RESUMO
BACKGROUND: Stereotactic radiosurgery (SRS) presents as a good treatment option for smaller, deep-seated brain metastases (BMs). This study aims to determine predictors of SRS failure for patients with non-small cell lung cancer BMs. METHODS: This was a retrospective study of single non-small cell lung cancer BMs treated using SRS. We included patients >18 years with a single, previously untreated lesion. Primary outcome was treatment failure, defined as BMs dimension increase above the initial values. Demographic, clinical, and radiological data were collected to study potential predictors of treatment failure. RESULTS: Worse rates of progression-free survival (PFS) were associated with heterogeneous contrast enhancement (18.1 ± 4.1 vs. 41.9 ± 4 months; P < 0.001). Better rates of PFS were associated with volumes <1.06 cm3 (log-rank; P = 0.001). Graded prognostic assessment was significantly associated with survival at 120 months (log-rank; P < 0.001). Karnofsky Performance Scale was evaluated in 3 strata: 90-100, 80, and ≤70. Mean survival rates for these strata were 31.8 ± 3.9, 10.6 ± 2.2, and 9.8 ± 2.3 months, respectively (log-rank; P < 0.001). There were no differences regarding presence of extracranial metastases, age, or lesion location. A multivariable logistic regression found that volume <1.06 cm3 was associated with higher survival rates at 10 years (odds ratio: 3.2, 95% confidence interval: 1.3-8.0). CONCLUSIONS: Contrast-homogeneous metastases and lesions <1.06 cm3 are associated with better rates of PFS. Karnofsky Performance Scale and graded prognostic assessment were associated with more favorable survival rates after 10 years. Volume <1.06 cm3 was the only significant predictor of survival in the multivariable analysis.