Comparison of a novel predictive model with AJCC staging for predicting cancer-specific survival and overall survival in senile (≥60 years) lung adenocarcinomas

Abstract

Abstract Objective: We hope to develop a simple, rapid, and comprehensive predictive model that can evaluate the prognosis of elderly patients with lung adenocarcinoma(LUAD). Methods: Basic and clinical data on 14,117 patients ≥60 years of age diagnosed with lung adenocarcinoma between 2010 and 2019 were retrospectively analyzed using the Surveillance, Epidemiology and End Results (SEER) database.Independent factors affecting patients' prognosis were identified by univariate and multivariate COX regression analyses, and Nomograms of overall survival (OS)and cancer-specific survival(CSS) at 1, 3, and 5 years were constructed based on the results of COX regression analyses.Using the Concordance-index (C-index), calibration curve, Receiver operating characteristic curve (ROC), and Decision curve analysis (DCA) to evaluate the performance of the Nomograms.We also validated our established model with a validation set of patients and finally compared it with the AJCC staging model. Results: We included a total of 14,117 patients, which were divided into a training set and a validation set. We used the chi-square test to compare the baseline data between the two groups, which was not statistically significant (P＞0.05); we analyzed the data from the training set using Cox univariate and multivariate regression, and found that gender, ethnicity, marital status, stage, treatment, and distant metastasis were significant independent prognostic factors for OS and CSS (P<0.05);The ROC curves were used to validate the training and validation set data after the construction was completed, and the AUC for 1, 3, and 5 years all reached above 0.75, in addition to the C-index;The consistency of the calibration curves for OS and CSS is well behaved and close to the 45°reference line;The models for OS and CSS were also analyzed using DCA, showing that the net clinical benefit of the models built in this study was higher in both the training and validation sets;Finally, we used the models built for OS and CSS to compare with the AJCC staging model, and we found that both our models outperformed the AJCC staging model in terms of predictive performance. Conclusion: This Nomogram More Accurately Predicts Prognosis in Elderly Lung Adenocarcinoma Patients.