Application Of CT Imaging Features And Radiomics In Preoperative Staging And Differential Diagnosis Of Clear Cell Renal Cell Carcinoma

The study was divided into two partsPart one CT Imaging Features and Risk Factors of High-risk c T1 Clear Cell Renal Cell Carcinoma Objective:To investigate the CT features and risk factors of high-risk c T1 clear cell renal cell carcinoma(cc RCC).Methods:The clinical and imaging data of 203 patients with c T1 stage clear cell renal cell carcinoma was collected.The CT imaging features and risk factors of high-risk clear cell renal cell carcinoma upstaging to p T3 a were analyzed retrospectively.Results:180(88.7%)were in p T1 stage and 23(11.3%)were upstaging to p T3 a of the203 patients with clear cell renal cell carcinoma.Univariate analysis showed that perirenal fascia thickening,tumor size,tumor morphology,tumor boundary,peritumoral collateral veins and pathological grade were related risk factors for p T3 a upstaging(P<0.05),patients with p T3 a had higher pathological grade,larger diameter,more blurred boundary,more irregular shape,more prone to peritumoral collateral veins and perirenal fascial thickening.Multivariate logistic regression analysis showed that perirenal fascia thickening(odds ratio(OR)=7.38,95% confidence interval(CI):2.33-23.40,P=0.001),tumor boundary(OR=8.74,95%CI:2.40-31.76,P=0.001),and peritumoral collateral veins(OR=7.98,95%CI:2.21-28.75,P=0.002)were independent risk factors for p T3 a upstaging.Conclusion : Fuzzy tumor boundary,perirenal fascia thickening and presence of peritumoral collateral veins are preoperative risk factors for high-risk c T1 renal clear cell carcinoma.Part two A CT-based Radiomics Nomogram for Differentiation of ≤4cm Clear Cell Renal Cell Carcinoma and Angiomyolipoma Without Visible FatObjective: CT based radiomics features combined with CT image features to establish radiomics nomogram and verify its discriminant effectiveness between ≤4cm clear cell renal cell carcinoma and renal angiomyolipoma without visible fat(AML.wovf).Methods: 149 patients with cc RCC(n=102)and AML.wovf(n=47)were retrospectively analyzed.The radiomics features were extracted from the CT images of the corticomedullary,parenchymal and excretory phases,and the radiomics score(Radscore)was calculated.Evaluate CT imaging features and general clinical factors to establish image features model.A comprehensive model was established by combining Radscore and independent image features,and the nomogram was drawn and evaluated from the aspects of fit,discrimination performance and clinical practicability.Results: The comprehensive model(area under curve(AUC),0.979;95%CI,0.962-0.996)and the radiomics features model(AUC,0.972;95%CI,0.949-0.994)showed better discrimination efficiency compared with the image features model(AUC,0.933;95%CI,0.897-0.970)(P=0.003 、 0.045).Decision curve analysis showed that the comprehensive model outperformed the radiomics features model and the image features model in terms of clinical usefulness.Conclusion: CT based radiomics nomogram is a non-invasive preoperative prediction tool,which shows good prediction effect in distinguishing AML.wovf and cc RCC,which may help clinicians to formulate accurate treatment plans.