on failure patterns of local progression (LP) and distant brain metastases (DBM). Methods Patients who underwent first-time SRS for newly diagnosed BMs (2011-2022) and repeat SRS (until June 2024) were included. Treatment courses were first-time or repeat SRS, with repeat SRS classified as LP and DBM. The primary endpoint was the 1-year local control rate (LCR). Secondary endpoints included survival and radiation necrosis (RN) incidence. Results Among 723 courses (427 patients, 4,524 BMs), 404 (141 patients, 2,924 BMs) met criteria. First-time SRS was performed in 141 courses (775 BMs), and repeat SRS in 263 (2,149 BMs), including 45 LP (38 patients, 224 BMs) and 218 DBM (126 patients, 1,925 BMs). The median age was 65 years, and 75.9% had lung cancer. LP had a longer interval from prior SRS (12.6 vs. 6.3 months, P < 0.001) but similar follow-up (12.4 vs. 13.7 months). The 1-year LCR was lower in LP (72.4% vs. 88.3%, P = 0.0022), though survival was similar (17.9 vs. 16.3 months). LP had higher RN incidence (20.6% vs. 5.7%, P < 0.001) and symptomatic RN (13.3% vs. 2.8%, P < 0.001). Multivariate analysis identified LP failure as a significant factor for increased local failure (subdistribution hazard ratio [SHR] 2.35, P = 0.039) and RN (SHR 3.41, P < 0.001). Conclusions Despite similar survival, LP failure was associated with lower LCR and higher RN incidence, highlighting the need for optimized repeat SRS strategies.