Predictive Value Of Inflammatory Markers And CT Imaging In Spread Through Air Space Of Operable Lung Adenocarcinoma

Objective: To study the changes of systemic inflammatory markers in patients with surgically treatable lung adenocarcinoma with spread through air space(STAS),including neutrophil-to-lymphocyte ratio(NLR),Platelet-to-lymphocyte Ratio(PLR),Lymphocyte-tomonocyte Ratio(LMR),Advanced Lung Cancer Inflammation Index(ALI),Prognostic Nutritional Index(PNI)and System Inflammation Index(SII),combined with the corresponding preoperative CT imaging and other data to establish a clinical prediction model,to explore the effect of various indicators on the occurrence of STAS in patients with operable lung adenocarcinoma The predictive value of the model was evaluated,and the predictive ability of the model was evaluated to provide a reference for subsequent clinical applications.Methods: A retrospective research analysis was used to collect 210 patients with lung adenocarcinoma confirmed by pathology report after surgical treatment in the Cardiothoracic Surgery Department of the Third Xiangya Hospital of Central South University from January 2018 to January 2021.They were divided into STAS negative group(167 cases)and STAS positive group(43 cases).The general clinical data,preoperative CT imaging data,pathological characteristics,and inflammatory indicators such as NLR,PLR,LMR,ALI,PNI and SII in the STAS negative group and STAS positive group were statistically analyzed and processed.Statistical methods such as independent samples t test and χ~2 test were used to analyze and compare between groups;the optimal cut-off value of each inflammatory index was determined according to the Receiver Operating Characteristic Curve(ROC),and the Area Under Curve(AUC)of each index was calculated.Multivariate Logistic regression was used to select CT imaging features and inflammatory indicators of STAS-related risk factors,calculate their Odds Ratio(OR),and draw Nomogram combined with R language Figures and calibration curves to build and evaluate clinical prediction models.Results: 1.In the statistical analysis of clinical data and pathological characteristics,there were significant differences in the degree of differentiation,lymph node metastasis,tumor stage,pathological subtype,presence or absence of micropapillary structure,capsule invasion and vascular invasion between the STAS-positive and STAS-negative groups.(P <0.05),the pathological features of patients with SATS-positive lung adenocarcinoma were more advanced,poorly differentiated,with micropapillary structure,with lymph node metastasis,and vascular and pleural invasion.The pathological subtypes were mainly acinar and papillary types;2.In the STAS positive group and the STAS negative group,the preoperative CT imaging features such as tumor location,density,shape,vascular bundle sign,edge feature and pleural traction sign were different and statistically significant(P <0.05),Most of the lesions in the STAS positive group of lung adenocarcinoma patients were located in the middle and lower lobes of the left lung.The proportion of positive symptoms(30.23%)was higher than that of the STAS negative group;3.In the STAS positive group,the NLR,PLR and SII levels were higher than those in the STAS negative group(P <0.05);the LMR,ALI and PNI levels were lower than those in the STAS negative group(P<0.05).The data analysis shows that among the various inflammatory indicators,the area under the NLR curve is the largest(AUC=0.83,95%CI:0.76～0.89),the best cut-off value is 2.07,its sensitivity is 86.05%,and its specificity is 67.07%.The second is ALI(AUC=0.81,95%CI:0.74～0.88),the best cut-off value is 35.37,its sensitivity is62.79%,and its specificity is 88.62%;4.Multivariate Logistic regression analysis showed that in the preoperative CT imaging features,tumor lesion density(OR=7.80,95%CI: 2.56-23.79),shape(OR=3.32,95%CI: 1.29-8.55),vascular bundles Signs(OR=4.80,95%CI: 1.44-15.94),border features(OR=3.27,95%CI: 1.28-8.36),and NLR ＞2.07(OR=5.69,95%CI: 1.70-9.05)in inflammatory markers),ALI ＞35.37(OR=0.22,95%CI: 0.07-0.67)were independent influencing factors of STAS in patients with operable lung adenocarcinoma;5.Based on the relevant combined indicators "tumor lesion density,shape,vascular bundle sign,edge features,and NLR,ALI",a STAS-related prediction model was constructed.The Nomogram nomogram constructed by the R language can improve the prediction model,and the calibration curve established therefrom has a good degree of coincidence with the ideal curve.Conclusion: Based on the density,shape,vascular bundle sign,edge features,and inflammatory indicators NLR and ALI of tumor lesions in preoperative CT imaging,the STAS prediction model for lung adenocarcinoma patients constructed in this study has good diagnostic prediction efficiency and high reliability,which can be applied to the preoperative evaluation of STAS in patients with clinical lung adenocarcinoma.