Abstract
BACKGROUND: The brain serves as the primary site for metastasis in patients with both non-small cell lung cancer (NSCLC) and small cell lung cancer (SCLC). The presence of lung cancer with brain metastasis (LCBM) is a debilitating condition associated with considerable morbidity and mortality. The objective of this study was to assess the incidence and survival rates of LCBM in the United States population. METHODS: We analyzed a total of 9,212 patients diagnosed with LCBM between 2010 and 2015, extracted from the Surveillance, Epidemiology, and End Results (SEER) database. Our analysis assessed the incidence, relative survival, and conditional survival (CS) of LCBM. We utilized the Kaplan-Meier method to estimate overall survival and determine CS at year y+x after x years of survival, following the formula CS(y|x) = CS(y+x)/CS(x). Prognostic factor selection was performed using the least absolute shrinkage and selection operator (LASSO) regression approach, and multivariate Cox regression was employed to demonstrate the impact of these predictors on outcomes and construct a CS-based nomogram. RESULTS: The overall age-adjusted incidence rate of LCBM was 5.82 cases per 100,000, with a slight decline observed during our study period. Patient relative survival showed a continuous decline with increasing age. CS analysis revealed that the 5-year CS rate for patients initially diagnosed with LCBM adjusted from 3% to 13%, 28%, 52%, and 73% over successive years of survival (1-4 years). Identified predictors included age at diagnosis, sex, race, tumor size, tumor grade, surgery, radiotherapy, and chemotherapy. These predictors, along with the CS formula, were employed to develop a CS-based nomogram for real-time prognosis prediction. Calibration curve, area under the time-dependent receiver operating characteristic (ROC) curve, concordance index (c-index), and decision curve analysis (DCA) demonstrated the model's strong predictive capabilities. CONCLUSIONS: This study deepened our understanding of LCBM patients, summarizing their epidemiological characteristics and CS patterns. We successfully developed a novel CS-based nomogram model for dynamic survival estimation, offering real-time and personalized prognostic information that is clinically valuable.