LASSO-based nomograms predict early death in small cell lung cancer (SCLC) patients with brain metastasis.

Background: Small cell lung cancer with brain metastasis (SCLC-BM) is associated with a poor prognosis and a high probability of death. However, to date, few models exist to predict early death (ED) in patients with SCLC-BM. This study aimed to construct nomograms to predict ED in patients diagnosed with SCLC-BM. Methods: The data of SCLC-BM patients were extracted from the Surveillance, Epidemiology, and End Results (SEER) database (2010-2015). The patients were randomly divided into training and test cohorts at a 7:3 ratio. Univariate regression was first performed to assess the significance of each variable. Least absolute shrinkage and selection operator (LASSO) regression was then applied to identify the most critical and minimum prognostic factors. Subsequently, a multivariate analysis was conducted to develop two nomograms to predict all-cause early death (ACED) and cancer-specific early death (CSED). The calibration and discriminative abilities of these nomograms were evaluated using receiver operating characteristic (ROC) curves and calibration curves. A decision curve analysis (DCA) was employed to assess the clinical applicability of the models. Results: Primary site, regional lymph node stage (N stage), race, age, tumor size, radiotherapy, and chemotherapy were identified as independent predictors of ACED, while race, age, tumor size, radiotherapy, and chemotherapy were identified as independent predictors of CSED. Nomograms were developed based on these variables. The area under the curve (AUC) values of the ROC curves for ACED were 0.837 in the training cohort and 0.813 in the test cohort, while those for CSED were 0.794 and 0.783, respectively. Based on the performance of the ROC, calibration, and DCA curves, these predictive models demonstrated favorable efficacy. Conclusions: This study developed nomograms to predict ACED and CSED in patients with SCLC-BM. The nomograms demonstrated clinical utility in ED prediction in the post-treatment planning period, and thus could assist oncologists in identifying high-risk patients after initial treatment planning has been established.