The New Immune Prognostic Model For Non-Small Cell Lung Cancer

Objective:Non-small cell lung cancer（NSCLC）is one of the deadliest malignant tumors in humans.The eighth edition of TNM staging is a reliable guide for NSCLC prognosis prediction and treatment decisions,but the postoperative survival of patients with the same TNM staging may also be significantly different.In view of the increasingly important role of the immune microenvironment in tumor diagnosis and treatment,by constructing an Immunoscore,it can provide a powerful supplement for TNM staging.However,the current Immunoscores that have been proposed in NSCLC have not achieved the desired effect in clinical applications.This study aims to build a more refined new Immunoscore with higher efficiency through statistical modeling by detecting the expression levels of various immune components in the NSCLC tumor microenvironment.The biological behavior of lung adenocarcinoma（LUAD）and lung squamous cell carcinoma（LUSC）is vastly different;thus,this study aimed to establish the Immunoscore（IS）in LUAD and LUSC separately.Finally,this study proposed a new staging system in NSCLC by combining TNM stages and Immunoscore.This study also combined more clinicopathological factors with Immunoscore to construct a clinical prognostic tool for evaluating survival probability of patients,which provide important reference information for the prognosis stratification and clinical decision-making of patients with NSCLC.Methods:This study selected 304 patients with resectable stage Ⅰ-Ⅲ NSCLC as the study subjects,including 204 adenocarcinoma and 100 squamous cell carcinoma.Using multi-color immunofluorescence staining technology,this study stained 17immunological markers on the pathological sections of these patients,combined them into 112 immune indicators and quantitatively analyzed their expression.This study analyzed the prognostic significance of the 112 immune indicators in the training cohort.Kaplan-Meier survival analysis and LASSO-COX regression analysis were used to screen the indicators.The grade scores and regression coefficients of the finally screened indicators were used to establish the Immunoscore(IS_NSCLC)in adenocarcinoma and squamous cell carcinoma separately.The IS_NSCLC were called IS_LUAD and IS_LUSCrespectively.This study tested the prognostic prediction efficacy of the Immunoscores through Kaplan-Meier survival analysis and the area under ROC curve（AUC）.Next,this study validates and evaluated the Immunoscores in the validation cohorts for squamous cell carcinoma and adenocarcinoma.Finally,this study combined the Immunoscore with tumor T and N stages to propose a new TN-I staging system in resectable NSCLC.This study also compared the efficacy of the Immunoscore with other prognostic models.On the other hand,this study combined more clinicopathological indicators with the Immunoscore to construct a nomograph model in lung adenocarcinoma and lung squamous cell carcinoma to evaluate the recurrence and death rates in NSCLC patients,which could improve the clinical application value of IS_NSCLC.Results:1.The constructed IS_LUAD contained three immune features:CD4⁺CD73⁺_{core of tumor（CT）},PD-L1_CT⁺,and IDO⁺_{invasive margin（IM）}.IS_LUSC contained two immune features:CD8⁺CD39^-CD73_CT^- and CD8⁺TIM-3_IM.⁺According to the expression level of the above features,this study constructed two formulas to obtain the immune score of each patient,where IS_LUAD=–(1.372×CD4⁺CD73_CT⁺ status)+(1.499×PD-L1_CT⁺status)+(1.014×IDO_IM⁺ status)and IS_LUSC=–(1.641×CD8⁺CD39^-CD73_CT^- status)+(1.319×CD8⁺TIM-3_IM⁺ status).2.In the training cohort,significant prognostic differences were found upon comparing low-IS_NSCLC patients with high-IS_NSCLCpatients.For LUAD,the 5-year disease-free survival（DFS）rates were 54.7%vs.8.1%and the 5-year overall survival（OS）rates were 82.4%vs.36%（all P<0.0001）.For LUSC,the 5-year DFS rates were74.0%vs.14.7%and the 5-year OS rates were 78.2%vs.17.6%（all P<0.0001）.In the validation cohort,the good prognostic predictive efficacy of IS_NSCLChas also been validated.Hierarchical analysis showed that IS_NSCLC is still an important prognostic marker of DFS or OS in each clinical pathological group.Multivariate analyses indicated that IS_NSCLCwas an independent indicator for prognosis.3.ROC analysis showed that the prognostic accuracy of IS_NSCLC in LUAD and LUSC was higher than that of TNM stages,suggesting that IS_NSCLC has a higher prognostic value than TNM.In addition,in LUAD and LUSC,ROC analysis showed that IS_NSCLC has a better prognostic value than the CD3/CD8 Immunoscore suggested by the Society for Immunotherapy of Cancer.4.Finally,this study combined IS_NSCLC with clinicopathological factors to establish a TN-I staging system and two nomogram models for clinical use.The LUAD cohort and LUSC cohort obtained the same combination of T,N,and I,forming a new NSCLC staging system.ROC curve analysis shows that the TN-I stages had better prediction accuracy than TNM stages.The nomogram is a method used to predict the3-year recurrence rate and 5-year mortality of patients with stage Ⅰ-Ⅲ LUAD/LUSC.Compared with IS_NSCLC or TNM staging,the AUC value of the nomogram model is larger.The calibration curve shows that the predicted probability and the observed probability are very consistent,and the decision curve analysis indicates the superior performance of the nomogram model.Conclusion:The newly established IS_LUAD and IS_LUSC were completely different,and were excellent indicators for the prognostic prediction of LUAD and LUSC respectively.The TN-I staging and nomogram model constructed on the basis of IS_NSCLC can effectively improve the prognostic accuracy and facilitate the clinical application.With the help of the TN-I staging system,the prognosis prediction,clinical decision-making and individualized treatment of lung cancer will be more accurate and reliable.