A Multi-Parametric Prognostic Model Based on Clinical Features and Serological Markers Predicts Overall Survival in Non-Small Cell Lung Cancer Patients with Chronic Hepatitis B Viral Infection

Abstract

features and serological markers to estimate overall survival (OS) in non-small cell lung cancer (NSCLC) patients with chronic hepatitis B viral (HBV) infection.Methods: The prognostic model was generated by using Lasso regression in training cohort. The incremental predictive value of the model to traditional TNM staging and clinical treatment for individualized survival was evaluated by concordance index (C-index), time-dependent ROC (tdROC), and decision curve analysis (DCA). A model risk score nomogram for OS was built by combining TNM staging and clinical treatment. Then we stratified patients into high and low risk subgroups according to the model risk score. Difference in survival between subgroups was analyzed using Kaplan–Meier survival analysis. Furthermore, correlations between the prognostic model and TNM staging or treatment were analysed.Results: The C-index values of the model for predicting OS were 0.769 and 0.676 in the training and validation cohorts, respectively, which were higher than that of TNM staging, and treatment, the tdROC curve and DCA also showed the model had good predictive accuracy and discriminatory power than TNM staging and treatment. And the nomogram shown some clinical net benefit. According to the model risk score, we divided the patients into low risk and high risk subgroups. The differences of OS rates were significant in the subgroups. Furthermore, the model was positive correlation with TNM staging and treatment. Conclusions: The proposed prognostic model showed favorable performance than traditional TNM staging and clinical treatment for estimating OS in NSCLC (HBV+) patients.

Construction of a prognostic model of lung adenocarcinoma based on machine learning

“…The process of this method is closely related, and compared with the prediction model constructed by solely using gene expression data or clinical data, the effect is better. However, the clinical information in the acquired data is not comprehensive, and better results will be obtained if the data such as CT images of lung cancer, serum markers and sufficient clinical data can be combined [14][15][16] . The data used in this study is from a public database, which has a long time, and there are many clinical data missing.…”