Survival prognostic nomogram for young metastatic non-small cell lung cancer: a study of the US SEER database and a Chinese cohort

Objective Young patients diagnosed with non-small cell lung cancer (NSCLC) present unique clinical, pathological, and genetic features, resulting in a highly heterogeneous patient population. The current TNM staging system is insufficient for accurately predicting their prognosis. This study aims to develop a nomogram model for survival prediction in young patients with metastatic NSCLC at initial diagnosis and further verify the effectiveness of the model. Methods This study enrolled 961 young patients diagnosed with metastatic NSCLC in the Surveillance, Epidemiology, and End Results (SEER) database between 2010 and 2017. The patients were allocated into a training cohort ( n = 673) and an internal validation cohort ( n = 288). An additional 215 patients from the Fourth Hospital of Hebei Medical University were included as a Chinese external validation cohort. Univariate and multivariate Cox regression analyses were conducted in the training cohort to identify independent risk factors influencing survival, which were used to develop a nomogram model. The model’s effectiveness was evaluated using C-index, calibration curve, receiver operating characteristic (ROC) curve, decision curve analysis (DCA) curve, and Kaplan–Meier survival curve. Results The multifactorial Cox regression model identified eight independent risk factors influencing overall survival (OS): race, marital status, histological type, T stage, N stage, liver metastasis, chemotherapy, and radiotherapy (all P < 0.05). These factors were incorporated into the nomogram, which achieved a C-index of 0.673 [95% confidence interval (CI) = 0.661–0.685]. The nomogram exhibited excellent prognostic value in both internal (C-index = 0.662, 95% CI = 0.643–0.681) and external (C-index = 0.724, 95% CI = 0.702–0.746) validation cohorts. In addition, calibration curves for 0.5-,1-, 2-, 3-, and 5-year OS probabilities showed close agreement between predicted and observed survival outcomes across various time points. Additionally, ROC curve analysis and Kaplan–Meier curves highlighted the robust discriminatory power of the model based on survival outcomes. Moreover, the DCA analysis revealed that the incremental net benefit of this model was significantly superior to that of the TNM staging system alone. Conclusions A nomogram model has been developed and validated to accurately predict the OS of young patients with metastatic NSCLC at initial diagnosis, demonstrating superior performance compared to the traditional TNM staging system. This model offers valuable guidance for precise predictions and making rational treatment decisions in clinical practice.