| OBJECTIVES:The purpose is to combine preoperative nutritional status and systemic inflammation to predict the prognosis of patients with esophageal carcinoma after radical surgery,and to evaluate the model construction,so as to obtain a model with clinical practicability and applicability.METHODS:This manuscript adopts a retrospective research method,in which 256 patients with stage Ⅰ~Ⅲ esophageal carcinoma in the Department of thoracic surgery of Subei people’s Hospital from January 1,2016 to February 31,2012 were included.The software of R 4.1.2 was utilized to acquire the survival best cut-off value and the confusion matrix of survival by using "pROC" package and "survminer" package for the continuity variables.SPSS statistics.26 statistical software was used to analyze the correlation of parameters,where the independent sample t-test or one-way ANOVA will be chosen if the continuous variables conform to the normal distribution,or the nonparametric rank sum test shall be selected.Pearson chi square test was used for categorical variables.The survival data were analyzed by R language 4.1.2 software,and the survival curve was obtained by Kaplan-Meier method.Univariate analysis of overall survival(OS)was performed by log-rank test.When p<0.2,the factors can be included in multivariate regression analysis.Through Cox regression analysis,the condition is forward stepwise regression,and the hazard risk ratio(HR),95%CI value and P value of each variable are generated.In the model comparison step,we tested the performance of the prediction model through the Area Under Curve(AUC)of Receiver Operating Characteristic Curve(ROC)and Decision Curve Analysis(DCA).Naturally,nomogram and Clinical Impact Curve(CIC)were plotted to assess the clinical practicability,applicability and net benefit of the model with the best diagnostic value.RESULT:A total of 256 patients with stage Ⅰ-Ⅲ esophageal squamous cell carcinoma were subsequently included in this study.Among them,they consisted of 63 female patients(75.4%)and 193 male patients(24.6%),with a median age of 65 years.None of the general characteristics were statistically significant.The results of perioperative examinations showed that the average count of Neutrophil was 194.0±67.7(109/L),the mean of Neutrophil-to-Lymphocyte Ratio(NLR)was 2.7±1.9,and the average value of Platelet-to-Lymphocyte Ratio(PLR)was 125.7±50.5.The average value of Skeletal Muscle Index(SMI)was 43.8±6.4(cm2/m2),and there was significant difference among the counting findings(P<0.05),whereas the other examination results were not significant.In addition,the average intraoperative bleeding(P=0.02),postoperative pathological TNM(P=0.002),vascular invasion(P=0.01),tumor size(P=0.01),postoperative radiotherapy and chemotherapy(P=0.02),and postoperative metastasis and recurrence(P<0.01)were also statistically significant.The nutritional index of SMI(P<0.01)and Cachexia Index(CIN)(P=0.01),the indicators of inflammation PLR(P=0.05)and NLR(P=0.001),the comprehensive indices Albumin Globulin Score(AGS)(P=0.03)and Albumin Globulin Score and Skeletal Muscle Index(CAS)were all positive(P<0.01).The correlation analysis of nutritional and inflammatory indices showed that the SMI value of patients was correlated with gender(P=0.03),positively correlated with erythrocyte count(r=0.3)and body weight(r=0.4),and negatively correlated with age(r=-0.2).In addition to SMI,albumin and globulin,CAS was also closely related to patient age(P=0.03,F=3.4),gender(P=0.04,X2=6.27,V=0.2),alcohol consumption(P=0.04,X2=6.5,V=0.2),lymphocyte count(P<0.01,F=6.8),NLR(P=0.03,F=3.7),PLR(P=0.04,F=5.6)and CIN(P=0.05,F=3.1).ROC curve shows that the AUC value of CAS is markedly better than NLR(AUC=0.53),PLR(AUC=0.55)and CIN(AUC=0.57),demonstrating that the evaluation effect of CAS is superior.The median follow-up of patients was 63 months(IQR:60~68 months),of which 160 cases survived at the end of follow-up,representing an overall survival rate of 62.5%.The one-year,three-year and five-year survival rates in G1 group were 100%,96.6%and 88.2%respectively;The one-year,three-year and five-year survivability rates in G2 group were 100%,91.5%and 69.8%separately;The 1-year,3-year and 5-year overall survival rates in G3 group were 96.8%,74.2%and 38.7%accordingly.Cox regression analysis revealed that postoperative recurrence and metastasis(HR=2.15,CI=1.16-4.00,P=0.02)and CAS(P<0.01)were independent contributors affecting the prognosis of patients.Subsequently,survival regression analyses were performed separately for SMI,CLN,PLR and AGS,which had a remarkable contribution to the postoperative survival of the patients in this study,and then a prognostic model was constructed.The results showed that both AGS,CIN,SMI and PLR,as well as CAS composite score had an impact on the prognosis of patients with esophageal cancer.However,only the CAS composite score,SMI and AGS could be independent factors affecting the survival of patients after esophagectomy.Two models were constructed for correlational comparison and validation.The results showed that the 5-year survival predicted by both the CAS composite score,SMI and AGS were similar to the actual survival,and the results were on the diagonal,indicating that the CAS composite evaluation model and the models constructed by SMI and AGS were consistent,and the CAS composite evaluation model was calibrated with good accuracy.The C-index of the prediction model fluctuates from 0.7 to 0.8.The model calibration shows that the accuracy of the model calibration is well established,and the survival rate forecasted by the model is not far from the actual situation.The clinical decision curve and clinical impact curve indicated a relatively high net revenue for the prediction model.CONCLUSION:1.The CAS score is more representative than NLR,PLR,CIN and other indicators to appraise the nutritional status and inflammation of patients with esophageal carcinoma.2.The CAS score can be combined with nutritional status and inflammation of patients with esophageal carcinoma to comprehensively evaluate the prognosis of patients undergoing radical esophagectomy.3.The prediction model including the CAS score has excellent accuracy,a high net revenue and favorable prediction function. |