Objective:To assess the value of ultrasound combined with peripheral blood inflammation indicators in predicting the WHO/ISUP classification of clear cell renal cell carcinoma(ccRCC).Methods:Ultrasound and clinical data were collected from 93 ccRCC patients confirmed by surgical pathology in our hospital,and these cases were divided into a high-grade group(n=25)and a low-grade group(n=68)according to WHO/ISUP grading.Univariate and multivariate logistic regression analysis were performed to screen for independent influencing factors predicting WHO/ISUP classification of ccRCC.A joint prediction model was then developed and presented in the form of nomogram.Its calibration,stability and net clinical benefit were evaluated with calibration curve,5-fold cross-validation and decision curve analysis(DCA)respectively.Results:According to univariate and multivariate logistic regression analysis,maximum diameter,abundant blood flow and platelet lymphocyte ratio(PLR)were identified as independent influencing factors for predicting the WHO/ISUP classification of ccRCC.(all P<0.05).The nomogram had an AUC of 0.925(0.868 to 0.982),with better predictive efficacy than each single variable(all P<0.05).The AUC of 5-fold cross-validation was 0.941,which was generally consistent with the nomogram and had clinical benefit.Conclusion:The prediction model constructed by ultrasound findings combined with peripheral blood inflammation indicators can effectively predict ccRCC WHO/ISUP classification,and the nomogram can visualize the prediction findings.
维普
万方数据