Prognostic Analysis And Prognostic Model Construction Of Patients With Stage ⅢA-N2 Non-small Cell Lung Cancer Based On SEER Database

Objective:To explore the prognosis factors affecting the cancer-specific survival rates of patients with stage IIIA-N2 non-small cell lung cancer and construct a prognostic Nomogram model for clinical treatment decisions.Methods:The information of patients diagnosed with stage IIIA-N2 non-small cell lung cancer diagnosed between 2013 and 2019 were downloaded from the National Cancer Institute Surveillance,Epidemiology,and End Results(SEER)database,and the basic information of the patients(race,sex,age,lateral nature of the primary site of the tumor,tumor size,histological classification,degree of differentiation),treatment methods(chemotherapy,radiotherapy,surgery,regional lymph node dissection,postoperative radiotherapy)and survival information(survival months,survival status)were collected.We use Rstudio software to take advantage of set.seed function and sample function to randomly divide all patients into training cohort(N=9321)and verification cohort(N=3996)according to the 7:3 ratio.In this study,"cancer-specific death" was used as the endpoint event,"death from other causes" as the competition event,and "survival months" as the survival time.The "rms","mstate","cmprsk","survival","pec","riskRegression",and "timeROC" packages were used for statistical analysis using Rstudio software.Univariate cancer-specific survival analysis was performed on the training cohort by competitive risk Gray test,survival rate differences were compared to find risk factors.Multivariate cancer-specific survival analysis was performed by competitive risk Fine-Gray test to screen independent risk factors.These independent risk factors were used to construct a cancer-specific survival prognosis Nomogram model for stage ⅢA-N2 non-small cell lung cancer.Area under the receiver operating characteristic curve(ROC curve)and the concordance index(C-index)were used to evaluate the discrimination of the model,and the calibration curve was used to evaluate the consistency of the model.Validation cohort patient information was substitute into the model,and AUC、C-index and calibration curves were used to validate internally.Results:13317 patients with stage ⅢA-N2 non-small cell lung cancer were enrolled.All patients were randomly divided into 70%(N=9321)in the training cohort and 30%(N=3996)in the validation cohort in a 7:3 ratio.The results of the Gray test of univariate competitive risk showed that gender,age,tumor size,pathology classification,degree of differentiation,chemotherapy,radiotherapy,surgery,regional lymph node dissection,and postoperative radiotherapy significantly affected the cancer-specific survival rate of patients with stage ⅢA-N2 non-small cell lung cancer(P<0.05).The Fine-Gray test showed that sex,age,tumor size,pathology classification,degree of differentiation,chemotherapy,radiotherapy,surgery,and regional lymph node dissection were independent prognostic factors for cancer-specific survival in patients with stage ⅢA-N2 non-small cell lung cancer(P<0.05).Based on these independent factors,the prognostic Nomogram model was constructed to predict cancer-specific survival at 1,3 and 5 years in patients with stage ⅢA-N2non-small cell lung cancer.In the training cohort and validation cohort,the 1-year,3-year,and 5-year AUCs were 0.711,0.665,0.631,0.711,0.659,and 0.633,and the 1-year,3-year,and 5-year C-indexes were 0.697,0.651,0.636,0.701,0.648,and 0.636,respectively.The calibration curves showed that the Nomogram model predicted cancer-specific survival rate and the actual cancer-specific survival rate were in good agreement.Conclusion:We constructed a cancer-specific survival prognosis Nomogram model for patients with stage ⅢA-N2 non-small cell lung cancer,which has good accuracy and reliability,and can provide a reference for clinical treatment decisions.