Analysis of Prognostic Factors and Risk Prediction in Brain Metastases: A SEER Population-Based Study

ZIXUAN YANG^*Qiang JiXun KangLili ZhouFeng ChenWenbin Li^*

• Capital Medical University, Beijing, China

This study investigates survival disparities and prognostic factors in patients with brain metastases originating from various primary cancers to facilitate risk stratification and enhance precision in diagnosis and treatment.Patients diagnosed with brain metastases between 2010 and 2018 were identified from the SEER database for analysis. Overall survival (OS) was evaluated using Kaplan-Meier curves and log-rank tests, complemented by multivariate Cox regression analysis. The impact of age on the risk and survival of brain metastases was examined using Restricted Cubic Splines (restricted cubic splines (RCS)RCS) in Cox regression models.A total of 55,094 patients diagnosed with brain metastases between 2010 and 2018 were retrospectively identified from the SEER database for inclusion in this study.The study included 55,094 patients diagnosed with brain metastases from 2010 to 2018. It was found that the median survival times were 2 months (95% CI: 2-3 months) for liver cancer, 3 months (95% CI: 3-4 months) for stomach cancer, and 5 months (95% CI: 4-5 months) for lung cancer. Survival was influenced by factors such as sex, age, primary cancer site, race, income, marital status, and treatment approaches. Surgical treatment notably decreased the mortality risk, with a hazard ratio (HR) of 0.49 (95% CI: 4-5 months 0.46-0.53) for lung cancer, 0.43 (95% CI:3-4 months 0.37-0.49) for kidney cancer, and 0.63 (95% CI: 5-7 months0.54-0.74) for breast cancer. The predictive model created with these variables achieved a C-index of 0.723 and 0.722 in the training and test sets, respectively, indicating vital accuracy. Calibration curves displayed minimal errors, and the area under the curve (AUC) values showed excellent performance at 3 months (training: 0.83, test: 0.83), 6 months (training: 0.80, test: 0.80), and 12 months (training: 0.77, test: 0.76).Brain metastases from liver, stomach, and lung cancers are linked to a poor prognosis. Surgical intervention significantly lowers mortality risk. The predictive model, which incorporates vital survival factors, demonstrates high accuracy and reliable performance, supporting the clinical management of patients with brain metastases.