Model Construction Of CT Image For Evaluation Of Postoperative Nutritional Indicators Of Esophageal Squamous Cell Carcinoma

Background: One of the most aggressive tumor types in the world is esophageal squamous cell carcinoma(ESCC),and nutritional symptoms are present in at least 60% of esophageal cancer patients,reducing their quality of life following surgery.However,at present,there are few nutritional risk screening tools at home and abroad that can be used to evaluate the prognosis of postoperative patients with esophageal cancer.Therefore,this study intends to construct a prognostic model of postoperative nutritional indicators of esophageal squamous cell carcinoma based on CT images to evaluate the prognosis.Methods: This study included 67 patients who underwent esophagectomy in the Department of Thoracic Surgery of Qingdao Municipal Hospital from October 2016 to December 2020.The patient information and related test data were collected through the clinical electronic medical record system of Qingdao Municipal Hospital,and the patients were followed up by telephone,and then the queue was divided into two parts: training set and verification set.This study evaluated the prognostic effects of traditional nutritional tools such as prognostic nutrition index(PNI),controlled nutrition status index(CONUT)and geriatric nutrition risk index(GNRI)on patients with esophageal squamous cell carcinoma.Construction of PPR indicators based on CT images: the paraspinal muscle area of the 12 th thoracic vertebra(T12)on CT images of the patients in the study was measured by INFINITT software before and 3-6 months after operation.The muscle index of T12 is defined as paraspinal muscle area(cm2)divided by height(m)squared.The postoperative /preoperative ratio of T12 muscle index(PPR)was calculated as follows: postoperative T12 muscle index divided by preoperative T12 muscle index.Firstly,the indexes with significant differences were screened out by univariate COX regression analysis(p<0.05).Then the significant differences were incorporated into multivariate COX regression analysis to select the independent prognostic factors affecting the survival of postoperative patients with esophageal squamous cell carcinoma.A new TNM-PPR nutrition assessment tool was constructed by using independent prognostic factors.Furthermore,an ESCC prognostic model based on the new scoring tool was proposed,and its differentiation,calibration and clinical applicability were analyzed.Results: At present,postoperative patients with esophageal squamous cell carcinoma have poor ability to use traditional nutrition assessment tools to predict prognosis.TNM stage(p<0.001)and PPR(p<0.001)were significantly correlated with postoperative outcome in esophageal cancer patients,according to univariate COX regression.TNM stage(p=0.0427)and PPR(p=0.0078)were identified as independent predictive variables for esophageal squamous cell carcinoma by multivariate COX regression analysis.Therefore,this study chose TNM stage and PPR as two indicators and put forward a new TNM-PPR scoring tool to simplify the model according to the number of indicators in the danger zone.Through the Kaplan-Meier survival curve analysis of the simplified TNM-PPR model,we could see that this tool can effectively predict the prognosis of postoperative patients with esophageal cancer(Log-rank p<0.001),which was better than the traditional nutrition assessment tool.The C index(C-index)of the nomogram,which is based on the evaluation tool,was 0.835(95% CI: 0.752–0.919),showing that the TNM-PPR model has a good capacity for prediction.The calibration curves verified internally and externally showed that it had a good fitting effect with the actual situation when judging the overall survival rate of 1 year and 2 years.ROC curve(Receiver operating characteristic curve),DCA(Decision Curve Analysis)curve and AIC(Akaike Information Criterion)value showed that this model had better differentiation and clinical applicability than traditional nutrition assessment tools.Conclusion: The performance of the traditional nutrition assessment instrument(PNI,CONUT,GNRI)in predicting the prognosis of patients after ESCC was unsatisfactory.Compared with the other three commonly used nutrition assessment tools,the prognosis model based on TNM-PPR proposed in this study had a better ability to predict the prognosis of postoperative patients with esophageal cancer.It could more accurately judge the survival status of postoperative patients,so as to adjust the nutritional treatment plan of patients in time.