Establishing A Risk Stratification Prediction Model For Gastrointestinal Stromal Tumors Based On Clinical And Spiral CT Arterial Enhancement Fraction And Ki67 Of Pathological Data

Objective:Based on the clinical and pathological data and spiral CT arterial enhancement fraction,we discussion the independent factors influencing the preoperative risk stratification of gastrointestinal mesenchymal tumor and constructed a prediction model and plotted nomogram for preoperative prediction of risk stratification of GIST patients.Methods:Two hundred and twenty-five patients with postoperative pathologically confirmed GIST admitted to Shanxi Cancer Hospital from January 2013 to May 2021 were included,randomly divided into training set(n=156)and validation set(n=69)in a ratio of 7:3,and patients were divided into non-high-risk and high-risk groups according to the risk classification criteria.The relevant evaluation indexes included,age,gender,tumor location,maximum tumor diameter,AEF value and Ki67,logistic regression analysis was used to screen the independent influencing factors of GIST risk grading,and the model was constructed and plotted in columns,ROC curve was used to evaluate the prediction discrimination of the model,calibration curve to evaluate the prediction accuracy of the model,and clinical decision curve Decision Curve Analysis was used to evaluate the clinical usefulness of the model.Results:The results of single-factor and multi-factor logistic regression analysis showed that the maximum tumor diameter,tumor location,AEF value and Ki67 were independent influencing factors for GIST risk classification.The AUC of the training set was 0.913(95% CI: 0.739-0.942)and the AUC of the validation set was 0.904(95% CI:0.731-0.938),which showed that the model had a good prediction discrimination.The calibration curve showed that the prediction curve of the model and the actual curve basically matched each other,indicating that its prediction accuracy was high.Finally,in terms of the usefulness of the model,the clinical decision curve of the model showed a net benefit of using the model,which has good clinical applicability.Conclusion:The maximum diameter,location,AEF value and Ki67 of GIST tumor are independent factors influencing the risk classification of GIST,and the GIST risk prediction model based on these indicators can help to predict the risk classification of GIST before surgery and provide a more reasonable basis for the selection of individualized treatment plan.