Analysis Of Risk Factors For Pathological Escalation Of Gastric Low-grade Intraepithelial Neoplasia After Endoscopic Resection And Establishment Of Prediction Model

Objective:Methods:The clinical data of 149 patients with gastric low-grade intraepithelial neoplasia who received endoscopic resection in the affiliated Hospital of Guilin Medical College from January 2016 to December 2021 were reviewed and collected.All patients were Collected with clinical data and endoscopic features,such as gender,age,infection with Helicobacter pylori(HP),location,the lesion of the size(largest diameter),type,mucosal congestion,appearance(ulceration,erosion,nodule),intestinal metaplasia and gastric atrophy.According to endoscopic resection pathological findings,patients were divided into escalation group and no escalation group.The independent risk factors of pathological escalation of gastric low-grade intraepithelial neoplasia after endoscopic resection were screened by univariate and multivariate Logistic regression analysis.Then,the selected independent risk factors were used to establish the line chart prediction model.Receiver operating characteristic curve(ROC),Area under curve(AUC)and calibration diagram were drawn to verify the prediction efficiency and accuracy of the line chart model;and the Boot-strap method was used for internal verification of Nomogram.Finally,the Decision Curve Analysis(DCA)was drawn to verify the clinical validity.Results: Among the 149 patients,60 had the final pathological escalated to high-grade intraepithelial neoplasia,early or advanced gastric cancer,the total upgrade rate was 40.3%.85 remain the same,4 had pathological degradation.According to multivariate logistic regression analysis,the long diameter of lesion ≥ 2cm,intestinal metaplasia,mucosal congestion,erosion/ulcer were an independent risk factor for pathological escalation of gastric low-grade intraepithelial neoplasia after endoscopic resection(P<0.05).The above independent risk factors were used to establish an individual line chart prediction model.After verification,it was found that the AUC value was 0.890 [95%CI:0.834-0.945,(P=0.000)];Brier score was 0.122;R2 was 0.573,which showed that the model had good differentiation and calibration.In the Bootstrap internal verification,the mean absolute error was 0.033,which indicated that the model performed well in internal verification.In the DCA curve,the predictive model performed well in terms of clinical efficacy.Conclusions: Long diameter of lesion ≥ 2cm,intestinal metaplasia,mucosal congestion and lesion surface erosion/ulcer were independent risk factors for pathological escalation after endoscopic resection of gastric low-grade intraepithelial neoplasia lesions.In this study,the nomogram of pathological escalation after endoscopic resection of gastric low-grade intraepithelial neoplasia lesions based on independent risk factors had a good predictive value and may played a guiding role in the clinical treatment of gastric low-grade intraepithelial neoplasia patients.