CT Differentiation Of Esophageal Neuroendocrine Carcinoma From Squamous Cell Carcinoma

Objectives:To evaluate clinical manifestations the differential CT features of esophageal neuroendocrine carcinomas(E-NECs)from squamous cell carcinomas(SCCs).Materials and methods:From January 2011 to November 2018,22 cases of E-NECs confirmed by menstruation pathology and immunohistochemistry in our hospital were collected,including 16 males and 6 females,with an average age of about 65.5 years old and an age range of 57-77 years old.As the control group,194 cases of E-SCCs confirmed by menstruation pathology and immunohistochemistry between January 2017 and October 2017 were collected.The 1:1 matching sampling was conducted between E-NECs and E-SCCs according to the age,gender and tumor stage of the patients.Finally,22 cases of E-SCCs were included,including 19 males and 3 females.The average age was about 63.5 years s ranging from 57 to 77 years.The clinical datas and CT imaging features of the two tumors were analyzed retrospectively.All 44 patients with E-NECs and E-SCCs underwent CT scanning in our hospital.All CT images were independently analyzed by two experienced radiologists,and consensus was reached through discussion when the review opinions were not uniform.The CT features of esophageal lesions were recorded from the following aspects:tumor location,the degree of wall thickening,the length of the long axis of the esophagus involved in the lesion,the ratio of the thickness of lesion to the length of the esophagus involved,the extent of the circumference of the esophagus infiltrated,the relationship between the length of lesion and the long axis of the esophagus(vertical,parallel),the boundary,the depth of invasion,the growth pattern(wall thickening,intra-extraluminal nodules),the shape of the air interface of the esophageal inwall around the tumor,homogeneity of lesions,presence of esophageal mucosa,degree of enhancement,enhancement pattern,lymph node metastasis and distant organ metastasis.Univariate statistical analysis was used to evaluate the CT performance difference between E-NECs and E-SCCs(chi-square test or Fisher's precise experience for categorized variables;independent t-test for continuous variables in accordance with normal distribution,Mann-Whitney U-test for non-normal distribution).The more important CT variables(P=0.1)were analyzed by multivariate binary logistic regression.The receiver operating characteristic curve(ROC)analysis was used to determine the optimal threshold for the size of the metastatic lymph nodes.Kappa was used to test the consistency of two observers for categorical variables and ICC was used to test the consistency of two observers for continuous variables.Results:There were statistically significant differences between E_NECs and E-SCCs in clinical symptoms(with or without dysphagia),the shape of the air interface of the esophageal inwall around the tumor,enhancement pattern,lymph node metastasis,the size of metastatic lymph nodes,and the enhancement pattern of hepatic metastatic tumor(P<0.05).The optimal threshold for lymph node size in both tumors was 2.05cm.The differences were no statistically significant between E-NECs and E-SCCs in the tumor location,the degree of wall thickening,the length of the long axis of the esophagus involved in the lesion,the ratio of the thickness of lesion to the length of the esophagus involved,the extent of the circumference of the esophagus infiltrated,the relationship between the length of lesion and the long axis of the esophagus(vertical,parallel),boundary,infiltration depth,growing methods(wall thickening or intraventricular tubercle),homogeneity of lesions,presence of esophageal mucosa,and degree of enhancement(P>0.05).Conclusions:CT plays role in the differentiation E-NECs from E-SCCs,and provides a strong basis for the selection of clinical treatment methods and prognosis analysis.