Application of enhanced CT-based nomogram in identifying Epstein-Barr virus-associated lymphoepithelioma-like gastric carcinoma from gastric adenocarcinoma
Objective To develop and evaluate a clinical and enhanced CT based nomogram for the differentiation of lymphoepithelioma-like gastric carcinoma associated with Epstein-Barr virus(EBV)from gastric adenocarcinoma.Methods The clinical and imaging data of 25 patients with pathohistology-confirmed lymphoepithelioma-like gastric carcinoma were retrospectively analyzed,and 50 patients with gastric adenocarcinoma were included as the control group.Univariate analysis and multivariate Logistic regression were used to screen the predictors for identifying the two types of tumors.A personalized nomogram was constructed,and the distinction degree,calibration degree,and clinical utility value of the nomogram were evaluated by drawing the ROC curve,calibration curve,and decision curve analysis(DCA).Internal validation was performed using the Bootstrap method.Results The location of the tumor(middle vs.upper,OR=1.866,95%CI:0.457-7.626;lower vs.upper,OR=14.454,95%CI:2.619-79.778),CT enhancement peak(OR=6.965,95%CI:1.796-27.016)and the lymph node status defined by CT(OR=4.242,95%CI:1.270-14.164)were independent differential factors in patients with lymphoepithelioma-like carcinoma.The AUC value of the nomogram was 0.815(95%CI:0.718-0.912).The nomogram showed a good calibration and a clinical benefit.For bootstrapping validation(internal validation)with 200 bootstrap-resampling,the nomogram showed good discriminatory accuracy and the AUC value of internal model validation was 0.823(95%CI:0.817-0.830).Conclusion The nomogram established in this study has great potential in identifying EBV related lymphoepithelioma-like gastric carcinoma from gastric adenocarcinoma and warrants clinical decision making.
NETL
NSTL
万方数据
