| Research background and purposeLung cancer is the most lethal tumor in the world,of which non-small cell lung cancer accounts for the majority?about 85%?.Non-small cell lung cancer is mainly composed of lung squamous cell carcinoma and lung adenocarcinoma.Advances in targeted therapy in recent years have significantly benefited patients with lung adenocarcinoma.EGFR mutations provide targets for targeted therapy.EGFR mutations are mainly present in lung adenocarcinoma,and only a small part is present in lung squamous cell carcinoma.Once non-small cell lung cancer patients are diagnosed with lung adenocarcinoma or lung squamous cell carcinoma,some patients or their families may want to know the survival time of the patients.At present,there are no individualized prediction models for lung squamous cell carcinoma and lung adenocarcinoma under different EGFR mutation states.Therefore,a prediction model that can help patients and their families to understand the disease is urgently needed.In recent years,some other solid tumors such as esophageal cancer,nasopharyngeal cancer,and breast cancer[1-3]can accurately predict the survival time of patients by constructing nomogram prediction models,and achieved this goal.The nomogram is built on the basis of a formula,substituting the variable to be studied into the equation,drawing multiple scales,and using a ruler to read the values in different situations.The nomogram is constructed by integrating multiple independent variables and the endpoints of the study?such as age,TNM stage,survival status,survival time?and other indicators,and based on statistical analysis,investigate the probability of a certain event in an organism with certain characteristics?such as the probability of developing diabetes,the probability of a three year survival time for a tumor patient,etc.?,this has played a role in promoting the individualized development of organisms[4-8].This study intends to construct a nomogram model of the overall survival of lung squamous cell carcinoma and lung adenocarcinoma patients under different EGFR mutation states,and verify its accuracy,to help patients and their families have a preliminary understanding of the condition,and make preparation for the big data analysis.Method1.Acquisition of clinical dataApply R3.5.1 software,install the readr package in R and the TCGAbiolinks package in bioconductor,run the program parameters under the TCGA database?GDC:https://portal.gdc.cancer.gov/?,load the above package,set the environmental parameters and program parameters,and then use R language to program,and select1026 NSCLC patients from 1991 to 2013,including 504 cases of lung squamous cell carcinoma and 522 cases of lung adenocarcinoma.2.Screening clinical dataScreen based on lung squamous cell carcinoma data of the same patient with complete age,gender,T staging,N staging,M staging,radiotherapy,drug treatment,survival time,survival status and other indicators,and a total of 420 cases of eligible data are selected.Screen based on lung adenocarcinoma data of the same patient with complete age,gender,EGFR mutation status,T staging,N staging,M staging,radiotherapy,drug treatment,survival time,survival status and other indicators,and 195cases of eligible data are selected.3.Analysis of COX resultsThis study uses COX regression method under SPSS software to analyze the survival of lung squamous cell carcinoma and lung adenocarcinoma respectively.The univariate analysis results could suggest potential risk factors related to patient survival?P<0.05?,and the multivariate analysis results could suggest independent risk factors related to patient survival?P<0.05?.4.Construction of nomogramsThe R program packages under R software such as rms,survival,foreign,etc.are used to establish the nomogram prediction models for the lung squamous cell carcinoma and lung adenocarcinoma.The total score is calculated by adding the scores of various predictive factors,and the corresponding 3-year and 5-year survival probabilities are obtained.5.Validation of nomogramsUsing R software packages under R software such as rms,survival,foreign,rmda,etc.,to calculate the C-index,draw a calibration curve?picture?and a decision curve for the lung squamous cell carcinoma and lung adenocarcinoma to verify the accuracy of the nomogram.ResultsIn lung squamous cell carcinoma,COX univariate analysis in lung squamous cell carcinoma shows that age(1.80*10-2)is a potential risk factor for overall survival?OS?in patients with lung squamous cell carcinoma.COX multivariate analysis shows that age(1.40*10-2)is an independent risk factor that affects the overall survival?OS?of patients with lung squamous cell carcinoma.In lung adenocarcinoma,COX univariate analysis shows that N stage(2.00*10-6)is a potential risk factor for overall survival?OS?in patients with lung adenocarcinoma.COX multivariate analysis shows that N stage(2.00*10-5)and EGFR mutation status(4.60*10-2)are independent risk factors that affect overall survival?OS?of patients with lung adenocarcinoma.The C-index of lung squamous cell carcinoma and lung adenocarcinoma are 0.635and 0.781,respectively.Lung squamous cell carcinoma and lung adenocarcinoma show that the calibration curve and the predicted curve are highly coincident in the 3 year and5 year survival calibration graphs.The decision curves for lung squamous cell carcinoma and lung adenocarcinoma have ideal net benefit rates over a large threshold range?lung squamous cell carcinoma:18%-58%,lung adenocarcinoma:4%-76%?.ConclusionsThe following conclusions can be drawn from this study:Lung squamous cell carcinoma1.Potential risk factors for overall survival?OS?in patients with lung squamous cell carcinoma:age.2.Independent risk factor for overall survival?OS?in patients with lung squamous cell carcinoma:age.3.The nomogram we constructed based on factors such as age,sex,T stage,N stage,M stage,radiotherapy,and drug treatment of lung squamous cell carcinoma patients can accurately predict the overall survival time of lung squamous cell carcinoma patients.Lung adenocarcinoma1.Potential risk factors for overall survival?OS?in patients with lung adenocarcinoma:N stage.2.Independent risk factors for overall survival?OS?in patients with lung adenocarcinoma:N stage and EGFR mutation status.3.The nomogram we constructed based on age,gender,T stage,N stage,M stage,EGFR mutation status,radiotherapy,and drug treatment of lung adenocarcinoma patients can accurately predict the overall survival of lung adenocarcinoma patients. |
PDF Full Text Request