RADT-01. COMPARATIVE STUDY OF LOCAL CONTROL FOR BRAIN METASTASIS BETWEEN PRE-FRAME AND POST-FRAME MRI METHODS OF GAMMA KNIFE STEREOTACTIC RADIOSURGERY

Abstract

Abstract INTRODUCTION Gamma Knife radiosurgery (GKRS) is an established treatment modality in the management of brain metastasis (BM). The standard treatment flow includes frame placement followed by obtaining a post-frame MRI for planning. However, there has been a shift in practice towards using the pre-frame MRI method. It shortens the frame-wearing time while improving quality of imaging/targeting by decreasing artifacts by screws. However, no previous studies have compared the GKRS outcome between the pre- and post-frame MRI methods. METHODS 134 patients (61 pre-frame and 73 post-frame) with 307 BMs treated with first time GKRS were reviewed retrospectively. We defined local failure (LF) as ≥ 20% increase in maximum diameter (RECIST criteria). Kaplan-Meier and Gray’s method were used to estimate and compare cumulative incidence rate of LF between the groups; multivariable analysis for time-to-local failure (TTLF) was performed using Cox frailty model. The number of lesions was intrinsically modeled in frailty model. RESULTS There was no significant difference between the groups for background variables including age, gender, primary cancer, performance status, and number of lesions. Post-frame group had significantly smaller tumors at baseline (p=0.004). The cumulative incidence rates of LF for pre- vs post-frame groups were 1.6% vs 5.5% at 6 month, 6.9% vs 9.9% at 12 month, and 10.4% vs 17.4% at 24 month, respectively, without significant differences (p=0.84). Using multivariate frailty Cox model adjusting for age, gender, and lesion size, extracranial metastasis (HR 4.13, 95%CI 1.48‒11.57, p=0.007) was the significant prognostic factor for TTLF, while the frame type was not significant (p=0.46). CONCLUSIONS This is the first report comparing the GKRS outcome of pre- and post-frame MRI methods. The results indicated that pre-frame MRI method conveys a comparable local control for BM while maintaining practical benefits such as shortened frame-wearing time, less-artifact imaging, and longer preparation time available for planning.