Risk Stratification and Clinical Outcomes for Surgically Treated Brain Metastasis: A Single Institutional Experience from China.
BACKGROUND Brain metastases (BMs) cause significant mortality in cancer patients. While surgical resection is standard for large or symptomatic lesions, accurate prognostic stratification remains challenging. Existing models often rely on historical data, failing to incorporate the survival impact of modern immune and targeted therapies. This study identifies independent prognostic factors and shares our institutional experience regarding surgical outcomes and multimodal management. METHODS We analyzed 301 patients who underwent surgical resection for BMs at Beijing Tiantan Hospital (2020-2022) using the NBTRC database. Univariate and multivariate Cox regression analyses were performed to identify predictors of overall survival (OS). A statistical model was constructed to visualize these parameters, and its internal discrimination and calibration were evaluated. RESULTS Multivariate analysis identified negative OS predictors in order of importance: infratentorial tumor location (HR 1.896, P<0.001), multiple metastases (HR 1.194, P=0.003), and advancing age (HR 1.018, P=0.035). Conversely, postoperative immune/targeted therapy (HR 0.361, P<0.001), adjuvant WBRT (HR 0.671, P=0.015), and higher discharge KPS (HR 0.983, P<0.001) were significant protective factors. An internally validated nomogram (C-index 0.692) confirmed these stratifications. CONCLUSION Based on our institutional experience, surgical resection of BMs has proven to be a safe and effective technique in our hands for carefully selected patients. Furthermore, strict risk stratification-focusing on tumor location, tumor burden, and the integration of modern systemic therapies-is essential for optimizing these real-world clinical outcomes.