Construction And Validation Of A Preoperative Risk Nomogram Prediction Model For Gastric Gastrointestinal Stromal Tumors

Objective:Gastrointestinal stromal tumors(GIST)carry a potential risk of malignancy and the treatment of GIST varies for different risk levels.Endoscopic resection is becoming a major option for the treatment of GIST,but it is generally accepted that surgical resection should be an option for GIST of high malignant potential.However,there is no systematic preoperative assessment protocol to predict the malignant potential of GIST.The aim of this study was to analyze the risk factors for the malignant potential of gastric GIST and to develop a reliable and clinically applicable preoperative nomogram prediction model to predict the malignant potential of gastric GIST.Methods:Patients with pathological diagnosis of gastric GIST at the Hospital from January 2015 to December 2021 were screened according to inclusion and exclusion criteria,and general information,clinicopathological features,EUS features,CT features and peripheral blood cell counts were retrospectively collected.The enrolled patients were randomly divided into a development group and validation group in the ratio of 7:3,with the former used for building the nomogram model and the latter for validation.The risk of gastric GIST was reclassified into low malignant potential group(very low and low risk)and high malignant potential group(moderate and high risk)according to the risk stratification criteria proposed by the National Institutes of Health(NIH)in 2008.Independent risk factors for high malignant potential gastric GIST were identified using univariate and multivariate logistic analysis.Based on these independent risk factors,a nomogram model predicting the malignant potential of gastric GIST was developed and the model will be validated in the validation group.The area under the curve(AUC)of the receiver operating characteristic(ROC)was used to evaluate the discriminatory ability of the prediction model,and the calibration curve and decision curve analysis(DCA)were used to evaluate the accuracy and clinical value of the model.Results:1.A total of 494 gastric GIST patients were included in this study,divided into a development group(n=345)and a validation group(n=149),with 16.5%(57/345)of high malignant potential gastric GIST in the development group and 16.1%(24/149)of high malignant potential gastric GIST in the validation group.2.In the development group,univariate logistic regression analysis identified gastric GIST patients with high malignant potential associated with tumor size(OR=3.55;95% CI: 2.63-4.8;P<0.001),tumor surface ulceration(OR=13.84;95% CI:7.16-26.76;P<0.001),EUS high-risk features(OR=3.78;95% CI: 2.1-6.81;P<0.001),CT growth pattern(exophytic growth OR=3.93;95% CI: 1.76-8.75;P=0.001;mixed growth OR=6.98;95% CI: 3.53-13.83;P<0.001),CT enhancement(moderate OR=2.19;95% CI: 1.06-4.53;P=0.035;obvious OR=2.02;95% CI: 0.97-4.19;P=0.06),platelet to lymphocyte ratio(PLR)(OR=2;95% CI: 1.11-3.59;P=0.021)and monocyte to lymphocyte ratio(MLR)(OR=4.73;95% CI: 2.58-8.7;P<0.001).Multivariate logistic regression analysis revealed tumor size(OR=2.65;95% CI:1.91-3.67;P<0.001),tumor surface ulceration(OR=4.56;95% CI: 1.86-11.17;P=0.001),CT growth pattern(exophytic growth OR=4.04;95% CI: 1.2-13.63;P=0.025;mixed growth OR=4.28;95% CI: 1.62-11.32;P=0.003)and Monocyte to lymphocyte ratio(MLR)(OR=2.51;95% CI: 1.01-6.19;P=0.047)were independent risk factors for high malignant potential gastric GIST.Based on the risk factors identified in the multivariate logistic regression analysis,a nomogram model was built to predict high malignant potential gastric GIST.The AUC of the model was0.932(95% CI: 0.890-0.974)and 0.922(95% CI: 0.868-0.977)in the development and validation groups respectively,the Hosmer-Lemeshow test for the development and validation groups were P=0.593 and P=0.856,respectively.The best cut-off value for the development group was 0.184,and the sensitivity and specificity at this value were 0.895 and 0.875,respectively;the cut-off value for the validation group was0.153,and the sensitivity and specificity at this value were 0.833 and 0.856,respectively.The calibration curves indicated good agreement between predicted and actual observed outcomes,while the DCA indicated that the nomogram model had clinical application.Conclusions:Tumor size,tumor surface ulceration,CT growth pattern and MLR are independent risk factors for high malignant potential gastric GIST,and nomogram model constructed based on these factors have a high ability to predict high malignant potential gastric GIST.It will help clinicians to identify individuals with high malignant potential and develop precise individual treatments for patients.