Effect and prediction of molecular subtype, metastatic site and treatment for early death in non-small cell lung cancer: A SEER-based study

Abstract

Abstract Background: early death in patients with non-small cell lung cancer remains an unresolved issue. Therefore, it is necessary to identify the risk factors for early death in non-small cell lung cancer patients. Purpose: The purpose of this study was to identify associated risk factors and develop a predictive nomogram for the early death of non-small cell lung cancer patients. Research method: We enrolled 51529 patients in SEER Database who were 60 years or older, diagnosed with primary non-small cell lung cancer, from 2010 to 2016. All patients were randomly divided into training and testing sets at 7:3. Multivariate COX proportional hazard regression was used to identify independent risk variables for early death. Use K-M survival curve to compare different molecular subtypes, metastatic sites, and treatment methods on the survival status of lung cancer patients. A predictive nomogram was developed to predict the 1-year overall survival (OS) and then validated by calibration curves, receiver operating characteristic (ROC) curves, and decision curve analysis (DCA). RESULTS: Eight independent variables, including race, sex, age, tumor size, AJCC M, molecular subtype, metastatic site and treatment were recognized by using multivariate COX proportional hazard regression model for identifying independent risk variables of early death about non-small cell lung cancer patients. By comprising these variables, a predictive nomogram was constructed in the training set cohort. In the nomogram of 1-year overall survival, the areas under the ROC curves were 0.781 (95% CI: 0.771–0.804) and for the training dataset as well as (95% CI: 0.761–0.821) for the validation dataset. Then, the consistency between the predicted and actual overall survival was confirmed by the calibration curves. Additionally, the results of the DCA indicated that the predictive model had a favorable clinical applicability. Conclusion: In the early stages of non-small cell lung cancer (NSCLC), different molecular subtypes, metastatic sites, and treatment methods can significantly affect the survival rate of patients. The nomogram model developed in this study provides an insightful and applicable tool for predicting the risk of early mortality in NSCLC patients. It can help clinicians identify patients at high risk of early death and tailor their treatment plans accordingly, potentially improving patient outcomes.